Electrocatalytic oxidation of formate by [Ni(PR 2NR′2)2(CH3CN)] 2+ complexes

Article abstract of DOI:10.1021/ja204489e

[Ni(P^R₂N^R′₂) ₂(CH₃CN)]²⁺ complexes with R = Ph, R′ = 4-MeOPh or R = Cy, R′ = Ph, and a mixed-ligand [Ni(P^R ₂N^R′₂)(PR′′ ₂NR′

Full text of DOI:10.1021/ja204489e

ARTICLE
pubs.acs.org/JACS
0
2 2
Electrocatalytic Oxidation of Formate by [Ni(PR NR ) (CH3CN)]2+
2
Complexes
Brandon R. Galan,‡ Julia Sch€offel,† John C. Linehan,‡ Candace Seu,† Aaron M. Appel,*,‡ John A. S. Roberts,‡
Monte L. Helm,# Uriah J. Kilgore,‡ Jenny Y. Yang,‡ Daniel L. DuBois,‡ and Clifford P. Kubiak*,†
‡Chemical and Materials Sciences Division, Pacific Northwest National Laboratory, Richland, Washington 99352, United States
†Department of Chemistry & Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla, California 02093,
United States
#Chemistry Department, Fort Lewis College, 1000 Rim Drive, Durango, Colorado 81301, United States
S Supporting Information
b
0
ABSTRACT: [Ni(PR NR ) (CH3CN)]2+ complexes with R = Ph, R0 = 4-Me-
2
2 2
0
R0
2
R00
OPh or R = Cy, R0 = Ph , and a mixed-ligand [Ni(PR N )(P 2NR 2)(CH3-
2
CN)]2+ with R = Cy, R0 = Ph, R00 = Ph, have been synthesized and characterized by
single-crystal X-ray crystallography. These and previously reported complexes
are shown to be electrocatalysts for the oxidation of formate in solution to produce
CO2, protons, and electrons, with rates that are first-order in catalyst and formate
at formate concentrations below ∼0.04 M (34 equiv). At concentrations above
∼0.06 M formate (52 equiv), catalytic rates become nearly independent of formate concentration. For the catalysts studied,
maximum observed turnover frequencies vary from <1.1 to 15.8 sꢀ1 at room temperature, which are the highest rates yet reported
for formate oxidation by homogeneous catalysts. These catalysts are the only base-metal electrocatalysts as well as the only
homogeneous electrocatalysts reported to date for the oxidation of formate. An acetate complex demonstrating an η1-OC(O)CH3
binding mode to nickel has also been synthesized and characterized by single-crystal X-ray crystallography. Based on this structure
and the electrochemical and spectroscopic data, a mechanistic scheme for electrocatalytic formate oxidation is proposed which
involves formate binding followed by a rate-limiting proton and two-electron transfer step accompanied by CO2 liberation. The
pendant amines have been demonstrated to be essential for electrocatalysis, as no activity toward formate oxidation was observed for
the similar [Ni(depe)2]2+ (depe = 1,2-bis(diethylphosphino)ethane) complex.
’ INTRODUCTION
One possible pathway for formate oxidation involves the het-
erolytic cleavage of the CꢀH bond. Recent work in one of our
The temporal variation in the availability of renewable energy
sources such as solar and wind makes storage of this energy in the
form of fuels attractive.1 Water-splitting is one attractive route for
energy storage, but the low volumetric energy density of H2 has
motivated efforts to store energy in the form of reduced carbon
products with much higher volumetric energy densities, such as
methanol or formic acid. Although formic acid has a lower energy
density than methanol, current formic acid fuel cells can run at
much higher concentrations than methanol fuel cells, with the
result that the performance of formic acid fuel cells is competitive
with the performance of methanol fuel cells.2ꢀ4
0
laboratories has shown that [Ni(PR NR ) (CH3CN)]2+ com-
2
2 2
plexes rapidly and reversibly promote the heterolytic cleavage of
H2 into a hydride and proton during H2 oxidation using the
nickel center as a hydride acceptor and the ligand as a proton
acceptor. There is also a considerable body of information on the
factors controlling the hydride acceptor abilities of this class of
complexes.9,10 In fact, metal diphosphine complexes were used to
determine the hydride donor ability of formate in acetonitrile
solutions (44 ( 2 kcal/mol),11 and this transfer of a hydride from
formate to generate CO2 and a metal hydride can be viewed as
Currently formic acid is mainly produced by carbonylation of
methanol followed by hydrolysis.5 However, there is also much
current work on the production of formic acid by direct electro-
chemical reduction of carbon dioxide or indirectly via water
electrolysis to form hydrogen followed by hydrogenation of CO2
to produce formic acid.6ꢀ8 Once viable methods become avail-
able for producing formic acid for energy storage, electrocatalysts
for the oxidation of formic acid will be of interest. At present,
however, the only electrocatalysts reported for this process are
heterogeneous catalysts based upon precious metals such as
platinum, palladium, or rhodium.9ꢀ17
the first step in the oxidation of formate. It was anticipated
0
that [Ni(PR NR ) (CH3CN)]2+ complexes would be electro-
2
2 2
catalysts for formate oxidation, as the hydride acceptor ability
0
of [Ni(PR NR ) (CH3CN)]2+ complexes is in the appropriate
2
2 2
range to accept a hydride from formate and can be systematically
varied, and because the pendant bases can facilitate the oxidation
of hydrides to protons.
Received: May 16, 2011
Published: June 21, 2011
r
2011 American Chemical Society
12767
dx.doi.org/10.1021/ja204489e J. Am. Chem. Soc. 2011, 133, 12767–12779
|

Journal of the American Chemical Society
ARTICLE
Chart 1
In this article we examine the influence of different phos-
due to immiscibility of water with benzonitrile, as well as the
potential variability of electrocatalytic rates with water concen-
tration, as recently reported elsewhere.16 A well-behaved, acet-
onitrile- and benzonitrile-soluble solid was ultimately isolated by
starting from aqueous tetrabutylammonium hydroxide and aqu-
eous formic acid, followed by extraction into ethyl acetate and
then crystallization. The product of this reaction was found to be
a 1:1 mixture of tetrabutylammonium formate and formic acid,
0
phorus and nitrogen substituents of [Ni(PR NR ) (CH3CN)]2+
2
2 2
complexes for catalytic formate oxidation and the role of a
pendant base in this electrocatalytic0 process. The synthesis and
characterization of new [Ni(PR NR ) (CH3CN)]2+ complexes
2
2 2
are described. The formation of nickel hydride species by reaction
with hydrogen as well as with formate is also described, and the
hydride donor abilities (ΔG°Hꢀ) of three new hydride complexes
are reported. Finally, the activities for electrocatalytic oxidation of
formate are discussed, both for the new complexes reported here
as well as for previously reported complexes.
noted as NBu4HCO2 HCO2H. Elemental analysis and 1H NMR
3
spectroscopy are consistent with this assignment, and a single-
crystal X-ray structure was collected and is in the Supporting
Information. Each formate and formic acid pair share one proton,
which appears at 18.7 ppm in the 1H NMR spectrum in CD3CN.
This solid, crystalline material was used as the source of formate/
formic acid for the reactions reported here.
’ RESULTS
Synthesis and Characterization of Ligands and Com-
plexes. The ligands used for this study differ in the substitution
on the phosphorus (R) and the0nitrogen atoms (R0) of the eight-
membered ring of the PR NR ligands, as shown in eq 1 in
Complexes [Ni(PPh2NPhOMe2)2(CH3CN)]2+, [Ni(PCy2NPh2)2-
(CH3CN)]2+, and [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+ react
with NBu4HCO2 HCO2H to form the respective nickel hydride
2
2
3
complexes, [HNi(PPh2NPhOMe2)2]+, [HNi(PCy2NPh2)2]+, and
[HNi(PCy2NPh2)(PPh2NPh2)]+. Formation of the hydride spe-
cies was confirmed by new signals in the 31P{1H} NMR spectra
and the presence of unambiguous hydride signals between ꢀ8
and ꢀ11 ppm in the 1H NMR spectra (see Supporting Informa-
tion for the 31P NMR spectra and a discussion of the 1H NMR
spectra for the hydride resonances). These hydride complexes
are unstable in solution under a nitrogen or argon atmosphere.
After 1 week, an acetonitrile solution of [HNi(PPh2NPhOMe2)2]+
forms a colorless precipitate, and the color of the solution
changes from yellow to red. The red color is attributed to the
Chart 1. The synthesis of the ligands was performed according to
literature procedures.12ꢀ15 The ligands used in this study were
the following: PPh2NPhOMe (1,5-di(4-methoxyphenyl)-3,7-di-
2
phenyl-1,5-diaza-3,7-diphosphacyclooctane), PPh2NPh2 (1,3,5,7-
tetraphenyl-1,5-diaza-3,7-diphosphacyclooctane), and PCy2NPh
2
(1,5-diphenyl-3,7-dicyclohexyl-1,5-diaza-3,7-diphosphacyclooctane).
Using these ligands, the two homoleptic complexes, [Ni(PPh
-
2
NPhOMe2)2(CH3CN)](BF4)2 and [Ni(PCy2NPh2)2(CH3CN)]-
(BF4)2, were prepared by the reaction of [Ni(CH3CN)6]-
(BF4)2 in acetonitrile with 2 equiv of the ligand, as shown in
eq 2 in Chart 1.12 A heteroleptic complex was prepared in two
steps. First, the complex [Ni(PCy2NPh2)(CH3CN)2](BF4)2 was
prepared by the reaction of [Ni(CH3CN)6](BF4)2 and 1 equiv of
the ligand PCy2NPh2 in acetonitrile. After isolation of the inter-
mediate [Ni(PCy2NPh2)(CH3CN)2](BF4)2, the second ligand,
PPh2NPh2, was coordinated in a second step (eq 3 in Chart 1),
formation of the starting complex [NiII(PPh2NPhOMe
)
2 2
(CH3CN)]2+, while the precipitate is the nickel(0) complex,
[Ni0(PPh2NPhOMe2)2] (both confirmed by 31P{1H} NMR). This
nickel(0) complex [Ni0(PPh2NPhOMe2)2] was synthesized and iso-
lated independently by the reaction of [NiII(PPh2NPhOMe2)2-
(CH3CN)]2+ with an excess of formate or by reaction with 2 equiv
of NaB(OMe)3H. The spectral data for [Ni0(PPh2NPhOMe2)2] are
provided in the Experimental Section section, and the crystal structure
yielding the mixed complex [Ni(PCy2NPh2)(PPh2NPh
)
2
(CH3CN)](BF4)2. The spectroscopic data are consistent with
reports of other homo- and heteroleptic nickel complexes with
phosphine ligands and are provided in the Experimental Section.
