Boosted UV Photodetection Performance in Chemically Etched ...oxidesemi.iphy.ac.cn/result/paper/2020/2020-zyhan-AOM.pdf · 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim in – 3

1901833 (1 of 8) © 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advopticalmat.de

Full PaPer

Boosted UV Photodetection Performance in Chemically Etched Amorphous Ga2O3 Thin-Film Transistors

Zuyin Han, Huili Liang,* Wenxing Huo, Xiaoshan Zhu, Xiaolong Du, and Zengxia Mei*

DOI: 10.1002/adom.201901833

the further promotion of devices’ perfor-mance.[7] As is well known to all, PPC phenomenon occurs in most oxide semi-conductor materials owing to the large quantity of oxygen vacancy (Vo) defects and high density of trap states.[8,9]

Phototransistors, one kind of three-ter-minal PD with one more terminal-gate to flexibly control the channel carriers’ trans-portation behavior, have been regarded as an alternative solution to improve the PD performance.[10,11] Phototransistor pos-sesses the intrinsic gain of transistors and regular photoconductors, which makes it possible to achieve both high light-to-dark current ratio and responsivity.[11,12] Last but not the least, the PPC phenom-enon is possibly eliminated by exerting a gate pulse, as Jeon et al. demonstrated in the three-terminal photosensor array with GIZO/IZO/GIZO channel.[5] For a-Ga2O3 UV PD or imaging applications, a research on fabrication and utiliza-tion of phototransistor architecture is

urgently needed to well suppress PPC and raise the response speed while retaining a high photoresponsivity as well. Mean-while, the controllable and selective etching of a-Ga2O3 channel to metals and other oxides is critical to the achievement of a low gate leakage current and good transfer characteristics.[13,14] Wet chemical etching of β-Ga2O3 has been demonstrated by using H3PO4 and H2SO4, respectively.[15,16] However, these strong acids will corrode metals and oxides readily, which bring a big trouble to device fabrication.

In this paper, we present bottom-gate a-Ga2O3 thin film tran-sistors (TFTs) and phototransistors where the a-Ga2O3 channels are selectively etched using tetramethyl ammonium hydroxide (TMAH) aqueous solution. Note that this new etching method owns the advantages of low cost, simple operation, good safety, and desirable compatibility with lithography. For the common bottom-gate Ga2O3 TFT on Si, the device with patterned channel exhibits superior transistor characteristics to the unpatterned one. A bottom-gate a-Ga2O3 phototransistor with interdigital finger-shaped source/drain (S/D) electrodes is prepared on quartz and applied to detect deep UV rays. It demonstrates typical transistor output and transfer characteristics with a high on/off ratio of ≈107. Meanwhile, an excellent photodetector performance appears under a 254 nm UV illumination, including a high light-to-dark ratio of 5 × 107 and responsivity of 5.67 × 103 A W−1. By applying a positive gate pulse, PPC in the a-Ga2O3 photo-transistors is effectively eliminated with a fast decay in 5 ms.

A three-terminal thin-film transistor (TFT) architecture is essential for photodetectors to reach a good balance between high responsivity and fast response speed. Bottom-gate amorphous Ga2O3 (a-Ga2O3) TFTs are fabricated to boost their UV photodetection properties. During the device fabrication process, a simple chemical-etching solution with the advantages of easy operation, low cost, and compatibility with traditional lithography process, is developed to selectively etch a-Ga2O3 films. The a-Ga2O3 channel etched device on Si manifests an effective suppression of the commonly observed gate leakage current. Meanwhile, a patterned a-Ga2O3 TFT on quartz shows an excellent n-type TFT performance with an on/off ratio as high as ≈107. It is further applied as a phototransistor, to diminish the persistent photoconductivity (PPC) effect while keeping a high responsivity (R) as well. Under the 254 nm UV illumination, the a-Ga2O3 phototransistor demonstrates a high light-to-dark ratio of 5 × 107, a high responsivity of 5.67 × 103 A W−1, and a high detectivity of 1.87 × 1015 Jones. Remarkably, the PPC phenomenon in a-Ga2O3 UV phototransistors is effectively suppressed by applying a positive gate pulse, which greatly shortens the decay time to 5 ms and offers a-Ga2O3 possible inroads into imaging applications.

