This is the submitted version of the journal article: Li, Jingkun; Zitolo, Andrea; Garcés-Pineda, Felipe A.; [et al.]. «Metal oxide clusters on nitrogen-doped carbon are highly selective for CO2 electroreduction to CO». ACS Catalysis, Vol. 11, issue 15 (August 2021), p. 10028-10042. DOI 10.1021/acscatal.1c01702 This version is available at https://ddd.uab.cat/record/271930 under the terms of the license Page 1 of 42 ACS Catalysis 1 2 3 Metal-oxide Clusters on Nitrogen-doped Carbon are highly Selective for CO2 Electroreduction 4 5 6 to CO 7 8 9 10 Jingkun Li1, Andrea Zitolo2, Felipe A. Garcés-Pineda3, Tristan Asset4, Mounika Kodali4, 11 12 13 PengYi Tang5, Jordi Arbiol,5,6 José Ramón Galán-Mascarós3,6, Plamen Atanassov4, Iryna V. 14 15 Zenyuk4, Moulay Tahar Sougrati1, and Frédéric Jaouen1* 16 17 18 19 1 ICGM, Univ. Montpellier, CNRS, ENSCM, 34090 Montpellier, France 20 21 22 2 Synchrotron SOLEIL, L’orme des Merisiers, BP 48 Saint Aubin, 91192 Gif-sur-Yvette, 23 24 France 25 26 3 Institute of Chemical Research of Catalonia, ICIQ, The Barcelona Institute of Science and 27 28 29 Technology, Av. Països Catalans, 16, 43007 Tarragona, Spain 30 31 4 Department of Chemical and Biomolecular Engineering, National Fuel Cell Research Center, 32 33 University of California Irvine, Irvine, 92697, United States 34 35 36 5 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and BIST, Campus 37 38 UAB, Bellaterra, 08193 Barcelona, Catalonia, Spain 39 40 6 ICREA, Pg. Lluís Companys 23, 08010 Barcelona, Catalonia, Spain 41 42 43 44 45 * Corresponding author: Frédéric Jaouen (frederic.jaouen@umontpellier.fr) 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1 ACS Paragon Plus Environment ACS Catalysis Page 2 of 42 1 2 3 Abstract 4 5 6 7 The electrochemical reduction of CO2 (eCO2RR) using renewable energy is an effective 8 9 approach to pursue carbon neutrality. The eCO2RR to CO is indispensable to promoting C-C 10 11 coupling through bifunctional catalysis and to achieving cascade conversion from CO2 to C2+. 12 13 14 This work investigates a series of M/N-C (M=Mn, Fe, Co, Ni, Cu and Zn) catalysts for which 15 16 the metal precursor interacted with the nitrogen-doped carbon support (N-C) at room 17 18 temperature, resulting in the metal being present as (sub)nano-sized metal oxide clusters in ex 19 20 situ condition, except for Cu/N-C and Zn/N-C. A volcano trend in their activity toward CO as 21 22 23 a function of the group of the transition metal is revealed, with Co/N-C exhibiting the highest 24 25 activity at –0.5 V vs. RHE; while Ni/N-C shows both appreciable activity and selectivity. 26 27 Operando XAS shows that the majority of Cu atoms in Cu/N-C forms Cu0 clusters during 28 29 30 eCO2RR; while the Mn/, Fe/, Co/, and Ni/N-C maintain the metal hydroxide structures, with a 31 32 minor amount of M0 formed in Fe/, Co/, and Ni/N-C. The superior activity of Fe/, Co/ and 33 34 Ni/N-C is ascribed to the phase contraction and the HCO3– insertion into the layered structure 35 36 37 of metal hydroxides. Our work provides a facile synthetic approach toward highly active and 38 39 selective electrocatalysts to convert CO2 into CO. Coupling with state-of-the-art NiFe-based 40 41 anodes in a full cell device, Ni/N-C exhibits >80% Faradaic efficiency toward CO at 100 mA 42 43 cm–2. 44 45 46 47 Keywords: CO2 electroreduction, metal oxide clusters, volcano trend, operando X-ray 48 49 absorption spectroscopy, phase contraction, nitrogen-doped carbon 50 51 52 53 54 55 56 57 58 59 60 2 ACS Paragon Plus Environment Page 3 of 42 ACS Catalysis 1 2 3 4 1. Introduction 5 6 7 The direct electrochemical CO2 reduction reaction (eCO2RR) to fuels and valuable chemicals, 8 9 powered by renewable energies, is an attractive pathway to establish a carbon-neutral energy 10 11 12 cycle. However, the implementation of CO2 electrolyzers is still impeded by: 1) low energy 13 14 efficiency due to the high overpotential needed to activate CO2;[1] 2) low Faradaic efficiency 15 16 (FE) due to competing hydrogen evolution reaction (HER) in aqueous electrolytes;[1] 3) low 17 18 19 current densities due to poor CO2 mass transport when using CO2-saturated aqueous 20 21 electrolytes as cathode feed.[2, 3] In the past decades, research in the field of eCO2RR has 22 23 focused on developing efficient and selective electrocatalysts via surface functionalization,[4, 5] 24 25 metal alloying,[6, 7] heteroatom doping,[8] facet/morphology regulation,[9, 10] and particle size 26 27 28 tailoring[11]. Recently, industrial scale current densities (>100 mA·cm–2) have been achieved 29 30 via enhanced mass transport of CO2 in gas-diffusion electrodes (GDEs).[12-16] 31 32 33 Depending on the nature of catalytic sites and the operating conditions, C1 (CO, CH4, CH3OH, 34 35 36 HCOOH, etc.), C2 (C2H4, CH3CH2OH, CH3COOH, etc.) and C3 (n-C3H7OH, etc.) products 37 38 have been reported for eCO2RR.[1] The state-of-the-art Faradaic efficiency (FE) of eCO2RR to 39 40 C1 products (CO or formate) is >90% over Zn,[17] Ag,[18] Au,[19] SnOx[20] and catalysts 41 42 43 comprising single-metal-atom sites.[21-25] However, only Cu-based catalysts promote the 44 45 production of C2+ chemicals but, hitherto, with low FEs (<50%) toward a particular C2+ 46 47 product.[6, 8-10] Only few studies have, to the best of our knowledge, reported the reduction of 48 49 CO2 toward a C2 product (acetic acid) over Cu-free catalysts and with significant selectivity.[26- 50 51 28] To date, the most efficient approach to favour C2+ products vs. CO, CH4 and formic acid is 52 53 54 the fine tuning of the surface structure of Cu, including the oxidation state, defects, crystal 55 56 facets and doping with a secondary metal.[6, 8-10, 29-32] The doping of Cu with Ag, Au and Zn 57 58 59 leans, partially, on the high selectivity for CO2-to-CO of such metallic surfaces and the 60 3 ACS Paragon Plus Environment ACS Catalysis Page 4 of 42 1 2 3 recognized importance of the CO intermediate in the formation of C2 products on Cu.[33-37]. 4 5 6 Doping Cu with Ag, Au and Zn can thus be described as introducing bifunctionality at the 7 8 atomic scale. However, the extent to which Cu can be doped by CO-selective sites is limited,[35, 9 10 36] and the doping simultaneously alters the geometric and electronic structures of the Cu 11 12 13 surface, modifying its selectivity.[6, 38-41] In addition, such advanced surface structures prepared 14 15 and optimized ex situ generally experience significant reconstructions during eCO2RR, either 16 17 due to the low electrochemical potential and/or due to operando produced CO.[42-44] 18 19 20 Recently, alternative bifunctional catalysis approaches have been investigated in which CO is 21 22 23 produced on one type of active sites and subsequently desorbed and transported to other sites 24 25 that catalyse C-C coupling.[4, 5, 36, 45] Compared to atomic-level bifunctionality, this approach 26 27 disentangles the effect of the two catalytic sites and allows optimizing them separately. For 28 29 [36] 30 example, E.H. Sargent’s group et al. demonstrated that metallic copper/metal-porphyrin 31 32 composite cathodes promote the C-C coupling, leading to higher peak Faradaic efficiency (FE) 33 34 of 41% toward ethanol compared to that of 29% over pure Cu. Similarly, P. Strasser’s group 35 36 37 reported that the ethylene production rate was 3-fold faster on Cu nanoparticles mixed with Ni- 38 39 N-C than on Cu nanoparticles alone.[45] 40 41 42 These examples show that CO-making sites do not need to be at atomic distance of the Cu 43 44 surface in order to enhance Cu selectivity towards C2 products. This opens the door to diverse 45 46 47 strategies at the cathode, electrolyzer and system levels in order to tune the product selectivity. 48 49 For example, CO/CO2 feed was shown to increase the selectivity towards ethylene compared 50 51 to pure CO2 feed.[45] The eCO2RR towards multi-carbon products might thus be achieved in 52 53 54 two stages from a cell engineering viewpoint as well, for example with a first electrolyzer cell 55 56 with a cathode designed to produce syngas from CO2, and the syngas serving as inlet gas for a 57 58 second electrolyzer cell where a Cu-based cathode transforms it selectively into C2 products.[46] 59 60 4 ACS Paragon Plus Environment Page 5 of 42 ACS Catalysis 1 2 3 In addition, complete electrochemical conversion of CO2-to-CO in the first stage would be 4 5 6 beneficial as it would allow carrying out electrochemical reduction of CO to C2+ products at 7 8 high pH (13-14) in the second stage, with more facile kinetics than in near-neutral pH.[2, 47] In 9 10 contrast, sustainably operating eCO2RR at high pH is impossible, due to the fast carbonation 11 12 13 of high-pH catholyte with CO2-gas feed.[2, 3] 14 15 16 Thus, it is of paramount importance to develop a robust, efficient and selective electrocatalyst 17 18 for eCO2RR to CO to realize the practical implementation of CO2 electrolyzers. Metal- 19 20 nitrogen-carbon (M-N-C, with M=Mn, Fe, Co, Ni, and Cu) materials that exclusively comprise 21 22 23 atomically-dispersed MNx sites have shown high selectivity toward CO (especially for M = Fe 24 25 or Ni).[21-25] Excessive metal content however leads to aggregation of a fraction of metal during 26 27 pyrolysis into particles, resulting in lower selectivity towards CO.[48-50] The synthesis of M- 28 29 30 N-C catalysts comprising only MNx sites is therefore not trivial and requires optimisation of 31 32 the content and nature of the metal, N and C precursors, as well as one or several pyrolysis 33 34 steps and, often, post-pyrolysis treatment to remove excess metal.[22-24, 51] This has drawback 35 36 37 on the life-cycle assessment if such catalysts were implemented in industrial-scale CO2 38 39 electrolyzers. MN4 macrocycles such as phthalocyanines and porphyrins adsorbed or attached 40 41 onto carbon allotropes have shown high selectivity for CO formation, comparable to that of M- 42 43 N-C materials.[52-54] However, the synthesis of such macrocycles is not ideal for large-scale 44 45 46 production of low-cost catalyst. Moreover, it has been reported that both MNx sites (in 47 48 pyrolyzed M-N-C) and MN4 sites (in macrocycles supported e.g. on carbon) are prone to 49 50 clustering during eCO2RR, as observed via operando X-ray absorption spectroscopy (XAS).[25, 51 52 55, 56] 53 While the Cu clusters formed at low potential re-dispersed into single-Cu-atom sites when 54 55 scanning the potential up, this observation raises the question whether the fine structure of M- 56 57 N-C materials with exclusive single-atom MNx sites is necessary for high CO2-to-CO 58 59 60 selectivity. 