Research on the Ga₂O₃/ZnGa₂O₄ mixed-phase films and solar-blind photodetectors prepared directly by annealing Zn alloying Ga₂O₃ films

Abstract

High-Zn-content doped β-Ga₂O₃ thin films were grown on c-plane sapphire substrates using metal–organic chemical vapor deposition (MOCVD), with subsequent annealing inducing a phase transition, converting the initially single-phase material into a Ga₂O₃/ZnGa₂O₄ mixed-phase structure. Furthermore, metal–semiconductor–metal structured solar-blind photodetectors were fabricated to evaluate the ultraviolet detection capabilities of these mixed-phase films. The detector performance was significantly enhanced following annealing. At an annealing temperature of 800 °C, phase separation occurred, and the improvement in device performance became more pronounced. Devices annealed in an argon atmosphere, undergoing phase separation, exhibited superior optoelectronic performance, achieving a low dark current of 0.137 pA at a bias voltage of 10 V. At the peak wavelength, the device also demonstrated a responsivity of 2058.54 A W⁻¹, a specific detectivity of 3.21 × 10¹⁵ cm Hz^1/2 W⁻¹, and a short decay time of 2.90 ms. The exceptional performance of mixed-phase devices is likely attributed to the large number of heterojunction interfaces between Ga₂O₃ and ZnGa₂O₄. This study investigates the transition of the thin film material from a single-phase to a mixed-phase structure, analyzing the associated changes in their physical properties and detector performance. Additionally, it expands the material design framework for Ga₂O₃-based ultraviolet photodetectors and introduces a novel strategy to enhance their performance.

---

Introduction

Ultraviolet detection technology holds significant market potential and covers diverse consumer electronics and industry applications, such as automotive, pharmaceutical, medical diagnostics, environmental monitoring and industrial process control, among others.¹ Ultraviolet light in the range of 200–280 nm in the solar spectrum is strongly absorbed by ozone in the atmosphere and cannot reach the surface of the Earth, so this band of ultraviolet is called the solar-blind ultraviolet band. As a result, detectors operating in these specific wavelengths exhibit lower noise levels and higher sensitivity.^2–4 Wide-bandgap semiconductor materials offer distinct advantages, including an intrinsic bandgap ideally suited for ultraviolet detection and excellent stability. These characteristics have made them the focus of growing interest and research in this field.^5–7 Typically, detectors fabricated from single-phase materials are more commonly used, as they generally offer superior performance compared to their mixed-phase counterparts.^8–10 However, a study conducted by our group in 2014 demonstrated that ultraviolet detectors made from mixed-phase MgZnO thin films could achieve performance comparable to that of single-phase devices.¹¹ Building on this, Han et al. conducted additional research, validating the ultraviolet response characteristics of mixed-phase MgZnO thin films.^12,13 In recent years, Ga₂O₃ has garnered significant attention due to its exceptional physical properties,¹⁴ leading to a growing body of research on solar-blind ultraviolet detectors.^15–21 Concurrently, studies on Ga₂O₃-based mixed-phase structural films have emerged as a prominent area of investigation. Researchers have successfully obtained mixed-phase Ga₂O₃ films by optimizing the growth conditions.^22,23 Liu et al. improved the sensitivity and signal-to-noise ratio of devices by utilizing amorphous/β-Ga₂O₃ films.²⁴ Alghareeb et al. grew mixed γ/β-Ga₂O₃ films on n-type Si, resulting in detectors with enhanced overall performance.²⁵ Furthermore, the incorporation of excess doping elements during the Ga₂O₃ growth process, beyond their solubility limits, can induce the formation of mixed-phase structures.^26,27 Fan et al. prepared a mixed-phase thin film by heating a mixture of Ga₂O₃ and SnO₂ powders in a tube furnace.²⁸ Qi et al. controlled the precursor ratio of In and Ga to grow (In_xGa_1−x)₂O₃ thin films, which exhibited coexistence of monoclinic and cubic phases.²⁹ Current research predominantly focuses on the direct synthesis of mixed-phase materials. In contrast, studies investigating the phase transition from single-phase to mixed-phase materials, as well as comparative analyses of the properties and device performance of single-phase versus mixed-phase materials, remain relatively scarce.

In this work, we employed metal organic chemical vapor deposition (MOCVD) to grow a single-phase Zn-alloyed Ga₂O₃ thin film with an approximate Zn composition of 15 at% on a sapphire (c-plane) substrate. Subsequently, the film was annealed to induce a phase transition, resulting in the formation of a mixed-phase structure. The Ga₂O₃ thin films were annealed in oxygen and argon atmospheres to investigate their structural and physical properties. Based on these films, a metal–semiconductor–metal (MSM) detector was fabricated to evaluate the impact of the Ga₂O₃/ZnGa₂O₄ mixed-phase structure on ultraviolet detection performance. The detector annealed at 800 °C in an argon atmosphere demonstrated superior detection performance. Under a 10 V bias, the dark current was below 1 pA, with a decay time of 2.90 ms. The responsivity at the peak wavelength was 2058.54 A W⁻¹. The substantial enhancement in device performance may be attributed to the formation of heterojunction interfaces between the mixed-phases within the thin films. Our findings in this work show promising prospects for Ga₂O₃/ZnGa₂O₄ mixed-phase photodetectors in solar-blind photodetection applications.

