Study by static SIMS, XPS and UPS of the adsorption of cyanogen on (100) Ni surfaces

Article abstract of DOI:10.1016/S0039-6028(00)00445-3

The interaction of cyanogen with (100) Ni surfaces at room temperature was studied using secondary ion mass spectrometry in a static mode, and photoemission spectroscopies (XPS and UPS). It has been shown that cyanogen is adsorbed on this nickel surface in a dissociative mode: atomic carbon and nitrogen are present on the surface as well as CN fragments. These adsorbed species have been characterized by XPS and UPS. Increasing the temperature in the 400 K range increases the rate of the dissociation reaction of CN into atomic carbon and nitrogen. Moreover carbon is found to dissolved into the bulk for temperatures as low as 475 K. No evidence has been found of a polymerized form of cyanogen molecules on this nickel surface, when the temperature is increased unlike the results obtained for the (111) Ni surface.

Full text of DOI:10.1016/S0039-6028(00)00445-3

Surface Science 458 (2000) 239–246
www.elsevier.nl/locate/susc
Study by static SIMS, XPS and UPS of the adsorption of
cyanogen on (100) Ni surfaces
R. Gouttebaron 1, S. Bourgeois *, M. Perdereau
Laboratoire de Recherches sur la R e´ activit e´ des Solides, UMR 5613, CNRS/Universit e´ de Bourgogne, UFR Sciences et Techniques,
BP 47870, 21078 Dijon Cedex, France
Received 16 January 2000; accepted for publication 16 March 2000
Abstract
The interaction of cyanogen with (100) Ni surfaces at room temperature was studied using secondary ion mass
spectrometry in a static mode, and photoemission spectroscopies ( XPS and UPS).
It has been shown that cyanogen is adsorbed on this nickel surface in a dissociative mode: atomic carbon and
nitrogen are present on the surface as well as CN fragments. These adsorbed species have been characterized by XPS
and UPS.
Increasing the temperature in the 400 K range increases the rate of the dissociation reaction of CN into atomic
carbon and nitrogen. Moreover carbon is found to dissolved into the bulk for temperatures as low as 475 K. No
evidence has been found of a polymerized form of cyanogen molecules on this nickel surface, when the temperature
is increased unlike the results obtained for the (111) Ni surface. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: Chemisorption; Cyanogen; Nickel; Secondary ion mass spectroscopy; Visible and ultraviolet photoelectron spectroscopy;
X-ray photoelectron spectroscopy
1
. Introduction
of a (110) Ni surface a c(2×2) structure is
obtained by LEED for a cyanogen coverage of
about one monolayer [1]. TDS and ARUPS
experiments indicate the presence of intact C N
The surface chemistry of small molecules on
transition and noble metal surfaces is the subject
of much investigation because of its relevance to
catalytic reactions. (CN) is with HCN the least
2
2
molecules in this monolayer phase. By ARUPS
five spectral bands are seen in the valence band
region of adsorbed C N due to emission from
2
complicated molecule containing the carbon–nitro-
2
g
2
gen triple bond. However, there are only few
studies of the adsorption of cyanogen on nickel
surfaces.
4s , 5s , 1p and 1p . This is the result of the
u
g
u
lowering of symmetry of the adsorbed C N and
2
2
a lifting of the degeneracy of the p states.
Molecules are oriented with their molecular axes
along [001] that is perpendicular to the troughs.
For exposures at temperatures lower than 250 K
Above 250 K dissociation of (CN) into CN
occurs upon adsorption. On a (111) Ni surface,
2
*
Corresponding author. Fax: +33 03 80 39 61 32 E-mail
address: sylvie.bourgeois@satie.u-bourgogne.fr (S. Bourgeois)
Present address: L.A.S.S.I.E, Materia Nova, Parc
adsorption of (CN) at 300 K [2] produces a
1
2
disordered surface structure. A dissociative chemi-
Scientifique Initialis, Avenue Nicolas Copernic, 7000 Mons,
Belgium.
sorption occurs, the CN group being more or less
0
039-6028/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S0039-6028 ( 00 ) 00445-3

