WorksPhysics and Astronomy
Degradiation of -GaO Schottky barrier diode under swift heavy ion irradiation
Abstract
The electrical characteristics and microstructures of -GaO Schottky barrier diode (SBD) devices irradiated with swift heavy ions (2096 MeV Ta ions) have been studied. It was found that -GaO SBD devices showed the reliability degradation after irradiation, including turn-on voltage Von, on-resistance Ron, ideality factor n and the reverse leakage current density Jr. In addition, the carrier concentration of the drift layer was decreased significantly and the calculated carrier removal rates were 5106 - 1.3107 cm-1. Latent tracks induced by swift heavy ions were observed visually in the whole -GaO matrix. Furthermore, crystal structure of tracks was amorphized completely. The latent tracks induced by Ta ions bombardments were found to be the reason for the decrease in carrier mobility and carrier concentration. Eventually, these defects caused the degradation of electrical characteristics of the devices. By comparing the carrier removal rates, the -GaO SBD devices were more sensitive to swift heavy ions irradiation than SiC and GaN devices.
Introduction
Monoclinic (-) is a traditional transparent conductive oxide materials and - based photodetectors are attracting interest as truly solar-blind deep ultraviolet photodetectors, since they exhibit cut-off wavelengths below 280 nm.[1, 2] This makes - attractive in the fields of new generation photoconductors, such as deep ultraviolet detectors, light-emitting diodes and lasers.
The research on - has been extremely hot in the past decade because of its new application in power electronic devices. - has not only excellent optical properties, but also a large bandgap of - eV and a high critical electric field strength of 8 MV/cm.[3] The large bandgap and the high critical electric field strength enables - based devices to operate at high temperature and high power. Furthermore, - can be prepared by melting method, which was the same as Si and sapphire substrate. Compared with SiC and GaN, the cost advantage of - further promotes its application in the field of high-power electronic devices.[4]
- devices will face huge challenges used in aerospace systems despite their excellent properties. The radiation environment in outer space comprises high-energy protons, electrons, neutrons, and heavy ions.[5] Then, the different types of damages can be formed in the devices after different particle irradiation. For electrons, protons and -rays irradiation, simple point defects are generally introduced in the wide band gap semiconductors.[6, 7] Heavy ions and fast neutrons mainly introduce point defects or cascade displacement damages by elastic collision with target atoms.[8] - is generally considered to be radiation hardness to displacement damage due to the high bond energy and large band gap.[9] According to the literature, the 4H-SiC single crystal was amorphous at fluence of 0.4 dpa (displacements per atom) for 4 MeV Xe ions irradiation,[10] but the saturate disorder state of - single crystal can be reached at a higher fluence of 0.6 dpa for 700 keV Sn ions irradiation.[11] Moreover, the irradiation response of carrier concentration in - Schotty barrier diode (SBD) is similar to that of GaN devices after irradiated by electrons and protons.[12]
Different from the above traditional particles that mainly introduce damages by interaction with the target atoms, the swift heavy ions (SHIs, MeV/u), one of the cosmic rays, mainly transfer energy to the target electrons through huge electronic energy deposition and target electrons further transfer the energy to the atoms through electron-phonon coupling.[13, 14] When electronic energy loss () is lager enough, a single swift heavy ion can cause local melting of material and introduce amorphous or recrystallized damage region during quenching. This damage region of nanometer in size is called latent track. In our previous study, it was found that amorphous latent tracks can be introduced in -GaO single crystal when the exceeded 17 keV/nm.[15] However, the effect of latent tracks on the electrical characteristics of -GaO devices is still not yet studied. Therefore, 2096 MeV Ta ions were used to irradiate -GaO SBD devices in this work and the role of latent tracks on the reliability degradation of devices was analyzed in detail.
