Model dialkyl peroxides of the fenton mechanistic probe 2-methyl-1-phenyl-2-propyl hydroperoxide (MPPH): Kinetic probes for dissociative electron transfer

Article abstract of DOI:10.1039/b305348b

Two dialkyl peroxides, devised as kinetic probes for the heterogeneous electron transfer (ET), are studied using heterogeneous and homogeneous electrochemical techniques. The peroxides react by concerted dissociative ET reduction of the O-O bond. Under heterogeneous conditions, the only products isolated are the corresponding alcohols from a two-electron reduction as has been observed with other dialkyl peroxides studied to date. However, under homogeneous conditions, a generated alkoxyl radical undergoes a rapid β-scission fragmentation in competition with the second ET resulting in formation of acetone and a benzyl radical. With knowledge of the rate constant for fragmentation and accounting for the diffuse double layer at the electrode interface, the heterogeneous ET rate constant to the alkoxyl radicals is estimated to be 1500 cm s^-1. The heterogeneous and homogeneous ET kinetics of the O-O bond cleavage have also been measured and examined as a function of the driving force for ET, ΔG_ET, using dissociative electron transfer theory. From both sets of kinetics, besides the evaluation of thermochemical parameters, it is demonstrated that the heterogeneous and homogeneous reduction of the O-O bond appears to be non-adiabatic.

Full text of DOI:10.1039/b305348b

View Article Online / Journal Homepage / Table of Contents for this issue
Model dialkyl peroxides of the Fenton mechanistic probe
-methyl-1-phenyl-2-propyl hydroperoxide (MPPH):
2
kinetic probes for dissociative electron transfer
David C. Magri and Mark S. Workentin*
Department of Chemistry, The University of Western Ontario, London, ON, Canada N6A 5B7
Received 14th May 2003, Accepted 22nd July 2003
First published as an Advance Article on the web 27th August 2003
Two dialkyl peroxides, devised as kinetic probes for the heterogeneous electron transfer (ET), are studied using
heterogeneous and homogeneous electrochemical techniques. The peroxides react by concerted dissociative ET
reduction of the O–O bond. Under heterogeneous conditions, the only products isolated are the corresponding
alcohols from a two-electron reduction as has been observed with other dialkyl peroxides studied to date.
However, under homogeneous conditions, a generated alkoxyl radical undergoes a rapid β-scission fragmentation
in competition with the second ET resulting in formation of acetone and a benzyl radical. With knowledge of
the rate constant for fragmentation and accounting for the diffuse double layer at the electrode interface, the
Ϫ1
heterogeneous ET rate constant to the alkoxyl radicals is estimated to be 1500 cm s . The heterogeneous and
homogeneous ET kinetics of the O–O bond cleavage have also been measured and examined as a function of
the driving force for ET, ∆GET, using dissociative electron transfer theory. From both sets of kinetics, besides the
evaluation of thermochemical parameters, it is demonstrated that the heterogeneous and homogeneous reduction
of the O–O bond appears to be non-adiabatic.
An alternative mechanism has been postulated to rationalise
the chemistry observed with the nonheme biomimetic catalysts.
Introduction
In recent years, there have been many claims of successfully
mimicking the elegant chemistry of monooxygenases, enzymes
that use molecular oxygen to selectively functionalise saturated
Substantial evidence supports alkoxyl radicals as the reactive
26–34
intermediate and not high-valent iron-oxo species.
The
initial step of this proposed mechanism involves homolysis of
1–9
hydrocarbons (eqn. (1)). Many of these ubiquitous enzymes
the O–O bond by electron transfer (ET) from the iron catalyst
1
0–12
13
such as the cytochrome P450s,
peroxidases,
and
30,34–36
to the hydroperoxide (eqn. (4)).
The generated alkoxyl
13
catalases contain an iron() protoporphyrin IX complex,
radical then abstracts a hydrogen atom from the saturated
hydrocarbon to afford a carbon radical (eqn. (5)). Under
aerobic conditions, the carbon radical reacts with dioxygen
1
4,15
whereas methane monooxygenases,
found only in methano-
trophic bacteria, contain a (µ-oxo) di-iron complex. Substantial
evidence suggests the active intermediate in the oxidative mech-
anism of heme enzymes is an oxo-iron() porphyrin π-radical
forming a peroxyl radical leading to alkane oxidation products
26–30,36
(
eqn. (6)).
In both of these mechanistic possibilities, the


ؒ
ϩ 11,16,17
cation [PFe ᎐O] .
n





catalyst, represented as Fe , can have either the Fe or Fe
oxidation state.
(
1)
(
2)
In attempts to mimic the aforementioned enzymes, most
biomimetic systems have utilised an Fe() or Fe() complex as
(3)
t
the catalyst and H O or a tertiary hydroperoxide (R OOH),
2
2
such as tert-butyl hydroperoxide (TBHP), as the oxidant. A
notable example is work by Que and colleagues who have
characterised and studied both mono-iron() and (µ-oxo)
di-iron() complexes containing tetradentate, tripodal tris-
(
(
(
4)
5)
6)
4–8,18–20
(
pyridylmethyl)amine (TPA) ligands.
The oxidation of
various saturated hydrocarbons in the presence of these iron
catalysts has been achieved and found to yield not only alcohols
(
eqn. (1)) but other products such as ketones and dialkyl per-
To differentiate biomimetic oxidations from oxidations
36
oxides. It was initially proposed that the mechanism of action
involved heterolysis of the hydroperoxide O–O bond generating
a high-valent iron-oxo species, Fe ᎐O or Fe ᎐O (eqn. (2)),
mediated by freely diffusing radicals —heterolytic (eqn. (2))
versus homolytic (eqn. (4)) O–O bond cleavage—Ingold,
5,6


Wayner, and co-workers devised the mechanistic probe
10
26
analogous to that observed with the monooxygenase enzymes,
2-methyl-1-phenyl-2-propyl hydroperoxide (MPPH).
The
the exact high-valent iron-oxo species dependent on the initial
oxidation state of the metal catalyst. Subsequently, the high-
valent iron-oxo species was presumed to abstract a hydrogen
atom from the hydrocarbon, and “insert” an oxygen atom into
characteristic feature of the MPPH probe is the rapid β-scission
8 Ϫ1
fragmentation (k ∼ 2.2 × 10 s ) of the corresponding alkoxyl
β
2
6
radical (eqn. (7)). The MPPH alkoxyl radical has a short life-
time (5 ns), much shorter than the tert-butoxyl radical, and too
short to abstract a hydrogen atom from a saturated hydro-
carbon. With MPPH, hydrogen abstraction of the alkane by the
MPPH alkoxyl radical does not occur (eqn. (5)) and instead the
radical fragments to acetone and a benzyl radical (eqn. (7)),
which is the only species subsequently oxidised. In the presence
of air, the benzyl radical reacts with dioxygen forming a
5,6
the substrate regenerating the catalyst (eqn. (3)). Indirect

evidence for the existence of a nonheme Fe ᎐O intermediate
21–23
has been invoked based on mechanistic studies,
but most
recently supported by a crystallographic and spectroscopic
2
4

study. On the contrary, evidence for a nonheme Fe ᎐O
7,8,25
intermediate is supported solely on mechanistic studies.
3
418
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3

View Article Online
Scheme 1 The competitive reaction pathways, reduction and β-scission fragmentation, available to the MPPH alkoxyl radical after dissociative
1
2
electron transfer to the dialkyl peroxides DMPP and BMPP. The rate constants k and k are used generically to represent either a heterogeneous
ET
ET
or homogeneous ET. In the text, a distinction is made by replacing the subscript ET with the subscript het or hom.
benzylperoxyl radical yielding a large quantity of benzalde-
hyde, whereas under an inert atmosphere, benzyl alcohol is the
major product.
dialkyl peroxides are an alkoxyl radical and an alkoxide
(eqn. (8)). Under heterogeneous conditions, the alkoxyl radical
is generated at the electrode surface at a potential much more
negative than its reduction potential. Because of the favourable
driving force for ET, generally in excess of Ϫ1.0 V, the alkoxyl
radical is reduced at the electrode interface before escaping into
the bulk solution. In only two examples has some other alkoxyl
21,22,28,29
(
7)
Despite the wealth of information known about the chem-
istry of the MPPH alkoxyl radical and the follow-up chemistry
reactivity been observed in competition with the second hetero-
39,40
geneous ET (eqn. (9)).
Intuitively, based on the large driving
(
eqn. (7)), little is known about the initial ET to the O–O bond
force for ET, it is assumed the rate of the heterogeneous ET to
the alkoxyl radical is extremely fast, faster than the rate of
diffusion away from the electrode surface; however, up to now
no attempt has been made to quantify the rate constant of this
process (eqn. (9)).
of MPPH (eqn. (4)). The usefulness of MPPH as a mechanistic
probe, in general, would be greatly enhanced if the rates of ET
from potential donors were known. Furthermore, knowledge
of the standard reduction potential of MPPH would allow the
feasibility for ET from potential donors to be predicted. By
applying theories of ET, the driving force of the reaction can be
related to the kinetics, and from such an analysis, other inherent
properties of MPPH can be determined.
One way of determining thermodynamic and kinetic inform-
ation about a substrate, such as standard reduction potentials
and kinetic rate constants, is by using various electrochemical
techniques. Unfortunately, hydroperoxides are prone to rapid
decomposition under reductive electrochemical conditions
because of the acidic peroxy proton, thus preventing such
information from being easily determined. On the other hand,
dialkyl peroxides lacking α-protons adjacent to the O–O bond
are robust to base-induced decomposition. Keeping these struc-
tural considerations in mind, we devised two dialkyl peroxides
as models for MPPH (see Scheme 1) in order to gain some
insight into its ET chemistry and intrinsic properties.
(
(
8)
9)
In this article, the dialkyl peroxides di-2-methyl-1-phenyl-2-
propyl peroxide (DMPP) and tert-butyl 2-methyl-1-phenyl-2-
propyl peroxide (BMPP) are studied using heterogeneous and
homogeneous electrochemical techniques in aprotic solvents.
These two peroxides were devised as kinetic probes for the
heterogeneous ET rate constant to alkoxyl radicals initiated by
dissociative electron transfer (eqn. (9)). Product studies were
conducted to investigate the partitioning ratio between the reac-
tion pathways available to the MPPH alkoxyl radical, reduction
and β-scission fragmentation, and from knowledge of the rate
The study of the ET reduction of peroxides and endo-
3
7–42
8
Ϫ1
peroxides is an active area of research by our group
and the
constant of β-scission (k ∼ 2.2 × 10 s ) we have quantified the
β
43–48
2
group of Maran and co-workers. Besides playing important
roles in a number of biological and chemical processes,
rate constant of the second heterogeneous ET (k het) to the
49–51
MPPH alkoxyl radical (eqn. (9)). In addition, the hetero-
1
1
peroxides and endoperoxides also provide intriguing model sys-
tems for testing Savéant’s dissociative electron transfer (DET)
geneous (k het) and homogeneous (k hom) ET kinetics of DMPP
and BMPP have been measured in N,N-dimethylformamide.
Cyclic voltammetry in conjunction with a thermochemical
cycle is used to evaluate the dissociative standard reduction
potential (EЊdiss) of each peroxide and used to relate the kinetics
52–54
theory.
In general, DET systems are characterised by the
concerted cleavage of a σ bond upon ET, resulting in the form-
ation of reactive species such as radicals and ions. In most case
studies, ET to O–O bonds has been observed to be a concerted
to the driving force for ET, ∆GET. We also provide further
3
7–44
37,38,44
dissociative process,
although examples of a stepwise dis-
have also been observed. Substantial
evidence that ET to the O–O bond is non-adiabatic,
and
44–48
sociative mechanism
accounting for this behaviour, evaluate a number of thermo-
chemical parameters for these two model compounds. The
knowledge gained from the study of these dialkyl peroxides has
relevance not only to the dissociative chemistry of peroxides in
general, but also to the employment of MPPH as a mechanistic
probe in biomimetic systems.
evidence has also revealed that ET to O–O bonds is non-
37,38,44
adiabatic.
This behaviour is manifested in ET rate con-
stants smaller than predicted by adiabatic ET models.
Regardless of the mechanistic route, concerted or stepwise,
upon O–O bond fragmentation the products from the ET to
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3419

