# Fileset

[article-2026-2-16.pdf](https://mdr.nims.go.jp/filesets/16373fdc-4509-43dd-81f8-6c140779f471/download)

## Creator

[Seiichi Kato](https://orcid.org/0000-0002-6427-5463), [Masayuki Okamura](https://orcid.org/0009-0003-3636-6653), [Tomomi Sawada](https://orcid.org/0000-0002-2335-4480), [Toshihide Nabatame](https://orcid.org/0000-0002-5973-0230), [Kazuhito Tsukagoshi](https://orcid.org/0000-0001-9710-2692)

## Rights

This article may be downloaded for personal use only. Any other use requires prior permission of the author and AIP Publishing. This article appeared in Seiichi Kato, Masayuki Okamura, Tomomi Sawada, Toshihide Nabatame, Kazuhito Tsukagoshi; Oxygen reduction for p-type TeOx thin-film transistor. Appl. Phys. Lett. 25 May 2026; 128 (21): 212109 and may be found at https://doi.org/10.1063/5.0329188.[In Copyright](http://rightsstatements.org/vocab/InC/1.0/)

## Other metadata

[Oxygen reduction for p-type TeOx thin-film transistor](https://mdr.nims.go.jp/datasets/d0224ae6-c558-4077-a514-3615f56698f8)

## Fulltext

1  Oxygen reduction for p-type TeOx thin film transistor 1  2 Seiichi Kato1, Masayuki Okamura1, Tomomi Sawada1, Toshihide Nabatame1, and Kazuhito 3 Tsukagoshi1 4  5 1 International Center for Materials Nanoarchitectonics (WPI-MANA), National Institute for Materials 6 Science (NIMS), Tsukuba, Ibaraki 305-0044, Japan 7  8 Keywords: oxide transistor, p-type TFT, TeOx, BEOL compatible, reducing agent, tungsten 9  10 Abstract 11  12 We investigated p-type TeOx thin films fabricated by vacuum deposition using a mixture of TeO2 13 powder and W powder in Al2O3 crucibles and evaluated its thin-film transistor (TFT) characteristics. 14 P-Type TeOx thin films were obtained when W was added at a weight ratio of 15% or more relative to 15 TeO2. However, the films became insulators and exhibited no electrical conductivity when the ratio 16 was 10% or less. X-ray photoelectron spectroscopy analysis revealed that the TeOx thin films contained 17 approximately 5% or more Te–Te bonds. Because W–O bonds exhibit a higher dissociation energy 18 than Te–O bonds, TeO2 was considered to be reduced by W. In addition, W was not present in the film 19 and was used solely for the reduction of TeO2. When a W boat was used as the container for heating 20 the evaporation source, quantitative analysis of the effects of the reduction by W was difficult. 21 Therefore, an Al2O3 crucible, which reacts poorly with TeO2, was used. The results showed that when 22 TeOx TFTs are fabricated by vacuum deposition, the addition of an appropriate amount of reducing 23 agent to the raw material is essential to form an adequate amount of Te–Te bonds. 24  25  26  27  28  29  30  31  32  33  34  35  36  2  The advancement of information technology (IT) society demands higher performance and lower 37 power consumption in semiconductors. To this end, the development of three-dimensional (3-D) 38 integration of logic and memory is widely considered necessary.1 Realizing this 3-D integration 39 necessitates the monolithic 3-D integration of back-end-of-line (BEOL)-compatible logic and memory 40 technologies based on complementary metal-oxide semiconductors (CMOSs).2,3 In BEOL technology, 41 processes must be performed below 400°C to avoid damaging the BEOL metal interconnect layers; 42 however, carrying out processes at such temperatures is difficult with Si.3 Amidst this, oxide 43 semiconductors are gaining attention as alternatives to Si. In addition, demand exists for introducing 44 CMOS technology into display circuits to achieve higher performance.4 Although CMOS devices 45 require both n-type and p-type oxide transistors, the lack of high-performance p-type oxide transistors 46 has prevented their practical implementation. Therefore, the realization of high-performance p-type 47 oxide transistors is strongly demanded. 