Sulphur uptake and exchange, HDS activity and structure of sulfided, Al2O3 supported MoOx, PdMoOx and PtMoOx catalysts

Article abstract of DOI:10.1039/b009047f

A growing interest is observed in the combined HDS-hydrogenation activity of noble metal catalysts, connected with emission limitations of both sulfur and aromatics content in exhaust gases. The bulk composition, structure, and valence state of an alumina supported PdMoO_x catalyst were studied by XRD, HRTEM, energy dispersive X-ray analysis, XPS, and ion scattering spectroscopy before and after sulfiding. Structural information, sulfur uptake, and exchange data were compared with the activity of these catalysts in HDS of thiophene. The sulfur uptake by alumina supported Pd-containing MoO_x exceeded substantially that of the Pt-containing and of the nonpromoted molybdena. Molybdena formed amorphous layers on the surface and transforms into MoS₂ patches upon sulfidation of PdMo. Palladium in PdMo reacted with H₂S to yield Pd₄S and Pd₁₆S₇ particles. Strong sulfidation increased the ratio of Mo(IV) and Mo(V) in comparison with samples, reduced in hydrogen. Thiophene HDS activity decreased upon "mild" sulfidation of PdMo and PtMo, but increased at "strong" sulfidation.

Full text of DOI:10.1039/b009047f

View Article Online / Journal Homepage / Table of Contents for this issue
Sulfur uptake and exchange, HDS activity and structure of sulÐded,
Al O supported MoO , PdMoO and PtMoO catalysts
2 3
x
x
x
Zoltan Paal,a Tamas Koltai,a (the late) Karoly Matusek,a Jean-Marie Manoli,b Claude Potvin,b
Martin Muhler,c¤ Ute Wildc and Pal Tetenyi*a
a Institute of Isotope and Surface Chemistry, CRC, Hungarian Academy of Science, Budapest,
P. O. Box 77, H-1525 Hungary
b L aboratoire de Reactivite de Surface, UMR 7609, casier 178, Universite P. and M. Curie,
F-75252 Paris Cedex 05, France
c Fritz-Haber Institut der Max Planck Gesellschaft, Berlin, Faradayweg 4-6, D-14195 Germany
Received 10th November 2000, Accepted 9th February 2001
First published as an Advance Article on the web 19th March 2001
The bulk composition, structure and valence state of an alumina supported PdMoO catalyst have been
x
studied by X-ray di†raction (XRD), high resolution transmission electron microscopy (HRTEM), energy
dispersive X-ray (EDX) analysis, X-ray photoelectron spectroscopy (XPS) and ion scattering spectroscopy (ISS)
before and after sulÐding. Results are compared with data from a previous study of PtMoO . Temperature
x
programmed reduction (TPR) data indicate that both catalysts contain Pd and Pt in metallic form with high
metal dispersion. Pd S and Pd S are formed upon sulÐdation of PdMoO opposite to PtS formation in
4
16 7
x
PtMoO . The sulfur uptake by alumina supported MoO , PdMoO and PtMoO was determined by using
x
x
x
x
H 35S tracer and compared with XPS data on the sulfur content of the sulÐded catalysts. EM data indicate,
that MoS forms patches on the support surface. The mean sulfur to metal ratio in the catalysts, as calculated
from sulfur uptakes, measured by radiotracer experiments, is equal with that measured by XPS for the topmost
2
2
layers in the case of MoO , substantially lower in the case of PdMoO and somewhat lower in the case of
x
x
PtMoO . The extent of sulfur heteroexchange and the ratio of mobile sulfur is much higher in the case of
x
molybdena alone than in the case of metal doped catalysts.
alumina supported PdMoO , PtMoO and CoMoO cata-
1
Introduction
x
x
x
lysts with respect to the amount of their sulfur uptake in the
initial stage of sulÐdation.14
A growing interest is observed recently in the combined
hydrodesulfurization (HDS)Èhydrogenation (HYD) activity of
noble metal catalysts,1h5 connected with emission limitations
of both sulfur and aromatics content in exhaust gases. Their
low sulfur tolerance6 represents, however, certain difficulty for
the application of noble metals as catalysts in HDS reactions.
Recent studies indicate,7,8 that adding of a second component
improves the sulfur tolerance of the noble metal. A drastic
drop of HDS activity of alumina supported Ru-, Pd-, Ir- and
PtMoO catalysts was only observed at the initial stage of
the sulÐdation, followed by a period of constant (and in the
cases of Ru- and Pd-containing catalysts even increasing)
activity upon further addition of H S.9 It was of interest
therefore to study the sulfur uptake capacity of alumina sup-
ported MoO , Pd- and PtMoO catalysts and compare those
In order to determine the sulfur uptake capacity and the
extent of sulfur exchange of Pd- and Pt-promoted and non-
promoted MoO /Al O and for their comparison with
x
2 3
alumina supported CoMoO and NiMoO catalysts we
x
x
applied the isotope tracer method in a gas-Ñow circulation
system, as developed and applied for study of NiMoO 15 and
x
CoMoO 16 catalysts.
x
The present paper reports also on the bulk and surface
analysis of PdMoO /Al O samples before and after sul-
x
x
2 3
Ðdation, analogous to that reported by us previously on
PtMoO /Al O :10 X-ray di†raction (XRD), electron micros-
x
2 3
copy (EM), X-ray photoelectron spectroscopy (XPS) and ion
scattering spectroscopy (ISS) were applied. XPS experiments
have been carried out with MoO /Al O catalyst too. Sulfur
2
x
x
x
2 3
with their structure, surface composition and HDS activity.
Structure and surface composition data of PtMo were taken
from a previous study.10
concentration data obtained by XPS analysis were compared
with sulfur uptakes, determined by the isotope tracer method.
Structural information, sulfur uptake and exchange data were
compared with the activity of these catalysts in HDS of
thiophene.
Kabe et al. reported in a recent study5 a substantial
increase of HDS and HYD activity of Al O supported sul-
2
3
Ðded Pd and Pt catalysts at increasing metal loading and at a
constant (0.25) S/Pd (or S/Pt) atomic ratio. This ratio was
substantially lower than that found by the same method of
sulfur uptake measurements for CoMo/Al O 11 and those
found by TPR/TPS for Al O supported Pd and Pt.12,13 Our
previous study indicated no substantial di†erence between
2
Experimental
2
3
2.1 Catalyst preparation and pretreatment
2
3
MoO /Al O (denoted throughout the paper as Mo) was pre-
x
2 3
pared in the Chemicals and Fuels Engineering Department,
University of Utah. A commercial Al O (surface area of 180
¤ Present address: Ruhr-Universitat Bochum, Lehrstuhl fur tech-
nische Chemie, Bochum, D-44780 Germany.
