Oxidation of 4-chloroaniline studied by on-line electrochemistry electrospray ionization mass spectrometry

Article abstract of DOI:10.1021/ac802563f

The oxidation of 4-chloroaniline (4-CA) has been studied by electrochemistry (EC) coupled on-line with electrospray ionization mass spectrometry (ESI-MS) using two electrochemical flow cells of different design. The experimental results, which generally v

Full text of DOI:10.1021/ac802563f

Anal. Chem. 2009, 81, 5180–5187
Oxidation of 4-Chloroaniline Studied by On-Line
Electrochemistry Electrospray Ionization Mass
Spectrometry
Camilla Zettersten,† Per J. R. Sjo¨ berg,† and Leif Nyholm*,‡
Department of Physical and Analytical Chemistry, Uppsala University, Box 599, SE-751 24 Uppsala, Sweden, and
Department of Materials Chemistry, Uppsala University, Box 538, SE-751 21 Uppsala, Sweden
The oxidation of 4-chloroaniline (4-CA) has been studied
by electrochemistry (EC) coupled on-line with electrospray
ionization mass spectrometry (ESI-MS) using two elec-
trochemical flow cells of different design. The experimen-
tal results, which generally verify previously suggested
oxidation pathways for 4-CA, also indicate the presence
of an up to now unrecognized comproportionation reac-
tion. The oxidation of 4-CA (m/z 128.2) was found to give
risetotheformationofbothanoxidizeddimer, 4-[(4-chloro-
phenyl)imino]-2,5-cyclohexadien-1-imine (m/z 217.2),
and a reduced dimer, 4-amino-4′-chlorodiphenylamine
(m/z 219.2), in addition to a dimer intermediate (m/z
253.2). The unexpected formation of the reduced dimer
is shown to stem from a comproportionation reaction
involving 4-CA and the oxidized dimer. The presence of
the latter reaction was clearly seen by comparing results
obtained with two thin-layer flow cells, both with conver-
sion efficiencies of 50% under mass transport controlled
conditions but of different design with respect to the
influence of the counter electrode reaction on the reaction
at the working electrode. The experimental results dem-
onstrate that the formation of the reduced dimer is favored
by a decrease in the local pH in the flow cell, and the
influence of the pH on the oxidation of 4-CA was also
investigated in the pH range between 2.0 and 6.0 using
off-line voltammetry. It is concluded that EC/ESI-MS is a
powerful tool for the study of the present type of reactions
and that studies of the reaction pathways of these systems
are best carried out under noncoulometric experimental
conditions as the latter facilitates the detection of reaction
intermediates. Comproportionation reactions, similar to
the reaction present in the 4-CA system, can also be
expected to be present during the formation of conducting
polymers such as polyaniline and polypyrrole.
studies regarding the early stages of the formation of polyaniline.
The electrochemical polymerization mechanism of aniline itself
is very complex and has been widely debated and extensively
studied over the last four decades.1,4-10 Due to the para-
substitution, the oxidation of 4-CA, unlike that of aniline, only gives
rise to the formation of dimers.1,2 It has been found1 that aromatic
amines, including 4-CA, undergo a rapid head-to-tail coupling
which in the case of 4-CA was reported to give rise to a 4-amino-
4′-chlorodiphenylamine in the oxidized form. The formation of this
oxidation product was assumed to take place via the formation of
a previously proposed11 intermediate. The oxidation products
formed upon the oxidation of aromatic amines in aqueous media
(at pH 4) have also been identified1 by electrochemical, spectro-
scopic, and quantitative chemical methods. Cyclic voltammetry
has been used9 to study the oxidation of aromatic amines in acidic
aqueous and nonaqueous media. It was found that the oxidation
in aqueous solutions gave rise to monocation radicals which
underwent tail-to-tail, head-to-tail, and head-to-head couplings to
yield benzidine, p-aminodiphenylamine, and hydrazobenzene,
respectively. In acetonitrile, aniline, on the other hand, was found9
to be oxidized in two one-electron steps through the formation of
a monocation. While the monocation did not undergo any coupling
reactions, the dication formed in the second oxidation step was
found to couple with aniline forming hydrazobenzene, which
instantaneously was oxidized to give azobenzene. Neither benzi-
dine nor p-aminodiphenylamine was formed during the oxidation
of aniline in acetonitrile.9 The electrochemical oxidation of 4-CA
in acetonitrile has likewise been studied.3 It was concluded that
the mechanism proposed by Bacon and Adams1 describing the
voltammetric oxidation of 4-substituted aniline derivatives in acidic
(1) Bacon, J.; Adams, R. N. J. Am. Chem. Soc. 1968, 90, 6596–6599
(2) Amatore, C.; Farsang, G.; Maisonhaute, E.; Simon, P. J. Electroanal. Chem.
1999, 462, 55–62
(3) Ka´da´r, M.; Nagy, Z.; Karancsi, T.; Farsang, G. Electrochim. Acta 2001,
46, 1297–1306
(4) Mohilner, D. M.; Adams, R. N.; Argersinger, W. J., Jr. J. Am. Chem. Soc.
1962, 84, 3618–3622
(5) Yang, H.; Bard, A. J. J. Electroanal. Chem. 1992, 339, 423–449
(6) Adams, R. N.; McClure, J. H.; Morris, J. B. Anal. Chem. 1958, 30, 471–
475
(7) Wawzonek, S.; McIntyre, T. W. J. Electroanal. Chem. 1967, 114, 1025–
1029
(8) Deng, H.; Van Berkel, G. J. Anal. Chem. 1999, 71, 4284–4293
(9) Sharma, L. R.; Manchanda, A. K.; Singh, G.; Vermal, R. S. Electrochim. Acta
1982, 27, 223–233
(10) Desideri, P. G.; Lepri, L.; Heimler, D. J. Electroanal. Chem. 1971, 32, 225–
234
(11) Daniels, D. G. H.; Saunders, B. C. J. Chem. Soc. 1951, 2112–2118
.
