Overoxidation of phenol by hexachloroiridate(IV)

Article abstract of DOI:10.1021/ic201897m

It has been previously established that the aqueous oxidation of phenol by a deficiency of [IrCl₆]^2-proceeds through the production of [IrCl₆]^3-and phenoxyl radicals. Coupling of the phenoxyl radicals leads prim

Full text of DOI:10.1021/ic201897m

Article
pubs.acs.org/IC
Overoxidation of Phenol by Hexachloroiridate(IV)
Na Song and David M. Stanbury*
Department of Chemistry and Biochemistry, Auburn University, Auburn, Alabama 36849, United States
S
* Supporting Information
ABSTRACT: It has been previously established that the aqueous oxidation of phenol by a
deficiency of [IrCl6]2− proceeds through the production of [IrCl6]3− and phenoxyl radicals.
Coupling of the phenoxyl radicals leads primarily to 4,4′-biphenol, 2,2′-biphenol, 2,4′-
biphenol, and 4-phenoxyphenol. Overoxidation occurs through the further oxidation of
these coupling products, leading to a rather complex mixture of final products. The rate
laws for oxidation of the four coupling products by [IrCl6]2− have the same form as those
+
for the oxidation of phenol itself: −d[IrIV]/dt = {(kArOH + kArO Ka/[H ])/(1 + Ka/
−
[H+])}[ArOH]tot[IrIV]. Values for kArOH and kArO have been determined for the four
−
substrates at 25 °C and are assigned to H2O-PCET and electron-transfer mechanisms,
respectively. Kinetic simulations of a combined mechanism that includes the rate of
oxidation of phenol as well as the rates of these overoxidation steps show that the degree
of overoxidation is rather limited at high pH but quite extensive at low pH. This pH-
dependent overoxidation leads to a pH-dependent stoichiometric factor in the rate law for
oxidation of phenol and causes some minor deviations in the rate law for oxidation of phenol. Empirically, these minor deviations
can be accommodated by the introduction of a third term in the rate law that includes a “pH-dependent rate constant”, but this
approach masks the mechanistic origins of the effect.
from coupling of the phenoxyl radical. These rates are then
incorporated into an overall mechanism for the oxidation of
phenol, which shows that overoxidation is responsible for many
of the reaction products. Moreover, it is shown that the degree
of overoxidation is pH-dependent and that this can account for
the observed deviations from the classical two-term rate law.
INTRODUCTION
■
Redox reactions in aqueous solution are often accompanied by
changes in protonation. When the reactions proceed via one-
electron steps, the phenomenon of proton-coupled electron
transfer (PCET) is often involved. The breadth of importance of
PCET is immense, and PCET is of great significance because it
is often an absolute criterion of reactivity.1 One approach to
gaining insight into PCET is to study reactions in which one of
the reactants is a typical inorganic outer-sphere electron-transfer
reagent with no acid/base properties: this constraint confines
much of the electron-proton coupling considerations to the
other reaction partner. The oxidation of phenols has become
central in developing the concepts of PCET, in part because of
its importance in revealing the function of tyrosine in redox
proteins but also because these reactions are quite amenable to
study. Accordingly, the oxidation of phenol by [IrCl6]2− has
been the focus of a classic publication,2 and it is now of interest
as a model for reactions where electron transfer occurs in
concert with proton transfer to the solvent: H2O-PCET.3−5
Despite its importance, the oxidation of phenol by [IrCl6]2−
presents certain difficulties. One of these is that the phenolic
products are a complex mixture,2 and it is unclear how the post-
rate-limiting steps lead to this mixture. It is also unclear whether
these latter stages in the reaction have any influence on the
measurements of the putative rate-limiting steps. Of further
concern is the pH dependence of the rates, which displays
deviations from the classic two-term rate law for such reactions5
and might provide additional evidence in support of a pH-
dependent rate constant, as has been reported previously for the
oxidation of phenol by [Ru(bpy)3]3+.6 Herein is reported a study
of the [IrCl6]2− oxidations of the four major products derived
EXPERIMENTAL SECTION
■
Reagents and Solutions. All commercial chemical reagents were
used as received except as noted. Ammonium hexachloroiridate(III)
monohydrate (IrIII), 2,6-dimethylacetanilide, chlorosulfonic acid, 3-
chloroperoxybenzoic acid, N-tert-butylphenylnitrone (PBN), α-(4-
pyridyl-N-oxide)-N-tert-butylnitrone (POBN), 2-methyl-2-nitrosopro-
pane (MNP), 5,5-dimethyl-1-pyrroline N-oxide (DMPO), deuterium
oxide, 3,5-dibromosulfanilic acid sodium salt, 2-iodophenol, 4-
hydroxyphenyl boronic acid, palladium(II) acetate, 1,1′-bis-
(diphenylphosphino)ferrocence (dppf), potassium carbonate, sodium
acetate anhydrous, cacodylic acid, and sodium hydroxide were
purchased from Sigma-Aldrich Chemical Co. Perchloric acid,
ammonium perchlorate, sodium chloride, ammonium chloride,
copper(II) nitrate trihydrate, acetic acid, monochloroacetic acid,
ethanol, diethyl ether, 1,4-dioxane, petroleum ether, ethyl acetate,
acetonitrile, hydrochloric acid, and hydrogen peroxide were from
Fisher Scientific Co. 2,2′-Biphenol, 4,4′-biphenol, and 4-phenoxyphe-
nol were commercially available from Acros Organics. Ammonium
hexachloroiridate(IV) (IrIV) was purchased from Alfa or prepared
according to the literature7 by the addition of ammonium chloride
(Fisher) to a solution of sodium hexachloroiridate(IV) hexahydrate.
Phenol (Fluka) was recrystallized from a 75% (w/w) water solution as
in the literature.8
Received: August 30, 2011
Published: November 16, 2011
© 2011 American Chemical Society
12762
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
All solutions were freshly prepared with deionized water provided
by a Barnstead NANO Pure Infinity ultrapure water system and
purged with argon gas prior to reaction to prevent potential
complications caused by O2. In order to increase the concentration
of the solution, 4,4′-biphenol or 2,4′-biphenol was first dissolved in
CH3CN or ethanol and then diluted with water to make a reaction
solution where less than 1% (w/w) organic solvent was present. The
ionic strength was adjusted by lithium perchlorate trihydrate (GFS)
and was approximately equal in both solutions to prevent Schlieren
effects. Selected buffer solutions (acetate, monochloroacetate, and
cacodylate buffers) were applied to control the pH if necessary.
Preparation of Sodium 3,5-Dibromo-4-nitrosobenzenesul-
fonate (DBNBS). This compound was synthesized from 3,5-
dibromosulfanilic acid sodium salt according to the method of Kaur
et al.9 and further purified using procedure type C of Hamilton et al.10
A mixture of 3,5-dibromosulfanilic acid (10 mmol), anhydrous sodium
acetate (10 mmol), 7.9 mL of a 30% aqueous hydrogen peroxide
solution, and 30 mL of glacial acetic acid was warmed gently to
dissolve the solid. The solution was stored at room temperature for 14
days. Then the crude yellow solid product was collected and washed
with glacial acetic acid, cold ethanol, diethyl ether/1,4-dioxane (1:1),
and cold ethanol again. A pale-yellow powder was obtained after
(NH4)2[IrCl6] has strong absorbance around 488 nm while the
corresponding product (NH4)3[IrCl6] does not, as shown in Figure S-2
(Supporting Information), all kinetic data were obtained by monitoring
the absorbance of IrIV at λmax (488 nm) with ε488 = (3.9 ± 0.1) ×
103 M−1 cm−1.14 In order to detect the DBNBS influence on the
reactions of phenol, we also observed the absorbance change over the
wavelength 400−525 nm at intervals of 25 nm. For the phenol and
4-phenoxyphenol reactions, the observed pseudo-first-order rate
constants were obtained from the fitting of kinetic traces over 5
half-lives to first-order exponential functions. For 4,4′-biphenol and
2,2′-biphenol, the observed initial rate constants were determined from
the slope of the linear regression of the logarithm of absorbance at 488
nm within the first half-life, while for 2,4′-biphenol, the rate constants
were obtained from double-exponential fits. Each reported observed
rate constant is the average of at least five shots. The Specfit/32,
version 3.0.15, global analysis system was applied to simulate the
reaction traces, and the GraphPad Prism 4 or 5 software was used to
analyze the rate law with 1/Y2 weighting.
