Cationic dirhodium(II,II) complexes for the electrocatalytic reduction of CO2 to HCOOH

Article abstract of DOI:10.1039/c6cc03253b

Two formamidinate bridged dirhodium(ii,ii) complexes with chelating diimine ligands L, [Rh₂(μ-DTolF)₂(L)₂]²⁺, were shown to electrocatalytically reduce CO₂ in the presence of H₂O. Analysis

Full text of DOI:10.1039/c6cc03253b

ChemComm
View Article Online
View Journal
COMMUNICATION
Cationic dirhodium(II,II) complexes for the
electrocatalytic reduction of CO2 to HCOOH†
Cite this:DOI: 10.1039/c6cc03253b
Suzanne E. Witt,a Travis A. White,a Zhanyong Li,b Kim R. Dunbar*b and
Claudia Turro*a
Received 18th April 2016,
Accepted 13th September 2016
DOI: 10.1039/c6cc03253b
www.rsc.org/chemcomm
Two formamidinate bridged dirhodium(II,II) complexes with chelating further with CO2 to form Rh(I) carbonyl compounds.19
diimine ligands L, [Rh2(l-DTolF)2(L)2]2+, were shown to electro- In contrast, mononuclear rhodium complexes that contain
catalytically reduce CO2 in the presence of H2O. Analysis of the diphosphine or polypyridyl ligands were shown to reduce CO2
reaction mixture and headspace following bulk electrolysis revealed to formate anions.11,20–24
H2 and HCOOH as the major products. The variation in relative
product formation is discussed.
Herein we present a bimetallic dirhodium(II,II) architecture
bridged by formamidinate ligands for the electrocatalytic reduction
of CO2. Each metal in these cationic complexes is chelated by a
Carbon dioxide represents an abundant source for the produc- diimine ligand, with the overall formula [Rh2(m-DTolF)2(dpq)2]2+
tion of fuels and chemicals needed in industrial processes. (1; DTolF = p-ditolylformamidinate, dpq = dipyrido[3,2-f:20,30-
Therefore, a facile and efficient means to convert CO2 into h]quinoxaline) and [Rh2(m-DTolF)2(phen)2]2+ (2; phen = 1,10-
useful chemical feedstocks and fuels is highly desirable.1,2 phenanthroline), whose structures are shown in Fig. 1. The excited
In recent years, research in the area of homogeneous electro- state properties of the series [Rh2(m-DTolF)2(L)2]2+, L = dpq, dppz
catalytic CO2 reduction has garnered substantial interest owing (dipyrido[3,2-a:20,30-c]phenazine), and dppn (benzo[i]dipyrido[3,2-
to its highly tunable and controllable nature.2–4 The distribu- a:20,30-h]quinoxaline) were previously investigated by us which
tion of products formed from electrocatalytic CO2 reduction revealed that these and related complexes are powerful reducing
is dependent on multiple factors, including the catalyst used, agents in the excited state,25–27 and that they function as efficient
acid concentration, and electron availability,1–4 such that a and robust electrocatalysts for H+ reduction.28 A proposed inter-
judicious choice of reaction components is required to attain mediate in the electrocatalytic H+ reduction cycle of these
the formation of desired products. The 2eꢀ/2H+ transformation complexes involves the formation of a RhI2I,III–H hydride species
of CO2 to HCOOH is an attractive process to provide a source of that may be exploited to afford CO2 reduction to HCOOH via CO2
formic acid for the textile, cleaning, and preservative industries. insertion into the Rh–H bond. In the present work, 1 and 2
In addition, this transformation can also provide a method for are explored for the electrocatalytic reduction of CO2 and the
the storage of hydrogen fuel in a condensed form.5
resulting products are reported.
