Regioselective S-O vs. C-O bond cleavage in sulfenate ester radical anions

Article abstract of DOI:10.1139/v05-164

The electron transfer (ET) reduction of benzyl benzenesulfenate ester (1) and tert-butyl benzenesulfenate ester (2) was investigated using electrochemical techniques. Analysis of the cyclic voltammetry of each compound suggests that the ET reduction proceeds via a stepwise dissociative mechanism. The voltammograms of 1 are similar to those of diaryl disulfides and it was found through controlled potential electrolysis (CPE) product studies that ET reduction leads to S-O bond cleavage. The voltammograms of 2 are dramatically different with a sharper dissociative wave occurring at a more negative peak potential. CPE experiments indicate products that result from ET leading to C-O bond cleavage in this case. DFT calculations of the singly occupied molecular orbitals (SOMOs) of 1 and 2 were performed and offer a rationale for the different reactivity of the two radical anions.

Full text of DOI:10.1139/v05-164

1473
Regioselective S—O vs. C—O bond cleavage in
sulfenate ester radical anions1
Donald L.B. Stringle and Mark S. Workentin
Abstract: The electron transfer (ET) reduction of benzyl benzenesulfenate ester (1) and tert-butyl benzenesulfenate es-
ter (2) was investigated using electrochemical techniques. Analysis of the cyclic voltammetry of each compound sug-
gests that the ET reduction proceeds via a stepwise dissociative mechanism. The voltammograms of 1 are similar to
those of diaryl disulfides and it was found through controlled potential electrolysis (CPE) product studies that ET re-
duction leads to S—O bond cleavage. The voltammograms of 2 are dramatically different with a sharper dissociative
wave occurring at a more negative peak potential. CPE experiments indicate products that result from ET leading to
C—O bond cleavage in this case. DFT calculations of the singly occupied molecular orbitals (SOMOs) of 1 and 2
were performed and offer a rationale for the different reactivity of the two radical anions.
Key words: sulfenate esters, dissociative electron transfer, electrochemistry, radical anions.
Résumé : On a étudié la réaction de réduction par transfert d’électron (TE) du benzènesulfénate de benzyle (1) et du
benzènesulfénate de tert-butyle (2) en faisant appel à des techniques électrochimiques. L’analyse de la voltampéro-
métrie cyclique de chaque composé suggère que la réduction par TE se produit par un mécanisme par dissociations
successives. Le voltampérogramme du composé 1 est semblable à celui des disulfures de diaryles et, sur la base
d’études des produits par électrolyse à potentiel contrôlé (EPC), on a trouvé que la réduction par TE conduit à une
coupure de la liaison S—O. Le voltampérogramme du composé 2 est dramatiquement différent et il présente une vague
dissociative qui se présente à un pic de potentiel beaucoup plus négatif. Les études d’EPC indiquent que, dans ce cas,
les produits qui résultent d’un TE proviennent d’une coupure d’une liaison C—O. Des calculs effectués sur la base de
la théorie de la densité fonctionnelle des orbitales moléculaires à occupation singulet (SOMO) des composés 1 et 2 et
ils permettent de présenter une rationalisation pour les réactivités différentes des deux radicaux anions.
Mots clés : esters de l’acide sulfénique, transfert d’électron avec dissociation, électrochimie, radicaux anions.
[Traduit par la Rédaction] Stringle and Workentin 1482
Introduction
fragment (3). The fragmentation pathway led to a homoge-
neous ET-fragmentation radical anion chain reaction that
was responsible for a potential dependence on the product
ratios of controlled potential electrolysis (CPE) experiments.
Both cases demonstrate the unique reactivity that can be ob-
served upon generation of radical anion species by electro-
chemical means, and how the reactivity can be controlled by
adjusting the potential at which they are formed. In this
sense, the electrochemical generation of radical anions
makes for ideal circumstances to investigate their reactivity.
The study of sulfenate esters developed as an extension to
our work on the ET reduction of peroxides (4–7) and endo-
peroxides (1–3, 8–10) and the work of others concerning the
ET reduction of diaryl disulfides (11–14). The ET reduction
of the O—O bonds in peroxides and endoperoxides and the
S—S bonds in diaryl disulfides are characterized by unusu-
A major focus of our group has been the study of dis-
sociative electron transfer (ET) reactions through electro-
chemical methods. The work provides an opportunity to
investigate many aspects of ET reactions, but also allows for
investigation into the reactivity of the intermediate species
formed upon ET. Studies of 9,10-diphenylanthracene endo-
peroxide found that upon ET reduction, O—O bond cleav-
age occurred generating an alkoxyl radical (1, 2). It was
observed that an O-neophyl type rearrangement at the
alkoxyl radical could occur in competition with its ET re-
duction. Similarly, the dissociative ET reduction of 3,3,6,6-
tetraaryl-1,2-dioxanes involved formation of a distonic radi-
cal anion (where the radical and anionic centres are spatially
and electronically distinct), which could be reduced or could
‡
ally large values for the intrinsic barrier for the ET (∆G0 ).
Received 27 April 2005. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 10 November
2005.
For the ET reduction of peroxides and endoperoxides the ET
is shown to cleave the O—O bond via a concerted dis-
sociative mechanism where electron uptake and bond cleav-
age are simultaneous. In this case, the large intrinsic barrier
is a result of a contribution from the bond dissociation en-
ergy of the fragmenting bond. In contrast, the reduction of
diaryl disulfides was shown to follow a stepwise dissociative
mechanism, with the formation of a radical anion intermedi-
ate that leads to fragmentation of the S—S bond in a second
D.L.B. Stringle and M.S. Workentin.2 Department of
Chemistry, The University of Western Ontario, London, ON
N6A 5B7, Canada.
1This article is part of a Special Issue dedicatted to organic
reaction mechanisms.
2Corresponding author (e-mail: mworkent@uwo.ca).
Can. J. Chem. 83: 1473–1482 (2005)
doi: 10.1139/V05-164
© 2005 NRC Canada

1474
Can. J. Chem. Vol. 83, 2005
Table 1. Data obtained from the CVs of 1 and 2 (1.0–3.0 mmol/L) in 0.1 mol/L TBAP–DMF solution.