.
.
.
The oxidation of 4-chloroaniline (4-CA) has been studied by
many authors and several reaction mechanisms have been
suggested during the years.1-3 One reason for the interest in this
system is that the oxidation of 4-CA is important from a
fundamental point of view as it can serve as a model system for
.
.
.
.
* Corresponding author. Phone: +46 (0)18 4713742. Fax: +46 (0)18 513548.
E-mail: Leif.Nyholm@mkem.uu.se.
.
† Department of Physical and Analytical Chemistry.
‡ Department of Materials Chemistry.
.
.
5180 Analytical Chemistry, Vol. 81, No. 13, July 1, 2009
10.1021/ac802563f CCC: $40.75  2009 American Chemical Society
Published on Web 05/28/2009

aqueous media was the main electrochemical oxidation route in
acetonitrile as well. 2-Amino-4′,5-dichloro-diphenylamine was iden-
tified3 as an oxidation product of 4-CA, and other oxidation
products (prepared by controlled potential coulometry) were also
isolated and identified by GC/ECD, GC/MS, and ESI-MS. It was
found that head-to-tail coupling dominated and that head-to-head
coupling was absent, in agreement with previous findings.1,4,5
Experimental evidence for the existence of radical cations formed
during the anodic oxidation of p-halogenoanilines has also been
presented2 as well as a reaction scheme for the oxidation of these
types of compounds, including 4-CA. An alternative mechanism,
however, was recently proposed,12 and it was concluded that no
single reaction mechanism alone can account for the obtained in
situ electrochemical electron spin resonance (ESR) and voltam-
metric results. It is, therefore, clear that additional research, for
example involving complementary techniques such as EC/ESI-
MS, is needed to obtain a better understanding of the oxidation
reaction mechanism for 4-CA.
In the present paper, on-line EC/ESI-MS and off-line cyclic
voltammetry were used to study the electrochemical oxidation of
4-CA. The EC/ESI-MS experiments were carried out with two
electrochemical flow cells both with conversion efficiencies of
about 50% under mass transport controlled conditions,13-15 and
the results obtained with the two cells were used to obtain new
information regarding the oxidation pathway for 4-CA. It is shown
that the oxidation pathway involves a previously unrecognized
comproportionation reaction and that EC/ESI-MS is a very
powerful tool in the study of the present type of electrochemical
reactions.
The detection of the electrochemically generated oxidation
products of 4-CA was made in the positive electrospray mode using
a PE-Sciex API I single quadrupole mass spectrometer (Concord,
ON, Canada) with an IonSpray interface. In the EC/ESI-MS
experiments, a thin layer EC flow cell (Bioanalytical Systems Inc.,
West Lafayette, IN) equipped with a 3 mm glassy carbon (GC)
disk electrode was employed in conjunction with a battery-powered
PalmSens potentiostat using the PalmScan software (version 1.42,
Palm Instruments BV, Houten, The Netherlands). The EC/ESI-
MS experiments were performed with two previously described13-15
EC flow cell setups referred to as the conventional and modified
thin-layer flow cell, respectively. A description of these cells and
additional experimental details are given in the Supporting
Information. A Bluetooth wireless connected HP iPAQ Pocket PC
(h2200 series, Hewlett-Packard Company) was utilized to control
the PalmSens instrument. In this way, the instrument could be
operated without risk for hazardous shocks although the thin-
layer EC flow cell and the PalmSens instrument were kept floating
at the potential induced by the ESI high voltage. A syringe pump
(Syringe Infusion Pump 22, Harvard Apparatus Inc., Cambridge,
MA) operating at a flow rate of 10 µL/min was utilized to pump
the solution through the flow cell and into the ion source of the
mass spectrometer. An isolation transformer (type PVM 440 0019,
Tufvassons, Sigtuna, Sweden) was used to float the syringe pump.
In this way, undesired backward currents from the ESI interface
held at high voltage (4.5 kV) to the pump could be avoided.
A 280 mm long (50 µm inner diameter and 185 µm outer
diameter) fused silica capillary (Polymicro Technologies, Phoenix,
AZ) was used to connect the thin-layer EC flow cell with the ESI-
MS instrument. The dead volume between the thin-layer EC flow
cell and the ESI emitter was about 0.55 µL, corresponding to a
transfer time of about 3.3 s at a flow rate of 10 µL/min. In the
experiments, the working electrode potential was changed in
increments of 0.2 V from 0 to +2.2 V vs Ag/AgCl. Experiments
were also performed during open circuit conditions for compari-
son. Background correction was made by mass spectra subtraction
of spectra recorded for 4-CA, FA/ACN solutions obtained under
open circuit conditions from spectra recorded for oxidized 4-CA,
FA/ACN solutions. Data were collected on a Macintosh computer
by the Tune2.5-FPU software, and the mass spectra were repro-
duced in Excel (Microsoft) after export/import of the data from
the Macintosh computer. To avoid electric shocks, particular care
should be taken when handling the floating EC/ESI-MS system when
the ESI high voltage is on.
