Aqueous biphasic hydroformylation in the presence of cyclodextrins mixtures: Evidence of a positive synergistic effect

Article abstract of DOI:10.1039/c2dt30646h

Mixtures of randomly methylated cyclodextrins of various sizes have been evaluated in the rhodium-catalysed hydroformylation of higher olefins in an aqueous biphasic medium. A marked positive non-linear effect on 1-tetradecene conversion is observed when the CD molar ratio in the mixture is modified. The formation of 2:1 ternary inclusion complexes between RAME-CDs and the olefin is supposed to be responsible for the extra conversion observed. The Royal Society of Chemistry.

Full text of DOI:10.1039/c2dt30646h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 8643
www.rsc.org/dalton
PAPER
Aqueous biphasic hydroformylation in the presence of cyclodextrins
mixtures: evidence of a positive synergistic effect†
Michel Ferreira,a,b,c François-Xavier Legrand,a,b,c Cécile Machut,a,b,c Hervé Bricout,a,b,c Sébastien Tilloy*a,b,c
and Eric Monfliera,b,c
Received 21st March 2012, Accepted 17th May 2012
DOI: 10.1039/c2dt30646h
Mixtures of randomly methylated cyclodextrins of various sizes have been evaluated in the rhodium-
catalysed hydroformylation of higher olefins in an aqueous biphasic medium. A marked positive non-
linear effect on 1-tetradecene conversion is observed when the CD molar ratio in the mixture is modified.
The formation of 2 : 1 ternary inclusion complexes between RAME-CDs and the olefin is supposed to be
responsible for the extra conversion observed.
In the field of aqueous organometallic catalysis, CDs are
mainly used as mass transfer promoters between aqueous and
Introduction
Cyclodextrins (CDs) are water-soluble oligosaccharides which
are able to form inclusion complexes with a large number of
molecules in aqueous media.1,2 Numerous studies have shown
that mixtures of such macromolecules can be more efficient in
terms of inclusion ability than the individual CD. For instance,
the commercial application of the well-known Febreze® product
uses cyclodextrins mixtures in order to be more efficient in eli-
minating odours.3 Another patent deals with the use of a mixture
of at least two different CDs selected among α-CD, β-CD and
γ-CD, with the aim of stabilising unsaturated fatty acids and to
significantly reduce the odour and the taste of the oleaginous
substance.4 In the academic literature, some examples are also
described. Indeed, the Loftsson group showed that addition of
small amounts of 2-hydroxypropyl-γ-cyclodextrin (HP-γ-CD) to
aqueous native γ-CD solution reduces the turbidity and increases
the complexation efficiency of various glucocorticoids (dexa-
methasone and hydrocortisone) leading to an increased guest
solubilisation.5,6 Formation of possible soluble larger CD aggre-
gates is supposed to be responsible for these effects. Enhance-
ment of aqueous solubility of a poorly water-soluble cyclic
undecapeptide7 and a potent glucocorticoid steroid8 using
respectively α-CD/HP-β-CD mixtures and α-CD/HP-γ-CD mix-
tures is along the same lines. Such synergistic effects between
different CDs have also been highlighted when mixed GC
stationary phases consisting of two different CD derivatives were
used in resolving enantiomers9–12 or in positional isomers com-
pounds separation.13,14
organic layers by forming inclusion complexes with the sub-
strate.15 For that matter, our group has recently showed that the
methylation degree of the randomly methylated-β-CD (RAME-
β-CD – Scheme 1), which is a crude mixture of several methy-
lated CDs, directly impacted activity of the 1-decene rhodium-
catalyzed hydroformylation. The catalytic activity was found to
gradually increase with the number of methyl groups on the
methylated-CD, whereas the chemoselectivity and regioselec-
tivity remained unchanged.16 This study has clearly demon-
strated that the CD mixture composition could affect the
catalytic activity.
In this context, the present paper focuses on the evaluation of
different RAME-CD binary mixtures in the hydroformylation
reaction: RAME-α-CD/RAME-β-CD mixture; RAME-α-CD/
RAME-γ-CD mixture and RAME-β-CD/RAME-γ-CD mixture.
