Efficient recycling of a chiral palladium catalytic system for asymmetric allylic substitutions in ionic liquid

Article abstract of DOI:10.1039/c1cc12157j

Chiral carbohydrate-based diphosphites were used for Pd-catalysed asymmetric allylic substitution (alkylation, amination, phosphination) in neat ionic liquids (ILs). Pyrrolidinium-based IL led to the best activities, allowing an efficient catalyst immobilization. In the allylic amination (TOF > 3100 h^-1), the catalyst could be recycled nine times preserving both activity and enantioselectivity.

Full text of DOI:10.1039/c1cc12157j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 7869–7871
www.rsc.org/chemcomm
COMMUNICATION
Efficient recycling of a chiral palladium catalytic system for asymmetric
allylic substitutions in ionic liquidw
a
b
b
c
´
Isabelle Favier, Angelica Balanta Castillo, Cyril Godard, Sergio Castillon,
b
a
a
´
Carmen Claver,* Montserrat Gomez* and Emmanuelle Teuma*
Received 14th April 2011, Accepted 18th May 2011
DOI: 10.1039/c1cc12157j
Chiral carbohydrate-based diphosphites were used for Pd-catalysed
asymmetric allylic substitution (alkylation, amination, phosphination)
in neat ionic liquids (ILs). Pyrrolidinium-based IL led to the best
activities, allowing an efficient catalyst immobilization. In the
allylic amination (TOF 4 3100 hꢀ1), the catalyst could be
recycled nine times preserving both activity and enantioselectivity.
Fig. 1 Chiral diphosphite ligands L1 and L2 (TBDPS = Si(tBu)Ph2).
Palladium-catalysed asymmetric allylic substitution reactions
using soft carbon nucleophiles is a well-studied process for
enantioselective formation of carbon–carbon and carbon–
heteroatom bonds, representing a very useful tool for organic
synthesis.1 In these processes, the use of chiral diphosphite
ligands straightforwardly prepared from a large pool of chiral
diols proved to be very efficient in enantioselective allylic
substitution reactions (ee 4 99% and increased reaction rates
due to the greater p-acceptor ability).2 In this context, carbo-
hydrate diphosphite derivatives have been very successfully
used in Pd-catalysed allylic alkylation and amination reactions3
and recently, impressive turnover frequencies together with
ee 4 99% were achieved with C2-symmetric carbohydrate
derived diphosphites in Pd-catalysed allylic alkylation and
amination reactions using dichloromethane as solvent.4 Using
chiral cationic diamidophosphite ligands, high enantioselectivity
was also achieved in the allylic amination of 1,3-diphenylallyl-
acetate in THF.5
most of these reports did not include the use of chiral ligands
and did not explore the possibility to reuse the catalyst.5,7 It is
now well established that imidazolium cations and their
associated anions form hydrogen bonds in solution, which can
considerably slow down allylic alkylation reactions in dialky-
limidazolium ionic liquids since the anions are not efficient
bases and thus do not deprotonate the nucleophile reagent,
slowing down the rate of nucleophilic attack.8
Here, we report the use of the chiral diphosphite ligands L1
and L2 derived from carbohydrates (Fig. 1) in the Pd-catalysed
asymmetric allylic alkylation, amination and phosphination
reactions of rac-3-acetoxy-1,3-diphenyl-1-propene (rac-I) in neat
imidazolium- and pyrrolidinium-based ionic liquids (Scheme 1).
The recycling of the catalyst dissolved in the ionic liquid phase
was also studied for the allylic amination reaction. The two
PdL catalytic systems (PdL1 and PdL2) were generated in situ
from [Pd(m-Cl)(Z3-C3H5)]2 in the presence of the appropriate
ligand (L/Pd = 1.25/1) (see the corresponding notez).
