Product studies and laser flash photolysis on alkyl radicals containing two different β-leaving groups are consonant with the formation of an olefin cation radical

Article abstract of DOI:10.1021/ja0042938

1-Bromo-2-methoxy-1-phenylpropan-2-yl (3) and 2-methoxy-1-phenyl-1-diphenylphosphatopropan-2-yl (4) were generated under continual photolysis from the respective PTOC precursors in a mixture of acetonitrile and methanol. The radicals undergo heterolytic fragmentation of the substituent in the β-position to generate the olefin cation radical (5). Z-2-Methoxy-1-phenylpropene (15) is the major product formed in the presence of 1,4-cyclohexadiene, and is believed to result from hydrogen atom transfer to the oxygen of the olefin cation radical, followed by deprotonation. Laser flash photolysis experiments indicate that reaction between 5 and 1,4-cyclohexadiene occurs with a rate constant of ～6 × 10⁵ M^-1 s^-1. 2,2-Dimethoxy-1-phenylpropane (18) is observed as a minor product. Laser flash photolysis experiments place an upper limit on methanol trapping of 5 at k <1 × 10³ M^-1 s^-1 and do not provide any evidence for the formation of reactive intermediates other than 5. The use of two PTOC precursors containing different leaving groups to generate a common olefin cation radical enables one to utilize product analysis to probe for the intermediacy of other reactive intermediates. The ratio of 15:18 is dependent upon hydrogen atom donor concentration, but is independent of the PTOC precursor. These observations are consistent with the proposal that both products result from trapping of 5 that is formed via heterolysis of 3 and 4.

Full text of DOI:10.1021/ja0042938

VOLUME 123, NUMBER 16
APRIL 25, 2001
© Copyright 2001 by the
American Chemical Society
Product Studies and Laser Flash Photolysis on Alkyl Radicals
Containing Two Different â-Leaving Groups Are Consonant with the
Formation of an Olefin Cation Radical
Brian C. Bales,† John H. Horner,‡ Xianhai Huang,§ Martin Newcomb,*,‡
David Crich,*,§ and Marc M. Greenberg*,†
Contribution from the Departments of Chemistry, Colorado State UniVersity,
Fort Collins, Colorado 80523, Wayne State UniVersity, Detroit, Michigan 48202, and
UniVersity of Illinois at Chicago, 845 West Taylor St., Chicago, Illinois 60607
ReceiVed December 19, 2000. ReVised Manuscript ReceiVed February 20, 2001
Abstract: 1-Bromo-2-methoxy-1-phenylpropan-2-yl (3) and 2-methoxy-1-phenyl-1-diphenylphosphatopropan-
2-yl (4) were generated under continual photolysis from the respective PTOC precursors in a mixture of
acetonitrile and methanol. The radicals undergo heterolytic fragmentation of the substituent in the â-position
to generate the olefin cation radical (5). Z-2-Methoxy-1-phenylpropene (15) is the major product formed in
the presence of 1,4-cyclohexadiene, and is believed to result from hydrogen atom transfer to the oxygen of the
olefin cation radical, followed by deprotonation. Laser flash photolysis experiments indicate that reaction between
5 and 1,4-cyclohexadiene occurs with a rate constant of ∼6 × 105 M-1 s-1. 2,2-Dimethoxy-1-phenylpropane
(18) is observed as a minor product. Laser flash photolysis experiments place an upper limit on methanol
trapping of 5 at k <1 × 103 M-1 s-1 and do not provide any evidence for the formation of reactive intermediates
other than 5. The use of two PTOC precursors containing different leaving groups to generate a common
olefin cation radical enables one to utilize product analysis to probe for the intermediacy of other reactive
intermediates. The ratio of 15:18 is dependent upon hydrogen atom donor concentration, but is independent of
the PTOC precursor. These observations are consistent with the proposal that both products result from trapping
of 5 that is formed via heterolysis of 3 and 4.
Olefin cation radicals are a family of reactive intermediates
that are involved in a broad range of chemical reactions. The
scope of reactions extends from synthetically useful transforma-
tions (e.g. cycloadditions) to DNA damage.1,2 The utility of
olefin cation radicals in organic synthesis stems from their high
rates of reaction (bimolecular rate constants can be as large as
109 M-1 s-1) with nucleophiles, alkenes, and dienes.3,4 For
synthetic purposes, these reactive intermediates are typically
generated via one-electron oxidation of alkenes. The range of
ion radicals that can be generated by this approach is limited
by the oxidation potential of the alkene, as well as by the fact
that the precursors themselves are reaction partners for the olefin
cation radical. A method for generating olefin cation radicals
containing varying reduction potentials from substrates other
* Address correspondence to these authors. E-mail: dcrich@uic.edu;
men@chem.wayne.edu; mgreenbe@lamar.colostate.edu.
† Colorado State University.
‡ Wayne State University.
§ University of Illinois at Chicago.
(1) (a) Bauld, N. L. Radicals, Ion Radicals, and Triplets; Wiley-VCH:
New York, 1987. (b) Schmittel, M.; Burghart, A. Angew. Chem., Int. Ed.
Engl. 1997, 36, 2550.
(2) (a) Greenberg, M. M. Chem. Res. Toxicol. 1998, 11, 1235. (b)
Meggers, E.; Dussy, A.; Scha¨fer, T.; Giese, B. Chem. Eur. J. 2000, 6, 485.
(3) Lew, C. S. Q.; Brisson, J. R.; Johnston, L. J. J. Org. Chem. 1997,
62, 4047.
(4) (a) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115, 6564.
(b) Schepp, N. P.; Johnston, L. J. J. Am. Chem. Soc. 1996, 118, 2872.
10.1021/ja0042938 CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/31/2001

3624 J. Am. Chem. Soc., Vol. 123, No. 16, 2001
Bales et al.
