Interrogation of surfaces for the quantification of adsorbed species on electrodes: Oxygen on gold and platinum in neutral media

Article abstract of DOI:10.1021/ja8050553

We introduce a new in situ electrochemical technique based on the scanning electrochemical microscope (SECM) operating in a transient feedback mode for the detection and direct quantification of adsorbed species on the surface of electrodes. A SECM tip generates a titrant from a reversible redox mediator that reacts chemically with an electrogenerated or chemically adsorbed species at a substrate of about the same size as the tip, which is positioned at a short distance from it (ca.1 μm). The reaction between the titrant and the adsorbate provides a transient positive feedback loop until the adsorbate is consumed completely. The sensing mechanism is provided by the contrast between positive and negative feedback, which allows a direct quantification of the charge neutralized at the substrate. The proposed technique allows quantification of the adsorbed species generated at the substrate at a given potential under open circuit conditions, a feature not attainable with conventional electrochemical methods. Moreover, the feedback mode allows the tip to be both the titrant generator and detector, simplifying notably the experimental setup. The surface interrogation technique we introduce was tested for the quantification of electrogenerated oxides (adsorbed oxygen species) on gold and platinum electrodes at neutral pH in phosphate and TRIS buffers and with two different mediator systems. Good agreement is found with cyclic voltammetry at the substrate and with previous results in the literature, but we also find evidence for the formation of incipient oxides which are not revealed by conventional voltammetry. The mode of operation of the technique is supported by digital simulations, which show good agreement with the experimental results.

Full text of DOI:10.1021/ja8050553

Published on Web 11/21/2008
Interrogation of Surfaces for the Quantification of Adsorbed
Species on Electrodes: Oxygen on Gold and Platinum in
Neutral Media†
Joaqu´ın Rodr´ıguez-Lo´pez, Mario A. Alpuche-Avile´s, and Allen J. Bard*
Center for Electrochemistry, Department of Chemistry and Biochemistry, Center for Nano- and
Molecular Science and Technology, The UniVersity of Texas at Austin, Austin, Texas 78712
Received July 10, 2008; E-mail: ajbard@mail.utexas.edu
Abstract: We introduce a new in situ electrochemical technique based on the scanning electrochemical
microscope (SECM) operating in a transient feedback mode for the detection and direct quantification of
adsorbed species on the surface of electrodes. A SECM tip generates a titrant from a reversible redox
mediator that reacts chemically with an electrogenerated or chemically adsorbed species at a substrate of
about the same size as the tip, which is positioned at a short distance from it (ca.1 µm). The reaction
between the titrant and the adsorbate provides a transient positive feedback loop until the adsorbate is
consumed completely. The sensing mechanism is provided by the contrast between positive and negative
feedback, which allows a direct quantification of the charge neutralized at the substrate. The proposed
technique allows quantification of the adsorbed species generated at the substrate at a given potential
under open circuit conditions, a feature not attainable with conventional electrochemical methods. Moreover,
the feedback mode allows the tip to be both the titrant generator and detector, simplifying notably the
experimental setup. The surface interrogation technique we introduce was tested for the quantification of
electrogenerated oxides (adsorbed oxygen species) on gold and platinum electrodes at neutral pH in
phosphate and TRIS buffers and with two different mediator systems. Good agreement is found with cyclic
voltammetry at the substrate and with previous results in the literature, but we also find evidence for the
formation of “incipient oxides” which are not revealed by conventional voltammetry. The mode of operation
of the technique is supported by digital simulations, which show good agreement with the experimental
results.
Introduction
will have an impact in this area. The operation of this technique
is based on the feedback mode of SECM, which is one of the
The electrochemical investigation of outer-sphere electron-
transfer reactions coupled to complicated homogeneous reactions
has become highly developed over the past decades because of
the utilization of powerful techniques like cyclic voltammetry
(CV), ultramicroelectrode (UME) voltammetry, and digital
simulations.1,2 However, understanding inner-sphere electro-
catalytic reactions is more difficult because of their greater
complexity and the difficulty in obtaining reliable quantitative
data about intermediates and products on electrode surfaces.
While a variety of electrochemical and surface spectroscopic
methods that can be employed with electrodes in solution have
been developed,3 it has been difficult to obtain the needed
versatility and sensitivity to address these kinds of reactions.
We propose here an in situ electrochemical technique based on
the scanning electrochemical microscope (SECM)4 for the
quantification of adsorbed species at electrodes that we hope
first developed5 and perhaps one of the most distinctive of this
electroanalytical tool. This is due to its ability to correlate an
electrochemical signal at the SECM tip to topographical,
chemical, and electrochemical features6 in such distinct systems
as the imaging of interdigitated arrays7 or the components of a
leaf,8 the detection of an immobilized enzyme,9 and the
heterogeneous kinetics of semiconductors.10 The SECM tip is
able to detect changes in the electrochemical signal by the
concurrent production and collection of the species of a mediator
pair (O/R, where only one of the species is present initially in
solution); variations in the diffusive flux of these species caused
by interaction with the substrate allow the current measured at
the SECM tip to respond accordingly. One of the most useful
features of SECM is the ability to calculate the tip current as a
function of different solution and interfacial phenomena, i.e.,
(5) Bard, A. J.; Denuault, G.; Lee, C.; Mandler, D.; Wipf, D. O. Acc.
Chem. Res. 1990, 23, 357–363.
† This paper is no. 61 of the series Scanning Electrochemical Microscopy.
(1) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley: New
York, 2001.
(2) Saveant, J.-M. Elements of Molecular and Biomolecular Electrochem-
istry; Wiley-Interscience: Hoboken, NJ, 2006.
(3) Electrochemical Interfaces. Modern Techniques for In-Situ Interface
Characterization; Abrun˜a, H. D., Ed.; VCH: New York, 1991.
(4) Scanning Electrochemical Microscopy; Bard, A. J., Mirkin, M. V.,
Eds.; Marcel Dekker: New York, 2001.
(6) Bard, A. J.; Fan, F. F.; Pierce, D. T.; Unwin, P. R.; Wipf, D. O.;
Zhou, F. Science 1991, 254 (5028), 68–74.
(7) Wipf, D. O.; Bard, A. J. Anal. Chem. 1992, 64, 1362–1367.
(8) Lee, C.; Kwak, J.; Bard, A. J. Proc. Natl. Acad. Sci. U.S.A. 1990, 87,
1740–1743.
(9) Pierce, D. T.; Bard, A. J. Anal. Chem. 1993, 65, 3598–3604.
(10) Horrocks, B. R.; Mirkin, M. V.; Bard, A. J. J. Phys. Chem. 1994, 98,
9106–9114.
