A high-performance solar blind photodetector based on spinel gallium oxide thin film supported by aluminum doping

Abstract

Metastable γ-phase gallium oxide (Ga₂O₃)-based semiconductor materials have recently garnered much attention for developing high-performance wide-band-gap optoelectronic devices. However, the difficulty in the synthesis of single-phase crystals and the poor thermal stability of metastable γ-Ga₂O₃ pose great obstacles for its potential applications. In this work, we experimentally demonstrate the epitaxial growth of γ-Ga₂O₃ thin films at 750 °C on α-Al₂O₃ (0001) substrates upon aluminum (Al) doping. A solar blind photodetector based on an Al-doped γ-Ga₂O₃ thin film was successfully fabricated, and its properties were investigated in detail. Benefiting from its high crystal quality and absence of oxygen vacancies in Al-doped γ-Ga₂O₃, the solar blind photodetector exhibited a very low dark current, fast decay time, high detectivity, and excellent wavelength selectivity, along with high stability and reproducibility. Notably, the present device exhibited no obvious performance degradation when the light intensity and bias voltage were increased, indicating good self-heating dissipation, which is beneficial for future practical applications. The γ-Ga₂O₃ thin film-based photodetector might open up new possibilities to obtain high-performance thin film-based solar blind ultraviolet optoelectronic devices.

---

1. Introduction

Solar-blind ultraviolet (UV) detection with a cutoff wavelength shorter than 280 nm has a vast and growing number of military and civil surveillance applications, such as missile tracking, short-range secure communication, UV astronomy, ozone hole monitoring, and corona detection.^1–5 Commercial solar blind photodetectors (PDs) are mainly based on vacuum photomultiplier tubes, which have obvious disadvantages, such as high operation power consumption, inflexibility, and difficulty in integration, thereby limiting their practical applications.⁶

Recently, solar blind PDs based on wide-bandgap semiconductors, such as AlGaN, MgZnO, diamond, and Ga₂O₃, have attracted great attention.^7–10 However, high-quality epitaxial AlGaN films are difficult to be prepared due to their crystal deterioration with increasing Al content and the breakdown of the single wurtzite phase of MgZnO with increasing Mg content, and diamonds cannot be used to detect the entire solar-blind UV region because of their non-tunable bandgap.^10,11 Ga₂O₃ with a direct bandgap of 4.9–5.3 eV lies sharply in the solar-blind spectrum region and also exhibits a flexible tunability in bandgap by alloying with different materials, making it one of the ideal candidates to fabricate solar-blind UV PDs.^12–15 Ga₂O₃ is known to show five crystalline phases named α, β, γ, δ and ε.¹⁶ Among these phases, the most stable monoclinic β-Ga₂O₃ has been widely studied. The solar-blind PDs based on Ga₂O₃ are concentrated on β-Ga₂O₃ because of its stability and ease of preparation. However, β-Ga₂O₃ has a monoclinic structure, which makes it difficult to grow high-quality β-Ga₂O₃ on hetero-substrates. It is challenging to find material lattices matching with the β-Ga₂O₃ lattice to grow epitaxial thin films for optoelectronic applications. Thus, it is meaningful to explore the potential application of other metastable materials with a hexagonal/cubic structure, which have the same structure as those of Al₂O₃, ZnO, SiC, GaN and MgAl₂O₄.^17–19 PDs based on other metastable phases have rarely been reported because of the difficulty in synthesizing single-phase crystals with thermal stability.

