Aluminum fluoride-supported platinum and palladium as highly efficient catalysts of n-pentane hydroisomerization

Article abstract of DOI:10.1002/cctc.201300840

Catalytic isomerization of n-pentane was studied over novel high surface area (HS) aluminum fluoride supported palladium and platinum catalysts, prepared by the nonaqueous fluorolytic sol-gel synthesis. HS-AlF₃ maintained well its nanoscopic, mesoporous, and highly acidic character at higher temperatures, up to 350 °C. Both Pt/HS-AlF₃ and Pd/HS-AlF ₃ catalysts, characterized by very small metal particle sizes (≈2 nm) maintained very good activity, stability, and selectivity towards isomerization (up to nearly 100 %) at up to 350 °C. Very strong acidity of HS-AlF₃ is the prerequisite for isomerization but the presence of the metal is also essential to maintain a high conversion level. The effect of metal-acid balance in Pd/HS-AlF₃ catalysts, important for bifunctional isomerization, was tested by screening three differently metal-loaded catalysts. The performance of the 0.5 wt % Pd/HS-AlF₃ appeared as good as that of 2.0 wt % Pd. The properties of the HS-AlF ₃ support, catalyst precursors, and catalysts were characterized by XRD, TEM, H₂ chemisorption, and FTIR photoacustic spectroscopy of adsorbed pyridine, and NH₃ temperature programmed desorption. The first three methods allowed evaluating the state of the metal in the reduced catalysts. The suppressed chemisorption of H₂ must result from different factors, such as carbon presence in Pd introduced during catalyst synthesis and activation, metal-acidic support interaction, and, partially, metal encapsulation in AlF₃ during sol-gel synthesis. Lewis and Bronsted acidity was detected in activated catalysts and both types of acid sites were assumed to play a role in bifunctional isomerization. NH ₃ effectively blocks the strongest acid sites and drastically limits the overall catalytic performance. Introduction of water to the reaction system also degrades, but to a lesser extent, the Pd(Pt)/HS-AlF₃ catalyst performance, most probably because of poisoning of only the strongest Lewis acid sites. A week-long performance: High surface area(HS) aluminum fluoride-supported palladium and platinum catalysts are used in the isomerization of n-pentane. The activity, stability, and selectivity of Pt/HS-AlF₃ and Pd/HS-AlF₃ catalysts is maintained at temperatures of up to 350 °C over reaction times of up to a week. Both Lewis and Bronsted acidity is detected in the activated catalysts, evidencing bifunctional isomerization. Copyright

Full text of DOI:10.1002/cctc.201300840

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300840
Aluminum Fluoride-Supported Platinum and Palladium as
Highly Efficient Catalysts of n-Pentane Hydroisomerization
[a]
[b]
[a, c]
Ostap Machynskyy, Erhard Kemnitz,* and Zbigniew Karpi n´ ski*
Catalytic isomerization of n-pentane was studied over novel
high surface area (HS) aluminum fluoride supported palladium
and platinum catalysts, prepared by the nonaqueous fluorolyt-
ic sol–gel synthesis. HS-AlF3 maintained well its nanoscopic,
mesoporous, and highly acidic character at higher tempera-
tures, up to 3508C. Both Pt/HS-AlF and Pd/HS-AlF catalysts,
precursors, and catalysts were characterized by XRD, TEM, H2
chemisorption, and FTIR photoacustic spectroscopy of ad-
sorbed pyridine, and NH temperature programmed desorp-
3
tion. The first three methods allowed evaluating the state of
the metal in the reduced catalysts. The suppressed chemisorp-
tion of H must result from different factors, such as carbon
3
3
2
characterized by very small metal particle sizes (ꢀ2 nm) main-
tained very good activity, stability, and selectivity towards iso-
merization (up to nearly 100%) at up to 3508C. Very strong
presence in Pd introduced during catalyst synthesis and activa-
tion, metal–acidic support interaction, and, partially, metal en-
capsulation in AlF during sol–gel synthesis. Lewis and Brønst-
3
acidity of HS-AlF is the prerequisite for isomerization but the
ed acidity was detected in activated catalysts and both types
of acid sites were assumed to play a role in bifunctional iso-
3
presence of the metal is also essential to maintain a high con-
version level. The effect of metal–acid balance in Pd/HS-AlF3
catalysts, important for bifunctional isomerization, was tested
by screening three differently metal-loaded catalysts. The per-
merization. NH effectively blocks the strongest acid sites and
3
drastically limits the overall catalytic performance. Introduction
of water to the reaction system also degrades, but to a lesser
formance of the 0.5 wt% Pd/HS-AlF appeared as good as that
extent, the Pd(Pt)/HS-AlF catalyst performance, most probably
3
3
of 2.0 wt% Pd. The properties of the HS-AlF support, catalyst
because of poisoning of only the strongest Lewis acid sites.
3
Introduction
As the current worldwide environmental concerns have intro-
duced regulations to reduce the level of aromatic compounds
issue. Recently we reported preliminary encouraging results of
catalytic studies on n-pentane isomerization over high surface
[1]
(benzene, in particular) in petrol according to WorldWide Fuel
area (HS)-AlF -supported palladium and platinum. Those stud-
3
Charter, isomerization of light normal paraffins to branched
paraffins successfully satisfies the demand for high octane mol-
ecules, and is, therefore, increasingly becoming a major option
for enhancing the research octane number/motor octane
number values. As normal pentane is one of the major compo-
nents of natural gasoline (C –C petroleum fraction), it should
ies were motivated by recent knowledge concerning prepara-
[2–4]
tion of nanoscopic aluminum fluoride
characterized by
highly specific surfaces, and, remarkably, by outstanding high
[5,6]
surface acidity.
These interesting properties suggest poten-
tial applications of HS-AlF in bifunctionally catalyzed reactions,
3
such as hydroisomerization of n-alkanes. Strong acidity of alu-
minum fluoride has already been combined with hydrogena-
tive properties of gold to create an example of bifunctional
5
7
necessarily be upgraded by converting it into isopentane.This
process is performed in industry using different types of cata-
lysts; however, call for yet more effective catalysts is still a live
[7,8]
catalysts in the reaction of menthol synthesis.
Our recent report revealed very good performances of HS-
AlF -supported platinum and palladium catalysts in n-pentane
3
[
a] M. Sc. O. Machynskyy, Prof. Z. Karpi n´ ski
Department of Catalysis on Metals, Institute of Physical Chemistry of PAS
ul. Kasprzaka 44/52, 01224 Warszawa (Poland)
Fax: (+48)223433333
[1]
hydroisomerization. In spite of its very strong acidity, metal-
free HS-AlF does not appear to be a very efficient isomeriza-
3
tion catalyst (Figure 4 in Ref. [1]). Likewise, platinum supported
E-mail: zkarpinski@ichf.edu.pl
on less acidic HS-MgF exhibits only moderate isomerization
2
[
b] Prof. E. Kemnitz
activity. Therefore, the superior isomerization performance of
Institut fꢀr Chemie, Humboldt-Universitꢁt zu Berlin
Brook-Taylor-Straße 2, 12489 Berlin (Germany)
Fax: (+49)03020937277
Pt(Pd)/HS-AlF catalysts have to be related to their bifunctional
3
character, that is, the appropriate balance between metal and
[9]
E-mail: erhard.kemnitz@chemie.hu-berlin.de
acid sites is essential for catalysis.
[
c] Prof. Z. Karpi n´ ski
Herein, we report results of a more detailed study in which
the impact of different parameters such as different catalyst
pretreatments on n-pentane conversion, isopentane selectivity,
and catalysts stability were investigated. Long-term kinetic
The Faculty of Mathematics and Natural Sciences - School of Sciences
Cardinal Stefan Wyszynski University in Warsaw
ul. Wꢂycickiego 1/3, 01938 Warszawa (Poland)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300840.
runs served for testing practical application of HS-AlF -support-
3
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 592

