Oxidation of Silicon by Ion Implantation and Laser Irradiation S

Oxidation of silicon by ion implantation and laser irradiation S. W. Chiang, Y. S. Liu, and R. F. Reihl

Citation: Appl. Phys. Lett. 39, 752 (1981); doi: 10.1063/1.92879 View online: http://dx.doi.org/10.1063/1.92879 View Table of Contents: http://apl.aip.org/resource/1/APPLAB/v39/i9 Published by the American Institute of Physics.

Related Articles The interface and surface effects of the bicrystal nanowires on their mechanical behaviors under uniaxial stretching J. Appl. Phys. 108, 074311 (2010) Polarization screening and induced carrier density at the interface of LaAlO3 overlayer on SrTiO3 (001) J. Appl. Phys. 108, 013701 (2010) Behavior of oxygen vacancies in BiFeO3/SrRuO3/SrTiO3(100) and DyScO3(100) heterostructures Appl. Phys. Lett. 94, 132905 (2009) Unsteady nanoscale thermal transport across a solid-fluid interface J. Appl. Phys. 104, 064306 (2008) In situ reflection high energy electron diffraction study of dehydrogenation process of Pd coated Mg nanoblades J. Appl. Phys. 104, 033534 (2008)

Additional information on Appl. Phys. Lett. Journal Homepage: http://apl.aip.org/ Journal Information: http://apl.aip.org/about/about_the_journal Top downloads: http://apl.aip.org/features/most_downloaded Information for Authors: http://apl.aip.org/authors

Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions Oxidation of silicon by ion implantation and laser irradiation s. W. Chiang, Y. S. Liu, and R. F. Reihl General Electric Corporate Research and Development. Schenectady. New York 12301 (Received 20 July 1981; accepted for publication 20 August 1981)

Q-switched laser irradiation has been used to anneal 0+ -implanted silicon to form Si02 layers. Results of differential Fourier-transformed infrared spectroscopy and transmission electron microscopy confirm the formation of oxide layers. Segregation of oxygen toward the surface was observed by secondary ion mass spectroscopy and correlated with resolidification velocities, which vary as a function oflaser energy densities. PACS numbers: 68.20. + t, 73.60.Hy, 79.20.Ds, 81.20.Cs

Among silicon integrated circuit processes, thermal depth profile. C/ ions were used as the primary ion source oxidation requires a processing temperature which is much to increase the oxygen yield, higher than those of other process steps. High-temperature The surfaces of laser-annealed regions exhibited green thermal oxidation causes stress, stacking faults, and dopant and purple colors, The diameter of the color-changed area redistribution in the materials employed and degrades the was about I mm, This was a visual indication that layers of device performance. Low-temperature silicon oxidation different structures were produced as a result of laser irra- techniques, such as high-pressure oxidation and plasma oxi dation, have been reported. 1-7 These two methods, as well as UNIMPLAN ED conventional thermal oxidation can be considered as being a single-step oxidation process, namely, the introduction of a 0+ IMPLANTED atom into Si and the reaction of a atom with Si are accom > plished simultaneously. An alternative method is an oxida c:r: « 2.0 J em- 2 tion process in which the introduction of 0 atom and the 0:: ~ CD reaction of 0 atom with Si take place in sequential steps. For cr « 2.3J em-2 example, using ion implantation, it has been shown that ox w u z ide can be formed by oxygen ion implantation followed by « l thermal annealing. X,9 Values of oxide dielectric constant, ox I- 2.5 J em-2 ::::!; (f) ide mobile charge density, oxide semiconductor fixed charge z « 0:: density, and oxide breakdown strength for the high-dose I- 900°C-annealed samples have been shown to be in excellent agreement with the thermally-grown oxides. 9 However, the anneal temperature required after ion implantation is as high as the conventional oxidation temperature. In this letter, we 1600 1400 1200 1000 800 600 400 (al 1 report a study of pulsed laser annealing of oxygen-implanted WAVE NUMBER ( em· ) Si layers to form an oxide layer. The main advantages oflaser annealing, namely the annealing of surface layers without 0+ IMPLANTED AND LASER ANNEALED appreciable heating of the whole wafer and the extremely MINUS 0+ IMPLANTED short heat treatment time, should ameliorate the problems ~ due to conventional thermal oxidation described above. 0:: « 0:: Specimens were prepared by implanting 50 ke V 0 + I- 2 CD 2.5Jem- ions into P-type (100) Si wafers at liquid nitrogen tempera 0:: « 17 2 ture. A fluence of 1 X 10 ions cm- was used with a low w u z 2 dose rate of ~ 3 f1A em - 2 to produce a continuous amor « 2.3 J em- CD 0:: phous Si surface layer as observed by transmission electron 0 22 3 (/) microscopy. A peak concentration of -1 X 10 ions cm- «CD 2.0Jem-2 at a depth of 0.1 f1m was produced. Laser annealing was carried out with a single 70-nsec Nd:Glass laser pulse at 1.06 f1m with a spot size of ~ 3 mm. The laser energy density was 2 varied in the range of 2.0--2.5 J cm- • Following laser irra diation, the samples were examined using differential Four 1600 1400 1200 1000 800 600 400 (b) ier-transform infrared (FTIR) spectroscopy. The micro WAVE NUMBER (em· l ) structure of irradiated samples was studied using the FIG. I. (al Fourier-transformed infrared spectra of 0 ' -implanted and laser transmission electron microcope (TEM). Secondary ion irradiated Si. (b) Differential FTIR spectra oflaser irradiated 0 • -implanted mass spectroscopy (SIMS) was used to examine the oxygen Si, using an 0+ -implanted sample as the reference spectrum.

Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions a - 0+ IMPLANTED, 50keV, I x10 17 10NS/cm 2 b -LASER ANNEALED, 2.0 J/cm 2 1023 c - LASER ANNEALED, 2.3 J/ cm 2

:z o >= ~ 10 20 Iz: u"-' :z 810 19 :z "-' <-? ~ X 0 18 10

10 17 L-----'_--:-'-_-:-'-::_-,'--_-"--_..L-_.l-~ 0.1 0.2 0.3 0.4 0.5 06 0.7 0.8

DEPTH (I'm J

FIG. 3. SIMS oxygen concentration profiles of 0+ -implanted and laser irradiated Si at 2.0 and 2.3 J cm - 2.

irregular features in Fig. 2(a) lie in the range of 10-50 nm, which corresponds to the impurity lateral diffusion length during the period in which solidification took place. The thickness variations of the oxide film have been confirmed by matching the dark and bright areas in the bright-field and FIG. 2. (al TEM bright-field image of silicon oxide film produced by 0+ dark-field images. One example of dark-field images at a implantatiaon and 2.5 J cm - 2 laser annealing. (b) TEM dark-field image higher magnification is shown in Fig. 2(b). In addition to the obtained from the diffuse scattering at 417 sin(O / A I::::: 1.6 A- '. light areas of 10--50 nm diameter, there are many tiny bright dots of -0.5 nm. Whether these tiny dots are clusters of diation. Differential FfIR was carried out to show the char small crystalline materials or arise from local fluctuations of acteristic Si-Q bond ofSi02• As shown in Fig. l(a), the ab silicon/oxygen concentrations could not be determined in sorption band between 900 and 1080 cm -I in 0+ -implanted this study using the conventional TEM technique. sample was shifted towards 1080 cm - I as the laser energy SIMS depth profiles of oxygen as a function of laser density increased from 2.0 to 2.5 J cm -2. By taking differen energy densities indicated surface segregation had occurred. tial FfIR spectra using an 0+ -implanted unannealed sam Figure 3 shows that the peak concentration increases from ple as the reference, the effects of laser irradiation on im 1 X 1022 cm-3 in the as-implanted sample to 1.2x 1022 and 22 planted samples could be distinguished. The differential 2.5 X 10 cm -3 in the laser-irradiated sample at 2.0 and 2.3 FTIR spectra showed an IR band at 1080 cm -I [Fig. l(b)], J cm -2, respectively. A SIMS depth profile of thermal oxide due to the Si-Q stretching mode ofSi02• 10 A broader shoul grown at 9oo·C was carried out for comparison and an oxy 22 3 der extending to 1250 cm -I was also observed, and became gen peak concentration of -2X 10 cm- was observed. more pronounced at 2.5 J cm -2. A discussion of this broader This confirms the formation ofSi02 in 0+ -implanted and shoulder band has been reported elsewhere. II laser-annealed silicon layer. In Fig. 3, the shift of the peak

