The Strecker reaction: Kinetic and equilibrium studies of cyanide addition to iminium ions

Article abstract of DOI:10.1039/b407853e

Kinetics studies are reported of the reactions of benzylidene benzylamine, 4a, and of benzylidene allylamine, 4b, with cyanide in aqueous buffers to give the corresponding a-aminonitriles. The results allow the calculation of values of rate and equilibrium constants for reaction of the iminium ions formed from 4a and 4b with cyanide ions. These values are compared with those, obtained from the hydrolysis reactions, for reaction of the iminium ions with hydroxide ions and with water. Comparison with some other iminium ions reveals that those formed from 4a and 4b are relatively unreactive due to the possibilities of charge delocalisation.

Full text of DOI:10.1039/b407853e

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
The Strecker reaction: kinetic and equilibrium studies of cyanide
addition to iminium ions†
John H. Atherton,a John Blacker,a Michael R. Crampton*b and Christophe Grosjeanb
a
Avecia Ltd., P.O. Box 521, Leeds Road, Huddersfield, UK HD2 1GA
Chemistry Department, Durham University, Durham, UK DH1 3LE
b
Received 26th May 2004, Accepted 23rd July 2004
First published as an Advance Article on the web 24th August 2004
Kinetics studies are reported of the reactions of benzylidene benzylamine, 4a, and of benzylidene allylamine, 4b,
with cyanide in aqueous buffers to give the corresponding a-aminonitriles. The results allow the calculation of
values of rate and equilibrium constants for reaction of the iminium ions formed from 4a and 4b with cyanide ions.
These values are compared with those, obtained from the hydrolysis reactions, for reaction of the iminium ions with
hydroxide ions and with water. Comparison with some other iminium ions reveals that those formed from 4a and 4b
are relatively unreactive due to the possibilities of charge delocalisation.
Similarly, high reactivity has been observed for related aliphatic
Introduction
iminium ions9 and for the N-nitrosoiminium ion, 3, where kH O
2
Reactions of the type shown in eqn. (1) are widely used in
for reaction with water10 was found to be 7 × 105 s−1.
synthesis.
RNH2 + RCHO + HNu → RNHCHRNu + H2O (1)
The Strecker reaction involves the use of cyanide as the
nucleophile, Nu, and leads to a-aminonitriles which may be
hydrolysed to give a-aminoacids.1
There is good evidence that in the Strecker reaction cyanide
and amine initially compete for the carbonyl compound with
cyanohydrin formation being favoured. The formation of
Here we report kinetic measurements for the reactions with
cyanide of benzylidene benzylamine, 4a, and benzylidene
allylamine, 4b, to give the corresponding a-aminonitriles, 6a
or 6b, respectively, as shown in Scheme 1. Our results allow the
calculation of the values of rate and equilibrium constants for
reaction of the corresponding iminium ions with cyanide ions
in water. In Scheme 1, kCN is the rate constant for reaction of
iminium ions with cyanide ions, and k−CN is the rate constant
for dissociation of 6. Measurement of the hydrolysis reactions
of the imines allows comparison of these values with those for
reaction with hydroxide ions or water.
a-aminonitrile follows more slowly and results from cyanide
attack on an iminium ion intermediate.2–5
Although iminium ions are known to be reactive inter-
mediates in many chemical and biochemical processes, including
the hydrolysis of imines,6,7 there is relatively little quantitative
information available for their reactions with nucleophiles in
water. Eldin and Jencks8 investigated the hydrolysis of some
ring-substituted anilinothioesters, 1, in the presence of added
thiolate ions. They assumed that the reverse reaction, k−1, was
diffusion controlled and obtained values of kH O for reaction
2
of the iminium ions, 2, with water in the range 106–108 s−1
depending on the nature of R.
† Electronic supplementary information (ESI) available: derivation of
eqn. (8). See http://www.rsc.org/suppdata/ob/b4/b407853e/
Scheme 1
T h i s j o u r n a l i s
©
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 6 7 – 2 5 7 1
2 5 6 7