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ed metal catalysts. To gain insight into a catalytic action of
acid sites of the support, the effects of ammonia and water
vapor were also studied. Obtained kinetic parameters (reaction
orders, apparent activation energies) were used for a short dis-
cussion of the probable mechanism of n-alkane isomerization
F
3
,unreduced Pt/HS-AlF
, and Pd/HS-AlF ) were investigated after
3 3
postfluorination by CHClF /N at 2508C for 5 h. Additionally, prior-
to-reaction and postreaction samples of Pt/HS-AlF and Pd/HS-AlF
were also tested. The method of sample pretreatment for FTIR-PAS
studies was described earlier. The catalyst samples were pretreat-
2
2
3
3
[12]
ꢁ1
ed under nitrogen flow (35 mLmin ) at 1508C. After 15 min, pyri-
on HS-AlF supported metals.
3
dine volumes of approximately 30 mL were added to the flow for
adsorption. Then, the samples were flushed at the same tempera-
ture during the next 15 min with nitrogen to remove physisorbed
pyridine. Then the FTIR spectra of samples with pyridine were re-
corded with a MTEC 200 photoacoustic cell (FTIR system 2000) Per-
kinElmer device at RT at wavenumbers between 4000 and
400 cm . The intensities of bands at 1454 cm (L-Py), 1490 cm
(L-Py and B-Py) and 1545 cm (B-Py) were used for identifying the
acidity of the catalysts.
Experimental Section
Synthesis and standard activation of Pd(Pt)/HS-AlF catalysts
3
ꢁ1
ꢁ1
ꢁ1
Synthesis of Pd/HS-AlF and Pt/HS-AlF was performed by using the
3
3
ꢁ1
nonaqueous fluorolytic sol–gel synthesis in dehydrated solvents,
[1]
[10,11]
described previously, based on reported procedures.
Briefly,
the synthesis proceeded in three steps. First, the fluorolysis of alu-
minum alkoxide in dehydrated THF was performed by addition of
Temperature-programmed desorption (TPD) of ammonia (lecture
bottle of Laborgas from Merck-Schuchardt, purity 99.96%) was em-
ployed for the determination of acidic sites, as described in
Ref. [13]. All stages of this experiment were: saturation with NH3
anhydrous HF followed immediately by addition of Pd(Pt)(acac) to
2
the formed sol, resulting in the preparation of 2.0 wt% metal-
loaded catalysts. In addition, three Pd/HS-AlF catalysts of different
3
(
at 1208C), release of weakly bound NH at 1208C, and TPD of NH
3
3
metal loadings (0.2, 0.5, and 2.0 wt%) were prepared by incipient
wetness impregnation using the acetonic solution of palladium
was monitored by mass spectrometry (M200 from Dycor-Ametek,
Pittsburgh, PA). After saturation with NH , the flow of NH was
3
3
chloride and preprepared AlF3 (C H O ) gel. Then, after solvents
ꢁx
3
7
3 x
II
stopped and after 0.5 h purge in He at 1208C, when the mass sig-
removal, the gas-phase postfluorination of the Pd(Pt) /AlF3
-
ꢁx
nals from NH returned to a negligible level, the temperature was
3
(
C H O ) was accomplished with CHClF . Finally, H reduction was
3 7 3 x 2 2
ꢁ1
increased at 108Cmin up to 6008C. For further presentation the
evolution of m/z 16 was selected because the release of m/z 17
would also result from water.
performed to obtain zerovalent metallic Pd(Pt) species. Standard
activation of catalysts involving postfluorination with CHClF2 (at
2
008C) and reduction (at 3508C, 1 h) are described in Ref. [1]. Dif-
ferent activation conditions were also used to check their effect on
the behavior of these materials and are indicated in the Results
section.
Catalytic conversion of n-pentane
The reaction of n-pentane (Aldrich, 99+, anhydrous, practical
purity 99.75%, as checked by GC) with hydrogen (purity 99.999%)
was performed in a tubular fused silica reactor equipped with a frit-
ted disk under atmospheric pressure, in a continuous-flow reaction
N adsorption–desorption studies
2
Measurements of surface areas and pore-size distribution were per-
formed by using an ASAP 2020 instrument from Micromeritics.
These studies are based on the BET and Barrett–Joyner–Halenda
methods, using nitrogen as adsorbate at ꢁ1968C. Prior to the mea-
surement catalyst samples were degassed at 200–3308C (depend-
ing on the maximum temperature used in activation) for 2 h in
vacuum.
[14]
system, as described elsewhere. The n-alkane was provided from
a saturator maintained at ꢁ21.88C to adjust its partial pressure at
8
.18 kPa. The partial-pressure ratio of hydrogen to hydrocarbon
was close to approximately 11.5. The reactor effluent was analyzed
by using a gas chromatograph (HP 5890 Series II with FID)
equipped with a 6 m squalane/Chromosorb P column and Chem-
Station. Product selectivities were calculated as the carbon per-
centage of n-pentane consumed in the formation of a given
product.
XRD, TEM, and hydrogen chemisorption
Special kinetic experiments were performed at lower overall con-
version of n-pentane. Results of such tests served for assessing ki-
netic parameters of this reaction. In addition, the effect of blocking
of catalyst’s active sites by ammonia and water was examined.
Longer catalytic tests at different reaction conditions (temperature,
Reduced and postreaction 2.0 wt% metal-loaded catalyst samples
were investigated by XRD (Siemens D5000, Ni-filtered CuKa radia-
tion) and high resolution (HR) TEM. HRTEM studies were performed
on a FEI Titan Cubed at 300 kV accelerating voltage, at the Institute
of Physics of PAS. For TEM, the samples were prepared by mechan-
ical grinding and dispersed in alcohol by using an ultrasonic gener-
ator. A drop of this suspension was placed on a Cu mesh covered
with carbon film and purified with plasma cleaner.
H /n-pentane ratio) and attempts to regenerate the catalysts were
2
also undertaken.
Hydrogen chemisorption on activated (i.e., postfluorinated and re-
duced) catalysts was performed in a pulse-flow fashion, permitting
determination of irreversibly bonded hydrogen.
Results
Catalyst characterization
The dependence of surface area and porosity of Pd/HS-AlF3
Acidity measurements: FTIR photoacoustic spectroscopy
and temperature-programmed desorption of ammonia
and Pt/HS-AlF on thermal pretreatment is shown in Table 1. It
3
is evident that both sol–gel prepared catalysts are sensitive to
high-temperature pretreatment, which results in a vast de-
crease in surface area and porosity, especially for samples pre-
treated at >3508C. Samples pretreated at 4008C have surface
The acidity of catalyst is represented by Lewis sites (L) and Brønst-
ed sites (B), which were indicated by FTIR photoacoustic spectros-
copy (FTIR–PAS) after pyridine adsorption. Three samples (HS-Al-
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 593

