# Fileset

[manuscript.pdf](https://mdr.nims.go.jp/filesets/048c24d0-567f-4aff-9ef1-e09cfa855342/download)

## Creator

[J. W. Liu](https://orcid.org/0000-0003-2580-7401), [T. Teraji](https://orcid.org/0000-0002-7731-0547), [B. Da](https://orcid.org/0000-0002-0785-8662), [Y. Koide](https://orcid.org/0000-0001-8321-9822)

## 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 J. W. Liu, T. Teraji, B. Da, Y. Koide; Calibration of binding energy and clarification of interfacial band bending for the Al2O3/diamond heterojunction. Appl. Phys. Lett. 2 September 2024; 125 (10): 101601 and may be found at https://doi.org/10.1063/5.0230817[In Copyright](http://rightsstatements.org/vocab/InC/1.0/)

## Other metadata

[Calibration of binding energy and clarification of interfacial band bending for the Al2O3/diamond heterojunction](https://mdr.nims.go.jp/datasets/ecc79f15-2ec8-4f0c-b53c-2e8e62a5dfef)

## Fulltext

Microsoft Word - manuscript71  Calibration of binding energy and clarification of interfacial band bending for 1 Al2O3/diamond heterojunction 2  3 J. W. Liu,1, a) T. Teraji,1 B. Da,2 and Y. Koide1 4 1Research Center for Electronic and Optical Materials, National Institute for Materials 5 Science (NIMS), 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan  6 2Research and Services Division of Materials Data and Integrated System, NIMS, 1-1 7 Namiki, Tsukuba, Ibaraki 305-0044, Japan  8  9 a) Author to whom correspondence should be addressed; electronic mail: 10 liu.jiangwei@nims.go.jp 11  12  13  14  15  16  17  18  19  20  21  22  23  24 2  Abstract 1 Due to the presence of an intrinsic C 1s peak in diamond, it is impossible to calibrate 2 its binding energies using the adventitious C 1s peak (284.8 eV) during X-ray 3 photoelectron spectroscopy measurement. The absence of accurate binding energy 4 measurement makes it challenging to determine the interfacial band bending for the 5 oxide/diamond heterojunction. To overcome this issue, a net-patterned gold (Au) mask is 6 applied to the boron-doped diamond (B-diamond) to suppress the charge-up effect and 7 calibrate the binding energy using the standard Au 4f peak (83.96 eV). The B-diamond 8 epitaxial layer shows downward band bending towards the surface with valence band 9 maximum of 0.85 eV. Upon the formation of Al2O3 using an ozone precursor through the 10 atomic layer deposition technique, the B-diamond continues to exhibit downward band 11 bending towards the Al2O3/B-diamond interface. However, the bending energy has 12 reduced, potentially attributed to the modification of the oxygen vacancies on the B-13 diamond surface by the ozone precursor during the Al2O3 deposition. 14  15  16  17  18  19  20  21  22  23  24 3  Carbon-related material of diamond has unique properties such as wide bandgap 1 energy, high thermal conductivity, high breakdown voltage, good chemical stability, and 2 radiation hardness.1-3 These properties make it suitable for applications in fields of high-3 power, high-frequency, high-temperature, and harsh environments. Diamond-based 4 electronic devices, such as diamond metal-oxide-semiconductor (MOS) capacitors4, 5 and 5 MOS field-effect transistors (MOSFETs),6-8 have been developed and show promise in 6 high-power, high-frequency, and high-temperature applications.  7 For fabricating high-performance diamond-based MOS capacitors and MOSFETs, 8 understanding the band configuration at the oxide/diamond interface is crucial.9 It can 9 help in optimizing the leakage current density and charge storage properties for the 10 diamond MOS capacitors. It can also help to understand the threshold voltage, charge 11 injection and extraction, and device stability for the diamond MOSFETs.10 Thus, 12 clarification the band configuration for the insulator/diamond interface is important for 13 designing and fabricating more efficient and reliable diamond-based MOS electronics. 