January 1, 2015 / Vol 40, No / OPTICS LETTERS 45 Demonstration of type-II superlattice MWIR minority carrier unipolar imager for high operation temperature application Guanxi Chen, Abbas Haddadi, Anh-Minh Hoang, Romain Chevallier, and Manijeh Razeghi* Department of Electrical Engineering and Computer Science, Northwestern University, Evanston, Illinois 60208, USA *Corresponding author: razeghi@eecs.northwestern.edu Received September 12, 2014; accepted November 13, 2014; posted November 24, 2014 (Doc ID 222761); published December 18, 2014 An InAs/GaSb type-II superlattice-based mid-wavelength infrared (MWIR) 320 × 256 unipolar focal plane array (FPA) using pMp architecture exhibited excellent infrared image from 81 to 150 K and ∼98% operability, which illustrated the possibility for high operation temperature application At 150 K and −50 mV operation bias, the 27 μm pixels exhibited dark current density to be 1.2 × 10−5 A∕cm2 , with 50% cutoff wavelength of 4.9 μm, quantum efficiency of 67% at peak responsivity (4.6 μm), and specific detectivity of 1.2 × 1012 Jones At 90 K and below, the 27 μm pixels exhibited system limited dark current density, which is below × 10−9 A∕cm2 , and specific detectivity of 1.5 × 1014 Jones From 81 to 100 K, the FPA showed ∼11 mK NEDT by using F/2.3 optics and a 9.69 ms integration time © 2014 Optical Society of America OCIS codes: (040.1240) Arrays; (040.1490) Cameras; (040.5160) Photodetectors; (040.5350) Photovoltaic; (110.3080) Infrared imaging; (250.0040) Detectors http://dx.doi.org/10.1364/OL.40.000045 The detection of mid-wavelength infrared (MWIR) radiation, whose radiation window is between and μm, has wide applications from military to civilian, such as aerial surveillance and reconnaissance, object identification in harsh environments, vascular and cancer detection, and industrial process monitoring As the requirements of the high-sensitivity, high-resolution, and multifunctional infrared imager keep increasing, the type-II InAs/GaSb superlattice (T2SL) has proved to be an attractive material system to meet all these requirements because of its unique flexibility in band structure engineering [1] In the recent past, the development of T2SL-based infrared detectors and imagers is very rapid and different achievements have been realized, such as innovative device structure designs [2–5], suppression of different dark current mechanisms [6,7], improvement of surface treatments [8–13], and realization of single- and dualband infrared imagers [14,15] Although the InAs/GaSb T2SL material system has shown great potential and has made rapid improvements over the years, it still needs to overcome several obstacles One of them is raising the operating temperature of the MWIR focal plane array (FPA) to remove the cryogenic cooling system To achieve high signal-to-noise ratio (SNR) in an infrared imager, photodetectors with high quantum efficiency (QE) and low dark current density are desired Realizing an MWIR infrared imager in high operating temperatures (HOT) is very challenging because near room temperature objects have much less irradiance in the MWIR regime, which means that the incoming MWIR photo-signal is much weaker than other infrared regimes Therefore, it is crucial to lower a detector dark current to have a high SNR ratio Tunneling, generation-recombination (G-R), and diffusion current are the three most common dark current mechanisms that need to be suppressed Heterojunction p-π-M-n architecture was introduced to suppress the tunneling current, and excellent MWIR FPA was achieved based on this architecture [7] Moreover, since diffusion current limited 0146-9592/15/010045-03$15.00/0 MWIR T2SL-based photovoltaic devices were desired for having lower dark current density, a minority carrier unipolar device was proposed, which was a hybrid structure between a photovoltaic and photoconductive device with a potential barrier to block the majority carrier transport but no built-in electrical field present [16] Since the minority carrier device does not have a built-in electrical field, the G-R current from the depletion region is supposed to be eliminated and the device should be diffusion limited, resulting in lower dark current density There are two types of unipolar device, the nBn [4] and the pMp [17,18] structures They are different in their type of minority carrier transport–holes for the nBn architecture and electrons for the pMp architecture Because of its type of minority carrier, pMp architecture usually has higher quantum efficiency with the same absorber thickness, and high performance FPA is expected However, because the p-type T2SL material is more sensitive to fabrication conditions, no FPA has been demonstrated based on the pMp architecture after the report of MWIR pMp structure single photodetector In this Letter, we report the fabrication and measurement results of a high performance MWIR FPA based on pMp architecture and its potential for HOT application The pMp architecture photodetector consists of a 1.5 μm thick InAs0.91 Sb0.09 etch stop layer, a 0.5 μm thick unintentional doped p-type