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Considering a relatively substantial metal loading (2.0 wt%),
the apparent absence of metal reflections in the sol–gel pre-
pared catalysts must be indicative of their very small size, typi-
cally ꢂ2 nm. Hydrogen chemisorption experiments did not
confirm such high metal dispersion (Table 2) suggesting that
the metal surface must be either partially blocked by organic
residues originating from the precursor or postfluorinationg
Table 1. BET specific surface areas and size of mesopores (dpore) of sol–gel
prepared HS-AlF supported palladium and platinum catalysts.
3
Catalyst
Pretreatment temperature
8C]
Surface area
d
pore
[nm]
2
ꢁ1
[
[m g
]
2
.0 wt% Pd/HS-AlF
3
200
261.3
216.6
182.8
167.0
86.9
<5
5
8
2
3
3
4
6
50
00
50
00
50
agent (CHClF ), or a considerable part of metal (introduced
9
2
23
35
6
8
8
during the sol–gel synthesis) is submerged in the support
matrix, inaccessible to chemisorption.
22.2
2.0 wt% Pt/HS-AlF
3
250
198.8
166.8
155.2
87.4
3
3
4
6
00
50
00
50
22
38
3
Table 2. Hydrogen chemisorption data of HS-AlF supported Pt and Pd
catalysts used in this work.
12.9
[
a]
Catalyst
H/Pt(Pd)
2
2
.0 wt% Pt/HS-AlF3 (sol–gel)
.0 wt% Pd/HS-AlF3 (sol–gel)
0.055
0.022
0.017
0.024
0.003
areas almost two times lower than those pretreated at 3508C.
Such effect is almost certainly detrimental for surface acidity of
nanoscopic aluminum fluoride.
2.0 wt% Pd/HS-AlF3 (impregnated)
0.5 wt% Pd/HS-AlF
.2 wt% Pd/HS-AlF
3
(impregnated)
(impregnated)
0
3
XRD studies of HS-AlF supported Pd and Pt supported sam-
[a] Irreversible uptake at 258C (for Pt) and 708C (for Pd).
3
ples after their sol–gel preparation, fluorination, and reduction
did not reveal any reflections from both crystalline phases of
aluminum fluorides and fcc phases of deposited noble metals.
The amorphous-like structure of the precious metals was also
maintained in the catalyst subjected to n-pentane conversion
performed at 3508C overnight. Similarly featureless XRD pro-
TEM studies of sol–gel-prepared 2.0 wt% Pd/HS-AlF3 and
2.0 wt% Pt/HS-AlF3 revealed the presence of very small
(ꢀ2 nm or less) metal particles differently distributed in alumi-
[1]
num fluoride, confirming our interpretation of the XRD pro-
files. The images in Figure 2 reveal that such high metal disper-
sion is maintained in the postreaction samples.
files were obtained for impregnated Pd/HS-AlF catalyst with
3
0
2
.2 and 0.5 wt% metal loading (Figure 1). The impregnated
.0 wt% Pd/HS-AlF exhibited very diffuse rather asymmetric
The distances of the {111} lattice planes (d111 =0.22 nm) for
metallic Pd (Figure 2, right top) and of the {111} lattice planes
3
reflections from the metal phase, which will be discussed later.
(
(
d111 =0.23 nm) for metallic Pt
right bottom) were calculated
from the respective lattice
fringes.
FTIR photoacoustic spectros-
copy (FTIR–PAS) employed for
characterizing the nature of acid
sites in reduced and postreac-
tion samples of Pt(Pd)/HS-AlF3
by using pyridine as a basic
probe molecule confirmed the
presence of strong acid sites. In
Figure 3 the vibration bands at
1
451 (u 9b), 1490 (u19a), and
1
ꢁ
1
1
610 cm (u ) are shown, char-
8a
acteristic for pyridine coordinat-
ed on Lewis acid sites (L-Py), and
at 1490 (u19b), 1540 (u19a), and
1
640 (u ) characteristic of
8a
Brønsted acid sites (B-Py). FTIR–
PAS spectra confirm that the in-
corporation of Pt(Pd) does not
influence the acidic character of
Figure 1. XRD patterns of variously loaded Pd/HS-AlF
3
catalysts (impregnation) after reduction at 3508C. Inset:
. A de-
[11]
HS-AlF3. It is obvious that all
XRD profile of 2.0 wt% Pd/HS-AlF catalyst after subtraction of the background from 0.2 wt% Pd/HS-AlF
3
3
presented samples, that is, HS-
tectable asymmetry of the (111) and (200) reflections is ascribed to a partial dissolution of carbon in Pd bulk (see
text).
AlF3 and unreduced, reduced,
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 594

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Time-on-stream behavior of Pd(Pt)/HS-AlF catalysts in
3
n-pentane conversion
The reaction of n-pentane conversion in excess of hydrogen
was performed at 3508C. By choosing such relatively high reac-
tion temperature it was possible to check the stability of cata-
lysts at most drastic reaction conditions. Pt/HS-AlF and Pd/HS-
3
AlF exhibited respective overall activity and also very high se-
3
lectivity towards isopentane formation (80–90%, no 2,2-dime-
thylpropane, mainly methane and butanes in remaining prod-
ucts), showing roughly a similar behavior to that recently re-
[
15–17]
ported for platinum supported on acid zeolites.
In our pre-
[1]
vious note we reported that platinum supported on the
much less acidic support HS-MgF was only slightly active in
2
this reaction. Simultaneously, HS-AlF alone was also much less
3
effective as compared to HS-AlF -supported metal catalysts
3
(
Figure 4 in Ref. [1]).
Catalytic performances of Pd/HS-AlF and Pt/HS-AlF catalysts
3
3
were monitored for a week under standard conditions and the
results (Figure 4) were compared with kinetic experiments per-
formed for a shorter period (one day).
Figure 2. STEM (left) and HR-TEM (right) images of postreaction samples of
Pd/HS-AlF (top) and Pt/HS-AlF (bottom).
3 3
Figure 4. Long-term performances of catalytic behavior of Pd/HS-AlF
Pt/HS-AlF catalysts. The course of shorter runs (one day) taken from Ref. [1]
is also marked in black to demonstrate the reproducibility of the results.
3
and
3
ꢁ
1
Weight/flow ratio=0.2 gcat/(0.69 mLmin ) n-C
T=3508C.
5 2 5
H12. H /n-C H12 =11.54,
Figure 3. FTIR–PAS spectra of pyridine (Py) adsorbed on HS-AlF
and Pt/HS-AlF catalysts. Peaks at (B) are indicative for Brønsted acidity, and
peaks at (L) for Lewis acidity.
3 3
, Pd/HS-AlF ,
3
In the case of Pd-based catalyst the conversion rate of n-
pentane after 7 d corresponds to the conversion rate at the
beginning of the experiment (Figure 4). However, if one com-
pares the maximum of conversion (at 10 h of time on stream)
and conversion after 7 d, the calculated deactivation rate is
and postreaction samples of Pd(Pt)/HS-AlF3 catalysts, have
both Lewis and Brønsted acid sites. No significant changes in
the acidity character are caused by running the catalysts in n-
ꢁ
1>
pentane isomerization. Notably, pure HS-AlF after postfluorina-
just 0.099% h . The results in Figure 4 also reveal that the
time-on-stream behaviors of overall conversion and isomeriza-
tion selectivity are opposite. The conversion increase corre-
sponds to the isomerization selectivity reduction and, con-
versely, the decrease in conversion is associated with the in-
3
[5]
tion exhibits exclusively extremely strong Lewis acid sites.
Consequently, the presence of Brønsted sites seems to be sur-
prising. However, owing to its extremely strong Lewis acidity,
these sites easily become coordinated by hydrogen donor mol-
ecules, for example, H O. Hence, the peak at approximately
crease in isomerization selectivity. In the case of Pt/HS-AlF , the
2
3
ꢁ
1
1545 cm is indicative of water formed during the reduction
largest decrease in overall conversion occurs during the first
few hours on stream. The simultaneous mild increase in iso-
merization selectivity demonstrates that the isopentane yield is
not much affected over a week of catalytic screening. Initial in-
crease in activity was not seen in this case. The course of initial
changes of conversion and selectivity to isomerization on Pt/
adsorbing to Lewis surface sites. This is relevant to further dis-
cussion on the reaction mechanism of n-pentane isomerization
(
see below). The FTIR–PAS spectra herein resemble those ob-
[
11]
tained by Patil et al., indicating appropriateness and good
reproducibility of catalyst synthesis.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 595

