Correlation of the rates of solvolysis of 3,4,5-trimethoxy- and 2,4,6-trichlorobenzoyl chlorides

Article abstract of DOI:10.3184/174751911X13192929905095

The introduction of methoxy groups into the 3- and 5-positions of 4-methoxybenzoyl chloride leads to a reduction in specific rates of solvolysis. An extended Grunwald-Winstein equation correlation for the specific rates of solvolysis of 3,4,5-trimethoxybenzoyl chloride gives sensitivities towards changes in solvent nucleophilicity (l value) of 0.29 and towards changes in solvent ionising power (m value) of 0.54. The low m value allows specific rates to be determined in highly ionising fluoroalcohol-H₂O mixtures. A parallel correlation of the specific rates of solvolysis of 2,4,6- trichlorobenzoyl chloride reveals that solvolyses in 100% and 90% ethanol or methanol do not appreciably follow the ionisation pathway indicated for solvolyses in the other solvents and it is proposed that, despite the two ortho-substituents, the addition-elimination pathway has become dominant.

Full text of DOI:10.3184/174751911X13192929905095

JOURNAL OF CHEMICAL RESEARCH 2011
RESEARCH PAPER 647
NOVEMBER, 647–653
Correlation of the rates of solvolysis of 3,4,5-trimethoxy- and
2,4,6-trichlorobenzoyl chlorides
Kyoung-Ho Park and Dennis N. Kevill*
Department of Chemistry and Biochemistry, Northern Illinois University, DeKalb, IL 60115-2862, USA
The introduction of methoxy groups into the 3- and 5-positions of 4-methoxybenzoyl chloride leads to a reduction in
specific rates of solvolysis. An extended Grunwald–Winstein equation correlation for the specific rates of solvolysis
of 3,4,5-trimethoxybenzoyl chloride gives sensitivities towards changes in solvent nucleophilicity (l value) of 0.29 and
towards changes in solvent ionising power (m value) of 0.54. The low m value allows specific rates to be determined
in highly ionising fluoroalcohol-H2O mixtures. A parallel correlation of the specific rates of solvolysis of 2,4,6-trichlo-
robenzoyl chloride reveals that solvolyses in 100% and 90% ethanol or methanol do not appreciably follow the ioni-
sation pathway indicated for solvolyses in the other solvents and it is proposed that, despite the two ortho-substitu-
ents, the addition–elimination pathway has become dominant.
Keywords: trisubstituted benzoyl chlorides, linear free energy relationships, Grunwald–Winstein equation, kinetics and
mechanism, substituent effects
One aspect of the present report involves a study of the rates of
solvolysis of 3,4,5-trimethoxybenzoyl chloride (1). There have
been extensive studies of 4-methoxybenzoyl chloride (2) sol-
volyses.1–6 In reactions of 1, the powerful resonance electron
donation of the para-methoxy group will be moderated not
only by its own electron withdrawing inductive effect but also
by the inductive effect of the two meta-methoxy groups. The
4-methoxybenzoyl chloride has been studied in 40 solvents but
for many of these solvents, including the important solvents
rich in 1,1,1, 3,3,3-hexafluoro-2-propanol (HFIP), values at
25.0 °C had to be estimated by anArrhenius-type extrapolation
of values obtained at lower temperatures. The retarding influ-
ence of the two meta-methoxy substituents allows the kinetics
of solvolysis of 1 to be studied directly at 25.0 °C over a much
wider range of solvents.
specific rates (first-order rate coefficients) for solvolysis of a
substrate RX in a given solvent and in the standard solvent
(80% ethanol–20% water v/v), respectively, m is the sensitivity
to changes in solvent ionising power (YX for a leaving-group
X), l is the sensitivity to changes in solvent nucleophilicity (NT
values based on solvolyses of the S-methyldibenzothiophe-
nium ion are usually used), and c is a constant (residual) term.
Tabulations of Yx14 and NT15 values are available.
The solvolyses of acyl chlorides can be considered in terms
of two mechanisms. An ionisation mechanism is shown as
Scheme 1. Indications are, however, that this mechanism is
frequently subject to an appreciable nucleophilic assistance
from the solvent. This can be pictured either in terms of nucleo-
philic solvation of a developing acylium ion7,8 or in terms of a
borderline SN1-SN2 pathway.1,2 An alternative pathway, involv-
ing nucleophilic addition to the carbonyl group is shown in
Scheme 2.
