Kinetics and mechanism of reaction between N-polyfluorophenylcarbonimidoyl dichlorides and tertiary amines in acetonitrile

Article abstract of DOI:10.1023/B:RUJO.0000010178.95945.a5

The reaction of N-polyfluorophenylcarbonimidoyl dichlorides with tertiary amines in acetonitrile afforded chloroamidines R²NC(Cl)=NAr _F and alkyl chloride. The precursor of the products is the corresponding quaternary ammonium salt [R³N+C(Cl)=NAr _F]-Cl^-. The rate of the salt formation is described by a second order equation; however with some amines a saturation effect was observed for the reaction rate with the growing amine concentration. This fact and also the influence of the amine and the substrate structure on the reaction rate suggests that reaction proceeds by addition-elimination mechanism with formation of a tetrahedral intermediate. The latter in the rate-limiting stage undergoes a stereomutation into an intermediate of a configuration favorable for conversion into a quaternary salt.

Full text of DOI:10.1023/B:RUJO.0000010178.95945.a5

Russian Journal of Organic Chemistry, Vol. 39, No. 8, 2003, pp. 1116 1124. Translated from Zhurnal Organicheskoi Khimii, Vol. 39, No. 8, 2003,
pp. 1187 1195.
Original Russian Text Copyright
2003 by Mikhailov, Savelova, Popov, Petrova, Platonov.
Kinetics and Mechanism of Reaction
between N-Polyfluorophenylcarbonimidoyl Dichlorides
and Tertiary Amines in Acetonitrile
V. A. Mikhailov, V. A. Savelova, A. A. Popov, T. D. Petrova, and V. E. Platonov
Litvinenko Institute of Physical Organic and Coal Chemistry,
National Academy of Sciences of Ukraine, Donetsk, 83114 Ukraine
Received November 7, 2001
Abstract The reaction of N-polyfluorophenylcarbonimidoyl dichlorides with tertiary amines in acetonitrile
afforded chloroamidines R2NC(Cl)=NArF and alkyl chloride. The precursor of the products is the cor-
responding quaternary ammonium salt [R3N+C(Cl)=NArF] Cl . The rate of the salt formation is described
by a second order equation; however with some amines a saturation effect was observed for the reaction rate
with the growing amine concentration. This fact and also the influence of the amine and the substrate
structure on the reaction rate suggests that reaction proceeds by addition elimination mechanism with
formation of a tetrahedral intermediate. The latter in the rate-limiting stage undergoes a stereomutation into
an intermediate of a configuration favorable for conversion into a quaternary salt.
N-Substituted
carbonimidoyl
dichlorides
tetrahedral intermediate II in the limiting stage [5,
8]. This kinetic pattern (k2> > k-1) arises apparently
due to easy elimination of HCl at the stages k2 and k4.
(RN=CCl2) are applied as electrophilic agents to the
synthesis of nitrogen-containing organic compounds,
in particular, nitrogen-containing heterocycles [1 3].
Their reactions with nucleophiles depending on reac-
tion conditions afford versatile products of replace-
ment at the carbon atom. This is illustrated by
Scheme 1 describing the intermediate and final
products arising in reaction of carbonimidoyl di-
chlorides with primary and secondary amines [4].
This opportunity is lacking in the reaction of di-
chlorides I with tertiary amines R3N. The tertiary
amines are known to serve as nucleophilic catalysts
in hydrolysis [11] and alcoholysis [12] of N-arylcarb-
onimidoyl dichlorides. The limiting stage of these
processes is a formation of quaternary ammonium salt
[RN=C(Cl) N+ R3] Cl [12]. Quaternary salts of
similar structure were isolated and characterized in
reactions of imidoyl chlorides (R1N=C(Cl)R2) with
N-heterocyclic amines in aprotic media [13]; kinetics
and mechanism of their formation was investigated
(see review [4]). However the information on the
reaction products obtained from carbonimidoyl di-
chlorides, in particular at the treatment with aliphatic
tertiary amines is very scanty [2].
In an aprotic medium secondary amines yield as
final product chloroamidine III; the chlorine atom in
the latter almost does not suffer substitution under
the conditions of the experiment [3, 5]. In a dioxane
water mixture [6] and in aqueous acetone [7] the
second chlorine atom in the chloroamidine III is
readily replaced, and the main reaction product is
guanidine IV.
The reaction of dichloride I with primary amines
both in aprotic [3, 8, 9] and protic [10] solvents
results in guanidine VI having as precursor carbo-
diimide V. Products V were observed in the reaction
mixture by spectral methods [3, 8, 9], and then they
were preparatively isolated.
