Aqueous phase reforming of glycerol using highly active and stable Pt0.05CexZr0.95-xO2 ternary solid solution catalysts

Article abstract of DOI:10.1016/j.apcata.2016.05.028

The aqueous phase reforming (APR) of glycerol is an attractive yet challenging pathway to convert abundant biomass into value added hydrogen. Pt catalysts have received attention due to their ability to produce hydrogen-rich gas under APR conditions. In t

Full text of DOI:10.1016/j.apcata.2016.05.028

Accepted Manuscript
Title: Aqueous phase reforming of glycerol using highly active
and stable Pt0.05CexZr0.95−xO2 ternary solid solution catalysts
Author: Afsaneh Sadat Larimi Mohammad Kazemeini Farhad
Khorasheh
PII:
S0926-860X(16)30306-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.05.028
APCATA 15892
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
28-2-2016
23-5-2016
29-5-2016
Please cite this article as: Afsaneh Sadat Larimi, Mohammad Kazemeini, Farhad
Khorasheh, Aqueous phase reforming of glycerol using highly active and stable
Pt0.05CexZr0.95-xO2 ternary solid solution catalysts, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.05.028
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Aqueous phase reforming of glycerol using highly active and
stable Pt0.05CexZr0.95-xO2 ternary solid solution catalysts
Afsaneh Sadat Larimi[a], Mohammad Kazemeini[a]*, and Farhad Khorasheh[a]*
[a] Department of Chemical and Petroleum Engineering, Sharif University of Technology, Tehran, Iran
* E-mail: kazemini@sharif.edu; khorashe@sharif.ir
1

Graphical abstract
2

Highlights

Pt0.05CexZr0.95-xO₂ ternary solid solutions were synthesized by a simple one-step method by our group
for the first time.


Pt0.05CexZr0.95-xO₂ ternary solid solutions are highly selective and stable for APR reaction.
Pt0.05Ce0.475Zr0.475O2 was found to be very active and stable in APR of glycerol, had the highest carbon
to gas conversion, H2 yield, H2 selectivity and stability.

This is the first application of the stable Pt0.05CexZr0.95-xO₂ ternary solid solutions not only for the APR
of glycerol but for any other catalytic reaction.
3

Abstract: The aqueous phase reforming (APR) of glycerol is an attractive yet challenging pathway to
convert abundant biomass into value added hydrogen. Pt catalysts have received attention due to their
ability to produce hydrogen-rich gas under APR conditions. In this work, the conversion of glycerol into
hydrogen is demonstrated using Pt0.05CexZr0.95-xO2 (x=0, 0.29, 0.475, 0.66 and 0.95) solid solution
catalysts. Both characteristic (XRD, BET, H2-TPR, CO-chemisorption, TEM and XPS) and reactivity
measurements were used to investigate the activity of the catalysts. Results indicated that reactivity
depended on the Ce/Zr ratio, which in turn affected the Pt oxidation state, active metal dispersion and
surface area, and particle size. Among these catalysts Pt0.05Ce0.475Zr0.475O2 showed the highest carbon to
gas conversion (95%), highest H2 yield (93%), highest H2 selectivity (98%) and at least 50 h of stability.
The overall catalytic performance decreased in the order Pt0.05Ce0.475Zr0.475O2 > Pt0.05Ce0.29Zr0.66O2 >
Pt0.05Zr0.95O2 > Pt0.05Ce0.66Zr0.29O2 > Pt0.05Ce0.95O2.
Keywords: PtCeZrO2 • Solid solution catalysts • Glycerol • Aqueous phase reforming • Renewable
hydrogen
Introduction
The environmental crises such as global warming and climate change caused by the fossil fuel
utilization, as well as the depletion of petroleum resources call for alternative energy sources.
Consequently, sustainable clean fuels generated from renewable sources, such as biomass, have come to
attention. In this sense, hydrogen is considered as one of the most promising forms of future clean
energy. The demand for hydrogen is also growing due to technological progress in the fuel cell industry
[1].
Aqueous-phase reforming (APR), developed by Dumesic and co-workers in 2002, is one of the
technically feasible approaches that has been proposed as an attractive new route for generating
hydrogen-rich fuel gas in a single step from various oxygenated hydrocarbons including glycerol,
ethylene glycol, sorbitol, glucose, methanol, and ethanol under relatively mild reaction temperatures of
200–250˚C and pressures in the range of 20–50 bar [2-8]. The current industrial process for hydrogen
production is the high-temperature steam reforming of hydrocarbons. APR has many advantages
compared with conventional steam reforming that include a higher energy efficiency and elimination of
undesirable reaction products that are formed at high temperatures. Additionally, the low reaction
temperatures used in APR favor the water-gas-shift reaction (WGS) resulting in a hydrogen-rich gas
with low CO content (<300 ppm) making it suitable for direct use in proton exchange membrane (PEM)
fuel cell systems [9].
4

According to the APR mechanism reported by Dumesic and co-workers [2-8], the desired catalyst in
APR should have highly activity for both the cleavage of the C-C bonds and for the WGS reaction
while at the same time be able to inhibit competing parallel and series reactions including C-O bond
cleavage and the methanation/Fischer-Tropsch reactions that would significantly reduce the hydrogen
yield. Of the metal catalysts employed in the APR process, the group VIII metals, particularly platinum,
have shown high activity for hydrogen production because of their high activity towards C-C bond
cleavage and WGS reactions and low selectivity towards the methanation reaction [2-4,10,11]. The
metal oxide carriers could significantly influence both the overall activity and the product selectivity in
APR reactions [4,11,12]. Catalyst stability in APR is a challenging issue [13]. When conventional metal
oxides such as Al2O3, ZrO2, SiO2, ZnO and TiO2 are used as support materials, their porous structures
could either deteriorate by sintering or their specific surface areas could decrease due to phase transfer
under APR conditions thus restricting their use as support materials [6,11,14]. Much is understood for
γ-Al2O3. Under the hydrothermal conditions, γ-Al2O3 is known to be hydrated and crystallized resulting
in crystalline boehmite and this process can affect the catalytic activity or the products formed [12].
Therefore, the support material should have a high hydrothermal stability to be used in the APR
reaction.
The use of ceria and zirconia based mixed oxide supports has been reported recently for a variety of
reactions including WGS [15-17] and preferential CO oxidation [18,19] with the mixed oxides
demonstrating a better performance compared with the individual oxides due to their higher
hydrothermal stability as well as oxygen storage capacity. In the case of APR reaction, Pt/γ-Al2O3
catalysts modified with ceria and ceria-zirconia show significantly higher H2 yield and selectivity
[20,21].
The APR of model compounds using a variety of nano-catalysts supported on metal oxide supports
has been reported by many investigators [6,11,12,22,23]. Decreasing the average diameter of the Pt
particles in the Pt/γ-Al2O3 catalysts increased the production of H2 and CO2 while CO production was
decreased [24] indicating that the rate of the water gas shift reaction increased with decreasing metal
particle size. These results provide valuable insight for development of new nano-catalysts. Catalyst
activity is highly influenced by the catalyst preparation method [9,25]. Dry or wet impregnation
methods that are most frequently used for preparation of APR catalysts could commonly result in the
agglomeration of the active phase on the support especially when a high loading of metal precursors are
used leading to the formation of large size particles.
Classifying multi-metallic catalysts into different structures including hetero-structures, core-shell
structures, solid solutions, ensembles and segregations, most investigated catalysts for the APR reaction
are hetero-structures. Different examples for bi-metallic catalysts are shown in Figure 1 [26]. In the
core-shell structure, different types of metal ions are reduced by their tendency for nucleation. First, the
5

