A radical-anion chain mechanism following dissociative electron transfer reduction of the model prostaglandin endoperoxide, 1,4-diphenyl-2,3- dioxabicyclo[2.2.1]heptane

Article abstract of DOI:10.1039/b809356c

The model prostaglandin endoperoxide, 1,4-diphenyl-2,3-dioxabicyclo[2.2.1] heptane (3), was investigated in N,N-dimethylformamide at a glassy carbon electrode using various electrochemical techniques. Reduction of 3 occurs by a concerted dissociative electron transfer (ET) mechanism. Electrolysis at -1.6 V yields 1,3-diphenyl-cyclopentane-cis-1,3-diol in 97% by a two-electron mechanism; however, in competition with the second ET from the electrode, the resulting distonic radical-anion intermediate undergoes a β-scission fragmentation. The rate constant for the heterogeneous ET to the distonic radical-anion is estimated to occur on the order of 2 × 10⁷ s^-1. In contrast, electrolyses conducted at potentials more negative than -2.1 V yield a mixture of primary and secondary electrolysis products including 1,3-diphenyl-cyclopentane-cis-1,3-diol, 1,3-diphenyl-1,3-propanedione, trans-chalcone and 1,3-diphenyl-1,3-hydroxypropane by a mechanism involving less than one electron equivalent. These observations are rationalized by a catalytic radical-anion chain mechanism, which is dependent on the electrode potential and the concentration of weak non-nucleophilic acid. A thermochemical cycle for calculating the driving force for β-scission fragmentation from oxygen-centred biradicals and analogous distonic radical-anions is presented and the results of the calculations provide insight into the reactivity of prostaglandin endoperoxides.

Full text of DOI:10.1039/b809356c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
A radical-anion chain mechanism following dissociative electron transfer
reduction of the model prostaglandin endoperoxide, 1,4-diphenyl-2,3-
dioxabicyclo[2.2.1]heptane
David C. Magri* and Mark S. Workentin*
Received 3rd June 2008, Accepted 19th June 2008
First published as an Advance Article on the web 28th July 2008
DOI: 10.1039/b809356c
The model prostaglandin endoperoxide, 1,4-diphenyl-2,3-dioxabicyclo[2.2.1]heptane (3), was
investigated in N,N-dimethylformamide at a glassy carbon electrode using various electrochemical
techniques. Reduction of 3 occurs by a concerted dissociative electron transfer (ET) mechanism.
Electrolysis at −1.6 V yields 1,3-diphenyl-cyclopentane-cis-1,3-diol in 97% by a two-electron
mechanism; however, in competition with the second ET from the electrode, the resulting distonic
radical-anion intermediate undergoes a b-scission fragmentation. The rate constant for the
heterogeneous ET to the distonic radical-anion is estimated to occur on the order of 2 × 107 s−1. In
contrast, electrolyses conducted at potentials more negative than −2.1 V yield a mixture of primary and
secondary electrolysis products including 1,3-diphenyl-cyclopentane-cis-1,3-diol,
1,3-diphenyl-1,3-propanedione, trans-chalcone and 1,3-diphenyl-1,3-hydroxypropane by a mechanism
involving less than one electron equivalent. These observations are rationalized by a catalytic
radical-anion chain mechanism, which is dependent on the electrode potential and the concentration of
weak non-nucleophilic acid. A thermochemical cycle for calculating the driving force for b-scission
fragmentation from oxygen-centred biradicals and analogous distonic radical-anions is presented and
the results of the calculations provide insight into the reactivity of prostaglandin endoperoxides.
Introduction
Prostaglandin endoperoxides are biologically important com-
pounds containing an O–O bond within
a strained 2,3-
dioxabicyclo[2.2.1]heptane ring structure.1 In particular, the
prostaglandin endoperoxide (PGH2) is the immediate precursor
to several classes of potent physiological regulators such as the
prostaglandins, prostacyclins and thromboxanes.2 At nanomolar
concentrations, these physiologicalregulators have manyprofound
effects including the activation of the inflammatory response, the
sensation of pain, the regulation of blood pressure, the coagulation
of blood, the induction of childbirth, and the regulation of the
sleeping and waking cycle.3
One of the biochemical pathways of PGH2, via the throm-
boxane synthase route, results in the formation of malondi-
aldehyde (MDA) and 12-hydroxyheptadecatrienoic acid (HHT).4
The mechanism is generally thought to result via b-scission
fragmentation of alkoxyl radicals following cleavage of the O–O
bond. One hypothesis is the mechanism is initiated by an inner-
sphere electron transfer (ET) from an iron-heme centre of an
enzyme.2 It is also plausible that MDA and HHT may result from
this type of dissociative ET to the O–O to yield a distonic radical-
anion, a reactive intermediate possessing a spatially separated
alkoxyl radical and alkoxyl anion, which may react via a b-scission
fragmentation.5–8
Radical-anions are gradually becoming accepted as biologically
important species.9 These reactive intermediates, possessing both a
charge and radical character, typically result from the transfer of a
single electron to a neutral molecule without any bonds breaking.
With endoperoxides the electron is usually accepted into the r*
orbital largely associated with the O–O bond, which cleaves to
yield the distonic radical-anion.10 Most radical-anion rearrange-
ments undergo a bond fragmentation that results in formation of
a single species containing a spatially separated radical and anion
[eqn (1)].11–14 Recently, we reported an uncommon fragmentation,
where a spatially separated radical-anion undergoes C–C bond
fragmentation to yield a localized radical-anion and a neutral
molecule [eqn (2)].5
(1)
(2)
In that study, we reported the heterogeneous ET reduction of the
bicyclic endoperoxides 1,4-diphenyl-2,3-dioxabicyclo[2.2.2]oct-5-
ene (1) and 1,4-diphenyl-2,3-dioxabicyclo[2.2.2]octane (2).5 Using
Department of Chemistry, The University of Western Ontario, London,
Ontario, Canada N6A 5B7. E-mail: mworkent@uwo.ca; Fax: (+1) 519-
661-3022; Tel: (+1) 519-661-2111 ext 86319
3354 | Org. Biomol. Chem., 2008, 6, 3354–3361
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
electrochemical techniques we evaluated previously unknown
kinetic and thermochemical data, delineated the reduction mech-
anism and provided insight into the fragmentation chemistry of
neutral biradicals and analogous distonic radical-anions. Among
many findings, we discovered that b-scission fragmentation of
the distonic radical-anion from 2 initiates an unprecedented
propagating radical-anion chain mechanism.
provide an estimate of the rate constant for ET from the electrode.
