Open Collections will be undergoing maintenance Monday June 8th, 2020 11:00 – 13:00 PT. No downtime is expected, but site performance may be temporarily impacted.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Annealing of GaA1As double heterostructures with homogeneous ruby laser light Brett, Michael Julian 1981

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1981_A6_7 B74.pdf [ 2.88MB ]
Metadata
JSON: 831-1.0085219.json
JSON-LD: 831-1.0085219-ld.json
RDF/XML (Pretty): 831-1.0085219-rdf.xml
RDF/JSON: 831-1.0085219-rdf.json
Turtle: 831-1.0085219-turtle.txt
N-Triples: 831-1.0085219-rdf-ntriples.txt
Original Record: 831-1.0085219-source.json
Full Text
831-1.0085219-fulltext.txt
Citation
831-1.0085219.ris

Full Text

ANNEALING OF GaAlAs DOUBLE HETEROSTRUCTURES WITH HOMOGENEOUS RUBY LASER LIGHT V ^ / MICHAEL JULIAN BRETT B . S c , Queen's U n i v e r s i t y , 1979 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF THE FACULTY OF GRADUATE STUDIES (Department of P h y s i c s ) We accept t h i s t h e s i s as conforming to the r e q u i r e d standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL 1981 (c) M i c hael J u l i a n B r e t t , 1981 by MASTER OF SCIENCE i n In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the U n i v e r s i t y o f B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the head o f my department or by h i s o r her r e p r e s e n t a t i v e s . I t i s understood t h a t c o p y i n g or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . Department o f Physics The U n i v e r s i t y of B r i t i s h Columbia 2075 Wesbrook P l a c e Vancouver, Canada V6T 1W5 D a t e March, 31, 1981 ABSTRACT The output from a ruby laser was homogenized, and used to laser anneal the active layer of GaAlAs double heterostructure wafers in an attempt to improve the radiative e f f i c i e n c y of the active layer. At anneal energies exceeding the estimated threshold for melting of the active layer, the radiative e f f i c i e n c y was reduced by a factor of two. Subsequently, semiconductor laser diodes fabricated from laser annealed heterostructures performed much worse than those fabricated from unannealed heterostructures. i i i TABLE•OF CONTENTS INTRODUCTION . . 1 CHAPTER TWO . . .' 5 2.1 History Of Annealing 5 2.2 Anneal Mechanisms 6 CHAPTER THREE 10 3.1 Ruby Laser 10 3.2 Laser Beam Homogenization 15 3.3 Annealing Technique 21 3.4 Photoluminescence Apparatus 24 CHAPTER FOUR 27 4.1 Heterostructure And Laser Diode Fabrication 27 4.2 Previous Annealing Experiments 31 4.3 Data 35 4.4 Data Analysis 43 CHAPTER FIVE ." 45 5.1 Amorphous S i l i c o n Films 45 5.2 Aluminum On Gallium Arsenide 48 CONCLUSION 51 APPENDIX A 53 APPENDIX B 56 REFERENCES 60 i v LIST OF TABLES Table I 42 LIST OF FIGURES Figure 1 4 Figure 2 1 2 Figure 3 1 2 Figure 4 • 1 3 Figure 5 -1-3 Figure 6 1 4 Figure 7 18 Figure 8 1 8 Figure 9 1 9 Figure 10 1 9 Figure 11 2 ^ Figure 12 2 3 Figure 13 2 6 Figure 14 2 9 Figure 15 2 9 Figure 16 30 Figure 17 3 3 Figure 18 3 4 Figure 19 3 8 Figure 20 3 9 Figure 21 • 4 0 Figure 22 4 1 Figure 23 4 7 Figure 24 5 0 Figure 25 • 5^ Figure 26 5 8 Figure 27 5 9 v i ACKNOWLEDGEMENTS I would l i k e to thank Dr. R.R. Parsons, my t h e s i s s u p e r v i s o r , and Dr. J.A. Rostworowski, for t h e i r continued support, advice and a s s i s t a n c e . I would a l s o l i k e to extend thanks to CM. Look, A.J. Springthorpe, J.C. Dyment and W.D. Westwood of Be l l - N o r t h e r n Research, f o r t h e i r c o l l a b o r a t i o n i n t h i s t h e s i s p r o j e c t and for t h e i r p a t i e n c e . F i n a l l y , thanks i s due to a number of f e l l o w students, p a r t i c u l a r l y U.O. Z i e m e l i s , R. McMahon, J . A f f i n i t o and J . Dahn, who were always w i l l i n g to give a s s i s t a n c e when asked. This work was funded i n part by grant 67-7804 from the Nat u r a l Sciences and Engineering Research C o u n c i l of Canada. 1 CHAPTER ONE; INTRODUCTION The use of high powered lasers to improve the surface properties of semiconductors was f i r s t demonstrated in 1974 [1]. Since then, pulsed and continuous wave lasers have been used to r e c r y s t a l l i z e amorphous semiconductor films [2], enhance the e l e c t r i c a l a c t i v i t y in ion-implanted material [3], create metal s i l i c i d e films on s i l i c o n [4], and regrow ion bombarded s i l i c o n to a single c r y s t a l [5]. In 1979, i t was shown that homogenization of the intensity p r o f i l e of a pulsed laser beam led to improved results from a laser anneal of s i l i c o n [6]. The intensity p r o f i l e of an unhomogenized beam may show fluctuations of up to 75% due to the multimode nature of radiation emitted from the resonant cavity of a laser [7]. In 1978, Parsons, Rostworowski, and Hutcheon of the University of B r i t i s h Columbia annealed single c r y s t a l GaAs with ruby laser l i g h t and generated an increase in the radiative e f f i c i e n c y of the band edge emission [8]. The success of t h i s work stimulated an extension of the experiment to include laser annealing of GaAlAs double heterostructure (DH) wafers, grown by collaborators Springthorpe, Dyment and Look of Bell-Northern Research (BNR) in Ottawa [9]. It was hoped that laser annealing would improve the radiative e f f i c i e n c y of the active layer, an ess e n t i a l component of the heterostructures. At BNR, semiconductor laser diodes are fabricated from DH wafers. The active layer of a DH wafer, shown in Figure 1, i s where stimulated emission of l i g h t occurs during laser diode operation [10]. Ruby laser i r r a d i a t i o n of a DH wafer i s predominantly absorbed in the active layer as shown [11]. At 2 high ruby laser powers, i t was hoped that only the active layer would be annealed, leading to improved radiative e f f i c i e n c y and c r y s t a l quality of the active layer and. enabling subsequent fa b r i c a t i o n of a superior semiconductor laser diode. However, the radiative e f f i c i e n c y of the DH material was d r a s t i c a l l y reduced by laser annealing. It i s now believed that damage to the active layer was induced partly through use of an inhomogeneous ruby laser beam. The main objective of th i s thesis study was to homogenize the output of a ruby laser and, thereby, to attempt an improvement of GaAlAs DH material through laser annealing. Photoluminescence, the luminescence generated by o p t i c a l e x c i t a t i o n , was used to analyse the e f f e c t s of a laser anneal. It i s an i d e a l l y suited technique, being non-destructive and supplying information such as radiative e f f i c i e n c y , bandgap, and dopant type. Fabrication, by BNR, of semiconductor laser diodes from annealed and unannealed DH material provided an "acid t e s t " of the usefulness of the laser anneal. Secondary research e f f o r t s included in t h i s report are regrowth of p o l y c r y s t a l l i n e s i l i c o n from amorphous s i l i c o n films on glass and an attempt to produce GaAlAs from laser annealing of Al films on GaAs substrates. These two investigations were not pursued in depth and are described only b r i e f l y . The structure of th i s report i s as follows. Chapter two describes the history and evolution of laser annealing and the physical mechanisms associated with the anneal. Chapter three outlines the apparatus common to a l l experiments, concentrating on the ruby laser, photoluminescence apparatus, and the 3 technique for beam homogenization. Chapter four describes the experiment with the heterostructures in d e t a i l . It includes spectral data, laser diode performance data, and explanation of the r e s u l t s . Chapter f i v e gives a brief description of the secondary experiments involving Al films and amorphous Si films and Chapter six serves as a summary and conclusion of thi s report. 4 RUBY LASER IRRADIATION I t 1 1 1 1 1 1 1 j jm \ CONFINING ! ! 1 LAYER t J * G a . 6 A I 4 A s | t t / 0 . 2 j jm > ACTIVE v///' // L A Y E R 'ABSORPTION^ G a 93 A l 07 A s V////X 1 Lim CONFINING LAYER G a 6 A I . 