A recent first-principles study on the energetics of the Ga₂O₃ polymorphs suggests that the differences in free energy between the β-phase and other metastable phases are small; therefore, the polymorph of Ga₂O₃ is selected depending upon the preparation conditions.^20–22 Moreover, doping is an important method to stabilize the crystal structure against thermal processes, making it a possible approach to obtain Ga₂O₃ with a thermally stabilized metastable phase and preventing the transition to the β-phase.^23–25 Acting as a satisfying high-quality Ga₂O₃ on hetero-substrates for developing high-performance wide-band-gap optoelectronic devices, metastable γ-phase Ga₂O₃ is considered here. Hiroyuki et al. have demonstrated the fabrication of Mn-doped γ-Ga₂O₃ at a temperature of 500 °C on α-Al₂O₃ (0001) substrates using the PLD method, which differs from the undoped film that shows the β phase.²⁶ Sam-Dong et al. have shown the enhancement of thermal stability of α-Ga₂O₃ films on sapphire substrates upon Al doping. The films can be grown at a temperature as high as 800 °C and can withstand an annealing temperature as high as 850 °C without forming the β phase.²³ Furthermore, Yue et al. have reported the synthesis of metastable γ-Ga₂O₃ nanoflowers with hexagonal nanopetals by the oxidation of metallic Ga in solutions.²⁷ The γ-Ga₂O₃ nanoflowers show excellent solar-blind detection performance; however, the poor thermal stability of the metastable phase still hinders the target application. Until now, the difficulty in synthesizing single-phase crystals and the poor thermal stability of the metastable γ-phase of Ga₂O₃ have greatly limited its potential applications in wide-band-gap high-performance optoelectronic devices.

In this work, the Al-doped γ-Ga₂O₃ thin film was epitaxially grown on α-Al₂O₃ (0001) substrates using the radio frequency magnetron sputtering technique, with the growth temperature successfully increased to 750 °C. A solar blind PD based on Al-doped γ-Ga₂O₃ thin film was successfully constructed. The PD shows excellent photoelectric properties. The γ-Ga₂O₃ thin film-based PDs might open up new possibilities for obtaining high-performance thin film-based solar blind UV optoelectronic devices.

2. Experimental section

2.1. Preparation and characterization of Ga₂O₃ thin films

The film samples were prepared on top of (0001) oriented single-crystal α-Al₂O₃ substrates using radio frequency (RF) magnetron sputtering. The Ga₂O₃ ceramic (π × 30² × 8 mm³, 99.99% purity) with/without embedded Al₂O₃ ceramic (π × 12.5² × 4 mm³, 99.99% purity) was used as a target for Al-doped/pure Ga₂O₃ thin film preparations, respectively. In the experiment, the Al₂O₃ target was covered at the glow circle position of the Ga₂O₃ target, and the incorporation amount was fine-tuned by varying the size of the coverage area. The base pressure in the chamber was 1 × 10⁻⁴ Pa, and the distance between the Ga₂O₃ target and the substrate was 6 cm. The flow rate of Ar (99.999% purity) gas was fixed to 24 sccm by a mass flow controller. The films were deposited at a working pressure of 1 Pa and RF power of 70 W for 2 h. The substrate temperature was fixed at 750 °C. The crystal structures of the as-grown films were investigated by a PANalytical X'pert PRO diffractometer using Cu Kα (λ = 1.5405 Å) radiation. The structural quality was estimated by X-ray rocking curve (XRC) measurements using Smart Lab X-ray diffraction (XRD). The surface morphologies were characterized by a Hitachi S-4800 field emission scanning electron microscope (FE-SEM). UV-visible (UV-vis) absorption spectra were taken using a Hitachi U-3900 UV-vis spectrophotometer. The elemental content was analyzed using X-ray photoelectron spectroscopy (XPS).

2.2. Fabrication and characterization of the photodetector

To construct a metal–semiconductor-metal (MSM) PD, a three-pair interdigital Au/Ti electrode was deposited on the film by RF magnetron sputtering, as illustrated in Fig. 1. The Au/Ti electrode is chosen because the affinity of Ga₂O₃ is close to the work function of Ti, making it easy to form ohmic contact. The width and length of the electrode fingers were 200 μm and 2800 μm, respectively. The spacing gap was 200 μm. A small point electrode (∼0.1 mm diameter) of indium metal was also pressed onto the Au/Ti electrode as connection points with a Cu wire. A Keithley 2450 source meter was utilized to measure the current–voltage (I–V) and time-dependent photoresponse of the Ga₂O₃ film-based PD. A UV lamp (∼7 W) with wavelengths of 254 nm and 365 nm was used as the light source, and the light irradiation power density was tuned by adjusting the distance between the light source and the sample.

