Open Access
Add to BibSonomy
Abstract
An avalanche photodetector (APD) based on the AlxIn1-xAsySb1-y digital alloy materials system has recently attracted extensive attention due to its extremely low excess noise. Device defects are a critical factor limiting the performance of APDs. In this work, we use low frequency noise spectroscopy (LFNS) to characterize the property of the defects in AlxIn1-xAsySb1-y APDs grown by molecular beam epitaxy (MBE) using the digital alloy technique. Based on low frequency noise spectroscopy results carried out before and after device oxidation, two surface defects and one bulk defect have been identified, which could provide useful information for the future optimization the material growth and device fabrication processes.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Avalanche photodetectors (APDs) have been widely deployed in applications such as single photon detection, optical communication, infrared imaging, and LIDAR detection [1–6]. The internal gain of APDs can provide a higher signal-to-noise (SNR) ratio than conventional p-i-n photodetectors, which results in enhanced optical sensitivity [1,6]. However, the impact ionization process responsible for the internal gain in APDs is stochastic in nature, leading to an additional source of shot noise called excess noise [7]. As a result, finding a material with low excess noise has been a critical issue in the APD research community. Silicon has demonstrated extremely low excess noise but is limited to applications in the visible region due to its intrinsic bandgap. InGaAs and InAlAs materials grown on InP are commonly used for APDs working in the 1.55µm wavelength band; however, InAlAs demonstrates a relatively larger excess noise than silicon [1]. There have been many attempts to improve the excess noise performance of III-V materials, such as using InAs [8–11], AlAsSb [12–14] or AlGaAsSb [15] as multiplication layer for low noise APDs. HgCdTe APDs also exhibit superior excess noise and is ideally suited for mid-wavelength infrared (MWIR) applications [16,17]. However, HgCdTe APDs need cryogenic cooling, which adds to the overall footprint and complexity of the detector package. And HgCdTe is incompatible with the global initiative to eventually phase out the use of mercury.
Recently, the AlxIn1-xAsySb1-y digital alloy materials system lattice matched to GaSb has demonstrated very low less excess noise. Additionally, by tuning the Al composition, the optical cutoff wavelength can be moved from the near infrared (NIR) to the MWIR [18,19]. Furthermore, AlxIn1-xAsySb1-y APDs have demonstrated very low dark current and high temperature stability, indicating that this digital alloy materials system is a promising candidate for high performance NIR and MIR APDs [18,20–22]. Since the material growth and device fabrication of the AlxIn1-xAsySb1-y material system is still at early development stage, the reported device performance could be further enhanced by identifying and eliminating defects within the AlxIn1-xAsySb1-y layers. In order to push the devices to operate in Geiger Mode for single photon detection, a detailed understanding of the defects in AlxIn1-xAsySb1-y layers is crucial, as these defects in the multiplication layer could play a key role in increasing not only the dark count, but also the after-pulsing probability, both of which are important figures of merit in single photon detection applications.
In this paper, we use low frequency noise spectroscopy (LFNS) to characterize three defect levels within AlxIn1-xAsySb1-y (x = 0.3) APDs which particular concentration is useful for 2um applications [21]. By comparing the noise features and defect properties of the devices before and after oxidation, two of them are identified as surface defects with activation energy of 0.16 and 0.21 eV respectively, indicating that further optimization of process flow can be achieved. The remaining defect with activation energy of 0.11 eV is found within the bulk material. The origin of these defects is discussed based on the LFNS measurements. The information found in this work could be helpful for future optimization of MBE growth and device processing.
2. Device structure and fabrication
An AlxIn1-xAsySb1-y p-i-n structure with x = 0.3 was grown on n-type Te-doped GaSb (001) substrates by solid source molecular beam epitaxy. The AlxIn1-xAsySb1-y layers were grown as a digital alloy of the binary constituents with a period thickness of 3nm and the following shutter sequence: AlSb, AlAs, AlSb, InSb, InAs, Sb. The bandgap of this material is estimated to be 0.58eV [19]. Further details of the AlxIn1-xAsySb1-y digital alloy growth technique are provided elsewhere [19]. The device cross section is shown in Fig. 1 below.