ions of the core are reduced. Metal atoms of the shell prefer to nucleate and grow surrounding the core,
since separate nucleation is difficult for them. Individual nucleation and growth of different kinds of
metal atoms can occur to form hetero-structures and segregations under certain circumstances. When
metal-metal bonds form, homogeneous mixture of different metals is made. According to the
distribution of distinct metal atoms either ensembles or solid solutions are produced. Catalysts in
present study are of solid solution type.
Figure 1. Different types of bi-metallic catalysts [26]
In the present study, the APR reaction has been investigated using a series of novel, highly active
and stable Pt0.05CexZr0.95−xO2 ter-metallic (ternary) solid solutions synthesized by controlled co-
precipitation technique, with emphasis on the influence of Zr4+ dopant on the properties and
performance of the catalysts. To this end, the catalysts were extensively characterized by using XRD,
BET, H2-TPR, CO-chemisorption, TEM, and XPS analyses. The catalytic activity, selectivity, and
stability over time of the catalysts were determined for the APR of glycerol. This is the first application
of Pt0.05CexZr0.95−xO2 ternary solid solution catalysts not only for the APR of glycerol but for any other
catalytic reaction.
Experimental Section
Catalyst preparation. All reagent chemicals were of analytical grade and used without further
purification. Distilled water was used to prepare all aqueous solutions. The novel nano-sized solid
solution catalysts which were studied here are Pt0.05CexZr0.95-xO2 (x=0, 0.29, 0.475, 0.66 and 0.95)
where x represented the nominal weight fraction of Ce in catalyst. Samples were prepared by controlled
precipitation or co-precipitation methods. Hexachloroplatinic acid, (H2PtCl6·6H2O), cerium nitrate
[Ce(NO3)3·6H2O] and zirconium oxychloride (ZrOCl2·8H2O) were used as precursor salts, and sodium
hydroxide (NaOH) was used as precipitant. The co-precipitation process was conducted by the dropwise
addition of 1 M solution of sodium hydroxide under constant stirring into an aqueous solution of
appropriate amount of precursor salts. The temperature was kept at 60˚C and to ensure complete
precipitation the addition of NaOH was continued until the pH reached a value of 12. To achieve the
desired structure of nano-catalysts, in addition to temperature and pH, the rate of addition of
precipitating agent was set so that the entire process of precipitation lasted 16 h. The mixture was
further aged at 25˚C for 12 h. Samples were then filtered and washed using hot distilled water until no
Cl¯ was detected by AgNO3 solution. The precipitate was then dried at 100˚C for 24 h and calcined at
6

600˚C for 5 h. For the purpose of comparison, pure CeO2 and ZrO2 samples were also prepared
following the same procedure.
Catalyst characterization. X-ray diffraction (XRD) patterns were obtained using a PW3040/60 X’Pert
PRO MPD (PANalytical) diffractometer equipped with a Cu Kα radiation source (λ= 1.5406 Å) and
operating at 40 kV and 40 mA. A continuous mode was used for collecting data in the 2θ range of 20˚ to
100˚ at a scanning speed of 10˚ min-1. Brunauer-Emmett-Teller (BET) surface areas of the catalysts
were evaluated from nitrogen adsorption-desorption isotherms at -196˚C using a Belsorp mini II
apparatus. Samples were gassed out at 120˚C for 15 h prior to analysis. Transmission electron
microscopy (TEM) micrographs were obtained using a Zeiss EM900 transmission electron microscope
with accelerating voltage of 80 keV. TEM images were analyzed using the Digital Micrograph software
package (Gatan, Inc.) with at least 100 particles imaged and analyzed for each sample to determine the
average diameter. H2 temperature-programmed reduction (H2-TPR) was performed on Micromeritics
TPR/TPD 2900 apparatus. The samples were first purged with helium at 300˚C for 90 min. After
cooling to room temperature, the flowing gas was switched to a 5 vol% H2/Ar and the catalyst was
heated from 25˚C to 800˚C with a ramp of 10˚C min-1. The amount of H2 consumption during the
process of reduction was measured by a thermal conductivity detector (TCD). Chemisorption uptakes
were measured using pulse methods (BELCAT-A, BEL, Japan). Prior to the chemisorption analysis,
samples were treated in O2 flow by increasing the temperature to 300˚C at 30˚C min-1 and holding at
300˚C for 15 min. The sample was then reduced in H2 flow at 500˚C for 1 h and subsequently cooled to
25˚C at 30˚C min-1 in order to remove the chemisorbed hydrogen. X-ray photoelectron spectroscopy
(XPS) analysis was performed on a Physical Electronics Systems Quantum 2000 spectrometer using
monochromatic Al Kα radiation.
Catalyst performance. APR of glycerol was carried out in a continuous tubular fixed bed reactor
(Figure 2). Glycerol was chosen as a model compound since it has a C:O ratio of 1:1 commonly found
in biomass-derived sugar compounds. 250 mg of the powder catalyst was packed into the reactor that
was a vertical stainless steel tube with an inner diameter of 5 mm. In order to hold the catalyst bed in
position, two plugs of quartz wool were used. The reactor was placed inside a tube furnace equipped
with type K thermocouple and a PID temperature controller. The reactor pressure was controlled using a
dual phase, diaphragm-type back pressure regulator (Equilibar, LLC). Prior to the start of each
experiment, the catalyst was reduced in-situ under H2 of 50 mL min-1 for 2 h at 500 ˚C using a
temperature ramp rate of 1 ˚C min-1. The reactor was subsequently purged with 100 mL min-1 flow of
N2 for 5 h. After cooling to the reaction temperature of 250˚C, the reactor pressure system was set to the
desired pressure of 50 bar. A Waters 510 HPLC digital pump was used to feed the glycerol solution
(WHSV of 2.45 h-1) into the reactor. Once the reactor system had stabilized (about 2 h after the start of
7