Z. Y. Han, Dr. H. L. Liang, Dr. W. X. Huo, X. S. Zhu, Prof. X. L. Du, Prof. Z. X. MeiBeijing National Laboratory for Condensed Matter PhysicsInstitute of PhysicsChinese Academy of SciencesBeijing 100190, P. R. ChinaE-mail: [email protected]; [email protected]. Y. Han, Dr. W. X. Huo, Prof. X. L. DuSchool of Physical SciencesUniversity of Chinese Academy of SciencesBeijing 100049, ChinaDr. H. L. Liang, Prof. X. L. Du, Prof. Z. X. MeiSongshan Lake Materials LaboratoryDongguan, Guangdong 523808, China

The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/adom.201901833.

1. Introduction

Recently, amorphous Ga2O3 (a-Ga2O3) has attracted increasing attention for its wide range of applications in deep ultra-violet (UV) photodetection, such as confidential space com-munication, imaging, flame detection, and missile warning systems.[1–4] So far two-terminal Ga2O3 photodetectors (PDs) are the most commonly investigated, but a slow response speed caused by persistent photoconductivity (PPC) effect[5,6] impedes

Adv. Optical Mater. 2020, 1901833

http://crossmark.crossref.org/dialog/?doi=10.1002%2Fadom.201901833&domain=pdf&date_stamp=2020-02-24

www.advancedsciencenews.com

© 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1901833 (2 of 8)


2. Results and Discussion

Herein, the etching behaviors of two a-Ga2O3 samples were investigated first in TMAH solutions with different TMAH con-centrations and temperatures. The a-Ga2O3 films are labeled as S1 and S2, which are sputtered under the condition with O2 and without O2 flux, respectively. Both of the samples are amor-phous, and the preparation details can be found elsewhere[1,2,3] and in the Supporting Information. S1 and S2 show different densities, solidly evidenced by the extracted values from X-ray reflection spectra (XRR) in Figure S1 of the Supporting Infor-mation, which are 5.32 and 4.84 g cm−3, respectively. That will definitely influence the etching rate, considering the fact of their different Vo defect densities.[8] The etching rates of S1 and S2 are plotted as a function of etching temperature in different TMAH concentrations in Figure S2 of the Supporting Informa-tion. The values are listed in Table S1 of the Supporting Infor-mation in detail. From these results, it can be found that the etching rates of S1 are slower than S2, which are attributed to its higher density and hence denser structure determined by XRR technique.

The selective etching ability of TMAH solution is very crit-ical for the fabrication of TFT devices, which is explored in the following experiments. The amorphous Al2O3 insulating layer was adopted and prepared by atomic layer deposition (ALD) at 200 °C with a thickness of 110 nm. After immersed in a 0.048% TMAH solution at 27 °C for 140 s, the depth profile of Al2O3 is almost unchanged, suggesting that it is hardly etched by TMAH in this low-level concentration. Similar results are observed in a moderate 0.12% concentration at 40 °C for 54 s (not shown here). On the other hand, a slight etching of Al2O3

occurs with a rate of 0.72 nm s−1 in a 0.24% TMAH solution at 60 °C for 15 s (Figure S3, Supporting Information). It illustrates that a selective etching rate ratio of a-Ga2O3 (S2) to Al2O3 is 17:1 in the 0.24% TMAH solution at 60 °C, which guarantees an adequately wide and easily controllable operation region during the fabrication of a-Ga2O3 TFTs. The variation of a-Ga2O3 and Al2O3 surface morphology after etched in a 0.24% TMAH solu-tion at 60 °C for 15 s is revealed by atomic force microscopy (AFM) images in Figure 1. The root-mean-square (RMS) sur-face roughness of the as-deposited a-Ga2O3 (Figure 1a) film is 0.55 nm in a 10 × 10 µm2 scanning area, slightly smaller than the 1.24 nm after etching (Figure 1b). In Figure 1c,d, the RMS values are 0.34 and 0.38 nm for unetched and etched Al2O3 films, respectively, indicating that Al2O3 surface is uniformly etched by TMAH solution under this condition. Together with the 17:1 etching rate ratio of a-Ga2O3 (S2) to Al2O3, the chem-ical etching with TMAH solution provides a strong guarantee for fabrication of three-terminal multilayered a-Ga2O3 devices.