48  49 High-performance n-type oxide transistors have been realized. InGaZnO was first discovered by 50 Kimizuka et al.5 and has been extensively researched 6–8 and commercialized as an n-type material. 51 InOx,9 InCO,10 InSiO, InWO, and InTiO11 have been widely studied and have achieved relatively stable 52 electrical characteristics. However, no p-type semiconductors suitable for practical application have 53 yet been reported. Cu2O, NiO, and SnO are known p-type semiconductors;12,13 however, they are 54 difficult to prepare as stable thin-film semiconductors because of problems such as property 55 degradation. Among these materials, SnO is the most viable option, with numerous research results 56 available. 14–18 However, distinguishing between n-type SnO2 and p-type SnO is difficult. Sputtering 57 and evaporation methods often result in mixed phases, leading to poor transistor characteristics (e.g., 58 low mobility and low on/off current ratios) when used as thin-film semiconductors. Achieving the 59 practical application of SnO thin-film semiconductors is extremely challenging because of the 60 difficulty in controlling these properties.19 Consequently, attention is increasingly shifting to p-type 61 Te-based thin films. 62  63 Metal Te thin films are known to exhibit thin-film transistor (TFT) characteristics when used as p-type 64 semiconductors. 20–25 Examples of TFTs that simultaneously achieve high mobility and a high on/off 65 ratio have been reported; however, they require special fabrication conditions such as cooling the 66 substrate to −80°C during deposition to achieve high mobility.26 TeO2 has been predicted to exhibit 67 stable characteristics as a p-type semiconductor,27 and it has been the subject of extensive research. 28–68 33 Recently, TFTs with high mobility and high on/off ratios were fabricated by doping TeO2 with Se, 69 attracting significant attention.34 The addition of Se to the thin film has been speculated to result in p-70 type doping. However, no fundamental investigation into the conductivity of tellurium oxide has been 71 reported, and no procedure or method for reliably producing semiconductor Te films has been 72  3  established. Te films are often prepared by vacuum deposition using TeO2 powder placed on a W boat 73 as the evaporation source. After the deposition, the W boat tends to undergo chemical reactions and 74 discoloration. Determining how much W reacts during this process is not straightforward; quantitative 75 analysis is difficult, and reproducibility is uncertain. In the present study, we used an Al2O3 crucible, 76 which reacts poorly with TeO2 and W, and fabricated thin films using a mixture of TeO2 powder and 77 W powder as the evaporation source. Our investigation of the effect of W revealed that the Te–Te 78 bonds within the film play a crucial role in the manifestation of TFT characteristics. 79  80 A Si wafer (low-doped n-type Si) with a 200 nm thermal oxide layer was cut into 15 mm squares, 81 cleaned with isopropyl alcohol and acetone, and then cleaned using an ultraviolet (UV) ozone cleaner. 82 A conventional thermal evaporator was used to deposit thin films onto the substrates. The evaporation 83 source was an Al2O3 crucible filled with a mixed powder of TeO2 (Sigma-Aldrich, 99%) and W 84 (Sigma-Aldrich, 99.9%). The average particle sizes of TeO2 and W were 20 μm and 10 μm, 85 respectively. The weight ratio of W to TeO2 ranged from 0% to 55%. The thin film grew at a rate of 86 0.1 nm/s, with a thickness of approximately 20 nm. The vacuum level at evaporation was 8×10−4 Pa 87 or lower, and the substrate was not cooled. The films were heat-treated at 225°C for 30 min in air after 88 the evaporation. Ni electrodes (100 nm-thick) were subsequently attached by deposition onto the TeOx 89 channels. The TFT characteristics were measured under vacuum using a vacuum probe station (ZEST-90 8000-3-VSAPI, Nagase Electronic Equipment Service). XPS measurements (Quantera SXM, Ulvac 91 Phi) and X-ray diffraction (XRD) measurements (SmartLab, Rigaku) were performed on TeOx films. 