Parameters
Scan rates (V/s)
0.1
1
1 + acid/MeOTsa
2
2 + acidb
–1.68
–1.67
–2.28
–2.29
Ep (V vs. SCE)
1.0
10.0
0.1
–1.76
–1.89
117
–1.75
–1.89
96
–2.36
–2.48
83
–2.35
–2.43
86
∆ = (Ep/2 – Ep) (mV)
1.0
10.0
124
156
107
91
127
164
95
96
110
170
129
144
0.58
1/2
Ip /ν (mA s1/2 V–1/2 c
)
RT
0.1
1.0
0.41
0.50
0.60
α =1.85
∆
F
0.38
0.31
0.35
0.52
0.37
0.38
0.50
0.37
0.28
0.55
0.49
0.41
⎛
⎞
⎟
RT ∂ ln v
⎜
αapp
=
⎜
⎟
⎠
nF ∂Ep
⎝
∂Ep/∂ log ν (mV)
n = Q/mF (F/mol)d
–83
1.1
–78
1.9
–103
1.1
–72
2.0
aAverage of multiple experiments where either acetanilide or TFE was used in at least two times excess in combination with at
least 1 equiv. of MeOTs.
bCVs preformed in the presence of at least a two times excess of acetanilide.
cCorrected for a 2 mmol/L solution.
dEquation determines the stoichiometry of an electroysis based on the charge introduced into the cell (Q), the mol of analyte (m),
and Faraday constant.
peroxides and disulfides. In this study the ET reduction of
benzyl benzenesulfenate ester (1) and tert-butyl benzene-
sulfenate ester (2), shown in Scheme 1, was examined.
These substrates vary only in the ester functionality, which
is ideal for probing the impact of this group on the reactiv-
ity. Both substrates are derivatives of benzenesulfenic acid,
which provides a convenient handle for comparison to diaryl
disulfides. The work reported here details the analysis of the
cyclic voltammetry (CV) and the CPE for each compound to
determine the mechanism of the initial ET and investigate
the reactivity of any radical anion species formed through
ET reduction.
Scheme 1.
S
S
O
O
1
2
step. For the disulfides, the large intrinsic barrier was shown
to be the result of contributions due to a significant S—S
bond stretching upon ET forming a loosely bound radical
anion intermediate. To further examine the factors that gov-
ern the stepwise vs. concerted mechanism in these types of
bonded systems, we initiated this study of the ET to
sulfenate esters, which contain a single S—O bond.
Results and discussion
Although isoelectronic with peroxides (O—O) and disul-
fides (S—S), the S—O bonded sulfenate esters might rea-
sonably be expected to show some differences in reactivity
because the relative size and electronegativity of the two at-
oms in the bond are different. Because the chemistry of a
compound initiated by ET and photochemical excitation of-
ten follow similar pathways, then it is constructive to note
that in the many investigations of the photochemistry of
sulfenate esters it is uniformly demonstrated that S—O bond
cleavage occurs upon excitation (15, 16), similar to both
peroxides and disulfides that undergo photolytic O—O and
S—S cleavage, respectively. This offers reason to suspect
that upon ET reduction these compounds would then un-
dergo S—O cleavage as well. However, a computational
study by Gregory and Jenks (17) suggests that compared to
R-O-O-R′ peroxides, sulfenic esters (R-S-O-R′) can have
much weaker R′—O bonds than the stronger central S—O or
the corresponding R—S bonds. This suggests that sulfenic
esters may indeed exhibit different reactivity compared to
ET reduction of benzyl benzenesulfenate ester (1)
The cyclic voltammetry of 1 was performed in dimethyl-
formamide (DMF) with 0.1 mol/L tetrabutylammonium per-
chlorate (TBAP) as supporting electrolyte. The pertinent CV
data for both 1 and 2 are listed in Table 1. A representative
CV of 1 in the absence of acid is shown in Fig. 1a. When
scanning cathodically, a broad wave with low peak current is
observed at –1.68 V vs. SCE at 0.1 V/s, which is assigned as
the dissociative wave of 1. This wave is irreversible at all
scan rates investigated (0.1–50.0 V/s). The peak current
corrected for scan rate and concentration is 107 µA for a
2 mmol/L solution. The peak current provides an indication
of the number of electrons involved in the reduction step.
The Ep shifts negatively with increasing scan rate by
83 mV/decade. The peak breadth also increases, from values
of 117 to 124 mV at 0.1 and 1.0 V/s, respectively, which
leads to the corresponding decrease in the transfer coeffi-
cient (α), with values of 0.41 and 0.38 at the same scan
© 2005 NRC Canada

Stringle and Workentin
1475
Fig. 1. Representative CVs of 1 in 0.1 mol/L TBAP–DMF at
0.1 V/s: (a) 1 (2.3 mmol/L) before (solid line) and after (dashed
line) introduction of 6.1 C of charge into the cell; (b) 1
(2.2 mmol/L) in the presence of acetanilide (40 mmol/L) and
MeOTs (40 mmol/L) before (solid line) and after (dashed line)
introduction of 10.3 C of charge into the cell.
Scheme 2. Proposed mechanism for the ET reduction of 1.
e
S
S
Ph
Ph
O
O
1
Ph
Ph
O
Ph
S
O
Ph
e
Ph
S
O
Ph
3
Ph
O
Ph
4
nate mode of cleavage (phenylthiyl radical and
phenylmethoxide anion) is less favoured as the phenylthiyl
radical is estimated to have a more negative reduction poten-
tial than the phenylmethoxyl radical (18). Regardless of the
mode of cleavage, the reduction potential of the radical
formed is far positive of the reduction potential of 1 and un-
dergoes a rapid ET to give the phenylmethoxide anion (4).
The reduction waves following the dissociative wave of 1
are due to products from the reactions involving the anionic
species 3 or 4 reacting with 1 or the medium. The addition
of an acid, such as acetanilide or 2,2,2-trifluoroethanol
(TFE), will result in the protonation of 4, eliminating any re-
actions of this anion with 1. The pKas of benzenethiol, acet-
anilide, and TFE as determined in DMSO are 10.3, 21.45,
and 23.45, respectively (19). Given the relative pKa values
for these compounds, it is apparent why acetanilide and TFE
have a limited capacity to protonate 3. Therefore, the addi-
tion of these acids does not hinder the ability of 3 to react
with 1, even when present in large excess. In this case, to
trap any 3 produced in the reduction of 1 a nucleophile trap
such as methyl p-toluenesulfonate (MeOTs) or methyl iodide
(MeI) is added.
The addition of both an acid and nucleophile trap is nec-
essary to observe the full peak current. For example, the ad-
dition of 2 equiv. of TFE results in an increase in the peak
current from 107 to 148 µA. The addition of 1 equiv. of
MeOTs caused a further increase from 148 to 164 µA. It was
also observed that addition of MeOTs alone was not capable
of stopping the parent–child reactions, as an increase in the
peak current of the dissociative wave is observed with the
addition of acetanilide to a cell containing 1 and a large ex-
cess of MeOTs. It would seem that both the anions gener-
ated from reduction of 1 are participating in some form of
parent–child reaction, therefore it is necessary to add both an
acid and a nucleophile trapping agent to observe the maxi-
mum peak current.
