GaSb superlattice based long-wavelength infrared detectors: Growth, processing, and characterization

Infrared Physics & Technology 54 (2011) 247–251

Contents lists available at ScienceDirect

Infrared Physics & Technology journal homepage: www.elsevier.com/locate/infrared

InAs/GaSb superlattice based long-wavelength infrared detectors: Growth, processing, and characterization Alexander Soibel, Jean Nguyen, Linda Höglund, Cory J. Hill, David Z. Ting ⇑, Sam A. Keo, Jason M. Mumolo, Mike C. Lee, Sarath D. Gunapala Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA

a r t i c l e

i n f o

Article history: Available online 25 December 2010 Keywords: Infrared detector InAs GaSb Antimonide Superlattice

a b s t r a c t We report growth, processing, and characterization of antimonide superlattice long-wavelength infrared photodetectors based on the complementary barrier infrared detector (CBIRD) design. We used photoluminescence measurements for evaluating detector material and studied the influence of the material quality on the intensity of the photoluminescence. We performed direct noise measurements of the superlattice detectors and demonstrated that while intrinsic 1/f noise is absent in superlattice heterodiode, side-wall leakage current can become a source of strong frequency-dependent noise. We developed an effective dry etching process for these complex antimonide-based superlattices that enabled us to fabricate single pixel devices as well as large format focal plane arrays. Ó 2011 Elsevier B.V. All rights reserved.

1. Introduction The type-II InAs/GaSb superlattice (SL) and InAs/GaInSb strained layer superlattice (SLS) promise absorption coefficients comparable to HgCdTe (MCT), uniformity, reduced tunneling currents, suppressed Auger recombination, and normal incidence operation [1,2]. Owing to the flexibility of the nearly lattice-matched InAs/ GaSb/AlSb material systems, many SL infrared detectors with advanced heterostructure architecture have been fabricated, including the nBn [3], the double heterostructure (DH) [4,5], the graded-gap W-superlattice based DH structure [6], the pMp structure [7], and the complementary barrier infrared detector (CBIRD) [8] and related structure [9]. Many of these detectors have exhibited very good performance [10] despite the short lifetimes in SL absorbers [11–13]. The growth and processing of these device structures present special challenges due to the complexity of these advanced heterostructures. In this paper, we focus on the development of CBIRD and present some aspects of material growth and characterization, device processing, and device characterization.

2. The complementary barrier infrared detector device structure The basic CBIRD device structure consists of a LWIR InAs/GaSb absorber SL sandwiched between an InAs/AlSb hole-barrier (hB) ⇑ Corresponding author. Address: Jet Propulsion Laboratory, M/S 302-231, 4800 Oak Grove Drive, Pasadena, CA 91109-8099, USA. Tel.: +1 818 354 1549; fax: +1 818 393 4663. E-mail address: [email protected] (D.Z. Ting). 1350-4495/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.infrared.2010.12.023

SL, and a MWIR InAs/GaSb electron-barrier (eB) SL. The hB SL and the eB SL are respectively designed to have approximately zero conduction and valence subband offset with respect to the absorber SL, i.e., they acts as a pair of complementary unipolar barriers [8] with respect to the absorber SL. A heavily doped n-type InAs0.91Sb0.09 adjacent to the eB SL acts as the bottom contact layer. Detailed results and discussion on this particular CBIRD device have been reported earlier [8,14]. In short, a CBIRD device with a 9.9 lm cutoff reached 300 K background limited infrared photodetection (BLIP) operation at 87 K, with a black-body BLIP D value of 1.1  1011 cm Hz1/2/W for f/2 optics under 0.2 V bias. Furthermore, CBIRD-based LWIR imaging focal plane array (FPA) have been recently demonstrated and results of this development were presented at a recent SPIE conference [15], as well as in an article by Gunapala et al. appearing elsewhere in this issue. 3. Material growth and characterization In growing CBIRD material for focal plane array applications, material quality and uniformity are important considerations. Photoluminescence (PL) measurements in combination with X-ray diffraction (XRD) and atomic force microscopy (AFM) are used for evaluation of the CBIRD detector material directly after growth. The PL peak wavelength serves as a good estimate of the cut-off wavelength of the detector (Fig. 1) and from the PL intensity and the full-width-half-maximum (FWHM) of the PL spectrum, information about the material quality is obtained [16–18]. An important aspect for FPA fabrication is the uniformity of the epi-material. PL intensity mapping generates valuable information about material uniformity across wafer.

