Detection of the short-lived cation radical intermediate in the electrochemical oxidation of N, N -dimethylaniline by scanning electrochemical microscopy

Article abstract of DOI:10.1021/ja511602v

The short-lived intermediate N,N-dimethylaniline (DMA) cation radical, DMA^?+, was detected during the oxidation of DMA in MeCN with 0.1 M tetra-n-butylammonium hexafluorophosphate. The detection was accomplished at steady state by scanning electrochemical microscopy (SECM) with ultramicroelectrodes using the tip generation/substrate collection mode. Cyclic voltammetry (CV) with a 2 mm Pt electrode indicates that DMA oxidation in acetonitrile is followed by a dimerization and two electrochemical reactions, which is consistent with previous results. The DMA^?+ intermediate is detected by SECM, where the DMA^?+ generated at the ca. 500 nm radius Pt tip is collected on a 5 μm radius Pt substrate when the gap between the tip and the substrate is a few hundred nanometers. Almost all of the DMA^?+ is reduced at the substrate when the gap is 200 nm or less, yielding a dimerization rate constant of 2.5 × 10⁸ M^-1·s^-1 based on a simulation. This is roughly 3 orders of magnitude larger than the value estimated by fast-scan CV. We attribute this discrepancy to the effects of double-layer capacitance charging and adsorbed species in the high scan rate CV.

Full text of DOI:10.1021/ja511602v

Article
pubs.acs.org/JACS
Detection of the Short-Lived Cation Radical Intermediate in the
Electrochemical Oxidation of N,N‑Dimethylaniline by Scanning
Electrochemical Microscopy
Fahe Cao, Jiyeon Kim, and Allen J. Bard*
Center for Electrochemistry, Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712, United
States
S
* Supporting Information
ABSTRACT: The short-lived intermediate N,N-dimethylaniline (DMA) cation
radical, DMA•+, was detected during the oxidation of DMA in MeCN with 0.1 M
tetra-n-butylammonium hexafluorophosphate. The detection was accomplished at
steady state by scanning electrochemical microscopy (SECM) with ultra-
microelectrodes using the tip generation/substrate collection mode. Cyclic
voltammetry (CV) with a 2 mm Pt electrode indicates that DMA oxidation in
acetonitrile is followed by a dimerization and two electrochemical reactions, which
is consistent with previous results. The DMA•+ intermediate is detected by SECM,
where the DMA•+ generated at the ca. 500 nm radius Pt tip is collected on a 5 μm
radius Pt substrate when the gap between the tip and the substrate is a few hundred
nanometers. Almost all of the DMA•+ is reduced at the substrate when the gap is
200 nm or less, yielding a dimerization rate constant of 2.5 × 108 M−1·s−1 based on
a simulation. This is roughly 3 orders of magnitude larger than the value estimated
by fast-scan CV. We attribute this discrepancy to the effects of double-layer capacitance charging and adsorbed species in the high
scan rate CV.
properties of polyaniline films have been presented in previous
studies.1−14 Early studies by Adams and co-workers13,15−17
indicated that the anodic oxidation of aniline and its derivatives
in aqueous and nonaqueous solutions usually occurred via a
series of reactions and resulted in a large number of products,
depending on the reaction conditions. Parker studied the
oxidation of DMA with different para-substituents by CV and
proposed that there was no reversible potential of DMA
oxidation because of the rapid consumption of the DMA•+ by
following chemical reactions. Based on linear sweep voltam-
metry, the second-order rate constants for the DMA•+ cation
radical reacting with acetate ion ranged from 3.2 × 106 M−1·s−1
for the p-methoxy-substituted cation radical to 3 × 109 M−1·s−1
for the corresponding p-nitro derivative.5 In our previous work,
FSCV was used to investigate the mechanism of the anodic
oxidation of DMA in acetonitrile solution.14 Based on FSCVs
and digital simulation, a rate constant of 6.3 × 105 M−1·s−1 was
found for the second-order cation radical coupling with a
following deprotonation reaction:
INTRODUCTION
■
In determining the mechanism of electrode reactions, it is often
helpful to detect and characterize the product of the initial
electron-transfer reaction. In many cases, the product of a one-
electron transfer reacts rapidly, for example by a proton transfer
or dimerization reaction. An ongoing challenge has been
increasing the speed of the electrochemical measurement so
that very unstable intermediates, e.g., those with lifetimes of less
than a microsecond, can be detected. One technique that has
been developed is fast-scan cyclic voltammetry (FSCV), where
an ultramicroelectrode (UME) is used to decrease contribu-
tions from capacitance. However, FSCV suffers from limitations
at a high scan rate because contributions from adsorbed species
and electrode surface processes interfere. Moreover, using a
single electrode for generating the species and detecting it
requires switching delays and complicates the measurement.
