Catalytic isomerization of hydrophobic allylarenes in aqueous microemulsions

Article abstract of DOI:10.1016/j.molcata.2010.12.002

In the course of our attempts to replace the traditional but environmentally disfavored organic solvents in organic processes by water, we studied the double bond isomerization of hydrophobic allylarenes, in aqueous microemulsions. The catalyst for these reactions was the rhodium-trichloride- Aliquat 336 ion pair encaged within hydrophobic silica sol-gel. During the entrapment of the rhodium compound, it was converted into supported catalytically active Rh(0) nanoparticles characterized by TEM and XPS studies. The transformations of the allylarenes to (E)- and (Z)-1-phenyl-1-propenes follows the first order rate law, and proved to be significantly affected by the electronic nature of the substrates and by the hydrophobicity of the sol-gel support. Upon completion of the isomerization the catalyst could be recovered and recycled at least six times without a decrease in the catalytic activity after the first catalytic run.

Full text of DOI:10.1016/j.molcata.2010.12.002

Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Editor’s choice paper
Catalytic isomerization of hydrophobic allylarenes in aqueous microemulsions
Diana Meltzera, David Avnira, Monzer Fanunb, Vitaly Gutkinc, Inna Popovc,
Reinhard Schomäckerd, Michael Schwarzed, Jochanan Bluma,∗
a Institute of Chemistry, The Hebrew University, Jerusalem 91904, Israel
b Center for Colloids and Surface Research, Al-Quds University, East Jerusalem 51000, Palestinian Authority
c Center for Nanoscience and Nanotechnology, The Hebrew University, Jerusalem 91904, Israel
d Institut für Chemie, Technische Universität Berlin, Strasse des 17. Juni 124, D-10623 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
In the course of our attempts to replace the traditional but environmentally disfavored organic solvents
in organic processes by water, we studied the double bond isomerization of hydrophobic allylarenes, in
aqueous microemulsions. The catalyst for these reactions was the rhodium-trichloride-Aliquat® 336 ion
pair encaged within hydrophobic silica sol–gel. During the entrapment of the rhodium compound, it was
converted into supported catalytically active Rh(0) nanoparticles characterized by TEM and XPS studies.
The transformations of the allylarenes to (E)- and (Z)-1-phenyl-1-propenes follows the first order rate law,
and proved to be significantly affected by the electronic nature of the substrates and by the hydrophobicity
of the sol–gel support. Upon completion of the isomerization the catalyst could be recovered and recycled
at least six times without a decrease in the catalytic activity after the first catalytic run.
© 2010 Elsevier B.V. All rights reserved.
Received 1 September 2010
Received in revised form
18 November 2010
Accepted 5 December 2010
Available online 14 December 2010
Keywords:
Heterogeneous catalysis
Isomerization
Microemulsion
Rhodium
Sol–gel processes
Sustainable chemistry
1. Introduction
2. Experimental
In the course of the attempts to replace harmful solvents used in
organic synthesis by water [1,2] we have developed a three phase
emulsion/solid state transport system (EST) that enables us to per-
form hydrogenation [3,4], hydroformylation [5] and C–C coupling
reactions of water insoluble substrates in aqueous media by recy-
clable transition metal catalysts [6,7]. We now find that the EST
system is also applicable to catalytic isomerization of hydropho-
bic olefins. A variety of industrially important processes involve
double bond migration reactions [8]. Examples are the synthe-
ses of various fatty acids, terpenes, alkaloids, and steroids. Double
bond isomerization is also a key step in the higher olefin man-
ufacturing processes and in the fabrication of various fragrances
[9]. As a model reaction we chose the double bond migration in
allylarenes that gives ultimately equilibrium mixtures of (E)- and
(Z)-1-propenylarenes shown in Scheme 1. Our studies revealed
several differences between the isomerization pathways in some
organic solvents and in aqueous microemulsions.
