記住我
In this study, we fabricated and characterized TCO/InGaAs Schottky barrier contact. We directly deposited TCO electrodes on n- and p-type InGaAs substrates to evaluate the Schottky barrier height (SBH) and the Ohmic contact resistivity from I–V and C–V measurements. The optical response of the TCO/p-InGaAs Schottky barrier contact up to 1800 nm was also investigated, demonstrating the high potential for efficient and broadband light detection in the visible to SWIR region.
N- and p-type In0.53Ga0.47As epilayers of 200 nm thickness and a nominal carrier density of 2 × 1017 cm−3 were grown on an InP (001) wafer by metal-organic chemical vapor deposition (MOCVD). Ce and H co-doped In2O3 having an electron mobility of ∼160 cm2/V s at a carrier density of ∼2 × 1020 cm−3 was used as the SWIR TCO electrode. First, amorphous TCO films of 110 nm thickness were grown using the reactive plasma deposition method through the liftoff process. Thereafter, the TCO films were crystallized by annealing at 250 °C. Further details about the TCO film can be found in Ref. 44. T. Maeda, K. Oishi, H. Ishii, W. H. Chang, T. Shimizu, A. Endoh, H. Fujishiro, and T. Koida, Appl. Phys. Lett. 119, 192101 (2021). https://doi.org/10.1063/5.0065776. Finally, Ni/Au layers were formed to serve as metal contacts, followed by post-metallization annealing (PMA) at 200 °C. Fabricated device structure is shown in Fig. 1(a). The sheet resistance of the TCO film was 18.7 Ω/sq., which is sufficiently low for a metal electrode. To characterize the TCO/InGaAs Schottky barrier contact, I–V and C–V measurements were performed at room temperature. Specific contact resistivity (ρc) was examined according to the circular transfer length method (CTLM) with the four-point probe using the heavily doped n-InGaAs layers having a donor density of ∼2 × 1019 cm−3. To reduce ρc, ultra-thin Ni layer from 0 to 3 nm thick were deposited on InGaAs at room temperature, and metallic Ni-InGaAs alloy layers were formed during the TCO crystallization process at 250 °C. For the photoresponsivity characterization, we previously developed a photocurrent measurement system with focused SWIR illuminations.4,13,144. T. Maeda, K. Oishi, H. Ishii, W. H. Chang, T. Shimizu, A. Endoh, H. Fujishiro, and T. Koida, Appl. Phys. Lett. 119, 192101 (2021). https://doi.org/10.1063/5.006577613. T. Maeda, H. Ishii, W. H. Chang, T. Shimizu, H. Ishii, O. Ohishi, A. Endo, and H. Fujishiro, Jpn. J. Appl. Phys., Part 1 59, SGGE03 (2020). https://doi.org/10.7567/1347-4065/ab5b4414. K. Oishi, H. Ishii, W. Chang, H. Ishii, A. Endoh, H. Fujishiro, and T. Maeda, Phys. Status Solidi A 218, 2000439 (2021). https://doi.org/10.1002/pssa.202000439 The monochromatic SWIR light illuminated the TCO/InGaAs Schottky barrier contact with a sensing area of 2.6 × 10−4 cm2. The incident power was measured with a Ge power meter and controlled using neutral density (ND) filters. The electrical performance of the TCO/InGaAs Schottky barrier contact during the photocurrent measurement was evaluated using a semiconductor device analyzer.The electrical parameters of the diode and the conduction mechanism are determined by analyzing the I–V curves of the TCO/InGaAs Schottky barrier contact. Typical I–V curves of n- and p-type InGaAs Schottky barrier contacts with the SWIR TCO electrode are shown in Fig. 1(b). Ohmic contact behavior was observed for TCO/n-InGaAs, implying a low effective Schottky barrier height (SBH). In contrast, the TCO/p-InGaAs contact exhibited typical rectifying characteristics with a high rectification ratio of ∼105, indicating significant Schottky barrier formation for hole carrier at the TCO/p-InGaAs interface.An ideal Schottky diode is defined as a junction having an ideal metal–semiconductor interface, with