Fully Transparent and High‐Performance ε‐Ga2O3 Photodetector Arrays for Solar‐Blind Imaging and Deep‐Ultraviolet Communication

Abstract

With the development of new broadband semiconductor materials and the expansion of solar‐blind deep‐ultraviolet (SBDU) optoelectronics, the demand for high‐performance fully transparent array integration has also increased. Herein, metal–semiconductor–metal (MSM)‐type fully transparent SBDU photodetector arrays (PDAs), based on commercially available ε‐Ga₂O₃ films, are constructed using indium tin oxide (ITO), indium gallium zinc oxide (IGZO), and aluminum zinc oxide (AZO) transparent conductive oxides as electrodes. The experimental results show that the overall performance of the fully transparent SBDU PDAs used in this study places them at the forefront of ε‐Ga₂O₃‐based photodetectors (PDs). All three PDs not only exhibit ultra‐high responsivity (286.2, 284.1, and 262.1 A W⁻¹), remarkable detectivity (4.73 × 10¹⁴, 5.06 × 10¹⁴, and 2.16 × 10¹⁵ Jones), and excellent stability, but also fast photoresponse time (5.6/7.2, 6.9/9.5, and 70/200 ms). It is found that the resistivity and carrier concentration of transparent conductive oxides significantly influence device performance. Consequently, it is demonstrated, for the first time, that fully transparent ε‐Ga₂O₃‐based SBDU PDAs can be used as ultraviolet imagers and signal receivers integrated into a homemade solar‐blind ultraviolet communication (SBUC) system to transmit text data. These results indicate great potential for future applications of fully transparent ε‐Ga₂O₃ thin‐film solar‐blind photodetectors in next‐generation artificial intelligence eyes, smart windows, light field cameras, and solar‐blind imaging, among others.

Introduction

During the past decade, due to extensive application in modern industrial agriculture and military intelligence integration development, transistors,^{[ ]} photodetectors,^{[ ]} light-emitting diodes,^{[ ]} and other power electronics have attracted widespread research.^{[ ]} Photodetectors are photoelectronic devices that convert optical signals (UV-IR) into electrical signals. The emergence of wide bandgap semiconductors has resulted in solar-blind deep-ultraviolet (SBDU) photodetectors (PDs) with extremely low background noise receiving considerable attention.^{[ ]} In general, SBDU PDs are constructed of semiconductor materials with a bandwidth greater than 4.4 eV, including Al _x Ga _x−1N,^{[ ]} ZnMgO,^{[ ]} SiC,^{[ ]} Ga₂O₃,^{[ ]} and diamond.^{[ ]} A new type of ultra-wide band semiconductor material, Ga₂O₃ with a wide bandgap of ≈4.9 eV, does not require a complex alloying process and possesses high chemical stability, and can therefore be considered a natural material for solar-blind signals detection. To date, various types of Ga₂O₃-based PDs have been proposed, including p-n junctions,^{[ ]} heterojunctions,^{[ ]} Schottky diodes,^{[ ]} avalanches,^{[ ]} metal–semiconductor–metal (MSM)^{[ ]} and photoelectrochemical^{[ ]} PDs. Among them, MSM-type SBDU PDs have been widely studied as their preparation is simple and they are easily integrated with electronic devices. Moreover, MSM-type Ga₂O₃-based PDs maintain a high performance with a low dark current, large light-to-dark ratio, excellent responsivity, and a fast photoresponse time.^{[ ]} These high-performance and easy-to-integrate MSM-type Ga₂O₃-based PDs present great potential in civil and military applications, including medical imaging, secure optical communication, environmental and fire monitoring, space exploration, and high-voltage leakage detection.^{[ ]}

As early as 2018, Chen et al. prepared an MSM-type β-Ga₂O₃ film-based PD array (4 × 4) to obtain clear target images in solar-blind imaging.^{[ ]} When the bias voltage is 45 V, the peak responsivity and detectivity of the PD array unit are ≈12.4 A W⁻¹ and 1.9 × 10¹² Jones, respectively. Recently, Qian et al. fabricated a large-scale high-uniformity 32 × 32 image sensor array, based on amorphous-Ga₂O₃ (a-Ga₂O₃) SBDU PDs.^{[ ]} The MSM-type a-Ga₂O₃ SBDU PD exhibits a super-high sensitivity of 733 A W⁻¹, fast photoresponse speed of 18 ms, ultra-high photo-to-dark current ratio of 3.9 × 10^7, and an excellent specific detectivity of 3.9 × 10¹⁶ Jones. Optical communication systems typically rely on light radiation to transmit information. Compared with traditional radio-frequency (RF) wireless communication, optical communication light signals easily bypass various terrain features by scattering and reflecting without interference, and thus can provide high-speed, ultra-low latency, green, and low-cost communication services.^{[ ]} In 2019, Lin et al. designed a high-performance solar-blind photodetector using a diamond wafer.^{[ ]} The device responsivity and response time are 13 A W⁻¹ and 1.3 μs/203 μs, respectively. Based on this, the authors used the novel diamond photodetector as the signal receiver to construct a solar-blind deep-ultraviolet communication (SDUC) system. Very recently, our group built a high-performance β-Ga₂O₃-based solar-blind photodetector with a remarkable dark current of 82 fA and fast photoresponse speed of 11/240 μs for the purposes of solar-blind deep-ultraviolet communication.^{[ ]} It is the first time that a signal receiver consisting of a β-Ga₂O₃-based PD integrated into a tailor-made SDUC system has been verified to transmit American Standard Code for Information Interchange (ASCII) codes. This proves the validity and applicability of β-Ga₂O₃ solar-blind deep-ultraviolet photodetectors for future use in secure communications. Among all the commonly identified polymorphs of Ga₂O₃, ε-Ga₂O_3, a metastable phase is regarded as the hexagon system belonging to the P63mc space group with high symmetry and low anisotropy crystal structure, which has unique advantages in the construction of optoelectronic devices.^{[ ]} However, there is no known research regarding the application of ε-Ga₂O₃ film-based photodetector arrays for solar-blind imaging and deep-ultraviolet communication.

