Large primary kinetic isotope effects in the abstraction of hydrogen from organic compounds by a fluorinated radical in water

Article abstract of DOI:10.1039/b405075d

Isotope effects have been measured for the abstraction of hydrogen from a series of organic substrates by the perfluoro radical, Na^{+ -}O₃SCF₂CF₂OCF₂CF ₂, in water. Both primary and secondary deuterium isotope effects were measured, with the primary isotope effects ranging in value from 4.5 for isopropanol to 19.6 for acetic acid. The values for the α- and β-secondary deuterium isotope effects were 1.06 and 1.035, respectively. It was concluded that tunneling contributes significantly to the production of the observed, large primary kinetic isotope effects in these C-H abstraction reactions.

Full text of DOI:10.1039/b405075d

View Article Online / Journal Homepage / Table of Contents for this issue
Large primary kinetic isotope effects in the abstraction of
hydrogen from organic compounds by a fluorinated radical
in water†
Joseph A. Cradlebaugh,a Li Zhang,a A. B. Shtarev,a Bruce E. Smart*‡b and
William R. Dolbier, Jr.*a
a Department of Chemistry, University of Florida, Gainesville, FL 32611-7200, USA
b DuPont Central Research and Development, Experimental Station, Wilmington,
DE 19880, USA
Received 5th April 2004, Accepted 26th May 2004
First published as an Advance Article on the web 24th June 2004
Isotope effects have been measured for the abstraction of hydrogen from a series of organic substrates by the
perfluoro radical, Naϩ ϪO3SCF2CF2OCF2CF2 , in water. Both primary and secondary deuterium isotope effects
ؒ
were measured, with the primary isotope effects ranging in value from 4.5 for isopropanol to 19.6 for acetic acid.
The values for the α- and β-secondary deuterium isotope effects were 1.06 and 1.035, respectively. It was concluded
that tunneling contributes significantly to the production of the observed, large primary kinetic isotope effects in
these C–H abstraction reactions.
Introduction
Recently we have been engaged in a broad study involving the
determination of rate constants of hydrogen abstraction from
ؒ
organic substrates by fluorinated radicals, Rf , first in non-polar
solvents,1 such as 1,3-bis-trifluoromethylbenzene (BTB), and
most recently in water.2–4 The rate constants for abstraction of
hydrogen from isopropanol in these two solvents are given
below to exemplify, first the rather large rate constants of such
processes, and secondly the modest rate-enhancing effect of the
polar solvent, water.1,3,4
studies in BTB and H2O, respectively, and they were therefore
used as the D-transfer agents in these respective studies. These
rate constants can be compared with their kH values for the
respective non-deuterated compounds: 4.9 × 105 MϪ1 sϪ1 for
t-BuMe2SiH (in BTB) and 3.3 ( 1.0) × 104 MϪ1 sϪ1 for THF
(in H2O),1 which allowed the isotope effects (kH/kD) for H versus
D abstraction from these two compounds to be calculated:
3.3 and 7.9, respectively – nothing overtly extraordinary.
A similar isotope effect (kH/kD = 3.1) was observed for another
silane, Et3SiH, in another non-polar solvent, 1,1,2-trichloro-
1,2,2-trifluoroethane,5 whereas the isotope effect for THF
versus THF-d8 in BTB was 7.0 at 25 ЊC.1
However, particularly in the water study, many other
prospective D-transfer agents were tested for possible use,
including CD3OD and CD3COCD3. During the course of such
testing, both the H-transfer and the D-transfer rate constants
were determined for these compounds, and the results were
surprising in that unusually large isotope effects (kH/kD = 11.4
and 17.0, respectively) were observed for the progressively
“slower” H-transfer agents. Of course, these two observed
isotope effect values, as well as the earlier mentioned value for
THF, derive from a combination of primary and secondary
deuterium isotope effects, but it will be seen that the actual
primary isotope effects, once corrected for the small secondary
effects, remain quite large.
These absolute rate constants were determined by competi-
tion studies in which the Rf radical abstracted hydrogen from
ؒ
the substrate in competition with abstraction of a deuterium
from a standard, carefully chosen deuterium transfer agent for
which the rate constant for D-transfer had been determined.