We have recently reported7 the correlations of the rates of
solvolysis of a series of dichlorobenzoyl chlorides, including
2,6-dichlorobenzoyl chloride (3) and 2,4-dichlorobenzoyl
chloride (4). We now report the specific rates of solvolysis
of 2,4,6-trichlorobenzoyl chloride (5) in a wide variety of
hydroxylic solvents. The consequences following from the
introduction of a third chlorine into the aromatic ring can
be considered either as regards its introduction into the para-
position of 3 or as regards its introduction into the remaining
ortho-position of 4.
Deprotonation of the first formed intermediate in Scheme 2
(general-base catalysis) allows progression to product by
an overall addition–elimination (association-dissociation)
pathway.1,2,8,16 The solvolyses of p-nitrobenzoyl chloride (6), in
all studied solvents except 97% HFIP, are believed to follow
this mechanism.8
Results and discussion
It has been shown8 that the specific rates of solvolysis of
benzoyl chloride and its derivatives over a wide range of
hydroxylic solvents can be conveniently correlated using an
extended form9–12 of the Grunwald–Winstein equation (1).13 In
Eqn (1), k and ko are the
The first-order rate coefficients (specific rates) for solvolysis
were obtained using a rapid response conductivity technique
for 31 solvents with 1 at 25 °C and for 28 solvents with 5 at
55.0 °C. These values are reported in Table 1. The solvents
used in the study consisted of ethanol, methanol and 2,2,2-
trifluoroethanol (TFE) and, for each, a series of mixtures with
log (k/ko)RX = lNT + mYx + c
(1)
Scheme 1
Scheme 2
* Correspondent. E-mail: dkevill@niu.edu

648 JOURNAL OF CHEMICAL RESEARCH 2011
Table 1 First order rate coefficients for solvolyses of 3,4,5-trimethoxybenzoyl chloride (1) at 25.0 °C and 2,4,6-trichlorobenzoyl
chloride (5) at 55.0 °C
Solventa
103k (1)/s-1
103 k(2)/s−1b
103 k(5)/s−1
k(5)/k(3)
100% EtOH
90% EtOH
80% EtOH
70% EtOH
60% EtOH
50% EtOH
100% MeOH
90% MeOH
80% MeOH
70% MeOH
60% MeOH
50% MeOH
90% Acetone
80% Acetone
70% Acetone
60% Acetone
50% Acetone
100% TFEg
97% TFEg
0.712 0.001
2.33 0.01
4.90 0.01
9.61 0.03
18.9 0.1
41.9 0.2
4.59 0.01
10.5 0.1
21.9 0.1
43.6 0.2
74.8 0.2
0.936
5.55
22.4
63.2
181
0.0162 0.0002
0.0496 0.0001
0.108 0.001
0.244 0.002
0.580 0.001
1.64 0.01
1.76
1.08
0.72
0.53
0.42
0.39
0.84
0.65
0.50
0.45
0.40
576d
9.57
36.0
124
0.0846 0.0009
0.218 0.001
0.523 0.001
1.28 0.01
320
834e
2400
3.05 0.01
127.0
1
0.154 0.001
0.703 0.003
7.99 0.04
14.5 0.1
2.67
11.6
50.9
212f
0.0111 0.0001
0.0441 0.0003
0.185 0.001
1.27
0.97
0.48
20.8 0.2
15.5 0.1
501
0.833 0.011
0.542 0.002
0.745 0.001
1.28 0.01
0.24
0.25
0.25
0.22
0.25
16.8 0.1
566h
90% TFEg
23.5 0.2
80% TFEg
34.7 0.3
70% TFEg
48.4 0.1
1800h
3.50 0.03
6.42 0.21
110
2.16 0.01
0.24
60% TFEg
67.6 0.2
50% TFEg
96.6 0.1
0.24
80T-20E
6.22 0.01
2.90 0.01
1.55 0.01
0.995 0.003
0.128 0.002
0.0465 0.0003
0.0219 0.0002
0.27
0.33
0.34
60T-40E
30.2
8.67
40T-60E
20T-80E
97% HFIPg
90% HFIPg
70% HFIPg
138
1
27x103i
12.7 0.1
6.34 0.03
6.18 0.02
0.23
0.26
0.26
84.0 0.2
81.3 0.1
a Prepared on a volume/volume basis at 25.0 °C, unless otherwise stated.