Furthermore, using a thermochemical cycle to calculate the
driving force for b-scission fragmentation from oxygen-centred
biradicals and analogous distonic radical-anions, including those
derived from 2,3-dioxabicyclo[2.2.1]heptane (4) and PGH2, we
provide new insight into the reactivity of prostaglandin endoper-
oxides.
Results
Synthesis
Compound 3 was synthesized according to Scheme 1. Ethyl 3-
benzoylpropionate from the esterification of 3-benzoylpropionic
acid,23 was reacted with acetophenone in dry toluene in the pres-
ence of excess sodium tert-butoxide at −78 ◦C. After 24 hours the
solution was extracted with water and warmed at 70 ◦C resulting
in the in situ loss of CO2 and precipitation of 1,4-diphenyl-1,3-
cyclopentadiene. Photo-oxygenation of the diene by irradiation of
O2 saturated solutions of 1 : 1 benzene–dichloromethane, contain-
ing a catalytic amount of meso-tetraphenylporphyrin (MTTP),
afforded 1,4-diphenyl-2,3-dioxabicyclo[2.2.1]hept-5-ene.24,25 The
saturated endoperoxide, 3, was prepared from the reduction
of 1,4-diphenyl-2,3-dioxabicyclo[2.2.1]hept-5-ene with excess di-
imide generated in situ from potassium azodicarboxylate in the
presence of acetic acid.22,26 The final product 3 was recrystallized
from ethanol as a crystalline white solid. The spectroscopic
characterization can be found in the Experimental section.
Thus far, we know of three common competing processes
available to the distonic radical-anion after ET reduction of
the O–O bond at an electrode: heterogeneous reduction, b-
scission fragmentation, and O-neophyl (1,2-phenyl migration)
fragmentation.5–8,15–17 In all electrochemical studies to date the
dominant pathway is reduction of the O–O bond to the distonic
radical-anion [eqn (3)] followed by reduction and protonation
to yield the cis-diol [eqn (4)]. However, in the electrochemical
studies of various diphenyl-substituted endoperoxides, the distonic
radical-anion undergoes a fragmentation reaction in competition
with the second heterogeneous ET.5–8
(3)
(4)
The fragmentation and rearrangement reactions of radical-
anions are often assumed to be similar to neutral radicals.
Recent work, mainly by Tanko and co-workers,5,11,12 has pro-
vided evidence that this assumption is not valid as both charge
and spin are contributing parameters governing the reactivity
of radical-anion rearrangements. This distinction is of great
importance considering many competing pathways, including
disproportionation, reduction and fragmentation, result in the
decomposition of prostaglandin endoperoxides. In particular,
the bridgehead hydrogen atoms are vulnerable to base-induced
decomposition.18,19 As a consequence, compounds that are readily
available, chemically stable and structurally simpler have routinely
been used as models.19,20 The common strategy has been to
substitute the hydrogen atoms at the bridgehead carbons of the
prostaglandin nucleus with phenyl rings, which prohibits base-
induced disproportionation and greatly increases the stability.18,21
In this paper, we study the bicyclic endoperoxide 1,4-diphenyl-
2,3-dioxabicyclo[2.2.1]heptane (3), as a model of a prostaglandin
endoperoxide.20–22 The compound was studied using a number of
electrochemical techniques including cyclic voltammetry, convo-
lution potential sweep voltammetry and preparative electrolysis
in order to elucidate the ET reduction mechanism and evaluate
previously unknown thermochemical parameters, notably the O–
O bond dissociation energy and the standard reduction potential.
The distonic radical-anion resulting from the concerted dissocia-
tive ET reduction of the O–O bond is observed to undergo a
competing b-scission fragmentation with direct reduction from
the electrode, and as observed previously with 2, initiates a
homogeneous propagating radical-anion chain mechanism. By
exploiting the dual reactivity of the distonic radical-anion, we
Scheme 1 The synthetic route and conditions used in the synthesis of the
bicyclic diphenyl-substituted endoperoxide 3. (a) NaOtBu, toluene, −78 ◦C
(b) ethyl 3-benzoylpropionate; D (c) O2, MTTP, hv, 1 : 1 benzene–CH2Cl2
(d) N2H2, CH2Cl2.
Cyclic voltammetry
The electrochemical reduction of 1,4-diphenyl-2,3-dioxabicyclo-
[2.2.1]heptane was studied by cyclic voltammetry using a glassy
carbon electrode in N,N-dimethylformamide (DMF) containing
0.10 M tetraethylammonium perchlorate (TEAP). The voltam-
mogram of 3, as shown in Fig. 1, is characterized by an electro-
chemically irreversible cathodic peak at all scan rates investigated
between 0.1 to 50 V s−1.
The peak potential, Ep, at 0.1 V s−1 is located at −1.34 V versus
SCE. At slow scan rates the peak width at half height, DEp/2
=
Ep/2 − Ep is uncharacteristically narrow with a peak width of
116 mV at 0.1 V s−1. Increasing the scan rate, v, is found to shift the
Ep negatively by 132 mV per log decade, resulting in the cathodic
wave broadening up to 187 mV at 10 V s−1. The transfer coefficient
determined from the peak widths using a = 1.857RT/(FDEp/2) are
found to decrease with v from 0.41 to 0.26. From the scan rate de-
pendence of the peak potential and a = 1.15RT/F[dEp/(dlog m)],
an average a of 0.23 is determined. The a values are consistent
with a concerted dissociative ET mechanism.27,28 On scanning back
after the forward reduction in the positive direction, poorly defined
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3354–3361 | 3355
©

View Article Online
Table 1 Cyclic voltammetry data for the dissociative ET reduction of 1,4-
diphenyl-2,3-dioxabicyclo[2.2.1]heptane in 0.10 M TEAP–DMF at 25 ◦
measured at a glassy carbon electrode
C
m/Vs−1
3
3 (acid)c
Ep/Va
0.1
1.0
10
0.1
1.0
10
−1.34
−1.52
−1.64
116
167
187
−1.50
−1.61
−1.73
183
198
222
DEp/2/mV
Fig. 1 Cyclic voltammograms of 1.7 mM of 3 in DMF containing
0.10 M TEAP in the absence (solid line) and presence (dashed line) of
10 equivalents of 2,2,2-trifluoroethanol at 0.2 V s−1. The inset highlights
the oxidative dip in the absence of acid.
a = 1.857RT/FDEp/2
0.1
1.0
10
0.41
0.29
0.26
0.26
0.24
0.22
a = 1.15RT/F(dEp/dlog m)
(dEp/dlog m)/mV−1b
n @ −1.6 V(no acid)
n @ −1.6 V(acid)
−132
0.23
1.97
nad
0.8
−116
0.25
nad
1.92
ndd
anodic peaks are observed between 0 and −0.8 V presumably
due to oxidation of the dialkoxide anion.15 When the switching
potential is extended past the end of the initial wave, which
is normally dictated exclusively by diffusion, an oxidative dip
is observed at −1.9 V. This feature of the voltammograms is
most noticeable at slower scan rates. Following the dip at more
negative scanned potentials there are other cathodic and anodic
peaks attributed to products resulting from the electrochemical
reduction of the 3. Repetitive cycling voltammetry was used to
probe the peaks in the vicinity of the dip as shown in Fig. 2. Two
redox couples are observed at −1.83 V and −2.00 V.
n @ −2.5 V
a Potentials are referenced versus ferrocene (0.475 V) versus SCE. b Based
on scan rates between 0.1 and 10 V s−1 c 10 mM 2,2,2-trifluoroethanol.
