Solvolysis of 1-alkyl-1-chloro-1-(4-methyl)phenylmethanes. Nucleophilic solvent intervention and extended YBnCl scale

Article abstract of DOI:10.1002/(SICI)1099-1395(199712)10:12<879::AID-POC879>3.0.CO;2-K

The solvolysis of 1-alkyl-1-chloro-1-(4-methyl)phenylmethanes (4a-d) in aqueous acetone, aqueous ethanol, aqueous methanol and ethanol-trifluoroethanol was studied. Grunwald-Winstein-type correlation analysis using the Y_BnCl scale suggests significant nucleophilic solvent intervention in the case of 1-chloro-1-(4-methyl)phenylethane (4a). Increasing bulkiness of the 1-alkyl substituent from methyl (4a) to ethyl (4b), to isopropyl (4c) and to tert-butyl (4d) resulted in a gradual change to limiting S_N1 mechanisms. The observed excellent linear correlations with Y_BnCl and the good solubility in high-water-containing binary solvents made 4d a suitable reference standard for deriving more Y_BnCl values. A positive azide salt effect was realized in the solvolysis of 4a but not 4d. A small decrease in the β-deuterium kinetic isotope effect from 4a to 1-chloro-1-(4-methoxyphenyl)propane (5) suggested the presence of additional stabilization of the benzylic cationic transition state. However, no relationship between k(CH₃)/k(CD₃) and the solvent effect was found. The superiority of employing the Y_BnCl scale over the combination of Y_Cl and I scales in the mehanistic study was observed.

Full text of DOI:10.1002/(SICI)1099-1395(199712)10:12<879::AID-POC879>3.0.CO;2-K

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)
SOLVOLYSIS OF 1-ALKYL-1-CHLORO-
1-(4-METHYL)PHENYLMETHANES. NUCLEOPHILIC SOLVENT
INTERVENTION AND EXTENDED YBnCl SCALE
KWANG-TING LIU,* LIH-WIE CHANG, DER-GANN YU, PANG-SHAO CHEN AND JU-TA FAN
Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, Republic of China
The solvolysis of 1-alkyl-1-chloro-1-(4-methyl)phenylmethanes (4a–d) in aqueous acetone, aqueous ethanol, aqueous
methanol and ethanol–trifluoroethanol was studied. Grunwald–Winstein-type correlation analysis using the YBnCl scale
suggests significant nucleophilic solvent intervention in the case of 1-chloro-1-(4-methyl)phenylethane (4a). Increasing
bulkiness of the 1-alkyl substituent from methyl (4a) to ethyl (4b), to isopropyl (4c) and to tert-butyl (4d) resulted in a
gradual change to limiting SN1 mechanisms. The observed excellent linear correlations with YBnCl and the good solubility
in high-water-containing binary solvents made 4d a suitable reference standard for deriving more YBnCl values. A positive
azide salt effect was realized in the solvolysis of 4a but not 4d. A small decrease in the ␤-deuterium kinetic isotope effect
from 4a to 1-chloro-1-(4-methoxyphenyl)propane (5) suggested the presence of additional stabilization of the benzylic
cationic transition state. Howver, no relationship between k(CH3 )/k(CD3 ) and the solvent effect was found. The
superiority of employing the YBnCl scale over the combination of YCl and I scales in the mehanistic study was observed.
© 1997 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 10, 879–884 (1997) No. of Figures: 3 No. of Tables: 6 No. of References: 27
Keywords: solvolysis; correlation analysis; azide salt effect; ␤-deuterium kinetic isotope effect; YBnCl; 1-alkyl-1-chloro-
1-(4-methyl)phenylmethanes
Received 31 March 1997; revised 9 May 1997; accepted 19 May 1997
INTRODUCTION
they proposed the use of new multiparameter equations (3
and 4) containing the aromatic ring parameter I13 to
interpret the solvolytic behavior of benzylic substrates:
Correlation analysis with the Grunwald–Winstein equation
(1)1 or (2)2 has long been employed as a diagnostic tool for
the study of solvent effects on solvolytic mechanisms:
log (k/k0 )=mY+hI
(3)
(4)
log (k/k0 )=mY
(1)
log (k/k0 )=mY+lN+hI
log (k/k0 )=mY+lN
(2)
However, several drawbacks and disadvantages of such
approaches have been pointed out recently.14 More work
seems to be needed to ascertain which one is more suitable
for interpreting solvolytic mechanisms.
