Open Access
Add to BibSonomy
Abstract
Mid-IR is a useful wavelength range for both science and military applications due to its low atmospheric attenuation and ability to be used for passive detection. However, many solutions for detecting light in this spectral region need to be operated at cryogenic temperatures as their required narrow bandgaps suffer from carrier recombination and band-to-band tunneling at room temperature leading to high dark currents. These problems can be alleviated by using a separate absorption, charge, and multiplication avalanche photodiode. We have recently demonstrated such a device with a 3-µm cutoff using Al0.15In0.85As0.77Sb0.23, as the absorber, grown on GaSb. Here we investigate Al0.15In0.85As0.77Sb0.23 as a simple PIN homojunction and provide metrics to aid in future designs using this material. PL spectrum measurements indicate a bandgap of 2.94 µm at 300 K. External quantum efficiencies of 39% and 33% are achieved at 1.55 µm and 2 µm respectively. Between 180 K and 280 K the activation energy is ∼0.22 eV, roughly half the bandgap of Al0.15In0.85As0.77Sb0.23, indicating thermal generation is dominant.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Mid-wavelength infrared (mid-IR) is a spectral range generally defined as 2- to 5-µm light and is useful for both science and military applications. Low atmospheric attenuation in this optical window is useful for astronomy and infrared imaging. It is also well suited for thermal detection, which is beneficial for night vision systems [1]. Three materials systems are common for mid-IR detection InAs [2], InSb [3], and HgCdTe [4]. By taking advantage of the process of impact ionization, APDs realized in these materials can achieve higher sensitivity than typical photodiodes. However, since these materials systems all have narrow bandgaps, they exhibit high dark currents at room temperature due to carrier recombination and as a result are typically operated at cryogenic temperatures to limit the effect. This limitation is inconvenient for certain applications as cryogenic coolers are complex and bulky. Additionally, the narrow bandgap leads to the onset of band-to-band tunneling at low electric fields. Although, this problem can be alleviated by growing thick structures to reduce the electric field, the dark current with typically suffer as a result.
The use of separate absorption, charge, and multiplication (SACM) structures, which decouple absorption and multiplication, has been successful in reducing the carrier recombination and band-to-band tunneling components of dark current in APDs. This design enables low electric fields in the absorber, limiting tunneling, and high electric fields in the multiplier to promote impact ionization. Recently, we have demonstrated an SACM APD with a cutoff of 3 µm [5] using the Al1-xInxAs1-ySby/GaSb digital alloy materials system. By varying the Al concentration, the bandgap energy of this material can be tuned for detection from the visible to ∼5 µm while maintaining a lattice match to the GaSb substrate. This materials system is direct bandgap up to an Al concentration of ∼80% [6]. APDs made in this materials system have demonstrated high gain and low excess noise [7–10], comparable to that of Si APDs. We have also demonstrated nBn photodetectors operating at 2 µm using Al0.30In0.70As0.67Sb0.33 [11,12], and 3 µm using Al0.15In0.85As0.77Sb0.23 [13]. The aforementioned 3-µm cutoff SACM also uses Al0.15In0.85As0.77Sb0.23 as the absorber. Here, we report material characteristics and device characteristics of Al0.15In0.85As0.77Sb0.23 that may be found useful for future photodetector designs incorporating the material for operation in the mid-IR. Material characteristics presented include refractive index, extinction coefficient, absorption coefficient, and the photoluminescence (PL) spectrum. Device characteristics of Al0.15In0.85As0.77Sb0.23 PIN photodiodes presented include current-voltage, capacitance-voltage, dark current density versus temperature, responsivity, activation energy, and ideality factor.
2. Growth and fabrication
Al0.15In0.85As0.77Sb0.23 PIN epitaxial layers were grown as a digital alloy on an n-type GaSb substrate via molecular beam epitaxy as described previously [6]. Figure 1(a) shows a schematic cross-section of the device. Figure 1(b) shows an X-ray diffraction pattern of the grown with the substrate and superlattice fringe peaks labeled. Figure 2 shows the photoluminescence of the crystal measured at several temperatures. At 78 K the peak is at 2.89 µm with a full width at half maximum of 45 meV (6.7 kT). At 300 K the peak is at 2.94 µm with a full width at half maximum of 81 meV (3.1 kT). The peak at 300 K falls between previously reported PL intensity peaks for Al0.09In0.91As0.31Sb0.69 (3.35 µm) and Al0.19In0.81As0.33Sb0.67 (2.64 µm) [6]. We suspect the bump present in the PL spectrum around 3.25 µm is due to atmospheric absorption. Ellipsometry was performed with a J.A. Woollam VASE Ellipsometer on the bare epitaxy prior to device fabrication. The wavelength range extended from 280 to 2450 nm with measurement angles from 65° to 75° in 5° increments. Figure 3(a) shows the measured Ψ and Δ at the three angles and their corresponding fits obtained using a B-spline fitting approach. Refractive indices, extinction coefficients, and absorption coefficients were extracted from the fits and are displayed in Fig. 3(b).
