- Email: [email protected]
6H-SiC(0 0 0 1) interface formation
Applied Surface Science 252 (2006) 5312–5315 www.elsevier.com/locate/apsusc
Study of the early stages of Cr/6H-SiC(0 0 0 1) interface formation I. Dontas *, S. Karakalos, S. Ladas, S. Kennou Department of Chemical Engineering, University of Patras, and FORTH/ICEHT, GR 26504, Patras, Greece Available online 24 January 2006
Abstract The early stages of the Cr/6H-SiC(0 0 0 1) interface formation at room temperature were investigated using XPS, LEED and work function (WF) measurements. Upon stepwise Cr evaporation in UHV up to a thickness of 5–10 monolayers (ML) at RT, the binding energy of the XPS Cr 2p3/2 core level peak shifted from 576.1 eV, at submonolayer coverage, to 574.7 eV (corresponding to metallic Cr) for the final Cr deposit, while the binding energies of the substrate XPS core level peaks remained stable. The WF exhibited a steep decrease of about 0.5 eV from the initial SiC substrate value, upon submonolayer coverage, but then increased gradually to saturation at a value of about 4.8 eV (polycrystalline Cr film with some chemisorbed oxygen). The growth of the ultrathin film was via 3D-cluster formation. The height of the Schottky barrier for the Cr/6HSiC(0 0 0 1) contact was found by XPS to be 0.5 0.1 eV. The results, generally, indicate the absence of any extended interfacial silicide-like interaction at RT. # 2006 Published by Elsevier B.V. PACS: 73.30.+y; 79.60.Dp; 82.80.Pv Keywords: Silicon carbide; Chromium; Ohmic metal contacts
1. Introduction Metal/SiC interfaces are important for high temperature and high power microelectronic devices [1–4] because of the high breakdown field and the wide band gap of SiC. Metal/SiC contact formation at room temperature has been studied for a large number of metals in recent years [5–18]. Cr/SiC is an interesting system for applications in SiC-based integratedcircuit technology; given that the Cr/Si contact was widely investigated in the past for Si semiconductor industry applications [19–26] and that SiC is a very promising material for the same applications area. Although there have been a number of reports on the interfacial reactions between thick Cr films and SiC at high temperatures [27–33], there is none investigating the initial steps of the Cr/SiC interface formation at RT. In this work, the very early stages of the Cr/6H-SiC(0 0 0 1) interface formation at room temperature were studied by X-ray photoelectron and Auger electron spectroscopy (XPS/XAES), low energy electron diffraction (LEED) and work function
* Corresponding author. Tel.: +30 2610 997856; fax: +30 2610 993255. E-mail address: [email protected] (I. Dontas). 0169-4332/$ – see front matter # 2006 Published by Elsevier B.V. doi:10.1016/j.apsusc.2005.12.066
(WF) measurements. The height of the interfacial Schottky barrier was obtained from the XPS data. 2. Experimental All measurements were performed in a conventional ultra high vacuum chamber (base pressure 9 10 10 mbar) equipped with a Leybold LHS-12 hemispherical electron energy analyser and a twin-anode X-ray gun for XPS. Photoelectrons were excited using the Al Ka line at 1486.6 eV. A Kelvin probe with a gold-plated vibrating reference electrode (piezoelectrically driven) was used to monitor WF changes. Single crystal 6H-SiC wafers, Si-face, from Cree Research Inc. [34] were used as substrates. The wafers were n-type, nitrogen-doped (3 1017 to 3 1018 cm 3 net doping density) and oriented 3–48 off-axis with respect to the ideal (0 0 0 1) plane towards the [11–20] direction. Before the samples were mounted onto the ultra high vacuum system, they were subjected to standard chemical cleaning to remove the native oxide. In UHV, the specimens were heated up to 1000 K. The substrates exhibited a good 1 1 hexagonal LEED pattern. A small XPS O 1s peak was observed in the XPS spectrum of the as prepared surface of the samples indicating oxygen coverage of about 1/2 ML.