These four complexes, [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2,
[Ni(PCy2NPh2)2(CH3CN)](BF4)2, [Ni(PCy2NPh2)(CH3CN)2]-
(BF4)2, and [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)](BF4)2, were
also characterized by X-ray diffraction studies, as discussed
below.
is shown below. Complexes [HNi(PCy2NPh2)2]+ and [HNi(PCy
-
2
N
Ph2)(PPh2NPh2)]+ exhibit similar behavior to [HNi(PPh
-
2
NPhOMe2)2]+. This disproportionation reaction is presumably
accompanied by the formation of hydrogen gas as shown in eq 4.
CH3CN
2+
2þ
2½HNiIIðPPh2NPhOMe2Þ ꢁþ
s
1 week
½NiIIðPPh2NPhOMe2Þ ꢁ
In order to explore the reactivities of the Ni(P2N2)2 com-
f
2
2
plexes, an acetonitrile- and benzonitrile-soluble source of formate
or formic acid was required. Aqueous formic acid was avoided
þ ½Ni0ðPPh2NPhOMe2Þ2ꢁ þ H2
ð4Þ
12768
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Figure 1. Thermal ellipsoid plots of the crystal structures of the complexes. Thermal ellipsoids are shown at the 50% probability level. For clarity,
hydrogen atoms and uncoordinated counterions and solvent molecules are omitted, and only the first carbon of the nitrogen and phosphorus
substituents are shown: [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2 CH3CN; [Ni(PCy2NPh2)2(CH3CN)](BF4)2; [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]-
3
(BF4)2 3/8 Et2O, where P(1) and P(2) belong to the PPh2NPh2 ligand and P(3) and P(4) belong to the PCy2NPh2 ligand; [Ni(PPh2NPhOMe2)2] THF; and
3
3
[Ni(PCy2NPh2)(CH3CN)2](BF4)2 CH3CN.
3
X-ray Diffraction Studies. Crystal structures of the four
dicationic complexes [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2,
[Ni(PCy2NPh2)2(CH3CN)](BF4)2, [Ni(PCy2NPh2)(CH3CN)2]-
(BF4)2, and [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)](BF4)2 as well
as the structure of the neutral complex [Ni(PPh2NPhOMe2)2] were
obtained, as shown in Figure 1, and selected bond lengths and
angles are in the Supporting Information. The crystals of the
three nickel(II) bis-diphosphine complexes were all grown from
acetonitrile solutions of the complexes layered with diethyl ether
at ꢀ35 °C. All three structures show a trigonal bipyramidal
coordination geometry around the nickel(II) center. In each of
these complexes, both P2N2 ligands are coordinated via the
phosphorus donor atoms to axial and equatorial positions on the
nickel(II). The third equatorial ligand is an acetonitrile molecule.
The bond lengths and angles, as well as the bite angles, are
comparable to those of previously reported structures,12 but
there are some differences in the boat/chair conformations of the
rings, likely due to packing effects.
coordination geometry around the nickel center. The square plane
is spanned by the two phosphorus atoms of the PCy2NPh2 ligand
and two nitrogen atoms of two coordinated acetonitrile mol-
ecules. Selected bond lengths and angles are in the Supporting
Information.
The nickel η1-acetate complex, [Ni(PPh2NBn2)2(OAc)](BF4)
(Bn = benzyl), was synthesized by adding 1 equiv of tetrabuty-
lammonium acetate (NBu4OAc) to a solution of [Ni(PPh2NBn2)2-
(CH3CN)](BF4)2 (synthesized as previously described)9 in
The structure of the zero-valent complex [Ni(PPh2NPhOMe2)2]
was determined with crystals grown from vapor diffusion of
pentane into a THF solution of the complex at ꢀ35 °C. As ex-
pected for a nickel(0) complex, the structure shows a tetrahedral
coordination geometry. The same coordination geometry was
observed for the recently published [Ni(PPh2NMe2)2] complex,
and in fact, the bond lengths and angles are very similar to those
reported in the literature.17
Figure 2. Perspective thermal ellipsoid plot of the crystal structure of
[Ni(PPh2NBn2)2(OAc)](BF4). Thermal ellipsoids are shown at the 50%
probability level. The Ni1ꢀO1 distance is 2.09 Å, whereas Ni1ꢀO2 is
3.56 Å. For clarity, hydrogen atoms and uncoordinated counterions are
omitted, and only the first carbon of the nitrogen and phosphorus
substituents is shown.
Crystals of [Ni(PCy2NPh2)(CH3CN)2](BF4)2 suitable for
X-ray diffraction were grown from an acetonitrile solution of
the compound layered with diethyl ether at room temperature.
The structure can be described as a distorted square planar
12769
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Electrochemical Studies. Representative cyclic voltammo-
grams for complexes [Ni(PPh2NPhOMe2)2(CH3CN)]2+,
[Ni(PCy2NPh2)2(CH3CN)]2+, and [Ni(PCy2NPh2)(PPh2NPh2)-
(CH3CN)]2+ in acetonitrile solutions are shown in Figure 3
by traces a, b, and c, respectively. They consist of two reversible
one-electron waves assigned to the Ni(II/I) and Ni(I/0) couples.
This behavior is consistent with previous studies of [Ni(diphos-
phine)2]2+ complexes, and the potentials for these couples are
listed in Table 1, together with related values from the literature.
It can be seen from Figure 3 and Table 1 that the potentials of the
Ni(II/I) couple are particularly sensitive to the nature of the
substituents on phosphorus, shifting to more positive potentials as
the number of cyclohexyl substituents increases.12,22ꢀ24 This posi-
tive shift is somewhat counter-intuitive, in that replacing phenyl
substituents on phosphorus with more electron-donating cyclo-
hexyl groups would be expected to shift the potentials more
negative on the basis of simple inductive effects. However, the
increased steric bulk of the cyclohexyl substituents produces larger
distortionstoward tetrahedralgeometry thatleadto a lower energy
for the lowest unoccupied molecular orbital (LUMO), as dis-
cussed elsewhere.22ꢀ24 The lowerenergy of the LUMO results ina
more positive potential for the Ni(II/I) couple.
Finally, it is of interest to compare the cyclic voltammogram of
the hydride complex [HNi(PPh2NPhOMe2)2]+ (Figure 4) to that
of [Ni(PPh2NPhOMe2)2(CH3CN)]2+ (Figure 3a). Scanning ano-
dically, [HNi(PPh2NPhOMe2)2]+ shows a well-defined irreversible
2eꢀ oxidation at ꢀ0.47 V. The return cathodic scan shows
reduction waves that correspond to the Ni(II/I) and Ni(I/0)
couples of [Ni(PPh2NPhOMe2)2(CH3CN)]2+. This result sug-
gests that the oxidation of [HNi(PPh2NPhOMe2)2]+ is followed by
rapid proton loss and a second electron transfer to generate
[Ni(PPh2NPhOMe2)2(CH3CN)]2+, which is reduced on the re-
turn scan. At variable scan rates of 50 mV/s to 20 V/s, the wave at
ꢀ0.47 V is irreversible, suggesting that proton transfer to the
pendant amine following this oxidation is fast (k > 50 sꢀ1).
Similar behavior is observed for [HNi(depe)2](BF4) (depe =
1,2-bis(diethylphosphino)ethane), but in that case the irrever-
sible oxidation occurrs at 0.0 V (the oxidation is irreversible at all
observed scan rates, up to 100 V/s). This result suggests that the
presence of a pendant base in [HNi(PPh2NPhOMe2)2]+ results in a
0.5 V negative shift in potential. This potential shift is attributed
to the pendant base assisting proton transfer upon oxidation
of [HNi(PPh2NPhOMe2)2]+, as previously observed for other
Figure 3. Cyclic voltammograms of (a) [Ni(PPh2NPhOMe2)2-
(CH3CN)]2+, (b) [Ni(PCy2NPh2)2(CH3CN)]2+, and (c) [Ni(PCy
-
2
NPh2)(PPh2NPh2)(CH3CN)]2+ Conditions: 0.8 mM solutions of the
complexes in acetonitrile, 0.2 M NBu4OTf as supporting electrolyte,
glassy carbon working electrode, scan rate of 50 mV/s. The cyclic
voltammogram for [Ni(PCy2NPh2)2(CH3CN)]2+ was recorded in a 20%
benzonitrile/acetonitrile solution. Potentials are referenced with respect
to the ferrocenium/ferrocene couple.
benzonitrile. The complex crystallizes as a distorted trigonal
bipyramid with the acetate bound through a single oxygen in the
equatorial plane, as shown in Figure 2. Bond distances and angles
are shown in the Supporting Information. The trigonal plane is
distorted, with one oxygenꢀnickelꢀphosphine angle being
122.05° and the other angle being 99.58°. There are no known
structures of analogous nickel(II) tetraphosphine acetate com-
plexes with which to compare directly. However, the NiꢀO bond
length of 2.092 Å is slightly longer (2.055 and 2.050 Å) than
those of other nickel(II) acetate complexes in the literature,18ꢀ21
with the NiꢀOꢀC bond angle being similar to those of other
nickel η1-acetate complexes.
Figure 4. Cyclic voltammograms of [HNi(PPh2NPhOMe2)2]+. Condi-
tions: 1.6 mM complex in acetonitrile, 0.2 M NBu4OTf as supporting
electrolyte, glassy carbon working electrode, scan rate of 100 mV/s, with
ferrocene as the internal reference at 0.0 V.
