In this paper, we report on the characterization of InAs/GaAsSb type-II superlattice long wavelength infrared photodiodes grown on InAs substrates by molecular-beam epitaxy and also present the device performance comparison with the superlattice devices grown on GaSb substrates. These devices with PIN structures had a 100% cutoff wavelength of 10 μm. The dark current density of InAs-based device at −30 mV reverse bias was 4.01 × 10−4 A/cm2 and the resistance-area product at zero bias (R0A) was 36.9 Ωcm2. The dark current density of GaSb-based device is higher more than one order of magnitude than that of InAs-based device. The temperature-dependence and bias-dependence of the dark current are studied experimentally and correlated to the theory. Good agreement was achieved between the measured I-V curves and the simulated ones, and between the experimental and theoretically predicted differential resistance values. Compared with InAs-based superlattice device, the generation-recombination current of GaSb-based device is larger and dominates in a wider temperature range due to shorter carrier lifetime and higher defect density.
© 2017 Optical Society of America
InAs/GaSb type-II superlattice (T2SL)-based photodetectors have been recognized as a viable alternative for imaging systems operating in the mid- and long-infrared (IR) spectral range (λ>3μm) due to the flexibility of bandgap engineering, reduced tunneling currents, low auger recombination rates, and similar absorption coefficient compared to the commonly used HgCdTe-based photodetectors. Moreover, T2SL materials are based on mechanically robust III-V material platform which improves the feasibility of the material and device processing.
High performance detectors based on T2SL with operation wavelengths spanning a wide IR range have been demonstrated by different groups [1–7]. However, despite these progresses, the InAs/GaSb T2SL photodetectors are still suffered from their short minority carrier lifetimes [8–10]. It was shown that the short minority carrier lifetime in the T2SL structures is mainly caused by the Shockley-Read-Hall (SRH) recombination [8, 10], which in turn is caused by defects in materials. Therefore, it is important and desired to grow low defect superlattice materials with high crystal perfection for device applications. The conventional InAs/GaSb superlattices are mostly grown on GaSb substrates with InSb as the interface layers. The typical growth temperature for GaSb-based InAs/GaSb superlattice is around 400 °C. Such a low growth temperature is probably the reason for high defect density and short minority lifetimes .
To obtain high quality T2SL materials and enhance the SRH lifetimes, we proposed and carried out the investigations of InAs-based InAs/GaAsSb superlattices [12,13]. By growing T2SL materials on InAs substrates, there is no need to grow InSb interface layers for strain compensation. This allows us to significantly increase the growth temperature of T2SL materials. A higher growth temperature is very helpful to promote two-dimensional epitaxy and improve the crystal quality, which can effectively reduce the probability of SRH recombination and increase the minority carrier lifetime. Moreover, InAs layer thickness mainly determines the cutoff wavelength of T2SL detectors. By growing the superlattices on InAs substrates, it will be very straightforward to increase the InAs layer thickness in each period to expand the cutoff wavelength. Therefore, the InAs-based InAs/GaAsSb superlattice could be another material choice for the long and very long wavelength infrared detectors.
In this paper, we compared the material and device performance of InAs-based InAs/GaAsSb and GaSb-based InAs/GaSb superlattices with the same PIN structure and similar cutoff wavelength, and demonstrated that the InAs-based superlattice devices had a higher performance and could be used for infrared detection.