EXPERIMENTAL SECTION
Electrochemistry/Electrospray Ionization Mass Spectrom-
etry (EC/ESI-MS). For the EC/ESI-MS experiments, a 5 µM
4-CA (98%, Sigma-Aldrich Chemie GmbH, Steinheim, Germany)
standard solution was prepared in 35/65 (% v/v) H2O/CH3CN
(i.e., 65% ACN) containing 20 mM formic acid (FA) (98-100%,
GR for analysis, Merck KGaA, Darmstadt, Germany) using a
4-CA stock solution. The latter was a freshly prepared 1 mM
solution in acetonitrile (ACN) (LiChrosolv, isocratic grade for
liquid chromatography, Merck KGaA, Darmstadt, Germany).
The stock and standard solutions of 4-CA were prepared
immediately prior to the experiments to avoid oxidation and
dimerization caused by the exposure of the sample to oxygen
(and light) although the solutions have been reported to be
stable.16 Blank solutions, containing 20 mM FA in 35/65
(% v/v) H2O/CH3CN, were also used in the ESI-MS experi-
ments. Deionized water was obtained from a Milli-Q Plus
system (Millipore Corp., Marlborough, MA). Caution: 4-CA is
highly toxic/carcinogenic and care should be taken during the
handling, analysis, and disposal of this substance.
Cyclic Voltammetry. The voltammetric experiments were
performed in 35/65 (% v/v) H2O/CH3CN with 1.0 mM solutions
-19
of 4-CA in pH 2.0, 4.0, and 6.0 Britton-Robinson buffers17
also containing 20 mM FA (Merck) and 0.5 M sodium nitrate
(>99.5%, Merck KGaA, Darmstadt, Germany). The cyclic voltam-
metry measurements were performed in quiescent solutions (with
a scan rate of 10, 125, or 250 mV/s), and a MacLab system
(ADInstruments, Castle Hill, Australia) connected to a personal
computer equipped with the MacLab Echem software (ADInstru-
(12) Streeter, I.; Wain, A. J.; Thompson, M.; Compton, R. G. J. Phys. Chem. B
2005, 109, 12636–12649
(13) Bo¨kman, C. F.; Zettersten, C.; Sjo¨berg, P. J. R.; Nyholm, L. Anal. Chem.
2004, 76, 2017–2024
(14) Eilers, G.; Zettersten, C.; Nyholm, L.; Hammarstro¨m, L.; Lomoth, R. Dalton
Trans. 2005, 6, 1033–1041
(15) Zettersten, C.; Lomoth, R.; Hammarstro¨m, L.; Sjo¨berg, P. J. R.; Nyholm, L.
J. Electroanal. Chem. 2006, 590, 90–99
(16) Chen, H.; Zhang, Y.; Mutlib, A. E.; Zhong, M. Anal. Chem. 2006, 78, 2413–
2421
.
.
(17) Pereira, F. C.; Stradiotto, N. R.; Zanoni, M. V. B. An. Acad. Bras. Cienc.
.
2002, 74, 425–432
(18) Davies, T. J.; Banks, C. E.; Bharathi, N.; Rusling, J. F.; France, R. R.;
Wadhawan, J. D.; Compton, R. G. Green Chem. 2002, 4, 570–577
(19) Peckova´, K.; Barek, J.; Moreira, J. C.; Zima, J. Anal. Bioanal. Chem. 2005,
381, 520–525
.
.
.
.
.
Analytical Chemistry, Vol. 81, No. 13, July 1, 2009 5181

Figure 1b) can most likely be ascribed to contaminants,20
including phthalates,21 as the latter peaks were also found to
be present under open circuit potential conditions without and
with 4-CA present in the solution with both flow cells. These
background peaks will therefore not be discussed further.
Figure 1b shows that the oxidation of 4-CA (m/z 128.2) gives
rise to oxidation products with peaks at m/z 217.2, 219.2, and
253.2. As opposed to the background peaks at m/z 141.3, 177.2,
205.3, 279.5, and 391.5, the peaks due to the oxidation products
(and the 4-CA peak) all exhibited isotopic patterns (see the
insets in Figure 1) in good agreement with the theoretical
chlorine isotopic pattern. Our previous findings13,15 have shown
that the influence of the electrochemical reactions inherent in
the electrospray ionization process on the appearance of the
mass spectra is negligible with the employed IonSpray interface.
This ensures that the oxidation products of 4-CA must have
been formed in the electrochemical cell. The latter is further
supported by the absence of any significant peaks at m/z 217.2,
219.2, and 253.2 in Figure 1a. There were also no ground points
between the electrochemical cell and the electrospray emitter
which could give rise to interfering electrochemical reactions
affecting the intensities of the electrochemically active species.
The oxidation products (with peaks at m/z 217.2, 219.2, and
253.2) formed upon the oxidation of 4-CA can be identified on
the basis of previous work1,2,4,5,12 in which reaction schemes for
the oxidation of 4-CA have been presented. The oxidation of 4-CA
has, thus, been shown to lead to the formation of a head-to-tail
dimer according to Scheme 1, in which it is presumed that the
oxidized dimer, DimOx, is formed via an electrochemical oxida-
tion (E) of 4-CA yielding radical cations in step 1 followed by
a chemical step (C1) involving radical-radical coupling produc-
ing an intermediate (2a). This intermediate was then assumed
to rearrange (C2) in step 2b giving rise to the total reaction 3.
It should be noted that the electrochemical step (E) in Scheme
Figure 1. EC/ESI-MS mass spectra obtained with the modified cell
for a potential of (a) 0.0 V before and (b) +1.45 V vs Ag/AgCl after
background correction, using a solution containing 5 µM 4-CA, 20
mM FA, and 65% ACN. The insets depict magnifications of the mass
spectra for the following m/z ratios (i) 128.2, (ii) 217.2 and 219.2,
and (iii) 253.2.
