ELSEVIER Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261
JOURNAL OF
ELECTRON SPECTROSCOPY
and Related Phenomena
XPS studies of V205, V6013 , VO 2 and V203
J. Mendialdua a, R. Casanova a'*, Y. Barbaux b
aLaboratorio de Superficies, Departamento de Fisica, Facultad de Ciencias, Universidad de Los Andes, M&ida, Venezuela
bLaboratorie de Catalyse H&&ogbne et homogbne, URA CNRS # 402, UniversitO des Sciences et Techniques de Lille, 59655 Villenueve
D'Ascq Cedex, France
First received 30 March 1994; in final form 4 November 1994
Abstract
A detailed XPS study of several oxides of vanadium is reported in this work, in an attempt to characterize clearly the
surface of these oxides. Several parameters, such as the FWHM of the V2p3/2 and Ols XPS peaks, their shape and
binding energy difference, have been utilized. The characterization is extended to these oxides following different sample
treatments. The effect of the presence or absence of adventitious carbon on the behaviour of the V205 sample is
investigated when the sample is heated in UHV. It is found that the X-ray beam has some effect on the properties of
both crystalline and polycrystalline V205 covered with Cd.
Keywords: Vanadium oxide; X-ray photoelectron spectroscopy
1. Introduction
In the field of heterogeneous catalysts, it is of
fundamental importance to know the state of the
surface of a given catalyst with high accuracy.
Vanadium pentoxide V205 is a widely used cata-
lyst in a variety of chemical reactions, for example,
the oxidation of orthoxylene to phthalic anhydride
[1,4]. There is great controversy about the surface
characterization of this catalyst, such that the
nature of its surface remains obscure. The surface
of this catalyst could be V205 or a lower vanadium
oxide.
In this work, we present a detailed XPS study of
several vanadium oxides in an attempt to identify
clearly the different oxidation states present on the
surface of these oxides. For this purpose we have
* Corresponding author.
analysed several parameters, including the FWHM
of the peaks V2p3/2 and O 1 s, their peak shape, and
the binding energy difference of these peaks.
2. Experimental
Four oxides, V205, V6013, VO2 and V203, were
studied in the present investigation. V203 was
studied in a polycrystalline form, prepared by
hydrogen reduction of polycrystalline V205 at
500°C for 24 h. The V203 stoichiometry was
determined from thermogravimetric and X-ray
diffraction studies. The polycrystalline V205
employed in the preparation of V203 was supplied
by the Catalysis Institute of Cracovia (Poland)
and was prepared by decomposition, in air at
500°C, of ammonium metavandadate [1]. The
specific surface area of the polycrystalline V203
0368-2048/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved
SSDI 0368-2048(94)02291-7
250 J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261
was 4.2m 2 g-l. This oxide was subjected to a reduc-
tion treatment at 500°C in a flow of H2/He gas
mixture, in the proportion 1:3, for 24 h. The thermo-
gravimetric measurements reveal that the weight loss
of this oxide proceeds to a stationary state and cor-
responds to the transformation V205 ~ V203. In
air and at room temperature, this sample regains
weight (approximately 0.5% of the total weight
loss) indicating a surface re-oxidation.
Our thermogravimetric studies of V205 at 550°C
in a mixture of O2/N2 reveal that this oxide crystal-
lizes with many non-stoichiometric defects, since
this sample continues to gain weight even after
60h of treatment. V205 was studied as a single
crystal (010) face and also in a polycrystalline
form. Taking into account the thermogravimetric
results, the V205 samples were subjected to an oxi-
dation treatment, in a stainless steel furnace, at
550°C under oxygen for 16 h in order to guarantee
a well-oxidized surface; the samples were transfered
to the spectrometer, under an inert atmosphere,
immediately after treatment. V6013 and VO 2 were
studied in polycrystalline form without being sub-
jected to any previous treatment. The V203 samples
were left in the XPS spectrometer in UHV for 14 h in
order to get rid of excess oxygen on the surface.