Fig. 1 Schematic of an MSM structure PD.

3. Results and discussion

The XRD patterns of undoped and Al-doped Ga₂O₃ thin films are shown in Fig. 2(a). The undoped Ga₂O₃ thin film formed on the α-Al₂O₃ substrate is a monoclinic β phase, as determined by XRD (JCPDS Card No. 43-1012). Besides the diffraction peaks of the Al₂O₃ substrate, diffraction peaks located at around 18.88°, 38.17°, and 58.88° are observed for the undoped film, which corresponds well to the β (2̄01), β (4̄02) and β (6̄03) lattice planes, respectively. The Al-doped Ga₂O₃ thin film formed on the α-Al₂O₃ substrate is a cubic γ phase, as determined by XRD (JCPDS Card No. 20-0426). Besides the diffraction peaks of the Al₂O₃ substrate, diffraction peaks located at around 18.67°, 37.85°, and 58.49° are observed for the Al-doped film, which corresponds well to the γ (111), γ (222) and γ (333) lattice planes, respectively. The X-ray rocking curves of γ-Ga₂O₃ (111) and β-Ga₂O₃ (2̄01) planes are shown in Fig. 2(b) and (c) for the Al-doped and undoped films, respectively, and correspond to the FWHM values of 3672 and 5184 arcsec. The result indicates that the crystal quality of the as-grown γ-Ga₂O₃ film is better than that of the β-Ga₂O₃ film, the preparation conditions of which are similar.

Fig. 2 (a) XRD patterns of undoped and Al-doped Ga₂O₃ films. X-ray rocking curves of (b) γ-Ga₂O₃ (111) for the Al-doped film and (c) β-Ga₂O₃ (2̄01)for the undoped film.

The top-view SEM morphologies of the Al-doped γ-Ga₂O₃ film and undoped β-Ga₂O₃ film are presented in Fig. 3(a). It is obvious that the γ-Ga₂O₃ and β-Ga₂O₃ films appear to form different surface structures on (0001) sapphire substrates. For the γ-Ga₂O₃ film, irregular pseudo-pyramid structures are distributed randomly on the surface, and the grain boundary is not clear; a similar surface structure has been observed for cubic-In₂O₃ on (0001) sapphire substrates in our previous study.²⁸ The surface structures of the β-Ga₂O₃ film are columnar-like, and a clear grain boundary can be observed. Fig. 3(b) shows the UV-vis absorbance spectrum of the Al-doped γ-Ga₂O₃ film and the undoped β-Ga₂O₃ film. It is evident that all the films have significant absorption edges at ∼250 nm, near the lower edge of the solar-blind region. Moreover, the bandgap of the semiconductor film can be evaluated by extrapolating the linear region of the plots of (αhν)² and (αhν)^1/2 versus hν for direct and indirect transitions, respectively. For β-phase Ga₂O₃, a direct bandgap semiconductor, the bandgap value was determined to be 4.94 eV, as shown in the inset of Fig. 3(b). As the transition type of γ-phase Ga₂O₃ is undefined, both the direct and indirect bandgaps were estimated for the Al-doped γ-Ga₂O₃ film. The direct and indirect bandgap values were determined to be 5.08 and 4.43 eV, respectively, which were slightly higher than those for undoped γ-Ga₂O₃ (5.0 and 4.4 eV, respectively).¹⁹

Fig. 3 (a) SEM images of the Al-doped γ-Ga₂O₃ and undoped β-Ga₂O₃ films. (b) UV-vis absorbance spectra of the Al-doped γ-Ga₂O₃ and undoped β-Ga₂O₃ films with the plot of (αhν)² and (αhν)^1/2 vs. hν in the inset.