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
HS-AlF3 catalyst were similar for short- and long-time tests
Catalytic tests at different H /n-pentane ratios (11.54 and
2
ꢁ
1
(Figure 4), with the rate of deactivation 0.096% h . The de-
5.77) revealed a more serious deactivation rate in the less H2-
rich reaction mixture. Attempts to regenerate the catalysts by
crease in catalyst activity is accompanied by an increase in iso-
merization selectivity. Considering the relatively high reaction
temperature (3508C) and hydrogen pressure of approximately
H treatment at 3508C for 1 h, or by a combined gas treatment
2
(O , at 4008C for 2 h, then postfluorination with CHClF at
2
2
1
bar, the activity of the catalysts was comparatively stable. We
can predict that an increase in the hydrogen pressure up to
0 bar (at industrial conditions) would slightly reduce the over-
2008C for 2 h, then H treatment at 3508C for 1 h) gave only
transient positive effects. After an initial increase, the overall
conversion returned to a previous stable level.
2
2
all conversion, but the catalyst stability would be considerably
good.
Appreciable conversion levels obtained during catalytic
screening at 3508C raise the question as to possible diffusion
limitations influencing the course of reaction. Such situations
often occur in testing zeolite-supported metal catalysts in
alkane conversions, when molecular traffic through zeolite
Kinetic parameters of n-pentane conversion on HS-AlF sup-
ported catalysts
3
The effects of reaction temperature and partial pressures of hy-
drogen and n-pentane were examined. To this reason, the re-
action was studied at much lower conversion level (<15%)
than in the other cases in which our intention was to achieve
good practical catalytic performance, that is, conversion of ap-
proximately 50% or even more. The results are shown in
Table 3.
[
18]
channels is slow. However, in our case, we disregarded such
a possibility because our catalytic materials were mesoporous
and consisted of very fine powder (sieve fraction 0.16–
0.25 mm). Instead, by examining the effect of a variable mass
of catalyst at constant space time, we recognized a considera-
ble effect of external diffusion (at 3508C). In addition, Arrhe-
nius plots constructed from experimental points col-
lected in the (descending) temperature range (350!
2508C) display two temperature regions character-
Table 3. Kinetic data for n-pentane conversion on Pd/HS-AlF
3 3
and Pt/HS-AlF catalysts.
ized by different levels of apparent activation energy
Figure 5). In the “low”-temperature region, that is,
between 250 and 3208C, the activation energies are
Catalyst Apparent reaction order with respect to
Apparent activation
(
[a]
[b]
[c]
ꢁ1
H
2
n-Pentane
energy [kJ mol
]
2
2
.0 wt% Pd/HS-AlF
.0 wt% Pt/HS-AlF
3
ꢁ0.37
ꢁ0.32
1.0
1.1
148ꢃ3.5
155ꢃ4.3
ꢁ1
higher than 100 kJmol , reproducing typical values
3
of activation energies in n-pentane isomerization
[
15]
[a] PH2 range=24.5–94.4 kPa; Pn-pentane =8.18 kPa. [b] Pn-pentane range=8.18–13.50 kPa;
over platinum supported on acidic carriers. On the
other hand, the high temperature region (i.e., 320–
P 2
H
=94.4 kPa. [c] Temperature range=280–3508C
3
508C) is characterized by very low activation energy
ꢁ
1
(
ꢀ35 kJmol ), clearly indicating the effect of diffu-
sion limitations.
n-Pentane isomerization on Pd/HS-AlF catalysts character-
ized by different metal loadings
3
Additional kinetic runs with Pt/HS-AlF3 performed on the
sample “passivated” at 3008C resulted in new experimental
data points in Figure 5 (~). The new points fit well in the Ar-
The effect of metal–acid balance in Pd-loaded HS-AlF -support-
3
rhenius plot constructed from data obtained after screening
from 3508C, indicating a similar level of catalyst’s stabilization
at both temperatures.
ed catalysts was examined by studying a series of various
metal-loaded catalysts prepared by impregnation (Figure 6).
Figure 5. Arrhenius plots for n-pentane conversion on Pd(Pt)/HS-AlF
structed from conversion data obtained at a gradually descending reaction
temperature (starting from 3508C). ~: Tinitial =3008C (for Pt/HS-AlF
).
3
con-
Figure 6. Time-on-stream behavior of variously metal-loaded Pd/HS-AlF
alysts prepared by impregnation (imp). Reaction conditions were as in
Figure 4.
3
cat-
3
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 596

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
TPD of NH from Pd(Pt)/HS-AlF and n-pentane conversion
Therefore, increasing pretreatment temperatures leads to a de-
3
3
by NH -pretreated Pd(Pt)/HS-AlF catalysts
crease in the acidity for both catalysts (Pd/HS-AlF and Pt/HS-
3
3
3
AlF3).
Two methods were used to characterize the acidic properties
of the samples: FTIR–PAS of adsorbed pyridine to differentiate
types of acid sites (see above) and the TPD of ammonia (NH3-
TPD) to characterize the strength distribution of acid sites of
the catalysts.
The results of NH -TPD encouraged us to perform kinetic ex-
3
periments with n-pentane conversion performed on the cata-
lysts for which part of the acidic sites were blocked by ammo-
nia at 3508C. On these catalysts, the strongest acid sites
should still be blocked by NH , and one would estimate possi-
3
The NH -TPD profiles of Pd(Pt)/HS-AlF catalysts subjected to
ble contributions of acid sites of weaker strength. The effect of
preadsorbed ammonia on the catalytic performance of Pd/HS-
AlF and Pt/HS-AlF is displayed in Figure 8. The drastic de-
3
3
standard postfluorination (at 2008C) and reduction at different
temperatures, from 250 to 4008C, are shown in Figure 7. The
3
3
crease in overall conversion upon treatment with NH is very
3
suggestive.
Figure 7. TPD of ammonia after pretreatment at different temperatures of
palladium- (left) and platinum-based (right) catalysts. Postfluorination tem-
perature of the catalysts was 2008C and reduction temperatures were 250,
300, 350, or 4008C.
Figure 8. The effect of ammonia on the catalytic activity of Pd/HS-AlF
and Pt/HS-AlF (right) in n-pentane isomerization. Steps were 1) treatment
with NH at 3508C; 2) reduction with H at 4008C; 3) (only for Pt) postfluori-
nation at 2008C (with CHClF ) and reduction with H at 3508C.
3
(left)
catalysts, especially those reduced at 350 and 4008C, are char-
acterized by similar broad desorption patterns. In the range of
3
3
2
2
2
150–5508C, the plots reveal basically two major peaks at ap-
proximately 2008C and 450–5308C. The first peak is much
wider than the other, spanning the range of 150–4008C, and
asymmetric on the high-temperature side. This feature sug-
gests the presence of sites with a wide range of weaker and
medium acid strength. The other peak, much sharper than the
first one, is observed at temperatures above 4508C, indicating
the presence of highly acidic sites. This feature is in line with
Effect of water treatment on the catalytic behavior of
Pd(Pt)/HS-AlF3
The aim of the pretreatment with water was to investigate
how water introduced onto the catalysts at 3508C would
[
2]
previous NH -TPD profiles from HS-AlF . The very sharp char-
change the catalytic performance of HS-AlF -supported pre-
3
3
3
acter of the peak may be caused by morphological changes in
cious metal catalysts. We presumed that such treatment would
largely modify the acidic properties of the support. Results of
HS-AlF , which also result in a vast decrease in specific surface
3
area after high-temperature treatment. It is observed that the
increase in reduction temperature shifts the location of this
high-temperature peak towards higher temperatures. However,
catalysts that were treated at lower temperature evidently
have a higher concentration (number) of strong acid sites.
water addition on the catalytic performance of Pd/HS-AlF and
3
Pt/HS-AlF catalysts are shown in Figure 9. Additional pretreat-
3
ments (reduction, fluorination) were also applied with the aim
to reproduce the initial catalytic behavior.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 597