To further corroborate the effect of TMAH solution on making device-quality a-Ga2O3 patterns, 400 nm thick Ga2O3 films were patterned and etched into interdigital fingers by UV-lithography and wet chemical etching using TMAH and H3PO4 solutions, respectively. Figure 2a,b presents their corresponding 3D laser scanning confocal microscope images, where well-defined patterns can be clearly recognized in both cases. The cross-sectional views of the two instances exhibit the same orthotrapezoid structure, which is favorable for metal deposi-tion on it. Besides, the ratios of the top to the bottom of the trapezoid are very similar as shown by the cross-sectional views in Figure 2a,b, demonstrating that the etching effect of TMAH solution is comparable with H3PO4. The conclusion is also well


Figure 1. AFM images of a) unetched a-Ga2O3 film (S2), b) etched a-Ga2O3 film (S2), c) unetched a-Al2O3 film, and d) etched a-Al2O3 film.




supported by the scanning electron microscope (SEM) images shown in Figure S4 of the Supporting Information. Thus, con-sidering the merit that TMAH barely etches metal and Al2O3, it is more preferred than H3PO4 in our work.

Based on the above-mentioned wet chemical etching tech-nique using TMAH solutions, the common bottom-gate TFTs with patterned and unpatterned a-Ga2O3 channels have been fabricated on commercial SiO2/Si substrates. Schematic struc-tures of the devices are exhibited in Figure 3a. The electrical characteristics are measured at a source-drain voltage (VDS) of 10 V with the source–gate voltage (VGS) sweeping from −100 to 200 V (Figure 3b). The TFT with patterned a-Ga2O3 channel exhibits a typical n-type transfer curve.[17,18] The gate current (IGS), generally defined as a leakage current, is in a very low level ≈10−10 A at the entire VGS sweeping range. However, the TFT with unpatterned a-Ga2O3 channel shows different trans-port behavior. An abrupt drop of the drain current (IDS) occurs in the transfer curve at a specific VGS, where the IGS begins to increase dramatically, indicating that the channel electrons start to transport vertically through SiO2 insulating layer and reach the gate electrode at a positive VGS. The abnormal curves imply that it is beneficial to use patterned a-Ga2O3 channel layer, which can be obtained by a combined processing with UV lithography and TMAH etching. As is reported previously, defect states including Vo and dangling bonds usually exist in SiO2 layer and at SiO2/Si interface.[13,19] These defects will play a role as path way for the electrons injected from the oxide

active layer.[13] At a positive VGS, the amount of accumulated electrons at the interface of a-Ga2O3/SiO2 will increase in the unpatterned a-Ga2O3 TFTs, as shown in the inset of Figure 3c. This will definitely enhance the probability of the electrons to inject into the SiO2 layer and transport through the inherent trap sites inside the dielectric layer, contributing to the uncon-ventional gate leakage current. From the equivalent circuit of unpatterned a-Ga2O3 TFT in Figure 3c, electrons will flow not only between the drain and source electrodes (IDS), but also through the vertical path between the S/D and gate electrodes (IGS or IGD), making SiO2 insulator not only a gate dielectric in TFT, but also an electron transport layer like a diode.[13] The IDS will decrease in the region where an adequate gate leakage current begins to inject into the drain electrode, resulting in the poor TFT transfer characteristics as depicted in Figure 3b. A suitable etching method to pattern a-Ga2O3 channel is conse-quently indispensable to avoid the appearance of large IGS.

To further improve the device performance, Al2O3 dielectric layer was used to fabricate a-Ga2O3 TFT with bottom-gate stag-gered structure on quartz substrate. The interdigital finger configuration, which is frequently adopted to effectively sepa-rate the photogenerated carriers in PDs,[1,20] is chosen as S/D shapes. The S/D electrodes have 15 pair fingers with 10 µm in width, 10 µm in spacing gap, and 145 µm in length. To match the interdigital S/D structures, a winding gate electrode is designed with a 1.5 µm overlap region at the edge of the S/D electrodes. A schematic structure of the device is exhibited


Figure 2. Laser scanning confocal microscope images of a-Ga2O3 film etched with a) TMAH solution and b) H3PO4 solution.




