much larger amounts of SiO2. Even at 500 °C, the surface
temperature remains much lower than temperatures com-
monly used to achieve SiO2 /Si interface formation. In addi-
tion, the SiO2 oxide layer is carbon free as previously sug-
gested for other SiC surfaces.2–5 Any direct C oxidation
would result in the formation of CO or CO2 which will de-
sorb from the surface into the vacuum as evident from the
lack of any related spectral feature.2–5 The changes observed
at the C 1s and Si 2p core levels would be more consistent
with the presence of a carbon interlayer bonded to SiO2 and
located at the SiO2 /SiC interface. Most significantly, the ox-
ide formed at 500 °C (⌬ESi 2pϭ4.5 eV) corresponds ex-
actly to the binding energy achieved for much thicker oxides
͑with a 144° Si–O–Si bond angle͒,11 despite the fact that
here, the oxide thickness is below 10 Å.
In conclusion, we have investigated the direct
SiO2 /-SiC͑100͒3ϫ2 interface formation by low molecular
O2 exposures using photoemission spectroscopy with syn-
chrotron radiation. Our results indicate that SiO2(⌬ESi 2p
ϭ3.9 eV) is already the dominant oxidation product at RT.
It shows that the nature of surface reconstruction plays a
major role in successful oxidation. Raising the surface tem-
perature by only few hundred degrees during O2 exposure
leads to ‘‘bulk-like’’ carbon free SiO2 oxide formation
(⌬ESi 2pϭ4.5 eV) already at small thicknesses. These find-
ings bring new and important insights in SiC surface passi-
vation as well as in low temperature oxide formation.
This work was supported by U. S. National Science
Foundation ͑NSF͒ under contract No. DMR 92-23710 and by
Northern Illinois University Graduate School Fund. It is
based upon research conducted at the Synchrotron Radiation
Center ͑SRC͒, University of Wisconsin-Madison which is
supported by NSF under contract No. DMR 92-12658. The
authors are grateful to C. Jaussaud and L. di Ciccio at LETI
FIG. 3. C 1s core level for clean and exposed to 10ϩ4 L of O2
-SiC͑100͒3ϫ2 surfaces at room temperature. The photon energy was
340 eV.
surfaces. The C 1s core level has been recorded in surface
sensitive mode at a photon energy of 340 eV. As can be seen
in Fig. 3, the C 1s core level is shifted by 0.6 eV to higher
binding energy upon exposure with a slight broadening only,
suggesting that the C atoms are somewhat involved in the
oxidation process. This cannot be related to graphite forma-
tion which would result in core level shifts larger than 2
eV.5,8 Also, it does not seem to indicate presence of direct
C–O bonds on the surface, in agreement with the picture
obtained from the O 2s core level showing only a single
component as for SiO2 /Si interface ͑Fig. 2͒. In fact, in this
case, much larger core level shifts would be expected.12
However, it is consistent with the formation of a complex
oxide as SiOC as suggested above from the Si 2p results.
Our above results support the picture of direct
SiO2 /-SiC͑100͒3ϫ2 interface formation already at RT
which, to our best knowledge, is the first observation of this
kind. Even on Si surfaces, direct dominant SiO2 formation
could be achieved at much higher temperatures only ͑above
900 °C͒. Interestingly, we do not observe a similar behavior
with other -SiC͑100͒ surface reconstructions like ͑2ϫ1͒
and ͑1ϫ1͒ which are much more difficult to oxidize than Si.9
This stresses the crucial importance of surface structure in
the oxidation process. In fact, the -SiC͑100͒3ϫ2 surface is
a much more open surface ͑when compared, e.g., to
Si͑100͒2ϫ1 or to Si͑111͒7ϫ7͒ which would allow O atoms
to interact more easily with the Si atoms. Furthermore, recent
theoretical calculations for Si-rich -SiC͑100͒ surfaces pre-
´
͑CEA-Technologies Avancees͒ for providing SiC͑100͒
samples. We thank the SRC staff for helpful assistance.
1 R. F. Davis, J. Vac. Sci. Technol. A 11, 829 ͑1993͒.
2 R. W. Kee, K. M. Geib, C. W. Wilmsen, and D. K. Ferry, J. Vac. Sci.
Technol. 15, 1520 ͑1978͒.
3 V. M. Bermudez, J. Appl. Phys. 66, 6084 ͑1989͒.
4 J. M. Powers and G. A. Somorjai, Surf. Sci. 244, 39 ͑1991͒.
5 M. Riehl-Chudoba, P. Soukiassian, C. Jaussaud, and S. Dupont, Phys.
Rev. B 51, 14300 ͑1995͒.
6 M. L. Shek, K. E. Miyano, Q. Y. Dong, T. A. Callcott, and D. L. Ederer,
J. Vac. Sci. Technol. A 12, 1079 ͑1994͒.
7 V. M. Bermudez and J. P. Long, Appl. Phys. Lett. 66, 475 ͑1995͒.
8 F. Semond, P. Soukiassian, P. S. Mangat, and L. di Cioccio, J. Vac. Sci.
Technol. B 13, 1591 ͑1995͒.
13
dicted very large Si–Si dimer bond lengths ͑Ϸ2.55 Å͒ in-
9 F. Semond, L. Douillard, D. Dunham, F. Amy, S. Rivillon, and P. Souki-
assian, recent results.
dicating weaken Si–Si bonds, especially when compared to
the Si͑100͒2ϫ1 surface having a Si–Si dimer bond length at
2.20 Å.14 This likely explains the much higher surface reac-
tivity to oxygen of the ͑3ϫ2͒ surface reconstruction when
compared to silicon surfaces.
10 F. J. Himpsel, F. R. McFeely, A. Taleb-Ibrahimi, J. A. Yarmoff, and G.
Hollinger, Phys. Rev. B 38, 6084 ͑1988͒.
11 F. J. Grunthaner, P. J. Grunthaner, R. P. Vasquez, B. F. Lewis, J. Maser-
jian, and A. Madhukar, J. Vac. Sci. Technol. B 16, 1443 ͑1979͒.
12 C. Kozlowski and P. M. A. Sherwood, Carbon 24, 357 ͑1986͒.
The surface reactivity is significantly enhanced when the
-SiC͑100͒3ϫ2 surface temperature is raised by few hun-
13
¨
¨
P. Kackell, J. Furthmuller, and F. Bechstedt, Appl. Surf. Sci., in press.
14 P. S. Mangat, P. Soukiassian, K. M. Schirm, Z. Hurych, L. Spiess, S. P.
Tang, A. J. Freeman, and B. Delley, Phys. Rev. B 47, 16311 ͑1993͒.
dred degrees ͑up to 500 °C͒ during O2 exposure leading to
2146 Appl. Phys. Lett., Vol. 68, No. 15, 8 April 1996 Semond et al.
130.18.123.11 On: Sun, 21 Dec 2014 10:57:23