XPS analysis was performed to elucidate the chemical compositions and oxygen deficiency in the as-grown films. The surface of the films was etched by Ar⁺ bombardment before the XPS measurement. Fig. 4(a) shows the wide survey spectrum in the binding energy range from −40 eV to 1350 eV to identify the elements in the Al-doped γ-Ga₂O₃ film. XPS peaks of Ga, O, Al and C were indicated in the Figure, and there were no other elements detected from the wide survey spectrum of the film. The charge-shift spectrum was calibrated using the C 1s peak at 284.8 eV. The Al 2p core level peak of the Al-doped γ-Ga₂O₃ film is presented in Fig. 4(b), and the energy peak center is observed at 71.1 eV. The peak position shows a significant shift compared to the Al 2p peak in pure Al₂O₃ film on the same substrate (74.5 eV). A similar phenomenon was also observed with a negative shift for the Al 2p peak during the formation of Al–O–Ga bonds.¹⁴ In their report, the peak position was noted at 73.9 eV in the (Al_0.22Ga_0.78)₂O₃ film; the significant shift in our result indicates that Al is present in the Al-doped γ-Ga₂O₃ film in the form of Al–O–Ga bonds. We have not yet identified whether the Al–O–Ga bonds are the key factors to obtain stabilized γ-Ga₂O₃, but they are a very significant sign to understand if the growing film is in the γ phase or not. The Ga 3d core level peaks of the Al-doped γ-Ga₂O₃ and undoped β-Ga₂O₃ films are shown in Fig. 4(c), with the energy peak centers observed at 19.7 eV and 20.1 eV, respectively. The O 1s core level peaks of the Al-doped γ-Ga₂O₃ and undoped β-Ga₂O₃ films are shown in Fig. 4(d), and the energy peaks are observed at 530.2 eV and 530.6 eV, respectively. The O 1s peaks can be divided into two components: I and II; peak I is attributed to lattice oxygen ions, and peak II is attributed to the oxygen ions in the oxygen vacancy region and surface defects of the film.^5,10 The intensity of oxygen vacancies and surface defects in the film has a significant negative influence on the performance of Ga₂O₃ thin film-based PDs, such as dark current and response speed.²⁹ The peak ratios of II/I were 1/5 and 1/3 for Al-doped γ-Ga₂O₃ and pure β-Ga₂O₃, respectively [Fig. 4(d)], which indicates a lower intensity of oxygen vacancies and surface defects in the Al-doped γ-Ga₂O₃ film than in the pure β-Ga₂O₃ film.

Fig. 4 (a) XPS wide survey spectrum and (b) core level of Al 2p for the Al-doped γ-Ga₂O₃ film. Core levels of (c) Ga 3d and (d) O 1s for Al-doped γ-Ga₂O₃ and undoped β-Ga₂O₃ films.

Fig. 5 shows the room-temperature I–V characteristics of the PD based on an Al-doped γ-Ga₂O₃ thin film in the dark, under 365 nm (100 μW cm⁻²) and 254 nm (with various light intensities) light illumination. The I–V curve in the dark is linear, indicating that the contact is of an ohmic type. The corresponding resistance is 19.5 GΩ, and the dark current is about 2.57 nA at a voltage of 50 V, as shown in Fig. 5(a), which is much lower than that of undoped β-Ga₂O₃ films (about 100 nA at a voltage of 10 V), as per our previous reports. The I–V curve under 365 nm light illumination indicates that the device is almost insensitive to 365 nm light illumination, as shown in Fig. 5(a). Meanwhile, the device has significant sensitivity for 254 nm light illumination, as shown in Fig. 5(b).

Fig. 5 (a) I–V characteristics of the Al-doped γ-Ga₂O₃ thin film-based MSM structure PD in the dark and under a 365 nm light illumination. (b) I–V characteristics of the device in the dark and under a 254 nm light illumination with various light intensities.

In order to further investigate the performance of the Al-doped γ-Ga₂O₃ thin film-based PD, the time-dependent photoresponse to solar-blind UV illuminations was investigated, as shown in Fig. 6. Fig. 6(a) and (b) show the time-dependent photoresponse of the device to 254 nm UV light illuminations with varied light intensities and bias voltages, respectively. The device shows excellent photoelectric performance, with the photocurrent gradually increasing with increasing incident light intensities and bias voltages. The larger the voltage applied, the higher the I_photo obtained, following the basic photoconductance rule, which works as a photo-sensitive resistance. The higher the light intensity, the higher the I_photo output, as more photo-generated carriers would be devoted to I_photo at a certain voltage. Another important performance parameter of solar blind PDs is the response time. To study the response time of the device, the current rising and decaying process can be derived by fitting the photoresponse curve with the following relaxation equation:³⁰

