The Co-Entrapment of a homogeneous catalyst and an ionic liquid by a sol-gel method: Recyclable Ionogel hydrogenation catalysts

Article abstract of DOI:10.1002/chem.200801809

Molecular hydrogenation catalysts have been co-entrapped with the ionic liquid [Bmim]NTf₂ inside a silica matrix by a sol-gel method. These catalytic ionogels have been compared to simple catalyst-doped glasses, the parent homogeneous catalysts, commercial heterogeneous cata-lysts, and Rh-doped mesoporous silica. The most active ionogel has been characterised by transmission electron mi-croscopy, X-ray photoelectron spectroscopy, and solid state NMR before and after catalysis. The ionogel catalysts were found to be remarkably active, recyclable and resistant to chemical change. 2009 Wiley-VCH Verlag GmbH & Cu. KGaA.

Full text of DOI:10.1002/chem.200801809

DOI: 10.1002/chem.200801809
The Co-Entrapment of a Homogeneous Catalyst and an Ionic Liquid by a
Sol–gel Method: Recyclable Ionogel Hydrogenation Catalysts
[
a]
[a]
[a]
[a]
Steven J. Craythorne, Kris Anderson, Fabio Lorenzini, Christina McCausland,
[
b]
[b]
[a]
[a]
Emily F. Smith, Peter Licence, Andrew C. Marr,* and Patricia C. Marr*
Abstract: Molecular hydrogenation
catalysts have been co-entrapped with
the ionic liquid [Bmim]NTf2 inside a
silica matrix by a sol–gel method.
These catalytic ionogels have been
compared to simple catalyst-doped
glasses, the parent homogeneous cata-
lysts, commercial heterogeneous cata-
lysts, and Rh-doped mesoporous silica.
The most active ionogel has been char-
acterised by transmission electron mi-
croscopy, X-ray photoelectron spectros-
copy, and solid state NMR before and
after catalysis. The ionogel catalysts
were found to be remarkably active, re-
cyclable and resistant to chemical
change.
Keywords: homogeneous catalysis ·
hydrogenation
· ionic liquids ·
sol–gel processes
Introduction
trapped in silica have been found to show resistance to poi-
soning, as a result incompatible reactions can be conducted
in tandem. Each of the following pairs, acidic and basic,
[9]
Homogeneous catalysts can be entrapped within solid matri-
ces to facilitate their separation from the products of reac-
tion and aid catalyst recycling. The matrix forms a porous
prison that prevents leaching, but allows the diffusion of the
substrate and product to and from the catalyst. Blum and
[
10]
chemocatalytic and enzymic
and oxidation and reduc-
[11]
tion, have been demonstrated in one pot reactions.
The emergence of ionic-liquid-mediated methods for the
preparation of mesoporous and microporous materials
led us to apply ionic liquids to the preparation of catalytic
[12–16]
[1]
Avnir have pioneered the entrapment of molecular cata-
lysts within silica by the sol–gel method. Catalytic materials
can be prepared to promote a wide variety of reactions in-
[17]
materials, firstly by the entrapment of metal nanoparticles
and subsequently by the entrapment of a molecular cata-
[
2]
[3]
[18]
cluding isomerization, disproportionation, hydrogenation
lyst. The resultant catalytic material can be either an aer-
[
4]
[5]
[12]
[14,19]
and dehalogenation, hydroformylation, Pauson–Khand
ogel or an ionogel.
tain entrapped ionic liquid.
Ionogels are silica gels that con-
[6]
[7]
[8]
coupling, epoxidation, and Heck coupling. Catalysts en-
The application of doped ionogel materials as catalysts
for rhodium-centred hydroformylation was recently demon-
[20]
strated. Vinylarenes could be selectively transformed into
branched aldehydes. The ionogel catalyst exhibited moder-
ate recycling ability.
[
a] S. J. Craythorne, Dr. K. Anderson, F. Lorenzini, C. McCausland,
Dr. A. C. Marr, Dr. P. C. Marr
School of Chemistry and Chemical Engineering and
Queenꢀs University Ionic Liquids Laboratories (QUILL)
Queenꢀs University Belfast, David Keir Building
Belfast, BT9 5AG (UK)
Ionic liquids have also been coated on the outside of inor-
ganic oxides to provide recyclable heterogenized catalysts,
these are often referred to as supported ionic liquid (phase)
[21–23]
catalysts and have been recently reviewed.
