Surface modification of ZrO2 nanopowder with oxovanadium species using slurry deposition and impregnation methods

Article abstract of DOI:10.1016/j.jallcom.2006.07.146

Structural differences in zirconia-supported vanadia, containing 3 mol% V₂O₅, obtained by slurry deposition and wet impregnation were elucidated using Raman, EPR and XRD methods, supported by Rietveld refinement. The preparation meth

Full text of DOI:10.1016/j.jallcom.2006.07.146

Journal of Alloys and Compounds 442 (2007) 302–305
Surface modification of ZrO2 nanopowder with oxovanadium
species using slurry deposition and impregnation methods
A. Adamskia,∗, P. Zapałaa, P. Jakubusb, Z. Sojkaa
a Faculty of Chemistry and Regional Laboratory of Physicochemical Analyses and Structural Research,
Jagiellonian University, Ingardena 3, 30-060 Krakow, Poland
b Institute of Chemistry and Environment Protection, Szczecin Technical University, Al. Piasto´w 42, 71-065 Szczecin, Poland
Received 17 June 2006; received in revised form 30 June 2006; accepted 14 July 2006
Available online 30 January 2007
Abstract
Structural differences in zirconia-supported vanadia, containing 3 mol% V2O5, obtained by slurry deposition and wet impregnation were eluci-
dated using Raman, EPR and XRD methods, supported by Rietveld refinement. The preparation method strongly determines both the heterogeneity
of the surface VOx species and incorporation of the V4+ ions into the zirconia matrix. Slurry deposition favors the formation of isolated or two-
dimensional oxovanadium clusters, whereas impregnation leads to agglomerated nanocrystalline V2O5. The stability of surface vanadium was
discussed in terms of its reducibility and changes in the temperature of the tetragonal to monoclinic ZrO2 phase transition, induced by surface
doping.
© 2007 Elsevier B.V. All rights reserved.
PACS: 61.66.−f; 81.05.−t; 81.16 Be; 81.07 Wx
Keywords: Oxide materials; Chemical synthesis; Slurry deposition; Impregnation; Vanadium ions speciation
1. Introduction
oxide to be deposited [3]. So far such procedure was used for
synthesis of supported MoO3 materials [4]. In this work, we
adapted this procedure for spreading V2O5 on the surface of the
tetragonal ZrO2 nanopowder.
Influence of the preparation method on the structure and
properties of composed nano-oxide materials is one of the most
important factors for tailoring them toward specific applications
[1]. For example, considering a two-component supported sys-
tem, where the major oxide plays the role of carrier, whereas the
minor one is a deposited functional phase (e.g., supported cata-
lysts, chemical oxide sensors), dispersion of the deposited oxide,
heterogeneity of surface entities and dissolution of the addi-
tive ions in the bulk, are strongly determined by the preparation
method and the treatment conditions as well.
Classic impregnation is one of the most popular methods of
deposition of one oxide onto another. This term denotes a pro-
cedure, where an excessive volume of the precursor solution
is contacted the carrier [2]. Slurry deposition is a more recent
preparation procedure, where practically insoluble oxide, func-
tioning as a support, is mixed with the slurry of a poorly soluble
2. Experimental
Single-phase tetragonal ZrO2 support (t-ZrO2) of enhanced thermal stability
was obtained by modified precipitation method from aqueous solutions of
0.6 M ZrOCl2·8H2O (Aldrich 99.99%) with 25% aqueous solution of ammonia
[5]. VOx/ZrO2 samples, containing nominally 3 mol% of V2O5, were prepared
by impregnation of the t-ZrO2 matrix with an aqueous solution of NH4VO3
(99%, Merck) and by deposition from a slurry consisting of 0.1455 g of V2O5
(99.6%, Merck) in 5 cm3 of water at 353–373 K. These samples are denoted
as imp-VOx/ZrO2 and sdp-VOx/ZrO2, respectively. All samples were dried at
373 K for 12 h. The impregnated samples were calcined in air at 873 K for 6 h.
The X-ray diffraction patterns were recorded by means of a DRON-3
(Bourevestnik, Russia) and a Philips PW 1820 diffractometers, equipped with
Fe and Ni filters, respectively, using Co K␣ and Cu K␣ radiations. For Rietveld
refinement (using the DBWS-9411 program [6]), the 2Θ scans were carried out
in 0.02◦ steps. High-resolution transmission electron microscopy (HR-TEM)
was carried out using a JEM-100CX II UHR instrument (JEOL) operating at
100 kV. Raman spectra were collected on a FTS 6000 spectrometer equipped
with a BIORAD accessory. The samples were excited with the 1064 nm line of a
diode-pumped Nd:YAG laser (Spectra Physics Model T108S) and the scattered
∗
Corresponding author. Tel.: +48 12 663 22 24; fax: +48 12 634 05 15.
E-mail address: adamski@chemia.uj.edu.pl (A. Adamski).
0925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2006.07.146