An important remaining challenge in this asymmetric reac-
tion is the recycling of the catalyst. In this context, the use
of a new reaction medium such as ionic liquids (IL) is of
critical importance. To date, only a few studies on Pd-catalysed
allylic substitutions in ILs were reported and in general, only
imidazolium-based ionic liquids were used.6 Furthermore,
The alkylation (NuH = A) and amination (NuH = B)
of rac-I were evaluated in 1-butyl-3-methyl imidazolium hexa-
fluorophosphate (IL1) and N-butyl-N-methyl pyrrolidinium
a
´
Universite de Toulouse, UPS, LHFA, 118 route de Narbonne,
31062 Toulouse, CNRS, LHFA UMR 5069, 31062 Toulouse,
France. E-mail: gomez@chimie.ups-tlse.fr, teuma@chimie.ups-tlse.fr;
Fax: +33 561558204; Tel: +33 561557738
b
c
´
´
`
Departament de Quımica Fısica i Inorganica, Universitat Rovira i
Virgili, C/ Marcel.li Domingo s/n 43007 Tarragona, Spain.
E-mail: carmen.claver@urv.cat; Fax: +34 977 559563;
Tel: +34 977 559575
´
Departament de Quımica Analıtica i Quımica Organica, Universitat
Rovira i Virgili, C/ Marcel.li Domingo s/n 43007 Tarragona, Spain
´
´
`
Scheme 1 Asymmetric allylic substitutions using dimethyl malonate
(A), benzylamine (B) and diphenylphosphine (C) as nucleophiles in
ionic liquids IL1 and IL2.
w Electronic supplementary information (ESI) available: Full experi-
mental details for the allylic substitutions and Scheme S1. See DOI:
10.1039/c1cc12157j
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 7869–7871 7869

View Article Online
Table 1 PdL1-catalysed allylic alkylation and amination of rac-I
in ILa
(R)-I and (S)-I is lower than that observed in dichloromethane
(in IL2: k(R)/k(S) = 8.6, data from entry 3 in Table 2; in
CH2Cl2, k(R)/k(S) = 49, data from entry 4 in Table 2).11
For amination reactions in IL2, PdL2 gave higher
asymmetric induction than that observed with PdL1, both
systems showing similar activity (entries 1 and 2, Table 3). In
contrast with the results obtained in allylic alkylation reactions,
both catalytic systems were more active in IL2 that in dichloro-
methane (entries 3 and 4 vs. 1 and 2, Table 3). However the
enantioselectivities were slightly lower in IL2 than in CH2Cl2,
following the same trend as that observed for the allylic
alkylation (Table 2).
Entry
Solvent
NuH
Conv.b (%)
eec (%)
1d,e
2d,e
3f
IL1
IL2
IL1
IL2
CH2Cl2
CH2Cl2
A
A
B
B
A
B
100e
100e
69
87
83
95(S)
92(S)
91(R)
82(R)
90(S)
499(R)
4f
5g
6
63
a
Results from duplicated reactions. Reaction conditions: rac-I/Pd =
b
100, 30 min at 40 1C. Determined by 1H NMR. Determined by
HPLC. rac-I/BSA/A = 1/3/3 with addition of a pinch of KOAc;
c
d
e
BSA = Bis(trimethylSilyl) Acetamide. Reaction time = 5 minutes.
f
g
rac-I/B = 1/1. From ref. 10: reaction time = 90 min; rac-I/Pd = 50.
For the allylic amination, the PdL2 catalytic system could
be recycled up to nine times without activity loss, preserving
the enantioselectivity up to eleven times (Fig. 2). The product
was quantitatively extracted with pentane after each run. It is
important to note that the palladium content on the extracted
organic product was less than 0.1 ppm (determined by ICP-MS)
for the first two runs and no palladium was detected for the
next runs. The activity loss after the ninth run is attributed to
the catalyst deactivation.
bis(trifluoromethylsulfonylamide) (IL2) using PdL1 as catalyst
(Table 1). Concerning the allylic alkylation, no remarkable
catalytic differences were observed in both ionic liquids
(entries 1 and 2, Table 1). In relation to the amination
reaction, higher activity was obtained in IL2 than that
observed in IL1 (entries 3 and 4, Table 1), probably due to
the formation of hydrogen bonds between the imidazolium
cation and benzylamine, decreasing then its nucleophilic
behaviour.8 Comparing with the catalytic data obtained in
dichloromethane (entries 5 and 6, Table 1), both alkylation
and amination were more active in IL2 than in CH2Cl2. With
the aim to avoid the plausible formation of carbene complexes
by hydrogen activation of the C(2)–H bond of the imidazolium
cation in basic medium,9 only the pyrrolidinium-based ionic
liquid IL2 was used in the subsequent experiments.
