- Email: [email protected]
MBE techniques for IV–VI optoelectronic devices
Prog. CrystalGrowthChamct.1979,Vol. 2, pp. 49-94.PergamonPre~ Ltd. Pdntedin GreetBritain.
MBE TECHNIQUES FOR IV-Vl OPTOELECTRONIC DEVICES* H. Holloway Ford Motor Company, Dearborn, Michigan 48121, U.S.A. and
J.N. Walpole Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, Massachusetts 02173, U.S.A. (Submitted July 1978)
ABSTRACT Recent development of hlgh-quality IV-VI optoelectronic devices grown by MBE has significantly increased the technological importance of this epltaxial technique in IV-VI materials. Despite the progress made, much is still unknown about the importance of crystal perfection for device performance. Criteria for crystal perfection required to minimize carrier recombination and optical losses in i.r. devices in IV-VI materials need to be established. In this paper the literature on IV-VI film growth by vacuum deposition techniques is briefly reviewed. MBE techniques used for growth of IV-VI materials on BaF 2 and SrF 2 substrates and for growth of PbSnTe on PbTe substrates are described. Emphasis is on the techniques used to deposit pseudobinary alloys with homogeneous composition. The use of foreign impurity dopants is also discussed. Criteria of quality and crystal perfection of epltaxlal layers needed for device performance are evaluated. In layers grown on IV-VI substrates, large dislocation densities due to lattice mismatch are normally present but may not be detrimental to device performance. Lattice-matched heteroepitaxlal systems are discussed. For insulating substrates, it is shown that the crystalline quality and the carrier mobility of layers grown on BaF 2 and SrF 2 are superior to those grown on alkali halide substrates and that the latter are unsuitable for making devices. Carrier mobility is not a good test of crystalline perfection, however, it is argued that device performance, particularly in the demanding requirements for low-noise photodiodes, is a sensitive measure of the epitaxial material quality. Thinfilm photodiodes grown by MBE and the hot-wall techniques are then reviewed in detail, including some recent unconventional * The Lincoln Laboratory portion of this work was sponsored by the Department of the Air Force and the Defense Advanced Research Projects Agency.
49
50
H. Holloway and J. N. Walpole devices in which the unique properties of thin films are exploited. Heterostructure optical waveguides are described briefly followed by a review of diode laser results obtained by MBE and the hot-wall technique. These results are for thin-film PbTe lasers grown on BaF 2 and for a variety of devices grown on I-VI substreates including single heterostructure, double heterostructure, homostructure, and distributed feedback lasers.
i.
INTRODUCTION
The first major application of the IV-VI semiconductors was the use of PbS layers as infrared-sensitive photoconductors in the 1940's. Most of the IV-Vl photoconductlveworkhas been with polycrystalline precipitates, which represent a divergent evolutlonary llne from the optoelectronlc devices that are described in the present article. However, the early studies of photoconductive PbS included vacuum deposition of thin films (1,2), which led to a demonstration of epltaxlal growth on NaCI (3) that may be regarded as the progeniotor of modern epitaxlal IV-VI devices. In a second wave of development during the 1960's, techniques were established for the growth of goodquality single crystals of IV-Vl semiconductors and for making p-n junction devices that are discussed here, The selection of subject matter for the present review involves some arbitrarily chosen boundaries because there is not a well-deflned separation between thin-film and bulk crystal growth techniques. At one end of the spectrum we have vacuum deposition of Dm thick films using a molecular beam that is generated with an effusion cell. At the other extreme, we find closed-tube transport down a temperature gradient to give crystals with cm dimensions either by spontaneous nucleation or with use of a seed crystal (4). This range of techniques includes both epitaxy on isostructural IV-VI substrates (or seeds) and or insulating, such as BaF2.* Much of the recent literature on growth of IV-VI thin films relates to a tehnique in which the source and the substrate are enclosed by a "hot-wall" that pipes most of the evaporant to the substrates. ** This method may be regarded as intermediate between molecularbeam epitaxy (MBE) *** and closed-tube transport. Thus, depending upon the wall temperature, the efflcleny of the piping and the approximation to equilibrium, growth conditions may vary between the extremes that are obtained with MBE and with closedtube transport. This intermediate character is also evident in the hot-wall layer thicknesses which may exceed IOO ~m (8), thereby approaching bulk crystal growth. In the following, the description of epitaxial growth techniques is restricted to two representative MBE systems that were developed by the authors and their colleagues to pursue their respective interests in epitaxial i.r. detectors and epltaxlal i.r. lasers. We have mostly omitted reference to extensive body of earlier work on MBE of IV-VI semiconductors because, despite its value for fundamental studies of the materials, the emphasis on alkali halid substrates led to layers whose crystal perfection was inadequate for p-~n juncation devices.
Insulating substrates are usually associated with tin films of the IV-VI semiconductors, but Pandy 5 has shown that closed-tube growth of (Pb,Sn) Te may be initiated with a BaF 2 seed. The hot-wall technique appears to have originated with Koller and Gohill (6) who used it to deposit ZnS. Subsequently it has been used extensively for IV-VI semiconductors (7-14). *** Molecular-beam epltaxy is a recent designation for a much older technique.
This
MBE Techniques for IV-Vl Optoelectronic Devices
51
name i s q u i t e d e s c r i p t i v e f o r t h e IV-VI s e m i c o n d u c t o r s where the beams a r e r e a l l y of m o l e c u l a r s p e c i e s . The term i s a l s o a p p l i e d ( 1 5 ) , though w i t h l e s s j u s t i f i c a t i o n , t o growth of m a t e r i a l s l l k e t h e I I I - V s e m i c o n d u c t o r s , where one works w i t h beams of the c o n s t i t u e n t e l e m e n t s . These studies have been reviewed by Zemel (15b). Epitaxial growth by the hot-wall technique has also been omitted because a recent review is available (16). However, for completeness we have included the results of recent hot-wall studies which indicate that this technique can yield devices whose performance is comparable to those made with MBE. We have excluded consideration of llquld-phase epltaxy because this would take us too far from the main area of our review, but we note in passing that this method has been shown to give high-performance IV-VI photodiodes (17,18) and lasers (19) on IV-VI substrates.
2.
TECHNIQUES FOR MOLECULAR-BEAM EPITAXY
Vacuum d e p o s i t i o n of t h e IV-Vl s e m i c o n d u c t o r s and t h e i r p s e u d o b i n a r y a l l o y s has m o s t l y been accomplished by e v a p o r a t i v e methods, a l t h o u g h some use has been made of f l a s h e v a p o r a t i o n (20) and of s p u t t e r i n g ( 2 1 ) . From t h e l i t e r a t u r e one may conclude t h a t almost any o f the s t a n d a r d vacuum d e p o s i t i o n t e c h n i q u e s w i l l y i e l d e p i t a x i a l l a y e r s , so t h a t t h e c h o i c e of method i s t o some e x t e n t a m a t t e r of t a s t e . However, e v a p o r a t i v e t e c h n i q u e s a r e p a r t l c u l a r l y a t t r a c t i v e b e c a u s e of t h e IV-VI s e m i c o n d u c t o r s have c o n v e n i e n t vapor p r e s s u r e s ( P ~ 0 . 1 T o r r a t 1000 K) and s u b l i m e p r e d o m i n a n t l y as d i a t o m i c m o l e c u l e s ( 2 2 - 2 4 ) . Even with evaporative techniques, care is needed when the deposit is a pseudobinary alloy. Use of a source of the alloy (25,26) does not, in general, give truly congruent evaporation. For example, Northrop (27) has shown that the gas phase in equilibrium with Pbl_xSnxTe is enriched in the SnTe component. The enrichment is roughly 30% over the range of compositions (x • O.2) that are suitable for i.r. detectors. With use, a Pbl_xSnxTe source that is operated near equilibrium will become depleted of SnTe and the energy gap of the deposit will drift towards larger values. This places a severe constraint upon the fraction of the source material that may be used if a uniform deposit composition is required. From Northrop's results a source with the composition Pb O 85Sno.15Te when operated near 7OOOC will initially give a deposit with composltion near Pbo.8oSno.2oTe corresponding to a cut off wavelength of approximately 11 ~m at 80 K. After use of 20% of the source material the source and deposit compositions may be estimated to be Pbo.86Sno.14Te and Pbo.81Sno.19Te, respectively.. This would lead to a decrease of the 80 K cutoff wavelength of the deposltied semiconductor by approximately 0.Spm.
E v a p o r a t i o n of a p s e u d o b l n a r y a l l o y w i t h o u t a d r i f t i n i t s c o m p o s i t i o n i s p o s s i b l e , p r o v i d e d an open c r u c l b l e r a t h e r t h a n a Knudsen c e l l us u s e d , and p r o v i d e d t h e source m a t e r l a l i s made homogeneous and i s i n a s o l i d ( r a t h e r t h a n g r a n u l a r or powered) form. D i f f u s i o n of t h e a l l o y c o m p o s i t i o n can t h e n become s u f f l c i e n t l y s m a l l compared to the rate of removal of the surface material that the solid never reaches equilibrium with the vapor. Electron microprobe analysis (28) of an angle-lapped (5°) 8-Bm film grown on a PbTe substrate at approximately 38OOC from a Pbo.88Sno.12Te alloy source showed a constant composition throughout the film of O.120 ~ 0.OO1SnTe** At the interface of the film with the substrate, the profile shown in Ylg. 1 was obtained. The errors in SnTe composition are based on the statistical scatter obtained in four different traces through the region. S t r a u s s (29) has r e p o r t e d l a r g e r v a r i a t i o n s i n f i l m c o m p o s i t i o n i n growth of PbSnSe f i l m s from p s e u d o b i n a r y a l l o y s . However, t h e s o u r c e i n g o t s were n o t s u b j e c t e d t o a homogenizing h l g h - t e m p e r a t u r e a n n e a l .
52
H. Holloway and J. N. Walpole
The error in distance of + 0.5 Bm is an estimate of the minimum error due to penetration of the microprobe-into the angle-lapped surface. Larger errors are likely. The lapping produces some smearing of the probed surface which may lead to additional errors. In any case, the transition region appears to be < IBm. Since this film was grown at IBm hr-Y, t h e 8 - h r growth time represents an unusually long period and the diffusion of the Pb-Sn composition should be the maximum normally encountered.* 14 13 12 i
I
•
~
I
ll ¢
10
C
8
O L ¢
o_
,
9
7 6
o
5
E
4
0 <
~
,
F-C CO
,
3 2 I
ar
|.
1 B
I
I
i
I
I
2
3
4
Oistonce Fig. i.
(micrometers)
Variation of alloy composition with distance near the interface of a Pbo.88Sno.12Te layer on a PbTe substrate as determined by electron microprobe analysis on an angle-lapped (5o) surface.. The uncertainty in distance is the minimum estimated error. The absolute SnTe composition by =2% although relative changes can be more accurately determined as indicated by the vertical error bars.
The vacuum system used for MBE of PbSnTe on PbTe substrates is shown schematically in Fig. 2. Ion pumping with a system pressure during growth of about 10-7 Torr was used. The substrate was radlently heated in a small furnace open at one end. The rate of deposition was controlled by a feedback signal to the source-heater power supply. The feedback signal was generated from the deposition rate onto a quartz-crystal film-thickness monitor. Rates of 1-3 Bm hr -I were used with substrate temperatures of 350 - 45OOC. The sources were contained in open pyrolytic boron nitride crucibles heated in a tantalum furnace. The distance between the sources and the substrate was about 20 cm. Separate sources were used for each type of layer grown. Each source was prepared form zone-melted (32) binary compounds Bicknell (30,31) has used X-ray techniques to investigate grading in PbSnTe films grown by a vapor epltaxy technique at higher substrate temperature (5OO-600oc). He finds somewhat more Pb-Sn composition diffusion.
MBE Techniques for IV-VI Optoelectronlc Devices
53
(PbTe and SnTe) weighed out in the proportions required to obtain the desired alloy composition.
IThermocoupl,e
11
~L
t
I
I
I1:!1-
'
illl
ThicknmessI oo.o,
,, Deposi Ratetion J Controller
I Gold/Quartz I Radiant Heater (vocuum" lo-TTorr)
I Power I Supply ,
~ ~-~ron Nitride
i~ Fig. 2.
I
Crucible
Schematic diagram of vacuum deposition system for epitaxial growth from PbSnTe alloy sources on PbTe substrates.
Before use, the sources were homogenized by sealing the constituents in an evacuated quartz ampoule and heating above the melting point for several hours followed by rapid quenching in water to room temperature. Subsequently, the ingot was annealed for 5 days below the melting point at 700oc and quenched again. Pieces of the polycrystalline ingot were then broken off and used as solid chunks in the source furnaces. When undopedsources were used, very low concentration films were obtained determined by waveguiding experiments discussed below. Foreign impurities used to control carrier type and concentration rather than controlling the from stoichlometry. Impurities are expected to diffuse much less than the from stoichiometry (33).
as were deviation deviation
For n-type films, Bi was added in controlled amounts and for p-type films TISe was added. Figure 3 shows the carrier concentration in the film compared to the concentration of dopant in the source. Above about 5 X 1018 cm -3 the film concentration appears to saturate for both types of dopants. The film concentration was measured using the carrier concentration dependence (34) of the infrared reflectlvlty minimum at the plasma edge. The u s e o f s e p a r a t e e f f u s i o n s o u r c e s f o r t h e c o m p o n e n t s o f p s e u d o b l n a r y a l l o y s * may a l s o b.e t r o u b l e s o m e b e c a u s e t h e r a t i o o f t h e c o m p o n e n t f l u x e s m u s t b e p r e c i s e l y controlled, I f o n e a s s u m e s a n e f f u s i o n c e l l t h a t o p e r a t e s a t a b o u t I 0 0 0 K, a t y p i c a l tmnperatur~ fluctuation o f + i K t o g e t h e r w i t h a s u b l i m a t l o n e n t h a p y o f a b o u t 50 kcal mole ~a gives a fluctuation ~ 2.4% i n the effusion rate. With two independent
54
H. Holloway and J. N. Walpole
sources the overall fluctuation in the ratio of the components is then + 5%. This is unacceptable in cases where the energy gap is a sensitive function of composition.
SUBSTRATETEMPERATURE425 °C GROWTH RATE 3p.m/hr Pbo.88Sno.12TeLAYERS
I
'E oo
.J~
[3
TISe r 7 ~ / ~
z
8i
U
0
Fig. 3.
If
I
2
I
I
I
I
I
I
I
1
4 6 8 I0 12 14 16 18 20 CONCENTRATIONIN SOURCE ( 1018cm-3)
Calibration curves of Bi and TISe doping efficiency in Pbo.88Sno.12Te films.
Composition fluctuations may be reduced by using a pair of thermally-linked cells (36) The idea is, that once the temperature fluctuations of the two cells have been minimized, their effect upon the composition can be further reduced by having them occur in phase. A change in the effusion rate of one of the components is then partially offset by a corresponding change in the effusion rate of the other component. For small temperature fluctuations the fractional change in the effusion rate of one of the components may be expressed as
J
J
~T RT 2
It is also possible to use separate effusion sources for the elemental constituents (35), but this seems unnecessarily complicated because, apart from the increase in the number of cources, the large range of volatility of the constituents makes radiation shielding between the sources necessary. (E.g. for growth of (Pb,Sn)Te the equilibrium vapor pressures of the constituent elements at IOO0-K are: Pb, 10-2 Torr; Sn, 5 x 10-8 Tort; and Te, 2 Tort. In contrast, the binary compounds PbTe and SnTe both have vapor pressures of about 0.i Tort at IO00-K).
MBE Techniques for IV-Vl Optoelectronic Devices
55
where AH s is the enthalpy of sublimation. For a pair of effusion cells that operate at the same temperature, but with independent random temperature fluctuations, the maximum change in the flux ratio is:
A • Jl
Jl
~22
'~ ~ + - -J2
max
(AHs, i+ AHS~2)
AT,
RT 2
where AH 8 1 and AH, ~ are the enthalpies of sublimation of the two components. If the temperature fluctuations are now constrained to occur in phase, the maxlmum change in the flux ratio is reduced to:
Jl
Jl
'b + - -
J2
"~ - J2
(AHsrl - AHs~2) AT. RT 2
max
The magnitude fo the effect may be estimated by again considering the deposition of Pbl_xSnxTe with Eg ~ 0.11 eV. The enthalpies of sublimation of PbTe and SnTe are 52.3 kcal mole -l, respectively. With a pair of effusion sources that operate at 1000 K with independent random fluctations of i K, the flux ratio can change by + 5 % giving Ax - + 0.01 and E 2 - + 6 meV. In contrast with in-phase temperature fluctuations, the fl~x ratio cha~ges-by about* ~ 0.3% to give Ax ~ ~ 6 X I0 -~ and AEg ~ + 0.4 meV. In practice, the temperature control can be improved to about 0.2 K with a corresponding further improvement in homogeneity. Schmatics of the deposition system and the double cell are shown in Figs. 4 and 5.
3.
THE CRYSTAL PERFECTION OF I V - V I EPITAXIAL LAYERS
The work described in this review has made use of IV-VI semiconductor substrates, principally for work with lasers, and insulating substrates, principally for work with photodiodes. There is relatively little prior information on the properties of the layers on IV-VI substrates, largely because good single crystal substrates are not readily available.** With IV-VI substrates, it is possible in principle to achieve a match in the lattice constants of the substrates and epitaxlal film. This may be technologically important as a means of reducing the density of the misfit dislocations that occur in such IV-VI structures (37,39). Figure 6 shows the lattice constant vs energy gap for the lead-tin chalcogenlde compounds and alloys of most interest. The lines representing the ternary alloys have been drawn as straight lines between the points representing the binary compounds and hence are only approximate. In a study of heteroepitaxy of Pbl-xSnxTe on {1OO} substrates of different alloy composition using both MBE and LPE (40) it was found that etchpit densities in layers several ~m in thickness were invariably greater than 107 cm -2 even for differences in tin composition as small as 2%. Homoepitaxy or nearly latticematched heteroepitaxy of *
There is a significant uncertainty here because the possible errors in the sublimation enthalpies are comparable to the 6 kcal value for their difference.
** The unsuitability of these conducting substrates for the transport measurements which have domlnated such work may also have had an influence.
•~
H. Holloway and J. N. Walpole
Bell jar
r
/
'/
Substrate Holder ~
S.S Mask - ~ Graphite Double Cell
Quartz tube Holder
\Tantalum
/
Heaters
/
Thermocoq
Ceramic SuPl~ Rods ( 2 )
Fig. 4.
Schematic diagram of a vacuum deposltion system for epitaxial growth of IV-VI semiconductors on fluorite structured substrates.
SUBSTRATE EFFUSION DOUBLE CELL GRAPHITE CASING QUARTZ OR ~ F QUARTZ OR CERAMIC PINS \ ~ E R A M I C RINGS
/ TANTALUM HEATER Scole: Double Size Fig. 5.
Isothermal double cell for evaporation of IV-VI pseudobinary alloys.
Pbo.88Sno.12Te on PbTeo.952Seo.048 (or vice versa) yielded etch-it densities
MBE Techniques for IV-Vl Optoelectronic Devices
57
comparable to that of the substrates used, < IO 4 cm- 2 . Lattice-matched heteoroepitaxy of PbSnSe on PbSSe has not been reported but appears particularly attractive since both these alloys have good metallurgical and mechanical properties especially compared to PbTeSe. To date, no optoelectronlc devices have been reported which were grown using lattice-matched heteroepitaxy. However, the device results to be discussed suggest that misfit dislocations may have little effect on minority carrier liftime in films* and there is evidence that recombination at hetero-lnterfaces can be quite small.
0.30~-- ~PbS
1:
°2°'J-s,
E
0.20 --
(.~ 0 Z W
BOUNDARY ~ OF SINGLE-PHASE ! ROCK-SALT e TURE
o.15> a. ,It
0.10
o_ or. o
--
15
f, Z
_o
/
3¢ I",.. I",,. "'
T; '°.,2s°°.,,
30
no.18Se
W 0 "it
ID INVERSION Pb_ __Sn___1
0.05
-- 1
' I
0.10i --
t
5.9
6.0
15
I
/
0.1~--
Fig. 6.
Z W -J W
I Sns'
6.1 6.2 6.3 LATTICE CONSTANT (AI
6.4
6.5
Variation of energy gap at 77 K (wavelength of emmission or absorption edge) with latlce constant for the lead-tin calcogenlde alloys.
Bickne11 (41) has shown t h a t m i s f i t d i s l o c a t i o n s i n vapor e p i t a x i a l PbSnTe have no e f f e c t on m i n o r i t y c a r r i e r d l s s u i o n l e n g t h .
layers in
58
H. Holloway and J. N. Walpole
•n contrast to the situation with IV-VI substrates, the growth of IV-Vl layers on insulating substrates has a substantial literature (15) that is mostly devoted to growth on alkali halide substrates, although occasional use has been made of other materials, such as CaF 2 and mica. This application of such semiconducting thin films to p-~ junction device technology demands a more critical approach to crystalline perfection than has been customary in the literature on epitaxial growth. Determination of the degree of epitaxy in IV-VI layers has been mostly by diffraction techniques that are insensitive to the spread in orientation that is associated with a small-grained mosaic structure. An early study (42) of PbS on NaCI demonstrated the presence of a mosaic structure with grain sizes of the order of I000 . Similar results were obtained later with PbTe on NaCI (43) which was found to have about 5 x iO I0 dislocation/ cm 2 that were mostly arranged to form grain boundaries with about 2000 spacing and about 2° misorientation. An electron mlcrograph that reveals the typical grainy structure of a PbTe layer on an alkali halide substrate is shown in Fig. 7.
Fig. 7.
Parlodlon replica of epitaxial PbTe grown on cleaved KCI. The region shown is about 40 ~m wide and the darker regions are PbTe grains that have been pulled out of the film during replication.
Despite the early indications of a mosaic structure in epitaxial IV-VI semiconductors on alkali halide substrates, much of the subsequent work ignored the crystal perfection of the layers and, consequently, has little relevance to p-n junction device applications. The situation resembles that in the early stages of development of epitaxial III-V semiconductors, such as GaAs, and earlier sonlnents (44) about the inadequacy of conventional diffration techniques for assessment of the crystal perfection of II-V layers apply with equal force to the IV-VI semiconductors. Such problems of characterization appear to be widespread with epitaxial semiconductors.
MBE Techniques for IV-Vl Optoelectronlc Devices
59
A commonly r e p o r t e d p r o p e r t y of IVIVI e p i t a x l a l l a y e r s i s the H a l l m o b i l i t y . Layers t h a t have been p r e p a r e d under a wide r a n g e of c o n d i t i o n s on v a r i o u s s u b s t r a t e s tend to e x h i b i t b u l k l i k e v a l u e s o f the H a l l m o b i l i t y a t 300 K and, t o a l e s s e r e x t e n t , at temperatures down t o 77 K. However the scattering lengths in these materials are too small for the carrier mobilities to give significant information about the crystal perfection. In the temperature range 300-77 K, the largest mobilltles that are observed with the lead chalcogenides follow,
~H ~
T-a' a ~ 5/2,
and this relationship has been interpreted in terms of scattering by phonons (45). At lower temperatures the mobility tends to saturate at values that have been attributed to scattering by ionized native defects (46). For the lead chalcogenldes, the phono-limited mobilities of both p-and n-type materials are about 2.5 x 104 cm2v-lsec -1 at 77 K. For typical carrier concentrations in the range 1016 - 1018 em -3 this corresponds to scattering lengths of only about 2000 . Thus, the 77 K Hall mobility cannot give information about departures form perfect crystallinity on a scale greater than a few thousand . Even at 4 K, where the Hall mobility of PbTe may be of the order of 106 cm2v-lsec -1, the scattering lengths still do not exceed a few inn. The utility of the mobility as a c r i t e r i o n f o r the p e r f e c t i o n of e p i t a x i a l l a y e r s i s f u r t h e r reduced by the f a c t t h a t even good d e v i c e - q u a l i t y b u l k c r y s t a l s f r e q u e n t l y f a l l to a t t a i n the f u l l p h o n o n - l i m i t e d m o b i l i t y a t 77 K. Thus, h i g h performance b u l k c r y s t a l PbTe p h o t o diodes have been made f r o m ~ - t y p e m a t e r l a l w i t h WB(77 K) - 1.0 x 104 cm2v-lsec -1 (47) r a t h e r t h a n the phonon l i m i t e d v a l u e . These d e p a r t u r e s form the p h o n o n - l i m i t e d m o b i l i t y appear to c o r r e l a t e w i t h the t h e r m a l h i s t o r y of the specimen and i t has been s u g g e s t e d t h a t they a r i s e from v a r y i n g degrees of i o n i z e d i m p u r t i y s c a t t e r i n g by complimentary p a i r s of n a t i v e d e f e c t s ( 4 6 ) . S i m i l a r e f f e c t s have been observed w i t h e p l t a x l a l PbTe and Pbo.gSno.2Te. In the evaluation of IV-VI semiconductors on insulating substrates as materials for p-~ junction devices it is important to verify that the layers are truly singlecrystalline, rather than the fine-gralned mosaics with some degree of epitaxial alignment that are commonly designated "slngle-crystal films". Beyond this point, characterization is made difficult by a lack of quantitative relationships between imperfections, such as low-angle grainboundaries, and the properties of p-n junctions, which are more easily related to such parmeters as carrier lifetime and diffusion length. Thus, at present the ultimate test of an epitaxial layer is provided by comparing the properties of a p-n junction in it with those of a similar junction in a bulk crystal of the same semiconductor. Epitaxial growth of Pbo.gSno.2Te on BaF 2 was first reported in 1970 (48). Subsequent work (49) showed that thse and similar layers of PbTe were remarkably free of lowangle grain boundaries. Figure 8 shows an electron micrograph of a replica of a PbTe layer grown on cleaved BaF 2. The most evident features are triangular pits that appear to arise at the intersection of dislocations with the layer surface.* The dislocation densities so estimated can vary quite widely from nearly 108 cm-2 to slgniflcantly less than 106 cm -2. Occasional low angle grain boundaries, as shown in Fig. 8, appear to arise by propagation of boundaries in the substrate. The typical mlsorientations across these isolated boundarles, as estimated from the dislocation spacing, are a few minutes arc and correspond with those found in some The dislocation plts may not occur if the substrate temperature is too low of the growth rate is too large. We have observed well-developed pits with growth at 1-2 tnnhr -I with substrate temperatures in the range 3OO-4OOOC.
6O
H. Holloway and J. N. Walpole
of the substrates.
Fig. 8.
Parlodion replica of epitaxlal PbTe grown on cleaved BaF 2. The region shown is about 40 um wide and the llne of dislocation pits corresponds to a tilt of a few minutes arc.
