Paired electrosynthesis at the femtoliter scale: Formation of 9,10-anthracenedione from the oxidation of anthracene and reduction of dioxygen

Article abstract of DOI:10.1021/ja952600h

A system is described in which the unique properties of a pair of microband electrodes are exploited both to initiate and to probe the products of a paired electrosynthesis using short-lived species. In particular, the oxidation of anthracene is coupled with the simultaneous reduction of dioxygen in acetonitrile to yield the anthracene radical cation and the superoxide anion. These latter react within a femtoliter scale volume to form initially 9,10-dihydro-9,10-epidioxyanthracene, which rearranges into 9,10-anthracenedione and, presumably, dihydrogen via an electron transfer catalyzed process.

Full text of DOI:10.1021/ja952600h

1482
J. Am. Chem. Soc. 1996, 118, 1482-1486
Paired Electrosynthesis at the Femtoliter Scale: Formation of
9,10-Anthracenedione from the Oxidation of Anthracene and
Reduction of Dioxygen
Christian Amatore* and Alan R. Brown
Contribution from the Ecole Normale Supe´rieure, De´partement de Chimie, URA CNRS 1679,
24 Rue Lhomond, 75231 Paris Cedex 05, France
ReceiVed August 2, 1995. ReVised Manuscript ReceiVed NoVember 15, 1995X
Abstract: A system is described in which the unique properties of a pair of microband electrodes are exploited both
to initiate and to probe the products of a paired electrosynthesis using short-lived species. In particular, the oxidation
of anthracene is coupled with the simultaneous reduction of dioxygen in acetonitrile to yield the anthracene radical
cation and the superoxide anion. These latter react within a femtoliter scale volume to form initially 9,10-dihydro-
9,10-epidioxyanthracene, which rearranges into 9,10-anthracenedione and, presumably, dihydrogen Via an electron
transfer catalyzed process.
Introduction
documented,5 while their synthetic applications remain rather
scarce. This fact stems from the normal practice of using such
electrodes for which the characteristic dimensions are less than
the width of the convection-free layer under the experimental
conditions used. The concomitant small total charge passed,
and thus chemical change wrought, has led to little enthusiasm
from synthetic chemists. Perhaps the most notable exception
has been in the use of colloidal suspensions of conducting
particles in an externally applied electric field,6 effectively as a
randomly dispersed array of spherical microelectrodes. It must
be borne in mind though, that the extraordinary properties of
microelectrodes can be exploited with an electrode of which
only one of the characteristic dimensions is small, leaving open
the possibility of bringing about macroscopic change in a
solution. There would be little point in performing a conven-
tional electrolysis at an electrode which was effectively an ornate
conventional electrode, but we propose that the system described
above can be achieved through the use of ultramicroelectrodes
to carry out a coupled electrosynthesis.7 Previously, the distance
between the anode and cathode in the undivided cells employed
and thus the distance which the electrogenerated reagents must
cross unchanged has been in the order of at least tens of
millimeters. Using a pair of closely spaced microband elec-
trodes for such a reaction leads to the definition of a reaction
volume where the diffusion layers of these two electrodes
overlap, as in previous studies involving the simultaneous
reduction and oxidation of a single compound on adjacent
microbands leading to electrochemically generated lumines-
cence.8
The reactions between strong nucleophiles and strong elec-
trophiles, though of profound fundamental and presumably
synthetic interest, have been little studied. This is due largely
to the difficulty of generating such powerful reagents together,
conditions favorable for the formation of one generally being
inimical to the other. There are however a few circumstances
under which such a combination becomes feasible. For instance,
photoinitiated electron transfer through excitation of a charge
transfer complex1 can lead to the formation of a strong
electrophile-nucleophile pair, which may or may not react
chemically.2 Evidently, the scope of this technique is limited
by the need to form a sizable concentration of the charge transfer
complex in the first place. A second alternative, also based on
the activation of inert precursors through electron transfer,
consists in pulsing the potential of an electrode alternately to
anodic and cathodic potentials to generate in turn an electrophile
and a nucleophile.3 However, the need to perform a series of
independant electrolyses implies that the solution used must be
of reasonable conductivity, which again may limit the scope of
the method. It occurred to us that this and other problems could
be resolved by employing, as before, two electrogeneration
potentials, separated not in time, but in space. The activated
products formed at each separated electrode could then react
together thanks to the merging of the diffusion layers extending
at each electrode. A similar notion has been applied previously
to studies at a rotating ring-disc electrode, where the products
of the separated electrode reactions are mixed by hydrodynamic
forces4 rather than by diffusion. In order that the highly reactive
and thus short-lived species which we wished to employ should
meet by diffusion and then react, the distance between the sites
where they are generated should be as small as possible. To
this end, the ultramicroelectrode seemed the tool of choice.