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
5313
Chromium was evaporated in UHV from a home-built evaporation source. After a stepwise deposition of chromium at room temperature, the Schottky barrier height was determined from the XPS data. After each deposition step, XPS/XAES and WF measurements were taken. The chromium coverage, in equivalent monolayers, was estimated by calibrating the source doses from the exponential decay of the Si 2p core level peak intensity, assuming a 2D growth at the early deposition stages, thus underestimating the coverage by about 20%. One monolayer (ML) corresponds to an average film thickness of 0.26 nm, or a Cr surface atom density of 2 1015 atoms cm 2. 3. Results and discussion Upon the early stages of Cr deposition at RT, the 1 1 LEED pattern of the initial surface weakened continuously with increasing background intensity and no new spots were observed. Thus at RT no Cr induced superstructures are formed. Fig. 1 shows the variation of the Cr 2p3/2/Si 2p peak intensity ratio with Cr coverage for two different samples from the same Si-face 6H-SiC wafer. The solid line corresponds to a layer-bylayer or Frank–van-der Moerve (FM) model and was calculated using attenuation lengths of 1.6 and 1.3 nm for the Si 2p and Cr 2p3/2 peak energies respectively in Cr [35]. Chromium growth seems to follow a 3D-model after the first monolayer, as the experimental points begin to fall clearly beneath the calculated solid line.
Fig. 2. Change of the WF upon chromium deposition.
Fig. 1. XPS Intensity ratio of the Cr 2p3/2 and Si 2p peaks as a function of Cr equivalent thickness, compared with the theoretically calculated layer-by-layer (FM) growth mode curve.
The change of the WF upon Cr deposition is shown in Fig. 2. In principle, it can be seen in Fig. 2 that the initial value of the WF is slightly higher than that corresponding to the clean 6HSiC(0 0 0 1) surface (4.5 0.1 eV) [10,13,18]. This is due to the residual adsorbed oxygen (1/2 ML) onto the surface during the sample preparation. For submonolayer Cr coverage, the WF decreases from its initial value (4.55 eV) and reaches a minimum of 4.0 eV at about 0.4 ML Cr. Then, the WF increases gradually to saturation at 4.9 eV. The initial steep decrease of the WF is partly attributed to Cr-substrate dipoles caused by a small charge transfer upon adsorption of Cr, probably on surface defect sites like steps, as in the case of the Cu/6H-SiC system [18]. In the latter case, the minimum of the WF was obtained for a Cu coverage of 0.15 ML. Chromium is much more reactive than Cu, so the initial interaction with the substrate extends beyond the steps and probably involves also the adsorbed oxygen. The fact that the WF reaches saturation at a Cr coverage of about 8 ML is consistent with the aforementioned 3D character of chromium film growth. Furthermore, the final WF values are higher by 0.3 eV than the value from the literature [22,36,37] for polycrystalline Cr films. This is probably due to further oxygen adsorption onto the Cr film surface during the deposition procedure. Fig. 3 focuses on the binding energy change of the Cr 2p core level doublet at various stages of Cr deposition, as determined from the respective XP spectra. For a coverage of 0.1 ML, the Cr 2p doublet is broad and the BE for Cr 2p3/2 at 576.2 eV could be related both to the small particle size of the deposit or to some degree of oxidation. The BE for Cr 2p3/2 decreased
5314
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
Fig. 3. The Cr 2p XPS peak at various stages of chromium deposition.
gradually to a saturation value of approximately 574.7 0.1 eV, whereas the splitting of the 2p doublet obtains a value of 9.3 0.1 eV after the completion of the first monolayer. These values are indicative of the presence of metallic Cr. The Cr 2p3/2 BE value is a little higher than the expected one for a metallic chromium film (574.4 eV) [38] and this could be due to the presence of adsorbed oxygen or OH from the residual atmosphere of the chamber. As a matter of fact, the O 1s core level peak intensity was increasing gradually during Cr deposition indicating newly adsorbed oxygen onto the growing Cr film of the order of a ML, as evidenced by the shift of the O 1s BE peak from 532.4 eV (SiOx, OH) to 531 eV (Cr–Ox, OH). Upon Cr deposition at RT, the C 1s and Si 2p core level peaks of the SiC substrate are slowly attenuated but their shape does not change significantly, indicating negligible chemical interaction with deposited Cr. The Si 2p peaks for the as prepared substrate and up to 9 ML of Cr are shown in Fig. 4. For submonolayer Cr coverage, a small shift in binding energy of about 0.1–0.2 eV is observed for both substrate peaks, attributed to an initial modification in surface band bending, related to the small charge transfer during the submonolayer stages of Cr deposition that was mentioned before. Upon further Cr deposition, the BE of the substrate peaks remains stable up to 1–2 ML and then returns progressively to its initial value, as the metal/SiC interface is formed. The Schottky barrier height, FB, for the Cr/SiC interface was calculated from the above XPS data using the method described
Fig. 4. The Si 2p XPS peak at various stages of chromium deposition.