2.1. Instruments
NMR spectra were recorded with either a Bruker DRX-400 or
a Bruker Avance II-500 in CDCl3, Infrared spectra were obtained
with a Bruker Vector 22-FTIR spectrometer. Mass spectral mea-
surements were performed with a Q-TOF-II spectrometer. ICP-MS
analyses were performed with a Perkin-Elmer model ELAN DRC
II instrument. A Micrometric ASAP 2020 instrument was used for
BET-N2, and BTH-N2 surface area and pore diameter measurements
of the sol–gel matrices. Analytical gas chromatographic determi-
nations were carried out with a Hewlett–Packard model Agilent
4890 D by using a 30 m long column packed with Carbowax 20M-
poly(ethylene glycol) in fused silica (Supelco 25301-U). Preparative
gas chromatographic separations were performed on a Varian
Aerograph instrument series 1400 equipped with a 2 m long 10%
Carbowax on Chromosorb W column. XPS measurements were per-
formed with a Kratos Axis Ultra X-ray photoelectron spectrometer.
Spectra were acquired with monochromated Al K␣ (1486.7 eV) X-
ray source with 0◦ take off angle. The pressure in the test chamber
was maintained at 1.5 × 10−9 Torr during the acquisition process.
High resolution XPS scans were collected for Rh 3d and C 1s peaks
with pass energy 20 eV. The XPS binding energy was calibrated
with respect to the peak position of C 1s at 285.0 eV. Data analy-
∗
Corresponding author. Tel.: +972 2 6585329; fax: +972 2 6513832.
E-mail address: jblum@chem.ch.huji.ac.il (J. Blum).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.12.002

D. Meltzer et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
9
Scheme 1.
sis was performed with Vision processing data reduction software
(Kratos Analytical Ltd.) and Casa XPS (Casa Software, Ltd.). Trans-
mission electron microscopy was done with Scanning Transmission
Electron Microscope (STEM) Tecnai G2 F20 (FEI Company, USA)
operated at 200 kV and equipped with EDAX-EDS for identification
of the elemental composition. Initial powders were dispersed in
ethanol and dropped onto a standard 400 mesh carbon coated cop-
per TEM grid. The average particle size of the sol–gel entrapped
RhCl3-Aliquat 336 was measured with an Olympus CX31 light
microscope operated at the magnification of 40.
stirred magnetically at 25 ◦C until a clear transparent microemul-
sion was formed. In some cases the addition of a few drops of the
propanol was necessary.
2.5. General procedure for the catalytic isomerization
One half of the above immobilized rhodium catalyst (contain-
ing usually 0.057 mmol of the rhodium compounds) was roughly
ground and admixed within a mini autoclave equipped with a
sampler together with one of the aforementioned freshly pre-
pared microemulsions. The autoclave was inserted in a preheated
thermostat. After stirring the reaction mixture at the desired tem-
perature, and for the necessary length of time, the autoclave was
cooled to 20 ◦C. The sol–gel material was filtered off, washed and
sonicated first for 30 min with TDW (30 ml) and then with hexanes.
The sol–gel-free filtrate was treated with NaCl (2 g), which caused
breakage of the microemulsion and phase separation. The aqueous
layer was extracted with hexanes (2 × 30 ml), the extract was dried
over MgSO4, concentrated and subjected to gas chromatographic
separation and analysis. Both the recovered sol–gel material and
the liquid phases were analyzed for Rh by ICP-MS. The dried sol–gel
was reused in a further catalytic run. Isomerization experiments in
several organic solvents were performed for comparison essentially
as described in the literature [12].
2.2. Chemicals
Tetramethoxysilane (TMOS), n-propyltrimethoxysilane, n-
octyltriethoxysilane and phenyltriethoxysilane were obtained
from ABCR-Glest, Inc. Allylbenzene, 4-allylanisole, methyltriocty-
lammonium chloride (Aliquat 336), cetyltrimethylammonium
bromide (CTAB), and rhodium trichloride trihydrate were pur-
chased from Sigma–Aldrich Chemical Company. Sodium dodecyl
sulfate (SDS) was obtained from Ridel de Haën; C12–C14 alcohols
polyoxyethylene ethers (Marlipal 24/70) were purchased from
Sasol Co. and sucrose laurate (L-1695-) from Mitsubishi-Kasai Food
Corp., Mie, Japan. 4-Allyltoluene, and 4-allylchlorobenzene were
prepared from allyl bromide and the corresponding bromoarenes
[10]. (E)- and (Z)-1-phenyl-1-propene, 1-(4-tolyl)-1-propene,
1-(4-methoxyphenyl)-1-propene
and
1-(4-chlorophenyl)-1-
3. Results and discussion
propene were prepared from the corresponding allylarenes by
polystyrene-bound dichlorobis(triphenylphosphine) ruthenium
(II) [11] followed by gas chromatographic separation of the
isomers.
