Study on the time-resolved detection performance of β-Ga₂O₃-based SBUV photodetectors: surface chemical analysis and the impacts of non-V_O factors

Abstract

The time-resolved detection performance of β-gallium oxide (β-Ga₂O₃)-based solar-blind ultraviolet (SBUV) photodetectors (PDs) are garnering extensive research attention. Generally, variations in the response and recovery rates of β-Ga₂O₃-based PDs have been attributed to oxygen vacancies (V_O), typically analyzed through X-ray photoelectron spectroscopy (XPS) O 1s spectra. However, in this study, we fabricated SBUV PDs with β-Ga₂O₃ films grown via metal–organic chemical vapor deposition (MOCVD). The surface chemical analysis of the obtained films reveals that the XPS O 1s spectra are unaffected by V_O. Additionally, by correlating film properties with PD performance, we demonstrate the significant role of non-V_O factors in shaping the time-resolved detection performance of these devices.

---

Introduction

SBUV, ranging from 200 to 280 nm, is notably devoid of background noise, making it exceptionally valuable for the detection of missiles, the prevention of fires, and the monitoring of ozone layer depletion.¹ β-Ga₂O₃ is celebrated for its intrinsic solar-blindness, outstanding thermal and chemical stability.^2,3 Moreover, the large-scale production of β-Ga₂O₃ crystals is feasible through both melt growth and epitaxy techniques,^4,5 which further cements its position as a material of choice for SBUV detection over other contenders such as Zn_xMg_1−xO, Al_xGa_1−xN, and diamond.⁶ Besides, β-Ga₂O₃ holds immense promise for applications in high-power devices and gas sensing technologies.^7,8

The quintessential PDs should embody high sensitivity, high signal-to-noise ratio, high spectral selectivity, high speed, and remarkable stability.⁹ However, the intrinsic persistent photoconductivity (PPC) of β-Ga₂O₃ poses a challenge to achieving the high-speed demands. In recent years, a surge of literature has focused on the time-resolved detection performance of β-Ga₂O₃-based PDs. Notably, Vu et al. have shown that enhancing the oxygen partial pressure during pulsed laser deposition processes can mitigate the PPC of β-Ga₂O₃-based PDs,¹⁰ Wang et al. dissected the impact of post-annealing on the response and recovery speeds of β-Ga₂O₃-based PDs,¹¹ while Deng et al. illustrated that surface plasma treatment can reduce the PPC of β-Ga₂O₃-based PDs.¹² Furthermore, the works of Chen and Li et al. have brought to light the potential of PPC in β-Ga₂O₃ for applications in brain-like artificial intelligence fields.^13,14 Collectively, current research predominantly attributes variations in time-resolved detection performance to oxygen vacancies (V_O).^{10–13,15–19} The elongation of response and recovery times, ascribed to V_O, is a trait shared by many oxide materials like tin dioxide and zinc oxide.^20–22 Compared to its oxide counterparts, β-Ga₂O₃ exhibits unique features, such as its ultra-wide indirect bandgap, suggesting more intricate mechanisms at play in its time-resolved response characteristics.² Additionally, XPS O 1s spectra are widely regarded as a key method for characterizing V_O, yet the interpretations of O 1s spectra remain a contentious issue to date.^23–26

In this paper, we fabricated SBUV PDs using β-Ga₂O₃ films grown by MOCVD. By fine-tuning the growth parameters of the β-Ga₂O₃ thin films, we delve into the influence of non-V_O factors on the time-resolved detection performance of the PDs and explore the correlation between the XPS O 1s spectra and V_O.

Experimental

Epitaxial growth

To meticulously analyze the relationship between the XPS O 1s spectra and V_O, the O/Ga flow ratio (VI/III ratio) was selected as the variable of interest. The epitaxial growth was conducted using a modified Emcore D180 MOCVD system, with 2-inch c-plane sapphire wafers as substrates. High-purity oxygen (O₂) and trimethylgallium (TMGa) served as the reactants. TMGa was transported by high-purity argon (Ar) through a bubbling method, with the pressure and temperature inside the bubbler maintained at 1200 mbar and 1 °C, respectively. The substrate temperature and growth pressure were controlled at 750 °C and 40 mbar, with the growth process lasting for a duration of 0.5 h. The flux of the reactants and the corresponding VI/III ratios are detailed in Table 1. The calculation of the VI/III ratios is based on the following equations:

P_MO = exp(a − b/T) (1)

n_Ga = F_Ar × P_MO/[V_m × (P_bub − P_MO)] (2)

n_O = 2F_O/V_m (3)

where P_MO denotes the vapor pressure of the organometallic source, with a and b representing constants 8.07 and 1703, respectively. T signifies the absolute temperature, n_Ga refers to the molar flow of TMGa, F_Ar is the flow rate of Ar, V_m is the molar volume of 22 414 cm³ mol⁻¹, P_bub is the pressure inside the stainless steel bubbler, n_O is the molar flow of O atoms, and F_O represents the flow rate of O₂.²⁷

Table 1 The calculated VI/III ratios for different samples

Sample	Flow of O2 (sccm)	Flow of Ar (sccm)	VI/III ratio (×10^3)
S1	200	10	5.6
S2	300	10	8.4
S3	400	10	11.2
S4	500	10	13.9
S5	600	10	16.7

---

Fabrication of SBUV PDs

The planar metal–semiconductor–metal (MSM) type SBUV PDs were fabricated utilizing the epitaxial β-Ga₂O₃ films. The Ti (15 nm)/Au (50 nm) electrodes were prepared by standerd lift-off process. For better metal/β-Ga₂O₃ contacts, an annealing process at 600 °C for a duration of 3 min was implemented.

Characterizations

The crystalline properties, surface morphology, and optical characteristics of the epitaxial films were characterized using X-ray diffraction (XRD, Rigaku, Ultima IV), atomic force microscopy (AFM, Veeco), a UV-vis spectrophotometer (Shimadzu, UV1700), and room-temperature photoluminescence (PL, Horiba iHR550 spectrometer), respectively. XPS (Thermo Fisher Scientific, Nexsa) was employed to investigate the surface chemical composition of the samples. The performance of the fabricated PDs was assessed using a Keysight B2902A precision resource/measurement unit. The spectral response was captured under a variable-wavelength xenon lamp.

Results and discussion

Crystalline properties

The crystalline properties of the β-Ga₂O₃ films were characterized using XRD. Fig. 1a displays the XRD 2θ-ω scan patterns for S1–S5, revealing only the peaks associated with the sapphire substrates and the β-Ga₂O₃ epi-layers (JCPDS no. 43-1012, ICSD 78-2426). Notably, the β-Ga₂O₃ peaks for sample S1 are significantly weaker than those of the other samples, indicating its inferior crystalline quality. Fig. 1b presents the rocking curves of the (2̄01) reflection for the β-Ga₂O₃ samples. The full width at half maximum (FWHM) of these curves suggests that samples S2 through S5 can be grouped together, with minimal variation in FWHM, indicative of similar crystalline qualities. In contrast, sample S1 exhibits a substantially higher FWHM, pointing to a higher density of dislocations and small-angle grain boundaries within the film.²⁸ The XRD data imply that a lower supply of O₂ adversely affects the epitaxial growth of β-Ga₂O₃ films.

Fig. 1 XRD patterns of the β-Ga₂O₃ films grown at different VI/III ratios: (a) 2θ-ω scan; (b) rocking curves.

Surface morphology

The AFM pictures of the β-Ga₂O₃ films is illustrated in Fig. 2. S1 displays a relatively smooth surface with a root-mean-square (RMS) roughness of merely 3.2 nm. As the VI/III ratio increases, the emergence of more hillocks is observed, along with a corresponding increase in RMS roughness. Notably, at the VI/III ratio of 16.7 × 10³, a proliferation of micro-grains becomes apparent, and the RMS roughness surges to 16.0 nm. This alteration in surface morphology is attributed to an enhanced nucleation density. The nucleation density at the initial stage of epitaxial growth can be calculated using the following formula:

N = A exp(−ΔG/RT) (4)

where N symbolizes the nucleation density, A is a constant, ΔG represents the activation free energy for nucleation, R is the Boltzmann constant, and T denotes the thermodynamic temperature of the substrate. ΔG is composed of the bulk free energy ΔG_V and the surface free energy ΔG_S, with ΔG_V being inversely related to the partial pressure of O₂, and ΔG_S being inversely related to the O₂ concentration.^29–31 In this study, the O₂ flow was increased to improve the VI/III ratio, leading to an increase in both the partial pressure and concentration of O₂. Consequently, the nucleation density increased with the enhanced VI/III ratio, resulting in both the increased number of hillocks and the rise in RMS roughness.

Fig. 2 AFM pictures of the β-Ga₂O₃ films grown at different VI/III ratios.

