Annealing-free Ohmic contact of β-Ga₂O₃via nitrogen plasma treatment

Abstract

With the advent of the artificial intelligence era, semiconductor materials with exceptional performance are increasingly in demand. Beta-gallium oxide (β-Ga₂O₃) is a promising candidate for advanced semiconductor applications; however, its limited contact performance between metal electrodes remains a critical restriction that impedes its full utilization. In this study, high-quality Ohmic contacts were established through a direct nitrogen plasma treatment, effectively preserving the crystallinity of β-Ga₂O₃ while enhancing its electrical performance. The carrier mobility of β-Ga₂O₃ reached levels up to 76 cm² V⁻¹ s⁻¹, approximately four times greater than the previously reported ranges (15–20 cm² V⁻¹ s⁻¹). The significantly improved on/off ratio (1.7 × 10¹⁰) can suppress the device malfunction due to leakage current and resolve existing structural limitations. The results of our investigation provide insights for the advancement of β-Ga₂O₃ as a cutting-edge semiconductor material for ultra-low-scale and multi-output integrated circuits.

---

Introduction

With the emergence of the artificial intelligence (AI) era and the imminent rise of high power processors that include circuits and amplifiers, the extreme miniaturization of circuit patterns, driven by high-density and down-scale demands in modern semiconductors, requires efficient current flowing with circuits in even smaller contact areas.^1–3 The foundation for on-device AI networking and high-bandwidth operations lies in achieving high-performance electronic devices, which starts with highly efficient metal via – semiconductor channel contacts.^4–6 Despite the potential of a material, if effective signal transmission between the material and the input line is not achieved, latency occurs, thus undermining the fundamental functionality of the circuit.^7,8 In this regard, beta-gallium oxide (β-Ga₂O₃), which is considered as a promising channel material for high-power semiconductors and diverse communication applications, can facilitate a high-performance multi-input/output channel with excellent electron mobility, subthreshold swing (SS), and ideality factor, meeting the need for high-capacity data transfer (in larger GB/s).^9–11 However, despite the significant advancements in the study of the electrical properties of β-Ga₂O₃, research on its contact with metals (via or electrode) remains limited. Without rapid thermal annealing (RTA), the formation of contacts between β-Ga₂O₃ and metals has rarely been successful.¹² Even with the RTA process, the Schottky barrier between the metal and β-Ga₂O₃ contact is typically high, and the contact resistance remains elevated, often requiring additional thermal treatment or high-concentration doping in the contact area.^13,14 The RTA process can lead to challenges such as metal vaporization of the electrode and degradation from the dopant or oxygen penetration at the interface. These issues impose significant constraints, limiting the contact formation with β-Ga₂O₃ in the front-end process, as such contact formation is not feasible during the middle- and back-end process.¹⁵ Therefore, research on low-temperature, high-efficiency contact improvements is essential to unlock the full potential of on-device applications with β-Ga₂O₃. For the extensive application of β-Ga₂O₃ in devices, the conventional thermal treatment approach to enhance contact characteristics has limitations for both front-end and back-end process integration. Annealing-related issues such as metal vaporization (e.g., Ti) are significant, and achieving a process integration margin is possible when contact processes perform well in integrated circuits. To establish stable performance in small areas without thermal treatment, further research is necessary.¹⁶

β-Ga₂O₃, with its ultra-wide bandgap of 4.8 eV and a high breakdown electric field of 8 MV cm⁻¹, is an excellent material applied in various electronic devices like high-performance Schottky barrier diodes (SBDs) and metal-oxide semiconductor field-effect transistors (MOSFETs).^17–19 Recently, epitaxy and chemical vapor deposition (CVD) methods, which support commercialization, have enabled high-quality large-area growth of β-Ga₂O₃.^20–22 Its high power allowance and current capacity make it suitable for a wide range of applications, i.e., from high to low power.²³ β-Ga₂O₃ possesses fine radiation hardness, eliminating the need for extra radiation passivation and optimization, which are essential for semiconductor applications in the aerospace industry.^24,25 β-Ga₂O₃ is highly suitable as an on-chip channel material capable of handling vast amounts of data transfer, signal processing, and computation required by AI semiconductor devices. From the perspective of device reliability, which is essential for repetitive calculations and data operations, β-Ga₂O₃ demonstrates excellent bias temperature instability (BTI) reliability and high breakdown voltage, thus providing stable operational tolerance even under extreme external conditions compared to traditional silicon. Although a simple deposition process can enable contact formation with metals without mismatch issues, additional contact formation processes are necessary to enhance performance.

