a direct aldol between 1 and R-methyl-â,γ-unsaturated
aldehyde 5b is complicated since the enantiomerically pure
aldehyde has not been reported and the diastereoselectivity
with the racemic aldehyde is modest.5
and [n-Bu3PH]BF4 as the palladium and phosphine sources,
respectively.11-13 These changes allowed us to control the
number and the nature of the ligands on the Pd center,14 as
well as avoid the use of air-sensitive n-Bu3P since the
tetrafluoroborate salt is an air-stable precursor of the phos-
phine. Under the modified conditions, using 10 mol % Pd-
(OAc)2, 10 mol % [n-Bu3PH]BF4, and 3 equiv of HCO2H/
Et3N15 (1:2) at 40 °C in CH3CN gave, after 5 min, complete
conversion with less than 3% of the internal olefin 8 in a dr
of 13:1. A summary of the optimization studies on 6b is
presented in Table 1.16 Using 20 mol % phosphine led to
lower conversion with a dr of 5:1 along with 17% of 8. The
use of a phosphine is required since no reaction was observed
in its absence (entry 3), and a 1:1 ratio of Pd:P was used
throughout the rest of the study. The conversion and the dr
were identical between room temperature and 40 °C.
However, running the reaction at 0-5 °C led to incomplete
conversion but an excellent dr (entry 5). Diluting the reaction
had a beneficial effect on the diastereoselectivity of the
reaction since a ratio of >15:1 was obtained (entry 6). The
palladium loading could be decreased to 2.5 mol % (entry
6-8), causing the reaction time to increase from 1 to 3 h.
When the scale was increased, it was found that for a
palladium loading of 2.5 mol % running the reaction at 40
°C was required in order to obtain complete conversion in a
reasonable time,17 although no effect on the dr was observed.
Optimized conditions [Pd(OAc)2 (2.5 mol %), [n-Bu3PH]-
BF4 (2.5 mol %), HCO2H/Et3N (1:2) (3 equiv), CH3CN (0.05
M), 40 °C] were used for the rest of the study. This study
represents the first use of phosphonium salts as a precursor
of phosphines in allylic substitution reactions.
Early experiments indicated the potential of our strategy
since reduction of 6a gave the desired product 7 in good
yield (75%) and moderate selectivity (5.7:1 dr) in favor of
the syn isomer along with 10% of the internal olefin 8 (eq
1).1 In this case, an R-substituted-γ-carboxy-R,â-unsaturated
aldehyde is used as a synthetic equivalent of an R-substituted-
â,γ-unsaturated aldehyde, which are rarely used in total
syntheses6 primarily because of olefin isomerization.7
Even though this strategy allows access to products of
broad interest, to be synthetically useful, higher selectivity
is needed as well as a determination of the scope of the
reaction. We have improved the diastereoselectivity (up to
>20:1 dr) by optimizing the reaction conditions, studied the
influence of protecting group, leaving group, olefin geometry,
and nearby substitution, and report the results of our
investigation herein.
The effects of the leaving group and protecting group were
also examined (Table 2).18 Under the optimized conditions,
the model substrate 6b gave 84% yield with a dr of >15:1
with less than 3% of 8. The substrate bearing a benzyl
carbonate (9) gave identical results.19 Using a bulkier TBS
group allowed the reaction to proceed in similar yield and
diastereoselectivity (entry 3). When R ) Ac, the desired
product 14 was obtained in lower yield with a low dr (entry
Initial efforts to optimize the original system (Pd2(dba)3
(5 mol %), n-Bu3P (20 mol %), HCO2NH4 (3 equiv), DMF,
50 °C) using 6a8 were unsuccessful. We turned our attention
to a more reactive leaving group hoping to improve the
diastereoselectivity by using milder conditions. Carbonates
are known to be an excellent leaving group in π-allyl Pd
chemistry,9 and applying the original conditions to 6b gave
incomplete conversion with a disappointing dr of 3:1.
Examination of different media10 showed that acetonitrile
was the solvent of choice and gave 7 with good conversion
with dr > 10:1 at 40 °C or room temperature.
(11) Mandai, T.; Matsumoto, T.; Tsuji, J.; Saito, S. Tetrahedron Lett.
1993, 34, 2513.
(12) Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295. The free
phosphine is liberated by the action of a base, Et3N in our case.
(13) Phosphonium salt can be easily synthesized (see ref 12) or,
alternatively, is available from Strem Chemicals.
To make the reaction simpler to run as well as resolve
some variability problems, we decided to employ Pd(OAc)2
(14) It was demonstrated that the dba ligands are not fully displaced by
n-Bu3P from Pd2(dba)3 (see ref 11) or by Ph3P from Pd(dba)2; see: Amatore,
C.; Jutand, A. Coord. Chem. ReV. 1998, 178-180, 511. Amatore, C.; Jutand,
A.; Khalil, F.; M’Barki, M. A.; Mottier, L. Organometallics 1993, 12, 3168.
(15) Formate salts (HCO2K, HCO2Cs) were not as effective, probably
due to their low solubility.
(5) (a) Roush, W. R. J. Org. Chem. 1991, 56, 4151. (b) Ahmar, M.;
Bloch, R.; Mandville, G.; Romain, I. Tetrahedron Lett. 1992, 33, 2501.
(6) For relevant examples of the use of R-substituted-â,γ-unsaturated
aldehyde in total synthesis, see: (a) Liu, P.; Jacobsen, E. N. J. Am. Chem.
Soc. 2001, 123, 10772. (b) Salamonczyk, G. M.; Han, K.; Guo, Z.-W.;
Sih, C. J. J. Org. Chem. 1996, 61, 6893. (c) Evans, D. A.; DiMare, M. J.
Am. Chem. Soc. 1986, 108, 2476.
(7) For a Lewis-acid-mediated approach to â,γ-unsaturated aldehydes
and their reactions in situ, see: (a) Lautens, M.; Maddess, M. L.; Sauer, E.
L. O.; Ouellet, S. G. Org. Lett. 2002, 4, 83. (b) Lautens, M.; Ouellet, S. G.;
Raeppel, S. Angew. Chem., Int. Ed. 2000, 39, 4079.
(16) The dr (ratio of syn/anti) of all reactions was estimated by 1H NMR
spectroscopy. For the purpose of this study, >15:1 dr indicates that the
other diastereomer was present in less than 6% (typically 4-6%), whereas
>20:1 dr specifies that the other diastereomer was not detectable.
(17) On a 0.2 mmol scale, the reaction with 2.5 mol % Pd at room
temperature led to almost complete conversion after 3 days, compared to
<1 h at 40 °C. On a 0.5 mmol scale, the reaction with 2.5 mol % Pd is
done in <2 h.
(8) All substrates used in this study are racemic.
(9) (a) Tsuji, J. Palladium Reagents and Catalysts. InnoVations in
Organic Synthesis. John Wiley & Sons: Chichester, 1995; pp 290-422.
(b) Pfaltz, A.; Lautens, M. ComprehensiVe Asymmetric Catalysis; Jacobsen,
E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer: New York, 1999; Vol. 2,
pp 833-884.
(18) Stereochemistry of the products was determined by comparison with
authentic samples or similar or previously reported compounds; see
Supporting Information for details.
(19) Under the optimized conditions, other leaving groups such as acetate
or formate (see ref 1c) gave incomplete conversion and no reaction,
respectively.
(10) Other solvents, including THF, DMF, or a combination of the two,
gave lower conversion with low dr.
3392
Org. Lett., Vol. 5, No. 19, 2003