Peroxyl radical clocks

Article abstract of DOI:10.1021/jo0601462

A series of peroxyl radical clocks has been developed and calibrated based on the competition between the unimolecular β-fragmentation (k _β) of a peroxyl radical and its bimolecular reaction with a hydrogen atom donor (k_H). These clocks are based on either methyl linoleate or allylbenzene and were calibrated directly with α-tocopherol or methyl linoleate, which have well-established rate constants for reaction with peroxyl radicals (k_H-tocopherol = 3.5 × 10^-6 M^-1 s^-1, k_H-linoieate = 62 M^-1 s^-1). This peroxyl radical clock methodology has been successfully applied to determine inhibition and propagation rate constants ranging from 10° to 10⁷ M^-1 s^-1.

Full text of DOI:10.1021/jo0601462

Peroxyl Radical Clocks
Bill Roschek, Jr., Keri A. Tallman,* Christopher L. Rector, Jason G. Gillmore,
Derek A. Pratt, Carlo Punta, and Ned A. Porter*
Department of Chemistry and Center in Molecular Toxicology, Vanderbilt UniVersity,
NashVille, Tennessee 37235
n.porter@Vanderbilt.edu
ReceiVed January 23, 2006
A series of peroxyl radical clocks has been developed and calibrated based on the competition between
the unimolecular â-fragmentation (kâ) of a peroxyl radical and its bimolecular reaction with a hydrogen
atom donor (kH). These clocks are based on either methyl linoleate or allylbenzene and were calibrated
directly with R-tocopherol or methyl linoleate, which have well-established rate constants for reaction
with peroxyl radicals (kH-tocopherol ) 3.5 × 106 M-1 s-1, kH-linoleate ) 62 M-1 s-1). This peroxyl radical
clock methodology has been successfully applied to determine inhibition and propagation rate constants
ranging from 100 to 107 M-1 s-1.
SCHEME 1. 5-Hexenyl Radical Cyclization Clock
Introduction
There are many methods for determining absolute rate
constants of radical-molecule reactions. The rotating sector,
flash photolysis, and pulse radiolysis approaches have provided
platforms for kinetic studies of autoxidation, radical polymer-
ization, and other important radical-molecule reactions.1 Indeed,
the determination of absolute rate constants has been the
cornerstone for recent advances in free-radical chemistry.
Knowledge of rate constants has been critical to understanding
the chemistry of oxygen-centered radicals. These species carry
the chain for radical reactions carried out under atmospheric
oxygen since alkyl radicals undergo diffusion-controlled reaction
with molecular oxygen to yield peroxyl radicals (eq 1). Peroxyl
radicals are therefore the predominant chain-carrying species
in the reaction of organic compounds with molecular oxygen,
and as such, the rate constants for peroxyl radical reactions have
been the focus of extensive investigation. Among the most
important reactions that peroxyl radicals undergo are atom-
transfer reactions with H-atom donors that can either serve to
propagate a chain reaction (kp, eq 2) or break it by forming a
radical that does not react with O2, e.g., phenolic antioxidants,
ArO-H (kinh, eq 3).2,3 The establishment of rate constants for
these important reactions stands today as a monument to the
exploration of reactivity by physical organic methods.
R• + O2 f ROO•
(1)
ROO• + R-H f ROOH + R•
(2)
(3)
ROO• + ArO-H f ROOH + ArO•
Radical clocks are an indirect method for determining
radical-molecule reaction kinetics by competition between a
unimolecular reaction having a known rate constant and a
bimolecular reaction with an unknown rate constant. For
example, the 5-hexenyl radical cyclization (Scheme 1) employs
a competition between a 5-exo-trig cyclization (kR) and abstrac-
tion of a hydrogen atom from a substrate A-H (kH). A simple
(1) (a) Burnett, G. M.; Melville, H. W. In Techniques of Organic
Chemistry; Friess, S. L., Lewis, E. S., Weissberger, A., Eds.; Interscience:
New York, 1963; Vol. VIII, Part II, Chapter 20. (b) Ingold, K. U. In Free
Radicals; Kochi, J. K., Ed.; Wiley: New York, 1973; Vol. 1, Chapter 2.
(c) Walling, C. Free Radicals in Solution; Wiley: New York, 1957. (d)
Howard, J. A. AdV. Free Radical Chem. 1972, 4, 49.
10.1021/jo0601462 CCC: $33.50 © 2006 American Chemical Society
Published on Web 03/14/2006
J. Org. Chem. 2006, 71, 3527-3532
3527

Roschek et al.