b Unless otherwise stated, from ref. 5.
c Values for 3 at 55.0 °C. from ref. 7.
d Ref. 1 gives value of 541.
e Ref. 1 gives value of 844.
f Ref. 1 gives value of 229.
g Solvent prepared on weight/weight basis.
h Values from ref. 2.
i Value from ref. 6 (obtained by Arrhenius extrapolation of values at −20.15 and −9.90 °C).
water. Specific rates of solvolysis were also determined in
mixtures of water with HFIP and acetone and in mixtures of
ethanol with TFE. Also contained within Table 1 are literature
values for the specific rates of solvolysis of 2 at 25.0 °C1,2,5,6
and the ratios of the specific rates for 5 relative to 37 at
55.0 °C.
It was possible for 1 to obtain specific rates of solvolysis at
25.0 °C over the full range of solvents usually included in
Grunwald–Winstein correlation studies. Surprisingly, in pure
ethanol, the specific rates were only slightly lower than for 2,
despite the addition of the two electron-withdrawing methoxy
substituents. As can be seen from the data reported in Table 1
for the solvolyses of 1 and 2, the differences become greater as
the water content increases and, for 50% ethanol, the retarda-
tion is by a factor of about 13. Similar trends are observed
for aqueous methanol and aqueous acetone. The two meta-
substituents lead to large retardations in TFE and in TFE–water
mixtures. The retardation is by a factor of 32 in 100% TFE,
and the value increases only slightly to 34 in 97% TFE (w/w)
and to 37 in 70% TFE (w/w).
solvent-induced change in mechanism, The correlation coeffi-
cient for the 22 point plot is reasonably good at 0.931, the
slope of the plot is 0.61 0.05, and the intercept is 0.01 0.06.
The slope is lower than might have been anticipated, showing
that the specific rates for solvolyses of 1 are less sensitive
towards solvent variation than those for 2. The correlation
values allow a specific rate value of 9713 to be calculated for
97% HFIP. This is only 36% of the recently estimated6 value.
Indeed, inspection of Fig. 1 shows that all the fluoroalcohol-
containing solvents lie below the correlation line.
Another approach to comparing the behaviour of 1 and 2
under solvolytic conditions is to apply the extended
Grunwald–Winstein equation to the specific rates of solvolysis
of each compound and to compare the sensitivity values (l and
m) obtained. Using the experimental specific rates at 25.0 °C
and the tabulated NT15 and YCl14 values, the l and m values for 1
can be compared to those previously reported 8 for 2 and with
new values for 2 incorporating in the analysis three recently
reported6 specific rates (Table 2). The sensitivities to changes
in solvent nucleophilicity are essentially identical and the
observed differences in response to solvent variation are due to
differences in the sensitivities towards changes in solvent
ionising power, with m values of 0.54 0.04 for 1 (Fig. 2) very
low for an ionisation pathway, and of 0.81 0.02 for 2, typical
for an ionisation pathway.
The behaviour can be examined in terms of a log–log plot
of the data for 1 against the data for 2. Plots of log (k/ko) for
an aroyl chloride against log (k/ko) for 2 have been shown
previously 3 to be a very useful probe for ionisation character
in the solvolyses of an aroyl chloride. Figure 1 shows a plot
of log (k/ko) for 1 against log (k/ko) for 2.8 There is some degree
of scatter but no constant trends that could correspond to a
The hydroxy and methoxy substituents are frequently put
forward as examples of substituents where the Hammett

JOURNAL OF CHEMICAL RESEARCH 2011 649
Fig. 1 Plot of log(k/ko) for the solvolyses of 3,4,5-trimethoxybenzoyl chloride (1) at 25.0 °C against log(k/ko) for 4-methoxybenzoyl
chloride (2) at 25.0 °C.
