Kinetics and mechanism of monomolecular heterolysis of commercial organohalogen compounds: XXXII. Solvent effects on activation parameters of heterolysis of 1-chloro-1-methylcyclopentane. Correlation analysis of solvation effects

Article abstract of DOI:10.1023/A:1023446808929

Kinetics of heterolysis of 1-chloro-1-methylcyclopentane in MeOH, BuOH, cyclohexane, i-PrOH, t-BuOH, tert-C₅H₁₁OH, γ-butyrolactone, MeCN, PhCN, PhNO₂, acetone, PhCOMe, cyclohexanone, and 1,2-dichloroethane at 25-50°C were studied by the verdazyl method. Correlation analysis of solvent effects on activation parameters of the reaction in 8 protic (additionally, AcOH and CF ₃CH₂OH) and 8 aprotic solvents together and separately in either group of solvents was performed. In all the solvents studied, two ΔH^≠-ΔS^≠ compensation effects were revealed.

Full text of DOI:10.1023/A:1023446808929

Russian Journal of General Chemistry, Vol. 72, No. 12, 2002, pp. 1882 1893. Translated from Zhurnal Obshchei Khimii, Vol. 72, No. 12, 2002,
pp. 1989 2001.
Original Russian Text Copyright
2002 by Dvorko, Koshchii, Prokopets, Ponomareva.
Kinetics and Mechanism of Monomolecular Heterolysis
of Commercial Organohalogen Compounds:
XXXII.1 Solvent Effects on Activation Parameters
of Heterolysis of 1-Chloro-1-methylcyclopentane.
Correlation Analysis of Solvation Effects
G. F. Dvorko, I. V. Koshchii, A. M. Prokopets, and E. A. Ponomareva
Kievskii politekhnicheskii institut National Technical University, Kiev, Ukraine
Received September 5, 2000
Abstract Kinetics of heterolysis of 1-chloro-1-methylcyclopentane in MeOH, BuOH, cyclohexane, i-PrOH,
t-BuOH, tert-C5H11OH, -butyrolactone, MeCN, PhCN, PhNO2, acetone, PhCOMe, cyclohexanone, and
1,2-dichloroethane at 25 50 C were studied by the verdazyl method. Correlation analysis of solvent effects
on activation parameters of the reaction in 8 protic (additionally, AcOH and CF3CH2OH) and 8 aprotic sol-
vents together and separately in either group of solvents was performed. In all the solvents studied, two
H
S
compensation effects were revealed.
The effect of a solvent on the rate of a chemical
process much depends on the mechanism of the reac-
tion [2] and physicochemical properties of the solvent,
such as polarity, polarizability, electrophilicity, nucleo-
philicity [3], and cohesion [4]. Solvent effects are
most frequently discussed by comparing the lo-
garithms of the rate constants and the free activation
energies of the reaction. Correlation of these values
with solvent parameters in terms of multiparameter
linear free energy equations gives insight into the
nature of the solvent effects [3, 5 9]. However, the
diagnostic power of such correlations is limited by the
sensitive to solvent, and S is mostly contributed by
solvation effects.
The rate of these reactions is determined by ioniza-
tion of the covalent bond [2, 3], which occurs via
consecutive formation of three ion pairs: contact,
loose, and solvent-separated [9, 16].
RX
R+X
R+
Reaction products.
X
R+ Solv X
It is assumed that the limiting stage involves inter-
action of the contact ion pair with solvent void [16]
(voids comprise 10% of the volume of a liquid
[17, 18]) to form the loose ion pair. The latter fast
converts into the solvent-separated ion pair which then
passes, also fast, into reaction products.
fact
temperature-dependent. At the same time,
sum function of the activation enthalpy and entropy
( G = H H S ), which are almost temperature-
that
logk
and
G
are
strongly
G
is a
independent in a given experimental range.
There has been little work on solvent effects on
In [10], we found G , H , and S values for
heterolysis of t-BuCl (E1 + SN1 reaction) in 8 protic
activation parameters, specifically on
H and S .
We can mention here the works on heterolysis of
t-BuCl [10, 11], alcoholysis of -ethylacryloyl chlo-
ride [12] and 3-phenylpropyl p-toluenesulfonate [13],
acylation of aniline [14], and homolysis of polymeric
peroxide of azelaic acid [15].
and 15 aprotic solvents and revealed no H
0.554) and G S (R 0.028) compensation effects.
Correlation analysis of solvation effects on the transi-
tion state showed that G Tr decreases with increasing
solvent dipolarity (polarity + polarizability), electro-
philicity, and nucleophilicity and increases with sol-
S (R
Of particular interest in terms of solvent effects are
monomolecular heterolysis reactions (SN1, E1, sol-
volysis) [5, 10, 11, 16], since their rate is highly
vent cohesion (R 0.91), while
crease with all the four parameters (R 0.81 andT0r .89,
respectively). The fact that correlates with the
H
and
S
in-
Tr
G
Tr
nucleophility parameter suggests nucleophilic solvent
assistance. According to [10], all kinds of solvation,
1
For communication XXXI, see [1].
1070-3632/02/7212-1882$27.00 2002 MAIK Nauka/Interperiodica

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
1883
both specific and nonspecific, increase H , thus
decreasing the reaction rate. The same factors increase
S and thus increase the reaction rate. Consequently,
the reaction rate is increased exclusively by the in-
crease in the activation entropy, and this effect is so
strong that it fully compensates for the increase in the
activation enthalpy.
thus increasing
effects on
parately for protic and aprotic solvents.
S
and
[17, 18]. Therefore, solvent
should be considered se-
H
S
A different pattern of solvent effects on
H
is
observed in alcoholyses of -ethylacryloyl chloride
(14 alcohols) [12] and 3-phenylpropyl p-toluene-
sulfonate (17 alcohols) [13]; moreover, different
patterns are characteristic of the two substrates. In the
first case, polarity and nucleophilicity increase H ,
whereas polarizabillity, electrophilicity, and cohesion
exert the opposite effect. In the second case,
decreases with increasing solvent nucleophilicity,
electrophilicity, and polarizability.