We contend that a measurement made at steady state and with
two electrodesone that generates the intermediate of interest
and one that collects (or detects) it as it traverses a small gap
as in scanning electrochemical microscopy (SECM), is a
simpler measurement that can be simulated with high accuracy.
We report here a study of a classic reaction, the oxidation of
N,N-dimethylaniline (DMA) to N,N,N,N-tetramethylbenzidene
(TMB), by the detection of the cation radical in MeCN and the
determination of its dimerization rate by SECM. The oxidation
of anilines in aqueous and nonaqueous solutions often leads to
the formation of a conducting film, and the anodic behavior of
anilines and the electrochemical, electrical, and physicochemical
•+
DMA − e− ⇌ DMA
(1)
2DMA•+ → TMB + 2H+
(2)
TMB − e− ⇌ TMB•+
(3)
Received: November 11, 2014
Published: December 5, 2014
© 2014 American Chemical Society
18163
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
TMB•+ − e− ⇌ TMB2+
(4)
FSCV has been used previously with other systems to resolve
single-electron-transfer steps in two-electron- and multi-
electron-transfer reactions and to detect unstable intermediates
in electrochemical reactions.18 As mentioned, a difficulty of
FSCV is the complexity of dealing with capacitance due to
double-layer charging and faradaic current due to adsorbed
species at high scan rates (v). Since the current caused by these
surface processes increases linearly with v, while that of
diffusion-controlled electrochemical reactions increases in
proportion to v1/2, the former dominates at larger v.18,19
A
rotating ring-disk electrode (RRDE) was also applied to study
the electro-oxidation of DMA in strong acidic aqueous
solution.17,20 Since construction of an RRDE is difficult, a
high rotation rate is required to effect interelectrode transport,
which is not required for SECM. The principles of SECM, i.e.,
generation at one electrode and collection at another, are
analogous to those employed in the RRDE study. However, the
approach curves or CVs at different fixed distances have the
advantage of larger collection efficiencies, larger fluxes between
electrodes, and the ability to vary continuously the interelec-
trode spacing.
SECM has been used as a powerful tool to detect
intermediates for over 20 years.18,21−23 Tip generation/
substrate collection (TG/SC) voltammograms for the reduc-
tion of acrylonitrile showed the electrogeneration of the anion
radical with a 5 μm diameter Au tip and a 60 μm diameter Au
substrate.21 Chang applied FSCV and SECM to investigate the
reduction of Sn(IV) bromide and detected the Sn(III)
intermediate with a 10 μm diameter Au tip and a 50 μm
diameter Au substrate.18 SECM with nanometer gaps was also
applied to study guanosine in dimethylformamide, and it was
possible to detect its cation radical with a 10 μm diameter
carbon tip and substrate in aqueous solution.23
In the current study, SECM was applied to study the electro-
oxidation of DMA in MeCN in conjunction with normal CV.
TG/SC mode with Pt UMEs of ca. 0.5 μm radius (a) for the tip
and a = 5 μm for the substrate was applied to detect the short-
lived cation radical formed in DMA electro-oxidation, and the
kinetic parameters were determined by fitting the collection
efficiency at different interelectrode distances to a simulation.
The operating principle of the TG/SC mode in SECM for the
DMA electrochemical oxidation study is illustrated in Figure 1.
Route 1 is the dimerization of DMA•+ radicals to form TMB,
and then TMB can be further oxidized at the tip at a larger d.
As d decreases, more and more radicals are captured, as shown
in route 2. To the authors’ knowledge, no nanogap SECM
experiments with tip and substrate electrodes this small (a = 0.5
μm tip and a = 5 μm substrate) have been conducted for the
detection of a cation radical in nonaqueous solution.
Figure 1. Schematic depiction of the collection of the unstable DMA•+
radical cation in TG/SC mode of SECM. When the tip is far from the
substrate, route 1, the dimerizationof DMA•+ to TMB leads to
electron-transfer reactions on the tip and substrate. For a smaller gap
between tip and substrate, route 2, more and more DMA•+ can be
captured by reduction at the substrate electrode.
Before the addition of DMA and EVD to the MeCN with TBAPF6
supporting electrolyte, the solution was purged with argon for about
10 min until the background CV (checked by potential sweeps
between 2.0 and −2.0 V) was flat. The solution was purged by argon
during the CV experiment. For the SECM experiments, a glovebag was
used to keep the system free of oxygen.24 When the collection
efficiency and approach curves were measured using SECM, no argon
was pumped into the glovebag. Otherwise, the glovebag was
continuously purged with argon.
Instrumentation and Measurement. CV and SECM experi-
ments were performed with a CHI920 SECM bipotentiostat (CH
Instruments, Austin, TX). Three electrodes were used for normal CV:
a Pt electrode (2 mm diameter) as a working electrode, a Pt wire as a
counter electrode, and a silver wire as quasi-reference electrode
(QRE). Four electrodes were used in SECM experiments: ca. a = 0.5
μm Pt UME with RG = 2 as a tip, a = 5 μm Pt UME with RG = 20 as a
substrate electrode, and the same counter and reference electrodes as
the CV experiments.