View Article Online
Table 1 Variation of kf with pH for 4a
Results and discussion
The imines 4a and 4b were prepared by reaction of
benzaldehyde with the appropriate amine in acetonitrile.
They show strong absorption in the UV region: kmax 250 nm,
 = 1.9 × 104 dm3 mol−1 cm−1. Reaction of the imines with
trimethylsilyl cyanide in damp acetonitrile yielded the corres-
ponding a-aminonitriles 6a and 6b. These showed no strong
absorption above 220 nm. Kinetic studies of the reactions
of the imines, 1 × 10−4 mol dm−3, with potassium cyanide,
0.001–0.020 mol dm−3, were made in aqueous buffers in the pH
range 6–10. The changes with time of the UV absorbance at
250 nm were used to monitor the reactions. Two processes were
observed: the faster, kfast, resulted in a decrease in absorbance,
while the slower, kslow, gave increases in absorbance. Specimen
traces measured at pH 9.01 are given in Fig. 1(a) and 1(b). The
fast and slow processes accurately followed first-order kinetics,
allowing the calculation of rate constants.
pH
kf/dm3 mol−1 s−1
kf(calc)a/dm3 mol−1 s−1
5.94
2.8
4.7
6.6
4.9
3.5
1.1
0.50
2.8
4.8
6.8
5.7
4.1
1.1
0.40
6.40
7.28
8.55
9.01
9.85
10.40
a Calculated from eqn. (7) with kCN = 6.7 × 103 dm3 mol−1 s−1
,
+
pKaHCN = 9.1 and pKaImH = 6.14.
the total, stoichiometric, concentration of a compound in all
its ionisation states. The rate constants for proton transfers
involved in the equilibria given in eqns. (3) and (4) are likely to
be rapid.
[H+ ][CN− ]
KaHCN
=
(3)
(4)
[HCN]
[Im][H+ ]
[ImH+ ]
+
KaImH
=
Hence the concentrations of iminium ions and cyanide ions
may be related to the stoichiometric concentrations leading to
eqn. (5).
[H+ ]
KaHCN










kfast = kCN [CN− ]stoich

+ k



+
(5)
ImH+
a
HCN 
Hyd
+








 [H ] +K
+ [H ]


K

a
At a given pH value this simplifies to eqn. (6), where kf is given
by eqn. (7).
kfast = kf[CN−]stoich + kHyd
(6)
(7)
[H+ ]
KaHCN













+
kf = kCN

ImH+
a
HCN 
+








 [H ] +K
+ [H ]


K

a
Values of kf were obtained from the slopes of linear plots
according to eqn. (6) of kfast versus the stoichiometric cyanide
concentration. The variation with pH is given in Table 1 for 4a
and in Fig. 2 for 4b.
Fig. 1 Absorbance versus time plots for the reaction of 4a with
potassium cyanide at pH 9.01. Concentrations of KCN are: 1, 0; 2,
0.001 mol dm−3; 3, 0.002 mol dm−3; 4, 0.004 mol dm−3; 5, 0.008 mol dm−3
;
6, 0.016 mol dm−3; and 7, 0.020 mol dm−3
.
The results are interpreted in terms of Scheme 1. The faster
process results in fractionation of the parent between amino-
nitrile formation and hydrolysis to form benzaldehyde, 7. Com-
pound 7 also absorbs at 250 nm,  = 1.5 × 104 dm3 mol−1 cm−1,
although less strongly than 4a or 4b. The trace in Fig. 1(a) with
no added cyanide shows that hydrolysis occurs at a measurable
rate under the conditions used. Our results in Fig. 1(a) indicate
that the increases in values of (Absorbance at completion) with
decreasing cyanide concentration are due to benzaldehyde
formation rather than to the reversibility of aminonitrile forma-
tion, i.e. the reverse process involving k−CN is negligible here.
Hence eqn. (2) may be written,
Fig. 2 Plot of log10 kf versus pH for the reaction of 4b with cyanide
in water. The full line is calculated according to eqn. (7) with
ImH+
velocity = kfast[Imine]stoich
kCN = 1.04 × 104 dm3 mol−1 s−1 and pKa
= 6.05.
= kCN[Imine H+][CN−] + kHyd[Imine]stoich
(2)
where kHyd is the overall rate constant for hydrolysis of the
imine in both its forms, 4 and 5, at a given pH. Here and else-
where in this paper the subscript ‘stoich’ is used to represent
The pKHa CN value for hydrogen cyanide is known11 to
be 9.1, so that the two unknowns in eqn. (7) are kCN and
+
KaImH . Experimental values for kf for 4a give a good fit with
2 5 6 8
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 6 7 – 2 5 7 1