the I–V relationship given by1515. E. H. Rhoderick and R. H. Williams, Metal-Semiconductor Contacts, 2nd ed. ( Clarendon, Oxford, 1988), pp. 89–129. I=I0 expq(V−IRS)nkBT1−expqVkBT,(1)where V is the applied bias voltage, kB is the Boltzmann’s constant, T is the absolute temperature, RS is the series resistance, and q is the electronic charge. The ideality factor, n, is determined by image force barrier lowering and thermionic-assisted tunneling and depends on the doping level and barrier height. Here, the saturation current I0 is given by I0=AA**T2 exp−qϕbkBT,(2)where A is the contact area, A** is the effective Richardson constant of p-InGaAs, and ϕb is the SBH. The Richardson constant of the p-type In0.53Ga0.47As is 61.9 A cm−2 K−2.1616. J. L. Veteran, D. P. Mullin, and D. I. Elder, Thin Solid Films 97(2), 187 (1982). https://doi.org/10.1016/0040-6090(82)90227-9 From the I–V curve of the TCO/p-InGaAs Schottky barrier contact, ϕbp, n, and RS were deduced to be 0.587 ± 0.01 eV, 1.16 ± 0.05, and 9.23 ± 0.05 Ω, respectively. The extracted n is smaller than that of other elemental metal/p-InGaAs contact counterparts,17,1817. L. Malacký, P. Kordos, and J. Novak, Solid-State Electron. 33, 273 (1990). https://doi.org/10.1016/0038-1101(90)90166-C18. J. Selders, N. Emeis, and H. Beneking, IEEE Trans. Electron. Device 32, 605 (1985). https://doi.org/10.1109/T-ED.1985.21985 which suggests the presence of fewer interfacial trap states and a thin interfacial layer between TCO and InGaAs.Considering the electrical energy band structure, TCO is a degenerate semiconductor having a high carrier concentration of ∼2 × 1020 cm−3. Therefore, a built-in field exists in the lightly doped InGaAs substrate (ND,A = ∼2 × 1017 cm−3), but not in the heavily doped TCO layer. Figure 1(c) shows plots of C and 1/C2 as a function of reverse-bias voltage for the TCO/p-InGaAs Schottky contact measured at 1 MHz. The depletion-layer capacitance of a Schottky barrier contact is given as follows:1919. H. H. Wieder, J. Vac. Sci. Technol., B 21(4), 1915 (2003). https://doi.org/10.1116/1.1588646 where εS is the semiconductor permittivity of the p-type In0.53Ga0.47As (εS = 13.4 ε0, where ε0 permittivity of the free space) and Vd0 is the diffusion potential at zero bias and is determined from the extrapolation of the linear 1/C2 − V plot to the V-axis. The carrier concentration NA is obtained from the slope. The barrier height can be calculated using the equation ϕb(C − V) = Vd0+Vp, where Vp is the potential difference between the Fermi-level and the top of the valence band of p-InGaAs, which can be calculated from the equation Vp = (kT/q)ln(NV/NA), where NV = 7.7 × 1018 cm−3 is the density of effective states in the valence band of the p-type In0.53Ga0.47As. As shown in Fig. 1(c), the estimated diffusion potential and barrier height of the TCO/p-InGaAs Schottky barrier diode are 0.480 ± 0.01 V and 0.567± 0.01 eV, respectively. Our results showed that the effective barrier heights estimated from I–V measurements agree closely with those obtained from C–V measurements. Figure 1(d) is the effective band diagram of the TCO/p-InGaAs Schottky barrier contact based on the C–V measurement. Here, the existence of an interfacial layer will be discussed later. A hole density of 2.7 × 1017 cm−3 was extracted for p-InGaAs, where Fermi-level lay at 0.087 eV above the valence band edge. The TCO is a highly degenerated, wide-bandgap semiconductor material whose work function of approximately 4.2 eV is smaller than the electron affinity of InGaAs.44. T. Maeda, K. Oishi, H. Ishii, W. H. Chang, T. Shimizu, A. Endoh, H. Fujishiro, and T. Koida, Appl. Phys. Lett. 119, 192101 (2021). https://doi.org/10.1063/5.0065776 According to the difference of Fermi-level energy between TCO and