Measurement of local reactivity at liquid/solid, liquid/liquid, and liquid/gas interfaces with the scanning electrochemical microscope: Principles, theory, and applications of the double potential step chronoamperometric mode

Article abstract of DOI:10.1021/jp972587i

A numerical model for scanning electrochemical microscopy (SECM) double potential step chronoamperometry (DPSC) has been developed and examined experimentally. The concept of this new mode of SECM is to generate a reactant in an initial potential step at a tip ultramicroelectrode (UME) positioned close to a target interface. The electrogenerated species diffuses from the tip to the interface, where it may be involved in a chemical process. The reactant is subsequently collected by electrolysis in a second potential step, and the form of the corresponding current-time curve provides information on the nature of the interaction between the initial tip-generated species and the interface. If the species is consumed in an irreversible interfacial process, the current flow during the second potential step is less than when the interface is inert with respect to the species of interest. The theoretical predictions are first examined with DPSC studies on the electrogeneration and collection of ferricyanide ions from aqueous ferrocyanide solutions, at a tip positioned close to aqueous/glass, aqueous/1,2-dichloroethane (DCE), and aqueous/air interfaces, as model examples of inert liquid/solid, liquid/liquid, and liquid/gas interfaces. The case of an active interfacial process is illustrated through studies of the electrogeneration and collection of Br₂, from aqueous sulfuric acid solutions of potassium bromide, at a tip positioned close to aqueous/DCE and aqueous/air interfaces. The transfer of Br₂ across these interfaces is found to be irreversible and effectively diffusion-controlled on the SECM time scale, putting a lower limit on the interfacial transfer rate constant of 0.5 cm s^-1. The experiments carried out at aqueous/air interfaces represent the firs demonstration that SECM can be used to probe liquid/gas interfaces, thereby further diversifying the range of novel environments that can be studied with this instrument.

Full text of DOI:10.1021/jp972587i

J. Phys. Chem. B 1997, 101, 10851-10859
10851
Measurement of Local Reactivity at Liquid/Solid, Liquid/Liquid, and Liquid/Gas Interfaces
with the Scanning Electrochemical Microscope: Principles, Theory, and Applications of the
Double Potential Step Chronoamperometric Mode
Christopher J. Slevin, Julie V. Macpherson, and Patrick R. Unwin*
Department of Chemistry, UniVersity of Warwick, CoVentry CV4 7AL, U.K.
ReceiVed: August 7, 1997X
A numerical model for scanning electrochemical microscopy (SECM) double potential step chronoamperometry
(DPSC) has been developed and examined experimentally. The concept of this new mode of SECM is to
generate a reactant in an initial potential step at a tip ultramicroelectrode (UME) positioned close to a target
interface. The electrogenerated species diffuses from the tip to the interface, where it may be involved in a
chemical process. The reactant is subsequently collected by electrolysis in a second potential step, and the
form of the corresponding current-time curve provides information on the nature of the interaction between
the initial tip-generated species and the interface. If the species is consumed in an irreversible interfacial
process, the current flow during the second potential step is less than when the interface is inert with respect
to the species of interest. The theoretical predictions are first examined with DPSC studies on the
electrogeneration and collection of ferricyanide ions from aqueous ferrocyanide solutions, at a tip positioned
close to aqueous/glass, aqueous/1,2-dichloroethane (DCE), and aqueous/air interfaces, as model examples of
inert liquid/solid, liquid/liquid, and liquid/gas interfaces. The case of an active interfacial process is illustrated
through studies of the electrogeneration and collection of Br2, from aqueous sulfuric acid solutions of potassium
bromide, at a tip positioned close to aqueous/DCE and aqueous/air interfaces. The transfer of Br2 across
these interfaces is found to be irreversible and effectively diffusion-controlled on the SECM time scale, putting
a lower limit on the interfacial transfer rate constant of 0.5 cm s-1. The experiments carried out at aqueous/
air interfaces represent the first demonstration that SECM can be used to probe liquid/gas interfaces, thereby
further diversifying the range of novel environments that can be studied with this instrument.
Introduction
application, the UME probe is positioned in a liquid phase, close
to the interface of interest at which the transfer process is initially
at equilibrium. The electrolysis of a target species at the UME
probe (in the potential step chronoamperometric mode) depletes
its concentration locally, which, in turn, provides the thermo-
dynamic force for the interfacial transfer process in the direction
of the phase containing the UME probe. This serves to enhance
the flux of electroactive material to the UME, compared to the
situation where the interface is inert, and the current flowing
provides quantitative information on the interfacial transfer
kinetics. Reversible reactions studied using this SECM equi-
librium perturbation mode include dissolution,15 adsorption/
desorption,14a and solvent extraction/stripping processes.13
In order to diversify the range of processes that can be studied
with SECM transient methods, we now consider SECM double
potential step chronoamperometry (DPSC) as a new methodol-
ogy for initiating and monitoring heterogeneous reactions at the
local level. When used in conjunction with UMEs in particular,
DPSC has proved to be a powerful approach for characterizing
the diffusion coefficients of electrogenerated species24 and the
lifetimes of transient species involved in solution processes.25
However, DPSC has not previously been used to investigate
the kinetics of processes confined to a target interface. The
basic concept of SECM DPSC is to employ the UME probe to
electrochemically generate a reactive species (B in Figure 1) in
an initial (forward) potential step for a fixed period. The species
could be generated through the electrolysis of an inert precursor
in the solution (A in Figure 1) or via the anodic dissolution of
a pure or amalgam metal electrode.
Scanning electrochemical microscopy (SECM) is continuing
to develop as a powerful device for probing a wide range of
interfacial processes with high spatial and temporal resolution,1
including redox reactions,2-10 ion transfer,11-13 adsorption/
desorption processes,14 and dissolution phenomena.15,16 A major
factor in the growing number of applications of SECM has been
the parallel development of new modes for which quantitative
mass transport models can readily be developed.