2
3
m2 g~1) was impregnated with an aqueous solution of
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543 1535
This journal is ( The Owner Societies 2001
DOI: 10.1039/b009047f

View Article Online
ammonium paramolybdate: the sample of Al O was grad-
ually wetted with distilled water, until the pore volume was
2.2 Surface analysis
2
3
Two methods: X-ray photoelectron spectroscopy (XPS) and
ion scattering spectroscopy (ISS) were used for surface
analysis. The XPS and ISS techniques and evaluation have
been described earlier.10,19 ISS used He` ions of 2 kV under
quasi-static conditions in repeated scans.19 ISS spectra are
shown without intensity normalisation. O 1s, Mo 2p, Pd 3d, C
1s, S 2p and Al 2s lines were monitored in the pass energy
mode using a Leybold LHS 12 MCD ESCA apparatus for
XPS. The binding energy (BE) shift of 6È7 eV due to electro-
static charging has been corrected to the Al 2p peak in Al O
Ðlled, i.e. the incipient wetness (0.5 cm3 g~1 Al O ) was deter-
2
3
mined. The concentration of the ammonium paramolybdate
solution was calculated and made up to give 12 wt.% of
MoO (D0.33 g cm~3), equivalent to 0.136 g MoO in 0.5
3
3
cm3 of water. The wet sample was dried, then calcined at 773
K overnight. Pd- and PtMoO /Al O (made up to give 3
wt.% of Pd or Pt metal ] 12 wt.% of MoO , denoted further
as PdMo and PtMo) were also prepared by the pore-Ðlling
x
2 3
3
method. (The speciÐc metal loadings of the catalysts were
chosen with the intention to compare sulfur uptake and cata-
lytic data with results obtained in previous experiments. The
catalyst with higher atomic % of Pd was prepared because of
the lower dispersion of metallic Pd.) Condea Pural Ng
alumina (surface area 180 m2 g~1, D0.65 cm3 g~1 of incipient
wetness) was impregnated with aqueous ammonium hepta-
molybdate solution (D0.45 cm3 g~1 alumina), followed by a
subsequent impregnation with PdCl or H PtCl solution
2
3
(BE \ 74.7 eV). Literature values of atomic sensitivity factors
were applied for quantitative determination of surface
composition.20,21
2.3 X-ray di†raction
X-ray di†raction (XRD) studies were performed on a Siemens
D500 automatic di†ractometer.10 The Ka wavelength of
copper was selected by a graphite di†racted beam monochro-
mator, the di†ractometer was adjusted for reÑection geometry.
The X-ray source was operated at 30 kV and 20 mA and
scanned at a rate of 0.15 degrees min~1 (in 2h). All data were
collected on a computer and analysed with Socabim Di†rac-
AT software. The identiÐcation of the di†erent phases was
made using the JPCDS library.
2
2
6
(D0.2 cm3 g~1 alumina). The samples were dried at room
temperature, calcined at 723 K and reduced at 673 K. Their
surface area was determined by N and CO adsorption
2
2
(BET). TPR and H /O titration of the samples were carried
2
2
out for the determination of their H -uptake and metal dis-
2
persion respectively. The ratio of the fraction of the alumina
surface not covered with MoO was determined by the
x
method of comparative chemisorption17 of CO on the Mo
2
sample and on the support. The bulk chemical composition of
2.4 Electron microscopy
the samples was determined by prompt gamma activation
analysis (PGAA) in the Budapest Neutron Center.18 Charac-
teristic data of the catalysts are collected in Table 1.
One of the following pretreatment procedures was applied
before tests:
The samples were ground, dispersed in ethanol in an ultra-
sonic bath and deposited on a thin carbon Ðlm supported on
a standard copper grid. HRTEM studies were performed on a
JEOL apparatus (JEM 100 CXII) equipped with a top entry
device and operating at 100 kV. The camera length was cali-
brated with polycrystalline gold powder. Energy dispersive
analysis of the emitted X-rays (EDX) was carried out on
selected areas.
Reduction. Treatment in H stream (30 NTP cm3 min~1)
2
for 120 min at 673 K (NTP \ normal temperature and
pressure).
2.5 Catalytic tests
““MildÏÏ sulÐdation. Heating in H stream as previously, fol-
2
lowed by injection of 6 H S pulses in 5 min intervals. The
A combined glass microcatalytic pulse reactor was used both
for in situ determination of thiophene conversion on reduced
and mildly sulÐded samples and for measurement of sulfur
uptake by radiotracer experiments.22
5 ] 10~3 cm3 of liquid thiophene pulses were injected into
hydrogen gas, Ñowing (30 NTP cm3 min~1) through a tubular
reactor (100 mm long, 3.5 mm in diameter). The temperature
of the reactor was kept constant and measured by a thermo-
couple at the external surface of the reactor. The GLC column
of n-octane/Porasil for analysis of the outÑow gases was con-
nected directly downstream to the reactor. The column outlet
stream was split into two parts, one of them being led to the
FID in order to analyse the products of thiophene HDS, the
other to a radioactivity detector (see section 2.6).
2
volume of one pulse was before catalytic tests: 10~4 cm3 at
298 K and 20 kPa (4.8 ] 1017 molecules of H S). The integral
2
sulfur uptake by this process was, as determined previously,13
D1.01 ] 1017, 1.2 ] 1017 and 0.9 ] 1017 S mg~1 for the Mo-,
Pd- and Pt-containing samples respectively. The volume of
one pulse before XPS, ISS, XRD and EM tests was 10~3 cm3
at 298 K and 60 kPa (1.44 ] 1019 molecules of H S).
2
““StrongÏÏ sulÐdation. Heating in H Ñow (60 NTP cm3
2
min~1) up to 673 K, Ñushing with H S/H (60 cm3 min~1, 6
2
2
vol% of H S), then Ñushing and cooling in He to 300 K.
Catalytic tests were carried out with samples of 10~2 g.
2
Samples of 0.15 g were taken for XPS, ISS, XRD and EM
tests, described in detail earlier.10 These measurements were
carried out ex situ, while catalytic tests with reduced and
mildly sulÐded samples were carried out in situ immediately
after the pretreatment process, as opposed to ex situ testing of
the strongly sulÐded samples.
Experiments with thiophene (““stronglyÏÏ sulÐded samples)
and with cyclohexene were performed in an identical micro-
catalytic apparatus without splitting the stream of the outÑow
gases. In these experiments the GCL column was packed with
18% squalane on Chromosorb P.
Table 1 Characteristic data of the samples
Catalyst
n
/1020 atoms g~1
n a
n b
Surf. area/m2 g~1
Dispersione Pd or Pt(%)
Free Al O surface (%)f
Mo
M
H
2 3
PdMo
PtMo
Mo
4.28
4.29
5.30
1.46
0.90
È
5.18
4.75
3.87
133c 129d
142c 141d
101 100
54 ^ 7
70 ^ 8
È
?