Generally, more stable metals with excellent electrical conductivity, such as Ti, Au, Ag, and Pt, are used as PD electrode materials. Due to the high cost and opaque nature of PDs fabricated from these metal electrodes, they cannot be applied in next-generation transparent optoelectronic technologies, such as secure communications, smart windows, artificial intelligence glasses, and light field cameras.^{[ ]} Therefore, transparent conductive oxides (TCOs) with high electrical conductivity and transparency have attracted widespread interest.^{[ ]} Compared to commonly used noble metals, TCOs are not only transparent, but are also an abundant earth resource, which, therefore, reduces costs. Moreover, due to the non-stoichiometric properties of the components, the optoelectronic properties of TCOs thin films can be adjusted over a wide range by doping and optimizing the preparation process.^{[ ]} These inherent characteristics make TCOs ideal for large-scale optoelectronics industries. For example, Wu et al. prepared an organic photodetector with a remarkable external quantum efficiency of 74%, an excellent responsivity of 0.64 A W⁻¹, and a satisfactory detectivity of 6.6 × 10¹² Jones, by using indium-doped zinc oxide electrodes.^{[ ]}

This study demonstrates three ε-Ga₂O₃-based SBDU photodetector arrays (PDAs), with different transparent conductive oxide electrodes, including indium tin oxide (ITO), indium gallium zinc oxide (IGZO), and aluminum zinc oxide (AZO). High-uniformity 5 × 4 arrays of fully transparent MSM-type SBDU PDs, based on post-annealed ε-Ga₂O₃ films, were constructed by radio frequency (rf) magnetron sputtering and masking processes. The photoelectric properties of the three fully transparent PDAs were systematically studied, and it is found that all three PDAs not only exhibit ultra-high responsivity, remarkable detectivity, and excellent stability, but also ultra-fast photoresponse time. Finally, we explored the potential applications of three ε-Ga₂O₃-based SBDU PDAs in solar-blind imaging and deep-ultraviolet communication.

Results and Discussion

Figure  shows the X-ray diffraction (XRD) pattern of the post-annealed ε-Ga₂O₃ film. It can be seen from the XRD pattern that three diffraction peaks located at 19.33°, 39.2°, and 60.2° correspond with the (002), (004), and (006) crystal planes of the ε-Ga₂O₃, respectively.^{[ ]} Apart from the three ε-Ga₂O₃ diffraction peaks (001), no other diffraction peaks are evident, indicating that the film annealed by argon at high temperature is a single-phase hexagonal gallium oxide. Figure  exhibits the ε-Ga₂O₃ film absorption spectrum, where it is obvious that the film absorption rate in the 300–800 nm band is almost zero, which is a powerful prerequisite for the next step in the preparation of a fully transparent device. Furthermore, the film has a steep absorption edge of around 280 nm, and the ε-Ga₂O₃ film optical bandgap is estimated to be approximately 4.9 eV from the Tauc plot.^{[ ]} In this study, three transparent electrodes of ITO, AZO, and IGZO are used to fabricate fully transparent MSM-type SBDU PDAs (5 × 4). The structure of a single PD is shown in the inset of Figure . At the same time, a surface scan X-ray energy dispersion diagram (EDS) of the ITO electrode single PD was obtained, as shown in Figure . Figure  displays the specific element mapping, showing that the PD is composed of ε-Ga₂O₃ thin film photosensitive material constituting a Ga/O element and ITO electrode material consisting of In/Sn/O elements. It is apparent that the In/Sn element map possesses a good boundary shape, indicating that the electrodes deposited by magnetron sputtering have good uniformity.^{[ ]}

[IMAGE OMITTED. SEE PDF]