We have also extended our studies on benzylic solvolyses
to several other systems. One of them was 1-alkyl-1-aryl-
A variety of scales of solvent ionizing power Y, a kind of
empirical solvent polarity scale,3 have been developed.4
Recently, new YBnX scales were suggested to be superior to
YX scales in the study of several secondary and tertiary
benzylic bromides,5 chlorides,6 p-nitrobenzoates7 and tosy-
lates.8
1-halomethanes.
Previous
work
on
both
Depression of the data point corresponding to log k
values measured in an isodielectric ethanol–trifluoroethanol
mixture9 from the linear plots was observed for the
solvolysis of several 2-aryl-2-propyl derivatives (1), and
nucleophilic solvent intervention was then deduced from the
variation in the electronic and steric factors of sub-
stituents.10, 11 Kevill and co-workers,12 on the other hand,
suggested a general preference for YX scales. In addition,
1-aryl-1-chloroethanes (2)15 and 1-aryl-1-tert-butyl-
1-chloromethane (3)16 under different reaction conditions
led to the conclusion that an SN1 mechanism occurred in
most cases. The kinetic solvent isotope effect on the
methanolysis of 2 suggested ion-pair formation with the
attack of the solvent molecule at the carbocation in the pre-
equilibrium.15e On the other hand, an SN2+ mechanism was
found in the acetolysis of 1-tert-butyl-1-chloro-
1-(4-methoxy)phenylmethane.16b Addition of thiourea in the
solvolysis of 1-chloro-1-phenylethane and 1-tert-butyl-
* Correspondence to: K.-T. Liu.
© 1997 John Wiley & Sons, Ltd.
CCC 0894–3230/97/120879–06 $17.50

880
K.-T. LIU ET AL.
change in 1-chloro-1-(4-methoxyphenyl)propane (5) were synthesized
1-chloro-1-phenylmethane resulted in
mechanism.17
a
by Grignard addition of arylmagnesium bromide to appro-
priate acyl chlorides, followed by reduction with sodium
borohydride and then chlorination with thionyl chloride.
Grignard reaction of methyl-d3-magnesium iodide with the
corresponding benzaldehyde, followed by the treatment
In this paper, we provide kinetic evidence to demonstrate
the presence of nucleophilic solvent intervention in the
solvolysis of 1-alkyl-1-chloro-1-(4-methylphenyl)methane
(4) and the importance of steric hindrance to this
effect. Moreover, an extended YBnCl scale to include the
values for high-water-containing solvents based on the
solvolytic reactivity of 1-tert-butyl-1-chloro-1-(4-methyl-
phenyl)methane (4d) is provided to circumvent the
limitation of the use of 2-aryl-2-chloroadamantane6 as a
reference standard.
RESULTS
The desired substrates were prepared by conventional
methods. 1-Alkyl-1-aryl-1-chloromethanes (4a–d) and
Table 1. Solvolytic rate constants for chlorides 4a–d
b
k(25 °C) (sϪ1
)
Solventa
4a
4b
4c
