£&& JAERI-CONF 95-005 ^•x—A^^L-4-%- f \JAER1\ S Vol.1
JAERI-Conf—95-005 (v<,l)
JP9507237 Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research INNOVATIVE LASER TECHNOLOGIES
tfM#*.
BteV' ii ' Xi :•*!& list &sm JJ
V,\
Energy Society of Japan 本レポートは，日本原子力研究所が不定 J~l に公'F IJ している研究報告書です。 入手の間合わせは，日本原子力研究所技1j:汁;')'報部情報資料課(干A*eo|UI^h*«, H*IR^*Bf5fiRfrttffiff/*ffl5tS#*e^» (T319-319-11 H 3&WCW8WSBJ~だ城県別 IJ 可郎東K 海村)あて.お I担し他しください。なお，このほかに財団法人原子力弘前会資料センター (干(T319-1319-11 1 茨域県郎珂3E«I&WJSIIIBIK»i郡東海村日本原子力研究所内)で佼写による実質頒布をおこなってW B -#m-Mmrm) -efs^c «t a§£««#£ i- £ o -c おります。fciJii".
This This This report report is is issued issue d irregularly. irregularly. Inquiries Inquiries Inquiries about abou t availability availability of of the the reports reports should should bc be addressed addressed to to Information Information Division. Division,
Department Department Department of of Technical Technical Information ,， JapanJapa n AtomicAtomic Energy Energy Research Research Institute ,， Tokai -・ mura. mura. mura, Naka-gun. Naka-gun , Ibaraki-kcn Ibaraki-kcn 319319-H・11. , Japan. Japan.
。© Japan Japan Atomic Atomic Encrgy Energy Rcscarch Researc h Institutc. Institute, 1995 I995 Publishcd Publishcd Published by by JAERI.JAERl, March March I991995 5 編集兼発行 日本原子力研究所 印en 刷m いばらさ印刷附^ ix ^ # f-n m m ^ "-*" — -»• • - • •_«•_. . W L ••. •• •. • ••••'•;:,•
March 23 (Wed), 1994
OPENING ADDRESS Mr. S. Shimomura (President of the JAERI, Japan)
Conference Room (Zuiun-no-ma)
Honorary Invited Lecture 1 "Lasers in perspective" Prof. A. L. Schawlow (Stanford Univ. USA) The 6th International Symposium on Advanced Nuclear Energy Research "INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY"
-" **feS5 Honorary Invited Lecture 2 "Improvement of lasers towards its theoretical limit in the light of advanced technology" Prof. H. Takuma (Univ. of Electro-Communications, Japan)
RECEPTION Opening (Dr. S. Matsuura, Chairperson, Organizing Committee of the Symposium, JAERI, Japan)
(Dr. E. Tachikawa, Secretariat of the Symposium, JAERI, Japan) The 6th International Symposium onAdvanced Nuclear Energy Research "INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY"
POSTER SESSION March 24 (Thu), 1994
-- Oral Introduction '] Chairperson I Dr. M. Hoshi (JAERI, Japan) ; Dr. H. Shiraishi (JAERI, Japan)
Conference Room (Tachibana-no-ma)
Poster Session Room (Ume-no-ma)
March 25 (Fri), 1994
Honorary Invited Lecture 3 "Impact of laser and nuclear technologies" Prof.V.S. Letokhov (Institute of Spectroscopy, Russia) The 6th International Symposium on Advanced Nuclear Energy Research "INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY"
PANEL DISCUSSION "Future of Laser Science and Nuclear Technology"
Chairperson Prof. C. Yamanaka (Himeji Institute of Technology, Japan)
-, •ul*.~,«~j^-*v^^
Panelists Dr. T. Arisawa Dr. J. J.Ewing Prof. V. S. Letokhov Prof. H. Takuma (JAERI, Japan) (Aculight Co., USA) (Institute of Spectroscopy, (Univ.ofElectro-communi- Russia) cations, Japan)
Question from the Audience CLOSING REMARKS (Dr. T. Kondo, JAERI, Japan) \ f&ZI THF fi»h IHTFRNATIONAL SYMPOSIUM on ADVANCED NUCLEAR ENERGY RESEARCH
INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY
•' ' ,. (A- Sohawlow/*ia;r^9V7sf-l<>:¥*»»la),---:'; '.
•. •,-.: -M 's..L.:etaKhbv. Bi/7 •:»«¥St5tmK)a).:.';.-, V";,
;-«
43
<$$0*Mvi&'ffl/
The Announcement Poster of the Symposium 0pai Sy^ JAERI..CONF JAERI..CONF JAERI-CONF 95..005 95-005 VoVol.l. 1
ProceedingsProceeding s of of theth 6th e 6th International International SymposiumSymposiu m ono n AdvancedAdvance d NuclearNuclea r EnergyEnerg y ResearchResearc h INNOVATIVE INNOVATIVE LASER TECHNOLOGIESTECHNOLOGIES IN IN NUCLEAR ENERGY
MarchMarc h 23-25 ,， 1994199 4 at at at Mito Mito Mito PlazaPlaza Hotel Hotel Mito ,， Ibaraki ，, JapanJapa n
Organized and Sponsored by by Japan Atom;cAtomic Energy Research Institute Institute
in in Cooperat;onCooperation with with Science and Technology AgencyAgency The Atom;cAtomic Energy Society 0' of JapanJapan The Japan Soc;etySociety of Applied Phys;csPhysics The Chemical Society 0' of JapanJapan ThThee ιaserLaser Soc;etySociety 0' of JapanJapan
日本原子力研究所 Japan JapanJapan Atomic Atomic Energy Energy Research Researc h Institute Institute Organizing Committee Chairperson:S.Matsuura(JAERI) Members of Committee: S.Arai (Kyoto Inst. Tech.) M.Date (JAERI) T.Fujioka (Tokai Univ.) S.Funahashi (JAERI) T.Hiraoka (JAERI) Y.Izawa (Osaka Univ.) Y.Kaneko (JAERI) M.Kishida (JAERI) I.Kobayashi (JAERI) T.Kondo (JAERI) H.Kuroda (Univ. Tokyo) K.Muraoka (Kyushu Univ.) S.Namba (Prof.emeritus, Osaka Univ., S.Saito (JAERI) Nagasaki Inst. Applied Science) N.Sasao (PNC) S.Sato (JAERI) K.Shiba (JAERI) R.Shimaki (LASER-J) K.Shimoda (Prof, emeritus, Univ. Tokyo) A.Suzuki (Univ. Tokyo) E. Tachikawa (JAERI) Y.Takashima (Prof, emeritus, H.Takuma (Univ. Electro-Communications) Tokyo Inst.Tech.,IRI) S.Tamura (JAERI) l.Tanaka (Prof, emeritus, Tokyo Inst. Tech., T.Tomimasu (FELI) National Inst. Academic Degrees) K.Toyoda (RIKEN) T.Tsujino (JAERI) N.Yamabayashi (JAERI) C.Yamanaka (Prof, emeritus, Osaka Univ., M.Yoshida (JAERI) Himeji Inst. Tech., ILT)
Secretariat of the Symposium: Enzo Tachikawa Director Department of Chemistry and Fuel Research Japan Atomic Energy Research Institute Tokai-mura, Ibaraki-Ken, 319-11, Japan Phone 0292-82-5487,6468 Fax 0292-82-6097 JAERI-CONF 95-005
Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -Innovative Laser Technologies in Nuclear Energy- March 23-25,1994, Mito Plaza Hotel, Mito, Ibaraki, Japan
Editorial Working Group Tokai Research Establishment Japan Atomic Energy Research Institute Tokai-mura, Naka-gun, Ibaraki-ken
(Received February 13,1995)
Records of the 6th International Symposium on Adovanced Nuclear Energy Research, Which focused on the subject of "Innovative Laser Technologies in Nuclear Energy", are contained in this issue. It consists of three honorary invited lectures, 33 invited papers which were orally Presented, 82 contributed papers which were presented on posters, and a panel discussion. The oral presentations were organized in the following four sessions, (1) Basic laser science, (2) Laser development, (3) Laser material processing and (4) Selective Photoreaction.
KeywordsrLaser Technology, Nuclear Technology, Basic Laser Science Laser Spectroscopy, Laser Development, Laser Material Processing Selective Photoreaction
I JAERJAERI-CON卜CONF95 F 95-00・005 5
第$66 m^tiwmimWii'yxv^A田原子力先端研究国際シンポジウム ーレーザーの聞く原子力技術ー
1994^1994 年 3 月n 2233 日B-2~255 日B 茨城県水戸市水戸プラザホテル2S*ft!|t*Fr|j frp-f 7*?*7-fr 論文集
日本原子力研究所東海研究所
編集ワーキンググループ
(1995 (l995#-2J年2 月3 13 13 日受理)0 £11)
第 6 田原子力先端研究国際シンポジウムが「レーザーの聞く原子力技術J のテーマのもと
に、fc, 1994^1994 年33 月fl 23-223~255 日にわたって開催された。本シンポジウムの主な内容は、Hi:*3fco-CP§ffi$tLfc0 #v ^V^WHL&fi^li, (1)(l)V-Hfレーザー-
基礎科学、SBI¥l-^J2)v-Hf-ffl5S,(3)v-¥~JpiXW4)V-¥-*Rl&-e*ojt(2) レーザー開発、 (3) レーザー加工及び'(4)レーザー光反応であった。またパネル討論0Sfe^*^f!ri& では、「原子力におけるレーザー科学の未来」のテーマで論じられた。 本論文集は、*HX^I±, 3636ffl«7)ffl#ISS[j|&X編の招待講演論文、, 8282«<*>#**-$g3eit;fc編のポスター発表論文、パネル討論の討論記録を収, v^^ftS&wfltSfrfBas-JR
録したものである。giLfcfew-eft a0
東海研究所:〒*«jH9fste3f: T 319-11 319-11 茨城県那珂郡東海村白方字白根^mmwmm&ij^&m-t2・4
ii ii II Preface
For the last six years the Japan Atomic Energy Research Institute has annually held an International Symposium on Advanced Nuclear Energy Research, the aim of which is to stimulate research activity in the nuclear field by exchanging ideas with scientists in other related fields. A specific theme was chosen for each of them, and the sixth symposium in March 1994 was entitled as "Innovative Laser Technologies in Nuclear Energy",
Since the invention about thirty years ago, the coherent light source, the laser, has contributed a great deal to many branches of basic science, and its excellent characteristics have also leo to some technological applications. This symposium was motived primarily by an expectation that the recent progress in laser science will not only accelerate the existing laser-related researches, but also enable newer applications in the nuclear science and technology.
The topics covered a broad range of research fields from basic physics and chemistry to applied areas such as laser processing and isotope separation. In addition, instrumental development was put on the agenda to discuss free electron laser,' solid state laser, and other types of lasers. During the three days of the symposium, about three hundred and sixty researchers from eight countries participated in enthusiastic discussion through the oral, poster and panel sessions.
This symposium was rather a unique one, where researchers from many different fields had a chance to exchange opinions. Although the presented views for the future of applied laser technologies were not totally optimistic in a short term prospect, general recognition was that promising possibilities exist in many directions, and that such a meeting was really useful to find new ideas or to advance practical applications. Yet, whether the symposium was truly successful must naturally depend on our effort to make productive use of the result documented here in the proceedings.
I now again express our thanks to the members of the organizing committee, the invited speakers, and all who participated in the symposium. The cooperation by Science and Technology Agency, the Atomic Energy Society of Japan, the Japan Society of Applied Physics, the Chemical Society of Japan, and the Laser Society of Japan is also appreciated. I would in addition like to acknowledge gratitude to Prof. C. Yamanaka who kindly helped us in editing the manuscript for the panel discussion.
Enzo Tachikawa Symposium secretary, Deputy Director General of Tokai Research Establishment, JAERI
III CONTENTS
Opening Address 1 S. Shimomura (President, JAERI) Session I Honorary Invited Lectures 1-1 Lasers in perspective 3 A.L. Sshawlow (Stanford University, USA) 1-2 Impact of laser and nuclear technologies 11 V.S. Letokhov (Institute of Spectroscopy, Russia) 1-3 Improvement of lasers towards its theoretical limit in the light of advanced technology 25 H. Takuma (University of Electro-Communications , Japan) Session II Basic Laser Science II— 1 Laser cooling and trapping of atoms 35 S. Chu (Stanford University, USA) II-2 Nuclear properties studied by laser spectroscopy < 44 T.T. Inamura (RIKEN, Japan) II-3 Time-resolved laser-induced fluorescence in the nuclear fuel cycle 54 C. Moulin, P. Decambox, P. Mauchien and A. Petit* (CEA, France) 11-4 Photoacoustic and photothermal spectroscopies 61 T. Sawada*, T. Kitamori and M. Nakamura (University of Tokyo, Japan) II-5 Laser induced fluorescence applied to studies of particle behaviour in high-temperature plasmas 67 K. Muraoka*, K. Uchino, T. Kajiwara, M. Maeda and T. Okada (Kyushu University, Japan) II-6 Physics of laser implosion 78 K.A. Tanaka (Osaka University, Japan) II-7 Intramolecular dynamics and unimolecular reaction. A case study of state-selectively excited nitrogen dioxide 87 S. Tsuchiya (Japan Women's University, Japan) II-8 Isotopic analyses using resonance ionization spectroscopy 97 T.J. Whitaker*, H. F. Arlinghaus and G. I. Bekov (Atom Sciences Inc., USA)
V 11-9 Atomic collisions related to atomic laser isotope separation 107 T. Shibata (JAERI, Japan)
11-10 Ultrashort-pulsc lasers and their applications 117 G. A. Mourou (University of Michigan, USA) 11-11 Infrared multiphoton dissociation processes of some halogenated hydrocarbons 126 A. Yokoyama*, K. Yokoyama, G. Fujisawa and T. Takayanagi (JAERI, Japan) Session III Laser Development III-l User issues at the Stanford Picosecond Free Electron Laser Center 137 T.I. Smith (Stanford University, USA) III-2 JAERI IR free electron laser program 145 E.J. Minehara*, R. Nagai, M. Sawamura, M. Takao, M. Sugimoto, S. Sasaki, M. Ohkubo, Y. Suzuki and N. Shikazono (JAERI, Japan) N. Kikuzawa (Kyushu University, Japan) J. Sasabe and Y. Kawarasaki (Hamamatsu Photonics Co., Japan) IH-3 High-average-power, diode-pumped solid state lasers for energy and industrial applications 153 W.F. Krupkc (Lawrence Livermore National Laboratory, USA)
III-4 Technological gas-dynamic C02 laser 163 V. V. Apollonov *, A. M. Prokhorov and V. I. Kislov (General Physics Institute, Russia) V. P. Malyavin and V. A. Feofilaktov (SIC "Luga" Moscow, Russia) III-5 Lasers for fusion energy • • • • • 171 J.F. Holzrichter (Lawrence Livermore National Laboratory, USA) III-6 Dye laser chain for laser isotope separation 179 D. Doizi*, L.A. Lompre and P. Arondel (CE Saclay, France) HI-7 Advances in tunable solid state lasers 186 J.J. Ewing*, D. Lowenthal, R. Mead, M. Bowers, C. Miyake and S. Tidwell (Aculight Co., USA) III-8 High power diode-array-pumped frequency doubled Nd:YAG laser 196 B. Le Garrec* and Ph. Feru (CE Saclay, France) III-9 High-power CO laser and its potential applications 202 S. Sato*, K. Takahashi, K. Shimamoto and Y. Takashima (Institute of Research and Innovation, Japan) K. Matsuda, S. Kuribayashi, 0. Noda, S. Imatake and M. Kondo (Mitsubishi Heavy Industries, Ltd, Japan)
VI 111-10 Compact soft X-ray laser development 208 T. Hara*, K. Ando and Y. Aoyagi (RIKEN, Japan)
III-l 1 Wavefront compensation applied to AVLIS laser systems 213 T. Gonsiorowski* and A. Wirth (United Technologies/AOA, USA)
Session IV Laser Material Processing
IV-1 Surface processing by high power excimer laser 223 M. Stehle (SOPRA, France)
IV-2 Laser ablation for the metal-oxide dun film deposition: Mechanisms and applications 229 T. Kawai (Osaka University, Japan) 1V-3 Multi-wavelength copper vapour lasers for novel materials processing application • 236 M Knovvles, J. Evans*, R. Foster-Turner, and A. Kearsley (Oxford Lasers Ltd and Oxford University, UK) rV-4 Surface modification of polymer materials by excimer-laser irradiation 246 S. Kawanishi *, M. Nishii, Y. Shimizu and S. Sugimoto (JAERI, Japan)
Session V Selective Photo reaction V-l Overview of selective photo-reaction 255 T. Arisawa (JAERI, Japan) V-2 Photochemical reactions of actinide ions 265 H. Tomiyasu (Tokyo Institute of Technology, Japan) V-3 Laser stimulated extraction of Pd from solution with uranyl 275 M. Vlasov and S. Mironov (Russian National Center I.V. Kurchatov, Russia) B. Krynetsky*, A. Kukhtenko, A. Prokhorov and A. Zhidkov (General Physics Institute, Russia)
V-5 LIS: The view from Urenco 280 K.R. Schneider (Urenco Ltd, UK)
V-6 New Japanese AVLIS R&D program 290 Y. Sato (LASER-J, Japan) V-7 Research initiatives toward the new horizon of molecular laser isotope • 298 separation at RIKEN K. Takeuchi*, Y. Kuga, J. Onoe and H. Tashiro (RIKEN, Japan)
VII V-8 Current status of MLIS program in PNC 304 O. Suto*, H. Yamaguchi, K. Tashiro, S. Kawakami and Y. Shimazaki (PNC, Japan) Session VI Panel Discussion Future of laser science and nuclear technology 313 Leader: C. Yamanaka (Himeji Institute of Technology, Japan) Panellist: H. Takuma (University of Electro-Communications , Japan) Panellist: T. Arisawa (JAERI, Japan) Panellist: V.S. Letokhov (Institute of Spectroscopy, Russia) Panellist: J.J. Ewing (Aculight Co., USA)
Closing Remarks 337 T. Kondo (Dept. Director General, JAERI)
Author Index Keyword Index
VIII Poster Session
Ila Basic Laser Science lla-Pl Beam cooling by using laser-undulator beat wave 339 Y. Kishimoto* and J. Koga (JAERI, Japan) T. Tajima and D.L. Fisher (University of Texas, Austin, USA) K. Mima (Osaka University, Japan) IIa-P2 Measurement of nuclear moments of radioactive nuclide, M3Pr (Ti/2=13.6d), by collinear laser-ion-beam spectroscopy 350 H. Iimura*, Y. Nakahara and S. Ichikawa (JAERI, Japan) M. Kubota (Toho University, Japan) T. Horiguchi (Hiroshima University, Japan)
Ha-P3 Development of a laser ion source for the TIARA-ISOL 358 M. Koizumi*, T. Sekine and A. Osa (JAERI, Japan) T. Horiguchi (Hiroshima University, Japan)
IIa-P4 Charge transfer cross section of gadolinium and neodymium 368 K. Ogura* and T. Shibata (JAERI, Japan) IIa-P5 Application of direct simulation Monte Carlo method for analysis of AVLIS evaporation process 375 A. Nishimura* (JAERI, Japan)
IIa-P6 Ion collection from laser-induced plasma by applying radio-frequency voltage 384 T. Shibata* and K. Ogura (JAERI, Japan)
IIa-P7 Photodissociation dynamics of dichloroethenes 393 K. Sato*, S. Tsunashima and H. Umemoto (Tokyo Institute of Technology, Japan) T. Takayanagi, K. Yokoyama and H. Shiraishi (JAERI, Japan)
IIa-P8 C02 laser photo-induced decomposition of ammoniated ammonium ions 402 Y. Ikezoe*, T. Soga, K. Suzuki, N. Moriyama and S. Ohno (JAERI, Japan)
IIa-P9 Laser-induced partial oxidation of cyclohexane in liquid phase 409 Y. Oshima*, X-W. Wu and S. Koda (University of Tokyo, Japan) Ila-PIO Laser-induced photochemical reaction of aqueous maleic acid solutions containing H2O2 • • • 418 Y. Shimizu*, S. Kawanishi andN. Suzuki (JAERI, Japan)
lib Laser Spectroscopy Ilb-Pl The spectroscopy of uranium within the "SILVA" program 425 R. Avril, A. Petit*, J. Radwan and E. Vors (CE Saclay, France)
IX llb-P2 Direct isotope ratio measurement of uranium metal by emission spectrometry on a laser-produced plasma 436 W. Pietsch, A. Petit* and A. Briand (CE Saclay, France)
Ilb-P3 High sensitivity detection of selenium by laser excited atomic fluorescence spectrometry using electrothermal atomization 444 U. Heitmann, A. Hese (Technische Universitat Berlin, Germany) G. Schoknecht (Institut fur Sozialmedizin und Epidemiologic, Germany) W. Gries* (Laser Analytical Systems GmbH, Germany)
Ilb-P4 The development of analytical system for tracer level elements using laser enhanced ionization method 454 H. Ueda*, T. Tamura and M. Yoshida (Toshiba Co., Japan) llb-P5 Resonance ionization spectroscopy of Pb 464 S. Hasegawa*, R. Ohmori, T. Masuda and A. Suzuki (University of Tokyo, Japan) llb-P6 Speciation of actinides in aqueous solution by time-resolved laser-induced fluorescence spectroscopy (TRLFS) 473 T. Kimura*, Y. Kato. G. Meinrath and Z. Yoshida (JAERI, Japan) G.R. Choppin (Florida State University, USA)
Ilb-P7 Development of optical fiber system for laser induced photoacoustic spectroscopy 487 T. Kihara*, S.I. Sinkov, S. Fujine and M. Maeda (JAERI, Japan) lie Spectrometric Measurement
Hc-P 1 New neutron detection methods using nuclear pumped laser 495 M. Nakazawa, M. Notani* and T. Iguchi (University of Tokyo, Japan) T. Kakuta, H. Yamagishi and M. Katagiri (JAERI, Japan)
IIc-P2 TV Thomson scattering system on JFT-2M 502 T. Yamauchi*, T. Shiina and Y. Ishige (JAERI, Japan) D. Dimock (Princeton University Plasma Physics Laboratory, USA)
IIc-P4 Fundamental study on laser manipulation of contamination particles with determinating shape, size and species 512 I. Shimizu* and T. Fujii (Ibaraki College of Technology, Japan)
IIc-P5 Explosives vapour identification in ion mobility spectrometry using a tunable laser ionisation source : a comparison with conventional 63Ni ionisation 521 A. Clark*, R.M. Deas, C. Kosmidis, K.W.D. Ledingham, A. Marshall and R.P. Singhal (University of Glasgow, UK)
X IIc-P6 UV laser induced photochemistry of nitrobenzene and nitrotoluene isomers 530 C. Kosmidis*, A. Clark, K.W.D. Ledingham, A. Marshall and R.P. Singhal (University of Glasgow , UK) IIc-P7 Laser light reflection measurement for in situ analysis of thin film growth process 540 N. Naito*, A. Takano, M. Matsuse, M. Kawasaki, M. Sumiya and H. Koinuma (Tokyo Institute of Technology, Japan) Ilia Free Electron Laser Illa-Pl Development of a compact FEL system and its technological applications for X-ray iteration 547 M. riijita*, K. Imasaki, J. Chen, H. Furukavva, S. Kuruma and C. Yamanaka (Institute for Laser Technology, Japan) M. Asakavva, N. Sakamoto, T. Yamamoto, N. Inoue, K.Mima and 5. Nakai (Institute of Laser Engineering, Japan) K. Nagami and Y. Tsunawaki (Osaka Sangyo University, Japan) T. Agari, T. Asakuma and N. Ohigashi (Kansai University, Japan) IIIa-P2 A waveguide FEL experiment using the coherent synchrotron radiation emitted from electron bunches 557 M. Asakawa*, N. Sakamoto, N. Inoue, T. Yamamoto, K.Mima and S. Nakai (Osaka University, Japan) J. Chen, M. Fujita, K. Imasaki and C. Yamanaka (Institute for Laser Technology, Japan) T. Agari, T. Asakuma and N. Ohigashi (Kansai University, Japan) Y. Tsunawaki (Osaka Sangyo University, Japan) IIIa-P3 FELI Iinac for IR- and UV-FEL facilities 565 T. Tomimasu*, Y. Morii, S. Abe, A. Koga, Y. Miyauchi, T. Keishi, E. Nishimura, K. Saeki, S. Sato, A. Kobayashi, A. Zakou, E. Tongu and A. Nagai (Free Electron Laser Research Institute Inc., Japan)
IIIa-P4 The RF system for FELI linac 575 Y. Morii*, S. Abe, T. Keishi and T. Tomimasu (Free Electron Laser Research Institute Inc., Japan) IIIa-P5 Optical cavity and transport system for FELI 582 K. Saeki*, E. Nishimura, A. Kobayashi, T. Keishi, S. Abe, S. Sato and T. Tomimasu (Free Electron Laser Research Institute Inc., Japan)
IIIa-P6 Vertical undulators for FIR- and IR-FELs at FELI 589 A. Kobayashi*, T. Keishi and T. Tomimasu (Free Electron Laser Research Institute Inc., Japan)
XI IIIa-P7 Development of high performance optics for free electron laser 597 F. Kitatani (PNC, Japan)
Illa-P8 Experimental study on infrared free-electron laser at University of Tokyo 604 R. Hajima*, T. Hara, H. Ohashi, M. Akiyama, S. Kondo, Y. Yoshida, T. Ueda, T. Kobayashi and M. Uesaka (University of Tokyo, Japan) E. Nishimura, K. Saeki, S. Abe, T. Keishi, A. Kobayashi, Y. Morii and T. Tomimasu (Free Electron Laser Research Institute Inc., Japan)
IIIa-P9 Development of a submillimeter free electron laser using a compact electro-static accelerator 610 Y. Kawamura*, S.H. Shu, T. Tanabe, D.J. Li and K. Toyoda (R1KEN, Japan)
Illa-P 10 High-power millimeter-wave free electron laser based on a recirculating electrostatic accelerator • • 617 B. Lee*, S. Kim, Y. Jeong, S. Cho and J. Lee (KAERI, Korea)
Illa-Pl 1 Permanent-magnet helical undulator for a millimeter-wave free electron laser 626 J. Lee*, Y. Jeong, B. Lee, S. Kim and S. Cho (KAERI, Korea)
Illb Gas / Solid State Lasers
IHb-Pl Copper vapor laser system development 637 C. Konagai*, N. Aoki, R. Ohtani, N. Kobayashi and H. Kimura (Toshiba Co., Japan)
IIlb-P2 Decrease in lower level density due to cooling of gas temperature by thermal dissociation of hydrogen in copper vapor laser 643 I. Watanabe, K. Hayashi, Y. Iseki, S. Suzuki, E. Noda and 0. Morimiya* (Toshiba Co., Japan) IIIb-P3 Low temperature operated copper vapor laser 653 A. Ohzu*, M. Kato, Y. Suzuki and T. Arisawa (JAER1, Japan)
IIIb-P4 Characteristics of the 100Hz Raman laser system developed at PNC for the MLIS process 661 Y. Miyamoto*, S. Kawakami and Y. Shimazaki (PNC, Japan) H. Tashiro (R1KEN, Japan)
IIIb-P5 Development of a tunable narrow-linewidth C02 laser for molecular laser isotope separation 668 A. Suda* and H. Tashiro (RIKEN, Japan) IIIb-P6 Development of high power lasers and their applications for nuclear engineering in IHI • • • 676 H. Kanazawa*, M. Uehara, M. Mori, Y. Taniu, S. Yamaguchi and H. Harashina (Ishikawajima -Harima Heavy Industries, Ltd., Japan)
XII IIIb-P7 Solid state pump lasers with high power and high repetition rate 686 M. Oba*, M. Kato andT. Arisawa (JAERI, Japan)
Hlb-P8 CW-25W solid-state green laser 695
T. Kojima* and K. Yasui (Mitsubishi Electric Co., Japan)
IIIb-P9 High average power Nd:YAG slab laser 702 B.Le Garrec* and P. Feru (CE Saclay, France) Illb-PIO Development of aNd:YAG slab laser 708 T. Kasai*, T. Iwasaki, T. Nagashima and S. Numata
(Fuji Electric Corporate Research and Development, Ltd., Japan)
IIIb-P12 Diode pumped lasers : beyond the kW barrier 715
L.R. Marshall (Light Solutions Co., USA)
IIIc Tunable Laser, Short-Pulse/ Short-Wavelength Lasers and Adaptive Optics
IIIc-Pl Development of AVL1S dye laser system 721 A. Sugiyama*, T. Nakayama, M. Kato and T. Arisawa (JAERI, Japan) IIIc-P2 Simulations of longitudinally pumped dye laser amplifier 730
K. Takehisa* and S. Takemori (Hitachi Ltd., Japan)
IIIc-P3 Characteristics of visible semiconductor laser as frequency-tunable laser 738
T. Fujii* and K. Nemoto (Central Research Institute of Electric Power Industry, Japan)
IIIc-P4 Continuously tunable diode lasers external cavity design options 748
T. Day, F. Luecke and M. Brownell* (New Focus Inc., USA)
IIIc-P5 Accurate frequency tuning mechanism by a wedge prism in the single mode tunable laser • • 767 D-K.Ko*, S.H. Kim and J. Lee (KAERI, Korea) IIIc-P6 Photophysical properties of some xanthylium salts performances under CVL pumping • • • • 777 D. Doizi* and L.A. Lompre (CE Saclay, France) M.C. Gazeau (LISA, France)
IIIc-P7 Characterization of a terawatt XeCl laser system 784 T. Hasama (Electrotechnical Laboratory, Japan) 3+ HIc-P8 Ultraviolet picosecond pulse amplification in new solid-state laser medium : Ce :LuLiF4 • • 789 N. Sarukura*, Z. Liu and Y. Segawa (RIKEN, Japan) K. Edamatsu, Y. Suzuki and T. Itoh (Tohoku University, Japan) V.V. Semashko, A.K. Naumov, S.L. Korableva, R.Y. Abdulsabirov and M.A. Dubinskii (Kazan State University, Russia)
XIII IIIc-P9 A compact femtosecond Ti:sapphire/KrF laser system 791 Z. Zhang* and T. Yagi (Institute of Research and Innovation, Japan) IIIc-PIO Characteristics of photoelectron triggered gas switch by KrF excimer laser 798 H. Kusama* and T. Yagi (Institute of Research and Innovation, Japan)
IIIc-P 11 Femtosecond laser triggering of electric discharges 804 T. Yagi*, Z. Zhang and H. Kusama (Institute of Research and Innovation, Japan) IIIc-P12 Demonstration of soft X-ray amplification by optical-field-induced ionization 810 K. Midorikawa*, Y. Nagata, S. Kubodera, M. Obara, H. Tashiro and K. Toyoda (R1KEN, Japan) IIIc-P13 Laser plasma X-ray for non-destructive inspection 817 T. Yagi* and H. Kusama (Institute of Research and Innovation, Japan)
IIIc-PH Spatio-temporal structure of an optical system including a photorefractive crystal 822 T. Nakajima*, Y. Ohishi, H. Ohashi and M. Akiyama (University of Tokyo, Japan) IIlc-P15 Laser beam-forming by deformable mirror for laser isotope separation 831 K. Nemoto*, T. Fujii and N. Goto (Central Research Institute of Electric Power Industry, Japan) IV Laser Material Processing IV-P1 UV laser ablation dynamics of liquids as revealed by nanosecond time-resolved spectroscopic and photographic measurements 839 Y. Tsuboi*, K. Hatanaka, H. Fukumura and H. Masuhara (Osaka University, Japan) IV-P2 Pulsed laser deposition of titanium nitride thin films with high electrical conductivity 849 M.B. Lee*, M. Kawasaki, M. Yoshimoto and H. Koinuma (Tokyo Institute of Technology, Japan) IV-P3 Improvement of corrosion resistance of carbon steel using chemical vapor deposition
from Cr(CO)6 and Mo(CO)6 with an ArF-excimer laser • 856 N. Okada, Y. Katsumura* and K. Ishigure (University of Tokyo, Japan) IV-P4 Mechanism study of laser implant-deposition of Si and its application to surface modification of stainless steel 304 863 K. Sugioka*, H. Tashiro and K. Toyoda (R1KEN, Japan) M. Jyumonji and H. Takai (Tokyo Denki University. Japan)
IV-P5 Diagnosis of laser ablated carbon particles measured by time-resolved X-ray absorption spectroscopy • 873 A. Miyashita* and 0. Yoda (JAERI, Japan) T. Ohyanagi and K. Murakami (University of Tsukuba, Japan)
XIV IV-P6 Dynamic behaviors of laser ablated Si particles 881 T. Ohyanagi* and K. Murakami (University of Tsukuba, Japan) A. Miyashita and 0. Yoda (JAERI, Japan) V Selective Photoreaction
V-Pl Laser photochemistry of lanthanide ions 891 N. Nakashima*, M. Kusaba and Y. Izawa (Osaka University, Japan) C. Yamanaka (Institute for Laser Technology, Japan) W. Kawamura (Mitsubishi Heavy Industries Ltd., Japan)
V-P2 Energy transfer and quenching processes of excited uranyl ion and lanthanide ions in solutions 898 T. Yamamura* and H. Tomiyasu (Tokyo Institute of Technology, Japan) V-P3 Neptunium separation process by using photochemical reduction in nuclear fuel reprocessing • • 906 G. Uchiyama*, T. Kihara, S. Hotoku, S. Fujine and M. Maeda (JAERI, Japan)
V-P4 Evaluations of photo-solution chemical behaviors of Pu and Np and application technology 916 Y. Wada and K. Morimoto* (PNC, Japan) H. Tomiyasu (Tokyo Institute of Technology, Japan) V-P5 Angular momentum in multi-step photo ionization 922 T. Yoshida*, H. Adachi, A. Kuwako, K. Nittoh and Y. Araki (Toshiba Co., Japan) T. Watanabe and I. Yoguchi (Tokyo Electric Power Company, Japan)
V-?6 Resonance ionization spectroscopy for AVLIS 930 M. Miyabe*, I. Wakaida and T. Arisawa (JAERI, Japan)
V-P7 Laser isotope separation using Ti: sapphire laser 938 K. Tamura*, M. Oba and T. Arisawa (JAERI, Japan)
V-P8 Detuning effect in multistep photo-ionization of atomic isotope of heavy element 948 I. Wakaida*K. Alcaoka, M. Miyabe, M. Ohba and T. Arisawa (JAERI, Japan) V-P9 Simulation of self-focusing of laser beam through medium with multi-step photo-ionization 955 K. Akaoka*, I. Wakaida and T. Arisawa (JAERI, Japan)
V-PIO Temperature profiles on the gadolinium surface during electron beam evaporation 963 H. Ohba* and T. Shibata (JAERI, Japan)
XV V-Pl 1 Laser isotope separation of gadolinium 970 H. Niki* (Fukui University, Japan) N. Aly, K. Koh, K. Nomaru, Y. kawa and S. Nakai (Osaka University, Japan) Y.-W. Chen and C. Yamanaka (Institute for Laser Technology, Japan) V-P12 Propagation of laser pulse in near resonant medium 980 K. Nomaru*, Y. Izawa and S. Nakai (Osaka University, Japan) Y.-W. Chen and C. Yamanaka (Institute for Laser Teclinology, Japan)
V-P 13 Laser isotope separation by infrared multiple-photon decomposition of ethers 987 T. Majima* and K. Sugita (RIKEN, Japan)
V-P 14 13C separation by IRMPD of halogenated difluoromethanes 997 P. Ma*, G. Chen, B. Wu, J. Liu, Y. jing and M. Chu (Qinghai Institute of Saline Lakes, P.R. China) S. Arai (Kyoto Institute of Technology, Japan) V-P 15 Selective excitation of branched vibrational ladder in uranium hexafluoride laser isotope separation 1006 M. Suzuki*, Y. Miyamoto, M. Hasegawa and Y. Shimazaki (PNC, Japan)
V-P16 Separation of boron isotopes by infrared laser 1019 K. Suzuki (JAERI, Japan)
Author Index ' • 1027 Keyword Index • 1031
XVI Opening Address Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY-
Opening Address for The 6th International Symposium on Advanced Nuclear Energy Research
Shozo Shimomura, President of JAERI
Good morning, ladies and gentlemen. I would first like to thank you for friendly participation in this symposium. I express, in particular, deep and sincere appreciation to Prof. Schawlow, Prof. Letokhov, and Prof. Takuma, who kindly accepted to give invited lectures, and also to other Speakers and chairpersons from both abroad and home.
Since the foundation in 1956, JAERI has been engaged in a variety of research work related to nuclear energy, such as reactor safety, fusion technology, development of a high-temperature engineering-test reactor, radiation technology, nuclear ship as well as various basic investigations to support these projects. In addition, we are constructing an 8-Gev synchrotron radiation facility, SPring 8, in cooperation with RIKEN, and further we have recently organized a research center in JAERI to promote studies in basic science.
The present symposium is the sixth of a series of symposia, which have been held to stimulate research and to explore direction of future activities. The themes have been on "Chemistry", "Accelerator", "Environment", "Materials", and "Neutron". This year we selected "Laser" as a theme.
As you know, maser and laser were invented by Prof.'s Townes, Basov, Prokhorov, and Schawlow about thirty years ago. Since then the novel light source has created new areas of spectroscopy, and has induced revolutionary advance in many branches of physics, chemistry, and other sciences. The reason is, of course, that the quality of light from laser is far superior than that from conventional sources in a number of essential points. I've also heard that extensive progress has been and is being made in many aspects of these advantageous characteristics.
In recent years, industrial applications are also expanding in the fields of optical communication, laser-aided measurement, materials processing, medical uses, and so forth. Needless to say, in the nuclear technology, application to isotope separation is considered promising.
The idea for optical separation of atomic isotopes seemed to exist already in
-1- early days of nuclear energy research. However, it was not until the development of laser that such method became technically feasible. JAERI has initiated research on Atomic Vapor Laser Isotope Separation in 1976 by recommendation of a committee of the Japan Atomic Energy Forum. We owe some of the participants today kind and helpful advice at that time. Since then we have been studying the AVLIS technology as a long-term project. We will continue the study in cooperation with Laser Atomic Separation Engineering Research Association of Japan.
Turning to another aspect of our activity, I would like to mention Free Electron Laser that is expected to be a powerful tunable laser of new type. We began research on FEL in 1988, and are currently constructing an apparatus for far infrared wavelength, operated with a superconducting RF linac. Successful tests on electron acceleration have been performed recently, and the lasing experiment is about to be started.
There are also other laser-related studies in JAERI. In the physical field, for example, laser is used in nuclear and plasma physics. Studies are also made on actinides chemistry, surface processing, synthesis, basic photochemistry, and so on.
I only referred to our research activities to introduce JAERI, but as you all know, extensive research is being carried out on laser science and technology in universities, public institutions, and in private companies. A broad range of laser-related subjects are to be discussed in this symposium. Following the honorary lectures this morning, a number of oral and poster presentations will be made. Further, a panel discussion entitled as " Future of Laser Science and Nuclear Technology" will be held to conclude this symposium.
I am afraid that the schedule is rather tight, and that only short time is allotted for each presentation. In addition, the number of participants are so large, to our happy surprise, that you may wish more space in the meeting room. We are sorry for the inconvenience you may feel, but I hope that fruitful discussion will nevertheless be made.
Before concluding my speech, I would like to acknowledge the support by Science and Technology Agency, the Atomic Energy Society of Japan, the Japan Society of Applied Physics, the Chemical Society of Japan, and the Laser Society of Japan, and also express appreciation to the cooperation by the members of the organizing committee. Thank you for your attention.
-2 — Session I Honorary Invited Lectures Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY -
Lasers in Perspective Arthur L. Schawlow Stanford University Stanford, CA 94305, U.S.A.
In the past 35 years, very many different kinds of lasers have been discovered, but relatively few of them have been seriously developed. Nevertheless, lasers with a very wide range of properties are now available. Output power levels range from microwatts to terrawatts. Pulse lengths range from a few femtoseconds (10~" seconds) to continuous waves. Frequency or wavelength stability can be better than one part in 10*2. The wavelengths obtainable range from the microwave region, through the infrared, visible and ultraviolet to soft x-rays.
According to a recent market survey (1) , the value of the lasers manufactured throughout the world is approximately one billion United States dollars per year. About one quarter of that is in the rapidly growing area of semiconductor diode lasers. For non-diode lasers, the three largest categories are materials processing, medical and research. The largest categories of lasers are. in dollar volume, carbon dioxide and lamp pumped solid state, followed by ion lasers. Diode pumped solid state lasers are growing rapidly. For semiconductor diode lasers, the largest applications are telecommunications, optical storage and material processing.
Research uses are a very small part of the diode laser market, in part because there are no widely used lasers powerful enough to be used like ion lasers for Raman spectroscopy and for pumping dye lasers.
Laser properties and applications
No one kind of laser shares all of the properties, and indeed each application requires only a restricted subset of them. There are still gaps in the coverage, for instance of wavelengths. Even moderate power of a few watts is obtainable at only a few wavelengths, not always the most desirable ones. Even where the need is apparent and we can see in principle how it might be fulfilled, it may take many years of scientific and technological discoveries and engineering to achieve. For
-3- instance, continuous-wave visible lasers of high average power have many obvious uses ranging from machining metals to large screen video displays. Visible light can be focused more sharply than the 10.6 micron infrared radiation from a carbon dioxide laser, and it is more strongly absorbed by many metals. While ionic gas lasers work beautifully up to 20 watts or so, their very low efficiency, with resulting heat disposal problems and damage to discharge tubes and windows, makes it very difficult to scale them to much higher power levels. The largest visible lasers are far from equaling carbon dioxide lasers for high, sustained power, but progress is now rapid. One way to produce visible (green) laser light is to start by using semiconductor laser diode arrays to pump a high- power neodymium-YAG or glass laser and then double the output frequency in a nonlinear optical crystal. This arrangement has recently been refined to produce a hand-held green laser pointer. For higher powers, many technical challenges remain, such as the need to find better ways to cool the diode lasers. For that purpose, microchannel water cooling seems promising, because in sufficiently small channels water does not show turbulence, and many channels can be used. All optical surfaces must be carefully coated to prevent reflection, with films that can sustain the high intensities to which they are subjected. Doubler crystals must have low absorption losses to minimize thermal distortion. Despite these difficulties, power levels in the kilowatt range seem to be attainable.
Spectroscopy
Until the advent of lasers, the narrowest spectral lines were produced by spontaneous emission or absorption in gases at low pressures. These were broadened by the Doppler effect from the thermal motion of the atoms. Thus the line width was of the order of v/c, (average atomic velocity divided by velocity of light), or typically about one part in 10^. The width could be reduced somewhat by cooling, but only as the square root of the absolute temperature, and for most substances that reduction was limited by condensation. Somewhat greater reduction could be achieved by using an atomic beam and observing it in a direction transverse to the motion of the atoms, but the number of atoms was greatly reduced. Laser cooling (2) now permits the thermal motions of gaseous atoms to be almost completely eliminated. For lasers, the line width of the source is ultimately limited by quantum fluctuations, or more often by less' fundamental factors
— 4 — such as gas discharge noise or the stability of the laser resonator. If a gas laser, for instance, has a resonator length of 1 meter, which is typical for laboratory argon lasers, a change in length of one nanometer would cause the output frequency to change by a part in 10^.
For really high wavelength stability, the laser can be locked to a resonance of a high Q resonator designed for maximum rigidity and minimum sensitivity to changes in temperature. Such an ultra-stable resonator needs to be protected from disturbances such as mechanical and acoustic vibrations. This can be done by suspending it from soft springs in a vacuum chamber. John Hall has shown that a laser's output frequency can be stabilized with such a resonator to about 1 Hertz (out of nearly 10^). However, additional uncertainties are produced by vibrations and air currents that cause fluctuations in the effective distance between the resonator and the laser. Ted Hansch and his associates in Garching (3) have been able to achieve an absolute laser frequency stability of 1000 Hertz in the visible, or nearly one part in 10^2.
For such high precision, wavelength measurements are not adequate. Wavelength-dependent phase shifts in the mirrors make uncertain the comparison of wavelengths for a fixed resonator. Measurements of the absolute frequency of the light wave are needed, and they can be done although with considerable difficulty. One can start with a precise methane-stabilized helium-neon laser at the infrared wavelength of 3.39 micrometers. Such lasers have been calibrated at national standards laboratories against the fundamental cesium atom microwave frequency standard. Further harmonic generation with the necessary microwave offsets can then produce light of a known frequency in the visible region. If this standard is close enough to the optical frequency to be measured, microwave mixers can determine the offset with great precision, and hence get the absolute frequency.
A newer method from Hansch's group in Garching, not yet fully developed, is to make use of frequency differences in the visible or near-visible spectral region.(4) In this technique, one starts with a laser at the frequency fj to be measured, and generates its third harmonic. A second laser is tuned to a frequency f2 so that its second harmonic coincides exactly with the third harmonic of fj. The exact coincidence can be assured
-5 — by phase locking these harmonics. Then 2f2 = 3fi, or ±2 ~ 3/2 fi, and the frequency difference between f2 and f^ is exactly 1/2 f^. Repeating the process n times, the frequency difference between fn and f^ can be reduced to a microwave frequency which can be measured against the cesium standard. Since this difference, fi " fn, is a known fraction of the original laser frequency, f j, the laser light's frequency can be accurately determined. This method avoids the need for several lasers in the infrared and far infrared regions. It does require a number of lasers, each phase-locked to the proper harmonic, but all of these lasers are in the same visible or near-visible spectral region. They can all be diode lasers if the diode lasers of the proper wavelengths can be obtained. With gas and optically pumped solid lasers, the available wavelengths were limited by the properties of the laser materials. With semiconductor diodes, nearly any wavelength could be made, but only a few are available commercially.
Eor nonlinear spectroscopy, such as saturation or two- photon spectroscopy, you may need more power than a stabilized laser can deliver. In order to obtain substantial power from a highly monochromatic laser, it may be useful to make a low-power ultra-stable laser and use its output to drive an amplifier or lock a more powerful laser to the same exact frequency.
Short pulse lasers
Techniques for generating ultra-short pulses of laser light are continuing to advance. (5) One very effective technique is Kerr mode locking, which uses the refractive index change at peak power to improve the matching of the laser to the pumping beam and the resonator. The shortest pulses require broad band amplification, and titanium-sapphire has been found particularly useful. Its bandwidth is great enough so that pulses as short as 3 femtoseconds might be attainable. That is approximately the length of one cycle at the 800 nanometer wavelength of the laser. To achieve the shortest pulses, care has to be taken to compensate for the dispersion of refractive index in the materials of the laser. Pulses as short as 6 femtoseconds have been reported, and a relatively simple titanium-sapphire laser with Kerr mode locking has generated pulses as short as 10 to 12 femtoseconds. When the light from femtosecond lasers is focused into a gas, very high harmonics are generated up to a hundred times the input frequency. If
- 6 — those harmonics are in phase, they may correspond to a pulse as short in the attosecond range (10"18 seconds), but as far as I know this has not yet been verified. Femtosecond pulses are so short that nothing macroscopic moves appreciably during their duration. Even a bullet moving at the speed of sound, 1000 meters per second, would travel only one nanometer in 10 femtoseconds. However, they are of the order of the duration of a vibrational cycle in molecules, and so have proved powerful tools for clarifying the molecular dynamics of chemical processes.
One potential application of lasers is in interferometers to detect gravitational waves. According to Einstein's general relativity theory, gravitational waves will be generated by such catastrophic events as the collapse of a star into a supernova or a black hole. These pulses would be about one millisecond in duration, and enormous amounts of energy would be radiated. Even so, as these events will occur only at astronomical distances from the earth, their effects on any detector are necessarily very small, so that extraordinarily high sensitivity will be needed to detect them. Such a wave would cause a fractional change in length, AL/L of about 10"18 if it originates in our galaxy, which will not happen very often. For gravitational collapses in a nearby galaxy, AL/L would be only about 10"22. Detecting the effects of such a gravitational wave would be extremely difficult, but it would provide a new kind of astronomical information even for events hidden from our view by intervening objects or dust.
One possible way to detect gravitational waves would be to use a very long interferometer of extremely high finesse. It would help to have two arms at right angles, as in the Michelson interferometer, since when the gravitational wave compresses one arm it would elongate the other. The longer the arms, the greater the sensitivity, and lengths of kilometers are being planned. The effective path length can be increased by making multiple traverses of the optical path between the mirrors, for which purpose mirrors of exceptionally low losses have been developed.
But even when all reducible sources of noise, such as from laser amplitude and frequency fluctuations and from mechanical vibrations, have been eliminated, there remains the noise from the discrete nature of photons. This can be minimized to some
- 7- extent by using rather powerful, as well as ultra-stable, lasers. However, to obtain the enormous sensitivity needed to detect gravitational waves. it may also be necessary to circumvent the quantum limit ordinarily imposed by fluctuations in the number of photons. This is a real, although difficult, possibility.
Non classical light
For many purposes, a light beam can be considered as a stream of photons arriving at the detector, which senses the intensity proportional to the number arriving per unit time. A good laser emits a coherent beam, in which the photons follow a Poisson distribution of arrival times, but one with minimum uncertainty. For a coherent beam, the fluctuation in the number of photons, n, arriving in a time interval is the square root of n. For many purpose this fluctuation is small enough so that it can be ignored. But for ultrasensitive detection, as for gravity waves, it can be reduced further to produce what is known as squeezed light, for which the photons follow each other at regular intervals. Squeezing can be achieved in a number of ways. Perhaps the simplest is to use a low-noise laser diode driven by a constant-current power supply. The electrons reach the diode junction at a constant rate, and for each one a photon is emitted.
The regularity of photon intervals in squeezed light, however, is very easily disturbed. For instance, at a semi- transparent beam splitter, the photons are randomly either transmitted or reflected, and either beam is noisier than the original. Photon detectors of less than 100% efficiency also introduce further noise.
It is possible to detect and measure light without absorbing any photons from it. This is done by passing the light beam through a transparent material whose refractive index changes with intensity, A probe beam of a different wavelength detects the change in refraction. Since the original light beam's amplitude is not changed, this measurement can be repeated many times for increased sensitivity.
It may also become possible to make a quantum optical repeater by coupling a semiconductor photodetector and a diode laser. (6) The quantum correlation between the photocurrent
-8- fluctuation of a semiconductor optical receiver and the output intensity fluctuation of a semiconductor light emitter driven by the photocurrent of the receiver has recently been demonstrated. The two fluctuations are correlated beyond the shot noise limit, which must be a quantum effect. This suggests the possibility that the light signal could be measured and exactly regenerated by such a "quantum optical repeater. However, thermal effects do add noise, and that must still be overcome.
Non locality
In recent years many non-classical properties of light and matter have been explored. Their results have ruled out many, perhaps all, local hidden variable theories. If two quantum objects, such as photons, are in entangled or mixed state, a measurement on one of them determines the probable results of a measurement on the other. For instance in the experiments by Aspect and his colleagues. (7) calcium atoms were excited from the ground ^-S0 level to a higher ^S0 level by a two-photon absorption transition. The excited atoms then returned to the ground state by emitting two photons. Since the initial and final singlet states each have zero angular momentum, the two photons emitted cannot carry away any angular momentum. They might both be linearly polarized in some undetermined direction, or they might be circularly polarized with opposite directions.
In their experiment, polarizers and photodetectors were placed in opposite directions from the emitting atoms, each about 6 meters away from the source. Coincidence in the arrival times showed when two photons from the same atomic cascade were observed. If the one polarizer was set for horizontally polarized light, and the other for vertical, we might expect classically to get some coincidences. That might happen if the emitted waves were linearly polarized at 45 degrees to the horizontal. But quantum mechanically, that does not happen. J.S. Bell had shown that averaged over all polarizer settings, quantum mechanics predicts a higher degree of correlation between the polarization of the photons than any local hidden variable theory. That is, the individual photons have to be thought of as having as polarization amplitudes the sum of the two possible polarizations. When the amplitudes are squared to get the probability, there is a cross product term in addition to the squares of the individual polarization probabilities.
-9- But the most remarkable result has come from an extension of this experiment, originally suggested by David Bohm. The detectors are far enough away that the time of flight for each photon is about twenty nanoseconds. During that time. a polarization modulator switches the setting of one polarizer in a pseudo-random way. Thus, although the polarizer settings are changed so rapidly that one detector cannot sense the setting of the other, the same degree of correlation is observed. Signals sent with the speed of light from one detector to the other would require a transmission time of 40 nanoseconds, and the time for polarization switching is only ten nanoseconds. These experiments appear to show that not only is the quantum state of the two photons described by a single wave function, but also that the information about the collapse of the wave function travels faster than light.
For an experimental physicist, these raise the question of whether that quantum information is transmitted with infinite speed or just with some very high velocity. Put another way, one may ask whether quantum mechanics which has been so thoroughly tested, is a local approximation and some propagation effects might be needed when it is applied over really long distances.
Conclusion
In conclusion, lasers are not only finding many uses in technology and medicine, but also they are proving to be stimulating tools to make possible many scientific investigations previously inconceivable. They are providing new insights into the nature of matter. both inanimate and biological, and into the nature of light itself.
References
(1) Stephen G. Anderson, Laser Focus World, January 1994, pages 62-76. (2) T.W. Hansch and A.L. Schawlow, Optics Communications 13. 68 (1975). (3) F. Schmidt-Kaler, D. Leibfried, M. Weitz, and T.W. Hansch, Phys. Rev. Lett. 70, 2261 (1993). (4) R. Wynands, T. Mukai. and T.W. Hansch, Optics Letters 17, 1749 (1992). (5) Henry C. Kapteyn and Margaret M. Murnane, Optics and Photonics News, 5. 20 (March, 1964) (6) Yoshihisa Yamamoto. Science, 263. 1394 (1994). (7) A. Aspect, Proceedings of International Conference on Atomic Physics. 8, 103, Plenum Press (1983)
— 10 — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research
-INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY-
IMPACT of LASER and NUCLEAR TECHNOLOGIES V.S.LETOKHOV
Institute of Spectroscopy Russian Academy of Sciences Triotzk, Moscow Region, 142092, RUSSIA
A brief review is given of the applications of laser spectroscopy techniques and laser separation processes at an atomic-molecular level in nuclear physics and technology.
Keywords: laser spectroscopy, laser separation, isotopes, nuclear isomers.
1. INTRODUCTION
A series of principally new technologies have been developed over the 20th century that have formed the basis for a revolution in modem science and engineering. While lagging in its evolution some
30-40 years behind the nuclear technology, the laser technology has already reached a level allowing it to exert a noticeable effect on other fields of science and technology, nuclear physics and technology in particular. It should be emphasized in this respect that the laser technology is now at its exponential development stage still far from saturation. And so its application trends observed to exist today will be enhanced many times over in the years to come.
Table 1 generally lists the main trends in the application of lasers in the nuclear technology, the types of lasers needed for particular applications being indicated. The lecture briefly considers only the first two trends based on the resonance laser- matter interaction, which enable one to implement laser detection (and diagnostics) and laser separation at an atomic-molecular level. The applications based on nonresonance laser- matter interaction are equally important but they fall outside the scope of my interests.
The laser methods for detection and separation at an atomic- molecular level are potentially capable of being successfully used at all the stages of the nuclear fuel cycle. First, the laser detection of the traces of radioactive and dangerous chemical species contaminating the environment can be employed at the nuclear power station itself, at the fuel- reprocessing and chemical conversion plants,
-11- and at the radioactive waste storage sites. Secondly, laser diagnostic techniques can be used to monitor the state of the critical reactor elements and to prospect for uranium deposits. Thirdly, laser separation
techniques can be employed to achieve a more efficient separation of uranium isotopes, to produce
monoisotopic materials for some construction reactor components, and probably to extract radioactive
isotopes for other applications in the chemical reprocessing of radioactive wastes.
Table 1. LASER SCIENCE and TECHNOLOGY for NUCLEAR TECHNOLOGY
Laser Spectroscopy _>. Laser Tunable LasersLaser Detection Fiber Optics and Diagnostics
Resonant Interaction of Laser Separation Pulsed High Average Laser Light with Atoms (Isotopes,....) Power Tunable Lasers and Molecules
Laser High Power Pulsed Processing of and CW Lasers Materials Nonresonant Interaction of Laser Light with Substance Laser High Energy and Thermonuclear """" Peak Power Pulsed Fusion Lasers
2. LASER SPECTROSCOPY
Unique properties of laser radiation (monochromaticity, temporal and spatial coherence, high
power, controllable pulse duration) has made it possible to revolutionize all the characteristics of optical
spectroscopy. It has now become possible to attain the ultimate of these characteristics (Table 2). All
these characteristics are potentially important for nuclear physics and technology applications. I will
restrict myself to the discussion of the possibilities of implementing the laser spectroscopy for study of
atoms with short- lived nuclei in ground and excited metastable state (nuclear isomers), detection of
atoms with very rare long- lived isotopes and traces of radioactive atoms.
- 12- Table 2. LASER SPECTROSCOPY
CHRACTER1ST1CS UMJTATJONS 1. Spectral Resolution - Homogeneous Width, - Interaction Time with Laser Wave -103- 1 Hz 2. Temporal Resolution - Period of Light Oscillations - 10-14-10'15s 3. Sensitivity - Atomic- Molecular - Single Atom Structure Molecule 4. Selectivity - Overlapping of Wings - Very Rare of Spectral Lines Isotopes 10"10- 10-20 5. Spatial Resolution - Light Wavelength, -0.1- 1.0 mem - de-Broglie Wavelength - 1-10 A 6. Remoteness - Length of Propagation - Cross- section of Scattering -0.1- 100 km
Laser spectroscopic techniques are greatly diversified (see monograph [1]), but a unique position among them is held by the laser resonance ionization spectroscopy (RIS) [2, 3] which allows one to achieve ultimate characteristics, specifically very high sensitivities and selectivities in detecting and separating of atoms and molecules. Figure 1 presents a schematic diagram of a laser resonance ionization spectrometer used to study the proportion of short- lived nuclei ( mean- square charge radius variation A
This method has been used at the Konstantinov Institute of Nuclear Physics and the Institute of
Spectroscopy to investigate the isotopic shifts and hyperfine structures of the isotopes of Nd (A= 132,
134- 142), Sm (A= 138- 145, 147, 149, 150, 152, 154), Eu (A= 138-115), Ho (A= 152- 165), and Tm
(A= 156- 172) and other rare- earth elements [5]. These measurements have yielded the mean square charge radii of the nuclei of these isotopes, their electromagnetic moments (with the exception of the even-even isotopes and the Eu isotopes), improved values of the spins of certain nuclei, and the isotopic behavior of the charge radius A of rare earth elements forN< 82 and 88< N < 94, i.e., on either side of the magic number N= 82. Fig. 2 shows the values of A
Sm, and the previously investigated [6] isotopes of Ba, Cs, Xe (using a different system). As can be
- 13 — seen, A
From the standpoint of nuclear physics, the success of laser optical spectroscopy in the study,
detection, etc. of nuclei is based on two properties: (1) the large cross section for the resonance
excitation of optical transitions in the electron shell (a0pt= X^/ 2n= 10**0 Cm2= 10^ barn) and (2) the
high intensity of even relatively modest laser beams (1 W/cm^ corresponds to an intensity 1= 10^
photon/ cm^s). This ensures the high rate of resonance excitation Wext= o"0ptI= 10" s"* of atoms in a
beam for given nuclear charge Z and given neutron number N.
Fig. 1 General view on laser resonance ionization spectrometer of Leningrad Institute of Nuclear Physics and Institute of Spectroscopy for study of short- lived isotopes generated by lGeV proton beam from accelerator.
Moreover, the scope of laser methods is actually much more extensive: it is possible to achieve
not only isotopically, but also isomerically selective excitation of atoms, which means that isotopic and
isomeric nuclei can be detected and separated by laser radiation [7], Selective laser photoionization of
nuclear isomers was first observed in on- line experiments on the hyperfine structure and isotopic shift
of the atomic lines of the radioactive isotopes of europium [5a].
— 14 — A
Number of neutrons Number of neuirons a b
Fig. 2 Variations of mean- square charge radius A
The nuclear isomers of samarium- 141g, 141m, and thulium- 164g, 164m have been separated
for the first [8] time by selective laser photoionization of atoms in an on- line experiment, using a proton
accelerator and a mass separator of radioactive isotopes. The principle employed was similar to that
illustrated in Fig. 3. Photoions with an excited nucleus of 141mSm or 164Tm, produced in isomer-
selective three- step photoionization, were extracted by the electric field from the region of interaction
between the laser radiation and the atomic beam, and were deposited on the cathode of a secondary-
electron channel multiplier. Fig. 3 shows the photoionization spectrum of a mixture of the Sm
isomers (laser wavelength tuned to the first step) together with an interpretation of the results. It is clear
that the photoionization spectrum contains well- resolved ions belonging to 141mSm. This means that,
when the laser frequency corresponding to the first step is tuned in the range 1-4 GHz, a beam of
photoions containing isomeric nuclei is produced in the direction perpendicular to the atomic beam. This
system can be used to investigate isomers with half- lives of the order of the time necessary to liberate
them from the target (down to 1 s).
- 15 — Ground slate Excited state "Sm,. of the nucleus ol the nucleus 1-1/2 /- tt/2
3/2 6161,SA J226s,6c»f •J/2
6751,5% ir'esspfy -2 -! 0 1 2 J * 5 6 GHZ 6004,2% m+g || | g III 111 I I I I , 7 S 11 13 15 A III II V*6*%f 2Fj.j ' Z 2'T 2' 2 2 ml 1 III III II a b Fig. 3 Resonance ionization spectrum (a) for the first step (X]) of excitation of '41sm with the ground (g) and excited (m) nuclear states. The diagram (b) shows the position of the lines of HFS due to ^'Sm atoms with ground- state (g) and isomeric (m) nuclei (from [8]).
There is relatively large number of very rare long-lived radioactive isotopes of cosmogenic and technogenic origin. In principle, any of the techniques of laser spectroscopy that are capable of detecting single atoms can be used for the highly selective detection of rare isotopes. The basic difficulty here is the attainment of maximum detection selectivity S, i.e., the detection of a small number N^ of atoms a very rare isotope A in the presence of a much larger number Ng of the main isotopes B (S= N-QI N^).
Selectivity is due to the small isotopic shift Avjs= v^- VQ of the spectral line resulting from one or several successive resonance transitions in the atom from ground to excited states. The width of the spectral line is the natural limitation on selectivity because of the overlap between the wings of the closely- spaced spectral lines of atoms A and B. However, the nature of this limitation is significantly different for different methods.
The most promising methods are based on resonance multistep excitation of the rare isotope in a multifrequency laser field, using the isotopic shift of several successive resonance transitions. The net result is that the selectivities S^ at the successive excitation and ionization steps are multiplied together.
Practical implementation of the principle of multiplication of selectivities encounters the difficulty that, for the most interesting long- lived isotopes, it is difficult to find an upward sequence of transitions with appreciable isotopic shifts, since only the ground state of the atom has an appreciable shift.
A universal way of overcoming this difficulty, and transforming the method of stepwise ionization into a real method of detection of rare isotopes, was proposed in [9]. The idea is to use collinear stepwise photoionization of a beam of accelerated atoms. The ions are accelerated and
-16- neutralized, and this is accompanied by the bunching of the longitudinal velocities of the atoms and, hence, by the removal of Doppler broadening during collinear excitation. The reduction in Doppler width is, in turn, accompanied by a Doppler shift of all the spectral transitions in accelerated atoms, which depends on the ion mass. This results in an artificial kinetic isotopic "mass" shift in all atomic transitions.
The method suggested in [9] was successfully implemented in a number of works. For example, a selectivity of 10^ was attained at the Institute of Spectroscopy in detecting the rare isotope 3 He
against the background of the abundant isotope ^He [10]. Based on this methods, an experimental setup was developed at the Mainz University for detecting the rare isotopes ^Sr and ^^Sr. Strontium- 90 is a most dangerous isotope, for it can accumulate in human bones and produce a high local radiation dose.
The detection limit attained amounted to 50- 10^ ^Sr atoms in the presence of 10*8 stable strontium
atoms in environmental samples [11]. So high a selectivity was achieved by combining preliminary mass
separation (selectivity over 1(P) with collinear resonance ionization.
To monitor the radioactive contamination of the environment specifically to reveal spreading
pathways and accumulation sites (those in the human body included), one has to detect radioactive
atoms at a very high sensitivity not attainable with the existing dosimeters. The RIS technique enables
one to do this. For example, the authors of [12] have managed not only to detect the ^21pr atom
(lifetime 22 min), but also to measure its optical spectrum, the number of the atoms in the sample being
a mere 105-106.
Plutonium is one of the most serious radioactive poisons in the environment, originating from
nuclear bomb test, nuclear power plants, and different type nuclear accidents in particularly Chernobyl
accident. Ultrasensitive and selective methods are required to study the ecological behavior and
migration in the environment. The detection technique actually used is a- spectroscopy, which has some
severe drawbacks: the sensitivity is limited (4- 10^ isotopes for "9pm)( fa& e]ement and isotope
selectivity is restricted because of other isotopes emit a- particles with similar energies, the
measurements take a long time, and 241pu can not be detected at all because of the absence of a-
emission.
-17- RIS method is an alternative, promising method. Its detection limit does not depend on the nuclear properties of the isotope under investigation such as the half- life and decay mode. The detection of trace amounts of Pu in environment samples was effected in [13] by the RIS technique used in combination with mass spectrometry (the RIS technique [2]). The sensitivity of detecting 239pu jn this experiment was as low as 10? atoms.
3. LASER SEPARATION
Tunable lasers made it possible selectively to excite practically any single quantum state of an atom or a molecule in the range 0.1- 10 eV. Systematic studies of the resonant interaction between laser radiation and matter have been under way since around 1970, mainly with a view to developing the laser isotope separation (LIS) process. Today, almost 25 years after the studies were commenced (see the review of these works in the monograph [14]), it has become clear that there exist at least three economically viable LIS methods (Fig. 4):
(1) the multistep isotope- selective photoionization of atoms, suggested in [15, 16];
(2) the two- step IR- UV isotope- selective photodissociation of molecules, proposed in [16,
17]; and
(3) the multiphoton IR isotope- selective photodissociation of polyatomic molecules, discovered in [18, 19].
fatf/Mfff A++ e" n-l+B h(D2
/iffli
a) b) c) Fig. 4 Economically viable methods of laser isotope separation: a) multi- step isotopically selective ionization of atoms; b) two- step IR-UV (vibrationally- mediated) isotopically selective photodissociation of molecules; c) multiple- photon IR isotopically- selective photodissociation of polyatomic molecules.
The multistep isotope- selective photoionization of atoms was successfully used in the
Lawrence Livermore Laboratory Program AVLIS for separating uranium isotopes [20J. The IR- UV
-18- isotope- selective photodissociation of molecules (UF5) was also studied for the purpose of uranium isotope separation at the Los- Alamos Laboratory [21]. And finally, the multiphoton (MP) IR isotope- selective photodissociation of polyatomic molecules is now at the stage of practical application to separation of isotopes of both light elements (e.g. ^C and ^C [22]) and uranium [23].
Since its decision in 1985 to forego continued development of gas centrifuge technology, the
Department of Energy of USA has spent hundreds of millions of dollars developing AVLIS technology to enrich uranium for commercial- grade nuclear fuel. The major AVLIS plant systems and facilities are
[24]:
(L) the chemical processing of uranium compounds to produce the metal feed;
(2) die process of dye lasers that generate red- orange light used to photoionize (by three- step excitation) the heated uranium vapor;
(3) the copper- vapor lasers that pump the process lasers,;
(4) the separators for enrichment of uranium in the 235U isotope;
(5) the chemical processing to convert the metal product to uranium oxide.
The AVLIS laser system (copper lasers with average power «10 kW and dye tunable lasers with average power »2 kW) has been exceptionally well engineered at LLNL. It is major advance in laser engineering.
The AVLIS technique is fairly universal and can be used to separate the isotopes of a series of elements., although a special R/D program is required for each element. At least three examples of
Applications of AVLIS technology can be given: Gd, Pu, Zr [24].
(1) The naturally occurring mixture of gadolinium oxides is now as turnable poison in essentially all building water reactors and many pressurized water reactors. Greater improvement with
Gd poison certainly could be achieved of ^Gd enriched material were available, particularly for fuel elements. Availability of Gd enriched to 80% ^^Gd would mean that Gd loading in the fuel could be reduced.
(2) Pu isotopes (238, 239, 242, 244) are unique and invaluable for powers sources, safeguards, and other special applications. The quantities required are tens or hundreds of kG for 2"Pu, tens of kG
-19- 2 2 for ^ Pu, and the order of lg for 244pu Production of these valuable isotopes is quite possible by
AVLIS.
(3) Most of hundreds power reactors in the world use 235TJ. enriched uranium oxide pollets clad in low- hafnium zirconium. All of these reactors pay heavy penalties with regard for performance
and economics because of neutron absorption by the "'Zr isotope and residual hafnium. AVLIS offers
the potential for removing this isotope and hafnium.
Let us finally emphasize that die multistep resonance ionization of atoms makes it possible to
separate not only elements (A- selectivity) and isotopes (A, N- selectivity), but also isobars where A=
Z+ N, A, Z, and N being the number of nucleus, protons, and neutrons in the nucleus, respectively, and
E is the isomeric nucleus excitation energy.
The molecular laser isotope separation (MLIS) technique is very an alternative of the AVLIS
technique. The MLIS technique is universal and applicable to all atoms that can be incorporate in
volatile molecular compounds, because isotope shifts inevitably manifest themselves in the vibrational
spectra of the molecules. Especially efficient is the MLIS technique based on the IR MP isotope-
selective photodissociation of molecules [18, 19].
Many experiments on isotopical selectivity of MP molecular dissociation have dius been
performed. These experiments have covered many isotopes, from light ones (hydrogen, deuterium,
tritium) to heavy ones (osmium, uranium) contained in very different molecules [22, 26]. Many of these
experiments became the basis of the laser isotope separation metiiods developing in numerous
laboratories of several countries. As an example, let us present more detailed data for the case of "C
isotope separation which was under development in the USSR through the cooperation of several
Institutes. (Institute of Spectroscopy, Kurchatov Institute of Atomic Energy and Institute of Stable
Isotopes).
To developed an economically viable method of isotope separation it is necessary first of all to
choose a polyatomic molecule which satisfies many requirements simultaneously: 1) high yield of MP
dissociation for a laser pulse energy fluence which is acceptable for optical windows of laser separation
cells (less than 2- 3 J/cm2); 2) high isotopical selectivity of MP dissociation for irradiation at CO2 laser
— 20- wavelengths; 3) low cost of initial molecular compound. In a present study of a large number of polyatomic molecules it was found that the molecule CF2HCI (Freon- 22) is optimal.
For reliable operation of the laser separation module it is very important to reduce the
requirements of the energy of the laser pulse performing the efficient IR MP dissociation. In this case,
first of all we can use the nonfocused beam of the CO2 laser and irradiate the molecular gas mixture in
a long separation call. Secondly, the probability of laser damage to the optical windows of the
separation cell can be diminished. The effective method of reaction where the first laser pulse performs
the isotopically- selective excitation of molecules in the vibrational quasicontinuum up to the
dissociation limit [26]. Two- frequency irradiation of the molecule CF2HCI improves the parameters of
carbon isotope separation significantly. Multifrequency irradiation of this model gives much better
results [27]. In this case very high selectivity of MP IR dissociation (up to 10^) and high yield of
dissociation can be achieved simultaneously. Figure 5 presents the dependencies of dissociation yield of
the CF2HCI molecule on the energy fiuence of the CO2 laser pulse in cases of 2- frequency, 3-
frequency and 4- frequency irradiation. Points of required degree of selectivity (100) of MP IR
dissociation are marked.
100 ~I TTTTT1 1 1 i 1 i"i 11| 1 1 1 1 1 rrr
CF2HCI ^
UJ >- S»IOO 2: 10 O < o 1050 ijQo o WAVENUMBER.CM-' to S2 o ' I t I 11 III B. O.OI 0.1 I 10 FLUENCE, J-CM-2
Fig. 5 Dependences of the dissociation yield for the CF2HCI molecule on the fiuence of the first laser pulse in the cases of (a) double-, (b) triple-, and (c) quadruple- frequency excitation. The arrows indicate the values of yield obtained at a selectivity of 10^ in the case of double-, triple-, and quadruple- frequency MP dissociation. Positions of laser frequencies for all cases are indicated on the linear IR absorption spectrum of CF2HCI (upper curve) (from [27]).
-21- This technique can also be used to separate other isotopes of interest in the nuclear technology, uranium isotopes in particular, by way of the IR MP dissociation of the UF6 molecule. For isotopes of heavy elements, vibrational isotope shift is relatively small, and so use should be made of gasdynamically cooled molecular beams in conduction with multiple- frequency IR excitation. It is also important to know the specific characteristics of the molecules in both low- and high- lying vibrational- rotational states. Table 3 lists the IR MP excitation parameters of polyatomic molecules that are critical to the optimum MLIS process.
Table 3. CRUCIAL (for MLIS) PARAMETERS of IR MPE/D of POLYATOMIC MOLECULES
PARAMETER PESULT
1. LOW-LYING RESONANT OPTIMAL FREQUENCIES for HIGH VIBRATIONAL- ISOTOPICAL SELECTIVITY of IR ROTATIONAL TRANSITIONS EXCITATION 2. VIBRATIONAL NUMBER of RESONANT VIBRATIONAL STOCHASTIZATION ENERGY EXCITATION STEPS ONSET 3. ABSORPTION SPECTRUM of OPTIMAL FREQUENCY and MINIMAL HIGHLY- ENERGY FLUENCE of IR VIBRATIONALLY EXCITED DISSOCIATING PULSE STATES
The most advanced MLIS program for uranium is under way at the AEC of South Africa [23].
Finally, we believe it quite possible to use laser selective photoprocesses for the processing of
nuclear wastes, specifically to (a) extract uranium and plutonium from liquid wastes, (b) return long-
lived fission products to the reactor for their nuclear transmutation to stable, nonradioactive isotopes,
(c) extract heat- generating radioactive elements, and (d) extract valuable metals, such as rhodium and
palladium.
4. CONCLUSION
(1) Laser control of nonproliferation of nuclear weapon using ultrasensitive and ultraselective
laser analytical methods;
(2) y- laser using laser separation of nuclear isomers [28] and laser search and study of
appropriate excited metastable nuclei [4- 6].
— 22 — REFERENCES
1) Demtroder W: Laser Spectroscopy (Springer- Verlag, Berlin) 1981
2) Letokhov V.S.: Laser Photoionization Spectroscopy (Academic Press, Orando) 1987
3) Hurst G.S. and Payne M.G.: Principles and Applications of Resonance Ionization
Spectroscopy (Adam Hilger, Bristol) 1988
4) Otten E.W. Investigation of Short- Lived Isotopes by Laser Spectroscopy (Harwood Acad.
Publ., Chur) 1989
5) a) Fedoseev V.N., Letokhov V.S., Mishin V.I. et al.: Optics Comm, 52, 24 (1984)
b) Zherikhin A.N., Kompanets O.N., Letokhov V.S. et al.: Zh. Eksp. Theor. Fiz. 86, 1249
(1984); c) Mishin V.I., Sekatskii S.K. et al. Zh. Eksp. Theor. Fiz. 93, 410 (1987); d) Alkhazov
G.D., Barzakh A.E., et al. Pis'ma ZhETF., 46, 136 (1987); e) Nuclear Physics, A477. 37
(1988); f) ibid, 337 (1990); j) Letokhov V.S., Mishin V.I. et al., J. Phys. G.: Nucl. Part. Phys.,
18,1177(1992)
6) Otten E.W.: Treatise on Heavy- Ion Science, vol. 8, ed. by D.A. Bromley (Plenum; New
York) p. 517 (1989)
7) Letokhov V.S.: Optics Comm., 2, 59 (1973)
8) Mishin V.I., Sekatskii S.K., Fedoseyev V.N. et al.: Optics Comm., 61., 383 (1987)
9) Kudriavtsev Yu.A. and Letokhov V.S. Appl. Phys. B29, 219 (1982)
10) Aseyev S.A., Kudriavtsev Yu.A., Letokhov V.S. and Petrunin V.V. Optics Lett., 16, 514
(1991)
11) Monz L., Hohmann R., Kluge H.-J. et al. Resonance Ionization Spectroscopy -1992, ed. by
CM. Miller and J.E. Parks (IOPP, Bristol) p. 225 (1992)
12) Andreev S.V., Letokhov V.S., and Mishin V.I. Phys. Rev. Lett., 59, 1274 (1987)
13) Urban F.- J., Deisenberger R., Herrmann G. et al. Resonance Ionization Spectroscopy- 1992,
ed. by CM. Miller and J.E. Parks (IOPP, Bristol) p. 233 (1992)
-23- 14) Letokhov V.S.: Nonlinear Laser Chemistry. Multiple Photon Exciation, (Springer, Berlin) p.
417(1983)
15) Letokhov V.S.: Soviet Patent No 784679. Appl. on March 30 (1970)
16) Ambartzumian R.V. and Letokhov V.S.: Appl. Opt. Ji, 354 (1972)
17) Letokhov V.S.: Soviet Patent No 784680. Appl. on March 30 (1970)
18) Ambartzumian R.V., Letokhov VS., Ryabov E.A. and Chekalin N.V.: Pis'ma ZhETF, 20, 597
(1974)
19) Ambartzumian R.V., Gorokhov Yu.A., Letokhov V.S. and Makarov G.N.: Pis'ma ZhETF, JI,
375 (1975) [JETP Lett. 21, 171 (1975)]
20) Paisner J.: Appl. Phys. 43B, 252 (1988)
21) Jensen R.J., Judd O'D.P. and Sullivan J.A.: Los Alamos Science, 3, No 1, 6 (1982)
22) Velikhov E.P., Baranov V.Yu., Letokhov V.S., Ryabov E.A. and Starostin A.N.: Pulsed CO-
Lasers and their Application for Isotope Separation (Publ. House "Nauka", Moscow) pp.
304 (in Russian) (1983)
23) Du Toit G.: The South African Mechanical Engineering. 42, 61, (Febr. 1992). Kemp D.M,
Bredell P.J., Ponelis A.A.and Ronander E. Bull, of the Research Lab. for Nuclear Research.
Spec. Issue l,p. 1 (1992)
24) Alternative Applications ofAVLIS Technology. Report. (Nat. Acad. Press, Washington D.C.)
(1991)
25) Alkhazov G.D., Barzakh A.E., Denisov V.P. et al. Nucl. Instr. and Meth. in Physics Research.
B69. 517 (1992); Mishin V.I., Fedoseyev V.N., Kluge H.-J., Letokhov V.S. et al. ibid., B73.
550(1993)
26) Bagratashvili V.N., Letokhov V.S., Makarov A.A.: Multiple Photon Infrared Laser
Photophysics and Photochemistry. (Harwood, London) p. 512 (1985)
27) Evseev A.V., Letokhov V.S., and Puretzky A.A: Appl. Phys. B3JL 93 (1985)
28) Letokhov V.S.: Zh. Eksp. Theor. Fiz. 64, 1555 (1973).
