Correlation of Pt-Re surface properties with reaction pathways for the aqueous-phase reforming of glycerol

Article abstract of DOI:10.1016/j.jcat.2011.11.015

The surface properties of Pt-Re catalytic nanoparticles supported on carbon following exposure to a hydrogen reducing environment and subsequent hydrothermal conditions have been studied using in situ X-ray photoelectron spectroscopy (XPS) and ammonia temperature-programmed desorption (TPD). These properties have been correlated with the catalyst selectivity for the aqueous-phase reforming of glycerol. We show that Pt in reduced Pt-Re/C becomes electron deficient, and a fraction of the Re becomes oxidized when the catalyst is subsequently exposed to hydrothermal reaction conditions. Oxidation of Pt-Re generates surface acidity, which drastically affects the reaction pathways. The acid site concentration, but not acid site strength, increases with Re loading. This acidity increase with Re addition favors C-O over C-C cleavage, which results in higher selectivity to liquid products and alkanes at the expense of hydrogen selectivity. We propose a model for the Pt-Re active site and the origin of acidity enhanced by the addition of Re.

Full text of DOI:10.1016/j.jcat.2011.11.015

Journal of Catalysis 287 (2012) 37–43
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Correlation of Pt–Re surface properties with reaction pathways for the
aqueous-phase reforming of glycerol
a,b,
Liang Zhang a, Ayman M. Karim a, Mark H. Engelhard a, Zhehao Wei b, David L. King a, , Yong Wang
⇑
⇑
a Institute for Integrated Catalysis, Pacific Northwest National Laboratory, 902 Battelle Blvd., Richland, WA 99352, United States
b The Gene and Linda Voiland School of Chemical Engineering and Bioengineering, Washington State University, Pullman, WA 99164, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The surface properties of Pt–Re catalytic nanoparticles supported on carbon following exposure to a
hydrogen reducing environment and subsequent hydrothermal conditions have been studied using
in situ X-ray photoelectron spectroscopy (XPS) and ammonia temperature-programmed desorption
(TPD). These properties have been correlated with the catalyst selectivity for the aqueous-phase reform-
ing of glycerol. We show that Pt in reduced Pt–Re/C becomes electron deficient, and a fraction of the Re
becomes oxidized when the catalyst is subsequently exposed to hydrothermal reaction conditions.
Oxidation of Pt–Re generates surface acidity, which drastically affects the reaction pathways. The acid
site concentration, but not acid site strength, increases with Re loading. This acidity increase with Re
addition favors C–O over C–C cleavage, which results in higher selectivity to liquid products and alkanes
at the expense of hydrogen selectivity. We propose a model for the Pt–Re active site and the origin of
acidity enhanced by the addition of Re.
Received 4 August 2011
Revised 21 November 2011
Accepted 22 November 2011
Available online 17 January 2012
Keywords:
Aqueous-phase reforming
Surface acidity
Pt–Re/C catalyst
Pt–Re interactions
C–C bond cleavage
C–O bond cleavage
Re oxidation
Ó 2011 Elsevier Inc. All rights reserved.
Ammonia TPD
1. Introduction
reports indicate that Re addition to Pt-based catalysts supported
on metal oxides increases activity for the WGS reaction. Two the-
Production of hydrogen through the aqueous-phase reforming
(APR) of biomass-derived oxygenated hydrocarbons (sugars, sugar
alcohols, polyols, etc.) is considered a promising catalytic process
and has attracted academic and industrial interest. Important
advantages of the APR process include its low energy consumption,
since APR proceeds at lower temperatures than conventional
steam reforming, and its compatibility with the use of wet feed-
stocks, avoiding any predrying step [1]. A variety of catalyst formu-
lations have been proposed and tested by several groups for the
APR process [2–10]. Precious metals such as Pt are preferred cata-
lysts due to their high selectivity in breaking C–C bonds [11]. The
catalyst activity in the APR reaction can be enhanced by doping
precious metals with other transition metals [9,10,12,13]. For
example, our recent work has shown that addition of Re to Pt/C
substantially increases conversion of glycerol in the aqueous phase
by at least one order of magnitude [14]. However, the role of Re
and nature of active sites are still not fully understood.
ories have been provided for the observed enhanced activity. The
first is Pt–Re alloy formation [15–17], and the second involves
the role of ReOx species [18,19]. One effect of Pt–Re alloy formation
that has been in debate is the strength of CO adsorption, with the
idea that too strong CO chemisorption increases CO site coverage
and decreases the availability of operating WGS sites. Therefore,
weaker adsorption strength could result in higher activity. There
have been both reports of stronger [15] and weaker CO adsorption
[13,20] compared to pure Pt. These contradictory reports could be
due to particle size effects. Ruettinger et al. reported that CO
adsorption is weaker on small Pt–Re particles and stronger on
the larger Pt–Re particles [16]. An alternative theory has been pro-
posed wherein ReOx species are present and provide a redox route
for the WGS reaction, in which ReOx is reduced by CO, generating
CO2, and re-oxidized by H2O, forming H2 [18].