92 Secondary-ion mass spectrometry (SIMS, PHI ADEPT-1010, Ulvac Phi) measurements were also 93 performed on a sample with 15% W to analyze its composition. In addition, the composition of the 94 powdered raw material remaining in the crucible after vapor deposition was analyzed by XRD 95 (Miniflex 600, Rigaku). 96  97 Figure 1 shows (a) photographs of a TeOx TFT and (b) an overall view of the TeOx TFT. The electrode 98 spacing ranges from 50 to 350 μm, which does not affect electrical transport properties. The film with 99 15% W was thinned using the focused ion beam (FIB) method and subsequently observed using cross-100 sectional transmission electron microscopy (TEM) and nano-beam electron diffraction (JEM-101 ARM200F, JEOL) [Figs. 1(c) and (d)]. The inset shows the nano-beam electron diffraction pattern of 102 the crystalline region. The majority of the film is amorphous, but some areas are crystallized. The 103 electron diffraction patterns reveal that these crystallized regions are metallic Te crystals, not TeO2 104 crystals. Measurements of the electrical properties reveal that the TFTs with a W ratio less than 10% 105 exhibited insulating behavior, whereas those with a W ratio of 15% or greater displayed transistor 106 characteristics. As an example, Fig. 1 shows the (e) output and (f) transfer curves for samples with 107 25% W and the (g) output and (h) transfer curves of the samples with 0% W. The TFT with 0% W is 108  4  an insulator, whereas that with 25% W exhibits p-type transistor characteristics. The field-effect hole 109 mobility and on/off current ratio were extracted from the transfer curves. Their relationships with the 110 weight ratio of W are plotted in Figs. 2(a) and 2(b). When the W ratio is 10% or less, the TeOx film 111 acts as an insulator and no current flows. However, when the W ratio is 15% or greater, the device 112 exhibits TFT characteristics. As a general trend, with increasing W ratio, the mobility increases and 113 the on/off ratio decreases. 114  115 Figure 3(a) shows high-resolution XPS spectra of the Te 3d core-level for the film with 25% W. The 116 main peaks are observed at the locations indicated by the red arrows (binding energy = 587, 576 eV), 117 and the satellite peaks are observed at the locations indicated by the blue arrows (binding energy = 118 583, 573 eV). These peaks originate from the 3d3/2 and 3d5/2 orbitals of Te, respectively. The two main 119 peaks originate from Te–O bonds, whereas the two satellite peaks originate from Te–Te bonds. 120 Therefore, the ratio of Te–O bonds to Te–Te bonds can be estimated from the intensity ratio between 121 the main peak and the satellite peak. The inset shows the XPS results for the 0% W sample before it 122 was annealed. Satellite peaks are scarcely observed, indicating that the sample contains few Te–Te 123 bonds. The intensity of each peak was estimated from its area, and the ratio of Te–Te bonds across the 124 entire observed region was calculated. Background processing was performed using the Tougaard 125 method. 35 The results for the as-grown film were omitted because its satellite peaks were extremely 126 broad, making analysis difficult. Figure 3(b) shows the relationship between the W weight ratio and 127 the ratio of Te–Te bonds. The white and black square values were calculated from the 3d5/2 and 3d3/2 128 peaks, respectively. The peak intensity ratios of the 3d5/2 and 3d3/2 peaks are nearly consistent. A trend 129 is observed where the Te–Te ratio tends to increase with increasing W concentration, indicating that 130 the greater the concentration of W, the greater the proportion of Te–Te bonds. For films with a W ratio 131 of 15% or greater, the Te–Te bond content is approximately 5% or higher; however, for films with a 132 W ratio less than 10%, the Te–Te bond content is less than 5%. Because conductivity is exhibited when 133 the W ratio is 15% or higher, a Te–Te bond content of 5% or more is required to attain TFT 134 characteristics. No peaks attributable to W were observed in the wide-scan (low-resolution) XPS 135 spectra of any of the films. In addition, the presence of W atoms could not be confirmed from the 136 SIMS measurement results, indicating that the film does not contain W atoms. 