Previously [12] we reported that catalytic double bond
migration in allyllic compounds by sol–gel entrapped
[(C8H17)3NCH3][RhCl4] (RhCl3-Aliquat 336) takes place smoothly
in toluene at 90 ◦C. After completion of the process the orange
colored support could be recycled. Unless a hydrophilic catalyst
was employed [13], the replacement of the organic solvent by
water proved to stall the isomerization, but addition of a sur-
factant, together with an alcoholic co-surfactant, resumed the
catalytic process. In the aqueous medium the sol–gel supported
catalyst turns black. TEM analysis indicated that this black sol–gel
contains Rh(0) nanoparticles (vide infra). XPS measurements of
the ceramic material revealed that prior to the catalytic process,
i.e., the immobilized [(C8H17)3NCH3][RhCl4], the rhodium catalyst
was in the +3 oxidation state showing binding energy (BE) peaks
at 308.67 and 313.41 eV corresponding to Rh(III) 3d5/2 and Rh(III)
3d3/2, respectively (cf. e.g., Refs. [14–16]). After the isomeriza-
tion process, the major BE peaks were centered at 307.62 and
312.36 eV, [Rh(0) 3d5/2 and Rh(0) 3d3/2] having shoulders at ∼308
2.3. Preparation of the sol–gel entrapped rhodium catalyst
A mixture of phenyltrimethoxysilane (0.45 ml, 2.26 mmol),
TMOS (3.6 ml, 24.2 mmol), RhCl3·3H2O (30 mg. 0.114 mmol) dis-
solved in triply distilled water (TDW, 2.0 ml), and Aliquat 336
(50 mg, 0.124 mmol) dissolved in MeOH (3.0 ml), was stirred at
room temperature for as long as possible (usually ca. 3 d). The gel
was allowed to stand at 25 ◦C for 3 d and then dried at 80 ◦C and
0.5 Torr for 24 h. The dry sol–gel matrix was washed and sonicated
with boiling CH2Cl2 (30 ml) and redried at 0.5 Torr at 80 ◦C for 5 h
before use. Both the ceramic material and the combined washings
were analyzed for their rhodium content by ICP-MS. In none of the
experiments did the washing contain >0.02 mg Rh.
For the entrapment of the RhCl3-Aliquat 336 ion pair within
octylated or propylated sol–gel, the hydrolysis of the TMOS and of
the alkyltrialkoxysilanes were carried out separately. The hydroly-
sis of octyltriethoxysilane (2.4 ml, 6.68 mmol) was performed with
EtOH (5.6 ml) and TDW (0.4 ml), and the hydrolysis of propy-
ltrimethoxysilane (1.17 ml, 6.64 mmol) was carried out with MeOH
(3.88 ml) and TDW (0.4 ml). After 24 h, TMOS (3.6 ml, 24.2 mmol)
together with RhCl3·3H2O (30 mg, 0.114 mmol) dissolved in TDW
(2.0 ml) and Aliquat 336 (50 mg, 0.124 mmol) in MeOH (2.4 ml) was
added to each of the hydrolyzed silanes.
and ∼312 eV [that corresponds to Rh(I) 3d5/2 and Rh(I) 3d3/2
]
and show distorted curves at 309 and 313 eV. The shoulders and
distorted curves indicate the presence of residual traces of Rh(I)
and Rh(III) species that are buried within the bulk sol–gel material,
and, therefore are not converted into Rh(0) nanoparticles. The XPS
analysis of the used heterogenized catalyst recovered from the
isomerization in toluene did not have any peaks that correspond to
Rh(0). The only signals were those of Rh(I) 3d5/2, Rh(I) 3d3/2, Rh(III)
3d5/2 and Rh(III) 3d3/2. This means that the reaction in toluene
does not involve the formation of metallic nanoparticles.
