Mediated electron transfer from lithium investigated voltammetrically in tetrahydrofuran: Why are some mediators more effective reducing reagents than others?

Article abstract of DOI:10.1002/poc.1232

A study of a range of aromatic molecules is investigated electrochemically to determine what makes an effective reducing mediator. With the aim of developing a better understanding of electron transfers (ETs) mediated from lithium in functional group reduction, a series of single ET reactions are reported in this paper. Typical reaction conditions involved the use of aromatic mediators such as naphthalene, anthracene, 4,4′-di-tert-butyl-1,1′- biphenyl (DBB) with lithium metal in tetrahydrofuran (THF) at -78 °C. The results of these experiments showed that some mediators were more effective reducing reagents than others. Cryoelectrochemical procedures are used to mimic the conditions of the SET reactions in order to investigate the exact nature and role of the mediator formed upon ET. It is demonstrated that electro-generated and stabilized radical anions of anthracene at -78 °C mediate the reduction of organic substrates, whereas the more reactive dianion is quickly protonated and therefore unable to act as an ET reagent; direct electrochemical reduction of the sulfide, phenyl 3-phenylpropylsulfide (RSPh) gives the thiol, thiophenol, and propyl benzene whereas mediated reduction gives the dimer, diphenyl disulfide and propyl benzene. The possibility to selectively reduce a substrate with either a single electron or with two electrons is possible by using either the radical anion (mediated) or via the direct electroreduction. DBB and naphthalene (both single electron accepting species only) have been found to be the most effective reducing reagents in this study. Anthracene and other two-electron accepting species only showed effective reducing ability when a stoichiometric amount of lithium was used therefore preventing the 'over-reduction' to the dianion. Copyright

Full text of DOI:10.1002/poc.1232

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2007; 20: 732–742
Published online 31 July 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/poc.1232
Mediated electron transfer from lithium investigated
voltammetrically in tetrahydrofuran: why are
some mediators more effective reducing
reagents than others?
Natasha C. L. Wood,1 Christopher A. Paddon,1 Farrah L. Bhatti,2 Timothy J. Donohoe2
1
*
and Richard G. Compton
1Physical and Theoretical Chemistry Laboratory, Oxford University, South Parks Road, Oxford OX1 3QZ, UK
2Chemistry Research Laboratory, Oxford University, Mansfield Road, Oxford OX1 3TA, UK
Received 22 March 2007; revised 16 May 2007; accepted 21 May 2007
ABSTRACT: A study of a range of aromatic molecules is investigated electrochemically to determine what makes an
effective reducing mediator. With the aim of developing a better understanding of electron transfers (ETs) mediated
from lithium in functional group reduction, a series of single ET reactions are reported in this paper. Typical reaction
conditions involved the use of aromatic mediators such as naphthalene, anthracene, 4,40-di-tert-butyl-1,10-biphenyl
(DBB) with lithium metal in tetrahydrofuran (THF) at ꢀ78 8C. The results of these experiments showed that some
mediators were more effective reducing reagents than others. Cryoelectrochemical procedures are used to mimic the
conditions of the SET reactions in order to investigate the exact nature and role of the mediator formed upon ET. It is
demonstrated that electro-generated and stabilized radical anions of anthracene at ꢀ78 8C mediate the reduction of
organic substrates, whereas the more reactive dianion is quickly protonated and therefore unable to act as an ET
reagent; direct electrochemical reduction of the sulfide, phenyl 3-phenylpropylsulfide (RSPh) gives the thiol,
thiophenol, and propyl benzene whereas mediated reduction gives the dimer, diphenyl disulfide and propyl benzene.
The possibility to selectively reduce a substrate with either a single electron or with two electrons is possible by using
either the radical anion (mediated) or via the direct electroreduction. DBB and naphthalene (both single electron
accepting species only) have been found to be the most effective reducing reagents in this study. Anthracene and other
two-electron accepting species only showed effective reducing ability when a stoichiometric amount of lithium was
used therefore preventing the ‘over-reduction’ to the dianion. Copyright # 2007 John Wiley & Sons, Ltd.
KEYWORDS: mediator; cryoelectrochemistry; anthracene; single electron transfer (SET); tetrahydrofuran (THF)
INTRODUCTION
reduced employing DBB and naphthalene1 as the medi-
ators;2–4 the reductive ring opening of oxygen-containing
benzo-fused heterocycles,5 structural modification of
carbohydrates,6 the reductive decyanation of nitriles,7 and
the coupling of carbonyl compounds8 have also been
reported. In addition, Donohoe et al. 9,10 have investigated
a number of applications of mediated syntheses including
the total synthesis of cylindricine A,11 the formation of
enantiopure pyrroline building blocks for natural product
synthesis,12 and in the partial reduction of hetero-
cycles.13,14 From an electrochemical perspective, redox
catalysis is gaining wide popularity particularly in
biosensor research, environmental applications, and in
biological investigations (proteins, DNA, etc.). The work
The role of mediators as homogeneous electron-transfer
(ET) reagents for the reductive cleavage of functional
groups is widely employed in synthetic organic chem-
istry. Typically, lithium metal is reacted with an arene
electron-transferring catalyst and substrate, the aromatic
radical anion being generated in situ (Scheme 1).
