The role of steric effects and acidity in the direct synthesis of iso-paraffins from syngas on cobalt zeolite catalysts

Article abstract of DOI:10.1002/cctc.201500777

This study focuses on the effects of the localization of Co species, zeolite structure, and acidity on the performance of Co bifunctional catalysts promoted with Pt for the direct synthesis of iso-paraffins from syngas. ZSM-5, MOR, and BEA were chosen as zeolites with different structures and pore diameters. The catalysts were prepared either by incipient wetness impregnation or by the mechanical mixing of the zeolite with a conventional silica-supported Co catalyst. The increase in the pore size and open character of the zeolite structure from ZSM-5 to BEA resulted in a higher fraction of Co located inside the pores of the catalysts prepared by impregnation. The catalytic performance was affected strongly by the zeolite acidity, pore structure, and Co distribution between the pores and the external surface. The selectivity to short-chain iso-paraffins is affected principally by the zeolite acidity, whereas the selectivity to long-chain branched hydrocarbons mostly depends on steric effects. Zeolite wins the fight: An increase in the size and open character of the zeolite pore structure from ZSM-5 to BEA results in a higher fraction of Co located inside the pores of the catalysts prepared by impregnation. The catalytic performance in Fischer-Tropsch synthesis is affected strongly by the zeolite acidity, pore structure, and Co distribution between the zeolite pores and the external surface.

Full text of DOI:10.1002/cctc.201500777

DOI: 10.1002/cctc.201500777
Full Papers
The Role of Steric Effects and Acidity in the Direct
Synthesis of iso-Paraffins from Syngas on Cobalt Zeolite
Catalysts
[
a]
[b]
[c]
[a]
Vijayanand Subramanian, Vladimir L. Zholobenko, Kang Cheng, Christine Lancelot,
[a]
[a]
[a]
[a]
Svetlana Heyte, Joelle Thuriot, SØbastien Paul, Vitaly V. Ordomsky,* and
[a]
Andrei Y. Khodakov*
This study focuses on the effects of the localization of Co spe-
cies, zeolite structure, and acidity on the performance of Co bi-
functional catalysts promoted with Pt for the direct synthesis
of iso-paraffins from syngas. ZSM-5, MOR, and BEA were
chosen as zeolites with different structures and pore diameters.
The catalysts were prepared either by incipient wetness im-
pregnation or by the mechanical mixing of the zeolite with
a conventional silica-supported Co catalyst. The increase in the
pore size and open character of the zeolite structure from
ZSM-5 to BEA resulted in a higher fraction of Co located inside
the pores of the catalysts prepared by impregnation. The cata-
lytic performance was affected strongly by the zeolite acidity,
pore structure, and Co distribution between the pores and the
external surface. The selectivity to short-chain iso-paraffins is
affected principally by the zeolite acidity, whereas the selectivi-
ty to long-chain branched hydrocarbons mostly depends on
steric effects.
Introduction
Ultra-clean liquid fuels can be produced from alternatives to
petroleum sources such as natural gas, biomass, and coal
reaction selectivity to specific hydrocarbon fractions is, there-
fore, a main challenge of FT synthesis.
through intermediate transformation into syngas (H +CO) and
Several methods were proposed to control the selectivity of
this reaction. FT synthesis can be combined with hydrocracking
and isomerization of long-chain hydrocarbons to restrict hy-
drocarbon distribution to a specific range. The isomerization
and cracking of FT hydrocarbons would lead to iso-paraffins or
diesel fuels constituted by C –C hydrocarbons. A possible
2
subsequent Fischer–Tropsch synthesis (FT) over iron- or Co-
[
1–3]
based catalysts.
Low-temperature FT synthesis over a Co-
containing catalyst is very attractive because of the high intrin-
sic activity of Co, higher conversion per single cycle, and
[2]
better catalyst stability compared to iron catalysts. FT synthe-
10
20
[
4]
sis is a “nontrivial surface polymerization reaction”, which
produces a broad distribution of linear hydrocarbons; the reac-
tion selectivity is described typically using the Anderson–
option to produce these fuels directly from syngas would be
to use hybrid or composite catalysts that contain both an
active FT component (Co, Fe, or Ru catalyst) and an acid cata-
lyst active in cracking and isomerization.
[5,6]
Schulz–Flory (ASF) distribution model.
The direction of the
As a result of their high acidity and stability, zeolite and
metal/zeolite materials have been proposed as some of the
most efficient catalysts for hydrocarbon cracking and isomeri-
[
a] Dr. V. Subramanian, Dr. C. Lancelot, Dr. S. Heyte, Dr. J. Thuriot, Prof. S. Paul,
Dr. V. V. Ordomsky, Dr. A. Y. Khodakov
UnitØ de catalyse et de chimie du solide (UMR 8181 CNRS)
UniversitØ Lille 1-ENSCL-EC Lille
[
3,7]
zation under conditions close to FT synthesis.
Hydrogenoly-
[8]
sis can also affect hydrocarbon distribution. An FT catalyst
can be integrated with a zeolite or metal/zeolite in several dif-
ferent ways: (1) in the dual-bed configuration of the catalytic
reactor with the FT catalyst in the first layer and the zeolite or
Bat. C3, CitØ Scientifique
5
9655 Villeneuve d’Ascq (France)
E-mail: vitaly.ordomsky@univ-lille1.fr
andrei.khodakov@univ-lille1.fr
[9,10]
[11–14]
metal/zeolite in the second;
(2) as core–shell catalysts;
[
b] Prof. V. L. Zholobenko
School of Chemistry and Physics
Keele University
(
3) as hybrid catalysts prepared by the mechanical mixing of
[
15,16]
the FT catalyst with the zeolite;
(4) by the introduction of
[
17–19]
Staffordshire, ST5 5BG (UK)
the FT active metal into the zeolite by impregnation.
[
c] Dr. K. Cheng
The dual-bed systems for direct iso-paraffins synthesis re-
quire high volumes and difficult process control. The synthesis
of core–shell catalysts usually involves rather sophisticated
State Key Laboratory of Physical Chemistry of Solid Surfaces
College of Chemistry and Chemical Engineering
Xiamen University
[
14]
preparation methods.
A number of publications have ad-
Xiamen 361005 (China)
dressed the direct synthesis of gasoline or middle distillates
over hybrid catalysts prepared by the mechanical mixing of
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500777.
ChemCatChem 2016, 8, 380 – 389
380
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[15,16,20–22]
supported Co catalysts with zeolites.