Asymmetric allylic phosphination can stand for an appropriate
means to synthesise enantiomerically enriched phosphines.12
In order to evaluate the catalytic activity of both systems
PdL1 and PdL2, allylic substitutions were performed in IL2
with a different molar ratio rac-I/Pd (Tables 2 and 3). For
alkylation reactions, PdL2 led to higher activity and enantio-
selectivity than the PdL1 catalytic system (entries 1 and 2,
Table 2). However, at a high rac-I/Pd ratio, the activity was
lower than that achieved in dichloromethane (entry 3 vs. 4,
Table 2). In addition, in IL2 the relative rate difference between
Fig. 2 PdL2-catalysed allylic amination recycling in IL2. Conversions
are represented in grey and ee in black.
a
Table 2 PdL1 and PdL2-catalysed allylic alkylation in IL2 and CH2Cl2
Entry
L
rac-I/Pd
Solvent
Conv.b (%)
ee(II)c (%)
ee(I)c (%)
TOF/hꢀ1
1
2
L1
L2
L2
L2
1000
1000
10 000
10 000
IL2
IL2
IL2
CH2Cl2
96
100
21
83(S)
92(S)
96(S)
98(S)
—
—
20(S)
99(S)
—
—
8400
22 000
3
4d,e
55
a
Results from duplicated reactions. Reaction conditions: rac-I/BSA/A = 1/3/3 with addition of a pinch of KOAc at 40 1C. Reaction time for
entries 1 and 2: 30 min and for entries 3 and 4: 15 min. Determined by 1H NMR. Determined by HPLC. Temperature reaction = 25 1C.
From ref. 4.
b
c
d
e
a
Table 3 PdL1 and PdL2-catalysed allylic amination in IL2 and CH2Cl2
Entry
L
rac-I/Pd
Solvent
Time/min
Conv.b (%)
ee(III)c (%)
TOF/hꢀ1
1
2
L1
L2
L1
L2
500
500
100
2500
IL2
IL2
CH2Cl2
CH2Cl2
60
60
15
60
81(62)d
75(53)d
55
88(R)
91(R)
499(R)
98(R)
3720e
3180e
220
3f
4f,g
16
400
a
b
c
Results from duplicated reactions. Reaction conditions: rac-I/B = 1/1 at 40 1C. Determined by 1H NMR. Determined by HPLC.
d
e
f
g
In brackets, conversion at 5 min. Determined at 5 min. Temperature reaction = 25 1C. From ref. 4.
c
7870 Chem. Commun., 2011, 47, 7869–7871
This journal is The Royal Society of Chemistry 2011

View Article Online
Table 4 Pd/L2-catalysed allylic phosphinationa
Entry Solvent Precursor
Time/h T/1C Conv.b (%) eec (%)
Notes and references
z General catalytic procedure: the catalytic precursor was generated
in situ from [Pd(Z3-C3H5)Cl]2 and the appropriate ligand13 (Pd/L =
1/1.25 for alkylation and amination, and Pd/L = 1/2.5 for phosphination)
in 1 cm3 of solvent for 30 minutes before adding the substrate rac-I
and the appropriate nucleophile (rac-I/A = 1/3, rac-I/B = 1/1 and
rac-I/C = 1/1). The mixture was stirred at 40 1C (amination and
alkylation reaction) or 25 1C (phosphination). At the end of the
reaction, the products were extracted with pentane and filtered over
celite. The solvent was then removed under reduced pressure and the
products analyzed by NMR and HPLC.