Surface chemical composition

To delve into the surface chemical composition of the fabricated β-Ga₂O₃ films, XPS was employed. To erase the effects of sample charging, all binding energies presented were calibrated using the C 1s peak at 284.6 eV, originating from surface contamination. Fig. 3a shows the full scan of S1, detecting signals solely related to Ga, O, and C. Fig. 3b details the O 1s spectra for samples S1 to S5, exhibiting pronounced shoulders and tailings on the high-energy side. A deconvolution of the spectra was performed using Voigt line-shape and Shirley baseline, with the FWHM of each component held equal. The O 1s spectra could be deconvoluted into three distinct components: a major component at 530.3 eV (O_I), a subordinate one at 531.9 eV (O_II), and a minor one at 532.9 eV (O_III). O_I and O_III can be distinctly attributed to Ga–O bonding and chemisorbed species, respectively, while the assignment of O_II remains a subject of debate.^23–26 In prior researches on the time-resolved detection performance of β-Ga₂O₃-based PDs, O_II has commonly been assigned to O²⁻ ions in oxygen-deficient regions due to V_O,^{10–13,15,16} even some directly attribute O_II to V_O.^{17–19,32,33} However, such interpretations have notable limitations. Firstly, O_II signals originate from the inner electrons of O atoms, but vacancies lack the electrons, thus precluding O_II from arising directly from V_O. Additionally, from S2 to S5, the O₂ flow was doubled, but the proportion of O_II increased from 9.8% to 13.8%, which is clearly in conflict with the view that O_II originates from oxygen-deficient regions due to V_O. Furthermore, V_O should exert direct effects on Ga but O spectra, yet the peaks of Ga 2p_1/2 and Ga 2p_3/2, as shown in Fig. 3c, are able to be deconvolute into a single component attributable to Ga–O bonding, with no correlation to O_II. Therefore, O_II is suggested to be independent of V_O. Based on our previous researches, O_II can be ascribed to surface lattice O atoms.^34,35 The binding energy for these O atoms at β-Ga₂O₃ surface lattice is elevated due to dangling bonds, leading to the formation of high-energy shoulders in the XPS spectra. Similarly, due to the absence of coordinated Ga atoms, such lattice O atoms also exist at defects like dislocations and grain boundaries. The proportions of O_II and the RMS roughness of the films show a strong correlation; as RMS roughness increases, the specific surface area of the samples enlarges, leading to an increase in the proportion of surface lattice O atoms. The only exception is sample S1, which has the lowest surface roughness but a high proportion of O_II. According to the XRD findings, more dislocations and small-angle grain boundaries exist within S1, contributing to the high proportion of O_II. After excluding O_III (chemisorbed species), the O/Ga atom ratios in the lattice calculated from O 1s and Ga 3d for samples S1 through S5 are 1.52, 1.50, 1.51, 1.50, and 1.47, respectively, closely matching the theoretical stoichiometric ratio of 1.5, with minor variations attributed to experimental error (within 5%). Despite an increase in O₂ flow from 200 sccm to 600 sccm, all corresponding VI/III ratios exceed 5000, indicating that all films were grown in significantly O-rich atmospheres. Thus, variations in O₂ flow had minimal impact on the O/Ga atom ratios. Additionally, certain studies have attributed O_II to C–O bonding.^36,37 However, attributing O_II to C–O bonds would require excluding it from the calculation of the O/Ga atom ratios in the β-Ga₂O₃ lattice; accordingly, the O/Ga atomic ratios for S1–S5 would be 1.31, 1.35, 1.34, 1.34, and 1.26, respectively. This inference, suggesting an increase in V_O despite a doubling of O₂ supply from S2 to S5, lacks both theoretical and experimental plausibility.

Fig. 3 XPS spectra of the β-Ga₂O₃ films grown at different VI/III ratios: (a) full scan; (b) O 1s; (c) Ga 2p.

Optical transmittance spectra

Fig. 4a displays the optical transmission spectra of the deposited β-Ga₂O₃ films. All samples exhibit sharp absorption edges ranging from 230 nm to 270 nm. Fig. 4b illustrates the optical bandgaps (E_gs) of the films, calculated using Tauc's plot method.³⁸ The E_gs of the β-Ga₂O₃ films are consistently 5.0 eV, underscoring their promising potential for SBUV PD applications.

Fig. 4 (a) Optical transmittance spectra of the β-Ga₂O₃ films grown at different VI/III ratios; (b) optical bandgaps of the films calculated by Tauc's plot method.