Analogous to the examination of Re promotion of Pt catalysts
for the WGS reaction, Kunkes and coworkers studied the structure
of reduced Pt–Re supported on carbon for the glycerol steam
reforming reaction. They proposed formation of a Pt–Re alloy
phase or a surface having close contact between the Pt and Re
phases and measured a lower CO binding energy compared with
Pt alone, thus allowing more sites to be available for reforming
[13,21]. However, the lower CO binding energy was measured on
APR of biomass-derived liquids normally involves reforming
followed by water gas shift (WGS) of the CO produced. Literature
⇑
Corresponding authors. Address: Institute for Integrated Catalysis, Pacific
Northwest National Laboratory, 902 Battelle Blvd., Richland, WA 99352, United
States.
E-mail addresses: david.king@pnl.gov (D.L. King), yongwang@pnl.gov (Y. Wang).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.11.015

38
L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
partially oxidized Pt–Re using 2% O2 at 100 °C followed by reduc-
tion at 70 °C prior to the CO-TPD [13].
for each TPD experiment. The catalyst was reduced in 50 sccm H2
at 280 °C for 1 h, and then the sample was purged with 50 sccm N2
for 30 min and the temperature cooled to 225 °C. N2 flow through
a water bubbler at room temperature was used to treat the catalyst
with water vapor at 225 °C through several pulses. The water partial
pressure in each pulse was about 20 Torr. The H2O pulsing was
repeated until no more H2O uptake was detected using the MS.
APR reactions were carried out in a single-channel microreactor
with active heat exchange to maintain near isothermal conditions
within the catalyst bed as previously described [14,24]. The micro-
channel reactor body was made from SS316 and was enclosed by
an oil-heating jacket with the oil being circulated using a Julabo
oil pump. The reaction temperature was controlled within 1 °C.
The total effective volume of the reactor channel was about
0.45 ml (0.635 mm ꢀ 12.7 mm ꢀ 55 mm). A charge of 200 mg cat-
alyst (60–100 mesh) was packed into the narrow slot of the micro-
channel reactor, and a stainless steel screen (ꢁ325 mesh) and a
stainless steel foam disk were placed on each side of the reactor
to maintain the catalyst particles in place during experiments.
Prior to the reaction test, the catalyst was reduced at 280 °C for
2 h in pure H2 (20 sccm) at atmospheric pressure and cooled down
to room temperature. After purging the system with N2 for 30 min,
the backpressure regulator was set at the reaction pressure
(425 psig), the aqueous feed solution comprising 10% glycerol (by
weight) was introduced using a HPLC digital pump (Series III) at
the desired feed rate, and heating of the catalyst bed was initiated.
When the reactor reached the reaction temperature of 225 °C, N2
flow was set at 20 sccm (N2 served as an internal analytical stan-
dard). The system was allowed to stabilize for about 2 h prior to
analysis of the reaction products. Gas product analysis was carried
out using a micro gas chromatograph (Agilent Micro GC 3000C)
equipped with four channels, one molecular sieve 5A column,
one Plot Q column, and two OV-1 columns. The liquid products
were collected and analyzed with a Waters high-performance
liquid chromatograph (HPLC). A Bio-Rad Aminex HPX-87H ion
exclusion column (300 mm ꢀ 7.8 mm) was used for analyte sepa-
ration. A 0.005 M H2SO4 aqueous solution was used as eluent for
this analysis at a flow rate of 0.55 ml/min. Carbon balance was typ-
ically greater than 95%.
In this work, we provide characterization of Pt–Re under condi-
tions similar to that encountered during APR in an effort to obtain
an improved understanding of the role of Re. We specifically focus
on investigation of the structure of Pt–Re/C catalysts under envi-
ronments that simulate APR reaction conditions and on the corre-
lation of catalyst structure to product distribution. XPS results
suggest that hydrogen reduction results in the formation of bime-
tallic Pt–Re with Pt slightly positively charged. Under an environ-
ment similar to APR reaction conditions, XPS shows that the Re
in the bimetallic is substantially oxidized. NH3 adsorption further
reveals that the Pt–Re interaction following exposure to water va-
por results in acidity generation that affects the reaction pathways.
We show that the prevalence of the dehydration pathways, relative
to decarbonylation (C–C bond cleavage), is related to the strength
and number of surface acid sites. Surface acidity on bimetallic cat-
alysts such as Pt–Re and Rh-Re has been recently reported [22,23].
