A generalized amino-modification strategy to boost CO2

14 Although Faraday efficiency (FE) for CO production of single-atom catalysts 15 immobilized on nitrogen-doped carbon supports (M-N/C) for CO2 electrocatalytic 16 reduction reaction (CO2RR) is generally over 90%, M-N/C catalysts demonstrate a poor 17 reaction current density, much worse than the current density of industrial level. Herein, 18 we first report a generalized strategy of amino-functionalized carbon supports to 19 regulate electronic structure of M-N/C catalysts (M=Ni, Fe, Zn) to significantly 20 increase current density of CO production. The aminated Ni single-atom catalyst 21 achieves a remarkable CO partial current density of 447.6 mA cm (a total current 22 density over 500 mA cm) with a nearly 90% CO FE at a moderate overpotential of 23 0.89 V, and especially CO FE can be maintained over 85% in a wide operating potential 24 range from -0.5 V to -1.0 V. DFT calculations and experimental researches demonstrate 25 that the superior activity is attributed to enhanced adsorption energies of CO2* and 26 COOH* intermediates caused by the change of electronic structure of aminated 27 catalysts. This work provides an ingenious method for significantly increasing current 28 density at industrial-relevant level of single-atom catalysts for CO2RR. 29


Introduction
Using renewable electricity to drive CO2 electrocatalytic reduction to high value-2 added chemical fuels can not only solve the problem of excessive CO2 emission, but 3 also can achieve direct conversion of intermittent electric energy to chemical energy, 4 which is of great significance to control carbon balance and optimize energy 5 consumption structure. 1,2 Due to the stable binding of C=O bond in CO2 molecules, 6 CO2 electrocatalytic reduction reaction (CO2RR) requires a high energy barrier, and 7 moreover, the competitive hydrogen evolution reaction (HER) is more favored than 8 CO2RR in kinetics. 3-5 Therefore, a large number of catalysts have been developed in 9 the study of CO2RR in recent years, including metals and oxides, 6-10 nonmetal, [11][12][13][14][15] 10 single-atom, 3, 16-26 and ultra-thin two-dimensional metal, [27][28][29] and molecule catalysts. 30

11
Among these catalysts, transition metal single-atom catalysts immobilized on applications. For example, CO partial current density of most Fe-N/C catalysts is less 19 than 15 mA cm -2 , 23, 37 and Zn-N/C is less than 20 mA cm -2 , 21,38 Ni-N/C is less than 25 20 mA cm -2 , 20, 24, 26, 39 Co-N/C is less than 35 mA cm -2 at a catalytic potential of -0.8 V (vs. 21 RHE). 25,35,40 The reason for a low current density is the limited adsorption capacity of 22 catalysts for reaction intermediates, resulting in the concentration of reactants adsorbed 23 on catalysts surface cannot meet the required concentration for reactions at more 24 negative catalytic potentials, thus CO2RR reaches mass transfer limit. Therefore, 25 enhancing mass transfer of reactions is the most effective way to increase current 26 density, such as modifying catalysts surface, 41, 42 introducing defects, 43 regulating 27 catalysts electronic structure, 19,44 or optimizing the reactors (using a gas-fed flow cell 28 or membrane-electrode assembly cell (MEA) to replace conventional H-type cell). 8