Experimental

Ga₂O₃ thin films alloyed with approximately 15 at% Zn were epitaxially grown on c-Al₂O₃ substrates via MOCVD. High-purity nitrogen was used as the carrier gas for diethylzinc (DEZn) and triethylgallium (TEGa), with 99.999% oxygen as the reactant gas. The source temperatures of DEZn and TEGa were maintained at −8 °C and 5 °C, respectively. The flow rates of DEZn, TEGa, and O₂ were set to 1, 5, and 250 sccm, respectively. Epitaxial growth was conducted at 700 °C and 740 Pa for 90 minutes. The samples were then annealed in O₂ and Ar atmospheres at 600, 700, 800, and 850 °C for 60 minutes, with a constant gas flow rate of 20 sccm. The crystal structure and phase composition of the thin films were characterized using X-ray diffraction (XRD) (Bruker D8, Cu Kα radiation, λ = 0.154 nm). The Zn content in the films was estimated using energy dispersive X-ray spectroscopy (EDS). The optical properties and surface morphology were examined with a UV-3101PC spectrophotometer, an atomic force microscope (AFM) (Bruker Multimode 8), and a scanning electron microscope (SEM) (Hitachi S-4800). Additionally, the elemental composition of the films was analyzed using X-ray photoelectron spectroscopy (XPS) (Thermo Scientific Escalab 250Xi), with charge calibration performed relative to the C 1s binding energy of 284.8 eV.

Au interdigital electrodes were fabricated on the thin film using photolithography to design the MSM structure. The current–voltage (I–V) and photocurrent (I–t) characteristics were measured using an Agilent B1500A semiconductor analyzer. The spectral response was evaluated with a 200 W UV-enhanced xenon lamp, a lock-in amplifier, and a monochromator for wavelength-dependent measurements. The response time of devices was characterized using a 244 nm wavelength light from an OPO (optical parametric oscillator) laser system. The laser was operated at a repetition rate of 10 Hz, with each pulse duration of 5 ns.

Results and discussion

The XRD patterns of the as-grown and annealed Zn-alloyed Ga₂O₃ thin films are shown in Fig. 1, and Fig. 1a shows a spectrum in the range of 2θ = 10–70°. For the as-grown Zn-alloyed Ga₂O₃ film, the sharp and strong peak at 2θ = 41.68° can be attributed to the (0006) diffraction peak of the c-plane sapphire substrate. The diffraction peaks at 2θ = ∼19.0°, ∼38.4°, and ∼59.2° correspond to the (−201), (−402), and (−603) diffraction peaks of the monoclinic structure of β-Ga₂O₃ (JCPDS no. 043-1012), respectively. To further confirm the crystalline phase of the as-grown thin film, we performed Raman spectroscopy analysis of Ga₂O₃, as shown in Fig. S1 (ESI†). Raman spectroscopy was performed on the as-grown film, and after accounting for the influence of the c-Al₂O₃ substrate, characteristic peaks of β-Ga₂O₃ were observed at 203.6, 351.0, and 661.5 cm⁻¹, which are consistent with previously reported values for the β-phase.^30–32 When the annealing temperature is between 600 °C and 700 °C, there is no significant change in the XRD diffraction peak of the obtained film, indicating that the crystal structure of the film has not changed at the annealing temperature of 600–700 °C. However, at annealing temperatures ranging from 800 °C to 850 °C, significant alterations in the diffraction peaks are observed, suggesting a marked transformation in the crystalline structure due to the higher annealing temperatures.

Fig. 1 (a) XRD patterns of the as-grown and annealed films under different conditions. (b) Enlarged sections of the XRD patterns at 2θ = ∼19.0°, ∼38.4°, and ∼59.2°.

To provide a clearer representation of the changes in the diffraction peaks, the ZnGa₂O₄ and β-Ga₂O₃-related peaks in the XRD spectra were enlarged, as shown in Fig. 1b. At an annealing temperature of 800 °C, the diffraction peak near 2θ = ∼19.0° exhibits significant broadening in the full width at half maximum, while the peaks at 2θ = ∼38.4° and ∼59.2° transition from single to double peaks. When the annealing temperature is further increased to 850 °C, the peak near 2θ = ∼19.0° also splits into two distinct peaks. This pattern is consistent regardless of whether the annealing process is conducted in an Ar or O₂ atmosphere. Comparing the twin peaks appearing in pairs around 2θ = ∼19.0°, ∼38.4°, and ∼59.2° with the JCPDS standard card, it can be found that the peaks with smaller angles are the (111), (222), and (333) peaks of cubic ZnGa₂O₄ (JCPDS no. 38-1240), and the peaks with larger angles are the (−201), (−402), and (−603) diffraction peaks of monoclinic β-Ga₂O₃. These findings suggest that annealing at 800 °C or higher facilitates the transformation of single-phase Zn-alloyed Ga₂O₃ thin films into mixed-phase thin films, consisting of both ZnGa₂O₄ and β-Ga₂O₃ phases. The diffraction peak of the monoclinic phase in the as-grown thin film shifts to a lower angle compared to the β-Ga₂O₃ peak observed after annealing. This shift in peak positions is likely due to lattice distortion induced by the incorporation of Zn atoms (with ionic radii of 0.74 Å for Zn²⁺ and 0.61 Å for Ga³⁺).³³ Moreover, the formation of a mixed-phase structure may be attributed to the Zn doping concentration in the as-grown film exceeding the threshold for single-phase Zn-doped Ga₂O₃, as reported in the literature, thus placing the material in a metastable state.^34,35 High-temperature annealing promotes the redistribution of dopant atoms and the relaxation of internal stresses within the film,^36,37 which together contribute to the transition from a single-phase to a mixed-phase structure.