4 A s SUBSTRATE G a A s Figure 1 . Structure of a double heterostructure wafer fabricated at Bell-Northern Research. Ruby laser l i g h t i s p r e f e r e n t i a l l y absorbed in the active layer. 5 ' CHAPTER TWO: LASER ANNEALING  2.1 History of Annealing Conventional thermal annealing of semiconductors and thin films involves heating the whole sample in an oven for times of the order of 30 minutes [12]. Typical annealing temperatures are 900°C for Si and 600°C for Ge [12]. The disadvantages of a thermal anneal are the growth of dislocations at high temperatures, the re d i s t r i b u t i o n of dopant concentration p r o f i l e s , warping and buckling of thinner samples, damage to the underlying substrate material, and the length of time required to process samples. Annealing with high powered lasers provides a means to introduce a large amount of energy to a l o c a l i z e d area in a short time and, therefore, i s an a t t r a c t i v e alternative to thermal annealing. The f i r s t annealing application of lasers was made in 1974 by Russian workers [1] to repair surface implantation damage in ion-implanted single c r y s t a l s i l i c o n . In subsequent work [3,5,13], annealing with Q-switched laser pulses was shown to be most e f f e c t i v e in stimulating c r y s t a l regrowth in a damaged, implanted region and in e l e c t r i c a l l y a c t i v a t i n g the implanted species. Due to the direc t application of laser annealing to the semiconductor processing industry, an explosion of research into both Q-switched and continuous wave (cw) laser processing has occurred [14], Examples of the broadening applications of laser processing are laser doping of semiconductors above their s o l i d s o l u b i l i t y l i m i t [15], laser thin f i l m reactions forming metal s i l i c i d e s [4], and laser a l l o y i n g of ohmic contacts to GaAs [16]. 6 P a r a l l e l to the above research has been the development of fl a s h lamp annnealing [17,18], with the use of a single incoherent l i g h t pulse from a high power flash lamp, and pulsed electron beam annealing [19,20], by a short (50 ns) pulse of 10 to 50 keV electrons. The rapid progress and interest in the f i e l d i s well i l l u s t r a t e d by the market a v a i l a b i l i t y of laser and pulsed electron beam wafer processing systems [21], already in use by semiconductor manufacturers. 2.2 Anneal Mechanisms The mechanism of a thermal anneal i s s o l i d phase epitaxy [22]. Enough thermal energy i s available to enable d i f f u s i o n of atoms to their l a t t i c e positions and impurities to their s u b s t i t u t i o n a l or i n t e r s t i t i a l p ositions. Crystal defects a r i s i n g from vacancies or di s l o c a t i o n s are removed through migration of atoms. This process i s well understood. Annealing of semiconductors using scanning cw lasers i s also f a i r l y well understood [23]. The model of thermal s o l i d phase regrowth is supported by data obtained using r e f l e c t i v i t y [24] and Rutherford backscattering measurements [25], which show the absence of a l i q u i d phase and the presence of an intermediate p a r t i a l regrowth stage. Also, no r e d i s t r i b u t i o n of implanted dopants occurs during the anneal [26]. An a n a l y t i c a l model has been presented [23] for s o l i d phase reactions induced by a scanning cw laser and interprets the effect of a laser in terms of a thermal anneal, with an e f f e c t i v e temperature and an e f f e c t i v e anneal time. Good agreement has been achieved between th i s model and experiment for the regrowth of amorphous Si and 7 the formation of Pd 2Si from a f i l m of Pd on a s i l i c o n substrate. The annealing mechanism during a pulsed laser anneal of semiconductor material i s not as well understood nor i s i t without controversy. Two models have been presented; 1) that annealing occurs through melting and l i q u i d phase regrowth [27,28] and 2) that the high intensity radiation creates a dense electron-hole plasma in the material, subsequently promoting i n t e r s t i t i a l migration [29]. A t y p i c a l pulsed laser anneal experiment would involve annealing an ion-implanted s i l i c o n wafer with a Q-switched ruby or yttrium-aluminum-garnet (YAG) 20 ns, 1 to 3 J/cm2 pulse [30,31]. In the melting model, the l i q u i d phase e p i t a x i a l (LPE) regrowth occurs once the implanted region melts and wets the underlying single c r y s t a l substrate [30]. Single c r y s t a l i s grown from the substrate c r y s t a l seed. When amorphous films on an amorphous substrate, such as Si on glass, are annealed, i t i s found that polycrystals are grown, not single c r y s t a l [32]. This observation supports the model of LPE regrowth from an underlying c r y s t a l seed. Direct evidence for melting has been produced by monitoring sample r e f l e c t i v i t y during a pulsed anneal [33]. A s o l i d - l i q u i d - s o l i d phase t r a n s i t i o n was found. Pulsed laser i r r a d i a t i o n i s absorbed by electron excitations near the surface of the sample, and the electron energy decays by c o l l i s i o n s with free electrons and thermal vibrations in times of the order of 1 ps. Energy i s transferred to the l a t t i c e in a c h a r a c t e r i s t i c time of about 100 ps [34]. Since the pulse duration, about 20 ns, i s much longer than.the heat transfer times, the laser l i g h t may be considered as a 8 dire c t heat source. Many authors [8,30,35] have used a one-dimensional heat equation to model the thermal e f f e c t s of a laser. An outline of such a c a l c u l a t i o n by Hutcheon [35] i s given in Appendix A. The non-thermal mechanism for pulsed laser annealing was o r i g i n a l l y proposed by Soviet workers in 1974 [36]. The high intensity laser excitation leads to a c a r r i e r density greater than 10 1 8 cm - 3 [37]. The excitation of c a r r i e r s i s known to strongly enhance the rate of d i s l o c a t i o n glide, point defect migration and i n t e r s t i t i a l d i f f u s i o n [30]. These enhancements are due to a weakening of the l a t t i c e , created by the large density of c a r r i e r s jumping across the energy gap from bonding to antibonding states. Since an amorphous phase may be viewed as c r y s t a l l i n e with a large d i s l o c a t i o n density [38], the accelerated quenching of dislocations by a hot e-h plasma generates the observed defect-free c r y s t a l formation during a pulsed anneal. This model i s supported by recent pulsed Raman temperature measurements of laser-heated c r y s t a l l i n e s i l i c o n [39-]. The l a t t i c e temperature was measured by Raman scattering from the 520 cm o p t i c a l phonon l i n e in s i l i c o n and found to be only 300°C for annealing power densities up to 1 J/cm 2. As these power densities would cause some amorphous to c r y s t a l l i n e conversion, the data supports a non-thermal mechanism of pulsed laser annealing. One could delve far deeper into the annealing controversy than has been done here. Throughout the report, the word "melting" i s continually used, showing the author's s l i g h t bias towards the thermal model. But, regardless of the regrowth 9 mechanism, there i s c e r t a i n l y a wealth of evidence demonstrating the use and p r a c t i c a l i t y of l a s e r a n n e a l i n g . The s u c c e s s f u l a p p l i c a t i o n s of l a s e r a n n e a l i n g have, i n p a r t , j u s t i f i e d the experiments o u l i n e d i n Chapters four and f i v e . 