Scheme 1
diffusible species are lacking. We report herein evidence
obtained via product analysis and LFP that supports the
generation of a diffusively free olefin cation radical as the major
reactive intermediate that is responsible for product formation
in the case of radical cation 5. The use of two radical precursors
containing different leaving groups in the â-position affirms that
a common intermediate, the olefin cation radical, is formed from
these species. Furthermore, the use of radicals containing
different â-leaving groups enables us to probe for the involve-
ment of reactive intermediates other than a diffusively free cation
radical using product studies. Although not all of the products
formed can be explained via trapping a diffusively free cation
radical by stable molecules, there is no positive evidence for
the involvement of other reactive intermediates in product
formation.
than alkenes would be highly desirable. Nature (in this instance
in the form of the effects of ionizing radiation on DNA) has
provided the impetus for exploring olefin cation radical genera-
tion via heterolysis from the â-position of an alkyl radical
(Scheme 1). The feasibility of this reaction has captured the
attention of computational chemists and experimentalists. We
wish to report herein an investigation, combining product studies
and laser flash photolysis (LFP) experiments, of the reactions
of two independent precursors to radical cation 5.
Results
At the time that this project was initiated, laser flash
photolysis and product studies had not yet appeared.10-15
Consequently, not knowing the minimum structural requirements
for olefin cation radical formation, a phenyl substituent was
incorporated at the â-position to maximize the driving force
for â-elimination from the initially formed alkyl radical (eq 1).
In 1975, γ-irradiation of DNA was proposed to give rise to
an olefin cation radical via phosphate monoester elimination
following C4′-hydrogen atom abstraction.5 Pulse radiolysis
experiments on model compounds, and more recent studies using
chemically modified oligonucleotides, as well as other models,
provided compelling evidence for olefin cation radical genera-
tion and/or a functionally equivalent nucleophilic substitution
reaction (SRN2′, Scheme 1).6-8 The latter process is analogous
to an SN2′ reaction, and has been identified in computational
experiments as a feasible mechanism for an alkyl radical
containing a leaving group in the â-position.9,10 These studies
alerted our groups to the possibility that alkyl radicals, which
are readily generated, containing leaving groups in the â-position
could serve as synthons for olefin cation radicals. Laser flash
photolysis studies have provided a wealth of spectroscopic
evidence for the formation of olefin cation radicals via phosphate
diester or halide elimination from the â-position of appropriately
substituted alkyl radicals.10-14 The observation of diffusively
free cation radicals in these processes is highly dependent upon
the polarity of the solvent.11-14 Evidence for the homolytic
substitution mechanism (SRN2′) is less ample. Distinguishing
mechanisms by product analysis is difficult, because trapping
of the olefin cation radical and SRN2′ substitution can yield
identical products.11,15 Furthermore, studies that reveal the ratio
of products resulting from caged processes relative to freely
In addition, an alkoxy group at a tertiary radical center was
included to model the C4′-position of the furanose ring in DNA
where the greatest evidence for â-elimination had been obtained.
It should be noted that it was not clear whether the R-methoxy
group in the alkyl radical would provide a thermodynamic
driving force for â-elimination. Comparing the effects of ether
substituents on the oxidation potentials of styrenes and other
alkenes, we estimated that the stabilization of the methoxy group
on the olefin cation radical (5) compared to styrene could be as
(5) Dizdaroglu, M.; von Sonntag, C.; Schulte-Frohlinde, D. J. Am. Chem.
Soc. 1975, 97, 2277.
(6) (a) Davies, M. J.; Gilbert, B. C. AdVances in Detailed Reaction
Mechanisms; JAI Press: Stamford, CT, 1991; Vol. 1, p 35. (b) Behrens,
G.; Koltzenburg, G.; Schulte-Frohlinde, D. Z. Naturforsch. 1982, 37c, 1205.
(7) (a) Strittmatter, H.; Dussy, A.; Schwitter, U.; Giese, B. Angew. Chem.,
Int. Ed. Engl. 1999, 38, 135. (b) Giese, B.; Dussy, A.; Meggers, E.; Petretta,
M.; Schwitter, U. J. Am. Chem. Soc. 1997, 119, 11130.
(8) (a) Peukert, S.; Giese, B. Tetrahedron Lett. 1996, 37, 4365. (b) Giese,
B.; Beyrich-Graf, X.; Burger, J.; Kesselheim, C.; Senn, M.; Scha¨fer, T.
Angew. Chem., Int. Ed. Engl. 1993, 32, 1742.
much as 10.3 kcal/mol.4b,16 Furthermore, the presence of the
methoxy group may facilitate migration of the phosphate group,
which could complicate kinetic and product analysis.17 Efforts
to reduce the possibility of migration resulted in examination
of a system containing a bromide (3) leaving group at the
â-position of the original alkyl radical, despite the possibility
that â-scission might compete with heterolytic cleavage.14,18 The
â-bromo radical (3) could undergo formal rearrangement via
heterolysis and subsequent trapping of the olefin cation radical
(9) (a) Zipse, H. Acc. Chem. Res. 1999, 32, 571. (b) Zipse, H. J. Am.
Chem. Soc. 1994, 116, 10773.
(10) Newcomb, M.; Horner, J. H.; Whitted, P. O.; Crich, D.; Huang,
X.; Yao, Q.; Zipse, H. J. Am. Chem. Soc. 1999, 121, 10685.
(11) Crich, D.; Huang, X.; Newcomb, M. J. Org. Chem. 2000, 65, 523.
(12) (a) Newcomb, M.; Miranda, N.; Huang, X.; Crich, D. J. Am. Chem.
Soc. 2000, 122, 6128. (b) Whitted, P. O.; Horner, J. H.; Newcomb, M.;
Huang, X.; Crich, D. Org. Lett. 1999, 1, 153.