View Article Online
Table 2 Variation of kslow with cyanide concentration for 4a in water,
pH 10.4 at 25 °C
[KCN]stoich
kslow/10−4 s−1
kslow(calc)a/10−4 s−1
0.002
0.004
0.008
0.016
0.020
5.4
4.7
3.2
2.1
1.8
6.1
4.7
3.3
2.0
1.7
a Values calculated from eqn. (8) with k−CN = 8.5 × 10−4 s−1, and with
kHydr = 2.5 × 10−3 s−1 and kf = 0.5 dm3 mol−1 s−1 which are the appropriate
values at this pH.
values calculated using kCN = (6.7 ± 1) × 103 dm3 mol−1 s−1 and
+
pKaImH = 6.14 ± 0.1. For 4b the values were kCN = (1.0 ± 0.2) ×
ImH+
104 dm3 mol−1 s−1 and pKa
= 6.05 ± 0.1. The variation in the
Fig. 3 pH profile for the hydrolysis of 4b. The solid line is calculated
nature of R in the remote side-chain from benzyl to allyl has
little effect on the values obtained.
from eqn. (9) with kH O = 0.18 s−1, kOH = 4.4 × 105 dm3 mol−1 s−1
,
2
+
pKaImH = 6.05, k−H O/kd = 188 dm3 mol−1 and k−OH/k−H O = 2.4 ×
2
2
The plateau between pH 6.5 and 8.5 in the plot shown in
Fig. 2 corresponds to the region between the pKa values of the
iminium ion, 5, and hydrogen cyanide. In this region, increas-
ing acidity increases the concentration of 5 but decreases the
concentration of free cyanide ions. Decreases in value of kf at
pH > 9 result from decreases in the concentration of 5 when the
free cyanide concentration is reaching the stoichiometric con-
centration. Correspondingly, decreases in kf at lower pH result
from decreases in the concentration of free cyanide ions when
the imine is largely protonated. Hence, the results are in accord
with the major reaction in the pH region 6–10 involving the
attack of cyanide ions on iminium ions.
The slower process shown in Fig. 1(b) results in the decom-
position to benzaldehyde. This was confirmed by isolation of
the dinitrophenylhydrazone derivative of 7. It should be noted
that, in the presence of cyanide, benzaldehyde is in equilibrium
with its cyanohydrin derivative11 which does not absorb at
250 nm. The decrease in final absorbance values in Fig. 1(b) with
increasing cyanide concentration reflect this equilibrium. Since
the equilibration of benzaldehyde and its cyanohydrin is rapid11
compared to formation of benzaldehyde, this process does not
affect the kinetics of decomposition. Treating the iminium ions
as steady-state intermediates leads to eqn. (8). The derivation is
given as ESI.†
10−8 mol dm−3
.
represented by the processes shown in Scheme 2. This leads7 to
eqn. (9),
kH O + kOH[OH− ]
2
kobs
=
(9)
ImH+
a
+






k
O[H+ ] + k
K


−OH 



−H2



1+
1+



[H ] 
kd





where the rate constants are defined by the processes shown
in the scheme. For 4a, values obtained are: kH O = 0.15 s−1, kOH
=
2
ImH+
2.0 × 105 dm3 mol−1 s−1, pKa
= 6.14, k−H O/kd = 95 dm3 mol−1
2
and k−OH/k−H O = 1.3 × 10−8 mol dm−3. Values for 4b are given
2
in Fig. 3.
kHydr
kslow = k−CN
(8)
kHydr + kf [CN− ]stoich
The values of kslow at pH 10.4 for 4a, in Table 2, lead to
a value for k−CN of 8.5 × 10−4 s−1. Values (not shown) at pH
9.01 yield a value of 9.3 × 10−4 s−1. An alternative method
for measuring k−CN directly was by following spectrophoto-
metrically, at 250 nm, the decomposition of 6a in buffer
solutions in the absence of cyanide. Values were found to be
independent of pH in the range 6.8–10.4 and give k−CN = (9.1 ±
0.5) × 10−4 s−1. Related measurements for 4b yield a value for
k−CN of (5.5 ± 0.5) × 10−4 s−1.
The hydrolysis reaction
Kinetic studies, using the change in absorbance at 250 nm,
were made of the hydrolyses of the imines 4a and 4b, 1 ×
10−4 mol dm−3, to give benzaldehyde and the corresponding
amine. Measurements were made in the pH range 1–12 using
dilute buffer solutions or solutions of hydrochloric acid,
although values in the pH range 3–6 were too fast to be deter-
mined. The pH profiles for 4a and 4b were similar and that for
4b is shown in Fig. 3. We did not investigate general acid/base
catalysis although effects are likely to be small at the low buffer
concentrations used.7
Scheme 2
The independence of kobs on pH in the range 8–12
corresponds to rate-limiting attack of hydroxide on the iminium
ion, 5. The increases in value in the pH range 8–6 represent the
increasing importance of reaction of 5 with water. In the pH
range 1–4, decomposition of the carbinolamine intermediate 8
is rate limiting, the decreases with increasing acidity representing
decreases in concentration of the reactive zwitterionic form.
Following the work of Jencks and co-workers6,7,12 the
mechanism of imine hydrolysis is well understood and may be
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 6 7 – 2 5 7 1
2 5 6 9