1H and 13C NMR spectra were acquired on a Bruker AV 400 MHz
spectrometer; chemical shifts in CDCl3 are relative to tetramethylsi-
lane, and those in D2O are relative to DSS. The melting points were
obtained using an Electrothermal IA 9100 digital melting point
apparatus. Cyclic voltammograms (CV) and Osteryoung square-wave
voltammograms were performed on a BAS 100B electrochemical
analyzer equipped with a BAS C3 cell stand and a purging and stirring
system; a glassy carbon electrode was the working electrode, Ag/AgCl
(3.0 M NaCl) was the reference electrode (E° = 0.205 V vs NHE),15
and a Pt wire was the auxiliary electrode.
1
drying. Yield: 30%. Mp: >300 °C. H NMR (D2O): δ 8.30 (s, 2H).
13C NMR (D2O): δ 119.10 (aryl CBr), 131.03 (aryl CH), 141.00 (aryl
CSO3Na), 147.61 (aryl CNO).
Preparation of Sodium 2,4-Dimethyl-4-nitrosobenzenesul-
fonate (DMNBS). This compound was prepared according to the
literature.11,12 2,6-Dimethylacetanilide (18 mmol) was added into 10
mL of cold ClSO3H with stirring. After reacting for 15 min at 5 °C, 1 h
at 25 °C, and 10 min at 40 °C, the mixture was poured onto cracked
ice to form a white solid of the sulfonyl chloride. Then this compound
was treated with concentrated HCl under reflux for 1 h before
neutralization by NaOH to produce sodium 2,6-dimethylaniline-3-
sulfonate. The sodium sulfonate (10 mmol) was dissolved in 30 mL of
methanol and oxidized by m-chloroperoxybenzoic acid (10 mmol)
with stirring for 1 h at room temperature. Then ether was added to
Quantum calculations of the electronic spectra were performed with
the Spartan '08 software package.16
RESULTS
■
Results for the oxidation of phenol by IrIV have already
appeared in preliminary form.5 Results for oxidation of the four
coupling products are shown here for the first time.
The kinetic traces for the consumption of 1 × 10−4 M IrIV in
its reaction with a large excess of phenol were obtained at various
p[H+] at 488 nm, the absorbance maximum of IrIV. Figure 1a
exhibits a typical kinetic trace for a reaction with 0.44 M phenol
in 0.05 M HClO4 (p[H+] = 1.3). Such kinetic traces do not give
good fits with either first- or second-order rate laws. As the
p[H+] increases, good-quality pseudo-first-order fits are obtained
at both p[H+] = 2.9 and 4.8 with 0.044 M phenol (Figure 1b,c).
At high p[H+] (=7.1), the reaction with 1.8 × 10−3 M phenol is
much faster and becomes non-pseudo-first-order again (Figure
1d). These results show a strong p[H+] effect on phenol
oxidation by IrIV. As shown below, the deviations at low pH arise
from inhibition by IrIII, while the deviations at high pH are due
to the absorbance of phenolic products.
1
precipitate the final product. Yield: 80%. H NMR (D2O): δ 2.56 (s,
3H), 2.75 (s, 3H), 7.54 (d, 1H), 8.13 (d, 1H).
Preparation of 2,4′-Biphenol. 2,4′-Biphenol was prepared
through a Pd-catalyzed Suzuki−Miyaura coupling reaction according
to the literature.13 Under dry nitrogen, 4-hydroxyphenyl boronic
acid (1.4 mmol), Pd(OAc)2 (0.14 mmol), bis(diphenylphosphino)-
ferrocene (dppf) (0.14 mmol), and K2CO3 (3 mmol) were added to a
solution of 2-iodophenol (1 mmol) in 10 mL of tetrahydrofuran. The
reaction was stirred under reflux for 1 day and monitored by thin-layer
chromatography. After the reaction completed, the crude product was
purified by column chromatography on silica gel (eluent PE/EtOAc =
3.5/1). Then, sublimation was performed under vacuum at 160 °C to
remove the nonvolatile metal residue impurities, and resublimation
was carried out at 110 °C to remove the volatile impurities. Yield: 50%.
Mp: 161.2−162.8 °C. 1H NMR (CDCl3; Figure S-1 in the Supporting
Information): δ 4.89 (s, 1H), 5.14 (s, 1H), 6.94−6.99 (m, 4H), 7.20−
7.24 (m, 2H), 7.34−7.36 (m, 2H).
Methods. A Corning 450 pH/ion meter was used with a Mettler
Toledo InLab 421 or InLab Semi-Micro-L combination pH electrode.
The reference electrode electrolyte was replaced with 3 M NaCl to
prevent the formation of KClO4 precipitate. Electrode calibrations at μ =
0.1 M (LiClO4) were carried out with 0.01−0.1 M perchloric acid. With
the known H+ concentration and pH reading, the activity coefficient γ
(=0.839 ± 0.04) was obtained from the equation p[H+] = pH + log γ,
where p[H+] is equal to −log [H+]. When the pKa of 4,4′/2,4′-biphenol is
measured, an alkaline error may occur with pH values higher than 11.
Thus, we calibrated the electrode with 1 × 10−3 M NaOH at μ = 0.1 M
(LiClO4) and found that the true pH is equal to the apparent pH plus 0.3.
All measurements were performed at 25.0 ± 0.1 °C. The kinetic
experiments were carried out on a Hi-Tech SF-51 stopped-flow
spectrophotometer with OLIS 4300 data acquisition and analysis
software. UV−vis spectra were monitored on a HP-8453 diode-array
spectrophotometer equipped with a Brinkman Lauda RM6 thermo-
statted water bath to maintain the temperature at 25 °C. Because
Kinetic Inhibition by IrIII. At p[H+] = 1.3, Figure 2a shows
the effects of adding a 5-fold excess of IrIII under conditions
otherwise identical with those in Figure 1a. The retarding effect
of IrIII at this pH is seen to be quite strong and clearly can be
expected to cause deviations from pseudo-first-order kinetics of
the type shown in Figure 1a. On the other hand, at p[H+] = 7.1,
no IrIII inhibition is observed with the addition of 5- and 10-fold
excess of IrIII (Figure 2b,c), and so the deviations from pseudo-
first-order kinetics at this pH (Figure 1d) are due to factors
other than inhibition by IrIII.
Basicity of IrIII. The basicity of IrIII was probed by measuring
the pH dependency of the cyclic voltammetry of 1 × 10−3 M IrIV
in the p[H+] range 0−2 at 1 M ionic strength (LiClO4). p[H+]
values were adjusted by HClO4 according to p[H+] = −log
[HClO4]. Reversible CVs (E1/2 = 0.717 ± 0.007 V vs Ag/AgCl)
were obtained over this pH range. No pH dependence of E1/2
was observed under all of these conditions, which implies that
12763
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
free of complications arising from general base catalysis. This
conclusion is in agreement with the recent report from Irebo -
et al. and Bonin et al. that general base catalysis in outer-
sphere phenol oxidation can be insignificant under certain
conditions.4,18,19
Spin Trapping. Several conventional spin traps (illustrated
in Scheme 1) were investigated for their effects on the kinetics
Scheme 1. Structure of Spin-Trapping Agents Tested
of the phenol/IrIV reaction in 0.05 M H+. As we have reported
previously, DBNBS is quite effective in this regard.5 Figure 3
illustrates the dramatically improved fit to a first-order rate law
achieved with only 10 mM DBNBS. The other spin traps PBN,
DMPO, POBN, and MNP barely affect the reaction, and
DMNBS is only partially effective (Table S-1 in the Supporting
Information). These results are not unexpected, given the
known specificity of DBNBS for phenoxyl radicals.20 Tests in
0.05 M H+ with concentrations of DBNBS ranging from 0.1 to
15 mM (Table S-2 in the Supporting Information) show that
the effects saturate at about 5 mM DBNBS. A standard
DBNBS concentration of 10 mM is used in the experiments
described below. The approach to saturation is expected to be
dependent on both the efficiency of phenoxyl radical
scavenging and the DBNBS dimerization equilibrium (K =
1.3 × 10−3 M).21
Figure 1. Kinetic traces of the oxidation of phenol by 1 × 10−4 M IrIV
at different p[H+] values. The lower boxes show the experimental
traces (solid lines) and the pseudo-first-order fits (dashed lines). The
upper boxes show the residuals in the fits. μ = 0.1 M (LiClO4); T = 25
°C. (a) [phenol]tot = 0.44 M; [HClO4] = 0.05 M. (b) [phenol]tot
=
0.044 M; p[H+] = 2.9 (0.02 M monochloroacetate buffer). (c)
[phenol]tot = 0.044 M; p[H+] = 4.8 (0.02 M acetate buffer). (d)
[phenol]tot = 1.8 × 10−3 M; p[H+] = 7.1 (0.02 M cacodylate buffer).
