Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Broadband Au/n-GaSb Schottky photodetector array with a spectral range from 300 nm to 1700nm

Open Access Open Access

Abstract

A broadband photodetector is becoming increasingly important as a key element for multicolor imaging. We proposed an Au/n-GaSb Schottky photodetector (PD) array with a wide spectral range from ultraviolet (UV) to short-wavelength infrared (SWIR). The PD was formed by deposition of a 5 nm-thick Au layer on the n-type GaSb substrate and subsequent mesa array formation. The fabricated PD array has shown uniform electrical characteristics and good rectifying behaviors. From the photoresponse measurement, the PD has shown uniformly high external quantum efficiency (EQE) over a spectral range of 300 nm to 1700nm. The value of EQE was 35% at 300 nm and exceeded 50% in the IR region. Furthermore, the PD has shown a rapid rise time of 1.44 µs from the transient photoresponse measurement.

© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction

Photodetectors (PDs) are an essential part of the core future technologies such as the Internet of Things (IoT) or autonomous driving. Recently, broadband imaging beyond the visible spectrum is of great interest. Through imaging in a wide spectral range including ultraviolet (UV) and infrared (IR) light, more advanced image processing is possible in many fields such as night recognition, deep-learning image recognition, and hyperspectral imaging. In general, three different sub-bands: 0.25 µm to 0.4 µm (UV), 0.45 µm to 1.0 µm (visible to near-IR), and 0.9 µm to 1.7 µm (near-IR to short-wavelength IR) are detected by GaN-, Si-, and InGaAs- based PDs, respectively [1]. In order to realize multicolor imaging with a single sensor, a wide spectral detection range is required for PDs. Various structures have been reported for broadband sensing: applying nanowire or 2-D materials [28], applying polymer photodetectors [1,912], or integration of more than two materials with different bandgap wavelengths [1317]. However, most of them have their problems such as slow response time, limited spectral range, fabrication complexity, and high expense, etc. To alleviate these issues, the facilliation of broadband PDs with a simple structure and high performances are highly important.

From this points of view, Gallium antimonide (GaSb) is a versatile III-V semiconductor for high-speed optoelectronic devices. GaSb is one of the suitable materials for broadband PDs, with a bandgap wavelength of 1.7 µm. However, most of the reported GaSb based photodetectors are mid-wavelength infrared (MWIR) or long-wavelength infrared (LWIR) devices with InAs/GaSb type-II superlattice structures [1820] or other epitaxially formed multilayer devices [2124]. On the other hand, to utilize the inherent GaSb property, we conceived the potential use of the GaSb Schottky diode as a broadband photodetector (PD), from our previous study on the investigation of junction characteristics of the Au/n-GaSb Schottky diode [25]. Schottky PDs would be a suitable candidate for broadband photodetectors due to the capability that converts the light absorbed at the surface of the device to the electrical signal with the quite simple fabrication process. Therefore, in this study, we demonstrated a broadband Au/n-GaSb Schottky photodetector array using a very thin 5 nm-thickness of Au layer. Regardless of an extremely thin thickness of the Au layer, electrical properties and the junction quality were maintained compare to the device with a 50 nm-thick Au [25]. The fabricated PD array has shown uniform external quantum efficiency (EQE) over a wide spectral range of 300 nm to 1700nm.

2. Structure and simulation

Au/n-GaSb Schottky PD consists of a simple structure with a thin Au layer deposited on a GaSb substrate to form a uniform Schottky barrier and extract the photocurrent. The photocurrent generation process of the Au/n-GaSb Schottky PD is illustrated in Fig. 1(a) with a band diagram. When the light with a shorter wavelength than the bandgap wavelength of the GaSb is absorbed in the active region of the PD, generated electron-hole pairs are converted tothe current. The thickness of the active region can be defined as the sum of the depletion region thickness (WD) and the diffusion length (lD): the distance that a carrier can move in the semiconductor before recombination.

 figure: Fig. 1.

Fig. 1. (a) Photocurrent generation process of the Au/n-GaSb PD expressed with a band diagram. (b) Simulation result of absorption at the GaSb active region according to the Au layer thickness. (c) Measured and simulated transmittance of 5 nm-thick Au layer. The measured value was divided by the transmittance of a bare glass wafer. (d) AFM images of a bare GaSb wafer and a 5 nm-thick Au deposited GaSb wafer.