14 There are several techniques used to investigate the band configurations at the 15 oxide/semiconductor interface, such as X-ray photoelectron spectroscopy (XPS),11, 12 16 capacitance-voltage measurement,13 current-voltage measurement,14 and Kelvin probe 17 force microscopy.15 Comparing other techniques, the XPS offers valuable advantages 18 such as providing chemical information, band bending analysis, non-destructive 19 measurements, and compatibility with different materials.  20 Band offsets for different oxide insulators on hydrogen-terminated diamond and 21 boron-doped diamond (B-diamond) have been demonstrated using the XPS technique.16, 22 17 However, the presence of the intrinsic C 1s peak in diamond makes it impossible to 23 calibrate the binding energy using the adventitious C 1s peak (284.8 eV), which makes it 24 4  difficult to clarify the interfacial band bending for the oxide/diamond heterojunctions.  1 In this letter, a net-patterned gold (Au) mask is formed on the B-diamond to suppress 2 the charge-up effect and allows for the calibration of the binding energies using the 3 standard Au 4f peak (83.96 eV). The surface band bending for the B-diamond and 4 interfacial band bending for the Al2O3/B-diamond heterojunction would be clarified. 5 Figure 1 illustrates the process of forming the net-patterned Au mask and the Al2O3 6 insulator on the B-diamond. The Ib-type (100) diamond substrate was immersed in a 7 solution of H2SO4 + HNO3 at 300 °C for 3 hours to clean the surface. The B-diamond 8 epitaxial layer was then grown using microwave plasma-assisted chemical vapor 9 deposition technique [Fig. 1(a)]. The microwave power, temperature, and chamber 10 pressure were maintained at 1.4 kW, approximately 1000 °C, and 18.6 kPa, respectively.18 11 The boron source used was the residual boron in the chamber from the previous B-12 diamond growth. The flow rates for the source gases of H2 and CH4 were set at 49 and 1 13 sccm, respectively. The thickness and boron atom concentration of the B-diamond 14 epitaxial layer was measured by secondary ion mass spectroscopy and found to be 825 15 nm and 4 × 1015 cm–3, respectively.  16 The B-diamond epitaxial layer was treated in an acid solution of H2SO4 + HNO3 at 17 300 °C for 3 hours again, resulting in a transformation of its hydrogen-terminated surface 18 to an oxygen-terminated surface. Following the treatment, the oxygen-terminated B-19 diamond epitaxial layer was cleaned sequentially with acetone, ethanol, and pure water. 20 The first-time high-resolution XPS measurement (Instrument: PHI Quantera SXM, 21 ULVAC-PHI) was conducted. Valence band and core level spectra were obtained for the 22 B-diamond without the Au mask. The X-ray source employed was monochromate Al Kα 23 (hv =1486.6 eV) with a take-off angle of 45°, a power output of 50 W, a measurement 24 5  area of 200 μm, a pass energy of 55 eV, and an energy step of 0.1 eV. 1 A laser lithography system and electron-gun evaporation equipment were utilized to 2 create the net-patterned Au mask on the B-diamond [Fig. 1(b)]. The B-diamond was 3 coated sequentially with a positive photoresist (LOR5A) and an image reversal 4 photoresist (AZ5214E) using a spin-coater. The spin speed and time for coating both 5 photoresists were set at 7000 rpm and 1 second, respectively. The LOR5A was baked at 6 180 °C for 5 minutes, while the AZ5214E was baked at 110 °C for 2 minutes. 7 Subsequently, it underwent exposure and development processes using a DL-1000 8 scanning maskless lithography system and a tetramethylammonium hydroxide (TMAH) 9 solution with a concentration of 2.38%. The developing time in the TMAH solution was 10 2.5 minutes. A 100 nm-thick Au mask and position correction cross patterns (“+”) were 11 formed on the B-diamond. The evaporation chamber pressure was approximately 10–6 Pa 12 with an evaporation rate of 2 Å/s. The second time XPS measurement was conducted to 13 obtain core level spectra for the B-diamond with the Au mask. 14 A 3 nm-thick Al2O3 insulator was deposited on the entire surface of the B-diamond 15 with the Au mask using an atomic layer deposition