On the mechanism of the Kabachnik-Fields reaction: Does a mechanism of nucleophilic amination of α-hydroxyphosphonates exist?

Full text of DOI:10.1134/S0012500808060037

ISSN 0012-5008, Doklady Chemistry, 2008, Vol. 420, Part 2, pp. 137–140. © Pleiades Publishing, Ltd., 2008.
Original Russian Text © E.D. Matveeva, N.S. Zefirov, 2008, published in Doklady Akademii Nauk, 2008, Vol. 420, No. 4, pp. 492–495.
CHEMISTRY
On the Mechanism of the Kabachnik–Fields Reaction:
Does a Mechanism of Nucleophilic Amination
of a-Hydroxyphosphonates Exist?
E. D. Matveeva and Academician N. S. Zefirov
Received December 25, 2007
DOI: 10.1134/S0012500808060037
The Kabachnik–Fields reaction, which is a three-
There is a huge literature on this reaction, which
component interaction of a carbonyl compound, an considers both preparative aspects [3, 4] and mecha-
amine, and a compound containing a P–H bond nism issues [4–8]. It is obvious that this process is mul-
(
Scheme 1), is the most important method for the syn- tistage and can involve different intermediates. Overall
thesis of α-aminophosphonates [1–4].
data on the possible reaction routes are shown in
Scheme 2. The complexity of the process also results
from the fact that different catalysts are used in this
reaction [3], while the initial components (an amine
as a base and a hydrophosphoryl compound as an
acid) can produce a wide variety of intermediate com-
plexes [4–11].
Scheme 1.
R1
R4O
O + H P O
R1 OR4
P O
R –NH +
R3 NH
3
2
R2
R O
5
R OR
2 5
Scheme 2.
A
B
R1
R1
R1
OH
R1
NHR3
+ R –NH2
3
+ R –NH
3
2
O
NR3
–
H O
2
R2
NHR3
2
R2
NHR3
R2
R2
1
3
6
R4O
R4O
+ H
R5O
R4O
F
C
G
H
P O
D
P
O
H
P
O
R4O
R5O
R5O
H
P
O
R5O
R O R
R O R
4
2
1
4
2
1
E
+
R3–NH2
H O
O P
OH
O P
NHR3
–
2
R O R
5
R O R
5
5
4
Nonetheless, there are two main reaction mecha-
According to one of the suggested mechanisms, at
nisms depending on what reagent (amine or hydrophos- the first stage, an amine is added to the carbonyl group
phoryl compound) reacts in the first stage with the car- to form corresponding imine 3 (route A→B→C,
bonyl compound (Scheme 2, routes A or D). Thus, we Scheme 2). At the second stage, a hydrophosphoryl
deal with a competitive addition of either N–H or P–H compound containing the P–H bond is added to the
fragments to the C=O bond; this dual reactivity was C=N bond of imine 3 (the Pudovik reaction [8, 10],
considered in terms of hard (N–H) and soft (P–H) bases route B, Scheme 2) to give α-aminophosphonate 4.
[
12].
One possible complication in the Kabachnik–Fields
reaction is also obvious: hydrophosphoryl compounds
can undergo addition to the C=O bond to produce α-
Moscow State University, Leninskie gory, Moscow, 119992
Russia
1
37

1
38
MATVEEVA, ZEFIROV
hydroxyphosphonate 5 (route D, theAbramov reaction) phosphite, and racemic 2-aminobutane yields the corre-
4]. The formation of compound 5 or the products of its sponding α-hydroxyphosphonate and a mixture of
[
subsequent transformations is the main side reaction.
Nonetheless, taking into account the reversibility of the
Abramov reaction [4–6], α-aminophosphonates 4 can
be obtained from α-hydroxyphosphonates 5 via route
D
5
diastereomeric α-aminophosphonates (in a 1.19 : 1
molar ratio). The Pudovik reaction of the preliminarily
prepared corresponding imine with the same amine
gives rise to a diastereomeric mixture of the same α-
aminophosphonates in a 1.52 : 1 ratio. Moreover, the
refluxing of α-hydroxyphosphonate with the same
amine for a long time also produces a mixture of dias-
A
B
C and, thus, α-hydroxyphosphonate
may be considered as a temporary depot for the car-
bonyl compound.