The voltammetry responds differently to the addition of
each reagent. Upon initial addition of acid the peak sharp-
ens, especially apparent at higher scan rates, with α values
ranging from 0.44 at 0.1 V/s to 0.53 at 1.0 V/s and then
down to 0.37 at 10.0 V/s. The Ep continues to shift nega-
tively as in the absence of acid by 74 mV/decade. The
dissociative wave sharpens even further with the addition of
MeOTs, most noticeably at the foot of the wave. This is car-
ried over to the values of α, ranging from 0.50 to 0.52 to
0.37 at 0.1, 1.0, and 10.0 V/s, respectively. The addition of
both reagents to the cell eliminates the shoulder on the tail
rates. A shoulder is present on the tail of this wave and fol-
lowing this is another small wave that appears to be quasi-
reversible. Both of the latter two features are present due to
the reduction of products formed from parent–child reac-
tions, where 1 (the parent) reacts with a basic or
nucleophilic anionic species (the child) generated from the
ET reduction of 1. Reversing to scan anodically, a large oxi-
dation near 0 V vs. SCE is observed and is attributed pri-
marily to the oxidation of phenylthiolate (3) to the
phenylthiyl radical (18). Because of the similarity in oxida-
tion potentials, this wave may also include the oxidation of
phenylmethoxide anion (4), and at slower scan rates this
wave does appear to be composed of two overlapping oxida-
tion peaks. The presence of 3 is confirmed by reversing the
scan direction to scan cathodically following this large oxi-
dation peak. A reduction wave attributable to diphenyl
disulfide is observed, though this wave is exceedingly broad
because of fouling of the working electrode by the putative
phenylthiyl radicals. The analysis of the CVs indicates that
the ET reduction occurs by the mechanism shown in
Scheme 2. The initial ET occurs to form a radical anion in-
termediate, which then cleaves at the S—O bond in a sepa-
rate step. The cleavage of the radical anion generates
phenylthiolate (3) and the phenylmethoxyl radical. The alter-
© 2005 NRC Canada

1476
Can. J. Chem. Vol. 83, 2005
Scheme 3. Products obtained from the CPE of 1.
Scheme 4. Parent–child reactions involved in the ET reduction
of 1.
OH
S
S
S
S
Ph
Ph
O
Ph
Ph
O
Ph
PhS
PhSSPh
3
1
7
4
8
5
6
7
S
S
Ph
O
PhSO
PhS
Ph
3
1
12
OH
Ph
O
H
O
O
O
S
S
S
S
H
Ph
Ph
O
Ph
Ph
PhS
4
1
3
13
6
8
9
10
phenyl disulfide (7), could be identified using GC by com-
parison with authentic samples. Benzyl phenyl sulfide (8), a
minor product, could also be identified by GC comparison.
The identities of these products were further confirmed by
mass spectrometry. The identities of the other minor prod-
ucts of reduction, benzylbenzenesulfinate (9), n-butyl phenyl
sulfide (10), and benzyl p-toluenesulfonate (11), are based
on their mass spectra. There were some minor products pres-
ent that could not be conclusively identified by their mass
spectra. Compounds 6–9 are products derived from parent–
child processes. The remaining products are attributed to
secondary reactions involving products of the reduction (3
and 4) with the solvent, electrolyte, or MeOTs. Because of
the significance of these secondary reactions, a similar elec-
trolysis was carried out, only switching to acetonitrile
(MeCN) as the solvent, which was then worked up as previ-
ously described. GC–MS analysis revealed that switching
the electrolysis solvent did not simplify the product mixture,
and that products produced from reaction with MeCN were
also present.
Based on the products observed in the DMF electrolysis,
the reactions shown in Scheme 4 have been postulated for
the parent–child processes. The first reaction is a nucleo-
philic attack at the sulfur atom of 1 by 3 to give diphenyl
disulfide (7) and 4. The reaction is believed to be occurring,
but is in essence inconsequential as 4 is produced by direct
ET to 1. Also, at the potential at which the electrolysis is
performed, any 7 formed by this reaction will be reduced to
give 2 equiv. of 3. This reaction will not affect either the
products observed or the charge required for complete elec-
trolysis. The second reaction involves attack at the benzyl
carbon of 1 by 3 to give the benzenesulfenate anion and
benzyl phenyl sulfide (8). This reaction accounts for the ob-
servance of 8 as a product from electrolysis and accounts for
the lower charge consumption, as the reduction potential of
8 is negative of the electrode potential during electrolysis.
The final parent–child reaction is the deprotonation of the
benzyl carbon of 1 to give benzyl alcohol (6), benzaldehyde
(13), and 3. Benzaldehyde was observed as a minor product
in the CPE experiments carried out in MeCN. The reduction
O
S O
O
11
of the dissociative wave and no reduction waves are ob-
served following it prior to the reduction of MeOTs occur-
ring around –2.1 V vs. SCE. This is evidence that it has
stopped any parent–child reactions and any products from
the reduction that have been trapped are not electroactive in
the potential window investigated. The large oxidation near
0 V vs. SCE also disappears, though it remains present in
the voltammetry of 1 until MeOTs is added, again confirm-
ing its assignment as the oxidation of 3.
Analysis of the data obtained from voltammograms per-
formed in the presence of both acid and a nucleophile trap-
ping agent allows an important conclusion concerning the
mechanism of reduction to be drawn. The values for the
peak width, peak current, and the shift in the peak potential
are all similar to an electrochemically reversible process
(i.e., the reduction of nitrobenzene) suggesting that the bond
cleavage, and not the ET, is the rate-determining step. This
implies that the ET reduction is occurring by a stepwise
dissociative mechanism.
CPE experiments carried out in the absence of both acid
and MeOTs required only 1.1 electron equiv., this value be-
ing the average obtained over several experiments. CVs
taken before and after complete electrolysis are shown in
Fig. 1a. Following complete reduction, the dissociative wave
has completely disappeared while a small irreversible wave
at more negative potentials is present. Scanning anodically,
it was found that the peak attributed to the oxidation of 3
significantly increased. The cell solution was acidified and
MeOTs was added following the electrolysis to aid in prod-
uct isolation. The majority of the DMF and TFE were re-
moved in vacuo with mild heating (60 °C). The residue was
then subjected to column chromatography. Analysis of the
fractions obtained by GC and GC–MS indicated the products
summarized in Scheme 3. The major products of the reduc-
tion, methyl phenyl sulfide (5), benzyl alcohol (6), and di-
°
potential of 13 is very close to that of 1 with E 13 = –1.85 V
vs. SCE, and it is suggested that the shoulder on the
dissociative wave is due to the reduction of 13 (Fig 1a). For
© 2005 NRC Canada

Stringle and Workentin
1477
Fig. 2. Representative CVs of 2 in 0.1 mol/L TBAP–DMF at
Scheme 5. Reactions accounting for the formation of 9.
