Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
The research areas include Solid-State Electronics, Quantum Electronics, ... tory, University of Illinois at Urbana-Champaign, began on October 1, 1986, ...
Typology: Exercises
1 / 96
This page cannot be seen from the preview
Don't miss anything!
FOR THE PERIOD
APRIL 1,^ 1988,^ THROUGH^ MARCH^ 31,^1989
FOR
CONTRACT N00014-84-C-
I
I
I
I
For the Period
U 1 April 1988 through 31 March 1989
I
Joint Services Electronics Program Contract N00014-84-C-
Monitored by the Office (^) of Naval Research
U
William Kenneth Jenkins JSEP Principal Investigator 3 Coordinated Science Laboratory
I
3 THE BOARD OF TRUSTEES OF THE UNIVERSITY OF ILLINOIS
AT URBANA-CHAMPAIGN Grants and Contracts Office 506 S. Wright Street Urbana, IL 61801
I
I
Reproduction in whole or in part is permitted for any purpose of the United States Government
Approved for public release. Distribution unlimited.
I
I
I
I
U
RESEARCH AREA: Solid-State Electronics
Unit I Crystal Growth from the Vapor Phase and Controlled Doping of Equilibrium 3 and^ Metastable^ Semiconductor^ Alloys:^ Ion/Surface^ Interactions^ ...........................^^3
Unit 2 Studies ot 1ransport Phenomena^ in^ Semiconductors^ .............................................^^8
Unit 4 Basic Studies of^ the^ Optical^ and^ Electronic Properties^ of^ Defects^ and Impurities in Compound Semiconductor Epitaxial Lavers and Related S up erlattices .................................................................................................. 11
Unit 5 Heterostructure Electronic Devices by Metalorganic Chemical Vapor Deposition (M O CVD ) ....................................................................... 18
I Unit 7 Computer-Aided Design of High-Performance Integrated Circuits with U ltra-Sm all Features .......................................................................... 21
Unit 24 Electronic and Transport Properties of Ultra-Low-Dimensional Structures^ ....................^27
I RESEARCH AREA: Quantum Electronics
3 Unit 8 Collective Electronic Transport in Quasi One-Dimensional Systems ........................... 32
Unit 9 An Investigation of Plasma and Chemi *.. P ocesses in Cylindrical M agnetron Plasm a Discharges .......^ ..........................................................^35
Unit 10 Excited State Chem istry in Gases ...................................................................... 38 I
RESEARCH AREA: Electromagnetics
Unit 11 Monolithic Millimeter-Wave Integrated Circuits with Microstrip 3 A n ten n as ...................................................................................................... 43
Unit 12 Investigation of Radar Scattering Characteristics of Controllable 3 Surface Shapes with Application to Low Observable Targets ........................... 48
I
I i
RESEARCH AREA: Information (^) Systems
Unit 13 Hig~h-Performance Testable Electronic Systemis (^) ..... .................................... 51
Unit 14 New Directions in Fault-Tolerant Computing ................................................ 4
Unit 15 Efficient Computation Techniques .................................................... (^) ...... 57
Unit 16 High-Resolution Sensor Array Processing ................................................... (^) 61I
Unit 17 Parallel VLSI Structures for Sensor Array Processing..................................... 65
Unit 18 Adaptive Algorithms for Identification, (^) Filtering, Control, and Signal Processing ......................................................................... (^68)
Unit 19 Distributed and Decentralized Systems ...................................................... (^711)
Unit 20 Robust Feedback Control of Nonlinear Systemns............................................ 76
Unit 21 Multiple-Terminal Digital Communication Systems ........................... ........... 79
Unit 22 Statistical Signal Processing in Comnunication Systems ................................... 86
RESE4RCH AREA: Electronics ResearchI
Unit 23 Basic Research in Electronics................................................................ (^893)
'II c e o 7
1's 11
1-)t 5t
ZI
Completion of the Uni-vrsity of Illinois EpiCenter (Unit 23)
in late 1988, construction of the University of Illinois EpiCenter was completed and a dedica- tion c--remony was held in November of 1ORS to commemorate its opening. The EpiCenter is a
wvorld-class facility that consists of seven MBE chambers interconnected by high-vacuum transfer tubes. Having the chambers,^ each^ of^ which^ is^ dedicated^ to^ a^ different^ type^ of material^ growth^ and characterization, interconnected by vacuum lines^ allows^ samples^ to^ be^ moved^ from^ one^ growth environment to another without external contamination. Since the planning for^ this^ new^ facility^ was
begun about five years ago^ and^ the^ constructi'w,^ has^ taken^ two^ years,^ the^ completion^ of^ this^ project^ is a major accomplishment. All of the JSEP discretionary funds provided under Unit 23 of the current contract were applied toward the purchase of several MBE chambers that will support future MBE research in the JSEP program. Since the^ EpiCenter^ is^ described^ in^ detail^ under^ Unit^ 23,^ further details will not be presented in^ this^ section.^^3
U
I
I
cI
C0~O
'0.02-I^ I
0. -1.0 -0.5 0.0 0.5 1.