The InAs-based T2SL device structures used in the study were grown on undoped, residually n-type (100) InAs substrates by solid-source molecular-beam epitaxy using valved cracker cells for arsenic and antimony. Details of the material growth have been explained previously . The device structures were grown at 480°C measured by a calibrated infrared pyrometer. First, a 1.0 μm Si-doped InAs buffer layer was deposited. Then, a 0.5 μm thick Si-doped 20ML InAs/9ML GaAsSb superlattice was grown, followed by a 2.5 μm thick non-intentionally doped 20ML InAs/9ML GaAsSb absorber region. Finally, a 0.5 μm thick Be-doped 20ML InAs/9ML GaAsSb superlattice was grown and capped with a 50 nm Be-doped GaSb top contact layer. The growth rates were 1ML/s for GaAsSb and 0.75ML/s for InAs, respectively. For strain compensation, we introduced 9% As into the GaSb layers to obtain GaAs0.09Sb0.91 ternary compound which was lattice-matched to the InAs substrates. The interfaces between the ternary and the binary layers were formed naturally by switching the source shutters without insertion of additional layers, avoiding the complicated interface scheme that was used for the MBE growth of T2SLs on GaSb substrates. The GaSb-based device structures consisted of a 0.5 μm thick Be-doped contact, a 2 μm thick nominally undoped absorber region and a 0.5 μm thick Si-doped contact. The total structure was sandwiched between a 1.0 μm p-doped GaSb buffer layer and a 50 nm n-doped InAs capping layer. For comparison with the InAs-based devices, one period of the absorber region had 12ML InAs and 9ML GaSb to have a similar cutoff wavelength. The structural quality and the lattice mismatch of these materials were characterized by high resolution X-ray diffraction (HRXRD).
These PIN structures were then processed into single-element devices using standard optical lithography and wet chemical etching. Ti/Pt/Au was used to define the top and bottom Ohmic contacts by e-beam evaporation and liftoff. The surfaces and sidewalls of the devices in this paper were not passivated by dielectric layers. The schematic of both InAs-based and GaSb-based devices were presented in Fig. 1. Optical response measurements of these devices were performed at 80 K. Current-voltage (I-V) measurements were carried out at temperatures ranging from 12 K to 180 K.
3. Results and discussion
Figure 2 shows the typical X-ray diffraction pattern of both InAs-based and GaSb-based device structures. Sharp diffraction peaks up to fifth order for the two kinds of superlattices can be observed clearly. The lattice mismatch between the superlattices and the substrates for both InAs-based and GaSb-based superlattices were 4.9 × 10−5 and 6.5 × 10−4, respectively. The full width at half maximum (FWHM) of the first-order peaks for InAs-based and GaSb-based superlattices were 25.2” and 21.6”, respectively. These results indicate that the superlattices grown on either InAs or GaSb substrates all have good crystal quality.
Figure 3(a) shows the optical spectra of InAs-based and GaSb-based superlattice photodiodes measured at 80 K. These devices have similar 50% cutoff wavelength at 9 μm and 100% cutoff wavelength at 10 μm. Figure 3(b) shows the calculated cutoff wavelength and the experimental results at 80 K as a function of InAs layer thickness in one period. It is obvious that the cutoff wavelength of InAs-based device is shorter than that of the GaSb-based device when they have the same InAs and GaSb layer thicknesses in one period. This was mainly caused that the bottom of InAs conduction band lines up below the top of the GaAs0.1Sb0.9 valence band with a break in the gap of 70 meV, while the “broken gap” is 150 meV in the InAs/GaSb-on-GaSb superlattice. Therefore, the cutoff wavelength moves to a shorter region. We have already shown that it is very straightforward to realize longer cutoff wavelength by simply increasing InAs layer thickness in the InAs-based T2SL structures .