Experimental details
The vertical β-GaO (001) drift layer (Sn: \times 10^{16}$ cm) of thickness 8 m was deposited by hydride vapor phase epitaxy on 1.5 mm bulk substrate (Sn: 3\times 10^{18}$ cm). The metal stack of Ti/Au was deposited on the whole back of substrate by E-beam evaporation and followed by the rapid thermal annealing at 500°C for 60 s under nitrogen atmosphere to form the Ohmic contact. The front side of drift layer was patterned by lift-off of E-beam deposited Schottky contacts Ni/Au (45 nm/65 nm). The diameter of Schottky contact was about 100 m. The structure of the schematic across section of -GaO SBD was shown in Figure 1 (a). The -GaO SBD devices were divided into three groups and named #1, #2 and #3, respectively.
Figure 1. (color online) (a) the schematic across section of the vertical β-GaO SBD and (b) the distribution diagram of electronic energy loss () and nuclear energy loss () of 2096 MeV Ta in β-GaO SBD.
Heavy ion irradiation experiment was performed at the Heavy Ion Research Facility in Lanzhou (HIRFL) in the Institute of Modern Physics (IMP), Chinese Academy of Sciences (CAS). The -GaO SBD devices without electrical bias were irradiated with 2096 MeV Ta ions in the vacuum chamber. The and nuclear energy loss () were calculated by SRIM 2013 code[16] and the detail distribution of and in -GaO SBD was plotted in Figure 1 (b). The range of 2096 MeV Ta ions in the device is about 50 m, reaching deep inside the substrate far away from the Metal-Semiconductor (M-S) interface. Due to the limited number of samples, cumulative irradiation was adopted in this work. The devices were irradiated for the first time with the fluence of – ions/cm, respectively. After irradiation, the irradiated samples were removed from the vacuum chamber for electrical properties measurement. Then they were continued to irradiated until the fluence up to – ions/cm, respectively. The specific irradiation parameters were listed in Table 1.
| Irradiation batches | #1 | #2 | #3 |
| 1st | ions/cm | ions/cm | ions/cm |
| 2nd | ions/cm | ions/cm | ions/cm |
Table 1. The irradiation fluence of the three groups of -GaO SBD in the first irradiation experiment and the total fluence after the second cumulative irradiation experiment.
Current density-voltage (J-V) and high frequency (1 MHz) capacitance-voltage (C-V) characteristics were measured by a Keithley 4200 Semiconductor Parameter Analyzer at room temperature. For each fluence, ten Schottky electrodes at least with almost identical electrical characteristics were analyzed. The normal behaviors of - and - are shown in the next section. The microstructure of -GaO SBD after irradiation was characterized by bright-field TEM using a Tecnai G2 F20 S-TWIN TEM (FEI, USA) at the accelerating voltage of 200 kV.
Results and Discussion
Figure 2 (a) shows the forward - characteristics and the differential on-resistance as a function of the voltage for -GaO SBD devices with different ion fluences. The results show that the forward current density decreases gradually with the increase in fluence. At the forward bias of 2 V, the maximum current density decreased from 327 A/cm to 83 A/cm and the increased from 3.8 m·cm to 13.7 m·cm at the fluence of ions/cm. When the ion fluence increased to and ions/cm, the -GaO SBD devices do not exhibit forward guide characteristics and the values reach to the order of M·cm (see Table 2). The reverse - characteristic also indicates the increase of reverse leakage current density as shown in Figure 2 (b). It suggests that Ta ions irradiation can significantly affect the - characteristics of the -GaO SBD devices and degrade the performance.

