Quantitative determination of the number of surface active sites and the turnover frequencies for methanol oxidation over metal oxide catalysts: Application to bulk metal molybdates and pure metal oxide catalysts

Article abstract of DOI:10.1006/jcat.2001.3289

The present work investigates the number and nature of the surface active sites and the catalytic activity of bulk metal molybdates and bulk metal oxides on methanol selective oxidation. Bulk metal molybdates were synthesized by coprecipitation and characterized by laser Raman spectroscopy, X-ray photoelectron spectroscopy, X-ray diffraction, and specific surface area analysis. The number of surface active sites (Ns) was determined by measuring the amount of methoxy species produced by methanol chemisorption on the catalysts at 100°C. The specific activity values (TOFs) were calculated by normalizing the reaction rate by the number of surface active sites. The significant differences in catalytic behavior of bulk metal molybdates and bulk metal oxides, and the surface molybdenum enrichment observed on metal molybdates, gave evidences that the surface of bulk metal molybdates may be composed only by molybdenum oxide species in a two-dimensional overlayer. It was not possible to establish the structure of surface molybdenum oxide species; however, it can be concluded that the structure of surface active sites species of bulk metal molybdates and monolayer supported oxide catalysts are different according to the dispersion of the Ns values of these systems. This work shows, for the first time in the literature, that bulk metal molybdates and monolayer supported molybdenum oxide catalysts possess similar activity and TOF in methanol selective oxidation. The number of active surface sites and the specific activity toward selective oxidation products (TOF redox), CO₂ (TOF basic), and dimethyl ether (TOF acid) of bulk metal oxides were also determined. In contrast to bulk metal molybdates, most bulk metal oxides catalyze the total combustion of methanol even at low methanol conversion.

Full text of DOI:10.1006/jcat.2001.3289

Journal of Catalysis 202, 268–278 (2001)
doi:10.1006/jcat.2001.3289, available online at http://www.idealibrary.com on
Quantitative Determination of the Number of Surface Active Sites and the
Turnover Frequencies for Methanol Oxidation over Metal Oxide Catalysts:
Application to Bulk Metal Molybdates and Pure Metal Oxide Catalysts
Laura E. Briand, ,1 Andrew M. Hirt,† and Israel E. Wachs‡
Centro de Investigacio´n y Desarrollo en Procesos Catal´ıticos, CONICET—Univ. Nacional de La Plata, Calle 47 No. 257, (1900) La Plata,
Buenos Aires, Argentina; †Materials Research Laboratories Inc., 290 North Bridge Street, Struthers, Ohio 44471; and ‡Zettlemoyer Center
for Surface Studies and Department of Chemical Engineering, Lehigh University, 7 Asa Drive, Bethlehem, Pennsylvania 18015
Received January 22, 2001; accepted May 23, 2001
tive oxidation over metal molybdate and bulk metal ox-
ide catalysts. Previous studies established the fundamen-
tals of the methanol chemisorption technique to determine
the number of surface active sites for methanol selective
oxidation on monolayer supported molybdenum and vana-
dium oxide catalysts (1–3). Spectroscopic analysis (infrared
and temperature programmed reaction spectroscopy) of
methanol adsorbed on monolayer supported molybdenum
and vanadium catalysts under several temperatures and al-
cohol partial pressures showed that methanol is molecu-
larly physisorbed at room temperature and dissociatively
chemisorbed as methoxy -OCH3 species at higher temper-
atures. At 100 C and 2000 ppm of methanol vapor in the gas
phase it ispossible to cover the surface ofan oxide with a sta-
ble monolayer of surface methoxy species. Surface methoxy
is the intermediate species in the formation of methanol
selective oxidation products (formaldehyde, methyl for-
mate and dimethoxymethane) and, therefore, the amount
of these species can be taken as the number of surface ac-
tive sites. The calculation of the catalytic activity per ac-
tive site (per methoxy) that is, the turnover frequency TOF
of monolayer molybdenum oxide supported on Al2O3,
TiO2, ZrO2, SiO2, MnO, NiO, Nb2O5, and Cr2O3 allowed
the comparison of the reactivity of this series of cata-
lytic systems (1–3). The present work extends the use of the
methanol chemisorption technique to bulk metal molyb-
date and bulk metal oxides catalysts. The interest in metal
molybdates and MoO3–metal molybdate mixtures began
in 1926 when the high efficiency of the MoO3/Fe2(MoO4)3
system to produce formaldehyde from methanol in excess
of oxygen was discovered (4). During the following years,
much research was devoted to understanding the catalytic
behavior and the nature of the active sites of molybdenum-
based catalytic systems.
The present work investigates the number and nature of the sur-
face active sites and the catalytic activity of bulk metal molybdates
and bulk metal oxides on methanol selective oxidation. Bulk metal
molybdates were synthesized by coprecipitation and characterized
by laser Raman spectroscopy, X-ray photoelectron spectroscopy, X-
ray diffraction, and specific surface area analysis. The number of
surface active sites (Ns) was determined by measuring the amount
of methoxy species produced by methanol chemisorption on the
catalysts at 100 C. The specific activity values (TOFs) were cal-
culated by normalizing the reaction rate by the number of surface
active sites. The significant differences in catalytic behavior of bulk
metal molybdates and bulk metal oxides, and the surface molyb-
denum enrichment observed on metal molybdates, gave evidences
that the surface of bulk metal molybdates may be composed only
by molybdenum oxide species in a two-dimensional overlayer. It
was not possible to establish the structure of surface molybdenum
oxide species; however, it can be concluded that the structure of sur-
face active sites species of bulk metal molybdates and monolayer
supported oxide catalysts are different according to the dispersion
of the Ns values of these systems. This work shows, for the first
time in the literature, that bulk metal molybdates and monolayer
supported molybdenum oxide catalysts possess similar activity and
TOF in methanol selective oxidation. The number of active surface
sites and the specific activity toward selective oxidation products
(TOF redox), CO2 (TOF basic), and dimethyl ether (TOF acid) of
bulk metal oxides were also determined. In contrast to bulk metal
molybdates, most bulk metal oxides catalyze the total combustion
c
2001 Academic Press
of methanol even at low methanol conversion.
Key Words: methanol chemisorption; metal molybdates; surface
active sites; turnover frequency; methanol oxidation; selective oxi-
dation; bulk catalysts; supported catalysts; metal oxides.
1. INTRODUCTION
The present investigation focuses on the quantification
of the number of surface active sites and methanol selec-
Sleight et al. performed spectroscopic studies on metha-
nol adsorbed on pure MoO3 and MoO3/Fe2(MoO4)3 that
combined with the reaction mechanism gave more in-
sights on the number of active sites and the nature of
1 To whom correspondence should be addressed. E-mail: briand@
dalton.quimica.unlp.edu.ar.