Results and discussion
The effect of various RAME-α-CD/RAME-β-CDs mixtures on
the catalytic activity has been studied in the rhodium-catalyzed
hydroformylation reaction of higher olefins, in an aqueous bi-
phasic system using Rh(acac)(CO)2 as catalyst precursor and
tris(m-sulfonatophenyl)phosphane trisodium salt (TPPTS) as a
water-soluble ligand (Scheme 1).
Substrate/CD/TPPTS/Rh molar ratios of 500/12/5/1 were used
as we have previously described that these conditions allowed
reaching high conversion of 1-decene in the presence of com-
mercial RAME-β-CD.17 Results concerning 1-decene, 1-tetra-
decene and 1-hexadecene are presented in Fig. 1.
In the case of 1-decene hydroformylation, a linear correlation
between olefin conversion and RAME-β-CD percentage in the
RAME-α-CD/RAME-β-CD mixture was obtained. On the other
hand, when 1-tetradecene and 1-hexadecene were used, the
conversion evolution as a function of the RAME-β-CD molar
aUniversité Lille Nord de France, F-59000 Lille, France
bUArtois, Unité de Catalyse et de Chimie du Solide (UCCS), F-62307
Lens, France. E-mail: sebastien.tilloy@univ-artois.fr;
Fax: +33 3 21 79 17 55; Tel: +33 3 21 79 17 54
cCNRS, UMR 8181, France
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c2dt30646h
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 8643–8647 | 8643

View Article Online
Scheme 1 Biphasic hydroformylation of higher olefins by a Rh–TPPTS system assisted by randomly methylated cyclodextrins mixtures.
Fig. 2 Influence of various RAME-CDs mixtures on 1-tetradecene
conversion in the hydroformylation reaction. Reaction conditions: Rh-
(acac)(CO)2 = 2.03 × 10−2 mmol, TPPTS = 0.105 mmol, cyclodextrin =
0.24 mmol, H2O = 5.75 mL, olefin = 10.17 mmol, 1500 rpm, T = 80 °C,
P(CO–H2) = 50 atm, time = 6 h.
obtained when RAME-α-CD or RAME-β-CD was used alone
(9% and 22% conversion respectively). However, with 1-hexa-
decene, the conversion profile is different. Indeed, the deviation
from the linear correlation is clearly less pronounced. Even so, a
19% maximum olefin conversion is reached for a composition
mixture of RAME-α-CD/RAME-β-CD equal to 25/75. In this
case, the maximum conversion is 1.2 times higher than the
theoretical conversion for a linear correlation and closed to the
value obtained with the sample containing only RAME-β-CD
(18% conversion).
Given that the strongest positive non-linear effect on the olefin
conversion was pointed out for 1-tetradecene, we decided to
perform experiments with this substrate in the presence of other
RAME-CDs combinations. Results concerning 1-tetradecene
biphasic hydroformylation using RAME-β-CD/RAME-γ-CD
and RAME-α-CD/RAME-γ-CD mixtures are shown in Fig. 2.
Fig. 1 Influence of the RAME-α-CD/RAME-β-CD mixture compo-
sition on the olefin conversion for different substrates in the hydroformy-
lation reaction. Reaction conditions: Rh(acac)(CO)2 = 2.03 × 10−2
mmol, TPPTS = 0.105 mmol, cyclodextrin = 0.24 mmol, H2O =
5.75 mL, olefin = 10.17 mmol, 1500 rpm, T = 80 °C, P(CO–H2) =
50 atm, time = 6 h (2 h for 1-decene).
ratio cannot be modelled in the same way. Indeed, a deviation
from the linear correlation, termed a positive non-linear effect,
was observed. Interestingly, this effect is characterized by a 29%
maximum olefin conversion for a specific composition mixture
(RAME-β-CD: 59%/RAME-α-CD: 41%) when the substrate was
1-tetradecene. This conversion value is about 2 times greater
than the one which should be observed for a linear correlation.