In the present study kinetics was investigated of
reaction between N-polyfluorophenylcarbonimidoyl
dichlorides RC6F4N=CCl2 [VII, R=4-CH3 (a), 4-F
(b), 4-Br (c), 4-CF3 (d), 4-CN (e), 4-NO2 (f)] with
tertiary amines R1R2R3N [VIII, trimethyl- (a), tri-
ethyl- (b), dimethylbenzyl- (c), dimethylbutyl- (d),
diethylmethyl- (e), diethylisopropyl- (f), and diethyl-
isobutylamines (g), N-methylmorpholine (h), N-ethyl-
morpholine (i), N-ethylpiperidine (j), diazabicyclo-
octane (k)] in acetonitrile [equation (2)].
The investigation of aminolysis kinetics of N-poly-
fluorophenylcarbonimidoyl dichlorides by conducto-
metric method revealed that although reaction pro-
ducts obtained from primary and secondary amines
were different the reaction mechanism involved the
same addition elimination with formation of the
1
In the H NMR spectrum of a mixture of N-penta-
fluorophenylcarbonimidoyl dichloride (VIIb) with
1070-4280/03/3908-1116$25.00 2003 MAIK Nauka/Interperiodica

KINETICS AND MECHANISM OF REACTION
1117
(1)
triethylamine (VIIIb) in CD3CN (see EXPERI-
MENTAL) appeared two quartets at 4.2 and 3.7 ppm
corresponding to the protons of the N-quaternized
group in the intermediate product IX and to the
protons of diethylamino group in chloroamidine X
respectively. In the course of the reaction the intensity
of the signal at 4.2 ppm first grows and then
diminishes, and the intensity of the quartet at 3.7 ppm
grows continuously. In 24 h when all triethylamine
In all cases the apparent pseudofirst order rate con-
stants k1 are constant in the course of the process and
independent of imidoyl dichloride VII concentration
indicating that in the given concentration range the
reaction is first order with respect to substrate. As a
rule the values k1 are proportional to amine
concentration evidencing that reaction is also first
order in amine. Thus for the majority of tertiary
amines the kinetics is similar to that observed for
primary and secondary amines [5, 8]. However some-
times (see Fig. 1) the linear relation k1 [B] (where
[B] is molar concentration of tertiary amine) is
violated. We will discuss this phenomenon later.
1
in the reaction mixture was consumed in the H NMR
spectrum were revealed ethyl chloride (XI) signals
(3.7 q, 1.47 t). Therefore the reaction of imidoyl
dichlorides VII with tertiary amines VIII can be
described as including two consecutive stages, as
follows.
In Table 1 are listed second-order rate constants
(k2) for reaction of dichloride VIIb with tertiary
amines VIII calculated proceeding from the linear
parts of relations k1 [B] . The effect on the reaction
rate of amine structure is fairly described by Taft
equation modified by Bogatkov, Popov, and Litvi-
nenko [16]:
*
logk2
=
(5.67 0.36)
(1.95 0.13)EN,
r 0.975, So 0.358, n 9,
(2.75 0.66
+
(3)
(2)
The kinetics of reaction between dichlorides VII
and amines VIII was studied in excess amine:
6
1
[VIII]> > [VII] 10 mol l . The reaction rate was
measured by conductometric method as in the pre-
viously studied reactions with primary and secondary
amines [5, 8]. Since according to Scheme 2 a ionic
product forms only in the first stage, the use of the
conductometric measurements makes possible to
evaluate just the rate of this stage. Under the condi-
tions of kinetic experiments the rate of conversion of
the quaternary salt IX into non-ionic products X and
XI may be neglected.
Fig. 1. Apparent rate constant of pseudofirst order ka1pp for
reaction of N-pentafluorophenylcarbonimidoyl dichloride
with triethylamine as a function of the nucleophile con-
centration [B], acetonitrile, 25 C.
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

1118
MIKHAILOV et al.
