Mechanistic comparison of aluminium based catalysts for asymmetric cyanohydrin synthesis

Article abstract of DOI:10.1016/j.tet.2010.01.004

A kinetic analysis of the asymmetric addition of trimethylsilyl cyanide to benzaldehyde using three aluminium based catalysts has been carried out. All three catalysts displayed rate equations, which were first order in trimethylsilyl cyanide concentratio

Full text of DOI:10.1016/j.tet.2010.01.004

Tetrahedron 66 (2010) 1915–1924
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Mechanistic comparison of aluminium based catalysts for asymmetric
cyanohydrin synthesis
*
Michael North , Pedro Villuendas, Courtney Williamson
School of Chemistry and University Research Centre in Catalysis and Intensified Processing, Bedson Building, Newcastle University, Newcastle upon Tyne. NE1 7RU, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A kinetic analysis of the asymmetric addition of trimethylsilyl cyanide to benzaldehyde using three
aluminium based catalysts has been carried out. All three catalysts displayed rate equations, which were
first order in trimethylsilyl cyanide concentration and zero order in benzaldehyde concentration. The
results are consistent with a common mechanism for effective asymmetric catalysis of cyanohydrin
synthesis, involving combined activation of the aldehyde by a Lewis acid and activation of the trime-
thylsilyl cyanide by a Lewis base. The mechanistic analysis was also applied to a magnesium-based
catalyst system to demonstrate its general applicability.
Received 18 September 2009
Received in revised form 27 November 2009
Accepted 4 January 2010
Available online 11 January 2010
Keywords:
Aluminium
Cyanohydrin
Kinetics
Ó 2010 Elsevier Ltd. All rights reserved.
Phosphine oxide
Lewis acid
Lewis base
1. Introduction
action of a particular catalyst with a particular cyanide source,1–4
however there has been no attempt to study the reaction more
Over the last fifteen years there has been an explosion in interest
in asymmetric cyanohydrin synthesis in which a chiral catalyst is
used to induce the asymmetric addition of a cyanide source to al-
dehydes or ketones (Scheme 1). The chiral catalyst may be a metal
derived Lewis acid,1,2 a synthetic organocatalyst,2–4 or an enzyme.5
Some of these processes have been commercialized for the syn-
generally. Such an investigation might reveal the relative importance
of acidic versus basic catalysis, activation of the carbonyl versus ac-
tivation of cyanide, and ultimately result in a generally applicable
mechanism or mechanisms, which are valid for all catalysts and
cyanide sources. In this manuscript, we report the results of a kinetic
study of three different, aluminium based, catalysts for the asym-
metric addition of trimethylsilyl cyanide to aldehydes.
thesis of
a-hydroxyacids and
b
-aminoalcohols.5–7
O
OX
catalyst
2. Kinetics
+ XCN
R
H
R
CN
Catalysts 1–3 were chosen for this study. Each of these aluminium
complexes can be prepared by a relatively short synthesis, requires
the presence of an external phosphine oxide for optimal activity and
is known to catalyse the asymmetric addition of trimethylsilyl cya-
nide to benzaldehyde, which was chosen as the test reaction. Thus,
these three catalysts would allow the influence of both the Lewis acid
and the Lewis/Brønsted base on the reaction mechanism to be pro-
bed. Whilst complexes 1–3 all have the same Lewis acidic metal, they
differ with respect to the internal base. Complex 1, which was de-
veloped by Shibasaki,12,13 possesses pendant phosphine oxides,
which have been proposed to act as Lewis bases and to activate the
trimethylsilyl cyanide. The utility of complex 1 has been shown by its
use in a number of total syntheses.14–17
Scheme 1. Catalytic asymmetric cyanohydrin synthesis.
Whilst the mechanism of achiral cyanohydrin synthesis using
basified hydrogen cyanide was elucidated by Lapworth over a cen-
tury ago,8–11 the mechanisms involved in asymmetric cyanohydrin
synthesis are more complex. Catalysts have been developed which
incorporate Lewis acidic and/or Lewis basic functionalities and
which will transfer cyanide from a wide range of cyanide sources
including: metal cyanides, trimethylsilyl cyanide, acyl cyanides,
cyanoformates, cyanophosphonates and other cyanohydrins.1–4
Many authors have studied or speculated on the mechanism of
Najera’s catalyst 2, retains the same axially chiral binol unit as
complex 1, but the pendant phosphine oxides are replaced by
* Corresponding author. Tel.: þ44 191 222 7128; fax: þ44 191 222 6929.
E-mail address: michael.north@ncl.ac.uk (M. North).
0040-4020/$ – see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.01.004

1916
M. North et al. / Tetrahedron 66 (2010) 1915–1924
tertiary amines, which are proposed to act as Brønsted bases to
activate and pre-organise hydrogen cyanide formed in situ from
trimethylsilyl cyanide and water from molecular sieves added to
the reaction mixture.18,19 The versatility of complex 2 has been
demonstrated by its ability to catalyse asymmetric cyanohydrin
1) NaH (2.5 eq.)
OH 2)Et2NCOCl (2.5 eq.)
OCONEt2
OCONEt2
OH
74%
synthesis from a wide range of different cyanide sources.20–26
A
4
closely related catalyst in which the diethylamino groups of com-
plex 2 are changed to morpholino groups has also been shown to
catalyse the asymmetric addition of trimethylsilyl cyanide to
aldehydes in the presence of a phosphine oxide additive.27,28
1) s-BuLi / TMEDA
2) NH4Cl (aq.)
CONEt2
NEt2
NEt2
1) LiAlH4
2) KF (aq.)
OH
OH
OH
OH
Ph2(O)P
ClAl
Et2N
63% from 4
O
O
O
O
ClAl
Et2N
tBu
CONEt2
5
6
Ph2(O)P
Scheme 2. Synthesis of BINOLAM.
1
2
tBu
Table 1
Synthesis of O-trimethylsilyl mandelonitrile using catalysts 1–3
Solvent Temp Time Conversion Eea
tBu tBu
O
Catalyst Co-catalyst
(mol %) (mol %)
(ꢁC)
(h)
(%)
(%)
N
O
O
N
N
1 (10)
CH2Cl2 ꢀ40
Toluene ꢀ20
CH2Cl2 ꢀ40
36
6
16
96
16
2
95 (91)b
99 (99)d
8029
93
16
75
83 (87)b
Al
Al
2 (10)c Ph3PO (40)
>99.5 (>99.5)d
N
O
O
3 (2)
1 (10)
1 (4)
1 (4)
1 (4)
2 (9)c
2 (9)c
2 (2)c
2 (2)
2 (2)
3 (1)
3 (1)
Ph3PO (10)
8929
69
15
4
tBu tBu
MePh2PO (40) CH2Cl2 ꢀ40
MePh2PO (40) CH2Cl2
0
MePh2PO (40) CH2Cl2 20
MePh2PO (16) CH2Cl2 20
MePh2PO (40) Toluene ꢀ20
2
68
3
tBu
tBu
6
99
99
3
MePh2PO (40) Toluene
MePh2PO (10) Toluene
MePh2PO (10) Toluene
0
0
0
4
4
4
99
63
49
95
65
60
MePh2PO (10) Toluene 20
Ph3PO (10)
Ph3PO (10)
4
16
6
79
46
Complex 3 whose catalytic activity was recently reported by our
group29 does not possess any pendant basic features, but still re-
quires the presence of triphenylphosphine oxide as a co-catalyst. A
number of related monometallic aluminium(salen) complexes have
also been shown to catalyse the asymmetric addition of trime-
thylsilyl cyanide to aldehydes and ketones, but in each case
a phosphine oxide30–33 or N-oxide34,35 co-catalyst is required. Alu-
minium(salen)alkoxides have also been found to be effective
asymmetric catalysts for the addition of cyanoformates to acylsi-
lanes, leading to cyanohydrin trimethylsilyl ethers after a 1,2-Brook
rearrangement.36,37 In this case no phosphine oxide co-catalyst was
required, but it was shown that the alkoxide could perform a similar
role, liberating cyanide from the cyanoformate.
CH2Cl2 ꢀ20
10029
10029
8029
6829
CH2Cl2
0
a
In all cases, the S-enantiomer of the cyanohydrin was obtained from the R-en-
antiomer of the catalyst.
b
Values in brackets are those reported in the literature.12,13
c
Reaction carried out in the presence of 4 Å molecular sieves.
d
Values in brackets are those reported in the literature.18,19
1-3 (cat)/
R3PO (cat)
Ac2O /
Sc(OTf)3 /
MeCN
O
OSiMe3
OAc
CN
+ Me3SiCN
Ph
H
Ph
CN
Ph
7
8
Scheme 3. Synthesis of (S)-mandelonitrile derivatives.
Catalyst 3 was prepared by the route we have previously
reported.38 Catalysts 1 and 2 are generated in situ from the binol
ligand and dimethylaluminium chloride.12–26 The binol ligand
needed for the preparation of complex 1 was prepared as previously
reported by Shibasaki,12,13 however, for the synthesis of the ligand
(BINOLAM) needed for the formation of Najera’s catalyst 2, we de-
veloped a three-step synthesis (Scheme 2) based on a bis-anionic
Fries rearrangement reported by Dennis and Woodward.39 Thus,
reaction of (R)-binol with N,N-diethylcarbamoyl chloride gave bis-
urethane 4, which on treatment with s-BuLi followed by acidic
work-up gave bis-amide 5. Reduction of amide 5 with lithium alu-
miniumhydride gave BINOLAM 6 in 47% overall yield.
Catalysts 1-3 were first tested for the asymmetric addition of
trimethylsilyl cyanide to benzaldehyde using conditions reported
in the literature to confirm the efficacy of in situ prepared catalysts
1 and 2 (Table 1). The enantiomeric excess of the O-trimethylsilyl
mandelonitrile 7 was determined by chiral GC analysis of O-acetyl
mandelonitrile 8 prepared by the method of Kagan40 (Scheme 3). In
each case, the activity and enantioselectivity of the catalyst was
comparable to that reported in the literature.
For catalyst 1, the kinetics had to be monitored at ꢀ40 ꢁC as at
higher temperatures the enantioselectivity of the reactions was
diminished to such an extent that it was not possible to be confident
that a catalysed reaction was being monitored rather than racemic
background reaction (Table 1). Whilst a reaction carried out in the
absence of phosphine oxide went to 95% conversion in 36 h at
ꢀ40 ꢁC (Table 1), addition of methyldiphenylphosphine oxide sig-
nificantly reduced the reaction rate but enhanced the enantiose-
lectivity as previously reported by Shibasaki.12,13 In the presence of
methyldiphenylphosphine oxide, reactions typically required four
days to go to completion. Never the less, kinetic experiments were
carried out in the presence of methyldiphenylphosphine oxide,
monitoring the reactions by 1H NMR spectroscopy in CD2Cl2 since
this allowed the kinetic influence of the Lewis base to be in-
vestigated. Shibasaki also reported12,13 that slow addition of tri-
methylsilyl cyanide enhanced the enantioselectivity of reactions,
but this was not compatible with our kinetic study, so all of the
trimethylsilyl cyanide was added at the start of the reaction.

M. North et al. / Tetrahedron 66 (2010) 1915–1924
1917
For catalyst 2 however, the standard conditions were hetero-
0.3
0.25
0.2
A
geneous due to the presence of 4 Å molecular sieves. To avoid this
complication, variations to the reaction conditions were sought
where the reaction would proceed in the absence of molecular
sieves. In addition, the triphenylphosphine oxide was changed to
methyldiphenylphosphine oxide to allow direct comparison with
results obtained using Shibasaki’s system. As can be seen from
Table 1, changing the phosphine oxide had no effect on the re-
activity or enantioselectivity of the catalyst system. Conducting the
reaction at an experimentally more convenient 0 ꢁC had only
a slightly detrimental effect on the enantioselectivity in this case.
Reducing the catalyst and phosphine oxide loadings to 2 mol % and
10 mol %, respectively did reduce both the yield and enantiose-
lectivity, though these were still significant. Omission of the 4 Å
molecular sieves at 0 ꢁC reduced the rate of reaction still further,
but had only a slightly detrimental effect on the enantioselectivity.
Finally, reactions carried out at 20 ꢁC proceeded to 79% conversion
in 4 h, albeit with a reduced enantioselectivity of 46%. These con-
ditions were adopted as the most convenient as they allowed re-
actions carried out in toluene-d8 to be readily monitored by 1H
NMR spectroscopy as discussed above for the Shibasaki system.
For reactions catalysed by complex 3, the standard conditions
were too slow to allow convenient monitoring of the reaction ki-
netics. Therefore, the reaction temperature was increased. At the
same time, the amount of catalyst 3 employed was decreased to
prevent complications caused by the UV or NMR spectrum of the
salen ligand interfering with the kinetic data. In this case however,
triphenylphosphine oxide was retained as the Lewis base as alter-
[Me3SiCN]=0.11M
[Me3SiCN]=0.21M
[Me3SiCN]=0.42M
[Me3SiCN]=0.63M
0.15
0.1
0.05
0
0
1000
2000
3000
4000
5000
6000
Time (minutes)
0.3
0.25
0.2
B
[Me3SiCN]=0.46M
[Me3SiCN]=0.86M
[Me3SiCN]=1.54M
0.15
0.1
0.05
0
natives were known to significantly reduce the reaction rate.29
A
0
50
100
reaction at ꢀ20 ꢁC still required 16 h to go to completion, but at 0 ꢁC
reactions were complete in 6 h and the enantioselectivity (68%) was
still acceptable. Under these conditions, the disappearance of benz-
aldehyde could be conveniently monitored from the UV absorbance
at 246 nm of aliquots removed from the reaction mixture as we have
previously reported for other metal(salen) complexes.41–48
Time (minutes)
0.3
0.25
0.2
C
[Me3SiCN]=0.21M
[Me3SiCN]=0.40M
[Me3SiCN]=0.82M
Duplication of kinetics experiments showed that they gave rate
data, which was reproducible within an error limit of ꢂ6%. Having
found reaction conditions under which the kinetics of asymmetric
cyanohydrin synthesis catalysed by each of complexes 1–3 could be
monitored, the complete rate equation for each catalyst system was
determined by systematically varying the initial concentration of
each component of the reaction. Reactions catalysed by each of
complexes 1–3 were found to obey the same rate equation (Eq. 1)
and to be first order in trimethylsilyl cyanide and zero order in
benzaldehyde. The overall first order nature of the reactions is il-
lustrated in Figure 1, and the dependence of the rate on the initial
concentration of trimethylsilyl cyanide is shown for each catalyst in
Figure 2. Experiments carried out at different initial concentrations
0.15
0.1
0.05
0
0
100
200
300
400
500
Time (minutes)
Figure 2. Demonstration that the rate of reactions catalysed by complexes 1–3 de-
pends on the initial concentration of trimethylsilyl cyanide. (A) Catalyst 1:
[1]¼0.025 M, [MePh2PO]¼0.093 M; (B) Catalyst 2: [2]¼0.005 M, [MePh2PO]¼0.035 M;
(C) Catalyst 3: [3]¼0.005 M, [Ph3PO]¼0.025 M.
Time (minutes)
-1
of benzaldehyde confirmed that for each of catalysts 1–3, the rate
was independent of the benzaldehyde concentration.
0
2000
4000
6000
-1.5
-2
y = -0.0003x - 1.377
R² = 0.9902
a
b
Rate ¼ kobs½Me3SiCNꢃ where kobsk½catalystꢃ ½R3POꢃ
(1)
(2)
-2.5
-3
y = -0.0015x - 1.3097
R² = 0.9886
so logðkobsÞ ¼ logðkÞ þ a logð½catalystꢃÞ þ blogð½R3POꢃÞ
The lack of dependence of the rate of reaction on the concen-
tration of benzaldehyde can be interpreted in two ways:
-3.5
-4
y = -0.0212x - 1.2257
R² = 0.9896
1. The aldehyde is only involved in the mechanism after the rate
determining step.
2. The catalytic cycle starts with rapid, reversible binding of the
aldehyde to the catalyst, resulting in formation of a catalyst–
aldehyde complex whose concentration remains constant
throughout the reaction (saturation kinetics).
-4.5
Figure 1. First order kinetics plots for catalysts 1–3. Catalyst 1: filled diamonds
([PhCHO]0¼0.25 M, [Me3SiCN]0¼0.42 M, [1]¼0.025 M, [MePh2PO]¼0.093 M). Catalyst
2: unfilled triangles ([PhCHO]0¼0.26 M, [Me3SiCN]0¼0.39 M, [2]¼0.005 M, [MePh2
-
PO]¼0.035 M). Catalyst 3: filled circles ([PhCHO]0¼0.25 M, [Me3SiCN]0¼0.40 M,
[3]¼0.005 M, [Ph3PO]¼0.025 M).

1918
M. North et al. / Tetrahedron 66 (2010) 1915–1924
To distinguish between these two possibilities, NMR experiments
with respect to catalyst concentration of 0.5 is entirely consistent
with a model in which the precatalyst is associated into dimers or
higher species, but the catalytically active species is monomeric.
Figure 4 shows that whilst reactions catalysed by complexes 2
and 3 were first order in phosphine oxide concentration, that cat-
alysed by complex 1 was zero order in phosphine oxide concen-
tration. Shibasaki has previously reported preliminary kinetic data
showing that the rate of reaction decreases in the presence of
phosphine oxide.12,13 This is consistent with our own preliminary
data reported in Table 1. However, the effect of varying the catalyst
to phosphine oxide ratio had not previously been reported. In all of
these reactions however, an excess (2–5 equiv) of phosphine oxide
with respect to binol–aluminium complex was employed. Thus, the
simplest explanation for the observed zero-order dependence is
that one equivalent of phosphine oxide is required to bind to the
aluminium of the active catalyst, changing its geometry to trigonal
bipyramidal as previously reported, and that the phosphine oxide
has no other role to play in the mechanism.
In contrast, the first order dependence on phosphine oxide
concentration observed for catalysts 2 and 3 (using 2–15 equiv of
phosphine oxide with respect to binol–aluminium complex in the
case of complex 2 and 10–30 equiv of phosphine oxide with respect
to aluminium(salen) complex in the case of complex 3) suggests
that the phosphine oxide has a second role in addition to co-
ordination to the aluminium of the catalyst. The most obvious such
role is to react with the trimethylsilyl cyanide as previously
reported by Corey.50,51 The different kinetic behaviour of complex
1 can then be accounted for by its internal phosphine oxides ful-
filling this role. In the case of catalyst 2, it should be noted however,
that whilst this analysis may be valid for the homogeneous con-
ditions employed in this work to study the kinetics, the optimised
conditions developed by Najera involved the addition of 4 Å mo-
lecular sieves to the reaction and it was shown that under these
conditions the active cyanating agent was hydrogen cyanide rather
than trimethylsilyl cyanide.18,19
were undertaken. For all three catalysts, the 1H and 13C NMR spectra
of benzaldehyde (in deuterated dichloromethane for catalysts 1 and
3 and in deuterated toluene for catalyst 2) showed no significant
changes when equimolar amounts of catalyst or when equimolar
amounts of catalyst and phosphine oxide were added. This strongly
suggested that there was no significant formation of catalyst–alde-
hyde complexes. To investigate the possibility that the catalyst–al-
dehyde complexes only form in the presence of trimethylsilyl
cyanide, reactions were carried out in the appropriate deuterated
solvent using 10 mol % of each catalyst and a 10% excess of benzal-
dehyde relative to trimethylsilyl cyanide. Analysis of the reaction
mixtures by 1H NMR spectroscopy again provided no evidence for
the presence of catalyst–aldehyde complexes. Thus, the kinetic data
for all three catalysts was interpreted on the basis that the aldehyde
is only involved in the mechanism after the rate determining step.
Having determined that all of the catalysts obey the same rate
equation, the order with respect to both aluminium complex
(Fig. 3) and phosphine oxide (Fig. 4) was determined for all three
catalysts by varying the concentrations of these two components
(Eq. 2). Figure 3 suggests that reactions catalysed by complexes 1
and 3 exhibit a first order dependence of the reaction rate on cat-
alyst concentration. This was confirmed by a plot of [cat] against
kobs which in both cases gave a straight line passing through the
origin. For catalyst 2 however, the order with respect to catalyst
concentration was found to be ca. 0.5. Najera has shown19 that the
asymmetric addition of trimethylsilyl cyanide to aldehydes cata-
lysed by complex 2/triphenylphosphine oxide does not show a non-
linear effect.49 In contrast, the addition of other cyanide sources to
aldehydes catalysed by complex 2 in the absence of triphenyl-
phosphine oxide does display a non-linear effect.20,22 This suggests
that in the absence of phosphine oxide, complex 2 forms aggregates
in solution and that one role of the triphenylphosphine oxide is to
dissociate these aggregates. We have previously shown as part of
a study of other catalyst systems,42 that the observation of an order
The full rate equations for catalysts 1–3 are summarised in
Eqs. 3–5.
log([cat])
-2
Catalyst 1 : rate ¼ k½1ꢃ½Me3SiCNꢃ
(3)
-3
-2.5
-2
-1.5
-1
y = 0.8048x - 1.034
R² = 0.9818
-2.5
-3
0:5
catalyst 1
catalyst 2
catalyst 3
Catalyst 2 : rate ¼ k½2ꢃ ½MePh2POꢃ½Me3SiCNꢃ
(4)
(5)
y = 0.438x - 2.4163
R² = 0.9946
Catalyst 3 : rate ¼ k½3ꢃ½Ph3POꢃ½Me3SiCNꢃ
-3.5
y = 1.359x - 1.2373
R² = 0.9713
-4
To further investigate kinetic similarities and differences
between catalysts 1–3 a variable temperature kinetics study was
carried out to allow the activation parameters to be determined.
Figure 5 shows the Arhenius plot for all three catalysts and the
resulting activation enthalpies, entropies and Gibbs energies are
given in Table 2, along with those of related titanium and vanadium
-4.5
Figure 3. Determination of the order with respect to catalyst concentration for com-
plexes 1–3.
0.008
y = 0.1145x + 0.0003
R² = 0.99
catalyst 1
catalyst 2
catalyst 3
1000/RT
-9
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0
-100.25
-11
0.35
0.45
0.55
y = -17.7x - 3.9243
R² = 0.968
-12
y = 0.0797x - 8E-05
R² = 0.9566
y = -54.698x + 9.2864
R² = 0.9936
-13
-14
-15
-16
-17
-18
y = -0.0007x + 0.0003
R² = 1
catalyst 1
catalyst 2
catalyst 3
0
0.02
0.04
0.06
0.08
0.1
y = -34.774x + 0.5073
R² = 0.984
[R3PO]
Figure 4. Determination of the order with respect to phosphine oxide concentration
for complexes 1–3. kobs data for catalyst 2 have been multiplied by 10 to allow all the
data to be plotted on the same scales.
Figure 5. Arhenius plots for complexes 1–3. Catalyst 1: [1]¼0.022 M, [MePh2
-
PO]¼0.084 M. Catalyst 2: [2]¼0.0027 M, [MePh2PO]¼0.015 M. Catalyst 3: [3]¼0.005 M,
[Ph3PO]¼0.025 M.

M. North et al. / Tetrahedron 66 (2010) 1915–1924
1919
Table 2
The data for catalyst 3 show that it has a very low enthalpy of
activation compared to any of the other aluminium or salen based
catalysts. Its entropy of activation is however more negative than
any of the other salen based catalysts, except for complex 10, which
is known to be a very slow catalyst. This is consistent with reactions
involving complex 3 requiring the presence of triphenylphosphine
oxide as a Lewis base, whilst reactions catalysed by complexes 9–12
do not require the addition of a separate Lewis base. Thus, in the
case of reactions catalysed by complex 3 more components have to
be brought together in the transition state resulting in a more
negative entropy of activation, which largely offsets its very low
enthalpy of activation. The overall result is that complex 3 is a less
active catalyst than the two very highly active catalysts (9 and 12),
but more active than any of the other catalysts.
Activation parametersa for catalysts 1–3 and 9–12
b
Catalyst
D
Hz (kJ molꢀ1
)
D
S
z (J molꢀ1
)
DG )
z (kJ molꢀ1
1
2
3
9
10
11
12
34.8 (ꢂ1.1)
54.7 (ꢂ2.7)
17.7 (ꢂ1.3)
35.9 (ꢂ3.2)
27.6 (ꢂ4.9)
32.5 (ꢂ3.2)
20.4 (ꢂ2.9)
ꢀ162 (ꢂ5)
ꢀ61 (ꢂ9)
79.0 (ꢂ1.1)
71.3 (ꢂ2.7)
60.0 (ꢂ1.3)
59.4 (ꢂ3.2)
77.8 (ꢂ4.9)
66.6 (ꢂ3.2)
57.5 (ꢂ2.9)
ꢀ155 (ꢂ5)
ꢀ86 (ꢂ12)
ꢀ184 (ꢂ17)
ꢀ125 (ꢂ12)
ꢀ136 (ꢂ10)
a
Error limits are calculated on the basis of a ꢂ6% error in all of the rate data used
to construct Figure 5.
b
At 273 K.
based salen complexes 9–12 determined previously.48 The overall
trend in Gibbs energies of activations matches the observed cata-
lytic activities of the complexes, with complexes 9 and 12 being the
most active (reactions complete in 1–2 h at 0.1 mol % catalyst
loading) and complexes 1 and 2 being least active (reactions require
10 mol % catalyst for 6–36 h).
3. Mechanistic analysis
The catalytic cycle previously proposed by Shibasaki13 to account
for catalysis by complex (S)-1 is shown in Scheme 4. This mecha-
nism is consistent with the zero-order dependence on phosphine
oxide concentration and the first order dependence on catalyst
concentration determined by our kinetics results. However, benz-
aldehyde is involved in the catalytic cycle (step A) before trime-
thylsilyl cyanide is involved (step B). Since NMR experiments
showed no evidence for the formation of a significant concentration
of a species such as 14, this is inconsistent with the kinetic data,
which show the reaction to be first order in trimethylsilyl cyanide
concentration and zero order in benzaldehyde concentration.
To make the catalytic cycle shown in Scheme 4 consistent with
the kinetic data, all, that is, required is that steps A and B be
interchanged so that the in situ assembled catalyst 13 reacts first (in
the rate determining step) with trimethylsilyl cyanide and sub-
sequently with benzaldehyde.
Comparison of the data for catalysts 1 and 2 shows that they have
similar Gibbs energies of activation (79 and 71 kJ molꢀ1, re-
spectively), but that this is made up of very different contributions
from the enthalpies and entropies of activation. Catalyst 1 has a very
negative entropy of activation, suggesting that it has a very highly
ordered transition state. In contrast, catalyst 2 has the least negative
entropy of activation of any of the catalysts (ꢀ61 J molꢀ1 compared
to ꢀ162 J molꢀ1 for catalyst 1). The significantly lower entropy of
activation of complex 2 compared to complex 1 is consistent with
deoligomerization of the precatalyst offsetting the entropic cost of
bringing the various reaction components together. In contrast,
catalyst 2 has the highest enthalpy of activation of any of the cata-
lysts (55 kJ molꢀ1) and this is 20 kJ molꢀ1 greater than the enthalpy
of activation of catalyst 1. This suggests that the diethylamino
groups in catalyst 2 are not as effective Lewis bases as the phosphine
oxide groups in catalyst 1. However, it is known that under the
optimal synthetic conditions (involving addition of 4 Å molecular
sieves), the active cyanating agent in reactions catalysed by complex
2 is hydrogen cyanide rather than trimethylsilyl cyanide.18,19 Under
these conditions, the diethylamino groups would be expected to act
as more effective Brønsted bases, thus accounting for the enhanced
activity of catalyst 2 under the optimal conditions.
Ph2(O)P
Ph2(O)P
ClAl
HO
HO
O
O
Et2AlCl
Ph2(O)P
Ph2(O)P
R3PO
tBu
tBu
Ph
P
Ph
Ph
O
Ph
Ph
P
Me3Si
O
tBu tBu
O
O
NC
Cl Al
Cl
N
N
O
O
N
N
O
O
O
O
Al
Ti
Ti
R3P
O
O
R3P
O
O
O
O
P
Ph
tBu tBu
O
Ph
P
Ph
13
Ph
Ph
PhCHO
A
tBu
tBu
9
OSiMe3
NC
Ph
Ph
P
Ph
O
Ph
O
P
Ph
Me3Si
Ph
O
O
O
N
N
O
H
H
NC
O
O
O
O
V
X
Me3SiCN
B
Cl Al
R3P
Cl Al
tBu
O
tBu
R3P
O
O
tBu
tBu
O
O
10: X = EtOSO3
P
Ph
P
Ph
11: X = Cl
Ph
Ph
12:
X = NCS
14
Scheme 4. Catalytic cycle for complex 1.

1920
M. North et al. / Tetrahedron 66 (2010) 1915–1924
The catalytic cycle proposed by Najera to account for catalysis by
alyst, but with external phosphine oxide activating the cyanide50,51
(Scheme 6). This would account for the first order dependence of the
rate of reaction on phosphine oxide concentration observed with
complex 2, and provided the formation of complex 17 was rate de-
termining would also be consistent with the zero-order dependence
of the rate on the benzaldehyde concentration.
Stereochemically, the mechanism shown in Scheme 5 involves
the addition of cyanide to the si-face of the coordinated benzal-
dehyde within complex 16. However, the si-face of the aldehyde in
complex 16 (and the analogous complex 18 in Scheme 6) also
appears to be the less hindered face for intermolecular addition to
occur on. Therefore, addition of cyanide to complex 18 would also
be expected to occur on the si-face of the coordinated aldehyde,
leading to the same stereochemical outcome ((R)-cyanohydrin
from (S)-catalyst) as that predicted by Scheme 5. This analysis was
supported by results reported by Najera using the aluminium
complex 19 of binol.19 Under a range of conditions, (S)-19 cata-
lysed the asymmetric synthesis of (R)-mandelonitrile trime-
thylsilyl ether, derived by addition of cyanide to the si-face of
benzaldehyde.
complex (S)-2 is shown in Scheme 5. However, this catalytic cycle is
for reactions carried out in the presence of 4 Å molecular sieves to
generate hydrogen cyanide in situ. Therefore, our kinetic results will
not be directly applicable to this catalytic cycle. It seems likely, that in
the absence of a water source, the actual catalytic cycle involving
catalyst 2 will resemble that shown in Scheme 4 for Shibasaki’s cat-
Et2N
HO
Ph
HO
HCN +
OSiMe3
NC
Et2N
Me2AlCN /
R3PO
NHEt2
Ph
H
CN
O
Et2N
O
O
O
O
Cl
Al
O
Me3SiCN
Al
O
Cl
R3P
R3P
O
ClAl
NEt2
NEt2
15
O
PhCHO
NEt2
NHEt2
Me3SiCN
H2O
Ph
H
19
Ph
CN
O
O
H
O
O
HCN
O
O
Cl Al
The order of 0.5 observed for the dependence of the rate on
catalyst concentration for complex 2 is consistent with previous
results reported by Najera. Thus, whilst the asymmetric addition
of trimethylsilyl cyanide to aldehydes carried out in the presence
of a phosphine oxide co-catalyst was found to show a linear re-
lationship between the enantiomeric excess of the catalyst and
the enantiomeric excess of the product,19 the asymmetric addition
of cyanoformates23 or cyanophosphonates20,22 to aldehydes
which occur in the absence of a phosphine oxide co-catalyst show
a pronounced non-linear relationship.49 This indicates that com-
plex 2 exists in toluene solution in equilibrium with oligomeric
species, but that these oligomers are completely broken down to
mononuclear complex 15 in the presence of a phosphine oxide co-
catalyst. Thus, multiple catalytically active species 15 are obtained
from each initially oligomeric complex 2. The same effect also
explains the remarkably small negative entropy of activation ob-
served for complex 2 (Table 2), since the coming together of
multiple molecules required to assemble the catalytically active
species in Schemes 5 and 6 is offset by the dissociation of the
initially oligomeric complex.
Cl Al
O
O
R3P
R3P
NEt2
NEt2
16
Scheme 5. Catalytic cycle for complex 2 in the presence of 4 Å molecular sieves.
15 + R3PO + Me3SiCN
Et2N
O
O
Me3SiCN
OSiMe3
NC
Al
O
Cl
R3PO + 15
R3P
+
Ph
R3P
NEt2
OSiMe3
NC
17
The kinetic data for complex 3 support a mechanism such as
that shown in Scheme 7, which we have previously proposed for
asymmetric cyanohydrin synthesis catalysed by complex 3 in the
presence of a phosphine oxide co-catalyst.29 This involves the use
of triphenylphosphine oxide to activate the trimethylsilyl cya-
nide50,51 and complexation of the resulting species to one of the
aluminium ions of complex 3 in the rate determining step of
the catalytic cycle, thus explaining the first order dependence of the
rate of reaction on the concentrations of catalyst, phosphine oxide
and trimethylsilyl cyanide. Coordination of benzaldehyde to the
other aluminium ion occurs after the rate determining step, and
subsequent intramolecular transfer of cyanide to the si-face of the
coordinated aldehyde (for reactions involving catalyst 3 derived
from (S,S)-diaminocyclohexane) leads to the formation of (R)-cya-
nohydrin trimethylsilyl ethers after silylation of the initially formed
aluminium coordinated cyanohydrin.
PhCHO
OSiMe3
R3P
R3P OSiMe3
NEt2
NC
NEt2
Ph
H
CN
Ph
O
O
H
O
O
O
O
Cl
Al
Cl Al
O
O
R3P
R3P
NEt2
NEt2
18
Scheme 6. Catalytic cycle for complex 2 in the absence of 4 Å molecular sieves.

M. North et al. / Tetrahedron 66 (2010) 1915–1924
1921
OSiMe3
carried out at various concentrations of complex 20 (Fig. 8) and
methyldiphenylphosphine oxide (Fig. 9) also showed that the re-
action was first order in both catalyst components. Thus, the full
rate equation for reactions catalysed by complex 20 is:
R3PO + Me3SiCN
Me3SiCN
R3P
NC
3
rate ¼ k½20ꢃ½MePh2POꢃ½Me3SiCNꢃ
(6)
O
Al(salen)
(salen)Al
CN
3 + R3PO
C
This is the same rate equation as that determined for reactions
catalysed by complex 3 (Eq. 5) and is consistent with a mechanism
such as that shown in Scheme 8 in which the formation of complex
N
Me3SiO
Ph
R3P
OSiMe3
PhCHO
Time (minutes)
-1.5
(salen)Al
O
Al(salen)
(salen)Al
O
Al(salen)
10
20
30
O
O
C
0
N
-2
-2.5
-3
CN
R3P
Ph
H
Ph
y = -0.0678x - 1.6882
OSiMe3
²
R = 0.9935
R3P OSiMe3
Scheme 7. Catalytic cycle for complex 3.
-3.5
-4
4. Application to magnesium-based catalyst system
Figure 6. First order kinetics plots for catalyst 20 ([PhCHO]0¼0.17 M,
[Me3SiCN]0¼0.17 M, [20]¼3.46 mM, [MePh2PO]¼3.44 mM).
In 1993, Corey reported a unique magnesium-based catalyst
system for asymmetric cyanohydrin synthesis.52 Bisoxazoline–
magnesium complex 20 was used as a chiral Lewis acid along with
bisoxazoline 21 as a chiral Lewis base to catalyse the asymmetric
addition of trimethylsilyl cyanide to aldehydes. To demonstrate
that the mechanistic analysis developed in this work had applica-
bility beyond aluminium complexes, it was decided to carry out
a kinetic analysis of asymmetric cyanohydrin synthesis catalysed by
complex 20.
0.2
[Me3SiCN]=0.17 M
[Me3SiCN]=0.35 M
0.15
0.1
0.05
0
[Me3SiCN]=0.52 M
[Me3SiCN]=0.69 M
CN
0
5
10
15
20
25
30
O
O
O
O
time (minutes)
N
N
N
N
Figure 7. Demonstration that the rate of reactions catalysed by complex 20 depends
on the initial concentration of trimethylsilyl cyanide ([PhCHO]0¼0.17 M,
[20]¼3.46 mM, [MePh2PO]¼3.44 mM).
Mg
Ph
Ph
Ph
Ph
Cl
20
21
1.4
1.2
Initial studies showed that the system developed by Corey was
not amenable to kinetic analysis. Trial reactions were initially car-
ried out at ꢀ78 ꢁC under the optimal conditions reported by Corey
(20 mol % 20 and 12 mol % 21). Unfortunately, manual sampling of
the reaction mixture was not possible as reaction occurred during
the sampling process. Monitoring of the kinetics by 1H NMR
spectroscopy was also not possible due to the mixed solvent system
of dichloromethane and propionitrile necessary to dissolve cata-
lysts 20 and 21. Attempts to carry out kinetic analyses at a higher
temperature (0 ꢁC) also failed as the catalyst system decomposed
during the reaction under these conditions.
y = 4.1018x - 0.05
1
R2 = 0.9856
0.8
0.6
0.4
0.2
0
0
0.1
0.2
0.3
[cat] (mM)
Figure 8. Determination of the order with respect to catalyst 20 concentration
([PhCHO]0¼0.17 M, [Me3SiCN]0¼0.17 M, [MePh2PO]¼3.44 mM).
To avoid these problems, a new catalytic system was developed
in which magnesium complex 20 was used as the chiral Lewis acid
in conjunction with methyldiphenylphosphine oxide as an achiral
Lewis base. This combination of catalysts was soluble in dichloro-
methane, and at 25 ꢁC the combination of (S,S)-20 (2 mol %) and
methyldiphenylphosphine oxide (8 mol %) catalysed the asym-
metric addition of trimethylsilyl cyanide to benzaldehyde with
100% conversion after 16 h, to giving (S)-mandelonitrile trime-
thylsilyl ether with 35% enantiomeric excess. Reactions carried out
in deuterated dichloromethane could be monitored by 1H NMR
spectroscopy.
3
2.5
2
y = 0.1682x + 0.55
R2 = 0.9887
1.5
1
0.5
0
0
5
10
15
[Me2PhP=O] (mM)
This catalyst system was again found to show overall first order
kinetics (Fig. 6), and reactions carried out at four different trime-
thylsilyl cyanide concentrations confirmed that the reactions were
first order in trimethylsilyl cyanide concentration (Fig. 7). Reactions
Figure 9. Determination of the order with respect to phosphine oxide concentration
for reactions catalysed by complex 20 ([PhCHO]0¼0.17 M, [Me3SiCN]0¼0.17 M,
[20]¼3.46 mM).

1922
M. North et al. / Tetrahedron 66 (2010) 1915–1924
22 is rate determining, which is closely analogous to the catalytic
cycle for catalyst 3 (Scheme 7). The mechanism shown in Scheme 8
is different to that proposed by Corey52 for the 20/21 catalyst sys-
tem, but this reflects the fact that methyldiphenylphosphine oxide
is a Lewis base, which can activate trimethylsilyl cyanide,50,51
whilst bisoxazoline 21 could function as a Brønsted base to activate
hydrogen cyanide generated in situ from trimethylsilyl cyanide.
450GC using a Supelco Gamma DEX 120 fused silica capillary col-
umn (30 mꢄ0.25 mm) with hydrogen as a carrier gas (flow rate
2.0 mL/min, column pressure 10 psi). Initial temperature 95 ꢁC, fi-
nal temperature 180 ꢁC, ramp rate 5.0 ꢁC/min.
6.2. BINOLAM 654,55
A solution of s-BuLi (1.8 mL of a 1.3 M solution in hexanes) was
added dropwise over 15 min to a stirred solution of 2,20-bis(N,N-
diethylcarbamoyloxy)-1,10-binaphthyl39 4 (500 mg, 1.3 mmol) and
TMEDA (0.31 mL, 2.1 mmol) in dry THF (10 mL) at ꢀ78 ꢁC under
nitrogen. The reaction was kept at ꢀ78 ꢁC for one hour, then
allowed to warm to room temperature and stir overnight. The
mixture was quenched with NH4Cl and the product extracted with
CH2Cl2 (3ꢄ50 mL). The combined organic phases were washed
with brine, dried (MgSO4) and the solvent evaporated to leave bis-
amide 5, which was used without further purification.
To a solution of the above prepared compound 5 in THF (20 mL)
at 0 ꢁC was added LiAlH4 (116 mg, 3.0 mmol). The reaction was then
heated to reflux overnight, quenched with a saturated KF solution
(1 mL), filtered through CeliteÒ and the solvent evaporated. The
residue was extracted with CH2Cl2, washed with phosphate buffer
and brine and the organic layer dried (MgSO4) and concentrated to
leave a yellow oil, which was purified by column chromatography
OSiMe3
R3PO + Me3SiCN
R3P
NC
20
CN
O
O
Me3SiCN
20 + R3PO
Cl
Mg
C
N
N
Me3SiO
Ph
Ph
CN
N
Ph
R3P
OSiMe3
22
CN
Cl
O
O
N
N
(CHCl3/MeOH, 25:1) to give BINOLAM 6 as a pale yellow solid
PhCHO 20
20
(205 mg, 63% from 4). Mp 140–141 ꢁC (lit.54 138–139 ꢁC); [
a]
D
Mg
O
Ph
Ph
CN
þ147.6 (c 0.5, CHCl3) (lit.55 þ147 (c 0.5, CHCl3)); dH(CDCl3) 1.01 (12H,
t J 7.2 Hz, CH3), 2.59 (8H, q J 7.2 Hz, NCH2CH3), 3.85 (2H, d J 13.5 Hz,
ArCH2), 4.13 (2H, d, J 13.5 Hz, ArCH2), 7.1–7.3 (6H, m, ArH), 7.68 (2H,
s, ArH), 7.78 (2H, d J 7.8 Hz, ArH).
20
PhCHO
Ph
R3P OSiMe3
Scheme 8. Catalytic cycle for complex 20 and MePh2PO.
6.3. O-Trimethylsilyl mandelonitrile using catalyst 1
5. Conclusions
To a solution of the substituted binol ligand (13 mg, 0.0182 mmol)
in dry CH2Cl2 (1 mL) under nitrogen, was added Me2AlCl (18 mL,
Although the rate equations for catalysts 1–3 and 20 differ in
detail, they are consistent with a common mechanistic basis for
asymmetric cyanohydrin synthesis. This involves activation of the
aldehyde by a Lewis acid and activation of the trimethylsilyl cya-
nide by a Lewis base. When a proton source is present in the re-
action mixture, the trimethylsilyl cyanide may be hydrolysed to
hydrogen cyanide, which can be activated by a Brønsted base rather
than a Lewis base. This dual activation of both components of the
reaction can be achieved by a single catalytic entity comprising
both Lewis acidic and basic sites or by two different catalytic spe-
cies and appears to be a general feature of the most effective cat-
alysts for asymmetric cyanohydrin synthesis.1–3 When the Lewis
acid and base are in separate molecules, the Lewis acid is usually
the source of chirality within the catalytic assembly as it will be
directly coordinated to the prochiral aldehyde, though examples
are known where both the Lewis acid and the base are chiral spe-
cies.52 Since simultaneous catalysis by Lewis acids and Lewis bases
can be applied to reactions other than cyanohydrin synthesis,53 the
results of this study may also have wider generality.
0.018 mmol, 1.0 M solution in hexanes). The resulting mixture was
cooled to ꢀ40 ꢁC, then benzaldehyde (21 mg, 0.192 mmol) and
Me3SiCN (34 mg, 0.345 mmol) were added and the reaction stirred
at ꢀ40 ꢁC for 36 h. The solution was then filtered and evaporated.
A sample was analysed by 1H NMR spectroscopy to determine the
conversion. To the rest of the residue, acetonitrile (1 mL), a few drops
of Ac2O and a catalytic amount of Sc(OTf)3 were added and the
mixture stirred at room temperature for five minutes. The reaction
was filtered through SiO2 and analysed by chiral GC to determine the
enantioselectivity.
6.4. O-Trimethylsilyl mandelonitrile using catalyst 2
To a suspension of (R)-BINOLAM 6 (11.4 mg, 0.025 mmol), Ph3PO
(28 mg, 0.1 mmol) and 4 Å molecular sieves (62 mg) in dry toluene
(1 mL) under nitrogen, was added Me2AlCl (25 mL, 0.025 mmol,
1.0 M solution in hexanes). The resulting mixture was cooled to
ꢀ20 ꢁC, then benzaldehyde (27 mg, 0.25 mmol) and Me3SiCN
(100
m
L, 0.75 mmol) were added and the reaction stirred at ꢀ20 ꢁC
for six hours. The solution was then filtered and evaporated. A
sample was analysed by 1H NMR spectroscopy to determine the
conversion. To the rest of the residue, acetonitrile (1 mL), a few
drops of Ac2O and a catalytic amount of Sc(OTf)3 were added and
the mixture stirred at room temperature for five minutes. The
reaction was filtered through SiO2 and analysed by chiral GC to
determine the enantioselectivity.
6. Experimental
6.1. Instrumentation and general methods
CH2Cl2 was dried by distillation from CaH2. All deuterated sol-
vents were purchased from GOSS chemicals. Chromatographic
separations were performed with silica gel 60 (230–400 mesh). All
UV spectra were recorded on a Biochrom Libra S12 spectrometer
(100–240 V). 1H NMR spectra were recorded on Bruker Avance 300,
Jeol ECS 400 or Jeol Lambda 500 MHz spectrometers at the tem-
peratures specified. Chiral GC analysis was carried out on a Varian
6.5. O-Trimethylsilyl mandelonitrile using catalyst 3
Complex 3 (11 mg, 0.0095 mmol) and Ph3PO (13 mg, 0.047 mmol)
were dissolved in CH2Cl2 (1 mL). Benzaldehyde (51 mg, 0.48 mmol)

M. North et al. / Tetrahedron 66 (2010) 1915–1924
1923
was added and the solution cooled to ꢀ40 ꢁC. Me3SiCN (75 mg,
0.76 mmol) was then added and the solution stirred at ꢀ40 ꢁC for
16 h. The solution was then passed through a short silica plug eluting
with CH2Cl2. The eluent was evaporated in vacuo and the residue was
analysed by 1H NMR spectroscopy to determine the conversion. The
residue was then dissolved in acetonitrile (1 mL), a few drops of Ac2O
and a catalytic amount of Sc(OTf)3 were added and the mixture
stirred at room temperature for 20 min. The reaction was filtered
through SiO2 and analysed by chiral GC to determine the
enantioselectivity.
evaporated. A sample was analysed by 1H NMR spectroscopy to
determine the conversion. The residue was then dissolved in ace-
tonitrile (1 mL), a few drops of Ac2O and a catalytic amount of
Sc(OTf)3 were added and the mixture stirred at room temperature
for 10 min. The reaction was filtered through SiO2 and analysed by
chiral GC to determine the enantioselectivity.
6.10. Kinetics study using catalyst 20
To a solution of cyanobox ligand52 in CD2Cl2 (2 mL) was added
iPrMgCl (2.0 M solution in Et2O) and the resulting solution was
stirred at room temperature for one hour under a nitrogen atmo-
sphere. Then, MePh2PO was added, followed by benzaldehyde and
Me3SiCN. The solution was transferred to a NMR tube and a 1H NMR
spectrum was recorded every few minutes over a period of one
hour depending on the concentrations of the various components.
6.6. Kinetics study using catalyst 1
To a solution of the substituted binol ligand and methyl-
diphenylphosphine oxide in CD2Cl2 (0.75 mL), Me2AlCl (1 equiv
relative to the amount of ligand) was added under N2 and the
mixture stirred at room temperature for one hour. Then, benzal-
dehyde was added and the sample transferred to a NMR tube fol-
lowed by the addition of Me3SiCN. The reaction tube was
immediately cooled with liquid N2 and transferred to the NMR
spectrometer. A 1H NMR spectrum was then recorded at ꢀ40 ꢁC at
appropriate intervals (every few minutes) over a period of ca. two
hours depending on the concentrations of the various components.
The quantity of each reagent varied depending on the concentra-
tion required. To construct the Arhenius plot, the reaction tem-
perature was varied between ꢀ40 and þ20 ꢁC.
References and notes
1. North, M.; Usanov, D. L.; Young, C. Chem. Rev. 2008, 108, 5146–5226.
2. Riant, O.; Hannedouche, J. Org. Biomol. Chem. 2007, 5, 873–888.
3. Davie, E. A. C.; Mennen, S. M.; Xu, Y.; Miller, S. J. Chem. Rev. 2007, 107, 5759–5812.
ˇ
´
4. Malkov, A. V.; Kocovsky, P. Eur. J. Org. Chem. 2007, 29–36.
5. Purkarthofer, T.; Skranc, W.; Schuster, C.; Griengl, H. Appl. Microbiol. Biotechnol.
2007, 76, 309–320.
6. Blacker, A. J.; North, M.; Belokon, Y. N. Chemistry Today 2004, 22, 30–32 (Chiral
Catalysis: C–C Coupling and Oxidation Supplement).
7. Blacker, J.; North, M. Chem. Ind. 2005, 22–25.
8. Lapworth, A. J. Chem. Soc. 1903, 995–1005.
9. Lapworth, A. J. Chem. Soc. 1904, 1206–1214.
6.7. Kinetics study using catalyst 2
10. Lapworth, A.; Helmuth, R.; Manske, F. J. Chem. Soc. 1928, 2533–2549.
11. Lapworth, A.; Helmuth, R.; Manske, F. J. Chem. Soc. 1930, 1976–1981.
12. Hamashima, Y.; Sawada, D.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 1999, 121,
2641–2642.
13. Hamashima, Y.; Sawada, D.; Nogami, H.; Kanai, M.; Shibasaki, M. Tetrahedron
2001, 57, 805–814.
14. Sawada, D.; Shibasaki, M. Angew. Chem., Int. Ed. 2000, 39, 209–213.
15. Sawada, D.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2000, 122, 10521–10532.
16. Takamura, M.; Yanagisawa, H.; Kanai, M.; Shibasaki, M. Chem. Pharm. Bull. 2002,
50, 1118–1121.
17. Nogami, H.; Kanai, M.; Shibasaki, M. Chem. Pharm. Bull. 2003, 51, 702–709.
18. Casas, J.; Na´jera, C.; Sansano, J. M.; Saa´, J. M. Org. Lett. 2002, 4, 2589–2592.
To a solution of (R)-BINOLAM 6 and methyldiphenylphosphine
oxide in toluene-d8 (0.75 mL), was added Me2AlCl under N2 and the
mixture stirred at RT for one hour. Then, benzaldehyde was added
and the sample was transferred to a NMR tube followed by addition
of Me3SiCN. A 1H NMR spectrum was then recorded at appropriate
intervals (every few minutes) over a period of ca. four hours
depending on the concentrations of the various components. The
quantity of each reagent varied depending on the concentration
required. To construct the Arhenius plot, the reaction temperature
was varied between þ20 and þ55 ꢁC.
´
´
19. Casas, J.; Najera, C.; Sansano, J. M.; Saa, J. M. Tetrahedron 2004, 60, 10487–10496.
20. Baeza, A.; Casas, J.; Na´jera, C.; Sansano, J. M.; Saa´, J. M. Angew. Chem., Int. Ed.
2003, 42, 3143–3146.
21. Casas, J.; Baeza, A.; Sansano, J. M.; Na´jera, C.; Saa´, J. M. Tetrahedron: Asymmetry
2003, 14, 197–200.
6.8. Kinetics study using catalyst 3
22. Baeza, A.; Na´jera, C.; Sansano, J. M.; Saa´, J. M. Chem.dEur. J. 2005, 11, 3849–3862.
23. Baeza, A.; Casas, J.; Na´jera, C.; Sansano, J. M.; Saa´, J. M. Eur. J. Org. Chem. 2006,
1949–1958.
Catalyst 3 and Ph3PO were dissolved in freshly distilled CH2Cl2
(1.75 mL) and the solution cooled to 0 ꢁC in an ice bath. A sample
24. Baeza, A.; Casas, J.; Na´jera, C.; Saa´, J. M. J. Org. Chem. 2006, 71, 3837–3848.
´
25. Baeza, A.; Najera, C.; Sansano, J. M. Eur. J. Org. Chem. 2007, 1101–1112.
(0.50 mL) was removed and diluted into CH2Cl2 (3.0 mL) to be used
as the reference sample to zero the spectrophotometer. Benzalde-
hyde was added to the reaction mixture and another sample
26. Baeza, A.; Sansano, J. M.; Saa´, J. M.; Na´jera, C. Pure Appl. Chem. 2007, 79, 213–221.
27. Qin, Y.-C.; Liu, L.; Pu, L. Org. Lett. 2005, 7, 2381–2383.
28. Qin, Y.-C.; Liu, L.; Sabat, M.; Pu, L. Tetrahedron 2006, 62, 9335–9348.
29. North, M.; Williamson, C. Tetrahedron Lett. 2009, 50, 3249–3252.
30. Kim, S. S.; Song, D. H. Eur. J. Org. Chem. 2005, 1777–1780.
31. Kim, S. S. Pure Appl. Chem. 2006, 78, 977–983.
(0.50 mL) removed and also diluted into CH2Cl2 (3.0 ml). The ab-
sorbance of the sample was measured at 246 nm to provide the t¼0
reading. Me3SiCN was added to the reaction mixture and the ki-
32. Kim, S. S.; Kwak, J. M. Tetrahedron 2006, 62, 49–53.
33. Zeng, Z.; Zhao, G.; Zhou, Z.; Tang, C. Eur. J. Org. Chem. 2008, 1615–1618.
34. Chen, F.-X.; Zhou, H.; Liu, X.; Qin, B.; Feng, X.; Zhang, G.; Jiang, Y. Chem.dEur. J.
2004, 10, 4790–4797.
35. Alaaeddine, A.; Roisnel, T.; Thomas, C. M.; Carpentier, J.-F. Adv. Synth. Catal.
2008, 350, 731–740.
netics monitored by taking samples (0.50 mL) and quenching them
into CH2Cl2 (3.0 ml) at appropriate time intervals over a period of
ca. six hours depending on the concentrations of the various
components. The quantity of each reagent varied depending on the
concentration required. To construct an Arhenius plot, the reaction
temperature was varied between ꢀ32 andþ23 ꢁC.
36. Nicewicz, D. A.; Yates, C. M.; Johnson, J. S. Angew. Chem., Int. Ed. 2004, 43,
2652–2655.
37. Nicewicz, D. A.; Yates, C. M.; Johnson, J. S. J. Org. Chem. 2004, 69, 6548–6555.
38. Mele´ndez, J.; North, M.; Pasquale, R. Eur. J. Inorg. Chem. 2007, 3323–3326.
39. Dennis, M. R.; Woodward, S. J. Chem. Soc., Perkin Trans. 1 1998, 1081–1085.
40. Norsikian, S.; Holmes, I.; Lagasse, F.; Kagan, H. B. Tetrahedron Lett. 2002, 43,
5715–5717.
41. Belokon, Y. N.; Caveda-Cepas, S.; Green, B.; Ikonnikov, N. S.; Khrustalev, V. N.;
Larichev, V. S.; Moscalenko, M. A.; North, M.; Orizu, C.; Tararov, V. I.; Tasinazzo,
M.; Timofeeva, G. I.; Yashkina, L. V. J. Am. Chem. Soc. 1999, 121, 3968–3973.
42. Belokon, Y. N.; Green, B.; Ikonnikov, N. S.; Larichev, V. S.; Lokshin, B. V.;
Moscalenko, M. A.; North, M.; Orizu, C.; Peregudov, A. S.; Timofeeva, G. I. Eur. J.
Org. Chem. 2000, 2655–2661.
6.9. O-Trimethylsilyl mandelonitrile using catalyst 20
To a solution of cyanobox ligand52 (1 mg, 0.003 mmol) in CH2Cl2
(2 mL) was added iPrMgCl (1.5
m
L of a 2.0 M solution in Et2O,
0.003 mmol) and the resulting solution was stirred at room tem-
perature for one hour under nitrogen atmosphere. Then,
a
MePh2PO (2.6 mg, 0.012 mmol) was added, followed by benzalde-
hyde (0.015 mL, 0.15 mmol) and Me3SiCN (0.02 mL, 0.30 mmol).
The solution was then stirred for 16 h before being filtered and
43. Belokon, Y. N.; North, M.; Parsons, T. Org. Lett. 2000, 2, 1617–1619.
44. Belokon, Y. N.; Green, B.; Ikonnikov, N. S.; North, M.; Parsons, T.; Tararov, V. I.
Tetrahedron 2001, 57, 771–779.

1924
M. North et al. / Tetrahedron 66 (2010) 1915–1924
45. Belokon, Y. N.; Clegg, W.; Harrington, R. W.; Young, C.; North, M. Tetrahedron
2007, 63, 5287–5299.
49. Satyanarayana, T.; Abraham, S.; Kagan, H. B. Angew. Chem., Int. Ed. 2009, 48,
456–494.
46. Belokon, Y. N.; Maleev, V. I.; North, M.; Usanov, D. L. Chem. Commun. 2006,
4614–4616.
47. Belokon, Y. N.; Clegg, W.; Harrington, R. W.; North, M.; Young, C. Inorg. Chem.
2008, 47, 3801–3814.
50. Ryu, D. H.; Corey, E. J. J. Am. Chem. Soc. 2004, 126, 8106–8107.
51. Ryu, D. H.; Corey, E. J. J. Am. Chem. Soc. 2005, 127, 5384–5387.
52. Corey, E. J.; Wang, Z. Tetrahedron Lett. 1993, 34, 4001–4004.
53. Ma, J.-A.; Cahard, D. Angew. Chem., Int. Ed. 2004, 43, 4566–4583.
´
´
48. Belokon, Y. N.; Clegg, W.; Harrington, R. W.; Maleev, V. I.; North, M.; Omedes
Pujol, M.; Usanov, D. L.; Young, C. Chem.dEur. J. 2009, 15, 2148–2165.
54. Saa, J. M.; Tur, F.; Gonzalez, J.; Vega, M. Tetrahedron: Asymmetry 2006, 17, 99–106.
55. Kitajima, H.; Ito, K.; Katsuki, T. Tetrahedron 1997, 53, 17015–17028.