SCHEME 2. Concept of the Peroxyl Radical Clock
peroxyl radical fragmentations have values that are complimen-
tary to the fragmentations associated with methyl linoleate.7
Therefore, we envisioned that these two substrates would
provide a range of values for peroxyl radical fragmentations
useful for the development of a series of clocks.
kinetic analysis reveals that the product ratio [1]/[2] is directly
proportional to kH/kR. Once kR for the unimolecular reaction
has been established, an experiment to determine the product
ratio for a given concentration of AH can be performed and
the unknown rate constant kH can be derived.
The radical clock method is ideally suited to applications
involving carbon-centered radicals due to the vast literature
documenting the kinetics of skeletal rearrangements of alkyl
radicals. As a result, dozens of clocks have been used for
determining reaction rate constants ranging from 10-1 to 1012
M-1 s-1.4 There is of course a requirement that radical clocks
must be “standardized” to an absolute rate constant. In the case
of radical cyclization, a number of methods have been used to
calibrate various clocks.
Given the relative ease with which conventional radical clock
experiments are conducted, it may prove useful to have radical
clock methods for the determination of rate constants for
reactions of peroxyl radicals. In fact, we have previously
reported studies that provide the basis for the development of
peroxyl radical clocks5 based on the autoxidation of lipids such
as methyl linoleate, and herein, we provide a detailed discussion
of the methodology and describe extended studies of these
clocks with a variety of H-atom donors.
The autoxidation of 5 is known to proceed by a series of
free-radical reactions, with the product distribution highly
dependent on the hydrogen atom donor (H-donor) present
(Scheme 3). When autoxidations of 5 were carried out in the
presence of R-tocopherol (>0.01 M) or other good H-donors,
only the kinetically controlled oxidation products (11-13) were
observed. Oxygen partitions itself across the three positions of
the pentadienyl radical (7) to form the nonconjugated (9) and
conjugated (8 and 10) peroxyl radicals with partitioning
coefficients of R and 1-R, respectively. Since â-fragmentation
of the conjugated peroxyl radicals (8 and 10) does not compete
with trapping by R-tocopherol (R-TOH), only the cis,trans-
conjugated hydroperoxides (11 and 12) are observed.8 In contrast
to the conjugated peroxyl radicals, the nonconjugated peroxyl
radical (9) undergoes a rapid â-fragmentation (kâΙ), reforming
the original pentadienyl radical (7).9 Competing with this
â-fragmentation is H-atom transfer (kH) to the peroxyl radical
to form the nonconjugated hydroperoxide 13. This competition
between â-fragmentation and H-atom transfer establishes the
basis for what we call the fast peroxyl radical clock (MLfast
)
derived from 5.
As is true with any clock, it is desirable to cover a broad
time domain to maximize possible applications. We describe
the development of a series of peroxyl radical clocks based on
the competition (Scheme 2) between a unimolecular peroxyl
radical rearrangement (kR) and a bimolecular H-atom transfer
(kH), where kH collectively refers to compounds that may
propagate (kp) or inhibit (kinh) radical chain reactions. By
modifying the structure of the clock (R1-H), we have been able
to vary the rate constant for the rearrangement reaction (kR),
providing a range of clocks amenable to studying peroxyl radical
When autoxidations of 5 are carried out in the presence of
H-donors that do not compete with kâI, only the thermodynamic
conjugated hydroperoxides (11, 12, 18, and 19) are observed
(Scheme 3). Although it is assumed that upon initial H-atom
abstraction the nonconjugated peroxyl radical (9) is formed, kâI
is the predominant pathway (relative to kH) under these reaction
conditions and there is a negligible amount of 9 present in
product mixtures. Only the conjugated peroxyl radicals (8 and
10) and their subsequent reactions are relevant in the thermo-
dynamically controlled autoxidation of 5. Upon formation of 8
and 10, one possible pathway is bond rotation followed by
â-fragmentation to form the new pentadienyl radical (14 and
15) with the trans,cis-conformation. Oxygen partitions itself at
the transoid or cisoid ends of 14 and 15 to form either the trans,-
cis- (8 and 10) or trans,trans-conjugated peroxyl radicals (16
reactions having rate constants between 100 and 107 M-1 s-1
.
Results and Discussion
Design of Peroxyl Radical Clocks. Methyl linoleate (5) and
allylbenzene (6) were chosen as substrates for the development
of a series of peroxyl radical clocks. The autoxidation mecha-
nism of methyl linoleate has been studied in great detail, and it
is known to involve â-fragmentations of several intermediate
peroxyl radicals.6 In addition, theoretical calculations have been
carried out on peroxyl radicals derived from linoleate as well
as those formed from allylbenzenes. The calculations suggest
that the rate constants associated with allylbenzene-derived
(6) (a) For the identification of the kinetic products, see: Tallman, K.