Table 1. Second-order rate constants k2 for reaction of N-pentafluorophenylcarbonimidoyl dichloride (VIIb) with
tertiary amines VIIIa k (acetonitrile, 25 C)
1
*
Amine
Range of amine concentration:
k2, l mol 1 s
na
[14]
EN [15]
1
[VIII] 104, mol l
VIIIa
VIIIb
VIIIc
VIIId
VIIIe
VIIIf
VIIIg
VIIIh
VIIIi
VIIIj
VIIIk
0.0292...3.28
56.6...3900
18.7...502
1.19...18.6
25.2...111
40.6...9340
224...3150
240...3050
503...10510
189...2570
0.0956...0.307
465 20
0.110 0.006
4.08 0.12
34.9 1.3
1.07 0.17
0.0380 0.0005
0.0164 0.0004
0.704 0.020
0.0179 0.0017
0.260 0.006
3200 120
7
31
21
7
3
10
7
5
4
7
5
0.0
0.3
0.215
0.13
0.20
1.54
3.80
2.23
2.2
3.0
0.38
0.325
0.67 [15] (0.13)b
0.57 [15] (0.03)b
0.28 [15]
0.39 [15] (-0.15)b
3.0
3.5
3.5
1.3
a
b
*
n is the overall number of points. Estimated
applied in this study for calculations along equation (3), (see text).
Table 2. Second-order rate constants k2 and first order one klim for reaction 4-substited polyfluorophenylcarbonimidoyl
dichlorides (VIIa f) with tertiary amines (VIIIb) and dimethylbenzylamine (VIIIc) (acetonitrile, 25 C)
VIIIb
VIIIb
Compd.
no.
R
[14]
k2 102, l mol 1 s
klim 102, s
k2, l mol 1 s
1
1
1
VIIa
VIIb
VIIc
VIId
VIIe
VIIf
4-CH3
4-F
4-Br
4-CF3
4-CN
4-NO2
-0.17
0.062
0.232
0.54
1.59
11
11
44
66
0.14
0.6
0.3
4
2
4
1.39
2.23
7.0
65
31
31
0.10
0.20
2.0
1.67
4.08
9.67
37.4
125
0.03
0.12
0.25
1.3
12
6
1
0.66 (0.89)a
0.778 (1.25)a
14
5
81
102
a
The values of are given in parantheses. [18].
*
where
characterizes the inductive effect of
secondary amines in reaction with the same substrate
has the following correlation parameters [8]:
substituents R1,R2, and R3 at the nitrogen, and EN
is the steric constant of the amine; the factors before
these variables characterize the sensitivity of the
reaction to the inductive and steric effect respectively.
The data of Table 1 for amines VIIIf and VIIIg are
excluded from the correlation because the correspond-
*
logk2
=
(7.35 0.34)
+ (2.03 0.12)EN,
r 0.983, So 0.315, n 14.
(5.09 0.33)
(4)
Our attention is engaged by the essentially smaller
2
*
values of logko and
in the reaction of tertiary
ing EN data are lacking. In calculations by equation
(3) the values
*
amines as compared to that of primary and secondary
amines. It was already noted that in the latter case a
mechanism was substantiated consisting in addition
elimination with a limiting stage k1 in Scheme 1. The
mentioned lower reactivity of the tertiary amines com-
pared to primary and secondary ones (at equal * and
EN) within the framework of the common mechanism
of all the three amine classes may be attributed to the
changed limiting stage of the reaction: k2 >> k for
primary and secondary amines (Scheme 1) 1 and
k2 << k 1 for tertiary amines (Scheme 5):
used for heterocyclic amines
VIIIh, i, k (see Table 1, in parentheses) were reduc-
ed by 0.54 compared with the published data [15].
This approach was justified for morpholine [5, 17]
and proved to be efficient for tertiary amines.
Evidently the increased nucleophilicity of morpho-
lines (and diazacyclooctane) is caused by intramol-
ecular (transannular) interaction of two heteroatoms.
The corresponding correlation equation describing
the combined effect of the structure of primary and
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

KINETICS AND MECHANISM OF REACTION
1119
(5)
Fig. 2. Reciprocal values of rate constants of pseudofirst
Applying the method of stationary concentrations
and taking into consideration the mass balance with
respect to tetrahedral intermediate XII of Scheme 5
we obtain the following expression for k1:
order 1/k1app in reaction of N-pentafluorocarbonimidoyl
dichloride with triethylamine as a function of reciprocal
amine concentration 1/[B], acetonitrile, 25 C.
(6)
The dependence of k1 on amine concentration [B]
predicted by equation (6) will be linear with the
constraint k + k2 >> k1[B]:
1
(7)
where k2 = k1k2/k-1 + k2; if k >> k2, then k2 = K1*k2,
1
where K*1 = k1/k .
Fig. 3. Values of logk2 (1, 2), logklim (3) as functions of
Hammett s constants for reaction of 4-substituted poly-
fluorophenylcarbonimidoyl dichlorides VII with dimethyl-
benzylamine (VIIIc) (1) and triethylamine (VIIIb) (2, 3)
(acetonitrile, 25 C).
