Evaluation of disaccharide-based ligands for Pd(0)-catalyzed asymmetric allylations

Article abstract of DOI:10.1016/j.tetlet.2007.03.102

A series of phosphines based on disaccharides containing a d-glucosamine framework were prepared and tested for their abilities as ligands for the palladium(0)-catalyzed asymmetric allylic allylation of racemic 1,3-diphenyl-2-propenyl acetate with various nucleophiles. In contrast to previous results exploiting monosaccharides, the iminophosphines generally afforded higher enantiomeric excesses, up to 99%.

Full text of DOI:10.1016/j.tetlet.2007.03.102

Tetrahedron Letters 48 (2007) 3569–3573
Evaluation of disaccharide-based ligands for
Pd(0)-catalyzed asymmetric allylations
a
b
b
b,
*
Sine A. Johannesen, Katarzyna Glegoła, Denis Sinou, Eric Framery and
a,
*
Troels Skrydstrup
a
Department of Chemistry, University of Aarhus, Langelandsgade 140, 8000 Aarhus C, Denmark
Equipe Synth e` se Asym e´ trique, ICBMS UMR CNRS 5246, Universit e´ Claude Bernard Lyon 1, 43,
boulevard du 11 novembre 1918, 69622 Villeurbanne Cedex, France
b
Received 15 February 2007; revised 13 March 2007; accepted 16 March 2007
Available online 21 March 2007
Abstract—A series of phosphines based on disaccharides containing a D-glucosamine framework were prepared and tested for their
abilities as ligands for the palladium(0)-catalyzed asymmetric allylic allylation of racemic 1,3-diphenyl-2-propenyl acetate with var-
ious nucleophiles. In contrast to previous results exploiting monosaccharides, the iminophosphines generally afforded higher enan-
tiomeric excesses, up to 99%.
ꢀ
2007 Elsevier Ltd. All rights reserved.
Asymmetric allylic alkylation catalyzed by palladium(0)
complexes represents a synthetically valuable protocol
for the construction of asymmetric carbon–carbon and
glycosides6 and oxazoline-thioglucose.7 However, the
most successful so far includes a phosphinoaryloxazo-
8
line, as well as a phosphinite-oxazoline ligand, both
1
9
carbon-heteroatom bonds. Although many asymmetric
derived from D-glucosamine.
catalytic systems have been described, Trost’s P–P
2
3
ligand and Pfaltz’s P–N ligand are amongst the best
ligands for these reactions providing enantioselectivities
as high as 95%. Nevertheless, for the achievement of
high catalytic activity as well as high enantioselectivity,
typically several experimental parameters require opti-
mization of which one is the ligand structure.
An alternative and structurally simpler carbohydrate-
based P–O ligand was introduced by Denis Sinou’s
group in 2003 represented by phosphine-amide 1 in Fig-
1
0
ure 1. This ligand is easily prepared from readily avail-
able 1,3,4,6-tetra-O-acetylglucosamine and provides
enantioselectivities up to 97% in the Pd(0)-catalyzed
allylic alkylation of 1,3-diphenyl-2-propenyl acetate
with a variety of nucleophiles. The acetyl groups on
the ligand were necessary for obtaining high ees in these
reactions, since similar experiments carried out with the
nonacetylated derivative 2 led to the formation of an
Within the last few years, there has been an increasing
interest in the application of chiral ligands prepared
from carbohydrates in asymmetric catalysis. The con-
4
figurational and conformational diversity among these
highly functionalized compounds provide excellent basis
for the preparation of new chiral ligands. The design
and fine-tuning of carbohydrate-based ligands is facili-
tated by the multiple functional groups within this class
of compounds. Several sugar ligands have already been
reported in the literature for asymmetric allylic alkyl-
ation including P–N, P–P, S–S, and S–X (X = P or N)
1
0
inactive catalyst.
OAc
O
OH
O
O
AcO
AcO
HO
HO
OAc
OH
NH
NH
5
O
type ligands derived from ferrocenylglucose, bis-thio-
Ph2P
Ph2P
Keywords: Disaccharide; D-Glucosamine; Iminophosphine; Asymmet-
ric allylation; Palladium.