2.4. Preparation of the microemulsions
In analogy to the isomerization of allylbenzenes by heterog-
enized catalysts in organic solvents [12,17–19] the reactions in
microemulsions were also converted after prolonged heating into
equilibrium mixtures of 1, 2 and 3, in which the residual compo-
A mixture of the allylarene (1.35 mmol), TDW (17–20 ml), the
appropriate surfactant (0.6–0.8 g and n-PrOH (1.6–2.0 ml) was

10
D. Meltzer et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
Table 1
Table 2
Isomerization of allylbenzene by immobilized RhCl3-Aliquat 336 in microemulsion
Substituent effect on the isomerization of allylbenzenes 1 by the heterogenized
under comparable experimental conditionsa.
RhCl3-Aliquat 336 under the EST conditionsa.
Entry
Hydrophobic function Rb
Composition of reaction mixture
after 3 h, %c,d
Entry
Substituent, R
105 M s−1
ks/ka
log(ks/ka)b
1
2
3
4
H
4-Cl
4-CH3
4-OCH3
5.74
2.64
10.51
13.08
1
0
0.460
1.831
2.279
−0.337
0.263
0.358
1, R = H
2, R = H
3, R = H
1
2
3
4
None
26.6
46.0
5.8
8.6
4.0
8.5
7.5
64.8
50.0
85.7
88.6
Phenyl
n-Octyl
n-Propyl
a
Reaction conditions as in Table 1 with phenylated sol–gel support.
ka and ks represent the reaction constants for the consumption of the unsubsti-
b
3.9
tuted and of the substituted allylbenzene, respectively.
a
Reaction conditions: Microemulsion of 1 (0.18 ml), SDS (0.65 g), n-PrOH (1.63 ml),
H2O (17.6 ml), sol–gel material prepared from tetramethoxysilane (TMOS, 1.8 ml),
either RSi(OMe)3 or RSi(OEt)3 (7 mmol), RhCl3·3H2O (15 mg), [(C8H17)3NCH3]Cl
(25 mg), 140 ◦C.
place with and without formation of Rh(0) particles, respectively,
the Hammett plots of both reactions (as well as by the RhCl3-
Dowex® 1 catalyzed process in 92% aqueous EtOH) are quite similar
having ꢁ values of the same order of magnitude [19].
b
In RSi(ORꢀ)3.
c
Average of at least two reactions that did not differ by 3%.
In the first run.
d
The reason for the generation of Rh(0) nanoparticles in the
aqueous microemulsion is not yet fully understood. In attempts
to elucidate possible factors that may be responsible for the con-
version of the immobilized Rh(III) into Rh(0) species in the aqueous
medium, we first omitted the PrOH co-surfactant from the reac-
tion mixture. However, the removal of the alcohol (which caused
the isomerization to slow down) still gave sol–gel entrapped Rh(0)
particles. On the other hand, experiments in toluene, in the presence
of large quantities of propanol, did not form any nanoparticles even
after prolonged heating at 140 ◦C within an autoclave. The possibil-
ity that the unsaturated substrate is responsible for the generation
of the nanoparticles, has also been disproved. The latter particles
were formed equally well when the allylbenzene was replaced by
either benzene, toluene or propylbenzene. Similar Rh(0) particles
appeared in the TEM micrographs when the isomerization was con-
ducted under ambient atmosphere or under N2.
The presence of surfactants in the reaction mixture led nearly
always to the generation of Rh(0) particles. An exception was the
case in which allylbenzene (1, R = H) was isomerized in toluene
under anhydrous conditions and the surfactant was the toluene
soluble Marlipal 24/70 (vide infra). Nanoparticles started, how-
ever, to accumulate upon addition of water to the stirred reaction
mixture. Thus, we assume that the aqueous surfactant is the main
reducing factor (cf., e.g. Ref. [24]). Although we have no means for
quantitatively measuring the amount of the randomly distributed
nanoparticles, we observed that both their abundance and the size
nent 1 exists only in a small amount. When for example, the starting
unsaturated hydrocarbon was 1, R = H, a mixture of 0.9% of 1, 7.5% 2
and 91.6% of 3 was generated after 20 h at 140 ◦C. However, unlike
in toluene the reactions in microemulsions were found to always
form Rh(0) nanoparticles.