Pt UME Fabrication. The 10 μm diameter platinum wire from Alfa
Aesar (Mard Hill, MA) and a borosilicate capillary with 1.5 mm outer
diameter and 0.75 mm inner diameter from FHC (Bowdoin, ME)
were used to fabricate the 10 μm diameter UME. All were prepared by
procedures described elsewhere.25 Approximately a = 0.5 μm Pt tips
with small RG (ratio of the radius of the tip that includes the glass
sheet to that of the Pt wire) of 2 were fabricated as reported elsewhere,
especially by the Amemiya group.26−29 A laser puller (P-2000, Sutter
Instruments Co., Novato, CA) with three different programs was used
to fabricate smaller Pt wires sealed in glass. The glass layer surrounding
the tip fabricated by laser puller was heated over the Pt coil heater of a
Micro Forge (MF-900, Narishige, Tokyo, Japan) to melt it and retract
it from the tip. A sharp tip with desirable inner and outer radii was
obtained by focused ion beam (FIB) milling (scanning electron
microscopy (SEM)/FIB FEI Strata DB235 SEM/FIB with Zyvex
S100) of the heat-annealed tip across the meniscus region of the glass
layer.30 The final result was a disk shaped a = 0.5 μm Pt tip with RG =
2 as shown Supporting Information, Figure S1a, which is an electron
beam image of the tip at 52° tilt.
Electrochemical characterization of the obtained sub-micrometer
electrode was performed by means of CV using a three-electrode
configuration. Assuming a non-recessed, disk-shaped sub-micrometer
electrode, estimation of the electrode radius can be done based on the
steady-state current from CV at a scan rate of 50 mV·s−1 in 1 mM
FcMeOH aqueous solution containing 0.1 M KNO3 as supporting
electrolyte, as shown in Supporting Information, Figure S1b. The
steady-state current of FcMeOH oxidation at 0.45 V vs Ag/AgCl was
EXPERIMENTAL SECTION
■
Chemicals and Materials. N,N-Dimethylaniline (DMA, 99%,
Acros Organics, New Jersey), ethyl viologen perchlorate (EVD, Sigma-
Aldrich, St. Louis, MO), and ferrocenemethanol (FcMeOH, 97%,
Acros Organics, NJ) were used as received. Tetra-n-butylammonium
hexafluorophosphate (TBAPF6, Sigma-Aldrich, St. Louis, MO) was
chosen as supporting electrolyte. Acetonitrile (MeCN, Acros Organics,
NJ) was used as received, but about 0.5 g of ICN Alumina N-Super 1
(ICN Biomedicals, Costa Mesa, CA) was added to the solution of
MeCN for further drying. Other chemicals were reagent grade and
were used as received. All aqueous solutions were prepared with
deionized Milli-Q water.
18164
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
149 pA, and the radius for the tip was calculated to be 0.49 μm based
on the following formula,31
iss = 4nFDca
where iss is the steady-state current, n is the number of electrons (n = 1
for FcMeOH oxidation), F is the Faraday constant (96 500 C·mol−1),
D is the diffusion coefficient (7.8 × 10−6 cm2·s−1 for FcMeOH in 0.1
M KNO3 aqueous solution32,33), c is the concentration of FcMeOH,
and a is the radius of the UME. This result is consistent with an SEM
image of the tip.
The average values are consistent with values determined from
approach curves to a conductive substrate,34 which agree very well
with the corresponding theoretical curve obtained from the COMSOL
fitting, as shown in Supporting Information, Figure S1c. Note that no
contact between the tip and smooth glassy carbon substrate occurs at a
very short distance d of L < 0.2 (d/a, L = 0.2 and d = 100 nm for a =
500 nm with RG = 2.1 in Figure S1c). The result confirms that the
FIB-milled tip is smooth and perpendicularly aligned with respect to
the tip length.
Figure 3. Collection of 1 mM EVD at different tip to substrate
distances. The tip potential is scanned from 0 to −0.5 V at a scan rate
of 20 mV·s−1, while the substrate potential is fixed at 0 V. The result
indicates low collection efficiency (66%) with the bigger gap (3.6 μm),
while it is 100% with the smaller gap (0.4 μm).