0.1 V/s: (a) 2 (2.4 mmol/L) before (solid line) and after (dashed
line) introduction of 6.0 C of charge into the cell; (b) 2
(2.3 mmol/L) in the presence of acetanilide (40 mmol/L) before
(solid line) and after (dashed line) introduction of 11.2 C of
charge into the cell and following the subsequent addition of
2.7 mmol/L MeOTs (dotted line).
O
O
S
S
S
Ph
Ph
Ph
Ph
O
O
S
Ph
O
Ph
Ph
Ph
S
Ph
12
1
14
4
O
O
S
S
(a)
Ph
S
Ph
O
Ph
PhS
4
14
9
3
O
S
2e−
Ph
PhSO PhS
12
14
3
these reasons it is suspected that 13 will be reduced to its
radical anion during the electrolysis of 1. Reduction of 13 in
aprotic solvent leads to the formation of 1,2-diphenylethane-
1,2-diol (20–25); however, this compound was not observed
as a product of our experiments. As the presence of acid is
required to observe the full peak current of the dissociative
wave of 1, there must be a parent–child reaction involving 4
and 1. Also, reaction between 4 and 1 to give 13 will result
in a net decrease in the charge required for complete elec-
trolysis, even if 13 is reduced to its radical anion. The prod-
ucts that were not conclusively identified were consistent
with products from reactions between 3 and 13 and could
account for the lack of products serving as evidence of the
presence of 13. On these grounds, we propose the reaction
shown in Scheme 4 as the most likely reaction between 1
and 4.
20 µA
(b)
reduction of
O
S
OMe
O
oxidation of
N
The formation of benzyl benzenesulfinate (9) cannot be
directly accounted for through a parent–child reaction. How-
ever, in the reaction leading to 8 the benzenesulfenate anion
(12) is generated. A possible reaction of this anion with 1 is
illustrated in Scheme 5. Attack at the sulfur atom of 1 by 12
generates phenyl benzenethiosulfinate (14), while releasing
4. A subsequent reaction between 4 and 14 could lead to the
formation of 9. Compound 14 could also be generated from
attack of 12 on 7 with the release of 3. The other primary
process available to 14 would be ET reduction, which would
lead to the generation of 12 and 3.
When the reduction is performed in the presence of both
MeOTs and acetanilide, a smooth conversion to phenyl
methyl sulfide and benzyl alcohol by two-electron reduction
is observed. It can be seen in Fig. 1b that no electroactive re-
agents remain following complete reduction of 1. Both
HPLC and GC results confirm quantitative formation of
phenyl methyl sulfide (5). GC was used to determine the
concentration of benzyl alcohol produced under these condi-
tions, though it was difficult to achieve consistent results,
even for the standards used for calibration. GC calibration
indicates that 76% of the maximum amount of benzyl alco-
hol is recovered when the reduction is carried out with acid
and MeOTs. These results were consistent with quantitative
formation of benzyl alcohol given the error associated with
these measurements. Considering that 5 is formed quantita-
tively and the two-electron stoichiometry for the complete
reduction, it is assumed that a less than quantitative return of
benzyl alcohol is primarily due to loss during the work-up.
O
1
0
-1
-2
-3
Analysis of the CV and CPE experimental data leads to
the conclusion that the ET reduction of 1 occurs via a step-
wise dissociative mechanism and the intermediate radical
anion cleaves at the S—O bond.
ET reduction of tert-butyl benzenesulfenate ester (2)
The CV experiments of 2 were performed in 0.1 mol/L
TBAP in DMF and the relevant voltammetric data are listed
in Table 1 for comparison to the CV data obtained for com-
pound 1. A representative CV of 2 in 0.1 mol/L TBAP–
DMF solution is shown in Fig. 2. A narrow, irreversible
wave is observed at –2.28 V vs. SCE at 0.1 V/s. A negative
shifting of the Ep with increasing scan rate is observed at a
rate of 103 mV/decade. The peak width increases with in-
creasing scan rate leading to decreasing values of the trans-
fer coefficient. The α values at 0.1, 1.0, and 10.0 V/s are
0.58, 0.50, and 0.37, respectively. These data are in accord
with the heterogeneous ET occurring by a stepwise
dissociative mechanism, although the shifting of the Ep is
higher than generally observed for systems proceeding by a
stepwise mechanism. The dissociative wave is the only peak
present on the cathodic scan, and reversing to the anodic
scan reveals only small oxidation waves, starting at –0.4 V
© 2005 NRC Canada

1478
Can. J. Chem. Vol. 83, 2005
Scheme 6. Mechanism for the ET reduction of 2.
Scheme 7. Parent–child reactions involved in the ET reduction
of 2.
e−
S
S
H
S
Ph
O
Ph
O
2
B
Ph
O
O
PhSO
BH
2
12
S
Ph
O
PhSO
12
O
S
O
S
S
Ph
Ph
Ph
O
Ph
S
e−
12
2
14
15
demonstrates that the parent–child process is highly effec-
tive when reduction is carried out on the electrolysis scale.
Figure 2a shows CVs of 2 taken before and after the elec-
trolysis. It can be seen that following electrolysis there are
no products that are electroactive at negative potentials.
