Use of kinetic isotope effects in mechanism studies. Isotope effects and element effects associated with Hydron-Transfer steps during alkoxide-promoted dehydrohalogenations

Article abstract of DOI:10.1021/ja970189n

The Arrhenius behavior of the primary kinetic isotope effect, (k(H)/k(D))(Obs) and (k(H)/k(T))(Obs), associated with the methanolic sodium methoxide-promoted dehydrohalogenations of m-ClC₆H₄C(i)HClCH₂Cl (I), m-CF₃C₆H₄C(i)-HClCH₂Cl (II) and p-CF₃C₆H₄C(i)HClCH₂F (III) has been used to calculate the internal-return parameters, a = k(-1)/K(Elim)(X), in a two-step mechanism featuring a hydrogen-bonded carbanion. This carbanion partitions between returning the hydron to carbon, k(-1), and the loss of halide, K(Elm)(X). Isotope effects at 25°C for I, (k(H)/k(D))(Obs) = 3.40 and (k(H)/ k(T))(Obs) = 6.20, and II, (k(H)/k(D))(Obs) = 3.49 and (k(H)/k(T))(Obs) = 6.55, result in similar values for a: a(H) = 0.59, a(D) = 0.13-0.14 and a(T) = 0.07. Smaller values of (k(H)/k(D))(Obs) = 2.19 and (k(H)/k(T))(Obs) = 3.56 for III are due to more internal return [a(H) = 1.9, a(D) = 0.50, and a(T) = 0.28] associated with the dehydrofluorination reaction. Calculation of k₁ ( k(Obs) [a + 1]) results in similar isotope effects for hydron transfer in these reactions: k₁(H)/k₁(D) = 4.74 and k₁(H)/K₁(T) = 9.20; II, k₁(H)/k₁(D) = 4.91 and k₁(H)/k₁(T) = 9.75; III, k₁(H)/k₁(D) = 4.75 and k₁(H)/k₁(T) = 9.17. Reactions of m-ClC₆H₄C(i)HBrCH₂Br and m-ClC₆H₄C(i)HClCH₂Br have very small amounts of internal return, a(H) = 0.05 and a(D) = 0.01, and (k(H)/k(D))(Obs) = 4.95 results in k₁(H)/k₁(D) = 5.11 The measured isotope effects are therefore due to differences in the amount of internal return and not in the symmetry of transition state structures for the hydron transfer, and the element effect, (k(HBr)/ k(HCl)) = 29, for m-ClC₆H₄CHClCH₂X is mainly due to the hydron-transfer step, k₁(HBr)/k₁(HCl) = 19, and not the breaking of the C-X bend. The kinetic solvent isotope effects, k(MeOD)/k(MeOH) ~ 2.5, are consistent with three methanols of solvation lost prior to the hydron-transfer step. The energetics associated with desolvation of methoxide ion are part of the measured reaction energetics of these systems.

Full text of DOI:10.1021/ja970189n

J. Am. Chem. Soc. 1997, 119, 9965-9974
9965
Use of Kinetic Isotope Effects in Mechanism Studies. 5.1
Isotope Effects and Element Effects Associated with
Hydron-Transfer Steps during Alkoxide-Promoted
Dehydrohalogenations
Heinz F. Koch,* Gerrit Lodder,* Judith G. Koch, David J. Bogdan,2a
Geoffrey H. Brown,2c Carrie A. Carlson,2a Amy B. Dean, Ronald Hage,2b
Patrick Han, Johan C. P. Hopman, Lisa A. James, Petra M. Knape,2b
Eric C. Roos,2b Melissa L. Sardina,2a Rachael A. Sawyer,2a Barbara O. Scott,2a
Charles A. Testa, III, and Steven D. Wickham2a
Contribution from the Department of Chemistry, Ithaca College, Ithaca, New York 14850, and
the Leiden Institute of Chemistry, Gorlaeus Laboratories, UniVersity of Leiden,
Leiden 2300RA, The Netherlands
ReceiVed January 21, 1997X
Abstract: The Arrhenius behavior of the primary kinetic isotope effect, (kH/kD)Obs and (kH/kT)Obs, associated with
the methanolic sodium methoxide-promoted dehydrohalogenations of m-ClC6H4CiHClCH2Cl (I), m-CF3C6H4Ci-
HClCH2Cl (II) and p-CF3C6H4CiHClCH2F (III) has been used to calculate the internal-return parameters, a ) k-1
/
kXElim, in a two-step mechanism featuring a hydrogen-bonded carbanion. This carbanion partitions between returning
the hydron to carbon, k-1, and the loss of halide, kX . Isotope effects at 25 °C for I, (kH/kD)Obs ) 3.40 and (kH/
Elim
kT)Obs ) 6.20, and II, (kH/kD)Obs ) 3.49 and (kH/kT)Obs ) 6.55, result in similar values for a: aH ) 0.59, aD
)
0.13-0.14 and aT ) 0.07. Smaller values of (kH/kD)Obs ) 2.19 and (kH/kT)Obs ) 3.56 for III are due to more internal
return [aH ) 1.9, aD ) 0.50, and aT ) 0.28] associated with the dehydrofluorination reaction. Calculation of k1 ()
kObs [a + 1]) results in similar isotope effects for hydron transfer in these reactions: I, kH1 /k1D ) 4.74 and k1H/k1T )
9.20; II, k1H/kD1 ) 4.91 and k1H/k1T ) 9.75; III, k1H/k1D ) 4.75 and kH1 /kT1 ) 9.17. Reactions of m-ClC6H4CiHBrCH2Br
and m-ClC6H4CiHClCH2Br have very small amounts of internal return, aH ) 0.05 and aD ) 0.01, and (kH/kD)Obs
)
4.95 results in kH1 /k1D ) 5.11. The measured isotope effects are therefore due to differences in the amount of internal
return and not in the symmetry of transition state structures for the hydron transfer, and the element effect, (kHBr
/
kHCl) ) 29, for m-ClC6H4CHClCH2X is mainly due to the hydron-transfer step, kHBr/kH1 Cl ) 19, and not the breaking
1
of the C-X bond. The kinetic solvent isotope effects, kMeOD/kMeOH ≈ 2.5, are consistent with three methanols of
solvation lost prior to the hydron-transfer step. The energetics associated with desolvation of methoxide ion are part
of the measured reaction energetics of these systems.
Our interest in alkoxide-promoted dehydrohalogenation reac-
tions began with measuring an extreme element effect, kHCl
carbanion, {C6H5C(CF3)CF2OCH2CH3}-, Via either a nucleo-
philic addition of ethoxide to C6H5C(CF3)dCF2 or a proton
abstraction from C6H5CH(CF3)CF2OCH2CH3. A mechanism
featuring two carbanion intermediates was postulated to explain
the hydrogen isotope effects that are associated with these
reactions.6,7 One carbanion is stabilized by a hydrogen bond
while the other one has no contact stabilization by either solvent
or the counterion.
/
kHF > 105, for the ethoxide-promoted eliminations of C6H5-
CHClCF2X, where X ) Cl or F. Kinetic studies of primary
kinetic isotope effects [PKIE] and Hammett F values led to the
suggestion that hydrogen-bonded carbanion intermediates are
formed on the reaction coordinate.3 Alkoxide-promoted elimi-
i
nations of C6H5CiHXCF2X [X ) Br or Cl and H ) H or D]
resulted in normal element effects, kHBr/kHCl ) 39-48, and
medium sized PKIE, kH/kD ) 2.3 to 4.5. These results could
be indicative of a concerted pathway; however, the Arrhenius
behavior of the isotope effects is inconsistent with an E2
mechanism.4 This raises the question about the origin of the
element effects associated with alkoxide-promoted dehydroha-
logenations in general.
That mechanism has been a model for our studies of alkoxide-
promoted dehydrohalogenations and alkoxide-catalyzed ex-
change reactions, Scheme 1. The extreme element effect
obtained for the ethoxide-promoted eliminations of C6H5-
CHClCF2Cl and C6H5CHClCF3 results from the loss of chloride
occurring from a hydrogen-bonded carbanion, HB, while the
loss of fluoride from -CF3 requires the formation of a free
carbanion, FC. Exchange of hydron from C6H5CiHClCF3
occurring 13 times faster than the loss of HF with a near unity
We became aware of the complexities of hydron transfer
between carbon and oxygen5 while studying the formation of a
* Address correspondence as follows: H. F. Koch, Ithaca College, and
G. Lodder, University of Leiden.
(3) Koch, H. F.; Dahlberg, D. B.; Toczko, A. G.; Solsky, R. L. J. Am.
Chem. Soc. 1973, 95, 2029.
(4) Koch, H. F.; Dahlberg, D. B.; McEntee, M. F.; Klecha, C. J. J. Am.
Chem. Soc. 1976, 98, 1060.
(5) Koch, H. F. Acc. Chem. Res. 1984, 17, 137.
(6) Koch, H. F.; Koch, J. G.; Donovan, D. B.; Toczko, A. G.; Kielbania,
A. J., Jr. J. Am. Chem. Soc. 1981, 103, 5417.
X Abstract published in AdVance ACS Abstracts, October 1, 1997.
(1) Part 4: Koch, H. F.; McLennan, D. J.; Koch, J. G.; Tumas, W.;
Dobson, B.; Koch, N.H. J. Am. Chem. Soc. 1983, 105, 1930.
(2) (a) Summer research student at Leiden from Ithaca College. (b)
Summer research student at Ithaca from the University of Leiden. (c)
Summer research student at Leiden from Oberlin College.
(7) Koch, H. F.; Koch, A. S. J. Am. Chem. Soc. 1984, 106, 4536.
S0002-7863(97)00189-3 CCC: $14.00 © 1997 American Chemical Society

9966 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Koch et al.
Scheme 1
ethoxide with all three systems are in Table 1, while those for
the reactions of methanolic sodium methoxide with YC6H4Ci-
HX′CH2X and YC6H4CiHX′CF2X are in Table 2. The reactions
of C6H5CH2CH2X were previously reported,11 and we measured
the kH/kD associated with C6H5CiH2CH2Cl to compare it to
that reported for C6H5CiH2CH2Br.12 Four systems used three
hydrogen isotopes to calculate internal-return parameters:
m-CF3C6H4CiHClCH2Cl, m-ClC6H4CiHClCH2Cl, p-CF3C6H4Ci-
HClCH2F, and m-ClC6H4CiHCBrCH2Br. Kinetic solvent iso-
tope effects were measured for C6H5CTClCF2Cl and C6H5-
CHClCH2Br by using mixtures of CH3OH and CH3OD, and
the Arrhenius behavior was studied for the dehydrochlorination
of p-CF3C6H4CHClCH2Cl and the protodetritiation of p-CF3C6H4-
CTClCF3.
k1H
kEXlim
X– + C
–OR
X
C
C– • • •H OR
C + HOR
X
C
C
H
k–H1
HB-h
k2H k–H2
X
FC
(kElim
)
kexc
X– + C
X
C
C– + HOR
C– + DOR
C
FC
k–D2
k2D
k–D1
k1D
X
C
X C C D
–OR
isotope effect is consistent with the loss of fluoride ion from
FC. Since Br- can leave from a hydrogen-bonded carbanion,
kHBr/kHCl values are normal.
