Heterolytic cleavage of a β-phosphatoxyalkyl radical resulting in phosphate migration or radical cation formation as a function of solvent polarity

Article abstract of DOI:10.1021/ol990054h

(Equation Presented) The 2-(diethylphosphatoxy)-2-(p-methoxyphenyl)-1,1-dimethylethyl radical (1) reacted to give the benzylic radical product from phosphate migration or a radical cation (or a mixture of the two) as a function of solvent. Smooth acceleration in rates of reactions of 1 in solvents of increasing polarity and consistent entropies of activation indicate that radical 1 reacts by common mechanism irrespective of the final products formed, specifically by initial heterolysis to a radical cation-phosphate anion pair.

Full text of DOI:10.1021/ol990054h

ORGANIC
LETTERS
1999
Vol. 1, No. 1
153-156
Heterolytic Cleavage of a
â-Phosphatoxyalkyl Radical Resulting in
Phosphate Migration or Radical Cation
Formation as a Function of Solvent
Polarity
Patrick O. Whitted,1a John H. Horner,1a Martin Newcomb,*,1a Xianhai Huang,1b and
,1b
David Crich*
Department of Chemistry, Wayne State UniVersity, Detroit, Michigan 48202, and
Department of Chemistry, UniVersity of Illinois at Chicago, 845 West Taylor Street,
Chicago, Illinois 60607-7061
men@chem.wayne.edu
Received April 26, 1999
ABSTRACT
The 2-(diethylphosphatoxy)-2-(p-methoxyphenyl)-1,1-dimethylethyl radical (1) reacted to give the benzylic radical product from phosphate migration
or a radical cation (or a mixture of the two) as a function of solvent. Smooth acceleration in rates of reactions of 1 in solvents of increasing
polarity and consistent entropies of activation indicate that radical 1 reacts by common mechanism irrespective of the final products formed,
specifically by initial heterolysis to a radical cation−phosphate anion pair.
The reactions of â-ester radicals in general are of interest
due to their mechanistic diversity.2 The â-phosphatoxyalkyl
radicals, a subset of â-ester radicals, are of further interest
because such radicals are intermediates in the cleavage of
DNA effected by anticancer agents such as bleomycin.3
â-Phosphatoxyalkyl radicals can react by phosphate migra-
tion, phosphoric acid elimination or nucleophilic substitution
with loss of phosphate from the adjacent position, a radical
equivalent of an SN2 or SN2′ reaction depending upon the
position of nucleophile attachment. The consensus views of
the mechanisms of these reactions are that two distinct
reaction manifolds are involved; migrations are concerted
processes, whereas elimination and substitution reactions
proceed through radical cations formed by initial heterolysis.2
Our groups have reported kinetic4,5 and labeling6,7 studies
of â-phosphatoxyalkyl radical reactions that, we believed,
were most consistent with concerted reactions. More recently,
however, we reported detailed studies of a competing
migration and elimination reaction of a â-phosphatoxyalkyl
radical that are not consistent with our original conclusions;
specifically, the results indicated that both products arose
from a common intermediate, an ion pair produced by
heterolysis, even in low polarity solvents.8 Herein, we report
(4) 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.
(5) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X.; Newcomb, M.;
Whitted, P. O. Tetrahedron 1999, 55, 3317-3326.
(6) Crich, D.; Yao, Q. W. J. Am. Chem. Soc. 1993, 115, 1165-1166.
(7) Crich, D.; Yao, Q. W.; Filzen, G. F. J. Am. Chem. Soc. 1995, 117,
11455-11470.
(1) (a) Wayne State University. (b) University of Illinois at Chicago.
(2) Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. W. Chem. ReV.
1997, 97, 3273-3312.
(8) Newcomb, M.; Horner, J. H.; Whitted, P. O.; Crich, D.; Huang, X.;
Yao, Q.; Zipse, H. Submitted for publication.
(3) Pogozelski, W. K.; Tullius, T. D. Chem. ReV. 1998, 98, 1089-1107.
10.1021/ol990054h CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/24/1999

