Investigation of lewis acid versus lewis base catalysis in asymmetric cyanohydrin synthesis

Article abstract of DOI:10.1002/chem.201001078

The asymmetric addition of trimethylsilyl cyanide to aldehydes can be catalysed by Lewis acids and/or Lewis bases, which activate the aldehyde and trimethylsilyl cyanide, respectively. It is not always apparent from the structure of the catalyst whether Lewis acid or Lewis base catalysis predominates. To investigate this in the context of using salen complexes of titanium, vanadium and aluminium as catalysts, a Hammett analysis of asymmetric cyanohydrin synthesis was undertaken. When Lewis acid catalysis is dominant, a significantly positive reaction constant is observed, whereas reactions dominated by Lewis base catalysis give much smaller reaction constants. [{Ti(salen)O}₂] was found to show the highest degree of Lewis acid catalysis, whereas two [VO(salen)X] (X = EtOSO₃ or NCS) complexes both displayed lower degrees of Lewis acid catalysis. In the case of reactions catalysed by [{Al(salen)}₂O] and triphenyl- phosphine oxide, a non-linear Ham- mett plot was observed, which is indicative of a change in mechanism with increasing Lewis base catalysis as the carbonyl compound becomes more electron-deficient. These results suggested that the aluminium complex/tri- phenylphosphine oxide catalyst system should also catalyse the asymmetric addition of trimethylsilyl cyanide to ke- tones and this was found to be the case.

Full text of DOI:10.1002/chem.201001078

FULL PAPER
DOI: 10.1002/chem.201001078
Investigation of Lewis Acid versus Lewis Base Catalysis in Asymmetric
Cyanohydrin Synthesis
Michael North,* Marta Omedes-Pujol, and Courtney Williamson[a]
Abstract: The asymmetric addition of
trimethylsilyl cyanide to aldehydes can
be catalysed by Lewis acids and/or
Lewis bases, which activate the alde-
hyde and trimethylsilyl cyanide, respec-
tively. It is not always apparent from
the structure of the catalyst whether
Lewis acid or Lewis base catalysis pre-
dominates. To investigate this in the
context of using salen complexes of ti-
tanium, vanadium and aluminium as
catalysts, a Hammett analysis of asym-
metric cyanohydrin synthesis was un-
dertaken. When Lewis acid catalysis is
dominant, a significantly positive reac-
tion constant is observed, whereas re-
actions dominated by Lewis base catal-
ysis give much smaller reaction con-
catalysis. In the case of reactions cata-
lysed by [{Al(salen)}2O] and triphenyl-
AHCTUNGTRENNUNG
phosphine oxide, a non-linear Ham-
mett plot was observed, which is indi-
cative of a change in mechanism with
increasing Lewis base catalysis as the
carbonyl compound becomes more
electron-deficient. These results sug-
gested that the aluminium complex/tri-
phenylphosphine oxide catalyst system
should also catalyse the asymmetric ad-
dition of trimethylsilyl cyanide to ke-
tones and this was found to be the
case.
stants. [{Ti
show the highest degree of Lewis acid
catalysis, whereas two [VO(salen)X]
ACHTUNGTRNE(NUGN salen)O}2] was found to
AHCTUNGTRENNUNG
(X=EtOSO3 or NCS) complexes both
displayed lower degrees of Lewis acid
Keywords: asymmetric catalysis
·
cyanohydrin · Hammett analysis ·
Lewis acids · Lewis bases
Introduction
Over the last decade, there has been a substantial increase
in interest in asymmetric cyanohydrin synthesis.[1] Most of
this work utilises trimethylsilyl cyanide as the cyanide
source and aldehydes as the electrophile to form non-race-
mic cyanohydrin trimethylsilyl ethers (Scheme 1), a reaction
which has been shown to be catalysed by chiral Lewis
acids[2] and chiral Lewis bases.[3] However, the relatively
complex nature of chiral catalysts means that they almost
always contain acidic and basic sites and it is then not clear
what the mode of action of the catalyst is. In addition, it is
well established that the most effective catalysts simultane-
ously employ acid and base catalysis to activate both com-
ponents of the reaction.[2c,4] An illustrative example of this is
the bis-N-oxide catalyst 1 developed by Feng et al. (see
Scheme 1. Asymmetric addition of trimethylsilyl cyanide to aldehydes.
Scheme 2).[5] The N-oxides within catalyst 1 are proposed to
act as Lewis bases to activate the trimethylsilyl cyanide,
whereas one of the two secondary amides acts as an acid to
activate the aldehyde, leading to the transition-state struc-
ture shown in Scheme 2.
*
For metal-based Lewis acids of general structure MLn ,
the situation is particularly mechanistically complex as the
ligands are only coordinated to the metal and dissociation of
one of the L* units will inevitably produce a chiral Lewis
base (L*) and increase the Lewis acidity of the metal. Thus,
for any metal-based catalyst the issue arises as to the rela-
tive importance of Lewis acid and Lewis base catalysis. A
particularly effective class of metal-based catalysts for the
asymmetric addition of trimethylsilyl cyanide to aldehydes
are complexes of the salen ligand 2.[2] We have shown that
the bimetallic titanium complex 3 is a highly efficient cata-
lyst[6] and subsequently developed the vanadium-based cata-
lysts 4a and 4b as even more enantioselective catalysts.[6b,7]
[a] Prof. M. North, M. Omedes-Pujol, C. Williamson
School of Chemistry and
University Research Centre in Catalysis and Intensified Processing
Newcastle University, Bedson Building
Newcastle upon Tyne, NE1 7RU (UK)
Fax : (+44)191-2226929
E-mail: Michael.north@ncl.ac.uk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001078.
Chem. Eur. J. 2010, 16, 11367 – 11375
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11367

nide as part of the formation of the catalytically active spe-
cies.[9] In addition, the phenolic oxygens of metal
(salen)
ACHTUNGTRENNUNG
complexes are known to be Lewis basic,[15] which provides
another opportunity for Lewis base catalysis in all com-
plexes 3–5. Thus, we decided to undertake a study to investi-
gate the relative importance of Lewis acidity and Lewis ba-
sicity in the catalytic activity of complexes 3–5 and this
paper presents the results of this study.
Results and Discussion
Since a Lewis acid will activate the aldehyde, whereas a
Lewis base activates the trimethylsilyl cyanide, a Hammett
plot[16] based on substituted aromatic aldehydes provides a
way of assessing the importance of Lewis acid and Lewis
base activation to the rate of asymmetric cyanohydrin syn-
thesis. Lewis acid catalysis implies that an aldehyde activat-
ed by coordination to the Lewis acid is involved in the rate-
determining step of the reaction. Therefore, if Lewis acid
catalysis is dominant, a Hammett plot with a large positive
reaction constant will be expected, since the Lewis acid
withdraws electron density from the carbonyl bond. Thus,
an aldehyde, which is already electron-deficient, will be
made even more reactive by coordination to the Lewis acid,
whereas a very electron-rich aldehyde will form a more
stable and hence, less reactive complex with the Lewis acid.
In contrast, a reaction proceeding entirely by Lewis base
catalysis would be expected to have a reaction constant of
zero, since the rate-determining step of the mechanism in-
volves activation of the trimethylsilyl cyanide and the alde-
hyde is only involved after the rate-determining step. Ham-
mett plots have previously been used to determine the
Lewis acidity of palladium complexes,[17] to study achiral
Lewis acid-[18] or base-catalysed[19] reactions and to study
asymmetric Lewis acid-catalysed,[20] Lewis base-catalysed[21]
and carbene insertion reactions,[22] but do not appear to
have previously been used to determine the relative impor-
tance of Lewis acid and Lewis base catalysis.