It has been proposed that the formation of HCOOH by the
Cyclic voltammograms of 1 and 2 were performed under a N2
reduction of CO2 proceeds via the insertion of CO2 into a metal– atmosphere in CH3CN and compared to those recorded in saturated
hydride bond, as the direct coordination of CO2 to the metal center CO2 solutions (Fig. 2). Under N2, 1 exhibits a RhI2II,II/II,II reduction
followed by reduction typically results in the formation of CO.6 at E1/2 = ꢀ0.38 V vs. Ag/AgCl; this couple is slightly shifted
Transition metal complexes of Fe, Co, Ni, Ru, Os, and Ir are known to a more negative potential E1/2 = ꢀ0.45 V vs. Ag/AgCl in 2.
to catalyze production of HCOOH from CO2,6–18 but molecular Similar shifts of this metal-centered process were observed in
Rh-based catalysts for CO2 reduction are less common.11,19–24
Rhodium hydride compounds react with CO2 and water to form
dihydrido bicarbonato Rh(III) complexes, which then react
a Department of Chemistry and Biochemistry, The Ohio State University, Columbus,
OH 43210, USA. E-mail: turro.1@osu.edu
b Department of Chemistry, Texas A&M University, College Station, TX 77843, USA.
E-mail: dunbar@mail.chem.tamu.edu
Fig. 1 Schematic representation of the molecular structures of 1 and 2.
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6cc03253b
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
(Fig. 2a), which can be attributed to protonation of the nitrogen
atom on the pyrazine portion of the reduced dpq ligand, as
previously reported in the presence of acid.28 As the potential is
scanned to more negative values, significantly greater catalytic
current enhancement is observed. The possibility that catalysis
occurs at the dpq ligand must be considered, since the pyrazine
moiety possesses a lone pair of electrons on each nitrogen atom
that may interact with a proton or the d+ carbon atom within
CO2. However, CV scans of [Ru(bpy)2(dpq)]2+, which possesses a
dpq ligand and lacks an open coordination site on the metal
center, exhibit no catalytic behavior and only a small current
increase attributed to ligand protonation (Fig. S1, ESI†). There-
fore, it may be concluded that the catalysis observed in 1 occurs
at one of the Rh centers and not on the dpq ligand. In addition,
this result provides further evidence that the current increase
observed at the first dpq ligand reduction is associated with
protonation of the ligand,28 since the addition of CO2 to CH3CN
is known to increase the acidity of the solution.24
For 2, the first phen0/ꢀ couple becomes irreversible in the
presence of CO2, with the current remaining unchanged (Fig. 2b,
dashed line). Given that the phen ligand does not possess
accessible nitrogen atoms, it is not expected to undergo protona-
tion or to interact with CO2. As such, no current increase or shift
in the peak potential is observed for this couple. The loss of
reversibility could be a signature of protonation or CO2 binding at
Fig. 2 Cyclic voltammograms of 0.5 mM (a) 1 and (b) 2 in 0.1 M TBAPF6/
CH3CN under N2 (solid lines) and CO2 (dashed lines).
the [Rh2(DTolF)2(L)2]2+ (L = dpq, dppz, dppn) series.25,28 The the Rh metal center, and the continued increase in current
ligand-based dpq0/ꢀ reduction in 1 is observed at E1/2 = ꢀ1.07 V following the first phen ligand reduction is evidence of catalytic
vs. Ag/AgCl, whereas the phen0/ꢀ couple appears at E1/2
=
behavior. For 1 and 2, a significantly greater catalytic current
ꢀ1.21 V vs. Ag/AgCl in 2, such that the phen reduction is more enhancement was observed when 3 M H2O was added to the
negative than that of dpq with a shift, DE, of 0.14 V. This shift is reaction mixture as a proton source (Fig. 3).