RESULTS AND DISCUSSION
■
CV of DMA in MeCN. Figure 2 shows the typical CVs of
DMA obtained at a 2 mm platinum disk electrode in MeCN
without a reversal peak in the negative scan, while several new
peaks are observed in the negative scan and the second positive
scan (Figure 2). In further scans, those peaks labeled I, II, III,
IV, and V are stable, indicating no polymer film formed on the
Pt electrodes. As shown in previous studies from Adams13,16,17
and our group,14 peaks II, IV and III, V represent two new
redox couples, namely TMB oxidation to form a stable dication
as shown in eqs 1−4 and route 1 in Figure 1. With additional
EVD, the curves between 0.4 and 1.2 V are almost the same as
shown in Figure 2a, while there is a well-defined redox reaction
of EVD in the potential range of 0 to −0.5 V vs Ag QRE that
does not interfere with DMA oxidation. Therefore, EVD can be
used as redox mediator for determining the distance between
the tip and substrate in nonaqueous solution. Previous research
has shown no direct evidence of the DMA•+ radical cation due
to the high rate constant of the dimerization reaction in eq 2
and the high charging current of FSCV. In the current study, we
applied the TG/SC mode of SECM with nanometer gap and
sub-micrometer Pt UME to (1) examine the mechanism of
DMA oxidation in detail and prove the existence of the DMA•+
cation radical at steady state and (2) estimate the lifetime of the
radical and the kinetic parameters of the homogeneous radical
ion coupling reaction.
Alignment of a = 0.5 μm Tip to a = 5 μm Substrate
Using EVD in Nonaqueous Solution. Typical CVs of 1 mM
EVD in MeCN with 0.1 M TBAPF6 as supporting electrolyte
for 0.5 and 5 μm radius Pt UMEs are shown in Supporting
Information, Figure S2. Well-defined diffusion-limited currents
of 322 pA and 3.4 nA for a = 0.5 and 5 μm Pt UMEs,
respectively, indicate that those two electrodes are well-
constructed, with no recess of the Pt and no gap between the
Pt and the glass. This conclusion is also confirmed by the UME
image and the approach curve shown in Supporting
Information, Figure S1. Due to the large background current
of a large substrate electrode in TG/SC experiments, a = 5 μm
Pt UME was chosen as a substrate to decrease the background
current relative to the very small a = 0.5 μm UME tip.
Combined with CCD camera microscopy (Infinity2-1, Caltex
lens VZ-400, Ontario, Canada), a typical x and y axis scanning
curve is shown in the Supporting Information, Figure S3, where
Figure 2. CVs of 0.5 mM DMA in acetonitrile (a) with 0.1 M tetra-n-
butylammonium hexafluorophosphate as supporting electrolyte and
(b) with 1.2 mM ethyl viologen perchlorate at a 2 mm diameter Pt
electrode at a scan rate of 50 mV·s−1.
containing 0.1 M TBAPF6 supporting electrolyte at low scan
rate (v = 50 mV·s−1) with and without 1.2 mM EVD, which
serves as a potential reference and redox mediator for further
approach experiments. The CVs show only one peak on the
first positive scan at a potential of ca. 0.92 V (vs Ag QRE)
18165
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
Etip and Esub are held at −0.45 and 0 V, respectively. When the
tip scans over the substrate, a significant positive feedback peak
for the tip can be observed, and a peak collection current on the
substrate is also observed both in x- and y-axis scans.23 The
substrate current is much smaller than the tip current due to
the large d between the tip and the substrate.
For the SECM approach experiment, EV2+ was electro-
reduced on the tip and regenerated at the substrate at fixed tip
and substrate potentials. A typical approach curve using EVD as
a redox mediator in MeCN solution without oxygen is shown
in Supporting Information, Figure S4. The Etip was held at
−0.45 V to reduce the EV2+, while the Esub was held at 0 V vs
Ag QRE, since the reduction of EV2+ at −0.45 V is diffusion
limited as shown in Supporting Information, Figure S2. The red
line represents a theoretical approach curve for a diffusion-
limited reaction, derived from a COMSOL simulation. As
shown in Figure S4, the experimental approach curve is well in
line with the theoretical curve, indicating EVD is a qualified
redox mediator for determining the tip to substrate distance
and aligning the tip to the substrate. Although there are some
slight fluctuations in the curve due to vibrations in the SECM
system, the fluctuations are not large enough to prevent good
alignment of both the tip and substrate. After aligning the tip
and substrate, the collection efficiency (CE) of EVD using the
TG/SC mode at different distances was measured. As shown in
Figure 3, in the case of a 3.6 μm gap, the CE is around 66%,
while the CE is 100% at a smaller gap of 400 nm, which is in
line with the simulation results. Due to the large size of the
substrate with respect to the small tip electrode and the low
scan rate of 20 mV·s−1, a relatively high CE can be found at a
large distance (3.6 μm). Significant positive feedback is seen in
the tip current, indicating the EVD is a well-defined redox
mediator for SECM in nonaqueous solution. Otherwise, the
configuration of the 0.5 μm tip and the 5 μm substrate meets
the requirements for nanogap tip to substrate separation, and
the substrate background current is small enough to observe
short-lived radicals using the TG/SC mode of SECM.
