n-Hg3In2Te6 Schottky contact

Accepted Manuscript Title: Effects of electrochemical sulfide passivation on electrical properties of Au/n-Hg3 In2 Te6 Schottky contact Authors: Qibin Guo, Li Fu, Haiyan Chen, Yapeng Li, Haizhao Zheng PII: DOI: Reference:

S0368-2048(16)30115-3 http://dx.doi.org/doi:10.1016/j.elspec.2017.03.019 ELSPEC 46659

To appear in:

Journal of Electron Spectroscopy and Related Phenomena

Received date: Revised date: Accepted date:

8-9-2016 23-1-2017 30-3-2017

Please cite this article as: Qibin Guo, Li Fu, Haiyan Chen, Yapeng Li, Haizhao Zheng, Effects of electrochemical sulfide passivation on electrical properties of Au/n-Hg3In2Te6 Schottky contact, Journal of Electron Spectroscopy and Related Phenomenahttp://dx.doi.org/10.1016/j.elspec.2017.03.019 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effects of electrochemical sulfide passivation on electrical properties of Au/n-Hg3In2Te6 Schottky contact Qibin Guo, Li Fu, Haiyan Chen, Yapeng Li, Haizhao Zheng

State Key Laboratory of Solidification Processing, School of Materials Science and Engineering, Northwestern Polytechnical University, Xi’an 710072, China

Highlights 1 This passivation method reduces dangling bonds on the MIT surface by forming In-S, Hg-S, Te-O bonds. 2 The Te-rich layer is removed by the dissolution of tellurides produced in sulfide-passivation process. 3 The work functions of MIT are 4.19 eV and 4.33 eV before and after passivation respectively due to the decrease of surface states. 4 The sulfide passivation for Au/MIT Schottky contacts can enhance the rectifying behavior and suppress the leakage current.

Abstract The influence of electrochemical sulfide passivation on the electronic properties of Au/n-Hg3In2Te6 Schottky contact was analyzed by photoelectron spectroscopy and I-V measurement. Through photoelectron spectroscopy, it was obtained that the element concentration of Te decreased and In-S and Hg-S bonds appeared in surface after passivation. Meanwhile, Fermi level was depinned due to the sulfide passivation, which is proved by the rise of work function. In addition, the leakage current of Au/n-Hg3In2Te6 Schottky contact reduced, owing to the decrease in the density of surface states. Thus this method is of effective capability to suppress leakage current and modify rectifying behavior.

Key words: electrochemical sulfide passivation, work function, Fermi-level pinning, surface states



Author to whom correspondence should be addressed. Electronic mail: [email protected]

1

1. Introduction As for Hg3In2Te6 (short for MIT) single crystal, it has unique properties, such as anti-irradiation performance, working at room temperature and the detecting ability in near-infrared waveband, which make it suitable for the fabrication of near-infrared photodetectors [1~5]. Generally, the performance of photodetectors based on MIT is especially sensitive to the chemical states and electronic structure of the surface [6]. The surface states at the surface of semiconductor can deteriorate the surface recombination and pin Fermi level, which can lead to the increase of surface leakage current and the degradation of detecting performance [7~9]. In order to improve the stability and reliability of photodetectors, many passivation methods have been applied to suppress surface states and cut down dangling bonds [10~13]. On the other hand, it can also decrease the density of interface states at the metal-semiconductor contact and weaken Fermi-level pinning [14~16]. Therefore, A suitable passivation technique can lower fixed surface charges to prevent the generation of inversion-layer and the accumulation of surface charges which can elevate surface leakage current including generation-recombination and tunneling current [17~19]. After chemically polishing by bromine solution, surface of telluride semiconductors are mostly found rich in Te, which is considered as a major cause of increased surface leakage current, but previous passivation methods have not adequately solved this dilemma [20~22]. Hu et al [23] carried out in-situ CdTe deposition and hydrogen plasma modification for HgCdTe, and this hybrid passivation mothed can reduce the trap-assisted tunneling current due to the removal of structural defects by hydrogen plasma modification. Lin et al [24] studied the SiO2 epitaxial film passivation on HgCdTe by direct photochemical-vapor deposition, and by this means a smaller interface state density and a higher thermal stability could be implemented. Wet chemical method of surface passivation is inherently simpler than other ways. Lee et al [25] comes to a conclusion that the quantum efficiency of CdS quantum dots with H2O2 wet passivation was two orders of magnitude higher than unpassivated samples, because the treatment could promote radiative recombination rate. Howard et al [26] and Hou et al [27] developed an anodic sulfide-passivation technique for AlGaInP and GaAs, and the results indicated that this electrochemical method could significantly reduce the surface recombination velocity and obtain a stable passivated surface over ordinary passivation methods. However, there were few papers about passivation methods for MIT surface. 2