(13) (a) Choi, S.-Y.; Crich, D.; Horner, J. H.; Huang, X.; Newcomb,
M.; Whitted, P. O. Tetrahedron 1999, 55, 3317. (b) Choi, S.-Y.; Crich, D.;
Horner, J. H.; Huang, X.; Martinez, F. N.; Newcomb, M.; Wink, D. J.;
Yao, Q. J. Am. Chem. Soc. 1998, 120, 211.
(16) The reduction potential of 5 was estimated by considering the
oxidation potential of 15, both of which are unknown. The upper limit for
the oxidation potential of 15 is believed to be 1.33 V (versus SCE), and is
based upon comparisons to related styrenes whose irreversible oxidation
potentials have been measured. The E1/2OX(SCE) of â-methyl-4-methoxy-
styrene ) 1.33 V.4b The E1/2OX(SCE) of R-trimethylsilyloxystyrene ) 1.32
V. See: Fukuzumi, S.; Fujita, M.; Otera, J.; Fujita, Y. J. Am. Chem. Soc.
1992, 114, 10271.
(14) Cozens, F. L.; O’Neill, M.; Bogdanova, R.; Schepp, N. J. Am. Chem.
Soc. 1997, 119, 10652.
(15) (a) Crich, D.; Huang, X.; Newcomb, M Org. Lett. 1999, 1, 225. (b)
Crich, D.; Gastaldi, S. Tetrahedron Lett. 1998, 39, 9377.
(17) Beckwith, A. L. J.; Crich, D.; Duggan, P. J.; Yao, Q. Chem. ReV.
1997, 97, 3273.
(18) Wagner, P. J.; Sedon, J. H.; Lindstrom, M. J. J. Am. Chem. Soc.
1978, 100, 2579.

Formation of an Olefin Cation Radical
J. Am. Chem. Soc., Vol. 123, No. 16, 2001 3625
Scheme 3a
spectral tables.22 The vinyl proton in the E-isomer is expected
to resonate downfield from the Z isomer. Osmylation of a Z/E
mixture of â-methylstyrene followed by etherification provided
the vicinal ether (17) as a mixture of diastereomers.23 The
dimethyl ketal (18) was prepared from phenylacetone.24 2-Meth-
oxy-3-phenylpropene (19) was prepared via titanocene mediated
methylenation of the respective methyl ester.22b The products
derived from hydrogen atom trapping of the alkyl radicals (20,
21) produced from the respective PTOC esters (1, 2) were
prepared from 22, which was obtained by sequential alkylation
and reduction of 2-methoxyacetophenone. Finally, 2-methoxy-
1-phenylpropane (23) was prepared from the corresponding
alcohol by alkylation.
a Key: (a) LDA, PhCHO, THF. (b) (PhO)2P(O)Cl, DMAP, CH2Cl2.
(c) CAN, H2O:CH3CN (10:1). (d) PPh3, 2,2′-dithiopyridine-1,1′-dioxide,
CH2Cl2.
by the bromide within the solvent cage, but it is highly unlikely
that the â-phosphate group would cleave homolytically.4a,19
Finally, the generation of alkyl radicals with different â-sub-
stituents could provide a means for distinguishing between
reactive intermediates that are responsible for product formation
(Vide infra).
Synthesis of Barton (PTOC, pyridine-2-thione-N-oxycar-
bonyl) Esters and Anticipated Products. PTOCs (1, 2) were
chosen as radical precursors, because these molecules undergo
homolytic bond scission initiated by long wavelength irradiation
to generate carboxyl radicals which decarboxylate on the
submicrosecond time scale.20 The bromo-PTOC (1) was syn-
thesized from ethyl lactate (6, Scheme 2). The benzylic bromide
UV-Irradiation of Bromo-PTOC (1) and Phospho-PTOC
(2). Samples containing 1 or 2 were freeze-pump-thaw
degassed and irradiated in a Rayonet Photoreactor containing
λmax ) 350 nm lamps for 20 min. The solvent mixture was
kept constant (MeOH:CH3CN, 4:1 (by volume), ET(30) ) 56.21
kcal/mol, [MeOH] ) 19.75 M) as the type and concentration
of other additives were varied.25 Although varying the concen-
tration of the nucleophile (MeOH) would be a useful mechanistic
probe, complications resulting from the strong dependence of
formation of diffusively free olefin cation radicals on solvent
polarity precluded this.10-14 The PTOC esters were stable at
room temperature in the dark for the period of sample prepara-
tion and photolysis, but did undergo some decomposition (up
to 30%) upon prolonged exposure (24 h) to these conditions.
The dimethoxy-PTOC (24) was the exclusive decomposition
product observed from 1 and 2. Photochemically induced
decomposition of independently synthesized 24 under the
conditions employed for 1 and 2 yielded the vicinal ether (17)
as a mixture of diastereomers as the sole product attributable
to the formation of 25 (eq 2).
Scheme 2a
a Key: (a) NaH, CH3I, THF. (b) LDA, BnBr, THF. (c) KOH, EtOH.
(d) NBS, CCl4. (e) PPh3, 2,2′-dithiopyridine-1,1′-dioxide, CH2Cl2.
was introduced by using NBS, and the crude carboxylic acid
obtained was transformed into the PTOC (1) due to the
propensity of 10 to undergo loss of HBr and CO2 to yield vinyl
ethers (15, 16). This facile decarboxylation/elimination dictated
the manner in which the phospho-PTOC (2) was prepared. To
circumvent this problem, following phosphorylation of the aldol
product (12), the ester was oxidatively cleaved, and the crude
acid was carried on to 2 using Mitsunobu conditions (Scheme
3).13a,21
Irradiation of 1 and 2 in the presence of independently
synthesized potential products revealed that only the vicinal ether
(17) was stable to the reaction conditions. Phenylacetone was
the major product in the absence of amine added as an acid
scavenger. All subsequent irradiations were carried out in the
The vinyl ether was prepared as a 1:1 mixture of geometrical
isomers (15, 16) by treating 10 with Hu¨nig’s base. The Z- (15)
and E- (16) isomers were separable by column chromatography
and their identities were established with 1H NMR, by compar-
ing the chemical shifts to the data in the literature and by using
(22) (a) Kriz, G. S.; Lampman, G. M.; Pavia, D. L. Introduction to
Spectroscopy, 2nd ed.; Sunders College, New York, 1996; p 473. (b) Gaudry,
M.; Marquet, A. Bull. Soc. Chim. Fr. 1969, 4178.