.
d na = not applicable, nd = not determined.
Constant potential electrolyses
At the potentials between −1.5 and −2.0 V in the absence of
acid, 3 consumes 1.97 mole equivalents of charge. The major
product is the cis-diol, 1,3-diphenyl-cyclopentane-1,3-cis-diol in
97% yield. In the presence of TFE, 1.92 F mol−1 of charge
is consumed, however the yield of cis-diol is quantitative. In
contrast, electrolysis at −2.5 V with no additional acid results
in coulometric values of only 0.8 F mol−1 charge consumed, and
the appearance of electroactive products at the expense of the
dissociative wave. Work-up of the electrolysis mixture revealed
several products by gas chromatography, as shown in Scheme 2,
including cis-diol (35%), 1,3-diphenyl-1,3-propanedione (10%),
trans-chalcone (15%) and 1,3-diphenyl-1,3-hydroxypropane (40%)
All compounds were confirmed by mass spectrometry.
Fig. 2 Cyclic voltammograms of 2.1 mM of 3. The dashed-line is the
initial reduction of the endoperoxide, followed by repetitively cycling
through the redox couples twice between the switching potentials of −1.6
and −2.3 V at 1.0 V s−1
.
A striking observation is that on addition of weak non-
nucleophilic acids there is a significant change in the voltammetry,
notably the initial dissociative wave, as illustrated by the dashed
line in Fig. 1. Table 1 compares the diagnostic voltammetry
parameters for the initial wave with and without excess acid. Weak
acids were chosen that are not electroactive within the potential
range investigated. In the presence of 5 equivalents of a non-
nucleophilic proton donor, such as 2,2,2-trifluoroethanol (TFE)
or acetanilide, the initial wave is much broader with a peak width
of 183 mV, the latter corresponding to an a value of 0.26. A
10% increase in the peak height, Ip is also observed. Addition of
more than five equivalents of TFE has no additional effect on
the initial peak height, while the peaks beyond the dissociative
wave continue to increase. Furthermore, in the presence of excess
acid the oxidative dip is not observed even at the slower scan
rates. It should be emphasized that reduction of the acid does not
occur within the potential window, and therefore, does not directly
contribute to the observed increase in peak current values.
Scheme 2 The products and yield from the reduction of 3 at −2.5 V.
Heterogeneous kinetics and thermochemical parameters
The heterogeneous ET kinetics and thermochemical parameters
for reduction of the O–O bond were evaluated by convolution
potential sweep voltammetry as shown in Fig. 3.29 The experiments
were performed in the presence of excess weak acid as this made
the voltammetry more reproducible. Only scan rates of 1.0 V s−1
and faster were usable because of non-Cottrell behaviour at the
slower scan rates, which results in lower limiting currents. A total of
11 background-subtracted cyclic voltammograms were recorded
between 1.0 and 20 V s−1. The background subtracted i–E curves
were transformed into sigmoidal-shaped I–E curves by use of the
convolution integral,29,30 and found to be in good agreement with
an error less than 3%. From the limiting currents, Ilim, and the
3356 | Org. Biomol. Chem., 2008, 6, 3354–3361
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
The E◦diss is related to the O–O bond dissociation energy (BDE)
by the equation E◦ = E◦
− BDFE/F derived from a
•
diss
ORRO–
thermochemical cycle where E◦
is the reduction potential
•
ORRO–
of the distonic radical-anion, F is the Faraday constant and the
BDFE is the bond dissociation free energy, which is the BDE
corrected for entropy (BDFE = BDE − TDS).10,31 The E◦
•
ORRO–
was approximated from the standard potential of the cumyl
alkoxyl radical equal to −0.12 V.32 Using the equation based on
E◦ the BDFE is 10 kcal mol−1.
diss
The BDE was determined using Save´ant’s concerted dissociative
ET model27,28 from the experimental DGo and the expression
ꢀ
ꢀ
DGo = (BDE + ksolvent + kinner)/4 where ksolvent is the solvent
reorganization energy and kinner is the inner reorganization energy
due to changes in the bond lengths and bond angles. An effective
radius approach was used to account for solvation of only the
portion of the molecule receiving the charge, while the rest of
the molecule acts to shield the charge from the solvent.33 The
Fig. 3 (a) Background-subtracted linear sweep voltammograms of
1.8 mM of 3 in DMF containing 0.10 mol L−1 TEAP (from left to right:
1.0, 2.0, 4.0, 10, 20 V s−1); (b) the convolution curves as a function of
scan rate (c) overlapping potential dependence of the log khet and (d) the
empirical relation ksolvent = 55.7/reff was used where ksolvent is given
27
in kcal mol−1 and reff is the effective molecular radius, given in A.
r
˚
potential dependence of aapp at 10 scan rates between 1.0 and 20 V s−1
.
eff is equal to 3.0 A resulting in ksolvent equal to 18 kcal mol−1. Using
˚
the ksolvent value (temporarily assuming kinner = 0) the BDE would
known area of the glassy carbon electrode, the diffusion coefficient
in the presence of 0.1 M TEAP was determined to be 6.5 ×
10−6 cm2 s−1. From the potential dependence of a for a concerted
dissociative ET mechanism (Fig. 3d), the E◦diss was determined by
extrapolation of the a–E data to a = 0.5, and the intrinsic barrier,
be 29 kcal mol−1.