View Article Online
Table 1 Cyclic voltammetry data for the cathodic reduction of DMPP and BMPP in N,N-dimethylformamide (DMF) and acetonitrile (ACN) in
the presence of 0.10 M TEAP at 25 ЊC measured at a glassy carbon electrode
a
a
a
a
DMPP
DMF
DMPP
ACN
BMPP
DMF
BMPP
ACN
Ϫ1
Scan rate/V s
0.1
b
E /V vs. SCE
Ϫ2.16
Ϫ2.27
Ϫ2.40
0.188
0.200
0.259
0.254
0.239
0.184
0.23
Ϫ2.12
Ϫ2.23
Ϫ2.37
0.192
0.211
0.248
0.249
0.226
0.192
0.23
Ϫ2.34
Ϫ2.46
Ϫ2.60
0.188
0.203
0.280
0.254
0.235
0.170
0.22
Ϫ2.30
Ϫ2.42
Ϫ2.60
0.194
0.221
0.260
0.246
0.221
0.184
0.22
p
1
0
0.1
1
0
0.1
.0
1
1
1
b
Ep/2 Ϫ E /V vs. SCE
p
.0
α = 1.857RT/F(Ep/2 Ϫ Ep)
1
0
.0
α = 1.15RT/F(dE /d(log ν))
Ϫ(dE /d(log ν))/mV
n (# of electrons)
p
127
2.0
129
2.1
134
2.0
133
2.0
p
c
a
b
Potentials are referenced versus ferrocene: 0.449 V and 0.475 V vs. SCE in ACN and DMF. dE /d(log ν) based on scan rates between 0.1 and 10 V
p
Ϫ1
c
s
. In presence of non-nucleophilic acids: n = 1.9 (2,2,2-trifluoroethanol), n = 2.2 (acetanilide).
Results
Cyclic voltammetry
The reduction of DMPP and BMPP was studied in solutions of
N,N-dimethylformamide (DMF) and acetonitrile (ACN) con-
taining 0.10 M tetraethylammonium perchlorate (TEAP) using
a glassy carbon working electrode at 25 ЊC. Cyclic voltammetry
of both peroxides is characterised by a broad and chemically
irreversible cathodic peak at all scan rates investigated between
Ϫ1
Ϫ1
0
.1 and 50 V s . The peak potentials, E , at 0.1 V s vs. SCE
p
are Ϫ2.16 and Ϫ2.34 V in DMF and Ϫ2.12 and Ϫ2.30 V in
ACN, respectively, for DMPP and BMPP. Increasing the scan
rate, v, shifts the E linearly to more negative potentials from
p
Ϫ1
Ϫ127 to Ϫ134 mV (log v) and broadens the cathodic wave
Ϫ1
(
2
peak width: ∆Ep/2 = Ep/2 Ϫ E ) from 188 mV at 0.1 V s to
p
Ϫ1
80 mV at 10 V s . The ∆E , and peak height, I , of the
p/2 p
voltammograms are unaffected by the addition of weak non-
nucleophilic acids such as 2,2,2-trifluoroethanol and acetan-
ilide. A summary of the voltammetry data in both solvents is
reported in Table 1.
An irreversible anodic peak, generated after the initial reduc-
tion, is observed in both solvents at roughly Ϫ0.2 V vs. SCE at
Ϫ1
0
.1 V s . It is attributed to the oxidation of alkoxides formed
from the initial reduction of the peroxide, as independently
confirmed for DMPP by an authentic sample of MPPOH in the
presence of tetraethylammonium hydroxide. In ACN, a second
anodic peak is observed at slightly more negative potentials on
increasing the scan rate and attributed to some other electro-
chemically generated base. These anodic peaks (only the more
positive peak in ACN was examined) were found to increase
with scan rate and to experimentally shift in the positive direc-
Ϫ1
Fig. 1 Cyclic voltammograms of 2.0 mM (a) DMPP at 1.0 V s and
Ϫ1
(
b) BMPP at 0.8 V s in 0.10 M TEAP–acetonitrile solution in the
absence (—) and presence (----) of 5 mM 2,2,2-trifluoroethanol.
Ϫ1
tion by 31.5 mV (log v) in accordance with a first order
55,56
irreversible process.
Cyclic voltammographs of DMPP and
½
current, i /ν , decreases with scan rate indicates the initial ET
is the rate-determining step and cannot be described using
Butler–Volmer kinetics. As a consequence, the standard
reduction potential (EЊdiss) and standard heterogeneous rate
constant (kЊhet) cannot be determined directly from the cyclic
voltammetry (vide infra).
p
BMPP in the absence and presence of the 2,2,2-trifluoroethanol
are shown in Fig. 1.
57
The transfer coefficient, α, is a diagnostic parameter, which
≠
reflects how the free energy of activation, ∆G , varies with the
driving force for ET, ∆GET. It can be used as an indication of
whether ET occurs by a stepwise (α ≅ 0.5) or a concerted mech-
5
2,53
anism (α << 0.5).
ways: i) from the scan rate dependence of the irreversible peak α
1.15RT/[dE /(dlog ν)] and ii) from the corresponding peak
In our analysis, α was determined two
Heterogeneous product studies
=
Constant potential electrolyses of both peroxides were carried
out with a glassy carbon rotating disk electrode (or a platinum
foil). The experimentally determined coulometry values are
p
56
width at each scan rate α = 1.857RT/(F∆Ep/2). The first
approach yields average values of 0.23 and 0.22 for DMPP and
BMPP; the second approach gives a decreasing range of α
values from 0.25 at 0.1 V s to 0.17 at 10 V s . A summary of
the α values is included in Table 1. Such low α values are an
indication that ET to DMPP and BMPP occurs by a concerted
Ϫ1
included in Table 1. In all cases, 2 F mol of charge is con-
Ϫ1
Ϫ1
sumed independent of solvent, electrode material and the pres-
ence of weak acid. Work-up of the electrolysed solution of
DMPP and quantification by GC yielded only 2-methyl-1-
phenyl-2-propanol (MPPOH) in an average isolated yield of
65%. In the case of BMPP, the recovered yield of MPPOH was
52,53
dissociative mechanism.
Furthermore, the scan rate
dependence of the E , and the fact that the normalised peak
p
3
420
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9

