Steering CO2 electroreduction toward ethanolproduction by a surface-bound Ru polypyridylcarbene catalyst on N-doped porous carbonYanming Liua,b, Xinfei Fanc, Animesh Nayakb, Ying Wangb, Bing Shanb, Xie Quana, and Thomas J. Meyerb,1

aKey Laboratory of Industrial Ecology and Environmental Engineering, Ministry of Education, School of Environmental Science and Technology, DalianUniversity of Technology, Dalian 116024, China; bDepartment of Chemistry, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599; and cCollegeof Environmental Science and Engineering, Dalian Maritime University, Dalian 116024, China

Contributed by Thomas J. Meyer, November 10, 2019 (sent for review May 6, 2019; reviewed by Andrew B. Bocarsly and Clifford P. Kubiak)

Electrochemical reduction of CO2 to multicarbon products is a sig-nificant challenge, especially for molecular complexes. We reporthere CO2 reduction to multicarbon products based on a Ru(II) poly-pyridyl carbene complex that is immobilized on an N-doped porouscarbon (RuPC/NPC) electrode. The catalyst utilizes the synergisticeffects of the Ru(II) polypyridyl carbene complex and the NPCinterface to steer CO2 reduction toward C2 production at lowoverpotentials. In 0.5 M KHCO3/CO2 aqueous solutions, Faradaicefficiencies of 31.0 to 38.4% have been obtained for C2 productionat −0.87 to −1.07 V (vs. normal hydrogen electrode) with 21.0 to27.5% for ethanol and 7.1 to 12.5% for acetate. Syngas is also pro-duced with adjustable H2/CO mole ratios of 2.0 to 2.9. The RuPC/NPC electrocatalyst maintains its activity during 3-h CO2-reductionperiods.

CO2 reduction | electrocatalysis | Ru(II) polypyridyl complex | porous carbon

Electrocatalytic reduction of CO2 to useful fuels and chemicalfeedstocks is a promising strategy for carbon utilization andgreenhouse gas mitigation. Among the CO2-reduction productsincluding CO, formate, methanol, methane, acetate, ethanol,etc., liquid multicarbon products such as ethanol and acetate aredesirable because of their high energy densities and economicvalues (1, 2). A variety of electrocatalysts have been explored forCO2 reduction, including metals (3, 4), metal oxides (5),heteroatom-doped carbon nanomaterials (6), molecular com-plexes (7–9), immobilized molecular complexes (10), and hybridcatalysts (11). Manipulation of morphology (12, 13), oxidationstate (14), and introduction of dopants (15), alloys (16), andsingle-metal atoms (17, 18) have been employed to controloverpotential, steer product distributions, enhance activity, andcontrol selectivity toward specific products. Significant progresshas been made, but, to date, the most common products for CO2electroreduction are CO and formate.Immobilized molecular complexes such as porphyrins (19, 20),

phthalocyanines (21), polypyridyl carbenes (22), and their hybridcatalysts (23, 24) have been investigated for electrochemicalreduction of CO2. They offer the merits of tailorable catalyticsites and molecular structures for enabling electrocatalytic per-formance optimization. By immobilizing molecular complexes onthe electrode surface, the catalysts are easy to reuse and showimproved CO2-reduction performance (25, 26). In previous works,we have demonstrated that the Ru(II) polypyridyl carbene complex[RuII(tpy)(Mebim-py)(H2O)]

2+ (tpy, 2,2′:6′,2″-terpyridine; Mebim-py,3-methyl-1-pyridyl-benzimidazol-2-ylidene) is an effective catalystfor electrochemical reduction of CO2 to CO with high selectivity insolution and on surfaces (22, 27, 28). CO2 reduction occurs byproton-coupled electron transfer through a carbene complex sta-bilized intermediate [RuII(tpy)(Mebim-py)(CO)]+ to give CO asthe product, which could reduce overpotential for CO2 reduction.A significant challenge that remains is development of electro-catalysts that steer CO2 reduction toward multicarbon productswith high selectivity at low overpotentials. Formation of C–C bonds