The thermogravimetric measurements were per-
formed in a Sartorius microbalance. The XPS spec-
tra were taken with two spectrometers: an AEI
ES200 spectrometer and a Leybold-Heraeus spec-
trometer, both operated in the fixed retarding ratio
mode. Non-monochromatic A1 Kc~ radiation was
used as the X-ray source, with a constant power of
300 W during the XPS measurements. The Leybold
spectrometer was fitted with Ar ion etching facil-
ities. The measured FWHM of Au4f7/2 in our spec-
trometers is 1.4 eV for the Leybold and 1.3 eV for
the AEI ES200.
The V205, V2013 and VO2 were provided by the
Crystallography Laboratory of Gand University
(Belgium). The surface of our samples looked very
uniform and we found that our sample surface
charged homogeneously in the XPS experiment.
3. Results and discussion
The wide use of the C ls level associated with
adventitious carbon as energy reference in XPS
studies of the oxides of vanadium, has contributed
to the scattering in the binding energy values of
some atomic energy levels in these materials, as
given in the literature, making an accurate
identification of their oxidation state difficult. It
has been shown [5] that for vanadium oxides the
energy reference mentioned above is not a good
reference, the best energy reference being that of
the Ols level at 529.8 eV for V205 and at 530 eV
for those oxides whose FWHM is larger than that
of V205. A comparison among the different results
found in the literature can be established if we
correct the energy reference and adopt as such
that of the Ols level. Mendialdua et al. [6] have
shown that in vanadium oxides the ratio No/N v
between the oxygen concentration No and the
vanadium concentration Nv does not allow differ-
entiation between the different oxidation states at
the surface of these oxides. This difficulty in obtain-
ing the degree of oxidation of these materials has
led us to utilize three parameters, with promising
results. These parameters are: the energy difference
(A) between the binding energy of the Ols level and
that of V2p3/2, the full width (A) at half maximum
(FWHM) of each of the former XPS peaks, and
their peak shapes. In Table 1 we present our
results and the corresponding spectra are shown
in Figs. 1 8.
3.1. V:05
The XPS spectra of a polycrystalline V2O 5
sample, obtained by grinding a single crystal, are
presented in Fig. 1, in both the as-received con-
dition and after it had experienced a sequence of
sample treatments in UHV. Spectrum l(a) corre-
sponds to the as-received polycrystalline sample:
the full width at half maximum for the Ols and
V2p3/2 peaks are 1.8 and 1.5 eV respectively.
After the sample was subjected to Ar ion bombard-
ment (20 min, 10 mA, 2 keV), the XPS spectrum
reveals that the FWHM of the Ols peak is 2.2 eV
while that of V2p3/2 is 3.8 eV (see Fig. l(b)). Sub-
sequently, the sample was re-oxidized in oxygen
(1 atm in situ) at 400°C for 14 h; the corresponding
XPS spectrum is shown in Fig. l(c); in this case the
full widths at half maximum for the peaks under
J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261
Table 1
Result of XPS studies of vanadium oxides
251
Bulk phase Sample treatment Surface phase E Ols/eV A E V2p3/2/eV ), A
V205 (single crystal) 16 h in 02 at 500°C V205 529.8 1.55 517 1.3 12.8
V205 (single crystal) Heating at 360°C in UHV V409 529.8 1.75 516.75 2.5 13.05
V205 (polycrystalline) 16 h at 1 atm 02 at 500°C V205 529.8 1.60 517 1.4 12.8
V205 (polycrystalline) Heating at 300°C in UHV V409 529.8 1.85 516.8 2.25 13.0
V205 (polycrystalline) Without contamination and State between V307 and V409 529.8 1.9 516.9 1.8 12.9