by Waldrop and co-workers [5]. In brief, FB = Eg–Efi, where Eg = 3.03 eV is the 6H-SiC band gap [5] and Efi is the interface Fermi level measured by XPS with respect to the top of the silicon carbide valence band; that is, Efi = EC 1s–(EC 1s–EV)o, where the quantity in parentheses is the C 1s to valence-band maximum binding energy difference for the clean SiC. The latter was measured at 281.4 0.1 eV by using XP valence band spectra from the as prepared surface (C 1s BE = 283.7 0.1 eV). From the measured C 1s binding energy after final Cr deposition (283.9 0.1 eV) a Schottky barrier height (SBH) of 0.5 0.1 eV is obtained, indicating an ohmic contact. We could not find any literature value for the Cr/6H-SiC Schottky barrier height. We found the same SBH only in a similar work for Cr on n-type Si(1 1 1) (a substrate exhibiting, in some cases, similar behaviour with Si-face 6H-SiC) [22] whereas some reference tables in a review paper by Porter and Davis [7] mention that generally Cr/SiC is an ohmic type contact. So, our result is consistent with both of these references. 4. Summary The very early stages of the Cr/Si-face 6H-SiC interface formation at RT were studied in detail by the combination of XPS and WF measurements. The Cr/SiC interface seems to grow following a 3D-model and the resulting interfacial Schottky barrier height was found by XPS to be 0.5 0.1 eV (ohmic contact). The results, generally, indicate the absence of
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
extended Cr–silicide-like interaction at RT, although at the initial stages of the interfacial layer growth there was a small Cr-substrate dipole formation (similar to that of Cu in [18]). Further investigation is necessary, including temperature dependent studies of the ultrathin Cr/SiC interface behavior. Acknowledgements We thank European Social Fund (ESF), Operational Program for Educational and Vocational Training II (EPEAEK II) and particularly the Program PYTHAGORAS, for funding the above work. References [1] J.W. Palmour, J.A. Edmond, H.S. Kong, C.H. Carter, Physica B 185 (1993) 461. [2] H. Morkoc, S. Strite, G.B. Gao, M.E. Lin, M. Burns, J. Appl. Phys. 76 (3) (1994) 1363. [3] C.H. Carter Jr., J.A. Edmond, J.W. Palmour, Mater. Sci. Eng. B 61–62 (1999) 1. [4] P.R. Chalker, Thin Solid Films 343–344 (1999) 616. [5] J.R. Waldrop, R.W. Grant, Y.C. Wang, R.F. Davis, J. Appl. Phys. 72 (10) (1992) 4757. [6] N. Lundberg, M. Ostling, Appl. Phys. Lett. 63 (22) (1993) 3069. [7] L.M. Porter, R.F. Davis, Mater. Sci. Eng. B 34 (1995) 83. [8] L. Li, I.S.T. Tsong, Surf. Sci. 364 (1996) 54. [9] A. Itoh, H. Matsunami, Phys. Stat. Sol. (a) 162 (1997) 389. [10] S. Kennou, A. Siokou, I. Dontas, S. Ladas, Diamond Rel. Mater. 6 (1997) 1424. [11] M.J. Bozack, Phys. Stat. Sol. (b) 202 (1997) 549. [12] I. Shalish, Y. Shapira, IEEE Electron Device Lett. 21 (12) (2000) 581. [13] I. Dontas, S. Kennou, Diamond Rel. Mater. 10 (2001) 13.
