Temperature-Dependent Study of β-Ga₂O₃ Solar-Blind Photodetectors

Firstly, Laser Molecular Beam Epitaxy (LMBE) was used to grow the β-Ga₂O₃ thin film on Al₂O₃ (0001) substrates. The base pressure in growing the chamber was 1×10⁻⁶ Pa. To explore the optimum growth condition, the growth temperature ranged from 650 °C to 850 °C and oxygen pressure ranged from 5×10⁻³ Pa to 5×10⁻¹ Pa. Then, the laser ablation was conducted at a laser fluence of 5 J cm⁻² with a repetition target rate of 2 Hz using a KrF excimer laser with a wavelength of 248 nm. To improve the film uniformity, the substrate was rotated during deposition, and the distance between the target and substrate was 5 cm. Through a scanning electron microscope (SEM), the growing rate of 0.28 Å/pulse could be observed. The thickness of the films was estimated to be about 200 nm. The β-Ga₂O₃ thin film was undoped. Subsequently, to fabricate a photodetector, using a shadow mask by radio frequency magnetron sputtering, an interdigital Ti/Au electrode was deposited on the film, and then it was thermally annealed at 300 °C for 10 min in the Ar atmosphere. The electrode fingers have a width of 100 μm, a length of 2800 μm, and a spacing gap of 100 μm. The schematic diagram of the interdigital electrodes of the photodetector is illustrated in Fig. 1.

Zoom In Zoom Out Reset image size

Figure 2 shows the XRD patterns of Ga₂O₃ thin films. The results indicate that all peaks are indexed as β-Ga₂O₃ (−201) and higher-order diffractions except for the substrate diffraction peaks. No other peaks apart from these are found, indicating a single (−201) plane orientation growth. Our previous work has investigated the impact of the electron beam on β-Ga₂O₃, and the results indicated that electron irradiation excites electron-hole pairs in β-Ga₂O₃ and increases traps in β-Ga₂O₃. ¹²

Zoom In Zoom Out Reset image size

Then, temperature-dependent experiments were conducted on β-Ga₂O₃ photodetectors. The typical temperature points were selected at 40 K, 100 K, 160 K, 220 K, 280 K, and 300 K. The voltammetry curves and photo response curves of β-Ga₂O₃ photodetectors were measured by Keithley 4200 at different temperature points. The photoresponse curve was obtained by measuring the current of the β-Ga₂O₃ photodetector as time changed under a switching interval of 50 s and a bias voltage of 10 V. The timer used in the experiment is GCI-73 multifunctional precision electronic timer, which can ensure the accuracy of the switching interval. The light source was generated by ENF-240C, which is manufactured by the American Spectronics company. The light intensity of the 254-nm illumination and 365-nm illumination was 310 μw cm⁻² and 300 μw cm⁻² at a position of about 6 inches, respectively.

The experimental results of the β-Ga₂O₃ photodetector are shown in Fig. 3. Specifically, Figs. 3a, 3c, and 3e respectively show the I–V characteristics curve of the β-Ga₂O₃ photodetector in a dark condition, under 365-nm light illumination, and under 254-nm light illumination. The step of temperature is 60 K. Figures 3b, 3d, and 3f show the current change corresponding to each temperature point under a bias voltage of 10 V in a dark condition, under 365-nm light illumination, and under 254-nm light illumination. The step of temperature is 20 K. It can be seen from Fig. 3 that as the temperature increased from 40 K to 300 K, the current of the β-Ga₂O₃ photodetector rose first and then fell. At the voltage of 10 V, the β-Ga₂O₃ photodetector had the maximum current at the temperature of 160 K under the dark condition and 365-nm light illumination. However, at the voltage of 10 V, the β-Ga₂O₃ photodetector had the maximum current at the temperature of 220 K under 254-nm light illumination. To find the turning point at which the current changes with temperature, the step of the temperature was adjusted from 60 K to 20 K. The current of the β-Ga₂O₃ photodetector at the voltage of 10 V was then measured at different temperatures under different light conditions. The results are shown in the Figs. 3b, 3d and 3f. It can be seen that the transfer selection point of the current of the β-Ga₂O₃ photodetector with the change of temperature is around 180 K under the three light conditions.