Hydrogen Isotope Effects Associated with the Ethoxide-
Promoted Dehydrohalogenation Reactions. The results in
Table 1 show a decrease in kH/kD for ethoxide-promoted
eliminations of HBr and HCl at 25 °C from the least reactive
to the most reactive: C6H5CH2CH2Br (7.53), C6H5CHClCH2-
Br (6.10), and C6H5CHBrCF2Br (4.51); and C6H5CH2CH2Cl
(7.43), C6H5CHClCH2Cl (4.24), and C6H5CHClCF2Cl (2.73).
According to theory, primary kinetic isotope effects result from
differences in C-iH and/or O-iH bond zero-point energies in
the ground state and the transition structure. Thus the Arrhenius
preexponential factors, A, should be equal for isotopically
labeled compounds,13 and kH/kD results from ∆EDa -H, kH/kD )
(AH/AD) exp(∆EDa -H/RT). The single temperature PKIE values
suggest a symmetrical transition structure for eliminations of
C6H5CiH2CH2X and later transition structures for the other
eliminations.14 However, there are significant differences for
the temperature dependence of the isotope effects as AH/AD
values range from 0.5 to 3.1. Arrhenius parameters associated
with C6H5CiH2CH2Cl are normal with ∆EDa -H ) 1.2 kcal/mol
and AH/AD ) 1.1. Although kH/kD is the same for C6H5CiH2-
Discussion and Results
Distinguishing between Concerted and Nonconcerted
Alkoxide-Promoted Dehydrohalogenations. When dehydro-
halogenations occur Via the hydrogen-bonded intermediate in
Scheme 1, observed rate constants are defined by
kObs ) [k1kXEllim]/[k-1 + kXElim
]
(1)
There are two extremes: (a) k-1 >> kX then kObs ) [k1/k-1
]
kXElim, and (b) kEXlim >> k-1 then kObs )Eklim1. For case a second-
order kinetics, a near unity PKIE, and a substantial element
effect are expected. Case b results in second-order kinetics, a
PKIE that obeys the Swain-Schaad relationship,8 kH/kD ) (kD/
kT)2.26. For a concerted reaction, both C-H and C-X bonds
are partially broken in the transition structure, and could give a
variable PKIE, which obeys the Swain-Schaad relationship,
as well as a variable element effect, which should be the same
for the loss of HX and DX.
CH2Br (7.53), the Arrhenius parameters differ with ∆ED-H
)
When neither extreme for eq 1 is applicable, all three rate
constants, k1, k-1, and kXElim, contribute to the experimental rate
constants and there is no single rate-determining step. The
measured PKIE will vary between the extremes mentioned
above; however, the Swain-Schaad relationship would not hold.
Streitwieser et al. made use of deviations from the Swain-
Schaad relationship and single temperature rate constants for
all three hydrogen isotopes to calculate an internal-return
parameter, a ) k-1/k2 [in our case k2 ) kXElim].9 This parameter
can be used to calculate k1 () kObs [a + 1]) for all three isotopic
reactions.
a
1.6 kcal/mol and AH/AD ) 0.51, which halves the kH/kD predicted
from ∆EDa -H. The latter results were interpreted as resulting
from an E2 mechanism that has quantum-mechanical tunnel-
ing.15
At the other extreme are the reactions of C6H5CiHBrCF2Br
[∆EDa -H ) 0.3 kcal/mol and AH/AD ) 3.1] and C6H5CiHClCF2-
Cl [∆EDa -H ) 0.1 kcal/mol and AH/AD ) 2.2], where the
calculated AH/AD accounts for 69% to 82% of the measured
kH/kD. High AH/AD values of 2.1 to 2.9 were also encountered
for reactions of C6H5CiHClCH2Cl, m-ClC6H4CiHClCH2Cl,
m-CF3C6H4CiHClCH2Cl, and C6H5CiHClCH2Br while those
obtained from reactions of m-CH3C6H4CiHClCH2Br,16
m-CH3C6H4CiHBrCH2Br, and C6H5CiHBrCH2Br are normal
with ∆EaD-H of 0.9 to 1.1, AH/AD of 0.8 to 1.4, and kH/kD of
The use of eq 2 and a similar one for (kH/kT)Obs gives an
alternate way to determine the values of a as well as the
energetics associated with internal return by modeling experi-
mental isotope effects obtained over a 25 to 50°C range:10
(11) DePuy, C. H.; Bishop, C. A. J. Am. Chem. Soc. 1960, 82, 2535.
(12) Saunders, W. H., Jr.; Edison, D. H. J. Am. Chem. Soc. 1960, 82,
138.
ln(kH/kD)obs ) [((ED1 - E-D1) - (EH1 - EH-1))/RT] +
1 + (AXElim/A-1) exp((E-D1 - EXElim)/RT)
(13) Schneider and Stern [Schneider, M. E.; Stern, M. J. J. Am. Chem.
Soc. 1972, 94, 1517] considered a variety of model reactions and showed
that AH/AD was always between 0.7 and 1.2 for a temperature range of
20-2000 K, and had an absolute minimum value of 0.5. Reference 10
includes some calculations where AH * AD.
(14) Since the F values of the other reactions are larger than those reported
in the phenethyl series, the smaller PKIE values would suggest later rather
than earlier transition structures.
ln
(2)
X
H
X
(
)
1 + (AElim/A-1) exp((E-1 - EElim)/RT)
Three systems were used in our analysis of alkoxide-promoted
dehydrohalogenations: C6H5CH2CH2X and various YC6H4-
CHX′CH2X and YC6H4CHX′CF2X, where X ) Br, Cl, or F
and X′ ) Br or Cl. Results from reactions of ethanolic sodium
(15) Kaldor, S. B.; Saunders, W. H., Jr. J. Am. Chem. Soc. 1979, 101,
7594.
(8) Swain, C. G.; Stivers, E. C.; Reuwer, J. T.; Schaad, L. J. J. Am.
Chem. Soc. 1958, 80, 5885.
(9) (a) Streitwieser, A., Jr.; Hollyhead, W. B.; Sonnichsen, G.; Pudjaat-
maka, A. H.; Chang, C. J.; Kruger, T. L. J. Am. Chem. Soc. 1971, 93,
5096. (b) The exponent for the Swain-Schaad relationship used in ref 9a is
2.344.
(16) The AH/AD value of 2.1 calculated for C6H5CiHClCH2Br [Table 3]
i
was obtained from six kinetic points over a 35 °C range for H ) H and
five kinetic points over a 50 °C range for iH ) D. The choice of measuring
the parameters for m-CH3C6H4CiHClCH2Br was made to compare two
dehydrohalogenations that occur at almost the same rates: m-CH3C6H4Ci-
HClCH2Br (1.2) and C6H5CiHClCH2Cl (1.0). We plan to redo the C6H5Ci-
HClCH2Br kinetics.
(10) Koch, H. F.; Dahlberg, D. B. J. Am. Chem. Soc. 1980, 102, 6102.

Use of Kinetic Isotope Effects in Mechanism Studies
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9967
Table 1. Rate Constants, Activation Parameters, and Hydrogen Isotope Effects for Ethanolic Sodium Ethoxide-Promoted Dehydrohalogenation
Reactions
k, M-1 s-1
(25 °C)
∆Hq,
kcal/mol
∆Sq, eu
(25 °C)
kH/kD or kD/kT
(25 °C)
∆Ea,
kcal
AH/AD or
AD/AT
temp range, °C
[no. of points]
compound
C6H5CH2CH2Fa
4.35 × 10-8
25.6
-6.4
60 to 80
C6H5CH2CH2Cl
C6H5CD2CH2Cl
C6H5CH2CH2Brb
C6H5CD2CH2Br
2.95 × 10-6
3.97 × 10-7
22.8 ( 0.1
24.0 ( 0.3
-7.4 ( 0.2
-7.4 ( 0.6
45 to 85 [10]
50 to 75 [5]
7.43
7.53
1.2
1.6
1.1
0.5
2.32 × 10-4
3.08 × 10-5
20.4 ( 0.3
22.0 ( 0.2
-6.6 ( 0.9
-5.3 ( 0.5
30 to 50 [3]
30 to 50 [3]
C6H5CH2CF2Clc
2.47 × 10-4
20.3 ( 0.1
-7.1 ( 0.5
25 to 70 [5]
C6H5CHClCH2F
C6H5CDClCH2F
2.07 × 10-5
9.43 × 10-6
23.1 ( 0.1
23.4 ( 0.1
-2.4 ( 0.3
-3.0 ( 0.3
30 to 80 [12]
45 to 85 [8]
2.20
2.59
4.24
4.17
4.11
5.70
6.11
5.88
5.83
0.3
0.4
0.3
0.3
0.3
1.0
0.7
0.9
1.1
1.4
1.2
2.5
2.5
2.9
1.2
2.1
1.4
0.8
p-CF3C6H4CHClCH2F
p-CF3C6H4CDClCH2F
C6H5CHClCH2Cld,e
C6H5CDClCH2Cl
4.43 × 10-3
1.71 × 10-3
20.4 ( 0.1
20.8 ( 0.1
-1.0 ( 0.3
-1.3 ( 0.4
20 to 50 [6]
20 to 50 [6]
1.39 × 10-3
3.28 × 10-4
20.7 ( 0.1
21.0 ( 0.2
-2.3 ( 0.3
-4.1 ( 0.5
-10 to 50
20 to 70
m-ClC6H4CHClCH2Cl
m-ClC6H4CDClCH2Cl
2.64 × 10-2
6.33 × 10-3
19.1 ( 0.1
19.4 ( 0.1
-1.6 ( 0.2
-3.4 ( 0.4
-15 to 25 [17]
10 to 50 [6]
m-CF3C6H4CHClCH2Cl
m-CF3C6H4CDClCH2Cl
5.79 × 10-2
1.41 × 10-2
18.8 ( 0.1
19.0 ( 0.1
-1.3 ( 0.4
-3.4 ( 0.1
-10 to 20 [3]
-10 to 40 [10]
m-CH3C6H4CHClCH2Br
m-CH3C6H4CDClCH2Br
3.06 × 10-2
5.37 × 10-3
17.9 ( 0.1
18.9 ( 0.1
-5.4 ( 0.2
-5.7 ( 0.2
-20 to 30 [11]
0 to 55 [10]
C6H5CHClCH2Br
C6H5CDClCH2Br
m-CH3C6H4CHBrCH2Brd
m-CH3C6H4CDBrCH2Br
C6H5CHBrCH2Brd
C6H5CDBrCH2Br
4.89 × 10-2
8.00 × 10-3
18.0 ( 0.1
18.7 ( 0.1
-4.3 ( 0.2
-5.5 ( 0.5
-10 to 25 [6]
-10 to 40 [5]
2.28 × 10-2
3.88 × 10-3
18.5
19.4
-3.8
-4.4
-10 to 35 [10]
0 to 50 [11]
3.07 × 10-2
5.27 × 10-3
18.1
19.2
-4.8
-4.5
-20 to 30 [6]
0 to 45 [6]
f
C6H5CHClCF3
4.46 × 10-7
4.64 × 10-7
3.84 × 10-4
5.53 × 10-3
29.8 ( 0.1
29.4 ( 0.3
27.1 ( 0.2
17.0 ( 0.2
12.2 ( 0.2
11.1 ( 0.9
16.6 ( 0.5
-11.8 ( 0.2
70 to 110 [4]
70 to 100 [3]
20 to 60 [9]
10 to 50 [6]
C6H5CHBrCF3
p-CF3C6H4CHClCF3
2,6-Cl2C6H3CHClCF2Cl
C6H5CHClCF2Cle
C6H5CDClCF2Cl
2.21 × 10-1
8.09 × 10-2
18.6 ( 0.2
18.7 ( 0.1
0.8 ( 0.5
-0.8 ( 0.3
-20 to 15
-15 to 25
2.73
4.51
0.1
0.3
2.2
3.1
C6H5CHBrCF2Cl
2.14 × 10-1
18.6 ( 0.2
0.7 ( 0.7
-25 to 15 [9]
C6H5CHBrCF2Br
C6H5CDBrCF2Br
8.30
1.84
14.9 ( 0.1
15.2 ( 0.1
-4.2 ( 0.4
-6.5 ( 0.4
-55 to -25 [8]
-35 to 0 [8]
a Reference 11. b Reference 12. c Reworked data from ref 17. d Corrected for wrong way elimination. e Reference 1. f Reference 35.