1
has been written for a pH value below the pKa value (i.e., 4.01)
of protonated 4-CA. As will be shown below, the oxidation of
4-CA can, however, also straightforwardly be carried out at pH
values above this pH value.
ments) was employed for the measurements. The working
electrode was a 3 mm glassy carbon disk electrode (Bioanalytical
Systems Inc.) while a platinum electrode was used as the counter
electrode. An Ag/AgCl reference electrode, or a stainless steel
quasi reference electrode (QRE), was used as the reference
electrode.
Based on Scheme 1, it is reasonable to assume that the peak
at m/z 217.2 corresponds to the oxidized dimer, DimOx, while
the peak at m/z 253.2 can be ascribed to the intermedi-
ate formed in reaction 2a. Although the isotopic pattern in the
region between m/z 253 and 257 differs somewhat from the
theoretical pattern, most likely due to the low signal intensities,
the pattern is in good agreement with the presence of a dimeric
molecule with two chlorine atoms, i.e., the formation of an
intermediate such as that previously proposed.2 In Figure 1b,
it is clearly seen that the intensity for the intermediate peak at
m/z 253.2 was much lower than for the other oxidation products
at m/z 217.2 and 219.2, respectively. The distribution of the
isotopic peaks in the m/z region between 216 and 222 (see inset
in Figure 1b) indicate the presence of an overlap between two
different species, corresponding to two different 4-CA “oxidation”
RESULTS AND DISCUSSION
EC/ESI-MS. Figure 1 exhibits mass spectra recorded with
the modified flow cell (which has a conversion efficiency of
about 50% under mass transport controlled conditions15) at a
potential of (a) 0.0 V and (b) +1.45 V vs Ag/AgCl after
background correction, respectively. As is seen in Figure 1a,
protonated 4-CA gave rise to a peak at m/z 128.2 with a typical
chlorine isotopic pattern while the background corrected
spectrum in Figure 1b features new peaks in the m/z range
between 216 and 222 (ii) and between 252 and 257 (iii),
respectively. The background correction in the latter case was
made by subtracting the mass spectrum for the nonoxidized
sample solution (i.e., 4-CA, FA, 65% ACN) from the mass
spectrum for the oxidized sample solution. The peaks seen at
m/z 141.3, 177.2, 205.3, 279.4, and 391.5 in Figure 1a (and in
products in this m/z range. Although the oxidized dimer, DimOx
,
contributes one of its isotopic peaks to the signal at m/z 219.2,
it is clear from the mass spectra that this effect cannot explain
(20) Ende, M.; Spiteller, G. Mass Spectrom. Rev. 1982, 1, 29–62
(21) Guo, X.; Bruins, A. P.; Covey, T. R. Rapid Commun. Mass Spectrom. 2006,
20, 3145–3150
.
.
5182 Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

Scheme 1. Electrochemical Oxidation of 4-CA Yielding Radical Cations (1) and the Follow-Up Chemical Reactions
Generating an Intermediate with m/z 253.2 (2a) and the Oxidized Dimer, DimOx, with m/z 217.2 (2b), Respectivelya
a The total reaction 3 was obtained as a sum of reactions 1, 2a, and 2b.
the full intensity of the m/z 219.2 peak. This was also verified
by carrying out an isotopic correction of the mass spectra (see
Table 1. Assignment of the Peaks in the Mass Spectra
Supporting Information). The intensities for m/z 219.2 in the
following figures hence refer to the resulting isotopic corrected
intensities. As has been previously demonstrated,1 the oxidized
dimer can be reduced to yield a corresponding reduced dimer
(DimRed), 4-amino-4′-chlorodiphenylamine, with m/z 219.2 at
potentials more negative than about +0.4 V vs SCE (saturated
calomel electrode). It is, therefore, reasonable to assume that
the peak at m/z 219.2 corresponds to DimRed even though the
presence of this species is puzzling as the flow cell in these
cases was operated at potentials for which the reduced dimer
should not be stable. Although the formation of the reduced
dimer as a result of the counter electrode reaction would be
possible in the conventional cell, this could obviously be ruled
out for the modified cell in which the counter electrode was
separated from the working electrode. The presence of the
reduced dimer was confirmed by comparing the theoretical
isotopic pattern for the oxidized and reduced dimers, i.e., 4-[(4-
chlorophenyl)imino]-2,5-cyclohexadien-1-imine and 4-amino-4′-
chlorodiphenylamine (which both contain one chlorine atom),
with the experimentally obtained spectra. We, therefore,
conclude that both the oxidized and reduced dimers are indeed
formed during the oxidation of 4-CA. Our results were,
therefore, similar to the previously published results8 showing
that about 75% of the dimers formed upon EC/ESI-MS oxida-
tion of aniline were in the oxidized state while 25% was in the
reduced state. The presence of reduced dimers upon the
oxidation of 4-CA consequently shows that the reaction scheme
presented in Scheme 1 cannot be complete since a reaction
generating the reduced dimer is missing. Our assignment of the
peaks in the mass spectra is given in Table 1.
m/z
label
name
128.2 4-CA
217.2 DimOx
4-chloroaniline
4-[(4-chlorophenyl)imino]-2,5-cyclohexadien-1-
imine
4-amino-4′-chlorodiphenylamine
219.2 DimRed
253.2 intermediate
formed upon the oxidation of 4-CA. While Chen et al.16 employed
a coulometric cell, our results and the results of Deng and Van
Berkel8 were obtained with noncoulometric thin-layer flow cells.