Luminescent propetries

Fig. 5a displays the PL spectra of S1–S5, acquired using a 240 nm (5.17 eV) laser as the excitation source. The spectra can be broadly delineated into three components: the self-trapped exciton (STE) recombination emission centered around 3.4 eV, the donor–acceptor pair (DAP) recombination emission near 3.1 eV, and a laser doubling peak at 2.58 eV (480 nm).³⁹ Consistent with the findings of the XRD analysis, the luminescence intensities of S2–S5 spectra exhibit close similarity, whereas the luminescence intensity of S1 is markedly diminished. This disparity can be ascribed to the inferior crystalline quality of S1, a large number of defects that serve as non-radiative recombination centers exist within the film.³⁸ To gain a more detailed understanding of these PL spectra, a deconvolution using a Gaussian line-shape was carried out. Fig. 5b illustrates the deconvolution for the spectrum of S1, which is composed of four distinct emission peaks at 3.8 eV (STE1), 3.4 eV (STE2), 3.0 eV (DAP1), and 2.7 eV (DAP2). Based on the research conducted by Ho et al., STE1 and STE2 originate from distinct exciton recombination within β-Ga₂O₃. Regarding DAP1 and DAP2, Ga interstitial (Ga_i) functions as the donor, while Ga–O vacancy pairs (V_Ga + V_O) and Ga vacancies (V_Ga) act as acceptors, respectively.⁴⁰ The deconvolution for the spectra of S2–S5 are provided in Fig. S1 (ESI†). Table 2 outlines the proportion of each emission peak. The results indicate that STE2 and DAP1 are the primary contributors to the spectra, with their proportions exceeding 34.0% each. The propotions of different emission peaks across each sample exhibits slight variations and does not display a discernible pattern.

Fig. 5 (a) The PL spectra of S1–S5; (b) the deconvolution for the PL spectrum of S1.

Table 2 The proportion of each emission peak for S1–S5

Sample	STE1 (%)	STE2 (%)	DAP1 (%)	DAP2 (%)
S1	3.6	38.8	36.2	21.3
S2	6.3	34.9	36.8	22.0
S3	4.5	36.3	37.9	21.3
S4	4.2	36.4	39.8	19.6
S5	3.4	36.4	38.1	22.0

---

SBUV photodetectors

MSM PDs were fabricated using the obtained β-Ga₂O₃ films, the schematic diagram of the PDs is shown in Fig. 6a.

Fig. 6 (a) The schematic diagram of the SBUV PDs; (b) the I₂₅₄ and I_dark of the PDs; (c) the I₂₅₄ measured at 20 V bias of the PDs.

I _dark and I₂₅₄. The current–voltage (I–V) characteristics of the PDs were measured in the dark and under 254 nm illumination, with an illumination power density of 1.75 mW cm⁻². The resulting I–V curves are shown in Fig. 6b. All PDs exhibit I₂₅₄/I_dark ratios on the order of 10⁵. The I₂₅₄ of each PD at 20 V bias are illustrated in Fig. 6c. The I₂₅₄ of device fabricated with S1 (DS1) is significantly lower the the others. Due to the inferior crystalline quality, DS1 harbors an increased number of defects which function as recombination centers. This results in the quenching of photo-generated carriers and leads to the diminished photo response of the PD.⁴¹

Time-resolved detection performance. The time-resolved detection performance of the fabricated PDs was measured under 254 nm illumination and 20 V bias. As depicted in Fig. 7a, all the PDs demonstrate excellent repeatability and stability, with the current rapidly responding to the toggling of SBUV irradiation. Table 3 presents the 10–90% rise time (t_r) and the 90–10% decay time (t_d) for each PD, averaging over five cycles. Two distinct phenomena are observed: (i) from DS1 to DS5, the t_r progressively increases, and (ii) the t_d values are relatively short, fluctuating around 0.1 seconds without a pronounced PPC effect. To ensure the reliability of these phenomena, we prepared two additional PDs with the same structure for each β-Ga₂O₃ film and conducted repetitive experiments. The experimental results have verified the reliability of the phenomena.

Fig. 7 (a) The time-resolved photocurrent response of DS1–DS5 measured under 254 nm illumination and 20 V bias; (b) the related t_r and the RMS roughness of the β-Ga₂O₃ films grown at different VI/III ratios.

Table 3 The average t_r and t_d of the SBUV PDs

PD	t r (s)	t d (s)
DS1	0.272	0.116
DS2	0.351	0.110
DS3	0.406	0.087
DS4	0.393	0.107
DS5	0.417	0.078

---

The prolongation of t_r or t_d is generally attributed to an increase in V_O.^{10–12,15–19} However, from S1 to S5, despite tripling the O₂ flow, the O/Ga atom ratio varied minimally, indicating that V_O is not the cause of the increased t_r. We believe that the increase in surface trap states leads to the extension of t_r. Dangling bonds in surface atoms are capable of forming surface trap states, which trap carriers and extend t_r.^42–44 In this study, the films were grown under extremely O-rich conditions, and the PDs performance was tested under ambient conditions. Therefore, the dangling bonds on the surface Ga atoms can be readily passivated by O₂, suggesting that surface traps are primarily composed of dangling bonds at the surface O atoms. With an increase in the VI/III ratio, the roughness of the obtained films increased, as shown in Fig. 7b, which in turn increased the surface area and the number of surface trap states.