where I₀ is the photocurrent in the steady state, A and B are the constant values, t is the time, and τ₁ and τ₂ are the relaxation time constants. As shown in Fig. 6(c) and (d), the photoresponse processes fit quite well. We note that the rising edge consists of two components (τ_r1 and τ_r2), and the decaying edge is steep, consisting of only one component. It has been estimated that the rise edge constants τ_r1 and τ_r2 are around 2.39 s/12.57 s, 1.52 s/8.16 s, 1.02 s/6.92 s and 0.88 s/5.89 s for the device under 20 μW cm⁻², 40 μW cm⁻², 60 μW cm⁻² and 80 μW cm⁻² light illuminations at a bias voltage of 50 V, respectively. In contrast, the rise edge constants τ_r1 and τ_r2 are estimated to be 1.22 s/8.86 s, 1.04 s/7.95 s, 0.90 s/5.84 s, 0.84 s/7.03 s and 0.71 s/6.05 s for the device under 100 μW cm⁻² light illuminations at bias voltages of 10 V, 20 V, 30 V, 40 V and 50 V, respectively. It is obvious that the rise speed increases with both light intensity and bias voltage. Generally, the fast-response component can be attributed to the rapid change in carrier concentration that occurs as soon as the light is turned on or off, while the slow-response component is caused by carrier trapping/releasing owing to the existence of oxygen vacancies and other surface defects in the Al-doped γ-Ga₂O₃ film. It is evident that, when there are plenty of photon-generated carriers under stronger intensities of light illumination, the trapping/releasing effect is easier to saturate, so the response time is shorter; when the applied bias voltage is higher, the photo-generated electrons and holes undergo a rapid separation and then transfer toward the corresponding electrodes, resulting in a faster response speed. The decay edge constants τ_d are estimated to be 0.44 s, 0.42 s, 0.41 s and 0.40 s for the device under 20 μW cm⁻², 40 μW cm⁻², 60 μW cm⁻² and 80 μW cm⁻² light illuminations at a bias voltage of 50 V, respectively. The decay edge constants τ_d are estimated to be 0.46 s, 0.43 s, 0.43 s, 0.41 s and 0.40 s for the device under 100 μW cm⁻² light illumination at bias voltages of 10 V, 20 V, 30 V, 40 V and 50 V, respectively. Due to the absence of oxygen vacancies in Al doped γ-Ga₂O₃ film, the trapping/releasing effect is a localization transport process without plentiful of photon-generated carriers when the light is off. It can hardly influence the photoconductance of the device, so the decaying edge is quite steep.

Fig. 6 Time-dependent photoresponse of the Al-doped γ-Ga₂O₃ thin film-based MSM structure PD measured under different conditions: (a) photoresponse of the device under different light intensities at 50 V; (b) photoresponse of the device at different bias voltages under 100 μW cm²; (c) experimental curve and fitted curve of the current rising and decaying process of the device under different light intensities at 50 V; (d) experimental curve and fitted curve of the current rising and decaying process of the device at different bias voltages under 100 μW cm²; (e) photoresponse stability of the device during a long time test and after storage for one month.

To quantitatively assess the device performance of the present solar blind photodetector, both responsivity (R_λ) and detectivity (D*) were calculated, as shown in Fig. 7(a) and (c) as a function of light intensity and bias voltage, respectively. R_λ, which is defined as the photocurrent generated per unit power of incident light on the effective area of a PD, is calculated by the following equation:²

R_λ = (I_light − I_dark)/(P_λ·S) (2)

where I_light is the light current, I_dark is the dark current of the device when illuminated with a light source, P_λ is the light intensity illuminated on the device, S is the effective illuminated area (S = 3.24 mm²), and λ is the wavelength of the illuminating light.