In particular, 4,40-di-tert-butyl-1,10-biphenyl (DBB)
and naphthalene have attracted much interest as being the
most effective mediators. In the past few decades many
examples of the use of these mediators has been reported.
For example, a range of functional groups which include
thioethers, allylic and benzylic thiols, acetals, disulfides,
tertiary alkyl chlorides, and heterocycles have been
15–17
and Parker,18 in particular, has signifi-
´
of Saveant
cantly developed the area of redox catalysis.
*Correspondence to: R. G. Compton, Physical and Theoretical Chem-
istry Laboratory, Oxford University, South Parks Road, Oxford OX1
3QZ, UK.
With a range of mediators available, both in synthetic
and in homogeneous redox catalysis, it is recognized that
some mediators are more ‘effective’ reducing reagents
E-mail: richard.compton@chem.ox.ac.uk
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742

MEDIATED ELECTRON TRANSFER
733
naphthalene (both single electron accepting species only)
have been found to be the most effective reducing
reagents in this study. Anthracene and other two-electron
accepting species only showed effective reducing ability
when a stoichiometric amount of lithium was used
therefore preventing the ‘over-reduction’ to the dianion.
In these cases, the formation of dimers was feasible.
Scheme 1.
EXPERIMENTAL
Reagents
than others. Given the structural similarities between
many of these mediators (polyaromatic compounds in
particular) the question arises of why this should be?
What makes a mediator effective as a reducing reagent?
Furthermore, the ability of some mediators to accept more
than one electron raises the interesting question of
whether it is possible to ‘selectively’ mediate with either
(a) a monoanion or (b) a dianion to yield, for example, a
monomeric or dimeric product, respectively. Finally, how
does a synthetic-mediated ET reaction compare to that of
a direct electrochemical reaction whereby the electrode
serves as the electron source? The aim of this
investigation is to answer these fundamental questions
which have been raised frequently in the literature,1,4,5,7
and to provide an insight into what makes a mediator
‘effective’.
Herein, a series of mediated reactions are reported in
this paper. Typical reaction conditions involved the use of
aromatic mediators such as naphthalene, anthracene,
DBB with lithium metal in tetrahydrofuran (THF) at
ꢀ78 8C. The results of these experiments showed that
some mediators were more effective reducing reagents
than others. For example, DBB and naphthalene gave
>70% yields of reduced phenyl 3-phenylpropylsulfide
and trans-1-(3-iodopropoxy)-4-(methoxymethoxy)cyclohe-
xane compared with anthracene and 2,20-dimethoxy-
1,10-binaphthyl which did not reduce the substrates but
did yield 9,10-dihydroanthracene in the former case.
These interesting observations raised the question as to
why this should be the case for a series of structurally
similar aromatic mediators?
Tetra-n-butylammonium perchlorate (TBAP) and tetra-
n-butylammonium hexafluorophosphate (TBAF) (Fluka),
anthracene, ferrocene, ferrocenium hexafluorophosphate,
and phenyl 3-phenylpropylsulfide (Aldrich) were used as
received and without any further purification. Anhydrous
THF was purified by filtration through two columns of
activated alumina (grade DD-2) as supplied by Alcoa,
employing the method of Grubbs and coauthors.20
Acetonitrile (ACN) (Fisher Scientific) was dried over
molecular sieves (BDH Laboratory Reagents) and kept
under argon before use. DBB and naphthalene (Aldrich)
were recrystallized from methanol before use.
2,20-Dimethoxy-1,10-binaphthyl was synthesized from
the corresponding alcohol (Aldrich) by methylation
(MeI, K2CO3, DMF). Using the corresponding carboxylic
acids (Aldrich), methyl naphthalene-2-carboxylate and
methyl naphthalene-1-caboxylate were prepared by
esterification (MeOH, H2SO4).
The halides, trans-1-(3-iodopropoxy)-4-(methoxy-
methoxy)cyclohexane (R-I), trans-1-(3-bromopropoxy)-
4-(methoxymethoxy)cyclohexane (R-Br), and trans-1-(3-
chloropropoxy)-4-(methoxymethoxy)cyclohexane (R-Cl)
were prepared from a general precursor which has been
described previously.21,22
Voltammetry
Herein, we used cyclic voltammetry (CV) in an
experimental setup designed to mimic the conditions of
the SET reactions in order to investigate the exact nature
and role of the mediator formed upon ET. Developing our
previous work with naphthalene19 we show, as a case
example, that electro-generated and stabilized radical
anions of anthracene at ꢀ78 8C mediate the reduction of
organic substrates, whereas the more basic dianion is
quickly protonated and therefore unable to act as an ET
reagent. For example, reductive mediation of the sulfide,
phenyl 3-phenylpropylsulfide (RSPh) yields the dimer,
diphenyl disulfide and the propyl benzene, whereas the
direct electrochemical reduction gives the thiol, thiophe-
nol, and propyl benzene. Therefore the possibility to
selectively reduce a substrate with either a single electron
or with two electrons becomes possible. DBB and
Voltammetric measurements were carried out on an
Autolab PGSTAT 20 (Eco-Chemie, Utrecht, Netherlands)
potentiostat. A three-electrode arrangement was used in
an air-tight, three-necked electrochemical cell. The cell
with solid electrolyte was dried in vacuo overnight before
solvent addition and electrochemical experiments. The
working electrodes employed were 5 and 12.5 mm
(radius) platinum disk-electrodes (Cypress Systems,
Inc., Kansas, US) and a 1 mm (diameter) platinum disk-
electrode housed in TeflonTM, with a large area, shiny
platinum wire (Goodfellow Cambridge Ltd, Cambridge,
UK) as the counter electrode. The working electrodes
were carefully polished before use on a clean polishing
pad (Kemet, UK) using 3, 1, and 1/10 mm diamond spray
(Kemet, UK). All working electrodes were carefully dried
prior to use. Before carrying out electrochemical
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