Short-Lived Ion Radical Detection by TG/SC Mode of
SECM with Nanogap Setup. Even though FSCV was used in
our previous work, the short-lived ion radical DMA•+ could not
be detected very well because of the fast second-order
irreversible chemical reaction and a large non-faradic current
at very high scan rates (10 000 V·s−1).14 With TG/SC, a
current is measured under a steady-state condition so that clear
evidence of DMA•+ reduction can be obtained with a
reasonably small gap between the tip and the substrate
electrodes, d. Figure 4a shows typical curves of the collection
of DMA•+ in SECM at different d with substrate current
baseline subtraction. The original collection curves are shown
in Supporting Information, Figure S5, which indicates that even
at 10 μm distance, the baseline of substrate current is not
horizontal. Baseline subtraction as shown in Figure S5 (green
line) was chosen to obtain those collection curves and net
collection currents on the substrate electrode as shown in
Figure 4. On the tip electrode, 0.4 mM DMA was electro-
oxidized to form the DMA•+ by sweeping the tip potential, Etip,
Figure 4. (a, top) Typical collection curves of 0.4 mM DMA oxidation
in acetonitrile solution at different gap distances (1.4, 0.8, and 0.2 μm).
(b, bottom) Tip and corresponding substrate collection currents of all
collection experiment when the tip potential is 1.3 V vs Ag QRE. The
tip potential was swept from 0.6 to 1.3 V vs Ag QRE at a scan rate of
50 mV·s−1 to electro-oxidize DMA, while the substrate potential was
held at 0.76 V vs Ag QRE to collect the DMA•+. The original
collection curves are shown in Supporting Information, Figure S5.
Baseline subtractions of substrate currents shown in Figure S5 were
chosen to obtain those curves and substrate collection currents.
from 0.6 to 1.3 V vs Ag QRE, while the substrate potential, Esub
,
was held constant at 0.76 V vs Ag QRE to reduce DMA•+ back
to DMA on the substrate UME. Esub is a key parameter for the
TG/SC mode of SECM. In this experiment, the DMA
oxidation peak potential is around 0.92 V, and an almost
constant oxidation background current is present when the
potential is less than 0.78 V. At a more negative potential, the
Figure 5. Experimental collection efficiency and the best fit of the
results with EC2EE model simulation. The second-order rate constant
(kc) of the irreversible dimerization reaction is 2.5 × 108 M−1·s−1.
18166
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
a
Table 1. Boundary Conditions in the SECM Simulation
no.
1
z coordinate
10 μm + d
r coordinate
0 ≤ r ≤ a
boundary conditions
CB = 0
∂CB
∂z
∂CA
∂z
DB
∂CD
= −D
A
∂CC
∂z
∂CE
∂z
DD
= −DC
+ DE
∂z
2
3
10 μm + d
a ≤ r ≤ rt1
rt1 ≤ r ≤ rt2
∂CA
∂z
∂CB
∂z
∂CC
∂z
∂CD
∂z
∂CE
∂z
D
= DB
= DB
= DC
= DD
= DE
= DE
= 0
= 0
A
(10 + d)rt2 − 50rt1
rt2 − rt1
40 − d
r +
∂CA
∂z
∂CB
∂z
∂CC
∂z
∂CD
∂z
∂CE
∂z
D
= DC
= DD
A
rt2 − rt1
4
5
50 μm
10 μm
rt2 ≤ r ≤ 50 μm
0 ≤ r ≤ a′
CA = 0; CB = CB*; CC−E = 0
CA = 0
∂CA
∂z
∂CB
∂z
D
= −DB
∂CC
A
∂CD
∂z
∂CE
∂z
DD
= −DC
+ DE
∂z
6
7
10 μm
a′ ≤ r ≤ rs1
rs1 ≤ r ≤ rs2
CA
∂CB
∂z
∂CC
∂z
∂CD
∂z
∂CE
∂z
D
= DB
= DC
= DD
= DD
= DE
= 0
= 0
A ∂z
10rs2
rs2 − rs1
−10
r +
∂CA
∂z
∂CB
∂z
∂CC
∂z
∂CD
∂z
∂CE
D
= DB
= DC
= DE
A
rs2 − rs1
∂z
8
0 μm
rs2 ≤ r ≤ 50 μm
r = 50 μm
r = 0
CA = 0; CB = CB*; CC−E = 0
CA = 0; CB = CB*; CC−E = 0
9
0 ≤ z ≤ 50 μm
10
10 μm ≤ z ≤ 10 μm + d
∂CA
∂r
∂CB
∂r
∂CC
∂r
∂CD
∂r
∂CE
∂r
D
= DB
= DC
= DD
= DE
= 0
A
a
r and z are the coordinates in the radial and normal directions to the electrode surface at its center, respectively. rt1, rt2, rs1, and rs2 are characteristic
parameters for the tip and the substrate as shown in Figure S6. Di and Ci are the diffusion coefficients and concentrations of the two species, and t is
time [i = A, B, C, D, and E; species i: (A) DMA•+, (B) DMA, (C) TMB, (D) TMB•+, (E) TMB2+].
dimerization product of DMA•+ can be reduced, as shown in
the CV curves in Figure 2a. Therefore, the substrate was held at
0.76 V. DMA•+ can be collected at the substrate electrode when
d is sufficiently small.