Different types of
zeolites such as ZSM-5, MCM-22, IM-5, ITQ-2, BEA, MOR, and Y
with and without noble-metal promoters have been used for
the preparation of such hybrid catalysts. A high ratio of
branched to linear paraffins (C /C ) has been observed over
iso
n
[
22]
hybrid catalysts that contain Co/SiO and Pd/BEA or Pd/Y. At
2
the same time, the slow diffusion of long-chain paraffins
through the pores of 10-membered ring zeolites, such as ZSM-
5
, leads to the preferential participation of the external surface
[
20]
acid sites, which results in a lower isomerization efficiency.
Impregnation is another method for the preparation of bi-
[17–13,23–25]
functional Co-zeolite catalysts.
A major drawback of
this procedure, however, could be the neutralization of the
zeolite acid sites by the introduced Co cations. This could
result in a simultaneous decrease in the number of acidic sites
available for hydrocarbon cracking and isomerization and
a lower extent of Co reduction as isolated Co ions in the zeo-
lite cationic sites are very difficult to reduce to metallic Co. The
choice of the best preparation method for bifunctional cata-
lysts for the direct synthesis of iso-paraffins from syngas would
depend on the zeolite acidity and its porous structure.
Figure 1. N
erence Co/SiO
therms are offset for clarity.
2
adsorption–desorption isotherms obtained at 77 K over the ref-
2
and Co-zeolite catalysts prepared by impregnation. The iso-
The main aim of this work is to elucidate the structural ef-
fects such as localization of Co species, zeolite pore structure,
and zeolite acidity on the properties of Co-zeolite bifunctional
catalysts prepared by impregnation or mechanical mixing and
the catalytic performance for the direct synthesis of iso-paraf-
fins from syngas. Zeolites with different porous structures, that
is, ZSM-5, mordenite (MOR), and beta-zeolite (BEA), have been
used in this work. The catalysts have been characterized by
low-temperature N2 adsorption, TEM, SEM, temperature-pro-
grammed reduction (TPR), and FTIR spectroscopy with ad-
sorbed pyridine (Py) and CO. The catalytic performance has
been evaluated on the REALCAT platform by using a Flowrence
high-throughput system (Avantium) equipped with 16 parallel
fixed-bed millireactors.
Impregnation with Co and Pt does not lead to any signifi-
cant changes in the N adsorption isotherms of ZSM-5, MOR,
2
and the silica support. Impregnation of the BEA zeolite leads
to the disappearance of the high-pressure hysteresis loop and
the appearance of a new sharp step in the isotherm at P/P0
ꢀ0.2. This is probably related to the rearrangement of small
zeolite crystallites in the agglomerates of the impregnated BEA
sample.
The textural properties are presented in Table 1. A significant
decrease in the zeolite surface area and pore volume is ob-
served after impregnation with Co and Pt. This decrease is
more significant in BEA (~50%) followed by MOR (~25%) and
ZSM-5 (17%). Notably, the surface area and pore volume of
Co/ZSM-5 only decrease slightly after impregnation relative to
the parent zeolite. The difference between the surface area of
Co/ZSM-5 and Co/MOR and their parent zeolites may be as-
signed to the effect of zeolite “dilution” in the impregnated
catalysts because of the presence of 20 wt% of Co. The 3D
pore structure of BEA (0.76”0.64 nm) is more open that of
MOR (0.65”0.7 nm; 1D) and ZSM-5 (~0.55 nm; 3D). A more
significant decrease in the BET surface area and micropore
volume of BEA compared to that of ZSM-5 and MOR (Table 1)
could be assigned, therefore, to a much easier diffusion of Co
species into the zeolite pores during the impregnation to lead
to a higher Co fraction in the BEA pores.
Results and Discussion
Characterization
The N adsorption–desorption isotherms for the catalysts pre-
2
pared by impregnation of the parent zeolites are shown in
Figure 1. The ZSM-5 and MOR samples display Type I iso-
therms, which exhibit a sharp uptake at a low relative pressure
followed by a plateau without a hysteresis. This type of iso-
therm is usually observed for microporous materials.
Our results are consistent with the recent report of Qureshi
[
27]
The BEA zeolite exhibits a combination of Type I and IV iso-
et al.
who found that the reduction in the support pore
therms, which shows a high N uptake at low relative pressure
volume after the introduction of the metal is more significant
for larger-pore supports for Pd catalysts. This is indicative of
the preferential localization of Pd inside the pores. The less sig-
nificant decrease in the support volume suggests the presence
of metal particles on the support outer surface.
2
and a hysteresis loop at a high relative pressure (P/P >0.8).
0
The adsorption of N in the range of P/P =0.8–1.0 can be ex-
2
0
[26]
plained by the filling of the intercrystalline pores. The amor-
phous silica support demonstrates mainly N2 adsorption at
high pressures because of the filling of silica interparticle
spaces. SEM micrographs of the ZSM-5, MOR, and BEA zeolites
To further confirm the influence of the zeolite pore size on
the Co distribution between the zeolite pores and external sur-
face, the impregnated catalysts were analyzed by TEM
(
Figure S1, Supporting Information) display agglomerates of
[
28]
relatively small crystallites (<100 nm).
(Figure 2). In agreement with previous data, TEM images of
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
381
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
Table 1. Catalyst characterization.
Sample
Co content
wt%]
Pt content
[wt%]
Si/Al
ratio
N
2
adsorption data
Acidity
[
d]
[
(Py adsorption)
[
a]
2
À1
[b]
3
À1
[c]
3
À1
À1
À1
S
BET [m g
]
V
tot [cm g
]
V
mic [cm g
]
Brønsted [mmolg
]
Lewis [mmolg ]
Co/SiO
SiO
2
20
–
0.1
–
–
–
211
307
0.85
1.3
–
–
0
0
0
0
2
Co/ZSM-5
ZSM-5
20
–
0.1
–
13
13
232
305
0.16
0.19
0.1
0.12
507
1090
1427
240
Co/MOR
MOR
20
–
0.1
–
8
8
284
404
0.21
0.21
0.13
0.17
345
921
275
452
Co/BEA
BEA
20
–
0.1
–
9
9
282
556
0.18
0.67
0.07
0.15
88
486
84
834
[
a] Specific surface area was measured using the BET method. [b] Total pore volume (TPV) was calculated from the amount of vapor adsorbed at a relative
pressure close to unity. [c] Micropore volume was calculated using deBoer t-plot method. [d] Concentrations of acid sites were calculated from the intensity
À1
of Py FTIR bands at n=1545 and 1450 cm . The measurements were performed with the catalysts reduced in H
2
at 4008C.
2 2
Figure 3. H -TPR profiles for the reference Co/SiO and Co-zeolite catalysts.
presence of Co, which is located inside the zeolite crystals. The
Co localization inside the zeolite pores could also explain a sig-
nificant decrease in the surface area and pore volume of the
BEA zeolite after the impregnation.
