A versatile fluorescence approach to kinetic studies of hydrocarbon autoxidations and their inhibition by radical-trapping antioxidants

Article abstract of DOI:10.1039/c2cc35214a

We report a simple coumarin-triarylphosphine conjugate that undergoes fluorescence enhancement upon reaction with hydroperoxides and demonstrate its use to follow autoxidations of 7-dehydrocholesterol (at 37 °C) and hexadecane (at 160 °C) and their inhibition by antioxidants. This journal is The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2cc35214a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 10141–10143
www.rsc.org/chemcomm
COMMUNICATION
A versatile fluorescence approach to kinetic studies of hydrocarbon
autoxidations and their inhibition by radical-trapping antioxidantsw
a
b
Jason J. Hanthorn,z Evan Haidasz,z Paul Gebhardtyb and Derek A. Pratt*ab
Received 19th July 2012, Accepted 26th August 2012
DOI: 10.1039/c2cc35214a
We report a simple coumarin–triarylphosphine conjugate that
undergoes fluorescence enhancement upon reaction with hydro-
peroxides and demonstrate its use to follow autoxidations of
7-dehydrocholesterol (at 37 8C) and hexadecane (at 160 8C) and
their inhibition by antioxidants.
conjugated to phosphines and operate similarly: oxidation of the
phosphine by ROOH precludes photoinduced electron transfer
(PeT) from the phosphine to the fluorophore, leading to an
enhancement in fluorescence. Two representative examples are
diphenyl-1-pyrenylphosphine (DPPP, 1)4,5 and Spy-LHP (2).6
Autoxidation is the radical chain process responsible for the
oxidative degradation of all hydrocarbons (R–H, Scheme 1),1
from polymers, lubricating oils and other petroleum-derived
products to the lipids that comprise the biological membranes
of the cells of living organisms. Given the importance of this
process in understanding the fate of industrial and biological
materials in an aerobic atmosphere, as well as its putative
involvement in the pathogenesis of degenerative disease, much
effort has been devoted to the detailed study of these reactions.
Autoxidations are followed either by consumption of O2 or
formation of the hydroperoxide product (ROOH). The former
is generally preferred since it can be done in real time by
differential pressure measurements,2 and the latter, which are
often carried out by either iodometric titration or HPLC/GC
following reduction of ROOH to ROH with PPh3,3 are tedious and
can be difficult to reproduce with good precision. Recently, several
(pro-)fluorescent molecular probes have been developed to detect
and/or image ROOH formation in real time in solution and/or
cells. These compounds generally feature known fluorophores
We sought to use these molecules in a microplate assay to
facilitate our studies of hydrocarbon autoxidations and their
inhibition by antioxidants, but quickly learned that these
compounds (and their synthetic precursors) suffer from very
poor solubility and are generally very expensive.7 Inspired by
the work of Bertozzi and co-workers,8 Xian et al.9 developed
the more accessible and soluble coumarin-based dye 3 for the
determination of S-nitrosothiols, but unfortunately we found
that reactions of 3 with ROOH are sluggish (k o 0.1 Mꢀ1
s
ꢀ1),
and that background oxidation with O2 can be problematic. Since
triphenylphosphine reacts much more quickly with ROOH,10 we
opted to include a phenyl spacer for our investigations. Moreover,
given the ease with which simpler 7-substituted coumarins can be
synthesized (by Koneovenagel condensation or Wittig olefination
and in situ lactonization of the cinnamate), we elected to focus on
this less-electron rich fluorophore. We considered 4 substitutions
(Scheme 2) in an attempt to optimize both PeT and rate of
reaction with ROOH.
The N,N-diethylamino-substituted coumarins had relatively
poor fluorescence quantum yields for both the phosphine (4–5)
and phosphine oxide (8–9) conjugates (Ff,ox and Ff,red
,
Table 1). The erosion in quantum yield compared to the
oxidized form of 3 (Ff,ox = 0.79 ꢁ 0.04)9 presumably arises
due to the increased conformational flexibility of the N,N-
diethylamino substituent as well as the intervening phenyl
linkage. Furthermore, the lack of any significant quenching of
the coumarin’s fluorescence by PeT from the phosphine moiety
indicated a mismatching of the energies of the phosphine and
coumarin HOMOs in this system. Substituting the N,N-diethyl-
amino substituent for a methoxy group provided a sharp increase
in quantum yield, and decreased the energy of the coumarin
HOMO energy, leading to a slightly more effective quenching
of the fluorescence. However, it was upon replacement of the
Scheme 1 Radical chain mechanism of hydrocarbon autoxidation.