2. Materials and methods
The catalysts were prepared by sequential incipient wetness
impregnation of tetra-amine platinum nitrate and perrhenic acid
(Alfa Aesar). A high surface area activated carbon support (Engel-
hard, SSA 540 m2/g, pore volume 0.42 ml/g) was selected for this
work. Specifically, for synthesis of the 3 wt%Pt/C catalyst, 20 g of
the dried carbon support was impregnated with a solution pre-
pared by dissolving 1.24 g of Pt(NH3)4(NO3)2 into 8.4 g of DI water
in a 30-ml glass vial, with shaking for at least 2 h prior to use. The
impregnated sample was dried at 110 °C in air for 2 h and then cal-
cined at 260 °C in air for 2 h with a ramp rate of 5 °C/min. Pt/C
catalysts with different loadings were prepared analogously. For
preparation of Pt–Xwt%Re/C catalysts (X ranges from 1 to 4.5),
the perrhenic acid solution was impregnated on the prepared Pt/
C in a subsequent step, with analogous pretreatment.
XPS measurements were made using a Physical Electronics
Quantum 2000 Scanning ESCA Microprobe. This system uses a fo-
cused monochromatic Al K
ical section analyzer. The X-ray beam (100
a
X-ray (1486.7 eV) source and a spher-
m diameter) is incident
l
3. Results and discussion
normal to the sample, rastering over a 1.3 mm by 0.2 mm rectangu-
lar area. The photoelectron detector was at 45° off-normal using an
analyzer angular acceptance width of 20° ꢀ 20°. High-energy-
resolution spectra were collected using a pass energy of 46.95 eV.
The energy scale of the analyzer was calibrated using sputter-
cleaned Cu and Ag foils. For the Ag 3d5/2 line, these conditions pro-
duced FWHM of better than 0.98 eV. The binding energy (BE) scale
is calibrated using the Cu 2p3/2 feature at 932.62 0.05 eV and Au 4f
at 83.96 0.05 eV from known standards. In an attempt to neutral-
ize any charge buildup during the analysis, the samples were
exposed to low-energy electrons (about 1 eV) and low-energy Ar+
ions. However, the surface charging could not be completely elim-
inated, and therefore, we used C 1s binding energy at 284.5 eV as a
reference to correct for the charging. The instrument is equipped
with a side chamber that allows in situ sample treatment with var-
ious gases with programmed temperature profiles and the transfer
of the samples to the XPS chamber without exposure to air. A
7-sample Mo holder is used to transfer the samples between the
XPS analysis chamber and the side chamber using a magnetic trans-
fer arm. Samples were sequentially treated with 100 sccm H2 flow
at 80 °C and 280 °C for 1 h and exposed to 100 Torr water vapor at
225 °C for 1 h. After each treatment, the samples were transferred
to the XPS chamber for spectra collection.
3.1. APR of glycerol on Pt/C and Pt–Re/C
As reported previously [13], Pt/C and Pt–Re/C behave differently
in aqueous-phase reforming of glycerol. Addition of Re to Pt signif-
icantly enhances conversion of glycerol as evidenced by more than
10 times increase in TOF of glycerol (based on CO chemisorption),
as shown in Fig. 1. As also can be seen in Fig. 1, addition of Re to Pt
changes the distribution of products. Specifically, the presence of
Re increases the selectivity toward Cþ2 alkanes and alcohols (etha-
nol and propanol), along with the production of carboxylic acids, at
the expense of hydrogen, carbon dioxide, and diols (ethylene glycol
(EG) and propylene glycol (PG)) selectivity. As a comparison, APR
experiments were also conducted on Re/C and a physical mixture
of Pt/C plus Re/C, Table 1. Re/C showed a much lower activity than
Pt/C, while the physical mixture of Pt/C and Re/C exhibited a sim-
ilar activity to Pt/C, indicating no synergistic effect by Re without
close contact between Pt and Re. In order to understand the role
of Re in changing the product distribution, we studied the surface
properties of Pt–Re, in particular under hydrothermal conditions.
3.2. Oxidation state change for Pt–Re/C catalysts
NH3 temperature-programmed desorption was carried out using
an Autochem II pulse chemisorption (Micromeritics) unit and mass
spectrometer (MS) analyzer (Pfeiffer); 100 mg of catalyst was used
To understand Pt–Re interaction on the Pt–Re/C catalyst during
the APR reaction, we examined the electronic structure of Pt and Re

L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
39
results in a clear shift of Pt0 binding energy, and the shift increases
with the amount of Re in the catalyst. After deconvolution of the Pt
4f peaks, the Pt 4f7/2 binding energy for the series of catalysts was
quantified and plotted against the fraction of Re in Pt–Re as shown
in Fig. 3. It can be seen that the Pt 4f7/2 binding energy shift is pro-
portional to the amount of Re in the catalyst. Some studies have
proposed that Pt and Re form an alloy upon reduction [13,28,29].