9
10.1021/ja8050553 CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 16985–16995 16985

A R T I C L E S
Rodr´ıguez-Lo´pez et al.
electrochemical kinetics, by digital simulation.4 We can then
consider the mediator pair to be an interrogation agent that
reports to the SECM tip about the state of the surface being
examined.
transient feedback mode. Some of these studies rely on the
purposeful integration of a reactive species into the system;
however, to the best of our knowledge, no SECM approach to
the detection and quantification of transient reaction intermedi-
ates at electrocatalytic and catalytic materials (i.e., potential or
nonpotential dependent) has been reported. This surface inter-
rogation method allows the detection and quantification of only
reactive adsorbed intermediates formed upon operation of the
substrate electrode, independent of their spectroscopic charac-
teristics and their electrochemical reactivity at the same
substrate. We hope this technique will ultimately shed some
light on inner-sphere mechanisms and be of use to a wider
audience of surface chemists and electrochemists. We introduce
the application of the technique with the quantification of gold
and platinum oxides at neutral pH because they represent a
classical example of the stable chemisorption of oxygen and
formation of surface oxides.28
The technique proposed here consists of approaching a SECM
tip to a small substrate, e.g., both 12.5 µm in radius, a, held at
a short distance, d, from each other of ca. 1 µm (L ) d/a e
0.1). The sensing mechanism is based on first bringing the
substrate to a potential to generate an intermediate or product
on the electrode surface and then taking the substrate to open
circuit and allowing the tip-generated member of the redox pair
to react chemically with an adsorbed species at the substrate
electrode. The tip current during this last stage reports the
amount of adsorbate to the SECM tip through a feedback loop.
The chemical aspects of this application, in a sense, are
analogous to modulated beam relaxation spectroscopy used in
vacuum for gas-solid studies of catalysis,11 as a generated
reactive species is allowed to interact with an adsorbate. In
another sense, however, it is a unique technique, as the reactive
species generation and the detection scheme are integrated into
the same electrochemical setup as enabled by the feedback mode
of SECM and correspondingly intended for its use in solution.
The SECM has been used previously for novel studies on a
variety of surfaces and interfaces, such as in the quantification
of intermediates released during an electrochemical reaction,12,13
in the study of the kinetics of adsorption14 and desorption15 at
pH-sensitive surfaces, and for the measurement of the steady-
state kinetics at a catalytic substrate16 or the binding of metal
ions at lipid monolayers;17 some studies have used a reactive
mediator to measure binding kinetics18 at self-assembled
monolayers.
Mode of Operation
The technique takes advantage of the contrast obtained in
the feedback operation mode of SECM. The SECM tip is a
UME disk sealed in an insulating material. When the tip is
biased to electrolyze a mediator at the diffusion-limited rate in
the bulk solution, iT,∞, and approached to a substrate, the tip
current becomes a function of the nature of the substrate and
the tip-substrate distance, d. The distance normalized to the
tip radius, a, or L ) d/a, gives a convenient way to describe
the steady-state current as a function of distance, i.e., an
approach curve. A substantial change in the SECM tip current
is observed when the distance between the tip and substrate is
less than one radius of the tip, L < 1. Such changes in current
are generally described in terms of dimensionless parameters
like IT ) iT/iT,∞, where iT is the actual tip current and iT,∞ is the
current far away from the substrate. At a given value of L, for
positive feedback, IT > 1 and for negative feedback IT < 1,
with intermediate regimes achieved by either changing condi-
tions such as the substrate potential under kinetically limited
conditions, decreasing the substrate size to a diameter of the
order of the tip or during short transients.
Total positive feedback is obtained when a substrate regener-
ates the mediator at a rate limited only by diffusion in the
tip-substrate gap. For example, for reduction at the tip (O + e
f R), the reaction at the substrate electrode would be the
opposite electrochemical reaction (R - e f O). While most
SECM studies are done at steady state, in the experiments
presented here, a transient chemical reaction of a species on
the substrate is used to generate the positive feedback loop.
Figure 1 illustrates the proposed mechanism for the surface
interrogation of a reducible adsorbed species at the substrate.
First, as depicted in Figure 1A, the substrate is pulsed or scanned
to a potential where oxidation occurs and an adsorbed species,
A, is formed; the tip at this time is kept at open circuit. The
solution contains the mediator in its oxidized state, O, which is
stable under these conditions and does not participate in any
reaction. Following this, as shown in Figure 1B, the substrate
electrode is taken to open circuit, and the tip is scanned or pulsed
to reduce O and generate R. After R diffuses across the
tip-substrate gap, it reaches the adsorbed species A and reacts
chemically with it. O is regenerated and A is consumed and
In this work we use transient SECM current measurements,
which have been implemented in SECM before, e.g., chrono-
amperometry19 and cyclic voltammetry.20,21 The study of the
effects of mediator surface diffusion,15,22,23 applications to the
study of lateral charge propagation in polymer films24,25 and
diffusion in monolayers,26 and the detection of reactive species
in living cells27 have been developed through the use of this
(11) Asscher, M.; Somorjai, G. A. Reactive Scattering. In Atomic and
Molecular Beam Methods; Scoles, G., Ed.; Oxford University Press:
New York, 1992; Vol. 2, pp 488-517.
(12) Yang, Y.; Denuault, G. J. Chem. Soc., Faraday Trans. 1996, 92, 3791–
3798.
(13) Sa´nchez-Sa´nchez, C. M.; Rodr´ıguez-Lo´pez, J.; Bard, A. J. Anal. Chem.
2008, 80, 3254–3260.
(14) Unwin, P. R.; Bard, A. J. Anal. Chem. 1992, 64, 113–119.
(15) Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1992, 96, 5035–5045.
(16) Selzer, Y.; Turyan, I.; Mandler, D. J. Phys. Chem. B 1999, 103, 1509–
1517.
(17) Burt, D. P.; Cervera, J.; Mandler, D.; Macpherson, J. V.; Manzanares,
J. A.; Unwin, P. R. Phys. Chem. Chem. Phys. 2005, 7, 2955–2964.
(18) Burshtain, D.; Mandler, D. J. Electroanal. Chem. 2005, 581, 310–
319.
(19) Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1991, 95, 7814–24.
(20) Zoski, C.; Luman, C. R.; Ferna´ndez, J. L.; Bard, A. J. Anal. Chem.
2007, 79, 4957–4966.
(21) Diaz-Ballote, L. F.; Alpuche-Aviles, M.; Wipf, D. O. J. Electroanal.
Chem. 2007, 604, 17–25.
(22) Lie, L. H.; Mirkin, M. V.; Hakkarainen, S.; Houlton, A.; Horrocks,
B. R. J. Electroanal. Chem. 2007, 603, 67–80.
(23) Slevin, C. J.; Unwin, P. R. J. Am. Chem. Soc. 2000, 122, 2597–2602.
(24) Mandler, D.; Unwin, P. R. J. Phys. Chem. B 2003, 107, 407–410.
(25) O’Mullane, A. P.; Macpherson, J. V.; Unwin, P. R.; Cervera-
Montesinos, J.; Manzanares, J. A.; Frehill, F.; Vos, J. G. J. Phys. Chem.
B 2004, 108, 7219–7227.
(26) Zhang, J.; Slevin, C. J.; Morton, C.; Scott, P.; Walton, D. J.; Unwin,
P. R. J. Phys. Chem. B 2001, 105, 11120–11130.
(27) Liu, B.; Rotenberg, S. A.; Mirkin, M. V. Anal. Chem. 2002, 74, 6340–
6348.