This suggests that the presence of the reduced dimer could be
due to a homogeneous reaction involving the oxidized dimer and
4-CA. In the subsequent sections, experimental evidence for such
a homogeneous reaction between 4-CA and the oxidized dimer
will be presented.
Figure 2 shows the average peak intensities (n ) 4) for the
background and isotope corrected peaks at selected m/z values
as a function of the applied potential for (a) the conventional and
(b) the modified thin-layer flow cell. (The background correction
was in this case made by setting the signal obtained at 0 V to
zero.) Upon the onset of the oxidation of 4-CA at potentials more
positive than about +0.3 V vs Ag/AgCl, the signal at m/z 128.2
clearly decreases as a result of the consumption of 4-CA. It is also
seen that the intensity for the oxidized dimer (m/z 217.2) starts
to increase at a potential of about +0.8 V while more positive
potentials (i.e., about +1.1 V vs Ag/AgCl) are needed for the
intensity of the reduced dimer (m/z 219.2) to increase for both
cells. Although a careful comparison of the results obtained with
the two flow cells is complicated by the possibility that the iR
drop within the cells differed slightly (it is reasonable to assume
that the iR drop was somewhat larger in the conventional cell), it
is clearly seen that the relative intensity of the reduced dimer
was significantly higher with the modified cell than with the
conventional cell. A closer examination of the data (see Supporting
Information) also reveals that the intensities for the oxidized dimer
were almost the same for both cells while the intensities for the
intermediate were significantly smaller for the conventional than
for the modified cell. The latter is interesting as it suggests that
the rate of formation of the oxidized dimer from the intermediate
One important clue in the search for the reaction giving rise
to the reduced dimer under oxidizing conditions is that only the
oxidized dimer (i.e., m/z 217) has been found16 when one employs
a coulometric flow cell in EC/ESI-MS oxidation studies of 4-CA.
These results were explained on the basis of the hypothesis that
the oxidation of 4-CA led to the formation of a reduced head-to-
tail dimer that in turn was oxidized prior to its detection by ESI-
MS. However, this explanation seems unlikely based on previous
work1,2,4,5,12 which indicates that no reduced dimer should be
Analytical Chemistry, Vol. 81, No. 13, July 1, 2009 5183

Figure 3. Average peak (isotope corrected) intensity ratios (n ) 4)
between the ESI-MS intensities obtained with the modified cell and
conventional cell as a function of applied potential for the species
labeled as follows: m/z 128.2 as [, m/z 217.2 as 9, m/z 219.2 as 2,
and m/z 253.2 as b.
experiments. The decrease in the local pH should be larger with
the modified cell than with the conventional cell as the reduction
of water at the counter electrode (situated only 16 µm away from
the working electrode) should result in an increased pH at the
latter electrode in the conventional cell. A lower pH with the
modified cell would also explain the higher signal intensities
obtained with the modified cell in general as a lower pH would
be expected to give rise to higher intensities in positive electro-
spray due to more efficient protonation of the detected species.
Our data, thus, indicate that the rate of the formation of the
reduced dimer is pH dependent. It is, therefore, interesting to
study the influence of the pH on the 4-CA oxidation and
dimerization reactions. Such investigations involving solutions with
different and controlled pH values, however, are not straightfor-
wardly carried out with EC/ESI-MS due to the fact that ion
suppression effects prevent buffers with sufficient buffer capacities
to be employed. As will be described in the next section, these
types of studies are, therefore, better carried out when one
employs off-line voltammetry.
Off-Line Cyclic Voltammetry. Off-line cyclic voltammetric
experiments were carried out to study the influence of the pH on
the oxidation of 4-CA. These experiments were carried out with
quiescent solutions in a glass cell initially using 5 mM solutions
of 4-CA also containing 20 mM FA, 65% acetonitrile, and 0.5 M
NaNO3. As is seen in Figure 4, the oxidation of 4-CA started at
potentials more positive than about +0.75 V versus Ag/AgCl and
the shape of the oxidation peak indicates that the oxidation of
4-CA, i.e., the electrochemical reaction 1 in Scheme 1, was mass
transfer controlled.22 This is further supported by the finding that
the 4-CA oxidation peak current ip was found to be proportional
to υ1/2 for the scan rates used in this study (i.e., 10, 125, 250
mV/s). No reduction peak was seen on the reverse scan in
accordance with the presence of a rapid follow-up chemical step
involving dimerization of the formed radical cations (i.e.,
reactions 2a and 2b in Scheme 1). At more negative potentials,
the formed oxidized dimer can, however, be seen to undergo
reduction to yield the reduced dimer. The latter can then be
oxidized on the subsequent anodic scan. This clearly shows that
reduced dimers should not be formed at the potentials used in
Figure 2. Average peak (background and isotope corrected) ESI-
MS intensities (n ) 4) as a function of the applied potential obtained
with (a) the conventional electrochemical cell and (b) the modified
electrochemical cell. The species have been labeled as follows: m/z
128.2 as [, m/z 217.2 as 9, m/z 219.2 as 2, and m/z 253.2 as b.
The m/z 128.2 signals were divided by a factor of 10 for clarity.
was lower in the modified cell. In addition, in Figure 2, it is seen
that the intensity of the reduced dimer also continued to increase
up to more positive potentials with the modified flow cell than
with the conventional cell. Figure 2, thus, clearly shows that
the oxidized dimer is formed prior to the reduced dimer and the
oxidized dimer, therefore, cannot be formed via oxidation of the
reduced dimer. It is also clear that more reduced dimer was
formed with the modified flow cell than with the conventional cell
and that this difference must be due to the influence of the counter
electrode reaction in the conventional cell as the conversion
efficiencies of the two cells were the same under the employed
experimental conditions.