Fig. 2a presents the UV-Vis transmission spectra of the different samples. Across the wavelength range of 300 to 800 nm, all samples exhibited an average transmission exceeding 85%. Moreover, a distinct optical absorption edge was consistently observed within the range of 240–250 nm. The absorption coefficient (α) of the thin film was calculated from the transmission spectrum data. From the obtained α values, (αhv)² and hν were subsequently derived. (αhv)² was then plotted versus hν, with (αhv)² on the vertical axis and hν on the horizontal axis. The linear region of the resulting curve was fitted to a straight line, and the intercept of this line with the abscissa axis provided the band gap (E_g) of the film.³⁸ The calculated band gap of the thin film was approximately 5.09–5.11 eV, as indicated in the inset of Fig. 2a. This indicates that both high-temperature annealing and material phase separation have minimal influence on the bandgap.

Fig. 2 (a) Transmission spectra of the as-grown and annealed films. The inset shows the corresponding plot of (αhv)²versus photon energy hν for the different processing conditions. (b) Root-mean-square (RMS) roughness of the films under various treatment conditions, as characterized by AFM.

The root-mean-square (RMS) roughness of both the as-grown film and films annealed under various conditions was measured using AFM, with the results shown in Fig. 2b. Detailed 3D surface morphology is provided in Fig. S2 and S3 in the ESI.† For annealing temperatures between 600 °C and 700 °C, the RMS roughness values of the films in both atmospheres exhibited a decreasing trend as the annealing temperature increased. These findings suggest that annealing effectively improves the crystalline quality of the films before phase separation occurs. However, at higher annealing temperatures (800–850 °C), the values gradually increase, which may be attributed to the formation of new grain boundaries caused by pronounced phase separation. Despite these variations in roughness, the changes remained within a narrow range of 1.25 nm to 2.5 nm, and the surfaces of the films remained relatively smooth and uniform.

Fig. 3 illustrates the surface SEM images of Zn-alloyed Ga₂O₃ thin films annealed under different atmospheres and temperatures, with cross-sectional SEM images provided as insets. For comparison, the SEM images of the as-grown films are presented in Fig. S4a and b in the ESI.† As the annealing temperature increases, the surfaces of the films under both atmospheres remain dense, without the formation of large grains. Cross-sectional SEM images reveal a sharp interface between the substrate and the epitaxial films, with the film thickness consistently maintained at approximately 110 nm. Notably, even in the presence of phase separation, no significant changes in surface morphology or the film thickness are observed. EDS analysis (see Fig. S5 and S6 in the ESI†) further confirms that the atomic ratio of Zn to Ga remains nearly constant before and after annealing, with the Zn content stabilized around 15.02%. These findings suggest that, although phase separation occurs during high-temperature annealing at 800–850 °C, no detectable loss of Zn or Ga elements from the films is observed.

Fig. 3 Surface SEM images of the films annealed under different conditions, with the corresponding cross-sectional SEM images shown in the insets. (a)–(d) Films annealed at various temperatures in an argon atmosphere. (e)–(h) Films annealed at different temperatures in an oxygen atmosphere.

Oxygen vacancies are a key factor affecting the performance of oxide semiconductor ultraviolet detectors.^39–41 To better understand their distribution within the thin films, we performed a detailed investigation using XPS. Fig. 4a presents the high-resolution O 1s core-level spectrum of the as-grown Zn-alloyed Ga₂O₃ thin film. The peak at approximately 530.68 eV (O_A) is attributed to lattice oxygen, while the peak at around 531.52 eV (O_B) is associated with oxygen vacancies.^42,43 The oxygen vacancy concentration in the as-deposited film was determined to be 24.2% based on the peak area ratio (O_B/(O_A + O_B)). Fig. 4b presents the high-resolution O 1s core-level spectra of films annealed at various temperatures under an Ar atmosphere. The oxygen vacancy concentrations showed a slight increase with annealing temperature, reaching 25.4%, 27.5%, 27.5%, and 28.1% at 600 °C, 700 °C, 800 °C, and 850 °C, respectively. In contrast, Fig. 4c shows the high-resolution O 1s core-level spectra of films annealed under an O₂ atmosphere. For these films, the oxygen vacancy concentrations exhibited a decrease as the annealing temperature increased, from 23.1% at 600 °C to 21.9%, 19.4%, and 16.7% at 700 °C, 800 °C, and 850 °C, respectively. Fig. 4d illustrates the evolution of the oxygen vacancy concentration under varying annealing conditions. With increasing annealing temperature, an Ar atmosphere promotes oxygen vacancy formation, while an O₂ atmosphere inhibits their formation. This divergence can be attributed to distinct annealing mechanisms: in oxygen-deficient environments, oxygen escapes from the film, creating vacancies, whereas in oxygen-rich environments, atmospheric oxygen interacts with the lattice and fills surface oxygen vacancies.⁴¹ These results demonstrate that while annealing in both Ar and O₂ atmospheres induces similar phase separation in the film, their effects on the oxygen vacancy concentration are entirely opposite. The fine Zn 2p spectrum exhibits a binding energy difference of approximately 23 eV between Zn 2p_1/2 and Zn 2p_3/2, while the fine Ga 2p spectrum shows a binding energy difference of approximately 26.9 eV between Ga 2p_1/2 and Ga 2p_3/2 (See Fig. S7, ESI†), both of which are consistent with reference values.⁴⁴ The binding energy difference between the Zn 2p_3/2 and Ga 2p_3/2 peaks serves as a parameter for differentiating ZnGa₂O₄ from a mixture of ZnO and Ga₂O₃.^45,46 The calculated binding energy difference of 95.9 eV closely aligns with the expected value for ZnGa₂O₄ formation.