Download Full Size | PDF

One possible technological issue would be the optical reflection at the surface. Therefore, the Au thickness should be optimized to guarantee high optical transmittance and low sheet resistance, because the top Au layer functions as an optical window and a top electrode at the same time. First, the rate of the absorption at the active region of an Au/n-GaSb Schottky diode according to the Au thickness was simulated using RSoft simulation tools. When a semiconductor is n-type, the thickness of the depletion can be expressed as [26],

$${W_D} = \sqrt {\frac{{2{\varepsilon _s}}}{{q{N_D}}}\left( {{\psi_{bi}} - V - \frac{{kT}}{q}} \right)}$$
where ɛs is the permittivity of a semiconductor, ND is an electron doping concentration, ψbi is a built-in potential and V is a voltage bias. At zero bias and room temperature (T=293 K), WD is calculated as 105 nm by substitution that ɛs of a GaSb is 15.7 ɛ0 [27], ND and ψbi are $8.12 \times {10^{16}}$ cm−3 and 0.500 V, respectively from section 4.2 for the Au/n-GaSb Schottky photodetector in this study. Diffusion length can be expressed as ${l_D} = \sqrt {{D_h}\tau } $ where Dh is diffusion coefficient of holes and τ is a recombination lifetime. By substitution of the hole diffusion coefficient of the substrate 27.8 cm2s−1 [28] and the Shockley-Read-Hall (SRH) lifetime of this device 2.6 ns (from section 4.2), lD can be calculated as 2.69 µm. Figure 1(b) shows the simulation result of the absorption rate at the GaSb active region of 2.79 µm from the above calculation, according to the Au layer thickness. As the thickness of the Au increases, it can be seen that the absorption decreases especially at the short wavelength. On the other hand, for metal films thinner than 10 nm, the properties of the films are known to be significantly dependent on the deposition method and properties of the substrate. Furthermore, the sheet resistance increases exponentially and the deposited metal film tends to be nonuniform when the thickness becomes thinner than 5 nm [29]. Therefore, we chose an Au thickness of 5 nm in this work. To evaluate the transmittance of the film, we measured the transmittance of a 5 nm-Au film deposited on a glass substrate. Here, the Au film was deposited by electron beam (E-beam) evaporation in high vacuum conditions. Transmittance was measured at the spectral range of 300 nm to 1800nm for normal incidence, and the transmittance of a bare glass substrate was also measured as a reference. In Fig. 1(c), a graph of measured transmittance was plotted and compared with the transmittance simulated from the properties of Au [30]. The Au layer has shown high transparency at the whole spectral range and it showed higher transmittance at shorter wavelengths. In Fig. 1(d), AFM images of a bare GaSb wafer and a 5 nm-thick Au film deposited GaSb wafer were depicted. Measured RMS roughness (Rq) is 0.471 nm for the bare GaSb and 0.550 nm for the Au deposited GaSb. Since the roughness is still low after the Au deposition, it was confirmed that a continuous Au film was formed.

3. Fabrication

A schematic diagram of the Au/n-GaSb Schottky PD unit cell is illustrated in Fig. 2(a). As a substrate, a tellurium doped n-type GaSb wafer supplied from Wafer Technology Ltd. was used. At first, a wafer was cut into about 1 cm×1 cm size chip. A GaSb chip was degreased in acetone and methanol and treated in concentrated HCl solution (37%) to remove native oxide [31]. The chip was rinsed using IPA to prevent oxidation [31], and immediately loaded at the E-beam evaporation chamber. A 5 nm-thick Au Schottky electrode was deposited by an E-beam evaporator on the front side. After the Au deposition, the PD array was patterned by photolithography. Mesa electrodes were formed by etching substrate except for the mesa areas which are defined by photoresist (PR) patterning. The thin Au layer and the GaSb substrate were etched by the inductively-coupled plasma-reactive ion etching (ICP-RIE) process, and the measured etching depth was approximately 450 nm. After the mesa formation, negative photoresist SU-8 (Kayaku Advanced Materials, Inc.) passivation layers were formed at the mesa edges to prevent the formation of leakage path by surface oxidation. Top metal pads were formed by deposition of Ti/Au layer and the area was defined by the lift-off process. Figure 2(b) shows an optical microscope image of the fabricated Au/n-GaSb Schottky PD array. The optical windows were squares with sizes of 700×700, 500×500, 300×300, 100×100, 75×75, 50×50, 30×30, and 15×15 µm2. The fabricated device was attached to a backplate by silver paste.

 figure: Fig. 2.