268
0021-9517/01 $35.00
c
Copyright
2001 by Academic Press
All rights of reproduction in any form reserved.

OXIDATION OVER METAL OXIDE CATALYSTS
269
surface intermediates (5–7). The authors demonstrated surface Bi/Mo concentration. The authors demonstrated
that methanol is adsorbed as undissociated molecules and that only well crystallized bismuth–molybdenum phases are
methoxy species only on a fully oxidized surface. Surface selective catalysts toward methanol reaction. The analysis
methoxy species are the intermediate species for formalde- of the near surface region by means of XPS spectroscopy
hyde production and the breaking of the carbon–hydrogen showed a comparable surface and bulk molybdenum con-
bond is the rate-limiting step in the reaction of surface centration before reaction with methanol.
methoxy groups over molybdates. This group observed that
The system SnO2–MoO3 was found to be active in
MoO3 was 2–4 times less active toward methanol oxida- methanol selective oxidation to formaldehyde and methyl
tion than Fe2(MoO4)3 when normalized by the number of formate (14). Weng et al. suggested the formation of
methanol adsorbed at room temperature. The observation SnMoO2/MoO3 when the mixture iscalcined at 873K. More
that methanol does not dissociatively chemisorb on coor- recently, Rivarola et al. suggested a synergetic effect bet-
dinatively saturated Mo ions (coordination number of 6), ween the oxides that enhance the catalytic activity of the
which are the predominant sites on orthorhombic MoO3, mixture based on the observation of the lower activity of
suggested that the activity of iron molybdate was related to the individual oxides (15). This effect would be the origin
the fact that all surface molybdenum atoms in the pseudo- of Mo5+ sites that would improve the redox capacity of
cubic Fe2(MoO4)3 have a coordination number less than the mixture. According to spectroscopic analysis, no new
6. Further kinetic studies of methanol selective oxidation compounds or modification of the surface composition are
to formaldehyde over ferric and ferrous molybdate, bis- produced when the mixture is calcined at 353 K or under
muth, chromium, aluminum, and heteropoly molybdates, reaction conditions. Pure bulk metal oxides and mixtures
demonstrated that a wide range of metal molybdates are of oxides have also received attention in methanol oxida-
active and selective on the process and that the same mech- tion literature. Spectroscopic studies of adsorbed methanol
anism of methanol adsorption and reaction is accompl- and temperature programmed desorption analysis demon-
ished (8).
strated that surface methoxy and formate species are gen-
An interesting, if not the only in situ electron microscopy erated upon exposure of metal oxides (Al2O3, ZrO2, TiO2,
study during methanol reaction over Fe2(MoO4)3, MoO3, NiO, Fe2O3, ZnO, CuO, Cu2O, Cr2O3, CeO2, Bi2O3, Co3O4,
and MoO3/Fe2(MoO4)3 was performed by Gai et al. (9). Mn2O3) to methanol vapor. The product distribution of
The experiments on ferric molybdate showed that under methanol reaction depends on the redox and/or acid–base
in situ reaction conditions using CH3OH, CH3OH–O2, and nature of the surface active sites; however, most pure metal
H2 environments, the catalyst is reduced to ferrous molyb- oxides produce COx above 200 C due to the decomposition
date (FeMoO4), -Fe2O3, and metallic Fe. The extent of the of surface formate species.
reduction is limited and the activity and selectivity of the
Wachs et al. performed in situ Raman spectroscopy and
catalyst are prolonged in the presence of gaseous oxygen. steady state kinetic measurements of methanol selective
Experiments on MoO3 using CH3OH–O2 showed the pres- oxidation over Al2O3, TiO2, ZrO2, Nb2O5, and supported
ence of suboxides and regions of ordered oxygen vacancies. molybdenum oxide catalysts (16). The bare oxide supports
In situ experiments carried out on iron molybdate–MoO3 possess low activity toward methanol selective oxidation
physical mixtures showed that the same changes described at 230 C. Dehydration of methanol to dimethyl ether and
for the individual components are produced in the mixture CO/CO2 generation were the major products of reaction.
under reaction conditions.
Similarly, methanol dehydrogenates to CO, CO2, and H2
Kinetics studies performed by Mann et al. and more re- over Cr2O3 at 300 C (17). In addition, dimethyl ether and
cently by Ivanov et al. indicated that a mixture of MoO3 water are produced on acid sites of chromium oxide.
and MnMoO4 possess activity and selectivity similar
In contrast, Fe2O3 and copper-based oxides possess se-
to MoO3/Fe2(MoO4)3 in methanol selective oxidation lective oxidation activity. Murakami et al. demonstrated
to formaldehyde (10a, 10b, 11). Fe2(MoO4)3/Cr2(MoO4)3/ that hematite catalyzed the production of small amounts
MoO3 mixture has also been proposed as suitable catalytic of formaldehyde, methyl formate, and carbon monoxide at
system to oxidize methanol to formaldehyde (12). The spec- 225 C and low conversions (18).
troscopic analysis of the catalyst before and after reaction
showed the modification of the composition due to the in- investigations since metallic copper and metallic silver were
teraction with methanol. the former industrial catalysts used to produce formalde-
Methanol reaction on copper was the subject of several
Wachs et al. physically and chemically characterized a hyde on a large scale. Studies of methanol chemisorp-
series of bismuth molybdates ( -Bi2Mo3O12, -Bi2Mo2O9, tion on copper(110)-oriented single crystal surfaces cov-
-Bi2MoO6, 0-Bi2MoO6, Bi38Mo7O78, etc.) by means of ered with chemisorbed oxygen and CuO powder showed
methanol reaction and spectroscopic analysis (13). Bismuth that methanol interacts with adsorbed oxygen to form sur-
molybdates are selective to formaldehyde (58–98% at low face CH3O species that decompose to formaldehyde and
methanol conversions) regardless of Mo coordination or H2 slightly above room temperature (19, 20).

270
BRIAND, HIRT, AND WACHS
Methanol steady state kinetics of methanol reaction over required amount of ammonium heptamolybdate was dis-
Bi2O3 showed the production of CO2 due to the basicity solved in 200 ml of distilled water and the solution was
of the oxide (13). Wachs et al. performed methanol re- added to the metal–citric-acid solution. The mixture was
action under differential conditions over a series of bulk dried in a steambath until a glassy texture was observed.
Bi2O3 with different specific surface areas. The authors The precursor was further dried in a vacuum oven at 70 C
detected combustion products (CO2, CO) as major prod- overnight, grounded, and calcined at 500 C for 24 h. The
ucts and formaldehyde in minor amounts regardless of the precursors of Al2(MoO4)3 and Zr(MoO4)2 were calcined at
specific surface area or the calcination temperature of the 600 and 700 C, respectively, in order to obtain crystalline
oxide. Similarly, CeO2, Co3O4, ZnO, and NiO catalyzed materials.
methanol combustion over a wide range of temperatures
and methanol conversions (21–24).