necessitates coupling reactions between a CO intermediate and/orintermediates from CO protonation (29–31). Assembling the Ru(II)polypyridyl carbene on an electrode surface that is capable ofC–C dimerization offers an attractive strategy for reducing CO2to multicarbon products.N-doped carbon nanomaterials have been widely used for

electroreduction due to their electrocatalytic activity and low cost.N-doped carbon electrodes have been shown to be capable of C–Cdimerization (6, 32). Combining the Ru(II) polypyridyl carbenecatalyst with an N-doped porous carbon electrode (RuPC/NPC)provides an appealing approach to facilitate CO2 electroreductiontoward multicarbon products. The immobilized Ru(II) polypyridylcarbene can provide atomically distributed active sites for electro-catalysis. The large surface area and porous structure of NPCfavors Ru(II) polypyridyl carbene attachment at catalytic inter-faces and exposes reactive sites.In the experiments described here, the pyrene-derivatized Ru(II)

polypyridyl carbene complex was attached to NPC by π–π stackingon the surface. As noted below, the catalytic results were notable inidentifying a greatly enhanced reactivity toward the formation ofethanol as a major product at relatively low overpotentials.

Results and DiscussionSynthesis and Characterization of the RuPC/NPC Electrode. The RuPC/NPC hybrid catalyst was prepared by attaching the pyrene-derivatizedRu(II) polypyridyl carbene (22, 27, 28, 33) on NPC bonded by π–π

Significance

Electrochemical reduction of CO2 can convert CO2 emission backto value-added fuels and chemicals and store renewable elec-tricity. Reducing CO2 to multicarbon products has attracted greatinterest because of their higher energy densities and associatedeconomic values. We report here a promising strategy forsteering CO2 electroreduction toward ethanol production byexploiting a surface-bound Ru polypyridyl carbene catalyst on anN-doped porous carbon electrode. We show the synergistic ef-fects of Ru polypyridyl carbene for CO intermediate productionwith a porous carbon for C–C coupling that could boost ethanolproduction at relatively low overpotentials. The strategy pro-vides insights on how to improve selectivity and efficiency forCO2 reduction toward multicarbon products.

Author contributions: Y.L. and T.J.M. designed research; Y.L., X.F., A.N., and Y.W. per-formed research; Y.L., X.F., A.N., Y.W., B.S., X.Q., and T.J.M. analyzed data; and Y.L., X.F.,A.N., Y.W., B.S., X.Q., and T.J.M. wrote the paper.

Reviewers: A.B.B., Princeton University; and C.P.K., University of California San Diego.

The authors declare no competing interest.

Published under the PNAS license.1To whom correspondence may be addressed. Email: tjmeyer@unc.edu.

This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplemental.

First published December 10, 2019.

www.pnas.org/cgi/doi/10.1073/pnas.1907740116 PNAS | December 26, 2019 | vol. 116 | no. 52 | 26353–26358

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

June

7, 2

021

http://crossmark.crossref.org/dialog/?doi=10.1073/pnas.1907740116&domain=pdfhttps://www.pnas.org/site/aboutpnas/licenses.xhtmlmailto:tjmeyer@unc.eduhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/cgi/doi/10.1073/pnas.1907740116

interactions between pyrene and the hexatomic carbon rings ofNPC (Fig. 1). The pyrene unit enables stable immobilization of Rupolypyridyl carbene on carbon surface (25). Briefly, the resol wassynthesized from phenol and formaldehyde polymerization underbasic conditions. The thermosetting resin was prepared by solventevaporation-induced self-assembly between resol, Pluronic F127(soft template), and dicyandiamide (nitrogen source) (34). Thesynthesized resin was pyrolyzed at 750 °C to obtain NPC. An RuPC/NPC electrode was prepared by loading the pyrene-derivatizedRu(II) polypyridyl carbene complex on NPC by π–π interactions.In brief, the NPC suspension (120 g/L in isopropanol) and RuPCsolution (100 μM in dimethyl sulfoxide [DMSO]-D6) were mixed(vol:vol = 10:1) by sonication for 10 min and stirring for 12 h,followed by drop-coating on a carbon fiber paper substrate. As acomparison, the NPC electrode was prepared with the samemethod without the added Ru(II) polypyridyl carbene.A transmission electron microscopic (TEM) image shows that

the NPC has a mesoporous structure (Fig. 2A). Its surface area,measured from N2 adsorption–desorption isotherms, is 302.4 m