1
2
3
IL2
IL2
IL2
IL2
IL2
[Pd2(dba)3]
Pd(OAc)2
[Pd(Cl)(C3H5)]2 24
[Pd(Cl)(C3H5)]2
[Pd(Cl)(C3H5)]2
24
24
40
40
40
25
25
25
65d
95
100
100
55
5
0
0
13
15
10
4
6
6
6
5e
6e
CH2Cl2 [Pd(Cl)(C3H5)]2
50
a
See Scheme S1 in ESI.w Results from duplicated reactions. Reaction
conditions: rac-I/C/L2/Pd = 20/20/1.25/1. Determined by 1H NMR.
c
b
d
Determined by chiral SFC. IV/V = 70/30. rac-I/C/L2/Pd =
e
1 (a) B. M. Trost and C. Lee, in Catalytic Asymmetric Synthesis,
ed. I. Ojima, Wiley-VCH, New York, 2000, vol. 8E, 593; (b) J. Tsuji,
Palladium reagents and Catalysts-Innovations in Organic Synthesis,
Wiley, Chichester, 1995; (c) B. M. Trost and M. L. Crawley, Chem.
Rev., 2003, 103, 2921; (d) Z. Lu and S. Ma, Angew. Chem., Int. Ed.,
2008, 47, 258.
200/200/2.5/1.
We then decided to study the allylic phosphination of rac-I
using diphenylphosphine as nucleophile, based on the
previously reported work by Togni et al.12a [Pd2(dba)3],
[Pd(m-Cl)(Z3-C3H5)]2 and Pd(OAc)2 were used as catalytic
precursors in the presence of ligand L2 (Table 4). In IL2 at
40 1C, the desired allylic phosphine IV was exclusively
obtained using palladium acetate and [Pd(m-Cl)(Z3-C3H5)]2
as catalytic precursors (entries 2 and 3, Table 4). How-
ever using [Pd2(dba)3], the side product corresponding to
1,3-(diphenylpropyl)diphenylphosphine, V, was also observed
(entry 1, Table 4). Unfortunately, no enantioselectivity was
induced in any case. At lower temperature and shorter reaction
time, the enantioselectivity increased up to 13% (entry 4,
Table 4). The enantioselectivity remained unchanged even at
lower conversion using lower Pd load (0.5 mol%) (entry 5 vs. 4,
Table 4). In fact, the catalytic behaviour in dichloromethane
was close to that observed in IL2 (entries 5 and 6, Table 4).
The low asymmetric induction obtained could be due to
the competition between phosphines, involved in the allylic
phosphination, and the diphosphite L2, in contrast to the
robust catalytic system containing Josiphos-type ligands.12a
In conclusion, the chiral palladium catalytic system containing
the diphosphite ligand L2 is highly active and induces high
enantioselectivities in pyrrolidinium-based ionic liquid in the
allylic alkylation and amination reactions. Its activity even
increases in relation to that obtained in dichloromethane for
the amination. For the first time, the Pd/L2 system could be
recycled up to nine times without activity loss, preserving the
enantioselectivity. The preliminary results obtained for the
allylic phosphination are very promising in terms of chemo-
selectivity, and encourage us to look for suitable catalytic
systems to induce enantioselectivity.
2 (a) M. Dieguez and O. Pamies, Acc. Chem. Res., 2010, 43, 312;
´
(b) K. N. Gavrilov, O. G. Bondarev and A. I. Polosukhin, Russ.
Chem. Rev., 2004, 73, 671.
3 (a) M. Die
70, 3363; (b) M. Die
Catal., 2005, 347, 1257; (c) A. Gual, S. Castillo
M. Dieguez and C. Claver, Dalton Trans., 2011, 40, 2852–2860.
4 A. Balanta Castillo, I. Favier, E. Teuma, S. Castillo
´
guez, O. Pamies and C. Claver, J. Org. Chem., 2005,
guez, O. Pamies and C. Claver, Adv. Synth.
n, O. Pamies,
´
´
´
´
n, C. Godard,
A. Aghmiz, C. Claver and M. Gomez, Chem. Commun., 2008, 6197.
´
5 S. E. Lyubimov, V. A. Davankov, M. G. Maksimova, P. V. Petrovskii
and K. N. Gavrilov, J. Mol. Catal. A: Chem., 2006, 259, 183.