The surface passivation for MIT is a complicated process because of the difference in chemically related properties of compound and the tendency of electrically active defects [28] to form at surface during the passivating process, yet there is few reports about this aspect. In order to enhance the performance of MIT-based infrared detectors and broaden the prospects for applications, we carried out the electrochemical sulfide-passivation method which does not require complex epitaxial processing on MIT surface. In this study, photoelectron spectroscopy was employed to examine the core level on the surface of MIT crystal after passivation, and the element concentration was obtained by the means of semi-quantitative analysis. Meanwhile, the effects of sulfide passivation on work function of MIT were also studied in detail. Furthermore, the electrical characteristics of Au/MIT contacts with and without passivation were measured by I-V experiments. Based on the Schottky model, the influence mechanism of sulfide passivation on the electrical properties of Schottky contact was analyzed.

2. Experiment In this study, the n-MIT bulk ingot was grown using the vertical Bridgman melt growth method and cut into wafers (5×5×1 mm3). The samples were mechanically polished and then chemically etched by dipping in the 2 vol% Br2-C3H7ON solution for 2 min, which resulted in the removal of surface damage layer. Then the wafers were successively rinsed with de-ionized water and blown dry with N2 flow. The electrochemical sulfide passivation setup consists of a platinum sheet as the cathode, a platinum sample clip as the anode, a constant voltage power supply, and ethylene glycol solution with saturated Na2S as electrolyte. A DC bias voltage of 30 volts was applied on the MIT samples fixed in the electrode clip for 20 min. After being rinsed in ethanol and de-ionized water successively for a short time (about 3 s), the samples were immediately put into the HHV auto 306 Thermal Evaporation instrument for the preparation of Au electrode at the pressure about 5×10-4 Pa. The Schottky contacts were fabricated by depositing Au film on the front surface of MIT wafers. In electrodes were deposited on the back side of the wafers to form ohmic contact. The electrical properties of the Au/MIT/In Schottky contacts were analyzed by the Agilent 4155C semiconductor parameter analyzer. The ESCALAB 250Xi XPS with high resolution of 0.45 eV was utilized to analyze the core-level binding energy and chemical composition of sulfide-passivation layer. The XPS instrument used in this experiment employs a monochromatised Al Kα X-ray source (hν=1486.7 eV). Meanwhile, the 3

measurement of work functions of the passivated and unpassivated MIT samples were carried out by UPS equipped with a He I light source (hν=21.2 eV) and a hemispherical energy analyzer. The Au 4f peak and the Fermi level of pure gold foil, as reference standards, were applied to calibrate all XPS and UPS spectra, respectively. The chemical states of each element in the passivated layer were investigated by the XPS depth profile. The sulfide-passivated MIT wafers were etched by Ar+ bombardment (at about 5×10−8 mbar) with a voltage of 1.0 kV and a current of 1.7 μA. The etching rate of the MIT surface was about 0.8 nm·min−1. Slightly Ar+ bombardment was applied to remove surface contamination. All samples were free of carbon after etching for 60 s.

3. Results and Discussion 3.1 The core level and element concentration of MIT samples after passivation Atomic concentrations and chemical states of MIT surface before and after the electrochemical sulfide passivation were investigated by XPS. As shown in Fig.1(a), the full spectrum shows that the intensity of Te peaks decreased significantly after sulfide passivation, especially Te 3d and Te 4d. The semi quantitative analysis of Fig.1(a) reveals that the Te content dropped from 45.3% to 0.8% after passivation, indicating that the generated telluride (TeO2 or TeO2-3) on the passivated MIT surface was dissolved during the passivation process. In Fig.1(b), the binding energy of Te 3d5/2 corresponding to Te-In and Te-O state were about 572.8 eV and 576.4 eV, respectively. As for MIT samples without passivation, a part of the Te on surface were oxidized into TeO2 by oxygen in the air. Only a minimal amount of TeO2 left on the passivated MIT surface, besides, the content of Te in Te-In state was much lower than the unpassivated surface.