heating at 550°C in UHV
V205 (polycrystalline) With contamination and V6013 529.8 2.0 516.2 3.3 13.6
heating at 500°C in UHV
V6013 (polycrystalline) Short annealing at 200°C V6OI3 530.0 2.2 516.5 3.4 13.5
in UHV
V6013 (polycrystalline) Oxidized in 02 at 300°C V3079. 529.8 1.8 516.9 1.8 12.9
for3h
VO 2 (polycrystaUine) Ion etching and annealing VO2 530.0 2.8 515.65 4.0 14.35
at 200°C in UHV
VO2 (polycrystalline) Oxidized in 02 at 300°C V307? 529.8 1.8 517.1 1.6 12.7
for5h
V203 (polycrystalline) Heating at 550°C for 14 h V203 530.0 2.0 515.15 4.8 14.84
in UHV
study, i.e. the Ols and V2p3/2, were 1.7 and 1.4 eV
respectively. We consider spectrum 1 (c) to be repre-
sentative of V205 fully oxidized at its surface (entry
3 in Table 1). From the XPS spectrum of Fig. l(d)
which corresponds to this sample after a short
annealing at 550°C (2min), we found that the
FWHM of the Ols peak changes very little
(it goes to 1.8 eV); A for the V2p3/2 peak is now 1.7
eV. After the temperature of the sample was kept at
550°C for 20 min, we obtained spectrum l(e) where
A(Ols) = 1.9 eV and (V2p3/2) -- 1.8 eV. (It is inter-
esting to note, by comparing spectra l(b) and
l(c), that in the reduced state the intensity of the
shake-up satellite associated with O ls tends to
decrease notably [7].) Spectrum l(f) corresponds
to the sample heated to 500°C for 2 h; in this case
the full widths at half maximum for the O ls and
V2p3/2 peaks were 2 and 3.3 eV respectively. As we
can see, it is possible to obtain an FWHM for the
V2p3/2 peak that is smaller than that reported in
the literature (see Table 2) when a well-defined
oxidized surface of V205 is prepared.
3.2. V6013
The XPS spectra of polycrystalline V6013 sample
are shown in Fig. 2. The sample was prepared by
grinding a single crystal. Spectrum 2(a) corre-
sponds to the sample after a short annealing
at 200°C in UHV. The obtained values of A for
the Ols and V2p3/2 peaks are 2.2 and 3.4 eV
respectively, which are notably higher than
those of V205. We consider this spectrum to be
representative of surface V6013 (entry 7 in
Table 1). After the sample was subjected to Ar ion
etching (20 min, 10 mA, 2 keV) the FWHMs of the
Ols and V2p3/2 change to 2.4 and 4.2 eV respec-
tively (this is shown in Fig. 2(b)); these widths
become 2.6 and 4.6 eV respectively, once the
sample is maintained at 300°C for 13 h in UHV
followed by Ar ion etching (20 min, 10 mA,
2 keV); this is represented in spectrum 2(c)). Sub-
sequently, the sample was annealed in oxygen at
300°C for 3 h; the corresponding spectrum is pre-
sented in Fig. 2(d); in this case, the widths of both
peaks are reduced to 1.8 eV. We can also observe in
this spectruman increase in the intensity of the Ols
shake-up. It is important to note how easy it is to
over-oxidize the surface of V6013.
3.3. V02
Fig. 3 shows several XPS spectra of a poly-
crystalline VO2 sample in both the as-received con-
dition and after being subjected to sample
treatments.
252 J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261
c:
o
v
>..
I....
z
I.U
l..-
z
IIII
............. • . . . . . . . . . . . " .
I I
....
I I I I
550 540 530 520
\
510
BINDING ENERGY(eV)
Fig. 1. XPS spectra of polycrystalline V205 showing the Ols and V2p3/2 peaks after a sequence of several sample treatments: (a)
as-received condition; (b) after Ar ion bombardment for 20 min; (c) after oxidation in 02 at 400°C for 14 h; (d) after a short annealing
(2 min) at 550°C; (e) sample temperature kept at 550°C for 20 min; (f) sample heated to 500°C for 2 h.