2
40
R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
parallel to the surface with both carbon and nitro-
gen atoms bonded to the surface. However, in a
theoretical study using a cluster model and the
atom superposition and electron delocalization
molecular orbital (ASED-MO) method [3], it is
found that for CN adsorption on (111) Ni, the
end-on (through the carbon) configuration is more
favoured than the lying-down or other configura-
tions. On this surface, upon heating the cyanogen
monolayer to 400 K [2], a (6×6) surface structure
is formed. This kind of ordering might result from
a polymerization of the cyanide groups to form
quadrupolar mass spectrometer in a static mode:
low energy (1 keV) and low density (6×1010
ions cm−2 s−1) primary argon ions were used in
order to work in static SIMS conditions [4].
Furthermore the total energy bandpass of the
selection–detection system was selected in order to
promote the superficial ions [10].
2.1.2. XPS and UPS experimental set-up
The experimental set-up used for XPS and UPS
experiments consisted of a preparation chamber
and an analysis chamber, the base pressure in the
chambers being in the 10−10 mbar range during
the experiments. In the preparation chamber the
sample can be dosed with different gases by mean
of a thin stainless steel tube with the out-gas being
positioned just in front of the sample surface. With
a tungsten filament mounted on a translational
motion the sample can be heated by the rear face
near 600 K.
(
C N ) .
2 3
Above 800 K all the CN are dissociated into
2
carbon and nitrogen atoms. Nitrogen atoms
desorbed from the surface as N molecules and
2
carbon atoms are dissolved into the bulk of the
sample.
In a previous study the adsorption of cyanogen
on a polycrystalline silver surface has been studied
by static secondary ion mass spectrometry (SIMS)
For XPS experiments the primary photon beam
was produced with a dual-anode X-ray tube
(Al Ka (1486.6 eV) and Mg Ka (1253.6 eV) radia-
tion). To obtain lower energy photons and thus to
work in the UPS mode an ultraviolet helium lamp
was used. In these two spectroscopic techniques
emerging photoelectrons were analysed with detec-
tion normal to the surface using a Riber Mac2
semi-imaging spectrometer having an angular
acceptance of about 22°. XPS spectra were cal-
ibrated to the Au 4f peak at a binding energy
[
4] and a reaction of polymerisation of cyanogen
has been demonstrated upon heating treatment
at 500 K.
In this work the adsorption of cyanogen on
nickel surfaces has also been studied by SIMS in
order to determine the adsorption mode of cyano-
gen on the (100) Ni surface and to check the
results concerning the CN polymerization on
nickel near 400 K [2]. SIMS, when used in a static,
low damage mode, can provide information on
the species present on the surface, with a particu-
larly high sensitivity. Moreover, by means of the
observed molecular ions, information on the
adsorption mode can be obtained [5–8]. XPS and
UPS experiments were also performed in order to
identify the chemical nature of the adsorbate and
to determine its adsorption geometry on the
surface.
7/2
of 84.0 eV. In XPS experiments, the working con-
ditions were chosen in order to have a total
resolution of the system equal to about 1.2 eV. In
this value both the broadness of the source radia-
tion and the analyser resolution were taken into
account. For UPS experiments the chosen total
energy resolution was about 150 meV as deter-
mined from the Fermi edge. During UPS experi-
ments the sample was polarized with a low negative
voltage of −15 V in order to increase the kinetic
energy of photoelectrons because of the poor sensi-
tivity of this spectrometer for electrons having a
kinetic energy lower than 10 eV.
2
2
2
. Experimental
.1. Methods
.1.1. Static SIMS experimental set-up
2.2. Sample preparation
The experimental ultrahigh vacuum apparatus
has been described in detail elsewhere [9]. SIMS
experiments were carried out with a Riber Q 156
Samples used were monocrystalline nickel sheets
10 mm×10 mm×2 mm prepared ex situ by elec-

