A continuous visible light spectrophotometric approach to accurately determine the reactivity of radical-trapping antioxidants

Article abstract of DOI:10.1021/acs.joc.5b02183

Inhibited autoxidations-monitored either by O₂ consumption or hydroperoxide formation-are the most reliable way to obtain kinetic and stoichiometric information on the activity of radical-trapping antioxidants (RTAs). While many comparatively simple "antioxidant assays" (e.g., the DPPH assay) have supplanted these in popularity, they are generally very poor substitutes since they often do not employ peroxyl radicals as the oxidant and do not account for both the kinetics and stoichiometry of the radical-trapping reaction(s). In an effort to make inhibited autoxidations as simple as the most popular "antioxidant assays", we have developed a spectrophotometric approach for monitoring reaction progress in inhibited autoxidations. The approach employs easily prepared 1-phenylbutadiene-conjugated or styrene-conjugated BODIPY chromophores (PBD-BODIPY or STY-BODIPY, respectively) as signal carriers that co-autoxidize along with a hydrocarbon substrate. We show that inhibition rate constants (k_inh) are accurately determined for a range of phenolic and diarylamine RTAs using this approach and that mechanistic experiments, such as kinetic isotope effects and kinetic solvent effects, are equally easily carried out. Moreover, synergistic interactions between RTAs, as well as the unconventional activity of diarylamine RTAs, are captured using this methodology. Lastly, we show that the approach can be employed for monitoring reactions in aqueous solution.

Full text of DOI:10.1021/acs.joc.5b02183

Featured Article
pubs.acs.org/joc
A Continuous Visible Light Spectrophotometric Approach To
Accurately Determine the Reactivity of Radical-Trapping Antioxidants
Evan A. Haidasz, Antonius T. M. Van Kessel, and Derek A. Pratt*
Department of Chemistry and Biomolecular Sciences, University of Ottawa, Ottawa, Ontario, K1N 6N5, Canada
S
* Supporting Information
ABSTRACT: Inhibited autoxidationsmonitored either by
O2 consumption or hydroperoxide formationare the most
reliable way to obtain kinetic and stoichiometric information
on the activity of radical-trapping antioxidants (RTAs). While
many comparatively simple “antioxidant assays” (e.g., the
DPPH assay) have supplanted these in popularity, they are
generally very poor substitutes since they often do not employ
peroxyl radicals as the oxidant and do not account for both the
kinetics and stoichiometry of the radical-trapping reaction(s).
In an effort to make inhibited autoxidations as simple as the
most popular “antioxidant assays”, we have developed a spectrophotometric approach for monitoring reaction progress in inhibited
autoxidations. The approach employs easily prepared 1-phenylbutadiene-conjugated or styrene-conjugated BODIPY chromophores
(PBD-BODIPY or STY-BODIPY, respectively) as signal carriers that co-autoxidize along with a hydrocarbon substrate. We show
that inhibition rate constants (kinh) are accurately determined for a range of phenolic and diarylamine RTAs using this approach and
that mechanistic experiments, such as kinetic isotope effects and kinetic solvent effects, are equally easily carried out. Moreover,
synergistic interactions between RTAs, as well as the unconventional activity of diarylamine RTAs, are captured using this
methodology. Lastly, we show that the approach can be employed for monitoring reactions in aqueous solution.
their putative role in maintaining human health and promoting
longevity, many methods have been developed to study RTAs.
The best approaches comprise kinetic studies of inhibited
autoxidations, where reaction progress is followed either by
consumption of one of the starting materials (O2) or the
formation of products (hydroperoxides).3−5 The former
approach is preferred as it can be carried out continuously,
but it involves specialized equipment to accurately assess the
pressure differential between inhibited and uninhibited
autoxidations carried out in parallel. The latter approach is a
discontinuous assay where hydroperoxides are determined in
aliquots removed from the reaction mixture. This is usually
done by laborious iodometric titrations or time-consuming
chromatographic analyses. Recently developed fluorogenic
probes are attractive alternatives, but the fact remains that
these are discontinuous assays.6
INTRODUCTION
Autoxidation, the radical-mediated chain reaction that converts
hydrocarbons to hydroperoxides (eqs 1 and 2, Scheme 1), can
■
Scheme 1. Chain-Propagating Steps of Hydrocarbon
Autoxidation (eqs 1 and 2) and Its Inhibition by a Phenolic
Radical-Trapping Antioxidant (eqs 3 and 4)
Over the years, the desire for more accessible and/or rapid
alternatives to inhibited autoxidations has prompted the
development of a myriad of assays of “antioxidant activity”.
Some of the most common include TROLOX equivalent
antioxidant capacity (TEAC),7 ferric reducing antioxidant
power (FRAP),8 and DPPH• quenching (eq 5).9−11 These
approaches are generally based on spectrophotometric deter-
mination of the position of the equilibrium between a test
antioxidant (a reducing agent) and an oxidizing agent that
be inhibited by the addition of radical-trapping antioxidants
(RTAs). The relevant inhibition reactions for phenolic RTAs
are shown in eqs 3 and 4 (Scheme 1). The inherent reactivity
of an RTA is characterized by the rate constant for its reac-
tion with peroxyl radicals (also known as the inhibition rate
constant, kinh) since this is the reaction that competes with the
rate-determining propagation step of the chain reaction (eq 1).
