The Influence of para Substituents in Bis(N-Heterocyclic Carbene) Palladium Pincer Complexes for Electrocatalytic CO2 Reduction

Article abstract of DOI:10.1021/acs.inorgchem.6b02213

The effect of modifying the pyridyl para position of lutidine-linked bis(N-heterocyclic carbene) Pd pincer complexes is studied both experimentally (R = OMe, H, Br, and COOR) and computationally, showing a strong effect on the first reduction potential of the complex and allowing the reduction potential to be tuned over a wide range in relation to the Hammett σ_p constant of the para substituent. The effect of the pyridyl para substituent on electron density of the metal center, frontier orbital energies, and dissociation energy of the trans ligand are also investigated in the context of reactivity with CO₂ through electrochemical characterization of the complexes under N₂ and CO₂ and controlled potential electrolysis experiments where CO₂ is reduced to CO.

Full text of DOI:10.1021/acs.inorgchem.6b02213

Article
pubs.acs.org/IC
The Influence of para Substituents in Bis(N-Heterocyclic Carbene)
Palladium Pincer Complexes for Electrocatalytic CO2 Reduction
Jeffrey A. Therrien and Michael O. Wolf*
Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, British Columbia V6T 1Z1, Canada
S
* Supporting Information
ABSTRACT: The effect of modifying the pyridyl para position of
lutidine-linked bis(N-heterocyclic carbene) Pd pincer complexes is
studied both experimentally (R = OMe, H, Br, and COOR) and
computationally, showing a strong effect on the first reduction potential
of the complex and allowing the reduction potential to be tuned over a
wide range in relation to the Hammett σp constant of the para
substituent. The effect of the pyridyl para substituent on electron
density of the metal center, frontier orbital energies, and dissociation
energy of the trans ligand are also investigated in the context of reactivity
with CO2 through electrochemical characterization of the complexes
under N2 and CO2 and controlled potential electrolysis experiments
where CO2 is reduced to CO.
monoxide ligands.5,11−13 Our contribution has been to further
INTRODUCTION
The development of more efficient methods for the reduction
of CO2 may enable its economical use as a source of carbon for
value-added chemicals and hydrocarbon fuels, in principle
■
investigate the pincer motif for CO2 reduction, first studied by
DuBois et al.14,15 with cationic Pd triphosphine complexes
capable of producing CO with high Faradaic efficiencies, by
incorporating a neutral redox-active lutidyl-linked bis-NHC
pincer ligand (C^N^C), also resulting in CO2 reduction
electrocatalysis.16 Furthermore, a tethered bimetallic triphos-
phine complex has been found to exhibit cooperative reactivity
with CO2 resulting in high reaction rates on the order of the
rate of [NiFe] CODH, though with a limited turnover
number.17 The lutidyl-linked bis-NHC ligand platform has
the modular capacity to be similarly extended and/or
incorporate additional functionalities to enhance reactivity
with CO2.
In our previous work, we have looked at modifications of the
NHC groups for Pd(C^N^C) complexes.18 Here, we
investigate, through a combined experimental and computa-
tional approach, modifications of the para substituent of the
bridging pyridyl donor to affect reactivity with CO2 by
modulation of the electron density of the metal center and
the ability of the monodentate trans ligand to dissociate.
Calculations have previously suggested the pyridyl moiety to be
the site of ligand redox activity upon reduction by one electron,
intimating its electrochemical significance.16 Similar para
substituent modifications for the bridging donor group in a
pincer ligand have previously been explored by Van Koten and
co-workers and Chirik and co-workers in relation to Hammett
para substituent constants (σp) with anionic 2,6-bis-
(dimethylamino)phenyl and neutral 2,6-bis(imino)pyridine
ligands, respectively, showing that such modifications can
allowing an anthropogenic carbon cycle where energy-dense
hydrocarbon fuels can be used in a sustainable, carbon-neutral
way.1−3 Approaching this challenge through the development
of molecular electrocatalysts is promising given the success seen
in the design of highly efficient hydrogen reduction electro-
catalysts4 and the existence of redox enzymes such as [NiFe]
carbon monoxide dehydrogenases ([NiFe] CODHs) which can
equilibrate CO2 and CO at extremely high rates under benign
conditions.5 The development of homogeneous electrocatalysts
for CO2 reduction has taken place over the past three decades,
with increasing interest over the past decade, contributing to
greater understanding of the factors required for reactivity with
CO2 and leading to electrocatalysts which can selectively
facilitate the two electron reduction of CO2 to CO or
HCOO−.6 Dihydrogen, H2, is often produced as a byproduct
due to the inherent presence of H+ in the CO2 reduction half-
reaction, thereby creating syngas which can be used in Fischer−
Tropsch processes to produce liquid hydrocarbon fuels.7
CO2 + 2e− + 2H+
→ CO + H2O
(−0.73 V vs Fc0/+in DMF)
2H+ + 2e− → H2 (−0.66 V vs Fc0/+in DMF)
Continuing work has been performed to further investigate
CO2 reduction with square-planar complexes containing
tetradentate ligands9,10 and octahedral Lehn-type complexes
typically containing a bipyridine (or related), halide, and carbon
Received: September 12, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
have a strong effect on redox potentials and catalytic
properties.19−22 Modification of the pyridyl para substituent
has not yet been performed for C^N^C pincers, with the
closest examples being pyridyl-linked bis-NHC (C−N−C)
pincers with the addition of a para carboxylate for surface
anchoring23 and reactivity of the para position in radical
species.24 The work reported here is the first example of an
investigation of the effects of modifying the pyridyl para
position for any C^N^C complex, especially in the context of
electrochemical properties and reactivity with CO2.
[Pd(C^N^Cp‑OMe)(CH3CN)](OTf)2. Abstraction was com-
plete within minutes at room temperature with R = OMe,
but required heating for 2 h in the case of R = Br, attributed to
the trans effect of the electronically modified pyridyl moiety
affecting lability of the halide.
The R = COOR species was synthesized according to
Scheme 2, where the alcohol groups of (4-bromopyridine-2,6-
diyl)dimethanol were protected by reaction with sodium
hydride and then chloromethyl methyl ether (MOM-Cl). The
use of weaker bases such as potassium carbonate was also
attempted but led to a complex mixture of products.
Additionally, protection of the alcohols by formation of methyl
ethers was attempted, leading to species susceptible to
controlled nucleophilic attack but difficult to deprotect, with
standard procedures with trimethylsilyl iodide, boron tribro-
mide, and even concentrated hydrobromic acid all failing. Only
extended heating (18 h) of a concentrated hydrobromic acid
solution at reflux led to deprotection of the methyl ethers. The
MOM-protected p-Br species was lithiated using 2 equiv of t-
BuLi at −78 °C in diethyl ether, producing a deep red solution
which was carboxylated by sparging with CO2(g), resulting in
an immediate color change to pale yellow. The resulting
carboxylate species was deprotected by stirring in 5 M HBr(aq)
at room temperature overnight. It should be noted that a
variety of lithiation procedures were attempted, and it was
found that the use of diethyl ether as the solvent was required
as the use of THF consistently resulted in an intractable
mixture of species under otherwise identical conditions,
perhaps due to the presence of three metalation directing
groups. Additionally, the reaction required t-BuLi for full
conversion; the use of n-BuLi consistently resulted in only
about 60% lithiation, indicating that the para-lithiated pyridine
species is less stable than typical sp2 carbanions as lithium-
halogen exchange reactions are equilibria determined by the
stability of the carbanion species involved.
RESULTS AND DISCUSSION
■
Synthesis and Characterization. A series of four
complexes with varied para substituents were prepared,
incorporating an electron-donating (ED) group (R = OMe;
σp = −0.27), electron-neutral group (R = H; σp = 0.00), mildly
electron-withdrawing (EW) group (R = Br; σp = +0.23), and
moderately EW group (R = COOR; σp = +0.45). The complex
[Pd(C^N^Cp‑H)Cl]BF4 was prepared previously,16 with the
other three complexes synthesized from the pyridine derivative
chelidamic acid, containing a hydroxyl group in the para
position and carboxylic acid groups in the ortho positions.
Following deprotection, an excess of phosphorus tribromide
was added to a suspension of the product in dichloromethane.