5.70 to 5.88 at 25 °C. The solvent properties of ethanol are
not as good as those for methanol, and for that reason a majority
of the more detailed investigations of the YC6H4CiHClCH2X
systems use methanolic sodium methoxide. Faster reactions in
ethanolic ethoxide are attributed to less internal return for the
larger ethanol.17 The kEtOH/kMeOH at 25 °C are 30 for p-CF3C6H4-
CHClCH2F, 11 to 17 for YC6H4CHClCH2Cl, 10 to 11 for
YC6H4CHClCH2Br, and 7 for C6H5CHBrCH2Br.
Comparison of the Element and Chlorine 35/37 Isotope
Effects Associated with Ethoxide-Promoted Dehydrohalo-
genations. The element effect was first proposed by Bunnett
in a 1957 paper “The ‘Element Effect’ as a Criterion of
Mechanism in Activated Aromatic Nucleophilic Substitution
Reactions”,18 and has also been used for studies of dehydro-
halogenation mechanisms.19,20 Use of kHBr/kHCl values replaces
the measurement of chlorine isotope effects, and should result
from differences of activation enthalpies for reactions with
similar activation entropies. Element effects associated with
the ethoxide-promoted dehydrohalogenations of the â-phenethyl
halides are good examples as ∆Sq ) -6 ( 1 eu, and values of
kHCl/kHF and kHBr/kHCl are due to ∆Hq differences: kHCl/kHF
)
68, ∆∆Hq ) 2.8 kcal/mol [HF - HCl]; kHBr/kHCl ) 79, ∆∆Hq
) 3.0 kcal/mol [HCl - HBr].
Values of kHBr/kHCl are similar for C6H5CHClCH2X (35) and
C6H5CHBrCF2X (39); however, the kHCl/kHF of 67 for C6H5-
CHClCH2X, which is quite normal, differs dramatically from
the kHCl/kHF of 1.8 × 105 for C6H5CHBrCF2X. The ∆Sq values
for the C6H5CHClCH2X series are all within 3 eu, and the
element effects are largely due to the differences in ∆Hq: ∆∆Hq
) 2.4 kcal/mol [HF - HCl] and ∆∆Hq ) 2.7 kcal/mol [HCl -
HBr]. In contrast, ∆Sq values for the C6H5CHBrCF2X series
differ significantly, favoring loss of HF [11 eu] over HCl [1
eu] and HBr [-4 eu]. The very large kHCl/kHF for C6H5-
CHBrCF2X comes from a much larger ∆∆Hq (10.8 kcal/mol)
that is partially offset by a more favorable ∆Sq for the
dehydrofluorination reaction. For kHBr/kHCl the more favorable
∆Sq for elimination of HCl from C6H5CHBrCF2Cl helps to
compensate for the slightly larger ∆∆Hq of 3.7 kcal/mol (HCl
- HBr).
(17) Koch, H. F.; Tumas, W.; Knoll, R. J. Am. Chem. Soc. 1981, 103,
5423.
(18) Bunnett, J. F.; Garbisch, E. W.; Pruitt, K. M. J. Am. Chem. Soc.
1957, 79, 385.
(19) Bartsch, R. A.; Bunnett, J. F. J. Am. Chem. Soc. 1968, 90, 408.
(20) Gandler, J. R. In The Chemistry of Double-bonded Functional
Groups; Patai, Ed.; John Wiley &Sons Ltd.: New York, 1989; p 742-
744.
The use of element effects to distinguish between an E2
mechanism and a multi-step process with internal return has
been questioned.21 In Table 3 chlorine isotope effects, k35/k37,

9968 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Koch et al.
Table 2. Rate Constants, Activation Parameters, and Hydrogen Isotope Effects for Methanolic Sodium Methoxide-Promoted
Dehydrohalogenation Reactions
k, M-1 s-1
(25 °C)
∆Hq,
kcal/mol
∆Sq, eu
(25 °C)
kH/kD or kD/kT
(25 °C)
∆Ea,
kcal
AH/AD or
AD/AT
temp range, °C
[no. of points]
compd
p-CF3C6H4CHClCH2F
p-CF3C6H4CDClCH2F
p-CF3C6H4CTClCH2F
C6H5CHClCH2Cla
C6H5CDClCH2Cl
1.49 × 10-4
6.81 × 10-5
4.19 × 10-5
23.4 ( 0.1
23.9 ( 0.1
24.2 ( 0.2
2.6 ( 0.3
2.5 ( 0.3
2.6 ( 0.6
25 to 65 [10]
25 to 60 [6]
35 to 63 [9]
2.19
1.63
0.5
0.3
1.1
0.93
1.25 × 10-4
3.26 × 10-5
21.7
22.6
-3.5
-3.4
25 to 70
30 to 75
3.83
0.9
0.95
m-ClC6H4CHClCH2Cl
m-ClC6H4CDClCH2Cl
m-ClC6H4CTClCH2Cl
1.60 × 10-3
4.71 × 10-4
2.58 × 10-4
21.1 ( 0.1
21.9 ( 0.1
22.3 ( 0.2
-0.5 ( 0.1
-0.4 ( 0.5
-0.0 ( 0.5
0 to 50 [22]
10 to 60 [9]
10 to 40 [3]
3.40
1.83
0.8
0.4
0.96
0.81
m-CF3C6H4CHClCH2Cl
m-CF3C6H4CDClCH2Cl
m-CF3C6H4CTClCH2Cl
3.34 × 10-3
9.58 × 10-4
5.10 × 10-4
20.3 ( 0.1
21.1 ( 0.1
21.5 ( 0.1
-1.7 ( 0.2
-1.7 ( 0.2
-1.5 ( 0.3
5 to 50 [9]
10 to 60 [11]
20 to 55 [8]
3.49
1.88
0.8
0.4
1.0
0.92
p-CF3C6H4CHClCH2Cl
CH3ONa/CH3OD
8.03 × 10-3
1.98 × 10-2
20.08 ( 0.04
19.34 ( 0.06
-0.7 ( 0.2
-1.5 ( 0.2
0 to 40 [7]
0 to 35 [6]
m-CH3C6H4CHClCH2Br
m-CH3C6H4CDClCH2Br
3.12 × 10-3
6.23 × 10-4
20.1 ( 0.1
21.0 ( 0.2
-2.5 ( 0.2
-2.8 ( 0.6
5 to 55 [10]
20 to 70 [9]
5.01
5.43
5.21
4.94
5.39
5.27
0.9
1.2
1.0
1.0
1.0
1.0
1.2
C6H5CHClCH2Br
C6H5CDClCH2Br
4.38 × 10-3
8.07 × 10-4
19.3 ( 0.1
20.5 ( 0.1
-4.6 ( 0.2
-3.8 ( 0.2
0 to 50 [12]
5 to 60 [11]
0.67
0.98
0.87
0.91
0.91
p-ClC6H4CHClCH2Br
p-ClC6H4CDClCH2Br
2.35 × 10-2
4.51 × 10-3
18.6 ( 0.1
19.6 ( 0.1
-3.4 ( 0.4
-3.4 ( 0.4
-10 to 40 [7]
0 to 50 [11]
m-ClC6H4CHClCH2Br
m-ClC6H4CDClCH2Br
4.58 × 10-2
9.28 × 10-3
17.8 ( 0.1
18.8 ( 0.1
-5.0 ( 0.3
-4.7 ( 0.3
-10 to 25 [6]
0 to 40 [6]
p-CF3C6H4CHClCH2Br
p-CF3C6H4CDClCH2Br
1.79 × 10-1
3.34 × 10-2
17.6 ( 0.1
18.6 ( 0.1
-3.1 (0.2
-2.9 ( 0.3
-25 to 20 [8]
-15 to 30 [8]
C6H5CHBrCH2Br
C6H5CDBrCH2Br
3.16 × 10-3
6.00 × 10-4
19.9
20.9
-3.3
-3.2
0 to 50 [6]
15 to 50 [6]
m-ClC6H4CHBrCH2Br
m-ClC6H4CDBrCH2Br
m-ClC6H4CTBrCH2Br
3.27 × 10-2
6.60 × 10-3
3.27 × 10-3
18.6
19.6
20.0 ( 0.1
-2.8
-20 to 25 [8]
0 to 40 [8]
20 to 45 [6]
4.95
2.02
1.0
0.4
0.90
1.1
-2.6
-2.8 ( 0.4
p-CF3C6H4CHClCF3
1.06 × 10-5
7.91 × 10-4
2.06 × 10-3
27.9 ( 0.1
25.42 ( 0.08
24.86 ( 0.13
12.3 ( 0.3
12.5 ( 0.3
12.5 ( 0.4
45 to 70 [7]
20 to 55 [8]
20 to 45 [6]
b
p-CF3C6H4CTClCF3
CH3ONa/CH3OD
C6H5CHClCF2Cl
C6H5CDClCF2Cl
C6H5CTClCF2Cl
1.73 × 10-2
7.60 × 10-3
4.91 × 10-3
20.2 ( 0.1
20.2 ( 0.1
19.6 ( 0.1
1.2 ( 0.3
-0.4 ( 0.3
-3.5 ( 0.3
-5 to 45 [14]
0 to 50 [13]
0 to 40 [5]
2.28
1.55
0.0
-0.6
2.3
4.7
C6H5CHBrCF2Br
C6H5CDBrCF2Br
C6H5CTBrCF2Br
8.19 × 10-1
2.23 × 10-1
1.27 × 10-1
16.7 ( 0.1
17.0 ( 0.1
17.1 ( 0.1
-3.0 ( 0.2
-4.6 ( 0.3
-5.1 ( 0.3
-40 to -5 [9]
-50 to 20 [19]
-30 to 25 [10]
3.67
1.76
0.3
0.1
1.6
1.3
a Reference 1. b Kinetics are for methoxide protodetritiation.