Scheme 2. Transition state for the asymmetric cyanohydrin synthesis cata-
lysed by the bis-N-oxide 1.
Most recently, we have shown that the combination of the
bimetallic aluminium complex 5 and triphenylphosphine
oxide forms an effective catalyst system.[8]
The mechanism of action of complexes 3–5 has been ex-
tensively investigated by a combination of kinetic studies
and identification of reaction intermediates.[6b,9–12] However,
this work has tended to focus on the role of the metal as a
Lewis acid and the origin of the asymmetric induction. How-
ever, it has, become increasingly apparent that Lewis base
catalysis is also of some importance in reactions catalysed
by complexes 3–5. This is most apparent in reactions cata-
lysed by aluminium complex 5, since addition of triphenyl-
phosphine oxide resulted in a substantial increase in the rate
and the enantioselectivity of cyanohydrin synthesis.[8] For va-
nadium-based complexes 4, the rate of reaction, but not the
degree of asymmetric induction, was found to depend on
the Lewis basicity of the X group.[10] The most active cata-
lyst 4b had the most Lewis basic X group (NCS), whereas
the complex with X=triflate was catalytically inactive de-
spite having the most Lewis acidic vanadium centre. It is
also known that the oxygen atom of the V=O bond has ap-
preciable Lewis basicity[13] and this can lead to the formation
of oligomeric complexes in solution,[6d,14] which are impor-
tant for formation of the most active catalysts as suggested
by the kinetic data. In the case of titanium complex 3, the
two bridging oxygen atoms are known to be Lewis basic, as
we have shown that they are silylated by trimethylsilyl cya-
In previous work,[8,9a,10b,12] we have determined the kinet-
ics of the asymmetric addition of trimethylsilyl cyanide to
benzaldehyde catalysed by complexes 3, 4a, 4b and 5 and
have shown that the reactions obey the rate equations
shown in Equations (1)–(4), respectively.
1:3
ð1Þ
rate ¼ k½3ꢀ ½Me3SiCNꢀ ¼ kobs½Me3SiCNꢀ
0:6
rate ¼ k½4aꢀ ½Me3SiCNꢀ½PhCHOꢀ ¼ kobs½Me3SiCNꢀ½PhCHOꢀ
ð2Þ
1:2
rate ¼ k½4bꢀ ½Me3SiCNꢀ½PhCHOꢀ ¼ kobs½Me3SiCNꢀ½PhCHOꢀ
ð3Þ
rate ¼ k½5ꢀ½Ph3POꢀ½Me3SiCNꢀ ¼ kobs½Me3SiCNꢀ
ð4Þ
It is apparent from Equations (1)–(4), that reactions cata-
lysed by complexes 3 and 5 follow first-order kinetics with
11368
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11367 – 11375

Asymmetric Cyanohydrin Synthesis
FULL PAPER
the reaction rate being independent of the benzaldehyde
concentration. In contrast, reactions catalysed by vanadium
complexes 4a and 4b follow second-order kinetics with the
reaction rate depending on the concentrations of benzalde-
hyde and trimethylsilyl cyanide.
All of the kinetics measurements throughout this work
were obtained by monitoring the UV absorbance at l=240–
315 nm of unreacted aldehyde in samples periodically re-
moved from reactions carried out at 08C, as we have previ-
ously described.[8,9a,10b,12] Every kinetics experiment was car-
ried out in duplicate and the average rate constant used to
construct the Hammett plot.[23] The aldehydes were chosen
to provide as wide a range of substituent constants as possi-
ble; though for titanium complex 3, highly electron-deficient
aldehydes (3,4-dichloro-, 3,5-difluoro- and 4-(trifluorome-
thyl)benzaldehyde) reacted so quickly (100% conversion in
ꢁ5 s) that it was not possible to monitor the kinetics and
this restricted the range of positive substituent constants
that could be studied. The results obtained with ten alde-
hydes by using 0.1 mol% of complex 3 as catalyst are shown
in Table 1. All of the aldehydes were found to follow the
Figure 1. Hammett plot for the asymmetric addition of trimethylsilyl cya-
nide to different benzaldehydes catalysed by complex
3
(y=
2.3715xꢂ0.0409, R2 =0.9427). For abbreviations see Table 1.
Scheme 3. Synthesis of cyanohydrin acetates and free cyanohydrins for
stereochemical analysis (Ts=para-toluenesulfonyl).
Table 1. Rate constants and enantioselectivities for the synthesis of cyanohydrin trimethylsilyl ethers catalysed
by bimetallic titanium complex 3.[a]
tained from 4-methoxybenzal-
dehyde was found to racemise
on hydrolysis, so the enantio-
meric purity was determined by
comparison of the specific rota-
tion of the trimethylsilyl ether
with literature data.[26] In each
case, the S enantiomer of the
cyanohydrin derivative was ob-
tained from complex 3 derived
Entry
Aldehyde
Abbrev.
ka [minꢂ1
]
kb [minꢂ1
]
kavg [minꢂ1
]
ee [%][b]
1
2
3
4
5
6
7
8
PhCHO
H
4-Cl
3-Cl
0.4506
1.8688
2.5318
0.5142
0.1471
0.0699
0.1971
0.2382
2.2268
0.4344
0.4060
1.9131
2.5923
0.6080
0.1201
0.0658
0.2006
0.2806
2.2496
0.3721
0.4283ꢃ0.0223
1.89095ꢃ0.02215
2.56205ꢃ0.03025
0.5611ꢃ0.0469
0.1336ꢃ0.0135
0.06785ꢃ0.00205
0.19885ꢃ0.00175
0.2594ꢃ0.0212
2.2382ꢃ0.0114
0.40325ꢃ0.03115
84
87
84
88
4-ClC6H4CHO
3-ClC6H4CHO
4-FC6H4CHO
4-MeC6H4CHO
4-MeOC6H4CHO
3,4-Me2C6H3CHO
4-MeSC6H4CHO
3-FC6H4CHO
3-MeC6H4CHO
4-F
4-Me
4-MeO
3,4-Me2
4-MeS
3-F
79
46[c]
57[d]
55
9
10
87
95
3-Me
from
(R,R)-diaminocyclohex-
[a] All reactions were carried out in duplicate (to give ka and kb, respectively) in dichloromethane at 08C with ane, which is consistent with
previous work.[6] The enantio-
meric excesses obtained for the
cyanohydrins confirm that the
[aldehyde]0 =0.5m, [Me3SiCN]0 =0.55m and [3]=0.5 mm. [b] Determined by chiral GC analysis of the cyanohy-
drin acetate unless stated otherwise. [c] Determined by comparison of the specific rotation with literature
1
data.[27] [d] Determined by H NMR spectroscopy in the presence of (R)-mandelic acid and dimethylaminopyr-
idine (DMAP).
reaction being monitored was
that catalysed by complex 3
rather than any uncatalysed re-
same first-order kinetic equation [Eq. (1)] previously found
for the use of benzaldehyde as substrate. As the data in
Table 1 show, the kinetics experiments showed good repro-
action leading to racemic products.
