Preparation of highly active and dispersed platinum nanoparticles on mesoporous Al-MCM-48 and their activity in the hydroisomerisation of n-octane

Article abstract of DOI:10.1002/chem.200800182

Platinum nanoparticles supported on Al-MCM-48 materials have been prepared. The resultant catalysts have been characterized by means of XRD, N₂ physisorption experiments, scanning electron microscopy (SEM), transmission electron microscopy (TEM

Full text of DOI:10.1002/chem.200800182

DOI: 10.1002/chem.200800182
Preparation of Highly Active and Dispersed Platinum Nanoparticles on
Mesoporous Al-MCM-48 and Their Activity in the Hydroisomerisation
of n-Octane
Juan M. Campelo,[a] Adam F. Lee,[b] Rafael Luque,*[a, c] Diego Luna,[a]
Jose M. Marinas,[a] and Antonio A. Romero[a]
Abstract: Platinum nanoparticles sup-
ported on Al-MCM-48 materials have
been prepared. The resultant catalysts
have been characterized by means of
XRD, N2 physisorption experiments,
scanning electron microscopy (SEM),
duction (TPR), and diffuse reflectance
infrared Fourier-transform spectrosco-
py (DRIFTS). The activity of these
nanoparticles has been tested in rela-
tion to the hydroisomerisation of n-
octane. The catalytic activities were
typically 50%, with selectivities in the
isomerisation process in excess of
70%, favouring the formation of the 3-
methylheptane isomer with respect to
the 2- and 4-methylheptanes.
Keywords: alkanes · heterogeneous
catalysis
·
isomerisation
·
transmission
electron
microscopy
nanoparticles · platinum
(TEM), temperature-programmed re-
Introduction
become an interesting research avenue for both academic
researchers and industrial partners.[2–5] Akhmedov et al.
have recently reviewed the salient advances with regard to
the selective isomerisation of longer C7 chain alkanes.[2] The
conversion of longer C7 chain alkanes is not trivial as there
are two different and competing pathways, namely isomeri-
sation and cracking. The conventional reaction mechanism
is monomolecular. Alkanes are dehydrogenated on the met-
allic phase, and the alkenes generated are protonated on the
acidic sites of the catalyst to form alkylcarbenium ions. The
reaction (either isomerisation or cracking of the alkane)
then takes place on the acidic sites, and the alkene generat-
ed in the process is hydrogenated at the nearby metallic
sites, affording the isomerised-fragmented n-alkane
(Scheme 1). Thus, it is rather challenging to obtain high se-
lectivities in favour of branched isomers without an appreci-
able selectivity in favour of cracking. The b-scission of long-
chain paraffins has been reported to be the main reason for
this phenomenon.[6] Consequently, the preparation of mate-
rials with improved activities and selectivities favouring the
isomerisation of long-chain paraffins is highly desirable. Hy-
droisomerisation reactions are generally performed over bi-
functional catalysts having a noble metal function, which is
usually involved in hydrogenation–dehydrogenation process-
Environmental regulations have restricted the content of ar-
omatic and alkenic hydrocarbons in gasolines, which have a
detrimental effect on their production as well as on their
octane numbers. Light isoalkanes are required for the pro-
duction of more environmentally friendly and high-octane-
number gasolines. n-Alkanes are isomerised to increase the
octane number in the naphtha (transformation of linear-
chain paraffins to branched isomers with high octane num-
bers) as well as to induce significant improvements in sever-
al physical properties of the gasoline, including the pour
point and viscosity.[1]
Much effort has been devoted to the isomerisation of
longer C7 chain alkanes in the last few years, and it has
[a] Prof. J. M. Campelo, Dr. R. Luque, Prof. D. Luna,
Prof. J. M. Marinas, Dr. A. A. Romero
Departamento de Química Orgµnica, Facultad de Ciencias
Universidad de Córdoba, Campus Universitario de Rabanales
Edificio Marie Curie, 14014 Córdoba (Spain)
E-mail: rla3@york.ac.uk
[b] Dr. A. F. Lee
Department of Chemistry, The University of York
Heslington, YO10 5DD, York (UK)
[c] Dr. R. Luque
ꢀ
es, and acidic sites, which are involved in C C skeleton rear-
Current address: Green Chemistry Centre of Excellence
The University of York
Heslington, YO10 5DD, York (UK)
Fax : (+44)190-443-2705
rangements. The major parameters influencing the reaction
selectivity are the pore structure[7–9] and the acidity.[10–12] The
key to the successful preparation of active and selective cat-
5988
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5988 – 5995

FULL PAPER
chosen Pt loading in samples (typically 0.5 wt%) was dissolved in ethanol
and subsequently mixed with the Al-MCM-48, stirring the mixture in a
rotary evaporator at 408C for a few hours. Ethanol acts as the reducing
agent in the deposition of the Pt nanoparticles.[26] The materials were
then calcined in air (60 mLminꢀ1) at 4008C for 2 h and then reduced in
hydrogen (100 mLminꢀ1) at 4008C for 3 h. The Pt/Al-MCM-48 catalysts
are denoted as 0.5%Pt-XAl, where 0.5 wt% is the approximate Pt load-
ing and X (40 to 15) is the theoretical Si/Al ratio (in the synthesis gel) of
the Al-MCM-48 support. Materials with different Pt loadings (ranging
from 0.5 to 5% Pt) were also prepared for comparative purposes. A
0.5 wt% Pt loading was chosen as we have previously reported this to be
ideal for obtaining catalysts with optimum activity and selectivity in
favour of the isomerisation process.[9]
Characterization of the materials: Powder X-ray diffraction patterns
(XRD) were recorded on a Bruker AXS diffractometer employing CuKa
radiation (l=1.5418 Š), over a 2q range from 2 to 108, with a step size of
0.028 and counting time per step of 1 s. Extended scans were performed
over a 2q range of 5 to 858, with a step size of 0.018 and a counting time
per step of 4 s.