— 24 — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY -
Improvement of Lasers Towards Its Theoretical Limit in the Light of Advanced Technology
Hiroshi Takuma Institute for Laser Science, University of Electro-Communications 1-5-1 Chofugaoka, Chofu, Tokyo 182 Japan
1. Introduction Performance of lasers have made a significant progress since the first appearance of ruby laser in 1960. In the present state of arts, the peak power of femt second lasers exceeds 2 TW even in commercially available systems, and present day target of the competition in high peak power generation club is 100 TW. The wavelength range of the fundamental output has also been extended from the long wavelength limit of a few millimeter to the short wavelength limit of 100 nm, and the amplification in the X-ray region has been confirmed on various atomic inner shell transitions, the shortest wavelength reaching around 40 nm. The single frequency output can now be obtained rather commonly by the use of commercially available units, and the spectral spread of the oscillation frequency in free-running lasers reaches to the order of a few kW. By stabilizing the laser frequency by locking it to the center of a high finesse Fabry-Perot resonator or sharp atomic or molecular lines reached to mHz relative and a few Hz absolute stability. The progress is not only in the laser performance but also in the efficiency and handiness of lasers. Solid state lasers pumped by diode lasers, for example, can be carried by hand and operated at any place where the power line is available. The improvements in the nonlinear optics has also been enormous. The efficiency of the second harmonic generation exceeded 50% for cw near infrared fundamentals. Such high efficiency is the result of the improvement in the purity of the output mode, introduction of new materials having high nonlinear susceptibilities, and also by the use of intra-cavity nonlinear processes. It should be noted that those progress were possible in the light of technical developments in materials and optical components. Especially development of low loss optical components and high power monolithic diode lasers were extremely important technical development which brought the performance of lasers toward its theoretical limit.
-25- In the present paper, the impacts of super mirrors, which are the reflectors having, high reflectivity and extremely low loss, and advanced semiconductor lasers and their arrays will be reviewed as examples of the effect of advanced techniques on the lasers.
2. Impact of Super Mirrors on Laser Technology One of the most important progresses in the optical technology which influenced the laser technology is the development of extremely low loss high reflectance mirrors. It is now not a dream to obtain a mirror with a reflectivity of 99.999% with a loss smaller than 5 ppm. A Fabry-Perot resonator made by such mirrors can have a finesse of several hundred thousand, and the transmission coefficient can be larger than the loss. A Fabry-Perot etalon with optimized coupling, that is, the coupling loss is equal to the internal loss of the resonator, the theoretical transmission coefficient should be 100 % at the resonant frequency. Even if such may be difficult to realize, a fairly high transmission coefficient can readily be obtained. Thus it is now not difficult to make a feedback loop to stabilized a laser at the resonant frequency of such a high finesse Fabry-Perot etalon, and the accuracy of locking the laser frequency is extremely high. As a fact to demonstrate it, locking stability on the order of shot noise limit has been demonstrated1 in low frequency region, by comparing the output frequencies of two lasers locked to the same frequency standard as shown in the example of Fig.2, using a setup as schematically shown in Fig.l.
L1 Fl AOM PM ^T|4:::[ZZl:::::Q:::a:% Temp. PZT Reference Fabry-Perot cavity
OSC F = 25.600, FSR = 750 MHz L2 A/2 PBS A/4 slow fast 20.0 MHz SERVO
L = 200 mm OPD and &+ 8PD's 4k OBM L3
Fig.l Schematics of a frequency stabilization
26- _ 10° S— III •i-rmj —I-|-|THII| —i—111 nii| i i iiiu| N -i
1» Ml J (a) 1 -; 10 F! -pi n 1 i * 10" -1 r
CD : - 3 •i 10- r (b) ~\ c ,.~ Shot noise limit : 10- 5 111 i mil i 111 • • tit •1-1 1 llll 101 102 103 104 10£ Fourier frequency [ Hz ]
Fig.2 Frequency spectrum of stabilized diode pumped YAG laser.
Root Allan variance of two Diode-pumped 77 Nd:YAG lasers locked to one/two FP cavities i 10 H llMI>1 i nni^ imwj minq miii^ IIUIM i nmq i mm
(a) o o 10-10 4 ~ 10 tr • cJ 0 3 D b" 10 |- C 10 -12 a D • g lO^ D (b) .2 -13 1 o no CM b 1 io t (c) 14 c 10- Fig.3 Stability of three types of lasers. 27- On the other hand, there is still much larger amount of instability in the absolute frequency. By observing the beat of two lasers locked respectively to separate Fabry-Perot etalons, it is found in most cases that the absolute frequency fluctuation is on the order of 10 Hz. The cause of much poorer absolute frequency stability means that the resonant frequency of the Fabry-Perot resonator is fluctuating by unidentified causes. There are several possible causes to fluctuate the center frequency of Fabry-Perot etalons It was found that isolation of acoustic vibration can improve the frequency stability of a gas laser. For high finesse etalon, the intensity of radiation in the Fabry-Perot etalon is extremely high, and even low loss dielectric thin films can easily be heated to fairly high temperature. Such a condition may cause thermal deformation of the reflecting surface, and can be most serious obstacle in increasing absolute stability of laser frequency. Figure 3 shows an experimental result on the stability comparison of a couple of lasers in various frequency locking condition. 100 8PD'S \/ 50 25 12.5 BS g:>:S::::Q::::>::: * ! ! I z(\" • S 9 i • v £ IfH^ m mm * • • •" * "C"' 50 ^ y : : N>i. X « • c^lThc""""*••"•*•••• • f ^^ IITHC* " " " * * ^rm m " * M Optical power divider (OPD) to an inductor Fig.4 Schematic illustration of detecting system for high power cw lasers which may be applied to form a high power beam combining many beams. -28- Another possible application of high quality reflecting surface is beam superposition. Because it is extremely difficult to generate a high power radiation in a single oscillator or amplifier, we want to superposition of many beams to form a single coherent high power beam. Such is impossible in the present state of arts, but of course theoretically possible. In order to realize it, we must have perfect beam splitters. It is also the same to measure the power of a high power beam, because the area of a detector should be unpractically large to do it without suffering saturation. It was demonstrated that power of a beam can be measured beyond the saturation limit by using array of detector and splitting the beam as shown in Fig.4. If the system schematically shown in Fig.4 is used in the inverse direction and the array of detector were replaced by array of oscillators, it should work as a beam combiner. It is also true that fundamental developmental works can improve the technology further. For example, an ultra-stable laser and high finesse resonator can detect a small amount of degeneracy breaking of cavity modes due to the deviation of the mirror surface from spherical surfaces. For example, it was found that TM01 and TM10 modes, which should be degenerate in a true axially symmetric cavity resonators, are observed to have different resonant frequencies by the use of ultra high finesse Fabry-Perot resonator and frequency stabilized lasers. 3. Future Prospects of Solid State Lasers Development of semiconductor laser technology brought a drastic progress of solid state lasers by using the semiconductor lasers as the pumping source. There are two important issues in the semiconductor laser pumping of solid state lasers. The first advantage is that the diode laser pumping defines the pumped volume so that the oscillation builds up in only one transverse mode. Under such a condition, true single frequency oscillation can be obtained without any special trick by designing a solid state laser to have very short cavity length. The second advantage of diode laser pumping is high efficiency. By choosing the output wavelength of a semiconductor laser at the most efficient place, we can use the diode laser photons efficiently to pump the solid state lasers. One minor problem in the diode laser pumped solid state lasers is that the most of the efficient solid state lasers generate invisible near infrared radiation. This is not a serious problem, because the quantum efficiency of modern near infrared detectors can be high, and also because efficiency of the second harmonic generation has been improved very much, exceeding 50% even for cw radiation in many cases. Such a high efficiency is established by employing intra-cavity — 29- frequency doubler which is resonant to the fundamental radiation, intensifying the fundamental beam intensity several hundred times or more. Such a progress on the solid state lasers, together with the development in the output power of diode laser arrays for solid state laser pumping and the consideration that a compact high power solid state laser can be adopted in automated machining systems, makes us convinced that the solid state laser can be an industrial laser for a broad range of mechanical machining. The only and very high obstacle in realizing such a condition is the extremely high price of the high power diode lasers and their arrays. Because such a technique has been developed for space born systems, their high price is rather intrinsic and very difficult to overcome unless a development of large market to satisfy the economical requirement in introducing a mass production line. On the other hand, expansion of the market is very difficult to realize without reducing the price. Thus it is extremely difficult to predict when, but in some near future most of the coating machine should be operation by lasers. 4. Advanced Semiconductor Diode Lasers If the present author were asked "Is diode laser pumped solid state laser the final shape of industrial laser?', the answer is "No, I don't think so". There is enough reason to assume that future semiconductor lasers will be able to do most of industrial applications of lasers. On the other hand, there is a large room for further development, of their performance, although excellent performance of modern diode lasers brought its characteristics very close to their theoretical limit. The inconvenience in using diode laser output is its strong axial asymmetry in angular divergence. However, development of manufacturing technique has improved the mode purity of diode lasers, and the output beam can be converted into a single parallel beam by using specially designed optical systems. One of the basic concepts is illustrated in Fig. 4. Another disadvantage of broad spectral line width of the output of diode lasers have been improved by the use of optical feedback from external frequency selective devices such as Fabry-Perot etalon or gratings. Mode hopping is considered to be intrinsic problem of semiconductor lasers. However, it can be removed by making on one of the cleavage surfaces of the diode laser by anti-reflection coating of extremely high quality, and installing the diode in an outside optical resonator. Because the emission spectrum of a semiconductor used to make diode lasers normally have very broad emission spectral width, we can expect to make a diode laser having broad frequency tuning -30- range without mode hoping. There have been several developmental works carried out in such a direction, and even commercial units have been introduced into research tool market.. Most of present commercial diode lasers are designed for the use by itself, but there must be a great deal of applications in which a diode having one end with a perfect anti-reflection coating is more suitable than conventional ones. One good example is an optically pumped rubidium or cesium clock which provides the secondary frequency and time standard. Application of such accurate frequency source is expected to be useful in high precision aviation control and it is expected that there are many other applications in which high accuracy time and frequency standard can be useful if it were maintenance free having a long MTBF. One serious problem which should be overcome for diode lasers to be employed in industrial applications. That is to increase the average output power. There is an intrinsic difficulty to obtain high power cw output from a monolithic diode laser although recent works made it possible to obtain more than a watt of output. However, the total output power can be as high as kW in diode arrays. A difficult problem here is how to have a high quality coherent beam by combining may beams coming out of an array of diodes. Injection locking of diodes to the same master oscillator seems to be a reasonable method to obtain simultaneous coherent oscillation of diode arrays, but the coupling between adjacent diode in an array complicates the characteristics of such an array laser. At present there is no well established method to combine the output of diodes in an array into one coherent beam. In spite of such difficulties, intensive works to develop such a laser system should be carried out. An encouraging fact is that the most of what are theoretically possible have been realized in the past if they are really needed. It may take a long time to realize it, and diode laser pumped solid state lasers will be used as the most important industrial lasers for along time until such ideal diode laser arrays are realized. 5. Advanced Applications of Diode Lasers Before high power diode laser applications come into realistic industrial field, one of the most important features of semiconductor diode lasers, tunability, should have a broad range of applications. Applications in spectroscopy is a direct use of tunability of diode laser, and several advanced technique which are presently only used in the frontiers of fundamental physics may have completely new applications. -31 — One of such possibilities is the application of atom optics, which is a new field opened by the use of advanced technique of laser cooling and trapping of neutral atoms. Because the Schrodinger equation of an atom freely moving in a space is exactly the same as the Maxwell equation of electromagnetic waves propagating in space, spatial distribution of the index of refraction of the latter corresponding to the potential distribution of the former. This means that the atoms behaves like light. Thus a new field called "atom optics" is now intensively investigated, and atom microscope corresponding to the electron microscope may valid as a new possible technique, by the use of long de Broglie wavelength established by laser cooling method. In order to make an image of an object by atoms, a field distribution having maximum potential on the axis is needed, but it is impossible to be made by dc electric field. However, resonant mode of a cylindrical microwave cavity of TM()] mode provides such a field. Such experiment has been tried by the use of neon metastable atoms, and focusing of abeam is demonstrated in a preliminary experiment' as shown in Fig.5. 59Bnm Lasor Cooled oocccccooo/"] IS5N0* ^ - • r'fc:: > UioroWnvc Cavity 3 mm KCP ^ss Fl. plate I |,0 camora Fig. 5 Schematics of the setup for focusing experiment of neon atom(left) and example of its results (right). Right upper: image without microwave field, and right lower: image with microwave field. 32- There are many other applications of laser cooling technique, such as the atom interferometry applied in high precision measurement of rotation, gravitational field, and others. In order to make those application of laser cooling, it is essential that such systems can be constructed by diode lasers. In order to carry out laser cooling in ideal condition, diode laser must have usable output on the order of 100 mW, and frequency stability better than 1 MHz. These figures are not far from present state of arts, and we see many laboratories where laser cooling experiments are carried out by the use of diode lasers, showing the feasibility of the use of diode lasers in laser cooling. There must be many new technical development by introducing diode lasers in the spectroscopy and its applications, and new advanced application are expected to be realized. References 1) N. Uehara and K. Ueda, Jpn. J. appl. Phys. 33,1628-1633 (1994). 2)N. Uehara and K. Ueda, Optics Lett., 19, 728-730(1994). 3)H. Takuma, K. Shimizu, and F. Shimizu, Quantum Optics VI, J. Harvey and D. Walls, Ed. (Springer Verlag, 1994). - 33- Session II Basic Laser Science Proceedings or the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- "Laser Cooling and Trapping of Atoms" Steven Chu Physics Department Stanford University, Stanford CA 94305 The basic ideas of laser cooling and atom trapping will be discussed. These techniques have applications in spectroscopy, metrology, nuclear physics, biophysics, geophysics, and polymer science. Keywords : Laser Cooling, Atom Trapping 1. Laser Cooling The field of laser cooling and trapping of atoms and other particles has blossomed during the last ten years and a short article such as this can not hope to adequately explain the major developments. This article will give a snapshot in time of the current state of the art in laser cooling, paying particular attention to those developments that this author believes will have general applicability. For a more thorough review of these advances, the interested reader is referred to a number of review articles.1,2 The electromagnetic forces used to manipulate atoms are feeble: for example, the average kinetic energy of an atom at room temperature is a thousand times greater than depth of our deepest optical traps. The rapid progress in atom manipulation techniques is based on the fact that atoms can be readily cooled in "optical molasses" to microkelvin temperatures. As originally conceived by Hansch and Schawlow in 19753, laser cooling is based on the Doppler effect for a two-level atom. If an atom is irradiated by counterpropagating laser beams tuned to the low frequency side of the atomic resonance, a moving atom will Doppler shift the beam opposing its motion to increase the scattering probability while the beam co- propagating with the atom will be frequency shifted away from the resonance. Thus, there will be a net scattering of photons opposing the motion of the atom with each scattering event transferring an average momentum of #k to the atom. By surrounding the atom with three sets of counterpropagating beams along the x,y,and z axes, a drag force opposing the velocity of the atom can be generated. This configuration of laser beams was predicted* to cool — 35- atoms to a minimum temperature of knT=/2l72 for a detuning Av of the laser equal to 172 in the limit of low laser intensity. The first three dimensional laser cooling of atoms was demonstrated by Chu, ct al., in 1985.5 The first measured temperatures and confinement times were consistent with the Doppler cooling, but a number of subsequent experimental findings were in strong disagreement with theoretical predictions including insensitivity to intensity imbalance between the counterpropagating beams,6 and anomalously long lived molasses in which the storage times of the atoms increased by a factor of 50 if the laser beams were misaligned.7 The biggest failing of two-level theory of optical molasses was uncovered by the NIST group. They found that sodium atoms could be cooled to temperatures at least six times colder than the predicted minimum temperature.8 This discovery prompted Dalibard and Cohen-Tannoudji9 and Chu and co-workers10 to develop a new theory of laser cooling. The basic idea of the new theory is based on a combination of optical pumping effects, induced light shifts, and the motion of atoms in light fields with polarization gradients. Polarization gradients can not be avoided in any three dimensional configuration of light beams. As a simple case, consider two counterpropagating, linearly polarized laser beams with orthogonal polarizations. The two light fields will generate regions in space where the local polarization has jg2,p2> with external momenta p, and p2 connected to an excited state |e> via light fields with quantum mechanical amplitudes A( and A2. In the case where A, = A2 and the energies - 36- / of the two ground states are equal, the state (lA 2)(Jg,,p, > - jg2,p2>) is "dark" in the sense that it is not be connected to the excited state by the radiation field. (This state is one of the new, re-diagonalized eigenstates of the system of jgi>, |g2>, and |e> after one turns on the light fields.) The deBroglie frequencies of the two states w, 2=E, 2/fi are not equal since E, 2 is the total energy (internal plus external) of the atomic states unless the magnitude of the momenta are the same. Thus, the "dark" state will oscillate into the orthogonal state (lA/2)(|g,,p,> + |g2,p2>) which is strongly coupled to the radiation field. As the atoms spontaneously scatter many photons, they may randomly scatter into a state where p, — -ftk and p2~ +#k so that the dark state remains dark for an appreciable amount of time. The width of the momenta distribution decreases the longer an atom remains in the radiation field since only atoms with momenta approaching the exact non-coupled dark state will not oscillate into the coupled state. If one considers the width of each of the momenta peaks about +hk as a measure of the "temperature" of the atoms, it is possible to cool the atoms to temperatures less than the recoil energy. The second technique, which uses velocity selective stimulated Raman transitions to push atoms towards the v=0 state,12 is similar to the frequency chirp method of cooling that was first employed to slow and stop atoms in an atomic beam.13 Stimulated Raman transitions can be extremely Doppler sensitive since the width of the transition is determined by the time of the transition (alternately the Rabi frequency of the driving light fields) and not the lifetime of the excited state.14 Thus, atoms cooled to optical molasses temperatures can be further cooled with Raman transitions by pushing the atoms towards the v=0 state. Consider an atom with ground states 11 > and 12 > and an excited state 13 >. Atoms are initially optically pumped into state | l,p>. A stimulated Raman transition moves a subset of the atoms into state 12,p+2/»k> where the direction of k is chosen so as the reduce the velocity of the atom. All atoms with velocities vy^O can be pushed towards the v=0 state. An atom put into the j2> state is optically pumped back into the state j l,p'>. In the spontaneous emission process, the new momenta p' may remain nea. v=0 or could be kicked out to a higher momentum state. If the atom remains near v=0, subsequent Raman pulses are designed not the excite the atom but if it receives a spontaneous kick to away from the v=0 state, a subsequent Raman pulse will have another chance to push it back to the v=0 state. This cooling process was applied to sodium pre-cooled with optical molasses to 35 /xK. Roughly 50% of the atoms were cooled to 100 nanoKelvin, corresponding to an increase in velocity phase space of an order of magnitude.12 Both of these cooling techniques can be extended to two and three dimensions. -37- 2. Atom Trapping The first two techniques used to trap atoms exploit either the permanent magnetic dipole or induced dipole moment of atoms. Since div B = 0 and div E = 4irp = 0 in a region of space free of charge, one can not construct a static E or B field that has a maximum value in a region of space void of charges or currents. Magnetic trapping of atoms15 is based on the idea that a local minimum in the a static magnetic can be made. If the magnetic moment is aligned anti-parallel to the magnetic field, the particle will be driven to the region of space where its energy W=-ji«B is minimized. As long as the particle moves slowly in the magnetic field, its quantum mechanical alignment with respect to the magnetic field is preserved. Magnetic traps have been used to confine spin-aligned hydrogen in order to avoid recombination of the hydrogen atoms into hydrogen molecules that occurs at the walls of a conventional cryostat.16 Atoms in their ground state do not have permanent electric dipole moments, and trapping must be done through an induced dipole moment. A focused laser beam produces a time varying electric field maximum at the focal point, and as long as the driving electric field is below the resonances of the atom, the induced dipole moment will be in phase with the E field. For this type of trap,17 the particle's energy is minimized by seeking regions of space where the electric field is strongest. If the laser field is tuned above the atomic resonance, the induced dipole moment will be aligned anti-parallel with the driving field and the atom will seek an E field minimum. Atomic traps based on a repulsive trampoline bowl of light and gravity have been proposed.18 If the dipole trapping light is tuned far from resonance, the probability of scattering photons is greatly diminished and the trap begins to behave like a purely conservative potential well.19 Sodium atoms with a resonant frequency at 589 nm have been recently trapped with light at 1.06 microns in our laboratory. We have also combined cooling via stimulated Raman transitions with this far off-resonant dipole trap and have cooled atoms along the axial direction of the trap to a velocity spread of 0.6fik/M. This method can be generalized to three dimensions in order to produce very cold, moderate density quantum gases.2 One of the most widely used atom traps is a hybrid magneto-optical trap.21 A weak magnetic field is used break the degeneracy of the Zeeman sub-levels of the atom. The slight shift in the energy levels cause counterpropagating a+ and a laser beams to scatter more photons in the direction towards B=0. This type of trap can be thought of as mostly optical molasses with a magnetic field to tell the atoms where to collect. A great advantage of the magneto-optic trap is that the laser intensity in each of the — 38- beams is on the order of a few milliwatts/cm2 so large volume traps can be made with low intensity lasers. Wieman and collaborators observed that the trap could collect atoms directly from a sealed, low-vapor pressure cell (~ 10"8 torr).22 The magneto-optic trap currently serves as a convenient starting point: as many as 4xl010 atoms at a density of ~ 10" atoms /cm3 can be collected within 0.5 second.23 Following a few millisecond exposure of polarization gradient molasses (accomplished by turning off the magnetic field) a dense sample of atoms at a temperature of k„T=~(3- 4flk)2/2M can then be used in a variety of experiments. Recently, 21Na with a half-life of 22.5 sec was trapped in a magneto-optic trap.24 21Na decays by positron emission to its "mirror" nuclei 21Ne and can be used in the precision test of the vector-axial structure of the weak interactions. The 21Na was produced with the reaction 24Mg (p,a) 21Na using a 25 MeV proton beam. The sodium atoms effusing from an atomic beam oven were collimated by transverse optical molasses, slowed by an opposing laser beam in an inliomogeneous magnetic field25 and brought to rest in a magneto-optic trap approximately 2 meters away from the production target. 3. Applications The first demonstration of laser cooling and trapping of atoms occurred less than a decade ago, and we have only begun to recognize the myriad of potential applications. With cold atoms, one can create an "atomic fountain" of atoms in which microkelvin atoms are tossed upwards in a ballistic trajectory.26 The energy levels of these atoms, free of any perturbation except gravity, can be measured with great precision because of the long observation time available. Our second generation atomic fountain measured the hyperfine splitting of the cesium ground state with a short term stability that exceeds the atomic clocks maintained by standards laboratories by an order of magnitude.27 An analysis of the systematic errors anticipated with an atomic fountain clock estimates that the absolute accuracy will be in the range of Ai>fa~ 10~15 to 10"16.28 Much more dramatic improvement is expected for fountain clocks that measure higher frequency transitions such as the 600 Ghz transition in Mg or near infrared or optical transitions in atoms such as magnesium, calcium and silver.29 A number of different atom manipulation techniques have been devised which are analogous to optical components such as lens, mirrors, gratings.30 Despite the proliferation of these methods, it is important to realize that "atom optics" can, in principle, be more powerful than photon optics. Photon optics is limited by a "brightness" conservation theorem -39- which states that optics can never be used to image a light source onto a image plane with an increase in the intensity of the light per unit bandwidth divided by the average divergence of the light. In contrast to photon optics, we can cool atoms and thereby increase the brightness of an atomic source. The first illustration of this idea was the "atomic funnel", where atoms from an atomic beam are captured by a magneto-optic trap, cooled and compressed in the trap, and then allowed to escape from the trap.31 The phase space density of the initial atom beam was increased by four orders of magnitude in the first version of this "phase space compressor", and another factor of 104 is possible if one incorporates all the tricks in laser cooling that have been realized after this work was completed. Atom optics have been used to construct atom interferometers.32 These devices are especially sensitive measuring devices when slow atoms are used in the atomic fountain geometry. For example, the acceleration due to gravity has been measured with a precision of 3 parts in 10"8,33 and we are working to improve the relative precision by three to four orders of magnitude and hope to measure g with an absolute precision of one part in 1010. Comparably sensitive gyroscopes and gravity gradiometers are being designed. A portable gravity meter or gravity gradiometer based on a compact Cs cell and diode lasers will find practical applications in oil and mineral exploration and the measurement of small land or ocean level changes that can not be resolved by the global positioning satellite system. Continued improvements in both cooling and trapping techniques may finally lead to the condensation of a dilute Bose gas or the creation of a degenerate Fermi gas. These novel quantum systems will provide unique opportunities to study the collective properties of fundamental quantum systems. Furthermore, the prospect of condensing a majority of the atoms into a single quantum state opens up the possibility of creating an atomic source with unprecedented properties analogous to a laser operating in the TEMQQ mode. Bose condensation can also be viewed as the theoretical limit to brightness enhancement of an incoherent atomic source. Optical traps based on a single focused laser beam can also trap particles between 0.02 to 10 jxm.34 Individual living cells and even organelles inside a cell can be manipulated and viewed simultaneously.35 Individual molecules can not be directly held by a laser beam at room temperature, but the attachment of micron-sized polystyrene "handles" to biological marcomolecules such as DNA 36 or actin or microtubule filaments37 have enabled researchers to quantitatively measure the mechanical behavior of individual molecules. Fundamental questions in polymer physics he also been addressed with these single molecule manipulation techniques.38 The ability to hold macroscopic particles with light opens up the possibility of — 40- suspending "target" materials in a manner such that the supporting mechanism (the light) will not interfere with the experiment. An example where a non-interacting support structure may be useful is the suspension of micro-spheres of tritium for hydrogen implosion experiments. The field of laser cooling and trapping is young. Much of the progress has been made within the last decade and there are no signs that progress in the field is waning. New trapping and cooling ideas will no doubt lead to exciting new applications. References: 1. For an excellent review of many aspects of laser cooling, see Laser Manipulation of Atoms and Ions, Proceedings of the International School of Physics "Enrico Fermi'', Course CXV1I1, eds. E. Arimondo, W.D. Phillips, and F. Strumia, (North Holland, Amsterdam, 1992). 2. For other reviews, see S. Chu, Science, 253, 861, (1991); Fundamental Systems in Quantum Optics, Les Ilouches, Session LII, 1990, eds. J. Dalibard, J.M. Raimond and J. Zinn-Justin (Elsevier Science Publishers, Amsterdam, 1992); C. Cohen-Tannoudji and W.D. Phillips, Phys. Today, 43, 33 (Oct 1990); the special issue of the J. Opt. Soc. Am. B6, eds. S. Chu and C. Wieman, (1989). 3. T.W. Hansch and A.L. Schawlow, Opt. Comm. 13, 68 (1975). 4. See, for example D. Wineland and W. Itano, Phys. Rev. A 20, 1521 (1979), and J. Gordon and A. Ashkin, Phys. Rev. A 21, 1606 (1980). 5. S. Chu, L. Hollberg, J.E. Bjorkholm, A. Cable, and A. Ashkin, Phys. Rev. Lett. 55, 48 (1985). 6. P.L. Gould, P.D. Lett, and W.D. Phillips, in Laser Spectroscopy VII, S. Svanberg and W. Persson, eds. (Springer-Verlag, Berlin, 1987). 7. S. Chu, M.G. Prentiss, A. Cable, and J.E. Bjorkholm, in Laser Spectroscopy VII, W. Persson and S. Svanberg, eds., (Springer-Verlag, Berlin, 1988) pp 64-67; Y. Shevy, D.S. Weiss, and S. Chu, in Spin Polarized Systems, S. Stringari, ed. (World Scientific, Singapore, 1989), pp 287-294. 8. P.D. Lett, R.N. Watts, C.I. Westbrook, W.D. Phillips, P.L. Gould, and H.J. Metcalf, Phys. Rev. Lett. 62, 1118 (1988). 9. J. Dalibard and C. Cohen-Tannoudji, J. Opt. Soc. Am. B 6, 2023, (1989). 10. P.J. Ungar, D.S. Weiss, E. Riis, S. Chu, J. Opt. Soc. Am. B 6, 2058, (1989). -41- 11. M. Kasevich and S. Chu, Phys. Rev. Lett. 69, 1741 (1992). 12. N. Davidson, H.J. Lee, M. Kasevich, and S. Chu, Phys. Rev. Lett. 72, 3158 (1994). 13. W. Ertmer, R. Blatt, J.L. Hall and M. Zhu, Phys. Rev. Lett. 54, 996 (1985). 14. M. Kasevich, D.S. Weiss, E. Riis, K. Moler, S. Kasapi, and S. Chu, Phys. Rev. Lett. 66,2297 (1991). 15. A. Migdall, J.V. Prodan, W.D. Phillips, T.H. Bergman, and H. Metcalf, Phys. Rev. Lett. 54, 2596 (1985). 16. H.F. Hess, G.P. Kochanski, J.M. Doyle, N. Masuhara, D. Kleppner, and T.J. Geytak, Phys. Rev. Lett. 59, 672 (1987). 17. S. Chu, J.E. Bjorkholm, A. Ashkin, and A. Cable, Phys. Rev. Lett. 57, 314 (1986). 18. M. Kasevich, D. Weiss, and S. Chu, Optics Lett. 15, 667 (1990); A.M. Steane, et al. Europhys. Lett. 14,231 (1991). 19. J.D. Miller, R.A. Cline, and D.J. Heinzen, Phys. Rev. A47, R4567 (1993). 20. H.J. Lee, C.S. Adams, N. Davidson, M. Kasevich, and S. Chu, to be published. 21. E.L. Raab, M. Prentiss, A.E. Cable, S. Chu, and D.E. Pritchard, Phys. Rev. Lett. 59, 2631 (1987). 22. C. Monroe, W. Swann, H. Robinson, and C.E. Wieman, Phys. Rev. Lett. 65, 1571 (1990). 23. K.E. Gibble, S. Kasapi, and S. Chu, Optics Letters 17, 526 (1992). 24. Z-T. Lu, era!., submitted to Phys. Rev. Letters, 1994. 25. J. Prodan, et al., Phys. Rev. Lett. 54, 992 (1985). 26. M. Kasevich, E. Riis, S. Chu and R. Devoe, Phys. Rev. Lett. 63, 612 (1989); A. Clairon, et al., Europhys. Lett. 16, 165 (1991). 27. K. Gibble and S. Chu, Phys. Rev. Lett. 70, 1771 (1993). 28. K. Gibble and S. Chu, Metrolgia 29, 201 (1992). 29. For discussions of potential atom candidates for clocks, see F. Strumia, in Laser Science and Technology, eds. A.N. Chester, V.S. Letokov, and S. Martellucci, (Plenum Press, N.Y., 1988) pp. 367-401; J.L. Hall, M. Zhu, and P. Buch, J. Opt. Soc. Am. B6, 2194 (1989). 30. For a review, see C.S. Adams, M. Segel, and J. Mlynek, Phys. Rep. 240, 143 (1994). 31. E. Riis, D.S. Weiss, K. Moler and S. Chu, Phys. Rev. Lett. 64, 1658 (1990). -42- 32. 0. Carnal and J. Mlynek, Phys. Rev. Lett. 66, 2689 (1991); D. Keith, C. Ekstrom, O. Turchette, and D. Pritchard, Phys. Rev. Lett. 66, 2693 (1991); F. Riehle, Th. Kisters, A. Witte, S. Helmeke, and Ch. Borde, Phys. Rev. Lett. 67, 177 (1991); M. Kasevich and S. Chu, Phys. Rev. Lett, 67, 181 (1991). 33. M. Kasevich and S. Chu, Appl. Phys. B 54, 321 (1992). 34. A. Ashkin. J.M. Dziedzic, J.E. Bjorkholm, and S. Chu, Opt. Lett. 11, 288 (1986). 35. A. Ashkin and J.M. Dziedzic, Science 253, 1517 (1987). 36. S. Chu, Science 253, 861 (1991). 37. K. Svoboda, C.F. Schmidt, B.J. Schnapp, and S.M. Block, Nature 365, 721 (1993); J. Finer, R.M. Simmons, and J. A. Spudich, Nature 368, 113 (1994). 38. T.T. Perkins, D.E. Smith and S. Chu, Science 264, 819 (1994); T.T. Perkins, S.R. Quake, D.E. Smith, and S. Chu, Science 264, 822 (1994). -43- Proceedings orthc 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- NUCLEAR PROPERTIES STUDIED BY LASER SPECTROSCOPY Takashi T. INAMURA Safety Center, RIKEN Wako-shi, Saitama, 351-01 Japan Magnetic dipolc and electric quadrupolc moments of atomic nuclei arc determined by means of laser spectroscopy model-independently. Lascr-rf techniques enable us to make precision measurement of these moments. The hfs anomaly, i.e., Bohr-Wcisskopf effect will be studied systematically at CERN .ISOLDE shortly. keywords: Hypcrfinc structure, Nuclear moments, Laser, RF, and Atomic-beam magnetic resonance techniques. 1. INTRODUCTION First, suppose we ask the question: Is there any compelling reason to believe that the nucleus is an assemblage of protons and neutrons which always orbit the nuclear center in a way analogous with the planetory model of atomic electrons orbiting the point-like nucleus in the Coulomb potential? The answer is "No". The prob lem is: These days, most nuclear physicists start their studies assuming that the nuclear shell theory is apriori. I am second to none in believing the nuclear shell theory, but I would like to remind you how the nuclear shell model was worked out [1, 2]. The hyperfine interaction played a decive role in establishing the nuclear shell model, providing nuclear spins, magnetic and electric quadrupole moments. The great advantage of hyperfine spectroscopy as a technique for nuclear study, lies in the fact that the electromagnetic interaction is the best understood interaction in present day physics; we arrive directly at nuclear data being free from the complications of the nuclear force problem. — 44- fluorescence \ V- Hhfi ee \ )NN laser Tfii A Fig. 1. Feynman diagram for the hyperfine interaction and laser-induced fluorescences. The advent of tunable dye-lasers have revolutionized hyper fine spectroscopy for nuclear study. Since 1980, the idea of laser spectroscopy for nuclear study has been firmly established [3]. It is, however, still to be fully exploited in order to explore the nuclear structure, for instance, the spacial distribution of nuclear magnetization. 2. HYPERFINE INTERACTION The interaction Hamiltonian of an atomic system is written as H = HCC+HNN + HCA, (1) r where Hec is the interaction for the fine structure (terms), HNN f° the nuclear structure, and HCA for the hyperfine interaction Hhn that is a purely electromagnetic interaction between atomic electrons and the nucleus. A Feynman diagram for the hyperfine Table 1. Energy scales of interactions in the atomic system. eV MHz 6 14 HNN » 10 3 x 10 8 HCC =1 3 x 10 HeA » 10-6 3 x 102 — 45- Pc^c pc: electronic charge density PN: nuclear charge density je: electronic current density JN: nuclear current density R: nuclear radius Fig. 2. The coordinates for the hyperfine interaction. interaction is shown in Fig. 1. When you shine this atom with lasers, you will observe induced fluorescence lights. Laser-induced fluorescence spectra provide the nuclear information through their hyperfine structures. Table 1 lists an energy scale for each interac tion in the atomic system. The energy of the hyperfine interaction is something like 10-6 eV. The hyperfine interaction is written as H [ [ P^^^.jf f JMiMdTcdTN (2) to first order in the non-relativistic approximation, from which we have the energy eigenvalue for X = 0 2 W(E0) = -Ze2(l/rc> + (2/3)irZe21 ye(0) |2 WF = A = -^Bj(0)/IJ, B = eQs02V/^2). (5) — 46- From A and B constants we can extract a nuclear magnetic dipople moment \i\ and a spectroscopic quadrupole moment Qs model-inde pendently. According to the prevailing nuclear model [4, 5], the spectroscopic quadrupole moment Qs is related to the nuclear- model quadrupole moment Qo (intrinsic quadrupole moment) as Qs = Qo[3K2 - I(I+l)]/(I+l)(2I+3), (6) where Qo = 3ZR2$/(5n)V2, (3 being the quadrupole deformation parameter. 3. EXPERIMENTAL MEASUREMENTS 3.1 Collinear laser spectroscopy Figure 3 shows a schematic drawing of the experimental setup for fast atomic-beam collinear laser spectroscopy at JAERI, Tokai [6]. The reason why fast atomic beams with a kinetic energy of a few ten keV are remarkably efficient in making high-resolu tion spectroscopy of atoms is described in Ref. [3]; at JAERI, Tokai the radioactive singly-charged ions, 143Prll, which were accelerated up to 20 keV and mass-separated, were guided into the interaction region (a cage in Fig.3) where they crossed the counter-propagating laser beams. Laser-induced fluorescences were collected with a controller with computer data acquisition system filters laser beam '~~~t moss-separated ^-0- 1 7 l^^l ion_ beam lens ^- coge \\~~\ mirror einzel lens Fig. 3. Schematic of the experimental setup for collinear laser spectroscopy at JAERI, Tokai. — 47- 143PrII 601.94 nm Ci £ r-t TH I IX, •JwftrtJw^ mimf^^^'ILto ^ , 4f [X,)6s1„[9/2Il/2)' _L J I 0 3 4 5 Laser Frequency (GHz) Fig. 4. Measured hyperfine spectrum of the transition 43 5G2—5H3 (X = 601.94 nm) in l PrII. spherical mirror and detected with a photomultiplier (Hamamatsu R2256); the overall detection efficiency was 10-5. Hyperfine tran sitions (F—F1) thus observed in 143PrII are shown in Fig. 4. According to Eq. (4), we have A and B constants for 143PrII as well as for stable 141PrII as summarized in Table 2. By refering to the known magnetic moment of 141Pr (I = 5/2), i.e., 143A/141A = 143 141 141 (HI/I)/ (HI/I), where HI= + 4.2754(5)HN [7], we have M\H = 143 +2.701 (4)HN- Similarly, we have Qs= +0.77(16) eb by refering to 141Qs [6]. This is the first determination of nuclear moments for the radioactive nuclide 143Pr (I = 7/2). Table 2. Hyperfine interaction constants A and B determined for 43 141prn and l PrII (in units of MHz). 43 Level 141A 143A 141B 1 B 3 4 4f ( l0)6p5H3 1029.5(7) 463.6(1.5) -14(2) 150(22) 4 4f3( l0)5d5G02 1703.6(7) 769.1(1.1) -11(5) 95(11) -48 — 3.2 Laser-rf double resonance spectroscopy The principle of laser-rf double resonance spectroscopy was first demonstrated by Rosner et al. [8] as an alternative high- resolution spectroscopic technique to the well-established magnetic resonance one [9]. This technique has been proven powerful and efficient in making high-resolution hyperfine spectroscopy [10]. Refractory elements, however, are left to be studied because their atomic beams are extremely difficult to produce. Here I shall present our laser-rf double resonance (LRDR) technique using an Ar-ion sputtering atomic beam source (Fig. 5) and precision measurements of the hyperfine structures in refractory elements [11-13]. Figure 6 shows the measured hyperfine spectra for Ta isotopes. Spectral lines a to f are from 181TaI; lines a and p are 15 from l80mTal. The natural abundance of I80nrra (T1/2>1.2 xlO y) is 0.012 %. In Fig. 7 the hyperfine transitions a and p are indicated and LRDR peaks for a and P are shown, where the fluorescence intensities were observed as a function of radiofrequency. Thus hyperfine constants A and B for I80mxa (i = 9) have been determin ed with unprecedented precision: A = 402.9466(10) MHz, and B = -1533.056 (15) MHz; and we have \i\ = +4.825(1 1)|1N and Qs = Fig. 5. Schematic of the LRDR setup at RIKEN. -49- r 4 6 8 10 12 16 Relative Frequency (GHz) Fig. 6. Measured hyperfine spectra of the transition ^Fyi- 4Di/2 (\ = 540.3 nm) in ISOmTal and ISlTal. 6300 , (a) :. 180m'J a 6100 -17/2 4Dl/2" 5900 -19/2 >S 5700 T 3336.2 3336.7 3337.2 4500 a (b) -21/2 RF o 4300 a -19/2 3 "F3G- -17/2 E 4100 1 RF. -15/2 3900 • 4318.8 4319.3 4319.8 RF Frequency (MHz) Fig. 7. Hyperfine splittings and laser-rf double resonances in 180mTaI. Laser-induced fluorescences were measured as a function of radiofrequency: (a) is for the hyperfine splitting 21/2—19/2 and the transition a; (b) for 17/2— 15/2 and p . +4.946(20) eb. The nucleus !80mTa [s considered to be deformed, so that the deformation parameter p is estimated to be 0.268(1) and Qo = +6.79(3) eb (I = K = 9 and see Eq. (6)). Since White and his collaborators identified the nuclide 180mTa in 1955 [14^ mjs exotic isotope has attracted much atten tion in astrophysics as well as in nuclear structure physics to date. It is of great importance to provide nuclear structure information in a model-independent way. — so- 3.3 Measurement of the Bohr-Weisskopf effect An effect of the spacial distribution of nuclear magnetization on the magnetic-dipole interaction constant A given by Eq. (5) was first studied theoretically by Bohr and Weisskopf in 1950 [15]. Experimentally, however, systematic precision measurements of this effect have not been conducted yet. The Bohr-Weisskopf effect is often called the hyperfine structure (hfs) anomaly compared to the hfs for the point-like nucleus (Eq. (4) and (5)): A = A0(l+e), (7) where Arj is the magnetic-dipole interaction constant for a point nucleus and e the Bohr-Weisskopf correction. We can rewrite Eq. (7) as A/A0 = <|Xi(r) | ¥e(r) | 2>/ji, | ¥c(0) I 2 = 1 + E. (8) This provides a powerful test of the nuclear wave function; and hopefully, it does a probe to study the nuclear magnetization, namely, the effect of quarks. It should be stated that EAO £ 1 kHz compared to the accuracy by ordinary laser-induced fluorescence measurements =100 kHz. Instead of E, usually we measure the differential Bohr-Weisskopf correction *A2 = £2 - ei for two isotopes FLUORESCENCE RF RF DETECTION Fig. 8. Schematic of the atomic-beam magnetic resonance setup for Bohr-Weisskopf effect measurement at CERN ISOLDE. The magnet C is a precison one with the field accuracy of the order of 106. Triple resonances are observed in the region A, B, and C. — 51- Ai and A2: A2M1 = 0i2fl2)(l+e2)/(m/Ii)(l+ei), (9) = (g2/gi)(l + ^2). Constants A and nuclear gyromagnetic ratios g have to be deter mined with an accuracy of the order of 10"5 or even better. We started an international collaboration to make systematic meaurement of the Bohr-Weisskopf effect at CERN ISOLDE [16]. Figure 8 shows a schematic drawing of the experimental setup. This is a fine example of the combination of modern precision techniques: laser, radiofrequency, and precision magnet. Details are described in Ref. [16]. To start with, we will measure the hfs anomaly for radioactive Cs isotopes provided by ISOLDE. 4. SUMMARY The hyperfine interaction reveals the nuclear properties such as spins I, magnetic dipole moments \i, electric quadrupole moments Q, size and even shape, model-independently. Lasers coupled with rf techniques will open up a "precision frontier" in nuclear physics: Sub-nucleon degrees of freedom, i.e., quarks, will be elucidated through precision measurements of the hfs anomaly. We have achieved an accuracy of the order of 10-5. Lastly, I would like to say, "Let's discover a big surprise in the precision frontier with laser-rf techniques. I agree with Jiirgen Kluge's spirit: "The event rate (of such a big surprise) is considerably higher when compared to the event rate of supernovae in our galaxy" [17]. I would like to thank all my colleagues for their help in our collaborative work. I am very much grateful to Dr. Hediki Iimura at JAERI, Tokai, and Dr. Masanori Wakasugi at RIKEN for valuable discussions. REFERENCES 1) Mayer M.G. and Jensen J.H.D.: "Elementary Theory of Nuclear Shell Structure", John Wiley & Sons, New York (1955). — 52 — Kopfermann H.: "Nuclear Moments", Pure and applied Physics 2, Academic, New York (1958). Otten E.W.: "Treatise on Heavy-Ion Science" Vol. 8 ed. D.A. Blomley, Plenum, New York, 517 (1989). Bohr A. and Mottelson B.R.: "Nuclear Structure" Vols. 1 and 2, W.A. Benjamin, Reading (1969 and 1975). Segre E.: "Nuclei and Particles", Addison-Wesley, New York, 308 (1982). Iimura H., Nakahara Y„ Ichikawa S., Kubota ML, and Horiguchi T.: to be published in Phys. Rev. A. Macfarlane R.M,, Burum D.P., and Shelby R.M.: Phys. Rev. Lett. 49., 636 (1982). Rosner S.D., Holt R.A., and Gaily T.D.: Phys. Rev. Lett. 3J., 785 (1975). Rabi I.I., Zacharias J.R., Millman S., and Kusch P.: Phys. Rev. 5_3_, 318 (1938). Childs W.J.: Phys. Rep. 2U., 113 (1992). Wakasugi M., Jin W.G., Inamura T.T., Murayama T., Wakui T., Kashiwabara T., Katsuragawa H., Ariga T., Ishizuka T., Koizumi M., and Sugai I.: Rev. Sci. Instrum. £4, 3487 (1993). Jin W.G., Wakasugi M., Inamura T.T., Murayama T., Wakui T., Katsuragawa H., Ariga T., Ishizuka T., and Sugai I.: to be published in Phys. Rev. A. Wakasugi M„ Jin W.G., Inamura T.T., Murayama T., Wakui T., Katsuragawa H., Ariga T., Ishizuka T., and Sugai I.: to be published in Phys. Rev. Lett. White F.A., Collins Jr.T.L., and Rourke F.M.: Phys. Rev. 9.6, 566 (1955). Bohr A. and Weisskopf V.F.: Phys. Rev. 7_7, 94 (1950). Duong H.T., Ekstrom C, Gustafsson M., Inamura T.T., Juncar P., Lievens P., Lindgren I., Matsuki S., Murayama T., Neugart R., Nilsson T., Nomura T., Pellarin M., Penselin S., Persson J., Pinard J., Ragnarsson I., Redi 0., Stroke H.H., Vialle J.L., and the ISOLDE collaboration: Nucl. Instrum. Methods A325. 465 (11793). 17) Kluge H.-J.: Hyp. Int. 2±, 287 (1992). - 53 — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- TIME-RESOLVED LASER-INDUCED FLUORESCENCE IN THE NUCLEAR FUEL CYCLE Christophe MOULIN*, Pierre DECAMBOX, Patrick MAUCHIEN, Alain PETIT CEA/DCC/DPE/SPEA/SPS/Analytical Laser Spectroscopy Laboratory. 91191 Gifsur Yvettecedex, FRANCE Time-Resolved Laser-Induced Fluorescence (TRLIF) is a very sensitive and selective method that has been used for actinides and lanthanides analysis in the nuclear fuel cycle. This technique has been used in different fields such as in geology, in the Purex process, in the environment, in the medical and in waste storage assessment. Spectroscopic data, limits of detection and results obtained in previously quoted fields are presented. Keywords : Actinides, Analysis, Fluorescence, Lanthanides, Laser, Time-resolved. 1. INTRODUCTION Fast and sensitive methods are still required in the nuclear fuel cycle for ultratrace determinations of actinides and lanthanides. Time-Resolved Laser-Induced Fluorescence (TRLIF) is a method of choice for such determinations and has been used for now more than 10 years at the CEA. The principle of this technique consists in pulsed laser excitation followed by temporal resolution of the fluorescence signal which leads to the elimination of unwanted short lifetime fluorescence (organic matters, Raman) as seen on figure 1. The main advantage of TRLIF aside rapidity (few minutes analysis) and sensitivity (LoD in the ng/1 range) is its triple selectivity : excitation selectivity by the proper choice of the laser excitation wavelength, emission selectivity since each fluorescent element gives a characteristic fluorescence spectrum and finally time resolution selectivity which characterizes the fluorescent element environment. Among the actinides and lanthanides, the ones that are fluorescent in solution are U(VI), Cm(III), Am(III) and Eu(III), Tb(III), Dy(III), Sm(III), Gd(III), Ce(III), Tm(III). These elements have been studied in different complexing media (nitric, phosphoric, sulphuric, carbonate, micellar) and analysed from the part per billion (10"6 g/1) to the part per trillion (10* 9 g/1) level and lower [1-6]. Fluorescence £ST^W Figure i [Uj: 50 [igtf in presence of organic matters Classical Fluorimetry 1000 (no time resolution) 590 X(nm) — 54- The different fields of interest in the nuclear fuel cycle where TRLIF has been used are in : - Geology, for ultratrace determination of uranium (ng/1 level) in waters from different localities in order to obtain a better understanding of uranium mobilization under natural reducing conditions and so to validate thermodynamic data [5]; - Environment, for uranium determination in water and soil samples in order to have a "cartography" of specific areas as well as speciation with the use of micellar enhanced ultrafiltration [7-9]; - All the different steps of the Purex process, in very complex matrices (Pu, Na, fission products) and for direct speciation in nitric acid [10-12]; - Medical surveillance for direct and fast determination of uranium in human urines. Hence, based on the uranium concentration in urine, metabolic models are established to estimate the body burden of uranium from persons exposed to uranium ambience [13] ; - Waste storage assessment, for the determination of complexing constants at very low level between actinides and organic matters. TRLIF allows to work directly at trace level and has been applied to curium, dysprosium and uranium [14-16]. Moreover, in Purex process control, remote TRLIF measurements are performed in gloves-box or shielded cell with the adaptation of fiber optics and optode [17] and is foreseen for environmental purposes [18]. 2. MATERIAL AND METHODS Apparatus : The experimental set-up is schematically shown in Figure 2. ^ LASER CB^JX 3C ritiEH oi'nr.s I'VSSIU • OI'TNOIIG £« tisa rONTHOL UNIT Figure 2 : TRLIF experimental set-up (dotted lines for remote TRLIF). A laser (N2 (337 nm), excimer (308 nm) or Nd-YAG (355 nm)) coupled or not with a dye laser is used as excitation source. The laser beam is focused into the cell of the spectrofluorometer "FLUO 200!" (DILOR, 244 rue des Bois Blancs, 59000 Lille, France) by a quartz lens. The radiation coming from the cell is focused on the entrance slit of the - 55 — Intensity fA.UJ 600 jA A A " A A y ^ v* A „ mMiL. , '^J^^t^m^ 450 490 530 570 X (ran) Figure 3 : Fluorescence spectrum of uranium at 0.5 ng/I in phosphoric acid (0.75M) TRLIF coupled with Micellar Enhanced UltraFiltration (MEUF) is used for uranium speciation in the environment. Hence, MEUF allows charge selectivity (affinity with the charged micelle surface) as well as size discrimination to a less extent as schematically shown on figure 4. With such features, it is possible to separate the different species present in solution and to perform speciation at very low level. Figure 4 : Schematic of MEUF and speciation of uranium in solution. Dotted line represents experimental data. [U] 1 mg/1. In the Piirex process in the presence of plutonium, americium, neptunium, fission products, sodium, organic effluents (for example in the TEO (Organic Waste Treatment) in order to determine the decontamination factor (DF) by using uranium as tracer in the organic phase (TBP, TPH)), directly in nitric acid to avoid the use of complexing reagents (such as H3PO4) which generate effluent. This last step allows also to perform direct uranium speciation in nitric + acid since the different uranium-nitric acid complexes (U02N03 , U02(N03)2) affect fluorescence spectrum (figure 5). This spectral modification permits spectral deconvolution and the determination of free nitrate in solution. - 56- monochromator (range covered 200 nm). The detection is performed by an intensified photodiodes (1024) array cooled by Peltier effect and positioned at the monochromator exit. Recording of spectra is performed by integration of the pulsed light signal given by the intensifier. The integration time adjustable from 0.3 to 90 s allows for variation in detection sensitivity. Time-resolution is obtained by the control unit that assures pulsed running of the intensifier and the photodiodes array. Measurements are adjustable with a delay from 0.1 to 999 us during a time of 1 to 99 us. All functions of the apparatus (temporal delay, gate width, integration time, reading of spectra, ...) are controlled by a PC-AT microcomputer. Analytical software automatically calculates concentrations by the standard addition or the initial fluorescence methods. 3. PERFORMANCES Table 1 presents convenient excitation and main emission wavelengths, lifetimes and limits of detection (in best complexing media) used in TRLIF for actinides and lanthanides determination in solution. ELEMENT ^excitation (nni) "•fluorescence 01'11) Lifetime ((.is) / Limit of medium detection (ng/l) URANIUM 337, 420 494-516-545-565 200 / H3PO4 0.0001 CURIUM 337, 385 612 200 / TTA 0.0001 AMERICIUM 507 697 0.03 / K2C03 1 EUROPIUM 337, 394 590-617-650-690 900 / TTA 0.001 SAMARIUM 337, 355, 380 565-600-650-710 80/TTA 0.01 TERBIUM 337, 355, 380 490-545-590-625 2100/EDTA 0.1 DYSPROSIUM 355, 390 485-580-670 20/EDTA 0.01 CERIUM 260 360 0.07 / H2S04 0.2 GADOLINIUM 275 313 3000 / H2S04 10 THULIUM 360 455 7 / K2C03 750 Table 1 : Main ^excitation, ^fluorescence lifetimes and limits of detection in TRLIF 4. RESULTS //; the environment, determination of uranium and curium at ultra low level are required for cartography purposes. Figure 3 represents the fluorescence spectrum of uranium at 0.5 ng/1 in phosphoric acid. This value, roughly corresponds to 107 molecules in the interaction volume. — 57- Fluorescence intensity (A.U.) 14000 • . HN03 1M 1000O • ft 6000 - IX M^Mi 2000 - life X. (nm) 1 1 1 1 i^ 450 490 530 570 610 650 Figure 5 : Uranium fluorescence spectra as a function of nitric acid concentration /// waste storage assessment, the knowledge of radioelement behaviour is of great importance. In particular, complexation reactions of radioelements with natural organic ligands such as humic substances (humic/fulvic acids) present in aquifer systems are important, because of the strength of complexes that humic substances form with cations, in particular trivalent elements like americium and curium. TRLIF is the only technique that allow to work directly at trace levels. The complexation of curium by humic substances leads to an increase of the curium fluorescence signal until saturation of the complexing sites occurs. From titration curves (figure 6), binding parameters such as the complexing capacity W and the conditional stability constant (3 (by non linear regression fit) are obtained. TRLIF as shown the influence of cation concentration on the interaction constant (3 as well as the independence of (3 with pH. I'luoa'sccncc Intensity (Cm! *-M-J ?. exi: : !>5~ r;iTi X fill : <)(:0 nit! Cm + l-:A As for speciation of uranium in the Purex process (directly in nitric acid), TRLIF is used for speciation of uranium in condition representative of future waste disposal sites as shown on -58- figure 7 where the fluorescence spectrum at pH 2 is only due to UCh2+ (with the four characteristic peaks but is drastically modified at pH 5 due to the presence of both UC>22+ and U02(OH)\ By spectral deconvolution, it is possible to determine the amount of each species present in solution and by doing so to perform speciation. Moreover, lifetimes of these different species are very different and allow in certain cases time resolution. U: 100 |i(./l dOO -11)0 500 r,?u &40 6GO liUO WO Figure 7 : Fluorescence spectra (normalized) of uranium at pH 2 and 5. I O.IM 5. CONCLUSION Time-Resolved Laser- Induced Fluorescence is a fast, sensitive and selective technique for fluorescent actinides and lanthanides determinations in the nuclear fuel cycle. These features have allowed to use TRLIF in various fields from ultratrace analysis to complexation studies (fluorescence titration, spectral deconvolution). Furthermore, the use of fiber optics and optodes allows remote measurements in hostile or difficult to access environment and is promising for in-line analysis in the reprocessing or for environmental monitoring. REFERENCES1 1) Direct uranium trace analysis in philoniuiu solutions by TRLIF. P. Mauchien, P. Decambox, C. Moulin, T. Berthoud, B. Kirsch. Anal. Chem. 60, 1296 (1988). 2) Direct determination of traces of lanthanides ions in aqueous solution by TRLIF. P. Mauchien, P. Decambox, C. Moulin, T. Berthoud, B. Kirsch. Anal. Chim. Acta. 220, 235 (1989). 3) Time-Resolved Laser-Induced Fluorescence for curium trace determination. C. Moulin, P. Decambox, P. Mauchien. Radiochim. Acta 48, 23 (1989). 4) Determination of curium at 0.1 ng/I levels in a mice ffar medium by TRLIF. C. Moulin, P. Decambox, P. Mauchien. Anal. Chim. Acta 254, 145 (1991). 5) Determination of uranium in solution at the ng/I by TRIJF: application to geological sitivey. C. Moulin, C. Beaucaire, P. Decambox, P. Mauchien. Anal. Chim. Acta. 238,291 (1990). 6) Americinm trace determination in aqueous and solid matrices by TRLIF. P. Thouvenot, S. Hubert, C. Moulin, P. Decambox, P. Mauchien. Radiochim. Acta. 61,15 (1993). -59- 7) TRUF studies of uranium-sodium dotlecyl sulfate. C. Moulin, P. Reiller, C. Beaucaire, D. Lemordant. Applied Spectro. 47, 2172 (1993). 8) Dual use ofinicelkf enJianced ultrafiltration and TRUF for tlie stucfy of uraiyl exclxoige at tlie suiface of alkybulfatemicelles. P. Reiller, C. Moulin, C. Beaucaire, D. Lemordant. J. Colloid. Interface. Sci. 163,81 (1994). 9) On the use of TRUF for uranium micelle interaction studies. C. Moulin, P. Reiller, C. Beaucaire, D. Lemordant. J. Colloid Interface Sci. 157, 411 (1993). 10) IRUF of UO-?* in nitric acid solutions. Preliminaiy results for on-line UO-?+ monitoring applications. H. Deniau, P. Decambox, P. Mauchien, C. Moulin. Radiochim. Acta 61,23 (1993). 11) IRIJFof UQ-r* in nitric acid solutions. Comparison between nitrogen and tripledNd-YAG laser. C. Moulin, P. Decambox, L. Couston, D. Pouyat. J. Nuclear Sci and Techno. 7,31 (1994). 12) S/xciation of uranium in nitric acid by TRUF applied to analytical determinations. L. Couston, D. Pouyat. C. Moulin, P. Decambox. Applied Spectroscopy. Submitted (1994). 13) Direct and fast determination of uranium in human urines by TRUF. C. Moulin, P. Decambox, P. Mauchien. Applied Spectro. 45, 116 (1991) 14) On the use of1IUJFfor interaction studies Ix'lween organicmatters cuid'actinides: apjrfication to Cm. C. Moulin, P. Decambox, P. Mauchien, V. Moulin, M. Theyssier. Radiochim. Acta. 52/53,119 (1991). 15) Complexation behaviour of Iminic substances toward'actinides studied by TRUF. V. Moulin, J. Tits, C. Moulin, P. Decambox, P. Mauchien, 0. de Ruty. Radiochim. Acta 58/59, 121 (1992). 16) Coinplexation beha\'ionr of uranium M'ith huniic substances studied by TRUF V. Moulin, C. Moulin, P. Decambox, I. Laszak. J. Environmental Sci. Techno. To be published (1994). 17) Uranium determination by Remote Time-Resolved Laser-Induced Fluorescence. C. Moulin, S. Rougeault, D. Hamon, P. Mauchien. Applied Spectroscopy 47, 2007 (1993). 18) Uranium speciation in solutions by TRUF. C. Moulin, P. Decambox, V. Moulin, I. Laszak. J. Environmental Sci. Techno. To be published (1994). 1 For simplification purposes, only references from the CEA are quoted in this paper - 60 — Proceedings of the 6th International Symposium on Advanced Nuclear Enci.ty Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Photoacoustlc and Photothermal spectroscopies Tsuguo SAWADA, Takehiko KITAMORI and Masato NAKAMURA Department of industrial Chemistry Faculty of Engineering The University of Tokyo 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan Tel.03-3812-2111, Fax.03-3815-6543 Photoacoustic and photothermal spectroscopy methods can be effectively applied to the analysis of microparticles in condensed matter. A more violent photo thermal conversion phenomenon of a particle, laser breakdown and accompanying plasma and acoustic emission, was applied to individual detection and analysis of ultrafine particles in ultrapure water. Laser-like nonlinear emission from the plasma was observed. Key words : Photoacoustic spectroscopy, Photothermal spectroscopy, Trace analysis, Nonlinear emission 1. Introduction Nowadays, the principle idea of the photoacoustic effect has been enlarged to that of a "photothermal" effect in which the basic physical phenomenon providing spectroscopical information is an energy conversion from photo-energy to thermal energy. Some methods to detect and analyze the photothermal effect have been proposed, such as thermal lens, optical beam deflection and laser induced grating methods. In addition to the methodological studies in this early period including theory and instrumentation, various applications of photothermal. spectroscopy have been developed over a wide range of science and technology, including physics, chemistry, biology, medicine, agriculture, electronics, semiconductor engineering, nuclear engineering and others. Many original articles and reviews have been pub lished for the respective fields, since it has now become too popularized to prepare an overall review. Hence, this paper focuses on recent progress in spectroanalytical application of photothermal spectroscopy to mainly condensed matter. -61- 2. Photothermal effect in the condensed phases Phsnomenological representations of typical photothermal phenomena in solid and liquid samples are illustrated in Fig. 1. Other, new photothermal phenomena are described in later sections. As shown in Fig. 1, intermittent optical excitation and relaxation of a sample generates a periodical variation of the temperature field, result ing in changes of pressure, stress, mass density, refractive index, and other mechan ical, thermal and optical parameters. Photoacoustic spectroscopy utilizes the acous tic, elastic and thermal waves, and detects them with a microphone and piezoelectric transducers as shown in the drawing. In photothermal deflection spectroscopy, a probe laser beam, which passes horizontally through the neighborhood of the excita tion beam irradiated surface, is deflected clue to diffraction or refraction by a refrac tive index wave, and the signal is measured from its deflection angle. Any surface deformation aiso deflects the reflected probe beam on the irradiated spot. Hereafter, these photothermal effects are briefly summarized with a general description to clarify the spectroanalytical, spectrochemical and physical meanings of the signals. Signal generation is phenomenologically classified into two processes as show in Fig. 2. The first process is a photothermal energy conversion and migra tion from optical energy to thermal energy. This energy conversion is achieved mainly through optical excitation and nonradiative relaxation. Other energy conver sion processes, such as a photochemical reaction, photosynthesis, and phase transi tion are also able to provide thermal energy as a source term of the signal. The generated thermal energy macroscopically migrates in the medium by thermal diffu sion, leading to formation of a heat distribution. This process, in which the optical energy is converted into thermal energy and the heat distribution is formed, is re ferred to as the photothermal process herein. When the excitation beam irradiation is intermittent, the heat distribution fluc tuates periodically. This, in turn, causes a fluctuation of the temperature field in the medium which then causes various mechanical and optical parameters to fluctuate. The second process is referred to as the thermodynamic process. These fluctuations of mechanical and optical parameters are detected by the individual photothermal spectroscopical method as mentioned in Fig. 1. - 62 — 3. Ultrafine Particle Analysis by Laser Breakdown Effect For an example of a more violent photothermal conversion process, laser breakdown and accompanying acoustic emission have been applied to individual detection and counting of ultrafine particles in liquids.'" An ultrafine particle smaller than 0.1 nm cannot be detected by the conventional laser scattering method, which is one of the most sensitive methods, because of the background due to medium Ray- leigh scattering, In place of the laser scattering method, a laser breakdown acoustic method was proposed, in which plasma formation from the ultrafine particle in liquids is induced by irradiation with a focused pulsed laser beam and the ultrafine particles are counted individually by detecting an acoustic pulse. Figure 3 shows a particle in a focused excitation beam. When the power density of the optical radiation exceeds the threshold of particle breakdown, the parti cle becomes a plasma and a strong acoustic emission due to explosive expansion and strong absorption of the plasma is induced. This plasma formation and acoustic emission process can be considered as one of the most violent photothermal energy conversion processes, because the plasma temperature is estimated to be at least 104K (several eV).3 The particle breakdown threshold is lower than those of liquid media and air bubbles, so only the particle can be broken down, and miscounting due to air bubbles and liquid breakdown can be avoided.' Using this method, it was demonstrated that 38 nm polystyrene ultrafine particles, which are one of the small est standard particles on the market, were counted individually at the number density level of ultrapure water, 102 particles per 1 ml as shown in Fig. 4. The minimum de tectable particle size, expected to be smaller than 10 nm, was at least one to two orders smaller than for the conventional laser scattering method. The acoustic pulse height showed a tendency to be proportional to the particle size,2 and atomic and ionic emission lines of particle component elements could be observed in the plasma emission.3 Therefore, laser breakdown acoustic spectrometry is expected to be a novel analytical method for ultrafine particles in liquids. Furthermore, we have found a non-linear optical phenomenon in the plasma emission. A line-like atomic emission line for hydrogen at 656.3 nm, which was measured in the forward direction plasma emission induced from a 0.3 urn polystyr ene particle, is shown in Fig. 5. The mechanism of this laser-like emission is not clear at the present time; however, it is considered to be a secondary laser emission from the underwater plasma due to non-equilibrium population kinetics of atomic levels in the plasma. This effect is expected to be a basis for another novel spectra- metric method for microparticles."' - 63- 4. References 1. Kitamori, T., Yokose, K., Suzuki, K., Sawada, T. & Gohshi, Y., Jpn. J. Appl. Phys., 27 (1988) L983. 2. Wu, J., Kitamori, T. & Sawada, T., J. Appl. Phys., 69(1991)7015. 3. Kitamori, T., Matsui, T., Sakagami, M. & Sawada, T., Chem. Lett., (1989) 2205. 4. Nakamura, M., Kitamori, T., & Sawada, T., Nature, 366, 138(1993) — 64 — a) Excitation Reflection Beam Beam Probe Seam / /—Microohone Deflection Beam // s Acoustic Wave **/ (R«fr»ctiv« Indix Wav«) -Deformation Probe Beam Thermal Diffusion <^^^, Region ^TTL, Thermoelastic Wave Solid Sample Piezoelectric Transducer b) Piezoelectric Transducer 4= Excitation Beam Ootical Window 7 Thermal Lens Effect Liauid Samole Fig. 1. Photothermal effects on and in condensed phase samples, (a) Solid sample, and (b) liquid sample [1]. Photolheimal process Thermodynamic process Honradiative Mass density lluctuation [Halation Pressur fluctuation Excitation Signal beam Photochemical Thermal diffusion reaction Refractive index fluctuation Phase transition Tension fluctuation i Fig. 2. Phenomenological representation of the photothermal signal generation process. Plosmn Breakdown Convex Lens /Acoustic Pulse Particle Beam Waist Excitation Beam Plasma Emission Fig. 3. Plasma formation by laser breakdown of an ulirafinc parlicle. -65- 1900 1800 1700- 1600- c o O 1500- 1400, 1300 300 600 900 Number Density of Added Particles (ml"') Fig. 4. Dependence of the breakdown acoustic pulse counts on the number density of the added 0.038 /tm polystyrene ultrafine particles. 650.0 nm J \ 656.3 nm 5- 1 ^ • 645 650 655 660 Wavelength (nm) Fig. 5. Forward direction plnstnn emission spcclnim of sin imdcrwutci plasma generated from a 0-3/mi polystyrene microsphere. — 66 — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- LASER INDUCED FLUORESCENCE APPLED TO STUDIES OF PARTICLE BEHAVIOUR IN HIGH-TEMPERATURE PLASMAS K. Muraoka, K. Uchino, T. Kajiwara, M. Maeda* and T. Okada* Department of Energy Conversion, Kyushu University, Kasuga Fukuoka 816, Japan Tel. 092-573-9611, Fax. 092-571-8013 *Department of Electrical Engineering, Kyushu University, Hakozaki Fukuoka 812, Japan Tel. 092-641-1101, Fax. 092-631-2790 Laser induced fluorescence (LIF) is the technique in which the laser frequency v is tuned to match a transition of a group of atoms or molecules, whose energy difference is £12, by a relation hv=E\2 (where h is the Planck's constant), and the resultant fluorescence is observed. By the resonant nature of the excitation, the fluorescence intensity is many orders of magnitude larger than other sources of scattering, such as Rayleigh, Raman, or Thomson scattering, if the relevant number densities are the same order of magnitude. Therefore, the LIF technique is very selective of species to be detected, sensitive and/or yields high spatial and temporal resolu tion. The availability of the tunable source (A>120 nm decided by the transmission of a LiF window) to match hv=E\i, namely A[nm]=1239.8/£i2[eV], dictates the usable ranges of the technique to £12^1 OeV. Therefore, the technique is useful for excitation from the ground levels of neutral or lowly ionized particles, otherwise excitation from the excited levels is neces sary. There are many kind of species at different ionization stages present in edge plasmas. These species are due to release from walls, supplies by gas puff and/or pellet injection, or re combination/charge exchange processes. In order to understand the particle behaviour and ul timately to control the plasma density profile and/or to reduce impurities, the density and veloc ity-distributions of these atoms, molecules and ions have to be measured. For this purpose, LIF should yield useful information. In addition, because electric and magnetic fields affect E\2, the values of these field strength should, in principle, be obtainable using LIF. In this presentation, we first review the principle of LIF, then give an overview of tunable laser sources, the crucial hardware for the experiment, and describe methods of calibration to obtain necessary information from the observed fluorescence, followed by the plasma mea surements which have already been conducted. Comments are made for the future perspective of LIF for high-temperature plasma diagnostics. Keywords: Laser Induced Fluorescence; Plasma Diagnostics. - 67- LASER INDUCED FLUORESCENCE APPLED TO STUDIES OF PARTICLE BEHAVIOUR IN HIGH-TEMPERATURE PLASMAS K. Muraoka, K. Uchino, T. Kajiwara, M. Maeda* and T. Okada* Department of Energy Conversion, Kyushu University, Kasuga Fukuoka 816, Japan *Department of Electrical Engineering, Kyushu University, Hakozaki Fukuoka 812, Japan J. Introduction Laser induced fluorescence (LIF) is the technique in which the laser frequency v is tuned to match a transition of a group of atoms or molecules, whose energy difference is £12, by a relation h\=E\2 (where h is the Planck's constant), and the resultant fluorescence is observed. By the resonant nature of the excitation, the fluorescence intensity is many orders of magnitude larger than other sources of scattering, such as Rayleigh, Raman, or Thomson scattering, if the relevant number densities are the same order of magnitude. Therefore, the LIF technique is very selective of species to be detected, sensitive and/or yields high spatial and temporal resolu tion. The availability of the tunable source (A>120 nm decided by the transmission of a LiF window) to match hv=E\2, namely A[nm]=1239.8/£i2[eV], dictates the usable ranges of the technique to £i2<10eV. Therefore, the technique is useful for excitation from the ground levels of neutral or lowly ionized particles, otherwise excitation from the excited levels is neces sary. There are many kind of species at different ionization stages present in edge plasmas. These species are due to release from walls, supplies by gas puff and/or pellet injection, or re combination/charge exchange processes. In order to understand the particle behaviour and ul timately to control the plasma density profile and/or to reduce impurities, the density and veloc ity-distributions of these atoms, molecules and ions have to be measured. For this purpose, LIF should yield useful information. In addition, because electric and magnetic fields affect £12, the values of these field strength should, in principle, be obtainable using LIF. In Section 2, we first review the principle of LIF and give an over-view of tunable laser sources, followed by the plasma measurements which have already been conducted in Section 3. Here, emphasis is given to studies of atomic hydrogen behaviour in high-temperature plas mas. Summary and comments are made for the future perspective of LIF for high-temperature plasma diagnostics, in Section 4. - 68 — 2. Principle ofLIF and tunable sources There are excellent textbooks on LIF (e.g., Demtroder, 1982). A good review was written (Bogen and Hintz, 1986) on the topic up to 1986, and recently a review was given on the topic for more recent work (K. Muraoka and Maeda, 1993), and interested readers may re fer to these publications. The first requirement for a LIF experiment is to match the laser wavelength to transitions of the atoms, molecules and ions to be detected. Usually, this requirement sets the limit of applicability of LIF to detection of various species. The subsequent optical, spectroscopic and detection system is common to usual optical emission or scattering experiments of plasmas. The relationship between the laser wavelength X [nm] and energy difference among the transition of species to be detected, En [eV], is given by X=1239.8 / £12. (0 Because atoms and ions in high temperature plasmas (and molecules in low temperature plas mas) are predominantly at the ground level, excitation from there is necessary. En for alkali atoms is around 2 eV, for most metal atoms around 5 eV, for light atoms (H,0, C, and others) around 10 eV and for ions substantially higher than the respective atoms. The corresponding excitations wavelengths are, from Eq. (1), in the visible, ultraviolet (UV) and vacuum ultravio let (VUV) regions, respectively. The tunable sources developed most extensively for the above purpose are dye lasers in these wavelength region. By the proper choice of organic dyes, spectral regions of visible, near infrared and UV wavelengths can be covered. A wide tunable range can be covered by the pumping of a high-power, short-pulse laser such as excimer, Nd:YAG (SHG,THG) and N2 lasers. The pulse dura- tions of these lasers are around 10 ns with a repeti ' 'mm. tion rate of 10-100 Hz. U(M{{((( IS 100 One example is shown in Fig. I (Lambda Physik GmbH, 1988), where dye laser output against wave 700 800 900 1 000 WAVELENGTH (nm) length is shown for com mercially available excimer- Fig. 1 Tunable ranges of commercial dye lasers pumped by laser pumped dye lasers. KrF or XeCl excimer lasers with a pumping energy of 150-300 Flashlamp-pumped mJ/pulse. — 69- dye lasers can operate over longer pulse widths , typically around 1 us, in the visible region. CW operation is possible by the pumping of an argon ion laser. The coherence of them is ex cellent, although the wavelength access over a wide spectral region is poor. In order to cover a wider spectral High-power H>~(a) rare-gas/halide (mainly shorter wavelength) region than excimer laser dye laser obtainable by dye lasers, as required for SHG in crystal detection of metal atoms and light atoms in THG in gas - (d) plasmas, various nonlinear frequency conver sion techniques have been developed. Figure Raman laser (H2, D2, CH4 etc.) 2 illustrates such a scheme, where tunable II i i i i i (a) sources based on an excimer laser are shown. m.iHMHhu^«Mum»a,^M (b) The visible region is covered by the excimer mwwwwtjww (c) laser pumped dye lasers, as shown above in "MMraKMir" (d) (e) Fig. 1, while the near UV spectral region can JL _l_ _1_ 100 200 300 400 500 600 be covered by second harmonic generation wavelength (nm) (SHG) in nonlinear optical crystals, such as Fig. 2 Wide range tunable sources based on a KDP, ADA, and BBO. In particular, high power excimer laser. Raman spectra (e) efficient UV generation down to 205 nm has are quasi-continuous. become very easy by the development of the new crystal BBO. Excimer lasers themselves have tunable range of around 1 nm centred on respective lasing wavelengths, such as 308 nm for the XeCl laser, 248 nm for the KrF laser and 193 nm for the ArF laser. In order to cover the VUV region, third harmonic generation (THG) and four-wave mixing techniques in gaseous media have been developed. Although the conversion efficiency from dye lasers by THG and four-wave mixing is extremely low at around 10"5~10"^, tunable VUV sources usable for LIF have become available such as for detection of C (165 nm) and H (121 nm). Also, Raman shifting of excimer laser outputs has proved useful for certain applications. These high power tunable sources are also useful for experiments of multi-photon excitation. Interested readers of the of the frequency conversion and Raman shifting may refer to a reference (Kaiser and Maier, 1972) for further details. Lately, new solid-state tunable sources such as Ti:sapphire laser and optical parametric oscillator (OPO) are rapidly being developed. Because these sources have wider tunable range, higher efficiency, and better reliability than those of dye lasers, they are most promising tunable sources in the next generation. 3 The laser power for saturation is given by / (vo)> {gi/(gi+g2)}*(8nhvo/c2)/g(vo) and is lower than 1 W for a laser beam diameter of a few mm and and for a single mode laser, in the visible and near UV wavelength region. In actual plasma diagnostics, however, a power level of more than 100 W is sometimes required to attain sufficient saturation, because of the broad- - 70- ening of the laser spectrum and the level quenching effects. It is easy to obtain the saturation condition for the visible or UV regions using these tunable sources, while it is usually difficult for the VUV region because of the small available laser power. Coarse tuning of tunable lasers is made using gratings or birefringent filters. For mea surements of spectral line profiles due to Doppler, Zeeman and Stark effects, a laser spectral line width of around 1 pm is sometimes required. For such a fine tuning, insertion of an intra- cavity etalon is used. In this respect, the RAPS (rapid frequency scan) laser (Honda et al., 1987) is noteworthy, where the spacing of intra-cavity elation is piezoelectrically driven to yield a spectral scan in a few us. This is useful for observation of a spectral line shape in a transient condition. 3. Plasma measurements Applications of LIF to measurements in plasmas are classified by the wavelength of the used tunable lasers as visible £ 400 nm), UV (400-200 nm), and VUV (< 200 nm). This is because there is the distinct difference in the excitation source for each region. In addition, the optical and detection components are mostly decided by the excitation wavelength, and the fac tors to be considered are accordingly different. 3.1 LIF using visible lasers Tunable sources in this wavelength region are readily available commercially as shown in Section 2. However, from Eq. (1), visible lasers correspond to E\i < 3 eV, and only atoms and ions of alkali metals and rare earth metals have transitions in this range of E12 from the ground levels. Because of this, basic plasma physics researches have been carried out in vapours and plasmas of these atoms (Bowles et al., 1992). Otherwise, such atoms have to be injected into high-temperature plasmas. A scheme to measure a magnetic field in a tokamak using an injected Li beam combined with LIF was carried out (West et al., 1987). Another possibility is to use a transition from an excited level to higher levels. The most notable among various such researches is the atomic hydrogen measurement in high-tempera ture plasmas, which will be described in the following. The energy level diagram of atomic hydrogen is shown in Fig. 3. First, LIF among ex cited levels, usually the Balmer series, has been performed. The measurements yield, after cali bration of the optical system, a population at an excited level. nH is calculated by a collisional- radiative model, which relates the ground and excited level densities using independent mea surements of election temperature and density. -71- The first successful measurements by 4 the scheme was performed on FT-1 Tokamak 3 A A „ at Ioffe Institute (Burakov et al., 1977). In v "a this experiment, various novel ideas were in A troduced, such as the two detection optics, one looking at the fluorescence and the other looking at the nearby plasma emission to be L a A (2x^Lp) subtracted from the former, and the introduc tion of a long (-20 us) flashlamp-pumped dye laser to have a long integration time to reduce fast fluctuations of background plasma radiations. The detection limit of n„ thus ob 1 tained was around 1*1014 m3, and radial Fig. 3 Low-lying energy levels of a hydrogen profiles of nH were obtained for various atom and wavelengths of some transitions. discharge conditions. Subsequently, the technique was applied to a mirror machine to study radial loss of hy drogen by charge exchange (Muraoka et al., 1985), and to a stellarator, and then to Heliotron E (Muraoka et al., 1990) and CHS (Compact Helical System) (Uchino et al., 1992) to study the detailed particle behaviour there. The last experiment is described in some length. The experimental arrangement for LIF on CHS is shown in Fig. 4. A flashlanp-pumped dye laser with an oscillator and an amplifier was used, which yielded 100 mJ at the Balmer-al- pha wavelength (Ha> 656.3 nm) with a pulse duration of 1 us TJT Optical Fibres FWHM (output power of 100 ^ ,JtLens kW). The spectral width of the laser light was adjusted by a Fabry-Perot etalon with an air-gap of 10 urn, and was -0.8 nm Laser Beam which corresponds to the Doppler broadening of hydrogen atoms at a temperature of 300 eV. The size of the observation volume Fig. 4 Experimental arrangement for measurements of was 30 mm in depth, 4 mm in atomic hydrogen densities on the CHS device. width and 50 mm in height. The detection solid angle was 2* 10"2 sr. The optical system was calibrated using Rayleigh scattering of the dye laser from nitrogen gas which filled the chamber. Saturation of the fluorescence signal from the plasma as a func- - 72- tion of laser power was checked, and the laser power density 17 mentioned above corresponded to 10 -r-1—i—i—J—i—i—i—i—i—i—i—i—i—f the saturation parameter being 16 5=4-6. The calibration and the 10 CO saturation assured an absolute I 15 • 6 10 determination of A«3 (=rt3s-"3, x a- where ri3S is the population 14 density at the level n=3 after the 10 saturated excitation by the laser 13 • and 113 is that without laser irradia 10 tion). The absolute value of A/J3, 900 1000 1100 1200 1300 1400 together with known values of R[mm] electron density nc and tempera 17 i i i—i | i—i—i—i—|—i-1—r ture re, yielded the population 10 • i i '-' i I •' • density at the ground level, no, (b) using a collisional-radiative (C-R) 10 16 model. Figures 5 (a) and (b) show "LCFS 15 rtH profiles measured by LIF K ti 10 tuned to Ha, at a cross section where the plasma is horizontally Local 14 elongated, for different chords as 10 '••••'• ' • -J 1 1 I 1 L. shown in the insets. Also shown 900 1000 1100 1200 1300 1400 are calculated profiles from the R [mm] DEGAS code (Heifetz, 1986). The calculations were performed Fig 5 „H profiles measured by Ha LIF for (a) a central for the case? of a localized wall chord and (a) a lower chord] as shown in the insets. source (local.) at the divertor Also shown ^e calculated profiles from the the DEGAS traces on the wall and of a simulation code for cases of a localized wall source uniformly distributed wall source (]oca]<) at divertor ^^ on the wan and 0f a uniformly (unif.), with the atomic energy set distributed wa]1 source (unif.). as eV (Muraoka et al., 1990). Because the calculated profiles of nH were in relative units, they were adjusted to best fit the experimental data. The fittings for Figs. 5 (a) and (b) are not independent but related, and we have to adjust the two profiles of the simulation simultaneously to the measured data. As can be seen from the figures, the fitting to the experimental points is better for the case of the - 73- uniform wall source. Based on the results, shown such as in Fig. 5 for different conditions, assisted by other diagnostic system consisting of arrayed detectors of Ha and Hp emissions and a CCD camera, detailed particle behaviour was discussed. 