A. Adamski et al. / Journal of Alloys and Compounds 442 (2007) 302–305
303
radiation was collected at 180◦ with 4 cm−1 resolution. The CW-EPR X-band
spectra were recorded at 77 K with a Bruker ELEXYS E-500 spectrometer, oper-
ating at the 100 kHz field modulation. The spectra simulation was performed
using the EPRsim32 program [7]. The samples were heated ex situ, with the rate
of 4 K/min, from room temperature to 1273 K in 100 K steps, and then kept for
0.5 h at each temperature, before measurement of the EPR spectra.
nating both spectra, also weak bands from m-ZrO2 at 191 cm−1
(Ag), 375 cm−1 (Bg) and cm−1 476 (Ag) can be found in the
Raman spectrum of imp-VOx/ZrO2 (Fig. 1a). The observed par-
tial destabilization of the metastable t-ZrO2 is thus, apparently
induced by the interaction with V2O5, produced on the support
surface upon the calcination [13].
The main bands of oxovanadium species present on the zirco-
nia surface can be clearly observed in the Raman spectra above
700 cm−1. Strong bands at 777 cm−1 and 991 cm−1, dominating
the spectrum of imp-VOx/ZrO2 (Fig. 1a) and a weak band at ca.
990 cm−1 in the spectrum of sdp-VOx/ZrO2 (Fig. 1b), reflect
significant differences in the vanadium dispersion in both
samples. The band at 777 cm−1 can be ascribed to νV–O–V mode
(B3u) in three-dimensional V2O5 nanocrystallites [14]. The band
around 990 cm−1 is characteristic of a symmetrical νV O mode
in a variety of the octahedral oxovanadium species [15,16]. In
case of the imp-VOx/ZrO2 sample, only the presence of a well
crystalline nano-V2O5 entities can explain the presence of such
sharp and intense band. Much weaker and broader feature at
990 cm−1 in the spectrum of sdp-VOx/ZrO2 reflects heterogene-
ity of the terminal V O groups, typical of the two-dimensional
oxovanadium surface clusters of relatively low crystallinity.
The differences in aggregation of surface vanadium species
can also be revealed by EPR, using the V4+ (3d1) ions, appearing
in the sdp-VOx/ZrO2 and imp-VOx/ZrO2 samples, as a paramag-
netic probe. The EPR spectra of both samples can be analyzed
in terms of the supposition of two signals: a broad structure-
less one, due to magnetically interacting clustered species, and
a narrow signal with the resolved hyperfine structure of iso-
lated V4+ centers. From the computer simulation of the spectra
recorded for the fresh samples (Fig. 2), the relative fraction of
the isolated surface vanadium was found to be 28% for imp-
VOx/ZrO2 and 48% for sdp-VOx/ZrO2, whereas the fraction of
the clustered oxovanadium species (including V2O5) was equal
to 63 and 19%, respectively, indicating much better dispersion
of vanadium in the sdp-VOx/ZrO2 sample.
3. Results and discussion
The XRD patterns recorded in the range of 10–70◦ for the
sdp-VOx/ZrO2 and the imp-VOx/ZrO2 dry samples are typical
of t-ZrO2 polymorph. An intense maximum centered at 35◦ can
be ascribed to (1 1 1) reflection of the tetragonal phase, whereas
weak maxima at 41◦ and 59◦ to (2 0 0)/(0 0 2) and (0 2 2)/(2 2 0)
reflections of t-ZrO2, respectively [8,9]. No traces of the mon-
¯
oclinic zirconia (m-ZrO2) with the diagnostic (1 1 1) and (1 1 1)
reflections, could be observed. The average size of particles,
assessed from the Scherrer analysis of the XRD line widths, was
equal to 6.360 0.024 nm. This remains in accordance with the
size of tetragonal nanodomains, determined in parallel HR-TEM
studies [5].
The absence of any reflections from V2O5 nanocrystals
implies the lack of an extended aggregation of the deposited
vanadia. Heating of the samples till 873 K did not lead to any
distinct changes in the diffractograms, confirming relatively
good thermal stability of t-ZrO2 in the presence of 3 mol% of
vanadia. However, Raman spectra (Fig. 1) exhibited distinct dif-
ference betweenthe imp-VOx/ZrO2 and sdp-VOx/ZrO2 samples,
in the range below 700 cm−1, where the lattice vibrations of
the support appear [10,11]. The factor group theoretical anal-
ysis predicts six Raman active modes (A1g + 2B1g + 3Eg) for
the tetragonal form and 18 (9Ag + 9Bg) active modes for the
monoclinic ZrO2 [12]. Careful inspection of the deconvoluted
diagnostic fragments of both spectra, revealed a small contribu-
tion of m-ZrO2 to the imp-VOx/ZrO2 sample. Apart from intense
bands characteristic of t-ZrO2 at 148 cm−1 (B1g), 266 cm−1
(Eg), 316 cm−1 (B1g), 450 cm−1 (Eg), 645 cm−1 (Eg), domi-
The parameters of the broad (ꢀBpp ≈ 17 mT) EPR sig-
nal of the surface oxovanadium clusters (g = 1.952 and
||
g = 1.979 for imp-VOx/ZrO2 and g = 1.950 and g = 1.973
⊥
||
⊥
for sdp-VOx/ZrO2) were almost insensitive to the prepa-
ration method. They remain in a good agreement with the
parameters found previously for partially reduced vanadia
[17]. The species containing magnetically isolated V4+ give
rise to an axial EPR signal with the well-resolved eight-line
hyperfine structure (due to 51V with I = 7/2 and 100% natural
abundance). The magnetic parameters, g = 1.930, g = 1.979
||
⊥
and |A | = 18.56 mT, |A | = 6.12 mT, for imp-VOx/ZrO2 and
||
⊥
g = 1.925, g = 1.977 and |A | = 17.96 mT, |A | = 6.45 mT for
||
⊥
||
⊥
sdp-VOx/ZrO2 were found to be distinctly different, depending
on the sample preparation method. Such signals, with g > g
⊥
||
and A > A , are characteristic of isolated vanadyl ions in a
||
⊥
square pyramidal or tetragonally distorted octahedral coordina-
tion with approximate C4v symmetry. The constituting ligands
can be formed by the oxygens of the ZrO2 matrix, adsorbed
water molecules or hydroxyls bonded to vanadium [18]. The
distortion of the surface vanadyl complex can be gauged by the
Fig. 1. Normalized Raman spectra of as prepared: (a) imp-VOx/ZrO2 and (b)
sdp-VOx/ZrO2 samples, recorded at ambient conditions.
factor B = (g − 2.0023)/(g − 2.0023) [18]. Substitution of the
||
⊥