in Figure 4a. Figure 4b shows the IDS–VGS transfer curves of the a-Ga2O3 TFT at different VDS in dark and under 254 nm light illumination, respectively. From the curves in dark, it can be seen that the on-state IDS increases proportionally as VDS increases from 0.1 to 10 V, and the off-state IDS remains at very low levels around 10−12 A. The extracted field-effect mobility (µFE) is only 0.04 cm2 V−1 s−1, which comes from the amorphous nature of Ga2O3 material and could be improved via doping etc. in the future. In spite of this, the device still exhibits decent electrical performance, such as a high on/off ratio of ≈107, a low subthreshold swing of 0.65 mV dec−1 and a moderate posi-tive threshold voltage (Vth) of 5 V. As to the transfer curves under UV 254 nm light illumination, IDS increases distinctly in the depletion region compared with the values in dark. Besides, the on-state photocurrent also increases proportionally as VDS increases from 0.1 to 10 V. At VDS = 10 V and VGS = 4 V, a high photocurrent of ≈10−4 A is obtained, leading to a large light-to-dark current ratio as high as 5 × 107. The ultrahigh rejection ratio is reasonable, because the depletion “off” state in dark is reversed into an “on” state due to the presence of numerous photogenerated electrons.[21] Figure 4c demonstrates the output curves of the same device with an excellent linear and satura-tion performance, indicating a good ohmic contact between the Ga2O3 channel and Ti/Au electrodes and no current crowding phenomenon. Time-dependent photoresponse behavior was evaluated with VGS set at 10 V and shown in Figure 4d. It can be obviously seen that IDS increases as VDS varies from 1, 10, to 20 V, which is consistent with our previous observation. Fur-thermore, the device demonstrates a relatively slow response to

the periodic UV illumination in all cases. Figure 4e presents the responsivity (R) as a function of VG at VDS = 10 V. R is cal-culated according to the following equation

photo dark=−

RI I

PS (1)

where Iphoto is the photocurrent, Idark is the dark current, P is the light intensity, and S is the effective illumination area of the channel. As VGS increases from −15 to 10 V, R increases to an ultrahigh value of 5.67 × 103 A W−1, illustrating an excellent ability of the transistor architecture on boosting the photore-sponsivity of a-Ga2O3 PDs. Based on this responsivity value, the

detectivity (D∗) expressed as /(2 )*12

dark1/2=D RS qI is calculated

to be 1.87 × 1015 Jones. A normalized photoresponse spectrum has been measured at VDS of 20 V and VGS of 10 V, as shown in Figure 4f. Note that the photoresponse peak is at 285 nm and shows a long tail till ≈360 nm. Considering the need for a higher carrier concentration in the channel, the a-Ga2O3 films are sputtered with no addition of O2 flux. In this case, a large quantity of VO defects are suspected existing in the channel layer, which definitely has an influence on the photoresponse performance.

After realizing a high responsivity, the three-terminal photo-transistor is further investigated regarding its effect on sup-pressing the PPC phenomenon with the assistance of the pulse gate voltage. As indicated by the arrows in Figure 5a, a strong PPC remains for dozens or even hundreds of seconds


Figure 3. a) Schematic structures of the TFTs with and without Ga2O3 patterned. b) IDS–VGS curves and IGS–VGS curves of the a-Ga2O3 TFTs with and without channel patterned, respectively. c) Equivalent circuit of a-Ga2O3 TFT without channel patterned.




in the a-Ga2O3 phototransistor working in a quasi-two-terminal configuration with VGS = 0 V, which seriously hinders its prac-tical applications in imaging arrays. To diminish the PPC, a

positive 20 V gate bias with a pulse width of 850 ms is applied while VDS keeps at 10 V. As demonstrated in Figure 5a, the IDS curve firstly shows an instantaneous sharp peak, and then


Figure 4. a) Schematic structure of the a-Ga2O3 phototransistor on quartz. b) IDS–VGS curves recorded at different VDS in dark and under UV 254 nm light illumination (45 µW cm−2). c) IDS–VDS curves recorded at different VG. d) Time-dependent photoresponse curves under UV 254 nm light illumi-nation (45 µW cm−2) at different VDS. e) Responsivity measured at different VGS under UV 254 nm light illumination (45 µW cm−2). f) Normalized photoresponsivity spectrum of the PD biased at VDS of 20 V and VGS of 10 V.

Figure 5. a) Suppression of the PPC with a positive gate pulse. b) Decay time tested by an oscilloscope.




immediately recovers to the pristine ≈10−10 A level once the gate voltage is reset to 0 V, suggesting that the PPC phenom-enon is well suppressed due to the effective gate control. The decay time τd, defined as the time during which the current decays from 90% to 10%, was tested by an oscilloscope with a series connection with Keithley 6487 picoammeter. τd largely decreases to 5 ms with the help of the phototransistor archi-tecture, as shown in Figure 5b. It should be noted that this strategy is not only repeatable, but also energy-efficient consid-ering such a low dark current (≈10−10 A) and a low gate bias (VGS = 0 V) are exerted.