3.2 LIF using UV lasers Tunable sources in this wavelength region are also readily available, where dye laser out puts are frequency up-converted (wavelength down-converted) using SHG, as shown in Sec tion 2 with a typical conversion efficiency of 10%. Usually, the available laser power is above the saturation of transitions shown in Section 2 and the LIF experiments can be performed al most as easily as in the visible regions. A>200 nm imposes, from Eq. (1), £12 < 6 eV, which means mat transition from the ground levels of metal atoms are in this wavelength region. The first attempt of LIF for high temperature plasma studies was indeed intended for metallic impurity detection in the TEXTOR toicamak, when the machine was envisaged as the one to study plasma-surface interactions, in particular the reduction of impurities by various means, such as surface modifications and a pumped limiter. Subsequently, similar experi ments were performed on the ASDEX tokamak (Schweer et al., 1982), and other machines. Interested readers may refer to these publications. 3.3 LIF using VUV lasers and multi-photon excitation As tunable sources in this wavelength region are not commercially available, we have to develop the sources themselves for various measurements. However, by various novel ideas, notably a recent development of a multi-photon excitation scheme, detection of light elements, such as H, C and O and various ions, has become feasible. The most important information to be obtained in this wavelength region is atomic hydro gen behaviour. The energy level diagram of atomic hydrogen has already been shown in Fig. 3. Because a tunable source for the Lyman alpha transition (121.6 nm) is available by THG of a dye layer output (364.8 nm) in Kr/Ar gas mixture, the first measurements were performed using the two-level system for the transition on TEXTOR (Bogen and Mertens, 1989; Mertens and Bogen, 1989). Although this scheme is straightforward and it is easy to convert the fluorescence intensity to atomic hydrogen density through calibration using the Rayleigh scattering, it is usually diffi cult to have a special optical system to reduce stray light necessary for the two-level system on large fusion machines. Therefore, a different scheme has recently devised. As shown in Fig. 3, the next longest wavelength in the Lyman series after the Lyman alpha is the Lyman beta - 74- (102.6 nm), which is beyond the cutoff of a LiF window. However, two-photon excitation is feasible, because a very large laser power is available at 205.1 nm, which is twice the Lyman beta wavelength. The emission can be generated by Raman shifting of an ArF laser emission at 193 nm with D2 gas, and the wavelength can be tuned over 1 nm at 205 nm. Figure 6 shows the experimental arrangement of the system on Heliotron E (Kajiwara e.t Tuning Unit >® w Lambda Physik EMG150MSC ^PMT L3 Heliotron E Observation Point PH V <|>0.4mm F: Filter, L1-L5: Lens, M1-M2: Mirror, PH: Pinhole, PMT: Photomultipliertube, Fig. 6 Experimental arrangement of two-photon excited LIF for atomic hydrogen detection on Heliotron E. al.y 1991). The laser characteristics were 10 mJ of output energy in an 8 ns FWHM pulse with a spectral width of about 30 pm. The spectral width was optimized for the density measure ments of atomic hydrogen at a temperature of several eV. The laser beam was focused at the observation point by a lens having a focal length off=S10 mm. Fluorescence signals were ob served by a photomultiplier through a pinhole having a diameter of 0.4 mm and a Ha filter. Even with this coaxial arrangement, the spatial resolution was found to be about 20 mm, and the system was calibrated through the comparison between perpendicular and coaxial observa tions to the laser beam. The measurements were performed at 30 mm outside the separatrix. Figure 7 shows typical traces of fluorescence signals for the two timings. At r=4 ms after the ECH initiation [Fig. 7 (a)], the hydrogen density at the observation point was high enough for the fluorescence signal to be recognized above the background Ha fluctuation. However, at r=12 ms, the fluorescence signal was not large enough to be clearly observed in a signal obser vation. Therefore, data from seven shots was accumulated and the result is shown in Fig. 7 (b). These results were put into the absolute calibration to yield atomic hydrogen densities of 3>=1016 nr3 for the former and 1 * 1016 nr3 for the latter. - 75- Based on the above proof-of-principle S FL Data (50060) 20 experiment, a plan is being drawn to study atomic hydrogen behaviour in the scrape-off J1 U- C\t KAJ V - Is* layers, especially in the region around divertors, uf4 h — -20 Vi J because their proper functioning is essential for \ c/> -40 * 1/ realization of thermonuclear fusion reactor. ,' - -luorescence J Sicina l -60 1 1 50 100 150 200 250 4. FURTHER STUDIES Time (ns) (a) SFL Data (50087-93) The application of LIF to plasma di 20 agnostics is in its juvenile stage, in that the study has been seriously pursued for less than a vAM/^-V > J\/ v -P ^ V f^ decade, and that it has not yet become indispens £ -20 f \[ J V able for plasma diagnostics. It has, however, / w -40 started to show potential for future promise. — / -60, The last comment is particularly true for studies 50 100 150 200 250 of hydrogen behaviour in the scrape-off layer of Time (ns) (b) high temperature plasmas. This is because these studies can not be pursued with other diagnostic Fig 7 Typical traces of fluorescence sig- means. nals (a) was obtained at /=4 ms after the For these measurements to become routine initiation of the ECH plasma, (b) was for plasma studies, maintenance-free operation of o^ned at t=\2 ms, and the integrated data the tunable sources is highly desired. In this 0f seven observations, respect, recent developments of tunable solid- state sources, such as alexandrite and Ti:sapphire lasers and OPO are noteworthy. High- power semiconductor lasers and new solid-state lasers pumped by them will also change basic structure of the laser system in the near future. Combining new developments of tunable lasers with measurements of physical quantities of plasma state which are not possible by other means, LIF will evolve into indispensable diag nostic technique for plasma studies during the next decade. The authors are indebted to Drs. P. Bogen, B. Schweer, Ph. Mertens, N. Rynn, and C. Honda for useful comments to the manuscript. They also wish to thank Dr M Bowden for cor recting the English of the manuscript. - 76- REFERENCES Bogen, P. and Hintz, E. (1986) in Physics of Plasma Wall Interactions in Controlled Fusion (Eds. Post, D. E. and Behrisch, R., Plenum, New York) p. 211. Bogen, P., and Mertens, Ph. (1989) Proc. 4th Int. Synip. Laser-Aided Plasma Diagnostics (Kyushu Univ.) p. 117. Bowles, J., McWilliams, R. and Rynn, N. (1992) Phys. Rev. Lett. 68, 1144. Burakov, V. S. et al. (1977) JETP Lett 26, 403. Demtroder, W. (1982) Laser Spectroscopy-Basic Concepts and Instrumentation (Springer, Berlin). Heifetz, D. B. (1986) in Physics of Plasma-Wall Interactions in Controlled Fusion (Eds. Post, D. E. and Behrisch, R., Plenum, New York) 17, 86. Honda, C, Maeda, M., Muraoka, M., and Akazaki, M.(1987) Rev. Sci. lustrum. 58, 759. Kaiser, W. and Maier, M. (1972) in Laser Handbook (Eds. Arecchi, F. T. and Schulz-Bubois, E. O., North-Holland, Amsterdam) 2, 1077. Kajiwara, T., Shinkawa, T., Uchino, K., Masuda, M., Muraoka, K., Okada, T., Maeda, M., Sudo, S., and Obiki, T. (1991) Rev. Sci. lustrum. 62, 2345. Lambda Physik GmbH (from a catalogue of 1988). Muraoka, K., Uchino, K., Itsumi, Y., Hamamoto, M., Maeda, M., Akazaki, M. et al. (1985) Jpn, J. Appl. Phys. 24, L59. Muraoka, K., Uchino, K., Maeda, M., Kajiwara, T., Matsuo, K., Okada, T., Honda, C, Suehiro, Y., Yano, N., Takeda, K., Hagiwara, H., Akazaki, M. et al. (1990) J. Nucl. Mater. 176&177, 231. Muraoka, K. and Maeda, M (1993) Plasma Phys. Contr. Fusion 35, 633. Nishimura, K., et al. (1990) Fusion Technol. 17, 86. Schweer, B., Bogen, P., Hintz, E., Rusbuldt, D., Goto, S. and Steuer, K. H. (1982) J. Nucl. Mater. 111&112, 71. Uchino, K., Takenaga, H., Kajiwara, T., Okada, T., Muraoka, K., Maeda, M. et al. (1992) J. Nucl. Mater. 196-198, 210. West, W. P., Thomas, D. M., deGrassie, J. S., and Zheng, S. B. (1987) Phys. Rev. Lett. 58, 2758. _ 77- Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Physics of Laser Implosion Kazuo A. Tanaka Institute of Laser Engineering, Osaka University Department of Electromagnetic Energy Engineering, Osaka University Suita, Osaka 565 Japan Tel:06-879-7232, Fax:06-877-4799 Shown are current status of physics related to laser inertial confinement fusion at the Institute of Laser Engineering, Osaka University. Plasma corona, energy transport, plasma fluid instability, and implosion physics are shown based on the recent experimental data. Two plasmon decay instability and stimulated Raman scattering are shown to be important nonlinear instabilities. Thermal electron energy transport is shown to have two different dependencies on the separation distance from the critical to ablation front. The fluid instability is measured and is consistent with the particle-in-cell simulation results. The implosion experiments show increased neutron production with improved laser uniformity. Key words: Laser Fusion, Cryogenic Target Implosion, Nd:glass Laser - 78- Physics of Laser Implosion Kazuo A. Tanaka Institute of Laser Engineering, Osaka University Department of Electromagnetic Energy Engineering, Osaka University Suita, Osaka 565 Japan Shown are current status of physics related to laser inertial confinement fusion at the Institute of Laser Engineering, Osaka University. Plasma corona, energy transport, plasma fluid instability, and implosion physics are shown based on the recent experimental data. Two plasmon decay instability and stimulated Raman scattering are shown to be important nonlinear instabilities. Thermal electron energy transport is shown to have two different dependencies on the separation distance from the critical to ablation front. The fluid instability is measured and is consistent with the particle-in-cell simulation results. The implosion experiments show increased neutron production with improved laser uniformity. 1. Introduction A key for laser inertial confinement fusion (ICF) is to keep spherically symmetric implosions to the final stage. In order to achieve this there are several different zones of plasmas at the target to be studied in detail. When the laser light is incident on the target, the intense laser light is absorbed at the critical density or may interact with plasmas nonlinearly, resulting in scattering, hot electrons, filamentations. One of the most typical nonlinear instability is stimulated Raman scattering, which could possibly scatter the incoming laser light or excite electron plasma waves accelerate thermal electrons to hot electrons. These hot electrons may preheat the fuel, affecting the implosion efficiency. Once the laser energy is converted to plasma energy via. the absorption, the energy should be transported to higher density of plasma in the shell, producing a large ablation pressure(<100 Mbar) to accelerate the shell. To know the lateral energy transport is essential for the shell to have uniform implosion speed. Effective lateral energy transport could smooth out the ablation pressure deviation created by the laser nonuniformity. The another important issue is the study of hydrodynamic instability. When the light plasma pushes the heavy shell, this state is hydrodynamically unstable during the shell acceleration. When the imploding shell starts stagnating the fuel and creating very high density regions(<1000 times the solid density), this state is also hydrodynamically unstable. 2. Experiments 2-1. Experimental Conditions The experiments were conducted with GEKKO XII glass laser system at X = 527 nm with random phase plates. Normal coherent laser was used for the most of experiments, while the amplified spontaneous emission (ASE) light was used for the fluid instability experiments[1]. Intensities were varied from 1013 W/cm.2 to 3x1014 W/cm2 with f/3 focusing lens. Targets were planer for fundamental studies such as corona instabilities, energy transport, and hydrodynamic instabilities. Implosion performance was studied using two — 79- different spherical targets: plastic shells with controlled pressure of deuterium gas to control the implosion convergence and cryogenically cooled foam shells with up to 100 atm. deuterium in the foam layer. 2-2. Corona Instabilities It has been found that two plasmon decay (TPD) instability is the main source of hot electrons and subsequent preheat of the shell in our experiment. [2] TPD instability is the one possibly excited at nJ4 of corona plasma, when the laser light is parametrically coupled with two oppositely going electron plasma waves with almost same frequencies. In our experiment other nonlinear instabilities such as stimulated Raman scattering are rather well suppressed since random phase plates have been inserted into each beam.[3] The random phase plate is a plate to divide the laser beam into more than 104 beamlets with 0 or p randomly distributed phases. With this plate the focal spot is created with overlapping these beamlets independently, reducing the intensity modulations within the focal spot. Shown in Fig. 1 is the relation of hot electron fraction versus TPD signals. TPD signal was monitored as 3co/2 scattered light of laser light(A.= 351 nm).[4] 1 1 r- --1 y • CDFT l.a-3.4vl014 t • A. CH 3.4ilON / + CMC 3.4xlON \—i 1 • Al 3.4X1014 r / / V[ ' / • ** ~ ' "• f t t / • / • / r / */ _ ./ / t 1 - / ' / \ i 1 10'3 10"2 10"1 Fraclion of Three Halves Harmonic Emission (%) Fig. 1 Hot electron vs. 3co/2 light fraction. In the graph, about 2% of incident energy appears to be converted to hot electrons of 15-20 keV temperature for plastic target. Shown in Fig. 2 is unique feature of stimulated Raman scattering (SRS) observed in our experiments. SRS is another parametric instability excited at and below nc/4, where the incident laser light decays into scattered light photon and electron plasma waves. Because of energy conservation, we could tell at which density SRS is excited by monitoring the scattered light frequency. Fig. 2 (a) shows SRS could be excited on the top of the density maximum or its neighborhood at a laser intensity l|_= 4x1014 W/cm2, showing that a relatively narrow spectral region shifts toward the longer wavelength side. In the experiment of Fig. 2(a), thin plastic plane is used to create such a density shape[5]. In the experiment of a spherical two layered target at Ii_=3x1014 W/cm2 with thin plastic (1 urn thick) coated on foam layer (50 urn) similar SRS was found to be excited as shown in Fig.2(b)[6]. Since the future targets could consists of several layers as ablator and fuel, this data indicates that the material and thickness should be chosen carefully to suppress SRS on the density maximum. - 80- 750 800 850 900 950 10001050 1100 Wavelength (nm) 700 750 800 850 900 950 10001050 1100 Wavelength (nm) Fig.2 (a) Stimulated Raman scattering indicated as SSRT excited on low density plasma plateau created in a planar target geometry with one laser beam, (b) SRS excited on two layered spherical target uniformly irradiated with 12 beams. 2-3. Energy Transport Thermal electron energy transport is important to know how much lateral laser intensity deviation could be smoothed out. We have studied this lateral energy transport by observing shock waves launched at the ablation front with imposing intentional laser intensity modulations in space on target[7]. The shock waves were spatially and temporally resolved by observing the back side of the target. The uniformity requirements on the laser beam could be eased by the degree of lateral heat conduction and plasma-hydrodynamic motions occurring between the absorption and the ablation surfaces. Theoretically the scaling of the smoothing on the standoff distance (D) between the absorption and the ablation surfaces and the intensity-modulation wavelength (L) is obtained in terms of heat conduction and plasma-fluid motions[8]. The experiments were conducted with use of a 0.53 ^im, 1ns (FWHM), Gaussian 1Q 9 shaped laser beam (I|_~2-8k10 W/cm ) focused onto thin polystyrene (CH) targets. The imposed intensity ratios of the nonuniformities were ten to one or two to one on targets, thus creating intensity modulations of 4x10"I3 and 4x1012 W/cm2, or 4x1013 and 2x1013 W/cm2, for example. The modulation wavelengths (L) of 50, 75,100, and 120 urn were controlled by the focusing lens and the incident laser energy. The thicknesses (T) of CH targets were changed from 10 to 25 u.m. The speeds of the fast and slow shock fronts were 3.0x106 cm/sec and 2.5x106 cm/sec and correspond to the highest (l|_=4x1013W/cm2) and lowest (l[_=4x1012W/cm2) intensity. The smoothing factor is defined as the ratio of perturbed ablation pressures with and without the thermal smoothing. Here the smoothing factor r is calculated using the 2 2 dependence [9] of Pa «IJ and Pa <* V as, -81- 11 2. '3 3 A 15P aU max.+1 a, min. OV a, max. a, min. r = 2 2 Pa 1% A v a, max. I a, max. a, min. 8P=P -P • 8V = V -V Wl 1 I I I I.L I M I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I p(-kD) o LL. °P 0.1 o o E 0.01 ' Illllllllllllllllllllllll1lll.il! 0.1 0.2 0.3 0.4 D/L Fig.3 D/L dependence of the obtained smoothing factor. The closed circles show the experimental results. The smoothing factor shows clear exponential decrease as D/L increases in Fig. 3. Also plotted are lines predicted by the cloudy-day model (r=exp(-kD)), the sound-wave model (r=exp(-1.34kD)), and the model independently — 82- combining the cloudy-day model and the sound-wave model (r=exp(-2.34kD)). The D/L dependence of the smoothing factor agrees with the simple cloudy-day model or the sound wave model for D/L smaller than 0.1. For D/L larger than 0.1 the smoothing factor is improved rapidly and approaches to T=exp(-2.34kD) fitting. [8] This dependence could be explained with considering the different speeds of energy transport by the ion sound and thermal conduction. Based on our evaluation, the sound wave is faster than the speed of thermal conduction. When the D/L is small, the sound wave mainly contributes to the smoothing, while both sound wave and thermal conduction contribute to the smoothing at D/L larger than 0.12. 2-4. Fiuid Instability We report the observation of RT growth of preimposed perturbations on foils accelerated by irradiation with amplified spontaneous emission (ASE) beams[10]. The focal.pattern is smoothed two dimensionally with ASE beams. Blank foils with no preimposed perturbations showed no observable growth when accelerated under the same conditions. In order to generate a smooth laser irradiation at 0.53 urn ASE beams were used with random phase plates (RPPs). The spectral width of the 0.53 |im ASE was about 0.3 nm. The smoothing time is estimated as a reciprocal of the frequency width to be about 3 ps. The divergence angle of 32 times diffraction limit broadens the speckle pattern produced by the RPP from ~3 \xm (laser) to -100 pm (ASE) in the target plane and smoothes the focal pattern significantly. 10r 12 3 4 5 Time (ns) Fig. 4 Total perturbations between experiments (shown with error bars) and simulations are compared. The line shows the simulation areal density, while the dotted line shows the displacement -83- To further improve the irradiation uniformity, we introduced angular dispersion of the spectrum in one dimension, which is essentially equivalent to the SSD technique. The targets were polystyrene foils with 25-u.m thickness and 500- um width, mounted on the rear side of Be washers with 50-u.m thickness and 500-u.m diameter hole. Sinusoidal amplitude modulations were imposed on the irradiation side of the target from 0.8 u.m to 2 urn with the lateral wavelengths of 30, 60, and 100 u.m. The instability was measured by x-ray back lighting method by making use of Mg Ly a line emission. The x-ray backlit images of RT unstable targets were measured with an x-ray framing camera with 83 ps frame time and 2 ns frame interval. Fig. 4 shows that pR growth measured with 25um thick target having a 60 urn modulation wavelength and 0.8 u.m initial amplitude. Also shown are the results from simulations by IZANAMI, a two dimensional particle-in-cell hydrodynamics code[11]. It is found that the measured evolution of the perturbation growth is in good agreement with the simulation. 2-5. Implosion Experiments A few years ago we demonstrated that the compressed density have reached as high as six hundred times the solid density from the plastic shell implosion experiments equivalent to values required for the ignition[12]. A next step is to achieve a hot spot at the center of the imploded core at the maximum compression by controlling the laser irradiation and the target shell uniformity. As mentioned in the introduction, the implosion experiments were conducted, using plastic and cryogenic foam shells. The former was used to study how much the implosion was kept in spherical symmetry. The latter was employed to study the implosion performance for the future targets[13]. The foam shell had a plastic ablator layer outside of the foam (thickness 10 ^m), typically 3 to 4 u.m. Each target was prepared at the target chamber, using a specially designed cryogenic system. This cryogenic system enables these targets to be irradiated at a desired cryogenic temperature. Once the target was prepared, a small shroud containing the target could be pulled down about 30 msec before the laser irradiation. The power and energy of the laser beams have been balanced within 5 % rms for the implosion experiments. Implosion performance could be measured by various observable parameters. Neutron yield is one of those and have been known to produce only several orders of magnitude less than the predicted by one dimensional simulations[14]. Figure 5 shows recent results of normalized neutron yield, the observed divided by the predicted neutrons, using foam cryogenic targets. The open circles were the results from target implosions with only energy balance of all twelve beams controlled within 5%, while the closed circles were from the ones with both energy and power balances controlled within 5% rms. One can see that the neutron production is improved almost an order of magnitude for the latter implosion experiments. -84 — 1 pI I I I I I I I I I I I I I I I I I I I 111111111111 ii Power & Energy balanced 5 %' 1 0,1 ©0 CD "(0 E Energy balance < 5j % I 0.01 • 0.001 !••• i i i 11 i 111 i i i 111 i 111 11 111 i i i i 111 11 0 10 20 30 40 50 60 70 Convergence Ratio Fig. 5 Normalized neutron yield (the observed divided by the simulated neutron) versus convergence ratios (the initial radius divided by the predicted compressed hot spot radius) of foam cryogenic targets. ACKNOWLEDGEMENT This work has been performed in collaboration with Drs. K. Nishihara, T. Yamanaka, H. Azechi, T.Endo, Y. Izawa, T. Jitsuno, M. Kado, T. Kanabe, Y. Kato, Y. Kitagawa, R. Kodama, K. Mima, N. Miyanaga, M. Nakai, S. Nakai, M. Nakatsuka, A. Nishiguchi, H. Nishimura, T. Norimatsu, H. Shiraga, H. Takabe, M. Takagi, and M. Tsukamoto. — 85- REFERENCES [I] H. Nakano et al., J. Appl. Phys. ,73, 2122 (1993). [2] K.A. Tanaka, M. Kado, R. Kodama, M. Tsukamoto, K. Mima, M. Nakai, A. Nishiguchi, K. Nishihara, T. Norimatsu, T. Yamanaka, C. Yamanaka, and S. Nakai, Japan-US Seminar on Physics of High Power Laser Matter Interactions, Kyoto, Japan Mar. 9-13, 1992, edited by S. Nakai and G.H. Miley (World Scientific Pub. Co.) [3] E.M. Campbell, Phys. Fluids B, 4, 3781 (1992) [4] R.P.Drake R.E. Turner, B.F. Lasinski, K.G. Estabrook, E.M. Campbell, C.L Wang, D.W. Phillion, E.A. Williams, W.L Kruer, Phys. Rev. Lett., 53, 1739 (1984). [5] W. Seka R. Bahr, R.W. Short, A. Simon, R.S. Craxton, D.S. Montgomery, A.E. Kubenshik, Phys. Fluids B, 4, 2232 (1992). [6] M. Tsukamoto, Ph.D. Thesis, Osaka University (1994), (in Japanese). [7] M. Kado, K.A. Tanaka et al., Institute of Laser Engineering Report No.9304p, Osaka University (1993). [8] W. M. Manheimer, D. G. Colombant, and J. H. Gardner, Phys. Fluids, 25, (1982). [9] C. E. Max, C. F. Mckee, and W. C. Mead, Phys. Fluids 23, 1620 (1980). [10] M. Katayama Ph.D. Thesis, Osaka University (1994), (in Japanese). [II] T. Yabe, A. Nishiguchi, and N. Ueda, Appl. Phys. Lett., 39, 222 (1981); A. Nishiguchi, and T. Yabe, J. Computational Phys., 52, 390 (1983). [12] H. Azechi, et al., Lasers and Particle Beams 9,193, (1991). [13] A. Richard, K.A. Tanaka, T. Kanabe, Y. Kitagawa, M. Nakai, K. Nishihara, T. Norimatsu, T. Yamanaka, Y. O. Fukuda, M. Katayama, M. Kado, T. Kawashima, C. Chen, M. Tsukamoto and S. Nakai, Phys. Rev. E., Feb. issue (1994). [14] H. Takabe , J. Plasma & Fusion Res., 69, 1285 (1993), (in Japanese). — 86- Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- INTRAMOLECULAR DYNAMICS AND UNIMOLECULAR REACTION. A CASE STUDY OF STATE-SELECTIVELY EXCITED NITROGEN DIOXIDE Soji TSUCHIYA Department of Chemical and Biological Sciences Japan Women's University Mcjirodai, Bunkyo-ku, Tokyo, 112 Japan Quantum levels of Hie quisibound uissociating NGj were observed by introduction of PHOFEX spectroscopy in which a yield of photofragment molecules in a specific quantum level is delected as a function of the photolysis energy. Thus, the unimolccular rate constant specific to each quantum level of excited N02 was determined which shows a stepwise increase when a new product stale channel opens. This is a direct experimental proof of the statistical theory of the unimolccular reaction process. Furthermore, the product stale distribution indicated quantum fluctuations associated with the individual quasibound cigenstatcs. Keywords: Stalc-to-state reaction, Laser-induced reaction, Molecular reaction dynamics 1. INTRODUCTION Investigation of the unimolccular reaction of polyatomic molecules selectively excited in a specific vibrational mode have attracted great interest, since it may answer the key question of how fast is the intramolecular vibrational rcdistribution(IVR) on a chemically significant time scale. If IVR occurs very fast, i.e., energy flows rapidly and randomly among all the vibrational modes, the unimolccular reaction proceeds statistically through the transition state from rcactants to products[l]. Then, the rate constant can be predicted by the statistical Ricc-Ramspcrgcr-Kasscl-Marcus(RRKM) theory [2] as k(E, J) = W(E-Eo, J)fh p(E, J), (1) -87- where the activated molecule contains the total energy E with the total angular momentum J, and p(E, J) is the density of the states of the activated molecules, and W(E-Eo, J) is the number of energy levels for vibration orthogonal to the reaction coordinate at the transition state whose lowest level energy is Eo (=activation energy). By application of recent laser techniques, it is possible to prepare a molecule in a certain cigenstate, and thus, the mode specificity of the unimolccular reaction can be tested by measuring the rate constants specific to respective cigenstates whose energies exceed the potential barrierEo of the transition state or by observation of the dependence of the product state distribution on the well-characterized initially excited level of the activated molecule. A large number of experiments have been performed so far to discuss the state-specificity. However, it is not always clear in most of these experiments how the unimolccular reactions arc related to the initially excited states or the vibrational modes. A clear observation of the state-specificity has been reported by the group of Moore in Bcrkclcy[3, 4]. An example is the photodissociation of HFCO which keeps quasibound states with high excitation in the out- of-planc bending modc[3]. If this mode is excited, the activated molecule dissociates slowly compared with the molecule having the same energy in other modes. This is interpreted as the fact that the out-of-planc bending mode is uncoupled with other modes. Another significant finding of this group is that the dissociation rate constants of D2CO in the energy region above the dissociation barrier fluctuate within three orders of magnitude though the averaged rate constant is in accord with the RRKM thcory[4]. This fluctuation is interpreted as a quantum statistical fluctuation which is necessary consequence of the complete IVR[5]. In order to investigate the state-specificity, it is necessary (1) to excite the relevant molecules to a specific cigenstate, (2) to measure the unimolccular rate constant from this particular level, and (3) to know the number of vibration-rotation levels of the transition state molecule as well as the density of states of the activated molecule. It is not easy to make an experiment to satisfy these requirements. In our group in University of Tokyo, Komaba, we succeeded to excite NO2 to a specific vibration-rotation state in the energy region which -88- exceeds the dissociation threshold, and furthermore, to observe the rate constant of the excited NO2 in each vibration-rotation state through homogeneous line-broadening of the photofragment cxcitation(PHOFEX) spectrum, which was taken by scanning the wavelength of the photolysis laser with monitoring a photofragment NO or O in a specific quantum statc[6]. We also could determine experimentally the density of state of the excited N02[7]. Thus, we could answer the question whether the dissociation reaction of NO2 proceeds in a dynamical or statistical mechanism. This paper summarizes our recent topics on NO2 photodissociation and discussions on unimolccular reaction dynamics. 2. EXPERIMENTAL In order to prepare NO2 in a single rovibrational cigenstate as an initial state of the unimolccular reaction, NO2 was cooled down rotationally in a supersonic jet formed by expansion of a NO2(0.2%)/Hc mixture through a pulsed nozzle under a stagnation pressure of 5 atm into a vacuum chamber. The rotational temperature of NO2 in the jet was estimated from the observed laser-induced fluorcsccncc(LIF) spectrum of NO2 which was excited in the lower energy region far below the dissociation threshold. The rotational temperature derived was about 0.9K, at which almost all NC>2(98%) molecules arc in the ground rotational level and the concentration of rotationally hot NO2 is only 2%. This implies that only one rotational transition R(0) is allowed in the photoexcitation of NO2. Thus, NO2 is prepared in the rotational level of A^=l of selected vibronic states. Two dye lasers(Lambda Physik FL3002E) were simultaneously excited by a XeCl excimer laser (Lambda Physik EMG101MSC); one dye laser output was used to excite and photolyzc NO2 and the other, which was frequency-doublcd(around 226 nm), was used to monitor the NO fragment. The resolution of the lasers is 0.5 enr1 in the low resolution measurement and 0.05 enr1 in the high resolution. — 89- The photolysis laser beam and the monitor laser beam collincarly countcrpropagatcd and perpendicularly crossed the jet. The photofragment NO produced from NO2 was detected by measuring the LIF signal of the A-X transitions. In order to record the PHOFEX spectrum of NO, the wavelength of the photolysis laser was scanned with fixing the monitor laser wavelength at each of the A-X vibration-rotation transitions of NO. 3. RESULTS LIF and PHOFEX spectra near the dissociation threshold In Fig. 1 shown is the observed LIF spectrum of NO2 below the dissociation threshold together with the PHOFEX spectrum above the dissociation threshold obtained by monitoring NO in the rotational ground level J=0.5 of the electronic and vibrational ground state. From this figure, the dissociation threshold is clearly identified, since as the excitation wavenumber is scanned towards higher energy, the LIF signal disappears suddenly at the threshold and the PHOFEX signal starts to emerge. In this study, the photolysis laser energy relative to the threshold, i.e. the excess energy is employed for discussion. Fig. 1 The high resolution LIF spectrum of N02 below the dissociation threshold and the PHOFEX spectrum of N02 above the dissociation threshold observed by monitoring the A" component of NOpn^; v = Q,J = 0.5). Adapted from Rcf. [6]. EXCESS ENERGY / cm"1 The lincwidths of the LIF spectral peaks below the dissociation threshold arc all about 0.055 cm-1 which is attributed to the laser resolution. Then, energy positions of these peaks mean the level energies of the excited NO2 in J=0.5. From the observed number of peaks the average density of states can be determined to be 8.2 /cm-1 just below the dissociation threshold. — 90 — However, since NO2 has odd electrons which cause the existence of two spin states for one vibration-rotation state, the density of the vibronic states should be 4.1 cm"1. A spectral structure is seen in the PHOFEX spectrum which indicates the existence of discrete states, while the lincwidth of each spectral peak is considerably broader than that in the LIF spectrum. It is reasonable to interpret that the level structure is similar in the energy region just below and above the dissociation threshold, and thus it is concluded that NO2 predissociatcs through its quasibound levels above the dissociation limit and this causes the homogenous line broadening. When the photolysis energy exceeds the dissociation threshold by 5.0 crrr1, the second lowest dissociation channel producing NO in the rotational level of J=0.5 and 1.5 opens. In Fig. 2, two PHOFEX spectra of J=0.5 and 1.5 for the A' component of the A doublet of product NO arc shown. In the energy region where only J=0.5 is produced, the widths of the spectral peaks arc narrow, while in the energy region where both of the two channels J=0.5 and 1.5 arc energetically allowed, the peak width becomes larger suddenly when the new product channcl(J=1.5) opens. Fig. 2 The PHOFEX spectra for J =0.5 and 1.5 of NO in A' component. The threshold 5.01 ±0.05 for the NO J =1.5 spectrum coincides with the rotational energy 5.015 cm-i of the J =1.5 level of free NO. Adapted from Ref. [6]. 0 5 10 EXCESS ENERGY/cm-1 The observed PHOFEX line-profile can be fitted by a Lorentzian function, and the FWHM -Tcan be interpreted as a dissociation rate constant k by the relation r=(hl2n)k. The determined rate constants scatter significantly to the extent of an order of magnitude, the smallest -91- rate constant of which is 8.5xl09 s-1. The cause of this scatter may not be attributed to the experimental accuracy. Product state distribution With a fixed wavenumber of the photolysis laser, the state distribution of NO or 0 can be observed. For example, the population ratio in the A' component vs the A" of the A doublet of NO can be determined by comparison of LIF intensities of R or P branch vsQ branch transitions. The results for J=0.5-6.5 rotational levels of product NO are given in Fig. 3 as a function of the photolysis energy. The observed fractions fluctuate considerably and no correlation is found among these oscillatory patterns. However, if the fraction of the A' component is averaged over a wide energy range, the average value converges to 0.5. This finding is common to the population distribution in rotational levels; the observed distribution is dependent considerably on the photolysis energy, while if the distribution is averaged in the wide photolysis energies, the distribution tends to be that of statistical degeneracies of respective levels. J=6.5 •W^H; J=5.5 •V^^ty/V^ J=4.5 Fig. 3 The fraction of the A' component Jt(A')A(A')+Jt(A") "A«?''vW* °0 50 100 150 EXCESS ENERGY I cm"1 4. DISCUSSIONS Dissociation rate constant Our results show that the dissociation rate constant k increases stepwise when a new product channel opens. This supports that there is no barrier — 92— on the potential surface of NO2 along the dissociation coordinate. If a barrier exists, the activation energy should be dependent on the product state channel. Thus, the transition state of the NO2 dissociation is located very close to the product state and is regarded as a "loose complex" in which the NO fragment may rotate freely. If this loose complex picture is assumed, the RRKM theory is essentially similar to the Phase Space Thcory(PST)[8], which is a theory to predict the product state distribution in a unimolccular reaction assuming statistical partitioning of the energy among all of possible product states. By counting the number of channels, W may be easily calculated; e.g., W=4 and 8 for J=0.5 and7=1.5, respectively in the case of J(N02)= 0.5. Since p can be treated as a constant in the narrow energy range close to the dissociation threshold. Therefore, when the new channel of product NO opens, W increases stepwise and consequently, k increases stepwise. To obtain an absolute value of k, it is necessary to estimate p. In the present experiment, p is directly determined from the observed LIF spectrum to be 4.1/cm-i. Owing to conservation of the total angular momentum, the rate constant based on the PST is dependent on the initial J of NO2. The calculated rate constant to produce NO(J=0.5) is 2.9xl09 and 5.8xl09 S'1 for 7(NO2)=0.5 and 1.5, respectively. These values are in accord with the observed rate constants within an order of magnitude. In PST, it is assumed that all of the angular momenta involved in the reaction process are spatially isotropic, and as a consequence, all M sublevels are equally populated. However, recent studies on the photodissociation processes have revealed that the angular momentum distribution of the fragment is often anisotropic[9]. Thus, the above value of W is possibly overestimated. Rotational and A doublet component distribution The state distribution of the photofragment is a reflection of the initial quantum state onto the final dissociative channels. The observed large fluctuation of the A doublet component distribution and the rotational distribution should be interpreted as the fact that the eigenfunctions of respective states are — 93- different significantly state to state. In other words, the fluctuation is a necessary consequence of complete mixing of vibrational states. The vibrational dynamics of NO2 has been proved to be chaotic from the nearest neighbor level spacing distribution, the Fourier transform of the observed spectra, and the peak intensity distribution etc. [7] These analyses indicate that there exists a strong correlation among the vibronic levels. Thus, the observed state-dependent partial rate constant to yield a certain quantum state is concluded to be a quantum fluctuation as proposed by Polik et ct/.[3,4] and Miller et al. [5]. The fluctuation of the rotational distribution was also reported by the group of Rcislcr [10]. However, in this experiment, the rotational hot bands and the limited laser resolution cause the initial excitation of NO2 to multiple quantum levels, so that the observed distribution is somehow averaged or smoothed. In high excess energies more than 100 enr1, it is quite difficult to choose a single quantum level of excited NO2. Nevertheless, they suggest a mechanism different from ours,/.e., an interference among overlapping resonances for the cause of the observed fluctuation. To clarify the coherent effect of the excited quantum states which couple with the dissociative channels, some new spectroscopic method is necessary to be applied. 'Loose'v j.'tight'complex The dissociation of photoexcitcd NO2 can be investigated by the time-resolved experiment, i.e., pulsed excitation of NO2 followed by detection of product NO through another timc-dclaycd light pulse. This timc-rcsolvcd detection method was applied by the group of Wittig[ll], who reported the rate constants just above the threshold and found two distinct steps of the rate constant at around 100 and 200 cm-1 of the excess energy. We interpreted this result in our papcr[6] as an opening of electronically excited 2 3 channels, NO( n3/2) and/or 0( Pi,o) according to our PHOFEX spectral data which support the PST. Contrary to this, the group of Wittig prefers a mechanism that the rate constant is enhanced when the excitation energy exceeds the bending vibrational quantum levels of the transition state. If this mechanism is adopted, the transition state is that of a tight complex which has necessarily a barrier along the reaction coordinate. However, our observed stepwise - 94 — increase of the rate constant just above the dissociation threshold contradicts their proposal. Thus, in order to accommodate this contradiction, they proposed that the transition state changes from 'loose' to 'tight' as the excitation energy incrcases[12]. This mechanism may be named a 'tightening' model. In reply to their commcnt[13], we referred to two recent theoretical papcrs[14,15]. These theoretical calculations of the potential energy surface of NO2 showed that there is no barrier for the lowest two channels of the NO rotation and that the barrier height increases while its position moves closer to the rcactant side as the bending quantum number increases. These predictions seem to explain successfully both the frequency and time domain experiments. However, it should be pointed out that the equivalence of these two experimental methods has not been proved yet. Further experimental trial, especially the measurement of the dependence of the rate constant on the pulse-width of the excitation laser would be desirable. 5. CONCLUSION (1) Up to the excess energy of 100 enr1, which is the upper limit of the present experiment, the dissociation reaction proceeds on a potential energy surface without a barrier. (2) The dissociation rate constant of NO2 in quasibound quantum levels increases stepwise when new product channel opens. This supports the statistical unimolccular reaction, i.e., the phase space theory. (3) The partial rate constants to produce NO in specific quantum states fluctuate considerably as a function of the excitation energy. This can be interpreted as a quantum fluctuation due to the complete IVR in excited NO2. (4) Further detailed experiment is necessary to make judgement on whether the transition state changes from 'loose' to 'tight' complex as the excitation energy increases. - 95- ACKNOWLEDGEMENT This work was conducted in Department of Pure and Applied Sciences, University of Tokyo under collaboration with Drs. J. Miyawaki and K. Yamanouchi, and was supported in part by a Grant-in-aid for Scientific Rcscarch(No. 02403006) from Ministry of Education of Japanese Government. REFERENCE [I] Crim F.F.: Science249,1387 (1990). [2] For example, Robinson P.J. and Holbrook K.A.: Unimolecular Reactions (Wiley, London, 1972). [3] Polik W.F., Guycr, D.R. and Moore C.B.:/. Chan. Phys. 92,3453 (1990). [4] Polik W.F., Guycr D.R., Miller W.H. and Moore C.B.:./. Chem. Phys. 92, 3471 (1990). [5] Miller W.H., Hernandez R., Moore C.B. and Polik W.F.: J. Chem. Phys. 93, 5657 (1990). [6] Miyawaki J., Yamanouchi K. and Tsuchiya S.:J. Chem. Phys. 99, 254 (1993). [7] Miyawaki J., Yamanouchi K. and Tsuchiya S.:J. Chem. Phys. 101, (1994) in press. [8] Pcchukas P. and Light LC.:J. Chem. Phys. 42, 3281 (1965). [9] Hall G.E. and Houston P.L.:Aiinu. Rev. Phys. Chem. 40, 375 (1989). [10] Hunter M., Rcbic S.A. and Rcislcr H.: J. Chem. Phys. 99, 1093 (1993). [II] Ionov S.I., Bruckcr G.A., Jaqucs C, Chen Y. and Wittig C.'.J. Chem. Phys.99, 3420 (1993). [12] Wittig C. and lonov S.I.: J. Chem. Phys. 100, 4714 (1994). [13] Miyawaki J., Yamanouchi K. and Tsuchiya S.:J. Chem. Phys. 100, 4716 (1994). [14] Klippcnstcin S.J. and Radivoycvitch T.:J. Chem. Phys. 99, 3644 (1993). [15] Katagiri H. and Kato S.: J. Chem. Phys. 99, 8805 (1993). -96- Proceedings ofthe 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - ISOTOPIC ANALYSES USING RESONANCE IONIZATION SPECTROSCOPY Tom J. WHITAKER, Heinrich F. ARLINGHAUS, and George I. BEKOV Atom Sciences, Inc. 114 Ridgeway Center Oak Ridge, TN 37830, USA Resonance Ionization Spectroscopy (RIS) is a laser technique which provides a sensitive and selective method of measuring the elemental concentration and/or i so topic distribution of trace materials. RIS techniques used at Atom Sciences, Inc. for elemental and isotopic analysis are described. Keywords: Resonance Ionization, Postionization, Laser, Elemental Analysis, Isotopic Analysis Introduction Resonance Ionization Spectroscopy (RIS) is a laser-based analytical technique in which wavelength- tunable lasers are used to efficiently and selectively ionize atoms of a specific analyte element. In RIS, the wavelength of the RIS laser is tuned to resonance with an electronic transition between the ground state (or a populated low-lying state) and an excited state. Often a second resonance step is used in which a second RIS laser is tuned to resonance with a transition which further excites the atom to an even higher excited state. The excited-state atoms are then ionized either by photoionization, electric field ionization, or collisional ionization, and the resulting ions are monitored by a particle detector. The steps of a typical RIS experiment would involve: (1) Vaporization and atomization of the sample (2) Suppression of ions formed in the vaporization/atomization process (3) Resonance ionization of the neutral atomic species (4) Detection of the RIS ions This paper describes steps 1 and 3; vaporization/atomization techniques and RIS laser schemes, in use at our lab. The discussion is limited to photoionization but interested readers should refer to the proceedings of the International Symposium on Resonance Ionization Spectroscopy [1] for a description of a wide variety of RIS techniques and applications. -97- Vaporlzatlon/Atonilzatlon Techniques RIS must be applied to atoms or molecules in the vapor phase. For bulk analyses of solids and liquids, thermal vaporization techniques arc simple and effective. However, if pulsed RIS lasers arc used, the mismatch between the continuous nature of thermal vaporization processes and the low duty cycle of pulsed lasers can lead to inefficient sample utilization and limit sensitivity. An alternative vaporization method is laser ablation. In this technique, the pulsed ablation laser can be synchronized with the RIS lasers to provide optimal overlap. However, care must be taken to use the proper intensity for the ablation laser so that the sample'is atomized. Another approach to pulsed vaporization is to use an ion beam to sputter atoms from the surface. We call RIS analyses based upon the above three methods of sample vaporization respectively, Thermal Atomization RIS (TARIS), Laser Atomization RIS (LAR1S), and Sputter-Initiated RIS (SIRIS). TARIS - TARIS is similar to graphite furnace atomic absorption spectrometry (GFAAS). Typically in GFAAS in that a hot reducing surface is used to vaporize and atomize the sample. However, TARIS differs from GFAAS not only in the fact RIS is used instead of atomic absorption, but also in that TARIS operates at low pressures so collisions and chemical reactions are negligible, greatly reducing matrix effects. This reduces sample preparation requirements to the point that some samples require only a simple acid digestion and some require no preparation at all. Not only is the time and cost of complex sample preparation avoided, but there is less chance to lose analyte in an incomplete preparation step or to introduce contaminants. This, combined with the selectivity and sensitivity of the RIS process, makes TARIS an excellent method for bulk analysis, especially for environmental and biological tissue samples where the substrate is complex and variable. Unfortunately, TARIS does not completely eliminate all matrix effects. Some effects are still present due to incomplete atomization of volatile molecular species. Just as in GFAAS, our current TARIS system uses a single stage of heating to accomplish both vaporization and atomization. Some molecules vaporize at temperatures too low to allow complete atomization. To eliminate this source of matrix effects, we plan in the near future to separate the vaporization and atomization functions into two stages so that the vaporization can be carried out normally, but atomization will occur on a separate hot surface, maintained at a temperature high enough to assure complete atomization but low enough to minimize surface ionization. This is similar to the two-stage atomizer developed by Bekov et al. [2] in which the atomic yield for pure rhodium was equal (within experimental error) to the yield for rhodium chlorides. This device should further reduce sample preparation requirements and decrease the need for internal standards. One disadvantage of TARIS is the mismatch between the continuous nature of the vaporization process and the pulsed RIS lasers. Because TARIS operates at low pressure, atoms leaving the hot atomization surface — 98- can pass through the laser interaction region between laser pulses. For example, a 0.5 cm laser beam from a 30 Hz laser would overlap only 2X10"4 Sr atoms emitted from a 2000°C source, even if the atomic beam were perfectly collimatcd. In spite of this limitation, TARIS is often orders of magnitude more sensitive than conventional techniques and its simplicity and low cost make it very useful for many bulk analyses. LARIS - Pulsed laser desorption (also called ablation) is an effective way of vaporizing material from a surface in a manner which can be synchronized with RIS lasers. By adjusting the delay between the ablation laser and the RIS lasers, the overlap of vaporized particles and RIS lasers can be optimized and the useful yield (number of ions detected per analyte atom vaporized) in LARIS can be on the order of a few percent. Another advantage which results from the use of a laser beam to vaporize the sample include the ability to focus the ablation laser to micrometer dimensions to provide microprobc capabilities. Also, the amount of material ejected on each laser pulse is much larger (over three orders of magnitude in many cases) than possible with pulsed ion beams and therefore the analysis of trace quantities can be performed more rapidly. It is important to be able to attenuate the ablation laser because desorption, atomization, and ionization are sensitive functions of intensity. In an excellent review of laser vaporization techniques, Dittrich and Wennrich [3] report that in the intensity range from ablation threshold (about 104 W cm"2) to 7xl07 W cm'2, the resulting vapor consists of polyatomic particles and that above 5xl08 W cm"2, the material is partially ionized. Within the range producing neutral atomic species, the yield is a nonlinear function of laser intensity. However, the neutral atomic yield is also a function of laser wavelength [4], pulse duration, target material, buffer gas composition and pressure, and surface morphology. Therefore, quantitative measurements using LARIS require not only careful control of laser intensity, but also calibration using standards similar in composition and surface morphology to the analyte. However, when used carefully, LARIS offers a sensitive microprobe technique well- suited for trace analysis of many materials, especially biological samples. Because of the high lateral resolution available and the ability to ablate material down to different depths, LARIS can rapidly produce three- dimensional images of the concentration of trace materials. S1RIS - Sputter-initiated RIS utilizes a pulsed ion beam to remove particles from a surface for RIS analysis. It has many of the benefits and attributes of LARIS but is much more controlled and reproducible, literally allowing depth profile data to be taken at sub-monolayer steps [5]. In our apparatus, we can use different ion guns to create ion spot sizes that range from over 100 /mi in diameter (using Ar+ ions) to spot sizes of less than 0.1 ftm (using a liquid Ga+ion gun). This extraordinary lateral and depth resolution can be used for creating three-dimensional images of trace concentrations in extremely small volumes, e.g. biological cells, semiconductor devices, or optical fibers. -99- TARIS, LARIS, and SIRIS all benefit from the separation of the ionization step from the atomization step. For example, SIRIS is similar to SIMS (Secondary Ion Mass Spectrometry) except that SIMS monitors secondary ions created by the sputtering process. These ions constitute 10"7 to 10'2 of the sputtered material, depending on the composition and morphology of the surface (i.e. matrix effects). As the ion yield changes over 5 orders of magnitude, the neutral yield varies by less than 1%, almost eliminating matrix effects for SIRIS measurements. Resonance Ionization Schemes Both the efficiency and selectivity of RIS stem from the resonance excitation step(s). Cross-sections for resonant excitation are orders of magnitude greater than those for nonresonant excitation and therefore relatively low-power lasers produce efficient excitation. Resonance wavelengths for most elements are well- separated compared to the bandwidth of commercial lasers and therefore, when the laser is tuned to resonance with the analyte element, it is well off-resonance with any potentially interfering species. This is the reason that RIS is extremely selective. We have previously demonstrated [6] that ionization efficiency for the selected element can be as much as 109 times higher than for the other elements in the sample. Molecular interferences can be a problem because their broad absorption bands may overlap the resonance, but the effect is quite small because the oscillator strength is spread out over the entire vibrational/rotational band, and the effect becomes negligible in most cases when a double resonance scheme is used. RIS selectivity helps simplify interpretation of analytical data by reducing or eliminating interferences and it helps maintain signal linearity by reducing space charge effects that would otherwise be present due to ionization of major constituents of the sample. The use of two resonance steps generally offers better selectivity than single-resonance schemes. Figure 1 shows generic single-resonance and double-resonance RIS schemes and their applicability. Scheme (a) in this figure shows a simple single-resonance scheme. This is the simplest but not necessarily the most desirable scheme. In the figure, Scheme (a) is shown using an infrared (IR) wavelength for the photoionization step. It is also possible to use visible (VIS) or ultraviolet (UV) wavelengths to perform photoionization, and this option is attractive because a single wavelength can be used for both the resonance and photoionization. However, photoionization cross-sections are much lower than those for resonance steps and therefore the photoionization laser intensity must be proportionally higher in order to efficiently ionize the atoms. High-intensity visible or UV laser beams can cause nonlinear ionization of interfering species but this is seldom a problem with high-intensity IR beams. Therefore, whenever selectivity is an issue, it is better to use IR wavelengths for the photoionization step. Scheme (a) is limited to very low ionization potential elements if -100- (a) 00 Ccj Cd) (e) Autoionizing Level ^„;t,.,,^,Y,;mm A3IR I fSKSiSKS*11 •HI X, A-2 VIS X3 IR "k3 VIS V 4 X V1S 2 x2 vis \9 IR i I 11 A, *1 vuv 7 * uv X, UV uv or or uv or VIS VIS VIS or VIS (a) Applicable to (b) Applicable to (c) Also applicable to (d) Applicable to (e) Also applicable to elements with low most elements. The most elements. elements with high elements with high double resonance Provides even greater Ionization potential Ionization potential Ionization potential offers greater selectivity than (b) but such as carbon or but also to elements such as alkali selectivity than a requires an extra krypton. with dense spectra metals. single resonance. tunable laser. such as actinides. Figure 1 Different implementations of the resonance ionization concept. Scheme (b) and (c) give the best selectivity and are applicable to over 85% of the elements. an IR wavelength is used for photoionization. Scheme (b) is the most versatile of 107 the schemes shown and it gives the best 108 • Implant, 10 keV, 10 & o i A Implant, 30 keV, 10' performance. Only H, He, C, N, O, F, Ne, P, •g 10 r O Implant, 25 keV, 10 te S, CI, Ar, Br, Kr, I, Xe, and Rn cannot be 10* • r Implant, 15 keV, 10 , vr.vf.vf.v?o",\d7 analyzed using Scheme (b) with commercial •g 10 t- a Homogeneousllomogci y Doped laser systems and frequency upconversion in & 10 nonlinear crystals. Figure 2 shows an example 101 r of the use of Scheme (b) to detect boron 1 10° sputtered from a Si substrate in the SIRIS 10-' Background = 0 counts c« a process. io- In 33,000 shots or < 2x10° atoms/cm8 4 RIS Scheme (c) in Figure 1 can be 10- •-' -• -i 1012 101" 1018 101B 1020 1022 used when autoionizing structure exists in the B Concentration (atoms/cm j ionization continuum. The cross-section for Figure 2 An example of Scheme (b) applied to the SIRIS measurement of boron in silicon. Note the dynamic range and autoionization is significantly larger than for linearity of the measurements. normal photoionization and therefore the -101- intensity of the laser docs not have to be as high. This means that visible wavelengths can be used without extensive nonlinear ionization of interfering species. Often, high-repetition rate pump lasers such as copper vapor lasers or excimer lasers can be used for this scheme. Scheme (d) is used for elements for which even the lowest-energy excited states require vacuum UV (VUV) wavelengths. It differs from Scheme (b) only in the experimental method used to generate the wavelength for the first transition. For wavelengths below about 195 nm, either four-wave mixing or anli-Stokes Raman must be used. This additional requirement makes Scheme (d) the most experimentally complex of the resonance ionization schemes. We have used this technique in our laboratory to measure the amount of 8lKr (a radioactive isotope with approximately 1 part in 1018 natural abundance in the atmosphere) in a water sample after several isotope enrichment steps [7]. Scheme (e) is generally used when the energy of the excited state structure is too high for Schemes (a), (b), or (c) and the researcher does not wish to go to the experimental complexity to set up Scheme (d). Examples are C, N, O, P, S, CI, Br, Kr, I, and Xe. The two-photon transition used in Scheme (e) is a second- order event and therefore its cross-section is quite low compared to the one-photon processes described in the other schemes. This means that high-intensity light is needed, and focussing of the laser beam is generally required. In fact, the intensity of the light needed for efficient two-photon excitation is usually sufficient to ionize the excited state and so Scheme (e) often requires only a single wavelength. This experimental simplicity is counterbalanced by the fact that the high-intensity light can cause nonlinear processes, including multiphoton ionization, to occur in many other species, leading to interferences. An interesting exception to the requirement of -« Q« high-intensity laser light for Scheme (e) occurs when a ^PuinSi.lOkeV one-photon allowed intermediate state is nearly .v , 'Dose = 4.7x109 cm'2 P 10" i. \ resonant with the laser wavelength. In this case, the § : Tr v \ "S I » ... T \ i two-photon excitation is resonantly enhanced and much jP °4j_| \\ lower intensities arc required (the photoionization step ~* j— 1ppb "Y may still require high intensity to efficiently ionize). .& i vr An example is shown in Figure 3 where we used a g 53 10 single 586.46 nm wavelength (17046.7 cm'1) in Scheme 0^ 20 4Q 60 80 100 120 (e) for RIS analysis of plutonium. The two-photon Depth (run) excitation was resonantly enhanced by the existence of ^^ 3 g. ^ imp]anied wjth 10 keV n*v a, 10„ a 9G, slate at 17045.8 cm"1 only 0.9 cm'1 below the atoms cm'2 and irradiated to produce an effective 239Pu dose of 4.7xl09 atoms cm'2. — 102 — virtual state at 1/2 the energy of the two-photon transition. Because of this enhancement, sub-part-per-billion measurements were possible with Scheme (e). Figure 3 also demonstrates an interesting phenomenon in two-photon excitation. A two-photon transition from a J=0 state to a J=l state is forbidden when the nuclear spin, /, is 0 (as it is in even isotopes of Pu) and the polarization is either pure linear or pure circular. However, odd isotopes of Pu have nonzero nuclear spin and the transition becomes allowed, explaining the good RIS signal we observed in Figure 3. Applicability As mentioned tibove, commercial laser systems using conventional nonlinear crystals to reach UV wavelengths can be used to apply RIS analyses to most of the elements. For example, Scheme (b) can be used on all elements except H, He, C, N, O, F, Ne, P, S, CI, Ar, Br, Kr, I, Xe, and Rn. Of these elements, Ar and F require four-wave mixing techniques illustrated in Scheme (d) to generate vacuum ultraviolet wavelengths, while the others except He and Ne can be done using Scheme (e) (although Scheme (d) may give better results). Only He and Ne cannot be done by any of the schemes shown in Figure 1. This is somewhat ironic since the first resonance ionization experiments performed at Oak Ridge National Laboratory by Hurst and coworkers, was on He [8]. However, these early experiments were performed on a metastable state of He and the limitations described above are for RIS analysis of ground-state atoms. Isotope Analyses Isotope shifts for most elements are small in comparison to the bandwidth of the pulsed lasers used in RIS experiments (the pulsed lasers used by Atom Sciences have 7-12 GHz bandwidth), and thus all the isotopes of the analyte will be essentially resonant with the RIS laser. In this case, isotopic analysis is achieved with a mass spectrometer (this is often called Resonance Ionization Mass Spectrometry, or RIMS). Time-of-flight mass spectrometers are especially well-suited for isotopic analysis of ions produced by pulsed RIS lasers because all the ions are detected on each pulse. Figure 4 shows time-of-flight data for SIRIS analysis of molybdenum isotopes. Each peak height is within 2% of the accepted value for the natural abundance of the respective isotope. One might expect that as long as the laser bandwidth is wider than the isotopic splitting, all isotopes will be ionized with equal efficiency. This has been shown to not be the case in several experiments where odd- even isotope anomalies have been reported [9]. Lambropoulos has explained in a series of papers [10] that this phenomenon is due to the coupling between hyperfine stales of isotopes with nonzero nuclear spin. If the - 103- coupling occurs at a rate that is a significant fraction of the time the average atom spends in the excited state, the ionization probability can be modified. The magnitude of this effect is dependent on several factors; the hyperfine splitting, laser pulse duration, laser polarization, ionization laser intensity, and ionization laser wavelength. Of course, laser bandwidth is also important if it is comparable to the isotope splitting. The good news is that it is possible to obtain 96 98 Mass (AMU) reproducible results if all these factors are controlled, Figure 4 Time-of-flight SIRIS spectrum of Mo isotopes. Amplitudes agree within 2% of the natural and therefore it is practical to use standards to calibrate abundance. the odd-even isotope effect for a particular RIS system. Another approach to isotopic analysis with RIS is to use narrow-bandwidth continuous wave lasers which can selectively excite specific isotopes of an analyte and sequentially scan over each isotope [11]. In combination with a mass spectrometer, this technique can permit detection of extremely small abundance isotopes. However, it too has some problems with anomalous peak height ratios between isotopes [12] related to optical pumping into specific magnetic substates of a hyperfine level, and care must be taken when quantitative work is required. RIS as a Light Detector Usually RIS is used to detect a small number of atoms or molecules by irradiating the vaporized sample with a large flux of photons. It is possible to reverse the process and detect a small flux of photons with a high density of atoms.The concept of using photon-induced ionization as a photon detector was first proposed in 1979 by Matveev and coworkers in the (then) USSR [13],[14] and it has received recent attention from Omenetto and coworkers [15]. Winefordner's group at the University of Florida has very recently demonstrated laser-induced ionization detection of 103 photons at 285.2 nm using pulsed lasers to ionize magnesium in an acetylene/air flame [16]. These techniques require the signal photon to produce a transition to an excited electronic state, typically from the ground state. This is followed by direct laser photoionization or laser excitation to a Rydberg state followed by field ionization or collisional ionization of the Rydberg atom. The RIS light detector can have extremely narrow bandwidth (sub-Doppler in some cases) and therefore it should be useful for detecting — 104- emission from atoms in a plasma or other narrow-band sources that would normally be difficult to detect because of bright, broadband background. Summary RIS is a powerful analytical tool for a number of applications. It has the advantage of providing extremely efficient and selective ionization. In cases where the sample size is limited, this results in sensitivity that is orders of magnitude better than conventional techniques in many cases. When used with laser ablation or ion beam sputtering, this efficiency allows excellent lateral and depth resolution of even trace materials. Another advantage is the lack of matrix effects. Because the ionization step occurs in the vapor phase at low pressures, many of the matrix effects that plague conventional techniques are reduced or absent in RIS analyses. RIS is applicable to almost all the elements and, when practiced with care, can be used for isotopic measurements as well. The major disadvantages of RIS analysis is the initial cost of the apparatus, the time required for analysis, and the fact that information is obtained for only a single element per laser system. Work is underway in our laboratory to provide a survey capability in TARIS and SIRIS by interleaving a nonselective ionization source between RIS laser pulses. In the case of TARIS, we will add an electron impact ionization source to the pulsed extraction region of the time-of-flight mass spectrometer. For SIRIS, we will simply keep pulsing the ion beam and monitor secondary ions. These additions will combine multi-component analysis with high-sensitivity measurements for a single element. Acknowledgement The authors wish to acknowledge the technical assistance of Chuck Joyner in generating much of the data contained in this document. References 1. Resonance Ionization Spectroscopy 1992. edited by Charles M. Miller and James E. Parks, Conference Series Number 128, published by the Institute of Physics Publishing, Bristol and Philadelphia (1992). Proceedings of earlier RIS conferences include Conference Series #114 (1990), #94 (1988), #84 (1986), and #71 (1984) by the same publisher. 2. G.I. Bekov, I.V. Kolpakov, V.N. Radaev, and V.A. Veselov, "Highly sensitive elemental analysis by laser photoionization spectrometry in vacuum", p. 265-270, Resonance Ionization Spectroscopy 1990. published by Institute of Physics, Bristol and Philadelphia, Conf. Ser. #114, edited by J.E. Parks and N. Omenetto, (1990). -105- 3. K. Dittrich and R. Wennrich, "Laser Vaporization in Atomic Spectroscopy", Prog. A nalyt. A torn. Spectrosc. 7 139-198 (1984), and references therein. 4. H.F. Arlinghaus and N. Thonnard, "Comparison of Atomization Processes: Trace Element Analysis Using RIS of Laser-Irradiated and Ion-Bombarded Biological and Metal Surfaces", Laser Ablation. Mechanisms and Applications. Lecture Notes in Physics #389 published by Springer-Verlag, Berlin, New York, Tokyo, edited by J.C. Miller and R.F. Haglund, Jr., (1991). 5. H.F. Arlinghaus, M.T. Spaar, N. Thonnard, A.W. McMahon, T. Tanigaki, H. Shichi, and P.H. Holloway, "Quantitative and sensitive profiling of dopants and impurities in semiconductors using sputter-initiated resonance ionization spectroscopy",/. Vac. Sci. Technol. A 11(4) 2317-2323 (1993). 6. D.W. Beekman and N. Thonnard, "Laser Ablation as an Atomization Source for Ultratrace Element Analysis Using Resonance Ionization Time-of-Flight Mass Spectrometry," Resonance Ionization Spectroscopy 1988. published by Institute of Physics Publishing, Bristol and Philadelphia, Conf. Ser. #94, edited by T.B. Lucatorto and J.E. Parks, 163-166 (1988). 7. N. Thonnard, R.D. Willis, M.C. Wright, WA. Davis, and B.E. Lehman, "Resonance ionization spectroscopy and the detection of 8IKr\ Nucl. Inslr. and Meth. in Phys. Res. B 29 398-406 (1987). 8. G.S. Hurst, M.G. Payne, S.D. Kramer, and J.P. Young, "Resonance ionization spectroscopy and one- atom detection", Rev. Mod. Phys. SI 767-819 (1979). 9. W.M. Fairbank, Jr., M.T. Spaar, J.E. Parks, and J.M.R. Hutchinson, "Anomalous odd- to even-mass isotope ratios in resonance ionization with broad-band lasers", Phys. Rev. A 40(4) 2195-2198 (1989). 10. P. Lambropoulos and Y. Lyras, "", Phys. Rev. A 40 2199 (1989). 11. TJ. Whitaker, B.A. Bushaw, and B.D. Cannon. "Laser-Based Techniques Improve Isotope Analysis", Laser Focus 24(2), 88-101, (1988). 12. B.D. Cannon, TJ. Whitaker, G.K. Gerke, and B.A. Bushaw. "Anomalous Linewidths and Peak Height Ratios in 137-Ba Hyperfine Lines". Appl. Phys. B 47, 201-206, (1988). 13. O.I. Matveev, N.B. Zorov, and Y.Y. Kuzyakov, J. A nalyt. Client. USSR 34 654 (1979). 14. A.A. Ganeev, O.I. Matveev, and S.E. Sholupov, J. A nalyt. Chem. USSR 43 1424 (1987). 15. N. Omenetto, B.W. Smith, and J.D. Winefordner, Spectrochim. Acta. B Special Supplement on Proceedings of the Symposium held at Pontifical Academy of Sciences 101 (1989). 16. G.A. Petrucci, R.G. Badini, and J.D. Winefordner, "Resonance detection of photons by atomic ionization", Resonance Ionization Spectroscopy 1992. Published by Institute of Physics, Bristol and Philadelphia, Conf. Ser #128, edited by CM. Miller and J.E. Parks, 333-336 (1992). -106- Proceedings of the 61I1 International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - Atomic Collisions related to Atomic Laser Isotope Separation Takemasa SHIBATA Department of Chemistry and Fuel Research Japan Atomic Energy Research Institute Tokai-mura, Ibaraki-ken, 319-11, Japan Atomic collisions are important in various places in atomic vapor laser isotope separation (AVLIS). At a vaporization zone, many atomic collisions due to high density have influence on the atomic beam characteristics such as velocity distribution and metastable states' populations at a separation zone. In the separation zone, a symmetric charge transfer between the produced ions and the neutral atoms may degrade selectivity. We have measured atomic excitation temperatures of atomic beams and symmetric charge transfer cross sections for gadolinium and neodymium. Gadolinium and neodymium are both lanthanides. Nevertheless, results for gadolinium and neodymium are very different. The gadolinium atom has one 5d electron and neodymium atom has no 5d electron. It is considered that the differences are due to existence of 5d electron. Keywords: Atomic collision, Laser isotope separation, Gadolinium, Neodymium, Atomic excitation temperature, Charge transfer 1. Introduction In atomic vapor laser isotope separation(AVLIS), one isotopic species in an atomic beam is selectively photoionized by pulse lasers and the ions are collected on electrodes. We have studied basic processes of AVLIS for gadolinium or uranium. These studies have made us recognize that atomic collisions play important roles in various places in AVLIS apparatus as shown in Fig. 1. Importance of the atomic collisions in AVLIS is summarized in Sec. 2. Recently, we have started the atomic collision studies for neodymium. The results were expected to be similar to those for gadolinium since both gadolinium and neodymium are lanthanides and the chemical properties are very similar. However, the results for the neodymium are very different - 107- from those for the gadolinium. Gadolinium atom has one 5d electron and neodymium atom has no 5d electron. The differences between gadoliniumm and neodymium are considered to be due to existence of 5d electron. Here it will be described that the existence of 5d electron plays important roles in collision processes in sections 3 and 4. 2. Roles of atomic collision processes in AVLIS 2.1 Collision processes in vaporization zone. Atomic vapor is produced by heating liquid metal such as uranium or gadolinium in a water-cooled crucible with an electron beam. Since evaporated atoms are ionized by the incident and reflected electron beams, the atomic vapor contains weakly ionized plasma[l,2,3,4]. The plasma must be removed from the atomic vapor[5] before flowing into the separation zone, otherwise enrichment factor decreases. Since atomic density is very high in the vicinity of the evaporation surface, the mean free paths of collisions between evaporated atoms are so short that the evaporated atoms have experience of many collisions which include elastic and inelastic scatterings. These collisions have influence on the atomic beam characteristics such as velocity distribution[6] and n-5x 10,2cm"3 I - 50 cm Charge Transfer 235U* + 238U - 235U + 238U+ n~1016cm-3 I - 0.025 cm Elastic Scat. U+U-.U + U Inelastic Scat, if + U - U + U + AE Ionization e + U -> e + IT Excitation e + U -. e + If Crucible Fig.l A schematic view of AVLIS apparatus and collision processes. - 108- metastable states' population[7,8,9] at the laser irradiation zone. Due to vapor expansion cooling, the atomic beam velocity at the separation zone is much higher than the thermal mean velocity at the vaporization surface temperature and has a narrow distribution. The atomic excitation temperature determined from the metastable states' population is far lower than the evaporation surface temperature due to high rate energy transfer between atomic excitation and translation in the atomic vapor expanding from the evaporation surface. The measured velocity[10,ll] is sometimes higher than the maximum velocity which can be attained by assuming that the total enthalpy containing the atomic excitation energy is adiabatically converted to the kinetic energy of the vapor motion. This fact suggests that the excitation energy of the atoms, which are excited the electron beams, contributes to the acceleration of the atomic vapor during expansion. 2.2 Collision processes in separation zone In the separation zone, the atomic density is low and the mean free path is much longer than the apparatus scale such as distance between collection electrodes. Therefore, only a small part of atoms in the separation zone collides with each other. However, scattered atoms are deposited on the collection electrodes and may degrade the enrichment factor[12,13]. The symmetric charge transfer process between the produced ions and the neutral atoms degrades the enrichment factor, since the cross section of symmetric charge transfer is much bigger than that of gas kinetics. Sputtering due to impact of the photoionized ions on the collection electrodes must also be considered in designing the separation zone. Moreover, one aimed isotope is photoionized by lasers through an autoionization level. Autoionization is a half of resonance collision of an electron with an ion. Distinct J-dependence of photoionization spectra of gadolinium atom[14] can be explained by a kind of selection rule of coupling between discrete and continuum states, where J is total angular momentum. 3. Atomic excitation temperature of gadolinium and neodyraium atomic beams The metastable states' populations in gadolinium atomic beams were measured by laser photoabsorption spectroscopy[7,8]. Moreover, metastable populations in gadolinium[9] and neodymium atomic beams were measured by the resonance photoionization. Figure 2 shows a schematic diagram of the experimental apparatus. Gadolinium or neodymium atomic beam was produced by the electron beam heating. The atoms between the collection electrodes were - 109- ionized by two-step two-photon resonance photoionization with pulse dye lasers. The gadolinium atom has five low-lying metastable levels. Figure 3 shows resonance ionization schemes which were used to measure the population distribution of three levels of 215cm"1 , 533cm"1 ,and 999cm"1 . The first laser pulse saturated the transitions between one of the metastable states and the common intermediate state of 23644cm"1. Then, the atoms in the intermediate state were ionized at the same ionization probability by the second laser pulse holding its frequency and intensity constant. Thus, the relative populations of metastable levels were obtained from the number of the photoionized ions detected by a Faraday cup. The distributions of the three levels including Ocm*1 or 1719cm'1 were also determined in the same way. The measured populations followed the Bolzmann distribution and the atomic excitation temperatures were determined. Figure 4 shows the atomic excitation temperature as a function of deposition rate monitored by a quartz crystal sensor. The excitation temperature of Gd is far lower than the evaporation surface temperature estimated from the deposition rate. -j HZ) Quart! Crystal Sensor Collection Collection Electrode 2 Electrode Y Foroooy Cups ID Laser Beam- Slit Elimination Electrode Electron Beom ^ / Crucible N Filament Fig.2 Schematic diagram of the experimental appratus. -110- -49952(AIL) •49603(IP) Fig.3 Ionization schemes to measure population distributions of three levels of 215, 533 and 999 cm-1 of gadolinium atoms. 7ir~v-,5—23644(: 0 = 4) AIL:Autoionization level IP: Ionization potential —1719(J=6) 999(J=5) 533(J=4) 215(J=3) 0(J=2) fs : Evoporotion surioce lemperoture p 2000 o L_ C5 c E CD *~ C o 9 » o *— xc i 1000 o> u E o 0 0 2 4 6 Deposition rate (nm/s) 4 Atomic excitation temperatures of gadolinium and neodymium atomic beams as functions of deposition rates monitored by a quartz crystal sensor. — Ill- The neodymium atom has five low-lying metastable levels. We measured relative populations of Ocm"'(J=4) and 1128cm"'(J=5) using the common inter mediate level of 23017cm"1 (J=4) and the autoionization level of 48533 cm"1. The excitation temperature of a Nd atomic beam obtained from the relative populations of Ocm"1 and 1128 cm"1 is also shown in Fig. 4. The evaporation surface temperature and evaporation rate were almost similar to those for Gd evaporation. However, the atomic excitation temperature of Nd was about 1400 K at high deposition rate, which is about three times higher than temperature of Gd. The gadolinium atom has ten valence electrons of 4f75d6sJ. The neodymium atom has six valence electrons of 4fJ6sJ. The collisional relaxation of gadolinium or neodymium occurs through inelastic collisions between fine structures, where the direction of spin and orbital angular momentum change. For gadolinium the direction of 6d electron easily changes by collisions and collisional relaxation easily occurs since 6d electron is slightly shielded by two 6s electrons. On the other hand, the neodymium atom has no 6d electron and SI states are formed by four 4f electrons which are strongly shielded by two 6s-electrons, because the orbit radius of the 4f-electrons is very short. Therefore, the direction of the spin and the orbital angular momentum is hard to change during collisions and collisional relaxation does not easily occurs. Thus, high collisional relaxation of gadolinium is considered to be due to existence of 5d electron. 4. Charge transfer cross sections of gadolinium and neodymium Since AVLIS was proposed about 20 years ago, it has been recognized that symmetric charge transfer is an important process. However, there are few experimental data for atoms of transition elements. We measured the cross sections of gadolinium[15] and neodymium[16] from the ratio between the number of product ions and that of primary ions extracted from a laser ion source. The experimental procedures are described in detail elsewhere[15,16]. Figure 5 shows obtained cross sections of gadolinium and neodymium. The theoretical prediction of charge transfer cross section of uranium by Sinha and Bardsley[17] is also shown. The measured cross section of gadolinium roughly agrees with theoretical prediction for uranium of which ionization potential is almost the same as gadolinium. However, the cross section of neodymium is about two times of that of Gd and impact energy dependence is also different. Usually symmetric charge transfer is resonant. However, for gadolinium or neodymium, not only resonant but also near-resonant charge transfer may -112- take HJ.ace, since atoms or ions of gadolinium or neodymium have low-lying raetastable levels. For the charge transfer between Gd(4f75d6s2) and GdM4f75d6s) which are in the ground states, two following reaction paths may take place depending on the state of spin of transferring 6s electron; 9 ,0 ,D ? Gd( Dj) +GdM D5'2H Gd'( Ds,2) +Gd( D2) + Ocm'' s 9 •* Gd'( D?-2) + Gd( D?) -2857cm" 1 The first and second reaction paths are resonant and near-resonant, respectively. For Nd(4f"6s£) and Nd* (4f"6s) in the ground states, following resonant and near-resonant reaction paths may also take place: Nd(5I;) + Nd'((5I06s,J=7/2) •» Nd' ((5I<)6s,J=7/2) +Nd(5I<) + 0 cm"' •» Nd4 ((6l4)6s,J=9/2) + Hd(5I«) - 513 cm'1 For gadolinium, near-resonant path does not occur in the ion impact energy range of 100-1000 eV, because the energy difference is about 0.3 eV. For neodymiura, on the other hand, energy difference in near-resonant reaction path is only 0.06eV. Therefore, it is considered that this near-resonant reaction occurs in the impact energy region of measurement and charge transfer cross section of Nd is bigger than that of Gd. ,-13 )o' 1 1 j , 1 1.1 - E : ,0. NdHNd c .2 "o CD - OT W CD • Il l 1 c + u + u ~~*-—-^_ ' GdHGd- (Sinha and Bardsley) CD ro - o 1 1 1 1 1 1 1 . , 1 1 t 1 1 T 1 1 1 1 10' 10J impact ion energy (eV) Fig.5 Measured cross sections for symmetrical charge transfer for gadolinium and neodymium as a function of the impact energy. The theoretical prediction of charge transfer cross section for uranium by Sinha and Bardsley[20] is also shown. - 113 — The Gd* and Nd* ion beam used for cross section measurement were produced by 2-step-l-wavelength resonance photoionization. From used photoionization schemes, produced Gd* and Nd* is considered to be 2857 cnr' 10 1 5 ( D?,j) and 1470 cnr (( Is }6s,J=9/2) or 1650cm"'((*I5 )6s,J=ll/2), respectively. The atomic beams used for measurement contain low-lying roetastable states as described in a previous section. Here, all possible collision paths of gadolinium and neodymium charge transfer were considered. Each cross section of possible reaction paths was calculated using the universal curve proposed by 01son[18]. Figure 6 shows total effective cross sections obtained by summing up the products of the cross section and the probability for each reaction path. As expected, the cross section of neodymium is bigger than that of Gd and energy dependence are very similar to experimental ones shown in Fig. 5, though calculated values are about 1.5 times smaller than the measured values. The energy difference between the Gd' levels of ,0D5.-2 and eD?.