734
N. C. L. WOOD ET AL.
experiments, the microdisk radius was electrochemþically
calibrated using literature methodology.23 A Fc/Fc PF6ꢀ
reference electrode was developed for use in THF and at
low temperatures.24 The temperature was monitored and
controlled by an external system (Julabo FT902,
JULABO Labortechnik GmbH, D-77960 Seelbach/
Germany). Typically the solutions were degassed
vigorously for 1 min using impurity-free argon (BOC
gases, Guildford, Surrey, UK) to remove any trace oxygen
and an inert atmosphere was maintained throughout all
analyses. All solutions were prepared under an atmos-
phere of argon using oven-dried glassware such as
syringes and needles used for the transfer of moisture
sensitive reagents and where required all mediator
solutions were prepared and stored at ꢀ78 8C. All
voltammetric measurements were performed inside a
Faraday cage in order to minimize any background noise.
Fresh lithium was prepared by submerging the wire in
pentane while scraping the oxidized surface with a knife.
The lithium metal was then cut directly into the mediator
solution under a stream of argon. The resulting blue–
green solution was cooled to ꢀ78 8C and allowed to stir
for 4 h. The starting material (0.50 mmol) in dry THF
(10 ml) was cooled to ꢀ78 8C and the mediator solution (2
equiv.) added rapidly. The reaction was allowed to stir at
ꢀ78 8C for a further 10 min and quenched by the
dropwise addition of saturated aqueous ammonium
chloride (5 ml). The solution was allowed to warm to
ambient temperature, poured into brine (20 ml), and
extracted with diethyl ether (2 ꢁ 30 ml). The organic
extracts were combined, dried over MgSO4, and
evaporated under reduced pressure.
Microdisk chronoamperometry
The current observed at a microdisk electrode can be
described in terms of the dimensionless current function
and time function:
Controlled-potential electrolyses
Controlled-potential, bulk electrolyses were carried out in
a two-compartment cell, the catholyte and anolyte being
separated by a sintered glass frit. The cathode was a
rectangular platinum plate (area ¼ 2 cm2) (Goodfellow,
UK), and the anode a platinum mesh housed within the
separate compartment with 0.1 M TBAP in THF. A Fc/
FcþPFꢀ6 reference electrode was used as the reference
electrode. All electrolyses were vigorously stirred using a
magnetic stirrer bar and the temperature maintained using
an external cooling system (Julabo FT902, JULABO
Labortechnik GmbH, D-77960 Seelbach/Germany). The
potential was held constant at the required reduction
potential of the electron transfer (ET) reagent (vs. Fc/
FcþPFꢀ6 ) determined from microdisk-cyclic voltammo-
grams.
I
f ¼
(1)
(2)
4FnDrc
4Dt
r2
t ¼
where I is the current, F is the Faraday constant, n the
number of electron(s) transferred, D the diffusion
coefficient, r the disk radius, c the bulk concentration,
and t the time. Microdisk chronoamperometry permits the
simultaneous determination of D and n provided no
coupled chemistry operates on the timescale of the
experiment. The time-dependent current response, I,
resulting from a diffusion-controlled reductive (or
oxidative if required) current after a potential step at a
microdisk electrode is given in Eqn (3):
Saturated aqueous ammonium chloride (2 ml) followed
by water (10 ml) was added to the electrolyzed solution
upon reaction completion. Diethyl ether (20 ml) was then
added and the two fractions separated. The organic
fraction was then washed with aqueous hydrochloric acid
(1 M, 8 ꢁ 10 ml), saturated sodium bicarbonate (10 ml),
and saturated aqueous sodium chloride (10 ml). The
organic layer was dried over magnesium sulfate and then
carefully evaporated under reduced pressure. The
I ¼ ꢀ4nFDcrfðtÞ
where Eqn (4) defines f(t):
(3)
ꢀ1=2
fðtÞ ¼ 0:7854 þ 0:8862tꢀ1=2 þ 0:2146eꢀ0:7823t
(4)
1
products were identified by H NMR, 13C NMR, and,
The above approximations [Eqns (1–4)] were derived
by Shoup and Szabo25 and describe the current response
to within an accuracy of 0.6% over all t. Experimentally,
the chronoamperometric experiment is run over a time
scale incorporating a transition from planar diffusion,
where necessary, LC-MS.
½
Single electron transfer (SET) reaction for
functional group reduction: general
procedure
with a I / D dependence, to steady-state hemispherical
diffusion with a I / D dependent behavior.26 Accordingly,
deconvolution of the parameters D and n is possible from
a single scan. Fitting is achieved using ORIGIN 6.0
(Microcal Software, Inc.) whereby inputting accurate
values for r and c, the software iterates through values of
D and n until the fit of the experimental data is optimized.