The TPR profiles of the reference catalyst and Co-zeolites
catalysts are shown in Figure 3. The TPR curve of the reference
Co/SiO catalyst exhibits two groups of peaks at 200 and 300–
2
[
29,30]
4
008C. In agreement with previous reports,
the first peak
2
Figure 2. TEM images of the Co/SiO , Co/ZSM-5, Co/MOR, and Co/BEA cata-
can be assigned to the reduction of Co O to CoO. The second
3
4
lysts.
peak at 300–4008C is mainly because of the reduction of CoO
to metallic Co. Notably, the TPR profiles of the impregnated
Co-zeolite catalysts display only a single group of peaks
around 350–4008C. The TPR profiles of the Co-containing zeo-
the reference Co/SiO demonstrate Co nanoparticles distribut-
2
[
31–34]
ed irregularly with a diameter in the range of 10–50 nm. In the
impregnated calcined Co/ZSM-5, polycrystalline zeolite parti-
cles are covered by large agglomerates of Co nanoparticles
lites are consistent with previous results
and could be at-
tributed to the reduction of small CoO species located either
x
on the external surface or inside the zeolite pores. The position
of these peaks depends on the zeolite structure. The principal
peaks in the TPR profiles of Co/BEA and Co/MOR samples are
located at lower temperatures than that of Co/ZSM-5. This
would lead to a higher fraction of metallic Co in the reduced
Co/BEA and Co/MOR catalysts. Notably, the Co/ZSM-5 zeolite
also exhibits a small peak at 6008C that can be assigned to the
reduction of smaller Co oxide particles located in the narrower
pores of this zeolite. The peaks attributed to the reduction of
(
100–200 nm). The metal does not seem to be present inside
the ZSM-5 zeolite pores. These results are consistent with the
preferential Co localization on the external surface of the ZSM-
5
zeolite.
The TEM images of Co/MOR demonstrate a significantly
higher incorporation of Co in the zeolite crystallites, although
some fraction of Co nanoparticles can be observed on the sur-
face of the zeolite crystals. TEM images of Co/BEA show that
the zeolite crystallites are almost totally dark because of the
2+
Co ions in the zeolite cationic positions are not observed;
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
382
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
the reduction of these species would take place above 8008C,
as has been shown previously.
(Figures 3 and 4), which suggest preferential Co localization on
the external surface of ZSM-5.
[31–34]
The acid and metal sites in the initial and Co-containing re-
duced samples have been studied by FTIR spectroscopy. Sever-
al bands have been distinguished in the spectra of samples in
the region of the OÀH stretching vibrations (Figure 4). The
The FTIR spectra recorded after Py adsorption on the re-
duced catalysts are presented in Figure 5. No IR bands attribut-
ed to adsorbed Py are observed in the spectrum of the Co/
Figure 5. FTIR spectra observed after the adsorption of Py over the parent
zeolites and Co-zeolite catalysts.
Figure 4. FTIR spectra in the OH region of the parent zeolites and Co-zeolite
catalysts.
SiO catalyst. This indicates the absence of both Brønsted and
2
Lewis acid sites (BAS and LAS) in this catalyst. With the zeolite
catalysts, the chemisorbed Py exhibits characteristic bands at
À1
À1
band observed at n=3743 cm is attributed to the external si-
n=1545 and ~1450 cm
(PyH ) and to Py adsorbed on LAS (which may include Co
ions), respectively. Py adsorbed on both the LAS and BAS also
assigned to the pyridinum ion
+
2+
lanol groups. This band has the highest intensity in silica and
À1
in the BEA zeolite. The band at n=3735 cm is observed only
À1
in the BEA zeolite and is usually attributed to the internal sila-
displays a band at n=1490 cm . The effective strength of the
À1
nol groups. The bands at n=3610 and 3664 cm are caused
BAS was evaluated from the variation of the intensity of the
À1
by the bridged acid hydroxyl groups (SiÀ(OH)ÀAl) and extra-
band at n=1545 cm as a function of the Py desorption tem-
framework AlOH species, respectively. The band at n=
perature. Py desorption proceeds more easily from the BAS of
BEA compared to that of MOR and ZSM-5 zeolites (Figure 6).
À1
3
610 cm has a significantly higher intensity in the spectra of
MOR and ZSM-5 compared to that in the spectrum of the BEA
À1
zeolite. The band at n=3782 cm is observed only in the
spectrum of BEA and it can be assigned to hydroxyl groups at-
tached to three-coordinate Al atoms, similar to those in alumi-
[35,36]
na.
Thus, the spectrum of BEA indicates the presence of
a considerable number of defect sites. Indeed, it has been
shown that the BEA structure can be dealuminated readily
during calcination with the deposition of extraframework Al
[37,38]
species inside the channels.
Impregnation with Co does not affect the FTIR spectrum of
the silica-supported catalyst significantly, and only a slight de-
crease in the intensity of silanol groups is observed. In con-
trast, the impregnation of the zeolites with Co leads to a major
decrease in the intensity of the FTIR bands attributed to the
bridged hydroxyl, silanol, and extraframework AlOH groups.
These bands disappear almost completely in the spectra of the
Co/BEA catalyst, which indicates a higher Co dispersion in this
zeolite and significant Co incorporation inside the zeolite
pores as well as a high coverage of the external surface of BEA
by Co species. The effect of impregnation on the intensity of
the OH groups is less pronounced for the MOR and ZSM-5
samples. Indeed, a moderate decrease in the intensity of the
bridged OH groups in the spectrum of the ZSM-5 zeolite after
the impregnation is consistent with a smaller fraction of Co
species located inside the pores of these zeolites. These results
Figure 6. Desorption of Py from the BAS of the parent zeolites measured by
FTIR with the increase of temperature.
The MOR and ZSM-5 zeolites demonstrate a similar effective
strength of the acid sites with a slightly higher desorption tem-
perature for MOR. These results are in line with previous am-
monia temperature-programmed desorption (TPD) data for the
[
39]
same zeolites.
The concentration of BAS and LAS calculated using the pub-
are in good agreement with the N adsorption and TEM data
lished extinction coefficients for the bands at n=1545 and
2
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
383
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
À1[40]
1
450 cm
are presented in Table 1. The results suggest that
lyst this band results from CO adsorption on Co cations in
amorphous Co silicates.
impregnation with Co and Pt leads to a significant decrease in
the concentration of BAS in the zeolites. The effect is more
pronounced for the BEA zeolite compared to MOR and ZSM-5.
