Kinetic solvent effects on peroxyl radical reactions

Article abstract of DOI:10.1039/b800369f

Kinetic solvent effects on peroxyl radical reactions are easily determined using a new peroxyester-based radical clock method. The Royal Society of Chemistry.

Full text of DOI:10.1039/b800369f

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Kinetic solvent effects on peroxyl radical reactionsw
Mukund Jha and Derek A. Pratt*
Received (in Cambridge, UK) 9th January 2008, Accepted 25th January 2008
First published as an Advance Article on the web 8th February 2008
DOI: 10.1039/b800369f
Kinetic solvent effects on peroxyl radical reactions are easily
determined using a new peroxyester-based radical clock method.
In an attempt to address these limitations, we have investi-
gated the decomposition of the homoconjugated peroxyester 9
as a source of the delocalized carbon-centred radical 2a, which
is oxygenated to the nonconjugated (3) and conjugated (4)
peroxyl radicals in Scheme 1 (eqn (1)). This approach makes
possible the determination of KSEs on the b-fragmentation of
peroxyl radicals which, in turn, allows the determination of
KSEs on the reactions of radical-trapping antioxidants with
peroxyl radicals.
The reactions of peroxyl radicals have long been of tremen-
dous interest. They play important roles in both biological and
industrial contexts, as intermediates in radical-mediated per-
oxidation reactions that are involved in the oxidative degrada-
tion of biological molecules, primary petroleum products, fine
chemicals and polymers. The mechanisms of the reactions
important in these processes have been extensively studied,
and more recent attention has focused on understanding
structure–activity relationships of radical-trapping antioxi-
dants (e.g. phenols and aromatic amines) which are useful
for inhibiting peroxyl radical-mediated oxidation.
ð1Þ
Various approaches, such as the rotating sector method,
flash photolysis, pulse radiolysis and inhibited autoxidations,
have yielded important kinetic and mechanistic information
for peroxyl radical–molecule reactions. However, these appli-
cations are generally restricted to a handful of research groups
with the necessary specialized equipment and expertise to
carry out the measurements. Moreover, the systematic study
of kinetic solvent effects (KSEs) on these reactions has been
difficult to carry out by these methods.1
Thermal decompositions of peroxyesters including 9 have
been extensively studied under anaerobic conditions and are
believed to occur with concerted rupture of the O–O and C–C
bonds, leading to an alkyl radical (phenylallyl 2a in the case of
9), an alkoxyl radical (t-butoxyl in the case of 9) and carbon
dioxide.5 When we subjected 9 to thermal decomposition
(37 1C, 4 h)6 under aerobic conditions in the presence of a
good H-atom donor (e.g. a-tocopherol, a-TOH), the same
non-conjugated and conjugated hydroperoxides (5a and 6a)
were formed as in the kinetically-controlled autoxidations of
1a that form the basis of the peroxyl radical clock approach
described by Porter and co-workers.2
As in the autoxidation of 1a,2 the ratio of 5a to 6a formed
upon decomposition of 9 (analyzed by GC following PPh3
reduction to 7a and 8a, respectively) was found to be depen-
dent on the concentration of a-TOH (Fig. 1). The rate con-
stant for b-fragmentation of 3a could be determined (from the
equation in the inset of Fig. 1) to be kb = 1.7(ꢁ0.1) ꢂ 105 sꢃ1
at 37 1C using kH = 3.5 ꢂ 106 Mꢃ1 sꢃ1 at 37 1C for a-TOH,2 in
good agreement with that determined in the autoxidation of 1a
An interesting alternative, recently suggested by Porter and
co-workers,2 employs a kinetic competition approach between
the unimolecular b-fragmentation of a non-conjugated peroxyl
radical (3) and its trapping in a bimolecular reaction (with
X–H, Scheme 1). While the utility of this approach has been
demonstrated for determining rate constants for H-atom
transfer (HAT) from simple phenols to peroxyl radicals in
benzene,2 it has a major limitation: it requires that Xꢀ, the
radical derived from X–H, carry the oxidation chain reaction.
This presents a problem for studying phenols that yield either
persistent (e.g. butylated hydroxy toluene, BHT)3 or highly-
stabilized (e.g. 6-amino-3-pyridinols)4 radicals or other X–H
that give rise to radicals that have difficulty carrying the chain
(e.g. aromatic amines). Furthermore, because this chain-pro-
pagating reaction is so slow (kp, Scheme 1), a large concentra-
tion of substrate is required (e.g. 2.6 M 1a) to ensure sufficient
oxidation products are formed to allow accurate, reproducible
rate constants to be determined; again precluding the study of
KSEs on these reactions.