The increase in electron deficiency of Pt observed in this work
may be a result of the formation of a Pt–Re alloy. It should be
emphasized that Re was impregnated on a presynthesized and cal-
cined Pt/C, and Pt particle size for catalysts with various Pt/Re ra-
tios is the same prior to Re addition. The size of Pt-associated
particles would be expected to increase upon addition of Re due
to the formation of the Pt–Re alloy. Therefore, the increase in bind-
ing energy shift is mainly due to the Pt–Re interaction and not due
to (a decrease in) Pt particle size. Additionally, EDS analysis has
been performed on individual particles in our Pt–Re/C catalysts
(EDS results on reduced catalyst are shown in the Supplementary
information, Figs. 2 and 3), and the Pt/Re ratio was uniform within
each particle, consistent with the formation of a Pt–Re alloy. It
should be noted that for a Pt–Re alloy, one would expect the trend
in Re 4f7/2 binding energy to be opposite to that of Pt 4f7/2. How-
ever, the Re 4f7/2 binding energy was higher for the Pt–Re/C cata-
lysts compared to the Re/C catalyst and was in the range of
40.8 eV to 41 eV, close to 4f7/2 of Re0 as reported by Shpiro et al.
[30]. It appears that Re addition results in both Pt and Re becoming
more electron deficient, possibly due to carbon–metal interactions,
wherein the extent of Pt electron deficiency is proportional to the
Re amount in the catalyst.
Fig. 1. TOF of glycerol (minꢁ1, based on CO chemisorption) and selectivity toward
products (based on carbon) for aqueous-phase reforming of 10% glycerol over 3%Pt/
C and 3%Pt–3%Re/C at 225 °C and 420 psig. The selectivity to H2 is defined as the
hydrogen produced relative to the maximum that could be produced if all the
reacted glycerol were converted to H2 and CO2. Space velocity (WHSV: gram
glycerol/gram catalyst/hr) was held at 1.5 for Pt/C and 15 for Pt–Re/C to achieve a
similar conversion level (30%).
in Pt–Re/C catalysts using XPS after hydrogen pretreatment fol-
lowed by exposure to water vapor. With the in situ treatment capa-
bility of the XPS instrument, we were able to track the changes in
the oxidation states of Pt and Re in Pt–Re/C catalysts after desig-
nated treatments without exposure to air. All spectra shown were
calibrated to C 1s at 284.5 eV. Fig. 2 shows Pt 4f7/2 and Re 4f7/2
spectra of a series of catalysts after reduction. The Pt 4f7/2 binding
energy of both calcined 3%Pt/C and 3%Pt–X%Re/C was measured at
72.85 eV (not shown), which was assigned to PtO [25]. When 3%Pt/
C was reduced at 280 °C, the Pt 4f7/2 binding energy shifted to
71.52 eV. This peak could be assigned to zero-valent Pt, although
it is 0.7 eV higher than bulk Pt metal, possibly due to metal–sup-
port interaction or a cluster size effect [26,27]. The addition of Re
In this study, the feed was 10 wt% glycerol in water. The molar
ratio of water to glycerol was about 50:1. Therefore, in the con-
densed phase at elevated temperature (200–250 °C for APR reac-
tion), the role of water on the Pt–Re interaction must be
considered. To simulate APR conditions, a stream of water vapor
was fed into the XPS side chamber where Pt–Re/C catalysts were
previously reduced. A dramatic change in oxidation state was
Table 1
Comparative conversion and selectivity values for the APR of glycerol over Pt/C, Pt–Re/C, and Re/C catalysts. Reaction conditions: 225 °C, 420 psig, WHSV = 6 hꢁ1
.
3%Pt/C unreduced
3%Pt/C reduced
3%Pt–3%Re/C unreduced
3%Pt–3%Re/C reduced
Physical mix reduced
3%Re/C reduced
Conversion
C–O/C–C
H2/CO2
8.2%
0.6
2.3
7.4%
0.6
2.3
6.7%
0.5
2.3
68.2%
1.1
2.0
10.1%
1.0
1.8
2.6%
33.8
Re 4f7/2 Reduced at 280oC
40.8
Pt 4f7/2 Reduced at 280oC
71.52
2.0
1.5
1.0
0.5
0.0
2.0
1.5
PtRe (3:1)
PtRe (3:3)
PtRe (1:3)
PtRe (3:3)
1.0
0.5
0.0
PtRe (1:3)
PtRe (0:3)
PtRe (3:1)
PtRe (3:0)
69
70
71
72
73
39
40
41
42
BE (eV)
BE (eV)
Fig. 2. XPS spectra of Pt/C, Re/C, and Pt–Re/C catalysts with different ratios (based on weight percentage of metal). All the catalysts were reduced at 280 °C. Dashed lines label
the binding energy for Pt0 4f7/2 (71.52 eV) in 3%Pt/C, and for Re0 4f7/2 (40.8 eV) in 3%Re/C, obtained by peak fitting.