p-InGaAs, the band of p-InGaAs bends downward to produce a built-in potential of 0.48 eV. The resulting depletion layer in p-InGaAs with the thickness of 51 nm can work as the absorption layer to extract the photogenerated carriers efficiently. As expected from the effective band diagram, we confirmed the Ohmic behavior for TCO/n-InGaAs contact and Schottky barrier formation for TCO/p-InGaAs contact. The SBH of 0.587 eV obtained from the I–V curves for the TCO/p-InGaAs Schottky diode is comparable with those for elemental metals, such as Ag (0.62 eV), Al (0.61 eV), and Ti (0.55 eV), whose work functions are 4.26, 4.28, and 4.33 eV, respectively.1717. L. Malacký, P. Kordos, and J. Novak, Solid-State Electron. 33, 273 (1990). https://doi.org/10.1016/0038-1101(90)90166-C Assuming that the barrier heights on n-type and p-type materials follow the relation ϕbn + ϕbp = Eg (0.752 eV), then ϕbn should be 0.185 eV. Since the charge neutral level ECNL of InGaAs is reported to be at ∼0.2 eV below the conduction band,1919. H. H. Wieder, J. Vac. Sci. Technol., B 21(4), 1915 (2003). https://doi.org/10.1116/1.1588646 the Fermi-level at the TCO/p-InGaAs interface may be pinned. However, it is difficult to clarify the Fermi-level pinning behavior at the TCO/InGaAs interface from the SBH value. Wang et al. stated that the Fermi-level pinning at the metal interface on p-type InGaAs is not as strong as that on n-type InGaAs.2020. R. Wang, M. Xu, P. Ye, and R. Huang, J. Vac. Sci. Technol., B 29, 041206 (2011). https://doi.org/10.1116/1.3610972 From the perspective of oxide metal, the high interface state density (Dit) at oxide/InGaAs interfaces is attributable to the Ga dangling bonds and/or As–As dimers created during the oxidation process at InGaAs surfaces. The passivation of trivalent oxides, such as Al2O3,2121. H. C. Chiu, L. T. Tung, Y. H. Chang, Y. J. Lee, C. C. Chang, J. Kwo, and M. Hong, Appl. Phys. Lett. 93, 202903 (2008). https://doi.org/10.1063/1.3027476 Ga2O3,2222. W. Jevasuwan, T. Maeda, N. Miyata, M. Oda, T. Irisawa, T. Tezuka, and T. Yasuda, Appl. Phys. Express 7, 011201 (2014). https://doi.org/10.7567/APEX.7.011201 or La2O3,2323. D. H. Zadeh, H. Oomine, Y. Suzuki, K. Kakushima, P. Ahmet, H. Nohira, Y. Kataoka, A. Nishiyama, N. Sugii, K. Tsutsui, K. Natori, T. Hattori, and H. Iwai, Solid-State Electron. 82, 29–33 (2013). https://doi.org/10.1016/j.sse.2013.01.013 with InGaAs surfaces has been reported to eliminate such dangling bonds and dimers because of the abrupt and chemical-bond-well-arranged interface between the trivalent oxides and InGaAs.2424. L. Lin and J. Robertson, Appl. Phys. Lett. 98, 082903 (2011). https://doi.org/10.1063/1.3556619 Similarly, trivalent In2O3 can passivate the interface states to reduce the Fermi-level pinning. The presence of ∼1 at. % hydrogen in TCO should also be considered because hydrogen can passivate the InGaAs surface to reduce Dit.2525. K. Tang, F. R. M. Palumbo, L. Zhang, R. Droopad, and P. C. McIntyre, ACS Appl. Mater. Interfaces 9, 7819 (2017). https://doi.org/10.1021/acsami.6b16232 Further investigation is required to verify the Schottky barrier characteristics in this unique TCO/InGaAs contact.A low specific contact resistivity (ρc) is desirable for all applications of optoelectronic devices. Although we investigated the Ohmic behavior for TCO/n-InGaAs (ND = ∼2 × 1017 cm−3) contact based on I–V characterization, the measured ρc of TCO/n+-InGaAs (ND = ∼2 × 1019 cm−3) contact using CTLM was 56.3 μΩ cm2. For further reduction of ρc, Ni-layer insertion between n-InGaAs and TCO has been examined. The Ni layer is well known as a wetting layer on semiconductors, which reduces the spiking of Ohmic contacts.2626. J. Lin, S. Yu, and S. E. Mohney, J. Appl. Phys. 114, 044504 (2013). https://doi.org/10.1063/1.4816097 To date, Ni-based Ohmic contact has been used to achieve a low ρc of 10−5–10−8 Ω cm2 for n-InGaAs by forming a uniform metallic Ni-InGaAs alloy layer upon annealing at temperatures as low as 250 °C.27–3027. L. Czornomaz, M. E. Kazzi, M. Hopstaken, D. Caimi, P. Machler, C. Rossel, M. Bjoerk, C. Marchiori, H. Siegwart, and J. Fompeyrine, Solid-State Electron. 