Kinetic isotope effects in hydrogen exchange of aromatic CH bonds in benzene, fluorobenzene, and nitrogen-containing heterocycles with solution of alkali metal amide in liquid ammonia

Article abstract of DOI:10.1023/A:1013134511332

The kinetic isotope effects in deuterium and tritium exchange in benzene, fluorobenzene, pyridine, pyridine N-oxide, and quinoline with a solution of an alkali metal amide in liquid ammonia k^NH3_D/k^NH3_T were determined, where k^NH3_D and k^NH3_T are the experimental rate constants of protodedeuteration and protodetritiation, respectively. The variation of the tritium exchange rates in benzene in going from NH₃ to ND₃ (k^NH3_T and k^ND3_T) was evaluated. The deviation of the observed ratios k^NH3_D/k^NH3_T and k^ND3_H/k^ND3_T from the maximum possible values corresponding to the Swan-Shaad equation suggests the reaction mechanism in which both elementary stages, ionization of the CH acid and diffusion separation of the complex of the carbanion with the ammonia molecule, are partially limiting. A small decrease in the secondary isotope effect of the solvent, defined as k^ND3_T/k^NH3_T, as compared to the theoretical maximum of 2.4, is assumed to be due to similar structures of the transition state and the equilibrium carbanion. A theoretical explanation was given for the observed dependences of the primary isotope effect of the substrate on pK_a for deutero (tritio) derivatives of fluorobenzene [4D(t) < 3D(t) < 2D(t)], for reactions in the series pyridine-2D(t) < pyridine-3D(t) < quinoline3D(t), and for hydrogen exchange in pyridine-2D(t) N-oxide (k^NH3_D/k^NH3_T ～ 1).

Full text of DOI:10.1023/A:1013134511332

Russian Journal of General Chemistry, Vol. 71, No. 7, 2001, pp. 1133 1143. Translated from Zhurnal Obshchei Khimii, Vol. 71, No. 7, 2001,
pp. 1200 1212.
Original Russian Text Copyright
2001 by Tupitsyn, Zatsepina.
Kinetic Isotope Effects in Hydrogen Exchange
of Aromatic CH Bonds in Benzene, Fluorobenzene,
and Nitrogen-containing Heterocycles with Solution
of Alkali Metal Amide in Liquid Ammonia
I. F. Tupitsyn and N. N. Zatsepina
Prikladnaya Khimiya Russian Scientific Center, St. Petersburg, Russia
Received April 19, 2000
Abstract The kinetic isotope effects in deuterium and tritium exchange in benzene, fluorobenzene,
pyridine, pyridine N-oxide, and quinoline with a solution of an alkali metal amide in liquid ammonia
NH NH
NH
NH
3
3/k
3
were determined, where kD and kT are the experimental rate constants of protodedeuteration
3
k
D
T
and protodetritiation, respectively. The variation of the tritium exchange rates in benzene in going from NH
3
NH
ND
NH NH
ND ND
3
3
and kT 3) was evaluated. The deviation of the observed ratios kD 3/k
3
and kH 3/k
to ND (k
from the
3
T
T
T
maximum possible values corresponding to the Swan Shaad equation suggests the reaction mechanism in
which both elementary stages, ionization of the CH acid and diffusion separation of the complex of the car-
banion with the ammonia molecule, are partially limiting. A small decrease in the secondary isotope effect of
ND NH
the solvent, defined as kT 3/k 3, as compared to the theoretical maximum of 2.4, is assumed to be due to
T
similar structures of the transition state and the equilibrium carbanion. A theoretical explanation was given
for the observed dependences of the primary isotope effect of the substrate on pK for deutero (tritio) deriv-
a
atives of fluorobenzene [4D(t) < 3D(t) < 2D(t)], for reactions in the series pyridine-2D(t) < pyridine-3D(t) <
NH NH
quinoline3D(t), and for hydrogen exchange in pyridine-2D(t) N-oxide (kD 3/kT
3
1).
The modern state of the mechanistic studies of
hydrogen isotope exchange of CH bonds in organic
molecules with basic reagents is contradictory. On the
one hand, owing to the maximum possible use of
kinetic isotope effects of various nature in combina-
tion with semiempirical correlations (like Brønsted or
Hammett equations), there has been a considerable
success in revealing the mechanism of reactions in
hydroxyl-containing solvents. As shown in numerous
studies, in these media the isotope exchange can be
a convenient test reaction for such protolytic processes
as enzymatic oxidation reduction, E2 elimination,
nucleophilic substitution of methane derivatives acti-
vated with nitro groups, etc. On the other hand, the
mechanism of the reaction in strongly basic highly
polar nonaqueous media such as solution of alkali
metal amide in liquid ammonia ( 26.7 at 60 C [1])
was practically beyond the scope of researchers’ inter-
est. At the same time, there are good grounds to ex-
pect that the hydrogen isotope exchange of organic
compounds in this medium will show different fea-
tures as compared to hydroxyl-containing media.
Indeed, the features of liquid ammonia (high capabil-
ity for solvation of metal cations as compared even
with the most efficient dipolar aprotic solvents [1]
and weak solvation of anions) suggest that hydrogen
isotope exchange in liquid ammonia, being an ionic
process, would be relatively weakly affected by the
medium and hence would be largely controlled by
forces acting between isolated reactants. The higher
protophilicity of amide ions dissolved in liquid am-
monia as compared to hydroxide ions in water follows
from simple comparison of their proton affinities in
1
the gas phase, kcal mol : 419 for NH and 384
2
for OH .
An additional argument in favor of different reac-
tion mechanisms in these reaction media is provided
by elemental chemical reactions. For example, nucleo-
philic dehalogenation of halobenzenes [2], 3- and
4
-chloropyridines [3], and 2-chloropyridine N-oxide
[4] in a solution of potassium amide in liquid ammo-
nia occurs chiefly by the aryne mechanism, whereas in
an alcoholic solution the major pathway is addition of
a nucleophile followed by elimination of the halide
ion [5].
In view of these facts, the goal of this work was to
reveal, as fully as possible, the features of the mech-
anism of deuterium and tritium exchange in a solution
of potassium (or sodium) amide in liquid ammonia
1
070-3632/01/7107-1133$25.00 2001 MAIK Nauka/Interperiodica

1134
TUPITSYN, ZATSEPINA
NH
Table 1. Rate constants of protodedeuteration (k 3) and
k /k = 2.1 0.1 (0 C, C
0.02 M). We obtained
0.05 M). The reason for
D
D
T
KNH2
NH
NH NH
3
protodetritiation (kT 3) and kD 3/kT ratio for deuterium
and tritium exchange of benzene-D(t) in a solution of
potassium amide in liquid ammonia (CKNH2 0.05 M)a
k /k = 1.9 0.1 (0 C, C
D
T
KNH2
this small difference is unclear, but it is of no princi-
pal importance. Along with measurements of the pri-
mary kinetic isotope effect, for revealing the exchange
mechanism it is important to study the kinetics of iso-
tope exchange in the catalytic system C H KND
NH
105,
NH
105,
3
3
Tempera-
ture,
k
k
D
T
NH3 NH3
k
/k
T
C
s 1
s 1
D
6
5
2
ND (liquid) and to study the secondary isotope effect
3
0
5
0
1.1 0.1
11 2
92 2
0.57 0.02
6.1 0.6
48 3
1.9
2.0
1.9
of the substrate defined as the ratio of the rates of
deuterium exchange of separate isotopomers with the
solvent in the same liquid ammonia solution [system
C D C H D KNH NH (liquid)]. These data are
2
4
6
6
6
5
2
3
a
Average values from 5 6 runs at each temperature.
listed in Tables 2 and 3.
