A wide ultraviolet spectra response photodetector based on epitaxial growth of highly-oriented ε-Ga₂O₃ crystal on diamond substrate

Abstract

In recent years, diamond has shown great potential in solar-blind ultraviolet (UV) photodetection due to its ultrawide bandgap (∼ 5.5 eV) and other superior semiconductor properties. However, the response region of diamond photodetector is usually smaller than 230 nm, which cannot cover the whole solar-blind region from 200 to 280 nm. In this work, an ε-Ga₂O₃/diamond photodetector with wide spectra responsivity from 210 (or even lower) to 260 nm was fabricated. High-quality ε-Ga₂O₃ film with columnar crystal was epitaxially grown on single crystalline CVD diamond substrate by pulse laser deposition (PLD). TEM characterization revealed that the ε-Ga₂O₃ film grew along the <001> orientation on diamond (100) substrate. The deep ultraviolet (DUV) photodetector based on the ε-Ga₂O₃/diamond structure showed a high light-to-dark ratio over 5.7 × 10⁴ and good linear response to the incident light power density from 10 to 400 mW/cm². Moreover, compared to other photodetectors, the fabricated ε-Ga₂O₃/diamond photodetector achieved high responsivity and wide spectra response region from 210 to 260 nm, with high solar-blind rejection ratio of 104 (R₂₄₀/R₂₈₀) and 165 (R₂₁₀/R₂₈₀), respectively. The extension of spectra region with high responsivity of the ε-Ga₂O₃/diamond photodetector can be attributed to the thin thickness of ε-Ga₂O₃ film (around 200 nm) and parts of the DUV light were absorbed by diamond. The high responsivity and wide spectra response region indicate the fabricated ε-Ga₂O₃/diamond photodetector can be used for the detection of ultraviolet in the most of the DUV region.

Keywords:

• Diamond
• ε-Ga₂O₃
• epitaxial growth
• deep ultraviolet photodetector
• wide spectra response

1. Introduction

The deep ultraviolet (DUV) with wavelengths from 200 to 280 nm is commonly referred to as the solar-blind UV range, as solar radiation in this wavelength region is effectively blocked by the ozone layer in the Earth’s atmosphere [1–3]. Owing to the low natural background, photodetectors operating in solar-blind region have high signal-to-noise ratios and low false alarm rates. Therefore, solar-blind photodetectors have many civil and military applications such as biological/chemical, sterilization and disinfection, missile tracking and environmental monitoring [4–6]. Unfortunately, most of the commercial semiconductor DUV photodetectors are made of silicon (Si), and the narrow bandgap (1.12 eV) makes Si photodetector vulnerable to visible light interference during DUV detection, thus limiting its applications. Diamond has emerged as a promising material for solar-blind photodetection due to its wide bandgap (E_g: ∼ 5.5 eV) and high radiation hardness, which can promise intrinsic invisible/solar blindness [7,8]. However, due to the limitation of the band gap and UV adsorption edge, diamond photodetectors only show response to DUV light with wavelengths smaller than 230 nm. Therefore, to realize the full (or nearly full) wavelength and high response to DUV light in a single photodetector, a semiconductor with smaller bandgap is necessarily to be employed to form heterostructure with diamond.

Gallium oxide (Ga₂O₃), as another ultrawide bandgap semiconductor, has also been widely used for DUV photodetection due to its wide bandgap (E_g: ∼ 4.9 eV), and other fantastic physical and electrical properties [9,10]. However, due to the decrease of adsorption coefficient (α) in the short wavelength region (< 240 nm) [11,12], the high spectra response regions of the Ga₂O₃ photodetectors are usually between 240 and 260 nm [13–15], which correspond exactly to the irresponsive region of diamond photodetectors in DUV region. Therefore, the combination of Ga₂O₃ and diamond has gained much interest in recent years, leading to the development of different fabrication methods of Ga₂O₃/diamond photodetectors [16,17]. Kim et al. transferred the β-Ga₂O₃ nanolayer onto the diamond substrate through mechanical exfoliation and fabricated a β-Ga₂O₃/diamond photodetector [16]. However, the fragility of β-Ga₂O₃ film limits the transferring of large-scale β-Ga₂O₃ films. Therefore, Chen et al. fabricated a β-Ga₂O₃/diamond photodetector by growing β-Ga₂O₃ film on diamond substrate through plasma enhanced chemical vapor deposition (PECVD) [17]. However, due to the low symmetric monoclinic structure of β-Ga₂O₃ (space group: C2/m) and lattice mismatch between β-Ga₂O₃ and diamond, the β-Ga₂O₃ film had a poor crystalline quality with a full width at half maximum (FWHM) of the rocking curve of over 2°, which leaded to a low responsivity (0.2 mA/W).