2·g−1

(Fig. 2B). The pore-size distribution curve (SI Appendix, Fig. S1)confirms that NPC has mesopores with sizes around 19.1 nm.The total pore volume of NPC is 0.32 cm3·g−1. The scanningelectron microscopic (SEM) image shows abundant pores on theNPC surface (Fig. 3A). After attaching the Ru(II) polypyridylcarbene on NPC, the surface of the RuPC/NPC hybrid catalystmaintains the porous structure (Fig. 3B). Energy-dispersive X-ray spectroscopic maps (Fig. 3 C and D) show the uniform dis-tribution of Ru and N on an RuPC/NPC hybrid catalyst, re-vealing that the Ru(II) polypyridyl carbene is homogeneouslydistributed on NPC.The N-containing species and surface content of NPC was

investigated by X-ray photoelectron spectroscopy (XPS). Asshown in the N 1s spectrum (Fig. 4A), pyridinic N (398.5 eV),pyrrolic N (400.1 eV), and graphitic N (401.2 eV) are observedon NPC (32). Among the 3 N species, pyridinic N is the main Nspecies for NPC, with a percentage of 49.9%, which has beenproved to be the most active N species for electrocatalysis (32).The total N content is 5.0 atomic% for NPC. In the XPS spectrumof the RuPC/NPC catalyst (Fig. 4B), the small peak centered at281.1 eV arises from Ru (35), showing that Ru(II) polypyridylcarbene has been successfully anchored to the surface.The amount of electrochemically active Ru(II) polypyridyl

carbene attached on NPC was estimated from cyclic voltammetrymeasurements on an RuPC/NPC hybrid electrode. Surface con-tent was evaluated by the expression Q = ΓnFA, where Q is thecharge obtained from redox peak integration, Γ is the amount of

electroactive Ru(II) polypyridyl carbene (mol·cm−2), n is the num-ber of electrons transferred, F is the Faraday constant (C·mol−1),and A is the electrode area (cm2). Fig. 4C shows that there are noredox peaks in the cyclic voltammogram (CV) for the NPC elec-trode from 0.6 to 1.3 V (vs. normal hydrogen electrode [NHE]),while reversible redox waves do appear at E1/2 = 1.05 V (vs. NHE)in the CV curve for the RuPC/NPC electrode (Fig. 4D). The wavesin the CV arise from the RuIII/RuII redox couple at a potentialsimilar to results found previously (27). The amount of Ru poly-pyridyl carbene loaded on the RuPC/NPC electrode is estimatedto be 1.5 ± 0.2 nmol·cm−2 based on peak area measurements of 3electrodes.

Electrocatalytic Reduction of CO2. The activity of RuPC/NPC to-ward electrochemical reduction of CO2 was examined by linear-sweep voltammetry in 0.5 M KHCO3 aqueous solutions (Fig.5A). Its current density in a CO2-saturated solution at potentialsmore negative than −0.62 V (vs. NHE) was notably enhancedcompared to an Ar-saturated solution, implying the reduction ofCO2 at a low onset potential by the RuPC/NPC electrode. Thecatalytic current was stable during 3 scans (SI Appendix, Fig. S2).CO2 reduction by the NPC electrode was also evaluated (SI Ap-pendix, Fig. S3). The net current density for CO2 reduction on NPCwas lower than on the RuPC/NPC under the same conditions,Fig. 1. Schematic illustration illustrating preparation of the RuPC/NPC electrode.

Fig. 2. TEM image (A) and N2 adsorption–desorption isotherm (B) of RuPC/NPC.

26354 | www.pnas.org/cgi/doi/10.1073/pnas.1907740116 Liu et al.