IrIII is not significantly basic even in 1 M H+. Bruhn et al.
observed a small but significant dependence of E1/2 over this
same pH range in H+/Na+ media;17 we attribute this effect to the
differing specific interaction coefficients of these two ions.
General Base Catalysis Tests. Tests for general base
catalysis of the reaction of phenol with IrIV were described
previously.5 In brief, they show that the reaction is not catalyzed
by low concentrations of the cacodylate buffer at pH 6.5, by low
concentrations of ammonia at pH 7, or by low concentra-
tions of phenoxide at pH 5.1. These results show that the
kinetic data presented herein on phenol oxidation by IrIV are
The effects of DBNBS are also significant at p[H+] = 7. The
results (Table S-3 in the Supporting Information) show that
DBNBS removes the deviations from pseudo-first-order
kinetics that are otherwise seen at this pH. The effect is
attributed to DBNBS interception of phenoxyl radicals, the
dimerization of which would lead to absorbing intermediates
(biphenoquinones), as shown below.
Figure 2. Comparative traces of the phenol reaction with added IrIII (dashed line) and no added IrIII (solid line). [IrIV]0 = 1 × 10−4 M; μ = 0.1 M
(LiClO4); T = 25 °C. (a) [phenol]tot = 0.44 M; [IrIII] = 5 × 10−4 M; [HClO4] = 0.05 M. (b) [phenol]tot = 1.8 × 10−3 M; [IrIII] = 5 × 10−4 M; p[H+] =
7.1 (0.02 M cacodylate buffer). (c) [phenol]tot = 1.8 × 10−3 M; [IrIII] = 1 × 10−3 M; p[H+] = 7.1 (0.02 M cacodylate buffer).
12764
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
Figure 4. Plot of kobs/[phenol]tot versus p[H+] in the absence of
DBNBS. [IrIV]0 = 1 × 10−4 M; μ = 0.1 M (LiClO4); T = 25 °C. The
p[H+] values between 2.4 and 3.4 were maintained by a 0.02 M
monochloroacetate buffer; the p[H+] values between 3.6 and 5.4 were
maintained by a 0.02 M acetate buffer; the p[H+] values between 5.6
and 7.0 were maintained by a 0.02 M cacodylate buffer. The solid line
is the fit to eq 2, and the dashed line is the fit to eq 3. Data are from
Table S-4 in the Supporting Information.
Figure 3. Trace of the phenol reaction with added DBNBS. The lower
box shows the experimental trace (solid line) and the pseudo-first-
order fit (dashed line). The upper box shows the residuals in the fit.
[phenol]tot = 0.44 M; [IrIV]0 = 1 × 10−4 M; [DBNBS] = 10 mM;
[HClO4] = 0.05 M; μ = 0.1 M (LiClO4); T = 25 °C.
6
−1
kArOH = 0.77 ± 0.03 M−1 s−1 and kArO = (8.0 ± 0.2) × 10 M
−
s−1. As discussed below, this rate increase induced by DBNBS is
attributed to differing stoichiometric factors.
Phenol Dependence. In results already published, it was
shown that the rate of oxidation of phenol by IrIV is first-order
in [phenol].5 This result was obtained in 0.05 M HClO4 with
10 mM DBNBS and also at p[H+] = 5.1 without DBNBS. The
conditional rate constants are 0.612 ± 0.001 at p[H+] = 1.3 and
106 ± 4 at p[H+] = 5.1, and they indicate that the kinetics are
sensitive to p[H+].
Inclusion of the k° Term. It has been reported that eq 2 is
inadequate to describe the reaction of phenol with [Ru-
(bpy)3]3+, and a “pH-dependent rate constant” (the k° term)
was introduced as in eq 3 to fit the data.6
p[H+] Dependence. A simple two-term rate law arises
from the assumption that both phenol and the phenolate anion
can react with IrIV, as shown by eqs 1 and 2.
(3)
(1)
When the pH-dependent data for oxidation by IrIV with no added
DBNBS are fit with eq 3, the optimized values are kArOH = 0.40 ±
6
−1
0.06 M−1 s−1, kArO = (4.9 ± 0.1) × 10 M , and k° = (6.5 ± 3) ×
10−3 M−1 s−1 (Table S-5 in the Supporting Information). This
fit yields slightly improved residuals, but the values of kArOH and
−
kArO are virtually unchanged. Over the p[H+] range 1−7, the
−
(2)
maximum contribution of the k° term to the total rate is 16%
and occurs around p[H+] = 2.7. Although this k° term has only
marginal statistical significance, we show below that it is a
consequence of pH-dependent stoichiometric factors.
Here, Ka is the acid dissociation constant of phenol, pKa,ArOH
=
9.79 at μ = 0.1 M.22 kArOH and kArO represent the reactivities of
phenol and the phenolate anion.
−
As we reported previously,5 in the presence of DBNBS, fitting
the data to eq 3 yields slightly improved residuals relative to eq 2
(Table S-6 in the Supporting Information). The fitted values for
Between p[H+] = 2.4 and 6.8, the oxidation of phenol by IrIV
yields good-quality pseudo-first-order fits, as shown in Figures
1b,c. This feature enables us to study the p[H+] effect on the
rate constants from p[H+] = 2.46 to 6.74 with 1 × 10−4 M IrIV
and (4.43−44.3) × 10−3 M phenol. Selected buffers were
employed to maintain the p[H+] values. The data are
summarized in Table S-4 in the Supporting Information, and
a plot of kobs/[phenol]tot versus p[H+] is shown in Figure 4.
A nonlinear least-squares fit of the data to eq 2 shows that the
rates conform to this two-term rate law with kArOH = 0.54 ±
−
kArOH and kArO are only slightly changed: kArOH = 0.59 ± 0.03
M−1 s−1, kArO = (7.3 ± 0.1) × 106 M−1, and k° = (2.0 ± 0.3) ×
−
10−2 M−1 s−1. The maximum contribution from the k° term is
28% and appears at the same p[H+] as that in the result without
DBNBS. The origin of the k° term in the presence of DBNBS is
currently unknown but is speculated to arise from a pH-dependent
overoxidation of the DBNBS/phenoxyl adduct.
0.02 M−1 s−1 and kArO = (5.0 ± 0.1) × 106 M−1 s−1.
−
Kinetic Isotope Effect (KIE). We have previously reported
on the deuterium KIE in the reaction between phenol and IrIV.5
It was measured by comparing the rates in a D2O solution to
those in normal H2O. The experiments were performed at
[H+] = 0.09 M in the presence of 10 mM DBNBS, and thus the
KIE refers to kArOH. The rates are independent of the pH under
these conditions, so the equilibrium isotope effect on Ka is not
an issue. The KIE value is 3.5 ± 0.3, which clearly indicates a
As described above, in the presence of 10 mM DBNBS, the
kinetic traces obey pseudo-first-order kinetics over the wider
p[H+] range of 1−7. Experiments above pH = 7 were not
performed because of the known instability of IrIV at high pH.23
The details of these experiments have been published,5 and
they show that the reaction obeys eq 2 with rate constants
about 50% greater than those without the addition of DBNBS:
12765
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
primary KIE and implies cleavage of the O−H bond in the rate-
limiting step. Although this result is taken as evidence of a
concerted PCET mechanism (see below), the measured KIE is
not extremely large, so it is conceivable that the measured kArOH
value includes a small contribution from a parallel sequential
PCET mechanism. Focusing on the (major) concerted PCET
component, the solvent (water) must be acting as the proton
acceptor because IrIII is not appreciably basic and no other
bases appear in the rate law. A similar KIE and mechanism were
recently reported for the oxidation of phenol by three
RuIII(bpy)3-type complexes3 and for the oxidation of hydroxyl-
amine by IrIV.24
Coupling Products. As we described in our previous
communication,5 the second step in the reaction between
phenol and IrIV is the self-reaction of the phenoxyl radicals.