A measured dark current density-bias voltage (J-V) curve and its associated dynamic resistance-area product (RA) and voltage (V) curve at 80 K for both InAs-based and GaSb-based T2SL devices were shown in Fig. 4. The dark current densities at −30 mV and the resistance-area product at zero bias (R0A) were 4.01 × 10−4 A/cm2, 36.9 Ωcm2 for InAs-based device and were 9.05 × 10−3 A/cm2, 1.4 Ωcm2 for GaSb-based device, respectively. It is obvious that the dark current density of GaSb-based PIN device is higher more than one order of magnitude than that of InAs-based device with the same device structure and similar cutoff wavelength. It should be noted that the absorber region thickness has almost no effect on the device’s dark current because the depletion region width of about 200 nm calculated according to our experimental parameters is far more thinner than that of the absorber region. In order to study the major components of the dark current at 80 K, the current-voltage characteristics of both the InAs-based and GaSb-based T2SL devices were modeled, as shown in Fig. 5. The modeled bandgap, carrier lifetimes, trap energy level, and defect density were 127.4 meV, 10 ns, 100 meV, 2.4 × 1012 cm−3 for InAs-based device and 124 meV, 1.5 ns, 84 meV, 6.8 × 1013 cm−3 for GaSb-based device, respectively. The fitted back-ground carrier concentration was 3.4 × 1015 cm−3 for InAs-based device and 4.5 × 1015 cm−3 for GaSb-based device, respectively. The effective masses were taken as me = 0.03m0 and mh = 0.4m0, where m0 was the electron mass in vacuum. The surface leakage current can be excluded according to our measured results that the dark current density is almost identical across multiple diode sizes. From the fitting parameters, we can see that though the reason why the carrier lifetime is lower than those reported [8–10] is still not well understood, the carrier lifetime of InAs-based device is nearly 7 times higher than that of GaSb-based device while the defect density of InAs-based device is more than one order lower than that of GaSb-based device. Table 1 presents the measured and fitted dark current density values at −30 mV. Due to the similar cutoff wavelength, the diffusion current density is closed to each other. However, the GR current density in GaSb-based device is far larger than that in InAs-based device, which is mainly caused by the shorter GR lifetime and the higher defect density according to the fitting parameters. The change of the band-to-band tunneling rate in Fig. 5 is mainly caused by the difference of both the modeled bandgap and the back-ground carrier concentration of the two structures. The high trap-assisted tunneling for GaSb-based device is mainly caused by the high defect density compared with that of InAs-based device.
Figure 6 shows the dark current densities measured at different temperatures (solid lines) and their correspondingly fitted values (dotted lines) for InAs-based and GaSb-based devices, respectively. It can be seen the fitted values are coincided well with the experimental ones for the whole range of temperatures and biases considered. Figure 7 shows the measured and theoretical R0A product values as a function of the inverse temperatures. According to the fitting, the R0A values of InAs-based device shows a diffusion-limited behavior for temperatures above 75 K, and then a GR-limited behavior from 65 to 75 K. Below 65 K, the values of R0A increases even more slowly, and the diffusion currents and the GR contributions can be made negligibly small as compared to the tunneling contributions. The activation energy of the GR current is determined to be 63 meV. For GaSb-based device, the dark current is mainly due to the GR for temperature above 65 K. Even at temperature up to 140 K, the GR still is the dominant mechanism. Below 65 K, the trap-assisted tunneling current begins to dominate. The activation energy of the GR current for the GaSb-based device is determined to be 62 meV. These extracted activation energy is consistent with half of those fitted bandgap. Compared with InAs-based superlattice device, the GR of GaSb-based device is larger and dominates in a wider temperature range, which may be caused by the shorter carrier lifetime and the higher defect density. This coincides with the fitting results obtained above. On the contrary, the limitation of GR current shows that the InAs-based T2SL materials may have lower defect density and longer carrier lifetime, which could make them be another choice for the long and very long wavelength infrared detectors.
In summary, two kinds of LWIR superlattice device structures were grown by MBE system on InAs and GaSb substrates, respectively. The FWHM of the first order satellite peak and the lattice mismatch were 25.2″, 4.9 × 10−5 for InAs-based superlattices and 21.6″, 6.5 × 10−4 for GaSb-based superlattices. The 50% cutoff wavelength of these devices was 9.0 μm. The dark current density at −30 mV bias and the resistance-area product at zero bias was 4.01 × 10−4 A/cm2, 36.9 Ωcm2 for InAs-based superlattice device and was 9.05 × 10−3 A/cm2, 1.4 Ωcm2 for GaSb-based superlattice device. Compared with the GaSb-based device, the dark current density of InAs-based device is lower more than one order of magnitude. Good agreement was achieved between the measured I-V curves and the simulated ones, and between the experimental and theoretically predicted differential resistance values. As expected, the GR current of GaSb-based device is larger a lot than that of InAs-based device and dominates in a wider temperature range due to shorter carrier lifetime and higher defect density.