Molybdenum trioxide was synthesized by thermal de-
composition of (NH4)6Mo7O24 4H2O (Alfa Aesar 99.9% )
The present work applies methanol chemisorption to at 300 C for 14 h. Bulk metal oxides studied in the present
quantify the number of surface active sites and presents work are commercial high purity chemicals (Alfa Aesar
a reliable comparison of the catalytic activities per ac- Puratronic 99.99% ).
tive site (turnover frequencies) of bulk metal molybdates,
monolayer molybdenum oxide catalysts, and metal ox-
2.2. Catalyst Characterization
ides with the aim of obtaining insights into the cata-
lytic behavior of supported and bulk molybdate catalytic
metal molybdate phases was determined by Raman spec-
systems. This work also deals with one of the fundamental
troscopy. The spectra were obtained under ambient con-
2.2.1. Laser Raman spectroscopy. The purity of bulk
ditions with an Ar+ ion laser (Spectra Physics Model
2020-50, excitation line 514.5 nm) delivering 15–40 mW of
incident radiation. The powdered solid (aprox. 100–200 mg)
was pressed into a thin wafer about 1 mm thick that was
mounted onto a spinning sample holder and rotated at
2000 rpm to avoid local heating effects. The scattered ra-
diation from the sample was directed into a Spex Triple-
mate spectrometer (Model 1877) coupled to a Princeton
Applied Research OMA III optical multichanneled ana-
lyzer (Model 1463) equipped with an intensified photodi-
questions of catalysis: the true surface composition of bulk
catalysts and its difference with bulk composition. The na-
ture of the surface of metal molybdates will be mapped
comparing the products distribution of methanol reaction
over MoO3 and pure bulk metal oxides.
2. EXPERIMENTAL
2.1. Catalysts Synthesis
Fe2(MoO4)3, NiMoO4, CoMoO4, MnMoO4, and Bi2-
Mo3O12 were synthesized by coprecipitation from (NH4)6-
Mo7O24 4H2O (Alpha Aesar Products, 99.9% ) and the
corresponding metal nitrate [Fe(NO3)3 9H2O, Ni(NO3)2-
6H2O, Co(NO3)2 6H2O, Bi(NO3)3 5H2O Alpha Aesar
Products, 99.9% ] or chloride (MnCl2 3H2O, Alpha Ae-
sar Products, 99.9% ) (25–29). A solution of ammonium
heptamolybdate was added dropwise to the metal nitrate
/chloride solution under stirring. The mixture was main-
tained at 80 C and pH 6.0 (pH 1.75 for Fe2(MoO4)3) by
adding HNO3 (1 M solution) or NH4Cl (1 M solution) as
needed. After aging for 3 h, the precipitate was filtered,
washed with distilled water, and dried at 100 C overnight.
The precursor was calcined at 500 C for 4 h.
ode array detector cooled to 243 K. The spectral resolution
1
and reproducibility are 2 cm
.
2.2.2. Specific surface area. The BET surface areas of
the samples were determined by N2 adsorption at 77 K on
a Quantasorb surface area analyzer (Quantachrome Cor-
poration, Model OS-9).
2.2.3. X-ray photoelectron spectroscopy. The near-
surface composition of bulk metal molybdates was investi-
gated using X-ray photoelectron spectroscopy (XPS). The
analysiswasperformed in a ModelDS800XPSsurface anal-
ysis system (Kratos Analytical Plc, Manchester, UK) that
operates with an X-ray beam of predominantly MgK or
9
AlK X-rays and a base pressure of 5 10 Torr. The
specimens for XPS analysis were prepared by pressing the
catalyst powder between a stainless steel holder and a pol-
ished single crystal silicon wafer.
CuMoO4, Zr(MoO4)2, Cr2(MoO4)3, ZnMoO4, Ce8-
Mo12O49, and Al2(MoO4)3 were synthesized through an
organic route from (NH4)6Mo7O24 4H2O (Alpha Aesar
Products, 99.9% ), the corresponding metal nitrate
2.2.4. X-ray diffraction. X-ray diffraction spectra of
[Cu(NO3)2 6H2O, Zn(NO3)2 6H2O, Cr(NO3)3 9H2O, chromium, cerium, aluminum, and zirconium molybdates
(NH4)2Ce(NO3)6, Al(NO3)3 9H2O Alpha Aesar Prod- were obtained with a Philips XRD-3720 equipment with an
˚
ucts, 99.9% ] or chloride(ZrCl4, Alpha Aesar Products, XRG-3100 X-ray generator (CuK , 1.54060–1.54439 A).
99.9% ) and citric acid [HOC(COOH) (CH2COOH)2 H2O Data acquisition and processing was performed with a DEC
Alfa Aesar 99.9% ] (30).
(Digital Equipment Corporation) Microvax 3100 computer
Fifteen grams of the metal containing chemical (nitrate/ system. The operating conditions were: voltage, 45 kV; cur-
chloride) were dissolved in 200mlofdistilled water to which rent, 30 mA; peak angle range, 2.00–90.02 ; sample time,
20 g of citric acid was added and stirred to dissolve. The 30.00 s/deg.

OXIDATION OVER METAL OXIDE CATALYSTS
271
temperatures for methanol oxidation over bulk metal ox-
ides: CeO2, 380 C; MnO, 300 C; Cr2O3, 290 C; NiO, 300 C;
CoO, 270 C; CuO, 330 C; Fe2O3, 300 C and ZnO, 380 C.
Bi2O3 is active at temperatures above 450 C.
Methanol conversion and the amount of products were
quantified with an on-line gas chromatograph (HP 5840)
equipped with TCD and FID detectors and two columns:
capillary column (CP-sil 5CB) for methylal, dimethyl ether,
methyl formate, and methanol analysis and a packed col-
umn (Carboxene-1000) for CO, CO2, O2, formaldehyde and
methanol analysis.
3. RESULTS
3.1. Bulk and Surface Characterization
of Metal Molybdates
The purity of MoO3 and bulk metal molybdates was
determined by Raman spectroscopy. Figures 1–3 show the
spectra of bulk metal molybdates. Pure MoO3 possess three
1
intense Raman signals at 996, 819, 668 cm and smaller
1
bands at 380, 339, 291, 246, 161, and 131 cm that have
been well documented in the literature for orthorhombic
crystalline alpha-MoO3 (data not shown). Bulk metal
FIG. 1. Raman spectra of bulk metal molybdates.