This self-reaction is currently believed to occur through C−C
and C−O coupling to generate the four major expected
product isomers in Scheme 2. According to pulse radiolysis
283 nm (ε250 = 9.0 × 103 M−1 cm−1 and ε283 = 5.2 × 103 M−1
cm−1). The spectrum of 4-phenoxyphenol exhibits a weaker and
broader absorption at 278 nm with ε278 = 2.2 × 103 M−1 cm−1.
pKa Values of 4,4′-Biphenol and 2,4′-Biphenol. Spectro-
photometric titration with NaOH was used to determine the
two pKa values for 4,4′-biphenol because controversial results
were found previously.26−28 At 0.1 M ionic strength (LiClO4),
3 × 10−5 M 4,4′-biphenol was titrated from pH = 6.5 to 12.6
with various amounts of NaOH, generating the series of UV−
vis spectra shown in Figure S-3 in the Supporting Information.
These spectra show a loss of absorbance at 260 nm due to
consumption of biphenol and a gain of absorbance at 288 nm.
The titration curve at 288 nm, after volume correction, is
sigmoidal, as shown in Figure 6. Satisfactory fits of this curve to
Scheme 2. Expected Dimerization Products of the Phenoxy
Radical and Their Composition According to Reference 25
Figure 6. Titration curve of 4,4′-biphenol at 288 nm after volume
correction. μ = 0.1 M (LiClO4); T = 25 °C. The solid line is the fit to eq 4.
experiments,25 the yields of these products are 25% for 4,4′-
biphenol, 18% for 2,2′-biphenol, 44% for 2,4′-biphenol, and 9%
for 4-phenoxyphenol. The remaining 4% is other products.
The UV−vis spectra of the four coupling isomers in H2O or a
0.2% ethanol/water mixture are shown in Figure 5. 4,4′-Biphenol
a model involving a single ionizable proton could not be
obtained. On the other hand, an excellent fit was obtained with
a model that included two ionizable protons, as in eq 4.
(4)
Equation 4 expresses the total absorbance (A) as a function of
the pH, pKa1, and pKa2, the total biphenol concentration (c),
and the molar absorptivities (ε) for all three protonated and
deprotonated forms. The result of the fit to eq 4 is shown in
Figure 6, and the derived values are pKa1 = 9.66 ± 0.05 and
4
pKa2 = 10.92 ± 0.4, εHOArArO = 2.4 × 10 M−1 cm−1, ε−OArArO
−
−
= 2.7 × 104 M−1 cm−1, and εHOArArOH = 8.0 × 103 M−1 cm−1.
The pKa1 value is very close to that reported by Jonsson et al.,27
and the pKa2 value is as expected for successive pKa values.
A similar titration of 2,4′-biphenol led to the complex series of
spectra shown in Figure S-4 in the Supporting Information. The
titration curve at 310 nm (Figure 7) is sigmoidal, but the curve at
285 nm clearly shows that three species are involved. A fit of
the absorbance data at 285 nm to eq 4 yields the following values:
Figure 5. UV−vis spectra of 4,4′-/2,2′-/2,4′-biphenol and 4-
phenoxyphenol in water or a 0.2% ethanol/water solution.
has strong absorption at 263 nm with a molar absorptivity (ε 263
)
of 2.1 × 104 M−1 cm−1. The maximum absorption wavelength is
red-shifted to 279 nm when the structure changes to 2,2′-
biphenol, and its ε279 value is 5.9 × 103 M−1 cm−1. Two
absorption bands are observed for 2,4′-biphenol: 250 and
pKa1 = 9.65 ± 0.04, pKa2 = 10.96 ± 0.06, εHOArArO = 8.5 × 103
−
M−1 cm−1, ε−OArArO = 6.1 × 103 M−1 cm−1, and εHOArArOH
=
−
5.2 × 103 M−1 cm−1. A fit at 310 nm yields pKa1 = 9.68 ±
0.04, pKa2 = 10.89 ± 0.05, εHOArArO = 4.8 × 103 M−1 cm−1,
−
12766
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
Figure 8. UV−vis spectra of products of 4,4′-/2,2′-/2,4′-biphenol
oxidation. [IrIV]0 = 2.5 × 10−5 M; μ = 0.1 M (LiClO4); T = 25 °C. The
solid line is 4,4′-biphenol oxidation: [4,4′-biphenol]0 = 2.5 × 10−5 M;
p[H+] = 5.5 (0.02 M acetate buffer). The short dashed line is 2,2′-
biphenol oxidation: [2,2′-biphenol]0 = 1 × 10−3 M; p[H+] = 4.0 (0.02
M acetate buffer). The long dashed line is the 2,4′-biphenol reaction:
[2,4′-biphenol]0 = 2 × 10−4 M; [HClO4] = 0.05 M. The absorption
with a maximum peak at approximately 400 nm is assumed as the
product spectrum of the oxidation reaction. The molar absorptivity, ε,
is calculated based on the stoichiometry that 1 equiv of
biphenoquinone product is obtained from 2 equiv of IrIV.
Figure 7. Titration curves of 2,4′-biphenol at 285 and 310 nm after
volume correction. μ = 0.1 M (LiClO4); T = 25 °C. The solid lines are
the fits to eq 4.
3
ε−OArArO = 8.8 × 10 M−1 cm−1, and εHOArArOH = 4.9 × 102 M−1
−
cm−1. The agreement between the pKa values at these two
wavelengths is excellent.
The two pKa values for 2,2′-biphenol are quite different from
those of the 4,4′ and 2,4′ isomers.27 This difference is attributed
to hydrogen bonding between the two O atoms, which is only
possible for the 2,2′ isomer. The pKa1 values for all four
coupling isomers are summarized in Table 1.
In a spectrophotometric titration at p[H+] = 5.5, 4.8 × 10−5
M
Overoxidation of Phenol. As shown below, IrIV oxidizes all
four of these coupling isomers rapidly, which leads to overoxidation
in the phenol reaction. A full account of the phenol reaction is thus
dependent on the details of each of these overoxidation pathways.
1. Oxidation of 4,4′-Biphenol. Qualitatively, the reaction of
IrIV with 4,4′-biphenol is signaled by the rapid loss of absorbance at
488 nm, which is characteristic of IrIV. A concurrent absorbance
increase occurs at 398 nm, and this absorbance decays at longer
time scales. A UV−vis spectrum of the 398 nm intermediate
generated from the reaction of 2.5 × 10−5 M 4,4′-biphenol with 2.5
× 10−5 M IrIV at p[H+] = 5.5 is shown in Figure 8. This absorbance
decays with an initial rate of −1.7 × 10−4 s−1, as shown in Figure S-5
in the Supporting Information. When the same experiment is
performed at p[H+] = 2.5, the maximum absorbance at 398 nm
and its initial decay rate are close to those obtained at p[H+] =
5.5, as shown in Table S-7 in the Supporting Information. The
species absorbing at 398 nm is assigned as 4,4′-biphenoquinone,
which was claimed to be observed at 400 nm in the oxidation of
phenol by other metal complexes.29,30 Our results imply that the
decomposition of 4,4′-biphenoquinone is almost pH-independ-
ent. The initial (maximum) absorbance of 4,4′-biphenoquinone
at 398 nm yields a molar absorptivity 5.2 × 104 M−1 cm−1 in an
aqueous solution, based on the assumption that 1 mol is
produced per 2 mol of IrIV; this value for ε398 is consistent with
the reported value (≈5 × 104 M−1 cm−1).31
IrIV was added to 1.5 mL of 4.8 × 10−5 M 4,4′-biphenol, generating
the titration curves in Figure S-6 in the Supporting Information. At
398 nm, the absorbance (after correction for dilution) rises linearly
to an abrupt end point, and after the end point, the absorbance is
IV
stable. The end point corresponds to a consumption ratio of nIr /
n4,4′‑biphenol of 2.5:1. Apparently, the 4,4′-biphenoquinone product
undergoes some further overoxidation in the presence of 4,4′-
biphenol. The absorbance at 488 nm (due to IrIV) remains
essentially zero (Figure S-6b in the Supporting Information) prior
to the same abrupt end point observed at 398 nm, implying that
IrIV is fully consumed up to the end point.