Regarding phenomenon ii, to better analyze the recovery behavior of the PDs, we compare the performance of DS3 with that of our reported PD based on β-Ga₂O₃ homepitaxial film (DH).⁴⁵ The FWHM of the XRD rocking curve for the homoepitaxial β-Ga₂O₃ film is less than 0.007°. As shown in Table 4, the t_r and t_d of DS3 are significantly shorter than those of DH. The growth temperatures, pressures, and VI/III ratios during the growth processes of S3 and the homoepitaxial film are identical, and the O/Ga atom ratio of the homoepitaxial film is 1.51 (Fig. S2 in ESI†), hence, it is unreasonable to assume a higher concentration of V_O in DH. Therefore, we believe that there are additional factors influencing the PPC effect of β-Ga₂O₃-based PDs, beyond V_O. Moreover, in our another study involving β-Ga₂O₃ films grown by MOCVD on sapphire wafers, with the XRD rocking curve FWHM of the β-Ga₂O₃ film decreased from 1.85° to 1.40°, yet the corresponding t_d of the PDs increased from 0.061 s to 0.537 s. These findings indicate that as the crystalline quality of the β-Ga₂O₃ film improves, the corresponding PD's PPC effect also intensifies. We attribute this to the intrinsic band structure characteristics of β-Ga₂O₃. Research on the band structure has indicated that β-Ga₂O₃ can be considered an indirect semiconductor in terms of recombination.^2,5,46,47 As a result, the transition of electrons from the conduction band minimum (CBM) to the valence band maximum (VBM) requires the involvement of phonons, and the direct band-to-band recombination rate is sluggish. Consequently, the relatively swift indirect recombination processes mediated by recombination centers expedites the recovery of β-Ga₂O₃-based PDs.³⁹ As the crystalline quality improves, defects like dislocations, which function as recombination centers within the film, diminish.^48–50 This diminishment leads to a higher proportion of direct recombining, thereby resulting in strong PPC. From this vantage point, the PPC of β-Ga₂O₃ is governed by a distinct mechanism compared to the commonly reported PPC of its amorphous countrapart, given that no existing studies have suggested the presence of an indirect bandgap in amorphous GaO_X.^51–53 Furthermore, although high crystalline quality can augment the PPC, it often results in higher photoelectric responsivity. Since both high photoelectric response and high response/recovery speed are crucial for an outstanding PD, achieving a balance between the two remains a challenge. Introducing a small number of deep-level impurities through doping engineering without compromising film quality may offer an effective solution, akin to doping gold atoms in silicon.⁵⁴

Table 4 Comparison of the performance of DS3 and DH measured at 20 V bias

PD	t r (s)	t d (s)	Ref.
DH	13.43	14.01	36
DS3	0.406	0.087	This work

---

Photoresponse spectra. Fig. 8 illustrates the photoresponse spectra of DS5, measured at a 20 V bias. The PD exhibits excellent selectivity for SBUV irradiation. The cutoff wavelength is established at 280 nm, with a marked increase in responsivity observed from 280 nm to 250 nm. The responsivity at 250 nm (R₂₅₀) and 285 nm (R₂₈₅) are 1.75 × 10⁻¹ A W⁻¹ and 5.81 × 10⁻⁴ A W⁻¹, respectively, yielding a R₂₅₀/R₂₈₅ ratio of 301.1. Table 5 displays the comparson of the parameters for DS5 and a part of reported β-Ga₂O₃-based PDs. Compared with other devices, DS5 has advantages of swift recovery and high I_photo/I_dark.

Fig. 8 The photoresponse spectrum of DS5 measured at 20 V bias.

Table 5 Comparison of the parameters for DS5 and a part of reported β-Ga₂O₃-based PDs

Type	I photo/Idark	R (A W^−1)	Rejection	t r	t d	Ref.
β-Ga2O3/MgO/Nb:SrTiO3 heterojunction (APD)	/	5.9 × 10^5	/	12.4 ns	41.7 μs	55
ZnO/β-Ga2O3 core–shell (APD)	/	1.3 × 10^3	/	20 μs	42 μs	9
β-Ga2O3/GaN heterojunction	6.1 × 10^4	3.05@0 V	R 254/R400 = 5.9 × 10^3	0.018 s	/	56
β-Ga2O3/SiC heterojunction	63.31	0.18	/	0.65 s/7.8 s	1.73 s/15.22 s	57
Vertical Schottky	360	0.16	R 222/R350 > 10^4	0.5 s	0.5 s	58
MSM	/	1.45	R 254/R365 ∼ 10.8	0.58 s/32.93 s	1.2 s/32.86 s	59
MSM	4	210	R 232/R320 = 5 × 10^4	3.2 s	1.4 s	60
MSM	>10^5	1.05	/	4.5	2.2	61
MSM	∼10^5	1.75 × 10^−1	R 250/R285 = 301.1	0.417 s	0.078 s	This work