A comparison of the kinetic and thermodynamic parameters
of 3 with our previous studies of 1 and 2 reveals some interesting
trends. The standard heterogeneous rate constant, k◦het, and stan-
dard reduction potential, E◦diss, are identical within experimental
error. The diffusion coefficient of 3 is slightly lower as might be
expected for a carbon skeleton one methylene unit less. However,
there is one noticeable difference. The intrinsic barrier of 3 is
much larger suggesting a BDE of 29 kcal mol−1 compared to
the average value of 20 kcal mol−1 determined for 1 and 2.5 This
difference may be attributed to a significant kinner contribution
from the relief of strain upon fragmentation of the O–O within the
bicyclo[2.2.1]heptane framework. The kinner is often assumed to be
negligible with respect to the magnitude of the BDE.27 However,
the BDE in endoperoxides are rather small. Considering the BDE
of 3 is expected to be similar to 1 and 2 on the order of 20 kcal
mol−1, kinner may contribute 9 kcal mol−1 to the intrinsic barrier.
ꢀ
DGo=, was determined from the slope of the linear regression.
In Table 2 the parameters determined from the convolution
analysis are summarized. As determined from the a–E plot, the
evaluated E◦ and DGo of 1 are −0.54 V and 11.8 kcal mol−1,
ꢀ
diss
respectively. The E◦diss is substantially more positive than the Ep by
over 0.8 V. The large overpotential observed in the cyclic voltam-
ꢀ
mograms is attributed to the large DGo that must be overcome
to stretch and break the O–O bond. The log k◦ is considerably
het
low with a value of −6.8 consistent with a totally irreversible
rate-determining electrode reaction characteristic of concerted
dissociative ET. The accuracy of the evaluated thermochemical
parameters in acidic media were adequately reproduced by digital
simulation of the dissociative waves using the data in Tables 1 and
2 for a two-electron concerted dissociative mechanism.
Discussion
The voltammetry of 3 is consistent with reduction of the O–O bond
by a concerted dissociative ET mechanism, where fragmentation
occurs simultaneously with electron uptake [eqn (3)]. At a
potential of −1.5 V, the major product is the corresponding 1,3-
diphenyl-cyclopentane-cis-1,3-diol in 97% yield. However, at more
negative potentials, in addition to the cis-diol, a number of other
compounds are identified: 1,3-diphenyl-1,3-propanedione, trans-
chalcone and 1,3-diphenyl-1,3-hydroxypropane.
Scheme 3 depicts the ET reduction mechanism of 3 considering
the electrode is the only electron source and in the absence of
the competing b-scission fragmentation, which accounts for the
majority of the 97% yield of 1,3-diphenyl-cyclopentane-cis-1,3-
diol. The near quantitative formation of the cis-diol 3c, results in
the consumption of two electrons per molecule. The first electron
is accepted into an r* orbital mainly localized on the O–O bond,
which in concert cleaves resulting in the formation of the distonic
radical-anion 3a with a negative charge on one oxygen atom and
an unpaired electron on the other. Subsequently, 3a generated in
Table 2 Data acquired by convolution sweep potential voltammetry of 3
in 0.10 mol L−1 TEAP–DMF at 25 ◦C at a glassy carbon electrode
Da/cm2 s−1
diss
6.5 × 10−6
E◦ b/V
−0.54
log (k◦het/cm s−1
)
−6.8
11.8
18
20
10
c
ꢀ
DGo=d/kcal mol−1
khete/kcal mol−1
BDEg/kcal mol−1
BDFEh/kcal mol−1
a Determined from the convoluted limiting current and the area of the
electrode using ferrocene with a diffusion coefficient 1.13 × 10−5 cm2 s−1
.
b Estimated error is 0.10 V and uncorrected for the double layer.
c Obtained by interpolation of the heterogeneous kinetics from a quadratic
fit at E◦diss. Estimated error of
0.5 from digital simulation. d DGo
ꢀ=
ꢀ=
determined from the slope of aapp vs. E plots and DGo = F/[8(da/dE)].
e
˚
khet = 55.7/reff; the calculated radius rAB of 4.6 A was determined from the
Stokes–Einstein equation and the diffusion coefficient. f Effective radius
g DGo = (BDE + ksolvent + kinner)/4
ꢀ
˚
of 3.0 A from reff = rB(2rAB − rB)/rAB
.
with kinner = 9 kcal mol−1
.
h BDFE = 23.06(E◦
− E◦diss).
• ORRO–
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3354–3361 | 3357
©

View Article Online
of 1,3-diphenyl-cyclopentane-cis-1,3-diol and 1,3-diphenyl-1,3-
propanedione, a radical-anion chain mechanism is postulated.5,7
Scheme 4 depicts two of the possible fates for 3a. As already
mentioned with respect to Scheme 3, the predominant pathway,
occurring in the vicinity of the electrode, is reduction of 3a to
the dialkoxide 3b, which is protonated in solution or upon work-
up to yield the cis-diol 3c. Alternatively, 3a could be protonated
in the reaction media, by 1,3-diphenyl-1,3-propanedione or upon
addition of an external acid, to yield the alkoxyl radical 3d, which
could be further reduced and subsequently protonated to yield 3c.
The formation of 1,3-diphenyl-1,3-propanedione, a product
with a lower molecular weight than the endoperoxide, results from
the b-scission fragmentation of 3a, which competes with reduction
from the electrode. The proposed mechanism in Scheme 4 also
accounts for the observed dip after the dissociative wave, the affect
of excess weak acid and the potential dependence on the electron
stoichiometry.
Scheme
3
The major mechanistic pathways for formation of
1,3-diphenyl-cyclopentane-1,3-cis-diol 3c by reduction of 3 at a glassy
carbon electrode.
At a potential of −1.6 V, 3% of 3a reacts by a competitive
b-scission fragmentation to yield 3f and ethylene via formation
of the ketyl radical 3e, a radical-anion with the spin and charge
no longer spatially separated. The radical-anion 3e, formed as a
consequence of the fragmentation, is a potential homogeneous
ET donor capable of reducing 3. The standard potential of 1,3-
diphenyl-1,3-propanedione 3f was measured to be −1.83 V versus
SCE as illustrated in Fig. 2. This is 1.3 V more negative than the
the vicinity of the electrode, could be reduced to the dialkoxide 3b
before diffusing away from the electrode, and then protonated in
the reaction solution or upon work-up to yield 3c. Alternatively,
3a could be protonated by a weak acid to yield the alkoxyl
radical 3d, which could be further reduced by the electrode before
subsequently being protonated to yield 3c. Either route from 3a
yields a near quantitivative amount of 3c. Due to electrostatic
interactions, the reduction potential of 3a is likely to be slightly
more negative than 3d, the latter is easily approximated to have a
value similar to the cumyl alkoxyl radical with a standard potential
of −0.12 V versus SCE in DMF.32 However, with the Ep located at
a potential of −1.5 V, the reduction of both 3a and 3d are predicted
to be thermodynamically favorable by at least 30 kcal mol−1.
