Aqueous-phase reforming of xylitol over Pt/C and Pt/TiC-CDC catalysts: Catalyst characterization and catalytic performance

Article abstract of DOI:10.1039/c3cy00636k

The aqueous phase reforming (APR) of xylitol was studied over five Pt/C catalysts. The correlation between physico-chemical properties of the catalysts and catalytic performance was established. The Pt/C catalysts have different textural properties as wel

Full text of DOI:10.1039/c3cy00636k

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Aqueous-phase reforming of xylitol over Pt/C and
Pt/TiC-CDC catalysts: catalyst characterization
and catalytic performance
Cite this: Catal. Sci. Technol., 2014,
4, 387
Alexey V. Kirilin,a Benjamin Hasse,b Anton V. Tokarev,a Leonid M. Kustov,c
Galina N. Baeva,c Galina O. Bragina,c Aleksandr Yu. Stakheev,c Anne-Riikka Rautio,d
a
Tapio Salmi,a Bastian J. M. Etzold,b Jyri-Pekka Mikkolaae and Dmitry Yu. Murzin
*
The aqueous phase reforming (APR) of xylitol was studied over five Pt/C catalysts. The correlation
between physico-chemical properties of the catalysts and catalytic performance was established. The
Pt/C catalysts have different textural properties as well as different mean Pt cluster sizes and surface
acidity. The average Pt cluster size was investigated by means of CO chemisorption as well as by TEM.
The reaction was found to be structure sensitive and TOF linearly increases with increasing average
Pt cluster size in the studied domain. The catalysts which possess higher surface acidity favoured higher
rates of hydrocarbon production. On the contrary the Pt/C materials with lower acidities generated
hydrogen with high selectivity and TOF.
Received 26th August 2013,
Accepted 4th November 2013
DOI: 10.1039/c3cy00636k
www.rsc.org/catalysis
contributions devoted to APR of sugar alcohols is signifi-
cantly less. The main reason for the latter case is a compli-
Introduction
Sustainable evolution of our society is impossible without
constant progress in such areas as chemical engineering and
technology. In the context of a sustainable society biomass
and biomass-derived products in the form of ready-to-use
chemicals are attracting a lot of attention all over the world.1,2
Hereby, much scientific effort was recently put into investiga-
tion of aqueous phase reforming (APR).3
cated product analysis, since for APR of xylitol or sorbitol
more than 25 products are present in the liquid phase25,31 in
non negligible amounts.
The reaction is generally performed at elevated pressures
(29–50 bar) and temperatures (498–523 K). Typically, the reac-
tion pressure is 4 bar above the boiling point of water at a
chosen process temperature. Various metals and support
materials have been tested for the catalytic activity in APR32
and so far Pt is the most used and studied metal possessing
the highest catalytic activity and selectivity towards the desired
products (hydrogen33 or alkanes34). Depending on the support
material and the process conditions (pH, acid additives), the
reaction can be steered towards enhanced hydrogen or hydro-
carbon formation. Different metal oxides4,14,18,32,35,36 can
be used as a support for Pt particles in the APR along with
zeolites21,23,37 and carbon.12,32,38 Carbon supports for Pt
catalysts in the APR process have recently attracted much
scientific attention. Such carbon supports as Vulcan active carbon
(Pt12,32,38,39 and Pt–M, M = Re, Ru, Os, Fe, Cu, Sn, Ir15), activated
carbon (Pt,23,40,41 Pt–Re42,43), carbon fibers (Pt and Pt–Ru),44
single-walled carbon nanotubes SWNTs (Pt and Pt–Co),45–48
ordered mesoporous carbon CMK-3 (Pt,49 Pt–Mn,50 Pt–Re51)
have been successfully investigated in the APR of polyols.
These supports possess high hydrothermal stability under
APR conditions (compared to the most investigated catalysts
Pt on γ-Al2O3, where the alumina support might slowly trans-
form into boehmite during water exposure at elevated tem-
peratures and pressures35,52), large surface area and good
APR is an important heterogeneously catalyzed reaction
which can be used for production of hydrogen and alkanes.
Because of that, APR has gained a lot of interest as a poten-
tial hydrogen and/or alkane supply from renewable biomass,
therefore being an alternative to fossil fuels. APR is the
process for transformation of aqueous solutions of polyols
(such as ethylene glycol,4–11 glycerol,12–19 xylitol20,21 or sorbitol22–26
)
and alcohols (methanol,27,28 ethanol28–30). Numerous publi-
cations can be readily found on ethylene glycol (EG), glycerol
and alcohol reforming. However, the number of scientific
a Laboratory of Industrial Chemistry and Reaction Engineering, Process Chemistry
Centre, Åbo Akademi University, Biskopsgatan 8, FI-20500 Åbo-Turku, Finland.
E-mail: dmurzin@abo.fi; Fax: +358 2 215 4479
b Institute of Chemical Reaction Engineering, University Erlangen-Nuremberg,
Egerlandstraße 3, DE-91058 Erlangen, Germany
c Zelinsky Institute of Organic Chemistry, Leninsky prosp. 47, RU-119991,
Moscow, Russia
d Microelectronics and Materials Physics Laboratories EMPART Research Group of
Infotech Oulu, University of Oulu, FI-90014 Oulu, Finland
e Technical Chemistry, Department of Chemistry, Chemical Biological Centre,
Umeå University, SE-901 87 Umeå, Sweden
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 387