View Article Online
Table 2 Data acquired from the convolution analysis of DMPP and BMPP in N,N-dimethylformamide (DMF) and acetonitrile (ACN) in the
presence of 0.10 M TEAP at 25 ЊC at a glassy carbon electrode
DMPP
DMF
DMPP
ACN
BMPP
DMF
BMPP
ACN
a b
2
Ϫ1
Ϫ6
Ϫ5
Ϫ6
Ϫ5
D
/cm s
4.87 × 10
Ϫ1.20
Ϫ6.7
1.16 × 10
6.68 × 10
1.47 × 10
c
EЊdiss /V d
Ϫ1.22
Ϫ6.5
12.4
Ϫ1.26
Ϫ6.8
11.8
Ϫ1.30
Ϫ6.3
10.5
Ϫ1
log (kЊ /cm s )
Ϫ1
het
≠
e
∆
G0 /kcal mol
10.8
a
These values are validated by the diffusion coefficient ratios (ACN/DMF), which are 2.38 and 2.20 for DMPP and BMPP in excellent agreement
b
Ϫ5
Ϫ5
2
Ϫ1
with the ratio between the viscosity of the solvents: 0.796/0.344 = 2.31. Ferrocene diffusion coefficients of 1.13 × 10 and 2.38 × 10 cm s in
c
DMF and ACN were used to calculate the area of the electrode from reference 38a. Estimated error is ± 0.20 V and uncorrected for the double layer.
d
Ϫ1
Estimated error ± 0.8 approximated by digital simulation of experimental CVs from 0.1 to 10 V s using the α values reported in Table 1 and the
e
≠
≠
experimental EЊdiss values. ∆G0 determined from the slope of α vs. E plots and ∆G0 = F/[8(dα/dE)].
similar. tert-Butanol, on the other hand, was not quantified due
solvents. The kЊhet was estimated from the second order fit to the
heterogeneous kinetics at EЊ . The kЊ are low with average
values of 2.0 × 10 and 4.0 × 10 cm s in DMF and ACN,
which results in the peak potential being more negative than the
to its high solubility in water and volatility, although it was
diss
het
Ϫ1
1
Ϫ7
Ϫ7
detected by H NMR spectroscopy as a singlet at 1.26 ppm.
Fragmentation products such as toluene or acetone were not
detected upon work-up suggesting under heterogeneous condi-
tions that β-scission fragmentation of the alkoxyl radicals does
not compete, at least to a measurable extent, with the second
heterogeneous ET. Admittedly, the isolated yields of MPPOH
are not quantitative. Further support for the lack of fragmen-
tation products was obtained by digital simulation of the
experimental voltammograms using the determined parameters
from Tables 1 and 2 and by monitoring the concentration pro-
files at several hundreds millivolts past the peak potential.
estimated EЊdiss by 0.8 V. A complete summary of the extracted
≠
0
parameters D, kЊhet, EЊdiss and intrinsic barrier, ∆G , is included
in Table 2. These values are consistent with those determined
38,43
for other dialkyl peroxides.
The experimental values were
finally checked by digital simulation of the cyclic voltam-
mograms including the second heterogeneous ET rate constant
2
(k het) determined in this study (vide infra) resulting in good
agreement between the experimental and simulated results.
8
Ϫ1
Simulations with k equal to 2.2 × 10 s predicted the form-
Homogeneous product studies
β
ation of alcohols in greater than 99% yield. An increase in k by
β
The indirect electrolysis of DMPP and BMPP was studied
using electrochemically generated outer-sphere ET radical
anion donors to determine if the β-scission fragmentation can
compete with the second ET under homogeneous conditions.
Mediators were carefully selected to eliminate the possibility of
one and two orders of magnitude, respectively, would yield
approximately 2% and 6% fragmentation products. Thus, the
coulometry, product analysis, and voltammetry are consistent
with the steps outlined in eqns. (10)–(12), with an overall
38,43,44
two-electron dissociative ET mechanism
eqn. (13).
described by
63
nucleophilic addition of the benzyl anion and studies were
performed in ACN because of a known complicating side reac-
64
tion in DMF. The results from the homogeneous product
studies of DMPP and BMPP are reported in Tables 3 and 4.
The data are reported as mole % based on the consumption of
peroxide. In all studies with DMPP no starting peroxide was
recovered. However, with BMPP and mediators at less negative
reduction potentials (EЊ > Ϫ1.5 V), which are not catalytic
because of the slower reaction rates, some unreacted peroxide
was recovered. Also reported in the tables are the standard
reduction potential of the donors in ACN and the electron
equivalents consumed in the indirect electrolysis. For all the
products the maximum possible yield is 100%, with the excep-
tion of MPPOH from DMPP, in which case the maximum
possible yield is 200%.
(
(
(
(
10)
11)
12)
13)
Heterogeneous kinetics
1
Convolution analysis was used to evaluate the kinetics (k het) of
the heterogeneous reduction of BMPP and DMPP at a glassy
carbon electrode in both DMF and ACN. The method has been
As can be seen in Tables 3 and 4, the indirect electrolysis is a
two-electron (n = 2) process. With donors of more negative
reduction potentials (EЊ < Ϫ1.7 V), the donor recovery was
always > 98%, with the exception of anthracene where a small
amount of dihydroanthracene was recovered. The major quan-
tifiable products from both peroxides are MPPOH and toluene,
3
7,38,41,43–48
successfully applied to a number of peroxide systems.
The experimental and theoretical details of the technique are
58–60
described in detail elsewhere.
Briefly, a series of high-quality
cyclic voltammograms were obtained between the scan rates of
.1 and 10 V s at 1 or 2 points per mV data acquisition and
the corresponding backgrounds subtracted. The normalised
Ϫ1
0
65
with bibenzyl as a minor product detected in trace amounts.
cyclic voltammograms (i–E curves) were then transformed into
sigmoidal-shaped curves by use of the convolution integral.
Formation of toluene and bibenzyl is conclusive evidence
β-scission fragmentation of the MPPH alkoxyl radical, gener-
ated upon the first ET, is in competition with a second ET. A
similar result was also observed from the decarboxylation of
the phenylacetoxyl radical from the chemically-initiated ET
61
The maximum limiting currents, Ilim, of the transformed curves
were in excellent agreement with a margin of error less than 3%.
From the determined Ilim and the known area of the glassy
carbon electrode (determined separately by analysis of ferro-
cene with literature diffusion coefficients), the diffusion co-
efficients, D, of BMPP and DMPP in the presence of 0.10 M
66
reduction of phenylacetyl peroxide.
The mass balance of phenyl-derived products is modest with
the accounted mass typically around 75% (range 60–80%). The
product loss is likely a consequence of the need to remove the
electrolyte from the electrolysis mixture before quantifying by
GC. In the heterogeneous electrolysis of DMPP, modest yields
of MPPOH were also obtained so the unaccounted mass in
the homogeneous studies is undoubtedly from unrecovered
TEAP in DMF and ACN were determined. From the i and I
lim
1
values, the potential dependence of both k
and α is evalu-
het
62
ated. EЊdiss were estimated from extrapolation of α–E plots
corresponding to α = 0.5. By this method, approximate values
of EЊdiss between Ϫ1.2 and Ϫ1.3 V were obtained in both
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3421

View Article Online
Table 3 Homogeneous product studies of DMPP with electrochemically generated radical anion donors in acetonitrile–0.10 M TEAP solution.
a
Yields are expressed in mole % based on consumption of DMPP
Donor
Anthracene DPA
Fluoranthene
Perylene
4m4mAzo
Azobenzene
EЊ/V
n
Donor recovered
MPPOH
Toluene
Bibenzyl
Ϫ1.965
Ϫ1.873
2.d2
100
122
36
Ϫ1.762
Ϫ1.674
2.d1
Ϫ1.478
2.b0
Ϫ1.335
2.0
1.9
— b
50
50
3
b, c
b, d
95
100
80
65
115
32
< 1
124
30
< 1
105
27
< 1
90
25
< 1
< 1
4
a
b
c 1
d
No DMPP was recovered in all experiments. Amount of donor recovered quantified by gas chromatography, H NMR, and cyclic voltammetry.
All other yields quantified by gas chromatography. DPA = diphenylanthracene; 4m4mazo = 4-methyl-4Ј-methoxyazobenzene.
Table 4 Homogeneous product studies of BMPP with electrochemically generated radical anion donors in acetonitrile–0.10 M TEAP solution.
a
Yields are expressed in mole % based on consumption of BMPP
Donor
Anthracene
DPA
Fluoranthene
Perylene
4m4mAzo
Azobenzene
EЊ/V
n
Donor recovered
MPPOH
Toluene
Bibenzyl
Ϫ1.965
Ϫ1.873
2.d0
Ϫ1.762
Ϫ1.674
2.d0
Ϫ1.478
— b
56
44
8
Ϫ1.335
2.0
1.8
— b
44
50
7
b, c
b, d
91
100
57
100
80
53
17
0
49
12
0
51
13
< 1
18
0
< 1
2
a
b
Unreacted BMPP was recovered in the experiments with 4m4mazo and azobenzene in 57 and 84% yield. Amount of donor recovered quantified by
gas chromatography, H NMR, and cyclic voltammetry. All other yields quantified by gas chromatography. DPA = diphenylanthracene;
m4mazo = 4-methyl-4Ј-methoxyazobenzene.
c 1
d
4
MPPOH. Despite the modest yields, comparison of the results
for both peroxides reveals [MPPOH]/[toluene] is approximately
constant with a ratio of 3.7 to 1 for donors with EЊ < Ϫ1.5 V.
In contrast, with 4-methoxy-4Ј-methylazobenzene and azo-
benzene the amount of donor recovered is no greater than 65%.
Comparison of the voltammetry of the redox couple before and
after the reaction revealed a significant decrease in the mediator
anodic current. MPPOH and toluene were still the major prod-
ucts; however, the amount of toluene decreases as the reduction
potential of the donor becomes less negative with ratios of
MPPOH to toluene as high as 17. With these two donors, GC
analysis also revealed other signals at later retention times.
These are attributed to products resulting from the destruction
once all the other rate constants for eqns. (14) through (20) are
1
known. Besides k hom, the only other rate constant not known
3
with certainty is k hom. It was estimated by examining the com-
petition kinetics between eqns. (18) and (19) as described in the
next section.
(
(
14)
15)
(16)
17)
(18)
19)
(20)
63
of the radical anion catalyst coupling with the benzyl radical.
(
From the studies with azobenzene two coupling products
were characterised by GC/MS with m/z of 272. No coupling
products were identified from the addition reaction between the
radical anion and the alkoxyl radicals. Alkoxyl radicals have
reduction potentials of ∼ Ϫ0.2 V so with all the donors used in
this study reduction, rather than coupling, is the more favour-
(
37,38
able pathway.
Having established the products from the indirect electrolysis,
the dissociative mechanism of DMPP and BMPP can be
adequately described by equations (14) through (20). After
reduction of the aromatic donor to the radical anion (eqn. (14))
the peroxide is reduced in a concerted ET (eqn. (15)) to yield an
alkoxyl radical, which can be subsequently reduced by another
radical anion molecule (eqn. (16)), or in the case of the MPPH
alkoxyl radical, rapidly undergoes the β-scission fragmentation
3
q Parameter measurements and the determination of k
hom
The competition between the homogeneous ET (eqn. (18)) and
coupling between a radical anion donor and the benzyl radical
(
eqn. (19)) can be expressed by the dimensionless parameter q:
21)
>> k , the major pathway is an ET and q is
(
eqn. (17)). The tert-butoxyl radical will react by reduction since
(
4
its β-scission fragmentation is considerably slower (8.3 × 10
s ). Upon fragmentation of the MPPH alkoxyl radical, acet-
Ϫ1 67
3
one and a benzyl radical are liberated (eqn. (17)). The benzyl
radical in the presence of a radical anion donor reacts com-
petitively either by reduction (eqn. (18)) or coupling with the
In the case k
hom
c
3
equal to 1.0. In the scenario where k >> k
only coupling
c
hom
between the radical and radical anion donor is observed and q
equals 0.0. A value of q in between these two extremes (0.0 < q
< 1.0) indicates a partitioning between the two processes.
Numerous studies with various carbon radicals including
63,68,69
donor (eqn. (19)) as previously reported.
Dimerization of
the benzyl radical (eqn. (20)) is a minor side reaction in
comparison to eqns. (18) and (19) as only a trace amount of
bibenzyl is observed on work-up, a testament to the low steady-
state concentration of the benzyl radical. Applying the method
74,75
75
76
alkyl,
between eqns. (18) and (19) is related to the potential difference
between the redox potential of the donor, EЊD/D , and of the
radical, in our case, the benzyl radical, EЊPhCH2ؒ/PhCH2
allyl and acyl, have demonstrated the competition
7
0–73
of homogeneous redox catalysis,
kinetics for dissociative
ؒ
Ϫ
1
77
electron transfer, k
(eqn. (15)), can be accurately evaluated
Ϫ
.
hom
3
422
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9