To interpret the mechanistic features of the ex-
change reaction under consideration, we use a kinetic
scheme taking into account the intermediate formation
of the primary and secondary contact complexes [7]:
with structurally related CH acids containing the same
exchange center, aromatic CH bond [benzene-D(t),
fluorobenzene-3D(t), fluorobenzene-4D(t), pyridine-
2
3
D(t), pyridine-3D(t), quinoline-2D(t), quinoline-
D(t), and pyridine-2D(t) N-oxide]. These compounds
k1
k2
cover a relatively broad range of kinetic CH acidity
and therefore are convenient objects for studying the
specific features of the multistage mechanism and
revealing the differences in the structures of the carb-
anionic transition states characterized by high locali-
zation of the excess negative charge. Among the
chosen objects, the most convenient model compound
is the simplest compound, benzene, whose hydrogen
atoms are equivalent. Therefore, benzene was studied
in most detail. We attempted to apply the modern
theoretical concept, that explains the whole set of
experimental data on the values and variations of the
primary kinetic isotope effect and Brønsted coefficient
in isotope exchange reactions of OH, NH, and CH
acids in hydroxyl-containing media [6, 7], to hydro-
gen isotope exchange of aromatic CH bonds in a solu-
tion of potassium amide in liquid ammonia. Most
attention was given to the following problems: estima-
tion of the contribution of the internal return effect
described by the ratio of the rate constants of the
C H + NL2
C
HNL2
C L + NLH .
Here L = H, D, or T; the constants k and k char-
C LNLH
k 1
(1)
1
1
acterize formation and decomposition of the primary
contact complex; k is the rate constant of the diffu-
sion-controlled process of separation of the leaving
solvent molecule from the primary complex and for-
mation of the secondary contact complex.
2
In terms of scheme (1), the overall rate constant of
deuterium exchange kD is given by relation (2):
k k
1
2
(2)
k =
.
D
k 1 + k2
The expected behavior of k can be different de-
pending on the ratio of the constants of the elemen-
tary stages. If k >> k , then k = k and we deal
with the reaction not complicated by internal return.
On the contrary, if the intermediate primary contact
complex is so reactive that it abstracts a proton from
the solvent molecule and this process occurs faster
than transformation into the secondary contact com-
D
2
1
D
1
competing stages k 1 and k [scheme (1)]; evaluation
2
of the extent of structural transformation of the initial
CH acid in the carbanionic transition state when
moving along the reaction coordinate; and study of
the variation of the reaction mechanism depending
on the molecular structure of the CH acid.
plex (k 1 >> k ), then there is a simple correlation
2
between the rate constant of the exchange reaction kD
Primary kinetic effect of the substrate and in- and rate constant of ionization of the CH acid k . The
1
ternal return. The average rate constants of proto-
dedeuteration (kD 3) and protodetritiation (kT 3) of
overall rate in this case is determined by the diffusion
rate (rate constant k ): k = (k /k )k = K k .
NH
NH
2
D
1
1
2
preeq 2
benzeneD(t) and the kinetic isotope effect of the reac-
Internal return seems to be an essentially dy-
namic process depending on many factors (structure
of the CH acid and reacting base, viscosity of the
reaction medium, temperature). Therefore, in many
cases detection of internal return plays a decisive
role in elucidation of the reaction mechanism.
NH NH
tion (kD 3/k 3) with a solution of potassium amide in
T
liquid ammonia are listed in Table 1. Previously, in
a paper on a similar problem [8], Shatenshtein et al.
reported for the reaction of benzene-D(t) with a solu-
tion of potassium amide in liquid ammonia the ratio
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

KINETIC ISOTOPE EFFECTS IN HYDROGEN EXCHANGE
1135
Table 2. Conditions and results of experiments on deuterodeprotonation (k NH D3) and deuterodetritiation (k TN D3) of benzene-t
in a solution of potassium amide in liquid deuterated ammonia (0 C, initial deuterium concentration 99 at. %, CKNH2
0
.05 M)
Experi-
ment
dura-
Relative intensity of indicated peak in mass
spectrum, %
b
ND
ND
3
3
C ,
k
k
T
b
H
0
2 c
Sb 102,c
m /
ms
S
10 ,
a
5
b
5
at. %
D
10 ,
10 ,
a
1
1
counts min counts min
1
1
s
s
tion, h
78
79
80
81
82
83
84
1.5
3.0
4.0
6.0
82
87
76
87
3.0
15.8
0.5
30.5 30.6 15.6
3.9
0.3 42.1
13
14
5.3 18.4 34.3 30.8 10.6 70.2
2.2 12.9 1.8 36.9 16.2 75.3
1.4 10.8 38.1 49.7 89.3
184
386
294
146
286
196
2.1
2.1
1.9
15.3
d
a
b
Here and hereinafter: (m , m ) number of moles of the solvent and substrate, respectively. (C ) Concentration of deuterium atoms
a
s
b
c
0
in benzene after the experiments.
(S , S ) Activities of benzene samples before and after the experiments, respectively.
b b
d
Equilibrium.
Table 3. Conditions and results of experiments on protodedeuteration of the isotope mixture C H D and C D (1 : 1) in
6
5
6 6
0.05 M]
a
a solution of potassium amide in liquid ammonia [system C H D + C D + NH (liquid) + KNH , C
6
5
6
6
3
2
KNH2
Tem- Experi-
Run pera- ment
no. ture, duration, ms
Relative intensity of indicated peak
in mass spectrum, %
k
k
C D
C H D
m /
C
,
C
,
6
6
6
5
k
/
C H D
5
a
C D
C H D
6
6
6
5
5
1
5
6
1
0 ,
10 ,
k
at. % D at. % D
1
C D
s
s
6
6
C
h
78 79 80
1.8 88.2
81
82
83
84
1
3.0 19.5 77.5
95.8b
87.7
75.9
71.1
90.0
94.3
83.2
82.3
14.7b
13.1
11.0
10.5
13.5
15.3c
13.2
12.8
1
2
3
4
0
0
0
3.0
6.0
7.0
44.9 11.2 78.8 0.7 3.0 12.7 36.9 46.7
55.3 34.1 65.9 3.6 12.6 28.1 35.0 20.0
60.9 37.1 62.9 6.1 17.0 31.1 31.6 14.1
0.81
1.1
1.1
7.0
1.1
1.2
1.3
9.0
1.3
1.1
1.2
1.3
25
0.25 33.4 18.8 81.2 0.8 2.2
.1 91.9 0.8
9.8 30.7 56.5
3.9 24.6 71.3
b
8
5
6
25
25
0.33 35.7 20.8 79.2 2.7 5.5 17.6 37.5 36.6
0.5 44.9 23.3 76.7 3.4 7.9 18.4 31.8 38.4
10.4
7.6
12.4
10.0
1.2
1.3
a
(
C
) Concentration of deuterium atoms in heavy isotopomers (C D , C D H, C D H ); (C
) concentration of deuterium
) rate constants of proto-
C D
6
6
6
5
6
4
2
C H D
6
6
6
5
atoms in light isotopomers present in the isotope mixture (C H D , C H D, C H ); (k
dedeuteration of the corresponding isotopic species. Concentration of deuterium atoms C
taken for run nos. 1 4. Concentration of deuterium atoms C
, k
and C
6
4
2
6
5
6
6
C D
C H D
6
6
6
5
b
in the isotopic mixture
C D
C H D
6
6
6 5
d
and C
in the isotopic mixture taken for run nos. 5 and 6.