Fig. 2. (a) A schematic image of the Au/n-GaSb Schottky PD. (b) An optical microscope image of the fabricated PD

Download Full Size | PDF

4. Measurement and analysis

4.1 Dark current measurement

To investigate electrical properties, we measured the dark current of the fabricated PD. Figure 3(a) shows dark currents of three arbitrary pixels with a 300×300 µm2 optical window, and it can be seen that the graphs are almost overlapped. Dark currents of pixels with different mesa sizes were measured and plotted in Fig. 3(b). Only the magnitude of the dark currents varies while the shapes of the graphs are the same according to the mesa sizes. In Fig. 3(c), dark current densities at −0.2 V were plotted as a function of the perimeter to area ratio to determine the contributions of the surface leakage current. The slope of the linear fitted line is $3.50 \times {10^{ - 5}}$ A/cm. When the dark currents were calculated by using this value, it was confirmed that the surface leakage is a major leakage source [32]. In order to reduce surface leakage current, more precise passivation of the GaSb especially the mesa edge [3342] will be needed.

 figure: Fig. 3.

Fig. 3. (a) Measured dark currents of pixels with 300×300 µm2 optical windows. (b) Dark currents measured at the pixels with different mesa sizes. (c) Dark current density measured at −0.2 V as a function of the perimeter to area ratio.

Download Full Size | PDF

4.2 I-V and C-V analysis

To figure out the properties of the Schottky junction and the GaSb substrate, current-voltage (I-V) and capacitance-voltage (C-V) analyses were conducted for a pixel with a 300×300 µm2 optical window. First, the actual doping concentration of the GaSb substrate and the Schottky barrier height was obtained by C-V analysis. The capacitance of the device was measured at a frequency of 500 kHz. For a Schottky diode, the relation between a junction capacitance CD and a voltage V can be expressed as follows [26].

$$\frac{1}{{{C_D}^2}} = \frac{{2[{{\psi_{bi}} - V - ({kT/q} )} ]}}{{q{\varepsilon _s}{N_D}}}$$
1/CD2 was plotted as a function of the applied reverse voltage in Fig. 4(a). Because the values of junction capacitance are more exact when the dark current is small, linear fitting was done by the points close to 0 V (−0.25 V ∼ 0.05 V). The doping concentration of the GaSb substrate of ND=8.12×1016 cm−3 was obtained from a slope of the line and the built-in potential ψbi = 0.500 V was obtained from an x-axis intercept. Schottky barrier height can be calculated from the equation below [26].
$${\phi _{bC}} = {\psi _{bi}} + \frac{{kT}}{q}\ln \left( {\frac{{{N_C}}}{{{N_D}}}} \right)$$
where ϕbC means Schottky barrier height obtained from C-V analysis, and NC is a conduction band state density of a semiconductor. The barrier height is obtained as 0.522 V by substitution of the NC = 2.5×1017 cm−3 for a GaSb at room temperature [27] and the value of ND = 8.12×1016 cm−3 obtained earlier.

 figure: Fig. 4.

Fig. 4. (a) The result of C-V analysis at the frequency of 500 kHz. Linear fitting was conducted by the points at −0.25 V to 0.05 V. (b) Comparison of the measured current with the I-V fitting results.

Download Full Size | PDF

I-V analysis was conducted by fitting the forward bias current from the method of our previous study [25]. From the fitting result of the previous study, the forward bias current of an Au/n-GaSb Schottky diode is almost composed of thermionic emission (TE) current and Shockley-Read-Hall (SRH) recombination current at room temperature. Therefore, forward bias current was fitted to the following equation [25,42,43] through the Matlab fitting tool.

$${J_{tot}} = {J_{TE}} + {J_{SRH}} = {A^\ast }{T^2}\textrm{exp} \left( { - \frac{{q\phi_b^j}}{{kT}}} \right)\left\{ {\textrm{exp} \left[ {\frac{{q({V - I{R_S}} )}}{{nkT}}} \right] - 1} \right\} + {J_{SRH(0)}}\left\{ {\textrm{exp} \left[ {\frac{{q({V - I{R_S}} )}}{{2kT}}} \right] - 1} \right\}$$
where Jtot, JTE, and JSRH means total, thermionic emission, and SRH recombination current density, A* is a Richardson constant, ϕbj means Schottky barrier height obtained from I-V analysis, RS is a series resistance of the diode, n is an ideality factor, and JSRH(0) is saturation current density of SRH recombination current. The temperature was 293 K and A* is 5.16×104 Am−2K−2 [27] for an n-type GaSb. The fitting variables are ϕbj, n, RS, and JSRH(0). The result of the fitting is expressed in Fig. 4(b) and obtained values are ϕbj = 0.567 V, n = 1.05, and JSRH(0) = 4.63 A/m2. The barrier height obtained from the I-V analysis agrees with the barrier height from the C-V analysis. Moreover, these values are consistent with the values when the Au thickness is 50 nm [25]. This result means that the Schottky junction is well-formed regardless of the extremely thin Schottky metal thickness which is much thinner than the electron mean free path in Au (about 38 nm) [44]. And the equation of JSRH(0) can be simplified as [26]
$${J_{SRH(0)}} \approx \frac{{q{W_D}{n_i}}}{{2{\tau _{SRH}}}}$$
where the intrinsic carrier concentration ${n_i}$ of a GaSb is 1.5×1012 cm−3 at room temperature [27] and τSRH is an SRH recombination lifetime. From the substitution of the value of WD = 105 nm obtained in section 2, the value of τSRH = 2.6 ns can be calculated for this device.