Fax : (+44)28-9097-6524
E-mail: a.marr@qub.ac.uk
In this paper we present hydrogenation catalysts prepared
by ionic liquid modified sol–gel processes involving the en-
trapment of homogeneous pre-catalysts. The catalytic mate-
p.marr@qub.ac.uk
[
b] E. F. Smith, Dr. P. Licence
School of Chemistry
rial doped with [RhCl ACHTUNGTRNEUN(G PPh ) ] exhibited excellent recycling
3 3
The University of Nottingham
University Park
Nottingham, NG7 2RD (UK)
without the need for pre-treatment and turnover frequencies
well above those achieved under homogeneous or conven-
tional heterogeneous conditions. This potent ionogel catalyst
was analyzed before and after recycling to assess any
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801809.
7094
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7094 – 7100

FULL PAPER
changes in structure resulting from the exposure to reaction
conditions. Some of the initial results were reported as a
communication.
mately 6 nm in diameter. The surface areas of the materials
vary considerably and are uniformly low for the ionogel ma-
terials, this is an indication of the entrapment of ionic liquid
within the pores of the material.
[18]
3
1
The P solid state NMR of a glass containing [Rh -
2
25]
[
AHCTUNGTRENNGNU( COD) AHCTUNGTNERGUN( dppm) ACHTUNGRETNNUG( m -Cl)]BF has been reported previously,
2 2 4
Results and Discussion
spectra were consistent with solution phase spectra suggest-
ing that the catalyst was entrapped in tact.
3
1
Catalyst preparation: The molecular catalyst precursors
The P solid state NMR of ionogel entrapped [RhCl-
ACHTUNGTREN(NUGN PPh ) ] (3) (Figure 1), d=35.4 ppm. This peak was assigned
[24]
[25]
4
[RhCl
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(PPh ) ]
and
[Rh2
A
H
U
T
E
N
N
(COD)
A
H
U
G
R
N
U
G
A
H
U
G
R
N
U
G
3
3
2
2
3 3
(COD=1,5-cyclooctadiene, dppm=bis(diphenylphosphino)-
to OPPh , the typical value for this resonance (31 ppm) was
3
methane) were entrapped in a silica glass (1 and 2) and an
ionogel (3 and 4) by sol–gel methods.
[18,25]
Preparing catalyst-doped silica glasses:
catalyst precursor, [RhCl(PPh ) ] for glass 1 and [Rh -
(COD) (dppm)(m -Cl)]BF for glass 2, was dissolved in the
the molecular
AHCTUNGTRENNUNG
3
3
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
ACHTUNGTRENNUNG
2
2
4
minimum quantity of dichloromethane and added to a solu-
tion of tetraethoxyorthosilicate (TEOS) dissolved in ethanol
containing a small quantity of water and benzyl amine.
After the gel had formed it was aged for a week to form a
glass, broken and extracted with dichloromethane under
reflux to remove any material that was not entrapped and
then ground to form a powder.
3
1
Preparing catalyst-doped ionogels: A [RhCl
ionogel (3) was prepared as reported previously.
(COD) (dppm)(m -Cl)]BF was entrapped in an ionogel by a
ACHTUNGTNERNUNG( PPh ) ] doped
3 3
Figure 1. P NMR spectrum of 3.
[18]
[Rh2-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
ACHTUNGTRENNUNG
2
2
4
modified version of the published procedure. [Rh ACHTNUGTRNEUNG( COD) -
2 2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(dppm)
methane under nitrogen and added to a stirred mixture of 3-
methyl imidazolium bistrifluoromethanesulfonimide
[Bmim]NTf ). Once the catalyst was added no further pre-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(m -Cl)]BF was dissolved in a minimum of dichloro-
shifted, owing the presence of the ionic liquid, this shift was
observed in a solution of OPPh in [Bmim]NTf and in the
solid state within an ionogel prepared containing entrapped
phosphine oxide. Loss of phosphine is expected during the
activation of the molecular catalyst and the observation of
2
4
3
2
(
2
cautions to exclude air were taken. Methanoic acid was
added to initiate gelation. Ionogel formation typically took
OPPh does not prove or disprove the existence of rhodium
3
3
0 min at room temperature. The gel was then aged for a
complexes or particles of metallic rhodium within 3. The
rhodium present in the sample was at a low concentration
(0.06%) and this, coupled with the mobile nature of the spe-
cies entrapped within the material, makes the solid state
NMR difficult to acquire and resolve, as a consequence any
multiplets exhibiting RhÀP coupling may be lost in the
week, broken up and extracted with dichloromethane under
reflux. The catalytic ionogel material (4) was dried and
ground to a powder.
Unused catalyst characterisation: The BET surface areas
and average pore diameters of the materials 1–4 are given
in Table 1 and compared with undoped materials 5 and 6.
For each method the average pore size was reduced by the
entrapment of the catalyst, this effect is greatest for the ion-
ogel. In each of the catalysts 1–4 the pore size is approxi-
noise.
