The Photodisintegration of Helium-3 at Photon Energies
THE PHOTODISINTEGRATION OF HELIUM-3
AT PHOTON ENERGIES OF 8.06 AND 9.17 MEV
JACK ROBERT MACDONALD
B.A.Sc•, The University of British Columbia, 1960
A THESIS SUBMITTED IN PARTIAL FULFILMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
in the Department of
PHYSICS
We accept this thesis as conforming to the
required standard
THE UNIVERSITY OF BRITISH COLUMBIA
September 1964 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of
British Columbia, I agree that the Library shall make it freely available for reference and study, I further agree that per• mission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that,copying or publi• cation of this thesis for financial gain shall not be allowed without my written permission.-,
Department of
The University of British Columbia,
Vancouver 8f Canada
Date OCT^g^ ~7 , (^4 The University of British Columbia
FACULTY OF GRADUATE STUDIES
PROGRAMME OF THE
FINAL ORAL EXAMINATION
FOB, THE DEGREE OF
DOCTOR OF PHILOSOPHY
of
JACK ROBERT MACDONALD
B„A„Sccj The University of British Columbia* 1960
TUESDAY, OCTOBER 6, 1964, AT 3;-00 P.M.
ROOM 12, HEBB BUILDING, (PHYSICS)
COMMITTEE IN CHARGE
Chairman; I. McT. Cowan
W. Mo Armstrong D,„L..Livesey K. L» Erdman Jo Mo McMillan C. Froese. J. Bo Warren
External Examiner: Leon Katz
Department of Physics
University of Saskatchewan THE PHOTODISINTEGBATION OF HELIUM-3 AT PHOTON ENERGIES OF 8.06 AM) 9,17 MEV.
ABSTRACT
The cross section for the photodisintegration of helium-3 has been measured at. gamma ray energies of
8,06 and 9.17 mev. The He3 ( section at 8,06 and 9.17 mev was found t.o be 0,493 t 0.066 and 0,723 t 0,087 millibarns respectively. The He3 ,n)2p reaction cross section at 9,17 mev was found to be 0.25 i 0,13 millibarns. These results are compared with other experimental work on the photodisintegration of helium-3 and tritium. The photodisintegration reaction was observed in a cylindrical gridded ionization chamber using a helium- 3^ methane, and argon gas mixture. The, C^3(ps y )_t\jl4 reactions at proton bombarding energies of 0,554 and 1.75 mev were used as the source, of gamma rays of well defined energy. The preparation of carbon-13 targets is discussed in detail. Theoretical calculations on the photodisintegration of mass 3 nuclei are summarized, Photodisintegration and electron scattering measurements are compared as, methods of determining the nature of the ground state wave function of the mass 3 system. GRADUATE STUDIES Field of Study: Physics Electromagnetic Theory G.M, Volkoff Waves .J.G, Savage Nuclear Physics J.B. Warren Elementary Quantum. Mechanics W, Opechowski Nuclear Reactions BtL, White Solid State Physics R, Barrie Special Relativity Theory W, Opechowski Introduction to Low Temperature Physics D.V, Osborne Group Theory Methods~ W. Opechowski Theoretical Nuclear Physics P,D, Kunz Advanced Quantum Mechanics P, Rastall Related Studies: Numerical Analys is T.E,. Hull Electronic Instrumentation F.K, Bowers PUBLICATIONS Gas Flow Regulator for an rf Ion Source, Rev. Sci. Inst. 33, 1111 (1962) Removal of Tritium from Helium-3, Rev. Sci. Inst. 34, 1280 (1963) Photodisintegration of Helium-3, Bull. Amer. Phys,, Soc. 8, 124 (1963) Photodisintegration of Helium-*3 near the threshold, Phys. Rev. 132, 1691 (1963) Simple Gas Circulation Pump, Rev. Sci. Inst. 35, 241 (1964) Simple Electron Bombardment Apparatus for Evaporating Boron, Rev. Sci. Inst. 35, 122 (1964) -i- ABSTRACT The cross section for the photodisintegration of helium-3 has been measured at gamma ray energies of 8.06 and 9.17 mev. The He (£,p)D reaction cross section at 8.06 and 9.17 mev was found to be 0.493 ± 0.066 and 0.723 ± 0.087 millibarns respectively. The He (tf,n)2p reaction cross section at 9.17 mev was found to be 0.25 ± 0.13 millibarns. These results are compared with other experimental work on the photodisintegration of helium-3 and tritium. The photodisintegration reaction was observed in a cylindrical gridded ionization chamber using a helium-3, methane, and argon gas mixture. The C13(p,^)N14 reactions at proton bombarding energies of 0.554 and 1.75 mev were used as the source of gamma rays of well defined energy. The preparation of carbon-13 targets is discussed in detail. Theoretical calculations on the photodisintegration of mass 3 nuclei are summarized. Photodisintegration and electron scattering measurements are compared as methods of determining the nature of the ground state wave function of the mass 3 system. -ii- TABLE OF CONTENTS Page Chapter I - INTRODUCTION 1 Chapter II - THEORETICAL ESTIMATES OF THE CROSS SECTION FOR THE PHOTODISINTEGRATION OF HELIUM-3 ... 4 A. Reaction Kinematics 4 B. Theoretical Estimates of the Cross Section 5 Chapter III - PHOTODISINTEGRATION OF 3-BODY NUCLEI- EXPERIMENTAL RESULTS 10 Chapter IY - APPARATUS AND EXPERIMENTAL ARRANGEMENT FOR THE PHOTODISINTEGRATION MEASUREMENT 17 A. Introduction 17 B. Cylindrical Gridded Ionization Chamber 17 1. Chamber Operation 17 2. Chamber Construction 18 3. Gas Mixtures 20 C. Energy Calibration of the Chamber 23 D. Chamber Background 27 E. Pulse Amplification and Analysis 29 F. Gramma Flux Measurement 30 G. Experimental Arrangement 31 Chapter V - PHOTODISINTEGRATION RESULTS AND ESTIMATION OF THE CROSS SECTION AT 8.06 AND 9.17 MEV.. 33 A. Spectrum at 8.06 mev 33 B. Spectrum at 9.17 mev - 2-body Breakup 33 C. Spectrum at 9.17 mev - 3-body Breakup 34 D. Analysis of the 2-body Data 35 E. Cross Section Calculation for the Reaction He3()r,p)D 36 -iii- Page Chapter Y (cont) F. Errors in the 2-body Cross Section 36 G. Estimation of the He3()T,n)2p Cross Section 38 Chapter 71 - DISCUSSION OF EXPERIMENTAL MEASUREMENTS AND THEORETICAL CALCULATIONS ON THE 3-NUCLEON SYSTEM 39 A. Photodisintegration of Helium-3 39 B. Theoretical Cross Section Calculations 42 C. Electron Scattering on Tritium and Helium-3 43 D. Conclusions 45 Appendix A - CROSS SECTION CALCULATIONS 47 A. Chamber Efficiency 47 B. Atom Density of Helium-3 49 C. Wall Effect 50 1. Wall Loss 50 2. Spectrum Shape due to the Wall Loss 51 3. 2-body Photodisintegration Spectrum Analysis .... 53 Appendix B - THE PREPARATION OF CARBON-13 TARGETS 56 Appendix C - THE REACTION C13(p,tt)N14 AND THE MEASURE• MENT OF THE GAMMA FLUX V. 60 A. Gamma Ray Counter Efficiency 60 B. The Reaction C13(p,tt )N14 as a Source of Gamma Rays 61 • 1. Introduction 61 2. 0.554 mev Resonance 62 3. 1.75 mev Resonance 63 4. Doppler Shift 64 C. Analysis of the Gamma Spectra 65 1. 0.554 mev Resonance 65 -iv- Appendix C (cont) Page 2. 1.75 mev Resonance 67 Appendix D - NEUTRON INDUCED BACKGROUND AND NEUTRON SHIELDING 69 Appendix E - PHOTODISINTEGRATION OE HE3 NEAR THE THRESHOLD by J. B. Warren, K. L. Erdman, L. P. Robertson, D. A. Axen, and J. R. MacDonald reprinted from The Physical Review, 132, 1691 (1963) following p.72 Appendix F - GAS FLOW REGULATOR FOR AN rf ION SOURCE, by B. L. White, L. P. Robertson, K. L. Erdman, and I. R. MacDonald reprinted from The Review of Scientific Instruments, 33, 1111 (1962) following p.72 Appendix G - REMOVAL OF TRITIUM FROM HE3 by K. L. Erdman, L. P. Robertson, D. Axen, and J. R. MacDonald reprinted from The Review of Scientific Instruments, 34, 1280(1963).. following p.72 Appendix H - SIMPLE GAS CIRCULATION PUMP by K. L. Erdman J. R. MacDonald, G. A. Beer, and D. A. Axen reprinted from The Review of Scientific Instruments, 35, 241 (1964) .... following p.72 Appendix I - SIMPLE ELECTRON BOMBARDMENT APPARATUS FOR EVAPORATING BORON by K. L. Erdman, D. Axen J. R. MacDonald, and L. P. Robertson reprinted from The Review of Scientific Instruments, 35, 122 (1964) .... following p.72 Appendix J - PHOTODISINTEGRATION OF ARG0N-40 AT 9.17 AND 17.7 MEV by M. A. Reimann, J. R. MacDonald, and I. B. Warren, preprint of article to be published in Nuclear Physics, (1964) following p.72 Bibliography 73 -v- LIST OF FIGURES to follow page 1. Energy Distribution for the Sum of the Energies of the Two Protons for the Reaction He3^ ,n)2p ...... 5 2. Theoretical Estimates of the Cross Section for the Reaction He3( t ,p)D 7 3. Theoretical Estimates of the Cross Section for the Reaction He^Of ,n)2p 7 4. He3(t ,p)D and T(t ,n)D Cross Sections from Eichmann (1963) 8 5. He3()f ,p)D Differential Cross Section at 90° in the Laboratory System 11 6. Cross Sections for the Reactions He3(Y*p)D and He3()f ,n)2p from Gorbunov and Varfolomeev (1964) .... 14 7. Ionization Chamber 18 8. Details of Collector Support 19 9. Electrode Voltage Supply 19 10. Charged Particle Ranges in Gases 20 11. Effect of Purification on Voltage Pulse Amplitude ... 24 12. Chamber Energy Calibration 27 13. Chamber Background 29 14. Experimental Arrangement 31 15. Photodisintegration Spectrum, 8.06 mev Gamma Rays ... 33 16. Photodisintegration Spectrum for the Reaction He3(y,p)D at a Gamma Ray Energy of 9.17 mev 33 17. Photodisintegration Spectrum for the Reactions He3( X ,n)2p and A40( T ,«*)S35 at a Gamma Ray Energy of 9.17 mev 34 18. Cross Section for the Reactions He3(V,p)D and T(tf,n)D 36 -vi- to .follow page 19. Angular Distribution of the Reaction Products for the Reaction He3(}p,p)D .' 39 20. Comparison of Total and Differential Cross Section Measurements 41 21. H0(a) vs. d and H2(d)/H0(d) vs. d 49 22. Wall Loss Spectra for the Reaction He3(tt ,p)D 52 23. Hypothetical Spectrum Shape due to Wall Loss 53 24. Target Preparation Apparatus 57 25. Yield of 9.17 mev Gamma Rays as a Function of Proton Energy 58 26. Gamma Counter Efficiency 60 27. Decay Schemes for the 8.06 and 9.17 mev Levels of N^-4 63 28. Gamma Spectrum from the Reaction C13(p, tt )N14 at the 0.554 mev Resonance 65 29. 6.14 mev Gamma Spectrum from the Reaction F19(p,e* )f)016 at the 340 kev Resonance 66 30. Gamma Spectrum from the Reaction C13(p,tt)N14 at the 1.75 mev Resonance 67 31. Cross Section for the Reaction He3(n,p)T 69 32. Beam Collimators and Neutron Shielding 70 -vii- LIST 0? TABLES to follow page I 90° Differential Cross, Section for He3()f,p)D from Bennan, Koester and Smith (1964) ... in text p.11 II 90° Differential Cross Section for He3(2T,p)D from Stewart, Morrison and O^Connell (1964) in text p.13 III Cross Section for the Reaction T(tf,n)D .. in text p.16 IV Chamber Dimensions 18 V Chamber Fillings in text p.23 VI Alpha Particle Backgrounds in text p.28 VII Data from Photodisintegration Spectra for He3()T,p)D 35 VIII He3()f,p)D Cross Section Data 36 IX Estimate of Errors 36 X Cross Section Errors in text p.37 XI Data for 3-body Cross Section at 9.17 mev 38 XII Measured Angular Distribution of the 9.17 mev Radiation at the 1.75 mev Resonance in text p.63 XIII Angular Distributions, Relative Intensities, and. Scintillation Counter Efficiencies .... to follow p.68 XIV (p,n) Thresholds to follow p.71 -viii- AC KNOWLEDGEMENTS I wish to express my sincere gratitude to Dr. K. L. Erdman for his kind supervision of this research projeot and his willing guidance throughout my research career. The assistance of Dr. J. B. Warren, Dr. B. L. White, Dr. G. M. Griffiths, Dr. G. Jones and Dr. M. McMillan proved invaluable and was greatly appreciated. I gratefully acknowledge the assistance of Dr. L. Robertson, Mr. D. Axen, Mr. M. Reimann and Mr. D. Healey in operating the Van de Graaff and in performing the measurements. Words fail to express my gratitude to my mother, my father and my wife for their endless patience and constant enc ouragement. I am deeply indebted to the National Research Council for the four scholarships held during the course of this work. -1- CHAPTER I Introduction Nuclear photodlsintegration reactions involve the interaction of the electromagnetic field of a photon with the charges, currents and electric and magnetic moments of a nucleus, which in turn are calculated from the wave functions describing the nuclear system. As the interaction of electro• magnetic fields with charge systems is well understood, the calculation of the radiative matrix elements and hence the reaction probability or cross section only involves assumptions concerning the initial and final state wave functions of the system. Furthermore, the knowledge gained about the wave functions through photodlsintegration experiments can in prin• ciple be used to obtain information about the nuclear forces. The comparison between theory and experiment for the photodlsintegration of the deuteron served to test the basic theory of photonuclear reactions. Accounts of experimental data on the deuteron photo-effect can be found in Wilkinson et al (1952) and Hulthen and Sugawara (1957). The agreement between experiment and theory is very good, largely because the deuteron is loosely bound. This allows the deuteron wave function to be well approximated by its known asymptotic form. Consideration of the photo-effect in the 3-nucleon system leads to a considerable difference in the methods of -2- theoretical analysis. Firstly, the 3-nucleon system is known to be tightly bound so that asymptotic wave functions may well be less reliable than in the case of the deuteron. Secondly, the wave functions for the ground state of the nucleus cannot be easily calculated from a potential. As a first approximation, the wave functions used to date have generally been chosen to have an analytical form which Is integrable. Usually these wave functions are justified by assuming a potential and calculating the binding energy of the system by a variational method. Photo• disintegration cross sections for the 3-nucleon system have been calculated with some of these assumed wave functions. The cross sections are very sensitive to the assumed form of wave function. 1 The need for accurate cross section measurements of the photo-effect for a 3-nucleon, system is obvious. Such measurements would supplement the experimental data of elastic scattering of electrons from tritium and helium-3 and provide the most sensitive test for 3-body wave functions. One's hope is that these measurements will give a more precise determina• tion of both initial and final state 3-nucleon wave functions and perhaps yield information about the nucleon-nucleon inter• action, if only as evidenced by the final state interaction. There would seem to be too many variables in the problem to allow one to make definite statements about a possible 3-body force. In view of the need for such measurements, an experi• ment to measure the photodisintegration cross section for -3- helium-3 was begun at U.B.C. as early as 1958. The first measurements on helium-3 were reported by Warren et al (1963). This thesis is concerned with the extension of the earlier work to higher photon energies and a preliminary investigation of the 3 particle breakup of helium-3. Theoretical estimates of the cross section for the photodisintegration of the 3-nucleon system are discussed in Chapter II. A resume of the experimental work done in other laboratories on the photo-effect in hellum-3 and tritium is given in Chapter III. The U.B.C. measurements are discussed in detail in Chapters IY and Y and Appendix E. A comparison between theory and experiment Is given in Chapter YI. -4- CHAPTER II Theoretical Estimates of the Cross Section for the Photodlsintegration of Helium-5 A. Reaction Kinematics For gamma ray energies below the pion production threshold, the possible photodlsintegration reactions of hellum-3 are 3 He + * —• p "D He3^ —• P + P+ n The Q-values for the 2 and 3-body breakup, Q2 and Qg, calculated from the mass values given by Everling et al (I960) are Q2 = -5.4-9316 ± O.OOOB3 ttev QB - - 7. 7! 9 ± O.OOJ mev If the momentum of the photon is neglected, the proton and deuteron in the 2-body breakup will have well defined fractions of the total energy E0 = (E^ -5.493) mev For the 3-body photodlsintegration, a continuum energy spectrum will be observed if only one or two of the reaction products are detected. The sum of the energies of the two protons, Eot), extends from a minimum energy of EQ^n= 1/3 (E<£- 7.719) mev to a maximum energy of -5- max Eob = (E^, - 7.719) mev. The minimum observed energy corres• ponds to the two protons emitted in the same direction with equal energies. The maximum observed energy corresponds to the two protons being emitted in opposite directions with the neutron having zero momentum. The observed energy distribution will depend on the density of final states of the reaction, and hence is propor• tional to the volume in phase space associated with the momen• tum distribution of the three particles. The probability of a given momentum distribution is in turn dependent upon the nuclear and coulomb interactions of the three particles. Robertson (1963) derives the shape of the energy distribution for the two protons in the 3-body brealaip neglecting final state interactions. Figure 1 shows this energy distribution. Deviations from this spectrum shape can be attributed to final state interactions. B. Theoretical Estimates of the Gross Section There have been five calculations on the photodisin• tegration of the 3-nucleon systems, tritium and helium-3. The calculations of Verde (1950), Gunn and Irving (1951) and Rossetti (1959) are quite similar and differ only in the choice of the form of the ground state wave function. In each case, the assumed ground state wave function was an S-state, totally symmetrical in the spatial co-ordinates of all three nucleons. (It i£[ known, from magnetic moment data, that the totally symmetric S-state component of the ground state wave function O-S l-O U5 SUM OF -me EMERSY OF THE TWO PROTONS RW* At GM-U-IA «*Y ENERGY OF 9-17 MEV FIGURE 1 Energy Distribution for the Sum of the Energies of the Two Protons for the Reaction He3( Jf ,n ) 2p -6- is dominant.) Final state interactions between the photodisin• tegration products were neglected. Furthermore, in the treatment of helium-3, the coulomb interaction between the two protons was treated as a perturbation, the triton and helium-3 wave functions thus being identical in zeroth order. Of course a coulomb barrier correction must be applied to the final cross section for helium-3. Verde (1950) and Gunn and Irving (1951) calculated the 2 and 3-body cross sections assuming a Gaussian wave function for the ground state of the triton: where 1T\: is the distance between the ifcJl and J**1 nucleons in the nucleus, and JLK is a size parameter found by equating the difference in binding energy of tritium and helium-3, 0.76 mev, to the coulomb energy of helium-3. This form of wave function is analytically attractive; however, the asymptotic form falls off faster than the correct solution. The final state wave function for 2-body breakup was taken as the product of the - 1 part of a plane wave for the nucleon and a pure S-state Gaussian function for the deuteron. That is, s y o< [ i^l» - cos kr3] expOolJ -7- where k is the momentum of the nucleon, r3 is the position co-ordinate of the nucleon and rlg is the relative co-ordinate of the proton and neutron in the deuteron. This calculation considers only the electric dipole o transition from the Sj. ground state to a continuum P-state. s Yerde shows that the magnetic dipole transition for the 2-body breakup is forbidden on the basis of symmetry arguments. For 3-body breakup, the magnetic dipole transition is not forbidden and Yerde gives an expression for the cross section using two plane waves for the final state wave functions. The cross sections for 2 and 3-body breakup are shown in Figures 2 and 3. Gunn and Irving (1951) have also evaluated the 2 and 3-body electric dipole photodlsintegration cross sections using ground state wave functions having a more suitable asymptotic behaviour. These wave functions are of the form This class of wave functions is known as Irving functions. Gunn and Irving evaluated (T(2-body) and ^(3-body) for n 5=1 1/2 although n = 1/4 gives the best value for the binding energy and coulomb energy difference for the 3-body system. Rossetti (1959) has evaluated (^(2-body) for n-0. In both cases the final state was assumed to be the Jt^l part of a PHOTOM EMERGV , Et (MEV) Figure 2 Theoretical Estimates of the Cross Section for the Reaction He ()f , p) D VERDE (,l950) ~ GAUSSIAN WAVE FUMCTIOM 3n .