Enantioselective hydrogenation of itaconic acid and its derivates with sol-gel immobilized Rh/BPPM catalysts

Article abstract of DOI:10.1016/j.molcata.2012.10.016

Itaconic acid and some of its derivates were hydrogenated with sol-gel entrapped Rh/BPPM catalysts in methanol solutions. The immobilization process was carried out by different gel building agents: hydrophilic tetramethyl orthosilicate Si(OMe)₄ (TMOS) and tetraethyl orthosilicate Si(OEt)₄ (TEOS), hydrophobic triethoxyphenylsilane PhSi(OEt) ₃/TMOS and trimethoxy(octyl)silane OcSi(OMe)₃/TMOS. The choice of the silane precursor influences the enantioselectivity and the rate of the reaction because of the hydrophobic interactions between catalyst, gel and substrate. The immobilized catalyst could be recovered and recycled several times under N₂-atmosphere. About 90-99% ee were achieved for the hydrogenation of itaconic acid to (S)-(+)-2-methyl succinic acid, and about 14% ee for the hydrogenation of dimethylitaconate to (S)-(+)-2-methyl-succinic acid dimethylester.

Full text of DOI:10.1016/j.molcata.2012.10.016

Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Enantioselective hydrogenation of itaconic acid and its derivates with sol–gel
immobilized Rh/BPPM catalysts
I. Volovycha, M. Schwarzea,∗, T. Hamerlaa, J. Blumb, R. Schomäckera
a Institut für Chemie, Technische Universität Berlin, Straße der 17. Juni 124, D-10623 Berlin, Germany
b Institute of Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904, Israel
a r t i c l e i n f o
a b s t r a c t
Article history:
Itaconic acid and some of its derivates were hydrogenated with sol–gel entrapped Rh/BPPM catalysts
in methanol solutions. The immobilization process was carried out by different gel building agents:
hydrophilic tetramethyl orthosilicate Si(OMe)4 (TMOS) and tetraethyl orthosilicate Si(OEt)4 (TEOS),
hydrophobic triethoxyphenylsilane PhSi(OEt)3/TMOS and trimethoxy(octyl)silane OcSi(OMe)3/TMOS.
The choice of the silane precursor influences the enantioselectivity and the rate of the reaction because
of the hydrophobic interactions between catalyst, gel and substrate. The immobilized catalyst could be
recovered and recycled several times under N2-atmosphere. About 90–99% ee were achieved for the
hydrogenation of itaconic acid to (S)-(+)-2-methyl succinic acid, and about 14% ee for the hydrogenation
of dimethylitaconate to (S)-(+)-2-methyl-succinic acid dimethylester.
Received 27 February 2012
Received in revised form 25 June 2012
Accepted 16 October 2012
Available online 22 October 2012
Keywords:
BPPM ligand
Heterogeneous catalysis
Hydrogenation
Chloro(1,5-cyclooctadiene) rhodium (I)
dimer
© 2012 Elsevier B.V. All rights reserved.
Sol–gel process
1. Introduction
solvents were already investigated by many research groups and
high enantioselectivities and fast reaction rates were obtained
The entrapment of organic and organometallic compounds into
sol gel matrices is an important technique for different applications,
e.g. photochemistry, chemical sensing and optics [1], biochemistry
and enzyme technology [2]. Furthermore, this technique allows for
immobilization of homogeneous catalysts in order to overcome the
major drawback of homogenous catalysis, the difficult separation of
the products and the catalyst after the reaction. Catalysts entrapped
in sol–gel matrices can be used in a variety of reactions: ligand-free
Heck and Suzuki couplings of aromatic compounds with Pd(OAc)2
catalyst [3,4], the RhCl3·H2O/Aliquat 336 catalyzed isomerization of
hydrophobic allylarenes [5], hydroformylation of styrene derivates
with [Rh(cod)Cl]2 [6] and disproportionation of dihydroarenes with
RhCl3·H2O/Aliquat 336 or Pd(OAc)2 catalysts [7]. The replacement
of organic solvents by aqueous solutions with different surfactants
(aqueous-micellar solutions or microemulsions) allows the more
environmentally friendly performance of different reactions [8].
In this contribution we investigate the enantioselective hydro-
genation of itaconic acid and some of its derivates by Rh/BPPM
catalysts entrapped within different sol–gel matrices (Scheme 1).
The homogeneously catalyzed hydrogenations of itaconic acid with
in situ generated Rh/BPPM catalyst in methanol and other organic
[9–11].
Also the hydrogenation of itaconic acid in aqueous solutions
and a mixture of aqueous-organic solvents [12] or the hydrogena-
tion of (Z)-methyl-␣-acetamidocinnamic acid in aqueous-micellar
solutions [13,14] were investigated earlier. High enantioselec-
tivities could be obtained for the hydrogenation of itaconic
acid and its derivates in aqueous-micellar solution with dif-
ferent homogeneous rhodium catalysts [15,16]. Because of the
use of surfactants or micelle forming agents, the catalysts could
be recovered by micellar enhanced ultrafiltration as shown by
Dwars et al. [17].