4d
100E
90E
80E
70E
60E
1·04ϫ10Ϫ5
1·22ϫ10Ϫ4
5·55ϫ10Ϫ4
1·61ϫ10Ϫ3
5·10ϫ10Ϫ3
9·78ϫ10Ϫ7 c
3·80ϫ10Ϫ4 9·86ϫ10Ϫ5 3·86ϫ10Ϫ6 d
1·27ϫ10Ϫ3 3·46ϫ10Ϫ4 1·18ϫ10Ϫ5
4·57ϫ10Ϫ3 1·37ϫ10Ϫ3 4·96ϫ10Ϫ5
2·56ϫ10Ϫ4
50E
40E
80A
70A
60A
50A
40A
100M
90M
80M
70M
60M
50M
100T
3·25ϫ10Ϫ5
2·42ϫ10Ϫ4
1·10ϫ10Ϫ3
4·97ϫ10Ϫ5
4·59ϫ10Ϫ7 e
3·02ϫ10Ϫ4 8·26ϫ10Ϫ5 2·73ϫ10Ϫ6 f
1·58ϫ10Ϫ3 4·98ϫ10Ϫ4 1·60ϫ10Ϫ5
1·15ϫ10Ϫ4
1·24ϫ10Ϫ4
6·48ϫ10Ϫ4
2·21ϫ10Ϫ3
1·74ϫ10Ϫ4 5·13ϫ10Ϫ5 2·57ϫ10Ϫ6 g
7·29ϫ10Ϫ4 2·36ϫ10Ϫ4 1·01ϫ10Ϫ5 h
3·68ϫ10Ϫ5
8·22ϫ10Ϫ3 2·89ϫ10Ϫ3 1·26ϫ10Ϫ4
4·25ϫ10Ϫ4
1·00ϫ10Ϫ1
3·88ϫ10Ϫ2 2·83ϫ10Ϫ2 1·27ϫ10Ϫ3
4·21ϫ10Ϫ3 2·33ϫ10Ϫ3 9·79ϫ10Ϫ5
5·34ϫ10Ϫ4 2·49ϫ10Ϫ4 1·04ϫ10Ϫ5 i
1·36ϫ10Ϫ3
80T–20E 9·43ϫ10Ϫ3
60T–40E 1·12ϫ10Ϫ3
70Tw
50Tv
1·59ϫ10Ϫ3
a Abbreviations: A, acetone; E, ethanol; M, methanol; T, 2,2,2-trifluoroethanol. The
numbers denote the volume percentage of the specific solvent in the solvent
mixture.
b Maximum error ±3%.
c Calculated from k=5.03ϫ10Ϫ4 sϪ1 (75 °C), k=1·70ϫ10Ϫ4 sϪ1 (65 °C) and
k=5·19ϫ10Ϫ5 sϪ1 (55 °C).
d Calculated from k=7·86ϫ10Ϫ4 sϪ1 (70 °C), k=1·64ϫ10Ϫ4 sϪ1 (55 °C) and
k=4·88ϫ10Ϫ5 sϪ1 (45 °C).
e Calculated from k=2·15ϫ10Ϫ4 sϪ1 (75 °C), k=7·25ϫ10Ϫ5 sϪ1 (65 °C) and
k=2·30ϫ10Ϫ5 sϪ1 (55 °C).
f Calculated from k=3·23ϫ10Ϫ4 sϪ1 (65 °C) and k=3·45ϫ10Ϫ5 sϪ1 (45 °C).
g Calculated from k=8·36ϫ10Ϫ4 sϪ1 (75 °C), k=2·93ϫ10Ϫ4 sϪ1 (65 °C),
k=1·09ϫ10Ϫ4 sϪ1 (55 °C) and k=3·13ϫ10Ϫ5 sϪ1 (45 °C).
h Calculated from k=3·65ϫ10Ϫ4 sϪ1 (55 °C), k=1·24ϫ10Ϫ4 sϪ1 (45 °C) and
k=3·59ϫ10Ϫ5 sϪ1 (35 °C).
i Calculated from k=3·25ϫ10Ϫ4 sϪ1 (55 °C) and k=1·11ϫ10Ϫ4 sϪ1 (45 °C).
© 1997 John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)

SOLVOLYSIS OF 1-ALKYL-1-CHLORO-1-(4-METHYL)PHENYLMETHANES
881
a
Table 2. Correlation analyses against YBnCl and NT
with thionyl chloride, yielded the ␣-trideuteriomethyl
compounds 4a-d3 and 5-d3 . Solvolytic rate constants within
±2% error were obtained by the conductimetric method
unless mentioned otherwise. Pertinent data are reported in
Table 1.
YBnCl
m
YBnCl and NT
Substrate
m
l
Good agreement of the rate constants with literature
4a
4b
0·734 (R=0·972, n=14) 0·834 0·146 (R=0·976, n=14)
0·735 (R=0·975, n=11) 0·817 0·098 (R=0·978, n=11)
0·756 (R=0.965, n=12)
data15a for 4a and 5 was found. Correlation analyses of log k
against YBnCl [equation (1)]6b and against YBnCl and NT
18
[equation (2)] were carried out and the results are given in
Table 2. Plots of log ks vs YBnCl for 4a and b are given in
Figures 1 and 2, respectively. Comparison of the correlation
analysis using equation (1) with YCl ,4 equation (2) with YCl
and NT,18 equation (3) with YCl and I13b and equation (4) with
YCl , NT and I is shown in Table 3, although only a limited
number of data are available for the multiparameter
correlation equation (4).
