Electrocatalytic reduction of CO2to CO and HCO2?with Zn(ii) complexes displaying cooperative ligand reduction

Article abstract of DOI:10.1039/d1cc03887g

Air-stable zinc(ii) pyridyl phosphine complexes, [Zn(κ²-2,6-{Ph₂PNMe}₂(NC₅H₃))Br₂] (1) and [Zn(κ²-2-{Ph₂PNMe}(NC₅H₃))Br₂] (2) are repo

Full text of DOI:10.1039/d1cc03887g

ChemComm
View Article Online
View Journal
COMMUNICATION
Electrocatalytic reduction of CO2 to CO and
ꢀ
HCO2 with Zn(II) complexes displaying
Cite this: DOI: 10.1039/d1cc03887g
cooperative ligand reduction†
Received 18th July 2021,
Accepted 13th August 2021
Patrick Berro,a Somayeh Norouziyanlakvan,a Gyaneshwar Kumar Rao,b
a
Bulat Gabidullina and Darrin Richeson
*
DOI: 10.1039/d1cc03887g
rsc.li/chemcomm
Air-stable zinc(II) pyridyl phosphine complexes, [Zn(j2-2,6-
{Ph2PNMe}2(NC5H3))Br2] (1) and [Zn(j2-2-{Ph2PNMe}(NC5H3))Br2]
(2) are reported and 1 was capable of electrocatalytic reduction
of CO2 at ꢀ2.3 V vs. Fc+/0 to yield CO/HCO2H in mixed water/
acetonitrile solutions. DFT computations support
a proposed
mechanism involving electron transfer reactions from a species
with the anionic PN3P ligand (‘‘Lꢀ/Zn(II)’’).
Any effort to balance the carbon cycle will rely on the
reduction of CO2, the pervasive by-product from combustion
with a well-established damaging impact on the environment.
However, Zn has documented roles in catalyzed reactions for
Conversion of CO2 to yield chemical feedstocks, either elec- capture and insertion of CO27–11 and ZnBr2 has been shown to
trochemically or photochemically, is a tantalizing goal and accelerate the reaction of PEt3/CH2I2 with CO2 to form CO.12
a
broad assortment of homogeneous transition-metal
Among molecular Zn complexes, A–D stand out as unique
complexes (e.g., Mn-, Fe-, Co-, Ni-, Cu-, Ru-, and Re-based catalysts for hydrogen generation and/or CO2 reduction. Tris
complexes), have been explored for electro- and photocatalytic (2-pyridylthio)methylzinc hydride, [k3-Tptm]ZnH A, catalytically
reduction of CO2.1–6 From a sustainability perspective, a generated H2 from triphenylsilane and performed catalytic hydro-
critical feature of this process is the use of catalysts that rely silylation of CO2.13 The square planar Zn(II) thiosemicarboazone
on earth-abundant elements.
complex B (‘‘Zn(DMTH)’’) can convert CO2 to formate via a methox-
Among the 3d metals, perhaps not surprisingly, Zn is nearly ide complex, Zn(HDMTH)(OCH3), which can insert CO2 yielding a
absent from the metals that have been used for CO2 reduction. methylcarbonate species. This carbonate can be subsequently
ꢀ
The key obstacle to using Zn is that it is found almost reduced to HCO2 by chemical or electrochemically generated
exclusively in the 2+ oxidation state and this redox innocence hydride sources.14 The electrochemical proton reduction to hydride
blocks the ability to undergo the fundamental oxidative addi- is key to this transformation. In fact, complexes B and C can
tion and reductive elimination steps typical of homogeneous generate H2 electrocatalytically from acetic acid through ligand-
catalyzed reactions.