The fact that both the Rh(III) precatalyst and the Rh(0) catalyze
(in organic media) the double bond migration in allylarenes, gave
rise to the question whether both species independently promote
the isomerization process in the microemulsion. As we have pre-
viously observed, the catalytically active RhCl3-Aliquat 336 can
be extracted from the sol–gel support with hot ethanol [20], we
treated our immobilized catalyst with the boiling alcohol after the
second, third and fourth run and found no Rh(III) species in any of
the extracts. Therefore, we conclude that unlike in toluene, under
the EST conditions only the Rh(0) nanoparticles are responsible for
the catalytic reactions.
It should be noted that although the generation of metallic
nanoparticles during isomerization of olefins is much less common
than during a variety of other catalytic processes [21], the dou-
ble bond migration in some cycloalkenes by Ni/Sn catalysts can be
regarded as a nanoparticle-catalyzed process [22].
In analogy to some other catalyses that take place under the
EST conditions [4,23], the isomerization of allylbenzene proved to
depend strongly on the hydrophobicity of the sol–gel support. The
hydrophobic sol–gel matrices were prepared by hydrolysis and co-
condensation of TMOS and RSi(ORꢀ)3, where R is the hydrophobic
group and Rꢀ is either CH3 or C2H5. Under comparable experimen-
tal conditions using (i) hydrophilic, (ii) phenylated, (iii) octylated
and (iv) propylated sol–gel, the rates of the double bond migra-
tion were in the order of iv > iii > i > ii (see Table 1). This order is
different than the one recorded in Pd(OAc)2-catalyzed Heck cou-
pling of bromobenzene and styrene in microemulsions where the
reaction does not take place at all by the catalyst within the non-
hydrophobic support and the best results are obtained within
phenylated sol–gel [23]. It is notable that in toluene the isomer-
ization of allylarenes by [(C8H17)3NCH3][RhCl4] proceeds within
hydrophilic sol–gel [12,20].
Under the conditions of Table 1 the catalytic transformation of
allylbenzene (1, R = H), as well as the conversions of the substituted
analogs, 1, R = Cl; 1, R = CH3 and 1, R = OCH3, follow the first order
rate law, and the initial rates were found to depend on the elec-
tronic nature of the substrates. Electron donating moieties proved
to enhance the rate and visa versa electron attracting groups. Rep-
resentative numerical data are summarized in Table 2, and can be
expressed in terms of a Hammett plot with ꢀ = −1.42 shown in
Fig. 1.
0.5
0.4
CH3
0.3
0.2
CH3O
0.1
H
0.0
-0.1
-0.2
-0.3
-0.4
Cl
0.2
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.3
σ
Fig. 1. Hammett plot for the isomerization reaction shown in Scheme 1 using
[(C8H17)3NCH3][RhCl4} in phenylated silica sol–gel at 140 ◦C as pre-catalyst. ka and
kb are the same as in Table 2. The ꢁ values were taken from Ref. [28]; Y = −1432X;
Although double bond migration in allylarenes by the RhCl3-
Aliquat 336 catalyst, under the EST conditions, and in toluene take
R
2 = 0.9891.

D. Meltzer et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
11
Fig. 2. TEM images of typical sol–gel supported sections after isomerization of allylbenzene by the [(C8H17)3NCH3][RhCl4} at sol–gel pre-catalyst: (a) in an aqueous
microemulsion, (b) in toluene.
Table 4
are strongly dependent on the reaction temperature. TEM measure-
Dependence of the isomerization of allylbenzene in aqueous microemulsion on the
ments indicated that during the isomerization of 1, R = H at 140 ◦C
recycling of the catalysta.
densely distributed Rh(0) particles, having diameters of 2–5 nm are
Run No.
Yield after 3 h, %b
formed (see Fig. 2), but at 100 ◦C their amount was significantly
reduced and their size shrunk to 0.7–1.7 nm. At 82 ◦C both the
formation of nanoparticles and the isomerization proceeded very
slowly. Seven days were required to obtain a mixture of 3.7% of 1,
R = H, 6.8% of 2, R = H and 89.5% of 3, R = H.