The requirements for the deuterium transfer agent were
stringent, requiring the agent to participate in creating a clean,
radical chain process that gave high conversions and excellent
mass balance: that is, high yields of the two products, with
virtually no observable side products. The rate constant for
D-transfer must also be of a magnitude that allowed a viable
competition between H- and D-transfer to be observed.
t-BuMe2SiD, with a measured kD of 1.5 ( 0.3) × 105 MϪ1 sϪ1,1
and THF-d8, with a measured kD of 4.2 ( 1.2) × 103 MϪ1 sϪ1,4
fulfilled these requirements admirably for the competition
These observations led to a more systematic examination of
† Electronic supplementary information (ESI) available: Tables of kin-
etic data and plots of kinetic data. See http://www.rsc.org/suppdata/ob/
b4/b405075d/
the primary and secondary kinetic isotope effects that derive
Ϫ
from hydrogen transfer to the Naϩ O3SCF2CF2OCF2CF2
ؒ
radical, 1, the results of which are given in Tables 1–3 below.
‡ DuPont Contribution No. 8366.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 0 8 7 – 2 0 9 1
2087

View Article Online
Table 1 Observed kinetic isotope effects for direct bimolecular competition between hydrogen and deuterium transfer from select organic com-
pounds to the Naϩ ϪO3SCF2CF2OCF2CF2 radical, 1, in water at 25 ЊC
ؒ
H-Substrate
D-substrate
kH/kD a (observed)
kH/104 MϪ1 sϪ1 b, c
kD/104 MϪ1 sϪ1 c
(CF3)2CHOϪNaϩ
Isopropanol
(CD3)2CHOH
THF
Ethanol
Ethanol
CH3CHDOH
Methanol
CHD2OH
Acetone
(CF3)2CDOϪNaϩ
(CD3)2CDOH
THF-d8
THF-d8
CD3CD2OH
CH3CD2OH
5.9 ( 0.4)
5.8 ( 0.1)
8.9 ( 0.3)
7.9 ( 0.4)
8.8 ( 0.2)
8.11 ( 0.12)
7.39 ( 0.04)
11.4 ( 0.4)
4.83 ( 0.02)
17.0 ( 0.4)
22.2 ( 0.7)
11 ( 3)6
4.8 ( 1.4)
3.7 ( 1.1)
3.3 ( 1.0)
1.2 ( 0.4)
1.2 ( 0.4)
—
0.18 ( 0.05)
—
0.006 ( 0.002)
0.002 ( 0.0006)
1.86 ( 0.6)
0.83 ( 0.25)
0.42 ( 0.13)
0.42 ( 0.13)
0.14 ( 0.04)
0.15 ( 0.04)
—
0.016 ( 0.005)
—
0.00035
CD3OH
Acetone-d6
CD3CO2H
Acetic acid
0.00009
a Errors in this column are standard deviations. b From reference 4, except where noted. c These LFP-derived absolute rate constants have an
estimated error limit of 30%.4
Table 2 Data used to determine secondary deuterium isotope effects
a
a
H-Substrate
D-Substrate
kH
kD
kH/kD (secondary)
kH/kD (per D)
(CH3)2CHOH
CH3CD2OH
CH3CHDOH
CHD2OH
(CD3)2CHOH
CD3CD2OH
4.8
0.148
kH/kD = 7.39
kH/kD = 4.83
3.74
0.136
1.28
1.088
1.07
1.10
1.04b (β)
1.03c (β)
1.07 (α)
1.05
a Rates are all expressed as k/104 MϪ1 sϪ1, and are 30%. b Obtained by taking the 6√ of 1.28. c obtained by taking the 3√ of 1.088.
Table 3 Primary isotope effects (after correction for α- and β-deuterium isotope effects)
a
a
Substrate
kH
kD
kH/kD (uncorr)
Corrections
kH/kD (primary)
(CF3)2CHOϪNaϩ
Isopropanol
THF
Ethanol
Ethanol
Methanol
Acetone
Acetic acid
116
1.86
0.83
0.42
5.9
5.8
7.9
8.8
8.1
11
none
5.9
4.5
4.84
6β (1.28)
1α,2β (1.14)
1α,3β (1.17)
1α (1.07)
2α (1.10)
2α (1.12)
2α (1.12)
3.34
6.9
1.24
0.136b
0.148c
0.016
0.00035
0.00009
7.5
7.6
10.0
15.2
19.6
1.24
0.184
0.0064
0.0024
17
22
a All these LFP-derived absolute rate constants have an estimated error of 30%.4 b CD3CD2OH as kD substrate. c CH3CD2OH as kD substrate.