Table I gives a comparison of some properties of the fluorite structured and the rock-salt structured substrates. Adequate lattice matches with the lead chalcogenides are obtained with both substrate types. The {Iii} cleavage of the fluorite structures leads to an unfamiliar growth habit for the rock-salt structured semiconductors, but this appears not to give problems provided that the vacuum system is sufficiently clean, (Insufficient care sometimes leads to the more familiar {I00} habit as a fibre texture). The fluorite structured substrates do give a superior thermal expansion match and this initially prompted the trial of BaF 2 substrates in an attempt to reduce damage when the layers were taken from a growth temperature near 700 K to an operating temperature near 80 K. However~ the thermal expansion mismatch does not appear to provide an explanation for the relatively grainy films that are obtained with alkali halide substrates. A significant feature may be the hydroscopic nature of the alkali halide surfaces. In contrast, cleavage surfaces of the fluorite homologues are stable in moist air and their low volatilities permit vacuum bake-out at relatively high temperatures (up to at leasts 800 K) to allow clean-up of air-cleaved surfaces. An unusual feature of the IV-VI layers on fluorite structured substates * is that epltaxy occurs with the substrate and deposit in a twinned rather than a parallel relationship, i.e. growth on the {Ill} BaF 2 surface gives a PbTe deposit that is rotated 180 ° about the {iii} axis relative to the substrate (51,52). The result
61
MBE Techniques for IV-VI Optoelectronic Devices may be interpreted in terms of an interface structure of the type
ABsB7
8 CsyA
8
C uByA8
C
PbTe
B~ r
where the stacking symbols (A,B,C for metals; a,8,7 for non-metals) have their usual significance. A detailed analysis shows that alternative interracial structures are less favoured both electrostatically and sterically (53). In general, Hall measurements of IV-VI epitaxial films on fluorite structured substrates give results that differ little from those reported for the best hulk single crystals. As discussed above, the attainment of large Hall mobilities is not an adequate criterion for the growth fo device-qualitymaterial. However, it is of interest that the low-temperature mobilites obtained with BaF 2 and SrF 2 substrates are significantly larger than those that have been reported for layers on alkali halide substrates. A typical example is the result for Pbs (54) that is shown in Fig. 9. For both n- and p-type layers the films give temperaturedependent mobilities that agree well with the best results obtained with bulk PbS. Surpirsingly, the mobility of the n-type layers is not significantly affected by growth from Pb-rich fluxes that appear to give Pb precipitates in the layers. A similar insensitivity to precipitates has been observed with n-type PbSe layers that were grown without a subsidiary Se source. Table I.
Compound
Properties of the lead chalcogenides and some substrates
Lattice Constant (
)
~ (K-I x 106 ) near 300 K
PbS
5.94
20
PbSe
6.12
19
PbTe
6.46
20
Cleavage
P(Torr) at 700 K
NaCI
5.64
39
{I00}
4.5 x 10 -7
NaBr
5.96
42
{i00}
3.7 x 10 -6
NaI
6.46
45
{100}
4.0 x IO-5
KCl
6.29
37
{i00}
2.2 x i0-6
KBr
6.59
38
{I00}
I.i x 10-5
K1
7.05
40
{100}
3.9 x 10-5
CaF2
5.40
19
{111}
~ 6 x 10 -20
SrF 2
5.80
18
{Iii}
~ 1 x 10-20
BaF 2
6.20
18
{111}
~ 3 x 10 -17
This applies to BaF 2 and SrF 2. While lattice match is not particularly critical, it appears that CaF 2 has too large a mismatch for successful epitaxy of the IV-Vl semiconductors. Studies in the Ford Laboratory have failed to give useful layers on CaF 2 substrates and this is in accord with published results for (Pb,Sn) Te on CaF 2 (50).
62
H. Holloway and J. N. Walpole
I05
,
I
III
I
!
,
,
I
..ik . . . . P - Z . z x K ) " cm -3
!
= J
I
"1
(AIIgaier& Sconion)
104 ~
~H
(cm-2V-I~.-i)
1031 300 Fig. 9.
llll i I00 T (OK)
30
I0
Hall mobilities of PbS layers grown on cleaved SrF 2 substrates. The curve shows the largest mobility that has been reported for a bulk crystal of PbS (Ref. 22).
With PbTe and (Pb,Sn)Te layers on BaF2, increases in the mobility have been obtained by annealing the films at relatively low temperatures (typically 3OO-4OO°C for 12-15 hr) (49). The increase was in the more-or-less temperature-independent saturation mobility that was attained at lower temperatures, rather than in the phonon-limited mobility, for which ~H ~ T-a with a ~ 5/2. The annealing results were interpreted in terms of a reduction in the ionized defect scattering that accompanies the pairwise recombination of native donors and acceptors. It was
sation is not invariably present in the layers. Thus, Fig.lO shows Hall data (54) for an as-grown 2.7 ~m thick n-type PbTe layer on BaF 2 that attains a low-temperature mobility of 5 x IO s cm2v-lsec-1. This appears to be the largest mobility that has been reliabily established for a IV-VI epitaxial layer.* +
Lo~ez-Otero (55) has reported low-temperature Hall mobilities of up to 2.5 x IO 6 cm-v-lsec -I in PbTe layers grown on BaY 2 and suggests that this result reflects the superiority of the hot-wall method over other deposition techniques. However, these results must be regarded as questionable because the large mobilities at
MBE Techniques for IV-Vl Optoelectronic Devices
63
13 K were obtained with specimens whose (56) mobilites at 77 K exceeded the commonly observed phonon-llmlted value by about a factor of two. Similar results with bulk single crystals of (Pb,Sn)Te (57) have been interpreted as a consequence o f i n h o m o g e n e i t y i n t h e specimens ( 5 8 ) .
106'U
I
'
I'I'I'
I
'
I
i, •0
_Speclmeh EW 271 105
.
-
i
~H.
J 0 m
B
i I
>
E U
i
:x: 10 4
:k
I B
•
-
o
-
I-
ro
-
_J ID
E v
0
--
T rr
m
.,J _J r
RoJUJo o le
103
•
o*
•
10 2
• o o° • ---
-
I
--
i
b-"
o ~
_~ t4. U. t~
8
•
_1
2xlO 2
I , 400
I 200
I,1,1 I I I00 80 60 40
,
I 20
Ii ~8
=
TEMPERATURE, T (*K) Fig. i0.
4.
Hall coefficient and Hall mobility of a 2.7 ~m-thick n-type PbTe layer on a BaF 2 substrate.
PHOTODIODE PERFORMANCE AS A CRITERION FOR CRYSTAL QUALITY
As pointed out earlier, the ultimate test of a semlconductlng material arises when one attempts to make devices with it. For discussion of the results that have been obtained with thin-film photodiodes it is convenient to consider the detectivity (D*), which is a standard figure of merit that is defined as a signal-to-noise ratio that has been noralized to make it independent of the incident power, the noise bandwidth, and the detector area. With, a w e l l - b e h a v e d p h o t o d i o d e a t z e r o b i a s i n t h e a b s e n c e o f s i g n i f i c a n t b l ack b o d y r a d i a t i o n form t h e background, t h e s i g n a l c u r r e n t depends upon t h e quantum e f f i c i e n c y (n) and t h e n o i s e c u r r e n t depends upon t h e J o hn so n n o i s e o f t h e j u n c t i o n r e s i s t a n c e . Under t h e s e c o n d i t i o n s we have
J
64
H. Holloway and J. N. Walpole
D* (Johnson) = n Ey
ReA ~ 4K2
where Ey is the photon energy. In most cases, the quantum efficiency of the photodiode is limited only by reflection losses and has a value in the range 0.4 - 1.0. The D* then d~pends on the zero-blas resistance area product (RoA) of the junction. This is a sensitive measure of the quality of the material because the junction resistance is inversely proportional to the saturation current, which may be substantially increased by defects that decrease the lifetime in the film. With large enough values of the resistance-area product, the Johnson noise current becomes smaller than the shot noise due to fluctuations in the rate of arrival of photons from the balckbody background. Under these conditions the D* becomes background limited with a value
D* (background) = I
~
n
'2'QB
where QB is the background photon flux in the energy range for which the photodiode is sensitive. The background-llmited D* depends upon the field of view (FOV) of the detector and also quite strongly upon the cut-off wavelength since this influences the value of QB. Table 2 shows some representative background-limited detectivities together with the reslstance-area products that must be exceeded for background limited operatuion with 180o FOV. Thus, the demonstration of background-limited D* at 180 ° FOV and at some convenient operating temperature (e.g. 77 K) is a convenient criterion for the attainment of material that is good enough to be useful. Further deCisions about quality may be based upon the extent to which the D* increases when the FOV is decreased (i.e. the extent to which RoA exceeds the value needed for background-limited operation at 180 ° FOV). Table 2.
Background-Limited Detectivities for Infrared Photodiodes %
Cut-off wavelength
D* (Background)
(~m)
(cmHz~W-I)
3 4
Equivalent RoA
(ohm em2)
1.2 x 1012
3900
2.6 x lO ll
115 15
5
1.2 x lO II
6
7.5 x I0 lO
4.1
8
4.6 x I0 I0
0.88
i0
3.8 x 1010
0.38
12
3.5 x I0 I0
0.22
t These detectivities are calculated for 180 ° FOV a t a wavelength near the cut-off with a typical reflection-loss-linLited quantum efficiency of 0.5. The corresponding values of RoA are those which, at 80 K, give equal contributions from the Johnson noise and ~he background noise. (This would give detectivities that are reduced by a factor of / ~ f r o m the background-limited values.)
~BE Techniques for IV-Vl Optoelectronic Devices 5.
THIN-FILM VI-VI SEMICONDUCTOR PHOTODIODES
While the choice of vacuum deposition technique for thin-film IV-VI semiconductors may be regarded as a matter of taste, the choice of insulating substrates for p-n junction applications is not. With one exception (discussed below) the very extensive body of work with alkali halide substrates has failed to yield p-n junction devices. In contrast, the fluorite structured substrate BaF 2 has given p-n junction devices whose performance is competitive with that of the best bulk crystal devices. Moreover, following the original work with PbTe on BaF 2 at the Ford Laboratory, similar devices have been obtained with a range of IV-VI semiconductors by the Ford group and by several other groups of workers who have used a variety of vacuum deposition techniques. Thus, the key feature is the choice of suhstrate and the successful development of a p-n junction technology may be attributed to the improvement in the crystal perfection that is described above. The sole example of significant performance from p-n junction devices made from IV-VI layers grown on alkali halide substrates appears to be the work of Schoolar who worked with PbS on NaCI. For junction formation it was necessary to dissolve away the substrate and use the PbS surface that had been in contact with the NaCI. PbS juncitons that were made by growing p-type PbS on n-type.epitaxial PbS layers gave quantum efficiencies around O.I and D*(4 ~m) = 109 cmHzIW -I when operated at 77 K (59). Subsequent work (60) with semltranspartent In metal barriers gave D * (3.8 Bm) ffi 1.5 x 1011 cmHzIW -I, which is only a factor of two below the background limit (assuming 180 ° FOV and a typical reflection-loss-limited quantum efficiency of 0.5). With the observation of high-quallty layers of PbTe and (Pb,Sn)Te on BaF 2 substrates (49) it became obvlouse to try farlcation of p-n junction devices with these materials The first high-performance thln-film IV-VI photodlodes and the first use of the metal barrier technique for IV-VI infrared photodlodes were reported by the Ford group in 1971 (61). Barrier layers of small-work-functlon metals, such as Pb, had been used previously by Nill et al. (62) to make bulk crystal lasers by electrostatically inverting the surfaces of p-type PbTe and (Pb,Sn)Te. Infrared response had been observed from the edges of these devices (63), but optical absorption in the barrier layer had precluded application as photodiodes.* The energy bands of a thin-film PbTe metal barrier photodiode, following the analysis by Walpole and Nill (65), are shown in Fig. II.** The BaF 2 substate is conveniently transparent for wavelengths up to about 12 ~m. This permits illumination of the junction via the substrate. Taking typical optical absorption lengths of the order of i um and typical minority carrier diffusion lengths of the order of I0 ~m, it is evident that a semiconductor layer that is a few ~m thick can give both good optical absorption and efficient collection of photogeneratedminority carriers.
The o r i g i n a l c o n f i g u r a t i o n f o r t h i n - f i l m PbTe p h o t o d i o d e s (61) was t h e simple c r o s s e d - s t r i p e a r r a n g e m e n t t h a t i s shown i n F i g . 12. When o p e r a t e d a t 77 K and 180 ° FOV, t h e s e d e v i c e s gave b a c k g r o u n d - l i m i t e d D * ' s . With r e d u c t i o n of t h e For t h e r e was an increase by a factor of four to a Johnson noise limit of D*(5 ~m) = 6 x 1011 Subsequent work (64)has shown that semitransparent In barriers, similar to those reported by Schoolar, may be used to make hlgh-performance bulk crystal Pbo. 8 Sno.2Te photodlodes.
** I n most cases we l a c k an u n e q u i v o c a l d e m o n s t r a t i o n t h a t t h e t h l n - f i l m p - ~ j u n c t i o n ~ a r e formed by e l e c t r o s t a t i c i n v e r s i o n , r a t h e r t h a n by s h a l l o w d i f f u s i o n of Pb i n t o , or c h a l c o g e n o u t o f , the s e m i c o n d u c t o r . D e s p i t e t h i s u n c e r t a i n t y the j u n c t i o n forming t e c h n i q u e i s of c o n s l d e r a b l e p r a c t i c a l i m p o r t a n c e .
B. Holloway and J. N. Walpole
o°il pb i
ENERGY ('v-az1[
b
ocm3
c8 vs
iF_ ........................
-0.4f
-0.6
-0.8
0 Fig. ii.
0.2 I
DEPTH (p.m)
0.4
Calculated band bending due to a Pb layer on p-type PbTe (p ~ 1017 cm-3).
cmHz½W -I. Subsequent work (54) has given Johnson noise limits as large as 1013 c ~ z ~ W -I for such thin-film PbTe devices at 77 K. Minor modifications of the crossed-stripe geometry gave the first demonstration of a thln-film injection leaser on an insulating substrate (66) (discussed elsewhere in this review) and also a junction field-effect transistor (67,68). Further studies showed that the performance of the PbTe photodiodes did not depend upon the technique that was used for junciton formation. Essentially similar detectivlties were obtained with p-n junctions that were made by proton bombardment (69) and by Sb ion implantation (70) of p-type PbTe layers. * Studies of the stability of the Pb-barrier devices have shown that their performance is retained after repeated cycling to cryogenic temperatures and after more than sufficient baking to outgas vacuum enclosures (~ I0 hr at 150°C). Further development from the original thin-film PbTe photodiodes has followed two main lines. First, the use of IV-VI alloys for specific applications in the 3-5 Bm atmospheric window and for reduction of the energy gap to permit operation in the 8-12 ~m atmospheric window. Secondly, refinement of the rather crude crossed-stripe configuaration to permit the delineation of and connection to arrays of small closel~ * To date only limited success has been reported with conventional diffusion techniques and with attempts to make p-n junctions by changing the conductivity type of t~e deposits during film growth. Callender (76) reported D*(IO Bm) = 4 x 109 C,LHz=W"I at 77 K for Pbo.82Sno.18Te devices that were made with a mesa technique applied to 15 ~m-thick layers on BaF 2. Lopez-Otero et a~. (72) reported D* = 1.4 x IO I0 c ~ z ½ W -I at 77 K for grown-ln p-'n junctions in epitaxial PbTe on BaF 2. In both cases the D* was an order of magnitude less than the background-limited value that has been exceeded by metal-battler devices.
MBE Techniques
for IV-Vl Optoelectronic Devices
67
spaced detector elements.
IR DETECTOR
FET
Pb
Pb
LASER Fig. 12.
C r o s s e d - s t r i p e c o n f i g u r a t i o n s f o r Pb b a r r i e r PbTe devices.
The major advance in thin-film 3-5 Bm detectors has been the development of devices whose operating temperature has been increased from 77 K to the intermediate temperature range, 170-2OO K, that is suitable for thermoelectric cooling. This permits application in lightweight hand-carrled thermal imaging systems. These increased operating temperatures greatly increase the requirements for device quality because of the exponential decrease of the junction resistance with temperature which reduces the Johnson-nolse-limited D*. Figure 13 shows the temperature-dependent spectralD* of a thin-film PbTe photodiode (73,74) operated with 180° FOV. The curves show several interesting features. First, there is a sequence of pronounced maxima and minima that arise from interference modulation of the quantum efficiency about the thlck-film limit of 0.6. This effect may be used to obtain quantum efflclencles as large as 0.9 for selected wavelengths. Secondly, as the oerpatlng temperature is increased from 80 K, the D* first increases and then decreases. This effect is due to a change in the dominant noise mechanism. At low temperatures the juncition resistance is sufficiently large that the Johnson noise is much smaller than the background noise. With increased temperature the increase in the energy gap shifts the cut-off to smaller wavelengths and reduces the effective background photon flux thereby increasing the background-limlted D*. At large enough temperatures the junction resistance decreases to a point where the device becomes Johnson noise limited. Further increase in temperature then reveals the temperature-dependance of this Johnsonnolse-llmitedD*.
68
H. Holloway and J. N. Walpole
1012
,
,
,
,
,
,
,
,
,
h
84 K
_ITOK.'~ ~ /
•r
-
°
"~.-
190 K~ , ~ . ~
?
~..
EW 3 3 5 - 4d
,o° Fig. 13.
2
3
4 X (Fm)
!I!l 5
6
Temperature dependence of the spectral detectivity of a thin-film PbTe photodiode.
The practically significant feature of Fig. 13 is the attainment of useful D*'s at temperatures that are suitable for thermoelectric cooling t For many 3-5 ~m applications the response of PbTe extends to long enough wavelengths. However, at 170 K the cut-off at about 4.8 ~m does not permit full use of the 3.5 ~m atmospheric window. This led to consideration of pseudobinary alloys of the Pb chalcogenides that are shown in Table 3. Table 3.
Optical Absorption Edges of Lead Chalcogenides
Material
Optical absorption edge (~m) at 77 K
at 170 K
PbS
3.7
3.3
PbSe
6.9
5.6
PbTe
5.7
4.8
t These results have been confirmed in part by McMahon (75).
MBE Techniques for IV-Vl Optoelectronic Devices
69
Following the demonstration of high performance t h l n - f i l m Pb b a r r i e r PbSe photodiodes (76) the film growth techniques were extended to PbSeo.8Teo. 2 (77) with an optical absorption edge near 5.4 ~m at 170 K. This material was used to make the devices whose temperature-dependent spectral D* are shown in Fig. 14 (73,78). Basically, these results are similar to those obtained with PbTe devices, but with the response extended to cover the whole of the 3-5 ~m spectral region. Subsequent work (78) has yielded PbSeo.sTeo. 2 devices with D*(5 ~m) = 1 X i0 II cmHzl/2W-1 when operated at 170 K with 18OO FOV. Under these conditions there are approximately equal contributions from the Johnson noise and the background photon shot noise. This performance appears to exceed that reported for any 3-5 ~m photodiode when operated at intermediate temperature. It is perhaps a measure of the maturity of thln-film photodiode technology that the Pb(Se,Te) pseudoblnary alloy system was not drlved from a bulk crystal application, but was first exploited with the thinfilm devices.
i0tt
170 K
']¢
A%~/I1,
.r
190 K
E S
Q i010 _
210K m~~l~
[
Spec'l'rol de'tectivil"y of PbSeo. s Teo.z ( t= 1.5/J,m)
I
I
2
I
I
3
I
I
I
4
6
I
I
I
8 I0
X(~m) Fig. 14.
Temperature dependence of the spectral dtectivlty of a thln-film PbSeo.sTeo.2 photodlode.
A further appllcatlon of the thln-film technology has been devised by Schoolar et ul. (13) who have developed high performance Pb(S,Se) photodlodes. In this work both sides of the BaF 2 substrates were coated with Pb(S,Se). One of the layers was used to make Pb barrier photodlodes. The layer on the other side of the substrate had a composition that was adjusted to give a sllghtly smaller energy gap and served as a long-wavelength pass filter. By careful adjustment of the compositions the response was reduced to a narrow range (~ 0.2 ~m wide at half maximum) that was defined by the cut-on of the filter and the cut-off of the photodiode.
70
H. Holloway and J. N. Walpole
Another major l l n e of development has been the e x t e n s i o n of t h l n - f i l m photodiode response to the 8-12 ~m atmospheric window by development of techniques f o r growth of Fbl_xSnxTe (x % 0.2) (48) and Pbl_xSnxSe (x ~ 0.07) on BaF 2 (79). Here, use of the (Pb,Sn)Te a l l o y system had many p r e c e d e n t s i n bulk c r y s t a l t e c h n o l o g y , but l i t t l e success had been r e p o r t e d with bulk c r y s t a l (Pb,Sn)Se photodiodes d e s p i t e e a r l y use of t h i s m a t e r i a l to e s t a b l i s h the e x i s t e n c e of the b a n d - c r o s s i n g pbenomenon (29,30). Figure 15 shows the spectra D*'s of two thin-film Pb-barrler (Pb,Sn)Se photodiodes for the 8-12 ~m spectral region (81). At 80 K with 180 ° FOV, such devices are background-llmited with reslstance-erea products up to 2ohm cm 2. With reduction of the FOV, a Johnson noise limit of D*(IO ~m) = 8 x 1010 emHz½W -I has been obtained. These results compare well with peak D*'s of 1 x iO II cmHz}W -I that have been obtained with bulk crystal (Pb,Sn)Te photodiodes at a reduced FOC (82). Approximately background-limlted D*'s near IO ~m have also been obtained with (Pb,Sn)Te In barrier devices (71,83) at 80 K with 180 ° FOV. Schoolar and Jenson (14) have also used (Pb,Sn)Se layers on BaF 2 substrates to make narrow response photodiodes by using the same two-layer technique that was described for Pb(S,Se).
I°"i'
I , I , I , I , I ,
I
,0 9
, 2
Fig. 15.
I, 4
I
I
6
8 X (~m)
I,
i I0
,
, 12
14
Spectral detectivities of (Pb,Sn)Se photodiodes at 80 K.
Development of a processing technology for thln-film IV-Vl photodiode arrays has been accomplished by Asch and co-workers (84) who have delineated the junction areas with windows in an insulating layer of vacuum-deposited BaF 2. Thus, the basic
MBE Techniques for IV-Vl Optoelectronic Devices
71
processes for thin-film array fabrication consist of four successive vacuum deposit ions. These give the semiconductor, the Pb ohmic contact, the BaF 2 insulator, and the Pb barrier, respectively. Mounting of the arrays for illumlnation through the BaF 2 substrates was done with a flip-chip technique that used a composite Cu foil/ glass/Cu header for thermal expansion compatibility. Some details of a thln-film array are shown in Fig. 16.
-'--] BoF Substrate ~ Evaporated8oF~
I
,I
,
Sputtered
Platinum
E~ :voporated Lead Semiconductor
Fig. 16.
Structure of a IV-VI semiconductor thln-film photodlode array.
72
H. Holloway and J. N. Walpole
6.
UNCONVENTIONAL THIN-FILM IV-Vl PHOTODIODES
The preceding section has emphasized the properties of thin-film IV-VI photodiodes whose performance may be compared that of bulk crystal devices to provide an assessment of the perfection of the epitaxial layers. For completeness we now review briefly some more recent developments in which the unique properties of thin films are exploited. One very characteristic property of the thin-film photodiodes is the interference modulation of the quantum efficiency whose influence is evident in Figs. 13-15. A detailed analysis (85) shows that with layers up to about 5 pm thick the quantum efficiency of IV-VI thin-film photodiodes should be limited only by reflection and transmission losses. The good agreement between the observed quantum efficiency of a PbTe metal barrier device and the calculated reflection-loss limit (RLL) is shown in Fig. 17. The result confirms the attainmnet of bulk-llke minority carrier diffusion lengths, and hence lifetimes, in the thin-filmmaterial. Exploitation of the interference effects may take two forms. First, by appropriate choice of the layer thickness the RLL at specific wavelengths may be increased to about 0.9 compared with about 0.5 for an uncoated conventional PbTe photodiode. This eliminates the need for antireflectlon coatings to optimize the detectivlty. Secondly by making use of combinations of IV-VI thin films with dielectric layers and metal reflectors the response of the photodiode may be made to resemble that of a combination of a conventional photodiode with an interference filter (85). Such devices may find application for low-cost evaluation of spectral signatures.
I
I
I
/,.,
I
/ li
#'
4> I
°
2
[
3
i
4
l
5
l
6
7
~.(Fm) Fig. 17.
Comparison of the measured and calculated quantum efficlencles of a thin-film PbTe photodiode at 80 K.
One significant disadvantage of conventional IV-VI semiconductor photodiodes is the
MBE Techniques for IV-Vl Optoelectronic Devices
73
i s the l a r g e j u n c t i o n c a p a c i t a n c e (of the o r d e r of 1 ~Fcm- 2 ) t h a t a r i s e s from t h e l a r g e d i e l e c t r i c c o n s t a n t s of t h e s e m a t e r i a l s . One approach t o c a p a c i t a n c e r e d u c t i o n i s r e d u c t i o n of the dopant c o n c e n t r a t i o n and t h i s has been d e m o n s t r a t e d w i t h b u l k c r y s t a l s of (PbjSn)Te by Andrews e t u Z . ( 1 7 ) . However~ s i n c e the c a p a c i t a n c e of a b r u p t j u n c t i o n o n l y d e c r e a s e s w i t h the s q u a r e r o o t of t h e dopant c o n c e n t r a t i o n , t h i s approach r e q u i r e s the growth of m a t e r i a l w i t h I015cm-3 c a r r i e r s o r l e s s . The t h i n - f i l m d e v i c e s o f f e r an a l t e r n a t i v e approach t h a t i s based on t h e p r o x i m i t y of the depletion region to the insulating substrate (86). The concept is shown in Fig. 18. With a concentional p--n junction a change in the bias causes a change in the width of the depletion region and thereby a change in the charge that is stored in the depletion region. This gives rise to a dynamic capacitance dQ/dF. However, with the thin-film device the p-n junction may be arranged to have the depletion region extend through to the insulating substrate. Under these conditions the depletion region can only accept or give up charge around its periphery. This leads to a reduction in the capacitance by the factor (peripher X layer thickness)/ (junction area). The capacitance reductlonmay be achieved by using layers that are thin enough (~ 0.6 Bm) that become pinched off when the depletion region is widened by back bias. For the latter mode it becomes necessary to reduce the I/f noise that is usually associated with IV-VI photodiodes that are operated in back bias (82). This has been achieved by careful cleaning of the PbTe surface.
ii I
N
I
DEPLETION
SUBSTRATE (a)
SUBSTRATE (b) F i g . 18.
Conceptual arrangement of the pinched-off photodiode. With increasing back bias, the depletion region edge moves to the position shown as a broken line in (a) for a conventional photodlode and in (b) for a plnched-off photodiode.
The t y p i c a l b l a s - d e p e n d e n t performance of a t h i n - f i l m P b T e p i n c h e d o f f p h o t o diode a t 80 K and 180 ° FOV i s shown i n F i g . 19. At zero b i a s t h e diode w i t h a r e a 6 x 10-~ cm2 has a c a p a c l t e n c e of 700 pF. With back b i a s g r e a t e r t h a n a b o u t O.15 v t h e c a p a c i t a n c e d e c r e a s e s t o a c o n s t a n t v a l u e of 70 pF t h a t a r i s e s from t h e c o n t a c t pads r a t h e r t h a n the j u n c t i o n . The 500 K blackbody c u r r e n t r e s p o n s i v i t y ( R I ) , which i s p r o p o r t i o n a l to the quantum e f f l c i e n c y ~ i s i n d e p e n d e n t of t h e b i a s ash t h e n o i s e remains c l o s e to t h a t c a l c u l a t e d f o r f l u c t u a t i o n s i n t h e background f o r back b i a s up to 0.35 v , a f t e r which 1 / f n o l s e becomes s i g n i f i c a n t . Thus, t h e D* r e m a i n s a t the b a c k g r o u n d - l i m i t e d v a l u e f o r b i a s e s t h a t p e r m i t an o r d e r o f magnitude r e d u c t i o n i n the d e t e c t o r c a p a c i t a n c e . L a r g e r a r e a p h o t o d i o d e s have
74
H. Holloway and J. N. Walpole
shown up to two o r d e r s of magnitude r e d u c t i o n i n c a p a c i t a n c e . I t i s of i n t e r e s t t h a t no b u l k c r y s t a l d e v i c e s e x i s t w i t h p r o p e r t i e s t h a t are comparable to the t h i n film prinched-off photodoides.