No reaction was observed after stirring overnight at room
temperature due to insolubility of 2,6-bis(hydroxymethyl)-
isonicotinic acid in the solvent, leading to the addition of THF
for additional solubility, which was unsuccessful, and finally to
the addition of N,N-dimethylformamide (DMF) wherein the
reaction occurred. After 4 h, methanol was added, and the
product was collected and purified by column chromatography,
yielding a pure, pale yellow oil. Through characterization by 1H
NMR and mass spectrometry (MS), the product was
determined to be 4-hydroxybutyl-2,6-bis(bromomethyl)-
isonicotinate and not the desired methyl ester species as a
result of the incorporation of ring-opened tetrahydrofuran to
the molecule. Given that the electronic properties are
equivalent to the methyl ester, this species was carried forward
and reacted with 2 equiv of 1-(n-butyl)imidazole to yield the
dibromide proligand salt, followed by coordination to
palladium through transmetalation from a silver carbene species
as described above. The reaction was sluggish, likely due to
increased solubility of the in situ silver carbene species
intermediate, and the crude product required additional
purification via column chromatography, yielding [Pd-
(C^N^Cp‑COOR)Cl]OTf in a low isolated yield but sufficient
purity for electrochemical study.
The R = Br and R = OMe species were synthesized according
to Scheme 1. For incorporation of the para-Br substituent,
chelidamic acid in chloroform was reacted with an excess of
phosphorus tribromide and bromine at reflux for 72 h,
producing a para-Br acid bromide which was quenched with
methanol to yield the para-Br methyl ester. The para-OMe
substituent was installed in a similar manner by reaction with
thionyl chloride and then methanol to form a para-hydroxyl
methyl ester species, followed by deprotonation and methyl-
ation with potassium carbonate and methyl iodide to yield the
para-OMe methyl ester, thereby avoiding complications with
pyridine−pyridone tautomerism. The methyl esters were then
reduced to alcohols with sodium borohydride in tetrahydrofur-
an (THF) and methanol, followed by conversion to the
respective 2,6-bis(bromomethyl)pyridine species by reaction
with phosphorus tribromide and then 2 equiv of 1-(n-
butyl)imidazole to yield the dibromide proligand salts in high
purity (Figures S1 and S4). Coordination to palladium
proceeded through transmetalation from a silver carbene
intermediate via reaction of the proligand with silver(I) oxide
followed by the addition of PdCl2(cod) and silver triflate,
y i e l d i n g [ P d ( C ^ N ^ C p ‑ B r ) C l ] O T f a n d [ P d -
(C^N^Cp‑OMe)Cl]OTf in good yields. These two species were
further functionalized by halide abstraction with silver triflate in
the presence of acetonitrile, producing the dicationic
acetonitrilo complexes [Pd(C^N^Cp‑Br)(CH3CN)](OTf)2 and
1
All complexes were characterized by H NMR, MS, and
attenuated total reflectance Fourier transform infrared (ATR-
FTIR) spectroscopy, with the IR spectra showing the presence
B
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 1. Synthetic Route to [Pd(C^N^Cp‑Br)Cl]OTf and [Pd(C^N^Cp‑OMe)Cl]OTf
Scheme 2. Synthetic route to [Pd(C^N^Cp‑COOR)Cl]OTf
o f c o o r d i n a t e d a c e t o n i t r i l e f o r [ P d -
( C ^ N ^ C p ‑ B r ) ( C H 3 C N ) ] ( O T f ) 2 a n d [ P d -
(C^N^Cp‑OMe)(CH3CN)](OTf)2. Microanalysis was used to
determine the percentage of H, C, N, and S in the bulk samples
of [Pd(C^N^Cp‑Br)Cl]OTf and [Pd(C^N^Cp‑OMe)Cl]OTf,
matching the calculated values. Insufficient material was
available for microanalysis in the case of [Pd-
(C^N^Cp‑COOR)Cl]OTf. Additionally, the growth of X-ray
quality crystals was attempted using numerous methods and
solvent combinations for [Pd(C^N^Cp‑Br)Cl]OTf and [Pd-
(C^N^Cp‑OMe)Cl]OTf. The complexes repeatedly condensed
as yellow gels, and no X-ray quality crystals were obtained.
The chemical shifts of the pyridyl meta-H atoms are sensitive
probes of the electron density of the pyridyl moiety of the
ligand, providing a strong linear correlation to the Hammett
para-substituent constant σp (Figure 1). The chemical shifts of
cationic chloride complexes vs dicationic acetonitrilo complexes
were also compared for each of R = H, Br, and OMe in
CD3CN, showing a consistent effect in each case where the
imidazol-2-ylidene, pyridyl, and pyridyl-CH2- protons were
shifted downfield by 0.12, 0.07, and 0.05 ppm, respectively, due
to decreased electron density on the metal center. The N-
methylene group of the butyl chain was shifted upfield by 0.27
ppm with the resonance changing from a pair of doublets of
triplets, due to chemical inequivalence and geminal coupling of
the diastereotopic protons (2JH,H = 13−14 Hz) resulting from
the presence of the adjacent chloride and the twisted
conformation imposed by the C^N^C ligand. Finally, it was
observed that in CD2Cl2 the acetonitrilo complexes displayed
broadened signals at room temperature, indicating an increased
rate of interconversion between conformers as compared to
that in CD3CN. All of the chloride complexes examined,
C
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
the pyridyl para-substituted complexes were electrochemically
irreversible with peak potentials determined by square-wave
voltammetry (Figure 2 and Table 1). These data indicate a very
Figure 1. Correlation of Hammett σp constant to chemical shift of
pyridyl m-H atom from 1H NMR spectra of complexes taken in
(CD3)2CO.
regardless of solvent, displayed sharp signals and diastereotopic
splitting indicating significantly slower fluxional processes.25,26
Reactivity of R = Br Species with Acid and Base. An
interesting effect was observed for as-synthesized [Pd-
(C^N^Cp‑Br)Cl]OTf, where a color change from pale yellow
to pink-red occurred upon the addition of a base to the
solution. It should be noted that pyridyl para-bromide groups
can be reactive, warranting further investigation as to the
stability of the product. For example, 4-bromopyridine slowly
self-condenses at room temperature to form ionic oligomers.27
The color change was initially observed with the addition of
potassium carbonate but also occurred with triethylamine and
aqueous sodium hydroxide. No solvent dependence was
observed. The addition of an acid, such as trifluoroacetic acid
or dilute hydrochloric acid, reversed the color change, leaving a
pale yellow solution. The addition of base followed by acid,
with concomitant color changes, could be carried out
repeatedly. This color change in the presence of a base did
not occur with any of the other complexes. Additionally, when
base was added to a solution of the para-Br proligand, no color
change occurred, ruling out the possibility of its cause being a
minor impurity brought forward from previous steps and
suggesting that the metal center may be involved.
Figure 2. Overlaid square-wave voltammograms (25 Hz) of 2 mM
solutions of complexes in 0.1 M [n-Bu4N]PF6/DMF under N2 using a
glassy carbon working electrode.
a
Table 1. Peak Potentials of Complexes versus Ferrocene0/+
Epc1
(V)
Epc2
(V)
Epc3
(V)
Epc4
(V)
complex
σp
[Pd(C^N^Cp‑OMe)Cl]+
−0.27
0.00
−1.91
−1.79
−1.63
−1.50
−2.55
−2.37
−2.11
−2.13
b
[Pd(C^N^Cp‑H)Cl]+
[Pd(C^N^Cp‑Br)Cl]+
[Pd(C^N^Cp‑COOR)Cl]+
0.23
−2.49
−2.32
−2.72
−2.51
0.45
a
Determined from square-wave voltammograms (25 Hz) for 2 mM
solutions in 0.1 M [n-Bu4N]PF6/DMF under N2 using a glassy carbon
b
working electrode. Ref 16.
strong electronic effect of the substituents on the reduction
potential, 5 times stronger than in a bis(amino)phenyl
[Ni(NCN)X] pincer complex where the HOMO and LUMO
span the complex (Figure S14), for example,19 and providing
further evidence of the first reduction potential being pyridyl-
localized.
When plotted against the Hammett σp constants of the
substituents (Figure 3), a strong linear correlation was observed
(Epc1 = 0.581σp − 1.77; R2 = 0.990), quantifying the electronic
sensitivity to the para substituent and providing an empirical
To investigate further, UV−vis spectra of an acetonitrile
solution of [Pd(C^N^Cp‑Br)Cl]OTf were collected before and
after the addition of potassium carbonate, revealing growth of a
broad absorption centered at 508 nm. No isosbestic point was
1
observed. Moreover, MS and H NMR spectroscopy did not
reveal any new signals or changes upon the addition of an
excess of base and then neutralization with an acid. On the basis
of these data, the color change is not due to a reaction with the
p-Br complex itself but is attributed to a trace impurity formed
during the coordination reaction to Pd, likely involving
modification at the para-bromo site.