An additional, but equally important, consideration is the
stoichiometry of the RTA-peroxyl reaction (n), which is
generally 2 for phenols due to the reactions in eqs 3 and 4.1,2
Given the importance of RTAs in the preservation of
petroleum-derived materials and consumer products, as well as
Received: September 16, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
Scheme 3. Synthesis of PBD-BODIPY
possesses a good absorbance and/or fluorescence.3−5 It must be
emphasized that peroxyl radicals are often not employed as the
oxidizing agent in these experiments and that they are generally
carried out in highly polar (alcoholic) solvents that are
necessary to solubilize the reagents. Such solvents are poor
models for either the confines of the lipid bilayer or petroleum-
derived materials wherein RTA activity is most important. In fact,
these reactions can be subject to kinetic solvent effects owing to
either H-bonding of the RTA or the contributions of electron-
transfer reactions from the conjugate base of the RTA.12 The
ideal approach to determine the reactivity of RTAs would
incorporate the desirable features of the above methods; it
would be rapid, operationally simple, utilize standard laboratory
equipment, and most importantly be capable of providing
accurate kinetic and stoichiometric information for reactions of
RTAs with peroxyl radicals.
phospholipid bilayer solubility, the solution-based experiments
described here were performed with the simpler BODIPY-
substituted 1-phenylbutadiene (hereafter PBD-BODIPY). This
compound is much more easily prepared (Scheme 3)16,17 and
manipulated than the commercially available (but exceedingly
expensive) C11-BODIPY581/591. The spectroscopic properties
of PBD-BODIPY have been described by others16,17 and are
highly similar to that of C11-BODIPY581/591 18 in particular, the
,
C11-BODIPY581/591 is a popular compound for monitoring
lipid oxidation in cell culture.13,14 Its fluorescence is shifted signi-
ficantly (from 595 to 520 nm) upon reaction with peroxyl
radicalspresumably as a result of reduced conjugation due to
peroxyl radical addition to the 4-phenyl-1,3-butadienyl (PBD) unit
followed by O2 addition to the resultant carbon-centered radical as
shown in Scheme 2 (although this has not been studied in detail).
absorption spectrum, which features a maximum at 591 nm and
emission spectrum with corresponding maximum at 608 nm.4
Under conditions similar to those commonly used in O2
consumption experiments, the AIBN-initiated (see Chart 1)
Chart 1. Radical Initiators Used in This Work
Scheme 2. Propagation Steps in the Autoxidation of
C11-BODIPY581/591
.
co-autoxidation of styrene (4.3 M in chlorobenzene) and PBD-
BODIPY (10 μM) at 37 °C, produced a linear increase in
fluorescence at 515 nm with time (λex = 485 nm, data not
shown). However, since PBD-BODIPY is expected to oxidize
in a similar manner to styrenevia addition of a peroxyl radical
to the phenylbutadienyl unit followed by O2 addition to the
resultant alkyl radical to afford a new peroxyl radical (as shown
for C11-BODIPY581/591 in eq 6)19the fluorescent product is
incorporated into a styrene-oxygen copolymer in solution,
precluding standardization of the signal.20 Therefore, we
elected to monitor the kinetics of PDB-BODIPY oxidation
and its inhibition by added RTAs using absorbance.
Representative examples are shown in Figure 1.
The reaction profiles in Figure 1 are strikingly reminiscent of
those obtained from inhibited autoxidations monitored by O2
consumption.3−5 Inhibition by the representative phenolic RTAs
1−4 followed the expected trend, where pyrimidinol 4 was most
reactive to peroxyl radicals, suppressing PBD-BODIPY oxidation
to the greatest extent, followed closely by 2,2,5,7,8-pentam-
ethylchroman-7-ol (PMC, 3) and then 3,5-di-tert-butyl-4-hydroxy-
anisole (2) and, last, 3,5-di-tert-butyl-4-hydroxytoluene (BHT, 1).
Although a simple comparison of the initial rates of PBD-
BODIPY consumption in the presence of the different RTAs
provides a good qualitative assessment of their relative RTA
activity, we sought to develop a method to obtain quantitative
data, i.e., the inhibition rate constant (kinh) for reaction of an
RTA with peroxyl radicals. In order to do so, the rate constant
for the competing reaction of PBD-BODIPY with peroxyl
radicals (kPBD‑BODIPY) must be known. This was easily deter-
mined by measuring its rate of consumption in co-autoxidations
with variable concentrations of PBD-BODIPY (assuming a
steady-state concentration of peroxyl radicals) via eq 7.3−5
Niki15 demonstrated that C11-BODIPY581/591 is oxidized in
azo-initiated co-autoxidations with methyl linoleate and that an
added antioxidant (2,6-di-tert-butyl-4-methylphenol, BHT) can
slow this process. Accordingly, we wondered if an assay based
upon the competitive oxidation of C11-BODIPY581/591 (or a suit-
able inexpensive analogue thereof) and an appropriate organic sub-
strate (to ensure that the process is a radical chain reaction) could
be developed to enable the accurate determination of the inhibition
rate constants (kinh) and stoichiometry (n) of reactions of RTAs
with peroxyl radicals. Our successful efforts are described below.
RESULTS AND DISCUSSION
■
Since the undecanoic acid substitution in C11-BODIPY581/591
simply provides fatty acid-like physical properties for
B
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
Doing so with 3 (e.g., using the data in Figure 1), which is
well-known to have n = 2, we obtain Ri = 2.2 × 10−9 M s−1
corresponding to ekd = 1.9 × 10−7 s−1. Although there are three
possible peroxyl−peroxyl radical termination reactions that may
occur (eqs 9−11), we can assume that eq 9 dominates since
styrylperoxyl radicals react very rapidly with each other
(kt = 2.1 × 107 M−1 s−1)23,24 and styrene is present in a
great excess compared to PBD-BODIPY.