However, an alternative mechanism is also consid- tereomeric α-aminophosphonates in a 1 : 1.59 ratio.
ered in the literature. This mechanism implies the pos-
sibility of obtaining compound 4 by the direct nucleo-
philic substitution of the hydroxy group in α-hydroxy-
phosphonate 5 by the amine component (nucleophilic
amination [4–8] of α-hydroxyphosphonates, route E,
Scheme 2). This mechanism was first suggested by
Kabachnik and Medved’ [1] by analogy with the Rodi-
onov and Strecker reactions. At the same time, Fields
The difference in diastereomeric composition is
explained by the authors [5, 6] as the existence of two
mechanisms of the Kabachnik–Fields reaction.
However, the observed difference in diastereomeric
composition can be explained by many factors such as
the difference in reaction conditions (heating for 2 or
8
h), the isomeric composition of the preliminarily
[
2] assumed the imine mechanism A
demonstrated the formal existence of route F (2
and route G (6
B
C and
obtained and intermediate imine, the possibility of
routes F and G, etc.
4)
4) by analogy with the Mannich
reaction. Moreover, he showed for one example the
impossibility of transformation 5 4 (route E). Thus,
the existence of the two above radically different mech-
anisms was postulated from the very beginning.
The most rigorous argument for the mechanism of
nucleophilic amination, route E, was obtained in [7]
in the study of the Kabachnik–Fields reaction for the
dialkyl phosphite–cyclohexylamine–benzaldehyde
Later, arguments for the mechanism of direct system. The authors found that the reaction pro-
nucleophilic substitution 5
4 (route E) were ceeded smoothly as a three-component process, but
obtained on the basis of kinetic [4–7] and stereochemi- specially obtained N-benzylidene-N-cyclohexyl-
cal evidence [8] and from the preparative study of the amine (3, R = Ph, R = H, R = cyclo-C H ) did not
1
2
3
6
11
mechanism of the Kabachnik–Fields reaction [7]. Let
us consider these arguments in more detail.
react with dialkyl phosphites even under catalytic
reaction conditions. Thus, there is classic proof for
the absence of a particular mechanism (in this case,
The kinetics of the reaction in some cases is total
third-order (first-order in each component) [5, 6].
imine route A
B
C, Scheme 2) because a spe-
Although the reaction often leads to the formation of a cially obtained would-be intermediate (imine 3 in
new chromophore (imine 3), the reaction with cyclo- this case) fails to produce a target product (α-amino-
hexylamine is accompanied by reduction in the optical phosphonate 4 in this case). 1
density in the region of the carbonyl chromophore,
Inasmuch as the nucleophilic amination mechanism
which was considered by the authors as the lack of a
new chromophore, imine 3 [7]. It should be noted that seems to be unlikely, we studied once more this reac-
the large excess of amine used in this work prevents the tion with participation of cyclohexylamine and benzal-
addition of hydrophosphoryl compound to imine (route dehyde. First of all, we carried out the Kabachnik–
C, 3
mation of a strong imine–amine complex [5] (in (benzene, formic acid (1 mol %), 20°C) [4, 7] and cat-
essence, aminal 6). Hence, it follows that route G
4) does not exist.
4). This was explained by the intermediate for- Fields reaction under both classic conditions [4, 6]
alytic conditions proposed by us in [3, 13, 14] (CH Cl ,
ambient temperature, 10 mol % of PcAlCl, 4 Å molec-
2
2
(6
t
We consider these arguments to be invalid. The ular sieves). In both cases, we obtained corresponding
absence of imine 3 can be explained by a particular α-aminophosphonate 9 (yields 90 and 98%, respec-
ratio of the rate constants of reactions 1
4 or even by the assumption of route F (2
Furthermore, it should be especially emphasized that
the rate constant of direct nucleophilic substitution
3 and tively) (Scheme 3). We found that specially obtained
3
4). α-hydroxyphosphonate 10 did not form amination
products when reacted with cyclohexylamine under
these conditions.