The kinetics results obtained by using thirteen aromatic
aldehydes and 0.1 mol% of the vanadium-based catalyst 4a
are shown in Table 2. In this case, the reactions were signifi-
cantly slower than those catalysed by complex 3, so more
electron-deficient aldehydes could be used (Table 2, en-
tries 11–13), which allowed the Hammett plot to be extend-
ed to more positive s values. All of the substrates obeyed
second-order kinetics (first-order in the aldehyde and in tri-
methylsilyl cyanide).[23] The kinetics results again showed
good reproducibility and when incorporated into a Hammett
plot (Figure 2) produced a good fit to a straight line with a
slope of 1.9. The enantiomeric excesses of the cyanohydrin
trimethylsilyl ethers were again found to always be
>50%,[23] confirming that the reaction being monitored was
the catalysed rather than uncatalysed reaction.
ducibility and when converted into
a Hammett plot
(Figure 1) produced a good fit to a straight line with a slope
of 2.4.
Where possible, the enantiomeric excesses of the cyano-
hydrin trimethylsilyl ethers shown in Table 1 were deter-
mined by chiral GC analysis[23] after conversion into the cor-
responding acetates by the method of Kagan (Scheme 3).[24]
In two cases, no separation of enantiomers was achieved by
chiral GC. Therefore, the enantiomeric excess of the cyano-
hydrin derived from 3,4-dimethylbenzaldehyde was deter-
1
mined by H NMR analysis[23] of the free cyanohydrin ob-
tained by hydrolysis of the acetate,[25] in the presence of (R)-
mandelic acid and DMAP.[26] The cyanohydrin derivative ob-
Chem. Eur. J. 2010, 16, 11367 – 11375
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11369

M. North et al.
Table 2. Rate constants and enantioselectivities for the synthesis of cyanohydrin trimethylsilyl ethers catalysed
In the case of asymmetric cy-
anohydrin synthesis catalysed
by the isothiocyanatovanadium
complex 4b, fourteen aldehydes
by vanadium complex 4a.[a]
Entry
Aldehyde
Abbrev.
ka [mꢂ1 minꢂ1
]
kb [mꢂ1 minꢂ1
]
kavg [mꢂ1 minꢂ1
]
ee [%][b]
1
2
3
4
5
6
7
8
PhCHO
H
4-Cl
3-Cl
0.0037
0.0078
0.0177
0.0053
0.0014
0.0010
0.0073
0.0166
0.0019
0.0014
0.0172
0.0344
0.0855
0.0042
0.0091
0.0213
0.0049
0.0013
0.0009
0.0068
0.015
0.0040ꢃ0.0003
0.0084ꢃ0.0007
0.0195ꢃ0.0018
0.0051ꢃ0.0002
0.00135ꢃ0.00005
0.00095ꢃ0.00005
0.0071ꢃ0.0003
0.0158ꢃ0.0008
0.00195ꢃ0.00005
0.00105ꢃ0.00035
0.01585ꢃ0.00135
0.0322ꢃ0.0022
0.08795ꢃ0.00245
86
86
72
89
4-ClC6H4CHO
3-ClC6H4CHO
4-FC6H4CHO
4-MeC6H4CHO
3,4-Me2C6H3CHO
4-BrC6H4CHO
3-FC6H4CHO
3-MeC6H4CHO
4-tBuOC6H4CHO
4-F3CC6H4CHO
3,5-F2C6H3CHO
3,4-Cl2C6H3CHO
were
used
as
substrates
(Table 3) and again second-
order kinetics was observed in
all cases.[23] However, 4-me-
4-F
4-Me
3,4-Me2
4-Br
68
65[c]
84
thoxybenzaldehyde
(Table 3,
3-F
88
86
58
77
entry 6) was found to display
an anomalously slow rate of cy-
anation in reactions catalysed
9
10
11
3-Me
4-OtBu
4-CF3
3,5-F2
3,4-Cl2
0.002
0.0007
0.0145
0.03
by complex 4b. This was attrib- 12
82
13
0.0904
78[c]
uted to coordination of the me-
[a] All reactions were carried out in duplicate (to give ka and kb, respectively) in dichloromethane at 08C with
[aldehyde]0 =0.5m, [Me3SiCN]0 =0.55m and [4a]=1.0 mm. [b] Determined by chiral GC analysis of the cyano-
hydrin acetate unless stated otherwise. [c] Determined by 1H NMR spectroscopy in the presence of (R)-man-
delic acid and DMAP.
thoxy group to the vanadium
ion of the catalyst, thus, inhibit-
ing the catalysis. Support for
this hypothesis came from reac-
tions that use 4-tert-butoxyben-
zaldehyde as substrate (Table 3,
entry 11), which when incorpo-
Table 3. Rate constants and enantioselectivities for the synthesis of cyanohydrin trimethylsilyl ethers catalysed
by vanadium complex 4b.[a]
rated into
a Hammett plot
(Figure 3) produced a reasona-
ble fit to a straight line that had
a slope of 1.6. The enantiomeric
excesses of the cyanohydrin tri-
methylsilyl ethers were all
>56%,[23] again confirming that
they were produced by a cata-
lysed rather than uncatalysed
process.
Table 4 presents the kinetic
data obtained for the asymmet-
ric trimethylsilylcyanation of
fifteen aldehydes by using
2 mol% of the aluminium-
based catalyst 5 and 10 mol%
of triphenylphosphine oxide as
catalysts. These reactions all
obeyed first-order kinetics,[23]
Entry
Aldehyde
Abbrev.
ka [mꢂ1 minꢂ1
]
kb [mꢂ1 minꢂ1
]
kavg [mꢂ1 minꢂ1
]
ee [%][b]
1
2
3
4
5
6
7
8
PhCHO
H
4-Cl
3-Cl
0.204
0.2452
0.3709
0.4866
0.2876
0.1058
0.0161
0.0734
0.1903
0.6663
0.2238
0.0746
0.5717
2.0289
0.4194
0.2246ꢃ0.0206
0.3872ꢃ0.0163
0.4925ꢃ0.0059
0.2890ꢃ0.0014
0.1171ꢃ0.0113
0.0174ꢃ0.0013
0.0762ꢃ0.0028
0.1638ꢃ0.0265
0.6562ꢃ0.0101
0.2170ꢃ0.0068
0.0736ꢃ0.0100
0.6015ꢃ0.0298
1.9676ꢃ0.0613
0.4219ꢃ0.0025
87
84
77
85
4-ClC6H4CHO
3-ClC6H4CHO
4-FC6H4CHO
4-MeC6H4CHO
4-MeOC6H4CHO
3,4-Me2C6H3CHO
4-MeSC6H4CHO
3-FC6H4CHO
3-MeC6H4CHO
4-tBuOC6H4CHO
4-F3CC6H4CHO
3,5-F2C6H3CHO
4-BrC6H4CHO
0.4035
0.4983
0.2904
0.1284
0.0187
0.0789
0.1373
0.6461
0.2102
0.0726
0.6312
1.9063
0.4243
4-F
4-Me
4-MeO
3,4-Me2
4-MeS
3-F
3-Me
4-OtBu
4-CF3
3,5-F2
4-Br
85
96[c]
76[d]
57
9
86
78
78
75
75
81
10
11
12
13
14
[a] All reactions were carried out in duplicate (to give ka and kb, respectively) in dichloromethane at 08C with
[aldehyde]0 =0.5m, [Me3SiCN]0 =0.55m and [4b]=1.0 mm. [b] Determined by chiral GC analysis of the cyano-
hydrin acetate unless stated otherwise. [c] Determined by comparison of the specific rotation with literature
data.[27] [d] Determined by H NMR spectroscopy in the presence of (R)-mandelic acid and DMAP.