40
L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
treated under conditions similar to the in situ XPS experiments. The
left panel in Fig. 6 shows the NH3 adsorption capacity for catalysts
reduced at 280 °C and exposed to 20 Torr water vapor (balance Ar
at atmospheric pressure) at 225 °C. To establish a baseline, blank
carbon was also treated in the same way as other catalysts and sat-
urated with ammonia; negligible ammonia desorption was de-
tected, as shown. It can be seen that the ammonia desorption
peak from Pt/C centers at 136 °C while ammonia desorption on
Pt–Re/C shows one peak centered at 188 °C with a shoulder at
108 °C for all catalysts. Based on the location of the major desorp-
tion peaks, the acid strength of Pt/C is less than that of Pt–Re/C. The
type or distribution of acid sites is difficult to determine, as charac-
terization by FTIR of adsorbed probe molecules such as CO, NH3, or
pyridine is not applicable to carbon-supported catalysts. However,
separate infrared studies on silica-supported Pt–Re suggest that
Brønsted acid sites form after the catalyst is first reduced followed
by exposure to steam (see Supplementary information, Fig. 4 and
related discussion). The number of acid sites was measured by
integrating the area under the TPD curves, and the results are
shown in the right panel of Fig. 6. The ammonia capacity increased
linearly with the loading amount of Re in Pt–Re/C catalysts. The ra-
tio of acid to metal sites was between 0.37 and 0.44 for the Pt–Re/C
catalysts, while it was 0.18 for the 3%Pt/C catalyst. The XPS results
on Re/C after reduction and exposure to steam had shown less oxi-
dation compared to Re in Pt–Re/C. Analogously, the Re/C showed
lower NH3 capacity than Pt–Re/C catalysts as shown in Fig. 6.
The XPS results, low ammonia capacity on Re/C, and the linear rela-
tion between the Re loading on Pt/C and the number of acid sites
all imply that the acid sites mainly evolved from Pt–Re interaction,
which could be ascribed to increasing number of surface-oxidized
Pt–Re. The introduction of surface acidity by Re was reported ear-
lier [22], and the addition of KOH was shown to neutralize the acid
sites [14,22]. More recently, Chia et al. showed that the acidity
introduced by ReOx was responsible for the enhanced C–O hydrog-
enolysis on Re–Rh/C catalysts [23]. However, we believe that the
acidity is introduced through a Pt–O–Re structure rather than a
ReOx species as will be discussed below. One of the possible struc-
tures of oxidized Pt–Re in the presence of water is shown in Fig. 7.
Pt–Re may be oxidized by dissociated water with OH attached to
Re. Because of the oxophilic nature of Re, Re–OH may function as
a Brønsted acid site, while nearby Pt could provide a Lewis acid site
due to its electron deficiency. Another evidence for oxidized Pt–Re
as acid site and active site resides in solubility of ReOx. It is well
known that Re VII readily dissolves in water. ICP analysis of the
Fig. 3. Binding energies of Pt 4f7/2 and Re 4f7/2 for Pt/C, Re/C, and Pt–Re/C catalysts
reduced at 280 °C.
subsequently found for both Pt and Re, as shown in Fig. 4. Pt was
partially oxidized by water in both Pt/C and Pt–Re/C (Fig. 4a and
b), indicating that Pt–O or Pt–OH might participate in glycerol con-
version. The left panel in Fig. 5 shows a shift in Pt 4f7/2 binding en-
ergy for Pt–Re/C catalysts with varying Pt/Re ratios. The Pt 4f7/2
binding energy of water vapor–treated catalysts follows a similar
trend as for the reduced catalysts, and the shift in Pt 4f7/2 binding
energy increases with the amount of Re. Water treatment also re-
sults in oxidation of Re in the Pt–Re catalysts. As shown in Figs. 4c
and 5 (right panel), Re is oxidized to a mixture of oxidation states,
which may include 2+, 4+, 6+, and 7+. The oxophilicity of Re has
been reported in the literature. A study by Fusy et al. showed that
water readily dissociated on the surface of Re at room temperature
with generation of atomic O and OH radicals [31]. Surprisingly,
after exposure to water vapor, Re/C is less oxidized than Re in
Pt–Re/C catalysts (right panel of Fig. 5). This suggests that addition
of Pt facilitates the oxidation of Re.
3.3. Surface acidity of oxidized Pt–Re/C
The oxidation of Pt and Re by steam has important implications
for the structure of Pt–Re under APR conditions. One possibility is
the generation of surface acidity. In order to measure the number
of surface acid sites and relative acid strength, we carried out
NH3 adsorption measurements on the Pt–Re/C catalysts with
different Pt/Re ratios. Prior to the base titration, the catalysts were
a
3.0
c
b
Pt 4f of 3%Pt/C
72.85
Pt 4f for 3%Pt3%Re/C
Re 4f for 3%Pt3%Re/C
45.5
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
71.52
71.7
72.85
3.5 40.88
Steam at 225oC
Steam at 225 oC
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Steam at 225 oC
2.5
2.0
1.5
1.0
0.5
0.0
Reduced at 280oC
Calcined
Reduced at 280 oC
Reduced at 280 oC
Calcined
Calcined
70
72
74
76
78
80
82
70
72
74
76
78
80
82
38
40
42
44
46
48
50
52
54
56
BE (eV)
BE (eV)
BE (eV)
Fig. 4. XPS spectra of 3%Pt/C and 3%Pt–3%Re/C catalysts acquired after sequential treatments: calcination at 260 oC, reduction at 280 °C, and exposure to steam at 225 °C. The
doublets in most of the spectra correspond to the 4f7/2 (lower energy) and 4f5/2 (higher energy) peaks. The left dashed line in the figures corresponds to the 4f7/2 values for Pt0
in 3%Pt/C and 3%Pt-3%Re/C in 4(a) and 4(b), respectively, and the 4f7/2 value for Re0 in 3%Pt-3%Re/C in 4(c), all obtained by peak fitting. The right dashed lines correspond to
Pt2 in bulk PtO and Re7+ in bulk Re2O7. A shift in the binding energy is clearly seen following addition of steam at 225 °C. A smaller but similar shift was seen with lower steam
partial pressure.