(route E, 5
4) was not determined in [5–7].
1
Note that, unfortunately, N-benzylidene-N-cyclohexylamine was
Stereochemical arguments are as follows [8]: the
Kabachnik–Fields reaction of benzaldehyde, dimethyl
not considered in the detailed study of the Pudovik reaction
[9−11].
DOKLADY CHEMISTRY Vol. 420 Part 2 2008

ON THE MECHANISM OF THE KABACHNIK–FIELDS REACTION
139
Scheme 3.
C H –CH=O + NH
C H –CH=N–C H
6 5 6
6
5
2
11
7
8
11
O
H
O
H
O
H
(C2H5O)2P
(C2H5O)2P
(C2H5O)2P
O
O
C H CH P(OC H )
2 5 2
C H CH P(OC H )
2 5 2
6
5
6
5
NC H
OH
6
11
1
0
9
Furthermore, we studied the reactions of corre- nance stabilization of an intermediate carbocationic
sponding imine 11. It was observed in [7] that the reac- center (R = alkoxy, dialkylamino groups) or, better,
tion of benzaldehyde with cyclohexylamine gave imine both factors should act together.
1
1, which was insoluble in the solvents used (benzene,
However, for R = R –NH–, i.e., when dehydration is
1
propanol-2) and precipitated (mp 138–142°C). It is
most important that the imine did not react with dialkyl
phosphite even under catalytic reaction conditions.
possible in principle resulting in the R'–N=C double
bond, the process shown in Scheme 4 can proceed via
the elimination–addition mechanism. The latter is most
These data, however, contradict the available data likely to be general for the reactions of this type.
since imine 11 was described as a liquid readily soluble
in organic solvents [15]. We have established that
freshly distilled benzaldehyde 7 smoothly reacts
with cyclohexylamine 8 to give imine 11 as an oil
distillable in vacuum. The reaction of imine 11 with
diethyl phosphite under both standard (benzene, for-
mic acid (1 mol %), 20°C, 12 h) and catalytic condi-
EXPERIMENTAL
N-Benzylidene-N-Cyclohexylamine (11)
(a) A mixture of 5 g (40 mmol) of freshly distilled
benzaldehyde and 4 g (40 mmol) of freshly distilled
cyclohexylamine was stirred in the presence of formic
acid (1 mol %) in 10 mL of anhydrous benzene for 5 h
under an argon atmosphere. The completion of the
reaction was monitored by TLC. The reaction mixture
t
tions (5 h, 20°C, 10 mol % of PcAlCl, benzene) pro-
duces α-aminophosphonate 9 (yields 90 and 95%,
respectively).
It is worth noting that some samples of benzalde- was distilled in vacuum to give 7 g (93%) of N-ben-
hyde react with cyclohexylamine to form an insoluble zylidene-N-cyclohexylamine 11, bp 172–174°C
precipitate with mp 138–142°C. The same product is (40 mmHg). Lit.: bp 250°C (760 mmHg), 174–175°C
1
readily obtained in the reaction of benzoic acid with
cyclohexylamine to give cyclohexylamine benzoate
(
(
41 mmHg) [15]. H NMR (CDCl , δ, ppm): 1.23–1.45
3
m, 3H, ring), 1.57–1.78 (m, 5H, ring), 1.84–1.88 (m,
H, ring), 3.18–3.25 (m, 1H, ring), 7.40–7.42 (m, 3H,
phenyl), 7.74–7.76 (m, 2H, phenyl), 8.33 (s, 1H,
CH=N).
(
mp 140–142°C). Thus, the experimental data that con-
2
firmed the mechanism of direct nucleophilic substitu-
tion of the hydroxy group and showed the absence of
reaction of imine 11 with phosphites are an artifact and
the mechanism of nucleophilic amination (route E), in
our opinion, has no experimental justification.
(
b) The reaction with benzaldehyde that was not
freshly distilled results under the same conditions in
certain samples in the formation of a precipitate with
mp 138–142°C. H NMR (CDCl , δ, ppm): 0.95–1.99
In summary, it should be noted that the issue under
consideration on the mechanism of nucleophilic amina-
tion has a general character. Indeed, many reactions
formally follow the scheme of nucleophilic substitution
of hydroxy group as shown in Scheme 4.