Electrochemical Behavior under N2 and CO2. The series
of synthesized complexes were electrochemically characterized
by cyclic and square-wave voltammetry in N2 and CO2 sparged
solutions. Previous work has shown that for the unsubstituted
Pd(C^N^C)X+ pincer complexes (where R = H), two
electrochemically irreversible one-electron reduction events
are typically observed. The first electron is suggested to be
localized on the pyridyl moiety, thereby weakening the Pd-
halide bond and enabling its dissociation, and the second
localized on the metal center.16,18 The first reduction events of
Figure 3. Linear correlation between first reduction potential of
Pd(C^N^Cp‑R)Cl+ complexes and Hammett σp substituent constant.
D
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
means of predicting the first reduction potential of complexes
with numerous other substituents, including R = NO2 (σp =
0.78; predicted Epc1 = −1.32 V), R = CF3 (σp = 0.54; predicted
Epc1 = −1.46 V), t-Bu (σp = −0.20; predicted Epc1 = −1.89 V),
and NMe2 (σp = −0.83; predicted Epc1 = −2.25 V). The use of
substituents from one side of the Hammett σp constant
spectrum to the other represents reduction potential tunability
over an expansive range of 1 V. Correlations with the individual
Swain-Lupton constants F (field inductive parameter) and R
(resonance parameter) were considered, with the first reduction
potentials correlating poorly to F (R2 = 0.19) and moderately
to R (R2 = 0.59) for the series.22,28,29
The second and later reduction events were less uniform
across the series. For the R = OMe and H complexes, only one
additional reduction event was observed in the potential range
with equivalent peak current to the first reduction event,
indicating another one-electron transfer. For the R = Br and
COOR complexes, however, at least three additional reduction
events occurred, with smaller currents than the first reduction
event, with the most prominent listed in Table 1. This is
attributed to the varying degrees of ligand dissociation from the
one-electron reduced species. With more electron-donating
(ED) para substituents there is a stronger trans effect and
increased charge on the metal center, thereby labilizing the
chloride and resulting in a fast chemical step after reduction and
full conversion to a new species able to be further reduced.
With electron-withdrawing (EW) para substituents, dissocia-
tion of the monodentate ligand is hindered, resulting in an
incomplete chemical step after reduction, and perhaps ion-
pairing, and consequently multiple species undergoing further
reduction at varying potentials. For the complexes studied here,
the scan rates were varied across the range of 50−1000 mV/s
and did not change the features in the cyclic voltammograms
after the first reduction potential or their relative proportions in
any significant measure.
which has been sufficiently hindered by a weaker electron
donor in the trans position, allowing the reduced species to be
reoxidized before undergoing a chemical reaction. This
emergence of reversibility when using a strong EW substituent
is further evidence of ligand dissociation being enabled upon
reduction.
Under CO2, the complexes displayed differing behaviors
(Figure 5). For the R = OMe and R = H species, significant
current enhancements, suggestive of reactivity with CO2, were
observed at potentials directly corresponding to each of the two
reduction potentials of the complexes under N2, with the
current enhancement at the second reduction potential with R
= OMe especially prominent. In the R = Br case, a current
enhancement was also observed at the first reduction potential
and also at more cathodic potentials, though at different or
shifted potentials when compared to the complex under N2.
Finally, with R = COOR, no current enhancement was
observed at the first reduction potential, and only a weak
enhancement at more cathodic potentials, with peaks different
or shifted relative to the complex under N2.
Under CO2 with 10 mM trifluoroacetic acid (TFA) added,
the current at the first reduction potential more than doubled
for all complexes except for the complex where R = COOR, in
which case only a small increase was observed (Figure 5). For
the species with R = OMe and R = H, the current increase
preceded the first reduction potential by 280 and 200 mV,
respectively, and in the case of R = OMe only, the current
increase is strongly localized at the first reduction potential. It
should be noted that under these conditions TFA, without any
complex present, has a peak potential of approximately −2.25
V, comprising or contributing to the features observed at −2.25
V in Figure 5A−C.16
Finally, the effect of adding the Lewis acid Mg2+ (as
Mg(ClO4)2), which has been shown to affect electrocatalytic
reactivity with CO2,12,16,18,30 was examined by cyclic
voltammetry resulting in no significant changes for complexes
with R = OMe, H, and COOR. For R = Br, a simplification of
the voltammogram was observed, where the second reduction
event became more prominent and the third and fourth
reduction events disappeared (Figure S15), attributed to
assistance with halide abstraction and/or passivation in solution
where ion-pairing may occur with the moderately weak EW
group R = Br.
For each complex, the first reduction event was analyzed by
cyclic voltammetry at scan rates of 50 to 5000 mV/s,
undergoing immediate reoxidation upon reduction and
showing electrochemical irreversibility in all cases except with
the strongest EW substituent, R = COOR, where electro-
chemical reversibility was observed at higher scan rates (Figure
4). This behavior is characteristic of an EC mechanism, which
in this case is attributed to ligand dissociation upon reduction
Density Functional Theory Modeling. Unreduced and
one-electron reduced N-methyl variants of the synthesized
complexes were modeled by density functional theory (DFT)
together with models of complexes with additional para
substituents (such as NMe2, t-Bu, and CF3). The goal of
these studies was to better understand the effects of the para
substituent on various molecular properties including energy
level and geometry of the frontier orbitals, natural population
analysis (NPA) charge on the metal center, Pd−Cl bond length
and dissociation energy, interaction energy of CO2 as it
approaches the reduced species, and energy of the nucleophilic
2
dz orbital of Pd.
Figure 4. Overlaid and normalized cyclic voltammograms of
[Pd(C^N^Cp‑COOR)Cl]OTf at varied scan rates.
E
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 5. Cyclic voltammograms of (A) [Pd(C^N^Cp‑COOR)Cl]OTf, (B) [Pd(C^N^Cp‑Br)Cl]OTf, (C) [Pd(C^N^Cp‑H)Cl]BF4, and (D)
[Pd(C^N^Cp‑OMe)Cl]OTf taken at 100 mV/s under N2, CO2, and CO2 with 10 mM trifluoroacetic acid present.
Figure 6. Calculated Pd−Cl bond length (left) and LUMO energy and Pd NPA charge (right) as a function of Hammett σp constant of para
substituent.
In each case throughout the series, the LUMO was calculated
to be localized primarily on the pyridyl group with some
secondary Pd character. The HOMO geometry was likewise
quite consistent across the series being localized on the NHC
π-orbitals and Pd dxz orbital, with the exception of the complex
with R = NMe2 where the HOMO is pyridyl-localized with
some Pd and carbene character (Figure S16). It was found that
the calculated LUMO energies for the unreduced complexes in
the series exhibited a good linear correlation with the Hammett
σp constant of the substituent (ELUMO = −0.0353σp − 0.0152;
R2 = 0.926), showing the direct electronic effect of the para
substituent upon the LUMO. Similarly, the NPA charge of Pd
also showed a strong linear correlation (QNPA = 0.0098σp +
0.2346; R2 = 0.989), showing the direct influence of the para
substituent on the electron density of the metal center (Figure
6). The Pd−Cl bond length was also tightly correlated with σp
(d(Pd,Cl) = −0.084σp + 2.3306; R2 = 0.985), consistent with the
trans-effect modulating the bond strength of the chloro ligand
according to the varied donor strength of the para-substituted
pyridyl group (Figure 6).
Given the strong correlation between σp and the first
reduction potential (Figure 3), the relationship between the
F
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
calculated LUMO energy and both the Pd NPA charge and the
first reduction potential was analyzed, showing linear relation-
ships with R2 = 0.974 and R2 = 0.972, respectively (Figure 7).
when reduced, and species with less ED substituents retained
their Pd−Cl bonding with varying barriers to dissociation
(Figure S18). Given that ligand dissociation is an important
step in the putative electrocatalytic cycle for CO2 reduction and
is an important factor in understanding reduction events
following the first reduction, additional calculations were
performed to investigate the energy barriers to halide
dissociation where the energy of each of the model complexes,
both unreduced and reduced by one electron, was monitored as
a function of Pd−Cl distance from 2.3 to 3.0 Å.
