Direct determination of equilibrium potentials for hydrogen oxidation/production by open circuit potential measurements in acetonitrile

Article abstract of DOI:10.1021/ic302461q

Open circuit potentials were measured for acetonitrile solutions of a variety of acids and their conjugate bases under 1 atm H₂. Acids examined were triethylammonium, dimethylformamidium, 2,6-dichloroanilinium, 4-cyanoanilinium, 4-bromoanilinium, and 4-anisidinium salts. These potentials, along with the pK_a values of the acids, establish the value of the standard hydrogen electrode (SHE) potential in acetonitrile as -0.028(4) V vs the ferrocenium/ferrocene couple. Dimethylformamidium forms homoconjugates and other aggregates with dimethylformamide; open circuit potentials (OCPs) were used to quantify the extent of these reactions. Overpotentials for electrocatalytic hydrogen production and oxidation were determined from open circuit potentials and voltammograms of acidic or basic catalyst solutions under H₂. For these solutions, agreement between OCP values and potentials calculated using the Nernst equation is within 12 mV. Use of the measured equilibrium potential allows direct comparison of catalytic systems in different media; it requires neither pK_a values, homoconjugation constants, nor the SHE potential.

Full text of DOI:10.1021/ic302461q

Article
pubs.acs.org/IC
Direct Determination of Equilibrium Potentials for Hydrogen
Oxidation/Production by Open Circuit Potential Measurements
in Acetonitrile
John A. S. Roberts* and R. Morris Bullock
Center for Molecular Electrocatalysis, Chemical and Materials Sciences Division, P.O. Box 999, K2-57, Pacific Northwest National
Laboratory, Richland, Washington 99352, United States
S
* Supporting Information
ABSTRACT: Open circuit potentials were measured for acetonitrile
solutions of a variety of acids and their conjugate bases under 1 atm H2.
Acids examined were triethylammonium, dimethylformamidium,
2,6-dichloroanilinium, 4-cyanoanilinium, 4-bromoanilinium, and 4-
anisidinium salts. These potentials, along with the pKa values of the
acids, establish the value of the standard hydrogen electrode (SHE)
potential in acetonitrile as −0.028(4) V vs the ferrocenium/ferrocene
couple. Dimethylformamidium forms homoconjugates and other
aggregates with dimethylformamide; open circuit potentials (OCPs)
were used to quantify the extent of these reactions. Overpotentials for
electrocatalytic hydrogen production and oxidation were determined
from open circuit potentials and voltammograms of acidic or basic
catalyst solutions under H2. For these solutions, agreement between
OCP values and potentials calculated using the Nernst equation is within 12 mV. Use of the measured equilibrium potential
allows direct comparison of catalytic systems in different media; it requires neither pKa values, homoconjugation constants, nor
the SHE potential.
“perfect” catalyst operating under steady-state conditions will
afford a wave with zero overpotential, i.e., a half-peak potential
coincident with the equilibrium potential EBH . A sigmoidal
INTRODUCTION
■
Research toward sustainable, cost-effective energy technologies
is focused on the interconversion of electrical and chemical
energy. Catalysts for this interconversion are characterized by
the relationship between the rate of catalysis, indicated by an
increased current on adding a catalytic substrate, and the
overpotential, defined as the difference between the potential of
an electrode mediating a given half-reaction and the equilibrium
potential for that half-reaction.1 The simplest reaction of this
kind interconverts a Brønsted acid with hydrogen and the
+ 2
response is not always obtained in these catalytic reactions;
however, provided the current increases rapidly over a limited
potential range, estimates of the half-peak potential Ep/2 may
be obtained from this data by various methods.3 The problem
+
of determining EBH is more difficult and thus contributes
substantially to the uncertainty in determining overpotentials.
The standard-state potential E°BH+ for reaction 1 may be
computed according to Scheme 1 with water or acetonitrile, as
those solvents have both a pKa scale and a value for the
potential E°H+ of the standard hydrogen electrode (SHE).
+
corresponding base (eq 1); EBH is its equilibrium (thermody-
namic) potential.
+
EBH
1
2
BH(+soln) + e− HooooooI B(soln)
+
H2(g)
(1)
Scheme 1. Thermochemical Cycle Relating E°BH+ with pKa
and EH°+
The cyclic voltammograms in Figure 1 show a molecular
electrocatalyst with no added acid and a larger sigmoidal wave
obtained on adding acid, indicating reduction of protons to
produce hydrogen. This response represents “pure kinetic”
conditions as described by Savea
́
nt,1 wherein the maximum
current depends on the rates of chemical steps proceeding at
steady state, i.e. with rates that do not vary with time. At a
potential of Ep/2 the catalytic current is at half its maximum
+
value; the overpotential at Ep/2 (the difference between EBH
and Ep/2) is characteristic of the catalyst under the conditions
employed. For electrocatalyst systems exhibiting kinetic
Received: November 9, 2012
Published: March 14, 2013
+
control, the comparison of Ep/2 with EBH is apt since a
© 2013 American Chemical Society
3823
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
+
under the same conditions, directly measured values of EBH
permit the straightforward and precise determination of
overpotentials.
Our results reported here also lead to a revised value for the
SHE in acetontrile: using data from four different acid−base
systems in acetonitrile and extrapolating to infinite dilution, we
have calculated E°H+ by the thermochemical cycle shown in
Scheme 1. The consistency among these results and the
observed adherence to eq 2 demonstrate that the OCP is the
equilibrium potential for reaction 1. The OCP measurement
may also be used to obtain equilibrium constants for acid−base
aggregation reactions, the reversible formation of solution
species by interaction of either an acid or a base with one or
more other molecules. We have extracted aggregation equilibrium
constants for [(DMF)H]+OTf− in acetonitrile and have used these
+
to calculate EBH
.
+
Comparison between values of EBH obtained by OCP
measurements and via the Nernst equation illustrates that
either method affords a reliable basis for overpotential
determination. However, reliable inputs to the Nernst equation
are not universally available, and improving catalyst perform-
ance and understanding catalyst−medium interactions may
involve a departure from well-characterized media. For
example, we recently reported a system consisting of a catalyst
dissolved in acidic ionic liquid−water mixtures of varying water
content, for which no estimates of pKa or E°H+ were available.8
Figure 1. Cyclic voltammograms of [Ni(PPh2NPh2)2](BF4)2 in
−
acetonitrile (nBu4N+PF6 ) under H2 (1 atm, black trace) with 1:1
[(DMF)H]+OTf−:DMF (0.40 M, red trace), collected with υ = 0.1 V
s−1. Fc+/0 is the half-wave potential of the ferrocenium/ferrocene redox
couple. The value of icat was selected as described in ref 3b. Ep/2
+
denotes the half-peak potential of the catalytic wave. EBH denotes the
+
We therefore determined the equilibrium potentials EBH for
equilibrium potential for reaction 1.
+
these mixtures by OCP measurements. Directly measuring EBH
obviates the need for the assumptions inherent in calculating
+
EBH , requires neither an estimate of EH°+ nor a pKa scale, and
permits comparison of catalysts operating in different solvents,
with different acid−base systems, on a strictly equal footing.
RESULTS
■
Platinum metal catalyzes both the oxidation and production of
H2 (reaction 1) in a variety of media with a wide range of acid−
base systems.9 This reaction is presented as involving one
cationic acid “BH+” and one neutral base “B,” but the methods
presented here also apply to other acid−base systems, such as
those consisting of neutral acids and anionic bases. Examining
OCP values as a function of acid and base concentrations for a
variety of acid−base systems in acetonitrile verifies that, at open
circuit, the potential of the platinum electrode is the
equilibrium potential of reaction 1.
+
The potential EBH is related to the standard state potential
EB°H+ by the Nernst equation (eq 2), where Q = αBαH 1/2/αBH
.
+
2
2.303RT
+
°
+
EBH = EBH
−
log(Q)
(2)
F
A robust pKa scale spanning more than 28 units exists for
acetonitrile.4 However, published values3a,5 of E°H+ range from
−0.034 to −0.26 V vs Fc+/0 (the formal potential of the
ferrocenium/ferrocene redox couple). As more efficient (lower
overpotential) catalyst systems are reported, this 230 mV
uncertainty (5.2 kcal mol−1 at 25 °C) becomes increasingly
First we describe the OCP measurement and factors
influencing reproducibility and establish that the [acid]:[base]
ratio influences the OCP as predicted by eq 2. We then report
the results of three types of studies: (1) We examine how the
OCP of 1:1 [acid]:[base] mixtures varies with concentration.
By extrapolating to infinite dilution using acids of known pKa,
we obtain an estimate of E°+ for acetonitrile (Scheme 1). The
acid−base systems examinHed in these studies are [(DMF)-
significant.3a Moreover, computing EBH using eq 2 and
+
assuming that the analytical concentrations of acid and base
are equal to their activities introduces errors in cases where acid
and base molecules aggregate, as with homoconjugation. These
effects are more pronounced at high substrate concentrations,
which are often required to obtain steady state kinetics in
electrocatalytic production or oxidation of hydrogen.