C D
C H D
6 5
6
6
NH
To find whether the isotope exchange of hydrogen
change in benzene-D(t) from the rate constants k 3,
D
NH
T
ND
ND
T
3
3
3
in benzene-D(t) with an alkali metal amide in liquid
ammonia is complicated by internal return, we esti-
mated the deviation of the exponent in the Swan
Shaad equation from the maximal value ymax = 3.34
corresponding to the case when the stage with the rate
k
, kH , and k
(0 C) found in this work,
where kD and kT are the rate constants of proto-
dedeuteration and protodetritiation, respectively ( re-
verse exchange); kH and kT are the rate con-
stants of deuterodeprotonation and deuterodetritiation,
respectively ( direct exchange, with solvent as deu-
terium donor).
NH
NH
3
3
ND
ND
3
3
constant k fully limits the reaction rate [9]:
1
log k NH D3/k TN D = y max log k DN H3/k 3.
NH
(3)
3
T
In contrast to the constants k ND H3, k 3, and k HN D3,
NH
T
Assuming, as usual, that the influence of the sol-
vent isotope effect on the primary kinetic isotope
effect is negligible, we determined, following the
above procedure, the exponent y for the isotope ex-
obtained from direct measurements (Table 1), the
ND
3
quantity kT was estimated indirectly on the basis
of the fact that the hydrogen isotope exchange in the
same reaction mixture occurs by two opposite path-
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

1136
TUPITSYN, ZATSEPINA
ways (Table 2). The first pathway is incorporation of
deuterium atoms from the solvent to the nondeuterated
benzene molecule; it prevails at short reaction times
approximately equal, because these stages involve
cleavage of bonds with a similar strength: CD(T) and
ND(T) [12]. Hence, the preequilibrium constant
Kpreeq = k /k , similar to k , tends to decrease with
(1.5 and 3 h, Table 2). In view of the large isotope
1
1
1
effect of the solvent, the loss of the tritium content in
the benzene molecule in this period can be neglected.
As for the second pathway involving exchange of tri-
tium in the benzene molecule with deuterium from the
solvent, we monitored its kinetics during a period
sufficient for the equilibrium in the back deuterium
exchange reaction to be arrained (4 6 h):
increasing temperature. As for the constant k , its
variation with temperature is governed by the close
2
correlation between k and the diffusion motion in
2
relatively bulky molecular complexes. Generally, the
lower the solution viscosity, the lower the potential
barrier to escape of the emerging solvent molecule
from the nearest environment of the carbanion and
hence the higher k . Naturally, variation with temper-
2
C D t + ND (liquid)
C D + ND t.
(4)
6
5
3
6
6
2
ature of the kinetic isotope effect under the influence
of the growing contribution of k will have the oppo-
2
Finally, the true deuterodetritiation rate constant
D
T
ND
site sign as compared to variation of K
/K
.
3
preeq preeq
k
was determined by correcting the apparent con-
T
The influence of temperature on k should be espe-
2
stant k (C D t ND ) for the secondary deuterium
T
6
5
3
cially significant in liquid ammonia which is charac-
terized by a high viscosity at low temperatures and a
strong temperature dependence of the viscosity [13].
Hence, the influence of two competing factors on the
kinetic isotope effect of hydrogen isotope exchange
of benzene with solution of potassium amide in liquid
ammonia causes the above-noted abnormal indepen-
isotope effect: k (C H D)/k (C D ) = 1.2 (Table 3).
D
6
5
D
6
6
ND
3
Then, kT = k (C D t ND )k (C H D)/k (C D ) =
T
6
5
3
D
6
5
D
6
6
5
1
2
2
1.2 10 5 = 2.4 10 s . The exponent y =
NH NH
D
3
3
.86 determined by substituting the ratios k /k
T
ND ND
3
3
and kH /k
in relation (3) is lower than the limit-
T
ing value ymax = 3.34 corresponding to the state when
the stage with the constant k in scheme (1), which is
NH3 NH3
1
dence of the kD /kT ratio from temperature. How-
sensitive to isotope substitution, is fully rate-determin-
ing. In other words, the kinetics of hydrogen isotope
exchange of benzene with a solution of potassium
amide in liquid ammonia is complicated in part by the
internal return.
ever, the relative contribution of the internal return to
the reaction mechanism cannot be estimated from the
NH3 NH3
temperature dependence of k /k . To estimate
D
T
T
the parameter
= k /k characterizing the extent of
T
1 2
internal return at tritium exchange, Streitwieser et al.
An additional unambiguous evidence of the internal
return is the zero temperature coefficient of the pri-
mary kinetic isotope effect in deuterium and tritium
[9] proposed the formula
[1
a (B 1
/ymax
1)]ymax = A + a (A
B), (5)
T
T
exchange of the aromatic CH bonds in benzene, found
y
y
NH NH
3
3
where y = 2.86; A = (k /k ) /(k /k ); B = (k /k )
by us: kD /k
= 1.9 2.0 at 0, 25, and 40 C
D
T
H
T
D
T
T
T
T
(
Table 1). In this connection, we should note the
K ; K is the equilibrium constant of the reaction
eq
eq
T
eq
following. Despite numerous data accumulated in the
recent decades that the primary kinetic isotope effect
in protolytic reactions can show no temperature de-
pendence [7, 10], or even the ratio k /k can increase
C H T + NH (liquid)
C H + NH T; K
1.2
6
5
3
6
6
2
[14].
Unfortunately, Eq. (5) for determining
cannot
T
D
T
be solved analytically, and it was solved numerically
by the iteration procedure. Knowing the extent of
internal return in tritium exchange, we can calculate,
without performing the experiment, the kinetic isotope
with temperature [11], the views that the kinetic iso-
tope effect should always monotonically decrease with
increasing temperature still prevail in the literature.
These views are based on the well-known theoretical
concept that the major contribution to the observed
isotope effect is made by the difference in the zero
vibration energies of the CD and CT bonds cleaving
in the course of the reaction. However, analysis shows
that this conclusion is valid only for exchange proc-
esses fully limited by the stage with the rate constant
k1 [scheme (1)]. The internal return can be the only
factor responsible for the abnormal behavior of the
ratio k /k , observed in this and some other [8, 10]
H
1
T
1
effect in the stage of hydrogen transfer (k /k ) and
the parameter
by Eqs. (6) and (7):
H
H
T
T
1)] 1,
k /k = k /k [1
a (K k /k
(6)
(7)
1
1
H
T
T
T
eq H
T
H
T
aH = a K k /k .