74, 71 (2012). https://doi.org/10.1016/j.sse.2012.04.01428. X. Zhang, H. X. Guo, X. Gong, C. Guo, and Y.-C. Yeo, ECS J. Solid-State Sci. Technol. 1, 82 (2012). https://doi.org/10.1149/2.014202jss29. J. Oh, S. Yoon, B. Ki, Y. Song, and H.-D. Lee, Phys. Status Solidi A 212(4), 804 (2015). https://doi.org/10.1002/pssa.20143171330. S. Kim, S. K. Kim, S. Shin, J.-H. Han, D.-M. Geum, J. P. Shim, S. Lee, H. Kim, G. Ju, and J. D. Song, IEEE J. Electron. Device Soc. 7, 869 (2019). https://doi.org/10.1109/JEDS.2019.2907957 The thickness of the Ni-InGaAs alloy can be controlled by adjusting the amount of deposited Ni. Therefore, “silicide-like” Ni-InGaAs alloyed contacts to the InGaAs channel were introduced as a self-aligned metallization method.24,25,2724. L. Lin and J. Robertson, Appl. Phys. Lett. 98, 082903 (2011). https://doi.org/10.1063/1.355661925. K. Tang, F. R. M. Palumbo, L. Zhang, R. Droopad, and P. C. McIntyre, ACS Appl. Mater. Interfaces 9, 7819 (2017). https://doi.org/10.1021/acsami.6b1623227. L. Czornomaz, M. E. Kazzi, M. Hopstaken, D. Caimi, P. Machler, C. Rossel, M. Bjoerk, C. Marchiori, H. Siegwart, and J. Fompeyrine, Solid-State Electron. 74, 71 (2012). https://doi.org/10.1016/j.sse.2012.04.014 However, a thick Ni layer or a resulting Ni-InGaAs alloy layer that is more than a few nm thick may degrade the transparency. To reduce ρc while retaining high transparency, we inserted an ultra-thin Ni layer between TCO and the n+-InGaAs layer. We investigated the relationship between ρc, transmittance, and ultra-thin Ni thickness. Figure 2 plots the measured ρc (solid symbols) and the transmittance at a wavelength of 1550 nm (open symbols) of the TCO/ultra-thin Ni-InGaAs alloy layer/n+-InGaAs structure as a function of Ni thickness from 0 to 3 nm. Evidently, transmittance monotonically decreases with increasing Ni thickness. The 3 nm thick Ni layer leads to about 50% reduction in transmittance. On the other hand, by inserting the Ni layer, the ρc drops sharply until it attains a constant value of 3.7 ± 0.1 μΩ cm2. Only the 0.3 nm thick Ni layer achieved significant ρc reduction with minimum transmittance degradation.Because the passage of electrical current through TCO/InGaAs interfaces is required during operation, the quality of interfaces has an immense impact on the performance of Schottky barrier contacts. We performed a high-resolution TEM analysis to understand the different TCO/InGaAs interfacial conditions before and after the insertion of the ultra-thin Ni layer, and how they affect the specific contact resistance. Bright-field (BF) cross-sectional TEM (XTEM), high-angle annular dark-field (HAADF) scanning TEM (STEM), and x-ray energy dispersive spectroscopy (XEDS) line scans of the TCO/InGaAs and the TCO/Ni(0.3 nm)/InGaAs are compared in Fig. 3. We observed a clear and flat interfacial structure between TCO and InGaAs, wherein TCO was well crystallized up to the interface. On the TCO side, a disordered but uniform interfacial layer of approximately 1.4 nm thickness was observed both in TCO/InGaAs and TCO/Ni/InGaAs structures. In the HAADF-STEM images, these interfacial layers exhibited a darker contrast, indicating a low mass density. The XEDS line scans of O, In, Ga, and As profiles reveal unique interfacial layer formation at the TCO/InGaAs interface. The O signal in TCO and the As signal in InGaAs show the step-like profiles