Preparation and application of polymer-supported π-allylpalladium complex as a chiral catalyst in the asymmetric allylic alkylation

Article abstract of DOI:10.1016/S1381-1169(01)00198-4

Chiral ferrocenylphosphine-imines A-D were synthesized and grafted to polystyrene to afford polymeric ligands PA-PD. Ligands A-D and PA-PD were used as ligands in the palladium-catalyzed asymmetric allylic alkylation. In the homogeneous reactions, the best result was obtained when the ligand A was used (99% yield and 82% ee). In the heterogeneous reactions, the highest ee value (73%) in the first run was obtained when PA was used as a ligand. Considering the enantioselectivity and reusability, the best result was obtained in the case of PC. The recycling of the polymer-supported π-allylpalladium catalyst gave unsatisfactory results: it can be reused three times with moderate enantioselectivities (60-70% ee).

Full text of DOI:10.1016/S1381-1169(01)00198-4

Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
Preparation and application of polymer-supported ␲-allylpalladium
complex as a chiral catalyst in the asymmetric allylic alkylation
H.-J. Park, Jin Wook Han, Hwimin Seo, Hye-Young Jang,
Young Keun Chung∗, Junghun Suh
School of Chemistry and Center for Molecular Catalysis, Seoul Nation University, Seoul 151-747, South Korea
Received 26 February 2001; received in revised form 2 May 2001; accepted 2 May 2001
Abstract
Chiral ferrocenylphosphine-imines A–D were synthesized and grafted to polystyrene to afford polymeric ligands PA–PD.
Ligands A–D and PA–PD were used as ligands in the palladium-catalyzed asymmetric allylic alkylation. In the homogeneous
reactions, the best result was obtained when the ligand A was used (99% yield and 82% ee). In the heterogeneous reactions,
the highest ee value (73%) in the first run was obtained when PA was used as a ligand. Considering the enantioselectivity and
reusability, the best result was obtained in the case of PC. The recycling of the polymer-supported ␲-allylpalladium catalyst
gave unsatisfactory results: it can be reused three times with moderate enantioselectivities (60–70% ee). © 2001 Elsevier
Science B.V. All rights reserved.
Keywords: Allylic alkylation; Enantioselective; Palladium catalyst; Polymer-supported
1. Introduction
area aim to narrow the gap between the homogeneous
and heterogeneous catalysis approaches. Recently, the
first asymmetric reaction with polymer-supported chi-
ral ␲-allylpalladium catalysts have been reported by
Yamamoto et al. [12]. Their catalyst was very stable
and could be reused four times without losing catalytic
activity. However, the enantioselectivities were rather
low (up to 47% ee).
The palladium-catalyzed allylic alkylation is
a synthetically important process in which new
carbon–carbon bonds are formed. With chiral cata-
lysts, high enantioselectivities have been often ob-
served. Ferrocenyl ligands represent one of the most
important classes for auxiliaries. They find application
in a variety of transition-metal catalyzed reactions.
We set-up a study aimed at incorporating ferrocenyl
ligands into polymeric structures. This field has been
pioneered by Hayashi et al., who first prepared pla-
nar chiral ferrocenyl phosphines [13]. The bidentate
The use of polymers as supports has found many
applications in catalysis [1–11]. The strategy of at-
taching a chiral ligand onto a polymer support offers
several advantages in catalytic asymmetric synthesis
over the use of homogeneous ligands. The advantages
include easy separation, potential recycling of expen-
sive chiral catalysts, and the possibility of carrying out
the desired transformation in a continuous-flow sys-
tem. Although chiral ligands attached to polymers are
useful and practical, there are still some shortcomings
such as lower catalytic activity and enantioselectivity
compared with those recorded for their homogeneous
counterparts. Thus, continuous research efforts in this
∗ Corresponding author. Tel.: +82-2-880-6662;
fax: +82-2-889-1568.
E-mail address: ykchung@plaza.snu.ac.kr (Y.K. Chung).