View Article Online
3
Table 5 Competition q parameters for the reactivity of the benzyl
radical with radical anion donors determined by the RDE and ultra-
microelectrode method in 0.10 M TEAP–DMF solution
EЊD/D
ؒ
Ϫ
of the donor where k
= k and q = 0.5. The curve
hom c
fitting analysis of the data is based on the quadratic-activation
relationship described by the Marcus theory of ET accounting
83
a
3
Ϫ1 Ϫ1 b
for diffusion (eqn. (22)):
Donor
q
k
hom/M
s
1
9
9
8
8
7
0
Fluoranthene
Perylene
0.95
0.85
0.70
0.40
0.15
0.00
1.9 × 10
5.7 × 10
2.3 × 10
6.7 × 10
1.8 × 10
1.0 × 10
(
22)
Tetracene
c
4
4
m4mazo
-Methoxyazobenzene
d
Azobenzene
where k is the diffusion-limited rate constant, K is the equi-
d d
a
librium constant for formation of the encounter complex, Z is
the collision frequency for an adiabatic ET, λ is the reorganis-
ation energy, and ∆GET is the free energy of the ET reaction.
Substitution of eqn. (22) into eqn. (21) and ∆GET = 23.06
Determined according to K. Daasbjerg, Acta Chem. Scand., 1993, 47,
b
3
1
98–402. Calculated from the measured q-value and eqn. (21) with k =
c
9
Ϫ1 Ϫ1
c
d
× 10 M
s . 4-Methyl-4Ј-methoxyazobenzene. Calculated using
q = 0.01.
(
EЊD/D
ؒ
Ϫ
Ϫ EЊPhCH2ؒ/PhCH2 ) yields an expression relating the q
Ϫ
Ϫ
78,79
parameter as a function of EЊD/D
ؒ :
Studies with radical probes,
flash photolysis method, indicate k = 1 × 10 M
and most recently by a laser
9 Ϫ1 Ϫ1
80
s
is a
c
reasonable approximation for coupling between aromatic radi-
cal anions and alkyl and benzyl radicals independent of the
(
23)
standard potential of EЊ
ؒ
Ϫ
. Thus, eqn. (21) provides a way of
D/D
3
evaluating k
once q is known. Most q parameter studies
hom
have utilised linear sweep voltammetry as the technique of
choice, as has been done with a number of benzylic-like radi-
In the curve fitting analysis the following parameter values
9
Ϫ1 Ϫ1
11
Ϫ1 Ϫ1
Ϫ1
77,81
were used: k = 1 × 10 M s , Z = 3 × 10
M
s , K = 1 M ,
c
d
cals.
We have measured the competition q-values for our
1
0
Ϫ1 Ϫ1
3
and k = 2 × 10
M
s . K was calculated from K = (4/3)πa N ,
d
d
d
a
series of donors using the rotating disk electrode (RDE) and
76,82
the expression for the stability constant of the encounter complex
when at least one reagent is uncharged where a is the centre-to-
centre distance between the reagents with radii of 3.5 Å. From
ultramicroelectrode method.
This relatively unused tech-
nique has many advantages over linear sweep voltammetry
including a more accurate determination of q, and can be used
Ϫ3
Ϫ1 Ϫ1 82
the fit of the data EЊ
2ؒ
Ϫ
2
= Ϫ1.49 ± 0.03 V and λ = 13.4
PhCH /PhCH
for slow reactions with a rate constant as low as 10
M
s .
Ϫ1
±
1.5 kcal mol . These values for the standard reduction poten-
The q-values were measured in N,N-dimethylformamide
from the average of several experiments with DMPP as the
tial and the reorganisation energy are in excellent agreement with
7
7,84,85
results from other techniques.
substrate. They are reported in Table 5 along with the values of
3
k
hom. The trend observed with the q-values parallels the results
Homogeneous kinetics
from the homogeneous product studies in Tables 3 and 4. For
example, in the product studies with fluoranthene no loss of the
mediator was observed by GC or cyclic voltammetry so q with
this donor is 1.0. A q-value of 0.95 was experimentally meas-
ured, which is within the error limit of the RDE and ultra-
microelectrode techniques. As the reduction potential of the
donor becomes more positive than fluoranthene, q is found to
decrease towards zero in accordance with the lower recovery
yields of donor in the product studies.
The outer-sphere ET kinetics from a series of electrochemically
generated radical anion donors were measured for DMPP and
BMPP in DMF using the methods of homogeneous redox
7
0–73
catalysis
and a potentiometric technique using a rotating
86,87
disk electrode.
In the first method, mediators were chosen
ؒ
Ϫ
with a reversible electron transfer couple (D/D ) with both
oxidised and reduced species stable on the voltammetric time
scale, and the reduction potential of the donor positive to the
peak potential of the peroxide. The reversible reduction wave of
the radical anion donor, D , is monitored (eqn. (14)) as a func-
tion of peroxide concentration and sweep rate. As increasing
concentrations of peroxide are added to the donor solution, the
reversible wave of the donor is transferred into a chemically
irreversible wave and the cathodic peak current increases
catalytically (eqn. (15)). Based on eqns. (14)–(20) and known
In Fig. 2, the q-values are plotted as a function of EЊD/D
The plot exhibits an S-shaped curve with q decreasing with
increasing EЊD/D . A rigorous non-linear regression analysis
ؒ
Ϫ
.
ؒ
Ϫ
ؒ
Ϫ
was used to extract the standard reduction potential of the
benzyl radical. Interpolation of the plot allows for the
q
determination of the half-wave potential, E ½, which is the
rate constants for the various steps, including the values of
3
k
determined by the q method, ET rate constants for the
hom
1
O–O cleavage (k hom) were determined by digital simulation. In
the simulations of the unsymmetrical peroxide, BMPP, the
fragments from dissociative ET were treated as different
entities. A set of side reactions with DMF was also included in
38,64
the analysis.
1
The homogeneous kinetics (k hom) measured by both tech-
niques are summarised in Table 6. Spanning 6 orders of magni-
tude, the rate constants increase as the standard potential of the
donor becomes more negative as a consequence of an increase
in the driving force for ET expressed as ∆GET = 23.06 (EЊD/D
EЊdiss). The EЊD/D , reported in Table 6, were measured by
ؒ
Ϫ
Ϫ
ؒ
Ϫ
cyclic voltammetry; the EЊdiss cannot be determined directly due
to the irreversible nature of concerted ETs. However, EЊdiss can
be estimated by at least two ways: (i) by convolution analysis as
previously outlined and (ii) by a thermochemical cycle yielding
38,88
eqn. (24).
Fig. 2 The competition parameter q for the benzyl radical as a
function of the EЊD/D of aromatic donors. The data were acquired
with DMPP as the substrate in DMF–0.10 M TEAP solution.
ؒ
Ϫ
(
24)
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3423

View Article Online
Table 6 Rate constants for electron transfer from electrochemically generated radical anion donors to DMPP and BMPP measured in DMF–0.10
M TEAP solution at 25 ЊC
a
EЊD/D
ؒ
Ϫ
(DMF) /
DMPP
∆GET/kcal mol
DMPP
log (k hom/M
BMPP
∆GET/kcal mol
BMPP
log (k hom/M
Ϫ1 b
1
Ϫ1 Ϫ1
Ϫ1 c
1
Ϫ1 Ϫ1
Donor
V vs. SCE
s
)
s )
d
Isoquinoline
Pyrene
Anthracene
Diphenylanthracene
Fluoranthene
Perylene
Ϫ2.158
Ϫ2.054
Ϫ1.934
Ϫ1.838
Ϫ1.730
Ϫ1.641
Ϫ1.548
Ϫ1.478
Ϫ1.423
Ϫ1.315
Ϫ1.256
Ϫ1.177
Ϫ1.100
Ϫ1.035
—
—
—
—
Ϫ16.8
Ϫ14.4
Ϫ11.6
Ϫ9.4
Ϫ6.9
Ϫ4.9
Ϫ2.7
Ϫ1.1
0.2
4.64
4.37
3.83
3.56
2.85
2.57
1.62
0.92
0.34
Ϫ0.39
—
Ϫ1.39
Ϫ1.70
Ϫ1.93
d
d
d
Ϫ12.8
Ϫ10.6
Ϫ8.1
Ϫ6.0
Ϫ3.9
Ϫ2.3
Ϫ1.0
1.5
4.18d
d
4.04
d
d
3.33d
d
3.10
d
d
Tetracene
2.04d
e
4
4
-Methyl-4Ј-methoxyazobenzene
-Methoxyazobenzene
1.46
e
e
0.91e
e
Azobenzene
0.10
2.7
e
3
2
,3Ј-Dimethoxyazobenzene
-Nitrobiphenyl
2.9
4.7
6.5
8.0
Ϫ0.15e
—
e
Ϫ1.32e
Ϫ1.30e
Ϫ1.70
5.8
7.6
9.1
e
Nitrobenzene
e
1
-Nitronaphthalene
a
b
c
EЊ measured in DMF versus ferrocene and corrected using EЊ
= 0.475 V. ∆GET calculated from EЊDMPP = Ϫ1. 38 V. ∆GET calculated from
Fc/Fcϩ
d
e
EЊBMPP = Ϫ1.43 V. Measured by homogeneous redox catalysis. Measured by a method using a rotating disk electrode.
Determination of the EЊdiss requires knowledge of the bond
dissociation free energy (BDFE), and the reduction potential
for the alkoxyl radical, EЊ
ؒ
Ϫ
, where F is the Faraday con-
RO /RO
stant (eqn. (24)). An estimate of the BDFE can be obtained
from the literature bond dissociation enthalpies (BDE) of
89
related peroxides and by including an entropy correction. The
value of EЊ can be estimated from cyclic voltammetry
ؒ
Ϫ
RO /RO
measurements provided the effect for the kinetic peak shift
88
is taken into account. Accounting for the peak shift,
EЊ (ACN) = Ϫ0.20 V and EЊ (DMF) =
ؒ
Ϫ
ؒ
MPPO /MPPO
Ϫ
MPPO /MPPO
Ϫ0.13 V vs. SCE. The standard potentials for the tert-butoxyl
radical were previously reported to be EЊ (ACN) =
ؒ
Ϫ
t-BuO /t-BuO
38
Ϫ0.30 V and EЊ
ؒ
Ϫ
(DMF) = Ϫ0.23 V vs. SCE. From
t-BuO /t-BuO
eqn. (24) and the data provided here, the EЊdiss for both DMPP
and BMPP are estimated to be within ± 0.10 V: EЊDMPP (ACN) =
Ϫ1.45 V, EЊDMPP (DMF) = Ϫ1.38 V, EЊ
and EЊBMPP (DMF) = Ϫ1.43 V vs. SCE.
(ACN) = Ϫ1.50 V,
BMPP
90
Likewise the thermochemical approach can be used to evalu-
ate the standard reduction potential of 2-methyl-1-phenyl-2-
propyl hydroperoxide, EЊMPPH. The BDE of hydroperoxides are
slightly larger than dialkyl peroxide ranging from 40 to 50 kcal
Ϫ1
mol . The BDE of MPPH was estimated from the heats of
formation for the homolytic cleavage of the O–O bond to be
Ϫ1 91,92
4
4.8 kcal mol .
Applying the same methodology, EЊMPPH
(
DMF) = Ϫ1.70 V, and EЊMPPH (ACN) = Ϫ1.77 V.
Discussion
Analysis of the kinetics
Knowing the values of EЊDMPP and EЊBMPP, the ∆G can be
ET
1
accurately calculated and plotted versus log (k hom) as shown in
Figs. 3a and 3b. The data are fit to a form of the Marcus expres-
5
2–54
sion derived for dissociative ET,
which relates the activation
Fig. 3 The homogeneous rate constants for reduction of (a) DMPP
and (b) BMPP measured by redox catalysis and chronopotentiometry.
The solid line through the data is the best fit according to eqns. (25) and
≠
free energy, ∆G , to the reaction free energy, ∆GET, and to the
Eyring expression (eqn. (25)). Savéant’s expression is similar to
≠
Ϫ1
the Marcus equation (eqn. (26)) with the exception that the
(26) with ∆G
0
= 13.0 kcal mol and a non-adiabatic log (κZhom) term.
≠
0
intrinsic barrier, ∆G , contains contributions from the BDE of
the cleaving bond (in this case the O–O bond) in addition to the
reorganisation energy, λ (eqn. (27)). Savéant’s theory, originally
applied to alkyl halides, assumed ET is adiabatic where the
transmission coefficient, κ, is equal to 1 and the pre-exponential
(
25)
11
Ϫ1 Ϫ1
factor, Z, is approximately 3 × 10
M
s . Previous studies by
(
(
26)
27)
our group and Maran and co-workers have revealed that ET
37,38,44
to peroxides and endoperoxides is non-adiabatic,
κ << 1, on the order of 0.01 to 0.02.
where
38
Using Savéant’s dissociative ET theory, two parameters can≠
be extracted directly from the homogeneous kinetic data, ∆G0
3
424
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9