R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
241
trochemical polishing. These nickel samples have
100) orientation. In the experimental set-up the
(
samples were submitted to classical preparation
consisting of repetitive cycles of Ar ion sputtering
(
500 eV, 1 mA cm−2) and heating (1000 or 600 K
depending on the experimental vessel). The purity
of the surface was checked by AES, SIMS or XPS.
The exposures of the surfaces to (CN) were
2
performed at room temperature unless otherwise
specified. Cyanogen was prepared in situ by ther-
mal decomposition at 630 K of AgCN [11,12].
The maximum pressure of the reacting species was
1
0−6 mbar as measured with ionization gauges
located near the pumping units. It should be noted
that this pressure value was underestimated by
about one order of magnitude owing to the intro-
duction of the gas just in front of the sample by
means of a capillary [4]. This fact has two conse-
quences: the exposure values, expressed in
Langmuir, are underestimated and cyanogen doses
are not directly comparable for the two experimen-
tal apparatuses (SIMS and XPS, UPS) because
the geometries are not the same.
Fig. 1. Evolutions of the negative secondary ionic intensities as
functions of the exposure to cyanogen.
3
. C N exposures studied by static SIMS
2 2
3.1. Results and discussion
A nickel sample is dosed with cyanogen at room
temperature between 0 and 120 L (experimental
conditions: (CN) pressure=10−8 mbar, exposure
2
time t=1000 s). After each 10 L exposure the
surface is analysed by static SIMS. Positive and
negative secondary ions such as Ni CN+,
2
Ni N+, Ni C+ and Ni(CN)−, CN− are detected
2
2
2
in addition to ions involving only nickel. The
evolutions of the intensities of negative secondary
ions as function of the cyanogen exposure are
presented in Fig. 1 and those of positive ions in
Fig. 2. It can be observed in Fig. 1 that saturation
of the surface in cyanide species occurs for cyano-
gen exposures of 20 L. Moreover, secondary ions
such as Ni CN+, Ni N+ and Ni C+ are attributed
2
2
2
to the presence on the nickel surface of CN(a),
N(a) and C(a) ((a) stands for adsorbed species).
Their detection at cyanogen exposures as low as
Fig. 2. Evolutions of the positive secondary ionic intensities as
functions of the exposure to cyanogen (#, Ni C+; $,
2
10 L (Fig. 2) reveals that a dissociative adsorption
Ni N+; dashed line, Ni CN+).
2
2

2
42
R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
of (CN) into CN(a), N(a) and C(a) takes place
2
on the (100) Ni surface upon cyanogen dosing at
room temperature. These static SIMS results are
in agreement with previous results obtained for
cyanogen adsorption on Ni (111) [2].
After saturation of the Ni (100) surface with
cyanogen at room temperature (total exposure
higher than 20 L), the temperature of the sample
is increased gradually in steps of 50 K up to 800 K
in order to study the evolution of cyanide species
adsorbed on the surface sample as a function of
its temperature. After each temperature increase,
the surface is analysed by static SIMS and the
SIMS spectra are characteristic of compounds still
adsorbed on the nickel surface at a given temper-
ature. This kind of experiment can be compared
with a ‘negative’ TPD study.
CN(a), C(a) and N(a) are observed to be
adsorbed on the surface because of the detection
of the same secondary ions as previously. The
Fig. 3. Evolution of the negative secondary ionic intensities as
a function of the temperature.
intensity of CN− and Ni(CN)− and the yields,
2
Ni CN+/Ni+, Ni N+/Ni+ and Ni C+/Ni+ are
2
2
2
2
2
2
plotted versus temperature in order to study the
evolutions of the quantity of CN(a), N(a) and
C(a) species adsorbed on the surface with temper-
ature (these yields can be related to the amount
of adsorbed species [6]).
the nickel surface occurs through the cracking of
adsorbed CN species to C(a) and N(a). Moreover,
during this heating treatment of the sample there
is no evidence of the appearance of new secondary
ions which would be the expression of the presence
on the surface of a polymerized compound of
cyanogen.
These evolutions are presented in Figs. 3 and 4,
respectively. Between 300 and 375K, no important
variations of the secondary ionic intensities, either
negative or positive is observed. Only a small
evolution is observed for the positive ions ratios
(
Fig. 4):
Ni CN+/Ni+
decreases
when
4. C N exposures studied by XPS and UPS
2
2
2 2
Ni N+/Ni+ and Ni C+/Ni+ increase. This can be
2
2
2
2
accounted for by a dissociation reaction of CN(a)
into C(a) and N(a). Moreover, between 375 and
4.1. Results and discussion
4
75 K, Ni(CN)− and CN− intensities decrease
In order to confirm the SIMS results and to
determine the adsorption geometry of CN frag-
ments on the nickel surface, studies by XPS and
UPS of the cyanogen/Ni (100) system have been
performed.
2
rapidly as well as the Ni CN+/Ni+ ratio as can
2
2
be seen in Figs. 3 and 4. The kinetics of the last
dissociation reaction increases. In the same temper-
ature range, Ni N+/Ni+ is multiplied by a factor
2
2
of 3 and Ni C+/Ni+ goes to zero (Fig. 4). This
2
2
last point can be accounted for by carbon dissolu-
tion into the bulk of the sample. Above 475 K, all
the cyanide species are desorbed from the nickel
surface and at 575 K the SIMS spectrum is charac-
teristic of a clean nickel surface.
4.1.1. XPS results
After obtaining a nickel surface free of contami-
nants, the surface is exposed two times to
10−7 mbar of cyanogen during 30 s (3 and 6 L of
cyanogen). On XPS spectra these exposures are
characterized by the appearance of a C 1s and a
Thus, the desorption of the cyanide species from