This study proposes high-quality contact formation in β-Ga₂O₃via straightforward nitrogen (N₂) plasma treatment at room temperature, without the need for thermal annealing. The nitrogen substitution mechanism effectively mitigates challenges associated with thermal treatment, while enabling the formation of complete contact. The nitrogen plasma-treated β-Ga₂O₃ nanoFETs with a small contact area exhibit high on/off ratios, high field-effect mobility, and low contact resistance. Through enhanced, high-quality contacts in small areas, β-Ga₂O₃ can emerge as an ideal on-device semiconductor material for next-generation multi-output/high-density data circuits.

Experimental details

A schematic of the β-Ga₂O₃ TLM device is shown in Fig. 1a. The β-Ga₂O₃ Sn-doped (001) surface orientation β-Ga₂O₃ substrates (Novel Crystal Technology, Japan) with an electron concentration of approximately 3 × 10¹⁸ cm⁻³ were cleaned with a sequential solvent cleaning process prior to device fabrication. The TLM device patterns were defined by standard photolithography, and N₂ plasma treatment was performed at room temperature at a flow rate of 10 sccm and an applied radio frequency power of 100 W for 30 s, using a plasma cleaner (Tergeo, Pie Scientific). The TLM electrodes were deposited on top of the N₂ plasma-treated β-Ga₂O₃ with e-beam evaporated Ti/Au (50/100 nm). To examine the feasibility of N₂ plasma treatment on a small contact area, a β-Ga₂O₃ nanoFET was fabricated using a mechanically exfoliated β-Ga₂O₃ flake. The β-Ga₂O₃ flake was dry-transferred onto SiO₂/p⁺-Si (300 nm/500 μm) with a back gate using transparent gel films (Gelpak) at 40 °C. The contact areas were defined by e-beam lithography followed by e-beam evaporation of a Ti/Au (50/100 nm) Ohmic electrode after N₂ plasma treatment.

Fig. 1 (a) Schematic of the transfer-length method (TLM) pattern fabricated on the β-Ga₂O₃ substrate with N₂ plasma treatment. (b) and (c) the I–V characteristics of the β-Ga₂O₃ TLM devices without and with N₂ plasma treatment, respectively.

Micro-Raman spectroscopy was employed to investigate the changes in the crystalline properties after the plasma treatment with a 532 nm diode-pumped solid-state laser (Omicron) under backscattering geometry. The differences in the thickness and surface roughness after the plasma treatment were measured using an AFM (NX10, Park System). XPS (Nexsa, ThermoFisher Scientific), with an Al Kα monochromator as the X-ray source, was performed to examine the changes of the chemical surface bonding states after the plasma treatment. The XPS data were calibrated using the reference C 1s peak (284.8 eV). Cross-sectional HR-TEM (JEM-ARM200F, JEOL) and EDS profiling were performed to investigate the structure of samples before and after the plasma treatment, and the specimens were prepared using the focused-ion beam technique (Helios 5UX). The electrical properties of the TLM and FET devices were analyzed using a semiconductor parameter analyzer (Agilent 4156B and 4155C, respectively).