2
9
The Si solid state NMR of ionogel entrapped [RhCl-
A
H
U
G
R
N
U
(PPh ) ] (Figure 2) contained a resonance at d=À111 ppm,
3
3
indicating Q4 species, d=À101 ppm signifying Q3 species
and d=À92 ppm representing Q2 species. As this spectrum
was the result of a cross-polarization experiment, the rela-
tive intensities are not reliable and Q4 is likely to be under-
represented, owing to the lack of nearby protons to cross-
polarize to. The C solid state NMR spectrum of 3 was of
poor quality, owing to the low concentration of C but reso-
nances were consistent with [Bmim]NTf2, as the spectrum
Table 1. BET surface area and pore diameter of the unused catalysts.
[
a]
[b]
Dopant
[RhCl(PPh
[Rh (COD)
[RhCl(PPh
[Rh (COD)
Method Surface area
Pore
13
1
3
1
2
3
4
5
6
7
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3
)
3
]
glass
glass
177
140
19
12
248
16
6.0
6.1
6.0
6.0
6.7
8.1
8.0
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(dppm)
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(m
2
2
-Cl)]BF
-Cl)]BF
4
4
[26]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3
)
3
]
ion. liq.
ion. liq.
glass
ion. liq.
ion. liq.
correlated closely with the solution phase spectrum.
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
(dppm)
A( m
C
H
T
U
N
G
T
R
E
N
N
U
N
G
Transmission electron microscopy (TEM) images of mate-
rial 3 (Figure 3) showed no sign of metallic rhodium. Ele-
mental analysis (EDX) showed traces of rhodium at the
limit of detection. It should be noted that low concentra-
none
none
Rh colloid from [RhCl
A( COD)]
H
U
G
R
N
N
2
563
2
À1
[
a] in m g ; [b] in nm.
Chem. Eur. J. 2009, 15, 7094 – 7100
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7095

A. C. Marr, P. C. Marr et al.
SiO and [Bmim]NTf in an approximate ratio of 12:1. Some
2
2
carbon contamination was detected; this may have resulted
from residual ethyl groups due to the incomplete reaction of
TEOS.
Catalyst testing: The materials 1–4 were tested as catalysts
for styrene hydrogenation (Scheme 1). The activity was
compared with the parent homogeneous catalyst and com-
2
9
Figure 2. Si NMR spectrum 3.
Scheme 1. The selective hydrogenation of styrene to ethyl benzene.
mercially available 5% Rh on carbon and 5% Rh on alumi-
na (Table 3). Routine controls were conducted to verify the
inactivity of the blank materials (5 and 6), the autoclave and
the reaction solution.
Table 3. Comparison of the styrene hydrogenation activity.
[
a]
À6
Catalyst
[Rh]
[
[Rh] [ꢂ 10 Convn TOF
Leach
À3
[b]
2
À1
[c]
wt.%] mol·dm
]
[%]
[ꢂ10 min ] [ppm]
1
2
3
4
0.059
0.360
0.060
0.070
–
7.8
48
8.4
25
200
200
75
69
98
30
n.a.
n.a.
20
3.8
32
3.2
5.3
4.5
<LOD
<LOD
<LOD
0.011
–
[
[
d]
A
H
U
G
R
N
U
G
A
H
U
G
E
N
N
(PPh
3
)
3
]
d,e]
A
T
G
R
N
N
2
A
H
U
G
E
N
N
2
A
T
N
T
E
U
G
–
A
H
U
G
R
N
N
2
4
Rh on carbon
Rh on alumina
7
5
5
220
660
15
99
98
25
1.2
0.40
4.3
0.550
0.320
0.014
0.057
[25]
[
a] weight percentage rhodium calculated from ICP/OES data ; [b] All
reactions
in
CH
2
Cl
2
,
T=1008C,
P
(1008C) =27 bar,
[styrene]=
À3
0
.793 moldm , stir rate 1200 rpm, reaction time 30 min, percentage con-
version of styrene to ethyl benzene; [c] the concentration of rhodium
[
25]
present in the product solution calculated from ICP/OES data, limit of
detection (LOD) 0.005 ppm ; [d] initial rate calculated by gas uptake;
Figure 3. TEM images of 3 before (top) and after (bottom) catalysis.
[
25]
[
e] quoted per rhodium.
tions of rhodium nanoparticles are hard to find by electron
microscopy.