I
Fig. 1. Conductance vs. Vd for GaAs/AIGaAs TIFET Device, I,, +5mA, Vgs IV
I
University of Illinois at Urbana-Champaign Coordinated^ Science^ Laboratory^3
I
U 3
U
TITLE: Crystal Growth from the Vapor Phase and Controlled (^) Doping of Equilibrium and Meta- stable Semiconductor (^) Alloys: Ion-Surface Interactions
J. E. Greene. Research Professor (^3) D. Lubben, Research Associate
3 F. Adibi, (^) Research Assistant P. Fons, Research Assistant B. Kramer, Research Assistant L. Markert, Research Assistant D. Mcintre, (^) Research Assistant D. Mce,Research Assistant J.-P. Noel, Research (^) Assistant R. Powell, Research Assistant M. A. Ray, Research Assistant I
SCIENTIFIC (^) OBJECTIVE:
The primary objective of this research program is to investigate energetic (^) particle-surface interactions (^) that control the nucleation and growth kinetics, chemistry, and physical (^) properties of alloy semiconductors (^) during vapor-phase crystal growth by UHV ion beam sputtering and accelerated-beam molecular beam epitaxy. In both of these growth techniques, (^) low energy ion- surface interactions allow an efficient coupling of kinetic energy (^) to the growth surface upon condensa- tion, thereby altering the surface reactivity as well as (^) adsorption and adatom diffusion kinetics allow- ing single crystal filn growth at lower temperatures, (^) much more precise control over dopant incor- poration probabilities and depth distributions, and the growth of unique metastable (^) alloys. This work is being pursued from both an analytical and an experimental point of view in order (^) to establish a detailed understan(ling of fundamental film growth mechanisms.
I
SUMMARY OF (^) RESEARCH:
Incorporation Probabilities (^) and Depth Distributions of Thermal and Accelerated Dopants in (^) Semi- conductors (^) Grown by MBE Most (^) common n- and p-type dopants used in bulk Si technology and many dopants used (^) in bulk GaAs provide problems (^) in MBE film growth due to either low dopant incorporation (^) probabilities and/or high surface-segregation rates. The problem (^) is especially acute in MBE Si. We have attacked these problems on two fronts. (^) (1) As discussed in the last two annual reports, we have developed a general time-dependent model, which combines thermodynamic and (^) kinetic elements, for describing the incorporation of thermal (^) dopants into films during deposition. (2) We have investigated the use of low-energy primary and secondary accelerated-ion doping (^) during MBE growth to demonstrate increases in ( bv more than five orders (^) of magnitude, abrupt doping profiles with no indication of segregati,-r,-i-1,,-ed (^) broadening, an!d ,--,,ont electrical properties. Primary-ion doping experiments
3 University of Illinois (^) at Urbana-Champaign Coordinated Science Laboratory
(r~k'I) i i % c '.t lllitdine. suirface. bulk, and ilhree incrinckhate .it, ) traniItin stat i tc idcl.