View Article Online
Paper
Catalysis Science & Technology
activity and hydrogen selectivity via APR. Nevertheless, there
are not enough systematic studies devoted to investigation
of the activity of Pt on different types of carbon supports.
Moreover, some important carbon materials such as Sibunit,53
which find a number of industrial applications, have not been
investigated at all. Sibunit is a synthetic carbon composite,
which is produced by pyrolysis of light hydrocarbons on a
granulated carbon black followed by steam activation at
973–1123 K of the composite obtained.54,55 A sponge-like system
including meso- and macropores whose dimensions depend
on the dispersion of the initial carbon black is formed.
Sibunit is a class of porous carbon composite materials com-
bining advantages of graphite (chemical stability and electric
conductivity) and mesoporous active carbon (high specific
surface area and mesopore volume). Therefore, this material
is an attractive support for the preparation of Pt catalysts.
The catalysts supported on Sibunit have been thoroughly
investigated in many reactions, including hydrogenation and
decarboxylation of fatty acids56–58 and hydrogenation59 and
oxidation of volatile organic compounds.60
Additionally, our focus was also on carbide-derived carbons.
The carbide-derived carbon (TiC-CDC) method allows one to
synthesize nanoporous carbons with highly reproducible
material properties. Hereby, e.g. pore size and pore volume,
specific surface area and carbon morphology can be tailored
towards special needs.61–63 These features make TiC-CDC a
promising model material in several fields of application,
such as supercapacitors, gas-storage and separation and as a
tailored catalyst support.64–68
Yet there is a lack of knowledge of the carbonaceous
catalyst features influencing selectivity and yield of the
desired products.32 Moreover, it is worth noting that APR
of such an important substrate as xylitol has not yet been
studied over Pt/C catalysts. In addition, there are no system-
atic investigations of APR of polyols revealing the influence
of the Pt cluster size on the reaction of higher polyols origi-
nating from biomass. It has been shown by Lehnert and
Claus13 that the Pt cluster size has a dramatic effect on
hydrogen production from glycerol. The catalysts with bigger
Pt clusters supported on alumina demonstrated higher TOF values
in terms of the hydrogen production. A recent study by
Kim et al.49 showed that in the case of Pt catalysts supported
on carbonaceous CMK the hydrogen production rate in APR of
ethylene glycol increases with an increase of the metal loading.
Structure sensitivity was therefore observed for C2–C3 polyols
only. For the first time, this phenomenon has been described
in the present work for polyols >C3. On top of that, there is no
systematic study explicitly demonstrating influence of surface
acidity of Pt/C on APR of polyols.
The main purpose of the present work was thus prepa-
ration, characterization of various Pt/C catalysts (supported
on TiC-CDC, Sibunit, active carbon (AC) and birch active
carbon (BAC)) and investigation of their catalytic perfor-
mance in APR of xylitol. Another goal was to correlate the
catalytic results and the physico-chemical properties of the
utilized materials.
Results and discussion
The textural properties of the Pt/C catalyst studied utilized in
the present work study are summarized in Table 1.
Pt/BAC as well as Pt/C (Degussa) possess large surface areas
exceeding 900 m2 g−1. Pt/TiC-CDC has a surface area of 850 m2 g−1.
The Pt/Sibunit catalysts prepared from (NH3)4Pt(HCO3)2
(denoted as Pt/Sibunit (I)) and H2PtCl6 (denoted as Pt/Sibunit
(II)) precursors have specific surface areas of 339 and 408 m2 g−1,
respectively. As can be noticed, the Pt/TiC-CDC sample has
the largest pore volume of 0.61 cm3 g−1 among the catalysts
tested; Pt/C (Degussa) and Pt/Sibunit (I) have similar pore
volumes: 0.52 and 0.53 cm3 g−1, respectively. Meanwhile, Pt/BAC
has a slightly lower pore volume than Pt/Sibunit (II) (0.47 and
0.45 cm3 g−1), as can be seen from Table 1. Pt/TiC-CDC as well
as Pt/C (Degussa) possess average pore diameters in the range
of 1.2–1.5 nm, whereas Pt/Sibunit (I) and Pt/Sibunit (II)
samples have mean pore sizes of 2.9 and 2.6 nm, respectively.
The average size of Pt clusters deposited on carbon supports
was investigated by means of pulse CO chemisorption. The
data are presented in Table 1. As can be seen, the size of the
Pt cluster for all the catalysts was below 10 nm. Pt/TiC-CDC
and Pt/Sibunit (I) have an average metal cluster size of 9.0 nm
which corresponds to the metal dispersion of 12%. The
Pt/Sibunit (II) catalyst prepared from H2PtCl6 has an average
metal cluster size of 2.8 nm (metal dispersion = 41%). The
difference in metal cluster sizes prepared from different metal
salts shows that the metal precursor has an impact on the
dispersion of the final catalyst. A slightly higher dispersion
(47%) was calculated for Pt/C (Degussa) which corresponds
to the mean Pt particle size of 2.4 nm. The highest dispersion
was observed in the case of the Pt/BAC catalyst which has an
average Pt cluster size of 1.5 nm and metal dispersion of
Table 1 Textural and surface properties of the Pt/C catalysts
CO uptake,
−1
Catalyst
Pt content, wt.% Surface area, m2 g−1 Pore volume, cm3 g−1 Mean pore diameter, nm μmol gcat
dPt,a nm D(Pt),a
%
Pt/TiC-CDC
Pt/C (Degussa)
Pt/Sibunit (I)
Pt/Sibunit (II)
Pt/BAC
2.8
5
5
5
5
850
910
339
408
890
110
0.61
0.52
0.53
0.47
0.45
0.25
1.4
1.5
2.9
2.6
1.2
8.2
18.3
119.6
31.3
104.5
191.5
74.6
9.0
2.4
9.0
2.8
1.5
3.9
12
47
12
41
73
29
Pt/Al2O3
5
a
Calculated based on the information provided by CO chemisorption.
388 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
73%. Generally, the larger surface area of the support facili-
tates the formation of smaller particles. However, as can be
seen from Table 1, both the type of the carbon support and
choice of the metal precursor contribute significantly to the
final properties of the catalyst. Furthermore, there are such
parameters as the point of zero charge (PZC), pH of deposi-
tion, and more specifically the driving force during the metal
deposition [(PZC − pH)/PZC], which can also have an influ-
ence on the properties of the catalytic materials.
All Pt/C catalysts were examined by means of transmission
electron microscopy (TEM) to determine the platinum cluster
size distribution on the carbon surface and to compare the
data with average cluster sizes obtained by CO chemisorption
to TEM.
The TEM images of all the catalysts as well as corresponding
particle size distributions are shown in Fig. 1. All the materials
showed monomodal and narrow distribution of nanoparticles.
The data obtained by TEM are in good agreement with the
data obtained by pulse CO chemisorption (Table 2) except
for two samples: Pt/TiC-CDC and Pt/Sibunit (I). For these
exceptions, the average cluster sizes calculated on the basis
of CO chemisorption data are higher than those obtained
from TEM. Minor deviations in the mean Pt cluster values
obtained by two different techniques can be caused by
several reasons.
For example, bigger Pt clusters are difficult to identify
when they melt and vanish into the background, especially
in the case of carbon supported catalysts, or when over-
laying smaller clusters melt in a bigger one. Moreover, during
TEM imaging selected parts of the catalytic surface are imaged
for calculations of the mean particle size and often do not
represent the whole particle size distribution. On the con-
trary, chemisorption provides the total volume of the carbon
monoxide adsorbed by Pt particles; therefore, all the particles
capable of adsorbing CO are taken into account. Recently the
similar issues were addressed by Wang et al.46,48 The values
for average Pt cluster size calculated from hydrogen and
CO chemisorption were much higher compared to the results
obtained by TEM and EXAFS. For instance, in a case of the
Pt/MWNT-r (multi-wall nanotubes) catalyst the Pt particle
size calculated from CO chemisorption was 5.26 nm, whereas
it was 2.2 1.2 and 2.3 1.4 nm according to EXAFS and
TEM, respectively. Moreover, Borchardt et al.65 reported
The Pt/TiC-CDC has an average pore size of 1.4 nm. There-
fore, pore blocking might be the reason for the lower disper-
sion observed by TEM similar to that of Pt/MWNT described
by Wang et al.46 Nevertheless, a final conclusion regarding
a discrepancy between two methods cannot be drawn due
to the influence of many factors described above. As for
Pt/Sibunit (I), the discrepancy can be caused by melting of
the Pt particles into the background. This in turn compli-
cates the identification of metal particles and correct size
determination. Additionally, for both catalysts prepared from
ammonia containing the Pt precursor the discrepancy between
CO chemisorption and TEM data was observed. Probably, both
materials are not completely reduced therefore higher values
are obtained by CO titration compared to TEM analysis.
Following the approach of Wang et al.,46 all turnover frequen-
cies are calculated on the basis of chemisorption data. This
technique is considered to be more relevant for catalysis
compared to TEM data.
Temperature-programmed reduction (TPR) profiles of the
Pt/C catalysts are shown in Fig. 2. The catalysts after prepara-
tion were reduced using various techniques. The data are
provided in the Experimental section. However, as can be
seen from the TPR curves formation of Pt oxide species may
occur. High-temperature hydrogen consumption peaks above
573 K can be assigned to hydrogen spill-over. Another reason
for it might be high temperature reduction (>573 K) of
the functional groups present on the carbon surface. Similar
hydrogen consumption zones were observed for Pt/CMK catalysts.51
Uptake of hydrogen was very low in the case of Pt/BAC, since
during preparation this catalyst was reduced chemically by
HCOOH. During storage Pt is partially oxidized as revealed by
the TPR curve: the peak at 398 K corresponds to the reduction
of PtOx species. Prior to all the experiments the catalysts were
pre-reduced under hydrogen flow for two hours at 523 K.
Therefore, the Pt oxidation state for all the Pt/C catalysts can
be regarded as Pt(0). However, Pt can interact with graphene
sheets in CDC. Foley and co-workers reported that electron-
rich graphene sheets interact with Pt nanoparticles and thus
stabilize them.70
Acidic properties of the materials were investigated by
NH3-TPD. For each sample five NH3-TPD curves at different
heating rates were recorded to calculate the total amount
of NH3 adsorbed and the heat of ammonia desorption. The
typical experimental curves recorded are shown in Fig. 3 for
the Pt/C (Degussa) catalyst.
The data obtained for the Pt/C catalysts are summarized in
Table 3, showing that different Pt catalysts possess different
acidic properties and heats of ammonia desorption.
Rather low values of the heats of desorption indicate that
all the catalysts possess weak acidic sites. As shown in Fig. 3a
for Pt/C (Degussa), the temperature of the peak maximum
(Tp) of ammonia desorption does not exceed 379 K which
implies that the catalyst does not contain strong acid sites or
acid sites of the moderate strength.71 However, the Pt/C catalysts
demonstrated differences in heats of NH3 desorption and
in the total amount of NH3 desorbed, thus indicating minor
Pt cluster size determined by TEM to be 2.5
0.6 nm,
whereas CO chemisorption data provided an even smaller
average cluster size of 1.5 nm, hence, both cases might take
place. The authors partially explained the disagreement in
the results obtained using different techniques by the fact
that some Pt particles cannot be “chemically” seen by probe
molecules. Another reason was a possible blocking of pores
by larger particles. Additionally partially reduced metal clusters
can be seen in TEM; however, they are not active in chemisorp-
tion of CO. In our case, for example, disagreement between
CO chemisorption and TEM results was observed for two
samples: Pt/TiC-CDC and Pt/Sibunit (I). Carbide-derived carbons
are microporous materials with a narrow pore size distribution.69
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 389