View Article Online
and κZhom. The intrinsic barrier (eqn. (27)) was determined
Ϫ1
using an average BDE of 37 kcal mol , typical of dialkyl
peroxides, and a calculated λ using a previously documented
37,38
approach.
Estimates for λ can be determined using expres-
sions by Marcus, Kojima–Bard and Savéant and the radius of
9
3
the donor, r , and the radius of the peroxide acceptor, rAB.
D
Ϫ1
Ϫ1
With the Bard approach, λ = 95[(2r ) ϩ (2rAB)≠ Ϫ (rD
ϩ
D
Ϫ1
Ϫ1
rAB) ], the λ’s are 14.9 and 14.8 kcal mol and ∆G0 = 13.0 kcal
mol for both peroxides. This value for the intrinsic barrier was
Ϫ1
used to provide reasonable estimates of log(κZhom) by a curve
fitting analysis of the homogeneous kinetics. As seen in Figs. 3a
and 3b, the curve fits are parabolic as predicted by the theories
1
of ET. The fit of the log (k hom) versus ∆GET plot gives values of
log (κZhom) = 10.3 ± 0.1 for DMPP and log (κZhom) = 10.1 ± 0.1
for BMPP. Removing the data points at the slowest end from
the best fit of the BMPP kinetics has a minimal effect on the
extracted parameters. However, if ET to the peroxides was
94
assumed to be adiabatic (κ = 1) and log (Zhom) = 11.7 then
≠
Ϫ1
Ϫ1
∆
G0 = 15 kcal mol . With an estimated λ of 14.9 kcal mol ,
Ϫ1
the BDE would be 46 kcal mol , which is significantly larger
than the accepted BDE of dialkyl peroxides. A comparison
between the theoretical adiabatic pre-exponential factor
and the determined non-adiabatic value reveals a discrepancy
ranging from 1.4 to 1.6 orders of magnitude. The effect of steric
hindrance on ET to the σ* orbital was found to contribute
0
.8 log units, estimated from comparison of the heterogeneous
38
reduction of di-tert-butyl peroxide with di-n-butyl peroxide.
However, if an EЊdiss value of 0.1 V more positive is used to
calculate the ∆GET, an average between the thermochemical and
convolution determined EЊdiss, log (κZhom) decreases by 0.8 log
units. But given the error in EЊdiss, the pre-exponential factors
determined from the curve fitting analysis are consistent with
the results obtained from the temperature dependence study of
38
di-tert-butyl peroxide.
Recently, Savéant has addressed the analysis of the hetero-
geneous activation-driving force data by a quadratic expres-
Fig. 4 The heterogeneous rate constants for reduction of (a) DMPP
and (b) BMPP measured by convolution analysis. The solid line
95
sion. Equation (25) is still valid, but eqn. (26) includes a
proportionality factor of 1.07 in the intrinsic barrier. A similar
curve fitting analysis was performed with the heterogeneous
kinetics obtained from the convolution analysis as shown in
Figs. 4a and 4b.
through the data is the best fit according to eqns. (25) and (26) with a
≠
∆G
and a non-adiabatic log (κZhet) term as described in the text.
0
8
Ϫ1
a rate constant of 2.2 × 10 s . Recall from Scheme 1 that both
parallel competing reactions are irreversible so the rate constant
ratio is equal to the ratio of the product yields [MPPOH]/[tolu-
ene] provided the contributions from the following chemistry
(protonations) are negligible. Based on the heterogeneous elec-
trolysis studies, we showed the ratio of MPPOH to toluene
produced after ET is in excess of 100 to 1. Equating the
product ratio to the rate constant ratio, and from the knowledge
The heterogeneous kinetics was related to ∆GET using the
reported EЊdiss in Table 2. Reasonable values were extracted by
≠
fitting for both ∆G0 and log (κZhet) simultaneously. For DMPP
and BMPP, the intrinsic barriers were evaluated to be 10.7 ± 0.1
Ϫ1
and 11.4 ± 0.1 kcal mol , similar to the values reported in
Table 2, and the log (κZhet) are Ϫ0.46 ± 0.01 and Ϫ0.14 ± 0.01,
respectively. Using the thermochemical EЊ to calculate ∆G
diss
ET
≠
Ϫ1
has the effect of lowering ∆G0 no more than 1 kcal mol and
of k , and by accounting for the diffuse double layer at the
β
log (κZhet) by only 0.3 log, while setting log (κZ ) to the adia-
electrode interface, an inherent property of heterogeneous ETs,
het
96
≠
Ϫ1
2
batic value 3.6, yields a poor fit with a ∆G0 of 17 kcal mol .
The experimentally determined log (κZhet) and the hetero-
geneous adiabatic log (Z) differ by at least 3.7 log units, again
suggesting the ET is non-adiabatic. The greater difference in the
experimental log (κZhet) from the adiabatic value, in com-
parison to the homogeneous data, may be due to the hetero-
geneous data covering a narrower driving force range. It is also
possible that a distance effect because of the diffuse double
layer at the electrode interface may also contribute to a lower
heterogeneous pre-exponential factor. As for the evaluation of
k
can be calculated and expressed as a rate constant in units
het
Ϫ1
of cm s .
The distance dependence at the electrode interface is
accounted for using the Gouy-Chapman model of the diffuse
56
double layer. The model postulates that ET occurs at the
closest possible distance to the electrode, that being the outer
Helmholtz plane. For solutions of ACN and DMF with a rel-
ative permittivity equal to 37 at 25 ЊC, the expression κ = 4.79 ×
7
½
10 zC* can be used to characterise the thickness of the diffuse
layer, where z is the charge magnitude of each ion of the electro-
Ϫ1
Ϫ1
λ, eqn. (27) yields values ranging from 6 to 9 kcal mol whereas
lyte, C* is the bulk electrolyte concentration in mol L and κ is
Ϫ1 56
including the proportionality factor of 1.07 yields values 3 kcal
given in cm . The reciprocal of κ is a measure of the diffuse
Ϫ1
mol lower.
double layer thickness. For a 0.10 M TEAP solution, κ is calcu-
Ϫ8
lated to be 6.6 × 10 cm. Therefore, the product of κ and the
2
het
value extracted from the competition kinetics yields k equal
2
Determination of k
het
Ϫ1
to 1500 cm s . Incidentally, the driving force for reduction of
From the heterogeneous product analysis and from the simu-
lated concentration profiles, qualitatively we can conclude that
the second heterogeneous ET to the alkoxyl radical occurs at a
rate much faster than the rate of β-scission fragmentation with
the alkoxyl radical is thermodynamically favourable, based on
the potential difference between the peroxide peak potential
and the reduction potential of the alkoxyl radicals, in excess of
Ϫ1
1.8 V or 41 kcal mol .
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3425