Fig. 1. (a) The schematic device cross-section for an Al0.15In0.85As0.77Sb0.23 PIN, (b) the X-ray diffraction pattern for Al0.15In0.85As0.77Sb0.23/GaSb.
Fig. 2. The photoluminescence versus wavelength for Al0.15In0.85As0.77Sb0.23/GaSb at different temperatures.
Fig. 3. (a) Measured Ψ (left axis) and Δ (right axis) with their corresponding fits (dashed), and (b) the extracted refractive indices, extinction coefficients (left axis), and absorption coefficients (right axis) for Al0.15In0.85As0.77Sb0.23/GaSb.
Following epitaxial characterization, circular mesas were defined using standard lithography techniques and were wet etched using a solution of C6H8O7:H3PO4:H2O2:H2O (10 g:6 mL:3 mL:60 mL). After forming the mesas, Ti/Au (10 nm/100 nm) contacts were deposited via e-beam evaporation on both the p- and n-contact layers.
3. Device characterization
Current-voltage curves were measured with an HP 4145 Semiconductor Parameter Analyzer in a cryogenic chamber cooled with liquid nitrogen. Figure 4(a) shows the dark current and photocurrent obtained using a temperature-stabilized 2-µm fiber-coupled laser diode at 200 K. For the photocurrent curve, ∼27 µW of light was incident on the top of the mesa. The light was focused with a fiber lens to a spot size smaller than the mesa diameter to ensure no sidewall illumination. A characteristic kink in the dark current curve is visible at -4.5 V, indicating the onset of tunneling. The photocurrent is constant out to the intersection of the photocurrent and dark current curves. Figure 4(b) shows the capacitance-voltage curve for a 200-µm-diameter device measured with an HP 3275A LCR meter at room temperature under blackout conditions. At a reverse bias of 2 V, the capacitance has leveled off, indicating the device is fully depleted. Dark current was measured in 20-K increments from 100 K to 300 K and is plotted as dark current density in Fig. 5(a). Figure 5(b) shows the dark current, under a -2 V bias at 300 K, for devices of different mesa sizes plotted against mesa diameter. The quadratic fit indicates the dark current is bulk-limited. The quadratic fit holds up until the onset of band-to-band tunneling at -5 V. Responsivity was measured at 300 K at 1.55 µm and 2 µm with temperature-stabilized fiber-coupled laser diode illumination. A reverse bias of 2 V was used to ensure full depletion of the device. At 1.55 µm and 2 µm the responsivity was 0.49 A/W and 0.53 A/W respectively, corresponding to external quantum efficiencies of 39% and 33% respectively. The theoretical external quantum efficiency, η, can be calculated with η = (1-R)(exp(-αW)), where R is the surface reflection, α is the absorption coefficient, and W is the depletion width (500 nm). Using the refractive index and absorption coefficient in Fig. 3(b) at 1.55 µm and 2 µm the theoretical external quantum efficiencies are 35% and 25% respectively. The difference between the theoretical and measured quantum efficiencies is likely due to uncertainty with fitting the optical constants. We used a B-spline approach for fitting as there is no reference model for Al1-xInxAs1-ySb. We also note that the depletion region is only 500-nm thick and there is no anti-reflection coating. From the optical constants in Fig. 3(b) we can calculate a normal incidence surface reflection at 1.55 µm and 2 µm of ∼31%. Therefore, with a proper 1%-reflectivity anti-reflection coating we can expect quantum efficiencies up to ∼57% and ∼48% at 1.55-µm and 2-µm illumination respectively.
Fig. 4. (a) Current-voltage characteristic at 200 K for a 150-µm-diameter Al0.15In0.85As0.77Sb0.23 PIN with photocurrent, under ∼27 µW of 2-µm illumination, and dark current. (b) A capacitance-voltage curve at 300 K for a 200-µm-diameter device.
Fig. 5. (a) Dark current density versus temperature curves from 100 K to 300 K in 20-K increments, and (b) dark current versus diameter curve with a corresponding quadratic fit at 300 K.