Spectrum 3(a) corresponds to the sample in the
as-received condition; in this case A(Ols) = 1.9 eV
and A(V2p3/2 ) ---- 3.1 eV. Fig. 3(b) gives the XPS
spectrum for this sample after being subjected to
Ar ion etching (20 min, 10 mA, 2 keV). The
obtained values of FWHM for the Ols and
V2p3/2 peaks are 2.8 and 4.2 eV respectively. The
temperature of the sample was raised from ambient
up to 200°C; in this condition the corresponding
spectrum is Fig. 3(c); it is found that A for the
V2p3/2 peak increases slightly up to 4.4 eV. The
sample temperature was kept at 200°C for 15 h
(Fig. 3(d)); the FWHM of V2p3/2 increases up to
4.6 eV. The sample was cooled to 35°C (Fig. 3(e));
in this case, the FWHMs of the Ols and V2p3/2
peaks were 2.6 and 4 eV respectively. We consider
spectrum 3(e) to be representative of surface VO2
(entry 9 in Table 1). Subsequently, the sample was
reoxidized in 02 at 300°C for 5 h; the correspond-
ing XPS spectrum is presented in Fig. 3(f) and the
determined full widths at half maximum for the
Ols and V2p3/2 are 1.8 and 1.6 eV, respectively.
We can observe, in the last spectrum, an increase
in the intensity of the shake-up satellite associated
with O ls. It can also be verified how readily the
surface of this oxide over-oxidizes, which has led
us to think that spectrum 3(a) does not correspond
to a surface state that is characteristic of VO2.
J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261 253
b,.
0
I - .
N
Z
w
F-
Z
//
.... c) .......................... "
__2bL-- - ~- - . _ J
Vz P3,'z
/i j, ....
• ....................
'""" " '" '" "": / ~:
\\._1-
5,50 540 L r 550 520 510
BINDING ENERGY(eV)
Fig. 2. XPS spectra of polycrystalline V6013. Ols and V2p3/2 peaks are shown after the sample has experienced a sequence of sample
treatments: (a) the sample was given a short annealing at 200°C in UHV; (b) after Ar ion etching; (c) sample maintained at 300°C for 13 h
followed by Ar ion etching (20 min); (d) sample was annealed in 02 at 300°C for 3 h.
3•4• 1/203
The XPS spectra of a polycrystalline V203
sample prepared as described in the experimental
section are presented in Fig. 4. In order to counter-
act the reoxidation effect of this sample (see Experi-
mental section), the sample temperature was kept
at 550°C for 14 h in UHV. The XPS spectrum for
this condition is shown in Fig. 4(a). The obtained
full widths at half maximum for the O 1 s and V2p3/2
peaks were 2 and 4.8 eV respectively. Spectrum 4(a)
is representative of surface V20 3 (entry 11 in
Table 1). Spectrum 4(b) corresponds to this sample
after being subjected to Ar ion etching (15 min, 10
mA, 2 keV); in this case the determined values of
FWHM for the above-mentioned peaks are 2.6 and
6 eV respectively•
From the former results and from Table 1, we
can observe that the FWHM for the V2p3/2 peak
increases as the degree of oxidation decreases, and
this is correlated to larger values of the binding
energy difference.
According to Tables 1 and 2 we can see that our
binding energy values agree quite well with those
reported by Sawatzki and Post [8], although their
values for the FWHM of V2p3/2 are larger than
ours, which could indicate a slightly reduced state
for their surface. The values given by Rao et al. [9]
would correspond to even more reduced states, and
their results for VO2 would correspond to a higher
254 J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249 261
E
0
>-
F-
Z
LiJ
I--
Z
hi OIs
II
II
I I
V2p312
,50 540 550 520 510
BINDING ENERGY (eV)
Fig. 3. Ols and V2p3/2 XPS region for polycrystalline VO 2 after a sequence of several treatments: (a) as-received condition; (b) after Ar
ion etching (20 min); (c) sample temperature raised from ambient to 200°C; (d) sample temperature kept at 200°C for 15 h; (e) sample
cooled to 35°C; (f) oxidation in 02 at 300°C for 5 h.
oxidation state; in fact once we have made the
energy corrections using their quoted reference,
we obtain 516.3 eV for the binding energy of the
V2p3/2 level, which actually corresponds to V2013.
The results reported by Werfel et al. [10] for
V205 are very close to ours, if the energy dif-
ference between the binding energies attributed
to the O ls level is taken into account. How-
ever, their values for the parameter A for V6013
and VO2 are larger that ours, which indicates
that their samples are very reduced at the surface.