View Article Online
Paper
Catalysis Science & Technology
Fig. 1 TEM images and corresponding Pt particle size distribution for a) Pt/BAC, b) Pt/C (Degussa), c) Pt/TiC-CDC, d) Pt/Sibunit (I); e) Pt/Sibunit (II).
390 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 2 Information on average Pt cluster sizes obtained from CO
chemisorption and TEM
Pt content,
wt.%
dPt, nm
(CO chem.)
Catalyst
dPt, nm (TEM)
Pt/TiC-CDC
Pt/C (Degussa)
Pt/Sibunit (I)
Pt/Sibunit (II)
Pt/BAC
2.8
5
5
5
5
9.0
2.4
9.0
2.8
1.5
3.3 1.0
2.8 1.0
3.4 0.6
1.9 2.1
2.0 1.8
Fig. 2 Temperature-programmed reduction profiles for Pt/C catalysts.
differences in the surface acidity of the samples. Generally
speaking, the surface acidity of carbon materials is deter-
mined by the surface chemistry of the carbons, i.e. by the func-
tional groups presenting on the carbon surface72 (e.g. carboxyl,
carbonyl and hydroxyl species). Moreover, higher oxidation
temperatures or a stronger oxidation agent result in more
oxidized carbon materials.73 Compared to other Pt/C catalysts
Pt/TiC-CDC demonstrated the lowest value of the heat of
desorption – 45 kJ mol−1. The Edes calculated for Pt/Sibunit (II),
prepared from hexachloroplatinic acid as a Pt source (54 kJ mol−1),
Fig. 3 NH3-TPD curves registered for Pt/C Degussa (a) and determination
of Edes (b).
Table 3 Acidic properties of Pt/C catalysts
Catalyst
Pt content, wt.% Acidity, μmol g−1 Edes, kJ mol−1
Pt/TiC-CDC
Pt/C (Degussa)
Pt/Sibunit (I)
Pt/Sibunit (II)
Pt/BAC
2.8
5
5
5
5
6
47
103
92
158
317
45
51
53
54
51
52
was almost the same as for Pt/Sibunit (I) (53 kJ mol−1
)
prepared from the hydroxocarbonate Pt complex. The Edes values
calculated for Pt/C (Degussa) and for Pt/BAC were somewhat
lower – 51 kJ mol−1. The total number of acid sites calculated
as the amount of ammonia desorbed is shown in Table 3. The
total acidity of the Pt/C catalysts decreases in the following
order: Pt/BAC > Pt/Sibunit(II) > Pt/Sibunit(I) > Pt/C(Degussa) >
Pt/TiC-CDC.
Thus, the Pt/TiC-CDC sample possessing the lowest value
of Edes has also the lowest acidity. On the other hand, Pt/BAC
containing the highest number of surface acid sites exhibited
Edes of 51 kJ mol−1, the same as that of Pt/C (Degussa). The
fact, that some catalytic materials exhibiting higher values of
Edes possess a lesser number of acid sites can be explained by
the difference in the nature of the carbon material, Pt source
and catalyst preparation technique.
Pt/Al2O3
5
Xylitol reforming comprises destruction of the initial molecule
into H2 and CO. The carbon monoxide formed further reacts
with water via water–gas shift reaction to form CO2 and
additional H2. The APR of xylitol can be therefore envisioned
as a reaction of one mole of xylitol with five moles of water
resulting in formation of eleven moles of H2 and five moles of
CO2 as shown in Scheme 1.
On top of that, the hydrogen formed can participate in
reactions with the initial polyol resulting in formation of
alkanes as well as other products.25,31 Therefore, the key
parameters to study when comparing the catalytic perfor-
mance of different catalysts are: conversion of xylitol, ability
of converting carbon into gas phase products, selectivity to
hydrogen (H2/CO2 ratio), and hydrocarbon selectivity.
Catalytic performance of the Pt/C in APR of xylitol
Catalyst stability studies. APR of xylitol solution was
investigated over the five Pt/C catalysts at 498 K and 29.3 bar.
An important issue to be addressed in APR is the catalyst
stability. There are several factors causing catalyst deactivation
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 391

View Article Online
Paper
Catalysis Science & Technology
are more favorable for selective production of hydrogen
because of less hydrogen consumption. As can be seen from
Fig. 5 displaying the xylitol conversion profile in the APR process
over Pt/TiC-CDC, the conversion of xylitol decreases from 58%
at 1.2 h−1 to 22% at 3.0 h−1 and then further to 11% at 3.6 h−1.
The H2/CO2 ratio indicating the selectivity to hydrogen
increases as conversion of xylitol decreases as displayed in
Fig. 5. At 58% conversion the calculated H2/CO2 ratio was 2
(being close to stoichiometric 2.2 as in Scheme 1); however,
at 11% conversion it increases significantly to ca. 3.
Comparison of different Pt/C catalysts in the APR of xylitol.
Catalysts exhibited a different ability of converting xylitol under
the same experimental conditions (498 K, 29.3 bar, 3.0 h−1)
and the results are collected in Fig. 6.
The highest conversion of xylitol was achieved in the
case of Pt/Sibunit (I) (32%). Pt/TiC-CDC was able to convert
22% of xylitol, whereas the xylitol conversion for Pt/BAC and
Pt/Sibunit (II) was 22 and 15%, respectively. The Pt/C (Degussa)
demonstrated the lowest value of the xylitol conversion among the
Pt/C catalysts investigated – 8%. Thus, the catalysts can be
ranked in the following order on the basis of xylitol conver-
sion: Pt/Sibunit(I) > Pt/TiC-CDC > Pt/BAC > Pt/Sibunit(II) >
Pt/C (Degussa).
Scheme 1 APR of xylitol representing the stoichiometry of H2 and
CO2 formed.
in APR:74 low hydrothermal stability of the support,52 leaching
of the supported metal and formation of coke are among them.
The catalysts were studied for stability performance for
at least 13 h time-on-stream (TOS). Pt/TiC-CDC was studied
for more than 120 h of TOS. All catalysts showed minor or
negligible deactivation with time-on-stream. The results are
summarized in Table 4. For instance, conversion of xylitol
over Pt/TiC-CDC changed from 58 to 47% and selectivity to H2
from 77 to 71% after 120 h TOS in the APR at 498 K, 29.3 bar
and WHSV = 1.2 h−1. For other Pt/C catalysts a similar trend
was observed, however, the changes in xylitol conversion and
selectivity to H2 did not exceed 10% compared to the initial
values observed at 4 h TOS.
Profile of xylitol conversion and H2/CO2 ratio. The formation
rates of the main APR products – H2 and CO2 – over Pt/TiC-CDC
are shown in Fig. 4. It is important to note that with
the decrease of conversion at higher space velocities the
H2/CO2 ratio increases. Therefore, the higher space velocities
Table
4
Time-on-stream behavior of Pt/C catalysts in the APR
of xylitol
Catalyst
TOS, h Xylitol conversion, % Selectivity to H2, %
Pt/TiC-CDC
Pt/C (Degussa)
Pt/Sibunit (I)
Pt/Sibunit (II)
Pt/BAC
10
120
4
16
4
13
4
16
4
58
47
57
55
73
72
57
53
76
74
77
71
67
67
63
62
62
61
60
58
Fig. 5 Conversion of xylitol (■) and H2/CO2 (▲) ratio in the APR process
over the Pt/TiC-CDC catalyst. Conditions: 0.5 g of catalyst, 498 K,
29.3 bar, N2 flow 30 ml min−1, 10 wt.% xylitol solution.
16
Fig. 4 Rate of H2 (light grey columns) and CO2 (dark grey columns)
formation in the APR of xylitol over Pt/TiC-CDC. Conditions: 0.5 g of
catalyst, 498 K, 29.3 bar, N2 flow 30 ml min−1, 10 wt.% xylitol solution.
Fig.
6 Conversion of xylitol over different Pt/C catalysts in APR.
Conditions: 498 K, 29.3 bar, N2 flow 30 ml min−1, 3.0 h−1
.
392 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
The conversion of the substrate in APR does not represent
a transformation per se into the target products, since trans-
formation of the initial feed may occur via several pathways,
including dehydrogenation and dehydration steps as has
been shown earlier for glycerol,17 xylitol25 and sorbitol75
(Scheme 2, on the basis of results reported earlier20,25). The
ratio H2/CO2 is of high significance in the APR process
displaying the potential of Pt/C catalysts to selectively pro-
duce hydrogen. Moreover, as has been mentioned above, the
hydrogen formation is accompanied with production of
CO2 in the APR process. The corresponding H2/CO2 ratios
at xylitol conversions ca. 10–12% are shown in Fig. 7. All
catalysts showed a H2/CO2 ratio above 2.2 with Pt/TiC-CDC
being the most selective (H2/CO2 = 2.95 at 10% conversion).
Slightly less selective behavior in terms of hydrogen produc-
tion was observed for Pt/C (Degussa) (2.65) and Pt/Sibunit (II)
(2.53) samples. The H2/CO2 ratio for Pt/BAC and Pt/Sibunit (I)
was 2.37 and 2.30, respectively. The H2/CO2 ratios observed
for Pt/C are above the theoretical ratio calculated from the
polyol reforming equation as represented in Scheme 2 for
xylitol. APR of polyol does not comprise only the reforming
reaction proceeding via the pathway illustrated in Scheme 1.
In fact, the reaction is more complex25 and includes several
transformations of the substrate resulting in the elimination
of H2 and CO, which are further transformed into CO2 and
H2O via WGS reaction and formation of oxygenates (Scheme 2).
Since APR reaction is not limited to the formation of H2 and
CO2, the H2/CO2 ratio might differ from the theoretical one.
In the case when a part of carbon is not converted into CO2,
the H2/CO2 ratio will be higher than the theoretical one based
on the reaction equation shown in Scheme 1.
Fig. 7 H2/CO2 ratio for different Pt/C catalysts in APR of xylitol.
Conditions: 0.5 g of catalyst, 498 K, 29.3 bar, N2 flow 30 ml min−1
,
10 wt.% xylitol solution, conversion ~ 10–12%.
The conversion of carbon to gaseous products in APR of xylitol
is shown in Fig. 8. As can be noted, the most active catalyst
is Pt/C (Degussa) with 13.6% of carbon converted into the
gaseous products at 10% of xylitol conversion. The Pt/BAC
catalyst was able to convert 10.5% of carbon, whereas the
corresponding values found for Pt/TiC-CDC, Pt/Sibunit (II)
and Pt/Sibunit (I) were 8.5%, 5.6% and 4.4%, respectively.
The following order on the basis of carbon converted into
the gas phase can be presented: Pt/C(Degussa) > Pt/BAC >
Pt/TiC-CDC > Pt/Sibunit(II) > Pt/Sibunit (I).
During APR of xylitol the formation of alkanes C1–C5 takes
place as a result of dehydration and further hydrogenation
reactions.20 Hydrocarbon formation versus WHSV in the APR
of xylitol over the Pt/TiC-CDC catalyst is presented in Fig. 9.
In general, all the catalysts have similar compositions of the
hydrocarbon mixtures formed with C1–C3 dominating in the
mixture. Butane was found in minor quantities as well as CO.
An important parameter in the APR process of polyols is
the ability of a catalyst to convert carbon into gas phase prod-
ucts, in other words, to cleave C–C bonds in the substrate.
Scheme 2 APR of xylitol: the main transformation pathways.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 393