at the TCO/InGaAs interface. In contrast, the In and Ga profiles show a distinct drop at the interface. Interestingly, the Ga profiles present a hump at the TCO side, which corresponds to the dark layer in the HAADF-STEM images of both structures. In the TCO/InGaAs contact, we also found a slightly brighter layer on the top of the InGaAs substrate in the HAADF-STEM image, which presents a contiguous lattice image of the InGaAs substrate. These observations indicate two types of interfacial layers exist between TCO and InGaAs. One is the In/Ga-rich InGaAs oxide layer at the TCO side and the other is the In/Ga-deficient InGaAs layer, as indicated in XEDS line scans. Our observations implied that In/Ga atoms preferentially incorporate into the In2O3 layer over the InGaAs crystals at the TCO/InGaAs interface. Although the exact origin of these layers is unclear, it is necessary to understand the thermodynamics of these systems. Before the TCO deposition, native oxides exist on the InGaAs surface. Further oxidation of InGaAs surface proceeds under the active oxygen ambient during the TCO deposition. The free Gibbs energy to from In2O3, Ga2O3, and As2O5 per mole O2 are −554, −666, and −313 kJ/mol, respectively.3131. R. Winter, P. Shekhter, K. Tang, L. Floreano, A. Verdini, P. C. McIntyre, and M. Eizenberg, ACS Appl. Mater. Interfaces 8, 16979 (2016). https://doi.org/10.1021/acsami.6b02957 The Gibbs free energy of In2O3 and Ga2O3 are more negative than that of As2O5. We expect that at the initial stage of the TCO deposition, In and Ga are preferentially oxidized rather than As, forming the In/Ga-rich InGaAs oxide layer under the TCO layer and remaining the In/Ga deficient layer in InGaAs substrate. These interface layers can have a strong impact on carrier transport at TCO/InGaAs Schottky barrier contacts. The incorporation of Ga in In2O3 increases the effective bandgap of In2O3, which is responsible for the high potential barrier for carriers. In addition, the In/Ga-deficient InGaAs layer may cause substantial losses in carrier transport by trapping carriers at defects and altered layers. The value of SBH in the TCO/p-InGaAs contact should be considered to include the electrical effect of these interface layers. In addition, since the formation of the interface layer relays significantly on the oxidation process, the value of SBH and contact resistance are expected to depend greatly on the TCO deposition conditions, that is, a substrate pretreatment, a deposition temperature, an oxidation atmosphere, and a crystallization temperature.We also observed two types of interfacial layers in TCO/Ni(0.3 nm)/InGaAs contact as shown in XEDS line scans. From TEM analysis, the Ni-InGaAs alloy layer exists conformably between TCO and InGaAs, while mainly In/Ga-diffused layer is observed above the Ni-InGaAs alloy layer similar to TCO/InGaAs contact. The Ni peak position agrees closely with the dent of the In and Ga profiles, which implies that Ni favorably reacts with As to form metallic nickel arsenide. Also, we observed that Ga atoms penetrate through the Ni layer and diffuse to the TCO. Therefore, As-rich Ni-InGaAs layer form directly on InGaAs without any oxide layer. It is reported that uniform NiAs-based compound formation adjacent to GaAs has previously been observed in Ni/GaAs contact and decreases the specific contact resistance.3232. Y.