137  138 Thin-film X-ray measurements were performed to examine the sample structures. Figure 4(a) shows 139 the diffraction patterns for films with W ratios of 15%, 35%, and 55%. The Miller indices are for the 140 metallic Te crystal. The three films exhibit current modulation in their gate-voltage change. Overall, 141 the films are amorphous; however, weak peaks corresponding to metallic Te crystals are observed. 142 Compared with the peaks of the sample with 15% W, those of the samples wither greater W ratios 143 (e.g., 35% and 55%) are slightly more intense. Therefore, the films with W ratios of 35% and 55% are 144  5  considered to contain more Te crystals than the film with a W ratio of 15%. No clear difference is 145 observed between the diffraction pattern for the film with 35% W and that of the film with 55% W. 146 The XPS results indicate that the amount of Te crystals increases with increasing W ratio in the raw 147 material. The XRD results show differences between the film with 15% W and that with 55% W but 148 not between the film with 35% W and that with 55% W. Figure 3(b) shows that the difference in the 149 Te–Te ratios between the films with 35% and 55% W is small compared with the difference between 150 the films with 15% and 35% W. Therefore, the lack of distinct differences in the XRD patterns is 151 reasonable. 152  153 Even when thin films were formed using pure TeO2 as the evaporation source, neither conductivity 154 nor semiconductor characteristics were observed. Introducing W into TeO2 generated an appropriate 155 proportion of Te–Te bonds, leading to conductivity and semiconductor characteristics. To adjust the 156 proportion of Te–Te bonds, we mixed W powder into the TeO2 powder. The W powder reduced the 157 TeO2, decreased the abundance of Te–O bonds, and increased the abundance of Te–Te bonds. The 158 dissociation energy of the Te–O bond (377 kJ/mol) is lower than that of the W–O bond (720 kJ/mol). 159 Therefore, the W–O bond is more stable. The oxygen in the Te–O bond is pulled toward the W, leading 160 to the phenomenon described above. 161  162 The powdered raw material remaining in the crucible after vapor deposition of the films with 45% W 163 was analyzed by XRD [Fig. 4(b)]. The black, blue, and red triangles in Fig. 4(b) indicate peaks 164 originating from TeO2, W, and WO3, respectively. The results show that WO3, which did not exist prior 165 to vapor deposition, is present. Potential sources of oxygen for W include TeO2 and/or the crucible's 166 Al2O3. The dissociation energy for the Al–O bond is 502 kJ/mol, which is greater than that for Te–O 167 but less than that for W–O. Therefore, TeO2 can supply oxygen to W more easily than Al2O3. In 168 addition, the W powder is thoroughly mixed with the TeO2 powder and only slightly contacts the Al2O3 169 crucible at the crucible surface. Therefore, the primary source of oxygen supplied to the W is the TeO2 170 powder. The following reaction is considered to have occurred within the crucible: 171  172 3TeO2 + 2W → 2WO3 + 3Te 173  174 The dependence of the XPS and XRD peak intensities on the W ratio indicates that increasing the W 175 ratio in the raw material increases the amount of Te–Te bonds and also increases the amount of metallic 176 Te crystals. Consequently, an increase in the conductive metal component within the TeOx film leads 177 to higher