Using these values of Ri and kt, the rate constant for the reac-
tion of PBD-BODIPY and peroxyl radicals can be calculated as
kPBD‑BODIPY = 2720 M−1 s−1. With this key value in-hand, the
data in Figure 1 can be analyzed using the well-established
kinetic expressions for inhibited hydrocarbon autoxidations,3−5
where kinh for the RTAs can be determined from the initial rate
of PBD-BODIPY consumption as in eq 12 and n from tinh as in
eq 13:51
Figure 1. Co-autoxidation of styrene (4.3 M) and PBD-BODIPY
(10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C (solid line) and
inhibited by 2 μM BHT (1), BHA (2), PMC (3), and 2-(N,N-
dimethylamino)-4,6-dimethyl-5-pyrimidinol (4). Reaction progress
was monitored by absorbance at 591 nm (ε = 139,000 M−1 cm−1).
k
PBD‐BODIPY[PBD‐BODIPY]Ri
nkinh[RTA]
−δ[PBD‐BODIPY]
δt
kPBD‐BODIPY
−δ[PBD‐BODIPY]
δt
=
=
Ri [PBD‐BODIPY]
(7)
2kt
(12)
(13)
tinhRi
n =
Thus, plotting the rate of PBD-BODIPY consumption versus
its concentration (as in Figure 2) yields a line where the slope is
given by a combination of kPBD‑BODIPY, √Ri, and √2kt (eq 7).
The rate of initiation, Ri, can be estimated from literature
data to be 2.0 × 10−9 M s−1 at 37 °C. Alternatively, it can be
determined readily under the exact experimental conditions
simply from the length of the inhibition period, tinh, of an
inhibited autoxidation with a known amount of inhibitor AH,
provided its stoichiometric number (n) is known, as in eq 8.
[RTA]
Doing so for RTAs 2−4 yields rate constants that are within
experimental error of those determined by O2 consumption (see
Table 1) and which are characterized by stoichiometric numbers
of n ∼ 2, as expected (vide supra). RTA 1 (BHT) was too slow
to effectively compete with PBD-BODIPY oxidation, precluding
determination of its corresponding kinh value (more on this later).
To demonstrate the utility of the approach for carrying out
mechanistic studies, we determined example kinetic isotope
effects (KIE) and kinetic solvent effects for the well-known
reaction of 3. The KIE for the reaction of 3 with peroxyl radi-
cals was obtained by the addition of 1% MeOH/MeOD to an
n[AH]
Ri =
tinh
(8)
Figure 2. Rate of PBD-BODIPY consumption as a function of PBD-BODIPY concentration in AIBN-initiated (6 mM) co-autoxidations of PBD-
BODIPY and styrene (4.3 M) in PhCl at 37 °C (left) and the corresponding representative raw data (right). Reaction progress was monitored by
absorbance at 591 nm (ε = 139,000 M−1cm−1).
C
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
Table 1. Inhibition Rate Constants and Stoichiometries Determined for Various RTAs Determined by Co-autoxidations of
a
PBD-BODIPY and Styrene
RTA
n
kinh (M−1 s−1
)
lit. kinh (M−1 s−1
)
1
2
3
too slow
1.8 0.2
2.0
too slow
1.4 × 104 (30 °C)24
1.1 × 105 (30 °C)24
3.8 × 106 (30 °C)24
(1.6 0.2) × 105
(3.8 0.2) × 106
(8.1 0.8) × 105
(6.6 0.4) × 105
(9.9 2.3) × 106
(4.4 0.6) × 106
too slow
b
d
3
1.9 0.1
2.0
7.5 × 105 (30 °C)
c
3
6.8 × 105 (30 °C)25
7.4 × 106 (30 °C)26
3.6 × 106 (30 °C)26
2.0 × 104 (65 °C)27
1.8 × 105 (37 °C)28
3.0 × 106 (37 °C)28
4
5
6
7
8
1.8 0.1
2.0 0.2
too slow
3.1 0.2
3.9 0.2
(3.7 0.2) × 105
(9.9 0.6) × 105
a
Determined by co-autoxidation of styrene (4.3 M) and PBD-BODIPY (10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C and inhibited by
b
1−2 μM RTA. Same as above, but containing 1% MeOD. The corresponding value for 1% MeOH was kinh = (3.6 0.2) × 106 M−1 s−1. The
literature value was determined for 1% D2O/H2O. Measured in CH3CN. Calculated from literature data.24,29
c
d
inhibited autoxidation. While MeOH had only a slight effect on the
inhibition rate constant, kinh(PhCl)/kinh(PhCl + MeOH) = 1.1,
addition of MeOD substantially lowered it, i.e. kinh(PhCl +
MeOH)/kinh(PhCl + MeOD) = kH/kD = 4.4, in good agreement
with the literature value of 5.129 and consistent with significant
O−H bond cleavage in the transition state of the reaction of
phenols with peroxyl radicals (eq 3). When the autoxidation was
carried out in acetonitrile in lieu of chlorobenzene, kinh(CH3CN) =
6.6 × 105 M−1 s−1 was obtained−a kinetic solvent effect of kPhCl
/
kMeCN = 5.8, in good agreement with the literature value of 5.6,24,25
and consistent with the existence of a hydrogen-bonded pre-
equilibrium.12
In order to further validate the current approach to assessing
RTA activity, we also sought to reproduce an example of
synergy between RTAs since this is often not picked up in
simple antioxidant assays.30 We chose an example from a recent
investigation on antioxidant synergism carried out with our
collaborators:26 the combination of a highly reactive pyridinol
RTA (5) with a much less reactive hindered phenol (2).