Fig. 4 O 1s core-level XPS spectra of films treated under different conditions. (a) As-grown films. (b) Films annealed in an Ar atmosphere. (c) Films annealed in an O₂ atmosphere. The peak at O_B is attributed to oxygen vacancy defects, while the peak at O_A corresponds to lattice oxygen in the film. The ratio of peak areas (O_B/(O_A + O_B)) serves as an indicator of the oxygen vacancy concentration within the film. (d) Changes in the oxygen vacancy concentration under various annealing atmospheres and temperature conditions.

To investigate the photoelectric properties of two groups of thin films, Au interdigitated electrodes were fabricated using photolithography and lift-off techniques. The electrodes were designed with a length of 500 μm, a width of 10 μm, an inter-finger spacing of 10 μm, and 25 pairs of fingers. These electrodes were incorporated into an MSM-structured ultraviolet photodetector. The schematic of the device structure is shown in Fig. 5a, with the effective illuminated area measured as 0.5 mm². Fig. 5b and c display the I–V characteristics of the annealed device in a dark environment. Devices fabricated using as-grown thin films exhibited a dark current of 32 pA under an applied bias of 10 V. After annealing, the dark current was drastically reduced to below 1 pA, indicating a significant enhancement in device performance. The I–V characteristics of the annealed devices under ultraviolet illumination (wavelength: 222 nm, power density: 0.78 mW cm⁻²) are shown in Fig. 5e and f. The data indicate that the photocurrent increases steadily with annealing temperatures ranging from 700 °C to 800 °C, irrespective of whether the annealing atmosphere is oxygen or argon. At 850 °C, the photocurrent decreases slightly, however, it remains higher than the value observed at 700 °C. To provide a clearer depiction of this trend, Fig. 5d plots the photocurrent at a bias voltage of 10 V as a function of the annealing temperature, offering an intuitive visualization of the relationship between annealing conditions and device performance. Oxygen vacancies are well-established as critical factors influencing device responsivity.³⁹ However, in this study, analysis of oxygen vacancy concentrations measured by XPS, in conjunction with the performance of devices annealed under different atmospheres, reveals that variations in device performance with annealing temperature are not primarily attributable to changes in oxygen vacancy concentrations. Combined with the previous XRD characterization results, it can be deduced that phase separation occurs in the thin film at 800 °C, coinciding with a significant increase in photoresponsivity. This improvement can be attributed to the formation of mixed-phase heterojunction interfaces, which enhance the generation and transport of photogenerated carriers.^47,48 At 850 °C, pronounced phase separation likely results in the formation of larger energy barriers at the heterojunction interfaces. These barriers impede the efficient transport of photogenerated carriers, thereby leading to a decline in photoresponsivity.²⁴

Fig. 5 (a) Schematic diagram of the MSM detector structure. (b) I–V characteristics of the detector based on the films annealed in an Ar atmosphere (dark environment). (c) I–V characteristics of the detector based on the films annealed in an O₂ atmosphere (dark environment). (d) Photocurrent response trends for different devices. (e) I–V characteristics of the detector based on the films annealed in an Ar atmosphere (illuminated with UV light, λ = 222 nm). (f) I–V characteristics of the detector based on the films annealed in an O₂ atmosphere (illuminated with UV light, λ = 222 nm).

To further investigate the response speed of the devices, the transient photoresponse measurement was carried out using an OPO laser system with a wavelength of 244 nm. The laser pulse width and the repetition rate are 5 ns and 10 Hz, respectively. The test results are shown in Fig. 6. In Fig. 6a, a good reproducibility of photodetectors based on the as-grown film can be clearly observed. To calculate the response time of the device, one cycle is selected, and its magnified image is shown in inset. The 10–90% rise time (defined as the time for the current increasing from 10% to 90% of the peak value) of the device is 0.02 ms. In addition, the 90–10% decay time (defined as the time for the current dropping from 90% to 10% of the peak value) of the device is 2.24 ms. Fig. 6b and c show the transient response curves of devices prepared based on films annealed under different atmospheres and temperatures. It can be seen that the repeatability of the transient response current does not change with annealing conditions. The 10–90% rise time of the devices is very short, around 0.02 ms, while the 90–10% decay time of the devices changes to some extent with the annealing temperature. Fig. 6d shows the trend of the 90–10% decay time with annealing temperature variation. At the same annealing temperature, annealing under oxygen is more conducive to shortening the descent time. When the annealing temperature is between 600 and 700 °C, the decay time is shorter than that of devices based on the as-grown films, and when the annealing temperature reaches 850 °C, the decay time significantly increases. This may be due to the formation of a large number of carrier traps at heterogeneous interfaces within the mixed-phase material, resulting in a significant increase in carrier lifetime.