Results
All of the isotope effects (kH/kD), both primary and secondary,
were obtained via simple, direct competition experiments
involving the photolysis of NaϩϪO3SCF2CF2OCF2CF2I in the
presence of mixtures of the respective protiated and deuterated
substrates, varying relative concentrations in order to establish
the linearity (pseudo first order nature) of the concentration
dependence of the substrates towards proton/deuteron transfer.
The conversions were 100%, with the mass balance of products
always being ∼90%. The linear relationship between the ratio of
RfH/RfD products and the ratio of protiated and deuterated
products defined the value of the isotope effect (kH/kD). The
derivation of both relative and absolute rate constants from
such data has been discussed in detail in earlier papers.1,4
A
typical plot of data for the reaction of 1 with CH3COCH3/
CD3COCD3 mixtures is provided in Fig. 1.
Fig. 1 Plot of [RfH]/[RfD] vs. [(CH3)2CO]/[(CD3)2CO].
CD2OH, both α- and β). The absolute values of the kH rate
constants were known from previous work.4 Therefore,
knowledge of the kH/kD values in Table 1 allows determination
of the values for kD for the given deuterated compound.
Secondary isotope effects
The “observed” kinetic isotope effects given in column 3 of
Table 1 in all cases except the first derive from a combination of
primary and secondary kinetic deuterium isotope effects, the
latter being either α- or β- in nature (or in the case of CD3-
Because of the multiple deuterium substitution that is present
in THF-d8, CD3CD2OH, CH3CD2OH, CD3OH, acetone-d6 and
CD3CO2H, it is necessary to have some measure of expected
α- and β-secondary deuterium isotope effects in order to be able
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 0 8 7 – 2 0 9 1
2088

View Article Online
to calculate the actual primary isotope effects for these sub-
strates. For example, in the case of CD3CD2OH, when the rate
constant, kD, for α-deuterium abstraction is measured, the
remaining α-deuterium and the three β-deuteriums give rise to
secondary isotope effects that lower the rate constant compared
to what it would have been had only hydrogens remained in the
radical product, as depicted in Fig. 2.
forming reactions.8,9 For example, in studies of the deazetation
of deuterium labeled azo compounds, Seltzer and coworkers
found that β-secondary deuterium isotope effects, which derive
largely from hyperconjugative interaction of the isotopically-
labeled site with the incipient β-radical site, were generally
much smaller for radical-forming reactions than for carbo-
cation-forming reactions. In contrast, α-secondary deuterium
isotope effects, which can be understood as deriving from a
change in hybridization at the radical-forming carbon atom,
have been found to be of similar magnitude to those deriving
from carbocation-forming reactions.10 Thus the average of our
two observed α-secondary deuterium isotope effects, kH/kD
=
1.06, is somewhat smaller than might have been expected based
on the limited data available in the literature for radical-forming
processes. However, the faster, exothermic hydrogen abstraction
reactions of our study should have much earlier transition
states than the endothermic dissociative reactions studied
by Seltzer, a fact consistent with the smaller α-secondary
deuterium isotope effect that we have observed.
Fig.
2
Depiction of transition state for D-abstraction from
pentadeuteroethanol.
In order to determine the values of the secondary isotope
effects, four additional competition studies were carried out.
These were (a) the intermolecular competition between
(CD3)2CHOH and THF-d8, (b) the intermolecular competition
between CH3CD2OH and CH3CH2OH; and intramolecular
competitions using (c) CH3CHDOH and (d) CHD2OH. The
data obtained from these experiments are contained in Tables 1
and 2.
The first two experiments allowed independent deter-
minations of the β-secondary deuterium isotope effect, the
values of which were calculated to be 1.28 and 1.088, for
isopropanol and ethanol respectively, or 1.04 and 1.03 per
deuterium in each case (using sixth route and cube root of
the overall isotope effects, respectively).