Fig. 7 (a) Responsivity and detectivity of the Al-doped γ-Ga₂O₃ thin film-based MSM structure PD as a function of bias voltage. (b) External quantum efficiency of the device as a function of bias voltage. (c) Responsivity and detectivity of the device as a function of light intensity. (d) External quantum efficiency of the device as a function of light intensity.

Detectivity, which is usually used to evaluate the smallest detectable of a PD, can be calculated as follows:³¹

where R_λ is the responsivity, A is the effective area of the present device channel (A = 3.24 mm²), e is the electronic charge, and I_dark is the dark current. It is clear that both the R_λ and D* values increase with increasing light intensity and bias voltage. Specifically, the responsivity and detectivity are as high as 0.26 A W⁻¹ and 1.79 × 10¹² under 100 μW cm⁻² light illumination at 50 V, respectively. Notably, the present device exhibits no obvious performance degradation with increasing light intensity and bias voltage, indicating that the device has good self-heating dissipation. Moreover, the external quantum efficiency (EQE) is also the key parameter to evaluate the performance of a PD, which is defined as the number of electrons probed per incident photon and can be calculated as follows:³²where R_λ is the responsivity, h is Planck's constant, c is the velocity of light, e is the electronic charge, and λ is the exciting wavelength of the solar blind light. The EQE gradually increases with increasing light intensity and bias voltage, reaching a maximum of 124.92% under 100 μW cm⁻² light illumination at 50 V, as shown in Fig. 7(b) and (d).

In order to investigate the spectral selectivity of the present device, the spectral responsivity was measured in the range of 225 to 600 nm, as shown in Fig. 8. It can be observed that the device shows excellent wavelength selectivity: it has highest responsivity at a wavelength of about 230 nm; as the wavelength increases, the responsivity gradually decreases and reaches a minimum value at about 280 nm. The rejection ratio of responsivity values for the solar blind region versus the solar visible region was estimated by dividing the responsivity at 230 nm by that at 280 nm. The excellent rejection ratio was evaluated at 60.98, indicating an excellent solar-blind UV response. The higher rejection ratio can be attributed to the better crystal quality and the absence of oxygen vacancies in Al-doped γ-Ga₂O₃.³³

Fig. 8 Normalized spectral selectivity of the Al-doped γ-Ga₂O₃ thin film-based MSM structure PD.

4. Conclusion

The Al-doped γ-Ga₂O₃ thin film was epitaxially grown on α-Al₂O₃ (0001) substrates by radio frequency magnetron sputtering, and the growth temperature was successfully increased to 750 °C. The solar blind UV PD based on the as-grown thin film exhibits excellent solar-blind UV photoelectric properties. Under 100 μW cm⁻² 254 nm light illumination at a bias voltage of 50 V, the device exhibits a rise and decay time of 0.71 s/6.05 s and 0.40 s, responsivity of 0.26 A W⁻¹, detectivity of 1.79 × 10¹², and an EQE of 124.92%, along with high stability and reproducibility. Remarkably, the present device exhibits no obvious performance degradation with increasing light intensity and bias voltage. It indicates that the device has good self-heating dissipation, which will be beneficial for future practical applications. Moreover, the device shows excellent wavelength selectivity with a solar blind/visible rejection ratio of 60.98, which is nearly blind to photons with wavelengths longer than 280 nm. These results suggest that γ-Ga₂O₃ thin film-based PDs can open up new possibilities for obtaining high performance thin film-based solar blind UV optoelectronic devices.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting this article are available from the corresponding author upon reasonable request.

Acknowledgements

National Key R&D Program of China (2023YFB2906000); National Natural Science Foundation of China (62375055); Guangdong Basic and Applied Basic Research Foundation (2023B1515020088); Basic and Applied Basic Research Project of Guangzhou Basic Research Program (2025A04J5417) and the Start-Up Funding of Guangdong Polytechnic Normal University (2022SDKYA013) are greatly acknowledged.