Fig. 6 Transient response curve of photodetectors under 244 nm pulsed laser irradiation. Insets are the enlarged images of a single pulse cycle. (a) Photodetectors based on the as-grown film. (b) Photodetectors based on films annealed in an Ar atmosphere. (c) Photodetectors based on films annealed in an O₂ atmosphere. (d) Trend of the 90–10% decay time with annealing temperature variations.

The responsivity spectra of the device were measured under a bias voltage of 10 V using a monochromator over a wavelength range of 230–500 nm, as shown in Fig. 7a and b. The responsivity of the detector increases with the annealing temperature and reaches its maximum value when the film transforms into a mixed-phase state and decreases as phase separation becomes more pronounced. When the annealing temperature is 800 °C, the device exhibits optimal photoelectric response performance. The peak responsivity is observed at approximately 244 nm, and the −3 dB cut-off wavelength is around 248 nm. These results indicate that the devices have good solar-blind ultraviolet photoresponse characteristics. When the annealing temperature is 800 °C, the responsivity of devices prepared based on films annealed under Ar and O₂ conditions is 2058.54 A W⁻¹ and 2364.01 A W⁻¹, respectively. In addition, the UV-visible rejection ratio, defined as the ratio between the peak responsivity and responsivity at 400 nm, is 3.39 × 10⁵ and 4.75 × 10⁵, respectively.

Fig. 7 Responsivity spectra of the devices based on (a) films annealed in an Ar atmosphere and (b) films annealed in an O₂ atmosphere.

The specific detectivity (D*), which serves as a measure of the signal-to-noise ratio, was determined using eqn (1)

where B denotes the measurement bandwidth (12.5 Hz), S_eff represents the effective illuminated area of the device (0.5 mm²), and NEP stands for the noise-equivalent power. To account for the low-frequency noise characteristics of the device, the detector was placed in a dark environment, and a long-term I–t curve was recorded. The dark current curve was analyzed using Fourier transform to determine the frequency–amplitude relationship, facilitating the extraction and evaluation of the corresponding noise spectral density curve. After fitting, the measured low-frequency noise of the device was found to be consistent with 1/f noise,⁴⁹ as shown in Fig. S8 (ESI†). The total noise current (i_noise) was obtained by integrating the noise spectral density curve, and the calculated i_noise values for devices annealed in two different atmospheres were found to be i_noise-Ar = 1.605 × 10⁻¹³ A and i_noise-O2 = 1.875 × 10⁻¹³ A. The relationship between the i_noise and NEP is expressed as: NEP = i_noise/R, where R represents the responsivity of the device.^50,51 Accordingly, eqn (1) can be rewritten as eqn (2)

The performance metrics of devices subjected to different annealing conditions are summarized in Table S1 (ESI†). A comparative analysis reveals that devices with mixed-phase films annealed at 800 °C exhibit superior overall performance. While an increase in the degree of phase mixing leads to a slight decline in device performance, it remains higher than that of devices with non-phase-separated thin films. These findings suggest that Zn alloying, followed by annealing to induce mixed-phase formation, is an effective strategy for optimizing the film structure and enhancing the optoelectronic performance of detectors. The performance metrics of devices annealed at 800 °C with mixed-phase structures, as summarized in Table 1, were compared to those of previously reported single-phase and mixed-phase devices. Notably, our devices exhibit significantly enhanced ultraviolet photodetection performance. This improvement is attributed to the formation of mixed-phase heterojunction structures induced by the annealing of β-Ga₂O₃ doped with high Zn content, which significantly enhances the performance of solar-blind ultraviolet photodetectors.

Table 1 Performance comparison of various Ga₂O₃-based ultraviolet detectors

Materials	Method	Bias (V)	Dark current (pA)	Responsivity (A W^−1)	Decay time (ms)	Detectivity (cm Hz^1/2 W^−1)	Ref.
Note:a M. E. means mechanical exfoliation.
β-Ga2O3	CVD	5	∼100	9.5 × 10^3	10	1.5 × 10^15	52
β-Ga2O3	M.E.^a	5	12.09	44.09	27.74	5.84 × 10^12	53
β-Ga2O3	MBE	10	4.2	83.3	690	1.35 × 10^15	54
β-Ga2O3	MOCVD	10	29	38.82	500	9 × 10^15	55
β-Ga2O3	MOCVD	20	0.635	72.35	8	—	56
β-Ga2O3	MOCVD	0	—	0.25	0.031	—	57
α-Ga2O3	MBE	−5	<1	0.5	0.238	9.66 × 10^12	58
β-Ga2O3:Zn	MOCVD	20	∼10	210	1400	—	59
β-Ga2O3:Zn	MOCVD	10	37	2.8 × 10^3	240	5.9 × 10^16	35
ZnGa2O4	PLD	5	8.56	1.97	400	—	60
ZnGa2O4	MOCVD	5	0.916	639	400	—	61
m-InGaO	CVD	10	0.03	0.97	260	4.62 × 10^14	29
m-Ga2O3/SnO2	CVD	40	11	0.32	1708	2 × 10^13	28
m-MgZnO	MBE	10	4	32	37	—	48
m-Ga2O3/ZnGa2O4	MOCVD	10	0.137	2058.54	2.90	3.21 × 10^15	This work
m-Ga2O3/ZnGa2O4	MOCVD	10	0.453	2364.01	2.32	3.15 × 10^15	This work