---

Conclusions

In summary, SBUV PDs were fabricated utilizing β-Ga₂O₃ films grown via MOCVD. Through surface chemical analysis, we demonstrate that the sub-intense component observed in the O 1s spectra is not associated with V_O, thereby rendering them inappropriate for characterizing V_O in the investigation of time-resolved detection performance of β-Ga₂O₃-based PDs. Regarding the time-resolved characteristics of the PDs, our comparative analysis between the PD performance and the film properties has shed light on the fact that, aside from V_O, the indirect bandgap of β-Ga₂O₃ also plays a pivotal role in the PPC of β-Ga₂O₃-based PDs, which distinguishes the PPC mechanism of β-Ga₂O₃ from that of its amorphous countrapart. Furthermore, we have identified that surface traps can lead to an extension of the t_r. Note that, due to the insignificant change in the concentration of V_O observed in this study, the discussion on the impact of V_O on the time-resolved response characteristics of β-Ga₂O₃-based PDs is not included in this study. This study is instrumental in guiding the optimization and design of β-Ga₂O₃-based SBUV PDs.

Author contributions

Zeming Li: conceptualization, methodology, writing – original draft, investigation, funding acquisition. Rensheng Shen: investigation. Wancheng Li: methodology, conceptualization. Teng Jiao: visualization. Yuchun Chang: funding acquisition. Hongwei Liang: funding acquisition. Xiaochuan Xia: supervision. Baolin Zhang: conceptualization, methodology.

Data availability

The data supporting this article have been included in the main text.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was funded by Liaoning province natural science foundation joint fund (doctoral research initiation project), Grant Number 2023-BSBA-028, the Fundamental Research Funds for the Central Universities, Grant Number DUT24BS008, Guangdong Basic and Applied Basic Research Foundation, Grant Number 2023A1515110064, the National Key Research and Development Program of China, Grant Number 2023YFB4503003, the National Natural Science Foundation of China, Grant Numbers 11975066 and 62104024, Dalian municipal bureau of science and technology, Grant Number 2020RT01, Special projects for industrial foundation reconstruction and high-quality development of manufacturing industry in 2022 (No. TC220A04A-49).