The catalysts based on entrapped [RhCl ACHTUNGTRNENUG( PPh ) ] (1 and 3)
3 3
The X-ray photoelectron spectroscopy (XPS) analysis and
comparison with the parent ionic liquid (Table 2), shows
that the surface of the material consisted almost entirely of
outperformed all other catalysts tested. The turnover fre-
quencies for 1 and 3 were 4–6 times greater than the initial
[25]
rate measured for the homogeneous catalyst. This reveals
that the Rh catalyst derived from [RhCl(PPh ) ] was entrap-
AHCTUNGTRENNUNG
3
3
ped in a reactive state and protected from decomposition by
the silica matrix.
AHCTUNGTNERGUN[ Rh ACHTUNGRTENNUNG( COD) CAHTUNGTRENN(NUG dppm) ACHTNUGTRENUNGN( m -Cl)]BF was entrapped (2 and 4)
2 2 2 4
intact and the catalytic activity observed within the matrix
approached that observed for the homogeneous catalyst.
Notably, the rate of hydrogenation fell during homogeneous
2
Table 2. Atomic percentages for [Bmim]NTf and 3 from XPS data.
Atomic [%]
Peak BE [eV]
A
H
U
G
R
N
N
[Bmim]NTf
2
3 unused
3 used
F 1s
O 1s
N 1s
C 1s
S 2p
Si 2p
686
530
399
284
167
101
26
14
12
41
7
9
48
4
18
2
13
40
6
22
4
catalytic tests of [RhCl
ACTHUNGRTENNUN(G PPh ) ] and [Rh2 ACHTUNGTNERUN(G COD) ACHTNUGTRENNNGU( dppm) ACHTUNGTRENNUNG( m -
3 3 2 2
–
18
15
[25]
Cl)]BF as the composition of the solution changed. The
4
7096
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7094 – 7100

Recyclable Ionogel Hydrogenation Catalysts
FULL PAPER
rate measured for 2 and 4 is an average and can be main-
tained in further catalytic runs.
and 4 showed no change in activity, whereas [RhCl ACHTUNGTRENNUNG( PPh ) ]
3 3
entrapped in glass (1) showed an increase in activity over 5
recycles. In contrast the commercial catalysts Rh on carbon
and Rh on alumina exhibited a considerable loss of activity
over the 5 runs. The robust activity of catalytic materials 1–4
is consistent with the concept of the matrix as a porous pro-
tector of the active species. The semi-permeable material
prevents contact with potential poisons such as other cata-
lyst molecules thus achieving better active site-isolation.
Metal leaching, as detected by inductively coupled
plasma/optical emission spectroscopy (ICP/OES) analysis of
the product solution is presented in Figure 5. Leaching from
The results presented for catalytic materials 1–3 are for
[18,25]
different catalyst batches to those reported previously,
and some differences in activity were observed. Differences
in activity are more acute for catalytic materials prepared
by conventional sol–gel methods (1 and 2). Catalysts similar
À1 [18]
to 1 can exhibit TOF as low as 180 min . In contrast, all
batches of ionogel catalyst 3 prepared thus far have activity
greater than the homogeneous catalyst [RhCl ACHTUNGTENRUNG( PPh ) ] under
3 3
the same conditions and the presence of the ionic liquid ap-
pears to stabilise the most active state of the catalyst. No
changes in selectivity were observed for entrapped catalysts
1
–4. Further reaction led to full conversion with no loss of
selectivity.
The activity of the heterogeneous catalysts, expressed as
turnover frequency, was low in comparison to the molecular
catalysts, this reflects the high concentration of rhodium
present. The heterogeneous rhodium catalysts tested exhib-
ited selective terminal alkene hydrogenation over short re-
action times. A previous batch of 5% Rh on carbon was
shown to be active for ring hydrogenation under similar con-
[18]
ditions.
Catalyst recycling: The ability of catalysts 1–4, Rh on
carbon and Rh on alumina to hydrogenate styrene to ethyl
benzene after recycling was tested. The hydrogenation reac-
tion was performed for each catalyst and then the product
solution removed and replaced with a fresh substrate solu-
tion. No attempt was made to regenerate or pre-treat the
catalysts. The results of styrene conversion to ethyl benzene
are presented in Figure 4.
Figure 5. Percentage of Rh leached from the solid catalysts.
entrapped catalysts 1–3 was below the limit of detection
(
LOD 0.005 ppm Rh in the product solution). Measurable
leaching was detected from 4, Rh on carbon and Rh on alu-
mina, in each case the heaviest leach was detected after the
first run, this value is reported in Table 3. In Figure 5 leach-
ing is expressed as a percentage of the rhodium present and
therefore the concentration of rhodium in the product is 30–
5
0 times less for [Rh ACHTNUGTRENNUGN( COD) ACHTNUREGTNNUNG( dppm) AHCUTNGERNNUG( m -Cl)]BF entrapped
2 2 2 4
in an ionogel than 5% Rh on alumina or carbon. The mag-
nitude of leaching is not great enough to explain the loss of
activity observed for the commercial catalysts and the most
active metal sites must be being lost. Leaching from 4 has
little effect on the activity and may originate from Rh con-
tamination, owing to [Rh CAHTUNGTRNEGNUN( COD) CTHGUNERTNNUNG( dppm) ACUTHNGTRNENUGN( m -Cl)]BF decom-
2A
2 2 4
position during the sol–gel process.