Nitrogen physisorption was measured with a Micromeritics ASAP2000
instrument at ꢀ1968C. Samples were degassed for 2 h at 1008C under
vacuum (p<10ꢀ2 Pa) and subsequently analysed. The linear part of the
BET equation (relative pressure between 0.05 and 0.22) was used for the
determination of the specific surface area. The pore size distribution
(PSD) was calculated from the adsorption branch of the N2 physisorption
isotherms employing the Barret–Joyner–Halenda (BJH) formula. The cu-
mulative mesopore volume, VBJH, was obtained from the PSD curve.
Scheme 1. n-Alkane hydroisomerisation reaction mechanism.
alysts lies in the selection of a mild acidity (to minimise the
production of cracking products) together with a highly
active metallic function (e.g. Pt, Pd, Ni), which improves the
selectivity in favour of isomerisation. Among the noble
metals, Pt has been shown to considerably increase the se-
lectivity in favour of isomerisation, and is therefore the
most suitable metal for the hydroconversion of n-alkanes.[2]
Various catalysts have been reported to be active in the
hydroisomerisation of C7 and longer C7 chain n-alkanes, in-
cluding materials based on MoO3,[13] Pt-WO3-ZrO2,[4,14–17]
Al-MCM-41,[18,19] zeolite-supported Pt,[20–23] and Pt–silicoalu-
minophosphates (SAPO) materials[8,9,24] as alternatives to
the traditional Pt and Re/Al2O3 catalysts. In particular, com-
pared to microporous zeolites, Al-MCM-41 materials have a
moderate acidity, and the incorporation of metallic sites into
these materials (e.g. Zr, Pt) produces solid catalysts that
have proved to be highly active and selective in the isomeri-
sation of C6 and C7 alkanes.[18,19] The optimum loading of
metal (Pt or Zr) in these materials was found to be close to
0.3 wt%, whereas Campelo et al. reported that a 0.5 wt% Pt
content gave catalysts with optimum activity and selectivity
in favour of isomerisation.[9] We have also recently reported
the preparation of catalytically active Al-MCM-48 solid
acids for use in various reactions.[25] These materials showed
great potential for use as catalysts and supports in a wide
range of applications.
Scanning electron micrographs (SEM) and elemental compositions of the
calcined samples were obtained using a JEOL JSM-6300 scanning micro-
scope with energy-dispersive X-ray analysis (EDX) at 20 kV. Samples
were coated with Au/Pd on a high-resolution SC7640 sputterer at a sput-
tering rate of 1500 V per minute, up to a thickness of 7 nm.
Transmission electron microscopy (TEM) micrographs were recorded on
an FEI Tecnai G2 fitted with a CCD camera for ease and speed of use.
The resolution was around 0.4 nm. Samples were suspended in ethanol
and immediately deposited on a copper grid prior to analysis.
Diffuse reflectance Fourier-transform infrared spectra (DRIFTS) were
recorded on a Bomem MB series instrument equipped with an environ-
mental chamber (Spectra Tech, P/N 0030–100) placed in the diffuse re-
flectance attachment (Spectra Tech, Collector). The environmental cham-
ber allows precise control of the temperature, pressure, and the environ-
ment of the sample, as well as ensuring reproducibility of the experi-
ments. The resolution was 8 cmꢀ1, and 256 scans were averaged to obtain
spectra in the 4000–400 cmꢀ1 range. Samples were dried at 1508C for
24 h, mixed with KBr to 15 wt%, placed in the environmental chamber
under a 20 mLminꢀ1 flow of air, heated to 3008C, and held at this tem-
perature for 1 h prior to measuring the spectra.
Surface acidity was measured in dynamic mode by means of a pulse chro-
matographic technique involving gas-phase (3008C) adsorption of pyri-
dine (PY; sum of Brønsted and Lewis acid sites) and 2,6-dimethylpyri-
dine (DMPY; Brønsted sites) as probe molecules, using a method previ-
ously described elsewhere.[25]
The experimental procedure used for temperature-programmed desorp-
tion of pyridine (PY-TPD) has also been reported previously.[25] The ap-
paratus used for pyridine adsorption/desorption was basically a modified
gas chromatograph. The catalyst (ca. 50 mg) was packed into a stainless
steel tube of length 100 mm and diameter 4 mm, and retained therein be-
tween quartz wool plugs. Pure N2 at a flow rate of 50 mLminꢀ1 was used
as the carrier gas. Prior to the adsorption experiments, the catalyst was
pretreated in situ by passing pure N2 through it at a flow rate of
50 mLminꢀ1 and heating from 50 to 4508C at 108Cminꢀ1 (the tempera-
ture was maintained at 4508C for 10 min). The temperature was then
lowered to 1008C and the adsorption experiment was carried out. Py-
TPD experiments were performed in the 50–7008C range after saturation
of the sample followed by purging in N2 at 1008C for 1 h.
Herein, we report the synthesis of platinum nanoparticles
supported on Al-MCM-48 and their catalytic performance
in the hydroisomerisation of n-octane.
Experimental Section
The preparation of the Al-MCM-48 has been reported previously.[25] The
synthesis of the supported Pt materials involved some different steps.