, o '2
I
1
Rz(SOOK)
I
I
m
]
D*(5.4pm)
N
_~
-I-
E
.~- lOll E ::t.
o
C
H
1
C~
I I
,o,o ,ol 0
Fig. 19.
Z nuise (calculated) I
I I00
I
I
I
200 5 0 0 4 0 0 BACKBIAS (mV)
i
_~ 10-2
500
Bias-dependent performance of an 0.6 vm-thick PbTe photodlode at 80 K and 180 ° FOV. The diode area is 6 x i0- g c m 2. The noise was measured at I kHz with I0 Hz bandwidth and the broken line shows the background noise current that was calculated from the d.c. background current.
An alternative approach to capacitance reduction using thin-film devices (87,89) is shown in Fig. 20. Here the n-region of a conventional photodiode is replaced by a matrix of small circular n-regions that act as collectors for photogenerated minority carriers from the intervening p-region. The insulating substrate acts as a potential barrier that confines the photogenerated carriers to the vicinity of the collectors, thereby reducing losses by recombination. Efficient collection is obtained with a collector spacing of up to two diffusion lengths (20-30 ~m for PbTe). With collector deameters in the range 2-5 Bm the reduction in junction area permits more than an order of magnitude reduction in junction capacitance. Figure 21 shows laser scans of such a lateral-collectlon photodiode (LCP) at different operating temperatures. This shows clearly the increase in collection efficiency as the device is cooled and the diffusion length increases. Under some conditions
MBE Techniques f o r IV-VI O p t o e l e c t r o n i c Devices
75
the saturation current of IV-VI diodes is generated withln the depletion region and in this case the junction resistance of the LCP is greater than that of a conventional device with a corresponding increase in the Johnson-noise-limited D*. This effect has been used to obtain a 35 K increase in the operating temperature
o f PbTe p h o t o d l o d e s . The o n l y c o m p a r a b l e b u l k c r y s t a l work a p p e a r s t o be by N o r e i k a e# aZ. (90) who u s e d c o l l e c t o r s w i t h a s t r i p e g e o m e t r y on b u l k c r y s t a l s o f (Pb,Sn)Te t o o b t a i n c a p a c i t a n c e r e d u c t i o n by a f a c t o r o f t h r e e .
t
n
J
t
n
J
~,e-h + 4 p - type TV'-vr
Ba F2 Substrate
Fig. 20.
Schematic arrangement of a lateral-collection photodiode.
7.
OPTICAL WAVEGUIDES
An optical waveguide structure consisting of a 6 ~m MBE layer of Pbo.92Sn O 08Te grown on a {100} oriented PbTe substrate followed by a 0.5 pm PbTe (cladding) layer as shown in Fig. 22 has been demonstrated (91). Striped guides were formed by growing over a SiO 2 mask which was later lifted off to remove the polycrystalline overgrowth except in the stripes where epitaxial growth occurred. The optical loss coefficient measured in these guides at 10.6 ~m wavelength for light polarized parallel to the plane of the llght-guldlng PbSnTe layer (Te polarization) was 7.8 cm-I at room temperature. Since room temperature free carrier absorption has been well characterized in PbTe (92) and is similar in Pbo.92Sno.o8Te (93), the loss coefficient of 7.8 cm-1 implies an upper limit of about 4 x 1016 cm -3 for the free carrier density of the layer. At 77 K the loss coefficient was too small to measure but was determined to be ~ 1.5 cm-1, the smallest optical loss coefficient measured to date in either bulk or epltaxlal PbSnTematerlal. For the orthogonal polarization (TM), losses were somewhat higher but the increased loss could be attributed to absorption in the metalllzatlon used on the surface of the 0.5 ~m PbTe cladding layer as part of the mountlng procedure for the waveguldes. The transmission scans of a CO 2 laser along the path indicated in Fig. 22 are shown in Fig. 23. The wavegulde transm~sslon is strongly reduced by coupling losses into the narrow guide, compared to transnLisslon through the substrate. The wavegulde results suggest the feasibility of fabricating by MBE monolithic integrated optical circuits incorporating active elements such as lasers, laser az~pllflersp and detectors. The control of wavegulde thickness, which can be tapered for coupling purposes, using shadowing techniques during growth, % is one of the advantages MBE offers for this purpose, in addition to the low as-grown % Tapered layers similar to those reported in GaAs MBE layers (94) can easily be grown in a similar manner in PbSnTe (95).
76
H. Holloway and J. N. Walpole
carrier densities necessary for low loss.
270K
170K
85K
Fig. 21.
Laser scans of a 320 ~m-square PbTe lateral- collection photodiode at 270 K, 170 K, and 85 K. The 5 ~m-dlameter collectors are on 20 ~m centers. The vertical displacement is proportional to the current response and the instrumental resolution is about I0 ~m.
Fig. 22.
Pbl_xSnTe Epilayer Scan Path
Typical Input Face
I 6.5p.m
Schematic of double heterostructure optical waveguides showing the path of scan by a lO.6-~m laser.
PbTe Epilayer
O
5.
t-'° O
I-4
O
= ¢,
O =r'
78
H. Holloway and J. N. Walpole
4.4 4.0
36 Z
0 3.2 o'J 2.8 Z
2.4
I--
2.0
I-. Z
1.6
~" W n
1.2 0.8
0.4 ','1 0,2
I 0.4
,
'
I 0,6
I
I 0.8
I
1.O
12
POSITION (ram) Fig. 23.
Transmission scans of TE and TMpolarized 10.6 ~m laser beams along the path indicated in Fig. 22 for a 4-~m-long waveguide at 77 K. Loss due to coupling is estimated at 95.2 and 92.3% for the TE and TMwaveguide modes, respectively. The guiding layer is located at the 0.2-mmposition.
8.
INJECTION LASERS
Lead-tin chalcogenide injection lasers (laser diodes) have become important research tools for tunable laser spectroscopy (96) in the infrared at wavelengths between about 4 ~m and 30 ~m (see Fig. 6). Their use as pollution monitoring and gas detection devices (97) has been demonstrated includlng their potential for remote heterodyne detection (98-100). The motivation for epltaxlal growth of these devices has been to achieve cw operation at liquid nitrogen temperature and above, rather than at the low temperature (~ 30 K) previously achieved with diffused p-n junctions in bulk-grown material.* An additlonal bonus of higher temperature operation is that, as the energy gap changes with temperature, a single device can be used at different temperatures to cover a large spectral region (400-500 cm -I) (101). Fine tuning of a single mode can be achieved at a given temperature by several means but most easily by varying diode current. The internal change of temperature with current changes the refractive index which shifts the requencies of the longitudinal cavity modes. The continuous tuning range of a mode is typically 1-2 cm-I In addition to the MBE and hot-wall lasers to be discussed, progress has also been made using llquld-phase epitaxy as mentioned, including the first cw operation of a lead-salt laser above 771((19).
MBE Techniques for IV-VI Optoelectronic Devices
79
The u s e o f h e t e r o s t r u c t u r e s i n I I I - V d i o d e l a s e r s t o r e d u c e l a s e r t h r e s h o l d and i n c r e a s e t h e o p e r a t i n g t e m p e r a t u r e s u g g e s t e d t h a t a s i m i l a r approach would be s u c c e s s f u l i n t h e IV-VI m a t e r i a l s ( 1 0 2 ) .
8.1.
Herterostructure lasers
F i g u r e 24 shows a t y p i c a l Pbl_xSnxTe d o u b l e h e t e r o s t r u c t u r e (DH) l a s e r c o n f i g u r a t i o n i n whlch an e p i t a x l a l l a y e r ( t h e a c t i v e r e g i o n ) l i e s b et w een t h e s u b s t r a t e and a n o t h e r l a y e r b o t h o f which have s m a l l e r Sn c o m p o s i t i o n s , l a r g e r b a n d - g a p s , and s m a l l e r r e f r a c t i v e i n d e x e s than t h e a c t i v e r e g i o n . If a p-n junction is also located within or at an edge of the active region, minority carriers injected at forward bias are confined within this region by the potential barriers resulting from the larger energy gaps at the edges. This carrier confinement is one of the features of DH lasers which, in principle, reduces ~ e threshold current density required for laser action provided that high recombination does not occur at the interfaces. The refractive index variation indicated in the figure forms a dielectric slab waveguide to guide (confine) the optical radiation within the active region where the optical gain is located. I n a s i n g l e h e r t o s t r u c t u r e (SH) l a s e r , o p t i c a l and c a r r i e r c o n f i n e m e n t a r e p r o v i d e d by a change i n a l l o y c o m p o s i t i o n on o n l y one s i d e of t h e a c t i v e r e g i o n . On t h e o t h e r s i d e o p t l c a l and c a r r i e r c o n f i n e m e n t a r e p r o v i d e d by a p - n j u n c t i o n o r a g r a d e d doping p r o f i l e as i n h o m o s t r u c t u r e l a s e r s . As w i l l be s e e n , i n t h e IV-VI m a t e r i a l s s t r o n g c o n f i n e m e n t can be o b t a i n e d i n t h i s way.
Electrical and
Thermal Contact
/
V / l / ~ / ~ ~ ~ / ~ ~ ~ ~ 1
]
\
,~:::~::~:::~.::~:~..~,~:.-~.::~.-..:~....::..:~:~..=~.:~,
( Not to scale)
n-Pb0 78 Sn0 22 Te =,,,,,,,,,,
..........
,,,n~,,,,,,,,
..........
~':'::'~;':~'?~";:"L'~"~=t"":':':':::~':T':~':":":~'~:';~L"
II I I mW~mHmHHmRDHmmmumm~mmHll
/ Electrical Contact F i g . 24.
S c h e m a t ic d l a g r a m o f a PbSnTe l a s e r grown w i t h a MgF2 growth mask. The v a r i a t i o n o f r e f r a c t i v e i n d e x n and e n e r g y gap Eg a r e shown a t t h e l e f t .
The f i r s t IV-VI h e t e r o s t r u c t u r e l a s e r was an u n u su al v a r i a t i o n from t h e c o n v e n t i o n a l s t r u c t u r e s d i s c u s s e d a b o v e . I t c o n s i s t e d o f a p-PbTe f i l m grown on a BaF 2 s u b s t r a t e
80
H. Holloway and J. N° Walpole
(66) with a Schottky barrier junction formed by an evaporated Pb strip. Ohmic contact to the film was obtained with sputtered Pt as indicated in Fig. 25. The lateral current flow required by the geometry of this structure leads to lateral voltage drops increasing with the sheet resistance of the film. Hence it is difficult to obtain uniform injection under the strip of Pb. In the devices reported, injection occurred along the edge of the Pb strip only and the active volume of the laser was concentrated within a diffusion length on each side of the edge. The BaF 2 substrate provided strong opitcal and carrier confinement on one side of the film while the Pb Schottky barrier or the interface with air adjacent to it provided optical confinement on the other. There was no significant lateral confinement within the film however. Carrier confinement at the Pb barrier would be characterized by the injection efficieny of the barrier, i.e. the ratio of current flow arising from injected electrons to the toal current flow. The injection efficieny of this type of junction appears to be good at low temperatures (62). Nevertheless, a significant Richardson's current flow could occur in this structure (63), and may prevent high temperature operation.
-H
8
Spectrometer Resolution
A
Pb
PbTe I
==6
-
o4
h.........~
z
o
,,idlL ~ A ~
A
6.0
5.8
5.6
5.4
5.2
5.0
4.8
(microns) Fig. 25
Spontaneous emission spectrum of a thin-film PbTe laser fabricated as shown in the insert. The theroretical fit (dashed curve) is for k-conserving parabolicband transitions.
Because of the transparent substrate, it was possible to collect a large fraction of the isotropically emitted spontaneous light from these devices so that the spontaneous emission spectrum could be measured. Few well-resolved data of this sort exist otherwise for IV-VI materials. Spectra obtained at 77 K showed a good fit to the theoretical spectrum for direct band-to-band transitions between parabolic bands as shown in Fig. 25. Although pulsed laser operation was observed in these devices, cw operation was precluded by the lack of adequate heat-slnking. Further improvements in these devices were reported later (IO3). Post-growth annealing of the films to increase
MBE Techniques for IV-Vl Optoelectronic Devices
81
carrier density and thus decrease sheet resistance and an etching technique used to form smooth parallel mirrors resulted in lower thresholds and cw operation was achieved near 10 K. Threshold current densities weree estimated to be about I KA/cm 2. The ability to etch smooth varallelmirrors in films ~rown on BaF 2 substrates may be a si2nificant advantage in device fabrication since mirror quality can be a serious problem in PbSnTe lasers because of the poor cleavage properties of the material (104). In spite of their promise, further development of IV-Vl lasers grown on BaF2 substrates has not been pursed to date. Subsequent efforts have made use of the IV-VI materials as substrates where the problem of lateral current flow is avoided and where heatsinking of the active layer is more straight forward. The first device of this type was an SH laser in which an n-type layer 3 ~m in thickness was grown on a p-type Pbo.88Sno.12Te substrate (105). The laser emission wavelength corresponded to the smaller band-gap PbSnTe substrate rather than the PbTe film as expected. The depth of the p-~z junction in this device was not established but most likely some diffusion during growth of an n-skin into the substrate on the order of I B m deep occurred. A pyrolytically deposited SiO 2 growth mask patterned with 50 ~m-wide striped openingswas used to obtain a stripe geometry. Polycrystalline PbTe grew over the oxide with crystalline growth in the stripes. The oxide did not adhere well to the substrate after growth so that it was not possible to mount the device with the PbTe layer bonded to the heat sink without shorting out the stripe. Instead, the Pbl_xSnxTe bulk side of the device was bonded to the heat sink. Despite the poor heatsinking, cw operation was obtained up to 65 K. In Fig. 26, the pulsed threshold current density as a function of temperature for a SH laser is compared with data from two low-threshold stripe-geometry homestructure lasers made by diffusion into bulk material. Singificant threshold reduction for the SH laser can be seen. At 77 K the threshold of 780 A/cm 2 is lower by a factor of 3 than any PbSnTe homostructure devices at that temperature reported to date. This result strongly suggests that high recombination at the film-substrate interface does not occur. The improvement of the SH laser over the diffused homostructures is due to the carrier confinement provided by the PbTe layer which prevents diffusion of injected holes to the contact where they would recombine. A previously unobserved phenomenon was seen in the I-V characteristics of these SH lasers. It has subsequently been observed in a number of injection lasers (i06-10~ including MBE DH lasers (101). At threshold the I-V characteristic abruptly changes from an exponential to a linear dependence. The change in slope or differential resistance (dV/d/) of an SH laser is shown versus d.c. bias current I at several temperatures in Fig. 27, but the effect is also obvious in the I-V characteristic itself as shown in Fig. 28 for an MBE DH laser at 77 K. The onset of strong stimulated emission at threshold reduces minority carrier liftime so that no further voltage (population inversion) can be applied to the junction and current flow is subsequently limited only by series resistance. As can be seen in Fig. 27 at 30 K and 47 K, there is sometimes a small drop in junction voltage at threshold which may be due to saturable absorption (i.e. less gain is required to maintain laslng action than to reach threshold due to the saturation of absorption processes). The junction voltage saturation implles virtually I00% incremental internal quantum efficiency above threshold (105, 109). It is a consequence of strong carrier confinement withing the optical field (in the lateral direction, by the use of stripe geometry, as well as in the direction transverse to the junction). It should be emphasized that the incremental internal efficiency above threshold is not related to the radiative quantum efflcency below threshold. Sleger et u~. (II0) have reported PbSSe SH lasers grown by the hot-wall technique in which the small band-gap active region is the epitaxial film rather than the substrate.
82
H. Holloway and J. N. Walpole
These d e v i c e s were o n l y o p e r a t e d p u l s e d a t 12 K, b u t had t h r e s h o l d c u r r e n t d e n s i t i e s of a b o u t 200 A/cm2 comparable t o h o m o s t r u c t u r e d e v i c e s , a g a i n i m p l y i n g no s e r i o u s n o n r a d i a t i v e recombination at the h e t e r o - i n t e r f a c e nor at m i s f i t d i s l o c a t i o n s gener a t e d by t h e 0.2% l a t t i c e mismatch. The r e l a t i v e l y low t h r e s h o l d a l s o i n d i c a t e s t h a t t h e l a s e r q u a l i t y o f t h e f i l m i n o t h e r r e g a r d s was comparable t o b u l k - g r o w n c r y s t a l s .
¢p_
z zp
Pbo6eSno t2Te
-
|
to /
0
Homojunction Stripe Geometry
J
i /
~
~
~Si.r~ile-_lqeterojunction
I
~
I
20
I
30
I
40
I
50
I
60
1
70
80
T ('K) F i g . 26
P u l s e d t h r e s h o l d c u r r e n t d e n s i t y of a SH PbSnTe l a s e r as a f u n c t i o n of t e m p e r a t u r e (T) compared w i t h two typical low-threshold dlffused-junctlon bulkcrystal devices.
DE[ l a s e r s grown by t h e h o t - w a l l t e c h n i q u e were f i r s t r e p o r t e d by McLane and S l e g e r (111) i n the PbSl-xSex a l l o y stem. T h r e s h o l d c u r r e n t d e n s i t i e s as low as 60 A/cm2 were o b t a i n e d f o r cw o p e r a t i o n a t 12 K, b u t no h i g h e r t e m p e r a t u r e d a t a were r e p o r t e d . P r e i e r e t a ~ ( l 1 2 ) a l s o u s i n g t h e h o t - w a l l t e c h n i q u e i n t h e PbSl-xSe x a l l o y stem, have s u b s e q u e n t l y o b t a i n e d DH l a s e r s w i t h low t h r e s h o l d a t and above 77 K. These d e v i c e s were made w i t h s t r i p e d c o n t a c t s to form a s t r i p e geometry as seen i n t h e i n s e t of Gig. 29. The p u l s e d t h r e s h o l d c u r r e n t d e n s i t y as a f u n c t i o n of t e m p e r a t u r e i p l o t t e d i n F i g . 29 f o r two Dtt l a s e r s and, f o r comparison, a d i f f u s e d h o m o s t r u c t u r e l a s e r . Above 30 K and 60 K, r e s p e c t i v e l y , t h e two DH l a s e r s show lower t h r e s h o l d current densities than the diffused laser. These d e v i c e s were c a p a b l e of cw o p e r a t i o n up t o 96 K. Higher t e m p e r a t u r e o p e r a t i o n c o u l d have b e e n a c h i e v e d w i t h a t h i n n e r PbS top l a y e r which a t 5 ~m t h i c k n e s s r e p r e s e n t s a l a r g e t h e r m a l impedance ( t h e
83
MBE Techniques for IV-VI Optoelectronic Devices
t h e r m a l c o n d u c t i v i t y of PbS a t 100 K i s a b o u t 0.07 W c~- 1 K -1) (112). L a t e r , s i m i l a r d e v i c e s w i t h 1.5 Um t h i c k top l a y e r s were o p e r a t e d cw up t o 120 K (113). 4.0
-tl
08
•~
G6
•o
o4
Single-Heterojunction P b T s - Pbol m Sno.tzTe L o w
4 . 2 eK
2 0 "K
0.2--
m
.--. 0
I 20
I 30
1 50
I
4o
I
60
I
7O
I (mA)
Fig. 27
The incremental resistance (dV/d/) of a Sh PbSn Te laser vs current I at several temperatures. The abrupt decrease in resistance occurs at ow leaser threshold.
At currents well above threshold, laser modes were seen over a sifficiently broad spectral region that slight inhomogeneity or grading of the PbSSe alloy composition seemed i~dicated.
I n t h e h o t - w a l l PbSSe r e s u l t s d i s c u s s e d above, c o n t r o l of c a r r i e r type and d e n s i t y was a c h i e v e d by an i n d e p e n d e n t l y c o n t r o l l e d Se vapor s o u r c e . While s i m i l a r c o n t r o l of type can be a c h i e v e d i n c o n v e n t i o n a l MBE ( 7 9 ) , i t i s d i f f i c u l t t o a c h i e v e h i g h p - t y p e c o n c e n t r a t i o n s needed f o r low r e s i s t a n c e c o n t a c t s . I n a d d i t i o n , d i f f u s i o n of t h e ~ - ~ j u n c t i o n can be r a p i d even a t t h e low t e m p e r a t u r e s of ~g3E growth as mentioned e a r l i e r . McLane and S l e g e r (111) e s t i m a t e d t h e i r c o n t r o l of j u n c t i o n p o s i t i o n t o b e no b e t t e r t h a n 1 ~m. For t h e s e r e a s o n s , the PbSnTe DH l a s e r s grown by MBE have b e e n doped w i t h f o r e i g n i m p u r i t i e s as d i s c u s s e d i n S e c t i o n 2. R e l a t i v e l y h i g h c o n c e n t r a t i o n s can be e a s i l y a c h i e v e d and j u n c t i o n d i f f u s i o n i s expected to b e s l o w e r f o r f o r e i g n d o p a n t s , as mentioned t h e r e . The PbSnTe DH l a s e r s were grown i n a s t r i p e geometry u s i n g a s t r i p e d growth mask o f MSF2 w i t h s t r i p e w i d t h s of 10 to 50 ~m. S u b s t r a t e s were T l - d o p e d FoTe (~ 1019 cm- 3 ) . A c t i v e r e g i o n s 1.5 ~m t h i c k were grown u s i n g Pbo.78SnO 22Te, Bi-doped to a c h i e v e a b o u t 6 X 1017 cm- 3 n - t y p e c o n c e n t r a t i o n . Cap l a y e r s 0.75 ~m t h i c k were grown u s i n g Pbo.88Sno.12'Te, Bi-doped to a c h i e v e 2 X 1018 cm-3 n - t y p e c o n c e n t r a t i o n . These d e v i c e s were c a p a b l e of ow o p e r a t i o n up t o 114 K (101). However, t h r e s h o l d c u r r e n t d e n s i t i e s were l a r g e p a r t i c u l a r l y f o r low t e m p e r a t u r e s as seen i n F i g . 30, where t h e t h r e s h o l d and t h e w a v e l e n g t h o f e m i s s i o n a r e p l o t t e d vs t e m p e r a t u r e .
84
H. Holloway and J. N. Walpole
Subsequent devices have had somewhat lower thresholds, ranging between 2 and 6 KA cm-2, but remaining large compared to diffused homostructure devices, the SH PbnTe devices, or the PbSSe devices discussed above. It can be easily determined from the wavelength of the emission at threshold and from the I-V characteristics such as that shown in Fig.28, that the voltage across the junction at threshold is within a few mV of the voltage required to establish population inversion. Since the current flow required to reach population inversion is clearly the dominant contribution to the threshold current (115), the high thesholds cannot be explained by optical losses. Instead there appears to be a large component of current due to surface leakage or nonradiative recombination. While misfit dislocations or interface recombination cannot be ruled out as the source of this current, the evidence of the SH lasers and the PbSSe lasers does not support hsi view (41), nor do the results obtained with epitaxial homostructure devices discussed below. Other possibilities include surface leakage, nonradiatlve recombination associated with the high levels of foreign impurities present (116), and nonradiative recombination due to a non-equilibrium, high concentration of native defects which have been proposed to explain low mobility in unannealed MBE films (49) of PbSnTe on BaF 2.
200
mA
-8-15276 I
m II
CURRENT
20 rnV VOLTAGE Fig. 28
The current-voltage characteristic of a DH PbSnTe laser at 77 K. Note the linear relationship above threshold at I = 600 mA.
The high threshold of the PbSnTe DH lasers grown by MBE do not significantly limit their performance since cw operation can be achieved for current densities over 20 KA cm-2. Adequate power and mode quality for high-resolution spectroscopy at 80 K has been demenstrated (117). The control of the thickness of the cap layer and the active layer, made possible by MBE, is a crucial factor in obtaining good performance since, as mentioned, the poor thermal conductivity of the lead-tin chaleogenides causes poor heat-sinking for junctions more than a few ~m deep. The MBE techniques used for the PbSnTe DH lasers have also proven extremely reproducible in numerous growth runs and the yield of devices with cw operation above 77 K from each wafer is high.
MBE Techniques for IV-VI Optoelectronic Devices
PLATED Au \
85
EVAPORATED In Cr
FL,-.'-'.-~-'~-~;'.~-'-'~ ~ S ~,~ p-PbS .j___~ ~ 1~m Cno, pl-PbS~_,Se,
L....................... Au
10 4
/
/
Cr
In
_~----2oo~,m
.-PbS
'
DIFFUSED LASER
%
E <:~
I0 ~
.= ...o
DH-LASERS
102 - -
10
Fig. 29
I
l
i
I
i
I
,
I
I
I
,I,I,I,
Comparison of pulsed threshold current densities of DH PbSnTe lasers with a bulk crystal diffused-junction laser as a function of temperature (from Eel. 112).
No other lasers grown by MBE or hot-wall techniques have been reported. There are several papers of interest in the Russian literature (116,118,119). However, no high performance at higher temperatures was mentioned and the growth techniques were not always specified but appear to involve vapor epltaxy. A disappointing aspect of the ~ E and hot-wall heterostructure lasers reported to date is that their external quantum efficienies have been low (less than a percent or two). Typical maixmum output power has been on the order of a few hundred ~W in PbSSe and about I00 BW in PbSnTe.* Much higher external efficlencies and output
86
H. Holloway and J. N. Walpole
powers have b e e n a c h i e v e d i n b u l k d i f f u s e d - j u n c t i o n l a s e r s (104, 106). A l t h o u g h the powers o b t a i n e d from h e t e r o s t r u c t u r e s a r e more t h a n a d e q u a t e f o r most a p p l i c a t i o n s , they are marginal for local oscillator applications in heterodyne detection.
20 18
14
12-10
12 .~
10 8 6
Jth L
U,J
IJJ
4
U 2
0
I
10
1
20
I
30
4O
I
50
I
60
I
"tO
1
80
I
90
1
100
I
110
OE T
t-U
0 120
HEAT-SINK TEMPERATURE ( K )
Fig. 30.
Threshold current density and emission wavelengths of a DH PbSnTe laser as a function of heat-sink temperature. The longest wavelength observed pulsed compared to the cw threshol~ wavelength gives a measure of the temperature increase in the active region.
The low external efficiency cannot generally b e attributed to low internal efficiency since the DH PbSnTe lasers which show junction-voltage saturation at threshold also have low external efficiency. Poor output coupling and scattering losses at the end mirrors due to poor mirror quality (104) and due to mode-coupling effects (121) nmy play a large role in the low efficiency. Also scattering losses and free-carrier absorption in the active region may be much larger in laser devices than in the waveguide structures discussed in section 4. The laser waveguides are much thinher, which leads to increased scattering due to roughness of the waveguide walls (103), and carrier densities in the lasers are much higher**. Optimized doping and the use of separate optical and carrier confinement structures might significantly improve external efficiency. An early DH Pbo.88Sno.12Te laser (120) with an undoped active region 6 ~m thick and an undoped PbTe cap 1 ~m thick grown on a Tl-doped PbTe substrate gave cw output power up to 9 ~f~. These devices were not capable of cw operation above = 30 K, however. ** Additional absorption above the classical free-carrier absorption may be present near the band edge in heavily doped materials (93,122).