The whole is then effectively a microreactor of only several
femtoliters volume, the bulk solution serving solely as a store
of reagents and products. Thus, starting from a mother liquor
containing suitably unreactive precursors, powerful nucleophiles
and electrophiles, or super-acidic and super-basic species, which
would normally react with the electrolyte solution before
The analytical uses of these electrodes are numerous and well
* To whom any correspondance should be addressed.
X Abstract published in AdVance ACS Abstracts, January 1, 1996.
(1) Hilinski, E. F.; Masnovi, J. M.; Amatore, C.; Kochi, J. K.; Rentzpis,
P. M. J. Am. Chem. Soc. 1983, 105, 6167-6168.
(2) Kochi, J. K. Angew. Chim. 1988, 27(10), 1227-1388.
(3) Mayeda, E. A.; Bard, A. J. J. Am. Chem. Soc. 1973, 95(19), 6223-
6226.
(5) Microelectrodes; Theory and Applications, Nato ASI; Montenegro,
M. I., Queiros, M. A., Daschbach, J. L., Eds.; Kluwer Academic Publish-
ers: Dordrecht, 1991.
(6) Fleischmann, M.; Ghorghchian, J.; Rolinson, D.; Pons, S. J. Phys.
Chem. 1986, 90(23), 6392-6400.
(7) Baizer, M. M. In Organic Electrochemistry, 3rd ed.; Lund, H., Baizer,
M. M., Eds.; Marcel Dekker, Inc.: New York, 1991; pp 1421-1430.
(8) Bartelt, J. E.; Drew, S. M.; Wightman, M. R. J. Electrochem. Soc.
1992, 139, 70.
(4) Maloy, J. T.; Bard, A. J. J. Am. Chem. Soc. 1971, 93(23), 5968-
5981.
0002-7863/96/1518-1482$12.00/0 © 1996 American Chemical Society

Paired Electrosynthesis at the Femtoliter Scale
J. Am. Chem. Soc., Vol. 118, No. 6, 1996 1483
meeting, may be generated simultaneously in close proximity,
and so combine chemically. It is the purpose of the present
work to demonstrate the feasibility of such a synthesis. For
this reason we have limited ourselves to the simple case of a
single pair of band electrodes. In order to bring about the bulk
transformation of a solution of large volume the use of an
electrode such as an interdigitated array with the same inter-
electrode gap but greater overall active length needs to be
envisaged, but was not required for the present investigation.
The synthetic reaction attempted was the coupling of the
anthracene radical cation with the superoxide anion to form, at
least initially, 9,10-dihydro-9,10-epidioxyanthracene (1). This
reaction was chosen on the grounds that the required oxidation
and reduction are both well known, as is the putative product
(1), although there appears to be no previous published
electrochemical study of the compound. The oxidation of
anthracene in the chosen solvent, acetonitrile, is chemically
irreversible on the normal voltammetric time scale due to a
reaction between the initially formed radical cation and the
solvent, although some doubt remains about the relevant pseudo-
first-order rate constant.9 The reduction of oxygen in acetonitrile
has long been known to give rise to the superoxide ion, which
is stable in dry acetonitrile on a time scale of minutes. The
apparent degree of reversibility depends strongly on the electrode
material used,10 hence the choice in the present investigation
of a gold cathode in the electrode assembly.
Figure 1. (a) Response to cyclic voltammetry at 200 mV s-1 of the
cathode (Au) of the double band array, the potential of the anode (Pt)
being fixed at +1.6 V: (curve A) anthracene 1.0 mM in argon saturated
solution; (curve B) dioxygen saturated solution; (curve C) anthracene
1.0 mM in dioxygen saturated solution (8.2 mM). The inset shows
the response of the system for a concentration of anthracene of 8.2
mM, conditions being otherwise identical to those for curve C, while
the bar in the lower right-hand quadrant shows the height of the wave
for the oxidation of anthracene under the same conditions. Solvent
was MeCN, NBu4BF4 0.2 M, at ambient temperature. (b) Responses
of the anode (lower trace) and cathode (upper trace) to linear sweep
voltammetry at the anode with the cathode being held at -1.2 V and
the conditions being otherwise identical to those for curve C in part
(a).
used as received, while 9,10-anthracenedione (Rhoˆne-Poulenc) was
recrystallized from chloroform. 9,10-Dihydro-9,10-epidioxyanthracene
was synthesized by literature methods14 and its identity and purity
confirmed by NMR spectroscopy.
