You are on page 1of 7

J. Phys. Chem.

C 2007, 111, 5805-5811 5805

Oxidation Reactions of a Series of Benzidines: Electrochemical Detection of Dimerization of


Cation Radicals and Examples of Potential Inversion Caused by Very Small Steric Effects

Norma A. Macı́as-Ruvalcaba and Dennis H. Evans*


Department of Chemistry, UniVersity of Arizona, Tucson, Arizona 85721
ReceiVed: January 30, 2007; In Final Form: February 21, 2007

It has been previously reported that cation radicals of some benzidines undergo dimerization. In the present
work dimerization equilibrium constants have been determined by UV-visible absorption spectroscopy for
the cation radicals of benzidines 1a-d. This dimerization reaction was shown to produce subtle but significant
effects in cyclic voltammograms recorded for oxidation of the benzidines. It is shown that cyclic voltammetry
is a simple and rapid technique to detect such reactions, whose quantitative evaluation must, however, be
accomplished by other means. The oxidation of certain N-peralkylated benzidines having substituents in the
3 and 3′ positions, 1e-i, occurs in a single, two-electron oxidation process due to the fact that the two standard
potentials for oxidation to the cation radical and dication, respectively, are very similar. In several cases,
potential inversion occurs. That is, the potential for removal of the second electron is less positive than that
for removal of the first. Causes of this potential inversion are discussed.

1. Introduction CHART 1
Benzidines, 1, are prime examples of simple redox systems
which, in nonaqueous media, undergo stepwise oxidation to
form the cation radical and dication forms.1,2 In a recent study
of o-dianisidine (1a, R ) R2 ) H; R1 ) OCH3), the two standard
potentials for the cation radical/neutral couple (E°1) and for the
dication/cation radical couple (E°2) were found to show normal
ordering of potentials with E°1 - E°2 ≈ -0.2 V, i.e., it was
more difficult to remove the second electron compared to the
first. By contrast, the N-permethylated derivative (1e: R ) CH3;
R1 ) OCH3; R2 ) H) displayed a single, two-electron oxidation
peak, i.e., the potentials were either inverted or strongly
compressed.3 The ethyl derivative (1f: R ) C2H5; R1 ) OCH3;
R2 ) H) also showed a single oxidation peak but this was
interpreted as being due to one-electron oxidation.3
In that work it was reported that oxidation of o-dianisidine
(1a: R ) R2 ) H; R1 ) OCH3) at the first oxidation peak in
dichloromethane resulted in the formation of the cation radical
perchlorate salt.3 A similar salt has been isolated from 3,3′,5,5′-
tetramethylbenzidine (1d) and it is described as the salt of a
dimer of the cation radical, which is a singlet (with a low-
lying triplet state). This dimer of the cation radical, A22+, where
A is the benzidine, was also shown to exist in solution.4
The dimer is thought to be a π-bonded species with a stacking
interaction between the two biphenyl ring systems. Thus,
oxidation of compounds such as o-dianisidine (1a) or 3,3′,5,5′-
tetramethylbenzidine (1d) should produce a mixture of mono-
meric cation radicals and their dimers. Will this weak inter-
action of the product of the electrode reaction cause discer-
nible changes in the current-voltage response? In this work
we have shown for the first time that the rapid, reversible
π-dimerization of the cation radicals of five different benzidines
can be readily detected by cyclic voltammetry and values of
the dimerization equilibrium constant can be extracted from the Theory for cyclic voltammetry with a fully reversible
data. dimerization following electron transfer has previously been
presented and discussed5,6 but it was more convenient for us to
* To whom correspondence should be addressed. E-mail: dhevans@ analyze our data by digital simulation. Another type of associa-
email.arizona.edu. tion reaction of cation radicals has been reported, especially
10.1021/jp070801p CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/24/2007
5806 J. Phys. Chem. C, Vol. 111, No. 15, 2007 Macı́as-Ruvalcaba and Evans