As another major Ga₂O₃ polymorph, ε-Ga₂O₃ is the second most stable phase and has many fantastic properties, like ferroelectric polarization [18,19]. Recently, the crystal structure of ε-Ga₂O₃ has been demonstrated to be composed of orthogonal nanoscale (about 5–10 nm) ordered κ-Ga₂O₃ cells (space group: Pna21) arranged 120° next to each other [20]. Therefore, the crystal structure of ε-Ga₂O₃ is regarded as orthographic (space group: Pna21) rather than the hexagonal (space group: P63mc) as previously thought. The orthorhombic crystal structure promises ε-Ga₂O₃ more suitable for epitaxial growth on hexagonal and cubic semiconductor substrates [21–23]. Therefore, due to better lattice compatibility, the high crystallographic symmetry indicates that it is possible to grow high quality ε-Ga₂O₃ crystal on single crystalline diamond substrates. In our previous work, an ε-Ga₂O₃/diamond photodetector with high thermal stability was fabricated by growing crystalline ε-Ga₂O₃ film on the diamond substate [24]. However, till now, there are few reports on the epitaxial growth of ε-Ga₂O₃ on diamond substrate, as well as the photodetector based on ε-Ga₂O₃/diamond heterostructure.

In this work, an ε-Ga₂O₃/diamond photodetector with nearly full spectra response in DUV region was fabricated. The high quality ε-Ga₂O₃ film was epitaxially grown on single crystalline diamond substrate by pulsed laser deposition (PLD). The crystal qualities of the ε-Ga₂O₃ film on the diamond substrate were detailly characterized by XRD and TEM, which revealed that the ε-Ga₂O₃ film was composed of a series of columnar crystals with in-plane rotational domains. Besides, the detailed structure of ε-Ga₂O₃/diamond interface was determined by a combination of high-resolution scanning TEM and first principles calculations. Based on the obtained ε-Ga₂O₃/diamond heterostructure, a DUV photodetector was fabricated by depositing Ti/Au electrodes on ε-Ga₂O₃ surface. Compared to other Ga₂O₃ photodetectors, the fabricated ε-Ga₂O₃/diamond photodetector showed high spectra response from below 210 nm to the wavelength about 260 nm. The fabrication of ε-Ga₂O₃/diamond photodetector with broad and high spectra response will push the investigation of the integration of ultrawide bandgap heterostructures in the future.

2. Materials and experiments

The CVD diamond (100) substrate with size of 7 × 7 × 0.3 mm³ used in this work is grew in our lab by micro-plasma chemical vapor deposition (MPCVD). The Raman peak center of the CVD diamond locates at about 1332.3 cm⁻¹ and the full width at half maximum (FWHM) is about 2.1 cm⁻¹. The ε-Ga₂O₃ film was deposited on diamond substrate by PLD system (SKY Technology Development Co., Ltd. Chinese Academy of Science) using a 248 nm KrF excimer laser. During ε-Ga₂O₃ growth, the oxygen pressure and growth temperature were maintained at 20 mTorr and 800 °C, respectively. The thickness of ε-Ga₂O₃ film can be controlled by changing the deposition cycling times. An ε-Ga₂O₃/diamond photodetector with metal-semiconductor-metal (MSM) structural interdigital electrodes (200 μm gap between electrodes) was then fabricated by depositing Ti (30 nm)/Au (50 nm) on the grown ε-Ga₂O₃ surface through electron beam deposition system (SKY Technology Development Co., Ltd. Chinese Academy of Science). An in-situ thermal annealing at about 450 °C was performed in vacuum to achieve good Ohmic contact between electrodes and substrate.