Department of Chemistry, Queen’s University, 90 Bader Crescent,
Kingston, Ontario, Canada K7L 3N6. E-mail:
pratt@chem.queensu.ca; Tel: +1 613 533 3287
w Electronic supplementary information (ESI) available: Experimental
procedures and raw data for Tables 1–3. See DOI: 10.1039/b800369f
Scheme 1
ꢄc
This journal is The Royal Society of Chemistry 2008
1252 | Chem. Commun., 2008, 1252–1254

View Online
Table 1 KSEs on the b-fragmentation of peroxyl radical 3a
Solvent
kH/Mꢃ1 sꢃ1a
kb/sꢃ1b
ac
Hexane
Benzene
Anisole
t-Amyl alcohol
Acetic acid
Ethyl acetate
Propionitrile
a
2.0 ꢂ 107
3.4 ꢂ 106
1.5 ꢂ 106
5.6 ꢂ 105d
8.8 ꢂ 105
2.3 ꢂ 105
3.8 ꢂ 105e
3.5(ꢁ0.4) ꢂ 105
1.4(ꢁ0.1) ꢂ 105
6.9(ꢁ0.1) ꢂ 104
3.7(ꢁ0.2) ꢂ 104
3.5(ꢁ0.2) ꢂ 104
2.1(ꢁ0.1) ꢂ 104
1.8(ꢁ0.2) ꢂ 104
0.77 ꢁ 0.01
0.76 ꢁ 0.01
0.77 ꢁ 0.01
0.76 ꢁ 0.01
0.75 ꢁ 0.01
0.73 ꢁ 0.01
0.70 ꢁ 0.01
Rate constants for reaction of a-TOH with cumylperoxyl radicals at
25 1C,9 used to derive kb and a as in Scheme 1. Determined at
b
c
25 1C. The partitioning of O2 upon addition to the phenylallyl
radical 2a as in Scheme 1. Determined in t-butanol. Determined
in acetonitrile.
d
e
Fig. 1 Ratio of 7a to 8a formed from the decomposition of 9 as a
function of [a-TOH] following 4 h at 37 1C in benzene (K,—) and
following 30 min irradiation (300 nm) at 25 1C in benzene (J,—).
and a = 0.76 ꢁ 0.01 at 25 1C, in very good agreement with the
data obtained at 37 1C (Fig. 1).
The rate constant for the reaction of a-TOH with cumylper-
oxyl radicals (kH) has been determined by laser flash photo-
lysis in several solvents,9 allowing us to determine the KSE on
the b-fragmentation of 3a.10 The results are shown in Table 1.
A significant KSE is observed on the b-fragmentation of 3a,
where as the polarity of the solvent increases, kb decreases. For
the selection of solvents we have examined in this study, the
decrease in kb is roughly 20-fold on going from the least polar
(hexane) to the most polar (propionitrile) solvent (Table 1).
The trend is reminiscent of the KSE observed on the
a-fragmentation of acyl radicals (eqn (2)), for which a smaller
effect is observed with a decrease in rate constant of roughly
4-fold on going from hexane to acetonitrile.11 These results
suggest that the peroxyl radical is better solvated than the
transition state (TS) for C–OOꢀ bond dissociation as the
polarity of the medium increases, as is believed to be the case
in the a-fragmentation of acyl radicals.11 The larger KSE on
peroxyl radical fragmentation is consistent with a greater
decrease in dipolar interactions12 with solvent on approaching
the TS (eqn (3)) as compared to acyl radical fragmentation
(eqn (2)).
(kb = 2.6(ꢁ0.3) ꢂ 105 sꢃ1),2 but with only 10 mM peroxyester
as a precursor to 2a. The values of a also agree well:
0.76 ꢁ 0.01 vs, 0.74 ꢁ 0.12 (at 37 1C).2
In order to minimize the incubation time necessary for
generation of sufficient radicals for clocking experiments,
photochemical decomposition of 9 was also carried out.