87
HBE T e c h n i q u e s f o r IV-VI O p t o e l e c t r o n i c Devices 8.2.
Homostructure l a s e r s
S t r o n g o p t i c a l and c a r r i e s c o n f i n e m e n t can be a c h i e v e d i n the l e a d - t i n c h a l c o g e n i d e s w i t h p r o p e r l y c o n t r o l l e d c a r r l e r - c o n c e n t r a t l o n p r o f i l e s (123). Because of t h e s m a l l e f f e c t i v e masses of b o t h e l e c t r o n s and h o l e s , r e l a t i v e l y l a r g e b u i l t - i n p o t e n t i a l s (band b e n d i n ~ o c c u r a t i n t e r f a c e s between low and h i g h c a r r i e r c o n c e n t r a t i o n . Also l a r g e f r e e - c a r r l e r c o n t r i b u t i o n s to the d i e l e c t r i c c o n s t a n t o c c u r b e c a u s e of t h e s m a l l e f f e c t i v e masses and t h e s m a l l energy gaps o r long w a v e l e n g t h s of o p e r a t i o n . D i f f u s e d - j u n c t i o n l a s e r s have b e e n o p e r a t e d ow r e c e n t l y up t o 6 1 K u s i n g c o n c e n t r a t i o n g r a d i e n t s f o r c o n f i n e m e n t (124). The d i f f u s i o n i s v e r y c r i t i c a l and d i f f i c u l t to control s i n c e t h e j u n c t i o n depth must be kept v e r y s m a l l f o r h e a t - s i n k l n g . E p i t a x i a l t e c h n i q u e s a r e b e t t e r s u i t e d f o r c o n t r o l l i n g c o n c e n t r a t i o n and depth of the j u n c t i o n . E p i t a x i a l h o m o s t r u c t u r e d e v i c e s c a p a b l e of cw o p e r a t i o n to 100 K have b e e n r e p o r t e d (125) grown by b o t h ~ E and LPE h a v i n g n + - n - p + doping p r o f i l e s . Bidoped a c t i v e r e g i o n s (n ffi 5 X 1017 cm-3) 1.5 pm t h i c k f o l l o w e d by Bi-doped capping l a y e r s (n + ffi 5 X 1018 cm-3) 0.75 pm t h i c k were grown on T l - d o p e d s u b s t r a t e s (p+ = 1019 cm-3). The HBE h o m o s t r u c t u r e s were grown i n PbTe and Pbo.78Sno.22Te, b o t h c o m p o s i t i o n s operati~tg cw to 100 K. T h r e s h o l d c u r r e n t d e n s i t i e s were comparable to t h e DH l a s e r s b u t o u t p u t powers were somewhat b e t t e r . F i g u r e 31 shows t h e cw s p e c t r a of a PbTe h o m o s t r u c t u r e a t 15 and 102 K w i t h 2.4 mN and 0.64mW o u t p u t , r e s p e c t i v e l y , predomln a n t l y i n a s l n g l e mode. The maximum o u t p u t o b s e r v e d from Pbo.78Sno.22Te homos t r u c t u r e s was ~ 0 . 4 mW a t 15 K and ~ O.12M~ a t 80 K.
T = 15K I = h84A
T= 102K
Ith = 0.12A
Ith • 1.3A
I ffi 2.18A
P = 0.64roW (Single End)
P = 2.4roW (Single End)
i
k
.
1680
L
_
_
1590
~
~
1600
1610
1950
1960
i
@32
I
I 6.28
I
I 6.24
WAVELENGTH (Fro)
F i g . 31.
i
l 6.20
[ 5.12
I
I
1970
I
1980 1990 WAVE NUMBER (cm -1)
WAVE NUMBER (cm-1) I
/
-
I
5.08
I
i
I
5.04
.
I
I
2000
WAVELENGTH (Fro)
S p e c t r a of a I~E h o m o j u n c t i o n PbTe l a s e r a t 15 and 102 K showing the r e l a t i v e l y h i g h s i n g l e - m o d e o u t p u t powers.
S i n c e t h r e s h o l d s a r e comparable t o DH l a s e r s and e x t e r n a l e f f i c i e n c i e s
i
500
are s t i l l
88
H. Holloway and J. N. Walpole
rather low, the homostructure results again imply that the lattice mismatch present in otherwise similar heterostructurelasers plays little role in limiting their performance. Nevertheless, if lattice matching should prove important, I~SE homostructures seem to offer the advantage of higher temperature operation without the lattice mismatch problem. 8.3.
Distributed feedback lasers
Figure 32 shows schematically the structure of a distributed feedback (DFB) laser which has been fabricated in MBE-grown PBSnTe layers (126, 127). The device is essentially the same as the DH lasers discussed above except that feedback is provided by Bragg reflection from the periodic grooves (grating) etched into the top layer rather than by reflections from the two ends of the device. The 800-A insulating MgF 2 film waspatterned as shown so that only one end of the mesa stripe was electrically contacted leaving an uncontacted length with high optical absorption to prevent reflective feedback from the other end of the mesa. The grating and the deeper grooves isolating the mesa stripe were obtained using photolithographic techniques and rf sputter etching.
[~
o
~
~' ~
60N'm
- 800 ~ .4~= 9.~.
I
/
1.1 • m
I ~
O.3~m DEPTH
0.5/~m
i.5~.m I
~'~" X --- 0
,~p.,..I~"
n ~ 2 X I018cm-3 (8i) n
~ 6 X 1017cm-3 (Bi)
p~ 1019 cm-3 (TL) Fig. 32.
Schematic of DFB double-heterostractare PbSnTe laser without metallization.
One advantage of the DFB laser is that is has significant mode discrimination, i.e. the longitudinal spatial modes of the cavity have different thresholds. The PbSn Te DFB lasers showed predominantly single-mode cw operation over a wide range of temperature and current. Another advantage of DFB lasers is that the frequency of the laser can be controlled by the grating periodicity since oscillation can occur only near the frequency at which Bragg reflection occurs. These advantages make DFB lasers potentially the ideal laser for many applications. The successful fabrication of the DFB structure shown in Fig. 32 is due largely to the ability to accurately control the thickness of the top layer to <0.i ~m. It is a device for which MBE techniques are ideally suited. The scanning electron micrographs shown in Fig. 33 depect two views of grating of I.I Bm periodicity with grooves 0.3 ~m deep. The grooves must be close to the active layer to obtain good
MBE Techniques for IV-VI Optoelectronic Devices
89
c o u p l i n g f o r Bragg r e f l e c t i o n . They must n o t , however, p e n e t r a t e t h e a c t i v e l a y e r o r c a r r i e r c o n f i n e m e n t would b e l o s t . The cap l a y e r was grown 0 . 5 ~m t h i c k , so t h a t c o n t r o l o f e t c h i n g p a r a m e t e r s a s w e l l as l a y e r t h i c k n e s s i f c r i t i c a l f o r good performance.
F i g . 33.
Scanning electron micrographs of a grating with l.l-vm periodicity, sputter etched to a depth of 0.3 ~m in PbSnTe.
9.
CONCLUSIONS
The optoelectronic device results which have been reviewed here demonstrate that hlgh-quality devices, competitive in performance with bulkcrystal devices, can be achieved in IV-VI MBE films. Indeed there are technological advantages and performance features discussed here which have not been demonstrated in bulkcrystal devices. These included the option of backside illumination of detector devices with transparent substrate; the compatibility with planar technology for smell
90
H. Holloway and J. N. Walpole
c l o s e l y - s p a c e d elements and for a p p l i c a t i o n s in i n t e g r a t e d o p t i c s ; and the c o n t r o l of l a y e r thicknesses which makes p o s s i b l e l a s e r devices with higher temperatures of cw o p e r a t i o n , DFE l a s e r s devices u t i l i z i n g t h i n - f i l m o p t i c s , and (because of the confinement provided by the proximity of an i n s u l a t i n g s u b s t r a t e ) devices such as the p i n c h - o f f diodes and the l a t e r a l c o l l e c t i o n devices. While HBE techniques have led to advances in device technology, i t i s also true t h a t the device r e s u l t s so obtain have given important information on the q u a l i t y of HBE films. The quantum e f f l c l e n c l e s and c u r r e n t - v o l t a g e c h a r a c t e r i s t i c s of j u n c t i o n devices give evidence about minority c a r r i e r l i f e t i m e , generation-recombination mechanisms, and i n t e r f a c e recombination. The g e n e r a l l y hish q u a l i t y with r e s p e c t to these p r o p e r t i e s found in the d e t e c t o r devices i s not found in many of the l a s e r s t r u c t u r e s discussed, as evidenced by the high t h r e s h o l d s . Also o p t i c a l a b s o r p t i o n , which can be low in passive I~E waveguides, appears to be high in the t h i n , h e a v i l y doped l a y e r s of the l a s e r s t r u c t u r e s , as evidenced by the low e x t e r n a l quantum e f f i c i e n c i e s . Although the device r e s u l t s suggest approaches f o r improvements and f u r t h e r progress is l i k e l y , there is need f o r more work to c o r r e l a t e c r y s t a l p e r fection (e.g. native defect concentrations, foreign impurities, misfit dislocations, i n t e r f a c e s t a t e s , e t c . ) with the p r o p e r t i e s s i g n i f i c a n t f o r device performance, i . e . c a r r i e r recombination and o p t i c a l l o s s e s . C r i t e r i a f o r good device performance are not e s t a b l i s h e d even f o r bulk c r y s t a l s and b e t t e r knowledge would g r e a t l y simplify the process of optimizing the growth and the design and f a b r i c a t i o n of devices. To this end, more s o p h i s t i c a t e d HBE systems with i n - s i t u a n a l y s i s equipment such as Auger and HEED instrumentation might prove f u l t f u l . The s t a t e - o f - t h e - a r t performance of IV-Vl o p t o e l e c t r o n i c devices 8 r o w n b y l ~ E is already more than adequate f o r many a p p l i c a t i o n s . Because of t h e i r demonstrated a p p l i c a t i o n s and t h e i r p o t e n t i a l for new a p p l i c a t i o n s , I v - v I ~ E devices are l i k e l y to remain t e c h n o l o g l c a l l v imvortant in the f u t u r e . REFERENCES
i. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15a. 15b. 16. 17. 18. 19. 20. 21.
H. Hinterberger, Z. Phys. 119, 1 (1942). L. Sosnowskl, J . Starklewicz and O. Simpson, ~ t t t r ~ , Lond. 159, 818 (1947). A.J. E11em-- and H. Wilman, Pros. Phys. Soy. Lond. 61, 164 ~ ' 9 4 8 ) . S.G. Parker, J. eZ4c~roeher,. Sos. 123, 920 (1976). H.K. Pandy, SoZ4d St. Commun.15, 449 (1974). L.R. Koller and H.D. C o g h i 1 1 , 7 . eleetr, oehem. Sos. 107, 973 (1960). P. Hudock, Trans. Am. Ins#. Mesh. ~ s . 239, 338 (1-~7). R.P. Bis, J.R. Dixon and J.R. Lowrey, J. v a t . $~i. Teohnol. 9, 226 (1972). A. Lopez-Otero and L.D. Haas, Thin S o l i d F4/~e 23, 1 (1974). K. Duh and H. P r e i s e r , Th4n S o l i d F4~m~ 27, 247~1975). I . Kasai, J . Hornung and J . Baars, J. e ~ t m o n . Mater. 4, 299 (1975). I . Kasai, D.W. Bassett and J . Hornung, J. appl. Phys. 47, 3167 (1976). R.B. Schoolar, J.D. Jensen and G.M. Black, Appl. Phys. L e f t . 31, 620 (1977). E.B. Schoolar and J.D. Jensen, Appl. Phys. L e t t . 31, 536 (1977-~. L.L. Chang and R. Ludeke, in: E p 4 ~ l Growth, P ~ t A, (ed.) J.W. Matthews, Academic Press, New York, San Francisco, London (1975). J.N. Zemel, in: Sol4d State Sumf~e S~en~e, (ed.) M. Green, Dekker, New York (1969). A. Lopez-Otero, Thin Solid Films 49, 3 (1978). A.M. Andrews J.A. Higglns, J.T. Longo, E.R. Gerner and J.G. Pasko, Appl, Phys. Left. 21 , 285 (1972). C.C. Wang and S.E. Hampton, Solid-St. Eleol-z~m. 18, 121 (1975). S.H. Groves, K.W. Hill and A.J. Strauss, Appl. ~ s . Left. 25, 331 (1974). T.F. Tao, I. Kasai and J.P. Butler, J. ~ . Sei. T e e ~ l . 6, 918 (1969). E. Krlkorlan J. ~ o S~. Te~hnol. 12, 186 (1975).
I~E Techniques f o r IV-VI Optoelectronic Devices 22. 23. 24. 25.
R.F. R.F. R.F. R.F.
P o r t e r , J. ohM. Phys. 34, Brebrick and A.J. S t r a u s s , Brebrick and A.J. Strauss, Bis, J. va~. Soi. Teah~o~.
91
583 (1961). J . oh~m. Phys. 40, 3230 (1964). J. ohem. Phys. 4 , 197 (1965). 7, 126 (1970).
26.
w.c. Chan, I n ~ d
27. 28. 29. 30. 31. 32.
D.A. Northrop, J . e~en#~oa~m. 6oc. 118, 1365 (1971). M.C. Finn and J.N. Walpole, unpubllshed. A.J. S t r a u s s , Phys. Rev. 157, 608 (1967). R.W. B i c k n e l l , I n ~ d PhYs. 17, 57 (1977). R.W. Bickne11, In3~az,ed Phys. 18, 133 (1978). T.C. Harman, in: Phys44s of I V ~ I Compounds and Alloys, (ed.) S. Rabii, Gordon & Breach, London (1974). J.N. Walpole and R.L. Guldl, in: Physics of IV-VI Compounds and A~loysj (ed.) S. Rabli, Gordon & Breach, London (1974). G. Dionne and J.C. Woolley, Phys. ReO. B6, 3898 (1972). D.L. Smith and V.Y. Pickhardt, J. eZeutTon. Muter. 5, 247 (1976). H. Holloway, D.K. Hohnke, R.L. Crawley and E. Wilkes, J. ~u. Sui. Teuhnol. O, O0(0OO). J.W. Matthews, Phil. Mug. 6, 1347 (1961). J.W. Matthews, Phil. Mug. 8, 711 (1963). K. Yagi, K. Takayanagi, K. Kobayaski, and G. Honjo, J. Cr~s. Gmmoth 9, 84 (1971) J.N. Walpole, S.H. Groves, A.R. Calawa and T.C. Harman in: Solid Sta~e Reseach Report, Lincoln Laboratory, M.I.T. (1974:4), p.13 DDC AD-AO02773/O. R.W. Bicknell, J. ~ac. Sci. Teuhnol. 14, 1012 (1977). J.N. Zemel, J.D. Jensen and R.B. Schoolar, Phys. Rev. 140A, 330 (1965). J.H. Myers, R.H. Morrlss and R.J. Deck, J. dppl. Phys. 42, 5578 (1971). H. Holloway, in: Use of Thin Films for Phys~al Invest~tions, (ed.) J.C. Anderson, Academic Press, New York (1966). R.S. Allgaler and B.B. Houston, Jr., Proc. Int. Conf. Phys. Semiuondaotors, p. 172, Exeter, 1962, Institute of Physics and Physical Society, London, (1962). E.M. Logothetls and H. Holloway, in: Physiu8 of IT-VI Compounds and AILg~8, (ed.) S. Rabii, Gordon & Breach, London (1974). J.P. Donnelly, T.C. Harman and A.G. Foyt, Apply. Phys. Left. 18, 259 (1971). H. Holloway, E.M. Logothetis and E. Wilkes, J. e'ppZ. Phys. 41, 3453 (1970). H. Holloway and E.M. Logothetls, J. appl. Phys. 42, 4522 (I~I). T.O. Fahrlnre and J.N. Zemel, d. vac. Sui. Teu~no-~. 7, 121 (1970). H.M. Manaserit and W.I. Simpson, J. ~eu%Pouhem. Sou. 122, 444 (1975). V.L. Lambert, J. appl. Phys. 46, 2304 (1975). D.K. Hohnke, H. Holloway and M.---D.Hurley, Thin Solid Film8 38, 49 (1976). H. Holloway, unpublished data. A. Lopez-Otero, J. appl. Phys. 48, 446 (1977). R.S. Allgaier and W.W. Scanlon,-~hys. Rev. III, 1029 (1958). G.A. Antcliffe and J.S. Wrobel, Mu~P. Res.~ll. 5, 747 (1970). C.M. Wolfe and G.E. Stillman, Appl. Phys. Left. 18~ 205 (1971). R.B. Schoolar, Appl. Phys. Lett. 16, 446 (1970). R.B. Schoolar, J. ~u. Sui. Teuhno'-~. 9, 225 (1972). E.M. Logothetls, H. Holloway, A.J. Varga and E. Wilkes, Appl. Phys. Left. 19, 318 (1971). K.W. Nil1, A.R. Calawa, T.C. Harman and J.N. Walpole, App,. Phys. Left. 16, 375 (1970). K.W. N i l l , J.N. Walt)ole. A.R. Calawa and T.C. Harman. in: The Phus~s o~ Semimentuls and ~ C , ro Semiuond~torsj (eds) D.L. Carter and R.T. Bate, Pergamon Press, Oxford (1971). H.B. Morris, R.A. Chapman, R.L. Guldi and S.C. Parker, paper presented at IRIS Detector Speu441#y Gz~up Meeting, Fort Monmouth, 1975. J.N. Walpole and K.W. Nil1, J. uppl. P~JS. 42, 5609 (1971). H. Holloway, W.H. Weber, E.M. Logothetis, A.J. Varga and K.F. Yeung, App,. Phys Lett. 21, 5 (1972).
33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
Phys. 14, 177 (1974).
92 67. 68.
H. Holloway and J. N. Walpole
H. Holloway, W.J. Johnson and A.J. Varga, unpublished data. H. Holloway, in: Physics of IV-VI Compounds and Alloys, (ed.) S. Rabli, Gordon & Breach, London (1974). 69. E.M. Logothetis, H. Holloway, A.J. Varga and W.J. Johnson, Appl. Phys. Lett. 21, 411 (1972). 70. J.P. Donnelly and H. Holloway, Appl. Phys. Lett. 23, 682 (1973). 71. R.E. Callender, paper presented at IRIS Detector ~eciaZty Group Meeting, Washington, D.C., 1973. 72. A. Lopez-Otero, L.D. Haas, W. Jantsch and K. Lischka, Appl. Phys. Lett. 28, 546 (1976). 73. Contract DAAKK 2-73-C-0225, Final Report, Febuary 1974. 74. H. Holloway, K.F. Yeung, A.E. Asch and D.A. Gorski unpublished data. 75. T.J. McMahon, Paper presented at IRIS Detector Specialt F Group Meeting, E1 Toro, 1974. 76. D.K. Hohnke and H. Holloway, Appl. Phys. Lett. 24, 633 (1974). 77. D.K. Hohnke and M.D. Hurley, J. appl. Phys. 47,--4975 (1976). 78. H. Holloway, D.K. Hohnke, K.F. Yeung, A.E. Asch and D.A. Gorski, unpublished data. 79. D.K. Hohnke and S.W. Kaiser, J. appl. Phys. 45, 892 (1974). 80. J.F. Butler, A.R. Calawa and T.C. Harman, App-~. Phys. Lett. 9, 427 (1966). 81. D.K. Hohnke, H. Holloway, K.F. Yeung and M.D. Hurley, Appl. ~hye. Lett. 29, 98 (1976). 82. W.H. Rolls and D.V. Edolls, Infrared Phys. 13, 143 (1973). 83. D.P. Mathur, Opt. Engng. 14, 351 (1975). 84. A.E. Asch, D.A. Gorskl, P.I. Zapella and H. Holloway, paper presented at IRIS Detector Specialty Group Meeting, Fort Monmouth, 1975. 85. H. Holloway, J. appl. Phys. (to be published). 86. H. Holloway and K.F. Yeung, Appl. Phys. Lett. 30, 210 (1977). 87. Contract DAAK02-75-C-OIO6, Final Report, June 1976. 88. H. Holloway, M.D. Hurley and E.B. Schermer, Appl. Phys. Lett. 32, 65 (1978). 89. H. Holloway, J. appl. Phys. (to be published). 90. A.J. Norelka, M.H. Francombe, W.J. Taei, R.N. Ghoshtagore and J.L. Wentz, paper presented at IRIS Detector Specialty Group Meeting, Colorado Springs, 1977. 9!. R.W. Ralston, J.N. Walpole, T.C. Harman, and I. Melngailis, Appl. Phys. Lett. 26, 64 (1975). 92. A.J. Strauss, J. Nonmetals i, 133 (1973). 93. C. Picketing, Solid St. Coleman. 22, 395 (1977). 94. J.L. Merz, R.A. Logan, W. Wiegmann, and A.C. Gossard, Appl. Phys. Lett. 26, 337 (1975). 95. J.N. Walpole, unpublished. 96. E.D. Hinkley, K.W. Nill, and F.A. Blum, in: Topics in Applied Phys@4s Vol. 2, Laser Spectroscopy of Atoms and Molecules, (ed.) H. Walter, Sprlnger-Verlag, Berlin, Heidelberg, New York (1976). 97. E.D. Hinkley, R.T. Ku, and P.L. Kelley, in: Topics in Applied Phys~s, Vol. 14, Laser Monitoring of the A~osphere, (ed.) E.D. Hinkley, Springer-Verlag, Berlin Heidelberg, New York (1976). 98. M. M,,~8, T. Kostiuk, S. Cohen, D. Buhl, and P.C. Von Thuna, Nature, Lond. 253, 514 (1975). 99. M.A. Frerking and D.J. Muehlner, Appl. Opt. 16, 526 (1977). I00. R.T. Ku and D.L. Spears, Opt. Lett. I, 84 (I~7). iO1. J.N. Walpole, A.R. C~lawa. T.C. Harman. and S.H. Groves, Appl. Phys. Lett. 28, 552 (1976). 102. K.J. Sleger, G.F. McLane, and D.L. Mitchell, in: Physics of IV-VI Compound8 and Alloys, (ed.) S. Rabii, Gordon & Breach, London (1974). 103. W.H. Weber and K.F. Yeung, J. ¢cppl. Phys. 44, 4991 (1973). 104. J.N. Walpole, A.R. Calawa, R.W. Ralston, and T.C. Harman, J. appl. Phys. 44, 2905 (1973). 105. J.B. Walpole, A.R. Calawa, R.W. Ralston, T.C. Harman, and J.P. MoVittle, Appl. Phys. Lett. 23, 620 (1973).
MBE Techniques for IV-VI Optoelectronic Devices 106. 107. 108. 109. IiO. Iii. 112. 113. 114. 115. 116.
R.W. Ralston, J.N. Walpole, A.R. Calawa, T.C. Harman, and J.P. McVittle,
J. appl. Phys. 45, 1323 (1974). T.L. Paoli and P.-'-A.Barnes, Appl. Phy8. Lett. 28, 714 (1976). K.J. Linden, K.W. Nill, and J.F. Butler, IEEE J. Quantum Electron. QE-13, 720 (1977). H.S. Sommers, Jr., Appl. Phys. Lett. 32, 547 (1978). K.J. Sleger, G.F. McLane, U. Steom, S.G. Bishop, and D.L. Mitchell, J. appl. Phys. 45, 5069 (1974). G.F. McL---aneand K.J. Sleger, J. electron, Mater. 4, 465 (1975). H. Preier, M. Bleicher, W. Riedel, and H. Maier, ~ppl. Phys. Lett. 28, 669 (1976). H. Preier, M. Bleicher, W. Riedel, and H. Maier, J. appl. Phys. 47, 5476 (1976) D.K. Honks and S.W. Kaiser, J. appZ. Phys. 45, 892 (1974). L.R. Tomasetta and C.G. Fonstad, IEEE J. Quantum Electron. QE-II, 384 (1975). L.N. Kurbatov, A.D. Britov, S.M. Karavaev, G.A. Kalyuzhnaya, M.I. Nikolaev, O.V. Pelevin, B.G. Girich, and T.F. Terekhovich, SoY. J. Quantum Electron. 7, 236
(1977). 117. 118. 119.
120. 121. 122. 123. 124. 125.
126. 127.
P.F. Moulton, D.M. Larsen. J.N. Waloole. and A. Mooradian. Opt. Lett. i. 51 (1977). L.N. Kurbatov, A.D. Britov, S.M. Karavaev, Y.I. Gorina, G.A. Kalyuzhnaya, and P.M. Starik, Soy. J. 6~z~ntT~mElectron. 5, 1137 (1975). A.D. Britov, S.M. Karavaev, G.A. Kalyuz~naya, Y.I. Gorina, A.L. Kurbatov, K.F. Kiseleva, and P.M. Starik, Soy. J. Quantum Electron. 6, 1217 (1976) and 6, 1384 (1976). J.N. Walpole, R.W. Ralston, A.R. Calawa, and T.C. l{arman, paper presented at the IEEE International Semiconductor Laser Conference, Atlanta (1974). R.W. Davies and J.N. Walpole, IEEE J. Quantum Electron. QE-12, 291 (1976). S.L. McCarthy, W.H. Weber, and M. Mikkor, J. a~rpl. Phys. 45, 4907 (1974). P.N. Asbuk, Ph.D. Thesis, M.I.T., 1975 (unpublished). W. Lo, IEEE J. Quantum Electron. QE-13, 591 (1977). J.N. Walpole, S.H. Groves, and T.C. Harman, IEEE Trans. Electron Devices ED-24, 1214 (1977). Also, Solid State REsearch Report, Lincoln Laboratory, M.I.T. (1977:4), p. 3, DDC AD-AO52463. J.N. Walpole, A.R. Calawa, S.R. Chinn, S.H. Groves, and T.C. Harman, Appl. Phys. Lett. 29, 307 (1976). J.N. Walpole, A.R. Calawa, S.R. Chinn, S.H. Groves, and T.C. Harman, Appl. Phys. Left. 30, 524 (1977).
94
H. Holloway and J. N. Walpole
THE AUTHORS
Dr. H. Holloway
Professor J. N. Walpole
Dr. Holloway was educated at Birkbeck College, University of London, where he received his B.Sc. (1956) and Ph.D. (1959) degrees in Chemistry. His research interests have included x-ray and electron diffraction, epitaxy -- particularly of compound semiconductors -- and the physics of thin-film devices. Currently, he is studying the properties of optoelectronic devices made with thin films of IV-VI Semiconductors. James Walpole was granted a B.S.E.E. degree in 1961 from Duke University, and then attended the Massachusetts Institute of Technology as a National Science Foundation Fellow, receiving his M.S. degree in 1962, and his Ph.D. degree in 1966, both in the Department of Electrical Engineering. His -~jor studies were in the areas of solid state physics and semiconductor devices. He was appointed Assistant Professor of Electrical Engineering at M.I.T. in 1966, and Associate Professor in 1971. Since 1972 he has been a staff member in the Applied Physics Group of the M.I.T. Lincoln Laboratory, where he is currently involved with development of GalnAsP heterostructure lasers. His previous research interests have centered around the physics of the lead-tin chalcogenide semiconductors, including studies of magnetoplasma wave propagation, diffusion, hot-electron conduction, metal-semiconductor contacts, and, in recent years, various aspects of injection lasers in these materials. He lives in Concord, Massachusetts with his wife and two children.