Catalyst generations: The ionic liquid recovered from the
ionogel catalyst preparation by extraction with dichlorome-
thane was used to prepare a second generation catalytic ma-
terial and recovered again and used for a third generation
catalytic ionogel. All three materials gave similar activity
and identical selectivity and exhibited leaching at or below
the limit of detection. The amount of ionic liquid incorpo-
rated into the ionogel in each case was 20%.
Figure 4. The conversion of styrene catalysed by recycled catalysts.
Entrapped catalysts 1–4 recycled well. No significant
losses of activity were observed for any of the entrapped
catalysts. Recycling of the ionogel catalyst 3 could be ex-
tended to 10 runs with no change in activity. Materials 2, 3
Chem. Eur. J. 2009, 15, 7094 – 7100
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7097

A. C. Marr, P. C. Marr et al.
Changing the reaction conditions: Catalytic ionogel 3 was
active for the hydrogenation of styrene under a range of
conditions. In toluene at 1108C and 7 bar total pressure the
effect of changing the particle size of the catalyst was inves-
tigated (Table 4). The stir rate was chosen as 1200 rpm.
catalysts and molecular catalysts entrapped by ionogel meth-
ods, a material was designed and synthesised to have the
characteristics of both. Material 7 was prepared by the iono-
gel entrapment of a Rh colloid, prepared by the decomposi-
tion of [RhCl
ACHTUNGTRENNUNG( COD)] at 758C under hydrogen in
2
[
Bmim]NTf . The resulting ionogel was aged and extracted
2
with dichloromethane to remove molecular species that
were not entrapped in the matrix. The dried and powdered
silica was analysed by BET, ICP and TEM. The BET analy-
sis (Table 1) suggested that the material produced did not
Table 4. Changing the particle size of catalyst 3.
[
a]
[b]
Particle size
conversion [%]
<
125
61
56
53
83
1
2
25–250
50–300
contain entrapped ionic liquid, as
563 m g was measured. The matrix was found to be highly
mesoporous with a pore volume of 1.08 cm g and hardly
a surface area of
2
À1
>
300
3
[
P
a] in mm separated by sieves [b] All reactions in toluene, T=1108C,
(1108C) =7 bar, [styrene]=0.872 moldm , stir rate 1200 rpm, 0.0500 g cat-
2
À1
À3
any micropores (micropore surface area 71 m g and pore
3
À1
alyst, reaction time 2 h, percentage conversion of styrene to ethyl ben-
zene.
volume 0.03 cm g ). The average pore size of the sample
was 8.0 nm. This material was analogous to Pd doped aero-
[17]
gel catalysts previously published.
XPS verified the ab-
Mixing was not rate limiting above 900 rpm. The increase in
rate for the largest particle size is consistent with a bulk cat-
alytic process.
sence of significant quantities of ionic liquid in the matrix
(N<0.5%, F<0.2%, S<0.1%). The rhodium particles in 7
were not resolved by TEM imaging (Figure 6).
Changing the substrate: In order to test the applicability of
our ionogel catalyst to the conversion of more complex sub-
strates, the hydrogenation of levoglucosenone was per-
formed (Scheme 2). In toluene at 1108C under 25 bar H2
90% of the levoglucosenone was converted in 20 h with
complete selectivity to alkene hydrogenation.
Figure 6. TEM image of 7.
Scheme 2. The selective hydrogenation of Levoglucosenone.
The Rh content of 7 was assessed by ICP (0.057 wt%)
and the material was used as a catalyst for the hydrogena-
tion of styrene (Scheme 1, Table 3). The activity of this het-
erogeneous catalyst (TOF 430 min ), was comparable with
the homogeneous catalysts, but less than that of the catalysts
Used catalyst characterisation: Catalytic ionogel 3 was ana-
lyzed by BET, XPS, solid state NMR and TEM after 5 cata-
2
À1
À1
lytic runs. The surface area by BET was 11 m g , with an
adsorption average pore diameter of 6.0 nm. The pore size
had not changed significantly and this is consistent with the
retention of catalytic activity. The surface area was signifi-
cantly reduced and this change in structure is reflected in
the XPS data for the used catalyst (Table 2). The XPS of the
prepared by the entrapment of [RhCl CAHTUNGTRNEUNG( PPh ) ]. The quantity
3 3
of rhodium found in the product was low, consistent with an
entrapped catalyst of low rhodium concentration. Catalyst 7
was recycled and used to hydrogenate styrene 5 times
(Figure 4). Heterogeneous catalyst 7 behaved in a similar
way to Rh on alumina, losing significant activity during the
first 3 recycles. The leach rate of rhodium from 7 (Figure 5)
was of a similar magnitude to that observed for catalyst 4.