an example of a â-phosphatoxyalkyl radical that reacts by
phosphate migration or fragmentation to a diffusively free
radical cation in a solvent-dependent fashion. The results
provide further evidence of a common pathway involving
initial heterolytic cleavage for seemingly different reactions.
Radical 1 contains the elements necessary for nanosecond
resolution kinetic studies in a variety of solvents employing
laser flash photolysis (LFP) methods. Laser irradiation (Nd:
YAG, 355 nm) of the PTOC ester precursor 25 results in
efficient cleavage of the N-O bond, giving the pyridine-2-
thiyl radical and an acyloxyl radical that rapidly decarbox-
ylates to give the desired radical 1 (Scheme 1). The pyridine-
Scheme 1
Figure 1. (A) Authentic spectra of benzyl radical 3 (open circles)
and radical cation 4 (filled circles); the spectra are not on the same
scale, see text for description of the methods used. (B-D) Time-
resolved spectra from reactions of radical 1 at 22 °C: (B) growth
in THF between 0.5 and 25.5 µs; (C) growth in acetonitrile between
0.3 and 2.5 µs; and (D) decay from products formed in water/
acetonitrile (10:90, v:v). The tick marks on the y axes of spectra
B-D are spaced at 0.01 AU.
2-thiyl radical has a long wavelength absorbance with λmax
at 490 nm,9 and UV spectra observed immediately after
photolysis of PTOC precursors contain this absorbance and
a bleaching centered at 360 nm from destruction of the
precursor. As shown in Scheme 1, diethyl phosphate migra-
tion in 1 would give the benzylic radical 3, whereas
heterolytic fragmentation would give radical cation 4.
Benzylic radical 3 was expected to have an absorbance at
about 320 nm.10 An authentic spectrum of radical 3 (Figure
1A) was obtained in the following manner. The PTOC ester
precursor11 5 was photolyzed in He-sparged acetonitrile
Radical 3 was formed “instantly” in both cases. The benzylic
radical was persistent over 4.5 µs in the He-sparged solution,
but decayed in the oxygen-containing solution over this
period. Subtraction of the nonsparged spectrum from the He-
sparged spectrum gave the residual spectrum that contains
only signals from 3.12 In passing, we note that the observation
of radical 3 in acetonitrile over 4.5 µs without detection of
radical cation 4 demonstrates that 3 is not a kinetically
competent intermediate for formation of 4 in this solvent
(see below).
An authentic spectrum of radical cation 4 was obtained
by photoejection from 1-(p-methoxyphenyl)-2-methylpropene
in acetonitrile upon 266 nm irradiation as described previ-
ously by Johnston.13 The UV spectrum of 4 is also shown
in Figure 1A; the absorbance centered at 390 nm is at slightly
longer wavelength than the absorbances reported for the
radical cations from p-methoxystyrene and anethole as
expected due to the additional methyl group(s) in 4.13,14 One
should note that the superimposed spectra of 3 and 4 in
solution and in acetonitrile solution containing oxygen, and
the UV spectra were monitored over several microseconds.
(9) Alam, M. M.; Watanabe, A.; Ito, O. J. Org. Chem. 1995, 60, 3440-
3444.
(12) This procedure was used previously to obtain a UV spectrum of a
benzylic radical from a PTOC ester precursor and is described in more
detail in ref 8.
(13) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115, 6564-
6571.
(14) Radical cation 4 also displayed an expected (ref 13) broad
absorbance at longer wavelengths; spectra to 700 nm are in Supporting
Information.
(10) Chatgilialoglu, C. In Handbook of Organic Photochemistry; Scaiano,
J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. 2; pp 3-11.
(11) The acid precursor to PTOC ester 5 was prepared by a reaction
sequence similar to that employed previously for production of the nor-
methoxy analogue (see ref 8). The synthetic intermediates and the carboxylic
acid had appropriate NMR spectra and elemental analyses, and PTOC ester
5 had appropriate NMR spectra.
154
Org. Lett., Vol. 1, No. 1, 1999