2.3. Methanol Chemisorption and Oxidation
The number of surface active sites was quantified by
methanol chemisorption. The catalysts were exposed to a
mixture of 2000 ppm of methanol vapor in He at 100 C and
the weight change was determined with a Cahn TGA mi-
crobalance (Model TG-131) coupled with a PC for temp-
erature and weight monitoring. The system allowed a
controlled flow of high purity gases: air for pretreatment,
helium and a mixture of methanol in helium for adsorption
experiments. A detailed flow diagram of the system and
the methanol chemisorption technique has been previously
published (3).
Methanol oxidation steady state kinetics was obtained in
a fixed-bed catalytic reactor under differential conditions.
The following operating parameters were used in order to
maintain methanol conversion below 10% for methanol re-
action over metal molybdates: sample weight, 10 mg; re-
1
action temperature, 380 C; flow rate, 100 cm3 (NTP) min
;
and feed gas composition methanol/oxygen/helium, 6/13/81
mol% . The activity of bulk metal oxides varies greatly be-
tween the different systems; therefore, the reaction tem-
perature was adjusted for each individual metal oxide to
reach a low methanol conversion. The following are the
FIG. 2. Raman spectra of bulk metal molybdates.

272
BRIAND, HIRT, AND WACHS
though Raman analysis demonstrated that metal molyb-
dates do not segregate bulk MoO3 or metal oxides, the sur-
face of these systems is generally nonstoichiometric due to
enrichment in molybdenum. Surface molybdenum concen-
tration increased after exposure to methanolunder reaction
conditions. These results were further confirmed compar-
ing the activity and selectivity on the methanol oxidation
reaction over metal molybdates with pure MoO3 and metal
oxides, which are presented below.
3.2. Surface Active Sites Quantification and Turnover
Frequency in Methanol Selective Oxidation
The number of surface active sites present in bulk metal
molybdates, pure MoO3, and metal oxides was deter-
mined by covering the surface with a stable monolayer of
chemisorbed methoxy species (1–3). Previous studies es-
tablished that the maximum surface methoxy coverage is
reached exposing the samples at 100 C to an atmosphere
of 2000 ppm of methanol in helium. The amount of surface
methoxy species can be calculated with the weight change
of the catalyst monitored with a microbalance.
The quantification of the amount of methoxy species that
can be formed on the catalyst surface allows the calcu-
lation of the catalytic activity per surface active site and
the direct comparison of the activity of different systems.
The specific catalytic activity (turnover frequency TOF) to-
ward methanol selective oxidation was calculated by deter-
mining the production rate of redox products (formalde-
hyde, methyl formate, and dimethoxymethane) normalized
per surface site. Table 2 presents the specific surface area,
surface density of active sites expressed as the amount of
FIG. 3. Raman spectra of bulk metal molybdates.
molybdate phases are free of excess molybdenum since no
MoO3 Raman signals were detected. Additionally, pure
CuO, Cu2O, NiO, ZnO, Cr2O3, CeO2, and MnO metal
oxides were analyzed by Raman spectroscopy. Table 1
presents the Raman bands obtained for those systems and
the signals of iron oxides, ZrO2, Al2O3, and bismuth oxides
from literature references (31–33). The absence of Raman
bands belonging to bulk metal oxides in metal molybdates
spectra ensures that no segregation of bulk metal oxides
occurs during calcination. Raman spectra of MnMoO4,
NiMoO4, CoMoO4, -Bi2Mo3O12, and Fe2(MoO4)3 coin-
cides with the signals previously reported in the literature
(26, 27, 29, 33, 34). The structures of chromium, cerium,
aluminum, and zirconium molybdates were primarily
analyzed by X-ray diffraction since no Raman spectra
are available in the literature. XRD spectra presented
in Fig. 4 belong to the following structures: Cr2(MoO4)3,
Ce8Mo12O49, Al2(MoO4)3, and Zr(MoO4)2 (35). Zinc
and copper molybdates were identified with commercial
ZnMoO4 and CuMoO4 (Alfa Aesar 99.9% ). No new sig-
nals or other modifications in the Raman spectra of metal
molybdates were observed after reaction with methanol.
Figure 5 shows surface molybdenum/metal ratio of metal
molybdates before and after reaction with methanol. Al-
TABLE 1
Raman Bands of Pure Bulk Metal Oxides
1
Metal oxide
Raman bands (intensity) [cm
]
CuO
Cu2O
NiO
123 (s), 158 (m), 294 (m), 462 (w), 632 (w)
122 (m), 154 (m), 219 (s), 633 (s)
502 (s)
ZnO
109 (m), 329 (m), 377 (m), 433 (s), 1171 (s)
550 (s), 600 (s)
450 (s)
648 (m)
662 (s), 532 (w)
Cr2O3
CeO2
MnO
Fe3O4
-Fe2O3
a
a
676 (s), 468 (w), 390 (w), 124 (s)
605 (m), 493 (w), 403 (m), 289 (s), 221 (s), 127 (m)
180 (s), 190 (m), 220 (w), 300 (w), 330 (m), 350 (m),
385 (m), 470 (s), 530 (w), 550 (w), 610 (m), 625 (s).
a
-Fe2O3
ZrO2
b
b
1
Al2O3
-Bi2O3
-Bi2O3
-Bi2O3
no Raman bands in the 100–1100 cm region
c
451 (m), 410 (w), 319 (m), 298 (m), 213 (m), 175 (w)
450 (w), 307 (m), 294 (s), 147 (vs)
570 (m)
c
c
Note. s, strong; m, medium; w, weak.
a Data from Ref. (31).
b Data from Ref. (32).
c Data from Ref. (33).

OXIDATION OVER METAL OXIDE CATALYSTS
273
FIG. 4. X-ray diffraction spectra of Cr2(MoO4)3, Ce8Mo12O49, A12(MoO4)3, and Zr(MoO4)2.
methoxy species per surface area (Ns), reaction rate, TOF, were produced under low methanol conversion reaction
and selectivity of bulk metal molybdates. Surface active site conditions. Further experiments performed on NiMoO4
density of copper and bismuth molybdates are probably and Fe2(MoO4)3 suggest that the specific activities and the
subject to error due to the low surface area determined surface active sites density of bulk metal molybdates are
with the BET method; however, this parameter does not independent of the specific surface area; however, the se-
influence TOF values.
lectivity toward formaldehyde varies slightly with surface
Bulk metal molybdates are active in methanol reac- area (Table 3).
tion at 380 C and highly selective toward formaldehyde.
Dimethyl ether was also detected while no carbon oxides methanol selective oxidation of pure bulk metal oxides was
also studied. The significant differences between the cata-
lytic behavior of metal molybdates and metal oxides in
methanol reaction are a key factor in uncovering the true
surface composition of bulk molybdates. Table 4 shows the
specific surface area, surface density of active sites (Ns),
turnover frequencies, and selectivity of pure bulk metal ox-
ides at low methanol conversion. Again, the low surface
areas of MnO, CoO, CuO, and Bi2O3 could be affecting
the surface active sites density values; however, the specific
activity TOF is independent of this parameter.