12770
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Table 1. Thermodynamic and Electrochemical Data in Acetonitrile and Benzonitrile
+
NiL2(CH3CN)2+ ligands
ΔG°Hꢀ (kcal/mol)
E1/2(NiII/I) (V vs Cp2Fe+/0
)
E(NiI/0) (V vs Cp2Fe+/0
)
pKa of HNiL2
PCy2NPh
PCy2NPh2, PPh2NPh
PPh2NPhOMe
63.7
60.5
58.6
59.0
61.4f
57.1
56.4f
60.7
68.4
ꢀ0.62 (ꢀ0.60) a
ꢀ0.76 (ꢀ0.76) a
ꢀ0.87 (ꢀ0.87) a
ꢀ0.84
ꢀ1.09 (ꢀ1.10) a
ꢀ1.05 (ꢀ1.07) a
ꢀ1.07 (ꢀ1.08) a
ꢀ1.02
17.3
16.5
17.4
16.3
13.8b
19.4
18.5b
21.2
18.0
2
2
2
PPh2NPh
c
2
PPh2NPhCF3
ꢀ0.74
ꢀ0.89
d
2
PPh2NBn
ꢀ0.94
ꢀ1.19
2
PPh2NMe
PCy2NBn
PPh2NBn2, dpppg
dppph
depeh
ꢀ0.98
ꢀ1.14
e
2
c
ꢀ0.80
ꢀ1.28
2
(ꢀ0.52)a
(ꢀ1.04)a
ꢀ0.12
ꢀ0.91
56.0
62.1
ꢀ1.16 (ꢀ1.13) a
ꢀ0.89 (ꢀ0.86) a
ꢀ1.29 (ꢀ1.30) a
ꢀ1.33 (ꢀ1.32) a
23.8
23.9
dmpph
a Values in parentheses are in benzonitrile. b Estimated from Ni(I/0) couple using pKa = ꢀ18.6 E1/2 = ꢀ2.65; see Figure 6. c Fraze et al.9 d Kilgore et al.16
e Yang et al.28 f Estimated from Ni(II/I) couple using ΔG°Hꢀ = 20.8 E1/2 + 76.8; see Figure 5. g Yang et al.29 h dppp = 1,3-bis(diphenylphosphino)-
propane, depe = 1,2-bis(diethylphosphino)ethane, dmpp = 1,2-bis(dimethylphosphino)propane. Berning et al.30
complexes containing diphosphine ligands with pendant amine
bases.25 It should be noted that the potentials for the nickel(II/I)
and nickel(I/0) couples for Ni(depe)22+ are actually an average
of 0.25 V more negative than for [Ni(PPh2NPhOMe2)2-
(CH3CN)]2+, suggesting that the presence of the pendant amine
may result in an even larger shift than suggested by simply
the difference between the corresponding hydride oxidation
potentials.
complex [HNi(PCy2NPh2)(PPh2NPh2)(CH3CN)]+, 60.5 kcal/
mol, is only slightly poorer than the hydride donor ability of
the homoleptic complex [HNiPPh2NPh2)2]+, 59.0 kcal/mol. The
0
pKa values for the [HNi(PR NR ) ]+ complexes were deter-
2
2 2
mined through thermochemical cycles using the hydride donor
abilities and the electrochemical potentials for the Ni(II/I) and
Ni(I/0) couples in acetonitrile, as previously reported for similar
complexes.31 The hydride donor abilities of the nickel complexes
show the expected linear trend with the Ni(II/I) couple, and the
pKa values for the metal hydrides also exhibit a linear trend with
the Ni(I/0) couple, as shown in Figures 5 and 6. These trends are
very similar to the trends previously observed for nickel bis-
diphosphine complexes with no pendant bases.30 The hetero-
leptic complex is not shown in Figures 5 and 6, but the data for
this complex are consistent with the trends observed for the
homoleptic complexes.
Thermodynamic Properties of Complexes [HNi(PPh
NPhOMe2)2]+, [HNi(PCy2NPh2)2]+, and [HNi(PPh2NPh2)(PCy
-
2
-
2
NPh2)]+. The hydride donor ability (ΔG°Hꢀ) in acetonitrile for
each of the three complexes was calculated from the observed
equilibrium concentrations of the hydride with the starting com-
plex using the thermodynamic cycle shown in Scheme 1.26,27 The
0
equilibrium between the hydride complex [HNi(PR NR ) ]+ and
2
2 2
0
1
[Ni(PR NR 2)2(CH3CN]2+ was measured by 31P{1H} and H
2
NMR spectroscopy using an appropriate base under 1.0 atm of
hydrogen (eq 5). The sum of reactions 5ꢀ7 is reaction 8 (the
heterolytic cleavage of the NiꢀH bond to form Hꢀ and the
corresponding Ni(II) complex), and the free energy associated
with this reaction is ΔG°Hꢀ, the hydride donor ability. The reverse
Reactivity toward H2 and CO2 using High Pressure NMR Spec-
0
troscopy. The three new bis-PR NR complexes [Ni(PPh
-
2
NPhOMe2)2(CH3CN)]2+, [Ni(PCy2NPh22)2(CH3CN)]2+, and
[Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+ did not demonstrate
any catalytic activity for CO2 reduction under high pressures of
CO2 and H2. The complexes were dissolved in acetonitrile-d3,
and the solutions were placed in a high-pressure PEEK NMR
tube.32,33 No reaction could be observed after applying 21 atm of
a 1:1 mixture of CO2/H2. Even the independently synthesized
hydride complex [HNi(PPh2NPhOMe2)2]+ did not react observa-
2
reaction corresponds to the hydride acceptor ability of
0
[Ni(PR2NR2)2(CH3CN)]2+. Larger values of ΔG°Hꢀ corre-
spond to better hydride acceptors and lower values to better
hydride donors. This information is summarized in Table 1,
together with data for selected additional complexes from the
literature.
bly with 21 atm of carbon dioxide. For complex [Ni(PCy
-
2
NPh2)2(CH3CN)]2+, the formation of a hydride species could
be observed without the addition of an external base, but no
evidence for formation of formate or formic acid could be
observed via NMR spectroscopy. These results are consistent
For the new complexes considered here, [HNi(PCy2NPh2)2]+ is
the poorest hydride donor, with ΔG°Hꢀ = 63.7 kcal/mol, and
[HNi(PPh2NPhOMe2)2]+ is the best hydride donor, with ΔG°Hꢀ
=
58.6 kcal/mol. The hydride donor ability for the heteroleptic
0
Scheme 1. Determination of Hydride Donor Abilities for [HNi(PR NR 2)2]+ Complexes
2
12771
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Figure 6. Correlation of the potential of the Ni(I/0) couple for
Figure 5. Correlation of the potential of the Ni(II/I) couple for
0
0
[Ni(PR NR 2)2(CH3CN)]2+ complexes with the pKa for the correspond-
[Ni(PR2NR 2)2(CH3CN)]2+ complexes with the hydride donor ability
for the corresponding hydrides. The designations for each point (R, R0)
indicate phosphorus and nitrogen substituents, respectively.
2
ing hydrides. The designations for each point (R, R0) indicate phos-
phorus and nitrogen substituents, respectively.
with thermodynamic data discussed in more detail below, where-
x
1=2
icat ¼ ncatFA½catalystꢁðDk½HCO2ꢀꢁ Þ
ð9Þ
in the transfer of a hydride to CO2 to generate formate (ΔGHꢀ
=
44 ( 2 kcal/mol)11 from a nickel hydride (ΔGHꢀ = 56ꢀ64 kcal/
!
1=2
mol in this study) is thermodynamically unfavorable.
F3
ip ¼ 0:4463
np3=2AD1=2½catalystꢁυ1=2
ð10Þ
Electrocatalytic Oxidation of Formate. The complexes
[Ni(PPh2NPhOMe2)2(CH3CN)]2+, [Ni(PCy2NPh2)2(CH3CN)]2+,
and [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+ were also exam-
ined for their activity for electrocatalytic formate oxidation. The
left-hand side of Figure 7 shows cyclic voltammograms of
[Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+ as a function of formate
RT
sffiffiffiffiffiffiffiffiffiffiffi
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
icat
ip
ncat
RT
x
¼
k½HCO2ꢀꢁ
ð11Þ
ð12Þ
3
0:4463 Fυnp
concentration (added as NBu4HCO2 HCO2H). The reversible
!
3
2
ꢀ
ꢁ
2
3
Fυnp
0:4463
icat
ip
wave at ꢀ1.33 V is due to the cobaltocenium/cobaltocene
couple, used as an internal standard. This compound was used
instead of ferrocene because the ferrocenium/ferrocene couple
overlaps the catalytic waves observed for formate oxidation. The
reversible wave at ∼ꢀ1.1 V is assigned to the Ni(I/0) couple as
discussed above. The reversible nature of this wave as the
concentration of formate increases implies that the Ni(I) and
Ni(0) species do not react directly with formate. The wave
at ∼ꢀ0.8 V is assigned to the Ni(II/I) couple. The latter wave
has an increase in current with increasing formate concentration,
shifts to more negative potentials, and assumes a shape more
closely resembling a plateau rather than a diffusional peak shape.
The negative shift in potential is consistent with a fast following
reaction, likely the binding of formate following the oxidation of
Ni(I) to form Ni(II). The increase in current upon oxidation of
the complex to the Ni(II) oxidation state and the plateau-shaped
wave are consistent with the electrocatalytic oxidation of formate.
The catalytic current enhancement (icat/ip) observed in the
presence of formate can be converted to a catalytic rate using
eqs 9ꢀ12 (icat and ip are the current in the presence and absence
of substrate, respectively). Dividing eq 9 by eq 10 gives eq 11,
wherein the terms for the area of the electrode (A), the diffusion
coefficient (D), and the concentration of the catalyst are all
eliminated. Equation 11 expresses icat/ip in terms of the number
of electrons in the catalytic process (ncat), the universal gas
constant (R), the temperature in Kelvin (T), Faraday’s constant
(F), the scan rate in V/s (υ), the number of electrons in the wave
in the absence substrate (np), and the observed rate constant for
the catalytic reaction (k).34ꢀ36 A catalytic turnover frequency
(TOF) can be calculated by converting eq 11 to eq 12 through a
simple rearrangement. The results of the kinetic studies are
summarized in Table 2, where the reported TOF is the highest
TOF observed, as calculated using eq 12.
x
TOF ¼ k½HCO2ꢀꢁ ¼
RT
ncat
A plot of the TOF of the catalyst versus the concentration of
formate (as NBu4HCO2 HCO2H) is shown on the right-hand
3
side of Figure 7. It can be seen that the TOF initially increased
linearly as a function of formate concentration, became indepen-
dent of formate concentration between ∼0.04 and 0.06 M, and
slowly decreased at higher formate concentrations. The initially
linear region of this plot implies a first-order dependence of the
TOF on formate concentration. The formate-independent re-
gion indicates a change in the rate-determining step from one
that is first-order in formate to one that is independent of formate.