used catalyst reveals a surface composition of 7:1 SiO :-
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
[Bmim]NTf . The change in surface area, surface composi-
2
tion and reduction in carbon contamination can be ex-
plained by an aging of the silica matrix and expulsion of the
3
1
29
unreacted ethoxy groups. P and Si solid state NMR data
for the used catalyst were identical to the unused catalyst
(
inlays in Figures 1 and 2). No dark spots indicative of rho-
dium metal particles were evident in the TEM images
Figure 3).
Conclusions
(
We have presented an example (3) of a catalytic ionogel
which gives higher activity than the parent homogeneous
catalyst and recycles extremely efficiently. The ionic liquid
modified sol–gel method of catalyst synthesis reduces gela-
Investigating the nature of entrapped catalysts, catalyst 7:
To better understand the difference between heterogeneous
7098
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7094 – 7100

Recyclable Ionogel Hydrogenation Catalysts
FULL PAPER
tion times with respect to conventional methods and affords
catalysts of reliable composition. Ionogel catalysts vary very
little from batch to batch and give excellent selectivity and
activity.
GC column used was a Restek rtx5 60 metres within a PerkinElmer
Turbo-Mass GC Mass Spectrometer.
Preparation of catalysts 3 and 4: Rhodium Catalysts Entrapped in an
À6
Ionogel Matrix: The catalyst precursor (4.0ꢂ10 mol) was dissolved in
the minimum amount of dichloromethane (ꢀ4 drops) and added to a so-
Our evidence suggests that the molecular catalyst was en-
+
À
3
3
lution of [Bmim]
A
H
U
G
R
N
U
G
2
]
(1 cm ) and TEOS (0.919 g, 1.0 cm , 4.41ꢂ
À3
trapped intact in the case of [Rh
A
H
U
G
R
N
U
G
A
H
U
G
R
N
N
(dppm)
A
H
U
G
R
N
U
G
10 mol, Sigma–Aldrich). The resulting solution was vigorously stirred
2
2
2
3
À2
for 20 min. After this time methanoic acid (2.44 g, 2.0 cm , 5.3ꢂ10 mol,
Sigma–Aldrich) was added dropwise, and the mixture stirred for 5 min.
The magnetic stirrer was removed. Gelation time was less than 30 min.
The material was aged for a period of one week and then continuously
extracted with boiling dichloromethane to remove excess ionic liquid and
catalyst. The material was ground to a fine powder with a mortar and
and activated in the case of [RhCl ACUTHNGTRNNEUN(G PPh ) ]. [RhCl AHCTUNGERTNNUN(G PPh ) ]
3 3 3 3
was not entrapped intact as OPPh was detected in the re-
3
sultant materials. The rhodium in these catalysts was activat-
ed during entrapment and the resultant materials were more
active than homogeneous [RhCl ACHTGNUTRNEUNG( PPh ) ]. The active catalyst
3 3
pestle.
was particularly complemented by the protective environ-
ment of the ionogel, evidenced by the excellent activity and
recycling of 3 compared with the inferior activity and recy-
cling of Rh doped mesoporous silica 7. The combination of
ionic liquid and solid oxide matrix forms stable pores that
contain the active species. The substrate and product can
readily diffuse to and from the catalyst, which is imprisoned
within a nanoscale environment.
As far as we can detect the chemical integrity of the ma-
terial consisting of metal complex, ionic liquid and silica
changes little during catalytic activity. The structure of the
surface of the matrix undergoes some changes, but this did
not affect the catalytic performance.
À5
Preparation of catalyst 7: [RhCl
A
H
U
G
R
N
U
G
2
(0.0238 g, 4.8ꢂ10 mol) in
+
À
3
[
Bmim]
A
H
U
G
R
N
U
G
2
]
2
for 1.5 h re-
+
À
3
3
À2
A
H
U
G
E
N
N
2
] (3 cm ) and TEOS (5.604 g, 6.0 cm , 2.69ꢂ10 mol) were added
3
1
2.0 cm , 0.32 mol) was added dropwise, the mixture stirred for 5 min
and the magnetic stirrer removed. Gelation occurred within an hour. The
material was aged for one week and then continuously extracted with
boiling dichloromethane to remove excess ionic liquid and catalyst. The
material was ground to a fine powder with a mortar and pestle.