R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
243
Fig. 4. Evolution of positive secondary ionic ratios as a function of the temperature.
N 1s peak positioned respectively at binding ener-
gies equal to 284.1 and 397.3 eV (Fig. 5).
C 1s and N 1s peak areas (I ) are propor-
our experimental conditions:
I
3N
s
E−1.
C 1s(N 1s)
C(N) C 1s(N 1s)
C 1s(N 1s)
tional to the number of atoms on the surface
In this formula the mean free paths of the C 1s
(
N
) multiplied by the ionization cross-section
C(N)
corresponding to the 1s level of these elements
and N 1s photoelectrons were not taken into
account because C and N atoms are located on
the sample surface. This formula has been applied
to our experimental results obtained after 3 and
(s
) and by the transmission function of the
C 1s(N 1s)
spectrometer which is proportional to E−1 under
6
L of cyanogen exposure (Table 1). C/N values
which are obtained (close to unity) confirm the
presence of cyanide species adsorbed on the nickel
surface. In these molecules all carbon atoms and
nitrogen atoms are equivalent because only one
C 1s peak and one N 1s peak are observed.
Moreover, under our experimental conditions a
cyanogen dose of 3 L is enough to saturate the Ni
(
100) surface with cyanogen because no differences
Table 1
Atomic ratios between carbon and nitrogen obtained from XPS
experiments
(
CN)
I
I
C/N with
2
C 1s
N 1s
exposure
L)
(counts s−1
(counts s−1
s
=1,
C 1s
N 1s
(
eV−1)
eV−1)
s
=1.8 [13]
3
6
1040
1085
2000
1912
0.98
1.07
Fig. 5. C 1s and N 1s XPS spectra for 3 L (dashed line) and 6 L
(
solid line) cyanogen exposures.