for cation radicals of aromatic hydrocarbons. This reaction is which was washed with dichloromethane, was obtained in 76.2%
the association of the cation radical with the neutral starting yield. mp >400 °C dec. Anal. Calcd for C12H10Cl3N2O4: C
material to form what is often called a “cation radical dimer”, 40.88; H 2.86; Cl 30.17; N 7.95; O 18.15. Found: C 39.74; H
A2•+.7 These “dimers” are often isolated as insoluble salts but 3.24; Cl 30.90; N 7.30; O 20.31.
they are also present in solution.7h We are not aware of examples The cation radical perchlorate of 1b (1b•+ClO4-) was
of the detection of A2•+ by voltammetric wave shape analysis, prepared as follows. o-Tolidine (1b) (150 mg) was electrolyzed
but hope to investigate the matter in future work. at 0.4 V vs Ag/AgNO3 under the same experimental conditions
The majority of cases where reversible dimerization has been as described above. A dark blue solid was obtained in 65.3%
shown to follow electron transfer are those in which σ-bonded yield. mp >400 °C dec. Anal. Calcd for C14H16ClN2O4: C
dimers are formed.8 In most such cases, the dimer can be 53.94; H 5.17; Cl 11.37; N 8.99; O 20.53. Found: C 54.12; H
detected, under appropriate conditions, through its separate 5.09; Cl 11.30; N, 8.98; O 20.34.
oxidation or reduction peak whose magnitude can ideally give The cation radical perchlorate salts of 1a (1a•+ClO4-) and
information about the dimerization equilibrium and rate con- 1d (1d•+ClO4-) were prepared following the same procedure
stants. In the case of the π-bonded dimers described in this work, as described above for 1b•+ClO4- and 1c•+ClO4-. The prepara-
the formation and dissociation of the dimers is so rapid that tion and identity of these radicals has been previously reported
separate voltammetric peaks for the dimer cannot be detected. by Douadi et al.3 and Awano and Ohigashi,11 respectively.
Instead, the fact that a dimer is formed is signaled via subtle 2.2. Electrochemical Cells, Electrodes, and Instrumenta-
changes in the voltammetric peak shape. tion. In general, this equipment was the same as described
As stated earlier, the N-permethylated derivative of o- earlier.10 The working electrode was a 0.3-cm diameter glassy
dianisidine (1e: R ) CH3; R1 ) OCH3; R2 ) H) shows a single carbon electrode whose area was determined to be 0.0814 cm2.
two-electron oxidation peak indicating that the potentials are The reference electrode was a silver wire immersed in 0.10 M
strongly compressed or perhaps inverted with E°1 - E°2 > 0.9 Bu4NPF6/0.010 M AgNO3 in acetonitrile. The potential of this
In the present work we have determined the extent of inversion reference electrode was periodically measured with respect to
for five substituted benzidines and have shown that the structural the reversible ferrocene/ferrocenium potential in acetonitrile, and
changes that underlie the inversion are quite subtle and depend all potentials reported in this work are with respect to ferrocene.
on very minor steric factors. The temperature of the jacketed cell was controlled with a
In this paper we report our first results on fast, reversible circulating bath. The reference electrode was at room temper-
dimerization reactions. Studies of other systems will be reported ature (nonisothermal operation). Voltammograms with only
later. solvent and supporting electrolyte were recorded and subtracted
from the voltammograms of the compounds to obtain background-
2. Experimental Section corrected data. All voltammetric experiments were conducted
in acetonitrile containing 0.10 M Bu4NPF6. The determination
2.1. Chemicals and Reagents. The solvent was acetonitrile of the magnitude of the solution resistance and its compensation
and the electrolyte was tetrabutylammonium hexafluorophos- were carried out as described.10 The resistance compensation
phate (Bu4NPF6). Sources and treatment of solvent and elec- was achieved as follows (total solution resistance, resistance
trolyte have been described.10 For preparation of the cation electronically compensated, resistance used in simulations): 25
radical salts, dichloromethane (99.8%, anhydrous (<0.001% °C (140 Ω, 120 Ω, 20 Ω), 0 °C (190 Ω, 130 Ω, 60 Ω), -18
water)) from Aldrich was used as received. °C (220 Ω, 200 Ω, 20 Ω).
3,3′,5,5′-Tetramethylbenzidine (1d) and o-tolidine (1b) (Al- Variation of the scan rate, V, was done in such a way that
drich) were recrystallized from benzene. N,N,N′,N′-Tetrameth- there was an approximately linear variation in log V. Thus, for
ylbenzidine (1j) (Aldrich) was recrystallized from methanol. scan rates between 0.1 and 30 V/s, the values chosen were 0.1,
o-Dianisidine (1a) (Aldrich) was twice recrystallized from water. 0.2, 0.3, 0.5, 1, 2, 3, 5, 10, 20, and 30 V/s (log V ) -1, -0.699,
N,N,N′,N′-Tetraphenylbenzidine (1k) (Aldrich) was used as -0.523, -0.301, 0, 0.301, 0.477, 0.699, 1, 1.301, 1.477).
received. 3,3′-Dichlorobenzidine (1c) was obtained by treatment Controlled potential coulometry was conducted as described
of an ether solution of the corresponding 3,3′-dichlorobenzidine earlier.12
dihydrochloride (Aldrich) with 60% aqueous NaOH followed 2.3. Determination of the Dimerization Equilibrium Con-
by evaporation of the organic phase. 4,4′-Bis(dimethylamino)- stant for the Cation Radicals of 1a-d. Measurements were
3,3′-dimethoxybiphenyl (1e), 4,4′-bis(diethylamino)-3,3′-dimethox- performed with a Spectral Instruments, Inc., CCD array UV-
ybiphenyl (1f), 4,4′-bis(dimethylamino)-3,3′-dimethylbiphenyl vis spectrophotometer equipped with a fiber-optic dip cell with
(1g), 4,4′-bis(dimethylamino)-3,3′,5,5′-tetramethylbiphenyl (1h), a path length of 1.03 ( 0.03 cm. The absorption measurements
and 4,4′-bis(diethylamino)-3,3′,5,5′-tetramethylbiphenyl (1i) were made at the λmax of the dimer (646-715 nm), in 0.10 M
were prepared by N-peralkylation of the corresponding ben- Bu4NPF6, Omnisolve acetonitrile (EMD) at 25, 0, and -18 °C.
zidines with dimethyl sulfate or diethyl sulfate according to a Equation 1 was fit to the apparent molar absorptivities,  ( )
procedure described in the literature.3 A/C where A is the observed absorbance at total concentration
The cation radical perchlorate of 1c (1c•+ClO4-) was prepared of cation radical, C), as a function of C.
as follows. 3,3′-Dichlorobenzidine (1c) (61.3 mg) in 50 mL of

[ ( )]
0.1 M Bu4NClO4 in dichloromethane was electrolyzed at 0.92
-0.5 + x0.25 + 2KdimC
V vs Ag/AgNO3. (Caution: Organic perchlorate salts cause  ) 0.5 - D +
irritation and may be harmful if swallowed. Contact with 4KdimC

[ ]
combustible materials, flammable materials, or powdered metals -0.5 + x0.25 + 2KdimC
can cause fire or explosion.) During electrolysis a dark deposit M (1)
was formed on the platinum gauze working electrode. The 2KdimC
consumption of charge corresponded to 1 F/mol. The dark solid
was dissolved in acetonitrile giving a green solution, the solvent Here Kdim is the dimerization equilibrium constant, D is the
was evaporated at 35 °C under vacuum, and a purple solid, molar absorptivity of the dimer, and M is the molar absorptivity
Oxidation Reactions of a Series of Benzidines J. Phys. Chem. C, Vol. 111, No. 15, 2007 5807

Figure 2. Spectra of 1.15 × 10-4 M o-dianisidine cation radical


perchlorate (1a•+ClO4-) in acetonitrile as a function of temperature:
-18, -15, -10, -5, 0, 5, 10, 15, 20, 25, 30, and 35 °C.