-C. Shih, M. Murakami, E. L. Wilkie, and A. C. Callegari, J. Appl. Phys. 62, 582 (1987). https://doi.org/10.1063/1.339860 The main difference from the TCO/InGaAs contact is that the metallic Ni-InGaAs alloy layer is directly contact on the top layer of InGaAs. We conclude that the ultra-thin (0.3 nm) Ni layer deposition between TCO and InGaAs effectively reduces the contact resistance and maintains transparency.InGaAs-based Schottky barrier photodiodes (SB-PDs) have advantages in terms of speed and long-wavelength detection owing to the built-in electrical field formed between the metal electrode and semiconductor contact. To validate the SWIR transparency in TCO, we measured the I–V characteristics under SWIR illimitation through the TCO electrode. The I–V characteristics for TCO/p-InGaAs SB-PDs measured under illumination at an excitation wavelength of 1550 nm with different incident powers and in the dark are shown in Fig. 4(a). Owing to the low dark current in reverse voltage due to high SBH, we observed a clear photoresponse, demonstrating front-side illuminated SWIR detection in the depletion layer of InGaAs just under the SWIR TCO electrode. As the incident power increased, the number of photogenerated charge carriers also increased, resulting in an enhanced photocurrent. The responsivity was determined to be 0.148 A/W from the linear relationship between the photocurrent and the incident power at VD = 0.1 V. The wavelength-dependent responsivity of the TCO/p-InGaAs SB-PDs in the wavelength range of 800–1800 nm is plotted in Fig. 4(b). Overall, the responsivity decreases gradually from 800 to 1800 nm, with 1800 nm being the cutoff wavelength for InGaAs. Compared to the spectral response of commercially available InGaAs PIN-PDs, the TCO/InGaAs SB-PDs exhibit much lower responsivity in the wavelength region of 1000–1800 nm, mainly due to the difference in absorption layer thickness. The width of the depletion layer formed in the intrinsic InGaAs layer of InGaAs PIN-PDs is generally designed to exceed 1–2 μm to ensure complete absorption of SWIR light, despite a trade-off between the light absorption and the response speed. On the other hand, the calculated depletion width xd of our TCO/p-InGaAs SB-PDs at VD = 0.1 V was 56.4 nm, which is significantly smaller than that of InGaAs-PDs. Notably, the responsivity of the TCO/InGaAs SB-PDs exceeds that of the InGaAs PIN-PDs at wavelengths smaller than 900 nm. In the NIR region, the responsivity of PIN-PDs is degraded since the light is absorbed by the highly doped contact layer where the photogenerated carriers cannot be extracted owing to the weaker built-in electric field and high recombination rate. In contrast, in the TCO/InGaAs SB-PDs, the visible and NIR light is also directly absorbed by the depletion layer immediately underneath the TCO, and the photogenerated carrier rapidly reaches the adjacent TCO with minimal access resistance, resulting in high responsivity with the high-speed response. Since the depletion layer formed in the TCO/InGaAs Schottky contact can absorb almost visible light up to the TCO bandgap, TCO/p-InGaAs SB-PDs can be used for broadband photodetection in the visible to SWIR region. We conclude that the TCO/InGaAs Schottky barrier contact would be an excellent route for broad wavelength detection and high-speed applications.We investigated the TCO/InGaAs Schottky barrier contact for optoelectronic applications in the visible to SWIR region. We found a high SBH of 0.587 ± 0.01 eV (I–V) and 0.567± 0.01 eV (C–V) for p-InGaAs. In contrast, the Ohmic contact behavior of 56.3 μΩ cm2 was observed for n-InGaAs. The insertion of an ultra-thin Ni layer between TCO and n+-InGaAs reduced the contact resistance while maintaining high transparency. We verified that TCO/p-InGaAs contact performed as SB-PDs sensing from the NIR to SWIR region in a front-side illuminating manner.
This work was supported by JST, CREST (Grant No. JPMJCR21C2), Japan, and Nano Processing Facility, AIST-NPF. T.K. would like to thank Jiro Nishinaga for sharing discussion on Ni insertion between InGaAs and TCO.
Conflict of Interest
The authors have no conflicts to disclose.