4.3 Photoresponse measurement

Photoresponse performance of the device was first evaluated by measuring photocurrent using a solar simulator and monochromator. The photocurrent was measured from a pixel with the largest optical window at the zero bias. The responsivity was obtained by dividing photocurrent by the light power, and the EQE was calculated from an equation below.

$$\eta = \frac{{h\nu }}{q}\frac{{{I_{ph}}}}{{{P_o}}}$$
where ν is a light frequency, Iph is a photocurrent, and Po is a power of the light. The graphs of the EQE and responsivity according to the wavelength are plotted in Fig. 5(a). From the graph, almost uniform EQE was obtained at the wavelength from 300 nm to 1700nm. Such a wide spectral range has not been reported for photodetectors with single p-n or p-i-n structures due to the strong light absorption in the doped region by free-carrier absorption. Next, photocurrent was measured with voltage sweep using laser diodes. Laser diode measurements were conducted at the pixel with a 300×300 µm2 optical window. Measured photocurrents are plotted in Figs. 5(b), (c), (d) for wavelengths of 638 nm and 1550 nm. In Fig. 5(e), measured photocurrents at 0 V for the light power were plotted on a log-log scale. To extract the dependence of the photocurrent on the light intensity, points in Fig. 5(e) were fitted to the power law: Iph = APα where P is light intensity, and A, α are coefficients depend on characteristics of a device. From the result of linear fitting on a log-log scale, α were obtained as 0.988 and 0.996 for 638 nm and 1550 nm, respectively. A value of α close to unity indicates there is almost no contribution of the defects in the diode [8]. By averaging EQEs calculated from the photocurrents at zero bias, 36.3% and 52.2% were obtained for 638 nm and 1550 nm, respectively. These values are consistent with the continuously measured EQE graph in Fig. 5(a). And through a comparison with the simulation result in Fig. 1(a), it can be estimated that more than 75% of photons passed through the Au layer were converted to electrons. Responsivity was calculated from the measured photocurrent and plotted according to the light power for 638 nm and 1550 nm in Fig. 5(f).

 figure: Fig. 5.

Fig. 5. (a) Spectral EQE and responsivity measured by a solar simulator and a monochromator at the pixel with a large optical window. (b) Photoresponse under dark and 638 nm laser diode illumination at the pixel with a 300×300 µm2 optical window. (c) Photoresponse under dark and 1550 nm laser diode illumination. (d) A log-scale graph of photoresponse for 638 nm and 1550 nm. (e) Power-law fitting results for 638 nm and 1550 nm illuminations. (f) Responsivity obtained from the laser diode measurement according to the light power.

Download Full Size | PDF

The response speed is one of the important figure-of-merit of photodetectors in many applications. To investigate the response speed of the fabricated PD, a square wave-formed light signal was illuminated onto a 300×300 µm2 optical window by modulation of a 638 nm laser diode using a function generator. The current signal from the PD was converted to the voltage signal through a current amplifier. Figure 6(a) reveals the photoresponse of pixels with different sizes by a 638 nm laser diode with a frequency of 50 kHz, which was monitored by an oscilloscope. In Fig. 6(b), the rise time and fall time of each pixel were extracted from the graphs in Fig. 6(a). As the pixel size decreased, the fall time hardly changed but the rise time showed a tendency to decrease. The overshoot also tends to decrease with a decrease in the pixel size. It would be due to the high sheet resistance of Au films for photocarrier extraction, indicating that further pixel scaling will significantly improve the transient characteristics of our devices. In Fig. 6(c), normalized photocurrents measured in modulation frequencies in a range of 500 Hz to 500 kHz are depicted. From the graph, 3 dB cutoff-frequency (f3dB) for a 300×300 µm2 window pixel of the fabricated device was obtained as 250 kHz. This speed is remarkably rapid compare to other broadband photodetectors applying nanowires, 2-D materials, or polymers [112]. The main factor of the time delay will be an R-C time constant because the diffusion speed of the generated photocarrier is very fast, as confirmed in section 2. The response speed of the device can be improved by reducing the pixel size or optimization of the doping concentration and the electrode structure.

 figure: Fig. 6.

