Chemical and structural characterization of V 2 O 5 /TiO 2 catalysts C. B. Rodella, P. A. P. Nascente, V. R. Mastelaro, M. R. Zucchi, R. W. A. Franco, C. J. Magon, P. Donoso, and A. O. Florentino Citation: Journal of Vacuum Science & Technology A 19, 1158 (2001); doi: 10.1116/1.1380720 View online: http://dx.doi.org/10.1116/1.1380720 View Table of Contents: http://scitation.aip.org/content/avs/journal/jvsta/19/4?ver=pdfcov Published by the AVS: Science & Technology of Materials, Interfaces, and Processing Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 http://scitation.aip.org/content/avs/journal/jvsta?ver=pdfcov http://oasc12039.247realmedia.com/RealMedia/ads/click_lx.ads/test.int.aip.org/adtest/L23/1291604448/x01/AIP/Hiden_JVACovAd_1640x440Banner_12_10and12_17_2013/1640x440_-_23874-BANNER-AD-1640-x-440px_-_USA.jpg/7744715775302b784f4d774142526b39?x http://scitation.aip.org/search?value1=C.+B.+Rodella&option1=author http://scitation.aip.org/search?value1=P.+A.+P.+Nascente&option1=author http://scitation.aip.org/search?value1=V.+R.+Mastelaro&option1=author http://scitation.aip.org/search?value1=M.+R.+Zucchi&option1=author http://scitation.aip.org/search?value1=R.+W.+A.+Franco&option1=author http://scitation.aip.org/search?value1=C.+J.+Magon&option1=author http://scitation.aip.org/search?value1=P.+Donoso&option1=author http://scitation.aip.org/search?value1=A.+O.+Florentino&option1=author http://scitation.aip.org/content/avs/journal/jvsta?ver=pdfcov http://dx.doi.org/10.1116/1.1380720 http://scitation.aip.org/content/avs/journal/jvsta/19/4?ver=pdfcov http://scitation.aip.org/content/avs?ver=pdfcov Redist Chemical and structural characterization of V 2O5 ÕTiO2 catalysts C. B. Rodella Departamento de Fı´sica e Ciência dos Materiais, Instituto de Fı´sica de Sa˜o Carlos, Universidade de Sa˜o Paulo, 13560-970 Sa˜o Carlos, SP, Brazil P. A. P. Nascentea) Centro de Caracterizac¸ão e Desenvolvimento de Materiais, Departamento de Engenharia de Materiais, Universidade Federal de Sa˜o Carlos, 13565-905 Sa˜o Carlos, SP, Brazil V. R. Mastelaro, M. R. Zucchi, R. W. A. Franco, C. J. Magon, and P. Donoso Departamento de Fı´sica e Ciência dos Materiais, Instituto de Fı´sica de Sa˜o Carlos, Universidade de Sa˜o Paulo, 13560-970 Sa˜o Carlos, SP, Brazil A. O. Florentino Departamento de Quı´mica, Instituto de Biocieˆncias, Universidade Estadual Paulista, 18618-000 Botucatu, SP, Brazil ~Received 29 September 2000; accepted 2 May 2001! A series of V2O5/TiO2 samples was synthesized by sol–gel and impregnation methods with different contents of vanadia. These samples were characterized by x-ray diffraction~XRD!, Raman spectroscopy, x-ray photoelectron spectroscopy~XPS!, and electronic paramagnetic resonance ~EPR!. XRD detected rutile as the predominant phase for pure TiO2 prepared by the sol–gel method. The structure changed to anatase when the vanadia loading was increased. Also, anatase was the predominant phase for samples obtained by the impregnation method. Raman measurements identified two species of surface vanadium: monomeric vanadyl (V41) and polymeric vanadates (V51). XPS results indicated that Ti ions were in octahedral position surrounded by oxygen ions. The V/Ti atomic ratios showed that V ions were highly dispersed on the vanadia/titania surface obtained by the sol–gel method. EPR analysis detected three V41 ion types: two of them were located in axially symmetric sites substituting for Ti41 ions in the rutile structure, and the third one was characterized by magnetically interacting V41 ions in the form of pairs or clusters. A partial oxidation of V41 to V51 was evident from EPR analysis for materials with higher concentrations of vanadium. ©2001 American Vacuum Society.@DOI: 10.1116/1.1380720# w ne p he an su e th o f ec on a e u c- en by ace rmal to ra- the ive n been the uble the he he con- the he ative d a l ys- the I. INTRODUCTION Vanadia supported on titania constitutes a well-kno catalytic system for selective oxidation reactions of o-xyle ammoxidation of hydrocarbon, and reduction of NOx with NH3. 