In [11], no
H
S and
G
S compensation
effects were revealed in heterolysis of t-BuCl in
15 protic and 16 aprotic solvents. However, in dipolar
H
aprotic solvents, the
H
S
compensation effect
takes place, while in protic, two such effects. It was
found that in all the solvents studied and in aprotic
solvents,
G decreases with increasing solvent po-
In benzoylation of aniline (17 aprotic solvents),
polarizability, electrophilicity, and nucleophilicity
decrease H . In acylation of aniline with carborane-
carbonyl chloride (20 aprotic solvents), nucleo-
philicity and cohesion decrease H , while polariza-
bility increase it [14]. In the first case, polarizability,
electrophilicity, and nucleophilicity decrease S ,
whereas polarity and cohesion increase it. In the
second case, polarity and electrophilicity increase S ,
and nucleophilicity and cohesion decrease it.
larity, polarizability, and electrophilicity or ionizing
ability (R > 0.95). In protic solvents, G decreases
with increasing ionizing ability (or electrophilicity)
and cohesion and increases with increasing nucleo-
phicility of the solvent (R > 0.95). The
H value
decreases with increasing polarity and polarizability
and increases with nucleophilicity (R 0.90) of protic
solvents and decreases with increasing polarity and
increases with increasing polarizability and electro-
philicity of aprotic solvents (R 0.92). The S value
in aprotic solvents increases with increasing nucleo-
philicty and electrophilicity and decreases with increas-
ing polarizability of the solvent (R 0.92).
The H and S values of each of the reactions
in question differently vary with solvent parameters,
both in protic and aprotic solvents. Thus, in solvolysis
of -ethylacryloyl chloride, electrophilicity decreases
H , while nucleophilicity increases it. In solvolysis
of 3-phenylpropyl p-toluenesulfonate, both these
parameters decrease H , whereas in solvolysis of
t-BuCl, electrophilicity has no effect on H , and
nucleophilicity decreases it. In the first case, solvent
polarity increases H , in the second, has no effect,
and in the third, decreases it. The only solvent para-
meter that exerts the same effect in all the three reac-
tions is polarizability: It operates to decrease H .
An even more intricate pattern is characteristic of
aprotic solvents.
Consequently, according to [11], nucleophilic
solvent assistance in heterolysis of t-BuCl is absent,
and, moreover, a negative effect of nucleophilic solva-
tion is observed in protic solvents, on account of the
solvent effect on H . Specific solvation effects that
control orientation of solvent molecules with respect
to substrate molecules increase S and increase the
reaction rate.
Thus, Abraham et al. [10] and Dvorko et al. [11]
have come to different conclusions as to the solvent
effect on the activation parameters of heterolysis of
t-BuCl. The latter referees explain this controversy by
the fact that Abraham et al. [10] used experimental
activation parameters for only 14 of 23 solvents, and
the H and S values for the other 9 solvents, as
well as certain G values were estimated by compar-
ing with data for t-BuBr and t-BuI. Moreover, the
At present we still have insufficient experimental
evidence for interpreting such a diversity of solvation
effects. However, it is already obvious that the solvent
effect on H and S is more specific than on G .
Proceeding with research into solvent effects on
the rate of heterolysis 1-halo-1-methylcycloalkanes,
we dwelt on the temperature effect on the rate of
heterolysis of 1-chloro-1-methylcyclopentane (I) in 6
protic (MeOH, BuOH, i-PrOH, t-BuOH, tert-
C5H11OH, cyclohexanol) and 8 aprotic solvents
( -butyrolactone, MeCN, PhCN, PhNO2, acetone,
PhCOMe, cyclohexanone, 1,2-dichloroethane). Cor-
relation analysis of the effect of solvent parameters
on the activation parameters of the reaction was
analysis of solvent effects on
H
and
S
in that
work included the whole set of solvents only. This is
incorrect because of the different effects of protic and
aprotic solvents on the activation entropy. Origination
of an ion pair in a heterolysis reaction is accompanied
by solvent structuring around the intermediate (elec-
trostriction effect), which decreases S . In the
strongly structured protic solvents, breakdown of
solvent structure occurs along with electrostriction,
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

1884
DVORKO et al.
v = k[I].
(3)
15
14
16
(a)
13 12 11
12
Kinetic experiments we performed by the verdazyl
method [20] using 1,3,5-triphenylverdazyl (II) as
internal indicator. The stoichiometric equation of the
reaction is as follows:
(b)
(d)
14 4 6 10
9
11
10
2 3
8
7
1
13
15
12
2
3
6
5
11
9
8
7
20
Ph
4
1
16
Me
9
N
N
N
N
+ 2
Ph
(c)
Cl
Ph
8
7
(4)
6
5
I
II
Ph
NH
Ph
0
30
60
90 120
150 180
Me
N
N
N+
N
N
Cl
S , J mol 1 K
1
+
Ph
+
N
N
Ph
Ph
Ph
III
IV
(a, b)
G
S
and (c, d)
H
S
dependences
(for solvent numbering, see Table 2).
It is assumed that verdazyl II rapidly and quanti-
tatively reacts with the solvent-separated ion pair of
the substrate, which allows one to follows the reaction
rate to be followed spectrophotometrically by the
performed by the Koppel Palm equation (1) [3, 5]
augmented with the cohesion energy density para-
2
meter
[4], as well as by Eq. (2).
decrease in the indicator concentration (
720 nm)
max
n2
1
1
[16, 20]. First verdazylium salt III and alkyl deriva-
tive of verdazyl II are formed, and the latter then
rapidly decomposes into 1-methylcyclopentene and
leucoverdazyl (IV). In protic solvents, chloride I
undergoes partial solvolysys. Therewith, HCl evolves
and rapidly and quantitatively reacts with verdazyl
to form compounds III and IV [20]. Therewith, the
stoichiometric relations in reaction (4) do not change.
= a0 + a1
+ a2
n2 + 2
2 + 1
2
+ a3E + a4B + a5
,
(1)
(2)
2
= a0 + a1ET(Z) + a2B + a3
.
Here
is a parameter to be correlated ( G , H ,
S ), is the dielectric constant of the solvent, n is
the refractive index, E and V are the empirical elec-
trophilicity and nucleophilicity parameters [3], ET and
The reaction rate is independent of the concentra-
tion of the indicator and on the nature of its substi-
tuents. The concentrations of the substrate and indi-
cator in the kinetic experiments were 0.01 0.10 and
Z are the solvatochromic ionizing ability parameters
2
[17],
= ( Hm
RT)/Vm is the self-association
energy of the solvent, Hm is the molar evaporation
heat, and Vm is the molar volume.
4
(1 3) 10 M, respectively, and the conversion of
the substrate was 0.001 1%.
Equations (1) and (2) are similar to each other,
since the ET and Z values are fairly accounted for by
the polarity, polarizability, and electrophilicity solvent
parameters [5]. Equation (2) has the advantage of the
applicability to a smaller number of experimental
points.