10 CHAPTER THREE: APPARATUS AND EXPERIMENTAL TECHNIQUE  3.1 Ruby Laser A l l laser annealing described in t h i s thesis was performed using the Korad model K-1QP ruby laser system shown in Figure 2. The Korad laser cooling unit c i r c u l a t e s d i s t i l l e d water through the laser head at a rate of 2 gal/min and regulates the temperature at 80 ± 1.5°F. The Korad Kl power supply consists of a high voltage power supply that charges a capacitor bank to an adjustable voltage between 3 and 5 kV. When the laser i s f i r e d , t h i s energy i s dumped to the xenon flash lamp, seen in the laser head in Figure 3. This h e l i c a l flash lamp produces uniform o p t i c a l pumping of the ruby rod, creating a population inversion of chromium ions in the rod and subsequent stimulated emission. The ruby laser has two modes of operation at 693 nm; free-running and Q-switched. In the free-running mode, the dye c e l l i s f i l l e d with methanol and becomes a simple 99% r e f l e c t i n g mirror. In t h i s mode, stimulated emission starts about 0.5 ms after the start of the pumping fl a s h and l a s t s about 0.5 ms [40], The emission depopulates the excited chromium impurities faster than the xenon pumping excites them, so the laser process must wait u n t i l population inversion i s again achieved. This results in the output shown in Figure 4, consisting of a large number of spikes, each spike l a s t i n g about a microsecond. To passively Q-switch the laser, cryptocyanine, a saturable dye, i s d i l u t e d with methanol in the dye c e l l . When emission occurs, the dye molecules i n i t i a l l y absorb the radiation and i n h i b i t laser operation. As the gain overcomes the absorption, 11 laser action begins, bleaching the dye molecules. The stored energy i s released in a giant pulse about 30 ns in duration, as shown in Figure 5. The maximum energy output in t h i s mode i s about 1 joule. Measurement of pulse energy and pulse shape may be made simultaneously as shown in Figure 6. A glass s l i d e r e f l e c t s 1/6 of the laser l i g h t to an MgO diff u s e r e f l e c t o r . This r e f l e c t i o n i s detected by a Korad K-Dl fast photodiode and may be viewed on a storage oscilloscope. Figures 4 and 5 were generated from t h i s procedure. A Korad K-J2 calorimeter absorbed the remaining 5/6 of the l i g h t , with the calorimeter response monitored by a nu l l voltmeter and displayed on a chart recorder. The c a l i b r a t i o n factor of the calorimeter, 55 J/mV, allowed c a l c u l a t i o n of the laser pulse energy. At energies less than 1 joule, the accuracy of the energy measurement decreases to about ±10% due to li m i t a t i o n s of the voltmeter and continual d r i f t i n g of the calorimeter output, created by fluctuations in the room temperature. The calorimeter i t s e l f was encased in insulating styrofoam, which helped reduce t h i s d r i f t i n g . 12 CAPACITOR B A N K HV P O W E R S U P P L Y T F L A S H L A M P " L A S E R H E A D L A S E R C O O L E R F i g u r e 2. A schematic of the Korad ruby l a s e r system. -99°/o R E F L E C T I N G M I R R O R S A P P H I R E O U T P U T R E F L E C T O R F i g u r e 3 . The ruby l a s e r head and output m i r r o r s . 0.25 TIME (ms ) 0.5 F i g u r e 4. Output of the ruby l a s e r o p e r a t i n g i n the fr e e - r u n n i n g mode. The l e n g t h of each i n d i v i d u a l pulse i s about l u s . Q - S W I T C H E D PULSE F i g u r e 5 . Output of the ruby l a s e r o p e r a t i n g i n the Q-switched mode. Photodiode Calor imeter Ruby Laser Lens Voltmeter \ Chart Recorder gure 6 . C o n f i g u r a t i o n used f o r measurement of ruby l a s e r p u l s e energy and shape. The g l a s s s l i d e t r a n s m i t s 5/6 of the l i g h t to the c a l o r i m e t e r . 15 3.2 Laser Beam Homogenization In the Introduction i t was mentioned that one of the motivations for this thesis project was the r e a l i z a t i o n that the ruby laser beam was very inhomogeneous in cross-section and that inhomogeneities led to detrimental e f f e c t s in annealing of thin films. Figure 9 shows how a f i l m of amorphous s i l i c o n on a glass substrate has been damaged by the multi-modal structure of an inhomogeneous beam. The numerous attempts to homogenize the beam are outlined below, along with the successful technique involving a bent d i f f u s i n g rod. Two standard techniques for homogenization were discounted almost immediately. They were; passing the beam through a plate of d i f f u s i n g glass and, placing a pinhole aperture at the centre of the lasing cavity to eliminate a l l but the lowest order mode. The former method resulted in very prominant speckle, beam divergence and i n s u f f i c i e n t mixing of the beam. The l a t t e r technique resulted in an intensity cross-section roughly Gaussian in shape, but with a drop in energy of an order of magnitude that reduced the maximum power available in the Q-switched mode to 0.2 J, too l i t t l e for the experiments outlined in the next chapters. A two foot section of an incoherent fibr e optics l i g h t guide, consisting of 1500 individual o p t i c a l f i b r e s in a bundle, was used as a beam mixer, coupling the laser beam in one end and out the other. Unfortunately, there appeared to be correlations between bundles of fibres at opposite ends, providing i n s u f f i c i e n t homogenization. This technique could s t i l l hold merit i f a well designed incoherent bundle were used. 16 Another attempt i n v o l v e d u s i n g a p l e x i g l a s s l i g h t p i p e , as suggested by Whitehead [41] and shown i n F i g u r e 7. The pipe had h i g h l y p o l i s h e d s u r f a c e s , e n a b l i n g many repeated i n t e r n a l r e f l e c t i o n s of a l a s e r beam t r a v e l l i n g down the pipe i n a d i r e c t i o n j u s t o f f the h o r i z o n t a l a x i s . The r e s u l t , not so s u r p r i s i n g i n h i n d s i g h t , was a myriad of images of the beam, o b v i o u s l y unacceptable. The f i n a l s o l u t i o n was a bent quartz d i f f u s i n g rod s i m i l a r to that used by C u l l i s et a l . [7], shown i n F i g u r e 8. The d i f f u s e r was f a b r i c a t e d from a s e l e c t e d , d e f e c t - f r e e , c i r c u l a r q u artz rod. The rod was bent a f t e r h e a t i n g with an oxygen-propane t o r c h , p o l i s h e d as w e l l as p o s s i b l e with s u c c e s s i v e l y f i n e r diamond g r i t on the output end and roughened with 30yum diamond g r i t on the input end. The roughened end d i f f u s e s incoming l i g h t , which i s t r a n s m i t t e d along the rod by repeated i n t e r n a l r e f l e c t i o n . The bend i n the rod mixes any a x i a l components of the beam and the f l a t , p o l i s h e d end enables maximum t r a n s m i s s i o n out of the rod. A comparison of a t h i n f i l m of s i l i c o n l a s e r annealed with and without the d i f f u s e r i s seen i n F i g u r e s 9 and 10. The c a l i b r a t i o n of the d i f f u s i n g rod t r a n s m i s s i o n was performed with a cw Spectra P h y s i c s Ar+ ion l a s e r and a S c i e n t e c h c a l o r i m e t e r as shown i n F i g u r e 11. I t was necessary to use a cw l a s e r because the ruby l a s e r would not supply enough energy to y i e l d accurate readings on the c a l o r i m e t e r . The maximum output from the c a l o r i m e t e r was recorded with and without the rod f o r 35 seconds of Ar l a s e r i l l u m i n a t i o n at 0.40 W. T h i s technique y i e l d e d t y p i c a l t r a n s m i s s i o n v a l u e s of 17 50 ± 5% for a d i f f u s i n g rod. From studies of laser annealed p o l y s i l i c o n thin films, i t was estimated that the intensity homogeneity was at least as good as ±10% across the beam diameter. Diffusing rods were also made of l u c i t e ( p l e x i g l a s s ) , having the d i s t i n c t advantage of being e a s i l y bent and polished due to the r e l a t i v e softness of the p l e x i g l a s s . However, after repeated use the plexiglass rods would slowly drop in transmission, t y p i c a l l y about 50 to 40% after 5 laser pulses. This was found to be due to formation of tiny microbubbles in the l u c i t e that scattered the laser l i g h t out of the rod. 18 2.5 c m F i g u r e 7. The p l e x i g l a s s l i g h t - p i p e used i n an attempt to homogenize the ruby l a s e r beam. 1 0.4 cm t Diffusive End 9 cm Polished End F i g u r e 8. Quartz d i f f u s i n g rod. Figure 9. An amorphous s i l i c o n f i l m annealed with an inhomogeneous laser beam. Magnification 25X. Figure 10. The even regrowth of a s i l i c o n f i l m annealed with a homogeneous laser beam. The darker, annealed region has a granular appearance. Magnification 25X. 20 A r + Laser 1 Pr ism Absorbing Surface \ C a l o r ime t e r T — l 7 » 1 / N 1 / X 1 / \ I / \ » / » i / Lens [Diffusing V R o d F i g u r e 11. C o n f i g u r a t i o n used f o r c a l i b r a t i o n of the d i f f u s i n g rod t r a n s m i s s i o n . 