After growth, the epilayer was fabricated into circular mesas using standard photolithography and chemically etched with a HCl:H2O2:H2O solution. Ti/Au ohmic contacts were deposited via electron beam evaporation. An SU-8 2000.5 coating was used for the device passivation.
3. Measurement results
3.1 Dark current performance
The dark current of a set of APDs with different diameters was measured in a probe station at room temperature by a semiconductor device analyzer as shown in Fig. 2. As shown in Fig. 2(a), the dark current varied significantly for the same size of device especially at relative large size (such as 500µm diameter). Variation in the dark current of the devices on a single chip may be due to material nonuniformity. Figure 2(b) replot the best devices picked from each size. These dark current values are close to these of the previous reported AlInAsSb APDs [23]. It is also noted that the dark current of these best-performance device is proportional to device diameter rather than device area, which suggests the dark current is dominated by surface leakage. The responsivity of the APD at -1V is measured with a grating monochromator from 500nm to 1800nm at -1V as shown in Fig. 3 based on a 150µm diameter device. The responsivity at 1550nm is 0.425A/W, corresponding to a thermal noise equivalent power of 6.15 × 10−14W/Hz1/2.
Fig. 2. The dark current versus diameter at -1 V at 300 K of the AlInAsSb (a) all the devices tested (b)devices with best performance at each size
In order to investigate the defect in the AlInAsSb APDs, we picked two devices with diameter of 500µm and 200 µm with relatively high dark current for varied temperature I-V measurement as well as low frequency noise analysis. The dark current for these two devices from 77K to 300K is shown in Fig. 4(a) and Fig. 4(b). The Arrhenius plot of the temperature dependent dark current at -1V is shown in Fig. 5, from which an activation energy (Ea) of 0.14eV was extracted, which is also close to that of previously reported AlInAsSb APD [23].
Fig. 4. I−V curves of a 500-µm device (a) and 200-µm diameter device (b) as a function of temperature
3.2 Low frequency noise spectroscopy analysis
After the dark current measurement, LFNS was used to investigate the defect characteristics of the AlxIn1-xAsySb1-y (x = 0.3) APDs. The current noise spectral density of low frequency noise can be mathematically expressed as the sum of generation-recombination (G-R) noise, 1/f noise, and white noise: [24–26]
where ${A_i}$ is the amplitude of G-R process, ${\tau _{oi}}$ is time constant of each G-R center, f is the frequency, B is the amplitude of the 1/f process, and C is the amplitude of the white noise [25–27]. To better visualize the Lorentzian feature in the measured noise spectrum, Eq. (1) is multiplied by f and plotted against temperature and frequency to suppress the 1/f background noise, which can be expressed as: The noise spectra were measured at different temperatures and bias regions in order to extract activation energy and cross section defect parameters. The activation energy can be determined by investigating how the Lorentzian peaks shift with temperature. More specifically, by fitting the measured noise spectrum in different temperature regions with Eq. (2), the lifetime ${\tau _{oi}}$ of each defect level can be evaluated [28,29]. Subsequently, the activation energy and the capture cross section can be calculated from the slope and the intercept of the Arrhenius plot of ln(${\tau _{oi}}{T^2}$) versus 1000/T.The low frequency noise spectra were measured at temperature ranging from 77 K to 300 K with a measurement step of 5 K or 6 K for 500µm and 200µm diameter devices. Figure 6 and Fig. 7 show the measured noise spectra at different temperature ranges where the G-R process related Lorentzian peak is observable in the devices. The time constant ${\tau _{oi}}$ corresponding to the carrier lifetime at each temperature was extracted from each noise spectrum via Lorentzian fitting. The Arrhenius plot of ln(${\tau _{oi}}{T^2}$) versus 1000/T is shown in Fig. 8. Defects with activation energies of 0.11eV and 0.16eV were found in 500µm diameter device, and defects with activation energies of 0.11eV and 0.21eV were found in the 200-µm diameter device. The defect with an activation energy of 0.11eV can be found in both devices, indicating it may be a bulk defect caused by material growth. The other defects (0.16 and 0.21 eV) could be surface-related defects introduced by fabrication, as the variation of activation energies of different devices is suggestive of a non-uniform process encountered during fabrication.