5315
[14] T. Suezaki, K. Kawahito, T. Hatayama, T. Fuyuki, Jpn. J. Appl. Phys. Lett. 240 (1A/B) (2001) L43. [15] F. Roccaforte, F. La Via, V. Raineri, L. Calcagno, P. Musumeci, Appl. Surf. Sci. 184 (2001) 295. [16] B. Schroeter, K. Komlev, W. Richter, Mater. Sci. Eng. B 88 (2002) 259. [17] F. La Via, F. Roccaforte, A. Makhtari, L. Calcagno, Microeltron. Eng. 60 (2002) 269. [18] I. Dontas, S. Ladas, S. Kennou, Diamond Rel. Mater. 12 (2003) 1209. [19] G. Ottaviani, J. Vac. Sci. Technol. 16 (5) (1979) 1112. [20] A. Franciosi, D.J. Peterman, J.H. Weaver, J. Vac. Sci. Technol. 19 (3) (1981) 657. [21] A. Franciosi, D.J. Peterman, F.A. Schmidt, C. Calandra, Phys. Rev. B 28 (12) (1983) 7000. [22] P. Wetzel, J.C. Peruchetti, D. Bolmont, G. Gewinner, J. Vac. Sci. Technol. A 5 (6) (1987) 3359. [23] L.F. Mattheiss, Phys. Rev. B 43 (15) (1991) 12549. [24] K.L. Hakansson, L.I. Johansson, A.N. Christensen, Phys. Rev. B 46 (3) (1992) 1535. [25] N. Benouattas, B. Tamaarat, R. Halimi, A. Mosser, Solid-State Electron. 43 (1999) 439. [26] R. Reiche, et al. Appl. Surf. Sci. 179 (2001) 316. [27] J.A. Edmond, J. Ryu, J.T. Glass, R.F. Davis, J. Electrochem. Soc. 135 (1988) 359. [28] K. Bhanumurthy, R. Schmid-Fetzer, Sens. Actuators A 51 (1995) 25. [29] K. Bhanamurthy, R. Schmid-Fetzer, Z. Metallkd. 87 (1996) 1. [30] F. Goesmann, R. Schmid-Fetzer, Mater. Sci. Eng. B 46 (1997) 357. [31] K. Bhanumurthy, R. Schmid-Fetzer, Composites: Part A 32 (2001) 569. [32] W.M. Tang, et al. Mater. Chem. Phys. 77 (2002) 236. [33] M. Hirai, C. Kamezawa, M. Iwami, Photon Fact. Activity Rep. 2002, #20 Part B, 2003, p. 87. [34] Cree Research Inc., 2810 Meridian Parkway, Suite 176, Durham, NC 27713, USA. [35] S. Tanuma, C.J. Powell, D.R. Penn, Surf. Interf. Anal. 21 (1993) 165. [36] D.E. Eastman, Phys. Rev. B 2 (1970) 1. [37] H.B. Michaelson, J. Appl. Phys. 48 (11) (1977) 4729. [38] C.J. Powell, Appl. Surf. Sci. 89 (1995) 141.
Study of the early stages of Cr/6H-SiC(0 0 0 1) interface formation I. Dontas *, S. Karakalos, S. Ladas, S. Kennou Department of Chemical Engineering, University of Patras, and FORTH/ICEHT, GR 26504, Patras, Greece Available online 24 January 2006
Abstract The early stages of the Cr/6H-SiC(0 0 0 1) interface formation at room temperature were investigated using XPS, LEED and work function (WF) measurements. Upon stepwise Cr evaporation in UHV up to a thickness of 5–10 monolayers (ML) at RT, the binding energy of the XPS Cr 2p3/2 core level peak shifted from 576.1 eV, at submonolayer coverage, to 574.7 eV (corresponding to metallic Cr) for the final Cr deposit, while the binding energies of the substrate XPS core level peaks remained stable. The WF exhibited a steep decrease of about 0.5 eV from the initial SiC substrate value, upon submonolayer coverage, but then increased gradually to saturation at a value of about 4.8 eV (polycrystalline Cr film with some chemisorbed oxygen). The growth of the ultrathin film was via 3D-cluster formation. The height of the Schottky barrier for the Cr/6HSiC(0 0 0 1) contact was found by XPS to be 0.5 0.1 eV. The results, generally, indicate the absence of any extended interfacial silicide-like interaction at RT. # 2006 Published by Elsevier B.V. PACS: 73.30.+y; 79.60.Dp; 82.80.Pv Keywords: Silicon carbide; Chromium; Ohmic metal contacts
1. Introduction Metal/SiC interfaces are important for high temperature and high power microelectronic devices [1–4] because of the high breakdown field and the wide band gap of SiC. Metal/SiC contact formation at room temperature has been studied for a large number of metals in recent years [5–18]. Cr/SiC is an interesting system for applications in SiC-based integratedcircuit technology; given that the Cr/Si contact was widely investigated in the past for Si semiconductor industry applications [19–26] and that SiC is a very promising material for the same applications area. Although there have been a number of reports on the interfacial reactions between thick Cr films and SiC at high temperatures [27–33], there is none investigating the initial steps of the Cr/SiC interface formation at RT. In this work, the very early stages of the Cr/6H-SiC(0 0 0 1) interface formation at room temperature were studied by X-ray photoelectron and Auger electron spectroscopy (XPS/XAES), low energy electron diffraction (LEED) and work function
* Corresponding author. Tel.: +30 2610 997856; fax: +30 2610 993255. E-mail address: [email protected] (I. Dontas). 0169-4332/$ – see front matter # 2006 Published by Elsevier B.V. doi:10.1016/j.apsusc.2005.12.066
