Inorganic Chemistry
Article
this charged pyridinium undergoes two stepwise reductions,
the first of which is proton-independent (Figure S5). CPE of
Fe-TPP with 1 equiv of 5 shows a similarly high rate of charge
consumption and current density as observed with additives
1−3 (Figures 5 and S22). However, after this initial increase,
the catalytic current decreases dramatically, and the cumulative
CO production is comparable to that of Fe-TPP alone, which
we attribute to the formation of radical anions that dimerize or
decompose catalyst. Figure 6 further shows that the log-
(TOF)−overpotential relationship for Fe-TPP is not signifi-
cantly changed upon addition of 4 or 5.
Additional synthetic NADH analogues were then inves-
tigated as additives to probe contributions of second-sphere
electron and proton relays to enhanced CO2 catalysis. CPE
experiments were performed using benzanilide 6 as a redox-
inactive additive that retains the amide substituent. Previous
work from our laboratory demonstrated that intramolecular
amide pendants can significantly enhance the reduction of CO2
to CO with proper second-sphere positioning.29 In agreement
with that study, the CPE experiments performed with Fe-TPP
and 1 equiv of 6 lead to improved rates of CO production (5.9-
fold increase) but not to the levels achieved with additives 1−3
seen in Figure 6 that possess both electron and proton transfer
capabilities (9.5- to 13.4-fold rate increases). Indeed, CVs of 6
in the presence of phenol do not show the ECEC behavior
observed with additives 1−3. Instead, a catalytic wave is
observed with a peak area that more than triples with
increasing phenol concentration, indicating that additive 6
behaves solely as a proton source through the amide NH
moiety (Figure S6). Finally, we tested pyridine amide additive
7, where the amide nitrogen atom is methylated to remove
benefits of hydrogen bonding and/or proton relays but retains
the redox-active reservoir. As observed with proton-only 6,
electron-only additive 7 increases charge consumption and
rates of CO production compared to Fe-TPP alone (8.4-fold)
but not to the same extent as NADH analogues 1−3 that have
the capacity for both electron and proton management
(Figures 6 and Figure S24). Interestingly, the enhanced
catalysis with additive 1 is not simply a summation of the
benefits provided by hydrogen bonding or proton donation
with 6 and by electron transfer with 7, suggesting a more
complex interplay. Nevertheless, these data suggest the
participation of both electron and proton relay components
in enhancing electrochemical CO2 reduction catalysis, as
NADH analogues that possess dual electron−proton redox
reservoir capabilities are superior to additives that only have
either electron or proton transfer capacity.
Finally, as NADH analogue 2 showed the highest activity
enhancements for electrochemical CO2 reduction, we
performed further dose-dependent CPE experiments with
this additive. Figure 7 compares relative current density and
charge accumulation values during 1 h of electrolysis at −2.4 V
with Fe-TPP under a CO2 atmosphere and 0, 1, 5, 10, or 15
equiv of additive 2. Addition of 1 equiv of 2 leads to a
significant enhancement in average current density to 4.19
mA/cm2, compared to 1.64 mA/cm2 with Fe-TPP alone.
However, increasing doses of 2 to 5, 10, or 15 equiv provides
only a modest further current density enhancement. We
speculate that this saturation behavior could be due to inherent
rate limitations of the parent Fe-TPP catalysis and/or
insufficient electrolysis to reduce enough 2 to couple to Fe-
TPP to be effective in the catalytic cycle, as the two
components must interact intermolecularly. Further experi-
Figure 7. CPEs of 0.5 mM Fe-TPP in the presence of 0−15 mM
additive 2 in 0.1 M TBAPF6 DMF electrolyte containing 500 mM
phenol under a CO2 atmosphere plotting (A) current density and (B)
charge accumulation over a period of 1 h at −2.4 V vs Fc/Fc+.
ments beyond the scope of this work are required to elucidate
such complex mechanistic details.
CONCLUDING REMARKS
■
In summary, we have presented a strategy for enhancing
electrochemical CO2 reduction catalysis through the use of
second-sphere additives that have dual electron−proton
reservoir capabilities. We establish this concept using synthetic
NADH analogues as electron−proton mediators with the
molecular CO2 reduction catalyst Fe-TPP. Indeed, although
extensive efforts have provided valuable design strategies for
facilitating second-sphere proton transfer for electrochemical
CO2 reduction, our findings show that such electrocatalysts
can be further augmented by adding electron transfer relays as
well. Specifically, the NADH-inspired nicotinamide-based
additives 1−3 undergo an ECEC reduction pathway to
subsequently serve as intermolecular electron−proton sources
for CO2 reduction catalyzed by Fe-TPP. Dihydropyridine
intermediates, which can be generated by chemical or
electrochemical means, can be reoxidized in air thermally or
through anodic electrochemistry back to the starting
nicotinamide to complete the electrochemical cycle.
A key design feature that emerges from this study is to
precisely match the redox potentials of the primary-sphere
molecular catalyst and second-sphere electron−proton medi-
ator to promote productive multielectron, multiproton
chemistry. For the two-electron/two-proton conversion of
CO2 to CO, catalytic rates for reduction can be augmented by
up to 13.4-fold in the presence of the electron−proton
mediator without sacrificing the high selectivity of Fe-TPP for
CO2 versus proton reduction. The cooperative second-sphere
approach complements work on CO2 hydrogenation catalysts,
where metal hydrides simultaneously deliver electrons and
protons.59,83−85 Indeed, few organic hydrides are strong
enough donors to reduce CO2 directly,73,76 and those that
can be regenerated encounter slow rates of hydride transfer
that limit integration into photo- or electrocatalytic cycles.75,86
This cooperative, bioinspired approach to electrocatalysis,
which leverages primary metal centers with secondary dual
electron−proton reservoirs, should be applicable to a broader
array of chemical transformations where bond activation and
catalytic turnover rely on controlling both electron and proton
inventories.
F
Inorg. Chem. XXXX, XXX, XXX−XXX