The slow decrease above ∼0.08 M formate is likely due to catalyst
0
decomposition. Loss of PR NR ligand from the nickel complex
2
2
has been observed at high formate concentrations after hours in
solution or in the presence of excess acetate, as indicated by the
0
appearance of free PR NR2 ligand in the 31P NMR spectrum.
2
The left-hand side of Figure 8 shows a series of cyclic voltam-
mograms recorded in benzonitrile solutions of [Ni(PCy2NPh2)-
(PPh2NPh2)(CH3CN)]2+ as a function of catalyst concentration
in the presence of excess formate (0.15 M NBu4HCO2
3
HCO2H). It can be seen that the catalytic current increases as
the concentration of the catalyst increases. On the right-hand
side of Figure 8 is a plot of the catalytic current (icat) as a function
of the catalyst concentration obtained from the data on the left-
hand side of the figure. The linear dependence of this plot on
catalyst concentration implies that the catalytic reaction is first-
order in catalyst. The data taken together implies that at low
formate concentrations, the rate-determining step includes the
reaction of formate with the catalyst, while at high formate
concentrations the rate of reaction is still first-order in catalyst
but appears to become independent of formate concentration.
However, decay of the TOF at high formate concentration, likely
12772
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Figure 7. Left: CVs of the titration of [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+, with NBu4HCO2 HCO2H. Right: Corresponding turnover frequency
3
versus [NBu4HCO2 HCO2H]. Conditions: 1.2 mM catalyst solution in benzonitrile, 0.2 M NBu4OTf as supporting electrolyte, glassy carbon working
3
electrode, scan rate 50 mV/s. Potentials are referenced with respect to the ferrocenium/ferrocene couple (0.00 V) using the cobaltocenium/cobaltocene
couple as a secondary reference (ꢀ1.33 V).
Table 2. Electrocatalytic TOFs and Thermochemical Data
a
+ b
+ b
NiL2(CH3CN)2+ ligands
TOF (sꢀ1
)
ΔG°Hꢀ (kcal/mol)b
pKa of HNiL2
pKa of free R0NH3
c
PPh2NMe
15.8
12.5
9.6
56.4 g
57.1
60.7
58.6
60.5
59.0
61.4 g
63.7
56.0
18.5 h
19.4
21.2
17.4
16.5
16.3
13.8 h
17.3
23.8
18.37 i
16.91 i
16.91 i
11.86 j
10.62 j
10.62 j
8.03 j
2
d
PPh2NBn
2
d
PCy2NBn
2
PPh2NPhOMe
8.7
2
PCy2NPh2, PPh2NPh
7.9
2
d
PPh2NPh
7.4
2
PPh2NPhCF3
3.4
e
2
PCy2NPh
<1.1
<0.4
10.62 j
2
depef
ꢀ
a Data collected in benzonitrile with ∼1 mM complex and 0.2 M NBu4OTf, where value shown is the maximum observed TOF. b Thermochemical
values in acetonitrile. c Yang et al.28 d Fraze et al.9 e Kilgore et al.16 f depe = 1,2-bis(diethylphosphino)ethane. Berning et al.30 g Estimated from Ni(II/I)
couple using ΔG°Hꢀ = 20.8 E1/2 + 76.8; see Figure 5. h Estimated from Ni(I/0) couple using pKa = ꢀ18.6 E1/2 = ꢀ2.65; see Figure 6. Izutsu.37
i
j Kaljurand et al.38
Figure 8. Left: CVs of the titration of NBu4HCO2 HCO2H with [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+. Right: icat vs the concentration of
3
[Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+. Conditions: 0.15 M NBu4HCO2 HCO2H solution in benzonitrile, 0.2 M NBu4OTf as supporting electrolyte,
3
glassy carbon electrode, scan rate 50 mV/s. Potentials are referenced with respect to the ferrocenium/ferrocene couple using cobaltocenium as a
secondary internal reference.
a result of catalyst decomposition, complicates the interpretation
of the formate concentration-independent region.
the formic acid was not expected to have a substantial effect on
catalysis. In order to evaluate the effect of the formic acid, the
Due to NBu4HCO2 HCO2H being a simple, soluble, crystal-
electrocatalytic formate oxidation was repeated for [Ni(PPh
-
2
3
NPhOMe2)2(CH3CN)]2+ with a solution containing NBu4H-
line solid, this was used as the formate source throughout the
present studies. This soluble salt does contain a co-crystallized
formic acid that is homoconjugated with the formate; however,
CO2 HCO2H with an equimolar amount of DBU to deproto-
3
nate the formic acid and thereby generate formate and
12773
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
[HNi(diphosphine)2]+ complexes are known to be in the range
of 57ꢀ67 kcal/mol,30 and the hydride donor ability of formate
anion in acetonitrile is 44 ( 2 kcal/mol.11 These data indicated
that the formate anion should readily transfer a hydride to nickel
to form CO2 and a nickel hydride, and that these hydrides are
the sam0e intermediates involved in the oxidation of H2 by [Ni-
(PR NR ) (CH3CN)]2+ complexes. As a result, it was antici-
2
2 2
pated that these complexes would be catalysts for formate
oxidation.
HCO2H f CO2 þ 2Hþ þ 2eꢀ
ð13Þ
0
A series of [Ni(PR NR ) (CH3CN)]2+ complexes were
2
2 2
synthesized, characterized, and tested for electrocatalytic oxida-
tion of formate. Systematic variation of the substituents at
phosphorus and nitrogen allowed control over the hydride donor
abilities and pKa values for the corresponding nickel hydrides,
which correlate with the reduction potentials for the Ni(II/I) and
Ni(I/0) couples, respectively (as shown in Figures 5 and 6).
These correlations provide estimates of the hydride donor
abilities and pKa values of the nickel hydrides from simple
electrochemical potentials, which are more easily measured.
The observed correlation is similar to the one previously
reported for nickel bis-diphosphine complexes lacking a pendant
base.30
Figure 9. Cyclic voltammograms for the titration of [Ni(depe)2]2+ with
NBu4HCO2 HCO2H. Conditions: 0.98 mM [Ni(depe)2]2+ in benzo-
3
nitrile, 0.2 M NBu4OTf as supporting electrolyte, glassy carbon working
electrode, scan rate 50 mV/s. Potentials are referenced with respect to
the ferrocenium/ferrocene couple using permethylferrocene as a sec-
ondary internal standard.
protonated DBU. The result of using 1 equiv of DBU per
NBu4HCO2 HCO2H was a very similar overall catalytic TOF of
3
10.6 sꢀ1 with DBU vs 8.5 sꢀ1 using NBu4HCO2 HCO2H
3
without DBU (see Experimental Section for additional details).
To determine that CO2 was indeed the product of the catalytic
reaction, a controlled potential electrolysis was carried out
using [Ni(PPh2NPhOMe2)2(CH3CN)]2+ as the catalyst (0.76 mM)
Our initial expectation was that the catalytic rates of formate
oxidation would parallel the hydride acceptor ability of the nickel
catalyst, where in this series [Ni(PCy2NPh2)2(CH3CN)]2+ is the
best hydride acceptor and [Ni(PPh2NPhOMe2)2(CH3CN)]2+ is
the worst. To test this hypothesis, we carried out a series of
kinetic studies that showed that the TOF for each catalyst was
dependent on formate concentrations below ∼0.04 M (34 equiv
in an acetonitrile solution containing 0.086 M NBu4HCO2
3
HCO2H at a potential ∼100 mV positive of the peak of the
catalytic current. After passing 16.3 C of charge (∼8 turnovers), the
headspace was analyzed by gas chromatography, and CO2 but no
H2 was observed. Quantification of the CO2 indicated a current
efficiency of 93 ( 5% for this reaction. The protons were pre-
sumably trapped by formate to produce formic acid.
NBu4HCO2 HCO2H) and that above this concentration the
3
TOF became formate concentration independent. In addition,
the onset of the catalytic current was observed near the Ni(II/I)
couple, consistent with no reaction of nickel(0) or nickel(I) with
formate, followed by immediate reaction after oxidation to
nickel(II). This interpretation is supported by NMR spectro-
scopic studies that demonstrate the nickel(II) complexes
[Ni(PPh2NPhOMe2)2(CH3CN)]2+, [Ni(PCy2NPh2)2(CH3CN)]2+,
[Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)]2+, and [Ni(PCy2NBn2)2]2+
all react with formate to form the corresponding hydride com-
plexes. However, the concentration range over which the catalytic
rates are dependent on formate concentration is small. At con-
For comparison, the complex [Ni(depe)2]2+ was also studied
to determine if a similar complex without a pendant amine would
be a catalyst for formate oxidation. As seen in Figure 9, addition
of increasing amounts of NBu4HCO2 HCO2H to acetonitrile
3
solutions of [Ni(depe)2]2+ results in the formation of [HNi-
(depe)2]+, which shows a 2eꢀ oxidation at ∼0 V vs Cp2Fe+/0
(Cp*2Fe+/0 used as the internal reference). The lack of current
enhancement beyond the 2eꢀ process at this [HNi(depe)2]+
oxidation indicates no measurable catalytic activity (<0.4 sꢀ1).
This result indicates that the pendant amines play an important
centrations above ∼0.06 M formate (52 equiv of NBu4HCO2
0
3
2+
role in the observed catalytic activity of Ni(PR2NR
complexes.
)
HCO2H), where catalysis is most likely to be performed under
practical conditions, the catalytic rate appears to become inde-
pendent of formate concentration.
2 2
For the pendant base-containing complexes tested, [Ni-
(PCy2NPh2)2(CH3CN)]2+ is the best hydride acceptor and yet
the least active catalyst. In fact, the catalytic rates correlate poorly
with the hydride acceptor abilities of these complexes, as shown
in Table 2. The failure of the catalytic rate to correlate with the
hydride acceptor ability is inconsistent with a rate-determining
step involving direct hydride transfer from formate to the nickel
center. The observation that [Ni(depe)2]2+ shows no observable
catalytic activity in the presence of excess formate (as shown in
Figure 9) indicates that the presence of the pendant amine is
essential for catalysis.