To conceptually depict the fast positive-bias-assisted PPC recovery mechanism, Figure 6 gives a schematic band diagram of the a-Ga2O3 TFT in the bias conditions of depletion and accumulation, respectively. As shown in Figure 6a, before UV illumination, the TFT works in the depletion region with a low off current since neither electric channel nor photogenerated carriers form in this case (see the corresponding dark current in Figure 4b). Under UV illumination, electrons are excited from the valence band into the conduction band and the deep-level neutral Vo defects are ionized to shallow donors Vo

2+ or Vo+,[22,23]

both of which make a big contribution to the current flowing between source and drain electrodes (see the photocurrent in Figure 4b). It is reported that the Vo

2+ states are surrounded by the outward relaxation of bonds with an energy barrier, [5,24] which impedes the neutralization of Vo

2+ states and contributes to the PPC effect. Moreover, as seen in Figure 6b, holes and the ionized shallow donors are trapped at the channel/dielectric interface, while electrons are pushed toward the bulk film and the back channel under the negative gate bias. Hence, after the light is turned off, the above-mentioned physical separation of Vo

2+ and photogenerated electrons caused by the negative gate

bias inhibits Vo2+ states rapidly recover back to the neutral Vo

states (Figure 6c), making the film still maintain the low resis-tivity. With a positive gate pulse bias, electrons readily accumu-late near the front channel (Figure 6d), corresponding to the transient increase of IDS in Figure 5. These accumulated elec-trons greatly facilitate the neutralization of Vo

2+ states and thus make the device quickly recover from the PPC state.

For a better comparison, some critical parameters are listed in Table 1.[10,11,20,25–31] It can be seen that the a-Ga2O3 pho-totransistor in our work demonstrates excellent photoelectric characteristics, including high responsivity and detectivity, as well as small τd. It can be obviously seen that a 5 ms decay time is almost at the best level in Ga2O3-based UV detectors, especially considering the simultaneously high responsivity and detectivity parameters. Besides, the presented a-Ga2O3 pho-totransistor with such good performance owns advantages of low-cost, easy fabrication and operation, which make it more promising for applications in UV detection.

3. Conclusion

In this work, a highly selective etching solution is firstly devel-oped for the preparation of a patterned a-Ga2O3 film by using TMAH. A distinct influence of the chemical etching on device performance is investigated by comparison of two common bottom-gate TFTs on SiO2/Si with a-Ga2O3 patterned and unpat-terned, respectively. Furthermore, a bottom-gate a-Ga2O3 TFT on quartz is fabricated, with the S/D electrodes in the interdig-ital finger shape. The TFT shows typical n-type oxide semicon-ductor TFT transfer curves with a high on/off ratio of ≈107 and excellent output characteristics. The device demonstrates a good


Figure 6. Schematic band diagrams of the a-Ga2O3 phototransistor, conceptually depicting the positive-bias-assisted PPC recovery mechanism. a) The device biased in depletion range. b) The device biased in depletion range with light illumination. c) After light illumination, the device biased in depletion range. d) After light illumination, the device under a positive gate pulse.




response to UV 254 nm light, such as a high light-to-dark ratio of 5 × 107 and responsivity of 5.67 × 103 A W−1. Besides, the PPC phenomenon in a-Ga2O3-based phototransistors is effec-tively suppressed by applying a positive gate pulse with a decay time as low as 5 ms. All the results suggest that the chemically etched a-Ga2O3 phototransistor is promising for high-perfor-mance UV photodetection and imaging applications.