?
43.5 ^ 5
a Pd or Pt. b Hydrogen uptake up to 673 K as measured by TPR. c As measured by N adsorption. d As measured by CO adsorption. e As
2
2
determined by O /H titration. f As determined by CO chemisorption (see Appendix).
2
2
2
1536
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543

View Article Online
Thiophene HDS and cyclohexene conversion tests were
carried out directly at the end of the pretreatment of the
sample, with the exception of ““stronglyÏÏ sulÐded samples,
being cooled after sulÐdation (to 300 K), then placed into the
reactor in He atmosphere and heated up to the temperature of
in Run 3 (Y ) indicated the extent of sulfur heteroexchange at
3
equilibrium Sexc (denoted further Sexc) and was calculated by
eq
the expression:
Y \ Sexc \ (I /I )m
3 so so
(4)
3
the respective tests in H .
Another set of experiments (Run 4) was the sulÐdation of a
fresh sample with nonradioactive H S/H under conditions
2
Thiophene and cyclohexene conversion tests were carried
2
2
out at 673 and 473 K, respectively. The degree of conversion
was determined by unreacted thiophene or cyclohexene per
pulse. The Ñow rate of the carrier gas stream was taken to be
equivalent to the Ñow rate of the pulses,23 and the number of
the converted molecules was calculated as percent of the Ñow
rate. The tests were repeated several times. The mean arith-
metic error was in the range of ^1È8%. The absence of di†u-
sion e†ect was veriÐed experimentally by parallel
measurements with two samples of di†erent catalyst mass: the
conversions were similar if measured at equal contact time.
similar to those of Run 1. The sample was exposedÈafter
evacuationÈto the same radioactive mixture as in Run 1. The
yield in this Run, Y , calculated by eqn. (1) should be equal to
4
the sum of reversible uptake and of sulfur heteroexchange at
adsorption/desorption and exchange equilibrium:
Y \ Y ] Y \ Srev ] Sexc
(5)
4
2
3
Studies with a CoMo catalyst veriÐed the method, as the
directly determined Y was equal to the sum of Y and Y
4
2
3
determined separately.16 It was found also, that Y (i.e. the
3
extent of exchange) was constant at P(H S) [ 4 kPa.
2
2.6 Sulfur uptake and exchange measurements
3
Results and discussion
The amounts of sulfur uptake by the catalyst and the extent of
sulfur heteroexchange between the sulÐded catalyst and the
gas phase were determined by gas phase radioactivity mea-
3.1 Surface analysis
Ion scattering spectroscopy (ISS) made it possible to obtain
information about the surface layer of PdMo. As opposed to
that of Pt,10 the ISS peak corresponding to palladium could
not be separated from that of Mo because of their closer
atomic masses. Hardly any decrease of the combined Pd/Mo
peak can be seen during repeated scans (Fig. 1). Integration,
however, indicated a drop of Pd/Mo intensity by up to 10%
between 3 and 29 scans. This means removal of some Mo and
Pd from the outermost layer by the sputtering e†ect of the
He` ions. The di†erence spectrum in Fig. 1 conÐrms this
minor change, together with the decrease of S and increase of
Al and O in the surface layer, in agreement with the XPS data
(see below).
The compositions measured from XPS line intensities of
Mo and PdMo are shown in Table 2 (omitting the major
component oxygen whose abundance was 57È60% as well as
the few percent carbon impurity). Table 2 contains for com-
parison analogous data for PtMo published previously10 and
recalculated by the method applied here for Mo and PdMo.
In situ reduction decreased the surface molybdenum concen-
tration in PdMo and so did both ““mildÏÏ and ““strongÏÏ sul-
Ðdation. It is remarkable that the Pd/Mo ratio was several
times lower in these surface and near to surface layers than
the bulk ratio: 0.02È0.08 vs. the bulk ratio of 0.34. Reduction
did not change the amount of Pd but sulÐdation more than
doubled the surface Pd percentage. The uptake of sulfur at
““strongÏÏ sulÐdation was considerable. The binding energy of
Pd 3d line was higher by a few tenths of eV than the literature
value for metallic Pd. Reduction decreased it slightly (D0.2È
0.3 eV) and sulÐdation shifted it back to higher values. Deter-
mination of the line position with higher accuracy was not
surements of H 35S. Data on sulfur uptakes at ““mildÏÏ sul-
2
Ðdation were published previously14 as determined from the
retained radioactivity of H 35S pulses at contacting with the
2
samples.
Total sulÐdation and sulfur exchange measurements were
carried out in a gas-Ñow circulation system. Detailed descrip-
tion of the apparatus and of the method is given in ref. 16.
Mixtures of H S (both 35S labelled and non-radioactive) with
2
H
of 53.33 kPa total pressure were prepared with deÐnite
2
(10 : 90 and 50 : 50) H S/H ratios, and introduced into a
2
2
previously evacuated system. The mixture circulated through
a tube, containing the catalyst sample (D24 mg) evacuated
previously at the temperature of the measurement. The radio-
activity of the H S/H mixture was measured in 5 min inter-
vals, using an Eu doped CaF scintillation crystal, which
formed one of the Ñat surfaces of a disc like cavity in a Ñow-
2
2
2
through radioactivity detector.24 The radioactivity measure-
ments were continued up to a constant value of the gas phase
counting rate. This took approximately 60 min. The treatment
was continued for another 60 min. No further changes were
observed in gas phase radioactivity during this time.
The amount of the total uptake of sulfur was determined in
Run 1 as the decrease of gas phase sulfur (Y ) calculated from
1
the values of initial and minimal counting rates in the gas
phase (I and I , respectively) by the expression:
so
1
Y \ St \ m (1 [ I /I )
so 1 so
(1)
1
where m is the number of H S molecules in the initial
so
2
2
H 35S/H mixture. St is the total amount of sulfur uptake cal-
2
culated by eqn. (1) and consists of both reversibly attached
and irreversibly bonded sulfur, viz.
St \ Srev ] Sirr
(2)
After treatment of the sample in vacuum (2 h) for removal of
reversibly bonded sulfur, the treatment of the sample was
repeated with the original H 35S/H mixture (Run 2). The
2
2
decrease of gas phase sulphur (Y ) in this case indicated the
2
amount of reversibly bonded sulfur only:
Y \ Srev \ m (1 [ I /I )
so 2 so
(3)
2
and Sirr was calculated as Y ÈY . The exchange between gas
1
2
and solid phase 35S did not a†ect the gas phase radioactivity
as 35S/(32S ] 35S) was equal in both phases.