Initial work by Kwak and Bard led to the development of a
finite-element model for the steady-state feedback response of
the UME tip for both conductive and inert surfaces,17 the results
of which have been widely used.1 The steady-state model was
subsequently developed to include the chronoamperometric
positive and negative feedback responses, simulated using both
a Krylov integrator18 and the alternating direction implicit finite
difference method (ADIFDM).19 The latter method was used
to investigate the effect of unequal diffusion coefficients of the
redox mediator couple on the chronoamperometric positive
feedback response.20 The effects of finite heterogeneous kinetics
and arbitrary substrate size on the feedback responses were also
considered.3 Models for the chronoamperometric feedback and
generation/collection modes for complex electrode reactions
involving coupled homogeneous kinetics have also been
introduced.21-23
The single potential step chronoamperometric mode has
proved to be a particularly powerful method for inducing and
monitoring reversible transfer processes at both liquid/solid
interfaces14a,15 and immiscible liquid/liquid interfaces.13 In this
During the forward step the electrogenerated species B
diffuses away from the probe and intercepts the interface. If B
interacts with the interface through either an irreversible or
X Abstract published in AdVance ACS Abstracts, November 15, 1997.
S1089-5647(97)02587-X CCC: $14.00 © 1997 American Chemical Society

10852 J. Phys. Chem. B, Vol. 101, No. 50, 1997
Slevin et al.
a solution precursor, may undergo irreversible first-order loss
at the interface. The model allows for arbitrary diffusion
coefficients of the precursor and reactant species.
The methodology is illustrated through several model systems.
For the case of an inert interface, the theoretical characteristics
are verified through studies of the electrogeneration and
collection of ferricyanide from a ferrocyanide solution at a Pt
UME positioned close to aqueous/glass, aqueous/DCE, and
aqueous/air interfaces. The investigation of reactive interfaces
is illustrated through studies of the transfer of Br2, electrogen-
erated from Br- in aqueous sulfuric acid at a Pt UME, across
aqueous/DCE and aqueous/air interfaces. The latter studies
serve to demonstrate that SECM can be used to probe physi-
cochemical processes at liquid/gas interfaces,26 as well as liquid/
liquid and liquid/solid interfaces studied hitherto.
Theory
We consider the situation where the solution initially contains
only a precursor species A that does not interact with the
interface. At time t ) 0, the potential of the UME probe is
stepped to a value to cause the following diffusion-controlled
electrolysis:
A ( ne- f xB
(1)
Figure 1. Schematic of the processes in the tip/substrate gap for SECM
DPSC using the irreversible transfer of a solute across a water/organic
(w/o) interface as an example. During the forward step, species B is
electrogenerated from A [a(i) and b(i)] in the aqueous phase. If B is
inert with respect to the target interface, it simply slowly leaks from
the tip/interface gap by hindered diffusion [a(i)]. Consequently, when
the potential is stepped to reverse the electrode reaction, much of B is
available for collection [a(ii)]. In contrast, if B is active with respect
to the w/o interface, as in b(i) and (ii) where it partitions into the
o-phase, there is a net depletion of B in the w-phase, resulting in a
smaller collector flux [b(ii)] compared to the inert interface case [a(ii)].
resulting in the generation of the reactive species B.
The transport of species A and B in the axisymmetric
cylindrical geometry of the SECM is governed by the following
general dimensionless diffusion equation (where the subscript
i denotes A or B):
∂C
∂τ
∂2Ci
∂Ci ∂2Ci
1
R
i ) λi
+
+
+
(2)
2
∂R
∂Z2
[
]
∂R
The dimensionless terms in eq 2 are defined as follows. R is
the radial coordinate in the axisymmetric cylindrical geometry
(starting at the center of the UME) normalized with respect to
the electrode radius, a. Z is the coordinate normal to the UME
surface normalized with respect to a. Dimensionless time is
defined by
reversible process [Figure 1b(i)], e.g., adsorption, absorption,
or a chemical reaction, its concentration profile will be modified
compared to the situation where there is no interaction and the
species simply slowly leaks out of the probe/interface gap by
hindered diffusion [Figure 1a(i)]. Consequently, when the
potential is reversed, in a final step to collect the species by
electrolysis, the flux of B at the UMEsand the corresponding
current-time characteristicsswill be strongly dependent on the
nature of the interaction of B with the target interface.
For the particular case of an irreversible process, illustrated
using the case of transfer (or absorption) in parts b(i) and b(ii)
of Figure 1, it may be anticipated that there will be a decrease
in the concentration of B in the probe/interface gap which, in
turn, will result in a diminution of the UME current during the
reverse potential step. After the mass transport problem is
solved for this situation, the reverse current-time characteristics
should thus provide quantitative information on the kinetics of
the local interfacial process.
The quantitative application of SECM requires that the
distance between the tip and target interface is known.1 An
attractive feature of the DPSC mode is that when the species
of interest is electrogenerated via the diffusion-controlled
electrolysis of a precursor in solution, which is inert with respect
to the target interface, the forward current-time behavior
provides unambiguous independent information on the probe/
interface separation,18,19,21 enabling the quantitative interpretation
of the reverse current-time characteristics.
τ ) tDA/a2
(3)
where DA is the diffusion coefficient of A. The parameter
λi ) Di/DA (i ) A or B)
(4)
reflects the fact that A and B may have different diffusion
coefficients. The concentrations of A and B have been
normalized with respect to the bulk concentration of A, [A]∞,
such that
CA ) [A]/[Α]∞
CB ) [B]/x[A]∞
(5)
(6)
The following boundary conditions are true following the initial
potential step at τ ) 0 and prior to the reverse step at τ ) τswitch
:
0 e τ < τswitch
,
Z ) 0, 0 e R e 1:
CA ) 0, -λB(∂CB/∂Z) ) ∂CA/∂Z (7)
In this paper, a numerical model for SECM DPSC is
developed using the ADIFDM. Specific consideration is given
to the case where the species of interest, electrogenerated at
the UME probe through the diffusion-controlled electrolysis of
0 e τ < τswitch
,
Z ) 0, 1 < R e RG:
∂CA/∂Z ) 0, ∂CB/∂Z ) 0 (8)

Measurement of Local Reactivity
J. Phys. Chem. B, Vol. 101, No. 50, 1997 10853
i/i(∞) ) -λ (π/2) 1(∂C /∂Z)Z)0R dR
(22)
0 e τ < τswitch
0 e τ < τswitch
0 e τ < τswitch
In eqs 8-11
,
,
,
R ) 0, 0 < Z < L:
∂CA/∂R ) 0, ∂CB/∂R ) 0 (9)
∫
B
B
0
The problem was solved numerically using a FORTRAN
program based on the ADIFDM algorithm.28 The interface
boundary conditions defined by eqs 11 and 19 are readily
incorporated in the general algorithm,21 with the consideration
of double potential steps requiring only minor modifications to
earlier single potential step programs.14a,15,20-23
R > RG, 0 < Z < L:
CA ) 1, CB ) 0 (10)
Z ) L, 0 < R e RG:
∂CA/∂Z ) 0, ∂CB/∂Z ) KCB (11)
Theoretical Results and Discussion
L ) d/a
(12)
(13)
The normalized DPSC response depends on L, K, λB, τswitch
,
and, if long-time behavior is considered, RG. The aim here is
to consider the general DPSC characteristics, and so we set RG
) 10 and λB ) 1 and investigate the effects of τswitch and K on
the current-time behavior for typical probe-to-interface separa-
tions.