centered pathways.15,16 In acetonitrile, these complexes undergo
ligand protonation by acetic acid to yield active catalysts, which
displayed catalytic reduction currents at ꢀ2.3 V and ꢀ3.29 V vs. Fc+/0
for B and C, respectively. Aside from B, complex D represents the
only other reported homogeneous Zn-centered catalyst for electro-
catalytic CO2 reduction. Cyclic voltammetry of this pyridylphosphine
ligated Zn complex showed an irreversible one-electron reduction at
ꢀ2.03 V vs. Fc+/0 and current enhancement, indicating electrocata-
lysis, of this reduction was observed under a CO2 atmosphere.17
When bulk electrolysis was carried out with D under a CO2 atmo-
sphere, CO formation was observed by mass spectrometry and IR
spectroscopy. However, no quantification of this reduction
a Department of Chemistry and Biomolecular Sciences Centre for Catalysis Research
and Innovation, University of Ottawa, 10 Marie Curie, Ottawa, ON K1N 6N5,
Canada. E-mail: darrin@uottawa.ca
b Department of Chemistry, Amity School of Applied Sciences, Amity University,
Haryana-122413, India
† Electronic supplementary information (ESI) available: Additional experimental,
crystallographic and computational details. CCDC 2097092 and 2097093. For ESI
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
d1cc03887g
This journal is © The Royal Society of Chemistry 2021
Chem. Commun.

View Article Online
Communication
ChemComm
product was reported. A partial mechanism employing the non- at d 27.7 ppm, an upfield shift from the free ligand at d 51.0 ppm.
coordinated phosphorus center to activate the CO2 by nucleophilic The crystals were of sufficient quality to carry out a single crystal
attack was proposed.
X-ray analysis and the results are represented in Fig. 1 (Fig. S2 and
Our investigation of the prospective of Zn complexes for Table S3, ESI†). Complex 2 crystallized with two molecules of
electrocatalytic CO2 reduction began with preparation of formula [Zn(2-{Ph2PNMe}(NC5H3))Br2] in the asymmetric unit and
compounds 1 and 2. The direct reaction of ZnBr2 with a the two displayed very similar bonding parameters. One of these
stoichiometric quantities of the neutral pincer ligand, molecules is displayed in Fig. 1, which confirms that complex 2
N,N0-bis(diphenylphosphino)-2,6-di(methylamino)pyridine possessed a four coordinate Zn(II) center bonded in a bidentate
2,6-{Ph2PNMe}2(NC5H3) (‘‘PN3P’’) yielded the new complex 1 as fashion by the diphenylaminopyridine (‘‘PN’’) ligand and two bro-
a colorless precipitate that was purified by crystallization from mide anions originating from the starting material. Overall, this
CH2Cl2. Complex 1 displayed 1H, 13C and 31P NMR spectra yields distorted tetrahedral Zn(II) centers. The PN ligand forms an
consistent with a symmetrical coordination of the ligand to Zn. essentially planar five-membered chelate ring, consisting of the
In particular, a single proton resonance for the N-Me groups as P–N–C–N ligand skeleton bonded to the Zn center with the ligand
well as symmetrical signals for the pyridyl and the PPh2 plane bisecting the Br–Zn–Br plane. The asymmetric unit gave Zn–N
moieties were observed. Definitive structural confirmation distances of 2.067(4) Å and 2.091(4) Å (average = 2.08 Å) and Zn–P
came through single crystal X-ray analysis of this complex distances of 2.4030(13) Å and 2.4045(13) Å (average = 2.40 Å) and
shown in Fig. 1 (Fig. S1 and Table S2, ESI†), which established these distances are comparable to reported bond distances. The
[Zn(k2-2,6-{Ph2PNMe}2(NC5H3))Br2] (1) as a four coordinate Zn–Br distances are nearly identical and displayed average values of
Zn(II) dibromide species.
2.34 Å, again comparable to reported values. While the six angles
In 1 the potentially tridentate ligand binds to Zn only around each Zn center in 2 ranged from 80.53(11)1 to 124.86(4)1, the
through the two phosphine groups that flank the central pyridyl averages of all of these approach an ideal tetrahedral angle
group while the pyridyl-N(2) center remained beyond bonding with values of 108.31 and 108.41, respectively. The limitation of the
distance to Zn at 2.680 Å. This group is directed toward the bite angle in the PN ligand led to the smallest angle in the five-
Zn(II) center, likely due steric restraints and electrostatic forces. membered ring being the P–Zn–N angle.