The proposed ET reduction mechanism of 3, including the
b-scission fragmentation, is depicted in Scheme 4. To account
for the experimental observations, including the formation
E◦ of 3. Therefore, the homogeneous reduction of 3 and 3a are
diss
both thermodynamically favourable by 30 kcal mol−1.
In Fig. 1 the cyclic voltammogram displays an oxidative dip
after the dissociative wave rather than a gradual decrease in
current dictated exclusively by diffusion. The dip occurs just
after the standard potential of 3f at −1.83 V. This feature in
the voltammogram, most noticeable at slower scan rates, is a
consequence of the competitive b-scission fragmentation. The
sudden drop in current reflects a decrease in the amount of 3
diffusing towards the electrode as the endoperoxide is intercepted
by 3e and homogeneously reduced to generate 3a and 3f. Any 3f
could be reduced by the electrode surface and react with more
incoming 3 thus propagating a chain reaction.
As shown in Scheme 5, excess weak acid affects the reduction
mechanism of 3 in two ways: i) by protonation of the distonic
radical-anion 3a and ii) by protonation of the radical-anion
3e. In the first case, the electrolyses of 3 in the presence of
10 mM TFE at potentials between −1.4 to −1.8 V, no 3f was
detected. The only isolated product was the cis-diol 3c. This
observation suggests that protonation of 3a hinders the b-scission
fragmentation of the alkoxyl radical 3d so that reduction from the
Scheme 4 The propagating radical-anion chain mechanism proposed
for the ET reduction of 3. 1,3-diphenyl-1,3-propanedione 3f acts as a
homogeneous electron donor at a reduction potential of −1.83 V.
Scheme 5 b-Scission fragmentation of the distonic radical-anion 3a to
the radical-anion 3e fragments faster than the alkoxyl radical 3d to the
carbon radical 3g.
3358 | Org. Biomol. Chem., 2008, 6, 3354–3361
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 3 Thermodynamic data for fragmentation of neutral biradicals and
electrode is quantitative. This observation supports the hypothesis
that both charge and spin affect the rates of radical-anion
rearrangements.11,12 The b-scission fragmentation of the alkoxyl
radical 3d must, therefore, be slower than b-scission fragmentation
of the distonic radical-anion 3a. Excess acid also affects the
reduction mechanism by protonating the radical-anion donor 3e,
which prevents homogeneous reduction of 3.
distonic radical-anions of select endoperoxides including PGH2
Thermochemical
parameters
2a
3
4
PGH2
DH◦
12.60
85.19
29.1
77.15
6.1
10.0
16.1
−16.1
25.08
−23.6
23.0d
−161.3
198.6
−220.5
23.0d
b
f(ROOR)
DS◦
b
(ROOR)
DG◦
−12.8
f(ROOR)
The applied potential also plays a role on the reduction
mechanism of 3. At potentials between −1.4 V and −2.0 V,
approximately 2 F mol−1 of charge is consumed during the
electrolysis. However, at −2.5 V only 0.8 F mol−1 of charge
is required for complete reduction of the endoperoxide. The
reasoning for this difference in the charge consumption can be
explained by the propagating radical-anion chain mechanism
shown in Scheme 4. The distonic radical-anion 3a can either be
reduced to the dialkoxide or fragment to yield radical-anion 3e,
which can act as a homogeneous ET donor to 3 yielding 3f and
another molecule of 3a. For every 3a generated, one molecule of 3f
can theoretically result to propagate the catalytic cycle. However,
1,3-diphenyl-1,3-propanedione 3f is not a robust catalyst as it
could be reduced to the radical-anion 3e and then protonated
by the alpha protons of 3f by a competing self-protonation
mechanism.34 However, at an applied potential of −2.5 V both
aryl ketones could be reduced to yield the dianion, which can
act as a two-electon donor, which inhibits the self-protonation
mechanism provided the heterogeneous ET to the radical-anion
3e is faster than the proton transfer from 3f. Efficient propagating
chain reactions with stable homogeneous donors require only
0.1 F mol−1 of charge.7,35,36 The competing reduction and self-
protonation mechanisms of 3f contribute to the higher electron
stoichiometry of 0.8 F mol−1 at −2.5 V as we similarly observed
with 2.5
BDFE
9.0
DG◦
−3.8
−0.6
−197.5
−89.6
−98.0
−4.6
e
f(• ORRO• )
DG◦
−73.4
−73.1
−2.8
−68.2
−87.8
−2.8
−89.6
−92.5
−4.6
b
f(diketone)
f
•
•
DG2frag( ORRO )
g
•
•
•
DG1( ORRO–/ ORRO )
h
•
DG3(C
=
−45.4
−42.2
−69
−69
O/( C–O–)
i
•
DGfrag( ORRO–)
−30.5
−48.4
−28.1
−33.6
a From reference 5. b Values at 298 K in kcal mol−1 except for DS◦
(ROOR)
in cal K−1 mol−1; group additivity estimates from the NIST Structures
and Properties Database, reference 37. c Estimated from reference 37 using
hydrocarbon models without O a◦toms and corrected using values from
reference 38. d BDFE = 23.06(E
− E◦diss); E◦ from reference
• ORRO–
diss
6. E◦
approximated as −0.20 V for 4 and PGH2 from reference
• ORRO–
32. e DG◦
= BDFE + DG◦
.
f Free energy for fragmentation
f(• ORRO• )
f(ROOR)
of the biradical: DG2frag( ORRO ) = DG◦
+ DG◦
− DG◦
;
;
•
•
•
•
f(diketone)
f(alkene)
f( ORRO )
DG◦f(alkene): DG◦
= −3.5 kcal mol−1; DG◦
= −205.9 kcal mol−1
f(ethylene)
f(HHT)
from reference 37. g DG1( ORRO–/ ORRO ) = FE1( ORRO–/ ORRO ); E1( ORRO–/ ORRO )
•
•
•
•
•
•
•
•
•
estimated from the value for the cumyloxy radical E◦
= −0.12 V;
•
• ORRO–
E◦
=−0.20 V reference 32. h DG3(C
= −nFE(C O/ C–O–); n = 1,
• ORRO–
O/ C–O–)
•
=
=
•
•
•
E2(C
= −1.97 V; E3(C
= −1.83 V; E4(C
= −3.0 V.