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Pd/HS-AlF catalysts also exhibited very low H/Pd chemisorp-
3
tion ratios. Interestingly, the XRD profile of activated (i.e., post-
fluorinated with CHClF and reduced at 3508C for 1 h) 2.0 wt%
2
Pd/HS-AlF3 prepared by impregnation exhibited fairly asym-
metric Pd(111) and Pd(200) reflections, with more developed
tails directed towards lower diffraction angles (inset in
Figure 1). It would be concluded that catalyst reduction at
3508C for 1 h was not fully sufficient for removal all carbon
species dissolved in palladium. Two sources of carbon entering
palladium bulk are considered. Carbon incorporation to palla-
dium bulk occurs during catalytic hydrodechlorination of
[22]
chloro(fluoro)carbons, therefore, it is also expected to occur
during postfluorination (with CHClF ) of our catalysts. Carbon
2
dissolved in palladium was also found in supported palladium
catalysts prepared from Pd acetylacetonate and subjected to
[23]
reduction in H at 3008C for 1 h. We used the same palladi-
2
um precursor, so the presence of carbon in our catalysts would
also result from the method of catalyst preparation. On the
other hand, our catalyst reduction conditions (3508C, 1 h) con-
stitute a compromise between full carbon removal and main-
taining the high acidity of the support. More severe reduction
conditions (higher temperature, longer duration) would result
in a serious detoriation of HS-AlF (see below for discussion of
3
3
Figure 9. The influence of water on the catalytic activity of Pd/HS-AlF (left)
catalytic results). Nevertheless, the XRD profile of the postreac-
tion sample of 2.0 wt% Pd/HS-AlF3 impregnated catalyst re-
vealed very symmetric reflections originating from pure palla-
dium phase (profiles not presented).
and Pt/HS-AlF
O; 2) H reduction; 3) postfluorination and reduction; 4) additional longer
4 h) postfluorination and reduction.
3
(right) in n-pentane conversion at 3508C. 1) Treatment with
H
2
2
(
Significant surface acidity in activated HS-AlF supported pal-
3
Discussion
ladium and platinum catalysts was confirmed by FTIR photoa-
coustic spectroscopy of adsorbed pyridine (Figure 5). Incorpo-
ration of Pt(Pd) did not influence the acidic properties of HS-
Featureless XRD spectra, TEM images (Figure 2), and BET specif-
ic surface area (Table 1) demonstrate that the non-aqueous
[11]
AlF in line with previous findings. Both, the metal-free HS-
3
sol–gel synthesis of HS-AlF -supported palladium and platinum
AlF -support and reduced Pd(Pt)/HS-AlF -samples contain
3
3
3
catalysts is a very effective approach to highly dispersed cata-
lytic metals (metal particle size ꢀ2 nm and less) supported on
nanoscopic aluminum fluoride of high Lewis acidity, high spe-
cific surface area, and suitable mesoporous structure. Such su-
perior characteristics are largely maintained during high-tem-
perature treatments, up to approximately 3508C, in good
agreement with earlier reported preparations. This tempera-
ture was intentionally selected for standard catalyst screenings
to study the stability of the catalysts under such severe reac-
tion conditions. It was found that the amorphous-like structure
was maintained in the catalyst subjected to n-pentane conver-
sion performed at 3508C for overnight.
Lewis acid sites. As mentioned, less expected Brønsted acid
ꢁ
1
sites, indicated by the peak at approximately 1545 cm in the
FTIR spectrum, would result from water formed during the re-
duction that is adsorbed on Lewis surface sites. In addition,
even traces of HF formed in the course of postfluorination
might be adsorbed at the strongest Lewis sites thus creating
[
11]
Brønsted acid sites. Since the acidity strength of HS-AlF is ex-
3
tremly high, even water from moist air can be adsorbed and
remains at the surface up to 300–3508C creating Brønsted
sites. Hence, there are several explanations possible for the for-
mation of the observed weak Brønsted sites. This issue will be
discussed further regarding the catalytic properties of Pd(Pt)/
However, very low H irreversible chemisorption levels in all
HS-AlF materials (see below).
2
3
investigated catalysts (Table 2) does not reflect the high metal
dispersions as implied by XRD and TEM. Electron deficiency of
The reaction of n-pentane conversion in excess of hydrogen
performed at 3508C (Figure 4) revealed that both Pt/HS-AlF3
small metal particles (caused by electron pulling by acidic AlF ,
and Pd/HS-AlF exhibit very respective overall activity and also
3
3
0
as evidenced by X-ray photoelectron spectroscopy of Pd /HS-
very high selectivity towards isomerization (80-90%, with
mainly methane and butanes as remaining products) showing
an approximately similar behavior to that recently reported for
[
19]
AlF3) and surface blocking by organic material (acetylaceto-
nate and carbonaceous residues not fully removed postsyn-
[15–17]
thetically after postfluorination with CHClF ) are considered as
platinum supported on acidic zeolites.
The results in
2
the main reasons for the suppression of H chemisorption. In
Figure 5 reveal 63% conversion at 3508C, 47% at 3208C, and
2
addition, metal encapsulation during aging of the AlF3 gel
cannot be fully disregarded (as found for sol–gel preparations
12% at 2808C for Pt/HS-AlF . Based on Figure 4, for the same
catalyst, the yield of isopentane nearly approaches the equilib-
3
[
20,21]
of Pt/sulfated zirconia),
although a series of impregnated
rium level at 3508C.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 598