1–3 The outstanding catalytic behavior of vanadia su ported on titania in comparison with that supported on ot oxides, such as SiO2 or Al2O3, is attributed to the strong support-active-phase interaction. Vanadia supported on tase presents higher activity and selectivity than vanadia ported on rutile. Vejux and Courtine4 proposed that a clos epitaxial crystallographic match between the structure of ~010! plane of V2O5 and the~001! plane of anatase leads t the spreading and preferential exposure of the~010! crystal- line active plane. Liettiet al.5 proposed that the formation o a stable VOx monolayer over anatase and the similar el tronegativities of titanium and vanadium as the main reas for apparent superior catalytic performance. There has been a lot of controversy about the nature functionality of the active sites of these catalysts. Howev we believe that the reaction mechanisms involving the s face V41/V51 redox pair contribute to the activity and sele tivity of the V2O5/TiO2 system. This hypothesis has be increasingly adopted in the literature.6–10 a!Electronic mail: nascente@power.ufscar.br 1158 J. Vac. Sci. Technol. A 19 „4…, Jul ÕAug 2001 0734-2101 Õ200 ribution subject to AVS license or copyright; see http://scitation.aip.org/term n , - r a- p- e - s nd r, r- The great disadvantage of the titania support prepared conventional ceramic processes is its low specific surf area. In addition, the anatase phase presents poor the stability at high temperature. Its partial transformation in rutile is thermodynamically favored but leads to a deterio tion of catalytic performance. Then, the characteristics of support are fundamental for the stabilization of the act phase in the V2O5/TiO2 catalysts and depend mainly o preparation methods. Several synthesis methods have employed, and of special importance are those in which active phase is dispersed by aqueous impregnation of sol salts or by nonaqueous impregnation of alkoxides into support.11–13 The main objective of these methods is t maximization of the catalyst effective region formed by t reagent/active-phase/support interface.14 However, in these impregnation methods, the dispersion, the active-phase centration, and the textural and structural properties of catalyst are limited by the characteristics of t support.13,15–17 The sol–gel method has been proposed as an altern route to synthesize catalysts with a high surface area an stable active phase.17–23This method permits a better contro of the textural and structural properties of the catalytic s tems, and an improved dispersion of the active phase on support.22,23 11581Õ19„4…Õ1158Õ6Õ$18.00 ©2001 American Vacuum Society sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 e b ce th n ur u a if c o- g by n, e in th 2 a a 0B io b te e ti 4 r. ith h a rg h b as ac ct K ncy he with by a- to ow- re- ra- e in oted the n- he tion, ease 1159 Rodella et al. : Chemical and structural characterization 1159 Redist In this work, we present the results of V2O5/TiO2 cata- lysts prepared by a modified sol–gel method. The purpos this investigation is to develop a preparation method capa of maximizing the vanadia dispersion on the titania surfa The conditions that favor the formation of rutile TiO2 are developed to show the influence of the active phase on V2O5/TiO2 structural and textural properties. We demo strate that the active phase permits the control of the text and structural properties, promoting materials with high s face area and enabling a mixed anatase/rutile support low calcination temperature. We have employed x-ray d fractometry ~XRD!, Raman spectroscopy, x-ray photoele tron spectroscopy~XPS!, and electron paramagnetic res nance~EPR! to characterize the V2O5/TiO2 catalysts. II. EXPERIMENTAL PROCEDURES TiO2/V2O5 samples with different contents of V2O5 ~1, 3, 6, and 9 wt %! were synthesized using sol–gel and impre nation methods. The