The data are shown in Figure 3, and are fully consistent with
results obtained by O2 consumption measurements.26 That is,
although the pyridinol is a fine RTA on its own, the hindered
phenol can only retard oxidation of PBD-BODIPY. However,
when the two RTAs are used together, their performance is
much better than their additive contributions. This arises
because the less reactive hindered phenol regenerates the more
reactive pyridinol in situ, as in eq 14. This equilibrium is
favorable since the O−H bond in pyridinol 5 is stronger than
the O−H bond in hindered phenol 2 (by 2.4 kcal/mol).26
Diarylamine RTAs were also considered in our method
validation. Representative inhibited autoxidation traces are
Figure 3. Co-autoxidation of styrene (4.3 M) and PBD-BODIPY
(10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C (solid line) and
inhibited by 2 (2 μM), 5 (1 μM), or 5 + 2 (1 μM + 2 μM). Reaction prog-
ress was monitored by absorbance at 591 nm (ε = 139,000 M−1 cm−1).
shown in Figure 4. Like BHT, diphenylamine was too slow to
effectively inhibit PBD-BODIPY oxidation. 4,4′-Dimethyldi-
phenylamine (7), representative of the industrial standard
dialkylated diphenylamines, displayed a pronounced inhibited
period, as did the recently described heteroaryl-containing
diarylamine 8.28 The rate constants obtained from fitting the
data were in good agreement with the literature values (see
Table 1).31 Interestingly, the stoichiometric numbers from
these experiments (n = 3.1 for 7 and n = 3.9 for 8) suggest that
D
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
state concentration of radicals and a concomitant increase in
the rate at which STY-BODIPY is consumed. However, slower
termination of chain-carrying radicals implies that cross-
termination between cumylperoxyl radicals and secondary
peroxyl radicals derived from STY-BODIPY will contribute
significantly to the steady-state concentration of radicals in the
autoxidation. Therefore, the direct determination of kSTY‑BODIPY
could not be carried out as we did to obtain kPBD‑BODIPY
.
Instead, kSTY‑BODIPY was determined from the inhibited
co-autoxidation of STY-BODIPY and cumene (Figure 5).
Figure 5. Co-autoxidation of cumene (3.6 M) and STY-BODIPY
(10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C (solid line)
and inhibited by 2 μM of BHT (1), (2) and diarylamines 6 and 7.
Reaction progress was monitored by absorbance at 571 nm
(ε = 128,000 M−1 cm−1).
Figure 4. Co-autoxidation of styrene (4.3 M) and PBD-BODIPY
(10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C (solid line) and
inhibited by 2 μM of 6, 7 and 8. Reaction progress was monitored by
absorbance at 591 nm (ε = 139,000 M−1 cm−1).
Since 2 is known to have kinh = 1.1 × 105 M−1 s−1 and a
stoichiometric number of n = 2,24 the rate constant could be
obtained simply by fitting the data to the standard kinetic
scheme for inhibited autoxidations modified to include the
contribution of chain propagation from the reaction of STY-
BODIPY with peroxyl radicals (see the Supporting Information
more than two peroxyl radicals are trapped by each molecule of
RTA, consistent with previous inhibited autoxidations moni-
tored by O2 consumption.28 The basis of this behavior remains
unclear, but is likely to be related to the catalytic behavior
exhibited by diarylamines at elevated temperatures.32,33
Monitoring Slower Autoxidations. Since PBD-BODIPY
reacts too quickly with chain-carrying peroxyl radicals to be able
to report on the potency of moderately reactive (and widely
used) RTAs such as BHT and diphenylamine, we sought to
develop a slightly less reactive analogue. We surmised that the
removal of one of the unsaturations in the 1-phenylbutadiene
sidechain in PBD-BODIPY should decrease its reactivity to
peroxyl radical addition significantly. We therefore prepared the
styryl analogue, STY-BODIPY, for use as the signal carrier in an
analogous co-autoxidation procedure employing cumene (kp =
0.34 M−1s−1)16
for complete details). Fitting of this data yields kSTY‑BODIPY
=
141 M−1 s−1. This rate constant indicates that STY-BODIPY is
significantly less reactive than PBD−PODIPY (recall kPBD‑BODIPY
=
2720 M−1 s−1)as expectedbut still more reactive than styrene
itself (41 M−1 s−1) and much more reactive than cumene (kp =
0.34),23 its co-autoxidation substrate.
Co-autoxidations of STY-BODIPY and cumene were well-
behaved and yielded inhibition rate constants for test RTAs
that were in very good agreement with values obtained from O2
consumption.35 As expected, the slower rate of probe oxidation
enabled the determination of rate constants for the slower
RTAs BHT (1) and diphenylamine (6), which did not yield
discernible inhibited periods in the autoxidation of the more
reactive probe, PBD-BODIPY (cf. Figures 1 and 4, respec-
tively).36 Some representative data are shown in Table 2.