system. The deposition involved the 16 use of Al(CH3)3 and ozone precursors at a temperature of 200 °C [Fig. 1(c)]. Subsequently, 17 a window was opened in the Al2O3 film to expose the Au mask by etching with the TMAH 18 solution at room temperature for 12 minutes [Fig. 1(d)]. The third time XPS measurement 19 was then performed to capture the core level spectra for the Al2O3 and B-diamond with 20 the Au mask. It should be noted that a 30 nm-thick Al2O3 film was deposited on a silicon 21 substrate to provide additional information on the valence band and core level spectra, 22 which can be calibrated with the adventitious C 1s peak (284.8 eV) and were crucial for 23 deducing the band offsets of the Al2O3/B-diamond heterojunction. 24 6  Figures 2(a) and 2(b) display the scanning electron microscopy image and scanning 1 X-ray image, respectively, of the Al2O3 (3 nm)/B-diamond with the Au mask. The B-2 diamond exhibits a net-patterned Au mask with 10 position correction cross patterns (“+”). 3 The side length of the square grid is 200 μm, and the center of the X-ray is positioned at 4 the Al2O3/B-diamond. The spectra for Au, Al2O3, and B-diamond can be simultaneously 5 obtained from this configuration. 6 Figure 3(a) presents the as-received C 1s spectra for the B-diamond without (black 7 line) and with (red line) the Au mask. The corresponding binding energies are measured 8 as 283.23 eV and 284.50 eV, respectively. The binding energy difference (∆E) of 1.27 eV 9 indicates the presence of a charge-up effect19 on the B-diamond epitaxial layer with the 10 low doping level. In Fig. 3(b), the as-received Au 4f7/2 spectra are shown for the B-11 diamond and the Al2O3/B-diamond samples with the Au mask. For the B-diamond with 12 the Au mask, the binding energy of the as-received Au 4f7/2 spectrum is measured as 83.97 13 eV, which closely matches the ideal value of 83.96 eV. This suggests that there is almost 14 no charge-up effect on the B-diamond with the Au mask during the XPS measurement. 15 However, for the Al2O3/B-diamond with the Au mask, the binding energy shifts to a 16 higher value of 84.15 eV. This shift arises from the accumulation of negative charges on 17 the sample surface due to the charge-up effect. Despite the presence of the Au mask, the 18 charge-up effect on the Al2O3/B-diamond cannot be eliminated.  19 Therefore, the presented results indicate the existence of a charge-up effect on the 20 low doping level B-diamond epitaxial layer, as evidenced by the significant binding 21 energy difference between the spectra without and with the Au mask. While the charge-22 up effect is minimal for the B-diamond sample with the Au mask, it remains pronounced 23 for the Al2O3/B-diamond sample, necessitating the calibration of peak binding energies 24 7  for accurate analysis. 1 Figure 4(a) shows the calibrated C 1s spectra for B-diamond and Al2O3(3 nm)/B-2 diamond with the Au mask. The main peak corresponding to carbon-carbon (C-C) bonds 3 in the C 1s spectrum of B-diamond is observed at a binding energy of 284.50 eV. Upon 4 depositing Al2O3 on B-diamond, the peak shifts to a lower binding energy of 284.15 eV, 5 resulting in a shift of ∆E = 0.35 eV. Additionally, a peak at approximately 284.5 eV in the 6 Al2O3(3 nm)/B-diamond spectrum is attributed to carbon-boron (C-B) bonds.20 In Fig. 7 4(b), the calibrated valence band spectra for B-diamond without an Au mask and Al2O3(30 8 nm)/Si are shown. Due to the potential impact of the Au mask on determining the valence 9 band maximum (VBM) of B-diamond, we opted to use B-diamond samples without an 10 Au mask for VBM determination. By extrapolating linear fittings for the leading edge to 11 the baseline, the valence band maximum (VBM) values are determined to be 0.85 eV for 12 B-diamond and 3.38 eV for Al2O3 (30 nm). Fig. 4(c) displays the calibrated Al 2p spectra 13 for the Al2O3(3 nm)/B-diamond and Al2O3(30 nm)/Si samples. The main peaks 14 corresponding to aluminum-oxygen (Al-O) bonds are observed at binding energies