(19) Steenken, S.; Goldbergerova, L. J. Am. Chem. Soc. 1998, 120, 3928.
(20) (a) Newcomb, M. Tetrahedron 1993, 49, 1151. (b) Barton, D. H.
R.; Crich, D.; Motherwell, W. B. Tetrahedron 1985, 41, 3901.
(21) Barton, D. H. R.; Samadi, M. Tetrahedron 1992, 48, 7083.
(23) Bassetti, M.; Floris, B. J. Chem. Soc., Perkin Trans. 2 1988, 227.
(24) Steenken, S.; McClelland, R. A. J. Am. Chem. Soc. 1989, 111, 4967.
(25) (a) Langhals, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 724. (b)
Reichardt, C. Chem. ReV. 1994, 94, 2319.

3626 J. Am. Chem. Soc., Vol. 123, No. 16, 2001
Bales et al.
Table 1. Product Distribution in the Presence of
1,4-Cyclohexadiene (CHD) and
Scheme 4
Scheme 5
3,3,6,6-Tetradeuterio-1,4-cyclohexadiene (d4-CHD) as a Function of
Solventa
% yield
PTOC
trap
solvent
MeOH
15
18
23
1
1
1
2
2
2
CHD
d4-CHD MeOH
d4-CHD d4-MeOH 11.3 ( 1.5 2.1 ( 0.2 2.9 ( 0.3
CHD
CHD
22.4 ( 1.2 3.7 ( 0.3 3.2 ( 0.2
10.0 ( 1.6 3.7 ( 0.5 2.5 ( 0.2
c
MeOH
MeOH
38.3 ( 2.2 4.4 ( 0.7
21.6 ( 1.6 4.4 ( 0.7
-
-
-
c
c
d4-CHD d4-MeOH 22.2 ( 1.9 2.3 ( 0.5
a [PTOC] ) 5 mM, [TBP] ) 10 mM, [CHD] ) [d4-CHD] ) 0.5 M.
b [MeOH] ) 19.75 M, [d4-MeOH] ) 19.73 M, CH3CN (20 vol %) is
a cosolvent. c Not observed.
presence of an acid scavenger that would not react with products
and intermediates resulting from photolysis. The requirements
of poor nucleophilicity and high oxidation potential to avoid
reaction with the olefin cation radical (5) led us to choose
hindered 2,6-di-tert-butylpyridine (TBP) as an acid scaven-
ger.16,26 Similarly, 1,4-cyclohexadiene (CHD) was chosen as a
potential hydrogen atom donor, because of concerns that thiols
would participate in electron transfer and/or nucleophilic
reactions with the olefin cation radical.27 Despite these precau-
tions, we found that the benzyl vinyl ether (19) underwent
significant decomposition (∼90%) during photolysis of the
radical precursors (1, 2).
Table 2. The Eeffect of Solvent Deuteration on Deuterium
Incorporation in Productsa
% 2H
PTOC
solventb
15
18
23
1
1
2
2
MeOH
d4-MeOH
MeOH
3.0 ( 0.6
2.2 ( 0.5
2.4 ( 0.4
2.0 ( 0.2
47.9 ( 0.8
63.1 ( 1.5
60.2 ( 0.4
76.9 ( 3.1
71.0 ( 0.7
82.1 ( 6.5
c
-
c
d4-MeOH
-
a [PTOC] ) 5 mM, [TBP] ) 10 mM, [d4-CHD] ) 0.5 M. b [MeOH]
) 19.75 M, [d4-MeOH] ) 19.73 M, CH3CN (20 vol %) is a cosolvent.
c Not observed.
Irradiation of 1 or 2 (5 mM) in the presence of TBP (10 mM)
and 1,4-cyclohexadiene (0.5 M) produced Z-2-methoxy-1-
phenylpropene (15) as the major product (Table 1). Vinyl ether
16 was not observed, and preferential formation of 15 is
consistent with molecular mechanics calculations that predict
∆∆Hf of 2.17 kcal/mol favoring the observed isomer.28 Smaller
amounts of dimethyl ketal (18) were also observed from both
PTOCs, but neither precursor yielded any vicinal ether (17).
The yields of 15 and 18 obtained from irradiation of the bromo-
PTOC (1) were consistently lower than those obtained from 2.
Control experiments in which samples of 1 were individually
spiked with 15 or 18 prior to photolysis revealed that no more
than 2% of 15 decomposed under the reaction conditions.29
These experiments also indicated that phenyl acetone was the
decomposition product of 15, but not the ketal (18). No evidence
for decomposition of 18 was obtained from these experiments.29
Photolysis of the bromo-PTOC also produced 2-methoxy-1-
phenylpropane (23) in ∼3% yield. Control experiments with
independently synthesized 21 suggest that 23 is produced from
the benzylic bromide under the photolysis conditions. The
product analogous to 21 expected from 2 (20) could not be
detected by GC, and was analyzed for by 1H NMR. Neither 20
nor 23 was observed upon irradiation of 2.30 Consistent with
the instability noted above, low, variable yields of terminal vinyl
ether (19) were also observed. It is difficult to quantitate how
much 19 is produced during the photolysis. On the basis of mass
balances, this vinyl ether could be formed in as great as 71%
and 56% yield from 1 and the phosphorylated PTOC (2),
respectively. However, we consider this unlikely, since 19 is
believed to be derived from the allylic radical (Scheme 5), and
this intermediate was not formed in high enough yield to be
detected by LFP (vide infra).