MBE TECHNIQUES FOR IV-Vl OPTOELECTRONIC DEVICES* H. Holloway Ford Motor Company, Dearborn, Michigan 48121, U.S.A. and
J.N. Walpole Lincoln Laboratory, Massachusetts Institute of Technology, Lexington, Massachusetts 02173, U.S.A. (Submitted July 1978)
ABSTRACT Recent development of hlgh-quality IV-VI optoelectronic devices grown by MBE has significantly increased the technological importance of this epltaxial technique in IV-VI materials. Despite the progress made, much is still unknown about the importance of crystal perfection for device performance. Criteria for crystal perfection required to minimize carrier recombination and optical losses in i.r. devices in IV-VI materials need to be established. In this paper the literature on IV-VI film growth by vacuum deposition techniques is briefly reviewed. MBE techniques used for growth of IV-VI materials on BaF 2 and SrF 2 substrates and for growth of PbSnTe on PbTe substrates are described. Emphasis is on the techniques used to deposit pseudobinary alloys with homogeneous composition. The use of foreign impurity dopants is also discussed. Criteria of quality and crystal perfection of epltaxlal layers needed for device performance are evaluated. In layers grown on IV-VI substrates, large dislocation densities due to lattice mismatch are normally present but may not be detrimental to device performance. Lattice-matched heteroepitaxlal systems are discussed. For insulating substrates, it is shown that the crystalline quality and the carrier mobility of layers grown on BaF 2 and SrF 2 are superior to those grown on alkali halide substrates and that the latter are unsuitable for making devices. Carrier mobility is not a good test of crystalline perfection, however, it is argued that device performance, particularly in the demanding requirements for low-noise photodiodes, is a sensitive measure of the epitaxial material quality. Thinfilm photodiodes grown by MBE and the hot-wall techniques are then reviewed in detail, including some recent unconventional * The Lincoln Laboratory portion of this work was sponsored by the Department of the Air Force and the Defense Advanced Research Projects Agency.
49
50
H. Holloway and J. N. Walpole devices in which the unique properties of thin films are exploited. Heterostructure optical waveguides are described briefly followed by a review of diode laser results obtained by MBE and the hot-wall technique. These results are for thin-film PbTe lasers grown on BaF 2 and for a variety of devices grown on I-VI substreates including single heterostructure, double heterostructure, homostructure, and distributed feedback lasers.
i.
INTRODUCTION
The first major application of the IV-VI semiconductors was the use of PbS layers as infrared-sensitive photoconductors in the 1940's. Most of the IV-Vl photoconductlveworkhas been with polycrystalline precipitates, which represent a divergent evolutlonary llne from the optoelectronlc devices that are described in the present article. However, the early studies of photoconductive PbS included vacuum deposition of thin films (1,2), which led to a demonstration of epltaxlal growth on NaCI (3) that may be regarded as the progeniotor of modern epitaxlal IV-VI devices. In a second wave of development during the 1960's, techniques were established for the growth of goodquality single crystals of IV-Vl semiconductors and for making p-n junction devices that are discussed here, The selection of subject matter for the present review involves some arbitrarily chosen boundaries because there is not a well-deflned separation between thin-film and bulk crystal growth techniques. At one end of the spectrum we have vacuum deposition of Dm thick films using a molecular beam that is generated with an effusion cell. At the other extreme, we find closed-tube transport down a temperature gradient to give crystals with cm dimensions either by spontaneous nucleation or with use of a seed crystal (4). This range of techniques includes both epitaxy on isostructural IV-VI substrates (or seeds) and or insulating, such as BaF2.* Much of the recent literature on growth of IV-VI thin films relates to a tehnique in which the source and the substrate are enclosed by a "hot-wall" that pipes most of the evaporant to the substrates. ** This method may be regarded as intermediate between molecularbeam epitaxy (MBE) *** and closed-tube transport. Thus, depending upon the wall temperature, the efflcleny of the piping and the approximation to equilibrium, growth conditions may vary between the extremes that are obtained with MBE and with closedtube transport. This intermediate character is also evident in the hot-wall layer thicknesses which may exceed IOO ~m (8), thereby approaching bulk crystal growth. In the following, the description of epitaxial growth techniques is restricted to two representative MBE systems that were developed by the authors and their colleagues to pursue their respective interests in epitaxial i.r. detectors and epltaxlal i.r. lasers. We have mostly omitted reference to extensive body of earlier work on MBE of IV-VI semiconductors because, despite its value for fundamental studies of the materials, the emphasis on alkali halid substrates led to layers whose crystal perfection was inadequate for p-~n juncation devices.
Insulating substrates are usually associated with tin films of the IV-VI semiconductors, but Pandy 5 has shown that closed-tube growth of (Pb,Sn) Te may be initiated with a BaF 2 seed. The hot-wall technique appears to have originated with Koller and Gohill (6) who used it to deposit ZnS. Subsequently it has been used extensively for IV-VI semiconductors (7-14). *** Molecular-beam epltaxy is a recent designation for a much older technique.
This
MBE Techniques for IV-Vl Optoelectronic Devices
51
name i s q u i t e d e s c r i p t i v e f o r t h e IV-VI s e m i c o n d u c t o r s where the beams a r e r e a l l y of m o l e c u l a r s p e c i e s . The term i s a l s o a p p l i e d ( 1 5 ) , though w i t h l e s s j u s t i f i c a t i o n , t o growth of m a t e r i a l s l l k e t h e I I I - V s e m i c o n d u c t o r s , where one works w i t h beams of the c o n s t i t u e n t e l e m e n t s . These studies have been reviewed by Zemel (15b). Epitaxial growth by the hot-wall technique has also been omitted because a recent review is available (16). However, for completeness we have included the results of recent hot-wall studies which indicate that this technique can yield devices whose performance is comparable to those made with MBE. We have excluded consideration of llquld-phase epltaxy because this would take us too far from the main area of our review, but we note in passing that this method has been shown to give high-performance IV-VI photodiodes (17,18) and lasers (19) on IV-VI substrates.
2.
TECHNIQUES FOR MOLECULAR-BEAM EPITAXY
Vacuum d e p o s i t i o n of t h e IV-Vl s e m i c o n d u c t o r s and t h e i r p s e u d o b i n a r y a l l o y s has m o s t l y been accomplished by e v a p o r a t i v e methods, a l t h o u g h some use has been made of f l a s h e v a p o r a t i o n (20) and of s p u t t e r i n g ( 2 1 ) . From t h e l i t e r a t u r e one may conclude t h a t almost any o f the s t a n d a r d vacuum d e p o s i t i o n t e c h n i q u e s w i l l y i e l d e p i t a x i a l l a y e r s , so t h a t t h e c h o i c e of method i s t o some e x t e n t a m a t t e r of t a s t e . However, e v a p o r a t i v e t e c h n i q u e s a r e p a r t l c u l a r l y a t t r a c t i v e b e c a u s e of t h e IV-VI s e m i c o n d u c t o r s have c o n v e n i e n t vapor p r e s s u r e s ( P ~ 0 . 1 T o r r a t 1000 K) and s u b l i m e p r e d o m i n a n t l y as d i a t o m i c m o l e c u l e s ( 2 2 - 2 4 ) . Even with evaporative techniques, care is needed when the deposit is a pseudobinary alloy. Use of a source of the alloy (25,26) does not, in general, give truly congruent evaporation. For example, Northrop (27) has shown that the gas phase in equilibrium with Pbl_xSnxTe is enriched in the SnTe component. The enrichment is roughly 30% over the range of compositions (x • O.2) that are suitable for i.r. detectors. With use, a Pbl_xSnxTe source that is operated near equilibrium will become depleted of SnTe and the energy gap of the deposit will drift towards larger values. This places a severe constraint upon the fraction of the source material that may be used if a uniform deposit composition is required. From Northrop's results a source with the composition Pb O 85Sno.15Te when operated near 7OOOC will initially give a deposit with composltion near Pbo.8oSno.2oTe corresponding to a cut off wavelength of approximately 11 ~m at 80 K. After use of 20% of the source material the source and deposit compositions may be estimated to be Pbo.86Sno.14Te and Pbo.81Sno.19Te, respectively.. This would lead to a decrease of the 80 K cutoff wavelength of the deposltied semiconductor by approximately 0.Spm.
E v a p o r a t i o n of a p s e u d o b l n a r y a l l o y w i t h o u t a d r i f t i n i t s c o m p o s i t i o n i s p o s s i b l e , p r o v i d e d an open c r u c l b l e r a t h e r t h a n a Knudsen c e l l us u s e d , and p r o v i d e d t h e source m a t e r l a l i s made homogeneous and i s i n a s o l i d ( r a t h e r t h a n g r a n u l a r or powered) form. D i f f u s i o n of t h e a l l o y c o m p o s i t i o n can t h e n become s u f f l c i e n t l y s m a l l compared to the rate of removal of the surface material that the solid never reaches equilibrium with the vapor. Electron microprobe analysis (28) of an angle-lapped (5°) 8-Bm film grown on a PbTe substrate at approximately 38OOC from a Pbo.88Sno.12Te alloy source showed a constant composition throughout the film of O.120 ~ 0.OO1SnTe** At the interface of the film with the substrate, the profile shown in Ylg. 1 was obtained. The errors in SnTe composition are based on the statistical scatter obtained in four different traces through the region. S t r a u s s (29) has r e p o r t e d l a r g e r v a r i a t i o n s i n f i l m c o m p o s i t i o n i n growth of PbSnSe f i l m s from p s e u d o b i n a r y a l l o y s . However, t h e s o u r c e i n g o t s were n o t s u b j e c t e d t o a homogenizing h l g h - t e m p e r a t u r e a n n e a l .
52
H. Holloway and J. N. Walpole
The error in distance of + 0.5 Bm is an estimate of the minimum error due to penetration of the microprobe-into the angle-lapped surface. Larger errors are likely. The lapping produces some smearing of the probed surface which may lead to additional errors. In any case, the transition region appears to be < IBm. Since this film was grown at IBm hr-Y, t h e 8 - h r growth time represents an unusually long period and the diffusion of the Pb-Sn composition should be the maximum normally encountered.* 14 13 12 i
I
•
~
I
ll ¢
10
C
8
O L ¢
o_
,
9
7 6
o
5
E
4
0 <
~
,
F-C CO
,
3 2 I
ar
|.
1 B
I
I
i
I
I
2
3
4
Oistonce Fig. i.
(micrometers)
Variation of alloy composition with distance near the interface of a Pbo.88Sno.12Te layer on a PbTe substrate as determined by electron microprobe analysis on an angle-lapped (5o) surface.. The uncertainty in distance is the minimum estimated error. The absolute SnTe composition by =2% although relative changes can be more accurately determined as indicated by the vertical error bars.
The vacuum system used for MBE of PbSnTe on PbTe substrates is shown schematically in Fig. 2. Ion pumping with a system pressure during growth of about 10-7 Torr was used. The substrate was radlently heated in a small furnace open at one end. The rate of deposition was controlled by a feedback signal to the source-heater power supply. The feedback signal was generated from the deposition rate onto a quartz-crystal film-thickness monitor. Rates of 1-3 Bm hr -I were used with substrate temperatures of 350 - 45OOC. The sources were contained in open pyrolytic boron nitride crucibles heated in a tantalum furnace. The distance between the sources and the substrate was about 20 cm. Separate sources were used for each type of layer grown. Each source was prepared form zone-melted (32) binary compounds Bicknell (30,31) has used X-ray techniques to investigate grading in PbSnTe films grown by a vapor epltaxy technique at higher substrate temperature (5OO-600oc). He finds somewhat more Pb-Sn composition diffusion.
MBE Techniques for IV-VI Optoelectronlc Devices
53
(PbTe and SnTe) weighed out in the proportions required to obtain the desired alloy composition.
IThermocoupl,e
11
~L
t
I
I
I1:!1-
'
illl
ThicknmessI oo.o,
,, Deposi Ratetion J Controller
I Gold/Quartz I Radiant Heater (vocuum" lo-TTorr)
I Power I Supply ,
~ ~-~ron Nitride
i~ Fig. 2.
I
Crucible
Schematic diagram of vacuum deposition system for epitaxial growth from PbSnTe alloy sources on PbTe substrates.
Before use, the sources were homogenized by sealing the constituents in an evacuated quartz ampoule and heating above the melting point for several hours followed by rapid quenching in water to room temperature. Subsequently, the ingot was annealed for 5 days below the melting point at 700oc and quenched again. Pieces of the polycrystalline ingot were then broken off and used as solid chunks in the source furnaces. When undopedsources were used, very low concentration films were obtained determined by waveguiding experiments discussed below. Foreign impurities used to control carrier type and concentration rather than controlling the from stoichlometry. Impurities are expected to diffuse much less than the from stoichiometry (33).
as were deviation deviation
For n-type films, Bi was added in controlled amounts and for p-type films TISe was added. Figure 3 shows the carrier concentration in the film compared to the concentration of dopant in the source. Above about 5 X 1018 cm -3 the film concentration appears to saturate for both types of dopants. The film concentration was measured using the carrier concentration dependence (34) of the infrared reflectlvlty minimum at the plasma edge. The u s e o f s e p a r a t e e f f u s i o n s o u r c e s f o r t h e c o m p o n e n t s o f p s e u d o b l n a r y a l l o y s * may a l s o b.e t r o u b l e s o m e b e c a u s e t h e r a t i o o f t h e c o m p o n e n t f l u x e s m u s t b e p r e c i s e l y controlled, I f o n e a s s u m e s a n e f f u s i o n c e l l t h a t o p e r a t e s a t a b o u t I 0 0 0 K, a t y p i c a l tmnperatur~ fluctuation o f + i K t o g e t h e r w i t h a s u b l i m a t l o n e n t h a p y o f a b o u t 50 kcal mole ~a gives a fluctuation ~ 2.4% i n the effusion rate. With two independent
54
H. Holloway and J. N. Walpole
sources the overall fluctuation in the ratio of the components is then + 5%. This is unacceptable in cases where the energy gap is a sensitive function of composition.
SUBSTRATETEMPERATURE425 °C GROWTH RATE 3p.m/hr Pbo.88Sno.12TeLAYERS
I
'E oo
.J~
[3
TISe r 7 ~ / ~
z
8i
U
0
Fig. 3.
If
I
2
I
I
I
I
I
I
I
1
4 6 8 I0 12 14 16 18 20 CONCENTRATIONIN SOURCE ( 1018cm-3)
Calibration curves of Bi and TISe doping efficiency in Pbo.88Sno.12Te films.
Composition fluctuations may be reduced by using a pair of thermally-linked cells (36) The idea is, that once the temperature fluctuations of the two cells have been minimized, their effect upon the composition can be further reduced by having them occur in phase. A change in the effusion rate of one of the components is then partially offset by a corresponding change in the effusion rate of the other component. For small temperature fluctuations the fractional change in the effusion rate of one of the components may be expressed as
J
J
~T RT 2
It is also possible to use separate effusion sources for the elemental constituents (35), but this seems unnecessarily complicated because, apart from the increase in the number of cources, the large range of volatility of the constituents makes radiation shielding between the sources necessary. (E.g. for growth of (Pb,Sn)Te the equilibrium vapor pressures of the constituent elements at IOO0-K are: Pb, 10-2 Torr; Sn, 5 x 10-8 Tort; and Te, 2 Tort. In contrast, the binary compounds PbTe and SnTe both have vapor pressures of about 0.i Tort at IO00-K).
MBE Techniques for IV-Vl Optoelectronic Devices
55
where AH s is the enthalpy of sublimation. For a pair of effusion cells that operate at the same temperature, but with independent random temperature fluctuations, the maximum change in the flux ratio is:
A • Jl
Jl
~22
'~ ~ + - -J2
max
(AHs, i+ AHS~2)
AT,
RT 2
where AH 8 1 and AH, ~ are the enthalpies of sublimation of the two components. If the temperature fluctuations are now constrained to occur in phase, the maxlmum change in the flux ratio is reduced to:
Jl
Jl
'b + - -
J2
"~ - J2
(AHsrl - AHs~2) AT. RT 2
max
The magnitude fo the effect may be estimated by again considering the deposition of Pbl_xSnxTe with Eg ~ 0.11 eV. The enthalpies of sublimation of PbTe and SnTe are 52.3 kcal mole -l, respectively. With a pair of effusion sources that operate at 1000 K with independent random fluctations of i K, the flux ratio can change by + 5 % giving Ax - + 0.01 and E 2 - + 6 meV. In contrast with in-phase temperature fluctuations, the fl~x ratio cha~ges-by about* ~ 0.3% to give Ax ~ ~ 6 X I0 -~ and AEg ~ + 0.4 meV. In practice, the temperature control can be improved to about 0.2 K with a corresponding further improvement in homogeneity. Schmatics of the deposition system and the double cell are shown in Figs. 4 and 5.
3.
THE CRYSTAL PERFECTION OF I V - V I EPITAXIAL LAYERS
The work described in this review has made use of IV-VI semiconductor substrates, principally for work with lasers, and insulating substrates, principally for work with photodiodes. There is relatively little prior information on the properties of the layers on IV-VI substrates, largely because good single crystal substrates are not readily available.** With IV-VI substrates, it is possible in principle to achieve a match in the lattice constants of the substrates and epitaxlal film. This may be technologically important as a means of reducing the density of the misfit dislocations that occur in such IV-VI structures (37,39). Figure 6 shows the lattice constant vs energy gap for the lead-tin chalcogenlde compounds and alloys of most interest. The lines representing the ternary alloys have been drawn as straight lines between the points representing the binary compounds and hence are only approximate. In a study of heteroepitaxy of Pbl-xSnxTe on {1OO} substrates of different alloy composition using both MBE and LPE (40) it was found that etchpit densities in layers several ~m in thickness were invariably greater than 107 cm -2 even for differences in tin composition as small as 2%. Homoepitaxy or nearly latticematched heteroepitaxy of *
There is a significant uncertainty here because the possible errors in the sublimation enthalpies are comparable to the 6 kcal value for their difference.
** The unsuitability of these conducting substrates for the transport measurements which have domlnated such work may also have had an influence.
•~
H. Holloway and J. N. Walpole
Bell jar
r
/
'/
Substrate Holder ~
S.S Mask - ~ Graphite Double Cell
Quartz tube Holder
\Tantalum
/
Heaters
/
Thermocoq
Ceramic SuPl~ Rods ( 2 )
Fig. 4.
Schematic diagram of a vacuum deposltion system for epitaxial growth of IV-VI semiconductors on fluorite structured substrates.
SUBSTRATE EFFUSION DOUBLE CELL GRAPHITE CASING QUARTZ OR ~ F QUARTZ OR CERAMIC PINS \ ~ E R A M I C RINGS
/ TANTALUM HEATER Scole: Double Size Fig. 5.
Isothermal double cell for evaporation of IV-VI pseudobinary alloys.
Pbo.88Sno.12Te on PbTeo.952Seo.048 (or vice versa) yielded etch-it densities
MBE Techniques for IV-Vl Optoelectronic Devices
57
comparable to that of the substrates used, < IO 4 cm- 2 . Lattice-matched heteoroepitaxy of PbSnSe on PbSSe has not been reported but appears particularly attractive since both these alloys have good metallurgical and mechanical properties especially compared to PbTeSe. To date, no optoelectronlc devices have been reported which were grown using lattice-matched heteroepitaxy. However, the device results to be discussed suggest that misfit dislocations may have little effect on minority carrier liftime in films* and there is evidence that recombination at hetero-lnterfaces can be quite small.
0.30~-- ~PbS
1:
°2°'J-s,
E
0.20 --
(.~ 0 Z W
BOUNDARY ~ OF SINGLE-PHASE ! ROCK-SALT e TURE
o.15> a. ,It
0.10
o_ or. o
--
15
f, Z
_o
/
3¢ I",.. I",,. "'
T; '°.,2s°°.,,
30
no.18Se
W 0 "it
ID INVERSION Pb_ __Sn___1
0.05
-- 1
' I
0.10i --
t
5.9
6.0
15
I
/
0.1~--
Fig. 6.
Z W -J W
I Sns'
6.1 6.2 6.3 LATTICE CONSTANT (AI
6.4
6.5
Variation of energy gap at 77 K (wavelength of emmission or absorption edge) with latlce constant for the lead-tin calcogenlde alloys.
Bickne11 (41) has shown t h a t m i s f i t d i s l o c a t i o n s i n vapor e p i t a x i a l PbSnTe have no e f f e c t on m i n o r i t y c a r r i e r d l s s u i o n l e n g t h .
layers in
58
H. Holloway and J. N. Walpole
•n contrast to the situation with IV-VI substrates, the growth of IV-Vl layers on insulating substrates has a substantial literature (15) that is mostly devoted to growth on alkali halide substrates, although occasional use has been made of other materials, such as CaF 2 and mica. This application of such semiconducting thin films to p-~ junction device technology demands a more critical approach to crystalline perfection than has been customary in the literature on epitaxial growth. Determination of the degree of epitaxy in IV-VI layers has been mostly by diffraction techniques that are insensitive to the spread in orientation that is associated with a small-grained mosaic structure. An early study (42) of PbS on NaCI demonstrated the presence of a mosaic structure with grain sizes of the order of I000 . Similar results were obtained later with PbTe on NaCI (43) which was found to have about 5 x iO I0 dislocation/ cm 2 that were mostly arranged to form grain boundaries with about 2000 spacing and about 2° misorientation. An electron mlcrograph that reveals the typical grainy structure of a PbTe layer on an alkali halide substrate is shown in Fig. 7.
Fig. 7.
Parlodlon replica of epitaxial PbTe grown on cleaved KCI. The region shown is about 40 ~m wide and the darker regions are PbTe grains that have been pulled out of the film during replication.
Despite the early indications of a mosaic structure in epitaxial IV-VI semiconductors on alkali halide substrates, much of the subsequent work ignored the crystal perfection of the layers and, consequently, has little relevance to p-n junction device applications. The situation resembles that in the early stages of development of epitaxial III-V semiconductors, such as GaAs, and earlier sonlnents (44) about the inadequacy of conventional diffration techniques for assessment of the crystal perfection of II-V layers apply with equal force to the IV-VI semiconductors. Such problems of characterization appear to be widespread with epitaxial semiconductors.
MBE Techniques for IV-Vl Optoelectronlc Devices
59
A commonly r e p o r t e d p r o p e r t y of IVIVI e p i t a x l a l l a y e r s i s the H a l l m o b i l i t y . Layers t h a t have been p r e p a r e d under a wide r a n g e of c o n d i t i o n s on v a r i o u s s u b s t r a t e s tend to e x h i b i t b u l k l i k e v a l u e s o f the H a l l m o b i l i t y a t 300 K and, t o a l e s s e r e x t e n t , at temperatures down t o 77 K. However the scattering lengths in these materials are too small for the carrier mobilities to give significant information about the crystal perfection. In the temperature range 300-77 K, the largest mobilltles that are observed with the lead chalcogenides follow,
~H ~
T-a' a ~ 5/2,
and this relationship has been interpreted in terms of scattering by phonons (45). At lower temperatures the mobility tends to saturate at values that have been attributed to scattering by ionized native defects (46). For the lead chalcogenldes, the phono-limited mobilities of both p-and n-type materials are about 2.5 x 104 cm2v-lsec -1 at 77 K. For typical carrier concentrations in the range 1016 - 1018 em -3 this corresponds to scattering lengths of only about 2000 . Thus, the 77 K Hall mobility cannot give information about departures form perfect crystallinity on a scale greater than a few thousand . Even at 4 K, where the Hall mobility of PbTe may be of the order of 106 cm2v-lsec -1, the scattering lengths still do not exceed a few inn. The utility of the mobility as a c r i t e r i o n f o r the p e r f e c t i o n of e p i t a x i a l l a y e r s i s f u r t h e r reduced by the f a c t t h a t even good d e v i c e - q u a l i t y b u l k c r y s t a l s f r e q u e n t l y f a l l to a t t a i n the f u l l p h o n o n - l i m i t e d m o b i l i t y a t 77 K. Thus, h i g h performance b u l k c r y s t a l PbTe p h o t o diodes have been made f r o m ~ - t y p e m a t e r l a l w i t h WB(77 K) - 1.0 x 104 cm2v-lsec -1 (47) r a t h e r t h a n the phonon l i m i t e d v a l u e . These d e p a r t u r e s form the p h o n o n - l i m i t e d m o b i l i t y appear to c o r r e l a t e w i t h the t h e r m a l h i s t o r y of the specimen and i t has been s u g g e s t e d t h a t they a r i s e from v a r y i n g degrees of i o n i z e d i m p u r t i y s c a t t e r i n g by complimentary p a i r s of n a t i v e d e f e c t s ( 4 6 ) . S i m i l a r e f f e c t s have been observed w i t h e p l t a x l a l PbTe and Pbo.gSno.2Te. In the evaluation of IV-VI semiconductors on insulating substrates as materials for p-~ junction devices it is important to verify that the layers are truly singlecrystalline, rather than the fine-gralned mosaics with some degree of epitaxial alignment that are commonly designated "slngle-crystal films". Beyond this point, characterization is made difficult by a lack of quantitative relationships between imperfections, such as low-angle grainboundaries, and the properties of p-n junctions, which are more easily related to such parmeters as carrier lifetime and diffusion length. Thus, at present the ultimate test of an epitaxial layer is provided by comparing the properties of a p-n junction in it with those of a similar junction in a bulk crystal of the same semiconductor. Epitaxial growth of Pbo.gSno.2Te on BaF 2 was first reported in 1970 (48). Subsequent work (49) showed that thse and similar layers of PbTe were remarkably free of lowangle grain boundaries. Figure 8 shows an electron micrograph of a replica of a PbTe layer grown on cleaved BaF 2. The most evident features are triangular pits that appear to arise at the intersection of dislocations with the layer surface.* The dislocation densities so estimated can vary quite widely from nearly 108 cm-2 to slgniflcantly less than 106 cm -2. Occasional low angle grain boundaries, as shown in Fig. 8, appear to arise by propagation of boundaries in the substrate. The typical mlsorientations across these isolated boundarles, as estimated from the dislocation spacing, are a few minutes arc and correspond with those found in some The dislocation plts may not occur if the substrate temperature is too low of the growth rate is too large. We have observed well-developed pits with growth at 1-2 tnnhr -I with substrate temperatures in the range 3OO-4OOOC.
6O
H. Holloway and J. N. Walpole
of the substrates.
Fig. 8.
Parlodion replica of epitaxlal PbTe grown on cleaved BaF 2. The region shown is about 40 um wide and the llne of dislocation pits corresponds to a tilt of a few minutes arc.