H]+OTf−:DMF, Et3NH+BF4 :Et3N, and four aryl-substituted
−
anilinium:aniline systems, all of which have been used as
substrates for hydrogen production/oxidation in acetonitrile
solution.3b,10 (2) We determine overpotentials at Ep/2 for
catalytic systems by voltammetry and OCP measurements,
using solutions of the same composition for both measure-
ments. This is the most precise method for anchoring
overpotential scales. (3) We examine OCP values of Et3N
solutions under hydrogen but without an explicitly added acid
and of [(DMF)H]+OTf− solutions under hydrogen but without
The open circuit potential (OCP) is a fundamental property
of any electrochemical cell. In 1965, Kolthoff and Thomas used
+
OCP values to determine EBH vs a AgCl/Ag reference
electrode for acetonitrile solutions containing sulfuric acid,
bisulfate, and hydrogen.5c OCP measurements have also been
used to determine EBH in water6 and in several protic ionic
+
liquids.7 As we demonstrate here, EBH may be determined by
+
OCP measurements for a number of different acid-base systems
in acetonitrile. In conjunction with catalytic studies performed
3824
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
any added base. These studies permit evaluation of over-
potentials for some published catalyst systems.3b,10
OCP Measurement. A schematic outlining the measure-
of the electrochemistry of hydrogen at polycrystalline and
single-crystal platinum surfaces was significantly improved
following the implementation of flame annealing as a method
of pretreatment,13 and we have found this practice to be
indispensible. Platinum electrodes were immersed in fresh aqua
regia, rinsed, heated in a hydrogen-air flame, cooled under
hydrogen, and then handled under N2. We used lengths of
platinum wire and found that they may be reconditioned
repeatedly. Conditioning requires corrosive conditions and high
temperatures, so commercially available glass- or plastic-encased
electrodes cannot be used. Deterioration of the electrode per-
formance, as indicated by systematic drift or sudden large changes
in the measured potential or increased signal noise, was diagnosed
by replacement with a fresh electrode. Reversible waves were
observed in potential sweep experiments on unstirred solutions.
However, OCP rather than voltammetric measurements were
chosen to avoid decomposition of MeCN at positive potentials.13
Selection of an appropriate reference couple depends on the
redox properties of the other constituent species. For example,
triethylamine and 4-anisidine are oxidized irreversibly near the
ment is shown in Figure 2. The analyte is an acetonitrile
Fc+/0 potential, so we used permethylferrocene (Fc*, E1/2
=
−0.503 V vs Fc+/0) instead in these cases. The reference
compound should be introduced in a redox state that is stable
+
at the open circuit potential: if redox occurs at EBH , the platinum
electrode will mediate this process until equilibrium is
reestablished, and the solution acid:base ratio will have changed
as a result. The most anodic OCP value measured in these studies
was −0.220 V vs Fc+/0, with 0.40 M [(DMF)H]+OTf−. An
electrode at this potential is sufficiently oxidizing to convert only
0.02% of the ferrocene to ferrocenium. Reference compound
concentrations were generally less than 1 mM for all experiments.
OCP as a Function of [Acid]:[Base]. Equimolar stock
solutions of the acid and base were prepared for each acid
studied. Two sets of OCP data were collected, the first with
aliquots of the acid added to the base, affording solutions with
[acid]:[base] varying from 0.1 to 1 and the second with
aliquots of base solution added to the acid. A representative
data set, collected with the acid 4-bromoanilinium tetrafluor-
Figure 2. Schematic of the four-electrode cell configuration used for
open circuit potential (OCP) measurements. The analyte solution
consists of an acid:base:hydrogen mixture of known composition with
an added internal reference compound such as ferrocene. Potentiostat
and potentiometer are shown as separate devices to illustrate the
principle of the measurement.
solution of acid, base, supporting electrolyte, and a reference
compound (ferrocene, Fc, or permethylferrocene, Fc*),
sparged with hydrogen. A cyclic voltammogram spanning the
reference couple is recorded using the glassy carbon working
electrode, the pseudoreference electrode, and the counter-
electrode. The OCP between the pseudoreference and plat-
inum wire electrodes is then measured. These measurements
afford the OCP vs Fc+/0, thereby canceling the unknown
junction potential5c,11 across the membrane separating the
analyte and reference compartments. This also places the OCP
on the scale employed for catalysis studies and recommended
4a
+
−
oborate (4-BrC6H4NH3 BF4 , pKa = 9.43 in acetonitrile,
0.50 M stock acid and base solutions), is shown in Figure 4.
by IUPAC for electrochemical studies in acetonitrile.12
representative OCP trace appears in Figure 3.
A
Figure 4. OCP values as a function of [base]:[acid], and of [acid]:
[base], in acetonitrile (nBu4N+PF6−, constant total ionic strength of 0.5 M).
For the data shown in filled red circles, 0.10 mL aliquots of a 0.10 M
Figure 3. OCP trace showing a range of 1.0 mV. Analyte: 0.020 M
[(DMF)H]+OTf−, 0.020 M DMF, 1.0 atm H2, in acetonitrile (0.5 M
−
nBu4N+PF6 ).
n
−
stock solution of 4-BrC6H4NH2 in acetonitrile (0.10 M Bu4N+PF6 )
+
were added to 1.00 mL of a 0.10 M solution of 4-BrC6H4NH3 BF4− in
Conditioning of the platinum electrodes is the most
acetonitrile; for the data shown in open blue circles, aliquots of the
significant factor in obtaining reproducibility. Understanding
acid solution were added to the base solution.
3825
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
The OCP values converge as equimolarity is approached
from either the acid-rich or base-rich solutions. These OCP
data, plotted vs log([base]/[acid]), along with data using
4-cyanoanilinium tetrafluoroborate (4-NCC6H4NH3+BF4−, pKa =
7.0, 0.040 M stock solutions),14 are shown in Figure 5. Linear
Table 1. OCP Values with [Acid] = [Base] = 4.0 mM for 1:1
Mixtures of Cationic Acids with Their Conjugate Bases in
Acetonitrile
a
−RT
OCP (V vs
Fc+/0
EH°+ (V vs
Fc+/0
acid
pKa
ln(Ka)
)
)
+
−
2,6-Cl2C6H3NH3 BF4
5.06
6.1
−0.299
−0.361
−0.701
−1.113
−0.317
−0.386
−0.731
−1.143
average
−0.023
−0.025
−0.032
−0.031
−0.028
−0.004
[(DMF)H]+OTf−
+
−
4-MeOC6H4NH3 BF4
11.86
18.82
Et3NH+BF4
−
standard deviation
a
Estimates of E°H+. Distilled from H2SO4, then from CaH2; see ref 16b.
and the structure and acid/base chemistry of the solutes.17 Reported
equilibrium constants for the 1:1 aggregation of an acid with its
conjugate base (homoconjugation, eq 3) in acetonitrile solution
range from KB H = 0−35 M−1 for aromatic and aliphatic ammonium
2
KB2
H
BH+ + B HooooooI B2H+
(3)
Figure 5. OCP as a function of log([base]/[acid]) for acids 4-
+
+
NCC6H4NH3 BF4− (filled red circles) and 4-BrC6H4NH3 BF4− (open
acids in acetonitrile18 and are somewhat larger for pyridines (up to
−
blue circles) in acetonitrile (nBu4N+PF6 , constant total ionic strength
2.7 × 102 M−1).19 Values of KB H can be quite large for the strong
2
of 0.5 M), collected under 1.0 atm hydrogen.
mineral acids (6.7 × 102−4.7 × 103 M−1),20 and the carboxylic
acids, sulfonic acids, and phenols (2.0 M−1−6.3 × 105 M−1).17a,21
Triethylamine and several anilines are reported not to undergo
homoconjugation.15,16,17a,19 Homoconjugation of [(DMF)H]+-
OTf− in acetonitrile, to our knowledge, has not been previously
documented.
fits to these plots have intercepts equal to the OCP values
predicted by eq 2 with Q = 1. Slopes of −2.303RT = −0.0591 V
per decade at 25 °C indicate that the assumption log([base]/
[acid]) = log(Q) holds; i.e. that acid−base aggregation
equilibria are not significant.
Plots of OCP vs log([base]/[acid]) for [(DMF)H]+OTf−
(Figure 6) show a sigmoidal dependence similar to responses
Slopes near −0.0591 V per decade were also obtained for
Et3NH+BF4 at three different stock solution concentrations
−
(Supporting Information Figure S1); however, data collected
using [(DMF)H]+OTf− (Supporting Information Figure S2)
showed a sigmoidal dependence of the OCP value on log([base]/
[acid]), suggesting that in this case, aggregation is significant.
Aggregation phenomena are examined in more detail below.
OCP Values as a Function of Concentration with
[Acid]:[Base] = 1:1. In order to estimate equilibrium
potentials in the limit of infinite dilution, we measured OCP
values for a variety of 1:1 acid−base mixtures under 1.0 atm
hydrogen with concentrations ranging from 0.032 mM to 0.50 M.
The ionic strength was maintained at 0.5 M by diluting with aceto-
n
nitrile (0.5 M Bu4N+PF6−). The acids studied were [(DMF)-
H]+OTf− (pKa = 6.1),15 2,6-dichloroanilinium tetrafluoroborate
(2,6-Cl2C6H3NH3+BF4−, pKa = 5.06),4a 4-anisidinium tetrafluor-
oborate (4-MeOC6H4NH3+BF4−, pKa = 11.86),4a and triethylam-
monium tetrafluoroborate (Et3NH+BF4−, pKa = 18.82).4a The
acetonitrile used for these experiments was purified by passage
through alumina, distillation from H2SO4, and then fractional
distillation from CaH2.16 The influence of solvent impurities on the
OCP response is discussed in detail in the Supporting Information;
data showing these effects are presented in Figure S3. Adherence
to eq 2 was obtained with concentrations ranging from 0.50 M to
4.0 mM; OCP values obtained with [acid] = [base] = 4.0 mM
were taken as estimates for the OCP at infinite dilution. Table 1
presents these OCP values, the pKa values of the acids used, and
the resulting estimates of E°H+ for acetonitrile (nBu4N+PF6−)
according to Scheme 1.