Photolysis of 1 or 2 (5 mM) in the presence of TBP (10 mM)
and d4-1,4-cyclohexadiene (d4-CHD, 0.5 M) showed an apparent
kinetic isotope effect of ∼2 for the formation of 15, while no
effect on the formation of 18 was observed (Table 1). Measure-
2
ment of the percent H incorporation in 15, 18, 19, and 23
showed incomplete isotopic incorporation when d4-CHD was
employed (Table 2).31 This led us to investigate the possibility
of hydrogen/deuterium transfer from methanol to one or more
of the intermediates formed upon photolysis of 1 and 2. The
isotopic content in these same products increased when the
PTOC esters were photolyzed in d4-MeOH (19.75 M), but were
unaffected by substituting deuterated acetonitrile for its protio
analogue. The effect of CHD concentration on the product yield
was also determined (Table 3). The proportional relationship
of product yields to CHD concentration indicates that CHD is
involved in formation of all of the observed products.
Laser flash photolysis (LFP) studies were conducted with
PTOC esters 1 and 2. As with other PTOC esters32-34 including
â-phosphatoxyalkyl radical precursors,10,12b,13 photolyses with
355 nm laser light cleaved the precursors readily. Photolyses
of either 1 or 2 in THF gave the time-resolved spectrum shown
Ox
(26) Trialkylamines were expected to have lower E1/2 than 15. For
example, E1/2Ox(SCE) of Et3N ) 0.78 V. See: (a) Liu, W.-Z.; Bordwell, F.
G. J. Org. Chem. 1996, 61, 4778. (b) Nelson, S. F.; Hintz, P J. J. Am.
Chem. Soc. 1972, 94, 7114. Pyridines are much less readily oxidized. The
E1/2Ox(SCE) of pyridine ) 2.12 V. See: CRC Handbook Series in Organic
Electrochemistry; Meites, L., Zuman, P., Eds.; CRC Press: Cleveland, OH,
1974; Vol. 1.
(27) The upper limit for E1/2Ox(SCE) of thiols ) 1.11 V. See: Asmus,
K.-D.; Bonifacic, M. in S-Centered Radicals; Alfassi, Z. B., Ed.; John Wiley
& Sons: New York, 1999. The E1/2Ox(SCE) of 1,4-cyclohexadiene ) 1.74
V. See: Shono, T.; Ikeda, A.; Hayashi, J.; Hakozaki, S. J. Am. Chem. Soc.
1975, 97, 4261.
(31) The rate constant for reaction of 1,4-cyclohexadiene with alkyl
radicals is <105 M-1 s-1. (Hawari, J. A.; Engel, P. S.; Griller, D. Int. J.
Chem. Kinet. 1985, 17, 1215.)
(32) Ha, C.; Horner, J. H.; Newcomb, M.; Varick, T. R.; Arnold, B. R.;
Lusztyk, J. J. Org. Chem. 1993, 58, 1194.
(33) Johnson, C. C.; Horner, J. H.; Tronche, C.; Newcomb, M. J. Am.
Chem. Soc. 1995, 117, 1684.
(28) Calculations were carried out using Spartan (version 5.0.3) at the
AM1 level of theory on an SGI Indigo.
(29) See the Supporting Information.
(30) The yield of 20 must be less than 5% based upon the detection
limit of the NMR experiment.
(34) Newcomb, M.; Horner, J. H.; Filipkowski, M. A.; Ha, C.; Park, S.
U. J. Am. Chem. Soc. 1995, 117, 3674.

Formation of an Olefin Cation Radical
J. Am. Chem. Soc., Vol. 123, No. 16, 2001 3627
Table 3. Product Distribution As a Function of 1,4-cyclohexadiene
presence of tris(p-bromophenyl)amine ((p-BrC6H4)3N) in aceto-
nitrile, the amine was oxidized to the corresponding aminium
cation radical. Using the molar extinction coefficient of the
aminium cation radical determined by triplet chloranil oxidation
in the same solvent, we estimated the concentration of aminium
cation radical produced at the end of the oxidation relative to
the concentration of the pyridine-2-thiyl radical byproduct
initially produced in the photolysis. In experiments with both
PTOC esters 1 and 2, the relative yield of aminium cation radical
was in the range 65-85%. This range represents a minimum
yield of 5 from radicals 3 and 4 because we do not know if
electron transfer to 5 from (p-BrC6H4)3N is the only reaction
pathway. In fact, when the same quantitation studies were
attempted with Ph3N, which is more easily oxidized than (p-
BrC6H4)3N, the yield of aminium cation radical was in the 40-
50% range. The latter results suggest that radical cation 5 reacted
in part with Ph3N by electrophilic addition, presumably to a
phenyl group. Irrespective of the details of the reaction pathways
with the amine, the yield of aminium cation radical produced
from (p-BrC6H4)3N indicates that heterolyses of 3 and 4 were
the major, if not only, primary reactions occurring in acetonitrile.
The kinetics of the primary reactions of radicals 3 and 4 were
studied in THF and in acetonitrile. Reactions of both in THF
were slow enough to follow by LFP with nanosecond resolution
methods. At 20 °C, radical 3 reacted with a rate constant of 2.8
× 105 s-1, and radical 4 reacted with a rate constant of 3.4 ×
106 s-1. In acetonitrile, both radicals gave radical cation 5
rapidly; the reaction of 4 at 20 °C was too fast to measure with
(CHD) Concentrationa
% yield
PTOC [CHD] (M)
15
18
23
15:18
1
1
1
2
2
2
0.5
22.4 ( 1.2 3.7 ( 0.3 3.2 ( 0.2 6.1 ( 0.6
18.2 ( 0.5 1.4 ( 0.3 3.0 ( 0.1 13.0 ( 2.8
0.25
0.125
0.5
0.25
0.125
14.1 ( 0.7 0.8 ( 0.1 2.6 ( 0.1 17.6 ( 2.4
b
38.3 ( 2.2 4.4 ( 0.7
33.6 ( 1.1 2.4 ( 0.1
26.4 ( 1.0 1.2 ( 0.2
-
-
-
8.7 ( 1.5
14.0 ( 0.7
22.0 ( 3.8
b
b
a [PTOC] ) 5 mM, [TBP] ) 10 mM. b Not observed.