Table 2 Correlation on the specific rates of solvolysis of several derivatives of benzoyl chloride using the extended Grunwald–
Winstein equation (1)
Substrate
T °C
25.0
na
lb
mb
cb
Rc
l/m
3,4,5 (MeO) 1
31d
27e,,f
37h
40i
0.32 0.05
0.34 0.04
0.31 0.05
0.30 0.04
0.29 0.03
0.29 0.07
0.35 0.06
1.46 0.16
0.32 0.02
0.29 0.08
0.40 0.06
1.78 0.08
0.54 0.04
0.59 0.03
0.81 0.02
0.83 0.02
0.87 0.02
0.74 0.05
0.86 0.05
0.41 0.05
0.87 0.03
0.62 0.07
0.82 0.06
0.54 0.04
0.09 0.05
−0.02 0.05
0.08 0.06
0.04 0.05
−0.08 0.05
0.10 0.07
−0.07 0.07
0.21 0.07
−2.80 0.07n
0.05 0.08
−0.21 0.08
0.11 0.37o
0.960
0.975
0.989
0.988
0.993
0.965
0.975
0.911
0.996
0.931
0.967
0.969
0.59
0.58
0.38
0.36
0.33
0.39
0.41
3.56
0.37
0.46
0.49
3.30
4 (MeO) 2g
25.0
37j
2,6 (Cl) 3k
2,6 (Cl) 3k
2,4 (Cl) 4k
2,4 (Cl) 4k
2,4,6 (Cl) 5
2,4,6 (Cl) 5
4 (NO2) 6g
55.0
55.0
55.0
55.0
55.0
55.0
25.0
30
26e
21l
10m
27
23e
34
a Number of solvents
b With associated standard error.
c Multiple correlation coefficient.
d F-test value of 159.
e Points for 100% ethanol, 100% methanol, 90% ethanol and 90% methanol removed from the
correlation.
f F-test value of 231
g From ref. 8.
h F-test value of 630.
i From refs. 6 and 8 (values for 97% HFIP, acetic and formic acids from ref.6); F-Test value of 772.
j As immediately above entry, but with 100% EtOH, 100% MeOH and 90% EtOH removed; F-test value of 1229.
k From ref. 7
l In EtOH–H2O, MeOH–H2O, acetone–H2O, and TFE–EtOH.
m In TFE–H2O and HFIP–H2O.
n Large negative value because the ko value is for the bimolecular pathway.
o Accompanied by the standard error of the estimate.
substituent constant can be appreciably solvent dependent.17,18
One notes immediately that 1 has three methoxy groups whose
interaction with the ring can be perturbed by solvation, as
opposed to only one in 2. Another factor to be considered
is that the 4-methoxy substituent in 1 is flanked by two
ortho-substituents, which will tend to give a steric hindrance to
its solvation not present in 2. For example, in nitromethane, a
cyano group present as a substituent in 1,3,5-trimethylbenzene
(mesitylene) requires, to describe its influence on aromatic
bromination, a σ+ value of 0.686 rather than the usual 0.562.
This was readily explained by the flanking methyl group giving
an appreciable hindrance to a solvation, which moderates the
electron-withdrawal by the meta-cyano group.19,20
In the present situation, steric hindrance to solvation of the
4-methoxy group present in 1 should lead to a faster ionisation
reaction, contrary to observation, and this cannot be the

650 JOURNAL OF CHEMICAL RESEARCH 2011
Fig. 2 Plot of log (k/ko) for solvolyses of 3,4,5-trimethoxybenzoyl chloride (1) in 31 solvents at 25.0 °C against (0.32 NT + 0.54 YCl).
Solvents: E, EtOH–H2O; M, MeOH–H2O; A, Acetone–H2O; TFE, TFE–H2O; HFIP, HFIP–H2O; T–E, TFE–EtOH.