248

A. Soibel et al. / Infrared Physics & Technology 54 (2011) 247–251

Fig. 1. Comparison of the QE at an applied bias of 0.15 V (left axis) and the PL spectrum of the same sample (right axis) showing a good correlation between the PL-peak and the cut-off wavelength of the detector.

In this study, the radial dependence of the PL intensity of three different wafers was investigated, all grown in the same growth run, symmetrically mounted at different positions on a three-wafer carrier. Samples were grown in a Veeco Gen III molecular beam epitaxy chamber. The sample holder is a molybdenum block with openings for three 50 mm wafers. Each wafer is isolated from the molybdenum block in the front by a pyrolytic boron nitride ring. The wafers are held in place with a tungsten snap ring. 50 mm clear sapphire wafers thermally isolate the GaSb wafers from the snap ring. The PL measurements were carried out at a temperature of 77 K using a Thermo-Fisher Fourier transform spectrometer, run in step-scan mode, with a liquid nitrogen cooled HgCdTe detector. A 658 nm laser diode (excitation power of 3.2 W/cm2) was used as excitation source. Two of the three wafers exhibited no or little variation in PL intensity across the 50 mm growth (Sb1995_W1 and Sb1995_W2 in Fig. 2a and b, respectively), while the third wafer exhibited a

large variation (Sb1995_W3, Fig. 2c), with a continuous decrease in the intensity from the center towards the edge. Since the wafers are mounted symmetrically on the wafer carrier, the observed variations should not be related to non-uniformity of the flux. One possible explanation for the observed non-uniformity is in substrate mounting. Occasionally a substrate slips out of the thin lip which holds it in place during growth and a small thermal gradient occurs. A second possibility is that the sapphire wafers used for thermal isolation develop thermal non-uniformities as As and Sb compounds slowly accumulate on the backside. From the X-ray data of the InAsSb contact layer, it is apparent that during MBE growth there was a small temperature gradient across one of the wafers (Sb1995_W3 wafer which exhibited variation of PL intensity). Although this temperature gradient is not discernable in the X-ray pattern for the superlattice, we can use variation in InAsSb stoichiometry as a thermal mapper of the wafer temperature during growth (Fig. 3). The PL non-uniformity can then be attributed to the temperature gradient across the wafer. This theory is corroborated by previous PL studies, which show that superlattice PL strength is strongly influenced by the growth temperature [19– 21]. For the particular CBIRD structures studied in this paper, a strong variation of the PL intensity is observed for growth temperatures in the range of 390–410 °C. These observations indicate that PL is a necessary tool for the evaluation of superlattice wafers designed for high performance detector application. PL is nondestructive, and provides critical information not available in standard materials analysis techniques such as XRD, surface scanning and AFM. The information contained in PL is useful not only for determining the quality of grown wafers, but also as a feedback mechanism for improving the quality control of the wafer growth process itself. 4. Leakage current and noise We performed direct measurements of the noise spectra of high performance SL heterodiodes based on a variant CBIRD design [9] at different operational conditions to understand the effects of dark

Fig. 2. PL spectra of CBIRD detector material measured along the radius of three different 200 wafers showing: (a) and (b) small variation of the PL intensity indicating high uniformity (c) a strong radial decrease of the PL intensity indicating a less uniform wafer. Distances are given relative to the center of the wafer.

A. Soibel et al. / Infrared Physics & Technology 54 (2011) 247–251

249

Fig. 3. X-ray diffraction measurements along the radius of Sb1995_W3, 0–20 mm from the center of the wafer. The layer peaks of the substrate, bottom barrier, top barrier and the InAsSb contact layer are labeled with the letters S, BB, T and C, respectively. The shift of the InAsSb contact layer peak serves as an indicator of the temperature gradient across the wafer during growth. The shift corresponds to a change in Sb-content from 8% at the center of the wafer to 8.5% 20 mm from the center.