of 15 72.99 eV and 74.57 eV, respectively. In the Al2O3(30 nm)/Si spectrum, a peak at 16 approximately 77.0 eV is attributed to surface aluminum-oxygen-hydroxide (Al-O-OH) 17 bonds.21 18 The valence band offset (VBO) of the Al2O3/B-diamond heterojunction can be 19 calculated using equation (1) below:  20 𝑉𝐵𝑂 =  (𝐸   −  𝐸 ) −  (𝐸  − 𝐸 )  (  )21 −  (𝐸  − 𝐸  )  (  ) 22 The term (𝐸   −  𝐸 )  represents the difference in binding energy between 23 the C 1s level and the VBM for the B-diamond, which has a value of 283.65 eV. The term 24 8  (𝐸  − 𝐸 )  (  ) represents the difference in binding energy between the Al 1 2p level and the VBM for a 30 nm thick Al2O3 layer, which is 71.19 eV. Finally, the term 2 (𝐸  − 𝐸  )  (  ) represents the difference in binding energy between the C 1s 3 level and the Al 2p level for a 3 nm thick Al2O3 layer, which amounts to 211.16 eV. By 4 substituting the given values into equation (1), the VBO for the Al2O3/B-diamond 5 heterojunction can be calculated as 1.30 eV, which is in good agreement with the previous 6 report.16 Additionally, based on the provided bandgap energies of diamond (5.47 eV) and 7 Al2O3 (7.2 eV),11 the conduction band offset (CBO) for the Al2O3/B-diamond 8 heterojunction can be inferred to be 0.43 eV. 9 Figure 4(d) illustrates schematic band diagram for the Al2O3/B-diamond 10 heterojunction, exhibiting a type I straddling band configuration. Based on a doping 11 concentration of 4 × 1015 cm–3, the Fermi level potential in the bulk B-diamond is 12 estimated to be approximately 0.37 eV.22 In the case of B-diamond, the presence of 13 oxygen vacancies or native defects results in positive charges on the surface, causing a 14 downward band bending towards the surface with the VBM of 0.85 eV. Upon the 15 formation of Al2O3 using an ozone precursor through the ALD technique, the B-diamond 16 continues to exhibit downward band bending towards the Al2O3/B-diamond interface. 17 However, the bending energy has reduced, potentially attributed to the modification of 18 the ozone precursor with regards to the oxygen vacancies. 19 In conclusion, a new technique was developed to obtain precise binding energies for 20 B-diamond and Al2O3/B-diamond. The band bending and band configuration for the 21 Al2O3/B-diamond were clarified. An Au mask was applied to the B-diamond to suppress 22 the charge-up effect and calibrate the binding energy using the standard Au 4f peak. The 23 B-diamond shows downward band bending towards the surface with the VBM of 0.85 eV. 24 9  Upon the formation of Al2O3 using an ozone precursor through the ALD technique, the 1 B-diamond continues to exhibit downward band bending towards the Al2O3/B-diamond 2 interface. However, the extent of bending has reduced. This technique is also applicable 3 for obtaining precise binding energies for other carbon-related materials. 4  5 This work is supported by the JSPS KAKENHI Projects (JP23K03966, 20H05661, 6 and JP20H00313), MEXT Q-LEAP (JPMXS0118068379), JST CREST (JPMJCR1773), 7 JST Moonshot R&D (JPMJMS2062), MIC R&D for construction of a global quantum 8 cryptography network (JPMI00316), and ARIM (JPMXP1223NM5006) of the Ministry 9 of Education, Culture, Sports, Science and Technology, Japan. 10  11  12 Data Availability Statements 13 The data that support the findings of this study are available from the corresponding 14 author upon reasonable request. 15  16  17  18  19  20  21  22  23  24 10  References 1 1. M. Geis, Mat. Res. Soc. Symp. Proc. 162, 15 (1990). 2 2. C. J. H. Wort and R. S. Balmer, Mater. Today 11, 22 (2008). 3 3. H. Umezawa, Mater. Sci. Semi. Proc. 78, 147 (2018). 4 4. J. Liu, M. Liao, M. Imura, H. Oosato, E. Watanabe, and Y. Koide, Appl. Phys. Lett. 5 102, 112910 (2013). 6 5. J. Liu, H. Oosato, B. Da, and Y. Koide, Appl. Phys. Lett. 117, 163502 (2020). 7 6. K. Hirama, H. Sato, Y. Harada, H. Yamamoto, and M. Kasu, Jpn. J. Appl. Phys. 51, 8 090112 (2012). 9 7. W. Wang, Y. Wang, M. Zhang, R. Wang, G. Chen, X. Chang, F. Lin, F. Wen, K. Jia, 10 and H. X. Wang, IEEE Electron Dev. Lett. 41, 585 (2020). 