(WF) measurements. The height of the interfacial Schottky barrier was obtained from the XPS data. 2. Experimental All measurements were performed in a conventional ultra high vacuum chamber (base pressure 9 10 10 mbar) equipped with a Leybold LHS-12 hemispherical electron energy analyser and a twin-anode X-ray gun for XPS. Photoelectrons were excited using the Al Ka line at 1486.6 eV. A Kelvin probe with a gold-plated vibrating reference electrode (piezoelectrically driven) was used to monitor WF changes. Single crystal 6H-SiC wafers, Si-face, from Cree Research Inc. [34] were used as substrates. The wafers were n-type, nitrogen-doped (3 1017 to 3 1018 cm 3 net doping density) and oriented 3–48 off-axis with respect to the ideal (0 0 0 1) plane towards the [11–20] direction. Before the samples were mounted onto the ultra high vacuum system, they were subjected to standard chemical cleaning to remove the native oxide. In UHV, the specimens were heated up to 1000 K. The substrates exhibited a good 1 1 hexagonal LEED pattern. A small XPS O 1s peak was observed in the XPS spectrum of the as prepared surface of the samples indicating oxygen coverage of about 1/2 ML.
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
5313
Chromium was evaporated in UHV from a home-built evaporation source. After a stepwise deposition of chromium at room temperature, the Schottky barrier height was determined from the XPS data. After each deposition step, XPS/XAES and WF measurements were taken. The chromium coverage, in equivalent monolayers, was estimated by calibrating the source doses from the exponential decay of the Si 2p core level peak intensity, assuming a 2D growth at the early deposition stages, thus underestimating the coverage by about 20%. One monolayer (ML) corresponds to an average film thickness of 0.26 nm, or a Cr surface atom density of 2 1015 atoms cm 2. 3. Results and discussion Upon the early stages of Cr deposition at RT, the 1 1 LEED pattern of the initial surface weakened continuously with increasing background intensity and no new spots were observed. Thus at RT no Cr induced superstructures are formed. Fig. 1 shows the variation of the Cr 2p3/2/Si 2p peak intensity ratio with Cr coverage for two different samples from the same Si-face 6H-SiC wafer. The solid line corresponds to a layer-bylayer or Frank–van-der Moerve (FM) model and was calculated using attenuation lengths of 1.6 and 1.3 nm for the Si 2p and Cr 2p3/2 peak energies respectively in Cr [35]. Chromium growth seems to follow a 3D-model after the first monolayer, as the experimental points begin to fall clearly beneath the calculated solid line.
Fig. 2. Change of the WF upon chromium deposition.
Fig. 1. XPS Intensity ratio of the Cr 2p3/2 and Si 2p peaks as a function of Cr equivalent thickness, compared with the theoretically calculated layer-by-layer (FM) growth mode curve.
The change of the WF upon Cr deposition is shown in Fig. 2. In principle, it can be seen in Fig. 2 that the initial value of the WF is slightly higher than that corresponding to the clean 6HSiC(0 0 0 1) surface (4.5 0.1 eV) [10,13,18]. This is due to the residual adsorbed oxygen (1/2 ML) onto the surface during the sample preparation. For submonolayer Cr coverage, the WF decreases from its initial value (4.55 eV) and reaches a minimum of 4.0 eV at about 0.4 ML Cr. Then, the WF increases gradually to saturation at 4.9 eV. The initial steep decrease of the WF is partly attributed to Cr-substrate dipoles caused by a small charge transfer upon adsorption of Cr, probably on surface defect sites like steps, as in the case of the Cu/6H-SiC system [18]. In the latter case, the minimum of the WF was obtained for a Cu coverage of 0.15 ML. Chromium is much more reactive than Cu, so the initial interaction with the substrate extends beyond the steps and probably involves also the adsorbed oxygen. The fact that the WF reaches saturation at a Cr coverage of about 8 ML is consistent with the aforementioned 3D character of chromium film growth. Furthermore, the final WF values are higher by 0.3 eV than the value from the literature [22,36,37] for polycrystalline Cr films. This is probably due to further oxygen adsorption onto the Cr film surface during the deposition procedure. Fig. 3 focuses on the binding energy change of the Cr 2p core level doublet at various stages of Cr deposition, as determined from the respective XP spectra. For a coverage of 0.1 ML, the Cr 2p doublet is broad and the BE for Cr 2p3/2 at 576.2 eV could be related both to the small particle size of the deposit or to some degree of oxidation. The BE for Cr 2p3/2 decreased
5314
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
Fig. 3. The Cr 2p XPS peak at various stages of chromium deposition.