5 ACS Paragon Plus Environment ACS Catalysis Page 6 of 42 1 2 3 Herein, we developed and studied a series of M/N-C (M=Mn, Fe, Co, Ni, Cu and Zn) catalysts 4 5 6 featuring highly-dispersed metal-oxide particles supported on nitrogen-doped carbon (N-C) 7 8 under ex situ condition, except for Cu/N-C and Zn/N-C, featuring mainly atomically dispersed 9 10 sites. Their catalytic activity and selectivity toward eCO2RR to CO was evaluated in a flow 11 12 13 cell with GDEs. A volcano trend in their activity toward CO as a function of the group of the 14 15 transition metal is revealed, with Co/N-C exhibiting the highest intrinsic activity at –0.5 V vs. 16 17 RHE; while Ni/N-C shows both high activity and selectivity. Operando XAS was further 18 19 exploited to investigate structural changes under working conditions. The majority of Cu atoms 20 21 22 in Cu/N-C forms Cu0 clusters during eCO2RR; while the Mn/, Fe/, Co/, and Ni/N-C maintain 23 24 the metal hydroxide structures, with a minor amount of M0 formed in operando with Fe/, Co/, 25 26 and Ni/NC samples. Then the M/N-C cathodes were coupled with state-of-the-art NiFe-based 27 28 29 anodes in PGM-free electrolyzer cells, with Ni/N-C exhibiting >80% FE toward CO at 100 mA 30 31 cm–2 at a cell voltage of ~4 V. 32 33 34 35 2. Results and discussion 36 37 38 2.1. Ex situ identification of the metal coordination in M/N-C 39 40 41 The metal coordination and presence of metal-oxide particles supported on N-C, is firstly 42 43 44 studied and revealed under ex situ condition for the series of materials developed in this work 45 46 and labelled as M/N-C (M=Mn, Fe, Co, Ni, Cu and Zn). The metal coordination and spectral 47 48 response is also compared with M-N-C catalysts previously studied at our laboratory, and 49 50 featuring exclusively atomically dispersed MNx sites. The synthesis of M-N-C catalysts was 51 52 53 described in a previous work.[25] Briefly, the well-mixed precursors of metal acetate, 1,10- 54 55 phenanthroline and ZIF-8 (Basolite® Z1200) were pyrolyzed in Ar at 1050 oC in flash mode. 56 57 In contrast, the M/N-C catalysts (M = Mn, Fe, Co, Ni, Cu) were synthesized via chemical 58 59 60 precipitation wherein the transition metals were deposited on N-C by dissolving M(II)Cl2 salt 6 ACS Paragon Plus Environment Page 7 of 42 ACS Catalysis 1 2 3 in a suspension of N-C in water/ethanol (see Experimental Methods for the detailed protocol). 4 5 6 The N-C support was prepared via pyrolysis of ZIF-8, as described further below. Then the 7 8 suspension was centrifuged, washed with water and let to dry in ambient conditions. The dry 9 10 powder was finally subjected to a mild heat treatment at 200°C for 2 h in flowing Ar, for 11 12 13 complete drying. The major difference between the synthesis of M-N-C and M/N-C is that the 14 15 metal atoms were subjected to a high-temperature pyrolysis for the former synthesis, leading 16 17 to the formation of atomically dispersed M-N4 sites, but not for the latter synthesis. The N-C 18 19 support was synthesized by pyrolysis of ZIF-8, with a first pyrolysis in flash mode in flowing 20 21 22 Ar at 1050°C for 1 h, followed by a second flash pyrolysis in flowing NH3 at 900°C for 10 23 24 min. The second pyrolysis was introduced to increase the N-content and surface basicity. As 25 26 will be shown later, the presence of nitrogen in the carbon substrate is key to the CO2RR 27 28 29 activity of the present M/N-C materials. It should be noted here that the N-C support contains 30 31 Zn-N4 sites (Figure S1). However, we label it as N-C for simplicity and also because the Zn- 32 33 N4 sites were previously demonstrated to be inactive towards CO2RR.[25] 34 35 36 37 The bulk metal contents in M/N-C materials were derived from the absolute absorption height 38 39 of XAS spectra, measured in transmission mode and for a fixed catalyst weight per cross 40 41 section of pellet. They are shown in Table S1, with metal loadings varying between 1.3 and 42 43 2.5 wt.%. This agrees with the metal contents obtained by X-ray photoelectron spectroscopy 44 45 46 (XPS) in the range of 0.5 to 2.2 wt.% (Table S2). All M/N-C materials also contain a significant 47 48 amount of Zn, derived from ZIF-8, that is comparable to the Zn amount in the N-C substrate 49 50 (Table S2). The Zn is present as atomically dispersed ZnN4 moieties, as shown by X-ray 51 52 53 absorption near edge spectroscopy (XANES) and extended X-ray absorption fine structure 54 55 (EXAFS) (Figure S1). The same ZnN4 site and similar Zn content were also observed after a 56 57 single pyrolysis in Ar of ZIF-8.[25] 58 59 60 7 ACS Paragon Plus Environment ACS Catalysis Page 8 of 42 1 2 3 The X-ray diffraction (XRD) patterns of M/N-C materials (Figure S2) exhibit only two broad 4 5 6 peaks typical for nanometric graphite-like domains present in amorphous N-C, and no 7 8 diffraction peaks related to metal-based crystalline structures. XRD is however not ideally 9 10 suited to identify the possible presence of metal clusters in such materials, due to i) the low 11 12 13 metal content on N-C (Tables S1-S2), ii) the small size of metal clusters, or iii) the amorphous 14 15 character of metal clusters due to the mild synthesis conditions. To obtain precise information 16 17 on the metal coordination and oxidation state in M/N-C, we then resorted to 57Fe Mössbauer 18 19 spectroscopy for Fe/N-C, and XAS at the metal K-edge for all M/N-C. 20 21 22 23 The Mössbauer spectrum acquired at room temperature shows only one quadrupole doublet 24 25 (Figure 1a), with isomer shift and quadrupole splitting values of 0.31 and 0.81 mm·s–1, 26 27 respectively (Table S3). While this doublet can generally be assigned to Fe3+ in high-spin (HS) 28 29 30 state,[57] Mössbauer measurement at room temperature cannot distinguish between HS Fe3+ in 31 32 amorphous or (sub)-nanometric Fe2O3 and HS Fe3+ in atomically-dispersed sites.[58, 59] In the 33 34 context of this study, Fe sites in Fe/N-C might exist as a result of possible trans-metalation 35 36 37 from ZnN4 moieties (present in N-C) to FeN4 moieties during the precipitation of Fe onto N-C 38 39 in aqueous solution,[60, 61] or as a result of Fe2+ adsorption on N- or O-functional groups on the 40 41 surface of N-C, followed by oxidation during the drying step in air. 42 43 44 57Fe We then performed Mössbauer spectroscopy at 5 K, taking advantage of the different 45 46 47 temperature-dependent behaviours of the Mössbauer spectrum of nanometric ferric oxides and 48 49 atomically-dispersed Fe-sites, the former changing from doublet to sextet signal below a 50 51 certain critical temperature,[62] while the latter remains a doublet even at 5 K.[59, 63] The 5 K 52 53 54 Mössbauer spectrum of Fe/N-C exhibits three sextets with slightly different IS-values and 55 56 hyperfine fields (HF) amounting to 96 % of the resonant area, as well as a minute amount of a 57 58 quadrupole doublet (4 %) (Figure 1b and Table S3). The IS- and HF values of the three sextets 59 60 8 ACS Paragon Plus Environment Page 9 of 42 ACS Catalysis 1 2 3 can be assigned to Fe2O3 with different particle sizes,[62, 64] while the doublet is assigned to HS 4 5 6 Fe2+, due to its high isomer shift of 1.07 mm·s–1.[65] Thus, the 5 K Mössbauer spectrum 7 8 demonstrates that the well-defined narrow doublet observed at room temperature is 9 10 overwhelmingly arising from amorphous or (sub)nano-sized Fe2O3. 11 12 13 1.000 14 1.000 15 0.995 (a) (b) 0.995 16 0.990 17 Transmission 18 0.985 0.990 19 0.980 20 0.975 0.985 21 0.970 22 0.980 0.965 Sextet 1 Sextet 2 Sextet 3 23 Doublet exp. fitted 24 0.960 0.975 25 -10 -5 0 5 10 -10 -5 0 5 10 26 Velocity (mm/s) Velocity (mm/s) 27 28 29 Figure 1. Ex situ 57Fe Mössbauer spectra for Fe/N-C measured at room temperature (a) and 5 30 31 32 K (b). 33 34 35 This is also well supported by ex situ XAS at the Fe K-edge. The XANES spectrum of Fe/N- 36 37 C highly resembles that of a commercial Fe2O3 powder as far as the edge position and the main 38 39 40 spectral features are concerned (Figure S3a). Moreover, the Fourier transform (FT) of the 41 42 EXAFS spectrum of Fe/N-C shows two prominent peaks at ~1.5 and ~2.5 Å (uncorrected for 43 44 phase shift), matching the Fe-O and Fe-Fe (labelled Fe-Feoxy) interactions in Fe2O3 (Figure 45 46 2a). Similar to our synthesis for Fe/N-C, R. Arrigo et al. deposited iron on a N-functionalized 47 48 49 carbon paper from nitrate salt and identified with XAS the iron species to be 50 51 oxides/oxyhydroxides.[26] We also note that the XANES and EXAFS spectra of Fe/N-C are 52 53 different from those of a Fe-N-C material from our laboratory, prepared by the direct pyrolysis 54 55 56 of Fe(II) salt, ZIF-8 and phenanthroline, comprising Fe exclusively as atomically-dispersed 57 58 FeNx sites that form at high temperature. The XANES spectra of Fe/N-C and Fe-N-C differ in 59 60 shape with a lower white line intensity, scaling with Fe-O interactions, for the latter (Figure 9 ACS Paragon Plus Environment ACS Catalysis Page 10 of 42 1 2 3 S3a). The FT-EXAFS spectra of Fe/N-C and Fe-N-C also differ, the latter only shows one 4 5 6 major peak at ~1.5 Å due to Fe-N (and Fe-O) interactions from (O2)-FeNx moieties,[66, 67] but 7 8 no peak at 2.5 Å (Figure 2a). 9 10 11 To investigate the metal coordination in the other M/N-C catalysts, we performed ex situ XAS 12 13 14 at their respective metal K-edges and compare spectral features to those of i) reference metal 15 16 oxide materials and ii) a series of M-N-C materials prepared at high temperature and 17 18 comprising only atomically-dispersed MNx sites.[25] The intensity of the white line of Co/N-C 19 20 XANES is consistent with that of the oxides, and the position of the edge is close to that of 21 22 23 Co2O3 (Figure S3b). The Ni/N-C XANES spectrum matches well the reference NiO spectrum 24 25 over a broad energy range (Figure S3c). Moreover, the position of the absorption edges 26 27 indicates that the average oxidation state of Ni in Ni/N-C is close to that of Ni in NiO (Ni2+). 28 29 30 The main peak in the FT-EXAFS spectra of Co/N-C at ~1.4 Å matches well the position of the 31 32 Co-O backscattering signal in Co3O4 while the second peak at ~2.3 Å matches the position of 33 34 the Co-Cooxy backscattering signal in Co3O4 and/or CoO (Figure 2b). This is even clearer for 35 36 37 Ni/N-C, with the positions of the two main peaks in its FT-EXAFS spectrum at ~1.5 and ~2.6 38 39 Å matching perfectly those for the Ni-O and Ni-Nioxy backscattering paths of NiO (Figure 2c). 40 41 Similarly as for Fe/N-C, the XANES and FT-EXAFS spectra of Co/N-C and Ni/N-C differ 42 43 from the corresponding spectra of Co-N-C and Ni-N-C materials from our laboratory, that 44 45 46 comprise only single-metal-atom MN4 moieties.[25] In the FT-EXAFS, the Co-N-C and Ni-N- 47 48 C reference materials show only one peak with slightly lower bond distance compared to the 49 50 main peak in M/N-C materials (1.3 vs. 1.4–1.5 Å), assigned to M-N bonds (Figure 2b-c, 51 52 53 compare red and blue curves). In the XANES spectra, the intensity of the white-line peak is 54 55 much weaker for Co-N-C and Ni-N-C compared to Co/N-C and Ni/N-C, respectively, 56 57 indicating a lower fraction of M-O bonds (Figure S3b-c, compare red and blue curves). In 58 59 60 10 ACS Paragon Plus Environment Page 11 of 42 ACS Catalysis 1 2 3 summary, the ex situ XAS spectra of Co/N-C and Ni/N-C demonstrate that the majority of 4 5 6 transition metals are in the form of amorphous or (sub)nanometric oxides. 