Results and discussion

Fig. 1a shows the fabrication process of the designed transfer-length method (TLM) pattern on N₂ plasma-treated β-Ga₂O₃ substrates. The TLM pattern was designed with lengths of 5, 10, 15, 20, 25, and 30 μm between the electrodes. The pristine sample, used as a control, was fabricated by depositing only Ti/Au (50/100 nm) metal for contact without any additional processes. In contrast, the modified sample underwent N₂ plasma treatment at 100 W and a flow rate of 10 sccm for 30 s after photolithography for metal patterning, followed by simultaneous metal deposition as with the pristine sample. RF power conditions of 50–80 W resulted in inconsistent or incomplete Ohmic contact formation, and power levels of 120–150 W did not yield further improvement in electrical performance compared to that under 100 W conditions. Based on a systematic investigation of RF power conditions, 100 W was selected as the optimal plasma treatment condition. Fig. 1b and c show the current–voltage (I–V) characteristics of the TLM pattern without and with N₂ plasma treatment, respectively. The pristine sample exhibited Schottky behavior, achieving a current of 0.69 mA mm⁻¹ at 0.1 V with a length of 30 μm, whereas after plasma treatment, an Ohmic contact was formed, with the current level improved by approximately two orders of magnitude. Fig. S1 (ESI†) shows the fitted data for calculating contact resistance, with the plasma-treated sample demonstrating a specific contact resistance of 2.66 × 10⁻⁴ Ω cm². Ultra-wide bandgap materials such as β-Ga₂O₃ are prone to Fermi level pinning due to surface states or defects, making Ohmic contact formation challenging. Additionally, β-Ga₂O₃ contains native oxygen vacancies, which can cause instability in electrical properties.^26,27 Given the analogous atomic radius of nitrogen and oxygen, the incorporation or substitution of nitrogen into oxygen vacancies presents a viable and promising approach.^28,29 The passivation effect on the surface from N₂ plasma treatment possibly contributed to the reduction of oxygen vacancies and the formation of a clear metal/semiconductor interface, achieving an Ohmic contact.³⁰

To examine surface changes triggered by the ion bombardment from direct N₂ plasma, atomic force microscopy (AFM) measurements were conducted to assess thickness and the root mean square (RMS) roughness. Fig. 2a shows the thickness profile of the β-Ga₂O₃ flake before and after N₂ plasma treatment. The thickness before plasma treatment showed 271 nm, and even after plasma treatment, the identical thickness was maintained, confirming that no etching occurred due to ion bombardment. The RMS surface roughness of the β-Ga₂O₃ flake decreased from 1 nm to 0.7 nm, suggesting that smoothing was achieved through vacancy passivation rather than surface etching. Previous studies have reported reductions in contact resistance by forming oxygen vacancies through Ar-based reactive ion etching (RIE). The main difference lies in the mechanism of defect passivation through N₂ plasma treatment, where RMS roughness decreases, as opposed to defect generation in Ar-based RIE, which results in a sharp increase in RMS roughness.³¹ Defect generation through etching is challenging to control and can cause undesired damage to the device, whereas defect passivation can create a suitable surface state for Ohmic formation without causing damage. Fig. 2b shows the changes in the Raman shift before and after N₂ plasma treatment. The Raman peaks of pristine β-Ga₂O₃ (black line) show 10 vibrational modes located at 111.6, 169.8, 200, 320.5, 347, 417.3, 477.6, 630, 652, and 766 cm⁻¹, which are typical Raman vibrational modes of β-Ga₂O₃ and in good agreement with the existing literature.³² The Raman-active vibrational modes of β-Ga₂O₃ can be classified into high-frequency modes (700–500 cm⁻¹) associated with the stretching and bending of GaO₄ tetrahedra, mid-frequency modes (480–310 cm⁻¹) related to the deformation of GaO₆ octahedra, and low-frequency modes (below 200 cm⁻¹) corresponding to the vibrations and translations of GaO₄ tetrahedra and GaO₆ octahedra chains. After N₂ plasma treatment, the same 10 vibrational modes were observed, and no intensity changes or new peaks appeared, thus confirming that N₂ plasma treatment did not affect the crystallinity of β-Ga₂O₃. However, the 320 and 347 cm⁻¹ peaks in the mid-frequency range, corresponding to A⁽⁴⁾_g and A⁽⁵⁾_g, were blue-shifted to 321.7 and 348.7 cm⁻¹. The A⁽⁴⁾_g and A⁽⁵⁾_g phonon modes are associated with the bending vibration mode, and the observed blue-shift is attributed to changes in the bond angles within β-Ga₂O₃₀.³³ Considering the reference, this blue-shift is likely due to the reduction of point defects like oxygen vacancies and the substitution of larger N atoms with O atoms, which increases bond strength and shifts the Raman peaks.³⁴

Fig. 2 (a) Atomic force microscope (AFM) height profile and (b) Raman spectra of the β-Ga₂O₃ obtained before and after N₂ plasma treatment. (c) Comparison of the Raman spectra of the mid-frequency modes, shown before (black line) and after (red line) N₂ plasma treatment.