Figure 2. (color online) (a) The forward - characteristics and the differential on-resistance before and after irradiation. (b) The reverse - characteristics before and after irradiation. The unit of Ta ions fluence is ions/cm.
| Parameters | Pre-irradiation | ions/cm | ions/cm | ions/cm | ions/cm | ions/cm | ions/cm | ||
| #1 | #2 | #3 | |||||||
| (V) | 0.75 | 0.78 | 0.75 | 0.75 | 0.78 | 0.76 | 0.78 | — | — |
| at 2 V (m cm) | 3.83 | 3.88 | 3.82 | 3.83 | 4.18 | 6.16 | 13.71 | ||
| 1.01 | 1.02 | 1.02 | 1.02 | 1.04 | 1.05 | 1.05 | — | — | |
| (eV) | 1.13 | 1.15 | 1.12 | 1.12 | 1.14 | 1.11 | 1.12 | — | — |
| at -20 V ( A/cm) | 2.00 | 2.00 | 2.05 | 3.12 | 3.45 | 3.75 | 3.97 | 4.16 | 4.25 |
| ( cm) | 1.81 | 1.79 | 1.81 | 1.81 | 1.74 | 1.15 | 0.54 | — | — |
| (cm) | — | — | — | 0 | — | — |
Table 2. The comparison of experimentally calculated values of -GaO SBD devices before and after 2096 MeV Ta ions irradiation
According to the thermionic emission theory, [17, 18] the relationship between the voltage and the current density can be described as:
where is the saturation current density, is the ideality factor, is Boltzmann’s constant, is the absolute temperature, is the effective Richardson constant (41.1 A/(cmK)), and is the Schottky barrier height. The parameters and can be estimated by fitting the linear region of the J-V curve and the detail electrical parameters of β-GaO SBD devices before and after irradiation were summarized in Table 2. In order to compare the variation of electrical parameters more intuitively, the increment of each parameter (the parameter value after irradiation minus the parameter value before irradiation) is shown in Figure 3. Since the β-GaO SBD devices do not exhibit forward guide characteristics when the fluence is up to and ions/cm, the variation of , and in Figure 3 (a) only covered in the fluence range from to ions/cm.
Figure 3. (color online) The increments of electrical parameters (turn-on voltage , ideality factor , Schottky barrier height , reverse leakage current density and on-resistance ) as a function of fluence before and after irradiation.
In Figure 3 (a), both of the turn-on voltage and ideality factor increased with fluence increasing, while the parameter showed little changes. The increase of ideality factor indicates that the current transport mechanism gradually deviates from the thermionic emission model. In general, the defects introduced by irradiation can lead to the increase of M-S interface state density and then other current transport mechanisms will participate in the process, such as tunneling.[19, 20] The defects can also act as the capture centers of carriers, resulting in the decrease of the carrier concentration and mobility.[21] Hence, the on-resistance value increased with the increase in fluence as shown in Figure 3 (b). In general, the reverse leakage current density can reflect the blocking characteristic of SBD. In Figure 3 (b), the increase of after irradiation indicates the degradation of blocking. This is mainly related to the reduced of carrier lifetime due to the increase of deep level recombination centers in the barrier region after irradiation.[22]
Figure 4 shows the C-V and 1/C²-V plots at a frequency of 1 MHz. The C-V relationship for a Schottky barrier is:[23]

Figure 4. (color online) C-V and 1/C²-V characteristics (1 MHz) of the devices after 2096 MeV Ta ions irradiation. The unit of Ta ions fluence is ions/cm².
where is the electron charge, is the area of the Schottky diode, is the dielectric constant (Ga₂O₃, ), is the built-in potential, stands for the carrier concentration in the drift layer. The carrier concentration can be extracted from the slope of the 1/C²-V curve and the results were listed in Table 2. Only the carrier concentrations in the drift layer were calculated with the fluence range of – ions/cm. The corresponding variation of normalized carrier concentration was summarized in Figure 5 (a).
Figure 5. (color online) (a) The normalized carrier concentration and (b) carrier removal rate in the drift -GaO layer after 2096 MeV Ta ions irradiation.