Author Contributions
Tatsuro Maeda: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Kazuaki Oishi: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal). Hiroto Ishii: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal). Hiroyuki Ishii: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal). Wen Hsin Chang: Data curation (equal); Formal analysis (equal); Supervision (equal); Validation (equal); Writing – review & editing (equal). Tetsuji Shimizu: Data curation (equal); Formal analysis (equal); Validation (equal). Akira Endoh: Supervision (equal); Validation (equal). Hiroki Inomata Fujishiro: Supervision (equal); Validation (equal). Takashi Koida: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Validation (equal); Writing – review & editing (equal).
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
1. T. Koida, H. Fujiwara, and M. Kondo, Appl. Phys. Express 1, 041501 (2008). https://doi.org/10.1143/APEX.1.041501, Google ScholarCrossref, ISI2. T. Koida, Y. Ueno, and H. Shibata, Phys. Status Solidi A 215, 1700506 (2018). https://doi.org/10.1002/pssa.201700506, Google ScholarCrossref3. T. Koida and J. Nomoto, Phys. Rev. Mater. 6, 055401 (2022). https://doi.org/10.1103/PhysRevMaterials.6.055401, Google ScholarCrossref4. T. Maeda, K. Oishi, H. Ishii, W. H. Chang, T. Shimizu, A. Endoh, H. Fujishiro, and T. Koida, Appl. Phys. Lett. 119, 192101 (2021). https://doi.org/10.1063/5.0065776, Google ScholarScitation, ISI5. J. H. Kim, S. S. Li, L. Figueroa, T. F. Carruthers, and R. S. Wagner, J. Appl. Phys. 64, 6536 (1988). https://doi.org/10.1063/1.342048, Google ScholarScitation, ISI6. D. G. Parker, P. G. Say, and A. M. Hansom, Electron. Lett. 23, 527 (1987). https://doi.org/10.1049/el:19870380, Google ScholarCrossref, ISI7. W. A. Wohlmuth, J.-W. Seo, P. Fay, C. Caneau, and I. Adesida, IEEE Photonics Technol. Lett. 9, 1388 (1997). https://doi.org/10.1109/68.623272, Google ScholarCrossref8. H. Algadi, C. Mahata, S. Kim, and G. K. Dalapati, J. Electron. Mater. 50(4), 1800 (2021). https://doi.org/10.1007/s11664-020-08565-1, Google ScholarCrossref9. N. Biyikli, I. Kimukin, O. Aytur, M. Gokkavas, M. S. Unlu, and E. Ozbay, IEEE Photonics Technol. Lett. 13, 705 (2001). https://doi.org/10.1109/68.930421, Google ScholarCrossref10. N. Biyikli, T. Kartaloglu, O. Aytur, I. Kimukin, and E. Ozbay, Appl. Phys. Lett. 79, 2838 (2001). https://doi.org/10.1063/1.1412592, Google ScholarScitation, ISI11. N. Biyikli, I. Kimukin, T. Kartaloglu, O. Aytur, and E. Ozbay, Appl. Phys. Lett. 82, 2344 (2003). https://doi.org/10.1063/1.1566459, Google ScholarScitation, ISI12. N. Biyikli, I. Kimukin, B. Butun, O. Aytür, and E. Ozbay, IEEE J. Sel. Top. Quantum Electron. 10, 759 (2004). https://doi.org/10.1109/JSTQE.2004.833977, Google ScholarCrossref13. T. Maeda, H. Ishii, W. H. Chang, T. Shimizu, H. Ishii, O. Ohishi, A. Endo, and H. Fujishiro, Jpn. J. Appl. Phys., Part 1 59, SGGE03 (2020). https://doi.org/10.7567/1347-4065/ab5b44, Google ScholarCrossref14. K. Oishi, H. Ishii, W. Chang, H. Ishii, A. Endoh, H. Fujishiro, and T. Maeda, Phys. Status Solidi A 218, 2000439 (2021). https://doi.org/10.1002/pssa.202000439, Google ScholarCrossref15. E. H. Rhoderick and R. H. Williams, Metal-Semiconductor Contacts, 2nd ed. ( Clarendon, Oxford, 1988), pp. 89–129. Google Scholar16. J. L. Veteran, D. P. Mullin, and D. I. Elder, Thin Solid Films 97(2), 187 (1982). https://doi.org/10.1016/0040-6090(82)90227-9,
留言 (0)