1
,
In the general case the dependence of k1 on amine
concentration would be curvilinear. The limiting
value of k1 designated hereinafter as klim is realized
under condition (k-1 + k2)<< k1[B]. In order to
estimate this value the expression (6) is transformed
in reciprocal coordinates:
For the reaction of N-pentafluorophenylcarbon-
imidoyl dichloride (VIIb) with the three above
amines the klim values are as follows:
(8)
Amine
VIIIb
VIIIf
VIIIg
Here the constant k2 plays the role of klim. The
curvilinear k1 as a function of amine concentration is
actually observed for three amines, VIIIb, f, g (see
an example in Fig. 1). This curve can be linearized in
the coordinates of equation (8) as shows Fig. 2. It is
probable that such form of dependence (6) would
have been observed with all the other amines if the
reaction rate would be studied at higher concentration
than that listed in Table 1. However it is difficult to
perform for the reaction rates are too high.
1
klim 103, s
K1*, l mol
22.3 2.1
4.9 3.1
3.16 0.08
12.0 0.4
3.34 0.17
4.9 0.4
1
As seen, the values K*1 estimated from the ap-
proximate expression k2 = K1*klim are of relatively
small absolute magnitude. It is clear that a significant
amine excess is required for the accumulation of
intermediate XII to become kinetically meaningful.
In order to support the above stated assumption
that the stage k2 in Scheme 5 is rate-determining we
studied the effect of substituent R in imidoyl di-
chlorides VII on the value of k2 in reaction with
triethylamine (VIIIb) and dimethylbenzylamine
(VIIIc), and also on the value of klim in reaction with
triethylamine (Table 2). The effect of R substituents
on the bimolecular rate constant k2 in reaction of
If the outlined version is true then the values klim
(k2) estimated by equation (8) characterize the uni-
molecular decomposition of intermediate XII in
Scheme 5, and the values k1 listed in Table 1 in a
sense correspond to combination of constants k1k2/k
1
+ k2 or in a simpler case (k >> k2) to a product
K*k , where K* is the equilibri1um constant of inter-
1 2
1
mediate XII formation.
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

1120
MIKHAILOV et al.
amine VIIIc with four dichlorides (VII, 4-R = CH3,
F, Br, CF3) fits to Hammett,s equation with excellent
statistical parameters:
The point for R = CN fits well into correlation
(9) if instead of Hammett,s equal to 0.66 [1, 18] is
used the value
0.89 [18] (see Fig.3, 1):
logk2 = ( 1.47 0.02)+ (1.91 0.06)
r 0.999 so 0.03 n 4
(9)
logk2
=
( 1.46 0.03) + (1.80 0.06)
r 0.998 so 0.053 n 5.
(10)
(11)
This means that a significant contribution to
transmission of the electronic effect of the substituent
provides an effect of direct polar conjugation result-
ing in stabilization of the tetrahedral intermediate:
0.66 and
0.78 and not
0.89 and
1.25
NO2
(Fig. 3, 2).NAO2nd for the process described with klim it
is possible to mark the change of sign ( < 0). It
was already mentioned that a notable decrease in
sensitivity of rate constant k2 to the influence of R
was also observed in reaction of amine VIIIc with a
substrate containing 4-NO2 substituent (Fig. 3, 1).
Such character of influence of substituent R in sub-
strate does not agree with a mechanism of addition
elimination (Scheme 5) with a limiting stage of tetra-
hedral intermediate decomposition (k2).
CN
4-Nitro group also should possess a similar effect
(
+ 1.25 [18]). However it significantly falls out
This inconsistency may be apparently removed re-
maining in the framework of a unique mechanism for
reactions of primary, secondary, and tertiary amines
if a formation is assumed on the reaction coordinate
of one more intermediate similar in structure to
intermediate XII in Scheme 5 or intermediate II in
Scheme 1 but unlike the final product. This assump-
tion is built on the stereochemical result of the amino-
lysis of N-pentafluorophenylcarbonimidoyl dichloride
(VIIb) with dibenzylamine in acetonitrile [19].
from the correlation (10) (see Fig. 3, 1). The cause
of this deviation will be discussed further.
A positive sign of in correlation (10) or (9) is
expectable and in a good agreement with the publish-
ed data on
various imidoyl chlorides with nucleophiles [4]. It
should be noted that in this reaction series no satura-
tion of rate was observed with increasing amine VIIIc
concentration.
values for bimolecular reactions of
The effect of substituents on k2 and klim values
for dichlorides VII reactions with triethylamine
(VIIIb) is more complicated (Fig.3, 2, 3), and for
six substituents basically cannot be described with
Hammett equation. The qualitative pattern of k2 and
It was established that the arising chloroform-
amidine (III), R = C6H5CH2) had Z and not E-con-
figuration (Scheme 11).