4c
0·835 (R=0·988, n=10) 0·830 Ϫ0·005 (R=0·988, n=10)
0·856 (R=0·983, n=11)
4d
0·886 (R=0·996, n=12)
0·884 (R=0·997, n=15)b
a Standard deviation is between 0·02 and 0·08.
b Using the data in Table 6.
Kinetic salt effects induced by adding sodium azide and
sodium perchlorate, respectively, were monitored titramet-
rically for 4a and d. The data are presented in Table 4.
For the study of the ␤-deuterium kinetic isotope effect,
the substrate pairs 4a and 4a-d3 and 5 and 5-d3 were
solvolyzed under identical experimental conditions and
were measured conductimetrically. The reproducibility of
the measured individual rate constant was found within
±1% error in triplicate experiments. The results are given in
Table 5.
DISCUSSION
The comparison of Tables 2 and 3 clearly indicates that the
correlation of logk with YCl is not good (R=0·92–0·96) and
shows a random change for substrates 4a–4d, while the
variations of both m and R show a reasonably increasing
trend from the correlation with YBnCl . The excellent linear
correlation19 (R=0·996–0·997) with m(YBnCl ) of about 0·9
observed in the case of 4d suggests that a limiting SN1
process is involved in the solvolysis. The m value decreases
and the correlation changes from excellent (R>0·99) to less
satisfactory (R=0·96–0·97) as the bulkiness of the ␣-alkyl
substituent decreases from tert-butyl to ethyl and methyl
groups. The solvolysis mechanisms are likely to be non-
limiting for 4a and 4b and to be a borderline case for 4c.
Steric hindrance is therefore probably an important factor
governing the solvolytic mechanism. The coalescence of
lines in the log k–YBnCl plots as the ␣-alkyl substituent
becomes bulkier, e.g. tert-butyl (Figure 3), could be
interpreted by the diminishing intervention of the nucleo-
philic solvent due to steric hindrance, in harmony with what
has been observed in the solvolysis of 1-alkyl-1-chloro-
1-phenylethanes.11b
Figure 1. Correlations of log k for 4a against YBnCl . Abbreviations
for solvents as in Table 1.
Splitting of lines made from data points measured in
different solvent systems was found in the log k–YBnCl plots
for 4a (Figure 1) and 4b (Figure 2), in which the line
defined by those measured in isodielectric and low nucleo-
philic ethanol–trifluoroethanol solvent systems was with the
smallest m value (0·822 for 4a and 0·784 for 4b). Obviously,
nucleophilic solvent intervention is present in those cases as
Figure 2. Correlations of log k for 4b against YBnCl . Abbreviations
for solvents as in Table 1.
© 1997 John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)

882
K.-T. LIU ET AL.
Table 3. Correlation analyses against YCl , NT and Ia
YCl
m
YCl
m
NT
I
Substrate
l
h
h/m
4a
0·732 (R=0·964, n=14) 0·690
Ϫ0·053
(R=0·965, n=14)
(R=0·976, n=13)
(R=0·978, n=13)
(R=0·942, n=11)
(R=0·973, n=11)
(R=0·980, n=11) 0·837
(R=0·950, n=10)
(R=0·980, n=10)
(R=0·980, n=10)
(R=0·952, n=15)
(R=0·983, n=14)
(R=0·984, n=14)
0·715
0·751
0·580
0·640
0·811
0·852
0·059
Ϫ0·070
4b
4c
4d
0·849 (R=0·943, n=12)
0·788
0·827
0·932
0·628
0·780
0·759
0·119
Ϫ0·263
0·869 (R=0·919, n=11) 0·618
0·895
0·892
0·871
0·877
0·973
0·983
Ϫ0·003
Ϫ0·217
0·944 (R=0·943, n=18)
0·806
0·945
0·999
1·003
1·12
1·06
1·12
0·077
a Standard deviations are 0·05–0·17 for m, 0·09–0·13 for l and 0·24–0·30 for h.
in the solvolysis of tertiary benzylic substrates.10, 11 Applica-
tion of the dual-parameter equation (2) with YBnCl and NT
(Table 2) indicated little improvement, although a decrease
in l could be found from 4a to b. With YCl and NT (Table 3),
even worse results including negative l values were
obtained. Therefore, factors other than the solvent nucleo-
philicity would be also influential. The outcome of the use
of equations (3) and (4) (Table 3) was not any better.