1381-1169/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S1381-1169(01)00198-4

152
H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
ligands of the type A are characterized by the presence
of two ligand atoms (P, N) imbedded in different and
easily modifiable steric and electronic environments.
We report herein the preparation of polymer-supported
ferrocenyl P, N-chelate ligands and their application
in the palladium-catalyzed allylic alkylation. Moberg
et al. also reported enantioselective allylic alkylation
using polymer-supported palladium catalysts [14].
Their catalytic system shows high enantioselectivities
(up to 80% ee). However, they did not discuss the
reusability of the catalytic system.
2.3. Ligand A
Molecular sieve (4 Å) in a 50 ml of two-neck
flask was flame-dried. To the flask were added 1
(0.20 g, 0.5 mmol), 20 ml of CH2Cl2, and benzy-
lamine (0.06 ml, 0.55 mmol). The resulting solution
was heated at reflux for 6 h. The color of the solu-
tion was changed from red to bright orange. After
the solution was cooled to room temperature, the
solution was filtered over a celite pad. The filtrate
was evaporated to dryness. Recrystallization of the
residue in Et2O/pentane yielded orange solids (0.16 g,
0.37 mmol, 74%). Melting point (mp) 132–135◦C; IR
ν(CN) 1645 cm−1; 1H NMR: δ 8.55 (s, 1H), 7.59–7.13
(m, 15H), 5.11 (s, 1H), 4.68–4.52 (d, 13.3 Hz, 2H),
4.50 (m, 1H), 4.11 (s, 5H), 3.86 (m, 1H) ppm. Anal.
found: C, 73.93; H, 5.38; N, 2.87. C30H26FeNP calcd.:
C, 74.13; H, 5.46; N, 2.85; [α]2D5 340 (c 0.1, CH2Cl2).
2. Experimental
2.1. General
1
The H NMR spectra were recorded in CDCl3 at
300 MHz (Bruker DPX-300) unless otherwise stated.
Combustion analysis (CE Instrument EA 1110) and
ICP-AES (Shimadzu ICPS-1000IV) were done at the
National Center for Inter-University Research Facili-
ties in Seoul National University. Enantiomeric excess
2.4. Ligand B
The same procedure as the synthesis of A was em-
ployed except NH2CH(C6H5)(C6H4CH3-p) instead of
benzylamine. Yield 69%; mp 146–148◦C; IR ν(CN)
1
were determined by H NMR using Eu(hfc)3 as the
chiral shift reagent. Merck Kieselgel 60H was used for
flash chromatography. Polystyrene resin cross-linked
with 1% divinylbenzene was used as a polymeric sup-
port. Compounds 1 and amines were prepared accord-
ing to the literature procedures [15–18]. Aminated
poly(styrene-co-divinylbenzene) (1% cross-linked,
200–400 mesh) was purchased from Aldrich Chem.
Co. Polymers, (aminomethyl)polystyrene (AMPS)
and 4-methylbenzhydrylamine (MBHA), were pre-
pared by the previously reported methods [19]. The
ee values were determined by 1H NMR using a chiral
shift reagent Eu(hfc)3.
1
1628 cm−1; H NMR: δ 8.51 (s, 1H), 7.56–7.02 (m,
19H), 5.37 (s, 1H), 5.11 (s, 1H), 4.46 (m, 1H), 4.00
(s, 5H), 3.80 (s, 1H), 2.30 (s, 3H) ppm. Anal. found:
76.96; H, 5.59; N, 2.43. C37H32FeNP calcd.: C, 76.92;
H, 5.63; N, 2.42; [α]2D7 109 (c 0.1, CH2Cl2).