A.; Roschek, B., Jr.; Porter, N. A. J. Am. Chem. Soc. 2004, 126, 9240 and
ref 5a. For the thermodynamic products, see: (b) Chan, H. W.-S.; Levett,
G. Lipids 1977, 12, 99. (c) Porter, N. A. Acc. Chem. Res. 1986, 19, 262.
(d) Porter, N. A.; Caldwell, S. E.; Mills, K. A. Lipids 1995, 30, 277.
(7) Pratt, D. A. Ph.D. Dissertation, Vanderbilt University, Nashville, TN,
December 2003.
(8) A value of 430 s-1 has been reported (see ref 5b), but we have
reevaluated the rate constants for fragmentation of the conjugated peroxyl
radicals.
(2) Ingold, K. U. Acc. Chem. Res. 1969, 2, 1.
(9) The mechanism for this â-fragmentation is under investigation. There
may be a contribution from a bis-allylic rearrangement through a radical-
dioxygen complex. Regardless of the mechanism, the kinetics remain
unchanged and does not influence the concept of the clock. For more
information regarding this radical-dioxygen complex, see: (a) Olivella,
S.; Sole´, A. J. Am. Chem. Soc. 2003, 125, 10641. (b) Mills, K. A.; Caldwell,
S. E.; Dubay, G. R.; Porter, N. A. J. Am. Chem. Soc. 1992, 114, 9689. (c)
Porter, N. A.; Mills, K. A.; Carter, R. L. J. Am. Chem. Soc. 1994, 116,
6690. (d) Porter, N. A.; Mills, K. A.; Caldwell, S. E.; Dubay, G. R. J. Am.
Chem. Soc. 1994, 116, 6697. (e) Lowe, J. R.; Porter, N. A. J. Am. Chem.
Soc. 1997, 119, 11534.
(3) (a) Cipollone, M.; Di Palma, C.; Pedulli, G. F. Appl. Magn. Reson.
1992, 3, 99. (b) Pedulli, G. F.; Lucarini, M.; Pedrielli, P.; Sagrini, M.;
Cipollone, M. Res. Chem. Intermed. 1996, 22, 1. (c) Lucarini, M.; Pedulli,
G. F., Valgimigli, L. J. Org. Chem. 1998, 63, 4497.
(4) (a) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317. (b)
Newcomb, M. Tetrahedron 1993, 49, 1151. (c) Newcomb, M.; Toy, P. H.
Acc. Chem. Res. 2000, 33, 449.
(5) (a) Tallman, K. A.; Pratt, D. A.; Porter, N. A. J. Am. Chem. Soc.
2001, 123, 11827. (b) Porter, N. A.; Wujek, D. G. J. Am. Chem. Soc. 1984,
106, 2626.
3528 J. Org. Chem., Vol. 71, No. 9, 2006

Peroxyl Radical Clocks
SCHEME 3. General Mechanism of Methyl Linoleate Oxidation
SCHEME 4. General Mechanism of Allylbenzene Oxidation
and 17) with partitioning coefficients of â and 1-â, respectively.
The conjugated peroxyl radicals can undergo â-fragmentation
to reform the original pentadienyl radical, denoted by kâII and
kâIII. Competing with this fragmentation is hydrogen atom
transfer to the peroxyl radical to form either the trans,cis- (11
and 12) or the trans,trans-conjugated hydroperoxides (18 and
19). This competition between hydrogen atom transfer and
â-fragmentation establishes our slow peroxyl radical clock
(MLslow).
Steady-state analysis of the mechanism presented in Scheme
3 leads to equations which describe the product distribution as
a function of oxygen partitioning, â-fragmentation, and [H-
donor] for autoxidations carried out under kinetically or
thermodynamically controlled conditions. Kinetically controlled
reactions (MLfast) are represented by eq 4, which describes the
product ratio as a function of R, kâΙ, kH, and [H-donor]. Equation
5 represents the thermodynamically controlled autoxidation
(MLslow), which describes the product distribution as a function
of â, kâII, kâIII, kH, and [H-donor].
