Direct Detection of Ion Pair Formation and Collapse in a Migration Reaction of a β-Phosphate Radical

Article abstract of DOI:10.1021/ol036052l

(Equation presented) In solutions of trifluorotoluene or toluene containing 2,2,2-trifluoroethanol, the β-phosphate radical heterolyzed to give a detectable ion pair, identified as a solvent-separated species. Rate constants for the radical fragmentation reaction forming the ion pair, for ion pair collapse, and for diffusive escape to free ions were measured. The kinetics and entropy of activation for fragmentation indicate that the rearrangement reaction occurs by a heterolysis pathway in all solvents.

Full text of DOI:10.1021/ol036052l

ORGANIC
LETTERS
2003
Vol. 5, No. 26
5055-5058
Direct Detection of Ion Pair Formation
and Collapse in a Migration Reaction of
a â-Phosphate Radical
Laurent Bagnol, John H. Horner,* and Martin Newcomb*
Department of Chemistry, UniVersity of Illinois at Chicago, 845 W. Taylor St.,
Chicago, Illinois 60607
men@uic.edu
Received October 20, 2003
ABSTRACT
In solutions of trifluorotoluene or toluene containing 2,2,2-trifluoroethanol, the â-phosphate radical heterolyzed to give a detectable ion pair,
identified as a solvent-separated species. Rate constants for the radical fragmentation reaction forming the ion pair, for ion pair collapse, and
for diffusive escape to free ions were measured. The kinetics and entropy of activation for fragmentation indicate that the rearrangement
reaction occurs by a heterolysis pathway in all solvents.
Heterolytic fragmentation reactions of radicals containing
leaving groups adjacent to the radical center produce alkene
radical cations under nonoxidative conditions. Such reactions
are known in diverse settings, including strand cleavage of
DNA by drugs under anaerobic conditions1 and organic
synthesis.2,3 The pathway for one-atom migrations of ester
groups in â-ester radicals might involve a radical heterolysis
reaction that gives a radical cation-anion pair followed by
ion recombination. For many such rearrangements, however,
activation parameters and stereochemical labeling studies
suggested that the reactions were concerted.2 Concerted
pathways for migrations also were found computationally,4
and, indeed, computational results indicate that concerted or
associative pathways are available for any reaction of a
radical with a â-leaving group, including eliminations and
nucleophilic substitutions.5
As synthetic applications of radicals with â-leaving groups
have increased in number,3 the mechanistic details of their
reactions have become a more important subject. A concerted
migration of a group would avoid side reactions, whereas
migration via an intermediate ion pair would compete with
reduction and nucleophilic capture of the radical cation. We
report here direct spectroscopic detection of formation and
collapse of ion pairs formed in a migration reaction of a
â-phosphate radical in high-polarity media. The results
indicate that a common heterolysis pathway for radical
isomerization exists in solvents of high and low polarity.
The â-phosphate radical 1 was produced in laser flash
photolysis (LFP) studies from the corresponding PTOC ester
precursor (Scheme 1 and Supporting Information).6,7 Radical
1 is known to rearrange to benzylic radical 2 in low- and
medium-polarity solvents without formation of detectable
intermediates.6 When radical 1 was produced in mixtures of
toluene or (trifluoromethyl)benzene (trifluorotoluene, TFT)
containing 2,2,2-trifluoroethanol (TFE) or 2,2,3,3,4,4,5,5-
(1) Burger, R. M. Chem. ReV. 1998, 98, 1153-1169.
(2) Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. W. Chem. ReV.
1997, 97, 3273-3312.
(3) Crich, D. In Radicals in Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2; pp 188-206.
(4) Zipse, H. J. Am. Chem. Soc. 1997, 119, 2889-2893. Zipse, H. J.
Am. Chem. Soc. 1997, 119, 1087-1093.
(6) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X. H.; Martinez, F. N.;
Newcomb, M.; Wink, D. J.; Yao, Q. W. J. Am. Chem. Soc. 1998, 120,
211-212.
(7) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X. H.; Newcomb, M.;
Whitted, P. O. Tetrahedron 1999, 55, 3317-3726.
(5) Zipse, H. Acc. Chem. Res. 1999, 32, 571-578.
10.1021/ol036052l CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/04/2003