View Article Online
Paper
Catalysis Science & Technology
Fig. 10 Rate of hydrocarbon formation for different Pt/C catalysts
in APR of xylitol. Conditions: 0.5 g of catalyst, 498 K, 29.3 bar, N2 flow
30 ml min−1, 10 wt.% xylitol solution, conversion ~ 10–12%.
Fig. 8 Conversion of carbon to gas phase products for different
Pt/C catalysts in APR of xylitol. Conditions: 0.5 g of catalyst, 498 K, 29.3 bar,
N2 flow 30 ml min−1, 10 wt.% xylitol solution, conversion ~ 10–12%.
xylitol conversion. The results indicating an increase of
hydrogen-to-carbon dioxide ratio at lower conversions in
APR over Pt/C are in line with the data previously reported for
Pt/Al2O3.25 Opposite to the H2/CO2 ratio, the selectivity to
alkanes is lower at higher space velocities for all the catalysts
studied. Since hydrogen formation is competing with hydro-
carbon formation in APR32 the trend of hydrocarbon selec-
tivity to increase with a decrease of hydrogen selectivity is
understandable.
Correlation between catalyst structure and catalytic
performance in APR of xylitol over Pt/C. All Pt/C catalysts
demonstrated a different behavior in the APR of xylitol in
terms of substrate conversion and selectivity towards the
main products such as H2 and CO2 as well as hydrocarbons.
The main reasons for determining catalytic behavior are the
textural and surface properties of the catalytic materials.
Based on CO chemisorption data (Table 1), corresponding
TOF values for Pt/C catalysts studied in the present work were
calculated. As a result a dependence of TOF for hydrogen pro-
duction on average Pt cluster size is displayed in Fig. 11. The
TOF increases linearly with an increase in the average size of
the Pt cluster in the APR of xylitol thus indicating that APR of
xylitol is a structure-sensitive reaction. Lehnert and Claus,
explaining the increase in the hydrogen formation rates
during APR of glycerol over Pt/Al2O3 catalysts with different
dispersions, proposed that adsorption and C–C cleavage of
polyol species preferably occurred on the face Pt atoms
rather than on edge and corner atoms. With an increase in
the cluster size the number of face atoms increases whereas
the number of corner and edge atoms should decline.77 The
volume of Pt clusters is proportional to r3, while the surface
area, to r2, where r is a radius of Pt particles. Therefore,
assuming that face Pt atoms are much active in APR than
edge and corner atoms, the catalytic activity, i.e. TOF should
increase linearly with an increase of the average size of the
Pt cluster. However, at a certain size of the Pt cluster the role
of corners and edges can be neglected and TOF reaches a
plateau. The results regarding the TOF dependence on the
average size of Pt clusters in xylitol APR are in very good
Fig. 9 Composition of hydrocarbons mixture in the APR of xylitol over
Pt/TiC-CDC. Conditions: 0.5 g of catalyst, 498 K, 29.3 bar, N2 flow
30 ml min−1, 10 wt.% xylitol solution.
The formation rates of all hydrocarbons as well as carbon
monoxide demonstrated a decreasing trend with an increase
in WHSV or in other words in a decrease of conversion.
Pentane concentration for all the experiments was below the
detection limit. It might be the case that pentane is not
formed since significantly higher surface acidity than those
possessed by the used carbon-supported catalysts is required.26,76
A comparative analysis of Pt/C catalysts in terms of hydro-
carbon formation is shown in Fig. 10. The hydrocarbon for-
mation rate in the case of Pt/BAC was more than 1.5 times
higher compared to both Pt/Sibunit samples and twofold
higher compared to Pt/C (Degussa). The lowest rate of hydro-
carbon formation was observed in the case of Pt/TiC-CDC
being almost four times lower in rate than that of Pt/BAC.
The catalytic data obtained for the Pt/C catalysts are
summarized in Table 5. As can be seen, for all the catalysts
with an increase in WHSV, the conversion of xylitol decreases.
Meanwhile the H2/CO2 ratio is higher at lower conversion of
the initial substrate, meaning less consumption of hydrogen
therefore the H2/CO2 ratio increases with a decrease in the
394 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 5 Catalytic data for APR of xylitol over Pt/C at different weight hour space velocitiesa
Pt/TiC-CDC
Pt/C Degussa
Pt/Sibunit (I)
Pt/Sibunit (II)
Pt/BAC
Pt/Al2O3
WHSV, h−1
Conversion, %
Cgas, %
H2/CO2
Salk, %
1.2
56
23.3
2.0
19
3.0
12
8.5
2.8
14
4.8
11
5.0
3.0
1
1.2
57
38.5
2.2
15
3.0
9
13.6
2.7
12
4.8
3
7.1
2.9
11
1.2
73
31.9
1.9
21
3.0
32
14.0
2.3
17
4.8
1
7.4
2.6
14
1.2
57
32.6
1.8
22
3.0
15
12.4
2.4
17
4.8
13
5.6
2.5
14
1.2
76
49.0
1.8
22
3.0
22
15.9
2.0
19
4.8
15
10.5
2.4
19
1.2
82
39.1
2.5
4.6
3.0
43
16.6
2.7
4.9
4.8
27
8.8
2.9
4.2
a
Conditions: 0.5 g of catalyst, 498 K, 29.3 bar, 10 wt.% xylitol solution, 30 ml min−1 N2 flow rate.
agreement with the results reported earlier for APR of glycerol
and ethylene glycol.
Another important catalyst characteristic bearing a signifi-
cant effect on the catalytic performance is the acidity of the
catalytic material. Generally, it is assumed that more acidic
supports facilitate the formation of hydrocarbons in APR
since the key step in alkane formation is dehydration of
polyol32 (Scheme 2). It has been shown recently that acidic
and basic properties of the support play a key role in hydrogen
formation in APR of glycerol.78 On top of that, the addition of
Re increases the acidity of Pt/C catalysts thus resulting in ele-
vated formation of alkanes from glycerol as demonstrated by
Zhang et al.43 Fig. 12 shows a clear correlation between the
acidic properties of the Pt/C catalysts elucidated by NH3-TPD
and catalytic performance in the APR of xylitol. As can be
seen from Fig. 12, with an increase in the total acidity of the
catalysts the selectivity to H2 decreases, while the rate of
hydrocarbon production increases, illustrating that forma-
tion of hydrogen and hydrocarbons are competing reactions
during APR. Thus, when more hydrogen is formed less
hydrocarbons can be produced. Zhang et al.43 demonstrated
that the ability of the catalyst to cleave C–O bonds (responsible
for hydrocarbons formation) is enhanced when more Re is
added to Pt (thus increasing the total acidity of the catalytic
material). Besides, Criftci et al.78 demonstrated that less
acidic supports are more favorable for selective production
of hydrogen. Additionally, higher selectivities to hydrogen
were observed in APR of glycerol when using such basic
supports for Pt as Mg(Al)Ox.79 Therefore, the correlation
Fig. 12 Dependence of alkanes formation rate and H2/CO2 ratio on a
surface acidity of Pt/C catalysts. Conditions: 498 K, 29.3 bar, N2 flow
30 ml min−1, conversion ~10–12%.
between the acidity of the catalytic materials investigated in
the present study is in a perfect correspondence with data
reported previously in the literature regarding APR of glycerol.
APR of xylitol over carbon- and metal oxide supported Pt
catalysts: effect of the support. In order to reveal the influence
of the support on the catalytic activity and selectivity towards
the main products a comparison of Pt/C and Pt/MOx catalysts
in the APR of xylitol under identical experimental conditions
was performed. The data are collected in Table 6. For com-
parison three carbon supported catalysts, Pt/TiC-CDC, Pt/C
(Degussa) and Pt/Sibunit (II) were selected. As for metal
oxide supported Pt catalysts, Pt/Al2O3 (Degussa) was chosen.
As can be seen from Table 6, Pt/C (Degussa) demonstrated
the highest conversion of carbon to gas phase products.
Carbon supported catalysts displayed in general higher
conversion of carbon into gas phase products at the same
xylitol conversion level than Pt supported on Al2O3. A similar
trend was observed by Kim et al.51 for APR of glycerol over
Pt–Re catalysts supported on carbonaceous materials CMK-3,
oxide supports as well as SiO2 at 523 K, 45 bar and 2.0 h−1
.
The order on the basis of carbon conversion into gas phase
products is the following: Pt/C (Degussa) > Pt/Sibunit (II) >
Pt TiC-CDC > Pt/Al2O3. Shabaker et al.4 reported higher
carbon turnover frequencies in APR of ethylene glycol: the
total carbon TOF (comprising CH4, CO2 and C2H6) in the case
of Pt/C was 4.88 min−1 whereas it was 3.07 min−1 for Pt/Al2O3.
The H2/CO2 ratio for the catalysts studied decreases in the
following order: Pt/Al2O3 > Pt/C (Degussa) > Pt/TiC-CDC >
Pt/Sibunit (II).
Fig. 11 Dependence of TOF versus mean Pt cluster size in APR of xylitol.
Conditions: 498 K, 29.3 bar, N2 flow 30 ml min−1, conversion ~10–12%.
(I) – Pt/BAC, (II) – Pt/C (Degussa), (3) – Pt/Sibunit (II), (4) – Pt/TiC-CDC.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 395