The crystal structure of ε-Ga₂O₃ film grown on single crystalline diamond substrate was characterized by X-ray diffraction (XRD, D8 Advance, Bruker, USA), and the microstructure of the ε-Ga₂O₃ film was characterized using an aberration-corrected transmission electron microscopy (TEM, Spectra 300, Themo Fisher Scientific, USA). The transmittance of ε-Ga₂O₃, diamond and ε-Ga₂O₃/diamond samples in the range from 190 to 800 nm were measured by UV-Visible spectrometer (Lambda 1050+, Perkin Elmer, USA). The I–V characteristic of the ε-Ga₂O₃/diamond photodetector was measured in a semiconductor parameter analyzer (Keithley 4200-SCS, Tektronix, USA) under the illumination of different ultraviolet light sources. For the scanning test, the voltage was increased from minus to positive values at a step of 0.2 V. The UV source used in this work is composed of a while light source (EQ99X LDLS, Energetiq, Holland) and spectrophotometer (iHR 320, Horiba, Japan).

3. Results and disscussion

XRD was performed to characterize the crystal phase and orientation of the as-grown Ga₂O₃ film on diamond substrate. According to the θ–2θ scan in Figure 1(a), besides the (400) plane diffraction peak of diamond substrate located at 119.4°, five obvious diffraction peaks at 19.1°, 38.8°, 59.8°, 83.5° and 112.7° can be also observed, which belong to the (002), (004), (006), (008) and (0010) planes of ε-Ga₂O₃ (space group: Pna2₁) [20]. The appearance of the only (001) series diffraction peaks of ε-Ga₂O₃ on the XRD spectra confirms that the as-grown ε-Ga₂O₃ film on diamond substrate has high purity and high crystal orientation. As shown in Figure 1(b), the measured FWHM of the ε-Ga₂O₃ (002) in the ω-scan mode is about 0.89°, which is much lower than the reported β-Ga₂O₃ film on diamond substrate grown by low pressure chemical vapor deposition (LPCVD) [25]. Figure 1(c) shows the comparison of the transmittance of diamond sample in the UV-visible region before and after ε-Ga₂O₃ film deposition. The transmittance of diamond is around 70% in the visible region, approximating to the theoretical value of intrinsic diamond. After ε-Ga₂O₃ deposition, when light is transmitted from ε-Ga₂O₃ side, the transmittance of ε-Ga₂O₃/diamond sample is limited below 70% in the region from 300 to 800 nm due to the existence of diamond substrate. Moreover, two absorption edges at 275 nm and 228 nm can be observed, which belong to the absorption edges of intrinsic ε-Ga₂O₃ film and diamond substrate, respectively. The insert Tauc plot analysis in Figure 1(c) revels that optical bandgap of ε-Ga₂O₃ film and diamond substrate are about 4.8 and 5.5 eV, respectively. The transmittance comparisons of the ε-Ga₂O₃/diamond samples with different ε-Ga₂O₃ thicknesses are shown in Figure S1 in Supplementary Information. With the increase of ε-Ga₂O₃ film thickness from about 10–200 nm, the transmittances of ε-Ga₂O₃/diamond samples, where the absorption edge of diamond occurred, decreased from about 40% to 8%. When thickness of ε-Ga₂O₃ is higher than 400 nm, the DUV below 230 nm is fully adsorbed by ε-Ga₂O₃ film, as shown in Figure S1.

Figure 1. (a) XRD θ-2θ scan of ε-Ga₂O₃ film grown on diamond substrate, (b) XRD ω-scan of around the ε-Ga₂O₃ (002) peak, (c) comparison of the UV-Vis transmittance spectra and Tauc plot (Inset) of ε-Ga₂O₃ film grown on diamond substrate (blue), diamond (green) and ε-Ga₂O₃ (orange). the thickness of ε-Ga₂O₃ film in (c) is about 225 nm.