The rapid decay process apparent in Figure 1A was too
fast for a bimolecular reaction (k > 1 × 107 s-1) and must
result from collapse of a short-lived ion pair. The slower
decay process had about the same rate constant (k ≈ 2 ×
105 s-1) as that reported for decay of diffusively free radical
cation 3, produced by photoejection from the corresponding
styrene in TFE solvent.8 The measured rate constants for
growth (kform) and fast decay (kdecay) of 3 are listed in Table
1. When diffusively free radical cation 3 was observed, we
Scheme 1
Table 1. Rate Constants for Reactions of 1a
octafluoropentan-1-ol, we observed rapid formation and
biexponential decay of a transient (Figure 1). Radical cation
% TFEb
ET(30)c
kform
kdecay kcollapse kescape τ (µs)d
10
5
2.5
1
OFPe
5%, PhMef
56.2
55.6
55.0
53.8
53.2
51.3
>200
180
150
115
100
90
18
22
35
50
70
50
7
16
31
50
70
50
11
6
4
<1
<1
<1
0.09
0.17
0.25
a Rate constants at 20 ( 2 °C in units of 106 s-1
.
b Percentage of TFE
by volume in TFT unless noted. c Measured ET(30) value. d Lifetime of ion
pair defined as 1/kescape.
e 2,2,3,3,4,4,5,5-Octafluoropentan-1-ol (0.2 M) in
TFT. f TFE (5% by volume) in toluene.
used the ratio of free radical cation to total radical cation
formed in the ion pair to divide the decay rate constant into
a collapse rate constant (kcollapse) and a diffusive escape rate
constant (kescape). Solvent polarity, as measured on the
ET(30) solvatochromatic scale,9 also is listed in the table.
As solvent polarity increased, the rate constants for formation
of radical cation 3 and for escape from the ion pair increased,
and the rate constant for ion pair collapse decreased.
The kinetic results and yields (see below) indicate the
complex ion pair model for radical heterolysis reactions
shown in Scheme 2. The heterolysis reaction gives a contact
Figure 1. Kinetic traces at λ ) 386 nm shown on two time scales
for reactions of radical 1 in TFT solutions containing, from the
top, 10, 5, 2.5, and 1% TFE.
3 was identified as the transient in both the fast and slow
decay processes. The same spectrum was obtained from both
decay processes and also from heterolysis of the analogous
â-mesylate radical (Figure 2 and Supporting Information),
and these spectra matched that reported for 3, which has λmax
at 386 nm.8
Scheme 2
ion pair (CIP) that can collapse to the original radical or its
isomer or can solvate to a solvent-separated ion (SSIP), and
the SSIP can return to the CIP or further solvate to give
diffusively free ions. The varying yields of transient radical
Figure 2. Spectra of radical cation 3. Blue: spectrum of diffusively
free 3 from reaction of the mesylate analogue of radical 1 in 2.5%
TFE in TFT. Green: spectrum of diffusively free 3 from radical 1
in 10% TFE in TFT (3×, decay spectrum from 0.3 to 4.5 µs).
Red: spectrum of ion-paired 3 from radical 1 in 10% TFE in TFT
(3×, decay spectrum from 46 to 200 ns).
(8) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115, 6564-
6571.
(9) Reichardt, C. Chem. ReV. 1994, 94, 2319-2358.
5056
Org. Lett., Vol. 5, No. 26, 2003