=
=
=
=
O/ C–O–)
O/ C–O–)
O/( C–O–)
i DGfrag( ORRO–) = DG1( ORRO–/ ORRO ) + DG2frag( ORRO ) − DG3(C
•
•
•
•
•
•
•
O/( C–O–).
cycle, the overall driving force for the fragmentation of the distonic
•
radical-anion DGfrag(
is given by eqn (5):
ORRO–)
•
•
•
• •
+ DG2frag(
•
ORRO )
DGfrag(
= DG1(
ORRO–)
ORRO–/ ORRO
)
•
− DG3(C
(5)
=
O/( C–O–)
•
•
•
•
and DG3(C
=
O/( C–O–)
The DG1(
values are calculated
ORRO–/ ORRO
)
from the oxidation potential of the distonic radical-anion and the
reduction potential of the ketone, respectively, from electrochem-
Thermochemical cycle
ically measurements or literature values.32 For DG2frag(
•
•
we
)
ORRO
Scheme 6 illustrates a thermochemical cycle that identifies three
parameters contributing to the driving force for fragmentation of
the distonic radical-anion:13 (i) the thermodynamic stability of the
delocalized distonic radical-anion as measured by the oxidation
potential of the biradical, (ii) the fragmentation of the biradical,
and (iii) the stability of the localized radical-anion as measured by
the reduction potential.
also require the Gibbs energies of formation of the biradical. It
was calculated according to eqn (6)
DG2frag(
= DG◦
+ DG◦
− DG◦
(6)
•
•
•
•
ORRO
)
f(diketone)
f(alkene)
f( ORRO )
where DG◦f(diketone), DG◦
and DG◦
are the Gibbs free
•
•
f(alkene)
f( ORRO
)
energies for the diketone, alkene and the biradical, respectively.
The Gibbs free energy for the biradical was calculated from
the free energy of the endoperoxide,37,38 and the experimentally
determined BDFE according to DG◦
) = BDFE + DG◦
.
•
•
f( ORRO
f(ROOR)
Table 3 summarizes the thermochemical values pertaining to the
previously studied diphenyl endoperoxide 2,5 and new data for 3,
4 and PGH2.
The enthalpies and entropies in Table 3 are directly from the
NIST database.37 The DH◦
of the bridgehead diphenyl-
f(ROOR)
substituted endoperoxides 2 and 3 are positive values in contrast to
the unsubstituted endoperoxides 4 and PGH2, which are negative,
while the DS◦(ROOR) tends to increase with the number of atoms and
degrees of freedom within each molecule. The DG◦
of 3 is a
f(ROOR)
positive value in contrast to the other three endoperoxides, whereas
the DG◦
of PGH2 is an order of magnitude more negative
Scheme 6 Thermochemical cycle for evaluating the driving force for
fragmentation of the distonic radical-anion 3a.
f(ROOR)
than 2 and 4. Comparison of the DG◦
•
•
for 3 and PGH2
)
(
ORRO
To simplify the approach, no correction has been made to
account for any possible interaction between the radical and anion
portions of the distonic radical-anion. From the thermochemical
predicts correctly that phenyl rings at the bridgehead positions
provide stability to the strained 2,3-dioxabicyclo[2.2.1]heptane
ring structure.18,21
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3354–3361 | 3359
©

View Article Online
From eqn (5), based on the thermochemical cycle in Scheme 6,
recently reported the first example of a competitive concerted-
stepwise dissociative ET mechanism for antimalarial G3-factor
endoperoxides,41 which potentially opens up an alternative mode
of action. Novel endoperoxides that react by a radical-anion
chain mechanism may be useful in designing antimalarial prodrug
models.42 Perhaps, homogeneous radical-anion chain mechanisms
may also be an unrealized mechanistic pathway in the reactivity
of prostaglandin endoperoxides.
the calculated DGfrag( ORRO–) for 3a is −48.4 kcal mol−1 in agreement
•
with the observed b-scission fragmentation from the distonic
•
radical-anion. Similarly, DGfrag(
for 4 and PGH2 are −28.1
ORRO–)
and −33.6 kcal mol−1 suggesting that b-scission fragmentation
from their corresponding distonic radical-anions is also thermo-
•
•
for the
)
dynamically favourable. From eqn (6), the DG2frag(
ORRO
biradicals of 2 and 3 are favourable with values of −73.1 and
−87.8 kcal mol−1. This is consistent with the observed b-scission
fragmentation products from the thermolysis and photolysis of
saturated bicyclic endoperoxides.1,39 The calculations also reveal
that the driving force for biradical b-scission fragmentation from 4
and PGH2, with values of −92.5 and −98.0 kcal mol−1, is also ther-
modynamically favourable. For all the endoperoxides examined,
both the oxygen-centred biradicals and distonic radical-anions are
predicted to react via b-scission fragmentation pathways.
The elaborate reduction mechanism deterred us from determin-
ing the rate constant for b-scission fragmentation of the distonic
radical-anions 3a using digital simulation. Rather we used the b-
scission fragmentation of the cumyl alkoxyl radical, with a known
rate constant of 7 × 105 s−1, as a starting reference point.40 This
value is thought to represent the lower limit for fragmentation of
3a. The known quantity of 1,3-diphenyl-1,3-propanedione formed
during the heterogeneous ET reduction can be exploited in order
to estimate the rate constant for ET to 3a. Using the experimental
product ratio of 97 : 3 [cis-diol 3c : 1,3-diphenyl-1,3-propanedione
3f] and 7 × 105 s−1 for kb, the ET rate constant is estimated to
occur with a rate constant of 2 × 107 s−1, which is the same as that
observed with 1,4-diphenyl-2,3-dioxabicyclo[2.2.2]octane with a
value of 3 × 107 s−1.
Experimental
Materials
N,N-Dimethylformamide (DMF) was distilled over CaH2 under
a nitrogen atmosphere at reduced pressure. Tetraethylammonium
perchlorate (TEAP) from Fluka was recrystallized three times
from ethanol and stored under vacuum. Flash chromatography
was performed with silica gel 60 (230–400 mesh ASTM) from
EM science. Reactions were developed in the chosen eluant
with Kieselgel 60 F254 TLC plates and viewed by ultraviolet or
visible light or iodine staining. Toluene was freshly distilled from
sodium and benzophenone. Photo-oxygenation reactions were
performed in solutions of spectroscopic grade dichloromethane
(Caledon) and benzene (EM Science). Other solvents and reagents
not specified were used without purification and obtained from
Aldrich.
Instrumentation
Melting points were recorded on an Electrothermal 9100 capillary
melting point apparatus and were corrected. UV-visible spectra
were recorded on a Varian Cary 100 Bio UV-visible spectrometer.