(28) Woods, R. Chemisorption at Electrodes. In Electroanalytical Chem-
istry; Bard, A. J., Ed.; Marcel Dekker: New York, 1976; pp 1-
162.
9
16986 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

SECM Quantification of Adsorbed Species on Electrodes
A R T I C L E S
experiment is that the SECM tip acts as both the titrant generator
and detector, greatly simplifying the design of this in situ
technique.
As depicted in Figure 1, the method assumes the use of a
substrate electrode of the same shape and size as that of the
SECM tip. There are a number of reasons for this choice, the
most important being the presence of positive feedback at large
unbiased electrodes29-32 and the need to approach the SECM
tip to very short distances. Positive feedback can occur in
SECM, even if the substrate is under open-circuit conditions,
when the substrate is larger than the UME disk of the SECM
tip.31 Previous SECM surface studies where this situation has
been of experimental relevance16,29 have used the configuration
shown in this work. When the microdisks of the tip and substrate
are of the same size and the substrate is allowed to rest at the
open-circuit potential, it is possible to go from a complete
negative feedback for very large values of the tip’s RG, e.g.,
10 (RG ) rg/a, where rg is the radius of the insulating glass
sheath), to a partial negative feedback for very small values of
RG, e.g., 1.1. It is desirable to attain complete negative feedback,
as this greatly improves the detection limits of the technique.
However, as stated earlier, an adequate positive feedback is not
present unless the distance between tip and substrate is very
small, which is experimentally challenging when RG is large.
A reasonable option is to establish larger values of positive
feedback that greatly enhance the contrast with only partial
negative feedback. In this case, both the tip and the substrate
are the same size and shape, with an RG between 1 and 2. These
tips, with a ) 12.5 µm, can be approached to distances within
1 µm, at which positive feedback with IT ) 6 (or higher at
smaller L-values) is typically attainable. Finally, different
techniques can be used to generate the titrant, R; we have chosen
to use CV, but the use of chronoamperometry is also shown
and will be developed in future research.
Figure 1. Schematics of the proposed mechanism for the surface inter-
rogation. (A) A reactive species is chemically or electrochemically adsorbed
at the substrate upon a potential scan or step while the tip is at open circuit.
(B) The substrate is taken to open circuit, and the tip generates the titrant,
which reacts at the surface of the substrate to support positive feedback at
the same tip. (C) Upon consumption of the adsorbate at the substrate, the
tip experiences negative feedback. (D) Expected current response at the tip
following the events depicted in panels A-C for an arbitrary electrode setup.
Simulation
We performed digital simulations coupling voltammetry to
the SECM geometry described in the previous section under
transient conditions using the finite element method provided
by the Comsol Multiphysics v.3.2 software (COMSOL, Inc.,
Burlington, MA). The simulation space is depicted in Figure 2,
as discussed in the Simulation Model section, with further details
in the Supporting Information. Basically, the problem was
reduced to an axi-symmetric 2-D problem involving two equal
size tips (a ) 12.5 µm and RG ) 1.5): the substrate tip with
the adsorbed layer of interest and the tip, where reduction of
the species O to R occurs. To treat the surface layer of A with
the Multiphysics software, we use a very thin slab subdomain
with a thickness of d′ < 0.01d in which the adsorbed species is
confined. Since the gap between the tips was taken as d ) 1.25
µm (L ) 0.1), the adsorbed layer slab was taken as 10 nm.
While this is thicker than an actual “surface layer” of adsorbate,
the time required for diffusion of reactant, R, to cross this layer
is so short (<1 µs) that this does not affect the results at the
time scales of interest (especially considering any surface
roughness of the substrate in an actual system).
transformed into a final product, P. Ideally, regeneration of O
should only be possible through reaction with A present at the
substrate, and the system would exhibit positive feedback only
for as long as this condition is met. As shown in Figure 1C, the
final state of the system is to show negative feedback, after all
the A is depleted, and O reacts at the collector electrode at the
rate governed by the hindered diffusion into the tip-substrate
gap. Experimentally, this is difficult to achieve, and as discussed
below, one requires a substrate with the same radius as that of
the tip to have an acceptable working system.
Figure 1D shows the expected voltammetric response for this
system (exaggerated for clarity) and compares it to the case of
positive and negative feedback. The tip potential is scanned
toward more negative potentials, and thus it is generating R,
which diffuses to the substrate and reacts at its surface to
regenerate O. Positive feedback persists as long as there is
something to react at the substrate, and once the adsorbed species
is consumed, the positive feedback drops and transforms into
negative feedback. Notice that an important peculiarity in this
Figure 3 shows representative results for the simulations at
different values of the kinetic constant, k (for the reaction of
(29) Macpherson, J. V.; Slevin, C. J.; Unwin, P. R. J. Chem. Soc., Faraday
Trans. 1996, 92, 3799–3805.
(30) Wipf, D. O.; Bard, A. J. J. Electrochem. Soc. 1991, 138, 469–474.
(31) Xiong, H.; Guo, J.; Amemiya, S. Anal. Chem. 2007, 79, 2735–2744.
(32) Zoski, C.; Simjee, N.; Guenat, O.; Koudelka-Hep, M. Anal. Chem.
2004, 76, 62–72.
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16987

A R T I C L E S
Rodr´ıguez-Lo´pez et al.
in a narrow potential region preceding the E0 of the mediator
and also exhibits very clear and distinguishable peaks (i.e., a
larger contrast between positive and negative feedback). The
advantage of the SECM technique compared to conventional
CV is that it uncouples the generation of surface species from
their detection and quantification. This allows one to detect
species that would be difficult to see by CV. While the
measurements described here were all made by scanning the
tip through the entire mediator wave, it would also be possible
to scan only through a portion of the wave and not attain a
diffusion-limited plateau. This would be useful, for example,
with a mediator whose reduction (or oxidation) occurs very close
to the solvent background. Figure S1 in the Supporting
Information shows some representative results of the simulations
upon variation of the scan rate and through the use of chrono-
amperometry.
Experimental Section
All solutions were prepared with deionized Milli-Q water.
Sulfuric acid (94-98%, Trace Metal grade), potassium nitrate,
potassium chloride, zinc metal, tris(hydroxymethyl)aminomethane
(TRIS, Molecular Biology grade), agar (Purified grade) and
o-phosphoric acid (85%) from Fisher Scientific (Fair Lawn, NJ);
silver nitrate (99+%,) silver wire (99.9%), and methyl viologen
dichloride hydrate (98%) from Aldrich (Milwaukee, WI); dibasic
potassium phosphate (“Baker Analyzed”) from J.T. Baker (Phil-
lipsbury, NJ); and hexaammineruthenium(III) chloride (99%,
Ru(NH3)6Cl3) from Strem Chemicals (Newburyport, MA) were used
as received.