After evacuation (as before) the sample was exposed to a
nonradioactive H S/H mixture (Run 3) of identical H S/H
Fig. 1 Ion scattering spectroscopy (ISS) of ““stronglyÏÏ sulÐded
Pd/MoO /Al O sample. Several subsequent scans provide a semi-
2
2
2
2
x
2 3
ratio with Runs 1 and 2. The treatment was continued until a
quantitative depth proÐling, as conÐrmed also by the di†erence spec-
maximum constant counting rate (I ) was reached. The yield
trum of scan 29 [ scan 3 (dotted line).
3
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543
1537

View Article Online
Table 2 Surface composition (XPS) of the catalysts after various treatments
Composition (atom%)a
Mo Pd or Ptb
Sample
Treatment
Al
S
PdMo
PtMo
Mo
None (as is)
None (as is)
None (as is)
7.45
8.10
4.0
0.15
1.30
È
92.4
90.6
96.0
È
È
È
PdMo
PtMo
Mo
Reduction in situ, 673 K
Reduction in situ, 673 K
Reduction in situ, 673 K
5.3
7.3
4.3
0.15
1.20
È
94.55
91.5
95.7
È
È
È
PdMo
PtMo
Mo
““MildÏÏ sulÐdation
È
““MildÏÏ sulÐdation
3.6
È
4.1
0.3
È
È
92.5
È
91.8
3.6
È
4.1
PdMo
PtMo
Mo
““StrongÏÏ sulÐdation
““StrongÏÏ sulÐdation
““StrongÏÏ sulÐdation
6.2
6.9
3.9
0.3
1.0
È
70.0
81.8
92.2
23.5
10.3
3.9
PdMo
PtMo
““StrongÏÏ sulf. ] ISS
““StrongÏÏ sulf. ] ISS
6.1
6.1
0.4
1.0
78.2
87.0
15.3
5.9
a Calculated from the intensities of Mo 3p, Pd 3d, S 2p and Al 2s lines (100%). Intensities of oxygen (the major component) and a few percent of
carbon have not been considered. b Experiments reported for PtMo are taken from ref. 10 and presented here for comparison.
possible due to the overall electrostatic charging (D5È7 eV)
involving also a possible di†erential charging25 arising on Pd
particles of di†erent size (see below, under Electron
microscopy). Reduction and sulÐdation did not change the
amount of Mo in surface layers of the Mo samples and a
moderate sulfur uptake was observed upon sulÐdation. The
ratio of sulfur to molybdenum, as determined by XPS, showed
close values after ““mildÏÏ and ““strongÏÏ sulÐdation.
(Table 3). The results for PdMo are not far from those
published earlier10 for PtMo catalysts. Comparison of valence
state data with those of the reduced in situ Mo sample indi-
cates a higher average degree of Mo reduction (e/Mo \ 1.12)
than those for PdMo and PtMo samples (Table 3). This seems
to contradict to some extent the generally accepted view on
the reduction inducing role of Group 8È10 metals as an expla-
nation of their promoting e†ect on molybdena.
Sulfur existed in two forms: storing the sample in air gave
rise to a minor sulfate component in addition to the major
chemical component: sulÐde.19 The sputtering e†ect of He`
ions during ISS sputtered away relatively more S in the sulÐde
state, indicating that these more sensitive speciesÈlikely
It is remarkable that whereas the e/Mo values were lower
after weak sulÐdation of PdMo and Mo than the respective
values upon in situ hydrogen treatment (presumably due to
exposing the sulÐded sample to air during its transfer), the
degree of Mo reduction after ““strongÏÏ sulÐdation was 1.2È1.3
for all three catalysts; higher than after their in situ reduction.
This indicated a substantially stronger reduction with H S/H
MoS (see below, under Electron microscopy)Èwere situated
2
in surface positions (Fig. 2).
2
2
The decomposition of the Mo 3d plus S 2s region made it
possible to estimate the amount of molybdenum components
in di†erent valence states.10 The best Ðt was obtained by using
three doublets corresponding to Mo(VI), Mo(V) and Mo(IV),
respectively. Fig. 3 illustrates the Ðtting procedure on the
example of the ““stronglyÏÏ sulÐded catalyst. The untreated
sample contained mostly Mo(VI) the amount of which
dropped after reduction and upon both sulÐdation treatments
than with H in agreement with a former observation.26 A
2
lower degree of Mo reduction was measured in strongly sul-
Ðded PdMo after ISS. Consequently, the Mo in topmost
layers was reduced to a higher extent than in deeper layers.
This is di†erent from PtMo samples with equal (D1.2) e/Mo
values before and after ISS of strongly sulÐded samples. More
Mo(IV)Èpresumably as MoS Èwas present on the surface of
2
the Pd doped sample and could thus be removed by He` ion
sputtering.
3.2 X-Ray di†raction
XRD detected Al O as the only crystalline phase in the Mo
2
3
sample.9 Calculations based on CO chemisorption data,17
2
indicated that about 80% of molybdena was present as a
Fig. 2 The S 2p peaks of ““stronglyÏÏ sulÐded Pd/MoO /Al O
Fig. 3 Line Ðtting to the Mo 3d plus S 2s region of ““stronglyÏÏ sul-
x
2 3
without any treatment (lower curve) and after ISS (as shown in Fig. 1),
together with lines Ðtted to the sulÐde and sulfate regions.
Ðded Pd/MoO /Al O sample. Similar Ðtting procedures were also
x
2 3
carried out with all samples (cf. Table 3).
1538
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543

View Article Online
Table 3 Line Ðtting of the Mo 3d regiona
Abundance, rel. (%)a
Mo(VI)
Sample
Treatment
Mo(V)
Mo(IV)
e/Mob
PdMo
PtMod
Mo
H c
2
77
95
71
23
5
29
È
È
È
0.23
0.05
0.29
None
H c
2
PdMo
PtMod
Mo
Reduction in situe
Reduction in situe
Reduction in situe
32
45
26
34
30
36
34
25
38
1.02
0.80
1.12
PdMo
Mo
““MildÏÏ sulÐdation
““MildÏÏ sulÐdation
56
52
27
32
17
16
0.61
0.64
PdMo
PtMod
Mo
““StrongÏÏ sulÐdation
““StrongÏÏ sulÐdation
““StrongÏÏ sulÐdation
36
36
3
10
9
22
54
56
55
1.18
1.15
1.3
PdMo
PtMod
““StrongÏÏ sulf. ] ISS
““StrongÏÏ sulf. ] ISS
48
31
15
16
37
53
0.89
1.22
a Fitting parameters: Ðxed Gaussian/Lorentzian ratio 0.3, fwhm \ 2.8 eV, *BE for Mo 3d [ 3d \ 3.1eV, max. BE values were found (self-
3@2
5@2
consistently) by the Ðtting program. The intensities of S 2s lines (in sulfate and sulÐde form) were considered (fwhm \ 3.8 eV). b e/Mo (the mean
degree of reduction): e/Mo \ 6 [ (&q ] Y ); Y : Mo valence state (IV, or V or VI), q : relative ratio of Mo(Y ). c Ex situ. d Data for PtMo from
Y
Y
ref. 10, included here for comparison. e At 673 K.