Figure 1. Cyclic voltammogram of 2.00 mM 3,3′-dichlorobenzidine,


1c, at 1.00 V/s and 25 °C (full curve). (A) Points: Digital simulation
based on simple, stepwise two-electron oxidation. Simulation parameter
values: E°1 ) 0.368 V, R1 ) 0.50, ks,1 ) 104 cm/s (reversible limit);
E°2 ) 0.599 V, R2 ) 0.50, ks,2 ) 0.42 cm/s; Kdisp ) 1.2 × 10-4, kf,disp
) 380 M-1 s-1, kb,disp ) 3.1 × 106 M-1 s-1; DA ) 1.89 × 10-5 cm2
s-1; DA+ ) DA2+ ) 1.15 × 10-5 cm2 s-1. (B) Points: Digital simulation
based on stepwise two-electron oxidation with fast, reversible dimer-
ization of the cation radical. Simulation parameter values: E°1 ) 0.406
V, R1 ) 0.50, ks,1 ) 0.16 cm/s; E°2 ) 0.559 V, R2 ) 0.70, ks,2 ) 0.23
cm/s; Kdisp ) 2.7 × 10-3, kf,disp ) 1.0 × 105 M-1 s-1, kb,disp ) 4.0 ×
107 M-1 s-1; Kdim ) 6.5 × 103 M-1, kf,dim ) 1.0 × 109 M-1 s-1; kb,dim
) 1.5 × 105 s-1; DA ) 1.70 × 10-5 cm2 s-1; DA+ ) DA2+ ) DAA2+ )
1.10 × 10-5 cm2 s-1. Slow, irreversible decomposition of dication
included with kf ) 0.17 s-1.

of the monomer cation radical. This equation was derived for Figure 3. Apparent molar absorptivity (absorbance/total concentration
of 1a•+ClO4-) at 715 nm for various concentrations of cation radical
the dimerization reaction in a manner analogous to that presented for three different temperatures (filled circles). Full curves: Fits of the
earlier for neutral/dication association.4c data according to eq 1 with molar absorptivities as shown and Kdim )
Owing to a slow decomposition of the cation radicals, the 1.76 × 103, 1.37 × 104, and 8.44 × 104 M-1 for 25, 0, and -18 °C,
solutions were prepared in the following manner. Stock solutions respectively.
of the cation radicals were prepared and maintained at ca. -20
°C. For experiments at -18 and 0 °C, increments of the stock oxidation of the benzidine to its cation radical and dication,
solution were added to acetonitrile at the indicated temperature according to reactions 2 and 3 (A is the benzidine, 1c).
to prepare solutions with increasing concentrations of cation
radical. Spectra were obtained quickly after each addition. For A•+ + e- h A E°1, ks,1, R1 (2)
experiments at 25 °C, separate solutions were prepared from
the -20 °C stock solutions and acetonitrile maintained at 25 A2+ + e- h A•+ E°2, ks,2, R2 (3)
°C and the spectra were obtained as quickly as possible after
each solution preparation. 2A•+ h A + A2+ Kdisp, kf,disp, kb,disp (4)
2.4. Calculations. Digital simulations were conducted with
DigiElch, version 2.0, a free software package for the Digital The points in Figure 1A are from a best-fit simulation of the
simulation of common Electrochemical experiments (http:// voltammogram, using reactions 2 and 3 along with dispropor-
www.digielch.de).13 The fitting routine in that program was used tionaton reaction 4. The simulation parameters, given in the
to establish the final best-fit parameter values for many of the figure caption, correspond to reversible electron transfers and
variables. a rapid, reversible disproportionation reaction. Small improve-
Complete geometry optimization and frequency calculations ments in the fit can be achieved by invoking unequal diffusion
were performed according to the density functional theory coefficients for the three species in the mechanism but the fit
(DFT), using the B3LYP/6-31G(d,p) level with the Gaussian can never be made perfect.
03 program.14 For each structure it was confirmed that there Although a reasonable fit between simulation and experiment
were no imaginary frequencies. For radicals, the corresponding was obtained in Figure 1A, some important discrepancies are
unrestricted (UB3LYP) method was used. apparent. The experimental current near the foot of peak Ia rises
much more sharply than predicted by the simulation (points).
3. Results and Discussion
In addition, the experimental current is less than that simulated
3.1. Detection of Dimerization of Cation Radicals by between peaks Ia and IIa and, in general, peaks IIa, IIc, and Ic
Voltammetric Peak-Shape Analysis. Figure 1A is a voltam- are somewhat sharper than indicated by the simulation. As
mogram of 2.00 mM 3,3′-dichlorobenzidine, 1c, in acetonitrile. mentioned in the introduction, dimerization of the cation radicals
Apparently the process is the simple, stepwise, two-electron of benzidines has been detected. To determine if such a reaction
5808 J. Phys. Chem. C, Vol. 111, No. 15, 2007 Macı́as-Ruvalcaba and Evans