a run), the reaction products were analyzed. The gaseous products were swept by N2 as a carrier gas and
analyzed continuously using an on-line Varian CP-3800 gas chromatograph that was equipped with a
Hayesep Q and a Molecular Sieve column. For each experiment, at least five gas samples were analyzed
and checked for relative standard deviations. Only four peaks representing H2, N2, CH4, and CO2, were
detected in the product gases. The liquid effluent collected in a gas-liquid separator was subsequently
analyzed for elemental carbon by a Shimadzu HPLC instrument equipped with a Rezex RCM-
Monosaccharide column using ultrapure deionized water as an eluent at 0.5 mL min-1. The catalyst
performance was evaluated based on H2 yield, H2 selectivity, CH4 selectivity, carbon conversion to gas,
and glycerol conversion which calculated according to:
All data are mean values of ≥3 experiments and error bars indicate one standard deviation.
Figure 2. Schematic of the reactor system
Results and Discussion
XRD. The XRD patterns for CeO2, ZrO2 and Pt0.05CexZr0.95-xO2 (x=0, 0.29, 0.475, 0.66 and 0.95)
samples are shown in Figure 3(A). A closer look in the 27-32˚ [Figure 3(B)] and 32-37˚ [Figure 3(C)]
regions allows one to observe that the pattern of Pt0.05Ce0.95O2 and Pt0.05Zr0.95O2 did not correspond to
any of the pure metal oxide standards and showed a slight shift of characteristic pure oxide peaks
towards higher angles. This suggests that some Zr4+ (0.84 Å) and Ce4+ (0.97 Å) ions had been displaced
by smaller Pt2+ (0.80 Å) or Pt4+ (0.63 Å) ions, leading to a contraction of CeO2 and ZrO2 lattices. Thus,
8

it should cause a decrease of the value of the lattice parameter (Table 1). Also, the pattern of the
Pt0.05CexZr0.95-xO2 (x=0.29, 0.475 and 0.66) nano-structures did not correspond to any of the
Pt0.05Ce0.95O2 and Pt0.05Zr0.95O2 catalysts. The XRD patterns of Pt0.05CexZr0.95-xO2 (x=0.29, 0.475 and
0.66) catalysts are similar to that of Pt0.05Ce0.95O2 but are shifted to higher angles with increasing Zr
content. This indicates that Ce4+ (0.97 Å) is partially substituted with Zr4+ (0.84 Å) in Pt0.05CexZr0.95-xO2
(x=0.29, 0.475 and 0.66) catalysts to form a ternary solid solution.
Table 1 summarizes the size of the crystallites and their lattice parameters as evaluated by Scherrer’s
equation (The shape factor was assumed to be 0.9. For cubic phase, lattice parameter is calculated using
plane with Miller indices (3 1 1) and for tetragonal phase, α = β and γ parameters are calculated using
planes with Miller indices (2 0 0) and (1 0 1), repectively.) Particle sizes in the nanometer range were
obtained for each material. Characteristic tetragonal and cubic phases were detected in pure ZrO2 and
CeO2, respectively. As shown in Table 1, nano-crystalline Pt0.05CexZr0.95-xO2 solid solutions had either
cubic (ICDD card No. 80-0785) or tetragonal (ICDD card No. 28-0271) phase depending upon ceria
content. Furthermore, the average grain size from Scherrer’s formula was used as reference. The
average crystallite size for the tetragonal phase gradually decreased from 9.2 nm (Pt0.05Zr0.95O2) to 4.8
nm (Pt0.05Ce0.475Zr0.475O2) with increasing x value, while that of the cubic phase increased from 6.6 nm
(Pt0.05Ce0.66Zr0.29O2) to 15.2 nm (Pt0.05Ce0.95O2), which can be attributed to the inhibition of crystal
growth for mixed oxides. The addition of ZrO2 into the lattice of CeO2 lowers the crystallinity of
Pt0.05CexZr0.95-xO2 solid solution.
Figure 3. (A) XRD patterns of catalysts and comparison of XRD patterns in (B) 27-32˚ region and (C) 32-37˚
region.
Table 1. Textural properties of catalysts measured by XRD and BET analyses.
BET. N2 adsorption-desorption isotherms of all samples are presented in Figure 4. The isotherms of
the ZrO2 and Pt0.05CexZr0.95-xO2 (x=0, 0.29, 0.475 and 0.66) showed a type IV pattern (IUPAC
classification) with a triangular hysteresis loop of type H2 that is typical of materials with mesoscale
pore network connectivity. The CeO2 and Pt0.05Ce0.95O2 samples showed typical type II isotherms with
H3 hysteresis loop. The pore size distributions of all samples were calculated from the N2 adsorption
isotherms using the BJH analysis. Unlike CeO2 and Pt0.05Ce0.95O2 catalysts, ZrO2 and Pt0.05CexZr0.95-xO2
(x=0, 0.29, 0.475 and 0.66) samples showed a narrow distribution.
Table 1 summarizes the BET specific surface area, pore volume, and pore diameter of all catalysts. It
can be seen that the surface areas of Pt0.05Ce0.95O2 and Pt0.05Zr0.95O2 catalysts were considerably greater
compared with those of ZrO2 or CeO2 which is consistent with the formation of Pt0.05Ce0.95O2 and
Pt0.05Zr0.95O2 solid solutions which corroborated the interpretation of the XRD results.
9