However, anodic scans reveal that a large oxidation peak is
present at 0.45 V vs. SCE at 0.2 V/s. Small peaks occurring
at slightly more negative potentials are also present. Elec-
trolysis performed with a large excess of acetanilide present
require 1.96 F/mol (i.e., 2 electron equiv.) for complete re-
duction indicating that only acid is necessary to effectively
suppress any parent-child reactions. Figure 2b shows the be-
fore and after CVs from the CPE of 2 with acid and it is
seen that again no reducible products are observed and the
large oxidation at 0.45 V vs. SCE seen in the voltammetry
following CPE without acid is present. There is a large peak
just negative of this oxidation, which is due to the oxidation
of the acetanilide anion. Figure 2b shows that the addition of
MeOTs following electrolysis will remove the oxidation
peak of the acetanilide anion along with the smaller oxida-
tion waves also present. The large wave at 0.45 V remains
unaffected by the addition of MeOTs. Washing of the cell
contents with an organic solvent and water proved inade-
quate in identifying the products of reduction as essentially
no products could be detected in the organic residue by GC
vs. SCE through to the largest oxidation wave occurring at
0.45 V vs. SCE at 0.2 V/s. The origin of these oxidations
will be discussed in the context of the electrolysis experi-
ments. It is sufficient at this point to note that unlike the
voltammograms for 1, only a small oxidation is present near
0 V vs. SCE. This is an indication that 3 is being produced
only in a limited quantity, and the addition of nucleophile
traps is deemed unnecessary when studying the electrochem-
istry of 2. In comparison, for the reduction of 1 a large oxi-
dation wave due to 3 is observed indicating a significant
amount of 3 is produced, leading to the conclusion that the
radical anion dissociates by S—O cleavage. Limited produc-
tion of 3 is an indication that S—O bond cleavage is not oc-
curring appreciably and that instead the radical anion of 2
dissociates by breaking the C—O bond as illustrated in
Scheme 6. In this scheme, compound 2 is initially reduced
by an ET to its radical anion, which then fragments in a sep-
arate step. The C—O bond is cleaved in the fragmentation
step resulting in the formation of benzenesulfenate anion
(12) and the tert-butyl radical. The standard reduction poten-
tial of this radical is estimated to be –1.54 V vs. SCE (18),
which is positive of the Ep of 2, therefore a heterogeneous
ET can reduce this radical to give the tert-butyl anion (15).
The addition of a weak, nonnucleophilic acid, such as ac-
etanilide or TFE, causes a significant increase in the peak
current. The peak current values, corrected for both concen-
tration and scan rate are 147 µA without acid and 170 µA
with acid for a 2 mmol/L solution. Again, some parent–child
interaction is responsible for a lower observed peak current
in the absence of a reagent to trap the reactive species. The
peak not only increased in current, but became narrower,
which was especially apparent at higher scan rates. This is
demonstrated in the value of the transfer coefficient, ranging
from 0.55 to 0.50 and 0.43 at scan rates of 0.1, 1.0, and
10.0 V/s, respectively. Also, the observed shift in the Ep with
increasing scan rate decreased from 103 to 72 mV/decade, a
value that is more consistent with ET by a stepwise
dissociative mechanism. The addition of acid did not dra-
matically affect the oxidation waves observed in the absence
of acid, as it appears that the products responsible for these
oxidation peaks are produced in the same quantity in the
presence and absence of acid.
1
or H NMR spectroscopy. In place of typical product analy-
sis, it was decided to further analyze the large oxidation
wave present in the CVs of the products of electrolysis.
Given that little to no 3 was observed in the CVs of 2,
C—O bond cleavage is believed to be occurring as shown in
Scheme 6. Based on the anions formed from this reduction,
the reactions illustrated in Scheme 7 are the proposed par-
ent–child reactions believed to be occurring. The first reac-
tion involves removal of a proton from the terminal methyl
groups of 2. The base involved in this reaction could be the
tert-butyl anion or more likely the basic species generated
from proton abstraction by the tert-butyl anion. Deprotona-
tion of 2 generates isobutylene and benzenesulfenate anion
(12). Therefore, the same product is observed in the volt-
ammetry whether formed from direct reduction or from de-
protonation of 2. This accounts for the observance of the
same oxidation wave following the electrolysis of 2 with or
without acid.
The second possible parent–child reaction involves
nucleophilic attack of 2 by 12. This process is identical to
that proposed in the formation of 9 from the ET reduction of
1, as previously shown. Again, phenyl thiobenzenesulfinate
(14) is produced only now with the tert-butoxide anion as
To confirm that C—O bond cleavage was occurring in the
ET reduction of 2, CPE experiments were necessary. Per-
forming CPE in the absence of acid required only 1.1 elec-
tron equiv. for the complete reduction of 2. This
© 2005 NRC Canada

Stringle and Workentin
1479
Fig. 3. Depiction of the SOMOs of the radical anions of 1 (left) and 2 (right) as determined by DFT calculations.
Table 2. Bond lengths (Å) calculated for the equilibrium con-
formers of 1 and 2 and their respective radical anions.
Scheme 8.
O
S
1·– (Å)
2 (Å)
2·– (Å)
S
Bond
1 (Å)
[O]
Ph
O
Ph
O
C—S
S—O
O—C
1.764
1.702
1.423
1.734
1.776
1.384
1.765
1.702
1.441
1.787
1.668
1.439
12
16
the leaving group. The presence of tert-butoxide is believed
to account for the small oxidation wave at –0.4 V vs. SCE in
the CVs of 2. Any 14 produced will likely be reduced to re-
turn to 12 and 3 (Scheme 5) and this reaction accounts for
the oxidation wave at 0 V vs. SCE attributed to the oxida-
tion of 3.
dation peaks of benzenesulfenate and benzenesulfinate sim-
ply coincide. Since benzenesulfenate is a highly unstable
species, it was not possible to obtain an authentic sample to
check this assertion.
However, the primary product from CPE of 2 in the pres-
ence and absence of acid is responsible for the oxidation
peak at 0.45 V vs. SCE. We attribute this peak to the oxida-
tion of benzenesulfinate (16), which is formed from the oxi-
dation of 12 (Scheme 8). Evidence in favour of this assignment
comes from the addition of sodium benzenesulfinate to the
cell solution following CPE of 2. The peak at 0.45 V in-
creased in peak current and no new peaks were observed,
while the Ep remained relatively constant indicating that the
peak was likely due to the oxidation of benzenesulfinate. In-
dependent CV experiments were performed that confirmed
that the Ep of the sodium benzenesulfinate coincided with
that of the oxidation wave observed in the electroysis of 2.
By varying the concentration of the sodium benzensulfinate
the current response was calibrated and the concentration of
16 formed in the electrolysis was determined. As much as
75% of 2 was converted to 16 in the electrolysis performed
without acid, and 16 was present in 96% of the starting con-
centration of 2 for CPE experiments performed with acid.
How the conversion of 12 to 16 occurs is not completely un-
derstood. Previous studies of the electrolysis of phenyl benz-
enethiosulfinate also reported a similar oxidation wave that
was attributed to 16 (26). In this study it was hypothesized
that a disproportionation reaction between two molecules of
12 was occurring to give 3 and 16. While it is possible that
this process occurs, the relative peak currents of the oxida-
tion peaks due to 3 and 16 indicate that it is not the primary
process leading to the generation of 16. It is difficult to
assign a particular pathway for the oxidation of benzenesulf-
enate. Oxidation by perchlorate is unfavourable and oxida-
tion by the solvent requires a significant amount of residual
water to be present. It should be recognized that the Ep of an
irreversible wave is not the most reliable means of identify-
ing an unknown present in solution and that the wave could
be due to the oxidation of benzenesulfenate and that the oxi-
DFT calculations of the SOMOs
It is interesting to compare the differences in the ET reac-
tions of 1 and 2 that result as a difference in the O group.