Figure 6. OCP as a function of log([base]/[acid]) for acid
−
[(DMF)H]+OTf− and base DMF in acetonitrile (nBu4N+PF6
,
constant total ionic strength of 0.5 M), collected under 1.0 atm
hydrogen. Molar ratios were adjusted as described in Figure 4. Acid
and base stock solutions were 0.50 M (filled red circles) or 0.10 M
(solid blue circles). Fits to experimental data (dashed lines) use the
homoconjugation model described in the Supporting Information.
observed by Coetzee in potentiometric titrations of nitrogen
bases.18 This dependence is reproduced by a regression model
accounting for homoconjugation (details are supplied in the
Supporting Information). According to this model, the sigmoidal
character should be more pronounced for higher analyte
concentrations, as observed experimentally. The resulting fits to
experimental data are shown in Figure 6. Refinement vs
Interactions between acidic and basic solution species to form
aggregates can influence equilibrium potentials for reaction 1 by
changing the activities of participant species; the significance
of this in the context of overpotential determination has been
noted.3a Aggregation depends sensitively on the both the solvent
3826
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
KBHA
experimental data using the homoconjugation model afforded an
−
BH+ + A HooooooI BHA
(4)
estimate of KB H = 49 M−1.
2
OCP values of 1:1 [(DMF)H]+OTf−:DMF mixtures depend
strongly on the total analyte concentration. Equation 3 alone
does not account for this dependence: the degree of
+
EBH vs log([base]/[acid]) but will not produce the sigmoidal
dependence shown in Figure 6. We therefore built a combined
regression model accounting for both eqs 3 and 4 (shown by the
dashed line in Figure 7). Other possible aggregation reactions
affording this dependence are considered in the Supporting
Information. These models permit interpolation of Q, allowing
OCP values to be estimated using eq 2. An example is provided
below. Our choice of nBu4N+PF6− as electrolyte for these studies
was motivated by its prevalence; in light of the ion pairing
interactions hypothesized for [(DMF)H]+OTf−, future study of
electrolyte effects on OCP values may prove useful.
+
homoconjugation does not affect the ratio αB/αBH , since
formation of the homoconjugate B2H+ sequesters one equiv
each of BH+ and B. Therefore, the OCP value should not vary
with analyte concentration. The dependence observed with
[(DMF)H]+OTf− (Figure 7) indicates a larger value of αB/αBH
+
+
Determination of EBH in Conjunction with Catalysis
Studies. The results above demonstrate that the OCP is the
+
equilibrium potential EBH for a system in which acids and
reducing equivalents interconvert with bases and hydrogen. While
the method is useful for examination of complex acid−base
equilibria, our primary goal is to measure equilibrium potentials
+
EBH for H2 production and oxidation under conditions employed
for catalysis, for the precise determination of overpotentials. Two
examples are presented below, hydrogen evolution with [Ni-
(PPh2NPh2)2](BF4)2 and hydrogen oxidation with [Ni(PCy2NBn2)2]-
(BF4)2.
Figure 8A shows voltammograms obtained with [Ni(PPh2NPh2)2]-
(BF4)2, before and after adding 1:1 [(DMF)H]+OTf−:DMF,
and after adding water, both in quantities affording maximum
turnover frequencies. Under H2 but in the absence of added acid
or water, two reversible one-electron reductions are observed,
assigned to the Ni(II/I) and Ni(I/0) couples (black trace). On
addition of 1:1 [(DMF)H]+OTf−:DMF (0.40 M), a catalytic wave
for the reduction of protons to H2 is observed near the Ni(II/I)
couple (red trace). Adding water (0.11 M) significantly increases
the catalytic current (blue trace). These features have been
discussed in detail elsewhere.3b Of primary importance for this
work is the measurement of overpotentials at Ep/2 for this catalytic
system.
Figure 7. OCP values as a function of log([acid]), equal to
−
log([base]), in acetonitrile (nBu4N+PF6 ; ionic strength of 0.5);
acid:base = 1:1 [(DMF)H]+OTf−:DMF. Data from the two most
dilute solutions (open circles) was not employed in the determination
of aggregation equilibrium constants. Acetonitrile was distilled first
from H2SO4, then from CaH2. The fit to experimental data (dashed
lines) uses the aggregation model based on eqs 3 and 4.
at higher concentrations, implying an aggregation reaction that
involves more BH+ than B.
Aggregation processes that would produce the observed
concentration dependence have been proposed for a variety
of acid−base systems in acetonitrile.4c,17a,18,19 Ion pairing (or
heteroconjugation, see eq 4) is the simplest case. This
equilibrium by itself will influence the slope of the plot of
Potentials for reversible H2 oxidation/production were
established by OCP measurements at Pt electrodes (Figure 8B)
−
Figure 8. (A) Cyclic voltammograms of [Ni(PPh2NPh2)2](BF4)2 in acetonitrile (nBu4N+PF6 ) under H2 (1 atm, black trace), with 1:1
[(DMF)H]+OTf−:DMF (0.40 M, red trace) and with both [(DMF)H]+OTf−:DMF and water (0.40 and 0.11 M respectively, blue trace), collected
with υ = 0.1 V s−1. Ep/2 denotes the half-peak potentials of the catalytic waves. EBH denotes the equilibrium potentials for reaction 1. (B) OCP of a
+
platinum electrode vs Fc+/0 as a function of time, used to determine EBH for the solution compositions denoted by the colors shown in part A.
+
3827
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
−
Figure 9. (A) Cyclic voltammograms of [Ni(PPh2NPh2)2](BF4)2 in acetonitrile (nBu4N+PF6 ) under H2 (1 atm, black trace), with [(DMF)H]+OTf−
(0.4 M, red trace) and with both [(DMF)H]+OTf− and water (0.4 and 0.11 M respectively, blue trace), collected with υ = 0.1 V s−1. Ep/2 denotes the
half-peak potentials of the catalytic waves. EBH denotes the equilibrium potentials for reaction 1. (B) OCP of a platinum electrode vs Fc+/0 as a
+
+
function of time, used to determine EBH for the solution compositions denoted by the colors shown in part A.
of solutions having the same composition as those used for
catalysis (Figure 8A). The OCP values in the presence and
absence of water differ by only 5 mV. These potentials, labeled
Overpotentials at Ep/2 for both sets of experiments,
determined using OCP measurements, eq 2, the method
reported by Evans and co-workers,5g and the method of Artero
and co-workers,3a are compared in Table 2. With the 1:1
acid:base mixture, overpotentials determined using OCP
measurements are 35−40 mV smaller than the estimates
afforded using eq 2 with the assumption that Q = 1, due to the
concentration effects discussed above. Accounting for the
aggregation reactions shown in eqs 3 and 4 affords an estimate
of Q = 5.2 (see the Supporting Information), and the revised
estimates using eq 2 are within 4 mV of the OCP-based values.
Overpotentials determined using OCP measurements are
considerably larger when only [(DMF)H]+OTf− is added to the
solution than when the 1:1 acid:base mixture is added. The value
+
EBH in Figure 8A, are subtracted from the half-peak potentials
of the catalytic waves to afford overpotentials at Ep/2 for
[Ni(PPh2NPh2)2](BF4)2 for these conditions. Turnover frequencies
measured in our current studies reproduce prior results obtained
in the absence of H2 (250 s−1 without added water and 560 s−1
with added water),3b indicating that hydrogen does not influence
the turnover frequency, consistent with the standard-state free
energy change of 9 kcal mol−1 for hydrogen addition to
[Ni(PPh2NPh2)2]2+ determined in previous studies.22
Most studies of electrocatalytic proton reduction to evolve
H2 are carried out by adding acids, but in the absence of added
conjugate base. We therefore characterized catalysis with
[(DMF)H]+OTf−, but without added DMF, the conjugate
base of the buffered pair (Figure 9), using the same catalyst,
acid, and water concentrations as shown above.
As expected, the OCP value is much more sensitive to added
water in this case, changing by 38 mV compared to 5 mV
for the buffered solution of acid. The OCP is stable both with
and without water, possibly due to basic impurities. Minor
impurities, while complicating the characterization of acid−base
equilibria or the estimation of E°+, do not influence over-
potential determinations, provideHd the solvent and other
constituents employed are identical for both catalysis and
OCP measurements. Since no base is initially present, the ratio
of activities of acid and base at the electrode will change with
time in an unstirred solution and may become quite different
from that of the bulk solution. Artero’s method for over-
potential determination addresses this situation, as is discussed
below.3a The cyclic voltammogram obtained without added
water shows a curve crossing that was not evident after water
was added; this behavior was reproducible over several repeated
scans and at different scan rates and catalyst concentrations and
may reflect an induction period associated with the initial
absence of base in the solution; this phenomenon is beyond the
scope of the present work.
+
of EBH estimated according to ref 3a produces overpotentials at
Ep/2 that are smaller than the OCP-determined value by 228
(without water) and 190 mV (with water).
Electrocatalytic oxidation of hydrogen was examined with
[Ni(PCy2NBn2)2](BF4)2 in acetonitrile (nBu4N+PF6 ) under
−
hydrogen with 1:1 and 1:10 Et3NH+BF4 :Et3N mixtures, and
−
with subsequent addition of water (Figure 10). Reduction of
[Ni(PCy2NBn2)2](BF4)2 in acetonitrile under N2 (black trace)
shows a reversible Ni(II/I) redox couple. Under H2, oxidation
(red trace) becomes irreversible and new waves are observed for the
isomers formed upon addition of H2 to [Ni(PCy2NBn2)2](BF4)2.23
Subsequent introduction of Et3N results in a catalytic wave
corresponding to the electrocatalytic oxidation of H2 (orange trace).