cation 3 with solvent polarity, apparent in Figure 1, require
that a very short-lived and undetected species, i.e., the CIP
with τ < 1 ns, exist but collapse to rearrangement product
in competition with formation of the SSIP. The detectable
transient ion pair with τ in the range of 100-250 ns is the
SSIP. Short lifetimes for the CIP, reflecting sub-nanosecond
solvation phenomena, and longer lifetimes for the SSIP are
expected.10
The SSIP lifetimes found in this work are exceptionally
long for a polar solvent, but they are similar to those found
in low-polarity solvents.10 We speculate that the solvent
mixtures “de-mixed” around the incipient ion pair to give a
high-polarity local solvent and a lower polarity bulk solvent.
This conclusion is supported by the observation of a highly
nonlinear plot of ET(30) versus solvent composition (Sup-
porting Information), which is a signature of solvent demix-
ing in the immediate vicinity of the zwitterionic salt used to
measure ET(30). Thus, high local solvent polarity resulted
in fast fragmentations of 1, whereas the diffusive escape rate
constants for the SSIP reflect the lower polarity of the bulk
medium.
is 10-20 times faster than solvation to the SSIP, which in
turn must be at least as fast as the measured rate of ion pair
formation listed in Table 1.
Variable-temperature studies were performed in a solution
of 5% TFE in toluene. In this solvent mixture, the ion pair
collapsed to product radical 2, and no diffusively free radical
cation 3 was formed. The rate constants for formation of 3
were described by log kform ) (11.0 ( 0.3) - (4.2 ( 0.4)/θ
(errors are 2σ, θ ) 2.3RT kcal/mol), and the rate constants
for ion pair collapse were described by log kdecay ) (13.6 (
0.5) - (8.0 ( 0.5)/θ. Note that, in this case, the “collapse”
reaction involves two processes, desolvation of a solvent-
separated ion pair and true collapse of the contact ion pair
to rearranged product 2, and the desolvation step must be
the slower reaction because CIP collapse has k′recom > 1 ×
109 s-1 (see above). Thus, the relatively high activation ener-
gy measured for the collapse reaction is the energy required
to desolvate the ions, and the large entropy term in this
process reflects liberation of tightly bound solvent molecules.
The kinetic results in this work combined with those for
rearrangement of radical 1 to radical 2 in low- and medium-
polarity solvents6 indicate that reactions of 1 proceed by
initial heterolytic fragmentation in all solvents. Figure 3
The ion pair model in Scheme 2 is consistent with the
observed yields of ion pairs. The maximum percentage yields
of radical cation 3 expected on the basis of measured rate
constants were obtained from kinetic simulations, and these
values are listed in the “predicted” column in Table 2. The
Table 2. Predicted and Observed Yields of Radical Cation 3a
% TFEb
ET(30)
predicted
found
partitioningc
10
5
2.5
1
OFPd
5%, PhMee
56.2
55.6
55.0
53.8
53.2
51.3
>80
75
68
59
44
40
16.5
13
2.5
2
4.5
<1:1.3
1:4
1:4
1:20
1:20
1:10
48
Figure 3. Rate constants for reactions of radical 1 at 20 °C from
ref 6 and the present work (red circles labeled with the solvent)
and from Table 1 (red double circles), rate constants for SSIP
desolvation (blue), and rate constants for ion pair escape (green).
The regression line for the rate constants for reactions of 1 shows
the 99% confidence interval in gray.
a Maximum percentage yield of 3 predicted from rate constants and
observed experimentally. b Percentage of TFE by volume in TFT unless
noted. c Partitioning ratio for the CIP between formation of the SSIP and
collapse to 2. d 2,2,3,3,4,4,5,5-Octafluoropentan-1-ol (0.2 M) in TFT. e TFE
(5% by volume) in toluene.
“found” column in Table 2 lists the observed percentage
yields of 3 in the ion pairs. The difference between these
two values is the percentage of 3 in the CIP that collapsed
to rearranged product 2 without solvating to the SSIP. From
these values, one can calculate the partitioning of the CIP
between collapse and solvation, and these calculated parti-
tioning ratios are given in Table 2. The CIP recombination
reaction was faster than solvation in all cases and became
increasingly efficient as the solvent polarity decreased. A
limit for the CIP recombination reaction in the three lower
polarity solvents in Table 2 is k′recom > 1 × 109 s-1; this
lower limit results from the requirement that CIP collapse
contains plots of rate constants as a function of solvent
polarity as determined by ET(30) values.9 The combination
of fragmentation rate constants in high-polarity media and
rearrangement rate constants in low- and medium-polarity
solvents displays an excellent correlation with ET(30),
indicating a common mechanism in all solvents. Similar
conclusions were reached for other â-phosphate radical
reactions, where rearrangement products and/or radical
cations were observed as a function of solvent polarity.11
A
common mechanism is also suggested by the entropic factor
found for reaction of 1. Specifically, the log A ) 11 value
(10) Zhou, J. W.; Findley, B. R.; Teslja, A.; Braun, C. L.; Sutin, N. J.
Phys. Chem. A 2000, 104, 11512-11521. Devadoss, C.; Fessenden, R. W.
J. Phys. Chem. 1990, 94, 4540-4549. Vauthey, E.; Parker, A. W.; Nohova,
B.; Phillips, D. J. Am. Chem. Soc. 1994, 116, 9182-9186.
(11) (a) Whitted, P. O.; Horner, J. H.; Newcomb, M.; Huang, X. H.;
Crich, D. Org. Lett. 1999, 1, 153-156. (b) Newcomb, M.; Horner, J. H.;
Whitted, P. O.; Crich, D.; Huang, X. H.; Yao, Q. W.; Zipse, H. J. Am.
Chem. Soc. 1999, 121, 10685-10694.
Org. Lett., Vol. 5, No. 26, 2003
5057