To initially explore and compare the effects of different electrodes on the photoelectric properties of ε-Ga₂O₃ films, three different fully transparent MSM-type PDAs (5 × 4) with ITO, IGZO, and AZO electrodes were fabricated, and a single PD was selected for optoelectronic performance evaluation. Figure  illustrates the current–voltage (I–V) characteristic curves of the three PDs in the dark and under 254 nm illumination with a light intensity of 1000 μw cm⁻². Surprisingly, the photocurrent of the PD with ITO electrodes (ITO-PD) reached 3.3 × 10⁻⁴ A at 10 V bias, and the photocurrent of the other two PDs also achieved 1.03 × 10⁻⁴ and 3.37 × 10⁻⁵A, respectively. Meanwhile, the photo-dark ratio of all three PDs is greater than 5 × 10⁴ at 10 V bias, and the photo-dark current ratio (PDCR) of PD with AZO electrodes (AZO-PD) is the largest, attaining 2.6 × 10⁵. Figure  depicts the single-period transient photoresponse characteristic curves (I–T) of the three PDs. The photocurrent of the SBDU PD increases or decays rapidly when the 254 nm light is switched on or off. Generally, the ability of a PD to track this fast-changing light signal is called the photoresponse speed. Using the second-order exponential equation to fit the transient photoresponse curves, the photoresponse speed of the three PDs can be quantitatively compared. The equation is as follows^{[ ]} 1 $\begin{array}{c}I=I_0+A{e}^{\raisebox{1ex}{$-\mathrm{t}$}\!\left/ \!\raisebox{-1ex}{$\tau 1$}\right.}+B{e}^{\raisebox{1ex}{$-\mathrm{t}$}\!\left/ \!\raisebox{-1ex}{$\text{τ}2$}\right.}\end{array}$ where I ₀ is the steady-state photocurrent, A and B are constants, and τ ₁ and τ ₂ are two relaxation time constants representing the fast and slow response times, respectively. In addition, the specific rise time (τ _r) and decay time (τ _d) of the three devices are fitted with second-order exponential equations, respectively, as displayed in Figure . Obviously, the photoresponse speeds of the PDs with ITO and IGZO electrodes (ITO-PD and IGZO-PD) are extremely fast, with τ _r1 and τ _d1 of 5.6/7.2 and 6.9/9.5 ms, respectively. Although the AZO-PD photoresponse time of 70/200 ms is slower than the other two PDs, it is above the average of previously reported solar-blind PDs.^{[ ]}

[IMAGE OMITTED. SEE PDF]

To further estimate and compare the optoelectronic performance of the three fully transparent MSM-type PDs, more systematic measurement and performance calculations were performed, as shown in Figure . Figure  displays the I–V curves of the PDs with ITO, AZO, and IGZO electrodes under 254 nm illumination and dark conditions, respectively. As seen, the I–V curves of AZO-PD and IGZO-PD display straight linear relationships, indicating that AZO/IGZO electrodes and ε-Ga₂O₃ film are good ohmic contacts, while ITO electrodes and ε-Ga₂O₃ film are back-to-back Schottky contacts. Meanwhile, the photocurrent of all three PDs increases with increased light intensity, which is related to the high light intensity promoting the generation of electron–hole pairs. The relationship between the specific photocurrent and light intensity at 10 V bias is shown in Figure . In general, this relationship conforms to the power-law function: $\mathrm{I}\propto P^{\text{θ}}$,^{[ ]} where P is the light intensity and θ is an exponent corresponding to the photogenerated carrier recombination. After fitting, it is obvious that the ITO-PD results in an exponent of 0.83, while the fitted curves of the IGZO-PD and AZO-PD attain two different exponent values: θ = 0.77/0.53 for low optical intensity density and θ = 0.42/0.28 for high optical intensity, respectively. These are slightly smaller than the ideal value (θ = 1). These results demonstrate that all devices suffer recombination losses, among which the recombination loss of ITO-PD is slight, while the recombination losses of IGZO-PD and AZO-PD are relatively larger. This may be related to defects formed by the electrode material and ε-Ga₂O₃ film.^{[ ]} Due to the increased recombination activity of photo-generated carriers as the light intensity increases, and a higher defect density existing between the electrodes and films compared to the ITO electrode, both IGZO-PD and AZO-PD have two fitting index values which are relatively smaller at high light intensities.^{[ ]} In addition to the photocurrent, the PDs PDCR also depends on the dark current and light intensity. The PDCR of all PDs augments with increasing light intensity, which corresponds to the change law of photocurrent with light intensity, as shown in Figure . However, the AZO-PD PDCR is the largest, due to the relatively small dark current, and is as high as 2.6 × 10⁵ under 254 nm light with an intensity of 1000 μW cm⁻².