The catalytic current is not appreciably affected by the presence of
added Et3NH+BF4− (0.1 equiv with respect to base, purple trace)
and increases when the acid and base are equimolar (green trace).
Water increases the catalytic current slightly (blue trace). The latter
enhancements are attributed to isomer redistribution, as discussed
elsewhere.24 Table 3 presents a comparison of overpotentials using
the different methods discussed herein.
The Nernst equation (eq 2) and the OCP method produce
+
values of EBH that agree within 12 mV, indicating that either
the direct measurement or eq 2 with suitable inputs may be
+
used to obtain reliable values of EBH . The method reported in
3828
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
Table 2. Overpotentials at Ep/2 (Bold) Evaluated by Several Methods, for Hydrogen Evolution Catalyzed by
n
−
[Ni(PPh2NPh2)2](BF4)2 in Acetonitrile (0.1 M Bu4N+PF6 )
1:1 [(DMF)H]+OTf−:DMF (0.40 M each)
[(DMF)H]+OTf− (0.40 M)
method
no added water
[H2O] = 0.11 M
no added water
[H2O] = 0.11 M
Ep/2 (V vs Fc+/0
)
−0.882
−0.865
−0.784
−0.784
+
open circuit potential (OCP; this work)
EBH (V vs Fc+/0
)
)
)
)
)
−0.424
0.458
−0.429
0.436
−0.224
0.560
−0.262
0.522
EH°+ not required
OP (V)
a
+
eq 2
EBH (V vs Fc+/0
−0.389
0.493
−0.389
0.476
b
b
EH°+ = −0.028 V vs Fc+/0
OP (V)
c
+
eq 2, accounting for aggregation
EH°+ = −0.028 V vs Fc+/0
EBH (V vs Fc+/0
−0.428
0.454
−0.428
0.437
OP (V)
d
+
ref 5g
EBH (V vs Fc+/0
−0.506
0.376
−0.506
0.359
−0.506
0.278
−0.506
0.278
EH°+ = −0.145 V vs Fc+/0
OP (V)
+
ref 3a
EBH (V vs Fc+/0
e
e
−0.452
0.332
−0.452
0.332
EH°+ = −0.070 V vs Fc+/0
OP (V)
a
b
c
Assuming that activities and concentrations are equal. Equation 2 is not defined for these conditions. See eqs 3 and 4, the associated text above,
and the Supporting Information for estimation of aggregation equilibrium constants. This value of E°H+ was amended in ref 5g from a value
presented in ref 5f. This method estimates EBH at the electrode for cases when catalyis is first order in [acid], and with no deliberately added base or
d
e
+
hydrogen.
ref 5g affords better agreement with the 1:10 acid:base mixture
than with the 1:1 mixtures, due to cancellation of errors
associated with the value of E°H+ employed and with the
assumption that the bulk solution composition does not
influence the ratio of acid to base at the electrode surface at
Ep/2. Adding water does not appreciably affect the OCP values
for these buffered solutions.
−
Figure 10. (A) Cyclic voltammograms of [Ni(PCy2NBn2)2](BF4)2 in acetonitrile (nBu4N+PF6 ) (black trace), under H2 (1.0 atm, red trace), with
Et3N (33 mM, orange trace), with Et3N and Et3NH+BF4− (31 and 3.1 mM respectively, purple trace; 20 mM for both, green trace) and with with
Et3N and Et3NH+BF4− (20 mM) and water (0.16 M, blue trace), collected with υ = 0.1 V s−1. Ep/2 denotes the half-peak potentials of the catalytic
waves. EBH denotes the equilibrium potentials for reaction 1. (B) OCP of a platinum electrode vs Fc+/0 as a function of time, used to determine EBH
+
+
for the solution compositions denoted by the colors shown in part A.
3829
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
Table 3. Overpotentials at Ep/2 (Bold) by Several Methods, for Hydrogen Oxidation Catalyzed by [Ni(PCy2NBn2)2](BF4)2 in
n
−
Acetonitrile (0.1 M Bu4N+PF6 )
Et3N (31 mM) +
Et3NH+BF4 (3.1 mM)
Et3N (20 mM) + Et3NH+BF4 (20 mM)
−
−
method
no added water
no added water
[H2O] = 0.11 M
Ep/2 (V vs Fc+/0
)
−0.773
−0.772
−0.767
+
open circuit potential (OCP; this work)
EBH (V vs Fc+/0
)
)
)
−1.203
0.430
−1.153
0.381
−1.150
0.383
EH°+ not required
OP (V)
a
+
eq 2
EBH (V vs Fc+/0
−1.200
0.427
−1.141
0.369
−1.141
0.374
EH°+ = −0.028 V vs Fc+/0
OP (V)
b
+
ref 5g
EBH (V vs Fc+/0
−1.258
0.485
−1.258
0.486
−1.258
0.491
EH°+ = −0.145 V vs Fc+/0
OP (V)
a
b
Assuming that activities and concentrations are equal. This value of E°H+, reported in ref 5g, amends a value originally appearing in ref 5f.
OCP Values of Solutions of [(DMF)H]+OTf− Under
Hydrogen with No Added Base and of Et3N Under
Hydrogen with No Added Acid. As mentioned above,
hydrogen production electrocatalysis is typically examined in
the absence of added base or hydrogen. To permit estimation
of overpotentials for cases where [(DMF)H]+OTf− was used as
the acid, we examined the OCP values of unbuffered solutions
of this acid under hydrogen, using acetonitrile purified as
described above. The acid concentration was varied in twenty
increments from 0 to 0.40 M. A plot of OCP vs log([[(DMF)H]+])
afforded a linear fit (Supporting Information, Figure S5A) with
R2 = 0.9983 (eq 5, with [[(DMF)H]+] in moles per liter).
Hydrogen is required for the open circuit measurement, and
employ base concentrations in this range), and then in ten
additional increments from 0.071 to 0.773 M. A plot of OCP
values vs log([Et3N]) afforded a linear fit (Supporting
Information, Figure S5B) with R2 = 0.9989 (eq 6 with
[Et3N] in moles per liter).
OCP (V vs Fc+/0) = −0.0548 log([Et3N]) − 1.407
(6)
The stability of these measurements and the linearity of OCP
vs log([Et3N]) is assumed to derive from the presence of trace
quantities of some proton source in solution. The measured
OCP with [Et3N] = 7.2 mM (the lowest concentration
measured) was −1.291 V vs Fc+/0. Rearranging eq 2 and
inputting EH°+ = −0.028 V vs Fc+/0 and the pKa value of 18.82
for Et3NH+, one obtains an estimate of [Et3NH+] = 0.02 mM.
Equation 6 may be used to estimate OCP values for
previously reported experiments. For example, hydrogen
oxidation catalyzed by [Ni(PCy2NBn2)2](BF4)2 was previously
OCP (V vs Fc+/0) = 0.0720 log([[(DMF)H]+]) − 0.192
(5)
so the use of eq 5 to estimate equilibrium potentials for experi-
ments carried out without base or hydrogen implies the
assumption that hydrogen itself does not influence Ep/2. This
assumption, which holds for [Ni(PPh2NPh2)2](BF4)2, should be
considered for each catalyst system examined.
reported in the presence of 0−8.6 mM Et3N, having Ep/2
=
−0.73 V vs Fc+/0 with [Et3N] = 8.6 mM.10c According to eq 6,
the OCP for these conditions would have been approximately
−1.29 V vs Fc+/0 affording an overpotential at Ep/2 of 0.56 V.
This series of measurements and the catalysis experiments
shown in Figure 9 were executed at different times with
different batches of solvent, substrate, and electrolyte. The
OCP value measured with [acid] = 0.40 M was −0.2195 V vs
Fc+/0 in the absence of catalyst, 5 mV positive of the value
obtained during the catalysis study. With acid but no Brønsted
base added deliberately to the solution, the value of Q in eq 2
will be close to zero at equilibrium, and since d ln(Q) = dQ/Q,
small variations in Q will cause large changes in ln(Q) and
the OCP will be very sensitive to variations in αB due to
adventitious base. This noise in the value of ln(Q) at very
large (or very small) values of Q underscores that the OCP
measurement, like any equilibrium determination, is more reliable
when the participant species are at similar activities. In addition, the
effect of an electrocatalytic process on the solution composition at
the electrode will be larger when Q is very large or very small. This
is examined further in the Discussion section.
DISCUSSION
■
For an electrocatalyst mediating a half-reaction in an electro-
chemical cell, energy efficiency is evaluated by the relationship
between current and overpotential. In homogeneous electro-
catalytic systems, the response of the current to a sweep in
potential often approaches a sigmoidal dependence, as shown in
Figure 1. Turnover frequencies are calculated straightforwardly
from catalytic currents by invoking a steady-state approxima-
tion, so catalytic studies often employ conditions in which the
current (and hence the rate) is pseudo-zero-order in substrate
(Figures 8−10).1,2b,3b,10 The value of EBH can and should
+
be obtained under the same conditions as those used to determine
turnover frequencies, ideally from the same solutions. However,
studies of electrocatalytic reactions are often conducted with the
catalytic substrates, but not the products, added deliberately to the
solution: for example, hydrogen production is often studied by
adding only acid (but no base) to the catalyst solution. While this
may afford steady-state kinetics (absent product inhibition,
induction, catalyst decomposition, or other effects), conditions at
the electrode may differ from those of the bulk solution and may
change with time, even in cases where a constant current is
We also measured OCP values for solutions of Et3N in
−
acetonitrile (nBu4N+PF6 ) under H2 with no deliberately added
proton source, again using the acetonitrile distilled first from
H2SO4 and then from CaH2, in order to estimate overpotentials
for previously reported hydrogen oxidation catalysis studies.