Gold (99.99+%) and platinum (99.99%) wires, 25 µm diameter,
from Goodfellow (Devon, PA) were used to fabricate the SECM
tips by procedures described elsewhere;4 the tips used in this study
had an RG e 1.5. All electrodes were polished prior to use with
alumina paste (0.05 µm) on microcloth pads (Buehler, Lake Bluff,
IL), sonicated for 15 min in water, and cycled between 0 and 1 V
versus NHE in 0.5 M sulfuric acid for 100 cycles to a constant
CV. Phosphate buffer solution (PBS, pH 7) was prepared by
adjusting the pH with o-phosphoric acid to a 0.3 M solution of
dibasic potassium phosphate. TRIS buffer (pH 7) was prepared by
adjusting the pH to a 0.3 M solution of TRIS base with sulfuric
acid. Chloride-free solutions of hexaammineruthenium(III) were
prepared by exchanging chloride for nitrate via potentiometric
titration with 10 mM silver nitrate versus a Ag/AgCl working
electrode; trace chloride was approximately 10 µM.
All solutions used in the electrochemical cell were bubbled with
argon prior to the experiment and kept under a humidified argon
blanket. A Ag/AgCl, saturated KCl reference electrode was used
for all experiments. To avoid chloride contamination from leakage
through the glass frit of the reference, a 1 M potassium nitrate-3%
w/v agar gel salt bridge was used. All potentials in this study are
referred to the NHE scale. In all experiments, the counter electrode
was a piece of 0.5 mm tungsten wire from Alfa Products (Danvers,
MA). The mediator concentration in all solutions used (either methyl
viologen or hexaammineruthenium(III)) was 1 mM.
SECM and other electrochemical measurements were done in a
CHI900 SECM station (CH Instruments, Austin, TX) with inchworm-
based positioning. The electrode setup, as shown in Figures 1 and
2, was achieved by inserting the substrate from below a drilled
Teflon electrochemical cell and positioning the SECM tip above
it. The tip was scanned to locate electrochemically the center of
the substrate electrode. It is critical to position the two UMEs
concentrically, and we were able to achieve this within (1 µm
lateral error. The aligning procedure based on tip generation/
substrate collection mode is described in detail in the Supporting
Information.
Figure 2. Schematics of the approach taken for digital simulations. For
this study, a ) 12.5 µm, d ) 1.25 µm, rg ) 18.75 µm, cO ) 1 mol/m3 (1
mM), and d′ ) 0.01 µm. Details of the simulation model are given in the
Simulation Model section and in the Supporting Information. Substrate is
above and SECM tip is below in this representation.
titrant R and A), and different amounts of adsorbed species A,
represented as its charge, Q. In Figure 3A, the kinetic constant
was varied over 3 orders of magnitude, and limiting behavior
under these conditions is found when k g 103 m3 mol-1 s-1
(mM-1 s-1), as shown by the overlapping of the voltammograms
with that of positive feedback before it reaches a maximum and
decays as the surface is depleted of A. This limiting behavior
establishes that the chemical reaction cannot regenerate O at a
rate faster than the diffusion of R to the substrate; this is denoted
as a titrant-limited situation. When k drops below this value,
there is a shift in the peak, and the behavior is kinetically limited.
Figure 3B shows the variation of the shape of the tip response
with the simulated charge; i.e., the total amount of A was entered
in the simulation under titrant-limited conditions (k ) 50 m3
mol-1 s-1). The system responds proportionally to the amount
of A following the positive feedback curve until A is depleted,
when it rapidly drops to negative feedback behavior. Upon
subtraction of the negative feedback curve to these tip responses
and integration of the current with time, the collection efficiency
(defined as the ratio of the charge recovered by the tip to the
input charge used in the initial charging of the substrate) is found
to be larger than 98%. This is another feature that can be
achieved only at very small tip-substrate distances when using
a substrate the same size as the tip, and it provides the advantage
of obtaining the titrated charge by direct integration of the tip
current without the need for using a calibration factor.
It is useful to have a titrant-limited system rather than a
kinetically limited one. As shown in both panels in Figure 3,
under titrant-limited conditions most of the collection takes place
Once the electrodes were positioned, their leads were connected
to a three-way switch that allowed setting the working electrode
9
16988 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

SECM Quantification of Adsorbed Species on Electrodes
A R T I C L E S
Figure 3. Representative results of digital simulations for the surface interrogation system showing linear sweep voltammograms for the first scan of the
tip, ν ) 50 mV/s and cO ) 1 mol/m3 (A) Variation of the reaction kinetic constant k at a fixed Q ) 8 nC. (B) Variation of the charge adsorbed at the
substrate Q at fixed k ) 50 m3 mol-1 s-1
.
2+
lead in the SECM to the tip, to the substrate, or as a dummy; e.g.,
if the substrate was selected, the tip and the dummy were at open
circuit. A typical measurement consisted of taking the substrate
electrode to the desired potential through a linear scan at 50 mV/s
and then turning off the cell at that potential, switching the working
electrode lead to the tip, and turning on the cell with the
programming for the tip scan. This scan was started at ap-
proximately 0.2 V versus the mediator E0 (0.5 s quiet time to allow
for double-layer charging) and run at 50 mV/s until -0.15 or -0.2
V versus E0; the time delay between the end of the substrate scan
and the beginning of the tip scan was typically 4 s. The tip was
run for two complete cycles within these potential limits to verify
the consumption of the adsorbed oxide and to obtain a blank for
subtraction of negative feedback. The charge neutralized was
calculated by numerical integration of the background- and baseline-
subtracted voltammograms.
immersed in a solution containing Ru(NH3)6 (chemically
3+
generated from a 1 mM solution of Ru(NH3)6 in PBS at pH
7 by passing through a Zn-packed syringe stored in an
Ar-flushed vial) and returned to the electrochemical cell with a
scan starting from the onset of the reduction wave, this oxide
reduction process was not observed (blue curve). If the anodized
electrode was instead immersed in a control solution of
Ru(NH3)63+, the reduction wave was still present (red curve)
with some losses that we attribute to some instability of the
oxide upon changing environments.
As described in the Experimental Section, the relative
positioning of the two electrodes is crucial to this experiment
and is done first by means of the generation/collection mode,
which is effective over longer distances, and then by positive
feedback. Figure 4B shows a typical lateral scan for locating
the center of the substrate electrode, in this case with the
substrate producing Ru(NH3)62+ while the tip is oxidizing it back
to Ru(NH3)63+; it is desirable to obtain symmetrical features in
the current profiles of both electrodes. The inset in Figure 4C
shows a typical approach curve once the centers of both
electrodes have been found and the electrodes aligned. A good
indicator for a successful alignment of the electrodes is the
collapse of the substrate and tip currents to the same curve,
which indicates nearly 100% collection efficiency (CE). Upon
a potential scan as shown in the main Figure 4C, the same high
CE should hold. Furthermore, obtaining such a scan confirms
that the electrodes are not physically touching (short-circuited).
The smoothly increasing current provided by the CV scan avoids
problems with overloading the potentiostat that were frequently
observed when a potential step was applied (chronoamperom-
etry).