The principal reaction is thus
(5)
Kinetic studies were performed with an excess of 4,4′-biphenol
over IrIV, but because of its low solubility, the initial concentrations
of 4,4′-biphenol were not high enough to ensure strictly pseudo-
first-order conditions. Thus, values of kobs were obtained from the
slope of the linear regression of the logarithm of absorbance at 488
nm within the first half-life (Figure S-7 in the Supporting
Information). Rates were determined at p[H+] = 1−7, and the
data are shown in Table S-8 in the Supporting Information. These
pH-dependent data conform to eq 2 (Figure 9), and they yield the
a
Table 1. Kinetic Data for the Reactions of IrIV with Phenol and Its Coupling Products
−1
substrate
phenol
kArOH, M−1 s−1
kArO , M s−1
pKa1
Ef(ArO•/ArO−), V
Ef(ArO•,H+/ArOH), V
−
b
f
0.77 ± 0.03
(8.0 ± 0.2) × 106
(6.5 ± 0.7) × 108
(4.0 ± 0.7) × 106
(1.3 ± 0.3) × 108
(1.1 ± 0.2) × 108
9.79
9.66
7.60
9.67
9.90
0.80
1.38
1.21
1.45
c
d
4,4′-biphenol
2,2′-biphenol
2,4′-biphenol
4-phenoxyphenol
(4.6 ± 0.4) × 104
6.6 ± 2.6
0.64
d
c
d
1.00
(4.0 ± 0.6) × 103
(1.2 ± 0.1) × 103
e
a
Ar = 4,4′-/2,2′-/2,4′-HOC6H4C6H4, 4-C6H5OC6H4. Rate constants otained by fitting the kobs results in Table S-8 in the Supporting Information to
b
c
d
e
eq 2 and with the Ka values constrained to the values given above. Reference 22. This work. Reference 27. The reported pKa1 = 9.81 at μ = 0.25
M in ref is converted to pKa1 = 9.90 at μ = 0.1 M by using the Davies equation. Ef corrected in this work from E° in ref 35.
44
f
12767
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
2,4′-biphenoquinone stoichiometry, the experimental spectrum
yields ε398 = 3.0 × 104 M−1 cm−1 and ε488 = 4.0 × 103 M−1
cm−1.
At p[H+] = 6.9 and 398 nm, the kinetic traces are triphasic,
showing a rapid rise in absorbance, a slower small-amplitude
fall, and then an even slower large-amplitude fall (Figure S-10
in the Supporting Information). The first phase occurs on the
same time frame as the initial absorbance loss at 488 nm (see
below). This triphasic behavior is thus attributed to the rapid
production of 2,4′-biphenoquinone in the first phase, followed
by its biphasic decay. Because of the relatively rapid decay in
the second phase, it was not possible to obtain an accurate
value for ε398 at this pH, a value of only 1.6 × 104 M−1 cm−1
corresponding to the maximum absorbance.
Figure 9. Plot of kobs/[substrate]tot versus p[H+]. [IrIV]0 = 2.5 × 10−5
M; μ = 0.1 M (LiClO4); T = 25 °C. At the p[H+] values between 1.0
and 2.2, p[H+] = −log [HClO4]; the p[H+] values between 2.4 and 3.6
were maintained by a 0.02 M monochloroacetate buffer; the p[H+]
values between 3.7 and 5.4 were maintained by a 0.02 M acetate
buffer; the p[H+] values between 5.5 and 7.1 were maintained by a
0.02 M cacodylate buffer. The solid lines are fits to eq 2.
At 488 nm, the kinetic traces exhibit biphasic decay. Double-
exponential fits were used to obtain the fast pseudo-first-order
rate constants for the consumption of IrIV (kobs) as well as the
slower first-order rate constants for the decay of 2,4′-
biphenoquinone (kobs,2) (shown in Table S-8 in the Supporting
Information). The p[H+] dependence of the IrIV reduction was
studied at p[H+] = 1−7. A fit of the values of kobs to eq 2
generated kArOH = (4.0 ± 0.6) × 103 M−1 s−1 and kArO = (1.3 ±
−
following rate constants: kArOH = (4.6 ± 0.4) × 104 M−1 s−1 and
0.3) × 108 M−1 s−1 (Table 1).
kArO = (6.5 ± 0.7) × 108 M−1 s−1 (Table 1).
−
4. Oxidation of 4-Phenoxyphenol. Kinetic studies of the
reaction of 0.3 mM 4-phenoxyphenol with 25 μM IrIV were
carried out in the range p[H+] = 1−7. The decay traces of IrIV
at 488 nm display excellent pseudo-first-order behavior under
these conditions (Table S-8 in the Supporting Information). A
fit of the data to rate law (2) led to the following parameters:
2. Oxidation of 2,2′-Biphenol. UV−vis spectroscopy was
used to detect the products of the reaction between 1 × 10−3
M
2,2′-biphenol and 2.5 × 10−5 M IrIV at p[H+] = 4.0. As shown in
Figure 8, a weak absorbance peak appears at 406 nm with a
broad shoulder observed in the range 450−600 nm. This
spectrum is attributed to the absorption of 2,2′-biphenoqui-
none, with ε406 = 1.4 × 104 M−1 cm−1 in an aqueous solution. It
decomposes slowly as shown in Figure S-8 in the Supporting
Information.
kArOH = (1.2 ± 0.1) × 103 M−1 s−1 and kArO = (1.1 ± 0.2) ×
−
108 M−1 s−1 (Table 1).
Evidence of Overoxidation of Phenol. As shown in
Figure 10, the reaction between 0.0443 M phenol and 1 × 10−4
The kinetic experiments on the oxidation of 2,2′-biphenol by
IrIV were performed by using stopped-flow spectroscopy to
monitor the loss of IrIV at 488 nm at p[H+] = 1−7 with at least
a 10-fold molar excess of biphenol over IrIV. However, because
of the absorbance of products at 488 nm and their subsequent
decay, deviations from pseudo-first-order kinetics were
observed at some p[H+]. Therefore, the initial rate constants
were calculated from the slopes of semilog plots of (A − A∞) vs
t. The reaction rates increase dramatically as the p[H+]
increases, although the oxidation is extremely slow at low
p[H+], as shown in Figure 9. Values for kArOH and kArO
−
obtained by fitting the data in Table S-8 in the Supporting
Information to eq 2 are 6.6 ± 3 and (4.0 ± 0.7) × 106 M−1 s−1,
respectively.
3. Oxidation of 2,4′-Biphenol. The products of the reaction
between 2 × 10−4 M 2,4′-biphenol and 2.5 × 10−5 M IrIV at
[HClO4] = 0.05 M were investigated by UV−vis spectroscopy.
As shown in Figure S-9 in the Supporting Information, an
intermediate that has an absorbance maximum at 398 nm is
produced rapidly, and then it undergoes decomposition with a
first-order rate constant of 7.2 × 10−3 s−1 (3.5 × 10−2 s−1 was
obtained by a stopped-flow instrument at 488 nm, as shown in
Table S-8 in the Supporting Information). The first UV−vis
spectrum obtained after mixing (∼10 s) is shown in Figure 8; it
has a main peak at 398 nm and a minor peak at 488 nm
(overlapping with that of IrIV). Assignment of this spectrum to
2,4′-biphenoquinone is supported by density functional theory
calculations at the B3LYP/6-31G* level that produce an
electronic spectrum with a major peak at ∼320 nm and a
secondary peak at 460 nm. With the assumption of a 2:1 IrIV/
Figure 10. UV−vis spectra of the reaction between 4.43 × 10−2
M
phenol and 1 × 10−4 M IrIV. p[H+] = 2.5 (0.02 M monochloroacetate
buffer); μ = 0.1 M (LiClO4); T = 25 °C; 5 s interval between spectra.