Catalytic reactions: The catalyst was weighed and added to the hydroge-
nation vessel under air. The autoclave was sealed, evacuated and gently
purged with nitrogen. The reaction solvent (dichloromethane or toluene)
and styrene were injected and the autoclave pressurized with hydrogen.
The mechanical stirrer was started and the reactor heated to the reaction
temperature. After the reaction time the reactor was cooled and depres-
surized. The resulting solution was syringed from the vessel and submit-
ted for analysis by gas chromatography and ICP/OES. The solid catalyst
was left in the vessel. For recycles the autoclave was resealed, evacuated
and gently flushed with nitrogen. Further styrene and solvent were inject-
ed to repeat the procedure. Experiments were conducted with either di-
The exact role of the molecular catalyst in templating the
matrix and retaining the ionic liquid is not known, but the
process is clearly dopant dependent and there is a fine bal-
ance between the synthesis of ionogel materials such as 3
and those devoid of ionic liquid such as 7.
3
3
À3
chloromethane (10 cm ) and styrene (1.0 cm , 8.7ꢂ10 mol) or toluene
(
3
3
À3
4.5 cm ) and styrene (0.5 cm , 4.4ꢂ10 mol). The reduction of levoglu-
cosenone was performed with the following quantities: 3 (0.050 g), levo-
glucosenone (0.2180 g, 0.002 mol), toluene (5.0 cm ). The reaction prod-
Experimental Section
3
uct was identified and quantified by NMR by comparison with published
spectra.
[
27,29,30]
General considerations: CH
cedures. Liquids for catalytic reactions were handles under a dinitrogen
atmosphere. [RhCl(PPh ], 5% Rh on carbon and 5% Rh on alumina
were purchased from Aldrich. [Rh and
2
Cl
2
was purified and dried by standard pro-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3 3
)
[
4
25]
2
A
T
N
T
E
N
G
2
A
T
G
E
N
G
A
H
U
G
R
N
N
2
-Cl)]BF
+
À[26]
2
ACHTUNGTRENNUNG[ NTf ]
[
Bmim]
[
18]
Acknowledgements
agents were used as commercially available. Entrapped catalysts 1, and
2
by Schleck and Sheldrake and prepared as previously reported. Surface
area was calculated by the BET method on a Micromeritics ASAP 2010.
The solid-state NMR was conducted by the EPSRC service at the Uni-
versity of Durham using a Varian UNITY Inova spectrometer with a
[
25]
were prepared as previously reported. Levoglucosenone was supplied
[
27]
We thank the EPSRC, CenTACat and the European Social Fund; Dr
David C. Apperley and the EPSRC solid state NMR service at the Uni-
versity of Durham; ASEP analytical services at Queenꢀs University Bel-
fast; Prof. Robert Mokaya, Dr. Mike Fay and Ignacio J. Villar Garcia at
The University of Nottingham; Prof. Brian Vincent and Ciarꢃn Martin at
The University of Bristol and Dr Gary Sheldrake and David Schleck at
Queenꢀs University Belfast.
7
.05 T (“300 MHz”) Oxford Instruments magnet. TEM were imaged on a
JEOL JEM-2000 FXII electron microscope operating at 200 keV. The
sample was suspended in methanol and 4 drops placed on a copper mesh
(
300 lines per inch mesh) and dried in air. XPS was carried out by the
Centre for Surface Chemical Analysis, University of Nottingham, UK
using a Kratos AXIS ULTRA with a mono-chromated Al ka X-ray
source (1486.6 eV) operated at 15 mA emission current and 10 kV anode
[1] J. Blum, D. Avnir, H. Schumann, CHEMTECH 1999, 29, 32–38.
[2] H. Sertchook, D. Avnir, J. Blum, F. Joç, A. Math, H. Schumann, R.
Weimann, S. Wernik, J. Mol. Catal. A 1996, 108, 153–160.
[3] A. Rosenfeld, J. Blum, D. Avnir, J. Catal. 1996, 164, 363–368.
[4] J. Blum, A. Rosenfeld, F. Gelman, H. Schumann, D. Avnir, J. Mol.
Catal. A 1999, 146, 117–122.