Effect of τswitch. The normalized current-time characteristics,
with K ) 0, for three different switching times, τswitch ) 0.01,
0.1, and 1.0, are shown in Figure 2. For each switching time,
three probe/interface separations, log(L) ) -1.0, -0.5, and 0.0,
have been considered. The characteristics for both the forward
RG ) rs/a
where d is the distance between the tip and the target interface
and rs is the radius of the probe (electrode plus glass insulator).
The boundary conditions have the following meanings.
Equations 7 and 8 denote that the electrolysis of A to B at the
UME occurs at a diffusion-controlled rate, but both species are
inert on the insulating sheath surrounding the electrode. Equa-
tion 9 is a consequence of the axisymmetric cylindrical
geometry, while eq 10 assumes that both species recover their
bulk solution concentrations outside the thin layer zone formed
by the probe and interface. This is a reasonable assumption
for the probes used practically in most SECM studies.1,15,17,21-23
Equation 11 reflects the activity of A and B at the interface
of interest. Species A is inert on this boundary, while B is lost
in a first-order irreversible interfacial process, characterized by
a normalized rate constant
and reverse steps are plotted together as |i/i(∞)| versus either
-1/2
τ-1/2 or (τ - τswitch
)
in order to emphasize the short-time
characteristics for the forward and reverse steps, respectively.
The form of the forward transient, when the probe is close
to an inert interface, has been discussed previously in connection
with single potential step chronoamperometry.18,19 In brief, at
sufficiently short times such that the diffusion field at the UME
is very small compared to the probe/interface separation, the
forward current-time response is identical to that for a
conventional UME in a solution. However, at times sufficient
for the diffusion field to intercept the interface [τ ≈ L2 ],
diffusion to and from the UME becomes hindered with the result
that the current that flows is lower than that, at a corresponding
normalized time, with the probe in free solution. The forward
current-time behavior in Figure 2a demonstrates that, at the
probe/interface distances considered for τswitch ) 0.01, hindered
diffusion is only observed at the closest distance of log(L) )
-1.0.
K ) ka/DB
(14)
where k (cm s-1) is a first-order heterogeneous rate constant.
For the period of the reverse step, in which B is electrolyzed
to A at the UME at a diffusion-controlled rate, it is only
necessary to evaluate the concentration profile for B by solving
eq 2 with i ) B subject to the following boundary conditions:
τ
switch e τ, Z ) 0, 0 e R e 1: CB ) 0
(15)
τ
switch e τ, Z ) 0, 1 < R e RG: ∂CB/∂Z ) 0 (16)
A comparison of the reverse transients, for the three distances
with τswitch ) 0.01, clearly shows that the effect of hindered
diffusion of B away from the UME (for log (L) ) -1.0) is to
increase the short-time current on the reverse transient. In
contrast, for log(L) ) -0.5 and 0.0, there is a strong concentra-
tion gradient of B driving this species away from the UME,
even during the reverse step, causing a smaller current to flow
at short times.
τ
switch e τ, R ) 0, 0 < Z < L: ∂CB/∂R ) 0 (17)
τ
switch e τ, R > RG, 0 < Z < L: CB ) 0 (18)
τ
switch e τ, Z ) L, 0 < R e RG: ∂CB/∂Z ) KCB (19)
The limiting case of an inert interface can be simulated with K
) 0 in eqs 11 and 19.
This trend is further illustrated in parts b and c of Figure 2.
Together with the data in Figure 2a, it is clear that the greater
the extent to which diffusion is hindered during DPSC measure-
ments, i.e., the closer the tip/interface separation and the longer
the switching time, the closer are the forms of the forward and
reverse current transients. Essentially, under conditions where
diffusion in and out of the probe/interface gap is strongly
impeded, there is a tendency for almost all of the material
generated in the forward step to be collected in the reverse step.
The effect described above is evident in parts a-c of Figure
3, which show the concentration profiles for species B for the
three distances of interest at three times just after a reverse
potential step at τswitch ) 0.1. For log(L) ) -1.0, it can be
seen that at the first time of interest [Figure 3a(i)], the gap
between the probe and interface, directly under the active
electrode, is filled with species B, which may be rapidly
recollected. Consequently, the reverse current is initially high
The aim of the model is to provide a solution for the current-
time behavior during the forward and reverse potential steps.
For the forward step, the current i, normalized with respect to
i(∞), the steady-state current for the diffusion-controlled elec-
trolysis of A with the tip at an effectively infinite distance from
the interface, is given by
i/i(∞) ) (π/2) 1(∂C /∂Z)Z)0R dR
(20)
∫
A
0
where27
i(∞) ) 4nFDAa[A]∞
(21)
In eq 21, F is Faraday’s constant. For the period of the reverse
step the normalized current ratio is given by

10854 J. Phys. Chem. B, Vol. 101, No. 50, 1997
Slevin et al.
readily be seen by comparing, for example, parts a(ii) and a(iii)
of Figure 3 with parts b(ii) and b(iii) of Figure 3 or by examining
the long-time reverse currents in parts b and c of Figure 2.