TABLE 1: Equilibrium Constants for the Dimerization Reaction of the Cation Radicals of Benzidines 1a-da
Kdim/M-1
compd 25 °C 0 °C -18 °C -18 °Cb D /M-1cm-1 M /M-1cm-1 λ/nm
1a•+ClO4 1.76 × 103 1.37 × 104 8.44 × 104 6.17 × 104 1.73 × 104 500 715
1b•+ClO4 6.10 × 103 1.10 × 105 5.94 × 105 4.19 × 105 4.00 × 104 ∼0 646
1c•+ClO4 6.50 × 103 4.07 × 104 1.15 × 105 9.20 × 104 3.60 × 104 ∼0 660
1d•+ClO4 5.37 × 103 1.19 × 104 3.22 × 104 2.72 × 104 3.56 × 104 160 661
a
Determined by UV-vis spectroscopy in 0.1 M Bu4NPF6 in acetonitrile. See the Experimental Section. Three-point van’t Hoff plots give ∆H°dim
) -13.5, -16, -10, and -6 kcal/mol for 1a-d•+, respectively. Estimated indeterminate error in equilibrium constants is (10%. b Determined in
the absence of Bu4NPF6.

will affect the simulation, it was added as a fast, reversible TABLE 2: Results of Controlled Potential Coulometry for
Benzidines, 1a-h and 1ja
dimerization and the result is shown in Figure 1B. Clearly there
is a profound improvement in the quality of the fit. It should color of oxidized
compd Eox/V n Ered/V % recovery solution
be emphasized that the only new reaction in the mechanism
used in Figure 1B compared to that in Figure 1A is the 1a 0.17 0.97 -0.10 40 olive green
dimerization, reaction 5. This mechanism, with the parameters 1b 0.22 0.93 -0.10 80 bright green
1c 0.50 0.96 0.20 43 dark green
listed in the caption of Figure 1, was found to provide adequate 1d 0.16 0.97 -0.10 95 bright green
fits for scan rates from 0.10 to 30 V/s. 1e 0.27 1.80 -0.30 22 dark olive green
1f 0.27 1.90 -0.20 77 dark purple
2A•+ h A22+ Kdim, kf,dim, kb,dim (5) 1g 0.34 1.40 -0.10 24 dark red
1h 0.43 1.53 ND ND red-orange
1j 0.15 0.99 -0.10 83 dark green
The value of Kdim used in the simulation in Figure 1B was a
Electrolyses conducted at ambient temperature. Eox and Ered are
6.50 × 103 M-1, a value determined by spectrophotometric the control potentials (vs ferrocene) for oxidation and subsequent
analysis (see below). When Kdim was allowed to vary to achieve reduction, respectively. n (Faradays per mole of reactant ) electrons
the best fit between simulation and experiment, the result was per molecule of reactant) for the oxidation. % recovery is the charge
Kdim ) 2.1 × 104 M-1. This rather large discrepancy is not recovered during the reductive electrolysis compared to charge passed
unexpected in view of the relatively small adjustments in the in the preceding oxidation. Only in the cases of 1d and 1j was the
shape of the voltammogram that are controlled by the magnitude original color restored upon reduction following oxidation. ND ) not
determined.
of Kdim. The spectrophotometric determination of Kdim is a more
direct measurement based, as it is, on a specific absorption band 3.3. Controlled Potential Coulometry. As the voltammetric
due almost entirely to the dimer. For this reason, Kdim was results for 1a-d will be analyzed by a model involving initial
evaluated by spectroscopy and the values so obtained were later one-electron oxidation followed by reversible dimerization of
used in the analysis of the voltammograms. the cation radicals, controlled potential coulometry was carried
3.2. Determination of the Dimerization Equilibrium Con- out at the first anodic peak to confirm that the overall reaction
stants for Benzidines 1a-d. Temperature-dependent absorption was a one-electron process. The results are summarized in Table
spectra of the cation radical of o-dianisidine, 1a•+, are shown 2. For compounds 1a-d, the oxidation requires one electron
in Figure 2. The absorption band near 700 nm, due almost within experimental error, but only in the cases of 1b and 1d
entirely to the dimer, is seen to grow from almost nil at 35 °C was the amount of charge recovered upon subsequent reduction
to prominence at -18 °C. The equilibrium constant for close to the oxidative charge. Loss of the cation radical was
dimerization was measured at 25, 0, and -18 °C from studies confirmed by cyclic voltammetric studies of the solution after
of the absorbance at 715 nm as a function of the concentration oxidative electrolysis. So, it would appear that the initial
of the cation radical, as described in the Experimental Section. oxidation reaction is formation of the cation radical, which is
The experimental results, along with fits according to eq 1, are followed in some cases by slower decomposition reactions.
shown in Figure 3 and the derived values of Kdim, D, and M Compounds 1e-h show a single two-electron oxidation
are shown in Table 1 along with results for the cation radicals process (see below). Electrolysis at potentials positive of the
of 1b-d. Spectra for the cation radicals of 1b,c (see ref 4c for oxidation peak requires 1.4-1.9 electrons/molecule and the
spectra of 1d•+) and plots like Figure 3 are given in the smaller percent charge recovered in the subsequent reductive
Supporting Information. electrolysis indicates that the dications are less stable than the
As can be seen in Table 1, the λmax for the dimer falls in the cation radicals (Table 2). Compound 1j undergoes stepwise
narrow range of 646-715 nm and the band is entirely due to oxidation and electrolysis at a potential between the two
absorption by the dimer for 1b and 1c (M ≈ 0) while a minor oxidation peaks requires one electron per molecule and the
contribution from the monomer is found for 1a and 1d. The subsequent reductive electrolysis indicates good recovery of the
value of Kdim for 1d•+ found at 25 °C, 5370 M-1, is in reasonable charge. Voltammograms of the solution after oxidative elec-
agreement with that reported (8700 M-1; recalculated according trolysis show almost complete formation of the cation radical.
to eq 1) earlier4c but the value of ∆H° reported there (for a Due to limited solubility, electrolysis of 1k was not attempted.
different range of temperatures) is more than twice that found 3.4. Fits of Voltammograms of 1a-d with Dimerization
in the present work. Equilibrium Constants Determined by Spectroscopy. Ex-
At -18 °C, values of Kdim were obtained in the absence and amples of fits of simulation to experimental voltammograms
presence of 0.10 M Bu4NPF6 (Table 1). The presence of the for 1a,b,d may be found in Figures 4-6. Each of these figures
electrolyte resulted in slightly smaller Kdim (15-30% lower). contains the experimental voltammogram (solid curve), the best-
For consistency, the values in the absence of electrolyte were fit simulation including dimerization of the cation radical (open
used in simulations of the voltammetric data at the three different circles) and the best-fit simulation without dimerization (dashed
temperatures. curve). (Data for 1c were shown in Figure 1.) The best-fit values
Oxidation Reactions of a Series of Benzidines J. Phys. Chem. C, Vol. 111, No. 15, 2007 5809