The microstructure of ε-Ga₂O₃ film on diamond substrate was characterized by TEM, as shown in Figure 2. It can be seen from the low magnification cross-section TEM image shown in Figure 2(a) that the thickness of ε-Ga₂O₃ film is around 200 nm. The ε-Ga₂O₃ film shows a lath-shaped contrast vertical to the surface of diamond substrate, indicating the grown ε-Ga₂O₃ film is polycrystalline composing of columnar structures, which was further verified by the selected area electron diffraction (SAED) pattern and high-resolution TEM (HRTEM) image, shown in Figure 2(b) and (c). The main diffraction spots (circled by green circles) in the SAED pattern shown in Figure 2(b) can be attributed to the ε-Ga₂O₃ grain oriented along the [010] zone axis, while other diffraction spots are attributed to the in-plane domain rotations of the ε-Ga₂O₃ crystal lattice (Figure S2 in supporting information). It is worth noting that all diffraction spots observed in the SAED formed lines perpendicular to the [001] crystallographic direction, indicating that the (001) planes of all ε-Ga₂O₃ grains are parallel to each other, and it is consistent with the XRD pattern showing high orientation consistency of the as-grown ε-Ga₂O₃. Figure 2(c) shows an HRTEM image of the ε-Ga₂O₃ containing two grains separated by the grain boundary marked in green. The interplanar spacings for both grains are determined to be 0.93 nm, which is consistent with the value expected for the (001) crystal planes of ε-Ga₂O₃ [20]. Due to the parallel alignment of the (001) crystal planes of ε-Ga₂O₃ with the diamond substrate, the perpendicular orientation of the grain boundary (marked by a green line) to the (001) planes of the two adjacent grains indicates the columnar crystal structure of the ε-Ga₂O₃ and also confirms the consistent orientation of the grains along the <001> direction. Scanning TEM (STEM) was carried out on the ε-Ga₂O₃ grain with [010] zone axis, as shown in Figure 2(d). The structure of the grain is consistent with the crystal model of ε-Ga₂O₃ viewed along the [010] direction as shown in Figure 2(e), confirming again the ε-phase of Ga₂O₃ on diamond substrate.

Figure 2. (a) Low magnification cross-section TEM image of ε-Ga₂O₃ film on diamond. (b) SEAD pattern taken from the region marked by the dashed circle in (a). (c) HRTEM image showing the grain boundary in the ε-Ga₂O₃ film. (d) STEM image showing the ε-Ga₂O₃ structure with the [010] zone axial at atomic resolution. (e) Enlarged image taken from the yellow rectangular region of (d). the insert in (e) shows the atomic model of ε-Ga₂O₃ viewed along the [010] direction. The red and blue balls represent Ga and O atoms, respectively.

In order to understand the growth mechanism of ε-Ga₂O₃ on diamond substrate, the microstructure of ε-Ga₂O₃/diamond interface was investigated by a combination of TEM characterizations and first principles calculations. The STEM image of the ε-Ga₂O₃/diamond interface is shown in Figure 3(a) where the (001) planes of ε-Ga₂O₃ are parallel to the (001) planes of diamond. It can also be observed that there is not an obvious interface layer between diamond and ε-Ga₂O₃ crystal. The suitable epitaxial growth of ε-Ga₂O₃ on substrates with cubic and orthorhombic structures is due to the orthorhombic structure of ε-Ga₂O₃. It should be noted that the lattice constant a_ε of ε-Ga₂O₃ is 5.04 Å [20], which is very close to the facial diagonal length of the unit cell of cubic diamond: $\sqrt{2}$a_D = 5.01 Å. The lattice constant mismatch between a_ε and $\sqrt{2}$a_D is smaller than 1%. However, for the lattice constant b_ε of ε-Ga₂O₃, it follows: 3/2 $\sqrt{2}$a_D < b_ε < 2$\sqrt{2}$a_D, as shown in Figure S3 in Supplementary Information. Therefore, beside of 120°, other rotation angles may occur during ε-Ga₂O₃ growth, while the pseudohexagonal 4H stacking of the O atoms is preserved, giving rise of the columnar structure of ε-Ga₂O₃ on diamond substrate [20]. The SAED pattern taken from the ε-Ga₂O₃/diamond interface is shown in Figure 3(b). The reflections of diamond marked by green circles and ε-Ga₂O₃ marked by yellow circles show that the (001) planes of the ε-Ga₂O₃ are parallel to the (001) planes of diamond, which matches well with the HRTEM image shown in Figure 3(a).

Figure 3. (a) STEM image of the ε-Ga₂O₃/diamond interface. (b) SAED pattern taken from the ε-Ga₂O₃/diamond interface. Diffraction spots marked by green and yellow circles represent diamond (with [110] zoon axis) and ε-Ga₂O₃ (with [010] zoon axis), respectively. (c) Atomic model of the ε-Ga₂O₃/diamond interface. The red and blue balls represent Ga and O atoms. The green and white balls represent the C atoms at the surface and inner of diamond substrate, respectively. (d) The calculated total energies of different atomic layers of ε-Ga₂O₃ contacting with (001) diamond substrate.