Notes and references

1. X. Hou, Y. Zou, M. Ding, Y. Qin, Z. Zhang, X. Ma, P. Tan, S. Yu, X. Zhou, X. Zhao, G. Xu, H. Sun and S. Long, J. Phys. D: Appl. Phys., 2020, 54, 043001.
2. J. B. Varley, J. R. Weber, A. Janotti and C. G. Van de Walle, Appl. Phys. Lett., 2010, 97, 142106.
3. Z. Liu, P.-G. Li, Y.-S. Zhi, X.-L. Wang, X.-L. Chu and W.-H. Tang, Chin. Phys. B, 2019, 28, 017105.
4. H. Kim, SN Appl. Sci., 2021, 4, 27.
5. Z. Li, T. Jiao, J. Yu, D. Hu, Y. Lv, W. Li, X. Dong, B. Zhang, Y. Zhang, Z. Feng, G. Li and G. Du, Vacuum, 2020, 178, 109440.
6. X. Chen, F.-F. Ren, J. Ye and S. Gu, Semicond. Sci. Technol., 2020, 35, 023001.
7. Y. J. Jeong, J. Y. Yang, C. H. Lee, R. Park, G. Lee, R. B. K. Chung and G. Yoo, Appl. Surf. Sci., 2021, 558, 149936.
8. T.-F. Weng, M.-S. Ho, C. Sivakumar, B. Balraj and P.-F. Chung, Appl. Surf. Sci., 2020, 533, 147476.
9. B. Zhao, F. Wang, H. Chen, Y. Wang, M. Jiang, X. Fang and D. Zhao, Nano Lett., 2015, 15, 3988–3993.
10. T. K. O. Vu, D. U. Lee and E. K. Kim, Nanotechnology, 2020, 31, 245201.
11. J. Wang, X. Ji, Z. Yan, S. Qi, X. liu, A. Zhong and P. Li, J. Alloys Compd., 2024, 970, 172448.
12. L. Deng, H. Hu, Y. Wang, C. Wu, H. He, J. Li, X. Luo, F. Zhang and D. Guo, Appl. Surf. Sci., 2022, 604, 154459.
13. X. Chen, W. Mi, M. Li, J. Tang, J. Zhao, L. Zhou, X. Zhang and C. Luan, Vacuum, 2021, 192, 110422.
14. X.-X. Li, G. Zeng, Y.-C. Li, Q.-J. Yu, M.-Y. Liu, L.-Y. Zhu, W. Liu, Y.-G. Yang, D. W. Zhang and H.-L. Lu, Nano Res., 2022, 15, 9359–9367.
15. L.-X. Qian, Z.-H. Wu, Y.-Y. Zhang, P. T. Lai, X.-Z. Liu and Y.-R. Li, ACS Photonics, 2017, 4, 2203–2211.
16. J. Wang, Y. Xiong, L. Ye, W. Li, G. Qin, H. Ruan, H. Zhang, L. Fang, C. Kong and H. Li, Opt. Mater., 2021, 112, 110808.
17. S. Feng, J. Li, L. Feng, Z. Liu, J. Wang, C. Cui, O. Zhou, L. Deng, H. Xu, B. Leng, X. Q. Chen, X. Jiang, B. Liu and X. Zhang, Adv. Mater., 2023, 35, 2308090.
18. L. Huang, Q. Feng, G. Han, F. Li, X. Li, L. Fang, X. Xing, J. Zhang and Y. Hao, IEEE Photonics J., 2017, 9, 1–8.
19. R. Xu, X. Ma, Y. Chen, Y. Mei, L. Ying, B. Zhang and H. Long, Mater. Sci. Semicond. Process., 2022, 144, 106621.
20. V. Brinzari, Appl. Surf. Sci., 2017, 411, 437–448.
21. M. Madel, F. Huber, R. Mueller, B. Amann, M. Dickel, Y. Xie and K. Thonke, J. Appl. Phys., 2017, 121, 124301.
22. Z. G. Yin, X. W. Zhang, Z. Fu, X. L. Yang, J. L. Wu, G. S. Wu, L. Gong and P. K. Chu, Phys. Status Solidi RRL, 2012, 6, 117–119.
23. Q. Cao, L. He, H. Xiao, X. Feng, Y. Lv and J. Ma, Mater. Sci. Semicond. Process., 2018, 77, 58–63.
24. P. Song, Z. Wu, X. Shen, J. Kang, Z. Fang and T.-Y. Zhang, CrystEngComm, 2017, 19, 625–631.
25. Z. Feng, L. Huang, Q. Feng, X. Li, H. Zhang, W. Tang, J. Zhang and Y. Hao, Opt. Mater. Express, 2018, 8, 2229–2237.
26. Y. J. Lee, M. A. Schweitz, J. M. Oh and S. M. Koo, Materials, 2020, 13, 434.
27. Z. Li, T. Jiao, D. Hu, Y. Lv, W. Li, X. Dong, Y. Zhang, Z. Feng and B. Zhang, Coatings, 2019, 9, 281.
28. C. Seitz, Z. G. Herro, B. M. Epelbaum, R. Hock and A. Magerl, J. Appl. Crystallogr., 2006, 39, 17–23.
29. C. Wu, D. Y. Guo, L. Y. Zhang, P. G. Li, F. B. Zhang, C. K. Tan, S. L. Wang, A. P. Liu, F. M. Wu and W. H. Tang, Appl. Phys. Lett., 2020, 116, 072102.
30. P. Bundle, Int. J. Electron., 1968, 24, 405–413.
31. X. Liu, X. Wu, H. Cao and R. P. H. Chang, J. Appl. Phys., 2004, 95, 3141–3147.
32. P. Li, H. Shi, K. Chen, D. Guo, W. Cui, Y. Zhi, S. Wang, Z. Wu, Z. Chen and W. Tang, J. Mater. Chem. C, 2017, 5, 10562–10570.