Zoom In Zoom Out Reset image size

The conductivity of n-type doped β-Ga₂O₃ can be expressed as $\sigma =nq\mu ,$ ¹⁴ where n represents the electron concentration, q represents the unit charge, and μ represents the carrier mobility. Among the three variables, only μ is strongly correlated with temperature. It has been shown that the electron mobility in Ga₂O₃ reaches its maximum value around 190 K, ¹³ which explains why the turning point is around 180 K when the photocurrent changes with temperature.

In addition to the current, photodetectors also have two very important parameters, namely photo-to-dark current ratio (PDCR) and photoresponse. PDCR represents the increase of the photocurrent relative to the dark current, and it is calculated as follows:

$$\begin{eqnarray}&&PDCR=\displaystyle \frac{{I}_{P{\rm{hoto}}}-{I}_{D{\rm{ark}}}}{{I}_{D{\rm{ark}}}}\end{eqnarray} \tag{ 1 }$$

where I_Photo is the photocurrent of the β-Ga₂O₃ photodetector, and I_Dark is the dark current of the β-Ga₂O₃ photodetector. ¹⁵

Photoresponse represents the ratio of the photodetector's input signal to its output signal, and the calculation formula of photoresponse is

$$\begin{eqnarray}&&{R}_{\lambda }=\displaystyle \frac{{I}_{P{\rm{hoto}}}-{I}_{D{\rm{ark}}}}{{A}_{d}{P}_{d}}\end{eqnarray} \tag{ 2 }$$

where A_d is the detective area, and P_d is the intensity of the UV light source. ¹⁶

In this study, the photo-to-dark current ratio and photoresponse of the β-Ga₂O₃ photodetector were calculated under different lighting conditions and different temperatures. The results are shown in Fig. 4.

Zoom In Zoom Out Reset image size

As shown in Fig. 4, the PDCR and photoresponse under 254-nm illumination are much larger than those under 365-nm illumination. Meanwhile, the photoresponse of the β-Ga₂O₃ photodetector reached a peak both under 365-nm illumination and under 254-nm illumination, but the peak of the photoresponse of the β-Ga₂O₃ photodetector varied with the temperature. Under 365-nm illumination, the peak point was at 160 K, which was lower than that under 254-nm illumination. This may be related to the redshift phenomenon of photodetectors at low temperature, ¹⁷ indicating that the wavelength sensitive to the β-Ga₂O₃ photodetector may become longer under a low temperature. With the increase in temperature, the PDCR of the β-Ga₂O₃ photodetector increased gradually under 365-nm illumination but decreased gradually under 254-nm illumination. This indicated that as the temperature increased, the photocurrent generated by 365-nm illumination gradually increased while that generated by 254-nm illumination gradually decreased. Experimental results have shown that point defects, including oxygen vacancy V_O, gap oxygen atom, gallium vacancy V_Ga and gap atom can be used as capture and recombination centers to reduce the collection efficiency of electric aurora carriers. Therefore, degradation of PDCR was caused. ^18,19 Combined with the changes in the photo-to-dark current ratio under different lighting conditions, the reason that the photo-to-dark current ratio at 254-nm light illumination decreased with the increase of temperature was that the point defects increased with the temperature.