Fig. 6. (a) Transient photoresponse of pixels with different sizes of optical windows with a modulation frequency of 50 kHz for 638 nm illumination. (b) Rise time and fall time of each pixel. (c) Relative photocurrent in terms of the laser modulation frequency ranging from 500 Hz to 500 kHz.

Download Full Size | PDF

In Table 1, the properties and performance parameters of the device were compared with broadband photodetectors reported in previous works. It can be seen that the Au/n-GaSb Schottky photodetector shows noticeable performance despite the simplest fabrication process.

Tables Icon

Table 1. Comparison with broadband photodetectors in previous papers

5. Conclusion

In this study, we proposed the broadband Au/n-GaSb Schottky photodetector with a wide spectral range from UV to SWIR. The PD was formed simply by deposition of 5 nm-thick Au layer on the GaSb substrate, and the mesa array was formed by photolithography and the subsequent ICP-RIE process. The fabricated PD showed good rectifying behavior and uniform dark current at each pixel. From the C-V analysis conducted with the frequency of 500 kHz, Schottky barrier height of the diode 0.522 V and the actual doping concentration of the GaSb substrate 8.12×1016 cm−3 were obtained. Next, I-V analysis was conducted by the method from the last study, and the fitting results including the junction parameters were consistent with the diode of which the Au thickness is 50 nm. From the spectral photoresponse measurement using a monochromator, uniformly high EQE was obtained over a spectral range of 300 nm to 1700 nm. Over 75% of internal quantum efficiency (IQE) was estimated by comparison with the simulation result. We expect that the device performance can be further improved by applying more precise passivation of the GaSb and the anti-reflection coating, etc. And the device has shown a rapid rise time of 1.44 µs compare to other broadband PDs from the transient photoresponse measurement. Because the Au/n-GaSb Schottky diode has advantages of both simple fabrication process and a wide spectral range, it has enough value as a candidate of the next-generation broadband photodetector. Furthermore, the spectral range of the GaSb-based MWIR or LWIR devices can be easily extended to the shorter wavelength range by deposition of a thin Au layer on the top GaSb surface.

Funding

Ministry of Trade, Industry and Energy (20006476); Brain Korea 21 (BK21) (FOUR); National NanoFab Center (OI project); G-CORE (N11210010); National Research Foundation of Korea (2019M1A2A2067928, 2020M3H4A3081735).

Disclosures

The authors declare no conflicts of interest

Data availability

Data underlying the results presented in this paper not publicly available at this time but may be obtained from the authors upon reasonable request.

References

1. X. Gong, M. Tong, Y. Xia, W. Chi, J. S. Moon, Y. Cao, G. Yu, C. Shieh, B. Nilsson, and A. J. Heeger, “High-detectivity polymer photodetectors with spectral response from 300 nm to 1450 nm,” Science 325(5948), 1665–1667 (2009). [CrossRef]  

2. Z. Liu, T. Lip, B. Liang, G. Chen, G. Yu, X. Xie, D. Chen, and G. Shen, “High-detectivity InAs nanowire photodetectors with spectral response from ultraviolet to near-infrared,” Nano Res. 6(11), 775–783 (2013). [CrossRef]  

3. A. Sharma, B. Bhattacharyya, A. K. Srivastava, T. D. Senguttuvan, and S. Husale, “High performance broadband photodetector using fabricated nanowires of bismuth selenide,” Sci. Rep. 6(1), 1–8 (2016). [CrossRef]  

4. T. Zhang, Z. Li, J. Wang, W. Kong, G. Wu, Y. Zheng, Y. Zhao, E. Yao, N. Zhuang, and L. Luo, “Broadband photodetector based on carbon nanotube thin film/single layer graphene Schottky junction,” Sci. Rep. 6(1), 1–8 (2016). [CrossRef]  

5. S. Du, W. Lu, A. Ali, P. Zhao, K. Shehzad, H. Guo, L. Ma, X. Liu, X. Pi, P. Wang, H. Fang, Z. Xu, C. Gao, Y. Dan, P. Tan, H. Wang, C. Lin, J. Yang, S. Dong, Z. Cheng, E. Li, W. Yin, J. Luo, B. Yu, T. Hasan, Y. Xu, W. Hu, and X. Duan, “A broadband Fluorographene photodetector,” Adv. Mater. 29(22), 1700463 (2017). [CrossRef]  