[IMAGE OMITTED. SEE PDF]

To quantitatively compare the device performance of the three fully transparent ε-Ga₂O₃ film-based solar-blind PDs, the PD responsivity (R) and detectivity (D *) under different light intensities at 10 V bias were calculated, as shown in Figure . R is defined as the effective photocurrent generated by incident light, per unit of light power, over the photodetector effective area, and can be expressed as^{[ ]} 2 $\begin{array}{c}R=\frac{I_{\text{photo}}-I_{\text{dark}}}{P_{\text{λ}}S}\end{array}$ where I _photo is the photocurrent, I _dark is the dark current, S is the effective light area (2.8 × 10⁻³cm²), and P _λ denotes the light intensity. The R of all three PDs decreases with increasing light intensity, and the maximum R of the PDs are 286.2, 284.1, and 262.1 A W⁻¹ under 254 nm illumination with light intensity (5 μW cm⁻²⁾ at 10 V, respectively, as shown in Figure . D * is related to the noise equivalent power and is used to evaluate the device's ability to detect weak signals in a noisy environment, and can be expressed by^{[ ]} 3 $\begin{array}{c}D=\frac{R\sqrt{S}}{\sqrt{2e{I}_{\text{dark}}}}\end{array}$ where e is the electron charge (1.6 × 10⁻¹⁹ C). Figure  illustrates the dependence of D * on light intensity. As observed for R as a function of light intensity, D * also decreases with increasing light intensity, and the three PDs possess ultrahigh D * of 4.73 × 10¹⁴, 5.06 × 10¹⁴, and 2.16 × 10¹⁵ Jones under 254 nm light with 5 μW cm⁻² light intensity at 10 V bias, respectively. As with the results of the power-law function fitting, the three PDs incur serious recombination loss, so the R and D * of the device gradually decreases with increasing light intensity.^{[ ]} Meanwhile, it is also noteworthy that the ITO-PD has the largest responsivity but the smallest detectivity, under 254 nm illumination with 5 μW cm⁻² light intensity due to it possessing the largest photocurrent and a large dark current. Repeatability and stability are also vital quality factors for solar-blind PDs. Figure  displays the transient photoresponse (I–t) characteristic curves of the ITO-PD with periodic turning on/off at 254 and 365 nm light. Even after multiple periods of turning the light on/off, the solar-blind PD still exhibits excellent photoresponse characteristics, indicating that the PD possesses outstanding repeatability and stability. Furthermore, the PD has almost no photo switching effect under 365 nm light, indicating the strong solar-blind ultraviolet characteristics and spectral selectivity of the device.^{[ ]} As shown in Figure S1, (Supporting Information), the I–t characteristic curve of the other two PDs periodically turned on/off 254 and 365 nm illumination at 10 V. It is obvious that they also show excellent solar-blind characteristic and stability

From the aforementioned, it is clear that the PD performance is basically related to photocurrent and response time. To explore the internal reasons for the small photocurrent and slow photoresponse time of the AZO-PD in the three PDs under the same conditions, the basic parameters of the three electrode materials were investigated (Table S1, Supporting Information). It is evident that the resistivity of the AZO electrode material reaches as high as 7.54 Ω cm, while the resistivity of the ITO and IGZO materials is only 0.034 and 0.068 Ω cm, respectively. In addition, the AZO electrode material carrier mobility is relatively low, at only 0.089 cm² V s⁻¹, while the ITO and IGZO materials carrier mobilities are 2.9 and 8.6 cm² V s⁻¹, respectively. Materials with low resistivity and high carrier mobility should be selected as the electrode material for collecting and transporting carriers. However, due to the AZO electrode material resistivity and carrier mobility being poor in this study, among the three PDs, the photocurrent of the AZO-PD is relatively small. Generally, the photoresponse time of a PD is not only dependent on the rapid generation and annihilation of photogenerated carriers, but is also related to the trapping/releasing of photogenerated carriers by defect states.^{[ ]} It can be seen from Figure  that the recombination loss of the AZO-PD is the largest of the PDs, which relates to a large number of defect states between the electrode and the ε-Ga₂O₃ film. These defect states may act as efficient photo-generated carriers trapping/releasing states, resulting in a slow PD photoresponse time.

To demonstrate the ultra-high performance of the three fully transparent solar-blind PDs, Table  shows the key parameters of MSM-type Ga₂O₃ PDs both in previous reports, and in this research. As shown in the table, all the PDs we prepared exhibit excellent performance, including a record ultra-high responsivity of more than 250 A W⁻¹ and ultra-fast millisecond level photoresponse time in MSM-type ε-Ga₂O₃ PDs. Furthermore, in contrast to previously reported PDs with noble metal electrodes, all three PDs in this work were prepared from low-cost transparent conductive oxides. Overall, the proposed fully transparent MSM-type PDs possess significant advantages of excellent performance, easy integration, and low cost, making them a promising option for next-generation fully transparent optoelectronic devices.