dominant influence of the introduction of the flanking
3,5-dimethoxy groups. An effect is needed which is modest in
ethanol and appreciable in aqueous-fluoroalcohols, with the
other solvents intermediate. Also, it is necessary to rationalise
why the major influence on the specific rates is reflected
entirely in a reduction of the sensitivity to changes in solvent
ionising power (m-value), with essentially no change in the
l-value. It appears that the transition state must involve about
the same degree of nucleophilic solvation but a reduced
heterolysis of the carbon–chlorine bond. Possibly, the expected
variations in solvation at the methoxy groups are, in the absence
of acidic solvents, secondary factors which change in intensity
only slightly in going from initial state to transition state, and
the dominant influence is indeed a tightening of the transition
state (reduced bond-breaking) and reduced specific rates due
to the electron-withdrawing influence of the methoxy groups
(σ = 0.115; σ+=0.047).19 Interaction of fluoroalcohol molecules
with the meta-methoxy groups of 1 will increase their elec-
tron- withdrawing power and increase the retardation relative
to the corresponding solvolyses of 2. At the opposite extreme
of the substitutent effects for 2 relative to 1, in ethanol, metha-
nol, and 90% ethanol, there are indications (points lie above
the plot in Fig. 2) for the superimposition of a bimolecular
component to the solvolyses of 1 in these solvents. In a previ-
ously reported analysis,8 a good two-term Grunwald–Winstein
analysis was obtained over all 37 solvents studied for 2. This
analysis is, however, improved if these three solvents are
omitted, with the F-test value rising from 630 to 971. A cor-
responding removal from the 40 solvent correlation, involving
addition of 97% HFIP, CH3COOH, and HCOOH6, leads to
an increase in the F-test values from 772 to 1229 (Table 2).
The component going by the addition–elimination pathway,
with the addition step rate determining, in these solvents
of relatively high nucleophilicity and relatively low ionising
power, would be favoured by the presence of the electron-
withdrawing meta-methoxy groups of 1, and the overall effect
would be a reduced ratio for the specific rates of 2 relative to
those of 1, as is observed (Table 1).
The lower specific rates allow an extension, at 25.0 °C, to
aqueous HFIP binary solvents. One can either use the solvoly-
ses of 1 as a substitute for the solvolyses of 2, if one wishes to
extend the solvent range for log (k/ko) values for direct linear
free energy relationship (LFER) plots, or one could use an
extension of Figure 1 to estimate, the log (k/ko) value for a
given solvent for 2 from the value obtained experimentally
for 1 and continue to use the solvolyses of 2 as the standard set
of solvolyses for use in probing the solvolyses of other acyl
chloride substrates.
It was established that, consistent with a previous observa-
tion21 for 2,6-dibromobenzoyl chloride, the solvolyses of
2,6-dichlorobenzoyl chloride (3) are slow at 25.0 °C and a
temperature of 55.0 °C was found to be a convenient one for
following the kinetics over a wide range of solvents.7 The same
temperature has been found to be convenient for the study
of the solvolyses of 2,4,6-trichlorobenzoyl chloride (5) and
the specific rates are reported in Table 1, together with the
corresponding ratios of the specific rates relative to those for
3 [k(5)/k(3)].
These specific rate ratios were found in TFE, six TFE–H2O
mixtures and three HFIP–H2O mixtures to be essentially
constant at 0.244 0.013 for the 10 solvents. This value is
consistent with the belief7 that the solvolyses involve a unimo-
lecular ionisation to an acylium ion. The electron-withdrawing
para-chloro substituent (σ = 0.227; σ+ = 0.114)19 when intro-
duced into 3 would be expected to cause a modest retardation,
as is observed. A two point determination based on hydrogen
or chloro in the para position (3 and 5) leads to values for ρ
of −2.7 and for ρ+ of –5.4. The ρ+ value is comparable with

JOURNAL OF CHEMICAL RESEARCH 2011 651
tabulated values19 of –4.1 to –4.6 for solvolyses of benzhydryl
chlorides and of –4.5 to –4.6 for solvolyses of tert-cumyl
(2-phenyl-2-propyl) chloride in ethanol or aqueous acetone.
Closer in character to the present study is the finding by
Bender and Chen 22 of a ρ+ value of –3.9 for the solvolyses of a
series of para-substituted 2,6-dimethylbenzoyl chlorides in
99% acetonitrile –1% water (by volume).