’ DISCUSSION
The development of inexpensive and energy-efficient electro-
catalysts for formate oxidation is essential for the utilization of
formic acid fuel cells (the anodic half-reaction involved is shown
in eq 13). In this reaction, formic acid is oxidized by two
electrons, with the release of CO2 and two protons. As with
other multi-electron and multi-proton processes, the manage-
ment of proton movement for catalysts of this type is of
paramount importance. In previous work it has been demon-
strated that pendant bases incorporated in the second coordina-
tion sphere of first-row transition metal complexes can serve as
proton relays for catalytic oxidation and production of H2 and for
O2 reduction.10,17 In addition, the hydride donor abilities of
To probe the binding of formate with the nickel center,
[Ni(PPh2NBn2)2(CH3CN)]2+ was synthesized as previously re-
ported9 and reacted with acetate to isolate a five-coordinate
12774
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
Scheme 2. Proposed Reaction Scheme for Electrocatalytic Formate Oxidation
nickel acetate complex. The nickel acetate complex is stable,
whereas the formate complex is not. This instability of the
formate complex is attributed to the possible net hydride transfer
to liberate CO2 and generate the nickel hydride. Higher con-
centrations of acetate led to almost complete displacement of the
phosphine ligands from nickel. This favorable binding of acetate
suggests that formate may also bind through oxygen to form a
complex analogous to the structurally characterized acetate
complex, [Ni(PPh2NBn2)2(OAc)]+ (Figure 2), as shown in step
B of the proposed catalytic cycle shown in Scheme 2.
For catalysts with a phenyl substituent at phosphorus, the
catalytic activity correlates with the pKa of the conjugate acid
Figure 10. Correlation of TOF for electrocatalytic formate oxidation
+
(R0NH3 ) of the free amine used to make the corresponding
+
with the pKa of the free primary ammonium (R0NH3 ) used to
0
PR NR ligand (see the blue data and line in Figure 10). This
0
0
synthesize the corresponding PR2NR ligands for each Ni(PR2NR 2)2-
2
2
2
2+
(CH3CN)]2+ complex. The data point labels in the figure refer to the R
and R0 substituents, respectively, and the trend line and corresponding
equation are only for R = Ph.
correlation and the lack of catalytic activity for Ni(depe)2
suggest that the rate-determining step requires the presence of
the pendant amine. In addition, the homoleptic complexes with
the bulkier cyclohexyl substituents at phosphorus exhibit lower
TOFs than their phenyl analogues, indicating that increased
steric bulk decreases catalytic activity. A mechanism that is con-
sistent with all of the above observations is shown in Scheme 2. In
of the resulting nitrogen protonated Ni(0) complex appears to be
fast and leads to the reformation of Ni(I), as in step E of Scheme 2.
Regardless of the precise mechanistic features, which are still
under investigation, a comparison of the TOFs observed for
formate oxidation by the nickel-based catalysts used in this study
with those of catalytic thermal formate oxidations show that the
rates observed here (from <1 to 16 sꢀ1) at room temperature are
comparable to those of the best molecular catalysts for formate
oxidation reported to date in the literature: ruthenium complexes
this scheme, the oxidation of Ni(I) to Ni(II) is followed by a fast
0
reaction with formate to generate [Ni(PR NR ) (O2CH)]+,
2
2 2
with formate binding through one oxygen. The steric retardation
of the catalytic rate for cyclohexyl vs phenyl substituents at
phosphorus may be the result of this equilibrium step being shifted
to disfavor formate association. As illustrated in steps C and D of
Scheme 2, the oxygen-bound formate complex is thought to
undergo a proton transfer from the carbon atom of formate to a
pendant nitrogen atom with a concomitant transfer of two
electrons to nickel and the release of CO2. It is this heterolytic
cleavage of formate that is proposed as the rate-limiting step in the
overall catalytic cycle. Subsequent deprotonation and re-oxidation
with TOFs of 3600 hꢀ1 (1 sꢀ1) at 40 °C and 18 000 hꢀ1 (5 sꢀ1
)
at 120 °C,7,39,40 and iridium complexes with TOFs as high as
14000 hꢀ1 (4 sꢀ1) at 90 °C.41,42 In addition, the failure to observe
any measurable catalytic currents for analogous nickel dipho-
sphine complexes lacking pendant amines illustrates the importance
12775
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
of the pendant amines in catalytic reactions involving multi-
electron and multi-proton processes.
X-ray structural determinations were mounted in polybutene oil on a
glass fiber and transferred on the goniometer head to the precooled
instrument. Crystallographic measurements were carried out on
either a Bruker P4 or platform diffractometer using Mo KR radiation
(λ = 0.71073 Å) in conjunction with a Bruker APEX detector. All
structures were solved by direct methods using SHELXS-97 and refined
with full-matrix least-squares procedures using SHELXL-97.43 All non-
hydrogen atoms are anisotropically refined unless otherwise reported; the
hydrogen atoms were included in calculated positions as riding models in
the refinement. Crystallographic data collection and refinement infor-
mation can be found in the Supporting Information.
’ SUMMARY AND CONCLUSIONS
Thermodynamic considerations of the hydride donor abilities
of the formate ion and the hydride acceptor abilities of
[Ni(diphosphine)2]2+ complexes suggested that hydride transfer
from formate to these nickel complexes should occur. In addi-
tion, pendant amines were incorporated in the diphosphine
ligands to assist in proton transfer reactions that were expected
to occur during0the catalytic reactions. Using these guidelines,
new [Ni(PR NR ) (CH3CN)]2+ and previously reported com-
X-ray Structural Analysis for [Ni(PPh2NBn2)2(OAc)](BF4). A
single red block (0.10 ꢂ 0.10 ꢂ 0.10 mm) was mounted using NVH
immersion oil onto a nylon fiber and cooled to the data collec-
tion temperature of 100(2) K. Data were collected on a Br€uker-AXS
Kappa APEX II CCD diffractometer with 0.71073 Å Mo KR radiation.
Unit cell parameters were obtained from 36 data frames, 0.5° Φ, from
three different sections of the Ewald sphere yielding a = 24.0216(11),
b = 17.6157(8), and c = 17.6157(8) Å, V = 6180.8(5) Å3. A total of 91 964
reflections (Rint = 0.1006) were collected (12749 unique) over θ = 1.43ꢀ
26.49°. The systematic absences in the diffraction data were consistent
with the centrosymmetric, orthorhombic space group, Pna2(1). The
data set was treated with SADABS absorption corrections based on
2
2 2
plexes were prepared that allowed a systematic variation in the
hydride acceptor ability of the metal and the basicity of the
0
pendant amine. These [Ni(PR NR ) (CH3CN)]2+ complexes
2
2 2
were demonstrated to be catalysts for the electrocatalytic oxida-
tion of formate, and mechanistic studies demonstrated that the
pendant amine plays an important role in the rate-determining
step that is thought to involve proton transfer from the formate
carbon atom to a pendant amine of the ligand. The catalytic TOFs
for these electrocatalysts are as high as any reported thermal
formate/formic acid oxidation catalysts, and to our knowledge, the
present catalysts are the first electrocatalysts to utilize a base metal,
nickel, rather than the more typical platinum, palladium, or
rhodium. Furthermore, these are the first homogeneous com-
plexes reported for the electrocatalytic oxidation of formate.
redundant multiscan data (G. Sheldrick, Bruker-AXS, 2001), Tmax/Tmin
=
1.00. The asymmetric unit contains one [Ni(PPh2NBn2)2(OAc)]+ cation,
one [BF4]ꢀ anion, and one molecule of Et2O solvent located on general
positions, yielding Z = 4 and Z0 = 1. All non-hydrogen atoms were refined
with anisotropic displacement parameters. All hydrogen atoms were treated
as idealized contributions. The goodness-of-fit on F2 was 1.022, with R1
(wR2) = 0.0429 (0.0889) for [Iθ > 2(I)] and with largest difference peak
and hole of 0.653 and ꢀ0.327 e/Å3.
’ EXPERIMENTAL SECTION
Synthesis of [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2. Kilgore et al.
very recently reported the synthesis of [Ni(PPh2NPhOMe2)2(CH3CN)]-
(BF4)2 using a similar procedure.16 For details of the synthesis used for
the present work, see the Supporting Information. Single crystals
suitable for X-ray diffraction were obtained after several days by layering
a concentrated acetonitrile solution with diethyl ether at ꢀ35 °C. 1H and
31P NMR data matched the previously reported data.16 13C{1H} NMR
(CD3CN, 20 °C, 125 MHz) δ [ppm]: 156.3 (CaromO), 145.6 (CaromN),
131.3 and 131.9 (CaromH, PPh), 129.6 (CaromP), 114.9 and 122.1
(CaromH, NPh), 59.1 (PCH2N), 55.5 (OCH3). IR (KBr) νmax [cmꢀ1]:
3057 (w), 3005 (w), 2938, (w), 2839 (w), 1511 (vs), 1438 (m), 1249
(s), 1191 (m), 1058 (s), 1035 (s), 890 (m), 839 (m), 816(w), 743 (m),
695 (m), 496 (m). Elemental analysis: C/H/N calcd for NiC62H67-
N5O4P4B2F8, 57.18/5.19/5.38; found, 57.16/5.32/5.61.
Materials and Methods. All chemicals were purchased from
commercial sources and used as received unless otherwise specified.
The complexes [Ni(PCy2NBn2)2](BF4)2, [Ni(PPh2NPh2)2(CH3CN)]-
(BF4)2, [Ni(PPh2NBn2)2(CH3CN)](BF4)2, [Ni(PPh2NMe2)2(CH3-
CN)](BF4)2, and [Ni(PPh2NPhCF32)2(CH3CN)](BF4)2 were prepared
as previously described.9,12,16,28 All manipulations were carried out using
standard Schlenk and glovebox techniques under an atmosphere of
nitrogen. Acetonitrile and tetrahydrofuran (THF) were sparged with
argon and dried over basic alumina with a custom dry solvent system.
Pentane and diethyl ether (Et2O) were distilled over sodium benzo-
phenone ketyl. 1H, 13C, and 31P NMR spectra were recorded on a Varian
300, 400, or 500 spectrometer or a Jeol ECA-500 spectrometer. The 1H
and 13C NMR spectra are referenced to TMS using the residual
resonances of the solvent. The 31P NMR shifts are referenced to H3PO4.