current and of the surface current on detector noise. We focused our study on two representative devices designated as d1 and d2, which were fabricated together by wet etching from the same CBIRD wafer (Sb1593). These devices have very similar differential resistance-area product of R0A = 1200 X cm2 (d1) and R0A = 1000 X cm2 (d2) at T = 77 K, but dark current in the device d2 is higher than in the device d1 (Fig. 4), and the higher dark current is attributed to detector mesa sidewall surface leakage current [22]. The bottom panel of Fig. 5 shows the current noise, in, of the device d1 at several applied bias voltages ranging from Vb = 0 V to Vb = 0.4 V. The noise spectra are relatively flat from 1 Hz to 5 kHz, showing the absence of 1/f noise in this device [23]. The shot noise in the device increases with an increase of the applied bias/current, as can be seen clearly from the noise spectral density at frequencies higher than 1 kHz. However, the general ‘‘flatness’’ of the noise spectra does not change with bias, and no onset of 1/f noise is observed. In contrast, the noise characteristics are profoundly different in the device d2 (Fig. 5, top). The noise amplitude is much larger than in device d1 and noise increases rapidly with the ap-

10

T=77K T 77K d1 d2

1 01 0.1 0 01 0.01

plied bias Vb, thus the noise in this device is much higher than can be expected from a simple theoretical estimation of the shot noise. Moreover, the noise spectra are frequency dependent even at zero bias. At Vb = 0.1 V and Vb = 0.17 V noise spectra have a 1=f 0:9 frequency dependence in the f = 100 Hz–4 kHz frequency range, but become almost frequency independent at lower frequencies. The observed noise has frequency dependence similar to that of 1/f noise in the limited frequency interval of f = 100 Hz–4 kHz and it becomes frequency independent at lower frequency. Such behavior is characteristic of flicker noise that is attributed to the surface states [24]. Indeed, the appearance of additional frequency noise associated with the surface states is consistent with the observation of surface leakage current that is also attributed to an electrical activity of the surface states. These results are a direct demonstration that intrinsically SL photodetectors do not exhibit 1/f noise. At the same time, our measurements clearly show that side-wall leakage current not only increases the shot noise by contributing to higher dark current but more importantly, it also introduces additional frequency-dependent noise, resulting in much higher noise in the detector. Since strongly frequency-dependent noise can be generated by side-wall leakage current, it is important to fabricate the high performance SL detectors and focal plane array (FPA) using the technology that can minimize the mesa side-wall leakage current. One way to achieve this result is by development of reliable sidewall passivation that can suppress the leakage current and prevent the onset of frequency-dependent noise. More details of the noise and gain measurement results on the CBIRD structure and a related device [9] can be found in Ref. [25].

1E-3 1E 4 1E-4

5. Dry etch processing

1E 5 1E-5 1E-6 E 1E-7 1E 7 -0.6 0.6 -0.4 0.4 -0.2 0.2 0.0

0.2

0.4

0.6

0.8

1.0

1.2

Vb(V) Fig. 4. Current–voltage (IV) characteristics of two CBIRD devices d1 and d2 at T = 77 K.

The CBIRD design resulted in significant improvements in dark current characteristics over homojunction SL detectors. Relative to the reduced dark current from within the detector, surface leakage current could now become the major contributor to the dark current. Fabrication must now accommodate this to maximize the full potential of the diodes. The etching technique for mesa isolation is important because it defines the interface at the sidewalls and forms a basis for any passivation or post-surface treatment meth-

250

A. Soibel et al. / Infrared Physics & Technology 54 (2011) 247–251

1

10

100

1000

1E-11

0V 0 .1 0 V 0 .1 7 V

0.5 5

Noisse (A No A/H Hzz )

1E-12 1E-13

Sb1593 d2 1E-11

0V 0 .2 0 V 0 .4 0 V

Sb1593 d1

1E-12 1E-13

1

10

100

1000

f(H z) Fig. 5. The current noise, in, vs. frequency of the devices d1 and d2 at several applied bias voltages as indicated on the graph. The dark current in the device d2 is higher than in the device d1, and the higher dark current, which is attributed to detector mesa sidewall surface leakage current, results in large frequency-dependent noise.