Even more significantly, this conversion is higher than those
8644 | Dalton Trans., 2012, 41, 8643–8647
This journal is © The Royal Society of Chemistry 2012

View Article Online
In both graphs, a positive non-linear relationship between
olefin conversion and RAME-CD mixture composition is still
observed. A 29% maximum conversion was obtained for a
RAME-β-CD: 62%/RAME-γ-CD: 38% mixture. In the RAME-
α-CD/RAME-γ-CD mixture, the best conversion equal to 14%
was obtained when CDs were added in the same ratio (50 : 50).
In both cases, the conversion was higher than those obtained
when RAME-CDs were used alone and the maximum conver-
sion value reached is about 1.7 times higher compared to linear-
ity. Consequently, the outstanding effect pointed out in the case
of RAME-α-CD/RAME-β-CD mixtures could be extended to
other CDs mixtures, showing a synergistic effect between CDs
towards 1-tetradecene conversion. This synergistic effect is gener-
ally applicable for this substrate and is not size cavity dependent.
In order to understand this positive non-linear effect origin,
two different hypotheses have been considered. This effect could
be due to the RAME-CDs mixtures surface activity or to a con-
sequence of specific inclusion complexes formation with the
substrate during the catalysis. Indeed, the outstanding effect of
RAME-β-CD on reaction rates during aqueous biphasic catalytic
reactions was attributed to its slight surface activity combined
with its deep hydrophobic host cavity that properly accommo-
dates the substrate.18 Tensiometry studies were realized in the
catalytic conditions i.e. on different 42 mM concentrated
RAME-α-CD/RAME-β-CD mixtures solutions. Results showed
that surface tension is still constant and close to 50 mN m−1
whatever the RAME-β-CD ratio in the mixture is, meaning that
the RAME-β-CD proportion variation did not significantly
modify the surface activity. The same behaviour was observed
for the RAME-β-CD/RAME-γ-CD and RAME-α-CD/RAME-
γ-CD mixtures (see ESI†). Consequently, the observed positive
non-linear effect could not be related to modifications in surface
activity. Hence, the formation of specific inclusion complexes
during the biphasic hydroformylation was taken into account.
Firstly, it must be pointed out that in the case of RAME-α-CD/
RAME-γ-CD mixture, the maximum olefin conversion was
obtained for the 50 : 50 specific composition (Fig. 2). This value
could be consistent with the formation of a 2 : 1 mixed inclusion
complex consisting of one RAME-α-CD and one RAME-γ-CD
complexing together one substrate molecule. This complex
would be more efficient than the 2 : 1 homo-complexes RAME-
α-CD/RAME-α-CD or RAME-γ-CD/RAME-γ-CD with one
substrate molecule. Ternary complexes between an aliphatic sub-
strate and two CDs have already been described. Indeed, the
crystal structure of the supramolecular complex between α-CDs
and decanoic acid was determined as a dimer of two cyclodex-
trin residues complexing the aliphatic chain of the carboxylic
acid, resulting in a 2 : 1 stoichiometry.19 Besides, linear func-
tional molecules, such as 12-aminododecanoic acid, were found
to form simultaneously 1 : 1 and 2 : 1 inclusion complexes with
α-CD in aqueous solution.20 In the same way, ternary complexes
forming 2 : 1 host–guest inclusion complexes involving β-CDs
have been reported in the literature with guests such as volatile
linear n-alkanes,21 long aliphatic α,ϖ-dicarboxylic acids,22–24 long
aliphatic diols25 and long linear aliphatic carboxylic acids.26
In our case, attempts to determine complexes stoichiometries
by headspace chromatography or NMR Job method were unsuc-
cessful due to the low volatility or the low water solubility of the
n-alkenes used. However, ternary complexes consisting of
Table 1 Linear-to-branched aldehyde ratio in the aqueous biphasic
hydroformylation assisted by RAME-α-CD/RAME-β-CD mixturesa
RAME-β-CD %
1-Decene
1-Tetradecene
1-Hexadecene
0
2.8
2.3
2.1
2.0
1.9
1.8
2.6
1.9
1.9
1.8
1.6
1.5
2.4
2.2
2.0
1.9
1.8
1.5
25
50
60
75
100
a Reaction conditions: Rh(acac)(CO)2 = 2.03 × 10−2 mmol, TPPTS =
0.105 mmol, cyclodextrin = 0.24 mmol, H2O = 5.75 mL, olefin =
10.17 mmol, 1500 rpm, T = 80 °C, P(CO–H2) = 50 atm, time = 6 h (2 h
for 1-decene).