The results from the intramolecular competition experiments
using CH3CHDOH and CHD2OH allowed two determinations
of the value of the α-secondary deuterium isotope effect. The
isotope effect for the CH3CHDOH system can be calculated
from the raw RfH/RfD integral as follows:
Primary isotope effects
With both the α- and the β-secondary isotope effects now in
hand, it becomes possible to convert the observed isotope
effects in Table 1 to the pure primary deuterium isotope effects
for each of the substrates. For isopropanol, ethanol, and
methanol the specific values of the secondary isotope effects
that were determined have been used, whereas for the THF,
acetone and acetic acid calculations, the per-deuterium value of
1.06 for the α- and 1.035 for the β-secondary isotope effects
were used. The calculated primary isotope effects are given in
Table 3.
Discussion
The values of the isotope effects for the latter three substrates
are considerably larger than what are generally considered to be
the theoretical limits for semiclassical H/D primary isotope
effects. It can be seen that the values exhibit a correlation with
the absolute rate constants for H-transfer, which vary over a
considerable range, with acetic acid being approximately 3000
times less reactive than isopropanol (per hydrogen, 9000
times!). The absolute rate constants themselves correlate with
the C–H BDE’s of the respective substrates.
These isotope effects appear to be real. The reactions are
clean, and good linear correlations with concentrations are
observed. One possible trivial explanation for the data might
have been that H-transfer from the solvent water could have
become increasingly competitive with H-transfer from the less
reactive substrates. However, the competition experiments for
acetone were run in both H2O and D2O, with the observed
kH/kD in D2O being 18.5 0.7, which indicates that H-transfer
from solvent water cannot be competitive with H-transfer from
the substrates under the conditions of the study.
1
However, a 500 MHz NMR spectrum of the synthesized
CH3CHDOH indicated that it contained a 1% diethyl ether
contaminant (4% integral of its CH2 groups versus 100% inte-
gral for the CHD of the deuteroethanol). This small amount of
contaminant will have a small, but not negligible impact upon
the (CF2H)/(CF2D) integral ratio, which when taken into
account yields a corrected value of 1.07 for the α-secondary
deuterium isotope effect for the ethanol system.7
The value of the isotope effect in the methanol system can
likewise be calculated as shown below:
Tunneling must be considered as a possible contributing
factor to such large isotope effects that are certainly outside of
the generally accepted semiclassical limits for hydrogen transfer
isotope effects, where one assumes that the zero-point-energy
difference associated with carbon–hydrogen stretching in the
substrate is completely lost in the transition state.11,12 Isotope
effects moderately larger than the ∼6.9 limit at 25 ЊC can be
accommodated by allowing weakening of bending modes, but
when such large values as those found in Table 1 are observed,
Although the absolute rate constants of Table 1 are not
reliable more than 30%, the data obtained in our competition
experiments have standard deviations generally in the 3–4%
range, which allows the values for the secondary isotope effects
to be credible. The relatively small β-effects are consistent with
the limited data that have been previously reported for radical-
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 0 8 7 – 2 0 9 1
2089

View Article Online
one is almost compelled to invoke tunneling.12 Therefore, as a
classical probe of tunneling, a temperature dependence study
was carried out for the acetone reaction, although the range of
temperatures that could be examined was limited severely by
the fact that water was the solvent. Thus the values determined
for kH/kD at 24, 56, and 80 ЊC were 16.6 ( 0.7), 12.6 ( 0.1), and
10.2 ( 0.5), respectively. An Arrhenius plot of this data (Fig. 3)
indicates a linear correlation between ln(kH/kD) and 1/T , with
Ea(D) Ϫ Ea(H) = 1.8 kcal molϪ1 and AH/AD = 0.80.
The kinetic studies were run in pyrex NMR tubes containing
a sealed capillary tube (CFCl3 in C6D6) as the internal standard.
For each kinetic study, a group of samples were prepared at the
same time. The NMR tubes were capped with rubber septa, and
wrapped with Teflon tape before chemicals were added. The
radical precursor, Naϩ ϪO3SCF2CF2OCF2CF2I,2 was used as a
stock solution (17.8% by weight) and added to the NMR tubes
with a micro-syringe. Quantities of the protiated and deuter-
ated substrates were added with syringes and weighed on the
balance. The samples were degassed by 3 cycles of freeze–
pump–thaw. After 19F NMR spectra were taken, they were
irradiated using a RPR-204 Rayonet photochemical reactor.