For the unreduced model species, there was, unsurprisingly,
an increasing energy penalty as the Pd−Cl distance increased to
3 Å, indicating that chloride dissociation will not occur under
typical conditions, though the energy penalty is marginally
smaller with stronger ED substituents (Figure 8). The one-
electron reduced species provided more interesting results,
where the modulation of the Pd−Cl bond strength due to the
varied pyridyl para substituents resulted in no energy barrier for
chloride dissociation for complexes with R = OMe and NMe2, a
small barrier with the R = H and Br species (where dissociation
with the R = H complex is more exogonic), and no dissociation
in the cases when R = COOR and CF3. Consequently, chloride
dissociation for the two-electron reduced species with R =
COOR and CF3 was modeled, showing a comparable energy
barrier to the one-electron reduced species of R = H and Br in
the case of R = COOR and a minimal barrier in the case of R =
CF3 (Figure 8). This is consistent with the behavior of the
synthesized complexes observed by cyclic voltammetry: first,
current enhancements in the presence of CO2 are only
observed for the complexes with R = OMe, H, and Br, but
not with R = COOR, given that loss or labilization of the ligand
in the fourth coordination site is required for reactivity, and
second, the existence of only a single additional reduction event
after the first, and of equal current to the first event, for the R =
OMe and H species indicates complete conversion to a single
reduced species, whereas with the more EW groups R = Br and
COOR, the results are multifarious, consistent with sluggish
and/or incomplete dissociation.
Figure 7. Calculated LUMO energy and Pd NPA charge vs first
reduction potential.
When data from other previously studied complexes were also
included, where R = H but the carbenes are benz- (weaker
overall donor) and phenanthro-annulated (stronger π-donor)
imidazol-2-ylidenes, the correlation between LUMO energy
and first reduction potential remained good (R2 = 0.954), but
the correlation with the NPA charge of Pd weakened
significantly (R2 = 0.107). This indicates that the LUMO
energies can be used in a more general way, with varied NHC
groups and/or para substituents, to predict the first reduction
potential of the complex (Epc1 = −10.51ELUMO − 1.90). The
NPA charges on Pd are strongly influenced by the electron-
donating ability of the NHC moiety which has only a tangential
effect on the LUMO and first reduction potential (Figure S17).
Calculated Chloride Dissociation Energy Barriers.
Strong correlations across the series, as described above for
the unreduced species, were not found for the reduced species
with regard to SOMO energy or NPA charges due to significant
geometry changes where complexes with more ED pyridyl para
substituents (NMe2, OMe, and t-Bu) freely dissociated chloride
Energetics of CO2 Interaction with Reduced Species.
Finally, the interaction of the reduced species with CO2 as it
approaches the metal center was modeled, both in an
Figure 8. Calculated energies of unreduced (left) and reduced (right) complexes as a function of Pd−Cl distance incremented from 2.3 to 3.0 Å
relative to the energy of the optimized structure where the Pd−Cl distance is 2.3 Å, with geometry relaxed at each step. Inset graph retains axes of
parent and shows energies of unreduced species from a Pd−Cl distance of 2.85−2.95 Å.
G
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
acetonitrile solvent field and vacuum, with the energy
monitored as a function of Pd-CO2 distance (Figures 9 and
Table 2. Calculated CO2 Bond Angles at Pd−CO2 Distance
of 2.0 Å
a
model species
in vacuum
in solvent field
R = COOR
R = Br
144.5°
140.1°
139.9°
139.7°
139.5°
141.8°
137.5°
137.3°
137.0°
136.7°
R = H
R = OMe
R = NMe2
a
A conductor-like polarizable continuum model (CPCM) solvent field
of acetonitrile was utilized.
to variable solvation energies, though the trend in activation
energy barrier is consistent (Figure S19). The differences
observed in the energy profiles across the series are attributed
mainly to the energetics of chloride dissociation, described
2
earlier, and to a lesser extent the energy of the dz orbital
(nucleophile) which the coordinating CO2 (electrophile)
interacts with, calculated by natural bond orbital (NBO)
methods (Figure 10).
Figure 9. Calculated energy profiles for approach of CO2 to reduced
complex in vacuum, starting from axial position at 3.25 Å.
S19). In each case the CO2 molecule is brought closer to the
complexes in 0.25 Å increments from a pseudoaxial position
beginning at a distance of 3.25 Å. The interaction is described
below with the solvent field present, though it is very similar in
vacuum. As CO2 is brought closer to the metal center, its
2
interaction with the dz orbital becomes stronger until at
approximately 2.75 Å the carbon atom is aligned in the axial
position above the square planar complex, with nearly equal
bond angles of ∼90° between each pincer donor atom, Pd, and
the C of CO2. At a Pd−CO2 distance of 2.50 Å for the
complexes with more strongly ED substituents (R = NMe2,
OMe, and H), the chloride is displaced away from the CO2
molecule and out of the plane of the pincer ligand, forming a
pseudotrigonal bipyramidal geometry. For the complex with R
= Br, this displacement occurs nearer to 2.25 Å, and for the R =
COOR case, it only occurs when forced to a Pd−CO2 distance
of 1.75 Å. From this point, as the Pd-CO2 distance is decreased
to smaller values, the chloride moves into the opposite axial
position with ∠(pyridyl N, Pd, Cl) ≈ 90°, whereas the trans-
coordinated CO2 has a bond angle of ∠(N, Pd, C) ≈ 105°,
forming an octahedrally distorted trigonal bipyramidal geom-
etry.
The energy profile of CO2 coordination to each of the
complexes in vacuum follows the trend in σp constants for the
pyridyl para substituents, showing a small energy barrier to
coordination of ∼8 kcal/mol and minimal overall energy
change for the complexes with strong EDGs R = NMe2 and
OMe, a larger barrier for the R = H and Br species of ∼13 kcal/
mol (with coordination being more endogonic for R = Br), and
no favorable coordination at all for the R = COOR complex,
with an energy barrier of ∼33 kcal/mol. The degree of CO2
activation was monitored from the bond angle of the
coordinated CO2 at a distance of 2 Å (Table 2). In a vacuum,
the bond angles for R = Br, H, OMe, and NMe2 are similar,
except for the complex with R = COOR, which shows a larger
angle. With a solvent field present, the trend was the same with
an increase in the degree of CO2 activation, due to stabilization
of the negative charge which accumulates on the oxygen atoms.
The interaction energy profiles for the species were more
difficult to directly compare with the solvent field present due
2
Figure 10. Calculated energies of dz orbitals by natural bond orbital
(NBO) methods.
Controlled Potential Electrolysis (CPE) Experiments.
The performance of the synthesized complexes for CO2
reduction was tested by CPE experiments with 2 mM solutions
of the species in 0.1 M [n-Bu4N]PF6/DMF sparged with CO2
and with 10 mM TFA added (pKa ∼ 6 in DMF).31,32 Other
acids have been tested previously, with TFA determined to be a
suitable match for these complexes.16,18 After electrolysis, the
headspace gas from the airtight cell was analyzed by gas
chromatography to quantify the H2 and CO produced. Each
species displayed satisfactory stability throughout the elec-
trolysis experiments: cyclic voltammograms after electrolysis
maintained all of the features of the complex before electrolysis
(Figure S20), no precipitates were formed, and there were no
unanticipated color changes.
The potentials chosen for electrolysis were determined with
reference to the reduction potentials of the complex under CO2
and with TFA added (see Figure 5). With the potential set in a
range where one-electron reduction of the complex occurs, the
percentage of CO in the gaseous products was 26% for the
complex with R = OMe, 8% for R = H,16 10% for R = Br, and
no significant activity was seen for R = COOR near its first
reduction potential (Table 3). Earlier investigations of this
system have verified CO2 as the source for the produced CO
H
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
a
a
Table 3. Results of CPE Experiments
Table 4. Results of CPE Experiments with Mg2+ Added
potential
(V)
charge
FE
FE
H2
ratio CO in
charge
passed
(C)
ratio CO in
produced
gas
species
passed (C)
CO
produced gas
potential
(V)
FE
FE
H2
species
CO
R = OMe
R = H
−2.00 V
−1.75 V
−1.80 V
−2.05 V
−2.50 V
6.0
13
26%
9%
74%
93%
94%
74%
75%
26%
8%
R = OMe + Mg2+
−2.00 V
−1.90 V
−1.80 V
−2.25 V
−2.05 V
6.0
7.0
6.0
7.6
2.5
21%
26%
8%
79%
78%
90%
80%
95%
21%
25%
8%
b
R = H + Mg2+
R = Br
6.0
3.0
3.0
11%
22%
22%
10%
23%
23%
R = Br + Mg2+
R = COOR
R = COOR
R = Br + Mg2+
13%
7%
14%
7%
R = COOR + Mg2+
a
Performed with 2 mM solutions in 10 mL 0.1 M [n-Bu4]PF6/DMF
a
Performed under CO2 with 2 mM solutions in 10 mL of 0.1 M [n-
under CO2, except for the complex with R = COOR where a 3 mL
solution was used. Potentials reported vs ferrocene0/+ and adjusted for
IR drop with a higher surface area electrode, including those previously
reported for R = H.16
Bu4]PF6/DMF with 25 mM Mg(ClO4)2, except for the R = COOR
complex where a 3 mL solution was used. Potentials reported vs
ferrocene0/+ and adjusted for IR drop with higher surface area
electrode, including those previously reported for the R = H species.16
b
through isotope labeling experiments.16 Faradaic efficiencies for
CO and H2 production are also reported, indicating nearly all of
the charge transfer resulting in the formation of those products.