1
Figure 2. Hammett plot for the asymmetric addition of trimethylsilyl cya-
nide to different benzaldehydes catalysed by complex 4a (y=
1.8724xꢂ0.1167, R2 =0.9252) For abbreviations see Table 2.
Figure 3. Hammett plot for the asymmetric addition of trimethylsilyl cya-
nide to different benzaldehydes catalysed by complex 4b (y=
1.5705xꢂ0.1448, R2 =0.8511) For abbreviations see Table 3.
11370
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11367 – 11375

Asymmetric Cyanohydrin Synthesis
FULL PAPER
Table 4. Rate constants and enantioselectivities for the synthesis of cyanohydrin trimethylsilyl ethers catalysed
constants between ꢂ0.3 and
+0.4 suggests that all substrates
are capable of coordinating to
the Lewis acidic titanium ion
and that, in doing so, they are
activated towards addition of
cyanide to the carbonyl group.
The reaction constant (+2.4)
observed for the addition of tri-
methylsilyl cyanide to alde-
hydes catalysed by complex 3 is
almost identical to that ob-
tained for the uncatalysed addi-
tion of cyanide to aldehydes
(+2.3) and for the hydrolysis of
methyl esters (+2.2),[16] indicat-
ing that nucleophilic addition to
the carbonyl bond is the rate-
determining step in each case.
The vanadium-based catalysts
4a and b gave linear Hammett
plots with reaction constants of
by the bimetallic aluminium complex 5.[a]
Entry
Aldehyde
Abbrev.
ka [minꢂ1
]
kb [minꢂ1
]
kavg [minꢂ1
]
ee [%][b]
1
2
3
4
5
6
7
8
PhCHO
H
4-Cl
3-Cl
0.0015
0.0021
0.0024
0.0019
0.0016
0.0007
0.0010
0.0011
0.0033
0.0015
0.0008
0.0020
0.0025
0.0018
0.0017
0.0016
0.0020
0.0019
0.0024
0.0015
0.0006
0.0012
0.0013
0.0028
0.0017
0.0010
0.0021
0.0035
0.0016
0.0019
0.00155ꢃ0.00005
0.00205ꢃ0.00005
0.00215ꢃ0.00025
0.00215ꢃ0.00025
0.00155ꢃ0.00005
0.00065ꢃ0.00005
0.0011ꢃ0.00005
0.0012ꢃ0.0001
0.00305ꢃ0.00025
0.0016ꢃ0.0001
0.0009ꢃ0.0001
0.00215ꢃ0.00005
0.0030ꢃ0.0005
0.0017ꢃ0.0001
0.0018ꢃ0.0001
64
61
54
69
4-ClC6H4CHO
3-ClC6H4CHO
4-FC6H4CHO
4-MeC6H4CHO
4-MeOC6H4CHO
3,4-Me2C6H3CHO
4-MeSC6H4CHO
3-FC6H4CHO
3-MeC6H4CHO
4-tBuOC6H4CHO
4-F3CC6H4CHO
3,5-F2C6H3CHO
4-BrC6H4CHO
3,4-Cl2C6H3CHO
4-F
4-Me
4-MeO
3,4-Me2
4-MeS
3-F
3-Me
4-OtBu
4-CF3
3,5-F2
4-Br
65
71[c]
68[d]
40
9
61
73
58
49
10
11
12
13
14
15
38
60
3,4-Cl2
50[d]
[a] All reactions were carried out in duplicate (to give ka and kb, respectively) in dichloromethane at 08C with
[aldehyde]0 =0.5m, [Me3SiCN]0 =0.55m, [5]=1.0 mm and [Ph3PO]=5.0mm. [b] Determined by chiral GC anal-
ysis of the cyanohydrin acetate unless stated otherwise. [c] Determined by comparison of the specific rotation
with literature data.[27] [d] Determined by 1H NMR spectroscopy in the presence of (R)-mandelic acid and
DMAP.
and 4-methoxybenzaldehyde (Table 4, entry 6) was again
found to react with an anomalously slow rate, whereas 4-
tert-butoxybenzaldehyde (Table 4, entry 11) gave kinetic
data that were more consistent with the other substrates.
For this catalyst system, it was impossible to fit the rate data
to a single linear Hammett plot (Figure 4). The enantiomer-
1.9 and 1.6, respectively (Figure 2 and Figure 3). These are
lower reaction constants than that found for the titanium-
based complex 3 and suggest that some Lewis base catalysis
may also occur in these reactions. However, Lewis acid cat-
alysis is still dominant as all aldehydes with substituent con-
stants of ꢂ0.3 to +0.7 fitted the Hammett plot, indicating
that they were all activated by coordination to the vanadium
ion. The small degree of Lewis base catalysis occurring in
both of these reactions may be due to the V=O bond,[13] and
the greater degree of Lewis base catalysis in the case of
complex 4b may be due to Lewis base catalysis by the iso-
thiocyanate counterion. There is a vast difference in the ab-
solute rate of catalysis by complexes 4a and 4b since the
second-order rate constant for catalysis by complex 4b is
100 times higher than the corresponding rate constant for
complex 4a.[10b] The Hammett data indicate that this differ-
ence in rate of catalysis is partly due to Lewis base catalysis
by the counterion, but may also be due to differences in the
oligomeric nature of species formed in the reaction mix-
ture.[6d,10b,11b,14]
Figure 4. Hammett plot for the asymmetric addition of trimethylsilyl cya-
nide to different benzaldehydes catalysed by complex 5. For abbrevia-
tions see Table 4.
The data for the aluminium-based catalyst 5 are clearly
very different to the data obtained for the titanium- and va-
nadium-based catalysts. This is not too surprising, since a
Lewis base (triphenylphosphine oxide) is specifically added
to reactions catalysed by complex 5. The failure of all of the
aldehydes to fit a linear Hammett plot (Figure 4) is consis-
ic excesses of the cyanohydrin derivatives obtained by using
complex 5 and triphenylphosphine oxide were generally
lower than those obtained by using the other catalysts, but
were still >35%,[23] thus, confirming that the kinetics being
monitored were for the catalysed reaction.
The Hammett data for asymmetric cyanohydrin synthesis
catalysed by complex 3 can simply be interpreted in terms
of catalysis entirely by Lewis acid catalysis. There is clearly
a linear correlation between how electron-deficient the alde-
hyde is and the rate with which it reacts (Figure 1). The
linear correlation observed for all aldehydes with substituent
tent with aluminiumACTHNUTRGNE(NUG III) being a weaker Lewis acid than ti-
tanium(IV) or vanadium(V). Thus, only the more electron-
rich aldehydes (those with s <+0.35) are capable of coordi-
nating to the metal and of being activated by Lewis acid cat-
alysis. The more electron-deficient aldehydes show evidence
of predominant Lewis base catalysis, since there is no corre-
lation between the substituent constant of the aldehyde and
the reaction rate. However, the aldehyde must still be at
least weakly bound to the aluminium, since asymmetric cya-
Chem. Eur. J. 2010, 16, 11367 – 11375
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11371

M. North et al.
nohydrin synthesis is occurring within the sphere of influ-
ence of the chiral salen ligands as evidenced by the asym-
metric induction observed (Table 4). In this system, both
Lewis acid and Lewis base catalysis are operative, with
Lewis acid catalysis being more important for electron-rich
aldehydes than for electron-deficient substrates.
The differences between the Hammett plots for catalysts
3–5 are emphasised in Figure 5 in which all of the data are
plotted on the same vertical axis. On this scale, all the data
by a mechanism that involves predominantly Lewis base
rather than Lewis acid catalysis.