View Article Online
Conclusions
capillary column. Product studies were quantified on a Hewlett
Packard 5890 Series II gas chromatograph interfaced to an
HP 3393A integrator with a 14 m × 0.32 mm × 0.25 µm HP
Electron transfer to O–O bonds is inherently slow, although the
peroxide bond is weak. Even for a highly driven ET, one must
consider that other chemical processes may be able to compete.
A comparison of the structurally related dialkyl peroxides,
dicumyl and di-tert-butyl peroxide, reveals the thermochemical
5
0ϩ capillary column. The densities of the peroxides were
determined from the mass of a known volume on an analytical
balance in a tared 10 µL Hamilton syringe. The syringe was
calibrated using 99ϩ% hexadecane with a density of 0.773 g
38,58
parameters and kinetics are similar
to those for the per-
Ϫ1
mL . The estimated error on the density measurements is 4%
oxides in this study, which is clearly evident when the kinetics
of each are plotted together. The significant difference with
DMPP and BMPP from their predecessors is in the follow-up
chemistry after cleavage of the O–O bond on accepting an
electron from a homogeneous donor. Heterogeneously, only
reduction of the MPPH alkoxyl radical was observed; homo-
geneously, a competition was observed between reduction and
β-scission followed by a second competition involving the
benzyl radical, which was characterised from product analysis
and by evaluating the competition q parameter. DMPP and
BMPP were primarily devised as kinetic probes to investigate
by the least-squares method.
Synthesis
Caution. Although no problems were encountered, dialkyl
peroxides and alkyl hydroperoxides are potentially explosive.
Syntheses were performed in a fumehood behind a safety
shield. The syntheses of di-2-methyl-1-phenyl-2-propyl per-
oxide (DMPP) and tert-butyl 2-methyl-1-phenyl-2-propyl per-
oxide (BMPP) were performed using a literature procedure with
99
some modifications by reacting 2-methyl-1-phenyl-2-propyl
bromide with the required hydroperoxide and silver trifluoro-
acetate. 2-Methyl-1-phenyl-2-propyl bromide was synthesised
from 2-methyl-1-phenyl-2-propanol with concentrated HBr and
the rate constant of the heterogeneous ET to alkoxyl radicals
2
(
k
het) liberated by DET. Now, with a thorough understanding
of the ET chemistry of DMPP and BMPP, these peroxides can
be used as mechanistic probes for the occurrence of DET. Stud-
ies utilizing these peroxides as mechanistic probes are currently
in progress. Aspects of the knowledge gained from this study
should assist those employing MPPH as a mechanistic probe by
providing more predictability power with a better understand-
ing of kinetics and thermodynamics.
100
LiBr.
2-Methyl-1-phenyl-2-propyl hydroperoxide (MPPH)
was synthesised from 2-methyl-1-phenyl-2-propanol using 50%
hydroperoxide and concentrated sulfuric acid
1
01
and from
99
2
-methyl-1-phenyl-2-propyl bromide. The crude dialkyl per-
oxide mixtures were subjected to NaBH in methanol to reduce
the major side product trifluoroacetic acid 1,1-dimethyl-2-
phenylethyl ester (identified by H, F, IR, MS and has a reduc-
tion potential of Ϫ2.39 V vs. SCE at 0.1 V s ), simplifying
4
1
19
Ϫ1
Experimental
purification by flash chromatography.
Materials
tert-Butyl 2-methyl-1-phenyl-2-propyl peroxide (BMPP). To
an oven-dried 100 ml 2-neck round-bottom flask containing a
Spectroscopic grade N,N-dimethylformamide was distilled
over CaH under a nitrogen atmosphere at reduced pressure.
Spectroscopic acetonitrile was fractionally distilled over CaH2
under a nitrogen atmosphere. The distilled solvents were stored
in glass bottles under an inert atmosphere in the dark. The
supporting electrolyte, tetraethylammonium perchlorate
2
1
Љ magnetic stir bar and immersed in an ice bath, 2-methyl-1-
phenyl-2-propyl bromide (5.0 g, 23 mmol) was added. tert-
Butyl hydroperoxide (2.5 g, 28 mmol), dried over Na SO , was
2
4
then added to the flask in slight excess followed by 20 mL of
pentane. Silver trifluoroacetate (5.3 g, 24 mmol) was added in
(
TEAP), was recrystallized three times from ethanol and stored
5
00 mg portions to the solution with stirring over a 30 min
under vacuum. The aromatic donors pyrene, anthracene, di-
phenylanthracene, fluoranthene, perylene, and 1-nitronaph-
thalene were of high purity and used as received. Isoquinoline
and nitrobenzene were distilled prior to use. Tetracene and
period and the reaction mixture stirred for 2 hours. The hetero-
geneous mixture was filtered into a 125 ml separatory funnel.
The solution was washed 3 times with 20 ml of water, dried over
Na SO , and the solvent removed by a rotary evaporator to give
2
4
2
4
-nitrobiphenyl were recrystallised from ethanol. 4-Methoxy-
Ј-methylazobenzene and 4-methoxyazobenzene were syn-
1
a crude mixture with BMPP in 65% yield by H NMR. The
mixture was subjected to NaBH in MeOH to reduce the tri-
4
thesised from the corresponding hydroxy-substituted azo-
benzene by reaction of methyl iodide with potassium carbon-
ate in acetone and recrystallised. 4-Hydroxy-4Ј-methylazo-
fluoroester product to the reagent alcohol (MPPOH). One mole
equivalent of NaBH to trifluoroester product was used. The
4
solution was allowed to stir immersed in an ice bath for 1 hour.
Afterwards 10 ml of 5% acetic acid were added to the solution,
which was stirred for an additional 15 minutes, and then
extracted 4× with 15 ml portions of CH Cl , washed 3× with
97
98
benzene and 3,3-dimethoxyazobenzene were prepared by
literature procedures. Trifluoroethanol was distilled, acetanilide
was recrystallised from water, and ferrocene sublimed prior to
use. All solvents and reagents not specified were used without
purification and are available from Aldrich. Flash chromato-
graphy was performed with silica gel 60 (230–400 mesh ASTM)
and fractions developed in hexanes on Kieselgel 60 F254 TLC
plates.
2
2
2
0 mL of water, dried over Na SO and the solvent removed by
2 4
a rotary evaporator. BMPP was purified and separated on silica
gel (1 : 100 ratio) by flash chromatography using 2% ethyl acet-
ate–98% petroleum ether (30–60 ЊC) in an overall yield of 50%.
Ϫ1
Clear, colourless liquid, density 1.01 g mL ; UV-vis (CH CN):
3
Ϫ1
Ϫ1
Ϫ1
λmax = 210 nm, ε = 8500 M cm , λ = 258 nm, ε = 200 M
cm ; νmax(NaCl)/cm 3064, 3029 (᎐C–H), 2977, 2933 (C–H),
max
Instrumentation
Ϫ1
Ϫ1
᎐
1
1
603 (C᎐C), 1198 (C–O), 871 (O–O); H NMR (CDCl ): δ 1.16
᎐
3
13
Nuclear magnetic resonance spectra were recorded on a Varian
Mercury spectrometer. H and C NMR spectra were recorded
at 400.1 and 100.6 MHz, respectively, with CDCl as the solvent
1
13
(
(
s, 6H), 1.23 (s, 9H), 2.84 (s, 2H), 7.16–7.27 (m, 5H); C NMR
CDCl ): δ 24.55, 26.68, 45.39, 78.51, 80.42, 125.95, 127.66,
3
3
1
30.68, 138.25; MS/EM found 222.162, calculated 222.162:
and are reported in ppm versus tetramethylsilane (δ = 0.00) for
1
13
m/z(intensity): 222(4), 131(29), 91(100), 73(75), 57(19).
H NMR and CDCl3 (δ = 77.00) for C NMR. Infra-red
spectra were recorded neat on a Bomem MB 100 FT-IR
spectrometer between NaCl plates. Mass spectrometry was
performed on a MAT 8200 Finnigan high resolution mass
spectrometer. GC/MS spectra were obtained with a Varian
Saturn 2000 GC/MS/MS interfaced to a Varian CP 3800 gas
chromatograph with a CD-Sil 5 CB-MS 30 m × 0.25 mm
Di-2-methyl-1-phenyl-2-propyl peroxide (DMPP). To an
oven-dried 100 ml 2-neck round-bottom flask containing a 1Љ
magnetic stir bar and immersed in an ice bath 2-methyl-1-
phenyl-2-propyl bromide (1.8 g, 8.4 mmol) and MPPH (1.4 g,
8.4 mmol) were added. To the flask was added approximately
3
426
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9

View Article Online
2
0 mL of pentane. Silver trifluoroacetate (1.5 g, 7.0 mmol) was
(5–12 times) with water (20mL), more so when DMF was the
added in 500 mg portions to the solution over a 15 min period
and the reaction mixture stirred for 3 hours. The work-up
procedure was the same as described for BMPP. DMPP was
separated and purified on silica gel (1 : 100 ratio) by flash
chromatography using petroleum ether (30–60 ЊC) in an overall
electrochemical solvent. At times, brine was added to break up
emulsions. An aliquot of the organic extract was analysed by
GC. The remaining solution was dried over sodium sulfate,
filtered and the solvent evaporated and the mass balance meas-
ured. For DMPP only one product was recovered: 2-methyl-1-
Ϫ1
1
13
yield of 30%. Pale yellow oil, density 1.13 g mL ; UV-vis
phenyl-2-propanol (MPPOH) in 65% yield. H and C NMR,
Ϫ1
Ϫ1
(
CH CN): λ = 210 nm, ε = 23000 M cm , λ = 258 nm,
and GC/MS were in agreement with those of an authentic
3
max
Ϫ1
max
Ϫ1
Ϫ1
1
ε = 370 M cm ; νmax(NaCl)/cm 3063, 3029 (᎐C–H), 2978,
934 (C–H), 1603 (C᎐C), 866 (O–O); H NMR (CDCl ): δ 1.18
s, 12H), 2.87 (s, 4H), 7.17–7.28 (m, 10H); C NMR (CDCl ):
sample. H NMR (CDCl , ppm): δ 1.23 (s, 6H), 1.39 (s, broad,
3
1
13
2
(
1H), 2.77 (s, 2H), 7.18–7.30 (m, 5H); C NMR (CDCl , ppm):
᎐
3
3
13
29.26, 49.75, 70.74, 126.34, 128.06, 130.29, 137.53. From prod-
3
1
δ 24.60, 45.60, 80.81, 126.01, 127.69, 130.72, 138.13; MS/EM
found 298.193, calculated 298.193: m/z(intensity): 298(1),
uct studies of BMPP, tert-butyl alcohol was detected by H
NMR as a singlet at 1.26 ppm. GC analysis of the cell contents
before work-up revealed no toluene. Electrolyses were per-
formed at least twice with each peroxide in each solvent with
minimum concentrations of 2 mM.
2
07(4), 149(8), 117(9), 91(100), 65(5).
Electrochemical procedures
Cyclic voltammetry was performing using a Perkin-Elmer PAR
Homogeneous product studies
2
83 or 263A potentiostat interfaced to a personal computer
with PAR 270 electrochemistry software. Experiments were per-
formed using a three electrode arrangement placed in a water-
jacket cell. A routine experimental setup is as follows: the cell
stored in an oven at 110 ЊC was placed in a copper Faraday cage
and purged with a continuous flow of high purity argon and
connected to a cooling bath maintained at 25 ЊC throughout the
entire experiment. The electrolyte, TEAP, was added to the cell
followed by 25 mL of the appropriate solvent and a ½Љ mag-
netic stir bar. The solution was bubbled with argon to expel
oxygen from the solution.
The cell setup was identical to the heterogeneous product
studies with the exception that a platinum foil was used for the
bulk electrolysis. Initially, 1–2 mM solution of mediator was
prepared and the reversible couple probed. Peroxide was added
to make a 2 mM concentration and the solution probed again
revealing an irreversible donor wave. The electrolysis was
conducted at approximately 100 mV past the cathodic peak
potential of the donor. The reaction was probed by cyclic
voltammetry and halted after disappearance of the peroxide
dissociative wave or after the catalytic wave of the donor was
once again completely reversible (I /I = 1). The electron count
The working electrode was a 3 mm diameter glassy carbon
rod (Tokai, GC-20) sealed in glass tubing. Prior to use it was
polished on a polishing cloth with diamond paste, rinsed and
sonicated in 2-propanol for 10 min and finally dried with a
stream of cool air. The working electrode was activated by
cycling 25 times between 0 and Ϫ2.8 V vs. SCE at a scan rate of
c
a
was typically 2.0 ± 0.2. Once the electrolysis was complete, 10
ml of water were added to the cell. The contents were extracted
with dichloromethane (3 × 8 mL) and washed 3 times with
water or brine (10 mL). The organic extract was dried over
sodium sulfate and filtered into a 25 mL volumetric flask. The
volume was topped-off and pentadecane added as an internal
standard. The sample was analysed by GC using calibration
curves consisting of a minimum of 5 points. Injections were
performed 2 or 3 times and found to be reproducible within 5%.
The temperature program was as follows: 30 ЊC for 3 min, ramp
Ϫ1
2
0
.2 V s . The counter electrode was a 1 cm Pt plate. The
reference electrode was a silver wire immersed in a glass tube
with a sintered end containing 0.10 M TEAP in the relevant
solvent. After each experiment, it was calibrated against the
ferrocene/ferricenium couple. In 0.10 M TEAP, the EЊ was
taken to be 0.475 and 0.449 V versus KCl saturated calomel
electrode (SCE) in DMF and ACN, respectively.
Ϫ1
Ϫ1
10 ЊC min to 250 ЊC for 10 min at a flow rate of 90 mL min .
1
In some cases, the sample was also analysed by GC/MS and H
The standard reduction potentials of the homogeneous
donors were determined in a standard cell at several scan rates
between 0.1 and 2.0 V s using 1–5 mM concentrations of
donor. Afterwards, 1–3 mM ferrrocene was added to the cell
and a series of cyclic voltammograms at identical sweep rates
were acquired. The standard potentials of ferrocene and the
aromatic donors were taken as the midpoint between the anodic
and cathodic peak potentials.
NMR (after evaporation). Identified in the reaction between
azobenzene and DMPP were two coupling products between
azobenzene and a benzyl radical: m/z(intensity): (isomer 1)
272(95), 195(7), 167(100), 165(34), 152(29), 105(11), 77(38);
(isomer 2) 272(58), 195(100), 180(32), 165(33), 152(18), 77(21).
Ϫ1
Kinetic measurements
Homogeneous rate constants were measured in DMF at 25 ЊC
by two methods. The faster kinetics were measured using the
set-up described above and the technique of homogeneous
redox catalysis. A cyclic voltammograph of the reversible
couple of the donor was recorded at 5–7 sweep rates between
Heterogeneous product studies
Constant potential electrolyses were performed with a Perkin-
Elmer PAR 263A potentiostat using the same cell setup as men-
tioned previously with the following exceptions: the working
electrode was an EDI101 RDE electrode with a 12 mm tip (or a
platinum foil) and the counter electrode was a platinum wire
immersed in a glass frit containing a 2 mm layer of neutral
alumina. A 3 mm glassy carbon electrode was used to probe the
concentration of peroxide. After cell setup, the electrolyte solu-
tion was pre-electrolysed at a potential over 100 mV negative to
the peroxide reduction peak potential. The experiments were
performed at approximately 100 mV negative to the peak
potential of the peroxide and continued until less than 3% of
the initial peak current remained. In some cases, the electrolysis
was continued until the initial peak current was reduced to the
initial background. The electron count was typically 2.0 ± 0.2.
Once the electrolysis was complete, dilute acetic acid or water
Ϫ1
0.1 and 2.0 V s at 1 and 2 mM concentrations of donor in the
absence of peroxide. Afterwards, measurements were per-
formed with at least 3 different concentrations of peroxide with
excess factors (the concentration ratio of peroxide to donor)
ranging from 0.5 with the most catalytic donors up to values of
20 with the least catalytic. Peak ratios were determined from the
cathodic peak current in the presence of peroxide relative to the
initial peak current in the absence of peroxide. The voltam-
metric behaviour of the donor peak currents was then simu-
lated using digital simulation software (Digisim 3.0 software by
Bioanalytical Systems Inc.). Kinetic measurements with each
peroxide were performed 2 or 3 times.
The slower kinetics were measured using a 3 mm rotating
disk electrode to monitor the decay of donor radical anion
upon addition of peroxide. A donor solution of 3–5 mM was
reduced at a Pt foil to generate a 1 mM solution of radical
(
10 mL) was added to the cell and the contents extracted with
dichloromethane (4 × 10 mL). The organic extract was washed
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3427