Assuming steady state in [ROOꢂ] the rate of autoxidation is given by:
d[ROOH]/dt = ꢀd[O2]/dt = Rp = kp[R–H]Ri1/2/2kt1/2
.
a Department of Chemistry, Queen’s University, 90 Bader Lane,
Kingston, Ontario, Canada K1N 6N5
b Department of Chemistry, University of Ottawa, 10 Marie Curie
Pvt., Ottawa, Ontario, Canada. E-mail: dpratt@uottawa.ca;
Fax: +1-613-562-5170; Tel: +1-613-562-5800 ext. 2119
w Electronic supplementary information (ESI) available: All experi-
mental procedures and raw data. See DOI: 10.1039/c2cc35214a
z These authors contributed equally to this work.
y Visting student from Westfalische Wilhelms-Universitat Munster.
¨
¨
¨
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 10141–10143 10141

View Article Online
Scheme 2 Synthesis of coumarin–triarylphosphine conjugates.17
Table Relevant spectral parameters of coumarin–phosphine
conjugates 4–7 and corresponding phosphine oxides (8–11) at 298 K
1
in MeOH
Fig. 1 Initial rates for the reaction of 7 (20 mM) with TetOOH in
MeOH (’) at 37 1C and t-AmOH (K) at 25 1C.
a
a
lex (nm)
lem (nm)
Ff,red
Ff,ox
Ff,ox/Ff,em
4, 8
5, 9
6, 10
7, 11
408
406
345
343
486
486
426
422
0.17
0.15
0.20
0.05
0.21
0.17
0.48
0.52
1.2
1.3
2.4
10.6
Scheme 1 caption). The use of compound 7 enabled us to
measure Rp in a 7-DHC autoxidation. Thus, azo-initiated15
oxidations of 7-DHC were carried out in 1,2-dichlorobenzene
in a 96-well microplate and the fluorescence of each well was
recorded at 4 minute intervals following dilution with MeOH
and addition of 7 (as a concentrated solution in CH3CN) in
order to determine [ROOH] (Fig. 2).17 The rate of ROOH
formation afforded a kinetic chain length (u = Rp/Ri) of
B400, which provided an oxidizability of 21.5 from which
a
Determined versus sodium difluorescein in methanol (Ff = 0.92).11
diphenylphosphine moiety with the more electron-rich di(p-tolyl)-
phosphine that the fluorescence enhancement upon oxidation
became useful (Ff,ox/Ff,red = 10.6). Moreover, 7 was indefinitely
stable in air at 25 1C in non-protic solvents, and underwent
negligible oxidation in protic solvents over several hours.
2kt could be calculated as 1.1 ꢃ 104 Mꢀ1 s
.
ꢀ1 18,19 These results
indicate that the high oxidizability of 7-DHC is therefore not
only due to its very high kp, but also its very low 2kt.
It is obvious from the ample kinetic data in the literature
that the rates of reactions of triarylphosphines and ROOH are
too slow for accurate determinations of ROOH concentration
based on a single ‘end-point’ fluorescence measurement (e.g.
k = 1.3, 1.0 and 0.4 Mꢀ1 sꢀ1 for reactions of n-, sec-, and tert-
BuOOH, respectively, with triphenylphosphine in EtOH at
25 1C).10,12 Therefore, in order to determine ROOH concen-
trations with greater accuracy and precision, a second order
rate constant for the reaction of 7 and a model hydroperoxide
was determined. Tetralin hydroperoxide (TetOOH) was chosen
as a representative secondary hydroperoxide (mimicking lipids
and other relevant autoxidation substrates) because it is stable
and can be recrystallized to very high purity.13 Fig. 1 shows the
initial rates for the reactions of 7 (20 mM) with TetOOH,
which were used to obtain k = 9.1 and 1.2 Mꢀ1 sꢀ1 in MeOH
at 37 1C and t-amyl alcohol (t-AmOH) at 25 1C, respectively
(see ESIw). With these rate constants in hand, ROOH con-
centrations of unknowns can be determined from the initial
rates of their reactions with 7.