Fig. 3 Influence of various mixtures composition on 1-tetradecene conversion considering the interaction between RAME-β-CD and TPPTS.
Reaction conditions: Rh(acac)(CO)2 = 2.03 × 10−2 mmol, TPPTS = 0.105 mmol, cyclodextrin = 0.24 mmol, H2O = 5.75 mL, olefin = 10.17 mmol,
1500 rpm, T = 80 °C, P(CO–H2) = 50 atm, time = 6 h.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 8643–8647 | 8645

View Article Online
RAME-α-CD, RAME-γ-CD and 1-tetradecene are suggested to
be formed and are thought to be more efficient in substrate trans-
fer between the two phases leading to the extra conversion
observed. These ternary complexes could explain the synergistic
effect observed for the specific 50 : 50 mixture composition.
This synergistic effect between cyclodextrins was also noted
in the other 1-tetradecene hydroformylation experiments
(RAME-α-CD/RAME-β-CD and RAME-β-CD/RAME-γ-CD
mixtures Fig. 1 and Fig. 2). However, the optimum mixtures
compositions were slightly shifted towards a higher RAME-
β-CD ratio (Fig. 1 and 2). This phenomenon could be related to
the formation of an inclusion complex between the TPPTS
ligand, used to dissolve the transition metal in the aqueous
phase, and RAME-β-CD.‡27,28 A part of RAME-β-CD is effec-
tively poisoned by the phosphane ligand, resulting in the shift
detected. If these inclusion complexes between TPPTS and
RAME-β-CD were taken into account, the optimum RAME-
β-CD percentage shifts from 59% to 51% for the RAME-α-CD/
RAME-β-CD mixture, and from 62% to 55% for the RAME-
β-CD/RAME-γ-CD mixture (Fig. 3). Consequently, 2 : 1 ternary
complexes are also formed with these mixtures. The formation
of inclusion complexes between TPPTS and RAME-β-CD was
supported by the linear/branched aldehyde ratio decrease
observed during all the experiments, when the RAME-β-CD pro-
portion in the mixture was increased (Table 1). As already
described, this phenomenon can easily be explained by modifi-
cations in the catalytic active species when RAME-β-CD is
added. In fact, the RAME-β-CD is able to dissociate TPPTS
from HRhCO(TPPTS)3 leading to the formation of low-coordi-
nated phosphane species responsible for this decrease in
regioselectivity.29
Sigma-Aldrich and Strem Chemicals, in their highest purity
and used without further purification. The sodium salt of tris-
(3-sodium sulfonatophenyl)phosphane (TPPTS – P(m-C6H4-
SO3Na)3) was synthesized as reported by Cornils et al.30 The
1
purity of TPPTS was carefully controlled. In particular, H and
31P{1H} NMR analysis indicated that TPPTS was only sulfo-
nated in the meta position and less than 1% of its oxide was
present. Ultrapure water was used in all experiments (Fresenius
Kabi, France; γ = 72.0 mN m−1 at 25 °C). Carbon monoxide–
hydrogen mixture (1 : 1) was used directly from cylinders
(>99.9% pure; Air Liquide). α-CD (Cavamax® W6), RAME-
β-CD (Cavasol® W7 M) and γ-CD (Cavamax® W8) were
purchased from Wacker Chemie GmbH in their pharmaceutical
grade and were used as received. Randomly methylated β-cyclo-
dextrin (RAME-β-CD) was a partially methylated cyclodextrin
and its degree of substitution was equal to 1.8 per glucopyranose
unit. RAME-α-CD and RAME-γ-CD were prepared from native
parent cyclodextrins by adapting a procedure reported by
Deratani et al.31 and their degrees of substitution were equal to
1.8 as well.