Independent quantification of the oxide was achieved with
conventional CV on the substrates to validate the proposed SECM
surface interrogation technique. In the case of Au, this process was
investigated in situ in a scan prior to the oxidation of the substrate
required to run the tip scans. In the case of Pt, separate measure-
ments were required because of the accessible potential window.
Results and Discussion
To show the possibility of allowing a chemical reaction to
maintain the positive feedback necessary for our measurements,
3+/2+
we conducted preliminary experiments with the Ru(NH3)6
pair (E0 ) 0.05 V; in this study E1/2 ) 0.01 V in PBS at pH 7
on either a Au or Pt electrode)33 to evaluate the feasibility of
the reduced form (Ru(NH3)62+) being able to reduce an
electrogenerated oxide on a gold electrode. In its simplest
scheme, we assume that the reaction (surface titration) in acid
media proceeds as follows:
Once the electrodes are positioned well, the oxidation of the
substrate and its subsequent interrogation can be performed.
Figure 4D shows a typical result obtained from oxidation of a
gold substrate at 1.4 V, followed by reduction by electrogen-
erated Ru(NH3)62+ with a positive feedback response in the tip.
The general features expected from the simulation are present:
the titration of the gold oxide by the reducing agent follows
the curve of positive feedback until the consumption of all of
the surface oxide causes a sharp decrease in the feedback current.
Most of the charge neutralization occurs well before reaching
E1/2 ) 0.01 V, and after this, the current drops to the level of
the expected negative feedback value (IT ≈ 0.4,31 which for
iT,∞ ) 4 nA gives iT ) 1.6 nA). A second cycle, without
oxidation of the Au, shows that the initially present oxide has
been consumed, as the clear peak observed during the first cycle
Au2O3 + 6Ru(NH3)62+ + 6H+ f 2Au + 6Ru(NH3)63+ + 3H2O
(1)
where three electrons are transferred per gold atom. Based on
the previous knowledge that the cyclic voltammogram of gold
in PBS at pH 7 shows the onset of oxidation at approximately
0.9 V, the reaction with Ru(NH3)62+ is allowed thermodynami-
cally (∆E0 g 0.85). This is further confirmed in Figure 4A,
where a gold tip was anodized to 1.2 V for 30 s; when the
potential scan was reversed, a characteristic reduction wave was
observed (green curve), but if an anodized electrode was
(33) Bard, A. J.; Fan, F.-R. F.; Mirkin, M. V. Scanning Electrochemical
Microscopy. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel
Dekker: New York, 1994; pp 243-373.
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16989

A R T I C L E S
Rodr´ıguez-Lo´pez et al.
Figure 4. Experimental features of the technique. All potentials versus NHE. (A) A gold tip is oxidized for 30 s at 1.2 V in 0.3 M PBS, pH 7. The
voltammograms correspond to the comparison of electroreduction just following oxidation (simple reversal, green curve) against the case when they are
dipped in solutions containing either Ru(NH3)62+ (blue curve) or Ru(NH3)63+ (dark red curve), taken back to the cell, and the same scan run. Initial potential
0.85 V, ν ) 50 mV/s, dip time 2 s. (B) Line scan of the tip over the substrate, showing a desirable symmetry in the electrochemical response; the electrodes
are centered at the correlated maxima of both curves. (C) Inset: Approach curve at the centered position in panel B, stopping at the highest current. The main
curve shows a CV at this position in which the substrate is being scanned from 0.2 to -0.2 V and the tip is held at 0.2 V. (D) Representative result of the
surface interrogation, showing the differences between the first and second cycles, ν ) 50 mV/s. The Au tip was oxidized for 30 s in 0.3 M TRIS buffer,
pH 7, at 1.4 V. (E) Comparison between the simulated and experimental curves at different neutralized charges. The simulation used k ) 50 m3 mol-1 s-1
experimental charges are within 10% of the simulated input for the surface oxide and correspond to the reduction of adsorbed oxygen on platinum.
;
does not appear. The second cycle voltammogram strongly
resembles the one expected for negative feedback, although there
is evidence that a very small amount of adsorbed oxygen is
present even at these potentials, as is shown and discussed in
the Supporting Information, Figure S2. For platinum oxide, the
simplest chemical equations that can be written are for the
reduction of the oxide obtained at high positive potentials, PtO2,
and for PtO, formed at lower potentials; eqs 2 and 3 hold for
acidic conditions:
Minor differences between the experiment and the simulation
may be caused by a slight misalignment of the electrodes, but
more importantly, different reaction kinetics between the titrant
and the different species and side reactions that may be involved
in the electron transfer required to reach complete reduction of
the oxide(s) to Pt. This is expected from the multiple number
of electrons (i.e., mediator molecules) required for completion
of the reaction. This may be the cause of the discrepancy
between the sharp drops in current in the simulated curves versus
the smooth drops observed in the experiment, which resembles
the slower kinetic curves shown in Figure 3A. Such a discrep-
ancy might also arise because of a coverage-dependent rate
constant, which we do not take into account in the simulation
model. Nonetheless, most of the charge neutralization is
achieved under the titrant-limited regime, as discussed previ-
ously in the simulation model, which allows a straightforward
quantification of the amount of oxide for both Pt and Au
electrodes.
Figure 5 shows a comparison of results of the SECM surface
interrogation and those conventionally obtained by the reduction
of gold and platinum oxides by CV in PBS at pH 7 in the
presence of 1 mM Ru(NH3)63+. Figure 5A shows the cyclic
voltammograms of gold at selected reversal potentials; the
amount of oxide is quantifiable above 1.0 V by integration of
the surface reduction wave present between 0.8 and 0.5 V upon
the cathodic scan. With CV, the oxidation of the substrate below
PtO2 + 2Ru(NH3)62+ + 2H+ f PtO + 2Ru(NH3)63+ + H2O
(2)
PtO + 2Ru(NH3)62+ + 2H+ f Pt + 2Ru(NH3)63+ + H2O
(3)
These equations can be further separated into their elementary
steps; however, as in the case of gold (eq 1), a discussion of
the mechanism for the reduction process is beyond the scope
of this study. Figure 4E presents a comparison of the experi-
mental and simulated results for the case of platinum oxide that,
as discussed later, exhibits only one peak in the collector scans.
The comparison is based on the charge that is input in the
simulation versus the amount calculated by integration in the
experiment. The simulation fits quite well with the experiment,
considering that the simulation takes into consideration only a
simple single electron transfer process.
9
16990 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

SECM Quantification of Adsorbed Species on Electrodes
A R T I C L E S
Figure 5. Results for the comparison of substrate voltammetry to the interrogation of Au (A-C) and Pt (D-F) in 0.3 M PBS, pH 7, with Ru(NH3)62+. All
potentials versus NHE. (A) Voltammetry of the Au substrate, ν ) 50 mV/s, showing curves at different reverse potentials. (B) Spectrum of tip scans for the
interrogation of Au electrodes taken to different potential limits, ν ) 50 mV/s. (C) Isotherm for oxygen on gold and comparison with ref 34. The inset shows
a closeup in the low potential limit. (D,E) Same as (A) and (B) but for platinum. (F) Comparison of the integrated charge found with the interrogation
technique versus the results of the electroreduction of the oxide in independent experiments. QH stands for the available electrochemical area of Pt, taken as
the charge integration for the hydrogen adsorption peak.