[Et3N] was varied in ten increments from 0 to 0.071 M (the
catalysis experiments exemplified by Figure 10 generally
+
observed. Precise determination of EBH (and hence overpotentials)
at the electrode using either the Nernst equation or the OCP
3830
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
method requires conditions wherein the activities of acid, base, and
hydrogen all adhere to a steady-state approximation.
Table 4. Reported and Present Values of E°H+ for Acetonitrile,
in Order of Precedence
As noted above, measuring overpotentials at the half-peak
potential Ep/2 of the catalytic wave is appropriate for characterizing
systems whose responses approach kinetic control, and this
standard has been adopted by many researchers.3a,5g A commonly
used method, formulated by Evans and co-workers,5g compares
EH°+ (V vs
Fc+/0
)
method
source
ref 5c
−0.034 OCP measurements using sulfuric acid:bisulfate
−0.26
−0.05
cyclic voltammetry of perchloric acid
ref 5f
E
p/2 to the standard state potential EB°H+ for reaction 1, computed
a thermochemical cycle constructed using the
potential of the hydrogen electrode in water
ref 5d
according to Scheme 1.5d This assumes that at Ep/2 in the region
+
near the electrode where heterogeneous redox occurs, αBH = αB
−0.07
−0.14
−0.07
a thermochemical cycle using hydride donor
abilities, pKa values, and 2-electron redox
potentials for a series of Pt and Ni hydrides
refs 5a and 5b
ref 5g
and αH2 = 1 (thus Q = 1; see eq 2), regardless of the initial
conditions. Increasing the analytical concentration of BH+ will
correction to the value presented in ref 5f to
account for incomplete dissociation of perchloric
acid
+
certainly increase αBH at the electrode, but may or may not
+
influence ip or Ep/2. Moreover, if only acid is added, αBH and αB
will become equal at some point during the potential sweep only if
catalysis is very rapid compared to substrate and product diffusion,
as with rapid catalysis under high catalyst loading (for example,
“total catalysis”).1 The potential at which this occurs does not
necessarily coincide with Ep/2. Artero and co-workers have also
addressed the issue of estimating Q at the electrode surface under
substrate diffusion-limited (i.e., nonsteady-state) conditions.3a The
method as described is limited to cases where the reaction is first-
order in acid and relies on E°H+, pKa, and the diffusion coefficients
of the acid, base, and hydrogen and, in cases where aggregation is
possible, the equilibrium constants for the aggregation reactions.
Moreover, estimating the solution composition at the electrode
under nonsteady-state conditions and taking this as the relevant
composition for overpotential determination neglects the work
required to establish the substrate and product concentration
gradients. The complexities and limitations associated with these
methods underscore the advantages of using a steady-state
approximation wherein Q at the electrode is essentially equal to
Q in the bulk solution throughout the course of the voltammetry
experiment.
correction to the value presented in ref 5f to
account both for incomplete dissociation of
perchloric acid and for concentration effects
ref 3a
a
−0.028 OCP measurements using a series of acids over a this work
range of solution compositions
a
Standard deviation is 4 mV across measurements from four different
acid−base systems.
Direct Measurement vs Calculation using the Nernst
Equation. OCP measurements require careful preparation but
no unusual equipment; due to the oxygen sensitivity of the
platinum surface, they are best carried out within a glovebox
having feedthroughs for hydrogen gas and potentiostat leads.
While the OCP measurement is more precise for overpotential
determination, the Nernst equation with suitable inputs will
+
provide more facile access to EBH . Table 3 presents
overpotentials at Ep/2 for hydrogen oxidation catalyzed by
[Ni(PCy2NBn2)2]2+ with acetonitrile as the reaction medium.
+
Values of EBH were obtained both by direct OCP measurement
and using eq 2 with EH°+ = −0.028 V vs Fc+/0 and the assumption
that the analytical concentrations of Et3N and Et3NH+BF4
−
In this paper, we establish that equilibrium potentials for
reaction 1, the interconversion of proton donors and electrons
with proton acceptors and hydrogen, may be measured using
the OCP method for acetonitrile electrolyte solutions. This
constitutes a general method for obtaining equilibrium
potentials as a basis for overpotential determination in catalytic
systems. The data presented here also yield the potential of the
SHE in acetonitrile vs the Fc+/0 couple. Our desire for accuracy
in this value motivated consideration of factors such as solvent
impurities and solute nonideality, effects that “cancel out” in an
overpotential determination. Here, precision may be obtained
by performing open circuit measurements and catalysis
measurements using solutions of the same composition,
prepared with material from the same sources.
Standard Hydrogen Electrode in Acetonitrile. The
potential EH°+ for the interconversion of a solvated proton at unit
activity with hydrogen, also at unit activity, is essential for
determining acidities, hydride donor abilities, and bond dissocia-
tion free energies using thermochemical cycles that include redox
reactions.5d The weak basicity of acetonitrile precludes direct
observation of this reaction; however, the established pKa scale
permits the estimation of EH°+ using Scheme 1 and examining the
OCP values of acid−base mixtures in the limit of infinite dilution.
The value obtained by this method is in good agreement with the
value reported in 1965 by Kolhoff and Thomas,5c also based on
OCP measurements but using sulfuric acid−bisulfate as the
analyte, and with thermochemical cycles reported by Wayner and
Parker and DuBois and co-workers (Table 4).
are equal to their activities. The agreement between these
methods is as expected and should be accurate inasmuch as the
activity−concentration assumption holds; this will be true in general
for acids that do not undergo homoconjugation or other aggrega-
tion reactions, such as Et3N, aniline, and para-toluidine.16a,18
Accounting for aggregation reactions becomes necessary
when their influence on equilibrium potentials is large
compared to the desired precision of the measurement. This
is shown in Table 2, listing overpotentials at Ep/2 for hydrogen
production mediated by [Ni(PPh2NPh2)2]2+ with buffered and
unbuffered [(DMF)H]+OTf− as substrate: Values of EBH
+
obtained by OCP measurements differ by 35−40 mV from values
obtained using eq 2 with the activity−concentration assumption.
+
It is also illustrated in Figure 7, showing values of EBH that span
50 mV for 1:1 mixtures of [(DMF)H]+OTf− and DMF at various
concentrations. Modeling of aggregation equilibria (eqs 3 and 4
above) allows interpolation of values of Q for the conditions used
for catalysis; calculations using eq 2 with these estimates of Q
+
afford values of EBH that agree within 5−10 mV of the direct OCP
measurements. Aggregation equilibrium constants are accessible
using a variety of analytical methods.17−21
Overpotential at Ep/2 as a Function of the Reaction
Medium. The OCP method is valuable for determining
equilibrium potentials, and thus comparing overpotentials in
different electrolyte solutions, including those for which there is
no pKa scale at present, e.g. fluorobenzene25 and in novel media
such as ionic liquids.7 For example, we have previously used the
OCP method to establish overpotential scales in mixtures of
3831
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
water with the ionic liquid dibutylformamidium bis-
(trifluoromethanesulfonyl)amide8 and may now make a direct
comparison between the overpotentials at Ep/2 obtained in
these media with those obtained in acetonitrile. The OCP of
the ionic liquid−water mixture having a mole fraction of 0.72 in
water was determined to be −0.219 V vs Fc+/0. Proton reduction
catalyzed by [Ni(PPh2NPh2)2](BF4)2 has an overpotential at Ep/2 of
0.420 V in this medium. For comparison, an OCP value of −0.224 V
vs Fc+/0 was measured for a 0.4 M solution of [(DMF)H]+OTf−
in acetonitrile without added water, affording an overpotential at
equilibrium ratio of acid and base activities may be estimated. In
other cases, information regarding complex acid−base equilibria
are obtained.
Establishing adherence to the Nernst equation is essential in
validating the OCP measurements. Use of OCP measurements
with new acid−base systems and new media therefore requires
examination of a range of different solution compositions.
Validation does not require a pKa scale or an estimate of the
SHE for the medium in question, only that measured potentials
respond to changes in solution composition according to the
Nernst equation.
E
p/2 of 0.560 V for the same catalyst (Table 2). With added water
(0.1 M), the OCP value was −0.262 V vs Fc+/0 and the over-
potential at Ep/2 was 0.522 V. This comparison underscores the
significance of the reaction medium not only in determining
catalytic rates (in this case faster in the ionic liquid by a factor of
10), but on overpotential as well (lower at Ep/2 by 100 mV). This
comparison also illustrates the generality of basing overpotential
The use of measured or, when appropriate, calculated values of
+
EBH as a basis for overpotential determination removes
assumptions regarding the activities of participant species at the
electrode surface and constitutes a rigorous basis for the
determination of overpotentials for electrocatalytic hydrogen
oxidation and production systems operating with all participants
at steady state. This allows comparison of overpotentials across
widely disparate reaction conditions, opening the way for systematic
exploration of the influence of solvent, substrate, hydrogen pressure,
temperature, and coreactant species on overpotential.