The tensile strain was calculated using Williamson-Hall analysis which relies on the principle that
the total peak broadening (βtot) is sum of size broadening (βsize) and strain broadening (βstrain).
Determining the particle size contributions to the line broadening from kλ/D cosθ (Scherrer equation)
and strain contributions to the line broadening from 4ε tanθ [27], we have βtot cosθ = kλ/D + 4ε sinθ in
which ε stands for strain. βtot cosθ was plotted with respect to 4 sinθ and strain is calculated from the
slope of the fitted line. The tensile strain for the Pt0.05Ce0.95O2 and Pt0.05Zr0.95O2 lattices were less than
those of ZrO2 and CeO2 which has been associated with lower surface energy and hence lower tendency
of the crystals to form agglomerates. Pt0.05Ce0.29Zr0.66O2 and Pt0.05Ce0.95O2 catalysts showed the highest
and lowest BET specific surface area, respectively. On the other hand, the pore volume increased with
increasing Zr content in the cubic and tetragonal structures. Upon Pt addition, pore diameter of CeO2
and ZrO2 had decreased about 30%. For Pt0.05CexZr0.95-xO2 (x=0.29, 0.475 and 0.66) catalysts the
average pore diameter decreased with increasing zirconia content.
Figure 4. N2 physisorption isotherms and corresponding pore size distributions.
H2-TPR. Since the redox characteristics of a catalyst is closely related to its activity, H2-TPR was
performed to determine the redox behavior of CeO2, ZrO2, and Pt0.05CexZr0.95-xO2 samples (Figure 5).
The ZrO2 support showed no reduction peaks up to 800˚C, while pure CeO2 showed one reduction peak
at around 560˚C which could be assigned to the reduction of surface ceria. There is no evidence of
peaks related to reduction of PtO (50˚C), PtO2 (100˚C) and with CeO2 (600˚C) in Pt0.05CexZr0.95-xO2
samples and thus all the observed peaks for Pt0.05CexZr0.95-xO2 samples are associated with the reduction
of mixed oxides [28].
The TPR profile of the Pt-containing CeO2 and ZrO2 were significantly modified. For Pt0.05Ce0.95O2
sample, the presence of a reduction peak at a lower temperature (170˚C) suggests that the reduction
behavior of CeO2 was further improved by introducing Pt2+ ions into the ceria lattice leading to higher
oxygen mobility and consequently easier bulk reduction of the sample. Similarly, Pt0.05Zr0.95O2 sample
shows a reduction peak at 270˚C due to the presence of Pt4+ ions in the zirconia lattice. For
Pt0.05CexZr0.95-xO2 (x=0.29, 0.475 and 0.66) catalysts, reduction peaks were shifted to higher
temperatures as compared with the Pt0.05Ce0.95O2 catalyst. Pt0.05CexZr0.95-xO2 (x=0.475 and 0.66)
catalysts had two reduction peaks at around 170 and 220˚C, attributed to introducing Pt2+ and Pt4+ ions
into the ceria-zirconia lattice, respectively. The detailed discussion of Pt oxidation state of all samples
will be reported in the section dealing with XPS analyses.
With the assumption that the initial Pt oxidation state in the catalyst is Pt4+ (PtO2), the nominal
hydrogen consumptions and the total hydrogen uptake from TPR are presented in Table 2. For the
Pt0.05Zr0.95O2 catalyst, it can be seen that the nominal hydrogen consumption is quite close to the actual
10

consumption suggesting that the reduction peak at 270˚C is due to the reduction of PtO2. On the other
hand, for Pt0.05CexZr0.95-xO2 (x=0.29, 0.475, 0.66 and 0.95) catalysts, the actual hydrogen consumptions
were significantly higher compared with nominal values probably due to the reduction of CeO2-ZrO2.
This indicates that the presence of Pt promotes the reduction of ceria via spillover of hydrogen from
platinum to the ceria thus shifting the peak for surface ceria reduction to lower temperatures. Also, the
reduction peaks of ceria-rich catalysts shifted towards lower temperatures in comparison with the
Pt0.05Zr0.95O2 sample indicating that the reduction of the catalysts strongly depends on the CeO2 loading.
Furthermore, the H2 consumption increased with increasing ceria content of Pt0.05CexZr0.95-xO2 solid
solutions.
Figure 5. H2-TPR profiles of the catalysts.
Table 2. Results obtained by CO chemisorption and H2-TPR analyses.
CO chemisorption. CO chemisorption results of Pt-containing samples are summarized in Table 2.
Pt0.05Ce0.475Zr0.475O2 and Pt0.05Ce0.95O2 samples showed the highest and lowest Pt dispersion,
respectively, and consequently Pt0.05Ce0.475Zr0.475O2 had the smallest while Pt0.05Ce0.95O2 had the largest
crystallite size.
TEM. The representative TEM images of fresh (as-calcined) catalysts are shown in Figure 6. The
particles did not show agglomeration, however, some zones with high and low concentration were
observed for the synthesized catalysts. The TEM images suggest that all catalysts exhibited a nano-
crystal nature and that the molar ratio of Ce/Zr had a significant effect on the particle size as indicated
by the particle size distribution provided for each sample. All catalysts demonstrated a narrow particle
size distribution with average particle sizes slightly greater than those calculated from XRD patterns.
Furthermore, the lack of evidence of metallic Pt particles in TEM images confirmed the formation of
ternary solid solution.
Figure 6. TEM images and corresponding particle size distribution of fresh catalysts.
Reactivity studies. An aqueous solution of 10 wt% glycerol was used as feed in catalytic reactor
tests performed at 250˚C, 50 bar, and WHSV of 2.45 h-1 for 50 h in a fixed bed reactor. The
experimental results are summarized in Table 3. The glycerol conversion over ternary solid solution
catalysts was slightly higher than those of Pt-containing CeO2 and ZrO2. Pt0.05Ce0.475Zr0.475O2 catalyst
had the highest activity in APR of glycerol, Pt0.05Zr0.95O2 displayed moderate activity, and Pt0.05Ce0.95O2
was the least active. The APR activity of Pt catalysts has been correlated with their WGS activity since
efficient CO removal from the catalyst surface is important at the relatively low reaction temperatures
11