304
A. Adamski et al. / Journal of Alloys and Compounds 442 (2007) 302–305
Fig. 2. X-band EPR spectra of: (a) imp-VOx/ZrO2 and (b) sdp-VOx/ZrO2 samples, recorded at 77 K before and after calcination at progressively increasing
temperatures.
numerical values gives B = 3.10 for imp-VOx/ZrO2 and B = 3.05
for sdp-VOx/ZrO2, indicating a shortening of the terminal V O
bond and simultaneous weakening of the oxygen ligands in the
equatorial plane for both samples, in comparison to those in
V2O5 single crystals (B = 4.9) [19].
The initial disappearance (sdp-VOx/ZrO2) or weakening
(imp-VOx/ZrO2) of the EPR signal due to the surface V4+ ions,
in the course of the calcination in air, reflects their progressive
oxidation to diamagnetic V5+ (3d0). Subsequent spontaneous
reduction, necessary for incorporation of the vanadium into the
bulk, was strongly related to the nature of oxovanadium surface
species. The agglomerated species, prevailing in the case of
imp-VOx/ZrO2, are more easily reduced than the isolated
oxovanadium centers [20]. As a result, the evolution of the EPR
spectra upon the heat treatment was much more dramatic than
in the case of sdp-VOx/ZrO2. Thus, reducibility of the surface
vanadium appears to be a key factor, controlling penetration
of the vanadium ions into the zirconia matrix. Furthermore,
there is a remarkable synergetic effect between the threshold
temperature of V4+ migration towards the bulk and the phase
transition of the tetragonal to monoclinic zirconia, in the case of
the imp-VOx/ZrO2 samples. Such phase transition, was found
not only to be accelerated in the presence of surface vanadium
by ca. 300 K, but it also makes the V4+ migration easier. In
case of the sdp-VOx/ZrO2 sample, because the well-dispersed
vanadium ions are hardly reducible below 1073 K (Fig. 2b),
their diffusion toward the bulk of the support was pronouncedly
retarded. Interestingly, upon reaching the temperature of 1173 K
it becomes quite rapid.
The sequence of the EPR spectra, recorded in the course
of the variable temperature calcination is shown in Fig. 2.
The axial signal, arising from the isolated V4+ surface centers,
diminished with the increasing temperature and simultane-
ously a new orthorhombic signal with gzz = 1.8723, gxx = 1.9759,
gyy = 1.9370, |Azz| = 16.30 mT, |Axx| = 5.59 mT, |Ayy| = 1.52 mT,
characteristic of V4+ isolated in the ZrO2, gradually devel-
oped [13]. The threshold temperature, Tth, at which this
signal emerged was distinctly different for imp-VOx/ZrO2
(Tth = 873 K) and sdp-VOx/ZrO2 (Tth = 1073 K) samples. It cor-
responds to dissolution of the isovalent V4+ ions in the ZrO2
matrix, leading to formation of a substitutional VxZr1−xO2
solid solution. This process is accompanied by a shrinkage
3
˚
of the m-ZrO2 unit cell volume from 140.899 0.003 A to
3
4+
˚
140.873 0.005 A , caused by the 13% difference in the V
and Zr4+ ionic radius, as inferred from Rietveld analysis of
the XRD data. Changes in the EPR spectra observed during
annealing of imp-VOx/ZrO2, suggest that formation of the solid
solution (T > 873 K) precedes the phase transition from t-ZrO2
to m-ZrO2, taking place at T > 973 K. Consequently, the V4+
ions are dissolved in the tetragonal matrix giving rise to an
axial signal, whereas for sdp-VOx/ZrO2, the opposite sequence
occurs. The V4+ ions are dissolved in the monoclinic matrix at
T > 1073 K(i.e., afterthephasetransitiontakesplace), producing
an orthorhombic EPR signal.
4. Conclusions
The way of V2O5 deposition onto tetragonal ZrO2 support
strongly influences the structure of surface oxovanadium
species, their dispersion and dimensionality. Isolated and