Firstly, the appropriate amount of tetraamine platinum(II) nitrate for the
Specific platinum surface areas and associated metal dispersions were
measured by H2 chemisorption using a Micromeritics 2700 PulseChemi-
Chem. Eur. J. 2008, 14, 5988 – 5995
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5989

R. Luque et al.
sorb surface area analyser. Samples were pre-reduced at 4008C for 1 h
and then cooled under an Ar stream prior to measurement by the pulse
chemisorption method. Temperature-programmed reduction (TPR) was
conducted on a Stanton-Redcroft STA750 thermal analyser under a
10 vol% H2/He stream at a total flow rate of 20 mLminꢀ1 and with a
ramp rate of 128Cminꢀ1 between room temperature and 8008C.
Catalytic reactions (typically with 0.1 g of catalyst) were carried out in a
continuous-flow system with PC control at 3758C, employing hydrogen
as carrier (25 mLminꢀ1) at 5 bar. The n-octane (50 mLminꢀ1, 0.31 mmol
minꢀ1) was fed into the system by means of a peristaltic pump. Reaction
products were identified and characterised by GC and GC-MS.
Results and Discussion
In general, the X-ray diffraction patterns of the Al-MCM-48
materials with supported platinum exhibited the characteris-
tic lines of MCM-48 materials (Figure 1).[25,27] The diffrac-
tion lines at high 2q angles (Figure 1, inset) can be assigned
to the typical fcc platinum metal by comparison with the
JCPDS card (No. 04-0802). The intensity of the Pt lines is
relatively low due to the extremely high dispersion and low
metal loading. Table 1 summarises the textural properties of
the Pt/Al-MCM-48 materials.
Figure 1. XRD patterns of 0.5%Pt-40Al material. The diffractogram in
the inset features the various diffraction lines of metallic platinum.
Pt/Al-MCM-48 materials display similar textural proper-
ties to those of the parent materials,[25] including high sur-
face areas and pore volumes, as well as pore diameters of
the order of 2 nm. Typical type IV isotherm profiles, charac-
teristic of mesoporous materials, were also obtained for all
of these materials. Elemental analyses of the mesoporous
supported materials showed that their Si/Al ratios remained
almost unaltered compared with those of the parent materi-
als.[25] Therefore, the platinum, which was present in very
low quantities (less than 1%), is believed to be incorporated
into the Al-MCM-48 materials. The interaction of the Pt
with the silicon or aluminum on the surface remains unclear,
although the presence of the Pt at such low loadings seemed
to increase the acidity of the materials (refer to Table 3 and
the PY-DRIFTS data).
TG-DTA and DRIFTS analyses of the materials bearing
supported Pt showed similar profiles to those of the parent
Al-MCM-48 materials.[25] SEM micrographs of Pt/Al-MCM-
48 (Figure 2) were similar to those of the Al-MCM-48 mate-
rials, showing particles of spherical morphology that were
very homogeneously distributed in terms of particle size and
texture.
Figure 2. SEM micrographs of 0.5%Pt-30Al at different magnifications;
a) ”20000 and b) ”30000.
Temperature-programmed reduction confirmed that only
reduced Pt was present in all of the samples, irrespective of
the metal loading or MCM composition. Decomposition of
the amine ligands of the tetraamine platinum(II) nitrate em-
ployed as precursor upon calcination may provide a reduc-
tive environment conducive to the reduction of the metal
nanoparticles and the formation of protons.[28] Specific metal
surface areas allowed the Pt dispersions to be calculated
(Table 2), and these exhibited only a weak dependence on
the Al content. The Pt nanoparticles were found to be well
dispersed at low metal loadings. Increased Pt loadings led to
materials with a relatively low metal dispersion (Table 2, en-
tries 6 and 7).
TEM micrographs of the Pt nanoparticles are shown in
Figure 3. The mesoporous materials exhibited a highly dis-
persed particle distribution with a relatively narrow particle
diameter range (Figure 3), in good agreement with the dis-
persion values obtained (Table 2). It should be noted that
H2 chemisorption measure-
Table 1. Textural properties including d211 spacing, unit cell parameter (ao), surface area (SBET), pore diameter
(DBJH), pore volume (VBJH), and wall thickness (e) of Pt/Al-MCM-48 materials.
ments necessitate a pre-reduc-
tion step at 4008C to remove
any passivating surface oxide;
this could potentially initiate
Pt sintering over weakly inter-
acting supports, leading to an
underestimation of the true
dispersion. However, the good
agreement between particle
size estimates for the high
[a]
Catalyst
d211 [nm] ao [nm] SBET [m2 gꢀ1
]
DBJH [nm] VBJH [mLgꢀ1
]
e [nm] Pt
loading[b]
Actual
Si/Al
[%]
ratio[b]
0.5%Pt-40Al 3.3
0.5%Pt-30Al 3.4
0.5%Pt-20Al 3.4
0.5%Pt-15Al 3.5
8.0
8.2
8.2
8.6
1261
1408
1009
1100
1.9
1.9
2.1
2.1
0.53
0.43
0.75
0.72
1.5
1.6
1.6
1.9
0.43
0.50
0.56
0.53
51
41
26
22
p
[a] Cubic unit cell parameter estimated as ao =d211 6. [b] The atom% of Pt and Si/Al ratio were determined
by elemental analysis (EDX).