11 8. W. Fei, T. Bi, M. Iwataki, S. Imanishi, and H. Kawarada, Appl. Phys. Lett. 116, 12 212103 (2020). 13 9. J. Robertson, J. Vac. Sci. Technol. B 18, 1785 (2000). 14 10. J. Goo, Q. Xiang, Y. Takamura, F. Arasnia, E. N. Paton, P. Besser, J. Pan, and M. 15 Lin, IEEE Electron Dev. Lett. 24, 568 (2003). 16 11. J. W. Liu, M. Y. Liao, M. Imura, and Y. Koide, Appl. Phys. Lett. 101, 252108 (2012). 17 12. J. Cañas, G. Alba, D. Leinen, F. Lloret, M. Gutierrez, D. Eon, J. Pernot, E. Gheerae18 rt, and D. Araujo, Appl. Surf. Sci. 535, 146301 (2021). 19 13. Y. Nemirovsky, G. Gordon, and D. Goren, J. Appl. Phys. 84, 1113 (1998). 20 14. Z. Xu, M. Houssa, S. D. Gendt, and M. Heyns, Appl. Phys. Lett. 80, 1975 (2002). 21 15. W. Melitz, J, Shen, A. C. Kummel, and S. Lee, Surf. Sci. Reports 66, 1 (2011). 22 16. A. Marechal, M. Aoukar, C. Vallee, C. Riviere, D. Eon, J. Pernot, and E. Gheeraert, 23 Appl. Phys. Lett. 107, 141601 (2015) 24 17. J. Liu, and Y. Koide, Sensors 18, 1813 (2018). 25 18. T. Teraji, T. Yamamoto, K. Watanabe, Y. Koide, J. Isoya, S. Onoda, T. Ohshima, L. 26 J. Rogers, F. Jelezko, P. Neumann, J. Wrachtrup, and S. Koizumi, Physica Status 27 Solidi A 212, 2365 (2015). 28 19. J. A. C. Santana, Quantitative Core Level Photoelectron Spectroscopy, Online 29 ISBN: 978-1-6270-5306-8, Morgan & Claypool Publishers (2015) 30 11  20. T. Yokoya, E. Ikenaga, M. Kobata, H. Okazaki, K. Kobayashi, A. Takeuchi, A. Awaji, 1 Y. Takano, M. Nagao, I. Sakaguchi, T. Takenouchi, K. Kobayashi, H. Kawarada, and 2 T. Oguchi, Phys. Rev. B 75, 205117 (2007). 3 21. W. Liang, K. J. Weber, D. Suh, J. Yu, and J. Bullock, 2013 IEEE 39th Photovoltaic 4 Specialists Conference, Tampa, FL, USA,16-21 June 2013. 5 22. A. T Collins, J. Phys.: Condens. Matter 14, 3743 (2002). 6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25 12  Figure captions 1  2 FIG. 1. Process of forming the Au mask and the Al2O3 insulator on the B-diamond: (a) 3 Diamond cleaning and B-diamond growth, (b) net-patterned Au mask, (c) Al2O3 4 deposition, and (d) opening windows for the Au mask. 5  6 FIG. 2. (a) Scanning electron microscopy image and (b) scanning X-ray image for 7 Al2O3/B-diamond with the Au mask. 8  9 FIG. 3. (a) As-received C 1s spectra for the B-diamond without (black line) and with (red 10 line) the Au mask, and (b) as-received Au 4f7/2 spectra for the B-diamond and the Al2O3/B-11 diamond samples with the Au mask.  12  13 FIG. 4. (a) Calibrated C 1s spectra for B-diamond and Al2O3(3 nm)/B-diamond with the 14 Au mask, (b) calibrated valence band spectra for B-diamond without an Au mask and 15 Al2O3(30 nm)/Si sample, (c) calibrated Al 2p spectra for the Al2O3(3 nm)/B-diamond and 16 Al2O3(30 nm)/Si samples, and (d) schematic band diagram for the Al2O3/B-diamond 17 heterojunction. 18  19  20  21 13   1  2  3 Liu et al., Figure 1 4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19 B-diamond Au Al2O3(a) (b) (c) (d)14   1  2  3 Liu et al., Figure 2 4  5  6  7  8  9  10  11  12  13  14  15  16  17 15   1  2  3 Liu et al., Figure 3 4  5  6  7  8  9  10  11  12  13  14  15  16  17 288 287 286 285 284 283 282 281 280    B-diamond without maskB-diamond with mask∆E=1.28 eVBinding energy (eV)Intensity (arb. units)(a) As-received C 1s86 85 84 83 82    B-diamond with maskAl2O3 (3 nm)/B-diamond with maskBinding energy (eV)Intensity (arb. units)(b) As-received Au 4f7/2∆E=0.18 eV16   1  2  3  4 Liu et al., Figure 4 5  6  7 288 287 286 285 284 283 282 281 280    Binding energy (eV)Intensity (arb. units)(a) Calibrated C 1sB-diamond with maskAl2O3 (3 nm)/B-diamond with mask∆E= 0.35 eVC-B20 18 16 14 12 10 8 6 4 2 0 -2  B-diamond without maskAl2O3(30 nm)/Si(b) Calibrated valance band0.85 eV3.38 eVBinding energy (eV)Intensity (arb. units)80 78 76 74 72 70  Al2O3 (3 nm)/B-diamond with maskAl2O3(30 nm)/SiBinding energy (eV)Intensity (arb. units)(c) Calibrated Al 2p∆E= 1.58 eVAl-O-OH(d) Band configurationCBMVBMEF0.35 eV0.37 eVCBMVBMCBO=0.43 eVVBO=1.30 eV0.85 eVB-diamondAl2O30.35 eV