In terms of the application of DPSC to the measurement of
interfacial kinetics, it follows that the technique will be most
sensitive to interfacial processes at the closest probe/interface
separations, as found for other SECM modes.1,3,7,9,10,13-15
Effect of K. Having identified that SECM DPSC will be
most sensitive to interfacial kinetics at the closest probe/substrate
separations, we consider the effect of interfacial kinetics on the
current-time response for log(L) ) -0.8, typical of a close
distance that is easily attainable practically in SECM. Figure
4 shows a set of transients, plotted as |i/i(∞)| versus τ-1/2
-1/2
(forward step) or (τ - τswitch
)
(reverse step) for τswitch )
1.0, with K ) 0, 0.1, 1, and 10. The forward transient is the
same for all of the kinetic cases, since the current response is
simply governed by the diffusion of the precursor species to
the UME. In contrast, the form of the reverse transient is seen
to be extremely sensitive to the value of K, with the current
increasingly diminished at all times as K increases.
The reason for the smaller currents observed during the
reverse potential step with K > 0 is clear from Figure 5, which
compares the evolving concentration profiles for species B at
several times during the forward step, up to and including the
switching time, for K ) 0 [Figure 5a] and K ) 1 [Figure 5b].
When species B is inert with respect to the target interface, the
forward step serves to fill the gap between the probe and the
interface (particularly in the region 0 < R < 1) with B [parts
a(i) and a(ii) of Figure 5], which is subsequently collected during
the reverse step [Figure 5a(iii)]. In contrast, when the interface
acts as a partial (irreversible) sink for B, the concentration of
this species is greatly depleted during the period of the forward
[parts b(i) and b(ii) of Figure 5] and reverse steps [Figure
5b(iii)], resulting in a smaller current flow when the potential
is reversed.
As with previous kinetic applications of the SECM, the time
scale of a measurement can be tuned to the region of interest
by varying the size of the probe [eq 14] and the probe/interface
separation. For the DPSC mode, the switching time may also
be used with good effect to enable the most sensitive kinetic
measurements. This point is illustrated in Figure 6, which shows
the current-time curves plotted as normalized current versus
-1/2
τ-1/2 (forward step) or (τ - τswitch
)
(reverse step) for τswitch
) 0.1 and 10 for various values of K. In conjunction with
Figure 4, these data demonstrate that the DPSC technique can
be made most sensitive to fast kinetics by employing a short
switching time, while slower interfacial reactions can be studied
most effectively by employing a longer switching time.
Figure 2. Forward (- - -) and reverse (-) current-time characteristics
plotted as |i/i(∞)| versus τ-1/2 (forward step) and (τ - τswitch
)
-1/2 (reverse
step) for τswitch ) 0.01 (a), 0.1 (b), and 1.0 (c). For each switching
time, data are shown for log(L) ) -1.0 (lowest current at longest times
on the forward step; highest current at shortest times on the reverse
step), -0.5, and 0.0 (highest current at longest times on the forward
step; lowest current at shortest times on the reverse step). In (a) only
two pairs of transients appear, since the characteristics for log(L) )
-0.5 and 0.0 are essentially identical on this switching time scale.
Experimental Section
Materials and Solutions. All aqueous solutions were
prepared from Milli-Q reagent water (Millipore Corp.). Fer-
rocyanide solutions contained 0.050 mol dm-3 potassium
ferrocyanide with 2.0 mol dm-3 potassium chloride (both
Analytical Reagent, Fisher, Loughborough, U.K.) as background
electrolyte. Bromide solutions contained 0.010 mol dm-3
potassium bromide (Analytical Reagent, Fisher) and 0.50 mol
dm-3 sulfuric acid (Analytical Reagent, Sigma-Aldrich, Gill-
ingham, U.K.) which served as background electrolyte. Aque-
ous solutions of bromine were prepared from 99.99% reagent
(Aldrich), with 0.50 mol dm-3 sulfuric acid as an electrolyte.
For immiscible liquid/liquid studies, DCE (HPLC grade, Sigma-
Aldrich) was employed as the organic phase.
[Figure 2b], but then rapidly decreases as B is consumed at the
UME [parts a(ii) and a(iii) of Figure 3].
For the larger probe/interface separations considered, there
are concentration gradients both normal and radial to the UME
that promote the diffusion of B away from the electrode during
the initial collection process [parts b(i) and c(i) of Figure 3].
The diffusion process away from the electrode is greatest, and
thus the initial reverse diffusion-limited current lowest, at the
largest probe/interface separation. However, at these larger
probe/substrate separations, a greater fraction of B is able to
survive to longer times, causing a higher current than at
equivalent times for close tip/substrate separations. This can
Apparatus and Instrumentation. Electrochemical measure-
ments were made using a two-electrode arrangement. A silver

Measurement of Local Reactivity
J. Phys. Chem. B, Vol. 101, No. 50, 1997 10855
Figure 3. Distribution of species B in the tip/substrate gap during the reverse potential step, for τswitch ) 0.1 with K ) 0. Data are shown for log(L)
) -1.0 (a), -0.5 (b), and 0.0 (c) at τ ) 0.10 (i), 0.11 (ii), and 0.15 (iii).
Figure 4. Forward (- - -) and reverse (-) current-time characteristics
plotted as |i/i(∞)| versus τ-1/2 (forward step) and (τ - τswitch
)
-1/2 (reverse
step) for τswitch ) 1.0 and log(L) ) -0.8. The reverse characteristics
are for K ) 0 (upper solid line), 0.1, 1, and 10 (lower solid line).
wire served as a quasi-reference electrode (AgQRE), and the
working electrode was a 25 µm diameter Pt disk UME. Two
designs of UME were employed: a conventional UME, which
has been described previously,15,29 and a new design of inverted
“submarine” UME,13,26 shown in Figure 7, to enable the study
of air/liquid interfaces by a UME positioned in the liquid phase.
The submarine UME was fabricated in a way similar to the
fabrication of a conventional UME, but the electrode was cut
short, ca. 15 mm, so that it could be accommodated in the cell,
and electrical connection was made with a flexible silicone
rubber-coated, stranded copper wire. The wire-electrode
connection was sealed with epoxy resin (Araldite Rapid, Ciba-
Geigy) such that no connecting wire was exposed to the solution.