+
scales on values of EBH measured by the OCP method for the
bulk solution as employed for catalysis. Use of the OCP method
with new media does require establishing that reaction 1 alone
governs the potential. Measurement across a range of solution
compositions and reconciling the results with the Nernst equation
accomplishes this goal. A single-point measurement is not
sufficient for this purpose. However, provided the range of
compositions examined encompasses those used for catalysis, the
data acquired, along with the model constructed, allows estimation
EXPERIMENTAL SECTION
■
Materials and Methods. Schlenk techniques or an inert-
atmosphere glovebox were used for all manipulations. Acetonitrile
(Alfa-Aesar, anhydrous, amine-free), dichloromethane (Fisher, not
stabilized), and diethyl ether (VWR, not stabilized) were purified by
sparging with nitrogen and passage through neutral alumina using a
solvent purification system (PureSolv, Innovative Technologies, Inc.).
Acetonitrile was further purified by short-path distillation (1 atm) from
sulfuric acid (99.999% metals basis, Aldrich) and then by fractional
distillation from calcium hydride (Aldrich) using a 50 cm Vigreux
column and a variable reflux ratio distillation head (Th = 80.4−80.5 °C).
The first 10% of distillate was discarded, and the distillation was
terminated at ∼90% completion.16b This acetonitrile was used for the
dilution studies described below. Tetrabutylammonium hexafluorophosphate
+
of EBH for any number of catalysis studies without requiring that
each new study be accompanied by a series of OCP measure-
ments.
A final point to be made regarding overpotential measurements
is that these values contain contributions that are intrinsic to the
catalyst (reaction kinetics and the redox properties of catalytic
+
intermidiates) and those that are not (factors influencing EBH ). In
the examples given above, the catalytic studies were often
conducted under conditions that would provide useful mechanistic
insights into the different catalysts under study. For example, the
use of triethylamine (pKa = 18.8 for Et3NH+ in acetonitrile)4c in
the studies of H2 oxidation using [Ni(PCy2NBn2)2](BF4)2 as the
catalyst is convenient in that deprotonation of the catalytic
intermediates is complete, because the pKa value of Et3NH+ is
much larger than the pKa values of the protonated catalytic
intermediates (approximately 13.4 for this particular catalyst).22
While the use of a strong base simplifies mechanistic studies and
enriches our understanding of these catalysts, it results in larger
extrinsic contributions to the overpotential. In principle, weaker
bases than triethylamine could be used with lower pKa values
(approaching 13.4 for [Ni(PCy2NBn2)2](BF4)2) and lower over-
potentials could be obtained. A precise, flexible, and universal
method for determination of overpotential should allow
practitioners to develop and share insights leading to the
optimization of this important parameter and ultimately to a
deeper understanding of the relationship between rate and
overpotential for molecular electrocatalysts.
−
n
+
−
(nBu4N+PF6 ) was prepared from Bu4N+I− and NH4 PF6 (Aldrich)
and purified by crystallization from saturated acetone solution.26
Ferrocene (Fc, Aldrich) was purified by sublimation. (Permethylferro-
cene Fc*, Aldrich) and permethylcobaltocenium hexafluorophosphate
(Strem) were used as received. Hydrogen (Oxarc UHP; 99.999%)
was purified by passage through a water/oxygen/hydrocarbon trap
(Restek 22464) and an indicating water/oxygen trap (Restek 22474)
and fed through the glovebox wall. Water was dispensed from a
Millipore Milli-Q purifier (18 MΩ cm−1) and sparged with nitrogen.
Dimethylformamide (DMF; Burdick and Jackson) was used as
received. Trifluoromethanesulfonic acid (HOTf; Aldrich, 99%) was
used as received and handled under nitrogen. [(DMF)H]+OTf− was
preprared by literature methods.27 Triethylamine (Et3N, Aldrich) was
fractionally distilled (Th = 88.7−88.8 °C). Tetrafluoroboric acid
etherate (HBF4·Et2O, Aldrich) was used as received and stored at −35 °C in
the glovebox. 4-Bromoaniline (4-BrC6H4NH2, Aldrich) and 4-cyanoaniline
(4-NCC6H4NH2, Aldrich) were sublimed. 2,6-Dichloroaniline (2,6-
Cl2C6H3NH2, Aldrich) and 4-anisidine (4-MeOC6H4NH2, Aldrich)
were sublimed then recrystallized from diethyl ether until colorless.
+
+
Et3NH+BF4−+, 4-NCC6H4NH3 BF4−, 2,6-Cl2C6H3NH3 BF4−, 4-
−
+
−
BrC6H4NH3 BF4 , and 4-MeOC6H4NH3 BF4 precipitated on slow
CONCLUSIONS
addition of HBF4·Et2O to an Et2O solution of ∼5% excess base. 4-
■
+
−
+
−
NCC6H4NH3 BF4 and 4-MeOC6H4NH3 BF4 were further purified
by recrystallization from acetonitrile/ether. [Ni(PPh2NPh2)2](BF4)2 and
[Ni(PCy2NBn2)2](BF4)2 were prepared by literature procedures.10c
Analytical Methods. Stock solutions were prepared as needed
using volumetric glassware and gastight syringes in the glovebox and
were used immediately. Solid and liquid solutes were quantitated by
mass. Electrochemical measurements were conducted using a CH
Instruments 620D potentiostat using the four-electrode cell shown in
Figure 2. The working electrode used for cyclic voltammetry was a
Open circuit potential (OCP) measurements are a powerful
method for the direct determination of equilibrium potentials
for the interconversion of acids and electrons with bases and
hydrogen. The measurements of acetonitrile solutions presented
here afford a revised value for the standard hydrogen electrode
(SHE) in this medium. The method works for a broad range of
different acid−base systems, and the results agree with values of
+
EBH computed using the Nernst equation for systems wherein the
3832
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
1 mm glassy carbon disk encased in polyether−ether−ketone (PEEK;
ALS), cleaned as needed inside the glovebox using a polishing pad
(Buehler MicroCloth) loaded with diamond paste (Buehler MetaDi II
0.25 μm) and lubricated with ethylene glycol (Aldrich), followed by
rinsing with acetonitrile. A fresh portion of the polishing pad was used
for each polishing operation. The counterelectrode used for cyclic
voltammograms was a 3 mm diameter glassy carbon rod (Alfa Aesar).
The pseudoreference electrode used for both cyclic voltammograms
and open circuit potenital (OCP) measurements was a silver wire
(Alfa Aesar; 1 mm dia., 99.9%) cleaned with abrasive paper, anodized
for 5 min in aqueous HCl (Aldrich), washed with water and acetone,
dried, and suspended in a glass tube containing neutral acetonitrile
OCP response. Fouling of platinum surfaces under hydrogen in
acetonitrile at potentials near that of the Fc+/0 couple is well-
documented13 and appears to involve formation of surface-bound
CN− groups. Reconditioning restored the OCP response.
OCP vs Fc+/0 as a Function of [Acid]:[Base]. These measure-
−
ments were carried out for [(DMF)H]+OTf−, Et3NH+BF4 , 4-
+
−
+
−
BrC6H4NH3 BF4 , and 4-NCC6H4NH3 BF4 . The general procedure
is illustrated using the following example: A 0.50 M stock solution of
[(DMF)H]+OTf− in acetonitrile was prepared by dissolving [(DMF)-
H]+OTf− (558.4 mg, 2.502 mmol) in acetonitrile and diluting to
5.0 mL. A stock solution of DMF (0.50 M) and nBu4N+PF6− (0.50 M)
was prepared by diluting DMF (182.3 mg, 2.494 mmol) to 5.0 mL
−
(nBu4N+PF6 , at the same ionic strength as the analyte solution) and
−
using acetonitrile (0.50 M nBu4N+PF6 ). A 1.0 mL portion of the acid
fitted with a porous Vycor disc using fluorinated ethylene propylene
heat-shrink tubing. Storage of the reference electrodes in a ball-and-
socket storage tube containing electrolyte solution and sealed with a
Kalrez O-ring preserved their function essentially indefinitely. The
potential of the reference electrode (Eref) vs Fc+/0 generally changed by
less than 25 mV over a given set of measurements.
solution and 0.10 mL of base solution were added to the cell. The
above measurement sequence was executed. A 0.10 mL aliquot of base
solution was then added to the cell, sparging was maintained for 1 min, and
the measurement sequence was repeated. The solution composition was
incremented in this way until the ratio of acid to base was 1:1. The entire
measurement sequence was then repeated using the same stock solutions,
beginning with 1.0 mL of base solution and 0.10 mL acid solution. A
1.0 mL portion each of the acid and base stock solutions were then diluted
The platinum wire electrodes used for OCP (Alfa Aesar; 99.95%,
1 mm diam, ∼50 mm length) were prepared by rinsing with acetone,
drying in air, immersing in freshly prepared aqua regia for
30 min, then rinsing with flowing deionized water (18 MΩ cm−1) for
5 min. Each wire was then clamped into a stainless steel hemostat and
heated to a uniform yellow-orange glow in a hydrogen-air flame
(Smith Little Torch, no. 6 tip, flame length ∼10 cm). Once the entire
wire was at the maximum temperature attainable (judged by color),
the flame was extinguished and the wire was allowed to cool in the
hydrogen stream until it no longer glowed. At this point, the wire was
placed in a screw-cap storage tube maintained under a positive flow of
nitrogen. After each wire was treated, the tube was closed and taken
into the glovebox. Each wire was useable for at least several
measurements, sometimes affording consistent data for up to a day.