Dow

nloa

ded

by g

uest

on

June

7, 2

021

https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/cgi/doi/10.1073/pnas.1907740116

consistent with enhanced CO2 reduction by the surface-boundRu(II) polypyridyl carbene.To probe for CO2-reduction products, bulk electrolysis was

performed on RuPC/NPC electrodes in CO2-saturated 0.5 MKHCO3 aqueous solutions with potentials from −0.87 to −1.17 V(vs. NHE). The liquid and gas products were detected by gaschromatography and 1H NMR. Ethanol, acetate, methanol, CO,and formate have been identified as CO2-reduction products forRuPC/NPC (SI Appendix, Fig. S4). To confirm that these prod-ucts originated from catalyzed CO2 reduction, bulk electrolysison the RuPC/NPC electrode was performed in Ar-saturated

0.5 M KHCO3 aqueous solutions at −0.97 and −1.07 V (vs.NHE). As expected, there was no evidence for carbon-containingcompounds after 1 h of electrolysis. The 13CO2-reduction ex-periment was performed at −0.97 V (vs. NHE), and the liquidproducts were analyzed by 1H NMR. In the 1H NMR spectrum(SI Appendix, Fig. S5), H–13C signals for ethanol, acetate,methanol, and formate with peak splitting were clearly observed,while H–12C signals were negligible, confirming that the productswere derived from CO2 reduction.Fig. 5B shows the Faradaic efficiency for the products that

appeared after CO2 reduction on the RuPC/NPC electrode for 1 h.In the data, it is notable that the Faradaic efficiency for ethanol(21.0 to 27.5%) is much higher than efficiencies for otherproducts at potentials from −0.87 to −1.07 V (vs. NHE). Forexample, at −0.97 V (vs. NHE), the efficiency for ethanol for-mation was 2.5 to 5.2 times higher than efficiencies for acetate,methanol, CO, and formate. The experimental observationsshow that RuPC/NPC has a high selectivity for ethanol pro-duction at relatively low overpotentials. The total efficiencies forthe multicarbon products, ethanol and acetate, were 31.0 to38.4% at −0.87 to −1.07 V (vs. NHE). When the potential wasshifted negatively from −0.87 to −1.17 V (vs. NHE), both thecurrent densities (SI Appendix, Fig. S6) and efficiencies forethanol and acetate increased initially and then decreased atpotentials more negative than −1.07 V (vs. NHE). Over the samepotential range, the current density and efficiencies for CO in-creased gradually. The electrolysis results show that ethanolproduction is favored on the RuPC/NPC electrode at less neg-ative potentials. Increasing the potential to or beyond −1.17 V(vs. NHE) caused CO to become the major CO2-reductionproduct to give syngas, CO, and H2 (SI Appendix, Fig. S7). Syngasis an important chemical feedstock for synthesizing bulk chemicalsand fuels such as methanol, acetic acid, and others. The molarratio of H2/CO syngas was 2.9:1 at −0.87 V (vs. NHE), whichdecreased gradually to 2:1 as potential shifted negatively from−0.87 to −1.17 V (vs. NHE) (Fig. 5C). However, it increased to2.4:1 at more negative potential (−1.17 V vs. NHR). The potential

Fig. 3. SEM images of NPC (A) and RuPC/NPC (B) and energy-dispersive X-ray spectroscopic maps of N (C) and Ru (D) on RuPC/NPC.

Fig. 4. N 1s XPS spectrum of NPC (A), Ru 3d and C 1s spectrum of RuPC/NPC (B), and cyclic voltammograms of NPC (C) and RuPC/NPC (D) electrodes in 0.1 MTBAPF6/MeCN at a scan rate of 10 mV·s

−1. a.u., arbitrary units.

Liu et al. PNAS | December 26, 2019 | vol. 116 | no. 52 | 26355

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

June

7, 2

021

https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplemental

dependent H2/CO ratios can meet the requirements for down-stream chemical and fuel production (36).The NPC electrode also reduced CO2 to ethanol, acetate,

methanol, formate, and CO, but the product distribution wasdifferent from the RuPC/NPC electrode. Fig. 5D compares C2product efficiencies between NPC and RuPC/NPC. Ethanol ef-ficiency on RuPC/NPC was enhanced by 1.9 to 2.2 times relativeto that on NPC, from −0.87 to −1.07 V (vs. NHE), but there wasno obvious difference in acetate efficiency with or without theRu(II) polypyridyl carbene catalyst. Since CO is the only productof CO2 reduction by the Ru(II) polypyridyl carbene (22, 27, 28),the high ethanol efficiency for RuPC/NPC must arise from thesynergistic effect of Ru(II) polypyridyl carbene and NPC at theinterface.The potential for CO2 reduction on RuPC/NPC is more pos-