To interpret the tip and substrate current curves at different
d values, as shown in Figure 4b, it is necessary to account for
the production of three oxidation products at the tip, DMA•+,
TMB•+, and TMB2+, and their reduction at the substrate as a
function of d. This is accounted for in the simulations described
below. However, the curves in Figure 4a yield direct evidence
that DMA•+ is produced at steady state using the TG/SC mode
of SECM.
The DMA oxidation currents on the tip and the
corresponding collection currents on the substrate (baseline
subtraction by linear fitting is shown in Figure 4a) with
different gap distances d and the tip current at 1.3 V vs Ag QRE
are shown in Figure 4b. The collection current of the DMA•+
on the substrate increases as d decreases from 1.4 to 0.2 μm,
indicating that more and more DMA•+ is captured. The tip
oxidation current is almost the same when the distance is high
than 0.5 μm. With the distance decreasing to 0.2 μm, an
obvious positive feedback effect can be found. The CE curve is
shown in Figure 5. As d decreases from 1.4 to 0.2 μm, the CE
increases from 0.17 to 0.90, indicating that almost all DMA•+ is
captured at d = 0.2 μm at steady state.
DMA Oxidation Reaction Mechanism and Kinetic
Parameters for the Irreversible Homogeneous Reaction.
When d is large, the oxidation product of DMA predominantly
forms TMB because of the high dimerization rate compared to
the DMA•+ arrival at the substrate (routes 1 vs 2 in Figure 1).
When d decreases to small enough values, more DMA•+ is
detected on the substrate. Thus, collection efficiency depends
not only on d, but also on kc, as shown in Figure 5. Simulations
were carried out with Multiphysics v4.2 software (COMSOL,
Inc., Burlington, MA) as described in the Supporting
Information. The simulation space was rendered in 2D axial
symmetrical mode for a tip with a radius of 0.5 μm, RG = 2, and
a substrate with a radius of 5 μm, RG = 3, vertically aligned in a
cylinder with a radius of 50 μm and a height of 50 μm, as
shown in Supporting Information, Figure S6. A depiction of the
EC2EE simulation model and corresponding boundary
conditions in the TG/SC mode are shown in Table 1, and
18167
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
constants k0 and transfer coefficients are assumed to be 1 cm·
s−1 and 0.5, respectively, consistent with one-electron transfers
for aromatic compounds in aprotic media.35 Comparing the
simulated and experimental data indicates that kc is about 2.5 ×
108 M−1·s−1 based on the entire d region from 1.4 to 0.2 μm.
The deviation between experimental data and simulation result
is small, and almost all generated current on the tip can be
collected on the substrate when d is 0.2 μm. For 0.4 mM DMA,
this indicates a half-life of the radical cation of around 10 μs,
which is too short to determine by FSCV.
Table 2. Reaction Mechanism and Corresponding Relevant
Time-Dependent Diffusion Equation in Cylindrical
a
Coordinates
Reactions and Parameters
B − e− ⇋ A
2A → C
E0 = 0.98 V, a = 0.5, k0 = 1 cm·s−1
kc = 2.5 × 108 M−1·s−1
C − e− ⇋ D
D − e− ⇋ E
E0 = 0.59 V, a = 0.5, k0 = 1 cm·s−1
E0 = 0.74 V, a = 0.5, k0 = 1 cm·s−1
For a second-order reaction, performing experiments at
different reactant concentrationsare an effective way to obtain
more information about the reaction mechanism. In the current
study, we have tried to collect the DMA•+ with higher and
lower concentrations of DMA than a single value, e.g., 1 mM
and 0.1 mM. However, for concentrations >0.4 mM, the
dimerization is too fast to obtain a good CE, even with a 200
nm gap. For lower concentrations, since the radius of tip is 0.5
μm, the corresponding tip and substrate collection currents are
too small (a few tens of picoamperes) to obtain reliable
measurements. We hope to achieve smaller gaps in the future,
which will allow studies over a larger range of concentrations.
