1
12
I. Preda et al. / Journal of Electron Spectroscopy and Related Phenomena 156–158 (2007) 111–114
1
s XAS spectra of the nanostructured NiO system studied here
which are completely consistent with the AFM data.
2
. Experimental
Theexperimentwasperformedinanultrahighvacuumcham-
ber located at the PM4 beam-line in the Synchrotron BESSY
Berlin). Successive depositions were performed and analyzed
(
by XAS. NiO was deposited at room temperature on HOPG
by reactive evaporation from a pure Ni filament. The evapora-
−
5
tion was performed in an oxygen atmosphere (5 × 10 Torr)
at room temperature. The evaporation rate was constant and
maintained low enough to allow the study of the early stages
of NiO growth. After deposition for large coverages, the
◦
sample was annealed at 400 C for 30 min in ultra high vac-
uum. More details on the preparation procedure can be found
elsewhere [4].
The XAS measurements were performed at the PM4 beam-
line of the BESSY II storage ring (Berlin). The spectra were
collected in the total electron yield detection mode at graz-
ing incidence to take advantage of the linear polarization of
the synchrotron light in order to enhance transitions towards
atoms in the direction to the surface normal (surface enhance-
ment). The estimated overall resolution of the plane grating
monochromator (PM4) was better than 100 meV at the O 1s
edge (530 eV). The spectra were normalized to the I0 current
coming from a fresh gold sample in order to correct the spec-
tra from the contamination of the optical elements and beam
losses.
Fig. 1. Ni 2p XAS spectra as a function of the NiO coverage.
state of the Ni atoms remains unchanged and the same as in bulk
NiO.
3
. Results and discussion
In order to corroborate the above results, we have compared
the Ni 2p XAS spectra with theory. According to de Groot et al.
[2,3], the spectra can be reproduced by calculating the 2p → 3d
multiplet transitions as a first step. In the case of NiO, these
transitions have to be projected in an octahedral crystal field of
certain strength. We have performed 2p → 3d multiplet calcula-
The Ni 2p XAS spectra as a function of the NiO coverage
are shown in Fig. 1. We have to note here that the coverages
have been calculated from the intensities of the XAS spectra
following conventional methods and assuming a layer-by-layer
way of growth at constant evaporation rate. Since the AFM data
show that the way of growth of NiO/HOPG is dominated by the
formation of nanometric planar islands, the coverage should be
understood as an indication of the equivalent material needed to
complete such layers. At first sight, all the spectra through the
series are similar. First of all we want to note that the spectrum
2+
tion for Ni ions in different symmetries: high spin Ni (Fig. 2a),
low spin Ni 2 (Fig. 2b) and Ni (Fig. 2c). The crystal field
parameter (10Dq) was set to 1.8 eV in all calculations. From
the comparison of the calculations with the experimental Ni 2p
spectra it is inferred that, indeed, the Ni atoms involved in the
+
3+
◦
2+
of the NiO thin film annealed at 400 C matches other published
experiment remain in the high spin Ni form.
Ni 2p spectra for bulk NiO [8,9]. Also, the spectra for large cov-
erages, prior to annealing, are almost identical. This is a clear
indication that a stoichiometric NiO thin film can be grown on
the HOPG substrate. However, the most important feature is
that the Ni 2p XAS spectra show the same line-shape during the
whole growth process. Even for low coverages, the structures
observed in the spectra are the same, although less defined and
slightly broader, than those for bulk NiO. These minor discrep-
ancies can be attributed to disorder and consequently to a lower
value of the crystal field. The spectra also slightly shift towards
higher photon energies as the coverage increases. Although XAS
measurements cannot be referred to the Fermi level, this shift is
directly related to the increase of the thickness of the NiO iso-
lating thin film. These results seem to indicate that the oxidation
Fig. 3 shows the O 1s XAS spectra as a function of the
coverage. Once again, the spectrum of the annealed thin film
is identical to other spectra published elsewhere for bulk NiO
[10,11]. The first peak at 530.5 eV is assigned to the hybridiza-
tion of O p states with Ni 3d (eg) states whereas the broad
peaks located at 536.0 and 539.0 eV correspond to O p states
hybridized with Ni 4s and Ni 4p, respectively. The spectra for
high coverages prior to annealing are also similar to that of bulk
NiO. This supports again the growth of a stoichiometric NiO
thin film on the HOPG substrate.
However, for low coverages (0.5 ML in Fig. 3) the spec-
trum differs strongly from the bulk spectrum. The single peak at
threshold, assigned to eg states in bulk NiO, is now split forming
a broad band. It is clearly seen that the initial broad band for the