All the high-pressure hydroformylation experiments were
carried out in a 25 mL stainless steel autoclave supplied by
Parr. Gas chromatographic analyses were carried out on a Varian
GC 3900 gas chromatograph equipped with a polydimethylsilox-
ane capillary column (25 m × 0.32 mm; 0.25 μm) and a flame
ionization detector (GC: FID).
General procedure for rhodium-catalyzed hydroformylation
experiments
All catalytic reactions were performed under nitrogen using stan-
dard Schlenk techniques. All solvents and liquid reagents were
degassed by bubbling nitrogen for 15 min before each use or
by two freeze–pump–thaw cycles before use. Rh(acac)(CO)2
(2.03 × 10−2 mmol), TPPTS (0.105 mmol) and chemically
modified cyclodextrin(s) (the final quantity of cyclodextrins was
equal to 0.24 mmol) were dissolved in 5.75 mL of degassed
water under an inert atmosphere. The resulting aqueous phase
and an organic phase composed of alkene (10.17 mmol) were
charged under an inert atmosphere into the 25 mL reactor, which
was heated at 80 °C. Mechanical stirring equipped with a multi-
paddle unit was then started (1500 rpm) and the autoclave was
pressurized with 50 bar of CO–H2 (1 : 1) from a gas reservoir
connected to the reactor through a high pressure regulator valve
keeping the pressure in the reactor constant throughout the whole
reaction. The reaction medium was sampled after 6 hours
(2 hours in the case of 1-decene) of reaction for GC analyses of
the organic phase after decantation. For kinetic measurements,
the time corresponding to the addition of CO–H2 was considered
as the beginning of the reaction.
Finally, selectivity towards aldehydes is highly maintained to
values around 98% in all entries, meaning that the composition
mixture has no influence on the chemoselectivity.
Conclusion
In this study, we have described for the first time that the cataly-
tic activity of the aqueous biphasic hydroformylation reaction
could be significantly increased by mixing different RAME-CDs
together. These results exhibit the formation of specific inclusion
complexes during the catalytic cycle. Given that the strongest
positive effect towards olefin conversion was always pointed out
for the mixture composition (50 : 50), these synergistic effects
between the CDs are related to the formation of 2 : 1 ternary
complexes.
Work is currently under way to exploit these synergistic
effects in other biphasic transition metal catalyzed reactions.
Experimental
General remarks
Acknowledgements
Organic compounds, alkenes and dicarbonylacetylacetonato
rhodium(I) were purchased respectively from Acros Organics,
This work was supported by the Centre National de la Recherche
Scientifique (CNRS). F.-X. Legrand is grateful to the Ministère
de l’Education et de la Recherche for financial support
(2007–2010). Roquette Frères (Lestrem, France) is gratefully
acknowledged for generous gifts of β-cyclodextrin.
‡Such inclusion complexes are not formed with RAME-α-CD or
RAME-γ-CD.
8646 | Dalton Trans., 2012, 41, 8643–8647
This journal is © The Royal Society of Chemistry 2012

View Article Online
17 T. Mathivet, C. Méliet, Y. Castanet, A. Mortreux, L. Caron, S. Tilloy and
E. Monflier, J. Mol. Catal. A: Chem., 2001, 176, 105.
18 L. Leclercq, H. Bricout, S. Tilloy and E. Monflier, J. Colloid Interface
Sci., 2007, 307, 481.
19 S. Rodríquez-Llamazares, N. Yutronic, P. Jara, U. Englert, M. Noyong
and U. Simon, Eur. J. Org. Chem., 2007, 4298.
20 K. Eliadou, K. Yannakopoulou, A. Rontoyianni and I. M. Mavridis,
J. Org. Chem., 1999, 64, 6217.
21 I. Sanemasa, Y. Wu, Y. Koide, M. Shigenaga, K. Ishibashi and
T. Deguchi, Bull. Chem. Soc. Jpn., 1993, 66, 1424.
22 S. Makedonopoulou and I. M. Mavridis, Acta Crystallogr., Sect. B:
Struct. Sci., 2000, 56, 322.