bottom contact, a 0.5 μm thick lightly n-doped larger bandgap M-structure [3], and a μm thick lightly p-type active region Silicon and Beryllium are used for the n-type and p-type dopant, respectively The p-contact and p-type active region have the same bandgap, and the schematic diagram of the device architecture is shown in Fig 1(a) The details of the superlattice design and doping level of each region can be found in [18] The total thickness of the entire pMp structure, including the InAsSb etch stop layer, is 6.2 μm After the material growth, an imaging FPA with 27 μm pixels and a pixel test chip (PTC) with pixels ranging from 27 to 400 μm were fabricated simultaneously © 2015 Optical Society of America 46 OPTICS LETTERS / Vol 40, No / January 1, 2015 Fig (a) Schematic diagram of the MWIR pMp photodetector under small reverse operation bias (b) The average dark current density of five 27 μm pixels of the pixel test chip processed with the FPA from 77 to 250 K The pixels were delineated by inductively coupled plasma (ICP) etching and citric-acid based wet etching, and had effective fill-factors of ∼80% After etching, top and bottom metal contacts were deposited simultaneously by electron beam metal evaporation, followed by a 600 nm thick SiO2 using plasma-enhanced chemical vapor deposition (PECVD) Small windows of dielectric layer covering the top and bottom contacts were opened by using CF4 :Ar plasma in an electron cyclotron resonance-reactive ion etching (ECR-RIE) system to remove the SiO2 Indium bumps were then deposited on the small window in a thermal evaporator The PTC was flip-chip bonded to a silicon fan-out, which allowed direct pixel characterization, while the FPA was hybridized to a 256 × 320 Indigo ISC9705 read-out integrated circuit (ROIC) Before completing the rest of the FPA fabrication, the electrical performance of the PTC was first characterized from 77 to 250 K to ensure good fabrication Shown in Fig 1(b) is the average dark current density of five 27 μm pixels At 150 K and −50 mV, the dark current density and differential resistance area product (RA) of the PTC are 1.2 × 10−5 A∕cm2 and 22 kΩ∕cm2 , respectively At 90 K and below, the dark current is limited by the system floor level of 100 fA, which creates the noise behavior in the dark current density curve at a low operation bias region From the evolution of the shape of the I–V curve, at 150 K and below, the pMp structure is not diffusion limited, and its dark current density is slightly higher than the one in Ref [17] This is because the surface leakage current still limits the performance of the MWIR photodetector at a low temperature, which is similar to the one reported in Refs [11,19] At 150 K and above, the diffusion current is the limiting dark current mechanism After the electrical characterization, the PTC and the FPA were both underfilled with epoxy, and their substrates were removed up to the InAsSb etch stop layer No antireflection coating was applied The optical response from backside illumination of 400 μm pixels from the PTC was characterized As shown in Fig 2(a), at 77 K, the QE of MWIR pMp photodetector saturates at −50 mV Fig (a) Quantum efficiency of the 400 μm pixel from the PTC at 77 K and −50 mV operation bias (b) The saturated QE at peak responsivity from 77 to 250 K (c) The evolution of peak detectivity with temperature The peak detectivity crosses the calculated BLIP line at 165 K because of the absence of a built-in electrical field to separate the photo-generated carriers The peak responsivity happens at 4.2 μm with a value of 2.4 A/W, and the QE at peak responsivity is 68.6% Fabry-Perot resonance is observed because of the cavity formed between the top-metal contact and the interface between the InAsSb etch stop layer and the air The extracted cavity length is ∼6.3 μm, which is about the same as the total thickness of the entire pMp structure, and the small difference is in the error tolerance of measurement and calculation Figure 2(b) shows the saturated QE at peak-responsivity wavelength between 4.2 and 4.9 μm from 77 to 250 K The QE stays in a similar level in the whole temperature range, while at a higher temperature, larger operation bias is required to fully extract the photo-generated carriers, which results in higher noise The peak specific detectivity (D ) at different temperatures and the background limited performance (BLIP) detectivity of 3.5 × 1011 Jones are shown in Fig 2(c) The BLIP temperature is determined as the temperature at which the detectivity of the device is equal to that of an ideal photodiode with 100% QE and a 2π field-of-view (FOV) in a 300 K background As the temperature increases, the peak detectivity of the pMp decreases from 1.5 × 1014 Jones at 77 K to 7.7 × 109 Jones at 250 K, and intersects with the BLIP detectivity at 165 K Because of the limited of the measurement system, at low operation bias, the dark current densities at 77 and 90 K are similar, which is the noise level of the system and results in a similar level of D The noise equivalent temperature difference (NEDT), a measure of the FPA