t][.11illiitcd (^) klexorptjon ( 151D) nmcasLIrClIleIt1N USiiie both thermal and .iccelcraitcd beamiis Sarlacc
Sh I-crc I(lnd l to dt.--ribe th le -nt~re '.cet of T st (V. I .R) (^) datat cmtreitnel% %%elI and~ pr(n-It ide % insi,_liit '111it~ a l I iolsattor ic h In ksiis and lo%% -Cnergp (^) 't)l/i :,t intera ti. IT 1', C!, ar. "or exam p!tha11t I. ii crpohiti( Mciclani'in fo accelrrait.d lIn mid Sb dopineL clialinie' diallllcall\ (^) *1l"lo%*
La .se'jkelllcrart d. and chiaractcri/Cd .a nsadjustable (^) grid. itilltiaptscratiirc 'ii '.ire I he, capathilit% i :'- clatke lateral mioton of the _,rid .illo.s~ hel icaerini- (^) ad ui~li1i I;ICnlI dUrIi>., "rCration. The -rid scparauoll (^) distance is also adjuistable 110-s/u ailo\ing indkepcnldcir u,)trl O%cr ilhc bcai current and \ oltagc x" hue lmainta-iiniiitiinuni11111 beam dil ereenClC. I XiCiC il licall idenCr.flcc. (^) a1Ild hiX as iniich as a factor of' three *r* a comfhlrahl.2 fixed grid ,( urcc, 'crc bIajiico - Fle dcpcendecec (^) of' the micasured bcami divergencc onl thet soure opcratine conlklilioii'. ailkl ;,rid ,coiir% (^) x\ crc exlie ae ilanl ion optics Ildl
3 1Electronic and] Optical Properties of Accelerated-Ion Doped NIBI" Si A Combination oil im lox-enicrgv and reflected high (^) enertv electron difIfractioni(kiliI and [IA-I). plait \ ic. and cross-sectional transmission (^) electron microscop\ (I N alld XIN), SIMNIS. tempratre depndetIflall mecasureinen ts, low-temiperatne il pholnriscelicc (0111 (^) (IT), anid deep-Iec l lransient spec (^) troscopv (Dl. TS) %\ereused to investigate the (^) structure, dopauit distributtion, and dlcc tri- cal properties of NIBI. Si( 100) films (^) growl, at 800 'C with either 200 eV In- or 150) eV Sb_ acclcrtcddopnt eams (^) The incorporation probabilities were found to be unity. A comiparison U (^) ct~ecn total kd(,ant concentrations deternilnedi *b* calibrated SfINS mecasuirements and dopaint con- centrations obtained from f lail data shows that both In an d Sb were incorpora1t ed into ,su~bs4tIrti (^) nla! electrically-active, sites. The concentration ranges (^) examined so far are 5xIl0 -2-x 10 cnir' for In. 16 , I%%xelI above! the equilibrium~l solid-soIlbilitV limit, and 10 -3x 10Cill for S, imore than anl order of mlagnIlitudeI abo(ve the highlest %-aluies obtainable bv thermal doigwithoul (^) trolWl!li1srcua
3 tion tree (^) Wvito iI..) indlication of residlual ion-induced damage.
Carrier mobilities were found to be equal to, or higher than, the best reported valuies for bulk Si. lin fact, hole mobilities measured (^) in In-t-doped films were the highest ever reported for In-doped Si and were much higher than mobilities for annealed hi-implanted Si. Temnperature-dependent ( 400 K) mobilities (^) of In +-doped films were , eii dc.;crit-ed b , theoret''-i -1lculations. 'Aith iio adjustable parameters, including lattice, ioniZed impurity, neutral Impurity, and hioil-ole scattering. The full effective mass tensor describing the non-isotropic andl (^) non-parabolic band struIC- I hire was used. D)oping profiles in Modulated structures (^) were ibrupt and in .situ ALS and RHI1xED analysis showed no indication of significant dopant surface accumu11.lations as were obser\ (^) ed during- the (^) growth of thermnally-doped films. Thus, uinder the present growth conditions, radiationl-inlduced
defects (^) were annealed (^) out at (^) a faster rate (^) than they (^) were produced resulting (^) in nok residual damage. (^) -S Recently, wve have obtained the first high quality photoluminescence ever reportedI from NI3E S films. In these experiments, we used a combination of low-temperature (4.2 K) IT and DLTS to
Icharacterize Si(100) films doped with E *.=200, 500, (^) and 1000 e V^75 As + ions (luring growth by molecular-beam epitaxy on n~ Sb-doped substrates at temperatures T, of 650 and (^) 800 'C:. Sharp no- phonion, (^) transverse-optical, and transverse-acoustic phonon assisted bound-exciton (BE) peaks associ- atedl with (^) As dopant species, together with broader, weaker, Sb)-related BE peaks, were the dlominant Pl. features obtained from 5-im-thick layers. No peaks ascribable (^) to residual ion-induced darnage were observed in films giet'.i at 650 'C vhi E, = 200 eV or T, 800 'C (^) w ith F., = 20,50 n 3 1000 cv'. No electron traps were observed by 1)1 S.