These results show the best performance for CO2 reduction
with an ED para substituent present, consistent with more
facile ligand dissociation, a more electron-rich metal center, and
For the R = H species, 65 mM Mg(ClO4)2 was used.
para substituent. Activity for reduction of CO2 near the first
reduction potential does follow the general trend that with the
most electron-donating para substituent, where facile dissoci-
ation of the trans ligand occurs and the metal center is more
electron rich, more CO is produced. With the most electron-
withdrawing para substituent there is little to no activity,
though there seems to be little difference between the R = H
and R = Br complexes. There is a trade-off between the ease of
ligand dissociation and nucleophilicity of the metal center on
the one hand and the reduction potential (and hence
overpotential) on the other, where the ligand must be labile
enough to easily dissociate upon reduction of the complex, but
after that point a stronger ED para substituent and trans effect
may not be necessary because the metal center is already
sufficiently electron rich to react with CO2. In addition to these
factors, the presence of coordinating anions (conjugate base
from proton source) and water generated during the reaction
will presumably affect competing reduction of H+ to differing
degrees than reduction of CO2, making prediction of overall
performance difficult amidst an interplay of various factors.
2
a higher energy dz orbital on Pd. It should be noted that when
operating at the potentials used in these experiments there is
also direct reduction of H+ at the electrode, without complex
present, accounting for a non-negligible fraction of the H2
produced.16
For the complex with R = COOR, electrolysis was performed
at potentials more cathodic than the first reduction potential,
−2.05 and −2.50 V, where current increases in the presence of
CO2 were observed (see Figure 5A). Equal performance for
CO2 reduction was observed in both cases with 22% CO
detected in the headspace gas, a result comparable to that of R
= OMe operating at −2.00 V. The recorded reduction
potentials and computational data discussed above for energies
of chloride dissociation and interaction with CO2 would suggest
that the R = COOR complex is operating in a regime of two-
electron reduction where ligand dissociation is facile (Figure 8)
2
and the dz orbital is of high energy, intermediate to that of the
CONCLUSIONS
R = OMe and R = NMe2 complexes when reduced by one
electron (Figure 10). These results raise the question of
whether the lutidyl-linked bis-NHC pincer motif is most
effective for reducing CO2 with a strong ED pyridyl para
substituent (e.g., OMe or NMe2) when operating near the first
reduction potential, or with a strong EW pyridyl para
substituent (e.g., COOR or NO2) when operating near, or
more cathodic than, the second reduction potential.
■
The electronic effect of modifying the pyridyl para-position of
lutidyl-linked bis-NHC Pd pincer complexes was investigated,
exhibiting a strong and consistent effect on the first reduction
potential of the complex in relation to the Hammett σp value of
the para substituent, consistent with DFT analysis of the
pyridyl-localized nature of the LUMO for these complexes. The
presence of a sufficiently strong EW group, R = COOR,
resulted in electrochemical reversibility of the first reduction
potential at high scan rates supporting the assignment of an EC
event where ligand dissociation occurs upon reduction. The
effect of the para substituents on electronic properties of the
metal center and ligand dissociation in the reduced species was
quantified by DFT modeling, showing the energy barrier to
dissociation to be strongly dependent on the σp value of the
substituent. Dissociation was found to be spontaneous in the
singly-reduced species with ED groups and significantly
hindered with moderately strong EW groups, having
implications for initial reactivity with CO2 and continued
activity where coordinating anions (conjugate base), water, and
CO are produced during the catalytic cycle. The presence of an
ED group, R = OMe, was found to improve CO2 reduction
relative to R = H when operating in a regime of initial one-
electron reduction of the complex, but the presence of a
moderately strong EW group, R = COOR, also improved CO2
reduction performance relative to R = H when operating in a
The addition of Mg2+ was also tested, as it has been shown to
affect reactivity with CO2 with this class of complexes and
others by acting as a Lewis acid to stabilize the coordinated
CO2 adduct, assisting its activation.12,16,18 Here, the only
observed improvement in CO production was for the R = H
species, reported previously (Table 4).16 Similar but slightly
decreased performance for CO production was observed in the
cases of the R = OMe complex at −2.00 V and the R = Br
complex at −1.80 V, with 21% and 8% CO detected,
respectively, in the gaseous products. For the R = Br species,
electrolysis was also performed at −2.25 V due to a reduction
feature becoming more prominent at that potential with Mg2+
present (Figure S15), resulting in 14% CO in the gaseous
products. Finally, the presence of the Lewis acid Mg2+
diminished performance in the case of the complex with R =
COOR, from 23% to 7% CO in the headspace gas.
In totality, the CPE results give a more intricate picture than
that of molecular properties tuned as a function of the pyridyl
I
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Frequency calculations were performed on all geometry optimized
structures to ensure that energy minima were achieved.
regime of two-electron reduction, and at similar potentials,
raising the question of whether this redox-active pincer
framework is most effective with a strongly ED or strongly
EW para substituent. In either case, the electrochemical
properties of the C^N^C pincer motif have been better
understood, providing an extendable platform where reduction
potentials can be predicted and tuned over a broad range, and
additional functionalities can be incorporated into the pyridyl,
NHC, and/or NHC N-substituents.
Synthesis. Chelidamic acid monohydrate (98%, Alfa Aesar), 1-(n-
butyl)imidazole (99%, Alfa Aesar), dichloro(1,5-cyclooctadiene)-
palladium(II) (99%, Strem), tert-butyllithium (1.7 M in pentane,
Aldrich), iodomethane (99%, copper stabilized, Sigma-Aldrich),
chloromethyl methyl ether (tech. grade, Aldrich), and silver
trifluoromethanesulfonate (99%, Aldrich) were used as received for
synthesis. [Pd(C^N^C)Cl]BF4 was previously synthesized.16
Synthesis of Proligands. 1,1′-((4-Methoxypyridine-2,6-diyl)bis-
(methylene))bis(3-butyl-1H-imidazol-3-ium) Dibromide,
C^N^Cp‑OMe·2HBr. To a solution of 2,6-bis(bromomethyl)-4-methox-
ypyridine (484 mg, 1.64 mmol) in tetrahydrofuran (15 mL) was added
1-(n-butyl)imidazole (0.46 mL, 3.5 mmol). The reaction solution was
stirred at room temperature overnight and then heated to reflux for 3
h, resulting in the formation of a precipitate which was collected by
centrifugation, washed with diethyl ether twice, and dried under a
EXPERIMENTAL SECTION
■
General. Unless otherwise specified, all reactions were performed
under nitrogen using standard Schlenk techniques, and solvents and
reagents were used as received from commercial sources. N,N-
Dimethylformamide (EMD Millipore Omnisolv), distilled acetonitrile
(Anachemia Accusolv), dimethyl sulfoxide (ACS grade, VWR), and
chloroform (HPLC grade, Fisher Chemical) were dried and stored
over 15% m/v 4 Å molecular sieves.33,34 Tetrahydrofuran was distilled
from sodium and benzophenone. Diethyl ether for reaction solvent use
was collected from a solvent purification system. Magnesium
perchlorate (Alfa Aesar) and trifluoroacetic acid (Aldrich) were used
as received for electrochemical experiments.
1
vacuum as a hygroscopic fine white powder (615 mg, 69% yield). H
NMR (300 MHz, (CD3)2SO): δ 0.91 (t, J = 7.4 Hz, 6 H), 1.27 (sxt, J
= 7.4 Hz, 4 H), 1.78 (quin, J = 7.3 Hz, 4 H), 3.88 (s, 3 H), 4.23 (t, J =
7.2 Hz, 4 H), 5.48 (s, 4 H), 7.14 (s, 2 H), 7.74 (t, J = 1.8 Hz, 2 H),
7.84 (t, J = 1.8 Hz, 2 H), 9.35 (s, 2 H). ESI-MS: m/z 462.6 ([M −
Br]+).