Table 1 Comparison of key parameters of MSM-type Ga₂O₃ solar-blind PDs

Materials	Array	electrode material	R [A W−1]	D* [Jones]	τ r/τ d [ms]	References
a-Ga2O3	No	Cr	1430@12 V	3.16 × 1015	>1000	[]
a-Ga2O3	Yes	Ti/Au	733@5 V	3.9 × 1016	1/18	[]
a-Ga2O3	No	Cr/Au	96.1@5 V	/	31/78	[]
a-Ga2O3	Yes	Au	8.9@12 V	3.3 × 1013	0.015/0.308	[]
α-Ga2O3	No	Ti/Au	11.5@12 V	1.0 × 1015	/42	[]
α-Ga2O3	No	Ni/Au	3.36@12 V	/	/	[]
β-Ga2O3	No	Au	9500@5 V	1.5 × 1015	10/10	[]
β-Ga2O3	No	Ni/Au	1.5@20 V	/	3330/400	[]
ε-Ga2O3	No	Ti/Au	1.368@5 V	9.13 × 1014	61.1/87.5	[]
ε-Ga2O3	No	Ti/Au	230@6 V	1.2 × 1015	/24	[]
ε-Ga2O3	Yes	ITO	286.2@10 V	4.73 × 1014	5.6/7.2	This work
ε-Ga2O3	Yes	AZO	262.1@10 V	2.16 × 1015	70/200	This work
ε-Ga2O3	Yes	IGZO	284.1@10 V	5.06 × 1014	6.9/9.5	This work

Subsequently, we explored the potential applications of large-area fully transparent ε-Ga₂O₃ film-based photodetector arrays in solar-blind imaging, as shown in the Figure  schematic diagram. The hollow mask engraved with “TCO” characters was irradiated under 254 nm light with a 400 μW cm⁻² light intensity, and three fully transparent PD arrays were placed under the mask. The ITO-PD, AZO-PD, and IGZO-PD arrays were shielded by three masks, namely “T”, “C”, and “O” characters, respectively. When solar-blind deep-ultraviolet light passes through the mask onto the PD array, the remaining PDs are kept in the dark or under weak UV light. Keeping the test system still, the current of each PD unit was recorded cell by cell through a pair of probes connected to the semiconductor parameter analyzer. The array uniformity is a basic requirement of integrated applications, and the uniformity of the three arrays is further discussed before measuring the solar-blind imaging. Figure  describes the dark current of each PD in an ITO-PD array. Obviously, the dark current of most PDs is about 10⁻⁹ A in magnitude, but the dark current at the corner of the three PD units is relatively large, which may be due to irreversible damage to the ε-Ga₂O₃ film during cutting in the later stages. Moreover, the dark current of PDs units in the AZO-PD and IGZO-PD arrays were also certified in dark conditions, as shown in Figure . Two fully transparent arrays exhibit excellent uniformity, and dark currents of ≈10⁻¹⁰ and 10⁻⁹ A magnitude, respectively. Meanwhile, the aforementioned method was used to conduct solar-blind imaging tests. Figure  shows a 2D dark current comparison diagram. Surprisingly, all arrays display clearer “T”, “C”, and “O” character shapes, indicating that these fully transparent solar-blind PD arrays have high-fidelity characteristics and have excellent application prospects in solar-blind imaging and machine vision.^{[ ]}

[IMAGE OMITTED. SEE PDF]

In recent years, with the introduction of a new generation of 5G broadband mobile communication technology, wireless spectrum resources have become increasingly scarce. Solar-blind ultraviolet communication (SBUC) not only demonstrates electromagnetic interference resistance and ultra-high signal-to-noise ratio, but also massively expands the communication spectrum. Due to the fast photoresponse speed and ultra-high responsivity of the three PDs, fully transparent ε-Ga₂O₃-based solar-blind PDs were prepared as optical signal receivers of an SBUC system. Figure  exhibits a flowchart of the solar-blind ultraviolet communication principle. First, the text data is converted into control commands by the driver (EP4CE6F17C8) to control turning the metal oxide semicoductor field effect transistor (MOSFET) (LR7843) on/off, corresponding to the on/off of the light emitting diode (LED) light source. Subsequently, the intermittent light (255.5 nm) from the LED is detected by the solar-blind photodetector and outputs high/low-level current signals. The photocurrent signal is then converted into a voltage signal and further amplified by the differential voltage preamplifier. The differential voltage signal input to the driver (analog-to-digital converter chip) converts it to a digital signal.

[IMAGE OMITTED. SEE PDF]

To evaluate the data transmission ability and stability of the SBUC system, three PD optical signal receivers were used to try to transmit the text data of three letters “TCO”. First, “TCO” is converted into ASCII by the program. The code only contains the digits “1” and “0”. The driver (EP4CE6F17C8) then converts the “1” and “0” into the control command corresponding to the MOSFET (LR7843) on/off function. To be able to see the digital signal before conversion, the corresponding real-time images displayed on the oscilloscope, are outputted under different modulation frequencies. Figure  shows the waveform signals of the three PDs SBUC systems at a transmission speed of 0.01 kbps. All the waveforms showed the same response as the digital “TCO” signal, confirming the suitability of all three devices as SBUC system receiving signals. However, at the same transmission frequency, the SDUC system based on the ITO-PD transmission waveform is the most obvious and orderly, indicating that this SDUC system transmission limit is the highest along with its late decoding accuracy. Comparing the relationship between the performance of three PDs and the SDUC system transmission, it is clear that the transmission speed of the SDUC system is directly related to the PD photoresponse speed as the signal receiving component. Therefore, the development of a PD with a faster photoresponse speed would be an effective way to increase the SDUC transmission rate. Alternatively, further optimizing the differential voltage preamplifier could also further increase the transmission rate.^{[ ]}