Figure 4. Voltammogram of 1.30 × 10-3 M o-dianisidine, 1a, at 1.00 Figure 6. Voltammogram of 1.18 × 10-3 M 3,3′,5,5′-tetramethylben-
V/s and 0 °C (full curve). Best fit simulation for two-step oxidation zidine, 1d, at 1.00 V/s and 0 °C (full curve). Best fit simulation for
without dimerization of the cation radical (dashed). Best fit simulation two-step oxidation without dimerization of the cation radical (dashed).
for two-step oxidation with fast, reversible dimerization of the cation Best fit simulation for two-step oxidation with fast, reversible dimer-
radical with simulation parameter values as summarized in Table 3 (a ization of the cation radical with simulation parameter values as
full list of parameter values is given in the Supporting Information) summarized in Table 3 (afull list of parameter values is given in the
(open circles). Supporting Information) (open circles).

TABLE 3: Summary of the Simulation Parameter Values


for the Reversible Dimerization of Benzidines 1a-d and 1j
at Three Different Temperaturesa
T/ E°1/ E°2/ E°1 - E°2/ Kdim/ kf,dim/ 106D/
compd °C V V V M-1 M-1 s-1 cm2 s-1
1a 25 0.092 0.241 -0.174 1.5 × 103 7.0 × 107 14.9
0 0.087 0.225 -0.204 1.4 × 104 6.0 × 108 10.8
-18 0.088 0.213 -0.223 8.4 × 104 1.7 × 109 8.6
1b 25 0.136 0.304 -0.169 6.1 × 103 1.3 × 108 14.4
0 0.124 0.278 -0.154 1.1 × 105 1.6 × 109 10.4
-18 0.127 0.269 -0.142 5.9 × 105 2.3 × 1010 8.7
1c 25 0.408 0.559 -0.149 6.5 × 103 1.0 × 109 15.8
0 0.392 0.548 -0.138 4.1 × 104 7.0 × 109 11.7
-18 0.380 0.542 -0.126 1.2 × 105 5.0 × 109 8.8
1d 25 0.080 0.254 -0.151 5.4 × 103 8.0 × 108 14.3
0 0.047 0.251 -0.156 1.2 × 104 2.8 × 109 10.5
Figure 5. Voltammogram of 3.52 × 10-3 M o-tolidine, 1b, at 1.00 -18 0.021 0.244 -0.162 3.2 × 104 3.0 × 109 7.1
V/s and -18 °C (full curve). Best fit simulation for two-step oxidation 1j 25 0.036 0.228 -0.192 0 15.7
without dimerization of the cation radical (dashed). Best fit simulation 0 0.027 0.212 -0.185 1.3 × 102 1.9 × 108 12.9
for two-step oxidation with fast, reversible dimerization of the cation -18 0.022 0.219 -0.197 2.1 × 102 4.4 × 108 9.90
radical with simulation parameter values as summarized in Table 3 (a a
full list of parameter values is given in the Supporting Information) E°1 and E°2 and D correspond to the average values obtained by
(open circles). simulation for sets of data obtained at two different concentrations of
the corresponding benzidine. Kdim are the values obtained by spectro-
photometric measurement and kf,dim correspond to the values that give
of the principal parameters for the dimerization mechanism good fits to the experimental voltammograms obtained at scan rates
are summarized in Table 3. For simulations without dimeriza- from 0.1 to 30.0 V/s and at two different concentrations. D is the
tion, the dimerization reaction was simply removed from the common diffusion coefficient for A, A•+, and A2+. See the Supporting
simulation and the potentials were readjusted to obtain best Information for DAA2+. Estimated indeterminate error for rate and
fit. Examples of additional simulations, at various tempera- equilibrium constants is (10% and (5 mV for potetntials.
tures, scan rates, and concentrations are given in the Suppor- scarcely affects the voltammogram. An example is given in
ting Information along with complete tables of simulation Figure 7 for 1j where KdimC is estimated to be only 0.7 at -18
parameters. °C. The fit with dimerization included (circles) is slightly
Figures 4-6 show that the fits with dimerization are definitely superior to that without dimerization (dashed) but there is no
superior to those without. Both fits, at any given temperature doubt that it would be impossible to conclude that dimerization
and concentration, were based on the parameter values that was occurring based solely on the voltammetric peak shape.
provided the best fits for scan rates from 0.1 to 30 V/s. That is, 3.5. Examples of Potential Inversion. Those N-peralkylated
for any given scan rate (e.g., those shown in Figures 4-6) benzidines that also contain substituents in the 3 and 3′ positions
somewhat improved fits might be obtained by further adjustment behave differently from benzidines containing NH2 groups.
of parameter values but such adjustments would give poorer Figures 8 and 9 are examples of voltammograms for 1f and 1g,
agreement at other scan rates. respectively. In each case, the two one-electron oxidation
For dimerization equilibrium, the fraction present as monomer processes seen with other benzidines are replaced by a single
is 0.5 when KdimC ) 1 (where C is the initial concentration of two-electron process with both electron transfers being relatively
monomer). Thus, it is not surprising that KdimC must be greater reversible. The points in Figures 8 and 9 are simulations based
than unity for most of the monomer to dimerize and to affect on two standard potentials, E°1 and E°2, that are quite similar.
the shape of the voltammograms significantly. The values of The disproportionation reaction (reaction 4) was not included
KdimC for the simulations in Figures 4-6 are 18, 2100, and 14, as it did not affect the simulations. A summary of the simulation
respectively. By contrast when KdimC is small, the dimerization parameters for 1e-h and 1k is given in Table 4 and a full listing
5810 J. Phys. Chem. C, Vol. 111, No. 15, 2007 Macı́as-Ruvalcaba and Evans