Experimental Section
The electronic electrochemical equipment used was of an entirely
conventional nature, the bipotentiostat being constructed in house. All
potentials are quoted relative to an aqueous SCE reference electrode,
and all voltammograms were recorded at a potential sweep rate of 200
mV s-1. Unless otherwise indicated, positive currents are cathodic.
The double band array used in the present study was constructed using
published techniques,11 as a “sandwich” of window glass, platinum foil
(Goodfellow, 99.95%), Mylar (Energy Beam Sciences, 2 µm thickness),
gold foil (Goodfellow, 99.95%), and a second layer of glass, bonded
with epoxy resin (Epon, type 828), hardened with 10% triethylamine-
tetramine (Aldrich, technical grade). The whole then consisted of a
platinum anode and a gold cathode, each of 5 µm thickness, separated
by 2.5 µm over a length of 4 mm. The array was polished using 1200
grit abrasive paper followed by 300 nm alumina suspension (both Presi)
on a felt cloth. The dimensions and correct mutual alignment of the
elements of the array were respectively determined and assured
voltammetrically12,13 and optically. Experiments were carried out in a
conventional electrochemical cell designed for use on a vacuum line,
the solution volume being 10 mL. A large platinum wire served as
counter electrode.
Tetrabutylammonium tetrafluoroborate was obtained from the reac-
tion of aqueous solutions of tetrabutylammonium hydrogen sulfate and
sodium tetrafluoroborate (both Aldrich) and recrystallized twice from
a methanol-water mixture before being dried under vacuum and used
at a concentration of 0.2 M as supporting electrolyte in all experiments
described. Solutions were saturated with dioxygen unless otherwise
stated. Double band experiments were carried out in the presence of
a small quantity of neutral alumina (Aldrich) previously heated under
vacuum to remove residual water. Acetonitrile (Aldrich, HPLC Grade)
was distilled over CaH2 under an atmosphere of dinitrogen before use.
Dioxygen and argon (both Air Liquide) were passed over CaCl2 and
anhydrous CuSO4 before use. Anthracene (Aldrich, Gold Label) was
Results and Discussion
Figure 1a shows three voltammograms, all produced by
holding the anode of the array at +1.6 V, a potential beyond
that required for the oxidation of anthracene, and performing
cyclic voltammetry at the cathode, thus simultaneously setting
in train synthetic reactions and probing the diffusion layer above
the array. Curve A shows the response in the presence of
anthracene and the absence of dioxygen, curve B that in the
presence of dioxygen and the absence of anthracene, and curve
C the result of the presence of both anthracene and dioxygen.
Curve C is in no way the resultant of the superposition of curves
A and B. The response of the system to the presence of
dioxygen and anthracene in equimolar quantities (8.2 mM) is
shown in an inset, the sharp peak shown on the forward scan
being present whenever the ratio of hydrocarbon to dioxygen
is high. The presence of such a feature in a steady-state
voltammogram can in no way result from a diffusional effect,
and must be due to a chemical reaction. A voltammogram
recorded under the same conditions as those for curve C but
with the anode held at +1.3 V, a potential at which anthracene
is not oxidized, was identical to curve B.
Perusal of curve C in Figure 1a shows that it contains three
interesting features, a plateau between -0.4 and -0.8 V, a sharp
rise in current at around -1 V, and an apparent return wave at
-0.73 V. The initial plateau occurs before the formal reduction
potential for dioxygen in acetonitrile, and is therefore not due
to an interaction with the superoxide ion and has been largely
neglected in this work, other than being subtracted when quoting
the height of the following wave. That the occurrence of this
second wave depends on the simultaneous oxidation of an-
thracene at the anode may be seen from Figure 1b, and it is
also worthwhile to note that the current recorded at the cathode
is approximately two and a half times that due to the oxidation
which gives rise to it. Similarly, that the superoxide ion may
(9) Murphy, M. M.; Stojek, Z.; O’Dea, J. J.; Osteryoung, J. G.
Electrochim. Acta 1991, 36(9), 1475-1484.
(10) Sawyer, D. T.; Chiericato, G.; Angelis, C. T.; Nanni, E. J.; Tsuchiya,
T. Anal. Chem. 1992, 54, 1720-1724.
(11) Bartelt, J. E.; Deakin, M. R.; Amatore, C. A.; Wightman, M. R.