To test this hypothesis, the asymmetric addition of trime-
thylsilyl cyanide to eight methyl ketones was investigated. A
reaction carried out with acetophenone as substrate at
ꢂ408C by using 1.6 equivalents of trimethylsilyl cyanide,
2 mol% of complex 5 and 10 mol% of triphenylphosphine
oxide gave a conversion of just 12% after a reaction time of
16 h. Increasing the reaction temperature to room tempera-
ture and increasing the reaction time to 48 h improved the
conversion to 85% and gave the (S)-cyanohydrin trimethyl-
silyl ether with 51% enantiomeric excess. Increasing the
amount of complex 5 to 4 mol% did not improve either the
conversion or the enantioselectivity. Hence, all subsequent
reactions were carried out by using 2 mol% of complex 5
and 10 mol% of triphenylphosphine oxide at room tempera-
ture for two days. As shown in Table 5 (entries 1–7), all of
Table 5. Asymmetric addition of trimethylsilyl cyanide to acetophenones
catalysed by aluminium complex 5 and triphenylphosphine oxide.
Entry
Ketone
Conversion [%][a]
ee [%][b]
1
2
3
4
5
6
7
8
PhCOMe
85
99
99
99
91
62
54
100
51 (S)
51 (S)
47 (S)
49 (S)
55 (S)
50 (S)
55 (S)
44 (S)
4-ClC6H4COMe
3-ClC6H4COMe
4-BrC6H4COMe
4-FC6H4COMe
4-MeC6H4COMe
4-MeOC6H4COMe
MeCH2CH2COMe
^
Figure 5. Superimposition of the Hammett plots for catalysts 3–5. [ =3
*
(c, y=2.3715xꢂ0.0409), & = 4a (g, y=1.8724xꢂ0.1167),
= 4b
~
(b, y=1.5705xꢂ0.1448); = 5 (a, y=0.3888xꢂ0.0131)].
[a] Determined by 1H NMR analysis of the reaction mixture. [b] Deter-
mined by 1H NMR spectroscopy in the presence of (R)-mandelic acid
and DMAP.[26]
for complex 5 appears to fit a linear correlation that is
almost horizontal compared to the data for catalysts 3, 4a
and 4b. This emphasises the high degree of Lewis base cat-
alysis in reactions catalysed by complex 5/triphenylphos-
phine oxide. The difference in slope (1) between catalysts 3
and 5 (2.4 and 0.4, respectively) is particularly marked, with
the data for the vanadium-based catalysts 4a and 4b having
an intermediate slope of 1.9–1.6.
Previously, we have shown that the titanium-based cata-
lyst 3 is capable of inducing the asymmetric addition of tri-
methylsilyl cyanide to ketones,[6b,c] but the vanadium-based
catalysts 4a and 4b do not accept ketones as substrates.[10b]
In light of the results presented above, this can be rational-
ised in terms of the titanium-based catalyst 3 being a suffi-
ciently good Lewis acid to active ketones, whereas the vana-
dium-based catalysts 4a and 4b are less Lewis acidic and so
cannot activate the ketone. The alternative mode of catalysis
involving Lewis base activation of the trimethylsilyl cyanide
is also apparently not sufficiently effective to allow com-
plexes 4a and 4b to catalyse the asymmetric addition of tri-
methylsilyl cyanide to ketones. Since the Hammett data for
the asymmetric addition of trimethylsilyl cyanide to alde-
hydes catalysed by complex 5 and triphenylphosphine oxide
suggested that Lewis base catalysis was significant for all
substrates, we reasoned that the complex 5/triphenylphos-
phine oxide catalyst system might be capable of catalysing
the asymmetric addition of trimethylsilyl cyanide to ketones
the acetophenone derivatives gave the corresponding cyano-
hydrin trimethylsilyl ether with an enantiomeric excess be-
tween 47 and 55%,[23] though the conversion varied signifi-
cantly between the substrates, with electron-rich aromatic
ketones giving low conversions (entries 6 and 7), whereas
halogenated acetophenones gave conversions >90%.
Pentan-2-one also gave the corresponding cyanohydrin tri-
methylsilyl ether with complete conversion under these con-
ditions, but with a reduced enantioselectivity of just 44%
(Table 5, entry 8).
Attempts to extend the chemistry to other ketones were
not successful as no conversion was observed with propio-
phenone, and 1-tetralone gave just 21% conversion (ee not
determined). Attempts were also made to use the vanadi-
um-based complexes 4a and 4b to catalyse the asymmetric
addition of trimethylsilyl cyanide to acetophenone in the
presence of triphenylphosphine oxide. However, no reaction
was observed. It may be that the triphenylphosphine oxide
coordinates to the vanadium(V) complex, thus, preventing it
from acting as a Lewis acid.
11372
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11367 – 11375

Asymmetric Cyanohydrin Synthesis
FULL PAPER
through a silica plug and eluted with CH2Cl2 and the solvent was evapo-
rated to allow the enantiomeric excess of the cyanohydrin trimethylsilyl
ether to be determined.
Conclusion
A Hammett-plot analysis based on the substrate that poten-
tially undergoes Lewis acid activation can be used to deter-
mine the relative importance of Lewis acid and Lewis base
catalysis in asymmetric catalysis. Application of this method-
General method for kinetics experiments involving vanadiumACHTUNGTRENNUNG(salen)
complexes 4a and 4b: A solution of catalyst 4a or 4b (1.96 mmol) in dry
CH2Cl2 (1.75 mL) was cooled to 08C in a water/ice bath. An aliquot
(0.5 mL) was taken and diluted with CH2Cl2 (3.5 mL) to be used as the
reference sample for the UV spectrophotometer. Then, freshly distilled
aldehyde (0.985 mmol) was added to the reaction and another aliquot
(0.5 mL) was taken and diluted with CH2Cl2 (3.5 mL). The absorbance of
the sample at the wavelength corresponding to lmax of the aldehyde was
recorded. Me3SiCN (116.5 mg, 1.182 mmol) was added to the reaction
and aliquots (0.5 mL) were taken and quenched with CH2Cl2 (3.5 mL) at
appropriate time intervals over a period of 1 to 2 h (depending on the
nature of the aldehyde) for reactions catalysed by complex 4b, and 8 h
for reactions catalysed by complex 4a. The remaining solution was then
passed through a silica plug and eluted with CH2Cl2 and the solvent was
evaporated to allow the enantiomeric excess of the cyanohydrin trime-
thylsilyl ether to be determined.
ology to a series of metalACTHUNRGTNE(NUG salen) complexes, which all cata-
lyse the asymmetric addition of trimethylsilyl cyanide to al-
dehydes, revealed major differences in the relative impor-
tance of Lewis acid and Lewis base catalysis, corresponding
to activation of the aldehyde and trimethylsilyl cyanide, re-
spectively. On the basis of this mechanistic analysis, the
combination of the bimetallic aluminium salen complex 5
and triphenylphosphine oxide was predicted to accept ke-
tones as substrates and this was found to be the case.