21 3.3 Annealing Technique The t y p i c a l setup of the ruby laser, semiconductor sample and optics during annealing i s shown in Figure 12. An autocorrelator was used to a l i g n the laser mirrors by centering r e f l e c t i o n s off successive surfaces on top of each other. The sample could be accurately positioned for laser annealing, with the use of a He-Ne alignment laser beam directed in from the back of the ruby laser and out onto the sample area, as shown in the figure. The sample to be annealed was placed as close to the end of the d i f f u s i n g rod as possible in order to obtain a well-defined, homogeneous anneal with l i t t l e beam divergence. A thin glass cover s l i p was sandwiched between the rod and the sample to protect the d i f f u s i n g rod from any vapourized material. Inspection of the s l i p would determine i f any material was lost during the anneal. Each laser pulse could be monitored by the photodiode and storage scope arrangement similar to that of Figure 6. To adjust the output pulse to a s p e c i f i c energy density, the d i f f u s i n g rod, sample and cover s l i p were removed and replaced by the Korad calorimeter shown in Figure 6. The laser was f i r e d repeatedly u n t i l a consistent pulse p r o f i l e and energy reading were established. The rod, sample, and cover s l i p were then inserted into the beam and the incident energy density, E, on the sample surface was calculated as in equation 1; E = E 0 Tt/A {1} . where E 0= energy reading from calorimeter T = rod transmission c o e f f i c i e n t t = cover s l i p transmission (taken as 0.96) A = cross-sectional area of the output end of the rod. 22 Two f a c t o r s p r o h i b i t e d the measurement of energy d i r e c t l y out of the rod. They were: 1) the Korad c a l o r i m e t e r had an i n t e r n a l a p e r t u r e which cut out part of the beam d i v e r g i n g from the rod, and 2) the S c i e n t e c h c a l o r i m e t e r had very poor response at the low e n e r g i e s and short p u l s e widths used. 23 I He-Ne Laser Diffuse Reflector Ruby Laser Glass Slide *To Photodiode and Scope Diffusing Rod F i g u r e 12. The c o n f i g u r a t i o n used during a l a s e r anneal. The He-Ne l a s e r enabled alignment of the rod, l e n s , and sample without having to f i r e the ruby l a s e r . 24 3.4 Photoluminescence Apparatus Analysis of the heterostructure and GaAs samples was done predominantly by photoluminescence (PL). A schematic of the PL system i s shown in Figure 13. Samples were excited by a Spectra Physics, model 165 Ar+ ion cw laser, operating between 50 and 250 mW. The Ar+ beam was passed through Corning f i l t e r s 4-72 and 4-96 to remove any IR components and also through an Ealing 486.1 nm bandpass f i l t e r so that only the 488 nm lasing l i n e was used for e x c i t a t i o n . T y p i c a l l y , the excitation spot on the sample surface was focussed to a diameter of 2 mm. Samples were mounted in an immersion-type l i q u i d He dewar in a free-hanging manner using t e f l o n tape to avoid stresses. The dewar could be f i l l e d with either l i q u i d helium or l i q u i d nitrogen, providing sample temperatures from 1.6 to 4.2 K and at 77 K. Luminescence was c o l l e c t e d from the excited surface with the use of a large concave mirror and then focussed onto the s l i t s ( t y p i c a l l y 300 /jm) of a Perkin-Elmer E l double pass monochromator. A Corning 3-66 f i l t e r was used to remove any spurious components below 560 nm. The PL l i g h t was detected by a Varian VPM-159 photomultiplier (InGaAsP photocathode) operated at dry ice temperature. A Nova 2 minicomputer enabled data acqu i s i t i o n in the photon-counting mode and also controlled the spectrometer, allowing substantial signal averaging. Data output c a p a b i l i t i e s included p l o t t i n g the spectrum on a chart recorder, as well as punching a permanent record of the spectrum on paper tape. Transmission spectra of the s i l i c o n thin films were made by placing the s i l i c o n sample between the monochromator and a collimated beam from a tungsten bulb . An estimate of the 2 5 system response and subsequent data c o r r e c t i o n c o u l d be made by o b s e r v i n g the spectrum of only the tungsten f i l a m e n t . These c o r r e c t i o n s w i l l be d i s c u s s e d i n Chapter 4. 2 6 Photo-multiplier Mono-chromator •Hlter /it i\\ lW i l * / I \ i i \ i i \ i i v / i > Aquisition Electronics Interface Minicomputer Dewar / Sample i y i A. / I Filters \ 1|-I/b Ar Laser F i g u r e 13. A schematic of the photoluminescence system. 27 CHAPTER FOUR; LASER ANNEAL OF HETEROSTRUCTURES  4.1 Heterostructure and Laser Diode Fabrication The composition of the GaAlAs double heterostructure wafers used in t h i s experiment i s shown in Figure 14. The layer thicknesses and aluminum concentrations are nominal values. The wafers were fabricated at BNR by l i q u i d phase epitaxy [42]. In t h i s process, the GaAs substrate was placed in a furnace and pulled sequentially under 3 separate melts containing appropriate GaAlAs solutions. At each melt, the furnace was cooled slowly to i n i t i a t e c r y s t a l growth over the preceeding layer. After the laser annealing and PL experiments, oxide s t r i p e laser diodes [10] were fabricated at BNR from the heterostructure wafers. The f i n a l laser diode structure i s seen in Figure 15. An insulating S i 0 2 layer was deposited over the top confining layer. A narrow str i p e etched into the S i 0 2 layer r e s t r i c t e d the gold contact to the stripe dimensions. The laser chips were cleaved from the wafers once a l l contacts were deposited. The laser chips could be cw operated by an applied forward bias i f they were bonded to heat sinks. During operation, c a r r i e r s are injected to the active region where radiative recombination occurs. Confinement of current and l i g h t i s maintained by the str i p e contact and waveguide structure of the active layer respectively. At high i n j e c t i o n currents the resonant cavity in the active layer, with the cleaved ends acting as 32% r e f l e c t i n g mirrors, provides the geometry for stimulated emission. A t y p i c a l plot of laser l i g h t output versus injected current i s shown in Figure 16. Below the 28 threshold for lasing (<100 mA), the laser acts as a l i g h t -emitting diode. The l i g h t output from a laser may degrade during device operation due to the formation of dark l i n e defects [42], which are highly non-radiative d i s l o c a t i o n networks emanating from defects present in the active layer. A more gradual degradation in e f f i c i e n c y i s caused by a uniform generation of defects in the active region. Ideally, laser annealing could improve the active layer c r y s t a l quality and radiative e f f i c i e n c y , leading to lower threshold and longer-lived lasers. 29 Thickness P - G a 6 A I 4 A s Confining 1pm = Ge Layer Active Q2pm p - G a 9 3 A , . 0 7 A s : G e Layer Confining 1 um n-Gag A I 4 A s :Te Layer 75pm n-GaAs Substrate F i g u r e 14. S t r u c t u r e of the double h e t e r o s t r u c t u r e wafers prepared at B e l l - N o r t h e r n Research by l i q u i d phase e p i t a x y . Gold Oxide Substrate Gold •Epitaxial Layers 500 um F i g u r e 15. C r o s s - s e c t i o n of a f a b r i c a t e d at B e l l h e t e r o s t r u c t u r e of semiconductor l a s e r diode -Northern Research from the F i g u r e 14. 