View Article Online
Paper
Catalysis Science & Technology
Table 6 Comparison between Pt/C catalysts and Pt/Al2O3 in the APR of xylitola
Catalyst
Pt content, wt.%
D,b
%
Carbon to gas, %
H2/CO2
Salk, %
TOFH2, min−1
TOFalk, min−1
Pt/TiC-CDC
Pt/C (Degussa)
Pt/Sibunit(II)
Pt/Al2O3
2.8
5
5
12
47
41
29
23.3
38.5
32.6
21.1
2.0
2.2
1.8
2.5
19
15
22
4.6
16.7
4.0
3.0
8.5
0.80
0.16
0.24
0.10
5
a
Conditions: 0.5 g of catalyst, 498 K, 29.3 bar, 10 wt.% xylitol solution, 30 ml min−1 nitrogen flow rate, conversion of xylitol ~55–57%.
Calculated based on CO chemisorption.
b
Likewise, Kim et al. observed the difference in the selectivity
APR of ethylene glycol. It has been demonstrated that the
conversion of ethylene glycol decreased in the following
order: Pt–Mn/CMK-3 > Pt–Mn/AC > Pt–Mn/Al2O3. The pro-
posed reason for superior behavior of Pt–Mn/CMK-3 in terms
of hydrogen production was connected to the ordered pores
and opened mesopores of the CMK-3 support structure. The
latter one provides easier access of the reactants and facili-
tates desorption of the products, i.e. (H2, CO2, alkanes) from
the catalytic material without diffusion resistance.80
to hydrogen for Pt–Re supported catalysts in APR of glycerol.51
Pt–Re supported on alumina was the most selective (77%)
catalyst; the SiO2-supported catalyst demonstrated a compa-
rable selectivity to the CMK-3-supported material being 61.7
and 61.6%, respectively. The lowest selectivity to H2 was
observed in the case of Pt–Re/AC which exhibited an SH2 of
only 57.1%.
Since production of hydrogen competes with formation of
hydrocarbons in APR, the catalysts which demonstrated the
lowest H2/CO2 values possessed the highest selectivity to alkanes.
Thus, the selectivity to alkanes decreases in the following order:
Pt/Sibunit (II) > Pt/TiC-CDC > Pt/C (Degussa) > Pt/Al2O3.
The results are in good correlation with the previous work on
APR of glycerol:51 selectivity to alkanes decreased in the follow-
ing order: Pt–Re/AC > Pt–Re/CMK-3 > Pt–ReSiO2 > Pt–Re/Al2O3.
Based on the data obtained from CO chemisorption, the
corresponding values of turnover frequencies of hydrogen and
alkane formation can be calculated (Table 6). The highest
value of TOFH2 was observed in the case of Pt/TiC-CDC catalyst
being equal to 16.7 min−1. Compared to Pt/TiC-CDC, Pt/Al2O3
Shabaker et al.4 reported efficient catalytic performance
of carbon supported catalysts over Pt/Al2O3 in the APR of
ethylene glycol in terms of hydrogen and hydrocarbons pro-
duction. The TOFH2 values for Pt/C were slightly higher than
those observed for Pt/Al2O3 in APR of EG. It is important
to note that the TOFalk values were also higher in the case of
Pt/C than for Pt/Al2O3.
The superior catalytic behavior of Pt/C catalysts, especially
Pt/TiC-CDC, compared to the alumina supported sample, can
be linked to the higher surface area of the carbon supported
materials and enhanced hydrothermal stability under severe
APR conditions. Similar to the case of CMK-3 based catalysts,50,51
the narrow pore size distribution inside carbide-derived
carbons as well as regular structure and high surface area of
carbons might facilitate APR thus resulting in enhanced
TOFH2 and TOFalk. Furthermore, the interactions between Pt
and carbide-derived support might enhance the catalytic
performance.70
demonstrated an almost twofold lower value of TOFH2
,
8.5 min−1. The TOF values of H2 generation for Pt/C (Degussa)
and Pt/Sibunit (II) were 4.0 and 3.0 min−1, respectively. It is
worth mentioning that Pt supported on carbon turned out to
be more catalytically active also in hydrocarbon production
via APR than Pt/Al2O3 as evidenced by the corresponding
TOFalk values presented in Table 6. Thus the TOFalk values in
APR of xylitol decrease in the following order: Pt/TiC-CDC >
Pt/Sibunit (II) > Pt/C (Degussa) > Pt/Al2O3.
Conclusions
The supported Pt–Re catalysts studied by Kim et al.51 were
placed in the following order based on the volumetric hydrogen
The APR of xylitol was investigated over five different carbon
supported Pt catalysts: Pt/TiC-CDC, Pt/C (Degussa), Pt/Sibunit
(I and II) and Pt/BAC.
production rates: Pt–Re/CMK-3 > Pt–Re/AC > Pt–Re/SiO2
>
Pt–Re/Al2O3. The authors noticed that carbon supports
(both CMK-3 and AC) showed a better performance in terms
of hydrogen production than alumina and silica supported
catalysts. After calculating TOFH2 based on the values of
hydrogen production rates and metal dispersion, the following
order can be presented: Pt–Re/AC > Pt–Re/SiO2 ~ Pt–Re/CMK-3 >
Pt–Re/Al2O3.
This order clearly shows the effect of support on hydrogen
production via APR of glycerol (TOFH2 values were calculated
using literature data and the equation described in the Experi-
mental section).
The catalysts demonstrated minor deactivation with time-
on-stream. The decrease in xylitol conversion after 120 h of
catalytic performance for Pt/TiC-CDC did not exceed 10%.
The catalysts possess different textural and acid–base
properties. Therefore, it was possible to study the effect of
surface acidity on the formation of hydrogen and alkanes.
For the first time the structure sensitivity in APR of polyols
containing more than three carbon atoms was found and
studied. The TOF linearly depends on the average size of the
Pt cluster and increases with an increase in the cluster size
within the studied cluster size domain. The results obtained
are in very good correlation with previously reported data on
Increased hydrogen production affected by the type of
support was also shown50 for bimetallic Pt–Mn catalysts in
396 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
APR of ethylene glycol and glycerol. A plausible explanation
for the TOF increase with an increase in Pt cluster size was
proposed. The highest TOF values in terms of hydrogen pro-
duction were observed for Pt supported on carbide-derived
carbon. This material has shown a high potential in the APR
of xylitol since the Pt/TiC-CDC catalyst exhibited both selec-
tive (H2/CO2 ~ 3) and effective production of hydrogen.
The TOFH2 values observed in the case of Pt/TiC-CDC were
twofold higher compared to those of Pt on alumina and more
than four times higher compared to those of platinum on
activated carbon and Sibunit. Moreover, the TOF for alkane
formation in the case of Pt/TiC-CDC was 0.80 min−1 which is
eight times more compared to that of Pt/Al2O3.
The results obtained from NH3-TPD studies revealed a
correlation between Pt/C acidity and catalytic performance. It
was found that catalytic materials bearing high acidity favor
the formation of hydrocarbons. On the contrary, Pt/C catalysts
possessing low (Pt/C (Degussa) or very low (Pt/TiC-CDC) acidity
were effective in hydrogen formation. Another important
advantage of Sibunit and carbide-derived carbon supports is
high mechanical stability. This in turn facilitates their use as
catalytic supports in continuous fixed-bed reactors.
was added until the total volume of the slurry was 50 ml. The
slurry formed was stirred for 2 h at 323 K. After impregnation
the sample was dried at 398 K in air for 12 h. The catalyst
was further calcined in air using the following temperature
program: 298 K → 623 K within 24 h, dwell at 623 K for 2 h.
The calcined sample was flushed with dry Ar and reduced in
hydrogen using the following temperature ramp: 298 K → 573 K
within 2 h and at 573 K for 1.5 h.
Preparation of 5 wt.% Pt/Sibunit from Pt(NH3)4(HCO3)2.
A solution containing 0.85 g of the Pt precursor in 50 ml of
distilled water was added to 8 g of Sibunit (fraction 0.16–0.4 mm).
Distilled water was added until the total volume of the slurry
reached 200 ml. The suspension obtained was stirred for
2 h at 323 K. The sample was dried, calcined and reduced
following the same procedure as described above for
Pt/Sibunit (from H2PtCl6).
5 wt.% Pt/BAC
The solution of Pt(NH3)4(HCO3)2 containing 0.85 g of the
complex in 50 ml of distilled water was added to the activated
carbon (from birch, 8 g). The platinum salt was added in
dissolved form. While monitoring the pH (acidic), the solu-
tion was neutralized by sodium carbonate for formation of
hydroxides precipitated on the support, which were then
readily reduced. The catalyst was reduced chemically using
formic acid. Sodium acetate, which prevents peptisation
(coagulation) of the metal according to,81 and the reducing
agent formic acid were added dropwise under stirring at
363 K. After stirring the mixture for 1 h, the solution was
cooled down and a sample was withdrawn for a qualitative
analysis to confirm that there were no chloride ions present
in the solution after the impregnation.
To conclude, carbon supports have great potential in APR,
since they are available, stable under hydrothermal condi-
tions and possess a variety of textural and surface properties,
which can be tuned to obtain the desirable catalytic material.
It has been shown that by the choice of carbon material, its
nature, pretreatment of the carbon support, as well as by the
choice of Pt source it is possible to steer the APR of polyols
towards sustainable hydrogen or hydrocarbon production.
Experimental part
Catalyst preparation
2.5 wt.% Pt/TiC-CDC (carbide-derived carbon)
All chemicals for catalyst synthesis (H2PtCl6, Pt(NH3)4(HCO3)2)
(>99.9%) were purchased from commercial suppliers and
used without further purification. Titanium carbide (TiC) was
purchased from Goodfellow with a purity >99.8% and a mean
particle size of 75 μm. The platinum precursor [Pt(NH3)4]Cl2
was purchased from Alfa Aesar with a purity of 99.9%; 65 wt.%
HNO3 was purchased from AppliChem as pure acid.
Preparation of the TiC-CDC support. Titanium carbide
(TiC, fraction 50–150 μm) was used as a precursor for the
preparation of the TiC-CDC support. Hereby the carbide was
subjected to chlorination in an alumina tubular reactor at
1473 K (CCl2 = 1.5 mol m−3, reactor diameter 3.2 cm, flow rate
of 3 cm s−1) using a mixture of Cl2 and Ar. After a chlorina-
tion time of typically 5 h, 5 g of carbide was converted into
1 g of TiC-CDC. The material obtained was then treated with
hydrogen for at least 30 min at the chlorination temperature
to remove residual chlorine and metal chlorides from
the pores of the carbon support and terminate its surface
with hydrogen.
The obtained carbon support was treated with 65 wt.%
HNO3 at 363 K for 2 h using 50 ml acid per 1 g TiC-CDC. The
carbon support obtained was filtered, washed with large
amount of distilled water until pH 7 and dried at 353 K
in air.64,82
5 wt.% Pt/Sibunit
Preparation of the Sibunit support. Sibunit was placed in
a glass reactor for calcination and calcined in a flow of dry
air (50 ml min−1) using the following procedure: from 298 K
to 373 K within 60 min, 373 → 433 K within 4 h, then 433 →
493 K within 4 h, 493 K → 553 K within 4 h, 553 K → 623 K
within 4 h, then cooling 623 K → ambient temperature
within 2 h. Calcined Sibunit was then crushed in a porcelain
mortar and sieved. A fraction of 0.16–0.4 mm was collected.
Preparation of 5 wt.% Pt/Sibunit from H2PtCl6. A solution
containing 0.546 g of H2PtCl6 in 14 ml of distilled water
(total Pt content 0.26 g) was added to 5 g of Sibunit (fraction
0.16–0.4 mm). Then an additional amount of distilled water
Preparation of 2.5 wt.% Pt/TiC-CDC. Impregnation was
performed with an ion adsorption method where the volume
of impregnation solution containing the catalyst precursor
material exceeds the pore volume of carbon material. The
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 397