5990
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5988 – 5995

Pt Nanoparticles
FULL PAPER
Table 2. Pt dispersion (obtained from H2 chemisorption) and average
nanoparticle sizes obtained from XRD (Scherrer equation; d=0.9l/
Bcosq) and TEM (averaging 50 particles from the TEM micrographs) of
various mesoporous Al-MCM-48-supported platinum materials.
sites in the materials might be present (even at low Pt load-
ings), which might be related to the decomposition of the Pt
amine precursor complex and its interaction with the surface
through the formation of the metal nanoparticles and pro-
tons.[28] On the other hand, the adsorbed aromatic bases, in-
cluding PY and DMPY, might also interact with the Pt spe-
cies through the formation of a complex, thereby increasing
the quantity of base adsorbed on the surface of the materi-
als. The metal–support interaction has also been reported to
influence the acidity of supported Pt materials.[28,30] Interest-
ingly, the platinised materials exhibited similar total (PY
data) and Brønsted (DMPY data) acidities (Table 3), re-
gardless of the Si/Al ratio, with the exception of the
0.5%Pt-20Al sample. Pt-SAPO materials display a much
higher density of acid sites compared to the Pt/Al-MCM-48
catalysts.
Entry Catalyst
Disper- Average particle
Average particle
size (estimated
sion
[%]
size (estimated
from XRD) d [nm] from TEM) [nm]
1
2
3
4
5
6
7
0.5%Pt-40Al
92
81
89
70
77
40
25
4.0
4.3
3.9
4.7
4.8
5.1
5.9
4.4
4.8
4.2
4.4
4.6
5.2
5.9
0.5%Pt-30Al
0.5%Pt-20Al
0.5%Pt-15Al
1%Pt-40Al
3%Pt-40Al
5%Pt-40Al
loading samples from the dispersion values in Table 2 (using
the methodology of ref. [29] and assuming an equal ratio of
(111):
N
N
Table 3. Surface acidity, measured by PY and DMPY titration (mmol
probe molecule per g of catalyst and mmol probe molecule per m2 at
3008C) and Brønsted-Lewis ratio (measured by PY-DRIFTS) of Pt/Al-
MCM-48 materials, compared to the values reported for Pt-SAPO mate-
rials.
suggests that the reported dispersions are accurate.[29] The
extremely high dispersion of the lowest loading (0.5% Pt)
sample indicates a significant number of sub-nanometre
clusters, which fall below the sensitivity of both powder
XRD and the TEM system. The nanoparticle morphology
was found to be spherical, with an average particle diameter
of around 4.4 nm, as calculated from the TEM micrographs.
These results are in good agreement with the particle sizes
estimated from the XRD patterns using the Scherrer equa-
tion (Table 2). An increase in Pt loading from 0.5 to 5 wt%
increased the average nanoparticle size (Table 2) and some
big metal clusters (ca. 10–15 nm) could be observed in the
materials with increased Pt loading (Figure 3B and C).
Catalyst
PY
[mmol
gꢀ1
PY
[mmol
mꢀ2
DMPY
[mmol
gꢀ1
DMPY
[mmol
mꢀ2
B/L ratio
(1508C,
DRIFTS)
G
U
G
ACHTREUNG
]
]
]
]
40Al
0.5%Pt-40Al
30Al
0.5%Pt-30Al
20Al
0.5%Pt-20Al
15Al
0.5%Pt-15Al
0.5%Pt-SAPO5
0.5%Pt-SAPO11
133
399
213
360
241
210
244
397
149
68
0.09
0.32
0.16
0.26
0.17
0.21
0.18
0.36
0.82
0.66
74
119
146
127
99
74
81
126
89
16
0.05
0.09
0.11
0.09
0.07
0.07
0.06
0.11
0.49
0.16
0.9
1.3
1.1
1.4
1.2
1.4
1.3
1.3
–
The catalysts exhibited surface acidities similar to those of
the parent materials,[25] although slightly more acidic com-
pared to Al-MCM-48 (Table 3). The incorporation of Pt
seemed to increase the acidity of the materials. Of note was
the unexpectedly high acidity of the Pt material with an Si/
Al ratio of 40 (0.5%Pt-40Al), which also exhibited the high-
est dispersion (92%, Table 2). Such an increase in acidity
compared to the parent Al-MCM-48 is, however, difficult to
rationalise. A synergistic effect between the Pt and acidic
–
Typical traces from PY-TPD experiments are shown in
Figure 4. The PY desorption trace was deconvoluted using a
Gaussian function with temperature as a variable. Experi-
mental data, the individual components obtained by decon-
volution, and the theoretical spectrum obtained by summing
the individual peaks (standard deviations <6%) for Pt/Al-
MCM-48 materials are compared to those of Al-MCM-48 in
Figure 4 (scaled-up for comparative purposes). Peaks pres-
ent at low temperature (around 210 and 2908C) can be as-
signed to weakly to moderately acidic sites, having a major
contribution to the acidity of the materials.[25] Peaks at
higher temperatures (ca. 460 and 6608C, respectively) can
be attributed to more strongly acidic sites (Brønsted and
Lewis) on the materials. In general, peaks were found to be
slightly shifted to higher temperatures in the supported Pt
materials compared to Al-MCM-48, and only a small de-
crease in peak intensity and contribution (area%) was
found for Pt/Al-MCM-48 materials compared to Al-MCM-
48 (Figure 4).[25]
PY-DRIFTS measurements were also performed to com-
plete the surface acidity studies of the materials. The results
Figure 3. TEM micrographs (”300000, 50 nm scale bar) of Pt/Al-MCM-
48 materials. A) 1%Pt-40Al; B) 3%Pt-40Al; C) 5% Pt-40Al.
Chem. Eur. J. 2008, 14, 5988 – 5995
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5991

R. Luque et al.
Figure 5. DRIFTS of adsorbed pyridine (1700–1400 cmꢀ1
) at 1508C.