In TFE–ethanol mixtures, the k(5)/k(3) ratios rise modestly
from 0.24 in TFE to values of 0.27, 0.33, and 0.34 with 20%
increments in the ethanol content (by volume). In mixtures
of ethanol, methanol, or acetone with water, with a 40%
water content, ratios of 0.42, 0.40, and 0.48, respectively, are
obtained. These values are also consistent with the electron-
withdrawing influence of the 4-chloro substituent on an ionisa-
tion mechanism. As the alcohol content increases so also does
the value for the ratio and in three instances it exceeds unity:
1.76 for 100% ethanol, 1.08 for 90% ethanol, and 1.27 for
80% acetone, indicating that, in these instances, the 4-chloro is
very modestly favouring reaction. We will return to these
observations after a consideration of the Grunwald–Winstein
treatment using Eqn (1).
The results from the application of Eqn (1) are shown in
Table 2 and the details of the correlation using the 27 data
points are shown in Fig. 3. An analysis only using solvent
ionising power (Eqn (1) without the lNT term) gives a poor
correlation with a slope of (m value) of 0.42 0.04 and a cor-
relation coefficient of 0.895. The correlation is considerably
improved when the full equation is used with an l value of
0.28 0.08, m value of 0.62 0.07, and c value of 0.05 0.08
(multiple correlation coefficient of 0.931). The l and m
values are little changed from those for 2,6-dichlorobenzoyl
chloride,7 which are also shown in Table 2.
100% methanol and 90% methanol lie considerably above the
correlation lines. Behaviour of this type can be an indication of
the superimposition of an appreciable contribution from an
alternative pathway in these solvents. Since these are solvents
of relatively high nucleophilicity and relatively low ionising
power, a bimolecular mechanism, such as in Scheme 2, is the
most likely candidate.
Supporting the incursion of an addition–elimination mecha-
nism in these solvents is the observation by Bender and
Chen22 that, in their study of the hydrolysis of para-substituted
2,6-dimethylbenzoyl chlorides, addition of tetramethylammo-
nium hydroxide led to the dominant pathway involving a
bimolecular attack by hydroxide ion, with a ρ value of +1.2.
This is consistent with the ρ value of +1.6 found23 for the
ethanolyses of a series of para and meta-substituted benzoyl
chlorides, except that the p-methoxy derivative lies above the
plot. In the present case, we have not increased the nucleophi-
licity of the attacking species22 but we have reduced, by
replacement of methyl-substituents by chloro-substituents, the
ability of the substrate to follow the ionisation pathway. Both
approaches would increase the possibility for a detectable
component from the addition–elimination pathway. Although
a recent publication has proposed (based on density functional
theory computations) that, for acetyl, chloroacetyl, or benzoyl
chloride solvolyses, a concerted one-step SN2 mechanism with
a variable transition state structure operates, 24 we continue to
favour a combination of the two pathways of Scheme 1 (with
nucleophilic solvation) and Scheme 2 (with general base
catalysis).
With the four deviating points mentioned above omitted,
new Grunwald–Winstein correlations [Eqn (1)] were carried
out not only for 5 but also for 3. These new correlations are
included in Table 2 and, also, a new plot corresponding to this
correlation of the solvolysis of 5 in 23 solvents, is presented in
Fig. 4. The correlations are considerably improved and the
sensitivity values (l and m) and the intercepts (c) are used to
calculate values for specific rates of solvolyses proceeding by
A comparison of Fig. 3 with the Fig. 1 of the earlier publica-
7
tion concerning the solvolyses of 2,6-dichlorobenzoyl chlo-
ride shows an almost identical pattern as regards the deviations
from the best-fit correlation line. In particular, in both instances,
the points for the solvolysis in 100% ethanol, 90% ethanol,
Fig. 3 Plot of log (k/ko) for solvolyses of 2,4,6-trichlorobenzoyl chloride (5) in 27 solvents at 55.0 °C against (0.29 NT +0.62YCl)
.
Solvents: E, EtOH–H2O; M, MeOH–H2O; A, Acetone–H2O; TFE, TFE–H2O; HFIP, HFIP–H2O; T–E, TFE–EtOH.

652 JOURNAL OF CHEMICAL RESEARCH 2011
Fig. 4 Plot of log(k/ko) for solvolyses of 2,4,6-trichlorobenzoyl chloride (5) in 23 solvents at 55.0 °C against (0.40 NT + 0.82 YCl)
.