All electrochemical experiments were carried out under an atmo-
sphere of nitrogen in a 0.2 M tetrabutylammonium trifluoromethane-
sulfonate (NBu4OTf) solution in benzonitrile using a CH Instruments
600 or 1100 series three-electrode potentiostat. The working electrode
was a glassy carbon disk (1 mm diameter), and the counter electrode was
a glassy carbon rod. A silver wire in electrolyte solution separated from
the working compartment by porous Vycor (4 mm, BAS) was used as a
pseudo-reference electrode. All potentials were measured using Cp2Fe
(0 V), Cp*2Fe (ꢀ0.50 V), or Cp2Co+ (ꢀ1.33 V) as an internal reference,
with all of the potentials reported vs the Cp2Fe+/0 couple. All stock
Synthesis of [Ni(PCy2NPh2)(CH3CN)2](BF4)2. Cyclohexylpho-
sphine (1.06 mL, 8 mmol) was placed in a Schlenk flask and was
transferred with dry ethanol (150 mL) into another flask containing
paraformaldehyde (0.46 g, 15 mmol). The reaction mixture was stirred
under reflux for 2 h until the suspension became a clear solution. Aniline
(0.73 mL, 8 mmol) dissolved in ethanol (20 mL) was added via cannula
to the reaction flask, and the reaction mixture was stirred for 2 h under
reflux. After 5 min the precipitation of a white fluffy solid was observed.
The reaction was cooled to room temperature, the solvent removed by
cannula filtration, and the precipitate washed twice with ethanol. The
resulting white solid was dried in vacuum and used without further
purification. Based on the initial cyclohexylphosphine, 0.8 equiv of
[Ni(CH3CN)6](BF4)2 (1.08 g, 3.2 mmol) in acetonitrile was added
to the crude ligand. The solution turned red and was stirred overnight.
The resulting solution was filtered and layered with diethyl ether. After
2 days, large red needles of [Ni(PCy2NPh2)(CH3CN)2](BF4)2 were
collected; yield, 59% (1.49 g, 1.9 mmol). Single crystals suitable for X-ray
diffraction were obtained from a CH3CN/Et2O solution at ꢀ35 °C.
1H NMR (CD3CN, 20 °C, 400 MHz) δ [ppm]: 7.51 (t, 4H, 3J = 8 Hz,
solutions of catalyst, NBu4HCO2 HCO2H, and electrolyte were freshly
3
prepared in a glovebox under a nitrogen atmosphere. UVꢀvis spectra were
collected on a Shimadzu UV-3600 instrument in 0.1 cm quartz cuvettes. IR
spectra were recorded as KBr pellets using a FT-IR Bruker Equinox 55
spectrometer. Elemental analyses were performed by Midwest MicroLab,
LLC, Indianapolis, IN, or by Atlantic Microlab, Inc., Norcross, GA.
Crystals of [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2, [Ni(PCy2NPh2)2-
(CH3CN)](BF4)2, [Ni(PCy2NPh2)(CH3CN)2](BF4)2, [Ni(PCy2NPh2)-
(PPh2NPh2)(CH3CN)](BF4)2, and [Ni(PPh2NPhOMe2)2] suitable for
12776
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
CaromH, NPh), 7.21 (d, 4H, 3J = 8 Hz, CaromH, NPh), 7.16 (t, 2H, 3J = 8
Hz, CaromH, NPh), 4.02 (d, 4H, 2J = 12 Hz, PCyCH2NPh), 3.58 (d, 4H,
2J = 12 Hz, PCyCH2NPh), 2.41 (m broad, 2H, PCy), 2.03 (s broad, 4H,
(w), 759 (w), 695 (w), 521 (vs). Elemental analysis: C/H/N calcd for
NiC58H83N5P4B2F8, 57.74/6.93/5.80; found, 57.88/6.92/6.11.
Synthesis of [HNi(PPh2NPhOMe2)2](BF4. Method a: [Ni(PPh
-
2
NPhOMe2)2(CH3CN)](BF4)2 (63.2 mg, 48.5 μmol) was dissolved in
acetonitrile in a glovebox, and this solution was added to a vial containing
excess sodium formate. The red solution turned yellow over 20 min.
The excess sodium formate was removed by filtration, and the
product was isolated by removing the solvent under vacuum. The
resulting solid was washed twice with pentane and dried under vacuum.
Method b: The hydride complex [HNi(PPh2NPhOMe2)2](BF4) can be
obtained by the same procedure using [Ni(PPh2NPhOMe2)2(CH3-
CN)](BF4)2 (101.6 mg, 78 μmol) and 1 equiv of sodium trimethox-
yborohydride (NaB(OMe)3H) (10.0 mg, 78 μmol) instead of sodium
P
Cy), 1.85 (m broad, 4H, PCy), 1.76 (d, 2H, 3J = 8 Hz, PCy), 1.52ꢀ1.41
(m, 8H, PCy), 1.36ꢀ1.31 (m broad, 2H, PCy). 31P{1H} NMR (CD3CN,
20 °C, 162 MHz) δ [ppm]: 9.8 (s).
Synthesis of [Ni(PCy2NPh2)(PPh2NPh2)(CH3CN)](BF4)2. The
PPh2NPh2 ligand12 was prepared using the same procedure as for PCy2NPh
2
described above, but with phenylphosphine (0.375 mL, 2.8 mmol),
paraformaldehyde (168 mg, 5.6 mmol), and aniline (0.256 mL, 2.8
mmol) as reagents. After filtration and drying of the resulting white solid
in vacuum, 0.8 equiv of [Ni(PCy2NPh2)(CH3CN)2](BF4)2 (0.85 g, 1.1
mmol) was added to the crude ligand via cannula. The reaction mixture
was stirred overnight, resulting in a dark red mixture. After filtration, the
solvent was removed, and the dark red solid was dried under vacuum.
The product was recrystallized by layering a concentrated acetonitrile
solution with diethyl ether at room temperature. Single crystals suitable
for X-ray diffraction were obtained by vapor diffusion of diethyl ether
into a concentrated solution of [Ni(PCy2NPh2)(PPh2NPh2)(CH3-
CN)](BF4)2 in acetonitrile at ꢀ35 °C; yield, 67% (0.90 g, 0.75 mmol).
1H NMR (CD3CN, 20 °C, 500 MHz) δ [ppm]: 7.85ꢀ7.82 (m, 2H,
CaromH, PPh), 7.72 (t, 1H, 3J = 8 Hz, CaromH, PPh), 7.61 (t, 1H, 3J = 8 Hz,
CaromH, PPh), 7.40ꢀ7.34 (m, 4H, CaromH, NPh), 7.25 (d, 2H, 3J = 9 Hz,
CaromH, NPh), 7.13 (d, 2H, 3J = 9 Hz, CaromH, NPh), 7.08ꢀ7.04 (m, 2H,
CaromH, NPh) 4.50ꢀ3.65 (broad, 2H, PPhCH2NPh, 4H PCyCH2NPh),
3.61 (d, 2H, 2J = 14 Hz, PPhCH2NPh), 1.60ꢀ1.44 (m broad, 5H, PCy),
1.32ꢀ1.19 (m broad, 2H, PCy), 1.11ꢀ1.04 (m, 1H, PCy), 0.92 (t broad,
1
formate. H NMR (CD3CN, 20 °C, 400 MHz) δ [ppm]: 7.62ꢀ7.61
(m, 8H, CaromH, PPh), 7.46 (t, 4H, 3J = 7.4 Hz, CaromH, PPh), 7.26 (t, 8H,
3J = 7.6 Hz, CaromH, PPh), 7.03 (d, 8H, 3J = 9.2 Hz, CaromH, NPh), 6.89
(d, 8H, 3J = 9.2 Hz, CaromH, NPh), 3.89 (d, 8H, 2J = 12.8 Hz, PCH2N),
2
3.78 (s, 12H, PCH2N), 3.42 (d, 8H, J = 13.2 Hz, PCH2N), ꢀ8.05
(quintet, 1H, 2J = 30.3 Hz, Ni-H). 31P{1H} NMR (CD3CN, 20 °C, 162
MHz) δ [ppm]: 15.1 (s).
Synthesis of [Ni(PPh2NPhOMe2)2. Method a: [Ni(PPh2NPhOMe2)2-
(CH3CN)](BF4)2 (63.2 mg, 48.5 μmol) was dissolved in acetonitrile and
added to a vial containing excess sodium formate. The red solution
turned yellow over 20 min, and a yellow precipitate formed.
The reaction was complete when the solution became transparent,
indicating complete precipitation of the product. The solvent was
removed by filtration, and the resulting solid, a mixture of NaBF4,
NaHCO2, and [Ni(PPh2NPhOMe2)2], was dried under vacuum and
washed three times with pentane. The product can be extracted with
THF. Single crystals suitable for X-ray diffraction were obtained by
vapor diffusion of pentane into a concentrated THF solution of
[Ni(PPh2NPhOMe2)2]. Method b: Complex [Ni(PPh2NPhOMe2)2] was
also obtained by the same procedure through the reaction of the
complex [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2 (101.6 mg, 78 μmol)
with 2 equiv of sodium trimethoxyborohydride (NaB(OMe)3H) (20.0
mg, 156 μmol) instead of sodium formate. 1H NMR (THF-d8, 20 °C,
300 MHz) δ [ppm]: 7.86 (t, broad, 8H, J = 6 Hz, CaromH, PPh),
2
1H, J = 12 Hz, PCy), 0.86ꢀ0.38 (broad, 1H, PCy). 13C{1H} NMR
(CD3CN, 20 °C, 75.5 MHz) δ [ppm]: 152.5 and 152.3 (CaromN), 40.8
(CaromH, NPh), 133.8 (m, Carom, PPh), 132.8 (CaromH, NPh), 131.6 (m,
Carom, PPh), 130.7 and 130.5 (broad, Carom, PPh), 124.1, 123.2, 120.7 and
119.4 (CaromH, NPh), 47.4 and 36.0 (broad, PCH2N), 28.5 and 27.8
(broad, PCy), 25.7 (PCy). 31P{1H} NMR (CD3CN, 20 °C, 121 MHz) δ
[ppm]: 8.40ꢀ7.40 (m) and 3.05ꢀ7.93 (m). IR (KBr) νmax [cmꢀ1]:
3389 (w), 3083 (w), 2986 (w), 2884 (w), 2822 (w), 1581 (s), 1485 (s),
1430 (vs), 1268 (m), 1226 (s), 1176 (s), 1131 (s), 969 (m), 894 (m),
857 (m), 770 (m), 718 (m), 685 (m), 514 (m), 459 (w). Elemental
analysis: C/H/N calcd for NiC58H71N5P4B2F8, 58.32/5.99/5.86;
found, 58.08/6.11/5.73.
