.,.x.:.:.x.:.::::::::.:.::::::,:.:.>
..,:,,.,,,
.-.:.:.:~::~:~.:~Ji:~~~::::::~::::.::::::::::::::,:,:::::::::::
:.::.:,,
“‘-‘.....A.
. . . ...._.............................,.,.,,,~
..
. . . .. .
.((.,./..
“.‘... . . . .. .. .. .... . . . . . .. . . .. .. .
ii....:
.,v...
#::;:.:‘:“.
Surface Science 279 (1992) 119-126
North-Holland
..
.,.,.,
,,,,,.
. .. . . . . .. . ..../_.
_,
,_
..“:‘.:.:z:8
surface science
..
.......‘.-“v.‘.‘.‘...,.(,
._, ./n.. .,.....,....
,.(.
z:.:.::::
.,.,
.,.......
:.:..
>..:.:.:::::::g
:,:.::::
.::::,
‘..“’
‘.‘I:.:
‘.Y
..,.,........
..v.....:
.,.....,......
““A”
,_:...
“‘A”“‘.
‘.‘+5
.....,.,,.
..,,.,.,:,~.~,,,.,.,,,,,~~~,,
...~~....
,i,,,.,,,
;::::::~
‘-‘.:.:‘:‘::.::::::~~~~:~::~,~~!:~:~.~.::::
:.:,:,
;::::...
X-ray photoelectron spectroscopic characterization
of ultra-thin silicon oxide films on a Mo( 100) surface
J.-W. He, X. Xu, J.S. Corneille
Department
of Chemistry,
Texas A&M
and D.W. Goodman
University, College Station,
*
TX 77843-3255, USA
Received 16 June 1992; accepted for publication 13 August 1992
Ultra-thin films of silicon oxides supported on a MoflOO) surface have been studied using X-ray photoelectron spectroscopy
(XPS). The films were synthesized by evaporating Si onto the MO surface in oxygen ambient and were subsequently characterized
using XPS with respect to the chemical states of silicon and the composition of the film. It has been found that the silicon oxide,
prepared at room temperature with a silicon deposition rate of N 1.2 A/min and an oxygen pressure of 2 x lo-’ Torr, consisted of
predominantly silicon dioxide with a small fraction of suboxides. Annealing to _ 1300 K yielded a stoichiometric film of SiO,. The
subozides are believed to further react with oxygen forming SiO, at an elevated temperature.
1. Introduction
Over the last decades, there have been numerous publications that have dealt with studies of
silicon oxide thin films using surface science spectroscopies [l-18]. These previous studies have
focused exclusively on the oxidation of silicon
single crystal surfaces, addressing important issues such as the mechanism and kinetics of the
thin film growth, structural and electronic properties of silicon dioxide, and the formation of
silicon suboxides. These investigations have been
mainly stimulated by important applications of
silicon oxides in the fabrication of electronic devices, for example, metal/ oxide/ semiconductor
transistors and SiO,/Si solar energy cells.
Silicon dioxide is also extensively used as a
catalyst support in the chemical industry. However, the surface chemistry of bulk silica has been
studied far less with surface science spectroscopies compared to SiO,/Si interfaces. This is
mainly because of the highly insulating and brittle
properties of silica which often cause experimen* To whom all correspondence should be addressed.
0039-6028/92/$05.00
tal difficulties such as surface charging, sample
mounting, sample heating and cooling. In order
to circumvent these difficulties and to exploit
microscopic processes on a silica surface using
surface analytical techniques, experiments have
been recently conducted to synthesize and characterize ultra-thin films of silicon oxide on a
refractory metal surface [19,20]. Furthermore, thin
films of SiO, prepared in this manner allow
much greater flexibility with respect to thermal
processing compared to SiO, films on silicon in
that films of SiO, on molybdenum are stable up
to 1500 K. Similar studies have been performed
on other catalytically important oxides such as
magnesium oxide [21] and iron oxides [22].