References

1. Y. Zhao, J. Zhang and D. Jiang, et al., Ultraviolet photodetector based on a MgZnO film grown by radio-frequency magnetron sputtering, ACS Appl. Mater. Interfaces, 2009, 1(11), 2428–2430.
2. W. Cui, D. Y. Guo and X. L. Zhao, et al., Solar-blind photodetector based on Ga₂O₃ nanowires array film growth from inserted Al₂O₃ ultrathin interlayers for improving responsivity, RSC Adv., 2016, 6(103), 100683–100689.
3. Z. Li, T. Yan and X. Fang, Low-dimensional wide-bandgap semiconductors for UV photodetectors, Nat. Rev. Mater., 2023, 8(9), 587–603.
4. X. Zhu, Y. Wu and Z. Pan, et al., Advancements in Ga₂O₃-based heterojunction ultraviolet photodetectors: Types, fabrication techniques, and integrated materials for enhancing photoelectric conversion efficiency, J. Alloys Compd., 2025, 1010, 177757.
5. Y. H. An, Z. S. Gao and Y. Guo, et al., One ε-Ga₂O₃-based solar-blind Schottky photodetector emphasizing high photocurrent gain and photocurrent-intensity linearity, Chin. Phys. B, 2023, 32(5), 058502.
6. Y. Li, T. Tokizono and M. Liao, et al., Efficient Assembly of Bridged β-Ga₂O₃ Nanowires for Solar-Blind Photodetection, ACS Appl. Mater. Interfaces, 2010, 20(22), 3972–3978.
7. Y. Gao, J. Yang and X. Ji, et al., Semipolar (1122) AlGaN-based solar-blind ultraviolet photodetectors with fast response, ACS Appl. Mater. Interfaces, 2022, 14(18), 21232–21241.
8. M. M. Fan, K. W. Liu and X. Chen, et al., Mechanism of Excellent Photoelectric Characteristics in Mixed-Phase ZnMgO Ultraviolet Photodetectors with Single Cutoff Wavelength, ACS Appl. Mater. Interfaces, 2015, 7(37), 20600–20606.
9. J. Ye, S. Jin and Y. Cheng, Photocurrent ambipolar behavior in phase junction of a Ga₂O₃ porous nanostructure for solar-blind light control logic devices, ACS Appl. Mater. Interfaces, 2024, 16(20), 26512–26520.
10. R. Jin, J. Liu and S. Chen, et al., Ultrasensitive Self-Powered Flexible Crystalline β-Ga₂O₃-Based Photodetector Obtained through Lattice Symmetry and Band Alignment Engineering, ACS Appl. Mater. Interfaces, 2024, 16(32), 42406–42414.
11. T. Oshima, T. Okuno and S. Fujita, Ga₂O₃ Thin Film Growth onc-Plane Sapphire Substrates by Molecular Beam Epitaxy for Deep-Ultraviolet Photodetectors, Jpn. J. Appl. Phys., 2007, 46(11R), 7217–7220.
12. Z. Y. Wang, G. Zhang and X. J. Zhang, et al., Polarization-sensitive artificial optoelectronic synapse based on anisotropic β-Ga₂O₃ single crystal for neuromorphic vision systems and information encryption, Adv. Opt. Mater., 2024, 12(29), 2401256.
13. T. Oshima, Y. Kato and M. Oda, et al., Epitaxial growth of gamma-(Al_xGa_1-x)O₃ alloy films for band-gap engineering, Appl. Phys. Express, 2017, 10(5), 051104.
14. H. Xu, Y. Weng and K. Chen, et al., Ultra-low BER encrypted communication based on self-powered bipolar photoresponse ultraviolet photodetector, Adv. Opt. Mater., 2024, 10, 2402238.
15. P. Li, L. Dong and C. Li, et al., Indium doping-assisted monolayer Ga₂O₃ exfoliation for performance-enhanced MOSFETs, Nanoscale, 2023, 15(28), 12105–12115.
16. R. Roy, V. Hill and E. Osborn, Polymorphism of Ga₂O₃ and the System Ga₂O₃-H₂O, J. Am. Chem. Soc., 1952, 74(3), 719–722.