The rate constants in 14 solvents at various tem-
peratures are given in Table 1. Table 2 lists our and
published values of log k25, G , H , and S in 16
solvents and required solvent parameters.
The figure shows the
dependences. Two compensation effects are well
defined: one for aprotic solvents (plots a and c) and
the second, for protic (plots b and d). No
G
S
and
H
S
In [1] we studied the kinetics of heterolysis of
chloride I at 25 C in 10 alcohols and 26 aprotic sol-
vents, and Takuchi et al. [19] reported kinetic data
for solvolysis of this substrate at various temperatures
in MeOH, EtOH, and AcOH, as well as at 25 C in
CF3CH2OH and various water ethanol mixtures. In
aprotic solvents, this reaction occurs by the E1 me-
chanism, while in protic, by the E1 + SN1 mechanism.
The reaction rate is always described by a first-order
equation (3).
G
S
and S compensation effects are observed for
H
all the 16 solvents. For 8 protic solvents, a good
correlation was obtained [Eq. (5)].
H
= (102000 1540) + (235 19) S ;
R 0.981, S 2700, F 156 (4.21), n 8.
(5)
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
1885
Table 1. Solvent and temperature effects on the rate of heterolysis of 1-chloro-1-methylcyclopentanea
1
1
Solvent
MeOH
t,
C
k 107, s
Solvent
t,
C
k 107, s
25.0
184 4b
320 3
Benzonitrile
25.0
0.186 0.004
0.386 0.002
0.579 0.015
1.12 0.01
29.5
35.0
40.5
44.5
25.0
30.0
35.0
40.0
44.5
25.0
31.5
35.0
43.5
49.5
30.0
35.0
40.0
45.0
25.0
30.0
35.0
41.0
45.0
25.0
25.5
30.0
35.0
40.0
44.0
25.0
30.5
35.0
45.0
50.5
25.0
29.5
35.0
39.5
43.0
48.5
25.0
30.5
35.0
39.5
44.0
646 1
1150 20
1770 10
13.0 0.1
25.3 0.7
45.2 0.1
83.9 0.4
140 5
4.80 0.01
11.8 0.4
17.1 0.5
44.2 0.2
95.5 0.1
2.00 0.02
BuOH
Nitrobenzene
Acetone
0.181 0.008
0.311 0.003
0.565 0.001
1.14 0.01
1.73 0.02
Cyclohexanol
0.0625 0.0003
0.0685 0.0007
0.128 0.001
0.248 0.002
0.578 0.001
0.847 0.001
0.0442 0.0002
0.0845 0.0003
0.137 0.004
0.345 0.002
0.586 0.002
0.0453 0.0001
0.0926 0.0002
0.168 0.003
0.319 0.009
0.472 0.007
0.808 0.005
0.106 0.002
0.227 0.004
0.400 0.010
0.705 0.002
1.16 0.03
i-PrOH
t-BuOH
25.0
30.0
35.0
39.0
44.0
25.0
31.0
35.0
40.0
46.0
50.0
25.0
30.5
35.5
40.0
41.0
46.0
25.0
30.5
34.5
39.5
45.5
49.5
25.0
29.5
34.5
40.0
44.5
14.3 0.2
28.7 0.2
49.7 1.0
79.4 0.1
126 4
Cyclohexanone
Acetophenone
4.16 0.05
7.13 0.03
10.6 0.3
15.1 0.1
23.3 0.2
31.5 0.1
1.50 0.06
2.70 0.02
4.58 0.01
6.61 0.01
7.35 0.02
10.7 0.5
2.47 0.01
4.52 0.01
7.13 0.01
11.7 0.2
21.3 0.3
32.6 0.1
2.00 0.02
3.84 0.02
6.86 0.01
12.1 0.3
20.4 0.5
t-AmOH
1,2-Dichloroethane
-Butyrolactone
Acetonitrile
a
b
5
5
1
Averages over 2 3 runs are given.
In [19], k 4.65 10 and k 86.1 10
s .
25
50
Here and heteinafter,
H
and G are in J/mol, F
value over critical).
are the apparent and critical (in parentheses) Fisher
criteria at a 95% confidence level (the reliability of
the model is confirmed by the excess of the apparent
The
G
S correlation for 8 protic solvents is
absent (R 0.78). Upon exclusion of the points for
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

1886
DVORKO et al.
Table 2. Solvent effects on the rate and activation parameters of heterolysis of 1-chloro-1-methylcyclopentane
log k251
G ,
kJ/mol
H ,
kJ/mol
S ,
ET,
kJ/mol
No.
Solvent
r
1
(k in s )
J mol 1 K
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
MeOHa
4.74
4.89
5.25
5.89
5.84
6.32
6.38
6.82
6.61
6.70
7.74
7.73
7.97
8.20
8.34
8.35
100 2
101
103
89 1.2
99
98
37 3
6
16
0.9996
1.0
1.0
232
214
217
210
203
196
184
175
185
193
176
176
175
177
173
171
AcOHb
EtOHc
BuOH
i-PrOH
Cyclohexanol
t-BuOH
tert-C5H11
-Butyrolactone
MeCN
PhNO2
PhCN
107 1
106 4
109 4
109 2
112 3
111 1
111 4
117 2
117 6
118 2
120 6
121 5
121 3
93.1 0.6
87.4 2.2
93.2 2.1
62.1 0.8
71.9 1.4
81.1 0.6
89.5 1.9
87.5 1.1
89.1 3.1
96.6 1.0
107.2 3.0
94.7 2.7
77.7 1.5
45 2
64 7
53 6
159 2
134 4
99 2
73 6
99 3
94 9
73 3
42 9
87 8
144 5
0.9999
0.9984
0.9992
0.9998
0.9984
0.9999
0.9989
0.9997
0.9971
0.9997
0.9980
0.9987
0.9995
1,2-Dichlorethane
Acetone
PhCOMe
Cyclohexanone
2
Z,
kJ/mol
E,
kJ/mol
B,
kJ/mol
,
No.