The inset shows the kinetic traces at 398 nm (solid line) and 488 nm
(dashed line).
M IrIV at p[H+] = 2.5 displays a loss of absorbance at 488 nm,
signaling the consumption of IrIV, and there is a concurrent rise
in absorbance at 398 nm that is assigned to the formation of
biphenoquinones arising from oxidation of the biphenol
coupling products. Figure S-11 in the Supporting Information
displays the absorbance at 398 nm on a longer time scale to
demonstrate the rise−fall character of the signal. A double-
exponential nonlinear regression fit generates two observed rate
constants: the fast one is 3.6 × 10−2 s−1, and it corresponds to
12768
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
the consumption of IrIV at 488 nm, while the slow one is 3.7 ×
10−3 s−1, which corresponds to the decay of 2,4′-biphenoqui-
none (Table S-7 in the Supporting Information).
(9)
DISCUSSION
■
The initial steps in the oxidation of phenol by IrIV are now well
established.2,5 First, depending on the pH, either phenol itself
or its conjugate base undergoes reversible one-electron
oxidation to produce the phenoxyl radical and [IrCl6]3−.
Next, the phenoxyl radicals undergo bimolecular C−C coupling
to produce biphenols and C−O coupling to produce
phenoxyphenol. Crucial evidence for this mechanism includes
the pH dependence of the rates, the kinetic inhibition by IrIII
and its removal by the spin-trap DBNBS, and the simulation of
these effects with a kinetic model employing realistic values for
the four redox rate constants and the radical coupling overall
rate constant. Early on, Cecil and Littler recognized that this
model is incomplete because it fails to account for the observed
production of diphenoquinone;2 they proposed that the initial
diphenol coupling products were oxidized further, generating
diphenoquinone. Our in situ UV−vis data provide further
evidence for this overoxidation mechanism. Our current studies
of the direct oxidations of the four principal phenoxyl coupling
products allow quantitative tests of the overoxidation
mechanism, and they provide an explanation of the apparent
pH-dependent rate constant (k° in eq 3).
(10)
Here, direct oxidation of the biphenol generates the
biphenosemiquinone through a concerted electron−proton-
transfer mechanism. This step is uphill and, hence, is shown as
reversible because E°(4,4′-HOC6H4C6H4O•,H+/4,4′-
HOC6H4C6H4OH) is 1.21 V [calculated from E°(ArO•,H+/
ArOH) = E°(ArO•/ArO−) + 0.059pKa with E°(ArO•/ArO−) =
0.64 V vs NHE27]. Oxidation of the conjugate base of biphenol
occurs through simple electron transfer; both of these are in
direct analogy with the oxidation of phenol. The fate of the
semiquinone could be either disproportionation, as in eq 9, or
further oxidation, as in eq 10. Disproportionation was reported
previously for the semiquinone in the absence of good
oxidants;30 although the disporportionation rate constant was
not determined, it can be expected to be large because the
reaction has a large driving force: E°(4,4′-biphenoquino-
ne,2H+/4,4′-biphenol) = 0.94 V vs NHE32 and E°(4,4′-
HOC6H4C6H4O•,H+/4,4′-HOC6H4C6H4OH) = 1.21 V vs
NHE. On the other hand, semiquinones are easily oxidized,34
and the oxidation of 4,4′-biphenosemiquinone is quite
favorable: E°(4,4′-biphenoquinone,H+/4,4′-biphenosemiqui-
none) = 0.67 V (as calculated from the above data). Moreover,
the semiquinone is relatively acidic with a pKa of 4,4′-
HOC6H4C6H4O• of 6.3.27 Clearly, a large rate constant can
be anticipated for ksemi. The low steady-state concentration of
the semiquinone ensures that its reaction with IrIV will be
dominant.
As shown in Figure 9, all four coupling products are oxidized
by IrIV with rates that are highly sensitive to the pH. Moreover,
they all (and phenol itself) have two-term rate laws, as given by
eq 2. A notable difference between the oxidations of the
biphenols and phenol is that the biphenols undergo net two-
electron oxidation to yield quinones, while phenol undergoes
net one-electron oxidation to yield radical coupling products.
Evidently, the biphenols achieve their two-electron oxidations
through one-electron oxidation of the biphenols and their
conjugate bases to form semiquinones, followed by the rapid
one-electron conversion of the semiquinones to the quinones.
Radical coupling occurs in the case of phenol because the
phenoxyl radical is not easily oxidized to its cation.
2,2′-Biphenol is more acidic than 4,4′-biphenol by about 2 pK
units. This difference is attributed to the formation of an
internal hydrogen bond between the two O atoms in the 2,2′-
The overall oxidation of 4,4′-biphenol by IrIV, as in eq 5, is
close to thermoneutral at pH = 1 because E°(4,4′-
biphenoquinone,2H+/4,4′-biphenol) = 0.94 V vs NHE32 while
E°(IrIV/IrIII) = 0.893 V.33 This reaction becomes more
favorable as the pH increases. In view of the overall
stoichiometry (eq 5) and the two-term rate law (eq 2), a
reasonable reaction mechanism is
−
biphenolate monoanion. The value of kArO is about a factor of
100 less for 2,2′-biphenol than for 4,4′-biphenol (for reasons
discussed below). The result of these two effects is that the
rates of oxidation of the two biphenol isomers are quite similar
at pH = 7. On the other hand, the value for kArOH is 10 000-fold
less for 2,2′-biphenol, so the pH-independent region of the rate
law is limited to only quite acidic conditions (Figure 9).
Despite these quantitative differences, we infer the same
qualitative mechanism for oxidation of the two substrates.
Unlike 2,2′-biphenol, 2,4′-biphenol is incapable of internal
hydrogen bonding and thus resembles 4,4′-biphenol quite
closely in its oxidation by IrIV.
(6)
(7)
(8)
4-Phenoxyphenol has a pKa value and values for kArOH and
−
kArO that are similar to those of 4,4′-biphenol, and hence its
rate−pH profile is similar. Upon oxidation, it yields the 4-
phenoxyphenoxyl radical rather than a semiquinone, and hence
the ultimate product is due to radical coupling.
The above considerations lead to the following mechanism
for the overall oxidation of phenol:
12769
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
The first step in this mechanism corresponds to the
reversible concerted proton-electron-transfer oxidation of
neutral phenol, and the second step is the reversible outer-
sphere electron transfer from the phenolate anion. The acid/
base reaction relating phenol and phenolate (eq 13) is assumed
(11)
to be at equilibrium because of rapid proton transfer. Values for
−
k
ArOH and kArO are directly measured as described above, while
−
values for k−ArOH and k−ArO are calculated from the forward
rate constants and the Ef values at μ = 0.1 M: Ef(IrIV/IrIII) =
0.893 V,33 Ef(C6H5O•/C6H5O−) = 0.80 V, corrected from
0.79 V at μ = 0.0 M35 by log γ = −Azi2μ1/2/(1 + μ1/2), and
Ef(C6H5O•,H+/C6H5OH) = 1.38 V [calculated from
Ef(ArO•,H+/ArOH) = Ef(ArO•/ArO−) + 0.059pKa]. Dimeriza-
tion of the phenoxyl radical, which can be partially rate-limiting
under certain conditions, forms the four major coupling
isomers: 4,4′-, 2,2′-, and 2,4′-biphenol and 4-phenoxyphenol.
Under the assumption that the dimerization rate constant
(12)
(13)
(14a)
(14b)
producing each species is proportional to its yield, kdim,4,4
,
(14c)
(14d)
k
dim,2,2, kdim,2,4, and kdim,pop in eqs 14a−14d are 2.88 × 108,
2.07 × 108, 5.06 × 108, and 1.04 × 108 M−1 s−1, respectively,
according to the reported overall 2kdim value36 and yield of each
isomer.25
Further one-electron oxidations of the phenol coupling
products are shown in eqs 15a−15d without reference to the
pH dependence evident in Figure 9. In our simulations
,
(15a)
(15b)
(15c)
generate simulated pseudo-first-order rate constants, kobs,sim
.