Irregular variation of l values and h/m ratios for 4a–d
suggested that those types of correlation including the
aromatic ring character I were insensitive to structural
change, and hence less useful in mechanistic studies than
the correlation with YBnCl
.
The study of salt effects20 with sodium azide and sodium
perchlorate indicated a significant rate enhancement for 4a
but not for 4d in aqueous acetone (Table 4). The observed
azide salt effects, k(NaN3 )/k(NaClO4 ) at equal salt-to-
substrate ratios, were 1·41 (0·04
1·63 (0·08 [salt]/0·005 [RCl]) in the case of 4a.
Solvolysis of 4d, however, showed only a negligible effect
of 1·03 at similar salt-to-substrate ratio (0·02 [salt]/
0·003 [RCl]). In agreement with the result of correlation
M
[salt]/0·005
M
[RCl]) and
M
M
M
M
Table 4. Salt effects in solvolyses of 4a and d
analysis with equation (2) (see above), the significant azide
salt effect suggests the involvement of nucleophilically
assisted process in solvolysis of 4a.
[Salt] k(NaN3 ) k(NaClO4 )
RN3
Substrate
(
M)
(sϪ1
)
(sϪ1
)
k(NaN3 )/k(NaClO4 ) (%)
The ␤-deuterium kinetic isotope effect has been con-
4ab
0
3·25ϫ10Ϫ5 3·25ϫ10Ϫ5
sidered
a
useful tool for elucidating solvolytic
0·04 5·78ϫ10Ϫ5 4·10ϫ10Ϫ5
0·08 7·21ϫ10Ϫ5 4·42ϫ10Ϫ5
1·41
1·63
20·8
31·4
mechanisms.21, 22 In a study of the solvolysis of 2, Shiner et
al.15a observed a small decrease in k(CH3 )/k(CD3 ) with
increasing electron-donating character of substituents from
hydrogen to p-methoxy and also with increasing electron-
withdrawing character of substituents from hydrogen to
p-nitro in the aromatic moiety. A change in the demand of
hyperconjugative stabilization from the ␤-methyl group to
4dc
0
3·45ϫ10Ϫ5 3·45ϫ10Ϫ5
0·02 4·09ϫ10Ϫ5 3·95ϫ10Ϫ5
1·03
a By titrametric method, maximum error is ±3%.
b 0·005
c 0·003
M
in 80% acetone at 25 °C.
in 60% acetone at 45 °C
M
Table 5. ␤-Deuterium kinetic isotope effects in solvolysis of 4a and 5 at 25 °C
Substrate Solventa
k(CH3 ) (sϪ1
)
k(CD3 ) (sϪ1
)
k(CH3 )/k(CD3 )
4a
4a
4a
4a
4a
5
100E
90E
80E
80A
70A
100E
90E
(1·044±0·012)ϫ10Ϫ5 (8·218±0·008)ϫ10Ϫ5
(1·223±0·017)ϫ10Ϫ4 (1·022±0·014)ϫ10Ϫ4
(5·551±0·028)ϫ10Ϫ4 (4·809±0·053)ϫ10Ϫ4
(3·252±0·016)ϫ10Ϫ5 (2·736±0·013)ϫ10Ϫ5
(2·421±0·024)ϫ10Ϫ4 (2·065±0·036)ϫ10Ϫ4
(2·168±0·008)ϫ10Ϫ2 (1·958±0·014)ϫ10Ϫ2
(2·453±0·004)ϫ10Ϫ1 (2·295±0·016)ϫ10Ϫ1
(7·213±0·023)ϫ10Ϫ2 (6·412±0·037)ϫ10Ϫ2
1·270±0·015
1·197±0·023
1·155±0·014
1·189±0·008
1·141±0·022
1·107±0·009
1·068±0·008
1·125±0·007
5
5
80A
a See Table 1 for abbreviations.