Conclusion

In summary, we demonstrate for the first time that transparent conductive oxides are potential excellent electrode materials to obtain high-performance ε-Ga₂O₃-based solar-blind deep-ultraviolet photodetectors. A proof-of-concept of the solar-blind imaging and deep-ultraviolet communication system was verified using three fully transparent ε-Ga₂O₃-based arrays, composed of easily-integrated MSM-type photodetectors. The results show that all three PDAs exhibit not only ultra-high responsivity, remarkable detectivity, and excellent stability, but also an ultra-fast photoresponse time. Finally, for the first time, we explored the potential applicability of three ε-Ga₂O₃-based PDAs in solar-blind imaging and deep-ultraviolet communication. Our research presents a method to fabricate high-performance, fully transparent, solar-blind, deep-ultraviolet PDAs and highlights their applicability in the next generation of fully transparent optoelectronics.

Experimental Section

Material Processing and Device Fabrication

The commercially available ε-Ga₂O₃ thin films from Beijing Gallium Group were selected, which were deposited on sapphire substrates by metal-organic chemical vapor deposition (MOCVD). Subsequently, ε-Ga₂O₃ films were annealed in a tube furnace at 500 °C for 2 h in argon. Three transparent electrode materials, including ITO, IGZO, and AZO, were then sputtered on the surface of ε-Ga₂O₃ films by magnetron sputtering and a masking process, respectively. High-uniformity 5 × 4 arrays of fully transparent MSM-type SBDU PDs were constructed, in which a single PD consisted of 5 pairs of interdigital electrodes with a length of 800 μm and a spacing of 30 μm. During the sputtering process, the sputtering chamber base vacuum pressure and argon flow rate were 5.0 × 10⁻⁴ Pa and 40 sccm, respectively. The working pressure, sputtering power, and sputtering time were 2.5/0.8/2.5 Pa, 120/50/120 W, and 10/15/10 min, respectively.

Material and Device Characterization

The crystal structures of ε-Ga₂O₃ films were analyzed by an X-ray diffractometer (Bruker D8 ADVANCE A25X) using Cu Kα lines (λ = 0.154 nm). The ε-Ga₂O₃ films absorption spectra were tested using a UV–vis–NIR spectrophotometer (U-4100). The electrical properties of the three transparent electrode materials were obtained by a Hall effect measurement system (HMS-3000, Ecopia). The surface morphology and elemental composition were obtained by field emission scanning electron microscopy (Thermofisher Apreo 2C). The photoelectric characteristics of ε-Ga₂O₃-based PDs, and the corresponding solar-blind imaging analysis, were measured by a Keithley 2450 source meter. A MOS5072 oscilloscope was applied to obtain the deep-ultraviolet communication system voltage signals. All characterization was conducted at room temperature.

Acknowledgements

The authors gratefully acknowledge support from the National Natural Science Foundation of China (Grant Nos. 11904041, 12104077), the Natural Science Foundation of Chongqing (Grant No. cstc2019jcyj-msxmX0237, cstc2020jcyj-msxmX0533, cstc2020jcyj-msxmX0557), the Science and Technology Research Project of Chongqing Education Committee (Grant Nos. KJQN201900542, KJQN202000511, KJQN202100501, KJQN202100540), and the College Students Innovation and Entrepreneurship Training Program of Chongqing City (Grant Nos. S202110637123, S202110637119, S202110637130, S202210637046, 2020051101165).

Conflict of Interest

The authors declare no conflict of interest.

Data Availability Statement

Research data are not shared.

References

Z. Y. Han, H. L. Liang, W. X. Huo, X. S. Zhu, X. L. Du, Z. X. Mei, Adv. Opt. Mater. 2020, 8, 1901833.

W. Y. Kong, G. A. Wu, K. Y. Wang, T. F. Zhang, Y. F. Zou, D. D. Wang, L. B. Luo, Adv. Mater. 2016, 28, 10725.

J. H. Kim, C. Y. Han, K. H. Lee, K. S. An, W. Song, J. Kim, M. S. Oh, Y. R. Do, H. Yang, Chem. Mater. 2014, 27, 197.

J. X. Chen, W. X. Ouyang, W. Yang, J. H. He, X. S. Fang, Adv. Opt. Mater. 2020, 30, 1909909.

X. Chao, F. Yan, Small 2017, 13, 1701822.

X. H. Chen, F. F. Ren, S. L. Gu, J. D. Ye, Photonics Res. 2019, 7, 381.

J. J. Xu, W. Zheng, F. Huang, J. Mater. Chem. C 2019, 7, 8753.

D. H. Wang, C. Huang, X. Liu, H. C. Zhang, H. B. Yu, S. Fang, B. S. Ooi, Z. Mi, J. H. He, H. Sun, Adv. Opt. Mater. 2021, 9, 2000893.