T
eq
1
1
The extents of internal return of triton and proton,
obtained by this procedure, are low: = 0.058 and
= 1.05 (0 C). Thus, we deal with a situation
T
D
T
H
works. Indeed, the temperature coefficient of the ki-
netic isotope effect in stages with the constants k1
already known for the hydrogen isotope exchange in
normal CH acids with an alkaline aqueous or methan-
olic solution (see, e.g., [7]): none of the three stages
and k , irrespective of the mechanistic features, is
1
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

KINETIC ISOTOPE EFFECTS IN HYDROGEN EXCHANGE
1137
of scheme (1) is definitely limiting, with the rate
(OD/OH secondary isotope effect). Namely, the ND2
ion in deuterated liquid ammonia is a stronger base
constants of the elementary stages k and k being
2
1
H
D
1
T
1
than the NH ion in nondeuterated liquid ammonia
comparable; k 1 > k > k . It is notable that even
2
(
Table 2), and the OD ion in D O is a stronger base
a weak internal return effect has a significant influ-
ence on the multistage mechanism of hydrogen iso-
tope exchange in benzene, decreasing the kinetic iso-
tope effect in the stage of ionization of the CH acid
2
than the OH ion in H O. Although the media being
2
compared differ significantly from each other in the
protophilic properties, the statistical thermodynamic
calculations taking into account small differences in
H
T
(
k /k ) from the theoretically calculated value of 19
1 1
the vibrational frequencies of H O and D O, as com-
to 12 [estimation by formula (6)]. There is also an-
other explanation for the decreased kinetic isotope
effect, taking into account the asymmetry of the three-
centered linear model of the transition state [15]. For
better understanding of this problem, it was desirable
to supplement the above data by studying of the sec-
ondary isotope effect of the solvent, characterizing the
energy and position of the transition state on the reac-
tion coordinate.
2
2
pared to NH and ND , give similar values of l and
3
3
D
H
K /K . Therefore, we believe that relationship (8)
eq eq
can be used to estimate the location on the reaction
coordinate of the transition state of hydrogen isotope
exchange of benzene with a solution of potassium
amide in liquid ammonia, and to evaluate the proper-
ties of this transition state. Knowing the ratio of the
rate constants of tritium exchange in a solution of
potassium amide in deuterated and light liquid
Secondary isotope effect of the solvent. Structure
and energy of the transition state of the reaction.
The goal of this part of our study was to apply the
semiquantitative interpretation of the secondary iso-
tope effect of the solvent, developed previously only
for protolytic processes in water or alcohol [7, 16, 17],
to isotope exchange of hydrogen in benzene with a
solution of potassium amide in liquid ammonia. One
of the main results of this approach was revealing a
direct correlation of the secondary isotope effect of the
ND NH
5
5
3
3
ammonia, kT /k
2
= (1.9 10 )/(0.9 10 ) =
.1 (0 C, CKNH2 0.05 M) (Table 2), we estimated by
Eq. (8) the coefficient for this reaction: = 0.85.
This value is close to that ( 0.8) obtained usually
T
for reactions in which the negative charge in the
forming carbanions is localized and its stabilization
is due to the s-orbital effect of the carbon atom of the
reaction center and the electrostatic interaction with
the other atoms of the molecule, including substitu-
ents (see, e.g., [18 22]). A characteristic feature of the
reaction in question, and also of similar reactions of
other normal CH acids, is also partial preservation of
the solvation of the reactants in the transition state,
which minimizes the contribution of the energy of
reorganization of the polar medium to the activation
energy.
solvent, defined as the ratio kT /kT , with the coeffi-
OD OH
cient in the Brønsted equation and with the maxi-
T
T
mum attaniable value of this ratio, (k /k
)
.
OD OH max
The latter, in turn, is determined as the product of the
deuterium fractionation factor in aqueous alkaline
solution l by the deuterium isotope effect manifested
H
D
in the ionic product of water, K
/K
. Then, we
water water
Variation of the primary kinetic isotope effect of
hydrogen isotope exchange of aromatic CH bonds
as influenced by the CH acidity of the substrates.
When discussing the results of this section, we used a
relatively versatile Eigen’s method for qualitative or
semiquantitative description of variation of the mech-
anism of hydrogen isotope exchange in aqueous solu-
tions as influenced by the reactant structure [6]. By
analogy with OH and NH acids, Eigen subdivided all
the CH acids into two groups: normal acids and pseu-
do-acids. The important distinctive feature of the
obtain [7]
T
T
T
T
k
/k
= (k /k
) = 2.4 .
(8)
OD OH
OD OH max
Relationship (8) was frequently used to evaluate
the extent of structural and energetic similarity be-
tween the transition state and the products of proton
transfer [7]. With data on the secondary isotope effect
of the solvent, this relationship reproduces, as a rule,
the values of that were determined from usual corre-
lation between the activation free energies (logKD)
normal CH acids is that, when logk of their reac-
and pK . In the limiting case of = 0, the CH bond
D
a
tions is correlated with pK , the so-called Eigen’s
remains intact in the transition state, whereas in the
opposite limiting case ( = 1) the transition state is
actually a carbanion.
a
curves are obtained, each of which can be approxi-
mated by linear sections with the coefficients equal
to unity and zero. In an Eigen’s curve there is a sharp
In this work we proceeded from an apparent as- break in going from the section with = 1 (relatively
sumption that the secondary isotope effect of the
solvent in the reaction of tritium exchange in a solu-
tion of potassium amide in liquid ammonia is due to
the same factor as in the similar reaction in water
narrow region where the difference pK between pKa
a
of the CH acids under consideration and pK of water
a
autoprotolysis differs from zero) to the section with
= 0 (a narrow region in the vicinity of pK = 0).
a
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

1138
TUPITSYN, ZATSEPINA
Table 4. Mean values of the rate constants of protode-
Available data for verifying the validity of the
above relationships as applied to variations of the
kinetic isotope effect of hydrogen isotope exchange in
benzene and its derivatives with solutions of amides
(NH2 or C H NH ) in nonaqueous solutions are
NH
NH
deuteration (k 3) and protodetritiation (kT 3) and the
D
NH NH
ratios kD 3/k
3
for deuterium and tritium exchange in
T
isotopically substituted fluorobenzenes in a solution of
a
sodium amide in liquid ammonia (m /m = 0.2)
Na
s
6
11
extremely scarce. Nevertheless, it is possible to ex-
plain in terms of the Eigen’s theory the data on the
kinetic isotope effect in deuterium and tritium ex-
change of a limited set of compounds (benzene and
its fluorinated derivatives) with cyclohexylamine,
catalyzed with lithium cyclohexylamide. The Brønsted
coefficient for this reaction is close to unity [23],
and hence we deal with normal CH acids. If, using
data in [24, 25], we consider the trend in variation of
k /k depending on pK or the partial rate factor
Tem-
pera-
ture,
C
NH
NH
3
3
k
k
T
D
NH3
D
k
/
5
5
fb
Compound
10 ,
10 ,
NH
3
k
T
1
1
s
s
Fluoro-
benzene-
10
0
7.6 0.2 2.9 0.1
16 2 6.3 0.2
17
2.6
3
D(t)
Fluoro-
benzene-
0
15
1.1 0.1 0.5 0.02
4.1 0.2 2.1 0.2
2.0
2.0
D
T
a
1.1
1.0
f = k (C H X)/k (C H ) (X is a substituent), we see
D 6 5 D 6 6
4
D(t)
that the kinetic isotope effect increases with decreas-
2
5
0
9.6 0.5 4.5 0.2
0.9 0.1
2.1
ing pK or with increasing logf in the following order:
a
Benzene-
D(t)
benzene-D(t) (k /k 1.6, pK 43, logf 0), fluoroben-
D
T
a
zene-4D(t) (k /k 1.9, pK 41.9, logf 0.03), fluoro-
D
T
a
benzene-3D(t) (k /k 2.0, pK 40, logf 1.95), trifluo-
D
T
a
a
Here and hereinafter, m is the number of moles of sodium
amide. (f) Factor of the partial rate of deuterium exchange.