-2 is determined by the coupling between 5d-electron and 6s electron of Gd'. On the other hand, the energy difference between the Nd* levels of ((*Ia }6s, J=7/2) and ((5L. )6s, J=9/2) is only 0.06 eV, since Nd4 has no 5d electron and the coupling between 4f-electrons and 6s electron is small. This small energy difference is the reason why charge transfer cross section of neodymium is bigger. Therefore, the existence of d-electron is important in the case of charge transfer, also. E c .o "u a> V) V) o u u CD W ID'14 c + ro G d + G d CD D> ro -! L_l_ 102 10J impact ion energy (eV) Fig.6 Calculated total effective cross sections for symmetrical charge transfer for gadolinium and neodymium. - 114 — 4. Conclusion It is briefly reviewed that atomic collisions are important in various places in AVLIS. We have measured atomic excitation temperature of atomic beams and charge transfer cross sections of gadolinium and neodymium. The atomic excitation temperature of the gadolinium atomic beam was 500 K, which is far lower than that of the evaporation surface. On the other hand, the atomic excitation temperature of neodymium atomic beam was about 1400 K, though the evaporation surface temperature was almost similar to that for gadolinium evaporation. Symmetric charge transfer cross sections for gadolinium and neodymium were obtained in the impact energy range of 100-1000 eV. The cross section of neodyraium is about two times of that of gadolinium and impact energy dependence is also different. The gadolinium atom has one 5d electron and the neodymiura atom has no 5d electron. These differences between gadolinium and neodymium are explained also due to existence of 5d electron in the gadolinium atom or ion. Uranium atom has one 6d electron in six valence electrons of 5f36d7s2. therefore, we can expect that the atomic collisions of uranium have the properties similar to those of gadolinium as for the atomic excitation temperature of the atomic beam and the symmetric charge transfer. References [l]H.0hba, T.Arisawa, A.Nishimura, K.Ogura, and T.Shibata, J.Vac.Soc.Jpn. (Shinku), 35(1992)282(in Japanese). [2]H.0hba, A.Nishimura, and T.Shibata, Jpn.J.Appl.Phys., 32(1993)5759. [3]H.0hba, K.Ogura, and T.Shibata, J.Vac.Soc.Jpn.(Shinku), 35(1992)1001(in Japanese). [4]T.Shibata, H.Ohba, and K.Ogura, J.Vac.Soc.Jpn.(Shinku), 36(1993)928(in Japanese). [5]H.0hba, A.Nishimura, K.Ogura, and T.Shibata, Rev.Sci.Instrum.,65(1994)657. [6]A.Nishimura, H.Ohba, and T.Shibata, J.Nucl.Sci.Technol., 29(1992)1054. [7]A.Nishimura, T.Arisawa, K.Ogura, H.Ohba, and T.Shibata, J.Vac.Soc.Jpn. (Shinku), 34(1991)639 (in Japanese). [8]A.Nishimura, T.Arisawa, H.Ohba, and T.Shibata, J.Vac.Sci.Technol., All (1993)1516. [9]K.0gura and T.Shibata, J.Phys.Soc.Jpn., 63(1994)834. [10]A.Nishimura, T.Arisawa, H.Ohba, K.Ogura, and T.Shibata, J.Vac.Soc.Jpn., -115- (Shinku), 35(1992)637 (in Japanese). [ll]K.Ogura and T.Shibata, J.Vac.Soc.Jpn.(Shinku), 36(1993)207 (in Japanese). [12]H.Kaburaki, A.Nishimura, H.Ohba, H.Yamamoto, and T.Shibata, J.Vac.Soc. Jpn. (Shinku), 3_4(1991)718 (in Japanese). [13]A.Nishimura, H.Ohba, and T.Shibata, JAERI-M 93-102(1993)(in Japanese). [14]K.0gura, T.Arisawa, and T.Shibata, JAERI-M 91-221(1992)(in Japanese). [15]K.Ogura and T.Shibata, J.Mass Spectrom.Soc.Jpn.(Shitsuryou Bunseki), 41(1993)37. [16]K.Ogura and T.Shibata, This proceedings. [17]S.Sinha and J.N.Bardsley, Phys.Rev., A14(1976)104. [18]R.E.Olson, Phys.Rev., A6(1972)1822. — 116- Proceedings of Ihc 6lh International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - ULTRASHORT-PULSE LASERS AND THEIR APPLICATIONS GCrard A. MOUROU Center for Ultrafast Optical Science University of Michigan 1006 l.S.T. Bldg., 2200 Bonistccl Blvd. Ann Arbor, MI 48109-2099 USA A revolution has occurred over the past ten years in our ability to generate, manipulate, and amplify ultrashort pulses. Laser pulses can be as short as a few femtoseconds (few optical cycles) and possess extreme power up to several tcrawatts. The applications of these pulses arc numerous in physics, chemistry, and biology, where they can be used to time resolve ultrafast events. We review in this article the state of the art in short pulse generation and amplification. Keywords: Ultrahigh-peak-power lasers, Ultrashort-pulse lasers, Chirped pulse amplification 1. INTRODUCTION Over the past few years a revolution has occurred in our ability to produce extremely high-power and high-intensity pulses. Whereas ten years ago a tablctop system with a beam size on the order of a centimeter squared could typically produce gigawatts the same size system can now produce 1000 to 10,000 times this peak power. Today the laser power is well into the tcrawatl regime with focused intensity greater than 1018 W/cm2. This vast improvement in laser peak power and intensity is the direct result of the technique of Chirped Pulse Amplification (CPA).1,2 Figure 1 describes the evolution of laser peak power since its inception. The different power jumps correspond to Q switching, mode locking, and finally CPA. Note that with CPA a dramatic improvement of three to four orders of magnitude has been obtained and should permit in the near future the generation of pulses limited by their theoretical peak power. This technique makes possible efficient energy extraction from superior energy storage materials by short pulses, without inducing undesired nonlinear effects in the amplifier system. This technique has been applied to (1) a wide variety of amplifying materials, i.e., Nd:glass,1-3 alexandrite,6 Ti:sapphirc,7-9 and LhSAF,10'11 with good energy storage characteristics, and (2) existing, large-scale Nd:glass systems, built for nanasccond-pulsc amplification as at the CEA Limcil in France12'13 or at Osaka in Japan.14 Record peak power of 55 TW (25 joules, 400 fs) were obtained at CEA Limeil. More ambitious performances arc contemplated at Lawrence Livermorc National Laboratory (LLNL), where the construction of a pctawatl (1 KJ/ps) system is underway. CPA had not only a major impact on the peak power of short-pulse systems, but also a similar one on their average power. As illustrated in Fig. 2, before CPA, short-pulse amplification systems had an average power of typically 10 mW. With CPA their average power is now on the order of 1 W, a factor- — 117- of-100 improvement. Here again the average power limit has not been reached and we expect that in the near future femtosecond amplification systems will produce an average power at the 10-W level. High average power is essential for high harmonic generation,15 ulirafast spectroscopy,lh medical17'18 and imaging19 applications. The production of a high-pcak-powcr pulse is a necessary but not sufficient condition to producing a high-irradiancc pulse on target. The brightness and ultimate focused irradiancc achievable with a laser pulse is determined by both the peak power and the spatial quality (divergence) of the pulse. By paying careful attention lo both linear and nonlinear aberrations in the laser and beam transport systems, it is now possible to produce nearly diffraction-limited, mullitcrawalt pulses. When focused, such pulses can achieve irradiancc between 1018 and 1019 W/cm2. Figure 3 illustrates the dramatic impact of CPA on focused intensity over the years. Scaling these systems up to the 100-TVV or even pctawatt (1000-TW) level will enable the study of laser-matter interaction at 1021 W/cm2and beyond. As the pulse duration decreases from the nanosecond to the picosecond and femtosecond regime it becomes increasingly difficult to extract the stored energy without causing unwanted nonlinear effects. This stems from the fact that the input flucncc (J/cm2) necessary to extract the optical stored energy has to be of the order of the saturation flucncc, Fs= hv/o, where h is the Planck constant, v the laser frequency and a the emission cross section. Trying lo extract the energy with picosecond-femtosecond pulses leads to wave-front distortion and filamcntalion20 due to the intensity-dependent index of refraction n = na+ ntf, where I is the beam intensity. The wave-front distortion, B, is equal to n2l elz (1) o To keep a diffraction-limited beam the value of B should not exceed 0.6, corresponding to wave-front distortion of TJT.21 The conditions for energy extraction and beam propagation can be simultaneously satisfied only with mediocre amplifying media possessing large emission cross sections on the order of 10" ^cm 2. These media, like dyes or excimers, have poor energy storage characteristics and require, for picosecond-femtosecond pulses, an input fluencc of the order of a mJ/cm2to extract their stored energy. 2 For femtosecond-picosecond pulses, this corresponds to an intensity /vof the order of a GW/cm , that is, below the threshold set by optical nonlincaritics. Consequently, before the advent of CPA they were the only choices possible for direct amplification of picosecond-femtosecond pulses. To reduce the size of laser systems, one wants to use good energy storage media, such as Nd:glass, alexandrite, Ti:sapphirc, and Li:SAF. These media have an emission cross section a ~ lO'^-lO"20 cm2 a thousand times smaller lhan dyes and excimcrs. Consequently they have a thousand limes better energy storage characteristics. However, these large saturation flucnccs, typically of a few joules, leads to extremely high saturation intensities /,.= FJT( T is the pulse duration), in the TW/cm2 range for pulses in - 118 — the picosecond-femtosecond regime. The B associated with these large intensities is a thousand times- above the recommended level, forbidding direct short-pulse amplification. To get around this seemingly insurmountable problem, that is, to keep the input flucncc as high as possible while maintaining the pulse intensity to the lowest level, the pulse is stretched by a large amount (100 to 10,000 x) prior to amplification. It is then amplified by a factor of 106-1010 and rccomprcsscd to its initial value. This technique was first demonstrated in 1985 and was called Chirped Pulse Amplification. 2. THE CHIRPED PULSE AMPLIFICATION TECHNIQUE The CPA technique in the optical regime involves some impressive manipulations (Fig. 4). First, a very short (ps-fs) and clean (over six to ten orders of magnitude) pulse is generated at the subnanojoulc level. Second, this pulse is stretched by a factor up to 104. Third, it is amplified by 10 to 11 orders of magnitude to the joule level. Fourth, it is rccomprcsscd by a factor of H)4 to its initial value. The end result must be a diffraction-limited pulse with an extremely high contrast ratio. These two attributes arc absolutely required to perform quality experiments at an intensity level greater than 1018 W/cm2. We will now describe the key elements that had to be developed over the past eight years which made Chirped Pulse Amplification the technique of choice for ultrashorl-pulsc amplification today. A. Generation of Short and Ultraclcan Pulses It is now possible to produce ultrashort pulses, using the Kcrr-lcns effect in Ti:sapphirc oscillators22 10-100-fs pulses arc routinely produced. Although these pulses arc extremely short, they have the tendency to exhibit a tail extending over a few hundred femtoseconds before the peak of the pulse. If this pedestal is greater than 1010W/cm2, it can produce a plasma on the target before the main pulse arrival, leading to questionable results. Before stretching, it can be necessary to further clean the oscillator pulse in a pulse cleaner. A high-contrast pulse cleaner has been demonstrated. It uses cross-phase modulation in a single-mode fiber located between two cross polarizers. A contrast enhancement of six orders of magnitude23 can be obtained. B. Stretching and Compression CPA demands the largest possible stretching/compression ratio to efficiently extract the stored energy without causing undesirable nonlinear effects, wave-front distortion, filamentation, and optical damage in the optical system. The first CPA] demonstration used the group velocity dispersion in fiber to stretch the pulse. The recompression was performed by using the negative group velocity dispersion provided by a diffraction grating pair. These stretcher/compressors arc not matched over all orders due mainly to the large third-order dispersion in the grating pair.24 The maximum compression ratio, R, that wc can produce is of the order of -r- where AX is the pulse bandwidth and "k is the laser wavelength. When R exceeds this value, incomplete recompression is pcrfomicd producing unwanted ripples before or — 119- after the main pulse. As an example, a picosecond pulse at 1.06 (.tin with a A\ of 20 A cannot be stretched and compressed by a factor R greater than 500. It quickly worsens for shorter pulses in the 100-fs range where R can not exceed 50. Using this embodiment, CPA seemed limited to R of the order ofl 00. In 1987, for optical communication applications, O. Martinez proposed a compressor with positive group velocity dispersion to compress pulses at 1.5 (im.25 In this regime, the fiber has negative group velocity dispersion and the pulses exhibit a negative going chirp. It is then necessary to rccomprcss the pulses with a device with positive group delay dispersion. To perform this operation Martinez proposed a compressor with positive group delay dispersion. This compressor is composed of a telescope with magnification of one between two antiparallcl diffraction gratings. It was recognized and demonstrated by M. Pcssot a a/.26 that this compressor at 1.5 |im could be, for wavelengths in the visible, a matched stretcher over all orders of a Tracy's compressor. Pcssot et al. ^ stretched and compressed an 80-fs pulse with R greater than 1000 without introducing any temporal distortion in the pulse. The discovery of this matched stretcher/compressor was a real breakthrough. It has been the key to ultrahigh-pcak-powcr generation and now equips all the Chirped Pulse Amplification systems. Extensive work is being done by a number of groups27 to further improve the stretching/compression ratio to 104 to 105. This very large ratio will be needed to produce pulses limited to their theoretical peak power. The high-/? stretcher/compressors arc based on low groove densities, i.e., 300 to 12(X) per mm, and refleciivc optics. C. Gain Narrowing Short pulses have a large Fourier spectrum. Because of the large overall gain involved in CPA, typically 1010 in Ihc small signal regime, Ihc pul.se .spectrum is reduced to a value ^^=^^m^r (2) where Acoa is the material gain bandwidth and Gdb((.oa) the gain. From this expression we sec how important it is to work with materials having a large gain bandwidth, such as Nd:glasscs, alexandrite, Ti:sapphirc, LiSAF, or a combination of those. A reduction in gain bandwidth has two undesirable effects. First, it reduces the stretched pulse duration, reducing the amount of cxiractable energy. Second, a narrower spectrum will lead to a longer compressed pulse. In first approximation ihc system output power is reduced according to the square of the gain narrowing. 3. THEORETICAL PEAK POWER Although formidable peak power well in the terawatt regime has already been obtained, it is interesting to note that we arc far from the theoretical limit. This limit can be estimated as the ratio of the maximum cxtractablc energy given by Fs, ihc saturation llucnce, over the minimum pulse duration imposed by the — 120 — material gain bandwidth. The peak power limit Ah per unit area, assuming a Au.T = .5 can be estimated as: This power also represents thai needed to produce a Rabi oscillation in the amplifying medium. The theoretical peak powers for different materials arc estimated in the following Tabic 1. As mentioned above, the generation of this power will require a stretching/compression ratio in the 105 regime. We can predict that in the near future 100-TW pulses will be achievable with very compact systems. It is also conceivable to think about cxawalt (1()18-W) pulses with largc-apcrturc (1-m) systems based on mixed Nd:glasscs. Table I: Theoretical peak powers Laser Type Cross Section A\ T ^tl. K)-20 cm 2 (nm) (fs) (TW/cm2) Nd:Glass Phosphate 4 22 80 60 Nd:Glass Silicate 2.3 28 60 100 Nd:Glass Combination 1.5 60 30 4000 Ti:Sapphirc 0.3 120 8 120 Alexandrite 1 100 10 2000 GvLiSAF 3 50 15 300 In conclusion, with chirped pulse amplification, the field of optical science is entering uncharted territory in physics. The intensities that can be generated arc such that during the interaction the electrons arc rclativistic, and x-ray radiation and formidable pressures due to pondcromotivc forces as high as gigabars can be produced. These ultrahigh-pcak-powcr lasers will undoubtedly have a profound impact on laser science. -121 — REFERENCES 1) Strickland D. and MourouG.: Opt. Commun. 56,219(1985). 2) Maine P., Strickland D., Bado P., Pcssot M., and Mourou G.: IEEE J. Quantum Electron. OE-24. 398 (1988). 3) Maine P. and Mourou G.: Opl. Lett. £, 467 (1988). 4) Fcrray M„ Lomprc L. A., GobcrtO., L'Huillicr A., Mainfray G., Manus C, and Sanchez A.: Opl. Commun. 75, 278 (1990). 5) Perry M. D., Patterson F. G., and Weston J.: Opt. Lett. 15, 381 (1991); Patterson F. G. and Perry M. D.: J. Opl. Soc.Amcr.Bft 2384 (1991). 6) Pcssot M., Squicr J.,Bado P., Mourou G., and Hartcr D.: IEEE J. Quantum Electron. 25,61 (1989); Pcssot M., Squicr J., Mourou G., and Hartcr D.: Opt. Lett. J4, 797 (1989). 7) Kmctcc J., Macklin J. J..and Young J. F.: Opt. Lett. J6, 1001 (1991). 8) Sullivan A., Hamster H., Kaptcyn H. C, Gordon S., White W., Nathcl H., Blair R. J., and Falcone R.W.: Opt. Lett. J5,1406(1991). 9) Dilmirc T. and Perry M.D.: Opl. Lcit. 18,426 (1993). 10) Bcaud P., Richardson M., Micsak E., and Chai B. T.: Opt. Lett. J8,1550 (1993). 11) White W. E., Hunter J. R., Van Wocrkom L., Ditmirc T., and Perry M. D.: Opt. Lett. 17,1067 (1992). 12) Sautcrct C, Husson D., Thicll G., Sczncc S., Gary S.. Migus A., and Mourou G.: Opt. Lett. J6,238 (1991). 13) Rouycr C, Mazataud E., Allais I., Pierre A., Sczncc S., Saulcrci C., Mourou G., and Migus A.: Opt. Lett. 18, 214 (1993). 14) Yamakawa K., Shiraga H., and Kato Y.: Opt. Lett. 16, 1593 (1991). 15) Li K., L'Huillicr A., Fcrray M., Lomprc L. A., and Mainfray G.: Phys. Rev. A 29,5751 (1989). 16) The reader is referred to Martin J.-L., Migus A., Mourou G., and Zcwail A. H., cds.: "Ultrafasl Phenomena VIII," Springer-Vcrlag, Berlin (1993). 17) Zyssct B., Fujimoto J. G., Puliafito C. A., Birngrubcr R., and Dculsch T. F.: Laser in Surgery and Medicine 9, 193 (1989). 18) Ho P. P., Wang L., Liang X., Galland P., Kalpaxis L. L., and Alfano R. R.: Optics and Photonics News 4,23 (1993) and references therein; L. Wang el al.. Science 253.769 (1991) and references therein. 19) Lcith E., Chen C, Chen H„ Chen Y., Dilworth D., Lopez J., Rudd J., Sun P. C, Valdmanis J., and Vosslcr G.: J. Opt. Soc. Am. A 2,1148(1992). 20) Shcn Y. R.: "Principles of Nonlinear Optics," and references therein, Wiley Intcrscicncc, New York (1984); Kocchncr W.: "Solid-State Laser Engineering, Third Edition," Springcr-Vcrlag, New York (1990). 21) Born M. and Wolf E.: "Principles of Optics," Pcrgamon, Oxford (1970). 22) Spcncc D. E., Kcan P. N., and Sibbctt W.: Opt. Lett. 16, 42 (1991). 23) TapidJ.L.and MourouG.: Opt. Lett. 17,136(1992). 24) Trcacy E. B.: IEEE J. Quantum Electron. OE-5.454 (1969). — 122 — 25) Martinez O. E.: IEEE J. Quantum Electron. OE-23.1385 (1987). 26) Pcssot M.: Opt. Commun. fi2,419 (1987). 27) Lcmoff B. E. and Barty C. P.: Opt. LcU. IS, 1651 (1993); Zhou J., Huang C.-P., Shi C, Mumanc M., and Kaptcyn H. C: Opt. Lett. 1&24 (1993). — 123- FIGURES theoretical limit pw TW chirped pulse amplification ,_ GW 3CD • mode locking o Q. MW ^ ' Q switching KW \ free-running 1960 1970 1980 1990 FIG. 1. Evolution of Laser Peak Power CD c CD 10-12 0.001 0.01 0.1 1 10 100 103 104 IQ5 106 107 108 repetition rate (Hz) > FIG. 2. Average Power of Femtosecond Lasers -124- 1020 10i5 _ T3 s o 10'0 1960 1970 1980 1990 FIG. 3. History of Laser-Matter Interaction 1*• oscillator stretcher amplifier compressor FIG. 4. Chirped Pulse Amplification — 125 — Proceedings ofilic 6ih International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - Infrared Multiphoton Dissociation Processes of Some Halogenated Hydrocarbons A. Yokoyama, K. Yokoyama, G. Fujisawa and T. Takayanagi Advanced Science Research Center, Japan Atomic Energy Research Institute Tokai-mura, Ibaraki-ken, 319-11, JAPAN We review our studies on mechanisms and dynamics of infrared multiphoton dissociation of CF3CHCIF, C2HC13) CBrF2CHClF, CBrF2CHBrF, CBrF2CBrClF, and c- C4F8 using a photofragmentation translational spectroscopy and ab initio MO calculations. In the case of CF3CHCIF, three-centered HCl elimination and C-Cl bond rupture occurred competitively as primary dissociation processes. The center-of-mass translational energy distribution of the products produced by the HCl elimination indicates that an exit barrier of several kcal/mol exists in the potential energy surface of the reaction. In the case of C2HCl3, three- and four-centered eliminations of HCl occurred competitively. Halogenated ethanes containing bromine atoms dissociated through a C-Br bond rupture channel. The halogenated ethyl radicals produced by the C-Br bond ruptures subsequently dissociated through a C- halogen bond rupture channel. The translational energy distributions for the C-halogen bond ruptures observed are almost consistent with those calculated by RRKM theory. In the case of c-C4Fg, dissociation of C-C4F8 to two C2F4 molecules was observed as a primary process. About 30% of the exit barrier for this reaction is converted to the translational energy of the products. This indicates that the dissociation proceeds through a loose transition state. This conclusion is supported by the transition state structure calculated by an ab initio MO method. Keywords : Infrared multiphoton dissociation, photofragmentation translational spectroscopy, Halogenated hydrocarbon, C02 laser 1. INTRODUCTION Infrared multiphoton dissociation (IRMPD) processes are important in the fields of both basic science and application to isotope separation. Lee and his co workers [1] have elucidated that the process occur under collision free condition by using a crossed laser-molecular beams method. This characteristic of IRMPD allows the study of unimolecular dissociation dynamics by measuring translational and/or internal energy distributions of nascent dissociation products under collision free conditions. A photofragmentation translational spectroscopy (PTS) is a powerful experimental technique for studying photodissociation mechanisms and dynamics of systems undergoing complex dissociation pathways including several primary and secondary dissociation channels [2]. We have studied mechanisms and dynamics of IRMPD of halogenated hydrocarbons by using the PTS. In this symposium, we — 126- present our studies on 1RMPD of 2-chloro-1,1,1,2-tetrafluoroethane (CF3CHCIF) [3], trichloroethene (CHCICCI2), l-bromo-2-chloro-l,l,2-trifluoroethane (CBrF2CHClF), 1,2-dibromotrifluoroethane (CBrF2CHBrF), l,2-dibromo-2-chlorotrifluoroethane (CBrF2CBrClF) and octafluorocyclobutane (C-C4F8). 2. EXPERIMENT A molecular beam apparatus used in this work has been described in detail previously [3]. Halogenated hydrocarbon beams except CF3CHCIF and C-C4F8 were produced by expanding a halogenated hydrocarbon/Ar mixture into a source chamber through a 0.1 mm nozzle. The CF3CHCIF and C-C4F8 beams were produced by expanding neat gases. The nozzle was heated to 230 - 320 °C to prevent from the formation of clusters and to enhance dissociation probabilities of molecules. The halogenated hydrocarbon/Ar mixtures were prepared by bubbling Ar through halogenated hydrocarbons at desired temperatures. The molecular beam was collimated by passing through two skimmers in differential pumping regions. The beam velocity distribution f(v) is approximated by a formula: 2 f(v) = Av2exp[-(v/a -S) ], where A is a normalization constant. The parameters a and S are tabulated in Table 1 along with beam conditions and laser wavelengths used for irradiation of the samples. The molecular beam was crossed perpendicularly with a laser beam. The laser beam was focused to 3x3 or 2x2 mm2 spots at the crossing point with a ZnSe lens. A TEA CO2 laser (Lumonics TEA-841) with a temporal pulse of a 100 ns spike followed by a 5 [is tail was used as the light source and operated at 15 Hz. Dissociation products were detected at beam-to-detector angles of 10° - 70° by a triply differential-pumped quadrupole mass spectrometer (EXTREL C-50) located 44.0 cm away from the interaction region of the laser and the molecular beam. 3. RESULTS AND DISCUSSION A. IRMPDofCF3CHClFandCHClCCl2 Most of data for IRMPD of CF3CHCIF were taken at laser fluence of 19 J/cm2. Signals of dissociation products were observed at m/e=31 (CF+), 32 (CHF+), + + + + 35 (C1+), 36 (HC1 ), 50 (CF2 ), 51 (CHF2 ), 69 (CF3+) and 82 (C2HF3 ). From the mass spectra of the dissociation products, it was found that CF3CHCIF molecules dissociate competitively through C-Cl bond rupture and three-centered HCl elimination reactions: CF3CHCIF -» CF3CHF + CI (1) -» CF3CF +HC1 (2) As shown in Fig. 1, dissociation products, HCl and CI, from these reactions were - 127— observed in the time-of-flight (TOF) spectra at m/e=36 and 35, respectively. The CF3CHF radical from reaction (1) was observed at m/e=31, 32, 50, 51, 69 and 82 (C2HF3+). At 19 J/crn^, the CF3CHF radicals from reaction (1) dissociated to two CF2 molecules, which were observed at m/e=31 and 50, by a secondary dissociation: CF3CHF -» CF2 + CF2 (3) Since the CF2 molecules were not observed at laser fluence of 3 J/cm^, this secondary process followed multiphoton absorption of the CF3CHF radicals. The branching ratio of reactions (1) to (2), which is defined as the ratio of the concentration of CI to that of HCl, is determined to be 0.36 from the relative concentrations of CI and HCl. The average excitation energy of the dissociating CF3CHCIF was estimated to be about 99 kcal/mol by comparing the center-of-mass translational energy distribution for reaction (1) to those calculated by Rice-Ramsperger-Kassel-Marcus (RRKM) theory as shown in Fig.2. The translational energy distribution for reaction (2), which is shown in Fig. 3, peaks at about 3 kcal/mol. This indicates that there exists an exit barrier of several kcal/mol on the potential energy surface of the reaction. Trichloroethene molecules were irradiated at laser fluence of 50 J/cm^. Signals of fragments were observed at m/e=24 (C2+), 35 (C1+), 36 (HC1+), 47 (CC1+), + 48 (CHC1+), 59 (C2C1+), 60 (C2HC1+) and 94 (C2Cl2 ). Since C2HCl2 was detected at m/e=48 and 60, it is clear that a primary dissociation channel is the following C-Cl bond rupture reaction: CHCICCI2 -» C2HCI2 + CI (4) .Average excitation energy of the dissociating CHCICO2 molecules is estimated to be about 4 kcal/mol above the critical energy (84 kcal/mol) of the C-Cl bond rupture by comparing observed translational energy distribution with those calculated by RRKM theory as shown in Fig. 4. Since HCl and C2CI2 were detected at m/e=36 and 94, respectively, HCl elimination also occurs competitively with reaction (4). TOF spectrum at m/e=36 in Fig. 5 shows a broad peak which consists of two components. These components come from slow and fast HCl molecules from different reaction channels. The translational energy distributions for these reaction channels are shown in Fig. 6. In the case of HCl elimination from CHCICCI2, the following three- and four-centered HCl eliminations can occur: CHCICCI2 -» :CCC12 + HCl (5) -* CC1=CC1 + HCl (6) Typically, there exists an exit barrier of several tens kcal/mol for the four-centered HCl elimination and the translational energy distributions peak at several kcal/mol [4]. On the other hand, the translational energy distribution for the three-centered HCl elimination from CF2CHCI peaks at 0 kcal/mol [4]. Therefore, the slow and fast components are reasonably ascribed to the three- and four-centered HCl eliminations, -128— respectively. It should be noted that the translational energy distributions for the three- centered HC1 elimination from the halogenated ethenes are very similar to those for C- halogen bond ruptures with no exit barrier. This seems to indicate that the exit barrier for the three-centered HC1 eliminations form the halogenated ethenes is very small in contrast with that for halogenated alkanes such as CHCIF2 (ref. 4) and CF3CHCIF described above. B. IRMPD of CBrF2CHClF, CBrF2CHBrF and CBrF2CBrClF All of these molecules were irradiated at laser fluences of 2 - 14 J/cm2. At low fluences of 2 -3 J/cm^, the following C-Br bond ruptures occurred as primary processes: CBrF2CHClF-> CF2CHC1F + Br (7a) CBrF2CHBrF-> C2HBrF3+Br (8a) CBrF2CBrClF^ C2BrClF3+Br (9a) When laser fluences were increased to about 14 J/cm^, halogenated ethyl radicals subsequently dissociated through the following C-halogen bond rupture channels: CF2CHC1F -» CF2CHF + C1 (7b) C2HBrF3 -> CF2CHF + Br (8b) C2BrClF3 -» CF2CC1F+Br (9b) In the case of CBrF2CBrClF, C-C bond rupture of CF2CC1F from reaction (9b) also occurred: CF2CC1F -* CF2 + CC1F (9c) Translational energy distributions for all the primary and secondary C-halogen bond ruptures peak at around 0 kcal/mol and decrease with increase of translational energy. These distributions are essentially consistent with those calculated by RRKM theory and indicate that there exists no or very small exit barrier for the C-halogen bond ruptures of both halogenated ethanes and ethyl radicals. Average excitation energies of the dissociating molecules at the low laser fluences are estimated from RRKM theory and tabulated in Table 2 along with the average translational energies and dissociation energies. Although average internal energies of the halogenated ethyl radicals exceed the C-halogen bond dissociation energies, the excess energies are not sufficient for producing me products with the average translational energies observed. Therefore, multiphoton absorption of the radicals should have occurred before dissociation. C. IRMPD of c-C4F8 Octafluorocyclobutane was irradiated at laser fluence of 17 J/cm^. Signals of + + + dissociation products were observed at m/e=31 (CF ), 50 (CF2 ), 81(C2F3H ) and — 129 — 100 (C2F4+). Since TOF spectra at all m/e shows the same peak, it is obvious that the dissociation product is C2F4 from the following reaction: c-C4F8 -» 2C2F4 (10) Translational energy distribution in Fig. 7 was determined by fittings of TOF spectra (Fig. 8) and a product angular distribution (Fig. 9). The distribution is similar to those for the three-centered HC1 elimination from CF3CHCIF and the four-centered HC1 elimination from CHCICCI2 described above. When there is an exit barrier on a potential energy surface of a reaction, a part of the barrier energy is released as translational energy of products. The fraction of the exit barrier released as the translational energy depends on the dynamics beyond the transition state. The fractions for the reactions observed in our studies are listed in Table 3 along with the fraction for the four-centered HC1 elimination from CH3CCI3 [4] and for the retro- Diels-Alder reaction of cyclohexene, which is the dissociation of cyclohexene to ethene and butadiene [5], We choose peak energy of translational energy distribution as the measure of the amount of the exit barrier released as translational energy, because the peak energy of the translational energy distribution does not change appreciably with the total excitation energy in the case of the dissociation of relatively large molecules [5], The fraction for the dissociation of C-C4F8 is small compared with that for the retro-Diels-Alder reaction of c-CgHjo. This is because the dissociation occurs through the loose transition state shown in Fig. 10 and asymmetric breaking of two C-C bonds causes rotational and vibrational excitation of the products. The fraction for the four-centered HC1 elimination from CHCICCI2 is the smallest. At the transition state the C-Cl bond extends long, while the C-H bond length (1.28A) is not so different from that of the parent molecule (1.08A) as shown in Fig. 11. Therefore, it is expected that potential energy is largely released as the vibrational and rotational energies of HC1 and the partitioning to the relative translation of the products is small. REFERENCES 1) M. J. Coggiola, P. A. Schultz, Y. T. Lee and Y. R. Shen, Phys. Rev. Letters 38, 17 (1977) 2) A. Yokoyama, X. Zhao, E. J. Hintsa, R. E. Continetti and Y. T. Lee, J. Chem. Phys. 92,4222(1990) 3) A. Yokoyama, K. Yokoyama and G. Fujisawa, J. Chem. Phys. (in press) 4) Aa. S. Sudbo, P. A. Schultz, Y. R. Shen and Y. T. Lee, J. Chem. Phys. 69, 2312 (1978) 5) X. Zhao, R. E. Continetti, A. Yokoyama, E. J. Hintsa and Y. T. Lee, J. Chem. Phys. 91,4118(1989) — 130 — 1.0 c 3 u reaction (1) reaction (2) PL! JOO 600 900 I20O 1500 Translational Energy (kcal/mol) Translational Energy (kcal/mol) Fig. 1 TOF spectra for dissociation Fig. 2 Translational energy distribution of CF3CHCIF for the C-Cl bond rupture of CF3CHCIF 1.0 1.0 0.8 -/ \ ^^^ 1 \ B | \ = 0.6 -1 \ ja • 1 \ u I \ 03 1 \ & 0.4 \ a- \ 0.2 n n 1 V 0 5 10 15 20 2 4 6 8 10 Translational Energy (kcal/mol) Translational Energy (kcal/mol) Fig. 3 Translational energy distribution Fig. 4 Translational energy distribution for the HC1 elimination from for the C-Cl bond rupture of -131- CF3CHC1F CHC1CC12 1.0 reaction(5) 08 cilcolatcd — — *reaction(6) mction(3) 0.6 mction(6) 1.0 0.4 0.2 \ y *" ~~ "** ^ •" 03 II 1 t i 11 ° 2 « £• o.< -1 t 1 t 1 1 I I II t ' \ X ' \ X 1 \V * ^ 500 1000 Flight time (ic) Translational energy (kcal/mol) Fig. 5 TOF spectra for the HCl elimination Fig. 6 Translational energy distribution form CHCICCI2 for the HCl eliminations 0 M0 <00 900 1200 1500 0 300 (00 900 1200 1500 Flight Time (us) 0 5 10 15 20 25 30 35 Translational Energy (kcal/mol) Fig. 7 Translational energy distribution Fig. 8 TOF spectra for the dissociation for the dissociation of C-C4F8 of the dissociation of C-C4F8 - 132- 7000 6000 " H1 HJ\ 3O00 4OO0 \l \ Count s 3000 M 2M0 • 1000 0 —i 1 >—1 1 1 . 1 c 1 . 1 c—L_~_ 0 10 20 30 40 JO 60 70 80 Beam-to-detector Angle (degree) CAS(4,4)/3-21G Fig. 9 Angular dependence of C2F4 Fig. 10 Transition state structure for the dissociation of C-C4F8 Fig. 11 Transition state structure of the HCl elimination from CHCICCI2 -133- Table 1. Parameters for beam velocity distributions and laser wavelengths used for irradiation of samples. Sample Nozzle Temp. Pressure (Torr) a S Laser Wavelength (°C) Sample Ar (m/s) (Urn) CF3CHCIF 250 200 0 177 3.4 9.294 CHCICCI2 320 50 150 136 5.6 10.675 CBrF2CHClF 230 95 205 110 5.3 10.303 CBrF2CHBrF 230 55 245 80 7.8 10.349 CBrF2CBrClF 270 30 270 77 8.7 9.305 C-C4F8 230 200 0 140 3.6 10.494 Table 2. Average translational energies, dissociation energies and average excitation energies for C-halogen bond ruptures of halogenated ethane and ethyl radicals. Reaction Laser fluence Average Dissociation a) Average*5) (J/cm2) translational energy excitation energy (kcal/mol) (kcal/mol) energy (kcal/mol) 7a 4.2 2.5 66.3 (65.6) 92 (26) 14.5 3.0 100 (34) 7b 4.2 1.5 25.2 14.5 2.0 8a 2.8 2.0 66.1 (65.4) 85 (20) 8b 14.5 3.0 11.2 9a 3.2 2.2 57.9 (57.4) 82 (25) 9b 14.5 3.5 15.5 a) Values are obtained by ab initio Molecular Orbital calculations at MP2/ECPDZP//HF/ECPDZP level. Values in parentheses indicate the critical energies at which the density of states is minimum. b) Values in parentheses indicate average excess energies. — 134 — Table 3. Conversion fractions of exit barriers to product translation Reaction Exit Barrier Peak Energy Conversion Reference (kcal/mol) (kcal/mol) Fraction (%) 0C4F8 -* 2C2F4 24 7.5 30 This work CHCICCI2 -» CClOCl + HCl 50 2.5 5 This work CH3CCI3 -» CH2=CC12 + HC1 42 5 9 (4) C-C6H10 -» C4H6 + C2H4 26 22 81 (5) — 135 — Session HI Laser Development Proceedings of the 6ih International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- User Issues at the Stanford Picosecond Free Electron Laser Center Todd I. Smith W.W. Hansen Experimental Physics Laboratory Stanford University Stanford, California 94305 USA Tel: (415) 723-1906, Fax: (415) 725-8311 Abstract Assembling a productive user facility around a Free Electron Laser (FEL) is a complex task. Reliable operation of the FEL is a necessary, but by no means sufficient, condition to ensure that the center will be able to attract and keep the interest of first rate researchers. Some other issues which are important include: center wavelength stability and ease of tuning, bandwidth control, amplitude and position stability, ability to select arbitrary sequences of micropulses, and real time availability of information of the FEL's important parameters (spectral width, center wavelength, micropulse length and energy, etc.). In addition, at the Stanford Center we have found that providing additional systems (conventional picosecond lasers synchronized to the FEL, an FTIR spectrometer, a confocal microscope,...) has been important. Keywords: Free Electron Laser, FEL, picosecond, infrared, user facility Introduction The Free Electron Laser (FEL), first developed almost twenty years ago1, is capable of delivering high power picosecond optical pulses at very high repetition rates and with broad tunability in the infrared. These capabilities, largely unmatched by conventional lasers, have lead to extensive proposals for FEL use in a wide range of experiments and applications2'3'4^. However, the relative expense and technical difficulties associated with FELs has limited the community to which they have been available. The Stanford Picosecond Free Electron Laser Center has been established as part of the U.S. Medical Free Electron Laser (MFEL) program to enlarge this community by promoting biomedical and materials research with the FEL. Issues associated with utilizing the FEL as a tool for serious scientific research will be discussed in the following sections. These include reliable operation of the FEL, stability of its operating characteristics, flexibility and ease of control of its parameters, and availability of complementary experimental facilities. — 137— FEL Operation Free Electron Laser Oscillator The basic configuration of light a linear accelerator based output a wiggler period i B„ = wiggler magnetic field FEL oscillator is given in light Fig 1. Fundamentally, it electron beam beam consists of two mirrors (kinetic energy T) forming an optical J wiggler resonator, and an electron beam which is co-linear with the optical beam inside • mirror separation: • 12.683459 ±0.0000001 m a 'wiggler'. The wiggler is a device which produces a Fig 1. Basic configuration of an FEL oscillator. The mirror spacing given is spatially periodic magnetic appropriate for the FEL at the Stanford Picosecond FEL Center. field transverse to the electron's direction of motion. The interaction between the electron beam and the optical beam inside the wiggler leads to gain6'7'8-9, and to oscillation if the gain exceeds the optical resonator losses. Wavelength of Emitted Radiation The wavelength of oscillation is determined by the energy of the electron beam, by the spatial period of the wiggler, and by the strength of the magnetic field in the wiggler. A straightforward explanation is that radiation emitted by an electron traveling through the wiggler has the same wavelength as the wiggler period, but greatly Doppler shifted (as shown in Fig. 2) since the electron is traveling at nearly the speed of light. The operating wavelength is given by: feBX X -• •[l + al) ,where a.. = 2y 2nm„c ) Fig 2. Snapshot of wavefronts emitted by a ielativistic electron as it oscillates down a < o /raw 2 wiggler. Since the electron is traveling at nearly In this equation, Y=(l+E/m0c ) is the usual Lorentz the speed of light, the radiation in the forward factor, E is the kinetic energy of the electrons, and direction is Doppler shifted. B0 and X.q are the amplitude and period of the wiggler field. The term containing aw is a result of the fact that the forward velocity of the electrons decreases as the strength of the transverse field increases. In a typical wiggler, aw~l, and Xq~2 cm. Thus, operation at a wavelength of 2 um requires energies of the order of 50 MeV. Longer wavelengths require less energy. The flexibility of an FEL comes from the fact that it gets its power and characteristics directly from an electron beam. Its temporal structure, its power output, and its wavelength are all controlled by the electron beam, and the electron beam can be precisely controlled through the sophisticated techniques developed for particle accelerators. Thus, the characteristics of the FELs optical beam can also be controlled. -138 — The electron beam which powers the Center's FELs is provided by a Superconducting Linear Accelerator10 (SCA). The accelerator produces a train of electron bunches which are adjustable in length from about one to three picoseconds, separated by 84.6 ns. These electron bunches, called micropulses, repeat within a longer time interval, called a macropulse. The macropulse can be adjusted from a few hundreds of microseconds to many milliseconds, and can repeat at nearly any frequency, limited by a maximum duty factor which depends on energy. A common mode of operation would consist of five millisecond macropulses, repeating at 10 Hertz. Reliability of Operation Reliable operation of an FEL is of paramount importance in attracting high quality researchers to the Center. It is true that the FEL's characteristics enable researchers to study systems which are otherwise inaccessible, and this does provide some understanding and patience from experimenters when faced with an complex system which can occasionally be temperamental. On the other hand, an researcher will quickly loose interest, and move on to other things if frequently disappointed. The Stanford Center's operating experience during 1993 is summarized in Table 1. The Table 1: STANFORD PS FEL CENTER FEL operated for three runs, each lasting 4 1993 Run Summary to 5 weeks, 24 hours a day. The operating wavelength ranged from 3 u,m to 6 ujn, as Start-up and Maintenance Time 845 hours requested by the experimenters. It is Experimental Shift Time 1,542 hours System development 225 hours important to note that 1500 hours of FEL FEL a nd Accelerator Science 365 hours time had been promised to users during Solid State and Surface Science 447 hours this period, 1500 hours were actually Molecular Materials and Chemistry 286 hours delivered. Based on this favorable Bio-Medical Science 219 hours experience, 1500 hours are planned again Total Run Time 2,387 hours this year. Yield 65% Table 1 also indicates the distribution of the FEL beam time among various classes of experiments. Note in particular that more than 40% of the beam time was provided to investigators who are not directly associated with the Stanford MFEL Center. -139 — FEL Characteristics Table 2 shows the range of FEL characteristics available for experimenters this last year, and it also indicaies the additional wavelength range to be covered when two additional wigglers are fully functional. The center wavelength of the FEL is ^able 2: FEL Characteristics actively stabilized11 to about 0.01% at Wavelength Range (Aug. 1993) 2-7 um 4 (.im, and is under direct control of the (Spring 1994) 2-15 (xm individual experimenters. Control is (Summer 1994) 2-100 urn accomplished by a system shown Center Wavelength Stability 0.01 % @ 4u. schematically in Fig 3. Some of the FEL Micropulse Length 1-8 ps output is directed through a dedicated Micropulse Separation 84.6 ns spectrometer and onto a position Energy per Micropulse sensitive detector. The output of the ~luJ detector is proportional to any deviation Macropulse Length 0-10 ms from the desired wavelength, and is then Macropulse Rep Rate 0-120 Hz used to provide negative feedback by controlling the energy of the Iinac. Control can be provided at frequencies up to about lKHz. RF AMPLITUDE REFERENCE SIGNAL Acousto-optic and electro-optic modulators AMPLITUDE allow single and multiple micropulse VARIABLE STABILIZATION ATTENUATOR selection for any of the experimental areas. MODULE Any desired set of micropulses from the FEL pulse train described earlier (84.6 ns KLYSTRON AMPLITUDE SIGNAL separation between micropulses) can be delivered. In addition, an acousto-optic modulator has been configured as a 'noise- RFIN eater', to stabilize the power delivered to the (JHHT)—{j^O1 experimental areas. The noise-eater reduces the 10% power fluctuations from the FEL Fig 3. Block diagram of active FTf L wavelength by an order of magnitude. stabilization system. User Control and Diagnostics Since the various operating parameters of the FEL can vary over such a wide range, researchers using the FEL beam need to know that the specific values which they need for a particular experiment are set and are being held constant. At the Stanford Center, a fraction of the light generated by the FEL is constantly being monitored in an area reserved for optical diagnostics. The remainder of the beam is delivered to the experimenters. - 140- Equipment in the diagnostic area measures the FEL's center wavelength, its optical spectrum, the micropulse length and energy, and the pointing stability of the beam in x, x\ y, and y'. At the present time the spectrum and the micropulse are measured by scanning a mechanical element, and therefore are averaged over many seconds. Systems are under development which will provide the data on a micropulse by micropulse basis. The pointing measurements are made with a arrangement of quadrature detectors and lenses providing the location of the centroid of the beam in transverse phase space. The data taking and presentation are controlled by PC clones running LabView, published by National Instruments. The information is displayed on repeater screens at every experimental station so that the researchers can monitor the state of the FEL, and verify that the desired conditions are present. Figs. 4 and 5 are black and white reproductions of the data displayed to experimenters while the FEL was in two significantly different states. In Fig 4, the FEL was operating with an extremely short micropulse (0.67 ps) at 4.11 \im. In Fig 5 the FEL was operating at 5.04 mm, with the micropulse lengthened to nearly 2.04 ps, to narrow the spectral width. The spectral curve is drawn by connecting discrete data points with straight lines. The autocorrelation curve is a best fit gaussian to the data. The JFELSpcctrumj.. I Pointing! 'Power' display should read in arbitrary units, as the detector is wavelength sensitive, and a correction has not yet -0 4-02 00 0.2 0 4 been included in the program. Several J Autocorrelation I features need to be discussed. First, note that the spectral width and 0.1 4.069 4.080 4.090 4.100 4.110 4.120 4 130 4.140 the micropulse length are nearly | Wavelength (microtia) [ -0.2 00 0.2 0S0G transform limited in both cases. (At 4.11 u,m, the spectral FWHM is 39 nm, Ef 4.1T001 O or 690 GHz (23 cm"1). The product AfAt is 0.46, which compares favorably Fig 4. Data presented to researchers with the FEL operating at 4.11 urn with a 670 fs micropulse. with the limiting value of 0.44. The product for the 5.04 (im example is 0.51.) Although this near ideal quality is routinely achieved in FELs, and is frequently exploited by the experimenters, it must be stressed that an FEL is quite capable of operating with a much more complex longitudinal behavior, in which a micropulse can contain several frequencies with a corresponding complicated temporal profile. The importance of the spectral diagnostic system for continuously monitoring Fig 5. Data displayed with the FEL operating at 5.04 nm, FEL should be clear. with a 2 ps micropulsc. -141- Second, the pointing data show as two clusters often dots each in the Figures. (In the actual display the dots are distinguished by color). One cluster represents the x,y position of the beam as a function of time. The other represents x',y'. Every few minutes the oldest dot is replaced with one representing the current position of the beam. The scale is approximately 100 ^m on a side for x-y, and 100 (Aradians for x'-y'. Our experience is that even though the FEL is located nearly 100 m from the experimental areas, it takes many hours for the clusters to drift perceptibly. Although it is not necessarily obvious from the figures, the display provides researchers with direct control of the operating wavelength of the Center's FEL. This is because the wavelength shown on the display is actually an input from the experimenters, using either a mouse or a keyboard. This value sets the wavelength controlling spectrometer (discussed above, and shown in Fig 3), and as long as the new value is within several percent of an "optimized" value, the FEL will follow it. Table 3: Wavelengths used in A recent experiment performed with the Center's FEL quantum well experiment will serve as an example of the tunability of the FEL, and Sample Sample of the importance of allowing the experimenters to select #3054 #2753 precisely the wavelengths of interest to them. The 4 \xm bam 5 urn bant 5 iim banc 12 13 experiment ' was a study of second harmonic 3.70 4.40 4.87 generation in an asymmetric quantum well (QW) 3.71 4.43 5.01 designed to be doubly resonant at about 4 urn. Table 3 is 3.73 4.45 5.06 a list of nearly 70 distinct wavelengths between 3.7 \im 3.76 4.50 5.08 and 6.7 um which were used over the course of a few 3.80 4.55 5.10 3.83 4.67 5.14 days of collecting their data! One reason for the large 3.84 4.75 5.33 number of wavelengths is that they were looking for 3.85 4.80 5.40 specific features, such as that shown in Fig 6. The figure 3.86 4.90 shows the second harmonic power from the sample at 3.87 4.95 6 iim banc 3.84, 3.85, and 3.86 \im, as a function of rotation angle 3.90 5.06 5.60 between the polarized FEL beam and the symmetry axis 3.91 5.10 5.63 of the GaAs substrate on which the QWs were grown. 3.92 5.14 5.63 The phase of second harmonic generated by the substrate 3.93 5.15 5.67 has an angular dependence, while that from the QW does 3.94 5.20 5.70 not, and the beating between the two is evident in the 3.97 5.25 5.80 4.00 5.33 5.84 figure. Also evident in the figure, and of even more 4.02 5.40 5.91 interest, is the fact that the phase of the second harmonic 4.03 5.49 6.00 generated by the QW is strongly wavelength dependent 4.04 6.13 and reverses sign at 3.85 (Am. The search for this phase 4.05 6 urn banc 6.20 reversal was a key element in the research plans of these 4.06 5.50 6.30 experimenters. Since they were able to monitor their 4.08 5.60 6.38 apparatus and change the FEL's wavelength 4.09 5.67 6.46 simultaneously, the search went fairly smoothly. If it had 4.10 5.70 6.66 been necessary to change wavelengths through an 4.11 5.80 intermediary, such as an official FEL operator, the search 4.12 5.84 4.14 6.01 would have been substantially more difficult. 4.16 6.30 4.18 6.66 -142- Additional Center Facilities and Equipment The center currently has eight experimental areas, each with an 8' by 10' optical bench, to which the FEL beam(s) can be delivered. Some of these areas are devoted to permanent installations of a specific nature, such as a pump-probe or photon-echo apparatus, a) X = 3.84um harmonic generation equipment, a multi- •A beam microscope, aThSapphire laser and a 0.20 dye laser which are synchronized with the * m / \ FEL for two color experiments. The other • % areas can be configured for specific 0.10 ! \ • i \ experiments. 0.00 V \ i Each experimental area is equipped with Jb) A. =V 3.85u m \J a computer and analog oscilloscope, and 1 1 several digital and high speed 0.08 oscilloscopes are available on a floating basis. A modest pool of specialized 0.05 •h. +> equipment, such as a spectrum analyzer, •rs A / a frequency synthesizer, a gated 0.03 integrator, a lock-in amplifier, etc., is available in the event of a sudden need. 0.00 c) k = 3.86um In order to make optimum use of the 0.15 FEL's unique capabilities, the Center has a number of additional facilities14. These 0.10 f•f- * s / \ include a Spectra-Physics Tsunami Ti:Sapphire laser tunable from 700 nm to 0.05 1000 nm, and a Nd:YLF pumped, cavity 0.00 dumped, dye laser tunable from 575 nm 90 180 270 360 to 700 nm. ij> (degrees) Fig 6. Second harmonic power generated by the quantum The Ti:Sapphire laser produces 80 fs, 13 well and the substrate as a function of angle and nJ pulses, while the dye laser produces wavelength. Note the relative phase reversal of the quantum well component between 3.84 um and 3.86 nm, and the 10 ps, 75 nJ pulses. Both lasers are zero at 3.85 um. synchronized with the FEL micropulses, allowing two-color experiments throughout the visible and near infrared spectrum. When the FEL is not in operation, each can be used for stand-alone experiments. The center also has a Bruker IFS-113v Fourier Transform Infrared Spectrometer (FTIR), installed in collaboration with Fisk University. This instrument has a useful spectral range from 5000 cm"1 to 10 cm"1, and a resolution of 0.03 cm"1. A 4.2 K Janis helium cryostst is installed in one channel of the spectrometer. The FEL beam can be delivered through a 3/4 meter monochromator to provide very narrow spectral width sample illumination. -143- Acknowledgment This work supported by the U.S. Office of Naval Research under contract number N00014-91-C-0170 References 1. LR. Elias, W.M. Fairbank, J.M.J. Madey, H. A. Schwettman, T.I. Smith, "Observation of Stimulated Emission of Radiation by Relativistic Electrons in a Spatially Periodic Magnetic Field," Phys Rev. Lett., vol 36, p.717, 1976. 2. J. Opt. soc. Amer. B, vol 6, no 5, 1989, Special Issue on Physics of Free Electron Laser Applications. 3. "Free-electron lasers and application'" in Proc. Soc. Photo-Opt. Instrum. Eng., vol 1227, 1990. 4. K-J Kim and A. Sessler, "Free-electron lasers; Present status and future prospects," Science, vol 250, no 1, pp. 88-93, 1990. 5. Dana D. Dlott and Michael D. Fayer, "Applications of Infrared Free-Electron Lasers: Basic Research on the Dynamics of Molecular Systems," IEEE J. of Quantum Electronics, vol 27, no. 12, pp. 2697-2713, 1991. 6. W. B. Colson, C. Pellegrini, A.Renieri, Laser Handbook: Free Electron Lasers. Vol 6. (North Holland, 1990) 7. C. A. Brau, Free Electron Lasers, (Academic Press 1990) 8. C. W. Roberson and P. Sprangle, "A Review of Free Electron Lasers", Physics of Fluids, Vol 1, 3 (1989). 9. H. P. Freund and T.M. Antonsen, Jr. Principles of Free Electron Lasers, Chapman and Hall, London 1992 10. C. M. Lyneis, M. S. McAshan, R. E. Rand, H. A. Schwetttman, T. I. Smith, and J. P. Tumeaure, "Unique Beam Properties of the Stanford 300 MeV Superconducting Recyclotron," IEEE Trans. Nucl. Sci. NS-28. No. 3, June (1981) pp. 3445-3448 11. "Feedback Stabilization of the SCA/FEL Wavelength," with A. Marziali, Nuclear Instruments and Methods 331, 52 (1993). "Demonstration of Wavelength Stabilization in a Free Electron Laser", to be published in IEEE Journal of Quantum Electronics, 1993. 12. E. L. Martinet, G. L. Woods, H. C. Chui, J. S. Harris, Jr. M.M. Fejer, C. A. Rella, B. A. Richman, "Free-electron laser nonlinear spectroscopy of doubly resonant (5.5-3.0 mm and 4.1-2.1 mm) InGaAs/AlGaAs asymmetric quantum wells", in SPIE Proceedings: Quantum Well and Superlattice Physics V, (1994), vol 2139. 13. H. C. Chui, E. L. Martinet, G. L. Woods, M.M. Fejer, J. S. Harris, Jr., C. A. Rella, B. A. Richman, H. A. Schwettman, "Doubly resonant second harmonic generation of 2.0 mm light in coulped InGaAs/AlAs quantum wells", submitted to Appl. Phys. Lett. 14. T.I. Smith, H. A. Schwettman, K. W. Berryman, and R. L. Swent. "Facilities at the Stanford Picosecond FEL Center," FEL Spectroscopy in Biology, Medicine, and Materials Science, H.A. Schwettman, Editor, Proc. SPIE 1854, 23 (1993) - 144- Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- JAERIIR FREE ELECTRON LASER PROGRAM E.J.Minehara, R.Nagai, M.Sawamura, M.Takao, M.Sugimoto, S.Sasaki, M.Ohkubo, N.Kikuzawa*, J.Sasabe**, Y.Suzuki, Y.Kawarasaki** and N.Shikazono Free Electron Laser Laboratory, Department of Physics, Japan Atomic Energy Research Institute 2-4 Shirakata-shirane, Tokai-mura, Naka-gun, Ibaraki-ken, 319-11 Japan *Department of Nuclear Engineering, Kyushu University 6-10-1 Hakozaki, Higashi-ku, Fukuoka-shi, Fukuoka-ken, 812 Japan **Hamamatsu Photonics Co., Research Center Hamakita-shi, Shizuoka-ken, 434 Japan A prototype for a quasi-cw, and high-average power free electron laser driven by a 15 MeV super-conducting rf linac has been developed, and constructed at Tokai, JAERI. Expected cryogenic( stand-by loss<3.5W at 4.5K) and +9 accelerating fields' performances( Eacc ~ 7MV/m and Q ~ 2 xlO ) of four JAERI superconducting accelerator modules have been demonstrated, and installed them in the FEL accelerator vault. Optical resonators and beam transport systems, which have been already assembled, are now under commissioning. A preliminary beam test of the JAERI superconducting rf linac FEL has been successfully performed to get an electron beam at around 15 MeV. Keywords: Quasi-CW, Free electron laser, superconducting rf linac l.INTRODUCTUION As well known, a laser consists of three major parts, i.e., a laser driver, gain media, and an optical resonator. Since the invention of the laser in 1950's, their efficiency, and average power level have been limited to very low their huge heat losses in the laser driver and gain media. If we could use an FEL instead of — 145 — the convennsional, we could neglect the heat loss in the gain media. Unfortunately, as long as the normal conducting accelerator were used to produce high energy electron beam as the laser driver, we could not realize the high efficiency and high average power level due to the heat losses in the accelerator as the FEL driver. In order to make a highly-efficient, and high average powered FEL, we resultantly have to minimize the heat losses in the accelerator. A developmental program[l,2] of the free electron laser(FEL) system for a far-infrared region from the wavelength of 20 |i.m to 50 |4.m or longer has been undertaken at Japan Atomic Energy Research Institute(JAERI), Tokai. The purpose of the present JAERI FEL program lies in constructing a very long pulse or quasi- continuous wave(cw) superconducting rf linac electron accelerator, and demonstrating a high-average power FEL in the far-infrared wavelength region. Because wall losses and required rf power become minimal in the superconducting accelerator cavity, we may realize a quasi-cw and high-current rf linac driver, and hence a high-average power laser. Each major part of the program including future plans has been reported in other papers[3-8] in detail. Here, we present an outlook of the program including the present status and schedule. 2.INJECTOR The injector of the JAERI FEL consists of a thermionic cathode electron gun with a pulsing grid, a sub-harmonic buncher(SHB), and a buncher. The accelerating voltage in the single gap electron gun is typically around 230KV, and the gun is usable from 200 to 250 KV. The cathode is mounted horizontally in a stainless-steel pressurized vessel with SFg gas to 2kg/cm in order to prevent break down across a 45 cm-long insulating ceramic tube of the gun. The accelerating gap electrodes are fabricated in a re-entrant geometry to increase the accelerating gradient. Optimization of the geometry was made by computer-modeling of electron beams using E-GUN[9]. The injector has been installed, and was commissioned late - 146- August 1991, and is now operated routinely. An extensive study of pulse characteristics as a function of injector parameters has resulted in sets of optimized operating conditions which minimize pulse width at a time focus point while maintaining the beam quality as good as possible. The characteristics typically obtained are as follows: an electron beam of 100mA or more with a 4ns bunch length was extracted from the gun at the accelerating voltage of 180-220KV. The beam was successfully compressed to less than 70 ps at around the time focus point by the bunching system[l]. 3. SUPERCONDUCTING RFLINAC The JAERI superconducting rf linac consists of two pre- accelerator modules of the single-cell cavity type and two main modules of the 5-cell cavity type. The resonant frequency of the cavities is 499.8MHz which is exactly the same with the buncher, and the sixth harmonic of SHB in the injector. We decided to choose a so-called DESY concept of the cavity geometry and fabrication technology refined by Siemens Energieerzeugung KWU for the JAERI FEL superconducting linear accelerator late September,1990. Design values of the accelerating o field strength and Q-value for the cavities are 5MV/m, and 2 x 10 , respectively. In 1992 Japanese fiscal year, we have successfully demonstrated expected cryogenic( stand-by loss<3.5W at 4.5K) and +9 accelerating fields' performances( Eacc—7MV/m and Q~2 xlO ) of four JAERI superconducting accelerator modules, and installed them in the FEL accelerator vault. As a main coupler was designed to have a variable coupling coefficient over 3 and half decades, we could inject not only low current but also high current electron beams into the accelerator module. Typical peak RF power for the coupler was measured up to the 150kW without trouble. The coefficient was designed to be adjusted by inserting a center conductor into the cavity. Three sets of the higher mode couplers were designed, and fabricated to suppress unwanted and harmful TE and TM modes having a higher resonance frequency. Two monitor couplers were -147- designed, and fabricated to use in monitoring and phase detecting in the feedback loop of a fast tuner. Slow and fast tuners were designed, and fabricated to tune a resonance frequency of the cavity in the module. The slow tuner consists of a stepping motor driver and an interface from the control system. The fast tuner consists of a piezo-electric actuator, a high voltage power supply, a feedback loop, and an interface from the phase detector and the control system. 4. CRYOSTAT AND REFRIGERATORS We have newly developed a multi-refrigerators system[5] integrated into the superconducting accelerator module cryostat to realize a highly-efficient system without any liquid coolant. A 4K closed-cycle He gas refrigerator mounted just above a liquid-He supply tower of the module was adopted to cool down and to recondense cold vapor of liquid He around a heat exchanger in the liquid He container. A 40K/80K two-stage closed-cycle He gas refrigerator, which was mounted in a vacuum vessel of the module was adopted to cool down the 40K and 80K heat shields and other major components of the cryostat. These two kinds of the refrigerators are available commercially in Japan and other countries. The 4K refrigerator suspended in a stainless-steel frame can be winched up and down to remove the heat exchanger out of the liquid He container, and to insert the exchanger into the container. Cooling capacity of the 4K refrigerator is 11W at 4.5K and 60Hz. The 40K and 80K heat shields are used to prevent heat invasion from outside into the liquid He container. In order to minimize heat loads to the container, the heat shields work as a thermal anchor, and make the return route having a temperature higher than 4K for all heat bridges from the outside. The 40K/80K refrigerator used here provides two cooling stages with a typical pair of temperature of 40K and 80K and heat load capacities of 40W and 120W , respectively. 5. RF POWER SOURCE - 148- One of the largest merit of a superconducting accelerating cavity is very low power loss, which makes it possible to use all- solid-state RF power amplifiers for all of the cavities[4]. Because the control voltage of the all-solid-state amplifiers is lower than that of a klystron and a tetrode, a more stable RF power is expected to be realized. We choose to use two sets of all-solid-state 50kW RF power amplifiers for the main accelerator modules. Two sets of the power supply have been already installed, and have been ready to use at the experimental area since the middle of last August. Performance of the rf power supplies has been preliminarily measured to be better than 1% of amplitude and within 1 degree of phase stability at an rf power level of 50kW or more. 6. ELECTRON BEAM TRANSPORT SYSTEM The energy of electron beams accelerated by the linac ranges from about 10 to 20 MeV. A conceptual design of the transport system was done by using the beam optics code TRACE-3D [10]. High current beams have to be fed to the undulator under isochronous and achromatic conditions for efficient lasing of FEL. Because of the large amount of charge density, space charge effects would become serious in a long transport line and a beam waist. Since the code could take into account partial space charge effects, the transport system has been investigated by using the code. A beam dump in preparation will be capable of handling about 40 u.A of average current or more, and 1 kW of beam power. Cooling of the dump is provided by air flowing in channels or pipes machined into an aluminum rod. About 30 cm-thick lead surrounds the dump to reduce the radiation levels during routine operation to natural background levels outside the shielding walls made of 150 cm-thick concrete. 7. HYBRID UNDULATOR A wedged-pole hybrid undulator will be used for the first lasing experiment of the JAERI FEL. The undulator was originally designed and built as a prototype undulator for the SPring-8 -149- project[ll]. This device is expected to generate brilliant photon beams of energy ranges around 10 keV by installation into the low emittance high energy storage rings such as the SPring-8[12]. In order to fit the undulator into the JAERI FEL system, the undulator was characterized by three-dimensional field calculation and two- dimensional field mapping. In order to characterize the undulator, a distribution of the multipoles was derived from the field distributions in the median plane of the undulator. A strip of the three-dimensional field distribution was obtained by using a conventional finite element method(FEM) calculation code ANSYS[13], An experimental distribution was obtained by field mapping with commercially- available three-dimensionally measuring equipment. Calculational and experimental distributions of the multipole components along the undulator axis were derived up to dodecapole components from the field distributions by a least- square fitting method. The calculated distributions of the multipoles quantitatively shows very good reproduction of the experimental distributions. 9. ACCELERATOR VAULT A new extension was completed to an old 5.5 MV electrostatic accelerator building as an FEL accelerator vault in March 1992. Two sets of the main accelerator module, the beam transport system, hybrid undulator, and opticals were installed inside the vault in thel992 Japanese fiscal year. 10. PRESENT STATUS AND SCHEDULE In January 1993, we have successfully demonstrated expected cryogenic( stand-by loss<3.5W at 4.5K) and accelerating fields' performances( Eacc > 5MV/m and Q > 2 xlO ) of four JAERI superconducting accelerator modules, and installed them in the FEL accelerator vault. In May 1993, Optical resonators and beam transport systems, which have been already assembled, are now under the beam test. A preliminary beam test of the JAERI -150 — superconducting rf linac has been successfully performed to get an electron beam at around 15 MeV. 11. SUMMARY In conclusion, we have presented the status and purpose of the JAERI quasi-cw, high-average power FEL program concerning the superconducting rf linac driver, and other FEL opticals. We reported our successful demonstration on the performances of the injector, rf power supplies, four JAERI superconducting accelerator modules, hybrid undulator, and Liquid He refrigerators, which have been installed for these three years. We are now active in transporting electron beams, doing some beam studies in the superconducting accelerator modules. Therefore, we hope to report upon the results in the very near future. ACKNOWLEDGEMENT The authors would like to thank Dr. M. Ishii of JAERI for their continuous encouragement and interests in this work. -151- REFERENCES [1] M.Sawamura et. al. ,Nucl. Instrum. Method A318 (1992)127. [2] M.Ohkubo et. al. ,Nucl. Instrum. Methods A296 (1990)270. [3]M. Takao, et al., in the Proceedings of Fourteenth International Free Electron Laser Conference, 1992, Kobe. [4]M. Sawamura, et al., ibid. [5]N. Kikuzawa, et al., ibid. [6]R. Nagai, et al., ibid. [7]M.Sugimoto, et al., ibid. [8]K. Sasaki, et al., ibid. [9]W.B.Herrmannsfeldt, SLAC Report-226, November 1979. [10]K. R. Crandall, et al., TRACE 3-D Documentation, LA-1054-MS, UC-32 and UC-28, 1987. [11]H. Kamitsubo, Nucl. Instr. and Meth. A303 (1991) 421. [12]S. Sasaki, et al., in the proceedings of Particle Accelerator Conference, 1991, San Francisco. [13]Swanson Analysis Systems, Inc. Reference manual of ANSYS- 386 Rev. 4.4. - 152 — Proceedings of the 6lh International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- HIGH-AVERAGE-POWER, DIODE-PUMPED SOLID STATE LASERS FOR ENERGY AND INDUSTRIAL APPLICATIONS William F. KRUPKE Laser Programs Directorate Lawrence Livermore National Laboratory University of California Livermore, CA 94550 Tel: (510) 422.5905 Progress at LLNL in the development high-average-power diode-pumped solid state lasers is summarized, including the development of enabling technologies.. Keywords: laser diode arrays, micro-optics, laser materials 1. INTRODUCTION Remarkable advances have been made recently in the performance of high- average-power, diode-pumped solid state lasers (DPSSLs), including output power level, efficiency, beam-quality, and wavelength-diversity, etc. It is now anticipated that DPSSLs will play an important and increasing role in emerging industrial and energy applications. The realization of high- performance DPSSLs, in turn, has been made possible by advances in many constituent technologies: 1) high-duty-factor, low-cost high-average-power semiconductor laser diode pump arrays; 2) novel low-cost array output beam microoptic elements; 3) new and/or improved laser and nonlinear optical materials, and; 4) advanced resonator/laser cavity mode-control techniques and components. Recent progress made at the Lawrence Livermore National Laboratory in each of these technology areas is summarized in this paper, along with DPSSL performance levels attained. DPSSL work at LLNL, summarized here, has been performed primarily by the group led by Rich Solarz, that includes Geogr Albrecht, Ray Beach, Brian Comaskey, Steve Velsko, Nils Carlson, Charles Hamilton, Steve Sutton, Mark Emanuel, Jay Skidmore, Bill Benett, Barry Freitas, and Pat Reichert. Novel laser materials developments have been performed principally by Steve Payne, Laura DeLoach, and Larry Smith. - 153 — 2. SEMICONDUCTOR LASER DIODE PUMP ARRAY TECHNOLOGY The relatively recent development of high-power, high-efficiency laser diode arrays is the quintessential technological advance that launched the current renaissance in high power solid state lasers. At LLNL, we have focused on the development of a suite of prototype high-average-power, high-duty-factor DPSSLs for use in various military and civilian applications: 1) 100-1000 watt, near-diffraction-limited, high-repetition-rate, free-running and Q-switched neodymium doped crystals lasers for industrial-scale precision- manufacturing applications, and as a source for soft x-ray lithography, 2) 100- 1000 watt, high-repetition-rate, Q-switched, harmonically-doubled and tripled crystal lasers for tactical military applications, as dye laser pump sources for advanced laser isotope separation applications, and for large scale precision manufacturing applications, 3) 100-watt, direct-diode-pumped tunable crystal lasers for remote-sensing applications, and in the longer term, 4) 105 to >106 watt-class crystal lasers for (military) theater missile defense and (civilian) inertial confinement fusion power reactor applications. These lasers generally require 2-D diode pump arrays producing up to -,500 watts of average power per cm2, comprised of linear bar arrays emitting up to ~100 watts of average power per cm, with duty-factors ranging from 25% to 100% (CW). For the mentioned DPSSLs to be economically viable, we estimate that a pump array OEM manufacturing unit cost below a few dollars per average watt will be necessary. We conclude that this is a feasible goal, provided: 1) the basic manufactured unit produces about 100 watts of average power with adequately long operating life (typically >10,000 hours or >10l0 shots); and 2) a low-cost, highly-reliable, mass-manufacturable cooler is developed that is amenable to 2-D stacking at a high packing density (~500 watts average/cm2). To implement this low-cost, high-duty-factor pump array strategy, we adopted and significantly further developed the silicon microchannel cooler approach first reported by Tuckerman and Pease [1]. Fabrication of LLNL's thin (~800 microns) silicon-based microchannel coolers (see Figure 1) draws on lithographic processing techniques developed for large-scale integrated circuit manufacture. The design of cooler packages, and the achieved performance of LLNL array packages have been reported [2-5]. Array packages (1.8 cm of diode arrays) made with circa 1992 processing fixtures and techniques are operated at 40 watts of average power with projected lifetimes >10,000 hours. A summary of our manufacturing methods has been reported [5,6] along with associated manufacturing unit costs. Using our 1992 fully vertically-integrated processing technology and know-how, we were able to produce 40 watt average power arrays at a unit cost of ~3$/average watt (were an annual production rate sustained at -140 kW average per year). The newer (1993) higher performance 140 watt average power array packages are produced at a unit cost of ~l$/average watt (were an annual production rate sustained at -420 kW average per year). More recent array packages using improved laser - 154- material produced at LLNL are now being operated at 100 watts CW from a one cm long bar array [5], (see Figure 2). 3. MICRO-OPTIC LENS TECHNOLOGY A second technology, crucial to the implementation of that new class of diode pumped solid state lasers possible only using unusually high pump fluxes (such as the so-called ground-state-depleted lasers [7] or various end-pumped lasers [8]), is that of low-cost micro-optical elements that are suitable for collimating the output beam from a linear bar array. Our approach to solving this problem is the "shaped fiber lens" [9]. Here (see Figure 3) a glass fiber preform is machined to the required (calculated) shape and pulled into a fiber (cylindrical) microlens with (typically) a 100 microns transverse dimension. The initial shape is maintained during the pulling process, the surface of the drawn lens is "fire-polished" to optical smoothness, and initial machining errors are reduced in magnitude by a factor equal to the ratio of initial to final size. These f/1 spherical-aberation-free micro-lenses are then attached to individual laser bar arrays (see Figure 4 [5]), collimating the "fast-axis" divergence angle (typically 60 degrees) of a laser bar to better than 10 milliradians. Linear bar arrays fitted with these cylindrical microlenses may be further integrated into 2-D array stacks (see Figure 5.) whose high average output power may be coupled to a fiber using a conventional lens, or be focused onto a small spot (laser rod end) using an optical "lens duct" (see Figure 6 [8]). Using the latter approach, pump fluxes in excess of 75 kW/cm2 have been achieved while end-pumping a tunable Cr:LiSAF rod laser [5, 10]. 4. HIGH AVERAGE POWER DPSSL DEVICES These high average power array technologies have been utilized to design, fabricate, and demonstrate a number of high-performance laser devices: 1. A kilowatt average power Nd:YAG laser (free-running repetitively pulsed ©2500 Hz) [11,12] 2. A 250 watt average power, electro-optically Q-switched diode- pumped Nd:YAG power oscillator [13] 3. A 100 watt average power, frequency-doubled Nd:YAG laser [13,14] 4. A 24 watt average power frequency-doubled Nd:YOS laser [15] -155 — 5. NOVEL DPSSL LASER MATERIALS To further exploit the unique high spectral brightness of 2-D laser diode pump arrays, we have performed guided searches for novel solid state lasers materials whose spectro-kinetic, optical, thermal, and mechanical properties are more favorable to the design and construction of lasers suitable for emerging applications. For example, the family of tunable chromium-doped colqunrite laser materials (e.g. Cr:LiSAF [16]) may be diode pumped directly [17,18] using ubiquitous AlGaAs laser diodes or now emerging InGaAlP "red" diodes. Such compact efficient tunable lasers may be used to perform a variety of applicationre (medicine, remote-sensing, etc.). For a second example the family of fluorapatites (FAP) [19] doped with selected rare earth ions appear to be superior laser gain materials for such diverse applications: 1. Gain medium (Yb:FAP [19,20]) in a DPSSL driver for an inertial fusion energy power reactor [21]; See Figure 7. 2. Gain medium (Nd:S-VAP [22]) in a minilaser for use in fiber optic telecommunications sytems, and in optical data storage systems. Figure 1. Advanced Modular Diode Laser MicroChannel Heatsink [4,5]. -156 — s S %% % • , \«- \s\> ^ ^j'};^*, \ *»^^S** *%^V- ^* ^'V'VX* ^"^ jv-V ««*SvJ* -i i ^ ^ " i, * 100.0 >? , —yi ^ * - , . / I * J > * - '" \l -; •-• " '• : If'™! j :^v$ i ;... TJ ! \«^>- f -'\\\*~{r- . J\ ^>^& ....; i ;... U* \ 1 !••• ^M f ; J * \ •? ^ W.*^ * ....; ,: ,f..J '^^'IVv / 50.0 ' r • j. ;.. / '* ;tf r- '" a •; — i / ; i t "• O.0 L. fc 00 100 0 .LOO j^l^u^^JnanitilyfcSwWjWdhjyw £>»»**&£***<*&' H'C.i J Figure 2. Output Power .From a One Centimeter Long Diode Bar 0.8382- 0.5406 Units = cm 0.75 -f 0.4492 I (a) (b Figure 3. a) Ellipical Immersion Lens Preform Shape and Dimensions; b) Electron Microscope Photograph of a Pulled Shaped Fiber Lens (100 micron lens thickness). - 157— Diode Laser ^ens Figure 4. Cross-sectional View of an LLNL Diode Array with a Collimating Cylindrical Shaped-Fiber Microlens Attached Figure 5. 2-D Stack of Microchannel-Cooled Linear Bar Arrays with Attached Shaped-Fiber Microlenses. - 158— .Diode assembly , Lens duct .Laser rod . Potartxtr "^ .Q-SYritCh Output coupler Figure 6. Optical Schematic of an End-Pumped DPSSL, Using a 2-D, Silicon MicroChannel Cooled Diode Pump Array, Fitted with Shaped Fiber Microlenses, and Coupled by a Lens Duct. Qas-eooled Gas-cooled gatnrn»dlum /poekola coll slabs Pulso Injoctlon (p~V Polorlzot Qo3-o0olod harmonic conversion Dlehrolc f^~~> Output pulso mirror _ • •• •' 3 Figure 7. Multi-pass Regenerative Amplifier Architecture for an Yb:FAP DPSSL Driver for an Inertial Fusion Energy Power Reactor [19-21]. - 159 — REFERENCES 1) D. Tuckerman and R. Pease, "Heat Transfer Micros tructures for Integrated Circuits", IEEE Electron Device Lett. ED-2,126 (1981). 2) D. Mundinger, R. Beach, W. Benett, R. Solarz, W. Krupke, R. Staver, and D. Tuckerman, "Demonstration of High-Performance Silicon MicroChannel Heat Exchangers for Laser Diode Array Cooling", Appl. Phys. Lett. 53,1030 (1988). 3) W. J. Benett, B. Freitas, R. Beach, D. Ciarlo, V. Sperry, B. Comaskey, M. Emanuel, R. Solarz, and D. Mundinger, "MicroChannel Cooled Heatsinks for High Average Power Laser Diode Arrays", Proceedings SPIE, Laser Diode Technology and Applications. 1634,453 (1992). 4) R. J. Beach, W. Benett, B. Freitas, D. Mundinger, B. Comaskey, R. Solarz, and M. Emanuel, "Modular MicroChannel Cooled Heatsinks for High Average Power Laser diode Arrays", IEEE J. Quantum Electronics 28,966 (1992). 5) R. J. Beach, M. A. Emanuel, W. J. Benett, B. L. Freitas, D. Ciarlo, N. W. Carlson, S. B. Sutton, J. A. Skidmore, and R. W. Solarz, "Improved Performance of High Average Power Semiconductor Arrays for Applications in Diode Pumped Solid State Lasers', SPIE OE/LASE 1994, Laser Diode Technology and Applications Conference. 6) W. F. Krupke, R. W. Solarz, R. J. Beach, M. A, Emanuel, and W. J. Benett, "The Economics of High-Duty-Factor, High-Average-Power Diode-Pumped Solid State Lasers (DPSSLs)", Conference Proceedings, LEOS 6th Annual Meeting, San Jose, CA, November 15-18 (1993). 7) W. F. Krupke and L. L. Chase, "Ground State Depleted Solid State Lasers: Principles, Characteristics, and Scaling Laws", Optics and Quantum Electronics, 22, SI (1990). 8) R. Beach, P. Reichert, W. Benett, B. Freitas, S. Mitchell, S. Velsko, J. Davin, and R Solarz, "Scalable Diode-End-Pumping Technology Applied to a 100 mj Q-s witched Nd:YLF Laser Oscillator", Optics Letters, 18,1326 (1993). 9) J. J. Snyder, P. Reichert, and T. Baer, "Fast Diffraction-Limited Cylindrical Microlenses", Appl. Opt., 30, 2743 (1991) 10) R. J. Beach, LLNL, private communication, December 1993. — 160- 11) B. J. Comaskey, G. Albrecht, R. Beach, S. Sutton, M. Emanuel, C. Petty, K. Jancaitis, W. Benett, B. Freitas, and R. Solarz, "A One Kilowatt Average Power Diode Pumped Nd:YAG Folded Zig-Zag Slab Laser", Diode Pumping of Average Power Solid State Lasers, Proceedings SPIE 1865,9 (1993). 12) B. Comaskey, R. Beach, G. Albrecht, W. Benett, B. Freitas, C. Petty, D. VanLue, D. Mundinger, and R. Solarz, High Average Power Diode Pumped Slab Laser", IEEE J. Quantum Electronics, 28, 992 (1992). 13) S. Velsko, C. Ebbers, B. Comaskey, G. Albrecht, and Scott Mitchell, "250 Watt Average Power Electro-Optically Q-Switched Power Oscillator", Technical Digest, Advance Solid State Lasers Conference, Salt Lake City, Feb. 1994. 14) S. Velsko, C. Ebbers, B. Comaskey, G. Albrecht, and S. Mitchell, "100 Watt Average Power at 0.53 Microns by External Frequency Conversion of an Electro-Optically Q-Switch,ed Diode Pumped Power Oscillator", Applied Phys. Letters, to be published. 15) B. Comaskey, G. Albrecht, S. Velsko, and B. Moran, "24 Watts Average Power at 0.53 Microns from an Externally Frequency Doubled Q- Switched Diode Pumped Nd:YOS Laser Oscillator", IEEE Journal of Quantum. Electronics, to be published. 16) S. A. Payne, L. L. Chase, H. W. Newkirk, L. K. Smith, and W. F. Krupke, "LiCaAlF6:Cr3+: A Promising New Solid State Laser Material", J. Quantum Electronics, 24, 2243 (1988). 17) S. A. Payne, W. F. Krupke, L. K. Smith, W. L. Kway, L. D. DeLoach, and J. B. Tassano, "752 nm Wing Pumped CnLiSAF Laser", J. Quantum Electronics, 28,1188 (1992). 18) R. Scheps, J. F. Myers, H. Serreze, A. Rosenberg, R. C. Morris, and M. Long, "Diode Pumped CnLiSAF Laser", Opt. Letters, 16, 820 (1991) 19) L. Deloach, S. A. Payne, L. L. Chase, L. K. Smith, W. L. Kway., and W. F. Krupke, "Evaluation of Absorption and Emission Properties ob Yb3+ Doped Crystals for Laser Applications, IEEE J. Quantum Electronics, 29, 1179 (1993) 20) S. A. Payne, L. K. Smith, L. D. Deloach, W. L. Kway, J. B. Tassano, and W. F. Krupke, "Laser Optical and Thermomechanical Properties of Yb- Doped Fluorapatite", IEEE J. Quantum Electronics, to be published. -161- 21) C. D. Orth, S. A. Payne, and W. F. Krupke, :Diode Pumped Solid State Laser Driver for Inertial Fusion Energy Power Plants", ICF Quarterly Report, Vol 3, Number 4, July-September, 1993, Lawrence Livermore National Laboratory (UCRL-LR-105821-93-4. 21) S. A. Payne, B. H. Chai, W. L. Kway, L. D. DeLoach, L. K. Smith, G. Lutts, R. Peale, X. X. Zhang, G. D. Wilke, and W. F. Krupke, "New High Cross Section Laser Crystal: Neodymium Doped Strontium Fluorovanadate", Post Deadline Paper PD-12, Conference on Lasers and Electro-Optics (CLEO), Baltimore, MD, May 1993. -162- Proceedings of the 6lh International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- TECHNOLOGICAL GAS-DYNAMIC C02 LASER V. V. Apollonov, A. M. Prokhorov, V. I. Kislov General Physics Institute, Moscow, Russia V. P. Malyavin, V. A. FeofilaJctov SIC "Luga" Moscow, Russia The self-contained technological gas-dynamic laser (SCTGDL-lOOkW) for various technological operations based on widely used liquid fuels will be dis cussed. The tasks enumerated require the autonomous movable technological laser complex production and the new application fields will be expanded for discussions. Keywords : gas-dynamic laser, liquid fuel 1. Self-contained technological gas-dynamic CO; laser 1.1. Purpose. The self-contained technological gas-dynamic laser (SCT GDL) is designed for technological operations (welding, cutting, surface hardening, etc.) without stationary power sources, compressed air and so on and consumes only widely used liquid fuel: kerosene and toluene. 1.2. SCT GDL arrangement The SCT GDL is located in a special container mounted on a trailer or other vehicle. The container consists of the following major compartments: A. a working compartment with: • a compressor unit with the aircraft turbojet engines; • a gas-dynamic laser, the main parts of which are combustion chamber, nozzle unit, resonator and diffusor; — 163- B. control compartment; C. fuel compartment. The dimensions of working compartment are: length,m 5 high,m 2.8 wide,m 2.5 The setup weight is 5000kg. 1.3. Method of operation. The working gas is formed during toluene (kerosene) burning in the air which passes into the generator unit from the compressor unit. Inside the optical resonator the inner power of the working gas expanded in the nozzle unit is partially transformed into laser emission. The ejection of the exhaust gas is made directly into the atmosphere. 1.4. SCT GDL parameters. Continuous emission power, kW 70-100 Angular divergence, rad 3-10-'1 Beam diameter, mm 90-120 Flow rate of kerosene, kg/min 40 Flow rate of toluene, kg/min 13.3 Original life, h 5000 Guaranteed life, h 1000 Onboard tanks capacity, m3 2.3 -164 — 1.5. SCT GDL servicing. A self-contained technological GDL operates for 30 min using onboard storage of fuel. A continuous operation is provided with additional fuel tanks. A SCT GDL operates continuously within 6 hours per shift with overall original life amounting to 5000 hours. Scheduled preventive maintenance is made after 1000 hours of operation. During overhaul repair a replacement of nozzle unit blades, resonator mirrors, separate units of aircraft engines and compressors is fulfilled. Servicing is provided within a year, after that the manufacture delivers the required spare parts, units out of operation within a month after the reception of an application and accomplishes the repair. It's price includes the price of spare parts and travel expenses. — 165 — 2. Pre-prototvpe technological gas-dvnamic laser (PT GPU. 2.1. Purpose. A pre-prototype technological gas-dynamic C02 laser is designed for optimization of optical systems for technological setups and for conduction of experiments on materials' welding, cutting and hardening. 