Fig. 6. The measured noise spectrum of a 500-µm diameter device at different temperatures, shown with Lorentzian fitting.
Fig. 7. The measured noise spectrum of a 200-µm diameter device under different temperatures, shown with Lorentzian fitting.
Fig. 8. Arrhenius plot of the defects in 500-µm diameter (red) and 200-µm diameter (blue) AlxIn1-xAsySb1-y (x = 0.3) APDs.
To further investigate the nature of the defects, we retested the same devices two months after the initial tests. It is believed that some surface oxidation happened within this two months period, as indicated in Fig. 9 by the increase in dark current. After these dark current measurements, the same devices were retested using LFNS. The measured noise spectrum versus temperature and Arrhenius plot of different device is shown in Fig. 10, Fig. 11, and Fig. 12. It is noted that only one defect with an activation energy of 0.16eV in the 500-µm diameter device and only one defect with an activation energy of 0.21eV in 200-µm diameter device were visible in LFNS measurements across multiple bias conditions and temperature ranges. It is believed that the defects associated with oxidation increased surface leakage current enough to overwhelm the G-R noise signature of the bulk defect originally found before oxidation, making the 0.11 eV defect undetectable. All LFNS defect findings are summarized in Table 1.
Fig. 9. Dark current versus temperature at -1 V before and after oxidation for the 500-µm and 200-µm diameter devices.
Fig. 10. The measured noise spectrum of the 500-µm diameter device at different temperatures after surface oxidation, shown with Lorentzian fitting.
Fig. 11. The measured noise spectrum of the 200-µm diameter device at different temperatures after surface oxidation, shown with Lorentzian fitting.
Fig. 12. Arrhenius plot of the defects in 500-µm diameter (red) and 200-µm diameter (blue) AlxIn1-xAsySb1-y (x = 0.3) APDs after surface oxidation.
Table 1. Detailed LFNS defect summary
Finally, we note that the study carried out in this work is based on the typically devices process flow using SU8 passivation, which is intended to to impede device degradation and oxidation. However, based on the measurement results shown above, the SU8 passivation cannot completely prevent this degradation process, suggesting the necessity of developing new passivation methods in the future. Moreover, to better understand the mechanism of device degradation, future works focus exposing unpassivated devices to oxygen flow to induce high level of surface oxidation.
4. Conclusion
In summary, LFNS was used to investigate the defects in the GaSb-based AlxIn1-xAsySb1-y (x = 0.3) digital alloy APDs. Three defects in two devices were found with activation energy of 0.16, 0.21, and 0.11 eV. Two of these defects (0.16 and 0.21 eV) are expected to be surface defects, and the other (0.11 eV) is believed to be a bulk defect. The defect characterization of these devices may help to further optimize and improve the AlxIn1-xAsySb1-y material growth and fabrication processes, leading to higher performance APDs in this materials system suitable for Geiger mode operation.
Funding
National Natural Science Foundation of China (61975121); Shanghai Sailing Program (17YF1429300); ShanghaiTech University startup funding (F-0203-16-002); National Key Research and Development Program of China (2019YFB2203400).
Acknowledgments
The authors would like to thank Yuan Yuan and Joe C. Campbell from the University of Virginia and Ann K. Rockwell, Stephen D. March, and Seth R. Bank from the University of Texas at Austin for their material growth and device process support.
Disclosures
The authors declare no conflicts of interest.