X-ray photoelectron spectroscopy (XPS) analysis was conducted to examine the surface chemical bonding states of β-Ga₂O₃ after N₂ plasma treatment. Fig. 3 shows the XPS data for the changes in the O 1s spectra of β-Ga₂O₃ samples treated with plasma at bias powers of pristine, 50, 100, and 150 W. The O 1s spectrum peak of the β-Ga₂O₃ film can be divided into three peaks: peak O(I) (530.7 eV) originates from Ga–O bonding, peak O(II) (531.3 eV) is attributed to O²⁻ ions in oxygen-deficient regions, and peak O(III) (532.4 eV) is related to chemisorbed hydroxyl and carbonate species.^35,36 The Gaussian peak O(II), related to oxygen deficiency, tends to decrease as plasma treatment power increases. This result indicates that N₂ plasma treatment is sufficient to achieve effective surface passivation of the β-Ga₂O₃ surface. Meanwhile, the Raman spectra shown in Fig. 2 confirm the absence of detectable lattice damage after plasma treatment. These findings suggest that the plasma-induced passivation occurs without introducing structural defects into the β-Ga₂O₃ lattice. Additionally, by calculating the area ratio of O(II)/(O(I) + O(II)), the trend in oxygen-related defects according to N₂ plasma bias power was confirmed. For plasma bias powers of pristine, 50, 100, and 150 W, and the area ratios of O(II)/(O(I) + O(II)) decreased progressively to 42.39%, 25.08%, 24.26%, and 22.58%, respectively. This indicates that oxygen vacancies are passivated by N₂ plasma, resulting in a reduction of these defects. Furthermore, in the XPS spectrum of Ga 2p, the Ga 2p^3/2 and Ga 2p^1/2 peaks tend to shift to lower binding energies as plasma power increases, indicating the formation of Ga–N bonds (Fig. S2, ESI†).^37,38

Fig. 3 X-ray photoelectron spectroscopy (XPS) spectra of the O 1s peak of plasma treated β-Ga₂O₃ substrates with bias powers of (a) pristine, (b) 50 W, (c) 100 W, and (d) 150 W.

To investigate the effects of N₂ plasma treatment on the β-Ga₂O₃ lattice and the interface with metal (Ti), high-resolution transmission electron microscope (HR-TEM) and energy dispersive spectroscopy (EDS) analyses were conducted. Fig. 4a shows a TEM image of the overall area at low magnification. The blue-dashed square marks the N₂ plasma-treated area, while the green dashed square indicates the untreated pristine area. In the plasma-treated region, a thin amorphous oxide layer (marked in white) is visible, whereas this layer is not observed in the pristine area. Fig. S3 (ESI†) shows a TEM image of a β-Ga₂O₃ sample with only metal deposition, without plasma treatment. An amorphous oxide layer was also formed in the pristine sample, which was noticeably thinner compared to the plasma-treated sample. Fig. 4b and c show high-magnification TEM images of the plasma-treated and untreated areas, respectively. Fig. 4b presents the interface between Ti and N₂ plasma-treated β-Ga₂O₃, in which the 3–4 layers of β-Ga₂O₃ located beneath the amorphous Ti maintain the original lattice structure. This structure is consistent with that of the untreated channel shown in Fig. 4c. However, one layer at the interface between Ti and β-Ga₂O₃ appears partially blurred, which is attributed to the formation of TiO_X resulting from the interaction between Ti and the O atoms in the passivated Ga₂O₃ layer during the Ti metal deposition process. To investigate the composition of the slightly blurred interfacial layer and the overlying Ti area, an EDS line profile was conducted (Fig. 4d). In Fig. 4d, the amorphous layer formed at the β-Ga₂O₃ interface appears in black and consists of Ti and O atoms, forming TiO_X. The formed TiO_X layer (approximately 3.6–3.9 eV for Ti₂O₃) has a relatively lower bandgap than β-Ga₂O₃ (approximately 4.8 eV), facilitating electron transport and aiding in Ohmic contact formation.³⁹ Additionally, a reduction in oxygen at the metal/semiconductor interface was observed, without diffusion of Ti atoms. This is likely because N₂ plasma treatment causes O and N substitution on the β-Ga₂O₃ surface, forming an oxygen layer that easily reacts with Ti. Fig. 4e–h shows EDS mapping images of N₂ plasma treated β-Ga₂O₃ regarding base, Ga, O, and Ti atoms, respectively. In the EDS mapping image of O atoms shown in Fig. 4g, oxygen has evidently disappeared from the β-Ga₂O₃ interface, and, consistent with the EDS line profile results, out-diffused into the Ti region. Fig. S4 (ESI†) shows the EDS mapping of pristine (untreated) β-Ga₂O₃. In Fig. S4 (ESI†), oxygen does not out-diffuse even after metal deposition, verifying that the formation of TiO_X is due to N₂ plasma treatment.