1,1′-((4-Bromopyridine-2,6-diyl)bis(methylene))bis(3-butyl-1H-
imidazol-3-ium) Dibromide, C^N^Cp‑Br·2HBr. 1-(n-Butyl)imidazole
(0.25 mL, 1.9 mmol) was added to a solution of 2,6-bis-
(bromomethyl)-4-bromopyridine (292 mg, 0.848 mmol) in tetrahy-
drofuran (10 mL), with the reaction solution stirred at room
temperature overnight and then heated to reflux for 3 h, resulting in
the formation of a precipitate which was collected by centrifugation,
washed with diethyl ether twice, and dried under a vacuum as a
1H NMR spectra were acquired using Bruker AV300 or AV400
spectrometers with chemical shifts referenced to residual solvent
signals. Mass spectra were acquired using a Waters LC-MS ESI-MS,
except for the acetonitrilo complexes which were acquired using a
Waters Micromass LCT ESI-MS. IR spectra were collected using a
PerkinElmer Frontier FT-IR Spectrometer with ATR attachment.
Gaseous products were analyzed using an SRI Model 8610C gas
chromatograph equipped with molecular-sieve columns, dual TCD
and FID detectors, and a methanizer.
1
hygroscopic fine white powder (414 mg, 82% yield). H NMR (400
MHz, CD2Cl2): δ 0.97 (t, J = 7.4 Hz, 6 H), 1.40 (sxt, J = 7.5 Hz, 4 H),
1.90 (quin, J = 7.5 Hz, 4 H), 4.39 (t, J = 7.4 Hz, 4 H), 5.76 (s, 4 H),
7.23 (t, J = 1.5 Hz, 2 H), 7.92 (s, 2 H), 7.97 (t, J = 1.5 Hz, 2 H), 10.79
(s, 2 H). ESI-MS: m/z 512.5 ([M − Br]+).
Electrochemistry. Electrochemical experiments were performed
using a Metrohm Autolab PGSTAT12 or Pine AFCBP1 potentiostat
in an airtight three-electrode cell with a 7 mm2 glassy carbon working
electrode (Bioanalytical Systems, Inc.), Pt mesh counter electrode, and
Ag wire pseudoreference electrode in a 0.010 M AgNO3 acetonitrile
solution separated from the bulk solution by a Vycor frit. Experiments
were performed under N2 or CO2 using 2 mM concentrations of the
complexes in 10 mL of anhydrous electrolyte solution unless otherwise
stated. Controlled potential electrolysis experiments used a glassy
carbon rod (Alfa Aesar) working electrode in a two compartment H-
cell, where the counter electrode compartment was separated from the
compartment containing the working and reference electrodes by
fritted glass. All glassy carbon electrodes were cleaned by successive
polishing with 1 μM, 0.3 μM, and 0.05 μM alumina paste, followed by
rinsing with water, sonication (5 min) in distilled water, and sonication
(5 min) in methanol. Electrolyte solutions were 0.10 M triply
recrystallized [n-Bu4N]PF6 in anhydrous N,N-dimethylformamide and
sparged with nitrogen prior to use. For experiments with CO2, the
solution was sparged with CO2 for 15 min, resulting in a concentration
of ∼0.20 M.35 Decamethylferrocene was added at the end of
electrochemical experiments as an internal standard, showing a
reversible redox couple at −404 mV vs Ag/AgNO3 and −476 mV
vs ferrocene/ferrocenium in 0.1 M [n-Bu4N]PF6/DMF.36 Cyclic
voltammograms were collected at a scan rate of 100 mV/s and square
wave voltammograms at a frequency of 25 Hz unless otherwise stated.
Computational Methods. DFT calculations were performed
using Gaussian 09 (Revision D.01) using the long-range and
dispersion-corrected ωB97xD hybrid functional without symmetry
constraints.37 Calculations for reduced species were performed using
unrestricted, open-shell wave functions. The D95(d) basis set was used
for all atoms except palladium, which employed the Stuttgart−
Dresden−Bonn quasi-relativistic effective-core potential and corre-
sponding correlation-consistent triple-ζ basis set, and Br, which used
the 6-31G(d) basis set.38,39 Calculations were performed with the
presence of a solvent reaction field produced by the conductor-like
polarizable continuum model (CPCM) unless otherwise stated.40
1,1′-((4-((4-Hydroxybutoxy)carbonyl)pyridine-2,6-diyl)bis-
(methylene))bis(3-butyl-1H-imidazol-3-ium) Dibromide,
C^N^Cp‑COOR·2HBr. 1-(n-Butyl)imidazole (70 μL, 0.53 mmol) was
added to a solution of 4-hydroxybutyl 2,6-bis(bromomethyl)-
isonicotinate (84 mg, 0.22 mmol) in tetrahydrofuran (5 mL) and
left to stir at room temperature overnight before being heated to reflux
for 4 h, resulting in the deposition of a viscous, pale yellow oil on the
glass surface. The solvent was removed by rotary evaporation with the
residue washed with pentane and then taken up in dichloromethane (2
mL), filtered, and dropped into pentane (3 mL) resulting in the
formation of a white precipitate which conglomerated as a viscous oil.
The oily residue was triturated with anhydrous diethyl ether and then
dried under a vacuum, resulting in bubbling and solidification of the
1
residue to yield a hygroscopic white powder (126 mg, 91% yield). H
NMR (400 MHz, (CD3)2SO): δ 0.91 (t, J = 7.3 Hz, 6 H), 1.27 (sxt, J
= 7.4 Hz, 4 H), 1.78 (quin, J = 7.3 Hz, 4 H), 1.84−1.91 (m, 2 H),
1.91−2.02 (m, 2 H), 3.62 (t, J = 6.5 Hz, 2 H), 4.20 (t, J = 7.2 Hz, 4 H),
4.35−4.44 (m, 2 H), 5.63 (s, 4 H), 7.70 (t, J = 1.7 Hz, 2 H), 7.82 (t, J
= 1.7 Hz, 2 H), 7.97 (s, 2 H), 9.22 (t, J = 1.7 Hz, 2 H). ESI-MS: m/z
612.1 ([M − OH]+), 488.0 ([M − (butylimidazole) − OH + H]+),
408.1 ([M-(butylimidazole) − OH − Br]+).
Synthesis of Palladium Complexes. [Pd(C^N^Cp‑OMe)Cl]OTf.
To a solution of proligand C^N^Cp‑OMe·2HBr (250 mg, 0.460 mmol)
in dimethyl sulfoxide (8 mL) was added silver(I) oxide (106 mg, 0.460
mmol) and 4 Å molecular sieves. The reaction mixture was heated to
55 °C, covered in foil, and left to gently stir for 24 h, resulting in the
formation of a light brown solid. The mixture was cooled to room
temperature before the addition of silver trifluoromethanesulfonate
(125 mg, 0.486 mmol) and then dichloro(1,5-cyclooctadiene)-
palladium(II) (131 mg, 0.460 mmol), followed by stirring at 30 °C
for 48 h. After centrifugation, the yellow supernatant solution was
concentrated by vacuum distillation at 55 °C. The oily residue was
taken up in dichloromethane (2 mL), filtered through Celite, and then
dropped into diethyl ether (12 mL), yielding a pale yellow precipitate
J
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
which conglomerated upon centrifugation to form a golden yellow oil.
The oil was triturated with diethyl ether, then taken up with
dichloromethane (2 mL) and washed with water (4 mL × 2), dried
with MgSO4, filtered, and concentrated to dryness under a vacuum,
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b02213.
■
S
1
yielding a pale yellow powder (182 mg, 59% yield). H NMR (400
MHz, CD3CN): δ 0.93 (t, J = 7.4 Hz, 6 H), 1.31 (sxt, J = 7.4 Hz, 4 H),
1.83 (quin, J = 7.4 Hz, 4 H), 3.95 (s, 3 H), 4.18 (dt, J = 13.3, 7.1 Hz, 4
H), 4.69 (dt, J = 13.3, 7.6 Hz, 2 H), 5.23 (d, J = 15.1 Hz, 2 H), 5.51 (d,
J = 15.1 Hz, 2 H), 7.09 (d, J = 1.7 Hz, 2 H), 7.22 (s, 2 H), 7.28 (d, J =
1.7 Hz, 2 H). ESI-MS: m/z 524.5 ([M − OTf]+). Anal. Calcd for
C23H31ClF3N5O4PdS·H2O: C, 40.01; H, 4.82; N, 10.14; S, 4.64.
Found: C, 40.05; H, 4.69; N, 9.96; S, 4.57.