The amount of surface active sites and the reactivity for
The number of surface active sites, specific activities, and
selectivity values of ZrO2 and Al2O3 have been taken from
literature references (16, 36). The activity of bulk metal
oxides was determined at different temperatures (typically
300 C) in order to maintain low methanol conversion val-
ues (see Section 2.3). It is expected that secondary reactions
do not affect the main reaction at low methanol conver-
sion and, therefore, the distribution of products reflects the
nature of the surface active sites.
Metal oxides possess high catalytic activity and selec-
FIG. 5. Surface molybdenum/metal atomic ratio of bulk metal molyb-
tivity toward total combustion of methanol even at low
conversion, with the exception of cerium, copper, and
dates: (᭜) theoretical ratio and experimental values obtained (᭿) before
and (᭡) after reaction with methanol.

274
BRIAND, HIRT, AND WACHS
TABLE 2
Surface Active Sites and Reactivity of Bulk Metal Molybdates and MoO3 Catalysts
toward Methanol Selective Oxidation
SBET
[m2/g]
Ns
[ mol/m2]
Reaction ratea
[ mol/m2 sec]
TOFb
Selectivity [% ]c
DME
1
Catalyst
[s
]
FA
DMM
Zr(MoO4)2
Ce8Mo12O49
MnMoO4
Cr2(MoO4)3
Al2(MoO4)3
NiMoO4
CoMoO4
CuMoO4
Fe2(MoO4)3
ZnMoO4
Bi2Mo3O12
MoO3
31.3
4.0
1.9
1.7
2.8
9.5
5.5
0.8
9.6
2.1
0.2
5.0
1.0
12.0
3.1
12.6
5.0
36.0
46.5
20.6
43.1
9.4
24.8
3.3
6.6
3.1
1.8
0.9
0.9
1.1
1.4
1.0
1.4
5.3
62.1
86.5
99.8
91.0
26.1
100.0
88.2
100.0
61.0
87.4
100.0
84.1
34.4
13.5
—
9.0
73.9
—
11.8
—
39.0
12.6
—
3.1
—
—
—
—
—
—
—
—
—
—
—
2.8
4.1
2.6
4.4
23.9
4.2
15.7
16.5
0.7
24.7
9.6
18.2
23.4
4.4
15.9
a Activity based on overall methanol conversion at 380 C.
b Turnover frequency based on methanol partial oxidation products (formaldehyde and dimethoxy methane); Reac-
tion temp., 380 C.
c Selectivity toward formaldehyde (FA), dimethyl ether (DME), and dimethoxy methane (DMM) at 380 C.
bismuth oxides that produce only formaldehyde and Al2O3 near surface region of a sample. The photoelectrons that
that catalyzes the dehydration of methanol to dimethyl escape from the solid, and are subsequently detected dur-
ether. Further experimentsofmethanolreaction over MnO, ing the analysis, originate from a narrow mean free path
˚
Cr2O3, NiO, CoO, Fe2O3, and CuO at 380 C (Bi2O3 is in- of 10–20 A for kinetic energies in the 15–1000 eV range
active below 450 C) showed that CO2 is the major product (37). In contrast, a reactive molecule only interacts with
of reaction.
the exterior surface of a catalyst and may be more reli-
able for the determination of the true surface composition
of a catalyst than a spectroscopic technique. Methanol is a
chemical probe molecule that can provide fundamental in-
formation about the number of surface active sites and the
distribution of different types of surface sites of metal ox-
ide catalysts. The methanol oxidation product distribution
reflects the nature of the surface active sites since redox,
acid, and basic sites yield formaldehyde (HCHO), dimethyl
ether (CH3OCH3), or CO2, respectively (38). Bulk metal
molybdates and MoO3 catalyze methanol selective oxida-
tion toward formaldehyde and dehydration to dimethyl
4. DISCUSSION
4.1. Surface Composition and Catalytic Activity
of Bulk Metal Molybdates
One of the fundamental questions of all catalytic stud-
ies about bulk compounds is the true surface composition
and its difference from the bulk composition. Regularly it
is assumed that the properties of the bulk reflect the state
of the surface. X-ray photoelectron spectroscopy is a valu-
able tool that measures the concentration of atoms in a
TABLE 3
Surface Active Sites and Reactivity of NiMoO4 and Fe2(MoO4)3 Catalysts with Varying Surface Areas
SBET
[m2/g]
Ns
[ mol/m2]
Reaction ratea
[ mol/m2 sec]
TOFb
Selectivityc
[% ]
1
Catalyst
NiMoO4
Synthesis method
[s
]
inorganic
inorganic
9.5
15.6
2.8
3.3
2.6
8.6
0.9
2.3
100.0
87.0
Fe2(MoO4)3
organic
inorganic
inorganic
1.5
3.9
9.6
6.7
8.3
4.2
26.7
23.3
9.6
2.3
1.8
1.4
58.0
64.9
61.0
a Activity based on overall methanol conversion at 380 C.
b Turnover frequency based on methanol selective oxidation products (formaldehyde and dimethoxy methane);
Reaction temp., 380 C.
c Selectivity toward methanol selective oxidation products at 380 C.

OXIDATION OVER METAL OXIDE CATALYSTS
275
TABLE 4
Surface Active Sites and Reactivity of Bulk Metal Oxides toward Methanol Selective Oxidation at Low Conversion
SBET
[m2/g]
Ns
[ mol/m2]
TOFa
Selectivity [% ]
1
Oxide (reaction temp.)
[s
]
FA
CO2
CO
DME
DMM
MF
b
ZrO2 (230 C)
39.0
2.4
0.8
3.0
180.0
1.1
0.4
0.3
21.4
5.3
0.2
1.1
4.8
1.6
12.4
5.6
4.4
2.2
7.0
3.2
0.3
2.0
0.1
—
14.0
—
20.5
59.6
—
17.4
34.2
—
—
—
—
4.7
—
—
—
—
—
22.9
—
trace
—
—
—
100.0
—
—
—
36.4
—
—
—
—
—
—
—
—
—
—
1.2
—
—
86.0
—
—
—
—
—
4.7
—
4.5
12.8
—
CeO2 (380 C)
MnO (300 C)
Cr2O3 (290 C)
1.3
31.0
7.1
0.0
6.4
24.7
15.9
3.3
18.2
24.7
100.0
79.5
35.7
—
82.6
61.2
100.0
57.9
32.7
100.0
b
Al2O3 (300 C)
NiO (300 C)
CoO (270 C)
CuO (330 C)
Fe2O3 (300 C)
ZnO (380 C)
Bi2O3 (450 C)
—
31.5
—
a Turnover frequency based on methanol partial oxidation products (formaldehyde, methyl formate and dimethoxy methane).