According to Figure 5 (a), it is clear that the carrier concentration in the drift -GaO layer shows little changes at the fluence of ions/cm. However, the carrier concentration decreased significantly with the fluence increased from to ions/cm. At the fluence of ions/cm, the normalized carrier concentration is only 30% that of the unirradiated samples. The acceptor-defects introduced by Ta ions result in the decrease of carrier concentration, further cause the increase of depletion width, and finally show that the capacitance in C-V measurement decreases with the increase of fluence. As the fluence increases further to and ions/cm, the excessively low carrier concentration is equivalent to the infinite width of the depletion layer and the Schottky barrier capacitance disappeared.
According to the carrier concentration, the carrier removal rate was calculated and the results were plotted as shown in Figure 5 (b). The carrier removal rate relates to the removal of carriers as deep traps which were introduced by radiation. It is related to the fluence and the decrease value of carrier concentration , through the equation:[6, 24]
The can provide a practical guide for estimating the degree of the degradation induced in the devices or materials for a given fluence of the common type of radiation. In this work, the calculated was cm for β-GaO SBD irradiated with Ta ions to the fluence of ions/cm and it reached saturation values of cm at the fluence of ions/cm. In general, is linear increasing with the fluence at the lower fluence. However, if most of the carriers are removed at a higher fluence, the excess defects will not contribute to the carrier removal effect any more. Thus, the relationship between and the fluence will not follow the linear relationship.[6]
The carrier removal rates of β-GaO based devices irradiated by different types of ions were summarized as shown in Figure 6 (a) (red symbols).[12, 25, 26, 27] Note that carrier removal rate is 406–728 for particles irradiation,[27] 300 for 10 MeV protons,[12] 4.9 for 1.5 MeV electrons[26] and 19–28 for 1.25 MeV neutrons[25] in -GaO SBD devices or rectifiers. However, the carrier removal rates for SHIs in this work are much higher. This indicates that the energetic Ta ions exhibit the highest carrier removal rates among these ions irradiation, and it can be explained by the damage type caused by SHIs.
Figure 6. (color online) (a) Carrier removal rate summary diagram in β-GaO (red symbols)[12, 25, 26, 27] and other types of GaN or SiC based devices (black symbols)[6, 22, 28, 29, 30, 31, 32] with different species and energies ion-irradiation. The shadow represents the energy regions of swift heavy ions. The data of red star are from our work. (b) The cross-sectional TEM image of β-GaO SBD irradiated with 2096 MeV Ta ions to a fluence of ions/cm. The irradiation direction is indicated by white arrows and the latent tracks parallel to each other are marked by red arrows.
The cross-sectional TEM of the -GaO SBD irradiated by 2096 MeV Ta ions to a fluence of ions/cm was shown in Figure 6 (b). It can be seen that the interface between Ni and N -GaO layer is sharp and there is little inter-mixing at the highest fluence of ions/cm. However, we found indication of latent tracks parallel to each other at N -GaO layer. In our previous work,[15] TEM results proved that one single 2096 MeV Ta ion introduced the amorphous latent track with a size of ~8 nm in -GaO single crystal. Considering the range of 2096 MeV Ta ions in -GaO SBD devices, the latent tracks can be introduced not only at the 8 m N layer, but also within the range of 40 m at the N layer. For a single swift heavy ion irradiation, the latent track along the ion trajectory is a nanometer-size amorphous region. For a single proton or particle irradiation, the introduced damage is isolated atomic-size point defects. Hence Ta ions exhibit the highest carrier removal rate.
Figure 6 (a) also summarized the carrier removal rates of GaN or SiC based devices including SBD devices and high electron mobility transistors (HEMTs).[6, 22, 28, 29, 30, 31, 32] It can be extracted from Figure 6 (a) that under the irradiation environment of high-energy electrons, protons and heavy ions which are mainly introduced displacement damages by elastic collision with the target atoms, the values of -GaO SBD or rectifier are similar to that of GaN or SiC based devices, indicating the excellent radiation hardness of -GaO devices. This can be attributed to the higher formation energy of vacancy defects in -GaO.[33, 34, 35] However, the degradation of -GaO SBD is more serious than that of SiC or GaN devices under the SHIs irradiation as the shadow shown in Figure 6 (a). In addition, SBD devices in our work are completely damaged under 2096 MeV Ta ions irradiation with fluence of . However, the GaN HEMTs reported by Hu et al. [36] are still functional after swift heavy Bi ions irradiation with energy of 1500 MeV to the fluece of .