The formation of Si02 was further confirmed using positions toward the surface suggests that oxygen atoms seg transmission electron diffraction. The diffraction patterns of regate during the melt and resolidification process by the laser-annealed samples contained halos of amorphous mate zone-refining effect. The peak concentration in the laser irra rials. The two distinguished halos, 41T sin (J / A = 1.6 and 5.1 diated sample at 2.3 J cm-2 was twice as high as that of the A - I, where (J is the scattering angle and A is the wavelength sample laser-annealed at 2.0 J cm - 2. The segregation ofoxy of incident beam, correspond closely to those of amorphous gen is understood by considering the variation of the liquid 12 Si02 reported byMozzi and Warrent. Some ill-defined ha solid interface velocity during recrystallization as a function los between 2.0 and 3.5 A - I were also observed in some of the annealing pulse energy. 15.16 The calculation shows the areas. By comparison, only amorphous Si was found in the average regrowth velocity at 2.0 J cm -2 is - 2 m/sec, which 2 0+ -implanted samples prior to laser annealing. The TEM is 15% higher than that at 2.3 J cm- • When the regrowth bright-field image of the oxide layer, Fig. 2(a), shows an ir velocity is reduced, less solute is trapped in the solid silicon; regular-shaped granular structure. This structure does not consequently, more oxygen atoms are observed to segregate appear to be similar to the regular cellular structure pre to the surface. 17 The tails of the depth profiles of the laser viously reported due to constitutional supercooling occur annealed samples shown in Fig. 3 reflect faster regrowth ve 13 14 ring in the resolidifying meit. • However, the sizes of the locity in the early stage of resolidification at the maximum

753 Appl. Phys. Lett., Vol. 39, No.9, 1 November 1981 Chiang, Liu, and Reihl 753

Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions molten depth. (The tail, between 0.22 and 0.35 /Lm, is more chern. Soc. 122, 1409 (1975). pronounced in the laser-annealed sample at 2.0 J cm- 2 'R. Champagne and M. Toole, Solid State Techno\. 20,62(1977). shown in Fig. 3.) 41. R. Ligenza, 1. App\. Phys. 36, 2703 (1965). '1. Kraitchman, 1. App\. Phys. 38, 4323(1967). An interesting observation is noted: a poly crystalline Si "M. A. Copeland and R. Pappu, App\. Phys. Lett. 19,199 (1971). layer was observed with TEM between the single-crystal Si 70. L. Pu1frey, F. G. M. Hathorn, and L. Young, 1. Electrochem. Soc. 120, substrate and the oxide layer formed after laser irradiation. 1529(1973). "1. Dylewski and M. C. 1oshi, Thin Solid Films 37,241 (1976). The existence of the polycrystalline Si layer suggests that "M. H. Badaw and K. V. Anand, 1. Phys. 0 10,1931 (\977). incomplete epitaxial regrowth which occurs even as the laser "'1. Wong, 1. App\. Phys. 44,5629 (\973). 2 energy density increases to 2.5 J cm- • It has been shown "Y. S. Liu, S. W. Chiang, and F. Bacon, App\. Phys. Lett. 38,1005(1981). that oxygen impurities could retard epitaxial regrowth in the lOR. L. Mozzi and B. E. Warren, 1. App\. Crystallogr. 2, 164(1969). "B. Chalmers, Principles of Solidification (Wiley, New York, 1964). solid epitaxy case. IX The role of 0 atom in the liquid-phase '4G. 1. van Gurp, G. E. 1. Eggermont, Y. Tamminga, W. T. Stacy, and 1. R. epitaxial growth ofSi, however, has not yet been well under M. Gijsbers, App\. Phys. Lett. 35, 273 (1979). stood and needs further investigation. "c. M. Surko, A. L. Simons, D. H. Auston, 1. A. Golovchenko, R. E. We would like to thank R. S. McDonald and D. V. Slu~her, and T. N. C. Venkatesan, App\. Phys. Lett. 34, 635(1979). ,oA. G. Cullis, H. C. Webber, 1. M. Poate, and A. L. Simons, Appl. Phys. Temple for FTIR spectrometric measurements, W. Katz for Lett. 36, 320 (1980). SIMS analysis, and L. M. Levinson for critical reading of the l7y. S. Liu, S. W. Chiang, and F. Bacon, "Laser and Electron-Beam Solid manuscript. Helpful discussions with J. W. Mayer are also Interactions and Material Processing," in the Proceedings ofthe Materials Research Society Annual Meeting, Vol. 1, Boston, Mass., 1980, edited by 1. acknow ledged. F. Gibbons, L. D. Hess, and T. W. Sigmon (North-Holland, New York, 1981), p. 117. '1. R. Ligenza and W. G. Spitzer, 1. Phys. Chern. Solids 14,131 (1960). '"E. F. Kennedy, L. Csepregi, and 1. W. Mayer, 1. App\. Phys. 48, 4241 lR. 1. Zeto, C. G. Thornton, E. Hryckowiam, and C. D. Bosco, 1. Electro- (1977).

754 Appl. Phys. Lett., Vol. 39, No.9, 1 November 1981 Chiang, Liu, and Reihl 754

Downloaded 05 Feb 2012 to 140.114.195.186. Redistribution subject to AIP license or copyright; see http://apl.aip.org/about/rights_and_permissions