10
Synthesis and characterizations of Ni-N 4 /C-NH 2 catalysts. Ni-N4/C-NH2 was 11 prepared by a two-step method. In brief, Ni-N4/C was obtained by pyrolyzing Ni doped 12 ZIF-8 precursor at high temperature firstly, and then Ni-N4/C was aminated by 13 impregnation and hydrothermal methods in ammonia water mixture to synthesize final for Ni K-edge of Ni foil, NiO, NiPc, and Ni-N 4 /C-NH 2 (g). Fitting for EXAFS spectrum of Ni-N 4 /C-4 NH 2 , inset is Ni-N 4 /C-NH 2 structure (h). 5 structure is observed for Ni-N4/C ( Figure S4). These results reinforce that Ni atoms still 6 retained atomic dispersion after amino-modification. Noteworthily, Ni-N4/C-NH2 7 exhibits similar D4h centro-symmetry to that of standard reference sample Ni 2+ Pc as 8 revealed by the sharp absorption peaks at 8337.5 eV ( Figure S5) in XANES spectrum, 9 implying that Ni-N4/C-NH2 has a similar planar local structure with Ni 2+ Pc, in which 10 Ni atoms are coordinated by four pyrrolic-type N atoms ( Figure S17). 3,54 Fitting results 11 of extended XAFS spectra also reveal that Ni center in Ni-N4/C-NH2 adopts a planar 12 structure with an average Ni-N coordination numbers of 3.8 ( Figure 1h, Table S1). As 13 shown in XANES spectrum, isolated Ni atoms exhibit a unique electronic structure with 14 valence states between Ni 0 (Ni foil) and Ni 2+ (Ni 2+ Pc) ( Figure S4), which is in good agreement with XPS results ( Figure S6). products were quantitatively analyzed by on-line GC and HPLC, respectively. 8 Electrocatalytic activity was first investigated by LSV tests (Figure 2a). Apparently, Ni-9 N4/C-NH2 demonstrates a larger current density in CO2-saturated electrolyte than that 10 in N2-saturated electrolyte. Moreover, a more positive onset reduction potential and a 11 larger current density of Ni-N4/C-NH2 in CO2-saturated electrolyte than that of Ni-N4/C 12 was observed, indicating Ni-N4/C-NH2 is indeed active for CO2RR. Then CO2RR that of Ni-N4/C is 98.1%. Interestingly, CO FE of Ni-N4/C-NH2 maintains over 90% in a wide operating potential window from 0.6 V to 0.9 V, which leads to a significant 1 enhanced CO partial current density (jCO, Figure 2c), reaching 63.6 mA cm -2 at 1.0 V, 2 which is 2.5 times that of Ni-N4/C. Additionally, a stable current density with CO FE 3 of over 90% was observed for Ni-N4/C-NH2 during a long-term CO2RR of 10 h at -0.8 4 V (Figure 2d).

5
To shed light on the effect of amino-modification on other single-atom catalysts, 6 we also modified Fe and Zn single-atom catalysts by the same method. Similar to Ni-7 N4/C-NH2, CO2 reduction activity of both catalysts, especially CO partial current 8 density, was significantly improved after being amino-modification ( Figure S9 and 9 S10), implying that amino-functionalized carbon supports is a generalized strategy for 10 boosting CO2RR activity of single-atom catalysts.  Figure S11). A maximum CO FE of 89.3% was observed for Ni-N4/C-NH2 at -0.9 V, 2 and especially note that CO FE keeps above 85% over a wide potential range from -0.5 3 V to -1.0 V (Figure 3b). Accordingly, Ni-N4/C-NH2 achieves a remarkable CO partial 4 current density of 447.6 mA cm -2 at 1.0 V, which is 7.0 times that in an H-type cell.

5
Compared with previously reported catalysts, Ni-N4/C-NH2 is one of the state-of-the-6 art single-atom catalysts for CO2RR (Detailed results of comparison of catalytic activity 7 are shown in Table S2). Although current density fluctuated regularly and slightly (~20  Figure S20).

18
To further elucidate the origin for high activity of Ni-N4/C-NH2, DFT calculations 19 were performed by using CP2K package, and the structural model of Ni-N4/C-NH2 is 20 shown in Figure S17. We first calculated the effect of amino-modification on free energies of intermediates in the reaction path. It can be seen that electrocatalytic 1 reduction of CO2 to CO includes a two-step electron-proton transfer process, involving 2 the key intermediate COOH* (protonation of oxygen atom in CO2) (Figure 4a).