3University of' Illinois at Urbana-Champaign (^) Coordinated Science Laboratory
I. oiccI of Vim.'- L-neruy~ 1lon/Sur tlace Interactionis or) fte Ntuleation and( ( roN th Kinetics of' I-illis I11t1,osittd (^) firnt the \ apor Phaise~
I hIitrr~~iiQIftrim-Cnil renliirenient, ) or piisticated thini-tiltu device and processirie teelin T I I.I J iin liI1p'iii, liii ehlni rter Conltroil ov 1r tl.iit (riit'licin i r' and tit I(r( IIrI( Ji It I, I- [1k tI l d i%, r' 1 k - Ie lt Js e i si ng Ilokker It itnIpc raitiurt Lrouwii t t liJ it ies I (m'- ell( 1 f_ on rri> i tion ofti t he stra(e and filln d.urulin1 deposition is presentl\ heinlk- *Used h* laboratoriesI (ilt -''rI1 Ii It aricit o1 beam ndt lnahsdfl r% t echiqueis Iii order to pro%id( iit (cvnotiplim- of kineti c iiere\ to the e'thsurtace^ ttt^ alter^ reactivit^ as,^ %%ell^ as^ adsorption,^ ada !,,II jltnsiit JllIILICIC,0ii. aiiA g-rtitll kinCtiCS.I ( lr crotp ltaN been 'tideihi ok-tnerg', ion/surlace interaction effectS that arecommnon toa
nI 'lie1d Ultra 1111,i0 %,aCLILIIm environmients^ to^ Prob~e^ 111CC^ Ia^ IIIStillns.^ The^ iiil^ results^ of t:ii' I ork >rcdescribed Ii the last an nual report. Ow~-i past %car, weC ha\Vee\ elop~ed %M1S and TSDI technliques I br metasuiring hi ndi iuoner- a -til hence it errini2 tvnidinz, sites, of low-enero\ accelerated conldensine, species. As a refereniceI P-1tnt 'A e ha'-e dcternred not onily the two-diiensional phase^ dia~raiu^ I-r^ Ini^ on^ Sil^ l0fl)'xl^ ilutlso h i 1,n112 'rc Ier' Of Inl Inl each Of the surlace phiases. This is the first^ timei^ both^ complete^ struictural Ind ciereitiihrmat ion hias been available for anyv metal verhavei sx^ stem^ on^ Si. I i- triw- nrt an1d oarl ace morphiohog of Ini o\erla~ ct, on Si( IOIWIsimrtict. were ivest iatcd as an c itk1,1 1 ILII),t rate^ temiperatUre^ and^ surface^ coverage^ ising^ low-enert_,^ and^ reflectiontlilteee rutlit tract ion as:cI, as Auger electron spectroscopy. IDesorpt ion kinetics of' adsorbed Init as lield -aIltllolae-ba desorption and tempileratuire-prograiniied desorption^ spectroseopies. hi,,mdiiii a', tfound to grow on Si(100) according(_ to a Stranski-Krastario mechanism with thle initial I, 'mii it cI %e cralI two ()- d imecn sIinalI phases preceding,, the nucleation and growth of thire.e- dnwiei Ihlliislanids. Binding energies anid^ frequenct^ factors^ were^ extracted^ front^ the^ desorptionI ittea cit I ll' t- isiln a uMo del I ased on first -order desorption from se\ eral in trdependent surface p)hase', Iirst ordhcr and teroth-order kinetics w ere observed for the total desorbing^ fIlux^ fronm^ coexist- i-isirtace hse Iie uist' Of lo%% *eerg\ Ini^ conidensing^ species^ wvas^ found^ to^ completely^ changze^ thle^ nucleation arid r'' thmcl an^ isnt.^ Nucleation^ denusit^ ies^ were^ increased^ bN^ several^ oirders^ of^ magnitude^.^ the^ stir- Idthil mI "tA,Is enhtanlced - and complete three-dimensional ok erlaver \ a obtainedI at much lo%%erI t', I-r Iitial measur ements of biniding, eneries as a functioni 0l ionur have been completed.
.JSIK-SIONSOREI) l'Ii1lIC.-\IONS
1) -u N Ihira;htita, .I.. Markert. Y.-W.^ IKin.^ J^ L.^ Cireiue.^ WV.A.^ Ni.^ J.^ Knall.^ Gi.^ V'.