1, R = H
2, R = H
3, R = H
1
2
3
4
5
46
70
71
74
73
4
3
4
3
4
50
27
25
23
24
The nanoparticles seem to be formed equally well when the sur-
factant is anionic (SDS), non-ionic [e.g., Marlipal® 24/70 or sucrose
laurate (L 1695)] or cationic [cetyltrimethylammonium bromide,
(CTAB)]. The nature of the surfactant has however, a different
influence on the efficiency of the catalytic isomerization. Some
typical conversions of allylbenzene by the Rh(0) nanoparticles in
microemulsions stabilized in the presence of SDS, Marlipal 24/70,
L 1695 or CTAB during 3 h at 140 ◦C, under comparable standard
reaction conditions, are summarized in Table 3. These data can be
rationalized by the different tendency of the surfactant micelles to
be absorbed by the catalyst-containing sol–gel. While the anionic
SDS is actually repelled by the partially negatively charged sil-
ica sol–gel, the cationic CTAB is attracted by the ceramic material
[25,26].
The heterogenization of the [(C8H17)3NCH3][RhCl4] ion pair
precatalyst by its entrapment within silica sol–gel, makes the cat-
alyst recyclable, indifferently if it acts in the form of a discrete
organometallic complex in toluene [12] or as supported Rh(0)
nanoparticles in aqueous microemulsions (Table 4). A significant
loss of catalytic activity takes place only in the first run. We assume
that this decrease in activity is associated with some morphologi-
cal changes in the sol–gel matrix (Si–O bond breaking) that takes
place at 140 ◦C by the water. Support for this assumption is pro-
vided by the BET measurements of the internal surface areas and
a
Reaction conditions as in Table 1, except that the sol–gel material was formed
in all cases from TMOS and PhSi(OEt)3.
b
The yields are the average of 2–5 experiments that did not differ by more than
2%.
the sizes of the pore diameters. The initial average internal sur-
face area of 264 m2/g was found to grow to 335 m2/g after the
fifth catalytic cycle. Likewise, the average pore diameter extended
˚
from 21 to 32 A (cf. a similar size increase during the Heck reac-
tion in water [6]). We recall that in toluene and in some other
organic solvents, the isomerization catalyst [(C8H17)3NCH3][RhCl4]
(entrapped in hydrophilic sol–gel) also loses some of its catalytic
activity. However, the decrease in organic solvents was shown to be
associated with pore entrance blocking by the substrate and prod-
ucts [20]. There the addition of a limited amount of water helps to
reopen the clogged pores, but addition of excessive water during
the catalytic process inhibits the isomerization.
In many catalytic reactions, mass-transport-limitations have
a strong impact on the reaction rate. Here, in the EST system,
the reaction takes place at the rhodium nanoparticles that are
entrapped in a porous hydrophobic sol–gel matrix. The allylben-
zene or the substituted analogs, dissolved in normal micelles of an
o/w-microemulsion, must be transported from the aqueous envi-
ronment to the active site of the catalyst. A scheme of this process
is shown in Fig. 3.
Table 3
Effect of the nature of the surfactant on the isomerization of allylbenzenea.
Entry
Surfactant
Composition of reaction mixture
after 3 h, %b
The Weisz–Prater criterion (Eq. (1)) [27] which is based on
macrokinetics and use to estimated the influence of diffusion on
the reaction rate by a calculation of the ratio of the characteristic
reaction time to the characteristic diffusion time is the Weisz mod-
ule ꢂ. For ꢂ-values lower than 1 the influence of diffusion can be
neglected.
1, R = H
2, R = H
3, R = H
1
2
3
4
SDS
Marlipal 24/70
L 1695
46
27
35
17
4
8
5.5
11.5
50
65
59.5
71.5
CTAB
a
Reaction conditions as in Table 1 except that the hydrophobic function was
phenyl and the surfactant was as indicated.
((n + 1)/2)((r · ꢀcat)/c0)
ꢂ =
(1)
Deff/Lc2
b
In the first run.

12
D. Meltzer et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
Fig. 3. Possible transport mechanism in the EST system (S: substrate, Rh: rhodium nanoparticles). KMW is the partition coefficient between the micelles and the continuous
aqueous phase and Diff stands for diffusion. The reaction rate R is a function of the experimental conditions, e.g. temperature, and depends on the local substrate concentration.