0.9 V shows no presence of the surface reduction wave when
reversing the scan, and presumably only double-layer charging.
This is consistent with previous observations of gold oxidation
in PBS at pH 7 reported by Oesch and Janata34 and used as a
reference for comparison in this study. Figure 5B shows the tip
voltammograms obtained by the surface interrogation technique
for increasingly positive potentials of the substrate. The feedback
response shows three clearly defined regions in this set of curves.
There is first a potential-dependent tip current signal that starts
when the substrate is subjected to potentials well negative of
0.9 V and grows in at more positive potentials, finally stabilizing
between 0.9 and 1.1 V. A following single peak then grows
until ES ) 1.4 V, and finally a higher double-peaked signal
appears for ES > 1.4 V vs NHE. The total surface coverage on
the substrate can then be obtained by integrating these tip
reduction currents, and Figure 5C shows the isotherm for oxide
coverage on the substrate versus ES. The equivalent charge of
a monolayer was obtained by the procedure suggested by Oesch
and Janata34 that consists in the measurement of the charge of
the surface reduction wave before the onset of a second
reduction of a species dubbed “oxide II” that appears between
0.4 and 0.2 V on the reverse scan and is taken to indicate the
completion of two monolayers of Au2O3, independent of the
roughness factor of the electrode. The comparison presented in
Figure 5C shows very clearly that the SECM-based technique
senses a reactive species adsorbed at less positive substrate
potentials where the CV of the substrate electrode does not, in
either results obtained here or those previously reported.34 We
believe this is evidence for the formation of the so-called
“incipient oxide” or singly oxygenated species such as Au-OH
(34) Oesch, U.; Janata, J. J. Electrochim. Acta 1983, 28, 1237–1246.
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16991

A R T I C L E S
Rodr´ıguez-Lo´pez et al.
or Au-O, which has recently been proposed to explain the
catalytic activity toward CO oxidation present at noble metals
at the liquid-solid35-38 and gas-solid39 interfaces. The cover-
age for the incipient oxide we observe is about 20% of a
monolayer, which is consistent with very early observations done
by capacitance measurements40 (∼15%), electroreflection tech-
niques41 (10-20%), and more recent studies using contact
electroresistance42 (∼15%); indirect evidence is provided in
photoelectrochemical43 and SERS44 studies by the enhanced
adsorption of radioactively labeled alkaline earth cations45 and
by AC voltammetry.46 An inherent feature of the surface
interrogation technique is its ability to measure directly the
charge that corresponds to such an adsorption feature, thus
avoiding more complex treatments to estimate coverage. In
considering the presence of these “incipient oxides”, we must
also take into account the possibility of titrating the accumulated
positive charge in the double layer as a result of bringing the
substrate to a potential more positive than the potential of zero
charge (pzc). A simple calculation assuming a reasonably high
value of the double-layer capacitance (30 µF/cm2) and a
roughness factor of 3, and reaching about 1 V difference from
the pzc, yields a value of 0.4 nC for double-layer charging,
which represents roughly 4% of the charge for a monolayer of
gold oxide. This is just slightly higher than our limit of detection,
as will be explained later, and while it may explain some effects
observed at open circuit or at very low potentials (see the
Supporting Information), it does not account for the amounts
of incipient oxide found.
As previously noted, Figure 5B shows a transition between
1.4 and 1.5 V from a single-peaked tip response to a double-
peaked one. From Figure 5C, this transition corresponds to the
coverage reaching more than half of a monolayer. The finding
of a change in behavior upon reaching discrete coverage values
has been treated in the literature as an indicator of the presence
or transformation of certain species.47 However, in the present
case it is difficult to discern between an actual surface behavior
and an artifact from the reduction mechanism. As will be
explained later, these features change with conditions such as
the type of buffer. From a kinetic viewpoint, the second peak
in the collector scans that appears at ES > 1.5 V actually shows
more facile kinetics for reaction with the reduced mediator than
the first one, indicating that the first process is conditioned for
its appearance; e.g., a sequential instead of a parallel mechanism
for the reduction of two different oxide species would have to
be invoked. This could involve, for example, a rearrangement
of the oxygen moieties,48 either spatially within the surface or
chemically, e.g., the addition of a proton.49
Figure 5D-F shows a similar experiment with a platinum
electrode in PBS at pH 7. The voltammograms that are obtained
under these conditions are less reversible than those found in
acid media,47 and in fact, as shown in the inset of Figure 5D,
the surface reduction wave is found in a region which overlaps
3+
the reduction of Ru(NH3)6 (E0 ) 0.05 V). Thus, the direct
measurement of platinum oxide reduction by CV had to be
carried out in independent experiments in the absence of
mediator. However, the SECM technique with mediator is
especially useful in electrochemical measurements where certain
surface processes are so irreversible that they are outside of the
potential window of the working electrode or electrolyte and
overlap with other waves. In such a case, direct electrochemical
quantification is difficult. In the surface interrogation technique
presented here, we can avoid this situation because the tip
operates under different conditions than the substrate and can
be made of a different material. Moreover, the electrochemical
irreversibility is not an issue for this technique, since the positive
feedback is maintained by a purely chemical reaction. For
example, one might think that, because the surface reduction
3+
wave of platinum oxide and the reduction of Ru(NH3)6
overlap, in the reaction of the platinum oxide with the titrant
2+
Ru(NH3)6 the ∆G value would not be negative enough to
cause the spontaneous reduction to occur. However, electro-
chemical irreversibility, as in the case of reduction of an oxide
at large overpotentials, is a kinetic effect, not a thermodynamic
one. Thus, the response of a platinum oxide to the titrant will
be mostly a function of the difference in their redox potentials
and will not be limited by the kinetics of the surface reduction
of the oxide. Figure 5E shows the tip voltammogram spectrum
for oxidized platinum; a clear difference from the spectrum of
gold is the appearance of only one peak at all potentials that
were studied but also that the rate of the titration reaction, Pt
oxide with Ru(NH3)62+, appears to be more sluggish than found
with Au, as inferred from the tailing of the peaks and the broader
peaks for comparable charges. Just as in the case of Au, an
analytical signal in the tip voltammograms is obtained below
the potential limit at which the direct CV surface reduction
shows up as appreciable current on the substrate voltammogram.
Hence we assign this to an “incipient” oxide formation for the
same reasons outlined previously for Au. Figure 5F shows this
in more detail in the inset, a non-normalized isotherm. Unlike
the case of Au, where a convenient strategy exists for normal-
izing the results, in the case of Pt traditionally the charge
integrations have been divided by twice the charge integration
for hydrogen adsorption (QH).28 Because the hydrogen adsorp-
tion wave is less well defined in PBS buffer, the normalization
procedure is less useful in PBS than it is in acidic media. An
average value of 4.5 nC was found for QH for the Pt electrodes
used based on measurements in acid media and can be used as
a guideline. A comparison can be made between reported results
and the isotherm shown in Figure 5F; it presents roughly an
equivalence of one monolayer in the 1.1-1.2 V potential range
(35) Hayden, B. E.; Rendall, M. E.; South, O. J. Am. Chem. Soc. 2003,
125, 7738–7742.