Solvent
(20)
n2D0
kJ l 1 mol
1
1
2
3
4
5
6
7
8
9
MeOH
AcOH
EtOH
BuOH
i-PrOH
Cyclohexanol
t-BuOH
tert-C5H11
-Butyrolactone
350
331
333
325
319
314
298
296
290
298
278
272
265
275
274
271
32.7
6.15
24.3
17.1
18.3
15.0
10.9
5.8d
39.0
37.5
34.8
25.2
10.4
20.7
17.4
18.3
1.3286
1.3716
1.3614
1.3992
1.3773
1.4674
1.3848
1.3859
1.4360
1.3416
1.5546
1.5282
1.4451
1.3588
1.5350
1.4510
62.3
61.1
48.5
43.1
33.6
28.9
21.8
22.6
12.1
21.8
0.8
3.3
12.6
8.8
2.9
2.1
2.61
1.66
2.81
2.76
2.82
2.89
2.95
2.95
2.48
1.91
0.8
1.85
0.48
2.68
2.42
2.89
941
427
703
552
565
515
460
460
695
594
477
515
411
393
464
431
10 MeCN
11 PhNO2
12 PhCN
13 1,2-Dichloroethane
14 Acetone
15 PhCOMe
16 Cyclohexanone
a
1
1
b
c
In [19], log k
5.34 [21].
4.33, log k
At 25 C.
3.06,
H
91.1 kJ/mol,
S
23 J mol
K
.
log k
3.51 [19]. log k
3.88 [19], log k 5
50
2
25
50
50
d
MeOH and t-BuOH we obtained Eq. (6).
The H
the decrease both in
S compensation effect is explained by
and in S .
H
G
= (102000 1220)
(79.8 18.1) S ;
The free terms in Eqs. (5) and (6) have the same
value. This value, G = H at S = 0, is equal
R 0.910, S 1840, F 19.4 (6.26), n 6.
(6)
to the potential energy of the reaction Er
102 kJ/mol [22, 23].
=
Further exclusion of the point for cyclohexanol
results in a fair correlation.
The
H
S correlation for 8 aprotic solvents is
G
= (102000 890)
(79.8 12.7) S ;
poor, R 0.893. Exclusion of the point for MeCN re-
sults in an approximate correlation.
R 0.964, S 1300, F 39.2 (9.12), n 5.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
= (118000 4980) + (300 52.1) S ;
1887
H
Consequently, at 38 C, the rate of heterolysis of
chloride I in all the protic solvents should be the same.
However, the log k 1/T dependences for protic sol-
vents reveal no isokinetic temperature. Thus, the same
rates of solvolysis of chloride I in MeOH and BuOH
are observed at 212 C (log k 67.1), in tert-
C5H11OH and MeOH, at 97 C (log k 15.9), in tert-
C5H11 and i-PrOH, at 55 C (log k 11.6), in t-BuOH
and cyclohexanol, at 21 C (log k 6.51), in tert-C5H11
and t-BuOH, at 125 C (log k 3.52), and in cyclo-
hexanol and i-PrOH, at 289 C (log k 1.63). Thus, the
presence of a compensation effect in a reaction series
not necessarily implies the fulfillment of the isokinetic
relationshiop.
R 0.932, S 3930, F 33.2 (4.95), n 7.
The same operation with -butyrolactone gives the
following correlation.
H
= (115000 4450) + (268 48) S ;
R 0.928, S 3700, F 30.8 (4.95), n 7.
A good correlation is obtained when both the sol-
vents are excluded.
H
= (118000 2310) + (288 24) S ;
The compensation effect points to a certain simi-
larity of solvation effects [23]. In our case this seems
obvious, as we observe two different compensation
effects in protic and aprotic solvents, where the condi-
tions of solvation of the transition state are much dif-
ferent. At the same time, it is solvation of the transi-
tion state in monomolecular heterolysis reactions
which determines changes in G , H , and S in
going from one solvent to another, since the solvation
energies of the ground state vary only slightly [10].
R 0.986, S 1820, F 140 (6.26), n 6.
The
G
S
correlation in aprotic solvents
is absent.
It is assumed that the presence of a compensation
effect points to the fulfillment of the isokinetic rela-
tionship which implies that there is some temperature
point in a given reaction series, at which the reaction
has the same rate in all the solvents [22 24].
This isokinetic temperature is found by the follow-
ing relationship:
In protic solvents, the most important is electro-
philic solvation of the substrate nucleofuge via forma-
tion of H complex A. This solvation is a driving force
of the heterolysis reaction [16, 25]. Along with a
linear quadrupole A, a cyclic quadrupole B may also
form [1]. This assumption is supported by the strong
decrease in S is going from primary to secondary
and further to tertiary alcohols (Table 2), which is ac-
counted for by steric hindrances to formation of sol-
vate B.
H
=
S .
The operator means the use of relative values of
activation parameters. With EtOH as standart solvent,
we have the following expression for protic solvents.
H
= 235 S ; R 0.981, S 2.7, n 8.
+
+
Me
+
+
H
+
R
+
H
R Cl
R Cl
R C N R Cl
+
Cl
OR
Cl
O=C
HC
O H
N C R
+
+
R
A
B
C
D
E
1
In dipolar aprotic solvents, too, cyclic C and
linear D quadrupoles may form. In this case,
decreases the activation entropy by 45 J mol 1 K
[27]. The values in acetone and cyclohexanone
suggest a quadrupole and a cyclic solvate E as transi-
decrease is indeed observed in -butyrolactone, MeCN, tion states in the first and second cases, respectively.
S
S
1
should decrease by about 85 J mol 1 K [26]. Such a
PhCN, PhNO2, and 1,2-dichloroethane. In the series
The formation of such cyclic complexes decreases
S
1
acetone acetophenone cyclohexane,
S
decreases
by 150 200 J mol 1 K [26, 28].
1
from 42 to 144 J mol 1 K . In acetone, a charge-
transfer transition complex is likely to be formed.
Actually, the coordination of one monodentate ligand
The presence of several compensation effects in
one reaction series may imply different solvation
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

1888
DVORKO et al.
1
mechanisms. Different compensation effects in one
and the same reaction series may be observed separa-
tely in protic and in aprotic solvents. For instance,
two compensation effects have been reported for
heterolysis of t-BuCl in protic solvents [11], as well
as for heterolysis of 7 -bromocholesterol benzoate in
aprotic solvents [29]. The same conclusions follow
from a comparison of H and S for heterolysis of
Ph2CHBr in aprotic solvents [30, 31].
t-BuCl ( 148, 164, and 193 J mol 1 K , respec-
tively) are much lower than for chloride I. This dif-
ference is presumably explained by hindrances to
formation of solvates C, D, and E.
Analysis of solvation effects in heterolysis of chlo-
ride I in 8 protic and 8 aprotic solvents in terms of
Eq. (1) gave a fair five-parameter correlation.
G
= (121000 15700) + (10600 34600)f( )
Analysis of solvent effects on activation parameters
for monomolecular heterolysis reactions shows that
the reason for the observation of several compensation
effects may lie in a narrow S scale [6, 10, 11, 16,
32, 33].