Anal. Chem. 1988, 60, 2169-2171.
(12) Fosset, B.; Amatore, C. A.; Bartelt, J. E.; Michael, A. C.; Wightman,
R. M. Anal. Chem. 1991, 63, 306-314.
(13) Deakin, M. R.; Wightman, R. M.; Amatore, C. A. J. Electroanal.
Chem. 1986, 215, 4961.
(14) Foote, C. S.; Wexler, S.; Higgins, R. J. Am. Chem. Soc. 1968, 90
(4), 975-981.

1484 J. Am. Chem. Soc., Vol. 118, No. 6, 1996
Amatore and Brown
Scheme 1
Figure 2. (a) Curve C of Figure 1a is repeated (curve B) for
comparison with a voltammogram (curve A, top left inset) recorded
under the same conditions, other than that anthracene was replaced by
9,10-anthracenedione, also at a concentration of 1.0 mM. (b) Volta-
mmetric response at a 500 µm diameter platinum disc of A, an argon
saturated 5 mM solution of 1, B a dioxygen saturated solution, and C
a dioxygen saturated 5 mM solution of 1 to cyclic voltammetry at 200
mV s-1
.
Scheme 2
be involved is shown by a careful examination of curve B in
Figure 1a which shows that the potentials concerned correspond
to the very foot of the reduction wave of dioxygen. There is
then a very small concentration of superoxide present in the
diffusion layer of the electrode.
Figure 2a shows the comparison of curve C from Figure 1a
(noted as curve B in Figure 2a) with the steady state wave for
the reduction of anthraquinone (9,10-anthracenedione, noted
ANQ) in the presence of dioxygen. The two waves are almost
identical, except for the previously noted sharp rise in current
in curve B. Such a feature is strongly suggestive of the
involvement of an electrocatalytic mechanism15 in which ANQ
is produced quantitatively in the diffusion layer as soon as a
threshold potential15 is reached. The expected product from
the reaction of the anthracene radical cation and superoxide is
1. This expectation together with the results detailed above
suggested that 1 was converted electrocatalytically into ANQ,
for which reaction we tentatively propose the mechanism16
reported in Scheme 1.
The sharp rise in current referred to earlier is greater than
would be expected on the basis of the reduction of a product
formed stoichiometrically from anthracene (see also Figure 1b).
We ascribe this effect to the presence of a redox catalysis
mechanism17 in which ANQ•- formed by reduction of ANQ
which is generated electrocatalytically from 1 (Scheme 1)
mediates the reduction of dioxygen. This suggestion is further
supported by the data presented in Figure 2b, which show the
voltammetric response at a 500 µm diameter platinum disc
electrode of 1 in argon saturated solution, curve A, of a saturated
solution of dioxygen alone, curve B, and of 1 in a saturated
dioxygen solution, curve C. At low concentrations of 1, both
the wave in this last curve and that due to the direct reduction
of dioxygen are visible. In brief, the response shown in curve
B in Figure 2a is very close to that obtained from an authentic
sample of 9,10-anthracenedione under the same conditions. We
propose that these observations come about as a result of an
electrocatalytic mechanism, the overall reaction being as shown
in Scheme 2 (where An•+ is formed at the anode).
Thus the wave in curve C of Figure 2b at around -1.00 V is
assigned to the reduction of dioxygen, mediated by 9,10-
anthraquinone formed itself Via the electron transfer catalyzed
reorganization of 1. This explains why the plateau of wave C
in Figure 1a and the cathodic current in Figure 1b are greater
than expected given the concentration of anthracene and
therefore the maximum concentration of anthraquinone formed.
In addition, the presence of a double return wave at -0.68 and
-0.44 V may be noted.
The peak at around -1 V in curve C of Figure 2b was found
to grow progressively as the concentration of 1 was increased.
The upper data set (open circles) in Figure 3a shows the
variation of this peak with the concentration of 1 added, as
measured at the gold electrode of the double band array,
operated in the single electrode mode. The lower set of data
(open triangles) shows the variation with the concentration of
anthracene of the height of the plateau following the sharp rise
in current at around -1.00 V (see Figures 1a and 2a), for
voltammograms recorded at the cathode under the same
conditions as those for curve C in Figure 1a. The nonlinear
dependance of the heights of these waves with respectively the
concentration of 1 (circles) or anthracene (triangles) present is
clear. A small quantity of anthracene gives rise to currents far
greater than would be predicted on the basis of the reduction
of a product formed from the anthracene radical cation and the
superoxide ion. The similarity, both in magnitude and form,
of these two sets of data, obtained from very different experi-
ments, is striking.