1
2
*
Corresponding authors. Tel.: +45 8942 3932; fax: +45 8619 6199
T.S.); e-mail: ts@chem.au.dk
(
Figure 1.
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.03.102

3570
S. A. Johannesen et al. / Tetrahedron Letters 48 (2007) 3569–3573
In this Letter, we report the influence of an additional
sugar ring at the anomeric position of these carbohy-
drate ligands in this palladium catalyzed allylic alkyl-
ation. In contrast to observations made with the
monosaccharide ligands, the disaccharide phosphine-
imines proved superior to the corresponding phos-
phine-amides, providing higher enantiomeric excesses.
In addition, the acetylation of the glucosamine unit
was not required for the obtention of high
enantioselectivities.
The analogous imino- and aminophosphine ligands 8–
10 and 11–13 were also synthesized in order to compare
their reactivity with 6 and 7, although previous work
with the monosaccharide revealed these P–P or P–N
1
1
11
types of ligands to be inferior to phosphine-amide 1.
Iminophosphines 8 and 9 were prepared in good yields
by the condensation of o-(diphenylphosphino)benzyl-
aldehyde with the corresponding D-glucosamine deriva-
tives 3 and 4 in toluene using MgSO as a drying agent
4
under inert atmosphere. The imino functionality was
then reduced with NaBH CN to provide amines 11
3
In a previous paper, we described the synthesis of three
and 12 with yields in the range of 40–81%. All ligands
derived from disaccharides were purified using silica
gel chromatography under inert atmosphere. Finally,
an amino- and iminophosphine ligands 10 and 13 were
synthesized from C6-deoxy disaccharide 5. Removal of
the benzyloxy group allows a preliminary assessment
of the role of this C6-substituent on the catalyst activity,
to determine whether the overall sterical bulk of the
D-glucose sugar is necessary, or whether there is a possi-
ble participation of a benzyl group in the shielding of
one of the two faces of the allyl palladium(II) cation
intermediate.
disaccharides 3, 4 and 5 all containing a free primary
1
2
amine (Scheme 1). Functionalization of these amines
with different phosphine derivatives gives access to a
variety of monophosphine ligands. Phosphine-amide
ligands 6 and 7 were obtained in yields of 64% and
7
0% respectively, from the coupling of o-(diphenyl-
phosphino)benzoic acid with the primary amine of
1
3
the disaccharide, using BOP as the coupling reagent
in the presence of triethylamine. Compared to phos-
phine-amide ligand 1 generated from D-glucosamine,
the two corresponding ligands 6 and 7 derived from
3
and 4, respectively, possess a bulky sugar unit in
the anomeric position of the D-glucosamine unit
instead of an acetate group. With the presence of the
O-benzyl groups on the D-glucose sugar, it was
expected that only a slight difference in the solubility
properties between the disaccharides would exist, there-
by allowing a comparison of the influence of such
ligands on the reactivity and enantioselectivity of
the catalyst with or without the acetyl groups on the
D-glucosamine unit.
The results of the palladium-catalyzed allylic alkylation
of racemic 1,3-diphenyl-2-propenyl acetate with di-
methyl malonate reaction applying the chiral disaccha-
ride-based ligands are summarized in Table 1. The
reactions were performed in THF (0.125 M) using
3
2 mol % of [Pd(g -C H )Cl] and 4–8 mol % ligand with
3
5
2
1
0
variations in both the reaction temperature and base.
Several interesting trends were noted from this study.
Of the three phosphine ligands 6, 8 and 11 bearing a
OH
O
OR
OR
O
OMe
OMe
BnO
O
BnO
O
Ph2P
O
BnO
O
O
BnO
O
RO
RO
RO
RO
NH2
NH
BOP, NEt3
THF
X
X
3
4
5
R = Ac, X = OBn
R = H, X = OBn
R = Ac, X = H
2
Ph P
6
7
R = Ac, X = OBn 64%
R = H, X = OBn 70%
O
Ph2P
MgSO4
Toluene
OR
OR
OMe
OMe
BnO
O
BnO
O
O
N
BnO
O
NaCNBH3
O
BnO
O
RO
RO
RO
RO
AcOH, MeOH
NH
X
X
Ph2P
Ph2P
8
9
1
R = Ac, X = OBn 80%
R = H, X = OBn 88%
0 R = Ac, X = H 88%
11 R = Ac, X = OBn 81%
12 R = H, X = OBn 46%
13 R = Ac, X = H 40%
Scheme 1.