The activation energy was extracted from the measurements in Fig. 5(a) using the relation,
where Idark is the measured dark current, T is the temperature in kelvin, Ea is activation energy, and kB is the Boltzmann constant. Figure 6 shows dark current density plotted against inverse temperature for reverse bias voltages from 1 V to 5 V in 1-V increments for a 150-µm-diameter device. Equation (1) was used to fit curves to the measured values and extract an activation energy for each bias voltage. The extracted activation energies are plotted in the inset. The bandgap for Al0.15In0.85As0.77Sb0.23 is known to be ∼0.4 eV [6]. These activation energies are approximately half the bandgap energy, which indicates that thermal generation dominates. At low temperatures under 4 V and 5 V reverse bias, the activation energy is likely dominated by trap-assisted tunneling current [14].Fig. 6. Dark current density versus inverse temperature for the Al0.15In0.85As0.77Sb0.23 PIN from -1 V to -5 V in 1-V increments with Eq. (1) fits (dashed). The inset shows the extracted activation energy versus bias.
The forward current-voltage characteristic was measured from 200 K to 380 K in 20-K increments using a Keithley 2400 SourceMeter. At each temperature, the measured curve was fit with Eq. (2) to extract the ideality factor. J is the measured forward current density, JS is the reverse saturation current density, q is the elementary charge, V is the applied voltage, n is the ideality factor, kB is the Boltzmann constant, and T is the temperature in kelvin.
Both JS and n were fit to the measured data using Eq. (2). Figure 7(a) shows the measured forward current density for a 200-µm-diameter device with fits shown as dashed lines. There is excellent agreement between the measured data and the fit curves. Figure 7(b) shows the extracted ideality factor plotted against temperature for two different mesa diameters, 150 µm and 200 µm. Two different size devices were measured to assess the role of surface effects. From 200 K to 300 K the ideality factors for the two sizes are the same and decrease with increasing temperature. This inverse trend between ideality factor and temperature has been previously reported in other structures [15,16]. However, above 300 K, the ideality factor increases with temperature and the two mesa diameters diverge. The sharp increase in ideality factor above 300 K likely results from the activation of surface effects, as indicated by the greater increase of the 200-µm-diameter device compared to the 150-µm-diameter device.
Fig. 7. (a) Forward current density versus bias for the Al0.15In0.85As0.77Sb0.23 PIN from 200 K to 380 K in 20-K increments with Eq. (2) fits (dashed), and (b) the extracted ideality factor versus temperature for a 150-µm-diameter device (square) and a 200-µm-diameter device (circle).
Fig. 8. Reverse saturation current density, Js, versus inverse temperature for the Al0.15In0.85As0.77Sb0.23 PIN with the corresponding activation energy fit using (1).
The fit Js is plotted versus inverse temperature in Fig. 8. Using the fit Js the activation energy was extracted using a similar method as used in Fig. 6. Under a similar temperature range as above the activation energy is 0.25 eV, similar to values above. At higher temperatures, above 300 K, the extracted activation energy is 0.48 eV. This activation energy is similar to the bandgap of Al0.15In0.85As0.77Sb0.23 indicating diffusion current dominates.
4. Conclusion
We have reported several material characteristics of Al0.15In0.85As0.77Sb0.23/GaSb. Refractive indices, extinction coefficients and absorption coefficients are presented along with a PL spectrum peaking at 2.94 µm at 300 K. Device parameters for Al0.15In0.85As0.77Sb0.23 PIN photodiodes are also presented. Dark current versus diameter measurements indicate the dark current of the fabricated PIN photodiodes is bulk-limited. External quantum efficiencies of 39% and 33% are achieved at 1.55 µm and 2 µm respectively. Extracted from the reverse dark current density, the activation energy between 180 K and 280 K is ∼0.22 eV, around half the bandgap energy of Al0.15In0.85As0.77Sb0.23, indicating thermal generation is dominant. Ideality factor is plotted versus temperature for two different mesa diameters. Between 200 K and 300 K the ideality factors decrease with increasing temperature. Above 300 K, the sharp increase in ideality factor and divergence between the two mesa diameters is likely caused by the activation of surface effects. These results may be useful in future photodetector designs using Al0.15In0.85As0.77Sb0.23 for operation in mid-IR.