Indeed, the reduced Co/ZSM-5 catalyst retains ꢀ50% of
Brønsted acidity, whereas the concentration of BAS in Co/BEA
is only 18% compared to that of the parent zeolite. In contrast,
a reversed order is observed for Py adsorption on LAS in the
impregnated reduced catalysts. The number of LAS in ZSM-5 is
increased significantly in the presence of Co and Pt in compari-
son with the parent sample. The effect is smaller in the case of
the MOR zeolite. At the same time, the parent BEA zeolite al-
ready demonstrates significant Lewis acidity because of its
defect-rich structure in comparison with ZSM-5 and MOR. Im-
pregnation with Co leads to a sharp decrease in the intensity
of the Py-LAS band in the spectrum of Co/BEA.
In the case of Co/ZSM-5 and Co/MOR, the intensity of the
À1
band at n=2205 cm is high because of CO adsorption over
a large number of cationic Co sites inside the zeolite pores.
The intensity of this band is significantly lower for Co/BEA.
These results complement our Py adsorption data, which show
a higher concentration of unreduced cations in ZSM-5 in com-
parison with that in MOR and BEA.
À1
The bands below n=2100 cm can be attributed to CO ad-
0
sorbed on reduced Co . The CO bands in the region of n=
À1
À1
2050–2070 cm show a ~10 cm blueshift for Co/zeolite sam-
0
ples as compared to COÀCo sites in the silica-supported cata-
[
42,43]
lyst.
This is probably because of the smaller particle size
and positive charge on the Co metal nanoparticles induced by
[
44]
their interaction with the zeolite BAS. The spectrum of the
À1
The decrease in the number of BAS in Co/ZSM-5, Co/MOR,
and Co/BEA relative to the parent zeolites can be accounted
for by the ion exchange of the bridging OH group protons for
Co/BEA catalyst shows an additional band at n=1887 cm
that can be assigned to the bridged Co-CO-Co. The band of
CO adsorbed on metallic Co sites exhibits the highest intensity
over Co/ZSM-5, which can be explained by CO adsorption on
highly dispersed Co on the ZSM-5 external surface.
2+
Co ions and by partial removal of the AlÀOH groups. Both
the ion exchange and presence of incompletely reduced CoO
x
species could explain the significant increase in Lewis acidity
over Co/ZSM-5 as Py can interact with cationic Co species. The
weak or negative effect of the introduction of Co on Lewis
acidity in BEA and MOR may be because of a more complete
Co reduction in these zeolites as confirmed by TPR. Further-
more, the presence of a significant amount of Co inside the
large-pore BEA and MOR (possibly in the metallic state) may
result in pore blocking, which hinders Py access to the LAS.
Thus, characterization of the impregnated Co-zeolite cata-
lysts using IR spectroscopy, electron microscopy, and adsorp-
tion–desorption measurements suggests a significant influence
of the zeolite porous structure on the Co distribution between
the external and internal surface of the zeolite and its reduci-
bility. Co species in the BEA zeolite are reduced more easily
compared to that of MOR and ZSM-5. A higher fraction of re-
duced Co is present inside the more open pore structure of
BEA. In addition, a higher concentration of stronger acid sites
is found in the Co/ZSM-5 catalyst compared to that of Co/MOR
and Co/BEA.
Indeed, N adsorption and TEM data suggest a much higher
2
Co concentration inside the pores of BEA and MOR.
The FTIR spectra of CO adsorbed on the Co/SiO catalyst
2
and impregnated zeolites are shown in Figure 7. The catalysts
Catalytic performance in FT synthesis
Catalytic activity
The results of catalytic tests for the studied Co catalysts, mea-
sured in a high-throughput Flowrence fixed-bed millireactor
system, are presented in Table 2 and Figures 8–11.
Hydrocarbons and water are the major reaction products,
whereas the CO selectivity was always lower than 2%. The ref-
2
erence Co/SiO catalyst exhibited stable activity during 40 h
2
with a conversion of ꢀ40% (Figure 8).
Notably, the catalysts prepared by the ion exchange of the
zeolites with cobalt nitrate have not shown any measurable ac-
tivity in FT synthesis. Indeed, FT synthesis occurs on Co metal
sites. The cationic metal sites present in the ion exchange posi-
tions of zeolites cannot be reduced under the reaction condi-
tions and do not participate in the FT synthesis.
Figure 7. IR bands that result from CO adsorption at room temperature on
the reference and composite catalysts (PCO =20 Torr)
À1
exhibit two IR bands at n=2190–2205 and 2060–2070 cm at-
tributed to CO adsorbed species. The bands in the region of
The catalytic performance of the samples prepared either by
impregnation or mechanical mixing is affected strongly by the
zeolite used and the preparation method. FT synthesis is a com-
plex reaction, and the reaction rate could be affected by the
number and intrinsic activity of the Co active sites, zeolite acid-
ity, and reagent and product diffusion. Notably, in this work, FT
synthesis on Co catalysts exhibits a different selectivity espe-
À1
n=2200–2205 cm can be assigned to CO adsorption on Co
cations; the stability of these bands suggests that they cannot
[41,42]
be caused by CO adsorption on the zeolite LAS.
This band
has a low intensity in the reference Co/SiO and is shifted to
2
lower frequencies in comparison with the zeolite-based cata-
[42]
lysts. In agreement with previous data, in the Co/SiO cata-
2
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
384
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
À1 CoÀ1
, T=2508C).
Table 2. Catalytic properties of hybrid and impregnated catalysts (P=2 MPa, H
Catalyst Selectivity [%]
–C12 normal
2
/CO=2, GHSV=34 Lh
g
X
CO [%]
Ratio
(iso/total)
CH
4
CO
2
C
2
–C
4
C
5
C
5
–C12 olefinic
C
5
–C12 iso
C
12+
C
5
–C12
C12+
Co/SiO
Co/ZSM-5
Co/SiO +ZSM-5
Co/MOR
Co/SiO
Co/BEA
Co/SiO
2
37.9
26.8
26.4
40.1
39.7
17.5
54.9
6.1
11.7
6.0
9.2
6.7
2.1
0.5
0.7
0.6
1.7
0.7
1.5
7.7
12.7
8.7
10.1
10.3
13.6
9.2
10.0
5.3
5.1
12.1
6.4
11.2
9.5
6.7
1.2
1.0
1.1
1.4
0.6
1.3
0.9
16.4
18.0
6.0
13.3
7.6
66.5
52.2
60.5
60.9
60.2
55.8
65
0.06
0.7
0.74
0.31
0.6
0.01
0.14
0.08
0.19
0.14
0.33
0.16
2
2
+MOR
+BEA
10.5
7.4
0.4
0.36
2
6.1
Hydrocarbon selectivities
The hydrocarbon distribution on the hybrid catalysts prepared
by mechanical mixing and impregnated samples is shown in
Table 2. The selectivity to methane is ꢀ6% over the reference
and hybrid catalysts. In comparison with the reference and
hybrid catalysts, the impregnated catalysts exhibit an almost
two times higher selectivity to methane. In contrast to the
hybrid samples, the impregnated catalysts contain some Co
[
45]
metallic sites inside the zeolite. On one hand, the slow diffu-
sion of long-chain hydrocarbons inside the zeolite pores and
their interaction with metal particles located inside the zeolites
would lead to hydrocarbon secondary hydrocracking and in-
crease methane production. On the other hand, Iglesia
[
46,47]
et al.
showed that diffusion limitations in the catalyst
pores could be more severe for CO in comparison with H . This
2
would lead to a higher effective H /CO ratio in the catalyst
2
Figure 8. CO conversion as a function of time (P=20 bar, H
2
/CO=2,
pores and respectively to higher selectivity to methane and
lighter hydrocarbons.