S. A. Johannesen et al. / Tetrahedron Letters 48 (2007) 3569–3573
3571
a
Table 1. Pd(0)-catalyzed allylic alkylation studies using disaccharide-based phosphine ligands
O
O
OAc
Ph
Time (h)
O
O
MeO
Ph
OMe
3
Pd2(η −allyl)2Cl2, Ligand
Ph
Pd/L
MeO
OMe
THF, Base
Ph
b
c
d
e
Entry
Ligand
Temp. (ꢁC)
Base
Conv. (%)
Yield (%)
ee (%) (config.)
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
6
6
1/1
1/2
1/1
1/2
1/1
1/2
1/1
1/2
1/1
1/2
1/1
24
48
24
24
24
24
24
24
24
24
96
24
24
24
72
72
24
48
48
24
24
24
60
60
25
25
25
25
25
25
25
25
25
60
25
25
60
25
60
25
25
60
25
60
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
BSA-KOAc
73
15
100
82
100
78
100
60
60
45
21
32
96
0
15
<5
53
37
70
95
44
91
65
—
98
77
98
68
98
55
54
38
—
—
93
—
—
—
48
–
56 (S)
44 (R)
68 (S)
86 (S)
39 (S)
31 (S)
81 (S)
81 (S)
47 (S)
49 (S)
85 (S)
68 (S)
88 (S)
—
8
8
11
11
10
10
13
13
9
1
1
1
1
1
1
1
1
1
1
2
2
2
K
2
K
2
K
2
K
2
K
2
K
2
K
2
CO
CO
CO
CO
CO
CO
CO
3
3
3
3
3
3
3
9
9
1/2
1/1
7
7
12 (S)
—
7
1/2
1/1
7
76 (S)
31 (S)
32 (S)
30 (S)
31 (S)
32 (S)
12
12
12
12
12
BSA-KOAc
K
2
K
2
K
2
K
2
CO
CO
CO
CO
3
3
3
3
68
91
–
1/2
88
a
b
c
[
1,3-Diphenyl-2-propenyl acetate]/[dimethyl malonate]/[Base]/[Pd] = 1/3/3/0.02; solvent [0.125 M].
Conversion determined by GC analysis.
Isolated pure product.
Determined by HPLC analysis (column Chiralpak AD 0.46 · 25 cm).
Determined by comparison with an authentic sample.
d
e
3
1
,4,6-triacetylated D-glucosamine unit (Table 1, entries
–6), the iminophosphine derivative 8 provided the
small increase in the ees followed (Table 1, entries 9
and 10).
highest enantioselecitivity (86% ee), exceeding that of
1
0a
To our surprise, iminophosphine ligand 9,14 displaying
free hydroxyl groups on the D-glucosamine framework
resulted in an exceptionally high reactivity and enantio-
selectivity (93% yield, 88% ee) after a reaction time of
24 h at 20 ꢁC (Table 1, entry 13). A ligand/Pd ratio of
2:1 was again the preferred combination for the
generation of an active catalyst. As in the previous
experiments, the corresponding phosphine-amide or
aminophosphine 7 and 12, respectively, proved to be less
optimal, providing either lower conversions or lower ees
(Table 1, entries 14–21). The excellent results obtained
with ligand 9 were unexpected, considering the inactivity
observed for the deacetylated phosphino-amide 2, there-
by providing support that an effective and potentially
water soluble D-glucosamine ligand may eventually be
possible.
the monosaccharide phosphine-amide 1 (83% ee).