Figure 1A are not on the same scale; the molar absorption
coefficient of radical cation 4 is ∼1 order of magnitude
greater than that of 3.
of a solution of 2 in acetonitrile and one of water/acetonitrile
(or water only) via syringe pumps immediately before the
LFP experiments were performed, and the kinetics in aqueous
acetonitrile solutions were obtained only at ambient tem-
peratures. Table 1 lists rate constants for reactions of 1 in
various solvents at ambient temperature.
Time-resolved UV growth spectra from reactions of radical
1 were obtained in THF and acetonitrile. The spectra were
produced by subtracting a short time spectrum from those
obtained at later times which removes the features present
immediately after photolysis. Positive signals are from
species forming with time, and negative signals are from
decaying species. The spectrum in THF (Figure 1B)5 shows
the growth of benzylic radical 3 and decay of the pyridine-
2-thiyl radical; no signal from radical cation 4 is apparent.
In acetonitrile solution, however, a strong signal from radical
cation 4 grew in (Figure 1C), and comparison of the 300-
320-nm regions in parts A and C of Figure 1 indicates that
benzylic radical 3 also was formed in acetonitrile. We
calculate an ∼35% yield of radical cation 4 from reaction
of 1 in CH3CN from a comparison of the ultimate intensity
of the 390-nm signal from the radical cation against the
instantaneous intensity of the 490-nm signal from pyridine-
2-thiyl and the known and measured extinction coefficients
of the two radicals.15
Table 1. Kinetics of Reactions of Radical 1 at (22 ( 2) °C
solvent
% 4a
k (s-1
)
THF
butanoneb
CH3CN
5% waterc
10% waterc
50% waterc
0
25
35
95
100d
100d
8 × 104
1.2 × 106
1.2 × 106
7 × 106
1.7 × 107
8 × 107
a Percentage of product 4 estimated from the ratio of the ultimate intensity
of the 390 nm signal from 4 to that of the pyridine-2-thiyl radical at 490
nm using the extinction coefficients in ref 15. b Despite drying, this solvent
might have contained some water as deduced from the apparent hydrolysis
of the PTOC ester upon standing. c Volume percent of water in water/
acetonitrile. d No signal was observed at 320 nm from radical 3.
Radical 1 reacted rapidly in aqueous acetonitrile solutions,
precluding useful growth spectra, but Figure 1D shows the
time-resolved decay spectrum of radical cation 4 produced
from reaction of 1 in a 10% water solution. Radical cation
4 appears to be the only product formed in the aqueous
acetonitrile reactions on the basis of the small 320-nm signal
in Figure 1D; in addition, the ratio of the intensities at 390
and 490 nm in aqueous acetonitrile solutions was about 4
times as great as that in acetonitrile, in qualitative agreement
with a high yield of 4 although the extinction coefficients
of the two radicals in these solvent mixtures are not known.
The kinetics of reactions of 1 were studied. PTOC ester 2
is stable in THF and in acetonitrile, and variable-temperature
studies were possible; the Arrhenius functions for reactions
in these solvents are given in eqs 1 and 2
The rate constants for reactions of 1 increase smoothly as
the solvent polarity increases from THF to aqueous aceto-
nitrile solutions, similar to the kinetic behavior observed in
other â-ester radical reactions. For example, the 15-fold
acceleration in reaction of 1 upon proceeding from THF to
acetonitrile is comparable to the acceleration in reactions of
radicals 6-8 in these solvents.4,5,8 The obvious difference
between radicals 6-8 and 1 is that the reactions of 6-8 do
not give different products in the different solvents; 6 and 7
react by diphenyl phosphate and trifluoroacetate migrations,
respectively, and radical 8 reacts in both solvents by a
combination of diphenyl phosphate migration and diphenyl-
phosphoric acid elimination reactions.4,5,8 The log A terms
in the Arrhenius functions (the entropic terms) for reactions
of radical 1 in THF and acetonitrile are similar to one another
as well as to those for reactions of radicals 6-8 in the same
solvents.4,5,8
log(k/s-1) ) (12.1 ( 0.8) - (9.7 ( 1.2)/2.3RT
log(k/s-1) ) (10.6 ( 0.7) - (6.1 ( 0.9)/2.3RT
(1)
(2)
for THF and CH3CN, respectively, where errors are at 2σ
and the Ea values are in kilocalories/mole. For reactions in
CH3CN, the rate constants for signal formation in the 312-
320- and 390-400-nm regions were comparable, suggesting
that relatively slow secondary reactions were not involved
in production of either product.16 Because PTOC esters
hydrolyze rapidly in aqueous solutions, studies in water/
acetonitrile mixtures were more difficult, requiring mixing
Both the kinetic solvent effects and the consistent entropies
of activation indicate that the fundamental mechanism of the
reaction of radical 1 does not change with solvent, even if
the products ultimately formed do. This is dramatically
illustrated by plotting the rate constants for reactions of 1
(15) The molar extinction coefficient for the 2-pyridinethiyl radical in
acetonitrile is 3200 M-1 cm-1 (ref 9). That for 4 is approximately 30 000
M-1 cm-1 as determined from oxidation of the styrene with the chloranil
triplet and comparison of the signal intensities for 4 and the chloranil radical
anion.
(16) We note, however, that this behavior only excludes a relatively slow
secondary reaction as the source of one product. It does not exclude the
possibility that two distinct reactions give rise to products 3 and 4 because
the kinetics are only reporting the total rate of loss of reactant radical 1,
irrespective of the number of reaction channels available.
(17) Reichardt, C. Chem. ReV. 1994, 94, 2319-2358. Skwierczynski,
R. D.; Connors, K. A. J. Chem. Soc., Perkin Trans. 2 1994, 467-472.
Org. Lett., Vol. 1, No. 1, 1999
155