It is important to point out that the oxidizability of 7-DHC
is 1100-fold higher than the polyunsaturated lipid (methyl)
linoleate and 3400-fold and 13 000-fold those of styrene and
cumene, respectively,20 the latter two of which are commonly
used as substrates in inhibited autoxidations to determine the
rate constant for the reaction of a radical-trapping antioxidant
with a peroxyl radical (kinh).2 Indeed, the autoxidation of
7-DHC could be inhibited by 2,2,5,7,8-pentamethyl-6-hydroxy-
chroman (PMHC, a truncated version of a-tocopherol, Nature’s
To demonstrate the quantitative utility of 7 to follow the
progress of hydrocarbon autoxidations, we carried out experiments
using both a representative biological lipid and industrial model
compound. 7-Dehydrocholesterol (7-DHC) was recently reported
by Porter and co-workers to be ‘‘one of the best chain-carrying
molecules that has been evaluated in free radical oxidations’’
with a propagation rate constant of kp = 2260 Mꢀ1 sꢀ1 14
.
However, the experiments used to determine kp were not
actually chain oxidations of 7-DHC, and therefore did not
allow for the determination of its oxidizability (kp/(2kt)1/2),
which requires knowledge of the rate of oxidation (Rp, see
Fig. 2 Autoxidation of 7-dehydrocholesterol (78 mM) in 1,2-dichloro-
benzene initiated with 52 mM of MeOAMVN (’), and its inhibition
with 4 mM of PMHC ( ) at 37 1C.
c
10142 Chem. Commun., 2012, 48, 10141–10143
This journal is The Royal Society of Chemistry 2012

View Article Online
of Ottawa. DAP also acknowledges the Canada Research
Chairs program.
Notes and references
1 K. U. Ingold, Chem. Rev., 1961, 61, 563–589.
2 G. W. Burton and K. U. Ingold, J. Am. Chem. Soc., 1981, 103,
6472–6477.
3 R. Jensen, S. Korcek, L. Mahoney and M. Zinbo, J. Am. Chem.
Soc., 1979, 101, 7574–7584.
4 K. Akasaka, T. Suzuki, H. Ohrui and H. Meguro, Anal. Lett.,
1987, 20, 731–745.
5 K. Akasaka, T. Suzuki, H. Ohrui and H. Meguro, Anal. Lett.,
1987, 20, 797–807.
6 N. Soh, T. Ariyoshi, T. Fukaminato, H. Nakajima, K. Nakano
and T. Imato, Org. Biomol. Chem., 2007, 5, 3762.
7 For example, Dojindo Molecular Technologies Inc. offer 2 for
$206 per mg (Mw 832.96, 90% purity).
8 G. A. Lemieux, C. L. de Graffenried and C. R. Bertozzi, J. Am.
Chem. Soc., 2003, 125, 4708–4709.
Fig. 3 Autoxidation of neat hexadecane initiated with
8 mM
9 J. Pan, J. A. Downing, J. L. McHale and M. Xian, Mol. BioSyst.,
2009, 5, 918.
TetOOH (’), and its inhibition with 1 mM of BHT ( ) at 160 1C.
10 R. Hiatt, R. Smythe and C. McColeman, Can. J. Chem., 1971, 49,
1707–1711.
11 X. F. Zhang, Q. Liu, H. Wang, Z. Fu and F. Zhang, J. Photochem.
Photobiol., A, 2008, 200, 307–313.
premier lipid-soluble radical-trapping antioxidant), and the rate of
the inhibited autoxidation could be used with the standard
formulae21 to determine kinh = 2.0 ꢃ 106 Mꢀ1
s
ꢀ1, in excellent
agreement with the literature value of 2.1 ꢃ 106 Mꢀ1 sꢀ1.2 The high
oxidizability of 7-DHC suggests that it is an excellent substrate for
kinetic studies of inhibited autoxidations, particularly for very
effective antioxidants,22,23 which are difficult with styrene.24
Although the fluorimetric determination of ROOH by reaction
with 7 is very useful for the rapid determination of autoxidation/
inhibition kinetics at ambient temperatures, its versatility is best
demonstrated by its application to follow autoxidations carried
out at elevated temperatures. These are impossible to study using
the conventional O2 uptake approach since the solutions must be
continuously oxygenated in a stirred flow apparatus to prevent
mass transfer of O2 from being rate-limiting3 and are therefore
conventionally monitored by iodometry or lengthy GC analyses
of reaction products.3,25 Enabled by 7 we were able to monitor the
TetOOH-initiated oxidation of hexadecane and its inhibition by
the common industrial antioxidant additive BHT (2,6-di-tert-
butyl-4-methylphenol) simply by removing aliquots from a stirred
flow reactor, cooling them to 25 1C, diluting them with t-AmOH,
loading them into a 96-well microplate and determining [ROOH]
in each aliquot from the increase in fluorescence intensity due to
reaction of the ROOH with 7 (added as a concentrated solution in
CH3CN).17 The results are shown in Fig. 3.