employed under APR conditions [16,29-31]. One explanation for this observation is that improved
surface oxygen mobility could enhance the rate of WGS reaction. The high APR activity observed in
Pt0.05CexZr0.95-xO2 catalysts could therefore be explained by considering that introducing the more
electronegative Zr atoms (compared with Ce atoms) in the ceria lattice could induce significant lattice
distortions thus increasing both the oxygen mobility and the concentration of oxygen vacancies in the
lattice [16,17,32] as was also evidenced by the H2-TPR (Figure 5) and XPS (Figure 8) results. Another
possible explanation for the high APR activity of Pt0.05CexZr0.95-xO2 catalysts could be their resistance to
the negative effect of CO2 re-adsorption that could lead to the formation of various carbonate species.
Pt0.05CexZr0.95-xO2 catalysts are less sensitive to irreversible deactivation by CO2 compared with
Pt0.05Zr0.95O2 and Pt0.05Ce0.95O2 catalysts [33].
Ternary solid solution catalysts produced higher hydrogen yield, carbon conversion to gas, and
hydrogen selectivity than binary ones except in the case of Pt0.05Ce0.66Zr0.29O2. For most of the catalysts,
the total H2 yields were lower than what was expected in terms of the stoichiometric ratio of H2 to CO2
in the overall reforming reaction (=7/3). This was due to the H2 consumption in the hydrogenation of
dehydrated reaction intermediates. The large H2/CO2 ratio of Pt0.05Ce0.95O2 was an exception that was
due to the low hydrogenation activity of this catalyst. The lowest H2 selectivity and yield, as well as
carbon conversion to gas was achieved by Pt0.05Ce0.95O2 which could be explained by its low surface
area and metal dispersion (Tables 1 and 2). Additionally, in the absence of Zr, interactions between the
CeO2 and Pt could be induced thus enhancing the oxidation of Pt [34] that would consequently lead to a
less active metal surface [35].
The product gas distributions from the catalytic reactor tests are presented in Table 3. H2, CO2, and
CH4 were the only gaseous products. No CO was detected in the product gases indicating that the CO
concentrations were below the GC detection limit (i.e., <100 ppm). Pt0.05Ce0.475Zr0.475O2 and
Pt0.05Ce0.29Zr0.66O2 catalysts led to H2-rich product gas streams containing 70 and 69 mol% H2,
respectively. The fraction of H2 in the product gases was strongly correlated with the active metal
surface area of the catalyst (Table 2) confirming the important effect of particle size on catalyst
performance. The gas product from APR over Pt0.05Ce0.475Zr0.475O2 catalyst contained less CH4
indicating a lower rate for methanation reaction. This catalyst also had the highest normalized rate of H2
-1
production, 65 cm3 gcat min-1. H2 production rate increased with increasing Pt dispersion as shown in
Tables 2 and 3. Low Pt dispersion, and consequently large particle size, favored the methanation
reaction.
Table 3. Performance of catalysts in APR of 10 wt% glycerol under conditions of 250˚C, 50 bar and WHSV of
2.45 h-1 for 50 h in a fixed bed reactor.
12

Contribution of this work with respect to the work of other researchers are summarized in Table 4.
Comparison between Table 3 and Table 4 shows higher H2 yield and H2 selectivity of present catalysts.
-1
H2 production rate in terms of cm3 gcat min-1 for Pt0.05CexZr0.95-xO2 solid solutions is 3 times higher
than those reported by Jeon and co-workers [36] for Pt/CeO2-ZrO2 hetero-structures in which Pt was
impregnated on CeO2-ZrO2 supports. Turnover frequency (TOF) of our solid solutions in the present
study were 18 times higher than those for hetero-structures reported by Ciftci and co-workers [37].
Table 4. Results of APR of glycerol under different experimental conditions
The reactor test results clearly indicated the advantage of ternary solid solution catalysts in the Pt-
catalyzed APR. Zr addition to Ce resulted in a significant increase in the active metal surface area of the
Pt catalysts (Table 2) which consequently resulted in improved H2 yield, H2 selectivity, CH4 selectivity,
and carbon to gas conversion such that each additional m2g-1 of active metal surface area resulted in an
approximate increase of 5.2% in the yield of H2, 2.5% increase in selectivity of H2, 3.2% increase in
carbon to gas conversion as well as 2.5% decrease in CH4 formation [Figure 7(A)]. Moreover, the
addition of Zr to Ce (up to 47.5 wt% of Zr in Pt0.05CexZr0.95-xO2 structure) resulted in the formation of
smaller particles which in turn led to improved catalytic performance [Figure 7(B)] suggesting that the
particle size of the catalyst might be one of the most important controlling factors for aqueous phase
reforming. Among the catalysts tested, Pt0.05Ce0.475Zr0.475O2 showed the best catalytic performance
which may be attributed to the highest Pt dispersion and the synergistic effect of smallest particle size
and large pore diameter (Tables 1 and 2) that was also consistent with higher active metal surface area
measured from CO-chemisorption (Table 2). These along with the XPS results to follow could account
for the favorable performance of this catalyst.
An inverse dependence of the TOF and metal dispersion on active metal surface area [Figure 7(C)]
and average particle size [Figure 7(D)] is observed, indicating that APR of glycerol is a structure-
sensitive reaction, i.e. adsorption and C–C cleavage of polyols preferably occurred on the face Pt atoms
rather than on edge and corner atoms. As the particle size increases, the number of face atoms increases
whereas the number of corner and edge atoms decrease [38]. Therefore, assuming that face Pt atoms are
much more active in APR than edge and corner atoms, the TOF should increase with an increase of the
average particle size. The results regarding the TOF dependence on the average particle size and active
metal surface area are in very good agreement with the results reported by other investigators [39,40].
Figure 7. Correlation between the results of aqueous phase reforming and (A, C) active metal (Pt) surface area
and (B, D) particle size. Aqueous phase reforming was performed using 0.05 mL h-1 of 10 wt% aqueous glycerol
as the feedstock (250 ˚C, 50 bar, 250 mg catalyst) with data averaged over 50 h of reaction.
13