Table 3. Chlorine and Hydrogen Isotope Effects and Element Effects Associated with Alcoholic Sodium Alkoxide-Promoted
Dehydrochlorination Reactionsa
compd
solvent
EtOH
EtOH
EtOH
k35/k37 [°C]
kH/kD [°C]
kHBr/kHCl [°C]
C6H5CH(CH3)CH2Cl
C6H5CD(CH3)CH2Cl
1.00590 ( 0.00013 [75]
1.00507 ( 0.00036 [75]
52 [75]
55 [75]
5.37 [75]
C6H5CH2CH2Cl
1.00580 ( 0.00034 [75]
C6H5CHClCH2Cl
C6H5CDClCH2Cl
1.00908 ( 0.00008 [24]
1.00734 ( 0.00012 [24]
35 [25]
24 [25]
4.24 [25]
3.83 [25]
2.77 [0]
C6H5CHClCH2Cl
C6H5CDClCH2Cl
1.00978 ( 0.00020 [21]
1.00776 ( 0.00020 [21]
35 [25]
25 [25]
MeOH
EtOH
C6H5CHClCF2Cl
C6H5CDClCF2Cl
1.01229 ( 0.00047 [0]
1.01003 ( 0.00024 [0]
66 [0]
39 [0]
C6H5CHClCF2Cl
C6H5CDClCF2Cl
1.01255 ( 0.00048 [20]
1.01025 ( 0.00043 [20]
47 [25]
29 [25]
MeOH
2.28 [25]
a Data from ref 1.
are compared with kHBr/kHCl values for the corresponding
eliminations with use of CH3CH2ONa/CH3CH2OH and CH3-
ONa/CH3OH. There is no correlation between the magnitude
. If these eliminations occur by a
concerted mechanism, the chlorine isotope effects should be
the same for loss of HCl or DCl; however, the amount of internal
return is different for protium compared to deuterium and this
results in (k35/k37)HCl * (k35/k37)DCl for C6H5CiHClCH2Cl and
C6H5CiHClCF2Cl.
of k35/k37 and kHBr/kHCl
If the breaking of a C-X bond is not the only reason for the
element effect, the ability of â-halide to enhance the hydron-
(21) Reference 1, pp 1935 and 1936.

Use of Kinetic Isotope Effects in Mechanism Studies
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9969
Table 4. Modeling the Arrhenius Behavior of the Primary Kinetic Isotope Effects Associated with Methanolic Methoxide Promoted
Dehydrohalogenation Reactions Based on Eq 2
temp, °C
kHObs, M-1 s-1
kDObs, M-1 s-1
kTObs, M-1 s-1
(kH/kD)obs
(kH/kD)cal
(kH/kT)obs
(kH/kT)cal
aH
aD
aT
m-ClC6H4CiHClCH2Cl (E-D1 - E-H1 ) 8.75; EH-1 - EEXlim ) -100 cal/mol; A2/A1 ) 2.0)
0
25
50
5.60 × 10-5
1.60 × 10-3
2.73 × 10-2
1.47 × 10-5
4.71 × 10-4
8.84 × 10-3
7.46 × 10-6
2.58 × 10-4
5.15 × 10-3
3.81
3.40
3.09
3.82
3.37
3.04
7.51
6.20
5.30
7.46
6.19
5.29
0.60
0.59
0.58
0.12
0.14
0.15
0.060
0.072
0.084
m-CF3C6H4CiHClCH2Cl (ED-1 - E-H1 ) 900; E-H1 - EXElim ) -100 cal/mol; A2/A1 ) 2.0)
0
25
50
1.32 × 10-4
3.34 × 10-3
5.13 × 10-2
3.37 × 10-5
9.58 × 10-4
1.62 × 10-2
1.68 × 10-5
5.10 × 10-4
9.12 × 10-3
3.92
3.49
3.17
3.98
3.50
3.15
7.86
6.55
5.63
7.93
6.55
5.57
0.60
0.59
0.58
0.11
0.13
0.14
0.056
0.068
0.079
p-CF3C6H4CiHClCH2F (ED-1 - E-H1 ) 850; EH-1 - EXElim ) -850 cal/mol; A2/A-1 ) 2.0 [uses tritium data])
25
50
75
1.49 × 10-4
3.44 × 10-3
5.06 × 10-2
6.73 × 10-5
1.64 × 10-3
2.54 × 10-2
4.22 × 10-5
1.05 × 10-3
1.63 × 10-2
2.21
2.10
1.99
2.20
2.10
2.02
3.53
3.31
3.10
3.55
3.27
3.04
2.1
1.9
1.7
0.50
0.50
0.50
0.27
0.28
0.30
m-ClC6H5CiHBrCH2Br (ED-1 - EH-1 ) 925; EH-1 - EXElim ) 1425 cal/mol; A2/A-1 ) 2.0 [uses tritium data])
0
25
50
1.68 × 10-3
3.26 × 10-2
4.03 × 10-1
2.89 × 10-4
6.60 × 10-3
9.27 × 10-2
1.36 × 10-4
3.27 × 10-3
4.82 × 10-2
5.81
4.94
4.35
5.81
4.97
4.36
12.4
9.97
8.36
12.4
9.97
8.26
0.036
0.045
0.054
0.0066
0.0095
0.013
0.0032
0.0049
0.0070
875 cal/mol,24 E-H1 - EEXlim ) -100 cal/mol, and internal-
return parameters of aH ) 0.59, aD ) 0.14, and aT ) 0.072 at
25 °C. If AXElim/A-1 is set at 3.0, the best fit is the same for the
kH/kD or kH/kT data, with kH/kD values that are within 0 to -0.3%
transfer step must play a partial role. To obtain the effect of
â-halide on hydron transfer, kinetics using all three hydrogen
isotopes are needed to calculate the amount of internal return,
a ) k-1/kXElim, which can be used to obtain a value for the
hydron-transfer step, k1 ) kObs[a + 1].
and kH/kT values within -1.1 to +1.5. The value for ED-1
-
EH-1 is 875 cal/mol and that for EH-1 - EXElim is -325 cal/mol,
with aH ) 0.58, aD ) 0.13, and aT ) 0.070 at 25 °C. Since
Internal Return Associated with the Hydron-Transfer
Step in Methoxide-Promoted Dehydrohalogenation Reac-
tions. When dehydrohalogenations occur by Scheme 1, the use
of eq 2 and one for (kH/kT)Obs allows the calculation of a values
as well as the energetics associated with internal return by
modeling experimental isotope effects obtained over a 25 to 50
°C range.10 For example, the methanolic sodium methoxide-
promoted eliminations of HCl and DCl from m-ClC6H4Ci-
HClCH2Cl were studied over a 50 °C range and resulted in kH/
AXElim/A-1 favors elimination more in the latter case, the EH
-
-1
EXElim favors the return by an additional 225 cal/mol; however,
the amounts of internal return are virtually the same for both
cases. We use AXElim/A-1 values of 2.0 for the rest of our
calculations.25
Although the value of ED-1 - E-H1 seems low, a similar result
is obtained for m-CF3C6H4CiHClCH2Cl with AXElim/A-1 ) 2.0:
ED-1 - EH-1 ) 900 cal/mol, E-H1 - EEXlim ) -100 cal/mol, kH/kD
within -0.6 to +1.5%, kH/kT within +0.9 to -0.6%, aH ) 0.59,
aD ) 0.13, and aT ) 0.068. The data are given for m-CF3C6H4-
CHClCH2Cl at nine temperatures between 5.0 and 50 °C, for
m-CF3C6H4CDClCH2Cl at ten temperatures between 15 and 60
°C, and for m-CF3C6H4CTClCH2Cl at eight temperatures
between 20 and 55 °C. Using the seven common temperatures
for all three compounds at 5 °C intervals between 20 and 50
°C gives the same results.
kD ) 3.40 (25 °C) with a ∆Ea ) 0.8 kcal/mol and AH/AD
)
0.96, Table 2. Although the tritium kinetics were only carried
out at three temperatures between 10 and 40 °C, calculated
deviations for ∆Hq and ∆Sq were similar to those obtained from
the deuterium data. Values of kH/kD and kH/kT at 0, 25, and 50
°C are obtained from the Arrhenius analyses. With use of a
set value for AEXlim/A-1 and starting values for E-D1 - E-H1 and
EH-1 - EEXlim 22 a program will systematically change the energy
,
terms to give the best fit for the kH/kD data, and calculate the
corresponding kH/kT. The process is repeated for a best fit for
kH/kT and calculation of the corresponding kH/kD values.23 With
AXElim/A-1 set at a new value, the process can be repeated.
A study with p-CF3C6H4CiHClCH2F at six common temper-
atures between 35 and 60 °C with AXElim/A-1 set at 2.0 results in
a similar value for ED-1 - E-H1 ) 850 cal/mol but a higher value
for EH-1 - EXElim ) -850 cal/mol. The kH/kD values are within
-0.5 to +1.0% and kH/kT are within -0.3 to -1.2%, with aH
With use of the data for m-ClC6H4CiHClCH2Cl and
AXElim/A-1 ) 1.0, calculated kH/kD values are within 0.5% of the
experimental values; however, corresponding kH/kT values are
24-29% high. The best fit for kH/kT (-0.4 to -1.3%) gives
kH/kD values that are low by 8 to 8.4%. When AEXlim/A-1 is
changed to 2.0, a best fit for the kH/kD (-0.3 to +0.6%) still
has kH/kT values that are high by 8.7 to 10.2%. However, the
kH/kT fit (-0.2 to -0.7%) gives kH/kD values within +0.3 to
) 1.9, aD ) 0.50, and aT ) 0.28 at 35 °C. The larger EH-1
-
EXElim is consistent with stronger C-F than C-Cl bonds. Since
the ED-1 - E-H1 for the three systems are within 50 cal/mol,
hydron transfer in the transition structures should be similar
for the elimination of hydrogen chloride and hydrogen fluoride.26
The lower isotope effect is usually attributed to the poorer
leaving group having a later transition structure. For our three
-1.6% of the experimental ones, and results in ED-1 - E-H1
)
(24) The value of E1D - E1H is always 46 cal/mol [∆HfD - ∆HHf )] larger
than ED-1 - EH-1. Other values for the equilibrium isotope effects are as
(22) The starting point for ED-1 - E-H1 is normally 1300 cal/mol. For
follows: ∆HTf - ∆HfD ) 20 cal/mol and ∆HfT - ∆HHf ) 66 cal/mol.