monomolecular layer on the alumina surface (see the
Appendix). This is in agreement with literature data, indicat-
ing that low (\20%) amounts of MoO form a nearly mono-
x
molecular, amorphous layer on the alumina surface.27
Fig. 4 shows the di†raction patterns of our PdMo catalyst
after three di†erent treatments. Di†raction lines for Al O and
2
3
Pd28 are the only ones appearing in a reduced catalyst, like in
previous studies in alumina supported PdMo and PtMo.9,10
A ““mildÏÏ sulÐdation resulted in the appearance of several new
lines. Panel b shows that they correspond to the pattern of
Pd S.29 At the same time, some residual metallic Pd still
4
remained (arrows). Both of these phases disappeared com-
pletely after ““strongÏÏ sulÐdation and the di†raction pattern
indicated Pd S as the main form of Pd.30 The mean particle
16 7
size of Pd applying the Scherrer formula was about 30È35 nm.
The crystallites of Pd S (lower panel) were slightly larger and
4
the size of Pd S particles (middle panel) reached 50 nm. The
16 7
behavior of palladium was di†erent from that observed in the
case of PtMo,10 containingÈeven after ““strongÏÏ sulÐdationÈ
mainly di†raction lines, corresponding to metallic Pt and less
intense ones of PtS. The Al O phase remained unchanged.
2
3
No lines corresponding to crystalline MoS could be
2
observed, although the increased background at 2h D 30È35¡
may indicate its presence. Its crystallinity was, however, prac-
tically zero.
3.3 Electron microscopy
Palladium appeared in two distinct forms: as very small crys-
tallites and as bigger aggregates of 80 to 300 nm. This latter
value is in fairly good agreement with XRD results. The EDX
analysis of sulÐded Mo and PdMo catalysts showed a rather
constant Mo/Al ratio (0.05È0.07) indicating a uniform dis-
tribution of molybdenum on the alumina support. This
analysis conÐrmed the uneven distribution of palladium par-
ticles in PdMo. The S/Pd ratios were typically between 0.13
and 0.35 on the weakly sulÐded and 0.5 to 1.5 on the strongly
sulÐded PdMo. The actual values showed large variation,
depending on the selected area and related also to the physical
appearance of Pd: more sulfur was observed in the vicinity of
larger Pd particles and less sulfur when the selected area con-
tained mostly well-dispersed metal particles. The correspond-
Fig. 4 X-Ray di†ractograms of Pd/MoO /Al O samples after dif-
x
2 3
ferent treatments. Upper panel: (a-1) alumina support, JPCDS
29-063; (a-2) ““as receivedÏÏ reduced material; arrows denote Pd metal
reÑections (JPCDS 05-681). Middle panel: (b) ““mildÏÏ sulÐdation;
ing S/Pd atomic ratios should be 0.25 for Pd S and 0.44 for
arrows denote Pd metal reÑections, lines represent reÑections of Pd S,
4
Pd S . This indicates excess sulfur after ““strongÏÏ sulÐdation.
16 7
4
(JPCDS 10-335). Lower panel: (c) ““strongÏÏ sulÐdation; lines represent
Its chemical form cannot be determined with certainty: one
reÑections of Pd
S
(JPCDS 30-884).
16 7
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543
1539

View Article Online
indication may be that on areas with high S/Pd ratios the
S/Mo ratios were low and vice versa. The appearance of ele-
mentary sulfur has been reported also on Co/MoO /Al O
x
2 3
catalysts.31 The S/Mo ratios of Pd-free areasÈin both mildly
and strongly sulÐded PdMo samplesÈwere equal to 1.2È1.3,
as indicated by EDX. This agrees well with the XPS data of
the Mo samples.
Fig. 5 shows an electron micrograph of a mildly sulÐded
PdMo catalyst. The presence of a large particle permitted us
to determine also selected area electron di†raction (SAED, see
inset). This corresponds to the crystal structure of Pd S.
4
Other, smaller Pd containing particles (either metal or sulÐde)
cannot be distinguished. An analogous micrograph of a
strongly sulÐded Pd is seen in Fig. 6a. Dark spots of di†erent
size might correspond to Pd containing particles: the di†rac-
tion pattern of the large grain in the middle shows the typical
di†raction pattern of Pd
guishable Pd particles (Fig. 6b) showed the characteristic
S
(inset). An area without distin-
16 7
layered structure of MoS . They represented patch-like struc-
2
tures with a thickness of 2È4 molecular layers.
3.4 Sulfur uptakes and exchange
Data on the amounts of sulfur uptake and the extent of sulfur
exchange are collected in Table 4. Reversible uptakes by
PdMo and PtMo were calculated at 4 kPa P(H S) as Y ÈY in
2
4
3
agreement with eqn. (5). The sum of Y and Y Èas separately
2
3
measured on PtMo at D27 kPa P(H S)Ègave 7.96 ] 1017 S
2
mg~1 for Y , whereas the value was 7.92 ] 1017 S mg~1 as
4
measured directly. The near to equal value of the directly
determined Y (6.87 ] 1017 S mg~1) with the sum of Y ] Y
4
2
3
(6.7 ] 1017 S mg~1) is also seen in the case of the Mo sample
Fig. 6 (a) Electron micrograph of
a
strongly sulÐded
at D4 kPa P(H S). These equalities indicate that all Y -values
2
Pd/MoO /Al O sample. The SAED (inset) taken from the larger
were obtained in equilibrium and eqn. (5) was valid.
x
2 3
dark spot in the middle corresponds to the di†raction pattern of
Pd S . A layered structure around this larger particle (likely MoS )
It is seen also from Table 4, that irreversible sulfur uptakes
16 7
2
do not depend on P(H S) above 4 kPa. This is in agreement
was seen on the negative (not reproducible on the positive picture). (b)
Another area of the same strongly sulÐded Pd/MoO /Al O sample
2
x
2 3
without noticeable Pd containing particles. Arrows denote a broken
layered structure, likely MoS which does not form a contiguous
2
overlayer.
with observation reported previously15,16 with respect to
NiMo and CoMo catalysts.
The degrees of sulÐdation are 0.73, 0.58 and 0.47 for PdMo,
PtMo and Mo, respectively, based on irreversible uptakes
observed at D27 kPa P(H S). Taking into account the sulfur
2
uptakes at stoichiometrically full sulÐdation (Sf) values, the
St/Sf ratio surpasses 1 (assuming MoS and PdS or PtS for Sf).
2
This indicates, that the reversibly bonded sulfur (or a part of
it) does not Ðt fully to this stoichiometry.