Best results were obtained with a ligand/Pd ratio of
:1 (Table 1, entry 4), suggesting that the imino ligand
2
binds to the palladium metal centre in a monodentate
fashion, in contrast to the bidentate nature of the phos-
phine-amide ligands. Interestingly, the configuration of
the obtained product is opposite to that observed for
1
0
the reactions exploiting ligand 1. Phosphine-amide 6
afforded lower ees compared to monosaccharide ligand
1
, suggesting that the addition of a bulky unit at C1
has a detrimental affect on the asymmetric induction
of these ligands. Whereas good conversions were ob-
served for aminophosphine 11 with a ligand/Pd ratio
of 1:1 (Table 1, entry 5), disaccharide amide ligand
proved less reactive (Table 1, entry 1). In any case, the
higher enantiomeric excesses observed for both phos-
phines with a ligand/Pd ratio of 1:1 supports a bidentate
role for these types of ligands. Allylation reactions per-
formed with the disaccharide ligand 10 lacking a C6-
benzyloxy group, afforded good ees suggesting that this
substituent is not important for either reactivity or
enantioselectivity (Table 1, entries 7 and 8). The higher
reactivity of the catalyst composed of a ligand/Pd
ratio of 1:1 is nevertheless puzzling. On the other hand,
a decrease in reactivity of the aminophosphine was ob-
served with ligand 13 compared to 11, whereas a
Finally, a series of Pd-catalyzed asymmetric allylation
reactions were performed with 1,3-diphenyl-2-propenyl
acetate and a variety of nucleophiles, in order to exam-
ine the scope of the iminophosphine ligand 9, the results
of which are illustrated in Table 2. In contrast to obser-
vations made with phosphine-amide 1, experiments per-
formed with nucleophiles such as a-methyl or a-
acetamido malonate led to both poor conversions and
enantioselectivities (Table 2, entries 1–3). Undoubtedly,

3572
S. A. Johannesen et al. / Tetrahedron Letters 48 (2007) 3569–3573
a
Table 2. Pd(0)-catalyzed allylic alkylation studies using iminophosphine ligand 9
OAc
Nu
3
Pd2(η −allyl)2Cl2, 9
NuH
Ph
Ph
THF, Base
Ph
Ph
b
c
d
e
Entry
NuH
Pd/L
Time (h)
Temp. (ꢁC)
Conv. (%)
Yield (%)
ee (%) (config.)
1
2
3
4
5
6
7
8
MeCH(CO
AcNHCH(CO
AcNHCH(CO
2
Me)
2
1/2
1/2
1/2
1/1
1/2
1/2
1/1
1/2
24
72
48
24
24
24
48
48
25
25
60
25
25
25
25
25
32
0
—
—
45
—
—
—
96
96
46 (R)
—
Me)
2 2
2
Me)
2
51
20
12
19 (R)
60 (R)
85 (R)
C
C
C
6
H
6
H
6
H
5
5
5
CH
CH
CH
2
2
2
NH
NH
NH
2
2
2
f
f
21
76 (R)
Morpholine
Morpholine
100
100
85 (R)
g
99 (R)
a
b
c
d
e
f
[
1,3-Diphenyl-2-propenyl acetate]/[dimethyl malonate]/[Base]/[Pd] = 1/3/3/0.02; solvent [0.125 M].
Conversion determined by GC analysis.
Isolated pure product.
Determined by HPLC analysis (column Chiralpak AD 0.46 · 25 cm).
Determined by comparison with an authentic sample.
3 2 2 3
Reaction performed in CH CN/H O (1/1) with K CO as base.
g
20
½
aꢀD ꢁ4.5 (c 0.5, CHCl
3
).
it appears that the metal complexes prepared with ligand
are sensitive to the sterical bulk of the nucleophile.
Gratifyingly, the amine nucleophiles such as benzyl
amine and morpholine provided high ees (85% and
of Science and Technology, the University of Aarhus
and the Carlsberg Foundation. And one of us (K.G.)
thanks the French ministry MENSR for a fellowship.