(ca. 104 s-1). This expectation was based on the 3 kcal/mol lower
radical stabilization energy (RSE) of a benzyl group as compared
to an allyl group and, hence, R-vinylbenzyl radical as compared
to the pentadienyl (or R-vinylallyl) radical. Indeed, we carried
out density functional theory calculations that suggested that
the C-OO• bond in 21 is 10.5 kcal/mol, compared to 7.4 kcal/
mol in 9 and 14.2 in 8 and 10.10 According to a published plot
of experimental rate constants for â-fragmentation vs calculated
C-OO• bond dissociation enthalpies,10 we estimated that the
rate constant for â-fragmentation of 21 would therefore be 9 ×
kâI
[11 + 12]
[13]
1 - R
R
1 - R
R
MLfast
:
)
+
(4)
(5)
(
)
kH[H-donor]
kH[H-donor] kâIII
[11 + 12]
[18 + 19]
â
MLslow
:
)
+
(
)
kâII 1 - â
kâII(1 - â)
The autoxidation of methyl linoleate provides rate constants
for the â-fragmentation clock reaction in the 106 and 102 s-1
regimes, vide infra. Allylbenzene (6) was also chosen as a
substrate since we expected that the nonconjugated peroxyl
radical derived from it (21, below) would have a rate constant
for â-fragmentation intermediate to those of the nonconjugated
(9) and conjugated (8 and 10) peroxyl radicals used in the MLfast
and MLslow clock reactions, respectively, suggesting it would
be useful for clocking reactions in the intermediate time regime
104 s-1
.
The mechanism for the autoxidation of allylbenzene (6),
shown in Scheme 4, is very similar to that of methyl linoleate
(5, Scheme 3). H-atom abstraction from the benzylic position
yields an R-vinylbenzyl radical (20), which is trapped by O2
generating the nonconjugated and conjugated peroxyl radicals
(10) These calculations were done at the (RO)B3P86/6-311G(d,p)//
(U)B3P86/6-311G(d, p) level of theory as in ref 11.
J. Org. Chem, Vol. 71, No. 9, 2006 3529

Roschek et al.
21 and 22, respectively.11 The fraction of R-vinylbenzyl radicals
trapped at the benzylic position and leading to the nonconjugated
peroxyl radical is defined as R, analogous to the partitioning of
O2 at the central position of the nonconjugated dienes. Similarly,
this nonconjugated peroxyl radical (21) undergoes â-fragmenta-
tion (kâΙ) in competition with H-atom transfer (kH), setting up
the intermediate peroxyl radical clock (AB). Kinetic analysis
of the mechanism in Scheme 4 (eq 6) leads to the same
relationship between product ratio, R, kâΙ, kH, and [H-donor] as
defined in eq 4 for the kinetically controlled autoxidation of
methyl linoleate.
TABLE 1. Experimental Rate Constants for â-Fragmentation (kâ)
and Partitioning Coefficients (r or â) for Peroxyl Radical Clocks
clock
kâ (s-1
)
R or â
MLfasta (5)
(25)
2.57 ((0.27) × 106
2.94 ((0.30) × 106
2.62 ((0.27) × 105
690 ((70), 50 ((5)
0.45 ((0.05)
0.46 ((0.04)
0.74 ((0.12)
0.686c
ABa
b
MLslow
a Calibrated with R-TOH. b Calibrated with methyl linoleate. c Previously
determined, ref 6a. Errors reported are 95% confidence limits of the rate
constant as determined by t tests. Error has been propagated to include
experimental error, as well as the error associated with literature values for
the propagation rate constant of R-TOH and methyl linoleate ((10%).
kâI
[23] kH[H-donor]
[24]
SCHEME 5. Hydrogen Atom Donors of Interest
1 - R
R
1 - R
R
AB:
)
+
(6)
(
)
Calibration of Clocks. The values for kâ and the oxygen
partition coefficients (R and â) must be derived from controlled
autoxidations of 5 or 6 in the presence of a H-atom donor with
a well-established kH. Thus, the MLfast and AB clocks were
calibrated by measuring the ratio of oxidation products as a
function of R-TOH concentration (kH ) 3.5 × 106 M-1 s-1 at
37 °C).12 The rate constant for R-TOH is too high to reliably
determine product ratios for the MLslow clock. Therefore, this
clock was calibrated using methyl linoleate itself (5, kH ) 62
M-1 s-1 at 37 °C) as the H-atom donor.13
Autoxidations of 5 (0.1-2.7 M depending on H-donor
present) or 6 (3.0-4.3 M), initiated by 2,2′-azobis-(4-methoxy-
2,4-dimethylvaleronitrile) (MeOAMVN), were carried out in
benzene or chlorobenzene and incubated at 37 °C in the presence
of varying concentrations of R-TOH. In the case of the MLslow
clock, R-TOH was omitted and the concentration of methyl
linoleate was varied. Conditions were chosen such that a
negligible amount of H-atom donor was consumed since the
kinetic analysis assumes a constant concentration of H-atom
donor. Higher concentrations of substrate and longer reaction
times were required for 6, as compared to autoxidations of 5,
to obtain sufficient yields of oxidation products for analysis.