Fig.1(c) shows that the In 3d5/2 core-level moves to the direction of high binding energy after sulfide passivation. Via XPS-peak-differentation-imitating analysis, the 3d5/2 core-level of the passivated MIT in three states were obtained: In-Te at 444.8 eV, In-S at 445 eV, and In-O at 445.6 eV, these results are in good agreement with the data from reported literatures [29~31]. More than half In atoms in the detected layer were bonded to S atoms, and few were bonded to O atoms. Only the In 3d5/2 peak at 444.8 eV in the state of In-Te appeared on the unpassivated MIT surface, which indicates that the In atoms were almost not sulfided or oxidized before passivation. In addition, Fig.1(d) illustrates 4

that Hg 4f core-level shifted towards high binding energy. It is concluded that Hg elements on passivated MIT surface were bonded to S or O atoms in the electrochemical passivation process. To monitor the change of chemical states of sulfide-passivation layer along depth direction, passivated MIT samples were etched by Ar+ beam in situ XPS measurement. The Te 3d core-level spectra in Fig.2 reveals that Te4+ peak at 576.3 eV (Te 3d5/2) and 586.7 eV (Te 3d3/2), Te0 peak at 572.8 eV (Te 3d5/2) and 583.2 eV (Te 3d3/2) are present for the samples with sulfide passivation. The Te4+ peaks are related to TeO2 or TeO2-3, and Te0 peaks are related to InTe and HgTe. On the surface of passivated MIT without etching, the Te 3d peak is extremely weak due to the low content of Te (about 0.8%). With the increase of etching time, the intensity of Te4+ peaks became stronger gradually until etching for 150 s, but from then on, the intensity decayed until Te4+ peak completely disappeared after etching for 630 s. The Te-depth analysis indicates that the content of TeO2 (or TeO2-3) increased firstly and then decreased as etching depth increases, which was caused by the dissolution of TeO2 (or TeO2-3) produced by passivation reaction.

As shown in Fig.3, In 3d core levels were studied by means of XPS-peak-differentation-imitating analysis. Fig.3(a) provides the In 3d5/2 core-level spectra of sulfide-passivated samples, and there are In-Te peak (444.8 eV), In-S peak (445.0 eV) and In-O peak (445.6 eV). The binding energy of In in the oxidation state (In-O peak) is +0.8 eV higher than In in the In-Te state, as shown in Fig.3(a), however, In in the sulfidation state (In-S peak) get a slight rise of about +0.2 eV. 50.4% of In atoms in the detected layer were bonded to S atoms after sulfide-passivation, and 14.2% of In atoms were bonded to O atoms, besides the rest of 35.4% kept the same state as In atoms in In-Te bonds. As shown from Fig.3(a) to Fig.3(f), the concentration of In atoms in the state of In-S decreases with the increase of etching time, and the concentration of In atoms in the state of In-Te presents the opposite case. The depth-analysis also reveals that In-O peak almost disappears after ething for 60 s, but in Fig.3(c), the intensity of In-O peak improves peculiarly after ething for 90 s, which is due to that O content suddenly increases at this time and excess O atoms are bonded with In atoms. This correlates well with the measured increase of O content over the same etching time, as illustrated in Table.1. Fig.4(a) illustrates that as the etching time increases up to 150 s, the binding energy of Hg 4f slightly shifts toward higher energy direction for about 0.1 eV compared to the time before etching, which demonstrates that Hg atoms were partially sulfided and oxidized during the passivating process. 5

Moreover, the binding energy of Hg 4f shifts toward lower energy direction as the etching time further increases from 150 to 630 s. Therefore, it is concluded that the concentrantion of Hg-S and Hg-O reduced over the same etching time, which correlates well with the decrease of S and O content as shown in Table.1. In addition, the presence of S at the MIT surface is confirmed by Fig 4(b) which indicates that S 2p core level have a pair of spin-orbit split doublet, 162.8 eV (S 2p1/2) and 161.6 eV (S 2p3/2) respectively.