The immobilization of the in situ generated complexes within
sol–gel materials allows a better and easier separation of these
catalysts from solution and facilitates their reuse, which greatly
improves the productivity of the catalyst. Different mesoporous sil-
ica materials can be used as support materials in order to adapt the
catalyst to the substrate of the reaction [18–20].
The reactions are carried out in methanol as organic solvent
and in aqueous solutions that contain surfactants to solubilize the
hydrophobic reactants in water.
We study the influence of different parameters (temperature,
reaction media, hydrophobicity of the substrate and support mate-
rial, type of the surfactant and amount of the catalyst) on the
reaction rate and enantioselectivity in order to derive a recipe for
an optimal immobilized catalyst.
∗
Corresponding author. Tel.: +49 3031424097; fax: +49 3031421595.
E-mail address: ms@chem.tu-berlin.de (M. Schwarze).
1381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2012.10.016

360
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
CH2
O
Rh/BPPM@sol-gel
30°C
CH3
O
O
R
H2
O
R
+
R
O
R
O
O
O
Scheme 1. Hydrogenation of itaconic acid (R = H), dimethyl (R = CH3), diethyl (R = C2H5) and n-dibutyl (R = C4H9) itaconates.
2. Experimental
catalyst was formed in situ from 10 mg [Rh(cod)Cl]2 (0.022 mmol)
and 20 mg BPPM (0.044 mmol) dissolved in 0.78 mL tetrahydro-
furan.
2.1. Chemicals
The hydrolyzed TMOS solution was mixed with the activated
catalyst and 0.83 mL ammonia solution (0.1 N) to catalyze the gela-
tion process. The gelation occurred after 20–30 min.
The immobilization procedure using [Rh(nbd)Cl]2 as precursor
was the same. Because of lower solubility of this rhodium precursor
in tetrahydrofuran, it was solubilized in 10 mL methanol.
(2S,4S)-1-Tert-butoxycarbonyl-4-diphenylphosphino-2-
(diphenylphosphinomethyl)pyrrolidine
(BPPM),
tetraethyl
orthosilicate (TEOS), trimethoxyphenylsilane (PhSi(OMe)3),
triethoxy(octyl)silane (OcSi(OEt))3, itaconic acid (IA), tetramethyl
orthosilicate (TMOS), (trimethylsilyl)diazomethane solution,
phosphorus and rhodium ICP standard solutions were obtained
from Sigma–Aldrich company. Diethyl itaconate (DEI) stabi-
lized with trichlorobenzene (TCB) and dibutyl itaconate (DBI)
were acquired from TCI Europe. Dimethyl itaconate (DMI),
chloro(1,5-cyclooctadiene) rhodium(I) dimer [Rh(cod)Cl]2 and
chloronorbornadiene rhodium(I) dimer [Rh(nbd)Cl]2 were
purchased from ABCR GmbH & Co., Karlsruhe, Germany.
The structures of catalysts and ligand are shown in
Scheme 2.
2.2.2. Entrapment in hydrophilic support
The entrapment of the Rh/BPPM catalyst in the modified support
could be realized with TEOS or a mixture of TMOS and TEOS under
N2-atmosphere. After the hydrolysis of 1.58 mL TEOS (9.88 mmol)
in 5.6 mL ethanol and 0.4 mL of distilled water for 24 h, the solution
was added to the catalyst solution (prepared similarly as described
in Section 2.2.1). Then a small amount of conc. triethylamine (3–5
drops) was added to the mixture to catalyze the gelation process.
The gelation of the catalyst occurred after about 24 h.
2.2. Immobilization of the rhodium catalyst
For the preparation of the mixed TMOS/TEOS immobilized cat-
alyst, the hydrolyzed TMOS solution (3.6 mL TMOS (23.0 mmol),
2.4 mL methanol (94.8 mmol) and 2.0 mL water (111 mmol)) was
added to the hydrolyzed TEOS solution.
The general procedure for the entrapment of the Rh/BPPM cat-
alyst using different gel building agents is shown in Scheme 3. In
every case, after preparation of the catalysts by the different proce-
dures described below (a–c), the catalysts were dried for 24 h in a
vacuum oven under a reduced pressure of 1000 Pa and a tempera-
ture of 30 ◦C, washed carefully 3 times with 10 mL boiling water and
dried again. The washing liquids were analyzed for their rhodium
and phosphorous content (catalyst leaching) with ICP-OES.
2.2.3. Entrapment in hydrophobic gel
The procedure for the immobilization of the catalyst in
hydrophobic supports is comparable to the immobilization of a
palladium catalyst for the Heck Coupling reaction as shown by
Rozin-Ben Baruch et al. [3]. A mixture of 2.1 mL octyl trimethoxy
silane (9.88 mmol) or 1.61 mL phenyltriethoxysilane (6.68 mmol)
was stirred for 24 h in 4.2 mL ethanol and 0.4 mL distilled water.
Then the hydrolyzed tetramethoxy orthosilicate solution was
added (Section 2.2.2) and stirred for 20 min. The combined solu-
tions were mixed with Rh/BPPM in tetrahydrofuran (prepared
similarly as described in Section 2.2.1) and the gelation was com-
pleted within 4–8 days.