Autoxidations in Water. Although PBD-BODIPY and
STY-BODIPY are highly lipophilic, the very small concen-
trations used in the styrene and cumene co-autoxidations
prompted us to consider if they would be useful for monitoring
the kinetics of inhibited autoxidations in aqueous media. Kinetic
data for the reactions of peroxyl radicals in water are relatively
sparse despite their purported importance in vivo. Most data are
Cumylperoxyl radicals undergo significantly slower termi-
nation reactions than styrylperoxyl radicals (i.e., kt = 2.3 × 104
M−1 s−1 vs. 2.1 × 107 M−1 s−1),34 resulting in a higher steady
E
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
measurements (2.4 × 106 M−1 s−1),42 a value of kSTY‑BODIPY
=
Table 2. Inhibition Rate Constants and Stoichiometries
Determined for Various RTAs Determined by
Co-autoxidations of STY-BODIPY and Cumene
7.9 × 103 M−1 s−1 was calculated.43 Using this value, the inhibition
rate constant for Trolox was measured to be kinh = 2.2 ×
105 M−1 s−1, in very good agreement with the value predicted
for α-tocopherol (4.9 × 105 M−1 s−1) using the relationship de-
veloped by Ingold and co-workers for kinetic solvent effects on
H atom transfer reactions (eq 15) and the constants k0inh = 7.2 ×
106 M−1 s−1, αH2 (α-TOH) = 0.37 and β2H (H2O) = 0.38.12,44
a
RTA
n
kinh (M−1 s−1
)
lit. kinh (M−1 s−1
)
1
2
6
7
2.2 0.1
2.1 0.1
2.5 0.1
2.4 0.1
(2.1 0.2) × 104
1.4 × 104 (30 °C)24
1.1 × 105 (30 °C)24
2.0 × 104 (65 °C)27
1.8 × 105 (37 °C)28
b
1.1 × 105
(1.7 0.2) × 104
(1.1 0.1) × 105
log kiSnh = log ki0nh − 8.3α2HβH
a
(15)
Determined by inhibited co-autoxidation of cumene (3.6 M) and
2
STY-BODIPY (10 μM) initiated by AIBN (6 mM) in PhCl at 37 °C
b
Under the reaction conditions, the stoichiometry of the
ascorbate/peroxyl reaction was found to be only ∼0.1. Although
ascorbate can, in principle, reduce two peroxyl radicals to form
dehydroascorbic acid (as in eqs 16 and 17), low stoichiometric
and inhibited by 2 μM antioxidant. Assumed in order to derive
kSTY‑BODIPY
.
derived from pulse radiolysis measurements where kinetics for
the reactions of RTAs with methylperoxyl, and more often the
highly reactive trichloromethylperoxyl, radicals have been
determined.37
Co-autoxidations of STY-BODIPY and THF as a 40%
(by volume) solution in water38 initiated by the water-soluble
azo initiator AAPH were very well behaved and consumption of
STY-BODIPY was expectedly inhibited by the water-soluble
antioxidants, Trolox (9) and ascorbate (10), see Figure 6.
numbers are commonly observed for ascorbateeither alone
(n → 0 as [ascorbate] → ∞)45 or when used in a synergistic
combination with a lipid-soluble compound such as α-TOH.46
The lower than ideal stoichiometry has been ascribed to the
autoxidation of ascorbate and/or the reaction of the ascorbyl
radical (anion) with O2 to yield a chain carrying hydroperoxyl
radical (superoxide, eq 18) in competition with the reaction of
the ascorbyl radical (anion) with another peroxyl radical.45,46
The methods described herein may prove useful in delineating
this reaction mechanism as well as others that have yet to be
studied in detail in aqueous media.
CONCLUSION
■
We have developed a novel, accurate, and operationally simple
method for the study of antioxidant reactivity by the inhi-
bited autoxidation approach. Reaction progress is monitored
continuously by following the consumption of an easily syn-
thesized BODIPY-conjugated probe by visible light spectro-
photometry. The inhibition rate constants measured for several
phenol and diarylamine RTAs using this approach are in
excellent agreement with literature values. The method de-
scribes the co-antioxidant behavior of RTA mixtures and also
the enhanced stoichiometry of suitably reactive diarylamines.
The method can be calibrated for use in solvents of different
polarities and/or H-bonding properties, allowing measurements
of kinetic solvent effects. Furthermore, deuterium kinetic
isotope effects are easily measured by the addition of 1%
MeOD/MeOH to the co-autoxidations. Overall, the approach
enables any researcher with access to a conventional UV−vis
spectrophotometer to carry out inhibited hydrocarbon auto-
xidations to obtain invaluable kinetic and stoichiometric data on
RTAs. It is our hope that this method will supplant popular
assays based on spectroscopic determinations of equilibria
between RTAs and persistent radicals or similarly dubious
assessments of “antioxidant activity”.
Figure 6. Co-autoxidation of THF (4.9 M) and STY-BODIPY
(10 μM) initiated with AAPH (1 mM) in water at 37 °C (solid line)
and inhibited by 2 μM of 9 or 20 μM of 10. Reaction progress was
monitored by absorbance at 562 nm (ε = 147,000 M−1 cm−1).
As was the case in the foregoing experiments, determination of
inhibition rate constants in this system requires knowledge of the
rate constant for reaction of STY-BODIPY with THF-derived
peroxyl radicals (kSTY-BODIPY). Accordingly, data from ascorbate-
inhibited autoxidations were fit to the standard kinetic scheme
for inhibited autoxidations of THF (kp = 4.4 M−1 s−1)39−41
modified to include the contribution of chain-propagation from
reaction of STY-BODIPY with the THF-derived peroxyl radi-
cals. Using the literature rate constant for the reaction of ascor-
bate and methylperoxyl radical obtained by pulse radiolysis
F
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
HRMS (EI - magnetic sector): calcd for C19H17BF2N2 322.14529, found
322.14342. λmax (PhCl) = 571 nm, ε = 128,141 M−1 cm−1.
EXPERIMENTAL SECTION
■
General Methods. All chemicals and solvents were purchased
from commercial suppliers and used without further purification.
Styrene and cumene were purified in batches. Styrene was extracted
three times with 1 M NaOH(aq) and washed with water before drying
with MgSO4 and filtering. It was then fractionally distilled under
vacuum, percolated through a column of silica, and stored under
nitrogen at −20 °C (it could be stored for up to 3−4 days prior to
use). It was further percolated through a column of ∼1/3 silica gel
layered on top of ∼2/3 basic alumina immediately prior to use.