Funding
Defense Advanced Research Projects Agency (W909MY-12-D-0008); Army Research Office (W911NF-17-1-0065).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Razeghi and B.-M. Nguyen, “Advances in mid-infrared detection and imaging: a key issues review,” Rep. Prog. Phys. 77(8), 082401 (2014). [CrossRef]
2. W. Sun, Z. Lu, X. Zheng, J. C. Campbell, S. J. Maddox, H. P. Nair, and S. R. Bank, “High-Gain InAs Avalanche Photodiodes,” IEEE J. Quantum Electron. 49(2), 154–161 (2013). [CrossRef]
3. J. Abautret, J. P. Perez, A. Evirgen, J. Rothman, A. Cordat, and P. Christol, “Characterization of midwave infrared InSb avalanche photodiode,” J. Appl. Phys. 117(24), 244502 (2015). [CrossRef]
4. O. Gravrand, J. Rothman, C. Cervera, N. Baier, C. Lobre, J. P. Zanatta, O. Boulade, V. Moreau, and B. Fieque, “HgCdTe Detectors for Space and Science Imaging: General Issues and Latest Achievements,” J. Electron. Mater. 45(9), 4532–4541 (2016). [CrossRef]
5. A. H. Jones, S. D. March, A. A. Dadey, A. J. Muhowski, S. R. Bank, and J. C. Campbell, “AlInAsSb Separate Absorption, Charge, and Multiplication Avalanche Photodiodes for Mid-Infrared Detection,” IEEE J. Quantum Electron. 108(19), 191108 (2022). [CrossRef]
6. S. J. Maddox, S. D. March, and S. R. Bank, “Broadly Tunable AlInAsSb Digital Alloys Grown on GaSb,” Cryst. Growth Des. 16(7), 3582–3586 (2016). [CrossRef]
7. M. E. Woodson, M. Ren, S. J. Maddox, Y. Chen, S. R. Bank, and J. C. Campbell, “Low-noise AlInAsSb avalanche photodiode,” Appl. Phys. Lett. 108(8), 081102 (2016). [CrossRef]
8. A. H. Jones, S. D. March, S. R. Bank, and J. C. Campbell, “Low-noise high-temperature AlInAsSb/GaSb avalanche photodiodes for 2-µm applications,” Nat. Photonics 14(9), 559–563 (2020). [CrossRef]
9. M. Ren, S. Maddox, Y. Chen, M. Woodson, J. C. Campbell, and S. Bank, “AlInAsSb/GaSb staircase avalanche photodiode,” Appl. Phys. Lett. 108(8), 081101 (2016). [CrossRef]
10. M. Ren, S. J. Maddox, M. E. Woodson, Y. Chen, S. R. Bank, and J. C. Campbell, “Characteristics of AlxIn 1−xAs ySb1− y (x:0.3−0.7) Avalanche Photodiodes,” J. Lightwave Technol. 35(12), 2380–2384 (2017). [CrossRef]
11. D. Chen, R. Wang, J. Andrew McArthur, X. Xue, A. H. Jones, S. R. Bank, and J. C. Campbell, “Demonstration of infrared nBn photodetectors based on the AlInAsSb digital alloy materials system,” Appl. Phys. Lett. 119(3), 031101 (2021). [CrossRef]
12. R. Wang, D. Chen, J. A. McArthur, X. Xue, A. H. Jones, S. R. Bank, and J. C. Campbell, “Al0.3InAsSb/Al0.7InAsSb Digital Alloy nBn Photodetectors,” J. Lightwave Technol. 40(1), 113–120 (2022). [CrossRef]
13. R. Wang, J. A. Mcarthur, S. R. Bank, and J. C. Campbell, “Infrared Al lt;sub gt;0.15 lt;/sub gt;InAsSb digital alloy nBn photodetectors,” J. Lightwave Technol. 1 (2022).
14. J. Huang, J. Huang, C. Zhao, C. Zhao, B. Nie, B. Nie, S. Xie, S. Xie, D. C. M. Kwan, X. Meng, Y. Zhang, Y. Zhang, D. L. Huffaker, W. Ma, W. Ma, and W. Ma, “High-performance mid-wavelength InAs avalanche photodiode using AlAs0.13Sb0.87 as the multiplication layer,” Photonics Res. 8, 755–759 (2020). [CrossRef]
15. J. Verschraegen, M. Burgelman, and J. Penndorf, “Temperature dependence of the diode ideality factor in CuInS2-on-Cu-tape solar cells,” Thin Solid Films 480-481, 307–311 (2005). [CrossRef]
16. M. K. Hudait and S. B. Krupanidhi, “Doping dependence of the barrier height and ideality factor of Au/n-GaAs Schottky diodes at low temperatures,” Phys. B (Amsterdam, Neth.) 307(1-4), 125–137 (2001). [CrossRef]