2.2. PT GDL arrangement. A pre-prototype technological gas-dynamic laser consists of (fig.l.): 1. the oxidizer supply and storage system; 2. the ballast supply and storage system; 3. the fuel supply and storage system; 4. generator (combustion chamber); 5. nozzle unit with the receiver; 6. optical resonator with a gas-dynamic window; 7. diffusor; 8. noise limiter; 9. parameters control and recording system. Fig.l. — 166 — 3. Field of application. 3.1. Metallurgy industry. A qualitative and quick slab cutting may be fulfilled with 70-^100 kW laser beams without strain hardening in the cut point, besides one may harden the working surface of the rolls and increase their original life 2-3 times. 3.2. Engineering industry and railway transport. In the production of large-scale machines and equipment laser beams can provide: • precise cutting of parts 50-100mm thick produced practically of all materials/ • a high-quality welding of thick-walled parts produced of various metals and different sorts of alloys; • hardening of large metal surfaces; • effective rails repair without their disassembly. 3.3. Oil and atomic industry. SCT GDL may be used for disassembly of derricks, trestles which became unnecessary, as well as during accidents and fires at oil wells. We consider promising the following applications: remote units disassembly of worked-out nuclear reactors on submarines, nuclear-powered ice- - 167— breakers, nuclear-powered electric stations, ferroconcrete blocks cutting included. 3.4. Oil films annihilation on water surface. As the result of accidents with tankers about 10-50 thousand tons of oil appear on the surface of the water amounting to several tens of square kilometers, with this one half and even more of it is covered with an oil film 10-100|.im thick that can't burn without assistance. By means of a medium-powered laser mounted on board a heavy helicopter, aircraft or vessel, it's possible to annihilate (burn out) such a film ecologically harmlessly with the productivity of 10-100 km2 for 2-3 days. Depending on means of transportation (helicopter, aircraft or vessel) annihilation of oil spots may be possible at sites located at the distance of 300-3000 km from technological systems basis. -168- CALCULATED RESULTS LIQUID FUEL HIGH POWER GAS DYNAMIC LASER Specific po»er pick up a' z 15 ^IQI o ^^^ | TN 10 ^ \ »T»0 \ps y.>.o 27 \ \ l_m m, m n • 1 1300 (500 1700 1900 2100 2300 T.,"K ^zn 1 Intermits* dlstrlbual'on In output uiTor plana 1— 1 - - 1 Ita, kV//cM» 3H 1 1 ! 1 1 _ ' 1 -iu IU— Ui u j^-aadynaalc tflnaow h = 0.3 MM X,Cm G(C,H,) • A kj/i V? zz 1 ;- T_~j O M=5.0 Gteo) = S4kf/» (3 3 ' JCu =220 G(«i) = 42 kg/, ri-r-i T. = 2000 *K s'. 1.0 LL 15 U 2.0 ?••* 3.0 M p. - 4.5 atm OniClL RZSOXiSOR oooLsd aimr -€ 6 * 1.2-10'pftA 8,= 2.4-lO!«A CD 9j = 6.0 •10" "A RX=40M meaaurtMnt ox chiracttristlca of CDL enlttance Ra = 200K ooolffd dlXraotloa grating a0(t)< 3-10'ua A\p(t)<1.3-l0"pAU focusing mirror CL/ter cylindrical optics 8<6-/0 Wave /ront structure -169- .Self-cuntaliiosct TiScl'inolofjlcal jtMS-tlj'Mftiriics: lfeStsr ilr enllector eocbustor rezomtlor ejector MttlUl Principle of resonator control In flow lasers (cas ana cnemical lasers!,, CH*n*CTtmsiiC5 ** SCT GOL Reference ' "nie Active resonator confiouralion for control of the output power of laser h- resonator length along the flow K-amplification coefficient Kp, Kp'- thresold amplifications; Kp''Kp The loss of low modes U 2-nd channel 1 ° ° stable unstable unstable under control Single channel configuration The amplitude distribution Problem formulation-. 1. To cetermine the law of the mirror deformation for control of — first mode resonator transparency without affecting the beam quality. secondmodB 2. To determine the parameters of the resonator and flexible mirrors, which provides effective control of laser characteristics. £>L. — no— Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- LASERS FOR FUSION ENERGY John F. Holzrichter Director, IR&D Lawrence Livermore National Laboratory Solid state lasers have proven to be very versatile tools for the study and demonstration of inertial confinement fusion principles. When lasers were first contemplated to be used for the compression of fusion fuel in the late 1950s, the laser output energy levels were nominally one joule and the power levels were 105 watts (pulse duration's of ICh3 sec). During the last 25 years, lasers optimized for fusion research have been increased in power to typically 100,000 joules with power levels approaching 1014 watts. As a result of experiments with such lasers at many locations, DT target performance has been shown to be consistent with high gain target output. However, the demonstration of ignition and gain requires laser energies of several megajoules. Laser technology improvements demonstrated over the past decade appear to make possible the construction of such multimegajoule lasers at affordable costs. Key words: Lasers, Fusion, Inertial Confinement The evolution of high power solid state lasers designed for the inertial fusion application started shortly after the invention1 and demonstration of the first laser by Maiman2 in 1960. The invention of the Nd:glass laser by Snitzer3 in 1961 made possible, in principle, the scaling of lasers to large sizes with concomitant high output energy and power. During the 1960s, the national laboratories of many nations—USA, France, Russia—experimented with laser irradiation of matter in various conditions. Later in the 1960s, researchers invented the needed laser technologies, which when optimized and assembled into systems consistent with propagation constraints, were able to heat matter to temperatures that began to interest fusion researchers. In the 1970s, our groups at Livermore, researchers at the University of Rochester and KMS Fusion, the Naval Research Laboratory, CILAS and the CEA Lemeil Laboratory in France, and the Lebedev in Moscow were able to obtain sufficiently accurate data on the generation of short pulses of 1.06 micron Nd laser light, on the amplification, the propagation, the isolation, the focusing, and the diagnostics that subsequent engineering of fusion lasers into large systems became possible. The development of accurate diagnostics with sufficient time resolution and dynamic range was critical to experimental progress and to subsequent theoretical interpretations during this decade. — 171- First Ruby Laser 1961 1M scale prototype amplifier 1J,103W. for Nova upgrade > 106 J and 1014 W. Figure 1: On the left, the first laser demonstrated by Maiman generated about 1 joule in 1 msec. On the right is shown a 1 /4 scale prototype amplifier for the Nova upgrade which will generate about 1.5 mega joules in 10 nsec. This is a one million fold increase in energy and eleven orders of magnitude increase in laser power over the 30 year period of high power laser design.4 10e 7 10 NIF "in 106 » O) 10" w 01 c 1000 w at 0 100 -A. 100 "5 1 10 10 O) 1 1970 1980 1990 2000 Figure 2: The laser energy (left vertical scale) used to irradiate increasingly large fusion targets is shown as a function of a 30 year development cycle. The associated increase in target scale is shown on the right vertical axis. The dotted line on the curve between now and the year 2000 shows that full scale experiments of target physics issues are being conducted using the Nova laser, but an integrated target demonstration will require a laser delivering energy exceeding 1 megajoule. -172 — Early attempts to raise the energy of short pulse, high power laser systems were thwarted by the self-focusing instability. This propagation instability is associated with the modification of an optical material's index of refraction as an intense laser beam transverses it. Thus, a local increase in the beam intensity because of diffraction rings, dust specks, etc. caused a local increase in the index of refraction, which caused the beam to locally experience a lens-like focusing, leading to a higher intensity downstream, and eventually filamentation and optical material breakdown. The problem with this instability is that, like most instabilities, it grew exponentially and thus was a potential limiting condition on the continued development of high power lasers. The near field amplitude growth rates are given by the equation I = IQ exp (B), where the "B" integral is equal to const A i (1*21 dl).5 100J.0.1nsec lw^Mlh^--1WiBaHtl Exit from spatial filter Fig 3: Early experiments showed the presence of beam break-up and loss of focusability. These experiments were conducted on the Cyclops laser facility a t LLNL in the early 1970s by Simmons et al.6 Once the problem of self-focusing was understood (and controlled using high gain amplifiers and multiple spatial filters), as well as understanding the problems of controlling parasitic (longitudinal and transverse) and isolation, one was able to begin to conduct precise laser plasma experiments. These high gain, high power single beam-line laser target systems were stable against parasitic oscillations, against prepulse damage to the targets, and were focusable and diagnosable. Having understood single beam lines, the designers turned their attention to the problems of multiple beam lines. -173- • Large apertures • High SB (Multiple spatial filters) " Multiple beams Diagnostics and control Ablative target implosions but major unresolved laser- plasma coupling issues Fig. 4: The Shiva laser was a 20 beam, 10 kilojoule, 20 terawatt fusion laser system that was designed to irradiate a hohlraum fusion target with 1 nsec, 1.06 u.m laser light, and to control implosion symmetry. Its construction was completed at the end of 1977. The Shiva laser was a very successful system and demonstrated the use of large diameter (20 cm) laser apertures, multiple spatial filters, multiple beam control and aiming, and it was used to demonstrate the concept of ablative target implosions. (Through the middle 1970s almost all implosions were of the exploding capsule-wall type, which were not capable of producing high density implosions of the fusion fuel.) However, the most significant experimental result of this decade was the realization that wavelengths shorter than 1 urn, and probably shorter than 0.5 urn, were needed to control laser plasma coupling instabilities. By the late 1970s, laser design and manufacturing had progressed to the point that new materials were available such as phosphate laser glasses and large KDP harmonic conversion crystals, as well as continued evolution in the mechanical, electrical, controls, and diagnostics of the unique electro-optic systems. A new 10-beam laser, Nova, was designed in 1978 and constructed in the early 1980s at the Lawrence Livermore Laboratory. A 2-beam prototype of Nova, called Novette, was used to validate design concepts and to validate the coupling of short wavelength radiation to the inertial fusion targets. This Nova laser is larger in diameter and is optimized for 3 nsec pulse duration and to produce an output of 100 kj at 1.05 |im, 50 kj at 0.53 |im, and 40 kj at 0.35 um At 1.05 urn each beam of Nova generated as much energy as the entire 20 beam Shiva system described above. Its cost on a per joule basis was 1/4 that of Shiva, and its controllability, maintainability, and diagnostics were far superior. A laser system, Gekko XII, based upon similar design principles was constructed in Osaka, Japan during this time period, and it too has been very successful in conducting inertial confinement fusion and laser plasma coupling experiments. -174- The most important physical constraints in the construction of these high performing systems were associated with the goal of high energy beam propagation as well as high power pulse propagation. This new operating condition is a consequence of extending the power levels of previous systems to longer time durations and, thus, the energy fluence through the optics could increase until breakdown. In addition, the long path lengths of these lasers, both longitudinally to the target, but also transversely (in up to 1 M diameter optics), allowed sufficient gain for new parasitic modes to develop that, if uncontrolled, would seriously limit the laser performance. . ___, . 1 issues wore: • Short wavelength (0.35 pm) ••. coupling ;,. • Both energy fluence and i power limited "^ • Pulso shaped and bandwidth flexible 1 m • Cost effective and t maintainable - • Management of complexity Fig 5: The Nova 10 beam laser system generates a 1.05 urn wavelength beam which is converted to 0.35 |xm directly before the target focusing lens. The pulse duration is variable from 0.1 nsec to greater than 3 nsec. Extended to "upgrade" aperture 0 • 1 1 L , 1 L.1 i !-,_. *_L. r~~~ ," "\ o o - * / axi s (mm ) > 300 1 V OL'...IA...'.!• • i1 ' '"' | i . i i • • • | i ' ' 1 0 100 200 300 400 H-axis (mm) 74cm diameter lull power fill factor 0.85 "smooth" profile Mil factor - 0.8 Fig. 6: The beam quality of modern lasers is such that a 5% peak to average beam amplitude "ripple" is seen at the output of the system. The above left picture shows a 74 cm diameter full power Nova beam with aperture usage of 0.8. On the right one sees an experimental beam from "Beamlet" for the upgraded Nova laser (NIF) which is "square" in aperture and has a fill factor of 0.85. - 175- The issues associated with designing and constructing the NIF (an extensive upgrade of the Nova laser) are strongly influenced by the wealth of new physics, engineering, and materials information obtained during the last 20 years of laser and target experimentation. The issues associated with the complexity of building the much larger system, which will have 192 laser apertures packaged 16 to 32 at a time, are those of cost and complexity. As a result, the accurate control and diagnostics of this system has deserved a great deal of attention. The dominant laser issues are associated with the efficient amplification and propagation of each laser beam at a higher fluence than has been used in the past. Operation at high fluence from a given aperture is desirable because the system costs are fixed for the mechanical hardware, controls, and diagnostics. More energy per unit fixed cost leads to an overall cost reduction when measured on a dollars/joule basis. Additional important laser design considerations are associated with controlling nonlinear parasitics covered by transverse Brilloin or Raman gain in the optical elements, with longitudinal parasitics associated with the very long propagation paths, with generating very accurately shaped temporal pulses, with efficient harmonic conversion, with purposefully generating and controlling the laser beam's spatial (Ak) propagation vectors and temporal (Aco) bandwidth. These issues have been satisfactorily resolved, and the beam amplitude picture shown in the right hand images in Fig. 6, taken from the "Beamlet" prototype, shows successful realization of this design. In addition, continuing target experiments on Nova are validating many of the assertions that controlled spatial and temporal bandwidth, at the scale of NIF targets, will produce the desired implosion quality. chamber Fig. 7: The proposed National Ignition Facility is a 2 mega joule, 500 terawatt laser for research into fusion target ignition and gain. It is an extensive upgrade to the presently operating Nova laser. The Beamlet laser shown on the right side, has been recently completed. Laser propagation experiments have validated the operating conditions of the proposed NIF laser operating points. -176 — Research on more effective ways to ignite laser fusion capsules continues because it is felt that the physics of capsule irradiation, capsule compression, ignition, and fuel burn will permit more efficient solutions than have been proposed so far. One of the most interesting is the "fast ignitor" concept of Tabak et al.7 In this configuration, a target is compressed as efficiently as possible (isochorically, with no central ignition condition required that reduces present target compression efficiency). When the desired fuel density is achieved, a very short, few-picosecond duration, laser pulse is propagated as close to the compressed fuel as possible, where the pulse generates an intense beam of hot electrons that ignites the fuel from the edge. Realization of this concept (see Perry et al.8) could reduce the cost or increase the yield from a fusion laser facility by up to an order of magnitude. >10,000:1 pulse compression I Fig. 8: Fast Ignitor: The lower image in this figure illustrates the concept of isochoric compression and edge ignition of a laser fusion target. The top image illustrates the large area pulse compression configuration being implemented on one beam of the Nova laser in order to test the physics of the "fast ignitor" concept. Since its invention in 1959 by Nuckolls et al.,9 the development of the inertial confinement fusion concept has been a remarkable achievement. The inventiveness of the target designers, the laser designers, the experimentalists, and the engineers have lead to a proposal for a system (NEF) that will demonstrate, without doubt, for the first time, high gain fusion in the laboratory. - 177- REFERENCES 1. A. L. Schawlow and C H. Townes, Phys. Rev. HI (6) 1940-1949 (1958). 2. T. H. Maiman, R. H. Hoskins, I. K. D'Haenens, C. K. Asawa, and V. Evtuhov, Phys.Rev.i, 1151-1157(1961). 3. E. Snitzer, "Optical Maser Action of Nd^+ in a Barium Crown Glass," Phys. fov. Lett. Z, 444 (1961). 4. J. F. Holzrichter, Nature 3_1£, 309-313 (1985). 5. J. B. Trenholme, The derivation of the far-field growth of small-scale, self- focusing instabilities; the only available reference is in the Laser Program Annual Report, Lawrence Livermore Laboratory, Livermore, CA UCRL-0021-76 (1977), p. 2-337. 6. W. W. Simmons, S. Guch, Jr., F. Rainer, and J. E. Murray, "Teraivatt Laser Performance," IEEE J. Quant. Elec. OE-II. 31D (1975). 7. M. Tabak, J. Hammer, M. Galinski, W. Kerr, S. Wilkes, M. Campbell, M. Perry, J. Woodworth, Physics Plasma (formerly Phys. Fluids) 1,1626 (1994) 8. M. Perry, G. Mourou, Science 261,917 (1994). 9. J. H. Nuckolls, L. L. Wood, A. R. Thiessen, and G. B. Zimmerman, Nature 239. 139-142 (1972). Acknowledgments: The author would like to acknowledge his colleagues at the Lawrence Livermore National Laboratory. In particular, the author would like to thank John Campbell for supplying much of the information on the Beamlet laser and the NIF design for this paper. The contributions of Richard Berger, Mark Henesian, John Hunt, Howard Lowdermilk, John Murray, Michael Perry, and John Trenholme in preparing this paper are appreciated. He would also like to cknowledge the remarkable accomplishments of Mike Campbell for experimental realization, John Emmett in laser design, and John Lindl and John Nuckolls in target design. - 178— Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- DYE LASER CHAIN FOR LASER ISOTOPE SEPARATION D. DOIZI*, L. A. LOMPRE and P. ARONDEL Centre d'etudes de Saclay, DPE/SPL, Bat. 467, F-91191 Gif sur Yvette Cedex, FRANCE . We describe a dye laser chain developped for the french Laser Isotope Separation project (SILVA). We report our latest technical developments and performances. Keywords: Isotope Separation, Dye lasers 1. INTRODUCTION Different processes used for uranium enrichment have been studied in France these last fifteen years and have led to the choice of SILVA as the most promising one. This process has the important advantage to be highly selective and is expected to take the lead over statistical processes such as gaseous diffusion. It involves two major subsytems: the separator and laser subsystems. The first one, the separator subsystem, is composed of a vaporization source -179- which delivers large amounts of atomic uranium vapor and of an extraction device to collect the product. The second one, the laser subsystem, interacts with the atomic vapor to selectively ionise U235, . The tiny difference between the absorption spectra of U235 and U238, which is the main physical property used to make the process selective, requires a very peculiar laser subsytem composed of dye lasers pumped by copper vapor lasers (CVL). The description of the dye laser subsystem will be the main purpose of this paper. 2. DESCRIPTION OF THE DYE LASER SUBSYSTEM The dye laser system developped for the french Laser Isotope Separation project ( SILVA) in CEA laboratories has now been working efficiently in our Process Pilot Facility A2 for more than six years. It consists of a complete four wavelength laser system used to selectively ionize U235. Each dye laser chain is composed of a master oscillator and two or three amplifying stages. The master oscillator shown on figure 1 is composed of an Hansen oscillator which generates a single mode tunable beam frequency stabilized with a ~ 10"7 relative precision obtained by means of a specially developped Fizeau wedge wavemeter. - 180- It is optically pumped by copper vapor lasers (CVL) through optical fibers.The output spectral width can be extended as desired from 50 MHz to a few GHz with an electrooptical modulator in order to cover the hyperfine structure of U235. The nominal specifications obtained with the oscillator are summarized here: Frequency stability : long term <50 MHz short term < 10MHz Tuning Range : 550 nm - 700 nm Divergence : diffraction limited E.S.A : < lO"7 Linewidth : ~ 50 MHz Pulse energy :~ 25 /J Repetition rate : 5 - 25 kHz Pulse length : ~ 50 ns The beam is then shaped to rectangular spatial profile in order to achieve efficient amplification. A great attention must be devoted to obtain an amplification process which doesn't modify the intensive properties of the oscillator beam. Amplifier stages have been scaled and designed using modelling procedures such as hydrodynamics, photophysics and optical modelling. Hydrodynamics studies were first conducted on a Reynolds similitude bench (magnification x 40) in order to carefully design the dye flow channel. At high repetition rates, thin boundary layers and absence of recirculation zones is obtained. This allows very few distortions of the wavefront of the dye laser -181 — beam to be amplified.To obtain such a satisfactory behavior, a great attention was also devoted to the inlet flow configuration which greatly determines the nature of the flow regime. We choose to work with a flow with high Reynolds numbers and low scales of turbulence. A good choice for the dye flow loop is also important. Different types of pumps were studied in order to minimize vibrations and to obtain very high flow rates. A great attention has also been devoted to the choice of the materials used to realise the dye flow loop in order to lower pollution effects. This allowed us to obtain long window lifetimes and low dye photodegradation. The active volume is composed of an organic dye dissolved in an alcoholic solution. Our main challenge is to enhance the pump to dye conversion efficiency in order to get the lower cost for the tunable photons. New dyes have been synthetized and tested on a specially designed small signal gain measurements testbench. We are studying the performances of new xanthylium salts and new rhodamines dyes. We expect to obtain conversion efficiencies over 50 % thanks to a better understanding of the different pathways for energy after an optical excitation. Particularly, we try to lower the absorption of the excited singlet states at the dye laser wavelength. The dyes are also photophysically and photochemically characterized (fluorescence spectra, fluorescence lifetime, absorption cross sections...) in order to predict their amplification performances by means of a code. This amplification code describes the dye with a three level system. Triplet states are neglected and the kinetic equations are solved using the steady state approximations. Optical modelling comprises two parts: - the propagation of the dye laser beam along the dye laser chain -132 — - the optical CVL pumping through optical fibers. The first part of the modelling uses a homemade diffraction and propagation code based on Fourier analysis. Wavefronts and intensities of the dye laser beams are calculated along the chain and compared with experiments. Figure 2(a,b) shows examples of the intensity of an experimental beam measured with a CCD camera (a) and a calculated one obtained by means of the code (b). Figuve 2 a) INTENSITY of a real beam (CCD camera + image analysis) b) CALCULATED INTENSITY AT THE EXIT OF THE POWER AMPLIFIER -183 — Low wavefront distortions (below X/5 RMS), high extraction efficiency (over 50%) are expected. The influence of the presence or absence of the pump beam, the change in repetition rates, the values of the renewal rates for the dye volume are analysed. Different ways to optically couple amplifiers have been studied. At the present time all the active components are optically relayed by telescopes. The second part of the modelling uses classical optical laws. The CVL pump beam at the exit of the optical fibers is imaged in two dimensions onto the amplifier tranverse windows. The optical combination is calculated to enable the easiest optical and mechanical settings. The nominal specifications of the dye amplifiers are summarized below: Energy: > 30 mJ Conversion efficiency: > 40 % A. S. E: < 3% over the entire spectrum Wavefront distortion: < X/5 RMS Repetition rate: 5-25 kHz Dye flow: rate 0.15-5 m3/ h pressure 0.5-3 bars - 184- 3. CONCLUSION In order to achieve our general assesment scheduled in 1996-1997, we are working on the extension of our Pilot Facility. An additive laser power (ten times higher) and a larger separator will be installed in this new facility named ASTER. This is a very good opportunity to take into account these new technical developments. The "all fiber" pump beam transportation allows a safe beam handling and decreases the optical components number. Such a system will be installed in ASTER. In research and development, work is still going on increasing the power of the CVL pump modules. More than 400 W have been obtained on prototype modules using solid state power supplies. In the dye laser domain, we are now working on our main challenges, i.e. to increase the pump to dye conversion efficiency, to run the chain at higher repetition rates, to deliver low distorted optical wavefronts at the entrance of the separator system. Preliminary experiments are in progress with adaptative optics to improve the spatial beam quality and to adjust the size of the laser beam. -185 — Proceedings or the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- ADVANCES IN TUNABLE SOLID STATE LASERS J. J. EWING, Dennis LOWENTHAL, Roy MEAD, Mark BOWERS, Charles MIYAKE, and Steven TIDWELL Aculight Corporation 40 Lake Bellevue, Suite 100 Bellevue, Washington 98005, USA Significant progress has been made in the development of near infra-red and visible solid state lasers that have broad tunability. The Titanium Sapphire laser, and now a revised look at an older technology, Optical Parametric Oscillators, all solid state lasers have extensive spectral coverage. With new materials, combination with other non-linear optical methods, and better control of the pump laser beam quality, OPO's provide a simple and powerful method for coherent generation in the UV, visible and 1R. Keywords: Tunable lasers, Non-Linear Optics, Optical Parametric Oscillators 1. INTRODUCTION For many applications, an all solid state laser source has exceptional promise. Compact size and minimal environmental impact are attractive features to end users. However, the solid state laser has traditionally been limited in power capability, especially in the mode of very high beam quality needed for frequency multiplication. Recent advances in the development of diode pumped lasers, which produce a smaller thermal load in the solid state medium, and the continued decrease in price of diode arrays has led to compact, efficient Nd laser with power of over 100 W output with high beam quality. Research on higher power 1 kW class diode pumped Nd lasers is in progress. In parallel to the development of higher power, high beam quality diode pumped solid state lasers, researchers and laser manufacturers have made significant advances in the technology and product offerings of tunable solid state lasers. All solid state solutions to almost every wavelength region now appear possible. For this paper tunable solid state lasers will be used to mean both real lasers such as TiiSupphiro mid Optical Parametric Generation, which is a tunable Non-Linear Optical phenomena which now can provide extremely broad spectral coverage. It is useful to compare these advanced all solid state techniques with the tunable laser of choice for most applications, the dye laser. Dye lasers have the advantage of broad visible spectral coverage, with the use of multiple dyes, and can be excited by solid state lasers, copper vapor lasers, excimer lasers as well as directly lamp pumped. There are many special niches -186— where the dye laser will continue as the tunable laser of choice. A top level comparison of dye laser technology, TkSapphire solid state tunable lasers, and Optical Parametric Generation is given in Table 1. In general the flowing liquid solution of dye lasers provides the laser designer with capability for higher power levels, while moderate powers, 1 to 10 W, suitable for numerous applications are now possible with solid state solutions. The dye system will use several different dyes to cover any spectral range, while the OPG methods will achieve this with a few crystals. The dyes for most wavelengths degrade with time and the solvents and dyes themselves can constitute an environmental and safety hazard. The all solid state system is more environmentally benign, though the increased optical intensities that characterize Non-Linear Optical process place more stringent requirements on the laser designer. TABLE 1. BRIEF COMPARISON OF TUNABLE SOLID STATE AND DYE LASERS General Property Dye Lasers Broad Bandwidth Optical Parametric Solid State Laser Generation Spectral Coverage Visible and Near IR Limited to near IR. 400 to 8000 nm with 3 multiple dyes, each- 700 to 1000 nm. crystals. UV with Non-linear 20nm typically UV/vis with NLO optics UV with NLO Operating Mode cw, pulsed, ultra cw, pulsed, ultra pulsed, and ultrashort as short to psec short to 10 converter for solid state laser femtosec Power Scaling and Flowing dye Conduction cooling Conduction cooling, power Heat Removal solution > 100 W Powers ~10's of W potential varies with material, 15 W demonstrated Energy Scaling mJto 10'sofJ mJ to ~ 5J, crystal up to 10 mJ, crystal size and size and drive laser drive laser limited limited Conversion 1 to 30 % , ~ 25% from green Varies, 1% to 60% Efficiency A. dependent source Other Features Dyes degrade Start up time makes Easily coupled to frequency Environmental, mixing difficult mixing Health, and Safety Direct diode Phase and conversion concerns pumping of Cr processes are dynamic lasers Among the solid state tunable lasers, both the most developed laser, Ti:Sapphire, and the long term promising concept of a direct diode pumped tunable laser such as CnLiSAF, the tuning range is broad but can not give full visible coverage. The Parametric Generation scheme in contrast has very broad tuning range, and is simpler to couple to other Non-Linear Optical processes such as frequency mixing. For this reason the future of Optical Parametric schemes is -187— very promising and research in this area has seen a tremendous rebirth of interest. When used with frequency multiplication of the primary "drive laser" and mixing schemes, the all non-linear optics technique has demonstrated coverage from .25 to 4 microns, and can be readily extended to the 8 micron region. This is can be done with just a few crystals, a marked difference to dye lasers. 2. Optical Parametric Oscillator Summary Optical Parametric processes are a non-linear optical phenomena that convert an input wavelength, called the driver wavelength, into two other colors commonly referred to as the signal and idler wavelengths. In practice the application may use the signal or the idler, and in the case of a "degenerate OPO" when the frequency is divided by two, and the signal and idler have the same color, both waves may be used. The output wavelengths are related to the input by a momentum conserving phase matching condition that relates the angles of the waves relative to the crystal axes and the wavelength and the indices of refraction along the crystal axes The indices are typically temperature dependent. A typical phase matching curve is shown in Figure 1, in this case for two IR materials AgGaSe2, and Zinc Germanium Phosphide, ZnGeP2- These are abbreviated as AGSe and ZGP. The "tuning curve" is wavelength dependent and given here for a 4um drive source. The generation of these curves and the wavelength range desired to be Phascmatching Curves for 7. = 4 \im ZnGcP =2 11 2'^inp ^ 12 ^ AgGaSe, 10 AgGaSc, •5 T c c00 / yp 40 50 60 70 6 8 10 12 Phascmatching Angle (Degrees) Idler Wavelength (nm) Figure 1. On the left, phase matching curves for two mid and far IR tunable OPO materials. On the right, the Non Linear Gain coefficient for ZGP. — 188- covered are needed before the crystal is cut and oriented for an end user. Unlike dye lasers, there are not widely available OPO tuning curves for various driver wavelengths and non-linear materials that are routinely seen hanging on the wall in a research laboratory. The dye laser clearly wins this "publicity" comparison. The second key element of the OPO is some estimate of the gain parameters. This is encapsulated in a effective non-linear coefficient that is related to the non-linear coefficients along the principal axes of the crystal projected onto the axis of propagation through the crystal. This material constant will vary for the wavelength used to drive the OPO and the wavelengths produced. The actual non-linear performance is often also summarized by a coefficient termed the "Non Linear Drive" [1,2] which is the effective net gain and saturation parameter. The gain coefficient C2, a material parameter, is multiplied by the medium length squared and the intensity that the crystal is driven at. The non-linear drive is related to the material parameters by the equation 2 2 2 1 1 2 C L I = Drive = 29.82x (deff) x (ry^n;)- ^ \\ )" L I where the refractive indices, ru s;j are those of the "pump" or driver wavelength, the signal and the idler and the wavelengths correspond to the signal and idler. It can be shown that the non linear drive relates to the conversion efficiency. In general the larger C2L2I, the greater the conversion efficiency. Regrettably L or I can not be driven up to arbitrarily large values. I is limited by damage to the materials, though in general it can be larger for short sub nano second pulses than for longer pulses. L is limited both by the size of crystal that the crystal growers can provide, and also the length over which the beams stop interacting. This phenomena, called "walkofT is due to the fact that the generated waves are angled to the pump to assure phase matching. Various walkoff compensation schemes, both macroscopic such as reversing the crystal orientation in multiple crystals, and microscopic schemes termed periodic poling. Some materials have a C2 so large that the OPO process is essentially superfluorescent in distances of under 1 cm. Many Optical Parametric Oscillator materials have a common problem; absorption of either the pump beam, the desired signal wavelength, or an idler wavelength. The absorption features of some of the newer OPO materials are shown in Figure 2. With all new materials, absorption properties can change in development of the growth process, especially in the transparent region where small impurities can increase the loss over the typical 0.1 to 0.5 % per cm loss that is seen in more mature materials. Often times the requirements of an application will drive the laser designer to use materials away from the region of best transmission. For example ZGP has very high non-linear drive and good transparency in the Mid-IR, but the logical laser drivers for this material, 2.1|.im from Tm or Ho lasers or from degenerately down-converted Nd, correspond to wavelengths of significant absorption. It is interesting to note that with BBO, KTA or KTP, and ZGP, that spectral coverage of all of the visible and most of the IR can be achieved. At lower powers, AGSe can complete the coverage in the IR. NLO plus OPO mixing schemes have demonstrated coverage from 250nm to the visible. — 189- ABSORPTION COEFFICIENT --- ZGP ZGP I KTA 1 BBO i BBQ ' i 1 1 KTA, ; i \ I in \ 1 in i u i i i ! 0.2 0.5 1 2 5 10 WAVELENGTH IN nm Figure 1. The absorption coefficient, a, for a few important OPO materials. Derived from published transmission data and materials lengths, a has units of cm"l for a loge attenuation law. The effects of reflectance have been included in the calculation. Data for materials is source and sample dependent. [3, 4, 5] 3. Thermal Considerations and Power Scaling The thermal load generated by absorption and finite thermal conductivity limit the thermal handling ability of these materials. A thermal figure of merit is defined to estimate the scaling of the first order thermal optical effect, thermal lensing. In real systems, higher order effects and optical and thermal anisotropy are important, but the thermal lens equation gives a rough comparison of materials. In certain approximations the optical power of the thermal lens is given by: 1/f = (Constant)x(Pa/ AL)x(L/K)x(dn/dT) Where f is the focal length, Pa is the absorbed power, A is the cross sectional area, L is the crystal length, K is the thermal conductivity, and dn/dT the temperature dependence of the index of refraction. For a simple comparison of the thermal capability of the various materials, we develop a thermal figure of merit, FOM, which is wavelength and pulse duration dependent. The power absorbed is related to the initial power, length, and absorption coefficient, a, at the wavelength(s) of interest. To make a fair comparison of the non-linear OPO materials we normalize to a constant non-linear drive, C^L^I. This value will vary by material and wavelength, but — 190- assuming a constant non linear gain equal for all materials being compared, the length term is removed from the equation to obtain 1/f = (const')x(P0/A)xFOM, where P0 is the incident power over the aperture irradiated, const' is ~ 4 and the FOM is given by: FOM = (a/K)x(dn/dT)x((C2I)-1/z) The intensity I is set at some operational range, limited by damage for each material and is a function of pulse duration. The FOM is obviously wavelength dependent in both a and the wavelength dependence of the damage, which relates to I. FOM is pulse'duration dependent through dependence of operating Intensity on the damage limit, and also in some cases where 2 photon absorption dominates in the pulse duration scaling of a. Table 2 on the following page gives the FOM for several materials at the wavelengths noted. A basic problem with the power scaling of Optical Parametric Generators is that the power production is limited by thermal focus, as well as higher order effects which often lead to beam break-up and damage. Moreover, many applications will drive the laser designer to use the material near an absorption edge, thus exacerbating the thermal limits inherent in a solid state material. As a rough comparison KTP and BBO in the visible, (and KTA operated away from the 4.5 to 5 micron edge) are roughly 10 to 20 X better (have less thermal lens) than KTA shifted to near its Mid-IR absorption edge, or ZGP used with a 2 micron drive laser. KTA or ZGP, with their losses, are each roughly 10X better than AGSe for generating outputs in the mid-IR spectral range. Measurements of thermal focus give specific guidance on the power limits of OPG crystals. AGSe excited with 2 micron radiation is limited to IR outputs of order of 2 to 5 W, though there is some belief that losses in the 2 micron range can be reduced. Average power handling of AGSe for processes that avoid the anomalous absorption at 2 micron may be a factor of 2 to 3 better. By comparison 2 micron driven ZGP should scale to the 10 W level, higher if driven at longer wavelengths. In the near IR spectral region, KTP and KTA can handle 100 W of drive power, and over 15 W has been demonstrated in a 1 micron driven degenerate KTP OPO. Experiments at higher powers have been driver limited. In the UV, linear absorption tells only a part of the story, as 2 photon absorption becomes the dominant loss mechanism, for example decreasing the FOM of BBO by an order of magnitude. Thus in a mixer to produce UV, average powers in the range of 5 Watts will start to show thermal problems. To work at or beyond these approximate thermal limits requires careful thermal and optical design. 4. Modeling of OPO's and Non-Linear Optical Systems With the resurgence of research interest in Optical Parametric generation, there has also been progress made in the development of models for these devices. Bowers and co-workers have developed a fully time dependent, model incorporating three spatial dimensions and the relevant - 191 - THERMAL FIGURE OF MERIT AND INPUT PARAMETERS MATERIAL AND (C2l)"^, a, K, dn/dT, FOM PARAMETERS cm"l cm"l W/cm °K oK-l BBO (vis/UV) 0.05 0.01 0.01 10"5 5xl0"7 ~10ns (~0.2if2PA) (2.5xlO-62PA) KTP (1 to 2^, 0.5 0.01 0.13 10"5 4xl0"7 ~10ns KTP (1 to 2^, 0.22 0.01 0.13 10"5 1.8xl0"7 -0.1ns KTA (1 to 3^, 0.22 0.02 0.13 10"5 3.6xl0"7 -0.1ns KTA (1 to 4.5n), 0.5 -0.61 0.13 10"5 2.4xl0"5 ~10ns KTA (1 to 4.5n), 0.22 -0.61 0.13 10-5 lxlO"5 -0.1ns ZGP (2.1 to 5n), 0.13 0.2 to 0.6+ 0.4 10"4 3 to 9xl0"6 ~5 ns ZGP (2.1 to 5n), 0.06 est. 0.2 to 0.6+ 0.4 lO"4 3to9xl0-6 -0.1ns AgGaSe2 (5to lOp.) 2 0.02 0.01 6xl0"5 24xl0"5 ~100 ns 5 5 AgGaSe2 (2.1 to 5n), 0.62 0.05 0.01 6xl0" 20x10' ~10ns 5 5 AgGaSe2 (2.1 to 5n), 0.22 0.05 0.01 6xl0" 7xl0" -0.1ns * Loss coefficient in ZGP is growth dependent in the 2 micron spectral range Table 2. Representative values of the thermal Figure of Merit for popular NLO materials. — 192- MMMtHy«>llH.MIMMUIMUIiJllliJIUm^ non-linear optical phenomena. [6] The model includes the effects of diffraction, walk off, cavity geometry, and includes capability to model curved and flat-flat OPO cavities. To date the model has been extensively applied to understand OPO performance in a seeded, 3 flat mirror ring cavity with KTP as the medium. [7] Energy outputs, spatial profiles, startup phenomena, reconversion and beam quality effects are all predicted with excellent comparison to experiments where measured. Behavior near threshold is modeled well. At several times above threshold there are small discrepancies between model seen in the KTP data. This is possibly due to other effects not included in the model, such as parasitics or possibly non-linear processes that are not parametric but are not currently in the model, for example 2 photon absorption of the pump or stimulated Raman scattering. Any given model data run requires about 1 to 2 minutes run time on a PC with a Pentium™ chip, somewhat longer on a 486 based computer. Comparison to other experiments has been limited to date. Good agreement, but not excellent agreement, was obtained for model predictions for published data on a 2 micron driven ZGP OPO. [8] The model predictions, and Mid-IR OPO's in practice, are sensitive to diffraction, and the initial simplifying assumption of a flat flat cavity undoubtedly reduced the effective recirculating flux in the ZGP OPO, whereas the confocal cavity of the experiment confined the generated light and extracted slightly more energy. 5. Generation of Visible and UV Near IR or visible parametric oscillators can be driven by the second harmonic of Nd:YAG, or the third harmonic as well. UV wavelengths are generated from these OPO by harmonic generation to provide tunable UV outputs. Alternately one can use the tunable laser output and mix it with the fundamental of the Nd laser or one of its harmonics. This technique can have much higher efficiency since the harmonic of the drive laser is typically generated at higher efficiency since there is no quantum defect. The latter technique has been used to generate blue light with very high efficiency in the mixing of Ti:Sapphire output with 1 micron output of a YAG laser. [9] The biggest drawback of mixing a Ti:Sapphire laser with YAG is the finite start up time, some times much longer than 10 ns. The OPO technique avoids some, but not all of this problem, with resultant simplification. The pulse duration matching of an OPO with the YAG or its harmonics is better, but could still be improved, as there is a finite number of OPO cavity round trips that are needed to bring the non-linear optical cavity into a mode of efficient extraction. This is area of OPO development where improvements are needed. One other very significant advantage of an OPO being mixed with the drive laser is that the tuning range of the OPO can be sufficiently large that there will be no holes in the spectrum of the OPO + mixer output. If the OPO driven by the second harmonic of a laser can be tuned from 1.5 times the energy of the drive laser down to the fundamental, the "idler" wave generated will be tuned from the fundamental to 0.5 the energy of the fundamental. When then mixed with the fundamental, and second and third harmonics, a very broad output range can be covered. Figure 4 shows the output of a BBO OPO and BBO OPA in the signal and idler wavelength regions for 100 mJ of green pumping the OPA and 25 mJ of green pumping the OPO. The data shows that the very broad spectral region from 700 nm to over 2000 nm is covered. The oscillator efficiencies for this OPO are about 30 %, though other OPO's can produce outputs at efficiencies of up to 65 %. The amplifier in this particular device is operating -193- at conversion efficiencies of order 35 % into the combination of signal and idler. Figure 5 also shows the UV generated by mixing the various signal and idler OPO / OPA waves with the second and third harmonic of YAG. Output energies in excess of 20 niJ throughout the UV have been generated in initial experiments. [10] 6. CONCLUSIONS Non-Linear Optics technology continues to make good progress. Power levels in certain materials will be able to be increased further, possibly to the 100 W level as has been achieved in second harmonic generation from YAG. Ideal candidates for such high powers would be KTP, KTA, and BBO when used in their regions of high transparency for both one and two photon absorption. Other materials will find increasing use in the Mid-IR and eventually the Far-IR, though the conversion efficiency from Nd lasers will for now be low due to the quantum defect. Novel techniques to recycle beams, such as ring converters with multiple non-linear optical materials, which have been applied to frequency multiplication are now beginning to be explored for tunable laser generation. New materials and improvements on existing materials continue to broaden the useful spectral range. For some applications the dye laser will continue as the laser of choice. However the steady improvement of solid state and non-linear optical devices points the way to continued use of all solid state solutions to a wider variety of applications. Idler Wavelength (nm) 2000 1600 1400 1200 1100 Figure 4 Measured visible and near IR outputs of a BBO OPO -«H a^ .--O' driven by a frequency OPA 100 mj Green Pump doubled Nd:YAG laser of the noted energies. The 5 30 outputs energy is for the sum of signal and idler. OPO 25 mJGreenPump 900 1000 Signal Wavelength (nm) -194- Figure 5. Initial UV mixing UV OUTPUT FROM MIXING STAGE data for a green driven OPO 40 mixed with the second and Mixing with 355 nm Mixing with 532 nm third harmonics of Nd:YAG. Q. The ouput data shows that S 30 /V %, e> jf* ^ one set of crystals can be \ / \ used to provide complete B \~* iS 20 r coverage of the near UV c from 250 nm to 400 nm. O ,0 Mixing 355 nm Mning355ri "i Mixing 532 nm | Mixing 532 nm with Signal with Idler 1wt h Signal with Idler 250 300 350 400 Output Wavelength (nm) - 195- Proceedings of the 6lh International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- HIGH POWER DIODE-ARRAY-PUMPED FREQUENCY DOUBLED ND:YAG LASER B. LE GARREC, Ph. FERU CEA/DCC/DPE/SPL Batiment 467 Centre d'Etudes de Saclay F-91191 GIF SUR YVETTE CEDEX FRANCE We report the demonstration of a transversally diode array pumped NdrYag laser using ten 15 watts CW linear diode arrays. At a 10 khz repetition rate the laser produces 5 watts average power at 532 nm when intracavity doubled with a KTP crystal leading to a 1.3 % electrical efficiency. Keywords: laser diode, Nd:Yag, second harmonic generation. 1. INTRODUCTION Efficient second-harmonic generation of high average power Nd:Yag laser is a difficult problem to solve. Non uniform temperature distribution in laser rods and thermally induced stress are leading to distorted wavefronts and low extra-cavity second harmonic efficiency. Intra-cavity second-harmonic generation [1] together with diode pumping [2] seems to be an alternate solution. We report the demonstration of a radially diode array pumped NdrYag laser. 2. DIODE PUMPED LASER MODULE Ten 15 watts cw diode laser arrays supplied by Spectra Diode Labs (SDL-3450-S) are optically coupled to a Nd:YAG rod (6mm diameter, 115 mm long) . Each 1 cm long diode array is mounted on a water cooled brass submount including the collimating cylindrical lens holder (see fig.l). The ten diode arrays are positioned at angles of 36° with respect to each other and arranged alternatively to pump the rod along 3 cm (see fig.2), each 1.5 cm segment of the rod is pumped by 5 diode bars. -196 — Nd:YAG LASER ROD Figure 1: diode array COUJMATINQ LENS mounted on a water cooled brass submount including the COOLINS WATER collimating cylindrical lens holder . LASEH DIODE ARRAY BRASS SUBMOUNT ND:YAG ROD CYUNDRICAL LENS r® © LINEAR DIODE ARRAY H Fig. 2. Cross sectional view of the diode pumped laser module showing the water cooled brass submount, the linear diode array and the collimating cylindrical lens. — 197- The rod is polished and is mounted inside a glass sleeve for water cooling. The outer part of the sleeve is partially coated with ten stripes of a highly reflective coating for pump light. Each stripe (1.5 mm wide and 15 mm long) is located on the other side of the sleeve with respect to the incoming light in order to reflect the diode light not absorbed by the rod at first pass. This geometry offers a number of advantages: a large fraction of the light is absorbed at both the 808.5 nm and the 812.5 nm absorption wavelengths of Nd3+ in YAG; the absorption band is larger (i.e a thermal drift of ± 1 Celsius degree does not affect the total absorptance). Our Yag rod was 0.8 % Nd3+ doped. At 25 A, the total emitted diode power is 145 watts and the pump density distribution exhibits a central peak as shown in the fluorescence profile (figure 3) at 1064 nm when recorded with a low aperture CCD camera [3]. This distribution has been verified through gain and threshold measurements in different 1064 nm CW cavity configurations. At the rod center, the total unsaturated E gain is equal to G0 = e °' = 1.25 ± 0.02 when the total pump power is 145 watts. -1 0 1 rod radius in mm Figure 3. Fluorescence profile at 1064 nm along the rod diameter, — 198— 3. LASER RESONATOR FOR XNTRA-CAVITY SHG A linear stable 50 cm long cavity including an acousto- optical Q-switch and a KTP non linear crystal has been designed (see figure 4). Both cavity mirrors are high reflectors at 1064 nm, the output coupler is flat and the other one is a lm radius concave mirror. An additional flat mirror high reflector at 532 nm is inserted in the cavity and acts as a dichroic output mirror. No wave plate is required because the KTP crystal itself acts like a retardation plate and the green output is highly polarized although the cavity is not polarized at 1064 nm. High reflector HlQh reflector @1064nm ©10B4nm 8*115 mm NttYag rod A.0 Q-switch KTP crystal ±- High reflector @ 532 nm Hlfljh transmitter ® 1084 nm te^ •^i 500 mm Figure 4: laser resonator for intra-cavity SHG. When repetitively Q-switched at a repetition rate between 7 and 12 khz, the laser produces more than 5 watts average power. The same results are obtained when the diode temperature is tuned to the 808.5 nm or the 812.5 nm absorption wavelength. An example is given figure 5 at 8 kHz. The laser pulse duration is ranging from 150 to 250 ns (FWHM) depending on pulse repetition rate, crystal length and cavity configuration. A maximum of 5.8 watts average power at 532 nm has been obtained at a 8 kHz repetition rate and a pulse duration of 24 5 ns (FWHM) at 145 watts pump 199 — power. The near field exhibits a smooth gaussian profile and the far field shows a 2.5 times diffraction limited profile. AVERAGE OUTPUT POWER AT 532 NM REPETITION RATE 8 kHz E y c CM If) 7*/ i • 3 a 1 . X < 20 40 60 80 100 120 U0 160 Total diode power in watts Fig. 5. Average output power at 532 nm and 8 kHz repetition rate as a function of total diode power with a 5 mm long KTP. The diode wavelength is kept constant at 808.5 nm. The slope efficiency is equal to 5.1 %. 4. FURTHER DEVELOPMENT. A new diode pumped laser module is under development. Thirty diode linear arrays are arranged radially around a single rod in a similar way as depicted above. With fifteen 20 watts cw diode laser arrays (SDL-3460-S), the total emitted power at 33 A is 300 watts and the folowing results have been obtained: - cw at 1064 nm: 82 watts in a 160 mm long stable cavity and a 10% transmission output coupler, - Q-switched at 10 kHz: 30 watts at 1064 nm in a 450 mm long stable cavity and a 10% transmission output coupler, - Q-switched at 8 kHz and intracavity SHG: 9 watts at 532 nm in a 450 mm long stable cavity. — 200 — REFERENCES 1) Ortiz M. , Fair J., Kuizenga D. : OSA Proceedings on Advanced Solid State Lasers 12, 361, 1992. 2) Hays A. , Marshall L. , Burnham R. :OSA Proceedings on Advanced Solid State Lasers .10., 255, 1991. 3) Le Garrec B., Feru Ph: ESI Publications Opto 92, 43 3, 1992. -201 — riucix-uings 01 1111: Din international symposium on Advanced Nuclcur Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- HIGH-POWER CO LASER AND ITS POTENTIAL APPLICATIONS Shunichi SATO, Kunimitsu TAKAHASHI, Kojiro SHIMAMOTO, and Yoichi TAKASHIMA Laser Laboratory, Institute of Research & Innovation 1201, Takada, Kashiwa-shi, Chiba 277, Japan Keiichi MATSUDA, Shizuma KUR1BAYASHI, Osamu NODA, Shigenori IMATAKE, and Motoe KONDO Takasago R&D Center, Mitsubishi Heavy Industries, Ltd. 2-2-1, Shinhama, Arai-cho, Takasago-shi, Hyogo 676, Japan The R&D program for the development of a high-power CO laser and its application technologies is described. Based on a self-sustained discharge excitation scheme, the available laser output has been successfully scaled to over 20 kW. The CO laser cutting experiments for thick metals have been performed in association with the decommissioning technologies development. Other potential applications, which include those based on photo chemical process, are reviewed. Recently demonstrated high-power tunable operation and room-temperature operation are also reported. Keywords: CO laser, cutting, decommissioning, tunable operation room-temperature operation 1. INTRODUCTION The CO laser can produce scalable output with high operating efficiency in the 5-Ltm-band spectrum region. Its shorter output wavelength, when compared with 10.6-u.m C02 laser, offers higher laser-to-materials coupling, and therefore, more efficient processing can be achieved especially in the field of metalworking such as cutting, welding, and surface treatment [1]. Potential of glass-fiber-based laser power delivery, which provides flexible processing, is another attractive issue in the CO laser applications [2]. The CO laser is featured, by not only its capability of producing high laser output but also, by its unique spectral characteristics. By introducing a line-selective cavity, tunable operation can be attained in the range of 4.9 - 6.3 p.m, which includes special lines used for isotope separation, crystal growth, etc. Although various excitation schemes have been investigated so far to obtain high laser output [3,4], only a limited number of apparatus have materialized for practical use. In our CO laser program, we first developed a 5-kW-class CO laser based on a self-sustained discharge excitation scheme [5], and this scheme has been scaled up to obtain a 20-kW-class output. In an early work with the 5-kW prototype laser, we demonstrated higher processing efficiency in — 202- cutting mild steel and stainless steel, with the CO laser than with C02 laser [6]. Attractive results were also obtained in processing of nonferrous metals [7]. Based on its excellent cutting capability, we consider that the CO laser can be an effective tool for decommissioning of nuclear power plants. For this purpose cutting of thick metals are now under test with the 20-kW CO laser developed. This paper reviews the current status of the MH1 / 1R1 CO laser program and potential applications of the CO laser. Recently demonstrated high-power tunable operation and room- temperature operation are also reported. 2. HIGH-POWER CO LASER DEVELOPMENT As mentioned above, we first developed a 5-kW-class CO laser and collected performance data [5], based on which we designed and constructed a scaled-up version ,i.e. a 20-kW-class laser. The specifications are given in Table 1 and the schematics of this apparatus are shown in Fig. 1. In the system, all components for gas circulation are installed in a 2.3-m-diameter, 6.8-m-long cylindrical vacuum chamber. Laser gas mixture is circulated with four axial blowers operating in parallel and cooled through two heat exchangers operating with liquid nitrogen (LN2) as coolant . The laser is excited by a transverse, self-sustained dc glow discharge which is produced between two-array hollow cathodes and a flat anode. An interelectrode gap length was selected to be the same as that of the 5-kW prototype laser, i.e. 6 cm, while a discharge length was scaled to around 4 m. Both a stable and an unstable resonators were tested, and it was confirmed that a steady, long-term operation was achievable at a 20-kW output power level even with a stable resonator. This is presumably due to lower absorption losses, both in bulk and coating of the ZnSe output coupler, at the CO-laser wavelength than at the C02-laser wavelength. The output performance is shown in Fig. 2. The output increases almost linearly with increasing discharge input and an output of 21 kW, which we believe to be the highest ever pure-cw CO laser output, is achieved with an electrical conversion efficiency of 22%. 3. APPLICATIONS OF HIGH-POWER CO LASER In early experiments with the 5-kW laser, we investigated cutting capability of the CO laser, and compared the results with that of the C02 laser. Figure 3 shows relations between maximum cutting speed and plate thickness for carbon steel and stainless steel [6]. A higher cutting performance is demonstrated with the CO laser than with the C02 laser. Although there is an uncertain factor relating to the difference in the beam quality of the lasers used in the experiment, higher cutting performance with CO laser has been confirmed in subsequent -203- experiments conducted by other groups [8,9]. Cutting of nonferrous metals was also performed and high quality cutting has been obtained for, e.g. a 10-mm-thick copper plate [7]. Based on these excellent cutting capability, we consider that the CO laser can be an effective tool for decommissioning of nuclear power plants. Cutting tests for thick metal plates have been performed for this purpose. An example is given in Fig. 4, where the stainless steel of over 100-mm thickness is successfully cut [10]. Cutting of much thicker plates are now under test by using the above-mentioned 20-kW CO laser. The CO laser has also been used by other groups for welding and surface treatment of metals [1,11] as well as crystal growth for A1203, YAG, etc. [1,12]. 4. RECENT TOPICS 4.1. High-power tunable operation Another attractive feature of the CO laser is its unique spectral characteristics, in addition to its available high laser output. By introducing a wavelength-selective cavity, tunable operation can be obtained in the spectral range of 4.9 - 6.3 u.m. Figure 5 shows the typical tuning characteristics and lines used for some potential applications based on photochemical processes [13-17]. Because there are no other high power laser sources in this wavelength region, we are encouraged to apply CO laser to these valuable applications. 4.2 Room-temperature operation Although high laser outputs have been achieved with various excitation schemes, operations at cryogenic temperatures have so far been required. However, recent efforts have enabled us to achieve an efficient, high-power room-temperature operation [18-20]. This has been accomplished by combining rf or microwave discharge excitation with convective gas cooling, but dc discharge excitation may be used. High specific laser outputs, which are comparable to those of typical industrial CO, lasers, have been achieved. Because of the scalability of the excitation scheme used, construction of a multikilowatt room-temperature CO laser is feasible. 5. SUMMARY The MHI / IRI CO laser program and the potential applications of the CO laser, as. well as some recent topics, have been briefly reviewed. The self-sustained discharge excitation scheme has been successfully scaled up to obtain an output over 20 kW. Such-high power CO lasers should be an effective tool for heavy metal processing, including decommissioning of nuclear power plants. Other applications which use the line-tunability of the CO laser, suchas isotope separation, crystal growth, etc. are also promising. The efficient room-temperature operation — 204- with high specific output has been demonstrated. Construction of a multikilowatt room- temperature CO laser is feasible. ACKNOWLEDGMENT The authors would like to acknowledge Prof. Tomoo Fujioka for his continuous encouragement and valuable discussions. REFERENCES 1) Maisenhalder F. : "Proc. of the Int. Symp. on Laser Advanced Materials Processing, Nagaoka, 1992" ed. Matsunawa A. and KatayamaS., Osaka, 43 (1992). 2) Sato S, Tanimito K et al.: Appl. Phys. Lett., 62 , 669 (1993). 3) Averin A. P. et al.: Sov. J. Quantum Electron., 13, 1391 (1983). 4) Billow H. von and Schellhom M.: Appl. Phys. Lett., 63, 287 (1993). 5) Sato S., Kondo M. et al. : Rev. Laser Eng., 18, 118 (1990) [in Japanese]. 6) Sato S., Shono S. et al.: "Proc. ICALEO '88, Santa Clara, 1992" (1988). 7)Wignarajah S., Shono S. et al. : "Proc. of the 5th Int. Symp. of the Japan Welding Society, Tokyo, 1990" 35 (1990). 8) Maisenhalder F. and Stohr M.: Laser und Optoelektronik, 23 , 40 (1991) [in German]. 9) Kanazawa H. et al. : Rev. Laser Eng,, 2J., 376 (1993) [in Japanese]. 10) Miya K., Sugihara ML, Onozawa T., Miwa T., and Matsumoto O. : "Proc. SFEN Int. Conf., 1992" (1992). 11) Rudlaff Th., Satani R. et al.: Laser und Optoelektronik, 23,46 (1991) [in German]. 12) Muckenschnabel J. et al.: Laser und Optoelektronik, 23,49 (1991) [in German]. 13) DeLeon R. L. and Rich J. W.: "Proc. LASERS'86" 37 (1986). 14) Rebello J. H. D., Stranb D. L., and Subramaniam V. V. : J. Appl. Phys., 72, 1133 (1992). 15) Abzianidze T. G. et al. : Sov. J. Quantum Electron., 16, 137 (1986). 16) Herman I. P. and Marling J. B. : J. Phys. Chem., 85., 493 (1981). 17) Garside B. K. et al. : Appl. Opt., 16, 398 (1977). 18) Luo X., Schafer J. H., and Uhlenbusch J.: Opt. Commun., 102, 65 (1993). 19) Uehara M. and Kanazawa H.: "Ext. Abstr. 54th Autumn Meeting of the Japan Society of Applied Physics, Sapporo, 1993" 907 (1993) [in Japanese]. 20) Shimizu K. and Sato S. : Jpn. J. Appl. Phys., 33, L180 (1994). — 205- Table I Specifications and operating conditions ofthe20-kW CO laser. Target performance Output power •> 20 kW Conversion efficiency >20% Excitation Type of discharge transverse dc glow discharge Discharge gap length 6.2 cm Discharge length A m (effective 3.7 m) Optical cavity Type stable / unstable Output window solid / aerodynamic Cavity length 8m Optical aperture 57.5 mm x 72 mm Operating gas conditions Gas mixtures CO/NJ/HC/O: Gas temperature 150 K Gas pressure 40 Torr Gas flow velocity 20 - 25 m/s DISCHARGE SECTION 25 I ' ' ' I \I ;I ' i I 5 20 - - HONEYCOMB 5 15 -j- - S /tf i 12 HEAT EXCHANGER o VACUUM VESSEL "•10 DC yy . : Ul : I //* '• HEATER 'Ms ' —•—22 m/s ./'s i —•—25 m/s . A I I 1 HEAT EXCHANGER i . • . . I ... iI ... I ... i ... •. 20 40 60 80 100 120 NOTE : Gai (low path Indicated BLOWER with arrowi DISCHARGE POWER (kW) (a) # 25 1 ' ' 1 ' ' ' 1 TR MIRROR INSULATOR PR MIRROR / CATHODE / ANO0E ! 0ATE VALVE O 20 ^1^.1 ...._ '• ~. Zs —*- 0PTOU. ATOUllK uj 1 | o t i rr 15 • / m Gas flow rate • 1 i z 10 J v o —•—22 m/s : 55 • 25 m/s c 1 • > i | z 1 HEAT EXCHANGER o u 20 40 60 80 100 120 DISCHARGE POWER (kWJ (b) Fig. 1 Schematics of the 20-kW CO laser. Fig. 2 Output performance of the 20-kW CO laser. — 206- 'a 'u 2.0 MIO STEEL STAINLESS STEEL CO _COj_ Max. cutting i»rt 2.0- «... umssKt* -52- fil. POUttf ilTOKSA> pauitlf irrgaatit Cgttirq abity (CO) e o , Cutting ability (COI •? s 15 t 1.0 \ i CO? (SkWI 1.0 1 1 \\ / COj U«WI 1 \Y / • < 0.5 / C\ CO UkVI 0.5 v \y /-• < 0 N.\ / \° Ok V-4—S 10 20 30 W> 50 60 70 60 10 20 30 40 PLATE THICKNESS (mm)' PLATE THICKNESS (mm) Fig. 3 Comparison of cutting capability between CO and C02 laser [6]. CO laser Laser power 11kw \ Malarial AISD04 In air cultlncj \ Focal length 1000mm \ Focusing point +10mm \ Assist Gas 0, 1 \ gas hlllMII 1MPa 200 \ \ -a \ OJ a 111CW/ w 100 Coaxial noula ^ aid* nezz i\ Aoeil length \ S ^ 508mm J S o i r I i i i i i i i. 30 40 50 60 70 80 90 100 110 120 130 140 150 Thickness (mm) Fig. 4 Cutting of stainless steel over 100-mm thickness with the CO laser beam [10]. 6.2 6.1 6.0 5.9 5.8 5.7 5.6 5.5 5.4 5.3 5.2 5.1 5.0 WAVE LENGTH ( pi m ) A CO V-V pumping applications (diamond growth, ,3C, 160 enrichment) B 235U enrichment (CRISLA) C ,3C, ,eO enrichment (multiphoton dissociation of COCI2) D deutrium, tritium enrichment (multiphoton dissociation of CHFa) E remote sensing of NO Fig. 5 Output spectrum of the tunable CO laser and its potential applications. — 207— • iv«-iiuHi£.> ui uic uui iiiiL-rnitiionui symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - Compact Soft X-Ray Laser Development Tamio HARA, Kozo ANDO and Yoshinobu Aoyagi The Institute of Physical and Chemical Research (Rikcn) Wako-shi, Saitama 351-01, Japan Using a pulse-train laser with only 2.3 J/cm pumping energy, we succeeded in observing ASE from two Li-likc Al Iincs^ (105.7 A and 154.7 A). The estimated gains for 105.7 A and 154.7 A were 2.0 cm"1 and 1.5 cm"1, respectively. The experiment of double pass amplification was carried out using a Mo/Si multilayer mirror with the reflectivity of 34 % for 154.7 A. When the shutter of the mirror was open, the line intensity doubled. Keywords: X-ray laser, Compact X-ray laser, Pulse-train laser, Li-likc Al line, Rccombining plasma 1. INTRODUCTION X-ray lasers have many promising applications such as X-ray microscopy, X-ray holography and X-ray lithography. However, the current X-ray lasers need huge laser systems as a pumping source. For many practical applications, a table-size soft X-ray laser should be developed. Recently we have observed soft X-ray amplified spontaneous emissions (ASE) in rccombining Al and Si plasmas produced by a low power driving laser which has a pulsc-train-likc pulse form [1-3]. In the rccombining plasma scheme, a key technology for down sizing of X-ray lasers is that a high density of multiple charged ions should be produced as efficiently as possible. Besides, the faster cooling of the electron temperature, Tc, of the plasma leads to the higher recombination rate and to the larger population inversion, because the recombination rate depends strongly on Tc. So the control of plasma production and recombination is important for improvements of pumping efficiency and development of a compact X-ray laser. 2. EXPERIMENTAL To study in detail the effect of multi-pulse irradiation on plasma production and on X-ray emission including ASE, we constructed a pulse-train glass laser system. Here a 100-ps laser pulse from a mode locked oscillator injects into an optical pulse stacker [4] to get an 8- pulsc train. This is doubled by a Michclson-typc system to make 16 pulses [5, 6]. Each pulse has almost the same peak power and the same interpulsc time. The pulse train was amplified by glass amplifiers and line-focused on an Al slab target by a lens system. For rapid cooling of the plasma, the width of the focus line was set at less than 40 fim. This — 208 — narrow width leads to a large volume ratio in plasma expansion. Temporal and spatial behavior of the emitted soft X-ray spectra were observed along the axis of a line plasma by an XUV flat-field spectrograph, which has an X-ray film or a photocathodc coupled to a streak camera. 3. PLASMA HEATING BY A PULSE-TRAIN LASER It is important to study abundances of charged ions as a function of time interval between pulses. Time integrated line intensities of two different ions (Al VI, Al XI) arc plotted as a function of intcrpulsc time for a 4-pulsc train laser in Fig. 1. Line intensities are normalized at the intcrpulsc time of 400 ps. It is apparent that the intensity of the Al XI 105.7 A line increases when the interpulsc time is decreased. The intensity of lower ionized Al VI line at 113.4 A, on the other hand, decreases as the interpulse time is decreased. This result shows that the intcrpulsc time has an important role in the production of highly charged ions. The variation of time integrated intensity of the Al XI 105.7 A line as a function of the number of pulses was investigated under the condition that the intcrpulse time was kept at 200 ps. The behavior of X-ray intensity divided by the number of laser pulses, which is related to the conversion efficiency from a pulse-train laser to X-ray shows that highly charged ions such as Al10+ arc produced efficiently by a train of successive short laser pulses. The above results lead to an understanding of the production mechanism of highly ionized atoms. Each pulse may act as a prc-pulsc for the next incoming pulses, of which energy is absorbed by the prc-formed plasma, and therefore raises the electron temperature effectively. As a result, successive heating of plasma may occur near the target surface with increasing number of pulses. In this consideration, interpulsc time is an important factor in order to attain higher electron temperature of the plasma, because such plasma heating must be done successively before the electron temperature decreases due to adiabatic expansion. The other advantage of the pulse-train laser is that the electron temperature raised effectively by successive heating in turn drops rapidly as soon as laser irradiation ceases. This is because the fall time of the train laser is the same as that of the last pulse in the train and is shorter than that of a smooth single pulse. Therefore the use of a pulse-train laser instead of a long pulse laser should be a powerful method to achieve high gain through recombination process. 4. ASE EXPERIMENT USING A PULSE-TRAIN LASER We have succeeded in observing the ASE signal of two Li-like Al lines (105.7 A 3d-5f and 154.7 A 3d-4f) with only 2.3 J/cm pumping energy from a pulse-train glass laser. - 209 — o Temporal behavior of an amplified 154.7 A line is shown in Fig. 2 for plasma lengths of 6 mm and 12 mm. The observation was made at the distance of 700 /nm from the target surface. It is evident that the intensity emitted from L= 12 mm plasma was higher than that from L = 6 mm plasma by a factor of 3 around the peak of the X-ray .intensity where amplification occurred strongly. The duration of gain is about 1 ns. A time integrated soft X-ray spectrum over the duration is shown in Fig. 3. After subtracting the background level • • • o o from the observed line intensities, the intensity ratio for 105.7 A and 154.7 A arc about 4. The estimated gains for 105.7 A and 154.7 A were 2.0 cm"1 and 1.5 cm"1, respectively. The experiment of double pass amplification was carried out at the same position using a Mo/Si multilayer mirror with the reflectivity of 34 % for 154.7 A. Fig. 4 and Fig. 5 arc a Q typical time history of the 154.7 A line and X-ray spectra integrated over gain duration, respectively. When the shutter of the mirror was open, the line intensity doubled. This result strongly supports the cxistancc of gain for the 154.7 A line. Based on these results, we built a new pulse-train Nd:YAG laser system and carried out ASE experiments on Li-like Al lines. A gain of 2 cm"1 at 105.7 A was observed with 1 J/cm pumping energy from the YAG laser at a distance of 200 fim from the Al target surface. It is observed that the gain coefficient is improved up to over 4 cm"1 when the laser intensity of the second half of the pulse train is reduced to 25 % of the first one. If an X-ray cavity is constructed with multilayer X-ray mirrors, a first X-ray oscillator will be realized, which leads to great improvement of coherence, pumping efficiency and output energy of X- ray lasers. 5. SUMMARY Using a pulse-train laser with only 2.3 J/cm pumping energy, we succeeded in observing ASE from two Li-like Al lines (105.7 A and 154.7 A). The estimated gains for 105.7 A and 154.7 A were 2.0 cm'1 and 1.5 cm"1, respectively. The experiment of double pass amplification was carried out using a Mo/Si multilayer mirror with the reflectivity of 34 % for 154.7 A. When the shutter of the mirror was open, the line intensity doubled. It is also observed that the gain coefficient is improved up to over 4 cm"1 when the laser intensity of the second half of the pulse train is reduced to 25 % of the first one. Development of compact X-ray lasers will promote new fields of practical applications. REFERENCES 1) HaraT., Ando K., Kusakabc N., Yashiro H. and Aoyagi Y.: Jpn. J. Appl. Phys., 28, L1010 (1989). — 210 — 2) Hani T., Ando K., Yashiro H. and Aoyagi Y.: "Proc. Uth Int. Conf. on X-Ray and Inner-Shell Processes" cd Carlson T. A., Krause M. 0. and Manson S T., American Institute of Physics, 197 (1990). 3) Yashiro H., Hara T., Ando K., Ncgishi F., Ido S. and Aoyagi Y.: Jpn. J. Appl. Phys., 3_LL92(1992). 4) Danson C. N., Edwards C. B. and Ross I. N.: Optics and Laser Technology, VL 99 (1985). 5) Hirosc H., Hara T., Ando K., Ncgishi F. and Aoyagi Y.: Jpn. J. Appl. Phys., 32,. L1538 (1993). 6) Hara T., Ando K., Negishi F., Yashiro H. and Aoyagi Y.,: "X-Ray Lasers 1990 " cd Tallcnts G. J., IOP Publishing Ltd, Bristol, 263 (1990). 2 T • , , ,- , I—,-r I , [L ' ' • i • • • . ' tn • c -Al XI 105.7 A Laser pulse 800 -< >• T=200 psec N=16 pulses E=2.3 J/cm z=0.7 mm 400 ILs© « 200 ^aixiJUu Afciita* 4 5 8 9 t (nsec) Fig. 2 Time history of the Al XI 154.7 A line observed at the distance of 0.7 mm from the target. -211- 12000 I— gooo 6000 3000 Wavelength (A) Fig. 3 Time integrated spectra for the plasma lengths of 6 mm and 12 mm, which were observed at the distance of 0.7 mm from the target. Laser pulse 1 ns 4000 T=200 psoc CD N=10 pulse 3000 ~F.=2.1 J/cm" MJTtnsir ©IP®05 z=0.7 mm 2000 Intensit y 1000 ./Wv Fig. 4 Time history of the Al XI 154.7 A line observed in the double-pass experiment using a Mo/Si multilayer mirror. 60000 T=200 psec N=16 pulse 45000 E=2.1 J/cm z=0.7 mm 30000 15000 Wavelength (A) Fig.5 X-ray spectra integrated over the gain duration. -212 — rroccuiings 01 inc om iniernational symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Wavefront Compensation Applied to AVLIS Laser Systems Thomas Gonsiorowski and Allan Wirth United Technologies / Adaptive Optics Associates 54 CambridgePark Drive, Cambridge, MA 02140-2308 USA Tel (617) 864-0201, FAX (617) 864-1348 ABSTRACT The efficiency of an AVLIS system depends upon the power density and uniformity of the laser system. Because of wavefront aberrations the realized beam quality is not ideal. Wavefront compensation provides a means to improve beam quality and system efficiency. INTRODUCTION Atomic Vapor Laser Isotope Separation (AVLIS) systems which employ a copper vapor laser (CVL) as the pump source for a dye laser (DL) amplifier are under development. In these systems there exist many effects which reduce optical quality of the laser beam and thus degrade AVLIS performance. First, the output beam leaving the DL amplifier chain is aberrated. These aberrations are due to errors in the CVL beam which "print through" the amplification process, as well as to optics and atmospheric effects in the DL optics and also to turbulence in the dye flow field. Additional aberrations occur in the DLbeam delivery optics, again due to distortions of the optical elements and turbulence within the beam path. Lastly, the path through the separation cell itself contributes to the aberration of the beam. Within the cell, in addition to the thermal and diffraction effects, the beam also is degraded by the atomic absorption and vapor density variations across the cell [1]. Real-time wavefront compensation offers the possibility for active maintenance of the laser beam phase both spatially and temporally. Real-time wavefront compensation systems have been in use for correction of atmospherically induced aberrations for two decades. Similar systems have been fabricated to compensate high energy laser systems for laser-internal beam and thermally induced errors. Thus, the technology presently exists that would allow correction of the beam quality in the AVLIS system. Such a system, capable of correcting only low order spatial and temporal aberrations, has been in operation at the Lawrence Livermore National Laboratory (LLNL) AVLIS system for some time [2]. — 213 — In this paper, the current performance capabilities of wavefront compensation systems are compared to the requirements of the AVLIS application. From this comparison, a concept for a beam control system is developed. Special attention will be paid to the key parameters of the adaptive optical system, namely correction bandwidth and number of channels. Finally, a low- cost diagnostic wavefront sensor which can be upgraded to support a complete active wavefront compensation system will be described. ADAPTIVE OPTICAL SYSTEMS An adaptive optical system includes a means by which the optical propagation path may be modified both spatially and temporally in real time. The use of an adaptive optical system to correct, in real time, for atmospheric disturbances in astronomical observations was proposed in the 1950s [3,4]. With the advent of laser devices another potential use of adaptive optics, to correct for aberrations of the laser source and the laser beam delivery system, was also proposed. In the last several decades extensive research and development of adaptive optical systems and the associated components technologies has occurred [5, 6, 7]. Most of this development has been for the purposes of astronomy or ground based surveillance/imaging of orbiting satellites. However, because the critical design parameters (i.e., number of channels, control bandwidth and aperture size) of those systems meet or exceed the needs of adaptive optics for a laser beam delivery system the developments are equally applicable. There are three principal elements of an adaptive optical system: a) a wavefront sensor which measures the error of the optical wavefront; b) a control processor which converts the measurements of the wavefront sensor into correction signals; and c) a deformable mirror which converts the correction signals to optical surface deformations which are thus applied to the optical wavefront. These elements are usually operated as a closed loop control system whereby the corrections introduced by the deformable mirror are subsequently measured by the wavefront sensor. The control algorithm attempts to minimize the errors measured by the wavefront sensor and in so doing minimizes the residuals errors on the optical wavefront leaving the deformable mirror surface. A simple block diagram of the adaptive optics for a laser beam delivery system is shown in Figure 1. The two parameters which most significantly define an adaptive optical system are the number of channels and the control bandwidth or update rate. The term channels refers most directly to the number of actuator elements in the deformable mirror. It also roughly represents the number of subapertures or sampling elements in the wavefront sensor. The correspondence is fuzzy because slight oversampling with the wavefront sensor is often used to improve system stability. An adaptive optical system designed for atmospheric - 214 — compensation would have 200-1000 channels and operate at 100-150 Hz. However, for correcting the errors in a laser beam delivery system 60-100 channels should suffice and control bandwidths probably would not exceed 80 Hz. Typically the wavefront sensor and deformable mirror must support update rates of 10-15 times the control bandwidth to insure adequate stability of the control system. LASER APPLICATION 1 \r> / WAVEFRONT >V, SENSOR // , BEAM . LASER J&> DEVICE /DELIVERY; ^ ' SYSTEM * "Vf7^ vv Y DEFORMABLE CONTROL MIRROR PROCESSOR Figure 1. A Simple Schematic of an Adaptive Optical System. The most common method for measuring the errors of a wavefront is by interferometry in which the wavefront to be tested is combined with a high quality plane wave. The resulting interference fringes encode the OPD of the test wavefront. Typically the wavefront sensor in an adaptive optical system does not have access to a high quality plane wave which is coherent with the test wavefront. In addition, the wavefront sensor rarely has sufficient signal to construct a plane wave internally (e.g., by spatially filtering a portion of its input). To overcome this constraint wavefront sensors interfere the input wavefront with a modified copy of itself (typically the copy is laterally offset or "sheared"), thereby encoding the phase information into an intensity fringe pattern [8, 9]. Such devices are referred to as shearing interferometers. Another method for measuring the wavefront errors derives from the classical Hartmann test [10]. In this approach, the aperture of the wavefront sensor is subdivided in many small subapertures. In each subaperture a simple lens form a focused spot the position of which encodes the errors of the incident wavefront. A pictorial representation of the Hartmann method is shown in Figure 2. The Hartmann sensor requires precise and stable alignment of the subaperture lenses and the detector plane which only became practical with the development of microlens array technology [11]. - 215- Whether the wavefront errors are measured using a shearing interferometer or a Hartmann sensor, the characteristic size and number of the sampling subapertures will be set by the detector pixels. Similarly the sampling rate and ultimately the control bandwidth of the adaptive optics are set by the detector frame rate. Thus high frame rate, large area focal plane array (FPA) detectors represent a critical technology for the development of practical adaptive optical systems. Advances in silicon FPA technology based on charge-coupled devices have yielded detectors with the necessary characteristics. A typical adaptive optical system might use a 1024x1024 FPA framing at 50-100 Hz. For atmospheric compensation systems higher frame rates are necessary and subsequently the array size is reduced. Figure 2. The Hartmann Wavefront Measurement Method. The second element of the adaptive optical system which we will discuss is the deformable mirror (DM). The critical function of the DM is to covert electrical signals into motion of an optical surface. Most often piezoelectric type materials are used to construct small actuators whose length can be modified by the application of voltage. These actuators are then bonded to thin membrane which is fabricated with a reflective, high quality optical surface on the opposing face [12]. Other technologies for deformable mirrors have been investigated with less success. Unlike FPAs which benefit from the miniaturization associated with integrated circuit technology, DMs are fabricated with relatively large scale components. Thus a typical 100 channel adaptive optical system would use a Hartmann sensor with subapertures measuring a few hundred microns but with a DM which has 5-10 millimeter actuator separation. Of particular importance to the application of adaptive optics to laser beam delivery systems is the power loading capacity of deformable mirrors. Deformable mirrors have been constructed with a series of water channels integrated into the mirror membrane to remove heat - 216 — deposited into the mirror under high power applications [13]. Such a deformable mirror could accommodate the power levels encountered in a typical materials processing laser system. Of the three primary elements of the adaptive optical system, the control processor is the most complex. This is because the control processor must perform a myriad of functions. In the real time data path the control processor performs a series of algorithms to convert the wavefront sensor measurements into control signals for the deformable mirror. These algorithms are depicted in Figure 3. First, the processor must convert the pixel intensities into a wavefront phase metric. For a shearing interferometer this conversion takes the form a multi- bin phase detection algorithm. In the case of a Hartmann sensor a centroid location algorithm is used. In all cases, the wavefront measurement data must be corrected for instrument errors. This may include correcting for response variations in the FPA as well as subtracting the phase errors introduced by the wavefront sensor optics. REAL TIME DATA PATH FROM CENTROID INSTRUMENT ERROR SENSOR LOCATION SUBTRACTION ST Q .x; PHASE ERROR CONTROL TO DM Q RECONSTRUCTION FILTERS USER INTERFACE and EXECUTION CONTROL Figure 3. Schematic of Control Processor Elements. Both a shearing interferometer and a Hartmann sensor result in a wavefront metric which is related to the derivative of the phase errors; therefore the second algorithm step involves reconstructing the phase errors from these derivatives. This is usually accomplished by an inverse matrix multiplication. The matrix includes not only the relation between sensor measurements and phase errors but also the relation between deformable mirror actuator motion and the optical phase introduced by the deformable mirror surface. The second relationship is included because moving a single actuator usually produces a non-localized mirror deformation. This deformable mirror characteristic is called the influence function. The output of the reconstruction algorithm represents the instantaneous errors at each of the deformable mirror actuators. The final algorithm step is a control filter which converts the -217— error signal to actuator commands. The filter algorithm is typically a simple integrator and low pass filter. However, the control algorithm may, in addition, remove or modify low order spatial modes such as tilt and focus either to apply these to separate special purpose active optical elements or to otherwise modify the system optical output. In addition to the real time data processing, the control processor must also coordinate software execution and system configuration. In typical operation, wavefront sensor and deformable mirror calibration data must be collected, post processed and stored for use within the real time algorithms. Also one or more sets of reconstruction matrices and control filter coefficients must be loaded and potentially switched during operation. Finally, given the complexity of an adaptive optical system, the control processor usually includes built-in diagnostics for the processing software as well as for the wavefront sensor and the deformable mirror. Because of the extremely high data throughput of the real time processing algorithms, a special purpose processing architecture is needed. Early adaptive optical systems relied on highly parallel analog electronic circuits to meet this demand. But the tremendous advances in digital integrated circuits has allowed more flexible digital signal processing architectures to meet the adaptive optics processing demands. Still these architectures rely on special purpose VLSI devices to handle the large amounts of data input/output associated with adaptive optics algorithms [14]. Only recently has commercial processing hardware become available which can meet the high data throughput demands of the adaptive optical system processing. APPLICATION TO AVLIS SYSTEMS Virtually all laser systems exhibit some degradation of their beam quality from the diffraction limited ideal. There are many potential causes for the corruption of the phase and intensity distribution of a laser beam. They may be broadly categorized as either static or temporally varying and further as to the relative strength of their impact on the phase and/or intensity of the laser beam. An example of a static defect might be a manufacturing error in one of the beam delivery optics. This leads to an error in the phase of the laser wavefront. Depending upon the effective optical propagation distance from the aberrating optic to the isotope separation cell such a phase error may appear as either a phase or intensity error in the beam. A typical example of a temporally varying beam error is the aberration induced by the heating of the optical element by the impinging laser beam. This heating bends the optical elements thus distorting of the laser wavefront. The time scales for these effects are typically -218- fairly slow (several seconds). Other thermally driven effects involve the generation of turbulence in fluid portions of the beam path. Depending on the type and amount of beam path conditioning these effects will exhibit shorter time scales on the order of 10-100 milliseconds. Table 1 below summarizes the typical characteristics of a number of aberrations that may affect a materials processing laser system. Comparison of adaptive optical system performance capabilities and the data in Table 1 shows that the range of aberration strengths and time scales that might be encountered in an AVLIS laser system are within the region for which adaptive compensation is possible and has been demonstrated. Table 1. Typical Aberration Characteristics. Aberration Type Strength (X) Time scale (sec) Example Optics Errors 0.25 - 2.0 Static Fabrication Tolerance Laser Quality 0.1 - 5.0 Static Beam Divergence Turbulence, Intracavity 0-1.0 0.01 -1 Dye flow field Turbulence, Beam Path 0.1 - 2.0 0.01 -1 Atmospheric Turbulence Thermal, Optics 0-30 0.01 -10 Bending due to Heating Thermal, Laser 0-2.0 0.01 -10 Mechanical drift of cavity The anticipated configuration of an AVLIS system which incorporates adaptive optical elements is shown in Figure 4. The primary adaptive compensation would correct the laser wavefront propagating in the isotope separation cell. This single compensation loop could correct both wavefront errors on laser beam leaving the DL and errors introduced by propagation through the separation cell. An optional, additional compensation loop might be used to correct for wavefront errors on the CVL beam before DL amplification. Such a loop could improve the beam quality and amplifier efficiency of the DL. Returning our attention to the characteristic time scales described in Table 1 and noting that beam path turbulence would present the most rapidly varying disturbance, we would expect to require an adaptive compensation control bandwidth of 80-100 Hz. Similar review of the disturbance strengths would suggest that optic bending will dominate the wavefront profile and might result in as much as 10 waves of OPD error. However, this error will be a low order spatial mode and will therefore drive the dynamic range of the adaptive optical system. The number of channels in the adaptive optical system will be governed by the disturbances with the highest spatial frequencies which again will most likely be beam path turbulence. Without detailed knowledge of the AVLIS system beam path conditioning we can also speculate that a 60-100 channel adaptive optical system would be required. — 219- SEPARATION CELL WAVEFRONT SENSOR \ WAVEFRONT CONTROL SENSOR PROCESSOR DYE LASER AMPLIFIER CONTROL PROCESSOR DEFORMABLE A MIRROR CVL DEFORMABLE MIRROR OPTIONAL -^^%^^^^$w~&?