33. Z. Xi, L. Yang, Z. Liu, S. Yao, L. Shu, M. Zhang, S. Li, Y. Guo and W. Tang, J. Phys. D: Appl. Phys., 2023, 57, 085101.
34. Z. Li, T. Jiao, W. Li, Z. Wang, Y. Chang, R. Shen, H. Liang, X. Xia, G. Zhong, Y. Cheng, F. Meng, X. Dong, B. Zhang, Y. Ma and G. Du, Appl. Surf. Sci., 2024, 652, 159327.
35. W. Li, C. Shen, G. Wu, Y. Ma, Z. Gao, X. Xia and G. Du, J. Phys. Chem. C, 2011, 115, 21258–21263.
36. H. Cui, Q. Sai, H. Qi, J. Zhao, J. Si and M. Pan, J. Mater. Sci., 2019, 54, 12643–12649.
37. R. Chen, D. Wang, J. Liu, B. Feng, H. Zhu, X. Han, C. Luan, J. Ma and H. Xiao, Cryst. Growth Des., 2022, 22, 5285–5292.
38. D. Hu, S. Zhuang, Z. Ma, X. Dong, G. Du, B. Zhang, Y. Zhang and J. Yin, J. Mater. Sci.: Mater. Electron., 2017, 28, 10997–11001.
39. L. Binet and D. Gourier, J. Phys. Chem. Solids, 1998, 59(8), 1241–1249.
40. Q. D. Ho, T. Frauenheim and P. Deák, Phys. Rev. B, 2018, 97, 115163.
41. E. Liu, B. Zhu and J. Luo, The Physics of Semiconductors, 2011, pp. 95–102, 132–143.
42. B. Mallampati, S. V. Nair, H. E. Ruda and U. Philipose, J. Nanopart. Res., 2015, 17, 176.
43. L. Li, E. Auer, M. Liao, X. Fang, T. Zhai, U. K. Gautam, A. Lugstein, Y. Koide, Y. Bando and D. Golberg, Nanoscale, 2011, 3, 1120–1126.
44. X. Zhao, P. Tu, J. He, H. Zhu and Y. Dan, Nanoscale, 2018, 10, 82–86.
45. Z. Li, T. Jiao, W. Li, G. Deng, W. Chen, Z. Li, Z. Diao, X. Dong, B. Zhang, Y. Zhang, Z. Wang and G. Du, Opt. Mater., 2021, 122, 111665.
46. C. Janowitz, V. Scherer, M. Mohamed, A. Krapf, H. Dwelk, R. Manzke, Z. Galazka, R. Uecker, K. Irmscher, R. Fornari, M. Michling, D. Schmeißer, J. R. Weber, J. B. Varley and C. G. V. d Walle, New J. Phys., 2011, 13, 085014.
47. H. Peelaers and C. G. Van de Walle, Phys. Status Solidi B, 2015, 252, 828–832.
48. M. Albrecht, A. Cremades, J. Krinke, S. Christiansen, O. Ambacher, J. Piqueras, H. P. Strunk and M. Stutzmann, Phys. Status Solidi B, 1999, 216, 409–414.
49. M. Albrecht, H. P. Strunk, J. L. Weyher, I. Grzegory, S. Porowski and T. Wosinski, J. Appl. Phys., 2002, 92, 2000–2005.
50. Y. Ohno, T. Taishi and I. Yonenaga, Phys. Status Solidi A, 2009, 206, 1904–1911.
51. Z. Zhang, X. Zhao, X. Zhang, X. Hou, X. Ma, S. Tang, Y. Zhang, G. Xu, Q. Liu and S. Long, Nat. Commun., 2022, 13, 6590.
52. R. Li, W. Wang, Y. Li, S. Gao, W. Yue and G. Shen, Nano Energy, 2023, 111, 108398.
53. R. Zhu, H. Liang, S. Hu, Y. Wang and Z. Mei, Adv. Electron. Mater., 2021, 8, 2100741.
54. X. Meng and C. Kang, The Physics of semiconductors, 1993, pp. 211–221.
55. Q. Zhang, N. Li, T. Zhang, D. Dong, Y. Yang, Y. Wang, Z. Dong, J. Shen, T. Zhou, Y. Liang, W. Tang, Z. Wu, Y. Zhang and J. Hao, Nat. Commun., 2023, 14, 418.
56. D. Guo, Y. Su, H. Shi, P. Li, N. Zhao, J. Ye, S. Wang, A. Liu, Z. Chen, C. Li and W. Tang, ACS Nano, 2018, 12, 12827–12835.
57. Y. Qu, Z. Wu, M. Ai, D. Guo, Y. An, H. Yang, L. Li and W. Tang, J. Alloys Compd., 2016, 680, 247–251.
58. F. Alema, B. Hertog, P. Mukhopadhyay, Y. Zhang, A. Mauze, A. Osinsky, W. V. Schoenfeld, J. S. Speck and T. Vogt, APL Mater., 2019, 7, 022527.
59. S. Oh, Y. Jung, M. A. Mastro, J. K. Hite, C. R. Eddy, Jr. and J. Kim, Opt. Express, 2015, 23, 28300–28305.
60. 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.
61. Y. Li, D. Zhang, R. Lin, Z. Zhang, W. Zheng and F. Huang, ACS Appl. Mater. Interfaces, 2018, 11, 1013–1020.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc03701d

---