6 Selected references: (a) L. Leclercq, I. Suisse, G. Nowogrocki and
F. Agbossou-Niedercorn, Green Chem., 2007, 9, 1097; (b) J. Ross,
W. Chen, L. Xu and J. Xiao, Organometallics, 2001, 20, 138;
(c) W. Chen, L. Xu, C. Chatterton and J. Xiao, Chem. Commun.,
1999, 1247.
7 (a) S. E. Lyubimov, V. A. Davankov and K. N. Gavrilov,
Tetrahedron Lett., 2006, 47, 2721; (b) I. Kmentova, B. Gotov,
E. Solcaniova and S. Toma, Green Chem., 2002, 4, 103;
(c) R. Moucel, K. Perrigaud, J.-M. Goupil, P.-J. Madec,
S. Marinel, E. Guibal, A.-C. Gaumont and I. Dez, Adv. Synth.
Catal., 2010, 352, 433.
8 (a) J. Ross and J. Xiao, Chem.–Eur. J., 2003, 9, 4900;
(b) L. Cammarata, S. G. Kazarian, P. A. Salter and T. Welton,
Phys. Chem. Chem. Phys., 2001, 3, 5192.
9 (a) L. Starkey Ott, M. L. Cline, M. Deetlefs, K. R. Seddon and
R. G. Finke, J. Am. Chem. Soc., 2005, 127, 5758; (b) J. Dupont and
J. Spencer, Angew. Chem., Int. Ed., 2004, 43, 5296; (c) D. Bacciu,
K. J. Cavell, I. A. Fallis and L.-l. OOi, Angew. Chem., Int. Ed., 2004,
43, 1277; (d) D. S. McGuinness, K. J. Cavell, B. F. Yates,
B. W. Skelton and A. H. White, J. Am. Chem. Soc., 2001, 123, 8317.
´
10 S. Jansat, M. Gomez, K. Philippot, G. Muller, E. Guiu, C. Claver,
S. Castillon and B. Chaudret, J. Am. Chem. Soc., 2004, 126, 1592.
´
11 k(R)/k(S) = ln[(1 ꢀ C/100)(1 ꢀ ee/100)]/ln[(1 ꢀ C/100)(1 + ee/100)],
where C = conversion and ee = enantiomeric excess of the
substrate (I), see: C.-H. Chen, Y. Fujimoto, G. Girdaukas and
C. J. Sih, J. Am. Chem. Soc., 1982, 105, 7294.
12 (a) P. Butti, R. Rochat, A. D. Sadow and A. Togni, Angew. Chem.,
Int. Ed., 2008, 47, 4878; (b) N. F. Blank, J. R. Moncarz, T. J.
Brunker, C. Scriban, B. J. Anderson, O. Amir, D. S. Glueck,
L. N. Zakharov, J. A. Golen, C. D. Incarvito and A. L. Rheingold,
J. Am. Chem. Soc., 2007, 129, 6847.
Authors are grateful to CNRS (Centre National de la
Recherche Scientifique), Universite Paul Sabatier, the Spanish
´
13 For the synthesis of ligand L1, see: G. J. H. Buisman,
M. E. Martin, E. J. Vos, A. Klootwijk, P. C. J. Kamer and
P. W. N. M. van Leeuwen, Tetrahedron: Asymmetry, 1995,
6, 719. For L2, see: M. R. Axet, J. B. Benett-Buchholz,
C. Claver and S. Castillon, Adv. Synth. Catal., 2007, 349, 1983.
Ministerio de Educacion y Ciencia (CTQ2007-62288/BQU,
Consolider Ingenio 2010, Intecat-CSD2006-0003 and ‘‘Ramon y
Cajal’’ fellowship to C.G.) and the Generalitat de Catalunya
(2009SGR116) for financial support.
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 7869–7871 7871