Infra-red spectra were recorded on a Bruker Vector 33 FT-
IR spectrometer on NaCl plates or in a solution cell. The IR
frequencies are reported in cm−1 and followed by a letter (w, m, or
s) designating the relative strength of the IR stretches as weak,
medium, or strong. Nuclear magnetic resonance spectra were
recorded on a Varian Mercury spectrometer and are reported
Conclusions
The strained bicyclic endoperoxide 1,4-diphenyl-2,3-dioxabicyclo-
[2.2.1]heptane (3) was studied as a representative model of a
prostaglandin endoperoxide using heterogeneous electrochemical
techniques. Reduction of the O–O bond occurs by a dissociative
ET mechanism with thermodynamic and kinetic parameters
consistent with our previous findings with other endoperoxides
that the kinetics are slow and the O–O bond is weak. However,
one noticeable distinction is that the intrinsic barrier includes
a significant inner reorganization energy contribution attributed
to the relief of ring strain. An interesting observation from the
thermochemical cycle in Scheme 6 and the data in Table 3 is
that fragmentation of distonic radical-anions are calculated to be
thermodynamically favourable. The thermochemical calculations
suggest that in many respects 3 is an appropriate model for PGH2.
This study also highlights another rare example of a radical-
anion chain mechanism initiated by the concerted dissociative ET
reduction of a bicyclic endoperoxide.5 Reduction of 3 yields a
distonic radical-anion 3a that undergoes a b-scission fragmenta-
tion in competition with ET reduction from the electrode. The
fragmentation results in the formation of ethylene and the ketyl
radical-anion 3e, which reduces the endoperoxide, thus propa-
gating a homogeneous radical-anion chain mechanism. However,
a competing self-protonation mechanism involving 1,3-diphenyl-
1,3-propanedione decreases the efficiency of the radical-anion
chain reaction. Nonetheless, the significance of this uncommon
mechanism should not be underestimated. In collaboration, we
1
in ppm. H and 13C NMR spectra were recorded at 400.1 and
100.6 MHz, respectively, with CDCl3 as the solvent. Spectra
1
are reported in ppm versus tetramethylsilane (dH = 0.00) for H
NMR and CDCl3 (dC = 77.00) for 13C NMR. Mass spectrometry
was performed on a MAT 8200 Finnigan high-resolution mass
spectrometer by electron impact (EI) and by chemical ionization
(CI) with isobutane.
Synthetic procedures
1,4-Diphenyl-2,3-dioxabicyclo[2.2.1]heptane. The target com-
pound was synthesized from 1,4-diphenyl-1,3-cyclopentadiene43,44
by photo-oxygenation to yield 1,4-diphenyl-2,3-dioxabicyclo-
[2.2.1]heptene, which was reduced with potassium azodicarboxy-
late to yield the desired product. It was recrystallized from ethanol
(110 mg, 34%) as white crystals. mp 110–112 ◦C; mmax(NaCl)/cm−1:
3055 m, 2984 w, 2944 w, 1604 w, 1497 w, 1450 m, 1340 m, 1267 m,
898 m, 742 s, 697 s; dH(400 MHz, CDCl3,, SiMe4): 2.32–2.38(2H,
m), 2.54–2.64 (2H, m), 2.72–2.78 (1H, m), 2.88–2.94 (1H, m), 7.33–
7.43 (6H, m), 7.48–7.53 (4H, m); dC(100 MHz, CDCl3,, SiMe4):
35.35, 52.70, 91.21, 126.59, 128.33, 128.49, 134.56; m/z(EI):
3360 | Org. Biomol. Chem., 2008, 6, 3354–3361
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
•
+
252(M , 8), 224(8), 220(22), 219(100), 115(6), 105(89), 91(15),
77(48), 51(15); Exact Mass: 252.1147 (calculated 252.1150).
9 Z. V. Todres, Organic Ion Radicals: Chemistry and Applications, Marcel
Dekker, New York, 2003.
10 F. Maran, D. D. M. Wayner and M. S. Workentin, Adv. Phys. Org.
Chem., 2001, 36, 85–166; F. Maran, D. D. M. Wayner and M. S.
Workentin, in Kinetics and Mechanism of the Dissociative Reduction
of C–X and X–X Bonds (X = O, S), ed. T. T. Tidwell and J. P. Richard,
New York, 2001.
Electrochemistry
Cyclic voltammetry was performed using either a Perkin-Elmer
PAR 283, or 263A potentiostat interfaced to a personal computer
equipped with PAR 270 electrochemistry software. The working
electrode was a 3 mm diameter glassy carbon rod (Tokai, GC-20)
sealed in glass tubing. The counter electrode was a 1 cm2 Pt plate.
The reference electrode was a silver wire immersed in a glass tube
with a sintered end containing 0.10 M TEAP in DMF. After each
experiment, it was calibrated against the ferrocene/ferricinium
couple at 0.475 V versus KCl saturated calomel electrode (SCE)
in DMF. Constant potential electrolyses were conducted with a
12 mm tipped glassy carbon rotating disk electrode (EDI101) with
a CTV101 speed control unit from Radiometer Analytical. The
experimental set-up was the same as previously reported.5
11 J. M. Tanko and J. P. Phillips, J. Am. Chem. Soc., 1999, 121, 6078–
6079.
12 J. P. Stevenson, W. F. Jackson and J. M. Tanko, J. Am. Chem. Soc, 2002,
124, 4271–4281.
13 M. Chahma, X. Li, P. Phillips, P. Schwartz, L. E. Brammer, Y. Wang
and J. M. Tanko, J. Phys. Chem. A, 2005, 109, 3372–3382.
14 J. M. Tanko, X. Li, M. Chahma, W. F. Jackson and J. N. Spencer, J. Am.
Chem. Soc., 2007, 129, 4181–4192.
15 R. L. Donkers and M. S. Workentin, Chem.–Eur. J., 2001, 7, 4012–4020.
16 R. L. Donkers and M. S. Workentin, J. Phys. Chem. B, 1998, 102,
4061–4063.
17 M. S. Workentin and R. L. Donkers, J. Am. Chem. Soc., 1998, 120,
2664–2665.
18 M. G. Zagorski and R. G. Salomon, J. Am. Chem. Soc., 1980, 80,
2501–2503.
19 R. G. Salomon, Acc. Chem. Res., 1985, 18, 294–301.
20 D. J. Coughlin and R. G. Salomon, J. Am. Chem. Soc., 1977, 99, 655–
657.
Heterogeneous electrolysis
21 R. G. Salomon and M. F. Salomon, J. Am. Chem. Soc., 1977, 99, 3501–
From the electrolysis of 1,4-diphenyl-2,3-dioxabicyclo[2.2.1]-
heptane at the peak potential was recovered 1,3-diphenyl-cyclo-
3503.