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Comparison with the behavior of Pt/HS-MgF , which ap-
in which the alkane is dehydrogenated on a metal phase, and
the alkene formed is protonated at Brønsted acid sites, yielding
2
peared only slightly active in this reaction, clearly argues for
the importance of surface acidity for providing the bifunctional
character of n-pentane isomerization. HS-MgF2 is much less
[26,27]
an alkylcarbenium ion.
It must be mentioned that the clas-
sic bifunctional metal-acid catalysts show hydrogen reaction
orders between ꢁ1 and 0, hydrocarbon reaction orders close
to 1, and apparent activation energy between 100 and
acidic than HS-AlF thus it is assumed to be also less effective
3
in creating carbenium-like species operating in the reaction
mechanism. As no adsorption of pyridine was observed over
ꢁ
1 [28–36]
150 kJmol ,
and similar kinetic parameters were also
[
11]
HS-MgF2, this clearly indicates the absence of stronger acid
found in this study (Table 3).
sites on this support. Our previous results with palladium sup-
The effect of metal–acid balance in HS-AlF supported Pd
3
[
14]
ported on less acidic supports (SiO , Al O ), for which which
catalysts on the isomerization of n-pentane was preliminary
tested by using differently metal-loaded Pd catalyst prepared
by impregnation. Their basic physicochemical characteristics
are given in Table 2 and Figure 2, and catalytic results are
shown in Figure 6. It is observed that during initial 150 min of
reaction, the overall conversion levels of 0.5 wt% 2.0 wt% cata-
lysts are similar to that by 2 wt% catalyst prepared by the sol–
gel method (Figure 4). Again, a mild maximum in conversion is
observed, but it appears at the very initial reaction time. Short-
ly thereafter, the behavior of impregnated catalysts changes
and the conversion of n-pentane over impregnated catalysts
started to decrease. Much longer evolution of overall conver-
sion observed on the sol–gel Pd catalyst would be rationalized
by a more difficult accessibility of metal particles, partly em-
2
2
3
the catalytic performance is limited to the metal function, re-
vealed very low n-pentane conversion (ꢀ1%), and a rather
nonselective course of reaction (cracking, isomerization, and
cyclization). Only if the Pd/Al O catalyst was pretreated at very
2
3
high temperature (reduction by H at 6008C for several hours),
2
the selectivity toward isomerization increased significantly, in-
[
14]
dicating the effect of Lewis acidity development. In the pres-
ent case, the HS-AlF -surface is almost free of hydroxyl groups
3
as a result of the postfluorination and consists mainly of Lewis
acid–base pair sites with very strong Lewis acid character. As it
was shown very recently by one of us, these Lewis sites are
[24]
strong enough for forming hydride-like surface species.
Based on this assumption, the isomerization would start from
[
25]
hydride abstraction on Lewis acid sites. However, as further
arguments based acidity and kinetic data show, some contribu-
tion from alkene protonation associated with Brønsted acidity
should also be considered (see below).
bedded in the AlF gel (Figure 4). Therefore, carbon removal
3
from impregnated Pd catalysts must be much faster. It is also
observed that the 0.2 wt% metal loading is not sufficient to
play a satisfactory role in bifunctional catalysis of n-pentane
conversion. Its behavior echoes the performance of the metal-
In Figure 4 it is shown that the overall activity of Pt/HS-AlF3
slightly decreases with time on stream, with concomitant in-
crease in isomerization selectivity. On the other hand, the activ-
free HS-AlF (Figure 4 in Ref. [1]), although the selectivity for
3
isomerization is still high.
ity of Pd/HS-AlF passes through a gentle maximum. Whereas
TPD of ammonia showed the presence of strong acid sites.
It is seen that the postreaction samples exhibit similar loca-
tions of TPD peaks, as those observed for prior-to-reaction cat-
alysts (Figure 7). However, the intensity of the high tempera-
ture peak is reduced after reaction. It appears that some stron-
ger acid sites are blocked by carbonaceous residues (coke). It
is also observed that catalyst precalcination at 4008C reduces,
but not fully eliminates, the amount of highly acidic sites (not
shown). Such sites are still detected in the catalysts subjected
to a prolonged reaction.
3
the behavior of Pt/HS-AlF is typical for the mild deactivation
3
of acid-supported Pt catalysts and has been frequently report-
ed, the maximum in conversion observed for Pd/HS-AlF was
3
rather unexpected. We believe that this high conversion may
result from incomplete decontamination of palladium during
reduction with hydrogen at 3508C for 1 h. As discussed above,
the procedures of catalyst synthesis and postfluorination with
CHClF can lead to massive incorporation of carbon in palladi-
2
um bulk. The time-on-stream behavior of Pd/HS-AlF shown in
3
Figure 4 would suggest that approximately half a day on
stream at 3508C is needed to remove all carbon from
palladium.
As mentioned before, NH -TPD runs revealed that practically
3
all strong acid sites should still be blocked by adsorbed ammo-
nia at 3508C. Therefore, only weaker/medium acid sites would
serve for n-pentane rearrangement performed on the catalysts
pretreated this way. The effect of such catalyst pretreatment
on catalytic activity is shown in Figure 8. The overall conver-
sion of n-pentane decreased drastically on both catalysts after
Slow catalyst deactivation at 3508C is most probably caused
by coking of acid sites of the support and not by fluorine
leaching (HF release). Deactivation at lower H /pentane ratios
2
was expected to proceed slower, but a reverse correlation was
observed (see the Supporting Information).
blocking the strong acid sites by NH . Therefore, it is evident
3
It is recalled that although the HS-AlF exhibits very high
that these strong acid sites play an important role in shaping
3
acidity, it is not an efficient catalyst for n-pentane isomerization
the overall attractive behavior of HS-AlF supported metal cata-
3
(
Figure 4 in Ref. [1]). It should deactivate rapidly because a syn-
lysts. Together with the overall conversion, the selectivity to
isopentane is also decreased after NH3 treatment. However,
after short time on stream the selectivity begins to develop,
reaching a significant level of approximately 95% for Pd/HS-
AlF and >40% for Pt/HS-AlF . It seems that our expectation
ergistic action of metal and acid sites is needed for maintain-
ing the isomerization performance. This fact in combination
with low isomerization activity of platinum supported on
much less acidic magnesium fluoride indicates that the reac-
tion of n-pentane isomerization performed on Pd(Pt)/HS-AlF3
catalysts occurs through the classical bifunctional mechanism,
3
3
that ammonia effectively blocks the strong acid centers and
limits the overall catalytic performance to metal-only catalysis
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 599

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
is working only for a very short reaction period. The noticeable
increase in isomerization selectivity (but not the same for both
catalysts) must be a result of restoring some functional surface
acidity during time on stream. Hence, it can be speculated that
at this temperature (3508C) and in the presence of the reac-
tant molecules ammonia easier desorbs from the strong Lewis
surface sites than in case of TPD conditions where just He is
used as carrier gas.
sites. The last finding confirmed several previous data report-
[42–44]
ing the poisoning of the active sites by water.
There are several good reasons to believe that water intro-
duced into our catalysts in their active (working) state should
rather destroy than develop the acidity of HS-AlF . Evidence
3
comes from theoretical considerations as well from
experiments.
Ab initio calculations of AlF –H O complex performed by
3
2
[45]
Both metals, platinum and palladium, exhibit considerable
activity in alkane conversion, but platinum is definitely a much
Ball resulted in determination of a minimum-energy struc-
ture (Scheme 1), in which tilted H atoms are directed towards
fluorine atoms.
[
37]
more active metal, not only because of a better hydrogena-
tion/dehydrogenation function, important in bifunctional iso-
merization, but also because of a better resistance to deactiva-
tion. Therefore, a higher isomerization selectivity shown by Pd/
HS-AlF cannot be assigned to metal-only action. It seems that
3
some, although not numerous, acid sites of NH -treated HS-
3
AlF are gradually liberated, playing an increasingly important
3
role in n-pentane isomerization. In the case of Pt/HS-AlF , the
3
share of restored surface acidity to the overall catalyst perfor-
mance is not as significant as for Pd/HS-AlF because platinum
3
contributes much more to the overall acidity than the less
active palladium.
After reduction of blocked catalysts at 4008C (treatment 2 in
Figure 8) the selectivity for isomerization increased to the ini-
tial value but the level of conversion on both catalysts only
3 2
Scheme 1. Schematics of the optimized geometry of the AlF –H O complex.
Adapted from Ref. [45].
gradually changed. Finally, the Pt/HS-AlF catalyst was addition-
3
ally fluorinated and reduced (treatment 3 in Figure 8). After
that, the conversion increased only to 5% and the selectivity
to isopentane approached approximately 60%. Evidently, this
pretreatment was not efficient in far-reaching recovery of sur-
Electrostatic interaction between the Al atom and the O
atom is not the only interaction between the molecules of AlF3
and H O. Significant electrostatic interaction between the hy-
2
drogen atoms of the water and the F atoms, in the eclipsing
orientation, allows for maximum electrostatic stability in the
complex. This molecular structure explains an inefficiency of
face acidity typical for untreated HS-AlF . n-Pentane hydroge-
3
nolysis occurred on metal sites still contributing to the overall
catalytic performance.
water bonded to AlF in creating substantial protonic acidity.
3
After knowing that only strong acid sites are active in the re-
action, it was interesting to investigate the nature of these
sites. As our PAS study of adsorbed pyridine revealed the pres-
Larger quantities of water (in polyhydrated aluminum fluoride)
would also not function for this purpose because of a relatively
[46]
stable extensive hydrogen bonding at the active sites.
ence of both Lewis and Brønsted acid sites in Pt(Pd)/HS-AlF , it
XPS studies of Makarowicz et al. showed that handling of
3
was of interest which sites are crucial for n-pentane isomeriza-
tion. To elucidate this question, our active catalysts were modi-
fied by addition of small quantities of water to see whether or
not such modification would result in noticeable changes of
the catalytic characteristics. Based on existing data showing
a possible effect of water addition to several acidic supports,
two options have to be taken into account. First, Lewis acid
sites will be converted into Brønsted ones, like in the case of
freshly fluorinated HS-AlF under a protecting Ar atmosphere
3
and in a glovebox exposes the samples to higher humidity
levels, so that Lewis sites are probably readily covered with OH
[47]
or H O surface species.
However, the authors argue that
2
these Brønsted acid species should be removed by treatment
at higher temperature before HS-AlF reagents can fully devel-
3
op their Lewis acidity.
Krahl et al. investigated surface acidity of HS-AlF by IR spec-
3
[
38]
[5]
alumina. A similar transformation was proposed in the case
of different acidic carriers. Appoximately sixty years ago,
Hindin and co-workers found an optimum isomerization activi-
ty of a silica-alumina catalyst after the addition of small quanti-
troscopy of adsorbed CO. A vast majority of adsorbed CO ex-
ꢁ
1
hibited large blue shifts >80 cm of the band of the CO
stretching frequency, corresponding to very strong Lewis
acidic adsorption sites. Adsorption of water at room tempera-
ture blocked the Lewis acid sites and generated Brønsted acidi-
ty. However, heating the sample under a certain partial pres-
sure of water followed by degassing results in a significant loss
of acidity. The latter appeared crucial in explaining the catalytic
[
39]
ties of water. More recently, Babou et al. suggested that the
surface sites on sulfated zirconia need small quantities of
[
40]
water to be active. In line with those findings, Kobe et al. re-
ꢁ1
ported that the addition of 75 mmolg of water to a sulfated
zirconia sample dehydrated at 5008C largely restored its high
behavior of Pt(Pd)/HS-AlF samples subjected to water treat-
3
[
41]
catalytic activity in n-butane isomerization.
On the other
ment at 3508C.
hand, excessive catalyst hydration led to poor catalytic perfor-
mance by causing extensive hydrogen bonding at the active
The results shown in Figure 9 reveal that H O treatment of
2
Pd(Pt)/HS-AlF catalysts at 3508C instantaneously reduces the
3
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 600