Figure 11a (data from Table S-11 in the Supporting
(15d)
(16a)
(16b)
Figure 11. Comparative pH dependence of the phenol reaction
experimental data and simulation results: (a) experimental data and
simulated results with overoxidation; (b) simulated results with and
without overoxidation; (c) simulated results with overoxidation and its
three-term rate law curve fit.
(16c)
(16d)
Information) shows that these results give an excellent fit to
the experimental pH dependence of the reaction. Figure 11b
compares the simulated results from the full mechanism with
12770
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
those obtained when overoxidation (eqs 15a−15d and 16a−
16d) is excluded from the mechanism, and it shows that
overoxidation increases the net rates at low pH but less so at
higher pH. As a result, overoxidation leads to a pH dependence
that deviates systematically from the simple two-term rate law
in eq 2. Figure 11c shows that the simulations with
overoxidation included give an excellent fit to the three-term
rate law (eq 3), with fitted values of kArOH = 0.38 ± 0.14 M−1
scavenge the phenoxyl radicals completely, preventing their
back-reaction with IrIII and ensuring that the rate-limiting step
in the oxidation by IrIV is the initial oxidation of phenol.
A remaining question is the origin of the pH dependence of
the [Ru(bpy) ]3+/phenol reaction reported by Sjodin et al.,
̈
3
which required the three-term eq 3.6 First, we note that Bonin
et al. were unable to reproduce the effect and found the two-
term eq 2 to be adequate.3 Second, the degree of overoxidation
to be expected depends on the phenol concentration, and this
concentration was not disclosed in the original report. Third,
the degree of overoxidation should also depend on the rate
constants for [Ru(bpy)3]3+ oxidation of the phenolic coupling
products, and these rate constants are unknown. However, it
can be shown that a significant degree of overoxidation should
occur in acidic media if favorable choices are made for the
reactant concentrations and rate constants.
s−1, kArO = (4.9 ± 0.2) × 106 M−1 s−1, and k° = (8.2 ± 7) ×
−
10−3 M−1 s−1. Although the statistical uncertainty in k° is large,
the value of k° is in good agreement with the value derived
from the experimental data. Evidently, the origin of the k° term
in the fit of the empirical rate law (3) to the experimental
results is the pH-dependent influence of overoxidation on the
rate of consumption of IrIV.
If overoxidation were not occurring, the rate constants in
eq 2 would agree exactly with those in the mechanism,
corresponding to a stoichiometric factor of unity in the rate law.
An effect of overoxidation is that the rate constants derived
from eq 2 are somewhat larger than the rate constants for the
elementary steps, which means that overoxidation introduces
stoichiometric factors greater than unity. These stoichiometric
factors, calculated as the ratio ksim(with overoxidation)/
Rate Constant Trends. Table 1 shows that the values for
kArO range from 4 × 106 to 6.5 × 108 M−1 s−1 for the five
−
phenolate ions considered in this study. These rate constants
correspond to electron-transfer reactions, and hence it is
reasonable to use the cross-relationship of Marcus theory, as
given in eqs 20−23, to rationalize their variations.38
k
sim(without overoxidation), decrease systematically from 1.75
(20)
at pH 2.46 to 1.10 at pH 5.36 (Table S-11 in the Supporting
Information). They also depend somewhat on the phenol
concentration. It is the pH dependence of the stoichiometric
factor that leads to the k° term in rate law (3).
Another consequence of the pH-dependent stoichiometric
factor is that the yields of the reaction products are also pH-
dependent. As shown in Table S-12 in the Supporting
Information, at pH 2.46 the simulated yield of the initial
coupling products per 1 mol of IrIV (corrected for the 1:2
stoichiometric ratio) is 9.6% and the yield of the overoxidation
products (corrected for the 1:4 stoichiometric ratio) is 90.4%.
However, at pH 5.36 these yields cross over to 72.9% and
27.1%, respectively.
(21)
(22)
(23)
In these equations, k12 is the cross-electron-transfer rate
constant (kArO = 8.0 × 106 M−1 s−1 for phenoxide), and k11
−
and k22 are the self-exchange rate constants of the ArO•/ArO−
and IrIV/IrIII redox couples, respectively. A value for k22 of 2 ×
39 and 1 × 1011 M−1 s−1 is
105 M−1 s−1 is used in the calculation,
In the presence of the spin-trap DBNBS, three more steps
are added to the above mechanism:
used for Z, the collision frequency.40 Zi and Zj are ionic charges
of the reactants, R is the ideal gas constant, and r is the center-
to-center distance between the two reactants when they are
approaching each other. The radii of [IrCl6]2− and C6H5O− are
4.141 and 2.5 Å, respectively, estimated from Corey−Pauling−
Koltun atomic models. μ is the ionic strength. wij is the
electrostatic energy between reactants i and j. If the distance r is
in angstroms and μ in molar, then w12 can be calculated
according to eq 23 in kilojoules per mole. With all of these
parameters and the experimental values of k12 and K12, k11 is
calculated from the above equations as 2.3 × 106 M−1 s−1. The
phenoxide anion and the phenoxyl radical are predicted to have
quite similar structures, with the largest difference being a 0.015
Å change in the C−O bond length.42 Such a small structural
change should not contribute significantly to the self-exchange
barrier. Estimates of the solvation contribution to the self-
exchange barrier (ΔGos⧧) are difficult to make because the
phenolate charge is highly localized on the oxygen end of the
(17)
(18)
(19)
As mentioned above, only the monomer of DBNBS can
scavenge the phenoxyl radical. Therefore, the dimerization step
of DBNBS with equilibrium constant Kdim,DBNBS of 1.3 × 10−3
M
should be included in our mechanism.21 The phenoxyl radical is
scavenged by DBNBS to form an adduct, which undergoes a
rapid oxidation by IrIV, as in eqs 18 and 19. Competition
between the scavenging of the phenoxyl radical by DBNBS and
its dimerization is evident in the kinetic saturation dependence
on [DBNBS], as shown in Table S-2 in the Supporting
Information. Simulations of this saturation effect at p[H+] =
1.3, performed with the mechanism in Table S-9 in the
Supporting Information, are sensitive to the value of kDBNBS. As
shown in Table S-13 in the Supporting Information, a good fit
is obtained when the rate constant of C6H5O•/DBNBS adduct
formation, kDBNBS, is 2.0 × 105 M−1 s−1. This rate constant is
large enough to enable 10 mM concentrations of DBNBS to
⧧
anion; however, this charge localization should cause ΔGos to
be larger than that for a spherical anion of comparable size. As a
result, the significant overall self-exchange barrier implied by
the k11 above value is attributed principally to the solvent
barrier. A significantly greater value for k11 of 1.9 × 108 M−1 s−1
was previously measured directly by electron spin resonance
line broadening.43 Apparently, the solvent barrier is reduced in
12771
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
(4) Bonin, J.; Costentin, C.; Louault, C.; Robert, M.; Saveant, J. M. J.
Am. Chem. Soc. 2011, 133, 6668−6674.
the actual self-exchange process, possibly through a weak
association between the radical and the phenoxide anion.
In the case of the 4,4′-biphenoxide anion [E°(ArO•/ArO−) =
(5) Song, N.; Stanbury, D. M. Inorg. Chem. 2008, 47, 11458−11460.
0.64 V],27 kArO is considerably larger than that for phenoxide
(6) Sjodin, M.; Irebo, T.; Utas, J. E.; Lind, J.; Meren
́
yi, G.; Åkermark,
̈
−
B.; Hammarstrom, L. J. Am. Chem. Soc. 2006, 128, 13076−13083.
̈
itself. Most of this increase can be ascribed to the greater
driving force for the reaction, but there is also some
(7) Kauffman, G. B.; Teter, L. A. Inorg. Synth. 1966, 8, 223−227.
(8) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon: New York, 1988.
contribution from a greater self-exchange rate constant (k11
=
4 × 107 M−1 s−1). A reduced self-exchange barrier can be
attributed to the delocalized electronic structure of the
semiquinone radical.
(9) Kaur, H.; Leung, K. H. W.; Perkins, M. J. J. Chem. Soc., Chem.