romethylbenzene-3D(t) (k /k 2.4, no data on pK ,
Na
D T a
b
logf 2.59), and fluorobenzene-2D(t) (k /k 3.0, pK
D T a
3
7.3, logf 5.43). Streitwieser and Mares [24] have not
As for pseudo-CH acids, they exhibit a wide linear
revealed a sharp decrease in the kinetic isotope effect
in the range of pK values in which the exchange rate
is determined not by the stage of ionization of the CH
bond but by another stage insensitive to the isotope
substitution of hydrogen in the substrate. According to
the theory, this pK range should be located in the
vicinity of pK of cyclohexylamine autoprotolysis,
4
correlation between logK and pK with the coeffi-
D
a
a
cient ranging from 0 to 1. Let us briefly discuss the
mechanistic features of deuterium exchange of normal
Eigen’s CH acids. It is believed at present that varia-
tions of are due to a change in the limiting stage of
the overall reaction mechanism. In particular, when
the Brønsted coefficient is equal to unity, two stages
partially determine the reaction rate: diffusion separa-
a
a
0.3 [26].
tion of the contact complexes (k ) and ionization of
2
The influence of substituents and heteroatoms on
the CH acid (k ) [scheme (1)] [7, 22]. In contrast, the
1
the kinetic isotope effect of deuterium exchange of
aromatic CH bonds in benzene with a solution of
potassium amide in liquid ammonia was not studied at
all. Shapiro et al. [27] compared, however, the relative
kinetic acidities in this catalytic system with previous-
ly published [24] data on the relative kinetic acidity of
systems that strongly differ in the protophilic power
and are characterized by = 1. The close coincidence
of the ratios f = [k (C H X)/k (C H )] suggests that
value = 0 corresponds to the mechanism in which
formation of the primary contact complex >CH B
is a fully equilibrium process, and the rate-determin-
ing stage is separation of the primary and secondary
contact complexes.
Study in this respect of variations of the kinetic
isotope effect of various protolytic processes in water
showed that, with a normal CH acid whose pK is
a
D
6
5
D
6
6
considerably greater than that of water autoprotolysis,
the behavior of benzene and its derivatives in this re-
action corresponds to the behavior of normal Eigen’s
CH acids.
the ratio k /k takes relatively low values (1.5 <
D
T
k /k < 2.2). In going to related CH acids whose pK
D
T
a
differs from that of water autoprotolysis to a lesser
extent, the ratio k /k gradually increases, with the
To find whether the trends in variation of the kinet-
ic isotope effect in hydrogen isotope exchange of
aromatic CH bonds with a solution of potassium (or
sodium) amide in liquid ammonia are consistent with
the Eigen’s theory developed for aqueous solutions,
we measured the variations of the isotope effect of the
reaction in benzene as influenced by the fluorine
substituent and its position in the ring. When studying
D
T
ionization of the CH acids remaining a partially limit-
ing stage. Fiinally, if pK of the CH acid is close to
a
that of the conjugated acid of the reacting base, the
ratio k /k decreases jumpwise, becoming close to
D
T
unity (1 < k /k < 1.1). In the latter case the limiting
D
T
stage is the diffusion-controlled transition from the
primary to the secondary contact complex.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

KINETIC ISOTOPE EFFECTS IN HYDROGEN EXCHANGE
1139
the hydrogen isotope exchange of fluorobenzene-3D(t) conditions when the exchange kinetics of the other
and fluorobenzene-4D(t), which is very fast, we used
as catalyst not potassium amide but less soluble (solu-
bility 0.012 M at 20 C) sodium amide. The results are
listed in Table 4. Their comparison with the above
data shows that the kinetic isotope effects of the hy-
compounds of the reaction series could not be studied
by common procedures. Furthermore, to evaluate the
kinetic isotope effect, it was necessary to estimate
preliminarily what part of the dissolved alkali metal
amide can be spent in side reactions. It was found in
drogen isotope exchange in benzene and deutero deriv- [28] that the color in the reaction mixtures originates
atives of fluorobenzene with solutions of alkali metal
amide in liquid ammonia and of lithium cyclohexyl-
from reversible addition of amide ions to the substrate
molecules. To elucidate the influence of this side
process on the kinetics of isotope exchange, we per-
formed special experiments involving simultaneous
measurement of the rate constants for the heterocycle
being tested and an indicator compound chemically
inert toward amide ions under the reaction conditions.
As such indicator we chose fluorobenzene-3D. The
conditions and results of experiments on concurrent
exchange of pyridine-2D and pyridine-3D with the in-
dicator are listed in Table 5. Comparison of Tables 4
and 5 shows that the exchange rate constants of both
CH acids are equal, within the experimental error, to
the values obtained for these compounds in control
experiments under the same conditions. Thus, the
potentially possible addition of amide ion to the aro-
matic ring has no effect on the kinetics of deuterium
exchange in pyridine.
amide in cyclohexylamine vary similarly: (k /k )
>
D
T ortho
(
k /k )
> (k /k )
(k /k )
. This fact
T C H
6
D
T meta
D
T para
D
6
seems natural, because with respect to the protophilic
properties the latter catalytic system is close to solu-
tions of amide ions in liquid ammonia. A note should
be made, however, about the isotope exchange of the
o-hydrogen atoms in fluorobenzene. The rate of this
reaction in the system NH NH (liquid) is too high,
2
3
NH NH
3
3
and we could not determine the ratio (kD /k )ortho
.
T
As seen from Tables 1 and 4, in both reaction series
NH NH
3
3
the ratio kD /k
tends to increase with increasing
T
relative kinetic acidity characterized by the factor f.
There can be two reasons for the observed increase in
the kinetic isotope effect: influence of F either on the
mechanism of isotope exchange, decreasing the inter-
nal return, or on the structure of the transition
state. To choose between these two alternatives, we
used the models involving the basic physics of proton
transfer in solution [18]. Indeed, previously the set
of data on this problem were obtained with these mo-
dels according to which the constant k 1 [scheme (1)]
is mainly related to the effective (statistical mean)
vibrational frequency of the nuclei in the solvent and
reactant molecules ( eff). Since the solvent molecules
prevail, they make the major contribution to eff.
Therefore, for protolytic processes in the same catalyt-
ic system, k 1 is practically the same for different
reactions. Thus, we can conclude that the above-noted
influence of the strength of the CH acids on the kinet-
ic isotope effect is mainly due to the fact that the
extremely asymmetrical structure of the transition
state in the case of benzene becomes less asymmetri-
cal in going to the reactions of m- and o-H atoms in
fluorobenzene.
An opposite result was obtained in concurrent ex-
change in mixtures of pyridine-2D N-oxide and ben-
zene-D (indicator) and of pyridine-2D N-oxide and
naphthalene-1D (indicator). The reaction of the indica-
tor hydrocarbon under the reaction conditions (25 C,
CKNH2 0.1 M) is fully suppressed, whereas in the
N-oxide the deuterium exchange is so fast that its
kinetics cannot be measured. This fact indicates that
under the reaction conditions potassium amide strong-
ly interacts with the molecules of the aromatic hetero-
cycle and that pyridine N-oxide is capable of active
hydrogen isotope exchange with the solvent even in
the presence of very low concentrations of amide ions.