View Article Online
Ϫ1
anion (2.5 F mol was added). The solution was stirred at 4500
14 Z. Hu and S. M. Gorun, in Biomimetic Oxidations Catalyzed by
Transition Metal Complexes, ed. B. Meunier, Imperial College Press,
London, 2000, pp. 269–307.
rpm and the stability of the radical anion monitored at one
minute intervals. While stirring, a greater than 10-fold excess of
peroxide was added to the solution and the exponential decay
of the anodic current was recorded for a minimum of one half-
life. Rate constants were extracted from the slopes of ln(I ) vs.
time plots discarding the initial portion of the decay curve. The
plots were linear with correlation coefficients greater than 0.99.
1
5 A. C. Rosenzweig and S. J. Lippard, Acc. Chem. Res., 1994, 27, 229–
2
36.
1
6 M. Sono, M. P. Roach, E. D. Coulter and J. H. Dawson, Chem. Rev.,
1996, 96, 2841–2887.
17 J. T. Groves and Y.-Z. Han, in Cytochrome P450: Structure,
nd
Mechanism and Biochemistry, 2 edn., Plenum Press, New York,
1
1995, pp. 3–48.
The slope of the line corresponds to Ϫ2k hom[ROOR]. Each
measurement was performed 2 or 3 times.
1
1
2
2
8 R. A. Leising, R. E. Norman and L. Que, Jr., Inorg. Chem., 1990, 29,
553–2555.
9 R. A. Leising, Y. Zang and L. Que, Jr., J. Am. Chem. Soc., 1991, 113,
555–8557.
2
Competition parameter q measurements
8
0 Y. Zang, J. Kim, Y. Dong, E. C. Wilkinson, E. H. Appelman and
L. Que, Jr., J. Am. Chem. Soc., 1997, 119, 4197–4205.
1 H. Miyake, K. Chen, S. J. Lange and L. Que, Jr., Inorg. Chem., 2001,
40, 3534–3538.
Measurements were performed using the technique developed
82
by Kim Daasbjerg using either an RDE or a microultra-
electrode. The RDE was a Radiometer Copenhagen EDI101
with a 3 mm tip connected to a CTV 101 speed control unit.
The microultraelectrode was a 10 µm Pt electrode from
Autolab. A donor solution of 2–5 mM in DMF was prepared
and vigorously purged with argon. A voltammogram was
recorded using an RDE spinning at 4500 rpm or a 10 µm Pt
microelectrode. The donor was subsequently reduced at a Pt foil
to generate a 2–3 mM solution of radical anion (4.8–7.2 F
22 S. J. Lange, H. Miyake and L. Que, Jr., J. Am. Chem. Soc., 1999, 121,
330–6331.
6
2
3 M. P. Jensen, S. J. Lange, M. P. Mehn, E. L. Que and L. Que, Jr.,
J. Am. Chem. Soc., 2003, 125, 2113–2128.
4 J.-U. Rohde, J.-H. In, M. H. Lim, W. W. Brennessel, M. R.
Bukowski, A. Stubna, E. Münck, W. Nam and L. Que, Jr., Science,
2003, 299, 1037–1039.
2
25 K. Chen and L. Que, Jr., Chem. Commun., 1999, 1375–
376.
Ϫ1
1
mol added). Afterwards, a series of voltammograms were
2
6 I. W. C. E. Arends, K. U. Ingold and D. D. M. Wayner, J. Am. Chem.
Soc., 1995, 117, 4710–4711.
7 D. W. Snelgrove, P. A. MacFaul, K. U. Ingold and D. D. M. Wayner,
Tetrahedron Lett., 1996, 37, 823–826.
28 P. A. MacFaul, K. U. Ingold, D. D. M. Wayner and L. Que, Jr.,
taken at one minute intervals to ensure the radical anion was
stable and no oxygen was present in the cell and then approx-
imately 0.5 mM peroxide was added to the radical anion solu-
tion with stirring. Voltammograms were recorded at regularly
timed intervals until no significant changes were observed after
2
J. Am. Chem. Soc., 1997, 119, 10594–10598.
9 P. A. MacFaul, I. W. C. E. Arends, K. U. Ingold and D. D. M.
Wayner, J. Chem. Soc., Perkin Trans. 2, 1997, 135–146.
2
3
consecutive scans. At the concentrations used, the fastest
reactions were complete within seconds whereas those with
azobenzene took approximately 8 minutes. Measurements were
performed at least 3 times.
3
0 P. A. MacFaul, D. D. M. Wayner and K. U. Ingold, Acc. Chem. Res.,
1
998, 31, 159–162.
3
3
1 C. Walling, Acc. Chem. Res., 1998, 31, 155–157.
2 K. Neimann, R. Neumann, A. Rabion, R. M. Buchanan and
R. H. Fish, Inorg. Chem., 1999, 38, 3575–3580.
Acknowledgements
3
3 A. Rabion, S. Chen, J. Wang, R. M. Buchanan, J.-L. Seris and
R. H. Fish, J. Am. Chem. Soc., 1995, 117, 12356–12357.
4 C. Walling, Acc. Chem. Res., 1975, 8, 125–131.
5 M. J. Perkins, Chem. Soc. Rev., 1996, 25, 229–236.
6 K. U. Ingold and P. A. MacFaul, in Biomimetic Oxidations
Catalyzed by Transition Metal Complexes, ed. B. Meunier, Imperial
College Press, London, 2000, pp. 45–89.
Financial support from the Natural Sciences and Engineering
Research Council of Canada (NSERC), the University of
Western Ontario (ADF Fund), the Canadian Foundation
for Innovation (CFI), the Ontario Research Challenge and
Development Fund (ORCDF), and the Premier’s Research
Excellence Award is gratefully acknowledged. DCM thanks
the provincial government and the Faculty of Graduate
Studies for an OGSST scholarship.
3
3
3
37 (a) R. L. Donkers and M. S. Workentin, Chem. Eur. J., 2001, 7,
4
012–4020; (b) M. S. Workentin and R. L. Donkers, J. Am. Chem.
Soc., 1998, 120, 2664–2665.
3
8 (a) R. L. Donkers, F. Maran, D. D. M. Wayner and M. S.
Workentin, J. Am. Chem. Soc., 1999, 121, 7239–7248; (b) M. S.
Workentin, F. Maran and D. D. M. Wayner, J. Am. Chem. Soc.,
References
1
995, 117, 2120–2121.
1
2
3
D. H. R. Barton and D. Doller, Acc. Chem. Res., 1992, 25, 504–512.
D. H. R. Barton and W. Chavasiri, Tetrahedron, 1994, 50, 19–30.
(a) D. H. R. Barton, Chem. Soc. Rev., 1996, 237–239; (b) D. H. R.
Barton, Synlett, 1996, 229–230.
39 R. L. Donkers, J. Tse and M. S. Workentin, Chem. Commun., 1999,
135–136.
40 D. L. B. Stringle, R. N. Campbell and M. S. Workentin, Chem.
Commun., 2003, 1246–1247.
4
5
6
T. Kojima, R. A. Leising, S. Yan and L. Que, Jr., J. Am. Chem. Soc.,
41 R. L. Donkers and M. S. Workentin, J. Phys. Chem. B, 1998, 102,
1
993, 115, 11328–11335.
4061–4063.
R. A. Leising, J. Kim, M. A. Pérez and L. Que, Jr., J. Am. Chem.
Soc., 1993, 115, 9524–9530.
42 D. C. Magri, R. L. Donkers and M. S. Workentin, J. Photochem.
Photobiol. A: Chem., 2001, 138, 29–34.
43 S. Antonello, M. Musumeci, D. D. M. Wayner and F. Maran, J. Am.
Chem. Soc., 1997, 119, 9541–9549.
44 S. Antonello, F. Formaggio, A. Moretto, C. Toniolo and F. Maran,
J. Am. Chem. Soc., 2001, 123, 9577–9584.
45 S. Antonello, F. Formaggio, A. Moretto, C. Toniolo and F. Maran,
J. Am. Chem. Soc., 2003, 125, 2874–2875.
J. Kim, R. G. Harrison, C. Kim and L. Que, Jr., J. Am. Chem. Soc.,
1
996, 118, 4373–4379.
7
8
K. Chen and L. Que, Jr., J. Am. Chem. Soc., 2001, 123, 6327–6337.
K. Chen, M. Costas and L. Que, Jr., J. Chem. Soc., Dalton Trans.,
2
002, 5, 672–679.
D. T. Sawyer, A. Sobkowiak and T. Matsushita, Acc. Chem. Res.,
996, 29, 409–416.
9
1
46 S. Antonello, M. Crisma, F. Formaggio, A. Moretto, F. Taddei,
C. Toniolo and F. Maran, J. Am. Chem. Soc., 2002, 124, 11503–
11513.
1
0 (a) P. R. Ortiz de Montellano, Cytochrome P450: Structure,
Mechanism and Biochemistry, Plenum Press, New York, 1995;
(
b) J. T. Groves, J. Chem. Educ., 1985, 62, 928–931.
47 S. Antonello and F. Maran, J. Am. Chem. Soc., 1999, 121, 9668–
1
1 (a) I. Schlichting, J. Berendzen, K. Chu, A. M. Stock, S. A. Maves,
D. E. Benson, R. M. Sweet, D. Ringe, G. A. Petsko and S. G. Sligar,
Science, 2000, 287, 1615–1622; (b) D. G. Kellner, S.-C. Hung,
K. E. Weiss and S. G. Sligar, J. Biol. Chem., 2002, 277, 9641–9644.
2 J. L. McLain, J. Lee and J. T. Groves, in Biomimetic Oxidations
Catalyzed by Transition Metal Complexes, ed. B. Meunier, Imperial
College Press, London, 2000, pp. 91–169.
3 B. Meunier, in Biomimetic Oxidations Catalyzed by Transition Metal
Complexes, ed. B. Meunier, Imperial College Press, London, 2000,
pp. 171–214.
9676.
48 S. Antonello and F. Maran, J. Am. Chem. Soc., 1997, 119, 12595–
12600.
49 D. A. Casteel, Nat. Prod. Rep., 1999, 16, 55–73.
50 D. A. Casteel, Nat. Prod. Rep., 1992, 9, 289–312.
51 Organic Peroxides, ed. W. Ando, John Wiley & Sons Ltd, New York,
1992.
52 J.-M. Savéant, in Advances in Physical Organic Chemistry,
ed. T. Tidwell, Academic Press, New York, 2000, vol. 35, pp. 117–
192.
1
1
3
428
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9