consistent with the presence of the pyrazine moiety in the dpq
Bulk electrolysis experiments were conducted at ꢀ1.40 V and
ligand, which serves to extend the p-system and renders it ꢀ1.60 V vs. Ag/AgCl for 3 hours to determine the products of
easier to reduce than phen. Similar shifts have been previously CO2 reduction (Table 1). Gas phase products from the head
reported extensively for Ru(II) complexes with these and related space of each sample were analyzed using gas chromatography
ligands,29–31 with the first ligand-based reduction appearing at following electrolysis. No CO production was detected but H2
0.18 V more negative potential in [Ru(bpy)2(phen)2]2+ as compared evolution was observed in each case (Fig. S2, ESI†). To determine
to [Ru(bpy)2(dpq)]2+.29,30 The metal-centered Rh2II,II/II,I couples are the products present in solution after each bulk electrolysis
observed at E1/2 = ꢀ1.55 V and E1/2 = ꢀ1.77 V vs. Ag/AgCl in 1 and experiment, the solvent was evaporated and the sample was
1
2, respectively (Fig. 2). The RhI2I,II/II,I reduction occurs at a potential reconstituted in D2O for H NMR analysis. In this manner, the
more negative than that of the first diimine ligand reduction in presence of HCOOH was confirmed as a singlet at B8.23 ppm,
each complex; this couple was observed at 0.22 V more positive which agrees with the 1H NMR spectrum of the formic acid
potential in 1 as compared to 2. This shift is attributed to the fact standard and differs from that of the formate anion standard,
that the reduced pyrazine moiety of the dpq ligand in 1 is further which has a chemical shift of B8.44 ppm (Fig. S3, ESI†). The
away from the metal, such that it contributes less electron density faradaic efficiencies, FEs, for 1 remained relatively constant for
to the Rh2 core than a reduced phen ligand in 2. The reduction of H2 production at ꢀ1.4 V and ꢀ1.6 V, 56–63%, whereas those for
the second dpq ligand in 1 is evident at E1/2 = ꢀ1.72 V vs. Ag/AgCl, HCOOH production decreased from 12% to 3.5%, respectively
but the reduction of the second phen ligand is not observed in the (Table 1). In contrast, the results indicate that no HCOOH is
cyclic voltammogram of 2, likely because it lies outside of the generated by complex 2 at ꢀ1.4 V, but 7.0% FE for formic acid
scanned solvent window. These assignments in CH3CN agree with formation is observed at ꢀ1.6 V vs. Ag/AgCl. In addition, the
those previously made by our group with DMF as the solvent.28
relative TON (turnover number) for the generation of HCOOH vs.
Cyclic voltammograms collected in CO2-saturated CH3CN, H2, RTON = TON(H2)/TON(HCOOH), decreases for 1 from ꢀ1.4 V
[CO2] B 0.28 M,32 display current enhancement with an onset to ꢀ1.6 V, but the trend is reversed for 2. In the latter case, no
at approximately ꢀ0.90 V for 1 and at ꢀ1.55 V for 2 vs. Ag/AgCl, HCOOH is formed at ꢀ1.4 V, but a TON of 10.8 ꢁ 0.4 was
consistent with CO2 reduction (Fig. 2, dashed lines). For 1, a measured at ꢀ1.6 V vs. Ag/AgCl. Bulk electrolyses under identical
small, non-catalytic current increase and a loss of reversibility conditions in the absence of catalyst were conducted as control
occurs at the first dpq ligand reduction in the presence of CO2 experiments at both ꢀ1.4 V and ꢀ1.6 V, which generated small
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
ChemComm
Communication
with 13CO2. Bulk electrolysis of 2 was performed at ꢀ1.6 V in
the presence of 13CO2 and the products were analyzed by 1H
1
and 13C NMR spectroscopy. In the H NMR spectra, a singlet
corresponding to H12COOH was observed, as well as a doublet
corresponding to H13COOH (Fig. S5, ESI†). The coupling con-
stant for the latter, J B 185 Hz, agrees well with the literature
value for H13COOH.36 The amount of H12COOH formed is
consistent with that measured for graphite exfoliation (2.2 mmol),
and the amount of H13COOH formed is similar to the values
measured for electrocatalytic CO2 reduction with 2 (12.4 mmol,
8.25 TON, 5.5 %FE; Table 1). Two singlets are observed in the
13C NMR spectra, which are consistent with reported values for
carbonate and HCOOH (Fig. S6, ESI†).37 These peaks are not
observed prior to electrolysis. The formation of carbonate during
electrolysis may account for the remaining 15–35% of charge.