2.2.1. Hydrochloric acid catalyzed entrapment in hydrophilic gel
The two step entrapment of the catalyst was carried out under
N2 atmosphere using the procedure described by Gelman et al. [21].
A mixture of 2.50 mL TMOS (16.92 mmol) in 3.50 mL methanol and
0.78 mL hydrochloric acid (prepared from 0.1 mL of 1.1 N HCl and
9.9 mL H2O) was stirred for 10 min at 25 ◦C. Separately the Rh/BPPM
Scheme 2. Structures of catalyst precursors and enantioselective ligand.

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
361
1) [Rh(cod)Cl]2 +BPPM+THF
EtOH/
H2O
(b) Rh/BPPM
@TEOS-derived
silica
hydrolized
TEOS
TEOS
2) NEt3 (24 h)
1) [Rh(cod)Cl]2+BPPM
+THF
(a)Rh/BPPM
@TMOS-derived
silica
hydrolized
TMOS
HCl/
MeOH
2) NH3(30 min)
[Rh(cod)Cl]2
+BPPM+THF
TMOS/
hydrolized
TEOS
(b) Rh/BPPM@
TMOS/TEOS-
derived silica
TMOS
TEOS
(TEOS/H2O/
EtOH)
(3-6 d)
MeOH/
H2O
hydrolized
PhSi(OEt)3
or OcSi(OMe)3
[Rh(cod)Cl]2
(c) Rh/BPPM@
OcTMOS- or
PhTMOS-derived
silica
+BPPM+THF
PhTMOS
hydrolized
TMOS
or
OcTMOS
(H2O,EtOH)
(4-8 d)
Scheme 3. Entrapment of the Rh/BPPM catalyst with different gel building agents (PhTMOS or OcTMOS = hydrolyzed TMOS + PhSi(OEt)3 or OcSi(OMe)3 solution).
2.3. Hydrogenation procedure
analyzed on a PC, from these results the substrate concentration
csubstrate and the conversion X were calculated.
The set up of the hydrogenation apparatus [22,23] is shown in
Scheme 4. About 1.0–1.4 g of the yellow-orange sol gel immobi-
lized catalyst, a desired amount of substrate and 95 mL of solvent
(methanol, water or aqueous-micellar solution) were added to a
stirred tank reactor and stirred at 400 rpm under N2-atmosphere
at the desired reaction temperature. The N2 was replaced by H2
(p = 1.1 × 105 Pa) without stirring and the reaction was started by
turning the stirrer on 800 rpm again. The reactions were performed
in semi-batch mode with hydrogen being permanently added to
the stirred tank reactor to achieve a constant total pressure of
1.1 × 105 Pa in reactor. The cumulative hydrogen consumption and
the pressure during the reaction were recorded using a Bronkhorst
flow meter and pressure controller (Bronkhorst Mättig GmbH,
Kamen, Germany), respectively (see Scheme 4). The sensitivity
of measured hydrogen consumption was 0.4 mL. The pressure
and the hydrogen flow during the reaction were registered and
A typical diagram of the experimental data obtained for the
hydrogenation of itaconic acid with a homogeneous Rh/BPPM cat-
alyst, is shown in Fig. 1.
After the reaction, the solution was analyzed by GC to deter-
mine the conversion and the selectivity and by ICP to check
for catalyst leaching. For recycling experiments, the immobilized
catalyst was left in the reactor under N2 atmosphere after the
reaction and reused directly again in several runs by adding new
substrate.
2.4. Instruments
The separation of catalyst and silica support was carried
out by microwave decomposition (p = 20 × 105 Pa, t = 35 min and
T = 200 ◦C) with a CEM Discover SP-D (sample preparation – diges-
tion) instrument (CEM GmbH, Camp-Lintfort, Germany). Before
this procedure, the catalysts were ball milled, and then mixed
with 12 mL of an HNO3/HCl/H2SO4 solution (ratio: 2eq/6eq/4eq).
After the microwave treatment, the white solid silica was removed
by filtration and the solution of detached catalyst was ana-
lyzed for rhodium and phosphorus content using a Varian 715-ES
Optical Emission Spectrometer (ICP-OES). Calibration of the instru-
ment was performed with commercial rhodium and phosphorus
standards.
The reaction conversion and the enantiomeric excess of the
hydrogenation of DMI were obtained by gas chromatography using
a Hewlett–Packard model HP 5710 equipped with a Lipodex E
column (Macherey-Nagel GmbH & Co. KG, Düren, Germany). The
following conditions were used for the analysis: Tinjector = 200 ◦C,
Tdetector = 250 ◦C, Toven = 80 ◦C and pcolumn = 60 × 103 Pa, carrier
gas = N2. For itaconic acid, before analysis the sample was treated
with (trimethylsilyl) diazomethane and then analyzed the same
way as described for DMI. For DEI and DBI, the conversion was
calculated from the GC results.
N2-BET specific surface area measurements of sol–gel immo-
bilized rhodium catalysts were obtained by
a Micromeritics
Gemini 1325 instrument. Transmission electron microscopy was
performed with a conventional LaB6-TEM Tecnai G220 S-TWIN
Scheme 4. Hydrogenation set up (Pl, pressure indicator; PC, pressure controller;
FM, flow meter; T, thermostat).