vanadium precursor, NH4VO3, was dis- solved in nitric acid to controlpH. Solution A was prepared by dissolving NH4VO3 in nitric acid (pH51) under 75 W ultrasonic vibrations for 1 min. Solution B was prepared diluting tetraisopropyl-orthotitanate (C12H28O4TiO) in iso- propyl alcohol at a molar ratio of 0.25. After solubilizatio solution A was added to solution B and a gel was form immediately. The gel was then vigorously stirred for 5 m under 75 W ultrasonic vibrations, and after a 24 h rest resulting gel was dried at 110 °C for 4 h, and calcined at 7 K for 16 h. Catalysts containing 6 and 9 wt % of V2O5 were prepared by impregnation of P-25 Degussa TiO2 ~50 m2/g! in solution A. The solid was gradually dried and calcined in manner which was similar to the previous method. X-ray powder diffraction patterns were obtained using automatic Rigaku Rotaflex diffractometer model RU 20 with Cu Ka radiation~40 kV/40 mA, 1.5405 Å! and a nickel monochromator filter. Raman spectra were obtained with 514.5 nm argon laser line, Spectra Physics 2020, a Jobin-Yvon U1000 dou monochromator with holographic gratings, and a compu controlled photon-counting system. The Raman spectra w obtained with an average of 100 scans, a spectral resolu of 4 cm21 at room temperature, and a power excitation of mW. XPS analysis was performed in ultrahigh vacuum~low 1027 Pa range! using a KRATOS XSAM HS spectromete Mg Ka (hn51253.6 eV) was used as the x-ray source, w a power determined by emission at 15 mA and 15 kV. T high-resolution spectra were obtained with an analyzer p energy of 20 eV. The samples were flooded with low-ene electrons from a flood gun to avoid charging effects. T binding energies were referenced to an adventitious car 1s line set at 284.8 eV. The Shirley background and a le square routine were used for peak fitting. The sensitivity f tors for quantitative analysis were referenced toSF 1s51.0. EPR experiments were carried out using bothX and Q bands on powdered samples. All EPR absorption spe were recorded as the first derivative of absorption at 77 JVST A - Vacuum, Surfaces, and Films ribution subject to AVS license or copyright; see http://scitation.aip.org/term of le . e - al r- t a - - - d e 3 n n le r- re on 5 e ss y e on t- - ra . The Q band spectra were taken with anE-line Varian Spec- trometer operating at 34 GHz. TheX-band~9.6 GHz! spectra were measured using a magnetic-field modulation freque of 85 kHz. Calibration ofg values was based on theg 51.9797 signal of a MgO:Cr31 marker. III. RESULTS AND DISCUSSION A. X-ray diffraction Figure 1 shows the typical XRD peaks attributed to t anatase, rutile, and brookite phases. Peaks associated crystalline V2O5 are not observed for all samples prepared the sol–gel method, indicating the efficient dispersion of v nadium in the titania matrix. Lietti et al.5 showed that the transformation of anatase rutile is only considered an efficient process for TiO2 samples calcined at temperatures higher than 973 K. H ever, the XRD results depicted in Fig. 1 indicate the p dominance of rutile for titania calcined at a lower tempe ture ~723 K!. This may be the consequence of a decreas the phase transition temperature of anatase to rutile prom by the above preparation method. In acidic preparations, hydrolysis of the titanium alkoxide is faster than the conde sation, producing samples with many hydroxyl groups. T thermal treatment of these samples causes dehydroxyla and forms Vac1O22 anionic vacancies.24 The high defect concentrations of anatase crystals contribute to the decr in the anatase to rutile phase transformation.24 FIG. 1. Evolution of the XRD patterns of V2O5 /TiO2 xerogel with the va- nadia concentration. sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 o ttr in h ic n as uc fo y n a y yl rr iO ay d a- in d r- ne t i- he ses, - ad se a- ux py na- on/ gna- ates The t ds t %, sup- –gel ples ma- r d 1160 Rodella et al. : Chemical and structural characterization 1160 Redist A substantial