itive than for the Ru(II) polypyridyl carbene in solution(−1.20 to −1.50 V vs. NHE) (27, 28) or as heterogenouscatalyst (−0.96 to −1.16 V vs. NHE) (22). Dimerization orprotonation of adsorbed CO have been reported as the mainpathways for C2 production (SI Appendix, Fig. S8) (30, 31, 37).CO2 reduction on Ru(II) polypyridyl carbene occurs throughinitial 2e− transfer to the ligands, followed by reaction with CO2to give [RuII(tpy)(Mebim-py)(COO2−)]0, and 1e−/1H+ reductionto give [RuII(tpy−)(Mebim-py)(COOH)]0. In the subsequent steps,it undergoes further reduction to give [RuII(tpy−)(Mebim-py)(CO)]+

as the intermediate (27, 28). The enhanced ethanol efficiency forRuPC/NPC may arise from that CO intermediates adsorbed at theRu polypyridyl carbene, [RuII(tpy−)(Mebim-py)(CO)]+, aretransformed to C2 products at the RuPC/NPC interface by C–Ccoupling between CO intermediates or intermediates from COprotonation. The porous structure of RuPC/NPC probably canfacilitate C–C coupling reaction via nanoconfinement of CO2 orCO2 reduction intermediates (38, 39). The results highlightedhere point to the strategy, assembling molecular complexes thatare active toward producing C1 intermediates on carbon materials

where C–C coupling can occur, being promising for steering CO2electroreduction toward multicarbon products.Achieving high stability is a significant challenge for heter-

ogenous molecular catalysis. For CO2 reduction by the RuPC/NPC electrode, electrochemical reduction was investigated overa 3-h electrolysis period at −0.97 V (vs. NHE) in 0.5 M KHCO3aqueous solution. During the electrolysis period, the currentdensity for RuPC/NPC was nearly stable except an initial de-crease (break-in period) (Fig. 6A). The Faradaic efficiency formulticarbon products, ethanol, and acetate was 33.1 to 37.3%for a 3-h experiment (Fig. 6B), and the efficiencies for otherCO2-reduction products presented no obvious change, sug-gesting that the RuPC/NPC electrode was stable during 3 h ofCO2 reduction. A Leach test with inductively coupled plasmaatomic emission spectroscopy analysis confirmed its stability(details are in SI Appendix). The RuPC/NPC hybrid catalyst ismore stable than the reported Ru(II) polypyridyl carbene het-erogeneous catalyst, which has a lifetime of ∼15 min under similarconditions (22).

ConclusionsA method is described here for constructing a heterogenousmolecular catalyst that steers electroreduction of CO2 towardC–C-bonded products, notably ethanol. It is based on anchor-ing a Ru(II) polypyridyl carbene complex on NPC. With thesynergistic effects of the Ru(II) polypyridyl carbene catalyst forCO intermediate production and NPC for C–C coupling,electrochemical reduction of CO2 to ethanol occurs with aFaradaic efficiency of 27.5% at relatively low overpotentials.Appearance of ethanol is in competition with a syngas mixtureof H2/CO at a mole ratio of 2.0 to 2.9. The RuPC/NPC elec-trocatalyst is stable toward CO2 reduction for a period of 3 h andadds a promising lead for the reduction of CO2 to multicarbonproducts.

Fig. 5. (A) Linear-sweep voltammograms for RuPC/NPC in Ar- or CO2-saturated 0.5 M KHCO3 aqueous solutions. (B) Faradaic efficiencies for ethanol, acetate,methanol, formate, and CO production on RuPC/NPC over a 1-h period. (C) Mole ratios of H2/CO produced on RuPC/NPC. (D) Faradaic efficiencies for C2products at NPC or RuPC/NPC electrodes over a 1-h period in 0.5 M KHCO3 aqueous solutions.