The electrode was mounted vertically, but inverted, in the cell
by inserting it into a small PTFE block through which two
parallel holes had been machined. This arrangement was then
Figure 5. Evolution of the concentration distribution of species B in
the tip/substrate gap during the forward potential step, up to and
including the reverse step (τswitch ) 1.0), for K ) 0 (a) and 1 (b). In
each case profiles are shown for τ ) 0.01 (i), 0.1 (ii), and 1.0 (iii)
(immediately after the potential was reversed).
connected to a conventional electrode holder, itself attached to
a micropositioner, via a glass capillary. The latter was inserted
through the second hole in the PTFE block. The submarine
electrode has been shown to function well in terms of current-
distance approach measurements, as described recently.13,26 Both

10856 J. Phys. Chem. B, Vol. 101, No. 50, 1997
Slevin et al.
data acquisition card (Lab PC card, National Instruments, Austin,
TX) was used to record current-potential and current-time
behavior. Some current-time transients were also measured
using a NIC310 digital storage oscilloscope (Nicolet, Coventry,
U.K.).
UMEs were positioned initially with a 431-2 stage (Newport
Corp., Fountain Valley, CA). Fine control in the z-direction
normal to the plane of the interface, with a spatial resolution of
around 0.1 µm over a 40 µm range, was achieved with a
piezoelectric translator, incorporating a strain gauge sensor
(translator Model P173 and controller P273, Physik Instrumente,
Waldbronn, Germany).
For experiments where the liquid/liquid or liquid/air interfaces
were probed, a one-piece cylindrical glass cell (40 mm diameter
and 50 mm depth) was used. The cell incorporated an optical
window on one side such that a zoom microscope fitted with a
CCD camera (maximum resolution 2.2 µm per screen pixel)
could be used to aid positioning of the UME near the interface.
For experiments concerned with confirming the theoretical
model with an inert glass disk substrate, a fully detachable
Plexiglas cell was used, comprising a body, lid, and base into
which the substrate could be securely fixed with its surface
perpendicular to the axis of the UME.
Experimental Results and Discussion
DPSC Characteristics at Inert Interfaces. In order to verify
the theoretical predictions, initial experiments were carried out
on the diffusion-controlled generation/collection of ferricyanide
ions from 50 × 10-3 mol dm-3 potassium ferrocyanide solutions
at a probe UME positioned close to (i) an aqueous/glass
interface, (ii) an aqueous/DCE interface, and (iii) an aqueous/
air interface. All of these are examples of interfaces that are
chemically inert with respect to the ferrocyanide/ferricyanide
couple. For these measurements, the tip potential was stepped
initially from 0.10 V, where no Faradaic processes occurred,
to 0.70 V, where the oxidation of ferrocyanide was diffusion-
controlled for a period tswitch typically of 200 ms. The potential
was then stepped back to 0.10 V to collect ferricyanide by
reduction to ferrocyanide.
Figure 6. Forward (- - -) and reverse (-) current-time characteristics,
plotted as |i/i(∞)| versus τ-1/2 (forward step) and (τ - τswitch
)
-1/2 (reverse
step) for log(L) ) -0.8: (a) τswitch ) 0.1 and K ) 0 (upper solid curve),
0.1, 1, 10, and 100 (lower solid curve); (b) τswitch ) 10 and K ) 0
(upper solid curve), 0.1, and 1 (lower solid curve).
Figure 8 shows a series of typical generation/collection
transients with the UME probe at various distances from the
aqueous/glass interface. In light of previous work, which has
demonstrated that the chronoamperometric oxidation of ferro-
cyanide under these conditions provides an excellent measure
of the probe/interface separation,15c the only variable involved
in fitting the forward transient data was the probe/interface
distance. A value of DFe(CN) ) 6.7 × 10-6 cm2 s-1, deduced
4-
6
from the steady-state diffusion-limited current for the oxidation
of ferrocyanide [eq 21], was employed to enable a comparison
to be made between experiment and theory for the forward step,
which was found to be in excellent agreement over the range
of distances investigated.
With the tip/interface separation fixed, the only variable
involved in the interpretation of the reverse transient behavior
3-
for this system is λFe(CN) . Experiment and theory were found
6
3-
to be in excellent agreement with λFe(CN) ) 1.13, which is
6
Figure 7. Schematic (not to scale) of the inverted submarine UME
design for investigating the interface between a liquid phase containing
the UME probe and a less dense immiscible liquid or gas phase.
the value expected, given the diffusion coefficients of ferro-
cyanide and ferricyanide under these conditions.30 The data in
Figure 8 both verify the numerical model for SECM DPSC and
demonstrate that measurements can readily be made on time
scales less than 200 µs with the experimental arrangement
described.
types of UME were characterized by a ratio of the overall probe
diameter to electrode diameter, RG ≈ 10.
Potentials were controlled with a triangular wave/pulse
generator (Colburn Electronics, Coventry, U.K.), and currents
were measured using a home-built current follower (gains of
10-5-10-9 A V-1). A Pentium 100 MHz PC equipped with a
Further to recent work, which has demonstrated that the
feedback10 and equilibrium perturbation13 modes can be used
to probe chemical processes at immiscible liquid/liquid inter-

Measurement of Local Reactivity
J. Phys. Chem. B, Vol. 101, No. 50, 1997 10857
Figure 9. Typical DPSC data (tswitch ) 200 ms) for the oxidation of
ferrocyanide (forward step; upper curve) and reduction of ferricyanide
(reverse step; lower curve) at a tip in aqueous solution positioned at a
distance of 3.4 µm from an aqueous/DCE interface. The solid lines
represent the experimental data, while the dashed lines correspond to
the theoretical characteristics. A value of DFe(CN) ) 6.7 × 10-6 cm2
4-
6
s-1 was employed to convert normalized time to real time, and λFe(CN)
) 1.13 was used for the simulation.
3-
6
Figure 10. Typical DSPC data (tswitch ) 200 ms) for the oxidation of
ferrocyanide (forward step; upper curve) and reduction of ferricyanide
(reverse step; lower curve) at a tip in aqueous solution, positioned at
a distance of 2.9 µm from an aqueous/air interface. The solid lines
represent the experimental data, while the dashed lines correspond to
the theoretical characteristics. A value of DFe(CN) ) 6.7 × 10-6 cm2
4-
6
s-1 was employed to convert normalized time to real time, and λFe(CN)
) 1.13 was used for the simulation.