The most intriguing result of changing the ester moiety is
the resulting change in the mode of fragmentation of the rad-
ical anions. The factors that result in the difference in cleav-
age reactions are not obvious. The same anionic fragment,
either phenyl thiolate (11) or benzenesulfinate (22), would
result from S—O and C—O cleavage of each compound, so
it is not related to a thermodynamic advantage imparted by
the stability of the anionic product. In both cases the carbon
radical formed from C—O cleavage would be relatively sta-
ble; estimates of the reduction potentials of the benzyl radi-
cal and the tert-butyl radical have them near –1.34 and
–1.54 V vs. SCE (18), respectively.
The difference seems to be related to differences in the
electronic configuration of the radical anion intermediate for
each of 1 and 2. The SOMO for each of these compounds
was determined by performing PM3 calculations to deter-
mine the equilibrium conformer of the radical anion and
then density functional theory calculations at the BLYP-3//6-
31** level were performed to arrive at the SOMO for each
compound. The results of these computations are depicted in
Fig. 3. The SOMO of the radical anion of 1 was largely as-
sociated with the S—O bond while the SOMO of the radical
anion of 2 was largely delocalized over the phenyl ring and
S atom. The respective bond lengths of the starting com-
pound and its radical anion determined by PM3 calculations
are given in Table 2. The ET to 1 causes significant length-
ening of the S—O bond, while the S—O bond of 2 shortens
upon ET. For 1 the majority of the additional charge is asso-
ciated with the S—O bond and S—O cleavage is the result.
For 2 the charge is delocalized onto the phenyl ring with
only a limited effect on the S—O and C—O bonds, and
© 2005 NRC Canada

1480
Can. J. Chem. Vol. 83, 2005
cleavage follows the most thermodynamically favourable
pathway. Gregory and Jenks (17) have performed calcula-
tions for the homolytic cleavage of disulfides, peroxides,
and sulfenate esters. While it was found that S—S and O—
O cleavage were the most energetically viable pathways for
disulfides and peroxides, the results indicated that fissure at
the C—O bond is the most thermodynamically favourable
mode of cleavage for a sulfenate ester. Our experimental re-
sults indicate that such an advantage is present in the radical
anion of 2 as well.
Regioselective bond cleavages were observed by Houmam
et al. (27) in their study of the ET reduction of benzyl
thiocyanates, where differences in the mode of cleavage were
observed for benzylthiocyanate and 4-nitrobenzylthiocyanate.
The differences in this case were related to the radical anion
intermediate, with benzylthiocyanate forming a σ* type radi-
cal anion akin to compound 1 and nitrobenzylthiocyanate
forming a π* radical anion intermediate as was observed for
2. In this case, the difference in the electron distribution of
the radical anions of 1 and 2 must be related to the nature of
the ester functionality as opposed to the aromatic moiety.
The electron-donating capacity of the tert-butyl group in-
creases the electron density on the O atom shifting the
SOMO onto the aromatic ring. The benzyl group has the op-
posite effect, decreasing the electron density at the O atom
and drawing the SOMO onto the S—O bond.
technologies GC 6890N gas chromatograph. HPLC was car-
ried out using a Waters 600 controller and Waters Delta 600
pump, a Waters 2487 dual absorbance detector all inter-
λ
faced to a PC using Millenium software, and a Hamilton
PRP-1 column.
Calculations were performed on
a PC using the
SPARTAN program, which performed the PM3 and DFT
calculations. The method for the calculations of the SOMOs
of the sulfenate ester radical anions involved a geometry op-
timization of the structure using semiempirical PM3 calcula-
tions followed by a single-point energy DFT calculation
using the B3LYP//6-31G** basis set.
Electrochemical apparatus and procedures
Cyclic voltammetry was performed using either a PAR
283 or PAR 263 model EG & G potentiostat interfaced to a
personal computer using the M270 DOS-based software. A
standard three electrode arrangement, with working, counter,
and reference electrodes, was employed. The cell was main-
tained at a constant temperature of 25 °C with continuous
purging using high purity Ar gas throughout the experiment.
The working electrode was a 3 mm diameter glassy carbon
rod (Tokai) sealed in a glass tube with epoxy resin. The
working electrode was cleaned prior to use by polishing with
0.25 µm diamond paste, rinsing with 2-propanol, and placed
in a sonicator in 2-propanol for 10 min. Electrochemical ac-
tivation of the electrode was performed by cycling several
times between 0 and –2.7 V at 0.2 V/s in the background so-
lution. The counter electrode was a platinum plate, which
was symmetrically placed under the working electrode. The
reference electrode consisted of a silver wire immersed in a
0.1 mol/L solution of the electrolyte in the appropriate sol-
vent contained in a narrow glass tube with a fine sintered
glass tip. The reference potential was calibrated following
each experiment using the ferrocene/ferrocenium couple
whose standard potential under our conditions was 0.464 V
vs. SCE. All reported potentials are referenced to SCE.
Conclusions
The ET reduction of benzyl benzenesulfenate ester (1) and
tert-butyl benzenesulfenate ester (2) has been investigated.
In each case, ET reduction generates a radical anion inter-
mediate that subsequently undergoes bond cleavage. The
bond that cleaves upon reduction, however, differs for the
two compounds. ET reduction of 1 leads to S—O bond
cleavage to give phenylthiolate and phenylmethoxide in an
overall two-electron process. Complicating secondary reac-
tions are effectively suppressed by addition of appropriate
trapping reagents. The ET reduction of 2 leads to C—O
cleavage of the radical anion. In the presence of acid the re-
duction proceeds to quantitatively generate benzenesulfen-
ate. The difference in the mode of cleavage for these two
radical anions is most likely related to the electron distribu-
tion in the radical anion. The additional charge on the radi-
cal anion of 1 is largely associated with the S—O bond,
leading to significant S—O bond stretching upon ET and ul-
timately, bond fragmentation. The additional electron on the
radical anion of 2 is largely associated with the phenyl ring
and S atom so the dissociation of the radical anion follows
the most thermodynamically favourable pathway, which in
this case is C—O cleavage.