Figure 1. (A) Time-resolved spectrum from photolysis of 2 in THF.
The traces are at 0.2, 0.3, 0.5, and 1.1 µs after irradiation with the
signal at 50 ns subtracted to give a baseline. (B) Time-resolved spectrum
from photolysis of 1 in acetonitrile. The traces are at 20, 30, and 100
ns after photolysis; the negative signal in the trace at 20 ns results
because the PM tube has not recovered from the scattered laser light
of the photolysis. The insets show typical kinetic traces with time in
ns.
our unit, but an approximate rate constant (k ) 8 × 107 s-1
)
was found for radical 3.
The results of the LFP studies with 3 and 4 are similar to
those found with related radicals containing â-leaving
groups.10,12b,13 Specifically, increases in solvent polarity were
found to result in accelerations in the rates of reactions and, in
some cases, changes from products resulting from concerted
reaction pathways or ion pair collapse pathways to diffusively
free radical cations. In the earlier work where the products were
sensitive to solvent polarity, we concluded that the reactions
occurred by initial heterolysis followed by collapse or dissocia-
tion of the ion pair, with the latter becoming increasingly
efficient as the polarity of the solvent increased.10,12b The
mechanistic details for 3 and 4 are discussed below, but one
can conclude from the LFP kinetic studies that no bimolecular
reactions of these radicals are fast enough to compete with the
unimolecular processes; in MeOH-acetonitrile (4:1), the life-
time of radical 3 is less than 1 ns, and radical 4 has a shorter
lifetime.
in Figure 1a, whereas photolysis of either in acetonitrile gave
the spectrum shown in Figure 1b. Photolysis of PTOC ester 1
in acetonitrile containing up to 1 M MeOH gave the same
spectrum as observed in neat acetonitrile. The spectrum in Figure
1b is quite similar to those of simple styrene radical cations4b
and the p-methoxy-â,â-dimethylstryene radical cation,12b and
we ascribe it to radical cation 5. The spectrum in Figure 1a
does not appear to be that of a benzylic radical that should have
λ
max at about 325 nm,35 but it is similar to that of the 2-methyl-
1-phenylallyl radical;13b we tentatively ascribe this spectrum to
the 2-methoxy-1-phenylallyl radical. The change in products
from radicals 3 and 4 with solvent polarity mirrors the behavior
seen with the 1,1-dimethyl-2-(p-methoxyphenyl)-2-(diethylphos-
phatoxy)ethyl radical that gave the benzylic radical migration
product in low-polarity solvents and the radical cation in high-
polarity solvents.12b Our interpretation of the behavior of radicals
3 and 4 is that heterolysis of each gave a radical cation-anion
pair that reacted by proton transfer to give the allylic radical in
low-polarity THF but dissociated to give diffusively free radical
cation 5 in the higher polarity solvents, acetonitrile and MeOH-
acetonitrile (4:1). Because all of the trapping studies conducted
in this work were performed in the highly polar solvent mixture
MeOH-acetonitrile (4:1), we conclude that radical cation 5 was
formed from radicals 3 and 4 under the conditions of the product
studies.
LFP studies of reactions of radical cation 5 were conducted
by following the decay of the signal at 368 nm in the presence
of potential reactants. From studies with varying concentrations
of 1,4-cyclohexadiene (CHD) in acetonitrile at 20 °C, a plot of
kobs versus concentration of CHD had a slope of k ) 6 × 105
M-1 s-1. We consider this to be an approximate rate constant
for reaction of CHD with 5 because biexponential decays were
observed at all concentrations of CHD studied. When radical 5
was produced in acetonitrile at ambient temperature in the
presence of varying concentrations of MeOH (up to 1 M), we
observed no increase in signal decay; the limit for the rate
The amount of radical cation 5 formed from radicals 3 and
4 could be evaluated in the following manner. The intensity of
the signal from 5 relative to that of the byproduct of the
photolysis reaction, the pyridine-2-thiyl radical, was the same
from both 3 and 4 suggesting a high-yield conversion to 5 from
both precursor radicals. When 5 was produced from 3 in the
constant for reaction of 5 with MeOH is k < 1 × 103 M-1 s-1
.
Discussion
Initially, a variety of mechanisms involving several reactive
intermediates was envisioned to explain the formation of the
products described above. Possible mechanisms for the forma-
(35) Chatgilialoglu, C. In Handbook of Organic Photochemistry; Scaiano,
J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. 2, pp 3-11.

3628 J. Am. Chem. Soc., Vol. 123, No. 16, 2001
Bales et al.
Scheme 6
PTOC precursors does not enable us to eliminate involvement
of mechanism 4 in product formation, in which deprotonation
of the highly acidic olefin cation radical (5) is effected by the
counterion in the solvent cage.1b,36 The small amount of
deuterium observed in 15 formed in the presence of d4-CHD is
also consistent with this cage process. LFP experiments provided
no evidence that the requisite allylic radical was formed from
3 or 4 in acetonitrile. Nonetheless, we believe the transient
observed in the less polar solvent THF was due to this radical,
and we cannot exclude formation of small amounts of the allylic
radical in more polar media.