The crystal structure of ε-Ga₂O₃ is complex, whose unit cell contains 16 Ga and 24 O atoms. The atoms in one unit cell can be clarified into 10 different layers vertical to the [001] direction. Thanks to the high-resolution STEM image, the atomic model of the ε-Ga₂O₃/diamond interface is shown in Figure 3(c). The atomic layers of ε-Ga₂O₃ are labeled as 1–10. One should be aware that it is difficult to determine the atomic layer close to the diamond by TEM characterizations due to its limited resolution. In this case, the first principle calculation was used to analyze the detailed structure of the ε-Ga₂O₃/diamond interface. The ε-Ga₂O₃/diamond interface models containing 140 atoms (96 C atoms, 17 Ga atoms and 27 O atoms) were built for calculations. A vacuum space of 12 Å was left along the Z-direction. The calculation was carried out using the Vienna ab initio Simulation Package (VASP) based on density functional theory, and the Projector-Augmented Wave (PAW) method together with the Perdew Burke Ernzerhof (PBE) formulation were employed [26]. The cutoff energy of the plane waves was set as 450 eV and the Monkhorst-Pack K-point mesh was 4 × 4 × 1. The diamond structure was fixed and the ε-Ga₂O₃ structure was fully relaxed until the Hellmann-Feynman force on all atoms was less than 0.05 eV/Å. The calculated total energy of the ε-Ga₂O₃/diamond interface system is shown in Figure 3(d). The total energy varies from −986 eV to −980 eV when the O atoms of ε-Ga₂O₃ contact with diamond substrate, lower than that when the Ga atoms of ε-Ga₂O₃ contact with diamond substrate (−973 eV to −962 eV). The calculated results show that compared to Ga atoms, the C atoms on diamond surface tend to form covalent bonds with O atoms due to the closer covalent radiuses of C (0.77 Å) and O (0.73 Å) atoms.

A solar-blind photodetector based on the obtained ε-Ga₂O₃/diamond heterostructure was fabricated by depositing Ti (30 nm)/Au (50 nm) interdigital electrodes on ε-Ga₂O₃ surface. As shown in Figure 4(a), the measured I–V curve was almost linear under no light illumination which indicates a good ohmic contact of Ti/Au electrodes on ε-Ga₂O₃ surface. The zero point of the I–V curve shifted to the negative voltage because of charging and discharging on the highly insulating ε-Ga₂O₃ film. At an applied voltage of 30 V, the ultralow dark current was about 4.3 × 10⁻¹³ A as shown in Figure 4(b). When exposed to 240 nm ultraviolet light, the ε-Ga₂O₃/diamond photodetector (effective area: ∼ 2.2 mm²) achieved a photocurrent of about 2.5 × 10⁻⁸ A for an applied bias of 30 V and a light-to-dark current ratio of more than 5.7 × 10⁴. The photoresponse of ε-Ga₂O₃/diamond photodetector under different incident light power densities is shown in Figure 4(c) where the photocurrent shows a linear relationship to the incident light power density from 10 mW/cm² to over 400 mW/cm². Moreover, the time response of the ε-Ga₂O₃/diamond photodetector is shown in Figure 4(d). The measured rise time (τ_up) and decay time (τ_decay) of the photodetector are about 0.64 s and 0.13 s when the current increases from 10% to 90% and decreases from 90% to 10% of the maximum current, respectively. The rise and decay times of the photodetector are similar to the reported values of other Ga₂O₃/diamond photodetectors [17,24].

Figure 4. (a) The I–V curve of the ε-Ga₂O₃/diamond photodetector, the insert is the schematic illustration of the fabricated ε-Ga₂O₃/diamond photodetector. (b) Photocurrent and dark current of the ε-Ga₂O₃/diamond photodetector. (c) Relationship of the photo response to the incident light power density. (d) Time response of the fabricated ε-Ga₂O₃/diamond photodetector.