View Article Online
Paper
Catalysis Science & Technology
charged precursor is bound and stabilized at the surface via
Coulomb forces83–85 at a distinct pH below or above the
point of zero charge (PZC) of the support. Hereby the
ratio of impregnation solution to carbon was adjusted to
50 mlsolution gTiC-CDC−1. The desired loading of platinum on
carbon was 2.5 wt.%. As the precursor material [Pt(NH3)4]Cl2
was dissolved in water leading to a cationic precursor
complex, the pH of the solution was adjusted to pH 3.5 using
HNO3. After an addition of acid functionalized TiC-CDC the
pH of the suspension was controlled and readjusted with
NaOH or HNO3 if necessary. The suspensions were stirred at
room temperature for 24 h and subsequently filtered. The
impregnated TiC-CDC filter cake was then washed with
200 ml of an HNO3 dilution of equal pH as the impregnation
solution and subsequently dried in an oven at 353 K overnight.
In the last step, the catalyst precursor on the impregnated
TiC-CDC was reduced at 673 K with 12 LN h−1 of 20 vol.% H2
in N2 in a tubular reactor.64,82 The real Pt content determined
by ICP-OES was 2.8% thus the subsequent designation (2.8%)
was used for Pt/TiC-CDC.
specific surface area the BET equation (Pt/TiC-CDC, Pt/Sibunit I
and II) or the Dubinin equation for microporous carbons was
applied (Pt/BAC, Pt/C (Degussa)).
The metal loading of all the catalysts (with the nominal wt%
loadings given in the catalyst descriptor) was determined by
ICP-OES using Perkin-Elmer Optima 4300.
The temperature-programmed reduction (TPR) (Micromeritics,
Autochem 2910) was measured by placing approximately 0.1 g
of Pt catalyst in a U-shaped tube, which was cooled to 298 K
in argon. The catalyst was reduced using 5% H2 in Ar with
the temperature being ramped from 298 K to 723 K at a rate
of 5 K min−1 and the hydrogen uptake monitored by a thermal
conductivity detector (TCD).
The NH3-TPD measurements were performed using
Micromeritics Autochem 2910 apparatus. The procedure for
determination of catalyst acidity by NH3-TPD is similar to the
one described by Souza and co-workers for Ni–Cu catalysts.86
The same method was recently applied to acidity determina-
tion of Pt– and PtRe/TiO2 catalysts.87 Prior to NH3 treatment
the catalyst sample (0.1 g) was dried in an oven at 373 K over-
night. The sample was then placed in a U-shape quartz tube
and reduced under a hydrogen flow (20 ml min−1) using the
following procedure: 5 K min−1 to 523 K, dwell for 2 h. The
catalyst was then flushed in a flow of He (20 ml min−1) for
30 min to remove hydrogen from the catalyst surface. The
sample was cooled down to ambient temperature and satu-
rated with NH3 (gas mixture 5% of NH3 in He) for 1 h. Then,
the gas mixture was switched back to He and the catalyst was
flushed for 30 min to remove physically adsorbed ammonia.
Temperature-programmed desorption was realized by heating
up to 498 K at 3 K min−1. Quantification of ammonia
desorbed was done by monitoring changes of the calibrated
TCD signal (calibrated with 5 vol.% of NH3 in He). The
number of acid sites was calculated through the measured
amount of ammonia desorbed from a sample. For determina-
tion of the heat of desorption, after the first desorption of
ammonia, a new saturation of a sample with ammonia was
performed. The sample was treated by mixture of ammonia
in helium as described above. The desorption was realized by
heating to 498 K at various heating rates (5, 10, 15, 20 K min−1).
Heat of desorption was calculated using a standard approach
which includes plotting Tp (temperature at maximum of desorption)
5 wt.% Pt/C (Degussa)
The commercial catalyst 5 wt.% Pt/C (F 1525 XKT/W) was
delivered by Degussa and used as received. The catalyst con-
tains 5 wt.% of Pt according to the specification.
5 wt.% Pt/Al2O3
The Pt/Al2O3 catalyst is a commercial catalyst provided by
Degussa (F 214 XSP). The catalyst contains 5 wt.% of Pt
according to the specification.
Catalyst characterization
Prior to characterization as well as catalytic measurement all
the catalysts were pre-dried overnight at 373 K in an oven in
air to remove moisture.
The metal dispersion was determined by CO pulse chemi-
sorption in an apparatus manufactured by Micromeritics
(Autochem 2900). The catalyst was reduced prior to the mea-
surement with the following program: 298–353 K at 10 K min−1
in He, dwell for 30 min, gas-switch to H2, 5 K min−1 to 523 K,
dwell for 2 h, followed by flushing for 60 min in He at 523 K to
remove surface hydrogen. Thereafter the catalyst was cooled to
ambient temperature and CO pulses were introduced utilizing
10 vol.% CO in He. For data evaluation a Pt/CO stoichiometry
2of 1 : 1 is assumed.
Transmission electron microscopy (TEM) measurements
were performed with LEO 912 Omega, voltage 120 kV. The
samples for TEM were prepared as a suspension of ethanol
and for calculating the diameter of particles ca. 500 particles
for each sample were taken.
2
versus ln(Tp /β) (where β corresponds to a heating rate) followed
by calculation of the slope, and then Edes [kJ mol−1].
Typical reaction procedure
For the APR studies reported in the present work, a continuous
fixed-bed reactor setup (stainless steel reactor, d = 4.8 mm,
l = 18 cm) equipped with a furnace was used. The reactor
scheme is shown in Fig. 13. In the standard experiment
the catalyst (0.5 g) was mixed with ca. 3 g of quartz sand
and loaded to the reactor. The catalyst was reduced prior
to the measurements with H2 using the following program:
298 → 523 K at 5 K min−1 in hydrogen for 2 hours, at a
The surface area and pore volume data for the Pt/C
catalysts were obtained by low-temperature N2 adsorption
using Micromeritics ASAP 2010. For determination of the
hydrogen flow rate of 30 ml min−1
.
398 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper

1 

r(H2 ) mol min


(3)
(4)
H2 / CO2 
1 
r(CO2 ) mol min


Selectivity to alkanes (%)
v(Calkanes
)
Salk

100%
v(Cin gas
)
Acknowledgements
Fig. 13 Continuous fixed-bed reactor applied for APR of xylitol.
The Academy of Finland in collaboration with the North
European Innovative Energy Research Programme (N-Inner)
and the Graduate School in Chemical Engineering (GSCE)
and Fortum Foundation are gratefully acknowledged for
financial support. In addition the COST-Action CM0903
(UbioChem) as well as the Bio4Energy programme in Sweden
are acknowledged. The authors gratefully acknowledge the
funding of the European Union Seventh Framework
Programme (FP7/2007-2013) within the project SusFuelCat
under grant agreement no. 310490 (http://www.susfuelcat.eu).
B.H. and B.E. acknowledge the funding of the German
Research Council (DFG), which supports the Cluster of Excel-
lence “Engineering of Advanced Materials” and the German
academic exchange service (DAAD). The work is a part of
the activities of Process Chemistry Centre (PCC) financed by
Åbo Akademi University.
The reaction was carried out at 498 K and 29.3 bar, at a
range of space velocities of 0.9–6.0 h−1. The weight hour space
velocity (WHSV) is defined as mass of substrate fed per mass
−1
of the catalyst per hour [gsub gcat h−1]. An aqueous solution
(10 wt.%) of xylitol was used as the feedstock and fed in a
continuous manner via an HPLC pump. Liquid samples were
withdrawn periodically and analyzed by means of high-
performance liquid chromatography (HPLC), applying an
injection volume of 2 μl, Aminex HPX-87H column, eluent
5 mM H2SO4, flow rate 0.6 ml min−1, 318 K, 70 min, applying
a refractive index (RI) detector to determine conversion
of xylitol.
The gaseous products were taken periodically and analyzed
by means of a micro-GC (Agilent Micro GC 3000A). The instru-
ment was equipped with 4 columns: Plot U, OV-1, Alumina
and Molsieve. The micro-GC was calibrated to perform quan-
titative analysis for the following gases: H2, CO2, CO, CH4,
linear hydrocarbons C1–C4 and 1 wt.% of He in N2 was
used as an internal standard. Moreover, the carbon balance
was monitored by means of total organic carbon analysis
(TOC instrument) and was confirmed to a degree of 95–100%
for all the measurements.
References
1 Thermochemical Conversion of Biomass to Liquid Fuels and
Chemicals, ed. M. Crocker, RSC Publishing, 2010.
2 C.-H. Zhou, X. Xia, C.-X. Lin, D. S. Tonga and J. Beltramini,
Chem. Soc. Rev., 2011, 40, 5588–5617.
3 J. C. Serrano-Ruiz, R. Luque and A. Sepulveda-Escribano,
Chem. Soc. Rev., 2011, 40, 5266–5281.
4 J. W. Shabaker, G. W. Huber, R. R. Davda, R. D. Cortright
and J. A. Dumesic, Catal. Lett., 2003, 88, 1–8.
5 R. R. Davda, J. W. Shabaker, G. W. Huber, R. D. Cortright
and J. A. Dumesic, Appl. Catal., B, 2003, 43, 13–26.
6 G. W. Huber, J. W. Shabaker and J. A. Dumesic, Science,
2003, 300, 2075–2077.
Calculation of the turnover frequency (TOF), selectivity to H2
and alkanes
Turnover frequencies for Pt/C catalysts were calculated using
the following equation:
7 J. W. Shabaker, G. W. Huber and J. A. Dumesic, J. Catal.,
2004, 222, 180–191.
8 G. W. Huber, J. W. Shabaker, S. T. Evans and J. A. Dumesic,
Appl. Catal., B, 2006, 62, 226–235.

1 


r(H2 ) mol min

1 

(I)

TOFH2 min


 (M) mol  D a.u.



9 X. Chu, J. Liu, B. Sun, R. Dai, Y. Pei, M. Qiao and K. Fan,
J. Mol. Catal. A: Chem., 2011, 335, 129–135.
10 J. Liu, B. Sun, J. Hu, Y. Pei, H. Li and M. Qiao, J. Catal.,
2010, 274, 287–295.
11 M. F. D'Angelo, V. Ordomsky, V. Paunovic, J. van der Schaaf,
J. C. Schouten and T. A. Nijhuis, ChemSusChem, 2013,
6, 1708–1716.
where r(H2) denotes the rate of H2 formation, v(M) – moles of
Pt and D – dispersion

1 

r(alk) mol min

(II)

1 

TOF min

alk

v(M) mol  D a.u.




where r(alk) denotes the rate of total alkane formation in the
gas phase, v(M) – moles of Pt and D – dispersion
12 R. R. Soares, D. A. Simonetti and J. A. Dumesic, Angew.
Chem., Int. Ed., 2006, 45, 3982–3985.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 399