The 19F NMR spectrum was taken again after 24 hours. The
acquisition time of NMR was at least 15 minutes to assure
accurate integration. The ratios of products were obtained
directly from the ratios of integration of the CF2H and CF2D
signals. The conversion and yield were obtained by comparison
of the integration of the CF2I peak in the starting material and
the (CF2HϩCF2D) peaks in the products relative to the internal
standard.
ؒ
Fig. 3 Plot of ln(kH/kD) for reaction of Rf radical, 1, with acetone vs.
1/T .
Procedure for measurement of the intramolecular isotope effects
of CH3CHDOH and CHD2OH
As Kreevoy has indicated,13 there is probably no unambigu-
ous way to demonstrate tunneling for reactions carried out
around 300 K.14 However, he and others contend that, even in
this temperature range, values of Ea(D) Ϫ Ea(H) > 1.20 kcal
The CH3CHDOH was prepared by a literature procedure,23
and, after purification, it was analyzed by 1H NMR. No signifi-
cant undeuterated ethanol could be detected (<1%), but a 1%
impurity of diethyl ether was detected (4% integral for its
CH2 groups versus the 100% due to the CHD group of
CH3CHDOH). The small impact of this impurity on the
measured [CF2H]/[CF2D] ratio was corrected for in the calcu-
lation of the α-secondary isotope effect. The CHD2OH was
purchased from Isotech, Inc.
For both CH3CHDOH and CHD2OH, two repetitive kinetic
studies were run in pyrex NMR tubes containing a sealed
capillary tube (CFCl3 in C6D6) as the internal standard, capped
with rubber septa, and wrapped with Teflon tape before the
chemicals were added.
In the typical case for CH3CHDOH, the IRfSO3Na in
water was used as a stock solution (17.8% by weight) and
added to the NMR tubes with a micro-syringe. Then, 80 µL of
monodeuteroethanol (14.3 × 10Ϫ4 moles) was added to each
tube. The samples were then degassed by three freeze–
pump–thaw cycles. After 19F NMR spectra of the two samples
were taken, the samples were irradiated using a RPR-204
Rayonet photochemical reactor. The 19F NMR spectrum
of each sample was taken again after 24 hours. The NMR
acquisition time was at least 15 minutes to assure accurate
integration. The product ratios were obtained, as usual, from
the ratios of integration of the CF2H and CF2D signals,
with multiple integrations being carried out for each. The
conversion and yield were obtained from the integration
of the CF2I peak in the starting material versus those of the
(CF2H ϩ CF2D) peaks in the products relative to the internal
standard.
molϪ1 will generally signal tunneling, as will values of AH/AD
<
1.0 for hydrogen transfer between massive, polyatomic donors
and acceptors.11–13 Although our Arrhenius data must be con-
sidered suspect because of the small temperature range and the
small number of data points, nevertheless the data are consist-
ent with these criteria for the involvement of tunneling in the
H-transfer process. Moreover, Truhlar has analyzed kinetic
ؒ
data for the related reaction of CF3 with CD3H within the
context of his calculations of the kinetic isotope effect for this
reaction,15 and he finds tunneling must be included in the calcu-
lation in order to attain agreement with the experimental data.16
Similarly, Roberto-Neto has analyzed isotope effect data for
the reaction of Cl with ethane,17 Michelsen18 and Hewitt19
ؒ
ؒ
ؒ
have similarly analyzed such data for the reaction of Cl with
methane, and Osman has examined the reaction of HO with
isopropanol,20 all concluding that tunneling effects make a
significant contribution to the H-transfer rate constants in these
reactions.
Carbon–hydrogen abstractions by alkyl radicals, and even
more so by perfluoroalkyl radicals, should involve relatively
sterically hindered transition states. It is recognized that
such sterically-constrained transition states for hydrogen
abstraction can lead to a steep rise in potential energy upon
close approach of the reactants to each other, resulting in a high
and thin potential barrier that can lead to large tunneling
contributions.12,21,22
Taking all of this into account, in the absence of a plausible
alternative explanation, we conclude that tunneling is probably
a significant factor in producing the large kinetic isotope
effects that are observed in our C–H abstraction reactions of
perfluoroalkyl radical 1 in water.
For CH3CHDOH, the average of the raw CF2H/CF2D ratios
was 7.39 ( 0.04).