To a two-neck flask containing the mediator under
investigation dry THF was added (typical concentration,
0.1 M) under argon, the solution was then cooled to 0 8C.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

MEDIATED ELECTRON TRANSFER
735
Tafel analysis
Kieselgel 60 (40–63 mm). Column fractions were moni-
tored by TLC.
Mass transport-corrected Tafel analysis of a voltammetric
wave is a method that may be used under certain
conditions for studying the kinetic nature of a hetero-
geneous ET.27 When these conditions are met, a plot of
ðF=RTÞðE ꢀ EfoÞ versus ln½ðIlim=IÞ ꢀ 1ꢂ will be linear in
both the limits of electrochemical reversibility and
irreversibility, where I is the steady-state current and
Ilim is the steady-state current in the limit of mas-
s-transport. For an electrochemically reversible ET this
plot will have a gradient of 1 and for an electrochemically
irreversible ET this plot will have a gradient of 1/a.28
RESULTS AND DISCUSSION
A series of SET reactions were performed using various
mediators as the ET reagents. The results, which are
discussed in the first section, showed that some mediators
were more effective reducing reagents than others. The
aim of the subsequent sections therefore was to answer
three key questions. First, why are some mediators more
effective than others? Second, is it possible to reduce a
substrate using either the monoanion or dianion form of a
mediator generated to transfer one or two electrons
selectively? Thirdly, how do the mediated and direct
electrolyses of mediation compare?
Analytical techniques
1H NMR spectra were recorded using a Bruker AV400
(400 MHz) spectrometer in CDCl3 and referenced to
tetramethylsilane (Si(Me)4) as an internal standard.
Signal positions were recorded in d (ppm) with the
abbreviations s, d, t, q, quint., sept., br, app., and m
denoting singlet, doublet, triplet, quartet, quintet, septet,
broad, apparent, and multiplet, respectively. All coupling
constants, J, are quoted in Hz.
Crude electrolysis mixture was analyzed by thin-layer
chromatography (TLC). TLC was performed on Merck
Kieselgel 60 F254 0.25 mm precoated aluminium-backed
silica plates. Compounds were visualized with UV light
and/or by staining with basic potassium permanganate
solution.
Single electron transfer experiments
A series of SET reactions were performed using halides
and sulfides as the organic substrates and with a range of
polyaromatic compounds as mediators. The structures of
the compounds used and the results of the reduction
reactions are given in Tables 1 and 2. The exact
experimental details have been described above and in
previous work.30 It was observed that DBB and
naphthalene (Table 1) used as the mediators gave the
best percentage yields of reduced products. For example,
the sulfide, phenyl 3-phenylpropylsulfide (RSPh), gave
propylbenzene (RH) in 74% yield with naphthalene and
[(3-{[trans-4-(methoxymethoxy)cyclohexyl]oxy}propyl)-
thio]benzene (R0SPh) gave trans-1-(methoxymethoxy)-
Flash chromatography was used to separate the
products following electrolysis and carried out according
29
to the method described by Still et al.
using Merck
Table 1. Voltammetric data for single electron transfer (SET) reagents with n ¼ 1 obtained at ꢀ78 8C in THF (0.1 M TBAP) using
a Pt microelectrode
SET experimental resultsa;
percentage yield of reduced
product (%)
E1/2 (V)þversus
D ꢁ 10ꢀ10
Tafel value
(mV)
ET reagent
DBB
Fc/Fc PFꢀ6
n
(m2 sꢀ1
)
ꢀ3.34
1 ꢃ 0.1
2.3 ꢃ 0.1
132
R-I ! R-H (60%) þ R-R (6%)
R0SPh ! R0-H (90%)
R-Cl ! R-H (92%)
R-Br ! R-H (81%)
O
X
R =
O
O
Naphthalene
ꢀ3.11
1 ꢃ 0.1
2.4 ꢃ 0.1
115
R-I ! R-H (62%)
RSPh ! R-H (74%)
R-Cl ! R-H (68%)
R-Br ! R-H (72%)
X
R =
a SET experiments performed using an excess of lithium.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

736
N. C. L. WOOD ET AL.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

MEDIATED ELECTRON TRANSFER
737
4-propoxycyclohexane (R0H) in 90% yield with DBB
(Table 1). Similar percentage yields of reduction were
also observed for iodides, bromides, and chlorides
(Table 1). In contrast to these results, all of the mediators
in Table 2 gave no reduction products under the general
conditions used (excess lithium, ꢀ78 8C). For example,
anthracene when used as a mediator for the reduction of
the sulfides and iodides described above did not act as an
ET reagent and gave only the partially reduced aromatic
compound: 9,10-dihydroanthracene as the product.
Following these data, an electrochemical procedure
was developed to mimic the conditions used in the SET
reactions and ultimately, investigate the properties of the
mediators.