As illustrated in Table 1, the content of O increases with prolonging of the Ar+ etching from 0 to 150 s, which is resulted from that MIT samples have reacted with H2O at anode. A small amount of H2O was unavoidably dissolved from hydration Na2S in electrolyte solution. According to reported literature, MIT reacted with O0 to form TeO2 or TeO2-3. However, previous passivation methods for telluride semiconductor materials mostly keep the presence of Te-rich layer as a major cause of the increased surface leakage current [21]. The composition-depth profile of passivated samples in this study demonstrates that the content of Te on the surface was extremely low, only about 0.8%, and increased with the increase of etching time. This phenomenon resulted from the dissolution of TeO2 and TeO2-3 on the passivated MIT surface in the electrolyte solution, which was additionally proved by the trend of variation in the content of Te4+ and O. The ratio of Te4+: Te0 increased from 0.56:1 to 2.51:1 along with etching time within 0 to150 s. Therefore, electrochemical sulfide passivation of MIT can be applied as an effective means to suppress the formation of Te-rich layer. In addition, there were only In, Te and Hg atoms on the surface of passivated samples after Ar ion beam sputtering for 630 s, and then the samples are considered to develop freshly cleaved surface.

3.2 The work function of MIT crystal after passivation. Fig.5 illustrates the typical UPS spectra, as well as the details of Fermi Edge and inelastic secondary electron cutoff on the surface of MIT with and without sulfide passivation. The inelastic cutoff can be distinguished in Fig.5(b) due to that the samples were taken for -5.0 V bias [32].The 6

positions of Fermi edge, namely the center of the slopes, are both at 26.18 eV, which were checked by Au thin films deposited on the partial area of MIT surface. The work functions are calculated by the formula [33]:

h    EF  Ecutoff

(1)

where φ is the work function, EF and Ecutoff are the kinetic energy of Fermi edge and the inelastic cutoff, respectively.

According to Eq (1), the work functions are calculated about 4.19 eV and 4.33 eV before and after sulfide-passivation, respectively. There is an increase about 0.14 eV of work function after passivation process. It is well known that the Fermi-level pinning is affected by many factors from the previous literature, such as the dangling bonds and Te-rich layer [14~16, 34~36]. Dangling bonds can be eliminated by S atoms which are bonded with In and Hg atoms, as demonstrated in Fig.3 and Fig.4. In addition, the surface defects are simultaneously reduced due to the dissolution of Te-rich layer. Thus, the density of surface states decreases and the effect of Fermi-level pinning is consequently alleviated, which leads to the increase of work function for the passivated surface of MIT samples. 3.3 The electrical properties of Au/ MIT Schottky contacts after sulfide-passivation.

From Fig.6, the ratio of current of Au/MIT contacts at 1 V to -1 V (I1v: I-1v) is raised from 1.8 to 5.1 after sulfide passivation. According to Cowely-Sze model, the more obvious rectifying behavior with passivation vice-versa demonstrates that the sulfide treatment has the efficacy to depin the Fermi level [37]. The leakage current at -1 V decreases by one order of magnitude after sulfide passivation, which is resulted from the reduction of interface states [38], fixed charges [39] and other surface defects [40] as proved above. In addition, the chemical composition of MIT surface is transformed from from Te-rich to Te-lack after passivation, which contributes to suppress the leakage current of Au/MIT Schottky contacts [41]. Note that there is a larger difference of electronegativity (short for EN) between In (EN=1.7) and S (EN=2.5), so that the bonding behavior of In atoms tends to shift from In-Te (Te: EN=2.1) bonds with covalent properties to In-S bonds with ionic properties on the sulfide-passivated MIT surface [33]. Likewise, Hg-S bond has more ionic property than Hg-Te bond. 7

As is known, the surface termination of the materials with more ionic property can cause a weaker disturbance of surface electronic structures, thus there is a lower density of surface states, and in this respect, the sulfide-passivation method can also alleviate Fermi level pinning of MIT [42].

4. Conclusions In this study, the core-level spectra show the presence of Te4+ peak after sulfide passivation, which indicates that Te atoms are oxidized to form Te-O bonds on the passivated surface of MIT samples. However, the dissolution of tellurides produced by sulfide treatment removes Te-rich layer that is considered as a major cause for the increase of surface leakage current. Besides, more than half of In atoms at surface are sulfided to form In-S bonds, which can be deduced by the shift of In 3d core level and the In-S decomposed peak at 445.0 eV. The binding energy of Hg 4f shifts toward higher energy direction for about 0.1 eV, which indicates the formation of Hg-S bonds. These results demonstrate that S atoms can passivate dangling bonds and give more ionic property to the MIT surface by means of forming In-S and Hg-S bonds, and consequently surface states are considerably reduced. The increase of 0.14 eV in work function after sulfide passivation results from the decrease of surface states, so that Fermi level pinning can be further alleviated. The sulfide passivation on MIT surface can effectively improve electrical properties of Au/MIT Schottky contacts by enhancing the effect of rectifying behavior and reducing the leakage current due to a lower density of surface states.