3
7.01ꢀ7.10 (m, 12H, CaromH, PPh), 6.99 (d, 8H, J = 9 Hz, CaromH,
N
Ph), 6.77 (d, 8H, 3J = 9 Hz, CaromH, NPh), 3.89 (d, 8H, 2J = 12 Hz,
2
Synthesis of [Ni(PCy2NPh2)2(CH3CN)](BF4)2. The ligand
PCH2N), 3.70 (s, 12H), 3.50 (d, 8H, J = 12 Hz, PCH2N), 3.45 (s,
12H). 31P{1H} NMR (THF-d8, 20 °C, 121.5 MHz)
δ [ppm]: 7.5 (s).
PCy2NPh was synthesized as described above for [Ni(PCy2NPh2)-
2
(CH3CN)2]2+. [Ni(CH3CN)6](BF4)2 (257 mg, 0.5 mmol) was dis-
solved in degassed acetonitrile (100 mL total) and transferred via
cannula into a flask containing the crude ligand (0.49 g, 1.0 mmol).
The reaction mixture immediately changed to red. After being stirred
overnight, the reaction mixture was canula filtered to remove unreacted
ligand, and the solvent was then removed by vacuum. The resulting red
solid was recrystallized from an acetonitrile solution of [Ni(PCy2NPh2)2-
(CH3CN)](BF4)2 layered with diethyl ether at room temperature.
Single crystals suitable for X-ray diffraction were obtained from a
concentrated acetonitrile diethyl ether solution of the compound at
Synthesis of [Ni(PPh2NBn2)2(OAc)](BF4) Et2O. A solution of
3
tetrabutylammonium acetate (14.5 mg, 0.049 mmol) in benzonitrile
was added to a solution of [Ni(PPh2NBn2)2(CH3CN)](BF4)2 (60 mg,
0.048 mmol) in benzonitrile. The solution immediately changed from
red to dark purple. Vapor diffusion of diethyl ether into the crude
reaction mixture at room temperature afforded dark red crystals suitable
for X-ray crystallography; yield, 38% (21 mg, 0.018 mmol). 31P{1H}
NMR (CD2Cl2, 20 °C): an AA0BB0 spectrum was observed with
resonances centered at 20.4 and ꢀ18.6 ppm.
1
Synthesis of NBu4HCO2 HCO2H. A solution of tetrabutylam-
ꢀ35 °C; yield, 73% (453 mg, 375 μmol). H NMR (CD3CN, 20 °C,
3
400 MHz) δ [ppm]: 7.35 (t, 8H, 3J = 8 Hz, CaromH, NPh), 7.10ꢀ7.04
(m, 12H, CaromH, NPh), 4.08 (d, 4H, 2J = 16 Hz, PCH2N), 3.82ꢀ3.73
(m, 8H, PCH2N), 3.56 (d, 4H, 2J = 16 Hz, PCH2N), 2.24 (t, 6H, J = 12
Hz, PCy), 2.10 (d, 4H, J = 12 Hz, PCy), 1.91ꢀ1.32 (m, 34H, PCy).
13C{1H} NMR (CD3CN, 20 °C, 300 MHz) δ [ppm]: 151.8 (CaromN,
3JPꢀC = 4 Hz), 130.4, 123.1, and 119.2 (CaromH), 49.3ꢀ49.1 and
47.4ꢀ47.1 (m broad, PCH2N), 39.7 (m, coord CH3CN), 28.9 (CCyP),
28.2 (broad, CCy), 27.6 (broad, CCy), 25.9 (CCy); the signal for the
quaternary carbon nucleus of the CH3CN could not be detected.
31P{1H} NMR (CD3CN, 20 °C, 162 MHz) δ [ppm]: 7.2 (s). IR
(KBr) νmax [cmꢀ1]: 3063 (w), 2958 (m), 2909 (m), 2849 (m), 1598
(m), 1496 (m), 1452 (w), 1272 (w), 1195(m) 1100 (s), 1021 (s), 890
monium hydroxide (3.8 M, 5.0 mL, 19 mmol) was placed into a
Schlenk flask with a stir bar. Formic acid (0.96 mL, 19.0 mmol) was
added dropwise over 5 min at room temperature and allowed to stir for
3 h. Water (0.5 mL) was added to reaction mixture, which was
then extracted with EtOAc (3 ꢂ 15 mL). The organic phase was dried
over MgSO4 and concentrated to give a colorless oil. Diethyl ether
(30 mL) was added and concentrated again, yielding a tacky white
solid. The solid was dried under vacuum for several hours and stored
in a glovebox. The product was crystallized by layering a saturated
solution of NBu4HCO2 HCO2H in THF with hexane; yield, 1.96 g
3
(6.8 mmol), 36%. 1H NMR (300 MHz, CD3CN) δ [ppm]: 18.7
(s, 0.8 H, HCO2H-O2CH), 8.55 (s, 1.7 H, HCO2H-O2CH), 3.22 (m,
12777
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
8.4 H, N(CH2CH2CH2CH3)4), 1.70 (m, 8.4 H, N(CH2CH2CH2CH3)4),
1.43 (sextet, 8.2 H, JHꢀH = 6.0 Hz, N(CH2CH2CH2CH3)4), 1.04
(t, 12.0 H, JHꢀH = 6.0 Hz, N(CH2CH2CH2CH3)4). Elemental analysis:
C/H/N calcd for C18H39NO4, C, 64.82; H, 11.79; N, 4.20; found, C,
64.22; H, 11.53; N, 4.29.
were recorded to determine catalytic currents. Plots of icat vs [cat] were
used to determine the order with respect to catalyst, as shown in
Figure 8.
Bulk Electrolysis. A multineck conical flask was used for the bulk
electrolysis experiment. A stainless steel mount was attached to one
neck, and to this mount a cylinder of reticulated vitreous carbon was
attached as the working electrode. Two of the other necks of the flask
were fitted with glass tubes terminating with 7 mm Vycor frits
(Princeton Applied Research). One tube was used for the reference
electrode and the other tube for the counter electrode. Both were filled
with electrolyte solutions (0.2 M NBu4OTf in benzonitrile), and the
reference cell contained a silver chloride-coated silver wire, whereas a
Nichrome wire was used in the countercell. With the fittings attached,
the cell had a total volume of 328 mL. The cell was filled with stock
General Procedure for Hydricity Determination by Hetero-
lytic H2 Cleavage: [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2. In a typical
experiment, [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2 (15 mg, 0.012 mmol)
and p-anisidine (6.0 mg, 0.049 mmol) were dissolved in 0.5 mL of dry
acetonitrile-d3 and transferred to an NMR tube equipped with a septum.
Hydrogen gas was bubbled through the solution for 20 min. After an
1
additional 10 min, the sample was analyzed by H and 31P NMR
spectroscopy. The weighted averages of the aromatic proton
chemical shifts of p-anisidine/p-anisidinium were used to deter-
mine the ratio of acid to base for the reference, and the ratio of
solutions of catalyst and NBu4HCO2 HCO2H to give 20 mL of a
+
[Ni(PPh2NPhOMe2)2(CH3CN)]2+ to [HNi(PPh2NPhOMe2)2] was deter-
3
solution that was 0.76 mM in [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2,
mined by 31P NMR spectroscopy. These ratios were then used to deter-
mine the equilibrium constant. The samples were run again after 24 h to
ensure equilibration. For [Ni(PCy2NPh2)2(CH3CN)](BF4)2, p-bro-
0.086 M in NBu4HCO2 HCO2H, and 0.2 M in NBu4OTf. Controlled-
3
potential coulometry was performed at ∼100 mV positive of the peak of
the catalytic current. After 16.3 C of charge had passed, samples of the
gas in the headspace of the flask were removed via a gastight syringe and
analyzed by gas chromatography. The percentages of N2 and CO2 in the
headspace were determined through calibration against gas standards of
a known composition. From these data, 5.3 mol of CO2 per mole of
catalyst was produced, a current efficiency of 93 ( 5% was calculated for
CO2 production, and no H2 was observed.
moaniline was used as the base, and for [Ni(PPh2NPh2)(PCy2NPh
(CH3CN)](BF4)2, aniline was used as the base.
Reactivity toward NBu4HCO2 HCO2H: General Procedure
)
2
3
for NMR Titration. In a glovebox, a solution of [Ni(PPh
-
2
NPhOMe2)2(CH3CN)](BF4)2 (5.0 mg, 0.0038 mmol) in 0.4 mL of
CD3CN was prepared in a 4 mL vial and transferred to a standard NMR
tube capped with a septum. A stock solution of NBu4HCO2 HCO2H
3
was prepared by dissolving 331 mg (0.99 mmol) in a second 4 mL vial
with 0.3 mL of CD3CN. Aliquots of the NBu4HCO2 HCO2H solution
3
(1.0 μL, 0.0033 mmol) were added to the NMR tube via a 10 μL gastight
syringe. The NMR tube was shaken thoroughly, and 1H and 31P NMR
spectra were acquired. This procedure was followed until all the starting
material was consumed. The 31P{1H} NMR spectra are shown in the
Supporting Information.
’ ASSOCIATED CONTENT
S
Supporting Information. Crystallographic structure for
b
NBu4HCO2 HCO2H and crystallographic data and CIF files
3
for NBu4HCO2 HCO2H, [Ni(PPh2NPhOMe2)2(CH3CN)]2+, [Ni-
3
Electrocatalytic Formate Oxidation: Order with Respect to
Formate. The catalyst was dissolved in a benzonitrile solution with 0.2
M NBu4OTf to make a solution with a total volume of 1.0 mL and a final
catalyst concentration of 1.0 ꢂ 10ꢀ3 M. The solution was purged with
N2 for 5 min, and an initial cyclic voltammogram was recorded. Aliquots
of NBu4HCO2 HCO2H were added, and cyclic voltammograms
were recorded to determine catalytic currents. Plots of TOF vs [NBu4-
HCO2 HCO2H] were used to determine the order with respect to
(PCy2NPh2)2(CH3CN)]2+, [Ni(PCy2NPh2)(CH3CN)2]2+, [Ni(PCy
-
2
N
Ph2)(PPh2NPh2)(CH3CN)]2+, and [Ni(PPh2NPhOMe2)2]; 31P{1H}
NMR spectra of the reactions of [Ni(PPh2NPhOMe2)2-
(CH3CN)]2+, [Ni(PCy2NPh2)2(CH3CN)]2+, and [Ni(PCy2NPh2)-
(PPh2NPh2)(CH3CN)]2+ with NBu4HCO2 HCO2H; detailed syn-
3
thetic procedure for [Ni(PPh2NPhOMe2)2(CH3CN)]2+. This material
is available free of charge via the Internet at http://pubs.acs.org.