increase in the anatase peak intensity is served as the vanadium loading increases. This can be a uted to the strong vanadium–titanium interaction which hibits the phase transformation of anatase to rutile. T vanadium atoms can occupy the anionic vacancies, wh were formed in the preparation method, and this occupa contributes to the stabilization of titania in the anatase ph Similar effects are related to the presence of impurities, s as sulfate and phosphate,5 or attributed to a chemical solid interaction between vanadia and anatase.25 In short, the structural nature of V2O5/TiO2 is determined by vanadia–titania interactions, which are more evident the V2O5/TiO2 system prepared from liquid precursors.3,26 In the case of the V2O5/TiO2 catalytic system prepared b the impregnation method, all XRD spectra show the prese of both anatase and rutile phases, as are expected from original structure of the titania support. Three very we peaks of crystalline V2O5 are present in the spectra. Man authors15,27–29reported the formation of monomeric vanad and polymeric vanadate species for vanadia loadings co sponding to less than the monolayer capacity of the T2 support. Increasing the vanadia loading above the monol capacity of the TiO2 support, crystallites of V2O5 are formed. For the P25 support, the monolayer capacity is estimate be 4.0 wt % V2O5, 14 but for our samples the monolayer c pacity is higher. B. Raman spectroscopy Figure 2 displays Raman spectra of the catalysts show a well-defined absorption band at 1030 cm21 and two broad bands at 900–960 and 770–850 cm21. The band at 1030 cm21 is attributed to monomeric vanadyls species bound rectly to the TiO2 support.30–34The number of bonds ancho ing the vanadyl group to the support cannot be determi from the Raman spectra. This assignment is based on FIG. 2. Raman spectra of the V2O5 /TiO2 xerogel dehydrated at 473 K unde 1023 Torr pressure. J. Vac. Sci. Technol. A, Vol. 19, No. 4, Jul ÕAug 2001 ribution subject to AVS license or copyright; see http://scitation.aip.org/term b- ib- - e h cy e. h r ce the k e- er to g i- d he similarity in the position of these bands and those for term nal VvO bonds of polyvanadate anions in solution. As t number of vanadium centers in the polyvanadates increa the number of terminal VvO groups per vanadium de creases to accommodateV–O–V linkages. The broad band in the region from 900 to 960 cm21, and centered at 930 cm21, is assigned to the terminal and internal VvO stretches,n(VvO), of polyvanadate groups, and the bro band centered at 822 cm21 is attributed to then~V–O–V! vibration of polymeric vanadates.34 Our Raman spectra do not exhibit the characteristic bands of crystalline V2O5 at 997 and 703 cm21.34 These results are in agreement with tho obtained by XRD and indicate an efficient dispersion of v nadium into the titania matrix. According to comparative studies carried out by Vej and Courtine4 using Fourier transform infrared spectrosco and Raman scattering, the vibrational frequencies of va date species do not depend strongly on hydrati dehydration treatments. The Raman spectra of samples prepared by the impre tion method are shown in Fig. 3. The 6 wt % V2O5 sample spectrum presents the bands of vanadyls and vanad groups, but for the 9 wt % sample these bands disappear. sharp and intense band at 998 cm21 and the broad band a 703 cm21, corresponding to crystalline V2O5, are clearly vis- ible for both samples. The intensities of these two ban increase as the vanadia content increases from 6 to 9 w indicating a decrease in the vanadium dispersion on the port surface. Contrary to the samples prepared by the sol method, the increase in vanadium content in the sam prepared by impregnation causes a partial oxidation of V41 to V51 surface ions. C. XPS analysis The XPS binding energies of the main peaks are sum rized in Table I. The O 1s satellite and the V 2p3/2 peaks FIG. 3. Raman spectra of V2O5 /TiO2 prepared by the impregnation metho dehydrated at 473 K under 1023 Torr pressure. sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 1161 Rodella et al. : Chemical and structural characterization 1161 JVST A - Vacuum, Redistribution subject to AV TABLE I. XPS characterization of the V2O5 /TiO2 samples. V2O5 ~wt %! Binding energies ~eV! O 1s Ti 2p3/2 V 2p3/2 Surface atomic ratio ~V/Ti ! 