XPS. In order to see if the solid solution is itself the catalytically active phase and Pt is not extruded
from the solid solution under the redox conditions of APR reaction, XPS measurements were performed
to determine the chemical state of platinum in the nano-catalysts. XPS spectra of the Pt(4f) core level
region in fresh (as-calcined) and used (after APR of 10 wt% glycerol under conditions of 250˚C, 50 bar
and WHSV of 2.45 h-1 for 50 h in a fixed bed reactor) catalysts are presented in Figure 8. All samples
exhibited two broad energy bands corresponding to the Pt 4f7/2 (at lower binding energies) and Pt 4f5/2
(at higher binding energies) core electrons, whose position strongly depended on the Zr/Ce ratio. The
binding energies of the Pt 4f7/2 and Pt 4f5/2 core electrons shifted toward higher values as the Zr content
of the sample increased as a result of the replacement of the less electronegative Ce by the more
electronegative Zr in the solid solution. The increase of the binding energies of Pt 4f electrons may be
ascribed to the electronic interactions between the Pt, Ce and/or Zr atomic orbitals, leading to partial
electron transfer from Pt to Ce and Zr. Such variation of the binding energy has also been reported for
other Pt-based bi-metallic nano-particles [41,42].
According to what is explained above, in the samples with greater Zr content, Pt has become
relatively poor in electron density. Such electron withdrawing from Pt would result in moderating the
strength of the Pt-CO bonds formed during APR reaction subsequently leading to an enhancement in
WGS reaction and a more facile removal of surface adsorbed intermediates [43] thus increasing the
APR reaction activity. In addition, the intensity of the Pt 4f peaks of the samples strongly depended on
the surface content of Pt as determined by CO chemisorption (Table 2). Considering the full width at
half maximum and 4f(7/2,5/2) spin-orbit splitting value, Pt (4f7/2,5/2) peaks in all samples were de-
convoluted into only two sets of spin-orbit doublets which could be assigned to Pt2+ and Pt4+. This
observation agrees well with the XRD and TEM results. Moreover, Pt has a multiple oxidation state
which is determined by the extent of defects formation in the final crystallized sample. Crystallization
of ZrO2 results in Zr in +4 oxidation state and with zirconia being highly irreducible, Pt can be
substituted in the Zr lattice in +4 oxidation state without any defect formation and thus Pt remains
mainly in +4 state in the Pt0.05Zr0.95O2 sample. Conversely, Pt in +2 state was dominant in Pt0.05Ce0.95O2
sample because both +3 and +4 oxidation states exist for Ce thus increasing the possibility of defect
formation in the case of Pt0.05Ce0.95O2. In Pt0.05CexZr0.95-xO2 (x=0.29, 0.475 and 0.66) catalysts the
proportion of Pt in +4 state increased with increasing Zr content.
Analysis of the spent catalysts after APR reaction indicated no significant changes in the oxidation
state of Pt or in the peak intensities. Ceria has high potential to facilitate the oxidization of Pt and
stabilizing it in its’ oxidized state because of the higher redox potential of the Ce4+/Ce3+ couple (1.61
eV) compared with those of the Pt2+/Pt0 (1.19 eV) and Pt4+/Pt2+ (1.05 eV) couples [44]. The introduction
of Pt into ceria–zirconia solid solution instead of impregnating Pt on ceria–zirconia mixed oxide
support, appears to be capable of stabilizing Pt in its’ oxide form against the APR redox conditions.
14

This is an important observation since it is well known that Pt segregates out from the Pt impregnated
ceria-zirconia catalysts under APR conditions [6,35,45].
Figure 8. High-resolution XPS spectra of Pt 4f region of fresh and spent catalysts.
Stability studies. The catalyst stability was evaluated in terms of the hydrogen yield [Figure 9(A)].
The catalytic performance remained constant during at least 50 h of the APR reaction of 10 wt%
glycerol under conditions of 250˚C, 50 bar, and WHSV of 2.45 h-1 indicating that the catalysts had
excellent stability without any degradation as was also evident from XPS analysis. The representative
TEM images and particle size distribution of the spent catalysts are shown in Figure 10. No evidence of
agglomeration was observed for the spent catalysts which had similar narrow size distributions and
average particle sizes close to those of the fresh samples. This was also confirmed by both XPS and
TEM results of the spent samples.
The Pt0.05Ce0.475Zr0.475O2 catalyst, which resulted in the highest H2 yield from the APR of glycerol in
the catalyst screening studies, was used to investigate the effects of WHSV on the reaction activities at
fixed catalyst loading (250 mg). Figure 9(B) presents the results from APR of glycerol solutions with
six different WHSV (0.12, 0.6, 1.2, 1.8, 2.4, 3 and 3.6 h-1). The APR of glycerol at WHSV=2.4 h-1
produced a significantly higher H2 yield than other WHSV values tested. At WHSV>2.4 h-1
methanation was mitigated and the WGS reaction was favored. At WHSVs less than or equal to 2.4 h-1,
sufficient contact between the glycerol reactant and the catalyst allowed the reforming reaction to take
place; the shorter contact time (i.e., for WHSV>2.4 h-1) only resulted in partial conversion of glycerol.
The stability of Pt0.05Ce0.475Zr0.475O2 catalyst was investigated by changing WHSV from 0.12 to 3.6 h−1
[Figure 9(C)]. This catalyst had excellent stability in all examined WHSV values during 50 h of the
APR reaction.
Figure 9. (A) Performance of catalysts during 50 h of continuous APR reaction using 10 wt% glycerol and
WHSV of 2.45 h-1, (B) The effect of the WHSV on H2 yield APR over Pt0.05Ce0.475Zr0.475O2 and (C) Performance
of Pt0.05Ce0.475Zr0.475O2 during 50 h of continuous APR reaction using various WHSV. All reactions conducted
using 250 ˚C, 50 bar and 250 mg catalyst.
Figure 10. TEM image and corresponding particle size distribution of spent catalysts.
Conclusions
Glycerol can be converted into H2-rich gas over Pt catalysts. Reactivity and characteristic
measurements on Pt0.05CexZr0.95-xO2 with varying compositions of Zr and Ce revealed notable
15

differences in the catalytic activities and product distributions from the APR reaction. Ternary solid
solution catalysts showed remarkably enhanced catalytic performance towards the APR of glycerol. Of
these catalysts, Pt0.05Ce0.475Zr0.475O2 with high Pt dispersion and small particle size, along with the large
pore diameter and high active metal surface area was found to be very active and stable in APR of
glycerol, had the highest carbon to gas conversion (95%), highest H2 yield (93%), highest H2 selectivity
(98%) and at least 50 h of stability. Achieving a highly active and stable catalyst for the aqueous phase
reforming of glycerol was the focus of our current investigation.
16