In Figure 3, the average peak intensity ratios (ECmodif/ECconv
,
n ) 4), for the different m/z species have been plotted versus
the potential to more clearly demonstrate the different behavior
of the two cells. Since all ratios were found to be higher than
unity, it was immediately evident that higher signal intensities
were always obtained with the modified thin-layer flow cell.
The other striking feature in Figure 3 is the dramatic increase
in the ratio for the reduced dimer for potentials more positive
than about +1.4 V, i.e., at potentials where the current versus
potential data indicated the onset of water oxidation. As the
oxidation of water results in a decrease in the pH at the working
electrode, this suggests that the formation of the reduced dimer
is favored by a decrease in the local pH. Such a decrease in the
pH as a result of the oxidation of water is not unexpected due to
the poor buffer capacity of the solution used in the EC/ESI-MS
(22) Bard, A. J.; Faulkner, L. R. Electrochemical methods , 2nd ed.; John Wiley
& Sons, Inc.: New York, NY, 2001.
5184 Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

findings.1,2,12 It has been shown2 that two protons and two
electrons are involved in this reaction which should yield a shift
in the peak potentials of -59 mV/pH in excellent agreement with
our results. Calculations of the ratio between the peak heights
for the reduction of the oxidized dimer and the oxidation of the
reduced dimer showed that the ratio was unity for all pH values
and scan rates used. The voltammetric results, hence, do not
provide any explanation as for why the signal intensity for the
reduced dimer should increase more than that for the oxidized
dimer when the pH is decreased. From the voltammetric results,
it is, however, possible to calculate the number of electrons, n,
involved in the oxidation of 4-CA on the basis of the following
22
well-known equation, ip ) (2.69 × 105)n3/2ADO1/2COυ1/2
.
By assuming a diffusion coefficient DO of 1.0 × 10-5 cm2/s
(which is relevant for aniline,23 p-toluidine,23 and 4-CA),24 the
n value was found to be 1.0 at 25 °C for a working electrode
area A of 0.07069 cm2 and a 4-CA concentration CO of 1.0 × 106
mol/cm3. This value is in excellent agreement with reaction 1
in Scheme 1 involving the formation of a 4-CA cation radical, as
well as other findings.1 From the present results, it is clear that
the oxidized dimer is indeed formed according to Scheme 1 and
not via the oxidation of an initially reduced dimer. It is also evident
that 4-CA is electroactive in both its protonated and deprotonated
form and that the rate of the creation of the cation radicals is
mass transport controlled and independent of the pH of the
solution for pH values between 2.0 and 6.0.
Figure 4. Cyclic voltammograms recorded with a solution containing
5.0 mM 4-CA in 20 mM FA, 0.5 M NaNO3, and 65% ACN at scan
rates of 10, 125, and 250 mV/s, respectively.
Generation of Reduced Dimers by Comproportionation.
The fact that the reduced dimer (i.e., m/z 219.2) should not be
formed at the positive potentials employed in the EC/ESI-MS
experiments suggests the presence of a previously unrecognized
comproportionation reaction between unreacted 4-CA and the
formed oxidized dimer as outlined in Scheme 2 below. The
comproportionation reaction (i.e., reaction 4), which gives rise to
the reduced dimer and additional radical cations, would explain
the presence of the reduced dimer in our experiments as well
as the absence of the reduced dimer in previous experiments16
when one employs a coulometric flow cell. Since our flow cells
have a conversion efficiency of about 50% under mass transport
controlled conditions, both the oxidized dimer and 4-CA will be
present in the solution after the passage of the flow cell. During
the transfer time of about 3.3 s (at 10 µL/min), during which the
solution is passed from the electrochemical cell to the ion source,
the oxidized dimer can, thus, react with 4-CA. This means that
our experimental conditions are well suited for the detection of
the influence of the comproportionation reaction while this
reaction clearly would go unnoticed when coulometric cells are
used. The fact that the reduced dimer could be seen in the mass
spectra suggests that the comproportionation reaction must be
sufficiently fast with respect to the transfer time (i.e., 3.3 s)
between the electrochemical flow cell and the electrospray emitter.
This is further supported by the finding that the ratio between
the signal intensities at m/z 219 and 217 remained constant when
transfer lines of different lengths (i.e., 500, 2300, and 6300 mm)
were employed between the electrochemical cell and the emitter.
It is likewise interesting to note that our experimental conditions
also enabled us to detect the m/z 253.2 intermediate proposed in
Figure 5. Cyclic voltammograms recorded for a solution of 1.0 mM
4-CA in Britton-Robinson buffers with pH 2.0, 4.0, and 6.0, respec-
tively, with a scan rate of 125 mV/s.