It is remarkable, that the Sexc values are practically equal at
4 and 27 kPa H S partial pressure, similar to that observed
2
for NiMo and CoMo.15,16 In contrast to the lower Sirr
Fig. 5 Electron micrograph of a mildly sulÐded Pd/MoO /Al O
uptakes by Mo, the Sexc values (and the Sexc/Sirr ratios) are
signiÐcantly higher, than in the case of PdMo and PtMo. This
indicates a much higher mobility of sulfur on the non-
promoted Mo in comparison with the noble metal promoted
x
2 3
sample. The SAED (inset) taken from the larger dark spot on the
left-hand side corresponds to the di†raction pattern of Pd S. Smaller
4
darker spots in the middle of the picture may represent partly sulÐded
Pd of Ðner dispersion (EDX showed positive but very scattering Pd
values).
MoO .
x
Table 4 Amounts of sulfur uptake and extent of S exchange (1017 mol mg~1) at 673 Ka
St(Y )
Srev(Y )
Sirrb
Sexc(Y )
Srev ] Sexc(Y )
1
2
3
4
Catalyst
I
II
I
II
I
II
I
II
I
II
PdMo
PtMo
Mo
13.9
10.5
8.8
15.0
12.6
12.1
5.1c
5.3c
4.2
7.7
7.1
7.2
8.9
5.2
4.6
7.3
5.5
5.0
1.2
0.9
2.5
1.4
0.8
2.6
6.3
6.1
6.9
È
7.9
9.7
a P(H S) ] P(H ) \ D 53 kPa. H S partial pressure; I D 4 kPa, II D 27 kPa. Catalyst mass: 24 mg. b Calculated as Y ÈY . c Calculated as
2
2
2
1
2
Y ÈY .
4
3
1540
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543

View Article Online
Table 5 Comparison of sulfur uptakes (in Sirr/&M unitsa) as mea-
derived from Tables 2 and 4. Data in Table 5 indicate that the
S/&M total is signiÐcantly higher in the case of PdMo than
that of PtMo, measured either by tracer, or by XPS experi-
ments. The S/Mo values measured by radiotracer and by
XPS, are in good agreement in the case of the Mo sample, in
sured by radiotracer and XPS
XPS
35S
Catalyst
Total
““MildÏÏ
““StrongÏÏ
After ISS
which the near (80%) to monomolecular layer of MoO is
x
transformed upon sulÐdation into MoS in part only. This is
2
PdMo
PtMo
Mo
1.41
1.03
0.91
0.92
È
1.00
3.62
1.39
1.00
2.51
0.83
È
in agreement also with the D50% of Mo(IV) valence state in
the strongly sulÐded Mo sample (Table 3). The Sirr/&M values,
measured by XPS, are higher in the case of the strongly sul-
Ðded PdMo and PtMo samples, than those calculated from
the radiotracer data. The Sirr/&M values decrease signiÐcantly
(30È40%) upon sputtering with He` at ISS: by 30È40%. It
follows from this that a substantially higher amount of sulfur
has been taken up by surface layers than in subsurface posi-
tions. It is remarkable, that in the case of Pt/Mo the Sirr/&M
value is lower after ISS than the mean value, calculated from
isotope tracer data, whereas the value for PdMo remains still
higher. It should be noted that the marked di†erences between
PdMo and PtMo cannot be explained with the substantially
lower surface concentration of Pd: the S/Mo ratios are also
much higher in the case of PdMo.
To sum up: it can be stated, that the amounts of sulfur
uptake (expressed in S/&M values) are nearly equal in the case
of nonpromoted Mo, calculated either from tracer or from
XPS data. The S uptakes, calculated by both methods, are
substantially higher by PdMo, than those by PtMo, but the
S/&M values are lower when calculated from the 35S tracer
measurements, than those observed by XPS. This can be
explained by the presumably much higher concentration of
sulfur and molybdenum in the surface and topmost layers.
a Sirr/(Mo ] Pd) or Sirr/(Mo ] Pt).
3.5 Comparison of S uptake data observed by XPS and
isotope tracer
Both XPS and isotope tracer experiments indicate substantial
uptakes of sulfur. Sulfur uptake data [expressed in S to
(Mo ] Pd) or to (Mo ] Pt) ratio, denoted as S/&M] for com-
parison of the two methods are collected in Table 5, as
Table
6
Thiophene HDS conversion (1017 mol mg~1 s~1) at
T \ 673 K; mass of catalyst: 0.1 g
In situ
reduced
Mildly
sulÐded
Strongly
sulÐded
Catalyst
PdMo
PtMo
Mo
CoMo32
NiMo-115,34
NiMo-215,34
NiW22,33
2.41
6.30
1.01
0.30
1.05
2.70
0.23
2.16
1.99
2.09
1.49
1.66
3.25
0.45
2.64
3.23
2.27
È
2.40
3.89
3.23
3.6 Thiophene HDS and cyclohexene conversion
Table 6 contains HDS conversion data on PdMo, PtMo and
Mo catalysts. There are collected also some data from our
previous studies with NiMo, NiW and CoMo, supported on
alumina.15,22,32h34. Data in Table 6 indicate an outstand
ingly high activity of PdMo, PtMo and NiMo-1 in reduced
form, much higher than that of the other catalysts. ““MildÏÏ
sulÐdation of PtMo results in a decrease of 78% in HDS
activity, practically no activity change is observed with PdMo
and a substantial increase is seen with Mo similiar to all other
samples in Table 1. This is in agreement with our previous
observation9 with Ru, Pd, Ir and Pt containing MoO /Al O
x
2 3
catalysts, indicating that even one H S pulse resulted in a 37,
2
67 and 51% decrease in the HDS activity of the Pt, Ru and Ir
containing Mo/Al O , respectively, followed with an activity
2
3
decrease of much lower degree upon subsequent injections of
H S pulses on PtMo and IrMo, whereas a small increase of
2
activity of PdMo and RuMo was observed with increasing
number of H S pulses. Alumina supported NiMo, CoMo and
2
NiW are more active, however, in the sulÐded than in the
reduced form. This is presumably due to the di†erence in the
physical state of the metals in reduced samples: a high ratio of
Fig. 7 HDS selectivities at conversion on catalysts of di†erent
pretreatment: ratio of hydrogenated to non-hydrogenated products.
Table 7 Conversion of cyclohexene (CH/) and product selectivities (%)a
Reduced sample
SulÐded sample/Proc 4/
Selectivity of
Converted
Selectivity of
Converted
Catalyst
CH/
CH
B
CH/
CH
B
PdMob
PtMob
Moc
85.65
100
0.46
95.87
99.83
0.0
4.13
0.17
100
100
27.15
21.15
1.74
99.0
96.3
100
89.7
1.00
2.9
0.0
Mod
0.47
0.0
4.45
10.3
a CH and B: cyclohexane and benzene respectively. Cyclohexene pulse: 2.9 ] 1018 mol. Flow rate of H : 60 NTP cm3 min~1. Mass of catalyst:
2
5 mg. b T : 473 K. c T : 493 K. d T : 508 K.