View Article Online
5
3 J.-M. Savéant, in Advances in Electron Transfer Chemistry, ed. P. S.
Mariano, JAI Press Inc., Greenwich, CT, 1994, vol. 4, pp. 53–116.
4 J.-M. Savéant, Acc. Chem. Res., 1993, 26, 455–461.
5 K. Daasbjerg, S. U. Pedersen and H. Lund, in General Aspects of the
Chemistry of Radicals, ed. Z. B. Alfassi, John Wiley & Sons, New
York, 1999, pp. 385–427.
6 A. J. Bard and L. R. Faulkner, Electrochemical Methods,
Fundamentals and Applications, 2nd edn., Wiley, New York, 2001.
7 The Butler–Volmer model of electrode kinetics implies the transfer
coefficient is a constant over a potential range, which is contrary to
the observed experimental results. See reference 56.
8 F. Maran, D. D. M. Wayner and M. S. Workentin, in Advances in
Physical Organic Chemistry, ed. T. Tidwell and J. P. Richard,
Academic Press, New York, 2001, vol. 36, pp. 85–166.
76 D. Occhialini, K. Daasbjerg and H. Lund, Acta Chem. Scand.,
1993, 47, 1100–1106.
77 R. Fuhlendorff, D. Occhialini, S. U. Pedersen and H. Lund,
Acta Chem. Scand., 1989, 43, 803–806.
78 S. U. Pedersen, T. Lund, K. Daasbjerg, M. Pop, I. Fussing and
H. Lund, Acta Chem. Scand., 1998, 52, 657–671.
79 S. U. Pedersen and T. Lund, Acta Chem. Scand., 1991, 45, 397–
402.
80 T. Lund, P. Christensen and R. Wilbrandt, Org. Biomol. Chem.,
2003, 1, 1020–1025.
81 A. Cardinale, A. A. Isse and A. Gennaro, Electrochem. Commun.,
2002, 4, 767–772.
82 K. Daasbjerg, Acta Chem. Scand., 1993, 47, 398–402.
83 R. A. Marcus and N. Sutin, Biochim. Biophys. Acta, 1985, 811,
5
5
5
5
5
2
65–322.
4 D. D. M. Wayner, D. J. McPhee and D. Griller, J. Am. Chem. Soc.,
988, 110, 132–137.
5 P. Hapiot, V. V. Konovalov and J.-M. Savéant, J. Am. Chem. Soc.,
995, 117, 1428–1434.
6 S. U. Pedersen and K. Daasbjerg, Acta Chem. Scand., 1989, 43,
01–303.
87 K. Daasbjerg, S. U. Pedersen and H. Lund, Acta Chem. Scand.,
5
6
6
9 J. C. Imbeaux and J.-M. Savéant, J. Electroanal. Chem., 1973, 44,
8
8
8
1
69–187.
1
0 J.-M. Savéant and D. Tessier, J. Electroanal. Chem., 1975, 61, 251–
2
63.
1
1 The background subtracted voltammograms are transformed to
sigmoidal-shaped i–E curves by use of the convolution integral and
3
the experimental current i.
The limiting
1
991, 45, 424–430.
current, Ilim, at the plateau of the sigmoidal-shaped i–E curves is
under diffusion control and defined as Ilim = nFAD C* where n is
the overall electron consumption, A is the electrode area, D is the
88 D. D. M. Wayner and V. D. Parker, Acc. Chem. Res., 1993, 26, 287–
½
294.
Ϫ1
89 A comparison between di-n-propyl (37.1 kcal mol ), di-isopropyl
Ϫ1
Ϫ1
diffusion coefficient, and C* is the substrate concentration.
2 (a) For an irreversible electrode process Ilim and i are related to k
(37.7 kcal mol ), and di-tert-butyl peroxide (37.4 kcal mol )
1
6
demonstrates that the BDE of acyclic peroxides is relatively
independent of substituent. The BDFEs of BMPP and DMPP
were estimated assuming the same BDE as DTBP and an entropy
correction of ∆S = 28.5 cal mol
BMPP and DMPP are calculated to be 28.9 kcal mol .
0 In the calculation of EЊBMPP, an average value for EЊ
used.
1 NIST Standard Reference Database; NIST Standard Reference
Database and Properties Database and Estimation Program, U. S.
Department of Commerce, Gaithersburg, MD, 2003.
2 S. W. Benson, Thermochemical Kinetics, 2nd edn., Wiley, New
York, 1976.
3 The radius of BMPP and DMPP was determined from the density
of the compound using the equation r (Å) = 10 [(3M/4πρN) ].
Upon ET the charge is localised on the oxygen atom so an effective
radius approach is used to better represent the size and molecular
portion where the electron is accepted. The effective radius was
het
by
; (b) Derivatisation of the ln
Ϫ1
Ϫ1
K
(at 25 ЊC). The BDFEs for
1
het
Ϫ1
k
allows the evaluation of the apparent transfer coefficient, α,
9
9
ؒ
Ϫ
was
RO /RO
uncorrected for the double layer.
.
6
6
3 T. Lund and H. Lund, Acta Chem. Scand., Ser. B, 1987, 41, 93–102.
4 N. T. Kjær and H. Lund, Acta Chem. Scand., 1995, 49, 848–852 .
The rate constant for hydrogen atom abstraction by the tert-butoxyl
9
9
6
Ϫ1 Ϫ1
radical was estimated to be 5 × 10 M
s .
6
5 No trans-stilbene was observed as reported in the indirect reduction
of some benzyl halides. See reference 63. Incidentally, dihydro-
anthracene and stilbene have the identical exact mass and similar
GC retention times.
–
1
3
53
8
6
6 C. L. Perrin, J. Wang and M. Szwarc, J. Am. Chem. Soc., 2000, 122,
4
569–4572.
eff
determined using rAB = [r (2r Ϫ r )/rAB], where rAB is the radius
B
AB
B
6
7 J. A. Howard and J. Lusztyk, Radical Reaction Rates in Liquids, ed.
H. Fischer, Landolt-Börnstein Group II, Springer-Verlag, Berlin,
vol. 18, subvol. D1, 1997.
8 C. P. Andrieux, A. Le Gorande and J.-M. Savéant, J. Am. Chem.
Soc., 1992, 114, 6892–6904.
9 H. Lund, K. Daasbjerg, T. Lund, D. Occhialini and S. U. Pedersen,
Acta Chem. Scand., 1997, 51, 135–144.
0 C. P. Andrieux, J. M. Dumas-Bouchiat and J.-M. Savéant,
J. Electroanal. Chem., 1978, 87, 39–53.
1 C. P. Andrieux, C. Blocman, J. M. Dumas-Bouchiat, F. M’Halla and
J.-M. Savéant, J. Electroanal. Chem., 1980, 113, 19–40.
2 J.-M. Savéant and K. B. Su, J. Electroanal. Chem., 1985, 196, 1–22.
3 L. Nadjo, J.-M. Savéant and K. B. Su, J. Electroanal. Chem., 1985,
of peroxide and rB is the effective radius of the alkoxide anion.
From the peroxide density radii, 4.44 and 4.71 Å for BMPP and
DMPP, the effective radii were calculated as 2.83 and 2.87 Å,
respectively. An average rD of 3.80 Å was used for the donors.
H. Kojima and A. J. Bard, J. Am. Chem. Soc., 1975, 97, 6317–6324.
6
6
7
7
2
½
9
4 Zhom = (rD ϩ rAB) (8πRT/µ) where r and rAB are the radius of the
donor and acceptor, and µ is the reduced mass. See reference 93.
95 J.-M. Savéant, J. Phys. Chem. B, 2002, 106, 9387–9395.
96 Zhet = (RT/2πM) where M is the molar mass.
97 K. Haghbeen and E. W. Tan, J. Org. Chem., 1998, 63, 4503–4505.
98 E. C. Horning, in Synthesis of Azobenzene, vol. 3, Wiley, London,
1955, p. 103.
99 P. G. Cookson, A. G. Davies and B. P. Roberts, J. Chem. Soc.,
Chem. Commun., 1976, 1022–1023.
D
½
7
7
1
96, 23–34.
4 D. Occhialini, S. U. Pedersen and H. Lund, Acta Chem. Scand.,
990, 44, 715–719.
7
7
100 H. Masada and Y. Murotani, Bull. Chem. Soc. Jpn., 1980, 53,
1
1181–1182.
5 D. Occhialini, J. S. Kristensen, K. Daasbjerg and H. Lund,
Acta Chem. Scand., 1992, 46, 474–481.
101 R. R. Hiatt and W. M. J. Strachan, J. Org. Chem., 1963, 28, 1893–
1894.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 4 1 8 – 3 4 2 9
3429