A) Al-MCM-48 and 0.5%Pt/Al-MCM-48 materials; a) 40Al, b) 0.5%Pt-
40Al, c) 15Al, and d) 0.5%Pt-15Al; B) 0.5%Pt-40Al at different Py de-
sorption temperatures; a) 1008C, b) 1508C, and c) 2008C.
while the rate-determining step is the rearrangement and/or
scission of the intermediates at the acidic sites.
Very low octane conversion was found at temperatures
below 3008C. Therefore, 3508C was chosen as the optimum
temperature for studying the activity of the Pt nanoparticles.
The typical n-octane conversion values for the Pt/Al-MCM-
48 systems were 50%, comparable to those of Pt-SAPO ma-
terials with similar Pt content (Table 4)[9] and higher than
that of the Al-MCM-48 parent material, in good agreement
with the acidity data (Table 3). The parent Al-MCM-48 ex-
hibited activities between 20 and 30%, and almost complete
selectivity in favour of the cracking products (mainly C2 to
C5). These materials proved to be very efficient in the crack-
ing of isopropylbenzene.[25]
Selectivities in the isomerisation process were higher than
70%, favoring the formation of the 3-methylheptane isomer
over the 2- and 4-methylheptanes (Tables 4 and 5). Interest-
ingly, 2-methylheptane is preferentially obtained with Pt-
SAPO catalysts. The aluminum content did not have a
marked effect on either the conversions or the selectivities,
in good agreement with the titration data (Table 3), which
showed the Lewis and Brønsted acidities to be very similar
in the supported Pt materials regardless of the Si/Al ratio.
Only a slight decrease in conversion was found for the
0.5%Pt-20Al catalyst as compared to the others (Table 4).
The 0.5%Pt-40Al catalyst was unexpectedly active in the
hydroisomerisation reaction on the basis of its high Si/Al
ratio. Nevertheless, the high activity of this material can be
correlated to its high acidity (Table 3, Figure 5). A higher se-
lectivity in favour of the C7-isomers was also found at longer
reaction times; this increased selectivity was achieved at the
expense of the selectivity in favour of cracking. This out-
come is most likely to stem from the decrease in the conver-
sion. The most plausible explanation is that the more strong-
ly acidic sites present in the support (mainly Brønsted), re-
Figure 4. TPD profiles of mesoporous MCM-48 materials: a) 40Al,
b) 0.5%Pt-40Al, c) 20Al, and d) 0.5%Pt-20Al.
are presented in Figure 5 and Table 3. The mesoporous sup-
ported catalysts exhibited several peaks due to pyridine
bound at Lewis sites (1618 and 1453 cmꢀ1), peaks due to
pyridine bound at Brønsted acid sites (1640 and 1543 cmꢀ1),
and a band at 1492 cmꢀ1 attributable to pyridine interacting
with both Lewis and Brønsted acid sites.[31] The desorption
of PY at increasing temperatures results in the removal of
both Brønsted-bound and Lewis-bound PY. The Pt materials
exhibited analogous spectra to that of Al-MCM-48, featur-
ing bands with similar intensities, except in the case of the
0.5%Pt-40Al sample, for which a large increase in the inten-
sity of the DRIFTS bands was found, in good agreement
with the PY titration data. The Brønsted-Lewis ratio in the
platinised materials was found to be very similar between
samples, regardless of the aluminum content (Table 3).
The activity of the nanoparticles was then investigated in
relation to the hydroisomerisation of n-octane. The hydroi-
somerisation of n-alkanes is a very complex catalytic pro-
cess, which involves several parallel, competing, and consec-
utive reactions, including hydrogenations, isomerisations,
and cracking reactions. Therefore, a great multitude of di-
verse products is generated (Scheme 1). The kinetics of
these processes may be subjected to various approximations.
The rate of reaction of the reactant is the key approximation
to take into account, as reported previously.[32] In an ideal
bifunctional catalyst, the metal phase only (de)hydrogenates,
5992
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5988 – 5995

Pt Nanoparticles
FULL PAPER
Table 4. Total conversions (XT, mol%) and selectivities in favour of cracking (SCK, mol%) and isomerisation
to methylheptanes (SMH, mol%), including 2-, 3-, and 4-methylheptanes, of different materials in the hydroiso-
merisation of n-octane.[a]
PCP mechanism, 3-methylhep-
tane will be formed in prefer-
ence to the 2-methyl isomer, in
good agreement with the re-
Catalyst
4 h reaction
SCK
12 h reaction
SCK
20 h reaction
SCK
XT
SMH
XT
SMH
XT
SMH
ported
experimental
re-
0.5%Pt-40Al
0.5%Pt-30Al
0.5%Pt-20Al
0.5%Pt-15Al
15Al
50
52
44
54
28
50
28
18
28
17
26
>95
79
75
69
75
70
–
48
49
36
52
25
45
25
17
27
14
21
>95
76
77
70
80
72
–
48
48
35
51
24
41
21
18
19
14
15
>95
65
79
73
81
80
–
sults.[38,39] We can therefore
conclude that the skeletal iso-
merisation of the intermediate
n-alkyl carbocations is likely to
proceed according to the PCP
mechanism as this would ac-
count for the 1:2:1 relative
ratio of the 2-, 3-, and 4-meth-
ylheptanes.