6. W. Gao, Z. Zheng, L. Huang, J. Yao, Y. Zhao, Y. Xiao, and J. Li, “Self-powered SnS1-xSex alloy/silicon heterojunction photodetectors with high sensitivity in a wide spectral range,” ACS Appl. Mater. Interfaces 11(43), 40222–40231 (2019). [CrossRef]  

7. J. W. John, V. Dhyani, S. Malty, S. Mukherjee, S. K. Ray, V. Kumar, and S. Das, “Broadband infrared photodetector based on nanostructured MoSe2-Si heterojunction extended up to 2.5 µm spectral range,” Nanotechnology 31(45), 455208 (2020). [CrossRef]  

8. D.-M. Geum, S. Kim, J. Kim, J. Kim, S. K. Kim, S.-Y. Ahn, T. Kim, K. Kang, and S.-H. Kim, “Arrayed MoS2-In0.53Ga0.47As van der Waals Heterostructure for high-speed and broadband detection from visible to shortwave-infrared light,” Small 17(17), 2007357 (2021). [CrossRef]  

9. X. Hu, X. Zhang, L. Liang, J. Bao, S. Li, W. Yang, and Y. Xie, “High-performance flexible broadband photodetector based on organolead halide perovskite,” Adv. Funct. Mater. 24(46), 7373–7380 (2014). [CrossRef]  

10. M. Gao, W. Wang, L. Li, J. Miao, and F. Zhang, “Highly sensitive polymer photodetectors with a wide spectral response range,” Chin. Phys. B 26(1), 018201 (2017). [CrossRef]  

11. H. Shin, J. Kim, and C. Lee, “Ternary bulk heterojunction for wide spectral range organic photodetectors,” J. Korean Phys. Soc. 71(4), 196–202 (2017). [CrossRef]  

12. Z. Zhao, C. Xu, L. Niu, X. Zhang, and F. Zhang, “Recent progress on broadband organic photodetectors and their applications,” Laser & Photonics Rev. 14(11), 2000262 (2020). [CrossRef]  

13. S. W. Seo, D. L. Geddis, and N. M. Jokerst, “3-D stacked thin-film photodetectors for multispectral detection applications,” IEEE Photonics Technol. Lett. 15(4), 578–580 (2003). [CrossRef]  

14. L. Menon, H. Yang, S. Cho, S. Mikael, Z. Ma, C. Reuterskiöld-Hedlund, M. Hammar, and W. Zhou, “Heterogeneously integrated InGaAs and Si membrane four-color photodetector arrays,” IEEE Photonics J. 8(2), 1–7 (2016). [CrossRef]  

15. C. Xie, V. Pusino, A. Khalid, A. P. Craig, A. Marshall, and D. Cumming, “Monolithically integrated InAsSb-based nBnBn heterostructure on GaAs for infrared detection,” IEEE J. Sel. Top. Quant. 24(6), 1–6 (2018). [CrossRef]  

16. E. T. Simola, A. DeLacovo, J. Frigerio, A. Ballabio, A. Fabbri, G. Isella, and L. Colace, “Voltage-tunable dual-band Ge/Si photodetector operating in VIS and NIR spectral range,” Opt. Express 27(6), 8529–8539 (2019). [CrossRef]  

17. D.-M. Geum, S. H. Kim, S. K. Kim, S. S. Kang, J. Kyhm, J. Song, W. J. Choi, and E. Yoon, “Monolithic integration of visible GaAs and near-infrared InGaAs for multicolor photodetectors by using high-throughput epitaxial lift-off toward high resolution imaging systems,” Sci. Rep. 9(1), 18661 (2019). [CrossRef]  

18. N. Gautam, H. S. Kim, M. N. Kutty, E. Plis, L. R. Dawson, and S. Krishna, “Performance improvement of longwave infrared photodetector based on type-II InAs/GaSb superlattices using unipolar current blocking layers,” Appl. Phys. Lett. 96(23), 231107 (2010). [CrossRef]  

19. A. Haddadi, R. Chevallier, A. Dehzangi, and M. Razeghi, “Extended short-wavelength infrared nBn photodetectors based on type-II InAs/AlSb/GaSb superlattices with an AlAsSb/GaSb superlattice barrier,” Appl. Phys. Lett. 110(10), 101104 (2017). [CrossRef]  

20. Z. Deng, D. Guo, C. G. Burguete, Z. Xie, J. Huang, H. Liu, J. Wu, and B. Chen, “Demonstration of Si based InAs/GaSb type-II superlattice p-i-n photodetector,” Infr. Phys. Technol. 101, 113–137 (2019). [CrossRef]  