Quantum-chemical calculations of chloroform-
amidine (III, R = C2H5) showed that the Z-isomer is
thermodynamically more favored [19]. Thus the
observed stereochemical result is due to a thermo-
dynamic control. A thermodynamic control of sub-
stitution stereochemistry was also observed in aryl-
aminolysis of N-benzenesulfonylbenzimidoyl chloride
[PhSO2N=C(Cl)Ph] in acetonitrile where as a limit-
ing stage was postulated the decomposition of a tetra-
klim dependence on
is in general analogous. For
four substituents (4-CH3, 4-F, 4-Br, 4-CF3) the trend
in substituent effect coincides, and corresponds to
the positive . But with electron-withdrawing sub-
stituents (4-CN and 4-NO2) the sensitivity of the rate
constant k2 to the R substituent effect sharply
diminishes (
0) even when are used values
CN
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

KINETICS AND MECHANISM OF REACTION
1121
hedral intermediate [20]. This explanation of the
stereochemical result assumes in its turn that in the
intermediate IIa an internal rotation is possible
providing the thermodynamically more stable con-
figuration IIb whence Z-isomer forms in the stage k3
(Scheme 11) ensuring the thermodynamic control
[19]. In this case either stage k2 or k3 may be the
limiting one.
the amine attack on the substrate, i.e. k2 = k1. This is
the case in the reactions of primary and secondary
amines [5, 8], as has been already mentioned before.
Under additional constraint k << k an expression
2
1
k2 = K*1k2 is valid. Under these conditions the value
klim is numerically equal to k2:
klim
=
k2
(16)
In a reversed situation when k3 << k and k2 the
2
Applying to Scheme 11 the method of stationary
concentration and taking into account the mass
balance for the intermediate products IIa and IIb we
obtain the following expression for the rate constant
of the pseudofirst order:
k2 is expressed by an equation
(17)
and klim by an equation
(12)
(18)
where K*2 is an equilibrium constant of stereomutation.
It follows from this equation that to the linear
part of the relation of k1 to [B] corresponds the
following expression for the second order rate
constant:
From the viewpoint of the above reasoning the
influence of substituents R contained in the substrate
on the processes described by the constants k2 and klim
may be interpreted as follows. In reaction of carb-
onimidoyl dichlorides (VII, 4-R = CH3, F, Br, CF3)
with amines VIIIb and VIIIc* the rate-determining
stage is the stereomutation of intermediate IIa into
IIb (stage k2). In this case the constant k2 is
determined by equation (15), and the constant klim by
(13)
and equation (14) for the limiting value of the pseudo-
first order rate constant:
(14)
equation (16). The estimation of parameter
from
three substituents** (4-R = CH3, Br, CF3) for the
As seen, the value k2 is a combination of five
bimolecular reaction with triethylamine (k2) gives a
constants corresponding to individual stages, and kli
m
value of 2.03 0.04 which is close to that for the
is a combination of three constants. This is a quite
sufficient reason to expect that the principle of
linearity of free energies might not be valid in the
total range of variation, at least with respect to k2.
reaction with dimethylbenzylamine VIIIc [see
equation (9)].
in
The estimation of parameter for rate constant of
the unimolecular reaction (klim = k2) using the same
points gave the value 2.32 0.38. The positive
means here that the stereomutation (k2) is facilitated
by electron-withdrawing substituents, and the sensi-
tivity to this effect is of the same order of magnitude
as observed in the stage of nucleophilic attack (cf.
with value for the stage k1 in reaction of substrates
VIIa f with dibenzylamine in acetonitrile at 25 C
that is equal to 2.11 [5]). Since from the above
reasoning follows that in the simplest case k2 = K*1 k2,
it is possible for reaction of dichlorides VIIa d with
However simpler versions of equations (13) and
(14) are also possible providing the rate-determining
stage is either the stereomutation of intermediate IIa
[stage k2 in Scheme 11] or the decomposition of
intermediate IIb to the final product (stage k3).
If k3 >> k , then
2
(15)
triethylamine to make a rough estimation of the value
*
The result is similar to that obtained in analysis of
Scheme 5, but it means that the rate-determining
stage is not the transformation of the primarily
formed intermediate into the reaction products, but its
stereomutation into the more thermodynamically
stable form IIb. And when an additional constraint
for the equilibrium of the intermediate formation:
1
*
0.5.