TABLE 4: Summary of Simulation Parameter Values for


1g-i and 1ka
compd E°1/V E°2/V E°1 - E°2/V 105 D/cm2 s-1
1e 0.124 0.123 +0.001 1.53
1f 0.064 0.104 -0.040 1.24
1g 0.253 0.208 +0.045 1.51
1h 0.391 0.305 +0.086 1.52
1ib 0.47 0.37 +0.10 1.36
1k 0.356 0.503 -0.147 c
a
Temperature: 25 °C. Potentials expressed with respect to the
ferrocenium/ferrocene potential. Diffusion coefficients assumed to be
identical for all species. With the exception of 1g, the E° and D values
correspond to the average from two concentrations. See Table 3 for
error estimates. b Approximate values. Poor fits at low scan rates. c Not
determined as this slightly soluble compound was studied with a
Figure 7. Voltammogram of 3.48 × 10-3 M N,N,N′,N′-tetramethyl- saturated solution.
benzidine, 1j, at 1.00 V/s and -18 °C (full curve). Best fit simulation
for two-step oxidation without dimerization of the cation radical
(dashed). Best fit simulation for two-step oxidation with fast, reversible oxidation states of the system. Structure 1e2+ shows how, in
dimerization of the cation radical with simulation parameter values as the dication, the dialkylamino groups will attempt to move into
summarized in Table 3 (a full list of parameter values is given in the the plane of the biphenyl system thus provoking steric stress
Supporting Information) (open circles). with substituents in the 3 and 3′ positions. The same tendency,
but weaker, is likely to be at work in the cation radicals.

Figure 8. Cyclic voltammogram of 3.94 mM N,N,N′,N′-tetraethyl-


3,3′-dimethoxybenzidine, 1f, at 1.00 V/s and 25 °C (full curve).
Simulation based on parameter values summarized in Table 4 (a full When the steric interactions in the dication are strong as in
list of parameter values is given in the Supporting Information) (open the dication of 3,6-bis(dimethylamino)durene (2), profound
circles). changes in structure occur. In the case of 22+, the N-methyl/
ring methyl interactions cause the central six-membered ring
to fold into a boat-form.15 Such significant changes in structure
cause large changes in orbital energies that eventually lead to
the removal of the second electron being easier than the first.
However, in the case of 1e2+ the calculated structure (see
the Experimental Section) does not differ in major ways
compared to the cation radical and neutral compound. The cation
radical and dication show less pyramidalization at nitrogen than
does the neutral and the dimethylamino groups tend to turn into
the plane of the adjacent ring system in the cations. Also, the
twist angle between the two biphenyl rings closes down in the
cation radical and dication compared to the neutral. Nevertheless,
the changes are small compared to the ring-folding that occurs
in 22+.
Figure 9. Cyclic voltammogram of 3.50 mM N,N,N′,N′,3,3′-hexam- However, the changes in structure on going from neutral to
ethylbenzidine, 1g, at 1.00 V/s and 25 °C (full curve). Simulation based cation radical to dication in 1e do bring about changes in the
on parameter values summarized in Table 4 (a full list of parameter energies of the species that lead to compression of E°1 - E°2
values is given in the Supporting Information) (open circles). to a value near zero (Table 4). This can be seen from the
computed (DFT) gas-phase energies of disproportionation
appears in the Supporting Information. The reactions are found (reaction 4) for 1e as compared with 1a. These are +95.26 kcal/
to be simple electron transfers with relatively large electron- mol (-4.13 V) for 1a and +73.80 kcal/mol (-3.20 V) for 1e.
transfer rate constants but with small differences in standard These gas-phase potential differences of several volts, indicative
potential, E°1 - E°2 being +0.001, -0.040, +0.045, and +0.086 of strongly disfavored disproportionation, are reduced in the
V for 1e-h, respectively. In most cases where E°1 - E°2 is solution phase to the range of tenths of volts by the solvation
severely compressed or even inverted (E°1 - E°2 > 0) there energies of the three species, particularly the very large solvation
are significant steric interactions in one or more of the three energy of the dication. Thus, the gas-phase disproportionation
Oxidation Reactions of a Series of Benzidines J. Phys. Chem. C, Vol. 111, No. 15, 2007 5811