b Ns, reaction rate (for TOF calculation) and selectivity values from Refs. (16, 36).
ether due to the presence of surface redox and acid sites, 500 C and react with methanol forming a volatile metal–
respectively. In contrast, the basic nature of the surface ac- methoxy complex at 200–250 C (39). This property of
tive sites of ZrO2, MnO, Cr2O3, NiO CoO, and ZnO cata- molybdenum species suggests that the surface molybde-
lyze the total combustion of methanol to CO2 even at low num enrichment of bulk metal molybdates is produced dur-
methanol conversion (see Tables 2 and 4). This observation ing calcination at 500 C and is increased due to the reac-
suggests that the activity toward methanol selective oxi- tion environment (6% methanol/oxygen–helium at 380 C).
dation of zirconium, manganese, chromium, nickel, cobalt, The thermal treatment of the precursors and the nature
and zinc molybdates is attributed to the presence of surface of the reactant play an important role in the formation
molybdenum oxide species. The catalytic behavior at low of the catalyst surface. Many publications stress the in-
methanol conversion of CeO2, CuO, Fe2O3, and the corre- fluence of certain parameters of the synthesis process like
sponding bulk metal molybdates (Ce8Mo12O40, CuMoO4, pH, temperature or atomic ratio of the components on the
and Fe2(MoO4)3) is alike. The presence of the metals on catalytic properties of mixed oxides; however, the present
the surface of these bulk metal molybdates cannot be dis- work shows that those parameters influence the bulk ho-
carded even though the XPS analysis showed a surface mogeneity rather than the surface of the metal molybdate
enrichment of molybdenum (see Fig. 5). In the particular catalysts(40). For example, Tables2and 3showthat the syn-
case of Al2(MoO4)3 the presence of aluminum on the sur- thesis in inorganic (i.e., NiMoO4, MnMoO4, Fe2(MoO4)3,
face is undoubted considering the selectivity to dimethyl CoMoO4, Bi2Mo3O12) or organic media (the rest of molyb-
ether of the catalyst and Al2O3. Finally, the high selectiv- date series) does not influence the specific surface area or
ity of Bi2Mo3O12 toward formaldehyde at 380 C, could be the surface composition of metal molybdates.
only attributed to surface molybdenum oxide species since
Bi2O3 is inactive at that temperature.
Finally, it is important to notice that the turnover fre-
quency of most metal molybdates are similar or even higher
The great difference between the specific activity val- than the TOF value of Fe2(MoO4)3, which is known to be
ues (TOFs) of bulk metal oxides and metal molybdates is one of the components of the catalytic system used to pro-
also marked. In general, metal oxides are more active in duce formaldehyde on an industrial scale. Moreover, with
methanol reaction and less selective toward partial oxida- the exception of zirconium and aluminum molybdates, all
tion products than MoO3 and bulk metal molybdates.
the metal molybdates investigated are more selective to-
The differences in catalytic behavior between MoO3 ward formaldehyde than iron molybdate.
and metal oxides and XPS results suggest that the surface
of metal molybdates is composed mainly of molybdenum
4.2. Comparison of Methanol Selective Oxidation
atoms. Moreover, the Raman spectra of metal molybdates
over Bulk Metal Molybdates and Monolayer
do not show MoO3 or new compounds after reaction, which
Supported Molybdenum Catalysts
indicates that molybdenum is covering the catalyst’s surface
probably in a single two-dimensional overlayer.
The catalytic activity of bulk and supported catalysts has
Studies of thermal- and reaction-induced spreading of been a matter of controversy in the catalytic field. The
MoO3 and V2O5 over metal oxides demonstrated that higher catalytic activity of supported than bulk catalysts
molybdenum and vanadium spontaneously migrate at 400– has been reported in the literature; however, this conclusion

276
BRIAND, HIRT, AND WACHS
surface coverage; however, the activity of bulk molybdate
catalysts appears to be somewhat dependent of Ns. Never-
theless, activity values are similar for both systems.
Figure 7 compares the turnover frequencies toward
selective oxidation products of monolayer molybdenum
supported catalysts and bulk metal molybdates with
the same components, such as, Zr(MoO4)2 and 3%
MoO3/ZrO2, MnMoO4 and 2% MoO3/MnO, NiMoO4 and
4% MoO3/NiO, Al2(MoO4)3 and 18% MoO3/Al2O3 and
Cr2(MoO4)3 and 3% MoO3/Cr2O3. The surprising similar-
ity of the turnover frequency of bulk and supported molyb-
denum based catalysts gives further proof that the surface
composition of these systems is alike although it is not
possible to conclude about the surface molybdenum oxide
structures.
FIG. 6. Comparison of reaction rate and surface active sites density
(Ns) of (᭡) bulk metal molybdates and (᭿) monolayer molybdenum oxide
supported catalysts.
4.3. Reactivity of Pure Bulk Metal Oxides
In line with typical catalytic research performed on bulk
metal oxide systems, the studies of the reactivity of metal
oxidesisfocused on reactant conversion and productsselec-
tivity. Kinetic experiments are typically performed mainly
at high temperatures and high conversion due to the appli-
cation of metal oxides in the catalytic abatement of air pol-
lutants. In the 1960s, some efforts were devoted to establish-
ing regularity in the catalytic behavior of pure metal oxides
in hydrocarbon oxidation (45–47). Correlations between
the catalytic activity and electronegativity, heat of forma-
tion, and acid–base property of the oxides were reported.
The present work will review those observations consider-
ing the catalytic activity per active surface site, TOF, of the
metal oxides.
was made on the basis of conversion of reactant to products
instead ofspecific activitiesdue to the uncertaintyregarding
the number of surface active sites of bulk catalysts (41–43).
The present work overcomes this problem since methanol
chemisorption enables the quantification of surface active
sites of bulk metal molybdates and the calculation of the
activity per surface active site. Previous work established
the density of surface active sites by methanol chemisorp-
tion technique and the turnover frequencies in methanol
selective oxidation of monolayer molybdenum supported
on TiO2, Al2O3, Nb2O5, NiO, MnO, ZrO2, Cr2O3, and SiO2
(3). In order to compare the catalytic activity of bulk metal
molybdates and monolayer molybdenum oxide supported
catalysts, the reaction rate and TOF values of supported
catalysts were extrapolated to 380 C considering that the
activation energy for methanol reaction over supported
molybdenum oxide catalysts is typically 20 kcal/mol (44).
Figure 6 compares reaction rate and surface active sites
density, Ns. The surface active sites density of both systems
shows quite a different behavior. Monolayer molybdenum
supported catalysts possess relatively constant Ns values.