For comparison the Zn–Npy distances in A are 2.076(3) Å and
The structural features of complexes 1 and 2 bear some
2.078(2) Å18 while those for C are 2.061(4) Å, 2.067(4) Å.17 resemblance to the catalytic Zn complexes A–D that have
Complex 1 displayed two identical Zn–P distances (Zn–P(1) demonstrated chemical and electrochemical catalysis for CO2
2.424(3) Å, Zn–P(2) 2.42(3) Å) and only slightly different Zn–Br reduction and H2 formation. The potential for similar reactivity
linkages (Zn–Br(1) 2.4016(13) Å, Zn–Br(2) 2.3868(15) Å). These from 1 and 2 began by performing cyclic voltammetry (CV)
bond distances are in-line with reported Zn bond distances.19,20 measurements, at cathodic potentials, of a 1.0 mM acetonitrile
The N(1) and N(3) centers in the ligand were planar with S solution of 1 [Zn(2,6-{Ph2PNMe}2(NC5H3))Br2] under N2 using a
angles equaling 3601. Although the Zn-centered angles range glassy carbon (GC) working electrode, a Pt counter electrode
from 101.92(8)1 to 115.50(5)1, the average for the six angles and a Ag pseudo-reference electrode, which was referenced to
around the Zn center are 109.51 giving a coordination geometry the ferrocene (Fc0/+) redox couple. Fig. 2 revealed two irrever-
that was appropriately categorized as pseudo-tetrahedral.
sible reduction peaks at ꢀ2.42 and ꢀ2.69 V vs. Fc+/0. Variation
A similar room temperature reaction of a toluene suspension of of the scan rate yielded linear i vs. (scan rate)1/2 which is an
ZnBr2 with N-(diphenylphosphino)-2-methylaminopyridine pro- indication of a freely diffusing species (Fig. S5, ESI†).22
duced [Zn(2-{Ph2PNMe}(NC5H3))Br2] (2) as a colorless precipitate.
A similar set of CV measurements were carried out on complex 2,
The 1H and 13C NMR spectra gave resonances and intensities [Zn(2-{Ph2PNMe}NC5H3))Br2] and are shown in Fig. 2 (Fig. S7 and S8,
consistent with a single environment for the ligand. Evidence for ESI†). Like complex 1, the CV for complex 2 showed two irreversible
coordination of the PPh2 group was given by the 31P NMR resonance reductions at ꢀ2.33 V and ꢀ2.73 V versus Fc+/0. The relationship of 1
and 2 can be seen in overlay of the CV’s.
In order to more clearly analyze the reduction behavior of 1
and 2 a computational study of these complexes and their
reduction products was undertaken. DFT optimization of
[Zn(2,6-{Ph2PNMe}2(NC5H3))Br2] (1), starting from the crystal-
lographic parameters, was performed with the B3LYP func-
tional, def2TZVP basis set and using the IEFPCM model for
solvation with acetonitrile. The resulting species was an excel-
lent match with the experimental structural features (Fig. S8
and S9, ESI†). These computations yielded frontier orbitals for
1 that displayed little contribution from Zn, with the HOMO
being a p molecular orbital of the ligand that was largely
localized on the py and NMe groups with the fragment orbital
allocation of 0.51 (py), 0.27 (NMe) (Fig. S10, ESI†). The LUMO
Fig. 1 Reaction scheme and structures of complexes 1 and 2.
was also ligand localized and mostly p in nature localized on
Chem. Commun.
This journal is © The Royal Society of Chemistry 2021

View Article Online
ChemComm
Communication
Fig. 3 Cathodic CV measurements of complex [Zn(2,6-{Ph2PNMe}2
(NC5H3))Br2] (1) in CH3CN with a GC electrode under N2 (blue), and under
a CO2 atmosphere with 2.5 M of added water (purple). All potentials are
referenced to Fc+/Fc.
Fig. 2 Overlay of the CV’s of complexes [Zn(2,6-{Ph2PNMe}2(NC5H3))Br2]
(1, orange) and [Zn(2-{Ph2PNMe}(NC5H3))Br2] (2, blue) in CH3CN with
tetrabutylammonium hexafluorophosphate (TBAHFP) supporting electro-
lyte using a glassy carbon (GC) working electrode, a Pt counter electrode.