© 1997 John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)

SOLVOLYSIS OF 1-ALKYL-1-CHLORO-1-(4-METHYL)PHENYLMETHANES
883
Table 6. Compilation of extended YBnCl values
varies significantly with solvent composition. Moreover, the
trends of the variation are not the same for different
substrates. For instance, the ratio k(CH3 )/k(CD3 ) in 90%
ethanol is greater than that in 80% acetone for 4a, but a
reversed order is found for 5. Therefore, a decrease in the
ratio k(CH3 (/k(CD3 ) cannot be considered a necessary
condition for nucleophilic attack of solvents. In other words,
a decrease in the ␤-deuterium kinetic isotope effect could
indicate a larger electronic stabilization due to electron-
donating substituents, but could not be considered
diagnostic for determining nucleophilic mechanisms. Varia-
tion of hyperconjugative stabilization with the solvent might
account for the observed correlation coefficient R of
0·976–0·978 in the regression with dual-parameter equation
(2) (Table 2) for the solvolysis of 4a and 4b.
The importance of nucleophilic solvent intervention in
the solvolysis of 1-alkyl-1-chloro-1-(4-methyl)phenyl-
methanes diminishes with increasing bulkiness of the
1-alkyl group. The present study confirms the limiting SN1
character for the solvolysis of 1-tert-butyl-1-chloro-
1-(4-methyl)phenyl-methane (4d) in acetone–water,
ethanol–water, methanol–water and ethanol–trifluoro-
ethanol solvent systems. Excellent linear mY correlations
with the YBnCl scale, R=0·996, are realized from Table 2.
Since the YBnCl scale was developed with the use of
2-chloro-2-(3-chloro)phenyladamantane and other 2-aryl-
2-chloroadamantanes, the value for high-water-containing
solvents is not available owing to their high reactivities and
low solubilities. Consequently, more values of the YBnCl
scale can be obtained from interpolation of the log k (4d)–
YBnCl plots (Figure 3). The data in 30% acetone, 30%
ethanol, 40% methanol or solvents containing more water
were not reliable, and were therefore discarded, because of
the low solubility of 4d in these solvents at 25 °C. A
compilation of available YBnCl values is given in Table 6.
b
b
Solventa
YBnCl
Solventa
YBnCl
100E
90E
80E
70E
60E
50E
40E
90A
80A
70A
60A
50A
40A
100M
90M
Ϫ1·61
Ϫ0·645
0·00
80M
70M
60M
1·12
1·72c
2·37c
2·97c
3·55
0·571
1·07
50M
100T
85T
70T
1·92c
3·57
2·72c
3·60
Ϫ2·28
Ϫ1·09
Ϫ0·259
0·518
1·23
50T
70Tw
3·62c
3·54
80T–20E
60T–40E
40T–60E
i-PrOH
HOAc
Formamide
2·42
1·18
0·18
2·33c
Ϫ2·71
Ϫ1·97
1·38
Ϫ0·253
0·585
a Abbreviations: A, acetone; E, ethanol; M, methanol; T, 2,2,2-trifluor-
oethanol; HOAc, acetic acid; i-PrOH, propan-2-ol. The numbers denote the
volume percentage of the specific solvent in the solvent mixture with the
exception of 70Tw, which denotes weight percentage.
b From Ref. 6b unless mentioned otherwise.
c This work.
EXPERIMENTAL
Capillary melting points were uncorrected. NMR spectra
were determined on a Bruker Model AM-300 or AC-200
instrument using tetramethylsilane as an internal standard.
IR spectra were measured on a Perkin-Elmer Model 983G
spectrometer.
Figure 3. Correlations of log k for 4d against YBnCl . The open
circles denote the points of interpolation. Abbreviations for solvents
as in Table 1.
the benzylic reaction in solvolysis and a partly nucleophilic
mechanism for substrates containing m-Br and p-NO2
substituents were concluded to be involved.15a The proposal
of a decreasing hyperconjugative effect with electron-
donating substituents was confirmed in the present study
(Table 5), and had also been found with certain tertiary
benzylic systems.23 However, the proposed relationship
between the decrease in k(CH3 )/k(CD3 ) and the presence of
nucleophilic solvent intervention15a is questionable. The
extrapolation of rate data from solvents of different ionizing
power and nucleophilicity to give ‘isotope effects corrected
to the value expected in 50% ethanol’15a seems to be
unreliable. In the present study, the data in Table 5 reveal
that the magnitude of ␤-deuterium kinetic isotope effects
Materials. Spectral-grade or reagent-grade solvents
were purified according to standard methods24 for kinetic
studies. Dry solvents for preparative purposes were freshly
distilled before use if necessary. Doubly deionized water
was used to prepare aqueous solvent mixtures for sol-
volysis. Starting materials for the preparation of substrates
were reagent-grade chemicals available commercially.