Fig. 4 Cross-sectional transmission electron microscopy (TEM) images of the (a) β-Ga₂O₃ structure showing both plasma-treated and untreated area, (b) interface of Ti/plasma-treated β-Ga₂O₃ (blue dash square), and (c) carbon-coated β-Ga₂O₃ (green dash square). (d) Energy dispersive spectroscopy (EDS) line profile across the interface between Ti and plasma-treated β-Ga₂O₃, along with the corresponding count per second. (e) High magnification TEM image of the interfacial region, where a TiO_X layer is formed. (f)–(h) EDS mapping of gallium, oxygen, and titanium corresponding to the region shown in (e).

Fig. 5 shows the electrical characteristics of a β-Ga₂O₃ MOSFET fabricated using the N₂ plasma treatment method. The β-Ga₂O₃ device was fabricated using a flake obtained by mechanical exfoliation. As shown in Fig. 5a, N₂ plasma treatment was applied only to the contact area of the exfoliated flake without annealing, creating a small contact area. The transfer and output characteristics of N₂ plasma-treated β-Ga₂O₃ MOSFET are shown in Fig. 5b and 5c, respectively. An extremely high on/off ratio of 1.7 × 10¹⁰ and a high field-effect mobility of 76 cm² V⁻¹ s⁻¹ were achieved, along with a low contact resistance of 13.1 kΩ μm calculated using the Y-function method. In modern semiconductor scaling down processes, the gate-induced drain leakage current (GIDL) poses a significant risk to the stable operation of transistors. This device maintains a leakage current level at the 100 fA level while achieving an extremely high on/off ratio of 1.7 × 10¹⁰ at V_DS = +20 V, and 8.8 × 10⁹ at V_DS = +5 V. Considering that the recommended and conventional value for existing silicon devices is 10⁶, and for β-Ga₂O₃ it does not exceed 10⁸, this demonstrates exceptionally high device stability.^40,41 In conventional semiconductor physics, the limited on/off ratio margin compels structures such as gate-all-around and buried channel array transistors with constructing heterojunction gates (like TiN and poly-Si) or creating deep shallow trenches. However, the record-high on/off ratio in our device eliminates the need for such impractical measures. Even in ultra-low-scale integration arrays, the high on-current can effectively prevent deterioration caused by GIDL characteristics caused by adjacent rows.

Fig. 5 Optical and electrical properties of the N₂ plasma-treated β-Ga₂O₃ field-effect transistor (FET). (a) Optical microscopy images of the fabricated β-Ga₂O₃ device, (b) V_GS–I_DS transfer curve, and (c) V_DS–I_DS output characteristics.

As shown in Fig. S5 and Table S1 (ESI†), the effect of N₂ plasma treatment on contact behavior was further evaluated under elevated temperature conditions. Consistent with typical temperature-dependent FET characteristics, the threshold voltage decreased, subthreshold swing increased, and both on-current and off-current increased with rising temperature. The improved contact behavior achieved through nitrogen plasma treatment appears to be largely independent of the thermal degradation of the channel itself, maintaining stable contact properties and sustaining increased on-current levels even under high-temperature operation. In addition, to evaluate the effect of additional annealing on the plasma-induced contact, transfer characteristics measured before and after rapid thermal annealing at 500 °C exhibited no significant changes (Fig. S6, ESI†). This indicates that the TiO_X layer formed on the nitrogen plasma-treated surface was already sufficient to achieve an Ohmic contact, and that additional TiO_X generated during annealing did not further enhance contact behavior.