---

The substantial enhancement in device performance observed at an annealing temperature of 800 °C may be attributed to the formation of heterojunction interfaces between the mixed-phases within the thin films. The operation mechanism of the solar-blind photodetector based on Ga₂O₃/ZnGa₂O₄ mixed-phase films is proposed and schematically presented in Fig. 8. Fig. 8a presents a schematic representation of the crystal phase structure within the mixed-phase film. Under thermal equilibrium conditions (Fig. 8b), the high density of interface states results in recombination centers being primarily localized at the heterojunction interface between Ga₂O₃ and ZnGa₂O₄, where electrons and holes overcome the potential barrier and recombine. When a bias voltage and ultraviolet illumination are applied, the energy band structure transitions to the configuration shown in Fig. 8c, and the interface states capture holes from the newly generated photogenerated carriers. The internal photoconductive gain of the detector can be described by eqn (3)

where τ represents the lifetime of photo-generated holes, and t_m denote the transit times of electrons. The presence of interface states extends the carrier lifetime, thereby significantly enhancing the photocurrent. Furthermore, the heterojunction interface broadens the built-in electric field and introduces a potential barrier, which restricts the drift motion of majority carriers and effectively suppresses the dark current, ensuring that it remains at a low level.⁴⁸ By optimizing the degree of phase mixing within the thin film, the response time can be controlled within an acceptable range, achieving a balance between the rapid response and high responsivity. These optimizations contribute to the enhanced overall performance of solar-blind ultraviolet detectors.

Fig. 8 Energy band diagrams and the carrier transport process in the mixed-phase detector. (a) Schematic representation of the crystal phase structure within the mixed-phase film. (b) Under thermal equilibrium conditions. (c) Under applied bias with ultraviolet light illumination. The schematics depict the flow of charge carriers, with blue spheres representing electrons and red spheres representing holes.

Conclusions

High-Zn-content doped β-Ga₂O₃ films were epitaxially grown on c-Al₂O₃ substrates using MOCVD. Annealing at temperatures above 800 °C induced a Ga₂O₃/ZnGa₂O₄ mixed-phase structure in the films. A photodetector with a simple MSM structure, fabricated from films annealed at 800 °C in an Ar atmosphere, demonstrated superior performance. Under a 10 V bias, the device exhibited a low dark current (0.137 pA), high photoresponse (2058.54 A W⁻¹), detectivity (3.21 × 10¹⁵ cm Hz^1/2 W⁻¹), and rapid response speed (decay time 2.90 ms). Through the analysis of oxygen vacancy concentrations under O₂ and Ar atmospheres and their correlation with device performance, we have ruled out oxygen vacancies as the primary factor influencing device behavior. Instead, the enhanced performance of the ultraviolet detector is attributed to the formation of a mixed-phase heterojunction interface. These findings contribute to a deeper understanding of mixed-phase material systems and provide important guidance for the development and optimization of high-performance ultraviolet detectors.

Author contributions

The manuscript was written through contributions from all authors. All authors have given approval to the final version of the manuscript.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article.

Conflicts of interest

The authors declare no competing financial interest.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (12204474, 12304111, 12304112, 62074148 and 11727902), the Jilin Province Young and Middle-aged Science and Technology Innovation Leaders and Team Project (20220508153RC), the Jilin Province Science Fund (20220101053JC), the Youth Innovation Promotion Association, CAS (2020225), and the Provincial Science and Technology Major Project of Jiangsu (BG2024030).