View Article Online
Table 3 Summary of rate and equilibrium constants
Cyanohydrin formation
It is known11 that in the presence of cyanide the hydrolysis pro-
duct benzaldehyde is in equilibrium with its cyanohydrin deriva-
tive. The extent of conversion was monitored using absorbance
measurements at 250 nm, where benzaldehyde absorbs strongly
but the cyanohydrin shows negligible absorbance. Owing to the
acid–base processes shown in Scheme 3, values of Kobs, defined
in eqn. (10), were found to be pH-dependent. The acidity
dependence is given by eqn. (11) where equilibrium constants
are defined by reference to the scheme.
4a
4b
kCN/dm3 mol−1 s−1
k−CN/s−1
6.7 × 103
9.1 × 10−4
7.4 × 106
8.1 × 103
2.0 × 105
0.15
1.0 × 104
5.5 × 10−4
1.8 × 107
1.6 × 104
4.4 × 105
0.18
KCN/dm3 mol−1
KHCN/dm3 mol−1
kOH/dm3 mol−1 s−1
kH O/s−1
2
+
pKaImH
6.14
6.05
has been estimated14 to be in excess of 1011 s−1, although this
value may be considerably reduced by electron releasing groups
in the aromatic ring or in the side-chain.15,16
The ‘low’ reactivity of 5 may be understood in terms of the
possibilities for charge delocalisation. As shown in Scheme 4,
5 may be regarded as carbocations stabilised both by amino
groups and by a benzene ring. Similar stabilisation has been
reported17 in iminium ions generated from some aromatic
enamines. The presence of the strongly electron withdrawing
nitroso group at the nitrogen atom of 3 greatly enhances its
reactivity relative to 5.
[Cyanohydrin]stoich
Kobs
=
(10)
(11)
[Benzaldehyde][HCN]stoich
+
COH
a





