5726
D. A. Rankic et al. / Tetrahedron Letters 51 (2010) 5724–5727
Table 3
Asymmetric Pd-catalyzed intermolecular arylation of 2,3-dihydrofuran (7) with phenyltriflate (8)
Pd(OAc)2 (3 mol%)
ligand (6 mol%)
OTf
Ph
Ph
O
O
O
DIPEA, benzene
40 ºC, 24 h
7
8
9
10
Entrya
Series
Ligand
3,30-Substituent
Conversion (%)
Product ratio (%ee, config)b
9
10
1
2
3
4
5
BINAPc
(S)-1a
(S)-1b
(S)-1c
(S)-1d
(S)-1e
H
32
56
100
100
68
91 (66, S)
99 (7, S)
99 (73, S)
97 (40, S)
96 (80, S)
9 (57, R)
1 (22, R)
1 (2, R)
3 (16, R)
4 (16, R)
OMe
OiPr
OBn
OPiv
6
7
8
9
xylBINAP
(S)-2a
H
100
100
100
100
66 (21, S)
98 (90, R)
98.5 (>99, S)
99 (83, R)
34 (83, R)
2 (45, R)
1.5 (77, S)
1 (6, S)
(Rax)-2b
(Sax)-2c
(R)-2d
ONp*
ONp*
OiPr
a
b
c
All catalysts were generated in situ by premixing Pd(OAc)2 and ligand for 30 min at rt in benzene.
Measured by chiral GC (80 °C for 2 min, increase 1 °C/min for 38 min. Cyclodex-B column).
Obtained from Ref. 4 for the purposes of comparison.
(R)-enantiomer.Upon substitution of the 3 and 30 positions of xylB-
INAP, improvements in both product ratio and enantioselectivities
were achieved (Table 3, entries 7–9). The OiPr-substituted xylBINAP
2d outperformed its BINAP counterpart 1c, affording product 9 in
83% ee instead of 73% ee, however the product ratios and enantiose-
lectivity of product 10 were comparable between the two ligands
(entry 9). This seemed to indicate that the effects of 3,30-disubstitu-
tion and the 3,5-dialkyl meta effect operate synergistically in the
case of ligand 2d and for this reaction type. The most impressive re-
sults were obtained with diastereomeric ligands (Rax)-2b and (Sax)-
2c. Both afforded product 9 nearly exclusively (98:2 ratio of 9 to
10). Moreover, catalysts based on these ligands afforded the highest
ee’s for any 3,30-disubstituted BINAP or xylBINAP ligands. (Rax)-2b
gave arylated 2,3-dihydrofuran (R)-9 in 90% ee and (Sax)-2c provided
product (S)-9 in >99% ee (S) (entries 7 and 8). Note that 3,4-dihydro-
furan arylation product 10 was also produced with the same sense of
chirality as the configuration of the ligand used in both cases, with
moderate enantioselectivities of 45% ee and 77% ee obtained using
(Rax)-2b and (Sax)-2c, respectively.
The fact that arylation products 9 and 10 were produced with
same absolute configuration when ligands (Rax)-2b and (Sax)-2c
were utilized in the Pd-catalyzed arylation of 2,3-dihydrofuran
provides important mechanistic insight into the intermolecular
Heck/Mizoroki arylation of 2,3-dihydrofuran (7). Although not well
understood, it was originally proposed that the mechanism of this
intermolecular Heck/Mizoroki reaction proceeds via a kinetic reso-
lution.13 The evidence for this was a connection between the prod-
uct ratio and enantiomeric purity of 9. Hayashi and co-workers
observed that as the ratio of 10 to 9 increased, the enantiomeric ex-
cess of 9 increased as well. Their best results for the enantioselec-
tivity of 9 were obtained when Pd(OAc)2 was used as a precatalyst
and Proton Sponge™ was used as a base instead of DIPEA. They re-
ceived a 9:10 ratio of 71:29 with 9 being produced in >96% ee and
10 being produced in 17% ee after a reaction time of 216 h (9 days).
They postulated that if the coordination of 2,3-dihydrofuran (7) to
the Pd/(S)-BINAP catalyst after oxidative addition of phenyl triflate
(8) occurred via the pro-S face, the resulting metal alkyl was capa-
ble of undergoing a b-hydride elimination/re-insertion/b-hydride
elimination sequence to produce 9. Conversely, they thought that
if coordination of 2,3-dihydrofuran occurred via the pro-R face of
the olefin, the migratory insertion product underwent a single, fac-
ile b-hydride elimination to eject (R)-product 10. This explanation
accounted for why arylated dihydrofuran 9 was always produced
with the (S)-configuration and 10 was always produced with the
(R)-configuration when (S)-BINAP was used as a ligand in this
reaction.
In our study, when diastereomeric 3,30-disubstitutedxylBINAPli-
gands (Rax)-2b and (Sax)-2c were used, the absolute configuration of
the chiral axis was directly transferred to both the arylated products,
compounds 9 and 10. Moreover, high enantiomeric excesses of
product 9 could be accessed even when minute amounts of the min-
or isomer 10 were produced. These two pieces of evidence appear to
contradict the kinetic resolution argument put forth by Hayashi and
co-workers. We are currently investigating this phenomenon and
the results will be reported in a subsequent publication.
In conclusion, we have demonstrated that the incorporation of
3,5-(dimethylphenyl) groups on phosphorus in combination with
3,30-substituents on the BINAP framework was not beneficial for
intramolecular Heck/Mizoroki reactions. However, when an inter-
molecular Heck/Mizoroki reaction was performed, that is, the ary-
lation of 2,3-dihydrofuran, superior ee’s, and conversions were
observed in comparison with 3,30-disubstituted BINAPs and the
parent BINAP. Most notably we observed an interesting phenome-
non when diastereomeric 3,30-diastereomeric xylBINAP ligands
(Rax)-2b and (Sax)-2c were utilized. With these ligands, we found
that this reaction proceeded with high enantioselectivity and high
specificity for the major isomer 9, even with comparatively shorter
reaction times than those reported by Hiyashi and co-workers13
(24 h vs 216 h, vide supra). The small amount of isomer 10 that
was generated was found to be of the same configuration as isomer
9, which is in contrast to what was previously observed with BIN-
AP. This suggests that the Heck/Mizoroki reaction might not pro-
ceed via a kinetic resolution when (Rax)-2b and (Sax)-2c are used
as ligands for the arylation of 2,3-dihydrofuran. We propose that
this might be due to an interaction between the naproxen ether
substituents and the aryl triflate upon oxidative addition. We have
evidence for a similar interaction in the Rh-catalyzed asymmetric
hydrogenations of dehydroamino acids which will be reported
elsewhere in due course.
Acknowledgments
The University of Calgary and the Natural Science and Engineer-
ing Council of Canada (NSERC) are thanked for funding. D.A.R. and
D.L. authors thank both NSERC and Alberta Ingenuity Fund for stu-
dentships. Thanks are also extended to Professor Warren Piers for
use of his chiral HPLC.