7 8 9 Regarding Cu/N-C, its XANES spectrum is dissimilar with those of CuO and Cu2O but 10 11 resembles more that of Cu-N-C (Figure S3d). Similarly, the FT-EXAFS spectrum of Cu/N-C 12 13 14 closely resembles that of Cu-N-C, with a single peak at ~1.4 Å and no secondary peak at the 15 16 position expected for Cu-Cuoxy backscattering (Figure 2d). This suggests that Cu might be 17 18 atomically dispersed as CuOx or CuNx sites. The latter label does however not necessarily 19 20 imply the exact same coordination as Cu in Cu-N-C (CuN4, formed at high temperature), and 21 22 23 we stress that the XANES spectra of Cu/N-C and Cu-N-C are slightly different (Figure S3d). 24 25 Transmetalation between ZnNx and Cu2+ in solution, forming CuNx sites, may be more facile 26 27 than with Fe2+, Co2+ or Ni2+ cations in solution due to steric or electronic effects, as reported 28 29 30 for metalation/transmetalation of metal-free porphyrins.[68] 31 32 33 1.0 34 1.4 (a) Fe-N/O (b) Co-N/O Fe-Fe Fe/N-C Co-Cooxy Co/N-C 35 1.2 Fe-Feoxy Fe-N-C Co-N-C 0.8 36 FeO CoO 37 1.0 Fe2O3 Co3O4 0.6 38 Co foil/5 |(R)| (A -3) |(R)| (A -3) o 0.8 Fe foil/3 o 39 0.6 0.4 40 41 0.4 42 0.2 0.2 43 44 0.0 0.0 0 1 2 3 4 5 6 0 1 2 3 4 5 6 45 46 2.5 1.2 (c) Ni-Nioxy (d) Cu-N/O Cu/N-C 47 Ni/N-C Cu-N-C 1.0 48 2.0 Ni-N-C Cu2O 49 Ni-N/O NiO CuO 0.8 |(R)| (A -3) |(R)| (A -3) 50 o 1.5 Ni foil/2 o Ni-N Cu foil/2 51 Cu-Cuoxy 0.6 52 1.0 53 0.4 54 0.5 55 0.2 56 0.0 0.0 57 0 1 2 3 4 5 6 0 1 2 3 4 5 6 58 o R (A) o R (A) 59 60 11 ACS Paragon Plus Environment ACS Catalysis Page 12 of 42 1 2 3 Figure 2. Non phase-shift corrected ex situ FT-EXAFS spectra of Fe/N-C (a), Co/N-C (b), 4 5 6 Ni/N-C (c) and Cu/N-C (d). For each metal, the FT-EXAFS spectra for metallic foils, metal 7 8 oxide powders and M-N-C powder catalysts with MN4 moieties are also presented as 9 10 references. 11 12 13 14 To further study the morphologies of the metal-rich clusters in this series of M/N-C catalysts, 15 16 we then resorted to aberration-corrected scanning transmission electron microscopy (STEM). 17 18 Metal-rich nanoparticles with irregular shapes are identified in Fe/, Co/ and Ni/N-C samples 19 20 (Figure 3a, c, e), in line with the M-Moxy interactions detected in EXAFS; whereas areas 21 22 23 without any obvious clusters/nanoparticles are also observed (Figure 3b, d, f, h). Notably, 24 25 Cu/N-C shows very few metal-rich clusters even after an extensive investigation on multiple 26 27 areas (Figure 3g), explaining the absence of Cu-Cuoxy interaction in the FT-EXAFS spectra 28 29 30 (Figure 2d). Moreover, single-metal-atoms are clearly seen in M/N-C (Figure S4); however, 31 32 we cannot conclusively say they are Co, Ni or Cu single-atom-sites, since the N-C support 33 34 contains a non-negligible amount of atomically dispersed ZnN4 moieties (Figure S1), and the 35 36 37 HAADF STEM cannot distinguish Zn from other metals. In summary, the transition metals in 38 39 M/N-C are present mainly as clusters of oxides, except for the case of Cu/N-C with CuNx or 40 41 CuOx moieties formed via transmetalation or adsorption. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 12 ACS Paragon Plus Environment Page 13 of 42 ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Figure 3. HAADF STEM images of Fe/N-C (a, b), Co/N-C (c, d), Ni/N-C (e, f) and Cu/N-C 26 27 (g, h). 28 29 30 The nitrogen speciation of the M/N-C catalysts was then investigated with XPS. The N1s 31 32 33 narrow-scan spectra were deconvoluted into five nitrogen species, namely pyridinic nitrogen, 34 35 amine or N in MNx, pyrrolic N (labelled N-H), graphitic N (including Ngr/N+ and bulk N-H, 36 37 i.e. N-H in the carbon plane) and N-Ox (Figure S5). The nitrogen chemistry of M/N-C catalysts 38 39 40 resembles that of the N-C substrate, with similar relative concentrations of different N species, 41 42 including in particular 15.8–18.9 relative % of N as “MNx” (Table S2). 43 44 45 As for the EXAFS and XANES data, we find it useful to compare the XPS N1s narrow-scan 46 47 spectra of M/N-C materials to those for M-N-C. Since the N1s spectra for M-N-C are also 48 49 50 similar for different 3d metals,[25] we show and discuss the N1s spectrum of Fe-N-C only as a 51 52 representative example. The N1s spectrum of Fe-N-C was fitted with the same N-species 53 54 (Figure S5f) but contains different relative concentrations of N-species compared to M/N-C 55 56 57 (Table S2). Specifically, the ratios of N-Pyridinic to MNx are 1.9–2.3 for M/N-C materials and 58 59 the N-C substrate as well, but only ~1.6 for Fe-N-C. The latter might indicate that pyridinic N 60 13 ACS Paragon Plus Environment ACS Catalysis Page 14 of 42 1 2 3 are preferentially utilized for forming FeNx sites during pyrolysis in the presence of Fe. In 4 5 6 contrast, similar N-Pyridinic/M-Nx ratios for all M/N-C materials and for the N-C substrate 7 8 suggests that pyridinic nitrogen did not bind metal cations during the metal salt precipitation, 9 10 and that the MNx signal in M/N-C can mainly be assigned to ZnNx sites that were already 11 12 13 present in the N-C substrate. Transmetalation of a fraction of ZnNx sites into CuNx sites would 14 15 also lead to a ratio of N-Pyridinic/M-Nx similar to that in N-C, and the XPS results are thus not 16 17 contradicting the XANES and EXAFS conclusions on Cu/N-C. 18 19 20 We also synthesized the Mn/ and Zn/N-C catalysts and their structures were studied with ex 21 22 23 situ XAS and XRD (Figure S6). The peaks at ~1.4, ~2.7 and ~4.8 Å in the FT-EXAFS 24 25 spectrum of Mn/N-C coincides with the Mn-O and Mn-Mnoxy interactions in manganese oxides, 26 27 while the additional peak at ~2.1 and 3.3 Å are also indicative of the presence of MnOOH 28 29 30 (Figure S6a). The remaining presence of a metal hydroxide in Mn/N-C and absence in the 31 32 other M/N-C samples is due to the significantly higher decomposition temperature of - 33 34 MnOOH (> 300 oC) compared to the metal (oxy)hydroxides for Fe, Co, Ni, Cu and Zn.[69] The 35 36 37 XRD pattern of Mn/N-C shows two peaks at ~33o and 36o attributed to Mn2O3 and Mn3O4, 38 39 respectively,[70] in addition to the two broad peaks arising from amorphous carbon (Figure 40 41 S6c). These results show that Mn/N-C contains manganese oxides/oxyhydroxides with 42 43 44 relatively larger particles sizes compared to other M/N-C, on the basis of their detection by 45 46 XRD. Regarding Zn/N-C, it contains both ZnN4 sites (already present in N-C) and zinc species 47 48 that were formed during the subsequent adsorption/precipitation process. As a result, the 49 50 51 XANES and FT-EXAFS of Zn/N-C closely resemble that of N-C, except for a minor peak at 52 53 ~2.9 Å attributed to Zn-Znoxy interaction as in ZnO. In contrast to Mn/N-C, the ZnO clusters 54 55 are not detected by XRD, suggesting amorphous or sub-nanometric ZnO clusters. 56 57 58 59 60 14 ACS Paragon Plus Environment Page 15 of 42 ACS Catalysis 1 2 3 In summary, ex situ XAS, 57Fe Mössbauer spectroscopy, XRD and STEM results demonstrate 4 5 6 the presence of highly dispersed transition metal oxides (MOx) supported on N-C (comprising 7 8 ZnNx sites) in M/N-C materials, except for Cu/N-C comprising atomically dispersed CuOx or 9 10 CuNx sites, with possible transmetalation from ZnNx to CuNx sites. 11 12 13 14 2.2 Activity and selectivity of M/N-C materials for eCO2RR 15 16 17 To measure the activity and selectivity, we resorted to an electrochemical flow cell with a GDE 18 19 cathode configuration, with CO2 gas feed at the back and, in this section, a continuous flow of 20 21 1 M KHCO3 as catholyte at the front (hydrophobic side) of the GDE. The anode and anolyte 22 23 24 utilized in this section were IrOx coated Ti plate and 1 M KHCO3, respectively. The Ag/AgCl 25 26 reference electrode was placed in the catholyte chamber. The gaseous products of eCO2RR 27 28 were identified and quantified under galvanostatic control (see details in Experimental 29 30 31 Methods). Figure 4 shows the Faradaic efficiencies (FE) for CO and H2 for all M/N-C 32 33 materials as well as for N-C, for different current densities ranging from 2 to 20 mA·cm–2. The 34 35 sum of the Faradaic efficiencies (FE) for CO and H2 is ≥ 90 % for all M/N-C materials, 36 37 38 indicating no or restricted formation of CO2RR liquid products. The FEs for CO an H2 vary 39 40 strongly both as a function of the nature of the metal in the catalyst and also as a function of 41 42 the current density (Figure 4). The trends of FEs vs. current density are shown in Figure 5a-d 43 44 for down-selected materials, namely Fe/, Co/, Ni/ and Cu/N-C. In contrast to Fe/ and Cu/N-C 45 46 47 with a bell shape in FE vs. current density (CO selectivity first increases with current density, 48 49 reaches a maximum and then decreases at even higher current density), Co/N-C and Ni/N-C 50 51 show increasing CO selectivity with current density up to 20 mA·cm–2 (Figure 5a-d). The 52 53 54 maximum selectivity towards CO is reached at 5 mA·cm–2 for Fe/N-C (ca 70 % CO) and Cu/N- 55 56 C (ca 50% CO) (Figure 5a and 5d), while the CO selectivity over Co/ and Ni/N-C reaches 73 57 58 and 90% at 20 mA·cm–2, respectively (Figure 5b-c). 59 60 15 ACS Paragon Plus Environment ACS Catalysis Page 16 of 42 1 2 3 Figure 5e shows the CO partial current density (JCO) vs. cathode potential for all M/N-C 4 5 6 materials and for N-C as well. All M/N-C catalysts show an earlier onset for eCO2RR to CO 7 8 than the N-C substrate, demonstrating the catalytic role of the metal species added to the N-C 9 10 substrate (itself containing ZnNx sites). Co/N-C shows the highest onset potential, > –0.3 V vs. 11 12 13 RHE, i.e. an absolute overpotential < 200 mV for eCO2RR to CO (Eo = –0.104 V vs. RHE[1]). 14 15 This onset potential is comparable to those reported for precious metal-based catalysts toward 16 17 eCO2RR to CO at similar pH.[19, 71, 72] The Fe/ and Ni/N-C materials have the second highest 18 19 onset potential, but while the polarization curves are almost overlapped at low current density 20 21 22 (< 2.5 mA·cm–2), they split at high current density with the Ni/N-C polarization curve showing 23 24 an exponential growth with decreasing cathode potential while the one for Fe/N-C shows a 25 26 nearly linear increase down to ca –0.55 V vs. RHE (Figure 5e). This is due to the sharp 27 28 29 decrease in FE(CO) for Fe/N-C at a total current density > 5 mA·cm–2 (Figure 5a), while the 30 31 FE(CO) of Ni/N-C continuously increases with increasing total current density (Figure 5c). 32 33 The JCO polarization curves of the Mn/N-C, Cu/N-C and Zn/N-C catalysts are similar and much 34 35 36 lower than those of Fe/N-C, Co/N-C and Ni/N-C (Figure 5e), due mostly to low FE(CO) on 37 38 those materials at current densities >5 mA·cm–2 (Figure 4c-d). In summary, these flow cell 39 40 measurements identify the Co- and Ni/N-C as the most promising catalysts in this series for 41 42 electrochemical CO production at high current densities. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 16 ACS Paragon Plus Environment Page 17 of 42 ACS Catalysis 1 2 3 2 CO H2 4 (a) 2 mA/cm (b) 5 mA/cm2 5 100 6 7 80 8 FE (%) 9 60 10 11 40 12 13 20 14 15 0 Mn Fe Co Ni Cu Zn NC Mn Fe Co Ni Cu Zn NC 16 2 2 (c) 10 mA/cm (d) 20 mA/cm 17 18 100 19 80 20 FE (%) 21 60 22 23 40 24 25 20 26 27 0 28 Mn Fe Co Ni Cu Zn NC Mn Fe Co Ni Cu Zn NC 29 Figure 4. The FEs at different current densities over M/N-C catalysts (M = Mn, Fe, Co, Ni, Cu 30 31 and Zn) at 2 (a), 5 (b), 10 (c) and 20 (d) mA·cm2. The measurements were performed in flow 32 33 34 cell under galvanostatic control for ~90 min in 1 M KHCO3 aqueous catholyte and anolyte at 35 36 room temperature, CO2 gas was fed at 50 mL min–1 at the back of the GDE. The cathode 37 38 catalyst loading was 1 mg·cm–2, the membrane was Tokuyama Neosepta® anion exchange 39 40 41 membrane, and the anode was a flat IrOx electrode. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 ACS Paragon Plus Environment ACS Catalysis Page 18 of 42 1 2 3 120 120 4 (a) CO H2 (b) 5 100 100 6 7 80 80 8 FE (%) FE (%) 9 60 60 10 11 40 40 12 20 20 13 14 0 0 0.5 1 2 5 10 20 0.5 1 2 5 10 20 15 Current density (mA/cm2) Current density (mA/cm2) 16 120 120 17 (c) (d) 18 100 100 19 20 80 80 FE (%) FE (%) 21 60 60 22 23 40 40 24 25 20 20 26 27 0 0 0.5 1 2 5 10 20 0.5 1 2 5 10 20 28 2 Current density (mA/cm ) 2 Current density (mA/cm ) 29 12 0.25 (e) (f) 30 31 15 N-C 10 jCO (mA/cm 2) @ -0.45 V  (V) @ -2.5 mA/cm 2 Mn/N-C 0.30 32 Fe/N-C 33 8 JCO (mA/cm 2) Co/N-C 34 10 0.35 Ni/N-C 6 35 Cu/N-C 36 Zn/N-C 4 0.40 5 37 38 2 0.45 39 0 40 0 -0.2 -0.4 -0.6 -0.8 Mn Fe Co Ni Cu Zn 41 Potential (V vs. RHE) 42 43 Figure 5. The Faradaic efficiencies for CO and H2 measured in a flow cell as a function of the 44 45 current density for the down-selected Fe/N-C (a), Co/N-C (b), Ni/N-C (c) and Cu/N-C (d), and 46 47 the derived polarization curves of CO partial current density vs. cathode potential for all 48 49 50 materials (e). (f) The volcano trends in eCO2RR activity for M/N-C catalysts: CO partial 51 52 current density at −0.45 V vs. RHE (left handside y-axis) and cathode overpotential at –2.5 53 54 mA·cm–2 (right handside y-axis). The CO partial current density and overpotential shown in 55 56 Figure 5f is extracted from Figure 5e by interpolation. The data were obtained under 57 58 59 galvanostatic control for ~90 min in 1 M KHCO3 at room temperature. CO2 gas was supplied 60 18 ACS Paragon Plus Environment Page 19 of 42 ACS Catalysis 1 2 3 at the back of the gas diffusion electrode whose microporous side was coated with 1 mg·cm-2 4 5 6 of catalyst, while the anode was IrOx. The potential and overpotential reported in e and f are 7 8 iR-corrected. 9 10 11 To identify trends, we then plotted the JCO at –0.45 V vs. RHE and the cathode overpotential 12 13 14 (η) at –2.5 mA·cm–2 (CO partial current density) vs. the atomic number of transition metal 15 16 (Figure 5f). Interestingly, with increasing atomic number from Mn to Co, JCO monotonically 17 18 increases, whereas with further increasing atomic number of the metal, the JCO decreases. A 19 20 similar trend is also apparent for the overpotential at –2.5 mA·cm–2. Similar volcano trends are 21 22 23 obtained if one uses the redox potential M2+/M as x-axis. Herein, we reveal a volcano trend in 24 25 the electrocatalytic activity for the CO2 to CO reduction over a series of catalysts characterised, 26 27 ex situ, with M/N-C materials comprising highly dispersed MOx sites in ex situ conditions, with 28 29 30 Co/N-C located at the volcano peak. This volcano trend is similar to that observed over M-N- 31 32 C catalysts with M-N4 sites.[25] 33 34 35 While the high activity and selectivity for eCO2RR to CO of the three most active catalysts, 36 37 38 Fe-, Co- and Ni/N-C, can be related to the (sub)-nano MOx structures identified ex situ, whether 39 40 the nitrogen plays a role during the synthesis or during eCO2RR of those M/N-C catalysts is 41 42 still elusive. Therefore, we prepared nitrogen-free FeOx/C material by precipitating FeOx on 43 44 Vulcan XC72 carbon black powder (labelled as FeOx/Vulcan), following the same protocol as 45 46 47 before. This reference material was characterised ex situ by XAS and its eCO2RR activity was 48 49 measured in the flow cell. The FeOx/Vulcan shows FT-EXAFS and XANES spectra similar to 50 51 those of Fe/N-C catalysts (Figure S7a-b). This demonstrates the presence of the metal as 52 53 54 (sub)nano-oxide, as intended. The Fe content measured from the edge jump of its XAS 55 56 spectrum (Table S1) confirm that FeOx/Vulcan has a metal content comparable to Fe/N-C. 57 58 Negligible eCO2RR to CO activity is observed for FeOx/Vulcan when tested in flow cell under 59 60 19 ACS Paragon Plus Environment ACS Catalysis Page 20 of 42 1 2 3 the same conditions as M/N-C, with H2 as the major product (Figure S7c). It has been reported 4 5 6 that Fe3+ species from iron oxyhydroxides supported on carbon were operando reduced into 7 8 metallic particles,[26] while the same Fe3+ species supported on N-C were only reduced to 2+ 9 10 oxidation state. Fe0 is probably responsible for the high selectivity toward H2 production of 11 12 13 FeOx/Vulcan, since Fe0 is not known to catalyse CO2 reduction selectively, even as 14 15 nanoparticles (it should be noted here that this might not be the case for other metals). These 16 17 results show that supporting the metal oxide particles on N-C is key to their CO2RR activity 18 19 and selectivity, due to similar Fe K-edge XAS spectra and Fe content identified ex situ for 20 21 22 Fe/N-C and FeOx/Vulcan, but different activity and selectivity (Figure 5 and S7). 23 24 25 Therefore, a synergy between N-defects or ZnNx and the MOx (sub)nano clusters seems to be 26 27 at play. R. Arrigo et al.[26] found from theoretical simulations that pyridinic N defects in N-C 28 29 30 lowers the interface energy between iron oxyhydroxide and carbon support, indicating the 31 32 possible direct interaction between the N-defects and the MOx species. A synergy between N 33 34 defects and FeOx sites was proposed to promote C-C coupling for acetic acid production.[26] 35 36 37 Moreover, P. Atanassov et al confirmed with near ambient pressure XPS that pyridinic and 38 39 hydrogenated (pyrrolic) nitrogen moieties, rather than Fe-N4 moieties, act as preferential 40 41 adsorption sites for CO2 over pyrolyzed Fe-N-C catalysts[73]. Thus, it is likely that the MOx 42 43 species and the neighbouring N-defects co-catalyse the eCO2RR. Moreover, further work will 44 45 46 be needed to elucidate whether Zn plays a role (direct or indirect) in stabilizing such hydroxides 47 48 particles or not. 49 50 51 2.3 Structural changes of MOx sites identified with operando XAS 52 53 54 55 To gain more insights in the fate of the metal-oxide nanoparticles supported on N-C, we 56 57 resorted to in operando XAS. It is widely acknowledged that metal-based electrocatalysts often 58 59 experience partial reduction and/or structural reconstruction during eCO2RR.[17, 26, 42, 74, 75] This 60 20 ACS Paragon Plus Environment Page 21 of 42 ACS Catalysis 1 2 3 can be due to the low electrochemical potential, operando produced H2 gas and/or operando 4 5 6 produced CO. The latter is known to facilitate reconstruction of metal surfaces, and 7 8 agglomeration of NPs.[43] For M-N-C materials prepared by pyrolysis and comprising only 9 10 MN4 moieties ex situ, we recently reported a decreased average oxidation state of Fe and Ni 11 12 13 within MN4 moieties and reduction of a large fraction of Cu2+Nx sites to Cu(0), via the 14 15 reversible formation of Cu (sub)nanoparticles, with decreasing electrochemical potential under 16 17 eCO2RR working conditions.[25, 55] To identify the local structures of transition metals sites 18 19 under operating conditions, we therefore conducted operando XAS on the series of M/N-C 20 21 22 catalysts. 23 24 25 We first studied the N-C substrate (comprising ZnNx moieties, Figure S1) at the Zn K-edge. 26 27 The Zn K-edge XANES spectra show only a slight decrease in white line intensity with 28 29 30 decreasing potential (Figure S8a-b). A similar behaviour is observed for Zn/N-C (Figure S8c- 31 32 d). By extrapolation, this suggests a slightly reduced average oxidation state of Zn in M/N-C 33 34 and M-N-C catalysts with decreasing potential. This can be assigned to the operando formation 35 36 37 of a very small amount of metallic Zn nanoparticles. Regarding Mn/N-C, the structure remains 38 39 unchanged till –1.0 V vs. RHE (Figure S8e-f). Overall, the operando observations with Zn and 40 41 Mn are in line with their Pourbaix diagrams, with no or minimal reduction of the metal expected 42 43 before –0.76 V and –1.2 V vs. RHE, respectively. 44 45 46 47 In contrast, the operando XAS spectra of Fe/, Co/, Ni/ and Cu/N-C are different from those 48 49 measured ex situ (Figure 6 and S9-10), in line with their much higher M2+/M redox potential. 50 51 To start with the simplest case, the Cu K-edge XANES spectrum of Cu/N-C shifts toward lower 52 53 54 energies with decreasing potential, corresponding to a decreased average oxidation state 55 56 (Figure 6a). This is accompanied by the appearance of a signal that can be assigned to the Cu- 57 58 Cu coordination, by comparison with the FT-EXAFS of a Cu foil (Figure 6b). This 59 60 21 ACS Paragon Plus Environment ACS Catalysis Page 22 of 42 1 2 3 demonstrates the reduction from Cu2+ to nearly Cu0 with an extensive formation of metallic 4 5 6 copper clusters during eCO2RR, even at 0 V vs. RHE, well above the onset of eCO2RR on 7 8 Cu/N-C (~ –0.4 V vs. RHE). The formation of Cu nanoparticles is reversible, as shown by the 9 10 spectrum of OCP-back overlapping with that of OCP (Figure S9a-b). A minor Cu-N/O peak 11 12 13 around 1.5 Å is visible at low potentials, indicating that a small fraction of copper atoms binds 14 15 with N or O surface groups from N-C (Figure 6b). The anchoring of some Cu atoms to both 16 17 the N-C support and the operando formed Cu0 nanoparticles could explain the restricted 18 19 particle growth and the observed reversibility with Cu re-dispersion at high potential. The 20 21 22 operando formation of Cu0 nanoparticles has been widely observed on other Cu-based catalysts 23 24 comprising isolated Cu atoms ex situ;[56, 76, 77] and these Cu0 nanoparticles very likely act as the 25 26 catalytically active species. Recently, T. Möller et al revealed that Cu nanoparticles supported 27 28 29 on carbon favour the formation of C1 product instead of the C2+ products generally observed 30 31 on unsupported Cu[78]. The selectivity towards C1 product with Cu nanoparticles is due to the 32 33 large mean interparticle distance for particles supported in low wt% on a high surface area 34 35 36 support material, in line with our finding that CO is the major product in this potential range 37 38 over the operando formed Cu0 supported on N-C. 39 40 41 For Ni/N-C, the XANES and FT-EXAFS spectra remain unchanged when the potential was 42 43 reduced from OCP to 0 V vs. RHE, and these spectra are close to those measured ex situ, 44 45 46 assigned to NiO clusters. From 0 to –0.5 V vs. RHE, the XANES spectra then show a gradual 47 48 decrease in white line intensity and increase in the pre-edge peak intensity at ~8334 eV (Figure 49 50 6c and S10a). The intensity of the pre-edge peak at 8334 eV stays the same at all potentials ≤ 51 52 53 –0.5 V vs. RHE, and is indicative of the formation of highly distorted octahedral or tetrahedral 54 55 Ni-Ox coordination in nickel hydroxides (Figure 6c and S10a).