1
*
In the case of this amine to the series of R in parentheses
should be added R = CN.
**
The point for 4-F-substituent strongly deviates from
,
Hammett s correlation by obvious reasons (Fig. 3, 2).
k2 >> k is valid then the rate-determining stage is
1
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

1122
MIKHAILOV et al.
Table 3. Rate constants of second order k2 and first
order klim of reaction between
N-pentafluorophenylcarbonimidoyl dichloride and
triethylamine in acetonitrile at various temperatures
Since the k2 value is a combination of rate con-
stants the calculated EA, H , and S also include
the corresponding parameters of the individual stages.
In the simplest case k2 = K1*k2. Therefore being in
possession of H and S values for stereomutation
stage (k2) we were able to calculate the enthalpy
( H*) and entropy ( S*) of intermediate IIa forma-
tion in Scheme 11.
Range of amine
2
k2 102, klim 10 ,
concentrations,
T, K
1
1
l mol 1 s
s
1
[B] 103, mol l
298
308
318
328
5.66...3410
4.02...3400
4.88...3210
5.88...3020
1.0 1.0 2.2
15.0 0.7 3.3
25.0 1.0 6.3
26.0 2.0 8.6
0.2
0.7
0.9
0.9
As seen, these figures should be regarded as rough
estimation for the calculation errors would be neces-
sarily high*. Nonetheless, it is possible to state that
the negative sign at H* evidences the exothermic
character of the formation equilibrium of intermediate
T*E (see further Scheme 21). The latter process occurs
virtually without entropy change ( S is a statistical
zero).
With growing acceptor character of substituents
(4-R= CN, NO2) the stage (k3) in Scheme 11
becomes rate-determining: k2 is described by equa-
tions (13) or (17), and klim by equations (14) or (18).
Since in all the above equations the rate constant k3 is
present characterizing the unimolecular decomposi-
tion of intermediate IIb (with negative ), this fact
Summing up the data of this study and those from
[5, 8] it is possible to build up a general scheme of
replacement mechanism of the first chlorine atom in
the N-substituted carbonimidoyl dichlorides by
primary, secondary, and tertiary amines (Scheme 21)
where in the general case R3N = R 1R2R3N.
will necessarily result in diminishing of
value at
attempts to build up Hammett,s correlations for the rate
constants k2 and klim till with a negative sign would
be obtained as has been actually shown by an example
of klim (Fig. 3, 3).
The difference between tertiary amines on the one
hand and primary and secondary amines on the other
hand consists in the existence of the tetrahedral inter-
mediate in the former case only in a zwitter-ionic
form T , and in the latter in the thermodynamically
more stable neutral form T0 forming from the T
through a fast proton transfer. This fact is decisive
for the relative ratio of individual stages rates
providing the same stereochemical result. In the
charged intermediate T (with tertiary amines) the
stereomutation of the primarily arising intermediate
TE into TZ form occurs (all other factors being equal)
significantly slower than the stereomutation of the
neutral intermediate T0E into T0 (with primary and
secondary amines). In the other words, the proton
transfer in the intermediate makes all stages after k1
more energetically favorable. Therefore in reactions
with primary and secondary amines the kinetic
behavior follows a single pattern with rate-determin-
ing stage k1 disregarding the structure of amine or
substrate. In contrast, for reactions with tertiary
In Table 3 are compiled the data on the tempera-
ture effect in reaction of N-pentafluorophenylcarb-
onimidoyl dichloride (VIIb) with triethylamine
(VIIIb) on the processes described with the rate
constants k2 and klim. The above reasoning indicates
that in this range at 25 C (see Fig. 3, 2 and 3) are
likely valid expressions (15) and (16) for k2 and klim
respectively.
In the temperature range studied both data for k2
and klim are well consistent with Arrhenius equation.
logk2
=
(4.33+ 0.46)
(1.57+ 0.14) 1/T
(19)
(20)
logklim = logk2 = (5.08+ 0.40)
+ 0.13) 1/T
(2.01
From parameters of equations (19) and (20) the
activation energy (EA), enthalpy ( H ), and entropy
( S ) were estimated for the processes described with
rate constants k2 and k2.
amines the stages k2 (k ) and k3 are more sig-
2
Process
EA,
kJ mol
H , ( H*)
kJ mol
S ( S*),
1
nificant from the energy viewpoint than k1, and thus
at certain reagents structure they become rate-
determining.
