Communications
optimal, diphenylmethanol also produced useful levels of
enantioselectivity (Table 1, entry 11). Results obtained with
primary benzyl alcohols (Table 1, entries 12–14) and isopro-
panol (entry 15) were less satisfactory.
Having thus established the basic parameters of the new
method, we proceeded to explore its substrate scope
(Table 2). Electron-donating or electron-withdrawing sub-
stituents on the phenyl ring had no appreciable effect on the
enantioselectivity (Table 2, entries 2–6). However, replacing
Scheme 3. Evidence of in situ racemization during KR. DCC=N,N’-
dicyclohexylcarbodiimide.
Table 2: Exploration of the substrate scope.[a]
DCC. For DKR we needed to generate stoichiometric
amounts of the chiral acyl donor. The use of 1.2 equivalents
of DCC produced the ester with a high ee value, but the
reaction was slow and accompanied by the formation of side
products (Table 1, entry 1). Pivalic anhydride[15b] reacted
more cleanly, but the reaction was also slow (Table 1,
entry 2). Fortunately, the use of benzoic anhydride, which is
Entry
R1
R2
By-product [%][b]
Yield [%][c]
ee [%]
1
2
3
Me
Me
Me
Me
Me
Me
Me
Me
Et
Ph
9
6
8
n.d.
5
90
85
89
92
93
91
88
87
62
71
13[e]
91
92
91
90
89
91
85
84
86
86
66
p-MeC6H4
p-MeOC6H4
p-MeOC6H4
p-ClC6H4
p-BrC6H4
PhCH2
CH3(CH2)7
Ph
4[d]
5
Table 1: Variation of the DKR protocol.
6
7
8
9
10
11
4
12
13
20
20
38
nBu
iPr
Ph
Ph
Entry
Activator
Solvent
ROH
Yield [%][a]
ee [%]
1
2
3
DCC
[D8]PhMe
[D8]PhMe
[D8]PhMe
[D8]PhMe
[D8]PhMe
CD2Cl2
CDCl3
C6D6
[D8]THF
CD3CN
[D8]PhMe
[D8]PhMe
[D8]PhMe
[D8]PhMe
[D8]PhMe
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
1-Np2CHOH
Ph2CHOH
35
35
91
49
87
67
61
88
66
60
97
93
99
94
89
89
91
91
[a] Performed on 0.10 mmol (Æ)-substrate, unless specified otherwise.
[b] Yield of the by-product (PhCO2CH(1-Np)2) was determined by
1H NMR analysis. [c] Yield of isolated product ((96Æ3)% pure by
1H NMR) unless specified otherwise. [d] Performed on 1.0 mmol scale
(see the Experimental Section). [e] Yield was determined by 1H NMR
analysis. n.d.=not determined.
Piv2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
Bz2O
4[b]
5[c]
6
À84
46
79
84
87
88
46
83
26
15
7
8
9
10
11
12
13
14
15
the phenyl group with a benzyl or n-octyl group proved to be
slightly detrimental (Table 2, entries 7 and 8). Replacing the
methyl group with a primary alkyl group also led to some loss
of enantioselectivity; the reaction proceeded more slowly,
thus resulting in lower yields of the desired product and the
concomitant increase in the yield of the benzoate ester by-
product (Table 2, entries 9 and 10). When a secondary alkyl
substituent was introduced, the benzoate ester became the
predominant product (Table 2, entry 11).
PhCH2OH
1-NpCH2OH
2-NpCH2OH
iPrOH
20
49
[a] Determined by 1H NMR analysis. [b] Catalyst 8 was used instead of 7.
[c] Catalyst 9 was used instead of 7. For structures of 7, 8, and 9 see
Figure 1. Bz=benzoyl, Np=naphthyl, Piv=2,2-dimethylpropanoyl,
THF=tetrahydrofuran.
We confirmed that both the lithium aluminum hydride
reduction and the trifluoroacetic acid catalyzed deprotection
of the DKR products occur in high yield (95% and 85%,
respectively) and without any loss of stereochemical integrity.
In the former reaction, di(1-naphthyl)methanol was also
recovered in 90% yield.
In conclusion, we have developed the first nonenzymatic,
enantioselective method for the DKR of a-(arylthio)- and a-
(alkylthio)alkanoic acids. In contrast to the existing enzymatic
protocol, it neither requires the prior conversion of the
substrates into their thioesters nor releases ethanethiol in the
course of the reaction, and can be directed towards either
enantiomer of the product simply by switching the absolute
configuration of the catalyst. Extension of the new DKR
process to other classes of acyl donors is under investigation
and will be reported in due course.
employed in Shiinaꢀs original protocol for the BTM-catalyzed
KR of arylpropanoic acids,[15a] proved to be optimal. Within
24 hours, the desired ester was obtained in 91% yield with
91% ee, accompanied by approximately 9% of the benzoate
ester by-product (Table 1, entry 3).
Variation of other reaction parameters was undertaken
next. (S)-HBTM (7; Figure 1; Table 1, entry 3) was superior to
(S)-BTM[14b] (8; entry 4) and (R)-Cl-PIQ[14c] (9; entry 5), a
result that is in agreement with our earlier study.[7c] Several
additional solvents were tested and they all produced some-
what lower ee values and yields than toluene (Table 1,
entries 6–10). Although di(1-naphthyl)methanol (Table 1,
entry 3), originally identified by Shiina et al.[15a] and subse-
quently used in our studies,[7c,10f] once again proved to be
5554
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 5553 –5555