The spectra response of the ε-Ga₂O₃/diamond photodetector under different ultraviolet wavelength is shown in Figure 5. As shown in Figure 5(a), with the decrease of incident ultraviolet wavelength, the photocurrent increased dramatically. Notably, the zero point shift disappeared on the I–V curves when the ε-Ga₂O₃/diamond photodetector was illuminated to the DUV light with wavelength smaller than 260 nm. It can be explained that when the illumination DUV wavelength was smaller than 260 nm, photocurrent would be generated in the ε-Ga₂O₃ layer and the photo-generated carrier balanced the defects in ε-Ga₂O₃ layer, leading to the disappearance of the zero point shift. The calculated responsivity is shown in Figure 5(b), in which the insert shows the logarithmic relationship of the responsivity of the ε-Ga₂O₃/diamond photodetector to incident wavelength at different applied voltages. For independent Ga₂O₃ and diamond photodetectors, the peak responsivity usually locates at around 250 and 225 nm, respectively. However, in this work, the responsivity of the ε-Ga₂O₃/diamond photodetector did not reduce like other ε-Ga₂O₃ photodetectors, when the incident wavelength decreased to 240 nm [27,28]. Instead, a second increase tendency can be observed on the responsivity curve when the ultraviolet wavelength decreased below 230 nm. When the ultraviolet wavelength decreased to 210 nm, the responsivity achieved the maximum value of 0.28 A/W for voltage of 30 V. The high spectra responsivity region of ε-Ga₂O₃/diamond photodetector (ε-Ga₂O₃: ∼200 nm) was extended to 260–210 nm or even lower. While for an ε-Ga₂O₃/diamond photodetector with 1 μm- thick ε-Ga₂O₃ film, as shown in Figure S4, the peak responsivity of the ε-Ga₂O₃/diamond photodetector locates at around 250 nm, and with the decrease of illumination wavelength from 250 nm, the responsivity decreased dramatically, which was also reported in Ref. [24].

Figure 5. (a) Photocurrent and (b) responsivity of ε-Ga₂O₃/diamond photodetector under different incident wavelength ultraviolet light at different applied voltage, the insert figure in (b) shows corresponding logarithmic responsivity. The (c) solar-blind and (d) visible rejection ratio of the ε-Ga₂O₃/diamond photodetector at different applied voltages, respectively. A schematic illustration of the electric field intensity in ε-Ga₂O₃/diamond photodetector with (e) 200 nm and (f) 1 μm ε-Ga₂O₃ layer.

Moreover, Figure 5(c,d) show the solar-blind and visible light rejection ratios of the ε-Ga₂O₃/diamond photodetector under different applied voltages. It can be observed that all the rejection ratio curves show an increase tendency to the applied voltage from 5 V to 30 V. With the decrease of ultraviolet wavelength, the solar-blind rejection ratios of the ε-Ga₂O₃/diamond photodetector increases from 104 (R₂₄₀/R₂₈₀) to 133 (R₂₂₅/R₂₈₀) and 165 (R₂₁₀/R₂₈₀) at 30 V, respectively. Meanwhile, the visible light rejection ratios increase from 738 (R₂₄₀/R₂₈₀) to 946 (R₂₂₅/R₂₈₀) and 1174 (R₂₁₀/R₂₈₀) at 30 V, respectively. Such rejection values are much higher than the reported values of other Ga₂O₃/diamond photodetectors [17,24]. The high solar-blind and visible-blind rejection ratios of the ε-Ga₂O₃/diamond photodetector indicate the good solar-blind properties of the proposed ε-Ga₂O₃/diamond photodetector. Moreover, compared to the Ga₂O₃/diamond photodetectors in Figure S4 and other reports, the wavelength response region of the proposed ε-Ga₂O₃/diamond photodetector extended from 230 nm to below 210 nm, which was never reported before. Considering the occurrence of two adsorption edges of ε-Ga₂O₃/diamond in the UV-Vis spectra in Figure 1(c), the wide spectra responsivity of the fabricated ε-Ga₂O₃/diamond photodetector is due to the ultraviolet response of diamond substrate.