The above experiments have not allowed determi-
nation of the true concentration of the NH ions in
2
solution. To estimate, at least approximately, the
factor of the partial exchange rate, we compared the
exchange kinetics in benzene-D and pyridine-2D
N-oxide. In the latter compound the rate of deuterium
exchange could be measured by common procedures
only at low temperature ( 37 C, Table 6). For deuteri-
Additional information on the mechanistic features
of hydrogen isotope exchange of aromatic CH bonds
can be, generally speaking, derived from the compari-
son of the influence of heteroatoms (N, NO) in the
aromatic ring on the kinetic CH acidity and kinetic
isotope effect with that of fluorine substituent. How-
ever, such a comparison involves methodical prob-
lems. First, the rates of hydrogen isotope exchange in
pyridine-3D(t) and pyridine-2D(t) N-oxide were kinet-
ically measurable only at low temperatures (from 50
to 25 C) in the presence of sodium amide, i.e., under
NH
3
um exchange in benzene-D as reference, kD was
determined by calculation from the experimental acti-
1
vation energy E 15.7 0.2 kcal mol and the preex-
a
ponential factor logA 7.8. For the exchange rate con-
NH
D
3
stant of benzene we obtained thus the value k
3
7
1
10 s , and for pyridine-2D N-oxide the exchange
factor f was greater than 4300.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

1140
TUPITSYN, ZATSEPINA
Table 5. Conditions and results of experiments on protodedeuteration of a mixture of pyridine-D with fluorobenzene-3D
1
a
(
m : m = 1 : 1) in a solution of sodium amide in liquid ammonia
s
i
i
b
h b
Tempe- Experiment
I ,
%
I ,
%
rel
NH3
D
105,
rel
NH3
D
105,
Run
no.
m /(m +
k
k
a
s
rature,
C
duration,
h
mi)
s 1
s 1
96
97
98
79
80
81
1
2
3
4
0
2.0
3.0
1.5
3.0
102.2
96.7
98.0
89.9
79.9
86.8
22.5
27.5
20.1
12.9
77.5
72.5
15.1
13.7
1.4
42.2
47.7
75.0
90.5
54.3
50.1
24.8
9.5
3.5
2.2
0.1
2.1
1.6
26
22
0.3
25
1.3
a
(
m ) Number of moles of the indicator compound. For run nos. 1 and 2: the heterocycle is pyridine-2D (initial relative intensity of
i
the peaks in the mass spectrum: 79, 37.0; 80, 61.0; 81, 2.0) and the indicator compound is fluorobenzene-3D (initial relative intensi-
ty of the peaks: 96, 40.7; 97, 59.2; 98, 0.1); for run nos. 3 and 4: heterocycle pyridine-3D (79, 4.9; 80, 90.5; 81, 4.6) and indicator
b
i
h
fluorobenzene-3D (96, 16.4; 97, 85.5; 98, 0.1). (I , I ) Relative intensities of the peaks of the indicator compound and hetero-
rel rel
cycle in the mass spectrum, respectively.
Table 6. Conditions and results of experiments on protodedeuteration (k ND H3) of pyridine-2D N-oxide in a solution of
sodium amide in liquid ammonia ( 37 C)
Relative intensity of indicated peak
Run
no.
Experiment
duration, h
in mass spectrum, %
kD 105,
m /m
m /m
Na s
a
s
s 1
9
5
96
97
98
Initial sample
86
129
15.4
52.6
71.8
79.6
46.9
21.8
3.6
0.5
6.3
1.4
1
2
0.2
0.25
0.18
0.22
1100
1400
To estimate the partial rate factors of deuterium
exchange in quinolines-2D and -3D, it was appropri-
2.2 10 4 M. Based on these data on the true concen-
tration of amide ions and on the results of comparative
ate to compare the exchange rates of quinolines-6D kinetic studies of deuterum exchange in quinoline-2D,
and -8D with those of naphthalenes-1D and -2D.
These results are listed in Table 7. It seems from their
comparison that the cyclic nitrogen atom strongly
deactivates the 6 and 8 positions of the ring with
respect to the hydrogen exchange of the aromatic CH
bond: In going from naphthalene-1D to quinoline-8D
quinoline-3D, and naphthalene-2D, we calculated the
relative exchange rates of heterocyclic compounds,
with the rate of hydrogen isotope exchange in benzene
at the same temperature and the same concentration
of amide ions taken as unity (Table 7). The average
rate constants of deuterium and tritium exchange in
some aromatic nitrogen-containing heterocycles, the
corresponding partial rate factors, and the kinetic iso-
tope effects required for subsequent discussion are
given in Table 8.
(or from naphthalene-2D to quinoline-6D), the rate
constant decreases by 2 3 orders of magnitude. How-
ever, actually this is not the case. According to pub-
lished data, the pyridine ring of quinoline has a very
weak effect on the reactivity of the annelated benzene
ring [29]. The drastic decrease in the deuterium ex-
change rate is due to formation of addition complexes
of quinoline with amide ions in solution, resulting in
disappearance of free amide ions. Assuming that the
exchange in the 6 position of quinoline should occur
at almost the same rate as in the 2 position of naph-
thalene and that the rate of deuterium exchange is
approximately proportional to the catalyst concentra-
However, before discussing data in Table 8, we
should examine the possibility of constructing the
general series including the deutero derivatives of
both fluorobenzene and heterocycles, as the structural
similarity of the compounds of these classes is not
evident. We believe that the general consideration of
the electronic interactions in substituted benzenes and
heteroaromatic compounds can be based on the tradi-
NH
3
tion, we found that actually to kD of quinoline-6D
corresponds the concentration of amide ions CKNH2
tional
analysis. In terms of this approach, various
types of electronic effects were distinguished and vari-
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

KINETIC ISOTOPE EFFECTS IN HYDROGEN EXCHANGE
1141
ous constants were obtained for the pyridine nitro-
gen atom and the N-oxide group as local structural
elements of the molecules [30]. Previously [31 33],
using these constants, the kinetic data on hydrogen
isotope exchange in substituted benzenes and the
heterocycles under consideration with alkali metal
amide in liquid ammonia were treated by correlation
analysis. In particular, a correlation was constructed
between the partial rate factors of o-substituents in the
Table 7. Average rate constants of protodedeuteration
NH
(k 3) of monodeutero-substituted naphthalenes and quino-
D
lines in a solution of potassium amide in liquid ammonia
(CKNH2 0.1 M)a
Temperature,
NH
5
1
3
Compound
k
10 , s
D
C
Naphthalene-1D
0
10
17 2
55 4
benzene rings and the
constants [32, 33]:
I
2
5
0
200 1
5 0.6
log fortho = 12.3 I
0.37; r 0.991, s 0.3, n 8. (9)
Naphthalene-2D
The influence of the N-oxide group on the acidity
of the adjacent aromatic CH bond follows relation (9)
with the induction constant of this group I(NO) 1.0,
following the conclusions in [30]. The correlation of
12
25
25
0
25
25
40
25
21 3
100 8
1.9 0.2
b
c
Quinoline-2D
Quinoline-3D
1.6 0.2
logfmeta with the
constants is somewhat worse:
20 3c
meta
Quinoline-8D
Quinoline-6D
0.75 0.02
3.5 0.4
0.22 0.01
15 2
log fmeta = 8.85 meta 0.16; r 0.964, s 0.4, n 7.
(10)
The point for pyridine-3D lies on the regression
line if the action of the induction effect of the N atom
is characterized by the value meta(3 N=) = 0.45 [30].