Based on the thermal spike model,[37] the latent track is formed through the material melting and quenching rapidly along the path of SHIs. Hence, thermodynamic properties and recrystallization ability of the target material are the main factors affecting the latent track formation.[38] The poor thermal conductivity and recrystallization ability of - make the threshold of latent track formation in - () is lower than that of SiC () and GaN (-).[15] Therefore, the damage introduced by SHIs in the whole - matrix has a greater impact on the degradation of - SBD devices than the damage in M-S interface.
Conclusions
In conclusion, we studied the degradation and the structure damages of - SBD devices after 2096 MeV Ta ions irradiation with the fluence range from to . Both the conducting and blocking characteristics were sensitive to ion irradiation. A strong reduction of the carrier was observed and the carrier removal rates were -. Furthermore, the amorphous latent tracks along the ions trajectories cross the whole area of drift layer, were responsible for the decrease in carrier concentration and mobility, and resulted in the deterioration of the - SBD devices. In addition, the damage introduced by SHIs in the whole - matrix has a greater impact on the degradation of - SBD devices than the damage in M-S interface. The serious degradation for - SBD indicates the worse radiation hardness of - based device to SHIs compared with SiC and GaN devices.
Acknowledgments
The authors would like to thank the accelerator staff of HIRFL of IMP for the technical support and for providing the characterization facilities. We also thank Dr. Lijun Xu at the Institute of Modern Physics, Chinese Academy of Sciences for support of the TEM measurement.
References
- [1]Pearton S, Yang J, Carey P, Ren F, Kim J, Tadjer M and Mastro M 2018 Appl. Phys. Rev. 5 11301
- [2]Ueda N, Hosono H, Waseda R and Kawazoe H 1997 Appl. Phys. Lett. 71 933–5
- [3]Zeng K, Vaidya A and Singisetti U 2018 IEEE Electron Device Lett. 39 1385–8
- [4]Reese S B, Remo T, Green J and Zakutayev A 2019 Joule 3 903–7
- [5]Claeys C and Simoen E 2002 Radiation Effects in Advanced Semiconductor Materials and Devices (New York: Springer Berlin Heidelberg)DOI
- [6]Pearton S J, Ren F, Patrick E, Law M E and Polyakov A Y 2016 ECS J. Solid State Sci. Technol. 5 Q35–60
- [7]Pearton S J, Deist R, Ren F, Liu L, Polyakov A Y and Kim J 2013 J. Vac. Sci. Technol. A 31 50801
- [8]Janousek B K, Yamada W E, Krantz R J and Bloss W L 1988 J. Appl. Phys. 63 1678–86
- [9]Yang G, Jang S-P, Ren F, Pearton S and Kim J 2017 ACS Appl. Mater. Interfaces 9 40471–6
- [10]Kerbiriou X, Constantini J-M, Sauzay M, Soriel S, Thomé L, Jagielski J and Grob J-J 2009 J. Appl. Phys. 105 73513
- [11]Wendler E, Treiber E, Baldauf J, Wolf S and Ronning C 2016 Nucl. Instrum. Methods B 379 85–90