In Eq. (1), Lc (cm) is the characteristic catalyst length, n is the
reaction order, r (mol s−1 g−1) is the reaction rate, ꢀcat (g/cm3)
is the density of the catalyst, Deff (cm2/s) is the effective diffu-
sion coefficient and c0 (mol/l) is the initial substrate concentration,
respectively. From light microscope measurements a mean char-
acteristic catalyst length of 0.063 cm is obtained for the sol–gel
material that has a density of approximately 1.9 g/cm3. From the
isomerization experiments, the reaction order was found to be
1 and a reaction rate of 7.8 × 10−8 mol s−1 g−1 was calculated for
an initial concentration of 0.068 mol/l of 1. Based on these values,
ꢂ was calculated for different situations: molecular diffusion and
micelle diffusion. For micelle diffusion (Deff ∼ 10−9–10−7 cm2/s) the
Weisz module would be extremely large (ꢂ ∼ 100–10,000) but in
case of molecular diffusion (Deff ∼ 10−5 cm2/s) values lower than
one will be obtained and mass transfer limitations can be excluded.
This estimation shows that the substrate is not able to approach
the catalyst by transportation through micelles, therefore a dif-
fusion of the individual substrate molecules is the only possible
transport mechanism. This is supported by the observation that
the rate responds to the effect of the substituents. In case of mass
transport limitations this behavior should not be observed. The par-
tition coefficient KMW of the allylbenzene between the micelles and
the continuous aqueous phase has a substantial influence on this
molecular diffusion process. This parameter itself is again strongly
affected by the size and polarity of the surfactants.
Comparison of the isomerization of allylbenzene in aqueous
microemulsions with that in organic solvents at identical tempera-
tures and stirring rates, with the same substrate concentration and
in the presence of the same amount of heterogenized precatalyst
reveals, that the reaction in the aqueous medium is the more effi-
cient one. However, it should be changed to noted, that while in
the reaction in the organic solvent no surfactant is present, and the
sol–gel matrix is hydrophilic, under the EST conditions these com-
ponents of the systems play a significant role. Some representative
experiments carried out in different solvents are summarized in
Table 5.
4. Conclusions
In aqueous microemulsions the sol–gel entrapped ion pair
[(C8H17)3NCH3][RhCl4] decomposes into Rh(0) nanoparticles. In
benzene and toluene no such decomposition could be detected,
although allylarenes were found to undergo isomerization to
1-aryl-1-propenes in both types of solvents. The reductive trans-
formation of the Rh(III) pre-catalyst to the Rh(0) species seems
to be affected by the aqueous surfactant that is used to solubi-
lize the hydrophobic allylarene. The efficiency of the double bond
migration in water is dependent on both the electronic nature
of the allylic substrate and that of the surfactant, as well as on
the hydrophobicity of the sol–gel matrix in which the rhodium is
encaged. It can be concluded that no mass transfer process inhibits
the reaction in the studied EST system. Since the reaction in the
aqueous medium proceeds faster than in benzene and in toluene,
the double bond isomerization, under EST conditions, can be useful
for the replacement of the conventional organic solvent by water.
Table 5
Dependence of the isomerization on the reaction mediuma.
Entry Reaction medium
Time
(h)
Composition of
reaction mixture
after 3 h, %b
Formation
of metallic
nanoparticles
1, R = H 2, R = H 3, R = H
1
2
3
4
5
Aq. microemulsion 14.5
Aq. microemulsion 20c
7.9
0.9
77.0
88.6
75.0
7.1
7.6
2.9
2.9
5.5
85.0
91.5
20.1
8.5
+
+
+
–
–
THF
14.5
14.5
14.5
Benzene
Toluene
19.5
a
Reaction conditions for entries 1 and 2 as in Table 1. Conditions for entries
Acknowledgement
3–5: allylbenzene (0.18 ml), solvent (18 ml), sol–gel material from TMOS (1.8 ml),
PhSi(OMe)3 (0.225 ml), RhCl3·3H2O (15 mg) and [(C8H17)3NCH3]Cl (25 mg), 140 ◦C.
b
In the first run.
We gratefully acknowledge the financial support of the trilat-
eral study by the Deutsche Forschungsgemeinschaft (DFG) through
grant SCHO687/8-1.
c
Since an increase of the reaction time did not affect anymore the composition of
the reaction mixture we postulate that the given data are those of the equilibrium
mixture.

D. Meltzer et al. / Journal of Molecular Catalysis A: Chemical 335 (2011) 8–13
13
References
[15] V.I. Nefedov, M.N. Firsov, I.S. Shaplygin, J. Electron Spectrosc. Relat. Phenom. 26
(1982) 65–79.