Figure 2. Voltammetric response of anthracene (3 mM) in
THF (0.1 M TBAP) using a 12.5 mm (radius) plꢀa1tinum micro-
electrode recorded at a scan rate of 10 mV s (at ꢀ78 8C)
Voltammetric analyses
Steady-state voltammetry, cyclic voltammetry,
and microdisk chronoamperometry of the med-
iators. Using microdisk voltammetry,31 the steady-state
voltammograms for the mediators used in the SET
reactions were obtained. Figures 1–3 show the voltam-
mograms for DBB, anthracene, and methyl naphthale-
ne-1-carboxylate, respectively, in THF (0.1 M TBAP) at
ꢀ78 8C. The half-wave potential, E1/2, and Tafel
slope27,28,31 for each compound are given in Tables 1
and 2. DBB and naphthalene showed a single wave
(Table 1) whereas the mediators in Table 2 all showed
double waves. Tafel analysis for anthracene (Table 2)
showed that the first wave waþs reversible (electrochemi-
cally) (E1/2 ¼ ꢀ2.47 vs. Fc/Fc PFꢀ6 ; 57 mV) whereas the
second wave is quasi-irreversible (E1/2 ¼ ꢀ3.14 vs. Fc/
FcþPFꢀ6 ; 94 mV). Voltammetry performed at low tem-
peratures (ꢀ78 8C) in THF allowed the full resolution of
the voltammetric waves for some mediators in Table 2 and
also for DBB (Table 1) from the solvent break-
down.22,30,32–34
Microdisk chronoamperometry26 was used to simul-
taneously determine the number of electron(s) trans-
ferred, n, to each mediator and the diffusion coefficient, D
in THF at ꢀ78 8C. The results are shown in Tables 1 and 2
for each mediator. DBB and naphthalene are n ¼ 1 ET
reagents (Table 1) while anthracene and 2,20-dimethoxy-
1,10-binaphthyl, for example (Table 2), are n ¼ 2 ET
reagents.
It was hypothesized at this stage that the dianions in
Table 2 were unable to act as mediators due to instability
of the reactive intermediate towards protonation. Further
investigations using CV at a macrodisk electrode were
performed. Figure 4 shows the voltammogram for
anthracene at ꢀ78 8C in THF (0.1 M TBAP). The first
peak is electrochemically reversible (Ep ¼ ꢀ2.72 V vs.
Fc/FcþPFꢀ6 ) while the second peak, as expected, was
chemically irreversible (Ep ¼ ꢀ3.43 V vs. Fc/FcþPF6ꢀ).
These results suggested that the dianion was being
Figure 3. Voltammetric response of methyl naphthalene-1-
carboxylate (3.3 mM) in THF (0.1 M TBAP) using a 5 mm
(radius) platinum microelectrode recorded at a scan rate
of 10 mV sꢀ1 (at ꢀ78 8C)
Figure 1. Voltammetric response of DBB (2.1 mM) in THF
(0.1 M TBAP) using a 13 mm (radius)ꢀp1latinum microelectrode
recorded at a scan rate of 10 mV s (at ꢀ78 8C)
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

738
N. C. L. WOOD ET AL.
protonated by the solvent and supported the results
obtained from both the SET reactions where anthracene
gave 9,10-dihydroanthracene as the only product
(Table 2).
Steady-state voltammetry, cyclic voltammetry,
and microdisk chronoamperometry of the or-
ganic substrates. In order to compare fully the
synthetic experiments electrochemically, the reduction
of various organic substrates by electro-generated media-
tors was required and under the same experimental condi-
tions (THF, ꢀ78 8C). As above, the steady-state voltam-
metry of the iodide (trans-1-(3-iodopropoxy)-4-(methoxy-
methoxy)cyclohexane) (R-I), (trans-1-(3-bromopropoxy)-
4-(methoxymethoxy)cyclohexane) (R-Br), and (trans-
1-(3-chloropropoxy)-4-(methoxymethoxy)cyclohexane)
(R-Cl) was recorded and is shown in Fig. 4. The order of
ease of reduction was found to be: R-I > R-Br > R-Cl.
The sulfide, phenyl 3-phenylpropylsulfide (RSPh) was
also analyzed voltammetrically and the voltammogram
recorded is shown in Fig. 5. The sulfide, [(3-{[trans-4-
(methoxymethoxy)cyclohexyl]oxy}propyl)-thio]benzene
(R0SPh) has been previously analyzed.22 Table 3 sum-
marizes the half-wave potential data and where applicable
Tafel slope data are given. Finally, microdisk chron-
oamperometry was used to determine both n and D. For
all of these organic substrates n ¼ 2. For all of these
compounds it can be observed that thþeir reduction
potentials are all >ꢀ3.2 V versus Fc/Fc PFꢀ6 and so
mediation using aromatic molecules at lower potentials is
advantageous.
Figure 5. Voltammetric response of phenyl 3-phenylpro-
pylsulfide (8.2 mM) in THF (0.1 M TBAP), using a 5 mm
(radius) platinuꢀm1 microelectrode and recorded at a scan
rate of 10 mV s (ꢀ78 8C)
that radical anion mediators in THF (ꢀ78 8C) are able to
act as ET reagents in contrast to dianions which rapidly
protonate. By using voltammetry, the potential at the
working electrode can be controlled accurately and
therefore an assessment of the anion and dianion
separately can be made. Moreover, each organic substrate
being an n ¼ 2 species allows the possibility for different
products based on direct versus mediated electrochem-
istry19 to be obtained.