362
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
Table 1
Specific surface area A of immobilized Rh/BPPM catalysts and support materials.
A (m2/g)
TMOS-derived silica
TEOS-derived silica
Rh/BPPM@TMOS-derived silica
Rh/BPPM@TEOS-derived silica
Rh/BPPM@PhTMOS-derived silica
Rh/BPPM@OcTMOS-derived silica
397
467
383
225
327
327
support materials were obtained from BET measurements and the
results are given in Table 1.
After immobilization of the Rh/BPPM catalyst, the specific
surface area decreases slightly from 397 m2/g to 383 m2/g, if
TMOS-derived silica is used and to 327 m2/g, if PhTMOS- or
OcTMOS-derived silica’s are used. For TEOS-derived silica we
observed a stronger decrease in the specific surface area. A rea-
son for the decreased specific surface area could be a blocking of
the smaller pores in the support material by the immobilized cat-
alysts. The average pore sizes of the catalysts were between 2.1
and 3.2 nm [5,7,26]. The TEM images of tetramethyl orthosilicate
immobilized rhodium catalyst showed in Fig. 2a indicate that the
catalysts used in the hydrogenation reaction of itaconic acid and
derivates are highly porous. Their average pore size is 1–2 nm and
the metal complexes are intercalated in the pore walls of the support
material, but cannot be indicated in Fig. 2a because of low loading
of the support with rhodium. However the energy dispersive X-ray
spectroscopy measurement EDX (Fig. 2b) showed small amounts of
metallic rhodium nanoparticles and large amounts of SiO2 support
in this sample.
3.2. Homogeneous vs. heterogeneous hydrogenation reaction
Fig. 1. Reaction profiles for the homogeneous hydrogenation of itaconic acid in
The hydrogenation of itaconic acid was carried out with a homo-
geneously dissolved Rh/BPPM catalyst and with an immobilized
Rh/BPPM catalyst on TMOS derived silica. The reaction rate for the
sol–gel immobilized Rh/BPPM catalyst is slower than for the homo-
geneous catalyst in methanol (Fig. 3), but the same enantiomeric
excess is obtained (ee = 90–97%). The turnover frequency (TOF) for
Rh/BPPM catalyst on TMOS derived silica was ∼2400 h–1 and about
3 times lower than for the homogeneous catalyst (TOF ∼ 6100/h),
because the reaction rate can be influenced by mass-transport lim-
itations within the heterogeneous sol–gel material or a decrease
in activity of the catalyst due to its encapsulation inside the silica
material.
methanol: (a) cumulative hydrogen consumption VH and hydrogen flow (dV/dt);
2
(b) conversion X and substrate concentration cIA. Reaction conditions: 15.4 mmol
substrate, 0.022 mmol [Rh(cod)Cl]2, 0.044 mmol BPPM, 100 mL solvent, 1.1 × 105 Pa
H2, 800 rpm, 30 ◦C.
instrument (FEI Company, USA) operated at 200 kV and equipped
with EDAX-EDS for identification of elemental compositions.
3. Results and discussion
3.1. Catalyst preparation and characterization
3.3. Temperature dependence of the hydrogenation reaction
The heterogeneously catalyzed hydrogenations of itaconic acid
and derivates were realized with sol gel immobilized and in situ
generated Rh/BPPM catalysts. The catalyst complex was obtained
by ligand exchange of cyclooctadiene (cod) or norbornadiene (nbd)
ligand in rhodium precursors by the enantioselective pyrrolidine
ligand BPPM. The use of the rhodium precursor containing norbor-
nadiene ligand is preferred in hydrogenation reactions because of
a faster ligand exchange rate [24,25].
The catalysts were prepared as described in Section 2 and
after successful entrapment of the rhodium into the silica
matrix, the support changes its color from white to yellow (see
Fig. S1 and Fig. S2 in supporting information). After the synthesis,
the washing liquids were analyzed by ICP-OES and the contents of
phosphorus and rhodium were determined to be <0.02 mg rhodium
(0.5% leaching) and <0.02 mg phosphorus (1.5% leaching) in 10 mL
of water, respectively.
To check for mass transport limitation, the temperature depend-
ence of the hydrogenation reaction was studied (Fig. 4). The
reaction rate increases with increasing temperature as expected.
From the slope of the regression line in the Arrhenius plot, the acti-
vation energy EA of the chemical reaction was calculated (Table 2).