2.5. Ligand C
The same procedure as the synthesis of A was
employed except 2 instead of 1. Yield 72%; mp
134–135◦C; IR ν(CN) 1640 cm−1; H NMR: δ 8.42
1
(s, 1H), 7.65–7.07 (m, 13H), 5.04 (s, 1H), 4.68 (s, 1H),
4.51 (s, 2H), 4.15 (s, 5H), 3.75 (s, 1H) ppm. Anal.
found: 61.66; H, 3.88; N, 2.25. C32H24FeNP calcd.:
C, 61.87; H, 3.88; N, 2.26; [α]2D5 15 (c 0.1, CH2Cl2).
2.2. Synthesis of 2
A procedure identical to that for 1 was applied. The
IR ν(CO) 1683 cm−1; 1H NMR (CDCl3): δ 10.15 (d,
2.13 Hz, 1H), 7.70–7.61 (m, 4H), 7.53–7.51 (d, 8.43
Hz, 2H), 7.31–7.22 (m, 2H), 5.13 (t, 1.34 Hz, 1H),
4.74 (t, 2.67 Hz, 1H), 4.26 (s, 5H), 4.00 (t, 1.42 Hz,
1H) ppm; 13C NMR (CDCl3): δ 193.1, 136.0, 135.9,
135.7, 133.3, 133.0, 132.8, 132.5, 126.3, 124.5, 76.4,
74.7, 73.7, 70.3, 70.0 ppm; HRMS for C25H17F6FeOP
calcd. 534.0271, found 534.0268; [α]1D5 130 (c 0.1,
CH2Cl2).
2.6. Ligand D
The same procedure as the synthesis of A was
employed except 2 instead of 1. Yield 80%; mp
139–141◦C; IR ν(CN) 1642 cm−1; H NMR: δ 8.47
1
(s, 1H), 7.72–6.79 (m, 17H), 5.33 (s, 1H), 4.92 (s,
1H), 4.45 (s, 1H), 4.08 (s, 5H), 3.70 (s, 1H), 2.30
(s, 3H) ppm. Anal. found: 66.01; H, 4.50; N, 1.96.

H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
153
C39H30F6FeNP calcd.: C, 65.65; H, 4.24; N, 1.96;
MBHA was used. In the same way as the above, the
contents of P, N-chelate ligand in PB–PD were deter-
mined: PB (0.13 mmol/g), PC (0.08 mmol/g), and PD
(0.06 mmol/g).
[α]2D5 199 (c 0.1, CH2Cl2).
2.7. Polymer-supported ligand PA
2.9. Catalytic reaction
Schlenk flask was flame-dried. To the flask AMPS
(1.0 g, 1 mmol/g) and 25 ml of benzene were added.
The polymer was allowed to be swollen for several
hours. A solution of A (80 mg, 0.2 mmol) in 5 ml of
benzene was transferred via cannula to the polymer
in benzene solution. The flask was fitted with a soxh-
let and the solution was heated at reflux for 12 h. Af-
ter the solution was cooled to room temperature, the
resin was filtered, washed successively with benzene
(10 ml), CH2Cl2 (10 ml), and methanol (20 ml), and
dried in vacuo to give bright orange solid PA.
The active catalysts were then prepared by reaction
of PA with [(C3H5)PdCl]2 dissolved in THF. After the
solution was stirred for 1 h and washed with a large ex-
cess of solvent, solid catalysts were isolated. Then the
content of Pd (0.18 mmol/g) coordinated to the solid
support PA was determined by the substraction of the
amount of Pd (determined by ICP-AES) in the filtrate
from the original amount of Pd. We presumed that the
contents of Pd coordinated to the solid support PA was
equal to the amount of the P, N-chelate ligand in PA.
A representative procedure for the allylation re-
action of rac-1,3-diphenyl prop-2-en-1-yl acetate is
as follows: to a mixture of polymer-supported chiral
␲-allylpalladium catalyst (2 mg, 0.005 mmol Pd) and
substrate (0.145 mg, 0.58 mmol), sodium malonate
(generated in situ by the reaction of NaH (30 mg) with
dimethyl malonate (0.085 ml, 0.74 mmol) in 10 ml of
THF at 25◦C was added. The mixture was stirred for
24 h, and then filtered. The progress of the reaction
was monitored by TLC. The catalyst was washed
with THF, and the filtrates were combined. The sol-
vent was removed under reduced pressure, and the
crude product was purified by silica gel column chro-
matography (hexane:diethyl ether = 10:1), giving
methyl-2-carbomethoxy-3,5-diphenylpent-4-enolate
1
in 86% yield. The ee values were determined by H
NMR using a chiral shift reagent Eu(hfc)3.