(36) Beltramo, G. L.; Shubina, T. E.; Koper, M. T. M. Chem. Phys. Chem.
2005, 6, 2597–2606.
(37) Spendelow, J. S.; Goodpaster, J. D.; Kenis, P. J. A.; Wieckowski, A.
J. Phys. Chem. B 2006, 110, 9545–9555.
(38) Assiongbon, K. A.; Roy, D. Surf. Sci. 2005, 594, 99–119.
(39) Kim, T. S.; Gong, J.; Ojifinni, R. A.; White, J. M.; Mullins, C. B.
J. Am. Chem. Soc. 2006, 128, 6282–6283.
(40) Schmid, G. M.; O’Brien, R. N. J. Electrochem. Soc. 1964, 11, 832–
837.
(41) Nguyen Van Huong, C.; Hinnen, C.; Lecoeur, J. J. Electroanal. Chem.
1980, 106, 185–191.
(42) Marichev, V. A.; Russ, J. Electrochem 1999, 35, 434–440.
(43) Watanabe, T.; Gerischer, J. J. Electroanal. Chem. 1981, 117, 185–
200.
47
based on QH and an equivalent value of approximately 2.3
for Q/2QH in the potential range above 2 V, although not a very
clear stabilization to a limiting value is seen.28 Clearly, however,
(44) DeSilvestro, J.; Weaver, M. J. J. Electroanal. Chem. 1986, 209, 377–
386.
(45) Hora´nyi, G. Electrochim. Acta 1991, 36, 1453–1463.
(46) Lertanantawong, B.; O’Mullane, A. P.; Surareungchai, W.; Somasun-
drum, M.; Burke, L. D.; Bond, A. M. Langmuir 2008, 24, 2856–2868.
(47) Conway, B. E.; Angerstein-Kozlowska, H. Acc. Chem. Res. 1981, 14,
49–56.
(48) Angerstein-Kozlowska, H.; Conway, B. E.; Barnett, B.; Mozota, J. J.
Electroanal. Chem. 1979, 100, 417–446.
(49) Pletcher, D.; Sotiropoulos, S. J. Chem. Soc., Faraday Trans. 1994,
90, 3663–3668.
9
16992 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

SECM Quantification of Adsorbed Species on Electrodes
A R T I C L E S
Figure 6. Interrogation of gold electrodes under different conditions. All potentials versus NHE. (A) Spectrum of tip scans in 0.3 M TRIS buffer, pH 7,
holding the substrate at the indicated potential for 30 s, ν ) 50 mV/s. (B) Isotherm for the conditions in (A). The insert shows a closeup in the low potential
limit. (C) Spectrum of tip scans in 0.3 M PBS, pH 7, using the MV2+/MV+ mediator pair, ν ) 50 mV/s. (D) Isotherm for the conditions in (C) and
comparison to ref 34.
the behaviors of both the surface reduction wave and the surface
interrogation plots correlate. Upon reaching the range 1.6-1.7
V, both curves exhibit a plateau (∼2QH) that continues with
further growth of the oxide; above these limits, the raw curves
from the surface interrogation show an increased background
which interferes with a clean integration of the voltammograms,
most likely because of suitable conditions for the production
of large amounts of oxygen coming from water oxidation or
even platinum dissolution.28
in the shape of the collector response with PBS and TRIS buffer,
the results in Figures 6B and 5C are similar. Note that near 1.8
V the isotherm reaches a limiting coverage of 2 monolayers,
and that the shape of this response corresponds very well to
that exhibited by the surface reduction wave. At low potentials,
a limiting value of 0.2 monolayer is reached at about 1.0 V,
even when the system is allowed to remain for a longer time at
this potential. This further supports our assumption that this
feature corresponds to the “incipient oxide”.
We also show in Figure 6 results obtained with another redox
mediator to support the independence of the surface character-
ization response on mediator used. We selected the methyl
The nature of the solution ions can affect the behavior of the
oxide films. Figure 6 shows the results of the surface interroga-
tion to gold oxidation when TRIS buffer is used in place of
PBS. Figure 6A displays the behavior of gold oxide (generated
viologen pair MV2+/+ (E0 ) -0.44 V, experimental E1/2
)
-0.46 V),50 whose reduced form is a stronger reducing agent
than Ru(NH3)62+. Figure 6C presents the tip scans for inter-
rogation of gold oxide in PBS at pH 7, and Figure 6D the
corresponding isotherm which shows good agreement with the
literature reference results34 and with the CV surface reduction
wave, although compared to Figure 5B (Ru(NH3)62+ titrant), a
slight excess charge is found at all potentials. This might be
the result of an increased reactivity toward a subsequent oxide
with the MV+ species or, more likely, to its increased sensitivity
to the presence of oxygen (either electrogenerated by water
oxidation or present residually during the experiment). As with
Ru(NH3)63+/2+, the presence of two processes in the tip scans
is observed with MV2+/+, although they are slightly different
in shape and proportions than the ones observed with
by holding the electrode for 30 s at the desired potential) in
2-
TRIS buffer; unlike PBS at pH 7, which contains the HPO4
/
-
H2PO4 pair which is negatively charged, TRIS contains the
+
(CH2OH)3CNH2/(CH2OH)3CNH3 pair which is positively
2-
charged, and in this case, SO4 is used as a counterion.
Changing the nature of the buffer will affect adsorption at the
positive polarization required for anodization of the gold
substrate, so that TRIS would be less likely to interact with the
electrode. The spectrum of tip voltammograms shown in Figure
6A contrasts clearly with the one shown in Figure 5B, as it
exhibits only one peak at low and moderate polarizations. From
the isotherm in Figure 6B, one can assign the appearance of
the shoulder in the tip scans at high polarizations to the
formation of the so-called “oxide II” which occurs after
completion of two monolayers and whose onset is the basis of
the isotherm normalization for gold.34 Despite the differences
(50) Fultz, M. L.; Durst, R. A. Anal. Chim. Acta 1982, 140, 1–18.
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16993

A R T I C L E S
Rodr´ıguez-Lo´pez et al.
Ru(NH3)62+. These results indicate that, although the mechanism
of the surface reduction may be different with different
mediators, the quantification of Au2O3 is similar (cf. Figures
5C and 6D), indicating complete detection of submonolayer
amounts of the oxide.
correspondence to the expected behavior remained. This new
analytical tool should be useful for the study of adsorption
phenomena at electrocatalysts and in the characterization of
surface modified materials, and studies in this area are under
way.
In the work described, we chose CV to obtain the tip response.