References
1. J. C. Campbell, “Recent Advances in Avalanche Photodiodes,” J. Lightwave Technol. 34(2), 278–285 (2016). [CrossRef]
2. X. Jiang, M. Itzler, K. O’Donnell, M. Entwistle, M. Owens, K. Slomkowski, and S. Rangwala, “InP-based single-photon detectors and geiger-mode APD arrays for quantum communications applications,” IEEE J. Sel. Top. Quantum Electron. 21(3), 5–16 (2015). [CrossRef]
3. P. Vines, K. Kuzmenko, J. Kirdoda, D. C. Dumas, M. M. Mirza, R. W. Millar, D. J. Paul, and G. S. Buller, “High performance planar germanium-on-silicon single-photon avalanche diode detectors,” Nat. Commun. 10(1), 1086 (2019). [CrossRef]
4. X. Meng, S. Xie, X. Zhou, N. Calandri, M. Sanzaro, A. Tosi, C. H. Tan, and J. S. Ng, “InGaAs/InAlAs single photon avalanche diode for 1550 nm photons,” R. Soc. Open Sci. 3(3), 150584 (2016). [CrossRef]
5. G. Buller and A. Wallace, “Ranging and three-dimensional imaging using time-correlated single-photon counting and point-by-point acquisition,” IEEE J. Sel. Top. Quantum Electron. 13(4), 1006–1015 (2007). [CrossRef]
6. B. Chen, Y. Wan, Z. Xie, J. Huang, N. Zhang, C. Shang, J. Norman, Q. Li, Y. Tong, and K. M. Lau, “Low dark current high gain InAs quantum dot avalanche photodetectors monolithically grown on Si,” ACS Photonics 7(2), 528–533 (2020). [CrossRef]
7. R. McIntyre, “Multiplication noise in uniform avalanche diodes,” IEEE Trans. Electron Devices ED-13(1), 164–168 (1966). [CrossRef]
8. A. R. Marshall, P. J. Ker, A. Krysa, J. P. David, and C. H. Tan, “High speed InAs electron avalanche photodiodes overcome the conventional gain-bandwidth product limit,” Opt. Express 19(23), 23341–23349 (2011). [CrossRef]
9. 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]
10. A. R. Marshall, J. P. David, and C. H. Tan, “Impact ionization in InAs electron avalanche photodiodes,” IEEE Trans. Electron Devices 57(10), 2631–2638 (2010). [CrossRef]
11. A. R. Marshall, C. H. Tan, M. J. Steer, and J. P. David, “Extremely low excess noise in InAs electron avalanche photodiodes,” IEEE Photonics Technol. Lett. 21(13), 866–868 (2009). [CrossRef]
12. X. Yi, S. Xie, B. Liang, L. W. Lim, J. S. Cheong, M. C. Debnath, D. L. Huffaker, C. H. Tan, and J. P. David, “Extremely low excess noise and high sensitivity AlAs 0.56 Sb 0.44 avalanche photodiodes,” Nat. Photonics 13(10), 683–686 (2019). [CrossRef]
13. X. Yi, S. Xie, B. Liang, L. W. Lim, X. Zhou, M. C. Debnath, D. L. Huffaker, C. H. Tan, and J. P. David, “Demonstration of large ionization coefficient ratio in AlAs 0.56 Sb 0.44 lattice matched to InP,” Sci. Rep. 8(1), 1–6 (2018). [CrossRef]
14. J. Xie, S. Xie, R. Tozer, and C. H. Tan, “Excess noise characteristics of thin AlAsSb APDs,” IEEE Trans. Electron Devices 59(5), 1475–1479 (2012). [CrossRef]
15. S. Xie, X. Zhou, S. Zhang, D. J. Thomson, X. Chen, G. T. Reed, J. S. Ng, and C. H. Tan, “InGaAs/AlGaAsSb avalanche photodiode with high gain-bandwidth product,” Opt. Express 24(21), 24242–24247 (2016). [CrossRef]
16. M. Kinch, J. Beck, C.-F. Wan, F. Ma, and J. Campbell, “HgCdTe electron avalanche photodiodes,” J. Electron. Mater. 33(6), 630–639 (2004). [CrossRef]
17. J. D. Beck, C.-F. Wan, M. A. Kinch, and J. E. Robinson, “MWIR HgCdTe avalanche photodiodes,” in Materials for Infrared Detectors, (International Society for Optics and Photonics, 2001), 188–197.