References
[1] S. Dunn, Int. J. Hydrogen Energy 27 (2002) 235-264.
[2] R. D. Cortright, R. R. Davda, J. A. Dumesic, Nature 418 (2002) 964-967.
[3] G. W. Huber, J. A. Dumesic, Catal. Today 111 (2006) 119-132.
[4] R. R. Davda, J. W. Shabaker, G. W. Huber, R. D. Cortright, J. A. Dumesic, Appl. Catal. B 56 (2005) 171-
186.
[5] R. R. Davda, J. A. Dumesic, Chem. Commun. (2004) 36-37.
[6] J. W. Shabaker, G. W. Huber, R. R. Davda, R. D. Cortright, J. A. Dumesic, Catal. Lett. 88 (2003) 1-8.
[7] J. W. Shabaker, R. R. Davda, G. W. Huber, R. D. Cortright, J. A. Dumesic, J. Catal. 215 (2003) 344-352.
[8] G. W. Huber, J. W. Shabaker, S. T. Evans, J. A. Dumesic, Appl. Catal., B 62 (2006) 226-235.
[9] R. L. Manfro, A. F. da Costa, N. F. P. Riberio, M. M. V. M. Souza, Fuel Process. Technol. 92 (2011) 330-
335.
[10] J. W. Shabaker, G. W. Huber, J. A. Dumesic, J. Catal. 222 (2004) 180-191.
[11] G. Wen, Y. Xu, H. Ma, Z. Xu, Z. Tian, Int. J. Hydrogen Energy 33 (2008) 6657-6666.
[12] A. Ciftci, B. Peng, A. Jentys, J. A. Lercher, E. J. M. Hensen, Appl. Catal., A 431 (2012) 113-119.
[13] K. Koichumanova, A. K. K. Vikla, D. J. M. de Vlieger, K. Seshan, B. L. Mojet, L. Lefferts, ChemSusChem
6 (2013) 1717-1723.
[14] G. Wen, Y. Xu, Z. Xu, Z. Tian, Catal. Lett. 129 (2009) 250-257.
[15] C. M. Kalamaras, D. D. Dionysiou, A. M. Efstathiou, ACS Catal. 2 (2012) 2729-2742.
[16] C. I. Vignatti, M. S. Avila, C. R. Apesteguia, T. F. Garetto, Catal. Today 171 (2011) 297-303.
[17] Y. T. Kim, S. J. You, E. D. Park, Int. J. Hydrogen Energy 37 (2012) 1465-1474.
[18] H. S. Roh, H. S. Potdar, K. W. Jun, S. W. Han, J. W. Kim, Catal. Lett. 93 (2004) 203-207.
[19] J. L. Ayastuy, M. P. Gonzalez-Marcos, A. Gil-Rodriguez, J. R. Gonzalez-Velasco, M. A. Gutierrez-Ortiz,
Catal. Today 116 (2006) 391-399.
[20] M. M. Rahman, T. L. Church, A. I. Minett, A. T. Harris, ChemSusChem 6 (2013) 1006-1013.
[21] M. L. Barbelli, F. Pompeo, G. F. Santori, N. N. Nichio, Catal. Today 213 (2013) 58-64.
[22] A. O. Menezes, M. T. Rodrigues, A. Zimmaro, L. E. P. Borges, M. A. Fraga, Renewable Energy 36 (2011)
595-599.
[23] Y. Guo, M. U. Azmat, X. Liu, Y. Wang, G. Lu, Appl. Energy 92 (2012) 218-223.
[24] A. Wawrzetz, B. Peng, A. Hrabar, A. Jentys, A. A. Lemonidou, J. A. Lercher, J. Catal. 269 (2010) 411-420.
[25] M. El Doukkali, A. Iriondo, J. F. Cambra, L. Jalowiecki-Duhamel, A. S. Mamede, F. Dumeignil, P. L. Arias,
J. Mol. Catal. A: Chem. 368 (2013) 125-136.
[26] V. Dal Santo, A. Gallo, A. Naldoni, M. Guidotti, R. Psaro, Catal. Today 197 (2012) 190-205.
[27] Y. T. Prabhu, K. V. Rao, V. S. S. Kumar, B. S. Kumari, World J. Nano Sci. Eng. 4 (2012) 21-28.
17

[28] P. Fornasiero, J. Kaspar, T. Montini, M. Graziani, V. Dal Santo, R. Psaro, S. Recchia, J. Mol. Catal. A:
Chem. 204-205 (2003) 683-691.
[29] A. Ciftci, D. A. J. M. Ligthart, A. O. Sen, A. J. F. van Hoof, H. Friedrich, E. J. M. Hensen, J. Catal. 311
(2014) 88-101.
[30] E. L. Kunkes, D. A. Simonetti, J. A. Dumesic, W. D. Pyrz, L. E. Murillo, J. G. Chen, D. J. Buttrey, J. Catal.
260 (2008) 164-177.
[31] D. A. Simonetti, E. L. Kunkes, J. A. Dumesic, J. Catal. 247 (2007) 298-306.
[32] S. Ricote, G. Jacobs, M. Milling, Y. Ji, P. M. Patterson, B. H. Davis, Appl. Catal. A 303 (2006) 35-47.
[33] M. W. Balakos, M. R. Madden, T. L. Walsh, J. P. Wagner, Proceedings of the AIChE Spring National
Meeting, New Orleans, LA, 2004.
[34] Q. Zhuang, Y. Qin, L. Chang, Appl. Catal. 70 (1991) 1-8.
[35] Y. F. Y. Yao, Ind. Eng. Chem. Prod. Res. Dev. 19 (1980) 293-298.
[36] S. Jeon, H. Ham, Y. Suh, J. W. Bae, RSC Adv. 5 (2015) 54806-54815.
[37] A. Ciftci, S. Eren, D. A. J. M. Ligthart, E. J. M. Hensen, ChemCatChem 6 (2014) 1260-1269.
[38] A. V. Kirilin, B. Hasse, A. V. Tokarev, L. M. Kustov, G. N. Baeva, G. O. Bragina, A. Y. Stakheev, A.
Rautio, T. Salmi, B. J. M. Etzold, J. Mikkolaae, D. Y. Murzin, Catal. Sci. Technol. 4 (2014) 387-401.
[39] K. Lehnert, P. Claus, Catal. Commun. 9 (2008) 2543-2546.
[40] T. Kim, H. Kim, K. Jeong, H. Chae, S. Jeong, C. Lee, C. Kim, Green Chem. 13 (2011) 1718-1728.
[41] J. B. Xu, T. S. Zhao, Z. X. Liang, J. Phys. Chem. C 112 (2008) 17362-17367.
[42] J. H. Zeng, J. Yang, J. Y. Lee, W. J. Zhou, J. Phys. Chem. B 110 (2006) 24606-24611.
[43] K. W. Park, J. H. Choi, B. K. Kwon, S. A. Lee, Y. E. Sung, H. Y. Ha, S. A. Hong, H. Kim, A. Wieckowski,
J. Phys. Chem. B 106 (2002) 1869-1877.
[44] J.A. Dean, Lange’s Handbook of Chemistry, fifteenth ed., McGraw-Hill, New York, 1998.
[45] H. J. Park, H. D. Kim, T. W. Kim, K. E. Jeong, H. J. Chae, S. Y. Jeong, Y. M. Chung, Y. K. Park, C. U.
Kim, ChemSusChem 5 (2012) 629-633.
18