In previous studies of silicon oxide thin films
on a MoQlO) surface using temperature programmed desorption (TPD), Auger electron spectroscopy (AES), electron energy loss spectroscopy
(ELS) and infrared reflection absorption spectroscopy (IRAQ [19,20], it was found that a SiO,
film can be synthesized by evaporating silicon in a
low background pressure of oxygen at a sample
temperature of N 300 K. The gas phase precursors for the SiO, film were reported to be SiO
0 1992 - Elsevier Science Publishers B.V. All rights reserved
120
J.-W. He et al. / Characterization of ultra-thin silicon oxide films on Mo(100)
which was produced by oxidative etching of the Si
source. The as-deposited films mainly consist of
SiO,, as indicated by the AES spectra. The SiO,
film annealed to - 1400 K exhibited electronic
and structural properties similar to those of vitreous silica. The thermal stability of the SiO, film
was found to depend on the film ;hickness. Annealing a thin SiO, film (< 50 A) to 1500 K
caused a reduction of the film by the MO substrate, forming the volatile products SiO and
MOO,.
Recently, we have further examined the silicon
oxide thin films grown on a Mo(100) surface with
X-ray photoelectron spectroscopy (XPS), addressing the effects that oxygen ambient pressure and
annealing temperature
have on the chemical
states of Si and the composition of the films. The
results are presented and discussed in this paper.
2. Experimental
The experiments were carried out in an ultrahigh vacuum chamber equipped with XPS, TPD
and low energy electron diffraction (LEED). A
detailed description of this chamber has been
given elsewhere (231.
The Mo(100) crystal was spot-welded onto a
0.2 mm Ta wire that allowed cooling of the crystal to 90 K and resistive heating to 1400 K. In
addition, the sample could be heated to 2300 K
using e-beam heating. A W-5%Re/W-26%Re
thermocouple was spot-welded onto the sample’s
back face for temperature
measurement.
The
Mo(100) surface was cleaned using the procedure
described in ref. [24].
The silicon source used for deposition was a
high purity Si block (- 1 x 1 X 3 mm3) wrapped
by a tantalum heating wire. The Si source was
degassed extensively prior to each dosage. The Si
oxide films, synthesized in this work, showed no
C, S or N impurities within the XPS and AES
detection limit (< 1 at%>. The thickness of the
films was zstimated using the electron mean free
path (23 A) in SiO, films for Mo(3d) photoelectrons, which have a kinetic energy of 1020 eV
with the Mg Ka X-ray source (1253.6 eV) [25,26].
The Si oxide deposition rate varies with both
oxygen pressure and the Si source temperature,
and owascontrolled at approximately either 0.6 or
1.2 A/min for this work.
In the present study of the SiO,/Mo(lOO) surfaces, the core level binding energies were measured with reference to the Mo(3d,,,) binding
energy at 277.7 eV [27]. The pass energy of the
spectrometer was 40 eV for this work, and the
substrate Mo(3d,,,) peak exhibited a full width
at half maximum (FWHM) of 1.4 eV which can
be taken as the instrumental resolution. The XPS
spectra in this paper were acquired using Mg Ka
X-rays (1253.6 eV>.
3. Results
Fig. 1 shows the Si(2p) XPS spectra for silicon
oxide thin films that were prepared at three different oxygen background pressures. The thicknesses of the films are estimated to be 36 A. At
an oxygen ambient pressure of 1 X lop8 Torr,
spectrum a exhibits a significant peak at 99.4 eV
that corresponds to silicon. This value is 0.2 eV
higher than the Si(2p) binding energy (99.2 eV)
from a pure silicon crystal [27]. The peak at
- 102.5 eV is broad, and is thus best decomposed
into two peaks using mixed Gaussian-Lorentzian
curve fitting. These two peaks at 103.2 and 101.7
eV correspond to oxidized silicon. As the oxygen
pressure increases (spectra b and c), the 103.2 eV
peak intensity increases along with a shift in its
peak position to higher binding energy, the intensity of the 101.7 eV peak first increases and then
decreases, and the intensity of the peak at 99.4
eV decreases considerably. In spectrum c, the
103.6 eV peak shows a binding energy similar to
that for silicon dioxide (103.4 eV>, and the peak
at 102.7 eV falls in the binding energy range of
suboxides (SiO,, with x < 2) [2-8,271. Therefore,
we assign these two peaks at 103.6 and 102.7 eV
to correspond to SiO, and suboxides, respectively. Previous studies also indicated that the
oxide films prepared at an oxygen pressure of
1,X 1O-5 Torr with a deposition rate of - 1
A/min consist predominantly of silicon dioxide,
as evidenced by AES and ELS spectra [19,20].