mobility, and an increase in the off-current causes a decrease in the on/off ratio. Therefore, 178 to obtain a high-performance TFT, the appropriate amount of W must be added. W is used solely for 179 the reduction of the raw material (TeO2) and does not enter the sample itself to cause any effect. The 180  6  melting point of W (>3380°C) is substantially higher than that of TeO2 (733°C); the W is therefore 181 unlikely to evaporate at the temperatures used for depositing TeO2 at a low rate of 0.1 nm/s. In addition, 182 films were prepared using powders of Si (799 kJ/mol) or C (1076 kJ/mol), which have Si–O and C–O 183 bond dissociation energies greater than that of Te–O, and their TFT characteristics were investigated 184 using the same method. TFT characteristics were achieved both for the sample mixed with 25% Si by 185 weight relative to TeO2 and for the sample mixed with 25% C by weight relative to TeO2. 186  187 The XRD and electron diffraction results indicate that samples exhibiting TFT characteristics contain 188 trace amounts of metallic Te crystals. In addition, the XPS measurements reveal that thin films 189 exhibiting TFT characteristics contain a certain amount of Te–Te bonds. These results suggest that the 190 presence of a moderate amount of Te–Te bonds is essential for achieving TFT characteristics. The 191 conventional method using W boats made quantitatively analyzing the influence of W difficult; 192 however, the method used in the present study enables accurate measurement of the amount of W in 193 the raw material. This approach both enables quantitative analysis and improves the reproducibility. 194 In addition, because there are no heating steps other than the 225°C heat treatment after vapor 195 deposition throughout the entire process, the proposed method is compliant with the requirement of 196 temperatures of 400°C or lower for BEOL-compatible systems. 197  198 In this work, we fabricated amorphous TeO2 thin films using vacuum evaporation with a low-reactivity 199 Al2O3 crucible. We mixed W powder into the TeO2 powder raw material and investigated the 200 relationship between the W ratio and electrical properties of the resultant films. When W was mixed 201 at a weight ratio of 15% or more relative to TeO2, films with electric conduction with p-type TFT 202 characteristics were obtained. However, when the W ratio was 10% or less, the material became an 203 insulator and did not exhibit TFT characteristics. The results of XPS analyses, thin-film XRD 204 measurements, and TEM observations revealed that films exhibiting TFT characteristics form fine Te 205 crystals and exhibit an increase in Te–Te bonds. A Te–Te bond content of at least 5% within the sample 206 is considered essential to attain TFT characteristics. In addition, because the entire process is 207 conducted below 400°C, it can be used with BEOL-compatible systems. 208  209 The authors are grateful to Prof. D. H. Lien, Dr. T. Onaya, and Dr. H. Yoshikawa for useful discussions. 210 This research was conducted with the support of NIMS Open Facility Surface and Bulk Analysis Unit 211 (SBU). 212  213 AUTHOR DECLARATIONS 214  215 Conflict of Interest 216  7   217 The authors have no conflicts to disclose. 218  219 Author Contributions 220  221 Seiichi Kato: Data curation (lead); Investigation (lead); Writing – original draft (lead); Writing – 222 review & editing (lead). Masayuki Okamura: Data curation (supporting). Tomomi Sawada: Data 223 curation (supporting). Toshihide Nabatame: Data curation (supporting); Investigation (supporting). 224 Kazuhito Tsukagoshi: Data curation (supporting); Investigation (equal); Writing – original draft 225 (supporting); Writing – review & editing (equal). 226  227 DATA AVAILABILITY 228  229 The data that support the findings of this study are available within the article. 