The CPE experiments were carried out using the same cell
set up protocol and a similar cell arrangement. In controlled
potential electrolysis, there is the inclusion of a second
working electrode, which is a 2.5 cm (dia.) × 0.8 cm (width)
glassy carbon disk, custom fit into a Teflon jacket that atta-
ches to a Radiometer ED-101 rotating disk assembly. A radi-
ometer CTV101 speed control unit was used to regulate the
velocity of the rotating disk electrode (RDE), which was
maintained at 4500 rpm for these experiments.
Materials
Dry solvents were obtained from an Innovative Technol-
ogy SPS-400-5 solvent purification system and were stored a
maximum of 5 days under an N2 atmosphere. Tetrabutyl-
ammonium perchlorate (Aldrich, Fluka) was purified by
multiple recrystallizations from 95% EtOH and the crystals
subsequently vacuum-dried at 60 °C overnight. The
recrystallized and dried electrolyte was then stored in a de-
humidifying cabinet. Benzyl alcohol, benzaldehyde, di-
phenyl disulfide, thioanisole, benzenethiol, sulfuryl chloride,
and m-chloroperbenzoic acid were purchased from Aldrich
and purified by standard procedures when necessary.
Experimental
General
The NMR spectra were obtained using a Varian Mercury
400 MHz spectrometer using the solvent peak (CDCl3,
1
7.28 ppm) as the reference. H and 13C NMR spectra were
recorded at 400.1 and 100.6 MHz, respectively. Mass spectra
were performed using a MAT 8200 Finigan high-resolution
mass spectrometer. GC was performed using an Agilent
© 2005 NRC Canada

Stringle and Workentin
1481
Benzyl benzenesulfenate ester (1)
Methyl phenyl sulfide (5)
The presence of 5 in the product mixtures could be in-
ferred from the observation of a peak in the GC chroma-
togram of the product mixture with the same retention time
as an authentic sample of 5 purchased from Aldrich. GC–
MS of this peak had the following major peaks: 124 (100%)
M+, 109 (47%) M+ – CH3, 91 (57%) M+ – SH, 78 (65%)
M+ – SCH2.
Compound 1 was prepared by the reaction of lithium
phenylmethoxide and benzenesulfenyl chloride as previously
reported (28). The crude product was a deep yellow oil,
which was purified by vacuum distillation at 100 °C and
>0.5 mmHg (1 mmHg = 133.322 4 Pa) using a Kugelrohr
1
apparatus to give a pale yellow oil. H NMR (CDCl3) δ:
7.40–7.43 (m, 2H), 7.26–7.32 (m, 4H), 7.17–7.22 (m, 2H),
7.09–7.15 (m, 2H), 4.70 (s, 2H). 13C NMR (CDCl3) δ:
140.13, 136.92, 128.96, 128.39, 127.32, 127.02, 126.69,
124.32, 79.50.
Benzyl alcohol (6)
The presence of 6 in the product mixtures was determined
from the observation of a peak in the GC chromatograms of
the products with the same retention time as an authentic
sample of 6 purchased from Aldrich. The corresponding
peak in the GC–MS had the following mass spectrum char-
acteristics: 108 (100%) M+, 91 (25%) M+ – OH, 79 (75%)
M+ – CHO.
tert-Butyl benzenesulfenate ester (2)
Compound 2 was prepared by reaction of tert-butyl alco-
hol and benzenesulfenyl chloride in the presence of pyridine
as previously reported (29). Purification was performed by
column chromatography using hexanes – ethyl acetate (4:1)
as the eluent to give a clear, colourless oil. 1H NMR
(CDCl3) δ: 7.66–7.70 (m, 2H), 7.49–7.53 (m, 3H), 1.56 (s,
9H). 13C NMR (CDCl3) δ: 146.46, 131.54, 128.90, 124.86,
82.87, 29.87.
Diphenyl disulfide (7)
The presence of 7 in the product mixtures was determined
from the observation of a peak in the GC chromatograms of
the products with the same retention time as an authentic
sample of 7 purchased from Aldrich. The corresponding
peak in the GC–MS had the following mass spectrum char-
acteristics: 218 (100%) M+, 185 (28%) M+ – HS, 154 (30%)
M+ – H2S2, 140 (10%) M+ – C6H6, 109 (99%) M+ – SC6H6,
65 (42%), 109 – CS.
Benzyl phenyl sulfide (8)
Compound 8 was prepared by the addition of a solution of
benzylbromide (5.9 mmol) in ~25 mL CH2Cl2 to a flask
containing benzenethiol (5.8 mmol) and pyridine (6.2 mmol)
dissolved in 25 mL of CH2Cl2. The reaction mixture was
washed with aq. K2CO3 solution, which was then washed
with CH2Cl2. The organic extracts were combined and
washed three times with H2O and dried over anhydr.
Na2SO4. The crude product was purified by column chroma-
tography with hexanes – ethyl acetate (10:1) as the eluent.
1H NMR (CDCl3) δ: 7.21–7.37 (m, 10H), 4.17 (s, 2H).
13C NMR (CDCl3) δ: 137.33, 136.28, 129.64, 128.73,
128.39, 127.08, 127.02, 126.21, 38.88.
Benzyl phenyl sulfide (8)
The presence of 8 in the product mixtures was determined
from the observation of a peak in the GC chromatograms of
the products with the same retention time as an authentic
sample of 8 prepared as previously described. The corre-
sponding peak in the GC–MS had the following mass spec-
trum characteristics: 200 (83%) M+, 109 (19%) M+ – C7H7,
91 (100%) M+ – C6H5S, 65 (45%), 109 – CS.
Benzyl benzenesulfinate (9)
Phenyl benzenethiosulfinate (14)
The presence of 9 in the product mixtures was determined
from the analysis of the GC–MS of the product mixture. The
peak assigned to this product had the following mass spec-
trum characteristics: 232 (1%) M+, 167 (6%) M+ – C5H5,
153 (2%) 167 – CH2, 125 (6%) M+ – C7H7O, 91 (100%)
M+ – C6H5O2S, 77 (23%) 91 – CH2, 65 (9%) M+ – C8H8O2S.
Compound 14 was prepared by the oxidation of diphenyl
disulfide with m-chloroperbenzoic acid according to ac-
cepted literature methods (30–32). Column chromatography
was employed for the purification using hexanes – ethyl ace-
1
tate (10:1) as the eluent. H NMR (CDCl3) δ: 7.64–7.68 (m,
2H), 7.47–7.56 (m, 5H), 7.41–7.46 (m, 1H), 7.34–7.40 (m,
2H). 13C NMR (CDCl3) δ: 143.96, 135.33, 131.50, 130.26,
129.24, 128.89, 127.52, 124.29.
n-Butyl phenyl sulfide (10)
The presence of 10 in the product mixtures was deter-
mined from the analysis of the GC–MS of the product mix-
ture. The peak assigned to this product had the following
mass spectrum characteristics: 166 (77%) M+, 152 (4%)
M+ – CH2, 123 (100%) M+ – C3H7, 109 (49%) M+ – C4H9,
91 (5%) M+ – C3H7S, 77 (11%) M+ – SC4H9, 65 (12%),
109 – CS.