This suboxide may contain several Si oxidative
J.-W. He et al. / Characterization of ultra-thin silicon oxide films on Mo(100)
B
:
oxidative states of Si. The spectral deconvolution
yields two more peaks at 102.7 and 99.9 eV,
which are characteristic of suboxides and silicon.
The integrated areas indicate that this film has
82% SiO,, 13% SiO, and 5% Si. Annealing this
film at 1023 K reduces the suboxide peak area to
zero and results in a symmetric peak for SiO,
with a small peak at 99.6 eV for Si. This clearly
indicates that annealing has eliminated the suboxides from the film. The SiO, species may have
been either oxidized to SiO,, desorbed into vacuum, or decomposed into Si and 0. The decomposition is unlikely because spectrum b shows no
101.9
rh
I
103.1
A
’
-
.
g-*
/
/’
i
;?\:
,;
-
“F4
-I
‘,c
‘1
: I
BINDING
.
.
Y
\
x0.3
99.4
F
106
121
i’
102
98
ENERGY
(eV)
Fig. 1. The Si(2p) XPS spectra of silicon oxide films (36 A) on
a Mo(100) surface. The films were prepared by evaporating Si
in an 0s pressure of 1 X 10-8 Torr (a), 5 X 10W8Torr (b) and
2 X 10W5Torr (c). The substrate temperature during the film
preparation was - 350 K. The oxide deposition rate was
approximately 1.2 IQmin.
states, as have been unambiguously observed in
the studies of SiO,/Si
interfaces using synchrotron XPS [2,6]. The present work is, however,
unable to further distinguish between these states
due to the resolution limitation of the XPS spectrometer.
Figs. 2 and 3 show the effect that annealing
temperature has on the silicon oxide films prepared at N 350 K. In fig. 2, the film deposited at
room temperature with an oxygen pressure of
2 X lo-’ Torr (spectrum a> is composed predominantly of SiO,. The spectral tail at the low binding energy side indicates the existence of the low
106
BINDING
102
ENERGY
98
(eV)
Fig. 2. The Si(2p) XPS spectra of silicon oxide films (36 & on
a Md100) surface. The films were prepared by evaporating Si
if an 0, pressure of 2~ lo-’ Torr with a dosing rate of - 1.2
A/min. The substrate temperature during the film preparation was - 350 K. The film was then annealed to 393 K (a),
1023 K (b) and 1373 K (c).
J.-W. He et al. / Characterization of ultra-thin silicon oxide film on MoilOOi
122
sity has been significantly decreased when compared to that of spectrum a. Fig. 3 demonstrates
that the film synthesized at relatively low oxygen
pressure has a considerable amount of suboxides
and Si. This is consistent with previous studies,
which showed that the film prepared at an oxygen
pressure less than 1 x 10m6 Torr contained silicon or silicon suboxides [19,201. Annealing the
film to elevated temperatures was found to remove the suboxide spectral features [19]. We will
discuss this SiO, in more detail in the following
section.