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
overall conversion nearly to zero, but after a few minutes the
activity approaches a stable conversion level, approximately
threefold lower than that of water-untreated catalysts. We hy-
pothesize that the introduced water “killed” Lewis acidity in
our catalysts, so the residual catalytic activity would be attrib-
uted to Brønsted acid sites. Comparison with the catalytic per-
Conclusions
Highly metal-dispersed high surface area (HS)-AlF -supported
3
platinum and palladium catalysts exhibited interesting proper-
ties in the hydroisomerization of n-pentane. Very high selectivi-
ty towards isopentane (80–90%) and appreciable conversion
levels are clearly indicative for the bifunctional character of our
catalysts. As confirmed in this study, a highly acidic character
formance of the Pt/HS-MgF catalyst (Figure 4 in Ref. [1]) clearly
2
indicates that the activity of water-pretreated Pt/HS-AlF (both
3
catalysts possess similar metal dispersions) cannot be attribut-
ed to exclusive action of metal sites. As seen in Figure 9, re-
gaining initial Lewis acidity by sole hydrogen treatment at
of nanoscopic HS-AlF is thought to play an essential role in ac-
3
tivating alkane molecules. Very stable catalytic performance of
metals supported on nanoscopic HS-AlF indicates a considera-
3
3
508C is rather inefficient. However, treating the water-affected
ble robustness of their structure and surface morphology. A
relatively low hydrogen chemisorption capability of our cata-
lysts suggests that the metal function of the catalysts should
be optimized to obtain better hydrogenation/dehydrogenation
catalysts with a fluorinating agent (CHClF ) gradually recovers
2
their initial performance. The latter might be indicative for par-
tial hydroxylation of surface AlꢁF sites and only a re-fluorina-
tion will restore the starting situation.
properties. Notably, optimization of HS-AlF supported catalysts
3
Comparison of overall conversion levels for water-treated
was not the aim of this study. Ammonia effectively blocks the
strongest acid centers and radically limits the overall catalytic
performance. Similar acid-site blockage by water is less harm-
ful. It seems that Brønsted acid sites owing to surface hydroxyl
group formation either by interaction of very strong Lewis sur-
face sites or hydrolysis of remaining alkoxide groups from non-
aqueous sol–gel synthesis participate in the bifunctional iso-
merization of n-pentane.
and untreated Pd(Pt)/HS-AlF catalysts would indicate that the
3
working state of reduced (untreated) catalysts encompasses
1
/3 of Brønsted and 2/3 of Lewis acid sites. The former sites
are detected by FTIR–PAS of adsorbed pyridine. The Brønsted
acidity probably originates from isolated hydroxyl groups in
catalysts formed by attraction of water from air thus creating
either Lewis sites by hydrolysis of remaining AlꢁOPr groups or
by interacting with strong Lewis sites. Owing to the electron-
withdrawing property of the highly electronegative fluorine
atoms, the OꢁH bond should be comparatively strong but hy-
drogen interaction with neighboring surface fluorine atoms
obviously overcompensates this effect resulting in finally weak
Brønsted acidity.
Acknowledgements
The work was realized within the International PhD Project Pro-
gram of the Foundation for Polish Science, co-financed from Eu-
ropean Regional Development Fund within Innovative Economy
Operational Program „Grant for innovation“—MPD/2009/1-B4.
O.M. expresses thanks to research staff of the Solid-State Chemis-
try and Heterogeneous Catalysis Group at Humboldt-Universitꢁt
zu Berlin for their care and support during his two research visits
in Berlin.
In conlusion, we believe that the bifunctional reaction mech-
anism including the combined action of metal and acid sites
well reflects the course of n-pentane isomerization on HS-AlF3
supported catalytic metals. The role of acid sites of the support
is twofold. First, the dominating Lewis acid sites would act
through hydride abstraction from hydrocarbon molecules and
form carbenium-like surface species. In agreement with the
“
carbenium” mechanism proposed by Kazansky et al., alkane
Keywords: acidity · isomerization · palladium · platinum ·
activation would, in principle, proceed on strong Lewis acid
supported catalysts
[
48]
sites of HS-AlF3. It has recently been found that such Lewis
sites are strong enough for forming hydride-like surface spe-
[1] O. Machynskyy, D. Łomot, K. Teinz, E. Kemnitz, Z. Karpi n´ ski, Catal.
Commun. 2012, 26, 235–238.
[
24]
cies. The second possibility of bifunctional catalysis involves
the formation of carbenium-like surface ions by protonation
[
[
[
[
2] E. Kemnitz, U. Groß, S. Rꢁdiger, C. S. Shekar, Angew. Chem. 2003, 115,
383–4386; Angew. Chem. Int. Ed. 2003, 42, 4251–4254.
3] S. K. Rꢁdiger, U. Groß, M. Feist, H. A. Prescott, S. C. Shekar, S. I. Troyanov,
E. Kemnitz, J. Mater. Chem. 2005, 15, 588–597.
4] S. Wuttke, S. M. Coman, J. Krçhnert, F. C. Jentoft, E. Kemnitz, Catal.
Today 2010, 152, 2–10.
5] T. Krahl, A. Vimont, G. Eltanany, M. Daturi, E. Kemnitz, J. Phys. Chem. C
4
(
at the Brønsted acid sites, that is, OꢁH groups, less abundant
than the Lewis acid sites) of alkene molecules that are generat-
ed at Pd(Pt) sites by dehydrogenation of n-pentane, rearrang-
ing them to isoalkyl species, in line with the classical mecha-
[
36]
nism of bifunctional isomerization. Another possibility is that
carbenium-like surface ions are formed thanks to the presence
2
007, 111, 18317–18325.
[6] M. Nickkho-Amiry, G. Eltanany, S. Wuttke, S. Rꢁdiger, E. Kemnitz, J. M.
[
49]
Winfield, J. Fluorine Chem. 2008, 129, 366–375.
of a protonic form of hydrogen, which spills over from the
metal to support sites. However, in the light of a negative
effect of hydrogen pressure found in kinetic experiments
[
7] A. Negoi, S. Wuttke, E. Kemnitz, D. Macovei, V. I. Parvulescu, C. M. Teo-
dorescu, S. M. Coman, Angew. Chem. 2010, 122, 8311–8315; Angew.
Chem. Int. Ed. 2010, 49, 8134–8138.
(Table 3) such a possibility is less likely here.
[8] S. M. Coman, P. Patil, S. Wuttke, E. Kemnitz, Chem. Commun. 2009, 460–
62.
9] A. Galadima, J. A. Anderson, R. P. K. Wells, Science World J. 2009, 4, 15–
2.
4
[
2
[
10] K. Murwani, E. Kemnitz, T. Skapin, M. Nickkho-Amiry, J. M. Winfield,
Catal. Today 2004, 88, 153–168.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 601