0.5%Pt-SAPO5
0.5%Pt-SAPO11
21
59
24
64
35
74
33
29
26
[a] T=3758C, p=5 bar, 0.1 g catalyst, WHSV=12.7 hꢀ1; the fact that the selectivities do not total 100 is due to
toluene (ethylbenzene was also obtained in small quantities).
sponsible for the formation of cracking products, are steadi-
ly deactivated with time due to the strong adsorption of car-
bonaceous species and/or the generated products (e.g.,
arenes and cyclohexane) on the catalyst surface.[33,34] Moder-
ately and weakly acidic sites, responsible for the isomerisa-
tion phenomenon, are less prone to such deactivation, so an
increase in the formation of C7-isomers is observed with a
concomitant decrease in the formation of cracking products.
Propane, butane, pentane, and hexane were obtained as
cracking products (mainly C2 to C4; Table 5). The relatively
low acidity of Al-MCM-48 materials accounts for the fact
that the formation of such products, which usually takes
place at the strongly acidic centres of the support, is not
prominent under milder reaction conditions. Increasing con-
centrations of cracking products start to appear at higher
temperatures, in good agreement with the classical consecu-
tive hydroisomerisation mechanism.[35] No multibranched
products were detected. Of note was the 1:2:1 relative ratio
of the 2-, 3-, and 4-methylheptanes (Tables 5 and 6). The
quantity of 3-methylheptane in the reaction product mixture
thus corresponds to the sum of the quantities of the 2- and
4-isomers produced. This ratio was found to be constant and
consistent for all of the Pt/Al-MCM-48 materials, regardless
of the aluminum content or the reaction time.
We then proceeded to investigate the effects of various
parameters, including the hydrogen pressure, temperature,
and Pt loading, on the performance of Pt/Al-MCM-48 in the
hydroisomerisation of n-octane.
The hydrogen pressure in the continuous-flow reactor was
found to be a critical parameter in determining the activity
and selectivity of the Pt nanoparticles. The activities of Pt/
Al-MCM-48 were investigated at four different hydrogen
partial pressures within the reactor (1, 3, 5, and 7 bar, re-
spectively, at 3758C; 20 h reaction). The results are summar-
ized in Table 6.
In general, an increase in the H2 pressure in the systems
increased the total conversion of starting material (the con-
version was doubled on going from 1 to 5 bar), regardless of
the aluminum content. Contrary to the observed trends,
Chaudhari et al. reported a decrease in activity with increas-
ing hydrogen partial pressures, as a result of rapid hydroge-
nation of the intermediate alkenes and carbocations, thereby
preventing further reactions.[19] Nevertheless, we believe that
an increase in the partial pressure of hydrogen activates the
metallic sites on the catalyst surface, improving their effi-
ciency in dehydrogenating the n-alkanes to n-alkenes (first
step) and hydrogenating the iso-alkenes to the final iso-alka-
nes, therefore providing enhanced reaction rates and im-
proved C8 conversions. The increased reaction rates also af-
fected the selectivity in favour of cracking products. The for-
mation of such products (including n-propane, butane, and
hexane) did not increase significantly on increasing the H2
pressure up to 4–5 bar. A further increase in the H2 pressure
(>5 bar) led to a higher selectivity in favour of cracking
products, especially C2 to C4
The most important step in the hydroconversion of n-al-
kanes is the rearrangement of the intermediate alkyl carbo-
cations (Scheme 1). The skeletal isomerisation of such car-
benium ions was originally proposed to involve propanated
cyclopropane (PCP) intermediates,[36,37] and this is generally
accepted for alkanes larger than n-butane. According to the
products. 3-Methylheptane was
still preferentially produced in
the reaction compared to 2-
and 4-methylheptanes.
Table 5. Product distribution and selectivities in favour of cracking (SCK, mol%), including selectivities for C2
to C4 (SC2–4, mol%), pentane (SC5, mol%), hexane (SC6, mol%), and isomerisation to methylheptanes (SMH
,
mol%), including 2-, 3-, and 4-methylheptanes (SXMH, mol%), of supported platinum materials in the hydro-
isomerisation of n-octane.[a]
The deactivation of the ma-
terials at atmospheric pressure
(1 bar) was notable. This may
be attributed to the formation
of carbonaceous species on the
catalyst surface, which contrib-
ute to the poisoning and/or
blocking of the active centres
Catalyst
SCK
SC5
SMH
S3MH
SOTHERS
SETBZ
SC2–4
SC6
S2MH
S4MH
STOL
0.5%Pt-40Al
0.5%Pt-30Al
0.5%Pt-20Al
0.5%Pt-15Al
0.5%Pt-SAPO11
10
18
7
15
22
5
8
7
8
10
3
2
3
3
1
22
19
22
20
28
37
33
39
35
18
15
13
14
15
13
6
3
5
4
7
1
–
3
–
1
[a] T=3758C; p=5 bar, 0.1 g catalyst, WHSV=12.7 hꢀ1; 4 h reaction.
Chem. Eur. J. 2008, 14, 5988 – 5995
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5993

R. Luque et al.
Table 6. Effect of the hydrogen pressure on the total conversion (XT, mol%) and selectivities in favour of cracking (SCK, mol%) and isomerisation to 3-
methylheptane (S3MH, mol%) and 2- and 4-methylheptanes (S2+4MH, mol%) of 3%Pt-40Al in the hydroisomerisation of n-octane.[a]
PH [bar]
4 h reaction
S3MH
12 h reaction
S3MH
20 h reaction
S3MH
2
XT
SCK
S2+4MH
XT
SCK
S2+4MH
XT
SCK
S2+4MH
1
3
5
7
38
56
70
78
29
25
42
52
31
34
26
21
31
33
25
22
36
55
67
75
26
22
39
43
33
35
27
26
33
35
27
25
34
53
64
72
23
24
35
35
32
35
29
30
32
33
29
29
[a] T=3758C, 0.1 g catalyst, WHSV=12.7 hꢀ1; the fact that the selectivities do not total 100 is due to toluene (traces of ethylbenzene were also ob-
tained).