18. S. R. Bank, J. C. Campbell, S. J. Maddox, M. Ren, A. K. Rockwell, M. E. Woodson, and S. D. March, “Avalanche photodiodes based on the AlInAsSb materials system,” IEEE J. Sel. Top. Quantum Electron. 24(2), 1–7 (2018). [CrossRef]
19. 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]
20. A. H. Jones, Y. Yuan, M. Ren, S. J. Maddox, S. R. Bank, and J. C. Campbell, “Al x In 1-x As y Sb 1-y photodiodes with low avalanche breakdown temperature dependence,” Opt. Express 25(20), 24340–24345 (2017). [CrossRef]
21. A. H. Jones, S. D. March, S. R. Bank, and J. C. Campbell, “Al x In 1-x As y Sb 1-y Separate Absorption, Charge, and Multiplication Avalanche Photodiodes for 2-µm Detection,” in 2019 IEEE Photonics Conference (IPC), (IEEE), 1–2.
22. A. H. Jones, A.-K. Rockwell, S. D. March, Y. Yuan, S. R. Bank, and J. C. Campbell, “High Gain, Low Dark Current Al0.8In0.2As0.23Sb0.77 Avalanche Photodiodes,” IEEE Photonics Technol. Lett. 31(24), 1948–1951 (2019). [CrossRef]
23. M. Ren, S. J. Maddox, M. E. Woodson, Y. Chen, S. R. Bank, and J. C. Campbell, “Characteristics of Alx In 1-x Asy Sb1-y(x: 0.3−0.7) Avalanche Photodiodes,” J. Lightwave Technol. 35(12), 2380–2384 (2017). [CrossRef]
24. B. K. Jones, “Low-frequency noise spectroscopy,” IEEE Trans. Electron Devices 41(11), 2188–2197 (1994). [CrossRef]
25. J. Huang, B. Chen, Z. Deng, Y. Gu, Y. Ma, J. Zhang, X. Chen, and J. Shao, “Deep levels analysis in wavelength extended InGaAsBi photodetector,” Semicond. Sci. Technol. 34(9), 095018 (2019). [CrossRef]
26. J. Huang, Y. Wan, D. Jung, J. Norman, C. Shang, Q. Li, K. M. Lau, A. C. Gossard, J. E. Bowers, and B. Chen, “Defect Characterization of InAs/InGaAs Quantum Dot p-i-n Photodetector Grown on GaAs-on-V-Grooved-Si Substrate,” ACS Photonics 6(5), 1100–1105 (2019). [CrossRef]
27. L. Zhu, J. Huang, Z. Xie, Z. Deng, L. Chen, C. Lin, and B. Chen, “Low-Frequency Noise Spectroscopy Characterization of HgCdTe Infrared Detectors,” IEEE Trans. Electron Devices 67(2), 547–551 (2020). [CrossRef]
28. W. Chen, B. Chen, J. Yuan, A. Holmes, and P. Fay, “Bulk and interfacial deep levels observed in In0. 53Ga0. 47As/GaAs0. 5Sb0. 5 multiple quantum well photodiode,” Appl. Phys. Lett. 101(5), 052107 (2012). [CrossRef]
29. W. Chen, B. Chen, A. Holmes, and P. Fay, “Investigation of traps in strained-well InGaAs/GaAsSb quantum well photodiodes,” Electron. Lett. 51(18), 1439–1440 (2015). [CrossRef]