For the hydrogenation of itaconic acid in methanol solution
the reaction rate is higher in water due to the higher solubil-
ity of hydrogen in methanol. Also the hydrogenation reaction
using the catalyst with 1:2 [Rh(cod)Cl]2:BPPM molar ratio is much
faster than with 1:4 molar ratio. The activation energies EA of the
hydrogenation with heterogeneous catalyst in methanol and also
in water or aqueous-micellar solutions (Table 2) is comparable
with the activation energies when using a homogeneous catalyst
(1 Rh(cod)SO3CF3:2BPPM) [22]. That means that the heteroge-
neously catalyzed hydrogenation is not mass transport limited
and the decrease of the reaction rate can be explained by the
part deactivation of the catalyst or encapsulation in silica mate-
rial. An influence of pore diffusion on the reaction rate would be
The loading of the heterogeneous catalysts were about 2–3 wt%
Rh/BPPM, with 0.1–0.2 wt% rhodium and the average particle size
was 200–500 ␮m. The specific surface areas of the catalyst and the

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
363
Fig. 2. (a) TEM image and (b) EDX spectra of typical TMOS based supported Rh/BPPM catalyst after the hydrogenation of itaconic acid in methanol.
indicated by a strong decrease EA. Also the variation of stirring
rates between 600 and 1200 rpm in homogeneous and heteroge-
neous reactions shows no increase of reaction rate with increasing
stirring intensity. That means the hydrogenation is not limited
through the transport between the gas and the liquid phase. Only
the reactions with stirring rates below 600 rpm are transport
limited.
All following hydrogenation reactions were realized at 30 ◦C,
were the reaction proceeds with a reasonable rate, due to good
catalyst activity and H2 solubility.

364
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
Table 2
The activation energies for the hydrogenation of itaconic acid.
a
b
[Rh(cod)Cl]2:ligand ratio
Solvent
EA, heterogeneous (kJ/mol)
eeheterogeneous (%)
EA,homogeneous (kJ/mol)
eehomogeneous (%)
1:4
1:2
1:4
Methanol
Methanol
Water
33.4
50.5
24.4
98
98
37
49.9
33c
95
38c
a
Rh(cod)2SO3CF3/BPPM [22].
[Rh(cod)Cl]2/BPPM.
H2O/SDS.
b
c
Fig. 3. Comparison between homogeneous and heterogeneous hydrogenation of
itaconic acid. Reaction conditions: 15.4 mmol substrate, 0.022 mmol [Rh(cod)Cl]2,
0.044 mmol BPPM, 100 mL methanol, 1.1 × 105 Pa H2, 800 rpm, 30 ◦C.
Fig. 5. Substrate variation in the hydrogenation with Rh/BPPM on TMOS-derived
silica in methanol: itaconic acid (IA), dimethyl itaconate (DMI), diethyl itaconate
(DEI), dibutyl itaconate (DBI). Reaction conditions: 15.4 mmol substrate, 0.022 mmol
[Rh(cod)Cl]2, 0.044 mmol BPPM, 100 mL solvent, 1.1 × 105 Pa H2, 800 rpm,
30 ◦C.
3.4. Substrate variation
The enantioselective Rh/BPPM@TMOS-gel catalyzed hydro-
genation reaction can be performed with different itaconic acid
derivates. The cumulative hydrogen consumption for the hydro-
genation of IA, DMI, DEI and DBI is shown in Fig. 5. For IA, the
reaction is very fast and full conversion is achieved within 25 min.
For DMI and DEI, full conversion is achieved only within 50–80 min,
but interestingly, the reaction with DEI is faster than with DMI. For
DBI, the conversion after 70 min is only 70%. From these results it is
clear that the conversion of the reaction increases with decreasing
hydrophobicity of the substrates (decreasing C-chain length). The
comparison to the same reactions with a homogeneous catalyst
also shows that the difference is caused by the reactivity of the
substrates. In order to examine the influence of the interactions
between the reactant and the catalyst support material a surface
modification with hydrophobic functional groups is introduced in
the next section.
3.5. Influence of the support materials
Besides the hydrophobic character of the substrate, the polar-
ity of the catalyst surface also influences the rate of the
hydrogenation reaction. The catalysts were prepared with dif-
ferent hydrophilic and hydrophobic sol gel building agents:
tetramethyl orthosilicate (TMOS), tetraethyl orthosilicate (TEOS),
hydrophobically modified TMOS (Ph/TMOS and Oc/TMOS). The
specific surface areas for the immobilized catalysts and sup-
port materials are shown in Table 1. The specific surface areas
of the different support materials, except for TMOS, are about
350 m2/g.
The hydrogenations of IA, DMI and DBI were carried out
with sol–gel supported Rh/BPPM catalysts with these different
hydrophobic surface modifications and the reaction profiles are
shown in Fig. 6.
The reaction rate with more hydrophilic substrates (IA, DMI)
increases if the hydrophilic surfaces are used (TMOS, TEOS). The
reaction rate was quite the same in the hydrogenation of more
hydrophobic DBI with the catalyst with hydrophobic support. The
use of Rh/BPPM catalyst immobilized in hydrophobically modified
silica is preferred in reactions with more hydrophobic substrates
for example DBI because of hydrophobic interaction between the
Fig. 4. Arrhenius plots of the hydrogenation of itaconic acid in water and methanol
at 1:2 and 1:4 Rh:BPPM ratio. Reaction conditions: 15.4 mmol substrate, 0.022 mmol
[Rh(cod)Cl]2, 0.044 mmol BPPM, 100 mL solvent, 1.1 × 105 Pa H2, 800 rpm.