Previous studies determined that the amount of surface ac-
tive sites of monolayer supported molybdenum catalysts is
influenced by polymerized or isolated molybdenum species
rather than the surface molybdenum coordination (octahe-
dral or tetrahedral) (3). Surface active site density of some
bulk metal molybdates are higher than molybdenum sup-
ported catalysts and do not correlate with the surface com-
position or the nature of the metal cation. The dispersion
observed in Ns values is probably related with the surface
structure of the molybdenum oxide species. Unfortunately,
this is a difficult issue to clarify since no spectroscopic tech-
nique is capable of revealing the surface structure of the ac-
tive sites of bulk catalysts. The activity of supported molyb-
denum oxide catalysts is independent of Ns at monolayer
Table 5showsredox, basicand acid TOFsat lowmethanol
conversion, heat of formation, and electronegativity of
FIG. 7. Comparison of specific activity (TOFs) and surface active sites
density (Ns) of (᭡) bulk metal molybdates and (᭿) monolayer molybde-
num oxide supported catalysts.

OXIDATION OVER METAL OXIDE CATALYSTS
277
TABLE 5
molybdates and tungstates that possess oxygen atoms in the
form of well-defined oxyanions. The oxides of the first two
groups activate and adsorb oxygen in the form of highly
active ionic radicals that can be considered as electrophilic
reactants. In the case of olefins, such electrophilic species
are responsible for the degradation of the carbon skele-
ton and total oxidation. The compounds of the third group
do not activate oxygen at the surface and practically no
chemisorption of oxygen is detected on nonreduced cata-
lysts. In conclusion, the high activity of CoO, MnO, Cr2O3,
and NiO (probably CuO could also be considered although
the semiconductor type is not known) toward methanol re-
action, mainly total oxidation, could be attributed to the
presence of highly reactive surface oxygen radicals. The
above-mentioned properties of these oxides suggest that
there should be a correlation between TOF values and the
ionization potential or the electron donor capacity of metal
oxides surface, which would depend on the distribution of
oxidation states of surface cations. This fundamental study
does not go further on metal oxides reactivity, but it is im-
portant to stress that methanol chemisorption and specific
activity values made it possible to reexamine previous the-
ories on metal oxides and are the key to understand the
catalytic behavior of these systems.
Redox, Basic, and Acid Turnover Frequency Values of Bulk Metal
Oxides toward Methanol Reaction
1H0/O atoma Electronegative
TOF [s
]
1
Oxide
[Kcal/mol]
M–O bondb
REDOXc BASICd ACIDe
CoO
57.5
92.0
89.6
38.5
58.4
66.2
83.4
122.7
1.69
1.99
1.99
1.67
1.71
1.45
1.43
2.65
24.7
31.0
7.1
15.9
6.4
3.2
18.2
1.3
12.8
8.0
11.9
—
1.5
—
—
—
—
—
—
1.9
—
—
MnO
Cr2O3
CuO
NiO
Fe2O3
ZnO
12.6
—
CeO2
a Heat of formation per oxygen atom of the oxide.
b Difference of electronegativities (Sanderson) between oxygen and
metal atoms.
c Turnover frequency based on methanol partial oxidation products
(formaldehyde, methyl formate, and dimethoxy methane).
d Turnover frequency based on CO2 production.
e Turnover frequency based on dimethyl ether production.
the metal–oxygen bond of a series of pure metal oxides.
Turnover frequency toward methanol selective products
(formaldehyde, methyl formate, and dimethoxy methane)
indicates the redox capacity, TOF toward CO2 reflects the
basic character and the TOF toward dimethyl ether indi-
cates the acid character of the surface active sites.
5. CONCLUSION
Moro-Oka et al. reported an increment of the catalytic ac-
tivity of metal oxides toward hydrocarbon oxidation with
This study demonstrated that methanol is a selective
the decrease of the metal–oxygen bond strength, which was chemical probe for the quantification and determination of
quantified with the heat of formation of the oxide per oxy- the nature of surface active sites of metal molybdates and
gen atom (45). The present work does not find a correlation bulk metal oxides. The determination of the density of sur-
between the redox or basic TOF with the strength of the face active sitesenablesthe calculation ofthe catalyticactiv-
M–O bond of the oxides. These results demonstrate that ity per active site TOF, which allows a reliable comparison
the rate determining step of methanol oxidation over metal of the catalytic behavior of bulk catalysts. These measure-
oxides does not involve the breaking of a surface metal– ments revealed that the surface of bulk metal molybdates is
oxygen bond. The electronegativity of the metal–oxygen primarily composed of molybdenum species approaching a
bond was also found to be responsible for alcohol dehy- surface two-dimensional layer (a monolayer) of molybde-
drogenation activity of metal oxide catalysts (46). Again num oxide species.
no influence of the electronegativity on TOF values was
observed in the present work.
The comparison of catalytic activity and TOF of bulk
metal molybdates and monolayer molybdenum oxide sup-
Metal oxides can be divided into three groups accord- ported catalysts demonstrated that both systems possess
ing to the surface electron density (48, 49). The first group similar activity and selectivity in methanol selective oxida-
comprises the oxides characterized by a high concentra- tion. The higher surface active sites density of bulk metal
tion of centers that can supply electrons to adsorbed oxy- molybdates than supported molybdenum catalysts suggests
gen molecules generating electron-rich species such as O . that these systems possess different surface molybdenum
These oxides possess p-type semiconductivity due to the structures. Bulk metal oxides showed that they are active
presence of transition metal cations in the valence state toward total oxidation rather than methanol selective ox-
with low ionization potential, such as CoO, MnO, Cr2O3, idation compared with bulk metal molybdates and mono-
and NiO. The second group is characterized by a low con- layer supported molybdenum catalysts. The specific activity
centration of electron donor centers, which adsorb oxygen of the bulk metal oxides does not correlate with parameters
as species O2 less rich in electrons. These are semicon- such as electronegativity or heat of formation, which have
ducting oxides of the n-type, such as ZnO and Fe2O3. The been used to explain the catalytic behavior of these systems
third group involves MoO3, WO3, Nb2O5, transition metal in the past.

278
BRIAND, HIRT, AND WACHS
ACKNOWLEDGMENTS
21. Roozeboom, F., Cordingley, P. D., and Gellings, P. J., J. Catal. 68, 464
(1981).
The authors are grateful to Consejo Nacional de Investigaciones Cientı´- 22. Roberts, D. L., and Griffin, G. L., J. Catal. 101, 201 (1986).
ficas y Tecnolo´gicas (CONICET) of Argentina for a postdoctoral fellow-
ship and the United States Department of Energy—Basic Energy Sciences
for financial support (Grant DEFG02-93ER14350).