A potentials are referenced to the Fc+/0 couple.
An investigation of the electrocatalytic ability of complex 1
toward CO2 reduction, was carried out by performing a CV
measurement of 1 under a CO2 atmosphere with H2O added as
a proton source. As shown in Fig. 3, a characteristic increase in
current was observed under these conditions, features consis-
tent with electrocatalytic CO2 reduction by 1. After electrolytic
reaction, the electrode was removed, rinsed and immersed in
fresh electrolyte/water solution containing no Zn complex and
no catalytic response was observed.
To definitively establish the catalytic behavior of complex 1,
controlled potential coulometry (CPC) experiments were per-
formed and the reactor headspace was examined by GC and
solution NMR analysis was employed to probe the formation of
condensed phase products such as formic acid (Fig. S25, ESI†).
CPC measurements were carried out at ꢀ2.3 V and ꢀ2.7 V vs.
Fc+/0 under a CO2 atmosphere with either 200 mL or 500 mL of
added water. The results of these investigations, presented in
Table 1 (Table S9, ESI†), documented the formation of COꢀand
H2 as the only observed gas phase products and HCO2 in
solution. Both formate and CO formation confirm the unique
electrocatalytic abilities of 1 as the only reported Zn-based
catalyst with quantified CO formation.17 Furthermore, in con-
trast to B, which reduced CO2 through a methylcarbonate
intermediate, 1 generated formate directly from CO2 and
H2O.14
the pendant Ph groups (56%) with some py contribution (22%)
(Fig. S11, ESI†). The lowest energy, unoccupied, orbital with
substantial Zn contribution was 0.2283 hartrees (6.21 eV) above
the HOMO and was clearly s antibonding with Zn s orbital
(Fig. S12, ESI†). Beginning with the optimized structure of
complex 1, an electron was added and a new DFT optimization
was performed (B3LYP/def2TZVP/IEFPCM/CH3CN). This led to
an anionic complex [Zn(2,6-{Ph2PNMe}2(NC5H3))Br2]ꢀ (1ꢀ)
(Fig. S13, ESI†). The optimized anion gave a complex that main-
tained a tetrahedral environment on Zn with two phosphine and two
bromo linkages. The anion yielded a Zn center that was less
symmetrical than in the parent 1. One of the Zn–P bonds showed
a slight decrease in Zn–P distance with an increase in bond order
compared to the other Zn–P linkage. Both Zn–Br increased slightly
in bond length. Consistent with these observations, the SOMO was
localized on the ‘‘shorter’’ Zn–P side of 1ꢀ (Fig. S14, ESI†). The
orbital array of 1ꢀ remains similar to 1 with the frontier orbitals
dominate by ligand centered p-type orbitals. The SOMO is, as
expected, localized on the phenyl substituents (67%). The SOMOꢀ1
(formerly HOMO) retains similar orbital characteristics distributed
predominantly on py (53%) and NMe (28%). According to these
computations, there is no significant frontier electron character on
the Zn center and that the added electrons reside in p orbitals on the
Ph and py groups.
A similar process was carried out for the optimization of 2
[Zn(2-{Ph2PNMe}NC5H3))Br2] and the corresponding anion 2ꢀ
(DFT/B3LYP/def2TZVP/IEFPCM/CH3CN). The optimized struc-
ture for 2 was an excellent reflection of the experimental X-ray
structure (Fig. S16 and S17, ESI†). For 2, the HOMO is a p
orbital on py with contribution from the bridging amine NMe
function (Fig. S18, ESI†) and the LUMO is of p character with
the majority localized on the py portion of the ligand (Fig. S19,
ESI†). For the single electron reduction product 2ꢀ was success-
fully computationally modelled (Fig. S21 and Table S8, ESI†).
Compound 2ꢀ displayed a SOMO that was PN ligand centered
(Fig. S22 and S23, ESI†). Again computational effort makes it
clear that there is no significant frontier electron character on
the Zn center and that the added electrons reside in p orbitals
on the Ph and py groups.