Table I gives a comparison of some properties of the fluorite structured and the rock-salt structured substrates. Adequate lattice matches with the lead chalcogenides are obtained with both substrate types. The {Iii} cleavage of the fluorite structures leads to an unfamiliar growth habit for the rock-salt structured semiconductors, but this appears not to give problems provided that the vacuum system is sufficiently clean, (Insufficient care sometimes leads to the more familiar {I00} habit as a fibre texture). The fluorite structured substrates do give a superior thermal expansion match and this initially prompted the trial of BaF 2 substrates in an attempt to reduce damage when the layers were taken from a growth temperature near 700 K to an operating temperature near 80 K. However~ the thermal expansion mismatch does not appear to provide an explanation for the relatively grainy films that are obtained with alkali halide substrates. A significant feature may be the hydroscopic nature of the alkali halide surfaces. In contrast, cleavage surfaces of the fluorite homologues are stable in moist air and their low volatilities permit vacuum bake-out at relatively high temperatures (up to at leasts 800 K) to allow clean-up of air-cleaved surfaces. An unusual feature of the IV-VI layers on fluorite structured substates * is that epltaxy occurs with the substrate and deposit in a twinned rather than a parallel relationship, i.e. growth on the {Ill} BaF 2 surface gives a PbTe deposit that is rotated 180 ° about the {iii} axis relative to the substrate (51,52). The result
61
MBE Techniques for IV-VI Optoelectronic Devices may be interpreted in terms of an interface structure of the type
ABsB7
8 CsyA
8
C uByA8
C
PbTe
B~ r
where the stacking symbols (A,B,C for metals; a,8,7 for non-metals) have their usual significance. A detailed analysis shows that alternative interracial structures are less favoured both electrostatically and sterically (53). In general, Hall measurements of IV-VI epitaxial films on fluorite structured substrates give results that differ little from those reported for the best hulk single crystals. As discussed above, the attainment of large Hall mobilities is not an adequate criterion for the growth fo device-qualitymaterial. However, it is of interest that the low-temperature mobilites obtained with BaF 2 and SrF 2 substrates are significantly larger than those that have been reported for layers on alkali halide substrates. A typical example is the result for Pbs (54) that is shown in Fig. 9. For both n- and p-type layers the films give temperaturedependent mobilities that agree well with the best results obtained with bulk PbS. Surpirsingly, the mobility of the n-type layers is not significantly affected by growth from Pb-rich fluxes that appear to give Pb precipitates in the layers. A similar insensitivity to precipitates has been observed with n-type PbSe layers that were grown without a subsidiary Se source. Table I.
Compound
Properties of the lead chalcogenides and some substrates
Lattice Constant (
)
~ (K-I x 106 ) near 300 K
PbS
5.94
20
PbSe
6.12
19
PbTe
6.46
20
Cleavage
P(Torr) at 700 K
NaCI
5.64
39
{I00}
4.5 x 10 -7
NaBr
5.96
42
{i00}
3.7 x 10 -6
NaI
6.46
45
{100}
4.0 x IO-5
KCl
6.29
37
{i00}
2.2 x i0-6
KBr
6.59
38
{I00}
I.i x 10-5
K1
7.05
40
{100}
3.9 x 10-5
CaF2
5.40
19
{111}
~ 6 x 10 -20
SrF 2
5.80
18
{Iii}
~ 1 x 10-20
BaF 2
6.20
18
{111}
~ 3 x 10 -17
This applies to BaF 2 and SrF 2. While lattice match is not particularly critical, it appears that CaF 2 has too large a mismatch for successful epitaxy of the IV-Vl semiconductors. Studies in the Ford Laboratory have failed to give useful layers on CaF 2 substrates and this is in accord with published results for (Pb,Sn) Te on CaF 2 (50).
62
H. Holloway and J. N. Walpole
I05
,
I
III
I
!
,
,
I
..ik . . . . P - Z . z x K ) " cm -3
!
= J
I
"1
(AIIgaier& Sconion)
104 ~
~H
(cm-2V-I~.-i)
1031 300 Fig. 9.
llll i I00 T (OK)
30
I0
Hall mobilities of PbS layers grown on cleaved SrF 2 substrates. The curve shows the largest mobility that has been reported for a bulk crystal of PbS (Ref. 22).
With PbTe and (Pb,Sn)Te layers on BaF2, increases in the mobility have been obtained by annealing the films at relatively low temperatures (typically 3OO-4OO°C for 12-15 hr) (49). The increase was in the more-or-less temperature-independent saturation mobility that was attained at lower temperatures, rather than in the phonon-limited mobility, for which ~H ~ T-a with a ~ 5/2. The annealing results were interpreted in terms of a reduction in the ionized defect scattering that accompanies the pairwise recombination of native donors and acceptors. It was
sation is not invariably present in the layers. Thus, Fig.lO shows Hall data (54) for an as-grown 2.7 ~m thick n-type PbTe layer on BaF 2 that attains a low-temperature mobility of 5 x IO s cm2v-lsec-1. This appears to be the largest mobility that has been reliabily established for a IV-VI epitaxial layer.* +
Lo~ez-Otero (55) has reported low-temperature Hall mobilities of up to 2.5 x IO 6 cm-v-lsec -I in PbTe layers grown on BaY 2 and suggests that this result reflects the superiority of the hot-wall method over other deposition techniques. However, these results must be regarded as questionable because the large mobilities at
MBE Techniques for IV-Vl Optoelectronic Devices
63
13 K were obtained with specimens whose (56) mobilites at 77 K exceeded the commonly observed phonon-llmlted value by about a factor of two. Similar results with bulk single crystals of (Pb,Sn)Te (57) have been interpreted as a consequence o f i n h o m o g e n e i t y i n t h e specimens ( 5 8 ) .
106'U
I
'
I'I'I'
I
'
I
i, •0
_Speclmeh EW 271 105
.
-
i
~H.
J 0 m
B
i I
>
E U
i
:x: 10 4
:k
I B
•
-
o
-
I-
ro
-
_J ID
E v
0
--
T rr
m
.,J _J r
RoJUJo o le
103
•
o*
•
10 2
• o o° • ---
-
I
--
i
b-"
o ~
_~ t4. U. t~
8
•
_1
2xlO 2
I , 400
I 200
I,1,1 I I I00 80 60 40
,
I 20
Ii ~8
=
TEMPERATURE, T (*K) Fig. i0.
4.
Hall coefficient and Hall mobility of a 2.7 ~m-thick n-type PbTe layer on a BaF 2 substrate.
PHOTODIODE PERFORMANCE AS A CRITERION FOR CRYSTAL QUALITY
As pointed out earlier, the ultimate test of a semlconductlng material arises when one attempts to make devices with it. For discussion of the results that have been obtained with thin-film photodiodes it is convenient to consider the detectivity (D*), which is a standard figure of merit that is defined as a signal-to-noise ratio that has been noralized to make it independent of the incident power, the noise bandwidth, and the detector area. With, a w e l l - b e h a v e d p h o t o d i o d e a t z e r o b i a s i n t h e a b s e n c e o f s i g n i f i c a n t b l ack b o d y r a d i a t i o n form t h e background, t h e s i g n a l c u r r e n t depends upon t h e quantum e f f i c i e n c y (n) and t h e n o i s e c u r r e n t depends upon t h e J o hn so n n o i s e o f t h e j u n c t i o n r e s i s t a n c e . Under t h e s e c o n d i t i o n s we have
J
64
H. Holloway and J. N. Walpole
D* (Johnson) = n Ey
ReA ~ 4K2
where Ey is the photon energy. In most cases, the quantum efficiency of the photodiode is limited only by reflection losses and has a value in the range 0.4 - 1.0. The D* then d~pends on the zero-blas resistance area product (RoA) of the junction. This is a sensitive measure of the quality of the material because the junction resistance is inversely proportional to the saturation current, which may be substantially increased by defects that decrease the lifetime in the film. With large enough values of the resistance-area product, the Johnson noise current becomes smaller than the shot noise due to fluctuations in the rate of arrival of photons from the balckbody background. Under these conditions the D* becomes background limited with a value
D* (background) = I
~
n
'2'QB
where QB is the background photon flux in the energy range for which the photodiode is sensitive. The background-llmited D* depends upon the field of view (FOV) of the detector and also quite strongly upon the cut-off wavelength since this influences the value of QB. Table 2 shows some representative background-limited detectivities together with the reslstance-area products that must be exceeded for background limited operatuion with 180o FOV. Thus, the demonstration of background-limited D* at 180 ° FOV and at some convenient operating temperature (e.g. 77 K) is a convenient criterion for the attainment of material that is good enough to be useful. Further deCisions about quality may be based upon the extent to which the D* increases when the FOV is decreased (i.e. the extent to which RoA exceeds the value needed for background-limited operation at 180 ° FOV). Table 2.
Background-Limited Detectivities for Infrared Photodiodes %
Cut-off wavelength
D* (Background)
(~m)
(cmHz~W-I)
3 4
Equivalent RoA
(ohm em2)
1.2 x 1012
3900
2.6 x lO ll
115 15
5
1.2 x lO II
6
7.5 x I0 lO
4.1
8
4.6 x I0 I0
0.88
i0
3.8 x 1010
0.38
12
3.5 x I0 I0
0.22
t These detectivities are calculated for 180 ° FOV a t a wavelength near the cut-off with a typical reflection-loss-linLited quantum efficiency of 0.5. The corresponding values of RoA are those which, at 80 K, give equal contributions from the Johnson noise and ~he background noise. (This would give detectivities that are reduced by a factor of / ~ f r o m the background-limited values.)
~BE Techniques for IV-Vl Optoelectronic Devices 5.
THIN-FILM VI-VI SEMICONDUCTOR PHOTODIODES
While the choice of vacuum deposition technique for thin-film IV-VI semiconductors may be regarded as a matter of taste, the choice of insulating substrates for p-n junction applications is not. With one exception (discussed below) the very extensive body of work with alkali halide substrates has failed to yield p-n junction devices. In contrast, the fluorite structured substrate BaF 2 has given p-n junction devices whose performance is competitive with that of the best bulk crystal devices. Moreover, following the original work with PbTe on BaF 2 at the Ford Laboratory, similar devices have been obtained with a range of IV-VI semiconductors by the Ford group and by several other groups of workers who have used a variety of vacuum deposition techniques. Thus, the key feature is the choice of suhstrate and the successful development of a p-n junction technology may be attributed to the improvement in the crystal perfection that is described above. The sole example of significant performance from p-n junction devices made from IV-VI layers grown on alkali halide substrates appears to be the work of Schoolar who worked with PbS on NaCI. For junction formation it was necessary to dissolve away the substrate and use the PbS surface that had been in contact with the NaCI. PbS juncitons that were made by growing p-type PbS on n-type.epitaxial PbS layers gave quantum efficiencies around O.I and D*(4 ~m) = 109 cmHzIW -I when operated at 77 K (59). Subsequent work (60) with semltranspartent In metal barriers gave D * (3.8 Bm) ffi 1.5 x 1011 cmHzIW -I, which is only a factor of two below the background limit (assuming 180 ° FOV and a typical reflection-loss-limited quantum efficiency of 0.5). With the observation of high-quallty layers of PbTe and (Pb,Sn)Te on BaF 2 substrates (49) it became obvlouse to try farlcation of p-n junction devices with these materials The first high-performance thln-film IV-VI photodlodes and the first use of the metal barrier technique for IV-VI infrared photodlodes were reported by the Ford group in 1971 (61). Barrier layers of small-work-functlon metals, such as Pb, had been used previously by Nill et al. (62) to make bulk crystal lasers by electrostatically inverting the surfaces of p-type PbTe and (Pb,Sn)Te. Infrared response had been observed from the edges of these devices (63), but optical absorption in the barrier layer had precluded application as photodiodes.* The energy bands of a thin-film PbTe metal barrier photodiode, following the analysis by Walpole and Nill (65), are shown in Fig. II.** The BaF 2 substate is conveniently transparent for wavelengths up to about 12 ~m. This permits illumination of the junction via the substrate. Taking typical optical absorption lengths of the order of i um and typical minority carrier diffusion lengths of the order of I0 ~m, it is evident that a semiconductor layer that is a few ~m thick can give both good optical absorption and efficient collection of photogeneratedminority carriers.
The o r i g i n a l c o n f i g u r a t i o n f o r t h i n - f i l m PbTe p h o t o d i o d e s (61) was t h e simple c r o s s e d - s t r i p e a r r a n g e m e n t t h a t i s shown i n F i g . 12. When o p e r a t e d a t 77 K and 180 ° FOV, t h e s e d e v i c e s gave b a c k g r o u n d - l i m i t e d D * ' s . With r e d u c t i o n of t h e For t h e r e was an increase by a factor of four to a Johnson noise limit of D*(5 ~m) = 6 x 1011 Subsequent work (64)has shown that semitransparent In barriers, similar to those reported by Schoolar, may be used to make hlgh-performance bulk crystal Pbo. 8 Sno.2Te photodlodes.
** I n most cases we l a c k an u n e q u i v o c a l d e m o n s t r a t i o n t h a t t h e t h l n - f i l m p - ~ j u n c t i o n ~ a r e formed by e l e c t r o s t a t i c i n v e r s i o n , r a t h e r t h a n by s h a l l o w d i f f u s i o n of Pb i n t o , or c h a l c o g e n o u t o f , the s e m i c o n d u c t o r . D e s p i t e t h i s u n c e r t a i n t y the j u n c t i o n forming t e c h n i q u e i s of c o n s l d e r a b l e p r a c t i c a l i m p o r t a n c e .
B. Holloway and J. N. Walpole
o°il pb i
ENERGY ('v-az1[
b
ocm3
c8 vs
iF_ ........................
-0.4f
-0.6
-0.8
0 Fig. ii.
0.2 I
DEPTH (p.m)
0.4
Calculated band bending due to a Pb layer on p-type PbTe (p ~ 1017 cm-3).
cmHz½W -I. Subsequent work (54) has given Johnson noise limits as large as 1013 c ~ z ~ W -I for such thin-film PbTe devices at 77 K. Minor modifications of the crossed-stripe geometry gave the first demonstration of a thln-film injection leaser on an insulating substrate (66) (discussed elsewhere in this review) and also a junction field-effect transistor (67,68). Further studies showed that the performance of the PbTe photodiodes did not depend upon the technique that was used for junciton formation. Essentially similar detectivlties were obtained with p-n junctions that were made by proton bombardment (69) and by Sb ion implantation (70) of p-type PbTe layers. * Studies of the stability of the Pb-barrier devices have shown that their performance is retained after repeated cycling to cryogenic temperatures and after more than sufficient baking to outgas vacuum enclosures (~ I0 hr at 150°C). Further development from the original thin-film PbTe photodiodes has followed two main lines. First, the use of IV-VI alloys for specific applications in the 3-5 Bm atmospheric window and for reduction of the energy gap to permit operation in the 8-12 ~m atmospheric window. Secondly, refinement of the rather crude crossed-stripe configuaration to permit the delineation of and connection to arrays of small closel~ * To date only limited success has been reported with conventional diffusion techniques and with attempts to make p-n junctions by changing the conductivity type of t~e deposits during film growth. Callender (76) reported D*(IO Bm) = 4 x 109 C,LHz=W"I at 77 K for Pbo.82Sno.18Te devices that were made with a mesa technique applied to 15 ~m-thick layers on BaF 2. Lopez-Otero et a~. (72) reported D* = 1.4 x IO I0 c ~ z ½ W -I at 77 K for grown-ln p-'n junctions in epitaxial PbTe on BaF 2. In both cases the D* was an order of magnitude less than the background-limited value that has been exceeded by metal-battler devices.
MBE Techniques
for IV-Vl Optoelectronic Devices
67
spaced detector elements.
IR DETECTOR
FET
Pb
Pb
LASER Fig. 12.
C r o s s e d - s t r i p e c o n f i g u r a t i o n s f o r Pb b a r r i e r PbTe devices.
The major advance in thin-film 3-5 Bm detectors has been the development of devices whose operating temperature has been increased from 77 K to the intermediate temperature range, 170-2OO K, that is suitable for thermoelectric cooling. This permits application in lightweight hand-carrled thermal imaging systems. These increased operating temperatures greatly increase the requirements for device quality because of the exponential decrease of the junction resistance with temperature which reduces the Johnson-nolse-limited D*. Figure 13 shows the temperature-dependent spectralD* of a thin-film PbTe photodiode (73,74) operated with 180° FOV. The curves show several interesting features. First, there is a sequence of pronounced maxima and minima that arise from interference modulation of the quantum efficiency about the thlck-film limit of 0.6. This effect may be used to obtain quantum efflclencles as large as 0.9 for selected wavelengths. Secondly, as the oerpatlng temperature is increased from 80 K, the D* first increases and then decreases. This effect is due to a change in the dominant noise mechanism. At low temperatures the juncition resistance is sufficiently large that the Johnson noise is much smaller than the background noise. With increased temperature the increase in the energy gap shifts the cut-off to smaller wavelengths and reduces the effective background photon flux thereby increasing the background-limlted D*. At large enough temperatures the junction resistance decreases to a point where the device becomes Johnson noise limited. Further increase in temperature then reveals the temperature-dependance of this Johnsonnolse-llmitedD*.
68
H. Holloway and J. N. Walpole
1012
,
,
,
,
,
,
,
,
,
h
84 K
_ITOK.'~ ~ /
•r
-
°
"~.-
190 K~ , ~ . ~
?
~..
EW 3 3 5 - 4d
,o° Fig. 13.
2
3
4 X (Fm)
!I!l 5
6
Temperature dependence of the spectral detectivity of a thin-film PbTe photodiode.
The practically significant feature of Fig. 13 is the attainment of useful D*'s at temperatures that are suitable for thermoelectric cooling t For many 3-5 ~m applications the response of PbTe extends to long enough wavelengths. However, at 170 K the cut-off at about 4.8 ~m does not permit full use of the 3.5 ~m atmospheric window. This led to consideration of pseudobinary alloys of the Pb chalcogenides that are shown in Table 3. Table 3.
Optical Absorption Edges of Lead Chalcogenides
Material
Optical absorption edge (~m) at 77 K
at 170 K
PbS
3.7
3.3
PbSe
6.9
5.6
PbTe
5.7
4.8
t These results have been confirmed in part by McMahon (75).
MBE Techniques for IV-Vl Optoelectronic Devices
69
Following the demonstration of high performance t h l n - f i l m Pb b a r r i e r PbSe photodiodes (76) the film growth techniques were extended to PbSeo.8Teo. 2 (77) with an optical absorption edge near 5.4 ~m at 170 K. This material was used to make the devices whose temperature-dependent spectral D* are shown in Fig. 14 (73,78). Basically, these results are similar to those obtained with PbTe devices, but with the response extended to cover the whole of the 3-5 ~m spectral region. Subsequent work (78) has yielded PbSeo.sTeo. 2 devices with D*(5 ~m) = 1 X i0 II cmHzl/2W-1 when operated at 170 K with 18OO FOV. Under these conditions there are approximately equal contributions from the Johnson noise and the background photon shot noise. This performance appears to exceed that reported for any 3-5 ~m photodiode when operated at intermediate temperature. It is perhaps a measure of the maturity of thln-film photodiode technology that the Pb(Se,Te) pseudoblnary alloy system was not drlved from a bulk crystal application, but was first exploited with the thinfilm devices.
i0tt
170 K
']¢
A%~/I1,
.r
190 K
E S
Q i010 _
210K m~~l~
[
Spec'l'rol de'tectivil"y of PbSeo. s Teo.z ( t= 1.5/J,m)
I
I
2
I
I
3
I
I
I
4
6
I
I
I
8 I0
X(~m) Fig. 14.
Temperature dependence of the spectral dtectivlty of a thln-film PbSeo.sTeo.2 photodlode.
A further appllcatlon of the thln-film technology has been devised by Schoolar et ul. (13) who have developed high performance Pb(S,Se) photodlodes. In this work both sides of the BaF 2 substrates were coated with Pb(S,Se). One of the layers was used to make Pb barrier photodlodes. The layer on the other side of the substrate had a composition that was adjusted to give a sllghtly smaller energy gap and served as a long-wavelength pass filter. By careful adjustment of the compositions the response was reduced to a narrow range (~ 0.2 ~m wide at half maximum) that was defined by the cut-on of the filter and the cut-off of the photodiode.
70
H. Holloway and J. N. Walpole
Another major l l n e of development has been the e x t e n s i o n of t h l n - f i l m photodiode response to the 8-12 ~m atmospheric window by development of techniques f o r growth of Fbl_xSnxTe (x % 0.2) (48) and Pbl_xSnxSe (x ~ 0.07) on BaF 2 (79). Here, use of the (Pb,Sn)Te a l l o y system had many p r e c e d e n t s i n bulk c r y s t a l t e c h n o l o g y , but l i t t l e success had been r e p o r t e d with bulk c r y s t a l (Pb,Sn)Se photodiodes d e s p i t e e a r l y use of t h i s m a t e r i a l to e s t a b l i s h the e x i s t e n c e of the b a n d - c r o s s i n g pbenomenon (29,30). Figure 15 shows the spectra D*'s of two thin-film Pb-barrler (Pb,Sn)Se photodiodes for the 8-12 ~m spectral region (81). At 80 K with 180 ° FOV, such devices are background-llmited with reslstance-erea products up to 2ohm cm 2. With reduction of the FOV, a Johnson noise limit of D*(IO ~m) = 8 x 1010 emHz½W -I has been obtained. These results compare well with peak D*'s of 1 x iO II cmHz}W -I that have been obtained with bulk crystal (Pb,Sn)Te photodiodes at a reduced FOC (82). Approximately background-limlted D*'s near IO ~m have also been obtained with (Pb,Sn)Te In barrier devices (71,83) at 80 K with 180 ° FOV. Schoolar and Jenson (14) have also used (Pb,Sn)Se layers on BaF 2 substrates to make narrow response photodiodes by using the same two-layer technique that was described for Pb(S,Se).
I°"i'
I , I , I , I , I ,
I
,0 9
, 2
Fig. 15.
I, 4
I
I
6
8 X (~m)
I,
i I0
,
, 12
14
Spectral detectivities of (Pb,Sn)Se photodiodes at 80 K.
Development of a processing technology for thln-film IV-Vl photodiode arrays has been accomplished by Asch and co-workers (84) who have delineated the junction areas with windows in an insulating layer of vacuum-deposited BaF 2. Thus, the basic
MBE Techniques for IV-Vl Optoelectronic Devices
71
processes for thin-film array fabrication consist of four successive vacuum deposit ions. These give the semiconductor, the Pb ohmic contact, the BaF 2 insulator, and the Pb barrier, respectively. Mounting of the arrays for illumlnation through the BaF 2 substrates was done with a flip-chip technique that used a composite Cu foil/ glass/Cu header for thermal expansion compatibility. Some details of a thln-film array are shown in Fig. 16.
-'--] BoF Substrate ~ Evaporated8oF~
I
,I
,
Sputtered
Platinum
E~ :voporated Lead Semiconductor
Fig. 16.
Structure of a IV-VI semiconductor thln-film photodlode array.
72
H. Holloway and J. N. Walpole
6.
UNCONVENTIONAL THIN-FILM IV-Vl PHOTODIODES
The preceding section has emphasized the properties of thin-film IV-VI photodiodes whose performance may be compared that of bulk crystal devices to provide an assessment of the perfection of the epitaxial layers. For completeness we now review briefly some more recent developments in which the unique properties of thin films are exploited. One very characteristic property of the thin-film photodiodes is the interference modulation of the quantum efficiency whose influence is evident in Figs. 13-15. A detailed analysis (85) shows that with layers up to about 5 pm thick the quantum efficiency of IV-VI thin-film photodiodes should be limited only by reflection and transmission losses. The good agreement between the observed quantum efficiency of a PbTe metal barrier device and the calculated reflection-loss limit (RLL) is shown in Fig. 17. The result confirms the attainmnet of bulk-llke minority carrier diffusion lengths, and hence lifetimes, in the thin-filmmaterial. Exploitation of the interference effects may take two forms. First, by appropriate choice of the layer thickness the RLL at specific wavelengths may be increased to about 0.9 compared with about 0.5 for an uncoated conventional PbTe photodiode. This eliminates the need for antireflectlon coatings to optimize the detectivlty. Secondly by making use of combinations of IV-VI thin films with dielectric layers and metal reflectors the response of the photodiode may be made to resemble that of a combination of a conventional photodiode with an interference filter (85). Such devices may find application for low-cost evaluation of spectral signatures.
I
I
I
/,.,
I
/ li
#'
4> I
°
2
[
3
i
4
l
5
l
6
7
~.(Fm) Fig. 17.
Comparison of the measured and calculated quantum efficlencles of a thin-film PbTe photodiode at 80 K.
One significant disadvantage of conventional IV-VI semiconductor photodiodes is the
MBE Techniques for IV-Vl Optoelectronic Devices
73
i s the l a r g e j u n c t i o n c a p a c i t a n c e (of the o r d e r of 1 ~Fcm- 2 ) t h a t a r i s e s from t h e l a r g e d i e l e c t r i c c o n s t a n t s of t h e s e m a t e r i a l s . One approach t o c a p a c i t a n c e r e d u c t i o n i s r e d u c t i o n of the dopant c o n c e n t r a t i o n and t h i s has been d e m o n s t r a t e d w i t h b u l k c r y s t a l s of (PbjSn)Te by Andrews e t u Z . ( 1 7 ) . However~ s i n c e the c a p a c i t a n c e of a b r u p t j u n c t i o n o n l y d e c r e a s e s w i t h the s q u a r e r o o t of t h e dopant c o n c e n t r a t i o n , t h i s approach r e q u i r e s the growth of m a t e r i a l w i t h I015cm-3 c a r r i e r s o r l e s s . The t h i n - f i l m d e v i c e s o f f e r an a l t e r n a t i v e approach t h a t i s based on t h e p r o x i m i t y of the depletion region to the insulating substrate (86). The concept is shown in Fig. 18. With a concentional p--n junction a change in the bias causes a change in the width of the depletion region and thereby a change in the charge that is stored in the depletion region. This gives rise to a dynamic capacitance dQ/dF. However, with the thin-film device the p-n junction may be arranged to have the depletion region extend through to the insulating substrate. Under these conditions the depletion region can only accept or give up charge around its periphery. This leads to a reduction in the capacitance by the factor (peripher X layer thickness)/ (junction area). The capacitance reductlonmay be achieved by using layers that are thin enough (~ 0.6 Bm) that become pinched off when the depletion region is widened by back bias. For the latter mode it becomes necessary to reduce the I/f noise that is usually associated with IV-VI photodiodes that are operated in back bias (82). This has been achieved by careful cleaning of the PbTe surface.
ii I
N
I
DEPLETION
SUBSTRATE (a)
SUBSTRATE (b) F i g . 18.
Conceptual arrangement of the pinched-off photodiode. With increasing back bias, the depletion region edge moves to the position shown as a broken line in (a) for a conventional photodlode and in (b) for a plnched-off photodiode.
The t y p i c a l b l a s - d e p e n d e n t performance of a t h i n - f i l m P b T e p i n c h e d o f f p h o t o diode a t 80 K and 180 ° FOV i s shown i n F i g . 19. At zero b i a s t h e diode w i t h a r e a 6 x 10-~ cm2 has a c a p a c l t e n c e of 700 pF. With back b i a s g r e a t e r t h a n a b o u t O.15 v t h e c a p a c i t a n c e d e c r e a s e s t o a c o n s t a n t v a l u e of 70 pF t h a t a r i s e s from t h e c o n t a c t pads r a t h e r t h a n the j u n c t i o n . The 500 K blackbody c u r r e n t r e s p o n s i v i t y ( R I ) , which i s p r o p o r t i o n a l to the quantum e f f l c i e n c y ~ i s i n d e p e n d e n t of t h e b i a s ash t h e n o i s e remains c l o s e to t h a t c a l c u l a t e d f o r f l u c t u a t i o n s i n t h e background f o r back b i a s up to 0.35 v , a f t e r which 1 / f n o l s e becomes s i g n i f i c a n t . Thus, t h e D* r e m a i n s a t the b a c k g r o u n d - l i m i t e d v a l u e f o r b i a s e s t h a t p e r m i t an o r d e r o f magnitude r e d u c t i o n i n the d e t e c t o r c a p a c i t a n c e . L a r g e r a r e a p h o t o d i o d e s have
74
H. Holloway and J. N. Walpole
shown up to two o r d e r s of magnitude r e d u c t i o n i n c a p a c i t a n c e . I t i s of i n t e r e s t t h a t no b u l k c r y s t a l d e v i c e s e x i s t w i t h p r o p e r t i e s t h a t are comparable to the t h i n film prinched-off photodoides.