À1
À1
GHSV=34 Lh
gCo , T=2508C) over the hybrid and impregnated catalysts.
Mostly linear hydrocarbons and very small amounts of iso-
paraffins were observed over the Co/SiO2 reference catalyst
(Figure 9, Table 2). This is consistent with previous re-
cially towards the long-chain hydrocarbons (C20+) as discussed
further in greater detail.
[
2,4,46,47]
ports
and could be attributed to the absence of strong
The hybrid catalysts prepared by the mechanical mixing of
acid sites in the silica support (Figure 5), which would initiate
formation of carbocations with subsequent skeletal isomeriza-
tion, possible b-scission, and the formation of short-chain iso-
Co/SiO with zeolites display different activities in comparison
2
with the reference catalyst. The CO conversion on Co/
[
48]
[46,47]
SiO +ZSM-5 is lower than that of the reference Co/SiO cata-
mers. In agreement with previous reports,
the olefin-to-
2
2
lyst. The conversion over Co/SiO +MOR is similar to that of the
paraffin ratio increases with an increase in the hydrocarbon
chain length (Figure 9). This could be because of the olefin sec-
2
reference catalyst, whereas the Co/SiO +BEA catalyst shows
2
a significantly higher activity than that of the reference Co/
ondary reactions in the Co/SiO catalyst: hydrogenation and
2
[
46,47]
SiO . Interestingly, the catalysts prepared by the impregnation
possible olefin insertion in the FT polymerization chain.
2
of the ZSM-5 and MOR zeolites demonstrate activity similar to
their counterparts prepared by mechanical mixing. This can be
explained by the preferential location of Co in these impreg-
nated catalysts mainly on the surface of zeolite crystals and
low Co content inside the zeolite pores. The Co/BEA zeolite
prepared by impregnation shows a lower FT reaction rate than
the hybrid catalyst prepared by mechanical mixing. The cata-
lytic results are consistent with CO adsorption data. Low CO
adsorption on metallic Co species in BEA zeolite could explain
the lower activity of the catalyst (Figure 7). Significantly higher
deactivation rates over impregnated Co/BEA and Co/MOR
were observed in comparison with that of the hybrid catalysts,
whereas the catalytic performance of Co/ZSM-5 was relatively
stable.
The probability of olefin secondary hydrogenation reactions
could be higher for long-chain hydrocarbons because of their
slow diffusion in FT catalysts.
The hydrocarbon selectivities observed on both impregnat-
ed and hybrid Co-containing zeolite catalysts are dependent
on the zeolite acidity, zeolite porous structure, and catalyst
preparation method. Indeed, the selectivity towards the lighter
C5–C12 olefins is much lower for both impregnated and hybrid
Co-zeolite catalysts as compared to Co/SiO (Table 2). This is
2
probably because of higher olefin reactivity on the zeolite acid
sites. The low olefin selectivity in the bifunctional Co-zeolite
catalysts coincides with the presence of significant amounts of
[
48]
iso-paraffins in the reaction products. Previous studies sug-
gest that paraffin isomerization on the metal-containing zeo-
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
385
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
BEA<MOR<ZSM-5 for the cata-
lysts prepared using either im-
pregnation
or
mechanical
mixing. The reason for the
higher isomerization activity of
MOR- and ZSM-5-supported cat-
alysts towards the short-chain
C5–C12 hydrocarbons in compari-
son with BEA could be the sig-
nificantly higher number of
strong BAS monitored in this
work by Py adsorption (Figures 5
and 6).
The hybrid Co/SiO +ZSM-5
2
and impregnated Co/ZSM-5 cat-
alysts show similar C –C olefin
5
12
and
paraffin
selectivities
(
Table 2). In both samples, Co is
located largely on the external
surface of the zeolite crystals,
hence, the zeolite acid sites are
only slightly affected by Co spe-
cies and, therefore, could be
available for the isomerization of
short-chain hydrocarbons. The
Co/MOR impregnated catalyst
2
Figure 9. Selectivity to olefins, paraffins, and isomers as a function of the chain length over the reference Co/SiO ,
hybrid, and impregnated Co-zeolite catalysts.
demonstrates
a
significantly
lower activity in the isomeriza-
tion of short-chain hydrocarbons
compared to the relevant hybrid
catalyst. This could be because
of the partial blocking of strong
acid sites by Co species in the
1
D channels of the impregnated
Co/MOR catalyst. Both impreg-
nated and hybrid Co BEA cata-
lysts exhibit a low C –C hydro-
5
12
carbon isomerization activity
Table 2). This should be attribut-
(
ed to the lower number of
strong BAS in the BEA zeolite
compared to that of ZSM-5 and
MOR. It can be concluded that
the rate of C –C hydrocarbon
5
12
isomerization is principally af-
fected by the number of strong
acid sites in the zeolite catalysts.
No noticeable effect of the zeo-
Figure 10. Distribution of isomers and paraffins in the liquid products over the Co-zeolite catalysts prepared by
mechanical mixing or impregnation.
lites proceeds by the intermediate formation of olefins. Our re-
sults show that light olefins, which are always present during
FT synthesis over Co catalysts, might participate in the isomeri-
zation reaction. The isomerization, however, seems to also in-
volve paraffins. Indeed, the amounts of iso-paraffins produced
are significantly higher than the amounts of olefins produced
lite pore sizes on the C –C hydrocarbon isomerization activity
5 12
has been detected.
Different selectivity results have been obtained for the C12+
hydrocarbons in comparison with the short-chain C –C paraf-
5
12
fins. For the reference Co/SiO , hybrid, and impregnated Co-
2
zeolite catalysts, the distribution of long-chain hydrocarbons
in conventional FT synthesis over the reference Co/SiO cata-
shows a maximum at ꢀC (Figure 10). In the Co-zeolite cata-
2
20
lyst. The ratio of isomers to the total amount of hydrocarbons
in the C –C range increases from 0.3 to 0.7 in the order
lysts prepared by either mechanical mixing or impregnation,
the distribution of C20+ hydrocarbons is rather different from
5
12
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
386
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
for short-chain hydrocarbons,
the isomerization of which is
mostly affected by the zeolite
acidity.