12 Soh et al. report k = 6400 Mꢀ1 sꢀ1 for the reaction of 2 and methyl
linoleyl hydroperoxides in EtOH at 37 1C.6 It is unclear why this reaction
is almost four orders of magnitude faster than the reaction of PPh3 and
sec-butyl hydroperoxide in EtOH (k = 1.0 Mꢀ1 sꢀ1 at 25 1C)9.
13 H. B. Knight and D. Swern, Org. Synth., 1954, 34, 90–91.
14 L. Xu, T. A. Davis and N. A. Porter, J. Am. Chem. Soc., 2009, 131,
13037–13044.
15 N. Noguchi, H. Yamashita, N. Gotoh, Y. Yamamoto, R. Numano
and E. Niki, Free Radical Biol. Med., 1998, 24, 259–268.
16 L. Xu, Z. Korade and N. A. Porter, J. Am. Chem. Soc., 2010, 132,
2222–2232.
17 See ESIw for experimental details.
18 The induction period (the first 8 minutes of the uninhibited
reaction in Fig. 1) is due to equilibration of the microplate to 37 1C.
19 The low 2kt, being similar to those of tertiary peroxyl radicals (e.g.
cumylperoxyl, 2kt = 4.5 ꢃ 104 Mꢀ1 sꢀ1), supports the mechanism
suggested by Porter and co-workers, wherein H9 and/or H14
abstraction followed by O2 addition to the termini of the two
resultant pentadienyl radicals generates four regioisomeric tertiary
peroxyl radicals at C5, C9 and C14 that carry on the radical chain
reaction from which the observed autoxidation products are
derived16
.
20 J. A. Howard, Adv. Free–Radical Chem. (London), 1972, 4,
49–174.
21 The inhibited autoxidation of hydrocarbons by a radical-trapping
antioxidant A–H is given by d[ROOH]/dt
{kp[R–H]Ri}/{2kinh[A–H]}, provided a chain length of >10, see
ref. 2.
=
ꢀd[O2]/dt
=
22 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.
23 J. J. Hanthorn, L. Valgimigli and D. A. Pratt, J. Am. Chem. Soc.,
2012, 134, 8306–8309.
24 Determining kinh from an inhibited autoxidation using the stan-
dard formulae21 requires that a chain reaction is maintained in the
inhibited period. Since this requires that the chain propagation
reactions of peroxyl radicals compete with their reaction with the
antioxidant, it is difficult for highly reactive antioxidants. The fact
that 7-DHC is so much more oxidizable than the typical substrates
implies that they will be better able to maintain the chain reaction.
25 J. Igarashi, R. Jensen, J. Lusztyk, S. Korcek and K. U. Ingold,
J. Am. Chem. Soc., 1992, 114, 7727–7736. These autoxidations
were self-initiated, so Ri was estimated based on [ROOH]
B 10 mM at t = 10 min. In our experiments, we know the initial
[ROOH] = 8 mM, exactly.
Using this procedure, the measured rate of hexadecane
autoxidation (7.3 ꢃ 10ꢀ5 M1/2
s
ꢀ1) is in satisfactory agreement
with that reported under similar conditions by Igarashi et al.
(15 ꢃ 10ꢀ5 M1/2
s
ꢀ1),25,26 as is the length of the inhibited period
observed with added BHT (B2500 s for 1 mM BHT vs.
B2600 s for 1.2 mM BHT).25 The capability to determine
[ROOH] at elevated temperatures is particularly useful as it
simulates conditions encountered by lubricants and polymers under
engine operating conditions or extrusion conditions, respectively.
We are in the throes of applying this approach to study the high-
temperature reactivities of our recently reported diarylamine
radical-trapping antioxidants, which are up to 200-fold more
reactive at ambient temperatures than current industry standards.23
This work was supported by the Natural Sciences and
Engineering Research Council of Canada and the University
26 The kinetics of the ROOH-initiated autoxidation of hexadecane obey
d[ROOH]1/2/dt = 12[R–H]kp{2ki/2kt}1/2 at low conversion, see ref. 25.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 10141–10143 10143