£8&^^*^^fc^$ffijf Figure 4. Proposed AVLIS Adaptive Optical System. United Technologies/Adaptive Optics Associates has recently delivered a diagnostic wavefront sensor and data processor for installation into an AVLIS laser system. This system uses a Hartmann vvavefront sensor to measures the incident wavefront on a 32x32 subaperture grid. The wavefront sensor utilizes a fast framing 256x256 silicon detector array which frames at up to 500 Hz. The subaperture data is stored and then processed to obtain high accuracy vvavefront error measurements. Since the instrument will initially be used as a diagnostic tool the data processing is performed on a general purpose computer workstation. In the future this system can be expanded with additional processing equipment and a deformable mirror to provide a complete real time adaptive optical compensation system. SUMMARY We have described the construction and performance capabilities of typical adaptive optical systems developed for atmospheric compensation. Upon reviewing the anticipated sources of beam degradation in an AVLIS system, we find that adaptive optics can compensate the laser wavefront in these systems. The anticipated requirements for such an adaptive optical system fall well within the current capabilities of adaptive optics technology. We have recently constructed a diagnostic instrument which can be upgraded to provide a full real time adaptive optical system of this type. We anticipate that the diagnostic instrument will enable evaluation of the key system parameters, namely the temporal and spatial frequencies of the optical — 220 — disturbances affecting AVLIS systems, which will ultimately govern the specification of the adaptive optical system requirements for this application. REFERENCES 1. Morioka, N., "New Japanese AVLIS Program", in Laser Isotope Separation, Proc. S. P. I. E., 1859, (1993). 2. Bass, I. L., Bonanno, R. E., Hackel, R.P., and Hammond, P.R., "High-average-power dye laser at Lawrence Livermore National Laboratory", Applied Optics, 31, pp. 6993- 7006 (1992). 3. Babcock, H. W., "The Possibility of Compensating Astronomical Seeing," Publ. Astronomical Society Pac, 65, pp. 229-236 (1953). 4. Babcock, H. W., "Deformable Optical Elements with Feedback," J. Optical Soc. Amer., 48, pp. 500-507 (1958). 5. Hardy, J. W., "Active Optics: A New Technology for the Control of Light," Proceedings of the IEEE, 66, pp. 651-697 (June 1978). 6. Tebo, A., "Adaptive Optics," OE Reports, 96 (December 1991). 7. Tyson, R. K., Principles of Adaptive Optics, Academic Press, San Diego, CA, 1991. 8. Wyant, J. C, Use of an AC heterodyne lateral shearing interferometer with real-time wavefront correction systems," Appl. Opt, 14, pp. 2622-2626 (1975). 9. Sandler, D. G.3 et. al, "Shearing Interferometer for laser guide-star atmospheric correction at large D/r0," J. Opt. Soc. Am., 11, No. 2 (Feb. 1994). 10. Hartmann, J., "Bemerkungen uber den Bau und die Justirung von Specktographen," Zt. Instrumentenkd., 20, No. 47 (1900). 11. Feinleib, J. M. and Schmutz, L. E., "High Speed/Low Light Wavefront Sensor System," United States Statutory Invention Registration, No. H615 (1989). 12. Ealey, M. A., "Low Voltage SELECT Deformable Mirrors," in Smart Structures and Materials: Active and Adaptive Optical Components and Systems II, Mark A. Ealey, Editor, Proc. S. P. I. E., 1920, pp. 91-102 (1993). 13. Lillard, R. L. and Heynau, H. A., "Cooled deformable mirror for ALPHA-LAMP integration experiment," in Smart Structures and Materials: Active and Adaptive Optical Components and Systems II, Mark A. Ealey, Editor, Proc.S. P. I. E., 1920, pp. 103- 114 (1993). 14. Dryden, C, et. al, "The HCP-100: A Novel Architecture for Real-Time Processing and Control," AOA Internal Publication, (1990). -221- Session IV Laser Material Processing rrocccaings 01 inc om international symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Surface Processing By High Power Excimer Laser Marc STEHLE Vice President Laser Department Manager SOPRA 26, rue Pierre Joigneaux, F92270 BOIS-COLOMBES, FRANCE Tel. (33.1) 47 81 09 49 Fax. (33.1) 42 42 29 34 Surface processing with lasers is a promissing field of research and applications because lasers bring substantial advantages : laser beams work at distance, laser treatments are clean in respect of environment consideration and they offer innovative capabilities for surface treatment which cannot be reached by other way. Excimer lasers are pulsed, gaseous lasers which emit in UV spectral range - the most common are XeCl (308 nm), KrF (248 nm), ArF (193 nm). From 1980 up to 1994, many of them have been used for research, medical and industrial applications such as spectroscopy, PRK (photo-refractive keratotomy) and micro-machining. According to this first generation of excimer lasers which will last as their average power is in the range of 1 watt up to 100 watts, some emerging applications have been investigated but they require more average power in order to be attractive for large scale industrial applications. Thus in the last six years, from 1987 up to 1993, efforts have been done in order to jump from 100 W average power up to 1 kW for XeCl laser at X = 308 nm. It was the aim of AMMTRA project in Japan as ETJ205 and EU213 Eureka projects in Europe. In this framework, SOPRA developped VEL (Very large Excimer Laser). In 1992, 1 kW (10 J x 100 Hz) millestone has been reached for the first time, this technology is based on X-Ray preionization and large laser medium (5 liters). Surface treatments based on this laser source are the main purpose of VEL Lasers. Some of them are given for instance : a) Turbine blades made with metalic substrate and ceramic coatings on the top, are glazed in order to increase corrosion resistance of ceramic and metal sandwich. b) Selective ablation of organic coatings deposited on fragile composite material is investigated in Aerospace industry. c) Chock hardening of bulk meiallic materials or alloys are investigated for automotive industry in order to increase wear resistance. d) Ablation of thin surface oxydes of polluted steels are under investigation in nuclear industry for decontamination. Keywords: Surface Processing, Excimer Laser, Glazing, Selective Ablation, Shock Hardening, Decontamination — 223 — Surface Processing By High Power Excimer Laser Marc STEHLE Vice President Laser Department Manager SOPRA 26, rue Pierre Joigneaux, F92270 BOIS-COLOMBES, FRANCE Tel. (33.1) 47 81 09 49 Fax. (33.1) 42 42 29 34 I) Surface processing with lasers is a promissing field of research and applications because lasers bring substantial advantages : laser beams work at distance, laser treatments are clean in respect of environment consideration and they offer innovative capabilities for surface treatment which cannot be reached by other way. Excimer lasers are pulsed, gaseous lasers which emit in UV spectral range - the most common are XeCl (308 nm), KrF (248 nm), ArF (193 nm). From 1980 up to 1994, many of them have been used for research, medical and industrial applications such as spectroscopy, PRK (photo-refractive keratotomy) and micro-machining. According to this first generation of excimer lasers which will last as their average power is in the range of 1 watt up to 100 watts, some emerging applications have been investigated but they require more average power in order to be attractive for large scale industrial applications. Thus in the last six years, from 1987 up to 1993, efforts have been done in order to jump from 100 W average power up to 1 kW for XeCl laser at X - 308 nm. It was the aim of AMMTRA project in Japan as EU205 and EU213 Eureka projects in Europe. In this framework, SOPRA developped VEL (Very large Excimer Laser). In 1992, 1 kW (10 J x 100 Hz) millestone has been reached for the first time, this technology is based on X-Ray preionization and large laser medium (5 liters). The first prototype has been described in former papers (Ref. 1) with technical description of its behaviour and the optimization based on extensive parametric approach. The principle of this laser is using an X-Ray preionization and L.C. inversion circuit with six parallel thyratrons. The X-Ray gun is a secondary emission electron gun, developed in collaboration with ONERA research Laboratory. This X-Ray gun has been described in earlier paper (Ref. 2). In order to refresh the laser discharge gas between two shots, an efficient loop of gas has been designed with a flow rate in the range of 2 m3/s. The vessel can work up to 7 Bars pressure. -224 — First results with longer bursts have been reported in 1993 (Ref. 3). Fig. 1 - SET UP The specifications of this laser are as following : the emitting wavelength is 308 inn with XeCl. The laser active medium is around 5 liters : (length = 1 meter, height = 8 cm, width = 6 cm). The maximum pressure of operation is up to 6 Bars, the pulse duration is typically 160 ns (FWITM and is shown on Fig. 2). • • ' • - • - .. . J i ' ...... j. • M/\ ; ? -I-'r-H- .,,-,-;, -!-:-+->.- -M4-I-- .^-H.r-i-!-H-r4:HHH-rV-;-i-H--!-!-!-i--f-i-t-i^- Ch2 +Wi Fig. 2 The maximum energy per pulse obtained is 15J in single shot operation at full high voltage, the wind tunnel flow rate is 2m3/sec, thus the speed of gas between both transverse discharge electrodes is 20 m/s. -225- ln the other part of the wind tunnel, speed of gas is only 2 m/s. The total consumption of twin blowers to move circulating gas is less than 3 kW which enables this laser to keep high plug efficiency over 2 percent including blowers. The SOPRA's approach of kW excimcr laser at k\V average power level is based on following choice : large energy per shot (over 10 Joules) at medium repetition rates (around 100 Hz) provided typical average power curves versus repetition rates. Sec Fig. 3. , Energy per pulse Average power (J ) (KW] 15 _^ 1,5 10 1 KW y _,---" y ,--'* y „••' y ,'' 0,5 5 y _,-' ys 0 y" i t i i i >— 20 40 60 80 100 Frequency (HZ) Fig. 3 As a partial conclusion, this choice of design of VEL 1 K laser gives to it the following position in comparison with conventional exciincr lasers. See Fig. 4. E/shor 0) 10 CLASSIC EXC LASERS 0,1 f rep. rale 100 Hz 1 kHz Fig. 4 — 226- II) Surface treatments based on this laser source arc the main purpose of VEL Lasers. Some of them are given for instance : a) Turbine blades made with metalic substrate and ceramic coatings on the top, arc glazed in order to increase corrosion resistance of ceramic and metal sandwich. b) Selective ablation of organic coatings deposited on fragile composite material is investigated in Aerospace industry. c) Shock hardening of bulk metallic materials or alloys arc investigated for automotive industry in order to increase wear resistance. d) Ablation of thin surface oxydes of polluted steels arc under investigation in nuclear industry for decontamination. In order to fulfill requirements of surface treatments based on use of VEL 1 K laser, efforts have been put on beam handling. As the beam profile dimensions arc 8 cm x 6 cm thus the power density on the optics, without focusing is only 200 inJ/cm2. This value is similar to conventional excimer lasers values. Only the diameter of optics had to be increased, but the lifetime is not reduced by any reason as thermal effects due to kilohertz repetition rate or by number of shots. In addition, different types of beam homogeneizers have been compared as beam shaping with cut cylindrical lenses, or cut prisms on one hand. On the other hand, fish eyes micro-lenses arrays have been successfully tested. Fig. 5 is giving beam profile after homogeneizer. Same quality, better than + or - 5 % is achieved in X or Y axis. '•-v.* V"*" •.iv.r, SUCI .1.7 !i J.-JU .I.H3 .Hit) 73 '.in CO I. JLHJ. A G.I. :2JL .O.I. 3 Hi Kl'iX 2'.V3 MIM Fig. 5 -227— Some applications require, optical fibers delivery. Multi-fibers coupling is under investigation. With energy at the entrance limited at 2J/shot, emitted energy at exit of fiber bundle over 1.1 Joule has been achieved, for length of optical fibers longer than 2 meters. At maximum, 10J/cm2 power density at output of optical fibers has been already demonstrated with silica optical fibers containing OH radical for better transmission in 308 nm UV. It is still too early to specify results of transmission for 20 m distances which should be necessary in nuclear industry if heavy contamination. Ill) CONCLUSION If wc assume that for stainless steel, oxyde thickness to remove for decontamination in nuclear industry is in the range of I to 5 urn, and if we assume that 1 to 5J/cm2 energy density is necessary to ablate such oxyde from the substrate, then a 1 kW excimer laser with multiple optical fibers delivery system will be capable to treat from 1 to 10 m2/hour. The process has to be optimized with better understanding of UV light ablation on surfaces, modelization and may be additional water interface to remove ablated debrics. VEL lasers are convenient demonstrator tools made for applied research. They can be used to demonstrate the feasibility of industrial applications of excimer laser surface treatments. IV) REFERENCES Ref. 1 : Marc STEHLE, Bruno GODARD, Pierre MURER, Francois-Xavier BROWN, Jean BONNET, Daniel PIGACHE "Innovative Way or Reaching 1 kW XeCl Excinier Goal : lOJ/shot Large Aperture Excimer Laser" LAMP'92 - 7/12 June 1992 Niigala, JAPAN Ref. 2 : Jean BONNET, Daniel PIGACHE, Patrick LABORDE, Marc STEHLE, "Secondary Emission Electron Gun for X-Ray Preionized Discharge Pumped XeCl Lasers" CLEO'89/Ballhuore, USA - 24/28 April 1994 Ref. 3 : Bruno GODARD, Pierre MURER, Marc STEILLE, Jean BONNET, Daniel PIGACHE - "Development of a 1 kW XeCl Laser" EUREKA EXCIMER LASER WORKSHOPlERLANGEN - April 1993 — 228 — rrocccaings 01 me om imcrnauonai symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- LASER ABLATION FOR THE METAL-OXIDE THIN FILM DEPOSITION :MECHANISMS AND APPLICATIONS Tomoji KAWAI ISIR-Sanken, Osaka University, Mihogaoka, Ibaraki, Osaka 567 Japan. ABSTRACT Laser ablation of Bi-Sr-Ca-Cu-0 superconductor has been studied by a dynamic luminescence spectroscopy. A mechanism is proposed in which only ions are produced at the initial stage followed by formation of neutral species in collision processes. Ferroelectric PbTiOg and LiNb03 films have been fabricated by the laser ablation in the presence of 02/03 gas and have shown large D-E hysteresis and surface acoustic wave properties, respectively. The control of atomic layer and sub- unit cell layer during the growth of metal oxide thin films has become possible by laser ablation under molecular beam epitaxy(Laser MBE) condition combined with in-situ surface science diagnostics, such as RHEED, AES, ISS and STM. By using this technique, two dimensional layered structure of high Tc cuprates have been artificially constructed- layer-by-layer. Evidences have been obtained for the atomic layer and sub-unit cell layer growth in the system of Bi2Sr2Can_-LCun0on+4 (n=l-10) and (Ca,Sr)Cu02 thin films. The atomically layered (Ca,Sr)Cu02 films thus prepared have shown resistivity drops and diamagnetic signals above 100K. Keywords: Laser Ablation, Superconductor, Luminescence Spectroscopy, Ferroelectric Films, Molecular Beam Eitaxy I. Laser ablation mechanism^' ' In order to elucidate the ablation mechanism of metal oxide materials, time-resolved luminescence spectra have been measured with nanosecond time resolution for a UV laser ablation of Bi-Sr- Ca-Cu-0 superconducting materials.(Fig.1) It has been revealed that in the early stages within 200 nsec after the irradiation, a wide continuum spectrum with self-absorption lines due to the electronic transitions from the ground states of Ca+, Sr+ and Sr is observed. The continuum emission due to plasma state decays with the lifetime of 70 nsec and Is replaced by emission lines of Ca+, Sr+, Ca and Sr which has decay times of around one microsecond. Time resolved spectroscopy of the plasma state produced at the very beginning within 50 nsec after the laser ablation has been also studied. The difference in the spectra among the crystalline and amorphous samples Indicated that the ablation products depend on the chemical structures of the irradiated samples. The experiments in the absence of the ambient gas has revealed that ionic species in the electronic ground states are mainly generated at the initial stage of the ablation, while atomic species are produced by collisions between the surrounding-gas molecules and plasma plume. — 229 — Ca+Sr+Ca Sr Cu : : 1 :: : 1 1 • • i: • t-5<.-l«Znsee i!i : : ~ ::i • : : | :: : 1 TH i : 1 : : : 1 •H : : :: i C"10«-192nsec 3 Hi • : 5 Si : tt^Mttv^^w^ ::l:;; 1 1 X : : I j i t-2(K-294nsic EH M W z ::i • i f: ;-i^r ^W^yj^^UMtlxJI^ Fig.l Time resolved emission w i i :: i t-454-542nsec spectra after the ablation EH of Bi Sr CaCu 0 (a)54-142ns, Z 2 2 2 8 H (b)104-192ns, (c)204-292ns, (d)454-542ns IMil .>>>A^j.,..ii~W J^»iA...tjii 400 450 500 550 WAVELENGTH/nm These results are summarized in a scheme in Fig.2. This scheme shows that ground state of ionic species is produced in a plasma at the initial stage within 50 nsec. After the colli sions among the species in the plasma as well as with the ambi ent gaseous molecule, ground states of neutral species are prod uced in the plasma. The isolated and excited neutral atoms and ions with longer life time are produced either by an absorption of light or by the recombinaion of multi-valent ions with ele ctrons. The atomic and ionic species containing these excited sp ecies reach the surface to form deposited layers in the ablation process. Initial stage hr(lj/cm2) Ablation Plasma Collision Piasma (tss70ns) - > A+(ion) C !/• A(neutral) , A+(ion) G (D ground state ground state Bi;5r£aCu208 A , A++hr - -A* , A*+ A**, A++++e • -A*, A** V A(*), A+(*) (long life time) Fig.2 Schematic mechanism of the laser ablation of Bi2Sr2CaCu208 — 230- II. Ferroelectric thin films 3,4) II-l. PbTi03 thin films: Second pulse irradiation effect By using ArF laser ablation method, PbTi03 thin films have been formed on SrTiO3(100) substrates at temperatures of 350C° to 400C". The target used for this experiment is a stoichiometric PbTi03 disk. So far, PbTi03 films have not been formed at temperatures lower than 500C°using other thin film techniques. The important points for the low temperature formation are the laser excitation of the substrate during the film growth and the lattice matching between the film and the substrate. The film thus formed on the SrTiOg substrate shows preferential orientation of the c-axis perpendicular to the substrate surface. The formation of PbTi03 have also been examined on Pt base electrodes or oxide superconductor (La.Sr^CuO^ electrodes. A second laser irradiation at the substrate surface during the growth is very effective for the crystallization of the films at low substrate temperature below 400C. The suitable energy density of the irradiation laser is in the range of 30 to lOOmJ/cm . X- ray diffraction patterns of PbTi03 show c-axis orientation with FWHM of rocking curve of 0.5 - 1.0 degree. These films exhibit large and clear' ferroelectric hysteresis loops.(Fig.3) The remanent polarization of the PbTi03 films is BOpC/cm which is almost the ideal value predicted by the calculation for this compound. 5} II-2. LiNb03 thin films The ArF laser ablation has been also applied to the formation of high quality epitaxial LiNb03 thin films on sapphire subst rates (001), (110) and (012). The epitaxial relationships as revea led by four axis x-ray diffraction are LiNbO 3(001)//sapphire(001), LiNb03//sapphire(110) and LiNbO 3(10O)//sapphire(012).(Fig.4) The LiNbO3(110) orientation is also obtained on (110) sapphire under the fast-deposition condi tion. The advantage of laser ablation method for this E(kv/cm) S H .a c 50 60 3S.cn m.n 28/deg Fig.3 D-E hysteresis and X-ray Fig.4 X-ray diffraction pattern diffraction pattern of a PbT103 and pole figure of LiNb03 on f^lm. sapphlre(012) . 231- compound is high purity of the LiNbC>3 film, because laser light irradiates only the targe t surface to produce Li and Nb atoms and ions without producing other contaminants. A surface acoustic wave (SAW) filters are fabricated on the films, and SAW properties are evaluated, The LiNb03 thin films on the sapphire substrates exhibited a h igh frequency SAW filter characteristics for the first time. The LiNb03 films prepared by laser ablation have high SAW velocities of propagation (5300 - 5600m/s) and good temperature coefficients (-34 to -80 ppm/Cf) . Ill.Layer-by-layer Growth of Superconducting thin films III-l. Formation of Bi2Sr2Can_1Cun02n+4 with n=4-10 and atomic layer control6,',8' Bi Sr Ca Cu on The growth of SrCu02 thin layer and 2 2 n-l n°2n+4 SrTiC>3 single crystal substrates have been studied by laser MBE equipped with RHEED (reflection high energy electron diffraction), AES and STM. (Fig.5) When Sr and Cu are supplied simultaneously to the substrates to form SrCu02, either a SrO or Cu02 layer is the surface layer depending upon the composition of the substrate surface to form a more stable interface, However the surface topmost layer is controlled by the sequence of the deposition when atomic layer-by-layer processes are applied By the successive supply of the subunit-cell layers of Bi2sr-2<-'an-lCun^2n+4 • "^e structures of not only n=l to 4 but also n=5-10 are formed.(Fig.6) The RHEED pattern shows streaks all through the growth, indicating two-dimensional layer growth in all the steps controlled within the monolayer accuracy. Emission spectroscope Fig-.5 A schmatic diagram of Laser MBE apparatus. — 232— We have observed scanning tunneling microscopy(STM) images and scanning tunneling sp ectra (STS) with atomic-scale resolution for a _SrTiO3(100) s urface annealed in ultrahigh vacuum at 1200C. ' AJ~5xvJ5-R26.6°surfa ce superstructure indicating oxygen vacancy ordering has been obse rved. The STS data provide evidence for a localized surface state arising from oxygen vacancies at 1.35eV below the Fermi level. STM images of the*T5x \T5~ structure correspond to the surface o rbital of the ordered Ti-oxygen vacancy complex. Atoms of Sr are dispersed by laser MBE on this particular surface to inves tigate the initial stages of the growth. The STM images show th at atoms are deposited very finely on the SrTi03 surface with the atomic height of Sr. The Sr atoms dispersed randomly at room t emperature, and move to form an epitaxial layer of Sr above 30 0C°. By the Laser ablation under the MBE condition, atomic height s urface morphology is obtained. BJ2Sr2Ca6Cu7QiB Fig.6 Crystal structure and X-ray diffraction pattern of Bi2Sr2Ca6Cu7018 9) III-2. Ca1_xSrxCu02 artificial lattices Thin films of Ca-i _xSrxCuC>2 have been prepared on SrTiOg (100) substratestrates by co-deposition of all elements(Ca,Selements(Ca, Srr anandd CuCu)) and successive stacking of atomic layers. When all the elements are deposited simultaneously, Ca1_xSrxCu02 films are formed whose resistivity-temperature is seraiconductive and magnetization temperature curve does not show any anomalies. By stacking atomic layers of Ca(Sr)0x and Cu0x, the infinite layer structure with superlattice modulation is possible. For the SrCuC>2 thin films, a diamagnetic transition and decrease of resistance have been observed at 65K and 90K. For the (Ca1_xSrx)Cu02 samples, the lattice constant c and electrical conductivity increase systematically with increasing Sr/Ca ratio. Clear diamagnetic signals at 90K and 120K have been observed in SrCuC^ and Ca0 4SrQ_gCuC>2 films, respectively. (Fig.7) In the CaQ 3Sr0 7CUO2 films, the onset of a small diamagnetic signal and a resistivity drop are also observed around 150K - 170K. In these thin films, systematic current density dependence of the resistivity was observed beloTv 150 - 170K. At the same temperature range, the films showed hysteresis of magnetization between zero-field cooling and field cooling process. -233— Temperature, T/K 60 120 180 240 3° 1 -T ' * ' ' ' ' ' • 4 , . o 8»*o»^QSj#s 8*2' 0 0 0 o 0 0 • Besides oxygen incorporation to thf Cal-xSrxCu02 films, selective doping' of Na1+ into Sr(Ca)'' site and periodic insertion of Bi20n layers to inject carriers to the Cu02 planes are possible. In this sense, laser molecular beam epitaxy is one of the most sophisticated methods for the formation of variety of superconducting artificial lattices. IV. CONCLUSION: Advantages of laser ablation methods The laser ablation method has many advantages for the thin film formation, especially for metal oxide thin films, over the other conventional methods, as follows; 1) Wide range of ambient gas pressure is applicable during film growth because laser light comes from outside of a vacuum chamber. 2) Photochemically activated species and/or substrate excitation are utilized for the low temperature film formation, as shown in the case of PbTi03 thin films. 3) This method has high controllability by choosing pulse number and pulse energy. 4) This method has versatility. Insulators, metals, ceramics and semiconductors are easily ablated. 5) The film is not contaminated since laser light ablate target alone. The example is shown for the LiNbOo films which shows good surface acoustic wave properties for the first time. 6) Very small composition change occurs between target and film as has been shown for superconducting materials. 7) High speed film formation is possible just by increasing laser intensity. 8) Even atomic layer control is possible by the laser ablation under MBE condition (Laser MBE) method. Thus, the laser ablation method will be used more widely for the formation of variety of functional thin films. References 1. H.Fukuraura et al, Jpn.J.Appl.Phys. , 28, L412(1989) -234- 2. H.Fukumura et al, Appl.Phys.Lett., 58, 2546 (1991). 3. H.Tabata et al, Jpn.J.Appl.Phys., 31, 2968 (1992). 4. H.Tabata et al, Appl.Phys-Lett., 59, 2354 (1991). 5. Y.Shibata et al, Appl.Phys.Lett. , 61, 1000 (1992). 6. T.Kawai et al, Nature, 349 ,200 (1991). 7. M.Kanai et al, Jpn.J.Appl.Phys., 31, L331 (1992). 8. H.Tanaka et al, Jpn.J.Appl.Phys., 32 , 1405 (1993). 9. X.Li et al, Jpn.J.Appl.Phys., 31, L217. L934 (1992). — 235- c. ~ iwiiiuiiuudi OjriII|)UMUni on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- MULTI-WAVELENGTH COPPER VAPOUR LASERS FOR NOVEL MATERIALS PROCESSING APPLICATION Martyn KNOWLES, Jason EVANS*, Richard FOSTER-TURNER and Andrew KEARSLEY Oxford Lasers Ltd, Abingdon Science Park, Abingdon, 0X14 3YR, United Kingdom. 'Department of Physics, Clarendon Laboratory, University of Oxford, Oxford, 0X1 3PU, United Kingdom. The copper vapour laser (CVL) is a high average power, short pulse laser with a multi-kilohertz pulse repetition rate. The CVL laser lines (511 nm & 578 nm) combined with the good beam quality and high peak power available from these lasers allow it to operate in a unique parameter space. Consequently, it has demonstrated many unique and advantageous machining characteristics. We have also demonstrated efficient conversion of CVL radiation to other wavelengths using non-linear frequency conversion, dye lasers and Ti:AL203. Output powers of up to 4 W at 255 nm have been achieved by frequency doubling. The frequency doubled CVL is inherently narrow linewidth and frequency locked making it a suitable source for UV photolithography. Slope efficiencies in excess of 25 % have been achieved with CVL pumped Ti:Al203 and dye lasers. These laser extend the wavelengths options into the red and infrared regions of the spectrum. The near diffraction limited beams from these tunable lasers can be efficiently frequency doubled into the blue and near UV. The wide range of wavelength options from the CVL enable a wide variety of materials processing and material interactions to be explored. A European consortium for Copper Laser Applications in Manufacture and Production (CLAMP) has been set up under the EUREKA scheme to coordinate the commercial and technical expertise currently available in Europe. Keywords: Copper Vapour Laser, Materials Processing, Dye Lasers, Titanium Sapphire, Frequency Doubling. 1. INTRODUCTION The copper vapour laser (CVL) is a pulsed visible laser (511 nm & 578 nm laser lines) which emits intense (50 - 500 kW), short (20 - 60 ns) pulses at high pulse repetition frequencies (2 - 32 kHz). Average powers from commercial units range from 10 - 120 W whilst laboratory devices have reached 750 W. Novel unstable resonator designs and amplifier configurations have yielded near diffraction limited beams which may be efficiently amplified to powers in excess of 2 kW [1] with typical wall-plug efficiencies of 1%. The requirement for multi-kilowatt dye laser systems for Atomic Vapour Laser Isotope Separation (AVLIS) of the fissionable isotope 235U in natural uranium necessitated multi-kilowatt visible pump lasers. The CVL was identified — 236 — as the only laser source that could be scaled to the required power levels in the timescale set to assess the feasibility of AVLIS. This research and development programme has led to the multi-hundred watt modules that are now available. These modules can be stacked in series to form very high power Master Oscillator Power Amplifier (MOPA) chains. At Lawrence Livermore National Laboratories, single MOPA systems routinely deliver 1.5 kW of near diffraction limited CVL radiation and the total facility is capable of over 10 kW. Oxford Lasers has installed a number of systems of over 250 W.. The unique set of laser parameters that characterize the CVL have enabled new and interesting results in materials processing. The extension of the wavelength range by frequency doubling and tunable lasers offers exciting possibilities in the study of materials interaction. 2. CHARACTERISTICS OF THE CVL RELEVANT TO MATERIALS PROCESSING As outlined above, the CVL has a set of characteristics which make it unique. The relevance of these features to materials processing will now be discussed. 2.1 Wavelength In radiation/material interaction the wavelength of the radiation determines the type and degree of interaction. The reflectivity of many metals at infra-red wavelengths is typically greater than 90% and for some metals such as copper it can be greater than 98%. However at the CVL wavelengths the reflectivity is much lower, typically 50 - 60%. A long pulse infra-red wavelength laser relies on a thermal interaction to melt the metal. Once the surface of the material is molten the fraction of the laser power coupled into the material increases to about 70%. The molten metal is then removed either by gas-assisted oxidation or melt shearing. Although this results in efficient cutting and drilling, the precision of the process is limited and the area around the initial irradiation point often suffers severe thermal damage owing to the low absorption of the solid. The thermal reaction makes it difficult to achieve very small drill holes or narrow kerf widths without creating a substantial heat affected zone (HAZ). The HAZ can produce weakening and distortion of the sample, sometimes rendering it useless. The absorption of UV radiation in most metals is very strong. The short high peak power pulse can under optimum conditions of high fluence, extended pulse length and helium atmospheres, ablate up to a few microns per pulse. However the low pulse repetition frequency limits the processing speed in applications such as single hole drilling. The high fluence and efficient coupling in most metals enables a single CVL laser pulse to remove material to a depth of the order 10 u.m. This combined with the high pulse repetition frequency results in a high material removal rate compared to an excimer in many applications. However the visible wavelength, aided by the short, high peak power, pulse, results in a "cold cut" or ablative process and a precision that cannot be matched by the IR lasers. -237- Certain non-metallic materials also interact more favourably with visible rather than infrared radiation. The shorter wavelength and high peak power of the CVL can be used to induce photochemical reactions resulting in a colour change in some polyester type materials used to package electronic components. Thus the CVL can be used for high speed, high definition alpha-numeric marking of components. Other non-metals respond better to IR radiation or UV radiation. For example many polymers are very effectively etched by UV radiation and an important application for CO2 lasers is cutting perspex. Extension of the CVL wavelengths using frequency doubling and tunable lasers would enable those applications which required UV wavelengths at high pulse repetition rates to be addressed. It has been found that 50 urn holes in thin polyester sheets can be drilled at a rate 2 orders of magnitude faster than the competing excimer based system [2]. 2.2 Pulsewidth The short pulsewidth of the CVL is advantageous in reducing HAZ. If the material is to be removed via evaporation then it is necessary to create extremely high temperature gradients in order to keep the total heat absorption of the component very low and at the same time to transform a large portion of the irradiated surface into the vapour phase. This requires an intensity 108 to 109 W/cm2 and therefore a pulsed laser is necessary to achieve these powers [3]. The short pulse serves to heat the irradiated volume to the vapourization temperature in a timescale shorter than that for the heat to diffuse into the surrounding region. The CVL shows a HAZ layer of less than 500 nm thickness in many metals [4]. 2.3 Repetition Rate The repetition rate and average power of the laser determines the rate of material removal and therefore the process speed. When a low pulse energy is required to avoid HAZ then it is clear that a high repetition rate laser is necessary to achieve a good process speed. The CVL has the highest average power . pulse repetition frequency product for any visible laser. 2.4 Beam Quality The minimum size that a laser beam may be focused to is determined by its wavelength and the beam quality. The minimum beam size for a given focal length lens and beam diameter is proportional to the wavelength. Therefore the CVL has significant advantages over CO2, and Nd:YAG lasers in terms of the minimum spot size to which it may be focused. Although close to diffraction limited beam quality may be obtained at very high power from a CO2 laser, its wavelength is 20 times greater than that of the CVL. At high powers thermal distortion of the laser crystal degrades the beam quality of the Nd:YAG laser to several times the diffraction limit. CVLs are commercially available [5] in modules with powers exceeding 100W and can be configured to deliver a large proportion of the power within the diffraction limit. Recent laboratory developments on CVL -238- resonators have increased the power within the diffraction limit to almost 100%. These resonators should be available commercially within the next few years. 3. WAVELENGTH EXTENSION OF THE CVL Although the 511 nm and 578 nm lines of the CVL have been demonstrated to be well suited for many applications the ability to extend the wavelength range offers some interesting options in materials interaction. The CVL is an excellent pump source for many tunable lasers and it can also be efficiently converted to UV wavelengths using non-linear frequency conversion. 3.1 CVL Pumped Dye Lasers The CVL is the pump laser of choice for dye lasers and has found applications in AVLIS, Resonance Ionization Mass Spectroscopy (RIMS), OH Radical Detection, Short Pulse Amplification and Photodynamic Therapy. The pulse width from the dye laser is approximately the same length as the CVL pulse and the beam quality near to the diffraction limit. The best efficiency is obtained from dyes lasing in the 550 - 700 nm range. The conversion efficiency for a low power (5 W), narrow bandwidth (<5 GHz) dye laser suitable for RIMS, OH radical detection and other spectroscopic applications is typically 25% [6]. Broad bandwidth dye lasers developed for medical applications but also suitable for materials processing and display applications typically show conversion efficiencies of 35%. 3.2 CVL Pumped Ti:Al203 Lasers The Laser Research Group at the Clarendon Laboratory Oxford have a world-lead in CVL pumped Ti:Al203 lasers. Output powers of greater than 5 W [7] at slope efficiencies of greater than 30% have been achieved and with injection seeding near transform limited bandwidths (<100 MHz) are possible with no reduction in output power. The laser was tunable from 690 to 950 nm. With different mirror sets it should be possible to extend this range as low as 660 nm and as high as 1050 nm. The low timing jitter of the Ti:Al203 laser pulse (+ 1ns) enables non linear frequency mixing of the CVL and Ti:Al203 laser radiation to generate tunable UV at shorter wavelengths than would be possible by frequency doubled Ti:Al203. Figure 1 shows the tuning range of the fundamental Ti:Al203 laser, the frequency doubled range and the sum-mixing range. 3.3 Non-linear Frequency Conversion of CVL Radiation Non-linear frequency conversion of CVL radiation generates narrow bandwidth, frequency locked UV pulses at multi-kilohertz repetition rates. Frequency doubling the 511 and 578 nm lines produces 255 and 289 nm whereas sum frequency mixing generates 271 nm radiation. Efficiencies of up to 25 % have been reported and output powers of up to 4 W have been generated — 239- [2]. The narrow linewidth of 0.002 nm and intrinsically stable wavelength makes the UV CVL ideally suited to high resolution photolithography. 4. MATERIALS PROCESSING WITH THE CVL A model developed by Lawrence Livermore National Laboratory [1] which compares the cutting speed and minimum kerf width in 6 mm stainless steel for different laser types has shown the highest productivity is obtained wiih CVL lasers and CVL pumped dye lasers. The model compares excimer lasers, Nd:YAG lasers, CO2 lasers and the projected values for Nd:YAG slab lasers. Experimental studies at Oxford Lasers with laser powers of up to 50 W average power, 200 kW peak power, has produced promising results. A wide range of metals including aluminium, brass, copper, chrome, gold, mild steel, stainless steel, nickel, stainless steel and titanium have been processed. Non- metals which have been successfully processed include carbon fibre, cement, ceramic, diamond, glass, polyimide, quartz and silicon. Figure 2 shows the drill time as a function thickness in copper using the 200 kW peak power system at a pulse repetition frequency of 7 kHz. The drill times are typical for hole drilling in metals although titanium, mild steel and stainless steel have slightly shorter drill times. Holes of 1 - 100 u.m diameter are usually drilled directly whereas larger holes are trepanned. The holes are characterized by a small heat affected zone (< 300 nm in copper and < 10 |im in titanium). Figure 3 shows a section of an array of 400, 000 holes of 10 urn diameter in 125 u,m thick aluminium. The drilling nature of the CVL is an ablative one. Material is either directly vapourized or explosively ejected as small molten globules. Assist gases are not necessary for melt shearing or oxidization as is the case for infrared lasers. The ablative nature of the CVL process allows blind holes with high aspect ratios to be drilled. The ablative nature also means that the cutting and drilling mechanism are very similar; cutting is essentially achieved by overlapping holes. Consequently the kerf widths and heat affected zones are very similar to those obtained with drilling. In addition to this it is possible to fabricate high aspect ratio blind slots which may be used for crack simulation in metal components. An example of a blind slot in stainless steel with a 65 u.m entrance diameter and depth of nearly 3 mm is shown in figure 4. Although ceramics are cut with speed and precision by CO2 and Nd:YAG lasers, the melt process tends to leave a glassified laser around the edge of the processed area. These substrates are often used in electronic circuit fabrication and the glassified layer can lead to adhesion problems for the conducting film. The ablative processing of the CVL results in a complete absence of glassification and with processing speeds comparable with similar power Nd:YAG lasers. Further developments of this process could lead to a major advantage to circuit fabricators. Other materials of interest to the electronics industry include silicon. Figure 5 shows the cutting speed of 300 urn thick polycrystalline silicon as a function of laser power. The highest speed achieved was 22 mm/s at a power of 50 W. Recent improvements to the beam quality of the laser may well improve upon this value. Figures 6 and 7 show a 350 urn side square and 200 urn - 240- diameter hole in 1 mm thick transparent natural diamond. The precision achievable with the CVL could be used in the fabrication of diamond heat sinks and dies for drawing wires. Initial trials with a frequency doubled CVL at 255 nm have shown excellent drilling properties in quartz, glass and polymers. As yet no materials processing applications studies have been undertaken in Oxford with the CVL pumped dye and Ti:Al203 lasers. However the wide wavelength range accessible would enable a thorough investigation of the processing of a given material as a function of wavelength with all other laser parameters, e.g. pulse width, repetition rate and power held nominally the same. This system could prove to be a potent tool for testing the processing capabilities of laser systems of the future. For example, the CVL pumped Ti:Al203 laser could be used to simulate a diffraction limited, high peak power, high pulse repetition rate laser diode array. 5. CONCLUSIONS The CVL is the new laser in materials processing. It has shown great potential as a tool for high precision manufacturing and is able to address applications that are not feasible with any other technique. Commercial Copper Laser Machining Systems are now commercially available. The ability to extend the wavelength range of the Copper Laser from 255 nm to beyond 1000 nm to be carried across a wide wavelength range at power levels not possible with existing laser sources. The output powers of Copper Lasers are more than 20 times higher than commercial frequency doubled Nd:YAG lasers. Recent breakthroughs and on-going research in Copper Lasers will continue to scale the efficiency, output power and beam quality of these lasers. REFERENCES [1] Hargrove R.S., Dragon E.P., Hackel R.P., Kantz D.D., Warner B.E., "SPIE Proceedings of Laser Isotope Separation", Vol 1859. 263, (1993) [2] Glover A.C., Coutts D.W., Ramsay D.J., Karabaic D.F., Brown D.J.W. Piper J.A., Proceedings of ICALEO Conference (1993). [3] Hetziger G., The influence of laser-induced plasma on materials processing, The Industrial Laser Handbook 1986, Ed. by D. Bekfirte, M. Levitt, PennWell Books, Tulsa, 108, (1986). [4] Bergmann H.W. & Hartmann M., "Laser Processing Proceedings", TMS Annual Meeting, Denver (1993). [5] Oxford Lasers Ltd ACL Data Sheet, Oxford Lasers Ltd, Abingdon Science Park, Abingdon, OX14 3YR, U.K. or Oxford Lasers Japan, 6 - 30 Kiba, 5- Chome, Koto-ku, Tokyo 135. [6] Evans I.J., D.Phil. Thesis, University of Oxford, (1993). — 241- [7] Knowles M.R.H. & Webb C.E., Optics Letters, 18_, 607, (1993). [8] Knowles M.R.H. & Webb C.E., Optics Communications, S9_, 493, (1992). Sum BBBJ^SH ThSapphire SH Dye SH CVL I 1 1 1 200 400 600 800 1000 Figure 1. Fundamental and Second Harmonic (SH) of the CVL and CVL pumped dye laser. Also tuning range of the CVL pumped Ti:Al203 laser, its' second harmonic and sum-mixing range. 60 -• 50 to g 40 - o> | 30 • 1 20 - 10 •- 0 -U 0 0.2 0.4 0.6 0.8 1 Thickness (mm Figure 2. Drill time as a function of thickness in copper with the 200 kW peak power. -242— 0> i* W'-A .'.V^to- it- '."••'•J IN' \>?Vf r ' ''-VSKI," "»^W/J * Ml KK L*-, Figure3. Matrix of 10 urn holes on a 40 urn pitch. 400, 000 holes were required in each unit with no blocked or oversized holes. All specifications were met with no holes greater than 12.5 urn (suction tested). (Courtesy of Creare Inc., Hanover, New Hampshire, U.S.A.) Figure 4. Blind slot in stainless steel. The slot entrance is 65 yum and the depth is 3100 |im. Note that there is no heat affected zone. — 243 — 10 20 30 40 50 Power (W) Figure 5. Cut speed of 300 urn thick polycrystalline silicon as a function of average laser power. Figure 6. Square hole of 350 urn cut in 1 mm thick diamond. -244- - + *M * m- j« : > v t ®». •Jresf'"' J JWSBKC'-''''- - i i a'•^SSESl&'iww*^- "£»&&• ' • 1 • 1 |F?0'SH"V„,1 • T. ?% */," ' . » r •*! * .; .A 013.4 -*> o i' n •lu 3 Figure 7. Circular hole of 200 \xm diameter cut in 1 mm diamond. -245 — Proceedings oflhc 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Surfacc Modification of Polymer Materials by Excimer-Laser Irradiation S. Kawanishi, M. Nishii, Y. Shimizu, and S. Sugimoto Osaka Laboratory for Radiation Chemistry Japan Atomic Energy Research Institute 25-1 Mii-minami, Neyagawa, Osaka 572, Japan The surface modification of fluoropolymcrs was studied with intense UV radiations from excimer lasers. The adhesive strength of polytetrafluorocthylcnc film was enhanced remarkably by KrF-lascr irradiation in air when a small amount of aromatic compounds was blended. Tctrafluorocthylenc-pcrfluoroalkyl vinyl ether copolymer was endowed with the high wettability by ArF-lascr irradiation in water dissolved carbon monoxide. The enhancement of the wettability was closely related to the decrease of the fluorine/carbon atomic ratio and the increase of the oxygen/carbon ratio. Keywords: Fluoropolymcr, Surface modification, Excimcr laser, Adhesiveness, Wettability 1. INTRODUCTION Fluoropolymer materials have many excellent properties such as thermally stability and solvent resistance with poor adhesiveness and wettability because of the chemical inactiveness. Therefore, the improvement of the surface of the fluoropolymcrs with keeping the excellent properties is very important from a viewpoint of the enlargement of applications. At present, the surface modifications, for example improvement of the adhesion and endowment with the wettability, of fluoropolymcrs has been carried out by an alkaline metallic solution1' and a plasma discharge2', but we could not be satisfied completely with these effects. Therefore, we have attempted to proceed efficiently the surface modification of typical fluoropolymers using an excimer laser cooperated with Kurabo Ltd.3"5) and Gunze Ltd.6) We chose polytetrafluorocthylcne (PTFE) and tctrafluoroethylcnc-perfluoroalkyl vinyl — 246 — ether copolymer (PFA) as fluoropolymers. Polytetrafluorocthylene (PTFE) is a typical fluorine resin which has such excellent heat resistance, chemical resistance and electrical properties. However, because of the chemical inactiveness, PTFE has poor adhesive properties. Therefore, the improvement of the surface properties of PTFE is very impor tant from a viewpoint of the enlargement of applications, for example, in which PTFE is laminated with various materials. On the other hand, PFA has advantages of the molten processing with excellent properties same as PTFE. The development of an effective method for the endowment with wettability has become a very important subject for the enlargement of applications such as medical-use. In this study, we report the effective improvement of adhesiveness and wettability for the surface of PTFE and PFA, respectively, using an intense UV-radiation from excimer lasers. These treatments are very convenient and the effects are more effective than ordinary engineering treatment such as chemical etching. 2. EXPERIMENTAL PROCEDURES Samples The PTFE film (0.5 mm thick) was prepared from the blended moldings of PTFE (Fluon G-163, Asahi-ICl Fluoropolymers Co., Ltd.) and aromatic polymers obtained by the sintering at 360 °C for 3 h using a peeling machine. Three kinds of the aromatic polymer used as additives to PTFE are aromatic polyester (APE, Ekonol E-101S, Sumitomo Chemical Co., Ltd.), polyetheretherketone (PEEK, Victrex PEEK 150-P, ICI) and polyimide (PI, Aurum, Mitsui Toatsu Chemicals). The chemical structures of the aromatic polymers are shown in Table I. The PFA film was prepared from the extrusion processing of PFA resin (Teflon PFA-350J Mitsui-Dupon Fluorochemical Co., Ltd.). These polymer films were cleaned in ethyl alcohol with a supersonic wave before irradiation. Table I Aromatic polymers blended to PTFE as UV absoeber. (a) Aromatic polyester 4°-^0^-co-rs- (b) Polyetheretherketone (c) Polyimide -^o^tor^X^tgr -247- Irradiations The polymer was irradiated with a KrF laser radiation (wavelength: 248 nm, flucncc: 150 mJ cm"2pulsc_1, frequency: 2-40 Hz) and an ArF laser radiation (wavelength: 193 nm, 13 mJ cm^pulsc"1, frequency: 50 Hz) from an excimer laser (Lumonics Hyper EX-460). The quantity of laser radiation was determined by using a calorimeter (Scicntcch 38- 4UV5). Measurement The adhesive strength of the PTFE film was evaluated by the measurement of the 180° peel strength of the film adhered on a stainless steel plate (SUS 304) with the cpoxy resin adhesive. The Wettability of the PFA film was evaluated from the contact angle towards water at room temperature. The chemical state of the polymer surface was analyzed using an x-ray photoelcctron spectroscopy (XPS) by Shimadzu ESCA 850S. Atomic ratios of carbon, oxygen and fluorine and the chemical bonding were obtained from the analysis of the Cls (300-280 cV), Ols (526-542 cV) and Fls (670-700 cV) XPS spectra. The polymer surface was observed by a scanning electron microscopy (SEM, Hitachi S-2400). 3. IMPROVEMENT OF ADHESIVENESS ON PTFE Polytetrafluorocthylcnc has no absorption in UV region, the chemical reaction hardly occurs in only PTFE even if the intense UV-radiation from excimcr lasers is irradiated to PTFE. In order to induce the chemical reaction three kinds of aromatic polymers shown in Table I were add to PTFE as UV absorber. The aromatic polymers have a strong absorption band in UV region and a high thermostability even at the molding temperature of PTFE. Figure 1 shows the adhesive strength of the PTFE film containing a small amount of APE, 1-20 wt%, as a function of the laser-irradiation energy. The adhesive strength of the PTFE film increased steeply with the irradiation energy in all the cases of 1-20 wt% ranges of APE content and furthermore the adhesive strength increased linearly with the APE content. In the case of the APE content of 20 wt%, the adhesive strength became remarkably large from 0.4 kg cm"1 to 3.0 kg cm"1 with the irradiation energy of 1.8 J cm-2 which is much larger than that, 1.2 kg cm"1, obtained by the chemical method. On the other hand, the adhesive strength of the PTFE film without APE was less than 0.02 kg cm"1 even at irradiation of 150 J cm"2. Therefore it was found that the adhesive -248- strength of the PTFE film containing ->—r a small amount of APE was 20% Aromatic polyester enhanced remarkably by KrF-lascr E o irradiation in air at room XT temperature, and also the adhesive strength is able to adjust in a wide C range with the irradiation energy. CD i_ •+-• These results indicate that the surface CO CD treatment by the laser irradiation is > *co CD favorable for the surface modification .c -a of PTFE. Effects on the co enhancement of the surface adhesive properties similar to APE were 0 4 8 observed in the PTFE containing Laser-irradiation energy (J/cm2) PEEK and PI. The PTFE surfaces irradiated by Fig.l Adhesive strength of PTFE the KrF-lascr radiation were irradiated by KrF laser. analyzed by an XPS and a SEM. Cls XPS spectra of the KrF laser irradiated and non-irradiated PTFE films containing 5 wt% APE are shown in Fig. 2. The Cls XPS spectrum of the non-irradiated PTFE film consisted the component due to -CF2- at 294.7 eV of the binding energy and a small part of the component due to APE and other impurities at 285-290 eV. With the laser irradiation up to 3 J cm"2, the peak at 294.7 eV remarkably decreased and those at 292.4 and 285-290 eV increased and also the other component at 290 eV increased 300 280 as the shoulder. These components Binding energy / eV were identified as -CHF-CHF- or - Fig.2 XPS specra of PTFE. a:non-irradiated, CF=CF-, -CH2-CF2-, and -C=0, b:KrF-laser irradiated. — 249- respectively. The atomic ratio F/C decreased from 1.93 to 0.3, on a contrary O/C increased from 0.02 to 0.08 with laser-irradiation energy of 3 J cm"2. These changes of the XPS spectra reveal that the release of the fluorine atoms, the formations of carbonyl group and ethylene linkage, and the carbonization occur on the surface of PTFE by the laser irradiation. Such chemical changes were not observed in the case of the PTFE film without APE. Similar chemical changes were observed in the cases of the PTFE con taining PEEK and PI. These results indicate that the chemical reactions and the laser ablation occur on the surface of the PTFE film containing the aromatic polymers by KrF-laser irradiation. Since the absorption coefficient of PTFE is very small in UV region and the aromatic polymers have a strong absorption band, most of the high intense UV light from the KrF laser would be absorbed in the aromatic polymer. It is therefore considered that the energy transfer from the aromatic polymer to PTFE and the laser ablation occur in the PTFE film containing the aromatic polymers, followed by the chemi cal reactions described above. From the surface observation by a SEM, many fine, ca. 0.1 \im, unevenness were observed on the PTFE surface. It is also considered that a remarkable enhancement in the adhesive strength of the PTFE film containing the aromatic polymers by KrF-lascr irradiation is attributable to both the chemical effect owing to the formation of the polar groups such as carbonyl group and ethylene linkage and the physical effect owing to the formation of the fine uncvenness by the carbonization. 4. ENDOWMENT WITH WETTABILITY ON PFA The PFA film was irradiated with an ArF laser in distilled water which various gases were dissolved in. Each gas was bubbles for 60 min at room temperature to be dissolved up to the saturated concentration. Figure 3 shows the change of the contact angle with ArF-laser irradiation. The contact angle of non-irradiated PFA film was 106 degrees, however, the contact angle decreased remarkably to 34 degrees by ArF-laser irradiation (energy: 208 J cm'2) in water dissolved carbon monoxide. Also by irradiation in water dissolved carbon dioxide or nitrogen, the contact angle considerably decreased to 48 degrees or 70 degrees. On the other hand, no change in contact angle of the PFA film was observed by irradiation (energy: 208 J cm"2) in water dissolved oxygen. Thus, it was revealed that the endowment with the wettability on the surface of the PFA film is influenced by the -250- 120 11 —I 1 1-—I - | — I- I r —I [' —r ' dissolved gases in water. We found that carbon monoxide ArF-laser irradiation dissolved in water plays an QJ - important role in the endowment T3, OQ . \ Plasma discharge D) with the wettability. The C 0$ endowment with the wettability •«—< ^\ Chemical treatment was hardly observed by KrF-lascr o as - O^^Q irradiation. 1=•4—c<1 40" o In the case of ArF-laser O irradiation in water dissolved 0 1 i < . . 1, i . i i .„!. , carbon monoxide, Cls XPS 0 100 200 spectra of the PFA film showed Laser-irradiation energy (J cm-2) five peaks, which were newly Fig.3 Wettability of PFA improved assigned to -C=C- (285.5 eV), by ArF laser. and -C=0 and =C-0-C- (288 cV), -COO (290 eV), -CF2- (294 cV) observed on non-irradiated 0.02 PFA. Furthermore, the peak area of the Ols XPS spectra increased, c o c and that of the Fls XPS spectra X3 o decreased by ArF-laser irradiation. c3 cc O O These results suggest that the release of fluorine atom, the sc 1- 0.01 c CD introduction of oxygen atom, and ° D) X the formation of ethylene linkage U- O occur on the surface of the PFA film by ArF-laser irradiation. 0 Figure 4 shows a relation 40 60 80 100 between the atomic ratios (F/C, Contact angle / deg. O/C) and the contact angle on the Fig.4 Wettability of PFA and surface of the PFA film. It is atomic concentrations. obvious that F/C decreased and Oroxygen/carbon ratio, O/C increased with the wettability ®:fluorine/carbon ratio. enhanced. In the case of irradiation in water dissolved carbon monoxide, O/C became 4 times and F/C became 1/3 times compared to non-irradiated PFA. Therefore, these results indicate that the -251- enhancement of the wettability, the release of fluorine atom, and the introduction of oxygen atom on the surface of the PFA film were closely correlate with each other. It can be considered that the remarkable enhancement of the wettability of the PFA film by ArF-lascr irradiation in water dissolved carbon monoxide is mainly attributable to the chemical effect owing to the formation of the polar groups such as ethylene linkage and carbonyl group. Further, the effect of depression of dissolved oxygen gas suggest the possibility of a radical reaction. Wc arc considering that the absorbed energy transfers from carbon monoxide to PFA and water, followed by the chemical reactions such as scission of the C-F bond at the interface between the PFA film and water, because the photon energy of ArF-lascr radiation is larger than the binding energy of C-F bond. But details arc now studying. 5. CONCLUSION The adhesive strength of PTFE film was enhanced remarkably by KrF-lascr irradiation in air when a small amount of aromatic polymer was blended with PTFE. The enhancement of adhesive properties are attributable to both the chemical effect owing to the formation of the polar groups and the physical effect owing to the formation of the fine uncvenness by the intense laser irradiation. Tctrafluoroethylcnc-perfluoroalkyl vinyl ether copolymer was endowed with the high wettability by ArF-laser irradiation in water dissolved carbon monoxide. The enhancement of the wettability was the formation of the polar groups with the release of fluorine atoms. These result shows the surface of fluoropolymcrs can be modified by excimcr laser conveniently and the modified properties are most excellent than ordinary engineering treatments such as chemical etching. References 1) E. R. Nelson, T. J. Kilduff, and A. A. Bcnderly, Ind. Eng. Chem., 50, 329 (1958). 2) H. Schonhorn and R. H. Hansen, J. Appl. Polym. Sci., 11, 1161 (1967). 3) M. Nishii, S. Sugimoto, Y. Shimizu, N. Suzuki, T. Nagasc, M. Endo, and Y. Eguchi, Chcm. Lett., 1992, 2089 (1992). 4) M. Nishii, S. Sugimoto, Y. Shimizu, N. Suzuki, T. Nagase, M. Endo, and Y. Eguchi, Chem. Lett., 1993, 1063 (1993). — 252- 5) M. Nishii, S. Sugimoto, Y. Shimizu, N. Suzuki, S. Kawanishi, T. Nagasc, M. Endo, Y. Eguchi, Polym. Mater. Sci. Eng., 68, 165 (1993). 6) A. Okada, Y. Ncgishi, Y. Shimizu, S. Sugimoto, M. Nishii, and S. Kawanishi, Chcm. Lett., 1993, 1637 (1993). — 253- Session V Selective Photoreaction Proceedings of the 6lh International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Overview of Selective Photo-reaction Takashi flrisaiva Japan Atomic Energy Research Institute Tokai-mura.Ibaraki-ken, 319-11,Japan Tel.0292-82-5918,Fax.0292-82-5572 Rbstract Selectiue reaction process Especially isotope separation is a key technology for the deuelopment of the technologies related to the nuclear energy. "Howeuer oifly a few species are separated on a production scale using the conuenuonal processes such as thermal diffusion, chemical eKchanae reaction and distillation for lighter isotopes ,and gas centrifuge and gaseous diffusion for uranium, fls these methods are based on statistical thermodunamics and haue law enrichment factors, they need repetitiue operations of separation uiith many separating units combined together. Electro-magnetic separation method known as the one with high separation factor can be applied to most of the elements ,but extremely low production rate is realized, which is uneconomical. From the aboue point of uiew, much attention has been paid to the laser process by which high selectivity is expected. This method can be applied to either gas .liquid or solid hase, and hiqn separation factors are basically realized only in gaseous phase. Since the Eeginning of the studies on isotope separation in early l97Bs, many ideas haue been proposer/for the selectiue photo-reaction process such as photoionization, muitiphoton dissociation and state selectiue chemical reaction. Rs a result of experimental and theoretical efforts, large scale production of some isotopes haue been intended. Production of deuterium Dy infrared multl-photodissociation method was studied aiming at the replacement of the conuentional dual temperature eHchange process, and lots oT experiments haue been achieued intensiuely for the uranium enrichment, fl stepwise selectiue photoionization method has also Been studied for the isotopic separation of many elements, especially uranium enrichment.To implement the laser processes on a large scale production system, aduanced performances are required not only for the tunable laser sustems bur also for many related technologies such as atomic/molecular source,pnoto-reactcr and extractor of products. Keywords:Laser Isotope Scparation.Selcctive Reaction,Photo-ionization,Photo-dissociation,SeIectivity 1.Introduction Tablet.Selectiue photoreaction by tunable IR lasers Selective photoreaction is Source Wavelength made based on the Laser Conversion "type of Reaction Working Materials spectroscopic difference CF3H,CFa2H,CF3CCIH. caused by the difference Halogenaled elements, isotopesjsomers Freonl23,CF3COH and isobars. In the field of hydrocarbon nuclear engineering, isotope Infrared Multi-photon CF2CI2,CF3I.CCI3,CF2C1H separation technique special attention is paid to TEA Dissociation Hexahalidc Si2F6.SF6,SeF6,WF6 the wide area ranging from CO, Tetrahalide SiF4.TiCl4 the uranium enrichment and isotope tailoring materials to Laser Trichloride BCI, the reprocessing of spent fuels. Among many Metal Zr(OC(CH3)3)4. conventional isotope Alkoxide Ge(OC(CH3)4) separation methods the laser method is most Others OsO. promising as a technology of Infrared Excitation +UV Ammonia NH, next generation in terms of Dissociation both energy efficiency and Halogenated selectivity. Laser Isotope Optical CC13H Separation experiments Pump Single-colored IRMPD hydrocarbon NH3 have long been achieved Hexafluoride UFfi since early 70's. In convenience the laser Raman Multi-colored IRMPD Hexafluoride UFA isotope separation methods p-H. CW Laser I nduced Thermal are divided according to the Hexafluoride SF< wavelength range used. Diffusion co2 One of these is the Halogenated Laser CCI2FH selective photoreaction by Laser Assisted Nozzle hydrocarbon tunable infrared laser and Method the other is the one made by Hexafluoride SF, tunable visible or ultra-violet laser. CO Laser Induced Chemical Hexafluoride UF, Laser Reaction (CRI SLA) 1.1 Selectiue Reaction with Tunable Infrared Lasers (1) TEA C02 laser based reaction Infrared Multi-photon Dissociation method (IRMPD) is most widely used by taking advantage of — 255 — discrete tunability of C02 laser irradiating special molecules which have resonant spectroscopic absorption peaks. Especially halogenated hydrocarbon, hexahalide, tetrahalide .trichloride and even metal alkoxide are used. Dissociation of molecule like ammonia by infrared excitation with ultraviolet excitation was also applied. In order to convert wavelength of C02 TEA laser, optically pumped NHo or CF4 lasers .especially Raman p-H? lasers are used for the dissociation of uranium hexafluoride based on the multi-colored IRMPD. (2) Reaction by the other types of lasers Continuous wave C02 laser is used for laser induced thermal diffusion and laser assisted nozzle method. In the very special case CO Laser is used for the laser induced chemical reaction (CRISLA). HEATER WMlffl Wavanumberf cm"' 1 HCI 00^7"^/ ^^^^^ fiacharound "101 USER UNIT IMC . loco 8 Wavenumberf cm"' 1 Fig. 1 Laser Induced Thermal Diffusion ) Fig.2 IR Multi-photon dissociation and hydrogen isotope eitchange process8^ UF -F Basic Reacti 5 UFSCI 235 235 UF6 +3hv3-> UF6 ->UF5CI + HF 5.33 it rn by CO laser 235 a =6x2\8 a V-T relaxation ../2 235| _/i UFfi + HCI -» ^UF. + HCI + =/ 0 \V-V scrambling ^UF, + ^UF, B5UF * Fig.3 CRISLfl method(IR Laser Induced 6 6 Photochemical Reaction) 1.2.\Selectiue Reaction with Tunabl^ Uisible or Ultrauiolet lasers Photochemical reaction by tunable visible\lasers is made by a dye laser.which can emit tunable photons with many colors of dyes. CW laser pump dye lasers are used for laser induced chemical reaction photochemical reaction in gaseous/U,quid phase and atomic beam deflection. Diode CW lasers dre also used for radiation pressure. Pulled laser are used for photo-dissociation, especially single photon dissociation and for laser induced chemical reaction, but most popular ana widely applicable selective reaction is a multistep photo'ipnization. (1) Single photon dissociation \ Single photon dissociation is seen in visible and ultra-violet wavelength region. Typical reaction systems are given forthe dissociation of symmetric tetrazine and formaldehyde. a.Dissociation by Visible Photon s-Tetrazine(H2C2N4) + hv -*HCN+N2 b.Dissociation by Ultra-Violet Photon H2CO+ h v -*H2+CO Combination of ultraviolet dissociation and cascade is proposed by LLNL1> Physical and — 256 — operational parameters are given as follows. Table 2. Selectiue photoreaction using tunable uisible/ultra-uiolet lasers Source Wavelength Laser Conversion Type of Reaction Working Materials Dye Laser Laser Induced Chemical Reaction IC1 CW Laser Photo-dissociation H2CO,C2H2N4 (Pump) Photochemical Reaction (Liq.) U02F2+HF+CH3OH Atomic Beam Deflection Ba CWLD Radiation Pressure Rb Pulse Photo-dissociation H2CO,Br2 Li+CCl FH,Yb+HCl,CSCl ,U(BH ) , Laser Dye Laser Laser Induced Chemical Reaction 2 2 4 4 (Pump) U02(HCOO)2H20 Multistcp Photoionization Li,Mg,K,Ca,Ti,Fe,Cu,Ru,Zr,Pd,Ba,Hg,Tl, Nd,Sm,Eu,Gd,Dy,Er,Tu,Yb,U,Pu Pulse Eu3+H 0 Laser Element Selective Photochemistry 2 And the following enrichment is expected. 13c 1.1 %->~70%(single stage),95%(2 stages) 180 0.208%->—40%(single stage),95%(2 stages) 17,0 0.038%->~7%(single stage),95%(3 stages) Table 3. Cascaded single photon dissociation process Physical parameters Operational Parameters Wavelength 290-355nm Laser power 129W Photoabsorption cross section 5xl0'17cm2 Laser PRF 4350Hz Spectroscopic selectivity 200(13C,17O),300(18O) Laser pulse width 40ns Quantum Yield 0.95 Temperature 300K Photolysis cell length 10m Geometry Number of passes 10 + \ELF + H2O" -»- -OH Reaction Coordinate ^cavenagr 340 360 380 400 420 440 460 480 500 320 WAVELENGTH (nm) Fig.4 Oxygen isotope separation in liquid Fig.5 Element selectiue photoreaction in liquid phase phase (2)Laser Induced Selectiue Chemical Reaction Photo-decomposition was made with Uranium Boro-hydride (-deuteride)by 2-step Scheme by — 257— E.R.Bernstein(COO-4118-1),in which 2-5% U(BH4)4/Hf(Zr)(BH4)4 was irradiated by tunable laser at low temperature to decompose it as follows;U(BH4)4+h v -*U(BH4)3.Photochemical reaction was also observed in Uranyl Formate in the following way2), +4 +2 U02(HCOO)2H20+h v -*U ,U02 (3)0ther reaction methods Laser induced sputtering of SF6 from the cold Kr matrix are made using TEA C02 laser ,and Isotopic Selectivity A3 was obtained when laser illuminated non-target molecules.3' Isotope separation in liquid phase was carried out for hydrogen isotope separation based on the photolysis of formaldehyde in liquid xenon"' 2+ 5 Oxygen isotope separation was made with U02 in methanol solution as follows. ) U02F2+2HF+CH3OH-+UF4+Organic Products Photochemical reaction selective to elements but not to isotope was achieved in liquid phase for europium.6' 3+ 2+ 2+ + Eu H20+h v -s.[Eu H20]*Eu +H +'OH 2 Eu^+SO.JA - —L.UVJV--»EuSO4, 2. Htomic Uapor Laser Isotope Separation As this method is based on the resonant multiple photoionization process, atoms can easily be ionized with low photon flux.which is related to the high utilization efficiency of photons, and the atomic spectrum is enough narrow so that the photons with tunable wavelengths are easily absorbed by the 235U without being absorbed by 238U atoms.Therefore ,this method . could achieve separation at high separation factor by one single operation. . could be economical. . could enrich uranium from the reprocessed uranium .which leads to the effective utilization of natural resources. Tabie4.MPI schemes for RIMS Cilliclot to dtpltfed UWIIIIB Element Ionization . Types of RlSScheme»a»erc AJ ««278 o,»o, "L*»I" Itilxr AJ 78930 2»,*«, J-,"-!*" Idtwdbbslgte to 7«40S> 2U.4U, "'*"' Al D tent IMircrihlt C 90820 2o,«., ntllei C 49306 2",4». 2",4U," Siltcllii J'oli Cd 12140 UL+U)4W| "|4»l*"l bokpt shift 2»,+u, Co 63364 ",+•>, U,4»,40, Q 54576 «'*»' U,4U, -,•>»,"' MiMIc Cnud slate a 31406 C,40, ».*».*». 2u,+u, state exciter ticiler Cu 62317 U,4U, »l*». u.+u," "Ulintriihld • "V t(reitad rialt 1 Linear ntlal nopiioloi IkluM itilt' Fe 63737 »l*»l ",*». 2u,+u, • "Ullicitelitilel GtotiM iloli Ge 6371} «,4U, *,+»,+», Ug-tUjiUj »°*U(Mri Hold u.+o.tu, Hj 84184 Enetj) Eicllolion Scheme "l'»l*"l Conceptual Drawing <,,**,+<*," Kr 12914 ».»»> u,«u,+ «a. 2u,*u, Ml 61671 -,•«," U,+u,iU| 2u,-u, Ni 61619 U, + o, u,+o,*u, Fig.6 Concept of B.ULIS process Pb 59820 ».*»i 2o,+Ul (I^U^CJ, SI 65748 «.•". VI,+!•, + », Zn 757W «l+«l+«I U,4J.,4U, &,+«, 2u,tu, Whole separation process is divided into four major subprocesses for convenience, ie. evaporation process for generating atomic beam, photoreaction process for ionizing isotopes selectively by resonant multistep photoionization method.ion extracting process for recovering isotopic ions from laser induced plasma, and light propagation process for propagating laser — 258 — beams in the optically deep atomic medium with the interaction between photons and specific elements. Some of the typical results from our experiments under the AVLIS (Atomic Vapor Laser Isotope Separation) process are presented8'. .I.Uersatile applicability Isotopes of almost all the elements except noble gases can be separated by AVLIS process. Normally visible laser process is more convenient than the ultraviolet laser process. Because in the shorter wavelength region special optical components must be sometimes developed. It is easy to gettunable visible sources using a variety of laser dyes,but tunable ultraviolet sources are obtained by non-linear optical crystals which converts wavelength into shorter one. These crystals suffers from thermal stress or degradation, when the laser power is increased. 1111111111 O 5 10 15 20 25 Laser Frequency Difference (GHz) Fig.7 Lithium isotope separation8) Fi98- Hyperfine structures of Gd8> Fig.6 shows the conceptual drawing of AVLIS process based on 3-step photo-ionization method ,in which foursubprocesses are indicated. Evaporation process includes the electron beam gun heating and production of atomic vapor. Avtr»g» Allowtd population Level* transition* distribution Photolon ballon (J * 0) 0% 100 Second axclttd (J «1) , . 13% 34% First MCIIPCMJ - 2) I , 1 Ground (J K 2) I 1 S3% mj (alactronlc) -i -1 0 t 2 m, 3 Selection rules for even Gd isotopes with no magnetic Herd and parallel linear polarization. c HI Population E Laval* AJIOWMI flow relatlvt tram itiom to ground iut> Photolonlutlon (J«0) 70% SacondnclIM (J*1) .V k fcr ?«* Flrslticll»d(Ja2) •7% Ground (J a 2) IB 111 100% mp (nuclitr ptui TTTT Tltt TTTi MTT f-fTT n. lUctronlc) -1 -0.5 0 0.5 1 1.5 m, t«Uctronlc) Dye laser frequency difference (GHz) Selection rules for odd Gd isotopes with no magnetic field and parallel linear polarization. 8 Fig.9 Ti isotope separation ^ Fig.IB Polarization selection for Gd8) -259 — 2.2 Selectiue MPI process Well-tuned tunable laser beams with different three colors illuminate the isotopic medium at the same time after combining each color. The specified isotopic atom at the ground state is excited to the selective excitation level discriminating the isotope shift. Excited atom is then further excited to the ionized state through the two-step transition.i.e. intermediate excitation and ionization. In the case that many atoms are thermally populated in their lower metastable states due to the high temperature evaporation.some of those states could be processed by adding another tunable laser beams excite them to the common selective excitation level from the ground state. Multiphoton ionization schemes of several atoms are listed in the NIST table.but precise spectroscopic studies are required for searching the economical isotope separation7' 2.3 Deuelopment of HULIS process in large scale (1) Process deuelopment The photoionization process includes all the above processes. Recovery process consists of the collection of depleted metal and the extraction of product ions.Light propagation process deals the interaction of long traveling laser beams and near resonant medium to increase the production rate. Fig.11 shows the process evaluation scheme to optimize the sub-processes to construct the whole integrated process system. Under the development program of evaporation process.data are taken on velocity of atoms in the vapor.intemal energy of atoms.spatial distribution of vapor ,vapor density and evaporation efficiency. For the photoionization process, spectroscopic data are taken on the energy levels, photo-absorption cross section, excited state lifetime, quantum number.isotope shift and hyperfine structure which are used for optimizing the combination of wavelength.photon flux,detuning. For ion extraction process.data are taken on the characteristics of laser induced plasma.charge exchange cross section .scattering rate, sputtering rate to obtain short extraction period of time and low enrichment loss. For the light propagation process .high density linear evaporator ,4- color tunable laser system are constructed to collect product ions onto the electrode and to evaluate the photon utilization. From these data energy efficiency and scale-up law for separative work are determined. • Quantitative analysis of separation parameters- • Scale-up law - Energy efficiency- (kgSWU/kWHT t t Separation experiment I Component Function - Quantity of product I • Energy efficiency - Enrichment I - Manufacturing cost 1 ¥ I - Maintenance Light Propagation Process . Multi-color tunable laser system - High density linear evaporator - Light propagation process In linear vapor medium • Collection of products onto electrode • Photon utilization I Evaporation Photo Ionization Ion extraction Component process process process development • Atomic velocity - Collection of • Characteristics • Copper vapor spectroscopic of laser Induced laser data plasma - Internal atomic • Combination • Charge- exchange - Dye laser temperature of wavelengths cross section - Vapor utilization • Optimum • Reduction of • Electron beam utltlzablllty photon flux extraction time evaporator • Evaporation - Optimum • Non-solcrctlve - Ion Extractor efficiency Frequency pick-up Vapor density . Detuning • sputtering - Depleted uranium oxtmotor Fig.l 1 Deuelopment of whole separation process (2) Euaporation process Several methods are considered for providing atoms into a photo-reactor. Those maybe a gas cell,long linear Knudsen cell.two-dimensional nozzle or electron beam gun evaporator with linear crucible. Among above devices an electron beam method is convenient because of its applicability to many kinds of elements with high melting point. In generating atomic vapor, following points are also considered;narrow Doppler width .generation of high density atomic beam .high directivity of atomic beam .well cooled internal energy states .high energy efficiency ,non-selective ionization of atoms by collision with electrons .behavior of electrons on the surface of liquid metal etc. — 260- The internal temperature decreases with the increase of the evaporation rate, and atoms can be cooled down more than 1500K. When feed metal is evaporated in the vacuum chamber by electron beam bombardment at very high temperature.which thermally excites atoms from the ground state to several lower metastable states. Highly populated lower metastable states are not appropriate for the efficient photoexcitation, because the laser beams illuminate atoms mainly in the ground state. Atomic vapor thus generated is cooled down by around 1000 K due to the adiabatic expansion.which increases the population of the ground state in the photoreaction zone. In the case that some lower metastable states are considerably populated even after cooling, some of the metastable states are used as target levels.Each population equilibrated at 2500K is around 70% (ground state) and 25% (lowest metastable state) in the case of uranium atom. With the 4-wavelength-3-step photo-ionization scheme more than 90% of feed atoms is ionized from both states. Heating the metal with high power electron beam may cause also thermal ionization and collisional ionization which dilutes the enriched photoions at the photo- reaction zone. (3) Photoionization process For AVLIS process, tunable visible lasers are used rather than UV because of their high energy conversion efficiency and easy to control the beam quality. As the visible photon has the energy of 2eV, normally 3-step photoionization scheme is employed to ionize the target isotopic atoms by taking into considerations the ionization potential around 6eV. This multistep process makes use of the real energy levels and is characterized by high ionization efficiency and high selectivity.To implement multi-step photoionization.data related to the energy levels and transition probability between those levels. The former data include the intermediate excitation and ionization levels,parity and total angular momentum represented by J-value. Photoionization includes the direct transition from the intermediate excitation level to the continuous energy spectrum and the autoionization to the high lying levels above ionization limit through the resonant interaction with the continuous spectrum of free electrons. The rate for the former process is faster than the latter.which makes it important to search for the autoionization levels with large photoionization transition probability. Data concerning transition include photoabsorption cross section caused by the electric dipole moment and the total integrated photoionization cross section calculated for the 3-step photoionization. As the former value especially control the total ionization rate, the most important issue is to find out the ionization scheme with large cross section. Spectroscopic data obtained so far are limited to the lower energy region and they were taken using hollow cathode lamps under poor resolution. Measurements ranges over the energy levels.transition probabilities between levelsjsotope shifts and hyperfine structures. Based on the fact that the oscillator strength is independent of isotopes, measurement is made at first stage for the isotopes without hyperfine structure and at the second stage isotope with hyperfine is measured after the characterization of the capability for ionization. From the view point of the restriction on the wavelength of the tunable lasers pumped by copper vapor lasers,4-wavelength-3step method is adopted, which ionize both the ground sate and the lowest metastable state through excitation levels located at 14000—18000cm"1,intermediate levels at32000~36000cm ,and ionization over the ionization limit 49958cm"1. Data to be taken concerning energy levels.total angular momentum(J-value),parity,autoionization width,hyperfine structure, isotope shift .photoabsorption cross section .lifetime, branching ratio ,etc. are not theoretically obtained for heavy elements even with the most advanced computer code. From this point of view JAERI has been collecting spectroscopic data on heavy metallic elements. Measurement method employed is mainly based on the resonant ionization using pulsed laser by which several kinds of optical properties can be obtained by scanning wavelength, power, pulse timing, polarization of tunable lasers. For the measurement which requires high resolution such as hyperfine structure stemming from the interaction between atomic nucleus and orbital electrons, Zeeman-effect, resonant fluorescent tunable lasers with narrow linewidth are employed. Spectroscopic studies was first done for 238U with zero nuclear spin to obtain data on level.c'ross section,J-value,lifetime and branching ration, and then optimization is made to give each wavelengths combination the integrated equivalent cross section. Succeedingly data are taken for235U with non-zero spin to obtain hyperfine structure and isotope shift for the further evaluation. Finally photoionization dynamics was studied by simulating the photoreaction process based on the quantum mechanical model to obtain both the optimum photon flux and detuning. Photoionization dynamics is generally described by quantum mechanical equations. However when ion extraction rate is fast or when the laser linewidth and absorption spectrum width are wide,coherent interaction is not much dominant. In such cases multistep equation system is roughly converted to the rate equation model without noticeable error.There are two typical methods. One is to use single mode laser with fixed frequency so that the highest ionization rate can be obtained by scanning the laser frequency. Second method is to modify the laser frequency during one single laser pulse so that every part of the atomic spectrum feels the laser frequency for the various laser frequency scanning rate and direction. As a result of calculation, frequency modification method can generally increase the ionization rate efficiently depending the actual - 267 — hyperfine structure. Especially ionization of atom with widely spread hyperfine is considerably improved. 