The distribution of long-chain
hydrocarbons is also affected by
the catalyst preparation method.
The fraction of the C20+ hydro-
carbons is much smaller for the
catalysts prepared by impregna-
tion compared with the mechan-
ically mixed hybrid catalysts
(
Figure 10). This indicates a great-
Figure 11. The ASF chain growth probabilities determined for hydrocarbons over the Co-zeolite catalyst prepared
by a) mechanical mixing or b) impregnation.
er role of hydrocracking for the
impregnated catalysts, which
contain a higher fraction of Co
species inside the zeolites crys-
that observed over the conventional Co/SiO2 catalyst. The
long-chain hydrocarbon selectivity is affected by both the zeo-
lite structure and the catalyst preparation procedure. Over Co/
tallites, particularly for the large-pore BEA and MOR. The close
proximity between the Co metallic species located in the zeo-
lite pores and zeolite acid sites favors secondary reactions of
long-chain hydrocarbons, that is, cracking and isomerization. In
contrast, spatial effects on the selectivity towards short-chain
n- and iso-paraffins are less substantial because of the relative-
ly fast diffusion of the C –C hydrocarbons in the catalyst
SiO +ZSM-5 and Co/ZSM-5, the distribution of the C20+ normal
2
hydrocarbons has a similar shape and similar chain growth
probability (a=0.87; Figure 11). The fraction of long-chain C20+
branched hydrocarbons is relatively small for both ZSM-5-
based catalysts prepared by impregnation and mechanical
mixing. The activity of these catalysts in the isomerization of
longer C20+ hydrocarbons is relatively low. This can be ex-
plained by diffusional constraints for the C20+ hydrocarbons in
the zeolite pores. Our data indicate that Co species in Co/ZSM-
5
12
pores.
This study provides new insights into the efficient control of
hydrocarbon selectivity in FT synthesis. In particular, the Co/
BEA catalysts prepared by impregnation exhibit a C –C selec-
10
20
tivity of ꢀ60% (Table 2, Figure 10), which is among the best
results reported. The methane selectivity remains rela-
[
3,8,13,49]
5
and Co/SiO +ZSM-5 are located mostly on the zeolite exter-
2
nal surface, which suggests that FT synthesis in these catalysts
occurs on the zeolite external surface and that only short-chain
hydrocarbons can diffuse and isomerize effectively inside the
pores of ZSM-5 zeolite. Indeed, the catalysts prepared using
ZSM-5 exhibit a higher activity towards the isomerization of
C5–C12 hydrocarbons.
tively low (<10%). This makes the Co catalysts supported by
large-pore BEA zeolites prepared by impregnation interesting
candidates for the direct synthesis of selected hydrocarbon
fractions (iso-paraffins, middle distillates) from syngas.
Conclusions
The fraction of C20+ long-chain hydrocarbons is significantly
lower for the MOR- and BEA-based zeolites (Table 2, Figure 10).
The distribution of long-chain hydrocarbons on both zeolites
shows noticeable deviations from the standard ASF model
Sample characterization and catalytic results indicate the
strong influence of the zeolite pore structure and acidity on
the formation of Co species and on the performance of Co-
zeolite bifunctional catalysts in Fischer–Tropsch synthesis. In
the catalysts prepared by impregnation, the distribution of Co
species between the pores and the external surface of ZSM-5,
MOR, and BEA is affected strongly by the zeolite pore struc-
ture. Such Co species can be present in the zeolite pores in
both the metallic and cationic state. A greater fraction of Co is
located inside the large-pore MOR and BEA zeolites.
(
Figure 11). It appears that larger zeolite pores favor the diffu-
sion of long-chain hydrocarbons into the zeolite pores with
their subsequent cracking. The most significant decrease in the
fraction of long-chain hydrocarbons is observed for the BEA-
based catalysts, which have the most open pore system
among the investigated zeolites. GC analysis is also consistent
with visual aspects of the liquid samples produced over differ-
ent catalysts. The wax produced over Co/BEA is liquid in com-
parison with the solid wax formed over Co/ZSM-5 and Co/MOR
The catalytic performance in Fischer–Tropsch synthesis ap-
pears to be a function of the Co distribution between the zeo-
lite pores and the outer surface, Co reducibility, and the zeolite
pore structure and acidity. In particular, the rate of hydrocrack-
ing and isomerization of hydrocarbons is dependent on both
the zeolite acidity and its pore structure. Different zeolite char-
acteristics have an effect on hydrocarbon selectivity for short-
and long-chain hydrocarbons. The selectivity in the C –C hy-
(
Figure 10). The fraction of the branched C12+ hydrocarbons is
much more significant for the BEA- and MOR-based catalysts
in comparison with the ZSM-5-based counterparts, although
ZSM-5 has a higher concentration of strong BAS compared to
BEA. For the long-chain hydrocarbons, our selectivity results
point to a more significant effect of open structure and larger
pore size of BEA on the catalyst hydrocracking and isomeriza-
tion activity rather than the effect of its acidity. This is different
5
12
drocarbon range is mostly affected by the number of strong
acid sites, and a higher concentration of iso-paraffins was ob-
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
387
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
served for ZSM-5 zeolite. For longer-chain hydrocarbons, the
zeolite activity in isomerization and cracking is mainly influ-
enced by spatial effects such as the size and open character of
the zeolite pore structure as well as the location of the Co spe-
cies. Large-pore zeolites exhibit a higher selectivity to long-
chain iso-paraffins and a lower selectivity to C20+ hydrocarbons.
The selectivity of the long-chain iso-paraffins is also favored by
the proximity and close contact between Co metallic species
and Brønsted acid sites.
The catalysts were reduced in an atmosphere of H at 4008C with
2
subsequent evacuation. The adsorption of CO over the impregnat-
ed catalysts was performed at RT (PCO was varied from 0 to
2
0 Torr). The adsorption of Py was performed at 1508C for 10 min.
The excess of probe molecules was evacuated at the same temper-
ature for 30 min. The evacuation was repeated in a stepwise fash-
ion every 508C up to 4508C.
Catalytic tests
CO hydrogenation was performed on the REALCAT platform by
[51]
Experimental Section
using a Flowrence high-throughput system (Avantium) equipped
with 16 parallel fixed-bed millireactors (d=2 mm) with plug-flow
hydrodynamics operated at a total pressure of 20 bar, 2508C, H2/
Catalyst preparation
CO=2 molar ratio, and gas hourly space velocity (GHSV) of
Commercial amorphous silica (CARIACT Q-10, Fuji Silysia) was used
for the preparation of a conventional silica-supported FT catalyst.