The temperature-dependent current-time (I–T) response of the β-Ga₂O₃ photodetector at 10 V under different UV light illumination conditions is shown in Fig. 5. The response-time characteristics of the β–Ga₂O₃ photodetectors at a temperature of 40 K–300 K by on/off switching under an applied bias of 10 V and under 365-nm and 254-nm light illumination are represented in Fig. 5. It can be observed that temperature has a considerable impact under different UV light illumination conditions. With the change of temperature, the response time and current of the photo will change correspondingly

Zoom In Zoom Out Reset image size

For a more detailed comparative study of response times, quantitative analysis of the current rise and decay process involves the following biexponential relaxation equation to fit the photoresponse curve. ²⁰

$$\begin{eqnarray}&&I={I}_{0}+C{e}^{-\displaystyle \frac{t}{{\tau }_{1}}}+D{e}^{-\displaystyle \frac{t}{{\tau }_{2}}}\end{eqnarray} \tag{ 3 }$$

where I₀ is the steady-state photocurrent; t is the time; C and D are the constant; ${\tau }_{1}$ and ${\tau }_{2}$ are two relaxation time constants. This equation indicates that there are two different components in the current rise and decay processes, and the magnitude of C and D means the proportion of each component. The time corresponding to a larger C and D is taken as the rise time and fall time of the β-Ga₂O₃ photodetector. Two time constants indicate that two different mechanisms are functioning during the current rise and decay process, which has been reported in many studies. ²¹ The first one is a fast band-to-band recombination mechanism in the bulk. It has a smaller time constant, and it can be used to measure the light response speed of the detector. The second one is attributed to the presence of surface states in the device. The surface states act as traps that slow down the response of the detector to the light pulse. This mechanism has a much larger time constant. ²² The rise and decay process of the β-Ga₂O₃ photodetector's photoresponse under different illumination conditions are fitted with formula 3. The smaller time constant in the rising process is denoted as ${\tau }_{r},$ and the smaller time constant in the decay process is denoted as ${\tau }_{d}.$ Then, the rise time and fall time of each temperature point were fitted by the first-order linear function, and the results are shown in Fig. 6.

Zoom In Zoom Out Reset image size

It can see from Fig. 6 that the slopes of the first-order linear fitting curves are all negative, indicating that the photoresponse time of the β-Ga₂O₃ photodetector decreases with the increase of temperature, no matter under 365-nm or 254-nm illumination, indicating that the higher the temperature, the faster the photoresponse rate. To further analyze the reason why the fast response time of the β-Ga₂O₃ photodetector decreased with the increase in temperature, the photoluminescence (PL) was exploited under the temperature from 40 K to 300 K.

The excitation photo of 365-nm and 254-nm illumination was used to irradiate the β-Ga₂O₃, and the PL spectrum form is shown in Fig. 7. As the temperature increased, the PL intensity of both excitation photos decreased significantly. PL intensity refers to the number of photons emitted by the device at the corresponding energy. A higher PL intensity indicates that more photons are excited by the corresponding energy and there will be fewer defects. Therefore, the result in Fig. 7 shows that the number of defects in the β-Ga₂O₃ photodetector increased with the temperature.

Zoom In Zoom Out Reset image size

To determine which defect in β-Ga₂O₃ changes with temperature, this study adopted the Alentsev-FOCK method ²³ to deconvolute the PL spectra shown in Fig. 7, and two representative temperature points of 40 K and 300 K were selected. The result is shown in Fig. 8. Under 365-nm excitation, the PL spectra were deconvoluted into three PL bands. At room temperature, the PL bands were centered at around 466 nm (∼2.64 eV), 527 nm (∼2.34 eV), and 570 nm (2.16 eV) (Fig. 8a). At the temperature of 40 K, the PL bands were centered at around 441 nm (∼2.79 eV), 467 nm (∼2.64 eV), and 511 nm (2.41 eV) (Fig. 8b). Under 254-nm excitation, the PL bands were centered at around 399 nm (∼3.09 eV), 438 nm (∼2.82 eV), and 534 nm (2.31 eV) (Fig. 8c) at room temperature. At the temperature of 40 K, the PL bands were centered at around 395 nm (∼3.12 eV) and 427 nm (∼2.89 eV) (Fig. 8d).