9
9
2
9%, respectively) with a ligand/Pd ratio of 2:1 (Table
, entries 5 and 9). Again, the monodentate nature of
References and notes
this ligand was confirmed by the lower ees obtained with
a ligand/metal ratio of 1:1. Additionally, with morphol-
ine as the nucleophile the reaction furnished the allylic
amine in an excellent yield of 96% after a reaction time
of 48 h at 25 ꢁC. On the other hand, with benzylamine
only low conversions were observed either in solvents
such as THF or acetonitrile/water mixtures. Possibly,
either complexation of the primary amine to the metal
centre may result in deactivation of the catalyst, or the
amine may be interfering with the Pd-bound imine func-
tionality of the ligand. In any case, it is interesting to
note the high selectivity displayed by this disaccharide
ligand 9 for the various substrates tested.
1
. (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96,
95–422; (b) Williams, J. M. J. Synlett 1996, 705–710; (c)
Helmchen, G. J. Organomet. Chem. 1999, 576, 203–214;
d) Pfaltz, A.; Lautens, M. In Comprehensive Asymmetric
Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H.,
Eds.; Springer: Tokyo, 1999; Vol. 2, p 833; (e) Trost, B. M.
Chem. Pharm. Bull. 2002, 50, 1–14; (f) Trost, B. M.;
Crawley, M. L. Chem. Rev. 2003, 103, 2921–2944; (g)
Trost, B. M. J. Org. Chem. 2004, 69, 5813–5837; (h)
Braun, M.; Meier, T. Angew. Chem., Int. Ed. 2006, 45,
3
1
5
(
6
952–6955; (i) Trost, B. M.; Machacek, M. R.; Aponick,
A. Acc. Chem. Res. 2006, 39, 747–760.
. Trost, B. M.; Van Vranken, D. L.; Bingel, C. J. Am.
Chem. Soc. 1992, 114, 9327–9343.
2
In conclusion, a series of disaccharide phosphine ligands
composed of a D-glucosamine unit have been prepared
and examined in asymmetric allylic alkylation. Surpris-
ingly, a Pd-complex with an imino phosphine derivative
of this disaccharide possessing free hydroxyl groups on
the glucosamine sugar proved to be the most active cat-
alyst with respect to conversion and enantioselectivity
with a preference for unsubstituted malonate and a sec-
ondary amine such as morpholine. Further work is now
underway to investigate the importance of the reducing
sugar end of this ligand and whether it can be substi-
tuted with a simple bulky alkyl group. Additionally,
work is in the process to systematically understand
how these carbohydrate ligands exert their chiral effect.
This work will be reported in due course.
3
. (a) Von Matt, P.; Pfaltz, A. Angew. Chem., Int. Ed. Engl.
1993, 32, 566–568; (b) Williams, J. M. J. Synlett 1996, 705–
710; (c) Helmchen, G. J. Organomet. Chem. 1999, 576,
203–214.
4
. (a) Dieguez, M.; Pamies, O.; Claver, C. Chem. Rev. 2004,
1
04, 3189–3215; (b) Dieguez, M.; Pamies, O.; Ruiz, A.;
Castillon, S.; Claver, C. Coord. Chem. Rev. 2004, 248,
165–2192.
2
5
6
. Nazarov, A. A.; Hartinger, C. G.; Arion, V. B.; Giester,
G.; Keppler, B. K. Tetrahedron 2000, 58, 8489–8492.
. Khiar, N.; Araujo, C. S.; Alvarez, E.; Fernandez, I.
Tetrahedron Lett. 2003, 44, 3401–3404.
7. (a) Hong-Wick, K.; Pregosin, P. S.; Trabesinger, G.
Organometallics 1998, 17, 3254–3264; (b) Barbaro, P.;
Currao, A.; Hermann, J.; Nesper, R.; Pregosin, P. S.;
Salzmann, R. Organometallics 1996, 15, 1879–1888.
8
9
. Glaser, B.; Kunz, H. Synlett 1998, 53–55.
. (a) Yonehara, K.; Hashizume, T.; Mori, K.; Ohe, K.;
Uemura, S. Chem. Commun. 1999, 415–416; (b) Yonehara,
K.; Hashizume, T.; Mori, K.; Ohe, K.; Uemura, S. J. Org.
Chem. 1999, 64, 9374–9380; (c) Hashizume, T.; Yonehara,
K.; Ohe, K.; Uemura, S. J. Org. Chem. 2000, 65, 5197–
5201.