Relevant Time-Dependent Diffusion Equations
2
∂CA
∂t
∂ CA
∂C
∂r
∂2CA
(1)
(2)
(3)
(4)
(5)
⎡
⎢
⎣
⎤
⎥
⎦
1
r
A
2
= D
+
+
− kcCA
A
∂r2
∂z2
2
∂CB
∂t
∂ CB
∂r2
∂C
∂r
∂2CB
∂z2
⎡
⎢
⎣
⎤
⎥
⎦
1
r
B
= D
+
+
A
2
∂CC
∂t
∂ CC
∂C
∂r
∂2CC
⎡
⎢
⎣
⎤
1
r
C
2
⎥
⎦
= DC
+
+
+ kcCA
∂r2
∂z2
2
∂2CD
∂z2
⎡
⎢
⎣
⎤
⎥
⎦
∂CD
∂t
∂ CD
∂r2
∂C
∂r
1
r
D
= DD
+
+
+
+
2
∂CE
∂t
∂ CE
∂C
∂r
∂2CE
⎡
⎤
1
r
E
⎢
⎥
= DE
CONCLUSION
∂r2
∂z2
■
⎣
⎦
The TG/SC mode of SECM, with the tip and the substrate
placed several hundred nanometers apart, was successfully
applied to detect the short-lived intermediate radical cation
DMA•+ in nonaqueous solution. This study represents the first
direct detection of the DMA•+ and indicates that the
electrochemical oxidation of DMA is followed by the
dimerization of the radical cation. Since the SECM measure-
ment can be conducted under steady state, but under a short
time window, without the problem of capacitive charging,
current that affect measurements like FSCV can be avoided. We
find that, combined with simulations based on the proposed
mechanism, the irreversible chemical reaction rate constant of
DMA•+ dimerization is 3 orders of magnitude larger than the
value estimated from FSCV.
Initial Condition, Completing the Definition of the Problem
b
t = 0, all r, z: CB = CB* = 0.4 mM, CA = 0, CC−E = 0, DA−E = 2.4 × 10−5 cm2/s
Etip = 1.3 V, Esub = 0.76 V
Current on the Tip (iT) and Substrate (iS)
a
a
⎛
⎜
⎝
⎞
⎟
⎠
⎛
⎜
⎝
⎞
⎟
⎠
∂CB
∂z
a
∂CC
∂z
iT = 2πFDB
r dr + 2πFDC
r dr
∫
∫
0
0
z=0
z=0
⎛
⎜
⎝
⎞
⎟
⎠
∂CD
∂z
+ 2πFDD
r dr
∫
′
0
z=0
a
a
⎛
⎜
⎝
⎞
⎟
⎠
⎛
⎞
∂CA
∂z
∂CC
∂z
′
⎜
⎝
⎟
⎠
iS = 2πFD
r dr − 2πFDC
r dr
∫
∫
A
0
0
z=0
z=0
a
a
⎛
⎜
⎝
⎞
⎟
⎠
⎛
⎞
⎟
⎠
∂CD
∂z
∂CE
∂z
′
′
⎜
− 2πFDD
r + 2πFDE
r
∫
∫
⎝
0
0
z=0
z=0
a
r and z are the coordinates in the radial and normal directions to the
ASSOCIATED CONTENT
■
electrode surface at its center, respectively. Di and Ci are the diffusion
coefficients and concentrations of the species i, and t is time [i = A, B,
C, D, and E; species i: (A) DMA•+, (B) DMA, (C) TMB, (D) TMB•+,
S
* Supporting Information
UME image and corresponding CV and approach curve,
alignment of tip and substrate, the original collection curve, and
COMSOL model report. This material is available free of
charge via the Internet at http://pubs.acs.org.
b
(E) TMB2+]. DMA diffusion coefficient data from our previous
work,14 while other species’ diffusion coefficients are assumed to be
the same as that of DMA.
AUTHOR INFORMATION
the reaction mechanism (a reversible electrochemical reaction
followed by a second-order irreversible chemical reaction and a
two-step electrochemical reaction, as shown in eqs 1−4). The
corresponding relevant time-dependent diffusion equations in
cylindrical coordinates are listed in Table 2. The COMSOL
report is available in the Supporting Information.
■
Corresponding Author
ajbard@mail.utexas.edu
Notes
The authors declare no competing financial interest.
The best fits of the CE curve as a function of d, with EC2EE
models for a given rate constant (kc) for the dimerization
reaction, are shown in Figure 5, along with the corresponding
experimental CE curve. In the simulation the TMB oxidation
reactions were assumed to be diffusion controlled with the
standard potentials of 0.98, 0.59, and 0.74 V vs Ag QRE for
DMA oxidation, TMB oxidation, and TMB•+ oxidation,
respectively, based on the CV in Figure 2 and CV simulation
by DigiElch,18 since the DMA oxidation reaction is not a
reversible reaction. Here, all electrochemical standard rate
ACKNOWLEDGMENTS
■
We acknowledge support of this research from the AFOSR
MURI (FA9550-14-1-0003) and the Welch Foundation (F-
0021). F.C. is especially grateful for the support of the China
Scholarship Council, Zhejiang University New Star Project and
National Natural Science Foundation of China (51171172).
F.C. thanks Dr. Hong Zhao for discussion of the COMSOL
simulation, Dr. Jinho Chang, Brent Bennett, Netzahualcoy
́
otl
Arroyo-Curras for help with the SECM experiment.
́
18168
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169

Journal of the American Chemical Society
Article
REFERENCES
■
(1) Yano, J. J. Electrochem. Soc. 1991, 138, 455.
(2) Weinberg, N. L.; Reddy, T. B. J. Am. Chem. Soc. 1968, 90, 91.