L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
41
Re 4f Exposed to Steam at 225oC
Pt 4f7/2 Exposed to Steam at 225oC
2.5
2.0
1.5
1.0
0.5
0.0
45.5
2.0
1.5
1.0
0.5
0.0
72.85
40.8
71.52
PtRe (3:1)
PtRe (1:3)
PtRe (3:3)
PtRe (1:3)
PtRe (3:3)
PtRe (3:1)
PtRe (3:0)
PtRe (0:3)
69
70
71
72
73
38 40
42 44 46 48
50 52 54 56
BE (eV)
BE (eV)
Fig. 5. XPS spectra of Pt/C, Re/C, and Pt–Re/C catalysts with different ratios (based on weight percentage of metal). All the catalysts were reduced at 280 °C and exposed to
water vapor at 225 °C. Left dashed lines correspond to the binding energies for Pt0 4f7/2 in 3%Pt/C and Re0 4f7/2 in 3%Re/C reduced at 280 °C, obtained by peak fitting. The right
dashed lines represent values for Pt2+ in bulk PtO and Re7+ in bulk Re2O7.
3Pt/C
2.50E-009
3Pt1Re/C
3Pt3Re/C
3Pt4.5Re/C
3Re/C
2.00E-009
1.50E-009
1.00E-009
5.00E-010
Carbon
0.00E+000
0
50 100 150 200 250 300 350 400 450
Temperature (oC)
Fig. 6. Adsorption of NH3 over Pt/C and Pt–Re/C catalysts with varying Pt/Re ratio along with Re/C: NH3 TPD on the left panel and amount of adsorbed NH3 as function of Re
loading in the right panel. The NH3 absorption on 3%Re/C is 0.03 mmol/g (not shown in the right panel), similar to 3%Pt/C. All the catalysts were reduced at 280 °C and
saturated with water vapor at 225 °C prior to NH3 adsorption.
spent 3%Pt–3%Re/C showed loss of about 50% of the Re; however,
the catalyst activity and selectivity were only slightly changed over
100 h (as shown in Fig. 5 in Supplemental materials). The number
of acid sites will be correlated with the catalyst activity and prod-
uct distribution below.
Following this concept, a schematic network of reaction pathways
is depicted in Fig. 8. To simplify the analysis and provide greater in-
sight into the effect of Re on the reaction pathway, we define a new
parameter of competition ratio to characterize the overall reaction
results. The competition ratio is defined as the ratio of dehydration
pathway events to decarbonylation pathway events and is calcu-
lated using the following equation, where the number of C–O
bonds cleaved is divided by the number of C–C bonds cleaved:
3.4. Effect of the surface acidity of Pt–Re/C on the reaction pathways
Several papers have reported the selectivity of gas products and
liquid products from APR of glycerol [13,32–34]. The complexity of
products makes the analysis of reaction pathways based on indi-
vidual products very difficult. We have previously proposed two
major competing reaction pathways over the Pt–Re surface [14]:
C–C cleavage through decarbonylation and C–O cleavage through
dehydration followed by hydrogenation (since we do not see any
unsaturated products, a hydrogenation step must also occur).
Dehydration
Decarbonylation
P
imoles of dehydrated producti ꢀdehydration steps to reach producti
¼
moles of CO2
The dehydrated products include those derived directly or indi-
rectly from one or more dehydration steps, while the decarbonyla-
tion pathway leads to the formation of CO2 via water gas shift of
Fig. 7. One possible structure of oxidized Pt–Re in the presence of water. Upon exposure to water, Pt–Re dissociates water with hydroxyl groups attached to Re. Because of the
oxophilic property of Re, OH tends to deprotonate. The dissociated proton may protonate adsorbed ROH to facilitate the dehydration reaction.