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
365
3.6. Use of different reaction media
Most homogeneously catalyzed reactions are carried out in
conventional organic solvents, but the use of water as solvent
allows more environmentally friendly processes. The main prob-
lem using water, as reaction medium is that many substrates are
hydrophobic and cannot be solubilized completely in water. The
addition of surfactants with concentrations higher than the critical
micelle concentration (cmc) results in the formation of micelles
able to solubilize the substrates, depending on the partition
coefficient, in their cores (hydrophobic) or in the palisade layer
(hydrophilic) [27]. By increasing the surfactant concentration to
higher amounts, the solubility of the substrate can be improved.
In Fig. 7a the reaction profiles for the hydrogenation of itaconic
acid in methanol, water and in aqueous-micellar solutions of
different surfactants are shown. As expected, because of the higher
solubility of hydrogen (4–5 times higher), the reaction is faster
in methanol than in the aqueous media, but the influence of the
surfactant on the hydrogenation rate of this hydrophilic substrate
is low. In the aqueous-micellar solutions the hydrogenation reac-
tions proceed with a similar rate as in water, because of the high
solubility of itaconic acid in water (ca. 40 g/L). For the solubilization
and hydrogenation of IA the surfactant is not required. For the
hydrogenation of DMI, the use of more hydrophobic nonionic
Fig. 6. Cumulative hydrogen volume for the hydrogenation of (a) itaconic acid,
(b) dimethylitaconate, (c) dibutylitaconate in methanol with Rh/BPPM catalyst@
different support materials. Reaction conditions: 15.4 mmol substrate, 0.022 mmol
[Rh(cod)Cl]2, 0.044 mmol BPPM, 100 mL methanol, 1.1 × 105 Pa H2, 800 rpm,
30 ◦C.
substrate and the catalyst surface. The same effect was seen in
a Heck reaction with Pd(OAc)2 and aromatic substrates [3]. The
advantages of the hydrophobically modified surfaces were already
shown in literature; the catalysts formed on these surfaces are more
stable (no leaching into the solution) because of more branched
pore structure [2].
Fig. 7. Cumulative hydrogen volume for the hydrogenation of (a) itaconic acid and
(b) dimethyl itaconate in methanol, water and aqueous-micellar solutions. Reac-
tion conditions: 15.4 mmol substrate, 0.022 mmol [Rh(cod)Cl]2, 0.044 mmol BPPM,
100 mL solvent, 1.1 × 105 Pa H2, 800 rpm, 30 ◦C.

366
I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
Table 3
Conversions, enantiomeric excess (ee) and turnover frequencies (TOF) for hydrogenation reactions in methanol and in aqueous-micellar solutions with homogeneous and
TMOS supported Rh/BPPM catalysts, respectively. a
Entry
Catalyst
Substrate
X (%)c
ee (%)c
TOF (1/h)b
Support
Solvent
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(nbd)Cl]2
[Rh(nbd)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(nbd)Cl]2
[Rh(nbd)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
[Rh(cod)Cl]2
IA
IA
IA
IA
IA
IA
IA
IA
IA
IA
IA
IA
DMI
DMI
DMI
DMI
DMI
DMI
DMI
DMI
DMI
100
100
100
100
25
99
98
99
98
37
35
2
95
23
88
30
2
6132
2390
3947
4426
126
201
111
377
163
565
195
251
592
398
964
369
121
464
235
508
300
–
SiO2
–
SiO2
SiO2
SiO2
SiO2
–
SiO2
–
SiO2
SiO2
–
SiO2
–
SiO2
SiO2
–
Methanol
Methanol
Methanol
Methanol
Water
44
H2O/methanol
H2O/SDSd
H2O/CTABe
H2O/CTABe
H2O/TX-100d
H2O/TX-100d
Microemulsionf
Methanol
Methanol
Methanol
Methanol
Water
100
100
100
100
45
100
100
56
100
100
53
79
62
61
12
63
5
2
20
10
33
10
H2O/SDSd
H2O/SDSd
H2O/TX-100d
H2O/TX-100d
SiO2
–
SiO2
96
63
a
Reaction conditions: 15.4 mmol substrate, 0.022 mmol [Rh(cod)Cl]2, 0.044 mmol BPPM, 100 mL solvent, 1.1 × 105 Pa H2, 800 rpm, 30 ◦C.
b
c
d
e
f
Rh content was obtained from ICP measurements, the turnover frequency is calculated as follows: TOF = (nProduct/nRh) × t.
All results were obtained from GC measurements.
10 g/L surfactant.
1 g/L surfactant.
The used microemulsion consists of 77 wt% cyclohexane, 5 wt% water, 9 wt% TX-100 and 9 wt% pentanol [14].
Triton X-100 (TX-100) surfactant (HLB = 13.5) or cationic
transport), which has already been reported in [8] for the Suzuki
and Heck coupling reactions with sol–gel immobilized palladium
catalysts in microemulsions.
Some results for the hydrogenation of itaconic acid
with similar catalysts from the literature are shown in
Table 4.