8

References [1] L. A. Kosyachenko, I. S. Kabanova, V. M. Sklyarchuk, O. F. Sklyarchuk, I. M. Rarenko, Phys. Status. Solidi. A. 206 (2009) 351. [2] M. N. Solovan, A. I. Mostovyi, V. V. Brus, E. V. Maistruk, P. D. Maryanchuk, Semiconductors. 50 (2016) 1020. [3] J. Sun, L. Fu, Y. Wang, J. Ren, Y. Li, W. Zhang, J. Zhu, J. Appl. Phys. 114 (2013) 083719. [4] O. G Grushka, Semiconductors. 50 (2016) 719. [5] Y. Li, L. Fu, J. Sun, X. Wang, J. Appl. Phys. 117 (2015) 085704. [6] G. Li, W. Wang, W. Yang, Y. Lin, H. Wang, Z. Lin, S. Zhou, Rep. Prog. Phys.79 (2016) 056501. [7] L. Zeng, C. Xie, L. Tao, H. Long, C.Tang, Y. H. Tsang, J. Jie, Opt. Express. 23 (2015) 4839. [8] R. J. A. Esteves, M. Q. Ho, I. U. Arachchige, Chem. Mater. 27 (2015) 1559. [9] G. Agostinelli, A. Delabie, P. Vitanov, Z. Alexieva, H. F. W. Dekkers, S. D. Wolf, G. Beaucarne, Sol. Energ. Mat. Sol. C. 90 (2006) 3438. [10] J. R. Weber, A. Janotti, P. Rinke, C. G. Van De Walle, Appl. Phys. Lett. 91 (2007) 142101. [11] M. Hjort, J. Wallentin, R. Timm, A. A. Zakharov, U. Håkanson, J. N. Andersen, E. Lundgren, L. Samuelson, M. T. Borgström, A. Mikkelsen, Acs. Nano. 6 (2012) 9679. [12] S. Koveshnikov, W. Tsai, I. Ok, J. C. Lee, V. Torkanov, M. Yakimov, S. Oktyabrsky, Appl. Phys. Lett. 88 (2006) 022106. [13] B. Marie, M. A. El, L. Nick, B. Guy, A. E. Zaghi, M. Marc, P. Jef, Adv. Energy. Mater. 5 (2015) 1401689. [14] M. M. Frank, S. J. Koester, M. Copel, J. A. Ott, V. K. Paruchuri, H. Shang, R. Loesing, Appl. Phys. Lett. 89 (2006) 112905. [15] M. Kuzmin, P. Laukkanen, J. Mäkelä, M. Tuominen, M. Yasir, J. Dahl, P. J. P. Marko, K. Kokko, Phys. Rev. B. 94 (2016) 035421. [16] R. T. Tung, Phys. Rev. Lett. 84 (2000) 6078. [17] V. F. Radantsev, Semicond. Sci. Tech. 8 (1993) 394. [18] G. Bahir, V. A. Bahir, V. G Bahir, D. R Bahir, A. Bahir, Appl. Phys. Lett. 65 (1994) 2725. [19] S. Luo, C. Eisler, T. H. Wong, H. Xiao, C. E. Lin, T. T. Wu, H. A. Atwater, Acta. Materialia. 106 (2016) 171.