Unlike conventional nitrogen ion implantation, which involves high-energy ion beams and thermal annealing, the nitrogen plasma treatment used in this study was performed at room temperature with low RF power (100 W), minimizing the possibility of lattice damage. Given the shallow penetration depth, the effect is limited to surface-level modifications such as trap passivation and oxygen vacancy compensation. The observed improvement in contact behavior is therefore attributed to surface state engineering rather than bulk doping or defect-induced charge trapping. This mechanism is further supported by AFM analysis and annealing results indicating stable and low-resistance contact formation.

Furthermore, the field-effect mobility, which is more than four times higher compared to that in conventional β-Ga₂O₃ devices, appears to result from the high carrier concentration in the contact area and low trap-induced scattering from passivated contact surface by N₂ plasma treatment.^42,43 This suggests that the previously achieved low mobility in β-Ga₂O₃ is attributed to poor contact quality rather than the intrinsic potential of β-Ga₂O₃. High mobility directly correlates with lower latency and faster operating speed, allowing the transmission of various signals with lower power, thereby expanding the range of electronic applications for β-Ga₂O₃. Despite the small contact area, the excellent performance of the device offers significant advantages in achieving larger margins in processes beyond contact formation in advanced low-scale applications. Additionally, the annealing-free process is highly compatible with various manufacturing processes, greatly enhancing process flexibility. The findings of our study can contribute to industrializing high-performance β-Ga₂O₃-based devices.

Conclusions

In summary, we proposed a novel annealing-free N₂ plasma treatment for achieving Ohmic contacts in β-Ga₂O₃ devices. This straightforward approach effectively reduced the contact resistance to 13.1 kΩ μm through a defect passivation mechanism. XPS analysis confirmed the modification of Ga–O and oxygen-related bonds induced by the N₂ plasma treatment, Raman spectroscopy demonstrated that the crystalline quality of the treated region was well preserved without any significant degradation. β-Ga₂O₃ nanosheet FETs incorporating the N₂ plasma-treated contacts exhibited an on/off ratio of 1.7 × 10¹⁰ and a field-effect mobility of 76 cm² V⁻¹ s⁻¹. The proposed highly effective and scalable approach to reducing contact resistance enables the full exploitation of the inherent potential of β-Ga₂O₃-based devices, optimizing their performance and enhancing their operational reliability.

Author contributions

Junghun Kim – investigation, visualization, formal analysis, writing – original draft. Hyoungwoo Kim – project administration, methodology. Woong Choi – formal analysis, methodology. Jihyun Kim – validation, resources. Dongryul Lee – conceptualization, supervision, data curation, writing – review & editing, and project administration.

Data availability

The data supporting the primary research of this article have been included within the main manuscript and as part of the ESI.† All referenced data are publicly available in the cited literature.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by the Korea Research Institute for Defense Technology Planning and Advancement (KRIT) grant funded by the Defense Acquisition Program Administration (DAPA) (KRIT-CT-22-046) and the K-Sensor Development Program (No. RS-2022-00154729), funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea)