General method for kinetics experiments involving bimetallic aluminium-
AHCTUNGTRENG(NUN salen) complex 5: Catalyst 5 (10 mg, 8.8 mmol) and Ph3PO (12 mg,
Experimental Section
43.2 mmol) were dissolved in freshly distilled CH2Cl2 (1.75 mL) and the
solution was cooled to 08C. A sample (0.50 mL) was removed and diluted
with CH2Cl2 (3.0 mL) to be used as the reference sample for the UV
spectrophotometer. Aldehyde (0.44 mmol) was added to the reaction
mixture and another sample (0.50 mL) was removed and diluted with
CH2Cl2 (3.0 mL). The absorbance of the sample at the wavelength corre-
sponding to lmax of the aldehyde was recorded. Me3SiCN (69 mg,
0.7 mmol) was added to the reaction mixture and the kinetics monitored
by taking samples (0.50 mL) and quenching them with CH2Cl2 (3.0 mL) at
appropriate time intervals over a period of approximately 6 h. The re-
maining solution was then passed through a silica plug and eluted with
CH2Cl2 and the solvent was evaporated to allow the enantiomeric excess
of the cyanohydrin trimethylsilyl ether to be determined. Characterising
data for known cyanohydrin trimethylsilyl ethers is given in the Support-
ing Information.
General: Dichloromethane was distilled over CaH2 under a nitrogen at-
mosphere immediately prior to use. For workup and chromatographic pu-
rification, commercial grade solvents were used. Trimethylsilyl cyanide
and benzaldehyde were distilled on a Bꢁchi B-580 Kugelrohr apparatus.
Benzaldehyde was freshly distilled prior to use. Other commercially
available chemicals (Alfa Aesar, Aldrich, Fluka, Riedel-de Haꢂn) were
used as received. Details of the instrumentation used to record the ana-
lytical data are given in the Supporting Information.
Synthesis of racemic cyanohydrin trimethylsilyl ethers
Racemic standards needed for enantiomeric excess analysis were pre-
pared as follows: An aldehyde (0.985 mmol) was added to a solution of
Bu4NSCN (30 mg, 0.1 mmol) in CH2Cl2 (1.75 mL) at room temperature.
To this solution, Me3SiCN (116.5 mg, 1.182 mmol) was added and the re-
action was stirred for 2 h. The solution then was passed through a silica
plug eluting with CH2Cl2 and the solvent evaporated to leave the racemic
cyanohydrin trimethylsilyl ether, which was used without further purifica-
tion.
Trimethylsilyloxy-(3,5-difluorophenyl)acetonitrile: [a]2D0 =ꢂ18.4 (c=1.0 in
1
CHCl3); H NMR (400 MHz, CDCl3, TMS): d=0.27 (s, 9H), 5.47 (s, 1H),
6.84 (tt, 3J(H,F)=8.8, 4J
ACTHNUTRGENNUG CAHTUNGTREN(NUGN H,H)=2.4 Hz, 1H), 6.9–7.1 ppm (m, 2H);
13C NMR (100 MHz, CDCl3, TMS): d=ꢂ0.4, 62.4, 104.8 (t, 2J
ACHTUNGTRENNUNG(C,F)=
Synthesis of cyanohydrin acetates: An unpurified cyanohydrin trimethyl-
silyl ether was dissolved in MeCN (2 mL), and Ac2O (1.5 mL, 1.58 mmol)
and ScACHTUNGTRENNUNG(OTf)3 (5 mg, 0.01 mmol) were added. The reaction was stirred for
30 min at room temperature, then the mixture was passed through a
short silica plug and eluted with MeCN. The resulting solution was used
for chiral GC analysis.[23]
25.1 Hz), 109.3 (d, 2J(C,F)=26.9 Hz), 118.2, 140.0 (t, 3J
ACTHNUTRGENNUG ACHUTTGNREN(NUGN C,F)=9.1 Hz),
ACTHNUTRGNEUNG
163.2 ppm (dd, 1J(CF)=249.4, 3J(C,F)=12.4 Hz); 19F NMR (376 MHz,
CDCl3, CFCl3): d=ꢂ107.6 ppm (t, 3J(FH)=7.5 Hz); IR (neat): n˜ =3096,
2962, 2903, 2243, 1626, 1602 cmꢂ1
; MS (ESI): m/z (%): 259 (25)
[M+H2O]+, 185 (100); HRMS (ESI): calcd for C11H13NOF2Si [M]+:
241.0735; found: 241.0731; ee determined by chiral GC analysis of the
corresponding acetate by using method 2:[23] Rt =14.8 min (R), Rt =
15.1 min (S).
Synthesis of cyanohydrins:[25] To a solution of a cyanohydrin acetate
(0.985 mmol) in ethanol (3 mL), p-TsOH·H2O (187 mg, 0.985 mmol) was
added, and the mixture was stirred at room temperature for 2 days. The
solvent was evaporated in vacuo and the residue was purified by column
chromatography by eluting with a gradient from 1:15 EtOAc/hexane to
1:6 EtOAc/hexane to give the cyanohydrin. To determine the enantio-
meric excess of the cyanohydrin, (R)-mandelic acid (2.74 mg, 18 mmol),
DMAP (1.73 mg, 18 mmol) and CDCl3 (0.6 mL) were mixed in an NMR
tube. The cyanohydrin (18 mmol) was then added and the solution ana-
Trimethylsilyloxy-(3,4-dimethylphenyl)acetonitrile: m.p. 31–328C; [a]D20
=
ꢂ24.1 (c=1.0 in CHCl3); 1H NMR (400 MHz, CDCl3, TMS): d=0.21 (s,
9H), 2.26 (s, 3H), 2.28 (s, 3H), 5.41 (s, 1H), 7.1–7.3 ppm (m, 3H);
13C NMR (100 MHz, CDCl3, TMS): d=ꢂ0.2, 19.6, 19.8, 63.6, 119.4, 123.9,
127.6, 130.1, 133.7, 137.4, 138.0 ppm; IR (ATR): n˜ =3017, 2960, 2924,
2239 cmꢂ1; MS (ESI): m/z (%): 251 (100) [M+H2O]+, 207 (95) [MꢂCN]+,
185 (40); HRMS (ESI): calcd for C13H19NOSi [M+H]+: 234.1314;
found:234.1305; ee determined by 1H NMR spectroscopy of the unpro-
tected cyanohydrin in the presence of (R)-mandelic acid and DMAP:
1
lysed by H NMR spectroscopy.[23,26]
General method for kinetics experiments involving bimetallic titanium-
ACHTUNGTRENNUNG(salen) complex 3: A solution of complex 3 (1.2 mg, 0.98 mmol) in dry
1H NMR (400 MHz, CDCl3, TMS): dH
ACHTUNTGRNENUG(R-CHCN)=5.36, dHACHTUNGTRENN(UGN S-CHCN)=
CH2Cl2 (1.75 mL) was cooled to 08C in a water/ice bath. An aliquot
(0.5 mL) was taken and diluted with CH2Cl2 (3.5 mL) to be used as the
reference sample for the UV spectrophotometer. Then, freshly distilled
aldehyde (0.985 mmol) was added to the reaction and another aliquot
(0.5 mL) was taken and diluted with CH2Cl2 (3.5 mL). The absorbance of
the sample at the wavelength corresponding to lmax of the aldehyde was
recorded. Me3SiCN (116.5 mg, 1.182 mmol) was added to the reaction
and aliquots (0.5 mL) were taken and quenched with CH2Cl2 (3.5 mL) at
appropriate time intervals over a period of 1 min to 1 h, depending on
the nature of the aldehyde. The remaining solution was then passed
5.31 ppm.