These measurements gave competitive and significant
(10–100 mmol) H2 gas formation (Table S9, ESI†). Furthermore,
Table
1
Controlled potential coulometry of 1 mM (15 micromoles)
solution of [Zn(2,6-{Ph2PNMe}2(NC5H3))Br2] (1) in acetonitrile under
CO2 atmosphere
Potential (V) Time Added water
CO (mmol)/ HCO2H FEa
vs. Fc+/0
(h)
(mL)/[H2O] (M) select.b
(mmol)
(%)
ꢀ2.3
ꢀ2.3
ꢀ2.3
ꢀ2.7
ꢀ2.7
5
5
11
5
200/0.7
500/1.9
500/1.9
200/0.7
500/1.9
14.2/1
0
70
69
64
74
58
49/0.78
80.9/0.79
24.7/1
13.5
21.8
0
5
143.1/0.76
44.8
a
Corrected for background H2 formation shown in Table S9 (ESI).
b
Select. = mmol CO/(CO + HCO2H) mmol.
This journal is © The Royal Society of Chemistry 2021
Chem. Commun.

View Article Online
Communication
ChemComm
complex 2, the lack of a free basic site (i.e. pyridyl) on the ligand
does not allow for the bifunctional activation of CO2 consistent
with the lack of reduction products from this species.
ꢀ
Electrocatalytic reduction of CO2 to yield CO and HCO2
presents a key process in the utilization of CO2 as a C1 source
for further synthesis of as a fuel. Significantly, this report
presents a unique multifunctional Zn complex that can effec-
tively catalyze both of these transformations in the presence of
water. This is unique among a rare set of Zn complexes that can
perform CO2 reduction. The observed reactivity is attributed to
a synergy between a reduced, basic ligand and the acidity of the
Zn(II) center. Our further efforts are focused on examining the
competitive water reduction to yield H2, investigating para-
meters to improve on product selectivity and on exploring the
role of pendant basic sites in bifunctional electroreduction.
Fig. 4 A proposed mechanism for the reduction of CO2 with complex
[Zn(2,6-{Ph2PNMe}2(NC5H3))Br2] (1).
Conflicts of interest
HCO2H appeared at 1.9 M H2O, concomitant with CO selectivity
dropping to E78%. Similar observations were made when a
more negative reduction potential was employed. Analogous
measurements carried out in the absence of complex 1 yielded
no CO2 reduction products and only at ꢀ2.7 V was a back-
ground generation of 2–3 mmol H2 observed. Finally, CPC
measurements made under an N2 atmosphere produced no
detected H2 at ꢀ2.3 V and only 3–6 mmol of H2 were observed at
ꢀ2.7 V. These CPC experiments documented the ability of the
Zn(II) complex 1 to electrocatalytically reduce both CO2, to give
CO and formic acid, and water to yield H2.
There are no conflicts to declare.
Notes and references
1 N. Elgrishi, M. B. Chambers, X. Wang and M. Fontecave, Chem. Soc.
Rev., 2017, 46, 761–796.
2 J. Bonin, A. Maurin and M. Robert, Coord. Chem. Rev., 2017, 334,
184–198.
3 H. Takeda, C. Cometto, O. Ishitani and M. Robert, ACS Catal., 2017,
7, 70–88.
4 Y. Yamazaki, H. Takeda and O. Ishitani, J. Photochem. Photobiol., C,
2015, 25, 106–137.
5 K. E. Dalle, J. Warnan, J. J. Leung, B. Reuillard, I. S. Karmel and
E. Reisner, Chem. Rev., 2019, 119, 2752–2875.
6 R. Reithmeier, C. Bruckmeier and B. Rieger, Catalysts, 2012, 2,
544–571.
7 S. Kissling, P. T. Altenbuchner, T. Niemi, T. Repo and B. Rieger, in
Zn in Catalysis. Applications in Organic Synthesis, ed. S. Enthaller and
X.-F. Wu, Wiley-VCH Verlag GmbH & Co., Weinheim, Germany,
2015, pp. 179–206.