230  231 REFERENCES 232  233 1 M.M. Sabry Aly, T.F. Wu, A. Bartolo, Y.H. Malviya, W. Hwang, G. Hills, I. Markov, M. Wootters, 234 M.M. Shulaker, H.-S. Philip Wong, and S. Mitra, “The N3XT approach to energy-efficient abundant-235 data computing,” Proc. IEEE 107(1), 19–48 (2019). 236 2 T. Srimani, A. Bechdolt, S. Choi, C. Gilardi, A. Kasperovich, S. Li, Q. Lin, M. Malakoutian, P. 237 McEwen, R.M. Radway, D. Rich, A.C. Yu, S. Fuller, S. Achour, S. Chowdhury, H.-S.P. Wong, M. 238 Shulaker, and S. Mitra, “N3XT 3D technology foundations and their lab-to-fab: omni 3D logic, 239 logic+memory ultra-dense 3D, 3D thermal scaffolding,” in 2023 International Electron Devices 240 Meeting (IEDM), (2023), pp. 1–4. 241 3 W. Wang, K. Li, J. Lan, M. Shen, Z. Wang, X. Feng, H. Yu, K. Chen, J. Li, F. Zhou, L. Lin, P. Zhang, 242 and Y. Li, “CMOS backend-of-line compatible memory array and logic circuitries enabled by high 243 performance atomic layer deposited ZnO thin-film transistor,” Nat Commun 14(1), 6079 (2023). 244 4 S.C. Jang, J. Shim, H. Jang, Y. Lee, S. Biswas, J. Park, H. Shin, H. Kim, S. Kim, and H.-S. Kim, 245 “Recent progress in the development of backplane thin-film transistors for information displays,” 246 Journal of Information Display 26(4), 341–356 (2025). 247 5 N. Kimizuka and T. Mohri, “Spinel, YbFe2O4, and Yb2Fe3O7 types of Structures for compounds in 248 the ln2O3 and Sc2O3-A2O3-BO systems [A: Fe, Ga, or Al; B: Mg, Mn, Fe, Ni, Cu, or Zn] at temperatures 249 over 1000°C”, J. Solid State Chem. 60 382-384 (1985). 250 6 K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, and H. Hosono, “Room-temperature 251 fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors,” 252  8  Nature 432(7016), 488–492 (2004). 253 7 H. Kunitake, R. Arasawa, T. Seki, R. Honda, H. Baba, D. Shimada, H. Kimura, R. Tokumaru, T. 254 Atsumi, K. Kato, S. Yamazaki, K. Ohshima, K. Tsuda, N. Matsumoto, T. Koshida, S. Ohshita, H. 255 Sawai, Y. Yanagisawa, and S. Saga, “A c-axis-aligned crystalline In-Ga-Zn oxide FET with a gate 256 length of 21 nm suitable for memory applications,” IEEE J. Electron Devices Soc. 7, 495–502 (2019). 257 8 N. Yoshida, J.P. Bermundo, Y. Ishikawa, T. Nonaka, K. Taniguchi, and Y. Uraoka, “Low temperature 258 cured poly-siloxane passivation for highly reliable a -InGaZnO thin-film transistors,” Applied Physics 259 Letters 112(21), 213503 (2018). 260 9 M. Si, Z. Lin, Z. Chen, X. Sun, H. Wang, and P.D. Ye, “Scaled indium oxide transistors fabricated 261 using atomic layer deposition,” Nat Electron 5(3), 164–170 (2022). 262 10 R. Kobayashi, T. Nabatame, T. Onaya, A. Ohi, N. Ikeda, T. Nagata, K. Tsukagoshi, and A. Ogura, 263 “Comparison of characteristics of thin-film transistor with In2O3 and carbon-doped In2O3 channels by 264 atomic layer deposition and post-metallization annealing in O3,” Jpn. J. Appl. Phys. 60(3), 030903 265 (2021). 266 11 S. Aikawa, T. Nabatame, and K. Tsukagoshi, “Effects of dopants in InOx-based amorphous oxide 267 semiconductors for thin-film transistor applications,” Appl. Phys. Lett. 103(17), 172105 (2013). 268 12 Z.-W. Shang, H.-H. Hsu, Z.-W. Zheng, and C.-H. Cheng, “Progress and challenges in p-type oxide-269 based thin film transistors,” Nanotechnology Reviews 8(1), 422–443 (2019). 270 13 G. Byeon, S.C. Jang, T. Roh, J.-M. Park, H.-S. Kim, and Y.-Y. Noh, “Recent progress in the 271 development of backplane thin film transistors for information displays,” Journal of Information 272 Display 24(3), 159–168 (2023). 273 14 I-Chung Chiu, Yun-Shiuan Li, Min-Sheng Tu, and I-Chun Cheng, “Complementary oxide–274 semiconductor-based circuits with n-channel ZnO and p-channel SnO thin-film transistors,” IEEE 275 Electron Device Lett. 35(12), 1263–1265 (2014). 276 15 Y.-S. Li, J.-C. He, S.-M. Hsu, C.-C. Lee, D.-Y. Su, F.-Y. Tsai, and I.-C. Cheng, “Flexible 277 complementary oxide–semiconductor-based circuits employing n-channel ZnO and p-channel SnO 278 thin-film transistors,” IEEE Electron Device Lett. 37(1), 46–49 (2016). 