Identification of products from the CPE experiments of 1
Product mixtures and authentic samples were subjected to
analysis by GC using a 15 m HP-5 column. The method
used to obtain the gas chromatograms is as follows: the ini-
tial temperature is 80 °C and upon initiation the temperature
is increased to 280 °C at a rate of 15°/min and held at
280 °C for 5 min. Following partial separation of the prod-
ucts into two major fractions by column chromatography,
the fractions of the product mixture were analyzed by GC–
MS. The method for the GC of these analyses is as follows:
the temperature is initially set at 60 °C and held for 3 min.
The temperature is then ramped to 300 °C by 15°/min.
Benzyl p-toluenesulfonate (11)
The presence of 11 in the product mixtures was deter-
mined from the analysis of the GC–MS of the product mix-
ture. The peak assigned to this product had the following
mass spectrum characteristics: 262 (12%) M+, 244 (3%)
M+ – H2O, 186 (3%) M+ – C6H5, 107 (100%) M+ – C7H7O2S,
91 (12%) M+ – C7H7O3S, 79 (52%), 107 – CO.
© 2005 NRC Canada

1482
Can. J. Chem. Vol. 83, 2005
13. S. Antonello, R. Benassi, G. Gavioli, F. Taddei, and F. Maran.
J. Am. Chem. Soc. 124, 7529 (2002).
14. S. Antonello, K. Daasbjerg, H. Jensen, F. Taddei, and F.
Maran. J. Am. Chem. Soc. 125, 14905 (2003).
15. W.H. Horspool. The chemistry of sulphenic acids and their de-
rivatives. Edited by S. Patai. John Wiley and Sons, Chichester.
1990.
16. (a) D.J. Pasto, F. Cottard, and C. Picconatto. J. Org. Chem. 59,
7172 (1994); (b) D.J. Pasto and G.L. L’Hermine. J. Org.
Chem. 55, 5815 (1990); (c) D.J. Pasto and G.L. L’Hermine.
Tetrahedron, 49, 3259 (1993); (d) D.J. Pasto and F. Cottard. J.
Org. Chem. 59, 4642 (1994).
17. D.D. Gregory and W.S. Jenks. J. Org. Chem. 63, 3859 (1998).
18. K. Daasbjerg, S.U. Pedersen, and H. Lund. In General aspects
of the chemistry of radicals. Edited by Z.B. Alfassi. John
Wiley and Sons, Chichester. 1999. p. 385.
19. F.G. Bordwell. Acc. Chem. Res. 21, 456 (1988).
20. S.G. Mairanovskii. Bull. Acad. Sci. USSR, 2003 (1961).
21. E. Laviron. Coll. Czech. Chem. Commun. 30, 4219 (1965).
22. L. Nadjo and J.-M. Savéant. J. Electroanal. Chem. 33, 419
(1971).
23. J.W. Hayes, I. Ruzic, D.E. Smith, G.L. Booman, and J.R.
Delmastro. J. Electroanal. Chem. 51, 269 (1974).
24. V.D. Parker and O. Lerflaten. Acta. Chem. Scand. Ser. B, B37,
403 (1983).
Benzaldehyde (13)
The presence of 13 in the product mixtures from CPE ex-
periments performed in MeCN was determined from the ob-
servation of a peak in the GC chromatograms of the
products with the same retention time as an authentic sample
of 13 purchased from Aldrich. The corresponding peak in
the GC–MS had the following mass spectrum characteris-
tics: 108 (100%) M+, 91 (26%) M+ – OH, 79 (73%) M+ –
CHO.
References
1. R.L. Donkers, J. Tse, and M.S. Workentin. Chem. Commun.
135 (1999).
2. R.L. Donkers and M.S. Workentin. J. Am. Chem. Soc. 126,
1688 (2004).
3. D.L.B. Stringle, R.N. Campbell, and M.S. Workentin. Chem.
Commun. 1246 (2003).
4. M.S. Workentin, F. Maran, and D.D.M. Wayner. J. Am. Chem.
Soc. 117, 2120 (1995).
5. S. Antonello, M. Musumeci, D.D.M. Wayner, and F. Maran. J.
Am. Chem. Soc. 119, 9541 (1997).
6. R.L. Donkers, F. Maran, D.D.M. Wayner, and M.S. Workentin.
J. Am. Chem. Soc. 121, 7239 (1999).
7. D.C. Magri and M.S. Workentin. Org. Biomol. Chem. 1, 3418
(2003).
8. M.S. Workentin and R.L. Donkers. J. Am. Chem. Soc. 120,
2664 (1998).
9. R.L. Donkers and M.S. Workentin. J. Phys. Chem. B, 102,
4061 (1998).
10. R.L. Donkers and M.S. Workentin. Chem. Eur. J. 7, 4012
(2001).
11. T.B. Christensen and K. Daasbjerg. Acta Chem. Scand. 51,
307 (1997).
12. K. Daasbjerg, H. Jensen, R. Benassi, F. Taddei, S. Antonello,
A. Gennaro, and F. Maran. J. Am. Chem. Soc. 121, 1750
(1999).
25. C.P. Andrieux, M. Grzeszczuk, and J.-M. Savéant. J. Am.
Chem. Soc. 113, 8811 (1991).
26. B. Persson. J. Electroanal. Chem. 86, 313 (1978).
27. A. Houmam, E.M. Hamed, P. Hapiot, J.M. Motto, and A.L.
Schwan. J. Am. Chem. Soc. 125, 12676 (2003).
28. E.G. Miller, D.R. Rayner, H.T. Thomas, and K. Mislow. J.
Am. Chem. Soc. 90, 4861 (1968).
29. L. Goodman and N. Kharasch. J. Am. Chem. Soc. 77, 6541
(1955).
30. L.D. Small, J.H. Bailey, and C.J. Cavalitto. J. Am. Chem. Soc.
69, 1710 (1947).
31. S.W. Bass and S.A. Evans. J. Org. Chem. 45, 710 (1980).
32. D.J. Pasto and F. Cottard. J. Am. Chem. Soc. 116, 8973 (1994).
© 2005 NRC Canada