3-
6
Figure 8. DPSC data for the oxidation of ferrocyanide (forward step;
upper curves) and reduction of ferricyanide (reverse step; lower curves)
at a tip in aqueous solution positioned at distances of 1.5 (a), 3.0 (b),
and 4.5 (c) µm from a glass surface. In each case the switching time
faces were viable. DPSC characteristics were obtained on
similar time scales, and the same range of tip/interface separa-
tions, as in the experiments described above. Experimental data
were found to be in good agreement with theory in all cases,
demonstrating that SECM can indeed be used to probe liquid/
gas interfaces. Typical results are shown in Figure 10 for a
tip/interface separation of 2.9 µm. It can be seen that the
experimental data are of a high quality and consistent with the
theoretical predictions.
was 200 ms. The solid lines are experimental data, while the dashed
4-
lines are the theoretical characteristics. A value of DFe(CN) ) 6.7 ×
10-6 cm2 s-1 was employed to convert normalized time t6o real time,
3-
while λFe(CN) ) 1.13 gave the best fit to the reverse current-time
6
behavior in each case.
faces, the experimental characteristics for the generation/
collection of ferricyanide in the DPSC mode at the aqueous/
DCE interface were examined. By use of the same aqueous
solution conditions outlined above (similar time scales and tip/
substrate separations between 1.5 and 9.0 µm), excellent
DPSC Characteristics at Active Liquid/Liquid Interfaces.
Br2 shows enhanced solubility in many organic solvents
compared to aqueous solutions.31 The partition coefficient of
Br2 in the DCE/0.5 mol dm-3 sulfuric acid system was
determined voltammetrically by first measuring the steady-state
limiting current for the reduction of 10 mmol dm-3 Br2 in 0.5
mol dm-3 sulfuric acid solution at a 25 µm diameter Pt UME
(iL1) and the corresponding limiting current after shaking the
solution with an equal volume of DCE (iL2). Under these
conditions, the coefficient defining the partitioning of Br2
between DCE and aqueous solution is then given by
agreement was found between experiment and theory.
A
representative case is illustrated in Figure 9, which shows the
DPSC characteristics for a switching time of 200 ms, with the
tip positioned 3.4 µm from the aqueous/DCE interface.
By use of the probe design shown in Figure 7, DPSC
experiments were carried out close to the aqueous/air interface
to determine whether SECM measurements at liquid/gas inter-

10858 J. Phys. Chem. B, Vol. 101, No. 50, 1997
[Br2]DCE iL1 - iL2
Slevin et al.
Kp )
)
(23)
iL2
[Br2]aq
A value of Kp ) 37((1) was deduced. This implies that Br2
electrogenerated through the oxidation of Br- in the aqueous
phase, close to an aqueous/DCE interface, will undergo ef-
fectively irreversible transfer across the interface, providing an
ideal model system for examining the DPSC technique as a
probe of irreversible interfacial kinetics.
In these experiments, Br- was first oxidised to Br2 in an initial
potential step, from 0.70 to 1.20 V, for a period in the range
10-200 ms, after which the potential was stepped to 0.70 V to
effect the diffusion-controlled collection of Br2 by reduction to
Br-. All experimental data obtained at tip/interface separations
from 2.5 to 9.1 µm, for the step times defined above, were found
to be in good agreement with the theoretical predictions for the
irreversible diffusion-controlled loss of Br2 at the interface. For
Figure 11. Typical DPSC data (tswitch ) 10 ms) for the oxidation of
bromide (forward step; upper solid curve) and reduction of bromine
(reverse step; lower solid curve) at a tip in aqueous solution positioned
at a distance of 2.8 µm from an aqueous/DCE interface. The upper
dashed line is the theoretical response for the forward step at the defined
-
the fitting of the experimental data, a value of DBr ) 1.80 ×
10-5 cm2 s-1 was employed, as deduced from steady-state UME
tip/interface separation, with DBr ) 1.8 × 10-5 cm2 s-1. The remaining
-
voltammetry [eq 21], together with DBr ) 9.4 × 10-6 cm2 s-1
,
2
dashed lines are the reverse transients for various transfer rate constants
measured using UME chronoamperometry.32 Both of these
values are in good agreement with those reported earlier under
similar solution conditions.33 It should be noted that the rapid
transfer of Br2 across the interface (and thus out of the tip/
interface gap) results in a dramatic decrease in the concentration
of Br2 in the aqueous phase. The formation of Br3-, through
the equilibration of Br2 and Br-,34 can thus be neglected so
that the basic theoretical model described herein is valid.
k with the values marked on the plot. A value of DBr ) 9.4 × 10-6
2
cm2 s-1 was used for the simulation.
A typical result for the shortest time step examined (tswitch
)
10 ms), at a tip/interface separation of 2.8 µm, is shown in
Figure 11. The data have been plotted as |i/i(∞)| vs t-1/2
-1/2
(forward step) or (t - tswitch
)
(reverse step). The reverse
current-transient is in good agreement with theory for a
diffusion-controlled process, particularly for (t - tswitch
-1/2
)
>
20 s-1/2. At longer reverse times, the experimental currents are
slightly higher than those predicted theoretically. This may be
attributed to a small effect of electrochemically induced back-
transfer of Br2 from DCE to the aqueous phase due to the
extensive depletion of Br2 in the aqueous phase during the
collection process. Nonetheless, the transfer of Br2 from the
aqueous phase to DCE, under the conditions of these experi-
ments, tends toward irreversibility and is effectively rate-limited
by diffusion of Br2 in the aqueous phase. Based on the data in
Figure 11, a lower limit of 0.5 cm s-1 can be assigned to the
interfacial rate constant for the transfer of Br2 from the aqueous
to the DCE phase under conditions where the latter phase is an
empty sink.
DPSC Characteristics at Active Liquid/Air Interfaces.