279 16 H. Yabuta, N. Kaji, R. Hayashi, H. Kumomi, K. Nomura, T. Kamiya, M. Hirano, and H. Hosono, 280 “Sputtering formation of p-type SnO thin-film transistors on glass toward oxide complimentary 281 circuits,” (2026). 282 17 J. Yang, Y. Wang, Y. Li, Y. Yuan, Z. Hu, P. Ma, L. Zhou, Q. Wang, A. Song, and Q. Xin, “Highly 283 optimized complementary inverters based on p-SnO and n-InGaZnO with high uniformity,” IEEE 284 Electron Device Lett. 39(4), 516–519 (2018). 285 18 Y. Ogo, H. Hiramatsu, K. Nomura, H. Yanagi, T. Kamiya, M. Hirano, and H. Hosono, “p-channel 286 thin-film transistor using p-type oxide semiconductor, SnO,” Applied Physics Letters 93(3), 032113 287 (2008). 288  9  19 Y.-L. Sun, T. Nabatame, J.W. Chung, T. Sawada, H. Miura, M. Miyamoto, and K. Tsukagoshi, 289 “Compositional changes between metastable SnO and stable SnO2 in a sputtered film for p-type thin-290 film transistors,” Thin Solid Films 807, 140548 (2024). 291 20 Y. Wang, G. Qiu, R. Wang, S. Huang, Q. Wang, Y. Liu, Y. Du, W.A. Goddard, M.J. Kim, X. Xu, P.D. 292 Ye, and W. Wu, “Field-effect transistors made from solution-grown two-dimensional tellurene,” Nat 293 Electron 1(4), 228–236 (2018). 294 21 G. Zhou, R. Addou, Q. Wang, S. Honari, C.R. Cormier, L. Cheng, R. Yue, C.M. Smyth, A. Laturia, 295 J. Kim, W.G. Vandenberghe, M.J. Kim, R.M. Wallace, and C.L. Hinkle, “High‐mobility helical 296 tellurium field‐effect transistors enabled by transfer‐free, low‐temperature direct growth,” Advanced 297 Materials 30(36), 1803109 (2018). 298 22 T. Kim, C.H. Choi, S.E. Kim, J.-K. Kim, J. Jang, S. Choi, J. Noh, K.-S. Park, J. Kim, S. Yoon, and 299 J.K. Jeong, “High-performance hexagonal tellurium thin-film transistor using tellurium oxide as a 300 crystallization retarder,” IEEE Electron Device Lett. 44(2), 269–272 (2023). 301 23 M. Kim, Y. Lee, K. Kim, G.-H. Pham, K. Kim, J.H. Jun, H. Lee, S. Yoon, H.J. Hwang, M.M. Sung, 302 and B.H. Lee, “Processes to enable hysteresis-free operation of ultrathin ALD Te p-channel field-effect 303 transistors,” Nanoscale Horiz. 9(11), 1990–1998 (2024). 304 24 H. Zhu, L. Fan, K. Wang, H. Liu, J. Zhang, and S. Yan, “Progress in the synthesis and application 305 of tellurium nanomaterials,” Nanomaterials 13(14), 2057 (2023). 306 25 T. Kawaguchi, Y. Tanaka, H. Sekiguchi, H. Hirai, K. Sato, and S. Kobayashi, “A novel method of 307 fabricating Te thin-film transistors,” Jpn. J. Appl. Phys. 25(3R), 449 (1986). 308 26 C. Zhao, C. Tan, D.-H. Lien, X. Song, M. Amani, M. Hettick, H.Y.Y. Nyein, Z. Yuan, L. Li, M.C. 309 Scott, and A. Javey, “Evaporated tellurium thin films for p-type field-effect transistors and circuits,” 310 Nat. Nanotechnol. 15(1), 53–58 (2020). 311 27 J. Robertson, X. Zhang, Q. Gui, and Y. Guo, “Amorphous TeO2 as p-type oxide semiconductor for 312 device applications,” Applied Physics Letters 124(21), 212101 (2024). 313 28 N. Devabharathi, S. Yadav, I. Dönges, V. Trouillet, and J.J. Schneider, “α‐TeO2 oxide as transparent 314 p ‐type semiconductor for low temperature processed thin film transistor devices,” Adv Materials Inter 315 11(16), 2301082 (2024). 316 29 S. Moufok, L. Kadi, B. Amrani, and K.D. Khodja, “Electronic structure and optical properties of 317 TeO2 polymorphs,” Results in Physics 13, 102315 (2019). 318 30 J. Robertson, X. Zhang, Q. Gui, and Y. Guo, “Amorphous TeO2 as p-type oxide semiconductor for 319 BEOL applications,” in 2024 International VLSI Symposium on Technology, Systems and Applications 320 (VLSI TSA), (IEEE, HsinChu, Taiwan, 2024), pp. 1–2. 321 31 S. Guo, Z. Zhu, X. Hu, W. Zhou, X. Song, S. Zhang, K. Zhang, and H. Zeng, “Ultrathin tellurium 322 dioxide: emerging direct bandgap semiconductor with high-mobility transport anisotropy,” Nanoscale 323 10(18), 8397–8403 (2018). 324  10  32 C. Niu, P. Tan, J.