The chlorides 4a–d and 5 were synthesized according to
the following conventional procedures: Grignard addition of
the corresponding arylmagnesium bromide to the appro-
priate acyl chlorides in tetrahydrofuran to give the ketones,
which were purified by column chromatography on silica
gel. The ketone was then reduced with sodium borohydride
© 1997 John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)

884
K.-T. LIU ET AL.
M.-L. Tsao, J. Org. Chem. 57, 3041 (1992).
9. J. Kaspi and Z. Rappoport, J. Am. Chem. Soc. 102, 3829
(1980).
10. (a) K.-T. Liu, P.-S. Chen, P.-F. Chiu and M.-L. Tsao,
Tetrahedron Lett. 33, 6499 (1992); (b) K.-T. Liu, P.-S. Chen,
C.-R. Hu and H.-C. Sheu, J. Phys. Org. Chem. 6, 122, 433
(1993).
11. (a) K.-T. Liu, L.-W. Chang and P.-S. Chen, J. Org. Chem. 57,
4791 (1992); (b) K.-T. Liu and P.-S. Chen, J. Chem. Res. (S)
310 (1994).
12. (a) D. N. Kevill and M. J. D’Souza, J. Phys. Org. Chem. 5, 287
(1992); (b) D. N. Kevill and M. J. D’Souza, J. Phys. Org.
Chem. 6, 642 (1993); (c) D. N. Kevill and M. J. D’Souza, J.
Chem. Res. (S) 332 (1993).
13. (a) D. N. Kevill, N. HJ Ismail and M. J. D’Souza, J. Org.
Chem. 59, 6303 (1994); (b) D. N. Kevill and M. J. D’Souza, J.
Chem. Soc., Perkin Trans. 2 973 (1995); (c) D. N. Kevill and
M. J. D’Souza, J. Chem. Res. (M) 1649 (1996).
in methanol to form the desired alcohol, followed by
chlorination with thionyl chloride in carbon tetrachloride.
The resultant chloride was purified by recrystallization or by
column chromatography on triethylamine-washed silica gel
with hexane–ethyl acetate as eluent. 1-Chloro-1-(4-methyl-
phenyl)ethane (4a),15a 1-chloro-1-(4-methylphenyl)propane
(4b),25 1-chloro-2-methyl-1-(4-methylphenyl)propane (4c)26
and 1-chloro-1-(4-methoxyphenyl)ethane (5)15a were
obtained in the pure liquid state and 1-chloro-2,2-dimethyl-
1-(4-methylphenyl)propane (4d)27 was isolated as a solid,
m.p. 28–29 °C. Their IR and proton and carbon NMR data
were found to be in line with the assigned structures.
The ␣-trideuteriomethyl compounds 4a-d3 and 5-d3 were
prepared by Grignard reaction of methyl-d3-magnesium
iodide (99% isotopically pure) with the corresponding
benzaldehyde, followed by the treatment of the resultant
alcohol with thionyl chloride.
14. (a) K.-T. Liu, J. Chem. Soc., Perkin Trans. 2 327 (1996); (b)
K.-T. Liu, C.-P. Chin, Y.-S. Lin and M.-L. Tsao, J. Chem. Res.
(S) 18 (1997).
Kinetic measurements. Conductimetric rate constants
15. For examples, see (a) V. J. Shiner, Jr, W. E. Buddenbaum, B. L.
Murr and G. Lamaty, J. Am. Chem. Soc. 90, 418 (1968); (b) Y.
Tsuno, Y. Kusuyama, M. Sawada, T. Fujii and Y. Yukawa, Bull
Chem. Soc. Jpn. 48, 3337 (1975); (c) J. P. Richard, M. E.
Rothenberg and W. P. Jencks, J. Am. Chem. Soc. 106, 1361
(1984); (d) D. J. McLennan, A. R. Stein and B. Dobson, Can.