Laser flash photolysis unambiguously supports trapping 5 by
CHD (k ) 6 × 105 M-1 s-1), which ultimately results in vinyl
ether (15) formation. Hydrogen atom transfer to 5 can occur to
carbon or oxygen (mechanism 5, Scheme 6). The small amount
of deuterium in 15 when d4-CHD is employed (Table 1) could
be indicative of a minor contribution of hydrogen atom transfer
to the R-carbon relative to the phenyl ring (path C). If hydrogen
atom abstraction by the R- and/or â-carbon of 5 provided a
major route to 15 (path C, Scheme 6), methanol trapping of 29
and/or 30 to yield the ketal (18) and/or vicinal ether (17),
respectively, was expected to be faster than deprotonation.37
Absence of nucleophilic trapping product 17, combined with
the formation of 18 in low yield, is inconsistent with hydrogen
atom abstraction by the carbon(s) of 5 providing a major
contribution to product formation.
tion of the major product, vinyl ether 15, are discussed below:
(1) â-scission from 3 and/or 4; (2) hydrogen atom donation to
3 and/or 4, followed by elimination of H-X; (3) migration of
“X” in the initially formed radical (3, 4), followed by sequential
hydrogen atom transfer and â-elimination of H-X (Scheme 4);
(4) formation and subsequent hydrogen atom donation to the
allyl radical formed via proton abstraction from 5 within the
solvent cage by the respective anion (phosphate diester, bromide)
(Scheme 5); and (5) sequential hydrogen atom abstraction
followed by deprotonation of the resulting cation. Hydrogen
atom abstraction could occur at either carbon or oxygen (Scheme
6).
The mechanism most consistent with these observations is
one in which 1,4-cyclohexadiene (CHD) transfers a hydrogen
(deuterium) atom to the oxygen of the olefin cation radical (5),
followed by deprotonation of the resulting protonated vinyl ether
(Scheme 6). Analogous hydrogen atom transfers to cation
radicals of ethers and sulfides have been observed in the gas
phase, and may be germane to reports of hydrogen atom
abstraction by fullerene cation radicals.38-40 Involvement of
CHD in the formation of the vinyl ether is indicated by the
effect of deuteration (Table 1), as well as the effect of CHD
concentration on the yield of 15 (Table 3). Substitution of d4-
CHD results in approximately a 2-fold reduction in the yield
of 15, independent of the PTOC precursor. The overall observed
KIE of ∼2 is a manifestation of primary and secondary effects,
and is in-line with those reported for hydrogen atom abstraction
by other cation radicals.41 In addition, the observation of
comparable KIEs for the formation of 15 derived from 1 and 2
indicates that the two PTOC esters yield 15 via a common
intermediate, the olefin cation radical (5). These results are fully
consistent with laser flash photolysis studies carried out on 1
and 2. Reactions of cation radicals in the gas phase led us to
investigate the possibility that methanol donates a hydrogen
atom to 5.38 However, LFP provided no evidence for scavenging
the transient assigned to 5, and established an upper limit for
trapping by methanol to be k ) 1 × 103 M-1 s-1. These
On the basis of the high energy of phosphate radicals,
mechanism 1 was deemed unlikely for 4.19 Although this
mechanism could not be excluded a priori as a contributor to
the formation of 15 from the â-bromo radical (3), the depen-
dence of the yield of 15 from 1 (and 2) on the concentration
and isotopic content of CHD is inconsistent with this unimo-
lecular pathway. Moreover, LFP experiments indicated that the
phosphate and bromo radicals produced equivalent yields of
olefin cation radical (5), suggesting the â-scission was not
occurring in either 3 or 4.
Product studies and LFP also ruled out mechanism 2 as a
viable option. Independent synthesis of 20 and 21 enabled us
to eliminate mechanism 2 as the source of 15. The phosphate
triester (20) was stable to the reaction conditions. The respective
benzylic bromide (21) was unstable, but yielded 23 and not 15.
Hydrogen atom transfer to 3 or 4 was shown to be untenable
on the basis of the absolute rate constants of heterolysis
measured for 3 and 4 (>107 s-1). Bimolecular trapping by CHD
could not possibly compete with these unimolecular processes.35
Independent synthesis could not be used to dismiss mecha-
nism 3 (Scheme 4), but further consideration is unnecessary
due to the determination by LFP that benzyl radicals are not
produced in appreciable yields from reactions of the PTOC
esters in acetonitrile. Furthermore, if mechanism 3 was respon-
sible for a significant fraction of 15, anticipated KIEs for the
elimination reaction would have yielded 15 with a much higher
level of deuterium incorporation from d4-CHD than was
observed (Table 2). Product studies do not allow us to rule out
a migration mechanism completely, on account of the presence
of small amounts of deuterium in 15 (Table 2). However, the
absence of benzyl radicals in the current LFP experiments (and
others) suggests that invoking this pathway would be an
unnecessary obfuscation.
(36) Baciocchi, E.; Bietti, M.; Lanzalunga, O. Acc. Chem. Res. 2000,
33, 243.
(37) Thibblin, A. J. Am. Chem. Soc. 1989, 111, 5412.
(38) (a) Biermann, H. W.; Morton, T. H. J. Am. Chem. Soc. 1983, 105,
5025. (b) Morton, T. H.; Beauchamp, J. L. J. Am. Chem. Soc. 1975, 97,
2355. (c) Berruyer-Penaud, F.; Bouchoux, G.; Tortajada, J. Rapid Commun.
Mass Spectrom. 1992, 6, 37.
(39) Chyall, L. J.; Byrd, M. H. C.; Kentta¨maa, H. I. J. Am. Chem. Soc.
1994, 116, 10767.
(40) (a) Siedschlag, C.; Luftmann, H.; Wolff, C.; Mattay, J. Tetrahedron
1999, 55, 7805. (b) Dunsch, L.; Ziegs, F.; Siedschlag, C.; Mattay, J. Chem.
Eur. J. 2000, 6, 3547.
Higher levels of deuterium content in the vinyl ether would
also be observed if significant amounts of 15 were formed via
hydrogen atom donation to the allylic radical (mechanism 4,
Scheme 5). The observation of small amounts of 19 from both
(41) Primary KIE measurements for McLafferty rearrangements, cation
radicals of ethers, and intermediates in the Loeffler-Freytag reaction vary
from 1.2 to 1.6. See: (a) Green, M. M. Tetrahedron 1980, 36, 2687. (b)
Morton, T. H. Tetrahedron 1982, 38, 3195.