(7460 26600)f(n) (0.288 0.062)E (0.258 0.965)B
2
(0.00959 0.0090)
;
R 0.952, S 2610, F 19.54 (2.91), n 16.
The errors in the polarity, polarizability, and nucleo-
philicity parameters show that these parameters has
no rate effect. Actually, the quality of the correlation
is practically almost unaffected on exclusion of these
parameters.
The rate of heterolysis of chloride I is about two
orders of magnitude higher that the rate of heterolysis
of t-BuCl. The rate increase in AcOH and primary
alcohols is mostly associated with a decrease in H ,
since the S values are roughly equal to each other.
This result points to a greater stability of the 1-methyl-
cyclopentyl cation compared with the tert-butyl cation
and to similarity of solvation effects in heterolysis of
both substrates. In i-PrOH, the H values for both
substrates (86 kJ/mol; here and hereinafter, values for
t-BuCl are given) are almost equal to each other, and
G
= (123000 2420)
(0.292 0.035)E
2
(0.00779 0.0050)
;
R 0.951, S 2310, F 62.03 (2.60), n 16.
In fact, depends only on the electrophilicity
G
the rate decrease in going from chloride I to t-BuCl is
1
explained by a decrease in
S
[ 106 J mol 1 K ].
of the solvent.
These data suggest that solvation of the transition
state in heterolysis of t-BuCl involves stronger steric
hindrances than in the case of chloride I. Therefore,
the rate of heterolysis of t-BuCl in i-PrOH is two
orders of magnitude lower compared with chloride I,
whereas the respective difference in primary alcohols
is only one and half orders of magnitude. In t-BuOH,
the rate decrease of two orders of magnitude in going
from chloride I to t-BuCl is explained by a sharp
G
= (119000 955)
(0.321 0.031)E;
R 0.942, S 2430, F 110.4 (2.44), n 16.
Using Eq. (2) for 16 solvents leads to a fair and an
approximate three-parameter correlations.
G
= (189000 6440)
(0.268 0.027)Z
2
+ (1.10 0.731)B + (0.000477 0.0050)
;
increase in
H (101 kJ/mol) with simultaneous in-
crease in S ( 67 J mol 1 K ). The steric hindrances
to solvation of the transition state of chloride I in
t-BuOH are so strong that here, compared with pri-
mary alcohols, quite a different solvate is formed,
which shows up in a sharp decrease in S . Of im-
portance is also the fact that t-BuOH is (again for
steric reasons) is a poorly structured solvent, and,
therefore, breakdown of the solvent structure should
only slightly affect S .
1
R 0.966, S 2020, F 56.27 (2.60), n 16.
G
= (179000 8570)
(0.356 0.057)ET
2
(0.510 1.026)B + (0.00372 0.0080)
;
R 0.926, S 2950, F 24.02 (2.60), n 16.
In both cases, cohesion has no rate effect.
G
= (189000 5730)
+ (1.10+0.703)B;
R 0.966, S 1940, F 91.36 (2.60), n 16.
= (177000 7510) (0.336 0.040)ET
(0.455 0.989)B;
R 0.924, S 2870, F 38.2 (2.60), n 16.
(0.266 0.020)Z
In -butyrolactone, MeCN, PhCN, and PhNO2, the
rate decrease of two orders of magnitude in going
from chloride I to t-BuCl is mostly contributed by an
increase in H , while the
S
values are close to
G
each other. In all the above solvents, apparently, sol-
vates C and D are formed. However, in acetone, aceto-
phenone, and cyclohexanone, the
S
values for
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
1889
(8)
The nucleophiicity of the solvent has a negative,
even if rather weak, rate effect, but the greatest con-
tribution is from ionizing ability solvent parameters.
G
= (122000 2280) + (35700+6110)f(n)
2
+ (1.77 0.287)B
(0.0357 0.0020)
;
R 0.993, S 632, F 93.72 (6.09), n 8.
G
= (187000 5970)
R 0.960, S 2030, F 163.5 (2.44), n 16.
= (177000 7280) (0.341 0.038)ET;
(0.254 0.020)Z;
Exclusion of the nucleophilicity parameters much
deteriorates correlation (7), making it fair.
G
G
= (133000 2720)
(0.219 0.083)E
R 0.923, S 2780, F 80.64 (2.44), n 16.
2
(0.0283 0.0060)
;
Thus, in the set of 8 protic and 8 aprotic solvents,
G decreases with increasing solvent electrophilicity
and ionizing ability. A weak negative effect of nucleo-
philic solvation is observed.
R 0.955, S 1410, F 26.14 (4.88), n 8.
Thus, in aprotic solvents, electrophilicity, ionizing
ability, and cohesion decrease G , whereas nucleo-
philicity and polarizability increase it. In this set of
solvents, too, parameters of Eqs. (1) and (2) are in-
The solvent parameters in Eqs. (1) and (2) for the
set of 8 protic and 8 aprotic solvents are independent
variables. The strongest collinearity [8] is observed
dependent variables. Collinearity between solvent
2
2
for the following pairs:
ET (R 0.707),
Z (R
parameters is slightly better in this case:
Z (R
f( )
2
2
2
2
2
0.650),
E (R 0.538), f(n) E (R 0.688), f( )
(R
0.805),
ET (R 0.703), f(n) E (R 0.802),
0.479). For the other pairs of parameters, R = 0.002
0.397.
(R 0.674); for the other pairs of parameters, R =
0.009 0.452.
Equation (2) for 8 aprotic solvents gives a good
and a fair three-parameter correlations.
Equation (2) for 8 protic solvents gives two fair
correlations.
G
= (176000 9540)
(0.283 0.061)ET
G
= (130000 31400)
(0.101 0.094)Z
2
2
+ (1.034 0.342)B
(0.0216 0.0040)
;
+ (4.93 2.58)B
(0.0085 0.0090)
;
R 0.989, S 777, F 61.63 (6.09), n 8.
R 0.967, S 1420, F 19.32 (6.09), n 8.
G
= (172000 19100)
(0.167 0.081)Z
G
= (121000 14900)
(0.113+0.067)ET
2
+ (1.49 0.599)B
(0.0228 0.0090)
;
2
+ (4.64 1.95)B
(0.0075 0.0070)
;
R 0.967, S 1360, F 19.25 (6.09), n 8.
R 0.975, S 1230, F 26.06 (6.09), n 8.
Exclusion of the nucleophilicity parameter deterio-
rates the correlation considerable.
Exclusion of the cohesion parameters deteriorates
the correlation.