View Article Online
Paper
Catalysis Science & Technology
13 K. Lehnert and P. Claus, Catal. Commun., 2008, 9, 2543–2546.
14 G. Wen, Y. Xu, H. Ma, Z. Xu and Z. Tian, Int. J. Hydrogen
Energy, 2008, 33, 6657–6666.
40 D. L. King, L. Zhang, G. Xia, A. M. Karim, D. J. Heldebrant,
X. Wang, T. Peterson and Y. Wang, Appl. Catal., B, 2010,
99, 206–213.
15 E. L. Kunkes, R. R. Soares, D. A. Simonetti and J. A. Dumesic,
Appl. Catal., B, 2009, 90, 693–698.
16 P. D. Vaidya and A. E. Rodrigues, Chem. Eng. Technol., 2009,
32, 1463–1469.
17 A. Wawrzetz, B. Peng, A. Hrabar, A. Jentys, A. A. Lemonidou
and J. A. Lercher, J. Catal., 2010, 269, 411–420.
18 A. O. Menezes, M. T. Rodrigues, A. Zimmaro, L. E. P. Borges
and M. A. Fraga, Renewable Energy, 2011, 36, 595–599.
19 D. Ö. Özgür and B. Z. Uysal, Biomass Bioenergy, 2011,
35, 822–826.
20 A. V. Kirilin, A. V. Tokarev, L. M. Kustov, T. Salmi,
J.-P. Mikkola and D. Yu. Murzin, Appl. Catal., A, 2012,
435–436, 172–181.
41 B. Meryemogl, A. Hesenov, S. Irmak, O. M. Atanur and
O. Erbatur, Int. J. Hydrogen Energy, 2010, 35, 12580–12587.
42 L. Zhang, G. Xia, Y. Yang, D. Heldebrant, D. King, Y. Wang
and L. F. Allard, Microsc. Microanal., 2010, 16 (Suppl. 2),
1428–1429.
43 L. Zhang, A. M. Karim, M. H. Engelhard, Z. Wei, D. L. King
and Y. Wang, J. Catal., 2012, 287, 37–43.
44 A. C.-C. Chang, R. F. Louh, D. Wong, J. Tseng and Y. S. Lee,
Int. J. Hydrogen Energy, 2011, 36, 8794–8799.
45 X. Wang, N. Li, L. D. Pfefferle and G. L. Haller, Catal. Today,
2009, 146, 160–165.
46 X. M. Wang, N. Li, J. A. Webb, L. D. Pfefferle and
G. L Haller, Appl. Catal., B, 2010, 101, 21–30.
21 T. Jiang, T. Wang, L. Ma, Y. Li, Q. Zhang and X. Zhang, Appl.
Energy, 2012, 90, 51–57.
47 X. M. Wang, N. Li, L. D. Pfefferle and G. L. Haller, J. Phys.
Chem. C, 2010, 114, 16996–17002.
22 G. W. Huber, R. D. Cortright and J. A. Dumesic, Angew.
Chem., Int. Ed., 2004, 43, 1549–1551.
23 Y. Xu, Z. Tian, G. Wen, Z. Xu, W. Qu and L. Lin, Chem. Lett.,
2006, 35, 216–217.
24 A. Tanksale, C. H. Zhou, J. N. Beltramini and G. Q. Lu,
J. Inclusion Phenom. Macrocyclic Chem., 2009, 65, 83–88.
25 A. V. Kirilin, A. V. Tokarev, E. V. Murzina, L. M. Kustov,
J.-P. Mikkola and D. Yu. Murzin, ChemSusChem, 2010,
3, 708–718.
48 X. M. Wang, N. Li, C. Wang, W. R. Swchartz, S. Lim,
T. Fadel, L. D. Pfefferle and G. L. Haller, ACS Catal., 2012,
2, 1480–1486.
49 T.-W. Kim, H.-D. Kim, K.-E. Jeong, H.-J. Chae, S.-Y. Jeong,
C.-H. Lee and C.-U. Kim, Green Chem., 2011, 13, 1718–1728.
50 H.-D. Kim, H. J. Park, T.-W. Kim, K.-E. Jeong, H.-J. Chae,
S.-Y. Jeong, C.-H. Lee and C.-U. Kim, Int. J. Hydrogen Energy,
2012, 37, 8310–8317.
51 H.-D. Kim, H. J. Park, T.-W. Kim, K.-E. Jeong, H.-J. Chae,
S.-Y. Jeong, C.-H. Lee and C.-U. Kim, Catal. Today, 2012,
185, 73–80.
26 Q. Zhang, T. Wang, B. Li, T. Jiang, L. Maa, X. Zhang and
Q. Liu, Appl. Energy, 2012, 97, 509–513.
27 R. D. Cortright, R. R. Davda and J. A. Dumesic, Nature, 2002,
418, 964–967.
52 R. M. Ravenelle, J. R. Copeland, W.-G. Kim, J. C. Crittenden
and C. Sieverse, ACS Catal., 2011, 1, 552–561.
28 Z. Tang, J. Monroe, J. Dong, T. Nenoff and D. Weinkauf,
Ind. Eng. Chem. Res., 2009, 48, 2728–2733.
53 F. Rodríguez-Reinoso and A. Sepúlveda-Escribano, Chapter 9 –
Porous Carbons in Adsorption and Catalysis, Handbook of
Surfaces and Interfaces of Materials, 2001, 309355.
54 V. F. Surobikin, G. V. Plaxin, V. A. Semikolenov, V. A. Likholobov
and I. J Tiunova, US patent, US4978649, 1990.
55 Y. I. Yermakov, V. F. Surobikin, G. V. Plaxin, V. A. Semikolenov,
V. A. Likholobov, A. L. Chuvilin and S. V. Bogdanov, React.
Kinet. Catal. Lett., 1987, 33, 435–440.
56 I. L. Simakova, O. Simakova, A. V. Romanenko and
D. Yu. Murzin, Ind. Eng. Chem. Res., 2008, 47, 7219–7225.
57 O. A. Simakova, P. A. Simonov, A. V. Romanenko and
I. L. Simakova, React. Kinet. Catal. Lett., 2008, 95, 3–12.
58 I. V. Deliy, I. L. Simakova, N. Ravasio and R. Psaro, Appl. Catal., A,
2009, 357, 170–177.
29 B. Roy, K. Loganathan, H. N. Pham, A. K. Datye and
C. A. Leclerc, Int. J. Hydrogen Energy, 2010, 35, 11700–11708.
30 A. V. Tokarev, A. V. Kirilin, E. V. Murzina, K. Eränen,
L. M. Kustov, D. Yu. Murzin and J.-P. Mikkola, Int. J.
Hydrogen Energy, 2010, 35, 12642–12649.
31 N. Li and H. Huber, J. Catal., 2010, 270, 48–59.
32 R. R. Davda, J. W. Shabaker, G. W. Huber, R. D. Cortright
and J. A. Dumesic, Appl. Catal., B, 2005, 56, 171–186.
33 R. R. Davda and J. A. Dumesic, Angew. Chem., Int. Ed., 2003,
42, 4068–4071.
34 G. W. Huber, J. Chheda, C. B. Barrett and J. A. Dumesic,
Science, 2005, 308, 1446–2079.
35 N. Luo, X. Fu, F. Cao, T. Xiao and P. Edwards, Fuel, 2008,
87, 3483–3489.
36 Y. Guo, M. U. Azmat, X. Liu, Y. Wang and G. Lu, Appl. Energy,
2012, 92, 218–223.
37 K. Murata, T. Isao and M. Inaba, React. Kinet. Catal. Lett.,
2008, 93, 59–66.
38 D. A. Simonenti, E. L. Kunkes and J. A. Dumesic, J. Catal.,
2007, 247, 298–306.
59 L. B. Okhlopkova, A. S. Lisitsyn, V. A. Likholobov, M. Gurrath
and H. P. Boehm, Appl. Catal., A, 2000, 204, 229–240.
60 O. P. Taran, E. M. Polyanskaya, O. L. Ogorodnikova,
C. Descorme, M. Besson and V. N. Parmon, Catalysis in Industry,
2010, 2, 381–386.
61 V. Presser, M. Heon and Y. Gogotsi, Adv. Funct. Mater., 2011,
21, 810–833.
62 M. Schmirler, T. Knorr, T. Fey, A. Lynen, P. Greil and
B. J. M. Etzold, Carbon, 2011, 49, 4359–4367.
39 E. L. Kunkes, D. A. Simonett, J. A. Dumesic, W. D. Pyrz,
L. E. Murillo, J. G. Chen and D. J. Buttrey, J. Catal., 2008,
260, 164–177.
63 M. Schmirler, F. Glenk and B. J. M. Etzold, Carbon, 2011,
49, 3679–3686.
400 | Catal. Sci. Technol., 2014, 4, 387–401
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
64 A. Schlange, A. R. dos Santos, B. Hasse, B. J. M. Etzold,
U. Kunz and T. Turek, J. Power Sources, 2012, 199, 22–28.
65 L. Borchardt, F. Hasché, M. R. Lohe, M. Oschatz, F. Schmidt,
E. Kockrick, C. Ziegler, T. Lescouet, A. Bachmatiuk,
B. Büchner, D. Farrusseng, P. Strasser and S. Kaskel, Carbon,
2012, 50, 1861–1870.
66 A. Silvestre-Albero, S. Rico-Francés, F. Rodríguez-Reinoso,
A. M. Kern, M. Klumpp, B. J. M. Etzold and S.-J. Alberro,
Carbon, 2013, 59, 221–228.
67 T. Knorr, P. Heinl, J. Schwerdtfeger, C. Körner, R. F. Singer
and B. J. M. Etzold, Chem. Eng. J., 2012, 181–182, 725–733.
68 F. Glenk, T. Knorr, M. Schirmer, S. Gütlein and B. Etzold,
Chem. Eng. Technol., 2010, 33, 698–703.
69 J. C. Palmer, A. Llobet, S.-H. Yeon, J. E. Fischer, Y. Shi,
Y. Gogotsi and K. E. Gubbins, Carbon, 2010, 48, 1116–1123.
70 R. Rajagopalan, A. Ponnaiyan, P. J. Mankidy, A. W. Brooks,
B. Yi and H. C. Foley, Chem. Commun., 2004, 2498–2499.
71 A. Corma, Chem. Rev., 1995, 95, 559–614.
76 T. Jiang, Q. Zhang, T.-J. Wang, Q. Zhang and L.-L. Ma,
Energy Convers. Manage., 2012, 59, 58–65.
77 R. Van Hardeveld and F. Hartog, Surf. Sci., 1969, 15, 189–230.
78 A. Ciftci, B. Peng, A. Jentys, J. A. Lercher and E. J. M. Hensen,
Appl. Catal., A, 2012, 431–432, 113–119.
79 D. A. Boga, R. Oord, A. M. Beale, Y.-M. Chung,
P. C. A. Bruijnincx and B. M. Weckhuysen, ChemCatChem,
2013, 5, 529–537.
80 Y.-G. Wang, L. Cheng, F. Li, H.-M. Xiong and Y.-Y. Xia,
Chem. Mater., 2007, 19, 2095–2101.
81 O. Hofstadt and K. Kartte, Patent EP0193889, 1986.
82 B. Hasse, F. Reißner, C. Dicenta, P. Hausmann and
B. J. M. Etzold, Carbide-derived carbon (TIC-CDC) as model
catalyst support – an exemplary study, 15-th International
Congress on Catalysis 2012, Munich, Germany, 2012.
83 P. Brunelle, Stud. Surf. Sci. Catal., 1979, 3, 211–232.
84 X. Hao, S. Barnes and J. R. Regalbuto, J. Catal., 2011,
279, 48–65.
72 P. Serp, and J. L. Figueiredo, Carbon Materials for Catalysis,
John Wiley & Sons 2009.
85 X. Hao, L. Quach, J. Korah, W. A. Spieker and J. R. Regalbuto,
J. Mol. Catal. A: Chem., 2004, 219, 97–107.
73 T. J. Bandosz, J. Jagiello and J. A. Schwarz, Anal. Chem.,
1992, 64, 891.
74 J. W. Shabaker, D. A. Simonetti, R. D. Cortright and
J. A. Dumesic, J. Catal., 2005, 231, 67–76.
75 Y. T. Kim, J. A. Dumesic and G. W. Huber, J. Catal., 2013,
304, 72–85.
86 R. L. Manfro, T. P. M. D. Pires, N. F. P. Ribeiro and
M. M. V. M. Souza, Catal. Sci. Technol., 2013, 3, 1278–1287.
87 A. V. Kirilin, A. V. Tokarev, H. Manyar, C. Hardacre, T. Salmi,
J.-P. Mikkola and D. Yu. Murzin, Aqueous phase reforming of
xylitol over Pt–Re bimetallic catalyst: effect of the Re addition
Catal. Today, 2013http://dx.doi.org/10.1016/j.cattod.2013.09.020.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 387–401 | 401