for ion pair formation in 5% TFE in toluene is the same as
the log A values found for rearrangement of 1 to 2 in low-
polarity (THF) and medium-polarity (CH3CN) solvents.6
The rate constants for collapse of the ion pair, although
limited in number, also appear to correlate with ET(30) values
(Figure 3). The measured values are the rate constants for
the slow step in conversion of the SSIP to product radical 2,
which, as noted above, is desolvation of the SSIP to give
the CIP. For acetonitrile, the extrapolated SSIP desolvation
rate constant would be kdesol ≈ 1 × 109 s-1, which is 2-5
times faster than diffusive escape rate constants for ion pairs
in this solvent.12 Thus, one would predict that, even if the
SSIP were formed in CH3CN, free ions would not be
observed in this solvent. Consistent with this prediction,
radical 1 rearranged to radical 2 in acetonitrile without
formation of a detectable amount of radical cation 3.6
The CIP partitioning ratios in Table 2 indicate that the
rate constants for collapse of the CIP will be much larger
than the rate constants for solvation to give the SSIP in low-
polarity solvents. If the CIP collapse rate constants exceed
rotational correlation times (ca. 1011-1012 s-1), then stereo-
chemical information would be preserved even when the
isolated radical cation is achiral; that is, a stereochemical
memory effect is predicted for the CIP in low-polarity
medium.13 Such behavior was found in studies of a system
related to radical 1 reported by Crich and co-workers. That
group found no stereochemical scrambling in benzene for
migration of a phosphate group that converted a homoallylic
radical to an allylic radical, whereas some stereochemical
scrambling was found when the reaction was conducted in
t-BuOH, a medium-polarity solvent.14
stereochemical memory effects mimicking concerted rear-
rangement reactions can be observed.
This work provides kinetic information for the very fast
ion pair phenomena involved in a radical heterolysis reaction
and establishes an ion pair model with some predictive
aspects. One should expect reactivity patterns similar to those
found for radical 1 in other radicals containing â-leaving
groups, with the rate constants for various processes offset
according to the stability of the incipient radical cation and
anion. Thus, the p-methoxyphenyl analogue of radical 1 gave
some diffusively free radical cation in medium-polarity
solvent, apparently due to smaller CIP collapse and SSIP
desolvation rate constants relative to those of 1, reflecting
the increased stability of the radical cation.11a In a similar
manner, the mesylate analogue of radical 1 gives high yields
of diffusively free radical cation 3 relative to those obtained
from 1 due to the increased stability of the anionic group.15
Reduced stability of either the radical cation or the anion is
expected to have the opposite effect, with slower overall
heterolysis reactions, less efficient production of diffusively
free radical cations, and an onset of stereochemical memory
effects in medium-polarity solvents. Stereoselective reactions
have provided the primary evidence for concerted migrations
of â-ester radicals,2 but they are readily rationalized in the
context of heterolysis reactions proceeding through short-
lived CIPs.16
Acknowledgment. This work was supported by a grant
from the National Science Foundation (CHE- 0235293).
Supporting Information Available: Synthetic details and
NMR spectra for the radical precursor to the mesylate
analogue of 1, description of LFP studies, and a plot of
measured ET(30) values for TFE mixtures in TFT. This
material is available free of charge via the Internet at
http://pubs.acs.org.
The combined results indicate that isomerization of radical
1 to radical 2 is likely to proceed by heterolysis reactions in
all solvents. Ion pairs and/or free ions are observed in high-
polarity solvents but not in solvents of medium polarity,
where the rate of SSIP desolvation exceeds the rate of escape
from the ion pair. In low-polarity solvents, the rate of CIP
collapse will exceed that of solvation to the SSIP, and
OL036052L
(15) Bagnol, L.; Horner, J. H.; Newcomb, M. Unpublished results.
(16) Unusual solvent character of the mixtures of TFE in aromatic
solvents found in this work also deserves note. The mixtures act as highly
polar solvents for ion-forming reactions but as low-polarity solvents for
ion pair escape processes. This behavior permitted direct detection of ion
pairs with exceptionally long lifetimes, a property that can be exploited in
studies of many ion pair reactions. We note that 5% of the polar solvent
TFE in toluene and in TFT appeared to result in a constant offset of the
polarity of the parent solvents (see Figure 3), suggesting that mixtures of
TFE in solvents less polar than toluene might increase the operating polarity
range of these exotic mixtures and provide media in which ion pairs have
even longer lifetimes than those found here.
(12) Gould, I. R.; Farid, S. Acc. Chem. Res. 1996, 29, 522-528. Taxil,
E.; Bagnol, L.; Horner, J. H.; Newcomb, M. Org. Lett. 2003, 5, 827-830.
Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc. 1990, 112,
4290-4301. Horner, J. H.; Taxil, E.; Newcomb, M. J. Am. Chem. Soc.
2002, 124, 5402-5410.
(13) Crich, D.; Ranganathan, K. J. Am. Chem. Soc. 2002, 124, 12422-
12423.
(14) Crich, D.; Yao, Q. W.; Filzen, G. F. J. Am. Chem. Soc. 1995, 117,
11455-11470. Crich, D.; Escalante, J.; Jiao, X. Y. J. Chem. Soc., Perkin
Trans. 2 1997, 627-630.
5058
Org. Lett., Vol. 5, No. 26, 2003