'A' Y)i IOns IdIUrie grow%\th hw mole()Ic ulIar beami CpIt axv .' A pI. J'h s. felt.-^ vol.^ -^. p- (5-JSs- l'/SR( -)
I 1Kntua^ lL^ NI^ A^ I^ asan,^ I^ I^ Suindgrn.^ A.^ Rockett^ ,^ L.^ Markert,^ a^ nd^ 1.^ F^.^ G;reenie,^ Indiunt !!.Cirpi rit nin itIi M1^ f.^ Si000)^ lavers^ using^ lovv-energ^ secondary^ and^ primar%^ ion^ implantation diuriitg- gri ikth. Ii Silicon Molecular Bemam Epitaxy 11,. I-C.- Bean and I.- J.- Showalter, V-ds I 'eirintinon NI. -1-lectrochem Soc. 1 988, p.- 4 17. (I SI1/SRC)
31 J P. Noel, N Ilirashita. L. C. Niarkert. Y.-W. Kimi. J F. (Treene., J. Kriall. Wv.-X. Ni. MI. A I asan. and J +.. Stinduen , T-lectrneal proiperties^ iif^ SI^ filmis^ (loped^ w~ith^200 eV^ Ili^ l^ ons^ duringI grok'%Ih ;,, molecular beauiii epitxv." 1. Appl. llh.,_ vol. 65. p. II,1989 CI(SIT/sR( I
[,.ni-vcrsitv of 1llinois at I -rI):naA- hanipruign Coordinaited Science I iLboratory
8I
TITLE: Studies of Transport Phenomena in Semiconductors
K. Hess, Research Professor n
I. C. Kizilyalli, Research Assistant .Higman,Research Assistanti S. Manion, Research Assistant
SCIENTIFIC OBJECTIVE: i This research involves the study of basic properties of semiconductors, semiconductor- heterolayers, new device concepts, and device simulation. Both theoretical and experimental methods are employed in each of these categories. We are examining a variety of hot electron phenomena and their effects on present and future device performance. The experimental studies concern mainly electronic transport in heterolayers (^) in high electric and high magnetic fields, while the theoretical studies are based on Monte Carlo methods (^) and, generally speaking, the use of large com- I putational resources.
SUMMARY OF RESEARCH: U In the past year, we have completed our research on the (^) hot electron heterojunction diode (H 2 ED), the velocity modulation (^) transistor (VMT), and basic investigations of hot electron effects in bulk AlGaAs, the high electron mobility transistor, and in real space transfer devices. This work has led to several publications and the demonstration (^) of microwave frequency operation of the H2ED. A major effort (^) has been made to explain and simulate transport at very high energies in silicon based devices. We have investigated the possibility of cold cathods. Cold cathods are p-n junction devices in which electrons are heated up so much that they can overcome the work function and be emitted out of the semiconductor material. The electron emission in such devices can be modulated
extremely fast and, therefore, presents opportunities that standard (^) hot cathods cannot offer. Since I the emission of electrons occurs at electron energies of 2eV or higher in the conduction band, these devices rffer a testing ground of high-field Monte Carlo theories and have revealed very (^) interesting physics, which has been described in an extensive publication [6]. In cooper- tion with R. W. Dutton (Stanford), we have applied our Monte Carlo codes to high- field transpo-i (^) - MOSFET's and have made a first attempt to investigate the possibility of merging standard d ir imulations and Monte Carlo methods. We have explained the detailed physics of I hot electrcr. ,iects such as substrate current and impact ionization. We have found that for devices with channel 1:;-.,i onger than one micrometer, a single Monte Carlo run on the basis of the field
profiles a! der., by standard simulations (PISCES) explains all experimental results well. We are currently (.'veloping methods for smaller featu-e sizes. The results have been (^) described in refer- ence [7].
I
University of Illinois at Urbana-Champaign Coordinated (^) Science Laboratory
1 9
The work on merging Monte Carlo investigations with standard device simulations has been per- formed in cooperation with the group of R. W. Dutton at Stanford. The work (^) on hot electrons in AlGaAs ', as in cooperation with K. Brennan (Georgia (^) Tech) and M. A. Littlejohn (North Carolina
i State).