Cumene was purified by the same method. Commercially available
phenolic antioxidants BHT, BHA, and PMC (1, 2, and 3) were
recrystallized prior to use. 6-(Dimethylamino)-3-pyridinol (4),
2-(dimethylamino)-4,6-dimethyl-5-pyrimidinol (5), and N2,N2-dimeth-
yl-N5-phenylpyrimidine-2,5-diamine (8) were synthesized by the
reported procedures.47,48 Diphenylamine (6) and di-p-tolylamine (7)
were purchased and used without further purification. UV−vis spectra and
kinetics were measured on a UV−vis spectrophotometer equipped with a
temperature controller unit and a thermostated 6 × 6 multicell holder.
Synthesis of PBD- and STY-BODIPY. Both compounds were
synthesized by literature procedures.16,17 Briefly, a solid mixture of
pyrrole-2-carboxaldehyde49 (1.0 g, 10.5 mmol), zinc dust (0.69 g,
10.5 mmol), and either cinnamyl triphenylphosphonium chloride or
benzyl triphenylphosphonium bromide (4.3 or 4.5 g, 10.5 mmol) was
heated to 100 °C under an inert atmosphere and stirred overnight.
The mixture was allowed to cool before it was diluted with CHCl3
(25 mL) and filtered. The filtrate was washed once with water and
once with brine before drying with MgSO4. Following filtration and
concentration under reduced pressure, the product was purified by
column chromatography on silica gel using a gradient of 20−30%
CHCl3 in hexanes.
The product phenylbutadienyl- or styryl-substituted pyrrole (0.53
or 0.46 g, 2.7 mmol) was then combined with 3,5-dimethylpyrrole-2-
carboxaldehyde (0.34 g, 2.7 mmol) in dry CH2Cl2 (135 mL) under an
inert atmosphere. POCl3 (0.27 mL, 2.9 mmol) was then added drop-
wise over several minutes, and the reaction was stirred overnight in the
dark at room temperature. Diisopropylethylamine (2.0 mL, 11.5 mmol)
was then added slowly, and the reaction was stirred for another 10 min,
followed by dropwise addition of BF3−OEt2 (1.4 mL, 11.2 mmol). The
reaction was stirred at room temperature until complete (∼1 h, as
judged by TLC), poured into a separatory funnel, and washed twice
with water and then a small amount of brine. The organic phase was
dried with MgSO4 and filtered and the solvent evaporated. Column
chromatography (100% CH2Cl2) removed the major impurities. A
second column (60% CH2Cl2 in hexanes) afforded pure product.
Due to the oxidizability of both PBD-BODIPY and STY-BODIPY,
2.0 mM stock solutions of each compound were prepared in either
1,2,4-trichlorobenzene or DMSO (depending on whether they are for
use in organic or aqueous solutions) and stored under nitrogen, frozen
at −20 °C. When stored in this manner, there was minimal back-
ground oxidation of the either compound over several months.
4,4-Difluoro-5,7-dimethyl-3-(4-phenyl-1,3-butadienyl)-4-bora-
3a,4a-diaza-s-indacene (PBD-BODIPY). Yield: 81 mg (9%). 1H NMR
(400 MHz; acetone-d6): δ 7.65 (d, J = 7.3 Hz, 2H), 7.51 (s, 1H),
7.43−7.32 (m, 5H), 7.21−7.15 (m, 2H), 7.06 (d, J = 4.7 Hz, 1H), 6.94
(d, J = 14.9 Hz, 1H), 6.30 (s, 1H), 2.58 (s, 3H), 2.33 (s, 3H). 19F
NMR (377 MHz; acetone-d6, PhCF3 standard): δ −143.16 (dd, J =
65.6, 32.8 Hz). HRMS (EI - magnetic sector): calcd for C21H19BF2N2
348.16094, found 348.15828. λmax (PhCl) = 591 nm, ε = 139162
M−1 cm−1. The poor solubility of PBD-BODIPY in either CDCl3,
DMSO, acetone, or benzene, combined with its high oxidizability,
prevented the acquisition of a well-resolved 13C NMR spectrum.
4,4-Difluoro-5,7-dimethyl-3-(2-phenylethenyl)-4-bora-3a,4a-
diaza-s-indacene (STY-BODIPY). Yield: 128 mg (15%). 1H (400 MHz;
acetone-d6): δ7.64 (t, J = 10.9 Hz, 3H), 7.52 (t, J= 8.2 Hz, 2H), 7.44
(t, J = 7.5 Hz, 2H), 7.36 (t, J = 7.3 Hz, 1H), 7.15(d, J = 4.3 Hz, 1H),
7.08 (d, J = 4.4 Hz, 1H), 6.29 (s, 1H), 2.57 (s,3H), 2.31 (s, 3H). 13C
NMR (75 MHz; acetone-d6): δ 159.2, 153.0, 143.6, 136.7, 135.3, 135.0,
128.93, 128.85, 128.71, 127.0, 123.2, 120.2, 119.0, 115.2, 14.0, 10.4. 19F
NMR (377 MHz; acetone-d6): δ −143.00 (dd, J = 69.6 Hz, 34.8 Hz).
Inhibited Co-autoxidations: Styrene/Cumene. Freshly purified
styrene or cumene (1.25 mL) was loaded into a 3 mL cuvette along
with 1.18 mL of PhCl. The cuvette was placed into the thermostated
sample holder of a UV−vis spectrophotometer and allowed to
equilibrate to 37 °C. A small aliquot (12.5 μL) of a 2.0 mM solution of
the BODIPY probe in 1,2,4-trichlorobenzene was added, followed by
50 μL of 0.3 M solution of AIBN in PhCl, and the solution was
thoroughly mixed. The absorbance at either 591 nm (PBD-BODIPY)
or 571 nm (STY-BODIPY) was monitored for 40−60 min to ensure
that the reaction was proceeding at a constant rate, after which 10 μL
of a 500 μM solution of the test antioxidant was added. The solution
was thoroughly mixed and the absorbance readings resumed.