The volatility of Br2 in aqueous solutions in contact with air
under open system conditions makes the process described
abovesin which bromine is electrogenerated close to an
interfacesan ideal one to investigate the use of SECM as a
probe of dynamic processes at liquid/gas interfaces. Experi-
ments were carried out using a protocol identical to that
described in the preceding section, with an inverted submarine
UME positioned at distances of 2.5-9.1 µm from an aqueous/
air interface. All experimental data for switching times of 10
ms and longer were found to agree well with the theoretical
predictions for the irreversible diffusion-controlled loss of Br2
at the aqueous/air interface using the diffusion coefficients
defined above.
Figure 12. Typical DPSC data (tswitch ) 10 ms) for the oxidation of
bromide (forward step; upper solid curve) and reduction of bromine
(reverse step; lower solid curve) at a submarine UME in aqueous
solution positioned at a distance of 2.5 µm from an aqueous/air interface.
The upper dashed line is the theoretical response for the forward step
at the defined tip/interface separation, with DBr ) 1.8 × 10-5 cm2
-
s-1. The remaining dashed lines are the reverse transients for various
transfer rate constants k with the values marked on the plot.
transient in Figure 12 lies slightly above the theory for a
diffusion-controlled process (particularly at short times), which
might indicate very slight interfacial control of the transfer
process. However, we were not able to go to sufficiently shorter
times or closer tip/interface separations to determine whether
there was a measurable interfacial resistance to the transfer
process. From the data in Figure 12, it may therefore be
deduced that the transfer of Br2 from aqueous solution to air is
a rapid process, which approaches a limiting rate controlled by
diffusion of Br2 in solution. The data allow a lower limit of
0.5 cm s-1 to be assigned to the rate constant defining the
interfacial transfer process.
Note that, in contrast to the reverse transient data for Br2
transfer from aqueous solution to DCE (Figure 11), the reverse
currents at longer times (t-1/2 < 20 s-1/2) in Figure 12 are closer
to the diffusion-controlled theoretical case for an irreversible
transfer process. This suggests that electrochemically induced
back-transfer of Br2 from air to aqueous solution, during the
collection process, is a negligible process. In part, this may be
attributed to the much higher transport rate of Br2 in air
A typical result for the shortest time step examined (tswitch
)
10 ms), at a tip/interface separation of 2.5 µm, is shown in
Figure 12. The data have been plotted as |i/i(∞)| vs t-1/2
-1/2
(forward step) or (t - tswitch
)
(reverse step). The reverse

Measurement of Local Reactivity
J. Phys. Chem. B, Vol. 101, No. 50, 1997 10859
(5) Wipf, D. O.; Bard, A. J. J. Electrochem. Soc. 1991, 138, L4.
(6) Mirkin, M. V.; Bulhoes, L. O. S.; Bard, A. J. J. Am. Chem. Soc.
1993, 115, 201.
(7) Pierce, D. T.; Unwin, P. R.; Bard, A. J. Anal. Chem. 1992, 64,
1795.
compared to DCE; i.e., once Br2 crosses the interface, it is
rapidly transported away and is thus not available for back-
transfer during the collection step.
(8) Pierce, D. T.; Bard, A. J. Anal. Chem. 1993, 65, 3598.
(9) (a) Mandler, D.; Bard, A. J. J. Electrochem. Soc. 1989, 136, 3143.
(b) Macpherson, J. V.; Slevin, C. J.; Unwin, P. R. J. Chem. Soc., Faraday
Trans. 1996, 92, 3799.
Conclusions
SECM DPSC has been shown to be a powerful technique
for investigating the chemical kinetics of tip-generated species
at liquid/solid, liquid/liquid, and liquid/air interfaces. The
technique complements earlier equilibrium perturbation transient
SECM methods by allowing the study of irreversible interfacial
chemical processes. The DPSC mode has been verified
experimentally with the model ferrocyanide/ferricyanide system
and has been used to show that the absorption of Br2 by both
air and DCE from aqueous solutions (under sink conditions) is
controlled by diffusion of Br2 in the aqueous phase, with
interfacial processes providing no detectable kinetic resistance
to transfer with the range of mass transfer rates attainable from
SECM.
(10) (a) Wei, C.; Bard, A. J.; Mirkin, M. V. J. Phys. Chem. 1995, 99,
16033. (b) Solomon, T.; Bard, A. J. J. Phys. Chem. 1995, 99, 17487. (c)
Tsionsky, M.; Bard, A. J.; Mirkin, M. V. J. Phys. Chem. 1996, 100, 17881.
(d) Shao, Y.; Mirkin, M. V.; Rusling, J. F. J. Phys. Chem. 1997, 101, 3202.
(11) (a) Kwak, J.; Anson, F. C. Anal. Chem. 1992, 64, 250. (b) Denuault,
G.; Troise Frank, M. H.; Peter, L. M. Faraday Discuss., 1992, 94, 23.
(12) (a) Scott, E. R.; White, H. S.; Phipps, J. B. J. Membr. Sci. 1991,
58, 71. (b) Scott, E. R.; White, H. S.; Phipps, J. B. Anal. Chem. 1993, 65,
1537. (c) Scott, E. R.; White, H. S.; Phipps, J. B. Pharm. Res. 1993, 10,
1699. (d) Scott, E. R.; White, H. S.; Phipps, J. B. J. InVest. Dermatol. 1995,
104, 142. (e) Macpherson, J. V.; Beeston, M. A.; Unwin, P. R.; Hughes,
N. P.; Littlewood, D. J. Chem. Soc., Faraday Trans. 1995, 91, 1407. (f)
Macpherson, J. V.; Beeston, M. A.; Unwin, P. R.; Hughes, N. P.; Littlewood,
D. Langmuir 1995, 11, 3959.
(13) Slevin, C. J.; Umbers, J. A.; Atherton, J. H.; Unwin, P. R. J. Chem.
Soc., Faraday Trans. 1996, 92, 5177.
There is considerable scope for further developing the SECM
DPSC mode. For example, to complement the systems in this
paper, we are currently investigating reversible interfacial
processes of tip-generated species using the DPSC mode. In
addition to generating reactants from solution precursor species,
as in the studies in this paper, it should be possible to locally
(14) (a) Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1992, 96, 5035. (b)
Yang, Y. F.; Denuault, G. J. Electroanal. Chem. 1996, 418, 99.