The conversions and enantioselectivities obtained for hydro-
genations of itaconic acid and its derivates with sol–gel
immobilized Rh/BPPM catalysts in methanol solution are compa-
rable with results in literature obtained with rhodium catalysts
immobilized on mesoporous silica. Some of the catalyst could be
recycled about 4–8 times without lost in activity and enantioselec-
tivity [32–34]. The hydrogenation reactions in aqueous solutions
are not very common. Here the main challenge is still to obtain
high enantioselectivity [29,30].
cetyltrimethyl-ammonium bromide CTAB (HLB = 12) with lower
HLB value (hydrophilic lipophilic balance) increases the reaction
rate compared to the reaction in water or aqueous solution of
hydrophilic SDS surfactant (HLB = 40).
The reaction rates of the hydrogenation of dimethyl itaconate
in methanol and in aqueous-micellar solution with TX-100 sur-
factant are comparable. But the main disadvantage of using of
micellar solutions for this reaction with heterogeneous catalyst is
the decreased enantioselectivity (Table 3). Only the reactions in
micellar solutions with homogeneous catalysts proceed with high
enantioselectivities (entries 8 and 10), because the immobilized
rhodium complexes are not protected against contact with water
by embedding them in the hydrophobic cores of the micelles. This
was also shown to be important in a variety of other hydrogenation
reactions [13,14,28].
The same phenomenon was reported by Jamis et al. [29,30] for
the hydrogenation of ˛-acetamidocinnamic acid and itaconic acid
with catalyst immobilized on sol–gel material or hexagonal meso-
porous silica (HMS) in water.
In our process the addition of methanol to an aqueous solution
(1:1) (entry 6) or the use of cyclohexane/TX-100/water/1-pentanol
microemulsion (entry 12), reported by [23] as alternative reaction
media, has increased the reaction rate because more organic sol-
vent was involved in the reaction and increased the solubility of
the hydrogen, but it did not improve the enantioselectivity. A suc-
cessful application of this concept is an EST process (emulsion-solid
3.7. Catalyst recycling
The replacement of homogeneous catalysts by sol–gel immobi-
lized catalysts allows their reuse and also decreases the costs of the
process. In case of an air sensitive in situ formed Rh/BPPM complex
the recycling procedure of the catalyst is not very easy. As can be
seen in Fig. 8a and b the catalyst could be immobilized in a gel from
tetramethyl orthosilicate and tetraethyl orthosilicate as gel build-
ing agent, and recycled 3 and 4 times with only small loss in activity
and enantioselectivity due to the leaching and catalyst deactiva-
tion. About 0.1 mg Rh (4 wt%) and about 0.4 mg P (30 wt%) were
Table 4
Hydrogenation of itaconic acid and derivates with immobilized catalysts (literature).
Substrate
Catalyst
Support
Solvent
X (%)
t (h)
ee (%)
Recycling times
3
Reference
IA
IA
DMI
DMI
[Rh(cod)Cl]2/(2S,4S)-BPPM
[RhEt2Cl]2/(EtO)3-Si-CH2-PMen2
[Rh(cod)2]+ BF4−/(R,R-MeDuPhos)
RhDuphos/RhBPE
[Rh(nbd)Cl]2/(R,R)-BDPBzPSO3
[Rh(cod)Cl]2-pincer complex
Ordered mesoporous silica (HMS)
Phosphinated silica (Kieselgel 100)
Al-MCM-41
Al-MCM-41/Al-MCM-48/Al-SBA-15
MCM-41
H2O
100
67
99
100
100
96
46
20
1
24
24
2
23
83
99
82–99
53
[29,30]
[31]
[32]
[34]
[20]
Methanol
Methanol
Methanol
Methanol
Ethanol
8
4
DMI
DEI (diethyl itaconate)
MCM-41
99
4
[33]

I. Volovych et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 359–367
367
surface of the catalyst. The immobilized catalyst could be recycled
several times with only minor decrease in activity and selectiv-
ity caused by deactivation and formation of black inactive Rh(0)
particles.
Acknowledgment
We gratefully acknowledge the financial support of this
study by Deutsche Forschungsgemeinschaft (DFG) through grant
SCHO687/8–1.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.molcata.2012.10.016.
References
[1] C. Rottman, G. Grader, Y. De Hazan, S. Melchior, D. Avnir, J. Am. Chem. Soc. 121
(1999) 8533–8543.
[2] H. Frenkel-Mullerad, D. Avnir, J. Am. Chem. Soc. 127 (2005) 8077–8081.
[3] A. Rosin-Ben Baruch, D. Tsvelikhovsky, M. Schwarze, R. Schomäcker, M. Fanun,
J. Blum, J. Mol. Catal. A: Chem. 323 (2010) 65–69.
[4] D. Tsvelikhovsky, J. Blum, Eur. J. Org. Chem. 2008 (2008) 2417–2422.
[5] D. Meltzer, D. Avnir, M. Fanun, V. Gutkin, I. Popov, R. Schomäcker, M. Schwarze,
J. Blum, J. Mol. Catal. A: Chem. 335 (2011) 8–13.
[6] K. Hamza, H. Schumann, J. Blum, Eur. J. Org. Chem. 2009 (2009) 1502–1505.
[7] T. Yosef, R. Schomäcker, M. Schwarze, M. Fanun, F. Gelman, J. Mol. Catal. A:
Chem. 351 (2011) 46–51.