-GUNN AND IRVING (V9S0" IRVLKL<» WAVE FUNCTION '/yx - 2.€> F (T (mb) 20 25 3o PHOTON EMERGY (MEV) FIGURE 3 'Theoretical Estimates of the Gross Section for the Reaction He3(V , n) 2p -8- plane wave and the deuteron wave function was of the form These results are also given in Figure 2. Eichmann (1963) has Improved the 2-body calculations by- considering quadrupole contributions to the cross section, by including a non-symmetrical component S' in the ground state wave function, and by Including final state interactions between the outgoing particle and the deuteron, Eichmann takes the symmetric part of the ground state wave function as a sum of z two terms of the form exp \\-/A { r\ +• r4| + r2|)] and chooses the parameters yj^ from a variational calculation of the binding and coulomb energies of tritium and helium-3. The mixed symmetry state S' is added with an amplitude ratio of % - O, \ , consistent with the neutron absorption by deuterium at low energies (Austern, 1.951, 1952). The final state wave function is taken as the product of wave functions for the deuteron ground state (the sum of two Gaussians) and the outgoing nucleon. The nucleon wave function Is assumed to be the Jl- 1 part of a plane wave in one calculation. The effect of final state interaction is calculated for the triton in a second calculation, where the interaction is calculated using a Serber potential with a Gaussian shape. The effects of including the mixed symmetry component, the quadrupole contribution and the final state interaction are shown in Figure 4. The mixed symmetry component contributes to 0.9- o.e- ' / V% 0.7- / / 0.6- (rr,b) 0-5- // MeBl*,p)D "^^0 0.4- // El (S-^P) "^O^ 03- // El CS-*-P) + E2 (S-*-D) "^r^ // El (.S+S'-^P) + E2 (S-t-S'-^D) 0.2- 0.1- 5.5 10 15 v io 25 30 PHOTON ENERGY (rAE\l) I.O-i / 0.9- / I ^ _ \ 0.9- 1 S ^N. \ W 06- 0.5- If 0.4- / ; T(t,n)D NN^ 0.3- 1 El (S-*-P) 0.2- jl £| (S-*- P) + FINAL STATE INTERACTIONS 0.1- (o'.zto \'o ife 20 2'S 3b PHOTOM ENERGY (MEV) FIGURE 4 He3()T , p)D and ?(t > n)D Cross Sections from Eichmann (1963) -9- the electric dipole transitions (S* —>• P) and lowers the energy of the cross section maximum by 10$, increasing the magnitude of the cross section by 10%. The electric quadrupole contribution has a dramatic effect on the > total . cross section as is shown in Figure 4, even for channel energies of 15 mev. The final state interaction increases the maximum cross section for the photodisintegration of tritium by 25% and causes the cross section curve to fall more rapidly above the maximum. Delves (1960, 1962) has calculated the 2-body cross section for helium-3 at low energies and the 3-body cross section for tritium including the effects of final state inter• actions in the latter case. These calculations are not applic• able to the present work and will not be discussed here. The theoretical calculations and experimental results for the photodisintegration of helium-3 will be compared in Chapter VI. -10- CBAPTER III Photodisintegration of 5-body Nuclei Experimental Results Apart from the U.B.C. results, there have been four experimental measurements reported on the photo-effect in helium-3 and one on tritium. Each of these will be discussed in some detail with particular emphasis on the experimental technique. Berman et al (1963, 1964) measured the 90° differen• tial cross section for the 2-body breakup at photon energies from 8.5 to 22 mev. The experiment was done with the brems- strahlung photon beam of a 22 mev betatron incident on a helium-3 gas target. Protons and deuterons from the 2-body breakup were counted in coincidence using two CsI(Tl) scintil• lators as detectors. Because the photon momentum is small at these energies, the proton and deuteron are emitted at approximately 180° to each other, with nearly equal momenta. Consequently, the energy of the proton, is twice that of the deuteron. This constitutes a unique signature for a 2-body event and also determines the energy of the photon which caused the event. 2-body events are thus easily separated from 3-body events for which no such angular correlation or kinematical restrictions apply. The photon flux per unit energy interval was determined by monitoring the total energy in the photon beam and assuming a brem^strahlung spectrum shape. -11- The results obtained by Berman et al are shown in Figure 5 and tabulated in Table I. TABLE I 90° Differential Cross Section for He3(^ ,p)D from Berman, Koester and Smith (1964) Gamma Ray Energy (mev) steradian (0^)90° 9 80.0 5.0 10 90.8 5.7 11 91.6 5.9 12 88.8 6.0 13 84.5 6.1 14 78.2 5.9 15 77.1 6.5 16 6,4.8 + 6.0 17 59.9 ± 5.8 18 59.7 ± 6.0 19 57.6 ± 6.2 20 57.5 ± 6.7 21 42.7 + 6.6 STEWART ET AL , YALE BERMAN ET AL. ILLIWOIS FlMCKH ET AL, HEIDELBERG .09 STeroduqn •06 •05-\ •o4- •cs- •02.- •01H -1——1 1 1 1 1 1 1 r- ~i 1 r —r~ -< 1— 10 15 20 25 30 GAMMA RAY ENERGY (MEV) FIGURE 5 He3()T , p)D Differential Cross Section at 90° in the Laboratory System -12- The errors quoted are certainly most optimistic in view of the continuous photon energy spectrum produced by the betatron. In addition to the 2-body data, Berman et al observed 3-body breakup in the cases where the two protons were emitted at 90° and 180° to each other. Although an attempt is made to extract a cross section for 3-body breakup from the data, the assumptions made are most likely incorrect, and we shall not reproduce the data here. Finckh et al (1963) also measured the 90° differential cross section for the 2-body breakup. Their technique was almost identical to that of Berman et al. A 31 mev bremsstrahlung beam was used to irradiate a helium-3 gas target. The reaction products were detected by CsI(Tl) scintillation counters. The absolute value of the cross section was determined by comparing the He ( )f ,p)D yield to the known yield from the reaction C12()f jnJC11. In this manner, uncertainties in the bremsstrahlung spectrum shape were eliminated. The results are shown in Figure 5. The estimated uncertainty in the absolute value of the cross section amounts to about • 1: 12%. Stewart et al (1964) have also measured the differ• ential cross section at 90° in the laboratory system for the 2-body, breakup of helium-3. Helium-3 contained in a cell was irradiated with, a 40 mev bremsstrahlung beam produced by an electron linear accelerator. The reaction products emitted at 90° to the photon beam passed through a thin window in the cell and were momentum selected by a quadrupole magnet. The magnet -13- focussed the particles onto two solid state transmission counters arranged to stop the deuterons in the first counter and protons in the second counter. The energy spectrum of the photons was determined by replacing the helium-3 with deuterium and normalizing the He (^JP)!* yield to the known D()f ,p)n yield. The results are given in Figure 5 and tabulated in Table II. TABLE II 90° Differential Cross Section for He3(tf ,p)D from Stewart, Morrison and O'Connell (1964) Gramma Ray Energy (mev) VdJl/90° /steradian 8.9 ± 0.4 83 13 10.3 ± 0.5 90 11 12.6 ± 0.8 87 ± 13 13.6 ± 1.0 90 ± 11 14.4 ± 1.2 85 ± 10 15.6 ± 1.3 74 ± 8 16.8 ± 1.5 70 ± 9 18.8 ± 1.8 64 ± 8 20.0 ± 2.0 59 ± 7 23.0 2.5 50 ± 6 27.7 ± 3.3 32 + 4 34.3 ± 4.3 19 ± 4 41.0 5.5 18 ± 8 46.1 ± 6.1 10 7 -14- Due to momentum considerations, the number of deuterons counted by the silicon detector should be equal to the number of protons produced in the 2-body breakup of hellum-3. Thus any "excess" protons could be attributed to 3-body breakup. Stewart et al measured the number of 3-body events as a function of proton energy and compared their results with a theoretical calculation using the Irving wave function. Becchi et al (1964) have used a 30 mev bremsstrahlung photon beam and solid state particle detectors to measure the 90° differential cross section for the 2-body breakup of helium-3. Their results suffer from poor statistics and are reported in terms of a total cross section, in spite of the fact that this is not what was measured. No information is tendered as to the method of determining the bremsstrahlung spectrum shape. Certainly, more information about the experimental technique is required before the measurements can be taken seriously. The work of Gorbunov and Varfolomeev (1963, 1964) together with the U.B.C. results represent the only measurements of the total cross section for the 2 and 3-body breakup of helium-3. Gorbunov and Varfolomeev used a cloud chamber contain• ing helium-3 and a bremsstrahlung beam with a maximum energy of 170 mev. Each reaction in the chamber was photographed and identified by its klnematical features. The total cross sections, 0"(2-body) and 6. The total yields for the two processes were found to be equal. The angular distribution of the 2-body \o 2o 3o 4o 50 60 70 80 9o loo GftMMA RAY EMERGY (t^6V) FIGURE 6 Gross Sections for the Reactions He3( )f, p)D and He3(Y , n)2p from Gorbunov and Varfolomeev (1964) -15- reactlon products was found to be I (e) = sin^e L I + (o.fc^io.io)cos© + (o.4-fc ± o,io) cos*e] 4 (0.03 ± o.oi) where all 2-body events are included. Experimental values of the integrated cross section were obtained. The integrated cross section due to electric dipole absorption can be written oo ( <£(2-body) =• (2G.5 ± W*) rnev-nob <£(3-boc|y) - (43. The E2 contribution to (T[2-body) is estimated to be 11 ± 4%. The total cross section for the 2-body photodlsinte• gration of tritium has been measured by Bosch et al (1964) using monoenergetic gamma rays from (n, tt) reactions. A tritium gas target was irradiated with gamma rays and the photo- neutrons were detected by a long counter. The cross section at gamma ray energies of 6.7, 7.6 and 9 mev was determined by comparing the yield from the T( tt,n)D reaction with the known yield from the D(tt ,p)n reaction. The assumption was made that -16- th e photoneutrons from tritium, like those from deuterium, display a Sin2© angular distribution. At these energies, this assumption is valid as electric dipole absorption will dominate. The results are given in Table III. TABLE III Cross Section for the Reaction T(tf",n)D Gamma Ray Energy (mev) Cross Section (mb) 6.7 .075 ± .012 7.6 .295 ± .035 9.0 .57 ± .07 -17- CHAPTER IV Apparatus and Experimental Arrangement for the Photodisintegration Measurement. A. Introduction The measurement of the total cross section for the reaction He3(}f,p)D at photon energies of 8.06 and 9.17 mev is an extension of previous work by Warren et al (Appendix E) and Robertson (1963) who measured the cross section at photon energies of 6.14, 6.97 and 7.08 mev. A brief description of the apparatus and experimental arrangement which Is described in Robertson (1963) is included here. B» Cylindrical Gridded Ionization Chamber 1. Chamber Operation When charged particles pass through a gas, they lose energy by ionizing the atoms of gas along their path. The 1 electron-positive ion pairs that are formed can be separated by an electric field and collected by the electrodes which define the field. A knowledge of the amount of charge collected by the electrodes and the energy lost by the ionizing particle per ion pair produced in the chamber gas leads to a measurement of the energy of the particle. In a pulse Ionization chamber, one electrode is connected to the voltage supply via a high impedance. As the ions are moving toward, and being collected by the -18- eleotrbde, a voltage pulse will be generated whose ampli• tude is a measure of the energy of the ionizing particle. Because the electron mobility in gases is of the order of 1000 times greater than that of heavy ions, the signal is measured at the positive electrode. To reduce further the collection time of the charge and hence the rise time of the pulse, a third electrode, a screening grid, can be placed between the positive and negative electrodes. This grid shields the positive electrode from the induction charge pulse of the electrons until the electrons have passed, through the grid and eliminates the induction charge pulse of the positive ions. Bunemann et al (1949) discuss the effects of such a grid in a parallel plate chamber and Robertson (1963) discusses the case of a cylindrical chamber. 2. Chamber Construetlon The chamber is shown in Figure 7 and the important dimensions are summarized in Table IT. The cylindrical wall, A, was made of 0.25 inch thick Alcan 65ST6 aluminum tubing, turned eccentrically on a lathe to reduce the thickness to 0.050 inches on one side. The seals between the cylinder and the end plates, B, were made using Kel-F 0-rlngs. Aluminum rings, C, were used to clamp the cylinder to the end plates. Field defining guard rings, E and F, were supported by insulated electrical terminals. Lava discs held the grid supports, G. The grid was made of FILLING SYSTEM FIGURE 7 Ionization Chamber TABLE IT Chamber Dimensions Dimension Inside radius of cylinder wall 5.72 ± .05 cm Radius of collector 0.0507 ± .0006 cm Mean radius of grid 0.385 ± .005 cm Diameter of grid wires 0.0058 cm Number of grid wires 15 Inner guard ring (radius to centre of ring) 1.75 cm Outer guard ring radius 3.65 cm Distance between grid supports, length of active volume 14.4 * .025 cm Inside length of chamber 21.5 cm Approximate total volume of chamber 2.2 liters Total volume in active region between guard rings 1480 ± 15 cc Volume of electrode support as a fraction of Vtot 2.20% Volume enclosed by grid as a fraction of V. . 0.45% tot Active volume 1455 ± 15 cc Measured chamber capacity 10 ± 2 pF Chamber capacity and amplifier input capacity 18 i 3 pF -19- 15 strands of equally spaced steel wire 0.004 in. diameter. The collector, J, made of 0,025 in.diameter German silver tubing, was supported at one end by a Kovar seal, M, attached to the electrode support structure and by a Kovar seal, P, at the other end. A glass bead, N, separated the collector from its supporting Kovar seal at one end, which was connected to the same voltage as the collector in order to eliminate electrical noise due to tracking across the glass bead. The complete inner structure was mounted on a s~teel frame, D, and centred within the cylinder. The cylinder wall, the end plates, and the guard rings were shielded from the positive electrode by a Mylar film, not shown in the dia• gram. The Mylar film was effective in reducing the chamber background due to natural radioaotivity in the structural material. The end of the collector from which the signal was taken was connected to P, which in turn was insulated from ground by another glass section, 0, Details of the signal end of the collector are shown in Figure 8. The electrode voltage supply is shown in Figure 9. A purifier containing a Ca-Mg eutectic mixture (Colli, 1952) was attached to the chamber at H* The gas was circulated through the purifier by a solenoid driven piston pump (Appendix H). The purifier removed hydrogen, oxygen, nitrogen, carbon dioxide and water vapour impurities which were detrimental to the operation of the chamber. FIGURE 8 Details of Collector Support F — — -0075y*F COLLECTOR |o M51 10 Mil E.H.T SUPPLY wvw vVWV SUPPORT M 0 TO 5.5 KV TEST PULSE INPUT COLLECTOR CASCODE INPUT STAGE OF SHIELD DYNATRON PREMA9L\F\ER 10 MSI 2pF 50 pF COLLECTOR GRID 2.5 M51 J_ 6KV DCWV IMNER FIELD I .02 /4 DEFINING' X 2 MSI RINGS I OUTER FIELD DEFINING :02/ FIGURE 9 Electrode Voltage Supply V0 = O. \GG \/GRvD -20- 3. Gas Mixtures The gas mixtures used in the chamber had to satisfy three criteria. a) The gas mixture had to be dense enough so that the reaction particle ranges were appreciably shorter than the diameter of the chamber. If this were not so, all the particles produced in the reaction would lose a portion of their energy in collision with the wall and consequently the voltage pulse height would not be indicative of the energy of the particle. The ranges of protons and deuterons in various gases are shown in Figure 10. It is obvious that helium-3 alone would not suffice to stop the reaction products which have energies of from 1 to 2 mev. For this reason, argon, a gas with a higher stopping power, was added. For such a mixture of gases, the range, R^^* is approximately given by where P^, Pg...are the partial pressures of the gases in atmospheres and R^, R2,...are the ranges in the gases at N.T.P. b) The gamma flux produces a large flux of electrons through the chamber. These electrons are a result of Compton, pair production, and photoelectric interactions between the photons and the material FIGURE 10 Charged Particle Ranges in Gases at 15° C and 76 cm Pressure. From Whaling (1956) -21- surrounding the chamber and the gas contained In the chamber. Because of the low stopping cross section of the electrons compared to heavy oharged particles, these electrons lose relatively little of their energy within the active volume of the chamber. However, if many electrons traverse the chamber within the resolving time of the electronics, large voltage pulses can occur due to the pile-up of small pulses. It is therefore important to keep both the rise and fall times of the pulses as small as possible. To reduce the collection time and the resolving time of the apparatus, methane was added to the argon and helium-3 . For a given collector voltage and gas pressure the methane