References

1. Z. Li, T. Yan and X. Fang, Nat. Rev. Mater., 2023, 8, 587–603.
2. C. Xie, X. T. Lu, X. W. Tong, Z. X. Zhang, F. X. Liang, L. Liang, L. B. Luo and Y. C. Wu, Adv. Funct. Mater., 2019, 29, 1806006.
3. X. Chen, F. Ren, S. Gu and J. Ye, Photon. Res., 2019, 7, 381–415.
4. X. He, R. Sun, X. Xu, H. Geng, S. Qi, K. H. Zhang and H. Long, ACS Appl. Mater. Interfaces, 2024, 16, 64146–64155.
5. U. Varshney, N. Aggarwal and G. Gupta, J. Mater. Chem. C, 2022, 10, 1573–1593.
6. J. Ding, P. Zhao, H. Chen and H. Fu, Appl. Phys. A, 2024, 130, 350.
7. J. Shi, J. Zhang, L. Yang, M. Qu, D. C. Qi and K. H. L. Zhang, Adv. Mater., 2021, 33, 2006230.
8. Z. Wang, R. Yu, X. Wang, W. Wu and Z. L. Wang, Adv. Mater., 2016, 28, 6880–6886.
9. P. Vashishtha, I. H. Abidi, S. P. Giridhar, A. K. Verma, P. Prajapat, A. Bhoriya, B. J. Murdoch, J. O. Tollerud, C. Xu, J. A. Davis, G. Gupta and S. Walia, ACS Appl. Mater. Interfaces, 2024, 16, 31294–31303.
10. L. Cheng, Y. Wu, W. Cai and W. Zheng, Mater. Today Phys., 2023, 36, 101164.
11. M. M. Fan, K. W. Liu, Z. Z. Zhang, B. H. Li, X. Chen, D. X. Zhao, C. X. Shan and D. Z. Shen, Appl. Phys. Lett., 2014, 105, 011117.
12. S. Han, H. Xia, Y. M. Lu, W. J. Liu, W. Y. Xu, M. Fang, P. J. Cao and D. L. Zhu, Nanotechnology, 2021, 32, 235202.
13. S. Han, D. Yue, P. Cao, W. Liu, Y. Zeng, M. Fang and D. Zhu, Opt. Mater., 2024, 147, 114612.
14. J. Zhang, J. Shi, D.-C. Qi, L. Chen and K. H. L. Zhang, APL Mater., 2020, 8, 020906.
15. C. Xie, X. Lu, Y. Liang, H. Chen, L. Wang, C. Wu, D. Wu, W. Yang and L. Luo, J. Mater. Sci. Technol., 2021, 72, 189–196.
16. S. Kim, Y. Yoon, D. Seo, J.-H. Park, D.-W. Jeon, W. S. Hwang and M. Shin, APL Mater., 2023, 11, 061107.
17. H. Huang, H. Yin, K. Han, Y. Wang, Z. Wang, X. Feng, Y. Zou, X. Zhou, G. Xu, X. Hou, X. Zhao and S. Long, Adv. Opt. Mater., 2024, 12, 2400788.
18. J. Zhang, X. Kuang, R. Tu and S. Zhang, Adv. Colloid Interface Sci., 2024, 328, 103175.
19. Z. Wang, G. Zhang, X. Zhang, C. Wu, Z. Xia, H. Hu, F. Wu, D. Guo and S. Wang, Adv. Opt. Mater., 2024, 12, 2401256.
20. C. Wu, F. Wu, H. Hu, S. Wang, A. Liu and D. Guo, Mater. Today Phys., 2022, 28, 100883.
21. C. Wu, T. Zhao, H. He, H. Hu, Z. Liu, S. Wang, F. Zhang, Q. Wang, A. Liu, F. Wu and D. Guo, Adv. Opt. Mater., 2024, 12, 2302294.
22. Y. M. Lu, C. Li, X. H. Chen, S. Han, P. J. Cao, F. Jia, Y. X. Zeng, X. K. Liu, W. Y. Xu, W. J. Liu and D. L. Zhu, Chin. Phys. B, 2019, 28, 018504.
23. S. Ganguly, K. Nama Manjunatha and S. Paul, Phys. Status Solidi RRL, 2023, 18, 2300296.
24. H. Liu, H. Li, S. Zhou, H. Zhang, S. Fan, Y. Cui, C. Kong, L. Ye, Y. Xiong and W. Li, Crystals, 2021, 11, 1111.
25. A. A. Mohammed and W. Foong Lim, Mater. Sci. Eng. B, 2024, 308, 117613.
26. F. Zhang, J. Sun, H. Li, J. Zhou, T. Sun, T. Fu, G. Xiao, Y. Chen, Y. Deng and J. Zhai, J. Electron. Mater., 2019, 48, 8061–8066.
27. Y.-C. Chen, G. Zeng, S.-T. Ding, C.-Y. Shi, X.-F. Zhao, D. W. Zhang and H.-L. Lu, ACS Mater. Lett., 2023, 5, 2852–2861.
28. M.-M. Fan, L. Cao, K.-L. Xu and X.-Y. Li, J. Alloys Compd., 2021, 853, 157080.
29. K. Qi, S. Fu, Y. Wang, Y. Han, R. Fu, C. Gao, J. Ma, H. Xu, B. Li, A. Shen and Y. Liu, J. Alloys Compd., 2023, 965, 171473.
30. B. M. Janzen, P. Mazzolini, R. Gillen, A. Falkenstein, M. Martin, H. Tornatzky, J. Maultzsch, O. Bierwagen and M. R. Wagner, J. Mater. Chem. C, 2021, 9(7), 2311–2320.
31. H. Cui, H. F. Mohamed, C. Xia, Q. Sai, W. Zhou, H. Qi, J. Zhao, J. Si and X. Ji, J. Alloys Compd., 2019, 788, 925–928.
32. C. Kranert, C. Sturm, R. Schmidt-Grund and M. Grundmann, Sci. Rep., 2016, 6, 35964.
33. S. Gao, W. Li, J. Dai, Q. Wang and Z. Suo, Mater. Res. Express, 2021, 8, 025904.