G
= (182000 15100)
(0.313 0.098)ET
G
= (157000 12800)
+ (2.87 1.40)B;
R 0.960, S 1410, F 29.06 (4.88), n 8.
= (135000 8070) (0.182+0.029)ET
+ (2.90 1.26)B;
R 0.967, S 1270, F 36.10 (4.88), n 8.
(0.184 0.033)Z
2
(0.0187 0.0070)
;
R 0.965, S 1260, F 33.47 (4.88), n 8.
G
= (166000 27000)
(0.134 0.114)Z
2
G
(0.0237 0.0120)
;
R 0.914, S 1940, F 12.7 (4.88), n 8.
(9)
Equation (1) for 8 aprotic solvents gives a good
and an excellent three-parameter correlations.
Exclusions of the nucleophilicity parameter renders
the correlations approximate (R 0.93).
G
= (132000 2100)
+ (1.034 0.466)B
R 0.980, S 1060, F 32.7 (6.09), n 8.
(0.195 0.063)E
Equation (1) for 8 protic solvents gives one good
three-parameter and three fair two-parameter correla-
tions.
2
(0.0304 0.0040)
;
(7)
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

1890
DVORKO et al.
(72200 11500)f( ) protic solvents, solvates A and B, moving along the
G
= (99900 6600)
reaction coordinate, consecutively convert into contact,
loose, and solvent-separated ion pairs and give reac-
tion products. The decrease in the reaction rate is
associated with the formation of a cyclic F [16] or a
linear G products of nucleophilic solvation of the
contact ion pair.
+ (69200 19000)f(n) + (8.04 1.02)B;
R 0.984, S 991, F 41.07 (6.09), n 8.
= (109000 8520) + (26300 31700)f(n)
(0.229 0.042)E;
G
R 0.957, S 1440, F 27.41 (4.88), n 8.
= (123000 6280) (16100+14600)f( )
(0.241 0.033)E;
R 0.961, S 1380, F 30.20 (4.88), n 8.
H
O
R
H
R+
Cl
O
R+Cl HOR
O R
H
G
F
G
Formation of solvate G in monomolecular hetero-
lysis reactions was discussed earlier [37 39]. The
assumptions made in those works are also consistent
with the fact that the polarizability of the solvent,
which appears to favors formation of complexes F
and G, too, exerts a negative rate effect.
G
= (121000 14900)
(0.113+0.067)E
2
+ (4.64 1.95)B
(0.0075 0.0070)
;
R 0.975, S 1230, F 26.06 (4.88), n 8.
The greatest contribution is from the electrophili-
city parameter.
The reason for the negative effect of nucleophilic
solvation in our set of aprotic solvents is the high
nucleophilicity of these solvents. Thus, -butyro-
lactone, acetone, acetophenone, and cyclohexanone
are as nucleophilic as alcohols, and the nucleophilicity
of PhCN and MeCN, too, is sufficiently high
(Table 2). The negative effect of nucleophilic solva-
tion may be associated with the formation of com-
plexes D and E. Stoichiometric evidence for the forma-
tion of complexes D in SN1 (E1) reactions has been
reported [40, 41].
G
= (116000 1420)
(0.249 0.033)E;
R 0.951, S 1410, F 57.11 (4.21), n 8.
Thus, in protic solvents,
creasing solvent electrophilicity or ionizing ability and
increases with increasing solvent nucleophilicity. The
strongest collinearity is characteristic of the following
pairs of parameters of Eqs. (1) and (2):
G
decreases with in-
2
ET (R
2
2
0.745),
Z (R 0.764), f( )
(R 0.719), B E (R
E (R 0.565); for the
2
2
0.716), f(n)
other pairs of parameters, R 0.118 0.453.
(R 0.584),
Correlation analysis of the effect of solvent para-
meters on the H of heterolysis of chloride I in the
set of 8 protic and 8 aprotic solvents in terms of
Eqs. (1) and (2) was unsuccessful (R 0.66). Exclusion
of the most deviating poins (acetone and t-BuOH) did
not improve the correlation.
Correlation analysis of solvation effects in hetero-
lysis of chloride I in 16 solvents showed that the re-
action rate is most strongly contributed by the electro-
philicity or ionizing ability of the solvent. Nucleo-
philic solvation exerts a negative rate effect, which is
characteristic both of the entire set of solvents and for
protic and aprotic solvents seprately. Earlier this effect
was observed in protic solvents only, in heterolyses
of t-BuX, 1-AdX, p-MeOC6H4CCl3, 2-bromo-2-
methyladamantane, PhCMe2Cl, Ph2CCl2, 1-bromo-1-
methylcyclohexane, and 1-bromo-1-methylcyclopen-
tane [6, 9, 11, 16, 21, 32, 34]. The same effect we
also found in heterolysis of chloride I in 12 protic
solvents [1].
The correlation by Eq. (1) for 8 aprotic solvents,
too, was poor (R 0.71). However, upon exclusion of
the point for cyclohexanone, one excellent, on good,
and two fair correlations were obtained.
H = (126000 2860) + (0.384 0.085)E + (4.15 0.71)B
2
(0.0878 0.0060)
;
R 0.993, S 1410, F 68.18 (8.94), n 7.
(10)
(11)
H
= (168000 13900)
(148000 47900)f(n)
The negative effect of nucleophilic solvation is
associated with the nucleophilic solvation of the
contact ion pair of the substrate, which is formed
before the limiting stage [16]. This stabilizes the
intermediate and hinders nucleofuge elimination by
the SN1 (E1) mechanism [35]. Nonequilibrium solva-
tion of the transition state takes place only [36]. In
2
(0.496 0.312)E
(0.0618 0.0110) ;
R 0.978, S 2430, F 22.35 (8.94), n 7.
H
= (151000 9740)
(80900 26500)f(n)
2
(0.0715 0.0110)
;
R 0.960, S 2850, F 23.36 (6.16), n 7.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
1891
H
= (124000 6810) + (3.98 1.71)B
celerated by formation of solvates C and D and
decelerated by formation of solvates E.
2
(0.0769 0.014)
;
Equation (1) for 7 protic solvents (MeOH ex-
cluded) gave a fair three-parameter correlation.
R 0.942, S 3400, F 15.91 (6.16), n 7.