M. M. Fan, K. W. Liu, Z. Z. Zhang, B. H. Li, X. Chen, D. X. Zhao, C. X. Shan, D. Z. Shen, Appl. Phys. Lett. 2014, 105, 011117.

S. P. Karanth, M. A. Sumesh, V. Shobha, J. Sirisha, D. Yadav, S. B. Vijay, K. V. Sriram, IEEE Trans. Electron Devices 2020, 67, 242.

Y. Xu, X. H. Chen, Y. F. Zhang, F. F. Ren, S. L. Gu, J. D. Ye, IEEE Electron Device Lett. 2020, 41, 997.

S. J. Pearton, J. C. Yang, P. Cary, F. Ren, J. Kim, M. J. Tadjer, M. A. Mastro, Appl. Phys. Rev. 2018, 5, 301.

S. Oh, C. K. Kim, J. Kim, ACS Photonics 2018, 5, 1123.

C. N. Lin, Y. J. Lu, X. Yang, Y. Z. Tian, C. J. Gao, J. L. Sun, L. Dong, F. Zhong, W. D. Hu, C. X. Shan, Adv. Opt. Mater. 2018, 6, 1800068.

H. B. Wang, H. Y. Chen, L. Li, Y. F. Wang, L. X. Su, W. P. Bian, B. S. Li, X. S. Fang, J. Phys. Chem. Lett. 2019, 10, 6850.

Y. Q. Zhang, Y. F. Wang, R. P. Fu, J. G. Ma, H. Y. Xu, B. S. Li, Y. C. Liu, J. Phys. D: Appl. Phys. 2022, 55, 324002.

R. C. Lin, W. Zheng, D. Zhang, Z. J. Zhang, Q. X. Liao, L. Yang, F. Huang, ACS Appl. Mater. Interfaces 2018, 10, 22419.

T. Zhang, Y. X. Shen, Q. Feng, X. S. Tian, Y. C. Cai, Z. Z. Hu, G. S. Yan, Z. Q. Feng, Y. C. Zhang, J. Ning, Y. Xu, X. Lian, X. Sun, C. Zhang, H. Zhou, J. Zhang, Y. Hao, Nanoscale Res. Lett. 2020, 15, 1.

X. H. Chen, Y. Xu, D. Zhou, S. Yang, F. F. Ren, H. Lu, K. Tang, S. L. Gu, R. Zhang, Y. D. Zheng, J. D. Ye, ACS Appl. Mater. Interfaces 2017, 9, 36997.

Z. X. Jiang, Z. Y. Wu, C. C. Ma, J. N. Deng, H. Zhang, Y. Xu, J. D. Ye, Z. L. Fang, G. Q. Zhang, J. Y. Kang, T. Y. Zhang, Mater. Today Phys. 2020, 14, 100226.

J. Wang, L. J. Ye, X. Wang, H. Zhang, L. Li, C. Y. Kong, W. J. Li, J. Alloys Compd. 2019, 803, 9.

L. J. Huang, Z. R. Hu, H. Zhang, Y. Q. Xiong, S. Q. Fan, C. Y. Kong, W. J. Li, L. J. Ye, H. L. Li, J. Mater. Chem. C 2021, 9, 10354.

C. Xie, X. T. Lu, Y. Liang, H. H. Chen, L. Wang, C. Y. Wu, D. Wu, W. H. Yang, L. B. Luo, J. Mater. Sci. Technol. 2021, 72, 189.

B. Zhao, F. Wang, H. Y. Chen, Y. P. Wang, M. M. Jiang, X. S. Fang, D. X. Zhao, Nano Lett. 2015, 15, 3988.

D. Kaur, M. Kumar, Adv. Opt. Mater. 2021, 9, 2002160.

H. W. Liu, S. R. Zhou, H. Zhang, L. J. Ye, Y. Q. Xiong, P. Yu, W. J. Li, X. Yang, H. L. Li, C. Y. Kong, J. Phys. D: Appl. Phys. 2022, 55, 305104.

Y. C. Chen, Y. J. Lu, Q. Liu, C. N. Lin, J. Guo, J. H. Zang, Y. Z. Tian, C. X. Shan, J. Mater. Chem. C 2019, 7, 2557.

Y. Qin, L. H. Li, Z. A. Yu, F. H. Wu, D. N. Dong, W. Guo, Z. F. Zhang, J. H. Yuan, K. H. Xue, X. S. Miao, S. B. Long, Adv. Sci. 2021, 8, 101106.