066. 561 Wavelength [nml Wavelength ( nm ) Photoionization Spectrum Wavelength Inm) Autoionizing Rydberg Series Fig. 12 Examples of photoionization spectrum8^ (4) Ian extraction process Generally speaking density of ions extracted with the static electric field from the relatively active plasma with the electron temperature of around 10eV should be lower than 1012ions/cm3,if effective extraction is expected. Space charge limited current is written by the Langmuir-Child's formula . On the other hand ion current is limited by the electron temperature of the plasma.Based on these experimental data we take the following measures to extractions effectively from laser induced plasma at low electron temperature. Highest allowable extraction speed is realized by applying voltage as high as possible to collect as'much as ions from the moving plasma at the similar speed with the neutral atom. Higher speed is realized by increasing electron temperature. Avoidance of sputtering from the once collected ions on the electrode due to high momentum of ions, interaction period as short as possible to reduce the chance for charge exchange and ionization collision are also important. From these points of view plasma heating can be considered to enhance the electron temperature, butcare must be taken notto heat up electron so high as the ionization limit. As the AVLIS process is usually applied to the separation of trace amount of isotopic components, the degree of ionization is low, even if 100% ionization of specified isotope is achieved, that is, the ions coexist with huge amount of neutral atoms.For this reason during the extraction of isotopically pure ions from the laser induced plasma.charge exchange easily occur by the collision between ions. The charge exchange cross section is obtained as a function of collisional energy for the reaction system of 238U + 235U+->- 238u+ + 235U , which is found to be similar to the semiempirical formula introduced by S.Sinha. Scattering of neutral atoms in the atomic beam is also important. We measure the adhesion of the neutral atoms as a function of atomic density. As a result of the experiment the adhesion rate of neutral atoms from the atomic beam onto the electrode is turned out to be mainly caused by two body scattering process. Special care should be taken in order to avoid charge exchange and scattering from the neutral atoms. (5) Light propagation process In the application of AVLIS process to the enrichment plant tunable laser beams interact with the specified atomic isotope in the long path of medium aiming at the enhancement of the reaction volume up to the order of m3 which is the product of cross sectional area and path length of the laser beam. This long propagation also enhances the photon utilization — 262 — efficiency. During the propagation in the reaction medium the wavefronts of the laser beams are degraded by nonlinear effectTaking into considerations the atomic beam velocity and the laser beam width, thermal blooming phenomenon does not occur because of the short pulse width, but the self (de-)focusing is observed due to the local change of the diffraction index.This phenomena are caused not only by the resonant photoreaction of the specified isotope but also by the near resonant interaction with nonspecified isotopes.After the self focusing and self trapping phenomena, laser beam might make a filamentation or decrease the upper state population by superradiance. These phenomena induce big changes in spatial distribution of laser beam intensity and frequency which brings out inefficiency in the separation process. We measured the change of a laser beam radius in a small detuning region of the tunahle laser frequency for the two level system in the medium with the long propagation distance. Measurement was preliminary made concerning the effect of light propagation on the temporal .frequency and spatial region using heat pipes. After the experiment we started to measure the light propagation effect through the long atomic vapor of uranium atoms as a function of detuning.Simulation of spatial distribution of laser energy propagating through the near resonant isotopic medium was analyzed based on the density matrix equation of motion combined with Maxwell's propagation equation. Distinguished deformation is observed. Fig.13 Experimental set-up for eualuating separatiue power (6) Eualuation of separatiue power Separator is a target cell for tunable laser beams in which long linear atomic beam is generated so that the laser beam could interact as much as atoms to increase the reaction volume. Rg13 shows the conceptual drawing of the electron beam bombardment type evaporator for generating metallic vapor. For most of atoms with high melting point, electron beam guns with either transverse type cathode or Pierce type cathode can be used. Linear cathode type has no scanning mechanism.whereas Pierce type has scanning mechanism to get uniformly distributed atomic density along the linear photoreactor. We constructed one meter long linear evaporator with scanning electron beam gun which is operated by two Pierce type electron beam guns. Cares must be taken so that the laser beam colors could cover all the atoms spread in the frequency field by inducing AC Stark effect to broaden spectrum, by chirping laser frequency or by operating lasers with multiple longitudinal modes. After optimization in each subprocess mentioned above, many parameters regarding operational conditions are fixed based on the simulation for the construction of the most appropriate separation process. Then the enrichment experiment is achieved by extracting small amount of product, from which total separative power is obtained. When we look at the distribution of the product along the long linear electrode.we find that around the central part of the electrode the production rate is highest due to the high adhesion rate of scattered atoms. The distribution of the recovered products along the linear electrode was obtained. In the collection of ions from the moving plasma with the linear electrode, neutral atoms tend to scatter with each other, which makes neutral atoms to attach to the electrodes and dilute enriched product. Separative power is calculated based on the experimental values.We also propose two unique parameters which can describe the process performance based on the above -263- separative power. One of them is the energy efficiency.which is defined by the separative power per unit energy consumption ,ie. gSWU/kWh. And the other one is the specific separative power.which is defined by the separative power per unit volume per pulse.Energy efficiency is naturally related to the energy consumption and specific separation power is related to the scale up performance. 1) MPI Scheme 2) Atomic Density Spatial 3) SOD Separative Power Specific 4 Distribution of Separative Power Products and Energy Enrichment Consumption Energy Efficiency 1) Laser Beam Intensity 2) Laser Beam Size 3) Laser Linewidth 4) Electrode Interval 5) Electrode Heiaht 5) Atomic Beam Slit 7) Propaaation Distance Fig.14 Procedures for eualuating separation process Ccst Fvjluotlo/i (optimum nlumlc dunslly) Ttrget Area for 70D j- T 35 "Imole merit tlton _ • 600 SO •H Left Cecltol in 500 :5 J/(kfl SWU/yl /I 20 •10O BBIH LffttTrocuclIn VlkgSVj) 300 IS X • Rljhl Efifergi| In 200 10 JAVg SWU) y 100 5 1 m 7 0.4 0.3 0.2 0.1 Ian Extraction Efficiency Q01 0.1 1 10 Specific Separative Power Fig.15 Example of cost estimation Fig16 status of whole process eualuation 3. Summary Fig.15 shows the typical example of cost estimation where cost depends on how efficiently ions are extacted from the plasma towards electrode under the high atomic density. Several typical results are plot in fig.16 in the plane of these two parameters to show the progress status of AVLIS technology. AVLIS process may lower the uranium enrichment cost.In order to implement the laser isotope separation process economically for various kinds of elements, innovations are expected especially in the development of compact and efficient light sources.the photo reaction process and the low loss recovery process. References 1) R.C.Stcrn.K.ESchcibncr.SPIE \bl. 1859- 25,1993 ;R.H.Pagc,S.C.Dropinski,E.F.Wordcn,.J.A,D.Stockdale,SPIE\bl. 1859-05 2)C.H.Muller,J.T.Portcr,E.S.Ensbcrg;GA-C17227 3) K.Wada,M.Aratam,O.Takahashi,S.Yabushita,Y.Ino,J. of Nuclc. Sci. Tcclmol. 27,5(1991)257 4) S.M.Frcund.W.B.Maicr II.R.F.Holland.W.H.Bcattie, J.Chcm.Phys. 5) C.K.Rofcr-Depoortcr,G.L.Depoorter,Chcm.Phys.Lclt. 61(3)1979) 6) Y.IIaas, G.Stcin, R.Eniic.Isrucl J.Chem. 10(l972)529;T.Donohuc, J.Chcm.Phys.,67( 1977)5402 7) E.B.Saloman.Proc.of thc5lh Int.Symposium on RIS and its Applications, 1990,45 8)T. Arisawa,Y.Suzuki,Y.Maruyama.K. Shiba, J. Phys.D. 15,1955(1982);T.Arisa\va,Y.Maruyama,Y.Suzuki, K.Shiba,Physica,115C(1983)266;Y.Maruyama,Y.Suzuki,T.Arisa\va,K.Shiba,Appl.Phys.B. 44(1987)16 3-166;T. Arisawa, Y. Suzuki, Y. Maruyama.K. Shiba.Chem. Phys. 81,473(1983);Y. Maruyama.M. Kato, A.Sugiyama, T.Arisawa, OpLCommun. 81(1991)67-108 35);YMaruyama,T.Arisawa,Thc Review of Laser Engineering, 14(1986)429-441 ;T.Arisawa ct al.,Optoelectronics-Devices and Technologies-, 8(1993)203-230 -264- Proceedings of ihe 61I1 International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Photochemical Reactions of Actinide Ions Hiroshi TOMIYASU Research Laboratory for Nuclear Reactors Tokyo Institute of Technology O-okayama, Meguro-ku, Tokyo 152, Japan This paper reviews the results of photochemical studies of actinide ions, which have been performed in our research group for past several years as follows: I) behavior of the excited uranyl(VI) ioni-6; ||) photo-reductions of the uranyl ion with organic and inorganic compounds7-^; |||) photo-oxidations of uranium(IV) and plutonium(lll) in nitric acid solutionsii-i3. Keyword: Photochemistry, Uranium, Reaction, Laser, NMR I INTRODUCTION Extensive studies have been reported on the photochemistry of uranyl ions for mainly basic photochemical interest. However, recently inteiest has been focused on the photochemical method of adjusting the valencies of actinide ions in the nuclear fuel reprocessing. This paper presents the results of photochemistry of actinide ions, mainly uranyl ion, in order to gain better understanding for the practical purpose. I I Behavior of Excited Uranyl Ion Deactivation Mechanism of the Excited Uranyl Ion Despite extensive studies on the life-times of the excited urany ion and its quenching rate constants, the quenching mechanism of the excited uranyl was much confused for a past decade. This might owe mainly to the dual luminescence of the uranyl ion as described below: I = Ijnexpt-M) + l02exp(-k2t) (1) where I01 and I02 refer to luminescence intensities at time = 0, and ^ and k2 are the -265- decay constants. We concluded that the dual luminescence was attributed to the 2+ emission from independent two species, the excited aqua ion *U02 and excited 2+ hydrolyzed species *(U02)2(OH)2 by the following reason -It was found in our studies that the 170 resonance of uranyl oxygen atoms, which was a single Lorentzian curve in strongly acidic solutions, revealed two distinct peaks at pH > 3. These two peaks were assigned to UOg2* and (U02)2(OH)22+ (expressed as (2,2) hereafter), and the rate constant for the exchange between these species were determined to be 59.3 and 93.8 s-i at a pH 3.15 for the forward and reverse processes, respectively. Since the above exchange is much slower than the luminescence decay, the species *U022+ and *(2,2) were considered to exist independently within the time scale of uranyl luminescence. The emission from 2+ *U02 and *(2,2) were observed in time resolved emission spectra, where the spectrum at earlier 0.2 n s corresponded to the emission of *\iOg+. while the final spectrum at 9.2 A s to that of *(2,2). The quenching of uranyl luminescence is of particular interest and many studies have been performed. For a deeper understanding of quenching reactions, we measured quenching rate constants for various complexes, U022+. (U02)2(OH)22+. U02(H2P04)2 and U02F42-, in the presence (kg) and absence (k0) of halogen and halogen-like anions. Figure 1 represents plots of log kq versus log ko. I n Fig. 1, the plots are linear for each quencher and all lines join at one point. This is explained well in terms of a linear free-energy relationship. In the absence of quenchers, the decay of excited uranyl complexes *U02(L), where L denotes a ligand such as H20 or F-, can be described by ko *U02(L) -> U02(L) (2) The linear free-energy relationship leads to the equation: AG0*(L) = AGE* + aAG(L) (3) Where AGo*(L) is the free energy of activation of reaction (2), AGE* is the free energy of activation for an excited complex, *U02(E), which is in the hypothetical^ highest energy state, AG(L) is the difference of free energies between *UO"2(L) and *U02(E), and a is a constant. In solutions containing quenching anions X-, the reaction can be expressed by — 266- kq *U02(L) + X- -> U02(L)+ X- (4) leading to the equation: AGq-(L)= AGq' + bxAG(L) (5) where A Gq*(L) is the free energy of activation of reaction (5), AGq* the free energy of activation for the quenching reaction of *U02(E) and bx refers to a constant, which differs as a function of quenchers. Elimination of AG(L) from (3) and (5) gives the following equation A Gq*(L) = (bx/a) A G0'(L) + constant (6) constant = AGq*- (bx/a)AGE* Equation (6) is consistent with In kq = (bx/a) In ko + constant (7) which is demonstrated in Fig. 1. The point in this figure where all the lines cross (we call this point as an isoquenching point) gives the maximum decay constant, which might be attributed to the decay of *U02(E), where a = b = 0, and A Go* and AGq* would have minimum values. 10 1 • 1 1 o o\ °\ O 0 ~ \° - 1 1 * 1 -1 0 1 AG2 /eV 2.0 3.0 4.0 5.0 6.0 7.0 2 1 3 -1 log *0 ^9- Plot of log *, vs. AG3; fc, •= 1.5X 10' mol" dm s , *_,/ Z = 0.25 and AC "(0) = 10.1 kJ mol"1. FIG- Plots or log *„ uersus uranyl complexes, k was determined for 2 Ihe following quenchers: (x) I", (•) SCN", IM Br" (D) Cl~ A UO S0 "°*;-3 ' - C, (U02MOH)r; D, UOa(HaPOJ,;' E, — 267- \ \ \ Quenching of the Excited Uranyl Ion by N-heteroaromatic Compounds The quenching of the excited uranyl ion by N-heteroaromatic compounds appear to be most interesting among similar quenching reactions of the excited uranyl ion by organic compounds. Quenching rate constants kq of uranyl emission were determined in aqueous solutions by measurements of lifetime of the excited uranyl ion by using the' following equation k=ko + k<,[Q] (1)' where k and ko refer to the decay constants of the excited uranyl ion in the presence and absence quenchers, respectively and [Q] is the concentration of N- heteroaromatic compound. It was found that kq decreased with the increasing 2+ vertical ionization potential lp. This suggests that the quenching of *U02 proceeds via an electron transfer mechanism with the quenchers. In order to explain the electron transfer mechanism, it is reasonable to presume that *U022+ interacts with N-heteroaromatic compound Q to form an exciplex or encounter complex [*U022+--Q], because of the 7r-bonding ability of the N-heteroaromatic compounds. Then the exciplex undergoes intramolecular electron transfer to form an unstable intermediate [*U02 +-Q +], which further undergoes reverse electron transfer to form U022+ and Q or simply decomposes to UO2 +• and Q +• On the basis of the above electron transfer mechanism, the quenching reaction can be written as a sequence of elementary steps as follows: 2+ + + *U02 + Q ^=^ [*U022+...Q] ^=^ [*U02 -Q ] k-1 k.2 / \. / k3 \i k4 2+ + + U02 + Q U02 +Q The steady state approximation leads to eq. (2) for the quenching rate constant kq. kq = k,/ {1 + (k.-,/ k2)}{1 + k.2/(k3/ M} (2) Since no photochemical products of U022+ , such as U02+ or U(IV), were observed on prolonged irradiation of the solutions containing UC>22+ and the quenchers, the k3 path is expected to be much faster than the k4 path, i.e., I<3 » l<4. On this basis, eq. 2 can be transformed into eq. (3) with k3 ~ Z kq = M1 + (k.1/Z)Kexp( A G2*/RT) + exp( A G2/RT)} (3) - 268- where AG2 is the free energy change involved in the electron transfer process from quencher to excited uranyl ion, AG2* is the activation free energy for the kz path and Z is the frequency factor. The value of AG2 were estimated from the following equation derived from Rehm and Weller + A G2 = E(Q/Q +) - E(U02 /U022+) - AEo.oCUC^) -e02/aa (4) + + where E(Q/Q ) and E(U02 /U022+ ) are the redox potentials of quenchers and uranyl ion respectively, AEo,o('U022+) is the energy of the excited uranyl ion and eo2/s a is a term which takes into account the solvent properties for ion separation. AG2* can be calculated by the Rehem-Weller empirical equation A G2* = A G2/2 + {(A G2/2 )2 + AG2*(0)2}i/2 (5) where AG2*(0) is the intrinsic barrier to electron transfer, i.e. the activation free energy when AG2' = 0. Figure 2 shows a plot of kq versus AG2for various N- heteroaromatic compounds using eq. (3). Although Fig. 2 shows some scatter, the calculated curve almost coincides with the experimental data. According to the Rehm-Weller correlation, the plot of log kg versus AG2 is expected to give a linear line with a slope of-1/2.3RT(16.9 eV-1) in the sufficiently endergonic AG2 region. In the quenching of uranyl emission with various series of quenchers, the values of -log kg /AG2 are reported to be much smaller (0.2- 2 eV-1) than 16.9 eV-1. In our study , the value of-log kq/AG2 in N-heteroaromatic compounds is nearly equal to 16.9 eV-1 and this is a special case in uranyl photoreactions. These results can be reasonably explained by the large it-bonding ability of N-heteroaromatic compounds towards "IIO^ yielding the formation of relatively stable exciplex (or an encounter complex), and AG2*(0) is nearly equal to the value predicted by the Rehm-Weller correlation. Energy Transfer from the Excited Uranyl Ion to Dysprosium Ion In solutions containing U022+ and Dy3+ , 414 nm laser irradiation resulted in the excitation of only U022+, because Dy3+ has no absorption at this wavelength. Under the above irradiation, the emission intensity l(t) at 570 nm showed a dual-exponential-function relation with time t: -269- l(t) = l(01)exp(-t / x 1) + l(02)exp(-t / x 2) where 1(01) and l(02) refer to the emission intensities at t = 0 for two excited species and x 1 and x 2 are the lifetimes of these species. The t = 0 is defined as the end of the irradiation pulse. The first term in the above equation can be assigned to the 2+ direct emission of excited uranyl ion *U02 and the second term to the emission of excited dysprosium ion *Dy3+ as mentioned below. The lifetime of term, x 2 = (2.5 - 3.0) x 10-6 s, which is independent of the concentration of Dy3+, agrees well with reported value for the lifetime of *Dy3+ . In contrast, the lifetime of the first term, x 1, decreases with increasing concentration of dysprosium ion, [Dy3+], and this behavior correlates well with the quenching of uranyl emission. The rate constant 2+ for the quenching of *U02 by Dy3+ was determined to be kq = 2.01 x 106 moM dm3 s-1 from the Stern-Volmer plot. The sensitized emission of Dy3+ was not observed in fluorospectrophotometric measurements in D20 solutions, but was observed in the differential spectrum at 0.5 us by the flash photolysis method. We also determined a decay constant at 578 nm from the time-resolved differential emission spectra which was in good agreement with the reported value for *Dy3+. These results clearly indicate that energy transfer takes place from the excited uranyl ion to the dysprosium ion. The sensitized process can be described by 2+ 2+ *3U02 + 6Dy3+ -> HJ02 + *4Dy3+ Laser Irradiation Nuclear Magnetic Resonance Laser Irradiation Nuclear Magnetic Resonance has been developed in our research group, which enable us to record high resolution NMR spectra under the irradiation of a laser. The 170 NMR spectra of uranyl oxygen atoms (oxygen atoms 2+ of U02 ) were measure by this method.The irradiation system is shown in Fig. 3, where a laser beam passes through the inside of temperature controller, which regulates the ample temperature, and irradiates an NMR sample from the bottom of sample tube with a flux 4 mm in diameter. By using this technique, the resolution of NMR spectra is kept unchanged during laser irradiation and the intensity of the irradiating laser is very stable. A JEOL JNM-FX 100 NMR spectrometer and a Spectra Physics Series 200 argon-ion laser were used. In our earlier papers, the 170 NMR spectra of uranyl oxygens were so narrow that three signals due to the — 270- isotopomers, i.e. peo-U-170], [170-11-170], and [180-U-170], were observed. Figure 4 shows the 170 NMR spectra of uranyl oxygens under the irradiation of a 488 nm laser as a function of output power (W), which is proportional to the laser intensity as calibrated by a radiometer. Two peaks were observed and assigned from the lower field as signals of [160-U-170] and p8O-LM70], respectively. It can bee seen in this figure that as output power, or laser intensity increased the signals shifted to the lower field, and their linewidths became broader, becoming a single peak at the high output powers. McGlynm and Smith reported that the long-lived excited state was triplet either 3nu or 3AU (with a preference for the former), and that the anisotropy of the paramagnetic susceptibility in the 3nu state was maximal parallel to the O-U-0 axis. This is consistent with the above result that the paramagnetic result is remarkable in the axal oxygens and little effect is observed in equatorial 2+ 2+ water. Since *U02 is in rapid equilibrium with U02 within the time scale of NMR measurements, i.e. the lifetime of *U022+ in acidic aqueous solutions is ca 10-6 s-1, the observed paramagnetism in NMR spectra should be proportional to the concentration of *U022+. SAMPLE TUBE (D) 10 Hz power/W PRISM 9- Scheme of laser-irradiation system. The system consists of four main components: (A) laser, (B) optical fibre, (C) temperature 4 controller and (D) sample tube. 9- "O NMR spectra of uranyl oxygens as a function of laser power (W). — 271- Ill Photo-reduction of Uranyl Ion Photo-reduction of the Uranyl Ion with Inorganic Compounds Photochemical reactions of the uranyl(VI) ion with halogen and halogen-like anions, I-, Br-, CI- and NCS-, were investigated in aqueous phosphoric acid solutions under the irradiation of nitrogen laser. The formation of U(IV) was observed in the reactions with I-, Br-, and NCS-, but not with CI-- The yield of U(IV) increases in the order, Br- < NCS- < I-. This order is the same as that of kq(X-) and the reverse of the standard potentials of these anions. The results are consistent with the electron transfer mechanism between excited uranyl ion and these halogen and halogen-like anions in the primary process. On the other hand, in the photo- reaction between U(IV) and Fe(phen)32+, since Fe(phen)32+ has strong absorbance in the visible region, i.e. the molar extinction coefficient is larger than 104 moM dm3cm-i at 510 nm, the decreasing rate of Fe(phen)32+ can be measured while holding the U(IV) concentration constant under the condition [U(VI)]0 » [Fe(phen)32+]0i where [ ]o denotes the initial concentration. Figure 5 shows plots of -ln(At - Af) versus tirr. where At and Af refer to the absorbances at tjrr and infinity, respectively, in the oxygen free and oxygen saturated solutions. In Fig. 5, the plot shows a curvature in the oxygen free solution, while is linear in the oxygen saturated solution. It was concluded clearly that the role of U(V) was important for this reaction, because the oxidation of U(V) by solvated oxygen changed the rate equation. Photo-reduction of the Uranyl Ion with Alkenes Photochemical reactions of U(VI) with alkenes, 1-heptene, 1-hexene, 1- octene, cycloheptene, cyclohexene, 1,4-cyclohexadiene, 2,4-hexadiene and 1,3- cycloheadiene, were performed by irradiating with a high pressure mercury lamp through a filter cutting off wavelength shorter than 400 nm in perchloric acid solutions. It was found that the quantum yields for the formation of U(IV) was larger in alkene with smaller vertical ionization potentials. The formation of U(IV) was not observed in 1-heptene, 1-hexene and 1-octene, but U(IV) was produced if NaN03 was added to the same solutions. Because of n-bonding ability of alkenes, it is reasonable to presume that *U(VI) interacts with an alkene to form an encounter complex or exciplex ['UO^+.-Q], which is subject to intramolecular electron transfer — 272 — from the alkene to *U(VI) to form an unstable intermediate [IIC^-CK] as described for the reaction between *U(VI) and N-heteroaromatic compounds. The + intermediate [UO2 +-Q+] either decomposes to form U022 or proceeds to the k3 path; hence only k4/(ks + Icj) contribute to the formation of U(V). The addition of nitrate to the solutions results in the formation of U(IV) even though nitrate is a good oxidizing reagent. It was reported that nitrate reacted with the propylene cation radial as + CH2=CH-CH2 + NO3- -» CH2 = CH-CH2ON02 + Based on the above reaction, since NO3- only slightly quenches *U022 , NO3- + + stabilizes U02 by quenching [U02 -O] and O, and as a result the addition of NO3- would assist the formation of U(IV). IV Photo-oxidation of Uranium (IV) and Plutonium(lll) in Nitric Acid Solutions Oxidation of U(IV) in nitric acid solutions was studied in dark and under the irradiation of Xenon light. Experiments were carried out in terms of the conventional spectrophotometric method. Figure 6 illustrates a pseudo first-order plot in dark, a plot of (At - Af) versus time. As seen in Fig. 6, the plot is linear only in the initial stage and tends to curve resulting from the acceleration of reaction. This curvature is attributed to the additional reaction path of U(IV) with nitrous acid, which is formed by the reduction of nitrate. As a matter of fact, reaction became much faster, if HN02 was previously added to the solution. It was found that the irradiation of light above 420 nm results in no gain for the apparent oxidation rate of U(IV). This obviously indicates that U(IV) is not sensitive for photo-reaction and that only excitation of NO3- should be taken account of. The above reaction is basically similar to the photo-oxidation of plutonium(lll) performed by Wada and coworkers. The purpose of this study is to separate Plutonium from solutions containing neptunium by using photochemical reactions. Under the irradiation of Hg lamp, Pu(lll) was selectively oxidized to Pu(IV) without changing Np(V). Acknowledgement The studies (Ref. 1-12) reviewed in this paper were performed in cooperation -273- with Dr. Yoon Yul Park of Pusan Junior College, Dr. Yasuhisa Ikeda of Institute of Research and Innovation, Mr. Yukio Wada of Power Reactor and Nuclear Fuel Development Co., and Masayuki Harada, Tomoo Yamamura and Koh Hatakeyama of author's colleagues in Tokyo Institute of Technology. The author wishes to thank them for their helpful cooperations. I I i 1 1 1 1.2 1.0 / / O.B /o - 0.6 2 4 1 1 1 1 1 1 time/10 s 0 10 20 30 40 50 60 Fig. 6 Pscudo first order plot for the oxidation of t|rr/min U (rV) by NOj" in dark (O) and the same plot for CU(Vl)]o = 5.0x 10-' mol-dm-'; the solution, in which HNO? ( 0.03M ) was previ [Fe(phcn)i*]o=5.0x 10"> mol-dm"'. ously added, (• ) under the condition : [U(IV)] = + Fig. 5 Plots of —ln{At—Af) vs. iirr in 0.009 M, [ NaN03 ] = 0.3 M, and ( H ] = 0.5 M at oxygen free (O) and oxygen satu at ionic strength = 3.5 y. *'.d at 40 °C. rated (•) 2 mol-dm"' phosphoric acid solutions at 25"C References 1) Park, Y.Y., Sakai, Y., Abe, R., IshiiT., Harada, M., Kojima, T., and Tomiyasu, H.: J. Chem. Soc, Faraday Trans., 86 (1990). 2) Yamaguchi, T., Harada, M., Park, Y.Y., and Tomiyasu, H.: J. Chem. Soc, Faraday Commun., 1621 (1990). 3) Harada, M., Sakamaki, S., Yamaguchi, T., and Tomiyasu, H.: Ninon Kagakukai Shi, 951 (1992), 4) Mizuguchi, K„ ikeda, Y., Harada, M., Park, Y.Y., and Tomiyasu, H.: J. Nucl. Sci. Technol., 3JL 542 (1993). 5) Yamamura, T., Park, Y.Y., Tomiyasu, H.: J. Alloy Compounds, IBS, 186 (1993). 6) Yamamura, T„ Iwata, S., Iwamura, S., and Tomiyasu, H.: to be published . 7) Park, Y.Y., Harada, M., Ikeda, Y., and Tomiyasu, H.: J. Nucl. Sci. Techno!., 2& 418 (1991) 8) Park, Y.Y. and Tomiyasu, H.: J.Photochem. Photobiol., 74, 11 (1993). 9) Park, Y.Y., Ikeda, Y., and Tomiyasu, H. RECOD'91, Sendai, Vol.il. 814(1991). 10) Park, Y.Y. and Tomiyasu, H.: J.Photochem. Photobiol., 64, 25 (1992). 11) Hatakeyama, K., Park, Y.Y., Tomiyasu , H. Wada, Y., and Ikeda, Y.:Global-93, Seattle, 1325 (1993). 12) Wada, Y., Goibuchi, T., Morimoto, K., and Tomiyasu, H.: Actinide-93, Santa Fe, 14 (1993). 13) Wada, Y., Wada, K., Goibuchi, T., and Tomiyasu , H.: J. Nucl. Sci. Technol (1994), in press. -274- Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- LASER STIMULATED EXTRACTION OF Pd FROM SOLUTION WITH URANYL. M.Vlasov*,B.Krynetsky,A.Kukhtenko,S.Mironov*,A.Prokhorov, A.Zhidkov. Laboratory of Laser Spectroscopy.General Physics Institute. Russian Academy of Science. Vavilov str.,38. Box 117333. Moscow. Russia. * Russian National Center I.V.Kurchatov. Have been investigated process of the extraction of metal palladium from solution Td +HC104+(UO^ ) by radiation eximer XeCl-laser ( x=308 ran ). By optimal parameters of solution efficiency of extraction was about 100%. Have been discussed processes reduction of palladium by resonance laser action. Keywords : Laser irradiation, Palladium, Extraction redox process. Recent demand of platinum -group elements has incraesd according to the grouth of high-technology industries, because it is considered that future consumption of such could greatly expand. For this reason , these metals are considered strategic materials. Thus, there is considerable incentive to recover them from nuclear fuel as a new source of noble metals. Thus , for example, the concentration of Pt- group metals can reach the values 4-19 kg/tonn [11, that is hundreds times the contens as the natural are deposit. Some Investigations of the technical feasibility of using fission product noble metals from spent nuclear fuel have been done by several facilities. But the traditional chemical methods of the extraction of these , rather inert metals are -275 — ineffective and result in the unwanted increase of the volumes of radioactive waste. For this reason the methods of laser photochemistry seem [2] to be espesially attractive and their application probably will be ablle to provide the precipitate of noble metalls extraction from nuclear waste. In this work the possibility of stimulation by laser precipitating of palladiom from the complex solution containing the ions U0*+ and Pd2+ was investigated. It is knowfy that under the treatment of acidic solutions by light the ions of uranyl turn into U4,+ states following the redox reactions [3,4].Ion U4+, as strong reduction agent can reduce the Pd2+ ions./The Pd2+ is stable state of palladium in acid solutions/to the neutral metallic state. As the sourse of radiation have been used excimer laser / x = 308 nm/ with output energy 70-80 mJ/pulse and repetition rate 10 Hz. The concentration of Pd2 + ions before and after treatment have been determined spectrometrically according to the peak of Pd absorption at the wavelength x = 380 nm. The spectra have been measvred after the centrifuging of solutions. Fig 1 shows the changes of the transmision spectrum of the solution palladium and uranium in time under the treatment by XeCl-laser. Its obviously, that bleach of solutions occurs in time in the region of 380 mm, where the absorption is mainly stimulated by the presence of palladium ions. — 276- The changes of the phase state of palladium - the appearence of metallic palladium - were observed visual. Fig 2. show yield of metallic paladium in time k = 1 - C/Co where C and Co -the concentrations of palladium before and after irradiaion and illustrate the influence of the organic reducing agent (ethyl alcohol). The presence of reducing agent increases drastically the rate of the process and the yield of metallic palladium. The fig 3 shows the results of the changes of the yield of palladium from the initial concentration of uranyl. The reduction of palladium in solutions under laser radiation comes about as two paths. The first one is conditioned by photo- redox process.The second-by photo substitution process. [2] By radiation of the solution of uranium the excitation of the triplet state of the ion uranyl U*0* + (S=1) followed by the changes of its valency occurs in the reaction with the reducer R. U*02 + + R ->U0!+ If 2 2 2U0++ 4H+-> U02++ U4++ 2H 0 2 2 2 The generated uranium ions reduced palladium to the neutral state. [5] Pd2+ + U4+ + H 0 -> Pd°! + U02++4H+ 2 2 Under light it's possible the direct reduction of the uranic as well, for example. U02+ + C H OH + 2H% hw -> U4+ + CH CHO + 2H 0 2 2 5 3 3 2 -277 — U02 + + hu -> U*02+-> U*+ 2 2 The reduction of palladium can occurs in the lack of uranyl and is bound up with the presence of the reducing agent. 2 + Pd '4(H20) + 2R + hw -> Pd(ROH) -> Pd°! Test experiments carried out without uranyl in the solution showed, that the process of photo-substitution was similar by effective to the redox process and, in principal,can be used for produce of metallic palladium from the solution without uranyl. References : 1) Bush R.P.: J.Platinum Metalls Rev. (London),35, 202,(1991). 2) Donahue T. .-Chemical and Biochemical Appl .of Lasers (N/Y, London,Toronto,),5,232,(1980). 3) Bell J.T.,Buxton S.R. J.Inorg.Nucl.Chem. (London),36, 1575,(1974). 4) Gurevitch A.I. and other : Pis'ma v ZTF, (Leningrad),14, 2124,(1988). 5) Koltunov V.S. and oth. Radiochimiya, (Moscow),3,59,(1990) — 278- Fig.lThe change of the spectre 330 380 430 480 630 680 630 680 730 before Irradiation l-rnln irradiation 3-min irradlotfon Initial solullon :|Pd2+)-0.01M/MUo2*]-0.05M/L.(HC]CK J-1M/L. (C2H50HH.1M/L * Fifl.2The plot of the yeld of palladium in tlmewith (1) and without (2) reducing agent. K. , — uo • 2< min B»d ]-0.01M/L,[UQ^ I-0.01M/L, '' [HCIO4 ]-1M/LIC2H5 OH]-0.*t/L Fio.3 The plot of ihe yeld of Pd from concentration of uranyl. 1-under irradiation of Ar-laser(20 min) • 2-undor Irradiation by XeCI-laserd min) 0.0E*00 1.0E-03 2.0E-03 3.0E-03 4.0E-03 6.0E-O0 -279- Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- LIS: The View from Urenco K R SCHNEIDER Urenco Limited, 18 Oxford Road, Marlow, Bucks. SL7 2NB, UK In parallel with developing gas centrifuge technology, the Urenco partners have undertaken research into laser isotope separation. The paper reviews progress that has been made and compares the costs, risks and potential of the technologies. Keywords: Gas centrifuge, Laser Isotope Separation, Uranium Enrichment 1. INTRODUCTION The Urenco centrifuge process for the enrichment of uranium is now well developed and economically proven. The centrifuge technology which forms the basis of Urenco's plants in the UK, the Netherlands and Germany, has evolved steadily over the last two decades since the Urenco partnership was formed. Current programmes of development will see this progress continue into the next century. In parallel with its centrifuge development, the Urenco partners have undertaken research into laser isotope separation techniques from as early as the 1970's. The UK partner has concentrated on the atomic route, the German partner on the molecular route and the Netherlands partner has investigated specific aspects of both processes. Urenco is therefore in a unique position to contrast and compare the costs, risks and potential of laser and centrifuge enrichment technologies. This paper considers the progress that Urenco has made and the reasons for its recent policy decision to rundown further laser-based R&D. 2. ATOMIC ROUTE LIS In the UK, work on methods of enriching isotopes using laser techniques started in 1974 within the UK Atomic Energy Authority. Both the molecular and atomic routes were studied. In 1983, a decision was taken to concentrate on the aiomic route as offering the greater economic potential. In 1986, a collaborative agreement was entered into by British Nuclear Fuels pic, the UK partner in Urenco, and the UKAEA. Subsequently the Urenco partners agreed to widen the scope of their collaborative R & D to include laser work in all three countries as well as development of centrifuge technology. — 280- By 1987, clear isotope selective ionisation had been consistently observed in experiments at UKAEA Harwellc,). The experiments used a laser system consisting of copper vapour lasers and dye lasers to climb a 3 x 2 ev ladder of excitation. Metallic uranium was evaporated in a small e-beam device known as EBV1. In parallel with further detailed studies, aimed at developing a complete theoretical understanding of the physical phenomena™, it was decided to invest in new facilities in order to take forward evaluations of technological feasibility and economic attractiveness. One such facility is known as the Laser Plant Development Facility (LPDF). The aims of the LPDF(3) were to demonstrate that a plant-like laser array, consisting of copper vapour lasers (CVL) and dye lasers, could be operated routinely, reliably and safely whilst providing laser light of the required specification at an economical cost. The facility was designed, constructed and commissioned in the period 1986-1990. It comprises four CVL master oscillator power amplifier (MOPA) chains which are multiplexed and used to pump three dye laser MOPA chains. The complete system has been demonstrated to work and approaching 100,000 hours of CVL operation have been logged to date. Whilst there is considerable confidence that a laser system capable of matching the stringent requirements can be developed (putting economic considerations aside for the moment), it is generally recognised that the more problematical aspects of the atomic route lie in the technology required to evaporate metal uranium, to collect the U335 ions and to handle the resultant product and tails streams in quantities of some hundreds of tonnes per year. Over the period of 1986-1990, a facility known as EBV2 was designed, constructed and commissioned for the purpose of evaluating this technology. Since then, the vessel, which is equipped with a strip electron gun with a power of 250 kW has been used to: • test and develop methods of ion collection at densities typical of plant, • test and characterise extended evaporation of uranium metal and mixtures, • study materials, coatings and corrosion rates in reflux experiments. As a result of this work, there exists an effective method of ion collection. There is also an understanding of how the efficiency of uranium evaporation can be improved, and the problems of handling liquid uranium eased, by the use of evaporants other than pure uranium, albeit at the price of an increase in complexity. For a device to operate with the reliability which would be expected in a production plant, the issue of corrosion is very important. The — 281 — partner in Netherlands had undertaken extensive investigations of potential materials and coatings for use with liquid uranium and the more promising have been tested for extended periods in EBV2. The result of this work was reviewed and evaluated at the end of 1992. The principal conclusions were: • although in the course of the project there had been many surprises there was still no reason to doubt that atomic route LIS on a production scale was feasible, • there remained however a considerable amount of development still to be done, • a realisation of the technology based on a 3-photon, CVL/dye laser system was unlikely to be competitive with advanced centrifuge technology. It was however decided to continue the project for a further two years in order to evaluate the longer term prospect of a laser technology based on 2-photon excitation and solid-state lasers. This drew particularly on spectroscopy work at Almelo and laser development at the Nederiands Centrum Voor Laser Research (NCLR). The evaluation of this work is considered later in the paper. 3. MOLECULAR ROUTE LIS Molecular Laser Isotope Separation (MLIS) has been investigated in Germany since 1971, starting on a research institute level. Since 1975 activities have been concentrated at Uranit (now Urenco Deutschland) in Julich(4). The key question to be answered in developing an economic MLIS process is whether it is possible to obtain single step enrichment at reasonable UF6 densities and laser fluences. Broadly, the aim has to be to demonstrate such a single step process at UF6 densities which exceed those in an atomic route by two orders of magnitude and with IR laser pulse energies below 1 J. The major objectives of the programme have been : • to establish the dependence of the separation performance on UF6 properties and laser parameters, • to optimise the design of nozzles capable of yielding the desired UF6 properties at the highest possible density, • to develop high repetition rate 16 (.im laser systems, -282- In comparison with the atomic route, there are formidable difficulties in establishing a reliable quantum mechanical theory for selective excitation, and even semi-empirical modelling is hard to perform. However, in contrast to the atomic route the selectivity is clearly dominated by the photophysics of a single molecule, rather than by bulk effects (eg non-selective pick-up, ion scrambling). This favoured a strategy of using molecular beam experiments in order to screen the parameter space quickly and to establish sets of promising parameter combinations. These sets were then used in a second experiment under process conditions of UF6 density in which weighable amounts of enriched UF5 product were sampled and directly analysed. In the molecular beam apparatus which used a circular nozzle expansion ofa mixture of 235 carrier gas with UF6 enriched to 50% U, it was demonstrated that the requirements for a single step process could be achieved. Separation experiments were then performed at process densities in a system which included a 20 cm wide pulsed slit nozzle in a closed loop, a photolysis chamber and a UF3 separation section. A typical experiment lasted less than one hour at a 2 Hz repetition rate and resulted in the production of milligram amounts of solid enriched UF5. In parallel to the separation experiments, investigations of UFfi slit nozzle expansions showed that it was possible to obtain improved density and condensation factors compared with conventionally designed nozzles. The laser system requirements have been satisfied by the in-house development of a 16 um system, based on TEA CO, laser modules and on multipath Raman cells working at 77 K, which deliver line-tuned pulses in excess of 1 J/pulse at around 3-5 Hz. In a production plant the requirement would be a repetition rate of 4 several 10 Hz and therefore high repetition rate C02 lasers would need to be developed and multiplexed. In 1992, a TEA CO, module achieved over 1 kHz and several J/pulse, and a room temperature Raman converter was constructed for operation at kHz repetition rate. As with the atomic route, considerable effort has been put into evaluating the feasibility and economic merit of realising this technology on a production scale. The principal conclusions are that: • single step enrichment is feasible, • the laser system requirements are achievable, • a reference plant design would attain SWU costs at best slightly below advanced centrifuge technology. -283- However, there remains a considerable amount of development still to be done and there are significant uncertainties, which have a major impact on economic prospects, regarding : • demonstration of simultaneous enrichment and stripping at plant conditions, • laser system performance. In 1992, the decision was taken, by Uranit, to run-down the MLIS project. 4. ADVANCED CENTRIFUGE TECHNOLOGY Following earlier pilot plant operation, the first generation of production machines were installed in Urenco enrichment plants starting in the mid-1970's. They significantly exceeded the original design expectations; indeed one of the plants is still operating. Since these first plants, Urenco has achieved ongoing improvements in the economics of the centrifuge process through the application of the basic concepts of: • large scale production of highly reliable centrifuges requiring no maintenance over an operational life exceeding 10 years, • progressive improvements in the separative power of an individual machine. The no-maintenance concept, together with very low power consumption, results in low operating costs whilst the improvements to each new generation of centrifuges have reduced the specific investment costs. The principal parameters influencing the separative power of a gas centrifuge are the speed of rotation and the length of the rotor. These in turn are determined primarily by the specific strength and specific modulus respectively of the rotor material with a further important factor, the centrifuge lifetime, being influenced by the operating environment (high stresses, elevated temperature, UF6). By making use of both improvements in material properties and by the introduction of new materials, the speed has been increased by a factor of nearly two and the length by a factor of more than five. The fourth generation machine which is currently being installed in plant has an output approximately 25 times that of the early machines, A fifth generation machine, to be ready for introduction into the plants during the second half of the 1990's, is being developed at present. It is this machine which forms the basis for comparing advanced centrifuge technology with laser enrichment technology. Experience over the past twenty years gives a basis for the confidence that the development targets for -284- the new machine will be achieved both economically and on time. Figure 1 illustrates that technical improvements have delivered increased outputs and reduced investment costs which have been the cornerstones of Urenco's competitiveness over the past two decades. In addition, machines have significantly exceeded the expectations on both failure rates and operating lifetime. 5. COMPARISON OF TECHNOLOGIES As has been pointed out elsewhere (5), the business reasons for introducing a new technology into a competitive market normally include : • an improvement in the product, • a response to market demand, • an improvement in competitive advantage. The enrichment market is an integral part of a mature fuel cycle based on acceptance and delivery of uranium hexafluoride. The product specifications are long established and firmly linked to qualification experience of the fuel fabricators. In comparing the technologies, it is Urenco's view that, for the foreseeable future, the enricher will have to comply with the existing fuel cycle and would have to bear the burden of any intermediate steps. If improvement in the final product is not a reason for introducing new technology, neither is an expectation of increased market demand; indeed, the market is likely to suffer from overcapacity well into the next decade. The primary motivation must therefore be that the new technology brings increased competitive advantage, which is either reflected in an increased market share because of lower prices or in better margins. The question for Urenco is then, in principal, quite simple: will either an atomic route or a molecular route laser process be achievable at a lower cost than the advanced centrifuge ? One of the problems in comparing the three technologies considered in this paper, is that they have very different cost structures. Figure 2 shows normalised estimates of capital and operating costs for plants of similar capacity. One consequence of these different cost structures is that financing assumptions (eg interest rates, amortisation) can become dominant in the evaluation. Another difficulty is that the basis for optimisation is different for the single step laser processes and the cascaded centrifuge process. For example, an atomic route SWU cost is strongly dependent on tails assay whereas, to first order, centrifuge SWU costs -285- are independent of tails assay (the assay being an issue of cascade configuration not machine performance). Therefore, although Figure 3 shows normalised estimates of SWU cost, it should be treated with some caution. In our evaluations we also consider UO, product cost for a range of feed costs. Even putting aside the above reservations, cost evaluations of reference plants are not by themselves a reliable basis for decision making. Other factors that have to be taken into account are: • Risk; The uncertainties in evaluation are much greater for both laser processes than for the next generation of advanced centrifuge (which will enter its qualification phase by the end of this year). For the atomic route these lie principally in reliability/lifetime issues associated with the separator vessel whereas for the molecular route they lie in the combination of enrichment and stripping at process conditions. Achievement of projected costs, lifetime and reliability are issues for both laser systems, particularly for somewhat speculative all solid-state technology assumed for the atomic route. • Capacity Planning; One of the features of Urenco's business is the match that has been maintained between installed capacity and sales; unlike its competitors. That additional capacity can be added in response to demand is proven for centrifuge plants. The financial penalties resulting from mismatch are considerable; Urenco's view is that issues of scale, modularity and timing favour centrifuge technology. • Development Status and Potential; Both laser processes would require several years of further development before even pilot-plant implementation and there is a high level of uncertainty about the associated R&D costs. Following a well proven route for development of a new machine, Urenco expects to be installing the fifth generation machine in plant in the late 1990's. Before the end of that programme a decision will be taken on whether to develop a sixth generation machine on the basis of well- understood costs. • The Fuel Cycle; In addition to consideration of the impact of any changes to a uranium hexafluoride based fuel cycle, it should be recognised that recycled uranium, with varying degrees of U,35 content, will become increasingly available in the next decade as major reprocessing plants come on — 286- line. It has been suggested that a laser process would be advantageous for such material. It is beyond the scope of this paper to discuss the issues arising, but Urenco has a proven ability to enrich recycled uranium arising from both gas-cooled and water-cooled reactors; indeed, since 1992, about 400tU has been processed for European and Japanese utilities in normal production operations. 6. CONCLUSIONS The technical results from both the atomic route and molecular route laser programmes have been encouraging and in fact many technical targets have been exceeded. However, in both cases, considerable importance was also attached to economic evaluation, based on a reference case design, in comparison with Urenco's future advanced gas centrifuges. The results from these studies have led Urenco to the conclusion that it will be very difficult for any commercial enrichment plant employing either laser process to be competitive with centrifuge technology. The comparison was with the next generation of Urenco centrifuge for which the late 1990's is a firm introduction date in contrast to the uncertain status of the laser processes. In 1992 a decision was made to stop further work on the molecular route and, recently, a corresponding decision has been made to run-down work on the atomic route. References (1)C Whitehead; Nuclear Technology International 1988 (ed. N Geary) (2) P T Greenland; Contemporary Physics (1990), vol 31, No. 6 (3) G J Dickinson; The Nuclear Engineer (1992), vol 33, No. 6 (4) G Meyer-Kretschmer; G Schweizer; International Symposium on Isotope Separation and Chemical Exchange Uranium Enrichment, Tokyo, 1990 (5) M H Schwartz, J J Steyn; The US Council for Energy Awareness International Enrichment Conference (1991) — 287- Figure 1 Centrifuge Technology Development Normalised Specific Investment Cost Normalised Separative Power 100 -, 30 80 - Qo I 20- 60 - 'SIC •S.P.' 40 10- 20 - 0 -l 1 1 1 r- '—I 1 1 1 r~ 1st 2nd 3rd 4th 5th 1st 2nd 3rd 4lh 5th Centrifuge Generation Centrifuge Generation Bmaa Figure 2 Capital and Operating Costs Laser-AR Laser-MR Centrifuge Laser-AR Laser-MR Centrifuge ICapital Cost lOperating Cost Figure 3 Comparative Cost Evaluation Laser - AR Laser-MR Centrifuge iRelativ'eSWUCost — 289 — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- NEW JAPANESE AVLIS R&D PROGRAM Y. SATO Laser Atomic Separation Engineering Research Association of Japan (LASER-J) 2-76, Aza-shirane, Oaza-shirakata, Tokai-mura, Ibaraki, Japan In LASER-J, the second phase of AVLIS R&D program has started since April 1993. At the first phase, we developed AVLIS plant components such as CVL, DL, and electron beam guns for a facility having about a one-ton SWU per year capacity and obtained engineering data related to this facility. During the second phase, the next five years or so, we will develop full scale AVLIS hardware. In this session, the R S D results of the first phase and the outline of the R&D program for the second phase will be presented. Keywords: CVL, DL, MOPA, DL beam quality, lifetime, DL wavelength stability 1. INTRODUCTION LASER-J completed the first phase of the AVLIS (Atomic Vapor Laser Isotope Separation) R&D program at the end of June 1992. During this phase, plant components were developed that included CVL and DL systems, and electron beam guns for a facility having a capacity of about one ton SWU per year. Now, the first fiscal year of the second R&D phase is nearing completion. In this session, I will present the results of the first phase and an outline the R & D program for the second phase, mainly addressing the laser systems. Before discussing these items, I will describe our mission for persons that may not be familiar with LASER-J. — 290- LASER-J, an acronym for Laser Atomic Separation Engineering Research Association of Japan, was founded in 1987. LASER-J is sponsored by twelve organizations, nine of which are electric utilities. The other sponsors are the Japan Atomic Power Industry, the Japan Nuclear Fuel Limited, and the Central Research Institute of Electric Power Industry. The main objectives of LASER-J's first R&D phase are: (1)to develop engineering-scale components that can be used in a prototype plant, (2) to construct the prototype plant, and (3)to conduct enrichment tests at the prototype plant. The Japan Atomic Energy Research Institute (JAERI), which has been developing the AVLIS process since 197S, is in charge of obtaining fundamental data regarding the AVLIS process and is working in close cooperation with LASER-J. I would like to mention a difference between LASER-J and most research organizations in other countries. Most, like LLNL (Lawrence Livermore National Laboratory) of the U.S.A. and the CEA (Commissariat a I'Energie Atomique) of France, are government organizations. LASER-J is a private organization. It is authorized under the Mining and Industry Technology Engineering Research Association Law by MITI (Ministry of International Trade and Industry) and the STA (Science and Technology Agency). The R&D performed has been done in close cooperation with contractors that are major Japanese industrial companies like Hitachi, Toshiba, Mitsubishi Heavy Industry and Mitsubishi Electric Co. This cooperation will allow the technologies developed to be easily transferred to industry. When LASER-J was founded, components needed for the AVLIS system that were available in Japan were limited to small-scale components. The performances of these devices were inadequate for use in an engineering-scale test. It was necessary to develop the major hardware components, such as lasers and electron beam guns, and to obtain firsthand experience related to the AVLIS system before beginning the design and construction of the prototype — 291- plant. Therefore, a preparatory test, which used available small-scale lasers and electron beam guns and so on, was performed in parallel with the development of engineering-scale components. Some enrichment tests and uranium recirculation tests were conducted near the end of the first phase. The results of the first R&D phase were reviewed and evaluated by our sponsors and the Japanese AEC (Atomic Energy Commission). The Japanese AEC issued an evaluation report concerning our work in August 1992. Based on this report and their own judgment, our sponsors directed LASER-J to proceed with the second R&D phase. This phase began in April 1993 and will continue for the next five years or so. Full-scale AVLIS hardware will be developed during this phase. Now, I will discuss the main topics. 2. DEVELOPMENT OF LASER SYSTEMS (1)CVL DEVELOPMENT A CVL is usually operated in a chain called a MOPA (Master Oscillator and Power Amplifier), which consists of an oscillator and several amplifiers. Large output power is a main requirement for AVLIS CVL systems. Figure 1 shows the CVL power level growth achieved by LASER-J. During the first phase, one important objective for the CVL development was to increase the output power. Generally, the output power of a CVL is proportional to its gain volume, which may be inferred from Fig. 2. Thus, our contractors have tried to increase the power by increasing the ceramic tube's diameter and length. In October 1992, we attained a maximum output power of 430 W (average 415 W ± 3% for three hours). And its value has already reached -292 — up to 485 W in December 1993. From our economical evaluation, we think about 500 W will be sufficient for a 1500 tSWU/y scale plant. Therefore, from the aspect of the output power, we hope we can produce a commercial-scale CVL- MOPA. It is very important that a CVL-MOPA in a commercial AVLIS plant have a long lifetime. At the first phase , we cumulatively operated CVL-MOPA for maximum 2300 hours, and at these operations, the lifetime of a thyratron, a switching device, and ceramic tube used for CVL-MOPA were confirmed. Based on these results, now on our CVL-MOPA, the limiting factor for the lifetime of a CVL is the high frequency switching device's lifetime. We have ever used a thyratron as a switching device, but a more reliable higher frequency switching device will be needed on a commercial plant. So, we have to develop a new type one. Fortunately, suitable semiconductor type switches offer a high level of reliability and a long lifetime. However, it is necessary for us to make further improvements. We are now studying how to extend the lifetime and increase the reliability of these devices. During the second R&D phase, it is also most important to decrease the loss of copper in the CVL during continuous operation and we are working to minimize this loss. There are other improvements needed. These include the need to improve the beam quality, efficiency, and stability. We hope we can make these improvements during the second phase. (2) DL DEVELOPMENT A DL is usually operated in a chain, the same as a CVL. A DL chain is also composed of an oscillator and several amplifiers. The main requirements for an AVLIS DL system are a large output power, high efficiency, a high beam quality, -293 — wavelength stability, and so on. When LASER-J started, the output power of DL-MOPA was only about 5 W. Generally DL's output power depends on CVL's one. As Fig.1 shows, DL- MOPA's output power had been raising up together with the increase of CVL- MOPA's one. At the end of the first phase R & D, we attained its output power of about 120 W in LASER-J's Tokai Test Facility, but this power is insufficient for a commercial-scale DL system, and must be increased. But for increasing DL power, there are other improvements needed also, such as how to prevent the thermal lens effect, and these also must be made during the second R&D phase. In DL development, the most different issue between the first phase and the second phase is the improvement level of DL beam quality. It seems that DL beam quality extremely contributes to an economical performance of AVLIS plant, because its improvement increases an amount of photons which are used for ionizing U235 atoms. Now, we think a rectangular and homogeneous DL beam is necessary for AVLIS processes, considering with a shape of ion extraction plates. However, for keeping DL beam rectangular and homogeneous, some new technology is needed, such as the image relaying for a rectangular beam and the hardware called an adaptive control mirror which compensates a distortional beam. Therefore, they are most important themes, and we are planning to study these items in our second R&D program. In a commercial AVLIS plant, a good DL beam wavelength stability is also required. We have attained its stability of ± 280 MHz for 170 hours in the first phase, but this performance has to be made much better, we think. It is considered that the limiting factor relating to DL wavelength stability are follows : (1)to let a lot of dye solution flow through a dye cell smoothly (2)to irradiate a dye cell with a stable high quality CVL beam (3)to control the length of resonators for compensating a wavelength. However, it seems that these problems are able to be solved by applying existing technology, and we will - 294- surely clear them in the second phase. In addition, we have other challenges, such as the improvement of efficiency and the decrease of ASE (Amplified Spontaneous Emission). These are also main themes in the second R&D phase. 3. THE SECOND PHASE OF THE AVLIS R&D PROGRAM As described in the introduction, in mid-1992 our sponsors directed us to proceed with the second R&D phase, which began in April 1993. Figure 3 shows the schedule of the program. Our main purpose at the second R&D phase is to improve the performance of all components to that needed for a commercial scale facility. As Fig.3 shows, during the first three years, the evaluation tests for verifying new technology described at the second paragraph in this session will mainly be done. Following those items, during the last three years or so, CVL-MOPA, DL-MOPA and an adaptive control mirror will be improved by the results of those tests. And then, those combination tests will be performed to confirm how our targets will be attained. — 295- 1000 100 o o > Fig. 1 CVL & DL Power Level Growth in L AS E R-J 3UU i <- PLANT-SCALE CVL /t / t / t / t 400 3---B ( F i g. 2 The Relation between CVL Power and Laser Tube Length -296- FY 19 93 1994 19 9 5 19 9 6 19 9 7 PRO DUCTION TEST CVL UNIT ^••<^.&-,v..----:, *•%;*.>..-..s*^.-- „. MAI(UFACTUR E ft TEST M 0 P A PRO DUCTION TEST DL OS C ,:,-,.-a>^y ,---,. PREPARATORY PRO TEST DUCT ION TEST AMP .' ;..-•• :•:.•:. • MANUF/ ICTURE TEST MO PA •^V^A,^*....-.. UNIT ADAPT I VE TE !; T MANUFACTURE CONTROL MIRROR .*:;•*•,*••'...• •.,•<•.'•; :'T —i COMB[NATION N > TEST ... vi'rs-..-; .•,'.;.!».--*•'••• Fig. 3 The Second Phase R & D Schedule — 297— Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- RESEARCH INITIATIVES TOWARD THE NEW HORIZON OF MOLECULAR LASER ISOTOPE SEPARATION AT RKEN Kazuo TAKEUCHI, Yoshikazu KUGA, Jun ONOE and Hideo TASHIRO The Institute of Physical and Chemical Research (RIKEN) Wako-shi, Saitama 351-01, Japan The R and D schedule of MLIS studies at Riken is briefly discussed. Ion-induced nucleation is found to be successful to enhance the growth of UF5 particles. keywords: Molecular laser isotope separation, Uranium hexafluoride, Uranium pentafluoride, Ion-induced nucleation, Ultrafine particle 1. INTRODUCTION In the molecular laser isotope separation process developed by RIKEN, gaseous UFg is cooled in a supersonic Laval nozzle to very low temperatures and is dissociated to UF5 and a F atom by multifrequency P-H2 Raman lasers at 16u.m. Since the beginning of the program of molecular laser isotope separation of uranium at Riken in 1985, our process has been utilizing solely the multiple photon dissociation (IRMPD) of UF6 cooled below 100K in a supersonic reactor, while the groups both in Los Alamos National Laboratory in the USA and Uranit in Germany adopted the method of selective IR excitation followed by UV dissociation at first, and later on switched to IRMPD. We have been quite aware of the advantages in IRMPD of UFg. Separation factors measured in our laboratory were found sufficiently high for the production of 3%-enriched uranium-235 at one stage. In 1985, RIKEN and Power Reactor and Nuclear Fuel Development Cooperation (PNC) officially agreed to transfer RIKEN's laser and process related technologies to PNC. In 1992, the Atomic Energy Commission of Japan — 298 — reviewed the progress and the future prospects of new uranium enrichment technologies including AVLIS, MLIS based on IRMPD and the chemical exchange process, and consequently authorized Riken and PNC to proceed to the new R and D stages of MLIS as shown in Fig.l. Under the renewed research cooperation agreement made, PNC takes the part of engineering development of Riken's version of MLIS, while Riken attempts to verify several new ideas to renovate the MLIS process. The outcome of these attempts will be reviewed by the commission. One of these is to investigate the crystallographic characteristics of ultrafine photoproduct particles of UF5 and enhance their growth by means of ion-induced nucleation. Such enhancement is expected eventually to reduce the size of the particle collector, to reduce the pumping duty and to result in the reduction of the enrichment cost. Fig.1 Research and Development for Molecular Laser-Isotope Separation of Uranium 1985 ~ 1987 1988 1992 1993 1997 1998 Development of High-Power, High-Repetition Laser Breakthrough Technologies (several hundred Hz) Laser (RIKEN) (RIKENt 1 1 Overall Verifiealion Engineering Verification (Data Base Reaction Process Breakthrough *> for Reaction Optimization) (RIKEN) Technolonies (RIKEN) Scientific Verification \ \ { 1 1 (RIKEN) Supply Verification of 100Hz Uranium ) Recovery (PNC Verification of 100Hz Uranium Enrichment System (II) Verification Enrichment System (I) IPNO (PNC) 2. CHARACTERIZATION OF UF5 ULTRAFINE PARTICLES The X-ray diffraction (XRD) pattern of the product (-10 Lim in diameter) was found to coincide with the authentic data of P-UF51) as shown in Fig.2. When the particle size is reduced to 10 ~ 20 nm, the XRD pattern still indicates that the crystal structure remains as P but the diffraction angles were systematically shifted to the lower side as shown in Fig.3. It must be noted that the deviations are much greater than the experimental error expected (~0.05°) in this XRD measurement. — 299 — The obtained lattice constants of the ultrafine UF5 particles in the a and b axes are compared in Table 1 with those of bulk. The results indicate that the ultrafine UF5 particles exhibit expansion by 2% in the a axis and by 5% in the b axis. It is worthwhile to note that the ultrafine particles of compounds show such expansion, since the data on lattice size changes of ultrafine particles in general are limited and since those for chemical compounds are even more scarce2'12) as shown in Table 2. (a) experimental =• c t: a w [a r A >h 1.1.. Lk...,. 1 rt (b) simulation Intensit y k • •M. h ... j,. ,. Table 1 Observed lattice constants of UFs a b P-UF5bulk 11.473 X" 5.208 A \ p-UFsUFP* 11.746 A 5.470 A expansion 2% 5% *)UFP:Ultraline Particles . **) experimental error: ± 0.007 A Table 2 Lattice constants of ultraline particles compared to bulk materials Bulk >• UFP Lattice constant Contraction" Expansion* Metal Ni, Cu, Co, Hg, Pb, Ca, Fe, Na, K Mg Compound AIF3 UFs •See for example, K. Kimura, Phase Transitions 24-26,493 (1990). C. Satoko, Microclusters (S. Sugano cds.), Springer, NY, p.142 (1986). -300 — 3. ENHANCEMENT OF UF5 PARTICLE GROWTH In order to test the feasibility of ion-induced nucleation, the following experiment was attempted. In a static cell containing the gas mixture of UF6, CH4 and Ar at room temperature, an cc-radiation source was placed. The mixture was irradiated by one shot from a KrF laser at 248nm through a CaF2 window attached to the cell, while the UF6 concentration was monitored with an IR spectrophotometer using the windows placed normal to the laser beam. After a sufficiently long time (-one day) for the particle growth to terminate, the moisture-sensitive UF5 ultrafine particles were taken out of the cell with care, and then observed via transmission electron microscopy. The results are show in Fig.4, which indicates that the primary particles in the case with the a- source are much larger in size than those without an a-source. Although the growth mechanism has to be studied further by theoretical and experimental approaches, the enhancement of the particle growth was clearly demonstrated. The processes of ion-induced nucleation may be as follows. Immediately after the laser pulse, a small fraction (~10-4 or less) of UF5 monomers is ionized due to either direct or collision-induced ionization by the a-source. The ionized condensation nuclei, which are presumably more stable than the neutral ones, attract partially polarized UF5 monomers by Coulomb force more efficiently than neutral condensation nuclei do. All of these processes may contribute to yield large primary particles. The characteristics of ion-induced particle growth are as follows. Since the fraction of ionized UF5 is small, the isotopic selectivity obtained in the nascent photoproduct (UF5) can _<_>_ (b) without a source 100 nm (a) with a source 100 nm Fig.4 Typical transmission electron micrographs of UFs particles - 301- hardly be deteriorated. The enhanced particle size reduces the duty of the particle collector, minimizes the size of the process compressor and eventually reduces the cost of MLIS. 4. CONCLUDING REMARKS We have shown the progress in the study of ion-induced particle growth of UF5. Another effort is to be made to investigate the negative role of cluster formation among UFg and rare gas molecules in the nozzle. Such a study will provide new viewpoints for better laser excitation schemes, nozzle design and cluster controlling methods leading toward higher separation factor, higher operating pressure and improved feasibility. Aside from the studies in process chemistry, some attempts to renovate the concepts of MLIS lasers have been tested in our group simultaneously. The outcomes of such research initiatives will be evaluated by the Atomic Energy Commission around the year 2000. LITERATURE CITED 1) J. Onoe, N. Uehara, Y. Iimura, T. Oyama, O. Suto, Y. Shimazaki and K. Takeuchi, J. Nucl. Mater., 207,205(1993). 2) N. Sato and K. Kimura, /. Am. Chem. Soc. iU, 4688(1990). 3) G. Apai, J.F. Hamilton, J. Stohr and A. Thompson, Phys. Rev. Lett. 43, 165(1979). 4) C. Brechignac, M. Broyer, Ph. Cahuzac, G. Delacretaz, P. Labastie, J.P. Wolf and L. Woste, Phys. Rev. Lett. 60, 275(1988). 5) P.A. Montano, J. Zhao, M. Ramanathan, G.K. Shenoy and W. Schulze, Z. Phys. D.V2, 103(1989). 6) K. Ohshima, S. Yatsuya and J. Harada, J. Phys. Soc. Jpn. 50,3071 (1981). 7) K. Ohshima and J. Harada, /. Phys. C L7, 1607(1984). 8) J. Harada and K. Ohshima, Surf. Sci. 106, 51(1981). 9) K. Ohshima, T. Yoshiyama and J. Harada, J. Phys. C18, 3073(1985). 10) K. Kimura, Phase Transitions 24-26, 493(1990). -302- 1 1) B. Delley, D.E. Ellis, A.J. Freeman, E.J. Baerends and D. Post, Phys. Rev. BTL, 2132(1983). 12) C. Satoko, in: Microclusters, vol. 4, eds. S. Sugano et al., Springer Ser. Mater. Sd.(Springer, New York, 1986)p.l42. -303 — Proceedings of the 6th Iniernaiional Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- CURRENT STATUS OF MLIS PROGRAM IN PNC Osamu SUTO, Hiromi YAMAGUCHI, Kiyoshi TASHIRO, Shigcaki KAWAKAMI and Yoshihiro SHIMAZAKI Nuclear Fuel Technology Development Division Power Reactor and Nuclear Fuel Development Corporation Tokai-mura, Naka-gun, Ibaraki-kcn, 319-11 Japan According to the Atomic Energy Long-Term Program revised by the Japan Atomic Commission in 1987, PNC has started R&D on MLIS since 1988. The main target of our program is to attain a high enrichment factor under engineering conditions that has been already achieved under experimental conditions by the Institute of Physical and Chemical Research. The test facility for enrichment tests under engineering conditions which consists of 100Hz laser systems at 16|Xm and a UF6 handling system has been developed and several tens of tests have been successfully carried out. Keywords: Molecular Laser Isotope Separation, UF6, Enrichment factor, Laser, 16um 1. INTRODUCTION Some uranium enrichment methods using lasers, which arc more efficient than conventional commercialized techniques, have been actively investigated during the past 20 years in several countries. Enrichment costs using lasers could be lower than those of conventional methods because these methods can enrich natural uranium to nuclear reactor grade in a single enrichment stage. One of these methods, the molecular laser isotope separation process (MLIS) is based on the isotope-selective photodissociation of gaseous UF6 to solid UFs, which can be removed from the gas flow by separators. The MLIS process can be divided into three main areas: — 304- • Cooling of UF6 molecules mixed with carrier gases in supersonic nozzle expansion. • Selective excitation and dissociation of UF6 by laser radiation at 16 )J.m. 0 Polymerization of UF5 molecules to UF5 particles and removal of UFs fine particles from the UF6 gas. From the standpoint of commercialization, ML1S has the great advantage that the feed material, UF6, is used in the current nuclear fuel cycle. Thus, wc have been developing this enrichment method. In Japan, the Institute of Physical and Chemical Research (ICPR) initiated the basic research on this process in 1982. According to the Atomic Energy Long-Term Program revised by the Japan Atomic Energy Commission (JAEC) in 1987, the Power Reactor and Nuclear Fuel Development Corporation (PNC) has been involved in the R&D in cooperation with IPCR and several industrial partners since 1988. This R&D includes the design and construction of a test facility for enrichment tests under engineering conditions as well as the basic technology development for the scientific and technical understanding of this process. This facility that consists of laser systems at 16ji.m and a UF6 handling system was completed at PNC Tokai-Works in 1990. The enrichment test was started in 1991 and recently, high isotopic selectivity has been demonstrated. In this presentation, the current status of PNC's MLIS program will be discussed. 2. DEVELOPMENT OF SUPERSONIC NOZZLES Wc developed a supersonic nozzle that can cool the UF6 gas to low temperature (less than 100K) during supersonic expansion without condensation. Moreover, the UF6 concentration in the supercooling area has to be high enough for commercialization. However, wc have to optimize a number of parameters such as nozzle shape, carrier gas, mixing ratio of UF6 and carrier gas and inlet pressure for the development of supersonic nozzles. Thus, wc first established rapid and reliable diagnostics methods of the characteristics of supersonic nozzles that arc based on high resolution infrared spectroscopy in the V3 Q-branch region of UF6 using tunable diode lasers. The UF6 rotational temperature is estimated by comparing the shape of the observed absorption spectrum of the V3 Q-branch (Fig.l) to the calculated spectrum (Fig.2). The UF6 vibrational temperature is estimated by comparing the absorption ratio of the hot Q-branch and the fundamental Q-branch (Fig.3) to the calculated value based on MaxwcII- Boltzman statistics. — 305 — The UF6 condensation is estimated by the saturation of the V3 Q-branch absorption as well as the light scattering method using an Argon-Ion laser. We already succeeded in the development of the laval nozzle which can cool UF6 to less than 60K without condensation (Fig.4) and further, we have been improving the characteristics of the nozzle by using these diagnostics methods. 3. ENRICHMENT EXPERIMENT IN SUPERSONIC FREE JETS OF UFe WITH A MPI/TOFMS Optimization of laser radiation parameters is extremely important in order to achieve a high enrichment factor. However, there are many parameters which include number of wavelengths, wavelength, bandwidth, frequency stability, delays, pulse shape and polarization. Besides there is no model which exactly simulates the UF6 multiphoton excitation exactly. One has to optimize these parameters by making a large number of enrichment experiments. However, it is very inefficient to investigate these optimum values by using the test facility because of its large consumption time. We have established an efficient experimental method that is based on enrichment experiments in supersonic free jets followed by multiphoton ionization timc-of -flight mass spectroscopy (MPI/TOFMS) of photodissociatcd UF5. As shown in Fig.5 and Fig.6, 16|Xm laser pulses are focused into a pulsed free jet of UFe diluted with carrier gas, and UF6 is photodissociatcd to UFs. UFs is selectively multiphotoionized at the second harmonic frequency (A,=532nm) from a Nd:YAG laser into various ions, mainly U2+. These ions arc accelerated in the electric field and reach the detector, multichanncl-photomultiplicr (MCP). For each radiation, the TOFMS spectra, the radiation energies and the delays arc recorded using digital storage oscilloscopes and then stored in a computer. Since the enrichment factor can be estimated by averaging 1000 individual spectra, wc can acquire the experimental results in a short time. The promising combinations of the laser radiation parameters resulting from these experiments arc tried in the enrichment tests using the test facility and the cut, 6, as well as enrichment factor, a, are then estimated. Although these experiments were just started, we have aircady achieved an enrichment factor high enough to produce the nuclear fuel. Hereafter, wc will actively carry out the enrichment experiments in supersonic free jets to optimize the laser radiation parameters. 4. ENRICHMENT TESTS USING THE TEST FACILITY The test facility for enrichment tests under engineering conditions, whiclt consists of laser — 306- systems at 16|im (Fig.7) and a UF6 handling system (Fig.8), was completed at the PNC Tokai-Works in 1990. The laser systems comprise two line-tunable TEA CO2 laser chains, one continuous- tunable TEMA CO2 laser chain and two multiple pass cells filled with para-hydrogen gas to convert the CO2 laser radiation to 16 |J.m by Raman scattering. The typical linc-tuncd 16 Jim pulse energy is about 600mJ and can be delivered stably. A weak TEMA CO2 laser pulse is also converted by Raman scattering with four-wave mixing and its continuously tuned 16(im pulse energy is about 40mJ. The handling system is a closed loop apparatus mainly comprising four sets of three roots pumps, supersonic laval nozzles (a 10 cm wide continuous flow nozzle and a 25 cm wide pulsed flow nozzle with piezoelectric drivers (Fig.9)) and two sets of UFs separators. Its throughputs arc several kilograms of UF6 per hour. The enrichment test was started in 1991. First a scries of tests have been done using two linc-tuncd 16jJ.m laser beams and several tens of milligrams of enriched UFs powder has been recovered by the low pressure multi-nozzle impactor (Fig.10). Recently, wc started the second stage enrichment tests using the continuously tunable TEMA CO2 laser chain as well as two linc-tunablc TEA CO2 laser chains. Even though only several tests have been carried out, wc have already confirmed an increase in selectivity. 5. FUTURE WORK The evaluation of the uranium enrichment technologies, MLIS as well as AVLIS (Atomic Vapor Laser Isotope Separation) and GCP (Gas Centrifuge Process), is to be made around the year 2000 by JAEC. Wc believe that the technologies and the description of the test facility presented in this paper enable us to optimize various parameters and achieve a high enough (X and 6 for commercialization till the C&R. — 307- § .O < Wave number -* 238 Fig. 1 Observed absorption spectrum of the v, Q-branch of UFfi. Wave number -* 238 Fig. 2 Simulated absorption spectrum of the v 3 Q-branch of UF6 v6 Q — branch I vz + v6 Wave number -• Fig. 3 Diode laser absorption scan in the v 3 Q-branch region. - 308 — wavenumder -* 0.1cm1 Fig. 4 Diode laser absorption scan in the v 3 Q-branch region of UF6 in the nozzle expansion. Reflection Fig. 5 Schematic diagram of enrichment experiments in supersonic free jets of UF6 with a MPI/TOFMS. -309- MPC MPI of UFs signal . MPD of UFe qi—tF d]—S=J W B n B pulsed free jet reflectron t0 PumP to pump Fig. 6 Schematic diagram of MPI/TOFMS. Fig. 7 16u.m laser systems. 