Thus, in aprotic solvents, is decreased by
H
H
=
(61100 33500) + (365000 117000)f(n)
cohesion and polarizability and increased by nucleo-
2
+ (0.90 0.15)E + (0.0514 0.022)
;
philicity. The increase in the reaction rate with in-
2
creasing
and f(n) is presumably explained by the
R 0.969, S 4880, F 15.6 (8.94), n 7.
formation of cyclic quadrupole C. Evidence for this
assumption comes from the fact that
2
nicely cor-
For 8 protic solvents, R 0.921. Equation (2) for 7
relates with the dipole moments of solvents [6], and
increasing solvent polarizability should favor quadru-
pole formation. The effect of solvent nucleophilicity
suggests that the negative effect of nucleophilic solva-
tion is associated with solvent effect on the activation
enthalpy of the reaction. Of interest is the effect of
solvent electrophilicity on H : on one hand, it favors
quadrupole C formation and thus accelerates the re-
action [Eq. (11)] and on the other, it favors solvate F
formation and thus decelerates the reaction [Eq. (10)].
protic solvents gives the following correlation.
H
=
(329000 211000) + (1.31 0.657)Z
2
+ (12.8 19.4)B
(0.067 0.094)
;
R 0.937, S 6920, F 7.23 (8.94), n 7.
The greatest contribution is from the Z parameter
(R 0.926).
The effects of protic solvent parameters on the
activation enthalpy, especially its increase with in-
creasing solvent electrophilicity and ionizing ability
seems paradoxical, since these parameters are con-
sidered to be primarily responsible for reaction ac-
celeration. However, as seen from Table 2, such re-
lations are the case. The same pattern has also been
observed in solvolysis of t-BuCl [11], where H and
simultaneously ET and E increased in the series i-
PrOH OctOH HexOH BuOH EtOH. It should be
noted that even though the activation enthalpy incre-
ases, the reaction rate increases rather than decreasing.
This increase in the reaction rate is associated with
the effect of the solvent on S . Thus, for 7 protic sol-
vents, after exclusion of the point for MeOH, we ob-
tain one good and one fair three-parameter equations.
Equation (1) for 7 aprotic solvents (cyclohexanol
excluded) gave an excellent three-parameter correla-
tion.
S
= (72.7 5.13)
(344 13.8)f(n)
2
+ (0.00540 0.0010)B
(0.000139 0.00001)
;
R 0.999, S 1.42, F 410.7 (8.94), n 7.
Exclusion of the nucleophilicity parameter gives a
good two-parameter correlation.
S
= (84.7 18.0)
(374 48.8)f(n)
2
(0.000127 0.00001)
;
R 0.978, S 5.27, F 43.1 (6.16), n 7.
S
=
(593 162) + (372 254)f( ) + (908 534)f(n)
+ (0.00396 0.00100)E;
The greatest contribution is from the polarizability
parameter (R 0.727).
R 0.963, S 22.1, F 12.64 (8.94), n 7.
Thus,
S
in aprotic solvents decreases with in-
creasing polarizability and cohesion.
S
=
(320 93.9) + (1300 364)f(n) (0.129 0.015)B
2
As shown above, the rate of heterolysis of chloride
I in aprotic solvents is increased by the electrophili-
city and cohesion of the solvent and decreased by po-
larizability and nucleophilicity [Eqs. (7) and (8)]. The
+ (0.000551 0.00001)
;
R 0.983, S 14.9, F 28.83 (8.94), n 7.
Equation (2) leads to a fair one-parameter correla-
tion.
revealed effects of solvent parameters on
H
and
S
suggests that positive solvation effects are as-
sociated with the effects of electrophilicity and co-
hesion on H , whereas negative solvation effects,
S
=
(1250 148) + (0.00373 0.00001)Z;
with the effects of polarizability on
S
and of
R 0.963, S 17.0, F 64.2 (4.95), n 7.
nucleophilicity on H . These conclusions are con-
sistent with the suggestion that the reaction is ac-
Thus, in heterolysis of chloride I in protic solvents,
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

1892
DVORKO et al.
polarizability and cohesion (or ionizing ability) in-
crease S , whereas nucleophilicity decreases it.
3. Pal’m, V.A., Osnovy kolichestvennoi teorii organi-
cheskoi khimii (Fundamentals of Qualitative Theory
of Organic Chemistry), Leningrad: Khimiya, 1977.
According to Eq. (9), the ionizing ability solvent
parameter increases the rate of heterolysis of chloride
I, while the nucleophilicity parameter decreases it. It
was found that a positive solvation effect is associa-
ted with the effect of electrophilic solvation on the
acivation entropy and a negative solvation effect,
by the effect of nucleophilic solvation on the activa-
tion enthalpy. The lack of rate effect of both the po-
larizability and cohesion parameters which increase
both H and S is explained by compensation.
4. Makitra, R.G. and Pirig, Ya.N., Zh. Obshch. Khim.,
1986, vol. 56, no. 3, p. 657.
5. Linear Free Energy Relationships, Chapman, N.B.
and Shorter, J., Eds., London: Plenum, 1972, p. 203.
6. Dvorko, G.F., Ponomareva, E.A., and Kulik, N.I., Usp.
Khim., 1984, vol. 43, no. 10, p. 948.
7. Makitra, R.G. and Pirig, Ya.N., Reakts. Sposobn. Org.
Soedin., 1978, vol. 15, no. 3, p. 352.
8. Abraham, M.H., Doherty, R.W., Kamlet, M.J., Har-
ris, J.M., and Taft, R.W., J. Chem. Soc., Perkin Trans.
2, 1987, no. 5, p. 913.
Similar solvation effects have been observed in
heterolysis of t-BuCl in 15 protic solvents [11]. In this
case, the reaction is decelerated by electrophilicity and
cohesion and decelerated by nucleophilicity, and,
therewith, positive solvation effects are explained by
the solvent effects on S , while negative, on H .
9. Dvorko, G.F., Tarasenko, P.V., Ponomareva, E.A.,
and Kulik, N.I., Zh. Org. Khim., 1989, vol. 25, no. 3,
p. 922.
10. Abraham, M.H., Grellier, P.L., Nasehzadeh, A., and
Walker, R.A.C., J. Chem. Soc., Perkin Trans. 2, 1988,
no. 6, p. 1717.
The nature of solvation effects in monomolecular
heterolysis reactions is controversial [9, 16, 42, 43].
This controversy not infrequently arises from the fact
that some authors used individual solvents [1, 5, 6,
8 11, 16, 20, 21, 29, 32 34], and others, solvent
mixtures [19, 25, 42 45], where the conclusions as to
the nature of solvation effects may be incorrect [16].