[8] R. Abu-Reziq, D. Avnir, J. Blum, Angew. Chem. Int. Ed. 41 (2002) 4132–4134.
[9] K. Achiwa, J. Am. Chem. Soc. 98 (1976) 8265–8266.
[10] H. Brunner, E. Graf, W. Leitner, K. Wutz, Synthesis (1989) 743–745.
[11] H. Brunner, W. Leitner, Angew. Chem. Int. Ed. Engl. 27 (1988) 1180–1181.
[12] F. Joó, Á Kathó, J. Mol. Catal. A: Chem. 116 (1997) 3–26.
[13] I. Grassert, E. Paetzold, G. Oehme, Tetrahedron 49 (1993) 6605–6612.
[14] R. Selke, J. Holz, A. Riepe, A. Börner, Chem. Eur. J. 4 (1998) 769–771.
[15] M. Schwarze, J.S. Milano-Brusco, V. Strempel, T. Hamerla, S. Wille, C. Fischer,
W. Baumann, W. Arlt, R. Schomäcker, RSC Adv. 1 (2011) 474–483.
[16] J.S. Milano-Brusco, H. Nowothnick, M. Schwarze, R. Schomäcker, Society 49
(2010) 1098–1104.
[17] T. Dwars, J. Haberland, I. Grassert, G. Oehme, U. Kragl, J. Mol. Catal. A: Chem.
168 (2001) 81–86.
Fig. 8. Recycling of the catalyst after the hydrogenation of itaconic acid in methanol:
(a) Rh/BPPM on TMOS-derived silica and (b) Rh/BPPM on TEOS-derived silica. Reac-
tion conditions: 15.4 mmol substrate, 0.022 mmol [Rh(cod)Cl]2, 0.044 mmol BPPM,
100 mL solvent, 1.1 × 105 Pa H2, 800 rpm, 30 ◦C.
[18] C.E. Song, S.-gi Lee, Chem. Rev. 102 (2002) 3495–3524.
[19] S. Sahoo, A. Bordoloi, S.B. Halligudi, Catal. Surv. Asia 15 (2011) 200–214.
[20] C. Bianchini, P. Barbaro, Top. Catal. 19 (2002) 17–32.
[21] F. Gelman, D. Avnir, H. Schumann, J. Blum, J. Mol. Catal. A: Chem. 146 (1999)
123–128.
[22] N. Weitbrecht, M. Kratzat, S. Santoso, R. Schomäcker, Catal. Today 80 (2003)
leached into the solution after the reaction. One part of the catalyst
is converted to small black inactive particles, which decrease the
activity, reaction rate and enantioselectivity. All recycling experi-
ments were done under N2 atmosphere to avoid the deactivation
of the catalyst by the oxygen from air.
401–408.
[23] J.S. Milano-Brusco, M. Schwarze, R. Schoma¨cker, Ind. Eng. Chem. Res. 47 (2008)
7586–7592.
[24] A. Preetz, H.-J. Drexler, C. Fischer, Z. Dai, Chem. Eur. J. 14 (2008) 1445–1451.
[25] A. Togni, C. Breutel, A. Schnyder, F. Spindler, H. Landert, A. Tijani, J. Am. Chem.
Soc. 116 (1994) 4062–4066.
[26] K. Hamza, J. Blum, Eur. J. Org. Chem. 2007 (2007) 4706–4710.
[27] K.T. Valsaraj, A. Gupta, L.J. Thibodeaux, D.P. Harrison, Water Res. 22 (1988)
1173–1183.
4. Conclusion
[28] A. Kumar, G. Oehme, J.P. Roque, M. Schwarze, R. Selke, Angew. Chem. Int. Ed.
Engl. 21 (1994) 2197–2199.
[29] J. Jamis, J.R. Anderson, R.S. Dickson, E.M. Campi, W.R. Jackson, J. Organomet.
Chem. 603 (2000) 80–85.
[30] J. Jamis, J.R. Anderson, R.S. Dickson, E.M. Campi, W.R. Jackson, J. Organomet.
Chem. 627 (2001) 37–43.
[31] A. Kinting, H. Krause, M. Capka, J. Mol. Catal. 33 (1985) 215–223.
[32] W.P. Hems, P. McMorn, S. Riddel, S. Watson, F.E. Hancock, G.J. Hutchings, Org.
Biomol. Chem. 3 (2005) 1547–1550.
The enantioselective hydrogenation of itaconic acid and
derivates can be realized in organic solvents like methanol or
in aqueous-micellar solutions of different surfactants. The reac-
tion in micellar solutions is more environmentally friendly but
tends toward lower enantioselectivities due to the poor interaction
between the catalyst and the substrate. Also the choice of different
support materials influences the performance of the catalysts. This
phenomenon could be explained by hydrophobic and hydrophilic
interactions between the substrate and the hydrophobic modified
[33] C. Pozo, A. Corma, M. Iglesias, F. Sánchez, Organometallics (2010) 4491–4498.
[34] A. Crosman, W. Hoelderich, Catal. Today 121 (2007) 130–139.