Unused electrodes remained active for several days if stored in the
glovebox. Excessive and increasing signal noise, leading ultimately to
a substantial instability of the OCP, were generally taken to indicate
deterioration of the electrode response, and the electrode was
replaced. Substitution of one electrode with another afforded the
same OCP value within 3 mV. The same set of four platinum wire
electrodes was reconditioned as described above and reused numerous
times with no effect on the OCP measurements. Contamination with
oxygen caused rapid deterioration of the OCP response: measure-
ments conducted after the O2 capacity of the O2/H2O traps had been
exhausted revealed a rapid deterioration of the signal-to-noise ratio.
This was interpreted as passivation of the platinum surface by oxygen.
Measurement of OCP vs Fc+/0. A new, oven-dried 1 dram shell
vial containing a stirbar and fitted with a polyethylene cap with
openings for the electrodes and gas inlet tube (Figure 2) was charged
with an analyte solution of the desired composition. One crystal of
ferrocene (<0.5 mg) was added to the solution. The hydrogen gas stream
was presaturated by bubbling through vigorously stirred acetonitrile and
let into the cell via PEEK tubing (0.063 in o.d. × 0.030 in i.d.). The cell
contents were sparged with hydrogen (5−10 bubbles per second) for
several minutes with stirring. The stirring was briefly paused, and the
hydrogen source tube was withdrawn to the headspace. A cyclic voltammo-
gram spanning the Fc+/0 couple was recorded at the glassy carbon working
electrode, using the reference electrode and the glassy carbon counter
electrode. Stirring and sparging were reestablished, and a 30-s trace of the
OCP between the reference and platinum wire electrodes was collected.
This measurement was repeated until the OCP value from t = 5−30 s
varied by less than 0.5 mV from the mean and showed no systematic drift
in potential; 1−3 traces were typical for each OCP measurement. The
equilibration time ranged from a few seconds to several minutes.
Temperatures within the glovebox were generally between 24 and
27 °C. The elevation of Richland, WA is 117 m (384 ft); pressure above
the solutions was assumed to be 1.0 atm for all experiments.
to a volume of 5.0 mL using acetonitrile (0.50 M Bu4N+PF6−) to afford
n
0.1 M stock solutions, and the measurement sequences were repeated. Data
for these experiments appears in Figures 4−6 and in Supporting
Information Figures S1 and S2.
OCP vs Fc+/0 as a Function of Concentration with [Acid]:
[Base] = 1:1. These measurements were executed using the redistilled
acetonitrile (see the Materials and Methods subsection) with
−
+
−
[(DMF)H]+OTf−, Et3NH+BF4 , 4-MeOC6H4NH3 BF4 , and 2,6-
Cl2C6H3NH3 BF4−. The following example shows the general
+
procedure. An analyte solution was prepared as follows: [(DMF)-
H]+OTf− (111.3 mg, 0.499 mmol) and DMF (36.3 mg, 0.497 mmol)
were combined and diluted to a volume of 1.0 mL, affording a solution
of 0.50 M analytical concentration in both [(DMF)H]+OTf− and
n
−
DMF. An electrolyte solution was then prepared, with Bu4N+PF6
(968.6 mg, 2.50 mmol) diluted to 5.0 mL. A crystal of ferrocene (<0.5
mg) was added to each. The above measurement sequence was
executed with the 0.50 M analyte solution. A 0.20 mL aliquot of this
solution was transferred to a new vial, and 0.80 mL of electrolyte
solution was added, affording a new analyte solution of 0.10 M acid
and base concentrations. The electrodes were rinsed, the cell cap fitted
to the new cell, and the measurement sequence was repeated. This
sequence was carried out over six sequential dilutions, affording
analyte concentrations ranging from 0.50 M to 0.16 mM. Data are
shown in Figure 7. For the purpose of estimating E°H+, the Fc+/0 and
Fc*+/0 redox potentials are assumed to be invariant with solution
composition.
+
Determination of EBH in Conjunction with Catalysis Studies:
Hydrogen Production Catalysis with [Ni(PPh2NPh2)2](BF4)2. Stock
solutions were prepared as follows: A 1.8 mg portion of [Ni-
(PPh2NPh2)2](BF4)2·CH3CN (1.5 mmol) was diluted to 2.0 mL in
−
acetonitrile (0.1 M nBu4N+PF6 ) and one crystal of Fc (<0.5 mg) was
added. An acid stock solution with [(DMF)H]+OTf− (447.1 mg, 2.00
mmol) diluted to 1.0 mL in acetonitrile was prepared, and a 1:1
acid:base stock solution was prepared by adding DMF (73.4 mg, 1.00
mmol) to 0.50 mL of the acid stock solution. Using a standard three-
electrode cell (1 mm diameter glassy carbon working electrode,
counter, and reference electrodes as described above), cyclic
voltammograms were recorded following (1) addition of 0.80 mL of
catalyst solution, (2) sparging for 2 min with acetonitrile-saturated
hydrogen (this cyclic voltammogram and the first were nearly
superimposable), (3) addition of 0.20 mL acid stock solution,
affording a turnover frequency (TOF) of 420 s−1,3b and (4) addition
of 2 μL deoxygenated water (TOF = 530 s−1). The resultant solution
was 0.40 M in acid and 0.1 M in water. Beginning with a 0.80 mL
solution of ferrocene in acetonitrile (0.1 M Bu4N+PF6−) without added
n
Deoxygenated water was added deliberately in some experiments
(up to 1 M) and had no effect on the stability of the OCP. Use of the
platinum wire as a working electrode for voltammetric determination
of the Fc+/0 redox potential resulted in immediate degradation of the
catalyst, the above sequence of additions was repeated in a separate cell
equipped for the four-electrode OCP experiment, with OCP and cyclic
voltammogram measurements as described above performed after each
addition. Data are shown in Figure 9. These experiments were then
3833
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
repeated using the 1:1 acid:base stock solution. Turnover frequencies were
250 s−1 with 1:1 acid:base (0.40 M) and 560 s−1 with added water (0.10
M). OCP values obtained by direct measurement of the solutions
containing the catalyst were within 10 mV of those measured with the
solutions without catalyst but were less stable.
Mayer, Dr. Aaron Appel, and Dr. Simone Raugei for invaluable
discussions.
REFERENCES
■
+
Determination of EBH in Conjunction with Catalysis Studies:
(1) Savea
́
nt, J.-M. Chem. Rev. 2008, 108, 2348−2378.
nt, J.-M. J. Am.
Hydrogen Oxidation Catalysis with [Ni(PCy2NBn2)2](BF4)2. Stock
(2) (a) Costentin, C.; Drouet, S.; Robert, M.; Savea
́
solutions were prepared as follows: [Ni(PCy2NBn2)2](BF4)2 (1.1 mg,
Chem. Soc. 2012, 134, 11235−11242. (b) Nicholson, R. S.; Shain, I.
Anal. Chem. 1964, 36, 706−723.
−
0.92 mmol) was diluted to 1.0 mL in acetonitrile (0.1 M nBu4N+PF6 )
and one crystal of permethylcobaltocenium hexafluorophosphate
(3) (a) Fourmond, V.; Jacques, P.-A.; Fontecave, M.; Artero, V. Inorg.
Chem. 2010, 49, 10338−10347. (b) Kilgore, U. J.; Roberts, J. A. S.;
Pool, D. H.; Appel, A. M.; Stewart, M. P.; DuBois, M. R.; Dougherty,
W. G.; Kassel, W. S.; Bullock, R. M.; DuBois, D. L. J. Am. Chem. Soc.
2011, 133, 5861−5872.
(4) (a) Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.;
Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019−1028.
(b) Kaljurand, I.; Rodima, T.; Leito, I.; Koppel, I. A.; Schwesinger,
(<1 mg; for the CoIII/II couple, E1/2 = −1.904 V vs Fc+/0) was added.
−
An acid stock solution with Et3NH+BF4 (47.6 mg, 0.252 mmol) and
−
nBu4N+PF6 (97.0 mg, 0.250 mmol) diluted to 5.0 mL in acetonitrile
was prepared. A base stock solution of Et3N (25.3 mg, 0.250 mmol)
−
diluted to 5.0 mL with acetonitrile (0.1 M nBu4N+PF6 ) was prepared.
Cyclic voltammograms were recorded following (1) addition of
0.50 mL catalyst solution, (2) sparging 2 min with acetonitrile-saturated
hydrogen, (3) addition of 1.0 mL base solution and 0.10 μL acid
solution (the solution was then 31 mM in base and 3.1 mM in acid;
TOF = 2.6 s−1), (4) addition of 0.9 mL acid solution (TOF = 3.1 s−1), and
(5) addition of 0.5 μL deoxygenated water (TOF = 4.6 s−1). The solution
was then 20 mM in both base and acid and 0.16 M in water. Beginning
with a 0.50 mL solution of permethylferrocene in acetonitrile (0.1 M
nBu4N+PF6−) without added catalyst, the above sequence of additions was
repeated in a separate cell equipped for the four-electrode OCP experiment,
with OCP and cyclic voltammogram measurements as described above
performed after each addition. Data are shown in Figure 10.
R. J. Org. Chem. 2000, 65, 6202−6208. (c) Kutt, A.; Leito, I.;
̈
Kaljurand, I.; Soovali, L.; Vlasov, V. M.; Yagupolskii, L. M.; Koppel, I.