the EC/ESI-MS experiments based on Scheme 1. As is seen in
Figure 4, the reduced and oxidized dimers were more easily seen
at high scan rates (e.g., 250 mV/s) due to the fact that there was
less time for the oxidized dimer to diffuse away from the electrode
surface in these cases. Cyclic voltammograms recorded for the
background electrolyte solution did not contain any oxidation or
reduction peaks in the relevant potential window. The overall 4-CA
oxidation behavior is consequently in good agreement with that
previously described in the literature.1,2,12
Figure 5 shows cyclic voltammograms recorded in Britton-
Robinson buffers at pH 2.0, 4.0, and 6.0, respectively. (No peaks
were seen in the cyclic voltammograms recorded in the corre-
sponding background electrolyte solutions.) It is seen that 4-CA
was electroactive and that the oxidation of 4-CA was diffusion
controlled at all three pH values, i.e., at pH values above and below
the pKa value (i.e., 4.0). The shift in the peak potential for the
oxidation of 4-CA, which shows that reaction 1 in Scheme 1 is
pH dependent, is likewise in agreement with the presence of
protonated 4-CA at pH 2.0 and deprotonated 4-CA at pH 6.0 as
well as in agreement with the smaller negative shift between pH
4.0 and 6.0 than between pH 2.0 and 4.0. More importantly, Figure
5 clearly demonstrates that the rate of the formation of the
oxidized dimer decreases with increasing pH. The negative shift
in the peak potentials for the reduction of the oxidized dimer and
the oxidation of the reduced dimer for increasing pH demonstrate
that protons are involved in these steps in agreement with previous
(23) Baird, R. S.; Bunge, A. L.; Noble, R. D. AIChE J. 1987, 33, 43–53
(24) Ukai, A.; Hirota, N.; Terazima, M. Chem. Phys. Lett. 2000, 319, 427–433
.
.
Analytical Chemistry, Vol. 81, No. 13, July 1, 2009 5185

Scheme 2. Proposed Comproportionation Reaction 4 Generating the Reduced Dimer, DimRed, with m/z 219.2 and
Two 4-CA Radical Cations as a Result of the Reaction between the Oxidized Dimer and 4-CA
Scheme 1 with both cells. This means that the rate of the
formation of the oxidized dimer must have been low enough to
allow the detection of this intermediate.
presence of the oxidized and reduced dimer in the 4-CA solution
in the first experiment above.
As higher intensities were obtained for the intermediate with
the modified cell while the intensities for the oxidized dimer were
about the same for the two cells, the results likewise indicate that
the rate of reaction 2b was lower with the modified cell. This is
in good agreement with the assumption that the pH was lower in
the modified cell as this would disfavor the formation of the
oxidized dimer via reaction 2b. Significant amounts of the reduced
dimer can only be expected to form once the oxidized dimer is
present in the flow cell at a sufficiently high concentration as the
formation of the reduced dimer relies on the presence of both
oxidized dimer and protonated 4-CA. This would explain the
observation that the reduced dimer appears at more positive
potentials than those needed to generate the oxidized dimer.
Since reaction 4 involves the protonated form of 4-CA, the
formation of the reduced dimer can also be expected to benefit
from a decreased pH. As water is expected to be oxidized at
potentials more positive than about +1.4 V, the local pH at the
working electrode should decrease at these potentials for both
flow cells, which can be expected to influence the rate of the
comproportionation reaction. (The local pH at the working
electrode should also be decreased as a result of the oxidation of
the protonated 4-CA according to reaction 1.) The observed
difference in the rate of the generation of the reduced dimer (i.e.,
m/z 219.2) can then be explained on the basis of pH differences
between the solutions in the cells due to different cell configura-
tions. The presence of a comproportionation reaction is, however,
not completely unexpected as we have recently shown13-15 that
such reactions can take place with the present flow cells and as
comproportionation has been shown25,26 to explain the so-called
nucleation loop present during the electropolymerization of
conducting polymers such as thiophenes. Based on previous
results,8 it is also reasonable to assume that comproportionation
reactions likewise are present during the oxidation of aniline. In
addition, it has been shown25,26 that autocatalytic processes
involving homogeneous comproportionation reactions take place
during the oxidation of benzene, pyrrole, and carbazoles, and it
was, therefore, concluded that this is a general phenomenon
characteristic for many chain-forming electropolymerization reac-
tions. While the latter authors based their conclusion on voltam-
metric results in combination with results obtained by simulations,
our experimental results include direct evidence for the formation
of a species (i.e., the reduced dimer) by comproportionation.
The presence of the comproportionation reaction was also
confirmed by carrying out experiments in which a solution of 50
µM 4-CA was oxidized in the (modified) flow cell at a potential of
+1.2 V and a flow rate of 10 µL/min while the oxidized solution
was collected in vials at the outlet of the cell. To one of the vials
(which all should contain 4-CA, oxidized dimer, and reduced
dimer), HCl was added to yield a pH of 1.0. An identical volume
of Milli-Q water was added to another vial to ensure that the
dilution of the concentrations in the two vials was the same. The
two vials were then left to stand for 30 min after which the ratio
between the signals at m/z 219 (the reduced dimer) and 217 (the
oxidized dimer) was calculated on the basis of the signals obtained
when injecting the solutions directly into the ESI-MS system (i.e.,
in the absence of the EC cell). The corresponding ratio was also
calculated for the case of direct injection of 4-CA solution that
had not been oxidized in the flow cell into the ESI-MS system.
The obtained results show that the ratio between the m/z 219
and 217 signals were significantly higher (i.e., 1.64 and 0.98) for
the oxidized solutions than for the 4-CA solution (ratio 0.66) and
that the highest ratio (1.64) was found for the pH 1.0 solution. It
should be pointed out that these ratios were not corrected for
isotopic effects and that the contribution from the oxidized dimer
(m/z 217) to the signal at m/z 219 due to the isotope effect should
be approximately 32%. The results, nevertheless, clearly show that
the oxidation of 4-CA gives rise to the formation of both the
oxidized and reduced dimer and that the rate of the formation of
the reduced dimer in a solution containing the oxidized dimer
and 4-CA increases when the pH is decreased. This latter is in
good agreement with the proposed comproportionation reaction
(without which no reduced dimer should be formed).
In another experiment, a sample of the same 50 µM 4-CA
solution, as in the experiment above, was injected into the ESI-
MS instrument after allowing the solution to stand overnight. It
was then found that the ratio between the signals at m/z 219 (the
reduced dimer) and 217 (the oxidized dimer) was significantly
higher compared to the ratio obtained shortly after the 4-CA
solution was prepared (the ratios were 0.80 and 0.66, respectively).