42
L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
Fig. 8. Major reaction pathways for aqueous-phase reforming of glycerol.
alcohols were found at higher conversion, and the dehydration of
alcohols and diols was found to be less favored compared to dehy-
dration of glycerol [14], leading to increased decarbonylation
selectivity at higher conversion. Also, in other experiments (not
shown), the dehydration selectivity (at similar conversion) during
APR followed the order ethanol < ethylene glycol < glycerol. We
found similar trends with the other Pt/C and Pt–Re/C catalysts
and extrapolated the data in order to compare the competition
ratio for the different catalysts at the same conversion. The results
are presented in Fig. 10, and it is clear that the dehydration/decarb-
onylation ratio is proportional to the number of acid sites. The
reaction medium is just mildly acidic (pH = 5–6), raising the possi-
bility that media pH could have an effect on the dehydration path-
ways. However, in a separate experiment (not shown), addition of
nitric acid showed no effect on Pt–Re/C catalyst activity or product
selectivity despite the low pH of 2 during the reaction. Fig. 10
shows a linear relation between C–O/C–C cleavage reactions and
number of surface acid sites, over a glycerol conversion range from
0.1 to 0.65. Therefore, it can be predicted that addition of base to
the feed should neutralize the surface acid sites and depress the
dehydration/hydrogenation pathways, increasing hydrogen pro-
ductivity. This was confirmed by our previous work of glycerol
reforming in the presence of KOH [14].
Fig. 9. Competition ratio (C–O/C–C) and H2 selectivity as function of conversion on
3%Pt–3%Re/C.
The question can be raised that the acidity of the system may be
provided by Re/C particles in which the Re is not associated with
Pt. We believe that the comparative data provided in Table 1, cou-
pled with the NH3 TPD results provided in Fig. 6, argue against this.
The activity and total acidity provided by 3%Re/C are too low to
contribute significantly to the overall results, and it appears that
the Re associated with Pt generates more significant acidity in this
system. Moreover, we measured by ICP the Pt and Re content of
3%Pt–3%Re/C after 1 week of tests and found that the Re content
decreased by nearly 50%, whereas the Pt content was virtually
unchanged. The corresponding catalyst activity decreased by about
10%, and the selectivity was nearly unchanged over this time (see
Fig. 5, Supplemental information). Since Re7+ (Re2O7) is soluble in
water, the effect of Re in contributing to C–O bond cleavage must
come from its association with Pt, wherein it is stabilized relative
to dissolution.
Fig. 10. Relation between C–O/C–C and number of acid sites on Pt/C and Pt–Re/C
catalysts with different Pt/Re ratios.
CO. Ethylene glycol, methanol, and CO are pure decarbonylation
products, whereas many products can be seen to be the result of
a combination of dehydration and decarbonylation. For example,
methane is considered a product derived from ethanol, and its for-
mation includes one dehydration step (either dehydration of glyc-
erol flowed by decarbonylation or dehydration of ethylene glycol)
along with two decarbonylation steps (producing of two moles of
CO2).
Fig. 9 shows the competition ratio (C–O/C–C) and H2 selectivity
as a function of conversion for the 3%Pt–3%Re/C catalyst. The con-
version was varied by changing the space velocity of the feed.
Clearly, the competition ratio decreases with increasing conver-
sion, and since the dehydration pathway is followed by hydrogena-
tion, thereby consuming hydrogen, the H2 selectivity also increases
with increasing conversion. This trend could be due to a pool of
intermediates formed at higher conversion, which are more
difficult to dehydrate. For example, more diols and monohydric
4. Conclusions
We have investigated the structure of Pt–Re/C catalysts with
varying Re loadings following reduction and exposure to a simu-
lated hydrothermal environment. When exposed to steam at ele-
vated temperature, the Pt–Re catalyst undergoes oxidization. Our

L. Zhang et al. / Journal of Catalysis 287 (2012) 37–43
[7] K. Murata, I. Takahara, M. Inaba, React. Kinet. Catal. Lett. 93 (2008) 59.
43
results suggest that the active site during the APR reaction is not
the Pt–Re alloy but is more likely to be based on oxidized Pt–Re.
The Pt–Re interaction under hydrothermal conditions results in
local surface acidity. The number of acid sites is proportional to
the amount of Re in the Pt–Re catalysts. The surface acidity facili-
tates the dehydration pathway, with the ratio of dehydration to
decarbonylation increasing linearly with number of surface acid
sites. Our results show the importance of surface acidity in control-
ling the reaction pathways during aqueous-phase reforming.
[8] A. Tanksale, C.H. Zhou, J.N. Beltramini, G.Q. Lu, J. Incl. Phenom. Macrocycl. 65
(2009) 83.
[9] X.M. Wang, N. Li, L.D. Pfefferle, G.L. Haller, Catal. Today 146 (2009) 160.
[10] A. Iriondo, J.F. Cambra, V.L. Barrio, M.B. Guemez, P.L. Arias, M.C. Sanchez-
Sanchez, R.M. Navarro, J.L.G. Fierro, Appl. Catal. B 106 (2011) 83.
[11] R.R. Davda, J.W. Shabaker, G.W. Huber, R.D. Cortright, J.A. Dumesic, Appl. Catal.
B 43 (2003) 13.
[12] G.W. Huber, J.W. Shabaker, S.T. Evans, J.A. Dumesic, Appl. Catal. B 62 (2006)
226.
[13] E.L. Kunkes, D.A. Simonetti, J.A. Dumesic, W.D. Pyrz, L.E. Murillo, J.G.G. Chen,
D.J. Buttrey, J. Catal. 260 (2008) 164.