6 sol–gel 530.7~78! 459.3 517.6 0.13 532.2~22! 9 sol–gel 530.5~88! 459.0 517.5 0.19 532.1~12! 6 impregnation 530.7~84! 459.4 517.1 0.18 532.0~16! 9 impregnation 530.3~79! 459.2 516.9 0.20 531.7~21! w 7 pe - d he , t in l– o ent of by mo- the ob- e f at y re, overlap, making it difficult to analyze the samples of lo vanadia concentration~1 and 3 wt %!. Two O 1s compo- nents, resolved by curve fitting, can be distinguished:~a! one between 530.3 and 530.7 eV and~b! the other between 531. and 532.2 eV. The number in parenthesis is the atomic centage of each contribution to the O 1s peak. These binding energies refer to Ti41– O and OH2, respectively, in agree ment with results obtained by Carleyet al.35 and Pouilleau et al.36 The binding energy values for Ti 2p3/2 correspond to Ti41 in an octahedral symmetry.37 The presence of adsorbe hydroxyl species~OH! can be due to water adsorbed on t surface of the catalysts. The V 2p3/2 binding energy values are similar for all samples, corresponding to V51 ions.24 However, an overlap can occur for V51 and V41 peaks. For the samples prepared by the impregnation method V/Ti ratios increase slightly as the percent of vanadium creases. In the case of the samples prepared by the so method, the V/Ti ratios increase with vanadium content Surfaces, and Films S license or copyright; see http://scitation.aip.org/term r- he - gel n the support is more pronounced, indicating an enhancem in the dispersion of the surface vanadium. The formation crystalline V2O5 was not detected in the sol–gel samples either XRD ~Fig. 1! or Raman spectroscopy~Fig. 2!. Thus, vanadium remained dispersed on the titania surface as nomeric vanadys groups and polymeric vanadates due to high surface area, which is characteristic of the solids tained by the sol–gel method.17–23 D. EPR analysis Figure 4 shows theQ-band EPR absorption derivativ spectra of the 1 wt % V2O5/TiO2 catalysts at 77 K. This EPR spectrum is well resolved and indicates the presence o least three families of V41 ions: two sets characterized b structured EPR signals~species A and B in Fig. 4! superim- posed on a broad signal centered atg'1.93 ~signal C!. Spe- cies A and B, which present a resolved hyperfine structu FIG. 4. EPRQ-band spectra of the 1 wt % V2O5 xerogel catalyst obtained at 77 K. The MgO:Cr31 (g51.9797) was used as a reference signal.~a! Ex- perimental spectra and~b! Simulated spectra: ~b1! hyperfine interactions and ~b2! dipolar interactions. sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 h te o g il s ia n fe m d th e le gh ta wo by ith is tio of en l o - e a- o n by o D ia. ace, ten- nt, we of o- ties ium tion the iO tal- pe- ric e of od but . At ng and y ies ., em. r. , B tal. n, 1162 Rodella et al. : Chemical and structural characterization 1162 Redist are related to magnetically isolated vanadium ions. T broad and structureless signal C is due to magnetically in acting V41 centers probably consisting of clusters or pairs vanadium ions close enough to cause dipolar broadenin the EPR line and smearing the51V hyperfine structure. Simi- lar spectra were previously observed for V2O5/TiO2 cata- lysts for both the anatase and the rutile phases.1,3,22,25,38–41 The experimental V41 EPR spectra of the V2O5/TiO2 catalysts were analyzed by numerically solving the Ham tonian spin using Lorentzian line shapes. The best fitting the experimental spectra were achieved for the Hamilton spin parameters summarized in Table II. As can be see Fig. 4, the simulated spectra closely reproduce the main tures, including the position and the intensities of the pro nent lines. The analysis of the EPR parameters