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543
1541

View Article Online
Pd and Pt is present on the catalyst surface in metallic form
(see Table 1). The presence of metallic nickel is indicated on
the surface of the NiMo-2 sample: this follows35,36 from the
fact, that this catalyst (and by our observation37 both Pd and
PtMo) was active in cyclohexane dehydrogenation under 573
K15 in contrast with our observation in the case of the
NiMo-1,15 CoMo37 and NiW37 samples. It should be borne
in mind that some earlier evidence9 indicated that Group
8È10 metals and molybdena acted separately at the initial
stage of sulÐdation and, presumably, in reduced form. Pd and
Pt metals act presumably as CÈS bond breaking catalysts, as
activation of thiophene occurs on these metals in accordance
with the mechanism proposed by the Yermakov group.38 This
could be presumably the caseÈin partÈwith the NiMo-2
sample, too; this explains its high HDS activity in reduced
form. The HDS activity of the metals drastically decreases (or
as measured by XPS, in the case of PdMo, and somewhat
higher in the case of PtMo. This follows from the much higher
concentration of sulfur on (in) the surface layer(s).
Sulfur exchange data indicate a substantially higher mobil-
ity of sulfur on nonpromoted Mo in comparison with that on
PdMo and PtMo. Thiophene HDS activity decreases upon
““mildÏÏ sulÐdation of PdMo and PtMo, but increases at
““strongÏÏ sulÐdation. An explanation of this phenomenon is
given, based on the observed previously separate catalytic
action of Group 8È10 metal and Mo/Al O in reduced form
2
3
and at the initial stage of sulÐdation.
Acknowledgements
The study was supported by the Hungarian National
Research Fund (OTKA 017051). The authors express their
gratitude to Professor F. E. Massoth (Department of Chemi-
cal and Fuels Engineering, University of Utah) for kindly sup-
plying the Mo/Al O sample, and to Professor G. Molnar and
disappears) upon exposure to H S (and/or thiophene) due to
2
poisoning. On the other hand, however, the formation of Mo-
and M(8-10)-sulÐdes [or MÈMoÈS sites] could cause a change
in the HDS mechanism: anion vacancies, formed at expense
of sulfur removal, take over the role of active sites,39h42 and
compensate to some extent in the case of PtMo, and over-
compensate, in case of the other catalysts, the activity loss due
to sulfur poisoning of the metallic phase. These vacancies form
2
3
Mr Zs. Revay (Institute of Isotope and Surface Chemistry,
Budapest) for carrying out the Prompt Gamma Activation
Analysis of the catalysts. The help of P. Beaunier and M.
Lavergne for TEM and EDX measurements is greatly appre-
ciated.
reportedly on the edges of MoS slabs and their formation is
2
promoted by added Co or Ni.43 Fig. 6b shows that the patchy
MoS entities provided many edge-sites where atomically dis-
Appendix
2
perse Pd could also create such active vacancies.
The characteristic di†erences in product selectivities of
HDS (Fig. 7) between Mo and Pd or PtMo verify this expla-
nation: butane ] tetrahydrothiophene formation selectivities
decreased signiÐcantly with sulÐdation of PdMo and PtMo,
whereas no marked di†erences are seen with Mo and with
NiMo.15 The e†ect of H S on the ratio Bu//Bu/ seems to
MoO coverage of Al O in the Mo/Al O sample
x
2
3
2 3
The amounts of CO chemisorbed on the catalyst (n ) and on
2
cat
the support (n ) were determined as in ref. 44, and were
supp
calculated both from adsorption and desorption data:
n
/mol g~1
supp
n
/mol g~1
cat
2
2
1
corroborate this explanation: its deÐnite decrease (3 ] 2.5) is
observed upon sulÐdation of Pd and Pt-containing catalysts.
As it was explained in detail,9 the blocking of sites for hydro-
genation (i.e. of metallic ones) reduces the Bu//Bu/ ratio.
Adsorption
Desorption
Mean
3.33 ] 1019
3.81 ] 1019
3.56 ] 1019
1.09 ] 1019
1.59 ] 1019
1.34 ] 1019
2
1
The high di†erences in HDS activities between Pd and Pt-
containing non-sulÐded samples might be a consequence of
the higher Pt dispersion, as seen from Table 1, whereas the
much higher sulfur tolerance of PdMo is due to the much
higher Mo/Pd ratio in comparison with Mo/Pt in surface
layers.
The fraction of uncovered alumina surface ( f ) was calculated
by the formula:45
f \ n /(n
] q )
cat supp
Al
where q , the weight fraction of the support in the catalyst
Cyclohexene conversion data are indicated in Table 7. A
high hydrogenation activity is observed on Pd and Pt pro-
moted samples, whereas the non-promoted Mo is a weak
dehydrogenating contact. SulÐdation resulted in a strong
decrease of activity of the Pd and Pt promoted samples, a
substantial hydrogenation activity still remained however in
part presumably due to the increased hydrogenation activity
of molybdena. No deÐnite conclusion could be drawn with
Al
sample, was 0.873. Then f \ 0.431 ^ 0.05, and the surface
coverage with MoO was 1 [ f \ 0.57 ^ 0.04. As the BET
x
area of the Mo sample was 101 ] 1018 nm2 (Table 1), the
area, occupied by MoO should be 57.6 ] 1018 nm2.
x
The degree of Mo reduction was 1.12 in the reduced Mo
sample (Table 3). Then the mean oxygen ion number, related
to one Mo` ion:
respect to the e†ect of H S on product selectivities.
x \ 1/2(6 [ e/Mo) \ 2.44
2
Assuming, that the surface occupied with molybdena equals
practically with the area of the oxygen ions (as the applicabil-
ity of this treatment has been proved45 successfully) we obtain
Conclusions
The uptake of sulfur by alumina supported Pd-containing
for the area of a monomolecular MoO
2.44
(a is the area of one oxygen ion). Accepting 0.132
layer, A
\ 2.44
MoO exceeds substantially that of the Pt-containing and of
MoO_
x
] n
a
the nonpromoted molybdena, as measured both by XPS and
Mo O_ O_
nm for oxygen ion radius46 a \ 0.054 74 nm2, and A
\
sulfur radiotracer experiments. Molybdena forms amorphous
layer(s) on the surface and transforms into MoS patches
upon sulÐdation of PdMo as seen by TEM as shown by elec-
tron di†raction. Palladium in PdMo reacts with H S to give
O_
MoO_
70.8 ] 1018 nm2. It follows from this, that the surface area,
2
experimentally
(63.4 ] 1018 nm2) reaches 82 ^ 6% of the monomolecular
MoO layer. Consequently the molybdenum oxide layer in
determined
by
CO
chemisorption
2
2
Pd S and Pd
S
particles. Strong sulÐdation increases the
2.44
4
16 7
Mo sample can be regarded as being near to monomolecular.
ratio of Mo(IV) and Mo(V) in comparison with samples,
reduced in hydrogen.