Inhibited Co-autoxidations: THF/H2O. Unstabilized THF
(1.0 mL) was loaded into a 3 mL cuvette along with 1.43 mL of
H2O. The cuvette was placed into the thermostated sample holder of a
UV−vis spectrophotometer and allowed to equilibrate to 37 °C. A
small aliquot (12.5 μL) of a 2.0 mM solution of the BODIPY probe in
DMSO was added, followed by 50 μL of 0.05 M solution of AAPH in
H2O, and the solution was thoroughly mixed. The absorbance at
562 nm was monitored for 10−20 min to ensure that the reaction was
proceeding at a constant rate, after which 10 μL of a 500 μM solution
of the test antioxidant was added. The solution was thoroughly mixed,
and the absorbance readings were resumed.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.5b02183.
■
S
UV−vis and NMR spectra of PBD-BODIPY and STY-
BODIPY, details of kinetic modeling and parameter
fitting (COPASI), and optimized geometries and
energies for computational results (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: dpratt@uottawa.ca.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the Natural Sciences and Engineering Research Council
of Canada (NSERC), the Canada Foundation for Innovation, and
the Canada Research Chairs program for financial support. E.A.H.
thanks the Government of Ontario for award of an Ontario
Graduate Scholarship, and A.V.K. thanks NSERC for an
Undergraduate Student Research Award.
REFERENCES
(1) Ingold, K. U. Chem. Rev. 1961, 61, 563−589.
(2) Ingold, K. U.; Pratt, D. A. Chem. Rev. 2014, 114, 9022−9046.
(3) Valgimigli, L.; Pratt, D. A. In Encyclopedia of Radicals in Chemistry,
Biology and Materials; Chatgilialoglu, C., Studer, A., Eds.; John Wiley &
Sons,: Chichester, UK, 2012.
■
(4) Li, B.; Pratt, D. A. Free Radical Biol. Med. 2015, 82, 187−202.
(5) Amorati, R.; Valgimigli, L. Free Radical Res. 2015, 49, 633−649.
(6) Hanthorn, J. J.; Haidasz, E.; Gebhardt, P.; Pratt, D. A. Chem.
Commun. 2012, 48, 10141−10143.
(7) Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.;
Rice-Evans, C. Free Radical Biol. Med. 1999, 26, 1231−1237.
(8) Benzie, I. F.; Strain, J. J. Anal. Biochem. 1996, 239, 70−76.
(9) Blois, M. S. Nature 1958, 181, 1199−1200.
(10) Huang, D.; Ou, B.; Prior, R. L. J. Agric. Food Chem. 2005, 53,
1841−1856.
(11) Foti, M. J. Agric. Food Chem. 2015, 63, 8765−8776.
G
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Featured Article
(12) Litwinienko, G.; Ingold, K. U. Acc. Chem. Res. 2007, 40, 222−
230.
(13) Kang, H. C.; Haugland, R. P. Long Wavelength Chemically
Reactive Dipyrrometheneboron Difluoride Dyes and Conjugates.
US5451663 A, 1993.
(14) Drummen, G. P.; van Liebergen, L. C.; Op den Kamp, J. A.;
Post, J. A. Free Radical Biol. Med. 2002, 33, 473−490.
(15) Yoshida, Y.; Shimakawa, S.; Itoh, N.; Niki, E. Free Radic. Res.
2003, 37, 861−872.
(36) The more reactive RTAs 3−5 and 8 completely inhibited the
co-autoxidation reaction, precluding determination of the inhibition
rate constants.
(37) von Sonntag, C.; Schuchmann, H.-P. Angew. Chem., Int. Ed. Engl.
1991, 30, 1229−1253. Neta, P.; Huie, R. E.; Ross, A. B. J. Phys. Chem.
Ref. Data 1990, 19, 413−513.
(38) The selection of this system was inspired by the work of
Valgimigli and Amorati communicated at the 2014 International
Congress on Hydrogen Atom Transfer (Rome).
(39) Howard, J. A.; Ingold, K. U. Can. J. Chem. 1969, 47, 3809−3815.
(40) Howard, J. A. Adv. Free Radical Chem. 1972, 4, 49−173.
(41) The propagation rate constant in THF/H2O was assumed to be
the same as in neat THF. Ri was determined from tinh in the
autoxidation inhibited by Trolox and assuming n = 2. See ref 24.
(42) Alfassi, Z. B.; Huie, R. E.; Kumar, M.; Neta, P. J. Phys. Chem.
1992, 96, 767−770.
(43) The rate constant for the reaction of STY-BODIPY with THF-
derived peroxyl radicals is significantly higher than that of STY-
BODIPY with cumylperoxyl radicals also in chlorobenzene, where it
was measured to be 2.0 × 103 M−1s−1 (versus 141 M−1s−1 for
cumylperoxyls). The difference (13.8-fold) can be rationalized by the
greater electrophilicity of the THF-derived peroxyl radical compared
to the cumylperoxyl radical and the significant polar effects known to
operate in the transition states of radical additions to alkenes. For a
discussion, see: Russell, G. A. In Free Radicals; Kochi, J. K., Ed.; Wiley:
New York, 1973; Vol. 1, pp 275−332.
(16) Kang, H. C.; Haugland, R. P. Ethenyl-Substituted Dipyrrome-
theneboron Difluoride Dyes and Their Synthesis. US5187288, 1993.
(17) De Cremer, G.; Roeffaers, M. B. J.; Bartholomeeusen, E.; Lin,
K.; Dedecker, P.; Pescarmona, P. P.; Jacobs, P. A.; De Vos, D. E.;
Hofkens, J.; Sels, B. F. Angew. Chem., Int. Ed. 2010, 49, 908−911.