(Table 6). However, the stabilities of the supported Pt cata-
lysts generally increase with increasing H2 pressure in the re-
actor (Table 6).[8] In any case, the materials proved to be
highly stable and the deactivation was almost negligible
(<10% of the initial catalytic activity) after 20 h of reac-
tion.
higher concentration of large Pt clusters in the materials
(Figure 3C).
Conclusion
The temperature also had an important effect on the con-
version and selectivity in these systems, increasing both the
n-octane conversion and the selectivity in favour of cracking
(Table 7), in good agreement with previously reported re-
sults.[8,9,18] These increases were only significant at H2 pres-
sures higher than 3 bar. 3-Methylheptane was obtained as
the main reaction product, with a selectivity of around 30%
at temperatures below 4008C. The platinum nanoparticles
were found to be highly stable, even at high reaction tem-
peratures (4008C and over).
The Pt loading was another key parameter influencing the
activity of the supported Pt materials (Table 8). An increase
in the Pt loading (from 0.5 to 4 wt%) provided higher cata-
lytic activities, but also led to the formation of increasing
quantities of cracking products (mainly C2 to C4). This in-
crease in selectivity in favour of short-chain alkanes was
more significant at higher hydrogen pressures in the system
(>5 bar). Pt loadings over 4 wt% gave reduced activities, in-
dicating a less well dispersed and heterogeneous particle dis-
tribution (with metal dispersions below 40%) as well as a
Highly active platinum nanoparticles dispersed on Al-
MCM-48 have been prepared. The supported materials dis-
played a very narrow particle size distribution. The materi-
als exhibited high activities and high selectivities in favour
of C7-branched isomers in the hydroisomerisation of n-
octane. The production of 3-methylheptane was found to be
favoured with respect to the 2- and 4-isomers, implying a
shape-selectivity effect in the mesoporous supported Pt cata-
lysts. The catalysts proved to be highly stable at high tem-
peratures and high hydrogen pressures. They therefore offer
a potentially interesting alternative to the traditional Pt, Pd,
and Re/Al2O3 supported catalysts.
Acknowledgements
This research was subsidized by grants from the Dirección General de In-
vestigación (Project CTQ2005-04080 and CTQ2007-65754/PPQ), the
Ministerio de Educación y Ciencia, FEDER Funds, and the Consejería
de Innovación, Ciencia y Empresa (Junta de Andalucia, Project FQM-
191).
Table 7. Effect of the temperature on the total conversion (XT, mol%) and selectivities in favour of cracking (SCK, mol%) and isomerisation to 3-methyl-
heptane (S3MH, mol%) and 2- and 4-methylheptanes (S2+4MH, mol%) of 3%Pt-40Al in the hydroisomerisation of n-octane.[a]
T [8C]
4 h reaction
S3MH
12 h reaction
S3MH
20 h reaction
S3MH
XT
SCK
S2+4MH
XT
SCK
S2+4MH
XT
SCK
S2+4MH
350
375
400
28
70
82
22
42
60
34
26
15
35
25
16
27
67
78
17
39
57
38
27
16
37
26
20
23
64
75
15
35
52
39
29
17
42
28
25
[a] p=5 bar; 0.1 g catalyst, WHSV=12.7 hꢀ1; the fact that the selectivities do not total 100 is due to toluene (traces of ethylbenzene were also obtained).
Table 8. Effect of the Pt loading of the material on the total conversion (XT, mol%) and selectivities in favour of cracking (SCK, mol%) and isomerisa-
tion to 3-methylheptane (S3MH, mol%) and 2- and 4-methylheptanes (S2+4MH, mol%) of Pt-40Al materials in the hydroisomerisation of n-octane.[a]
Pt (%)
4 h reaction
S3MH
12 h reaction
S3MH
20 h reaction
S3MH
XT
SCK
S2+4MH
XT
SCK
S2+4MH
XT
SCK
S2+4MH
0.5
1
3
50
54
70
61
18
32
42
37
37
31
26
29
38
30
25
26
48
50
67
57
17
32
39
35
39
31
27
28
39
30
26
29
48
49
64
55
18
20
35
36
39
38
32
29
38
37
30
27
5
[a] T=3758C, p=5 bar, 0.1 g catalyst; WHSV=12.7 hꢀ1; the fact that the selectivities do not total 100 is due to toluene (traces of ethylbenzene were also
obtained).
5994
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 5988 – 5995

Pt Nanoparticles
FULL PAPER
[22] R. Roldan, F. J. Romero, C. JimØnez-Sanchidrian, J. M. Marinas, J. P.
Gomez, Appl. Catal. A 2005, 288, 104–115.
[23] P. N. Kuznetsov, J. Catal. 2003, 218, 12–23.
[24] Z. Wang, Z. Tian, F. Teng, G. Wen, Y. Xu, Z. Xu, L. Lin, Catal. Lett.
2005, 103, 109–116.
[25] J. M. Campelo, D. Luna, R. Luque, J. M. Marinas, A. A. Romero,
J. J. Calvino, M. P. Rodríguez-Luque, J. Catal. 2005, 230, 327–338.
[26] V. L. Budarin, J. H. Clark, R. Luque, K. Milkowski, D. J. Macquar-
rie, R. J. White, PCT Patent, WO 2007104798 A2 20070920, 2007.