Synthetic procedures for proligand precursors and
1
acetonitrilo complexes, H NMR spectra of proligand
precursors, proligands, and complexes, ATR-FTIR
spectra of complexes, additional cyclic voltammograms,
LUMO geometry diagrams, plots of ip vs ν1/2 for
complexes, TOF and overpotential calculations, DFT
calculation sample input, and atomic coordinates of
geometry-optimized species (PDF)
[Pd(C^N^Cp‑Br)Cl]OTf. To a solution of proligand C^N^Cp‑Br·2HBr
(251 mg, 0.422 mmol) in dimethyl sulfoxide (7 mL) was added
silver(I) oxide (100 mg, 0.432 mmol) and 4 Å molecular sieves. The
reaction mixture was heated to 55 °C, covered in foil, and left to gently
stir for 24 h, resulting in the formation of a light brown solid. The
mixture was cooled to room temperature before subsequent additions
of silver trifluoromethanesulfonate (114 mg, 0.443 mmol) and
dichloro(1,5-cyclooctadiene)palladium(II) (121 mg, 0.422 mmol),
followed by stirring at 30 °C for 48 h. After centrifugation, the red
supernatant solution was concentrated by vacuum distillation at 55 °C,
leaving a wet solid. The residue was taken up in dichloromethane (4
mL), washed with water (4 mL × 2), dried with magnesium sulfate,
filtered, and the solvent removed, yielding an orange solid. The crude
solid was dissolved in dichloromethane (1 mL) and dropped into
diethyl ether (10 mL), forming a light precipitate which was collected
by centrifugation as a sticky orange solid. Further drying under a
vacuum at 60 °C for 24 h resulted in a yellow-orange powder (218 mg,
72% yield). 1H NMR (400 MHz, CD3CN): δ 0.93 (t, J = 7.3 Hz, 6 H),
1.31 (sxt, J = 7.4 Hz, 4 H), 1.83 (quin, J = 7.5 Hz, 4 H), 4.18 (dt, J =
13.5, 7.1 Hz, 2 H), 4.67 (dt, J = 13.5, 7.7 Hz, 2 H), 5.29 (d, J = 15.2
Hz, 2 H), 5.54 (d, J = 15.2 Hz, 2 H), 7.10 (d, J = 1.7 Hz, 2 H), 7.29 (d,
J = 1.9 Hz, 2 H), 7.92 (s, 2 H). ESI-MS: m/z 571.9 ([M-OTf]+). Anal.
Calcd for C22H28BrClF3N5O3PdS: C, 36.63; H, 3.91; N, 9.71; S, 4.44.
Found: C, 36.89; H, 4.10; N, 9.45; S, 4.06.
AUTHOR INFORMATION
Corresponding Author
*Phone: (604) 822-1702. Fax: (604) 822-2847. E-mail:
mwolf@chem.ubc.ca.
ORCID
Michael O. Wolf: 0000-0003-3076-790X
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
■
This research was supported by the Natural Sciences and
Engineering Research Council of Canada (NSERC). WestGrid
(www.westgrid.ca) and Compute Canada−Calcul Canada
(www.computecanada.ca) are acknowledged for access to
computational resources for DFT calculations. Prof. David
Wilkinson is thanked for use of the gas chromatograph. J.A.T.
thanks NSERC for scholarship funding.
[Pd(C^N^Cp‑COOR)Cl]OTf. To a solution of proligand C^N^Cp‑COOR
·
REFERENCES
2HBr (117 mg, 0.186 mmol) in dimethyl sulfoxide (5 mL) was added
silver(I) oxide (48 mg, 0.21 mmol) and 4 Å molecular sieves. The
reaction mixture was heated to 55 °C, covered in foil, and left to gently
stir for 36 h. After cooling to room temperature, silver
trifluoromethanesulfonate (52 mg, 0.202 mmol) and dichloro(1,5-
cyclooctadiene)palladium(II) (53.0 mg, 0.186 mmol) were added,
followed by stirring at 30 °C in darkness for 72 h. Afterward, the
mixture was centrifuged and the solvent removed from the supernatant
solution by vacuum distillation at 55 °C. The residue was taken up in
dichloromethane (1.5 mL) and washed with water (3 mL × 2), then
dried with magnesium sulfate, filtered, and dropped into diethyl ether,
resulting in the formation of an off-white precipitate which
conglomerated to form a viscous oil. The oily residue was washed
and triturated with diethyl ether then dried under a vacuum, leading to
bubbling and solidification. A pale yellow solid was obtained (84 mg)
which required further purification by flash chromatography (SiO2,
80:12:8 CH3CN/H2O/KNO3(aq)). The first fraction collected was
purified through another column (SiO2, 90:6:4 CH3CN/H2O/
KNO3(aq)), the solvent removed under a vacuum at 50 °C, and the
residue taken up in acetone (2 mL) and then saturated with potassium
trifluoromethanesulfonate. After stirring at room temperature for 1 h,
the solvent was removed by rotary evaporation and the residue
extracted with dichloromethane, which was then washed with water,
dried with magnesium sulfate, filtered, and concentrated to dryness
under a vacuum. A pale yellow powder was obtained (4.0 mg, 3%
■
(1) Inglis, J. L.; MacLean, B. J.; Pryce, M. T.; Vos, J. G.
Electrocatalytic Pathways towards Sustainable Fuel Production from
Water and CO2. Coord. Chem. Rev. 2012, 256, 2571−2600.
(2) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois,
D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; et al.
Frontiers, Opportunities, and Challenges in Biochemical and Chemical
Catalysis of CO 2 Fixation. Chem. Rev. 2013, 113, 6621−6658.
(3) Mikkelsen, M.; Jørgensen, M.; Krebs, F. C. The Teraton
Challenge. A Review of Fixation and Transformation of Carbon
Dioxide. Energy Environ. Sci. 2010, 3, 43.
(4) DuBois, D. L. Development of Molecular Electrocatalysts for
Energy Storage. Inorg. Chem. 2014, 53, 3935−3960.
(5) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M.
Electrocatalytic and Homogeneous Approaches to Conversion of CO2
to Liquid Fuels. Chem. Soc. Rev. 2009, 38, 89−99.
(6) Qiao, J.; Liu, Y.; Hong, F.; Zhang, J. A Review of Catalysts for the
Electroreduction of Carbon Dioxide to Produce Low-Carbon Fuels.
Chem. Soc. Rev. 2014, 43, 631−675.
(7) Leckel, D. Diesel Production from Fischer−Tropsch: The Past,
the Present, and New Concepts. Energy Fuels 2009, 23, 2342−2358.
(8) Pegis, M. L.; Roberts, J. A. S.; Wasylenko, D. J.; Mader, E. A.;
Appel, A. M.; Mayer, J. M. Standard Reduction Potentials for Oxygen
and Carbon Dioxide Couples in Acetonitrile and N,N-Dimethylfor-
mamide. Inorg. Chem. 2015, 54, 11883−11888.
1
yield). H NMR (400 MHz, (CD3)2CO): δ 0.93 (t, J = 7.5 Hz, 6 H),
(9) Costentin, C.; Drouet, S.; Robert, M.; Saveant, J.-M. A Local
Proton Source Enhances CO2 Electroreduction to CO by a Molecular
Fe Catalyst. Science 2012, 338, 90−94.
(10) Thoi, V. S.; Kornienko, N.; Margarit, C. G.; Yang, P.; Chang, C.
J. Visible-Light Photoredox Catalysis: Selective Reduction of Carbon
Dioxide to Carbon Monoxide by a Nickel N-Heterocyclic Carbene-
Isoquinoline Complex. J. Am. Chem. Soc. 2013, 135, 14413−14424.
1.30−1.39 (m, 4 H), 1.83−1.92 (m, 4 H), 1.92−1.98 (m, 2 H), 2.03−
2.09 (m, 2 H), 3.68 (t, J = 6.1 Hz, 2 H), 4.31 (dt, J = 14.2, 7.5 Hz, 2
H), 4.42 (s, 1 H), 4.46 (t, J = 6.0 Hz, 2 H), 4.70 (dt, J = 13.6, 7.8 Hz, 2
H), 5.85 (d, J = 15.2 Hz, 2 H), 5.96 (d, J = 15.2 Hz, 2 H), 7.38 (d, J =
1.7 Hz, 2 H), 7.61 (d, J = 1.7 Hz, 2 H), 8.42 (s, 2 H). ESI-MS: m/z
594.1 ([M − OH]+).
K
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(11) Hawecker, J.; Lehn, J.; Ziessel, R. Efficient Photochemical
Reduction of CO 2 to CO by Visible Light Irradiation of Systems
Containing Re(bipy) (CO) 3 X or Ru(bipy) 3 2+ − Co 2+
Combinations as Homogeneous Catalysts. J. Chem. Soc., Chem.