3. Results and discussion
2.8. Preparation of polymer-supported
ligands PB, PC, and PD
3.1. Preparation of planar chiral ferrocenyl P,
N-chelate ligand
The same procedure as the preparation of PA was
applied. For the synthesis of PC, polymer AMPS was
used and for the synthesis of PB and PD polymer
Recently, we reported the use of planar chiral
imino-phosphine chromium ligands in palladium-
Scheme 1.

154
H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
catalyzed asymmetric allylic alkylation [20]. In the
reaction, high yields up to >98% ee were obtained.
Thus, first we attempted to use the planar chiral
imino-phosphine chromium ligands in the prepara-
tion of polymer-supported catalysts (For the first run,
the yield and enantioselectivity were 35 and 12%,
respectively). However, due to the facile decompo-
sition of the chromium tricarbonyl moiety during
the catalytic reaction, planar chiral imino-phosphine
chromium ligands were not used in the prepara-
tion of the polymer-supported catalyst system. In-
stead, ferrocenyl ligands A–D were used as ligands
(Scheme 1).
(2)
The syntheses of 4-methylbenzhydrylamine
polystyrene (MBHA) and aminomethyl polystyrene
(AMPS) were previously described by Hudson et al.
The planar chiral P, N-ligands A–B were synthe-
sized from 1 (S)-␣-(diphenylphosphino) ferrocenecar-
boxaldehyde (Eq. (1)).
(1)
[19]. These polymers have been reported to possess
excellent swelling characteristics. After careful iso-
lation of the light orange solid-supported ligands,
they were characterized by infrared spectroscopy and
elemental analysis. The active catalysts were then
prepared by reaction of the polymer-supported ligand
with [(allyl)PdCl]2 dissolved in THF. Scheme 2 shows
the polymeric ligands synthesized in this study. The
ICP-AES analysis indicated that the contents of Pd
were highly dependent upon the polymer-supported
ligand: 0.18 mmol/g for PA, 0.13 mmol/g for
PB, 0.08 mmol/g for PC, and 0.06 mmol/g for
PD.
Compound 1 was previously prepared by Kagan
et al. [15]. Treatment of 1 with primary amine in the
presence of molecular sieves (4 Å) gave A and B in
74 and 69% yields, respectively. Due to the hydrolysis
of the imino group, chromatographic purification of A
and B on a silica gel column led to the conversion of
A and B to 1. Thus, A and B were purified by recrys-
tallization. In the same way, C and D were obtained
in 72 and 80% yields, respectively, from 2.
3.2. Preparation of polymeric ligands
To retain the same chiral environment as in the
homogeneous catalysis, the polymer was introduced
through the phenyl group on the imnino moiety. The
polymer-supported ligands PA–PD were prepared by
reaction of amino-functionalized polystyrenes with an
excess of ferrocenecarboxaldehyde in toluene at reflux
for 12 h (Eq. (2)).
3.3. Catalytic reaction
Before investigating the polymer-supported cat-
alysts, the reaction of the corresponding homoge-
neous catalysts was studied. The palladium-catalyzed
asymmetric allylic alkylation of rac-1,3-diphenyl

H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
155
Scheme 2.