of 1e is almost 1 V less unfavorable that that of 1a in qualitative 81, 1177-1182. (c) Pragst, F.; Ziebig, R.; Seydewitz, U.; Driesel, G.
agreement with the fact that solution-phase disproportionation Electrochim. Acta 1980, 25, 341-352. (d) Pragst, F.; Janda, M.; Stibor, I.
Electrochim. Acta 1980, 25, 779-783. (e) Bruno, P.; Caselli, M.; Traini,
is unfavored for 1a but is close to favorable for 1e (Table 4). A. J. Electroanal. Chem. 1980, 113, 99-111. (f) Parker, V. D.; Hammerich,
This discussion is based on the relation ∆G°disp ) F(E°1 - E°2). O. Acta Chem. Scand. 1981, B35, 341-347. (g) Lerflaten, O.; Parker, V.
Thus, even the relatively minor structural changes in 1e affect D. Acta Chem. Scand. 1982, B36, 225-234. (h) Margaretha, P.; Parker, V.
D. Acta Chem. Scand. 1982, B36, 260-262. (i) Kashti-Kaplan, S.; Hermolin,
the energies of the three oxidation states such that dispropor- J.; Kirowa-Eisner, E. J. Electrochem. Soc. 1981, 128, 802-810. (j)
tionation can become favorable, corresponding to potential Hermolin, J.; Kirowa-Eisner, E.; Kosower, E. M. J. Am. Chem. Soc. 1981,
inversion. 103, 1591-1593. (k) Osteryoung, J.; Talmor, D.; Hermolin, J.; Kirowa-
Eisner, E. J. Phys. Chem. 1981, 85, 285-289. (l) Rusling, J. F. J.
Electroanal. Chem. 1981, 125, 447-458. (m) Ryan, M. D.; Swanson, D.
4. Conclusion D.; Glass, R. S.; Wilson, G. S. J. Phys. Chem. 1981, 85, 1069-1075. (n)
Amatore, C.; Pinson, J.; Savéant, J. M. J. Electroanal. Chem. 1982, 137,
In this work it has been shown that fast, reversible dimer- 143-148. (o) Amatore, C.; Pinson, J.; Savéant, J. M. J. Electroanal. Chem.
ization of the cation radicals of benzidines can be detected by 1982, 139, 193-197. (p) Houser, K. J.; Bartak, D. E.; Hawley, M. D. J.
cyclic voltammetry through careful analysis of wave shapes. Am. Chem. Soc. 1973, 95, 6033-6044. (q) Koppang, M. D.; Woolsey, N.
The benzidines whose cation radicals undergo dimerization F.; Bartak, D. E. J. Am. Chem. Soc. 1984, 106, 2799-2805. (r) Koppang,
M. D.; Woolsey, N. F.; Bartak, D. E. J. Am. Chem. Soc. 1985, 107, 4692-
are all unalkylated at nitrogen except for N,N,N′,N′-tetra- 4700. (s) Evans, D. H.; Jimenez, P. J.; Kelly, M. J. J. Electroanal. Chem.
methylbenzidine for which weak dimerization is detected 1984, 163, 145-157. (t) Gennaro, A.; Romanin, A. M.; Severin, M. G.;
at low temperature. The dimerization equilibrium constants Vianello, E. J. Electroanal. Chem. 1984, 169, 279-285. (u) Parker, V. D.
Acta Chem. Scand. 1983, B37, 871-877. (v) Amatore, C.; Garreau, D.;
have been determined independently by UV-visible absorption Hammi, M.; Pinson, J.; Savéant, J. M. J. Electroanal. Chem. 1985, 184,
spectra of solutions of cation radical salts that were synthe- 1-24. (w) Mendkovich, A. S.; Michalchenko, L. V.; Gultyai, V. P. J.
sized in this work. For several N-peralkylated benzidines that Electroanal. Chem. 1987, 224, 273-275. (x) Eliason, R.; Hammerich, O.;
also contain ring substituents, the individual standard potentials Parker, V. D. Acta Chem. Scand. 1988, B42, 7-10. (y) Savéant, J. M. Acta
Chem. Scand. 1988, B42, 721-727. (z) Starichenko, V. F.; Efremova, N.
for the two steps of oxidation are very similar so that the V.; Shteingarts, V. D. IzV. Akad. Nauk SSSR Ser. Khim. 1988, 2170-2171.
benzidines are oxidized in a single, reversible, two-electron (aa) Gultyai, V. P.; Rubinskaya, T. Ya.; Mendkovich, A. S.; Rusakov, A.
process. Reasons for this potential compression have been I. IzV. Akad. Nauk SSSR Ser. Khim. 1987, 2812-2814. (ab) Mendkovich.
A. S.; Churilina, A. P.; Rusakov, A. I.; Gultyai, V. P. IzV. Akad. Nauk
discussed. SSSR Ser. Khim. 1991, 1777-1782. (ac) Rubinskaya, T. Ya.; Mendkovich,
A. S.; Lisitsina, N. K.; Yakovlev, I. P.; Gultyai, V. P. IzV. Akad. Nauk
Acknowledgment. This research was supported by the SSSR Ser. Khim. 1993, 1735-1738. (ad) Gultyai, V. P.; Lisitsina, N. K.;
National Science Foundation, Grant CHE 0347471. Ignatenko, A. V.; Mendkovich, A. S. IzV. Akad. Nauk SSSR Ser. Khim.
1991, 873-877. (ae) Mendkovich, A. S.; Churilina, A. P.; Mikhalchenko,
L. V.; Gultyai, V. P. IzV. Akad. Nauk SSSR Ser. Khim. 1990, 1492-1495.
Supporting Information Available: Temperature-dependent (af) Mikhalchenko, L. V.; Mendokovich, A. S.; Gultyai, V. P. IzV. Akad.
spectra of cation radical salts of 1b and 1c, apparent molar Nauk SSSR Ser. Khim. 1985, 2158. (ag) Smie, A.; Heinze, J. Angew. Chem.,
absorptivity vs total concentration of cation radical salts of 1b- Int. Ed. Engl. 1997, 36, 363-367. (ah) Tschuncky, P.; Heinze, J.; Smie,