For CHD2OH, the average of the raw CF2H/CF2D ratios was
4.83 ( 0.02).
Tables of kinetic data and plots of kinetic data for all kinetic
experiments are available in the ESI.†
Experimental
All deuterated compounds were purchased in >98% purity
from Isotech, Inc, except CH3CHDOH and (CD3)2CHOH,
which were synthesized by reduction of acetaldehyde by
LiAlD4, and reduction of acetone-d6 by LiAlH4, respectively.
Acknowledgements
Support of this research in part by the National Science
Foundation is acknowledged with thanks. We are grateful to
Dr. Ion Ghiviriga for assistance with NMR analyses of deuter-
ated substrates. Insightful discussions with Dr. K. U. Ingold
were also much appreciated.
General procedure for competition kinetic studies
The kinetics were carried out in a manner similar to the earlier
studies in BTB, and the analysis of the data was also analogous.1
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 0 8 7 – 2 0 9 1
2090

View Article Online
11 R. P. Bell, The Tunnel Effect in Chemistry, Chapman and Hall,
References
London, 1980.
1 A. B. Shtarev, F. Tian, W. R. Dolbier, Jr. and B. E. Smart, J. Am.
Chem. Soc., 1999, 121, 7335.
2 L. Zhang, W. R. Dolbier, Jr., B. Sheeller and K. U. Ingold, J. Am.
Chem. Soc., 2002, 124, 6362.
3 L. Zhang, The Addition and Hydrogen Abstraction Reactions of a
Fluorinated Radical in Aqueous Solution, Ph. D. Dissertation,
University of Florida, Gainesville, 2001.
4 L. Zhang, J. Cradlebaugh, G. Litwinienko, B. E. Smart, K. U. Ingold
and W. R. Dolbier, Jr., Org. Biomol. Chem., 2004, 2, 689.
5 W. R. Dolbier, Jr. and X. X. Rong, J. Fluorine Chem., 1995, 72,
235.
12 L. Melander and W. H. Saunders, Jr., Reaction Rates of Isotopic
Molecules, John Wiley and Sons, New York, 1980.
13 Y. Kim and M. M. Kreevoy, J. Am. Chem. Soc., 1992, 114, 7116.
14 K. U. Ingold, Pure Appl. Chem., 1984, 56, 1767.
15 Y. P. Liu, D. H. Lu, A. Gonzalez-Lafont, D. G. Truhlar and B. C.
Garrett, J. Am. Chem. Soc., 1993, 115, 7806.
16 T. E. Sharp and H. S. Johnston, J. Chem. Phys., 1962, 37, 1541.
17 O. Roberto-Neto and F. B. C. Machado, Theor. Chem. Acc., 2001,
107, 15.
18 H. A. Michelsen and W. R. Simpson, J. Phys. Chem. A, 2001, 105,
1476.
6 J. Cradlebaugh, L. Zhang, G. R. Shelton, G. Litwinienko, B. E.
Smart, K. U. Ingold and W. R. Dolbier, Jr., Org. Biomol. Chem.,
2004, DOI: 10.1039/b405075d (preceding paper).
7 The method used for calculation of the “corrected” value of the
isotope effect can be found in the ESI†.
8 S. Seltzer and E. J. Hamilton, Jr., J. Am. Chem. Soc., 1966, 88, 3775.
9 A. Tsolis, P. P. Hunt, J. K. Kochi and S. Seltzer, J. Am. Chem. Soc.,
1976, 98, 992.
19 G. D. Boone, F. Agyin, D. J. Robichaud, F.-M. Tao and S. A. Hewitt,
J. Phys. Chem. A, 2001, 105, 1456.
20 N. Luo, D. C. Kombo and R. Osman, J. Phys. Chem. A, 1997, 101,
926.
21 E. S. Lewis and L. H. Funderburk, J. Am. Chem. Soc., 1967, 89,
2322.
22 H. Wilson, J. D. Caldwell and E. S. Lewis, J. Org. Chem., 1973, 38,
564.
10 S. Seltzer and S. G. Mylonakis, J. Am. Chem. Soc., 1967, 89,
6584.
23 A. J. Gellman, M. T. Buelow, S. C. Street and T. H. Morton, J. Phys.
Chem. A, 2000, 104, 2476.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 2 0 8 7 – 2 0 9 1
2091