reduced the rise time of the pulses by a factor of two. The requirements of a high stopping power for heavy charged particles and a low stopping power for electrons cannot be simultaneously satisfied. However, with the faster rise times achieved with the addition of methane, the increase of electron noise with higher gas pressures could be tolerated. e) The gas mixture chosen had to produce satisfactory energy resolution. This criterion set a lower limit of 3.2 microseconds on the amplifier differ• entiation time constant and an upper limit on the total pressure in the chamber. -22- The total amount of gas which could be contained in the chamber was limited by two factors. First, the thin wall of the chamber could only withstand pressures up to 10 atmospheres, an obvious upper limit on the total pressure. However, at these high pressures, the ion density along an ionization track becomes so great that recombination of the electron-positive ion pairs may take place. In a uniform electric field this effect results in a lower pulse height for a fixed energy ionizing particle. In a non-uniform field, such as that which exists in a cylindrical chamber, the recombination effect is greater in low field regions near the wall with a resulting spread in pulse heights, and a subsequently poorer energy resolution. The amount of recombination which takes place is a function of the specific ionization of the ionizing particle, the mixture of gases in the chamber and the ratio of the electric field to total gas pressure. In order to keep the recom• bination to a minimum for the proton and deuteron tracks in the chamber, the total pressure was limited to a value of 8 atmospheres. Consideration of the foregoing criteria led to the chamber fillings given in Table V. -23- TABLE V Chamber fillings. Gas pressures in Atmospheres at 0°C Helium-3 Methane Argon 8.06 mev run 2.24 0.218 5.32 9.17 mev run (2-body 1.515 0.251 5.08 breakup) 9.17 mev run (3-body 1.41 0.19 1.36 breakup) C. Energy Calibration of the Chamber The rate of energy loss from a charged particle in a gaseous media has three components, one due to ionization, one due to excitation and one due to elastic interaction. The total rate of energy loss is given by d* ax dx u cU where n^, nm and nq refer respectively to ionizing, exciting and elastic collisions with energy coefficients i, m and q.. The average energy lost by the charged particle through all processes per ion pair formed is -24- For a given gas or mixture of gases,W is nearly independent of the nature of the charged particle or its velocity. However, values of w vary considerably for different gases or mixtures of gases. The inclusion in a gas of a contaminant able to destroy metastable states by an ionizing collision (for example, argon in neon) will increase n^ at the expense of nm and thus decrease the value of w. For a given value of w and a given number of ion pairs, the voltage pulse height is determined by the fraction of the electrons which are ultimately collected. When all the electrons are collected, saturation has been achieved. In general, two processes inhibit complete collection. If electronegative atoms or molecules are present in the ionization chamber, electrons may attach to them and the heavy negative ion thus formed migrates slowly toward the collector and is removed from the fast component of the voltage pulse. For a given gas, the probability of electron attachment depends on the energy with which the electron strikes the electronegative atom or molecule. Roughly speaking, this probability is decreased as the electron velocity increases; that is, as the electric field increases, or the gas pressure decreases. Figure 11 shows a graph of voltage pulse height CIA versus collector voltage for 5.3 mev alpha particles from PQ . Curve "A" was taken from data before the mixture was purified, hence with a probable admixture of electronegative gases. Curve "B" is taken from data after purification and indicates that saturation is much more complete. AO Q BEFORE PUR\FVI^G _ - looo 2ooo 3ooo 4ooo COLLECTOR VOLTAGE (VOLTS) FIGURE 11 Effect of Purification on Voltage Pulse Amplitude -25- The second process which inhibits complete collection is columnar recombination of electrons and positive ions along an ionization track. The amount of recombination is a function of the density of ion pairs along the track and the magnitude of the electric field. For a given field, the ion density depends on the total pressure and on the specific ionization of the charged particle. Since the specific ionization is different for protons, deuterons and alpha particles, it is possible to achieve complete collection for proton and deuteron tracks while having incomplete collection for alpha particle tracks. Such an effect has been observed in the chamber used in this experiment. The Q-value for the photodisintegration of argon-40 into an alpha particle and a sulphur-36 nucleus Is -6.764 mev. This reaction has been observed at U.B.C, (Appendix 3"), at a gamma ray energy of 9.17 mev. The argon was placed in a gridded cylindrical ionization chamber and irradiated with gamma rays from the proton bombardment of car- bon-13 (Appendix C). The known energy of the alpha particles, 2.41 mev, agreed very well with the energy of the alpha group in the speotrum as established in terms of a plutonium 239, 5,15 mev alpha souroe. Because: pure argon was used, complete collection of the electrons was achieved at usable values of collector voltage. As Robertson (1963) shows, the addition of methane and/or helium to argon increases the colleotor voltage required to achieve saturation. For the gas mixtures used in -26- the helium-3 experiment, oomplete collection for the alpha particles was not achieved. Because argon was used as a stopping gas in the chamber, the photoalphas from argon were observed during the photodisintegration of helium-3 experiment. The energy of the alpha group was predicted correctly by the 5.3 mev Po210 calibration source over a wide range of gas pressures. However, the energy of the proton-deuteron group from the photodisinte• gration of helium-3 was not predicted correctly at all pressures. The agreement was reasonable, though not exact, for the low pressure runs of Warren et al, whereas at the pressures used in the present experiment the 3.67 mev photodisintegration group was found at an energy which would be interpreted as 210 being 5.0 mev as predicted by the P0 calibration source in the chamber. Although saturation is not achieved at high pressures for the alpha tracks, saturation is much more complete for the proton-deuteron tracks. That is, a greater percentage of electrons is collected. At lower pressures, collection is more complete for the alpha tracks and consequently the agree• ment in the energy scale between the alpha particles and the protons and deuterons is better. When collection in the chamber is not complete, a given voltage pulse height corresponds to different energies, depend• ing upon whether the ionizing partiole was an alpha particle or a proton. For this reason, the energy scale for the photodis• integration of helium-3 was determined using .the reaction -27- He (n,p)T. The voltage pulse amplitude from a pulse generator (Robertson, 1957) was calibrated in terms of the thermal neutron capture peak at 0.765 mev. The pulse generator was then used to determine the energy scale for each run. A further check was provided by Irradiating the chamber with 7.12 mev gamma rays from the proton bombardment of fluorine-19. The position of the thermal neutron peak, the 7.12 photodisintegration peak and the 9.17 mev photodisintegration peak are shown in Figure 12. The energy scale is linear over a wide range of energies. The position of the 5.3 mev o< peak is also shown and the disagree• ment in the two energy scales is obvious. D. Chamber Background One of the problems associated with a low count rate experiment such as the photodisintegration of helium-3 is that of background. If the time taken to accumulate sufficient photodisintegration data is long, a considerable number of counts will be contributed to the spectrum by the natural alpha activity from the impurities in commercially available materials. Table VI, taken from Sharp (1955), lists the alpha particle backgrounds from various materials. For the work of Warren et al, the interior of the chamber was coated with an aqueous dispersion of graphite known as aquadag. The background in the chamber of total area 1000 cm3 compared favourably with the figure listed in Table VI. Furthermore, the majority of counts were in the region of 4 to 6 ENERGY SCALE FOR PROTONS AND DEUTERONS (wev) FIGURE 12 Chamber Energy Calibration -28- TABLE VI Alpha Particle Backgrounds p Material Alphas per cm per hour of energy greater than 250 kev Machined Copper 0.09 Commercial Brass 0.05 Mild Steel 0.03 Commercial Aluminum 0.31 Solder 28 Aquadag (Graphite) 0.07 Lead 60 2 Air from room 32 per 100 cm per hour Cylinder argon 0 mev, well above the energy of the photodisintegration products and background was no problem. For the higher energy photo• disintegration measurements, the energy region from 4 to 6 mev for alpha particles corresponded to a proton-deuteron energy of from 2.7 to 4 mev. As the photodisintegration products have an energy of 3.67 mev for 9.17 mev gamma rays, it was necessary to reduce the chamber background considerably. The technique which eventually proved satisfactory ?/as to line the chamber with a polyester film produced by du Pont under the trade name "Mylar". This is a tough, transparent -29- material with a low vapour pressure, composed of 62.5% carbon, 4.2% hydrogen and 33.3% oxygen by weight. A thickness of —3 ? 3.6(10) cm or 5 mg/cnr was found sufficient to stop a 7 mev alpha particle. The Mylar was coated with a thin film of gold, evaporated on both surfaoes, to render the Mylar conducting. This was necessary so that the Mylar did not accumulate charge and distort the field within the chamber. The success of this technique is illustrated in Figure 13 which shows the effect of inserting the Mylar film. The background with the Mylar was 0.02 alphas per cm2 per hour, uniformly distributed over an energy range from 0 to 7 mev. E. Pulse Amplification and Analysis The voltage pulses appearing at the collector of the chamber were amplified by a Dynatron Preamplifier Unit and a Dynatron Main Amplifier Type 1430A. This amplification system provided a wide range of integration and differentiation time constants for pulse shaping. The pulse shape was chosen to provide optimum energy resolution while keeping the pulse width as short as possible to reduce electron pile-up. The integra• tion time constant was set at 8.0 microseconds and the differentiation time constant at 3.2 microseconds for all of the experimental data. The pulses from the main amplifier were fed to a Nuclear Data Model ND103, 256 channel pulse height analyzer for analysis and storage. The time taken for the kicksorter (pulse height analyzer) to record each pulse is ('25+ ) microseconds NUMBER OF COUNTS (ARBITRARY SCALE") U H . 1 1 I I I I 1 I I L_ -30- where N is the channel in which the pulse Is stored. The total time during which fche analyzer will not accept a pulse is referred to as "dead time". If the dead time is an appreciable fraction of the total time of the run, a correction must be made to account for those pulses which were not analyzed. The kicksorter records the dead time so that the true number of pulses N which appeared at the kicksorter input is given by N - -I-N' T-t where T is the total time of the run, t is the dead time and N' is the number of pulses accepted by the kicksorter. During this experiment, the dead time never exceeded 2% of the total time. F» G^mma Flux Measurement The integrated gamma flux was measured with a 2§ inch diameter by 4^ inch long Nal(Tl) scintillation counter. The measurement of the efficiency for this crystal is described in Appendix C. The output of the photomultiplier was fed through a cathode follower to a Nuclear Data Dual Amplifier Model ND501. This amplifier also served as a single channel analyzer. The discriminators of the single channel analyzer were set so that St• all pulses corresponding to -ry mev energy release in the crystal were fed to a Model U.B.C.-NP-ll scalar and counted. Pulses from a pulse generator (Robertson, 1957), fed Into the cathode -31- follower of the counter, were used to set the discriminators. The pulse generator voltage level was calibrated in terms of gamma ray energy release in the crystal by using the full energy peak in the 2.62 mev gamma ray spectrum of a Ra Th source. The gamma ray speotrum was continuously recorded by feeding the output of the ND501 amplifier into a Nuclear Data Model ND 120, 512 ohannel pulse height analyzer. In this manner, a constant cheok was maintained on the bias level, and the speotrum shape was continuously monitored. Typical gamma spectra are shown in Appendix C. The calculation of the gamma flux through the chambers in terms of the gamma flux recorded by the solntillatlon counter is discussed in Appendix A. G. Experimental Arrangement The experimental arrangement is shown in Figure 14. The target and Ionization chamber were surrounded by a castle of wax and cadmium to shield the chamber from neutrons. The castle wall consisted of 6 inches of wax and 0.15 inch thick cadmium sheathing. The only dlreot path into the castle was the If inoh diameter hole for the beam tube. The chamber was supported In an adjustable rack whioh sat on the bottom of the castle. The carbon-13 targets were prepared by thermally crack• ing methyl iodide onto .002 Inch thick platinum backings. The technique is discussed in Appendix B, The platinum target WATER COOLING IONIZATION. CHAMBER WAX CADMIUM COLD TRAP -5 ^777 ///' PROTON BEAM PYREX BEAM TUBE FROM VAN DE GRAAFF /// / ^/— V"//' TAvRGET FIGURE 14 Experimental Arrangement -32- backing was soldered with indium onto the target support which consisted of a loop of 3/16 inch copper tubing. The target was cooled by passing water through the copper tubing,and beam currents up to 30 microamps produced no deterioration of the target. The scintillation counter was placed outside the castle wall in a position such that the only material between the counter and target was the castle wall itself. The percentage of gamma rays absorbed in the castle wall was determined by measuring the difference in gamma flux per microcoulomb of proton current with and without the.oastle. The measured trans• mission factors for 8.06 and 9.17 mev gamma rays agreed with those estimated from the gamma ray attenuation coefficients for paraffin and cadmium. -33- CHAPTER V Photodisintegration Results and Estimation of the Cross Section at 8.06 and 9.17 mev A. Spectrum at 8.06 mev The spectrum obtained for the reaction He ,p)D at a gamma ray energy of 8.06 mev is shown in Figure 15. The deter• mination of the energy scale is discussed in Chapter IY. The background shown in Figure 15 and used in the cross section analysis was entirely due to residual radioactivity in the walls of the chamber. When the helium-3 in the chamber was replaced by helium-4, the spectrum in the region of Interest was the same, with and without a gamma flux present. This was the case for both 8.06 and 9.17 mev gamma rays and justifies the use of the time dependent (gamma ray independent) background in the 2-body breakup cross section analysis. B. Spectrum at 9.17 mev - 2rbpdy Breakup A typical spectrum obtained at a gamma ray energy of 9.17 mev is shown in Figure 16. The energy scale applies only to the photodisintegration of helium-3 as discussed in Chapter IY« The background is again due to residual radioactivity in the chamber and is much lower than for the 8.06 mev run because of the effectiveness of the Mylar shield (see Chapter IY), which was inserted after the 8.06 mev run was completed. © © PHOTODISINTEGRATION SPECTRUM 4- • • TIME DEPENDENT BACKGROUND 3 x x Ra- Be NEUTRON SOURCE 2oo- I60- CHAMBER FILLING (PRESSURES !4o- IN ATMOSPHERES AT 0°C) h 4 z 120- Me3 - 2.24- D 3 O C H4. " 0. 218 U loo- 2-j ARGON - 5.32 U- o 8o- io H or 6o- IU CO 6 40- Z 4- 20- ~~r~ —r 1— 1 1— 1 o to 2o 30 4-0 50 "io" 70 KICKSORTER CHANNEL r -1 o 3 4- ENERGY T(MEV ) z FIGURE 15 Photodisintegration Spectrum, 8.06 mev Gamma Rays -O PHOTO DISINTEGRATION! SPECTRUM • TIME DEPENDENT BACKGROUND X X Ra-Be NE.UTROM SOURCE IOO- 9o ^«L,TIOO 80- CMAhABER FILLING (PRESSURES ATMOSPHERES AT O ° C) 70- 6 He3 - 1.515 50- ARGoN ~ S.08 40- l \ \ \ 30- i \ k T 2o- I I lo- T T 1 T —I— 4rO 50 6o —r— 80 VO 20 30 CHKNNIEL, KlCXSORTER TO T T" r 2 3 4- o ENERGY (MEV) FIGURE 16 Photodisintegration Spectrum for the Reaction He3(Y, p)D at a Gamma Ray Energy of 9.17 mev -34- C. Speotrum at 9.17 mev - 5-body Breakup The 2-body photodisintegration data was obtained with a total gas pressure of 6.846 atmospheres (at 0°C). The high pressure was required to reduce the wall loss to a reasonable value. As a result of this high gas density, the gamma ray induced electron spectrum extended to 2 mev. The 3-body breakup spectrum, which extends from 0.49 to 1.47 mev, was thus totally obscured by the electron noise. In order to see the 3-body spectrum, the total gas pressure was reduced to 2.96 atmospheres. Figure 17 shows the 3-body spectrum obtained for the chamber filling indicated. The electron end point has been reduced to 0.57 mev as a result of the lower total pressure and because the gamma flux was decreased to reduce pile-up. The dominant contribution to the background in the region of the photodisintegration spectrum was due to neutron induced disintegrations (Appendix D). The background shown in Figure 17 was obtained with the proton beam on the carbon-13 target but off resonance, so that the 9.17 mev gamma flux was negligible. As the neutron induced back• ground was critically dependent on the proton beam energy, half the background was run 12 kev above the resonance at 1.75 mev and the other half at 12 kev below resonance. Events from the 2-body breakup of helium-3 can also be seen in Figure 17. However, the wall loss for the chamber filling used was 76% so there is no well-defined full energy peak in the spectrum. As a consequence of the lower gas pressure, the photo-alpha group from argon-40 can also be seen -o PHOTOOIS INTEGRAT\ON SPECTRUM + 25 _ . BEA\M DEPENDENT BMIKGROUND 50- CHAMBER FILLING (PRESSURES IN 4o- ATMOSPHERES AT 0° C) He* - 1.4-1 I- ? 