The OfIs) spectra for a silicon oxide film are
shown in fig. 4. After the film preparation at 330
K, the O(ls) spectrum indicates a single peak
with a binding energy of 533.0 eV. This peak
exhibits a symmetric lineshape with a full width at
half maximum (FWHM) of 2 eV, identical to that
reported for a thick SiO, film grown on a silicon
surface [IS]. Annealing this film to 1373 K causes
a 0.3 eV shift in the peak maximum with no
noticeable change in the peak width or peak
shape.
105
BINDING
101
97
ENERGY
(ev)
Fig. 3. The Si(‘2pf XPS spectra of silicon oxide
a Mo(100)
i,” an 0,
A/min.
tion was
surface. The films were prepared
pressure of 5 X IO-’
The substrate
- 350
films (36 A) on
by evaporating
Si
Torr with a dosing rate of - 1.2
temperature
during the film prepara-
K. The film was then annealed to 373 K (a),
803 K (b) and 1198 K Cc).
significant increase in the Si peak area. Further
annealing to 1373 K eliminates the silicon peak
with a highly symmetric peak of SiO, remaining.
Fig. 3a is the XPS spectrum for a silicon oxide
film that was dosed at relatively low OZ pressure
(5 x lo-’ Torr). This film consists of 26% SiO,,
38% SiO, and 36% Si, as estimated from the
areas of the deconvoluted peaks. Annealing to
803 K causes a significant increase in the SiOz
peak intensity with a concurrent decrease in the
intensities of SiO, and of Si. Spectrum c, collected after annealing the film to 1198 K, exhibits
virtually no suboxide peak, and the Si peak inten-
540
526
533
BINDING
ENERGY
(eV)
Fig. 4. The O(ls) XPS spectra of a silicon oxide film (100 ,&I.
The film was prepared by evaporating Si at an oxygen pressure of 1 X lo-’ Torr with a dosing rate of - 0.6 .&/min at a
sample temperature of 300 K. The film was then annealed to
the indicated temperatures.
J.-W He et al. / Characterization of ultra-thin silicon oxide jilms on Mo(100)
z
!z
nealing temperature of N 1300 K, we assign the
ratio corresponding to the spectra at 1480 K to be
two, the atomic ratio of 0 to Si in SiO,. In the
calculation of the ratios, the contribution of oxygen from the MO substrate has been subtracted.
The O(1.s) binding energies are 533.0 eV for SiO,
and 530.7 eV for 0 chemisorbed on Mo(lOO), and
thus are easily discernable at a low coverage of
SiO,.
Finally, we present the Mo(3d) XPS spectra
for a SiO,/Mo(lOO) surface at several annealing
temperatures. The Mo(3dgJ peak has a FWHM
of 1.3 eV and a binding energy of 227.7 eV for all
the spectra. These values remain unchanged in
the 1023-1473 K annealing temperature range.
Deposition of SiO, and subsequent annealing
cause no measurable change in the MO substrate.
The intensity increase in the 1473 K spectrum is
due to the desorption of SiO, [20].
SiPp
E-l
103.9
0%
zal
E-
103.7-
ifi
CL
.;
103.5
G
::
533.4
-
-
533.2
-
533.0
-
:
01s
3
2 s
E501
z
s
m
i;
400
123
600
ANNEAL
600
1000
1200
TEMPERATURE
1400
!
1600
(K)
Fig. 5. The change of Si(2p) and O(k) binding energies and
the ratio of the Si(2p) integrated intensity to that of O(k) for
a SiO, film (100 A) on a Mo(100) surface. The SiO, film was
prepared by evaporating Si at an ?xygen pressure of 1 X lo-’
Torr with a dosing rate of - 0.6 A/min. The sample temperature during the SiO, deposition was - 300 K.