[16] C. Furlani, G. Mattogno, G. Polzonetti, G. Braca, G. Valentini, Inorg. Chim. Acta
69 (1983) 199–205.
[17] A. Zocan, Y. Sasson, J. Blum, J. Org. Chem. 46 (1981) 255–260.
[18] M. Setty-Fichman, J. Blum, Y. Sasson, M. Eisen, Tetrahedron Lett. 35 (1994)
781–784.
[19] M. Setty-Fichman, Y. Sasson, J. Blum, J. Mol. Catal. A: Chem. 126 (1997) 27–36.
[20] A. Rosenfeld, D. Avnir, J. Blum, J. Chem. Soc., Chem. Commun. (1993) 583–584.
[21] D. Astruc (Ed.), Nanoparticles and Catalysis, Wiley-VCH, Weinheim, Germany,
2008.
[22] P. Lesage, J.-P. Candy, C. Hirigoyen, F. Humbolt, M. Leconte, J.-M. Basset, J. Mol.
Catal. A: Chem. 117 (1996) 303–309.
[23] A. Rozin-Ben Baruch, D. Tsvelikhovsky, M. Schwarze, R. Schomäcker, M. Fanun,
J. Blum, J. Mol. Catal. A: Chem. 323 (2010) 65–69.
[24] M. de Dios, F. Barroso, C. Tojo, in: M. Fanun (Ed.), Surfactants Science Series,
Vol. 144. Microemulsions: Properties and Applications, CRC Press/Taylor and
Francis Inc., Boca Raton, 2009, pp. 451–464.
[25] C. Rottman, S. Garda, Y. De Hazan, S. Malchior, D. Avnir, J. Am. Chem. Soc. 121
(1999) 8533–8543.
[26] E. Tyrode, M.W. Rutland, C.D. Bain, J. Am. Chem. Soc. 130 (2008) 17434–17445,
and references cited therein.
[1] W. Wei, C.C.K. Keh, C.-J. Li, R.S. Varma, Chem. Technol. Environ. Policy 6 (2004)
250–257, and references cited therein.
[2] U.M. Lindström (Ed.), Organic Reactions in Water: Principles Strategies and
Applications, Blackwell, Oxford, 2007.
[3] R. Abu-Reziq, D. Avnir, J. Blum, Angew. Chem. Int. Ed. 41 (2002) 4132–4134.
[4] R. Abu-Reziq, J. Blum, D. Avnir, Chem. Eur. J. 10 (2004) 958–962.
[5] R. Abu-Reziq, D. Avnir, J. Blum, Eur. J. Org. Chem. (2005) 3640–3642.
[6] D. Tsvelikhovsky, J. Blum, Eur. J. Org. Chem. (2008) 2417–2422.
[7] D. Tsvelikhovsky, D. Pessing, D. Avnir, J. Blum, Adv. Synth. Catal. 350 (2008)
2856–2858.
[8] W.A. Herrmann, M. Prinz, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homo-
geneous Catalysis with Organometallic Compounds, Wiley-VCH, Weinheim,
Germany, 2002, pp. 1119–1130.
[9] A. Scarso, M. Colladon, P. Sgarbossa, C. Santo, R.A. Michelin, G. Strukul,
Organometallics 29 (2010) 1487–1497, and references cited therein.
[10] C.D. Hurd, M.T. Bollman, J. Am. Chem. Soc. 56 (1934) 447–449.
[11] A. Zoran, Y. Sasson, J. Blum, J. Org. Chem. 46 (1981) 255–260.
[12] J. Blum, A. Rosenfeld, N. Polak, O. Israelson, H. Schumann, D. Avnir, J. Mol. Catal.
A: Chem. 107 (1996) 217–223.
[13] H. Schumann, V. Ravindar, L. Meltser, W. Baidossi, Y. Sasson, J. Blum, J. Mol.
Catal. A: Chem. 118 (1997) 55–61.
[14] J.L.G. Fierro, J.M. Palacios, F. Tomas, Surf. Interface Anal. 13 (1988) 25–32.
[27] P.B. Weisz, C.D. Prater, Adv. Catal. 6 (1954) 143–196.
[28] T. Matsui, H.C. Ko, G. Hepler, Can. J. Chem 52 (1974) 2906–2911.