30 CH4. - 0.19 o ARGON - \«B6 20- He3(Y,p)D ^ - - —(5 *~ — • - - •* - — -1— —1 1 1—« o 10 20 30 40 50 Go 70 So 9o KlCKSORTER CHKNNEL T "T~ T 2. ENERGY (MEV) FIGURE 17 Photodisintegration Spectrum for the Reactions He3()f , n)2p and A40(ft,cx) S36 at a Gamma Ray Energy of 9.17 mev -35- at an energy correctly predicted by the thermal neutron capture peak (see Chapter IY). D« Analysis of the 2-body Data The data for the 8.06 and 9.17 mev runs is summarized in Table VII. The gamma flux was calculated as shown in Appen• dix C, and correction was made for absorption In the wax and cadmium shield where applicable. Run 2 at 9.17 mev was made without the shield in place. The absorption in the aluminum chamber wall was calculated to be 1% in all cases. The photodisintegration spectra show a full energy peak at E0 = (Ey. - 5.493) mev and a low energy tail. This low energy tail is due to those events in which one of the reaction products strikes the wall or passes into the end region of the chamber. This effect is termed "wall loss" and is a function of the size and shape of the chamber and the range of the reaction products. The total yield is the sum of the number of counts In the full energy peak and the number of counts in the wall loss tail. The method of determining the total yield is explained in detail in Appendix A. Briefly, the total yield is equal to the number of counts above an energy 2EQ/3 plus a calculated wall loss correction for those counts in the energy interval from E0/3 to 2E0/3. Table VII shows the two contributions to the total yield. A correction to the total yield due to the "dead time" of the kicksorter is explained in Chapter IV. TABLE VII Data from Photodisintegration Spectra for He3(t , p)D 9.17 mev 8.06 mev Run 1+ Run 2 Time (hours) 8.606 3.889 5.617 Proton beam current (microamps) 16 25 23 Proton Energy (mev) 0.730 1.760 1.760 Total number of 8.06 or 9.17 mev gamma rays from target. (Includes correction for absorption in castle wall) 9.19(10)9 5.32(10)9 10.87(10)9 Total number of counts from energy ^^ E0 = Ey - Q , SE = half width of peak 938 427 512 Time dependent background in energy interval ^ to EQ + SE 543 78 93 Number of photodisintegrations in intervalto EQ + SE * 397 . 356 421 True number of photodisintegrations in peak, Np 360 267 309 Partial wall loss correction 25 33 38 Total yield 422 389 459 *Includes correction for "dead time" of the kicksorter + Spectrum shown in Figure 16 -36- g E. Gross Section Calculation .£&£ th£ Reaction He (r .n)D It is shown in Appendix A that the cross section (f is given in terms of the total yield by H(d)/p&\.S where H(d) is the effective path length-solid angle product for a source a perpendicular distance d from the chamber wall, is the atom density of helium-3 in the chamber, and Ny. is the total number of gammas emitted by the source. The cross section values are given in Table VIII together with experimentally measured values of d, pWg and ~&, The errors quoted are the total errors in each measured value. All the U.B.C. 2-body photodisintegration data is shown in Figure 18. The 90° differential oross section measure• ments of Berman et al (1963) and Stewart et al (1964) are shown for comparison. At these energies, the photodisintegration is predominantly due to electric dipole absorption -with a resulting sin 6 angular distribution of the reaction products. Hence, the differential cross sections have been transformed to total cross sections by multiplying by 8TT/3. The results of BSsoh et al (1964) for the photodisintegration of tritium are plotted as a function of the energy of the reaction produots above the threshold energy. F. Errors in the 2-body Cross Seotion The estimated probable errors, excluding the statisti• cal variation in the number of counts observed, are summarized in Table IX. The error in determining the scintillation counter CV3H A T0 0.74 A 0.6A T* (mb) 0.54 ©J. 0.4- © U.B.C. DATA • BERMAN ET AL 03 ^ STEWART ET AL V BOSCH ET AL ' T(Y; n)[ 0.2' THRESHOLD 0.1 1 0 ENERGY ABOVE THRESHOLD (MEV) 2 3 4-5 , i , i . i . u_ 7 Q 9 10 PHOTON ENERGY (MEV) FOR HeHtf^D FIGURE 18 Cross Sections for the Reactions He (V , p)D and T(tf , n)D TABLE VIII He3(£, p)D Cross Section Data Run Yield d H(d) cr ( from ( cm ) (cm- PH. V (mb) Table VII) sterad.) (atoms/cm^) 4n 8.06 mev 422 + 31 2.79 ± 0.1 19.7 ±0.5 (6.02 ± 0.08)(10)19 (7.22 ± 0.74)(10)8 0.493 ± 0.066 9.17 mev 389 ± 21 0.95 ± 0.1 34.9 ± 2.0 (4.07 ± 0.07)(10)19 (4.17 ± 0.43)(10)8 0.658 ± 0.089 Run 1 9.17 mev 459 ± 24 3.91 ± 0.1 18.29± 0.2 (4.07 ± 0.07)(10)19 (8.53 ± 0.88)(10)8 0.723 ± 0.087 Run 2 *Includes correction for absorption in chamber wall. TABLE IS Estimate of Errors 1. Error in determining gamma flux at the scintillation counter a. Affecting each run Counter solid angle "±.1.5% Setting bias point ± 0.5% b. Affecting 8.06 mev and Run 1, 9.17 mev Attenuation in paraffin and cadmium shield ± 2% Total ± 2.6% c. Systematic error in scintillation counter efficiency ± 10% Total error in gamma flux at the scintillation counter ± 10.3% 2. Error in determining gamma flux through the ionization chamber a. Due to uncertainty of the angular distribution of 9.17 mev gamma rays ± 2.8% b. Due to uncertainty of relative intensity of 8.06 and 9.17 mev ± 1% gamma rays Total ± 3.0% 3. Error in H(d) due to measurement of d 8.06 mev ± 2.5% 9.17 mev, Run 1 ± 5.7% 9.17 mev, Run 2 ± 1.1% 4. Error in p 'H e 8.06 mev ± 1.3% 9.17 mev ± 1.7% 5. Error in yield due to uncertainty of wall loss correction ± 1% -37- solid angle is caused by inaccuracies in the measurement of the counter area and the target,to counter distance. The error in determining the gamma flux through the ionization chamber is considered in Appendix C. The seemingly large uncertainty in is explained in Appendix A. The errors for each run, including statistical errors in the yield measurement, are given in Table X. The error Is calculated by adding the individual errors in quadrature. In each case the systematic error in the scintillation counter efficiency dominates the total error. • TABLE X Cross Section Errors Probable Error Neglecting Total Run the Uncertainty in Scintil• Probable lation Counter Efficiency Error 8.06 mev ± 8.4% ± 13.3% 9,17 mev-Run 1 ± 9.0% ± 13.5% 9.17 mev-Run 2 ± 6.5% ± 12.0% The errors in the absorption of the gamma rays by the chamber wall and the error in assuming this absorption to be constant over the surface of the chamber are negligible. The -38- absorption of the gamma rays by the glass wall of the beam tube is assumed to be the same for the scintillation counter and the chamber. The error in this assumption is less than 0.4%. The error in assuming a point source of gamma rays rather than a distributed source is also less than 1%. G. Estimation of the He ( ,n)2p Cross Section Using the data in the spectrum shown in Figure 17, a preliminary value for cf (3-body) can be obtained. .Unfortunately the thermal neutron capture peak from the reaction He^(n,p)T completely obscures the 3-body spectrum from 0.49 to 1.0 mev. Furthermore the wall loss and end effect for the 2-body events (Appendix A) give a significant contribution to the total yield in the region from 1.0 to 1.5 mev. Table XI summarizes the data for the 3-body photo• disintegration run. If one assumes that the spectrum shape is given by Figure 1, (that is, that there are no final ,state interactions), the cross section calculated from the data is The uncertainty quoted includes a ± 50% error due to statistical fluctuations in the total number of counts and 10% contributions due to uncertainties in gamma flux and background subtraction. TABLE XI Data for 3-body Cross Section at 9.17 mev Gamma Ray Source: C13(p,^)N14 at 1.75 mev resonance Time £.091 hours Proton beam current 15 mlcroamps Total number of 9.17 mev R gamma rays, Ny * 1,72(10) 4n 3-body photodisintegration data from Figure 17 Total counts 1.0 to 1.7 mev 99 Backgrounds: Beam dependent, neutron induced 29 2-body wall effect 36 2-body end effect 14 Net number of counts from 1.0 to 1.7 mev 20 Data for cross section calculation 3.79(10)19 H(d), d = 2.5 cm 24.5 cm-steradians 3-body wall loss < 10%, neglected * Includes correction for attenuation in castle wall. -39- CHAPTER VI Discussion of Experimental Measurements and Theoretical Calculations on the 5-Hueleon System A. Photodisintegration of Helium-5 The total cross section measurements for the reaction He (}T,p)D reported in this thesis and by Warren et al (1963) are not Inconsistent with the measurements of Gorbunov and Varfolomeev (1964). Direct comparison of these total cross section measurements with the 90° differential cross section data of Berman et al (1963, 1964), Stewart et al (1964) and Finckh et al (1963) cannot be made without knowledge of the angular distribution,of the reaction products as a function of photon energy. Gorbunov and Varfolomeev (1964) have measured the angular distribution but report only the angular distribu• tion for all 2-body events summed over photon energies from 5,5 to 170 mev. This angular distribution is shown as curve A of Figure 19. Eichmann (1963) has calculated the angular distribu• tion at a photon energy of 19.5 mev, with and without an electric quadrupole contribution to the dominant electric dipole absorption. For pure electric dipole absorption the angular distribution I(Q) is equal to sin2© . The effect of the electric quadrupole contribution is shown as curve B of Figure 19, where the curve has been normalized to the same peak height MEASURED A*J CALCULATED ANGULAR D\STR\FCUTION AT A PHOTON ENERSY OC 19.5 M£V/ FOR CLECTRVC DI POLE. AND ELECTRIC QUADRUPOLE A^SORRTION. FROM EICrtMA^IN (V9631) RELATIVE YIELD 0 CM FIGURE 19 Angular Distribution of the Reaction Products for the Reaction He (ft , p)D -40- asthe results of Gorbunov and Varfolomeev. The assumption can be made that the angular distribu• tion at photon energies riear 20 mev is approximated by curve A of Figure 19. This assumption Is partially justified by the calculated effect of the electric quadrupole contribution on the differential cross section. In this case the total cross section, cross section by Eq 6-1 where I(©) is given by Eq 6.-2 I(©) is suitably normalized so that 4/n steradians From equations 6-1 and 6-2 we have -41- Figure EO shows the total cross section data and an average of the 90° differential cross section data transformed by equation 6-3 to a total cross section. Figure 20 indicates that the total and 90° differential cross sections are in poor agreement and are inconsistent with the angular distribution given by equation 6-2. It is thus unlikely that both sets of data are correct. Consideration of the experimental techniques favour acceptance of the work of Stewart et al (1964) and Finkh et al (1963) over the measurements of Gorbunov and Varfolomeev (1964). In the region of the cross section maximum (Egv= 10 to 12 mev), electric quadrupole absorption is almost certainly negligible and the resulting angular distribution is P sin 0. Consequently, the total cross section is given by Eq 6-4 The maximum total cross section calculated using equation 6-4 and the differential cross section data of Stewart et al (1964) and Berman et al (1964) is max PHOTON. ENERGY (MEV) FIGURE 20 Comparison of Total and Differential Cross Section Measurements -42- B» Theoretical Cross Section Calculations Berman et al (1964) have obtained a good fit to their 90° differential cross section data using an Irving wave function with size parameter 1/^M = 2o6 F as the ground state wave function for helium-3 (Chapter II). They have considered only electric dipole absorption and consequently have under• estimated the total cross section by approximately 10% for photon energies greater than 20 mev. The Gaussian wave function first proposed by Yerde (1950) was rejected on the basis of their data. If the size parameter in the Gaussian wave function is chosen to fit the maximum cross section, the position of the peak is too high in energy; and if the maximum in the cross section is fitted to the oorrect energy, the corresponding cross section is much too large to fit the experimental data. From the work of Eichmann (1963), it is clear that the electric quadrupole contribution is significant and must be included in transforming differential cross section data into total cross section values. Furthermore, final state inter• actions would appear to Increase the maximum value of cross section by as much as 25% and cause the cross section to fall more rapidly at higher photon energies. As neither of these effects have been considered by Beiman et al, it is doubtful if the theoretical fit to their data has much significance. Gorbunov and Varfolomeev (1964) attempt fits to their total cross section data using both Irving wave functions and the theoretical estimations of Eichmann (1963). In neither case -43- is there reasonable agreement between theory and experiment. In addition to electric quadrupole contributions and final state interactions, Eichmann (1963) has included the mixed symmetry S' state with a 1% probability in the Initial state wave function for helium-3. Such an admixture is expected to be present to account for the neutron capture by deuterium (Austern, 1952) and has the effect of decreasing the energy of the cross section maximum by 20% and increasing the maximum value of the cross section by 10%. The possibility of an admixture of D-state in the ground state wave function of the 3-nucleon system has not been considered, although magnetic moment data call for its inclusion (Sachs, 1953). Electric dipole absorption from the D-state to a continuum P-state would account for the isotropic component of the angular distribution observed by Gorbunov and Yarfolomeev. To date, the deuteron wave function has been taken as a pure S-state although the measured quadru• pole moment of the deuteron indicates a 4% probability for a D-3tate component. Furthermore, the final state /interactions have Included only central forces; the influence of a tensor force on the final state wave functions has not been calculated. It is probable that all of these contributions to the cross section are significant and must be Included in further theoret• ical calculations. C. Electron Scattering on Tritium and Helium-3. The electromagnetic structure of tritium and helium-3 -44- has been the subject of much recent experimental and theoretical investigation. Collard et al (1963) have measured the charge density and magnetic moment form factors for both tritium and helium-3. The distribution of magnetic moment for helium-3 is more compact than that of charge, whereas the form factors for tritium appear to be approximately equal and furthermore equal to the magnetic moment factor for helium-3. The behavior of the form factors as a function of momentum transfer can be deduced from 3-nucleon wave functions. Several workers have attempted to fit the data using various forms of wave functions for the ground state of the 3-nucleon system (Schiff et al, 1963; Levinger, 1963; Schiff, 1964; Srivastava, 1964; Melster et al, 1964; Grlffy, 1964; Krueger and Goldberg, 1964). Several combinations of S1 or D-state admixtures to the dominant S-state . with exponential, Gaussian and Irving wave functions yield equally reasonable agreement with the experimental data. Schiff (1964) has proposed a 4% S'-state admixture to account for the difference in oharge and moment form factors for helium-3. Krueger and Goldberg (1964) include a D-state admix• ture and Griffy (1964) indicates that a T = 3/2 admixture to the dominant T = 1/2 component in the ground state wave function could also account for the difference in form factors. It is apparent that existing theoretical fits to the electron scattering data are too insensitive to determine 3-body wave functions. It is also evident that theoretical calculations on the photodisintegration of tritium and helium-3 are far more -45- sensltive to the choice of ground state wave functions. D. Conclusions The experimental teohnique used in the present cross section measurements has the advantage of using monoenergetic photons. An extension of these measurements using the 17.6 mev 7 8 gamma radiation from the Li (p,fr)Be reaction would serve to differentiate between the results of Gorbunov and Varfolomeev (1964) and the 90° differential cross section data. The reaction could not be observed in the ionization chamber used in the present work as the wall loss would be prohibitive. Prelim• inary calculations indicate that the experiment could be done using solid state oounters. The 2-body photodisintegration cross sections at 8.06 and 9.17 mev agree very well with the cross section calculated from the inverse reaction D(p,tf" )He3 using the principle of detailed Balance (Lai, 1961; G. Bailey and G. M. Griffiths, private communication). The results also agree with the work of Bosch et al (1964) on the photodisintegration of tritium and strengthen the hypothesis of charge independence of nuclear forces. The 3-body cross section for the photodisintegration of helium-3 at 9.17 mev is found to be (0.25 ± 0«13)mb, in agreement with Gorbunov and Varfolomeev (1964). The 3-body energy spectrum did not show significant deviation from the shape predicted by phase space calculations in the region where -46- the sum of the proton energies is greater than 1 mev. This would indicate that final state coulomb interactions for the 3-body breakup at a photon energy of 9.17 mev are not signifi• cant . -47- APPENDIX A Cross Section Calculation A. Chamber .Efficiency The number of photodisintegrations which occur in the active volume of the ionization chamber has been calculated by Robertson (1963). Consider a small element of volume dV in the chamber, of area dA and thickness dx at a distance x from the gamma source. The areal density of helium atoms in this volume element is given by pHe^x where p^£ is the atom density of helium in the chamber. If the attenuation in passing through the gas between the source and volume element is neglected, the number of gamma rays which pass through dV is Jir. Ny. is the total number of gamma rays emitted by the source with an angular distribution a I (cose) = aQ + o^cose + a2cos e+. where 4-TT sWQoSons The number of photodisintegrations in dV is given.in terms of -48- the cross section tlx, Kco^eV^^./o .d* dA can be written as x2d5c where dSc is the solid angle subtended by dA at the source. The total yield from the active volume is Qy), Chamber Active Volume If the source Is a perpendicular distance b 4- d from the chamber centre axis where b is the radius of the chamber we define functions n Hn(d) = || cos ed5ld> Chamber AcJive Volume In terms of these functions, equation A-l becomes Yield = f • pH€ ' ^ ' H (d) Eq A-2 where H(d) - aQ H0(d) -f eL± H-^d) 4- a2 H2(d)4 .... Robertson has evaluated Hn(d) by numerical Integration for n= 0,2,4- and BlacJtmore (private communication) has repeated the calculation with greater accuracy. The results for n = 0,2 are shown in Figure 21.for the chamber at 90 to the direction of the proton beam. It should be noted that the graph given in Robertson's thesis corresponding to Figure 21 is Incorrect. B. Atom Density of Helium-3 In the course of the three years taken to investigate the cross section for the photodisintegration of helium-3, the gas mixture of argon, methane, and helium-3 was cycled in and out of the chamber several times. During one period, the chamber developed a leak and part of the gas was lost. In all cases, when the gas was transferred, accurate pressure readings were taken using a March "Master Test", Model 210 pressure gauge so that the atom density of helium-3 was always known to within 1%, In certain cases, the argon and methane were separated from the hellum-5 by freezing out these gases at a temperature of 65°K in a trap cooled by solid nitrogen. The vapour pressure of argon at 65°K is 0.44 psia and this residual pressure of argon was accounted for in the pressure corrections. At the time of the 8.06 mev runs, the chamber contained 2,24 ± 0.03 atmospheres of helium-3 at 20°C. This is equivalent to an atom density of (6.02 ± 0.08)(10)19 atoms per cm3. The pressure in the chamber during the 9.17 mev 2-body disintegra• tion runs was 1.515 ±. 