This work was supported by the National Natural Science Foundation of China (Grant Nos. 61505237, 61176082, 61290302, 61534006 and 61505235), the National Key Research and Development Program of China (Grant No. 2016YFB0402403) and the Shanghai Natural science foundation (Grant Nos. 15ZR1445600 and 16ZR1447900).
References and links
1. D.-Y. Ting, C. J. Hill, A. Soibel, S. A. Keo, J. M. Mumolo, J. Nguyen, and S. D. Gunapala, “A high-performance long wavelength superlattice complementary barrier infrared detector,” Appl. Phys. Lett. 95(2), 023508 (2009). [CrossRef]
2. B. Nguyen, D. Hoffman, P. Delaunay, and M. Razeghi, “Dark current suppression in type II InAs/GaSb superlattice long wavelength infrared photodiodes with M-structure barrier,” Appl. Phys. Lett. 91(16), 163511 (2007). [CrossRef]
3. M. Razeghi, E. K. Huang, B.-M. Nguyen, S. Abdollahi Pour, and P.-Y. Delaunay, “Type-II antimonide-based superlattices for the third generation infrared focal plane arrays,” Proc. SPIE 7660, 76601F (2010). [CrossRef]
4. M. Razeghi, A. Haddadi, A. M. Hoang, E. K. Huang, G. Chen, S. Bogdanov, S. R. Darvish, F. Callewaert, and R. McClintock, “Advances in antimonide-based Type-II superlattices for infrared detection and imaging at center for quantum devices,” Infrared Phys. Technol. 59(13), 41–52 (2013). [CrossRef]
5. A. Khoshakhlagh, L. Höglund, D. Z. Ting, C. J. Hill, S. A. Keo, A. Soibel, J. Nguyen, and S. D. Gunapala, “High performance long-wave type-II superlattice infrared detectors,” J. Vac. Sci. Technol. B 31(3), 03C122 (2013).
6. C. Asplund, R. Marcks von Würtemberg, D. Lantz, H. Malm, H. Martijn, E. Plis, N. Gautam, and S. Krishna, “Performance of mid-wave T2SL detectors with heterojunction barriers,” Infrared Phys. Technol. 59(6), 22–27 (2013). [CrossRef]
7. R. Rehm, M. Walther, J. Schmitz, F. Rutz, J. Fleißner, R. Scheibner, and J. Ziegler, “InAs/GaSb superlattices for advanced infrared focal plane arrays,” Infrared Phys. Technol. 52(6), 344–347 (2009). [CrossRef]
8. D. Donetsky, S. Svensson, L. E. Vorobjev, and G. Belenky, “Carrier lifetime measurements in short-period InAs/GaSb strained-layer superlattice structures,” Appl. Phys. Lett. 95(21), 212104 (2009). [CrossRef]
9. D. Donetsky, G. Belenky, S. Svensson, and S. Suchalkin, “Minority carrier lifetime in type-2 InAs-GaSb strained-layer superlattices and bulk HgCdTe materials,” Appl. Phys. Lett. 97(5), 052108 (2010). [CrossRef]
10. B. C. Connelly, G. D. Metcalfe, H. Shen, and M. Wraback, “Direct minority carrier lifetime measurements and recombination mechanisms in long-wave infrared type II superlattices using time-resolved photoluminescence,” Appl. Phys. Lett. 97(25), 251117 (2010). [CrossRef]
11. S. P. Svensson, D. Donetsky, D. Wang, H. Hier, F. J. Crowne, and G. Belenky, “Growth of type II strained layer superlattice, bulk InAs and GaSb materials for minority lifetime characterization,” J. Cryst. Growth 334(1), 103–107 (2011). [CrossRef]
12. F. Wang, J. Chen, Zh. Xu, Y. Zhou, and L. He, “InAs-based type-II superlattice long wavelength photodetectors,” Proc. SPIE 9755, 975519 (2016). [CrossRef]
13. F. Wang, J. Chen, Zh. Xu, Y. Zhou, and L. He, “InAs-based InAs/GaAsSb type-II superlattices: Growth and characterization,” J. Cryst. Growth 416, 130–133 (2015). [CrossRef]