The values reported by Anderson [1] for V205
correspond to reduced surfaces; the results for
V203, taking into account the width of the
V2p3/2 level and the corrected value for its binding
energy (516.7 eV), indicate an oxidized surface.
The results obtained by Larsson et al. [11]
and Blauw et al. [12] correspond to reduced
surfaces.
Our binding values for V203 are very close to
those of Rao et al. [9] who have used the same
sample preparation procedure as in this work.
The results of Werfel et al. [10] indicate a very
reduced surface (A = 17.2 eV) with a corrected
binding energy for the V2p3/2 level at 512.8 eV
which is very close to the binding energy of metallic
J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261 255
.,=_
c
b.
o
>-
Z
hl
I"-
Z
O|S
L I I I L
550 540 530 520 510
BINDING ENERGY(eV)
Fig. 4. Ols and V2p3/2 XPS spectral peaks ofV205 after several consecutive sample treatments: (a) sample temperature kept at 550°C
for 14 h in UHV; (b) after Ar ion bombardment (15 min).
vanadium reported by Sawatzki and Post [8]. These
authors reported values for V203 which would cor-
respond to oxidized surfaces (A = 14.4 eV) and a
corrected binding energy for V2p3/2 of 515.6 eV.
For the case of V205, in which vanadium exhib-
its the highest degree of oxidation, we can see that
the use of FWHM and the parameter A (or the
corrected binding energy for the V2p3/2 level),
allows us to differentiate easily between well-
oxidized and reduced states of the surface. How-
ever, to distinguish between V6013. VO2, V203,
and their reduced or oxidized surfaces, it is neces-
sary to study in detail the structure and width of the
peaks.
After making a detailed comparison of the XPS
spectra of V205, V6013, VO 2 and V203, after dif-
ferent treatments we can observe that the FWHM
values of the Ols and V2p3/2 peaks increase in the
above order. However, the broadening experienced
by the O 1 s peak is less that than of V2p3/2, which
can reach up to three or four times the correspond-
ing width in V205.
We should point out that in V205 oxygen
occupies three different sites whose Madelung
potentials are very different [13]; consequently,
according to the ionic model, three oxygen peaks
should be observed in the XPS spectrum. This is
not the case, and what is found experimentally is
that the width of the Ols peak in V205 is even
smaller than in V6013, VO2, and V203. This sug-
gests a more covalent model and the presence of
mechanisms in addition to those of Madelung.
256 J. Mendialdua et al./Journal of Electron Spectroscopy and Related Phenomena 71 (1995) 249-261
L ._
0
m
@
C
H
V2p3/2
\
01s
vep3/2
BINDING ENERGY (eV)
Fig. 5. Comparison between the XPS spectra of (a) V6013 after a short heating at 200°C in UHV; and (b) V205 at 500°C in UHV.
Vanadium occupies three different sites in V6013
with charges [2] 4.13(e), 4.65(e) and 4.36(e), and
seven different types of oxygen whose charges
range from 1.94(e) to 2.17(e), i.e. a range of
variation of 0.23(e) (in contrast in V205 where
the range is only 0.1(e)). Taking into account
these differences we should expect that the Ols
peak width in V203 is larger than in V205.
VO2 has two different types of oxygen, but the
differences in the V-O bond lengths are smaller
and their Madelung potentials differ only by 0.3
eV [2]. However, it is found experimentally that
the O ls peak is wider than the corresponding
peak in V205. We believe that the broadening of
the Ols peak is not due to the presence of contami-
nants (H20, CO, etc.)
本文档为【XPS studies of V2O5, V6O13, VO2 and V2O3】,请使用软件OFFICE或WPS软件打开。作品中的文字与图均可以修改和编辑,
图片更改请在作品中右键图片并更换,文字修改请直接点击文字进行修改,也可以新增和删除文档中的内容。
该文档来自用户分享,如有侵权行为请发邮件ishare@vip.sina.com联系网站客服,我们会及时删除。
[版权声明] 本站所有资料为用户分享产生,若发现您的权利被侵害,请联系客服邮件isharekefu@iask.cn,我们尽快处理。
本作品所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用。
网站提供的党政主题相关内容(国旗、国徽、党徽..)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。