Trimethylsilyloxy-(4-tert-butoxyphenyl)acetonitrile: [a]2D0 =ꢂ17.4 (c=1.0
in CHCl3); 1H NMR (400 MHz, CDCl3, TMS): d=0.22 (s, 9H), 1.36 (s,
3
3
9H), 5.46 (s, 1H), 7.02 (d, JACTHNUTRGENNU(G H,H)=8.4 Hz, 2H), 7.36 ppm (d, JACHTUNGTRENNUNG(H,H)=
8.4 Hz, 2H); 13C NMR (100 MHz, CDCl3, TMS): d=ꢂ0.2, 28.8, 63.4,
79.0, 119.3, 124.2, 127.2, 130.9, 156.4 ppm; IR (neat): n˜ =3063, 3036, 2978,
2904, 2240, 1608, 1508 cmꢂ1; MS (ESI): m/z (%): 278 (50) [M+H]+, 276
(70), 242 (100); HRMS (ESI): calcd for C15H23NO2SiNa [M+Na]+:
300.1396; found: 300.1371; ee determined by chiral GC analysis of the
Chem. Eur. J. 2010, 16, 11367 – 11375
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11373

M. North et al.
corresponding acetate by using method 5:[23] Rt =72.9 min (R), Rt =
73.7 min (S).
Chem. Soc. 1999, 121, 3968–3973; b) Y. N. Belokon, B. Green, N. S.
Ikonnikov, M. North, V. I. Tararov, Tetrahedron Lett. 1999, 40,
8147–8150; c) Y. N. Belokon, B. Green, N. S. Ikonnikov, M. North,
T. Parsons, V. I. Tararov, Tetrahedron 2001, 57, 771–779; d) Y. N.
Belokon, W. Clegg, R. W. Harrington, M. North, C. Young, Inorg.
Chem. 2008, 47, 3801–3814.
Trimethylsilyloxy-(3,4-dichlorophenyl)acetonitrile: [a]2D0 =ꢂ15.9 (c=1.0
in CHCl3); 1H NMR (400 MHz, CDCl3, TMS): d=0.26 (s, 9H), 5.44 (s,
1H), 7.31 (dd, 3J
8.4 Hz, 1H), 7.57 ppm (d, 4J
CDCl3, TMS): d=ꢂ0.3, 62.4, 118.4, 125.4, 128.3, 131.0, 133.3, 133.7,
ACHTUNGTRENNUNG ACHTUNGTRENNUNG ACTUHNGTRENNUGN
(H,H)=8.4, 4J(H,H)=2.0 Hz, 1H), 7.50 (d, 3J
AHCTUNGTRENNUNG
[7] a) Y. N. Belokon, M. North, T. Parsons, Org. Lett. 2000, 2, 1617–
1619; b) M. North, M. Omedes-Pujol, Tetrahedron Lett. 2009, 50,
4452–4454.
136.3 ppm; IR (neat): n˜ =3094, 3025, 2961, 2901, 2242, 1595, 1568 cmꢂ1
;
MS (ESI): m/z (%): 296 (100) [M+Na]+; HRMS (ESI): calcd for
C11H13NOCl2SiNa [M+Na]+: 296.0041; found: 296.0026; ee determined
[8] M. North, C. Williamson, Tetrahedron Lett. 2009, 50, 3249–3252.
[9] a) Y. N. Belokon, B. Green, N. S. Ikonnikov, V. S. Larichev, B. V.
Lokshin, M. A. Moscalenko, M. North, C. Orizu, A. S. Peregudov,
G. I. Timofeeva, Eur. J. Org. Chem. 2000, 2655–2661; b) Y. N. Belo-
kon, A. J. Blacker, P. Carta, L. A. Clutterbuck, M. North, Tetrahe-
dron 2004, 60, 10433–10447; c) Y. N. Belokon, W. Clegg, R. W. Har-
rington, E. Ishibashi, H. Nomura, M. North, Tetrahedron 2007, 63,
9724–9740.
[10] a) Y. N. Belokon, V. I. Maleev, M. North, D. L. Usanov, Chem.
Commun. 2006, 4614–4616; b) Y. N. Belokon, W. Clegg, R. W. Har-
rington, V. I. Maleev, M. North, M. Omedes Pujol, D. L. Usanov, C.
Young, Chem. Eur. J. 2009, 15, 2148–2165.
[11] a) Y. N. Belokon, M. North, V. I. Maleev, N. V. Voskoboev, M. A.
Moskalenko, A. S. Peregudov, A. V. Dmitriev, N. S. Ikonnikov, H. B.
Kagan, Angew. Chem. 2004, 116, 4177–4181; Angew. Chem. Int. Ed.
2004, 43, 4085–4089; b) Y. N. Belokon, W. Clegg, R. W. Harrington,
C. Young, M. North, Tetrahedron 2007, 63, 5287–5299.
1
by H NMR spectroscopy of the unprotected cyanohydrin in the presence
1
of (R)-mandelic acid and DMAP: H NMR (400 MHz, CDCl3, TMS): dH-
ACHTUNGTRENNUNG(R-CHCN)=5.30 ppm, dHACHTUNGTREN(NUNG S-CHCN)=5.22 ppm.
General method for the asymmetric addition of trimethylsilyl cyanide to
ketones: Ph3PO (12 mg, 0.04 mmol) and catalyst 5 (10 mg, 8.4 mmol) were
dissolved in CH2Cl2 (1.0 mL) and ketone (0.42 mmol) was added in one
portion. Me3SiCN (66 mg, 0.67 mmol) was then added and the resulting
solution was stirred at room temperature for 48 h. After this time, the re-
action mixture was passed through a silica plug and eluted with CH2Cl2.
The solvent was removed under reduced pressure and the residue puri-
fied by column chromatography by eluting with Et2O/petrol to give the
trimethylsilyl protected cyanohydrin product. To determine the enantio-
meric excess, the cyanohydrin trimethylsilyl ether was dissolved in
CH2Cl2 (2 mL) and hydrochloric acid (1 mL) was added in one portion.
The reaction was stirred vigorously at room temperature for one hour,
then the organic layer was separated and the aqueous layer was washed
with CH2Cl2 (2ꢃ2 mL). The combined organic layers were washed with
brine (2 mL) and dried over MgSO4. The solvent was removed under re-
duced pressure to leave the free cyanohydrin as an oil. The product was
dissolved in CDCl3 and (R)-mandelic acid (100 mg, 0.66 mmol) and a cat-
alytic amount of DMAP were added. The sample was then analysed by
1H NMR spectroscopy.[23] Characterising data for the cyanohydrin trime-
thylsilyl ethers is given in the Supporting Information.
[12] M. North, P. Villuendas, C. Williamson, Tetrahedron 2010, 66, 1915–
1924.
[13] a) B. Cashin, D. Cunningham, J. F. Gallagher, P. McArdle, T. Hig-
gins, J. Chem. Soc. Chem. Commun. 1989, 1445–1446; b) B. Cashin,
D. Cunningham, J. F. Gallagher, P. McArdle, T. Higgins, Polyhedron
1989, 8, 1753–1755; c) N. F. Choudhary, P. B. Hitchcock, G. J. Leigh,
S. W. Ng, Inorg. Chim. Acta 1999, 293, 147–154; d) B. Cashin, D.
Cunningham, P. Daly, P. McArdle, M. Munroe, N. N. Chonchubhair,
Inorg. Chem. 2002, 41, 773–782.
[14] a) M. Mathew, A. J. Carty, G. J. Palenik, J. Am. Chem. Soc. 1970, 92,
3197–3198; b) A. Hills, D. L. Hughes, G. J. Leigh, J. R. Sanders, J.
Chem. Soc. Dalton Trans. 1991, 61–64; c) D. L. Hughes, U. Kleinkes,
G. J. Leigh, M. Maiwald, J. R. Sanders, C. Sudbrake, J. Chem. Soc.
Dalton Trans. 1994, 2457–2466; d) M. Kojima, K. Nakajima, M. Tsu-
chimoto, M. Tanaka, T. Suzuta, Y. Yoshikawa, J. Fujita, Chem. Lett.
1994, 949–952; e) S. A. Fairhurst, D. L. Hughes, U. Kleinkes, J.
Leigh, J. R. Sanders, J. Weisner, J. Chem. Soc. Dalton Trans. 1995,
321–326; f) K. Yamamoto, K. Oyaizu, E. Tsuchida, J. Am. Chem.
Soc. 1996, 118, 12665–12672; g) K. Oyaizu, K. Yamamoto, K.
Yoneda, E. Tsuchida, Inorg. Chem. 1996, 35, 6634–6635; h) K. Na-
kajima, M. Kojima, S. Azuma, R. Kasahara, M. Tsuchimoto, Y. Ku-
bozono, H. Maeda, S. Kashino, S. Ohba, Y. Yoshikawa, J. Fujita,
Bull. Chem. Soc. Jpn. 1996, 69, 3207–3216; i) R. Kasahara, M. Tsu-
chimoto, S. Ohba, K. Nakajima, H. Ishida, M. Kojima, Inorg. Chem.
1996, 35, 7661–7665; j) N. F. Choudhary, N. G. Connelly, P. B. Hitch-
cock, G. J. Leigh, J. Chem. Soc. Dalton Trans. 1999, 4437–4446;
k) E. Tsuchida, K. Oyaizu, E. L. Dewi, T. Imai, F. C. Anson, Inorg.
Chem. 1999, 38, 3704–3708; l) M. Tsuchimoto, G. Hoshina, N. Yosh-
ioka, H. Inoue, K. Nakajima, M. Kamishima, M. Kojima, S. Ohba, J.
Solid State Chem. 2000, 153, 9–15; m) M. Tsuchimoto, E. Yasuda, S.
Ohba, Chem. Lett. 2000, 562–563; n) D. M. Boghaei, S. Mohebi, J.
Mol. Catal. A 2002, 179, 41–51; o) P. B. Chatterjee, D. Mandal, A.
Audhya, K.-Y. Choi, A. Endo, M. Chaudhury, Inorg. Chem. 2008,
47, 3709–3718.
Acknowledgements
The authors thank the EPSRC for financial support and a studentship (to
M.O.P.).
[1] a) M. North, Synlett 1993, 807–820; b) M. North, Tetrahedron:
Asymmetry 2003, 14, 147–176; c) J.-M. Brunel, I. P. Holmes, Angew.
Chem. 2004, 116, 2810–2837; Angew. Chem. Int. Ed. 2004, 43, 2752–
2778.
[2] a) T. R. J. Achard, L. A. Clutterbuck, M. North, Synlett 2005, 1828–
1847; b) N. H. Khan, R. I. Kureshy, S. H. R. Abdi, S. Agrawal, R. V.
Jasra, Coord. Chem. Rev. 2008, 252, 593–623; c) M. North, D. L.
Usanov, C. Young, Chem. Rev. 2008, 108, 5146–5226.
[3] S. E. Denmark, G. L. Beutner, Angew. Chem. 2008, 120, 1584–1663;
Angew. Chem. Int. Ed. 2008, 47, 1560–1638.
[4] For recent reviews of asymmetric cooperative catalysis see: a) D. H.
Paull, C. J. Abraham, M. T. Scerba, E. Alden-Danforth, T. Lectka,
Acc. Chem. Res. 2008, 41, 655–663; b) S. Matsunaga, M. Shibasaki,
Bull. Chem. Soc. Jpn. 2008, 81, 60–75; c) M. Shibasaki, S. Matsuna-
ga, Synlett 2008, 1583–1602; d) J. Gawronski, N. Wascinska, J.
Gajewy, Chem. Rev. 2008, 108, 5227–5252; e) C. Schneider, Angew.
Chem. 2009, 121, 2116–2118; Angew. Chem. Int. Ed. 2009, 48, 2082–
2084; f) M. Shibasaki, M. Kanai, S. Matsunaga, N. Kumagai, Acc.
Chem. Res. 2009, 42, 1117–1127; g) C. Nꢄjera, J. M. Sansano, J. M.
Saꢄ, Eur. J. Org. Chem. 2009, 2385–2400.
[15] a) S. J. Coles, M. B. Hursthouse, D. G. Kelly, A. J. Toner, N. M.
Walker, J. Organomet. Chem. 1999, 580, 304–312; b) J. Sun, M.
Yang, F. Yuan, X. Jia, X. Yang, Y. Pan, C. Zhu, Adv. Synth. Catal.
2009, 351, 920- 930.
[16] A. Williams, Free Energy Relationships in Organic and Bio-organic
Chemistry, RSC, Cambridge, 2003.
[5] a) Y. Wen, X. Huang, J. Huang, Y. Xiong, B. Qin, X. Feng, Synlett
2005, 2445–2448; b) B. Qin, X. Liu, J. Shi, K. Zheng, H. Zhao, X.
Feng, J. Org. Chem. 2007, 72, 2374–2378.
[17] M. Gasperini, F. Ragaini, Organometallics 2004, 23, 995–1001.
[18] a) S. Otto, F. Bertoncin, J. B. F. N. Engberts, J. Am. Chem. Soc. 1996,
118, 7702–7707; b) N. Xie, R. A. Binstead, E. Block, W. D. Chan-
[6] a) Y. N. Belokon, S. Caveda-Cepas, B. Green, N. S. Ikonnikov, V. N.
Khrustalev, V. S. Larichev, M. A. Moscalenko, M. North, C. Orizu,
V. I. Tararov, M. Tasinazzo, G. I. Timofeeva, L. V. Yashkina, J. Am.
11374
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11367 – 11375

Asymmetric Cyanohydrin Synthesis
FULL PAPER
dler, D. G. Lee, T. J. Meyer, M. Thiruvazhi, J. Org. Chem. 2000, 65,
1008–1015; c) D. Landini, A. Maia, C. Pinna, J. Chem. Soc. Perkin
Trans. 2 2001, 2314–2317; d) J. Choudhury, S. Roy, J. Mol. Catal. A
2008, 279, 37–46.
[23] Data included in the Supporting Information.
[24] S. Norsikian, I. Holmes, F. Lagasse, H. B. Kagan, Tetrahedron Lett.
2002, 43, 5715–5717.
[25] T. Sakai, K. Wang, T. Ema, Tetrahedron 2008, 64, 2178–2183.
[26] L. S. Moon, R. S. Jolly, Y. Kasetti, P. V. Bharatam, Chem. Commun.
2009, 1067–1069.
[27] J. Brussee, E. C. Roos, A. Van der Gen, Tetrahedron Lett. 1988, 29,
4485–4488.
[19] D. Tichit, M. H. Lhouty, A. Guida, B. H. Chiche, F. Figueras, A.
Auroux, D. Bartalini, E. Garrone, J. Catal. 1995, 151, 50–59.
[20] M. P. Doyle, M. Valenzuela, P. Huang, Proc. Natl. Acad. Sci. USA
2004, 101, 5391–5395.
[21] S. E. Denmark, T. Bui, J. Org. Chem. 2005, 70, 10393–10399.
[22] L. A. Dakin, P. C. Ong, J. S. Panek, R. J. Staples, P. Stavropoulos,
Organometallics 2000, 19, 2896–2908.
Received: April 23, 2010
Published online: August 16, 2010
Chem. Eur. J. 2010, 16, 11367 – 11375
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11375