Absorption spectra collected before and after electrolysis are
inconsistent with catalyst decomposition (Fig. S7, ESI†). Slight
spectral changes and a small increase in the baseline can be
attributed to the formation of a black precipitate known to arise
from graphite exfoliation.
The clear trend of the increase in TON for H2 production for
1 and 2 as the applied bias is increased to more negative
potential is observed. A turn-on of the HCOOH production is
observed when the potential is increased from ꢀ1.40 V to
ꢀ1.60 V for 2. In this complex, at ꢀ1.40 V, the first phen ligand
has been reduced, which is not proposed as the active catalytic
species. It is only when the applied potential begins to
approach the RhI2I,II/II,I reduction that HCOOH is detected. This
explanation is supported by the low TON measured for H2
production for 2 at ꢀ1.4 V compared to ꢀ1.6 V and by the
values for complex 1. The RhI2I,II/II,I couple occurs at B220 mV
more positive potential in 1 than in 2, so catalysis is possible at
both ꢀ1.4 V and ꢀ1.6 V as the active catalytic species has been
generated. Therefore, the activities should be compared using
ꢀ1.4 V for 1 and ꢀ1.6 V for 2, since these values are B200 mV
beyond the potential at which the active species is formed in
each catalyst. Comparing these values, it is clear that 2 is a
more active catalyst than 1, however the selectivity for CO2 is
Fig. 3 Cyclic voltammograms of 0.5 mM (a) 1 and (b) 2 in 0.1 M TBAPF6/
CH3CN purged with N2 (solid lines), CO2 (dashed lines), and with CO2 and
3 M H2O (dotted lines).
Table 1 Turnover number (TON) and percent Faradaic efficiency (%FE)
following bulk electrolysis of 1 and 2 in the presence of CO2 and H2O at
various potentialsa
H2
HCOOH
TON
b
Complex Potential/V TON
%FE
%FE
RTON
1
2
ꢀ1.40
ꢀ1.60
ꢀ1.40
ꢀ1.60
48.7 ꢁ 8.1 56 ꢁ 4 4.3 ꢁ 1.1 12 ꢁ 4 0.09
187 ꢁ 15 63 ꢁ 6 4.9 ꢁ 1.1 3.5 ꢁ 2 0.03
17.8 ꢁ 1.8 49 ꢁ 6
0
0
0
118 ꢁ 8.0 77 ꢁ 4 10.8 ꢁ 0.4 7.0 ꢁ 0.4 0.09
a
[Complex] = 0.5 mM, 0.1 M TBAPF6/CH3CN with B0.28 M CO2 and
3 M H2O; held at each potential for 3 hours conducted in triplicate.
b
Relative TON, RTON, TON(HCOOH)/TON(H2).
amounts of H2 and HCOOH (see the ESI†). The quantities very similar. The lower activity observed in 1 is consistent with
obtained from these experiments were subtracted from those our previous work, in which we showed that the protonation of
obtained from bulk electrolysis in the presence of a catalyst in the nitrogen atoms on the reduced dpq ligand consumes the
order to calculate TON values and %FE.
substrate, but that species is not catalytic.28
In addition, bulk electrolysis at ꢀ1.6 V under N2 in the
As mentioned previously, the bulk electrolysis data shows
presence of a catalyst was performed to confirm CO2 as the that both complexes produce significantly more H2 than
source of carbon for HCOOH formation and the details of HCOOH, indicating selectivity for the reduction of protons over
the results appear in the ESI.† For both 1 and 2 a small amount CO2. If HCOOH formation proceeds through CO2 insertion into
of HCOOH was detected which can be attributed to the graphite a Rh–H hydride bond, that process must occur before the metal
electrode interacting with H2O in solution. The electrochemical center is protonated a second time to evolve H2. Therefore, the
exfoliation of graphite has been documented throughout the relative concentrations of protons and CO2 in solution are
literature.33–35 When water is reduced at the electrode, hydroxyl expected to play a significant role in the outcome of electrolysis.