A. Adamski et al. / Journal of Alloys and Compounds 442 (2007) 302–305
305
clustered species were detected in the samples obtained by
slurry deposition, whereas agglomerated oxovanadium and
nanocrystalline V2O5, dominated on the surface of samples
prepared by slurry deposition. The dispersion state of surface
vanadium strongly influences its reducibility and the threshold
temperature of incorporation of the resultant V4+ ions into the
bulk of the ZrO2 matrix.
[6] R.A. Young, A.C. Larson, C.O. Paiva-Santos, User’s Guide to Program
DBWS-9807a for Rietveld Analysis of X-ray and Neutron Powder Diffrac-
tion, Schools of Physics, Georgia, Institute of Technology, Atlanta, GA,
U.S.A. (2000).
[7] T. Spalek, P. Pietrzyk, Z. Sojka, J. Chem. Inf. Model. 45 (2005) 18–29.
[8] K.T. Jung, A.T. Bell, J. Mol. Catal. A 163 (2000) 27–42.
[9] G. Katz, J. Am. Ceram. Soc. (1971) 531.
[10] T. Hirata, J. Phys. Chem. Solids 56 (1995) 951–957.
[11] K. Matsui, H. Suzuki, M. Ohgai, H. Arashi, J. Ceram. Soc. Jpn. Int. Ed. 98
(1990) 15–19.
[12] T. Hirata, E. Asari, M. Kitajima, J. Solid State Chem. 110 (1994) 201–
207.
Acknowledgement
This work was financially supported by The Polish State
Committee for Scientific Research (KBN) within the project
3 T09 147 26.
[13] A. Adamski, Z. Sojka, K. Dyrek, M. Che, Solid State Ionics 117 (1999)
113–122.
[14] P. Clauws, J. Broeckx, J. Vennik, Phys. Stat. Sol (b) 131 (1985) 459–473.
[15] M. Schraml, W. Fluhr, A. Wokaun, A. Baiker, Ber. Bunsenges. Phys. Chem.
93 (1989) 852–857.
References
[16] T.R. Gilson, O.F. Bizri, N. Cheetham, J. Chem. Soc., Dalton Trans. (1973)
291–294.
[1] H.H. Kung, E.I. Ko, Chem. Eng. J. 64 (1996) 203–214.
[2] J.A. Schwarz, C. Contescu, A. Contescu, Chem. Rev. 95 (1995) 477–510.
[3] E. Hillerova´, M. Zdrazˇil, Appl. Catal. A 138 (1996) 13–26.
[4] T. Klicpera, M. Zdrazˇil, Catal. Lett. 58 (1999) 47–51.
[5] A. Adamski, P. Jakubus, Z. Sojka, Nukleonika 51 (Suppl. 1) (2006)
S27–S33.
[17] Y. Kera, Y. Matsukaze, J. Phys. Chem. 90 (1986) 5752–5755.
[18] V.K. Sharma, A. Wokaun, A. Baiker, J. Phys. Chem. 90 (1986) 2715–2718.
[19] A. Adamski, Z. Sojka, K. Dyrek, M. Che, G. Wendt, S. Albrecht, Langmuir
15 (1999) 5733–5741.
[20] S. Albrecht, G. Wendt, G. Lippold, A. Adamski, K. Dyrek, Solid State
Ionics 101–103 (1997) 909–914.