1
1
J mol 1 deg
k2
k2
K1
30.0 2.7
38.5 2.4
27.5 2.5
36.0 2.2
8.5 3.4
170 18
156 12
14.0 24
*
Absolute value of errors in H* and S* were calculated as in
[21].
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

KINETICS AND MECHANISM OF REACTION
1123
(21)
EXPERIMENTAL
p-toluenesulfonyl chloride was added in amount of
3 5 wt% with respect to amine. and the mixture was
kept for several hours. Then the mixture was heated
in a flask with a reflux condenser, and the liberated
trimethylamine was passed through a column packed
with solid potassium hydroxide and was absorbed by
acetonitrile. The concentration of the solution obtain-
ed was measured by acid base titration. The working
solutions were prepared by dilution.
1H NMR spectra were registered on spectrometers
Gemini 200 (200 MHz) and Tesla BS467 (60 MHz).
The preparation of carbonimidoyl dichlorides was
described in [1, 22].
The triethylamine was heated with p-toluene-
sulfonyl chlorides (about 5 wt%) for several hours,
then it was twice distilled and the middle fraction was
separated. The latter was refluxed for several hours
over sodium metal and then distilled on a column
40 cm long separating the fraction of bp 89 89.5 C
(publ. bp 89.5 C). In going to a new portion of tri-
ethylamine several kinetic measurements were per-
formed for reaction with N-pentafluorophenylcarb-
onimidoyl dichloride in acetonitrile at 25 C (the most
thoroughly studied reaction). If the results obtained
did not fit to the previous data the purification was
repeated.
1,4-Diazabicyclo[2,2,2]octane was sublimed in a
vacuum from metallic sodium and immediately after
sublimation it was dissolved in acetonitrile. The
concentration of the solution obtained was measured
by acid base titration. The working solutions were
prepared by dilution.
The acetonitrile for kinetic measurements was
twice subjected to rectification on a column over
phosphorus pentoxide and over freshly calcined
potassium carbonate with separating the middle
fraction. The electroconductivity of such solvent was
Amines VIIIc j were purified by the same proce-
dure as triethylamine. Their boiling points were
consistent with the published data.
1
no more than 1 10-6 Ohm m .
Reaction of N-pentafluorophenylcarbonimidoyl
dichloride with triethylamine (preparative run).
To a solution of 0.6 g (2.3 mmol) of N-pentafluoro-
phenylcarbonimidoyl dichloride in 3 ml of anhydrous
Trimethylamine was obtained from its hydro-
chloride by treating with aqueous alkali, and it was
collected as a solution in acetonitrile. To this solution
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003

1124
MIKHAILOV et al.
ether was added at stirring 0.92 g (9.1 mmol) of tri-
ethylamine in 3 ml of ether. The mixture was stirred
at room temperature for 6 h. The precipitate was
separated, the filtrate was evaporated. The residue
was distilled in a vacuum (138 139 C at 10 mm Hg).
We obtained 0.62 g (90%) of N,N-diethyl-N-penta-
fluorophenylchloroformamidine. The characteristics
of the compound coincide with those published in [5].
5. Mikhailov, V.A., Kolesnikova, I.V., Popov, A.A.,
Petrova, T.D., Platonov, V.E., and Savelova, V.A.,
Zh. Org. Khim., 1989, vol. 25, p. 1030.
6. Hegarty, A.F. and Digman, K.J., J. Chem. Soc.,
Perkin Trans., II, 1975, p. 1046.
7. Bacaloglu, R. and Korodi, T., Rev. Roum. Chim.,
1979, vol. 24, p. 1165.
8. Mikhailov, V.A., Popov, A.A., Savelova, V.A.,
Kolesnikova, I.V., Petrova, T.G., and Platonov, V.E.,
Zh. Org. Khim., 1989, vol. 25, p. 1683.
9. Kolesnikova, I.V., Petrova, T.D., Platonov, V.E.,
Ryabicheva, T.G., Mikhailov, V.A., Popov, A.A.,
and Savelova, V.A., Zh. Org. Khim., 1989, vol. 25,
p. 1689.
10. Korodi, T. and Bacaloglu, R., Bul. Sti. Sitehn. Inst.,
politehn. Timisoara, Ser. Chim., 1979, vol. 24, p. 84;
Ref. Zh. Khim., 1981, 21B963.
Reaction of N-pentafluorophenylcarbonimidoyl
dichloride with triethylamine (1H NMR experi-
ment). Dry deutroacetonitrile used in the run was
obtained by distillation of CD3CN over phosphorus
pentoxide in a distillation device dried for 4 h at
180 C in a drying cabinet just before use. The
spectrum of thus prepared solvent contains only the
signal from the residual protons of the methyl group.