2
44
R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
are observed in the spectra between 3 and 6 L
cyanogen exposures.
When an Ru (100) surface is exposed to cyano-
gen [14] at 100 K the presence of molecular cyano-
gen is characterized by an N 1s peak at a binding
energy of 401.3 eV. This peak shifts to 397.6 eV
when the temperature is increased to 250 K. This
shift can be attributed to the dissociation of
(
CN) into CN on the Ru (100) surface. Moreover,
2
an XPS study of the cyanogen/Si (100) and Si
111) systems [15] has shown that at low temper-
(
ature (100 K), a multilayer of molecular cyanogen
is present on these surfaces and is characterized
by photoelectrons peaks C 1s and N 1s positioned
at binding energies equal to 287.6 and 399.8 eV,
respectively. At about 250 K, this multilayer is
desorbed and a monolayer of molecular cyanogen
is adsorbed on the silicon surface. These cyanogen
molecules interact more strongly with the sub-
strate, this interaction inducing a 1.8 eV shift of
the C 1s and N 1s peaks towards the lower binding
energies: C 1s at 285.8 eV and N 1s at 398 eV. At
Fig. 6. UPS spectra using He II radiation of a clean Ni (100)
surface (a) and after dosing of the surface to 3 L of cyanogen
(b). A zoom of the zone corresponding to binding energies
between −4 and −14 eV is also shown.
6
50 K, the dissociation of (CN) into 2(CN) is
2
evidenced by a shift of C 1s to a binding energy
equal to 285.4 eV. The total dissociation of cyano-
gen into atomic carbon and nitrogen on Si(100)
and (111) surfaces which occurs at 800 K corres-
ponds to the appearance of a C 1s peak at 283.5 eV
and of a N 1s peak at 397.6 eV.
the Ni 2p line, of a significant charge transfer from
nickel atoms to CN fragments.
4.1.2. UPS results
When comparing the results of this work on
the cyanogen/Ni (100) system with results in the
literature, the C 1s and N 1s spectra obtained here
are characteristic of the presence on the surface of
carbon and nitrogen atoms and/or of CN frag-
ments in which the triple bond would be weakened
because of an interaction of C and N atoms with
the substrate. However, it should be noted that
cyanogen adsorption on Ni (100) induces no
energy shift and no shape modification of the
Ni 2p peak. This peak remains positioned at a
binding energy of 852.6 eV which is characteristic
of metallic nickel [16] and its FWHM remains
equal to 1.62 eV. Although the sensitivity to sur-
face nickel atoms is not so good in this experiment
because of the detection of the photoelectrons in
a direction normal to the surface, no direct evi-
dence can be found in these XPS results concerning
UPS experiments using He I and He II radiation
are performed in parallel to XPS analysis. A UPS
spectrum characteristic of a clean Ni (100) surface
is presented in Fig. 6a using He II primary radia-
tion. It is composed mainly of a thin structure at
0.2 eV which can be attributed to a surface
emission phenomena [17] and of the satellite peak
at 6 eV below the Fermi level.
−
When this surface is dosed with 3 L of cyanogen
Fig. 6b) a large decrease in the intensity of the
(
band at −0.2 eV is observed as well as a new
structure between −5 eV and −14 eV characteris-
tic of the adsorbate. In order to identify more
precisely this new structure a difference spectrum
between these two spectra has to be done. In doing
this difference a problem arises from the presence
of the nickel satellite peak at −6 eV that is to say
in the same energy range as the photoemission
3/2

R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
245
Table 2
Ionization cross-sections for carbon and nitrogen [18]
He I
He II
s
s
s
s
1.230
1.170
1.086
1.875
4.351
C 2s
N 2s
C 2p
N 2p
6.128
9.588
sented in Table 2 [18]. However, the energy width
of the structure in the difference spectrum is too
large to be attributed only to the atomic 2p bands
of carbon and nitrogen. For this reason photoemis-
sion bands characteristic of molecular orbitals 4s,
1p and 5s of CN fragments should also be present
between −11 and −5 eV [19–21]. However, it is
impossible to go further and to attribute exactly
to each molecular orbital of CN fragments a
binding energy in the difference spectrum.
Fig. 7. Difference spectrum between the UPS spectrum of a
nickel surface saturated with cyanogen and the UPS spectrum
of a clean Ni (100) surface.
5. Conclusion
bands of adsorbates. This problem was solved by
suppressing the contribution of this satellite signal
in the two UPS spectra before subtraction. The
result of this mathematical treatment between ‘ini-
tial spectrum: clean Ni (100)’ and ‘final spectrum:
Ni (100) saturated with 3 L of cyanogen’ is pre-
sented in Fig. 7.
The results of this study by SIMS, XPS and
UPS of the cyanogen/Ni (100) system have shown
that at room temperature cyanogen is adsorbed
on a Ni(100) surface in a dissociative mode and
that reactions (1) and (2) take place on the nickel
surface.
Referring to SIMS results described previously
where the presence of C and N atoms has been
demonstrated on the nickel surface and using UPS
results obtained in different studies on transition
metal carbides and nitrides, the peaks at −13.5 eV
and at −12 eV can be ascribed respectively to the
(CN) 2CN(a),
(g)
CN(a)C(a)+N(a).
(1)
(2)
2
This dissociative adsorption is characterized with
SIMS by the presence of secondary ions such as
Ni CN+, Ni C+, Ni N+, CN− and Ni(CN)−.
2
2
2
2
2s bands of atomic nitrogen and atomic carbon
The appearance on the surface of C(a), N(a)
and of CN fragments in which carbon and nitrogen
atoms interact strongly with the substrate atoms
is characterized by the presence in the XPS spectra
of a C 1s peak at a binding energy of 284.1 eV
and of a N 1s peak at 397.3 eV with a ratio C/N
equal to unity after correction of the peak areas
by the corresponding relative sensitivity factors.
UPS results confirm the dissociative adsorption
reaction of cyanogen on Ni(100) surface and the
presence on the surface of C(a), N(a) and CN(a).
The atomic 2s bands of carbon and nitrogen are
adsorbed on the sample surface. Because of the
detection of these 2s bands the 2p bands character-
istic of carbon and nitrogen atoms have to be
present in the difference spectrum and be posi-
tioned at binding energies 5–6 eV lower than the
corresponding 2s bands. That is why in the struc-
ture characteristic of the adsorbate molecular
orbitals presented in Fig. 7, it is possible to locate
these 2p bands around −7 eV, the contribution of
the N 2p band being greater than that of the C 2p
band as regards the ionization cross-sections pre-