BET
XRD
a
b
Crystallite size (nm)
Lattice parameter (Å)
BET surface
area (m2 g-1)
Pore volume
(cm3 g-1)
Pore diameter
(nm)
Catalyst
Lattice type
cubic
tetragonal
α = β
γ
Crystallinity (%)
CeO2
Pt0.5Ce0.95O2
Pt0.5Ce0.66Zr0.29O2
Pt0.5Ce0.475Zr0.475O2
Pt0.5Ce0.29Zr0.66O2
Pt0.5Zr0.95O2
cubic
cubic
cubic
tetragonal
tetragonal
tetragonal
tetragonal
15.1
15.2
6.6
-
-
-
-
-
-
-
5.42
5.41
5.38
3.73
3.67
3.61
3.62
5.42
5.41
5.38
5.27
5.22
5.09
5.12
74
66
53
48
40
40
31
13
38
91
81
102
101
64
0.06
0.12
0.15
0.13
0.13
0.17
0.16
18
12
7
6
5
4.8
5.2
9.2
11.7
7
10
ZrO2
a
Crystallite size is obtained from Scherrer's equation.
Lattice parameter is obtained from Bragg's Law.
b
19

CO chemisorption
H2-TPR
CO uptake
(μmol CO gcat
Mean particle size
(nm)
Pt surface area
Pt dispersion
(%)
H2 consumption
Nominal H2 consumption
-1
(mmol H2 gcat )
a
b
c
-1
-1
-1
)
(m2 gcat
)
(mmol H2 gcat
)
Catalyst
CeO2
Pt0.5Ce0.95O2
Pt0.5Ce0.66Zr0.29O2
Pt0.5Ce0.475Zr0.475O2
Pt0.5Ce0.29Zr0.66O2
Pt0.5Zr0.95O2
-
-
-
-
8
1.3
1.5
2.1
1.7
1.2
0.5
0.3
0.51
0.51
0.51
0.51
0.51
0.51
0.51
21
38
101
90
49
-
13.9
7.6
2.9
3.2
6
1
2
5
4
2
-
15
39
35
19
-
ZrO2
a
-
Mean particle size= Vchem*SF*MW/C
b
c
Pt surface area= Vchem*NA*SF*σm
Pt dispersion=60*C/( Vchem*NA*SF*σm*ρ)
Vchem=Chemisorped volume, SF=Stoichiometry factor (in this case=1), MW=Pt atomic weight, C=wt% of Pt in sample, NA=Avogadro number, σm=Pt cross
section area, ρ=Density of Pt
20

Selectivity (%)
Gas product (mol %)
Glycerol
conversion (%)
Carbon to gas
conversion (%)
c
H2 production rate
(cm3 gcat-1 min-1)
f
a
b
H2 yield (%)
TOF (min-1)
d
e
Catalyst
H2
CO2
CH4
H2
CH4
13
8
Pt0.5Ce0.95O2
99.2
99.4
99.8
99.5
99.3
81
87
95
93
91
71
78
93
89
85
87
89
98
95
93
67
67
70
69
68
29
30
30
30
30
4
3
0
1
2
49
54
65
62
59
105
64
29
31
55
Pt0.5Ce0.66Zr0.29O2
Pt0.5Ce0.475Zr0.475O2
Pt0.5Ce0.29Zr0.66O2
Pt0.5Zr0.95O2
a
2
3
5
Glycerol conversion=amount of glycerol removed from liquid phase/amount of glycerol in feedstock*100
Carbon to gas conversion=number of C atoms in gas product/number of C atoms fed into the reactor*100
H2 yield= H2 produced experimentally/7/glycerol molar flow rate fed into the reactor*100
H2 selectivity=2*H2 produced experimentally/total number of H atoms in gas product*100
CH4 selectivity=CH4 produced experimentally/total number of C atoms in gas product*100
TOF= H2 production rate/moles of chemisorped CO*Stoichiometry factor (in this case=1)
b
c
d
e
f
21

Reaction conditions
a
b
Catalyst
Temperature (˚C)
Pressure (bar)
Glycerol wt%
H2 yield (%)
H2 Selectivity (%)
Reactor type Ref.
3 wt% Pt/3 wt% CeO2/Al2O3
1.1 wt% Pt/4 wt% CeO2/1 wt% ZrO2/α-Al2O3
1.49 wt% Pt/CeO2
0.75 wt% Pt/CeO2
5 wt% Pt/CeO2-ZrO2 (Ce/Ce+Zr=0.4)
220
250
225
225
250
225
30
44
23
27.6
45
5
10
1
5
10
10
56
20
27
21.4
8.5
26.5
88
81
-
58.9
74.3
-
fixed-bed
batch
batch
fixed-bed
fixed-bed
autoclave
20
21
22
23
35
36
2.76 wt% Pt/Ce0.25Zr0.75O2
a
25
H2 yield= H2 produced experimentally/H2 produced theoretically*100
b
H2 selectivity=2*H2 produced experimentally/total number of H atoms in gas product*100
22