The silica-supported catalyst (20 wt% Co and 0.1 wt% Pt) was pre-
pared by incipient wetness impregnation using aqueous solutions
of Co(NO ) ·6H O (Sigma–Aldrich) and [Pt(NH ) ](NO ) (Sigma–Al-
À1 À1
3
4 LgCo h . The catalyst loading was 50 mg per reactor in the
case of impregnated catalysts. In the case of hybrid catalysts, the
same amount of Co and zeolite (by mass) was placed in the reac-
tor, which corresponded to 90 mg of the catalyst. Before the cata-
3
2
2
3 4
3 2
lytic tests, all the samples were activated in a flow of H at atmos-
2
drich). The ZSM-5, MOR, and BEA zeolites with Si/Al ratios of 13, 8,
and 9, respectively, were provided by Zeolyst.
pheric pressure for 10 h at 4008C. During the activation step, the
À1
temperature ramp was 38Cmin . After activation, the catalysts
The Co-containing zeolites were prepared by incipient wetness im-
were cooled to 1808C, and a flow of premixed syngas was intro-
duced gradually into the reactor. When the pressure attained
20 bar, the temperature was increased slowly to 2508C. The gas-
eous reaction products were analyzed by online GC. The analysis
of permanent gases was performed by using a Molecular Sieve
column and a TCD. CO and C –C hydrocarbons were separated
pregnation of the zeolite with Co(NO ) ·6H O and [Pt(NH ) ])(NO ) .
3
2
2
3 4
3 2
The Co and Pt content was 20 and 0.1 wt%, respectively. No wash-
ing of the impregnated samples was involved in the catalyst prep-
aration. The prepared catalysts were calcined in air at 4508C. The
hybrid catalysts were prepared by the mechanical mixing of differ-
ent zeolites with the silica-supported catalyst. The ratio of zeolite
to Co in the mechanical mixtures was 4:1 w/w, that is, the same as
that in the impregnated catalysts. For the catalytic tests, the Co
and Pt loadings in the reactor were exactly the same in the im-
pregnated samples and in hybrid catalysts prepared by mechanical
mixing.
2
1
4
by using a PPQ column and analyzed by using a TCD. The C –C
5
12
hydrocarbons were analyzed by using a CP-Sil5 column and
a flame ionization detector. High-molecular-weight products were
collected at atmospheric pressure in vials heated at 608C. CO that
contained 5% helium was used as an internal standard to calculate
the CO conversion. The product selectivity (S) was reported on
a carbon atom basis.
Characterization
Acknowledgements
The BET surface area, pore volume, and average pore diameter
were determined by N low-temperature adsorption by using a Mi-
2
cromeritics ASAP 2000 automated system. The samples were de-
gassed under vacuum at <10 mmHg at 3008C for 4 h before N2
The authors gratefully acknowledge the support of the French
National Research Agency (ANR-11-IS09–0003 and ANR-12-BS07–
physisorption. The total pore volume (TPV) was calculated from
0
029). The REALCAT platform benefits from a Governmental sub-
vention administrated by the French National Research Agency
ANR) within the frame of the “Future Investments” program (PIA)
the amount of vapor adsorbed at a relative pressure close to unity
by assuming that the pores are filled with the condensate in the
liquid state. The catalyst external surface area and micropore
(
[50]
with the contractual reference “ANR-11-EQPX-0037”. The Nord-
Pas-de-Calais Region and the FEDER are acknowledged for their
financial contribution to the acquisition of the equipment of the
platform. The Centrale Initiative Foundation is also acknowl-
edged for its financial support.
volume were calculated using the deBoer t-plot method.
Electron micrographs were obtained by using a MIRA-LMH (Tescan)
SEM equipped with a field emission gun. The TEM observations of
the samples were obtained by using a Tecnai instrument equipped
with a LaB6 crystal operated at 200 kV. Before the analysis, the
samples were dispersed by ultrasound in ethanol for 5 min, and
a drop of the suspension was deposited onto a carbon membrane
on a 300 mesh copper grid.
Keywords: acidity · cobalt · hydrocarbons · platinum · zeolites
The catalyst reducibility was studied by using a TPR system
[
1] B. H. Davis, M. L. Occelli, Advances in Fischer–Tropsch Synthesis Catalysts,
and Catalysis in Stud. Surf. Sci. Catal. 2006, 163, 1–420
equipped with a thermal conductivity detector (TCD). The samples
À1
were reduced in a flow of 5% H /Ar (60 mLmin ) and heated up
to 9008C at a rate of 108Cmin .
[2] A. Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 2007, 107, 1692–
2
À1
1744.
[
3] Q. Zhang, K. Cheng, J. Kang, W. Deng, Y. Wang, ChemSusChem 2014, 7,
251–1264.
4] H. Schulz, Catal. Today 2014, 228, 113–122.
The FTIR spectra were recorded by using a Nicolet ProtØgØ 460
1
À1
FTIR spectrometer at 2 cm resolution. Before the measurements,
[
À2
the catalysts pressed in self-supporting discs (~10 mgcm ) were
[5] G. Henrici-OlivØ, S. OlivØ, Angew. Chem. Int. Ed. Engl. 1976, 15, 136–141;
Angew. Chem. 1976, 88, 144–150.
activated in the IR cell attached to a vacuum line at 4008C for 4 h.
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
388
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[
[
[
6] C.-X. Xiao, Z. P. Cai, T. Wang, Y. Kou, N. Yan, Angew. Chem. Int. Ed. 2008,
7, 746–749; Angew. Chem. 2008, 120, 758–761.
7] R. Sethuraman, N. N. Bakhshi, S. P. Katikaneni, R. O. Idem, Fuel Process.
Technol. 2001, 73, 197–222.
8] X. Peng, K. Cheng, J. Kang, B. Gu, X. Yu, Q. Zhang, Y. Wang, Angew.
Chem. Int. Ed. 2015, 54, 4553–4556; Angew. Chem. 2015, 127, 4636–
[30] B. Ernst, A. Bensaddik, L. Hilaire, P. Chaumette, A. Kiennemann, Catal.
Today 1998, 39, 329–341.
[31] C. Resini, T. Montanari, L. Nappi, G. Bagnaco, M. Turco, G. Busca, F. Bre-
gani, N. Notaro, G. Rocchini, J. Catal. 2003, 214, 179–190.
[32] G. Bagnasco, M. Turco, C. Resini, T. Montanari, M. Bevilacqua, G. Busca,
J. Catal. 2004, 225, 536–540.
4
4
639.