The simulation of the optical intensity and electric field distribution in ε-Ga₂O₃/diamond photodetector with different ε-Ga₂O₃ thickness were performed and compared. According to traditional physical theory, the light intensity decays exponentially with the light penetration depth. When the ε-Ga₂O₃ film was 200 nm or thinner, the DUV light can transmit the ε-Ga₂O₃ layer to the diamond substrate, which will generate the photocurrent in the diamond substrate (Figure S5). However, with the increase of ε-Ga₂O₃ film thickness, nearly all the DUV light with wavelength smaller than 230 nm was adsorbed by the ε-Ga₂O₃ layer, leading to the small or little photocurrent in diamond. Moreover, the distribution of the electric field in ε-Ga₂O₃ and diamond layers also reveal that with the increase of ε-Ga₂O₃ thickness, the electric field intensity becomes weaker in the diamond substrate, as shown in Figure 5(e,f) and Figure S6. Besides the above two reasons, quantum efficiency is also an important influence factor. The single crystalline diamond has much higher quantum efficiency than the columnar crystalline ε-Ga₂O₃ layer in the DUV region. Therefore, the extension of the spectra response region of the ε-Ga₂O₃/diamond photodetector was caused by above many reasons. The wide spectra response region indicates the ε-Ga₂O₃/diamond photodetector can be used for photodetection in the most of the DUV region.

4. Conclusion

In this work, an ε-Ga₂O₃/diamond photodetector with high spectra responsivity through the region from 210 nm to 260 nm was fabricated by epitaxially depositing ε-Ga₂O₃ film on diamond substrate by PLD. The microstructure of ε-Ga₂O₃ film gown on diamond has been shown to be the columnar crystal that rotating along the z axis. Due to the better lattice compatibility and crystal matching of ε-Ga₂O₃ and diamond, the ε-Ga₂O₃ film has good crystal quality with the FWHM of the ε-Ga₂O₃ (002) rocking curve of 0.89°. In comparison to other independent photodetectors based on ε-Ga₂O₃ or diamond materials, the fabricated ε-Ga₂O₃/diamond heterojunction photodetector achieved a significantly wide spectra response, from below 210 nm to about 260 nm. Moreover, the high rejection ratios for the responsivity for wavelengths of 280 nm and 400 nm were obtained, which demonstrated a high selectivity and sensitivity of the ε-Ga₂O₃/diamond photodetector in the most of the DUV region. The high responsivity and wide spectra region of the ε-Ga₂O₃/diamond photodetector shows great potential applications in DUV detection.

5. Associated content

Supporting Information: (I) Comparison of the UV-visible transmittance of ε-Ga₂O₃/diamond samples with different ε-Ga₂O₃ thicknesses. (II) The diffraction spots in the SAED pattern of ε-Ga₂O₃ film. (III) A potential contacting model of ε-Ga₂O₃ on diamond substrate according to calculation. (IV) The photo responsivity of the ε-Ga₂O₃/diamond photodetector with thick ε-Ga₂O₃ film. (V) Comparison of the optical intensity in ε-Ga₂O₃/diamond photodetector. (VI) Comparison of the electric field in ε-Ga₂O₃/diamond photodetector.

Author contributions

Q. Y.: conceived, measured photoresponse and wrote the manuscript; W. W.: grew ε-Ga₂O₃ film; W. Z.: gave direction on the ε-Ga₂O₃ growth; M. Q.: fabricated the device; M. Y. and Z. J.: helped measure photoresponse, B. W.: performed TEM characterization, C.-T. L. and K. N.: supervised the project; K. C.: developed the simulation; K. W. A. C.: supervised and corrected the manuscript; J. C.: TEM analysis and developed the theoretical model; N. Jiang: supervised and corrected the manuscript. All authors contributed to the discussions.

Acknowledgements

The authors thank Bingxue Zhang and Mingxin Hao (Qianwan Institute of CNITECH) for assistance with the characterization of ultraviolet-visible (UV-Vis) spectrophotometer and TEM sample preparation.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors are grateful for the financial support by the National Natural Science Foundation of China (52075527, 52102055), the National Key R&D Program of China (2022YFB3706602, 2021YFB3701801), Ningbo Key Scientific and Technological Project (2022Z191), the Yongjiang Talent Introduction Programme of Ningbo (2021A-037-C, 2021A-108-G), Science and Technology Major Project of Ningbo (2021ZDYF020196, 2021ZDYF020198) and the Project of Chinese Academy of Science (ZDKYYQ2020001) and Ningbo 3315 Innovation Team (2019A-18-C).