gradually to a saturation value of approximately 574.7 0.1 eV, whereas the splitting of the 2p doublet obtains a value of 9.3 0.1 eV after the completion of the first monolayer. These values are indicative of the presence of metallic Cr. The Cr 2p3/2 BE value is a little higher than the expected one for a metallic chromium film (574.4 eV) [38] and this could be due to the presence of adsorbed oxygen or OH from the residual atmosphere of the chamber. As a matter of fact, the O 1s core level peak intensity was increasing gradually during Cr deposition indicating newly adsorbed oxygen onto the growing Cr film of the order of a ML, as evidenced by the shift of the O 1s BE peak from 532.4 eV (SiOx, OH) to 531 eV (Cr–Ox, OH). Upon Cr deposition at RT, the C 1s and Si 2p core level peaks of the SiC substrate are slowly attenuated but their shape does not change significantly, indicating negligible chemical interaction with deposited Cr. The Si 2p peaks for the as prepared substrate and up to 9 ML of Cr are shown in Fig. 4. For submonolayer Cr coverage, a small shift in binding energy of about 0.1–0.2 eV is observed for both substrate peaks, attributed to an initial modification in surface band bending, related to the small charge transfer during the submonolayer stages of Cr deposition that was mentioned before. Upon further Cr deposition, the BE of the substrate peaks remains stable up to 1–2 ML and then returns progressively to its initial value, as the metal/SiC interface is formed. The Schottky barrier height, FB, for the Cr/SiC interface was calculated from the above XPS data using the method described
Fig. 4. The Si 2p XPS peak at various stages of chromium deposition.
by Waldrop and co-workers [5]. In brief, FB = Eg–Efi, where Eg = 3.03 eV is the 6H-SiC band gap [5] and Efi is the interface Fermi level measured by XPS with respect to the top of the silicon carbide valence band; that is, Efi = EC 1s–(EC 1s–EV)o, where the quantity in parentheses is the C 1s to valence-band maximum binding energy difference for the clean SiC. The latter was measured at 281.4 0.1 eV by using XP valence band spectra from the as prepared surface (C 1s BE = 283.7 0.1 eV). From the measured C 1s binding energy after final Cr deposition (283.9 0.1 eV) a Schottky barrier height (SBH) of 0.5 0.1 eV is obtained, indicating an ohmic contact. We could not find any literature value for the Cr/6H-SiC Schottky barrier height. We found the same SBH only in a similar work for Cr on n-type Si(1 1 1) (a substrate exhibiting, in some cases, similar behaviour with Si-face 6H-SiC) [22] whereas some reference tables in a review paper by Porter and Davis [7] mention that generally Cr/SiC is an ohmic type contact. So, our result is consistent with both of these references. 4. Summary The very early stages of the Cr/Si-face 6H-SiC interface formation at RT were studied in detail by the combination of XPS and WF measurements. The Cr/SiC interface seems to grow following a 3D-model and the resulting interfacial Schottky barrier height was found by XPS to be 0.5 0.1 eV (ohmic contact). The results, generally, indicate the absence of
I. Dontas et al. / Applied Surface Science 252 (2006) 5312–5315