30 _ 4 ? E 01 LU o /Lasing Q_ 2 / Mode OUTPUT A y i LED M o d e ^ ^ ^  1 I t h =100mA i i 50 100 150 CURRENT (mA) Figure 16. Typical output c h a r a c t e r i s t i c s of a semiconductor laser diode, showing a threshold for lasing of 100 mA. 31 4.2 Previous Annealing Experiments A brief description of the motivation behind the experiments on heterostructures was given in the Introduction. Here I describe in greater d e t a i l the previous annealing experiment performed by Parsons et a l . [11] on pulsed laser annealing of heterostructure wafers. Ruby laser l i g h t was found to be ideal for annealing the active Ga < 1 5Al 0 7As layer. This s u i t a b i l i t y i s i l l u s t r a t e d by Figure 17, showing that p r e f e r e n t i a l absorption occurs in the active layer due to an energy gap less than the energy of the ruby l i g h t . Calculations [11] have shown that 23% of the l i g h t incident on the top confining layer would be absorbed in the active layer whereas less than 1% i s absorbed in the confining layers. A c a l c u l a t i o n of the temperature p r o f i l e following a 20 ns Q-switched ruby laser pulse showed that the heat generated does not have time to reach the surface of the exposed confining layer and thus an increase in absorption due to bandgap shrinkage could be neglected. It was estimated that the incident threshold energy density required to melt the active layer was 0.15 J/cm2 (Appendix A). Parsons et a l . laser annealed heterostructures at energy densities above and below that required for melting. They found that even at a low energy anneal (0.05 J/cm2) the photoluminescence (PL) associated with the active layer was not observed, whereas strong PL was observed in unannealed samples. Their results are seen in Figure 18. The PL peaks observed from the active layer were i d e n t i f i e d as band-to-band (b-b) recombination of free electrons and free holes and band-to-32 acceptor (b-A) recombination of f r e e e l e c t r o n s and bound h o l e s . PL from the c o n f i n i n g l a y e r was i d e n t i f i e d as b-A recombination. The d e s t r u c t i o n of the r a d i a t i v e p r o p e r t i e s of the a c t i v e l a y e r d u r i n g l a s e r a n n e a l i n g was a t t r i b u t e d i n pa r t to gross damage a s s o c i a t e d with the inhomogeneity of the l a s e r beam. Proof i s by means of example; l o o k i n g back at F i g u r e s 9 and 10 shows the uneven damage induced on s i l i c o n by an inhomogeneous beam compared with the c l e a n , even r e c r y s t a l l i z a t i o n c r e a t e d by a homogeneous beam of equal energy d e n s i t y . 33 Q_ < 0 >-0 on LU z LU - -Ruby Laser Confining Confining 1.79 eV 1.92 eV Active 1.92 eV 1.50 eV Substrate 1.42 eV HETEROSTRUCTURE LAYERS F i g u r e 17. The ruby l a s e r i s i d e a l l y s u i t e d f o r d e p o s i t i n g energy i n the a c t i v e l a y e r . The l i g h t energy i s gr e a t e r than the a c t i v e l a y e r energy gap but l e s s than the c o n f i n i n g l a y e r energy gap. 34 D O O C >-CO LU I-Active L. b-b b-A T = 77K BNRDH 10 0.00 & 0.07 J cm -2 BNRDH 8 0.155 J c m - 2 Confining L. b-A ± 1.4 1.7 2.0 PHOTON E N E R G Y ( e V ) Figure 18. Previous photoluminescence results from a heterostructure annealed with an inhomogeneous laser beam. Notice the extinction of active layer photoluminescence at anneal energies above 0.15 J/cm2. 35 4.3 Data In t h i s section, the data from the homogeneous annealing experiment on double heterostructure wafers are presented. Six heterostructure wafers, each about 1 cm2 in area, were annealed with homogenized Q-switched ruby laser pulses of various i n t e n s i t i e s above and below the estimated melting threshold (0.15 J/cm 2). The experimental setup for the anneal was shown before in Figure 12. Only 1/2 of each wafer was annealed, allowing comparison of PL between annealed and unannealed areas on the same wafer. After annealing, PL studies of each wafer were made at 77K, as described in Chapter 3 and shown in Figure 13. Ar+ laser excitation and luminescence c o l l e c t i o n were made from the e p i t a x i a l side of the heterostructures. A PL spectrum was also taken from the substrate side of a wafer for comparison. In order to enable correction of the spectra for system response, the spectrum of a tungsten filament was also recorded. To obtain a corrected spectra, the data points of the o r i g i n a l spectrum were divided by the system response to the filament. No corrections were made for the emission spectrum of the tungsten filament i t s e l f , as i t i s known for the blackbody radiation law [43] that the spectral emission w i l l vary slowly and smoothly by about 20% over the short intervals (1.4 to 1.6 eV) of importance here. Figure 19 shows three spectra; the PL from an unannealed heterostructure, the PL from a heterostructure i r r a d i a t e d at 0.20 J/cm2 (above the melting threshold) and the system response. The or i g i n of each band in the PL i s c l a r i f i e d in Appendix B. The mechanism of PL from the e p i t a x i a l layers i s as 36 follows. Using o p t i c a l absorption data [44], i t was calculated that 63% of the Ar+ excitation l i g h t was absorbed in the top 0.1 urn of the 1.0 thick confining layer. As the photo-excited non-equilibrium c a r r i e r s diffused across the e p i t a x i a l layers, radiative recombination occurred and the PL spectra c h a r a c t e r i s t i c of each material was seen. Notice that in an expanded PL spectra of an unannealed sample seen in Figure 20, peaks associated with the confining layers are very weak. Previous PL spectra taken at 4K [11] show a strong confining layer PL and a weak active layer PL due to a short c a r r i e r d i f f u s i n g length at this temperature. However, at 77K, a longer c a r r i e r d i f f u s i o n length and good radiative e f f i c i e n c y of the active layer create strong active layer PL and weak confining layer PL, as observed. Thus the active layer PL i s dependent on the d i f f u s i o n properties of photo-excited c a r r i e r s across the p-Ga 6 A 1 ^ As layer, the perfection of the interface and the radiative e f f i c i e n c y in the active layer. Before analyzing the PL r e s u l t s , i t should be noted that the substrate PL shown in Figure 21 i s shifted to lower energies than the active layer PL due to the bandgap broadening presence of Al in the active layer. Any weak PL from the substrate (which i s unli k e l y due to the depth of the e p i t a x i a l layers) w i l l not i n t e r f e r e in the interpretation of the active layer PL. The PL data of Figure 18 are t y p i c a l of those recorded from a l l wafers. Band positions or r e l a t i v e band i n t e n s i t i e s did not change as a result of annealing. However, in samples annealed above the estimated threshold for melting, 0.15 J/cm2, the active layer PL dropped by a factor of two.• This e f f e c t i s well 37 i l l u s t r a t e d by Figure 22. The PL of the b-b t r a n s i t i o n i s recorded, as t h i s is the t r a n s i t i o n responsible for lasing [10]. The error bars indicate the uniformity and r e p e a t i b i l i t y of the PL measurements. Although there i s no d i r e c t evidence of melting, the sudden decrease of the peak intensity in the region 0.1 to 0.2 J/cm2 correlates with the estimated threshold for melting. The PL from the confining layers was very weak and not always observed, so no interpretations were made from the anneal behavior of these bands. Thus, an above threshold anneal reduces the active layer PL by about a factor of two. Previous PL data [11] from heterostructures annealed with an inhomogeneous laser beam showed a v i r t u a l elimination of'active layer PL after annealing at energies above threshold. After the PL measurements were made, the wafers were sent to Bell-Northern Research for fabrication of oxide s t r i p e laser diodes. The performance of devices fabricated from annealed and unannealed material i s given in Table I. Approximately 82% of the unannealed device chips gave stimulated emission at room temperature with an average current threshold of 152 mA. Although a l l annealed devices gave spontaneous emission (emission in the LED mode), only some devices annealed at 0.07 and 0.18 J/cm2 actually lased, with noticeably higher lasing thresholds. 