Table 1
Use of ligands A–D in Pd-catalyzed allylic alkylation
Entry
Base
L∗
Solvent
Time (h)
Yield (%)a
ee (%)b
1
2
3
4
5
6
7
8
BSA
NaH
BSA
NaH
BSA
NaH
BSA
NaH
A
A
B
B
C
C
D
D
CH2Cl2
DMF
CH2Cl2
DMF
CH2Cl2
DMF
CH2Cl2
DMF
4
5
7.5
5.5
4
7
6
7
99
32
86
96
97
40
73
35
82
80
74
61
76
68
80
40
a Isolated yield.
b The ee values were determined by 1H NMR using chiral shift reagent, Eu(hfc)3.
prop-2-en-1-yl acetate with dimethyl malonate was
carried out (Eq. (3) and Table 1).
ues (74–82%) were rather insensitive to the ligand,
and the conversion yields (73–99%) were slightly
(3)
The nucleophile was generated from dimethyl mal-
onate in the presence of N,O-bis(trimethylsilyl)aceta-
mide (BSA) and a catalytic amount of KOAc or
sodium hydride. Our preliminary results showed that
CH2Cl2 was the preferred choice for the reaction
medium for BSA and DMF for NaH. The absolute
configuration of the product was assigned as R from
specific rotation determination and comparison with
literature data. When BSA was used, the ee val-
dependent upon the ligand. However, when NaH was
used, the ee values (40–80%) and conversion yields
(32–96%) were highly dependent upon the ligand.
The best result was obtained when the ligand A and
BSA were used. Thus, neither the increase of the
steric bulkiness of imine nor the introduction of the
electron-withdrawing group helps increase the ee
value and conversion rate.

156
H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
Table 2
Use of ligands PA–PD in Pd-catalyzed allylic alkylationa
Entry
Run
Base
L∗
Solvent
Time (day)
Yield (%)a
ee (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1
2
3
1
2
1
2
1
2
1
2
3
1
2
1
2
3
1
BSA
BSA
BSA
NaH
NaH
BSA
BSA
NaH
NaH
BSA
BSA
BSA
NaH
NaH
BSA
BSA
BSA
NaH
PA
PA
PA
PA
PA
PB
PB
PB
PB
PC
PC
PC
PC
PC
PD
PD
PD
PD
CH2Cl2
CH2Cl2
CH2Cl2
DMF
2
2
2
2
2
3
3
3
3
2
3
5
2
3
3
5
7
3
95
64
24
72
31
82
82
86
86
57
55
44
24
14
51
50
15
52
67
59
59
73
66
27
25
67
64
70
66
60
69
21
61
61
18
57
DMF
CH2Cl2
CH2Cl2
DMF
DMF
CH2Cl2
CH2Cl2
CH2Cl2
DMF
DMF
CH2Cl2
CH2Cl2
CH2Cl2
DMF
a Isolated yield.
b The ee values were determined by 1H NMR using chiral shift reagent, Eu(hfc)3.
The polymer-supported palladium catalyst-cataly-
zed reaction of rac-1,3-diphenyl prop-2-en-1-yl ac-
etate with dimethyl malonate was carried out (Table 2).
The absolute configuration of product, (R), is the
same as would be anticipated from the results of
the homogeneous reaction. When BSA was used as
a base, in the first run the yields (51–95%) and ee
values (27–70%) were highly dependent upon the P,
N-chelate ligand. Comparing the yields and ee values,
the best result was obtained when PA was used as
a ligand. However, when the catalyst was reclaimed
by filtration from the reaction mixture and was then
reused with a new batch under identical conditions
in the second run, the yield was decreased to 64%
with a 59% ee value. In the third run, the yield was
further decreased to 24% but the ee value was the
same as for the second run. As the run continued, the
solution after filtration became brown and eventually
dark brown. When we considered only the enantios-
electivity, the best result was obtained in the case of
PC: 70% in the first run, 66% in the second run, and
60% in the third run. When NaH was used as a base,
the yields (24–86%) were highly dependent upon the
ligand but the ee values (57–73%) were relatively
insensitive. The highest ee value (73%) in the first
run was obtained when PA was used as a ligand.
This is less than the value of 82% obtained in the
homogeneous system. For the second run, the best
result was obtained when PB was used as a ligand.