The use of chronoamperometry would perhaps be more desirable
for routine quantification, and we have also explored this
approach (Figure S3 in the Supporting Information). CV
provided us with diagnostic features of the voltammograms,
reproducible results, and good figures of merit; in the case of
gold our estimated limit of detection (taken as 3 times the
average standard deviation of our measurements) is 0.24 nC,
which for the roughness factors found in our gold electrodes
(typically between 2 and 3) represents 2.1-3.4% of a mono-
layer. This figure is also equivalent to 49 µC/cm2 of geometric
area (16-25 µC/cm2 of real area considering the above-
mentioned roughness factors). Furthermore, because of the
overlapping between the positive feedback scans and the tip
response, we were able to verify experimental aspects such as
the alignment of the electrodes. Chronoamperometry does not
allow such straightforward diagnostics, and we found it to be
less reproducible; however, these results are preliminary and
its final implementation is yet to be undertaken.
Simulation Model
The total simulation space consists of two relevant subdo-
mains: the first one corresponds to the usual SECM setup for
two UMEs (a substrate and a tip, both with a ) 12.5 µm and
RG ) 1.5) that are being approached, and the second one
represents the thin slab described previously. The gap between
the tips is set to 1.25 µm, which implies L ) 0.1. In the first
subdomain, we consider only a diffusion problem in which the
reduced species R is being generated at the tip from the initially
present oxidized species O. Thus, for transient conditions
following Fick’s second law as required for implementing linear
voltammetry, we have eqs 4 and 5:
∂c
O ) DO 2cO
(4)
(5)
∂t
∂c
R ) DR 2cR
∂t
where cO and cR are the concentrations of oxidized and reduced
species, respectively, and are functions of r, z, and time t. DO
and DR are the diffusion coefficients of these species and are
given the value of 5.5 × 10-10 m2/s, which corresponds to the
Conclusions
A new electrochemical in situ technique to study surfaces
based on the transient feedback mode of SECM for the
quantification of adsorbed species at electrodes has been
described. Here an electrogenerated titrant reacts with the
electrogenerated adsorbate at the substrate electrode at open
circuit and generates a positive feedback response at the SECM
tip only while the adsorbate is present. The technique is
analogous to a directed redox titration, and because of its SECM
feedback characteristics, the experimental setup is simplified
by having a SECM tip behave as both titrant generator and
detector. Its operation was verified through the study of stable
electrogenerated oxide species on the surface of Au and Pt. The
experimental results are in general agreement with the known
behavior of these systems and with the results from digital
simulations, thus confirming the proposed mode of operation.
Isotherms for the coverage of oxygen species with potential
on Au and Pt are presented at neutral pH. The results obtained
by the proposed SECM surface interrogation agree on a first
level very well with those previously reported and obtained by
conventional CV of the substrate. However, they provide deeper
insight into the formation of species that may, in fact, be at too
small a level or obscured by other processes occurring during
a conventional potential scan on a substrate. For example, our
results strongly suggest the formation of “incipient oxides” on
Au and Pt at less positive potentials that are detected by virtue
of the sensitivity of the titration of the substrate surface and
that are not observed on the cathodic part of the substrate CV.
This feature offers the advantage of detecting species at the
potential at which they are formed by means of a chemical
reaction and not at the potential at which they become
electrochemically detectable. The detection of the gold oxides
Ru(NH3)6 cation.51 The bulk concentration of species is set
3+
to the one used during the experiment, cO* ) 1 mol/m3 and
cR* ) 0 mol/m3 (1 mol/m3 ) 1 mM); the adsorbed species A
is not present in this subdomain. At the tip boundary, z ) 0, 0
< r < a, we write the inward flux condition following
Butler-Volmer kinetics, eqs 6 and 7, and the linear potential
sweep in eq 8:
0
0
-DO cO ) -k° e-Rf(E-E )cO(0, r) + k° e(1-R)f(E-E )cR(0, r)
(6)
0
0
-DR cR ) k° e-Rf(E-E )cO(0, r) - k° e(1-R)f(E-E )cR(0, r)
(7)
(8)
E ) Ein - νt
where k° is the heterogeneous rate constant, chosen to be 0.1
m/s, avoiding thus any kinetic complication, R the symmetry
factor equal to 0.5; f, which is equal to 38.94 V-1 at 298 K,
and E0 the standard potential of the reaction. E is the potential
and relates the scan to time t by means of the scan rate ν, which
is set to 0.05 V/s, and the initial potential Ein. The second
subdomain consists of the thin slab in which the adsorbed
species A is contained and the reduction reaction shown in eq
9 occurs; its thickness was set to 0.01 µm.
k
R + AfO
(9)
This element is considered to be permeable for both O and R,
and their diffusion coefficients remain the same as in the first
subdomain. The adsorbed species A, however, is not mobile in
the thin slab; this is a reasonable assumption, as any surface
diffusivity is irrelevant to the dimensions of the substrate and
assumed to be negligible in the time scale of the experiment.
Because of this, eqs 10 and 11 for O and R contain both
diffusive and kinetic components, while eq 12 is a purely kinetic
2+
was tested with two different titrants, Ru(NH3)6 and MV+,
thus confirming the operation principle. While there were small
differences in the reduction mechanisms as observed in the
collector responses, the quantification in both cases was very
similar. Changing other conditions, such as the buffer composi-
tion, also influenced the shape of the response; however, the
(51) Baur, J. E.; Wightman, R. M. J. Electroanal. Chem. Int. Electrochem.
1991, 305, 73–81.
9
16994 J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008

SECM Quantification of Adsorbed Species on Electrodes
A R T I C L E S
function. The kinetic component is assumed to be a second-
order process associated with a kinetic constant k.
approximation used above in terms of a more widely used
quantity as the surface coverage ΓA is provided in the Supporting
Information.
The current read at the tip is obtained by integrating the
normal diffusive flux of the oxidized species at its surface and
multiplying by a geometric argument to solve for the geometry,
as shown in eq 13. In all simulations we assume a one-electron
transfer, n ) 1, for both R to A and O to R.
∂c
O ) DO 2cO + kcRcA
(10)
(11)
(12)
∂t
∂c
R ) DR 2cR - kcRcA
∂t
∂cO(r, 0)
∂c
itip
)
r)a 2πnFDOr
dr
(13)
A ) -kcRcA
∫
r)0
∂z
∂t
Acknowledgment. The support of this research by the National
Science Foundation (CHE 0451494) and the Robert A. Welch
Foundation (F-0021) is gratefully acknowledged. J.R.-L. thanks
Secretar´ıa de Educacio´n Pu´blica of Me´xico for a SEP-DGRI
scholarship support.
Because this thin slab is a two-dimensional representation
of a three-dimensional object, just as the rest of the elements
in the simulation, it is now possible to specify the amount of
adsorbate in terms of a concentration. This concentration is
calculated as the number of moles nA of adsorbed species
divided by the volume element represented by the whole thin
slab, as shown in Figure 2. While for the simulation it is
necessary to calculate nA, it is much easier to refer to this
quantity in terms of the charge it carries; thus, the charge Q is
equal to nA times Faraday’s constant F ) 96 485 C/mol. A
convenient way to normalize the results obtained through the
Supporting Information Available: Digital simulations cou-
pling cyclic voltammetry to the SECM geometry under transient
conditions using the finite element method. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA8050553
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16995