Solvents: E, EtOH–H2O; M, MeOH–H2O; A, Acetone–H2O; TFE, TFE–H2O; HFIP, HFIP–H2O; T–E, TFE–EtOH (100E, 90E, 100M, 90M are
omitted from the correlation; they are added to the plot to show the extent of their deviations.)
the ionisation mechanism. These are expressed below as the
percentage contribution of this pathway to the overall experi-
mental specific rate for the four omitted solvents. For 5, the
percentages of overall reaction calculated as proceeding by the
ionisation mechanism are 5% in 100% ethanol, 10% in 100%
methanol, 26% in 90% ethanol and 22% in 90% methanol. The
corresponding values for 3 in 26 solvents are 13% for 100%
ethanol, 14% for 100% methanol, 49% for 90% ethanol and
27% for 90% methanol. The contributions from the ionisation
pathway are somewhat larger for 3 than for 5 because the
introduction of the para-chloro substituent will, as indicated
by literature ρ values from related Hammett correlations,
slightly retard the ionisation pathway19,22,25 and slightly
accelerate the addition–elimination pathway.22,23,25
The kinetic measurements at 55.0 °C for solvolyses of 5 can
also be compared with those7 for 2,4-dichlorobenzoyl chloride
4. In aqueous ethanol, methanol, and acetone with less than
50% water content, 4 reacts faster than 5 (solubility problems
prevent studies in more aqueous mixtures). These observations
are, however, strongly reversed for solvolyses in water–
fluoroalcohol mixtures.
Conclusions
In ethanol, the 3,4,5-trimethoxybenzoyl chloride (1) reacts
marginally slower than 4-methoxybenzoyl chloride (2). The
extent to which 2 is favoured increases as water is added and
the rate ratio favouring 2 increases to a value of about 35 in
97-70% TFE. A direct linear free energy relationship (LFER)
correlation of log (k/ko) for 1 against log (k/ko) for 2 gives a
reasonably linear plot with a slope of 0.61 0.05. An extended
Grunwald–Winstein treatment (Eqn (1)) of the specific rates of
solvolysis of 1 shows that it has essentially the same l value
as for solvolyses for 2,8 and the source of the differences in
behaviour lies in very different m values of 0.54 ( 0.04) for 1
and 0.81 ( 0.02) for 2. It is suggested that the dominant influ-
ence differentiating the behaviour of 1 and 2 is a tightening of
the transition state due to the electron-withdrawing influence
of the two meta-methoxy groups. The lower specific rates
for 1 allow an extension to direct measurements in HFIP-H2O
solvents at 25.0 °C. These values could not be obtained2,6 ear-
lier due to a fast reaction of 2. Use of 1 as the substrate extends
the range of solvents which can be used in LFER correlation
of related solvolyses against log (k/ko) for para-substituted
methoxybenzoyl chlorides.
The 2,4,6-trichlorobenzoyl chloride (5) reacts, at 55.0 °C,
about one quarter as fast as 2,6-dichlorobenzoyl chloride 3 in
100% TFE, 97-50% TFE–H2O, and 97–70% HFIP–H2O (Table
1). The differences become smaller in pure alcohols and in
mixtures of water with ethanol, methanol, or acetone and in
100% and 90% ethanol, and in 80% acetone, the solvolyses of
5 are marginally faster than of 3. It is proposed that, due to the
steric hindrance from the ortho-substituents, in the majority of
solvents the ionisation pathway (Scheme 1) is dominant for
both 3 and 5 but in solvents of low ionising power and rela-
tively high nucleophilicity an appreciable contribution from
the addition–elimination pathway (Scheme 2) is operative.
We have shown previously7 that an analysis of specific rates
for solvolysis of 4 at 55.0 °C in terms of Eqn (1) leads to a
pronounced duality of mechanism, with 21 solvents favouring
the addition–elimination pathway (Scheme 2) and the ten
aqueous-fluoroalcohol solvents favouring the ionisation
pathway (Scheme 1). The details are included in Table 2. This
is consistent with a considerably lower steric hindrance to
nucleophilic attack with only one ortho-substituent. For 5, in
the presence of the second ortho-substituent, the majority of
the solvents now solvolyse with the ionisation pathway being
dominant.

JOURNAL OF CHEMICAL RESEARCH 2011 653
2
3
4
5
6
7
T.W. Bentley and H.C. Harris, J. Chem. Soc., Perkin Trans.2, 1986, 619.