Formation of an Olefin Cation Radical
J. Am. Chem. Soc., Vol. 123, No. 16, 2001 3629
Scheme 7
With use of the upper limit for MeOH trapping (k < 1× 103
M-1 s-1) and the observed rate constant for CHD (0.5 M)
trapping (k ) 6 × 105 M-1 s-1), the expected ratio of 15:18
would be g15, provided this were the sole pathway for the
formation of 18.
experiments are supported by product studies in which the yield
of 15 was unaffected by substituting d4-MeOH for MeOH (Table
1).
Summary
Product studies suggest that both â-substituted radicals (3,
4) generate freely diffusible olefin cation radical 5 in a solution
composed of methanol-acetonitrile (4:1). Laser flash photolysis
experiments carried out in neat acetonitrile unequivocally
demonstrate the formation of 5 in equal amounts from 3 and 4.
The majority of 15 is believed to be formed by hydrogen atom
transfer to the oxygen of 5 by CHD, which reacts with
diffusively free cation radical at k ) 6 × 105 M-1 s-1. While
such reactions are well precedented in the gas phase, they have
at most been rarely observed in solution.38-41 Low yields of
ketal (18) are also produced. Laser flash photolysis experiments
and product studies indicate that nucleophilic trapping of 5 by
methanol (<1 × 103 M-1 s-1) is slow compared with hydrogen
atom transfer by 1,4-cyclohexadiene (CHD). The increased
“radical” reactivity and decreased “cation” reactivity seen with
5 are important features of radical cations that can be exploited
in synthetic applications, and the high radical reactivity in
hydrogen atom abstraction might be important in understanding
the details of other radical cation reactions with hydrogen atom
donors. LFP provides no evidence for the formation of any
reactive intermediates other than olefin cation radical, and imply
that both products (15, 18) are formed from reaction with 5.
Product studies support this proposal. By employing cation
radical precursors containing different leaving groups, we are
able to utilize the measurement of product ratios (15:18) to probe
for the involvement of reactive intermediates other than 5. The
observation that 1 and 2 produce ratios of 15:18 that are within
experimental error of each other supports the proposal that both
products are derived from a common reactive intermediate, the
olefin cation radical.
Several pathways can also be envisioned for ketal (18)
formation: (1) SRN2′ reaction with the respective â-substituted
radicals (Scheme 7); (2) methanolysis of the product (27)
resulting from sequential migration of “X” in the initially formed
radical (3, 4) and hydrogen atom transfer (Scheme 4); (3)
nucleophilic trapping of the olefin cation radical (5), followed
by hydrogen atom transfer to the benzylic radical (28, path B,
Scheme 6); and (4) hydrogen atom transfer to 5 at the benzylic
position, followed by nucleophilic trapping of the oxocarbenium
ion (29, path C, Scheme 6).
Ketal formation via the SRN2′ mechanism (mechanism 3,
Scheme 7) or via migration of X (mechanism 4, Scheme 4) in
the originally formed radicals (3, 4) (mechanism 4) is incon-
sistent with laser flash photolysis experiments. Both pathways
require the formation of benzyl radicals from 3 and 4 in
competition with production of radical cation 5. We conserva-
tively estimate that benzylic radical formation in acetonitrile
was <10% that of radical cation 5, and production of benzylic
radicals should be even less competitive in the more polar
solvent mixture MeOH-acetonitrile (4:1). Furthermore, forma-
tion of 18 via pathways 1 and 2 (above) requires derivation of
this product from reactive intermediates other than the olefin
cation radical (5). Product studies show that the ratio of 15:18
is identical within experimental error over a range of CHD
concentrations (Table 3). If 18 were formed from an intermediate
other than olefin cation radical (5) the product studies require
that the intermediates produced upon irradiation of 1 and 2
fortuitously be trapped at the same rate constant. A simpler
proposal is that product 18 is derived from reaction of CHD
and/or MeOH with olefin cation radical 5.
This interpretation is consistent with the product studies.
Formation of 18 from olefin cation radical can also be reconciled
with LFP studies if one allows for the possibility that hydrogen
atom transfer to the carbon atom of 5 occurs to a minor extent
(in competition with transfer to oxygen) and occurs exclusively
at the less hindered benzylic position. Atom transfer from either
CHD or MeOH is consistent with isotopic analysis of 18 (Table
2). The subsequently formed carbocation would then be trapped
by the nucleophile.
Regioselective nucleophilic trapping of 5 by MeOH, followed
by hydrogen atom transfer may also contribute to the formation
of 18. As per above, isotopic labeling studies indicate that 28
produced would have to be trapped by CHD and MeOH (Table
2).42 The fact that LFP experiments were only able to place an
upper limit on the reaction of 5 with MeOH does not rule out
this possibility, given the high [MeOH] (19.75 M) employed.
Experimental Section
Detailed experimental procedures are provided in the Supporting
Information.
Acknowledgment. We are grateful for research support from
the National Institutes of Health (CA60500 to D.C., GM-54996
to M.M.G.) and the National Science Foundation (CHE-9986200
to D.C., CHE-9732843 to M.M.G., CHE-9981746 to M.N.).
Mass spectra at CSU were obtained on instruments supported
by the NIH shared instrumentation grant GM-49631.
Supporting Information Available: Experimental proce-
dures for the synthesis of PTOC esters and products derived
from them as well as procedures for the photolysis of PTOC
esters and subsequent analysis, including product stability studies
(PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
(42) The absolute bimolecular rate constant for reaction between CHD
and benzyl radical was estimated to be <102 M-1 s-1 (see ref 31.). The
respective rate constant for reaction of a benzyl radical with methanol has
not been reported, but is expected to be less than that for CHD. We believe
that the high concentrations of the reagents ([CHD] ) 0.5 M, [CH3OH] )
19.75 M) help overcome these kinetic barriers.
JA0042938