0.02 atmospheres and the atom density of helium-3 was (4.07 ± 0.07)(lO)19 atoms per cm3. Both of these figures were checked by repeating the work of Warren et al " T 0.1 0.2 0.3 0.4- 0.5 0.6 d/v''3 FIGURE ?IQ CHAMBER AT 90 TO PROTON BEA.M DIRECTION FIGURE 21 b FIGURE 21 HQ(d) vs. d and H2(d)/H0(d) vs. d for Chamber of Volume V and Diameter to Length Ratio 0.783 -50- (Appendix E) at 7.12 mev and using the known cross section at this energy to determine pHg . The agreement is well within the ± 10% experimental error of the comparison measurement. During the 9.17 mev 3-body run the atom density of helium-3 at 20°0 was (3.79 ± .04)(10) . G. Wall Effect At a given gamma ray energy Ey. and for a fixed total density of gas in the chamber, the photodisintegration products have a definite range, R. A fraction of these produots will collide with the cylindrical wall or pass out of the active volume defined by the guard rings at the end of the chamber. The effect of this "wall loss" is to remove counts from the full energy peak at EQ= (E^ - Qj mev to a lower energy in the spectrum. On the other hand, a fraction of ionization tracks which originate outside the sensitive volume will extend into the active region. These events will also appear in the low energy part of the spectrum and are termed "end effect". The combination of the processes which add counts to the low energy section of the spectrum is termed wall effect. 1. Wall Loss For an accurate measurement of the number of reactions occurring in the active region, it Is necessary to know the magnitude of the wall loss and its influence on the shape of the spectrum. Let the fraction of tracks of length R cm originating in the active volume which either strike the wall or extend into the end region of the chamber be P(R). -51- For a cylinder of radius b cm and length L cm, Robertson (1963) has calculated P(R) to be Eq A-3 He approximates the cylinder by a plane and shows this approximation to be valid for R P0(R) will be greater than P(R) since the number of events taking place near the wall (close to the gamma source) will be greater than for a uniform distribution. Robertson has calculated the limits on PQ(R) for the geometry used in the experiment. He finds P(R) <-. P0(R) < (.12 P(R) 2, Spectrum Shape due to the Wall Loss A fraction of the total number of counts in the photo- ' disintegration spectrum will appear below the full energy peak at energy EQ. Part of this low energy tail is due to wall loss; the fraction P(R) of events in the active volume -52- which collide with the wall or extend into the end region of the chamber. The remainder of the tail is due to end effect. In the 2-body breakup of helium-5, the proton and deuteron have energies 2EQ/3 and EQ/3 respectively. If the ranges are short compared to the wall curvature, only one of the reaction products can collide with the wall. The maximum energy lost in such a collision is 2E0/3. Alterna• tively, the end effect will contribute counts from zero energy to a maximum energy of 2EQ/3 in the spectrum. Batchelor et al (1955) discuss the shape of such a spectrum and these results are adapted by Robertson to the chamber used in this experiment. Figure 22 shows the calculated wall loss spectrum shape for the 8.06 and 9.17 mev 2-body runs. The end effect contribution is not shown. The following assumptions were made in the calculation. a. The curvature of the wall can be neglected. b. The track distribution is isotropic. c. The reactions are uniformly distributed throughout the chamber. d. The stopping power dE/dx is constant. The first three approximations are as valid for the spectrum shape calculation as they were for the wall loss calculation. The maximum wall loss correction was obtained for the 9.17 mev results and amounted to 29%. The error in this wall TOTAL WALL LOSS P(R)=l5.9°/« PARTIAL WALL LOSS P(R)- 6.5%, V O a of UJ r . (0 $ loo- s: i O.S56 \.7»2 2.568 ENERGY (MEV) WALL LOSS SPECTRUM FOR E^ = 8.06 MEV TOTAL WALL LOSS P(R)= 29% in PARTIAL WALL LOSS P'(R) • \0.°) % »- ui Pi * °5 uair ar 1654 CO H 5 o or z 1.226 2.4-52 3.677 ENERGY (MEV) e 1 7MEV/ WALL LOSS SPECTRUM FOR EY 9- " FIGURE 22 Wall Loss Spectra for the Reaction He (£ , p)D -53- loss correction is (-22% - 10%) which would result in an error of (-6% ± 3%) in the total yield. An energy dependent dE/dx alters the shape of the spectrum within the energy intervals E0/3 to 2EQ/3 and 2E0/3 to E0 but does not appreciably change the number of counts in each interval. To reduce the error in the wall loss correction and to eliminate the effect of an energy dependent ^/dx, the total yield was calculated in terms of a yield for E > E0/3 and a partial wall loss. This has the further advantage of removing the end effect from the calculation. " 3. 2-body Photodisintegration Speotrum Analysis The method of spectrum analysis is best explained with reference to the hypothetical spectrum shown in Figure 23. K(E) is the number of counts per unit energy so that the total number of counts NQ is given by Eo+SE Eq A-4 Eo/3 Due to the finite energy resolution, the full energy peak has a width 2SE which buries a part of the wall loss spectrum in the peak. We assume that the finite energy resolution does not affect the shape of the wall loss spee- trum to an appreciable degree. The number of counts in the peak Kp, excluding the wall loss contribution, is given by -54- Eq A-5 where the wall loss removes NQP(R) counts from the peak. The actual number of counts in the peak is denoted by Np'. Np' is the sum of Np and the wall loss contribution to the peak. The wall loss contribution is that fraction of the total number of wall loss counts for which the energy is greater than E0 - BE . This fraction is essentially the ratio of the area under the wall loss spectrum for energy greater than E0~ SE to the total area and is given by 0 + £0/3 Since the total number of wall loss events is NQP(R), Np' is given by Combining equations A-5 and A-6 and solving for NQ yields Equation A-7 gives the total yield NQ in terms of the measured number of counts in the peak and the calculated -55- wall loss. In order to reduce the dependence of the total yield on the calculated wall loss and spectrum shape, we calculate the total yield in terms of a yield for E >2"^'5/3 and a partial wall loss P*(R). P* (R) is the fraction of the total number of counts which appear in the spectrum at an energy less than ^^/^» P'(R) is given by Hence the total yield N0 Is given by N0--^ N(E)^E + ( *® • Np Eq A-8 where Np is given by equation A-6. The difference between the total yield as calculated using equation A-7 and the total yield calculated from equation A-8 was always less than 2%. It is assumed that equation A-8 gives a more accurate value of the yield. -56- APPENDIX B The Preparation of Carbon-15 Targets Commercially available C^3 targets are produced by separating the C-1-3 isotope in a mass spectrometer and allowing 15 the C beam to hit a target backing. This technique has certain inherent weaknesses. The C is only loosely attached to the target backing and is not able to withstand bombardment by proton beams greater than a few microamps. The targets so produced are non-uniform and consequently have a low yield relative to the calculated yield based on the measured target thickness. Further• more, the thickness of such targets is limited to a few kev for 'l mev protons. For these reasons it was necessary to produce C13 targets free of these shortcomings. Phillips and Richardson (1950) have reported a technique whereby benzene"is thermally "cracked" on silver foils using a high temperature oven. In an alternative procedure they used an induction heater to crack methyl iodide, CH3I, on nickel disks. Seagrave (1952) prepared carbon targets by passing current through a tantalum strip in a chamber containing CH^ or CHgl. Holmgren et al (1954) and McCormick et al (1961) report success with similar, techniques. The method used in the present work is simple, efficient, and reproducible. Methyl iodide, enriched to 59.5% 13 C was obtained from Merck, Sharp and Dohme of Canada Limited. The CHgl was purchased in a sealed glass container (11) which was -57- fitted with a flanged pyrex top and attached to the apparatus shown in Figure 24. The apparatus consisted of a valve assembly (8); (9), (10) and a pyrex tube (7) in which a 0.002 in thick platinum strip was suspended (6). The platinum strip was heated by ac current from a 5 volt isolating transformer fed by a 110 volt variac. The output of the transformer was connected to the Kovar seals (1) and the brass plate (15). The apparatus was made vacuum tight by means of a neoprene 0-ring(2) and a flange assembly (5),(4). The procedure used for making targets is outlined below. 1. A clean platinum strip 1.4 cm wide x 7.5 cm long was spot welded to 0.040 in. nickel wire (5) and to a nickel plate (12) and suspended in the pyrex tube. 2. The tube assembly was evacuated to a pressure of 2(10) mm, of Hg. through a pumping port (13) and the platinum was outgassed at 800°C (red heat) by passing 40 amperes ac current through it. 3. The tube was closed to the pump and the valve to the CHgl was also closed. For the first target, the fragile glass seal on the vial containing the CHgl was broken using an iron rod held within the tube with a permanent magnet. This allowed the volume of tube up to the Viton A valve diaphragm to become saturated with CH I vapour at a pressure of 30 cm of Hg. FIGURE 24 Target Preparation Apparatus. Scale - Full Size -58- 4. With the platinum heated to 600°C (dull red heat), the valve to the CH3I was opened and the vapour diffused into the craoking chamber. Thermal decomposition of the CHgl proceeded satisfactorily and the carbon deposited on the platinum. The copious quantities of gaseous iodine released in the process did not condense on the hot platinum but deposited on the sides of'the pyrex tube which was cooled with an air blower. After 5 minutes of cracking, the valve to the CHgl was closed and the iodine was pumped off. This procedure could be repeated many times, depending upon the thickness of target desired. Each five minute run deposited about 100 micrograms/cm2 of carbon on each side of the platinum in the form of an extremely tough layer. If the platinum was cooled quickly, the carbon layer blistered and could be removed from the platinum intact. If the platinum was cooled slowly, the carbon adhered to the platinum which then served as a target backing. Approximately 45% of the carbon in the CHgl was deposited on the platinum. The "one run" targets were 10 to 15 kev thick to 1.8 mev protons and withstood beam currents of 35 micro amps for hundreds of hours when suitably cooled. The excitation function for such a target is shown in Figure B5. The maximum yield is 7.4 gemma rays of energy 9,17 mev per / * 9 (10) protons* This compares favourably with the theoretical 1.740 1.750 1.7(60 1.770 PROTON ENERGY (MEV) FIGURE 25 Yield of 9.17 mev Gamma Rays as a Function of Proton Energy -59- g yield of 7.5 gamma rays per (10) protons as calculated using the measured cross section and resonance width (Appendix C). A C15 target 600 kev thick to 1 mev protons was prepared by peeling the carbon off the platinum and placing several layers in a gold target holder. This target produced a maximum yield of 5.8 photons of energy 8,06 mev per (10)9 protons of energy 730 kev. The yield calculated from the measured cross section and resonance width is 6.0 photons per 9 (10) protons. -60- APPENDIX C The. Reaction C13(p, )N14 and the Measurement of the Gamma Flux A. Gamma Ray Counter Efficiency The measurement of the efficiency of the 2-3/4 in. diameter by 4-| in. long Nal(Tl) scintillation counter is discussed by Singh (1959), Griffiths et al (1962), Larson (1957) and Robertson (1963). The gamma counter efficiency for gamma rays of energy Ey> is given in terms of the number of counts in the gamma spectrum above an energy bias E^/2 relative to the total number of gamma rays impinging on the counter. The •efficiency is written £(EY ; E^> /2) and is shown in Figure 26 as a function of Ey • Robertson (1963) derives expressions for the efficiency in terms of an energy bias other than Ey. /2, and this straightforward calculation will not be repeated here. There is one essential difference between the work of Robertson (1963) and the present work. The absolute efficiency of the scintillation counter had been measured at a photon energy of 6.14 mev and could be extrapolated to 6.92 and 7.12 mev with little error. Such extrapolation is not sufficiently reliable for the 8.06 and 9.17 mev gamma rays used in the present experiment, and the factor which limits the accuracy of the photodisintegration cross sections at these energies is knowledge of the gamma counter efficiency. It would be worth• while therefore, to measure the efficiency of the gamma counter FIGURE 26 Gamma Counter Efficiency for Half Energy Bias - 2 3/4 in diam. by 4 1/2 in Long Nal(Tl) Crystal -61- at an energy greater than 6.14 mev. This could be done in the following way. Two steps are required. Firstly, the absolute efficiency of the counter could be measured at 4.433 mev by counting the gamma rays to the ground state of C12 from the first excited state at 4,433 15 12* mev. This level can be populated by the reaction N (p,°<)C • The gamma yield from the consequent de-excitation can be determined exactly by counting the associated alpha particles with a solid state counter, providing the alpha and gamma angular distributions and the alpha-gamma angular-correlation are taken into account, l? Secondly, one could populate the 16.11 mev level in C resonanceby the proto. nThi bombardmens level decayt ofs 90%a througt the hwell-know the 4.43n 316 me0 vke levev l mentioned above, and gamma rays of energy 11.7 and 4.43 mev can be observed in coincidence. Hence the determined efficiency at 4,43 mev would lead directly to an experimental measurement of the efficiency at 11,7 mev. B, The Reaction C15(p, T )H14 as a Source of Gamma Rays 1« Introduction In order to obtain an accurate measurement of the cross section for the photodisintegration of helium-3 at gamma ray energies of 8.06 and 9,17 mev, we must know the angular distributions and relative intensities of the gamma rays produced by the proton bombardment of C^5 at the 0,554 -62- and 1,75 mev resonances. As is shown in Appendix A, the angular distribution has a large effect on the factor Hn(d) which enters directly into the cross section calcula• tion. The relative intensities of cascade gamma rays must be known In order to calculate the 8,06 or 9,17 mev gamma flux from the recorded gamma spectra. 2, 0,554 mev Resonance The reaction C^Cp,^ )N"*"4 was first studied, by Curran et al (1939), They observed a resonance in the gamma yield at a proton energy of 0,57 mev and determined the maximum energy of gamma ray to be 8.5 mev. Subsequent work by Lauritzen and Fowler (1940), Fowler et al (1948), and Seagrave (1952) fixed the energy of the resonance at 0,554 mev and the energy of the gamma ray to the ground state of at 8.06 mev, Seagrave also measured the resonance width PR and obtained a value of 32,5 kev. Devons and Hine (1949) measured the angular distribution of the 8.06 mev gamma rays from the 8,06 (.554)- mev level and found it to be isotropic. This is consistent with the level being formed by S-wave protons as expected. Subse• quent workers have Investigated the de-excitation of the 8,06 (.554)- mev level in detail (Woodbury et al, 1953; Hird et al, 1954; Lehman et al,1956; Wilkinson and Bloom, 1957; Broude et al, 1957). The accumulated data has been compended by Ajzenberg-Selove and Lauritzen (1959) and is -63- shown in Figure 27. 3. 1,75 mev Resonance Day and Perry (1951) found a sharp resonance in the reaction C13(p,}f )N14 at a proton energy of 1.75 mev, yield• ing 9.17 mev gamma rays. Several workers (Table XII) have since measured the angular distributions or the anisotropy A for the 9.17 mev to ground state transition. The results of Rose et al (1960) are used in this present work. TABLE XII Measured Angular Distribution of the 9.17 mev Radiation at the 1.75 mev resonance Anisotropy = Angular Distribution Yield(0°) -i Yield(90°) Day and Perry (1951) - 0.399± 0.013 Woodbury et al (1953^ - 0.48 ±0.03 Rose et al (1960) 1 - (0.55± 0.02)cos2e - 0.55+0.02* \Q+\ <0.02 Segal et al (1961) 1 - (0.59 ±0.03)oosf 6 - 0.56± 0.04* + (0,03 ± 0.03)cos*G ^Calculated from the. measured angular distribution. Several people have measured the relative intensities of the cascade gamma rays produced In the de-excitation of the 9.17- 4- 8.06- SQI CM CM t—r m VI +» +• 0D 7.03- ort CM 00 in 6.44- 3 5.Q3. 