When we compare the results of the first run for each
polymer-supported ferrocenyl P, N-chelates, the best
yield was obtained for polymer PA in CH2Cl2 and
the best optical yield for polymer PA in DMF. Thus,
as in the case of a homogeneous system neither the
increase of the steric bulkiness of imine nor the intro-
duction of electron-withdrawing group help increase
the ee value and conversion rate.
4. Conclusion
In conclusion, we have prepared the polymer-
supported ␲-allylpalladium catalyst for the asymmet-
ric allylic alkylation. High ee’s and conversions were
obtained in the first run but decreased in the next runs.
The catalyst can be reused three times with moderate
enantioselectivities (60–70% ee).
Acknowledgements
This work was supported by grant No. 2000-2-
12200-001-1 from the Basic Research Program of the

H.-J. Park et al. / Journal of Molecular Catalysis A: Chemical 174 (2001) 151–157
157
Korea Science and Engineering Foundation (KOSEF)
and (1999-1-122-001-5) and the KOSEF through
the Center for Molecular Catalysis. Jin Wook Han,
Hwimin Seo and Hye-Young Jang thank the Brain
Korea 21 fellowships.
[9] B. Altava, M.I. Burguete, B. Escuder, S.V. Luis, R.V. Salvador,
J. Org. Chem. 62 (1997) 3126.
[10] C. Dreisbach, G. Wischnewski, U. Kragl, C. Wandrey, J.
Chem. Soc., Perkin Trans. 1 (1995) 875.
[11] Y. Uozumi, H. Danjo, T. Hayashi, Tetrahedron Lett. 39 (1998)
8303.
[12] M. Bao, H. Nakamura, Y. Yamamoto, Tetrahedron Lett. 41
(2000) 131.
[13] T. Hayashi, K. Yamamoto, K. Kumada, Tetrahedron Lett.
(1974) 4405.
References
[1] P.H.U. Herkens, H.C.J. Ottenheijm, D.C. Rees, Tetrahedron
53 (1997) 5643.
[14] K. Hallman, E. Macedo, K. Nordstrom, C. Moberg,
Tetrahedron: Asym. 10 (1999) 4037.
[2] P.H.H. Herkens, H.C.J. Ottenheijm, D.C. Rees, Tetrahedron
52 (1996) 4527.
[15] O. Riant, O. Samuel, T. Flessner, S. Taudien, H.B. Kagan, J.
Org. Chem. 62 (1997) 6733.
[3] D. Gravert, K. Janda, Chem. Rev. 97 (1997) 489.
[4] S. Kobayashi, Chem. Soc. Rev. 28 (1999) 1.
[5] T. Li, K.D. Janda, J.A. Ashley, R.A. Lerner, Science 264
(1994) 1289.
[16] O. Riant, O. Samuel, H.B. Kagan, J. Am. Chem. Soc. 115
(1993) 5835.
[17] A. Pohland, H. Sullivan, J. Am. Chem. Soc. 75 (1953)
5898.
[6] A. Vidal-Ferran, N. Bampos, A. Moyano, M.A. Pericas, A.
Riere, J.K.M. Sandera, J. Org. Chem. 63 (1998) 6309.
[7] D. Seebach, R.E. Marti, T. Hintermann, Helv. Chim. Acta 79
(1996) 1710.
[18] F.J. Weiberth, S.S. Hall, J. Org. Chem. 51 (1986) 5388.
[19] J.H. Adams, R.M. Cook, D. Hudson, V. Jammalamadaka,
M.H. Lyttle, M.F. Songster, J. Org. Chem. 63 (1998)
3706.
[8] L. Canali, E. Cowan, H. Deleuze, C.L. Gibson, D.C.
Sherrington, Chem. Commun. (1998) 2561.
[20] H.-Y. Jang, H. Seo, J.W. Han, Y.K. Chung, Tetrahedron Lett.
41 (2000) 5083.