23 S. Makedonopoulou, A. Tulinsky and I. M. Mavridis, Supramol. Chem.,
1999, 11, 73.
24 S. Makedonopoulou and I. M. Mavridis, Carbohydr. Res., 2001, 335,
213.
25 T. Bojinova, H. Gornitzka, N. L. Viguerie and I. Rico-Lattes, Carbohydr.
Res., 2003, 338, 781.
26 S. Makedonopoulou, J. Papaioannou, J. Argyroglou and I. M. Mavridis,
J. Inclusion Phenom. Macrocyclic Chem., 2000, 36, 191.
27 E. Monflier, S. Tilloy, C. Méliet, A. Mortreux, S. Fourmentin, D. Landy
and G. Surpateanu, New J. Chem., 1999, 23, 469.
Notes and references
1 K. A. Connors, Chem. Rev., 1997, 97, 1325.
2 H. Dodziuk, Molecules With Holes – Cyclodextrins, Wiley-VCH Verlag
GmbH & Co. KGaA, Weinheim, 2006.
3 T. Trinh, J. P. Cappel, P. A. Geis, M. L. McCarty, D. Pilosof and
S. Zwerdling, World Patent WO 96/04937, 1996.
4 M. Skiba, M. Lahiani, F. Bounoure and P. Dechelotte, World Patent WO
2007/107679, 2007.
5 P. Jansook and T. Loftsson, Int. J. Pharm., 2008, 363, 217.
6 P. Jansook, G. C. Ritthidej, H. Ueda, E. Stefánsson and T. Loftsson,
J. Pharm. Pharmaceut. Sci., 2010, 13, 336.
7 B. Malaekeh-Nikouei, H. Nassirli and N. Davies, J. Inclusion Phenom.
Macrocyclic Chem., 2007, 59, 245.
8 B. Malaekeh-Nikouei, S. A. Sajadi Tabassi, G. Gerayeli, M. A. Salmani
and A. Gholamzadeh, J. Inclusion Phenom. Macrocyclic Chem., 2011,
72, 383.
9 D. Kreidler, H. Czesla and V. Schurig, J. Chromatogr., B: Anal. Technol.
Biomed. Life Sci., 2008, 875, 208.
10 M. Y. Nie, L. M. Zhou, Q. H. Wang and D. Q. Zhu, Chromatographia,
2000, 51, 736.
11 J. Magnusson, L. G. Blomberg, S. Claude, R. Tabacchi, A. Saxer and
S. Schürch, J. High Resolut. Chromatogr., 2000, 23, 619.
12 W. A. König, D. Icheln, T. Runge, B. Pfaffenberger, P. Ludwig and
H. Hühnerfuss, J. High Resolut. Chromatogr., 1991, 14, 530.
13 X. Y. Shi, Y. Q. Zhang and R. N. Fu, Anal. Chim. Acta, 2000, 424, 271.
14 S. Qi, P. Ai, C. Wang, L. Yuan and G. Zhang, Sep. Purif. Technol., 2006,
48, 310.
15 F. Hapiot, A. Ponchel, S. Tilloy and E. Monflier, C. R. Chim., 2011, 14, 149.
16 F.-X. Legrand, M. Sauthier, C. Flahaut, J. Hachani, C. Elfakir, S. Fourmentin,
S. Tilloy and E. Monflier, J. Mol. Catal. A: Chem., 2009, 303, 72.
28 E. Monflier, S. Tilloy, L. Caron, J. M. Wieruszeski, G. Lippens,
S. Fourmentin, D. Landy and G. Surpateanu, J. Inclusion Phenom.
Macrocyclic Chem., 2000, 38, 361.
29 E. Monflier, H. Bricout, F. Hapiot, S. Tilloy, A. Aghmiz and
A. M. Masdeu-Bultó, Adv. Synth. Catal., 2004, 346, 425.
30 R. Gärtner, B. Cornils, H. Springer and P. Lappe, DE Pat 3235030,
1982.
31 J. Ropenga, A. Deratani, J. F. Letavernier and D. Majou, FR Pat
2693728, 1992.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 8643–8647 | 8647