26356 | www.pnas.org/cgi/doi/10.1073/pnas.1907740116 Liu et al.

Dow

nloa

ded

by g

uest

on

June

7, 2

021

https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplementalhttps://www.pnas.org/cgi/doi/10.1073/pnas.1907740116

Materials and MethodsPreparation of RuPC/NPC Electrode. A total of 60.0 mg of NPC was added into10.0 μL of Nafion (5 wt%) and 490.0 μL of isopropanol. RuPC dissolved at

DMSO-D6 (100 μM, 50 μL) was added into the suspension. After sonicationfor 10 min and stirring for 12 h, the suspension was drop-coated oncarbon-fiber substrate with a catalyst loading of 0.03 g·cm−2. The pre-pared RuPC/NPC electrode was washed by isopropanol to remove Rupolypyridyl carbene, which was not anchored on NPC, and vacuum driedat 80 °C.

Electrochemical Experiments. All of the electrochemical tests were conductedin a 3-electrode system at room temperature. RuPC/NPC or NPC was used asthe working electrode, and Pt foil was used as the counterelectrode. Theamount of Ru polypyridyl carbene anchored on the RuPC/NPC electrode wasmeasured by cyclic voltammograms in Ar saturated in 0.1 M TBAPF6/MeCNwith Ag as reference electrode. The potential was converted to vs. NHE byusing ferrocene for calibration.

All CO2-reduction experiments were tested with Ag/AgCl as referenceelectrode, and the potentials were converted to vs. NHE by using theequation of Evs. NHE = Evs. Ag/AgCl + 0.2 (V). The activity of RuPC/NPC forelectrocatalytic CO2 reduction was examined by linear-sweep voltammo-grams in CO2- or Ar-saturated 0.5 M KHCO3 aqueous electrolyte (scan rate of10 mV·s−1). CO2 bulk electrolysis was performed at −0.87 to approximately−1.17 V (vs. NHE) in a gas-tight H-type 2-chamber cell separated by Nafion117 membrane, which was filled with CO2-saturated 0.5 M KHCO3 aqueoussolution.

Product Analysis. After bulk electrolysis, gas samples were drawn from theheadspace of the gas-tight cell and injected into gas chromatography(Varian 450) equipped with a thermal conductivity detector. The liquidproducts were measured by a 1H NMR spectrum (Bruker B600) and gaschromatography (Shimadzu, catalog no. GC-2010) equipped with a flameionization detector (FID). For 1H NMR spectra, solvent suppression wasapplied for CO2-reduction product analysis to reduce the intensity ofwater peak. The samples were collected into 10% D2O with DMSO as in-ternal standard for quantification. The amount of ethanol produced wasconfirmed by gas chromatography (Shimadzu, catalog no. GC-2010)equipped with an FID detector and DB-Wax column (30 m × 0.25 mm ×0.50 μm).

The Faradaic efficiency (FE) for CO2 reduction was calculated via theequation of FE = nNF/Q, where n is the number of electron transferred forCO2 reduction to products, N is the molar quantity of CO2-reduction prod-ucts (mol), F is the Faraday constant (C·mol−1), and Q is the amount of chargepassed through the cell (C).

Data Availability. All data are included in the main text and SI Appendix.

ACKNOWLEDGMENTS. This work was supported by the US Departmentof Energy, Office of Basic Energy Sciences Award DE-SC0015739; andNational Natural Science Foundation of China Grants 21707016 and 51708085.Y.L. was supported by the State Scholarship Fund from the ChinaScholarship Council.

1. O. S. Bushuyev et al., What should we make with CO2 and how can we make it? Joule2, 825–832 (2018).

2. C. G. Morales-Guio et al., Improved CO2 reduction activity towards C2+ alcohols on atandem gold on copper electrocatalyst. Nat. Catal. 1, 764–771 (2018).

3. R. G. Mariano, K. McKelvey, H. S. White, M. W. Kanan, Selective increase in CO2electroreduction activity at grain-boundary surface terminations. Science 358, 1187–1192 (2017).