1
3
(
(
five m, 10H, ring), 2.91–2.97 (m, 1H, ring), 7.37–7.44
m, 3H, phenyl), 8.00–8.02 (m, 2H, phenyl), 8.51 (br s,
+
3
H, NH ). The salt obtained from benzoic acid and
3
cyclohexylamine has mp 140–142°C and an identical
H NMR spectrum.
Scheme 4.
1
R
R
δ+ δ–
H—Z +
OH
Z + H2O.
Diethyl (cyclohexylamino)(phenyl)methylphosphonate
In particular, in addition to the Kabachnik–Fields
reaction, these reactions are the formation of acetals
and aminals and the Mannich, Rodionov, and Strecker
(9)
(a) A mixture of 0.4 g (2 mmol) of freshly distilled
reactions. Because OH is a poor leaving group, for the N-benzylidene-N-cyclohexylamine 11 and 0.25 mL
nucleophilic substitution to occur, either acid catalysis (2 mmol) of diethyl phosphite in 5 mL of anhydrous
is required or group R should provide very strong reso- benzene in the presence of 1 mol % of formic acid was
DOKLADY CHEMISTRY Vol. 420 Part 2 2008

1
40
MATVEEVA, ZEFIROV
stirred at ambient temperature for 12 h under an argon
atmosphere. The completion of the reaction was moni-
tored by TLC. The reaction mixture was chromato-
graphed on a silica gel column using a dichlo-
romethane–methanol (20 : 1) mixture as the eluent to
ACKNOWLEDGMENTS
We are grateful to Prof. R.A. Cherkasov and
Prof. I.V. Galkin for helpful discussions and to
T.A. Podrugina for running the experiment.
This work was supported by the Russian Foundation
for Basic Research (project no. 05–03–33054), the
Council for Grants of the President of the Russian Fed-
eration for Support of Leading Scientific Schools (grant
no. NSh–2552.2006.3), the targeted program “Devel-
give 0.60 g (90%) of aminophosphonate 9 as a viscous
1
oil. H NMR (CDCl , δ, ppm): 1.10 (t, 6H, ëç ), 1.26–
3
3
1
.37 and 1.51–1.89 (both m, 10H, ring), 2.33 (br s, 1H,
NH), 3.74–4.14 (three m, 4H, éëç ), 4.19 (d, 1H, CH,
2
2
Jçê = 22.3 Hz), 7.40–7.42 (m, 3H, phenyl), 7.24–7.41 opment of the Scientific Potential of Higher Education”
1
3
of the Ministry of Education and Science of the Russian
Federation (project no. RNP.2.1.1.7779), and the Divi-
sion of Chemistry and Materials Science of the RAS
(m, 5H, phenyl). ë NMR (CDCl , δ, ppm): 16.21 and
3
1
6.47 (both d, ëç ); 24.40, 26.06, 32.00, and 34.37
3
3
(4C, ring); 53.42 (d, C1, ring, J = 15.4 Hz); 57.60 (d,
ëê
(
program no. 1).
1
ëç, J = 152.19 Hz); 62.58 and 63.10 (both d,
ëê
éëç ); 127.62, 128.32, 128.44, and 136.80 (s, C
2
REFERENCES
3
1
arom.). ê NMR (CDCl , δ, ppm): 24.24.
3
1
. Kabachnik, M.I., Medved’, T.Ya., Kabachnik, M.I., and
(
b) A mixture of freshly distilled imine 11 (0.4 g,
Medved’, T.Ya., Dokl. Akad. Nauk SSSR, 1952, vol. 83,
no. 5, pp. 689–717.
2
5
mmol) and 0.25 mL (2 mmol) of diethyl phosphite in
mL of anhydrous benzene in the presence of 5 mol %
2. Fields, E., J. Am. Chem. Soc., 1952, vol. 74, p. 1528.
t
of PcAlCl was stirred at ambient temperature for 5 h
under an argon atmosphere. The completion of the
reaction was monitored by TLC. The reaction mixture
was chromatographed on a silica gel column using a
dichloromethane–methanol (20 : 1) mixture as the elu-
ent to give 0.62 g (95%) of aminophosphonate 9 identi-
cal to the above sample.