T.W. Bentley and I.S. Koo, J. Chem. Soc., Perkin Trans.2, 1989, 1385.
T.W. Bentley and C.S. Shim, J. Chem. Soc., Perkin Trans.2, 1993, 1659.
K.-T. Liu and H.-I. Chen, J. Chem. Soc., Perkin Trans.2, 2000, 893.
T.W. Bentley and H.C. Harris, Int. J. Mol. Sci., 2011, 12, 4805.
K.-H. Park and D.N. Kevill, J. Phys. Org. Chem., 2011, 24, in press (doi:
10.1002/poc. 1851).
Comparing the calculated specific rates, using the l and m
values from entries 4 and 8 of Table 2, with the experimental
values, the percentages of reaction proceeding with ionisation
have been estimated for both 3 and 5 in 100% and 90%
ethanol, and in 100% and 90% methanol.
8
9
D.N. Kevill and M.J.D’Souza, J. Phys. Org. Chem., 2002, 15, 881.
S. Winstein, E. Grunwald and H.W. Jones, J. Am. Chem. Soc., 1951, 73,
2700.
Experimental
The 3,4,5-trimethoxybenzoyl chloride (1, 98%) and the 2,4,6-trichlo-
robenzoyl chloride (5, 97%) were obtained from Sigma-Aldrich
Chemical Company and they were used as received. Solvents were
purified as described previously.11 The determination of the kinetics of
10 F.L. Schadt, T.W. Bentley and P.v.R. Schleyer, J. Am. Chem. Soc., 1976, 98,
7667.
11 D.N. Kevill and S.W. Anderson, J. Org. Chem., 1991, 56, 1845.
12 D.N. Kevill and M.J. D’Souza, J. Chem. Res., 2008, 61.
13 E. Grunwald and S. Winstein, J. Am. Chem. Soc., 1948, 70, 846.
14 T.W. Bentley and G. Llewellyn, Prog. Phys. Org. Chem., 1990, 17, 121.
15 D.N. Kevill Advances in quantitative structure-property relationships, ed.
M. Charton, JAI Press, Greenwich, CT, 1996, pp. 81–115.
16 T.W. Bentley and H.C. Harris, J. Org. Chem., 1988, 53, 724.
17 J. Shorter, Correlation analysis of organic reactivity, Research Studies
Press, New York, NY, 1982, pp. 169–170.
18 M.H. Aslam, N.B. Chapman, M. Charton and J. Shorter, J. Chem. Res.,
1981, (S) 188, (M) 2301.
19 J.E. Leffler and E. Grunwald, Rates and equilibria of organic reactions,
Wiley, New York, NY, 1963, pp. 203–210.
7
the solvolysis reactions and the calculation of the specific rates of
solvolysis7,26 were as previously reported.
We thank Professor Malcolm J. D’Souza (Wesley College,
Dover, Delaware, USA) for updating his previous analyses
by incorporating recently reported additional specific rates of
solvolysis of p-methoxybenzoyl chloride (2). This article is
abstracted, in part, from the PhD Dissertation of Kyoung-Ho
Park, Northern Illinois University, May 2009.
Received 22 September 2011; accepted 14 October 2011
Paper 1100901 doi: 10.3184/174751911X13192929905095
Published online: 22 November 2011
20 G. Illuminati, J. Am. Chem. Soc., 1958, 80, 4941.
21 S.C.J. Olivier, Rec. Trav. Chim., 1929, 48, 227.
22 M.L. Bender and M.C. Chen, J. Am. Chem. Soc., 1963, 85, 30.
23 A. Kivinen, The chemistry of acyl halides, ed. S. Patai, Interscience, New
York, 1972, p. 193.
24 F. Ruff and O. Farkas, J. Phys. Org. Chem., 2011, 24, 480.
25 B.D. Song and W.P. Jencks, J. Am. Chem. Soc., 1989, 111, 8470.
26 H.J. Koh, K.L. Han, H.W. Lee and I. Lee, J. Org. Chem., 1998, 63, 9834.
References
1
T.W. Bentley, G.E. Carter and H.C. Harris, J. Chem. Soc., Perkin Trans.2,
1985, 983.

Copyright of Journal of Chemical Research is the property of Science Reviews 2000 Ltd. and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.