22 D. J. Coughlin, R. S. Brown and R. G. Salomon, J. Am. Chem. Soc.,
1979, 101, 1533–1539.
23 B. S. Furniss, A. J. Hannaford, P. W. G. Smith and A. R. Tatchell,
Vogel’s Textbook of Practical Organic Chemistry - 3-Benzoylpropanoic
Acid, John Wiley & Sons, New York, 1989.
24 K. Gollnick and G. O. Schenck, in 1,4-Cycloaddition Reactions, ed.
J. Hamer, Academic Press, New York, 1967, pp. 255–343.
25 Y. Takahashi, K. Wakamatsu, S. Morishima and T. Miyashi, J. Chem.
Soc., Perkins. Trans. 2, 1993, 243–253.
26 W. Adam and H. J. Eggelte, J. Org. Chem., 1977, 42, 3987–3988.
27 J.-M. Save´ant, in Advances in Electron Transfer Chemistry, ed. P. S. Mar-
iano, Greenwich, 1994, vol. 4, pp. 53–116.
pentane-cis-1,3-diol as
a
white solid in 97% yield;
mmax(NaCl)/cm−1: 3375 s (broad), 3088 m, 3060 m, 3030 m,
2966 m, 2938 m, 2858 w, 1665 w, 1602 w, 1495 m, 1449 s, 1406 m,
1262 m, 1227 m, 1100 m, 1071 s, 878 m, 759 s, 700 s; dH(400 MHz,
CDCl3, SiMe4): 2.42–2.57 (6H, m), 3.43 (2H, s, br), 7.25–7.31
(2H, m), 7.34–7.41 (4H, m), 7.50–7.55 (4H, m); alcohol peak
verified by deuterium exchange; dC(100 MHz, CDCl3,, SiMe4):
41.85, 55.83, 83.92, 124.94, 127.12, 128.36, 145.77; m/z(EI):
237(M − 17, 15), 236(M − 18, 41), 133(7), 105(100), 91(5), 77(13);
m/z(CI): 238(20), 237(100), 236(42), 220(6), 219(39), 218(33),
177(6), 105(52), 77(7); Exact Mass: (M − 18) 236.1204 (calculated
236.1201).
28 J.-M. Save´ant, Adv. Phys. Org. Chem., 2000, 35, 117–192.
29 J. C. Imbeaux and J.-M. Save´ant, J. Electroanal. Chem., 1973, 44, 169–
187.
30 J.-M. Save´ant and D. Tessier, J.Electroanal. Chem. Interfacial Elec-
trochem., 1975, 61, 251–263.
31 D. D. M. Wayner and V. D. Parker, Acc. Chem. Res., 1993, 26, 287–294.
32 R. L. Donkers, F. Maran, D. D. M. Wayner and M. S. Workentin,
J. Am. Chem. Soc., 1999, 121, 7239–7248.
Acknowledgements
This work was financially supported by the Natural Sciences and
Engineering Research Council of Canada, the Government of
Ontario (PREA) and the University of Western Ontario. DCM
thanks the Ontario Government for an OGSST postgraduate
scholarship. Doug Hairsine is thanked for performing the mass
spectroscopic measurements.
33 The effective radius reff was calculated using the expression reff
=
rB(2rAB − rB)/rAB where rAB is the molecular radius, and rB is the radius
of the charged portion.27 rAB was determined from the Stokes–Einstein
equation and the diffusion coefficient determined from the convolution
experiments. rB was approximated using the cumyl alkoxide anion.
34 C. Amatore, G. Capobianco, G. Farnia, G. Sandona`, J.-M. Save´ant,
M. G. Severin and E. Vianello, J. Am. Chem. Soc., 1985, 107, 1815–
1824.
35 J.-M. Save´ant, Acc. Chem. Res., 1980, 13, 323–329.
36 J. Pinson and J.-M. Save´ant, J. Am. Chem. Soc., 1978, 100, 1506–1510.
37 S. E. Stein, R. L. Brown and Y. A. Mirokhin, in NIST Standard
Reference Database 25, Gaithersburg, MD, 1991.
38 S. W. Benson, in Thermochemical Kinetics: Methods for Estimation of
Thermochemical Data and Rate Parameters, ed. John Wiley and Sons
Ltd., New York, 1976.
39 W. Adam, Acc. Chem. Res., 1979, 12, 390–396.
40 D. V. Avila, C. E. Brown, K. U. Ingold and J. Lusztyk, J. Am. Chem.
Soc., 1993, 115, 466.
Notes and references
1 Organic Peroxides, ed. W. Ando, John Wiley & Sons Ltd., New York,
1992.
2 C. Martin and V. Ullrich, in Prostaglandins, Leukotrienes and other
Eicosanoids: from Biogenesis to Clinical Applications, ed. F. Marks and
G. Fu¨rstenberger, Wiley-VCH, Weinheim, New York, 1999, pp. 89–108.
3 D. J. Voet and J. G. Voet, Biochemistry, Wiley, Hoboken, NJ, 3rd edn,
2004.
41 F. Najjar, C. Andre´-Barre`s, C. Lacaze-Dufaure, D. C. Magri, M. S.
Workentin and T. Tze`dakis, Chem.–Eur. J., 2007, 13, 1174–1179.
42 P. M. O’Neill, P. A. Stocks, M. D. Pugh, N. C. Araujo, E. E. Korshin,
J. F. Bickley, S. A. Ward, P. G. Bray, E. Pasini, J. Davies, E. Verissimo
and M. D. Bachi, Angew. Chem., Int. Ed., 2004, 43, 4193–4197.
43 N. L. Drake and J. R. Adam, J. Am. Chem. Soc., 1939, 61, 1326–
1329.
4 J. P. Plastaras, F. P. Guengerich, D. W. Nebert and L. J. Marnett, J. Biol.
Chem., 2000, 275, 11784–11790.
5 D. C. Magri and M. S. Workentin, Chem.–Eur. J., 2008, 14, 1698–1709.
6 R. L. Donkers and M. S. Workentin, J. Am. Chem. Soc., 2004, 126,
1688–1698.
7 D. L. B. Stringle and M. S. Workentin, Chem. Commun., 2003, 1246–
1247.
44 L. G. Greifenstein, J. B. Lambert, R. J. Nienhuis, G. E. Drucker and
G. A. Pagani, J. Am. Chem. Soc., 1981, 103, 7753–7761.
8 R. L. Donkers, J. Tse and M. S. Workentin, Chem. Commun., 1999,
135–136.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3354–3361 | 3361
©