17. X. C. Xia, Y. P. Chen and Q. J. Feng, et al., Hexagonal phase-pure wide band gap epsilon-Ga₂O₃ films grown on 6H-SiC substrates by metal organic chemical vapor deposition, Appl. Phys. Lett., 2016, 108(20), 202103.
18. S. Fujita, M. Oda and K. Kaneko, et al., Evolution of corundum-structured III-oxide semiconductors: Growth, properties, and devices, Jpn. J. Appl. Phys., 2016, 55(12), 1202A3.
19. T. Oshima, T. Nakazono and A. Mukai, et al., Epitaxial growth of gamma-Ga₂O₃ films by mist chemical vapor deposition, J. Cryst. Growth, 2012, 359(15), 60–63.
20. S. Yoshioka, H. Hayashi and A. Kuwabara, et al., Structures and energetics of Ga₂O₃ polymorphs, J. Phys.: Condens. Matter, 2007, 19(34), 346211.
21. H. Nishinaka, D. Tahara and M. Yoshimoto, Heteroepitaxial growth of epsilon-Ga₂O₃ thin films on cubic (111) MgO and (111) yttria-stablized zirconia substrates by mist chemical vapor deposition, Jpn. J. Appl. Phys., 2016, 55(12), 1202BC.
22. S. X. Ge and Z. Zheng, Phase controllable synthesis of unusual epsilon-Ga₂O₃ single crystals with promising luminescence property, Solid State Sci., 2009, 11(9), 1592–1596.
23. S. D. Lee, Y. Ito and K. Kaneko, et al., Enhanced thermal stability of alpha gallium oxide films supported by aluminum doping, Jpn. J. Appl. Phys., 2015, 54(3), 030301.
24. R. Lorenzi, N. V. Golubev and E. S. Ignat'eva, et al., Defect-assisted photocatalytic activity of glass-embedded gallium oxide nanocrystals, J. Colloid Interface Sci., 2022, 608(3), 2830–2838.
25. G. Yang, H. A. Ni and H. D. Liu, et al., The doping effect on the crystal structure and electrochemical properties of LiMn_xM_1−xPO₄ (M = Mg, V, Fe, Co, Gd), J. Power Sources, 2011, 196(10), 4747–4755.
26. H. Hayashi, R. Huang and H. Ikeno, et al., Room temperature ferromagnetism in Mn-doped gamma-Ga₂O₃ with spinel structure, Appl. Phys. Lett., 2006, 89(18), 181903.
27. Y. Teng, L. X. Song and A. Ponchel, et al., Self-Assembled Metastable gamma-Ga₂O₃ Nanofl owers with Hexagonal Nanopetals for Solar-Blind Photodetection, Adv. Mater., 2014, 26(36), 6238–6243.
28. W. Cui, X. L. Zhao and Y. H. An, et al., Direct charge carrier injection into Ga₂O₃ thin films using an In₂O₃ cathode buffer layer: their optical, electrical and surface state properties, J. Phys. D Appl. Phys., 2017, 50(13), 135109.
29. D. Guo, P. Li and Z. Wu, et al., Inhibition of unintentional extra carriers by Mn valence change for high insulating devices, Sci. Rep., 2016, 6(1), 24190.
30. D. Y. Guo, Z. P. Wu and P. G. Li, et al., Fabrication of beta-Ga₂O₃ thin films and solar-blind photodetectors by laser MBE technology, Opt. Mater. Express, 2014, 4(5), 1067–1076.
31. W. Y. Kong, G. A. Wu and K. Y. Wang, et al., Graphene-beta-Ga₂O₃ Heterojunction for Highly Sensitive Deep UV Photodetector Application, Adv. Mater., 2016, 28(48), 10725.
32. J. M. Wu and W. E. Chang, Ultrahigh Responsivity and External Quantum Efficiency of an Ultraviolet-Light Photodetector Based on a Single VO₂ Microwire, ACS Appl. Mater. Interfaces, 2014, 6(16), 14286–14292.
33. A. S. Pratiyush, S. Krishnamoorthy and S. V. Solanke, et al., High responsivity in molecular beam epitaxy grown beta-Ga₂O₃ metal semiconductor metal solar blind deep-UV photodetector, Appl. Phys. Lett., 2017, 110(22), 221107.

---