Into a freshly heated NMR tube was charged 0.140 g
(0.54 mmol) of N-pentafluorophenylcarbonimidoyl
dichloride and 0.4 ml of deuteroacetonitrile. The
mixture was frozen in the liquid nitrogen, and
0.07 ml (0.05 g, 0.5 mmol) of triethylamine was
added. The ampule was cautiously warmed till the
lower layer melted, and then the mixture was stirred
and became light-brown. Then the ampule was frozen
again and it warmed already in the probe of the
NMR spectrometer. Beside the resonances belonging
to triethylamine (1.0 t and 2.6 q ppm) in the spectrum
11. Bacaloglu, R. and Korodi, T., Ostrogovich G., Rev.
Roum. Chim., 1979, vol. 24, p. 933.
12. Bacaloglu, R., Korodi, T., Florescu, A., and Ostrogo-
vich, G., Rev. Roum. Chim., 1977, vol. 22, p. 877.
13. Drizhd, L.P., Kryuchkova, E.N., Litvinenko, L.M.,
Savelova, V.A., and Chotii, K.Yu., Ukr. Khim. Zh.,
1984, vol. 50, p. 11948.
14. Spravochnik khimika (Chemist,s Handbook), Moscow:
Khimiya, 1964, vol. 3.
15. Bogatkov, S.V., Belova, N.A., and Medved,, S.S.,
Reakts. Sposobn. Organ. Soedin., 1975, vol. 12,
p. 267.
appears
a
signal from imidoylammonium salt
(4.2 ppm q) and a lot more weak signal from chloro-
formamidine (3.7 ppm q). The upfield part of the
spectrum is overlapped with stronger signals of tri-
ethylamine. The intensity of imidoylammonium salt
resonance grew within 1 h, and then diminished; the
intensity of chloroformamidine signal steadily
increased. In 3 days in the mixture only signals of
chloroformamidine (1.25 t, 3.7 q ppm) and those of
ethyl chloride (1.47 t, 3.7 q ppm) were registered.
16. Bogatkov, S.V., Popov, A.F., and Litvinenko, L.M.,
Reakts. Sposobn. Organ. Soedin., 1969, vol. 6,
p. 1011.
17. Perepichka, I.F., Popov, A.F., Kostenko, L.I., and
Piskunova, Zh.P., Reakts. Sposobn. Organ. Soedin.,
1986, vol. 33, p. 320.
18. Gordon, A.J. and Ford, R.A., The Chemist,s Compa-
nion, New York: Wiley, 1972.
19. Mikhailov, V.A., Yufit, D.S., Balabanov, E.Yu.,
Struchkov, Yu.T., Popov, A.A., Savelova, V.A.,
Kolesnikova, I.V., Petrova, T.D., and Platonov, V.E.,
J. Phys. Org. Chem., 1993, vol. 6, p. 319.
20. Yufit, D.S., Struchkov, Yu.E., Mikhailov, V.A.,
Drizhd, L.P., and Savelova, V.A., Izv. Akad. Nauk
SSSR, Ser. Khim., 1988, p. 783.
21. Kassandrova, O.N. and Lebedev, V.V., Obrabotka
rezul, tatov nablyudenii (Processing of Measurements
Results), Moscow: Nauka, 1970, p. 88.
22. Savchenko, T.I., Petrova, T.D., Platonov, V.E., and
Yakobson, G.G., Zh. Org. Khim., 1979, vol. 15,
p. 1018.
The procedure used in kinetic measurements was
described before [23].
REFERENCES
1. Savchenko, T.I., Kolesnikova, I.V., Petrova, T.D.,
and Platonov, V.E., J. Fluor. Chem., 1983, vol. 22,
p. 439.
2. Petrova, T.D. and Platonov, V.E., Usp. Khim., 1988,
vol. 57, p. 405.
3. Kolesnikova, I.V., Petrova, T.D., Platonov, V.E.,
Mikhailov, V.A., Popov, A.A., and Savelova, V.A.,
J. Fluor. Chem., 1988, vol. 40, p. 217.
4. Savelova, V.A. and Popov, A.F., Zh. Org. Khim.,
1999, vol. 35, p. 821.
23. Litvinenko, L.M., Mikhailov, V.A., Savelova, V.A.,
and Drizhd L.P., Zh. Org. Khim., 1986, vol. 22,
p. 2343.
RUSSIAN JOURNAL OF ORGANIC CHEMISTRY Vol. 39 No. 8 2003