- [12]Yang J, Chen Z, Ren F, Pearton S, Yang G, Kim J, Lee J, Flitsiyan E, Chernyak L and Kuramata A 2018 J. Vac. Sci. Technol. B 36 11206
- [13]Komarov F F 2003 Physics-Uspekhi 46 1253–82
- [14]Aumayr F, Facsko S, El-Said A S, Trautmann C and Schleberger M 2011 J. Phys. Condens. Matter 23 393001
- [15]Ai W, Xu L, Nan S, Zhai P, Li W, Li Z, Hu P, Zeng J, Zhang S, Liu L, Sun Y and Liu J 2019 Jpn. J. Appl. Phys. 58 120914
- [16]Ziegler J F and Biersack J P Http://www.srim.org/ [2013]
- [17]Wang L, Nathan M I, Lim T, Khan M A and Chen Q 1996 Appl. Phys. Lett. 68 1267–9
- [18]Shur M 1990 Physics of Semiconductor Devices (New Jersey, United States: Prentice Hall, Inc.)DOI
- [19]Kumar A, Singh R, Kumar P, Singh U, Kandasami A, Karaseov P, Titov A and Kanjilal D 2018 J. Appl. Phys. 123 161539
- [20]Kumar S, Katharria Y, Batra Y and Kanjilal D 2007 J. Phys. D. Appl. Phys. 40 6892
- [21]Manikanthababu N, Tak B, Kunche P, Sarkar S, Kandasami A, Kanjilal D, R. Barman S, Singh R and Panigrahi B 2020 Appl. Phys. Lett. 117 142105
- [22]Yang Z, Ma Y, Gong M, Li Y, Huang M, Gao B and Zhao X 2017 Nucl. Instrum. Methods B 401 51–5
- [23]Neamen D A 2011 Semiconductor Physics and Devices: Basic Principles (Boston, United State: McGraw-Hill Co.)
- [24]Jun B and Subramanian S 2003 IEEE Trans. Nucl. Sci. 49 3222–9
- [25]Polyakov A Y, Smirnov N B, Shchemerov I V, Vasilev A A, Yakimov E B, Chernykh A V, Kochkova A I, Lagov P B, Pavlov Y S, Kukharchuk O F, Suvorov A A, Garanin N S, Lee I-H, Xian M, Ren F and Pearton S J 2020 J. Phys. D. Appl. Phys. 53 274001
- [26]Yang J, Ren F, Pearton S, Yang G, Kim J and Kuramata A 2017 J. Vac. Sci. Technol. B 35 31208
- [27]Yang J, Fares C, Guan Y, Ren F, Pearton S, Bae J, Kim J and Kuramata A 2018 J. Vac. Sci. Technol. B 36 31205
- [28]Kumar V, Maan A and Akhtar J 2018 Phys. Status Solidi A 215 1700555
- [29]Ma G, Zhang Y, Li H, Liu C, Qi C, Wei Y, Wang T, Dong S and Huo M 2019 2019 IEEE 26th International Symposium on Physical and Failure Analysis of Integrated Circuits (IPFA) pp 1–5
- [30]Mikelsen M, Grossner U, Bleka J H, Monakhov E V., Svensson B G, Yakimova R, Henry A, Janzén E and Lebedev A A 2008 Mater. Sci. Forum 600–603 425–428
- [31]Omotoso E, Meyer W E, Auret F D, Diale M and Ngoepe P N M 2016 Phys. B Condens. Matter 480 196–200
- [32]Yang J, Li H, Dong S and Li X 2019 IEEE Trans. Nucl. Sci. 66 2042–7
- [33]Blanco M, Sahariah M, Jiang H, Costales A and Pandey R 2005 Phys. Rev. B 72 184103
- [34]Miceli G and Pasquarello A 2015 Microelectron. Eng. 147 51–4
- [35]Gao F and Weber W J 2002 Nucl. Instrum. Methods B 191 504–8
- [36]Hu P P, Liu J, Zhang S X, Maaz K, Zeng J, Zhai P F, Xu L J, Cao Y R, Duan J L, Li Z Z, Sun Y M and Ma X H 2018 Nucl. Instrum. Methods B 430 59–63
- [37]Toulemonde M, Bouffard S and Studer F 1994 Nucl. Instrum. Methods B 91 108–23
- [38]Rymzhanov R A, Medvedev N, O’Connell J H, Janse van Vuuren A, Skuratov V A and Volkov A E 2019 Sci. Rep. 9 3837