sensitivity, was measured by varying the blackbody temperature between 20°C and 30°C and mounting a MWIR Janos ASIO lens with F-number of 2.3 The signal and noise of each pixel were calculated as the mean and standard deviation over 100 frames From the NEDT distribution shown in Fig 3, the median value of NEDT using 9.69 ms integration time was 11 mK at 81 K and 15 m K at 110 K The NEDT distributions from 81 to 110 K not show any tail in logarithmic scale, which is an indication of uniform processing over the whole array This results in high operability of the FPA, where 98% of pixels are operable As shown in the inset January 1, 2015 / Vol 40, No / OPTICS LETTERS 47 camera is 1.6 × 1012 Jones, and the BLIP temperature is 150 K DBLIP θ Fig NEDT histogram of the FPA at 81 K and 110 K Inset: the minimum achievable NEDT from 81 to 140 K of Fig 3, the minimum achievable NEDT does not vary significantly below 120 K, changing from 11 mK at 81 K to 23 mK at 120 K At 120 K and above, the NEDT starts to increase and reaches 93 mK at 140 K because of the fast increment of the dark current density, which generates higher noise As shown in Fig 4, under 9.69 ms integration at 81 K, an excellent infrared human body image was obtained, and from 81 to 120 K, the image quality is excellent Although the quality of the image becomes poor at higher temperatures, human blood veins can still be clearly observed as the temperature rises to 150 K, which indicates its high potential for HOT application The decrement of the image quality at higher operating temperature attributes to the increment of the dark current density The difference between the calculated BLIP temperature and the FPA camera is because of the difference of the FOV The FOV of the FPA camera is 24.5 degrees and its BLIP detectivity can be calculated from Eq (1), where θ is a cone angle With BLIP detectivity at 2π FOV of 3.5 × 1011 Jones, the BLIP detectivity of the FPA Fig Excellent infrared image obtained at 81 K with 9.69 ms integration time Human blood veins can be seen up to 150 K DBLIP 2π sin θ∕2 (1) In summary, a high-performance MWIR minority carrier unipolar FPA using pMp architecture based on InAs/GaSb type-II superlattice was demonstrated with 98% operability Excellent infrared human body images were obtained from 81 to 120 K with a BLIP temperature at 150 K with 24.5 degrees of FOV The success of the MWIR FPA using pMp architecture opens another possibility to raise the T2SL infrared camera operation temperature The authors would like to acknowledge the Walter P Murphy fellowship, DARPA, the Army Research Laboratory (ARL), the Night Vision and Electronic Sensor Directorate (NVESD), the Air Force Research Laboratory, and NASA for their support, interest, and encouragement References Y Wei and M Razeghi, Phys Rev B 69, 085316 (2004) E H Aifer, J G Tischler, J H Warner, I Vurgaftman, W W Bewley, J R Meyer, J C Kim, L J Whitman, C L Canedy, and E M Jackson, Appl Phys Lett 89, 053519 (2006) B.-M Nguyen, M Razeghi, V Nathan, and G J Brown, Proc SPIE 6479, 64790S (2007) J B Rodriguez, E Plis, G Bishop, Y D Sharma, H Kim, L R Dawson, and S Krishna, Appl Phys Lett 91, 043514 (2007) D Z.-Y Ting, C J Hill, A Soibel, S A Keo, J M Mumolo, J Nguyen, and S D Gunapala, Appl Phys Lett 95, 023508 (2009) H Mohseni, V I Litvinov, and M Razeghi, Phys Rev B 58, 15378 (1998) S A Pour, E K Huang, G Chen, A Haddadi, B.-M Nguyen, and M Razeghi, Appl Phys Lett 98, 143501 (2011) E Plis, J B Rodriguez, S J Lee, and S Krishna, Electron Lett 42, 1248 (2006) R Rehm, M Walther, F Fuchs, J Schmitz, and J Fleissner, Appl Phys Lett 86, 173501 (2005) 10 F Szmulowicz and G J Brown, Infrared Phys Technol 53, 305 (2010) 11 G Chen, B.-M Nguyen, A M Hoang, E K Huang, S R Darvish, and M Razeghi, Appl Phys Lett 99, 183503 (2011) 12 G Chen, A M Hoang, S Bogdanov, A Haddadi, S R Darvish, and M Razeghi, Appl Phys Lett 103, 223501 (2013) 13 G Chen, A M Hoang, and M Razeghi, Appl Phys Lett 104, 103509 (2014) 14 M Razeghi, “Focal plane arrays in type-II superlattice,” U.S patent 6,864,552 (March 8, 2005) 15 E K Huang, M A Hoang, G Chen, S R Darvish, A Haddadi, and M Razeghi, Opt Lett 37, 4744 (2012) 16 S Maimon and G W Wicks, Appl Phys Lett 89, 151109 (2006) 17 B.-M Nguyen, S Bogdanov, S A Pour, and M Razeghi, Appl Phys Lett 95, 183502 (2009) 18 B.-M Nguyen, G Chen, A M Hoang, S A Pour, S Bogdanov, and M Razeghi, Appl Phys Lett 99, 033501 (2011) 19 G Chen, E K Huang, A M Hoang, S Bogdanov, S R Darvish, and M Razeghi, Appl Phys Lett 101, 213501 (2012) ... 1, 2015 Fig (a) Schematic diagram of the MWIR pMp photodetector under small reverse operation bias (b) The average dark current density of five 27 μm pixels of the pixel test chip processed with... backside illumination of 400 μm pixels from the PTC was characterized As shown in Fig 2(a), at 77 K, the QE of MWIR pMp photodetector saturates at −50 mV Fig (a) Quantum efficiency of the 400 μm pixel... of the limited of the measurement system, at low operation bias, the dark current densities at 77 and 90 K are similar, which is the noise level of the system and results in a similar level of