A.; Engelmann, G.; Kossmehl, G. J. Electroanal. Chem. 1997, 433, 223-
d, simulation parameter values used to fit voltammograms for 226. (ai) Huebler, P.; Heinze, J. Ber. Bunsenges. 1998, 102, 1506-1509.
1a-d and 1j, comment on the magnitude of dimerization rate (aj) Heinze, J.; Rasche, A. Electrochem. Commun. 2003, 5, 776-781. (ak)
constants, simulation parameter values used to fit voltammo- Mazine, V.; Heinze, J. J. Phys. Chem. A 2004, 108, 230-235. (al) Macı́as-
grams of 1e-h and 1k, respectively and additional examples Ruvalcaba, N. A.; Telo, J. P.; Evans, D. H. J. Electroanal. Chem. 2007,
600 294-302.
of fits of simulation to experimental voltammograms for 1a-
(9) (a) For a recent example of potential inversion see ref 9b. For
d, 1j, and 1e-h. This material is available free of charge via another, with extensive references to earlier examples, see ref 9c. (b) Gruhn,
the Internet at http://pubs.acs.org. N. E.; Macı́as-Ruvalcaba, N. A.; Evans, D. H. Langmuir 2006, 22, 10683-
10688. (c) Macı́as-Ruvalcaba, N. A.; Evans, D. H. J. Phys. Chem. B 2006,
References and Notes 110, 5155-5160.
(10) Macı́as-Ruvalcaba, N. A.; Evans, D. H. J. Phys. Chem. B 2005,
(1) Hand, R. L.; Nelson, R. F. J. Am. Chem. Soc. 1974, 96, 850-860. 109, 14642-14647.
(2) Debrodt, H.; Heusler, K. E. Electrochim. Acta 1980, 25, 545-549.
(11) Awano, H.; Ohigashi, H. Synth. Met. 1989, 32, 389-394.
(3) Douadi, T.; Benabid, S.; Cariou, M. Electrochim. Acta 2003, 48,
2659-2665. (12) Macı́as-Ruvalcaba, N. A.; Evans, D. H. J. Electroanal. Chem. 2005,
(4) (a) Note that the material was incorrectly formulated as a charge- 585, 150-155.
transfer complex between the dication and the neutral compound in the (13) (a) Rudolph, M. J. Electroanal. Chem. 2003, 543, 23-29. (b)
earlier work.4b,c (b) Awano, H.; Ogata, T.; Murakami, H.; Yamashita, T.; Rudolph, M. J. Electroanal. Chem. 2004, 571, 289-307. (c) Rudolph, M.
Ohigashi, H. Synth. Met. 1990, 36, 263-266. (c) Awano, H.; Ohigashi, H. J. Electroanal. Chem. 2003, 558, 171-176. (d) Rudolph, M. J. Comput.
Bull. Chem. Soc. Jpn. 1990, 63, 2101-2103. (d) Awano, H.; Ichihara, O.; Chem. 2005, 26, 619-632. (e) Rudolph, M. J. Comput. Chem. 2005, 26,
Sawada, K.; Ohigashi, H. Ber. Bunsenges. Phys. Chem. 1996, 100, 1700- 633-641. (f) Rudolph, M. J. Comput. Chem. 2005, 26, 1193-1204.
1705.
(5) Savéant, J. M.; Vianello, E. Electrochim. Acta 1967, 12, 1545- (14) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
1561. M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
(6) Shuman, M. S. Anal. Chem. 1970, 42, 521-523. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
(7) (a) Howarth, O. W.; Fraenkel, G. K. J. Am. Chem. Soc. 1966, 88, Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
4514-4515. (b) Chiang, T. C.; Reddoch, A. H.; Williams, D. F. J. Chem. Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Phys. 1971, 54, 2051-2055. (c) Kotowski, S.; Grabner, E. W.; Brauer, Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
H.-D. Ber. Bunsenges. Phys. Chem. 1980, 84, 1140-1145. (d) Enkelmann, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
V.; Morra, B. S.; Kröhnke, Ch.; Wegner, G.; Heinze, J. Chem. Phys. 1982, Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
66, 303-313. (e) Terahara, A.; Ohya-Nishiguchi, H.; Hirota, N.; Oku, A. Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
J. Phys. Chem. 1986, 90, 1564-1571. (f) Oyama, M.; Mitani, M.; Washida, Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
M.; Masuda, T.; Okazaki, S. J. Electroanal. Chem. 1999, 473, 166-172. A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
(g) Oyama, M.; Masuda, T.; Mitani, M.; Okazaki, S. Electrochemistry 1999, Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
67, 1211-1213. (h) Kochi, J. K.; Rathore, R.; Le Maguères, P. J. Org. Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
Chem. 2000, 65, 6826-6836. (i) Small, D.; Zaitsev, V.; Jung, Y.; Rosokha, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
S. V.; Head-Gordon, M.; Kochi, J. K. J. Am. Chem. Soc. 2004, 126, 13850- Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
13858. (j) Komaguchi, K.; Nomura, K.; Shiotani, M.; Lund, A.; Jansson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
M.; Lunell, S. Spectrochim. Acta, Part A 2006, 63, 76-84. Revision B.05; Gaussian, Inc.: Pittsburgh PA, 2003.
(8) (a) Yeh, L.-S. R.; Bard, A. J. J. Electroanal. Chem. 1976, 70, 157- (15) Gruhn, N. E.; Macı́as-Ruvalcaba, N. A.; Evans, D. H. J. Phys.
169. (b) Yildiz, A.; Baumgärtel, H. Ber. Bunsenges. Phys. Chem. 1977, Chem. A 2006, 110, 5650-5655.

You might also like