4. W. Luo, J. Zhang, M. Li, A. Züttel, Boosting CO production in electrocatalytic CO2reduction on highly porous Zn catalysts. ACS Catal. 9, 3783–3791 (2019).

5. Q. Li et al., Tuning Sn-catalysis for electrochemical reduction of CO2 to CO via thecore/shell Cu/SnO2 structure. J. Am. Chem. Soc. 139, 4290–4293 (2017).

6. Y. Liu, S. Chen, X. Quan, H. Yu, Efficient electrochemical reduction of carbon dioxideto acetate on nitrogen-doped nanodiamond. J. Am. Chem. Soc. 137, 11631–11636(2015).

7. M. D. Sampson, C. P. Kubiak, Manganese electrocatalysts with bulky bipyridine li-gands: Utilizing Lewis acids to promote carbon dioxide reduction at low over-potentials. J. Am. Chem. Soc. 138, 1386–1393 (2016).

8. C. Costentin, S. Drouet, M. Robert, J. M. Savéant, A local proton source enhances CO2electroreduction to CO by a molecular Fe catalyst. Science 338, 90–94 (2012).

9. M. L. Clark, P. L. Cheung, M. Lessio, E. A. Carter, C. P. Kubiak, Kinetic and mechanisticeffects of bipyridine (bpy) substituent, labile ligand, and brønsted acid on electro-catalytic CO2 reduction by Re(bpy) complexes. ACS Catal. 8, 2021–2029 (2018).

10. S. Lin et al., Covalent organic frameworks comprising cobalt porphyrins for catalyticCO2 reduction in water. Science 349, 1208–1213 (2015).

11. F. Lei et al., Metallic tin quantum sheets confined in graphene toward high-efficiencycarbon dioxide electroreduction. Nat. Commun. 7, 12697 (2016).

12. M. Liu et al., Enhanced electrocatalytic CO2 reduction via field-induced reagentconcentration. Nature 537, 382–386 (2016).

13. A. S. Hall, Y. Yoon, A. Wuttig, Y. Surendranath, Mesostructure-induced selectivity inCO2 reduction catalysis. J. Am. Chem. Soc. 137, 14834–14837 (2015).

14. S. Gao et al., Partially oxidized atomic cobalt layers for carbon dioxide electro-reduction to liquid fuel. Nature 529, 68–71 (2016).

15. Y. Zhou et al., Dopant-induced electron localization drives CO2 reduction to C2 hy-drocarbons. Nat. Chem. 10, 974–980 (2018).

16. T. T. H. Hoang et al., Nanoporous copper-silver alloys by additive-controlled electro-deposition for the selective electroreduction of CO2 to ethylene and ethanol. J. Am.Chem. Soc. 140, 5791–5797 (2018).

17. C. Zhao et al., Ionic exchange of metal-organic frameworks to access single nickel sitesfor efficient electroreduction of CO2. J. Am. Chem. Soc. 139, 8078–8081 (2017).

18. K. Jiang et al., Isolated Ni single atoms in graphene nanosheets for high-performanceCO2 reduction. Energy Environ. Sci. 11, 893–903 (2018).

19. I. Hod et al., Fe-porphyrin-based metal-organic framework films as high-surfaceconcentration, heterogeneous catalysts for electrochemical reduction of CO2. ACSCatal. 5, 6302–6309 (2015).

20. Z. Weng et al., Electrochemical CO2 reduction to hydrocarbons on a heterogeneousmolecular Cu catalyst in aqueous solution. J. Am. Chem. Soc. 138, 8076–8079 (2016).

21. N. Han et al., Supported cobalt polyphthalocyanine for high-performance electro-catalytic CO2 reduction. Chem 3, 652–664 (2017).

Fig. 6. Current density (A) and Faradaic efficiencies (B) for CO2 reduction onRuPC/NPC over a 3-h period in 0.5 M KHCO3 aqueous solutions (−0.97 V vs. NHE).

Liu et al. PNAS | December 26, 2019 | vol. 116 | no. 52 | 26357

CHEM

ISTR

Y

Dow

nloa

ded

by g

uest

on

June

7, 2

021

https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.1907740116/-/DCSupplemental

22. Y. Wang, S. L. Marquard, D. Wang, C. Dares, T. J. Meyer, Single-site, heterogeneous

electrocatalytic reduction of CO2 in water as the solvent. ACS Energy Lett. 2, 1395–1399 (2017).