C. N. Lin, Y. J. Lu, Y. Z. Tian, C. J. Gao, M. M. Fan, X. Yang, L. Dong, C. X. Shan, Opt Express 2019, 27, 29962.

S. R. Zhou, X. Peng, H. W. Liu, Z. F. Zhang, L. J. Ye, H. L. Wen, Y. Q. Xiong, L. B. Niu, F. L. Chen, L. Fang, C. Y. Kong, W. J. Li, X. Yang, H. Zhang, Opt. Mater. Express 2022, 12, 327.

S. Li, J. Y. Yue, X. Q. Ji, C. Lu, Z. Y. Yan, P. G. Li, D. Y. Guo, Z. P. Wu, W. H. Tang, J. Mater. Chem. C 2021, 9, 5437.

S. Ju, A. Facchetti, Y. Xuan, J. Liu, F. Ishikawa, P. D. Ye, C. W. Zhou, T. J. Marks, D. B. Janes, Nat. Nanotechnol. 2007, 2, 378.

H. Kim, K. R. Chauhan, J. Kim, E. H. Choi, Appl. Phys. Lett. 2017, 110, 101907.

J. Vidal, F. Trani, F. Bruneval, M. Marques, S. Botti, Phys. Rev. Lett. 2010, 104, 136401.

J. L. Wu, L. H. Lai, Y. T. Hsiao, K. W. Tsai, C. M. Yang, Z. W. Sun, J. C. Hsieh, Y. M. Chang, Adv. Opt. Mater. 2021, 10, 202101723.

Y. C. Oshima, E. Víllora, Y. Matsushita, S. Yamamoto, K. Shimamura, J. Appl. Phys. 2015, 118, 085301.

T. Wang, H. L. Liang, Z. Y. Han, Y. X. Sui, Z. X. Mei, Adv. Mater. Technol. 2021, 6, 2000945.

H. Kan, W. Zheng, H. Chen, R. C. Lin, J. T. Luo, F. Huang, ACS Appl. Mater. Interfaces 2020, 12, 6030.

D. Y. Guo, Y. L. Su, H. Z. Shi, P. G. Li, N. Zhao, J. H. Ye, S. L. Wang, A. P. Liu, Z. W. Chen, C. R. Li, W. H. Tang, ACS Nano 2018, 12, 12827.

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

D. Y. Guo, Z. P. Wu, Y. H. An, X. C. Guo, X. L. Chu, C. L. Sun, L. H. Li, P. G. Li, W. H. Tang, Appl. Phys. Lett. 2014, 105, 023507.

C. Xie, X. T. Lu, M. R. Ma, X. W. Tong, Z. X. Zhang, L. Wang, C. Y. Wu, W. H. Yang, L. B. Luo, Adv. Opt. Mater. 2019, 7, 1901257.

Z. X. Zhang, L. H. Zeng, X. W. Tong, Y. Gao, C. Xie, Y. H. Tsang, L. B. Luo, Y. C. Wu, J. Phys. Chem. Lett. 2018, 9, 1185.

L. B. Luo, G. A. Wu, Y. Gao, L. Liang, C. Xie, Z. X. Zhang, X. W. Tong, T. Wang, F. X. Liang, Adv. Opt. Mater. 2019, 7, 1900272.

K. Arora, N. Goel, M. Kumar, M. Kumar, ACS Photonics 2018, 5, 2391.

X. H. Hou, H. D. Sun, S. B. Long, G. Tompa, T. Salagaj, Y. Qin, Z. F. Zhang, P. Tan, S. J. Yu, M. Liu, IEEE Electron Device Lett. 2019, 40, 1483.

S. Li, D. Y. Guo, P. G. Li, X. Wang, Y. H. Wang, Z. Y. Yan, Z. Liu, Y. S. Zhi, Y. Q. Huang, Z. P. Wu, W. H. Tang, ACS Appl. Mater. Interfaces 2019, 11, 35105.

Y. C. Chen, Y. J. Lu, M. Y. Liao, Y. Z. Tian, Q. Liu, C. J. Gao, X. Yang, C. X. Shan, Adv. Funct. Mater. 2019, 29, 1906040.

G. Qiao, Q. Cai, T. C. Ma, J. Wang, X. H. Chen, Y. Xu, Z. G. Shao, J. D. Ye, D. J. Chen, ACS Appl. Mater. Interfaces 2019, 11, 40283.

A. S. Pratiyush, S. Krishnamoorthy, S. V. Solanke, Z. B. Xia, R. Muralidharan, S. Rajan, D. Nath, Appl. Phys. Lett. 2017, 110, 221107.

Y. Qin, L. H. Li, X. L. Zhao, G. Tompa, H. Dong, G. Z. Jian, Q. M. He, P. J. Tan, X. H. Hou, Z. F. Zhang, S. J. Yu, H. D. Sun, G. W. Xu, X. S. Miao, K. H. Xue, S. B. Long, M. Liu, ACS Photonics 2020, 7, 812.

G. B. Cen, Y. J. Liu, C. X. Zhao, G. H. Wang, Y. Fu, G. H. Yan, Y. Yuan, C. H. Su, Z. J. Zhao, W. J. Mai, Small 2019, 15, 1902135.