against the ET(30) solvent polarity scale17 (Figure 2).
Although the products change from exclusively 3 in THF to
exclusively 4 in 10% water in acetonitrile, the linear behavior
of the plot indicates that the mechanism is unaltered to the
point of the rate-determining step.
for collapse to products or (2) further solvation of the SSIP
to give free ions is competitive with return of the SSIP to
the CIP. Diffusively free radical cation was not formed in
the reactions of radical 8 in acetonitrile, but the methoxy-
substituted styrene radical cation 4 is appreciably more stable
than the radical cation from â,â-dimethylstyrene as judged
by comparison of the oxidation potentials of styrenes18 and
the fact that radical 1 reacts in CH3CN at ambient temper-
ature about 20 times faster than its nor-methoxy analogue,
the 2-(diethylphosphatoxy)-2-phenyl-1,1-dimethylethyl radi-
cal.5 Thus, either possibility (or both) appears reasonable,
and a distinction between the two might require picosecond
dynamic studies. The kinetic behavior of radical 8 led us to
conclude, however, that the rate constant for evolution of
the SSIP to free ions (kC in Scheme 2) was likely to be highly
sensitive to solvent polarity.8
The solvent-dependent behavior of radical 1 provides a
striking confirmation that aryl-substituted â-phosphatoxyalkyl
radicals (and, by analogy, other â-ester radicals2) can react
by initial heterolysis even when the final products are those
of ester group migration. On the basis of the similarities of
the kinetic behaviors and log A terms for several â-aryl-â-
ester radicals in solvents of widely different polarities,2,4,5,8
we believe the mechanistic picture in Scheme 2 is likely to
apply for all cases where the incipient radical cation is at
least as stable as a styrene radical cation. This includes simple
R-oxy-â-phosphatoxy radicals such as those formed by
hydrogen-atom abstraction from the sugars in DNA.19 The
mechanisms of reactions of â-ester radicals containing no
cation-stabilizing groups, for which concerted pathways have
been found computationally,20 remain an open question.
Figure 2. Rate constants for reactions of radical 1 plotted against
the ET(30) solvent polarity scale. The butanone value was not used
in the regression line; see note in Table 1.
Our recent studies of radical 8 indicated that the reaction
proceeded as shown in Scheme 2.8 Initial heterolysis gives
a contact ion pair (CIP) of the radical cation and phosphate
anion. Rapid collapse of the CIP competes with (possibly
diffusion-controlled) evolution to a solvent-separated ion pair
(SSIP), and the SSIP partitions back to the CIP or further
separates to give diffusively free ions. In the case of radical
8, a short-lived diffusively free radical cation was detected
as a minor product in aqueous acetonitrile solutions but not
in acetonitrile.
Acknowledgment. This work was supported by grants
from the National Institutes of Health (GM56511 to M.N.
and CA60500 to D.C.) and the National Science Foundation
(CHE-9614968 to M.N. and CHE-9625256 to D.C.). We
thank Dr. Seung-Yong Choi for experimental assistance.
Scheme 2
Supporting Information Available: Spectra of radical
cation 4 from 300 to 700 nm. This material is available free
of charge via the Internet at http://pubs.acs.org.
OL990054H
(18) The oxidation potential of p-methoxystyrene is ∼0.4 V lower than
that of styrene; see: Akbulut, U.; Toppare, L.; Tu¨rker, L. Makromol. Chem.
1983, 184, 1661-1667.
Radical 1 appears to react by the same scheme. In THF,
collapse to the migration product is the major secondary
process. In acetonitrile, either (1) the rate constant for
solvation of the CIP to give the SSIP is comparable to that
(19) For example, ethyl vinyl ether and styrene have the same oxidation
potentials in acetontrile. See: Katz, M.; Riemenschneider, P.; Wendt, H.
Electrochim. Acta 1972, 17, 1595-1607.
(20) Zipse, H. J. Am. Chem. Soc. 1997, 119, 2889-2893.
156
Org. Lett., Vol. 1, No. 1, 1999