Commun. 1981, 142−143.
(10) Hamilton, L.; Nielsen, B. R.; Davies, C. A.; Symons, M. C. R.;
Winyard, P. G. Free Radical Res. 2003, 37, 41−49.
(11) Kanetani, F.; Yamaguchi, H. Bull. Chem. Soc. Jpn. 1981, 54,
3048−3058.
(12) Konaka, R.; Sakata, S. Chem. Lett. 1982, 411−414.
(13) Amblard, F.; Govindarajan, B.; Lefkove, B.; Rapp, K. L.;
Detorio, M.; Arbiser, J. L.; Schinazi, R. F. Bioorg. Med. Chem. Lett.
2007, 17, 4428−4431.
The 2,2′-biphenoxide anion is considerably more difficult to
oxidize [E°(ArO•/ArO−) = 1.00 V]27 than phenoxide, but the
−
two substrates have quite similar kArO values. A large self-
exchange rate constant of 3 × 108 M−1 s−1 is required by these
data. It is conceivable that the internal hydrogen bonding in the
2,2′ isomer reduces the degree of hydrogen bonding with the
solvent and thus leads to a greater k11 value. Corresponding
discussions of the 2,4′-biphenoxide and 4-phenoxyphenoxide
rates will require determination of the relevant E° values.
We have previously argued that the direct oxidation of
phenol by IrIV (eq 11, kArOH) involves proton transfer to the
solvent in concert with electron transfer (H2O-PCET).5 This
mechanistic assignment was based largely on the significant
solvent deuterium KIE, the high acidity of the ArOH•+ radical
cation, and the low basicity of IrIII. This conclusion is
strengthened by the CV measurements described above,
which show that IrIII is not significantly protonated in 1 M
H+. Further support for this H2O-PCET mechanism is
provided by a linear free-energy relationship that relates the
rates of oxidation of phenol to the E° values for IrIV and a set of
three RuIII oxidants.3 It seems reasonable to assign a H2O-
PCET mechanism to all of the phenol reactions in Table 1.
Recently, Bonin et al. have developed a theoretical treatment of
CPET reactions with water (and other bases) as the proton
acceptor;4 this theory predicts, in the absence of other effects,
that these reactions should display a typically Marcusian
dependence of the rates on the driving forces. Qualitatively,
Table 1 shows that this expectation is met in comparing phenol
with 4,4′-biphenol and in comparing 4,4′-biphenol with 2,2′-
biphenol. However, 2,2′-biphenol reacts 9 times faster than
phenol, even though it is 0.07 V more difficult to oxidize. This
apparent contradiction may signal the importance of several
other variables in the theory of H2O-PCET.
(14) Sarala, R.; Stanbury, D. M. Inorg. Chem. 1990, 29, 3456−3460.
(15) Sawyer, D. T.; Sobkowiak, A.; Roberts, J. L. Electrochemistry for
Chemists, 2nd ed.; John Wiley and Sons: New York, 1995; p 192.
(16) Spartan ’08; Wavefunction, Inc.: Irvine, CA, 2008.
(17) Bruhn, H.; Nigam, S.; Holzwarth, J. F. Faraday Discuss. Chem.
Soc. 1982, 74, 129−140.
(18) Bonin, J.; Costentin, C.; Robert, M.; Saveant, J. M. Org. Biomol.
Chem. 2011, 9, 4064−4069.
(19) Irebo, T.; Reece, S. Y.; Sjodin, M.; Nocera, D. G.;
̈
Hammarstrom, L. J. Am. Chem. Soc. 2007, 129, 15462−15464.
̈
(20) Zalomaeva, O. V.; Trukhan, N. N.; Ivanchikova, I. D.;
Panchenko, A. A.; Roduner, E.; Talsi, E. P.; Sorokin, A. B.; Rogov,
V. A.; Khodeeva, O. A. J. Mol. Catal. A 2007, 277, 185−192.
(21) Ide, H.; Hagi, A.; Ohzumi, S.; Murakami, A.; Makino, K.
Biochem. Int. 1992, 27, 367−372.
(22) Martell, A. E.; Smith, R. M.; Motekaitis, R. J. NIST Critically
Selected Stability Constants of Metal Complexes Database, version 7.0;
U.S. Department of Commerce: Gaithersburg, MD, 2003.
(23) Fine, D. A. Inorg. Chem. 1969, 8, 1014−1016.
(24) Makarycheva-Mikhailova, A. V.; Stanbury, D. M.; McKee, M. L.
J. Phys. Chem. B 2007, 111, 6942−6948.
(25) Ye, M.; Schuler, R. H. J. Phys. Chem. 1989, 93, 1898−1902.
(26) Das, T. N. J. Phys. Chem. A 2001, 105, 5954−5959.
(27) Jonsson, M.; Lind, J.; Meren
́
yi, G. J. Phys. Chem. A 2002, 106,
4758−4762.
(28) Jonsson, M.; Lind, J.; Meren
́
yi, G. J. Phys. Chem. A 2003, 107,
5878−5879.
(29) Al-Ajlouni, A.; Bakac, A.; Espenson, J. H. Inorg. Chem. 1993, 32,
5792−5796.
(30) Papina, A. A.; Koppenol, W. H. Chem. Res. Toxicol. 2007, 20,
ASSOCIATED CONTENT
* Supporting Information
Thirteen tables of kinetic data, four figures of spectra, six figures
of kinetic data, and one figure of spectrophotometric titration
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
2021−2022.
■
(31) Al-Ajlouni, A. M.; Shawakfeh, K. Q.; Rajal, R. Kinet. Catal. 2009,
S
50, 88−96.
(32) Pelizzetti, E.; Mentasti, E. J. Inorg. Nucl. Chem. 1977, 39, 2227−
2230.
(33) Margerum, D. W.; Chellappa, K. L.; Bossu, F. P.; Burce, G. L. J.
Am. Chem. Soc. 1975, 97, 6894−6896.
(34) Neta, P.; Grodkowski, J. J. Phys. Chem. Ref. Data 2005, 34, 109−
199.
(35) Lind, J.; Shen, X.; Eriksen, T. E.; Meren
1990, 112, 479−482.
(36) Tripathi, G. N. R.; Schuler, R. H. J. Chem. Phys. 1982, 88, 253−
255.
(37) Binstead, R. A.; Jung, B.; Zuberbuhler, A. D. Specfit/32 Global
Analysis System, version 3.0; Spectrum Software Associates:
Marlborough, MA, 2000.
AUTHOR INFORMATION
Corresponding Author
*E-mail: stanbury@auburn.edu.
■
́
yi, G. J. Am. Chem. Soc.
ACKNOWLEDGMENTS
We thank the NSF for support of this research.
■
REFERENCES
(38) Zuckerman, J. J. Inorganic Reactions and Methods; VCH:
■
Deerfield Beach, FL, 1986; Vol. 15, pp 13−47.
(1) Huynh, M. H. V.; Meyer, T. J. Chem. Rev. 2007, 107, 5004−5064.
(2) Cecil, R.; Littler, J. S. J. Chem. Soc. B 1968, 1420−1427.
(3) Bonin, J.; Costentin, C.; Louault, C.; Robert, M.; Routier, M.;
(39) Hurwitz, P.; Kustin, K. Trans. Faraday Soc. 1966, 62, 427−432.
(40) Ram, M. S.; Stanbury, D. M. J. Phys. Chem. 1986, 90, 3691−
3696.
́
Saveant, J.-M. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 3367−3372.
12772
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773

Inorganic Chemistry
Article
(41) Sun, J.; Stanbury, D. M. J. Chem. Soc., Dalton Trans. 2002, 785−
791.
́
(42) McDonald, W. J.; Einarsdottir, O. J. Phys. Chem. A 2008, 112,
11400−11413.
(43) Schuler, R. H.; Neta, P.; Zemel, H.; Fessenden, R. W. J. Am.
Chem. Soc. 1976, 98, 3825−3831.
(44) Stradins, J.; Hasanli, B. J. Electroanal. Chem. 1993, 353, 57−69.
12773
dx.doi.org/10.1021/ic201897m|Inorg. Chem. 2011, 50, 12762−12773