extended Cr–silicide-like interaction at RT, although at the initial stages of the interfacial layer growth there was a small Cr-substrate dipole formation (similar to that of Cu in [18]). Further investigation is necessary, including temperature dependent studies of the ultrathin Cr/SiC interface behavior. Acknowledgements We thank European Social Fund (ESF), Operational Program for Educational and Vocational Training II (EPEAEK II) and particularly the Program PYTHAGORAS, for funding the above work. References [1] J.W. Palmour, J.A. Edmond, H.S. Kong, C.H. Carter, Physica B 185 (1993) 461. [2] H. Morkoc, S. Strite, G.B. Gao, M.E. Lin, M. Burns, J. Appl. Phys. 76 (3) (1994) 1363. [3] C.H. Carter Jr., J.A. Edmond, J.W. Palmour, Mater. Sci. Eng. B 61–62 (1999) 1. [4] P.R. Chalker, Thin Solid Films 343–344 (1999) 616. [5] J.R. Waldrop, R.W. Grant, Y.C. Wang, R.F. Davis, J. Appl. Phys. 72 (10) (1992) 4757. [6] N. Lundberg, M. Ostling, Appl. Phys. Lett. 63 (22) (1993) 3069. [7] L.M. Porter, R.F. Davis, Mater. Sci. Eng. B 34 (1995) 83. [8] L. Li, I.S.T. Tsong, Surf. Sci. 364 (1996) 54. [9] A. Itoh, H. Matsunami, Phys. Stat. Sol. (a) 162 (1997) 389. [10] S. Kennou, A. Siokou, I. Dontas, S. Ladas, Diamond Rel. Mater. 6 (1997) 1424. [11] M.J. Bozack, Phys. Stat. Sol. (b) 202 (1997) 549. [12] I. Shalish, Y. Shapira, IEEE Electron Device Lett. 21 (12) (2000) 581. [13] I. Dontas, S. Kennou, Diamond Rel. Mater. 10 (2001) 13.
5315
[14] T. Suezaki, K. Kawahito, T. Hatayama, T. Fuyuki, Jpn. J. Appl. Phys. Lett. 240 (1A/B) (2001) L43. [15] F. Roccaforte, F. La Via, V. Raineri, L. Calcagno, P. Musumeci, Appl. Surf. Sci. 184 (2001) 295. [16] B. Schroeter, K. Komlev, W. Richter, Mater. Sci. Eng. B 88 (2002) 259. [17] F. La Via, F. Roccaforte, A. Makhtari, L. Calcagno, Microeltron. Eng. 60 (2002) 269. [18] I. Dontas, S. Ladas, S. Kennou, Diamond Rel. Mater. 12 (2003) 1209. [19] G. Ottaviani, J. Vac. Sci. Technol. 16 (5) (1979) 1112. [20] A. Franciosi, D.J. Peterman, J.H. Weaver, J. Vac. Sci. Technol. 19 (3) (1981) 657. [21] A. Franciosi, D.J. Peterman, F.A. Schmidt, C. Calandra, Phys. Rev. B 28 (12) (1983) 7000. [22] P. Wetzel, J.C. Peruchetti, D. Bolmont, G. Gewinner, J. Vac. Sci. Technol. A 5 (6) (1987) 3359. [23] L.F. Mattheiss, Phys. Rev. B 43 (15) (1991) 12549. [24] K.L. Hakansson, L.I. Johansson, A.N. Christensen, Phys. Rev. B 46 (3) (1992) 1535. [25] N. Benouattas, B. Tamaarat, R. Halimi, A. Mosser, Solid-State Electron. 43 (1999) 439. [26] R. Reiche, et al. Appl. Surf. Sci. 179 (2001) 316. [27] J.A. Edmond, J. Ryu, J.T. Glass, R.F. Davis, J. Electrochem. Soc. 135 (1988) 359. [28] K. Bhanumurthy, R. Schmid-Fetzer, Sens. Actuators A 51 (1995) 25. [29] K. Bhanamurthy, R. Schmid-Fetzer, Z. Metallkd. 87 (1996) 1. [30] F. Goesmann, R. Schmid-Fetzer, Mater. Sci. Eng. B 46 (1997) 357. [31] K. Bhanumurthy, R. Schmid-Fetzer, Composites: Part A 32 (2001) 569. [32] W.M. Tang, et al. Mater. Chem. Phys. 77 (2002) 236. [33] M. Hirai, C. Kamezawa, M. Iwami, Photon Fact. Activity Rep. 2002, #20 Part B, 2003, p. 87. [34] Cree Research Inc., 2810 Meridian Parkway, Suite 176, Durham, NC 27713, USA. [35] S. Tanuma, C.J. Powell, D.R. Penn, Surf. Interf. Anal. 21 (1993) 165. [36] D.E. Eastman, Phys. Rev. B 2 (1970) 1. [37] H.B. Michaelson, J. Appl. Phys. 48 (11) (1977) 4729. [38] C.J. Powell, Appl. Surf. Sci. 89 (1995) 141.