[H ] + K



Kobs = K1
HCN 
+


[H ] + K

a
Measurements at pH 6.9 where Kobs ≈ K1 gave a value for K1 of
300 ± 10 dm3 mol−1. Values of Kobs at pH 10.2 and 11.6 were
27 ± 3 and 7.6 ± 1 respectively. Using the known value of 9.1
for pKaHCN these values lead via eqn. (11) to a value of pKaCOH
of 10.75 ± 0.1. Using eqn. (12) the value for K2 is found to be
7 ± 0.5 dm3 mol−1.
KaCOH
KaHCN
K2 = K1
(12)
These values are close to those reported previously.11
Scheme 4
Experimental
The imines 4a and 4b were prepared by reaction of equimolar
amounts of benzaldehyde and the appropriate amine in
acetonitrilefollowedbyevaporationof thesolventunderreduced
pressure. 4a was also available as a commercial specimen. 4a dH
(400 MHz, CD3CN): 4.78 (2H, s, CH2), 7.3–7.8 (10H, m, ArH),
8.47 (1H, s, CH); m/z (EI) 195. 4b dH (400 MHz, CD3CN): 4.19
(2H, m, CH2N), 5.14 (2H, dd, CH2), 6.05 (1H, m, CH), 7.4–7.7
(5H, m, ArH), 8.31 (1H, s, CHN); (M + H)+ m/z (CI) 146.
The a-aminonitriles 6a and 6b were prepared in solution by
reaction of the appropriate imine with trimethylsilyl cyanide
(1 equiv.) in damp acetonitrile. 6a dH (CD3CN): 2.5 (1H, m, NH),
3.84 (2H, m, CH2), 4.80 (1H, d, J = 9.6 Hz, CH), 7.2–7.5 (10H,
m, ArH). Coupling was observed between NH and hydrogens on
adjacent carbon atoms. 6b dH (C6D6): 3.05 (2H, d, CH2N), 4.18
(1H, d, J = 9.2 Hz, CHCN), 4.98 (2H, dd, CH2CH), 5.57 (1H, m,
CHCH2), 7.0–7.3 (5H, m, ArH); (M − H)+ m/z (EI) 171.
Other materials and solvents were the purest available
commercial specimens. The pH values of aqueous buffers
(phosphate, borax or bicarbonate, ca. 0.05 mol dm−3) were
measured using a Jenway 3020 pH meter.
UV-vis spectra and kinetic measurements were made at
25 °C with a Perkin Elmer Lambda 2 spectrophotometer or a
Shimadzu UV-2101 PC spectrophotometer. The substrate was
added to the reaction mixtures as a concentrated solution in
acetonitrile so that the final solvent composition was 99/1 (v/v)
water–acetonitrile. Rate constants were measured under first-
order conditions and are precise to ±5%. Benzaldehyde was
characterised as a product of hydrolysis, in more concentrated
solutions, as the 2,4-dinitrophenylhydrazone derivative, m.p.
238 °C, lit,18 237 °C.
Scheme 3
Comparisons
Values of rate and equilibrium constants are collected in Table 3.
They show that cyanide addition to the iminium ions 5 is an
extremely favourable process with values of KCN (=kCN/k−CN
)
being ca. 107 dm3 mol−1. Changing the group R from Ph to
CHCH2 has little effect on values of rate and equilibrium
constants. The use of eqn. (13)
KaHCN
KHCN = KCN
(13)
ImH+
a
K
allows the calculation of values of ca. 104 dm3 mol−1 for KHCN
the equilibrium constant for addition of hydrogen cyanide to the
imines 4. Knowledge of these values is important in the context
of the Strecker synthesis of aminonitriles.
Our results show that reactivity for the iminium ions decreases
in the order OH− > CN−  H2O. This order of nucleophilicity
reactivity is that commonly found for attack at cationic
centres.13
,
Safety Cyanide is poisonous and must be treated with
caution.
Nevertheless, the iminium ions 5 are relatively unreactive
References
when compared with some other iminium ions. Thus values for
1 D. Enders and J. P. Shilvock, Chem. Soc. Rev., 2000, 29, 359.
2 J. Taillades and A. Commeyras, Tetrahedron, 1974, 30, 2493.
3 M. Bejaud, L. Mion and A. Commeyras, Bull. Soc. Chim. Fr., 1976,
1425.
kH O for reaction with water for 2, 106–108 s−1, and 3, 7 × 105 s−1,
2
are many orders of magnitude higher than those for 5. Also, the
value of kH O for reaction of the benzyl carbocation with water
2
2 5 7 0
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 6 7 – 2 5 7 1

View Article Online
4 A. Commeyras, J. Taillades, L. Mion and M. Bejaud, Inf. Chim.,
1976, 158, 199.
12 W. P. Jencks, Prog. Phys. Org. Chem., 1964, 2, 63.
13 C. D. Ritchie, Acc. Chem. Res., 1972, 5, 348.
14 T. L. Amyes and J. P. Richard, J. Am. Chem. Soc., 1990, 112, 9507.
15 J. P. Richard, T. L. Amyes and M. M. Toteva, Acc. Chem. Res., 2001,
34, 981.
16 J. P. Richard, T. L. Amyes, V. Jagannadham, Y.-G. Lee and D. J.
Rice, J. Am. Chem. Soc., 1995, 117, 5198.
17 P. Y. Sollenberger and R. B. Martin, J. Am. Chem. Soc., 1970, 92,
4261.
5 G. Moutou, J. Taillades, S. Benefice-Madouet, A. Commeyras,
G. Messina and R. Mansani, J. Phys. Org. Chem., 1995, 8, 721.
6 E. H. Cordes and W. P. Jencks, J. Am. Chem. Soc., 1962, 84, 832.
7 E. H. Cordes and W. P. Jencks, J. Am. Chem. Soc., 1963, 85, 2843.
8 S. Eldin and W. P. Jencks, J. Am. Chem. Soc., 1995, 117, 4851.
9 S. Eldin, J. A. Digits, S.-T. Huang and W. P. Jencks, J. Am. Chem.
Soc., 1995, 117, 6631.
10 A. Vigroux, A. J. Kresge and J. C. Fishbein, J. Am. Chem. Soc.,
1995, 117, 4433.
11 W.-M. Ching and R. G. Kallen, J. Am. Chem. Soc., 1978, 100, 6119.
18 A. I. Vogel, Textbook of Practical Organic Chemistry, 5th edn.,
Brian S. Furniss, Antony J. Hannaford, Peter W.G. Smith and
Austin R. Tatchell, ed., Longman, London, 1989.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2 , 2 5 6 7 – 2 5 7 1
2 5 7 1