Fig. 5 shows the plots of the Si(2p) and O(1.s)
binding energies and the ratio of the integrated
intensity of 0 to that of Si as a function of
annealing temperature. The silicon oxide film was
synthesized in an oxygen pressyre of 1 X lo-’
Torr with a dosing rate of 0.6 A/min at room
temperature. Upon annealing to 800-1300 K, the
O(ls) binding energy shifts 0.3 eV and the Si(2p)
shifts 0.4 eV toward a higher binding energy. The
integrated intensity ratio of 0 to Si shows an
increase in the 1100-1350 K annealing temperature range. Since all the experimental evidence,
including the XPS spectra in fig. 2 and the TPD,
AES and ELS spectra in previous work [19,20],
have indicated that both suboxides and Si have
been totally eliminated from the film at an an-
4. Discussion
One of the primary goals of synthesizing silicon oxide on a MO surface is to prepare a stoichiometric silica film that can be used for model
catalyst studies. Figs. l-3 indicate that such a
silica film can be obtained by evaporation of Si
with an oxygen pressure ,of 2 X lo-’ Torr at a
deposition rate of N 1.2 A/min, followed by annealing the film to 1300 K. At a relatively low
oxygen pressure ( < 5 X lo-* Torr), the film has a
significant amount of suboxides and silicon in
coexistence with SiO,. Fig. 1 further indicates
that the film synthesized at room temperature
contains more than 10% of suboxides even with
an oxygen pressure of 2 X 1O-5 Torr. A further
increase in oxygen pressure produces little improvement in the film composition. Annealing to
1300 K, however, eliminates the suboxides and
silicon and results in a film consisting of only
SiO,. The silica film has been suggested to assume a vitreous SiO, structure in which the basic
[SiO,] blocks are networked via bridging oxygen
atoms [19,20]. This film showed no LEED pattern
and thus lacks long range order.
The suboxide species formed during oxidation
of silicon surfaces have been extensively studied
124
J.-W. He et al. / Characterization of ultra-thin &cm
with a variety of spectroscopies including XPS
[Z-9], ELS 13,10-121 and IRAS [13-151. It is
generally believed that the SiO, species are located at the SiO,/Si interface with a thickness of
a few angstroms. In studies of oxygen adsorption
on Si surfaces, suboxide species were also observed for a monolayer surface oxide [8,11]. An
iuvesti~ation of silicon oxidation by means of ion
implantation at room temperature has indicated
the formation of a thin layer inside a Si crystal
which contains a mixture of SiO,, SiO, and Si [4].
The SiO, species, investigated by both high resolution XPS spectra [7,83 and theoretical calculations [S], consist of all four of the Si oxidative
states, Si’+, Si2+, Si”+ and Si” with the Si(2p)
core level shifts of 0.6, l&2.8 and 4.5 eV, respectively. Therefore, we believe that the suboxide
film prepared at room temperature in this work
may contain all four oxidative states. The shift in
the maxima of the oxide peaks towards a higher
binding energy (fig. I), as the oxygen pressure
increases, reflects the increase in the relative
population of the higher oxidative states. At an
annealing temperature
of 1300 IS, the film is
present as pure SiO,. The peak in fig. 2c indicates a chemical shift of 4.4 eV relative to the
binding energy of Si (99.2 eV). This value is in
good agreement with those (4.3-4.5 eV> of SiO,
reported in previous pubIicati~ns [4,7,8,27].
The XPS peaks corresponding to silicon in
figs. t-3 exhibit binding energies of 99.4-99.9
eV, a few tenths of an eV higher than that of
single crystal Si (99.2 eVl [27]. This difference
may be due to the fact that the silicon atoms are
~nt~rrn~ed with the silicon oxides and thus are
electronically perturbed with respect to bulk Si. It
is also possible that the Si peaks are a convolution of Si and Si’ +. This would also shift the
maximum of the resultant peak to a higher binding energy.
The binding energy and symmetric lineshape
of spectrum c in fig. 2 indicate that the film
consists of only SiO,. The stoichiometry was calcufated using the integrated intensities of Si and
0, normalized by the atomic sensitivity factors
[27]. This calculation yielded an 0 to Si atomic
ratia of 2 : 1, further confirm@ the existence of a
singIe phase of SiO, in the film. The stoichiome-
oxide films on Mo(100)
try of the film prepared at room temperature
without annealing is estimated to be 1.7: 1, consistent with a film which has a composition of
82% SiO,, 13% suboxide and 5% silicon as shown
in the fig, 2a.