Commun. 1983, 9, 536−538.
(12) Sampson, M. D.; Kubiak, C. P. Manganese Electrocatalysts with
Bulky Bipyridine Ligands: Utilizing Lewis Acids to Promote Carbon
Dioxide Reduction at Low Overpotentials. J. Am. Chem. Soc. 2016,
138, 1386−1393.
(13) Stanton, C. J.; Machan, C. W.; Vandezande, J. E.; Jin, T.;
Majetich, G. F.; Schaefer, H. F.; Kubiak, C. P.; Li, G.; Agarwal, J. Re(I)
NHC Complexes for Electrocatalytic Conversion of CO2. Inorg. Chem.
2016, 55, 3136−3144.
(14) DuBois, D. L.; Miedaner, A. Mediated Electrochemical
Reduction of CO2. Preparation and Comparison of an Isoelectronic
Series of Complexes. J. Am. Chem. Soc. 1987, 109, 113−117.
(15) DuBois, D. L.; Miedaner, A.; Haltiwanger, R. C. Electrochemical
Reduction of Carbon Dioxide Catalyzed by [Pd(triphosphine)
(solvent)](BF4)2 Complexes: Synthetic and Mechanistic Studies. J.
Am. Chem. Soc. 1991, 113, 8753−8764.
(16) Therrien, J. A.; Wolf, M. O.; Patrick, B. O. Electrocatalytic
Reduction of CO 2 with Palladium Bis-N-Heterocyclic Carbene Pincer
Complexes. Inorg. Chem. 2014, 53, 12962−12972.
(17) Steffey, B. D.; Curtis, C. J.; DuBois, D. L. Electrochemical
Reduction of CO2 Catalyzed by a Dinuclear Palladium Complex
Containing a Bridging Hexaphosphine Ligand: Evidence for
Cooperativity. Organometallics 1995, 14, 4937−4943.
(18) Therrien, J. A.; Wolf, M. O.; Patrick, B. O. Polyannulated Bis(N-
Heterocyclic Carbene)palladium Pincer Complexes for Electrocatalytic
CO 2 Reduction. Inorg. Chem. 2015, 54, 11721−11732.
(19) van de Kuil, L. A.; Luitjes, H.; Grove, D. M.; Zwikker, J. W.; van
der Linden, J. G. M.; Roelofsen, A. M.; Jenneskens, L. W.; Drenth, W.;
van Koten, G. Electronic Tuning of arylnickel(II) Complexes by Para
Substitution of the Terdentate Monoanionic 2,6-Bis[(dimethylamino)-
methyl]phenyl Ligand. Organometallics 1994, 13, 468−477.
(20) Slagt, M. Q.; Rodríguez, G.; Grutters, M. M. P.; Klein Gebbink,
R. J. M.; Klopper, W.; Jenneskens, L. W.; Lutz, M.; Spek, A. L.; van
Koten, G. Synthesis and Properties Ofpara-Substituted NCN-Pincer
Palladium and Platinum Complexes. Chem. - Eur. J. 2004, 10, 1331−
1344.
(21) Darmon, J. M.; Turner, Z. R.; Lobkovsky, E.; Chirik, P. J.
Electronic Effects in 4-Substituted Bis(imino)pyridines and the
Corresponding Reduced Iron Compounds. Organometallics 2012, 31,
2275−2285.
(22) Hansch, C.; Leo, A.; Taft, R. W. A Survey of Hammett
Substituent Constants and Resonance and Field Parameters. Chem.
Rev. 1991, 91, 165−195.
Field and Resonance Components. J. Am. Chem. Soc. 1983, 105, 492−
502.
(29) Swain, C. G.; Lupton, E. C., Jr. Field and Resonance
Components of Substituent Effects. J. Am. Chem. Soc. 1968, 90,
4328−4337.
́
(30) Bhugun, I.; Lexa, D.; Saveant, J.-M. Catalysis of the
Electrochemical Reduction of Carbon Dioxide by Iron (0) Porphyrins.
Synergistic Effect of Lewis Acid Cations. J. Phys. Chem. 1996, 100,
19981−19985.
(31) Roy, K.; Popelier, P. L. A. Predictive QSPR Modeling of the
Acidic Dissociation Constant (Pk a) of Phenols in Different Solvents.
J. Phys. Org. Chem. 2009, 22, 186−196.
(32) Maran, F.; Celadon, D.; Severin, M. G.; Vianello, E.
Electrochemical Determination of the pKa of Weak Acids in N,N-
Dimethylformamide. J. Am. Chem. Soc. 1991, 113, 9320−9329.
(33) Burfield, D. R.; Smithers, R. H. Desiccant Efficiency in Solvent
Drying. 3. Dipolar Aprotic Solvents. J. Org. Chem. 1978, 43, 3966−
3968.
(34) Williams, D. B. G.; Lawton, M. Drying of Organic Solvents:
Quantitative Evaluation of the Efficiency of Several Desiccants. J. Org.
Chem. 2010, 75, 8351−8354.
(35) Gennaro, A.; Isse, A. A.; Vianello, E. Solubility and
Electrochemical Determination of CO2 in Some Dipolar Aprotic
Solvents. J. Electroanal. Chem. Interfacial Electrochem. 1990, 289, 203−
215.
(36) Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay,
P. A.; Masters, A. F.; Phillips, L. The Decamethylferrocenium/
Decamethylferrocene Redox Couple: A Superior Redox Standard to
the Ferrocenium/Ferrocene Redox Couple for Studying Solvent
Effects on the Thermodynamics of Electron Transfer. J. Phys. Chem. B
1999, 103, 6713−6722.
(37) Chai, J.-D.; Head-Gordon, M. Long-Range Corrected Hybrid
Density Functionals with Damped Atom-Atom Dispersion Correc-
tions. Phys. Chem. Chem. Phys. 2008, 10, 6615−6620.
(38) Martin, J. M. L.; Sundermann, A. Correlation Consistent
Valence Basis Sets for Use with the Stuttgart−Dresden−Bonn
Relativistic Effective Core Potentials: The Atoms Ga−Kr and In−
Xe. J. Chem. Phys. 2001, 114, 3408.
(39) Andrae, D.; Haußermann, U.; Dolg, M.; Stoll, H.; Preuß, H.
̈
Energy-Adjusted Ab Initio Pseudopotentials for the Second and Third
Row Transition Elements. Theor. Chim. Acta 1990, 77, 123−141.
(40) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. Energies,
Structures, and Electronic Properties of Molecules in Solution with the
C-PCM Solvation Model. J. Comput. Chem. 2003, 24, 669−681.
(23) Urgoitia, G.; SanMartin, R.; Herrero, M. T.; Domínguez, E.
Palladium NCN and CNC Pincer Complexes as Exceptionally Active
Catalysts for Aerobic Oxidation in Sustainable Media. Green Chem.
2011, 13, 2161.
(24) Yu, R. P.; Darmon, J. M.; Milsmann, C.; Margulieux, G. W.;
Stieber, S. C. E.; DeBeer, S.; Chirik, P. J. Catalytic Hydrogenation
Activity and Electronic Structure Determination of Bis(arylimidazol-2-
Ylidene)pyridine Cobalt Alkyl and Hydride Complexes. J. Am. Chem.
Soc. 2013, 135, 13168−13184.
́
(25) Miecznikowski, J. R.; Grundemann, S.; Albrecht, M.; Megret, C.;
̈
Clot, E.; Faller, J. W.; Eisenstein, O.; Crabtree, R. H. Outer Sphere
Anion Participation Can Modify the Mechanism for Conformer
Interconversion in Pd Pincer Complexes. Dalton Trans. 2003, 5, 831−
838.
(26) Hahn, F. E.; Jahnke, M. C.; Gomez-Benitez, V.; Morales-
Morales, D.; Pape, T. Synthesis and Catalytic Activity of Pincer-Type
Bis(benzimidazolin-2-Ylidene) Palladium Complexes. Organometallics
2005, 24, 6458−6463.
(27) Feast, W. J.; Tsibouklis, J. The Self-Condensation of 4-
Bromopyridine. Polym. Int. 1994, 35, 67−74.
(28) Swain, C. G.; Unger, S. H.; Rosenquist, N. R.; Swain, M. S.
Substituent Effects on Chemical Reactivity. Improved Evaluation of
L
DOI: 10.1021/acs.inorgchem.6b02213
Inorg. Chem. XXXX, XXX, XXX−XXX