5.69- 5.69- 5.10- -.9I (O) 4 - o m ! + 3.95- 3.95- m vO 231 2.31- X N1* 8.0k MEV LEVEL DECAY 9.17 MEV LEVEL DECAY FIGURE 27 Decay Schemes for the 8.06 and 9.17 mev Levels of Nx . Only those levels involved in the decay of the 8.06 and 9.17 mev states are shown. Broken arrows denote uncertain transitions and parentheses denote uncertain assignments. -64- 9,17(1*75)-mev resonanoe level in N14 (Seagrave, 1952; Wood• bury et al, 1953; Rose et al, 1959; Rose, 1960), The results of Rose (I960) are reproduced in Figure 27 and are used in the calculation of the gamma flux in the present work, 4. Doppler Shift When a nucleus in an excited state of energy EQ, moving with velocity v, decays by emission of a gamma ray to the ground state, the energy of the gamma ray is given by Eq CI where 14 is the mass of the nucleus and 0 is the angle between v and the direotion of emission of the gamma ray. The first term in equation C-l is the Doppler shift, the second term is the recoil shift, For the decay of the 9,17 14 mev excited state in N , the recoil shift is 3,2 kev. Since the ionization chamber was placed at 90° to the proton beam and hence to ^, the Doppler shift in the present case is zero. The maximum Doppler shift occurs at 8=0° or 180° and when v has the maximum value allowed by the kinematics of the reaction; that is, when decay takes place before the -65- nueleus slows down in the target. The magnitude of this maximum shift for the decay of the 9.17 mev state is £ 40 kev. A knowledge of the stopping time X and the stopping oross section £ for the nucleus leads to a direct measure• ment of the lifetime of the excited state. Such a shift in gamma ray energy would result in a corresponding shift in the 2-body photodisintegration peak in the ionization chamber. However, the expected shift for the 9.17 mev gamma rays is too small to be observed and an estimate of the lifetime of the state is not possible by this technique. C. Analysis of the Gamma Spectra 0*554 mev Resonanoe The gamma spectrum obtained by bombarding the thick target described in Appendix B with 0.730 mev protons is shown in Figure 28. The de-excitation of the 8.06(.554)-mev 14 level in N yields 8.06, 5.7 and 4.11 mev gamma rays in the ratio 82:11:4. The 5.7 mev radiation is due to the sum of the 5.69 -*-0 and 8.06->-2.31 transitions which cannot be separated in the scintillation counter spectra. Any aniso• tropy in the 5.7 mev component can be neglected at the angle of observation used in the present work. The total number of counts above 4.03 mev In the spectrum shown in Figure 28 is given by M * [o.82£(8.O6;4:0B) + 0.O4r£(5j;4-.03)-»-OJl£(4.l\;4:O5)]N' NUMBER OF COUNTS -66- where ^(E^ ; Eb) is the scintillation counter efficiency for gamma rays of energy E^ at an energy bias E^ and N' is the total number of 8.06, 5.7 and 4.11 mev gamma rays impinging on the counter. £(5.7;4.03) was determined using the efficiency to a half energy bias, £(5.7; 2.85)3and an assumed spectrum shape. The gamma ray spectrum from the 340 kev resonance of 16 the reaction F19(p,©0no which yields 98% 6.14 mev radiation (Dosso, 1957) was assumed to give the shape of the spectrum for the 5.7 mev radiation and is shown in Figure 29. The ratio of the number of counts above 4.03 mev to the number above 2.85 mev was 0.82 ± 0.01. Hence £(5.7; 4.03) is given in terms of the efficiency to a half energy bias of 2.85 mev by E(5.7;4-.03)= 0.82 £(5.7; 2.85) = 0.6OB ± 0.008 where £,(5.7; 2.85) is given by Figure 26. The uncertainty in £(5.7; 4.03) does not include the systematic error in £ (5.7; 2.85). As seen In Figure 28, the 4.11 mev radiation is easily accounted for by extrapolation of the 8.06 and 5.7 mev tail, again using an assumed spectrum shape. The ratio of the number of counts above 4.03 mev contributed by the 4.11 mev radiation to the total number above 4.03 mev was found to be 0.023 ± 0.003. Hence the ratio of the number of counts above 4.03 mev due to the 8.06 mev radiation to the total IIOO -, ENERGY (MEV) FIGURE 29 6.14 mev Gamma Spectrum from the Reaction F19( p,=*Y)016 at the 340 kev Resonance -67- number above 4.03 mev is 1 - 0.04 £(5.7}4.03) - 0.023 - 0,953 ± 0,011 The number of 8,06 mev gamma rays incident on the counter is given in terms of the number of counts N above 4,03 mev in the spectrum by ^cou^Ter _ 0.953 N 8-0G E(8.06;*h03) = \.2B Kl 2, 1.75 mev Resonance 13 The gamma spectrum from the proton bombardment of C at 1.75 mev is shown in Figure 30. As well as the primary component at 9,17 mev, the speotrum contains contributions from 7,03, 6,44, 5,7 and 5,1 mev gamma rays. In order to determine the number of 9,17 mev gamma rays from the spec• trum the contribution from the other gamma rays must be subtracted. Two methods of subtraction were used. In the first case, the 9,17 mev contribution was assumed to have the same shape as the pure 8.06 mev spectrum obtained previously. Using this shape and the known efficiency £,(9.17; 4,585), the number of 9,17 mev gamma rays incident on the counter is given in terms of the number of counts N above 4,585 mev by Sooo 4ooo Y- z 3ooo- o (J BIAS 2ooo I ooo I 1 1 1 1 1 1 1 1 1 90 lOO UO I20 <3o 14-0 150 160 170 \QO KlCKSORTER CHANNEU i 1 1 1 1 1 1 1 1 1 1 1 5 6 7 8 9 10 ENERGY (MEV) FIGURE 30 Gamma Spectrum from the Reaction C13(p,Y)N14 at the 1.75 mev Resonance -68- counter N = 1.156N 9.17 In the second method, the efficiencies £(E£;4«585) for each gamma ray were calculated using the known efficiencies EfE^E^y^) as given in Figure 26 and assuming the spectrum shape to be as given in Figure 29. In terms of the relative intensities R^O) of the gamma rays at an angle © to the proton beam, the number of 9.17 mev gammas incident on the counter is given by 2 Rt(e) £(Ei; 4-.SS5) counter N 9.17 R9.l7(e) £(ai7;+.505) Table XIII shows the values of E^tG) and (Ei;4.585) as calculated. Using these values one obtains, counter N - 1.162N 9.17 The two methods are in good agreement and a value counter N = 1.16N was used in calculating the 9.17 mev 9.17 gamma flux. TABLE XIII Angular Distributions, Relative Intensities and Scintillation Counter Efficiencies for the Gamma Rays from the 9.17(1.75)-mev Resonance Level in N- Relative Intensity Gamma Ray Ri(G) Angular Distribution integrated over Energy, Ej. 9=152.3° £(Ei; 4.585) 4 TT steradians £(Ei ;f) 9.17 1 - 0.55cos28 100 70 0.782 0.782 7.03 isotropic 3 3 0.767 0.662 6.44 1 + 1.6cos2© 7 8.5 0.76 0.606 - l.lcos^© 5.7 1 - 0.55cos2© 3 2.1 0.744 0.518 5.1 1 - 0.32cos2e 3 2.5 0.728 0.223 -69- APPENDIX D Neutron Induced Background and Neutron Shielding One of the main experimental difficulties associated with the measurement of the cross section for the photodisin• tegration of helium-3 was the problem of neutron-Induced back- ground. The reaction He (n,p)T has a positive Q-value of 0.765 mev. The cross section as a function of neutron energy is shown in Figure 31. Because of the high cross section and beoause of the unique energy release for this reaction, several workers have used helium-3 filled Ionization chambers and proportional counters as neutron spectrometers (Batchelor et al, 1955; Mills et al, 1962; Freeman and West, 1962; Brown, 1964; Sayres and Coppola, 1964; Friedas and Chrien, 1964). As discussed in Chapter IV, the thermal neutron capture by helium-3 served as an energy calibration for the ionization chamber. The neutron disintegrations were distributed through• out the chamber and produced counting conditions similar to those met in the photodisintegration measurement. In all other respects, the advantages of the chamber as a neutron detector became disadvantages for the photodisintegration work. For the 2-body photodisintegration at 8.06 and 9.17 mev, the thermal neutron capture peak at 0,765 mev did not inter• fere with the photodisintegration spectrum. Hence fast neutron disintegration resulting in an energy release in the chamber of greater than 1.5 mev was the only source of neutron Induced FIGURE 31 Cross Section for the Reaction He ( n, p)T. From Hughes and Schwartz (1958) -70- background whioh proved troublesome. Elaborate paraffin and cadmium shielding eliminated this background. The neutrons were moderated by the paraffin and absorbed by the cadmium which has a high capture cross section for thermal and epithermal neutrons. For the 3-body photodisintegration at 9.17 mev, the thermal neutron peak fell in the middle of the 3-body spectrum (Figure 1, Chapter II), and all neutrons, thermal or fast, interfered with the measurement. The main source of neutrons, and that most easily eliminated, was in the exit seotion of the Yan de Graaff column and in the 90° analyzing magnet vacuum box where the neutrons were produced by the deutron bombardment of deuterium and carbon. The deuterons arise from the 0.0156% natural deuterium contamina• tion of the hydrogen used in the Yan de Graaff ion source whioh results in deuterium ions being accelerated along with the proton beam. The 90° analyzing magnet selects the mass one proton beam; the mass two and mass three beams (Hg, D and HD ) strike a tantalum strip in the magnet box (Figure 32). The number of neutrons produced in the column and magnet box was kept to a minimim by periodically cleaning the tantalum collimator in the accelerating column and the tantalum beam catcher in the magnet box. The neutrons which were produced were prevented from entering the chamber by wax shielding walls and a wax and cadmium castle around the chamber and target. For the 8.06 mev runs at a proton energy of 0,730 mev, the proton beam itself produced no neutrons and the precautions IONIZATION CHAMBER VAN OE. GRAAFF COLUMN WAX AND CADMIUM CASTL.E TANTALUM. COLLIMATOR WAX AMD CAOMIUM J WALL 90 ANALYZING PLATlWut MA&MET TUBE LlNf MASS 2 BEAt-1 EL&CTROST AT I | TANTALUM FOCUSSING COLLIMATORS' r PROTON BEAM 5 PYRE.H. TUBE " TANTALUM COLD TRAP BEAM CATCHER CARBON- 13 TARGET ON PLATINUM RACKING FIGURE 32 Beam Collimators and Neutron Shielding -71- outlined above were adequate. However, for the 9.17 mev runs, the proton beam energy was 1.75 mev, well above the (p,n) threshold for most materials, as shown in Table XIV. By introduc• ing collimators in the horizontal pyrex beam tube, the proton beam was prevented from striking the glass, a major source of neutrons from (p,n) reactions. In a similar manner, the proton beam scattered off the platinum target backing was shielded from the beam tube wall by a 0.001 inch thick tantalum liner. Such extra precautions effectively eliminated all neutrons which resulted in an energy release in the chamber greater than 1.5 mev and thus were adequate for the S-body photodisintegration at a photon energy of 9.17 mev. Such was not the case for the 3-body runs at 9.17 mev. All attempts to completely eliminate the thermal neutron peak failed. Table XIV shows that there are very few materials suitable for use as a target backing, as beam collimators, or as a beam tube liner for which the (p,n) threshold is less than -1.75 mev. Carbon and nickel are obvious choices. However, impurities in both materials produced a neutron flux which resulted in 20 thermal neutrons counted by the chamber per 15 (6)(10) protons incident on the material. The concentration of impurities need not be greater than one part in (10) to account for this neutron count rate. The ionization chamber is "black" to thermal neutrons and the observed neutron count rate represents a thermal neutron flux of only 0.1 neutrons per cm per minute through the chamber for the proton beam currents used. TABLE XIV (p,n) Q-values. Calculated from the Mass Values of Everling et al (i960) Isotope % Natural (p,n) Q-value Isotope % Natural (p,n) Q-value Isotope X Natural (p,n) Q-value Abundance Abundance Abundance Li7 92.48 -1.646 Mg26 11.3 -4.808 Ca42 0.64 -6.660 Be9 100 -1.854 Al2? 100 -5.607 Ca44 2.13 -4.432 28 4 B10 18.83 -4.56 Si 92.28 -14.545 Sc 5 100 -2.827 BH 81.17 -2.764 Si2? 4.67 -5.745 1^46 7.95 -8.082 12 C 98.89 -18.453 p31 100 -6.219 Ti47 7.75 -3.695 32 48 C13 1.11 -3.003 S 95.06 -13.812 Ti 73.45 -4.805 i4 N 99.6 -5.934 S33 0.74 -6.231 Ti49 5.51 -1.383 016 34 99.8 -16.360 S 4.18 -5.774 Ti50 5.34 -2.991 19 35 F 100 -4.038 CI 75.4 -6.039 V51 99.75 -1.534 Ne20 90.5 -16.121 CI37 24.6 -1.598 Cr50 4.41 -8.433 40 52 Ne22 9.2 -3.622 A 99.6 -2.274 Cr 83.46 -5.485 23 53 Na 100 -4.875 K39 93.1 -7.638 Cr 9.54 -1.370 24 4 Mg 78.6 -14.80 K41 6.9 -1.220 Cr5 2.61 -2.15 Mg25 10.1 -5.074 Ca40 96.92 -14.768 Mn55 100 -1.014 TABLE XIV (continued) Isotope % Natural (p,n) Q-value Isotope % Natural (p,n) Q-value Isotope % Natural (p,n) Q-value Abundance Abundance Abundance Fe54 5.90 -9.622 Ga69 60.0 -3.02 Zr96 2.8 -0.500 56 71 93 Fe 91.52 -5.383 Ga 40.0 -1.015 Nb 100 -1.265 Fe57 2.245 -1.649 Ge70 20.45 -7.32 Mo92 15.05 -7.200 Cc59 100 -1.857 Ge72 27.41 -5.143 Mo94 9.35 -5.102 Ni58 67.8 -9.317 Ge?3 7.77 -1.155 Mo95 15.78 -2.439 Ni60 26.2 -6.912 Ge?4 • 36.58 -3.347 Mo96 16.56 -3.760 Ni61 1.2 -3.076 Ge76 7.79 -1.760 Mo9? 9.60 N.C. Ni62 3.7 -4.62 As75 100 -1.648 Mo98 24.60 -2.500 Ni64 1.1 -2.46 Br79 50.57 -2.404 Mo100 9.68 N.G. Cu63 69.1 -4.149 Br81 49.43 -1.030 Rh103 100 -1.342 65 8 Cu 30.9 -2.131 Rb 5 72.15 -1.890 Pd102 0.8 N.C. 64 8 104 Zn 48.87 -7.84 Rb ? 27.85 -0.510 pd 9.3 -5.062 66 105 Zn 27.62 -5.96 Zr90 51.46 -6.900 pd 22.6 -2.800 67 9 Zn 4.12 -1.781 Zr l 11.23 -2.380 Pd106 27.1 -3.753 68 92 108 Zn 18.71 -3.70 Zr 17.11 -2.860 Pd 26.7 -2.619 94 110 Zn70 0.69 -1.435 Zr 17.40 ^1.550 Pd 13.5 -2.230 TABLE XIV (continued) Isotope % Natural (p,n) Q-value Isotope % Natural (p,n) Q-value Isotope % Natural (p,n) Q-value Abundance Abundance Abundance Agl07 51.35 -2.223 SnU? 7.67 -2.602 W184 30.64 N.C. 109 118 186 Ag 48.65 -0.940 Sn 23.84 N.C. W 28.64 -3.300 Gdl°6 1.22 N.C. Sn119 8.68 -1.361 Pt-192 0.78 -4.020 Cd108 0.89 -5.880 Sn120 32.75 -3.505 Pt194 32.8 -3.348 Cd1*0 12.43 -4.742 Sn122 4.74 -2.37 Pt195 33.7 -1.053 CdUl 12.86 -2.020 Sn124 6.01 -1.380 Pt196 25.4 -2.570 Cd112 23.79 -3.398 Sb*2* 57.25 -0.399 Pt198 7.23 -1.440 CdH3 12.34 -0.473 Sb123 42.75 -0.820 Au197 100 N.C. 204 CdH4 28.81 -2.203 jl27 100 -1.480 Pb 1.37 -5.068 116 33 206 Gd 7.66 -1.370 Csl 100 -1.273 pb 26.26 -4.383 113 140 207 In 4.16 -1.466 Ce 88.45 -4.042 Pb 20.82 -3.183 115 142 208 In 95.84 -0.282 Ce 11.1 -1.480 pb 51.55 -3.659 Snll2 1.01 N.C. Pr141 100 -2.582 Bi209 100 -2.68 114 181 Sn 0.68 N.C. Ta 100 -0.982 y238 99.28 -0.633 115 182 Sn 0.35 N.C. W 26.31 N.C. 116 Sn 14.28 -5.48 W183 14.28 N.C. -72- Of all the materials used as collimators and target backings, gold, platinum and reactor grade graphite produced the lowest neutron count rate in the chamber (20 thermal neutrons per milliooulomb of proton beam on target). Stopping the proton beam outside the castle reduced the count rate to less than 100 neutrons per coulomb of stopped beam. The number of neutrons produced by (£ ,n) reactions in the material surrounding the target and chamber was at least an order of magnitude lower than those produced by (p,n) reactions. That is, the presence of a 9.17 mev gamma flux did not increase the neutron flux to a measurable extent over the neutron flux observed with no gamma rays. The 3-body spectrum shown in Chapter V was obtained using a platinum target backing and a platinum beam tube liner. A self-supporting carbon-13 target would allow the proton beam to pass through the target and be stopped outside the castle. Preliminary runs with no target but with the beam passing through the castle indicate that neutron count rates of less than 200 neutrons per coulomb of proton beam can be attained. APPENDIX E Reprinted from THE PHYSICAL REVIEW, Vol. 132, No. 4, 1691-1692, 15 November 1963 Printed in U. S. A. Photodisintegration of He3 near the Threshold* J. B. WARREN, K. L. ERDMAN, L. P. ROBERTSON,! D. A. AXEN.J AND J. R. MACDONALDJ Physics Department, University of British Columbia, Vancouver, British Columbia, Canada (Received 1 July 1963) The total cross section for the reaction He3 (y,p)T> has been measured at gamma-ray energies of 6.14, 6.97, and 7.08 MeV. The cross section was found to be 0.102, 0.298, and 0.307 mb at the three energies. The experimental cross-section values are compared with those of the inverse reaction D(p,y)Hc?, as an accurate check on the principle of detailed balance. HE photodisintegration of He3 has been observed parts He3 by freezing the tritium at 4.2°K. The charac• T by Cranberg1 and Berman et al? This letter teristics of the ionization chamber and the purifica• describes the measurement of the total cross section of tion technique will be published in a separate the reaction He3(Y,/>)D (Q=— 5.493 MeV) at gamma- communication.3 ray energies of 6.14, 6.97, and 7.08 MeV. The experi• Two identical chambers were used in the experiment mental cross-section values are compared with those of and were placed symmetrically on either side of the the inverse reaction as an accurate check on the principle gamma-ray source. One chamber contained He3, of detailed balance. methane, and argon, the other contained He4, methane, The reaction was observed in a cylindrical, gridded and argon. The outputs of the two chambers were fed ionization chamber of active volume 1.485 liters. The through separate amplifying systems into separate chamber contained 1.05 atm of He3, 0.0187 atm of halves of the memory of a model ND 103 Nuclear methane, and 1.36 atm of argon. The methane was added Data pulse-height analyzer. In this way, all effects not to reduce the resolving time of the chamber to less than specific to the He3 could be monitored. Elaborate wax 2 usee, whereas the argon served as a stopping gas for and cadmium shielding was required to attenuate the the photodisintegration products. It was necessary to thermal neutron background and thus reduce the cap• reduce the tritium contamination in the He3 to eliminate ture of thermal neutrons by the He3. the electron background due to the /3 decay of tritium. The gamma rays were produced by bombarding CaFs The tritium contamination was reduced from one part targets with protons accelerated by a 3 MeV Van de 6 3 10 tritium per 10 parts He to five parts tritium per 10 Graaff generator. Table I shows the relative yields and 1301 673.5 KEV RESONANCE 120 995 KEV RESONANCE IIO IOO 90 SO• He3+ Y FIG. 1. Photodisinte• TO gration spectra. 60 '7.12 M«V 80 40 SO 20 IO- I I I 0.5 1.0 1.5 '•>.«•'€ N ERG Y IN MEV * Research supported by a grant from Atomic Energy ,of panada, Ltd. t Holder of a National Research Council Studentship 1960-:<>2f.Present'address: Rutherford High,Energy Laboratory, Oxford, England. ' •''•*'£•'•;'. • • ' ""; " '•' J's'iSi?./ i Holder of a National Research Council Studentship 1961-63. 1 L. Cranberg, Bull. Am. Phys. Soc. 3, 173 (1938)." «i. •-.•..->:,•>-- ...... 8 B. L. Berman, L. J. Koester, Jr., and J. H. Smith, Phys. Rev. Letters 10, 527 (1963). ' K. L. Erdman, L. P. Robertson, D. A. Axen, and J. R. MacDonald, Can. J. Phys. (to be published). 1691 1692 WARREN, ERDMAN, ROBERTSON, A XEN, AND MACDONALD TABLE II. Cross section for the reaction He?