(15) Amatore, C.; Pinson, J.; Save´ant, J-M.; Thie´bault, A. J. Electroanal.
Chem. 1980, 107, 59-74.
(16) Further evidence for the occurrence of such a mechanism was
obtained from the electrochemistry of (1) at a platinum electrode. The
results obtained, which are not presented here, indicate that this compound
is quantitatively converted into ANQ (E°ANQ ) 0.86 V Vs SCE) within the
diffusion layer following its one-electron reduction, provided that the
potential sweep rate used is less than 2 V s-1. This reaction is completely
frozen at much shorter time scales (>200 Vs-1), allowing the real
voltammogram of 1 (E°1 ) 0.90 V Vs SCE) to be recorded. For intermediate
scan rates, the voltammetry of 1 shows trace crossings characteristic of
electron transfer catalyzed reactions.15 The response recorded at low sweep
rate (see Figure 2a, curve A) is then superficially similar to that obtained
from an authentic sample of ANQ itself, the apparent reversibility of the
reduction then being due to the reversibility of the reduction of this
compound.
(17) For straightforward electrochemical redox catalysis see: (a) An-
drieux, C. P.; Dumas-Bouchiat, J. M.; Save´ant, J-M. J. Electroanal. Chem.
1978, 87, 39-54. For a mechanism involving redox catalysis through a
mediator generated in an electrocatalytic reaction, as in the present case,
see: (b) Amatore, C.; Oturan, M. A.; Pinson, J.; Save´ant, J-M.; Thie´bault,
A. J. Am. Chem. Soc. 1984, 106, 6318.
The dependance of the height of the second plateau in the
forward scan of curve C of Figure 1a with the concentration of
dioxygen in the solution, for three concentrations of anthracene,
is shown in Figure 3b. The amount of dioxygen present was

Paired Electrosynthesis at the Femtoliter Scale
J. Am. Chem. Soc., Vol. 118, No. 6, 1996 1485
Scheme 3
cation radicals and electron-rich species are known18 to be
predominantly electron transfer in nature, this most likely occurs
Via an intermediate complex of excited neutral species formed
upon electron transfer from the superoxide ion and anthracene
cation radical. Subsequent to the easy reduction of 1, 9,10-
anthracenedione is formed electrocatalytically and mediates the
reduction of dioxygen.
Figure 3. (a) Variation of peak current measured at the gold electrode
of the double band electrode as a function of the concentration of the
analytes; 1 + O2 (circles) or anthracene + O2 (triangles). (b) Variation
of plateau height (current in µA) of curve C in Figure 1a with the
concentration (in mM) of dioxygen in solution for the three concentra-
tions of anthracene: 1 (solid circles); 5 (solid diamonds), or 10 mM
(open squares).
One objection that could be raised is the fact that the reductive
process on the cathode of the double band appears to be
triggered at potentials where the reduction of dioxygen is only
incipient. However, once a single molecule of 1 has been
formed it will itself be reduced, leading catalytically Via the
formation of the radical anion of 9,10-anthracenedione to the
further reduction of dioxygen to superoxide. This superoxide
is then available to react with the anthracene radical cation, and
a catalytic cycle is established, leading to a sharp rise in current.
Such autocatalytic mechanisms are well known experimentally
and their theoretical kinetics well documented.19 The overall
reaction is shown in Scheme 3 in which the reactions detailed
in Schemes 1 and 2 are implicit. The close resemblance
between the curves in Figure 2a may now be seen in terms of
the reactions outlined in Scheme 3. In curve B, the “explosive”
rise in current is triggered at around -1 V. Following this, the
response tracks that for 9,10-anthraquinone at a similar con-
centration as almost all the anthracene is converted to the
quinone form. Comparison of the plateau heights of the two
curves in Figure 2a indicates a yield of around 90%.
varied by the partial degassing of a dioxygen saturated solution.