(3) Roy, P. R.; Saha, M. S.; Okajima, T.; Ohsaka, T. Electrochim. Acta
2006, 51, 4447.
(4) Roy, P. R.; Okajima, T.; Ohsaka, T. J. Electroanal. Chem. 2004,
561, 75.
(5) Parker, V. D.; Tilset, M. J. Am. Chem. Soc. 1991, 113, 8778.
(6) Park, S. G.; Trulove, P. C.; Carlin, R. T.; Osteryoung, R. A. J. Am.
Chem. Soc. 1991, 113, 3334.
(7) Oyama, N.; Ohsaka, T.; Shimizu, T. Anal. Chem. 1985, 57, 1526.
(8) Oyama, M.; Higuchi, T. J. Electrochem. Soc. 2002, 149, E12.
(9) Ocon, P.; Herrasti, P. J. Mater. Sci. 1991, 26, 6487.
(10) Lapin, E.; Jureviciute, I.; Mazeikiene, R.; Niaura, G.;
Malinauskas, A. Synth. Met. 2010, 160, 1843.
(11) Kirchgessner, M.; Sreenath, K.; Gopidas, K. R. J. Org. Chem.
2006, 71, 9849.
(12) Hand, R.; Nelson, R. F. J. Electrochem. Soc. 1970, 117, 1353.
(13) Mizoquchi, T.; Adams, R. N. J. Am. Chem. Soc. 1962, 84, 2058.
(14) Yang, H. J.; Wipf, D. O.; Bard, A. J. J. Electroanal. Chem. 1992,
331, 913.
(15) Seo, E. T.; Nelson, R. F.; Fritsch, J. M.; Marcoux, L. S.; Leedy,
D. W.; Adams, R. N. J. Am. Chem. Soc. 1966, 88, 3498.
(16) Galus, Z.; Rowland, F. S.; White, R. M.; Adams, R. N. J. Am.
Chem. Soc. 1962, 84, 2065.
(17) Galus, Z.; Adams, R. N. J. Am. Chem. Soc. 1962, 84, 2061.
(18) Chang, J. H.; Bard, A. J. J. Am. Chem. Soc. 2014, 136, 311.
(19) Bard, A. J.; Faulkner, L. R. Electrochemical methods: fundamentals
and applications; John Wiley & Sons, Inc.: New York, 2001.
(20) Neubert, G.; Prater, K. B. J. Electrochem. Soc. 1974, 121, 745.
(21) Zhou, F. M.; Bard, A. J. J. Am. Chem. Soc. 1994, 116, 393.
(22) Bard, A. J.; Fan, F. R. F.; Kwak, J.; Lev, O. Anal. Chem. 1989, 61,
132.
(23) Bi, S. P.; Liu, B.; Fan, F. R. F.; Bard, A. J. J. Am. Chem. Soc. 2005,
127, 3690.
(24) Combellas, C.; Ghilane, J.; Kanoufi, F.; Mazouzi, D. J. Phys.
Chem. B 2004, 108, 6391.
(25) Fan, F.-R. F.; Demaille, C. In Scanning Electrochemical
Microscopy; Mirkin, M. V., Bard, A. J., Eds.; Marcel Dekker, Inc.:
New York, 2001; p 75.
(26) Zhang, B.; Galusha, J.; Shiozawa, P. G.; Wang, G. L.; Bergren, A.
J.; Jones, R. M.; White, R. J.; Ervin, E. N.; Cauley, C. C.; White, H. S.
Anal. Chem. 2007, 79, 4778.
(27) Kim, J.; Kim, B. K.; Cho, S. K.; Bard, A. J. J. Am. Chem. Soc.
2014, 136, 8173.
(28) Katemann, B. B.; Schuhmann, T. Electroanal 2002, 14, 22.
(29) Kim, J.; Izadyar, A.; Nioradze, N.; Amemiya, S. J. Am. Chem. Soc.
2013, 135, 2321.
(30) Ishimatsu, R.; Kim, J.; Jing, P.; Striemer, C. C.; Fang, D. Z.;
Fauchet, P. M.; McGrath, J. L.; Amemiya, S. Anal. Chem. 2010, 82,
7127.
(31) Bard, A. J.; Mirkin, M. V. Scanning Electrochemical Microscopy;
Marcel Dekker: New York, 2001.
(32) Wain, A. J.; Pollard, A. J.; Richter, C. Anal. Chem. 2014, 86,
5143.
(33) Miao, W. J.; Ding, Z. F.; Bard, A. J. J. Phys. Chem. B 2002, 106,
1392.
(34) Shao, Y. H.; Mirkin, M. V. J. Phys. Chem. B 1998, 102, 9915.
(35) Kojima, H.; Bard, A. J. J. Am. Chem. Soc. 1975, 97, 6317.
18169
dx.doi.org/10.1021/ja511602v | J. Am. Chem. Soc. 2014, 136, 18163−18169