10 mol % of the catalyst and three additives: water (30 mol
%), butyllithium (10 mol %), and tri(2,6-dimethoxy-
phenyl)phosphine oxide (10 mol %). Most recently, Na´jera
et al. have shown that an aluminum binol complex would
catalyze the asymmetric addition of methyl cyanoformate
to aldehydes at room temperature.8 In this case, 10 mol %
of the catalyst along with 4 Å molecular sieves were required
to produce cyanohydrin carbonates with up to 80% enan-
tiomeric excess.
Table 1. Addition of Ethyl Cyanoformate to Benzaldehyde
Catalyzed by Complex 1
temp (°C)
1 (mol %)
time (h)
completion (%)
ee (%)
-85
-73
-40
-40
25
-40
-40
1
1
1
0.1
0.1
5
19
48
19
72
148
18
<3
100
100
<3
<3
100
94 (S)
83 (S)
95 (S)
93 (S)
10
51
100
While the result at -73 °C was encouraging, the long
reaction time was felt to be impractical, so the effect of
increasing the amount of catalyst was investigated to see if
a similar enantiomeric excess could be obtained at a
temperature where the rate of reaction was faster. Gratify-
ingly, use of 5 mol % of the catalyst at -40 °C resulted in
the complete formation of (S)-mandelonitrile ethyl carbonate
with 95% enantiomeric excess after just 18 h. This combina-
tion of catalyst mol %, reaction temperature, and product
enantiomeric excess is a significant improvement on any of
the previously known catalysts and was taken to be the
optimal conditions for the use of complex 1.
The addition of ethyl cyanoformate to other aldehydes was
then investigated under these optimized conditions,10 and the
results are shown in Table 2. Electron-rich aromatic alde-
hydes were found to be excellent substrates for this reaction,
giving cyanohydrin carbonates in high chemical yield and
with excellent enantiomeric excesses. Thus, all three isomers
of methoxybenzaldehyde and 4-methylbenzaldehyde gave
products with 94-99% enantiomeric excess. Cinnamalde-
hyde was also found to be an excellent substrate giving a
cyanohydrin carbonate with 94% enantiomeric excess. The
introduction of an electron-withdrawing trifluoromethyl
group onto the aromatic ring resulted in a very rapid reaction,
albeit to give a product with a lower enantiomeric excess,
possibly due to a competing uncatalyzed addition. 4-Chlo-
robenzaldehyde was, however, an excellent substrate, giving
the corresponding cyanohydrin ethyl carbonate in high yield
and with 94% enantiomeric excess. For all of these reactions
with aromatic or R,â-unsaturated aldehydes, the use of 2
equiv of ethyl cyanoformate was necessary in order for the
reactions to be complete in less than 20 h. The quantity of
ethyl cyanoformate used could be reduced to just 1.2 equiv,
though this resulted in extended reaction times of 45-68 h.
In view of this precedent, we decided to investigate the
use of ethyl cyanoformate as a potential cyanide source for
use in conjunction with catalyst 1 (Scheme 1). This would
have the advantages of being less expensive than the use of
trimethylsilyl cyanide, while avoiding the heterogeneous
reaction conditions needed for reactions employing potassium
cyanide and of being a completely atom-economical reaction.
In addition, cyanohydrin carbonates are more stable toward
unwanted hydrolysis than cyanohydrin trimethylsilyl ethers.
In this communication, we present our results.
Scheme 1
For the initial study, benzaldehyde was selected as the
substrate, and the addition of ethyl cyanoformate catalyzed
by complex 1 was studied under various conditions as shown
in Table 1. When 1 mol % of the catalyst was used at -85
°C in dichloromethane, no product was detected. However,
raising the temperature to -73 °C resulted in complete
reaction after 48 h to give (S)-mandelonitrile ethyl carbonate
with a highly encouraging 94% enantiomeric excess.9
Increasing the temperature further increased the rate of
reaction but at the expense of a reduction in the enantiomeric
excess of the product. Attempts to reduce the amount of
catalyst to 0.1 mol % (the optimal amount for the addition
of trimethylsilyl cyanide to benzaldehyde2) gave unsatisfac-
tory results even at room temperature.
(10) Typical Experimental Procedure. A stirred solution of benzalde-
hyde (0.48 mL, 4.7 mmol) in dichloromethane (20 mL) and (R)-1 (5 mol
%, 0.264 g, 0.22 mmol) was cooled to -84 °C, and EtOCOCN (0.93 mL,
9.4 mmol) was added in one portion. The yellow solution was then allowed
to warm to -40 °C and then stirred vigorously for 19 h. To remove the
catalyst, the filtrate was passed through a pad of silica eluting with
dichloromethane. The solvent was removed in vacuo and the resulting
orange-brown liquid dried in vacuo to give the crude cyanohydrin carbonate,
which could be microdistilled to give mandelonitrile ethyl carbonate as a
clear liquid (0.87 g, 90%): ee 95% (determined by chiral GC using a γ-CD
(7) (a) Tian, J.; Yamagiwa, N.; Matsunaga, S.; Shibasaki, M. Angew.
Chem., Int. Ed. 2002, 41, 3636-8. (b) Tian, J.; Yamagiwa, N.; Matsunaga,
S.; Shibasaki, M. Org. Lett. 2003, 5, 3021-4.
butyryl, fused silica capillary column (30 m × 0.25 mm) with hydrogen as
20
the carrier gas); [R]D -20.1 (c 1.8 CHCl3) [lit.11 [R]20 +16.2 (c 2.8,
D
(8) Casas, J.; Baeza, A.; Sansano, J. M.; Na´jera, C.; Saa´, J. M.
Tetrahedron: Asymmetry 2003, 14, 197-200.
CHCl3) for (R)-enantiomer with 94% ee]; δH 1.26 (3H, t J 7.2), 4.21 (2H,
q J 7.2), 6.19 (1H, s), 7.2-7.5 (5H, m).
(11) Tian, J.; Yamagiwa, N.; Matsunaga, S.; Shibasaki, M. Angew. Chem.,
Int. Ed. 2002, 41, 3636-3638.
(9) All enantiomeric excesses were determined by chiral GLC and are
accurate to (3%.
4506
Org. Lett., Vol. 5, No. 23, 2003