For each point the concentration was determined voltammetri-
cally by comparison with the height of the wave obtained on a
single band for the saturated solution. The anode was then
reconnected and the voltammogram re-recorded in double band
mode. The data have been presented without subtracting the
height of the preceding plateau. This procedure may be justified
if it is assumed that the process giving rise to this plateau is
suppressed as the mechanism causing the following sharp front
is triggered, then the height of this peak should be measured
from the baseline. The apparent return wave in curve C of
Figure 1a or 2a may then be interpreted as the point at which
the mechanism responsible for the sharp current rise is supressed
and the system attempts to return to steady-state behavior. This
interpretation is further justified by the fact that this “return
wave” was never observed to pass below the zero current line
under any circumstances. The data presented demonstrate that
both dioxygen and anthracene play a large part in the response
of the array. It can be seen that the behavior of the system is
consistent over the concentration ranges studied for dioxygen
and anthracene, which spanned a large part of the total solubility
range for each compound. Furthermore, although due to the
scale the data sets in Figure 3a appear to reach a plateau, in
fact the limiting currents continue to increase with the concen-
tration of anthracene or 1 present.
Conclusion
The use of paired microband electrodes allows the coupling
of a strong nucleophile with a strong electrophile, something
which has, up to now, been difficult, due to the instability of
such species and the incompatibility of the conditions required
(18) Koike, K.; Thomas, J. K. J. Chem. Soc., Faraday Trans. 1992, 88
(2), 195-200.
(19) Amatore, C.; Pinson, J.; Save´ant, J. M.; Thie´bault, A. J. Am. Chem.
Soc. 1981, 103, 6930-6937.
(20) One of the referees pointed out the similarity between the proposed
technique and photoinitiated intermolecular electron transfer. Indeed, as
noted in the Introduction section, here a reactive ion pair is “constructed”
from individual ionic components created at each electrode, while in CT
photochemistry, the ion pair results from direct photochemical activation
of the CT complex. The advantage of the electrochemical method described
here is the obviation of the need to form an initial charge transfer complex
between acceptor and donor. Such complexes are, in most cases, only
weakly bound, and thus exist at low concentration. In passing the electron
instead by an external circuit considerable gains in conversion rate and
efficiency may be expected. Also, the use of electrochemical conditions
vs photochemical ones may lead to different overall reactions because of
different followup reactions of the initial product, as evidenced here by the
electron transfer catalyzed conversion of 1 into ANQ.
The arguments outlined above allow a coherent picture to be
drawn of the processes involved in the response of the double
band. Curves A and B of Figure 2a are almost certainly due to
the same electrochemical reaction. The forms of the voltam-
mograms are very close, and the similarity of the dependencies
of the forward peak heights on the concentration of either
anthracene or 1 as shown in Figure 3 is remarkable. This is
explained, in the case of the double band results, by the
successful synthesis of 1. Since reactions between aromatic

1486 J. Am. Chem. Soc., Vol. 118, No. 6, 1996
Amatore and Brown
to generate them. Anthracene and dioxygen have been directly
coupled to form 9,10-anthracenedione, something which would
be difficult to achieve by conventional chemical techniques.20
Moreover, a chemical reaction has been carried out uniquely
through the imposition of a spatially structured reactivity in
solution.
The results presented are directly transposable to interdigitated
arrays (IDA)21,22 as the diffusional regime is similar at these
devices. This promising technique will thus be brought to
maturity, allowing the use of electrodes of greater overall active
area, and so lead to preparative scale reactions.
The technique of paired electrosynthesis may now be
extended beyond the present realm, where the requirement that
the electrogenerated reagents be sufficiently unreactive to cross
a conventional electrochemical cell unchanged is a major
constraint. In addition, although we used an arrangement where
current was passed through a distant counter electrode the
technique would be equally applicable were the current to be
passed uniquely between adjacent microbands, in a way very
reminiscent to what occurs in pit corrosion. Since the associated
Ohmic drop would then be much reduced, solutions of low
conductivity11 could then be exploited.
Acknowledgment. This work has been supported in part
by the CNRS (URA CNRS 1679) and by the Ecole Normale
Supe´rieure (MESR). A.R.B. would like to thank the Engineer-
ing and Physical Sciences Council for a Western European
Fellowship.
JA952600H
(21) Such large electrode arrays have been produced through a variety
of methods5 for analytical purposes, the most robust and thus those most
suited to an eventual preparative application probably being those produced
by photolithographic techniques22 where interelectrode spacers of silicon
dioxide as slim as 0.5 µm have been reported.
(22) See e.g.: (a) Aoki, K.; Morita, M.; Niwa, O.; Tabei, H. J.
Electroanal. Chem. 1988, 256, 269. (b) Chidsey, C. E.; Feldman, B. J.;
Lundgren, C.; Murray, R. W. Anal. Chem. 1986, 58, 601.
The present work was developed to test and establish the
principle of the spatial segregation of chemical reactions based
on the small diffusion layers that are inherent to microelectrodes.