-3

10

Shallow Wet Shallow Dry D Deep D Dry

2

Darkk Cu Da Currren nt Den D ns sity y [A [A/c cm m]

ods. There are several different techniques used for mesa isolation, a common one being chemical wet etch [8,26,27]. However, the large degree of undercut and concave sidewall profiles from wet etching becomes unacceptable for small pixel sizes needed for FPA fabrication. A solution to the wet etching problems was high-density plasma etching, which is anisotropic due to the plasma sheath and ionized gas directionality. Chlorine-based plasma chemistry [28,29] yields fast etch rates and smooth morphologies. While this approach has been successful in the past, the CBIRD performance can be degraded by the mesa side wall surface roughness from plasma damage and preferential etching that stems from the low volatility of the InClx etch products formed during chlorinebased plasma etch. In this work, we studied the effects of four different etch techniques on performance of CBIRD detectors [30]. In comparing detectors fabricated from a third CBIRD by different etching techniques, we found that the best option was the combination BCl3/ Cl2/CH4/H2/Ar ICP dry etch. Fig. 6 compares the combination dry etch to wet etch yielding similar dark current densities with near-vertical and smooth sidewalls observed under SEM. Furthermore, the combination dry etch showed minimal change in dark current when scaling the detector size down towards smaller areas with a relatively flat behavior slope in the perimeter-to-area vs. resistance-area product RAeff plot. To verify the high surface quality, the leakage current for the same device with a different amount of exposed sidewall surface area was compared. The fact that the amounts of leakage current were the same for the shallow (1 lm) etched device compared to the deep etched (4 lm) device justifies the excellence of the combination dry etch. There is only a slight difference in I–V curve at lower biases (<175 mV), possibly due to the fact that in the shallow etched device, the bottom contact is closer to the depletion region. The amount of preferential etching was found to be a factor in lowering the dark current. The periodic structure of the superlattices adds an additional complication in requiring the same etch rates of the individual InAs, GaSb, and AlSb layers in order to prevent irregularity with the crystal surface termination. A differential etching rate would create ripples along the sidewalls, thereby becoming electrically active sites for generating additional dark current [21]. The electrically active sites, oxides, byproducts, and contaminants all contribute to dark current through interfacial states. Interface traps located within the band gap are sources for

-4 4

10

-5

10

-6

10

-7 7

10

-0.2 02

-0.1 01

0 0.0 0

0 0.1 1

0 0.2 2

0 0.3 3

0 0.4 4

0 0.5

V lt Voltage [V] Fig. 6. Comparison of dark current density curves for 200 lm square diodes measured at T = 77 K for an un-passivated CBIRD device with 10.3 lm cutoff.

carriers to tunnel through, inducing surface recombination and trap-assisted tunneling dark current. Another contributing factor is the presence of a non-zero surface potential. With a non-zero surface potential at the sidewall interface, band-bending occurs at the surface, and this results in accumulation/inversion of the majority carriers that then can create conductive leakage pathways parallel to the surface [21]. For LWIR (k > 10 lm) diodes, the amount of band bending becomes appreciable relative to the band gap of the device [31], so achieving near-zero surface potential can come from improper termination of the crystal lattice at the semiconductor-air interface where native oxides attach to unsatisfied bonds and form a secondary compound. Each gas in the combination BCl3/Cl2/CH4/H2/Ar ICP dry etch served a purpose. The methane/hydrogen is believed to be advantageous in the formation of a thin polymer layer, serving as a pseudo-passivation material whose low fixed charge properties have been shown to be beneficial to SL detector performance [32]. The BCl3 was introduced due to its efficacy in removal of unwanted native oxides and re-deposited byproducts; both of which have been shown to be contributing factors to high dark currents [33]. Finally, a slight amount of Cl2 was added to increase the etch rate and prevent erosion of the SiNx hard mask.