[14] D.L. King, L. Zhang, G. Xia, A.M. Karim, D.H. Heldebrant, X. Wang, T. Peterson, Y.
Wang, Appl. Catal. B (2010).
Acknowledgements
[15] Y. Sato, K. Terada, S. Hasegawa, T. Miyao, S. Naito, Appl. Catal. A 296 (2005) 80.
[16] W. Ruettinger, X.S. Liu, X.M. Xu, R.J. Farrauto, Top. Catal. 51 (2008) 60.
[17] D.A. Simonetti, E.L. Kunkes, J.A. Dumesic, J. Catal. 247 (2007) 298.
[18] K.G. Azzam, I.V. Babich, K. Seshan, L. Lefferts, Appl. Catal. B 80 (2008) 129.
[19] O.M. Daniel, A. DeLaRiva, E.L. Kunkes, A.K. Datye, J.A. Dumesic, R.J. Davis,
ChemCatChem 2 (2010) 1107.
[20] T. Ebashi, Y. Ishida, Y. Nakagawa, S. Ito, T. Kubota, K. Tomishige, J. Phys. Chem.
C 114 (2010) 6518.
[21] J.L. Xiao, R.J. Puddephatt, Coord. Chem. Rev. 143 (1995) 457.
[22] D.L. King, L. Zhang, A. Karim, Y. Wang, Prepr. – Am. Chem. Soc. Div. Pet. Chem.
56 (2011) 41.
The authors acknowledge financial support from the US Depart-
ment of Energy, Office of Energy Efficiency and Renewable Energy.
The XPS experiments were carried out at the Environmental and
Molecular Sciences Laboratory, a user facility of the Department
of Energy, of the Pacific Northwest National Laboratory.
Appendix A. Supplementary material
[23] M. Chia, Y.J. Pagan-Torres, D. Hibbitts, Q.H. Tan, H.N. Pham, A.K. Datye, M.
Neurock, R.J. Davis, J.A. Dumesic, J. Am. Chem. Soc. 133 (2011) 12675.
[24] C.S. Cao, G. Xia, J. Holladay, E. Jones, Y. Wang, Appl. Catal. A 262 (2004) 19.
[25] A.S. Arico, A.K. Shukla, H. Kim, S. Park, M. Min, V. Antonucci, Appl. Surf. Sci. 172
(2001) 33.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jcat.2011.11.015.
[26] W. Eberhardt, P. Fayet, D.M. Cox, Z. Fu, A. Kaldor, R. Sherwood, D. Sondericker,
Phys. Rev. Lett. 64 (1990) 780.
References
[27] S. Hüfner, G.K. Wertheim, Phys. Rev. B 11 (1975) 678.
[28] A.S. Fung, M.R. McDevitt, P.A. Tooley, M.J. Kelley, D.C. Koningsberger, B.C.
Gates, J. Catal. 140 (1993) 190.
[29] W.K. Shiflett, J.A. Dumesic, J. Catal. 77 (1982) 57.
[30] E.S. Shpiro, V.I. Avaev, G.V. Antoshin, M.A. Ryashentseva, K.M. Minachev, J.
Catal. 55 (1978) 402.
[31] J. Fusy, M. Alnot, J. Jupille, P. Pareja, J.J. Ehrhardt, Appl. Surf. Sci. 17 (1984) 415.
[32] N.J. Luo, X.W. Fu, F.H. Cao, T.C. Xiao, P.P. Edwards, Fuel 87 (2008) 3483.
[33] G.D. Wen, Y.P. Xu, H.J. Ma, Z.S. Xu, Z.J. Tian, Int. J. Hydrogen Energy 33 (2008)
6657.
[34] A. Wawrzetz, B. Peng, A. Hrabar, A. Jentys, A.A. Lemonidou, J.A. Lercher, J. Catal.
269 (2010) 411.
[1] R.R. Davda, J.W. Shabaker, G.W. Huber, R.D. Cortright, J.A. Dumesic, Appl. Catal.
B 56 (2005) 171.
[2] J.W. Shabaker, G.W. Huber, R.R. Davda, R.D. Cortright, J.A. Dumesic, Catal. Lett.
88 (2003) 1.
[3] F.Z. Xie, X.W. Chu, H.R. Hu, M.H. Qiao, S.R. Yan, Y.L. Zhu, H.Y. He, K.N. Fan, H.X.
Li, B.N. Zong, X.X. Zhang, J. Catal. 241 (2006) 211.
[4] A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, M.B. Guemez, R.M. Navarro, M.C.
Sanchez-Sanchez, J.L.G. Fierro, Top. Catal. 49 (2008) 46.
[5] P.N. Kechagiopoulos, S.S. Voutetakis, A.A. Lemonidou, I.A. Vasalos, Ind. Eng.
Chem. Res. 48 (2009) 1400.
[6] X.H. Liu, K. Shen, Y.G. Wang, Y.Q. Wang, Y.L. Guo, Y. Guo, Z.L. Yong, G.Z. Lu,
Catal. Commun. 9 (2008) 2316.