(25) Although the best fit is the same when based on either the kH/kD or
H
loss of HF and HCl the starting point for E-1 EX is 500 cal/mol and
i
i
-
Elim
for loss of HBr 1800 cal/mol.
kH/kT data for AXElim/A-1 ) 3.0, the overall fit with kH/kT data with
(23) We thank Duncan McLennan, Auckland University in New Zealand,
for a copy of his program that models the Arrhenius behavior of primary
kinetic isotope effects for a reaction that features an internal-return
mechanism. The program makes use of eq 2 and a similar one for protium
and tritium. The original program, in Fortran, was converted to Turbo Pascal
for use with a PC, and can be obtained from J. G. Koch.
AXElim/A-1 ) 2.0 is better.
(26) When the observed rate constants are corrected for internal return,
the isotope effects associated with the proton transfer step, k1, are within
7% at 25 °C: m-ClC6H4CiHClCH2Cl, kH/kD ) 4.73; m-CF3C6H4CiHClCH2-
Cl, kH/kD ) 4.92; p-CF3C6H4CiHClCH2F, kH/kD ) 4.57.

9970 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Koch et al.
slightly to kD1 Br/kD1 Cl ) 18. Therefore the major part of the
element effect associated with this dehydrohalogenation is due
to a â-bromide enhancing the hydron-transfer step, k1, more
than a â-chloride.30
Kinetic Solvent Isotope Effects [KSIE]. It has been
proposed that KSIE are due to differences in the number of
CH3OiH solvating methoxide ion and those solvating the
transition structure.31
systems, the lower experimental PKIE is due to greater amounts
of internal return for the dehydrofluorination than for the
elimination of hydrogen chloride.
Calculations in the YC6H4CiHClCH2Br systems have been
hampered due to the unsuccessful syntheses of pure YC6H4-
CTClCH2Br. The use of the corresponding dibromides, YC6H4Ci-
HBrCH2Br, also presented experimental problems because of
a competing wrong way elimination that forms YC6H4CiHd
CHBr.27 The reaction of m-ClC6H4CiHBrCH2Br resulted in less
than 1% of m-ClC6H4CHdCHBr, but that increases to 3 to 6%
for m-ClC6H4CDdCHBr. Therefore the tritium kinetics could
result in 6% to 12% of m-ClC6H4CTdCHBr constantly being
formed, and require special treatment of the experimental data
for m-ClC6H4CTBrCH2Br.
R-H + MeO-‚‚‚(iHOCH3)m f
{R‚‚H‚‚O(Me)‚‚‚(iHOCH3)p}q- + (m - p)CH3OiH
A fractionation factor, φOMe ≈ 0.74,32 for CH3OD implies
the equilibrium abundance of deuterium in the m hydrogens of
a solvated MeO- is ca. 0.74 times that in the bulk solvent. The
value of m is taken as 3 which has methanol at all three possible
ligand sites. At least one methanol must be lost to form the
encounter complex, R-H···-O(Me)···(iHOCH3)p, prior to proton
transfer. Calculation of the KSIE by eq 3 determines the
minimum effect, kOD/kOH ≈ 1.3, when a single methanol is
released to solvent (φ ≈ 1.0) and the transition structure
fractionation factors equal those for methoxide, φq ) φOMe. A
maximum effect has all three methanols released to solvent and
results in kOD/kOH ≈ 2.5.
Kinetic samples for protium and deuterium compounds are
analyzed by using gas chromatography with an internal standard.
Experimental rate constants at eight temperatures, -20 to 25
°C, result in a good Arrhenius plot for m-ClC6H4CHBrCH2Br
with EHa ) 19.26 ( 0.03 kcal/mol that shows little change
when the rate constants are corrected for wrong way elimination.
However, the EDa of 20.35 ( 0.05 kcal/mol obtained from the
experimental rates for m-ClC6H4CDBrCH2Br decreases slightly
to EDa ) 20.24 kcal/mol when corrected for the formation of
m-ClC6H4CDdCHBr. This results in a ∆Ea of 1.0 instead of
1.1 kcal/mol with AH/AD increasing from 0.76 to 0.90. Kinetic
samples from the reaction of m-ClC6H4CTBrCH2Br will have
a growing infinity point due to the formation of m-ClC6H4-
CTdCHBr in competition with m-ClC6H4CBrdCH2. Therefore
the experimental counts had to be corrected for the formation
of m-ClC6H4CTdCHBr using the rate data for the formation
of m-ClC6H4CDdCBr from m-ClC6H4CDBrCH2Br. The cor-
rected rate constants gave a six point Arrhenius plot over a 25
°C range [20 to 45 °C] that resulted in ETa ) 20.6 ( 0.1 kcal/
mol and AT ) 4.155 × 1012.
m
k
OD/kOH ) (φq)p/(φOMe
)
(3)
Reactions of C6H5CTClCF2Cl and C6H5CHClCH2Br are 2.5
times faster in pure CH3OD than in CH3OH. A complete study
of KSIE should include kinetic runs in mixtures of CH3OD and
CH3OH as well as in the pure solvents. Equation 4, where n is
the mole fraction of CH3OD, allows calculation of predicted
rate enhancement for these mixtures:
kOa D/kOH ) 1/[1 - n + 0.74n]3
(4)
Rate constants from the Arrhenius analyses are used to
calculate kH/kD and kH/kT at 0, 25, and 50 °C. With
AXElim/A-1 set at 2.0, values for E-D1 - E-H1 ) 925 cal/mol and
EH-1 - EXElim ) 1425 cal/mol are calculated for the best fit of
the experimental results.28 Calculated kH/kD values are within
0.0 to +0.4%, kH/kT are within 0.0 to -1.2%, and internal-
return parameters are aH ) 0.045, aD ) 0.0095 and aT ) 0.0049
at 25 °C. In the absence of tritium data, the Arrhenius behavior
of m-ClC6H4CHClCH2Br and m-ClC6H4CDClCH2Br gives a
good fit, using eq 2 with AXElim/A-1 ) 2.0, E-D1 - E-H1 ) 925
The experimental values at 25 °C for the two systems are
within 4% of those predicted by eq 4: [kOn D × 103 M-1 s-1
,
(n)] C6H5CTClCF2Cl, 12.3 (1.0), 8.59 (0.67), 7.31 (0.50), 6.44
(0.33), and 4.91 (0.0); C6H5CHClCH2Br, 11.0 (1.0), 8.09 (0.75),
6.67 (0.50), 5.54 (0.25), and 4.38 (0.0).
The temperature dependence of the KSIE for methanol was
also measured for the dehydrochlorination of p-CF3C6H4-
CHClCH2Cl and the hydrodetritiation reaction of p-CF3C6H4-
CTClCF3. Values of kOD/kOH are 2.76 [0°], 2.47 [25°], and 2.23
[50°] for p-CF3C6H4CHClCH2Cl, and 2.84 [0°], 2.60 [25°], and
2.42 for p-CF3C6H4CTClCF3. The KSIE in these cases result
mainly from ∆∆Hq of 0.6 kcal/mol with the ∆∆Sq less than
0.7 eu, Table 2. Therefore the energetics associated with the
desolvation of methoxide ion are part of the measured reaction
energetics and should be the same for these reactions studied
in methanol.
cal/mol, EH-1 - EEXlim ) 1350 cal/mol, aH ) 0.05 and aD
)
0.01.29 The experimental kH/kD values at 25 °C for the two
systems are 4.95 and 4.94.
The experimental element effects, kBr/kCl, associated with the
methoxide-promoted dehydrohalogenations of m-ClC6H4Ci-
HClCH2X can now be corrected for internal return. The (kHBr
/
kHCl
step of k1HBr/k1HCl ) 19, while (kDBr/kDCl
)
exp of 29 results in an element effect for the proton-transfer
)
exp
of 20 is reduced
Activation Entropies. To compare rates of bimolecular
reactions that differ by several orders of magnitude, we routinely
measure kinetics over a 30-50 °C range. To anticipate what
to expect for two molecules reacting in the rate-limiting step,
(27) A similar problem was encountered with C6H5CTClCH2Cl where
the competitive formation of C6H5CTdCHCl interfered with measurement
of good tritium kinetics; however, m-ClC6H4CiHClCH2Cl gave negligible
amounts of m-ClC6H4CTdCHCl and resulted in good kinetics.
(28) The starting point for EH-1 - EEXlim is 1800 cal/mol for the
dehydrobrominations. With A2/A-1 ranging from 1.5 to 3.0, ED-1 - E-H1
(30) A referee pointed out that the element effect on k1 suggests negative
ion hyperconjugation in the transition structure since the electronegativity
of Cl Vs Br should give the reverse order. The same would hold true for
the case of F Vs Cl. A publication that appeared after this manuscript was
submitted gives additional computational evidence for negative ion hyper-
conjugation by a â C-F [Sanuders, W. H., Jr. J. Org. Chem. 1997, 62,
244].
values are from 925 and 950 cal/mol, EH-1 - EEXlim values are from 1200
and 1500 cal/mol, and internal-return parameters at 25 °C are aH ) 0.046
( 0.003, aD ) 0.0097 ( 0.008, and aT ) 0.0049 ( 0.0003.
(29) Using only eq 2 and the temperature dependence of kH/kD will give
good modeling of the experimental results with a much larger range of
AXElim/A-1 values. Since the measurement of tritium rates is not always
(31) Gold, V.; Grist, S. J. Chem. Soc. (B) 1971, 2282.
convenient, the temptation to use only hydrogen and deuterium data must
(32) (a) More O’Ferrall, R. A. Chem. Commun. 1969, 114. (b) Gold,
V.; Grist, S. J. Chem. Soc. (B) 1971, 1665. (c) Gold, V.; Morris, K. P.;
Wilcox, C. F. J. Chem. Soc., Perkin Trans. 2 1982, 1615.
be suppressed until a consistent value of AXElim/A-1 is found for several
similar systems.