{y,p)D. Total probable error including Experi- uncertainties in cr mental relative gamma- (MeV) (mb) error ray yields 6.14 (873-keV resonance) 0.109 9% 14% (935-keV resonance) 0.102 6% 7% 6.97 0.298 5% 16% 7.08 0.307 5% 8% The gamma flux was measured with an accurately calibrated Nal(Th) scintillation counter. The efficiency for this counter had been previously measured by Grif• fiths, Larson, and Robertson7 at a gamma-ray energy of 6.14 MeV._ 0.5 l.o I.S The photodisintegration spectra at the two fluorine resonances are shown in Fig. 1. Below 1.2 MeV, the ^REACTION * [E " 5.493] Y MEV spectra from the two resonances are essentially identical. FIG. 2. Comparison of cross sections for the inverse In calculating the cross sections at 6.92- and 7.12- MeV reactions Ke?(y,p)T) and D(p,y)Ke'. gamma-ray energies, no attempt was made to separate the peaks. Instead, the cross section is given in terms of angular distributions of the gamma rays from the reac• a mean gamma-ray energy assuming the ratios of 6.92- tion F19(^,a7)016 at the two proton bombarding energies and 7.12-MeV gamma rays as given in Table I, and used in the experiment.4-6 taking into account the rate of change of cross section with change in gamma-ray energy. The experimental TABLE I. Relative yields and angular distribution of cross section values are shown in Table II. gamma rays. The inverse reaction, D(/>,7)He3, has been studied by several workers.7-11 Figure 2 shows the results of the Relative gamma- ray yields application of the principle of detailed balance to the 11 -^proton 6.14 6.92 7.12 experimental measurements of Griffiths. The photo• (keV) MeV MeV MeV p—y angular distribution disintegration cross-section values are shown for com• 873.5 73% 20% 7% 6.14 MeV 1-0.1 cos^fl parison. The agreement is well within experimental 6.92 MeV 1+0.51 cos=0-O.22 cos'0error . 7.12 MeV 1+0.622 cosV 935 75% 3% 22% Isotropic at all energies 7 G. M. Griffiths, E. A. Larson, and L. P. Robertson, Can. J. Phys. 40, 402 (1962). 8 W. A. Fowler, C. C. Lauritsen, and A. V. Tollerstrup, Phys. «J. M. Freeman, Phil. Mag. 41, 1225 (1950). Rev. 76, 1767 (1949). 6 H. J. Martin, \V. A. Fowler, C. C. Lauritsen, and T. Lauritsen8 D. H., Wilkinson, Phil. Mag. 43, 659 (1952). Phys. Rev. 106, 1260 (1957). 10 G. M. Griffiths and J. B. Warren, Proc. Phys. Soc. (London) 6 R. W. Peterson, W. A. Fowler, and C. C. Lauritsen, A68Phys,. 78 Rev1 .(1955) . 96, 1250 (1954). 11 G. M. Griffiths (private communication). lM'l-'WWIA, if Reprinted from THE REVIEW OF SCIENTIFIC INSTRUMENTS, Vol. 33, No. 10, 1111-1112, October, 1962 Printed in U. S. A. Gas Flow Regulator for an rf Ion Source* when the power supply voltages for the oscillator are set at fixed values, the power coupled into the plasma is B. L. WHITE, L. 1'. ROHERTSON,| K. L. ERDMAN, directly proportional to the gas pressure in the discharge AND J. R. MACDONALDI tube. This note describes an apparatus which stabilizes Physics Department, University of British Columbia, the plasma density by maintaining the oscillator power Vancouver S, B. C, Canada output constant at its optimum level by means of a servo (Received June 28, 1962) system controlling the pressure in the discharge tube (Fig. 1). A signal proportional to the oscillator power HE beam extracted from an ion source for injection regulates the flow of gas into the discharge tube. into a Van de Graaff accelerator must be stable in T The gas flow rate is controlled by the power dissipation intensity and focus for long periods of time. The intensity in the thermal leak heater (Fig. 2). As the heater power and focus are both functions of two externally controllable is increased from zero, the leak temperature rises, and the TO ACCELERATOR TUBE differential expansion of the brass case and the stainless 1 RF steel rod allows a gap to form between the ball and its OSCILLATOR OSCILLATOR PLATE SUPPLY A seat, the size of the gap determining the leak rate from OSCILLATOR the high pressure reservoir into the discharge. The set COIL screw is adjusted so that at a heater power input of 15 W the leak just begins to open; it will then shut off tight at ION SOURCE OAS FLOW REGULATOR OSCILLATOR PLATE SUPPLY i DIFFERENCE VR AMPLIFIER TO OAS BOTTLE FIG. 1. Block diagram of regulator system. ion source parameters, the plasma density and the ex• tracting field. Geometrical factors affecting the beam FIG. 3. Servo circuit. usually remain fixed while the ion source is operated, and the extracting field is easily regulated by conventional a power input of about 6 W. In the present system, how• means. This leaves the plasma density as the factor whose ever, the leak is not required to act as a tight shutoff, stability will ultimately determine the beam stability. since the various gas supply bottles are shut off separately The plasma density is in turn a function of the power with solenoid valves. coupled into the discharge from the rf oscillator, and of The servo circuit (Fig. 3) is driven by an error signal the gas pressure within the discharge tube. However, derived by comparing the voltage Vi .developed across Ri (by passing through Ri a fraction of the direct current in the rf oscillator plate circuit) to the reference voltage VJI developed across the Zener diodes D3 and Di. The comparison is made in the difference amplifier Ti, after > o o o e the dc level of Vi has been shifted by means of the Zener diodes Di and D-.. The emitter-followers T» and T3 provide successive stages of power gain. The gain feedback product oooo o/o o o ©/o © around the loop (oscillator power level —> difference am• plifier —> power amplifier —> thermal leak—> gas pressure © —> oscillator power level) is about —10, which is sufficient to stabilize the discharge conditions so that the beam current onto the target of the accelerator is maintained constant to within a few percent over periods of many FIG. 2. Thermal leak. Body made of brass. The leak is mounted in an evacuated case for thermal stability and minimum power con• hours. sumption. (1) Set screw, (2) soft solder vacuum seal, (3) Nichrome heater, 22.5 il, (4) stainless steel rod, (5) asbestos, (6) steel ball, * Research supported by a grant from Atomic Energy of Canada, (7) ball seat, taper formed with drill tip, ball forms tight shut-off Ltd. seat by pressure deformation of brass. f Recipient of a National Research Council of Canada Studentship. APPENDIX G Rojirintcd from Tut: REVIEW OF SCIENTIFIC INSTRUMENTS, Vol. 34, No. 11, 1280-1281, November. 1963 Printed in U. S. A. Removal of Tritium from He3* K. L. ERDMAN, • L. P. ROBERTSON,! D. AxEN,f AND J. R. MACDONAI-Df Physics Department, University of British Columbia, Vancouver 8, B.C., Canada (Received 22 July 1963) OMMERCIALLY available He3 has a tritium con• liquid He4. Two German silver tubes 0.278 cm in diameter C centration of approximately 2 parts in 10s. As and 30 cm long, containing spiral inserts to increase the tritium is beta active the concentration must be reduced to effective tube, length, were soldered together to serve as a less than 1 part in 1010 before He3 can be used in ionization heat exchanger. The trap in the cryostat had a volume of chambers or proportional counters. Hanson et al.1 report 6 cm3. The transfer pump was constructed of commercially concentrations of 7 parts in 10" by distilling the He3 at available micro-finished Shelby steel tubing. The piston 1°K. Elliott2 reports concentrations of 1 part in 10" by was fitted with a double O-ring seal and the portion of the absorbing the tritium in pyrophoric uranium. We have upper cylinder below the piston was connected to a reser• found that similar concentrations are obtainable by adding voir filled with argon such that the argon pressure always approximately 1% hydrogen to act as a carrier gas and exceeded the He3 pressure above the piston. This pre• then freezing the hydrogen and tritium from the mixture caution insured that no nitrogen or oxygen would con• at 4.2°K. taminate the He3 and destroy its quality as a counter gas. A schematic diagram of the purification system is shown The piston was driven by compressed air admitted to the in Fig. 1. The helium cryostat was of standard design. The lower of the two cylinders. He3 was precooled in a liquid nitrogen bath above the The purification procedure was as follows. The system TO GEIOER COUNTER 6 SYSTEM FIG. 1. Schematic of He* pur• ification system. TO AIR CONTROL FOR TRANSFER PUMP DRIVE HELIUM CRYOSTAT 2 N OTES •was evacuated and tested for leaks. The cryostat was achieved with a filling of 2.5 cm Hg of helium and 10 cm cooled to liquid helium temperature with the vacuum Hg of methane. The counting was done with a Berkeley system connected to ensure that no leaks developed during scaler model 2001. The background activity was measured the cooling process. Carrier hydrogen was then added to with a He4-methane mixture. The counter had to be etched the impure He3. With the valve to the transfer pump after a few runs to remove the background due to tritium closed, the gas from the bottle was leaked slowly into the occluded at the walls. trap in the cryostat until the pressure in the system came 6.5 liters of He3 were purchased from the Monsanto to equilibrium. This usually took 2 to 3 min. The initial Corporation with a quoted tritium concentration of 1.7 pressure in the He3 bottle was slightly below atmospheric parts in 108. With the addition of 1% hydrogen to the He3, pressure. The valve to the bottle was closed and the gas in the tritium concentration was reduced to 6±2 parts in 10". the trap was transferred to the clean storage bottle using Approximately 2.5 liters of liquid helium were evaporated the transfer pump. The process was repeated until the He3 in the process. bottle was evacuated. * Research supported by a grant from Atomic Energy of Canada During the purification, samples of the purified gas were Ltd. 3 t Recipient of a National Research Council of Canada Studentship, let into a Geiger counter of 25-cm active volume and the 1961-63. activity.measured. With 1900 V applied to the central 1 E. R. Hanson, H. H. Otuski, L. Passell, W. H. Lien, and N. E. Phillips, Rev. Sci. Instr. 30, 591 (1959). electrode of the counter, a Geiger plateau of 150 V was 2 M. J. W. Elliott, Rev. Sci. Instr. 31, 1218 (1960). APPENDIX H Reprinted from THE REVIEW OE SCIENTIFIC INSTRUMENTS, Vol. 35, No. 2, 241, February 1964 Printed in U. S. A. Simple Gas Circulation Pump* K. L. ERDMAN, J. R. MACDONALD,! G. A. BEER,*;'AND D. A. AXEN| Physics Department, University of British Columbia, Vancouver 8, British Columbia, Canada (Received 10 October 1963) ' | VHE standard method of purifying gases in ionization chambers, proportional counters, and gas scintilla• tion counters is to pass the gas over hot calcium-magnesium eutectic mixtures.1'2 One usually depends on convection currents to circulate the gas through the purifier. For efficient circulation of the gas, the tubes leading to and from the purifier must be large, and thus the volume of purifier and associated connections may represent a large fraction of the total volume of the chamber. This note describes a pump with a small volume and relatively high rate of gas circulation. The details of construction of the pump are shown in Fig. 1. The cycle is as follows: (1) -The piston is shown in the position with the solenoid energized. The solenoid is energized by a pulse of 0.3-sec duration that drives the piston upwards. During this part of the cycle the upper valve is closed, pushing the gas out of the pump cylinder, and the lower valve opens to admit gas to the lower part of the pump cylinder. (2) When the voltage pulse is removed from the solenoid, the piston falls under gravity and is stopped by the spring. The lower valve closes and the upper valve opens, allowing the gas to leak into the space above the piston. The piston clearance in the cylinder is 0.002 in. and negligible leakage occurs during the stroke. The cycle is repeated once per second. The total dis• placement of gas per cycle is 4 cm3 for the size pump shown. One can thus pump 14 liters/h. The pump may be operated continuously for several months without excessive wear or loss of efficiency and has been tested at pressures up to 11 atm. SCALE - INCHES * Research supported by a grant .from Atomic Energy of Canada, FIG. 1. Gas circulation pump. All joims are silver soldered. (1) Ltd. copper tube; (2) adjustable clamp to hold solenoid; (3) soft iron cap, t Recipient of a National Research Council of Canada Studentship slotted for removal; (4) mild steel valve; (5) soft iron piston; (6) ac (1961-1963). or dc solenoid; (7) stainless steel cylinder; (8) piano wire spring; (9) X Present Address: University of Saskatchewan, Saskatoon, screw cap, slotted; (10) 3 equally spaced ports; (11) Neoprene O-ring; Saskatchewan, Canada. (12) 6 equally spaced screws; (13) mild steel valve; and (14) copper ' O. Ruff and H. Hartmann, Z. Anorg. Allgem. Chem. 121, 167 tube. (1922). 2 N. Colli and U. Facchini, Rev. Sci. Instr. 23, 39 (1952). APPENDIX I Reprinted from THK RKVIKW OK SCIENTIFIC INSTRUMENTS, Vol. 35, No. 1, 122-123, January 1964 rrintcd in U. S. A. Simple Electron Bombardment Apparatus for Evaporating Boron* K. L. EROMAN, D. AXEN,! J. R. MACDONALD,! AND L. P. ROBERTSON! Physics Department, University of British Columbia, Vancouver S, B. C, Canada (Received 22 July 1963; and in final form, September 16, 1963) ' | %HIS note describes an electron bombardment appara- 20 mA current at 20 kV. The glass tee was cooled with a tus which has proven to be successful in preparing Dayton model 2C 610 air blower. The evaporator was evaporated films of high resistivity materials. Tempera• attached to a mercury diffusion pump which produced a tures in excess of 2500°C were obtained. residual pressure of approximately 2X IO-5 mm Hg. The evaporator (Fig. 1) was enclosed in a commercially Two threaded brass rods served as filament terminals. available1 2-in. Pyrex tee connection (1). This design has One of the rods (2) was attached to the brass plate (3) at the advantages that no high voltage vacuum feedthrough the top of the glass tee. The other (4) was admitted through and no water-cooling connections are required. a Kovar seal (5). The upper brass plate was at ground The filament power was supplied by a 115-A 5-V trans• potential. Two 0.010-in.-thick tantalum plates \ in. apart, former. The high voltage was supplied by a Scientific (6), served as focusing electrodes for the electron beam. Electric model P3-238 dc power supply rated to deliver The hole through the center of these plates was XTT in. in diameter. The filament (7) consisted of a single turn of O.OlO-in.-diam tungsten wire 2 in. long. The material to be evaporated was placed in a well at the top of a carbon rod (8) \ in. in diameter and \\ in. long. For the reasons dis• cussed by Hill,2 the carbon rod was attached to a brass Sylphon bellows (9) movable in all three directions by the adjusting screw (10). The metal backing (11), onto which the film was deposited, was clamped to the grounded filament electrode approximately l^- in. above the upper tantalum plate. The high voltage lead was connected directly onto the brass plate (12) at the lower end of the glass tec. Initially, the carbon rod was outgassed by applying 5 kV to the lower plate and enough filament current to produce approximately 1 mA of electron current and then pumping the system for \ h. A typical set of operating conditions during the evaporation of boron, which requires tempera• tures above 2500°C, was filament current, 6A; accelerating voltage, 5 kV; electron current, .20-30 mA; and pressure, 4Xlfrs nun Hg. The apparatus was operated under these conditions for periods up to 3 h. Under these conditions red diffraction rings appeared on the film at a rate of one per minute. This rate of deposition was extremely sensitive to the position of the boron. The best results were obtained when the boron was situated approximately ^ in.'below the lower tantalum plate. < * Research supported by a grant from Atomic Energy of Canada Ltd. ! Recipient of a National Research Council of Canada Studentship 1961-63. 1 Pyrex tec available from Corning Glass Works, Corning, New ••'-Sifork.2 ' FIG. 1. Schematic vertical section of electron bombardment apparatus. H. A. Hill, Rev. Sci. fnstr. 27, 10S6 (1956). APPENDIX J Charged Photoparticles from Argon M.A. Reiraann, J.R. MacDonald and J.B. Warren Abstract 40 j. Charged photoparticles from A have been observed in a gridded ionization chamber irradiated with 9.17 MeV and 17.71 MeV gamma rays. Partial and total cross sections 40 39 4o 36 are given for the reactions A(r,p) Cl and A(j<,a) S. I. Introduction Because it is relatively well understood, the electromagnetic interaction suggests itself as a powerful tool for the investigation of nuclear structure. Thus quanta of radiation have the advantage over particles as participants in nuclear reactions, in that the interaction Hamiltonian is known. Over the past two decades this know• ledge has been utilized in the construction of nuclear models enjoying considerable success in accounting for the giant dipole absorption resonance observed in nuclear reactions . The advance of our understanding of the mechanism involved in photonuclear disintegration following the large amount of experimental work which has been done^ , has been hampered by the lack of a suitable source of gamma radiation, and shortcomings in instrumentation for monitoring the available sources. Bremsstrahlen have been used effectively to establish the shape of the dipole resonance, and to reveal the gross features of the energy distribution of the photo• disintegration products. The results have been used to estimate level densities in heavier nuclei, and have lead to descriptions of the disintegration process in terms of both statistical6^ and direct photoeffect7^ models. In order to assess the pertinence of these models, and in order to examine the fine structure, both of the dipole absorption cross section as a function of energy and of the energy spectrum of emitted photoparticles, the photodisintegration must be initiated by a monochromatic source. To this end the photodisintegration of argon at 17.6 MeV excitation was investigated by Wilkinson and 8) Carver 9 ] Our present knowledge of the reaction Q-values ', as pointed out by Spicer*0^, does hot admit the inter• pretation placed by Wilkinson and Carver upon their results. For this reason the reactions 40A(^,p)39Cl and 4oA(^,a)36S have been re-examined at a photon energy Of 17.71 MeV and the cross sections determined. In addition the photoalpha cross section was measured at a photon energy of 9.17 MeV. 2. Experimental The energy spectra of charged photoparticles were obtained using a gridded ionization chamber. The chamber is described by Hay and Warren11 \ but in the present case the interior was lined with a 35 um thickness of Mylar with sufficient gold evaporated onto it to reduce the resistivity of the inner surface to approximately a thousand ohms per cm. This technique was found more effective than a graphite coating for reducing background due to activity in the chamber walls. The chamber was filled with high grade welding argon containing less than 0.02% of impurities, these being oxygen, nitrogen, hydrogen and carbon dioxide. After filling, the impurity content was further reduced by convection circulation of the gas through a heated sidearm containing a calcium-magnesium ' 12) "" '" eutectic mixture With an argon pressure of 6.80 atm a few hours of purification in this manner resulted in an energy resolution of 4% for 5.15 MeV alpha particles from a Pu239 calibrating source. The calcium-magnesium eutectic contained radioactive 222 impurities which emitted the isotope Rn . This isotope was identified from the;peaks produced in the chamber back• ground corresponding to the energies predicted by the decay scheme for this nucleide. While these peaks in turn provided a useful check on the energy scale set by the calibrating source, the background activity could be reduced to negligible proportions for the duration of the experimental work by beginning the runs within a few hours of filling the chamber with fresh gas. Voltage pulses from the chamber were amplified by a Dynatron Preamplifier Unit modified to incorporate a type 7586 Nuvistor as a first stage, followed by a Dynatron Main Amplifier 1430 A. Differentiating and integrating time constants of 8 us were used for pulse shaping, and pulses were analysed in 128 channels of a Nuclear Data Model ND 103 256 channel pulse height analyser. Gamma flux measurements were made with a 6.98 cm diameter by 11.4 cm long Nal(Tl) crystal. The efficiency of this scintillation counter has been determined to within 15% at 17.7 MeV, and is known to within 10% at 9.17 MeV13-15). The gamma spectra were also monitored by a Nuclear Data Model ND 120 256 channel pulse height analyser. The 9.17 MeV gamma rays were obtained by bombarding a 13C target 30 keV thick with 1.75 MeV protons from the University of British Columbia 3 MeV Van de Graaff accelerator. The target was prepared by cracking isotopically enriched methyl iodide onto a platinum backing16K The yield of 9.17 MeV gamma rays was determined by spectrum stripping to separate out the 6.44 MeV component16). For the 9.17 MeV work, the ionization chamber was situated at 90° to the proton beam. Gamma rays of energy 17.64 MeV were obtained from the reaction ^LiCp,?')