Next, mediation by anthracene and coulometric
experiments was required in order to confirm the theory
Mediated electrochemistry and simulations
Using a similar procedure to our work involving
naphthalene as a mediator in THF (ꢀ78 8C)19 exper-
iments involving the sulfide, phenyl 3-phenylpro-
pylsulfide (RSPh) and anthracene as the mediator were
Figure 4. Voltammetric response of (a) trans-1-(3-iodopro-
poxy)-4-(methoxymethoxy)cyclohexane (R-I) (3 mM); (b)
trans-1-(3-bromopropoxy)-4-(methoxymethoxy)cyclohexane
(R-Br) (3 mM); and (c) trans-1-(3-chloropropoxy)-4-(methoxy-
methoxy)cyclohexane (R-Cl) (3 mM) in THF (0.1 M TBAP)
using a 5 mm platinum electrode. Scan rate: 10 mV sꢀ1
(ꢀ78 8C)
Figure 6. Cyclic voltammetry of anthracene (2.5 mM), at
ꢀ78 8C in THF (0.1 M TBAP) using a platꢀin1um electrode
(1 mm) recorded at a scan rate of 250 mV s
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

MEDIATED ELECTRON TRANSFER
739
Table 3. Organic substrates used in electrochemical analyses: Electrochemical parameters obtained from voltammetric
experiments performed at ꢀ78 8C
E1/2 (V)þversus
D ꢁ 10ꢀ10
Tafel
value (mV)
ET reagent
Fc/Fc PFꢀ6
n
(m2 sꢀ1
)
Phenyl 3-phenylpropyl sulfide
ꢀ3.53
ꢀ3.25
ꢀ3.6
2 ꢃ 0.3
3.6 ꢃ 0.2
2.0 ꢃ 0.1
2.1 ꢃ 0.2
1.9 ꢃ 0.3
128
207
—
S
R-I
2 ꢃ 0.2
2 ꢃ 0.3
2 ꢃ 0.2
O
X
X
X
R =
O
O
O
O
O
R-Br
R =
O
O
R-Cl
ꢀ3.7
—
R =
O
performed. If our theory above was correct, then the first
wave should mediate whereas the second wave would not.
The steady-state voltammetry for anthracene with and
without RSPh (at various concentrations) in THF
(ꢀ78 8C) is shown in Fig. 7(A). It was observed that
the wave height (E1/2 ¼ ꢀ2.47 V vs. Fc/FcþPF6ꢀ) corre-
sponding to the first ET increased upon additions of the
sulfide. These data are recorded in Table 4. In an
analogous experiment the mediation was performed at
ꢀ50 8C in THF and the results are shown in Fig. 7(B). As
for the ꢀ78 8C case, the wave increased upon addition of
the sulfide (Table 4).
If the proposed mechanism for the mediated ET
follows the well-established electrochemical EC mech-
anism, it may be successfully simulated by DIGISIM
software. The data used in the simulations (as for previous
investigations using naphthalene)19 were obtained from
the electrochemical analyses described above: E1/2, n, c,
and D values (Tables 2 and 3). For the ꢀ50 8C simula-
tions, the diffusion coefficient, D, for anthracene was
determined from the limiting current at a given tempera-
ture on the steady-state voltammogram of anthracene
reduction using Ilim ¼ 4nFDrc, where Ilim is the limiting
current, F is Faraday’s constant (96 485 C molꢀ1), r the
electrode radius, and c the concentration. D for the sulfide
was determined by calculating the activation energy, Ea,
for diffusion from a plot of ln Ilim vs: 1=T at a range of
temperatures followed by an Arrhenius-type calculation
to obtain D at ꢀ50 8C. Using these parameters, which also
included the radius r of the electrode (5 mm), initial
concentrations of species (1 mM anthracene, fixed initial
concentration RSPh), temperature T (ꢀ78 8C, ꢀ50 8C),
transfer coefficient a values of 0.5, and scan rate v equal to
that of the experimental scan rate, simulations were
performed in order to obtain the kinetic parameters: ks and
kf (heterogeneous and homogenous rate constants,
respectively). The fits obtained at ꢀ78 and ꢀ50 8C are
shown in Fig. 8(A) and 8(B), respectively.
From these simulations, ks and kf values of ca. 1 cm sꢀ1
and 60 L molꢀ1 sꢀ1 were obtained for mediation by
anthracene at ꢀ78 8C. At ꢀ50 8C, ks and kf values of
1 cm sꢀ1 and 100 L molꢀ1 sꢀ1 were determined. The
limiting currents obtained by simulations are shown in
Table 4 and show excellent agreement to those obtained
experimentally. At higher temperatures, mediation was
not observed and no catalytic increases in the current
wave height were observed.
Coulometric experiments
To support the simulation and synthetic evidence that
anthracene mediates as the monoanion but not as the
dianion a number of electrolyses were performed.