, o '2
I
1
Rz(SOOK)
I
I
m
]
D*(5.4pm)
N
_~
-I-
E
.~- lOll E ::t.
o
C
H
1
C~
I I
,o,o ,ol 0
Fig. 19.
Z nuise (calculated) I
I I00
I
I
I
200 5 0 0 4 0 0 BACKBIAS (mV)
i
_~ 10-2
500
Bias-dependent performance of an 0.6 vm-thick PbTe photodlode at 80 K and 180 ° FOV. The diode area is 6 x i0- g c m 2. The noise was measured at I kHz with I0 Hz bandwidth and the broken line shows the background noise current that was calculated from the d.c. background current.
An alternative approach to capacitance reduction using thin-film devices (87,89) is shown in Fig. 20. Here the n-region of a conventional photodiode is replaced by a matrix of small circular n-regions that act as collectors for photogenerated minority carriers from the intervening p-region. The insulating substrate acts as a potential barrier that confines the photogenerated carriers to the vicinity of the collectors, thereby reducing losses by recombination. Efficient collection is obtained with a collector spacing of up to two diffusion lengths (20-30 ~m for PbTe). With collector deameters in the range 2-5 Bm the reduction in junction area permits more than an order of magnitude reduction in junction capacitance. Figure 21 shows laser scans of such a lateral-collectlon photodiode (LCP) at different operating temperatures. This shows clearly the increase in collection efficiency as the device is cooled and the diffusion length increases. Under some conditions
MBE Techniques f o r IV-VI O p t o e l e c t r o n i c Devices
75
the saturation current of IV-VI diodes is generated withln the depletion region and in this case the junction resistance of the LCP is greater than that of a conventional device with a corresponding increase in the Johnson-noise-limited D*. This effect has been used to obtain a 35 K increase in the operating temperature
o f PbTe p h o t o d l o d e s . The o n l y c o m p a r a b l e b u l k c r y s t a l work a p p e a r s t o be by N o r e i k a e# aZ. (90) who u s e d c o l l e c t o r s w i t h a s t r i p e g e o m e t r y on b u l k c r y s t a l s o f (Pb,Sn)Te t o o b t a i n c a p a c i t a n c e r e d u c t i o n by a f a c t o r o f t h r e e .
t
n
J
t
n
J
~,e-h + 4 p - type TV'-vr
Ba F2 Substrate
Fig. 20.
Schematic arrangement of a lateral-collection photodiode.
7.
OPTICAL WAVEGUIDES
An optical waveguide structure consisting of a 6 ~m MBE layer of Pbo.92Sn O 08Te grown on a {100} oriented PbTe substrate followed by a 0.5 pm PbTe (cladding) layer as shown in Fig. 22 has been demonstrated (91). Striped guides were formed by growing over a SiO 2 mask which was later lifted off to remove the polycrystalline overgrowth except in the stripes where epitaxial growth occurred. The optical loss coefficient measured in these guides at 10.6 ~m wavelength for light polarized parallel to the plane of the llght-guldlng PbSnTe layer (Te polarization) was 7.8 cm-I at room temperature. Since room temperature free carrier absorption has been well characterized in PbTe (92) and is similar in Pbo.92Sno.o8Te (93), the loss coefficient of 7.8 cm-1 implies an upper limit of about 4 x 1016 cm -3 for the free carrier density of the layer. At 77 K the loss coefficient was too small to measure but was determined to be ~ 1.5 cm-1, the smallest optical loss coefficient measured to date in either bulk or epltaxlal PbSnTematerlal. For the orthogonal polarization (TM), losses were somewhat higher but the increased loss could be attributed to absorption in the metalllzatlon used on the surface of the 0.5 ~m PbTe cladding layer as part of the mountlng procedure for the waveguldes. The transmission scans of a CO 2 laser along the path indicated in Fig. 22 are shown in Fig. 23. The wavegulde transm~sslon is strongly reduced by coupling losses into the narrow guide, compared to transnLisslon through the substrate. The wavegulde results suggest the feasibility of fabricating by MBE monolithic integrated optical circuits incorporating active elements such as lasers, laser az~pllflersp and detectors. The control of wavegulde thickness, which can be tapered for coupling purposes, using shadowing techniques during growth, % is one of the advantages MBE offers for this purpose, in addition to the low as-grown % Tapered layers similar to those reported in GaAs MBE layers (94) can easily be grown in a similar manner in PbSnTe (95).
76
H. Holloway and J. N. Walpole
carrier densities necessary for low loss.
270K
170K
85K
Fig. 21.
Laser scans of a 320 ~m-square PbTe lateral- collection photodiode at 270 K, 170 K, and 85 K. The 5 ~m-dlameter collectors are on 20 ~m centers. The vertical displacement is proportional to the current response and the instrumental resolution is about I0 ~m.
Fig. 22.
Pbl_xSnTe Epilayer Scan Path
Typical Input Face
I 6.5p.m
Schematic of double heterostructure optical waveguides showing the path of scan by a lO.6-~m laser.
PbTe Epilayer
O
5.
t-'° O
I-4
O
= ¢,
O =r'
78
H. Holloway and J. N. Walpole
4.4 4.0
36 Z
0 3.2 o'J 2.8 Z
2.4
I--
2.0
I-. Z
1.6
~" W n
1.2 0.8
0.4 ','1 0,2
I 0.4
,
'
I 0,6
I
I 0.8
I
1.O
12
POSITION (ram) Fig. 23.
Transmission scans of TE and TMpolarized 10.6 ~m laser beams along the path indicated in Fig. 22 for a 4-~m-long waveguide at 77 K. Loss due to coupling is estimated at 95.2 and 92.3% for the TE and TMwaveguide modes, respectively. The guiding layer is located at the 0.2-mmposition.
8.
INJECTION LASERS
Lead-tin chalcogenide injection lasers (laser diodes) have become important research tools for tunable laser spectroscopy (96) in the infrared at wavelengths between about 4 ~m and 30 ~m (see Fig. 6). Their use as pollution monitoring and gas detection devices (97) has been demonstrated includlng their potential for remote heterodyne detection (98-100). The motivation for epltaxlal growth of these devices has been to achieve cw operation at liquid nitrogen temperature and above, rather than at the low temperature (~ 30 K) previously achieved with diffused p-n junctions in bulk-grown material.* An additlonal bonus of higher temperature operation is that, as the energy gap changes with temperature, a single device can be used at different temperatures to cover a large spectral region (400-500 cm -I) (101). Fine tuning of a single mode can be achieved at a given temperature by several means but most easily by varying diode current. The internal change of temperature with current changes the refractive index which shifts the requencies of the longitudinal cavity modes. The continuous tuning range of a mode is typically 1-2 cm-I In addition to the MBE and hot-wall lasers to be discussed, progress has also been made using llquld-phase epitaxy as mentioned, including the first cw operation of a lead-salt laser above 771((19).
MBE Techniques for IV-VI Optoelectronic Devices
79
The u s e o f h e t e r o s t r u c t u r e s i n I I I - V d i o d e l a s e r s t o r e d u c e l a s e r t h r e s h o l d and i n c r e a s e t h e o p e r a t i n g t e m p e r a t u r e s u g g e s t e d t h a t a s i m i l a r approach would be s u c c e s s f u l i n t h e IV-VI m a t e r i a l s ( 1 0 2 ) .
8.1.
Herterostructure lasers
F i g u r e 24 shows a t y p i c a l Pbl_xSnxTe d o u b l e h e t e r o s t r u c t u r e (DH) l a s e r c o n f i g u r a t i o n i n whlch an e p i t a x l a l l a y e r ( t h e a c t i v e r e g i o n ) l i e s b et w een t h e s u b s t r a t e and a n o t h e r l a y e r b o t h o f which have s m a l l e r Sn c o m p o s i t i o n s , l a r g e r b a n d - g a p s , and s m a l l e r r e f r a c t i v e i n d e x e s than t h e a c t i v e r e g i o n . If a p-n junction is also located within or at an edge of the active region, minority carriers injected at forward bias are confined within this region by the potential barriers resulting from the larger energy gaps at the edges. This carrier confinement is one of the features of DH lasers which, in principle, reduces ~ e threshold current density required for laser action provided that high recombination does not occur at the interfaces. The refractive index variation indicated in the figure forms a dielectric slab waveguide to guide (confine) the optical radiation within the active region where the optical gain is located. I n a s i n g l e h e r t o s t r u c t u r e (SH) l a s e r , o p t i c a l and c a r r i e r c o n f i n e m e n t a r e p r o v i d e d by a change i n a l l o y c o m p o s i t i o n on o n l y one s i d e of t h e a c t i v e r e g i o n . On t h e o t h e r s i d e o p t l c a l and c a r r i e r c o n f i n e m e n t a r e p r o v i d e d by a p - n j u n c t i o n o r a g r a d e d doping p r o f i l e as i n h o m o s t r u c t u r e l a s e r s . As w i l l be s e e n , i n t h e IV-VI m a t e r i a l s s t r o n g c o n f i n e m e n t can be o b t a i n e d i n t h i s way.
Electrical and
Thermal Contact
/
V / l / ~ / ~ ~ ~ / ~ ~ ~ ~ 1
]
\
,~:::~::~:::~.::~:~..~,~:.-~.::~.-..:~....::..:~:~..=~.:~,
( Not to scale)
n-Pb0 78 Sn0 22 Te =,,,,,,,,,,
..........
,,,n~,,,,,,,,
..........
~':'::'~;':~'?~";:"L'~"~=t"":':':':::~':T':~':":":~'~:';~L"
II I I mW~mHmHHmRDHmmmumm~mmHll
/ Electrical Contact F i g . 24.
S c h e m a t ic d l a g r a m o f a PbSnTe l a s e r grown w i t h a MgF2 growth mask. The v a r i a t i o n o f r e f r a c t i v e i n d e x n and e n e r g y gap Eg a r e shown a t t h e l e f t .
The f i r s t IV-VI h e t e r o s t r u c t u r e l a s e r was an u n u su al v a r i a t i o n from t h e c o n v e n t i o n a l s t r u c t u r e s d i s c u s s e d a b o v e . I t c o n s i s t e d o f a p-PbTe f i l m grown on a BaF 2 s u b s t r a t e
80
H. Holloway and J. N° Walpole
(66) with a Schottky barrier junction formed by an evaporated Pb strip. Ohmic contact to the film was obtained with sputtered Pt as indicated in Fig. 25. The lateral current flow required by the geometry of this structure leads to lateral voltage drops increasing with the sheet resistance of the film. Hence it is difficult to obtain uniform injection under the strip of Pb. In the devices reported, injection occurred along the edge of the Pb strip only and the active volume of the laser was concentrated within a diffusion length on each side of the edge. The BaF 2 substrate provided strong opitcal and carrier confinement on one side of the film while the Pb Schottky barrier or the interface with air adjacent to it provided optical confinement on the other. There was no significant lateral confinement within the film however. Carrier confinement at the Pb barrier would be characterized by the injection efficieny of the barrier, i.e. the ratio of current flow arising from injected electrons to the toal current flow. The injection efficieny of this type of junction appears to be good at low temperatures (62). Nevertheless, a significant Richardson's current flow could occur in this structure (63), and may prevent high temperature operation.
-H
8
Spectrometer Resolution
A
Pb
PbTe I
==6
-
o4
h.........~
z
o
,,idlL ~ A ~
A
6.0
5.8
5.6
5.4
5.2
5.0
4.8
(microns) Fig. 25
Spontaneous emission spectrum of a thin-film PbTe laser fabricated as shown in the insert. The theroretical fit (dashed curve) is for k-conserving parabolicband transitions.
Because of the transparent substrate, it was possible to collect a large fraction of the isotropically emitted spontaneous light from these devices so that the spontaneous emission spectrum could be measured. Few well-resolved data of this sort exist otherwise for IV-VI materials. Spectra obtained at 77 K showed a good fit to the theoretical spectrum for direct band-to-band transitions between parabolic bands as shown in Fig. 25. Although pulsed laser operation was observed in these devices, cw operation was precluded by the lack of adequate heat-slnking. Further improvements in these devices were reported later (IO3). Post-growth annealing of the films to increase
MBE Techniques for IV-Vl Optoelectronic Devices
81
carrier density and thus decrease sheet resistance and an etching technique used to form smooth parallel mirrors resulted in lower thresholds and cw operation was achieved near 10 K. Threshold current densities weree estimated to be about I KA/cm 2. The ability to etch smooth varallelmirrors in films ~rown on BaF 2 substrates may be a si2nificant advantage in device fabrication since mirror quality can be a serious problem in PbSnTe lasers because of the poor cleavage properties of the material (104). In spite of their promise, further development of IV-Vl lasers grown on BaF2 substrates has not been pursed to date. Subsequent efforts have made use of the IV-VI materials as substrates where the problem of lateral current flow is avoided and where heatsinking of the active layer is more straight forward. The first device of this type was an SH laser in which an n-type layer 3 ~m in thickness was grown on a p-type Pbo.88Sno.12Te substrate (105). The laser emission wavelength corresponded to the smaller band-gap PbSnTe substrate rather than the PbTe film as expected. The depth of the p-~z junction in this device was not established but most likely some diffusion during growth of an n-skin into the substrate on the order of I B m deep occurred. A pyrolytically deposited SiO 2 growth mask patterned with 50 ~m-wide striped openingswas used to obtain a stripe geometry. Polycrystalline PbTe grew over the oxide with crystalline growth in the stripes. The oxide did not adhere well to the substrate after growth so that it was not possible to mount the device with the PbTe layer bonded to the heat sink without shorting out the stripe. Instead, the Pbl_xSnxTe bulk side of the device was bonded to the heat sink. Despite the poor heatsinking, cw operation was obtained up to 65 K. In Fig. 26, the pulsed threshold current density as a function of temperature for a SH laser is compared with data from two low-threshold stripe-geometry homestructure lasers made by diffusion into bulk material. Singificant threshold reduction for the SH laser can be seen. At 77 K the threshold of 780 A/cm 2 is lower by a factor of 3 than any PbSnTe homostructure devices at that temperature reported to date. This result strongly suggests that high recombination at the film-substrate interface does not occur. The improvement of the SH laser over the diffused homostructures is due to the carrier confinement provided by the PbTe layer which prevents diffusion of injected holes to the contact where they would recombine. A previously unobserved phenomenon was seen in the I-V characteristics of these SH lasers. It has subsequently been observed in a number of injection lasers (i06-10~ including MBE DH lasers (101). At threshold the I-V characteristic abruptly changes from an exponential to a linear dependence. The change in slope or differential resistance (dV/d/) of an SH laser is shown versus d.c. bias current I at several temperatures in Fig. 27, but the effect is also obvious in the I-V characteristic itself as shown in Fig. 28 for an MBE DH laser at 77 K. The onset of strong stimulated emission at threshold reduces minority carrier liftime so that no further voltage (population inversion) can be applied to the junction and current flow is subsequently limited only by series resistance. As can be seen in Fig. 27 at 30 K and 47 K, there is sometimes a small drop in junction voltage at threshold which may be due to saturable absorption (i.e. less gain is required to maintain laslng action than to reach threshold due to the saturation of absorption processes). The junction voltage saturation implles virtually I00% incremental internal quantum efficiency above threshold (105, 109). It is a consequence of strong carrier confinement withing the optical field (in the lateral direction, by the use of stripe geometry, as well as in the direction transverse to the junction). It should be emphasized that the incremental internal efficiency above threshold is not related to the radiative quantum efflcency below threshold. Sleger et u~. (II0) have reported PbSSe SH lasers grown by the hot-wall technique in which the small band-gap active region is the epitaxial film rather than the substrate.
82
H. Holloway and J. N. Walpole
These d e v i c e s were o n l y o p e r a t e d p u l s e d a t 12 K, b u t had t h r e s h o l d c u r r e n t d e n s i t i e s of a b o u t 200 A/cm2 comparable t o h o m o s t r u c t u r e d e v i c e s , a g a i n i m p l y i n g no s e r i o u s n o n r a d i a t i v e recombination at the h e t e r o - i n t e r f a c e nor at m i s f i t d i s l o c a t i o n s gener a t e d by t h e 0.2% l a t t i c e mismatch. The r e l a t i v e l y low t h r e s h o l d a l s o i n d i c a t e s t h a t t h e l a s e r q u a l i t y o f t h e f i l m i n o t h e r r e g a r d s was comparable t o b u l k - g r o w n c r y s t a l s .
¢p_
z zp
Pbo6eSno t2Te
-
|
to /
0
Homojunction Stripe Geometry
J
i /
~
~
~Si.r~ile-_lqeterojunction
I
~
I
20
I
30
I
40
I
50
I
60
1
70
80
T ('K) F i g . 26
P u l s e d t h r e s h o l d c u r r e n t d e n s i t y of a SH PbSnTe l a s e r as a f u n c t i o n of t e m p e r a t u r e (T) compared w i t h two typical low-threshold dlffused-junctlon bulkcrystal devices.
DE[ l a s e r s grown by t h e h o t - w a l l t e c h n i q u e were f i r s t r e p o r t e d by McLane and S l e g e r (111) i n the PbSl-xSex a l l o y stem. T h r e s h o l d c u r r e n t d e n s i t i e s as low as 60 A/cm2 were o b t a i n e d f o r cw o p e r a t i o n a t 12 K, b u t no h i g h e r t e m p e r a t u r e d a t a were r e p o r t e d . P r e i e r e t a ~ ( l 1 2 ) a l s o u s i n g t h e h o t - w a l l t e c h n i q u e i n t h e PbSl-xSe x a l l o y stem, have s u b s e q u e n t l y o b t a i n e d DH l a s e r s w i t h low t h r e s h o l d a t and above 77 K. These d e v i c e s were made w i t h s t r i p e d c o n t a c t s to form a s t r i p e geometry as seen i n t h e i n s e t of Gig. 29. The p u l s e d t h r e s h o l d c u r r e n t d e n s i t y as a f u n c t i o n of t e m p e r a t u r e i p l o t t e d i n F i g . 29 f o r two Dtt l a s e r s and, f o r comparison, a d i f f u s e d h o m o s t r u c t u r e l a s e r . Above 30 K and 60 K, r e s p e c t i v e l y , t h e two DH l a s e r s show lower t h r e s h o l d current densities than the diffused laser. These d e v i c e s were c a p a b l e of cw o p e r a t i o n up t o 96 K. Higher t e m p e r a t u r e o p e r a t i o n c o u l d have b e e n a c h i e v e d w i t h a t h i n n e r PbS top l a y e r which a t 5 ~m t h i c k n e s s r e p r e s e n t s a l a r g e t h e r m a l impedance ( t h e
83
MBE Techniques for IV-VI Optoelectronic Devices
t h e r m a l c o n d u c t i v i t y of PbS a t 100 K i s a b o u t 0.07 W c~- 1 K -1) (112). L a t e r , s i m i l a r d e v i c e s w i t h 1.5 Um t h i c k top l a y e r s were o p e r a t e d cw up t o 120 K (113). 4.0
-tl
08
•~
G6
•o
o4
Single-Heterojunction P b T s - Pbol m Sno.tzTe L o w
4 . 2 eK
2 0 "K
0.2--
m
.--. 0
I 20
I 30
1 50
I
4o
I
60
I
7O
I (mA)
Fig. 27
The incremental resistance (dV/d/) of a Sh PbSn Te laser vs current I at several temperatures. The abrupt decrease in resistance occurs at ow leaser threshold.
At currents well above threshold, laser modes were seen over a sifficiently broad spectral region that slight inhomogeneity or grading of the PbSSe alloy composition seemed i~dicated.
I n t h e h o t - w a l l PbSSe r e s u l t s d i s c u s s e d above, c o n t r o l of c a r r i e r type and d e n s i t y was a c h i e v e d by an i n d e p e n d e n t l y c o n t r o l l e d Se vapor s o u r c e . While s i m i l a r c o n t r o l of type can be a c h i e v e d i n c o n v e n t i o n a l MBE ( 7 9 ) , i t i s d i f f i c u l t t o a c h i e v e h i g h p - t y p e c o n c e n t r a t i o n s needed f o r low r e s i s t a n c e c o n t a c t s . I n a d d i t i o n , d i f f u s i o n of t h e ~ - ~ j u n c t i o n can be r a p i d even a t t h e low t e m p e r a t u r e s of ~g3E growth as mentioned e a r l i e r . McLane and S l e g e r (111) e s t i m a t e d t h e i r c o n t r o l of j u n c t i o n p o s i t i o n t o b e no b e t t e r t h a n 1 ~m. For t h e s e r e a s o n s , the PbSnTe DH l a s e r s grown by MBE have b e e n doped w i t h f o r e i g n i m p u r i t i e s as d i s c u s s e d i n S e c t i o n 2. R e l a t i v e l y h i g h c o n c e n t r a t i o n s can be e a s i l y a c h i e v e d and j u n c t i o n d i f f u s i o n i s expected to b e s l o w e r f o r f o r e i g n d o p a n t s , as mentioned t h e r e . The PbSnTe DH l a s e r s were grown i n a s t r i p e geometry u s i n g a s t r i p e d growth mask o f MSF2 w i t h s t r i p e w i d t h s of 10 to 50 ~m. S u b s t r a t e s were T l - d o p e d FoTe (~ 1019 cm- 3 ) . A c t i v e r e g i o n s 1.5 ~m t h i c k were grown u s i n g Pbo.78SnO 22Te, Bi-doped to a c h i e v e a b o u t 6 X 1017 cm- 3 n - t y p e c o n c e n t r a t i o n . Cap l a y e r s 0.75 ~m t h i c k were grown u s i n g Pbo.88Sno.12'Te, Bi-doped to a c h i e v e 2 X 1018 cm-3 n - t y p e c o n c e n t r a t i o n . These d e v i c e s were c a p a b l e of ow o p e r a t i o n up t o 114 K (101). However, t h r e s h o l d c u r r e n t d e n s i t i e s were l a r g e p a r t i c u l a r l y f o r low t e m p e r a t u r e s as seen i n F i g . 30, where t h e t h r e s h o l d and t h e w a v e l e n g t h o f e m i s s i o n a r e p l o t t e d vs t e m p e r a t u r e .
84
H. Holloway and J. N. Walpole
Subsequent devices have had somewhat lower thresholds, ranging between 2 and 6 KA cm-2, but remaining large compared to diffused homostructure devices, the SH PbnTe devices, or the PbSSe devices discussed above. It can be easily determined from the wavelength of the emission at threshold and from the I-V characteristics such as that shown in Fig.28, that the voltage across the junction at threshold is within a few mV of the voltage required to establish population inversion. Since the current flow required to reach population inversion is clearly the dominant contribution to the threshold current (115), the high thesholds cannot be explained by optical losses. Instead there appears to be a large component of current due to surface leakage or nonradiative recombination. While misfit dislocations or interface recombination cannot be ruled out as the source of this current, the evidence of the SH lasers and the PbSSe lasers does not support hsi view (41), nor do the results obtained with epitaxial homostructure devices discussed below. Other possibilities include surface leakage, nonradiatlve recombination associated with the high levels of foreign impurities present (116), and nonradiative recombination due to a non-equilibrium, high concentration of native defects which have been proposed to explain low mobility in unannealed MBE films (49) of PbSnTe on BaF 2.
200
mA
-8-15276 I
m II
CURRENT
20 rnV VOLTAGE Fig. 28
The current-voltage characteristic of a DH PbSnTe laser at 77 K. Note the linear relationship above threshold at I = 600 mA.
The high threshold of the PbSnTe DH lasers grown by MBE do not significantly limit their performance since cw operation can be achieved for current densities over 20 KA cm-2. Adequate power and mode quality for high-resolution spectroscopy at 80 K has been demenstrated (117). The control of the thickness of the cap layer and the active layer, made possible by MBE, is a crucial factor in obtaining good performance since, as mentioned, the poor thermal conductivity of the lead-tin chaleogenides causes poor heat-sinking for junctions more than a few ~m deep. The MBE techniques used for the PbSnTe DH lasers have also proven extremely reproducible in numerous growth runs and the yield of devices with cw operation above 77 K from each wafer is high.
MBE Techniques for IV-VI Optoelectronic Devices
PLATED Au \
85
EVAPORATED In Cr
FL,-.'-'.-~-'~-~;'.~-'-'~ ~ S ~,~ p-PbS .j___~ ~ 1~m Cno, pl-PbS~_,Se,
L....................... Au
10 4
/
/
Cr
In
_~----2oo~,m
.-PbS
'
DIFFUSED LASER
%
E <:~
I0 ~
.= ...o
DH-LASERS
102 - -
10
Fig. 29
I
l
i
I
i
I
,
I
I
I
,I,I,I,
Comparison of pulsed threshold current densities of DH PbSnTe lasers with a bulk crystal diffused-junction laser as a function of temperature (from Eel. 112).
No other lasers grown by MBE or hot-wall techniques have been reported. There are several papers of interest in the Russian literature (116,118,119). However, no high performance at higher temperatures was mentioned and the growth techniques were not always specified but appear to involve vapor epltaxy. A disappointing aspect of the ~ E and hot-wall heterostructure lasers reported to date is that their external quantum efficienies have been low (less than a percent or two). Typical maixmum output power has been on the order of a few hundred ~W in PbSSe and about I00 BW in PbSnTe.* Much higher external efficlencies and output
86
H. Holloway and J. N. Walpole
powers have b e e n a c h i e v e d i n b u l k d i f f u s e d - j u n c t i o n l a s e r s (104, 106). A l t h o u g h the powers o b t a i n e d from h e t e r o s t r u c t u r e s a r e more t h a n a d e q u a t e f o r most a p p l i c a t i o n s , they are marginal for local oscillator applications in heterodyne detection.
20 18
14
12-10
12 .~
10 8 6
Jth L
U,J
IJJ
4
U 2
0
I
10
1
20
I
30
4O
I
50
I
60
I
"tO
1
80
I
90
1
100
I
110
OE T
t-U
0 120
HEAT-SINK TEMPERATURE ( K )
Fig. 30.
Threshold current density and emission wavelengths of a DH PbSnTe laser as a function of heat-sink temperature. The longest wavelength observed pulsed compared to the cw threshol~ wavelength gives a measure of the temperature increase in the active region.
The low external efficiency cannot generally b e attributed to low internal efficiency since the DH PbSnTe lasers which show junction-voltage saturation at threshold also have low external efficiency. Poor output coupling and scattering losses at the end mirrors due to poor mirror quality (104) and due to mode-coupling effects (121) nmy play a large role in the low efficiency. Also scattering losses and free-carrier absorption in the active region may be much larger in laser devices than in the waveguide structures discussed in section 4. The laser waveguides are much thinher, which leads to increased scattering due to roughness of the waveguide walls (103), and carrier densities in the lasers are much higher**. Optimized doping and the use of separate optical and carrier confinement structures might significantly improve external efficiency. An early DH Pbo.88Sno.12Te laser (120) with an undoped active region 6 ~m thick and an undoped PbTe cap 1 ~m thick grown on a Tl-doped PbTe substrate gave cw output power up to 9 ~f~. These devices were not capable of cw operation above = 30 K, however. ** Additional absorption above the classical free-carrier absorption may be present near the band edge in heavily doped materials (93,122).