(18) Naguib, Y. M. Anal. Biochem. 1998, 265, 290−298.
(19) DFT calculations using the B3LYP/CBSB7 method indicate
that the most favorable position for peroxyl addition is adjacent to the
phenyl ring, yielding an allyl-BODIPY radical at least 7.4 kcal/mol
more stable than those radicals derived from addition of the peroxyl
radical at any of the other three positions of the 1-phenylbutadiene
moiety. See the Supporting Information for further details.
(20) Moreover, in some preliminary inhibited co-autoxidations the
fluorescence intensity was significantly suppressed even when no/
minimal inhibition was expected to take place. This was likely the
result of fluorescence quenching by some RTAs and/or their oxidation
products.
(44) Snelgrove, D. W.; Lusztyk, J.; Banks, J. T.; Mulder, P.; Ingold, K.
U. J. Am. Chem. Soc. 2001, 123, 469−477.
(45) See: Wayner, D. D. M.; Burton, G. W.; Ingold, K. U. Biochim.
Biophys. Acta, Gen. Subj. 1986, 884, 119−123 and references cited
therein.
(21) Hammond, G. S.; Sen, J. N.; Boozer, C. E. J. Am. Chem. Soc.
1955, 77, 3244−3248.
(22) Van Hook, J. P.; Tobolsky, A. V. J. Am. Chem. Soc. 1958, 80,
779−782.
(23) Howard, J.; Ingold, K. Can. J. Chem. 1967, 45, 793−802.
(24) Burton, G. W.; Doba, T.; Gabe, E.; Hughes, L.; Lee, F. L.;
Prasad, L.; Ingold, K. U. J. Am. Chem. Soc. 1985, 107, 7053−7065.
(25) Valgimigli, L.; Amorati, R.; Petrucci, S.; Pedulli, G. F.; Hu, D.;
Hanthorn, J. J.; Pratt, D. A. Angew. Chem., Int. Ed. 2009, 48, 8348−
8351.
(46) Doba, T.; Burton, G. W.; Ingold, K. U. Biochim. Biophys. Acta,
Lipids Lipid Metab. 1985, 835, 298−303. Barclay, L. R. C.; Locke, S. J.;
MacNeil, J. M. Can. J. Chem. 1983, 61, 1288−1290.
(47) Nam, T.-G.; Nara, S. J.; Zagol-Ikapitte, I.; Cooper, T.;
Valgimigli, L.; Oates, J. A.; Porter, N. A.; Boutaud, O.; Pratt, D. A.
Org. Biomol. Chem. 2009, 7, 5103−5112.
(48) Hanthorn, J. J.; Valgimigli, L.; Pratt, D. A. J. Org. Chem. 2012,
77, 6908−6916.
(49) Ryskiewicz, R. M.; Silverstein, E. E.; Willard, C. Org. Synth.
1956, 36, 74.
(50) Hanthorn, J. J.; Pratt, D. A. J. Org. Chem. 2012, 77, 276−284.
(51) Alternatively, the data in Figure 1 can be fit numerically to the
standard kinetic scheme for inhibited autoxidations3−5 modified to
include the contribution of chain propagation from the reaction of
PBD−BODIPY with peroxyl radicals in order to derive kinh for the
RTAs (see the Supporting Information for complete details). The two
approaches yield indistinguishable data.
(26) Valgimigli, L.; Bartolomei, D.; Amorati, R.; Haidasz, E.;
Hanthorn, J. J.; Nara, S. J.; Brinkhorst, J.; Pratt, D. A. Beilstein J.
Org. Chem. 2013, 9, 2781−2792.
(27) Brownlie, I. T.; Ingold, K. U. Can. J. Chem. 1966, 44, 861−868.
(28) Hanthorn, J. J.; Amorati, R.; Valgimigli, L.; Pratt, D. A. J. Org.
Chem. 2012, 77, 6895−6907.
(29) Burton, G. W.; Ingold, K. U. J. Am. Chem. Soc. 1981, 103, 6472−
6477.
(30) Equilibrium methods (such as those used in many of the most
common “antioxidant activity” assays) only reflect the reaction
stoichiometry and show only additive relationships between multiple
antioxidants.
(31) The literature value for 828 was determined using a peroxyl
radical clock approach,50 which was found to overestimate the rate
constants for some of the more reactive diarylamines relative to data
obtained from inhibited autoxidations (by a factor of 2−3).28
(32) Jensen, R. K.; Korcek, S.; Zinbo, M.; Gerlock, J. L. J. Org. Chem.
1995, 60, 5396−5400.
(33) Haidasz, E. A.; Shah, R.; Pratt, D. A. J. Am. Chem. Soc. 2014,
136, 16643−16650.
́
(34) Enes, R. F.; Tome, A. C.; Cavaleiro, J. A. S.; Amorati, R.; Fumo,
M. G.; Pedulli, G. F.; Valgimigli, L. Chem. - Eur. J. 2006, 12, 4646−
4653.
(35) Diarylamine 7 showed reduced stoichiometry in these
experiments relative to the PBD-BODIPY/styrene co-autoxidations
(n = 2.4 vs 3.1). Secondary alkyl radicals (or the alkoxyamines derived
from their combination with diarylnitroxides) are necessary for either
the retro-carbonyl-ene or homolysis/dis-proportionation reaction to
complete the catalytic cycle.27,28 Similarly, the more reactive
diarylamine 8 completely inhibited the reaction and had a
stoichiometric number, n = 2.3, also down considerably from 3.9 in
the PBD-BODIPY/styrene co-autoxidations.
H
DOI: 10.1021/acs.joc.5b02183
J. Org. Chem. XXXX, XXX, XXX−XXX