U JSEP-SPONSORED PUBLICATIONS:
[11 (^1) C. Kizilyalli and K. Hess, "Ensemble Monte Carlo simulation of velocity modulation transistors (VMT) and real space transfer (NERFET, CILINT) devices," Superlatt. (^) and Microstruct., vol. 4, no. 3, pp. 287-288, 1988. (ARO/JSEP)
3 [2] J. Kolodzey, (^) J. Laskar, T. K. Higman, M. A. Emanuel, J. J. Coleman, and K. Hess,
"Microwave frequency operation of heterostructure hot electron diodes," IEEE Electron Dev. Lett., vol. 9, no. 6, pp, 272-274, 1988. (NSF/JSEP)
[3] M. A. Emanuel, T. K. Higman, J. M. Higman, J. M. Kolodzey, J. J. Coleman, (^) and K. tless, "Theoretical and experimental investigations (^) of the heterostructure hot electron diode," in Proc. Fifth Int. Conf. on Hot Carriersin Seinicond., Boston, MA, 1 _f); also, Solid-State Electron, vol. 31, no. 3/4, pp. 589-592, 1988. (ONR/JSEP)
[4] K. F. Brennan, D. H. Park, K. Hess, and M. A. Littlejohn, "Theory of the velocity-field relation in AlGaAs," J. Appl. Phys., vol. 63, no. 10, pp. 5004-5008, 1988. (JSEP)
[5] J. M. Higman, I. C. Kizilyalli, and K. Hess, "Nonlocality of the electron ionization coefficient in n-MOSFET's: (^) An analytic approach," IEEE Electron Dev. Lett., vol. 9, no. 8, pp. 399-401,
[6] J. M. Higman, K. Kim, K. Hess, T. van Zutphen, and H. M. J. Boots, "Monte Carlo simulation of Si and GaAs avalanche emitting dioaes," J. Appl. Phys., vol. 65, no. 3, pp. 1384-1386,
3 [7] J. M. Higrnan, K. Hess, C. G. Hwang, and (^) R. W. Dutton, "Coupled Monte Carlo-drift diffusion analysis of hot-electron effects in MOSFET's," IEEE Trans. Electron Dev., vol. 36, no. 5, pp. 930-937, 1989. (JSEP)
[8] D. W. Bailey, C. J. Stanton, M. A. Artaki, K. Hess, F. W. Wise, and C. L. Tang, "Ensemble Monte Carl( simulations of femtosecond energy (^) relaxation of photoexcited electrons in bulk GaAs," Solid-State Electron., vol. 31, no. 3/4, pp. 467-470 1988. (ONR)
U [9] I. C. Kizilyalli, K. Hess, T. Higman, M. Emanuel, and J. J. Coleman, "Ensemble Monte Carlo
Carriers in Semicond., 1987; also, Solid-State (^) Electron., vol. 31, no. 3/4, pp. 355-357, 1988.
University of Illinois at Urbana-Champaign (^) Coordinated Science Laboratory
I
TITLE: Basic Studies of the Optical (^) and Electronic Properties of Defects and Impurities in Com- pound Semiconductor Epitaxial Layers and Related Superlattices
G E. Stillnan. Research Professor
I
SCIENTIFIC PERSONNEL AND TITLES:
3 V. M. Robbins, Research Assistant
S. S. Bose, Research Assistant M. A. IHaase, Research Assistant M. H. Kim, Research (^) Assistant N. Pan, Research Assistant A. Reed, Research Assistant B. Lee, Research Assistant
1. Szafranek, Research Assistant D. Sengupta, Research Assistant
The objective of this research unit is to contribute (^) to our understanding of impurity incorpora- tion mechanisms, sources, (^) and defects and to improve our understanding of the influence of growth conditions on impurities and defects in semiconductor materials that will be important (^) for new multiple-layer compound heterostructure devices. (^) It includes developing new characterization tech- niques that will extend the range of impurity concentrations over which quantitative analysis is possi- ble. These techniques (^) will lead to better control of high-purity growth and accurate doping levels in epitaxial layers grown by metalorganic chemical vapor deposition (MOCVD), molecular beam epitaxy (MBE), and chemical beam epitaxy (CBE) or gas source molecular (^) beam epitaxy (GSMBE) growth techniques that are most important for the preparation -,f multiple layer heterostructures for high- speed electronic and optoelectronic devices.
The capability (^) of donor identification in n-type and certain p-type high-purity GaAs samples using low-temperature magnetophotoluminescence with a superconducting solenoid and a one-meter, double slit grating spectrometer was (^) established. By correlating photothermal ionization measure- ments with the magneto-photoluminescence (^) measurements on the same samples, the ls-2p 0 and Is- 2p-1 two-electron transitions of the neutral donor-bound excitons (D*,X), corresponding to the com- mon donor species in GaAs, have been identified. This technique has been extended to the identification of the residual (^) donor species, Si, S and Ge, in high-purity undoped p-tpe epitaxial GaAs, (^) grown by metalorganic chemical vapor deposition and arsenic trichloride vapor phase techniques, using the magnetic splittings of "two-electron" replicas of donor-bound exciton transitions at low temperature (-1.8K) (^) and at a high magnetic field (9.0 T). This technique permits identification (^) of donors in certain high-purity p-type GaAs samples that have sufficient neutral donor-bound exciton recombination but in which donor species cannot (^) be identified
University of Illinois at Urbana-Champaign Coordinated Science Laboratory
by photothernal ionization (^) spectroscopy or anN other technique.