The disappearance of the SiO, species from
the film upon annealing is 1ikeIy via reaction with
oxygen within the film to form SiO,. This reaction is apparent by the increase in the SiO, peak
intensity and a concurrent decrease in the SiO,
peak intensity (figs. 2 and 31, while the stoichiometry of the film remains unchanged in the 4001000 K annealing temperature range (fig. 5). The
interaction between SiO, and 0 upon annealing
has also been observed in previous studies of the
oxidation of Si crystals [4,61. The oxygen involved
in the post-reaction in the present work was likely
trapped in the film during SiO, synthesis I.51or
was contributed by the decomposition of [SiO,]
moieties [19,20]. The increase in the film stoichiometry at 1100-1400 K is associated with the
disappearance of the Si peak in fig. 2. The Si
appears to have interacted with SiO, to form
volatile SiO that immediately desorbed [20].
The Oils) spectra of the silicon oxide film
show a single peak with a binding energy of 533.0
eV and a FWHM of 2 eV, both identical to the
values reported for bulk SiO, [18]. It seems that
the oxygen involved in the SiO, post-oxidation
has a binding energy similar to that of 0 in SiO,
and thus is not discernable in the present XPS
spectra. Annealing the film to 1373 K causes a 0.3
eV shift toward a higher binding energy (figs. 4
and 5). The Si(2p) peak also shifts 0.4 eV upwards (fig. 5). The variation of O(ls) and Si(2p)
binding energies has been measured in detail as a
function of the thickness of the SiO, films on Si
[183. The present study also found that the Si(2p)
and O&s) binding energies change with the fifm
thickness. These shifts can be attributed to several physical processes such as extra-atomic reIaxation due to a high defect concentration [18],
band bending in the SiO, film that causes a static
surface potential change [28,29], or a local charging effect as the small islands of SiO, nucleate
into large clusters.
Fig. 6 shows that upon film deposition and
subsequent annealing of the sample, the Mo(3d)
J.-W He et al. / Characterization of ultra-thin silicon oxide films on Mo(IO0)
125
stoichiometric SiO, film can0 be obtained with a
Si deposition rate of N 1.2 A/min at an oxygen
pressure of 2 X 10e5 Torr at 300 K, followed by
an annealing of the film to N 1300 K. The film
prepared at room temperature without annealing
was found to contain a small fraction of suboxides
and silicon which undergo post-oxidation at an
elevated temperature,
forming either SiO, or
volatile SiO.
MoJd
Acknowledgement
236
229
222
BINDING ENERGY (eV)
Fig. 6. The Md3d) XPS spectra of a Mo(100) surface covered
with a silicon oxide film (36 A). The film was prepared by
evaporating Si in an 0.z pressure of 2 x 10m5 Torr with a
dosing rate of - 1.2 A/min at a sample temperature of
- 350 K. The sample was then annealed to the indicated
substrate temperatures.
XPS spectra show no measurable change either
in the peak binding energy or in the peak FWHM.
This rules out any significant interaction between
the MO substrate with the SiO, film at a substrate temperature below 1100 K. Previous TPD
studies have indicated that the SiO, was reduced
into volatile SiO by the substrate MO in the
1400-1700 K temperature range, and the products SiO and MOO, desorbed promptly [20]. The
present XPS spectra detected no significant
molybdenum oxide remaining after annealing the
film to 1473 K and thus are consistent with the
reduction mechanism previously proposed [20].
5. Summary
We have reported the experimental results regarding the synthesis and characterization of ultra-thin silicon oxide films on a Mo(100) surface
using XPS. It was shown that the composition of
the film, prepared by evaporating Si in an oxygen
ambient with a substrate temperature of * 300
K, varies strongly with the oxygen pressure. A
We acknowledge with pleasure the partial support of this work by the Department of Energy,
Office of Basic Energy Science, Division of
Chemical Sciences.