(15) (a) Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1994, 98, 1704.
(b) Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1994, 98, 3109. (c)
Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1995, 99, 3338. (d)
Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1995, 99, 14824. (e)
Macpherson, J. V.; Unwin, P. R. J. Phys. Chem. 1996, 100, 19475. (f)
Macpherson, J. V.; Unwin, P. R.; Hillier, A. C.; Bard, A. J. J. Am. Chem.
Soc. 1996, 118, 6445.
(16) (a) Mandler, D.; Bard, A. J. J. Electrochem. Soc. 1990, 137, 2468.
(b) Mandler, D.; Bard, A. J. Langmuir 1990, 6, 1489. (c) Mandler, D.;
Bard, A. J. J. Electrochem. Soc. 1990, 137, 1079. (d) Meltzer, S.; Mandler,
D. J. Electrochem. Soc. 1995, 142, L82. (e) Wipf, D. O. Colloids Surf. A
1994, 93, 251.
(17) Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1221.
(18) Bard, A. J.; Denuault, G.; Friesner, R. A.; Dornblaser, B. C.;
Tuckerman, L. S. Anal. Chem. 1991, 63, 1282.
(19) Bard, A. J.; Unwin, P. R.; Wipf, D. O.; Zhou, F. Am. Inst. Phys.,
Conf. Proc. 1992, 241, 235.
(20) Martin, R. D.; Unwin, P. R. J. Electroanal. Chem., in press.
(21) Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1991, 95, 7814.
(22) Zhou, F.; Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1992, 96, 4917.
(23) Demaille, C.; Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1996, 100,
14137.
(24) Macpherson, J. V.; Unwin, P. R. Anal. Chem. 1997, 69, 2063.
(25) (a) Andrieux, C. P.; Hapiot, P.; Save´ant, J. M. J. Phys. Chem. 1988,
92, 5992. (b) Andrieux, C. P.; Save´ant, J. M. In InVestigations of Rates
and Mechanisms of Reactions; Bernasconi, C., Ed.; Wiley: New York, 1986;
Vol. 6, 4/E, Part 2, p 305.
inject several types of metal ions, e.g., Cu2+, Ag+, Zn2+, Pb2+
,
Cd2+, etc., into the gap between the probe and the interface
through the anodic dissolution of metal or mercury amalgam
UMEs. We plan to use this approach as a means of studying
the adsorption and absorption of metal ions at solid/liquid and
immiscible liquid/liquid interfaces.
SECM has been widely used to probe physicochemical
processes at liquid/solid interfaces1 and is proving to be a
valuable tool for investigating charge and molecular transfer
processes at immiscible liquid/liquid interfaces.10,13 The studies
herein build further on recent work26 by demonstrating that
SECM can be used to study liquid/gas interfaces through the
application of a simple submarine UME design. We anticipate
that this is an area with considerable promise for further
development. In particular, the use of SECM to noninvasively
probe chemical and diffusion processes in Langmuir monolayers
at the air/water interface is readily envisaged.
(26) The first studies of liquid/gas interfaces by SECM were reported
in the following. Unwin, P. R.; Slevin, C. J. Abstr.sElectrochem. Soc. 1996,
96-2, 810.
(27) Saito, Y. ReV. Polarogr. 1968, 15, 177.
(28) Peaceman, D. W.; Rachford, H. H. J. Soc. Ind. Appl. Math. 1955,
3, 28.
Acknowledgment. We thank the EPSRC, the Royal Society
Paul Instrument Fund, and Zeneca for support of this work.
Helpful discussions with Dr. John Atherton and John Umbers
(Zeneca Huddersfield Works) are much appreciated.
(29) Bard, A. J.; Fan, F.-R. F.; Kwak, J.; Lev, O. Anal. Chem. 1989,
61, 132.
(30) (a) Adams, R. N. Electrochemistry at Solid Electrodes; Marcel
Dekker: New York, 1969; p 219. (b) Martin, R. D.; Unwin, P. R. Submitted
for publication.
(31) Seidell, A.; Linke, W. F. Solubilities: Inorganic and Metal-Organic
Compounds, 4th ed.; American Chemical Society: Washington, DC, 1958;
Vol. 1.
(32) (a) Winlove, C. P.; Parker, K. H.; Oxenham, R. K. C. J. Electroanal.
Chem. 1984, 170, 293. (b) O’Hare, D.; Winlove, C. P.; Parker, K. H. J.
Biomed. Eng. 1991, 13, 304. (c) Denuault, G.; Mirkin, M. V.; Bard, A. J.
J. Electroanal. Chem. 1991, 86, 27.
References and Notes
(1) For reviews see, for example, the following. (a) Bard, A. J.; Fan,
F.-R.; Pierce, D. T.; Unwin, P. R.; Wipf, D. O.; Zhou. F. Science 1991,
254, 68. (b) Bard, A. J.; Fan, F.-R. Faraday Discuss. 1992, 94, 1. (c) Bard,
A. J.; Mirkin, M. V.; Fan, F.-R. In Electroanalytical Chemistry; Bard, A.
J., Ed.; Marcel Dekker: New York, 1993; Vol. 18, p 243. (d) Arca, M.;
Bard, A. J.; Horrocks, B. R.; Richards, T. C.; Treichel, D. A. Analyst 1994,
119, 719. (e) Unwin, P. R.; Macpherson, J. V. Chem. Ind. 1995, 874. (f)
Mirkin, M. V. Anal. Chem. 1996, 177A.
(33) Compton, R. G.; Stearn, G. M.; Unwin, P. R.; Barwise, A. J. J.
Appl. Electrochem. 1988, 18, 657, and references therein.
(34) Griffith, R. O.; McKeown, A; Winn, A. G. Trans. Faraday Soc.
1932, 28, 101.
(2) Wipf, D. O.; Bard, A. J. J. Electrochem. Soc. 1991, 138, 469.
(3) Bard, A. J.; Mirkin, M. V.; Unwin, P. R.; Wipf, D. O. J. Phys.
Chem. 1992, 96, 1861.
(4) Engstrom, R. C.; Small, B.; Kattan, L. Anal. Chem. 1992, 64, 241.