A. J. Org. Chem. 2006, 71, 2829−2838. (d) Leito, I.; Kaljurand, I.;
Koppel, I. A.; Yagupolskii, L. M.; Vlasov, V. M. J. Org. Chem. 1998, 63,
7868−7874. (e) Kutt, A.; Rodima, T.; Saame, J.; Raamat, E.; Maemets,
̈
̈
V.; Kaljurand, I.; Koppel, I. A.; Garlyauskayte, R. Y.; Yagupolskii, Y. L.;
Yagupolskii, L. M.; Bernhardt, E.; Willner, H.; Leito, I. J. Org. Chem.
2010, 76, 391−395.
(5) (a) Curtis, C. J.; Miedaner, A.; Ellis, W. W.; DuBois, D. L. J. Am.
Chem. Soc. 2002, 124, 1918−1925. (b) Ellis, W. W.; Raebiger, J. W.;
Curtis, C. J.; Bruno, J. W.; DuBois, D. L. J. Am. Chem. Soc. 2004, 126,
2738−2743. (c) Kolthoff, I. M.; Thomas, F. G. J. Phys. Chem. 1965, 69,
3049−3058. (d) Wayner, D. D. M.; Parker, V. D. Acc. Chem. Res. 1993,
26, 287−294. (e) Curtis, C. J.; Miedaner, A.; Raebiger, J. W.; DuBois,
D. L. Organometallics 2004, 23, 511−516. (f) Daniele, S.; Ugo, P.;
Mazzocchin, G.-A.; Bontempelli, G. Anal. Chim. Acta 1985, 173, 141−
148. (g) Felton, G. A. N.; Glass, R. S.; Lichtenberger, D. L.; Evans, D.
H. Inorg. Chem. 2006, 45, 9181−9184.
OCP vs Fc+/0 as a Function of [[(DMF)H]+OTf−] with No
n
−
Added Base. An electrolyte solution was prepared with Bu4N+PF6
(779.0 mg, 2.011 mmol) diluted to 10.0 mL using redistilled
acetonitrile. 0.8 mL of electrolyte solution was transferred to a new
vial and sparged with hydrogen and one crystal of ferrocene was
added. An analyte solution was prepared by diluting [(DMF)H]+OTf−
(444.9 mg, 1.993 mmol) to 1.0 mL using redistilled acetonitrile. The
CV-OCP measurement sequence described above was performed after
each of twenty additions of 10.0 μL analyte solution, affording analyte
concentrations ranging from 0.025 to 0.40 M. Data are shown in
Supporting Information Figure S4A.
(6) Kinoshita, K.; Madou, M. J. J. Electrochem. Soc. 1984, 131, 1089−
1094.
OCP vs Fc+/0 as a Function of [Et3N−] with No Added Acid. A
1.0 mL portin of the above electrolyte solution was transferred to a
new vial and sparged with hydrogen and one crystal of ferrocene was
added. The CV-OCP measurement sequence described above was
performed after each of ten additions of 1.0 μL Et3N, and after each of
ten additions of 10.0 μL Et3N, affording analyte concentrations
ranging from 0.0072 to 0.70 M. Data are shown in Supporting
Information Figure S4B.
(7) Bautista-Martinez, J. A.; Tang, L.; Belieres, J. P.; Zeller, R.; Angell,
C. A.; Friesen, C. J. Phys. Chem. C 2009, 113, 12586−12593.
(8) Pool, D. H.; Stewart, M. P.; O’Hagan, M.; Shaw, W. J.; Roberts, J.
A. S.; Bullock, R. M.; DuBois, D. L. Proc. Natl. Acad. Sci. U.S.A. 2012,
109, 15634−15639.
(9) Appleby, A. J. In Comprehensive Treatise of Electrochemistry;
Conway, B. E., Bockris, J. O. M., Yeager, E., Khan, S. U. M., White, R.
E., Eds.; Plenum Press: New York, 1983; Vol. 7, p 173−240.
(10) (a) Yang, J. Y.; Bullock, R. M.; Shaw, W. J.; Twamley, B.; Fraze,
K.; Rakowski DuBois, M.; DuBois, D. L. J. Am. Chem. Soc. 2009, 131,
5935−5945. (b) Appel, A. M.; Pool, D. H.; O’Hagan, M.; Shaw, W. J.;
Yang, J. Y.; Rakowski DuBois, M.; DuBois, D. L.; Bullock, R. M. ACS
Catal. 2011, 1, 777−785. (c) Wilson, A. D.; Newell, R. H.; McNevin,
M. J.; Muckerman, J. T.; Rakowski DuBois, M.; DuBois, D. L. J. Am.
Chem. Soc. 2006, 128, 358−366. (d) Kilgore, U. J.; Stewart, M. P.;
Helm, M. L.; Dougherty, W. G.; Kassel, W. S.; DuBois, M. R.; DuBois,
D. L.; Bullock, R. M. Inorg. Chem. 2011, 50, 10908−10918. (e) Wiese,
S.; Kilgore, U. J.; DuBois, D. L.; Bullock, R. M. ACS Catal. 2012, 2,
720−727.
ASSOCIATED CONTENT
* Supporting Information
■
S
Figures S1-S5 as mentioned in the text and aggregation refine-
ment models. Data and graphs presented in Excel format. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: john.roberts@pnnl.gov.
■
(11) Barry, P. H.; Diamond, J. M. J. Membrane Biol. 1970, 3, 93−122.
(12) (a) Gagné, R. R.; Koval, C. A.; Lisensky, G. C. Inorg. Chem.
Notes
1980, 19, 2854−2855. (b) Gritzner, G.; Kuta, J. Pure Appl. Chem.
̊
The authors declare no competing financial interest.
1984, 56, 461−466.
(13) Climent, V.; Feliu, J. J. Solid State Electrochem. 2011, 15, 1297−
1315.
ACKNOWLEDGMENTS
■
(14) Appel, A. M.; Lee, S.-J.; Franz, J. A.; DuBois, D. L.; Rakowski
DuBois, M.; Twamley, B. Organometallics 2009, 28, 749−754.
(15) Kolthoff, I. M.; Chantooni, M. K., Jr. J. Am. Chem. Soc. 1965, 87,
1004−1012.
(16) (a) Coetzee, J. F.; Cunningham, G. P.; McGuire, D. K.;
Padmanabhan, G. R. Anal. Chem. 1962, 34, 1139−1143. (b) Forcier,
G. A.; Olver, J. W. Anal. Chem. 1965, 37, 1447−1448.
This research was supported as part of the Center for
Molecular Electrocatalysis, an Energy Frontier Research Center
funded by the US Department of Energy, Office of Science,
Office of Basic Energy Sciences. Pacific Northwest National
Laboratory is operated by Battelle for the US Department of
Energy. The authors thank Dr. Daniel DuBois, Prof. James M.
3834
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835

Inorganic Chemistry
Article
(17) (a) Coetzee, J. F. Prog. Phys. Org. Chem. 1967, 4, 45−92.
(b) L.N., B.; Petrov, S. I. Russ. Chem. Rev. 1972, 41, 975−990.
(c) Leito, I.; Rodima, T.; Koppel, I. A.; Schwesinger, R.; Vlasov, V. M.
J. Org. Chem. 1997, 62, 8479−8483. (d) Kaupmees, K.; Kaljurand, I.;
Leito, I. J. Phys. Chem. A 2010, 114, 11788−11793. (e) Augustin-
Nowacka, D.; Makowski, M.; Chmurzynski, L. Anal. Chim. Acta 2000,
418, 233−240.
(18) (a) Coetzee, J. F.; Padmanabhan, G. R. J. Am. Chem. Soc. 1965,
87, 5005−5010. (b) Coetzee, J. F.; Padmanabhan, G. R.; Cunningham,
G. P. Talanta 1964, 11, 93−103.
(19) Augustin-Nowacka, D.; Chmurzynski, L. Anal. Chim. Acta 1999,
̃
381, 215−220.
(20) Kolthoff, I. M.; Chantooni, M. K. J. Chem. Eng. Data 1998, 44,
124−129.
(21) (a) Canel, E.; Tasţ ekïn, M.; Atakol, O.; Kilic,̧ E. Turk. J. Chem.
2003, 27, 77−83. (b) Fujinaga, T.; Sakamoto, I. J. Electroanal. Chem.
1977, 85, 185−201.
(22) Fraze, K.; Wilson, A. D.; Appel, A. M.; Rakowski DuBois, M.;
DuBois, D. L. Organometallics 2007, 26, 3918−3924.
(23) Wiedner, E. S.; Yang, J. Y.; Chen, S.; Raugei, S.; Dougherty, W.
G.; Kassel, W. S.; Helm, M. L.; Bullock, R. M.; Rakowski DuBois, M.;
DuBois, D. L. Organometallics 2011, 31, 144−156.
(24) O’Hagan, M.; Ho, M.-H.; Yang, J. Y.; Appel, A. M.; DuBois, M.
R.; Raugei, S.; Shaw, W. J.; DuBois, D. L.; Bullock, R. M. J. Am. Chem.
Soc. 2012, 134, 19409−19424.
(25) Liu, T.; DuBois, D. L.; Bullock, R. M. Nature Chem. 2013, 5,
228−233.
(26) Fry, A. J. In Laboratory Techniques in Electroanalytical Chemistry,
2 ed.; Kissinger, P. T., Heineman, W. R., Eds.; Marcel Dekker, Inc.:
New York, 1996; p 469−483.
(27) Favier, I.; Dunach, E. Tetrahedron Lett. 2004, 45, 3393−3395.
̃
3835
dx.doi.org/10.1021/ic302461q | Inorg. Chem. 2013, 52, 3823−3835