34. E. Chikoidze, C. Sartel, I. Madaci, H. Mohamed, C. Vilar, B. Ballesteros, F. Belarre, E. del Corro, P. Vales-Castro, G. Sauthier, L. Li, M. Jennings, V. Sallet, Y. Dumont and A. Pérez-Tomás, Cryst. Growth Des., 2020, 20, 2535–2546.
35. X. Sun, K. Liu, X. Chen, Q. Hou, Z. Cheng, J. Yang, Q. Ai, Y. Zhu, B. Li, L. Liu and D. Shen, J. Mater. Chem. C, 2023, 11, 3227–3234.
36. S. S. Shankar, U. Manzoor, D. K. Kim, M. Islam and A. S. Bhatti, PLoS One, 2014, 9, e86418.
37. H. Lee, S. Kim, H. Ahn, K. Kim and M. Yang, Cryst. Res. Technol., 2020, 56, 2000149.
38. S. S. Kumar, E. J. Rubio, M. Noor-A-Alam, G. Martinez, S. Manandhar, V. Shutthanandan, S. Thevuthasan and C. V. Ramana, J. Phys. Chem. C, 2013, 117, 4194–4200.
39. L. Dong, R. Jia, B. Xin, B. Peng and Y. Zhang, Sci. Rep., 2017, 7, 40160.
40. J. Wang, X. Ji, Z. Yan, S. Qi, X. Liu, A. Zhong and P. Li, J. Alloys Compd., 2024, 970, 172448.
41. M. Narayanan, A. P. Shah, S. Ghosh, A. Thamizhavel and A. Bhattacharya, Appl. Phys. Lett., 2023, 123, 172106.
42. H. Huang, L. Wang, H. Zhou, H. Xing, L. Wang, W. Zhang, K. Tang, J. Huang and L. Wang, ACS Appl. Mater. Interfaces, 2024, 16, 63808–63817.
43. L. Gu, H.-P. Ma, Y. Li, A.-F. Wang, W.-J. Chen, Z.-R. Tang, Y. Shen, F. Y. Sun, J.-T. Zhu and Q.-C. Zhang, Appl. Surf. Sci., 2023, 641, 158502.
44. X. Duan, F. Yu and Y. Wu, Appl. Surf. Sci., 2012, 261, 830–834.
45. A. R. Phani, S. Santucci, S. DiNardo, L. Lozzi, M. Passacantando, P. Picozzi and C. Cantalni, J. Mater. Sci., 1998, 33, 3969–3973.
46. Y. Xin, Z. Gao, X. Shang, J. Wu, D. Yu, J. Xiu and Z. Li, J. Alloys Compd., 2023, 933, 167760.
47. D. Li, R. Deng, Y. Li and D. Jiang, ACS Appl. Mater. Interfaces, 2024, 16, 45156–45165.
48. M.-M. Fan, K.-W. Liu, X. Chen, X. Wang, Z.-Z. Zhang, B.-H. Li and D.-Z. Shen, ACS Appl. Mater. Interfaces, 2015, 7, 20600–20606.
49. Q. Chen, J. W. Yang, A. Osinsky, S. Gangopadhyay, B. Lim, M. Z. Anwar, M. A. Khan, D. Kuksenkov and H. Temkin, Appl. Phys. Lett., 1997, 70, 2277–2279.
50. H.-Y. Lee, J.-T. Liu and C.-T. Lee, IEEE Photonics Technol. Lett., 2018, 30, 549–552.
51. Y. Fang, A. Armin, P. Meredith and J. Huang, Nat. Photonics, 2018, 13, 1–4.
52. 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 and T. Y. Zhang, Mater. Today Phys., 2020, 14, 100226.
53. S. Qi, J. Liu, J. Yue, X. Ji, J. Shen, Y. Yang, J. Wang, S. Li, Z. Wu and W. Tang, J. Mater. Chem. C, 2023, 11, 8454–8461.
54. J. Tian, X. Song, S. Yang, B. Chen, S. Lei, X. Zhang and L.-X. Qian, IEEE Trans. Electron Devices, 2024, 71, 3746–3752.
55. C. Zhang, K. Liu, Q. Ai, X. Huang, X. Chen, Y. Zhu, J. Yang, Z. Cheng, B. Li, L. Liu and D. Shen, J. Phys. Chem. C, 2022, 126, 21839–21846.
56. Y. Yuan, Z. Li, X. Hou, X. Zhao, M. Ding, S. Yu, Z. Wang, J. Liu, G. Xu, Z. Jia, X. Tao, W. Mu and S. Long, J. Alloys Compd., 2023, 969, 171596.
57. Z. Cai, X. He, K. Wang, X. Hou, Y. Mei, L. Ying, B. Zhang and H. Long, Small Methods, 2023, 8, 2301148.
58. X. Chen, Y. Xu, D. Zhou, S. Yang, F.-F. Ren, H. Lu, K. Tang, S. Gu, R. Zhang, Y. Zheng and J. Ye, ACS Appl. Mater. Interfaces, 2017, 9, 36997–37005.
59. F. Alema, B. Hertog, O. Ledyaev, D. Volovik, G. Thoma, R. Miller, A. Osinsky, P. Mukhopadhyay, S. Bakhshi, H. Ali and W. V. Schoenfeld, Phys. Status Solidi A, 2017, 214, 1600688.
60. C.-C. Yen, A. K. Singh, H. Chang, K.-P. Chang, P.-W. Chen, P.-L. Liu and D.-S. Wuu, Appl. Surf. Sci., 2022, 597, 153700.
61. A. Sood, F.-G. Tarntair, Y.-X. Wang, T.-C. Chang, Y.-H. Chen, P.-L. Liu and R.-H. Horng, Results Phys., 2021, 29, 104764.

---

Footnote

† Electronic supplementary information (ESI) available: AFM, EDS, and performance of films annealing under different conditions. See DOI: https://doi.org/10.1039/d5tc00710k

---