23. J. Choi et al., A porphyrin/graphene framework: A highly efficient and robust electro-

catalyst for carbon dioxide reduction. Adv. Energy Mater. 8, 1801280 (2018).24. X. Zhang et al., Highly selective and active CO2 reduction electrocatalysts based on

cobalt phthalocyanine/carbon nanotube hybrid structures. Nat. Commun. 8, 14675(2017).

25. B. Reuillard et al., Tuning product selectivity for aqueous CO2 reduction with a

Mn(bipyridine)-pyrene catalyst immobilized on a carbon nanotube electrode. J. Am.Chem. Soc. 139, 14425–14435 (2017).

26. A. S. Varela, W. Ju, P. Strasser, Molecular nitrogen-carbon catalysts, solid metal or-ganic framework catalysts, and solid metal/nitrogen-doped carbon (MNC) catalysts

for the electrochemical CO2 reduction. Adv. Energy Mater. 8, 1703614 (2018).27. Z. Chen et al., Splitting CO2 into CO and O2 by a single catalyst. Proc. Natl. Acad. Sci.

U.S.A. 109, 15606–15611 (2012).28. P. Kang, Z. Chen, A. Nayak, S. Zhang, T. J. Meyer, Single catalyst electrocatalytic re-

duction of CO2 in water to H2+CO syngas mixtures with water oxidation to O2. EnergyEnviron. Sci. 7, 4007–4012 (2014).

29. H. Xiao, T. Cheng, W. A. Goddard, 3rd, Atomistic mechanisms underlying selectivitiesin C1 and C2 products from electrochemical reduction of CO on Cu(111). J. Am. Chem.

Soc. 139, 130–136 (2017).

30. W. Luo, X. Nie, M. J. Janik, A. Asthagiri, Facet dependence of CO2 reduction paths onCu electrodes. ACS Catal. 6, 219–229 (2015).

31. R. Kortlever, J. Shen, K. J. Schouten, F. Calle-Vallejo, M. T. Koper, Catalysts and re-action pathways for the electrochemical reduction of carbon dioxide. J. Phys. Chem.Lett. 6, 4073–4082 (2015).

32. J. Wu et al., A metal-free electrocatalyst for carbon dioxide reduction to multi-carbonhydrocarbons and oxygenates. Nat. Commun. 7, 13869 (2016).

33. J. J. Concepcion et al., Catalytic water oxidation by single-site ruthenium catalysts.Inorg. Chem. 49, 1277–1279 (2010).

34. J. Wei et al., A controllable synthesis of rich nitrogen-doped ordered mesoporouscarbon for CO2 capture and supercapacitors. Adv. Funct. Mater. 23, 2322–2328 (2013).

35. V. Q. Nguyen et al., Unprecedented self-organized monolayer of a Ru(II) complex bydiazonium electroreduction. J. Am. Chem. Soc. 138, 9381–9384 (2016).

36. D. Pakhare, J. Spivey, A review of dry (CO2) reforming of methane over noble metalcatalysts. Chem. Soc. Rev. 43, 7813–7837 (2014).

37. T. T. Zhuang et al., Steering post-C-C coupling selectivity enables high efficiencyelectroreduction of carbon dioxide to multi-carbon alcohols. Nat. Catal. 1, 421–428(2018).

38. T. T. Zhuang et al., Copper nanocavities confine intermediates for efficient electro-synthesis of C3 alcohol fuels from carbon monoxide. Nat. Catal. 1, 946–951 (2018).

39. C. Zhang et al., Capillary effect-enabled water electrolysis for enhanced electro-chemical ozone production by using bulk porous electrode. J. Am. Chem. Soc. 139,16620–16629 (2017).

26358 | www.pnas.org/cgi/doi/10.1073/pnas.1907740116 Liu et al.

Dow

nloa

ded

by g

uest

on

June

7, 2

021

https://www.pnas.org/cgi/doi/10.1073/pnas.1907740116