for signals A an shows that species A differs from species B mainly in value of the51V hyperfine constants~Ai andA'!, which are definitely higher for species A. The fact thatgi.g' andAi .A' for both species suggests that V41 ions associated with signals A and B are located in sites with octahedral symm try, substituting the Ti41 in the titania matrix of the rutile structure.22,39–43This may be due to the fact that the ruti phase of TiO2 is tetragonal, with two nonequivalent Ti41 ions per unit cell, where each ion is surrounded by six nei boring oxygen atoms, giving rise to an orthorhombic crys field at the titanium ion position. Therefore, there are t nonequivalent Ti41 ions whose positions can be occupied substitutional V41 ions.41 The spectra of the catalysts with lower loading~1–3 wt % V2O5, not shown in this work! follow the general trends observed in Fig. 4. The EPR linewidths of the sample w the higher V2O5 concentration are significantly broader. Th change in EPR line shape might be due to the interac between V41 ions resulting from increasing concentration vanadia. The signals associated with the parallel compon of the isolated vanadium~A and B! can be observed in al samples, but in theX band the distinction between these tw signals is not as clear as in theQ band. Signal C, correspond ing to vanadium pairs, which according to th literature1,3,22,25,38–41can be located in the rutile or the an tase phase, giving rise to EPR lines with the sameg factor.27 A detailed analysis enables us to extract the linewidth signal C as 200 G for the catalysts with lower loading a 250 G for the 9 wt % V2O5 catalysts. This result is consistent with the interpretation given XRD, Raman spectroscopy, and XPS for the catalysts tained by the sol–gel and impregnation methods. XR shows that the TiO2 rutile structure~which contributes to the TABLE II. Hamiltonian spin parameters for the V41 ions obtained by com- puter simulation of the spectra. Signal gi g' Ai(G) A'(G) DHpp(G) A 1.956 1.901 152.37 39.25 12.5 B 1.953 1.905–1.910 141.30 25–39 7.1 C 1.926 200 J. Vac. Sci. Technol. A, Vol. 19, No. 4, Jul ÕAug 2001 ribution subject to AVS license or copyright; see http://scitation.aip.org/term e r- f of - of n in a- i- B e - - l n ts f d b- A and B signals! decreases with the increase of vanad Raman spectroscopy, which probes the catalyst surf shows that the vanadyl (V41) and vanadate (V51) groups are present in all samples prepared by sol–gel. The EPR in sities of signals A and B level off for higher vanadia conte and the intensity of signal C increases significantly. Thus, conclude that a partial oxidation of V41 to V51 takes place for higher concentrations of V2O5. The oxidation of V41 in rutile has been reported in the literature.22 IV. CONCLUSIONS The sol–gel method employed for the preparation V2O5/TiO2 catalysts allows a better control of their comp sition, homogeneity, dispersion, and structural proper than the impregnation method. The increase of vanad concentration induces the anatase formation. The forma of V2O5 crystallites was not observed in all samples. For samples obtained by the impregnation method, the T2 structure is independent of the vanadia loading, and crys line V2O5 was observed. Raman spectra identified two s cies of surface vanadium: monomeric vanadyl and polyme vanadates to xerogels and crystalline V2O5 in samples pre- pared by impregnation. XPS results indicated the presenc V51. The V/Ti atomic ratios showed that the sol–gel meth favored the vanadium deposition on the titania surface, some vanadium was incorporated into the rutile structure least three families of V41 ions were identified: two isolated V41 ions in locations with octahedral symmetry, substituti Ti41 in the rutile structure, and magnetically interacting V41 ions, present in pairs or clusters, giving rise to a broad unresolved EPR line. For 9 wt % V2O5, a partial oxidation of V41 ~paramagnetic! to V51 ~diamagnetic! was observed by EPR. The quantity of V41 ions in the samples obtained b the impregnation method is very small. ACKNOWLEDGMENTS This work has been supported by the Brazilian agenc FAPESP and CNPq. 1F. Cavani, G. Centi, E. Foresti, F. Trifiro`, and G. J. Busca, J. Chem. Soc Faraday Trans.84, 237 ~1988!. 