Carbon is a ubiquitous impurit in essentially all epitaxial growth techniques for gallium arsenide. Although carbon^ is^ from^ Group^4 and^ is^ potentially^ amphoteric,^ it^ has^ never^ been ob,;er ,ed as a donor in GaAs However, it is a common residual acceptor as well as a useful inten-
tional p-type dopant, especially for thin epitaxial layers and multiple quantum well structures because of its low diffusion coefficient.
We have studied the source of residual (unintentional) carbon impurities in MBE GaAs by intentional injection of CO into an MBE svstem. High-purity Si-doped MBE GaAs layers grown with and without the intentional introduction of CO gas have been characterized by Hall effect measure- ments, photoluninescence, and photothermal ionization (^) spectroscopy. The results indicate that CO itself is not the source (^) of residual C acceptor impurities in MBE GaAs samples. The observations of
the correlation of residual C impurity incorporation with the residual CO gas in the MBE growth chamber suggest that the partial pressure of CO, Pco, (^) gives a quantitative indication of background levels of unidenitified hydrocarbons that are the source of C (^) acceptors.
Our results, combined with the earlier results discussed above, show that the C source in MBE reactors that is responsible f'r- the residual C acceptors in MBE (^) GaAs samples is not CO but prob- ablx other hydrocarbons. The elimination of these residual (^) hydrocarbons in the MBE reactor was
claimed to be an important step for the growth of high-purity GaAs. Further experiments would be desirable to confirm this hypothesis. Since the hydrocarbons, including CH 4 , are desorbed relatively faster than CO, their background levels can eventually reach or go below the detection limit of the mass spectrometer. Even though the partial pressures of hydrocarbons are near or below the detec-
tion limit, the C impurity concentration can still be high if the sticking coefficients of the hydrocar- bons are near unity. (^) The residual partial pressure of CO is a qualitative indication of the cleanliness of the growth reactor and of the background (^) levels of hydrocarbons that are the sources of C accep- tors.
In MOCVD growth, C is (^) an intrinsic impurity because of the metal organic sources, so it is important to be able to minimize the incorporation of these intrinsic C impurities (^) except when they
are desired for doping of the epitaxial layers.
Metalorganic chemical vapor deposition (MOCVD) of epitaxial layers of III-V compounds is among the most widely used growth technique. As greater understanding and control of (^) the MOCVD growth process have been achieved, improvements in surface morphology, layer thickness uniformity,
interface quality, and purity have led to ongoing improvements in device quality and performance.
One important area that is not fully understood is the process of residual C acceptor incorpora- tion in GaAs grown by the MOCVD technique using TMGa and AsH 3. There is some question
about which chemical species carries the C impurity to the epitaxial layer. The methyl groups of the TMGa molecules are widely believed to be the source of carbon. On the other hand, (^) hydrocarbon impurities were detected in the residue from fractionally (^) distilled TMGa. Growths using redistilled
TMGa [21] were found to have lower C acceptor concentrations [21]; thus, these hydrocarbon resi- dues may be a source of carbon. The residual C acceptor concentration in GaAs has (^) been found to be dependent on several growth parameters: V/III ratio [22-27], substrate temperature [23-27], sub- strate orientation [26,27], and the reactor pressure (28]. These are all important pieces of information needed to understand the sources and incorporation (^) mechanisms of residual C; however, a full under-
standing of C incorporation has proved difficult based on these facts alone.