Acknowledgements
We are grateful for generous financial support from the
Danish Natural Science Research Council, the Ministry

S. A. Johannesen et al. / Tetrahedron Letters 48 (2007) 3569–3573
3573
1
0. (a) Tollabi, M.; Framery, E.; Goux-Henry, C.; Sinou, D.
Tetrahedron: Asymmetry 2003, 14, 3329–3333; (b) Kono-
vets, A.; Glegola, K.; Penciu, A.; Framery, E.; Jubault, P.;
Goux-Henry, C.; Pietrusiewicz, K. M.; Quirion, J.-C.;
Sinou, D. Tetrahedron: Asymmetry 2005, 16, 3183–3187.
1. Glegoła, K.; Johannesen, S. A.; Skrydstrup, T.; Sinou, D.;
Framery, E., in preparation.
2. Johannesen, S. A.; Pedersen, B. O.; Duus, J.; Skrydstrup,
T. Inorg. Chem., in press.
3. Benzotriazol-1-yl-oxy-tris-(dimethylamino)-phosphonium-
(t, 1H, J = 8.0 Hz), 2.5 (br s, 1H), 1.87 (br s, 1H), 1.71 (br
s, 1H). C NMR (CDCl , 100 MHz) d (ppm) 163.7 (d,
3
1
3
J = ꢁ16.1 Hz), 139.5, 139.3 (d, J = ꢁ16.8 Hz), 138.6,
138.4, 138.2 (d, J = ꢁ20.6 Hz), 137.5 (d, J = ꢁ8.4 Hz),
137.0 (d, J = ꢁ9.9 Hz), 134.6, 134.4, 134.2 (2C), 134.0,
133.8, 130.8, 129.3, 129.1, 129.0, 128.9 (2C), 128.7, 128.6,
128.4, 128.1, 127.9 (2C), 127.7, 101.3, 98.5, 80.5, 79.3,
77.2, 76.7, 75.6 (2C), 73.8, 73.4, 70.3, 70.2, 68.5, 62.5, 55.6.
1
1
1
1
3
1
P NMR (CDCl
3
, 162 MHz) d (ppm) d ꢁ9.9. ES-HRMS
+
C H NNaO P [M+Na ]; calcd: 920.3540; found,
5
3
56
10
hexafluorophosphate.
4. Data for iminophosphine 9. H NMR (CDCl
920.3539.
1
3
, 400 MHz) d
15. For some examples of allylic substitution with morpho-
line, as well as structural assignments, see: (a) Nemoto, T.;
Masuda, T.; Akimoto, Y.; Fukuyama, T.; Hamada, Y.
Org. Lett. 2005, 7, 4447–4450; (b) Faller, J. W.; Wilt, J. C.
Org. Lett. 2005, 7, 633–636; (c) Hamada, Y.; Seto, N.;
Takayanagi, Y.; Nakano, T.; Hara, O. Tetrahedron Lett.
1999, 7791–7794; (d) Constantieux, T.; Brunel, J.-M.;
Labande, A.; Buono, G. Synlett 1998, 49–50, See also Ref.
10a.
(
ppm) d 8.70 (d, 1H, J = 3.9 Hz), 7.81 (dd, 1H, J = 3.7,
7.4 Hz), 7.41–7.18 (m, 27H), 6.91 (dd, 1H, J = 4.3,
7.4 Hz), 4.88 (d, 1H, J = 11.2 Hz), 4.81 (d, 1H, J =
11.2 Hz), 4.79 (d, 1H, J = 12.3 Hz), 4.64 (d, 1H, J =
12.3 Hz), 4.60 (d, 1H, J = 7.8 Hz), 4.55 (d, 1H,
J = 3.5 Hz), 4.49 (d, 1H, J = 12.1 Hz), 4.43 (d, 1H,
J = 11.9 Hz), 3.83–3.73 (m, 2H), 3.72–3.63 (m, 2H),
3
.54–3.36 (m, 5H), 3.33 (s, 3H), 3.20–3.11 (m, 2H), 2.94