This is consistent with the higher C-H BDEs predicted for these
compounds.7
The products of methyl linoleate oxidation were measured
by HPLC analysis of the hydroperoxides for the MLfast clock,
whereas the oxidation products for MLslow were reduced to the
corresponding alcohols with triphenylphosphine. These analysis
conditions provided optimal separation of the oxidation products.
In addition to methyl linoleate, autoxidations were also carried
out with cis,cis-6,9-pentadecadiene (25), which has been shown
to give nearly identical results.6a Whereas the oxidation products
from 5 were analyzed by HPLC, the products derived from 25
were reduced to the corresponding alcohols, identified by
comparison to authentic samples, and analyzed by GC. The
hydroperoxides formed upon autoxidation of 6 were reduced
to the corresponding alcohols with triphenylphosphine and
subsequently analyzed by HPLC or GC. The nonconjugated and
conjugated alcohols observed under these conditions were
identified by comparison to authentic samples. Utilizing these
various substrates, we take advantage of complimentary methods
of analysis.
In addition to measuring the product ratio derived from each
oxidation, the concentration of H-donor was also monitored by
HPLC. The kinetic expressions require that the reactions be
carried out under pseudo-first-order conditions and hence a
constant concentration of H-donor during the course of the
reaction. Reaction conditions were chosen so that less than 5%
of H-donor was consumed.
The product ratios for each clock were plotted as a function
of H-donor concentration (i.e., MLfast: trans,cis-conjugated/
nonconjugated, AB: conjugated/nonconjugated, and MLslow
:
trans,cis-conjugated/trans,trans-conjugated). When the data was
fit by eqs 4-6, the values for the partitioning coefficients (R)
and â-fragmentations (kâI, kâII, and kâIII) were obtained (Table
1). The values for MLfast are on the order of 106 s-1, whereas
the kâI for AB is slightly less. For the MLslow clock, the
fragmentation rate constants are 690 and 50 s-1 for kâII and
kâIII, respectively. From the data it is clear that these substrates
provide a wide range of values for kâ and, hence, can be used
to clock a wide variety of H-donors.
Application of Peroxyl Radical Clocks. To demonstrate the
utility of the peroxyl radical clocks, we investigated the rate of
hydrogen atom abstraction from a series of compounds (Scheme
5). Many of these compounds have been studied elsewhere so
their rate constants are known. This provides an opportunity to
validate the radical clock method. The H-atom donors we have
chosen to study vary in structure and span a wide range of rate
(11) Although most of the spin resides in the aromatic ring upon
generation of the R-vinylbenzyl radical, we have drawn the radical
delocalized across the allyl moiety since trapping is only observed at the
allyl group. We have attempted to trap benzyl radicals at the ortho and
para positions with no success. See: Pratt, D. A.; Mills, J. H.; Porter, N.
A. J. Am. Chem. Soc. 2003, 125, 5801.
(12) The rate constant for R-TOH at 37 °C was obtained from a linear
interpolation of the two flanking rate constants at (a) 30 °C (3.2 × 106
M-1 s-1) (Burton, G. W.; Doba, T.; Gabe, E. J.; Hughes, L.; Lee, F. L.;
Prasad, L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053) and (b) 50
°C (4.1 (( 0.4) × 106 M-1 s-1): Pratt, D. A.; DiLabio, G. A.; Brigati, G.;
Pedulli, G. F.; Valgimigli, L. J. Am. Chem. Soc. 2001, 123, 4625.
(13) The rate constant for methyl linoleate has been reported as 62 M-1
s-1 at 30 °C (Howard, J. A.; Ingold, K. U. Can. J. Chem. 1967, 793). A
comparison of kp/(2kt)1/2 values at 30 °C and 37 °C (Cosgrove, J. P.; Church,
D. F.; Pryor, W. A. Lipids 1987, 22, 299) reveals comparable values. Until
the absolute rate constant for methyl linoleate is measured at 37 °C, we
assume the value to be approximately 62 M-1 s-1
.
3530 J. Org. Chem., Vol. 71, No. 9, 2006

Peroxyl Radical Clocks
TABLE 2. Rate Constants for Hydrogen Atom Donors Using
Peroxyl Radical Clock Methoda
0.05-20 times the rate of the clock reaction. In addition, recall
that the clock substrates are H-atom donors themselves.
Therefore, the concentration of H-atom donor of interest must
be chosen so that the reaction rate of the substrate with
propagating peroxyl radicals does not compete with the H-atom
donor of interest.
Confidence in the analytical protocol can be provided by
application of multiple clocks and/or multiple analytical meth-
ods. Although most of the antioxidants clocked with AB and
diene 25 were analyzed by GC, the analysis could also be carried
out with HPLC. In contrast, most analyses of methyl linoleate
oxidation mixtures were by HPLC. Multiple analytical methods
offer the advantage that the clocking experiments can be carried
out by the protocol of choice for a particular antioxidant.