4. Experimental SectionDevice Fabrication: Common-gated a-Ga2O3 TFTs were fabricated

on SiO2 (300 nm)/Si substrates, which were ultrasonically cleaned in acetone and isopropyl alcohol successively and then blown dry by pure nitrogen gas. An a-Ga2O3 channel layer (25 nm) was deposited by RF-sputtering technique using a Ga2O3 ceramic target (5 n pure) at room temperature and patterned by UV-lithography followed by wet chemical etching in TMAH solution. The ITO (100 nm) S/D electrodes were prepared by UV-lithography and lift-off process, where the film deposition was carried out in a sputtering chamber. Bottom-gate staggered a-Ga2O3 TFTs were fabricated on quartz substrates with interdigital finger S/D electrodes. The same cleaning method was performed on quartz substrates. First, Cr film (35 nm) was deposited by RF-magnetron sputtering and wet etched after the lithography to form bottom gate electrode. Then, Al2O3 insulator with a thickness of 110 nm was grown on the gate metal at 200 °C in ALD system and patterned by the combined process of UV-lithography and wet etching in H3PO4 solution. After that, a 25 nm a-Ga2O3 layer was synthesized by RF-magnetron sputtering, which is also patterned into discrete rectangles with dimensions of 175 µm × 600 µm via the same method as Al2O3 layer except that the etching solution is the selective TMAH. Next, interdigital finger S/D electrodes are defined by UV-lithography technique. The S/D electrodes have 15 pair fingers with 10 µm in width, 10 µm in spacing gap, and 145 µm in length. At last, a sequential deposition of Ti and Au layers were proceeded in the RF-magnetron sputtering system with a thickness of 20 and 80 nm, respectively, which are further patterned by the following lift-off process.

Device Characterization: The film thickness was measured by a stylus profiler (KLA-Tencor P-6 Stylus Profiler). The surface morphology and roughness were evaluated by AFM (Bruker Dimension EDGE). The surface morphology of two patterned a-Ga2O3 samples etched by TMAH and H3PO4 was probed by laser scanning confocal microscope (OLYMPUS LEXT OLS5000) and SEM (Zeiss Sigma 300), respectively. The electrical properties of the a-Ga2O3 TFTs were analyzed using a Keithley 4200 semiconductor characterization system and Keithley 6487 picoammeter. A handheld UV 254 nm lamp (ZF-5) and a Xe lamp with

Omni-λ 180i grating spectrometer were adopted as the light sources in time-dependent and wavelength-dependent photoresponsivity measurements, respectively.

Supporting InformationSupporting Information is available from the Wiley Online Library or from the author.

AcknowledgementsThis work was supported by the National Natural Science Foundation of China (Grant Nos. 11675280, 11674405, 61874139, 11875088, and 61904201).

Conflict of InterestThe authors declare no conflict of interest.

Keywordsamorphous gallium oxide, persistent photoconductivity, phototransistors, thin film transistors, wet chemical etching

Received: October 30, 2019Revised: January 14, 2020

Published online:

[1] S. Cui, Z. Mei, Y. Zhang, H. Liang, X. Du, Adv. Opt. Mater. 2017, 5, 1700454.

[2] S.-J. Cui, Z.-X. Mei, Y.-N. Hou, Q.-S. Chen, H.-L. Liang, Y.-H. Zhang, W.-X. Huo, X.-L. Du, Chin. Phys. B 2018, 27, 067301.

[3] S. Cui, Z. Mei, Y. Hou, M. Sun, Q. Chen, H. Liang, Y. Zhang, X. Bai, X. Du, Sci. China: Phys., Mech. Astron. 2018, 61, 107021.

[4] N. Kumar, K. Arora, M. Kumar, J. Phys. D: Appl. Phys. 2019, 52, 335103.

[5] S. Jeon, S.-E. Ahn, I. Song, C. J. Kim, U.-I. Chung, E. Lee, I. Yoo, A. Nathan, S. Lee, K. Ghaffarzadeh, J. Robertson, K. Kim, Nat. Mater. 2012, 11, 301.


Table 1. Comparison of the main parameters for the reported Ga2O3 photodetectors.

Materials and structures R [A W−1] D* [Jones] Decay time–τd [ms] Ref.

β-Ga2O3 phototransistor 3 × 103 1.3 × 1016 30 [10]

a-GaOx phototransistor 4.1 × 103 2.5 × 1013 >4 × 105 [11]

a-GaOx MSM 70.26 1.26 × 1014 200 [20]

β-Ga2O3 MSM 150 – 300 [25]

β-Ga2O3 MSM 96.13 – 78 [26]

α-Ga2O3/ZnO heterostructures 1.1 × 104 9.66 × 1012 0.238 [27]

β-Ga2O3 MSM 0.903 – <3 × 103 [28]

Au/β-Ga2O3 nanowire array 6 × 10−4 – 64 [29]

α-Ga2O3 MSM 11.5 1 × 1015 42 [30]