[79, 80] In contrast, in FT-EXAFS 56 57 a new intense peak progressively appears at ~2.5 Å from –0.4 V to –1.0 V vs. RHE (Figure 58 59 60 S9d), and that can be assigned to Ni-Nioxy interaction in Ni hydroxide. In addition, the 22 ACS Paragon Plus Environment Page 23 of 42 ACS Catalysis 1 2 3 appearance of a shoulder at ~2 Å at–1.0 V vs. RHE (Figure 6c), a position corresponding to 4 5 6 Ni-Ni interaction in the metallic Ni foil, suggests the minor formation of metallic Ni at the 7 8 lowest potential. Overall, the data suggests the transformation from (sub)nano-NiO ex situ to 9 10 (sub)nano-Ni(OH)2 at –0.4 V vs. RHE, and then clustering to larger Ni(OH)2 nanoparticles at 11 12 13 even lower potential. Moreover, there is a phase contraction of the material at low potential 14 15 compared to Ni(OH)2, as evidenced by the shorter Ni-O and Ni-Nioxy bond distances compared 16 17 to those of Ni(OH)2 reference (e.g. at –0.5 and –1.0 V vs. RHE, Figure 6d and Figure 7). We 18 19 note that this phase contraction has been widely observed for nickel-based catalysts during 20 21 22 oxygen evolution reaction (OER),[81, 82] attributed to the oxidized nature of Ni at high potentials. 23 24 We for the first time, to the best of our knowledge, report similar phase contraction of nickel 25 26 hydroxides, but at reducing potentials. The lattice contraction may be due to cation vacancies 27 28 29 introduced by partially reduced nickel, and/or due to the intercalation of water and bicarbonate 30 31 ions into the layered structures of nickel hydroxides.[83-85] Upon reverting the potential from – 32 33 1 V vs. RHE to OCP (OCP-back curves), the XANES and FT-EXAFS spectra were largely 34 35 36 restored to the initial OCP spectrum (Figure S9c-d), suggesting the reversibility of this process. 37 38 39 The absorption edge in the XANES spectrum of Co/N-C remains largely unchanged with 40 41 potential from OCP to 0 and –0.3 V vs. RHE, while the white line intensity slightly increases 42 43 (Figure 6e and Figure S9e). The latter can be explained by a transition from cobalt oxide to 44 45 46 cobalt hydroxide phase, as can be seen by comparison to the white line intensities for the 47 48 reference spectra for the latter (compare Figure S3b and 6e). The XANES spectra at –0.5 and 49 50 –1.0 V vs. RHE show a decreased white line intensity and an increased pre-edge feature 51 52 53 (~7710–7714 eV) intensity, indicative of the partial reduction of Co into Co0 species (Figure 54 55 6e and S10b). The broad Co-O peak in the corresponding FT-EXAFS spectra seen at high 56 57 potential at ~1.5 Å (with a shoulder ~1.7 Å) splits into two distinct peaks at 1.2 and 1.7 Å at – 58 59 60 0.5 V vs. RHE and below (Figure 6f). These two peaks are possibly assigned to Co-O 23 ACS Paragon Plus Environment ACS Catalysis Page 24 of 42 1 2 3 interaction from octahedral Co2+ and Co-O or Co-C interaction from tetrahedral Co2+, 4 5 6 respectively. It should be noted here that splitting of a FT-EXAFS peak can also be explained 7 8 by the destructive interferences due to the (dis)appearance of phases. Tetrahedral CoX4 9 10 structure has been reported for layered Co(OH)2, where in the Co2+ coordinates with O and/or 11 12 13 C from intercalated water and HCO3– cations between CoO6 layers.[86] The intercalated water 14 15 and cations also induce the phase contraction of Co(OH)2, as evidenced by the shorter Co-O 16 17 and Co-Cooxy peak in Co/N-C than in reference Co(OH)2 (Figure 6f). The critical role of 18 19 HCO3– cations, instead of gas-phase CO2, during eCO2RR has been observed over Au and 20 21 22 Cu.[87, 88] Interestingly, the intensity of the peak arising from Co-Cooxy in the FT-EXAFS 23 24 spectra first increases at –0.5 V vs. RHE induced by the clustering of Co sites; and then 25 26 decreases at –1.0 V vs. RHE, probably due to the potential-induced reversible amorphization 27 28 29 of cobalt oxide reported previously.[89] The XANES spectrum reverts to the spectrum of 0 V 30 31 vs. RHE at the end of the series of applied potentials, while the Co-O peak in the FT-EXAFS 32 33 spectrum remains split, different from the initial OCV FT-EXAFS (Figure S9e-f). This 34 35 36 suggests that the intercalated structure was maintained for at least some time after removing 37 38 the potential bias. 39 40 41 Regarding Fe/N-C (Figure 6g), there is a negative shift in the position of the absorption edge 42 43 of Fe K-edge XANES spectra of ~2 eV from OCP to 0 V vs. RHE, indicative of a Fe3+ to Fe2+ 44 45 46 reduction; and then the XANES spectra remain the same till –0.4 V vs. RHE (Figure S9g). 47 48 When the potential is further lowered to –1.0 V vs. RHE, the white line intensity gradually 49 50 decreases, accompanied with an increased pre-edge peak (~7115 eV) intensity (Figure S10c). 51 52 53 This enhanced pre-edge peak is indicative of the partial reduction of Fe2+ to Fe0 species. As for 54 55 Co/N-C, the Fe-N/O peak in the corresponding FT-EXAFS spectra also splits into two peaks 56 57 at 1.1 and 1.8 Å (Figure 6h). This effect was observed previously by R. Arrigo et al. as well.[26] 58 59 60 The FT-EXAFS Fe-Feoxy peak first decreases in intensity till –0.4 V vs. RHE due to a decreased 24 ACS Paragon Plus Environment Page 25 of 42 ACS Catalysis 1 2 3 particle size of iron hydroxide (Figure S9h); and then shifts toward lower R value below –0.5 4 5 6 V vs. RHE, due to phase contraction (Figure 6h and S9h). The changes in oxidation states 7 8 observed in the XANES spectra of Fe/N-C are largely reversible when the potential is reverted 9 10 back to OCP (Figure S9g), while the split Fe-O peak due to the phase contraction remain to 11 12 13 some extend (Figure S9h). 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25 ACS Paragon Plus Environment ACS Catalysis Page 26 of 42 1 2 3 4 1.4 2.5 (a) (b) 5 1.2 Cu foil/5 6 2.0 7 1.0 ormalized (E) 8 R)| (A ) 1.5 -3 0.8 o 9 OCV -1.0 V 10 0.6 1.0 0V Cu foil 11 0.4 -0.5 V Cu(OH)2 12 0.5 0.2 13 14 0.0 0.0 15 8980 9000 9020 9040 9060 0 1 2 3 4 5 6 1.6 1.4 16 (c) (d) 17 1.4 1.2 Ni foil/5 18 1.2 1.0 ormalized (E) 19 1.0 R)| (A ) 20 -3 0.8 o 21 0.8 0.6 22 0.6 OCV -1.0 V 23 0.4 0V Ni foil 0.4 24 -0.5 V Ni(OH)2 0.2 0.2 25 26 0.0 0.0 27 8340 8360 8380 8400 0 1 2 3 4 5 6 28 1.0 (e) (f) 29 Co foil/2 1.5 0.8 30 31 ormalized (E) R)| (A ) 32 0.6 -3 o 1.0 33 34 0.4 35 0.5 OCV -1.0 V 36 0V Co foil 0.2 -0.5 V Co(OH)2 37 38 0.0 0.0 39 7700 7720 7740 7760 7780 0 1 2 3 4 5 6 40 1.0 41 (g) (h) 1.2 Fe foil/3 42 0.8 normalized (E) 43 |(R)| (A -3) 44 0.8 0.6 o 45 46 OCV -1.0 V 0.4 47 0.4 0V Fe foil 48 -0.5 V Fe(OH)2 0.2 49 50 0.0 0.0 51 7100 7120 7140 7160 7180 0 1 2 3 4 5 6 52 Energy (eV) R (A) o 53 Figure 6. The operando XANES and FT-EXAFS spectra for M/N-C materials (M = Fe, Co, 54 55 56 Ni, Cu) and their comparison to reference spectra of metal-hydroxides and metallic foils. 57 58 Operando XANES spectra for Cu/N-C (a), Ni/N-C (c), Co/N-C (e) and Fe/N-C (g). Operando 59 60 26 ACS Paragon Plus Environment Page 27 of 42 ACS Catalysis 1 2 3 FT-EXAFS spectra for Cu/N-C (b), Ni/N-C (d), Co/N-C (f) and Fe/N-C (h) and comparison to 4 5 6 the spectra for metal-hydroxides or metallic foils. 7 8 9 In summary, the majority of Cu atoms in Cu/N-C forms Cu0 clusters during eCO2RR; while 10 11 the Mn/, Fe/, Co/, and Ni/N-C maintain the metal hydroxide structures, with a minor amount 12 13 14 of M0 formed in Fe-, Co-, and Ni-based samples. The unexpected information from operando 15 16 XAS is that the vast majority of Ni, Co and Fe nano-oxides were not reduced to metallic 17 18 particles, even at –1 V vs. RHE, which is 550 to 750 mV lower than the values expected 19 20 according to the Pourbaix diagrams. In contrast, most metal atoms in Ni-, Fe- and Co/NC 21 22 23 catalysts stay as oxides, but with a phase contraction (Figure 7). The M-O peaks in the FT- 24 25 EXAFS of Fe/ and Co/N-C split into two peaks due to tetrahedral M2+ coordination in the 26 27 intercalated layer, indicating that the HCO3– is inserted into the layered structure of metal 28 29 30 hydroxides (Figure 7). Recently, two independent groups demonstrated the critical role of 31 32 HCO3– ions, through equilibrium exchange with dissolved CO2, rather than the supplied CO2, 33 34 as the primary source for eCO2RR.[87, 88] Thus, the inserted HCO3– cations and/or the phase 35 36 37 contraction of metal oxides at working potentials may be responsible for the superior activity 38 39 of Fe-, Co- and Ni/N-C catalysts. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 27 ACS Paragon Plus Environment ACS Catalysis Page 28 of 42 1 2 3 1.0 (c) 1.0 (d) 1.4 (e) 4 OCV OCV 1.2 OCV 0.8 -0.4 V 0.8 -0.4 V 5 -1.0 V -1.0 V 1.0 -0.4 V -1.0 V 6 |(R)| (A -3) Co(OH)2 R)| (A ) Fe(OH)2 R)| (A ) -3 0.6 0.6 -3 o o o 0.8 Ni(OH)2 7 0.6 0.4 0.4 8 0.4 9 0.2 0.2 0.2 10 0.0 0.0 0.0 11 0 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 o 4 5 6 o o 12 R (A) R (A) R (A) 13 Figure 7. The scheme illustrating the phase contraction and structural distortion of M(OH)2 14 15 induced by the intercalation of HCO3– (modified from Ref. 86). The structures of M(OH)2 16 17 18 without (a) and with (b) cathodic potential bias. Species for M, OH, HCO3–, and H2O are 19 20 represented by orange, red, green, and blue spheres, respectively. The insertion of HCO3– leads 21 22 to the distortion of MO6 octahedra and the appearance of tetrahedral MO4 with different M-O 23 24 25 bond distances, as evidenced by the splitting of M-O peaks. The insertion of HCO3– is also 26 27 accompanied with shortened M-O and M-Moxy bond distances. The splitting of M-O peaks with 28 29 decreasing potential of Fe/N-C (c) and Co/N-C (d), as well as the shortened M-O and/or M- 30 31 Moxy bond distances in Fe/N-C (c), Co/N-C (d) and Ni/N-C (e) vs. M(OH)2 references are 32 33 34 shown from selected operando FT-EXAFS. 35 36 37 2.3 Optimized flow cell with M/N-C cathodes and NiFe-based anodes 38 39 40 The intrinsic activity and selectivity of Ni/N-C toward CO2RR to CO are comparable to those 41 42 43 of the state-of-the-art Au, Ag, and M-N-C catalysts (Table S4). M/N-C and M-N-C materials 44 45 hold great potential to lower the cost of CO2 electrolyzer by replacing the expensive precious 46 47 metal-based catalysts (e.g., Au and Ag). Moreover, the advantage of M/N-C over M-N-C lies 48 49 50 not only in the facile and controlled synthesis of the former, but also the possible reactivation 51 52 of M/N-C materials after degradation (reduction), since the reconversion of metallic 53 54 nanoparticles into metal hydroxide form should be easier than the reconversion of a metallic 55 56 cluster back into MNx sites. However, even with the most promising M/N-C (Fe, Co, or Ni) 57 58 59 cathode and flat IrOx anode, the total cell voltage is above 3 V at 20 mA/cm2 (Figure S11), 60 28 ACS Paragon Plus Environment Page 29 of 42 ACS Catalysis 1 2 3 with a large overpotential likely arising also at the anode side. To decrease the overpotential 4 5 6 for the oxygen evolution reaction (OER) and remove precious metals from the cell, we 7 8 implemented a high surface area Ni-Fe oxyhydroxide anode prepared on a Ni foam (NF) via a 9 10 phosphidation process, labelled as NF/Ni2P/Ni-FeOx (see synthesis in Experimental 11 12 13 Methods)[90]. 