38 b-A b-b / n o t \ / annealed \ ACTIVE LAYER <D y PHOTOLUMINESCENCE seal \ T=77K (linear / / 0.20 \ > r- / / J-cnf2 \ INTENSI X / / \ / / X / / \ / / \ / / \ / / — / sCs^ - — — £ystem response 1 I l 1 1 1.3 1.5 1.7 PHOTON ENERGY(eV) Figure 19. Typical photoluminescence spectra from homogeneously annealed double heterostructure wafers. At anneal i n t e n s i t i e s > 0.15 J/cm2, the luminescence was halved. 39 F i g u r e 20. Photoluminescence from the e p i t a x i a l l a y e r s , showing the c o n f i n i n g l a y e r bands o c c u r r i n g at higher e n e r g i e s . 40 W c >-Ul Z LU r -z S u b s t r a t e , P L / »y \ Epitaxial ' \S ide PL T = 7 7 K J I i n 1.2 1.4 1.6 PHOTON ENERGY (eV) F i g u r e 21. Photoluminescence spectrum of the s u b s t r a t e of a h e t e r o s t r u c t u r e wafer. The band peaks do not c o i n c i d e with those from the e p i t a x i a l l a y e r s . 41 ° -£ if) g O ^ o > X h-Q_ »—' CO < 1x1 CL LU 0 ESTIMATED MELTING THRESHOLD 0.15 J e r r i 2 1 1 0.1 0.3 0.5 0.7 ANNEAL ENERGY DENSITY ( J e m " 2 ) F i g u r e 22. Behavior of the peak i n t e n s i t y of the b-b a c t i v e l a y e r photoluminescence as a f u n c t i o n of the l a s e r anneal energy d e n s i t y . The region of degradation of the photoluminescence c o r r e l a t e s with the estimated t h r e s h o l d f o r m e l t i n g . Table I. Effe c t of a homogeneous laser anneal on device performance. Anneal Intensity (J/cm 2) Number of Devices Y i e l d % Average I t h at 300K (mA) 0.00 51 82 152 0.07 10 20 220 0.15 10 0 -0.18 11 27 165 0.22 10 0 -0.28 10 0 -0.30 10 0 — 43 4.4 Data Analysis The behavior of laser thresholds after annealing may be interpreted through the change in PL e f f i c i e n c y and the threshold equation for lasing [45]; J t h = (d/N)[B" 1(L- 1lnR- 1+a)+J] {2} where J t h = threshold current density for lasing d = laser cavity width L = laser cavity length R = mirror r e f l e c t i v i t y N = internal quantum e f f i c i e n c y , proportional to the radiative recombination rate a = absorption constant B = constant, but decreasing with increasing temperature J = constant, but increasing with increasing temperature. A halving of the PL intensity w i l l halve N and double the threshold. But at higher laser thresholds, heating of the device a f f e c t s the constants B and J, driving the threshold even higher again, with the result that lasing i s u n l i k e l y to be obtained at room temperature operation [42], In other work on GaAlAs heterostructures, Henry et a l . [8] used a cavity-dumped Ar+ laser with 18 ns pulses to generate dark l i n e defects in the active layer. They found that the experimental threshold for damage was an incident power of 6 to 12 MW/cm2 and that damage was due to l o c a l i z e d melting and e p i t a x i a l r e c r y s t a l l i z a t i o n of non-radiative material containing large numbers of point defects and d i s l o c a t i o n loops. However, no mention was made of the Ar+ laser beam homogeneity. The threshold for PL deterioration in Figure 22 was 0.1 to 0.2 J/cm2, corresponding to 5 to 10 MW/cm2, agreeing with the damage threshold of Henry. Thus one mechanism of damage to the heterostructures was probably melting and r e c r y s t a l l i z a t i o n of an active layer containing defects and d i s l o c a t i o n s . Another 44 damage mechanism may have been through r e d i s t r i b u t i o n of A l i n the a c t i v e l a y e r d u r i n g the molten phase, due to the hig h s e g r e g a t i o n c o e f f i c i e n t of A l i n GaAs [46]. Furthur damage may have been in t r o d u c e d by thermal s t r e s s e s at the a c t i v e l a y e r and c o n f i n i n g l a y e r boundaries due to d i f f e r e n t i a l a b s o r p t i o n and he a t i n g of these l a y e r s . S e l l et a l . [44] re p o r t that G a h x A l x A s compounds with low x c o n t r a c t more r a p i d l y than those with high x, as the temperature i s lowered. From the data c o l l e c t e d i n t h i s experiment i t i s c l e a r that homogeneous l a s e r a n n e a l i n g of h e t e r o s t r u c t u r e m a t e r i a l at i n t e n s i t i e s near or above the m e l t i n g t h r e s h o l d of the a c t i v e l a y e r damages the a c t i v e l a y e r and degrades i t s r a d i a t i v e e f f i c i e n c y , subsequently causing very d e l e t e r i o u s e f f e c t s on device performance. 45 CHAPTER FIVE: SECONDARY LASER ANNEALING EXPERIMENTS  5.1 Amorphous S i l i c o n Films Interest in producing p o l y s i l i c o n films on glass substrates was generated from possible application to fabrication of charge-coupled device arrays produced at Bell-Northern Research [47]. Because films of amorphous Si on glass are cheap and easy to make, i t was hoped that laser annealing of these films would result in inexpensive p o l y c r y s t a l l i n e material. Research by other workers [2] has shown that single c r y s t a l Si cannot be regrown on an amorphous substrate due to lack of a c r y s t a l seed for an e p i t a x i a l growth process. Undoped, amorphous, hydrogenated s i l i c o n films (a-Si:H) were supplied by W. Westwood of BNR. The films were 1 yum thick and deposited on Corning 7059 glass substrates by plasma deposition from silane ( S i H + ) . Annealing was performed with the ruby laser in the Q-switched mode at energy densities between 0.1 and 1.0 J/cm2. These annealing energies are less than those used for bulk s i l i c o n [30] (about 2.5 J/cm2) as the sample i s a thin f i l m on an insulating substrate. Results of transmission studies at 77K using a tungsten filament as the source are shown in Figure 23. The data have been corrected by dividing by the system response to the tungsten filament alone. The bands are due to interference of l i g h t within the f i l m , the band maxima given by the standard r e l a t i o n for normally incident l i g h t [43]; mX = 2n(A)t {4} where m = order number of the fringe X = l i g h t wavelength t'= f i l m thickness n(X)= r e f r a c t i v e index of the f i l m . 46 As the r e f r a c t i v e indices of amorphous and p o l y c r y s t a l l i n e s i l i c o n films have been shown to be nearly equal [48], the i d e n t i c a l spacing of bands in both the v i r g i n and annealed films show that no material loss (through vapourization) occurred during the anneal. This was also supported by an o p t i c a l microscope examination of the cover s l i p separating the d i f f u s i n g rod and the f i l m . The smaller amplitude of the interference structure in the annealed fi l m i s probably due to the granular structure of the f i l m as seen in Figure 10, providing a roughened surface not favourable for interference fringe generation. That the laser anneal produced s i l i c o n of a more c r y s t a l l i n e quality i s well indicated by the s h i f t in o p t i c a l bandgap seen in Figure 23. The absorption edge of the amorphous Si:H i s in the region of 1.9. eV, agreeing with other experimenters [49]. However, the absorption edge of the annealed Si:H has been shifted to lower energies in the region of 1.1 to 1.2 eV. Since the bandgap of single c r y s t a l Si at 77K is 1.18 eV [50], these data support the b e l i e f that a p o l y c r y s t a l l i n e s i l i c o n f i l m has been produced. Furthur analysis of the films, such as X-ray d i f f r a c t i o n and H a l l measurements, were not made. 47 J I I I L_ 1.2 1.4 1.6 1.8 2.0 ENERGY (eV) F i g u r e 23. Transmission s p e c t r a of unannealed and l a s e r annealed a-Si:H on g l a s s taken at 77K. Note the o p t i c a l bandgap s h i f t to lower e n e r g i e s i n the annealed f i l m , i n d i c a t i v e of a more c r y s t a l l i n e m a t e r i a l . 