anions may oxidize bulk graphite, corroding the electrode Work is currently underway to improve the selectivity for CO2
surface.33–35 However, the yield of HCOOH produced in the reduction, such as increasing the pH of the reaction mixture.
absence of CO2 is very low, clearly showing that complexes 1 Additionally, since the identity of the diimine ligand appears to
and 2 act as catalysts for this reaction.
have little effect on the selectivity, the ongoing work will focus
To confirm that the carbon in HCOOH originates from the on manipulating the bridging ligand and blocking one or both
electrocatalytic reduction of CO2, experiments were conducted axial positions.38
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
In conclusion, [Rh2(m-DTolF)2(L)2]2+ (L = dpq, phen) complexes
were shown to exhibit electrocatalytic activity under a CO2
atmosphere. The products HCOOH and H2 were formed upon
bulk electrolysis of the complexes in acetonitrile in the presence of
CO2 and water. Further analysis indicated that catalysis occurred at
the metal center, and not on the diimine ligand. Complex 1 is a less
active catalyst than 2, likely because the pyrazine moiety of the dpq
ligands in 1 becomes protonated and consumes the substrate. The
active catalytic species, Rh2II,I, however, is formed at a more positive
potential in 1, so catalysis may occur at a lower overpotential than
in 2. Both complexes are selective for proton reduction to H2 under
the current conditions.
15 S.-N. Pun, W.-H. Chung, K.-M. Lam, P.-H. Chan, K.-Y. Wong, C.-M.
Che, Y.-Y. Chen and S.-M. Peng, J. Chem. Soc., Dalton Trans., 2002,
575–583.
16 M. D. Rail and L. A. Berben, J. Am. Chem. Soc., 2011, 133, 18577–18579.
17 P. Kang, C. Cheng, Z. Chen, C. K. Schauer, T. J. Meyer and
M. Brookhart, J. Am. Chem. Soc., 2012, 134, 5500–5503.
18 S. T. Ahn, E. A. Bielinski, E. M. Lane, Y. Chen, W. H. Bernskoetter,
N. Hazari and G. T. Palmore, Chem. Commun., 2015, 51, 5947–5950.
19 T. Yoshida, D. L. Thorn, T. Okano, J. A. Ibers and S. Otsuka, J. Am.
Chem. Soc., 1979, 101, 4212–4221.
20 S. Slater and J. H. Wagenknecht, J. Am. Chem. Soc., 1984, 106,
5367–5368.
21 C. M. Bolinger, B. P. Sullivan, D. Conrad, J. A. Gilbert, N. Story and
T. J. Meyer, J. Chem. Soc., Chem. Commun., 1985, 796–797.
22 C. M. Bolinger, N. Story, B. P. Sullivan and T. J. Meyer, Inorg. Chem.,
1988, 27, 4582–4587.
23 P. Paul, B. Tyagi, A. K. Bilakhiya, M. M. Bhadbhade, E. Suresh and
G. Ramachandraiah, Inorg. Chem., 1998, 37, 5733–5742.
24 C. Caix, S. Chardon-Noblat and A. Deronzier, J. Electroanal. Chem.,
1997, 434, 163–170.
The authors are grateful for the generous support of the
U.S. Department of Energy, Office of Science, Office of Basic
Research (CT: DE-SC0010542; KRD: DE-SC0010721).
25 Z. Li, N. A. Leed, N. M. Dickson-Karn, K. R. Dunbar and C. Turro,
Chem. Sci., 2014, 5, 727–737.
26 (a) P. M. Bradley, B. E. Bursten and C. Turro, Inorg. Chem., 2001, 40,
1376–1379; (b) P. K.-L. Fu, P. M. Bradley and C. Turro, Inorg. Chem.,
2001, 40, 2476–2477.
Notes and references
1 M. Aresta and A. Dibenedetto, Dalton Trans., 2007, 2975–2992.
2 J. Qiao, Y. Liu, F. Hong and J. Zhang, Chem. Soc. Rev., 2014, 43, 27 D. A. Lutterman, N. N. Degtyareva, D. H. Johnston, J. C. Gallucci,
631–675. J. L. Eglin and C. Turro, Inorg. Chem., 2005, 44, 5388–5396.