14 15 16 The porous backbone of NF is observed in the micro-computed tomography (CT) (Figure S12a) 17 18 and scanning electron microscopy (SEM) (Figure S13a) images, and the NiFe-based OER 19 20 catalyst was decorated homogeneously on the Ni foam (Figure S12a). The XRD pattern 21 22 23 exhibits two strong peaks assigned to NF, two minor peaks arising from Ni2P, and no peak of 24 25 the Ni-Fe phase (Figure S13b); while the presence of Fe, P and O is clearly evidenced in 26 27 energy dispersive X-ray spectroscopy (EDX) and electron energy loss spectroscopy (EELS) 28 29 30 mapping (Figure S13c and S14a). This indicates the amorphous nature of the Ni-Fe 31 32 oxyhydroxide in the catalyst, further supported by high resolution (HR)-TEM images, showing 33 34 a crystalline phase of hexagonal Ni2P in the core and another amorphous phase on the edges 35 36 37 (Figure S14b-c). The OER activity of NF/Ni2P/Ni-FeOx was first evaluated with rotating disk 38 39 electrode (RDE) in KOH (0.1 and 1 M) and KHCO3 (1 M) in comparison with IrO2 supported 40 41 on Ni foam. Figure S15a-c demonstrates that the OER activity of NF/Ni2P/Ni-FeOx is superior 42 43 to that of IrO2 in all electrolytes, and that the overpotential for the former at 10 mA·cm–2 44 45 46 reduces with increasing pH, reaching ~240 mV in 1 M KOH (Figure S15b). Moreover, the 47 48 overpotential at 10 mA·cm–2 was further reduced by ~100 mV after an accelerated stability test 49 50 (AST) (Figure S15d). No obvious changes in the morphology and structure of NF/Ni2P/Ni- 51 52 53 FeOx were observed by X-CT, SEM, XRD, and HRTEM (Figure S12-13 and S16). Its 54 55 impressive activity and stability make it appealing to implement in electrolyzers, and a 56 57 significant decrease in total cell voltage up to ~1 V was observed after replacing IrO2 anode 58 59 60 29 ACS Paragon Plus Environment ACS Catalysis Page 30 of 42 1 2 3 with NF/Ni2P/Ni-FeOx in a CO2 electrolyzer with Ni-N-C as cathode and 1 M KHCO3 as 4 5 6 electrolyte (Figure S11). 7 8 9 We investigated the activity and selectivity of Fe/, Co/ and Ni/N-C cathodes with NF/Ni2P/Ni- 10 11 FeOx anode (Figure 8). Since the anode catalyst is more active at higher pH, we used 1 M 12 13 14 KOH instead of KHCO3 as anolyte and catholyte in the following study. Co- and Ni/N-C 15 16 exhibit similar cell voltages for a given current density, lower than that of Fe/N-C. However, 17 18 Ni/N-C exhibits a higher FE than Co/N-C, reaching a FE(CO) >90% at 50 mA·cm–2. A FE(CO) 19 20 >80% is even maintained over Ni/N-C at 100 mA·cm–2 total current density, with a total cell 21 22 23 voltage of ~4 V. It should be noted here that a large portion of the cell voltage comes from the 24 25 iR-drop due to the thick layers of electrolyte in the commercial cell. The total cell resistance is 26 27 > 25 Ω·cm2 as measured by high frequency resistance, while that estimated for the liquid 28 29 30 chambers is ~10 Ω·cm2 (from circa 2 cm thickness of anolyte + catholyte, electrolyte 31 32 conductivity of ~200 mS·cm–1 for 1 M KOH). Thus, the total cell voltage could be further 33 34 reduced through cell design, e.g. reducing the thickness of each electrolyte chamber possibly 35 36 37 down to zero-gap configuration, or increasing the temperature. 38 39 40 When the total current density was increased from 100 to 150 mA·cm–2, a sharp decrease in FE 41 42 was observed. Similar behaviour has been observed over Ni-N-C as well. This is probably due 43 44 to CO2 depletion at the catalytic sites, leading to competitive HER. This can be improved by 45 46 47 increasing the CO2 backpressure. P. Strasser et al.[16] reported a stable current density as high 48 49 as 200 mA·cm–2 over Ni-N-C catalysts with the same flow cell, by closing the CO2 gas out to 50 51 force the CO2 passing through the GDE, resulting in pressurized CO2 gas flow. Moreover, we 52 53 54 also noticed some particles in the anolyte, indicating the mechanical instability of the anode 55 56 during strong O2 evolution. This suggests that improved mechanical robustness of the anode 57 58 during strong gas evolution is also important for cell stability. 59 60 30 ACS Paragon Plus Environment Page 31 of 42 ACS Catalysis 1 2 3 4 (a) CO H2 (b) 100 100 5 6 80 80 7 8 FE (%) FE (%) 60 60 9 10 40 40 11 12 20 20 13 14 0 0 15 2 5 10 20 50 2 5 10 20 50 16 Current density (mA/cm2) Current density (mA/cm2) 17 120 -6 18 (c) (d) 19 100 Fe/N-C Co/N-C -5 20 Ni/N-C Cell voltage (V) 21 80 22 -4 FE (%) 23 60 24 -3 25 40 26 -2 20 27 28 0 -1 29 5 10 20 50 100 150 0 20 40 60 80 100 120 140 160 30 2 Current density (mA/cm ) Current density (mA/cm2) 31 32 33 34 Figure 8. The Faradaic efficiencies of flow cell measurements with NF/Ni2P/Ni-FeOx anode 35 36 and M/N-C cathodes (Fe (a), Co (b) and Ni (c)), and (d) the polarization curve of cell voltage 37 38 vs. total current density. Data were acquired in flow cell with 1 M KOH aqueous catholyte and 39 40 41 anolyte at room temperature, and the CO2 gas was fed at the back of GDE. 42 43 44 In summary, we have synthesized a series of M/N-C catalysts with highly dispersed MOx 45 46 particles supported on nitrogen-doped carbon, and evaluated their activity and selectivity 47 48 49 toward eCO2RR to CO in a flow cell with GDE. We reveal a volcano-shaped trend in the 50 51 electrocatalytic activity for CO over M/N-C catalysts, which is similar to that reported for M- 52 53 N-C catalysts with atomically-dispersed M-N4 sites. Operando XAS shows that the metal 54 55 (hydr)oxide clusters in M/N-C largely remain as contracted hydroxides at –1 V vs. RHE in 56 57 58 Fe/N-C, Co/N-C and Ni/N-C, with the parallel formation of a minor amount of metallic species. 59 60 These metal hydroxides experience phase contractions and the insertion of H2O/HCO3– into 31 ACS Paragon Plus Environment ACS Catalysis Page 32 of 42 1 2 3 their layered structures. Last, flow cell measurements with M/N-C cathodes and state-of-the- 4 5 6 art NiFe-anodes were performed, and Ni/N-C exhibits >80% FE toward CO at 100 mA/cm2 7 8 with a total cell voltage of ~4 V. 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 32 ACS Paragon Plus Environment Page 33 of 42 ACS Catalysis 1 2 3 ASSOCIATED CONTENT 4 5 6 7 AUTHOR INFORMATION 8 9 10 Corresponding Author 11 12 13 * To whom correspondence should be addressed. Email: frederic.jaouen@umontpellier.fr 14 15 16 17 Author Contributions 18 19 20 J.L. and F.J. designed, synthesized and physically characterized the materials. F.A.G.-P., P.T, 21 22 J.A., and J.R.G.-M. processed and characterized the NiFe-based OER catalysts and electrodes. 23 24 P.T. and J.A. conducted the HRTEM and STEM-EELS characterization of NiFe-based anodes. 25 26 27 J.L. and F.J. also performed the eCO2RR electrochemical measurements and product analysis. 28 29 T.A., and P.A. conducted the STEM characterizations of M/N-C materials. I.Z. performed the 30 31 X-CT experiments and analysis. J.L. and A.Z. conducted the ex situ and operando XAS 32 33 34 experiments. T.A., M.K. and P.A. conducted the XPS measurement and analysis. M.T.S. 35 36 performed the Mössbauer spectroscopy measurements and analysis. J.L., A.Z. and F.J. wrote 37 38 and edited the manuscript with input from all authors. All authors read and approved the final 39 40 41 version of the manuscript. 42 43 44 Funding Sources 45 46 47 48 The research leading to these results has received funding from the A-LEAF Project, which is 49 50 51 funded by the European Union’s H2020 Programme under Grant Agreement No.732840. ICN2 52 53 54 and ICIQ acknowledge funding from the FEDER/Ministerio de Ciencia e Innovación, 55 56 57 Agencia Estatal de Investigación (projects ENE2017-85087-C3 and RTI2018-095618-B- 58 59 60 I00); the Generalitat de Catalunya (2017 SGR 327 and 2017- SGR-1406) and by the CERCA 33 ACS Paragon Plus Environment ACS Catalysis Page 34 of 42 1 2 3 4 Programme / Generalitat de Catalunya. ICN2 and ICIQ are supported by the Severo Ochoa 5 6 7 program from Spanish MINECO (Grants No. SEV-2017-0706 and CEX2019-000925-S). 8 9 10 11 Notes 12 13 14 15 The authors declare no competing financial interest. 16 17 18 19 Supporting Information. 20 21 The following files are available free of charge. 22 23 24 Physical (XRD, STEM, XPS), ex situ and operando XAS, and electrochemical (FEs and 25 26 27 polarization curves obtained from flow cell tests) characterizations of M/N-C catalysts for 28 29 eCO2RR or NiFe-based catalysts for OER (PDF) 30 31 32 ACKNOWLEDGMENT 33 34 35 36 We acknowledge Synchrotron SOLEIL (Gif-sur Yvette, France) for provision of synchrotron 37 38 radiation facilities at beamline SAMBA (proposal number 20190681). The authors thank Dr. 39 40 David A. Cullen for assistance in obtaining atomic resolution HAADF-STEM images, which 41 42 was conducted at the Center for Nanophase Materials Sciences, which is a DOE Office of 43 44 45 Science User Facility. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 34 ACS Paragon Plus Environment Page 35 of 42 ACS Catalysis 1 2 3 REFERENCES 4 5 1.Fan, L.; Xia, C.; Yang, F.; Wang, J.; Wang, H.; Lu, Y. Strategies in Catalysts and Electrolyzer 6 Design for Electrochemical CO2 Reduction toward C2+ Products. Sc. Adv. 2020, 6, eaay3111. 7 2.Ma, M.; Clark, E.L.; Therkildsen, K.T.; Dalsgaard, S.; Chorkendorff, I.; Seger, B. Insights 8 into the Carbon Balance for CO2 Electroreduction on Cu using Gas Diffusion Electrode 9 Reactor Designs. Energy & Environ. Sci. 2020, 13, 977-985. 10 3.Burdyny, T.; Smith, W.A. CO2 Reduction on Gas-Diffusion Electrodes and Why Catalytic 11 Performance Must Be Assessed at Commercially-Relevant Conditions. Energy & Environ. Sci. 12 13 2019, 12,1442-1453. 14 4.Nam, D.-H.; Nam, D.-H.; Luna, P.D.; Rosas-Hernández, A.; Thevenon, A.; Li, F.; Agapie, 15 T.; Peters, J.C.; Shekhah, O.; Eddaoudi, M.; Sargent, E.H. Molecular Enhancement of 16 Heterogeneous CO2 Reduction. Nat. Mater. 2020, 19, 266-276. 17 5.Wei, X.; Yin, Z.; Lyu, K.; Li, Z.; Gong, J.; Wang, G.; Xiao, L.; Lu, J.; Zhuang, L. Highly 18 Selective Reduction of CO2 to C2+ Hydrocarbons at Cu/Polyaniline Interfaces. ACS Catal. 2020, 19 10, 4103-4111. 20 21 6.Clark, E.L.; Hahn, C.; Jaramillo, T.F.; Bell, A.T. Electrochemical CO2 Reduction over 22 Compressively Strained CuAg Surface Alloys with EEnhanced Multi-Carbon Oxygenate 23 Selectivity. J. Am. Chem. Soc. 2017, 139, 15848-15857. 24 7.Calle-Vallejo; F.; Koper, M.T.M.; Bandarenka, A.S. Tailoring the Catalytic Activity of 25 Electrodes with Monolayer Amounts of Foreign Metals. Chem. Soc. Rev. 2013, 42, 5210-5230. 26 8.Zhou, Y.; Che, F.; Liu, M.; Zou, C.; Liang, Z.; Luna, P.D.; Yuan, H.; Li, J.; Wang, Z.; Xie, 27 28 H.; Li, H.; Chen, P.; Bladt, E.; Quintero-Bermudez, R.; Sham, T.-K.; Bals, S.; Hofkens, J.; 29 Sinton, D.; Chen, G.; Sargent, E.H. Dopant-Induced Electron Localization Drives CO2 30 Reduction to C2 Hydrocarbons. Nat. Chem. 2018, 10, 974-980. 