References
[ll N.F. Mott, S. Rigo, F. Rochet and A.M. Stoneham,
Philos. Mag. 60 (1989) 189.
121M. Niwano, H. Katakura, Y. Takeda, Y. Takakuwa, N.
Miyamoto and M. Maki, J. Vat. Sci. Technol. A 10 (19921
339.
[31 N. Aoto, E. Ikawa, N. Endo and Y. Kurogi, Surf. Sci. 234
(1990) 121.
[41 V.A. Labunov and P.V. Protasevich, Nucl. Instrum.
Methods B 39 (1989) 466.
151A. Feldman, Y.N. Sun and E.N. Farabaugh, J. Appl.
Phys. 63 (1988) 2149.
161W. Braun and H. Kuhlenbeck, Surf. Sci. 180 (1987) 279.
[71 F.J. Grunthaner, P.J. Grunthaner, R.P. Vasquez, B.F.
Lewis and J. Masetjian and A. Madhukar, Phys. Rev.
Lett. 43 (1979) 1683.
181 G. Hollinger and F.J. Himpsel, Phys. Rev. B 28 (1983)
3651.
191A. Bianconi and R.S. Bauer, Surf. Sci. 99 (1980) 76.
[lOI T. Adachi and C.R. Helms, Appl. Phys. L&t. 35 (1979)
15.
[ill H. Ibach and J.E. Rowe, Phys. Rev. B 9 (1974) 1951.
[12] J. Derrien and M. Commandre, Surf. Sci. 118 (19821 32.
[13] W.C.M. Claassen, R.W.A.H. Schmitz and J. Dieleman,
Appl. Surf. Sci. 36 (19891 240.
[14] M. Nadamura, Y. Mochizuki, K. Usami, Y. Itoh and T.
Nozaki, Solid State Commun. 50 (1984) 1079.
[15] P.G. Pai, S.S. Chao, Y. Takagi and G. Lucovsky, J. Vat.
Sci. Technol. A 4 (1986) 689.
[16] S.H. Garofalini, J. Chem. Phys. 78 (1983) 2069.
[17] V.M. Bermudez, Surf. Sci. Lett. 230 (1990) L155.
[18] J. Finster, D. Schulze, F. Bechstedt and A. Meisel, Surf.
Sci. 152/153 (1985) 1063.
126
J.-W. He et al. / Characterization
of rdtra-thin silicon oxidefilms
[19] X. Xu and D.W. Goodman, Appl. Phys. Lett., in press.
[20] X. Xu and D.W. Goodman, to be published.
1211 M.-C. Wu, J.S. Corneille, S.A. Cesar, J.-W. He and D.W.
Goodman, Chem. Phys. Lctt. 182 1199if 472.
[22] J.S. Corneille, J.-W. He and D.W. Goodman, to be
published.
1231 R.A. Campbell and D.W. Goodman, Rev. Sci. &strum.
63 f1992) 172.
1241 M. Grunze, H. Ruppender and 0. Eishazly, J. Vat. Sci.
Technol. A 6 (1988) 1266.
f25] R. Flits& and S.I. Raider, J. Vat. Sci. Technoi. 12 (1975)
305.
on Mo(lOOl
[26] M.F. Hochella and AH. Carim, Surf. Sci. 197 (1988)
L260.
I271 CD. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder.
G.E. Muilenberg, Eds. Handbook of X-ray Photoelectron
Spectroscopy (Perkin-Elmer, Eden Prairie, MN, 1979).
[28] R.W. Grant, S.P. Kowalczyk, J.R. Waldrop and WA.
Hill, in: The Physics of MOS Insulators, Eds. G. Lucovsky, S.T. Pantelides and F.L. Galeener (Pergamon, New
York, 1980) p. 202.
[29j W.F. Egelhoff, Surf. Sci. Rep. 6 (1987) 253.