2K. V. R. Chary, G. Kishan, T. Bhaskar, and C. J. Sivaraj, J. Phys. Ch B 102, 6792~1998!. 3L. Dall’Acqua, M. Baricco, F. Berti, L. Lietti, and E. Giamello, J. Mate Chem.8, 1441~1998!. 4A. Vejux and P. Courtine, J. Solid State Chem.23, 93 ~1978!. 5L. Lietti, P. Forzatti, G. Ramis, G. Busca, and F. Bregani, Appl. Catal. 3, 13 ~1993!. 6G. Centi, E. Giamello, D. Pinelli, and F. Trifiro`, J. Catal.130, 220~1991!. 7T. Mongkhonsi and L. Kershenbaun, Appl. Catal., A170, 33 ~1998!. 8A. Andersson, J. Catal.76, 144 ~1982!. 9H. Hausinger, H. Schmelz, and H. Kno¨zinger, Appl. Catal.39, 267 ~1988!. 10I. E. Wachs, L. E. Briand, J-M. Jehng, L. Burcham, and X. Gao, Ca Today57, 323 ~2000!. 11C. R. Dias, M. F. Portela, and G. C. Bond, J. Catal.157, 344 ~1995!. 12A. M. Efstahiou and K. Fliatura, Appl. Catal., B6, 35 ~1995!. 13G. T. Went, L. J. Leu, and A. Bell, J. Catal.134, 479 ~1992!. 14I. E. Wachs, G. Deo, B. M. Weckhuysen, A. Andreini, M. A. Vuurma M. Boer, and M. D. Amiridis, J. Catal.161, 211 ~1996!. 15G. C. Bond, Appl. Catal., A157, 91 ~1997!. 16B. Grzybowska-Swierkosz, Appl. Catal., A157, 263 ~1997!. sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40 er . A S . . B rts, ron . . C. 1163 Rodella et al. : Chemical and structural characterization 1163 Redist 17O. Zegaoui, C. Hoang-Van, and M. Karroua, Appl. Catal., B9, 211 ~1996!. 18G. M. Pajonk, Appl. Catal.72, 217 ~1991!. 19M. Schneider, M. Maciejewski, S. Tschudin, A. Wokaun, and A. Baik J. Catal.149, 326 ~1994!. 20C. Hoang-Van, O. Zegaoui, and P. Richat, J. Non-Cryst. Solids225, 157 ~1998!. 21M. A. Reiche, E. Ortelli, and A. Baiker, Appl. Catal., B23, 187 ~1999!. 22R. D. Gonzalez, T. Lopez, and R. Gomez, Catal. Today35, 293 ~1997!. 23M. A. Cauqui and J. M. Rodrı´gues-Izquierdo, J. Non-Cryst. Solids147Õ 148, 724 ~1992!. 24B. Xia, H. Huang, and Y. Xie, Mater. Sci. Eng., B57, 150 ~1999!. 25M. C. Paganini, L. Dall’Acqua, E. Giamello, L. Lietti, P. Forzatti, and G Busca, J. Catal.166, 195 ~1997!. 26B. M. Reddy, B. Chowdhurry, I. Ganesh, E. P. Reddy, T. C. Rojas, and Fernández, J. Phys. Chem. B102, 10178~1998!. 27K. V. Narayana, A. Venugopal, K. S. Rama Rao, V. Venkat Rao, Khaja Masthan, and P. Kanta Rao, Appl. Catal., A150, 269 ~1997!. 28G. C. Bond and S. F. Tahir, Appl. Catal.71, 1 ~1991!. 29I. E. Wachs and B. M. Weckhuysen, Appl. Catal., A157, 67 ~1997!. 30G. T. Went, L.-J. Leu, R. Rosin, and A. T. Bell, J. Catal.134, 492~1992!. JVST A - Vacuum, Surfaces, and Films ribution subject to AVS license or copyright; see http://scitation.aip.org/term , . . 31V. Luca, S. Thomson, and R. F. Howe, J. Chem. Soc., Faraday Trans93, 2195 ~1997!. 32G. Busca, L. Marchetti, G. Centi, and F. Trifiro`, Langmuir2, 568~1986!. 33F. Prinetto, G. Ghiotti, M. Occhiuzzi, and V. Indovina, J. Phys. Chem 102, 10316~1998!. 34C. C. P. Forzatti and G. Busca, J. Catal.116, 586 ~1989!. 35A. F. Carley, G. Spoto, P. R. Chalker, J. C. Riviere, and M. W. Robe J. Chem. Soc., Faraday Trans. 1183, 351 ~1987!. 36J. Pouilleau, D. Devilliers, H. Groult, and P. Marcus, J. Mater. Sci.32, 5645 ~1998!. 37N. J. Price, J. B. Reitz, R. J. Madix, and E. I. Solomon, J. Elect Spectrosc. Relat. Phenom.98Õ99, 257 ~1999!. 38A. Aboukaı̈s, C. F. Aissi, M. Dourdin, D. Courcot, M. Guelton, E. M Serwicka, E. Giamello, F. Geobaldo, A. Zecchina, A. Foucault, and J Vedrine, Catal. Today20, 87 ~1994!. 39A. Davidson and M. Che, J. Phys. Chem.96, 9909~1992!. 40H. J. Gerritsen and H. R. Lewis, Phys. Rev.119, 1010~1960!. 41F. Kubec and Z. Sroubek, J. Chem. Phys.57, 1660~1972!. 42M. Grätzel and R. F. Howe, J. Phys. Chem.94, 2566~1990!. 43R. Gallay, J. J. van der Klink, and J. Moser, Phys. Rev.34, 3060~1986!. sconditions. Download to IP: 186.217.234.225 On: Tue, 14 Jan 2014 15:32:40