One more reason for the controversy is that some
authors used multiprameter equations [3, 5, 6, 11 16,
21, 32 34], and others, in one- and two-parameter
Grunwald Winstein equations [19, 42 45] which are
of limited utility [16, 33]. However, even with indi-
vidual solvents and multiparameter equations, dif-
ferent solvation effects are frequently observed with
different substrates and solvent sets. The present work
shows that the reason for this phenomenon may lie in
different quantitative and qualitative effects of solvent
parameters on activation enthalpies and entropies.
11. Dvorko, G.F., Zaliznyi, V.V., and Ponomarev, N.E.,
Zh. Obshch. Khim., 2002, vol. 72, no. 9, p. 1501.
12. Makitra, R.G., Marshalok, G.A., Pirig, Ya.N., and
Yatchishin, I.I., Available from VINITI, Moscow,
1984, no. 407-84.
13. Sendega, R.V., Makitra, R.G., and Pirig, Ya.N., Kinet.
Katal., 1986, vol. 27, no. 4, p. 789.
14. Makitra, R.G. and Pirig, Ya.N., Available from
VINITI, Moscow, 1989, no. 6003-V89.
15. Kucher, R.V., Gavryliv, E.M., Zhukovskii, V.L., Ma-
kitra, R.G., Pirig, Ya.N., and Turovskii, A.A., Dokl.
Akad. Nauk USSR, 1989, no. 11, p. 31.
16. Dvorko, G.F., Ponomarev, N.E., and Ponomare-
va, E.A., Zh. Obshch. Khim., 1999, vol. 69, no. 11,
p. 1835.
17. Reichardt, C., Solvents and Solvent Effects in Organic
Chemistry, Weinheim: VCH, 1988, 2nd ed.
EXPERIMENTAL
18. Moura-Ramos, J.J., J. Solution Chem., 1989, vol. 18,
Reagents and solvents were synthesized and puri-
fied as described in [1]. Kinetic measurements were
performed in the temperature-controlled cell of an
SF-26 spectrophtometer [11, 20]. Calculations by
Eqs. (1) and (2) were performed by the least-squares
procedure using the Spss package; confidence level
95%.
no. 10, p. 957.
19. Takeuchi, K., Ohga, Y., Ushino, T., and Takasuka, M.,
J. Phys. Org. Chem., 1997, vol. 10, no. 3, p. 717.
20. Dvorko, G.F. and Ponomareva, E.A., Usp. Khim.,
1991, vol. 60, no. 10, p. 2089.
21. Dvorko, G.F., Koshchii, I. V., Pervishko, T.L., and
Ponomareva, E.A., Zh. Obshch. Khim., 2000, vol. 70,
no. 6, p. 973.
REFERENCES
22. Hammett, L., Physical Organic Chemistry, New York:
McGraw Hill, 1970, 2nd ed.
23. Exner, O., Chem. Listy, 1973, vol. 67, no. 2, p. 135.
24. Isaaks, N.S., Physical Organic Chemistry, New York:
Wiley, 1992.
1. Dvorko, G.F., Koshchii, I.V., Prokopets, A.M., and
Ponomareva, E.A., Zh. Obshch. Khim., 2002, vol. 72,
no. 11, p. 1902.
2. Ingold, C.K., Structure and Mechanism in Organic
Chemistry, Ithaca: Cornell Univ., 1969, 2nd ed.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002

KINETICS AND MECHANISM OF MONOMOLECULAR HETEROLYSIS... : XXXII.
1893
25. Bentley, T.W. and Schleyer, P.v.R., Adv. Phys. Org.
Chem., 1977, vol. 14, no. 1, p. 1.
35. Ulrich, Z., Chem. Ztg., 1984, vol. 108, no. 2, p. 381.
36. Gertner, B.J., Wilson, K.R., Zichi, D.A., Lee, S., and
Hynes, J.T., Faraday Discuss. Chem. Soc., 1988,
no. 85, p. 297.
26. Hoffmann, R.W., Auf Klarung von Reaktionsme-
chanismen, Stuttgart: Thieme, 1976.
27. Gordon, J.E., The Organic Chemistry of Electrolyte
Solutions, New York: Wiley, 1975.
37. Ions and Ion Pairs in Solvolysis Reaktions, Szwarc, M.,
Ed., New York: Wiley, 1974, vol. 2, p. 248.
28. Maskill, H., Physical Basis of Organic Chemistry,
38. Dannenberg, J.J., J. Am. Chem. Soc., 1976, vol. 98,
Oxford: Oxford Univ. Press, 1993.
no. 20, p. 6261.
29. Yakhimovich, R.I., Ponomarev, N.E., and Dvor-
ko, G.F., Zh. Obshch. Khim., 1988, vol. 58, no. 4,
p. 881.
39. Okamoto, K., Pure Appl. Chem., 1984, vol. 56, no. 12,
p. 1798.
30. Ponomareva, E.A., Pervishko, T.L., and Dvorko, G.F.,
Reakts. Sposobn. Org. Soedin., 1981, vol. 18, no. 2,
p. 323.
40. Weiner, H. and Sneen, R.A., J. Am. Chem. Soc., 1965,
vol. 87, no. 2, p. 287.
41. Okamoto, K., Nitta, I., Dohi, M., and Shingu, H., Bull.
Chem. Soc. Jpn., 1971, vol. 44, no. 2, p. 320.
31. Ponomareva, E.A., Pervishko, T.L., and Dvorko, G.F.,
Reakts. Sposobn. Org. Soedin., 1982, vol. 19, no. 1,
p. 18.
42. Bentley, T.W. and Llewellyn, G., Progr. Phys. Org.
Chem., 1990, vol. 17, no. 1, p. 121.
32. Dvorko, G.F., Zhovtyak, V.N., and Evtushenko, N.Yu.,
43. Kevill, D.N. and D’Sourza, M.J., J. Phys. Org. Chem.,
Zh. Obshch. Khim., 1989, vol. 59, no. 7, p. 1600.
1992, vol. 5, no. 3, p. 287.
33. Dvorko, G.F., Zaliznyi, V.V., and Ponomarev, N.E.,
44. Bentley, T.W., Carter, G.T., and Roberts, K., J. Org.
Chem., 1984, vol. 49, no. 26, p. 5183.
Zh. Obshch. Khim., 2002, vol. 72, no. 10, p. 1644.
34. Dvorko, G.F., Cherevach, T.V., Kulik, N.I., and Pono-
marev, N.E., Zh. Obshch. Khim., 1994, vol. 64, no. 6,
p. 979.
45. Liu, K.-T. and Shen, H.C., J. Org. Chem., 1991,
vol. 56, no. 9, p. 3021.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 72 No. 12 2002