2
46
R. Gouttebaron et al. / Surface Science 458 (2000) 239–246
[
[
3] Z. Xu-Yan, S. Dan Hua, C. Pei Lin, Surf. Sci. 223
1989) 393.
4] S. Bourgeois, R. Gouttebaron, M. Perdereau, Surf. Sci.
95 (1997) 356.
positioned at binding energies of −12 eV and
13.5 eV respectively, the corresponding 2p bands
around −7 eV and the bands characteristic of
molecular orbitals of CN between −11 and
(
−
3
[5] C.M. Greenlief, P.L. Radloff, S.A. Akhter, J.M. White,
−5 eV. These last UPS results concerning the
Surf. Sci. 186 (1987) 563.
[
[
[
6] A. Brown, J.C. Vickerman, Surf. Sci. 124 (1983) 267.
7] V. Matolin, E. Gillet, Surf. Sci. 238 (1990) 75.
8] M. Barber, J.C. Vickerman, Surf. Sci. 68 (1977) 130.
adsorbed CN fragments can be related to that
obtained in the case of cyanogen adsorption on a
palladium surface [19] where it was deduced that
the CN fragments are adsorbed parallel to the
metal surface.
[9] S. Bourgeois, M. Perdereau, J. Spectrosc. Electron. 6
(1981) 335.
10] S. Bourgeois, M. Perdereau, Le Vide, Les Couches Minces
[
3
8 (1983) 183.
Increasing the temperature in the 400 K range
increases the rate of the dissociation reaction of
CN into atomic carbon and nitrogen. Carbon is
found to be dissolved into the bulk for temper-
atures as low as 475 K. Moreover, no polymeriza-
tion reaction of cyanogen on this nickel surface is
detected upon heating treatments unlike the results
obtained for the cyanogen adsorption on (111) Ni
[
[
11] F.P. Netzer, Surf. Sci. 61 (1976) 343.
12] R. Gouttebaron, Ph.D. Thesis, University of Bourgogne,
1996.
[
13] J.H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8
(
1976) 129.
[
[
14] N.J. Gudde, R.M. Lambert, Surf. Sci. 124 (1983) 372.
15] Y. Bu, L. Ma, M.C. Lin, J. Phys. Chem. 99 (1995) 1046.
[16] Auger and X-ray Photoelectron Spectroscopy, D. Briggs,
M.P. Seah (Eds.), Practical Surface Analysis, second ed.,
Vol. 1 Wiley, New York, 1990.
[
[
[
[
[
2] or on silver [4].
17] T.T.A. Nguyen, R. Cinti, J. Micros. Spectrosc. Electron.
(1976) 157.
18] J.J. Yeh, I. Lindau, At. Data Nucl. Data Tables 32 (1)
1985).
19] M.G. Ramsey, G. Rosina, F.P. Netzer, Surf. Sci. 218
1989) 317.
1
(
References
(
20] M.G. Ramsey, G. Rosina, F.P. Netzer, Surf. Sci. 217
(1989) 140.
[
[
1] N. R o¨ sch, T. Fox, F.P. Netzer, M.G. Ramsey, D. Stein-
m u¨ ller, J. Chem. Phys. 94 (1991) 3276.
2] J.C. Hemminger, E.L. Muetterties, G.A. Somorjai, J. Am.
Chem. Soc. 101 (1979) 62.
[21] D.W. Turner, C. Backer, A.D. Backer, C.R. Brundle,
Molecular Photoelectron Spectroscopy, Wiley, New
York, 1970.