Notes and references

1. C. X. Xue, Y. C. Chiu, T. W. Liu, T. Y. Huang, J. S. Liu, T. W. Chang, H. Y. Kao, J. H. Wang, S. Y. Wei, C. Y. Lee, S. P. Huang, J. M. Hung, S. H. Teng, W. C. Wei, Y. R. Chen, T. H. Hsu, Y. K. Chen, Y. C. Lo, T. H. Wen, C. C. Lo, R. S. Liu, C. C. Hsieh, K. T. Tang, M. S. Ho, C. Y. Su, C. C. Chou, Y. D. Chih and M. F. Chang, Nat. Electron., 2020, 4, 81–90.
2. J. Lee, P. Asenov, R. Rhyner, E. Kao, S. M. Amoroso, A. R. Brown, X. W. Lin and V. Moroz, IEEE Trans. Electron Devices, 2024, 71, 1893–1899.
3. L. Ye, Z. Wang, T. Jia, Y. Ma, L. Shen, Y. Zhang, H. Li, P. Chen, M. Wu, Y. Liu, Y. Jing, H. Zhang and R. Huang, Sci. China Inf. Sci., 2023, 66, 200407.
4. D. D. Sharma, R. Blankenship and D. Berger, ACM Comput. Surv., 2024, 56, 1–37.
5. A. Liu, X. Zhang, Z. Liu, Y. Li, X. Peng, X. Li, Y. Qin, C. Hu, Y. Qiu, H. Jiang, Y. Wang, Y. Li, J. Tang, J. Liu, H. Guo, T. Deng, S. Peng, H. Tian and T. L. Ren, Nano-Micro Lett., 2024, 16, 119.
6. S. Batool, M. Idrees, S. T. Han, V. A. L. Roy and Y. Zhou, Small, 2023, 19, 2206550.
7. Z. Zhou, J. Wei, Y. Luo, K. A. Clark, E. Sillekens, C. Deakin, R. Sohanpal, R. Slavík and Z. Liu, Nat. Electron., 2023, 6, 694–702.
8. S. Sadulla, Prog. Electron. Commun. Eng., 2024, 1, 13–18.
9. Y. Zhang, R. Zhu, W. Huo, H. Liang and Z. Mei, Nano Lett., 2024, 24, 14398–14404.
10. J. H. Kim, H. J. Lee, H. J. Kim, J. Choi, J. H. Oh, D. C. Choi, J. Byun, S. E. Ahn and S. N. Lee, Adv. Electron. Mater., 2024, 10, 2300863.
11. S. Sun, C. Wang, S. Alghamdi, H. Zhou, Y. Hao and J. Zhang, Adv. Electron. Mater., 2024, 2300844.
12. S. Wu, Z. Liu, H. Yang and Y. Wang, Crystals, 2023, 13, 1045.
13. K. Tetzner, A. Thies, P. Seyidov, T. S. Chou, J. Rehm, I. Ostermay, Z. Galazka, A. Fiedler, A. Popp, J. Würfl and O. Hilt, J. Vac. Sci. Technol., A, 2023, 41, 043102.
14. L. Q. Zhang, W. Q. Miao, X. L. Wu, J. Y. Ding, S. Y. Qin, J. J. Liu, Y. T. Tian, Z. Y. Wu, Y. Zhang, Q. Xing and P. F. Wang, Inorganics, 2023, 11, 397.
15. Z. Yu, X. Yang, X. Zhao, Y. Li, S. He and Y. Wang, J. Vac. Sci. Technol., B, 2024, 42, 062202.
16. S. G. Jung, J. K. Kim and H. Y. Yu, IEEE Trans. Electron Devices, 2022, 69, 930–935.
17. W. Li, D. Saraswat, Y. Long, K. Nomoto, D. Jena and H. G. Xing, Appl. Phys. Lett., 2020, 116, 192101.
18. K. Zeng, A. Vaidya and U. Singisetti, Appl. Phys. Express, 2019, 12, 081003.
19. S. J. Pearton, F. Ren, M. Tadjer and J. Kim, J. Appl. Phys., 2018, 124, 220901.
20. S. B. Anooz, R. Grüneberg, C. Wouters, R. Schewski, M. Albrecht, A. Fiedler, K. Irmscher, Z. Galazka, W. Miller, G. Wagner, J. Schwarzkopf and A. Popp, Appl. Phys. Lett., 2020, 116, 182106.
21. M. J. Tadjer, F. Alema, A. Osinsky, M. A. Mastro, N. Nepal, J. M. Woodward, R. L. Myers-Ward, E. R. Glaser, J. A. Freitas, A. G. Jacobs, J. C. Gallagher, A. L. Mock, D. J. Pennachio, J. Hajzus, M. Ebrish, T. J. Anderson, K. D. Hobart, J. K. Hite and C. R. Eddy, J. Phys. D, 2021, 54, 034005.