3 C. Costentin, M. Robert and J.-M. Saveant, Chem. Soc. Rev., 2013, 42, 28 T. A. White, S. E. Witt, Z. Li, K. R. Dunbar and C. Turro, Inorg. Chem.,
2423–2436. 2015, 54, 10042–10048.
4 E. E. Benson, C. P. Kubiak, A. J. Sathrum and J. M. Smieja, Chem. 29 J. V. Caspar and T. J. Meyer, Inorg. Chem., 1983, 22, 2444–2453.
Soc. Rev., 2009, 38, 89–99.
5 S. Moret, P. J. Dyson and G. Laurenczy, Nat. Commun., 2014, 5, 1–7.
30 A. Delgadillo, P. Romo, A. M. Leiva and B. Loeb, Helv. Chim. Acta,
2003, 86, 2110–2120.
6 J. R. Pugh, M. R. M. Bruce, B. P. Sullivan and T. J. Meyer, Inorg. 31 A. A. Vlcek, E. S. Dodsworth, W. J. Pietro and A. B. P. Lever,
Chem., 1991, 30, 86–91. Inorg. Chem., 1995, 34, 1906–1913.
7 L. Chen, Z. Guo, X.-G. Wei, C. Gallenkamp, J. Bonin, E. Anxolabehere- 32 Y. Tomita, S. Teruya, O. Koga and Y. Hori, J. Electrochem. Soc., 2000,
Mallart, K.-C. Lau, T.-C. Lau and M. Robert, J. Am. Chem. Soc., 2015,
137, 10918–10921.
8 M. R. M. Bruce, E. Megehee, B. P. Sullivan, H. H. Thorp, T. R. O’Toole,
147, 4164–4167.
33 C. T. J. Low, F. C. Walsh, M. H. Chakrabarti, M. A. Hashim and
M. A. Hussain, Carbon, 2013, 54, 1–21.
A. Downard, J. R. Pugh and T. J. Meyer, Inorg. Chem., 1992, 31, 34 A. M. Abdelkader, A. J. Cooper, R. A. W. Dryfe and I. A. Kinloch,
4864–4873. Nanoscale, 2015, 7, 6944–6956.
9 H. Ishida, K. Tanaka and T. Tanaka, Organometallics, 1987, 6, 181–186. 35 Z. Y. Xia, G. Giambastiani, C. Christodoulou, M. V. Nardi, N. Koch,
10 P. Kang, T. J. Meyer and M. Brookhart, Chem. Sci., 2013, 4, 3497–3502.
11 S. C. Rasmussen, M. M. Richter, E. Yi, H. Place and K. J. Brewer,
Inorg. Chem., 1990, 29, 3926–3932.
12 H. Ishida, H. Tanaka, K. Tanaka and T. Tanaka, J. Chem. Soc., Chem.
Commun., 1987, 131–132.
13 J. P. Collin, A. Jouaiti and J. P. Sauvage, Inorg. Chem., 1988, 27,
1986–1990.
E. Treossi, V. Bellani, S. Pezzini, F. Corticelli, V. Morandi, A. Zanelli
and V. Palermo, ChemPlusChem, 2014, 79, 439–446.
36 S. M. Ng, C. Yin, C. H. Yeung, T. C. Chan and C. P. Lau, Eur. J. Inorg.
Chem., 2004, 1788–1793.
37 Z. Chen, C. Chen, D. R. Weinberg, P. Kang, J. J. Concepcion,
D. P. Harrison, M. S. Brookhart and T. J. Meyer, Chem. Commun.,
2011, 47, 12607–12609.
˜
14 C. Arana, S. Yan, M. Keshavarz-K, K. T. Potts and H. D. Abruna, 38 J. D. Aguirre, D. A. Lutterman, A. M. Angeles-Boza, K. R. Dunbar and
Inorg. Chem., 1992, 31, 3680–3682.
C. Turro, Inorg. Chem., 2007, 46, 7494–7502.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016