β-Ga2O3 (grown with N2O) MSM 26.1 1.25 × 1013 180 [31]

a-Ga2O3 phototransistor 5.67 × 103 1.87 × 1015 5 This work




[6] H. X. Jiang, J. Y. Lin, Phys. Rev. Lett. 1990, 64, 2547.[7] D. Ye, Z. Mei, H. Liang, L. liu, Y. Zhang, J. Li, Y. Liu, C. Gu, X. Du,

Sci. Rep. 2016, 6, 26169.[8] H. Liang, S. Cui, R. Su, P. Guan, Y. He, L. Yang, L. Chen, Y. Zhang,

Z. Mei, X. Du, ACS Photonics 2019, 6, 351.[9] A. Jacnotti, C. G. Van de Walle, Phys. Rev. B 2007, 76, 165202.

[10] Y. Qin, H. Dong, S. Long, Q. He, G. Jian, Y. Zhang, X. Zhou, Y. Yu, X. Hou, P. Tan, Z. Zhang, Q. Liu, H. Lv, M. Liu, IEEE Electron Device Lett. 2019, 40, 742.

[11] Y. Qin, S. Long, Q. He, H. Dong, G. Jian, Y. Zhang, X. Hou, P. Tan, Z. Zhang, Y. Lu, C. Shan, J. Wang, W. Hu, H. Lv, Q. Liu, M. Liu, Adv. Electron. Mater. 2019, 5, 1900389.

[12] W. Kim, K. S. Chu, Phys. Status Solidi A 2009, 206, 179.[13] J. Lee, K.-H. Lim, Y. S. Kim, Sci. Rep. 2018, 8, 13905.[14] K. Seong, K. Kim, S. Y. Park, Y. S. Kim, Chem. Commun. 2013, 49,

2783.[15] S. Ohira, N. Arai, Phys. Status Solidi C 2008, 5, 3116.[16] T. Oshima, T. Okuno, N. Arai, Y. Kobayashi, S. Fujita, Jpn. J. Appl.

Phys. 2009, 48, 040208.[17] T. Kamiya, K. Nomura, H. Hosono, Sci. Technol. Adv. Mater. 2010,

11, 044305.[18] R. A. Street, Adv. Mater. 2009, 21, 2007.

[19] E. Oudot, M. Gros-Jean, K. Courouble, F. Bertin, R. Duru, N. Rochat, C. Vallée, J. Vac. Sci. Technol., A 2018, 36, 01A116.

[20] L.-X. Qian, Z.-H. Wu, Y.-Y. Zhang, P. T. Lai, X.-Z. Liu, Y.-R. Li, ACS Photonics 2017, 4, 2203.

[21] H. S. Bae, M. H. Yoon, J. H. Kim, S. Im, Appl. Phys. Lett. 2003, 83, 5313.

[22] H. Lu, X. Zhou, T. Liang, L. Zhang, S. Zhang, IEEE J. Electron Devices Soc. 2017, 5, 504.

[23] A. Janotti, Appl. Phys. Lett. 2005, 87, 122102.[24] S. Lany, A. Zunger, Phys. Rev. B 2005, 72, 035215.[25] Y. Xu, Z. An, L. Zhang, Q. Feng, J. Zhang, C. Zhang, Y. Hao, Opt.

Mater. Express 2018, 8, 2941.[26] K. Arora, N. Goel, M. Kumar, ACS Photonics 2018, 5, 2391.[27] X. Chen, Y. Xu, D. Zhou, S. Yang, F. Ren, H. Lu, K. Tang, S. Gu,

R. Zhang, Y. Zheng, J. Ye, ACS Appl. Mater. Interfaces 2017, 9, 36997.[28] F.-P. Yu, S.-L. Ou, D.-S. Wuu, Opt. Mater. Express 2015, 5, 1240.[29] X. Chen, K. Liu, Z. Zhang, C. Wang, B. Li, H. Zhao, D. Zhao,

D. Shen, ACS Appl. Mater. Interfaces 2016, 8, 4185.[30] X. Hou, H. Sun, S. Long, G. S. Tompa, T. Salagaj, Y. Qin, Z. Zhang,

P. Tan, S. Yu, M. Liu, IEEE Electron Device Lett. 2019, 40, 1483.[31] D. Zhang, W. Zheng, R. C. Lin, T. T. Li, Z. J. Zhang, F. Huang,

J. Alloys Compd. 2018, 735, 150.


Documents

Boosted UV Photodetection Performance in Chemically Etched ...oxidesemi.iphy.ac.cn/result/paper/2020/2020-zyhan-AOM.pdf · 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim in – 3