A. Soibel et al. / Infrared Physics & Technology 54 (2011) 247–251

6. Conclusions We have shown that photoluminescence can be a very useful tool for determining material quality of SL detector wafers. We have demonstrated that while intrinsic 1/f noise is absent in superlattice heterodiode, side-wall leakage current can become a source of strong frequency-dependent noise. Finally, we have developed a dry etch technique that enables the fabrication of small pixel without impeding their performance. These results are important steps in advancing the state-of-the-art of Sb-based infrared detector technology. Acknowledgments The authors thank S. Bandara for helpful discussions, and M. Tidrow, R. Liang, M. Herman, E. Kolawa, and P. Dimotakis for encouragement and support. The research described in this publication was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. Government sponsorship acknowledged. References [1] D.L. Smith, C. Mailhiot, Proposal for strained type II superlattice infrared detectors, J. Appl. Phys. 62 (1987) 2545–2548. [2] C.H. Grein, P.M. Young, H. Ehrenreich, Minority carrier lifetimes in ideal InGaSb/InAs superlattices, Appl. Phys. Lett. 61 (1992) 2905–2907. [3] J.B. Rodriguez, E. Plis, G. Bishop, Y.D. Sharma, H. Kim, L.R. Dawson, S. Krishna, nBn detectors based on InAs/GaSb type-II strain layer superlattice, Appl. Phys. Lett. 91 (2007) 043514. [4] J.L. Johnson, L.A. Samoska, A.C. Gossard, J.L. Merz, M.D. Jack, G.R. Chapman, B.A. Baumgratz, K. Kosai, S.M. Johnson, Electrical and optical properties of infrared photodiodes using the InAs/Ga1x In x Sb superlattice in heterojunctions with GaSb, J. Appl. Phys. 80 (1996) 1116–1127. [5] B.-M. Nguyen, D. Hoffman, E. K.-W. Huang, P.-Y. Delaunay, M. Razeghi, Background limited long wavelength infrared type-II InAs/GaSb superlattice photodiodes operating at 110 K, Appl. Phys. Lett. 93 (2008) 123502. [6] I. Vurgaftman, E.H. Aifer, C.L. Canedy, J.G. Tischler, J.R. Meyer, J.H. Warner, E.M. Jackson, G. Hildebrandt, G.J. Sullivan, Graded band gap for dark-current suppression in long-wave infrared W-structured type-II superlattice photodiodes, Appl. Phys. Lett. 89 (2006) 121114. [7] B.-M. Nguyen, S. Bogdanov, S. Abdollahi Pour, M. Razeghi, Minority electron unipolar photodetectors based on type II InAs/GaSb/AlSb superlattices for very long wavelength infrared detection, Appl. Phys. Lett. 95 (2009) 183502. [8] D. Z.-Y. Ting, C.J. Hill, A. Soibel, S.A. Keo, J.M. Mumolo, J. Nguyen, S.D. Gunapala, A high-performance long wavelength superlattice complementary barrier infrared detector, Appl. Phys. Lett. 95 (2009) 023508. [9] C.J. Hill, A. Soibel, D.Z.-Y. Ting, S.A. Keo, J.M. Mumolo, J. Nguyen, M. Lee, S.D. Gunapala, High temperature operation of long-wavelength infrared superlattice detector with suppressed dark current, Electron. Lett. 45 (2010) 1089–1090. [10] W.E. Tennant, Rule 07’’ revisited: still a good heuristic predictor of p/n HgCdTe photodiode performance?, J Electron. Mater. 39 (2010) 1030–1035. [11] D. Donetsky, S.P. Svensson, L.E. Vorobjev, G. Belenky, Carrier lifetime measurements in short-period InAs/GaSb strained-layer superlattice structures, Appl. Phys. Lett. 95 (2009) 212104. [12] D. Donetsky, G. Belenky, S. Svensson, S. Suchalkin, Minority carrier lifetime in type-2 InAs–GaSb strained-layer superlattices and bulk HgCdTe materials, Appl. Phys. Lett. 97 (2010) 052108.