21. J. Bowers, A. Srivastava, C. Burrus, M. DeWinter, M. Pollak, and J. Zyskind, “High-speed GaInAsSb/GaSb PIN photodetectors for wavelength to 2.3 µm,” Electron. Lett. 22(3), 137–138 (1986). [CrossRef]  

22. L. Jun, S. Hang, Y. Jin, J. Hong, G. Miao, and H. Zhao, “Design of a resonant-cavity-enhanced GaInAsSb/GaSb photodetector,” Semicond. Sci. Technol. 19(6), 690–694 (2004). [CrossRef]  

23. D. Lackner, O. J. Pitts, S. Najmi, P. Sandhu, K. L. Kavanagh, A. Yang, M. Steger, M. L. W. Thewalt, Y. Wang, D. W. McComb, C. R. Bolognesi, and S. P. Watkins, “Growth of InAsSb/InAs MQW on GaSb for mid-IR photodetector applications,” J. Crystal Growth 311(14), 3563–3567 (2009). [CrossRef]  

24. J. Li, A. Dehzangi, G. Brown, and M. Razeghi, “Mid-wavelength infrared avalanche photodetector with AlAsSb/GaSb superlattice,” Sci. Rep. 11(1), 1–7 (2021). [CrossRef]  

25. J. Jang, J. Song, S. S. Lee, S. Jeong, B. J. Lee, and S. Kim, “Analysis of temperature-dependent I-V characteristics of the Au/n-GaSb Schottky diode,” Mater. Sci. Semicond. Process. 131, 105882 (2021). [CrossRef]  

26. S. M. Sze and K. K. Ng, Physics of Semiconductor Devices (John Willey & Sons, 2006), Chap. 1-2

27. M. Levinshtein, S. Rumyantesv, and M. Shur, Handbook Series on Semiconductor Parameters (World Scientific, 1996), Vol. 1

28. O. Madelug, U. Rössler, and M. Schulz, “Gallium antimonide (GaSb), hole mobility,” in Group IV Elements, IV-IV and III-V Compounds. Part b – Electronic, Transport, Optical and Other Properties (Springer, 2002).

29. J. Yun, “Ultrathin meatl films for transparent electrodes of flexible optoelectronic devices,” Adv. Funct. Mater. 27(18), 1606641 (2017). [CrossRef]  

30. P. B. Johnson and R. W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]  

31. Z. Y. Liu, D. A. Saulys, and T. F. Keuch, “Improved characteristics for Au/n-GaSb Schottky contacts through the use of a nonaqueous sulfide-based passivation,” Appl. Phys. Lett. 85(19), 4391–4393 (2004). [CrossRef]  

32. A. Osinsky, S. Gangopadhyay, B. W. Lim, M. Z. Anwar, and M. A. Khan, “Schottky barrier photodetectors based on AlGaN,” Appl. Phys. Lett. 72(6), 742–744 (1998). [CrossRef]  

33. P. Barman, U. N. Roy, and S. Basu, “Improved junction properties of Au-n-GaSb Schottky diodes after chemical modification of GaSb surfaces,” Mater. Lett. 10(4-5), 203–206 (1990). [CrossRef]  

34. P. S. Dutta, K. S. Sangunni, H. L. Bhat, and V. Kumar, “Sulphur passivation of gallium antimonide surfaces,” Appl. Phys. Lett. 65(13), 1695–1697 (1994). [CrossRef]  

35. M. Pérotin, P. Coudray, L. Gouskov, H. Luquet, C. Llinarès, J. J. Bonnet, L. Soonckindt, and B. Lambert, “Passivation of GaSb by sulfur treatment,” J. Electron. Mater. 23(1), 7–12 (1994). [CrossRef]  

36. P. S. Dutta, K. S. R. Koteswara Rao, H. L. Bhat, and V. Kumar, “Effect of ruthenium passivation on the optical and electrical properties of gallium antimonide,” J. Appl. Phys. 77(9), 4825–4827 (1995). [CrossRef]  

37. G. Eftekhari, “Effects of sulfur passivation of GaSb on the thermal stability of Al/n-GaSb contacts,” Jpn. J. Appl. Phys. 35, 564–567 (1996). [CrossRef]  

38. D. M. Murape, N. Eassa, J. H. Neethling, R. Betz, E. Coetsee, H. C. Swart, J. R. Botha, and A. Venter, “Treatment for GaSb surfaces using a Sulphur blended (NH4)2S/(NH4)2SO4 solution,” Appl. Surf. Sci. 258(18), 6753–6758 (2012). [CrossRef]  