As 4-CA is easily oxidized by oxygen, the oxidized dimer can be
expected to form, but in the absence of the comproportionation
reaction, the reduced dimer should not be formed under oxidizing
conditions. These results likewise show that the reduced dimer
is indeed formed as a result of the formation of the oxidized dimer
in agreement with the presence of the proposed comproportion-
ation reaction. The oxidation of 4-CA by oxygen also explains the
(25) Heinze, J.; Rasche, A.; Pagels, M.; Geschke, B. J. Phys. B 2007, 111, 989–
997
(26) Dietrich, M.; Heinze, J.; Krieger, C.; Neugebauer, F. A. J. Am. Chem. Soc.
1996, 118, 5020–5030
.
.
5186 Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

As the comproportionation should give rise to a loss of oxidized
dimer and a generation of reduced dimer, it can be expected that
the shape of the voltammogram also should be affected by this
effect. There was, however, no sign of such an effect in our
voltammetric data as the ratio between the peak heights for the
oxidation of the reduced dimer and the reduction of the oxidized
dimer was always close to unity for all scan rates, even at pH 2.0.
This is, however, not surprising as the uncertainties in these ratios
were relatively large and as the amount of reduced dimer formed
via the comproportionation should be much smaller than the
amount formed as a result of the reduction of the (remaining)
oxidized dimer. The fact that the formation of the reduced dimer
was evident in the EC/ESI-MS experiments clearly demonstrates
that the EC/ESI-MS approach is a very good complement to
voltammetric investigations of the 4-CA system. As is evident from
a comparison of the results obtained with our on-line EC/ESI-
MS approach and those previously obtained3 when one employs
off-line coulometric oxidation of 4-CA prior to ESI-MS, the obtained
oxidation products depend on the experimental conditions. These
two approaches, thus, provide complementary information. Our
on-line approach is clearly better suited for the detection of
intermediates and chemical follow-up reactions such as reaction
4 while the off-line coulometric method, which allows much longer
oxidation times to be used, favors studies of reactions taking place
on a longer time domain in solutions containing high concentra-
tions of oxidation products. The presence of different reactions
on different time scales was in fact recently demonstrated12
experimentally as it was found that both fast radical-radical
reactions (detected by cyclic voltammetry) and slower radical-4-
CA reactions (detected by ESR) take place upon the oxidation of
4-CA. Another difference between our experimental conditions and
those previously used is that the concentration of 4-CA used in
our experiments (i.e., 5 µM) was significantly lower than those
generally used in previous studies.1-3,7,12,27 This should affect the
rates of the reactions 2a and 4 in Schemes 1 and 2, respectively,
and the detection of the intermediate in our experiments was most
likely facilitated by this effect. EC/ESI-MS is, therefore, a very
powerful tool in the study of the present type of systems involving
electrochemical steps coupled to chemical follow-up reactions
involving dimerization reactions.
formed upon the oxidation of 4-CA with delay times between the
electrochemical cell and the ion source of about 3 s. On the basis
of the results obtained with on-line EC/ESI-MS and off-line
voltammetry, it can be concluded that the proposed reaction
schemes need to be complemented with a previously unrecog-
nized comproportionation reaction. This comproportionation reac-
tion, which involves 4-CA and the oxidized dimer (i.e., 4-[(4-
chlorophenyl)imino]-2,5-cyclohexadien-1-imine) formed upon
dimerization of the radical cations of 4-CA, gives rise to the
reduced dimer (i.e., 4-amino-4′-chlorodiphenylamine). Analogous
comproportionation reactions can also be expected to be present
during the oxidation of other similar molecules such as in aniline,
pyrrole, and thiophene. The present results, therefore, provide
additional support for the hypothesis that the so-called nucleation
behavior seen during the electropolymerization of, e.g., thiophenes
in fact is due to comproportionation reactions and that this
behavior is typical for many systems involving chain-forming
electropolymerization reactions. In addition to the oxidized and
reduced dimers, a peak corresponding to a previously proposed
but undetected intermediate was also observed in the present
mass spectra. It is concluded that electrochemical flow cells with
conversion efficiencies lower than 100% under mass transport
controlled conditions are generally better suited for studies of
these kinds of reaction pathways than coulometric flow cells. The
results likewise show that different results can be obtained with
different flow cells in which the influence of the counter electrode
reactions varies and that this can be used to obtain additional
information about the reaction pathway. On-line EC/ESI-MS with
flow cells of different designs consequently constitute a very
valuable tool for the study of the oxidation of 4-CA in particular
and of electropolymerization reactions in general.
ACKNOWLEDGMENT
We thank The Swedish Research Council (Grant 621-2005-
5493) for financial support.
SUPPORTING INFORMATION AVAILABLE
Information regarding the experimental conditions, isotope
corrected mass spectra intensities, and average peak intensities
as a function of the potential for the two different cell setups, as
well as cyclic voltammograms recorded with different scan rates
at pH 2.0. This material is available free of charge via the Internet
at http://pubs.acs.org.
CONCLUSIONS
The present results clearly demonstrate that on-line EC/ESI-
MS enables straightforward identification of the oxidation products
Received for review December 4, 2008. Accepted May 18,
2009.
(27) Ka´da´r, M.; Taka´ts, Z.; Karancsi, T.; Farsang, G. Electroanalysis 1999, 11,
809–813
.
AC802563F
Analytical Chemistry, Vol. 81, No. 13, July 1, 2009 5187