Use of Kinetic Isotope Effects in Mechanism Studies
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9971
Table 5. Rate Constants and Activation Parameters for Methanolic Sodium Methoxide-Promoted Dehydrochlorination Reactions
compd
k, M-1 s-1 (0 °C)
∆Hq, kcal/mol
∆Sq, eu (25 °C)
temp range, °C [no. of points]
o-CH3C6H4CHClCF2Cl
o-CF3C6H4CHClCF2Cl
2,6-Cl2C6H3CHClCF2Cl
2-Cl-6-FC6H3CHClCF2Cl
o-ClC6H4CHClCF2Cl
p-CH3C6H4CHClCF2Cl
m-CH3C6H4CHClCF2Cl
o-FC6H4CHClCF2Cl
1.60 × 10-5
4.21 × 10-5
8.78 × 10-5
1.74 × 10-4
1.94 × 10-4
2.28 × 10-4
4.01 × 10-4
7.03 × 10-4
7.15 × 10-4
8.21 × 10-4
1.16 × 10-3
1.64 × 10-3
1.75 × 10-3
5.47 × 10-3
1.25 × 10-2
1.60 × 10-2
1.79 × 10-2
3.71 × 10-2
9.14 × 10-2
2.10 × 10-1
7.67 × 10-1
2.88
21.81 ( 0.08
18.90 ( 0.03
18.60 ( 0.03
19.25 ( 0.05
19.44 ( 0.06
20.94 ( 0.03
20.48 ( 0.07
19.43 ( 0.04
20.04 ( 0.07
18.27 ( 0.07
20.16 ( 0.07
18.24 ( 0.08
19.03 ( 0.10
19.00 ( 0.04
18.71 ( 0.03
18.21 ( 0.02
18.41 ( 0.09
18.37 ( 0.09
17.84 ( 0.06
16.67 ( 0.10
15.77 ( 0.03
15.77 ( 0.04
-0.5 ( 0.2
-9.2 ( 0.1
-8.8 ( 0.1
-5.1 ( 0.2
-4.2 ( 0.2
1.6 ( 0.1
25 to 65 [8]
20 to 60 [9]
10 to 60 [8]
10 to 45 [8]
15 to 40 [6]
5 to 45 [12]
0 to 40 [8]
1.1 ( 0.3
-1.6 ( 0.2
0.6 ( 0.2
-5.6 ( 0.3
2.0 ( 0.3
0 to 40 [6]
C6H5CHClCF2Cl
2,6-F2C6H3CHClCF2Cl
p-FC6H4CHClCF2Cl
0 to 45 [10]
-5 to 30 [6]
-10 to 30 [6]
-5 to 25 [7]
-10 to 20 [8]
-20 to 20 [15]
-25 to 5 [7]
-25 to 10 [9]
-30 to 5 [6]
-25 to 0 [6]
-30 to 0 [8]
-45 to -20 [3]
-61 to -20 [4]
-65 to -30 [6]
2,4-Cl2C6H3CHClCF2Cl
2,4-F2C6H3CHClCF2Cl
p-ClC6H4CHClCF2Cl
m-FC6H4CHClCF2Cl
m-ClC6H4CHClCF2Cl
3,4-F2C6H3CHClCF2Cl
m-CF3C6H4CHClCF2Cl
p-CF3C6H4CHClCF2Cl
m-NO2C6H4CHClCF2Cl
p-CNC6H4CHClCF2Cl
3,5-(CF3)2C6H3CHClCF2Cl
-4.3 ( 0.3
-1.3 ( 0.3
0.8 ( 0.1
1.4 ( 0.1
0.1 ( 0.1
1.0 ( 0.3
2.4 ( 0.3
2.2 ( 0.2
-0.4 ( 0.4
-1.2 ( 0.1
1.5 ( 0.2
the ∆Sq values obtained from the reactions of methanolic sodium
methoxide with ring-substituted â,â-difluorostyrenes, YC6H4-
CHdCF2 + CH3O- f {YC6H4CHCF2OCH3}-, were used.33
The average ∆Sq for ten ring substituents used to calculate F )
3.56 ( 0.03 (25 °C) is -14.3 ( 0.8 eu, the average ∆Sq for
seven compounds not used to calculate rho is slightly lower,
-15.7 ( 0.8 eu, and the average ∆Sq for ten compounds with
ortho substituents is -17.3 ( 1.2 eu. Experimental ∆Sq values
calculated from the Arrhenius plots routinely have a standard
deviation of less than 0.3 eu. Although methoxide-promoted
dehydrochlorinations of YC6H4CHClCF2Cl are bimolecular
reactions, the average ∆Sq ≈ 1 eu for the non-ortho substituted
compounds is much larger than that for the alkene reactions,
Table 5.
It was suggested that ∆Sq values vary with the amount of
internal return and more favorable activation entropies occur
as the amount of internal return increases.34 The kH/kD of 1.1
to 1.2 for alkoxide-catalyzed exchange and elimination reactions
of C6H5CiHClCF3 and C6H5CiH(CF3)2 suggest a very large
amount of internal return associated with these reactions, and
result in ∆Sq values greater than 10 eu.35 On the other hand,
ethoxide-promoted dehydrohalogenations of C6H5CH2CH2X [X
) Br or Cl] have substantial kH/kD values, no internal return,
and ∆Sq values of about -7 eu. The stereochemistry of the
elimination can play a major role in determining ∆Sq. Cristol
and co-workers measured values of ∆Sq for the elimination of
four isomers of benzene hexachloride, C6H6Cl6s1.0, 3.6, 6.5,
and 20.2 eu.36 The 20.2 eu value is for the only isomer unable
to undergo antielimination as all the chlorines are trans to each
other.
major conformers of C6H5CHClCF2Cl.37 Therefore the leaving
halogen might not be anti to the encounter complex, CH3O-‚‚
H-C, and would require a 60° rotation to obtain the proper
alignment.38 Rotational barriers in highly halogenated com-
pounds are between 9 and 14 kcal/mol,39 and this could lead to
a third step.40 Elimination from C6H5CH2CF2Cl, which has two
benzylic hydrogens, results in a decrease of ca. 8 eu in ∆Sq
compared to C6H5CHClCF2Cl. This is attributed to dehydro-
chlorination occurring with substantially less or no internal
return.
The presence of bulky ortho substituents could hinder internal
return, and cause a decrease in the values of ∆Sq. Seven
substituents were used to determine F ) 3.69 ( 0.05 at 0 °C,41
and the average ∆Sq equals 1.0 ( 1.0 eu for the 13 non-ortho-
substituted YC6H4CHClCF2Cl or Y2C6H3CHClCF2Cl in Table
5. This changes dramatically for the nine compounds with one
or two ortho substituents as ∆Sq values range from -0.5 to -9.2
eu. Values for o-CF3C6H4CHClCF2Cl [-9.2 eu] and 2,6-
Cl2C6H3CHClCF2Cl [-8.8 eu] are lowest, while 2,4-F2C6H3-
CHClCF2Cl [-1.3 eu] and o-CH3C6H4CHClCF2Cl [-0.5 eu]
result in the highest values. There is internal consistency: two
o-F compounds [-1.6 and -1.3 eu] and two o-Cl compounds
[-4.3 and -4.2 eu]. Values are larger with two ortho
substituents: 2,6-difluoro [-5.6 eu], 2-chloro-6-fluoro [-5.1
eu], and 2,6-dichloro [-8.8 eu]. The trifluoromethyl group
[-9.2 eu] is the largest and the only problem seems to be the
o-methyl [-0.5 eu].42 Further work is in progress dealing with
the effects of ortho substituents.
(37) Drysdale, J. J.; Phillips, W. D. J. Am. Chem. Soc. 1957, 79, 319.
(38) The methoxide-promoted dehydrohalogenation of C6H5CiHBrCFClBr
i
i
gives only anti-eliminations of HCl and HBr [Koch, H. F. In Compre-
hensiVe Carbanion Chemistry, Part C; Buncel, E., Durst, T., Eds.;
Elsevier: Amsterdam, 1987; pp 333 and 334.]
We have been unable to model the temperature behavior for
the isotope effects associated with eliminations from C6H5Ci-
HClCF2Cl or C6H5CiHBrCF2Br. The major conformer of C6H5-
CHBrCF2Br (>98%) has the bromine atoms anti to each other,
and the H-C-C-Cl is not antiperiplanar in either of the two
(39) (a) Thompson, D. S.; Newmark, R. A.; Sederholm, C. H. J.
Chem.Phy. 1962, 37, 411. (b) Weigert, F. J.; Winstead, M. B.; Garrels, J.
I.; Roberts. J. D. J. Am. Chem. Soc. 1970, 92, 7359.
(40) One of the referees commented that this would be at odds with the
Curtin-Hammett principle. Although this is true, the energetics of the
rotational barriers and the competition with internal return may well be an
example of the breakdown of this principle, which does not have a large
amount of experimental backing.
(33) The measurement of kinetic solvent isotope effect for these reactions
has not been as extensive. However, the few that we have measured are
about 2.0.
(34) See ref 17, pp 5425-27.
(41) Substituents used were as follows: p-CH3, m-CH3, H, p-Cl, m-F,
m-Cl, and m-CF3. The F value for the reaction of YC6H4CHdCF2 increases
to 3.92 ( 0.04 at 0 °C.
(35) Koch, H. F.; Dahlberg, D. B.; Lodder, G.; Root, K. S.; Touchette,
N. A.; Solsky, R.L.; Zuck, R. M.; Wagner, L. J.; Koch, N. H.; Kuzemko,
M. A. J. Am. Chem. Soc. 1983, 105, 2394.
(42) We are at a loss to explain the ∆Sq of -0.5 for o-methyl. The
compound has been synthesized three times by different students and always
results in virtually the same value for ∆Sq.
(36) Cristol, S. J.; Hause, N. L.; Meek, J. S. J. Am. Chem. Soc. 1951,
73, 674.

9972 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Koch et al.
Conclusions. Our working model for alkoxide-promoted
dehydrohalogenation and hydron-exchange reactions is Scheme
1. For reactions with methanolic methoxide, the process starts
after methoxide ion loses one molecule of solvation which
allows it to form an encounter complex with the carbon acid.