3
3
formate (see eqs 11 and 12 and Figure 7). The peak current (ip) is the
observed current for the Ni(II/I) in the absence of formate. The catalytic
current (icat) is the observed peak or plateau current observed in the
’ AUTHOR INFORMATION
Corresponding Author
aaron.appel@pnl.gov; ckubiak@ucsd.edu
presence of NBu4HCO2 HCO2H.
3
Electrocatalytic Formate Oxidation: Titration of [Ni(PPh
-
2
NPhOMe2)2(CH3CN)](BF4)2 with NBu4HCO2 HCO2H and DBU.
3
[Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2 (6.5 mg, 0.005 mmol) was dis-
solved in a 5.0 mL volumetric flask with 0.2 M NBu4OTf benzonitrile
’ ACKNOWLEDGMENT
solution. In a 1.0 mL volumetric flask, NBu4HCO2 HCO2H (333 mg,
1.0 mmol) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU, 150 μL,
1.0 mmol) were dissolved in 0.2 M NBu4OTf benzonitrile solution.
Funding by Deutsche Forschungsgemeinschaft (J.S.), the NSF
GRFP (C.S.), and by the Helios Solar Energy Research Center,
which is supported by the Director, Office of Science, Office of
Basic Energy Sciences of the U.S. Department of Energy under
Contract No. DE-AC02-05CH11231, is gratefully acknowl-
edged. B.R.G., J.C.L., A.M.A., and D.L.D. were supported by
the U.S. Department of Energy, Office of Science, Office of Basic
Energy Sciences, Division of Chemical Sciences, Biosciences and
Geosciences. J.A.S.R., M.L.H., U.J.K., and J.Y.Y. were supported
as part of the Center for Molecular Electrocatalysis, an Energy
Frontier Research Center funded by the U.S. Department of
Energy, Office of Science, Office of Basic Energy Sciences, under
FWP 56073. Pacific Northwest National Laboratory is operated
by Battelle for the U.S. Department of Energy.
3
Aliquots of the NBu4HCO2 HCO2H/DBU solution were added to
3
1.0 mL of the [Ni(PPh2NPhOMe2)2(CH3CN)](BF4)2 stock solution in
5 equiv increments. Cyclic voltammograms were recorded to determine
catalytic currents. Rates and TOF were similar to those of the [Ni(PPh
-
2
NPhOMe2)2(CH3CN)](BF4)2 with NBu4HCO2 HCO2H in the absence
3
of DBU, specifically 10.6 sꢀ1 with DBU vs 8.5 without DBU.
Electrocatalytic Formate Oxidation: Order with Respect
to Catalyst. The order with respect to catalyst was determined by
titration of a 0.15 M NBu4HCO2 HCO2H solution with catalyst,
3
both dissolved in benzonitrile with 0.2 M NBu4OTf. The solution was
purged with N2 for 5 min, and an initial cyclic voltammogram was
recorded. Aliquots of catalyst were added, and cyclic voltammograms
12778
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779

Journal of the American Chemical Society
ARTICLE
’ REFERENCES
(31) 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.
(32) Yonker, C. R.; Linehan, J. C. J. Organomet. Chem. 2002, 650,
249–257.
(1) A Modular Approach to the Development of Molecular Electro-
catalysts for H2 Oxidation and Production Based on Inexpensive Metals;
Wiley-VCH: Weinheim, 2010; pp 165ꢀ180.
(2) Rice, C.; Ha, S.; Masel, R. I.; Wieckowski, A. J. Power Sources
2003, 115, 229–235.
(33) Yonker, C. R.; Linehan, J. C. Prog. Nucl. Magn. Reson. Spectrosc.
2005, 47, 95–109.
(3) Song, C. Catal. Today 2002, 77, 17–49.
(34) Nicholson, R. S.; Shain, I. Anal. Chem. 1964, 36, 706–723.
(35) Saveant, J. M.; Vianello, E. Electrochim. Acta 1965, 10, 905–920.
(36) Savꢀeant, J. M.; Vianello, E. Electrochim. Acta 1967, 12, 629–646.
(37) Izutsu, K. (International Union of Pure and Applied Chemistry,
Commission on Electroanalytical Chemistry). Acidꢀbase dissociation
constants in dipolar aprotic solvents; Blackwell Scientific: Oxford, 1990.
(38) Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.;
Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019–1028.
(39) Boddien, A.; Loges, B.; Junge, H.; Beller, M. ChemSusChem
2008, 1, 751–758.
(40) Morris, D. J.; Clarkson, G. J.; Wills, M. Organometallics 2009,
28, 4133–4140.
(41) Himeda, Y. Green Chem. 2009, 11, 2018–2022.
(42) Himeda, Y.; Miyazawa, S.; Hirose, T. ChemSusChem 2011,
4, 487–493.
(4) Zhu, Y.; Ha, S. Y.; Masel, R. I. J. Power Sources 2004, 130, 8–14.
(5) Reutemann, W., ; Kieczka, H. Formic AcidꢀUllmann’s Encyclo-
pedia of Industrial Chemistry; Wiley-VCH: Weinheim, 2002.
(6) Reda, T.; Plugge, C. M.; Abram, N. J.; Hirst, J. Proc. Natl. Acad.
Sci. U.S.A. 2008, 105, 10654–10658.
(7) Enthaler, S.; von Langermann, J.; Schmidt, T. Energy Environ. Sci.
2010, 3, 1207–1217.
(8) Fellay, C.; Dyson, P. J.; Laurenczy, G. Angew. Chem., Int. Ed.
2008, 47, 3966–3968.
(9) Fraze, K.; Wilson, A. D.; Appel, A. M.; Rakowski DuBois, M.;
DuBois, D. L. Organometallics 2007, 26, 3918–3924.
(10) Rakowski DuBois, M.; DuBois, D. L. Chem. Soc. Rev. 2009,
38, 62–72.
(11) (a) DuBois, D. L.; Berning, D. E. Appl. Organomet. Chem. 2000,
14, 860–862. (b) Curtis, C. J.; Miedaner, A.; Ellis, W. W.; DuBois, D. L.
J. Am. Chem. Soc. 2002, 124, 1918–1925.
(12) 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.
(43) Sheldrick, G. M. SHELXL 97ꢀProgramm fur die Verfeinerung
von Kirstallstrukturen; Universitat Gottingen, 1997.
(13) M€arkl, V.; Jin, G. Y.; Schoerner, C. Tetrahedron Lett. 1980, 21,
1409–1412.
(14) Kane, J. C.; Wong, E. H.; Yap, G. P. A.; Rheingold, A. L.
Polyhedron 1999, 18, 1183–1188.
(15) Bobrov, S. V.; Karasik, A. A.; Sinyashin, O. G. Phosphorus Sulfur
and Silicon and the Related Elements 1999, 146, 289–292.
(16) Kilgore, U. J.; Roberts, J. A. S.; Pool, D. H.; Appel, A. M.;
Stewart, M. P.; Rakowski DuBois, M.; Dougherty, W. G.; Kassel, W. S.;
Bullock, R. M.; DuBois, D. L. J. Am. Chem. Soc. 2011, 133, 5861–5872.
(17) Yang, J. Y.; Bullock, R. M.; Dougherty, W. G.; Kassel, W. S.;
Twamley, B.; DuBois, D. L.; DuBois, M. R. Dalton Trans. 2010, 39,
3001–3010.
(18) Gonzꢀalez-Arellano, C.; Gutiꢀerrez-Puebla, E.; Iglesias, M.;
Sꢀanchez, F. Eur. J. Inorg. Chem. 2004, 1955–1962.
(19) Drew, J.; Hursthouse, M. B.; Thornton, P. J. Chem. Soc., Dalton
Trans. 1972, 1658–1660.
(20) Nagataki, T.; Tachi, Y.; Itoh, S. Chem. Commun. 2006,
4016–4018.
(21) Burling, S.; Elliott, P. I. P.; Jasim, N. A.; Lindup, R. J.; McKenna,
J.; Perutz, R. N.; Archibald, S. J.; Whitwood, A. C. Dalton Trans.
2005, 3686–3695.
(22) Miedaner, A.; Haltiwanger, R. C.; DuBois, D. L. Inorg. Chem.
1991, 30, 417–427.
(23) Raebiger, J. W.; Miedaner, A.; Curtis, C. J.; Miller, S. M.;
Anderson, O. P.; DuBois, D. L. J. Am. Chem. Soc. 2004, 126, 5502–5514.
(24) Nimlos, M. R.; Chang, C. H.; Curtis, C. J.; Miedaner, A.; Pilath,
H. M.; DuBois, D. L. Organometallics 2008, 27, 2715–2722.
(25) Curtis, C. J.; Miedaner, A.; Ciancanelli, R.; Ellis, W. W.; Noll, B. C.;
Rakowski DuBois, M.; DuBois, D. L. Inorg. Chem. 2003, 42, 216–227.
(26) Berning, D. E.; Noll, B. C.; DuBois, D. L. J. Am. Chem. Soc.
1999, 121, 11432–11447.
(27) Wayner, D. D. M.; Parker, V. D. Acc. Chem. Res. 1993, 26,
287–294.
(28) Yang, J. Y.; Bullock, R. M.; Dougherty, W. G.; Kassel, W. S.;
Twamley, B.; DuBois, D. L.; Rakowski DuBois, M. Dalton Trans. 2010,
39, 3001–3010.
(29) Yang, J. Y.; Bullock, R. M.; Shaw, W. J.; Twamley, B.; Fraze, K.;
RakowskiDuBois, M.;DuBois, D. L.J. Am. Chem. Soc. 2009, 131, 5935–5945.
(30) Berning, D. E.; Miedaner, A.; Curtis, C. J.; Noll, B. C.;
Rakowski DuBois, M. C.; DuBois, D. L. Organometallics 2001, 20,
1832–1839.
12779
dx.doi.org/10.1021/ja204489e |J. Am. Chem. Soc. 2011, 133, 12767–12779