22. G. Seryogin, F. Alema, N. Valente, H. Fu, E. Steinbrunner, A. T. Neal, S. Mou, A. Fine and A. Osinsky, Appl. Phys. Lett., 2020, 117, 262101.
23. M. A. Mastro, A. Kuramata, J. Calkins, J. Kim, F. Ren and S. J. Pearton, ECS J. Solid State Sci. Technol., 2017, 6, 356–359.
24. A. Azarov, J. G. Fernández, J. Zhao, F. Djurabekova, H. He, R. He, Ø. Prytz, L. Vines, U. Bektas, P. Chekhonin, N. Klingner, G. Hlawacek and A. Kuznetsov, Nat. Commun., 2023, 14, 4855.
25. J. Kim, S. J. Pearton, C. Fares, J. Yang, F. Ren, S. Kim and A. Y. Polyakov, J. Mater. Chem. C, 2019, 7, 10–24.
26. J. B. Varley, J. R. Weber, A. Janotti and C. G. Van de Walle, Appl. Phys. Lett., 2010, 97, 142106.
27. M. D. McCluskey, J. Appl. Phys., 2020, 127, 101101.
28. G. Li, A. Abliz, L. Xu, N. André, X. Liu, Y. Zeng, D. Flandre and L. Liao, Appl. Phys. Lett., 2018, 112, 253504.
29. H. Wang, H. Zhan, Y. Zhou, Y. Wu, X. Chen, H. Wang and J. Kang, Nanoscale, 2016, 8, 4381–4386.
30. K. M. Niang, B. C. Bayer, J. C. Meyer and A. J. Flewitt, Appl. Phys. Lett., 2017, 111, 122109.
31. P. Kachhawa, A. K. Jain and N. Chaturvedi, Phys. Status Solidi A, 2023, 220, 2200722.
32. C. Kranert, C. Sturm, R. Schmidt-Grund and M. Grundmann, Sci. Rep., 2016, 6, 35964.
33. K. Zhang, Z. Xu, S. Zhang, H. Wang, H. Cheng, J. Hao, J. Wu and F. Fang, Phys. B, 2021, 600, 412624.
34. R. Rao, A. M. Rao, B. Xu, J. Dong, S. Sharma and M. K. Sunkara, J. Appl. Phys., 2005, 98, 094312.
35. X. Hou, X. Zhao, Y. Zhang, Z. Zhang, Y. Liu, Y. Qin, P. Tan, C. Chen, S. Yu, M. Ding, G. Xu, Q. Hu and S. Long, Adv. Mater., 2022, 34, 2106923.
36. 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.
37. J. Wang, X. Ji, S. Qi, Z. Li, Z. Yan, M. Li, X. Yan, A. Zhong, C. Lu, X. Qi and P. Li, J. Mater. Chem. C, 2023, 11, 6202–6211.
38. Y. Zhang, X. Chen, Y. Xu, H. Gong, Y. Yang, F. F. Ren, B. Liu, S. Gu, R. Zhang and J. Ye, ACS Appl. Electron. Mater., 2020, 2, 808–816.
39. J. Bae, H. Y. Kim and J. Kim, ECS J. Solid State Sci. Technol., 2017, 6, 3045–3048.
40. N. Moser, J. McCandless, A. Crespo, K. Leedy, A. Green, A. Neal, S. Mou, E. Ahmadi, J. Speck, K. Chabak, N. Peixoto and G. Jessen, IEEE Electron Dev. Lett., 2017, 38, 775–778.
41. J. Kim, M. A. Mastro, M. J. Tadjer and J. Kim, ACS Appl. Mater. Interfaces, 2018, 10, 29724–29729.
42. J. Kim, H. Kim, I. Kang, J. Kim, S. Ko, J. Bae and J. Kim, J. Mater. Chem. C, 2024, 12, 6193–6200.
43. H. Zhou, M. Si, S. Alghamdi, G. Qiu, L. Yang and P. D. Ye, IEEE Electron. Dev. Lett., 2017, 38, 103–106.

---

Footnote

† Electronic supplementary information (ESI) available: Calculating contact resistance, XPS spectra of Ga 2p peaks, and TEM and EDS mapping of β-Ga₂O₃ without plasma treatment (Fig. S1–S4). See DOI: https://doi.org/10.1039/d5tc00129c

---