bond. The competition between a loss of â-Cl and internal
return has an activation energy slightly favoring internal return
and an A factor slightly favoring the loss of chloride, and this
results in the moderate amounts of internal return, kH-1 ≈ 0.6
kCEllim. Dehydrobromination has a much more favorable activa-
tion energy for loss of â-Br as well as the favorable A factor
MeO-‚‚(HOCH3)2 + H-C-C-X f
(CH3OH)2‚‚(Me)O-‚‚H-C-C-X
which results in a negligible amount of internal return, k-H1
≈
0.05 kBElrim. Since the internal return does not affect the overall
rate of dehydrobromination, the reactions of m-ClC6H4CHClCH2-
Br and m-Cl6H4CHBrCH2Br could be classified as E2 mecha-
nisms by at least one operational definition: that the transition
structure for the rate-limiting step has the â-hydrogen partially
but not completely transferred and the C-X bond partially but
not completely broken. This allows for the formation of some
intermediates along the reaction pathway; howeVer, the inter-
mediates would not contribute to the kinetics or rate of the
oVerall process.48
The KSIE associated with the reaction, kOD/kOH ≈ 2.5 at 25
°C, suggests the two remaining methanols are lost before
reaching the transition structure leading to a hydrogen-bonded
carbanion, kH1 . This intermediate partitions between losing
â-halide, kXElim, or returning hydron to a negatively charged
carbon, kH-1
:
k1H
kXElim
X-C-C-H‚‚‚-OR y z X-C-C-‚‚H-OR
8
Since a â-Br is more effective than a â-chloride in accelerat-
ing the hydron-transfer step, it should carry more of the negative
charge and has a weaker bond to carbon.49 Removing the
benzylic halide would place more charge on â-bromide and
could result in even less internal return. The Arrhenius behavior
of the isotope effect obtained for ethoxide-promoted dehydro-
brominations of C6H5CiH(CH3)CH2Br was shown to be con-
sistent with a reaction proceeding by Scheme 1. Swain-Schaad
does not hold, (kH/kT) * (kD/kT)3.26, and the temperature
kH-1
X- + CdC + HOR
The hydrogen-bonded carbanion not only has the original C-H
bond broken, but also has an elongated C-X bond. The latter
conclusion comes from the analysis of chlorine isotope effects
associated with elimination reactions of C6H5CiHClCH2Cl and
C6H5CiHClCF2Cl.43
Chloride and bromide can leave from the hydrogen-bonded
intermediate as can fluoride when it leaves from a -CH2F;
however, when fluoride is part of a trifluoromethyl group, each
C-F bond is stronger than that in a -CH2F. This requires the
formation of the free carbanion before F- can depart to form
an alkene. The near unity isotope effect for the exchange
reactions of p-CF3C6H4CiHCF3, kD/kT ) 1.08 at 25 °C,44 [similar
values are reported for C6H5CiH(CF3)2 and C6H5CiHCClCF3],34
means that hydron transfer is not occurring in a rate-limiting
step for the exchange reaction. Since the exchange is over 60
times faster than the elimination reaction, the barrier for the
formation of a hydrogen-bonded intermediate is lower than that
for loss of fluoride. The acidity of the benzylic C-H bonds
should be about the same for p-CF3C6H4CHClCF3 and
p-CF3C6H4CHClCF2Cl,45 and result in similar rates of formation
for the respective hydrogen-bonded intermediates. The element
effect, kHCl/kHF of 1.5 × 105,46 has a ∆∆Gq of 7 kcal/mol, and
this is not an unreasonable value for the breaking of a hydrogen
bond.47
dependence was modeled by using ABr /A-1 ) 0.1, E-D1
-
Elim
EH-1 ) 1340 cal/mol, and E-H1 - EBr ) 1950 cal/mol.50 The
Elim
Br
fit is good with aH ) 0.39 at 25 °C; however, the A /A-1
Elim
favoring internal return is inconsistent with the AElim/A-1 used
for the analyses in this study. The Arrhenius behavior, AH/AD
) 0.4, has been attributed to quantum-mechanical tunneling
occurring during an E2 mechanism.15 Some preliminary studies
with C6H5CiH(CH3)CH2Cl did not give satisfactory kinetics over
a reasonable temperature range to see if this system behaves
similar to C6H5CiH2CH2Cl. Work is in progress with
YC6H4CiH(CH3)CH2Cl compounds with electron withdrawing
groups to enhance reactivity and obtain better rate measure-
ments.
Experimental Section
Materials. All starting materials were purchased from Aldrich.
Solutions of alcoholic sodium alkoxide were made from the reaction
of sodium with anhydrous alcohol. Deuterated alcohols were purchased
from Aldrich and used without further purification. NMR spectra were
recorded with several instruments. Early 60-MHz 19F spectra were
For reactions of m-ClC6H4CiHClCH2X, a â-bromine acceler-
ates the formation of the hydrogen-bonded carbanion by a factor
1
obtained from a Varian HA-60 IL and H spectra were obtained with
of ≈19 over â-chlorine. The experimental element effect, kHBr
/
a Varian T-60. Recent spectra have been obtained by using a Varian
XL 300-MHz FT-NMR. A Varian VISTA system with 6 ft 0.25 or
0.06 in. packed columns with 15-20% SE-30 or OV 101 on 60/80
Chromasorb W was initially used to analyze kinetic samples. A HP
4600 with a 30 m RSL-150 polydimethylsiloxane (OV 101) with 0.53
mm i.d. and 1.2 µm film thickness is currently being used and gives
much better results.
kHCl ) 29, is largely due to the proton-transfer step, k1HBr/kH1 Cl
) 19, rather than the formal breaking of the carbon-halogen
(43) Reference 1, p 1934, has a full discussion of this with additional
references.
(44) The rate of exchange for p-CF3C6H4CDClCF3 with methanolic
sodium methoxide at 25 °C is 8.58 × 10-4 with activation parameters of
∆Hq ) 25.90 ( 0.08 and ∆Sq ) 14.3 ( 0.3. Justin C. Biffinger, unpublished
results.
(45) Relative rates of methoxide-promoted dehydrobromination of C6H5-
CHBrCF2Br, C6H5CHBrCFClBr, and C6H5CHCFBr2 vary by a factor of 4,
with the slowest being one diastereomer of C6H5CHBrCFClBr and the
fastest being the other diastereomer [Koch, H. F. In ComprehensiVe
Carbanion Chemistry: Ground and Excited State ReactiVity; Buncel, E.,
Durst, T., Eds.; Elsevier: Amsterdam, 1987; Part C, Chapter 6, p 330-
335].
Synthesis of YC6H4CiHClCH2F. A three-step synthesis started with
YC6H4MgBr and FCH2CtN to give a ketone, YC6H4COCH2F, which
was reduced by NaBiH4 after purification by column chromatography.
The YC6H4CiHOHCH2F was converted to YC6H4CiHClCH2F by
reaction with (C6H5)3P and CCl4. Yields for the first step ranged from
60% to 80%. Although yields for the last two steps were variable,
they can each be carried out in >90% yield.
(46) The rate of dehydrofluorination of p-CF3C6H4CHClCF3 at 25 °C and
activation parameters are given in Table 2. The activation parameters for
p-CF3C6H4CHClCF2Cl and rate at 0 °C are given in Table 5. The rate at
25 °C is 1.58 M-1 s-1.
(47) (a) Sano, T; Tatsumoto, N.; Niwa, T.; Yasunaga, T. Bull. Chem.
Soc. Jpn. 1972, 45, 2669. (b) Sano, T.; Tatsumoto, N.; Mende, Y.; Yasunaga,
T. Bull. Chem. Soc. Jpn. 1972, 45, 2673.
(48) Saunders, W. H., Jr. Acc. Chem. Res. 1976, 9, 19.
(49) A referee states: “It is assumed that the faster reaction with â-Br
than â-Cl is a result of more bond breaking to halogen at the transition
state: it seems at least possible that much of this rate effect could come
from relief of two serious gauche interactions as the R-carbon moves from
sp3 to sp2 hybridization.”
(50) Reference 11, p 6106.

Use of Kinetic Isotope Effects in Mechanism Studies
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9973
1. p-CF3C6H4COCH2F. 4-Bromobenzotrifluoride (12.4 g, 55
mmol) in 100 mL of ether was added dropwise to 1.3 g (55 mmol) of
Mg in 25 mL of ether. After the addition was completed, the reaction
mixture was refluxed for 30 min and then allowed to cool to room
temperature prior to the dropwise addition of 2.95 g (50 mmol) of
fluoroacetonitrile in 50 mL of ether. The reaction mixture was cooled
to 0 °C before adding 50-75 mL of 1 M HCl, which dissolves the
precipitate formed during the reaction. After separation, the aqueous
layer was extracted twice with 50 mL of ether. The combined ethereal
layers were washed with water until neutral and 50 mL of saturated
CaCl2 and the dried over MgSO4. Ether was removed by a rotary
evaporator. The remaining organic material was purified by using a
column of silica gel and 20% ether/hexane to yield 6.6 g (32 mmol,
CHdCH2 was purchased. Deuterium and tritium incorporation for â,â-
difluorostyrenes still made use of dithianes.
1. m-ClC6H4CD(OH)CH3. m-Chloroacetophenone (24.8 g, 161
mmol, 1.0 equiv) was added dropwise to a solution of 1.97 g (47 mmol,
1.17 equiv) of NaBD4 dissolved in 30 mL of 95% ethanol at a rate to
keep the reaction below 50 °C. A white solid is formed during the
addition. After 45 min of stirring, 60 mL of 3 M HCl was added to
dissolve the precipitate. The solution was heated (to evaporate ethanol)
until two distinct layers formed. After the layers separated, the aqueous
layer was washed twice with 25 mL of CH2Cl2. The combined organic
layers were washed four times with 25 mL of cold water to remove
any remaining ethanol and dried over MgSO4. The CH2Cl2 was
removed by a rotary evaporator, and the resulting 18.6 g (118 mmol,
73% yield) of m-ClC6H4CD(OH)CH3 was 99% pure according to gas
chromatography. The proton MNR at 60 MHz had signals at 1.3 (3H,
s), 4.2 (1H, s), 7.1 (3H), and 7.2 ppm (1H) for the aromatic protons.
2. m-ClC6H4CDdCH2. m-ClC6H4CD(OH)CH3 (4.48 g, 28 mmol,
1.0 equiv), 2.61 g of KHSO4, 0.05 g of 4-tert-butylcatechol, and a
magnetic stir bar were placed in a 50-mL single neck 24/40 round
bottom flask fitted with a Claisen head that had a vacuum takeoff with
a 50-mL round bottom flask receiver. Heat was applied by a 100 mL
heating mantle filled with sand around the reaction flask. Maximum
voltage was applied to rapidly raise the temperature of the sand to >200
°C over a period of 8-10 min. A vacuum of 160 to 180 mmHg was
applied and the receiver was immersed in a Dewar with liquid nitrogen.
Distillate came over between 60 and 75 °C. The 2.85 g (20 mmol) of
m-ClC6H4CDdCH2 was separated from the codistilled water, and the
GC gave the purity of 98%. The proton MNR at 60 MHz had signals
for the vinyl hydrogens at 5.1 (1H), 5.5 (1H), 7.1 (3H), and 7.2 ppm
(1H) for the aromatics. A second run started with 7.04 g of m-ClC6H4-
CD(OH)CH3 and 5.42 g of KHSO4 and resulted in a yield of only 48%.
It has been our experience that the best yields are obtained when under
5 g of starting alcohol is used.
Synthesis of YC6H4CiHClCH2Cl. Initially the chlorination of
styrenes was carried out in CCl4; however, this method can form up to
30% YC6H4CHClCHCl2 as an impurity.1 An alternate method with
dimethylformamide, DMF, as the solvent was used for most of the
YC6H4CiHClCH2Cl in this study.54 This gave a significant reduction
in the amount of the trichloride (2-5%). These small amounts of
YC6H4CHClCHCl2 did not interfere with the kinetics since YC6H4-
CCldCHCl and YC6H4CHdCCl2 formed by dehydrohalogenation had
different retention times in the gas chromatographic analysis of kinetic
samples. Since it is 30 times more reactive, the trichloride does not
interfere with the dichloride tritium kinetics and only results in a slightly
higher infinity point activity due to the small amount of YC6H4-
CTdCCl2 formed.
m-ClC6H4CDClCH2Cl. m-ClC6H4CDdCH2 (2.8 g, 20 mmol) was
dissolved in 15 mL of DMF in a 50 mL round-bottom flask with a stir
bar and placed in an ice bath. Chlorine (1.4 g, 20 mmol) was condensed
into the reaction flask. The reaction mixture was stirred for 30 min at
room temperature before a short path distillation head replaced the reflux
condenser prior to vacuum distillation. m-ClC6H4CDClCH2Cl (2.9 g,
14 mmol, 69% yield) was collected between 108 and 120 °C at 7
mmHg. The 300 MHz 1H MNR has an AB pattern for the -CH2Cl at
3.9 ppm with JA-B ) 11 Hz. The -CH2Cl of m-ClC6H4CHClCH2Cl
has an ABX pattern at 3.9 to 4 ppm with JA-X ) 6 Hz, JB-X ) 8 Hz,
and JA-B ) 11 Hz, and the benzylic proton has a d of d at 5.0 ppm.
m-ClC6H4CDClCHCl2 has a singlet at 5.95 ppm, and m-ClC6H4-
CHClCHCl2 has two doublets at 5.2 and 6.0 ppm with a J ≈ 8 Hz.