[
28]
potential - 150 W. Further details of the procedure and data are includ-
ed in the supplementary data. The ICP/OES samples were prepared as
[
25]
previously reported and run on a Perkin–Elmer Optima 4300DV ICP/
OES. The ICP was calibrated using 4 standards of corresponding metal
(
Rh or Ru): blank, 0.10 mg/L, 1.00 mg/L, 10.00 mg/L; and in conjunction
with a number of prepared standards ranging in concentration from
.01 mg/L to 10 mg/L. The catalytic tests were conducted using a Parr
[5] J. D. Ribeiro de Campos, R. Buffon, New J. Chem. 2003, 27, 446–
451.
[6] K. H. Park, S. U. Son, Y. K. Chung, Tetrahedron Lett. 2003, 44,
2827–2830.
[7] R. B. Pellegrino, R. Buffon, J. Braz. Chem. Soc. 2004, 15, 527–531.
[8] K. Hamza, R. Abu-Reziq, D. Avnir, J. Blum, Org. Lett. 2004, 6,
925–927.
0
3
Series 5500 high pressure stainless steel 50 cm Compact Laboratory Re-
actor stirred by a magnetically driven four blade impeller. Reaction time
was started once the reactor had reached the reaction temperature. The
GCMS were performed by ASEP, The Queenꢀs University of Belfast; the
Chem. Eur. J. 2009, 15, 7094 – 7100
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7099

A. C. Marr, P. C. Marr et al.
[
9] F. Gelman, J. Blum, D. Avnir, Angew. Chem. 2001, 113, 3759–3761;
Angew. Chem. Int. Ed. 2001, 40, 3647–3649.
[22] A. Riisagera, R. Fehrmann, M. Haumann, P. Wasserscheid, Top.
Catal. 2006, 40, 91–102.
[
10] F. Gelman, J. Blum, D. Avnir, J. Am. Chem. Soc. 2002, 124, 14460–
4463.
[23] V. I. Pꢅrvulescu, C. Hardacre, Chem. Rev. 2007, 107, 2615–2665.
[24] J. F. Young, J. A. Osborn, F. H. Jardine, G. J. Wilkinson, J. Chem.
Soc. Chem. Commun. 1965, 131–132.
[25] F. Lorenzini, K. T. Hindle, S. J. Craythorne, A. R. Crozier, F. Mar-
chetti, C. J. Martin, P. C. Marr, A. C. Marr, Organometallics 2006,
25, 3912–3919.
[26] P. Bonhꢆte, A. P. Dias, N. Papageorgiou, K. Kalyanasundaram, M.
Gratzel, Inorg. Chem. 1996, 35, 1168–1178.
[27] G. N. Sheldrake, D. Schleck, Green Chem. 2007, 9, 1044–1046.
[28] E. F. Smith, F. J. M. Rutten, I. J. Villar-Garcia, D. Briggs, P. Licence,
Langmuir 2006, 22, 9386–9392.
1
[
[
11] F. Gelman, J. Blum, D. Avnir, New J. Chem. 2003, 27, 205–207.
12] S. Dai, Y. H. Ju, H. J. Gao, J. S. Lin, S. J. Pennycook, C. E. Barnes,
Chem. Commun. 2000, 243–244.
[
[
[
13] C. J. Adams, A. E. Bradley, K. R. Seddon, Aust. J. Chem. 2001, 54,
6
79–681.
14] M.-A. Nꢄouze, J. Le Bideau, P. Gaveau, S. Bellayer, A. Vioux,
Chem. Mater. 2006, 18, 3931–3936.
15] H.-F. Wang, Y.-Z. Zhu, X.-P. Yan, R.-Y. Gao, J.-Y. Zheng, Adv.
Mater. 2006, 18, 3266–3270.
[
[
16] A. Karout, A. C. Pierre, J. Non-Cryst. Solids 2007, 353, 2900–2909.
17] K. Anderson, S. CortiÇas Fernꢃndez, C. Hardacre, P. C. Marr, Inorg.
Chem. Commun. 2004, 7, 73–76.
[29] K. Kadota, T. Kurusu, T. Taniguchi, K. Ogasawara, Adv. Synth.
Catal. 2001, 343, 618–623.
[30] E. Keinan, N. Greenspoon, J. Am. Chem. Soc. 1986, 108, 7314–
7325.
[
18] S. J. Craythorne, A. R. Crozier, F. Lorenzini, A. C. Marr, P. C. Marr,
J. Organomet. Chem. 2005, 690, 3518–3521.
[19] F. Shi, Q. Zhang, D. Li, Y. Deng, Chem. Eur. J. 2005, 11, 5279–5288.
[20] K. Hamza, J. Blum, Eur. J. Org. Chem. 2007, 4706–4710.
[21] C. P. Mehnert, Chem. Eur. J. 2005, 11, 50–56.
Received: September 2, 2008
Revised: March 19, 2009
Published online: June 16, 2009
7100
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7094 – 7100