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[
11] P. T. Patil, A. Dimitrov, H. Kirmse, W. Neumann, E. Kemnitz, Appl. Catal. B
008, 78, 80–91.
12] V. Quaschning, J. Deutsch, P. Druska, H.-J. Niclas, E. Kemnitz, J. Catal.
998, 177, 164–174.
[32] M. Guisnet, V. Fouche, Appl. Catal. 1991, 71, 307–317.
2
[33] A. Hollꢂ, J. Hancsꢂk, D. Kallꢂ, Appl. Catal. A 2002, 229, 93–102.
[34] T. Matsuda, K. Watanabe, H. Sakagami, N. Takahashi, Appl. Catal. A 2003,
242, 267–274.
[
1
[
[
[
13] M. Skotak, D. Łomot, Z. Karpi n´ ski, Appl. Catal. A 2002, 229, 103–115.
14] M. Bonarowska, Z. Karpi n´ ski, Pol. J. Chem. 2006, 80, 1877–1887.
15] T. Matsuda, H. Kodama, H. Sakagami, N. Takahashi, Appl. Catal. A 2003,
[35] ꢇ. B. Demirci, F. Garin, J. Mol. Catal. A 2002, 188, 233–243.
[36] Y. Ono, Catal. Today 2003, 81, 3–16.
[37] a) J. R. Anderson, N. R. Avery, J. Catal. 1966, 5, 446–463; b) M. Boudart,
L. D. Ptak, J. Catal. 1970, 16, 90–96; c) Z. Karpi n´ ski, T. Ko s´ cielski, J. Catal.
1980, 63, 313–323.
[38] See, for example, H. Knçzinger, Stud. Surf. Sci. Catal. 1985, 20, 111 – 125
(Catalysis by Acids and Bases (Eds.: B. Imelik, C. Naccache, G. Coudurier,
Y. Ben Taarit, J. C. Vedrine), Elsevier, Amsterdam, 1985).
[39] S. G. Hindin, A. G. Oblad, G. A. Mills, J. Am. Chem. Soc. 1955, 77, 535–
538.
[40] F. Babou, G. Coudurier, J. C. Vedrine, J. Catal. 1995, 152, 341–349.
[41] J. M. Kobe, M. R. Gonzalez, K. B. Fogash, J. A. Dumesic, J. Catal. 1996,
164, 459–466.
[42] C. Morterra, G. Cerrato, F. Pinna, M. Signoretto, G. Strukul, J. Catal. 1994,
149, 181–188.
[43] C. Morterra, G. Cerrato, F. Pinna, M. Signoretto, J. Phys. Chem. 1994, 98,
12373–12381.
[44] R. A. Keogh, R. Srinivasan, B. H. Davis, J. Catal. 1995, 151, 292–299.
[45] D. W. Ball, J. Phys. Chem. 1995, 99, 12786–12789.
[46] M. Krossner, G. Scholz, R. Stçsser, J. Phys. Chem. A 1997, 101, 1555–
1560.
248, 269–278.
[
[
[
[
16] M. C. Alvares-Galvan, A. Brito, F. J. Garcia-Alvarez, V. A. dela PeÇa O’Shea,
E. Borges, B. Pawelec, Appl. Catal. A 2008, 350, 38–45.
17] C. M. Lꢂpez, V. Sazo, P. Pꢃrez, L. V. Garcꢄa, Appl. Catal. A 2010, 372, 108–
1
13.
18] for example, H. Liu, G. D. Lei, W. M. H. Sachtler, Appl. Catal. A 1996, 137,
67–177.
1
19] P. T. Patil, Nanoscopic Metal Fluoride Based Novel Solid Catalysts, PhD
Thesis, Humboldt-Universitꢅt zu Berlin, 2008, p. 41 (http://edoc.hu-ber-
lin.de/dissertationen/patil-pratap-tukaram-2008-12-19/PDF/patil.pdf).
20] Z. Paꢆl, U. Wild, M. Muhler, J.-M. Manoli, C. Potvin, T. Buchholz, S.
Sprenger, G. Resofszki, Appl. Catal. A 1999, 188, 257–266.
21] J.-M. Manoli, C. Potvin, M. Muhler, U. Wild, G. Resofszki, T. Buchholz, Z.
Paꢆl, J. Catal. 1998, 178, 338–351.
22] a) E. J. A. X. van de Sandt, A. Wiersma, M. Makkee, H. van Bekkum, J. A.
Moulijn, Catal. Today 1997, 35, 163–170; b) E. J. A. X. van de Sandt, A.
Wiersma, M. Makkee, H. van Bekkum, J. A. Moulijn, Appl. Catal. A 1997,
[
[
[
1
55, 59–73; c) W. Juszczyk, A. Malinowski, Z. Karpi n´ ski, Appl. Catal. A
1
998, 166, 311–319.
[47] A. Makarowicz, C. L. Bailey, N. Weiher, E. Kemnitz, S. L. M. Schroeder, S.
Mukhopadhyay, A. Wander, B. G. Searle, N. M. Harrison, Phys. Chem.
Chem. Phys. 2009, 11, 5664–5673, and references of N. M. Harrison’s
group therein.
[
[
[
23] M. Krishnankutty, M. A. Vannice, J. Catal. 1995, 155, 312–326, especially
Figure 4 on p. 316.
24] M. Ahrens, G. Scholz, Th. Braun, E. Kemnitz, Angew. Chem. 2013, 125,
5
436–5440; Angew. Chem. Int. Ed. 2013, 52, 5328–5332.
[48] V. B. Kazansky, V. Yu. Borovkov, A. V. Zaitsev in Proceedings of the Ninth
International Congress on Catalysis, Calgary, 1988, Vol. 3 (Eds.: M. J. Phil-
lips, M. Ternan), Chemical Institute of Canada, Ottawa, 1988, pp. 1426–
1433.
[49] A. Zhang, I. Nakamura, K. Aimoto, K. Fujimoto, Ind. Eng. Chem. Res.
1995, 34, 1074–1080.
25] for example, J. C. Duchet, D. Guillaume, A. Monnier, C. Dujardin, J. P.
Gilson, J. van Gestel, G. Szabo, P. Nascimento, J. Catal. 2001, 198, 328–
3
37.
26] G. A. Mills, H. Heinemann, T. H. Milliken, A. G. Oblad, Ind. Eng. Chem.
953, 45, 134–137.
[
1
[
[
[
[
[
27] P. B. Weisz, E. W. Swegler, Science 1957, 126, 31–32.
28] J. H. Sinfelt, H. Hurwitz, J. C. Rohrer, J. Phys. Chem. 1960, 64, 892–894.
29] J. H. Sinfelt, Adv. Chem. Eng. 1964, 5, 37–74.
30] F. Chevalier, M. Guisnet, R. Maurel, C. R. Acad. Sci. Ser. IIc 1976, 282, 3–5.
31] M. Belloum, C. Travers, J. P. Bournonville, Rev. Inst. Fr. Pet. 1991, 46, 89–
Received: October 4, 2013
Revised: November 18, 2013
Published online on January 2, 2014
107.
ꢀ
2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 592 – 602 602

Copyright of ChemCatChem is the property of Wiley-Blackwell and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.