251

[13] J. Pellegrino, R. DeWames, Minority carrier lifetime characteristics in type II InAs/GaSb LWIR superlattice n+ pp+ photodiodes, Proc. SPIE 7298 (2009) 72981U–4. [14] D. Z.-Y. Ting, C.J. Hill, A. Soibel, J. Nguyen, S.A. Keo, J.M. Mumolo, M.C. Lee, B. Yang, S.D. Gunapala, Antimonide superlattice barrier infrared detectors, Proc. SPIE 7419 (2009) Â;74190B. [15] D.R. Rhiger, R.E. Kvaas, S.F. Harris, B.P. Kolasa, C.J. Hill, D.Z. Ting, Characterization of barrier effects in superlattice LWIR detectors, Proc. SPIE 7660 (2010) 76601N. [16] R.K. Ahrenkiel, Minority carrier lifetime in III–V semiconductors, in: R.K. Ahrenkiel, M.S. Lundstrom (Eds.), Semiconductors and Semimetals, vol. 39, Academic Press Inc., San Diego, 1993, pp. 40–150. [17] L. Pavesi, M. Guzzi, Photoluminescence of AlxGa1xAs alloys, J. Appl. Phys. 75 (1993) 4779–4842. [18] C.S.R. Kurtz, J. Bajaj, D.D. Edwall, S.J.C. Irvine, Infrared photoluminescence characterization of long-wavelength HgCdTe detector materials, Semicond. Sci. Technol. 8 (1993) 941–945. [19] L. Burkle, F. Fuchs, J. Schmitz, W. Pletschen, Control of the residual doping of InAs/(GaIn)Sb infrared superlattices, Appl. Phys. Lett. 77 (2000) 1659–1661. [20] C.L. Canedy, W.W. Bewley, C.S. Kim, M. Kim, I. Vurgaftman, J.R. Meyer, Dependence of type II ‘‘W’’ mid-infrared photoluminescence and lasing properties on growth conditions, J. Appl. Phys. 94 (2003) 1347–1355. [21] H.J. Haugan, S. Elhamri, F. Szmulowicz, B. Ullrich, G.J. Brown, W.C. Mitchel, Study of residual background carriers in midinfrared InAs/GaSb superlattices for uncooled detector operation, Appl. Phys. Lett. 92 (2008) 071102-1– 071102-3. [22] Device sidewall surface leakage current is a well-known phenomenon in these devices and is typically evaluated from a dark current density dependence on an area/perimeter ratio. (see P.-Y. Delaunay et. al., Appl. Phys. Lett. 91 (2007) 091112). We attributed the extra dark current to the sidewall surface leakage based on the dark current vs. area/perimeter measurements performed for similar structures. [23] A small hump seen in the noise spectra near 100 Hz is attributed to the instrument noise since the hump size and shape is independent of detector bias. [24] A.V.D. Ziel, Noise: Source, Characterization, Measurements, Prentice-Hall, 1970 (Chapter 6). [25] A. Soibel, D. Z.-Y. Ting, C.J. Hill, M. Lee, J. Nguyen, S.A. Keo, J.M. Mumolo, Sarath D. Gunapala, Gain and noise of high-performance long wavelength superlattice infrared detectors, Appl. Phys. Lett. 96 (2010) 111102. [26] C.L. Canedy, E.H. Aifer, J.H. Warner, I. Vurgaftman, E.M. Jackson, J.G. Tischler, S.P. Powell, K. Olver, J.R. Meyer, W.E. Tennent, Controlling dark current in typeII superlattice photodiodes, Infrared Phys. Technol. 52 (2009) 326–334. [27] A.D. Hood, A.J. Evans, A. Ikhlassi, D.L. Lee, W.E. Tennant, LWIR strained-layer superlattice materials and devices at teledyne imaging sensors, J. Electron. Mater. 39 (2010) 1001–1006. [28] E. K.-W. Hunag, D. Hoffman, B.-M. Nguyen, P.-Y. Delaunay, M. Razeghi, Surface leakage reduction in narrow band gap type-II antimonide-based superlattice photodiodes, Appl. Phys. Lett. 94 (2009) 053506. [29] F. Fuchs, U. Weimar, E. Ahlswede, W. Pletschen, J. Schmitz, M. Walther, InAs/ Ga1xInxSb infrared superlattice photodiodes for infrared detection, Proc. SPIE 3287 (1998) 14–21. [30] J. Nguyen, A. Soibel, David Z.-Y. Ting, C.J. Hill, M.C. Lee, S.D. Gunapala, Low dark current long-wave infrared InAs/GaSb superlattice detectors, Appl. Phys. Lett. 97 (2010) 051108. [31] M. Herrera, M. Chi, M. Bonds, N.D. Browning, Joseph N. Woolman, R.E. Kvaas, S.F. Harris, D.R. Rhiger, C.J. Hill, Atomic scale analysis of the effect of the SiO2 passivation treatment on InAs/GaSb superlattice mesa sidewall, Appl. Phys. Lett. 93 (2008) 093106. [32] A. Hood, P.-Y. Delaunay, D. Hoffman, B.-M. Nguyen, Y. Wei, M. Razeghi, V. Nathan, Near bulk-limited R0A of long-wavelength infrared type-II InAs/GaSb superlattice photodiodes with polyimide surface passivation, Appl. Phys. Lett. 90 (2007) 233513. [33] A. Gin, Y. Wei, J. Bae, A. Hood, J. Nah, M. Razeghi, Passivation of type II InAs/ GaSb superlattice photodiodes, Thin Solid Films 447–448 (2004) 489–492.