Phenyl 3-propylsulfide (RSPh) was reduced in the
presence of anthracene on a preparatory scale using
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

740
N. C. L. WOOD ET AL.
Figure 8. A: Overlay of (a) the theoretical simulation result
(solid line) with (b) (dashed line) the RSPh response (23 mM,
scan rate 5 mV sꢀ1, 5.0 mm (radius)). Parameters used for the
simulations were as follows: T ¼ ꢀ78 8C, a ¼ 0ꢀ.51,
Figure 7. A: Voltammetric response of (a) anthracene
(1 mM) in THF (0.1 M TBAP), using a 5 mm (radiuꢀs)1platinum
microelectrode recorded at a scan rate of 10 mV s and with
phenyl 3-phenylpropylsulfide (b) 12 mM (5 mV sꢀ1), (c)
16 mM (2 mV sꢀ1), and (d) 31 mM (2 mV sꢀ1) at ꢀ78 8C. B:
Voltammetric response of (a) anthracene (1 mM) in THF
(0.1 M TBAP), using a 5 mm (radius) platinum microelectrode
recorded at a scan rate of 10 mV sꢀ1 and with phenyl
3-phenylpropylsulfide (b) 9.4 mM (5 mV sꢀ1) (c) 18.3 mM
(5 mV sꢀ1), and (d) 26.8 mM (5 mV sꢀ1) at ꢀ50 8C
Danthracene ¼ 2.9 ꢁ 10ꢀ6 cm2 sꢀ1, DRSPh ¼ 3.6 ꢁ 10ꢀ6 cm2 s
,
ks ¼ 1 cm sꢀ1, kf ¼ 60 L molꢀ1 sꢀ1. B: Overlay of (a) the theor-
etical simulation result (solid line) with (b) (dashed line) the
RSPh response (18 mM, scan rate 5 mV sꢀ1, 5.0 mm (radius)).
Parameters used for the simulations were as follows:
T ¼ ꢀ50 8C,
a ¼ 0.5,
Danthracene ¼ 5.7 ꢁ 10ꢀ6 cm2 sꢀ1
,
DRSPh ¼ 8 ꢁ 10ꢀ6 cm2 sꢀ1, ks ¼ 1 cm sꢀ1, kf ¼ 95 L molꢀ1 sꢀ1
Table 4. Comparison of limiting currents observed at different temperatures and concentrations of RSPh with simulated
currents
Temperature (K)
kf (L molꢀ1 sꢀ1
)
[RSPh] (mM)
Ilim (experimental) (nA)
Ilim (simulated) (nA)
ꢀ78 8C
60
0
0.58
0.67
0.75
0.82
0.84
1.13
1.48
1.61
1.73
1.79
0.58
0.73
0.76
0.80
0.84
1.13
1.48
1.62
1.72
1.81
12.01
16.12
23.16
30.79
0
9.37
18.3
26.83
34.94
ꢀ50 8C
95
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

MEDIATED ELECTRON TRANSFER
741
Table 5. Controlled-potential, bulk electrolyses in THF (0.1 M) at ꢀ78 8C
Number of
electron(s), n
Potential versus
Starting material
Mediator
Fc/FcþPFꢀ6 (V)
Product(s) isolated
Phenyl 3-phenylpropylsulfide (RSPh)
RSPh
Anthracene
RSPh
Anthracene
Anthracene
None
1
2
2
2
ꢀ2.7
ꢀ3.5
ꢀ3.3V
ꢀ3.7
Phenyl disulfide (PhS-SPh)a
No reduction
9,10-Dihydroanthracene
Thiophenol; propylbenzene
(PhSH; Ph(CH)2CH3)a
trans-1-(Methoxymethoxy)-4-
propoxycyclohexane (R-H)a
None
trans-1-(3-Iodopropoxy)-4-
(methoxymethoxy)cyclohexane (R-I)
None
2
ꢀ3.2
Products isolated.
a Characterization of PhS-SPh and RH has been described previously.19,30
controlled-potential electrolysis. For electro-generation
of the radical anion of anthracene the potential was held at
ꢀ2.7 V (vs. Fc/FcþPF6ꢀ) (Table 2). Notably, this value is
lower than the value for the direct reduction of the sulfide,
Table 3. The results of the electrolysis are shown in
Table 5. The indirect reduction of RSPh gave the product
diphenyl disulfide (PhS-SPh) as obtained using naphtha-
lene19 as the mediator and distinct from the thiol which is
obtained from direct reductions of the sulfide, RSPh
(Table 5).
Following this, a second electrolysis was performed
whereby the potential was held constant at ꢀ3.7 V (vs. Fc/
FcþPFꢀ6 ) corresponding to the formation of the dianion of
anthracene. No reduction products were obtained from
this experiment.
Finally, the potential of a solution of anthracene only
was held constant at ꢀ3.7 V (vs. Fc/FcþPF6ꢀ). The results
are shown in Table 5. The formation of 9,10-dihydro-
anthracene was evident with the protons likely to be
coming from both the supporting electrolyte and solvent.
These data support our finding above in that the dianion is
too reactive within the timescale required for mediation,
whereas the radical anion is stabilized to some extent by
the lower temperature, enough to mediate.
a direct electrochemical reaction whereby the electrode
serves as the electron source?
First, in regard to the behavior of the different
mediators, naphthalene and DBB are known to be by
far the most effective mediators. Table 1 shows that the
radical anions of these species are thermodynamically
more powerful reducing reagents than the aromatic arenes
of all the other mediators investigated as shown by their
formal potentials (Tables 1 and 2). It is likely that this is a
reason for their greater effectiveness. Second, the use of a
dianion to bring about a mediated two-electron process
appears to be synthetically of no use since the dianions are
much more basic species than the monoanion, so that
attempts at mediation at the dianion level result in the
abstraction of protons from the solvent in preference to
the reduction. Third, direct electroreduction of iodides
and sulfides is a two-electron process, contrasting to the
one ET seen using mediators. This complementarity of
direct and mediated reduction is a potentially useful
synthetic strategy.