310- l tinriH K\a * •*• • < mil, Fig. 8 UF6 handling system. Fig. 9 Supersonic pulsed flow nozzle with a 25cm wide slit. — 311- Fig. 10 Enriched UF5 powder recovered by the low pressure multi-nozzle impactor. — 312 — Session VI Panel Discussion — Future of Laser Science and Nuclear Technology — Proceedings of the 6th International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY - PANEL DISCUSSION Future of Laser Science and Nuclear Technology C.YAMANAKA: The title of the panel discussion is the Future of Laser Science and Nuclear Technology. In this session the four panel members will each give a presentation. After the panel members have finished their presentations, the members will be seated at the table and begin discussion with the members of the audience. Now, I would like to introduce myself and speak about the scope of this panel meeting. My name is Yamanaka, Director of Institute for Laser Technology and President of Himeji Institute of Technology. I have been working for laser fusion for 20 years to develop the Gekko XII laser and to promote the laser interaction with nuclear engineering. Today's panel members are Dr. Takuma of the University of Electro-Communication, Dr. Arisawa of JAERI, Dr. Letokhov of the Institute of Spectroscopy in Moscow, and Dr. Ewing of the Aculight Company. As you know, the laser technology has these four kinds of special features: one is temporal control of the laser by which a very short pulse can be produced; another is spatial control by which you can have a very precise interferometry and measurement; and at the same time, energy application and wave application can be performed in the fields of several technologies. As an example, temporal control can have an ultra short-pulse. It can provide us with a new field of laser technology. Also spatial control - you can have data processing, holography, interferometry, and so on. Thus, there is a wide range of new applications. And for energy applications, as you can see, material processing is one of the main themes. The properties of lasers can have wide applications in nuclear engineering. As for, wave application, spectroscopy, environmental applications, and nonlinear optics - many wide ranges are open. This document of the panel discussion was restored from a tape. The program committee has to state that there may be some errors introduced by the restoration procedure. -313 — This morning I heard a very interesting presentation by Dr. Letokhov. As you see, technology transfer and conversion in the future is a very important topic. For a long time new technology has been developed by military demands, and this is a very powerful way to develop scientific technology, because you can invest a huge amount of research money. You don't need any market at all for this purpose, and you can draw the fantasy of ideas. But nowadays, the situation has changed and so those technologies should transfer to civilian technologies. It means that some spin-off is expected. But the spin-off is very difficult scheme, because in the case of a fancy military technology, there is no consideration about the market, no consideration about efficiency, and no consideration about the cost. So this kind of thing, to transfer to civilian technologies, is very difficult. On the other hand, spin-on, that is civilian technology used for military applications, is very easy, because you can buy big quantities at reasonable costs. But the reverse way, spin-off, is a very difficult task. In Japan, during the war, all industries were for military. But just after the war we had nothing to do with the military use. So people got into civilian applications. We had no other way at that time. We have crucial experiences; this is a good example for the United State and Russia, and in this context, we want to have a discussion here today. The same topic - this is what Dr. Letokhov presented this morning. You see, nuclear engineering was developed very quickly because of the applications of this technology to the weapon - the atomic bomb. And laser technology has been going up exponentially. This growth rate is perhaps due to SDI, and other military demands. Now it is approaching, quite a bit, saturation. But in this area, very strong mutual interaction is expected to find a way to survive in future. I just show general scope. The first one is material processing; perhaps several people will be discussing in detail. It is the industrial application of high-power lasers. Next, laser isotope separation, about which Dr. Arisawa talked in detail. And another topic is the extremely bright short pulse femto second laser. You can have 1021 W/cm2. This is enormous electric field and you can produce 1012 V/m, which is 100 times bigger than that in the atomic field. So you can have many applications in this kind of field. Also the X-ray lasers, they have very interesting features. The water window is obtained between the carbon absorption edge and the oxygen edge -ranges from 10 to 40A. This window is very interesting for biophysics. Those things we should talk about. When discussing X-ray lasers, we must touch on synchrotron orbital radiation. This is very efficient for producing short wavelength light. The FEL is very closely related, and these are the coming topics. Further, the laser accelerator, the beat wave acceleration, this is one of the interesting topics. And nuclear excitation by laser; this is nuclear excitation by electronic transition, which can have very — 314 - interesting schemes concerning 7-ray lasers. In addition, nuclear spectroscopy can also be done by a laser. Now, I want to make three comments from my concerns, because this is a seminar for atomic energy research and related industries. As you know, the maintenance of nuclear reactors is a crucial issue. Heat exchangers, small pipes, occasionally develop small cracks. How to repair these cracks? You may position a sleeve over the crack and then you weld the sleeve in place using a YAG laser system and optical fiber to send laser light to the sleeve position. As you see, the tubes have a very small diameter, say 2 cm or so, and this device is positioned at the defect point, rotates, and welds the sleeve to the tube very nicely. This is one of the new material processing schemes that utilize the laser. I must say something about synchrotron orbital radiation. We are building the SPring 8 machine in the Nishiharima area. This is one of the three biggest synchrotron orbital radiation sources in the world - eight GeV and it has about 60 beam lines. It will be completed in three years. I want to show you the pictures. Yes, this is the Linac - a one GeV electron beam enters the synchrotron accelerator, and leaves for the ring here - you can see several beams. It will deliver 1A light, and so is a useful tool for the analysis of materials. Moving to another topic, this is our glass laser system, Gekko XL You see this is a 12 beam system that sends beams to the target and compresses the fuel. We have succeeded in compressing the fuel up to 600 times more than liquid density. This is the target chamber; you see over here the fuel pellet, put it here, and 12 beams come down to compress it. This is a picture showing the compression. This is a framing picture taken with an exposure time of 80 picoseconds. The interval between frames is 170 picoseconds. You can see that very tight compression of the core has been accomplished. By this method, laser fusion is coming very near to its goal. As you see, this is the density and the temperature, and the laser imploded plasma comes up here - at 600 times normal density. The solar interior here, and white dwarf interior are shown here. So you can have an astronomical experiment in the laboratory. This is one of the interesting features of fusion research. In this graph the abscissa is laser energy and the ordinate is neutron yield, and Gekko XH, and NOVA, green laser come up here. May be in a few decades, we will be able to get to the break-even condition. As a privilege of the coordinator, I covered the nuclear fusion research in brief. I now want to call Professor Takuma to make his presentation. I will introduce his career. Dr. Takuma was graduated from Tokyo University in 1953. He went to the NBS Colorado in 1964, stayed a couple of years there, and returned to a professorship at Tokyo University in - 315- 1967. For a while he had a job at JEOL Company, and he is now a professor of the University of Electro-Communication and Director of the Institute of Laser Science. He will talk about basic laser research. H.TAKUMA: Thank you very much Professor Yamanaka. Let me start from the conclusion part of my talk I gave the day before yesterday. I concluded that the next generation lasers will be laser diode-pumped solid state lasers, including MOPA systems and fiber lasers. And finally the pure diode laser will come. These conclusions are obtained from my experience that is limited to pure research using lasers. I have been interested in the application of lasers in various fields including laser isotope enrichment and also laser fusion. But I would like to remind you that most of the early applications of laser technology, in the decade of 1960s, were in pure research. The most efficient application of lasers was in pure research, mainly in physics. And also I would like to ask you to remember that the lasers approach to its theoretical limit, for example, in our institute we have a stabilized solid state laser, a diode laser pumped YAG laser, whose stability is much below the Schawlow-Townes limit. This is the theoretical limit, that is shot noise of classical light. If we squeeze the light, of course, we can go further, but not much. This is actually very close to its theoretical limit. Such progress is the result of rather pure research. So, even though weapon research is not as powerful a motivation as it used to be, after the end of cold war, still the basic research, scientific research, is one of the strong motivations to develop new technology. So let me point out some of the future applications of lasers in physics, which will probide a good chance to develop new technology. First of all, atomic and molecular physics was the earliest field in which the laser was very efficiently used. And what is left in this field? I think that QED verification, quantum electrodynamics, is still left as a field to be investigated more closely. Very nice work was done on the 1S-2S Lamb shift, and it is ever progressing. But it takes a very long time, because even in this simple system very complicated physics is involved. We must solve the problems one by one by making very careful investigations. And because of a basically different interaction, we need to study muonium, that is the hydrogen atom made by a muon and an electron. In addition, positronium, and also the investigation of heavy hydrogen-like atoms are very interesting. That is a heavy nucleus having only one electron, so the spectrum is essentially hydrogen-like but it is quite different. If the atom is quite heavy, the relativistic effect is very strong. So we need to investigate the quantum mechanical relativistic effect by investigating such kinds of atoms. Why hydrogen atoms? Because this system is the only system we can treat rigorously. Next CPT violation is a very important problem still left. That means the charge, spatial -316 — and time reversal symmetry. And for this, anti-hydrogen research is ideal. People are very eagerly working on this, but have had no success yet. However, I think they will soon succeed in producing anti-hydrogen and will trap it. Also the laser is very conveniently used to polarize nuclei. Experiments with polarized nuclei are also very interesting for CPT-type experiments. The forbidden transition intensity measurement reveals the very complicated mixing of electronic states; that is also a very important subject. Permanent dipole moment measurement also is important. Those are all very closely related to high resolution and high sensitivity laser spectroscopy. Speaking of exotic atoms like anti-hydrogen, muonium, and positronium, I should mention that recently Professor Yamazaki of the Institute for Nuclear Study of the University of Tokyo found an interesting complex that is an anti-proton, electron and helium nucleus, making a helium-like atom. This has the anti-proton replacing one electron of helium. The existence of this kind of complex was verified by laser spectroscopy at CERN very recently. A laser is useful also in nuclear and high energy physics. First of all, laser particle acceleration is a very interesting future tool to get high energy. I think the innovation of this new scheme of acceleration is more important than developing SSC. It may be theoretically possible to accelerate the particle in a rather short distance. This is a rather old concept, since in the early 1960s Professor Shimoda published a paper with an idea for accelerating an electron in a very narrow hole through a crystal, for example. The nonlinear optics of nuclei I have proposed is a very interesting subject. For example, laser induced r radiation and laser induced fission, etc. In order to have new knowledge about this, I'd like to propose an experiment with a particle beam accelerator. Such experiments must be very interesting, and we can think about, generally speaking, two types of arrangements: a high power laser and a particle beam accelerator. One is a DC field, that is the non-resonant case. In this case, the second-order Stark effect is the main effect on the nucleus. For example, two naked nucleus accelerated and collided with each other in a very strong laser field. The laser field can be made very strong, because we have a supermirror. There is no loss in the vacuum, so we can take advantage of supermirror in order to make the field inside the resonator very strong. We can make a resonator having a finesse, for example, of 100,000 and more. So the radiation intensity inside the focal region is enormous. Under such strong field, the reaction must be very strongly influenced. Another scheme is the resonance field. If the heavy ion beam is accelerated with very high speed, the relativistic Doppler shift makes the laser wavelength - looking from the coordinate fixed to the particle, very short. It is in the wavelength of 7 -rays. It can be tuned by adjusting the angle of the laser beam, so one has an angular tuned nuclear spectroscopy scheme. This must also be very interesting. Thus, the combination of a high peak power laser and a particle -317- accelerator, not of protons or electrons, but a heavy ion accelerator must be very interesting. Then let me go into my favorite field that is the laser manipulation of particles. Professor Letokhov is one of the very early pioneers of this field. We are very much interested, and as I have shown in my talk, we succeeded in the focusing of the atoms by a microwave cavity. This kind of work must also be interesting. By cooling and trapping neutral atoms, we can manipulate the neutral atoms in any way we like, to bring them to some other place or to see the Bose condensation. In our laboratory we study both Lithium 7 and Lithium 6, and those are the most promising atoms to be cooled to high density and collected at high density, because the escape rate is lowest in these atoms. I don't know what the practical application of Bose condensation is. However, it is one of the pioneering fields we must exploit in the near future. In addition, atom optics must be very useful in high accuracy interferometry and also as a surface probe. For pure research, we have a lot of opportunities to see quantum mechanical effects directly. I briefly described what kind of applications in physics are possible in the near future. Even in this limited field there are so many things we can do in the future, and for doing such research, we need significant technical development; that may be one of the motivations. Thank you very much. C.YAMANAKA: Thank you very much, Professor Takuma. I will call the next panel member, Dr. Takashi Arisawa. He will talk about laser applications for nuclear engineering. He graduated Department of Mechanical Engineering, Yokohama National University in 1969. Now he is a leader of laser isotope separation team in JAERI. T.ARISAWA: Yes, I would like to present the laser application in nuclear energies. I am not going to give philosophical discussions like the previous presenter. I just picked the specific technology in which lasers might be used in the future from a practical point of view. I divided the application field into several groups. First of all, diagnostic. For diagnostic, of course, a radioactive element is very easy to detect using a 7 -ray or alpha particle. But in some cases, as Professor Letokhov presented this morning, for example, the trace radioactive strontium isotopes, are very difficult to measure. So we need to have a very special technique using RIMS, Resonant Ionization Mass Spectrometry. We can measure a very small trace amount of materials and elements using this technology, and this is very important from the viewpoint of environmental safety. This technology can also be applied to very sensitive detection of TRU, transuranic atoms. In the US, there is a problem, a radioactive gas hazard caused by radon gas. So in this case, making use of the fact that radon becomes lead in the final stage, RIMS can also be used to detect small amounts of lead to obtain a precise analysis of radon gas. - 318- Another application is process control. In process control, for example, in an enrichment plant, including the gaseous centrifuge process, it is indispensable to monitor or to control the concentration or density of the materials, UF6 in the case of an enrichment plant. Similarly, in a reprocessing plant, people want to monitor liquid density and the concentration of elements, both precisely and remotely. So in these cases, a tunable laser, a portable multicolor tunable laser, would be very useful. The next group is material processing. In the fuel fabrication, laser is already useful in many places. In this case, I think the fiber delivery is required, and also high power lasers are required. When we can get a high power laser, it can be used for disassembling the fuel assemblies. In a reprocessing plant, they are now using a shearing machine to cut and chop the rod assembly. A laser can be used, but average power should be higher than 10 kW and also fiber delivery is required. When lasers having a power of 100 kW or more are available, we can use these lasers for the decommissioning of nuclear facilities, to chop the structural material into pieces, and this can be done in places that require remote access. In this case also, fiber delivery is required because these things must be cut underwater. For surface decontamination, nowadays lasers are already used for this purpose. A pulsed laser is used to remove dirty particles from the surface. I shall move to the next group, which is the selective photo-reaction. For this application, we have to have a very high quality tunable laser with narrow linewidth, stability, and reliability. Multi-colored tunable lasers are especially convenient for the production of fuel for fission reactors, that is low enrichment uranium. As I mentioned this morning, multi-photon, multi-step photo-ionization or infrared multi-step photo-dissociation - these two methods are used. There was a very pessimistic presentation this morning by Dr. Schneider of URENCO about the future of laser uranium enrichment, but cost of enrichment depends on many things. Cost depends on the economic situation of the country and the social and industrial framework. It is not an easy thing to calculate the cost precisely at this moment. Anyway we will continue to develop this technology. Yes, Japan will continue laser isotope separation technology development, both atomic and molecular. The next item I would like to point out is heavy water production. This would be very easy if you use IR, infrared multi-photon dissociation. Ontario Hydro in Canada has been working on this development. You can also use formaldehyde as a working material using single photon dissociation. Another application is the removal of hazardous elements, for example, tritium, from nuclear waste. But the concentration is really low, so we need a very efficient separation system. In this the laser plays a very useful role. Halogenated hydrocarbons with low dissociation thresholds can particularly be used for this purpose based on the results of infrared multi-photon - 319 — dissociation. In the future, in the reprocessing plant, photo-redox methods might be utilized. In this case, a tunable laser with a linewidth that is not too narrow can be used. And for stable isotope separation; isotopic tailored materials might be useful, for example, for fusion reactors in order to reduce induced radioactivity, which is harmful in the long term. In that casec, chloide compounds of some structual could be used as working material based on infrared multi-photon dissociation. And also for the future fusion reactor, whether it is a Tokamak or laser fusion system, we need lithium and tritium separation. Next group is the application to energy sources. I think there is no need to say anything about laser fusion. A direct nuclear pumped laser may be applied to transmit high power energy to a remote area, for example, into space. So a reactor may be used to transmit power optically. This is a very special case. Some of these ideas are dreams, but some of them can be realized in the near future. We have to take account of the timing and the cost effectiveness when we develop technologies in the nuclear engineering field. But as you know, for the development of nuclear technology, we have to take account safety also. So the costs are rather high in the nuclear energy development compared with ordinary industry. The cost of a laser photon is also high. So I think the laser photon application to nuclear energy is very promising, if we combine the costs of these applications. We can not develop laser photon technology without its application to the nuclear energy field. Thank you very much. C.YAMANAKA: Thank you very much Dr. Arisawa. Now I would like to call Professor Letokhov. He graduated Moscow Physical Technical Institute in 1963 and has belonged to the Lebedev Physical Institute for some time. He is the head of the Laser Spectroscopy Department of the Institute of Spectroscopy in Moscow. He received the Lenin Prize in 1978. He is so famous that there is no need to introduce him anymore. I would like Dr. Letokhov to talk about the future scope of laser application to science. V.S.LETOKHOV: Let me start by continuing the philosophical discussion about how we can predict, foresee, applications of fundamental science. This is a very important point - Professor Yamanaka mentioned this just now. In this plot you can see, in arbitrary units, time scale and activity. It is general; may be it can be considered for laser technology, and for others. The green line is more or less natural development - natural growing without artificial enhancement, that is, artificial acceleration like the nuclear bomb, Star Wars - SDI -, and others. It is fashion-like in a way. Really, in science sometimes people are working in a narrow field - too many people are -320 — working in a narrow field, and start to heat up some few ideas. It seems to me the number of scientists, at least in the physics area, are sometimes more than necessary. And all of the ideas will be exhausted, but technology is still not ready to absorb these ideas, because technology has its own rate of the growth, in an internal natural way. And as a result, you have a kind of gap, - it looks like a fall or failure, - how to say in English, a kind of negative result. But it is not necessary; it was because of too great an involvement of scientists working in a fashionable field. It looks like a catastrophe. But this is a gap between fundamental science and technology. It is quite artificial - it is not natural. Such a phenomena is known for a long time - there always has been a gap between science and technology. And we are working in the real world and should consider this point. This is a kind of sociology of science. After such a philosophical introduction let me say a few words, more technical words, just to continue my previous discussion in this morning. This was my last transparency about the laser separation process, which is very, very universal, and we can exploit this technology quite well from many respects. Now I want to demonstrate to you how we can separate nuclear isomers. This shows our experiment in the Leningrad, now St. Petersburg, Nuclear Physics Institute using a dye laser. It is a case of samarium; samarium generated in the excited states and ground states, the nuclear excited states being produced by the proton bombardment. Because of the different nuclear scheme, hyperfine structures are different for the first step. And we can see such a structure. Here is a mixture of the atoms in their ground states and those in the excited state, isomeric samarium. And you can see that here, only the ground state nuclei can absorb light. I suppose here is the metastable nuclei and you can separate this. And of course, we are able to detect nuclear properties, decay time, and other characteristics. This is an example how you can use this technique for separation. One of my dreams in 1973 was the separation of nuclei for a 7 -ray laser. But again I don't want to invest any money in this field, because we would need 10 more years of fundamental research before we would be able not only to separate the excited/unexcited nuclei, but also to arrange preparation of crystals and develop techniques for the reduction of the non-homogeneous (Bolnikov? ) Mossbauer line. In this field it is now possible to buy a (Nilanikov defectless crystal to convert amorphous material to the crystal material? ). And the next step will be to develop new ideas - we need new ideas how to reduce the Mossbauer line, because the defects of crystals still are able to broaden this Mossbauer line a million times. This should be reduced at least a factor of several thousand. This is the same kind of experiment we have done for thulium, thulium on-line separation. Metastable nuclei and ground state nuclei; you can see here this structure which has lines belonging to the thulium atom in the excited metastable nuclei. We can separate and study the properties of this. It is another example. -321 — And now I change, because we are talking about separation a lot in this symposium - it is right, because it is a very important technology for the future. I should mention about energy consumption. We still have a huge challenge for good inventive people to achieve the thermodynamic limit concerning the consumption of energy for separation. Energy of separation is determined by the thermodynamics - this value - this is a ratio of isotopes and for UF6, of course, it must be around 100 times more than kT for room temperature. 10"2 electron volts for this isotope, and for this they are 100 times more. But we are very far from this limit, very far. Even the best centrifuge technology needs, to separate the atom, about 105 electron volts. The best laser technology - 10\ 103. We still should think of some other approach, a nonstatistical approach for separation. I myself am now working, or spending time on a computer at home. In cold winter time - it was quite cold in Moscow during the last two months, -20 °C. I couldn't drive, because the road was blocked by snow. And I was thinking during this time about some new approach. This is one approach based on the near field effect. I want to use the gradient force; this is the near field, and this is a hole, a small hole, its size being 10 times less than the optical wavelength, two-pi less. And here, the laser light is coming in this direction to penetrate, and you have a gradient force as a result of the gradient of intensity. It is another type of force, not based on the spontaneous emission, but based on the stimulated absorption. This photon is supposed moving in this direction; and the emission in another direction. As a result, a part of the momentum will be transferred to the atom. This atom current can be in this direction and can be focused in an empty place without a real focusing device. You can foresee several potential applications. First of all, it is very useful for lithography, potentially of course, because it still has huge chromatic aberration here. According to our estimation, you can focus on a small size like 1 nm, few nanometers, not more. This is one possibility for the future to control atoms by laser light. And also we can make manifolds of any structure, and can deposit in a pattern in which we want to do this. And also this is the realization of Maxwell's Demon, because the gradient force is resonant force, sensitive to the type of particle and velocity of particle. As a result we can have cross-section for acceptance, cross-section of penetration of atom of a specific sort or a specific isotope. It will be higher for a resonant isotope, not for the other. According to our calculation we can enhance the cross-section several times. It is also a way to control penetration of the particle through the structure, through the filter like this. These are just a few examples of potential future applications. Let me underline and emphasize that in spite of the 30 year history of laser technology and laser application, we are still in an exponential phase. We can expect many more magnificent results in the future. Even for, let's say for example, the isotope separation by the gradient force. - 322- Here, the gradient force uses a very small energy for separation; stimulated absorption and stimulated emission, and you keep energy in the field. But of course, it is a fantasy, we need maybe 10-20 years more for the development of this technology. What is most important and challenging for me is that there is still a huge area. You can after five years, after ten years, you can again find another new interesting idea. It means that laser science is still growing. It is not converted totally to the technology, to engineering. It is a nice field for work and I think you can foresee a lot of nice application in nuclear technology. Let me stop, so I do not make a fantasy, because I want to visit Japan again and want to keep some material for my next visit. Thank you very much for your attention. C.YAMANAKA: Thank you very much Dr. Letokhov. Now I will call Dr. Ewing. He graduated from the University of California, Riverside in 1964. He attended the University of Chicago and, after getting his Ph.D. degree, he joined AVCO Everett Research in 1972. And he is very famous for the discovery of the excimer laser. He was for a while at the Lawrence Livermore Laboratory, and in 1979 he joined Mathematical Science Northwest, which was later renamed STI Optronics. He became president of STI in 1992. He recently set up Aculight, where he is also the president. He will talk about the future prospects of lasers. J.J.EWING: Thank you. I arri just going to make a few brief comments along the theme that our company is really organized along, and that is that solid state lasers have a very bright future in certain particular areas. And I want to talk about one of the things that came up in discussion yesterday - about when diode prices might come down enough to where the solid state revolution is really more complete. I'll start with this chart which I showed yesterday, which is the concept that with parametric oscillators and other techniques - frequency doubling, frequency halving an so on - that one can have very extensive coverage of the spectrum. And here, this is plotted with sort of relative energy against wavelengh. And optical parametric oscillators, neodymium lasers and so on can do that. And this is sort of relative power. But we also hear or we have seen discussions of, for example, diode-pumped lasers possibly eventually being used for laser fusion. So one could also make a plot, for example, of energy possibility there and they go from the sort of millijoule class devices that are in the marketplace right now up to these megajoule kind of things that would be appropriate for fusion. So as well as covering the wavelength, spectral region, there will be coverage of all of the different energy regions, if this dream comes true. So the concept, as I showed the other day is here- some kind of high powered diode arrays with coupling optics into some kind of high power neodymium laser. And high power today in the marketplace might be a number like 5 W and may be eventually even larger. -323- So the real question is- people have been saying this for 10 years or more - that diode-pumped lasers are just around the corner. It is the old light at the end of the tunnel routine and some people have never believed it. But I think the time is coming and the cost of the diode arrays has been dropping steadily. This is a chart that has been shown in a variety of different forms; it gets updated every year and it's subject to interpretation, argument, and so on, about what are the real prices that are on here. But the conclusion is that the selling price of high power diode arrays has been dropping steadily. These are years and these are costs and in this particular one, $/mJ or .S/W average for what are called the quasi-CW arrays, and here is the cost in the CW arrays, which really relate to a lot of the commercial applications right now. Very early on the costs were driven by research programs that were funded for a variety of reasons to make lasers that might have some tactical military applications or other things like that. But the cost came down substantially when people started building research lasers that were capable of putting out 100 mJ or a Joule and people had to buy hundreds of bars per year. Now you will see, and I will give this caution again that there are a variety of figures of merit that people might talk about in terms of cost - dollars, yen or whatever you want to think in terms of - per bar, per millijoule, per Watt average or per Watt peak, and just always make sure you understand what it is that someone is talking about and what the conversion factors are. Quasi-CW bars are for Q-switched or long-pulsed operation. And the CW arrays are used for most of the current products. There is an error bar on each of these numbers because it depends on the wavelength you are getting; are you pumping thulium; are you pumping neodymium, etc. It depends on a variety of other factors, including how many of them you are buying at any time. So, how low do the prices need to get before this revolution will be more significant? In my mind, it is a very simple comparison. And it's one of the things that makes diode-pumped lasers both attractive but difficult in terms of where they will eventually become more readily accepted. The biggest difference between a diode-pumped neodymium laser and a conventional neodymium laser is that you are using a very efficient, small and reliable, but very expensive pump source. You are using that in place of a very cheap lamp. The very cheap lamp happens to come with a big power supply, whereas the little diode happens to work with a small power supply. Now, the thing is you are replacing effectively a large power supply with an expensive diode, so you are competing against an established technology. And power and cooling systems, whether you are building them, buying them, or whatever, you can go to any number of places, and find out those things will cost something on the order of 50-100 yen/W to build or buy. And, yes, there are some scaling factors depending on the size that you want to have. Now the - 324- overall efficiency ratio of a diode-pumped laser to a lamp-pumped solid-state laser is in the range of 10 to 1, and again someone can argue that there are some lamp-pumped solid-state lasers with 3% efficiency. I am not here to argue that; it is just to give a rough number. So the cost of the diode arrays needs to get into the range of $10/W for substantial use. The power supply goes down a factor of 10 and that is the money you are saving. And additionally, the end-user may care about some feature of small size or something else, but in the end, it's how much he has to put out in money to get some accomplishment to either please his funding agent in the National Science Foundation or whoever, or to do separative work, or to melt steel, or whatever it is. Typical diode arrays have per bar average powers of 15-20 W, and again there are variations in design here. This implies the bar price needs to be in the range of $100-200/bar before one can have substantial conversion - before the price will be low enough to where people will choose the diode-pumped one over the lamp-pumped one for all applications. So, that basically says that the sales prices roughly have to come down somewhere between a factor of five or ten from where they were in 1993. And, if you go back to this chart, you can see roughly what the trend line has been. And you say, 'Okay, when will it come down this next factor of five or ten?'. And there will be markets that develop as we come down this curve. But that basically says we need to be down here somewhere; and that says somewhere around 1995 or so is when the diode bar prices in principle could get down low enough. Now if you really look at these things carefully, you will see there might be inflection points, and this one might say it's the year 2000, so you need to be careful. And this is history and doesn't say anything about the future. And the real question is 'What drives the cost down?' Well, what drives the cost down in anything like this is demand. Demand is of course related to markets. But, here are just some projections I have used of what the number of bars per year needs to be to bring the diode array prices down. So they have been dropping on a curve like this. And what I have got plotted here is something where it's in bars/year, going from less than a hundred bars, where you were just making a few prototypes and they might cost 10,000 a bar, a piece, up to order of 100,000 or more bars per year. And you can use that factor of roughly 20 W or so average power to go to how many kilowatts of diode arrays need to be made each year. Your prices right now in this particular plot are in this range, depending on what wavelength you want, how many you are buying, and so on. And there is about a factor of 2-3 variation in the price. So these are cost-volume estimates from historical prices and sort of total market estimates and a variety of different industry sources. And there are variations.on that theme and how you might come up with a cost for what this is going to be. But the real point is that the volume that is required to achieve the $10/W or $200/bar is really quite substantial. You need to — 325- be doing something like 100,000 bars per year or 2 MW of output. Now you could look at that and be very frightened by that and say we will never get there, because this implies something like 400 1-kW lasers per year. And that says to me very clearly that industrial applications are needed. And they have to be significant volume. Now there might be uncertainties in here - is it 400 1-kW lasers or 100 1-kW lasers, but it is something in that ball park. And another thing that says is it is not going to be little 1 W lasers that will drive the cost down. And that is really different from what has happened with the video disc playing arrays, because there unit volumes have gone up and the prices has gone down based on very, very large numbers. One of the problems is that there are just not that many bars that need to be made in this high power range. So really, I guess concluding points that I want to make is - rather than being, in a sense, technology-driven and saying what are some of the great sciences or technologies or whatever that we can do, we really need to be looking at end-users and applications and say 'Where are these applications going to be?' And one of the things that I found interesting just in looking at some of the viewgraphs that were presented, I realized that the answer to some of this is really sort of here already; that a number of us probably use software like Microsoft Word or something like this to create charts. There are probably more people in the past year that have bought Microsoft Word or Word for Windows than have bought a little laser to do a projection. It is a curious thing and when you think about where money might flow and get some return, there are other things that have come along - for example, software and computers themselves, laptops and so on - that are really changing in a dramatic way. And you really have to get into this area of very, very large production. And other places in lasers where that has happened, for example in helium-neon lasers; that has really been driven by the fact that He-Ne lasers got out of the laboratory and got into the supermarket. That drove the cost down because production went up. So I am very interested in knowing where might we be making a couple hundred 1 kW or a couple of thousand 100 W lasers. Thank you. C.YAMANAKA: Thank you very much Dr. Ewing. So the panel members, please go up to the stage to start discussion with the audience. The four panellists presented several different topics and I would like the participants to make a question or comment about those items. As you know, this is a very interesting meeting to think about the future, say, one decade, two decades. Anyone making a comment, please state the name of the person you are addressing. QUESTION FROM AUDIENCE: Dr. Arisawa has shown a very nice viewgraph showing various — 326- different applications and laser qualities going with it. I really appreciate that table. In the table you mentioned at various places 'narrow bandwidth'. Can you explain that a little more quantitatively, 'narrow' meaning how narrow in terms of relative bandwidth? T.AR1SAWA: Yes, for laser isotope separation, the bandwidth depends on the spectrum. So, in the case of uranium, uranium-235 has hyperfine structures, while uranium-238 has no hyperfine structures, and there is the isotope shift; so we have to have narrow linewidth which can recognize the difference between the two isotopes. I should say that 1 GHz is okay for such isotopes, but in other cases, in other cases of a small isotope shift, you have to have a very narrow linewidth - maybe several tens of megahertz, something like that. CONTINUATION OF QUESTION: I am a free-electron laser man, apparently ignorant of all these things. How does it translate in terms of relative bandwidth, say suppose you use a molecular approach, then you use infrared radiation? Is the relative bandwidth of 10'3 enough or does it need to be 10"5? T.ARISAWA: Okay, in JAERI, we are developing an FEL in the infrared region, which might be used for molecular laser isotope separation. I am not in charge of this development, and I am not sure what the linewidth of the free electron laser is. I myself am working on atomic vapor laser isotope separation, and the free electron laser people say that in the next step they will develop an ultraviolet or visible wavelength free electron laser. So, if it is successfully developed, maybe in two or three years, I am not sure, but we will have visible light, which has narrow linewidth. CONTINUATION OF QUESTION: I'm still trying to find out your requirement. What bandwidth would you like to have? T.ARISAWA: The requirement is, as I said before, if you apply the laser to uranium isotope separation, we need 1 GHz - at least 1 GHz. Although I didn't explain this morning, in order to ionize efficiently uranium-235 with hyperfine structure, normally we need to have several tens of megahertz in linewidth, and we modulate the laser frequency within the pulse. That is called chirping. So in that case, I think 50 megahertz should be required. AUDIENCE: Thank you very much. J.J.EWING: I would like to make a comment about the last question as well. I am certainly not an expert in molecular isotope separation either. But in addition to the question of requirement for spectral widths, I think I heard in earlier talks a requirement for a relatively high fluence per pulse in a kinetic time associated with the multi-photon excitation dynamics in the J/cm2 range. And I think the kinetic time is something less than microseconds; it's my impression. So that I believe the natural waveform from the laser tends to be high repetition rate micro-pulses with -327- perhaps 100 nanosecond interpulse time. So one would be necessary to accumulate micropulses within one microsecond to achieve J/cm2 instantaneous fluence. That would be my observation. V.S.LETOKHOV: Let me add a little bit to your information. Actually, for heavy molecule like UF6 the threshold of dissociation by (viewing for light?) is 30 times less than you thought. Not Joule, 30 ml is enough. Nevertheless, it's quite a big energy, fluence, for the free electron lasers, but it is not as much as needed for the separation of a more simple molecule. C.YAMANAKA: Any other questions, the topic is moving to the free electron laser. So perhaps Prof. Smith may have some comment on the frequency stability of free electron lasers. COMMENT FROM AUDIENCE: I wasn't aware that this was supposed to be a free electron laser conference, but it appears to be turning into one, at least temporarily. One comment I'd like to make as a personal observation is that this conference has shown how many different areas there are in which light is needed and that I, for instant, was not aware of. And I suspect that that may be true of others as well, so I am very grateful for the opportunity to be here and become educated myself a bit about this. The free electron laser has the capability of doing an awful lot of things. The light that is produced is basically governed by the electron beam. And the electron beam can be produced in many, many, many different ways. A very common way now, because it is easy and natural, is by using a linear accelerator that of course produces micropulses. There are van de Graaff machines which produce long pulses in the microsecond range; induction machines have been used which are intermediate between those two. I would say that if an application came along that required specific characteristics with enough money behind it, I suppose, to be crass, there might be a strong effort to meet those demands. And of course this is all said with the full realization that the free electron laser has been promising an awful lot for an awful long period of time. And as yet, the highest average power FEL that I am aware of is something like 10 W. Yet there are serious proposals; we have one, other people have others to produce 1 kW demonstrations in the near future. There is a superconducting accelerator in the United States, CEBAF, which is attempting to get funding for a program which would lead to a 100 kilowatt UV free electron laser with fairly narrow bandwidth. It is an example of the kind of directions which things can go. And the projected cost per Watt delivered for that machine is less than one cent. So it is intended for an industrial application where it would be very competitive with any other process. C.YAMANAKA: Thank you very much. Any other questions from the field? QUESTION FROM AUDIENCE: I have a question to Dr. Ewing. My name is Tomoo Fujioka of Tokai University. Yes, your scheme of all solid-state lasers, the tunable solid-state lasers, is very -328- attractive, and I believe that it will be very widely used in the future in laboratories. So I believe you are going to be much richer in the future. But if we consider industrial applications; yes, I remember in 1968 at IQEC conference in Miami - Professor Shimoda and Professor Yamanaka attended there - some papers appeared describing very high efficiency frequency-doubled YAG lasers, which attained more than 50% efficiency. At that time they said 'Good-bye argon ion laser'. From that time, all the people believed that for blue or green light, frequency-doubled YAG laser will be used instead of argon ion laser. But now, many people are still using the argon ion laser; according to my limited knowledge, no industrial application of the frequency- doubled YAG laser is there. And the major reason is the creation of the color center in the crystals, nonlinear crystals. So I wonder if we overcome that defect, or if high power nonlinear lasers can be used in the future. So my question is 'Do you have some good ideas to become much richer to overcome that defect?'. That is my question. J.J.EW1NG: So I really thank you Professor Fujioka for your concern of my wealth. This is very important and should not be taken lightly. There are a number of comments that you have made and they're all important and they all are considerations in this. There are applications for each of these lasers, and I think it would be foolhardy for anyone to imagine, for example, that neodymium may in fact replace C02, because there are an awful lot of things that C02 does, does it very well, and it does it now; and it doesn't do it at sometime in the future. There are people doing welding, cutting, and so on and they had 10-20 kW 10-20 years ago. And people have put those into real industrial applications. I think that a key thing in bringing a new technology, be it a laser or biotech or anything, to the marketplace is to be sensitive to what the market really needs. I think your example of frequency-doubled YAG and argon ion lasers is a very beautiful example of that. The basic reason that YAG doesn't replace argon ion lasers is that, at the 90% level, 90% of the units of argon ion lasers sold are sold for blue applications. And those blue applications are not accessible to 532. That is something where we still get research money to - we are just starting on a project here shortly to look at nonlinear techniques that are more advanced than simply doubling to make blue and UV. And that still is a research area, but it really responds to a need that's real. That people in the printing industry, who use argon ion lasers say, a) it is established, b) they are inexpensive, and c) they know what to do, and the whole thing is centered around blue and UV wavelengths. And they are not going to change. You could say the same thing about the fourth harmonic of YAG and KrF. For example, there is a very beautiful potential business, that is, industrial KrF lasers in photolithography. It may never happen in volumes because lamps are so good for doing photolithography anyway. But 266 is not 248, and that is a significant point. And it just really needs to be looked at. And, when you then start saying 266 - how would you instead make it be 248? It is a different - 329 — scheme. You really have to have a look at the application and really understand what the end-user wants. And I think all too often the laser world has suffered from the fact -1 am sure that I've been one of those people out there doing it - that we have this really great thing, please use it, please use it. In the end, that doesn't solve the problem of someone who wants to cut something, weld something, or do some particular process. You really have to understand his needs and have a laser technology that fits those. Now you also mentioned color centers and lifetime of nonlinear crystals and so on. That is an issue - it depends on wavelength, depends on where you are going. Some of those things are getting better with time as new crystals come along. Things look better in certain areas. Even the same crystal, KTP for example, it really depends on how it is grown. So there is a lot of science that needs to be still done there. And another, I think, problem has been that really doing extensive lifetimes of solid-state nonlinear components hasn't been possible, because the drive lasers haven't been there with ultrahigh rep rates and the reliability that is needed. That stuff is coming now, and I believe that technology will drive, shall we say, the reliability engineering of the nonlinear crystals. But there will be work to do; I have no doubt about that. C.YAMANAKA: Any other comments? There are various kinds of lasers, and Dr. Ewing has proposed the diode-pumped solid-state laser can have all possible applications and potential to make money. May be some other lasers are very efficient. So does anyone want to say something about other kind of lasers, and protest this kind of proposal? X-ray lasers and some kind of new COMMENT FROM THE AUDIENCE: Yes, diode-pumped lasers should be a very good laser compared to the flashlamp-pumped YAG lasers. It is undoubtedly. But I think you told us the end-user's demand is low cost, number one is low cost. So in that sense, in future the cost of the diode arrays will decrease - it can be overcome. But now we don't know how long it will take. And another point is the beam quality; the beam quality is very, very important. When the power level is low, the beam quality of the YAG laser is diffraction-limited. But if it exceeds a few or several hundred Watts, kilowatts or 10 kW level, the inherent defects of the big crystal decrease the beam quality very severely. So for such applications, even the diode-pumped YAG laser cannot be used for very precise material processing applications, I think. Professor Yamanaka asked for some other laser candidates. I am proposing, only my group is proposing a very fine beam quality laser, the COIL laser, the chemical oxygen iodine laser. Many of the people here don't know the iodine laser, but in Nagaoka - may I show one viewgraph? Please note this is not my company. This is Nagaoka ALEC company, where they have the - 330- first industrial COIL laser. The COIL laser is the chemical laser, but very safe chemical laser unlike an HF or DF laser. This is a photograph of the first lkW industrial chemical lasers, installed at ALEC company in Nagaoka city. It is very huge, but please remember that the first 1 kW C02 laser - the paper was presented also in the IQEC conference in 1968 - the length was 100 m. In this sense this laser is very small compared to that. V.S.LETOKHOV: What is the price of the laser? COMMENT FROM AUDIENCE: I don't know, please ask ALEC. Cost is very expensive, because it is the first one. All of the participants here, including the participants from foreign countries, can test this laser in Nagaoka. This is a 1 kW laser and the beam quality is ideal, i.e. diffraction-limited. The transmission through fibers is much better than YAG lasers, because the wavelength is 1.316 micron. C.YAMANAKA: Any questions about this kind of chemical laser? Any questions from the application field? How about the life of the laser media? COMMENT FROM AUDIENCE: This is a chemical laser. If we provide the fuel, life is forever, I believe. I am optimistic. C.YAMANAKA: There have been several kinds of chemical lasers; for instance the Max Planck was using the iodine laser for the fusion applications. But they stopped the experiment quite recently, because maintenance is a rather hard task. Maybe the glass laser is much easier to operate. QUESTION FROM THE AUDIENCE: Just a comment. I believe there have been great advances in semiconductor lasers which have given rise to this great renewal in solid-state lasers. But I think in parallel with that, there has been excellent work done on mode control of highly loaded solid-state media, particularly, the work that has been done in Germany by Veppers group and several other groups examining conventional stable/unstable resonators, very sophisticated gradient mirrors and the implementation of them. In addition to that, there is also the practical realization now of practical use of phase conjugate mirrors, not only for a single pulse, but for repetition rate lasers. And I would like to leave my own opinion with this committee that the prospects for near diffraction limited solid-state lasers in the multi-kilowatt range is quite to be expected in the near term. Again it is not going to be easy - research is continuing, but I think good progress is being made in this field as well. Now, I have a question to Professor Takuma. You made the very interesting suggestion and observation of the productive marriage that can be made by laser science and heavy particle accelerators. This observation has been made by some individuals also in the United States. I think that what I have seen as a laser scientist has been a very slow recognition on the part of heavy particle physicists as to the benefits that can be accrued. So my question really is - How -331- can we as laser scientists in the community promote the prospect to translate these ideas into practical commitments to proceed in this research area? H.TAKUMA: Well, thank you very much. I welcome this kind of question. I am very much interested personally in the marriage of the high intensity laser and heavy ion, accelerated heavy ions. At our institute we have three professors, and one of them is now responsible for the special grants of the Ministry of Education, Science, and Culture on Heavy Ion Science. The group of people working on the atomic physics is now interested in the multiply-ionized heavy ions. They are producing it in the superconductor, magnet and confining the plasma in a very narrow diameter and making the temperature very high, they can produce multiply-ionized, rather heavy ions very efficiently. And they are also interested in the cyclotron, electron-cooled cyclotron type experiments. So really, the laser physicist and the particle beam physicist are now very closely working - just started. So I see a bright future. And also there is an interesting experiment, as I briefly introduced, done at the CERN. This was done by Professor Yamazaki. He is the Director of the Institute for Nuclear Study of the University of Tokyo. He is interested in the exotic atoms, or exotic complex including the anti-particles. And a very detailed spectroscopic study is now being made using tunable lasers. So I think it is really starting now. And his next project is to construct a heavy ion accelerator and the interaction with a laser beam is already included in it. But the experiment under the very strong field -1 mean the combination of strong laser field and high energy heavy ion beam is my proposal. Thank you very much. V.S.LETOKHOV: Can I a little add about that. There is another combination of the laser and nuclear physics which will be important maybe in the next 20 years. That is connected to the multiply- charged ions, and you have there nucleons, not the bare but a few electrons around the nucleons, and as a result, nuclear transition will have satellites, red and blue shifted satellites, because they are so-named mixed transition with simultaneous change of the electron shell energy and nuclear energy, nuclear excitation energy. That is a lot potentially to make in the future, not now. The combination of the T-ray laser and X-ray laser. What you need is not the total inversion of population, but a kind of the partial inversion of the population. The small amount of the excited nuclei can generate the light, or can generate the shifted satellite which will not be absorbed because the transition will be to excited state or excited electron shell. It is enough to have inversion population only for electron shell to generate nuclear transition. It is a new physics. Now because of the progress of the table top femtosecond laser with high energy like one Joule, you can now produce the laser generated multi-charged ion plasma. And you may observe such light which has never been observed actually. We have a kind of potential for new physics here. -332- C.YAMANAKA: Are there any questions and comments? Maybe for a 7 -ray laser proposal? Or some chemical applications and processing? QUESTION FROM AUDIENCE: I want to ask Dr. Arisawa if reprocessing of spent nuclear fuel and partitioning of high level waste may be an important subject after laser isotope separation. For example, this morning Russian scientist, Krynetsky gave a lecture about Pd separation from solution. I think you have some ideas for laser applications to partitioning and reprocessing. T.ARISAWA: Yes, now the reprocessing plant is being operated by the PUREX method using solutions. I am thinking of a new method to reprocess spent fuel, which should use the so-called 'dry process' based on the atomic vapor laser isotope separation. There are many kinds of isotopes produced in spent fuel. We have an electron-beam gun to heat up material as is done in the uranium vapor laser isotope separation. So if we apply this technology to the separation of the multi-element system like spent fuel, we can simplify the whole process of reprocessing. The one used today is a very complex system. We first sort the elements into some groups, based on the vapor pressure. The volatile ones can easily be extracted from the spent fuel. We can separate some elements through such a evaporation/distillation system. Then after sorting out those elements, we can apply a method similar to the atomic vapor laser separation method to the evaporated elements. But we need to have a lot of basic spectroscopic data. Scientific investigation is required to implement this technology in the future. But I think this will help the process of reprocessing. This is only a preliminary study or, a conceptual study. Is this the answer to your question? C.YAMANAKA: Okay, thank you very much. One more question or comment? QUESTION FROM THE AUDIENCE: I have a question for Professor Letokhov. Just now I think fission power is an extremely important supply of energy in Russia. How can lasers in the near future be properly used in the nuclear area in Russia? V.S.LETOKHOV: That section is not my field, but I can say to you there are nuclear reactors working at 25% of their capacity, because we overproduced the energy. But the government can not stop. Coal is too expensive, and nuclear energy is much cheaper than coal energy - but the coal mines cannot be closed because that would cause unemployment for millions of people. As a result, we have not only too many nuclear warheads, but we have too many - nuclear power stations. The stations are being used at only 25% of their capacity. Can you imagine? Nuclear energy is the cheapest energy in our country, as you know. All the rumors about inefficiency, or bad safety standards of Russian reactors are not correct. It is a question of their exploitation, and Chernobyl was a special social event. I can say to you privately, not to you as an audience; this event was not connected with nuclear technology, but it was connected with the human factor. That is the point. -333- But what we are seriously thinking about is to use the lasers for diagnostics, for diagnosis of the crucial elements of a nuclear power plant. This seems to me a very efficient way, because we can enhance the lifetime of the nuclear reactors. C.YAMANAKA: Okay, thank you very much. The time is very close to the end. So at last, I would like to make a summary of this session. We have been discussing laser applications to nuclear energy. One of the conclusions is that it is very difficult to make money using lasers. But in the future, we would like to develop these kinds of things. Professor Takuma commented that the development of laser physics can be a strong incentive for the development of applications. As the science frontier - gravity wave detection, squeezed states, chaos, cooling, accelerator, nuclear excitation by electron transition, and also inertial confinement fusion - these kinds of scientific frontiers reflect back on the development of the laser, vice versa, these are chicken and egg relations to elevate the level of laser technology. The stability, coherency, power, and the ultra short pulse are very important items to enhance development of the laser. As for the nuclear engineering by using laser, material processing is very widely applied, and an example is nuclear reactor repairing - this is a very important issue especially in Japan. Because after twenty years of operation, reactors require some maintenance. And laser isotope separation has been endeavored. Nowadays, enriched uranium is so surplus in the market that the development of this technology is meeting some difficulty. But science itself, technology itself - they are very interesting, and we are pursuing R and D of this technology. The laser nuclear fusion reactor is a distant goal, but this is also very important here in Japan. In the middle of the next century, this technology can provide a resolution for the energy problem for humankind. As with laser applied sciences; spectroscopy, especially concerning trace amounts of atoms, and laser micro-spectroscopy- Professor Letokhov discussed in much detail. This is a very interesting branch of fundamental research. And nanotechnology for chemical reaction control can be performed by an ultra short pulse laser; this can also have very wide applications in physical chemistry. Futher, a new type of laser - the X-ray laser is so popular now by virtue of high-power-laser pumping, and also the T-ray laser comes up to the topics; that is due to the nuclear excitation by laser electronic transition. These kinds of field, we must think about for the long-term program. Synchrotron orbital radiation and the free electron laser are very important tools, which expand the wavelength and also the power. They are much expected for material research as -334 — well as nuclear engineering. The diode-pumped lasers are highly efficient, and possibly to be very cheap. This topic was very enthusiastically discussed. They can have a strong power for the field of nuclear fusion also. The X-ray, UV, infrared, and far-infrared lasers in various ranges of wavelengths were compared. The prospects for these kinds of lasers are very hopeful. So as a summary of this meeting, Professor Schawlow gave a very interesting talk on the quantum nature of light and matter, and Professor Takuma presented a talk on diode-pumped solid state lasers. Basic laser science, laser development, laser material processing, and selective photoreactions - these were main topics of the symposium for four sessions where 34 papers were presented. At the poster session almost 80 papers were presented, and in total 121, scoping the present and the future application of lasers to nuclear engineering. One of the most important issues is the photon cost. This is a very familiar picture of the photon cost and quantum yield. As you see for chemical reactions, mercury lamps, C02 and KrF are indicated. The photon cost of laser is so high, that the only practical applications are isotope separation and pharmaceutical use. For these applications, the high photon cost is no problem. But for general industrial applications, photon costs should be almost ten or 100 times lower. This will be a very important issue in the future. Laser fields for research are expanding, and Professor Letokhov described this as an exponential increase. This is the markets for lasers and photoelectronics. The 1980 level of optical industry was about $1 billion and it will increase to $100 billion in 2000. This field will be a very strong industry of the world. I should say the next century will be the Photon Age, when the technology of main stream changes from electronics to photonics which has a very promising future. As you see, in 1980's laser applications extended widely, but mainly in the conventional fields of the compact disc, the printer, laser processing, medicine, and so on. But in 1990's the second generation started, and new laser applications utilize the energy of laser light in spatial and temporal applications. And a new technology begins to develop in various fields, such as quantum electronics, applications based on laser cooling techniques, biotechnology, laser mechatronics, and also in nuclear engineering. So these kinds of things should be envisioned for the next one or two decades; we must make preparations to get them to the society of the new century. Thank you very much for your cooperation and discussion. Now I would like to adjourn this session. - 335- Closing Remarks Proceedings of the 6ih International Symposium on Advanced Nuclear Energy Research -INNOVATIVE LASER TECHNOLOGIES IN NUCLEAR ENERGY- Closing Remarks for The 6th International Symposium on Advanced Nuclear Energy Research Tatsuo Kondo, Dept. Director General, JAERI Ladies and gentlemen, it is now time to close the 6th International Symposium on Advanced Nuclear Energy Research. I think that thanks to your cooperation, we had a pleasant and successful symposium,. The participants were many indeed, on the average about 300 per day, including 30 foreign guests. On the first day we listened to the lecture by Prof. Schawlow, who began with a review on the existing various kinds of lasers, and showed how these have contributed to the recent advances in the spectroscopy. Some novel applications of lasers were then described as well as experiments on the verification of quantum mechanics. I think his talk has impressed everyone in some way or another. Prof. Letokhovgave a lecture on laser methods for separating or detecting various nuclear species. He also talked about some potential applications of lasers to studies related to the nuclear technology. Since he has been an excellent pioneer in the field of laser science, I am sure that the directions pointed to by him are worth exploration. Prof. Takuma spoke of the progress in laser performance, which has been made possible by various advanced techniques. In particular, he explained how improvement has been achieved on the solid state lasers. His lecture convinced us that this kind of lasers will be further improved in the near future not only in their extreme capability but also in their usefulness for technological applications. In the afternoon of the 23rd, lectures been given about various basic investigations. Most of these were spectroscopic studies, and ranged from purely basic ones, such as those on laser cooling, nuclear physics, short-pulse lasers, and chemical-reaction dynamics to more application-oriented ones like LIF, PAS, and RIS spectroscopic studies on various subjects. In addition, physics of laser implosion was discussed. On the second day we had first the session for laser development, where presentations have been made on free electron lasers, high-power gas or solid lasers, and tunable lasers. We also listened to lectures on a soft-X-ray — 337- laser and that on the adaptive optics. I think these represented various important aspects of the recent progress in lasers that may be used for the nuclear technology or for other applications. In the next session we had presentations on the surface processing, surface modification, and the synthesis of functional materials. Though these new techniques may seem a little distant from the nuclear field at present, I presume it is quite possible that they may find good applications in the future. In the poster session there were 84 presentations covering various topics related to each oral session. I cannot summarize these in short, but I think many of you must have enjoyed fruitful discussion there. In Session V this morning, we had reviews and reports on selective photoreactions. Two of them were about laser application to the downstream chemistry, while the rest of them were on the atomic or the molecular laser isotope separation. I think the former is at the stage of promising proposal, while the latter is approaching to the stage of substantiation. We are aware of the problems of the latter, and I hope relevant difficulties will be overcome in the near future with the aid of advance in the laser technology. In the panel discussion just before, various interesting and suggestive remarks have been made on the future of lasers, laser sciences, and laser applications to the nuclear technology. As summarized by Prof. Yamanaka, we have many areas of researches to make further progress. To conclude, we had a good symposium "Innovative Laser Technologies in Nuclear Energy". I think many researchers in the nuclear field, including staff of JAERI, have learned much from this symposium. I am sure it will give rise to significant results from now on. Since this symposium is not regular about the topic, we do not know when we will have another meeting on laser. But, I hope to see you again someday. (Maybe when nuclear-powered laser or gamma-ray laser is selected as a topic? Anyway,) I again thank you very much for your cooperation. Last, some of you will join the technical to JAERI in Tokai, and there is an announcement from our staff later. Thank you and good-bye. — 338 — Author Index Author Index Fukumura.H. IV-pl Furukawa.H. ffla-pl CA] Abdulsabirov.R.Y . IDc-p8 CG] Abe,S. ffla-p3, ina-p4,ffla-p5 Gazeau.M.C. fflc-p6 Adachi.H. V-p5 Gonsiorowski.T. ffl-11 Agari,T. IDa-pl, IIIa-p2 Goto.N. HIc-pl5 Akaoka,K. V-p8, V-p9 Gries.W. Hb-p3 Akiyama.M. IDa-p8, IIIc-pl4 Aly.N. V-pll CH] Ando.K. ffl-10 Hajima.R. ffla-p8 Aoki.N. IIIb-pl Hara.T. ffl-10, ma-p8 Aoyagi.Y. ffl-10 Harashina.H. fflb-p6 Apollonov.V.V. ffl-4 Hasama.T. fflc-p7 Arai.S. V-pl4 Hasegawa.M. V-pl5 Araki.Y. V-p5 Hasegawa.S. Hb-p5 Arisawa.T. V-l.inb-p3(nib-p7,fflc-pl Hatanaka.k. IV-pl V-p6, V-p7.V-p8.V-p9 Hayashi.K. fflb-p2 Arlinghaus.F. n-8 Heitmann.U. Db-p3 Arondel.P. ffl-6 Hese,A. Db-p3 Asakawa,M. IHa-pl, IIIa-p2 HoIzrichterJ.F. ffl-5 Asakuma,T. ffla-pl. IIIa-p2 Horiguchi.T. na-p2,IIa-p3 Avril.R. Ilb-pl Hotoku.S. V-p3 [B] CI] Bekov.G.I. n-8 Ichikawa.S. 0a-p2 Bowers.M. ffl-7 Iguchi.T. Oc-pl Briand.A. IIb-p2 Iimura.H. IIa-p2 Brownell.M. mc-p4 Ikezoe,Y. 0a-p8 Imasaki.K. ffla-pl, ffla-p2 CC] Imatake.S. ffl-9 Chen.G. V-pl4 Inamura.T.T. n-2 Chen,J. IDa-pl, ma-p2 Inoue.N. nia-pl. ffla-p2 Chen,Y-W. V-pll, V-pl2 Iseki.Y. fflb-p2 Cho.S. IDa-plO, Illa-pll Ishige,Y. Ec-p2 Choppin.G.R. IIb-p6 Ishigure.K. IV-p3 Chu.M. V-pl4 Itoh.T. HIc-p8 Chu.S. D-l Iwasaki.T. nib-pio Clark.A. IIc-p5, IIc-p6 Izawa.Y. V-pl, V-pll, V-pl2 CD] CJ] Day.T. mc-p4 Jeong.Y. Ha-plO, Hla-pll Deas.R.M. Ic-p5, IIc-p6 Jing,Y. V-pl4 Decambox.P. n-3 Jyumonji.M. IV-p4 Dimock.D. Uc-p2 Doizi.D. ffl-6, fflc-p6 CK] Dubinskii.M.A. mc-p8 Kajiwara.T. H-5 Kakuta.T. nc-pi CE] Kanazawa.H. IHb-p6 Edamatsu.K. I0c-p8 Kasai.T. nib-pio EvansJ. IV-3 Katagiri.M. nc-pi Ewing,J.J. rn-7 Kato.M. DIb-p3. fflb-p7, nic-pl Kato.Y. Eb-p6 CF] Katsumura.Y. IV-p3 Feofilaktov.V.A. ffl-4 Kawai.T. IV-2 Feru.Ph. ffl-8, fflb-p9 Kawakami.S. V-8, fflb-p4 Fisher.D.L. Ila-pl Kawamura.W. V-pl Foster-Turner.R. IV-3 Kawamura.Y. DIa-p9 Fujii.T. IIc-p4, nic-p3, fflc-pl5 Kawanishi.S. IV-4, Da-plO Fujine.S. Ilb-p7, V-p3 Kawarasaki.Y. ffl-2 Fujisawa.G. n-n Kawasaki.M. IIc-p7, IV-p2 Fujita,M. ffla-pl, ffla-p2 Kearsley.A. IV-3 Keishi.T. IDa-p3, IIIa-p4, !Ha-p5, ffla-p6 Masuda.T. Db-p5 Kihara.T. nb-p7, V-p3 Masuhara.H. IV-pl Kikuzawa.N. ra-2 Matsuda.K. m-9 Kim,S.H. mc-p5 Matsuse.M. Dc-p7 Kim,S. ma-plO, Ula-pll Mauchien.P. n-3 Kimura.H. nib-pi Mead.R. m-7 Kimura.T. IIb-p6 Meinrath.G. Db-p6 Kishimoto.Y. Ua-pl Midorikawa.K. nc-pi2 Kislov.V.I. m-4 Mima.K. Ila-pl, ffia-pl, IIIa-p2 Kitamori.T. n-4 Minehara.EJ. m-2 Kitatani.F. IDa-p7 Mironov.S. V-3 Knowles.M. IV-3 Miyabe.M. V-p6,V-p8 Ko,D-K. mc-p5 Miyake.C. m-7 Kobayashi.A. ma-p3, IIIa-p5, IUa-p6 Miyamoto.Y. nib-p4,V-pl5 Kobayashi.N. Illb-pl Miyashita,A. IV-p5,IV-p6 Kobayashi.T. ffla-p8 Miyauchi.Y. HIa-p3 Koda.S. Ua-p9 Mori.M. HIb-p6 Koga.A. UIa-p3 Mori.Y. ma-p3 KogaJ. Ila-pl Morii.Y. HIa-p4, ma-p8 Koh.K. V-pll Morirniya.O. nib-p2 Koinuma.H- Uc-p7, IV-p2 Morimoto.K. V-p4 Koizumi.M. IIa-p3 Moriyama.N. IIa-p8 Kojima.T. IHb-p8 Moulin.C. n-3 Konagai.C. inb-pi Mourou.G.A. n-io Kondo.M. ra-9 Murakami.K. IV-p5,IV-p6 Kondo.S. IIIa-p8 Muraoka.K. n-5 Korableva.S.L. mc-p8 Kosmidis.C. IIc-p5, Dc-p6 CN] Kmpke.W.F. m-3 Nagai.A. HIa-p3 Krynetsky.B. V-3 Nagai.R. m-2 Kubodera.S. inc-pi2 Nagami.K. nia-pl Kubota.M. IIa-p2 Nagashima.T. nib-pio Kuga.Y. V-7 Nagata.Y. nic-pi2 Kukhtenko.A. V-3 Naito.N. nic-p7 Kuribayashi.S. m-9 Nakahara.Y. Ha-p2 Kuruma.S. IHa-pl Nakai.S. Hla-pl, EIa-p2, V-pll, V-pl2 Kusaba.M. V-pl Nakajima.T. nic-pi4 Kusama,H. rac-pio, nic-pii, uic-pi3 Nakamura.M. n-4 Kuwako.A. V-p5 Nakashima.N. V-pl Nakayama.T. fflc-pl CL] Nakazawa.M. De-pi Le Garrec.B- m-8, inb-p9 Naumov.A.K. nic-p8 Ledingham.K.W.D). Nemoto.K. nic-p3, nic-pis IIc-p5, Hc-p6 Niki,H. V-pll Lee.B. IDa-plO, Illa-pll Nishii.M. IV-4 Lee,J. IDa-plO, Illa-pl 1, IIIc-p5 Nishimura.A. Ha-p5 Lee.M.B. IV-p2 Nishimura.E. nia-p3, nia-p5 Letokhov/V.S. 1-2 Nittoh.K. V-p5 Li.DJ. IDa-p9 Noda.E. nib-p2 Liu J. V-pl4 Noda.O. m-9 Liu.Z. IIIc-p8 Nomaru.K. V-pll, V-pl2 Lompre,L.A. m-6, nic-p6 Notani.M. nc-pi Lowenthal.D. m-7 Numata.S. nib-pio Luecke.F. mc-p4 CO] CM] Oba,M. nib-p7, V-p7 Ma.P. V-pl4 Obara.M. nic-pi2 Maeda.M. n-5,IIb-p7, V-p3 Ogura.K. na-p4, IIa-p6 Majima.T. V-pl 3 Ohashi.H. nia-p8, nic-pl4 Malyavin.V.P. m-4 Ohba,H. V-plO Marshall.A. IIc-p6 Ohba,M. V-p8 Marshall.L.R. IIIb-pl2 Ohigashi.N. ffla-pl, HIa-p2 Ohishi.Y. IIIc-pl4 Suzuki.K. na-p8, V-pl6, Ohkubo.M. ffl-2 Suzuki.M. V-pl5 Ohmori.R. IIb-p5 Suzuki ,N. Ila-plO Ohno,S. Ha-p8 Suzuki.S. IIIb-p2 Ohtani.R. nib-pi Suzuki.Y. m-2,mb-p3, inc-p8 Ohyanagi.T. IV-p5, F/-p6 Ohzu.A- IIIb-p3 CT] Okada.N. IV-p3 Tajima.T. Ila-pl Okada.T. n-5 Takahashi.K. m-9 OnoeJ. V-7 Takai.H. IV-p4 Osa.A. IIa-p3 Takano.A. nc-p7 Oshima,Y. IIa-p9 Takao.M. m-2 Takashima.Y. m-9 CP] Takayanagi.T. n-ll,IIa-p7 Petit.A- n-3,nb-pi, nb-p2 Takehisa.K. nic-p2 Pietsch.W. IIb-p2 Takemori.S. nic-p2 Prokhorov.A. V-3 Takeuchi.K. V-7 Prokhorov.A.M. m-4 Takuma.K. 1-3 Tamura.K. V-p7 CR] Tamura.T. Db-p4 Radwan,J. Ilb-pl Tanabe.T. Hla-p9 Tanaka,K.A. n-6 CS] Taniu.Y. DIb-p6 Saeki.K. ma-p3, IQa-p5 Tashiro.H. V-7,fflb-p4, Inb-p5,nic-pl2, Sakamoto.N. IHa-pl, IHa-p2 IV-p4, Sarukura.N. IHc-p8 Tashiro.K. V-8 SasabeJ. ffl-2 Tidwell,S. m-7 Sasaki,S. ffl-2 Tomimasu.T. IIIa-p8 Sato.K. IIa-p7 Tomiyasu.H. V-2,V-p2, V-p4 Sato.S. m-9, HIa-p3, lDa-p5 Tomimasu.T. nia-p3, IIIa-p4, HIa-p5, ma-p6 Sato.Y. V-6 Tongu.E. 01a-p3 Sawada.T. n-4 Toyoda.K. ffla-p9, IIIc-pl2 Sawamura.M- m-2 Toyoda.K. IV-p4 Schawlow,A-L. 1-1 Tsuboi.Y. IV-pl Schneider,K.R. V-5 Tsuchiya.S. n-7 Schoknecht,G. Ub-p3 Tsunashima.S. 0a-p7 Segawa.Y. fflc-p8 Tsunawaki.Y. Illa-pl Sekine,T. IIa-p3 Tsunawaki.Y. 0Ia-p2 Semashko.V.V. IHc-p8 Shibata,T. H-9, Ila-p4, IIa-p6, V-plO Cu] Shiina.T. IIc-p2 Uchino.K. n-5 Shikazono.N. m-2 Uchiyama.G. V-p3 Shimamoto.K. m-9 Ueda.H. Eb-p4 Shimazaki.Y. V-8,IIIb-p4( V-pl5 Ueda,T. nia-p8 Shimizu.I. IIc-p4 Uehara.M. HIb-p6 Shimizu.Y. IV-4, Ila-plO Uesaka.M. HIa-p8 Shiraishi.H. Ua-p7 Umemoto.H. Ha-p7 Shu.S.H. IHa-p9 Singhal,R.P. Uc-p5 [V] Sinkov.S.I. IIb-p7 Vlasov.M. V-3 Smith.T.I. m-i Vors.E. Hb-pl Soga,T. Ua-p8 Stehle.M. IV-1 CW] Suda.A. IIIb-p5 Wada,Y. V-p4 Sugimoto.M. m-2 Wakaida.I. V-p6, V-pB, V-p9 Sugimoto.S. IV-4 Watanabe.1. HIb-p2 Sugioka.K. IV-p4 Watanabe,T. V-p5 Sugita.K. V-pl3 Whitaker.TJ. n-8 Sugiyama.A. mc-pl Wirth,A. ffl-11 Sumiya.M. Hc-p7 Wu.B. V-pl4 Suto.O. V-8 Wu,X-W. Ha-p9 Suzuki, A. IIb-p5 m Yagi.T. inc-p9, fflc-plO, IIIc-pl 1, mc-pl3 Yamagishi.H. IIc-pl Yamaguchi.H. V-8 Yamaguchi.S. IIIb-p6 Yamamoto.T. Illa-pl, IIIa-p2 Yamamura.T. V-p2 Yamanaka.C. IHa-pl, IHa-p2, V-pl, V-pl 1, V-pl2 Yamauchi.T. Hc-p2 . Yasui.K. IIIb-p8 Yoda,0. IV-p5, IV-p6 Yoguchi.I. V-p5 Yokoyama.A- 11-11 Yokoyama.K. EI-11, Ua-p7 Yoshida.M. IIb-p4 Yoshida.T. V-p5 Yoshida.Y. IIIa-p8 Yoshida.Z. Ilb-p6 Yoshimoto.M. IV-p2 CZ] Zakou.A- IIIa-p3 Zhang.Z. mc-p9, lDc-pll Zhidkov.A. V-3 Keyword Index Keyword Index CD] Decommissioning m-9 [A] Decomposition V-pl3 Ablation processing IIIb-p6 Decontamination rv-i Actinides II-3 Deformable mirror IIIc-pl5 Adaptive optics Iilc-pl5 Delay time IIIc-plO Adhesiveness IV-4 Density matrix equation V-p9, Vp-15 Advanced reprocessing V-p4 Detector IIc-pl Ammonia IIa-p8 Detuning effects V-p8 Ammonia laser V-pl6 Diamond like carbon nia-p7 Amplifier Illb-pl, IIIc-pl Dichloroethenes IIa-p7 Analysis n-3 Diffraction pattern IIc-p4 Analytical system IIb-p4 Diode-pumped solid-state laser Angular momentum V-p5 IIIb-pl2 Anharmonic-splitting V-pl5 Direct simulation Monte Carlo method ASE fflc-pl, HIc-p2 IIa-p5 Atom trapping II-l Dissociation IIIb-p2 Atomic lead IIb-p5 DL beam quality V-6 Atomic vapor laser isotope separation DL wavelength stability V-6 IIa-p5, V-p6, V-p9 DLC IIIa-p7 V-pl2 Dye laser V-6, IIIc-p2,111-6, IV-3 Autoionizing transition V-p8 Dye oscillator-amplifier system IIIc-p7 CB3 Beam-forming IIIc-pl5 (E) Boron isotopes V-pl6 Electric discharge IIlc-pll Branching ratio V-p6 Electro-static accelerators Brewster angle IIc-p7 IIIa-p9 Electron beam evaporation [CD V-plO C-13 enrichment V-pl4 Electron injector ffla-p3 Carbon cluster IV-p5 Electrostatic accelerator Illa-plO Carbon monoxide laser IIIb-p6 Elemental analysis II-8 Carbonate complexation IIb-p6 Emission spectrum IIb-p2, IIb-p6 CCD IIc-p2, V-plO Energy transfer IIa-p5, V-p2 Charge transfer Ha-p4 Enrichment factor V-8, V-pl5 Chirp-pulse V-pl2 Epitaxial thin films IV-p2 Chirped Pulse Amplification ESA IEc-pl 11-11 Evaporation IIa-p5 Cluster ion IIa-p8 Evaporation surface V-plO Cm CD) IIb-p6 Excimer laser rv-l, IV4,na-p9, CO laser ra-9 Ila-plO, IUb-p6, fflc-plO, C02 laser n-ll, IIa-p8,nib-p5 rv-p4 Coherent synchrotron radiation Excitation V-pl5 ffla-p2 Excited uranyl ion V-p2 Cold plasma IIIc-pl2 Experiment IIIa-p8 Compact X-ray laser 111-10 Explosives IIc-p5 Continuous wave nib-p8 Extraction redox process V-3 Copper Vaper Laser IV-3, nib-pl, HIb-p2 IIIb-p3, IV-p3 CF] Ila-pl Corrosion resistance IV-p3 Feedback system Cr(CO) IV-p3 Femtosecond laser mc-pll 6 IV-2 Cross-section V-p6 Ferroelectric films Cutting ra-9, nib-p6 Fluorescence II-3 CVL V-6, IIIc-pl Fluoropolymer IV-4 CW Q-switch IIIb-p7 Free Electron Laser ni-i Frequency chirping Cyclohexanol na-p9 IIIc-p3 Ionization V-p5 Frequency control lllc-pl IR FEL llla-pl Frequency doubling IV-3, HIb-pl2 ISOL (Isotope Separator On-Line) Frequency tuning mechanism IIa-p3 IIlc-p5 Isotope control IIb-p2 Fusion ID-5 Isotope selectivity V-pl4 Isotope separation IIa-p3, III-6, EIb-p5, (G) IIlc-p3 Gadolinium IIa-p4, Ha-p5, V-p6, Isotope shift V-p8, V-pll V-plO, V-pll Isotopic analysis II-8 Gain IIIc-pl Gas centrifuge V-5 CJ) Gas temperature IIIb-p2 J-value V-p6 Gas-dynamic laser m-4 JFT-2M IIc-p2 Glazing m^.IV-l.ffla-pS, Jitter IIlc-plO IIIa-p4. IIIa-p5, nia-p6, IIIa-p8, ffla-p9, IDa-plO, CK) Illa-pll Klystron IIIa-p4 Grid pulser IIIa-p3 KrF IIIc-pll Group velocity dispersion V-pl2 CD Lj, m absorption edge IV-p6 CH) Lanthanide ion V-pl, V-p2 Halogenated hydrocarbon Lanthanides II-3 n-n Laser ablation IIb-p2,IV-2,IV-p5, He-Ne laser llc-p7 1V-P6 Hg lamp V-p4 Laser beam V-p9 High brightness Ulb-plO Laser beat wave lla-pl High efficiency IIlb-p3 Laser cooling IM High electrical conductivity Laser CVD IV-p3 IV-p2 Laser diode Ec-p7, ffl-8, nib-p7, High performance optics HIa-p7 IIIb-pl2, High power IIIb-p8 Laser diode arrays III-3 High-lying state V-p6 Laser dyes IIIc-p6 Human whole blood samples Laser excited atomic fluorescence spectrometry IIb-p3 IIb-p3 Hydration number IIb-p6 Laser implant-deposition IV-p4 Hydrogen IIIb-p2 Laser induced photoacoustic spectroscopy Hydrogen bromide IIIb-p3 IIb-p7 Hydrogenated amorphous silicon thin film Laser ionisation IIc-p5 IIc-p7 Laser irradiation V-3 Hydrolysis IIb-p6 Laser isotope separation V-5, IIa-p4, IIa-p6, Hydroxy acid Ila-plO Hb-pl, mb-p4, IIIc-pl5, Hyperfine structure IIa-p2, V-p8, V-pll V-p7, V-pll, V-pl3, V-pl5 Image intensifier IIc-p2 Laser manipulation IIc-p4 In situ monitoring IIc-p7 Laser materials III-3 Inertial confinement III-5 Laser plasma IIIc-pl3 Infrared 1E-1, IIIa-p8 Laser pumping IIIc-p6 Infrared multiphoton dissociation Laser spectroscopy V-p5 11-11, V-pl3, V-pl4, Laser-induced reaction II-7 V-pl5 Lasers III-5 Intracavity Illb-p8 Li-like Al line 111-10 Ion collection IIa-p6 Lifetime V-6, IIb-p6, V-p6 Ion mobility spectrometry Light irradiation V-p4 IIc-p5 Lightning Illc-pll Ion source IIa-p3 Linear accelerator HIa-p3, IEa-p8 Ion-induced nucleation V-7 Liquid ablation dynamics IV-pl Liquid fuel III-4 Optical fiber IIb-p7, Illb-plO Long pulse IIIa-p4 Optical parameric oscillator Longitudinal pumping IIIc-p2 IIIb-pl2 Low temperature IIIb-p3 Optical-field-induced ionization Luminescence spectroscopy IIIc-pl2 IV-2 Oscillator Illb-pl, nic-pl Oscillator strength V-p6 CM] Maleate Ila-plO CP] Mass spectroscopy IIc-p6 P-polarized light IIc-p7 Material processing rv-3, nib-p6, nib-pio Palladium V-3 Matrix effect IIb-p4 Para-H2 fflb-p4 Maxwell's equation V-p9 Partial oxidation IIa-p9 Metastable states IIa-p5 PAS ffla-p7 Micro-optics m-3 Permanent magnet IIIa-p6 Mo(CO)6 IV-p3 Permanent-magnet helical undulator Modelocked oscillations Ma-p9 Illa-pll Modulator IIIa-p4 Phase space structure Ila-pl Molecular Beam Eitaxy IV-2 Photo-acoustics spectroscopy Molecular laser isotope separation IIla-p7 V-7, V-8 Photo-excitation V-p5 Molecular reaction dynamics Photocathode Illa-pl II-7 Photochemical ablation IV:pl MOPA V-6 Photochemical reduction V-p3 Multichannel quantum defect theory Photochemical technology IIb-p5 V-p4 Multiphoton ionization IIa-p7 Photochemical valency adjustment Mutual separation V-p4 V-p4 Photodissociation V-l, IIa-p7, Hc-p6, CN] Photoelectron trigger IIIc-plO Nanosecond photography IV-pl Photofragmentation translation spectroscopy Narrow linewidth IIIb-p5 11-11 Nd:YAG ID-8, IIIb-p8, IIIb-pl2 Photoionization V-l, IIa-p4, V-p6 Nd:YLF IIIb-pl2 Photon V-p4 Near-resonant V-pl2 Photoorganic synthesis Ila-plO Neodymium IIa-p4 Photorefractive effect IIIc-pl4 Neptunium V-p3, V-p4 Photothermal ablation IV-pl Neutron IIc-pl Picosecond III-l Ni-63 ionization IIc-p5 Planck's law V-plO Nitrobenzene Hc-p6 Plasma IIa-p6, IIc-p2 Nitrotoluene IIc-p6 Plasma chemical vapor deposition NMR V-2 IIc-p7 Non-destructive inspection Plutonium V-p4 IIIc-pl3 Polarization V-p5 Non-Hamiltonian manipulation Postionization II-8 Ila-pl Potassium titanyl phosphate Non-invasive elimination IIc-p4 IIIb-p8 Nonlinear optics IIIb-pl2 Pressure tuning IIIc-pl Nonlinear thomson scattering Propagation V-p9 ffla-pl Pulse compression V-pl2 Nuclear fuel reprocessing IIb-p7 Pulse-train laser 111-10 Nuclear moments IIa-p2 Pulsed laser deposition IV-p2 Nuclear pumping IIc-pl Nuclear structure IIa-p2 CQ] Quasi-CW III-2 CO] Quenching V-p2 Optical cavity IIIa-p5 Optical chaos nic-pi4 Optical damage IIIa-p7 CR] Subpicosecond HIc-p7 Radiation intensity V-plO Superconducting rf linac 1II-2 Radiation pressure IIc-p4 Superconductor IV-2 Radio-Frequency IIa-p6 Supersonic helium gas bombardment Radioisotope IIc-p4 1V-P6 Raman laser IIIb-p4 Surface modification IV-4, IV-p4 Rapid beam cooling Ila-pl Surface processing IV-1 Rate equation IIIc-pl Surface treatment IIlb-p6 Reacceleration IIIa-p3 Reaction V-2 CT] Reaction mechanism Ila-pl 0 Tandem mass spectrometer Recombining plasma 111-10 IIa-p8 Redox reaction V-p4 TEA-C02 laser IIIb-p4, V-pl3 Reprocessing V-p3 Temperature profile V-plO Resonance ionization II-8, Ha-p3 Terawatt XeCl Laser system Resonance ionization spectroscopy IIIc-p7 IIb-p5 Thomson scattering IIc-p2 Resonance photoionization Ti:sapphire laser V-p7 lla-p6 Time-resolved II-3 RF Hla-p4 Time-resolved measurement RIS V-p6 IV-p5 Room-temperature operation Time-resolved spectroscopy lE-9 IV-pl Rubidium V-p7 Titanium nitride IV-p2 Ruby laser IIc-p2 Titanium sapphire IV-3 Trace element detection of selenium CS] IIb-p3 S-type beam transport system Tracer level IIb-p4 Hla-p3 Translation IIa-p5 Second harmonic generation Transport system IIIa-p5 1D-8, nib-p8 Transuranium V-p4 Selective ablation IV-1 Transverse pumping IIIc-p2 Selective reaction V-l Trapped Particles Ik-p4 Selectivity V-l, Ila-plO TRLFS IIb-p6 Self-focusing V-p9 Tunable laser IIIb-p5, inb-p7, IIIc-p3 Semiconduction film llc-p7 Tunable operation I1I-9 "idmiconductor laser IIlc-p3 Two-dimensional IIIc-pl Separation V-p3, V-p4 Two-photon reduction V-pl Shape and size of particles IIc-p4 [U] Shock Hardening 1V-1 U(V) IIb-p6 Si deposition IV-p4 UF6 V-8,inb-p4 Si doping IV-p4 Ultrafast phenomenna Illc-p9 Simulation V-p9 Ultrafine particle V-7 Single mode IIIc-pl Ultrahigh-peak-power lasers Slab HIb-p7 U-11 Slab laser Ulb-plO Ultrashort-pulse lasers U-11 Small particles Hc-p4 Ultraviolet laser pulse UIc-p9 Smoke particles Hc-p4 Undulator Ila-pl, Illa-plO Sodium vapor V-p9 Unidirectional ring oscillator Soft x-ray laser HIc-pl2 IIIc-pl4 Solid-state laser IIIb-p7, Ulb-p8, fflb-pl2 Uranium V-2, IIb-p2, V-p6 Spatio-temporal structure IIIc-pl4 Uranium enrichment V-5 Speciation Hb-p6 Uranium hexafluoride V-7, V-pl5 Spectral width HIa-p3 Uranium pentafluoride V-7 Stainless steel !V-p4 Uranium spectroscopy Ilb-pl State-to-state reaction II-7 Uranium vapor V-p9 Stochasic cooling Ila-pl User facility UI-1 (V) Vertical undulator IIIa-p6 Vibrational modes V-pl5 VUV spectral region IIb-p3 CW] Waveguide FEL IIIa-p2 Wavelength limitation IIIa-p3 Wedge prism IIIc-p5 Welding IIIb-p6 Wettability IV-4 CX] X-ray mc-pl3 X-ray absorption spectroscopy !V-p6 X-ray laser m-10, nic-pl2 X-ray sources Illa-pl XAFS lV-p5 CY] YAG HIb-p6