The mechanisms governing C incorporation can be understood in terms of the chemical reac- tions. The curve in Figure (^1) is a line of constant concentration of surface adsorbed, AsH 3 *. This
line represents the minimum AsH 3 * concentration needed to prevent deleterious effects. Concentra- tions of AsH 3 * (^) below this minimum lead to increased carbon incorporation and/or instabilities in the
carrier concentration. The source of carbon can be understood in terms of the concentration of adsorbed (CH 3 ) 3 Ga on the substrate surface and the reactions shown in Figure 2. At low growth temperatures, reaction (2) proceeds at a relatively low rate allowing a large number of (CII.),Ga
molecules to reach the substrate. But if the V/Ill ratio is increased, the abundant AsIl,* (^) will react I
University of Illinois at Urbana-Champaign Coordinated Science Laboratory 3
I
141
I
I
(CH 3 ) 3 Ga AsH 3
0 0 AsH 2 + H
(CH 3 ) 3 Ga-I AsH; +
AsHi ' 3
I
1
I
Fig. 2. Drawing of chemical reactions influencing carbon acceptor incorporation.^ I
I I I I I I
University of Illinois at Urbana-Champaign (^) Coordinated Science (^) LaboratoryI
I
with the (CH 3 ) 3 Ga. the driving reaction (5) at a high (^) rate and reducing [(C11 3 ) 3 Ga]. This results in reduced carbon incorporation. When the growth temperature (^) is increased, more (CH 3 ) 3 Ga is depleted from the gas via reaction (2) and fewer (CH 3 ) 3 Ga reach (^) the substrate; hence, a smaller V/Ill ratio is sufficient to provide the requisite quantity of (^) AsH 3 * to drive reaction (5) at a high rate. These reactions account for the low-temperature "arm" of Figure 1. Further increases (^) in the growth temperature will eventually increase the rate of reaction (1) and reduce the concentration of adsorbed AsH. Increasing the V/Ill ratio will provide the requisite concentration of adsorbed (^) AsH 3 to react with [(CH 3 )3Ga1 and reduce C incorporation. These (^) reactions account for the high-temperature ".arm" of Figure 1. Thus, (^) C incorporation requires that (^) an uncracked TMGa molecule (^) adsorb on the
substrate and remain there unreacted. Reactions that interfere^ with^ this^ process^ result^ in^ reduced^ C acceptor incorporation. These results are direct evidence that (^) the residual carbon acceptor concentration is controlled by two chemical reaction mechanisms: (1) the removal of the first methyl group (^) from gas phase TMia, (^) and (2) the removal of the first hydrogen atom from surface adsorbed AsH 3. An (^) analysis of the chemical reaction shows that surface adsorbed TMGa is (^) the source of residual carbon acceptors in GaAs grown by MOCVD with TMGa and AsH 3 sources. These chemical reactions have been controlled to reduce the C acceptor concentration, resulting in high-mobility GaAs epitaxial (^) layers. I
3 [11 (^) S. S. Bose, M. H. Kim, and G. E. Stillman, "Identification of residual donors in high (^) purity undoped p-type epitaxial GaAs by magneto-photoluminescence," (^) Appl. Phys. Lett., vol. 53, pp. 980-982, Sept. 1988. (DARPA/AFOSR/JSEP/NSF-CDR)
[2] M. J. McCollum, M. H. Kim, S. S. Bose, B. Lee and G. E. Stillman, (^) "High purity epitaxial indium phosphide grown by the hydride technique," (^) Appl. Phys. Lett., vol. 53, pp. 1868-1870, Nov. 1988. (NSF-ECS/NSF-DMR/JSEP/NSF-CDR)
[3] B. Lee, S. S. Bose, M. H. Kim, G. E. Stillman, and A. R. Calawa, "Spectroscopic studies of the influence of intentionally (^) increased CO partial pressure on the Incorporation of residual C impurities in MBE (^) GaAs," Appl. Phys. Lett., vol. 53, pp. 2438-2440, Dec. 1988. (JSEP/NSF)
[4] A. D. Reed, S. S. Bose, (^) and G. E. Stillman, "Residual carbon acceptor incorporation in gallium arsenide grown by metalorganic chemical vapor deposition," Appl. Phys. Lett., vol. 54, pp. 1262-1264, Mar. 1989. (JSEP/DARPA)
[5] B. Lee, K. Arai, B. J. Skromme, S. S. Bose, (^) T. J. Roth, J. A. Aguilar, T. Lepkowski, N. C. Tien, and G. E. Stillman, "Spectroscopic (^) studies of the influence of oxygen partial pressure on the incorporation of residual silicon impurities in vapor (^) phase epitaxial gallium arsenide," J. Appl. Phys. (to appear). (^) (AFOSR/JSEP/DARPA/NSF-CDR)
[6] B. Lee, S. S. Bose, M. H. Kim, A. D. Reed, and G. E. Stillman, "Orientation (^) dependent amphoteric behavior of group (^) IV impurities in the molecular beam epitaxial and vapor phase epitaxial growth (^) of GaAs," J. Cryst. Growth (to appear). (NSF-DMR/NSF-CDR/JSEP)
I
SUniversity (^) of Illinois at Urbana-Champaign Coordinated Science Lahoratory