A final caVeat concerns the fact that the clocks are themselves
oxidizable compounds, a requirement of the approach. This can
give rise to problems in the analysis if sufficient care is not
taken to ensure the appropriate purity of the clock. Commercial
methyl linoleate may have substantial amounts (1-2%) of
oxidation products as received and blank analyses should always
be performed to ensure that clock substrates are uncontaminated
by the very products formed in the clock experiment.
antioxidant
clock
exptlb kH (M-1 s-1
)
lit.c kH (M-1 s-1
)
27
MLfast (5)
MLfast (25)
MLfast (5)
MLfast (25)
MLfast (5)
MLfast (25)
AB
9.35 ((1.50) × 106
1.53 ((0.29) × 107
2.68 ((0.48) × 106
3.90 ((0.64) × 106
4.56 ((0.76) × 106
6.15 ((0.92) × 106
2.32 ((0.51) × 105
3.10 ((0.60) × 105
2.37 ((2.03) × 104
1.88 ((0.77) × 104
1.71 ((0.64) × 104
3.22 ((0.32) × 103
384 ((46)
8.6 × 106 (50 °C)d
3.8 × 106,e
5.7 × 106,e
28
29
30
32
31
31
33
34
35
36
37
6
8.5 × 104,e
AB
AB
na
1.4 × 104,e
MLslow
MLslow
MLslow
MLslow
MLslow
MLslow
MLslow
1.4 × 104,e
1.6 × 104 (65 °C)f
3.1 × 103,e
397g
265 ((33)
362g
na
367 ((42)
5.5 ((0.6)
10h
a Errors reported are 95% confidence limits of the rate constant as
determined by t tests. Error has been propagated to include experimental
error, as well as the error in kâ for each clock. Rate constants for MLfast
were derived from eq 4, AB from eq 6, and MLslow from eq 5. b Values
from clocks at 37 °C. c Literature values at 30 °C unless otherwise noted.
d Taken from ref 12b. e Taken from ref 12a. f Taken from ref 14. g Taken
from ref 15. h Taken from ref 16.
Conclusion
constants and this provides the opportunity to explore the general
application of the radical clock method. Some of the compounds
we chose to study fall into the antioxidant class given their
ability to inhibit autoxidation processes (kinh), whereas others
are known to propagate radical chemistry (kp). Regardless of
their classification (kinh or kp), the fundamental reaction is
H-atom transfer and the kinetics remain unchanged.
The peroxyl radical clock experiments were carried out in
the same manner as described above for the calibration of each
clock, substituting R-TOH with the compound of interest. The
clock was chosen on the basis of the expected rate constants
for each H-donor, using a clock with a kâ that best matched the
expected kH. If the kH could not be estimated, then multiple
clocks are used to find the best match. The faster antioxidants
were clocked with MeLin (5) or the diene (25) utilizing kâI
(MLfast) and the product ratios measured by HPLC or GC,
respectively. Allylbenzene (6) was used for the intermediate
antioxidants and MLslow was used for the compounds expected
to have the lowest kH values. The product ratio was determined
as a function of the H-donor concentration and plotted in the
same way as described above. Using eqs 4-6 for the appropriate
clock and substituting the derived values for the partitioning
coefficients and kâ from Table 1, the values for kH were
determined (Table 2).
For those compounds whose inhibition rate constants are
available in the literature, those obtained by the clock method
are in good agreement with previously reported values taking
into account the temperature differences. While literature values
have been measured in experiments that widely vary in solvent,
temperature, and method, the values reported here were derived
from experiments carried out in benzene or chlorobenzene at
37 °C by a single method. As a result, comparison of the trends
in kH is straightforward and reliable.
The clock approach involves a competition leading to two
different products and the experimental imperative is a deter-
mination of the relative amounts of these two products formed
at a known concentration of H-atom donor. As a general guide,
a unimolecular clock reaction should be selected to time H-atom
transfer reactions having bimolecular rate constants that are
Three peroxyl radical clocks were developed based on the
competition between the unimolecular â-fragmentation of a
peroxyl radical with bimolecular hydrogen atom transfer. These
clocks were calibrated with two common hydrogen atom donors,
R-TOH and methyl linoleate, which have well-established rate
constants for reaction with peroxyl radicals. The clocks are
amenable to studying a variety of hydrogen atom transfer
reactions with rate constants that range from 100 to 107 M-1
s-1
.