9

[20] K. T. Chen, D. T. Shi, H. Chen, B. Granderson, M. A. George, W.E. Collins, A. Burger, R. B. James, J. Vac. Sci. Technol. A. 15 (1997) 850. [21] W. B. Sang, K. S. Wang, J. H. Min, J. Y. Teng, Q. Zhang, Y. B. Qian, Semicond. Sci. Tech. 20 (2005) 343. [22] M. J. Mescher, T. E. Schlesinger, J. E. Toney, B. A. Brunett, R. B. James, J. Electron. Mater. 28 (1999) 700. [23] W. D. Hu, X. S. Chen, Z. H. Ye, W. Lu, Appl. Phys. Lett. 99 (2011) 091101. [24] J. D. Lin, Y. K. Su, S. J. Chang, M. Yokoyama, F. Y. Juang, J. Vac. Sci. Technol. A. 12 (1994) 7. [25] W. Lee, H. Kim, D. R. Jung, J. Kim, C. Nahm, J. Lee, B. Park, Nanoscale. Res. Lett. 7 (2012) 1. [26] A. J. Howard, C. Ashby, J. A. Lott, R. P. Schneider, R. F. Corless, J. Vac. Sci. Technol. A. 12 (1994) 1063. [27] X. Y. Hou, W. Z. Cai, Z. Q. He, P. H. Hao, Z. S. Li, X. M. Ding, X. Wang, Appl. Phys. Lett. 60 (1992) 2252. [28] G. Li , W. Wang, W. Yang, H. Wang, Surf. Sci. Rep. 70 (2015) 380. [29] Y. Tao, A. Yelon, E. Sacher, Z. H. Lu, M. J. Graham, Appl. Phys. Lett. 60 (1992) 2669. [30] W. K. Liu, W. T. Yuen, R. A. Stradling, J. Vac. Sci. Technol. B. 13 (1995) 1539. [31] A. K. Tkalich, V. N. Demin, V. P. Zlomanov, J. Solid. State. Chem. 116 (1995) 33. [32] T. Yang, M. Wang, C. Duan, X. Hu, L. Huang, J. Peng, F. Huang, X. Gong, Energ. Environ. Sci. 5 (2012) 8208. [33] D. Cahen, A. Kahn, Adv. Mater. 15 (2003) 271. [34] C. A. Mead, W. G. Spitzer, Phys. Rev. 134 (1964) A713. [35] W. E. Spicer, I. Lindau, P. R. Skeath, C. Y. Su, P. W. Chye, Phys. Rev. Lett. 44 (1980) 420. [36] R. R. Lieten, S. Degroote, M. Kuijk, G. Borghs, Appl. Phys. Lett. 92 (2008) 2106. [37] A.V. Thathachary, K.N. Bhat, N. Bhat, M.S. Hegde, Appl. Phys. Lett. 96 (2010) 2108. [38] B. J. Coppa, R. F. Davis, R. J. Nemanich, Appl. Phys. Lett. 82 (2003) 400. [39] S. Taniwaki, H. Yoshida, K. Arafune, A. Ogura, S. Satoh, Y. Hotta, J. Vac. Sci. Technol. A. 34 (2016) 061506. [40] Y. H. Lee, J. M. Wu, Y. L. Chueh, L. J. Chou, Appl. Phys. Lett. 87 (2005) 172901. [41] K. Nomura, A. Takagi, T. Kamiya, H. Ohta, M. Hirano, H. Hosono, Jpn. J. Appl. Phys. 45 (2006) 4303. [42] J. Wu, Y. Wu, C. Hou, M. Wu, C. Lin, L. Chen, Appl. Phys. Lett. 99 (2011) 253504.

10

Fig.1. The XPS Spectra of the surface of MIT sampels with and without sulfide passivation: (a) full spectrum; (b) Te 3d core-level; (c) In 3d core-level; (d) Hg 4f core-level

Fig.2. The depth analysis of Te 3d core-level spectra on the surface of sulfide-passivated MIT samples.

11

Fig.3. The In 3d core-level spectra of the sulfide-passivated MIT samples: (a) without Ar+ etching; (b) etching for 60 s; (c) etching for 90 s; (d) etching for 150 s; (e) etching for 330 s; (f) etching for 630 s.

Fig.4. The depth analysis of Hg 4f core-level spectra (a) by Ar+ ething for 60 s,90 s,150 s,330 s,660 s, and S 2p peak (b) on the surface of passivated MIT.

12

Fig.5. The UPS spectrum of the MIT (a), the details of inelastic secondary electron cutoff (b)， and Fermi Edge (c) with and without sulfide passivation in kinetic energy scale.

Fig.6. I–V measurement of Au/MIT Schottky contacts with and without sulfide-passivation.

13

Table1. Chemical composition of the MIT samples etched by Ar ion beam Time of Ar+

XPS surface element content (atm%)

etching

S

In

Te

Hg

0

42.2

26.9

0.8

60 s

37.2

43.4

90 s

27.9

150 s

Ratio of

Ratio of

O

Te/In

Te/Hg

12.0

6.5

0.03

0.07

7.3

4.8

7.3

0.17

1.52

38.7

11.9

4.4

17.3

0.31

2.70

13.8

34.5

21.3

4.6

25.8

0.62

4.63

330 s

2.9

33.0

50.3

8.7

5.2

1.52

5.78

630 s

0

34.2

55.7

10.1

0

1.63

5.51

14