^e at the 441 keV resonance. A different target was used for each run, these being produced by evaporating metallic lithium onto a copper backing. The target thickness varied from 30 KeV to 60 keV. The ionization chamber was at 0° to the proton beam. At this position the Doppler shift is + 67 keV, giving photons of energy 17.71 MeV. x The ratio of 17.7 MeV to 14.8 MeV gamma rays was obtained by spectrum stripping. For this purpose the spectral shape of the 17.7 MeV component was assumed to be similar to that of the 17.1 MeV component from the reaction ^B(p,y)^C at a bombarding energy of 1.2 MeV. At the scintillation counter, positioned at 138° to the proton beam, the ratio of 17.6 to 14.8 MeV gamma yields was found to vary from 1,87 +_ 0.1 for a 30 keV target to 1.78 + 0.1 for a 60 keV target. These results are consistant with those obtained by Mainsbridgei7). The ratio of yields increases by approximately 4% when observed at 0° to the proton beam, due to the anisotropies involved1®). The gamma flux through the chamber was calculated accordingly. 3. Results A calculation of the cross section for the events observed requires knowledge of the efficiency of the ionization chamber. This is a geometric calculation comprising integration over the sensitive volume of the chamber to establish the mean gamma flux-path length product. This calculation was 19) carried out by Axen and Robertson 'as a function of chamber dimensions and target position. The number of events associated with each peak identified in the photoparticle energy distributions was obtained by correcting the number of counts in the peak for time dependent background, wall effect and, in the 17.7 MeV work, estimated contribution due to 14.8 MeV radiation. Wall effect is due to two causes, and results in the accumulation of counts in the spectra at energies lower than actually associated with the charged particles causing them. The most important contribution to wall effect is the wall loss. This results if the charged particles originating from a photodisintegration event do not lose all their energy by ionization of the gas in the sensitive volume of the chamber, either because of collision with the chamber wall or because they leave the sensitive volume and proceed into the dead spaces at the ends of the chamber. The other contribution to wall effect is the end effect. This refers to events taking place in the dead spaces giving rise to charged particles which enter the sensitive volume after losing some of their energy outside it. The calculation of the wall effect is also a geometric one, involving the chamber dimensions and the particle ranges, and has been treated in detail for cylindrical chambers by Robertson19). The necessary corrections to the cross sections varied from 41% for the most energetic protons to 6% for the least energetic alphas. The predominating source of uncertainty in our values for the cross sections is the gamma spectrometer efficiency. The experimental uncertainties if the gamma counter efficiency were known exactly, are also indicated with the experimental results in table 1. A considerable uncertainty, +_ 15% for the 9.17 MeV results, is due to the relatively large background subtraction involved. This uncertainty is smaller for the 17.71 MeV work. Other sources of uncertainty, such as knowledge of the angular distributions of the gamma rays, have been considered in estimating the possible error of our results, but these are much smaller than the two already mentioned and will not be discussed further. 3.1 Results at 9.17 MeV The energy distribution of charged photoparticles from argon irradiated with 9.17 MeV gamma rays is shown in fig. 1. We identify the peak at 2.29 MeV with the reaction 40 *3fi A(y,a) S leaving the residual nucleus in its ground state. We obtain a Q-value of =6.88 + 0.1 MeV for this reaction, in 9) agreement with Everling ejt aJL , who obtain -6.81 +_ 0.01 MeV. The total cross section for the reaction is 31 +_ 6 ub. 3.2 Results at 17.71 MeV The irradiation with lithium gamma rays gave the charged photoparticle spectra shown in fig. 2. Fig. 2(a) comprises spectra obtained with 6.80 atm and 3.40 atm of argon pressure respectively, and fig. 2(b) shows results obtained at 7.01 atm with improved energy resolution. The identification of peaks B and C as photoalpha peaks is confirmed by comparison of the spectra in fig. 2(a). Both give the same cross section within experimental error when wall effect corrections appropriate to alpha particles are made for peaks B and C, and appropriate to protons are made for peak D. Different assignments would not give such agreement. Because of the width of the 14.8 MeV component of the gamma flux, its contribution to the photoparticle spectrum is distributed over a large energy interval and hence does not influence the structure significantly. 20) Furthermore, the bremsstrahlung work of McPherson et al shows the 40A(y,p)39Cl cross section at 15 MeV to be down by a factor of six from that at 18 MeV. We assume that this reduction in the cross section is entirely due to the shape of the dipole resonance, so we use the same factor in computing the yield of photoalphas due to the 14.8 MeV radiation from that measured at 17.71 MeV. Thus the number of counts due to the 14.8 MeV gammas was estimated for each possible reaction by taking the number of counts due to the 17.71 MeV gammas for the same reaction corrected for background but not wall effect, reducing it by the McPherson et al factor of six, and further reducing it in proportion to the ratio of 14.8 to 17.71 MeV gamma rays. For the purpose of subtraction from the charged particle spectrum, the approximation was made that these counts were evenly distributed over a 2 MeV interval centred on the appropriate energy. In no case did this correction exceed 5%. The results at 17.71 MeV are summarized in table 1. 4. Discussion If the photonuclear process in argon proceeds largely through the formation of a compound nucleus, the ratio of partial cross sections for decay by charged particle emission through various channels should simply be that of the respective penetrabilities. It would there• fore be expected that the partial cross section for 40 39 A(JSP) cl to the ground state of °9C1 would be larger than that to individual excited states. We therefore suppose that the peak £ of fig. 2 represents transitions to more than one excited state of Cl not resolved in our energy distribution. This possibility is supported by Endt and van der Leun2*^ who show levels in ^9C1 at 0.36 and 0.8 MeV. On the other hand, the supposition that peak E represents two or more levels in ^Cl makes the partial cross section for the ground state transition appear larger than predicted from statistical considerations; larger by an extent dependent upon the spin and parity assignments for the excited states, and dependent upon how much lower than classical the Coulomb barrier height is due to fuzziness of the nuclear surface22 Since the predictions of the ratio of penetrabilities can only be considered as approximate, it is not possible to estimate to what extent this partial cross section ratio has been influenced by contributions from a direct photo- effect. The ratio of the partial cross section to the ground 36 state of S by alpha emission and to the ground state of 39 Cl by proton emission at 17.71 MeV excitation was observed to be an order of magnitude smaller than predicted from the penetrabilities. This might in part be due to a small formation factor23) for alpha particles in the compound nucleus, but this could not account for the entire dis• crepancy. In the absence of known selection rules operating 36 against the alpha transition to the ground state of S, the existence of another low lying level in ^Cl contributing to peak D of fig. 2 may be indicated. Alternatively, the cross section for alpha emission to the ground state of 36S may have been observed at an excursion below its average value, demonstrating the existance of excitation energy dependent cross section fluctuations as described by Ericson24*25). We are indebted to the National Research Council of Canada for financial support and for awarding student• ships to two of us (M.A.R. and J.R.M.). We wish to thank Dr. E. Vogt for his most helpful comments, and Mr. E.W. Blackmore for reprogramming the gamma flux-path length calculations for the ionization chamber. Figure 1 Figure 2(a) Figure 2(b) Table 1 Experimental results at 17.71 MeV Spectrum peak Reaction Remarks Q-value Cross section MeV mb 40 36 36, A A(r,a) S To ground state of - 6.6+0.4 0.08+0.05 40 36 B A(r,a) S To excited state(s) of -10.1+0.1 0.35+0.06 36s at e& 3.5 MeV 4 °A(r,«)36S To excited state(s) of -11.3+0.1 0.40+0.08 ,36s at 4.7 MeV 40., ,36c A(^,a) S Total cross section 0.83+0.16 (+ 0".09) 40 39 oq D A(r,p) Cl -12.46+0.1 1.15+0.2 To ground state of Cl 39 E 40A(r,p)39C1 -12.99+0.1 1.26+0.2 To excited states of Cl at 0.5 MeV 40 39 39 F,G A(r,p) Cl To excited states of C1 4.75+1.0 40 39 A(r,P) Cl Total cross section 7.16+1.26 (+0.5*6) The uncertainties in brackets are estimated on the assumption that the efficiency of the gamma counter is known exactly. FIGURE CAPTIONS Fig. 1: The photodisintegration of 40A at 9.17 MeV showing the photoalpha peak. The background, shown normalized to the same running time, was obtained with the proton energy off the target resonance. 40 Fig. 2: The photodisintegration of A at 17.71 MeV. The group at A and the peaks B and C we attribute to photoalpha particles leaving " S in its ground and lower excited states respectively. Peaks D and E and the group G, F are identified with proton transitions to the ground, first and higher excited states of39 C1. REFERENCES 1. D.H, Wilkinson, Physica 22 (1956) 1039 2. G.E. Brown and M. Bolsterli, Phys.Rev.Letters 3 (1959) 472 3. G.E. Brown, L. Castillejo and J.A. Evans, Nuclear Physics 22 (1961) 1 4. W. Brenig, Nuclear Physics 22 (1961) 14 5. Photonuclear Reactions, International Atomic Energy Agency, Vienna, 1964(Bibliographical Series No.10) 6. V.F. Weisskopf and D.H. Ewing, Phys.Rev. 57 (1940) 472 7. E.D. Courant, Phys.Rev. 82 (1950) 703 8. D.H. Wilkinson and J.H. Carver, Phys.Rev. 83 (1951) 466 9. F. Everling, L.A. KOnig, J.H.E. Mattauch and . . A.H. Wapstra, Nuclear Physics 18 (1960) 529 10. B.M. Spicer, Phys.Rev. 100 (1955) 791 11. H.J. Hay and J.B. Warren, Can.J.Phys. 37 (1959) 1153 12. L. Colli and U. Fachinni, Rev.Sci.Inst. 23,(1952) 39 13. P.P. Singh, Ph.D. Thesis, University of Brit.Col.(1959) 14. G.M. Griffiths, E.A. Larson, and L.P. Robertson, Can.J.Phys. 40 (1962) 402 15. E.A. Larson, M.A. Thesis, University of Brit.Col. (1957) 16. J.R. MacDonald, Ph.D. Thesis, University of Brit.Col. (1964) 17. B. Mainsbridge, Austr.J.Phys. 13 (1960) 204 18. B. Mainsbridge, Nuclear Physics 21 (1960) 1 19. L.P. Robertson, Ph.D. Thesis, University of Brit.Col. (1963) 20. D. McPherson, E. Pederson and L. Katz, Can.J.Phys. 32 (1954) 593 P.M. Endt and C. van der Leun, Nuclear Physics 34 (1960) 1 J.M.C. Scott, Phil.Mag. 45 (1954) 441 Physics of the Nucleus, M.A. Preston (Addison Wesley, Reading, Mass. 1962) p. 536 T. Ericson, Phys.Rev.Letters 5 (1960) 430 T, Ericson, Adv. ln Phys. (Supplement to the Phil.Mag.) 9 (1960) 425 . -73- BIBLIOGRAPHY Ajzenburg-Selove, F.A. and Lauritsen, T. (1959), Nuclear Physios 11, 1. Austerne, N. (1951), Phys. Rev. 83, 672. Austerne, N. (1952), Phys. Rev. 85, 147. v Batchelor, R., Aves, R., and Skyrme, T.H.R. (1955), Rev. Sci. Inst. 26, 1037. Becchi, C, Manuzio, G.E., Meneghetti, L., and Vitale, S. (1964), Physics Letters 8, 322. Berman, B.L., Koester, L.J., and Smith, J.H. (1963), Phys. Rev. Letters 10, 527. Berman, B.L., Koester, L.J., and Smith, J-»H. (1964), Phys. Rev. 135. B117. Bosch, R., Lang, J., Miller, R., and wSlfli, W. (1964), Physics Letters 8, 120. Broude, C, Green, L.L., Singh, J.J., and Willmott, J.C. (1957), Phil. Mag. 2, 499 and 1006. Brown, W.K. (1964) , Nuclear Instruments and Methods 26_, 1. Bunemann, 0., Cranshaw, T.E., and Harvey, J.A. (1949), Can. J. Research A27, 191. Collard, H. and Hofstadter, R. (1963), Phys. Rev. 151, 416. Collard, H., Hofstadter, R., Johansson, A., Parks, R., Ryneveld, M., Walker, A., Yearian, M.R., Day, R.B., n and Wagner, R.T. (1963), Phys. Rev. Letters 11, 132. Colli, L. and Facchini, U. (1952), Rev. Sci. Inst. 23, 39. Curran, S.C., Dee, P.I., and Petrzilka, V. (1939), Proc Roy. Soc. (London) 169, 269. Day, R.B. and Perry, I.E. (1951), Phys. Rev. 81, 662A. Delves, L.M. (1960), Phys. Rev. 118, 1318. Delves, L.M. (1962), Nuclear Physics 29, 268. -74- Devons, S. and Hine, M.G.N. (1949), Proc. Roy. Soc. (London) A 199, 56. Dosso, H.H. (1957), M.A. Thesis, University of Brit. Col. Eichmann, V.E. (1963), Zeits. Phys. 175, 115. Everllng, F., Konlg, L.A., Mattauoh, J.H.E., and Wapstra, A.H. (1960), Nuclear Physics 18, 529. Finckh, E«, Koslek, R., Lindenberger, K.H., Meyer-Berkhout, U., N\lcker, N., and Schl\ipmann, K. (1963), Physics Letters .7, 271. Fowler, W.A., Lauritsen, CC, and Lauritsen, T. (1948), Revs. Mod. Phys. 20, 236. Freeman, I.M. and West, D. (1962), Nuclear Physios 38, 89. Friedes, J.L. and Chrien, R.E. (1964), Rev. Sol. Inst. 35, 469. Gorbunov, A.N. and Yarfolomeev, A.T. (1963), Physics Letters 5, 149. Gorbunov, A.N. and Varfolomeev, A.T. (1964), Physics Letters 11 137. Griffiths, G.M., Larson, E.A., and Robertson, L.P. (1962), Can. J. Phys. 40, 402. Grlffy, T.A. (1964), Physics Letters 11, 155. Gunn, J.C and Irving, J. (1951), Phil. Mag. 42, 1353. Hird, B., Whitehead, C, Butler, J., and Collie, CH. (1954), Phys. Rev. 96, 702. Holmgren, H.D., Blair, J.M., Famularo, K.F.j Stratton, T.F., and Stuart, R.V. (1954), Rev. Sci. Inst. 25, 1026. Hughes, D.T. and Schwartz, R.B. (1958), Brookhaven National Laboratory Report BNL 325. Hulthen, L. and Sugawara, M. (1957), Handbuch der Physik, ed. by S. Flugge, (Springer-Verlag, Berlin) Vol 34, 1. Krueger, D.A. and Goldberg, A. (1964), Phys. Rev. to be published. Lai, M. (1961), Ph.D. Thesis, University of Brit. Col. -75- Larson, E.A. (1957), M.A. Thesis, University of Brit. Col. Lauritsen, CC. and Fowler, W.A. (1940), Phys. Rev. 58, 193. Lehman, P., Leveque, A., and Pick, R. (1956), Comptes Rendus 243, 743. Levinger, J.S. (1963), Phys. Rev. 131, 2710. McConaick, R.D. and McCormick, J.D. (1961), Nuclear Instruments and Methods 13, 147. Meister, N.T., Radha, T.K., and Schiff, L.I. (1964), Phys. Rev. Letters 12, 509, Mills, W.R., Caldwell, R.L., and Morgan, I.L. (1962), Rev.Sci. Inst. 33, 866. Phillips, G.C. and Richardson, J.E. (1950), Rev.Sci. Inst. 21, 885. Robertson, L.P. (1957), M.A. Thesis, University of Brit. Col. Robertson, L.P. (1963), Ph.D. Thesis, University of Brit. Col. Rose, H.J. (1960), Nuclear Physics 19, 113. Rose, H.J., Trost, W., and Reiss, F. (1959), Nuclear Physics 12, 510. Rose H.J., Reiss, F., and Trost, W. (1960), Nuclear Physics 21, 367. Rossetti, C. (1959)j Nuovo Cimento 14, 1171. Sachs, R.G. (1953), Nuclear Theory, (Addison-Wesley, Cambridge, Mass.) Sayres, A. and Coppola, M. (1964), Rev. Sci. Inst. 35, 431. Schiff, L.I., Collard, H., Hofstadter, R., Johansson, A., and Yearian, M.R. (1963), Phys. Rev. Letters 11, 387. Schiff, L.I. (1964), Phys. Rev. 133, B802. Seagrave, J.D. (1952), Phys.Rev. 85, 197. Segal, R.E., Daughtry, J.W., and Olness, J.W. (1961), Phys. Rev. 123, 194. -76- Sharpe, J. (1955), Nuclear Radiation Detectors, (Methuen and Co. Ltd. London). Singh, P.P. (1959), Ph.D. Thesis, University of Brit. Col. Srivastava, B.K. (1964), Phys. Rev. 133, B545. Stewart, J.R., Morrison, R.C., and O'Connell, J.S. (1964), Tale Electron Accelerator Laboratory, Internal Report. Verde, M. (1950), Helv. Phys. Acta 23, 453. Verde, M. (1951) Nuovo Cimento 8, 1. Warren, J.B., Erdman, K.L., Robertson, L.P,, Axen, D.A., and MacDonald, J.R. (1963), Phys. Rev. 152. 1691, and Bull. Amer, Phys. Soc. 8, 124. Whaling, W. (1956), Handbuch der Physik, ed. by S. Flugge, (Springer-Verlag, Berlin) Vol. 34, 193. Wilkinson, D.H,. Barnes, CA,, Carver, J.H,, and Stafford, G.H. (1952), Phys. Rev, 86, 359. Wilkinson, D.H, and Bloom, S.D. (1957), Phil. Mag. 2, 63. Woodbury, H.H., Day, R.B., and Tollestrup, A.V. (1953), Phys. Rev. 92, 1199.