31 9.Luo, W.; Nie, X.; Janik, M.J.; Asthagiri, A. Facet Dependence of CO2 Reduction Paths on 32 Cu Electrodes. ACS Catal. 2016, 6, 219-229. 33 10.Tang, W.; Peterson. A.A.; Varela, A.S.; Jovanov, Z.P.; Bech, L.; Durand, W.J.; Dahl, S.; 34 Nørskov, J.K.; Chorkendorff, I. The Importance of Surface Morphology in Controlling the 35 36 Selectivity of Polycrystalline Copper for CO2 Electroreduction. Phys. Chem. Chem. Phys. 2012, 37 14, 76-81. 38 11.Loiudice, A., Lobaccaro, P.; Kamali, E.A.; Thao, T.; Huang, B.H.; Ager, J.W.; Buonsanti, 39 R. Tailoring Copper Nanocrystals towards C2 Products in Electrochemical CO2 Reduction. 40 Angew. Chem. Int. Ed. 2016, 55, 5789-5792. 41 12.Yin, Z.; Peng, H.; Wei, X.; Zhou, H.; Gong, J.; Huai, M.; Xiao, L.; Wang, G.; Lu, J.; Zhuang, 42 L. An Alkaline Polymer Electrolyte CO2 Electrolyzer Operated with Pure Water. Energy & 43 44 Environ. Sci. 2019, 12, 2455-2462. 45 13.Dinh, C.-T.; Burdyny, T.; Kibria, M.G.; Seifitokaldani, A.; Gabardo, C.M.; Arquer, F.P.G.; 46 Kiani, A.; Edwards, J.P.; Luna, P.D.; Bushuyev, O.; Zou, C.; Quintero-Bermudez, R.; Pang, 47 Y.; Sinton, D.; Sargent, E.H. CO2 Electroreduction to Ethylene via Hydroxide-Mediated 48 Copper Catalysis at an Abrupt Interface. Science 2018, 360, 783-787. 49 14.Xia, C.; Zhu, P.; Jiang, Q.; Pan, Y.; Liang, W.; Stavitski, E.; Alshareef, H.N.; Wang, H. 50 51 Continuous Production of Pure Liquid Fuel Solutions via Electrocatalytic CO2 Reduction using 52 Solid-Electrolyte Devices. Nat. Energy, 2019, 4, 776-785. 53 15.Chen, C.; Kotyk, J.F.K.; Sheehan, S.W. Progress toward Commercial Application of 54 Electrochemical Carbon Dioxide Reduction. Chem 2018, 4, 2571-2586. 55 16.Möller, T.; Ju, W.; Bagger, A.; Wang, X.; Luo, F.; Thanh, T.N.; Varela, A.S.; Rossmeisl, 56 J.; Strasser, P. Efficient CO2 to CO Electrolysis on Solid Ni–N–C Catalysts at Industrial 57 Current Densities. Energy & Environ. Sci. 2019, 12, 640-647. 58 59 60 35 ACS Paragon Plus Environment ACS Catalysis Page 36 of 42 1 2 3 17.Jeon, H.S.; Sinev, I.; Scholten, F.; Divins, N.J.; Zegkinoglou, I.; Pielsticker, L.; Cuenya, 4 5 B.R. Operando Evolution of the Structure and Oxidation State of Size-Controlled Zn 6 Nanoparticles during CO2 Electroreduction. J. Am. Chem. Soc. 2018, 140, 9383-9386. 7 18.Dutta, A.; Morstein, C.E.; Rahaman, M.; López, A.C.; Broekmann, P. Beyond Copper in 8 CO2 Electrolysis: Effective Hydrocarbon Production on Silver-Nanofoam Catalysts. ACS 9 Catal. 2018, 8, 8357-8368. 10 19.Cao, Z.; Zacate, S.B.; Sun, X.; Liu, J.; Hale, EE.M.; Carson, W.P.; Tyndall, S.B.; Xu, J.; 11 Liu, X.; Liu, X.; Song, C.; Luo, J.-H.; Cheng, M.J.; Liu, W. Tuning Gold Nanoparticles with 12 13 Chelating Ligands for Highly Efficient Electrocatalytic CO2 Reduction. Angew. Chem. Int. Ed. 14 2018, 57, 12675-12679. 15 20.Yuan, T.; Hu, Z.; Zhao, Y.; Fang, J.; Lv, J.; Zhang, Q.; Zhuang, Z.; Gu, L.; Hu, S. Two- 16 Dimensional Amorphous SnOx from Liquid Metal: Mass Production, Phase Transfer and 17 Electrocatalytic CO2 Reduction towards Formic Acid. Nano Lett. 2020, 4, 2916-2922. 18 21.Ju, W.; Bagger, A.; Hao, G.-P.; Varela, A.S.; Sinev, I.; Bon, V.; Cuenya, B.R.; Kaskel, S.; 19 Rossmeisl, J.; Strasser, P. Understanding Activity and Selectivity of Metal-Nitrogen-Doped 20 21 Carbon Catalysts for Electrochemical Reduction of CO2. Nat. Commun. 2017, 8, 944. 22 22.Li, X.; Bi, W.; Chen, M.; Sun, Y.; Ju, H.; Yan, W.; Zhu, J.; Wu, X.; Chu, W.; Wu, C.; Xie, 23 Y. Exclusive Ni–N4 Sites Realize Near-Unity CO Selectivity for Electrochemical CO2 24 Reduction. J. Am. Chem. Soc. 2017, 139, 14889-14892. 25 23.Pan, Y.; Lin, R.; Chen, Y.; Liu, S.; Zhu, W.; Cao, X.; Chen, W.; Wu, K.; Cheong, W.-C.; 26 Wang, Y.; Zheng, L.; Luo, J.; Lin, Y.; Liu, Y.; Liu, C.; Li, J.; Lu, Q.; Chen, X.; Wang, D.; 27 28 Peng, Q.; Chen, C.; Li, Y. Design of Single-Atom Co–N5 Catalytic ite: aA obust 29 eElectrocatalyst for CO2 Reduction with Nearly 100% CO Selectivity and Remarkable 30 Stability. J. Am. Chem. Soc. 2018, 140, 4218-4221. 31 24.Wang, X.; Chen, Z.; Zhao, X.; Yao, T.; Chen, W.; You, R.; Zhao, C.; Wu, G.; Wang, J.; 32 Huang, W.; Yang, J.; Hong, X.; Wei, S.; Wu, Y.; Li, Y. Regulation of Coordination Number 33 over Single Co Sites: Triggering the Efficient Electroreduction of CO2. Angew. Chem. Int. Ed. 34 2018, 57, 1944-1948. 35 36 25. Li, J.; Prslja, P.; Shinagawa, T.; Fernandez, A.J.M.; Krumeich, F.; Artyushkova, K.; 37 Atanassov, P.; Zitolo, A.; Zhou, Y.; García-Muelas, R.; LopÀez, N.; Perez-Ramírez, J.; Jaouen, 38 F., Volcano Trend in Electrocatalytic CO2 Reduction Activity over Atomically Dispersed 39 Metal ites on niNtrogen-Doped Carbon. ACS Catal. 2019, 9, 10426–10439. 40 26.Genovese, C.; Schuster M.E.; Gibson E.K.; Gianolio, D.; Posligua, V.; Grau-Crespo, R.; 41 Cibin, G.; Wells, P.P.; Garai, D.; Solokha, V.; Calderon, S.K.; Velasco-Velez, J.J.; Ampelli, 42 C.; Perathoner, S.;Held, G.; Centi, G.; Arrigo, R. Operando Spectroscopy Study of the Carbon 43 44 Dioxide Electro-Reduction by Iron Species on Nitrogen-Doped Carbon. Nat. Commun. 2018, 45 9, 935. 46 27. De, R.; Gonglach, S.; Paul, S.; Haas, M.; Sreejith, S.S.; Gerschel, P.; Apfel, U.-P.; Vuong, 47 T.-H.; Rabeah, J.; Roy, S.; Schöfberger, W. Electrocatalytic Reduction of CO2 to Acetic Acid 48 by a Molecular Manganese Corrole Complex. Angew. Chem. Int. Ed. 2020, 59, 10527-10534. 49 28.Lv, K.; Lv, K.; Fan, Y.; Zhu, Y.; Yuan, Y.; Wang, J.; Zhu, Y.; Zhang, Q. Elastic Ag- 50 51 Anchored N-Doped Graphene/Carbon Foam for the Selective Electrochemical Reduction of 52 Carbon Dioxide to Ethanol. J. Mater. Chem. A 2018, 6, 5025-5031. 53 29.Wang, Y.; Wang, Z.; Dinh, C.-T.; Li, J.; Ozden, J.; Kibria, M.G.; Seifitokaldani, A.; Tan, 54 C.-S.; Gabardo, C.-M.; Luo, M.; Zhou, H.; Li, F.; Lum, Y.; McCallum, C.; Xu,Y.; Liu, M.; 55 Proppe, A.; Johnston, A.; Todorovic, P.; Zhuang, T.-T.; Sinton, D.; Kelley, S.O.; Sargent, E.H. 56 Catalyst Synthesis under CO2 Electroreduction Favours Faceting and Promotes Renewable 57 Fuels Electrosynthesis. Nat. Catal. 2020, 3, 98-106. 58 59 30.Liang, Z.-Q.; Zhuang, T.-T.; Seifitokaldani, A.; Li, J.; Huang, C.-W.; Tan, C.-S.; Li, Y.; 60 Luna, P.D.; Dinh, C.-T.; Hu, Y.; Xiao, Q.; Hsieh, P.-L.; Wang, Y.; Li, F.; Quintero-Bermudez, 36 ACS Paragon Plus Environment Page 37 of 42 ACS Catalysis 1 2 3 R.; Zhou, Y.; Chen, P.; Pang, Y.; Lo, S.-C.; Chen, L.-J.; Tan, H.; Xu, Z.; Zhao, S.; Sinton, D.; 4 5 Sargent, E.H. Copper-on-Nitride Enhances the Stable Electrosynthesis of Multi-Carbon 6 Products from CO2. Nat. Commun. 2018, 9, 3828. 7 31.Luna, P.D.; Quintero-Bermudez, R.; Dinh, C.-T.; Ross, M.B.; Bushuyev, O.S.; Todorović, 8 P.; Regier, T.; Kelley, S.O.; Yang, P.; Sargent, E.H. Catalyst Electro-Redeposition Controls 9 Morphology and Oxidation State for Selective Carbon Dioxide Reduction. Nat. Catal. 2018, 10 1, 103-110. 11 32.Mistry, H.; Varela, A.S.; Bonifacio, C.S.; Zegkinoglou, I.; Sinev, I.; Choi, Y.-W.; Kisslinger, 12 13 K.; Stach, E.A.; Yang, J.C.; Strasser, P.; Cuenya, B.R. Highly Selective Plasma-Activated 14 Copper Catalysts for Carbon Dioxide Reduction to Ethylene. Nat. Commun. 2016, 7, 12123. 15 33.Ting, L.R.L.; Piqué, O.; Lim, S.Y.; Tanhaei, M.; Calle-Vallejo, F.; Yeo, B.S. Enhancing 16 CO2 Electroreduction to Ethanol on Copper–Silver Composites by Opening an Alternative 17 Catalytic Pathway. ACS Catal. 2020, 10, 4059-4069. 18 34.Ren, D.; Ang, B.S.-H.; Yeo, B.S. Tuning the Selectivity of Carbon Dioxide Electroreduction 19 toward Ethanol on Oxide-Derived CuxZn Catalysts. ACS Catal. 2016, 6, 8239-8247. 20 21 35.Morales-Guio, C.G.; Cave, E.R.; Nitopi, S.A.; Feaster, J.T.; Wang, L.; Kuhl, K.P.; Jackson, 22 A.; Johnson, N. C.; Abram, D.N.; Hatsukade, T.; Hahn, C.; Jaramillo, T.F. Improved CO2 23 Reduction Activity towards C2+ Alcohols on a Tandem Gold on Copper Electrocatalyst. Nat. 24 Catal. 2018, 1, 764-771. 25 36.Li, F.; Li, Y.C.;Wang, Z.; Li, J.; Nam, D.-H.; Lum, Y.; Luo, M.; Wang, X.; Ozden, A.; 26 Hung, S.-F.; Chen, B.; Wang, Y.; Wicks, J.; Xu, Y.; Li, Y.; Gabardo, C.M.; Dinh, C.-T.; Wang, 27 28 Y.; Zhuang, T.-T.; Sinton, D.; Sargent, E.H. Cooperative CO2-to-Ethanol Conversion via 29 Enriched Intermediates at Molecule–Metal Catalyst Interfaces. Nat. Catal. 2020, 3, 75-82. 30 37.Montoya, J.H.; Shi, C.; Chan, K.; Nørskov, J.K. Theoretical Insights into A CO 31 Dimerization Mechanism in CO2 Electroreduction. J. Phys. Chem. Lett. 2015, 6, 2032-2037. 32 38.Vasileff, A.; Xu, C.; Jiao, Y.; Zheng, Y.; Qiao, S.-Z. Surface and Interface Engineering in 33 Copper-Based Bimetallic Materials for Selective CO2 Electroreduction. Chem 2018, 4, 1809- 34 1831. 35 36 39.Liu, K.; Ma, M.; Wu, L.; Valenti, M.; Cardenas-Morcoso, D.; Hofmann, J.P.; Bisquert, J.; 37 Gimenez, S.; Smith, W.A. Electronic Effects Determine the Selectivity of Planar Au–Cu 38 Bimetallic Thin Films for Electrochemical CO2 Reduction. ACS Appl. Mater. & Interfaces 39 2019, 11, 16546-16555. 40 40.Kim, D.; Resasco, J.; Yu, Y.; Asiri A.M.; Yang, P. Synergistic Geometric and Electronic 41 Effects for Electrochemical Reduction of Carbon Dioxide Using Gold–Copper Bimetallic 42 Nanoparticles. Nat. Commun. 2014, 5, 4948. 43 44 41.Huang, J.; Mensi, M.; Oveisi, E.; Mantella, V.; Buonsanti, R. Structural Sensitivities in 45 Bimetallic Catalysts for Electrochemical CO2 Reduction Revealed by Ag–Cu Nanodimers. J. 46 Am. Chem. Soc. 2019, 141, 2490-2499. 47 42.Jeon, H.S.; Timoshenko, J.; Scholten, F.; Sinev, I.; Herzog, A.; Haase, F.T.; Cuenya, B.R. 48 Operando Insight into the Correlation between the Structure and Composition of CuZn 49 Nanoparticles and their Selectivity for the Electrochemical CO2 Reduction. J. Am. Chem. Soc. 50 51 2019, 141, 19879-19887. 52 43.Eren, B.; Zherebetskyy, D.; Patera, L.L.; Wu, C.H.; Bluhm, H.; Africh, C.; Wang, L.W.; 53 Somorjai, G.A.; Salmeron, M. Activation of Cu (111) Surface by Decomposition into 54 Nanoclusters Driven by CO Adsorption. Science 2016, 351, 475-478. 55 44.Kim, Y.-G.; Baricuatro, J.H.; Javier, A.; Gregoire, J.M.; Soriaga, M.P.The Evolution of the 56 Polycrystalline Copper Surface, First to Cu (111) and then to Cu (100), at A Fixed CO2RR 57 Potential: A study by Operando EC-STM. Langmuir 2014, 30, 15053-15056. 58 59 45.Wang, X.; Araújo, J.F.; Ju, W.; Bagger, A.; Schmies, H.; Kühl, S.; Rossmeisl, J.; Strasser, 60 P. Mechanistic eaction paPthways of Enhanced Ethylene Yields during Electroreduction of CO 37 ACS Paragon Plus Environment ACS Catalysis Page 38 of 42 1 2 3 2–CO Co-Feeds on Cu and Cu-Tandem Electrocatalysts. Nat. Nanotechnol, 2019, 14, 1063- 4 5 1070. 6 46. Luna, P.D.; Hahn, C.; Higgins, D.; Jaffer, S.A.; Jaramillo, T.; Sargent, E.H., What Would 7 it Take for Renewably Powered Electrosynthesis to Displace Petrochemical rocesses? Science 8 2019, 364, eaav3506. 9 47.Jouny, M.; Hutchings, G.S.; Jiao, F. Carbon mMonoxide Electroreduction as An Emerging 10 Platform for Carbon Utilization. Nat. Catal. 2019, 2, 1062-1070. 11 48.Huan, T.N.; Ranjbar, N.; Rousse, G.; Sougrati, M.; Zitolo, A.; Mougel, V.; Jaouen, F.; 12 13 Fontecave, M. Electrochemical Reduction of CO2 Catalyzed by Fe-NC Materials: A Structure– 14 Selectivity Study. ACS Catal. 2017, 7, 1520-1525. 