23. Christoskova, St. G., Stoyanova, M., Georgieva, M., and Mehandzhiev,
D., Appl. Catal. 173, 95 (1998).
24. Ai, M., J. Catal. 54, 223 (1978).
25. Alessandrini, G., Cairati, L., Forzatti, P., Villa, P. L., and Trifiro, F., in
“Proceedings of the 2nd International Conference on the Chemistry
and Uses of Molybdenum,” p. 186. Oxford, 1976.
REFERENCES
26. Ozkan, U., and Schrader, G. L., J. Catal. 95, 120 (1985).
27. Ozkan, U., Gill, R. C., and Smith, M. R., J. Catal. 116, 171 (1989).
28. Schrader, G. L., Sivrioglu, U., and Basista, M. A., in “Proceedings
of the 4th International Conference on the Chemistry and Uses of
Molybdenum,” p. 415. Climax Molybdenum Co., Golden, CO, 1982.
29. Ozkan, U., and Schrader, G. L., Appl. Catal. 23, 327 (1986).
30. Machiels, C. J., Chowdhry, U., Harrison, W. T. A., and Sleight, A. W., in
“Solid State Chemistry in Catalysis” (R. K. Grasselli and J. F. Brazdil,
Eds.), p. 103. Am. Chem. Soc., Washington, DC, 1985.
1. Burcham, L. J., and Wachs, I. E., Catal. Today 49, 467 (1999).
2. Burcham, L. J., Wachs, I. E., and Briand, L. E., in “Proceedings of the
16th Meeting of the North American Catalysis Society,” p. 170. 1999.
3. Briand, L. E., Farneth, W. E., and Wachs, I. E., Catal. Today 62, 219
(2000).
4. Adkins, H., and Peterson, W. R., J. Am. Chem. Soc. 53, 1512 (1931).
5. Cheng, W. H., Chowdhry, U., Ferretti, A., Firment, L. E., Groff,
R. P., Machiels, C. J., Mc Carron, E. M., Ohuchi, F., Staley, R. H.,
and Sleight, A. W., in “Proceedings of the 2nd Symposium of IUCCP,”
pp. 165–181. Dept. Chemistry, Texas A & M Univ. Press, College Sta-
tion, 1984.
31. Jehng, J-M., Wachs, I. E., Clark, F. C., and Springman, M. C., J. Mol.
Catal. 81, 63 (1993).
32. Wachs, I. E., Catal. Today 27, 437 (1996).
33. Hardcastle, F. D., and Wachs, I. E., J. Phys. Chem. 95, 10763 (1991).
34. Hill, Jr., C. G., and Wilson III, J. H, J. Mol. Catal. 63, 65 (1990).
35. X-ray diffraction signals coincides with the following JCPDS-Inter-
national Centre for Diffraction Data files (1997): 23-0764 for Al2
(MoO4)3; 20-1468 for Zr(MoO4)2, 20-310 for Cr2(MoO4)3 and 31-330
6. Machiels, C. J., Cheng, W. H., Chowdhry, U., Farneth, W. E., Hong, F.,
Mc Carron, M. E., and Sleight, A. W., Appl. Catal. 25, 249 (1986).
7. Holstein, W. L., and Machiels, C. J., J. Catal. 162, 118 (1996).
8. Machiels, C. J., and Sleight, A. W., in “Proceedings, 4th International
Conference on the Chemistry and Uses of Molybdenum” (H. F. Barry
and P. C. H. Mitchell, Eds.), p. 411. Climax Molybdenum Co., Ann
Arbor, MI, 1982.
9. Gai, P. L., and Labun, P. A., J. Catal. 94, 79 (1985).
10. (a) Mann, R. S., and Hahn, K. W., J. Catal. 15, 329 (1969); (b) Ind.
Eng. Chem. Process Des. Develop. 9, 43 (1970).
for Ce8Mo12O49
.
36. Badlani, M., Master of science thesis, Lehigh University, Bethlehem,
PA, 2000.
37. Niemantsverdriet, J. W., in “Spectroscopy in Catalysis: An Introduc-
tion.” VCH, New York, 1995.
38. Tatiboue¨t, J. M., Appl. Catal. A: General 148, 213 (1997).
39. Wang, C.-B., Cai, Y., and Wachs, I. E., Langmuir 15, 1223 (1999).
40. Trifiro, F., Catal. Today 41, 21 (1998).
41. Klissurski, D., Pesheva, Y., Abadjieva, N., Mitov, I., Filkova, D., and
Petrov, L., Appl. Catal. 77, 55 (1991).
42. Niwa, M., Yamada, H., and Murakami, Y., J. Catal. 134, 331 (1992).
43. Yang, T-J., and Lunsford, J. H., J. Catal. 103, 55 (1987).
44. Louis, C., Tatiboue¨t, J-M., and Che, M., J. Catal. 109, 354 (1988).
45. Moro-Oka, Y., Morikawa, Y., and Ozaki, A., J. Catal. 7, 23 (1967).
46. Trifiro, F., and Pasquon, I., J. Catal. 12, (1968) 412.
47. Krylov, O. V., and Fokina, E. A., in “Proceedings of the 4th World
Congress on Catalysis,” Moscow, 1968, preprint No. 909.
48. Bielanski, A., and Haber, J., Catal. Rev.-Sci. Eng. 19, 1 (1979).
49. Fierro, J. L. G., and Garcia De La Banda, J. F., Catal. Rev.-Sci. Eng.
28, 265 (1986).
11. Ivanov, K. I., Litcheva, P., and Klissurski, D., Collect. Czech. Chem.
Commum. 57, 2539 (1992).
12. Klissurski, D. G., Rives, V., Pesheva, J., Mitov, I., and Abadzhjieva, N.,
Catal. Lett. 18, 265 (1993).
13. Arora, N., Deo, G., Wachs, I. E., and Hirt, A. M., J. Catal. 159, 1
(1996).
14. Weng, T., and Wolf, E. E., Appl. Catal. 96, 383 (1983).
15. Valente, N. G., Cadus, L. E., Gorriz, O. F., Arru´a, L. A., Rivarola,
J. B., Appl. Catal. 153, 119 (1997).
16. Hu, H., and Wachs, I. E., J. Phys. Chem. 99, 10911 (1995).
17. Yamashita, K., Naito, S., and Tamaru, K., J. Catal. 94, 353 (1985).
18. Yamada, H., Niwa, M., and Murakami, Y., Appl. Catal. 96, 113 (1993).
19. Wachs, I. E., and Madix, R. J., J. Catal. 53, 208 (1978).
20. Poulston, S., Rowbotham, E., Stone, P., Parlett, P., and Bowker, M.,
Catal. Lett. 52, 63 (1998).