3. Zefirov, N.S. and Matveeva, E.D., ARKIVOC, 2008,
vol. 2008, part (i), pp. 1–17.
4
5
6
. Cherkasov, R.A. and Galkin, V.I., Usp. Khim., 1998,
vol. 67, pp. 940–968.
. Galkina, I.V., Galkin, V.I., and Cherkasov, R.A., Zh.
Obshch. Khim., 1998, vol. 68, pp. 1469–1475.
. Galkina, I.V., Sobanov, A.A., Galkin, V.I., and Cherka-
sov, R.A., Zh. Obshch. Khim., 1998, vol. 68, pp. 1465–
1
468.
(c) A mixture of 0.25 g (2 mmol) of freshly distilled
benzaldehyde, 0.2 g (2 mmol) of freshly distilled cyclo-
hexylamine, and 0.25 mL (2 mmol) of diethyl phos-
phite in 5 mL of anhydrous benzene in the presence of
7. Galkina, I.V., Zvereva, E.R., Galkin, V.I., Sobanov, A.A.,
and Cherkasov, R.A., Zh. Obshch. Khim., 1998, vol. 68,
pp. 2224–2227.
. Dimukhametov, M.N., Bayandina, E.V., Davy-
dova, E.Yu., Gubaidullin, A.T., Litvinov, I.A., andAlfon-
sov, V.A., Mendeleev Commun., 2003, p. 150.
8
1
mol % of formic acid was stirred at ambient temper-
ature for 12 h under an argon atmosphere. The comple-
tion of the reaction was monitored by TLC. The reac-
tion mixture was chromatographed on a silica gel col-
umn using a dichloromethane–methanol (20 : 1)
mixture as the eluent to give 0.56 g (85%) of amino-
phosphonate 9 identical to the above sample.
9
. Sobanov, A.A., Zolotukhin, A.V., Galkina, I.V.,
Galkin, V.I., and Cherkasov, R.A., Zh. Obshch. Khim.,
2
006, vol. 76, pp. 442–451.
1
1
0. Sobanov, A.A., Zolotukhin, A.V., Galkin, V.I., Mos-
tovaya, O.A., Cherkasov, R.A., and Pudovik, A.N., Zh.
Obshch. Khim., 2003, vol. 73, pp. 921–926.
(d) A mixture of 0.25 g (2 mmol) of freshly distilled
1. Sobanov, A.A., Zolotukhin, A.V., Galkin, V.I., Cherka-
sov, R.A., and Pudovik, A.N., Zh. Obshch. Khim., 2002,
vol. 72, pp. 1141–1144.
benzaldehyde, 0.2 g (2 mmol) of freshly distilled cyclo-
hexylamine, and 0.25 mL (2 mmol) of diethyl phos-
phite in 5 mL of anhydrous benzene in the presence of
1
1
2. Gancarz, R., Tetrahedron, 1995, vol. 51, no. 38,
t
5
mol % of PcAlCl and 0.5 g of 4 Å molecular sieves
p. 10 627.
was stirred at ambient temperature for 5 h under an
argon atmosphere. The completion of the reaction was
monitored by TLC. The reaction mixture was chro-
matographed on a silica gel column using a dichlo-
romethane–methanol (20 : 1) mixture as the eluent to
give 0.66 g (98%) of aminophosphonate 9 identical to
the above sample.
3. Matveeva, E.D., Podrugina, T.A., Prisyazhnoi, M.V., and
Zefirov, N.S., Izv. Akad. Nauk, Ser. Khim., 2007, vol. 56,
no. 4, pp. 768–774.
1
1
4. Matveeva, E.D. and Zefirov, N.S., Zh. Org. Khim., 2006,
vol.42, pp. 1254–1255.
5. Collins, D. and Graymore, J., J. Chem. Soc., 1957,
pp. 9–10.
DOKLADY CHEMISTRY Vol. 420 Part 2 2008