15 49.Fan, X.; Peng, Z.; Ye, R.; Zhou, H.; Guo, X. M3C (M: Fe, Co, Ni) Nanocrystals Encased in 16 Graphene Nanoribbons: An Active and Stable Bifunctional Electrocatalyst for Oxygen 17 Reduction and Hydrogen Evolution Reactions. ACS Nano 2015, 9, 7407-7418. 18 50.Jin, Q.; Ren, B.; Li, D.; Cui, H.; Wang, C. In Situ Promoting Water Dissociation Kinetic of 19 Co Based Electrocatalyst for Unprecedentedly Enhanced Hydrogen Evolution Reaction in 20 21 Alkaline Media. Nano Energy 2018, 49, 14-22. 22 51.Zhao, C.; Dai, X.; Yao, T.; Chen, W.; Wang, X.; Wang, J.; Yang, J.; Wei, S.; Wu, Y.; Li, 23 Y. Ionic Exchange of Metal–Organic Frameworks to Access Single Nickel Sites for Efficient 24 Electroreduction of CO2. J. Am. Chem. Soc. 2017, 139, 8078-8081. 25 52.Ren, S.; Joulié, D.; Salvatore, D.; Torbensen, K.; Wang, M.; Robert, M.; Berlinguette, C.P. 26 Molecular Electrocatalysts Can Mediate Fast, Selective CO2 Reduction in A Flow Cell. Science 27 28 2019, 365, 367-369. 29 53.Zhang, Z.; Xiao, J.; Chen, X.-J.; Yu, S.; Yu, L.; Si, R.; Wang, Y.; Wang, S.; Meng, X.; 30 Wang, Y.; Tian, Z.-Q.; Deng, D. Reaction Mechanisms of Well‐Defined Metal–N4 Sites in 31 Electrocatalytic CO2 Reduction. Angew. Chem. Int. Ed. 2018, 57, 16339-16342. 32 54.Costentin, C.; Drouet, S.; Robert, M.; Savéant, J.-M. A Local Proton Source Enhances CO2 33 Electroreduction to CO by A Molecular Fe Catalyst. Science 2012, 338, 90-94. 34 55.Karapinar, D.; Huan, N.T.; Sahraie, N.R.; Li, J.; Wakerley, D.; Touati, N.; Zanna, S.; 35 36 Taverna, D.; Tizei, L.H.G.; Zitolo, A.; Jaouen, F.; Mougel, V.; Fontecave, M. Electroreduction 37 of CO2 on Single ‐ Site Copper ‐ Nitrogen ‐ Doped Carbon Material: Selective Formation of 38 Ethanol and Reversible Restructuration of the Metal Sites. Angew. Chem. Int. Ed. 2019, 58, 39 15098-15103. 40 56.Weng, Z.; Wu, Y.; Wang, M.; Jiang, J.; Yang, K.; Huo, S.; Wang, X.-F.; Ma, Q.; Brudvig, 41 G.W.; Batista, V.S.; Liang, Y.; Feng, Z.; Wang, H. Active Sites of Copper-Complex Catalytic 42 Materials for Electrochemical Carbon Dioxide Reduction. Nat. Commun. 2018, 9, 415. 43 44 57.Jaouen, F. Heat ‐ Treated Transition Metal ‐ NxCy Electrocatalysts for the O2 Reduction 45 Reaction in Acid PEM Fuel Cells. In Non‐Noble Metal Fuel Cell Catalysts; Wiley‐VCH Verlag 46 GmbH & Co. KGaA, 2014, 29-118. 47 58.Mineva, T.; Matanovic, I.; Atanassov, P.; Sougrati, M.-T.; Stievano, L.; Clémancey, M.; 48 Kochem, A.; Latour, J.-M.; Jaouen, F. Understanding Active Sites in Pyrolyzed Fe–N–C 49 Catalysts for Fuel Cell Cathodes by Bridging Density Functional Theory Calculations and 57Fe 50 51 Mössbauer Spectroscopy. ACS Catal. 2019, 9, 9359-9371. 52 59.Sougrati, M.T.; Goellner, V.; Schuppert, A.K.; Stievano, L.; Jaouen, F. Probing Active Sites 53 in Iron-Based Catalysts for Oxygen Electro-Reduction: A Temperature-Dependent 57Fe 54 Mössbauer Spectroscopy Study. Catal. Today, 2016, 262, 110-120. 55 60.Menga, D.; Ruiz‐Zepeda, F.; Moriau, L.; Šala, M.; Wagner, F.; Koyutürk, B.; Bele, M.; 56 Petek, U.; Hodnik, N.; Gaberšček, M.; Fellinger, T.-P. Active‐Site Imprinting: Preparation of 57 Fe–N–C Catalysts from Zinc Ion–Templated Ionothermal Nitrogen ‐ Doped Carbons. Adv. 58 59 Energy Mater. 2019, 9, 1902412. 60 38 ACS Paragon Plus Environment Page 39 of 42 ACS Catalysis 1 2 3 61.Mehmood, A.; Pampel, J.; Ali, G.; Ha, H.Y.; Ruiz‐Zepeda, F.; Fellinger, T.-P. Facile Metal 4 5 Coordination of Active Site Imprinted Nitrogen Doped Carbons for the Conservative 6 Preparation of Non‐Noble Metal Oxygen Reduction Electrocatalysts. Adv. Energy Mater. 2018, 7 8, 1701771. 8 62.Kamali-M, S.; Ericsson, T.; Wäppling, R. Characterization of Iron Oxide Nanoparticles by 9 Mössbauer Spectroscopy. Thin Solid Films 2006, 515, 721-723. 10 63. Li, J.; Sougrati, M.T.; Zitolo, A.; Ablett, J.M.; Oğuz, I.C.; Mineva, T.; Matanovic, I.; 11 Atanassov, P.; Huang, Y.; Zenyuk, I.; Cicco, A.D.; Kumar, K.; Dubau, L.; Maillard, F.; Dražić, 12 13 G.; Jaouen, F. Identification of durable and non-durable FeNx sites in Fe–N–C materials for 14 proton exchange membrane fuel cells. Nat. Catal. 2021, 4, 10-19. 15 64.Zboril, R.; Mashlan, M.; Petridis, D. Iron (III) Oxides from Thermal Processes Synthesis, 16 tructural and Magnetic Properties, Mössbauer Spectroscopy Characterization, and 17 Applications. Chem. Mater. 2002, 14, 969-982. 18 65.Fultz, B. Mössbauer spectrometry. In Characterization of materials, Wiley‐VCH Verlag 19 GmbH & Co. KGaA, 2002, 1-21. 20 21 66. Li, J.; Ghoshal, S.; Liang, W.; Sougrati, M.-T.; Jaouen, F.; Halevi, B.; McKinney, S.; 22 McCool, G.; Ma, C.; Yuan, X. J.; Ma, Z.-F.; Mukerjee, S.; Jia, Q. Structural and Mechanistic 23 Basis for the High Activity of Fe−N−C Catalysts toward Oxygen Reduction. Energy Environ. 24 Sci. 2016, 9, 2418−2432. 25 67. Zitolo, A.; Goellner, V.; Armel, V.; Sougrati, M.-T.; Mineva, T.; Stievano, L.; Fonda, E.; 26 Jaouen, F. Identification of Catalytic Sites for Oxygen Reduction in Iron-and Nitrogen-Doped 27 28 Graphene Materials. Nat. Mater. 2015, 14, 937-942. 29 68.Hötger, D.; Abufager, P.; Morchutt, C.; Alexa, P.; Grumelli, D.; Dreiser, J.; Stepanow, S.; 30 Gambardella, P.; Busnengo, H.F.; Etzkorn, M.; Gutzler, R.; Kern, K. On-Surface 31 Transmetalation of Metalloporphyrins. Nanoscale 2018, 10, 21116-21122. 32 69.Lee, J.A.; Newnham, C.E.; Stone, F.S.; Tye, F.L. Thermal Decomposition of Managanese 33 Oxyhydroxide. J. Solid State Chem. 1980, 31, 81-93. 34 70.Li, Y.; Liu, L.-D.; Liu, L.; Liu, Y.; Zhang, H.-W.; Han, X. Efficient Oxidation of Phenol 35 36 by Persulfate Using Manganite as A Catalyst. J. Mol. Catal. A: Chem. 2016, 411, 264-271. 37 71.Hyun, G.; Song, J.T.; Ahn, C.; Ham, Y.; Cho, D.; Oh, J.; Jeon, S. Hierarchically Porous Au 38 Nanostructures with Interconnected Channels for Efficient Mass Transport in Electrocatalytic 39 CO2 Reduction. Proc. Natl. Acad. Sci. U.S.A 2020, 117, 5680-5685. 40 72.He, Q.; Lee, J.H.; Liu, D.; Liu, Y.; Lin, Z.; Xie, Z.; Hwang, S.; Kattel, S.; Song, L.; Chen, 41 J.G. Accelerating CO2 Electroreduction to CO Over Pd Single ‐ Atom Catalyst. Adv. Funct. 42 Mater. 2020, 30, 2000407. 43 44 73.Asset, T.; Garcia, S.T.; Herrera, S.; Andersen, N.; Chen, Y.; Peterson, E.J.; Matanovic, I.; 45 Artyushkova, K.; Lee, J.; Minteer, S.D.; Dai, S.; Pan, X.; Chavan, K.; Barton, S.C.; Atanassov, 46 P. Investigating the Nature of the Active Sites for the CO2 Reduction Reaction on Carbon- 47 Based Electrocatalysts. ACS Catal. 2019, 9, 7668-7678. 48 74.Handoko, A.D.; Wei, F.; Jenndy, Yeo, B.S.; She, Z.W. Understanding Heterogeneous 49 Electrocatalytic Carbon Dioxide Reduction through Operando Techniques. Nat. Catal. 2018, 50 51 1, 922-934. 52 75.Zhu, Y.; Wang, J.; Chu, H.; Chu, Y.-H.; Chen, H.M. In Situ/Operando Studies for 53 Designing Next-generation Electrocatalysts. ACS Energy Lett. 2020, 5, 1281-1291. 54 76.Ma, L.; Hu, W.; Mei, B.; Liu, H.; Yuan, B.; Zang, J.; Chen, T.; Zou, L.; Zou, Z.; Yang, B.; 55 Yu, Y.; Ma, J.; Jiang, Z.; Wen, K.; Yang, H. Covalent Triazine Framework Confined Copper 56 Catalysts for Selective Electrochemical CO2 Reduction: Operando Diagnosis of Active Sites. 57 ACS Catal. 2020, 8, 4534-4542. 58 59 77.Nam, D.-H.; Bushuyev, O.S.; Li, J.; Luna, P.D.; Seifitokaldani, A.; Dinh, C.-T.; Arquer, 60 F.P.G.; Wang, Y.; Liang, Z.; Proppe, A.H.; Tan, C.S.; Todorović, P.; Shekhah, O.; Gabardo, 39 ACS Paragon Plus Environment ACS Catalysis Page 40 of 42 1 2 3 C.M.; Jo, J.W.; Choi, J.; Choi, M.-J.; Baek, S.-W.; Kim, J.; Sinton, D.; Kelley, S.O.; Eddaoudi, 4 5 M.; Sargent, E.H. Metal–Organic Frameworks Mediate Cu Coordination for Selective CO2 6 Electroreduction. J. Am. Chem. Soc. 2018, 140, 11378-11386. 7 78.Möller, T.; Scholten, F.; Thanh, T.N.; Sinev, I.; Timoshenko, J.; Wang, X.L.; Jovanov, Z.; 8 Gliech, M.; Cuenya, B.R.; Varela, A.S.; Strasser, P. Electrocatalytic CO2 Reduction on CuOx 9 Nanocubes Tracking the Evolution of Chemical State, Geometric Structure, and Catalytic 10 Selectivity using Operando Spectroscopy. Angew. Chem. Int. Ed. 2020, 59, 17974-17983. 11 79.Adak, S.; Hartl, M.; Daemen, L.; Fohtung, E.; Nakotte, H. Study of Oxidation States of the 12 13 Transition Metals in a Series of Prussian Blue Analogs using X-Ray Absorption Near Edge 14 Structure (XANES) Spectroscopy. J. Electron Spectros. Relat. Phenomena 2017, 214, 8-19. 15 80.Gardner, T.H.; Spivey, J.J.; Kugler, E.L.; Campos, A.; Hissam, J.C.; Roy, A.D. Structural 16 Characterization of Ni-Substituted Hexaaluminate Catalysts using EXAFS, XANES, XPS, 17 XRD, and TPR. J. Phys. Chem. C 2010, 114, 7888-7894. 18 81.Friebel, D.; Louie, M.W.; Bajdich, M.; Sanwald, K.E.; Cai, Y.; Wise, A.M.; Cheng, M.-J.; 19 Sokaras, D.; Weng, T.-S.; Alonso-Mori, R.; Davis, R.C.; Bargar, J.R.; Nørskov, J.K.; Nilsson, 20 21 A.; Bell, A.T. Identification of Highly Active Fe Sites in (Ni, Fe)OOH for Electrocatalytic 22 Water Splitting. J. Am. Chem. Soc. 2015, 140, 1305-1313. 23 82.Jiang, J.; Sun, F.; Zhou, S.; Hu, W.; Zhang, H.; Dong, J.; Jiang, Z.; Zhao, J.; Li, J.; Yan, W.; 24 Wang, M. Atomic-Level Insight into Super-Efficient Electrocatalytic Oxygen Evolution on 25 Iron and Vanadium Co-Doped Nickel (oxy) ydroxide. Nat. Commun. 2018, 9, 2885. 26 83.Scheinost, A.C.; Sparks, D.L. Formation of Layered Single-and Double-Metal Hydroxide 27 28 Precipitates at the Mineral/Water Interface: A Multiple-Scattering XAFS Analysis. J. Colloid 29 Interface Sci. 2000, 223, 167-178. 30 84.Hall, D.S.; Lockwood, D.J.; Bock, C.; MacDougall, B.R. Nickel Hydroxides and Related 31 Materials: A Review of Their Structures, Synthesis and Properties. Proc. R. Soc. A 2015, 471. 32 85.Chen, Y.; Bylaska, E.J.; Weare, J.H. Weakly Bound Water Structure, Bond Valence 33 Saturation and Water Dynamics at the Goethite (100) Surface/Aqueous Interface: Ab Initio 34 Dynamical Simulations. Geochemical Trans. 2017, 18, 3. 35 36 86.Ma, R.; Liu, Z.; Takada, K.; Fukuda, K.; Ebina, Y.; Bando, Y.; Sasaki, T.; Tetrahedral 37 Co(II) Coordination in α-Type Cobalt Hydroxide: Rietveld Refinement and X-Ray Absorption 38 Spectroscopy. Inorg. Chem. 2006, 45, 3964-3969. 39 87.Dunwell, M.; Lu, Q.; Heyes, J.M.; Rosen, J.; Chen, J.G.; Yan, Y.; Jiao, F.; Xu, B. The 40 Central Role of Bicarbonate in the Electrochemical Reduction of Carbon Dioxide on Gold. J. 41 Am. Chem. Soc. 2017, 139, 3774-3783. 42 88.Zhu, S.; Jiang, B.; Cai, W.-B.; Shao, M. Direct Observation on Reaction Intermediates and 43 44 the Role of Bicarbonate Anions in CO2 Electrochemical Reduction Reaction on Cu Surfaces. 45 J. Am. Chem. Soc. 2017, 139, 15664-15667. 46 89.Bergmann, A.; Martinez-Moreno, E.; Teschner, D.; Chernev, P.; Gliech, M.; Araújo, J.F.; 47 Reier, T.; Dau, H.; Strasser, P. Reversible Amorphization and the Catalytically Active State of 48 Crystalline Co3O4 during Oxygen Evolution. Nat. Commun. 2015, 6, 8625. 49 90.Zhou, H.; Yu, F.; Sun, J.; He, R.; Chen, S.; Chu, C.-W.; Ren, Z. Highly Active Catalyst 50 51 Derived from a 3D Foam of Fe(PO3)2/Ni2P for Extremely Efficient Water Oxidation. Proc. 52 Natl. Acad. Sci. U.S.A 2017, 114, 5607-5611. 53 54 55 56 57 58 59 60 40 ACS Paragon Plus Environment Page 41 of 42 ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 TOC 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 41 ACS Paragon Plus Environment ACS Catalysis Page 42 of 42 1 CO M 2 O N 3 CO2 H2O 4 ACS Paragon Plus Environment Inserted Anion 5 6