48 5.2 Aluminum on Ga11iurn Arsenide By laser annealing down onto a thin Al f i l m on a GaAs substrate, i t was hoped to momentarily melt both materials to form a Ga,.xAlxAs graded bandgap structure upon r e c r y s t a l l i z a t i o n . This technique has been used previously to form metal s i l i c i d e compounds by laser annealing a f i l m of metal deposited on single c r y s t a l s i l i c o n [4]. The GaAs substrate was single c r y s t a l , n-type and doped with Si at 1.5 X 10 1 S cm"2. After cleaning the surface with a standard etch [51] of 2% bromine in methanol, 0.1 of Al was deposited on the substrate by planar magnetron sputtering. Part of the Al f i l m was laser annealed with the ruby laser operating in the free-running mode in the setup of Figure 12. It was hoped that the longer free-running pulse length would generate a longer-lived melt of Al and GaAs. An energy density of 3.8 J/cm2 was used, being midway between lower energy pulses showing no v i s i b l e effect on the Al f i l m , and higher energy pulses which result in mechanical damage to the substrate. Before PL measurements were taken, selected areas of the unannealed Al f i l m were etched off with a weak HN03 solution to expose the GaAs substrate. PL spectra were then taken of the unannealed GaAs and the annealed Al on GaAs. The r e s u l t s , uncorrected for system response, are shown in Figure 24. The PL bands are readily i d e n t i f i e d [52] as b-b recombination at 1.49 eV and. the broader band at 1.17 eV i s attributed to a Ga vacancy-Si donor complex. The spectra of unannealed GaAs and annealed Al on GaAs are v i r t u a l l y i d e n t i c a l , with no appearance of structure at higher energies which could 49 be due to GaAlAs. Optical microscope examination of an annealed spot on the Al f i l m showed ragged edges of Al f i l m surrounding the exposed GaAs substrate. Microscope examination of the cover s l i p separating the d i f f u s i n g rod and sample during the anneal showed small flecks of aluminum deposited on the glass. Thus, i t appears that the Al f i l m had been "blown o f f " the GaAs substrate. The Al f i l m has a lower melting point (930K) than the GaAs (1510K) and t h i s anneal behavior could possibly be explained by a model of melting and vapourization of a thin f i l m of Al on an insulating substrate. 50, ' F i g u r e 24. Photoluminescence s p e c t r a of an annealed A l f i l m on GaAs and of the GaAs s u b s t r a t e . No presence of a higher bandgap GaAlAs a l l o y was d e t e c t e d . 51 CONCLUSION Annealing of GaAlAs double h e t e r o s t r u c t u r e m a t e r i a l with homogenized, Q-switched ruby l a s e r l i g h t was not s u c c e s s f u l i n improving the r a d i a t i v e y i e l d of the a c t i v e l a y e r . R a d i a t i v e e f f i c i e n c y was reduced by a f a c t o r of two f o r h i g h energy anneals above the estimated t h r e s h o l d f o r me l t i n g of the a c t i v e l a y e r . Subsequently, e l e c t r o l u m i n e s c e n t l a s e r s f a b r i c a t e d from the annealed m a t e r i a l performed much worse than those made from unannealed m a t e r i a l . However, the p o t e n t i a l value of ann e a l i n g with a homogenized beam i s demonstrated by the comparison to v i r t u a l e l i m i n a t i o n of the r a d i a t i v e a b i l i t y of the a c t i v e l a y e r when an inhomogeneous beam was used. P o s s i b i l i t i e s s t i l l e x i s t f o r the s u c c e s s f u l a p p l i c a t i o n of l a s e r a n n e a l i n g to h e t e r o s t r u c t u r e m a t e r i a l . For example, l a s e r a n n e a l i n g the wafer while the s u b s t r a t e i s furnace heated (perhaps 900K) would s i g n i f i c a n t l y slow the r a t e of c o o l i n g of the a c t i v e l a y e r and may r e s u l t i n fewer trapped d e f e c t s . A l s o , an anneal using-m u l t i p l e p u l s e s of lower energy d e n s i t y may be worth i n v e s t i g a t i o n . Attempted pr o d u c t i o n of a GaAlAs a l l o y by a f r e e - r u n n i n g l a s e r anneal of A l f i l m s on GaAs r e s u l t e d i n the A l f i l m being removed from the GaAs c r y s t a l s u r f a c e . A complete i n v e s t i g a t i o n of t h i s idea should i n c l u d e v a r i a t i o n s i n pulse l e n g t h , f i l m t h i c k n e s s and s u b s t r a t e temperature. Q-switched ruby l a s e r a n n e a l i n g of amorphous S i f i l m s on g l a s s has produced a more c r y s t a l l i n e m a t e r i a l , i n d i c a t e d by lowering of the o p t i c a l bandgap to a value c l o s e t o t h a t of s i n g l e c r y s t a l S i . These p o s i t i v e r e s u l t s produced some 52 enthusiasm with collaborators at Bell-Northern Research but time l i m i t a t i o n s did not allow furthur investigation. Other p o s i t i v e aspects of the laser annealing experiments in general were that they tended to generate more ideas and questions than they solved. These included possible techniques to completely eliminate laser speckle effects and an experimental method to test the melting model of pulsed laser annealing. To conclude, i t i s appropriate to mention that the double heterostructure experiment enabled publication of a research paper and presentation of the work at the 1980 annual meeting of the Materials Research Society [53]. Also, ideas stimulated by the research have led to a patent application for a "Thin Film Laser Power Indicator". 53 APPENDIX A: LASER HEATING MODEL A simple heating model has been proposed by Hutcheon [54] for heating of the active layer of a double heterostructure wafer during Q-switched ruby laser i r r a d i a t i o n . The model gives an estimate of the threshold for melting of the active layer. Figure 25 shows a sketch of the e p i t a x i a l layers with boundary conditions included. It was assumed that the confining layers absorbed no radiation (laser energy < bandgap) and that the top and bottom surfaces of the wafer remained at room temperature. Similar to other models [30], the one-dimensional heat equation was used to find the temperature T; where c = 344 J/kg p = 5.37 X 10 3 kg/m3 R = 0.3 ( r e f l e c t i v i t y ) K = 5.7 X lO'T' 1- 2 5 (thermal conductivity) <* = 30 m"1 (absorption constant) I 0 (t)= incident 20 ns laser pulse. The middle term represents the heating of the active layer due to absorbed l i g h t whereas the right term i s the heat loss by conduction into the confining layers. Thus the l e f t term i s the t o t a l heat flow into the active layer. Knowing the incident l i g h t pulse I D ( t ) , the heat equation was integrated numerically by finding the temperature p r o f i l e at t=0, then incrementing by a small time interval and improving on the p r o f i l e . It was found that for an incident pulse of 0.15 J/cm2, the temperature of the active layer would reach i t s melting point (1500 K).. Henry et a l . [8] have performed a similar c a l c u l a t i o n for a GaAlAs double heterostructure and found a power of 5.2 MW/cm2 was necessary for melting. They stated that their value was {4} 54 probably an underestimate. The estimate by Hutcheon corresponds to 7.5 MW/cm2, thus the two values agree to b e t t e r than 50%. 55 T=300K Tr 3 0 0 K Ut) Confining Layer Ac t i ve yy 'Layer/ / / Confining Layer No Absorpt ion •7///, Absorption^ '////, No Absorption -> z F i g u r e 25. E p i t a x i a l l a y e r s of a h e t e r o s t r u c t u r e wafer showing the boundary c o n d i t i o n s used i n the m e l t i n g model. 