Data on total sulfur uptakes measured by the radiotracer
method are in agreement with those measured by XPS
resulting in a similar sequence: PdMo [ PtMo [ Mo. The
S/&M values are equal in the case of Mo, substantially higher,
References
1
B. H. Cooper, A. Stanislaus and P. N. Hannerup, Hydrocarbon
Process., 1993, 72, 83.
1542
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543

View Article Online
2
3
W. R. A. M. Robinson, J. A. R. Van Veen, V. H. J. de Beer and
R. R. van Santen, Fuel Process. T echnol., 1999, 61, 103.
H. R. Rheinhoudt, R. Troost, A. D. van Langeveld, S. T. Sie,
J. A. R. van Veen and A. Moulijn, Fuel Process. T echnol., 1999,
61, 117.
21 Practical Surface Analysis, ed. D. Briggs and M. P. Seah, Wiley,
Chichester, 1990, app. 6, vol. 1, p. 635.
22 T. I. Koranyi, M. Dobrovolszky, T. Koltai, K. Matusek, Z. Paal
and P. Tetenyi, Fuel Process. T echnol., 1999, 61, 55.
23 D. Bassett and H. W. Habgood, J. Phys. Chem., 1960, 64, 767.
24 U. Schroder and N. H. Schoon, J. Catal., 1993, 143, 369.
25 Z. Paal, R. Schlogl and G. Ertl, Catal. L ett., 1992, 12, 331.
26 T. I. Koranyi and Z. Paal, Surf. Sci., 1991, 52, 285.
27 Y. C. Xie and Y. Q. Tang, Adv. Catal., 1990, 37, 1.
28 JPCDS File 5-0681.
4
5
6
7
8
9
M. H. Pinzon, L. I. Merino, A. Centeno and S. A. Giraldo, Stud.
Surf. Sci. Catal., 1999, 127, 97.
T. Kabe, W. Oian, Y. Hirai, L. Li and A. Ishihara, J. Catal., 2000,
190, 191.
J. Barbier, E. Lamy-Pitare, P. Marecot, J. P. Boitiaux, J. Cosnys
and F. Verna, Adv. Catal., 1990, 37, 279.
29 JPCDS File 10-0335.
30 JPCDS File 30-0884.
S. K. Purnell, J. R. Chang and B. C. Gates, J. Phys. Chem., 1993,
97, 4196.
31 J. B. Gunter, O. Marks, T. I. Koranyi and Z. Paal, Appl. Catal.,
P. C. Largentier and N. S. Figoli, Ind. Eng. Chem. Res., 1997, 36,
2543.
1988, 39, 285.
32 T. Koltai, PhD Thesis, Veszprem University, Hungary, 1999.
33 T. Koltai, V. Galsan and P. Tetenyi, React. Kinet. Catal. L ett.,
1999, 67, 391.
M. Dobrovolszky, K. Matusek, Z. Paal and P. Tetenyi, J. Chem.
Soc., Faraday T rans., 1993, 89, 3137.
10 Z. Paal, P. Tetenyi, M. Muhler, J.-M. Manoli and C. Potvin, J.
Chem. Soc., Faraday T rans., 1998, 94, 459.
11 W. Qian, A. Ishihara, G. Wang, T. Tsuzuki, M. Godo and T.
Kabe, J. Catal., 1997, 170, 286.
34 P. Tetenyi and V. Galsan, React. Kinet. Catal. L ett., 2000, 70,
265.
35 A. Balandin, Adv. Catal., 1958, 10, 96.
36 P. Tetenyi, in Surface and Defect Properties of Solids, Chemistry
Society, Burlington House, London, 1976, vol. 5, p. 81.
37 P. Tetenyi and V. Galsan, React. Kinet. Catal. L ett., 2001, to be
published.
12 P. J. Mangnus, A. Bos and J. A. Moulijn, J. Catal., 1994, 146,
437.
13 J. Shabtai, N. K. Nag and F. E. Massoth, J. Catal., 1987, 104,
413.
14 M. Dobrovolszky, Z. Paal and P. Tetenyi, Appl. Catal. A, 1996,
38 A. N. Startsev, V. A. Burmistrov and Yu. I. Yermakov, Appl.
Catal., 1988, 45, 191.
142, 159.
39 H. Topsoe and B. S. Clausen, Catal. Rev. Sci. Eng., 1984, 26, 395.
40 R. Prins, V. H. J. de Beer and G. A. Somorjai, Catal. Rev. Sci.
Eng., 1989, 31, 1.
15 T. Koltai, M. Dobrovolszky and P. Tetenyi, Stud. Surf. Sci.
Catal., 1999, 127, 137.
16 T. Koltai, F. E. Massoth and P. Tetenyi, React. Kin. Catal. L ett.,
2000, 71, 85.
41 F. E. Massoth, J. Catal., 1977, 47, 316.
42 F. Bataille, J.-L. Lemberton, P. Michaud, G. Perot, M. Vrinat, M.
Lemaire, E. Schulz, M. Breysse and S. Kasztelan, J. Catal., 2000,
61, 103.
17 W. S. Millman, Koii-Ichi Segawa, D. Smrz and W. K. Hall, Poly-
hedron, 1986, 5, 169.
18 G. Molnar, T. Belgya, I. Dabolczi, B. Zs. Fazekas, A
Š
. Veres, I.
43 H. Topsoe, Proc. 12th ICC 2000, Studies in Surf. Sci. and Catal.,
Elsevier, Amsterdam, 2000, vol. 130, part A, p. 1.
Bikit, J. Kiss and J. Ostor, J. Radioanal. Nucl. Chem., 1997, 215,
111.
44
45 L. Tamaska, A
1988, 65, 285.
A
Š
. Redey, J. Goldwasser and W. K. Hall, J. Catal., 1988, 113, 82.
19 J.-M. Manoli, C. Potvin, M. Muhler, U. Wild, G. Resofszki, T.
Buchholz and Z. Paal, J. Catal., 1998, 178, 338.
20 J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben,
Handbook of X-Ray Photoelectron Spectroscopy, Perkin-Elmer
Co., Eden Prairie, MN, 1992.
Š
. Redey and P. Tetenyi, React. Kinet. Catal. L ett.,
46 R. C. Weast, M. J. Astle and W. H. Beyer, Handbook of Chemistry
and Physics, 66th edn. CRC Press, Boca Raton, FL, 1987.
Phys. Chem. Chem. Phys., 2001, 3, 1535È1543
1543