Irradiation at 300 nm for 30 min was found to be sufficient
to generate comparable amounts of products for accurate and
reproducible analysis (completion is only a few percent under
these conditions). However, the photolytic decomposition of 9
resulted in a slightly different product profile compared to its
thermal decomposition. Along with the formation of the
expected alcohols (7a and 8a) following work-up, two new
products were identified in the chromatograms, corresponding
to the diastereomers of 1,2-epoxy-3-hydroxy-3-phenylpropane
(10). Indeed, under the photolytic conditions, the O–O bond in
6a is homolytically cleaved to yield the corresponding alkoxyl
radical, which can undergo a 3-exo cyclization to a resonance-
stabilized benzyl radical intermediate in competition with
HAT from a-TOH (or any H-donor, X–H).7 This alkyl radical
can then undergo subsequent reactions with O2, a-TOH and
then PPh3 (on work-up) to yield 10 (Scheme 2). Epoxides
expected from an analogous cyclization of the alkoxyl radical
derived from 5a were not observed,8 presumably since either a
primary alkyl radical or dearomatized intermediate would
arise. Thus, in experiments carried out by photochemical de-
composition of 9, the ratios of products are corrected for
epoxide formation, i.e., the ratio on the y-axis in Fig. 1 is
actually [7a]/([8a] + [10]). This affords kb = 1.4(ꢁ0.1) ꢂ 105 sꢃ1
ð2Þ
ð3Þ
Since we have now determined kb for 3a in several solvents,
the peroxyl radical clock approach can be applied to determine
KSEs on peroxyl radical–molecule reactions. For our initial
studies we have chosen a small, but representative set of
antioxidants of varying structure and reactivity: 4-methylphe-
nol (11), 2,4,6-trimethylphenol (12), 2,6-di-tert-butyl-4-
methylphenol (13) and diphenylamine (14). First, to validate
the approach, we restricted our studies to benzene since
comparable literature data were available obtained using
conventional approaches (Table 2). For 11–13, the rate con-
stants obtained using the peroxyester approach were within a
factor of 2 of the literature data, confirming its accuracy. For
14, our rate constant is roughly 3.5-fold greater than the
literature value. However, it should be pointed out that the
Scheme 2
ꢄc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 1252–1254 | 1253

View Online
Table 2 Rate constants for reactions of 11–14 with peroxyl radicals
at 25 1C in benzene using the peroxyester-based radical clock ap-
proach. Literature values at 30 1C are presented for comparison
Table 3 KSEs on the reactions of 11–14 with peroxyl radicals given
as the slopes of correlations of log kSH vs, bH2 (see ESIw) compared with
that expected from eqn (4)
Slope (R)
a2H
ꢃ8.3aH2
11
12
13
14
ꢃ5.9 (0.98)
ꢃ4.9 (0.99)
ꢃ2.9 (0.98)
ꢃ2.5 (0.99)
0.573
0.374
0.216
0.324
ꢃ4.8
ꢃ3.1
ꢃ1.8
ꢃ2.7
Clock kH/Mꢃ1 sꢃ1
Lit. kH/Mꢃ1 sꢃ1
Ref.
11
12
13
14
4.4(ꢁ0.1) ꢂ 104
1.7(ꢁ0.3) ꢂ 105
3.0(ꢁ0.2) ꢂ 104
5.3(ꢁ0.6) ꢂ 104
2.6 ꢂ 104
8.5 ꢂ 104
1.9 ꢂ 104
1.5 ꢂ 104
13
14
14
15
DPPHꢀ have been described as solvent-independent,19,20 the
reactivities of peroxyl radicals may be expected to display some
solvent dependence, as they are subject to significant dipole–
dipole12 and H-bonding21 interactions.
This work was supported by the Natural Sciences and
Engineering Research Council of Canada and Queen’s
University. DAP also acknowledges the Canada Research
Chairs program.
Notes and references
Scheme 3
1 G. Litwinienko and K. U. Ingold, Acc. Chem. Res., 2007, 40, 222–230.
2 B. Roschek, Jr, K. A. Tallman, C. L. Rector, J. G. Gillmore, D. A.
Pratt, C. Punta and N. A. Porter, J. Org. Chem., 2006, 71, 3527–3532.
3 Despite its inclusion in ref. 2 and our own extensive efforts, we have
repeatedly failed to obtain a rate constant for BHT by this method.
4 M. Wijtmans, D. A. Pratt, L. Valgimigli, G. A. DiLabio, G. F. Pedulli
and N. A. Porter, Angew. Chem., Int. Ed., 2003, 42, 4370–4373.
5 P. D. Bartlett and R. R. Hiatt, J. Am. Chem. Soc., 1958, 80,
1398–1402.
literature value was estimated from an inhibited autoxidation
of styrene for which no well-defined inhibition period was
observed.15
Until now, only KSEs for reactions of alkoxyl radicals (i.e.
t-butoxyl) and hydrazyl radicals (i.e. 2,2-diphenyl-1-picryl-
hydrazyl, DPPHꢀ) with phenols have been systematically
determined.1 The KSEs observed in these reactions are domi-
nated by the H-bonding interaction between the phenols
(ArOH) and H-bond accepting solvents (S), which effectively
‘tie up’ the phenolic O–H, preventing abstraction of the
H-atom by the radical (Yꢀ, Scheme 3).
6 We suspect that some (if not all) of the decomposition we observe may
be induced by a-TOH, which yields the same products as homolysis.