J. Chem. 64, 1201 (1986); (e) I. Lee, W. H. Lee and H. W. Lee,
J. Phys. Org. Chem. 6, 361 (1993); (f) T. W. Bentley, M.
Christl, R. Kemmer, G. Llewellyn and J. E. Oakley, J. Chem.
Soc., Perkin Trans. 2 2531 (1994).
were measured for general solvolytic studies. The con-
ductivity cells were placed in
temperature variation of ±0·02 °C.
1ϫ10Ϫ4–2ϫ10Ϫ5
was used. In some cases, a small
a
thermostat with
a
A
solution of
M
amount (0·1%) of 2,6-lutidine was added to the solution to
prevent curvature of the rate constant plot. For studying the
kinetic isotope effect, the solvolyses of the isotopic pairs
were run side-by-side in the same thermostat. The potentio-
metric titration method was employed in the study of salt
16. For examples, see (a) S. Winstein and B. K. Morse, J. Am.
Chem. Soc. 74, 1133 (1952); (b) T. Kinoshita, H. Ueda and K.
Takeuchi, J. Chem. Soc., Perkin Trans 2 603 (1993).
17. K. Yatsugi, I. Akasaka, Y. Tsuji, S.-H. Kim, S.-D. Yoh, N.
Sugiyama, M. Mishima, M. Fujio and Y. Tsuno, Tetrahedron
Lett. 35, 135 (1994).
effects. The concentration of substrate was 0·003–0·005 .
M
ACKNOWLEDGEMENTS
We are grateful to the National Science Council, Republic
of China, for financial support of this research, and to Mr
Hung-I Chen for checking some rate data.
18. D. N. Kevill and S. W. Anderson, J. Org. Chem. 56, 1845
(1991).
19. For definition, see H. H. Jaffe, Chem. Rev. 53, 191 (1953).
20. For examples of pioneering work, see (a) D. J. Raber, J. M.
Harris, R. E. Hall and P. v. R. Schleyer, J. Am. Chem. Soc. 93,
4821 (1971); (b) R. A. Sneen, Acc. Chem. Res. 6, 46 (1973).
21. C. J. Collins and N. S. Bowman (Eds) Isotope Effects in
Chemical Reactions, ACS Monograph No. 167. Van Nostrand
Reinhold, New York (1970).
REFERENCES
1. E. Grunwald and S. Winstein, J. Am. Chem. Soc. 70, 846
(1948).
2. S. Winstein, E. Grunwald and H. W. Jones, J. Am. Chem. Soc.
73, 2700 (1951).
3. For review, see: C. Reichardt, Solvents and Solvent Effects in
Organic Chemistry, 2nd ed. Chapt. 7. VCH, Weinheim (1988).
4. For the latest review, see T. W. Bentley and G. Llewellyn, Prog.
Phys. Org. Chem. 17, 121 (1990).
5. K.-T. Liu and H.-C. Sheu, J. Chin. Chem. Soc. 38, 29 (1991).
6. (a)K.-T. Liu, H.-C. Sheu, H.-I Chiu and C.-R. Hu, Tetrahedron
Lett. 25, 3611 (1990); (b) K.-T. Liu and H.-C. Sheu, J. Org.
Chem. 56, 3021 (1991).
22. D. E. Sunko and W. J. Hehre, Prog. Phys. Org. Chem. 14, 205
(1983).
23. K.-T. Liu and M.-Y. Kuo, Tetrahedron 44, 3523 (1988).
24. D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, 3rd ed. Pergamon Press, New York (1988).
25. S. Cabiddu, A. Maccioni, M. Secci and F. Sotgiu, Ann. Chim.
(Rome) 63, 675 (1973); Chem. Abstr. 82, 72594n (1975).
26. NV Naarden International, Neth. Pat. Appl 78 02 038 (1978);
Chem. Abstr. 90, 71951e (1979).
7. K.-T. Liu, H.-I Chen and C.-P. Chin, J. Phys. Org. Chem. 4,
463 (1991).
8. K.-T. Liu, J.-S. Yang, S.-M. Chang, Y.-S. Lin, H.-C. Sheu and
27. M. R. Nair and S. W. Nair, Indian J. Chem. 19B, 1091 (1980);
Chem. Abstr. 91, 180459n (1980).
© 1997 John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 879–884 (1997)