87
HBE T e c h n i q u e s f o r IV-VI O p t o e l e c t r o n i c Devices 8.2.
Homostructure l a s e r s
S t r o n g o p t i c a l and c a r r i e s c o n f i n e m e n t can be a c h i e v e d i n the l e a d - t i n c h a l c o g e n i d e s w i t h p r o p e r l y c o n t r o l l e d c a r r l e r - c o n c e n t r a t l o n p r o f i l e s (123). Because of t h e s m a l l e f f e c t i v e masses of b o t h e l e c t r o n s and h o l e s , r e l a t i v e l y l a r g e b u i l t - i n p o t e n t i a l s (band b e n d i n ~ o c c u r a t i n t e r f a c e s between low and h i g h c a r r i e r c o n c e n t r a t i o n . Also l a r g e f r e e - c a r r l e r c o n t r i b u t i o n s to the d i e l e c t r i c c o n s t a n t o c c u r b e c a u s e of t h e s m a l l e f f e c t i v e masses and t h e s m a l l energy gaps o r long w a v e l e n g t h s of o p e r a t i o n . D i f f u s e d - j u n c t i o n l a s e r s have b e e n o p e r a t e d ow r e c e n t l y up t o 6 1 K u s i n g c o n c e n t r a t i o n g r a d i e n t s f o r c o n f i n e m e n t (124). The d i f f u s i o n i s v e r y c r i t i c a l and d i f f i c u l t to control s i n c e t h e j u n c t i o n depth must be kept v e r y s m a l l f o r h e a t - s i n k l n g . E p i t a x i a l t e c h n i q u e s a r e b e t t e r s u i t e d f o r c o n t r o l l i n g c o n c e n t r a t i o n and depth of the j u n c t i o n . E p i t a x i a l h o m o s t r u c t u r e d e v i c e s c a p a b l e of cw o p e r a t i o n to 100 K have b e e n r e p o r t e d (125) grown by b o t h ~ E and LPE h a v i n g n + - n - p + doping p r o f i l e s . Bidoped a c t i v e r e g i o n s (n ffi 5 X 1017 cm-3) 1.5 pm t h i c k f o l l o w e d by Bi-doped capping l a y e r s (n + ffi 5 X 1018 cm-3) 0.75 pm t h i c k were grown on T l - d o p e d s u b s t r a t e s (p+ = 1019 cm-3). The HBE h o m o s t r u c t u r e s were grown i n PbTe and Pbo.78Sno.22Te, b o t h c o m p o s i t i o n s operati~tg cw to 100 K. T h r e s h o l d c u r r e n t d e n s i t i e s were comparable to t h e DH l a s e r s b u t o u t p u t powers were somewhat b e t t e r . F i g u r e 31 shows t h e cw s p e c t r a of a PbTe h o m o s t r u c t u r e a t 15 and 102 K w i t h 2.4 mN and 0.64mW o u t p u t , r e s p e c t i v e l y , predomln a n t l y i n a s l n g l e mode. The maximum o u t p u t o b s e r v e d from Pbo.78Sno.22Te homos t r u c t u r e s was ~ 0 . 4 mW a t 15 K and ~ O.12M~ a t 80 K.
T = 15K I = h84A
T= 102K
Ith = 0.12A
Ith • 1.3A
I ffi 2.18A
P = 0.64roW (Single End)
P = 2.4roW (Single End)
i
k
.
1680
L
_
_
1590
~
~
1600
1610
1950
1960
i
@32
I
I 6.28
I
I 6.24
WAVELENGTH (Fro)
F i g . 31.
i
l 6.20
[ 5.12
I
I
1970
I
1980 1990 WAVE NUMBER (cm -1)
WAVE NUMBER (cm-1) I
/
-
I
5.08
I
i
I
5.04
.
I
I
2000
WAVELENGTH (Fro)
S p e c t r a of a I~E h o m o j u n c t i o n PbTe l a s e r a t 15 and 102 K showing the r e l a t i v e l y h i g h s i n g l e - m o d e o u t p u t powers.
S i n c e t h r e s h o l d s a r e comparable t o DH l a s e r s and e x t e r n a l e f f i c i e n c i e s
i
500
are s t i l l
88
H. Holloway and J. N. Walpole
rather low, the homostructure results again imply that the lattice mismatch present in otherwise similar heterostructurelasers plays little role in limiting their performance. Nevertheless, if lattice matching should prove important, I~SE homostructures seem to offer the advantage of higher temperature operation without the lattice mismatch problem. 8.3.
Distributed feedback lasers
Figure 32 shows schematically the structure of a distributed feedback (DFB) laser which has been fabricated in MBE-grown PBSnTe layers (126, 127). The device is essentially the same as the DH lasers discussed above except that feedback is provided by Bragg reflection from the periodic grooves (grating) etched into the top layer rather than by reflections from the two ends of the device. The 800-A insulating MgF 2 film waspatterned as shown so that only one end of the mesa stripe was electrically contacted leaving an uncontacted length with high optical absorption to prevent reflective feedback from the other end of the mesa. The grating and the deeper grooves isolating the mesa stripe were obtained using photolithographic techniques and rf sputter etching.
[~
o
~
~' ~
60N'm
- 800 ~ .4~= 9.~.
I
/
1.1 • m
I ~
O.3~m DEPTH
0.5/~m
i.5~.m I
~'~" X --- 0
,~p.,..I~"
n ~ 2 X I018cm-3 (8i) n
~ 6 X 1017cm-3 (Bi)
p~ 1019 cm-3 (TL) Fig. 32.
Schematic of DFB double-heterostractare PbSnTe laser without metallization.
One advantage of the DFB laser is that is has significant mode discrimination, i.e. the longitudinal spatial modes of the cavity have different thresholds. The PbSn Te DFB lasers showed predominantly single-mode cw operation over a wide range of temperature and current. Another advantage of DFB lasers is that the frequency of the laser can be controlled by the grating periodicity since oscillation can occur only near the frequency at which Bragg reflection occurs. These advantages make DFB lasers potentially the ideal laser for many applications. The successful fabrication of the DFB structure shown in Fig. 32 is due largely to the ability to accurately control the thickness of the top layer to <0.i ~m. It is a device for which MBE techniques are ideally suited. The scanning electron micrographs shown in Fig. 33 depect two views of grating of I.I Bm periodicity with grooves 0.3 ~m deep. The grooves must be close to the active layer to obtain good
MBE Techniques for IV-VI Optoelectronic Devices
89
c o u p l i n g f o r Bragg r e f l e c t i o n . They must n o t , however, p e n e t r a t e t h e a c t i v e l a y e r o r c a r r i e r c o n f i n e m e n t would b e l o s t . The cap l a y e r was grown 0 . 5 ~m t h i c k , so t h a t c o n t r o l o f e t c h i n g p a r a m e t e r s a s w e l l as l a y e r t h i c k n e s s i f c r i t i c a l f o r good performance.
F i g . 33.
Scanning electron micrographs of a grating with l.l-vm periodicity, sputter etched to a depth of 0.3 ~m in PbSnTe.
9.
CONCLUSIONS
The optoelectronic device results which have been reviewed here demonstrate that hlgh-quality devices, competitive in performance with bulkcrystal devices, can be achieved in IV-VI MBE films. Indeed there are technological advantages and performance features discussed here which have not been demonstrated in bulkcrystal devices. These included the option of backside illumination of detector devices with transparent substrate; the compatibility with planar technology for smell
90
H. Holloway and J. N. Walpole
c l o s e l y - s p a c e d elements and for a p p l i c a t i o n s in i n t e g r a t e d o p t i c s ; and the c o n t r o l of l a y e r thicknesses which makes p o s s i b l e l a s e r devices with higher temperatures of cw o p e r a t i o n , DFE l a s e r s devices u t i l i z i n g t h i n - f i l m o p t i c s , and (because of the confinement provided by the proximity of an i n s u l a t i n g s u b s t r a t e ) devices such as the p i n c h - o f f diodes and the l a t e r a l c o l l e c t i o n devices. While HBE techniques have led to advances in device technology, i t i s also true t h a t the device r e s u l t s so obtain have given important information on the q u a l i t y of HBE films. The quantum e f f l c l e n c l e s and c u r r e n t - v o l t a g e c h a r a c t e r i s t i c s of j u n c t i o n devices give evidence about minority c a r r i e r l i f e t i m e , generation-recombination mechanisms, and i n t e r f a c e recombination. The g e n e r a l l y hish q u a l i t y with r e s p e c t to these p r o p e r t i e s found in the d e t e c t o r devices i s not found in many of the l a s e r s t r u c t u r e s discussed, as evidenced by the high t h r e s h o l d s . Also o p t i c a l a b s o r p t i o n , which can be low in passive I~E waveguides, appears to be high in the t h i n , h e a v i l y doped l a y e r s of the l a s e r s t r u c t u r e s , as evidenced by the low e x t e r n a l quantum e f f i c i e n c i e s . Although the device r e s u l t s suggest approaches f o r improvements and f u r t h e r progress is l i k e l y , there is need f o r more work to c o r r e l a t e c r y s t a l p e r fection (e.g. native defect concentrations, foreign impurities, misfit dislocations, i n t e r f a c e s t a t e s , e t c . ) with the p r o p e r t i e s s i g n i f i c a n t f o r device performance, i . e . c a r r i e r recombination and o p t i c a l l o s s e s . C r i t e r i a f o r good device performance are not e s t a b l i s h e d even f o r bulk c r y s t a l s and b e t t e r knowledge would g r e a t l y simplify the process of optimizing the growth and the design and f a b r i c a t i o n of devices. To this end, more s o p h i s t i c a t e d HBE systems with i n - s i t u a n a l y s i s equipment such as Auger and HEED instrumentation might prove f u l t f u l . The s t a t e - o f - t h e - a r t performance of IV-Vl o p t o e l e c t r o n i c devices 8 r o w n b y l ~ E is already more than adequate f o r many a p p l i c a t i o n s . Because of t h e i r demonstrated a p p l i c a t i o n s and t h e i r p o t e n t i a l for new a p p l i c a t i o n s , I v - v I ~ E devices are l i k e l y to remain t e c h n o l o g l c a l l v imvortant in the f u t u r e . REFERENCES
i. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15a. 15b. 16. 17. 18. 19. 20. 21.
H. Hinterberger, Z. Phys. 119, 1 (1942). L. Sosnowskl, J . Starklewicz and O. Simpson, ~ t t t r ~ , Lond. 159, 818 (1947). A.J. E11em-- and H. Wilman, Pros. Phys. Soy. Lond. 61, 164 ~ ' 9 4 8 ) . S.G. Parker, J. eZ4c~roeher,. Sos. 123, 920 (1976). H.K. Pandy, SoZ4d St. Commun.15, 449 (1974). L.R. Koller and H.D. C o g h i 1 1 , 7 . eleetr, oehem. Sos. 107, 973 (1960). P. Hudock, Trans. Am. Ins#. Mesh. ~ s . 239, 338 (1-~7). R.P. Bis, J.R. Dixon and J.R. Lowrey, J. v a t . $~i. Teohnol. 9, 226 (1972). A. Lopez-Otero and L.D. Haas, Thin S o l i d F4/~e 23, 1 (1974). K. Duh and H. P r e i s e r , Th4n S o l i d F4~m~ 27, 247~1975). I . Kasai, J . Hornung and J . Baars, J. e ~ t m o n . Mater. 4, 299 (1975). I . Kasai, D.W. Bassett and J . Hornung, J. appl. Phys. 47, 3167 (1976). R.B. Schoolar, J.D. Jensen and G.M. Black, Appl. Phys. L e f t . 31, 620 (1977). E.B. Schoolar and J.D. Jensen, Appl. Phys. L e t t . 31, 536 (1977-~. L.L. Chang and R. Ludeke, in: E p 4 ~ l Growth, P ~ t A, (ed.) J.W. Matthews, Academic Press, New York, San Francisco, London (1975). J.N. Zemel, in: Sol4d State Sumf~e S~en~e, (ed.) M. Green, Dekker, New York (1969). A. Lopez-Otero, Thin Solid Films 49, 3 (1978). A.M. Andrews J.A. Higglns, J.T. Longo, E.R. Gerner and J.G. Pasko, Appl, Phys. Left. 21 , 285 (1972). C.C. Wang and S.E. Hampton, Solid-St. Eleol-z~m. 18, 121 (1975). S.H. Groves, K.W. Hill and A.J. Strauss, Appl. ~ s . Left. 25, 331 (1974). T.F. Tao, I. Kasai and J.P. Butler, J. ~ . Sei. T e e ~ l . 6, 918 (1969). E. Krlkorlan J. ~ o S~. Te~hnol. 12, 186 (1975).
I~E Techniques f o r IV-VI Optoelectronic Devices 22. 23. 24. 25.
R.F. R.F. R.F. R.F.
P o r t e r , J. ohM. Phys. 34, Brebrick and A.J. S t r a u s s , Brebrick and A.J. Strauss, Bis, J. va~. Soi. Teah~o~.
91
583 (1961). J . oh~m. Phys. 40, 3230 (1964). J. ohem. Phys. 4 , 197 (1965). 7, 126 (1970).
26.
w.c. Chan, I n ~ d
27. 28. 29. 30. 31. 32.
D.A. Northrop, J . e~en#~oa~m. 6oc. 118, 1365 (1971). M.C. Finn and J.N. Walpole, unpubllshed. A.J. S t r a u s s , Phys. Rev. 157, 608 (1967). R.W. B i c k n e l l , I n ~ d PhYs. 17, 57 (1977). R.W. Bickne11, In3~az,ed Phys. 18, 133 (1978). T.C. Harman, in: Phys44s of I V ~ I Compounds and Alloys, (ed.) S. Rabii, Gordon & Breach, London (1974). J.N. Walpole and R.L. Guldl, in: Physics of IV-VI Compounds and A~loysj (ed.) S. Rabli, Gordon & Breach, London (1974). G. Dionne and J.C. Woolley, Phys. ReO. B6, 3898 (1972). D.L. Smith and V.Y. Pickhardt, J. eZeutTon. Muter. 5, 247 (1976). H. Holloway, D.K. Hohnke, R.L. Crawley and E. Wilkes, J. ~u. Sui. Teuhnol. O, O0(0OO). J.W. Matthews, Phil. Mug. 6, 1347 (1961). J.W. Matthews, Phil. Mug. 8, 711 (1963). K. Yagi, K. Takayanagi, K. Kobayaski, and G. Honjo, J. Cr~s. Gmmoth 9, 84 (1971) J.N. Walpole, S.H. Groves, A.R. Calawa and T.C. Harman in: Solid Sta~e Reseach Report, Lincoln Laboratory, M.I.T. (1974:4), p.13 DDC AD-AO02773/O. R.W. Bicknell, J. ~ac. Sci. Teuhnol. 14, 1012 (1977). J.N. Zemel, J.D. Jensen and R.B. Schoolar, Phys. Rev. 140A, 330 (1965). J.H. Myers, R.H. Morrlss and R.J. Deck, J. dppl. Phys. 42, 5578 (1971). H. Holloway, in: Use of Thin Films for Phys~al Invest~tions, (ed.) J.C. Anderson, Academic Press, New York (1966). R.S. Allgaler and B.B. Houston, Jr., Proc. Int. Conf. Phys. Semiuondaotors, p. 172, Exeter, 1962, Institute of Physics and Physical Society, London, (1962). E.M. Logothetls and H. Holloway, in: Physiu8 of IT-VI Compounds and AILg~8, (ed.) S. Rabii, Gordon & Breach, London (1974). J.P. Donnelly, T.C. Harman and A.G. Foyt, Apply. Phys. Left. 18, 259 (1971). H. Holloway, E.M. Logothetis and E. Wilkes, J. e'ppZ. Phys. 41, 3453 (1970). H. Holloway and E.M. Logothetls, J. appl. Phys. 42, 4522 (I~I). T.O. Fahrlnre and J.N. Zemel, d. vac. Sui. Teu~no-~. 7, 121 (1970). H.M. Manaserit and W.I. Simpson, J. ~eu%Pouhem. Sou. 122, 444 (1975). V.L. Lambert, J. appl. Phys. 46, 2304 (1975). D.K. Hohnke, H. Holloway and M.---D.Hurley, Thin Solid Film8 38, 49 (1976). H. Holloway, unpublished data. A. Lopez-Otero, J. appl. Phys. 48, 446 (1977). R.S. Allgaier and W.W. Scanlon,-~hys. Rev. III, 1029 (1958). G.A. Antcliffe and J.S. Wrobel, Mu~P. Res.~ll. 5, 747 (1970). C.M. Wolfe and G.E. Stillman, Appl. Phys. Left. 18~ 205 (1971). R.B. Schoolar, Appl. Phys. Lett. 16, 446 (1970). R.B. Schoolar, J. ~u. Sui. Teuhno'-~. 9, 225 (1972). E.M. Logothetls, H. Holloway, A.J. Varga and E. Wilkes, Appl. Phys. Left. 19, 318 (1971). K.W. Nil1, A.R. Calawa, T.C. Harman and J.N. Walpole, App,. Phys. Left. 16, 375 (1970). K.W. N i l l , J.N. Walt)ole. A.R. Calawa and T.C. Harman. in: The Phus~s o~ Semimentuls and ~ C , ro Semiuond~torsj (eds) D.L. Carter and R.T. Bate, Pergamon Press, Oxford (1971). H.B. Morris, R.A. Chapman, R.L. Guldi and S.C. Parker, paper presented at IRIS Detector Speu441#y Gz~up Meeting, Fort Monmouth, 1975. J.N. Walpole and K.W. Nil1, J. uppl. P~JS. 42, 5609 (1971). H. Holloway, W.H. Weber, E.M. Logothetis, A.J. Varga and K.F. Yeung, App,. Phys Lett. 21, 5 (1972).
33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66.
Phys. 14, 177 (1974).
92 67. 68.
H. Holloway and J. N. Walpole
H. Holloway, W.J. Johnson and A.J. Varga, unpublished data. H. Holloway, in: Physics of IV-VI Compounds and Alloys, (ed.) S. Rabli, Gordon & Breach, London (1974). 69. E.M. Logothetis, H. Holloway, A.J. Varga and W.J. Johnson, Appl. Phys. Lett. 21, 411 (1972). 70. J.P. Donnelly and H. Holloway, Appl. Phys. Lett. 23, 682 (1973). 71. R.E. Callender, paper presented at IRIS Detector ~eciaZty Group Meeting, Washington, D.C., 1973. 72. A. Lopez-Otero, L.D. Haas, W. Jantsch and K. Lischka, Appl. Phys. Lett. 28, 546 (1976). 73. Contract DAAKK 2-73-C-0225, Final Report, Febuary 1974. 74. H. Holloway, K.F. Yeung, A.E. Asch and D.A. Gorski unpublished data. 75. T.J. McMahon, Paper presented at IRIS Detector Specialt F Group Meeting, E1 Toro, 1974. 76. D.K. Hohnke and H. Holloway, Appl. Phys. Lett. 24, 633 (1974). 77. D.K. Hohnke and M.D. Hurley, J. appl. Phys. 47,--4975 (1976). 78. H. Holloway, D.K. Hohnke, K.F. Yeung, A.E. Asch and D.A. Gorski, unpublished data. 79. D.K. Hohnke and S.W. Kaiser, J. appl. Phys. 45, 892 (1974). 80. J.F. Butler, A.R. Calawa and T.C. Harman, App-~. Phys. Lett. 9, 427 (1966). 81. D.K. Hohnke, H. Holloway, K.F. Yeung and M.D. Hurley, Appl. ~hye. Lett. 29, 98 (1976). 82. W.H. Rolls and D.V. Edolls, Infrared Phys. 13, 143 (1973). 83. D.P. Mathur, Opt. Engng. 14, 351 (1975). 84. A.E. Asch, D.A. Gorskl, P.I. Zapella and H. Holloway, paper presented at IRIS Detector Specialty Group Meeting, Fort Monmouth, 1975. 85. H. Holloway, J. appl. Phys. (to be published). 86. H. Holloway and K.F. Yeung, Appl. Phys. Lett. 30, 210 (1977). 87. Contract DAAK02-75-C-OIO6, Final Report, June 1976. 88. H. Holloway, M.D. Hurley and E.B. Schermer, Appl. Phys. Lett. 32, 65 (1978). 89. H. Holloway, J. appl. Phys. (to be published). 90. A.J. Norelka, M.H. Francombe, W.J. Taei, R.N. Ghoshtagore and J.L. Wentz, paper presented at IRIS Detector Specialty Group Meeting, Colorado Springs, 1977. 9!. R.W. Ralston, J.N. Walpole, T.C. Harman, and I. Melngailis, Appl. Phys. Lett. 26, 64 (1975). 92. A.J. Strauss, J. Nonmetals i, 133 (1973). 93. C. Picketing, Solid St. Coleman. 22, 395 (1977). 94. J.L. Merz, R.A. Logan, W. Wiegmann, and A.C. Gossard, Appl. Phys. Lett. 26, 337 (1975). 95. J.N. Walpole, unpublished. 96. E.D. Hinkley, K.W. Nill, and F.A. Blum, in: Topics in Applied Phys@4s Vol. 2, Laser Spectroscopy of Atoms and Molecules, (ed.) H. Walter, Sprlnger-Verlag, Berlin, Heidelberg, New York (1976). 97. E.D. Hinkley, R.T. Ku, and P.L. Kelley, in: Topics in Applied Phys~s, Vol. 14, Laser Monitoring of the A~osphere, (ed.) E.D. Hinkley, Springer-Verlag, Berlin Heidelberg, New York (1976). 98. M. M,,~8, T. Kostiuk, S. Cohen, D. Buhl, and P.C. Von Thuna, Nature, Lond. 253, 514 (1975). 99. M.A. Frerking and D.J. Muehlner, Appl. Opt. 16, 526 (1977). I00. R.T. Ku and D.L. Spears, Opt. Lett. I, 84 (I~7). iO1. J.N. Walpole, A.R. C~lawa. T.C. Harman. and S.H. Groves, Appl. Phys. Lett. 28, 552 (1976). 102. K.J. Sleger, G.F. McLane, and D.L. Mitchell, in: Physics of IV-VI Compound8 and Alloys, (ed.) S. Rabii, Gordon & Breach, London (1974). 103. W.H. Weber and K.F. Yeung, J. ¢cppl. Phys. 44, 4991 (1973). 104. J.N. Walpole, A.R. Calawa, R.W. Ralston, and T.C. Harman, J. appl. Phys. 44, 2905 (1973). 105. J.B. Walpole, A.R. Calawa, R.W. Ralston, T.C. Harman, and J.P. MoVittle, Appl. Phys. Lett. 23, 620 (1973).
MBE Techniques for IV-VI Optoelectronic Devices 106. 107. 108. 109. IiO. Iii. 112. 113. 114. 115. 116.
R.W. Ralston, J.N. Walpole, A.R. Calawa, T.C. Harman, and J.P. McVittle,
J. appl. Phys. 45, 1323 (1974). T.L. Paoli and P.-'-A.Barnes, Appl. Phy8. Lett. 28, 714 (1976). K.J. Linden, K.W. Nill, and J.F. Butler, IEEE J. Quantum Electron. QE-13, 720 (1977). H.S. Sommers, Jr., Appl. Phys. Lett. 32, 547 (1978). K.J. Sleger, G.F. McLane, U. Steom, S.G. Bishop, and D.L. Mitchell, J. appl. Phys. 45, 5069 (1974). G.F. McL---aneand K.J. Sleger, J. electron, Mater. 4, 465 (1975). H. Preier, M. Bleicher, W. Riedel, and H. Maier, ~ppl. Phys. Lett. 28, 669 (1976). H. Preier, M. Bleicher, W. Riedel, and H. Maier, J. appl. Phys. 47, 5476 (1976) D.K. Honks and S.W. Kaiser, J. appZ. Phys. 45, 892 (1974). L.R. Tomasetta and C.G. Fonstad, IEEE J. Quantum Electron. QE-II, 384 (1975). L.N. Kurbatov, A.D. Britov, S.M. Karavaev, G.A. Kalyuzhnaya, M.I. Nikolaev, O.V. Pelevin, B.G. Girich, and T.F. Terekhovich, SoY. J. Quantum Electron. 7, 236
(1977). 117. 118. 119.
120. 121. 122. 123. 124. 125.
126. 127.
P.F. Moulton, D.M. Larsen. J.N. Waloole. and A. Mooradian. Opt. Lett. i. 51 (1977). L.N. Kurbatov, A.D. Britov, S.M. Karavaev, Y.I. Gorina, G.A. Kalyuzhnaya, and P.M. Starik, Soy. J. 6~z~ntT~mElectron. 5, 1137 (1975). A.D. Britov, S.M. Karavaev, G.A. Kalyuz~naya, Y.I. Gorina, A.L. Kurbatov, K.F. Kiseleva, and P.M. Starik, Soy. J. Quantum Electron. 6, 1217 (1976) and 6, 1384 (1976). J.N. Walpole, R.W. Ralston, A.R. Calawa, and T.C. l{arman, paper presented at the IEEE International Semiconductor Laser Conference, Atlanta (1974). R.W. Davies and J.N. Walpole, IEEE J. Quantum Electron. QE-12, 291 (1976). S.L. McCarthy, W.H. Weber, and M. Mikkor, J. a~rpl. Phys. 45, 4907 (1974). P.N. Asbuk, Ph.D. Thesis, M.I.T., 1975 (unpublished). W. Lo, IEEE J. Quantum Electron. QE-13, 591 (1977). J.N. Walpole, S.H. Groves, and T.C. Harman, IEEE Trans. Electron Devices ED-24, 1214 (1977). Also, Solid State REsearch Report, Lincoln Laboratory, M.I.T. (1977:4), p. 3, DDC AD-AO52463. J.N. Walpole, A.R. Calawa, S.R. Chinn, S.H. Groves, and T.C. Harman, Appl. Phys. Lett. 29, 307 (1976). J.N. Walpole, A.R. Calawa, S.R. Chinn, S.H. Groves, and T.C. Harman, Appl. Phys. Left. 30, 524 (1977).
94
H. Holloway and J. N. Walpole
THE AUTHORS
Dr. H. Holloway
Professor J. N. Walpole
Dr. Holloway was educated at Birkbeck College, University of London, where he received his B.Sc. (1956) and Ph.D. (1959) degrees in Chemistry. His research interests have included x-ray and electron diffraction, epitaxy -- particularly of compound semiconductors -- and the physics of thin-film devices. Currently, he is studying the properties of optoelectronic devices made with thin films of IV-VI Semiconductors. James Walpole was granted a B.S.E.E. degree in 1961 from Duke University, and then attended the Massachusetts Institute of Technology as a National Science Foundation Fellow, receiving his M.S. degree in 1962, and his Ph.D. degree in 1966, both in the Department of Electrical Engineering. His -~jor studies were in the areas of solid state physics and semiconductor devices. He was appointed Assistant Professor of Electrical Engineering at M.I.T. in 1966, and Associate Professor in 1971. Since 1972 he has been a staff member in the Applied Physics Group of the M.I.T. Lincoln Laboratory, where he is currently involved with development of GalnAsP heterostructure lasers. His previous research interests have centered around the physics of the lead-tin chalcogenide semiconductors, including studies of magnetoplasma wave propagation, diffusion, hot-electron conduction, metal-semiconductor contacts, and, in recent years, various aspects of injection lasers in these materials. He lives in Concord, Massachusetts with his wife and two children.