39. M. V. Lebedev, E. V. Kunitsyna, W. Calvet, T. Mayer, and W. Jaegermann, “Sulfur passivation of GaSb(100) surfaces: comparison of aqueous and alcoholic sulfide solutions using synchrotron radiation photoemission spectroscopy,” J. Phys. Chem. C 117(31), 15996–16004 (2013). [CrossRef]  

40. L Zhao, Z. Tan, R. Bai, N. Cui, J. Wang, and J. Xu, “Effect of sulfur passivation on GaSb metal-oxide-semiconductor capacitors with neutralized and unneutralized (NH4)2S solutions of varied concentrations,” Appl. Phys. Express 6(5), 056502 (2013). [CrossRef]  

41. D. Tao, Y. Cheng, J. Liu, J. Su, T. Liu, F. Yang, F. Wang, K. Cao, Z. Dong, and Y. Zhao, “Improved surface and electrical properties of passivated GaSb with less alkaline sulfide solution,” Mater. Sci. Semicond. Process. 40, 685–689 (2015). [CrossRef]  

42. D. Donoval, M. Barus, and M. Zdimal, “Analysis of I-V measurements on PtSi-Si Schottky structures in a wide temperature range,” Solid St. Electron. 34(12), 1365–1373 (1991). [CrossRef]  

43. D. Donoval, A. Chvála, R. Šramatý, J. Kováč, E. Morvan, Ch. Dua, M. A. DiForte-Poisson, and P. Kordoš, “Transport properties and barrier height evaluation in Ni/InAlN/GaN Schottky diodes,” J. Appl. Phys. 109(6), 063711 (2011). [CrossRef]  

44. D. Gall, “Electron mean free path in elemental metals,” J. Appl. Phys. 119(8), 085101 (2016). [CrossRef]  

Data availability

Data underlying the results presented in this paper not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1.
Fig. 1. (a) Photocurrent generation process of the Au/n-GaSb PD expressed with a band diagram. (b) Simulation result of absorption at the GaSb active region according to the Au layer thickness. (c) Measured and simulated transmittance of 5 nm-thick Au layer. The measured value was divided by the transmittance of a bare glass wafer. (d) AFM images of a bare GaSb wafer and a 5 nm-thick Au deposited GaSb wafer.
Fig. 2.
Fig. 2. (a) A schematic image of the Au/n-GaSb Schottky PD. (b) An optical microscope image of the fabricated PD
Fig. 3.
Fig. 3. (a) Measured dark currents of pixels with 300×300 µm2 optical windows. (b) Dark currents measured at the pixels with different mesa sizes. (c) Dark current density measured at −0.2 V as a function of the perimeter to area ratio.
Fig. 4.
Fig. 4. (a) The result of C-V analysis at the frequency of 500 kHz. Linear fitting was conducted by the points at −0.25 V to 0.05 V. (b) Comparison of the measured current with the I-V fitting results.
Fig. 5.
Fig. 5. (a) Spectral EQE and responsivity measured by a solar simulator and a monochromator at the pixel with a large optical window. (b) Photoresponse under dark and 638 nm laser diode illumination at the pixel with a 300×300 µm2 optical window. (c) Photoresponse under dark and 1550 nm laser diode illumination. (d) A log-scale graph of photoresponse for 638 nm and 1550 nm. (e) Power-law fitting results for 638 nm and 1550 nm illuminations. (f) Responsivity obtained from the laser diode measurement according to the light power.
Fig. 6.
Fig. 6. (a) Transient photoresponse of pixels with different sizes of optical windows with a modulation frequency of 50 kHz for 638 nm illumination. (b) Rise time and fall time of each pixel. (c) Relative photocurrent in terms of the laser modulation frequency ranging from 500 Hz to 500 kHz.

Tables (1)

Tables Icon

Table 1. Comparison with broadband photodetectors in previous papers

Equations (6)

Equations on this page are rendered with MathJax. Learn more.

W D = 2 ε s q N D ( ψ b i V k T q )
1 C D 2 = 2 [ ψ b i V ( k T / q ) ] q ε s N D
ϕ b C = ψ b i + k T q ln ( N C N D )
J t o t = J T E + J S R H = A T 2 exp ( q ϕ b j k T ) { exp [ q ( V I R S ) n k T ] 1 } + J S R H ( 0 ) { exp [ q ( V I R S ) 2 k T ] 1 }
J S R H ( 0 ) q W D n i 2 τ S R H
η = h ν q I p h P o
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.