Experimental Section
Materials. The initiator, 2,2′-azobis(4-methoxy-2, 4-dimeth-
ylvaleronitrile) (MeOAMVN), was dried under high vacuum for 2
h. Benzene and chlorobenzene were passed through a column of
neutral alumina and stored over 4 Å molecular sieves. Clocks:
Methyl linoleate (5) was chromatographed on silica (10% EtOAc/
hexanes) prior to use. The synthesis of 6,9-pentadecadiene (25) was
previously reported.6a Allylbenzene (6) was chromatographed on
silica (hexanes) immediately before use to remove any oxidation
products. Antioxidants: R-TOH (26) was purified by flash chro-
matography (10% EtOAc/hexanes, silica), protected from light. It
is crucial that the R-TOH be purified prior to use and dried overnight
under high vacuum. NMBHA (32)17 and the benzofuran (29)11a were
synthesized by literature procedures. The synthesis of the pyrimi-
dinol antioxidant (27) and 37 is reported in the Supporting
Information. All other antioxidants used in this study were
commercial and purified if necessary. Oxidation products: The
oxidation products of MeLin (5) and diene (25) have been
(14) Howard, J. A.; Ingold, K. U. Can. J. Chem. 1965, 2724.
(15) Porter, N. A.; Lehman, L. S.; Weber, B. A.; Smith, K. J. J. Am.
Chem. Soc. 1981, 103, 6447.
(16) Howard, J. A.; Ingold, K. U. Can. J. Chem. 1967, 793.
(17) (a) For the synthesis, see: Coates, R. M.; Firsan, S. J. J. Org. Chem.
1986, 51, 5198. For related studies, see: b) Punta, C.; Rector, C. L.; Porter,
N. A. Chem. Res. Toxicol. 2005, 18, 349. (c) Minisci, F.; Recupero, F.;
Cecchetto, A.; Gambarotti, C.; Punta, C.; Faletti, R.; Paganelli, R.; Pedulli,
G. F. Eur. J. Org. Chem. 2004, 109. (d) Amorati, R.; Lucarini, M.; Mugnaini,
V.; Pedulli, G. F.; Minisci, F.; Recupero, F.; Fontana, F.; Astolfi, P.; Greci,
L. J. Org. Chem. 2003, 68, 1747.
J. Org. Chem, Vol. 71, No. 9, 2006 3531

Roschek et al.
characterized in a previous publication.6a The oxidation products
of allylbenzene (6) were purchased.
compound and the concentration of H-atom donor monitored over
time. The oxidation was stopped by the addition of excess phenol
(not the phenol being monitored), followed by the addition of the
internal standard. The hydroperoxides were reduced to their
corresponding alcohols with PPh3, and the samples were diluted
with acetonitrile or methanol (1.8 mL). HPLC analyses were carried
out with a C-18 column using methanol and water as the mobile
phase. H-atom donors 26-29 were analyzed using isocratic
conditions (26, 100:0; 27, 50:50, 28, 70:30; and 29, 70:30
MeOH,H2O). H-atom donors 5, 6, and 30-37 were analyzed using
a gradient (MeOH/H2O) 90:10 for 5 min, ramp to 75:25 in 10 min,
ramp to 20:80 in 15 min, ramp to 90:10 in 10 min. Consumption
was monitored at a wavelength >260 nm to ensure no peak
contamination from the oxidation products. When consumption of
NMBHA was monitored, acetonitrile was used instead of methanol
due to solubility.
General Procedure for Clocking Experiments. The samples
were prepared as described above for each clock in benzene and
varying the hydrogen atom donor concentration (see the Supporting
Information for specific information). The samples were incubated
at 37 °C for 4 h (MLfast and diene), 1-24 h (AB), or 1 h (MLslow).
Following the oxidation, BHT (50 mM) and internal standard (5
mM benzyl alcohol for MLfast, 5 mM tetradecane for diene, 5 mM
General Procedure for MeLin and Diene Clock Calibrations
(MLfast). Stock solutions of MeLin or diene (1.0 M), MeOAMVN
(0.1 M), and R-TOH (1.0 M) were prepared in benzene. For samples
containing high concentrations, neat R-TOH (2.2 M) was used.
Samples were prepared in 1.0 mL autosampler vials with a total
reaction volume of 100 µL. It is important to add the solutions in
the following order to avoid premature oxidation: R-TOH (0.05-
1.8 M), linoleate or diene (0.10 M), MeOAMVN (0.01 M) and
diluted to 100 µL with benzene. The sealed samples were then
incubated at 37 °C for 4 h.