4
0
a
b
NH
5
6 runs at each temperature. The true constant k
0.1 M, 25 C is 8.6 10
D the factor f was taken equal to 6. The true constant k
3
at
D
Note that the choice of the
and
constants for
I
meta
3
1
C
s
, f 50; for naphthalene-
KNH2
analysis of this reaction series is governed by the fact
that the negative charge arising on the hydrogen atom
in the transition state cannot efficiently interact with
the system, and the resonance effects appear to be
much less important than the inductive effect.
c
NH
2
3
D
3
1
at C
0.1 M, 0 C is 7.2 10
; f 840 and 2700, respectively.
s
, and at 25 C, 9
KNH
2
2
1
s
10
lated, with aromatic heterocycles this is not always the
case. For example, although the accelerating effect of
Since substituted benzenes and aromatic nitrogen-
containing heterocycles can be covered by general replacement of the 3-CH bond in the benzene molecule
correlations only when the reaction mechanism does
not change cardinally in going from one compound to
another, it is appropriate to discuss the kinetic isotope
effect in both groups of compounds from a common
viewpoint. In this connection, it should be noted first
that, whereas variations of the factor f and kinetic
isotope effect in deuterated fluorobenzenes are corre-
by a nitrogen atom (Table 8) is significant (f 120), the
number and accuracy of measurements are insufficient
to reliably differentiate the kinetic isotope effect of
hydrogen exchange in both compounds (1.9 and 2.1,
respectively). With hydrogen exchange in quinoline-
3D(t), the pattern is somewhat different. The kinetic
isotope effect increases to a considerably greater ex-
Table 8. Average rate constants of protodedeuteration (k ND H3), protodetritiation (k TN H3), and the ratio k DN H3/kT for deute-
NH
3
rium and tritium exchange of some aromatic nitrogen-containing heterocycles in a solution of alkali metal amide in liquid
a
ammonia
NH
105,
NH
105,
NH
NH
3
3
3
3/k
D T
Compound
Tempera-
ture,
Alkali metal
amide
k
k
f
k
D
T
C
s 1
s 1
Pyridine-2D(t)
0
NaNH2
2.1 0.1
4.7 0.2
3.7 0.1
1.5 0.1
4.0 0.1
1.8 0.2
2.2
1.3
1.2
2.1
1.1
1.4
2.7
10
Pyridine-3D(t)
37
37
40
15
NaNH2
NaNH2
KNH2
KNH2
120
4300
Pyridine-2D(t) N-oxide
Quinoline-2D(t)
Quinoline-3D(t)
1400 70
1300 70
26 2b
18 2b
50
8.1 0.6b
3.0 0.1b
1800 900
a
b
5
6 runs at each temperature. Apparent rate constants, without taking into account the loss of amide ions from solution as a result
of interaction.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

1142
TUPITSYN, ZATSEPINA
tent as compared to benzene or pyridine-3D(t). An
increase in f by a factor of 1800 is accompanied by
the same increase in the ratio kD /kT as that ob-
is totally lacking (k ND H3/k
NH
T
1.1) (Table 8). The
3
fact that introduction of the NO group considerably
enhances the CH acidity and simultaneously drastical-
ly decreases the kinetic isotope effect strongly sug-
gests that the above anomaly is due to the Eigen’s
NH NH
3
3
served in going from benzene-D(t) to fluorobenzene-
3D(t) (f 17). Despite the lower sensitivity of the kinet-
ic isotope effect to increase in the rate in the case of
pyridine-3D(t) and quinoline-3D(t), we conclude that,
condition according to which in the point pK the
isotope effect is close to unity. To check this assump-
a
similar to 3-F substituent in benzene, incorporation of tion, it is necessary, however, to estimate pK of pyri-
a
the cyclic nitrogen atom into the aromatic system of
the ring (1 position) does not alter the mechanism of
hydrogen isotope exchange in quinoline-3D but only
shifts the position of the transition state at the same
reaction mechanism.
dine-2D N-oxide as CH acid. The difficulty of deter-
mining this quantity is due to the lack of reliable esti-
mates for the factor f of deuterium exchange in pyri-
dine-2D(t) N-oxide (see above). Therefore, we re-
stricted our task to estimating the order of magnitude
of pK rather than determining its value. For this
a
Certain problems arise in interpretation of the
purpose, we applied to the case of the acid base equi-
librium in liquid ammonia the approach used previ-
abnormal decreases in both the rate and the ratio
NH NH
3
3
k
/k
for hydrogen isotope exchange in pyridine-
D
T
ously [23, 24] for calculating pK of mono- and poly-
a
2
D(t) and quinoline-2D(t) (Table 8). The factors f and
NH NH
fluorinated benzenes in a solution of cesium cyclo-
hexylamide in cyclohexylamine. Streitwieser et al.
3
3
the ratios kD /k
1.1 1.4 for these compounds
deviate from data obtained for the other examined
compounds. As the constant of the cyclic nitrogen
T
[
23, 24] proceeded from the fact that the coefficient
I
of the Brønsted relationship in this catalytic system
atom is relatively large ( I 0.45 [30]), the behavior of
pyridine-2D(t) in hydrogen isotope exchange cannot
be described in terms of the existing views on the
electron-acceptor inductive effect of nitrogen on the
is 1, and pK of the parent compound, benzene, is 43.
a
This fact allowed calculation of pK of each particular
a
compound belonging to the examined reaction series
using a simple formula pK (C H X) = pK (C H )
a
6
5
a
6
6
adjacent reaction center. Nevertheless, the low values
NH NH
logf. Since the relative kinetic acidities of the aromat-
ic CH bonds in benzene and its derivatives do not
change in going from the above system to a solution
of potassium amide in liquid ammonia [27], we esti-
3
3
of f and kD /k , which are difficult to explain in
T
terms of a purely inductive effect, can be interpreted
assuming that in the examined compounds the kinetic
isotope effect in the course of rearrangement of the
intramolecular structure is determined by superposi-
tion of the inductive effect and an effect of the oppo-
site sign originating from direct interaction of the lone
electron pairs of the nitrogen atom and the anionic
carbon atom in the adjacent position of the ring [30].
mated the upper level of pK of pyridine-2D N-oxide:
a
pKa 38. The latter value is close to pK of autopro-
a
tolysis of liquid ammonia (pK 35). Hence, in the case
a
of pyridine-2D N-oxide the reaction mechanism can
be different, with the diffusion-controlled separation
of the primary and secondary contact complexes (rate
The latter effect is manifested not only in the kinetics
NH NH
constant k ) becoming the limiting stage.
3
3
2
of deuterium exchange or in the kD /kT values but
also in variations of various physical quantities [inten-
sities of stretching vibrations of aromatic CH
Thus, in this work, with quite a different kinetic
material as compared to [7, 20 22] and with a strong-
ly protophilic medium, we confirmed the applicability
of the Eigen’s theory to the mechanism of hydrogen
isotope exchange of aromatic CH bonds.
3
5
bonds in the IR spectra, resonance frequencies ( Cl)
in the NQR spectra, etc.], i.e., in cases when the indi-
cator center lies in the aromatic ring plane. Judging
from the partial rate factor of deuterium exchange in
quinoline-2D, the effect of the lone electron pair is
considerably weakened relative to that in pyridine-2D1
EXPERIMENTAL
(
Table 8). Nevertheless, the kinetic isotope effect in
Manipulations with solutions of alkali metal
amides in liquid ammonia were performed in special
apparatus resistant to high pressures and preventing
access of moisture and oxygen. Commercial deute-
rated organic compounds were used; they were puri-
fied and dried before experiments. Tritium-labeled
deuterated compounds were prepared via the corre-
sponding organomagnesium derivatives which were
hydrolyzed with deuterium and tritium oxide. Also,
the majority of nitrogen-containing heterocycles-D,t
this reaction remains low as compared to the hydrogen
exchange in benzene.