9] T.-S. Zhao, J. Chang, Y. Yoneyama, N. Tsubaki, Ind. Eng. Chem. Res. 2005,
4, 769–775.
[33] X. Wang, H. Chen, W. M. H. Sachtler, Appl. Catal. B 2001, 29, 47–60.
[34] G. Espinosa, J. M. Domínguez, P. Morales-Pacheco, A. Tobon, M. Aguilar,
J. Benítez, Catal. Today 2011, 166, 47–52.
[
4
[
[
10] F. G. Botes, W. Bçhringer, Appl. Catal. A 2004, 267, 217–225.
11] B. Zeng, B. Hou, L. Jia, J. Wang, C. Chen, D. Li, Y. Sun, Catal. Sci. Technol.
[35] J. Pierez-Pariente, J. Sanz, V. FornØs, A. Corma, J. Catal. 1990, 124, 217–
223.
2
013, 3, 3250–3255.
[36] C. Paz›, A. Zecchina, S. Spera, A. Cosma, E. Merlo, G. Spanò, G. Girotti,
Phys. Chem. Chem. Phys. 1999, 1, 2627–2629.
[37] J. C. Jansen, E. J. Creyghton, S. L. Njo, H. van Koningsveld, H. van Bek-
kum, Catal. Today 1997, 38, 205–212.
[
[
[
[
12] G. Yang, C. Xing, W. Hirohama, Y. Jin, C. Zeng, Y. Suehiro, T. Wang, Y. Yo-
neyama, N. Tsubaki, Catal. Today 2013, 215, 29–35.
13] X. Li, J. He, M. Meng, Y. Yoneyama, N. Tsubaki, J. Catal. 2009, 265, 26–
3
4.
14] R. Xie, D. Li, B. Hou, J. Wang, L. Jia, Y. Sun, Catal. Commun. 2011, 12,
80–383.
15] N. Tsubaki, Y. Yoneyama, K. Michiki, K. Fujimoto, Catal. Commun. 2003,
, 108–111.
[38] M. Maache, A. Janin, J. C. Lavalley, J. F. Joly, E. Benazzi, Zeolites 1993, 13,
419–426.
[39] F. Lónyi, J. Valyon, Thermochim. Acta 2001, 373, 53–57.
[40] M. Guisnet, P. Ayrault, C. Coutanceau, M. F. Alvarez, J. Datka, J. Chem.
Soc. Faraday Trans. 1997, 93, 1661–1665.
3
4
[
[
[
16] Z.-W. Liu, X. Li, K. Asami, K. Fujimoto, Catal. Commun. 2005, 6, 503–506.
17] S. Bessell, Appl. Catal. A 1995, 126, 235–244.
18] D. J. Koh, J. S. Chung, Y. G. Kim, Ind. Eng. Chem. Res. 1995, 34, 1969–
[41] K. Chakarova, K. Hadjiivanov, Microporous Mesoporous Mater. 2009, 123,
123–128.
[42] A. Y. Khodakov, J. Lynch, D. Bazin, B. Rebours, N. Zanier, B. Moisson, P.
Chaumette, J. Catal. 1997, 168, 16–25.
[43] M. Jiang, N. Koizumi, T. Ozaki, M. Yamada, Appl. Catal. A 2001, 209, 59–
70.
[44] J. Ansorge, H. Foerster, J. Catal. 1981, 68, 182–185.
[45] S. Sartipi, K. Parashar, M. J. Valero-Romero, V. P. Santos, B. van der Lin-
den, M. Makkee, F. Kapteijn, J. Gascon, J. Catal. 2013, 305, 179–190.
[46] E. Iglesia, S. C. Reyes, R. J. Madon, S. L. Soled, Adv. Catal. 1993, 39, 221–
302.
1
975.
19] K. Jothimurugesan, S. K. Gangwal, Ind. Eng. Chem. Res. 1998, 37, 1181–
188.
[
[
[
[
1
20] A. Martínez, S. Valencia, R. Murciano, H. S. Cerqueira, A. F. Costa, E. F. S.
Aguiar, Appl. Catal. A 2008, 346, 117–125.
21] A. Martinez, J. Rollan, M. A. Arribas, H. S. Cerqueira, A. F. Costa, E. F. S.
Aguiar, J. Catal. 2007, 249, 162–173.
22] Z.-W. Liu, X. Li, K. Asami, K. Fujimoto, Fuel Process. Technol. 2007, 88,
1
65–170.
[47] E. Iglesia, Appl. Catal. A 1997, 161, 59–78.
[
[
23] S. Bessell, Appl. Catal. A 1993, 96, 253–268.
24] K. Cheng, L. Zhang, J. Kang, X. Peng, Q. Zhang, Y. Wang, Chem. Eur. J.
[48] V. G. Komvokisa, S. Karakouliaa, E. F. Iliopouloua, M. C. Papapetroua, I. A.
Vasalosa, A. A. Lappasa, K. S. Triantafyllidis, Catal. Today 2012, 196, 42–
55.
2015, 21, 1928–1937.
[
[
[
[
[
25] S. Sartipi, M. Alberts, M. J. Meijerink, T. C. Keller, J. Perez-Ramirez, J.
Gascon, F. Kapteijn, ChemSusChem 2013, 6, 1646–1650.
26] A. Simon-Masseron, J. P. Marques, J. M. Lopes, F. R. Ribeiro, I. Gener, M.
Guisnet, Appl. Catal. A 2007, 316, 75–82.
27] Z. S. Qureshi, P. B. Sarawade, M. Albert, V. D’Elia, M. N. Hedhili, K. KÅhler,
J.-M. Basset, ChemCatChem 2015, 7, 635–642.
28] J. Hong, E. Marceau, A. Y. Khodakov, A. Griboval-Constant, C. La Fon-
taine, V. Briois, Chem. Eur. J. 2012, 18, 2802–2805.
[49] C. Xing, G. Yang, M. Wu, R. Yang, L. Tan, P. Zhu, Q. Wei, J. Li, J. Mao, Y.
Yoneyama, N. Tsubaki, Fuel 2015, 148, 48–57.
[50] J. H. de Boer, B. C. Lippens, B. G. Linsen, J. C. P. Broekhoff, A. van den
Heuvel, T. V. Osinga, J. Colloid Interface Sci. 1966, 21, 405–414.
[51] http://avantium.com/uploads/brochures/avantium-chemicals-
flowrence.pdf.
29] W. Chu, P. A. Chernavskii, L. Gengembre, G. A. Pankina, P. Fongarland,
A. Y. Khodakov, J. Catal. 2007, 252, 215–230.
Received: July 12, 2015
Published online on November 3, 2015
ChemCatChem 2016, 8, 380 – 389
www.chemcatchem.org
389
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim