Approach to (R)- and (S)-ketone cyanohydrins using almond and apple meal as the source of (R)-oxynitrilase

Article abstract of DOI:10.1016/S0957-4166(97)00110-9

The synthesis of aliphatic and aromatic (R)-ketone cyanohydrins through the addition of hydrogen cyanide to the corresponding ketones and the synthesis of the (S)-enantiomers through the kinetic resolution of racemic ketone cyanohydrins has been studied in the presence of almond or apple meal. Substrate tolerance of the (R)-oxynitrilases towards ketones (R¹R²C=O) is highly restricted compared to that of structurally similar aldehydes, reactivity following the order of H>Me>>Et for R². In the case of aromatic methyl ketones reactivity difference (C₆H₅>>p-Me-C₆H₄ for R¹) is notable.

Full text of DOI:10.1016/S0957-4166(97)00110-9

Tetrahedron: As~mmetr3, Vol. 8. No. 10. pp 1551-1557. 1997
(~) 1997 Elsevier Science Ltd. All rights reserved.
Printed in Great Britaio
PP e r g a m o n
~
PII: S0957-4166(97)00110-9
0957-4166/97 $1700 + 0.00
Approach to (R)- and (S)-ketone cyanohydrins using almond and
apple meal as the source of (R)-oxynitrilase
Eero Kiljunen a and L i i s a T. K a n e r v a b'*
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
bDepartment of Medical Physics and Chemistry, University of Turku, PL 123, FIN-2052t Turku, Finland
Abstract:
The synthesis of aliphatic and aromatic (R)-ketone cyanohydrins through
the addition of hydrogen cyanide to the corresponding ketones and the synthesis of the
(S)-enantiomers through the kinetic resolution of racemic ketone cyanohydrins has been
studied in the presence of almond or apple meal. Substrate tolerance of the (R)-oxynitrilases
towards ketones (RIR2C=O) is highly restricted compared to that of structurally similar
aldehydes, reactivity following the order of H>Me>>Et for R2. In the case of aromatic
methyl ketones reactivity difference (C6H5>>p-Me-C6H 4 for R 1) is notable. (~) 1997
Elsevier Science Ltd
The synthesis of optically active aldehyde cyanohydrins by oxynitrilase-catalysed addition of
hydrogen cyanide to prochiral aldehydes has been the goal of extensive and successful studies during
recent years, t-16 The synthesis of optically active ketone cyanohydrins, on the other hand, has been
less studied. (R)-Oxynitrilases [E.C. 4.1.2.10] isolated from almonds (Prunus amygdalus) and flax
(Linum usitatissimum) were found to catalyse the enantioselective addition of hydrogen cyanide to
methyl or ethyl alkyl ketones, yielding the corresponding (R)-ketone cyanohydrins.16-18 In the case
of the flax enzyme, however, enantiopurities and absolute configurations remain mostly unresolved.16
The almond enzyme was also useful for the preparation of some aliphatic (S)-ketone cyanohydrins
through kinetic resolution. 19 (S)-Oxynitrilase [E.C. 4.1.2.37] isolated from Manihot esculenta and
overexpressed in E. coli shows high potential (ee of the order of 90% or lower) for the preparation
of aliphatic and even aromatic (S)-ketone cyanohydrins. 13 Unfortunately, the enzyme preparation is
not commercially available. On the other hand, the commercial (S)-oxynitrilase [E.C. 4.1.2.11 ] from
Sorghum bicolor does not accept ketones as substrates.7
Relaying on the (R)-selective almond enzyme, we have now exploited almond meal for the
preparation of (R)-ketone cyanohydrins 2a-h through the addition of hydrogen cyanide with the
corresponding prochiral ketone (Scheme 1A) and (S)-ketone cyanohydrins 2a-i through the kinetic
resolution of the corresponding racemic cyanohydrin with or without the presence of aldehyde 3
(Scheme IB). As previously shown, apple meal (another rich source of (R)-oxynitrilase) is more
favourable than almond meal for sterically hindered aldehydes (e.g., pivalaldehyde) as substartes.8
Accordingly, some apple meal-catalysed additions were also studied in this work.
Enzymatic kinetic resolution in the presence of hydrolytic enzymes is the most commonly used
biocatalytic way to optically active alcohols.2° The method was previously carried out in the case
of aldehyde cyanohydrins by lipases.21"22 Unfortunately, lipases rarely accept tertiary alcohols as
substrates. 2° To our knowledge, the microbial hydrolysis of ketone cyanohydrin acetates on incubation
with grown cells of Pichia miso IAM 4682 is the only method reported for the preparation of the
less reactive (S)-ketone cyanohydrin acetates.23 This further underlines the importance of oxynitrilase
methods for the preparation of the enantiomers of ketone cyanohydrins.
* Corresponding author. Email: lkanerva@utufi
1551

1552
E. KILJUNEN and L. T. KANERVA
O
Almond or
OH
Apple meal ,.
RI ----~..,,u Re
[-A'] R J J ~ R2
HCN
CN
1
Diisopropyl ether
(R)-2
a
R1 =Bu, R2 = Me
R l = i B u , R2 =Me
d
e
R I = n-Hexyl, Re = Et
R l = P h , R2 =Me
g
h
R t = Ph, R2 = Et
b
R I = P h , R2 = C H C I2
c Rl=n-Pentyl,Re=Me
f
Rl=p-Me-Ph,Rz =Me
i
Ri=iPentyl,R2 =Me
Almond or
Apple meal
40
O:
R'N R2
O___~
R1
R2 + R3CHO
,
Re, ..... RI
CN
+
+
R3
H
CN
CN
Diisopropyl ether
Buffer
O
2
3
(S)-2
I
4
Scheme I.
Results and discussion
The synthesis of (R)-cyanohydrins 2a-h in the presence of almond or apple meal was studied in
diisopropyl ether containing tartrate buffer [2% (v/v), 0. l M, pH 5.4] at 5°C (Scheme 1A, Table 1). In
the synthesis, the corresponding ketone reacts with hydrogen cyanide which freely evaporates from a
solution in diisopropyl ether in one compartment of the reaction vessel and ends up to the other where
it dissolves into the reaction mixture (previously called method 2).7,8 Under such conditions, chemical
condensation to racemic cyanohydrins is effectively suppressed except in the case of dichloromethyl
ketone lh where the electron deficiency of the carbonyl carbon favours nonenzymatic condensation.
Aliphatic methyl ketones la-c are transformed to the corresponding (R)-cyanohydrins with high
enantiopurity (ee 96-99%). Compared to the almond meal-catalysed synthesis of stericaily similar
aldehyde cyanohydrins reactivity now is considerably lower. Thus, the almond meal-catalysed synthesis
of (R)-2-hydroxyheptanenitrile from hexanal (corresponding to lc with R2=H) proceeds to 72%
conversion within 42 hours (ee 97%) compared to 18 days needed for 76% conversion in the case of
methyl ketone lc (R2=Me) as a substrate (Table 1; row 5).8 Fortunately, reactivity can be considerably
increased by increasing the enzyme content (Table 1; row 2 compared to row 1). Reactivity turns
even worse when aromatic methyl ketones le and f serve as substrates although enantioselectivity still
remains considerably high. Moreover, the methyl substituent in the benzene ring significantly affects
reactivity. As a comparison, the corresponding aldehydes le and f (R2=H) under the otherwise same
conditions have practically reacted within 1 and 2 days, respectively.7 Ethyl ketones ld and g are
almost unreactive under the reaction conditions, indicating that the ability of a carbonyl compound to
act as a substrate for the almond enzyme decreases with increasing size of the group R2. Clearly, apple
meal as a catalyst does not improve reactivity and/or enantioselectivity for the synthesis of ketone
cyanohydrins. This is against the previous results for the formation of aldehyde cyanohydrins where
apple meal was found to accept even sterically hindered aldehydes.8
In synthetic chemistry, the oxynitrilase-catalysed formation of optically active cyanohydrins is a
general goal. 1-3 In nature, however, oxynitrilases typically catalyse the decomposition of cyanohydrins,
mandelonitrile (Scheme 1B, 4; R3=C6H5) being the natural substrate of the almond enzyme. The
ability of the almond enzyme to catalyse the decomposition of acetone cyanohydrin is also clearly
demonstrated.~8,15 In the present work, the reversible nature of oxynitrilase catalysis has been studied
for the kinetic resolution of racemic cyanohydrins 2a and e in the presence of almond meal in

(R)- and (S)-Ketone cyanohydnns
1553
Table 1. Synthesis of(R)-2a-h in diisopropyl ethe~ in the presence of almond or apple me~ (150 mg/mmolsubstrate)at 5°C
Substrate
Source ofenzyme Time/h Conversion/% ee/%
la
la
lb
lb
lc
ld
le
le
le
If
Almond
Almondb
Almond
Apple
168
24
60
73
45
40
76
2
96
99
97
99
98
76
95
90
93
80
20
20
0
192
192
432
336
120
312
168
504
144
96
Almond
Almond
Almond
Almond
Apple
5
14
13
2
Almond
Almond
Apple
lg
lg
lh
lh
<1
1
Almond
Apple
144
96
67
69
5
~Contains tartrate buffer [2% (v/v),0.1 M, pH 5.4]. b500 mg/mmol substrate.
diisopropyl ether containing tartrate buffer [7% (v/v), 0.1 M, pH 5.4] at 25°C (Scheme 1B in the absence
of 3). The (R)-enantiomer then selectively decomposes to ketones la and e leaving the (S)-enantiomer
of 2 unreacted. Unfortunately, the ee values of 37% obtained for (S)-2a and 88% obtained for (S)-
2e (conversions 53 and 57%, respectively) are disappointingly low. As an explanation, the ketone
1 enzymatically reacts back to the (R)-cyanohydrin 2, making the preparation of the corresponding
highly enantiopure (S)-cyanohydrin impossible. Thus, e.g., for the resolution of 2a the drop of ee is
from 46 to 37% at the conversion range t¥om 35 to 53% although the enantiopurity of the less reactive
(S)-counterpart is expected to increase with conversion. Clearly, there is a need to trap the hydrogen
cyanide delivered during the resolution process.
For decades, synthetic chemists have exploited an interchange between a ketone cyanohydrin and an
aldehyde for the preparation of lower water-soluble cyanohydrins in the presence of a base.24 Under
the conditions where the chemical interchange is negligible compared to the corresponding almond
(or apple) meal-catalysed reaction, the hydrogen cyanide delivered by the reactive (R)-enantiomer of
2 is accepted by aldehyde 3 (Scheme IB). As a further demand, the formed aldehyde cyanohydrin 4
must be stable enough under the conditions where the (R)-cyanohydrin 2 is turned to ketone 1, leaving
(S)-2 unreacted. For that purpose, various aldehydes as hydrogen cyanide acceptors were studied for
the almond meal-catalysed resolution of racemic 2a in diisopropyl ether in the presence of tartrate
buffer [7% (v/v), 0.1 M, pH 5.4] at 25°C (Table 2). Compared to lower aliphatic aldehydes 3 (R3 from
Me to n-pentyl) benzaldehyde is not a good acceptor as can be seen from a low ee (43%) for the less
reactive (S)-2a. In the presence of almond meal, benzaldehyde yields (R)-mandelonitrile (the natural
substrate for the enzyme) which in the reversible enzymatic reaction produces hydrogen cyanide as
another decomposition product.25 Thus, hydrogen cyanide is not totally bound to benzaldehyde. As
a result, there is a complicated equilibrium between the various components of the reaction mixture,
leading to the racemization of the enantiomericaily enriched (S)-2a with time. The unexpectedly low
enantiopurity of (R)-4 (Table 2), on the other hand, can be explained by a chemical interchange between
ketone and aldehyde cyanohydrins (Table 3). Acetaldehyde or propionaldehyde (3; R3 Me or Et) seem

1554
E. KILJUNEN and L. T. KANERVA
Table 2. The almond meal-catalysed resolution of racemic 2a in the presence of aldehyde 3 in diisopropyl ethera at 25°C
Aldehyde
R~
(S)-2a
ee/%
43
(R)-4
Time/h Conversion/%
ee/%
Ph
312
144
165
165
165
51
53
51
50
50
Pent
Pr
96
92
85
44
96
Et
99
Me
97
"Contains 0.1 M tartrate buffer [7% (v/v), pH 5.4].
Table 3. The chemical interchange of racemic 2a, e and f with acetaldehyde in diisopropyl ethera at 25°C
Table 3. The chemical interchange of racemic 2a, e and f
with acetaldehyde in diisopropyl ethed at 25 °C.
Substrate
Time/h
144
Conversion/%
2a
2e
2f
3
23
9
144
144
aContains tartrate buffer [7% (v/v), 0.1 M, pH 5.4].
to be the most advantageous acceptors of hydrogen cyanide because the obtained cyanohydrins 4 then
can be extracted into water or distilled at low temperature.
According to the results of Table 4 for the kinetic resolution of racemic cyanohydrins 2a-i, lower
aliphatic methyl ketones 2a, b and i and acetophenone 2e can be resolved by almond (or apple)
meal with short-chain aliphatic aldehydes as hydrogen cyanide acceptors. The method indicates the
importance of trapping the hydrogen cyanide with an aldehyde. Thus, the enantiomeric excesses of
2a and e can be increased from ca. 40 to >95% and from 88 to 98%, respectively, using acetaldehyde.
Branching in R1 somewhat enhances reactivity (Table 4; rows 1 and 3 or 4 and 13). Ethyl ketones 2d
and g as well as dichloromethyl ketone 2h lead to racemic products, indicating again that substrates
with R2>Me are poorly accepted by (R)-oxynitrilase. In accordance with the data in Table 1, there is
practically no effect on enantioselectivity when almond meal as a catalyst is replaced by apple meal.
As a conclusion, the enantiomers of ketone cyanohydrins can be prepared using aliphatic and
aromatic methyl ketones as substrates for almond and apple meal in diisopropyl ether in the presence
of minor amounts of tartrate buffer (0.1 M, pH=5.4). The (R)-enantiomers are obtained introducing
gaseous hydrogen cyanide into the reaction mixture (method 2 in the connection of aldehyde
cyanohydrinsT,8). The method to the (S)-enantiomers is based on the enzymatic interchange between
the (R)-enantiomer of the racemic ketone cyanohydrin and an aldehyde (corresponds to method 1 in the
case of aldehyde cyanohydrins5-7). In this method, the (S)-enantiomer is properly enriched only if the
enzymatic decomposition of another interchange product, (R)-aldehyde cyanohydrin, can be suppressed
under the reaction conditions. Substrate tolerance of the oxynitrilase enzymes is considerably restricted
compared to that observed in the case of sterically similar aldehyde cyanohydrins.
Experimental
Chemicals
Defatted almond meal was purchased from Sigma. Apple meal was prepared from apple seeds as
previously described. 8 Diisopropyl ether and commercial aldehydes and ketones were obtained from

(R)- and (S)-Ketone cyanohydrins
1555
Table 4. The almond or apple meal-catalysed resolution of racemic 2a-i in the presence of aldehyde 3 in diisopropyl ethera
at 25°C
Aldehyde
R3
(S)-2
eel%
97
99
98
67
2
Substrate
2a
Time/h
165
48
Conversionb/%
Me
Pr
50
50
49
53
1
2b
2b
Me
Pr
94
2e
672
432
72
2d
Me
Me
Et
2e
58
60
57
32
90
25
35
40
98
59
39
73
1
2f
336
408
408
144
168
144
432
2f
Me
Me
Me
Et
2f
2g
2h
0
2h¢
2i
Me
Pr
5
96
~Contains tartrate buffer [7% (v/v), 0.1 M, pH 5.4]. bUnreacted 2 in the reaction
mixture. CUsing apple meal as a catalyst.
Aldrich or Merck and were distilled before use. Neat hydrogen cyanide was prepared by adding
potassium cyanide solution dropwise into dilute sulphuric acid7'8 and leading the hydrogen cyanide
formed into dry diisopropyl ether. The etheral solution was stored in a freezer. Racemic ketone
cyanohydrins were prepared from the corresponding ketones via a reaction with trimethylsilylcyanide
and ZnI2 followed by desilylation with 3 M HCI. 26
IH (TMS as an internal standard) and 13C NMR spectra were measured on a Lambda GX 400
spectrometer in CDC13. MS spectra were recorded on a VG Analytical 7070E instrument equipped with
a VAXstation 3100 M76 system. The enantiomeric excess of the aldehyde and ketone cyanohydrins
were determined after acylation with acetic, propionic or hexanoic acid anhydride by the chiral GLC
method using Chrompack CP-Cyclodextrin-13-2,3,6-M-9 column.7,8 Optical rotations were measured
using a JASCO DIP-360 polarimeter.
Method
Synthetic method
In a typical experiment, almond meal (150 mg) was mixed with diisopropyi ether (16 ml) in the
compartment A of the two-compartment reaction vessel. Tartrate buffer (330 lal, 0.1 M, pH 5.4) and
a freshly distilled ketone (1.0 mmol) were added. The reaction mixture was stirred at 5°C. The other
compartment B contained diisopropyl ether (10 ml) and hydrogen cyanide (10 mmol) at 5°C. The
reaction was followed by taking samples at intervals and using the chiral GLC method.7'8 The reactions
for the formation of (R)-cyanohydrins 2a--c were stopped by filtring off the enzyme. The solvent was
evaporated under reduced pressure. In the other cases the products were not separated from the reaction
mixture. The spectroscopic data were obtained from chemically prepared racemic cyanohydrins.

1556
E. KILJUNEN and L. T. KANERVA
Resolution methods
In a typical experiment, the enzymc prcparation (150 mg) was mixed with diisopropyl ether (16
ml) and tartrate buffer (1.2 ml, 0.1 M, pH 5.4). One of the racemic ketone cyanohydrins 2a-i (1.5
mmol), and in the case of the interchange reactions freshly distilled aldehyde 3 ( 1.0 mmol) were added.
The reactions were followed as described above. In the case of racemic cyanohydrins 2a and 2e the
interchange reactions were performed in a gram-scale in the presence of acetaldehyde 3 (R3=Me).
The reactions were stopped by filtering off the enzyme preparation. The unreacted aldehyde and the
corresponding cyanohydrin 4 (R3=Me) were washed out. The organic layer was dried over Na2SO4
and the solvent was evaporated under reduced pressure. The isolated yields were calculated taking the
conversion to correspond 100% theoretical yield.
(R)- 2-Hydroxy-2-methylhexanenitrile C7H13NO 2a
The product corresponding to 82% conversion; [o¢]~5s=+5.1 (CHCI3, c=5.2).
(S)-2-Hydroxy-2-methylhexanenitrile C7H13NO 2a
Yield 45%; [~x]25=-7.4 (CHCI3, c=6.6). 1H NMR (CDC13): 8 0.95 (t, 3H, CH3); 1.20-1.30 (m,
2H, CH2); 1.30-1.50 (m, 2H, CH2); 1.56 (s, 3H, CH3) 1.75-1.85 (m, 2H, CH2); 13C NMR (CDCI3):
8 14.19 and 22.46 (CH3); 26.35, 27.73 and 41.50 (CH2); 68.76 (C); 122.01 (CN). MS: m/z 126 (M- 1,
0.1), 112 (3), 100 (5), 94 (1), 85 (5), 84 (6), 71 (100), 68 (10), 58 (20), 57 (38), 43 (24) and 41 (12).
(R)-2-Hydro,o,-2,4-dimethylpentanenitrile C7HI3NO 2b
The product corresponding to 45% conversion; loci(is=+4.4 (CHCI3, c=3.5). IH NMR (CDC13): 8
1.09 (d, 6H, CH3); 1.65 (s, 3H, CH3) 1.70 (d, 2H, CH2); 1.96 (m, 1H, CH); 13C NMR (CDCi3): 8
15.28, 23.63 and 23.68 (CH3), 28.89 (CH2), 49.80 (CH), 68.07 (C), 122.01 (CN). MS: m/z 127 (M,
0.1), 108 (2), 100 (5), 94 (9), 85 (6), 82 (5), 71(100), 57 (45), 43 (95), 41 (53) and 39 (24).
(R)-2-Hydro©'-2-methylheptanenitrile CsHI5NO 2c
The product corresponding to 76% conversion; [o¢]~5s=+6.0 (CHCI3, c=5.1); [tx]~5 (1it17)=+2.7
(CHC13, c=1.12). IH NMR (CDCI3): 8 0.92 (overl., t 3H, CH3); 1.20-1.55 (m, 6H, CH2); 1.61 (s,
3H, CH3); 1.72-1.80 (m, 2H, CH2); 13C NMR (CDC13): 8 13.92 and 22.42 (CH3); 23.93, 27.70,
31.45 and 41.73 (CH2); 68.77 (C); 122.04 (CN). MS: rn/z 140 ( M - I , 0.2), 126 (0.5), 114 (2), 108
(1), 99 (2), 95 (1), 85 (1), 71 (68), 68 (10), 58 (22), 55 (10), 43 (100) and 41 (20).
2-Hydro©'-2-ethyl-2-octanenitrile C1oHIgNO 2d
1H NMR (CDCI3): 8 0.90 (t 3H, CH3); 1.11 (t, 3H, CH3); 1.29-1.40 (m, 6H, CH2); 1.47-1.55
(m, 2H, CH2); !.67-1.88 (m, 4H, CH2); 13C NMR (CDC13): 8 8.23 and 14.02 (CH3); 22.52, 23.89,
29.07, 31.57, 33.26 and 39.71 (CH2); 72.92 (C); 121.26 (CN). MS: m/z 169 (M, 0.2), 154 (6), 141
(6), 113 (19), 97 (28), 85 (100), 72 (14), 57 (18) and 43 (32).
(S)-2-Hydroxy-2-methyl-2-phenylacetonitrile CgH9NO 2e
Yield 40%; [tx]25= - 1.8 (CHCI3, c=9.8). IH NMR (CDCI3): 8 1.90 (s, 3H, CH3); 2.98 (s, IH, OH);
7.21-7.23 (m, 2H, Harom); 7.45-7.48 (m, 2H, Harom); 13C NMR (CDCI3): ~i 31.08 (CH3); 70.76 (C);
121.42 (CN); 2"124.44, 2"129.01,129.26 and 139.22 (Carom). MS: m/z 147 (M, 35), 132 (100), 120
(12), 105 (40), 77 (40), 54 (12), 51 (20), 43 (16) and 39 (8).
2-Hydroxy-2-methyl-2-( 4'-methylphenyl)acetonitrile CIoH11NO 2f
IH NMR (CDC13): 8 1.87 (s, 3H, CH3); 2.37 (s, 3H, CH3); 3.07 (s, 1H, OH); 7.39-7.46 (m, 2H,
Harom); 7.58-7.61 (m, 2H, Harom); 13C NMR (CDCi3): /5 21.09 and 30.92 (CH3); 70.61 (C); 121.59
(CN); 2"121.59, 2"129.60, 137.83 and 139.22 (Carom). MS: m/z 161 (M, 25), 146 (100), 134 (25),
119 (70), 91 (65), 65 (24), 51 (12), 43 (25), 39 (20) and 32 (12).

(R)- and (S)-Ketone cyanohydrins
2-Hydroxy-2-ethyl-2-phenylacetonitrile C1oH11NO 2g
1557
IH NMR (CDCI3): 8 1.04 (t, 3H, CH3); 1.98-2.18 (qq, 2H, CH2); 2.90 (s, IH, OH); 7.37-7.45
(m, 3H, Harom); 7.55-7.58 (m, 2H, Harom); 13C NMR (CDCI3): 8 8.79 and 36.89 (CH3); 75.48 (C);
120.71 (CN); 2"125.07, 2"128.96, 129.29 and 139.78 (Carom). MS: rn/z 161 (M, 20), 132 (100), 105
(40), 77 (30), 51 (20) and 39 (6).
2-Hydroxy'-2-(dichloromethyl)-2-phenylacetonitrile C9H7CI2NO 2h
IH NMR (CDCI3): 8 3.94 (s, H, OH); 5.79 (s, IH, CH); 7.45-7.50 (m, 3H, Harom); 7.66-7.68 (m,
2H, Harom); 13C NMR (CDCI3): 8 22.86 (CH); 76.01 (C); 126.41 (CN); 128.95, 130.67, 133.79 and
138.50 (Carom). MS: m/z 219 (M+I, 0. l), 217 (M-I, 0.5), 215 (M-3, 1.0), 163 (0.5), 144 (1), 132
(45), 125 (5), 105 (100), 77 (50), 51 (26) and 39 (5).
2-Hydroxy- 2,5-dimethyl hexanenitrile CsHI5NO 2i
IH NMR (CDCI3): 8 0.95 (d, 6H, CH3); 1.10-1.90 (m, 7H); 1.61 (s, 3H, CH3); 13C NMR (CDCI3):
8 22.42 (2"CH3); 27.72, 27.95, 33.05, 39.72 (CH2) 68.86 (C), 121.96 (CN). MS: m/z 140 ( M - l ,
0.3), 126 (3), 108 (10), 99 (12), 83 (5), 81 (14), 71 (100), 58 (34), 56 (54), 43 (88) and 41 (6).
Acknowledgements
We thank the Technology Development Centre (TEKES) for financial support.
References
1. Kruse, C.G. in Chirality in lndusto', Collins, A.N.; Crosby, J. Eds. John Wiley & Sons Ltd, 1992,
ch. 14, p. 279.
2. Effenberger, F. Angew. Chem. Int. Ed. Engl. 1994, 33, 1555.
3. Kanerva, L.T. Acta Chem. Scand. 1996, 50, 234.
4. Zandbergen, P.; van der Linden, J.; Brussee, J.; van der Gen, A. Synth. Commun. 1991, 21, 1387.
5. Huuhtanen, T.T.; Kanerva, L.T. Tetrahedron: Asymmetry 1992, 3, 1223.
6. Kiljunen, E.; Kanerva, L.T. Tetrahedron: Asymmetry 1994, 5, 311.
7. Kiljunen, E.; Kanerva, L.T. Tetrahedron: Asymmetry 1996, 7, 1105.
8. Kiljunen, E.; Kanerva, L.T. Tetrahedron: Asymmetry 1997, 8, in press.
9. Klempier, N.; Griengl, H.; Hayn, M. Tetrahedron Lett. 1993, 34, 4769.
10. Klempier, N.; Pichler, U.; Griengl, H. Tetrahedron: Asymmetry 1995, 6, 845.
11. Schmidt, M.; Herv6, S.; Klempier, N.; Griengl, H. Tetrahedron 1996, 52, 7833.
12. Effenberger, E; Eichhorn, J.; Roos, J. Tetrahedron: AsymmetO' 1995, 6, 271.
13. F6rster, S.; Roos, J.; Effenberger, F.; Wajant, H.; Sprauer, A. Angew. Chem. Int. Ed. Engl. 1996,
35, 437.
14. Danieli, B.; Barra, C.; Carrea, G.; Riva, S. Tetrahedron: Asymmetry 1996, 7, 1675.
15. Ognyanov, V.I.; Datcheva, V.K.; Kyler, K.S.J. Am. Chem. Soc. 1991, 113, 6992.
16. Albrecht, J.; Jansen, I.; Kula, M.-R. Biotechnol. Appl. Biochem. 1993, 17, 191.
17. Effenberger, E; H6rsch, B.; Weingart, E; Ziegler, T.; Kfihner, S. Tetrahedron Left. 1991, 32, 2605.
18. Effenberger, E; Heid, S. Tetrahedron: Asymmeto, 1995, 6, 2945.
19. MenEndez, E.; Brieva, R.; Rebolledo, E; Gotor, V. J. Chem. Soc., Chem. Commun. 1995, 989.
20. Kanerva, L.T. in Enz3,matic Reactions in Organic Media, Koskinen, A.M.P.; Klibanov, A.M., Eds.,
Chapman & Hall, London, 1996, ch. 7, 170.
21. Kanerva, L.T.; Kiljunen, E.; Huuhtanen, T.T. Tetrahedron: Asymmetry 1993, 4, 2355.
22. Kanerva, L.T.; Rahiala, K.; Sundholm, O. Biocatalysis 1994, 10, 169.
23. Ohta, H.; Kimura, Y.; Sugano, Y.; Sugai, T. Tetrahedron 1989, 45, 5469.
24. Mowry, D.T. Chem. Rev. 1948, 42, 189.
25.30% of (R)-mandelonitrile (1 mmoi) decomposes to hydrogen cyanide and benzaldehyde within
24 hours in diisopropyl ether containing tartrate buffer [7% (v/v), 0.1 M, pH 5.4] and almond meal
(150 mg); only 4% of butyraldehyde is obtained from (R)-butanenitrile under the same conditions.
26. Gassman, P.G.; Talley, J.J. Tetrahedron Lett. 1978, 40, 3773.
(Received in UK 24 February 1997)