After 4 h, the oxidation was stopped by the addition of BHT
(50 µL of 1.0 M solution in hexanes), followed by the addition of
the internal standard (5 mM benzyl alcohol for linoleate and 5 mM
tetradecane for diene). The linoleate samples were diluted to 1.0
mL with hexanes and analyzed by HPLC (0.5% i-PrOH/hexanes,
1 mL/min, detection at 207 nm) as their hydroperoxides. The
response factors for the nonconjugated and conjugated hydroper-
oxides are 0.77 and 1.09, respectively, relative to benzyl alcohol.
The diene samples were reduced with PPh3 (50 µL of 1.0 M
solution/hexanes) and analyzed by GC (100-180 °C at 5 °C/min,
180-280 °C at 20 °C/min, 10 min). The response factors for the
nonconjugated and conjugated alcohols are 1.05 and 1.08, respec-
tively, relative to tetradecane. The ratio of products (conjugated:
nonconjugated) was plotted versus the R-TOH concentration. The
values for R and kâ were derived from the y-intercept and slope,
respectively, using eq 4.
benzyl alcohol for AB, and 5 mM cinnamyl alcohol for MLslow
)
were added to each sample. MeLin samples (MLfast) were diluted
to 1 mL with hexanes and analyzed by HPLC (0.5% i-PrOH/
hexanes, 1 mL/min, detection at 207 nm). Oxidation mixtures from
MLslow clocking experiments were reduced with PPh3 and analyzed
by the same HPLC method. Oxidation mixtures of the diene and
AB were reduced with PPh3 (50 mM) and analyzed by GC: diene
(100-180 °C at 5 °C/min, 180-280 °C at 20 °C/min, 10 min),
AB (75 °C, 5 min-150 °C at 5 °C/min, 150-280 °C at 25 °C/
min, 2.8 min). The rate constants kH were derived from eqs 4-6
for the respective clock.
Error Analysis. All errors reported are 95% confidence limits
of the rate constant as determined by t tests, i.e., best fit ( (standard
error)(t*). The standard error was determined from SigmaPlot or
Origin and multiplied by t*. The value for t* was calculated in
Excel using the equation TINV (0.05, df), where df is the degrees
of freedom. The degrees of freedom were calculated by subtracting
the number of parameters (2) from the total number of data points.
The error in the literature values for the kH of R-TOH and methyl
linoleate ((10%) was also propagated into the error.
General Procedure for Allylbenzene Clock Calibrations (AB).
The oxidations were carried out in the same manner as described
for the MLfast clock with allylbenzene (2.26 M), MeOAMVN (0.02
M), and R-TOH (0.20-1.07 M) in benzene. Allylbenzene was also
calibrated in chlorobenzene to determine if there is a solvent effect
on the reaction. The samples were incubated at 37 °C for 4 h. The
oxidation was stopped by the addition of excess BHT, followed
by the addition of the internal standard (10 mM benzyl alcohol).
The hydroperoxides were reduced to their corresponding alcohols
with PPh3, and the samples were diluted with hexanes (1.8 mL).
HPLC analysis was carried out using the same conditions described
for the methyl linoleate clock. The response factors for the
nonconjugated and conjugated alcohols are 1.17 and 0.40, respec-
Ι
tively, relative to benzyl alcohol. The values for R and kâ were
derived from the y-intercept and slope, respectively, using eq 6.
General Procedure for Slow MeLin Calibration (MLslow). The
oxidations were carried out in a similar manner as the MLfast
calibrations using methyl linoleate (0.3-2.7 M) and MeOAMVN
(0.01 M) in chlorobenzene for 1 h at 37 °C. In this case, methyl
linoleate was its own H-atom donor. The oxidations were stopped
by the addition of excess BHT, reduced with PPh3, and followed
by the addition of the internal standard (5 mM cinnamyl or cumyl
alcohol). HPLC analysis was carried out using the same conditions
described for the MLfast clock. The response factors for the trans,-
cis- and trans,trans-conjugated alcohols are 0.37 relative to
cinnamyl alcohol and 0.27 and 0.26, respectively, relative to cumyl
alcohol. The values for kâII and kâΙII were determined from the slope
and the y-intercept, respectively, using eq 5.
Acknowledgment. This project was supported by funding
from the National Institutes of Health and the National Science
Foundation. K.A.T., C.L.R., J.G.G., and B.R. also acknowledge
support from the Center in Molecular Toxicology, Vanderbilt
University.
Supporting Information Available: Experimental procedures
for the synthesis of H-atom donors; plots of calibrations, H-atom
donor consumption and clocking experiments, and NMR spectra
of all synthesized compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
General Procedure for H-Atom Donor Consumption Experi-
ments. Oxidations were carried out as described above for each
JO0601462
3532 J. Org. Chem., Vol. 71, No. 9, 2006