7 From the ratio of 10 to 8a determined as a function of [a-TOH] we
estimate a rate constant for the cyclization of the cinnamoxyl radical
9
of 5.1 ꢂ 106 sꢃ1 from k = 2.8 ꢂ 108 Mꢃ1 sꢃ1 for a-TOH + ROꢀ
.
8 By comparison with authentic materials synthesized indepen-
dently, as in X. Xie, G. Yue, S. Tang, X. Huo, Q. Liang, X. She
and X. Pan, Org. Lett., 2005, 7, 4057–4059.
Ingold and co-workers have expressed the rate constant for
HAT from phenol to a radical in a given solvent (kSH) in terms
of the H-bond accepting ability of the solvent (given by its b2H
parameter16) and the H-bond donating ability of the phenol
(given by its aH2 parameter17) as in eqn (4).18 This equation is
believed to be valid for the prediction of KSEs on HAT
reactions since they are assumed to be independent of the
nature of the abstracting radical, i.e. any interaction of the
abstracting radical with the solvent does not substantially
affect its reactivity.18 However, since eqn (4) was derived from
data for reactions of phenols with t-butoxyls and DPPHꢀ, we
wondered whether the reactions of phenols with peroxyls—
arguably the most important radical reaction undergone by
phenols—would also obey the relationship.
9 L. Valgimigli, J. T. Banks, J. Lusztyk and K. U. Ingold, J. Org.
Chem., 1999, 64, 3381–3383.
10 While cumylperoxyl is a tertiary peroxyl and 3a is a secondary peroxyl,
rate constants determined by flash photolysis using cumylperoxyl are
indistinguishable from those determined by inhibited autoxidations
whose chain reactions are carried by secondary peroxyls, see ref. 9.
11 Y. P. Tsentalovich and H. Fischer, J. Chem. Soc., Perkin Trans. 2,
1994, 729–733.
12 The dipole moment in benzylperoxyl has been determined to be
(2.4 ꢁ 0.2) D, see R. W. Fessenden, A. Hitachi and V. Nagarajan,
J. Phys. Chem., 1984, 88, 107–110, and peroxyl radicals are
predicted to be highly polarizable by solvent, see P. Aplincourt,
M. F. Ruiz-Lopez, X. Assfeld and F. Bohr, J. Comput. Chem.,
1999, 20, 1039–1048.
13 J. A. Howard, in Radical Reaction Rates in Liquids, ed. W.
Martienssen, Landolt-Bornstein, Springer, New York, 1996, vol.
¨
18, pp. 1–434.
log kSH = log kH0 ꢃ 8.3aH2 bH2
(4)
14 G. W. Burton, T. Doba, E. Gabe, L. Hughes, F. L. Lee, L. Prasad
and K. U. Ingold, J. Am. Chem. Soc., 1985, 107, 7053–7065.
15 M. Lucarini, P. Pedrielli, G. F. Pedulli, L. Valgimigli, D. Gigmes
and P. Tordo, J. Am. Chem. Soc., 1999, 121, 11546–11553.
16 M. H. Abraham, P. L. Grellier, D. V. Prior, P. P. Duce, J. J. Morris
and P. J. Taylor, J. Chem. Soc., Perkin Trans. 2, 1989, 699–711.
17 M. H. Abraham, P. L. Grellier, D. V. Prior, J. J. Morris and P. J.
Taylor, J. Chem. Soc., Perkin Trans. 2, 1990, 521–529.
18 D. W. Snelgrove, J. Lusztyk, J. T. Banks, P. Mulder and K. U.
Ingold, J. Am. Chem. Soc., 2001, 123, 469–477.
19 D. V. Avila, C. E. Brown, K. U. Ingold and J. Lusztyk, J. Am.
Chem. Soc., 1993, 115, 466–470.
20 L. Valgimigli, K. U. Ingold and J. Lusztyk, J. Org. Chem., 1996,
61, 7947–7950.
21 V. Mugnaini and M. Lucarini, Org. Lett., 2007, 9, 2725–2728.
Plots of log kH vs, solvent bH2 parameter for 11–14 were
linear with good correlation coefficients (Table 3), confirming
that KSEs on the reactions of peroxyl radicals with phenols (and
14) are dominated by H-bonding of the phenolic O–H to the
solvent. While it is gratifying that the slopes of these correlations
for 11–13 are indeed proportional to aH2 , they are larger than
expected from eqn (4) (ꢃ8.3a2H, see Table 3). Thus, there would
appear to be a slightly larger solvent dependence of peroxyl
radical reactions with phenols as compared to t-butoxyl radicals
or DPPHꢀ. While the reactivities of t-butoxyl radicals and
ꢄc
This journal is The Royal Society of Chemistry 2008
1254 | Chem. Commun., 2008, 1252–1254