For comparison, a parallel set of measurements were carried
out with [Zn(2-{Ph2PNMe}(NC5H3))Br2] (2), which displayed
enhanced cathodic current in the presence of H2O with no
additional enhancements when the atmosphere was changed
to CO2 (Fig. S26, ESI†). Initial CPC measurements of 2 at ꢀ2.6 V
in acetonitrile with added water (0.7 M) and under a CO2
atmosphere gave 22 mmol H2 and no indications of CO2
reduction products.
8 S. Klaus, M. W. Lehenmeier, C. E. Anderson and B. Rieger, Coord.
Chem. Rev., 2011, 255, 1460–1479.
9 M. Kato and T. Ito, Inorg. Chem., 1985, 24, 504–508.
A proposed mechanism for the electrocatalytic reduction of 10 M. Kato and T. Ito, Inorg. Chem., 1985, 24, 509–514.
11 M. N. Collomb-Dunand-Sauthier, A. Deronzier and R. Ziessel,
J. Phys. Chem., 1993, 97, 5973–5979.
12 R. Dobrovetsky and D. W. Stephan, Angew. Chem., Int. Ed., 2013, 52,
CO2 by complex 1 is presented in Fig. 4. The first step in this
proposal is based on a ligand-centered reduction and on the
proximity of the free pyridyl base and a Lewis acidic Zn
center.12,17 The increased basicity of the reduced ligand
together with the additional coordination by the zinc ion
facilitate the binding of CO2 and formation of pyridyl-N coor-
dinated CO2 intermediate.21,22 The bound and activated CO2
then proceeds along two alternate pathways to yield either CO
or HCOOꢀ.23 In one case, protonation leads to a metallacar-
boxylic acid (A), which is commonly invoked intermediate in
reduction of CO2 to yield CO. Subsequent protonation coupled
2516–2519.
13 W. Sattler and G. Parkin, J. Am. Chem. Soc., 2012, 134, 17462–17465.
14 S. P. Cronin, J. M. Strain, M. S. Mashuta, J. M. Spurgeon,
R. M. Buchanan and C. A. Grapperhaus, Inorg. Chem., 2020, 59,
4835–4841.
15 A. Z. Haddad, B. D. Garabato, P. M. Kozlowski, R. M. Buchanan and
C. A. Grapperhaus, J. Am. Chem. Soc., 2016, 138, 7844–7847.
16 S. P. Cronin, A. Al Mamun, M. J. Toda, M. S. Mashuta, Y. Losovyj,
P. M. Kozlowski, R. M. Buchanan and C. A. Grapperhaus, Inorg.
Chem., 2019, 58, 12986–12997.
17 E. S. Donovan, B. M. Barry, C. A. Larsen, M. N. Wirtz, W. E. Geiger
and R. A. Kemp, Chem. Commun., 2016, 52, 1685–1688.
with reduction leads to water elimination, release of CO and 18 W. Sattler and G. Parkin, J. Am. Chem. Soc., 2011, 133, 9708–9711.
19 J. Burt, W. Levason and G. Reid, Coord. Chem. Rev., 2014, 260, 65–115.
20 R. E. Desimone and G. D. Stucky, Inorg. Chem., 1971, 10, 1808–1812.
21 J. N. Gayton, Q. Li, L. Sanders, R. R. Rodrigues, G. Hill and
J. H. Delcamp, ACS Omega, 2020, 5, 11687–11694.
22 P. Deria, S. Li, H. Zhang, R. Q. Snurr and T. Hupp, Chem. Commun.,
2015, 51, 12478–12481.
23 R. Francke, B. Schille and M. Roemelt, Chem. Rev., 2018, 118,
4631–4701.
closes that leg of the cycle. A path leading to formate complex
(B) was envisioned to be directly linked to electrochemical
generation of hydrogen, a process with literature precedent.14
Alternatively, formation of a transient Zn–H species and inser-
tion of CO2 could result in complex B.13 Substitution of formate
with bromide completes this portion of the cycle. In the case of
Chem. Commun.
This journal is © The Royal Society of Chemistry 2021