56 APPENDIX PHOTOLUMINESCENCE BAND IDENTIFICATION The d i r e c t bandgap dependence of Gaj_ xAl xAs for x < .45 (direct bandgap) as a function of x and temperature T has been modelled by [42]: E = 1.519 - (5.405 X 10" 4T 2)/(T+204) + 1.24x {5} where T i s in K and E i s in eV. The bandgap diagram for GaAs i s shown in Figure 26, with t y p i c a l values for donor and acceptor s i t e s [52], From th i s data and equation {5}, one readily i d e n t i f i e s the peaks associated with active layer PL and confining layer PL, as i d e n t i f i e d in Figure 19 and 20. In the confining layer PL, a band i s observed at 1.92 ± .01 eV and a narrower band at 2.02 ± .01 eV. These peaks agree reasonably well with previously published results by Kaneko et a l . [55], who for x = 0.36 and T = 77K found corresponding peaks at 1.89 and 2.01 eV. Kaneko et a l . attributed the peaks to band-acceptor (b-A) recombination (1.92 eV band) and b-b recombination (2.02 eV band). The confining layer PL was observed only in some wafers at higher Ar+ laser excitation powers. When present, the PL provided an accurate measurement of the aluminum concentration, x = 0.36, in the range of the expected nominal concentration x = 0.4. In Figure 18, the higher energy peak of the active layer PL at 1.53 ± .005 eV i s readily i d e n t i f i e d as b-b recombination of free electrons and free holes. This peak i s c h a r a c t e r i s t i c of GaAs and i s reported for p-type GaAs:Ge by numerous workers at 1.51 eV [56-59]. Also indicative of the b-b t r a n s i t i o n are the results shown in Figure 27; that the PL of the b-b peak increases r e l a t i v e to the other bands as the excitation power i s 57 increased [56]. The peak s h i f t from that of pure GaAs yiel d s an estimate of the Al concentration in the active layer (see equation {5}). Thus a s h i f t of .02 ± .005 eV implies an aluminum concentration of 2 to 3%, somewhat less than that of the nominal value, 7%. The band centred at 1.48 ± .005 eV is most probably b-A recombination. The acceptor l e v e l for Ge has been reported to be between 0.030 and 0.042 eV in GaAs [52]. The observed peak separation between the b-A and b-b bands i s 0.05 ± 0.01 eV, in agreement with the higher estimate of the acceptor l e v e l in GaAs. The preceeding i d e n t i f i c a t i o n of the active and confining layer PL bands agrees with the previous work with BNR heterostructures [11], although the bands observed in th i s experiment are broader. A broadening of bands i s usually i n d i c a t i v e of a higher dopant concentration [56]. 58 Conduction Band E Q =1.508 1 eV J Erf 0 . 0 0 5 eV GaAs at 77 K JEA= 0 . 0 3 5 eV Valence Band F i g u r e 26. Values of the energy gap, and donor and acceptor e n e r g i e s f o r GaAs. 59 Fi g u r e 27. Photoluminescence s p e c t r a of the a c t i v e l a y e r . The peak i n t e n s i t y of the b-A peak has been normalized, showing the r e l a t i v e i n c r e a s e of the b-b peak with i n c r e a s i n g Ar l a s e r e x c i t a t i o n . REFERENCES 1. E. Shtyrkov, I. Khaibullin, M. Zaripov, M. Galyatudinov, and R. Bayazitov, Sov. Phys. Semicond. 9, 1309 (1976). 2. J . Schott, Laser and Electron Beam Processing of Materials , Proc. Materials Research Society 1980, (New York: E l s e v i e r ) . 3. 0. Kutakova and L. Streltsov, Sov. Phys. Semicond. 10, 265 (1976) . 4. Z. Liau, B. Tsaur, J . Mayer, Appl. Phys. Lett. 34, 221 (1979) . 5. J . Muller, A. Grob, R. Stuck, J. Grob and P. S i f f e r t , Appl. Phys. Lett. 33, 287 (1978). 6. A. C u l l i s , H. Webber and N. Chew, Royal Signals and Radar Establishment Newsletter 3, 10 (1979). 7. A. C u l l i s , H. Webber and P. Bailey, J. Phys. E 12, 688 (1979). 8. C. Henry, P. Petroff, R. Logan and F. Mer r i t t , J. Appl. Phys 50, 3721 (1979). 9. F u l l address: Bell-Northern Research, PO Box 3511, Station C Ottawa, Ontario, Canada, K1Y 4H7. 10. J.C. Dyment, F.P. Kapron and A.J. Springthorpe, IEEE ED24, 995 (1972). 11. R.R. Parsons, J.A. Rostworowski, A.J. Springthorpe and J.C. Dyment, Laser and Electron Beam Processing of Materials , Proc. Materials Research Society 1979, (New York: Academic Press). 12. J. Mayer, L. Erikson and J. Davies, Ion Implantation in  Semiconductors , (New York: Academic Press, 1970). 13. R. Young, C. White and G. Clarke, Appl. Phys. Lett. 3_2, 139 (1978) . 14. See, for example, Laser and Electron Beam Processing of  Materials , Proc. Materials Research Society 1979, (New York: Academic Press). 15. J. Bean, H. Leamy, J . Poate, G. Rozgonyi, J . Van der Z i e l and J. Williams, J. Appl. Phys. 50, 881 (1979). 16. S. Margalit, D. Fakete, D. Pepper, C. Lee, A. Yariv, Appl. Phys. Lett. 33, 346 (1978). . 17. R. Cohen, J . Williams and L. Feldman, Appl. Phys. Lett. 33, 751 (1978). 61 18. H. Bomke, H. Berkowitz, M. Harmatz, S. Kronenberg and R. Lux-, Appl. Phys. Lett. 33, 955 (1978). 19. S. Lau, W. Tseng, M. Nicolet and J . Mayer, Appl. Phys. Lett. 33, 235 (1978). 20. T. Inada, K. Tokunage and S. Taka, Appl. Phys. Lett. 3_5, 546 (1979) . 21. Maxwell Laboratories, Inc., San Diego, C a l i f o r n i a . 22. S.M. Edelglass, Engineering Materials Science , (New York: Ronald Press, 1966) . 23. R.B. Gold and J.F. Gibbons, J. Appl. Phys. 51, 1256 (1980). 24. D. Auston, J. Golovchenko, P. Smith, C. Surko and T. Venkatisan, Appl. Phys. Lett. 3_3, 539 (1978). 25. J . Williams, W. Brown, H. Leamy, J . Poate, J . Rodgers, D. Rousseau, G. Rozgonyi, J. Shelnutt and T Sheng, Appl. Phys. Lett. 33, 542 (1978). 26. A. Gat and J. Gibbons, Appl. Phys. Lett. 3_3, 389 (1978). 27. J . Narayan and C. White, Laser and Electron Beam Processing  of Materials , Proc. Materials Research Society 1979, (New York: Academic Press), p.65. 28. D. Kim, R. Shah and D. Crosthwait, J. Appl. Phys. 51, 3121 (1980) . 29. J. Van Vechten, Laser and Electron Beam Processing of Materials , Proc. Materials Research Society 1979, TNew York: Academic Press), p.79. 30. P. Baeri, S. Campisano, G. Foti and E. Rimini, J . Appl. Phys. 50, 788 (1979). 31. A. Fowler and R. Hodgson, Appl. Phys. Lett. 36, 914 (1980). 32. N. Johnson, D. Biegelsen, D. Bartelink, D. Moyer and H. Singh, Laser and Electron Beam Processing of Materials , Proc. Materials Research Society 1980, (New York: Else v i e r Press) . 33. D. Auston, C. Surko, T. Venkatesan, R. Slusher, J . Golovchenko, Appl. Phys. Lett. 33, 437 (1978). 34. M. Von Allmen, Laser and Electron Beam Processing of Materials , Proc. Materials Research Society 1979,"TNew York: Academic Press), p.6. 35. J.A. Rostworowski, R.R. Parsons and D.G. Hutcheon, Appl. Phys. Lett. 35, 934 (1980). 62 36. I. Khaibullin, E. Shtyrkov, M. Zaripov, R. Bayazitov, M. Galyatudinov, Radiation E f f e c t s 36, 225 (1978). 37. E.J. Yoffa, Laser and Electron Beam Processing of Materials Proc. Materials Research Society 1979") (New York: Academic Press), p.59. 38. S. Edwards, M. Warner, P h i l . Mag. A, 40, 257 (1979). 39. A. Compaan, H. Lo, Laser and Electron Beam Processing of  Materials , Proc. Materials Research Society 1980, (New York: E l s e v i e r Press). 40. M.J. Beesley, Lasers and Their Applications , (London: Taylor and Francis, 1976), p.98. 41. Lome Whitehead, private communication. 42. H. Casey and M. Panish, Heterostructure Lasers-Part B , (New York: Academic Press). 43. E. Hecht, A. Zajac, Optics , (Reading: Addison-Wesley, 1974) . 44. D. S e l l and H. Casey, J. Appl. Phys. 45, 800 (1974). 45. H. Kressel, J . Butler, Semiconductor Lasers and  Heterojunction LED's , (New York: Academic Press, 1977). 46. M.B. Panish and M. Ilegems, Progress in S o l i d State  Chemistry , (New York: Pergamon Press, 1972) , Vol 7. 47. M. King, D.D. Smith, Telesis 2, 17 (1980). 48. C. Cullen and 0. Wang, ed., Heteroepitaxial Semiconductors  for Electronic Devices , (New York: Springer-Verlag, 1978), p.187. 49. M.H. Brodsky, ed., Amorphous Semiconductors , (Berlin: Springer-Verlag, 1979), p.294. 50. J. Millman and C. Halkias, Integrated Electronics , (New York: McGraw-Hill, 1972), p.29. 51. J . Vossen and W. Kern, ed., Thin Film Processes , (New York: Academic Press, 1978). 52. R. Willardson and A. Beer, ed., Semiconductors and  Semimetals, Vol 8 , (New York: Academic Press, 1972). 53. M.J. Brett, J.A. Rostworowski, R.R. Parsons, A.J. Springthorpe, CM. Look and J . C Dyment, Laser and Electron  Beam Processing of Materials , Materials Research Society 1980, (New York: Elsevier Press). 54. D.G. Hutcheon, 1979, unpublished. 63 55. K. Kaneko and M. Ayabe, J . Appl. Phys. 51, 6337 (1980). 56. F. Rosztoczy, F. Ermanis, I . Hayashi and B. Schwartz, J . Appl. Phys. 41, 264 (1970). 57. H. K r e s s e l , F. Hawrylo and P. L e f u r , J . Appl. Phys. 39, 4059 (1968). 58. H. K r e s s e l and M. Ettenberg, Appl. Phys. L e t t . 23, 511 (1973) . 59. E. Rao and N. Duhamel, I n s t . Phys. Conf. Ser. No. 46, 360 (1979) . 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0085219/manifest

Comment

Related Items