As already noted, the regular increase in the rate of
the hydrogen isotope exchange in pyridine-2D(t) N-
oxide exhibiting the highest constant among all the
examined CH acids confirms the assumed common
induction mechanism of activation of the reaction
center. At the same time, the isotope effect in isotope
exchange with pyridine-2D(t) N-oxide is the lowest or
I
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001

KINETIC ISOTOPE EFFECTS IN HYDROGEN EXCHANGE
1143
were prepared from the corresponding halo derivatives
by reduction with zinc dust in deutero(tritio)sulfuric
acid. Quinoline-2D(t) was prepared by decarboxyla-
tion of quinaldic acid. Anhydrous liquid ammonia and
deuteroammonia (94 99 at. %) were used.
(Thermal Properties of Ammonia: Handbook),
Moscow: Izd. Standartov, 1978, p. 143.
14. Tupitsyn, I.F., Semenova, N.K., and Avdulov, G.I., in
Radioisotopes in the Physical Sciences and Industry,
Vienna: IAEA, 1962, pp. 255 257.
1
1
1
1
5. Westheimer, F.H., Chem. Rev., 1961, vol. 61, no. 3,
pp. 265 273.
6. Steffa, L.J. and Thornton, E.R., J. Am. Chem. Soc.,
The rates of deuterium and tritium exchange were
measured in solutions containing usually 100 150 mol
of a solvent per mole of an organic compound. The
deuterium concentration was determined with an MI-
1
967, vol. 89, no. 23, pp. 6149 6156.
7. Gold, V. and Grist, S., J. Chem. Soc., Perkin Trans. 2,
972, no. 1, pp. 89 95.
1201 mass spectrometer at an ionizing electron energy
1
of 12 14 eV. The tritium activity was measured using
a scintillation counter with two coaxial photomultipli-
ers operating in the coincidence mode. The scintilla-
tion procedure was used in two versions: (a) samples
of organic compound before and after experiments
were introduced directly into the scintillation solution;
8. Shapiro, I.O., Fizicheskaya khimiya. Sovremennye
problemy (Physical Chemistry. Modern Problems),
Moscow: Khimiya, 1987, pp. 129 164.
9. Lin, A.C., Chang, A.C., Dahlberg, Y., and Kresge, A.J.,
J. Am. Chem. Soc., 1983, vol. 105, no. 16, pp. 5380
1
2
5
386.
(
b) tritium-labeled ammonia was burned over CuO,
0. Aroella, T., Arrowsmith, C.H., Hojatti, M., and
Kresge, A.J., J. Am. Chem. Soc., 1987, vol. 109,
no. 23, pp. 7198 7199.
and the resulting water was dissolved in the scintil-
lator. The measurement conditions were strictly fixed.
The rate constants of deuterium and tritium ex-
change, calculated by a first-order equation, had
stable values in all the systems studied.
21. Kresge, A.J. and Powell, M.F., J. Org. Chem., 1986,
vol. 51, no. 6, pp. 819 822.
22. Argyrou, A. and Washabaugh, N.W., J. Am. Chem.
Soc., 1999, vol. 121, no. 51, pp. 12054 12062.
23. Streitwieser, A., Scannon, P.J., and Niemeyer, K.,
J. Am. Chem. Soc., 1972, vol. 94, no. 22, pp. 7936
7937.
24. Streitwieser, A. and Mares, F., J. Am. Chem. Soc.,
1968, vol. 90, no. 3, pp. 644 648.
25. Stratakis, M., Wang, P.J., and Streitwieser, A., J. Org.
Chem., 1996, vol. 61, no. 9, pp. 3145 3150.
26. Streitwieser, A. and Guibe, B., J. Am. Chem. Soc.,
1978, vol. 100, no. 14, pp. 4523 4534.
REFERENCES
1
2
3
. Malykhin, E.V. and Shteingarts, V.D., Zh. Ross. Khim.
O va, 1999, vol. 43, no. 1, pp. 49 56.
. Huisgen, R., Mack, W., Herbig, K., and Ott, N.,
Chem. Ber., 1960, vol. 93, no. 2, pp. 412 424.
. Hertog, H.J., Pieterse, M.J., and Burman, D., Recl.
Trav. Chim. Pays-Bas, 1963, vol. 82, no. 12,
pp. 1173 1178.
4
5
6
7
8
. Martens, P.J. and Hertog, H.J., Recl. Trav. Chim.
Pays-Bas, 1964, vol. 83, no. 6, pp. 621 630.
. Illuminatti, N., Adv. Heterocycl. Chem., 1965, vol. 4,
pp. 285 314.
. Eigen, M., Angew. Chem. Int. Ed. Engl., 1964, vol. 3,
no. 1, pp. 1 19.
. Washabaugh, M.W. and Jencks, W.P., J. Am. Chem.
Soc., 1989, vol. 111, no. 2, pp. 683 692.
. Shatenshtein, A.I., Yakushin, F.S., and Arshino-
va, M.I., Kinet. Katal., 1964, vol. 5, no. 6, pp. 1000
27. Shapiro, I.O., Ranneva, Yu.I., and Shatenshtein, A.I.,
Zh. Obshch. Khim., 1979, vol. 49, no. 9, pp. 2030
2036.
28. Kane, A.A. and Tupitsyn, I.F., in Khimiya i tekhno-
logiya izotopov (Chemistry and Technology of Iso-
topes), Leningrad: Khimiya, 1967, pp. 232 236.
29. Heterocyclic Compounds, Elderfield, R.C., Ed., New
York: Chapman and Hall, 1950. Translated under the
title Geterotsiklicheskie soedineniya, Moscow: Ino-
strannaya Literatura, 1955, vol. 4, pp. 83 96.
30. Zatsepina, I.F., Kirova, A.V., and Tupitsyn, I.F.,
Reakts. Sposobn. Org. Soedin., 1968, vol. 5, no. 1,
pp. 70 87.
1
007.
9
. Streitwieser, A., Hollyhead, W.B., Pudjaatmaka, A.H.,
and Owens, P.H., J. Am. Chem. Soc., 1971, vol. 93,
no. 20, pp. 5088 5096.
31. Tupitsyn, I.F., Zatsepina, N.N., Kirova, A.V., and
Kapustin, Yu.M., Reakts. Sposobn. Org. Soedin., 1968,
vol. 5, no. 3, pp. 601 611.
1
1
1
0. Thibbin, A., Onicido, I., and Ahlberg, P., Chem.
Scripta, 1982, vol. 19, no. 3, pp. 145 148.
1. Shapiro, I.O., Ranneva, Yu.I., and Shatenshtein, A.I.,
Zh. Org. Khim., 1986, vol. 22, no. 6, pp. 1332 1333.
2. Bell, R.P., The Proton in Chemistry, Ithaca: Cornell
Univ. Press, 1973, 2nd ed. Translated under the title
Proton v khimii, Moscow: Mir, 1974, pp. 291, 297.
3. Teplofizicheskie svoistva ammiaka: Spravochnik
32. Ranneva, Yu.I. and Shatenshtein, A.I., in Korrelya-
tsionnye uravneniya v organicheskoi khimii (Correla-
tion Equations in Organic Chemistry), Tartu: Tartusk.
Gos. Univ., 1962, vol. 1, pp. 136 149.
33. Kamrad, A.G., Reakts. Sposobn. Org. Soedin., 1968,
vol. 5, no. 3, pp. 701 710.
1
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 71 No. 7 2001