-Y. Lin, L. Long, Z. Lin, Y. Zhang, H. Wang, G.D. Wilk, and P.D. Ye, “First 325 demonstration of BEOL wafer-scale all-ALD channel CFETs using IGZO and Te for monolithic 3D 326 integration,” in 2024 IEEE International Electron Devices Meeting (IEDM), (IEEE, San Francisco, 327 CA, USA, 2024), pp. 1–4. 328 33 P. Tan, C. Niu, Z. Lin, J.-Y. Lin, L. Long, Y. Zhang, G. Wilk, H. Wang, and P.D. Ye, “Wafer-scale 329 atomic layer-deposited TeOx/Te heterostructure p-type thin-film transistors,” Nano Lett. 24(40), 330 12433–12441 (2024). 331 34 A. Liu, Y.-S. Kim, M.G. Kim, Y. Reo, T. Zou, T. Choi, S. Bai, H. Zhu, and Y.-Y. Noh, “Selenium-332 alloyed tellurium oxide for amorphous p-channel transistors,” Nature 629(8013), 798–802 (2024). 333 35 S. Tougaard, “Quantitative analysis of the inelastic background in surface electron spectroscopy,” 334 Surface & Interface Analysis 11(9), 453–472 (1988). 335  336  337 Figure Captions 338  339 Figure 1 340 (a) Photograph of a TeOx TFT. (b) Overall view of the TeOx TFT fabricated on a 15 mm square 341 Si substrate. The electrode spacing ranges from 50 to 350 nm. (c) Cross-sectional TEM image and (d) 342 a corresponding enlarged view of the sample with 15% W. The inset shows the nano-beam electron 343 diffraction pattern of the crystalline region. The majority of the film is amorphous, but some areas are 344 crystallized. Analysis revealed that the film was a metallic Te crystal. (e) Output and (f) transfer curves 345 for samples with 25% W and (g) output and (h) transfer curves for samples with 0% W. The sample 346 with 0% W is an insulator, whereas the sample with 25% W exhibits p-type transistor characteristics. 347  348 Figure 2 349 Correlation between the W-to-TeO2 weight ratio and (a) the field-effect hole mobility and (b) the on/off 350 current ratio. When the W ratio is 10% or less, the film acts as an insulator and no current flows. 351 However, when the W ratio is 15% or greater, the film exhibits TFT characteristics. When the W ratio 352 is 15% or greater, as a general trend, the mobility increases and the on/off ratio decreases with 353 increasing W ratio. 354  355 Figure 3 356 (a) High-resolution Te 3d core-level spectra for the sample with 25% W. The main peaks are observed 357 at the locations indicated by the red arrows (binding energy = 587, 576 eV), and the satellite peaks are 358 observed at the locations indicated by the blue arrows (binding energy = 583, 573 eV). These peaks 359 originate from the 3d3/2 and 3d5/2 orbitals of Te, respectively. The two main peaks originate from Te–360  11  O bonds, whereas the two satellite peaks originate from Te–Te bonds. The inset shows the XPS results 361 for the sample with 0% W prior to annealing. Satellite peaks are scarcely observed, indicating that 362 almost no Te–Te bonds are present. (b) The relationship between the weight ratio of W and the ratio 363 of Te–Te bonds. The white and black square values were calculated from the 3d5/2 and 3d3/2 peaks, 364 respectively. A trend is observed where the intensity ratio tends to increase with increasing W ratio. 365 That is, the greater the amount of W, the greater the proportion of Te–Te bonds. 366  367 Figure 4 368 (a) X-ray diffraction pattern of samples in which the W-to-TeO2 weight ratio is 15%, 35%, and 369 55%. The Miller indices are for the metallic Te crystal. All of the films exhibit TFT characteristics. (b) 370 X-ray diffraction pattern of the raw material powder remaining in the crucible after evaporation of the 371 film with 45% W. The black, blue, and red triangles indicate peaks originating from TeO2, W, and 372 WO3, respectively. WO3, which did not exist prior to vapor deposition, is present. W was oxidized by 373 consuming O2 from its surroundings. 374  375  376  377  378  379  380  381  382  383  384  385  386  387  388  389  390  391  392  393  394  395  396  12  Figure 1 397  398   399 Figure 2 400  401   402  403  13  Figure 3 404  405   406  407 Figure 4 408  409  410