1
64%) of a yellow oil that was used in the next step. The H MNR at
60 MHz has a doublet due to the -CH2F at 5.5 ppm with a JH-F of 47
Hz, and the usual pattern for para substitution in the aromatic region.
2. p-CF3C6H4CHOHCH2F. 2-Fluoro-4′-trifluoroacetophenone (3.0
g, 14.6 mmol, 1.0 equiv) in 10 mL of ethanol was added dropwise to
a solution of 0.3 g (7.3 mmol, 2.0 equiv) of NaBH4 in 50 mL of ethanol
and warmed to 50 °C for 30 min. Addition of 25 mL of 3 M HCl
dissolved the precipitate formed during the reaction, and the resulting
mixture was added to 200 mL of ice water and extracted four times
with 50 mL of CH2Cl2. The combined methylene chloride layers were
washed three times with ice water and dried over MgSO4. Removal
of the CH2Cl2 by a rotary evaporator left 3.7 g of residue that consisted
of ≈3 g of product and ≈0.7 g of ethanol. This mixture was used in
the last step, and the amount of (C6H5)3P used was adjusted for any
ethanol in the sample.
3. p-CF3C6H4CHClCH2F. The 3.7-g product from the previous
reaction was refluxed with 6.5 g (24.8 mmol) of (C6H5)3P and 50 mL
of CCl4 for 4 h. A precipitate formed and a GC of the CCl4 layer
showed no residual alcohol.51 The CCl4 was removed by distillation,
and the remaining mixture was codistilled with water. The two layers
of the distillate were separated after adding 10 mL of CH2Cl2, and the
aqueous layer was extracted twice with 10 mL of CH2Cl2. The
combined CH2Cl2 layers were dried over MgSO4 prior to removal of
the CH2Cl2 by a rotary evaporator. The remaining 3.0 g of colorless
liquid were used for kinetics. The 1H MNR at 60 MHz has a doublet
of doublets for the -CH2F at 4.6 ppm with a JH1-H2 of 6 Hz and a
JH2-F of 47 Hz, and a multiplet for the benzylic proton at ≈5.0 ppm,
which overlaps with the upfield part of the -CH2F signal. The 19F
NMR at 200 MHz has a doublet of triplets at 211.7 ppm with a JH2-F
of 47 Hz and JH1-F of 13 Hz. The 1H 60-MHz spectrum of p-CF3C6H4-
CDClCH2F is much cleaner with -CH2F at 4.6-4.7 ppm and d with
JH-F ) 47 Hz.
Synthesis of Styrenes. Substituted styrenes and â,â-difluorostyrenes
are the precursors of the dichloro, dibromo, and bromochloro com-
pounds used in this study. â,â-Difluorostyrenes were made by
following a procedure similar to that reported for the synthesis of 2-(p-
chlorophenyl)pentafluoropentene (method II).52 This is a Wittig type
reaction of (C6H5)3PCF2CO2Li [made in situ from ClCF2CO2Li and
(C6H5)3P] and the substituted benzaldehyde with dimethylformamide
as solvent. The phosphonium salt is used in n excess [from 1.3 to 2.0]
to ensure complete reaction of the aldehyde. After the reaction is
complete, the alkene is codistilled with water from the dark reaction
mixture. The lower layer is separated and dried over CaSO4. The
yields range from 40% to 60%, and the alkenes are pure enough for
the next step without further purification.
The synthesis of YC6H4CiHdCH2 [iH ) D or T] started with a
reduction of YC6H4COCH3 with NaBiH4 to make YC6H4CiHOHCH3,
which was dehydrated using KHSO4 to give the substituted styrene,
YC6H4CiHdCH2.53 When iH ) H, a Grignard reaction of YC6H4MgBr
and CH3CHO yielded YC6H4CHOHCH3 directly. The dehydration
occurred with variable yields, but was favored for deuterium or tritium
incorporation over our old procedure Via a dithiane to obtain the
YC6H4CiHO followed by a Wittig reaction.1 When possible, YC6H4-
Synthesis of m-ClC6H4CHBrCH2Br. m-ClC6H4CHdCH2 (3.5 g,
27 mmol) and 35 mL of CCl4 were placed in a 125 mL Erlenmeyer
flask with a magnetic stir bar. A solution of 4.3 g (27 mmol) of Br2
and 2 mL of CCl4 was added dropwise, and the resulting mixture was
stirred overnight. After the CCl4 was removed the product was purified
by column chromatography by using a silica gel column and a 1:1
ratio of hexanes:CH2Cl2. The 4.0 g (49% yield) of m-ClC6H4CHBrCH2-
Br had a mp of 42.5 to 44 °C. The 300 MHz NMR has an ABX pattern
for the non-aromatic protons: HX at 5.1 ppm (JA-X ) 5 Hz, JB-X ) 11
Hz); HA at 4.0 ppm (JA-B ) 11 Hz); HB at 3.9 ppm. The m-ClC6H4-
(51) The extent of reaction has also been followed by the disappearance
of the OH proton using proton NMR.
(52) Herkes, F. E.; Burton, D. J. J. Org. Chem. 1967, 32, 1311.
(53) Roberts, R. M.; Gilbert, J. C.; Rodewald, L. B.; Wingrove, A. S.
An Introduction to Modern Experimental Organic Chemistry, 2nd ed.; Holt,
Rinehart and Winston, Inc.: New York, 1974; pp 323 and 324.
(54) De Roocker, A.; De Radzitzky, P. Bull. Soc. Chim. Belges 1970,
79, 531.

9974 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Koch et al.
CDBrCH2Br resulted in the absence of the signal at 5.1 ppm, with an
AB pattern at 4.0 ppm (2H, JA-B ) 11 Hz).
into about 100 mL of dilute aqueous HCl and extracted with 0.25 to 1
mL of CCl4 with use of a 125 mL separatory funnel with a Teflon
stopcock. After vigorous shaking for about a minute, the layers were
allowed to separate [usually waiting 15-20 min] and the CCl4 layer
was analyzed with an automated gas chromatographic instrument. A
typical run consisted of nine samples and an infinity point. The
advantage of using gas chromato-graphy to analyze the samples is that
both the disappearance of substrate and the appearance of products can
be followed.
Synthesis of m-ClC6H4CDClCH2Br. A 250 mL Erlenmeyer flask
was charged with 26 mL of water, 26 mL of 12 M HCl, and a magnetic
stir bar and placed in an ice/water bath. While the solution was stirred
small portions of 17.5 g (126 mmol) of m-chlorostyrene and 17.3 g
(126 mmol) of N-bromosuccinamide were added alternately and formed
a second layer.55 The aqueous layer was washed three times with 25
mL of CH2Cl2 and combined with the original organic layer. This was
washed once with 25 mL of 5% NaHCO3, twice with 25 mL of 20%
NaS2O5, and three times with water and dried over MgSO4. The CH2-
Cl2 was removed by a rotary evaporator prior to vacuum distillation.
After preliminary distillation, the combined cuts were carefully
redistilled and a center cut, 138-140 °C at 12-13 mmHg, was still
found to contain about 17% of the dibromide, m-ClC6H4CDBrCH2Br.
The 60 MHz 1H NMR had the -CH2Br at 3.9 for m-ClC6H4CDClCH2-
Br and at 4.0 for m-ClC6H4CDBrCH2Br.
Kinetics of Tritium Compounds. The detritiohalogenation and
exchange reactions were run with <10 µL substrate in 12-15 mL
methanolic sodium methoxide. Aliquots were taken with a 1 mL glass
or plastic syringe fitted with a 4-in. stainless steel needle. Nine points
and an infinity sample were taken. Samples for a given run were of
the same size [0.85 to 0.95 mL] and were injected into 75 mL of dilute
aqueous HCl and extracted with 10.0 mL of toluene by using a 125
mL separatory funnel with a Teflon stopcock. The mixture was shaken
for the same length of time [30-60 s] and allowed to stand for 20 min
before the layers were separated. The toluene layer was dried for 20
min with Drierite. A 5.0-mL sample of the dried toluene was mixed
with 5 mL of a scintillation cocktail [INSTA-FLUOR (TM) or OPTI-
FLUOR (R)] and counted with a Packard Tri-Carb 4640.
Since neither the m-ClC6H4CDBrCH2Br nor the reaction products,
m-ClC6H4CBrdCH2 or m-ClC6H4CDdCHBr, interfere with the GC
analysis, this sample was used for kinetic runs. The product ratios for
the synthesis of m-ClC6H4CHClCH2Br was similar to that for the
deuterium compound.
Synthesis of YC6H4CHClCF2Cl. The chlorination reaction of C6H5-
CHdCF2 dissolved in CCl4 has been described in detail.56 The 19F 60
MHz NMR spectra of all the YC6H4CHClCF2Cl shows a clear ABX
Acknowledgment. We are grateful to the National Science
Foundation [NSF Grant Nos. CHE-8316219, CHE-870650, and
INT-8212693], the donors of the Petroleum Research Fund,
administered by the American Chemical Society, and the
William and Flora Hewlett Foundation of Research Corporation
for grants to support this research. Support for the Ithaca College
students working summers in Leiden was received from grants
from the Faculty of Mathematics and Natural Sciences of Leiden
University and the Ithaca College Dana Foundation Internship
Program. We wish to thank Prof. Andrew Streitwieser for
encouragement and helpful suggestions. H.F.K. also thanks Prof.
William H. Saunders for fruitful discussions while at the
University of Rochester as a Dana Visiting Professor of
Chemistry.
pattern: FA at 57-58 ppm, FB at 61-62 ppm, JA-H ≈ 7 Hz, JB-H
≈
10 Hz, and JA-B ) 162-164.
Kinetics of Hydrogen and Deuterium Compounds. Dehydroha-
logenation kinetics were carried out under pseudo-first-order conditions
with ratios of alkoxide to substrate larger than 15 to 1. A solution of
alcoholic sodium alkoxide solution [30-40 mL ≈ 0.3 M] in a 50-mL
Erlenmeyer capped with a serum stopper was allowed to reach
temperature in a bath held to ( 0.1 °C. The reactant, 25-75 µL, and
a standard, 15-50 µL, were mixed and injected into the solution by a
1-mL glass Luer-lok syringe equipped with a 4 in. stainless steel needle.
Disposable glass or plastic 5-mL syringes fitted with 4-in. stainless
steel needles were used to take 3-4-mL samples that were injected
(55) Buckles, R. E.; Long, J. W. J. Am. Chem. Soc. 1951, 73, 998.
(56) Burton, D. J.; Anderson, A. L.; Takei, R.; Koch, H. F.; Shih, T. L.
J. Fluorine Chem. 1980, 16, 229.
JA970189N