Zoom In Zoom Out Reset image size

In fact, the meaning of the deconvoluted PL bands is considered to be related to the recombination of a bound exciton with the hole trapped at one of the defects in β-Ga₂O₃ ²⁴ Specifically, the band peaking at around 3.0 eV is ascribed to (V_Ga+V_O) disvacancies in the (1−) charge state, whereas the 2.7 eV PL emission correlates well with V_Ga with (2−) charge state. Similarly, the 2.4 eV PL emission is attributed to neutral oxygen interstitial (O_i ⁰) defects. ²⁵ Therefore, under 365-nm excitation, the defect was V_Ga ²⁻ at the temperature of 40 K. When the temperature increased gradually, the defect of O_i ⁰ occurred. Under 254-nm excitation, the defects were V_Ga ²⁻ and (V_Ga+V_O)¹⁻ at the temperature of 40 K. The defect of O_i ⁰ occurred when the temperature increased.

The generation and recombination mechanism of photo-generated carriers of the photodetector is shown in Fig. 9, which mainly involves the following four processes. Process 1: Under UV excitation, electron-hole pairs go directly from the valence band to the conduction band. Process 2: electron-hole pairs go to the defect-conduction band and then go to the conduction band. Process 3: with the illumination off, the electrons created in the conduction band recombine with the holes through the recombination centers. Process 4: the electrons go directly to the valence band to recombine with the holes. Under 254-nm UV illumination, photo-generated carriers are mainly from process 1 and only a few from process 2. However, under 365-nm UV illumination, photo-generated carriers are only from process 2. This is why the PDCR at 254-nm UV illumination is much larger than the PDCR at 365-nm UV illumination. The influence of temperature on the β-Ga₂O₃ detector is mainly in process 2. As the temperature increased, the defects of O_i ⁰ increased, indicating that the number of indirect transitions would increase. This explains why the photoresponse time becomes shorter under 365-nm illumination. Since the direct transition process also includes a small part of process 2. The photoresponse time also becomes shorter under 254-nm illumination, and the variation magnitude is much smaller than that in the indirect transition. This can be seen from the fact that the slope of Fig. 6a is much larger than that of Fig. 6b.

Zoom In Zoom Out Reset image size

This paper conducted experiments to investigate the temperature-dependent properties of the β-Ga₂O₃ photodetector. The temperature of the experiment is from 40 K to 300 K. Under different light illumination conditions, when the temperature increased from 40 K to 300 K, the maximum current of the β-Ga₂O₃ photodetector was obtained at around 180 K, which is related to the carrier mobility of β-Ga₂O₃. Under 365-nm illumination, the photo-to-dark current ratio (PDCR) of the β-Ga₂O₃ photodetector increased from 0.25 to 2.56 when the temperature increased from 40 K to 300 K, and the photoresponse peak was at 160 K. Under 254-nm illumination, the PDCR of the photodetector decreased from 4986.63 to 609.29 when the temperature increased from 40 K to 300 K, and the photoresponse peak was at 220 K. These results indicated that as the temperature increased, the release of impurities or traps in the bandgap led to an increase in the number of photo carriers generated under 365-nm illumination and a corresponding decrease in the number of photo carriers generated under 254-nm illumination. The change of the photoresponse peak indicated that the photodetector had a redshift at low temperatures. Meanwhile, the fast photoresponse time under 365-nm and 254-nm illumination decreased as the temperature increased. This indicated that the defect concentration of Ga₂O₃ increased gradually with the temperature, leading to a faster response time of the photodetector. Our research provides a theoretical basis for the application of β-Ga₂O₃ photodetectors in the space environment in the future.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 12075069, 61771167, 11775061, and 11805045), the key project of Science and Technology of Sichuan Province (Grant No. 2019YFSY0028), and the project of State Key Laboratory of Intense Pulsed Radiation Simulation and Effect (Grant Nos. SKLIPR1912and SKLIPR2015).