Acknowledgements
The SET reactions were repeated using the iodide and
sulfides listed in Tables 1–3 using a stoichiometric
amount of lithium. The results showed that, although not a
high percentage yield of reduced product, anthracene and
2,20-dimethoxy-1,10-binaphthyl, for example, can med-
iate at the monoanion level.
C. A. P. and F. L. B. thank EPSRC for project studentships
(Grant GR/T05011/01).
REFERENCES
1. Yus M, Herrera RP, Guijarro A. Chem. Eur. J. 2002; 8: 2574.
2. Yang A, Butela H, Deng K, Doubleday MD, Cohen T. Tetrahedron
2006; 62: 6526.
3. Yus M, Foubelo F. Targets Heterocycl. Syst. 2002; 6: 136.
4. Yus M, Martinez P, Guijarro A. Synthetic Commun 2003; 33: 2365.
5. Yus M, Moreno B, Foubelo F. Synthesis 2004; 7: 1115.
6. Soler T, Bachki A, Falvello LR, Foubello F, Yus M. Tetrahedron
Asymmetry 2000; 11: 493.
7. Guijarro A, Yus M. Tetrahedron Lett. 1994; 50: 3447.
8. Zhao H, Li D-J, Deng L, Liu L, Guo Q-X. Chem. Commun. 2003;
4: 506.
9. Bhatti FL, Donohoe TJ. In 232nd ACS National Meeting, San
Francisco, US, 2006.
CONCLUSIONS
We posed three questions at the outset of this paper. First,
what makes a mediator effective as a reducing reagent?
Second, is it possible to ‘selectively’ mediate with either
(a) a monoanion or (b) a dianion to yield, for example, a
monomeric or dimeric product, respectively. Third, how
does a synthetic-mediated ET reaction compare to that of
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc

742
N. C. L. WOOD ET AL.
10. Donohoe TJ, House D. J. Org. Chem. 2002; 67: 5015.
11. Johnson DJ, Donohoe TJ, Bamford MJ. In 232nd ACS National
Meeting, San Francisco, US, 2006.
12. Thomas RE, Donohoe TJ, Linney ID. In 232nd ACS National
Meeting, San Francisco, US, 2006.
13. Donohoe TJ, Johnson DJ, Mace LH, Thomas RE, Chiu JYK,
Rodrigues J, Compton RG, Banks CE, Tomcik P, Bamford MJ,
Ichihara O. Org. Biomol. Chem. 2006; 4: 1071.
14. Donohoe TJ, Harji RR, Cousins RPC. Tetrahedron Lett. 2000; 41:
1327.
21. Paddon CA, Bhatti FL, Donohoe TJ, Compton RG. J. Phys. Org.
Chem. 2007; 115–121.
22. Paddon CA, Bhatti FL, Donohoe TJ, Compton RG. J. Electroanal.
Chem. 2006; 589: 187.
23. Compton RG, Welford PJ, Brookes BA, Wadhawan JD, McPeak
HB, Hahn CEW. J. Phys. Chem. B 2001; 105: 5253.
24. Paddon CA, Compton RG. Electroanalysis 2005; 21: 1919.
25. Shoup D, Szabo A. J. Electroanal. Chem. 1982; 140: 237.
26. Paddon CA, Silvester DS, Bhatti FL, Donohoe TJ, Compton RG.
Electroanalysis 2007; 19: 11.
´
15. Andrieux CP, Saveant J-M, Su KB. J. Phys. Chem. 1986; 90: 3815.
16. Andrieux CP, Blocman C, Dumas-Bouchiat JM, M’Halla F,
27. Albery WJ. Electrode Kinetics. Clarendon Press: Oxford, 1975.
28. Streeter I, Compton RG. Electrochim. Acta 2007; 52: 4305.
29. Still WC, Kahn M, Mitra A. J. Org. Chem. 1978; 43: 2923.
30. Paddon CA, Bhatti FL, Donohoe TJ, Compton RG. Chem. Comm.
2006; 32: 3402.
´
Saveant J-M. J. Electroanal. Chem. 1980; 113: 19.
´
17. Andrieux CP, Dumas-Bouchiat JM, Saveant J-M. J. Electroanaly-
tical Chem. 1978; 87: 39.
18. Parker VD, Tilset M, Hammerich O. J. Am. Chem. Soc. 1987; 197:
7905.
19. Burasov AV, Paddon CA, Bhatti FL, Donohoe TJ, Compton RG.
J. Phys. Org. Chem. 2007; 144–150.
31. Bard AJ, Faulkner LR. Electrochemical Methods: Fundamentals
and Applications (2nd edn). John Wiley & Sons: New York, 2001.
32. Fietkau N, Paddon CA, Bhatti FL, Donohoe TJ, Compton RG.
J. Electroanal. Chem. 2006; 131–141.
20. Pangborn AB, Giardello MA, Grubbs RH, Rosen RK, Timmers RJ.
Organometallics 1996; 15: 1518.
33. Fakhr A, Mugnier Y, Laviron E. Electrochim. Acta 1983; 28: 1897.
34. Mugnier Y, Laviron E. J. Electroanal. Chem. 1980; 180: 375.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 732–742
DOI: 10.1002/poc