[27] J. C. Vartuli, K. D. Schmitt, C. T. Kresge, W. J. Roth, M. E. Leono-
wicz, S. B. McCullen, S. D. Hellring, J. S. Beck, J. L. Schlenker, D. H.
Olson, E. W. Sheppard, Chem. Mater. 1994, 6, 2317–2326.
[28] M. Lashdaf, V. V. Nieminen, M. Tiita, T. Venäläinen, H. Österholm,
O. Krause, Microporous Mesoporous Mater. 2004, 75, 149–158.
[29] P. Gallezot, G. Bergeret, Handbook of Heterogeneous Catalysis, Vol.
2 (Eds.: G. Ertl, H. Knçzinger, J. Weitkamp), Wiley-VCH, Wein-
heim, 1997, 441.
[1] N. A. Cusher, P. Greenouch, J. P. K. Rolfe, J. A. Weiszmann, in
Handbook of Petroleum Refining Processes (Ed.: R. A. Meyers),
McGraw-Hill, New York, 1986.
[2] V. M. Akhmedov, S. H. Al-Khowaiter, Catal. Rev. 2007, 49, 33–109.
[3] Z. B. Zhang, A. Kamo, T. Yonedo, T. Komatsu, T. Yashima, Appl.
Catal. A 1997, 159, 119–132.
[4] E. Iglesia, D. G. Barton, S. L. Soled, S. Miseo, J. E. Baumgartner,
W. E. Gates, G. A. Fuentes, G. D. Meitzner, Stud. Surf. Sci. Catal.
1996, 101, 533–540.
[5] S. I. Miller, US Patent 5135638, 1992.
[6] Y. Liu, Y. Guan, C. Li, J. Lian, G. J. Gan, E. C. Lim, F. Kooli, J.
Catal. 2006, 244, 17–23.
[7] P. Meriaudeau, V. A. Tuan, V. T. Nghiem, S. Y. Lai, L. N. Hung, C.
Naccache, J. Catal. 1997, 169, 55–66.
[8] J. M. Campelo, F. Lafont, J. M. Marinas, J. Chem. Soc. Faraday
Trans. 1995, 91, 1551–1555.
[9] J. M. Campelo, F. Lafont, J. M. Marinas, J. Chem. Soc. Faraday
Trans. 1995, 91, 4171–4177.
[30] G. Xiong, N. Yao, W. Han, S. Sheng, W. Yang, Cuihua Xuebao 2005,
26, 759–762.
[10] F. Alvarez, F. R. Ribeiro, G. Perot, C. Thomazeau, M. Guisnet, J.
Catal. 1996, 162, 179–189.
[11] L. B. Galperin, Appl. Catal. A 2001, 209, 257–268.
[12] K. C. Park, S. K. Ihm, Appl. Catal. A 2000, 203, 201–209.
[13] F. C. Meunier, F. Cavallaro, T. Le Goaziou, A. Goguet, C. Rioche,
Catal. Today 2006, 112, 64–67.
[14] D. G. Barton, S. L. Soled, G. D. Meitzner, G. A. Fuentes, E. Iglesia,
J. Catal. 1999, 181, 57–72.
[15] E. Iglesia, S. L. Soled, G. M. Kramer, J. Catal. 1993, 144, 238–253.
[16] P. Perez-Romo, C. Potvin, J. M. Manoli, G. Djega-Mariadassou, J.
Catal. 2002, 205, 191–198.
[31] F. M. Bautista, J. M. Campelo, A. Garcia, D. Luna, J. M. Marinas,
A. A. Romero, J. A. Navio, M. Macias, J. Catal. 1994, 145, 107–125.
[32] R. Beecher, A. Voorhies, P. Eberly, Ind. Eng. Chem. Prod. Res. Dev.
1968, 7, 203–209.
[33] A. Chica, A. Corma, J. Catal. 1999, 187, 167–176.
[34] J. F. Kriz, T. D. Pope, M. Stanciulescu, J. Monnier, Ind. Eng. Chem.
Res. 1998, 37, 4560–4569.
[35] P. Sanchez, F. Dorado, M. J. Ramos, R. Romero, V. JimØnez, J. L.
Valverde, Appl. Catal. A 2006, 314, 248–255.
[36] F. E. Condon, Stepwise reaction on separate catalytic centers. Catal-
ysis , Vol. 6 (Ed.: P. H. Emmet), Reinhold, New York, pp. 121.
[37] D. M. Brouwer, H. Hogeveen, Prog. Phys. Org. Chem. 1972, 9, 179–
184.
[17] J. Xu, J. Y. Ying, Angew. Chem. 2006, 118, 6852–6856; Angew.
Chem. Int. Ed. 2006, 45, 6700–6704.
[18] I. Eswaramoorthi, V. Sundaramurthy, N. Lingappan, Microporous
Mesoporous Mater. 2004, 71, 109–115.
[19] K. Chaudhari, T. K. Das, A. J. Chandwadkar, S. Sivasanker, J. Catal.
1999, 186, 81–90.
[38] J. Weitkamp, Ind. Eng. Chem. Prod. Res. Dev. 1982, 21, 550–558.
[39] A. de Lucas, J. L. Valverde, P. Sanchez, F. Dorado, M. J. Ramos,
Appl. Catal. A 2005, 282, 15–24.
[20] A. Hollo, J. Hancsok, D. Kallo, Appl. Catal. A 2002, 229, 93–102.
[21] S. Gopal, P. G. Smirniotis, Appl. Catal. A 2003, 247, 113–123.
Received: January 29, 2008
Published online: May 15, 2008
Chem. Eur. J. 2008, 14, 5988 – 5995
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
5995