3
Encouragingly, after being modified with amino groups, free energy (ΔG) for COOH* 4 formation decreases by 0.57 eV, whereas ΔG to form H* formation (intermediate for 5 H2 formation) increased by 0.38 eV (Figure 4a and Figure S18), demonstrating Ni-6 N4/C-NH2 is favorable for CO formation, which coincides well with our experimental 7 results. A higher energy barrier to form COOH* than to form CO* are observed for 8 both catalysts, indicating CO2 activation to form COOH* is the RDS, which agrees well 9 with the Tafel slope ( Figure S14). Furthermore, a higher selectivity to CO of Ni-N4/C-10 NH2 also can be inferred from a more positive difference value of limiting potentials 11 (ΔPlimit) for CO2RR and HER ( Figure S19). Besides, we calculated charge distribution 12 to investigate the effect of amino-modification on electronic structure of catalysts.  Figure S15 and S16). Additionally, the enhanced adsorption energy of COOH* can be In conclusion, we provided an ingenious amino-modification method for 5 significantly increasing the current density of CO2RR of Ni single-atom catalyst at 6 industrial level, demonstrating a remarkable CO partial current density of 447.6 mA 7 cm -2 with a nearly 90% CO FE. Even more remarkably, amino-modification can not 8 only increase the current density of Ni single-atom catalyst but also has been identified 9 as a general strategy to enhance the current density of other single-atom catalysts, such

Synthesis of Ni-N 4 /C-NH 2 .
In a general procedure, as-prepared Ni-N 4 /C (0.10 g) was firstly 28 mechanically mixed with carbamide (5.00 g) by an agitated mortar for 30 min, and then the fine 29 powder mixture was pyrolyzed at 450 °C for 3 h with a heating rate 5 °C min-1 in a stream (15 30 sccm) of NH 3 . Next, the as-obtained catalyst was impregnated with the mixed solution of ammonia 31 (25%) and ethanol (V:V=2:1, 50 mL) and stirred for 24 h. Subsequently, the obtained precipitate 32 was dispersed in ammonia water (35 mL) and then was sealed in a Teflon-lined stainless steel 33 autoclave and heated to 150 °C for 12 h. Finally, the suspension was centrifuged, and then the precipitate was washed with deionized water and dried in vacuum at 60 °C for 24 h to obatain Ni-1 N 4 /C-NH 2 catalyst. As-prepared catalysts were directly used without any post-treatment. 2 Preparation of working electrodes. The substrate electrode using in H-type cell was fabricated by 3 carbon cloth (1×1 cm), which was sonicated in hydrochloric acid (10 M), acetone, and deionized 4 water for 30 min, respectively. Typically, powder catalysts (12.0 mg) were dispersed in Nafion 5 perfluorinated resin solution (5 wt%, 120 μL) and isopropanol (600 μL) by ultrasonication for 30 6 min to form homogeneous catalysts ink. Then the as-prepared catalyst ink was sprayed onto the 7 carbon cloth several times with a 20 μL pipette, using 360 μL for each electrode. Finally, the 8 obtained composite electrodes were dried at 50 °C for 12 h, and the catalyst loading is 3.0 ± 0.1 mg 9 cm -1 . 10 Carbon paper (Freudenberg H14C9) with a Micro Porous layer (MPL) and hydrophobic treatment 11 (PTFE) was used as gas diffusion layer (GDL). Powder catalyst was coated on the MPL face of 12 GDL by the same method as the carbon cloth electrode using in H-type cell, and the catalyst-13 supported GDL was used as GDE in flow cell. 14

Electrochemical experiments in H-type cell. Electrochemical experiments were performed in a 15
gas-tight H-cell containing 25 mL of electrolyte, which was separated by a proton exchange 16 membrane (Nafion N117). Ag/AgCl (saturated KCl) electrode and Pt plate (1×1 cm) were used as 17 reference and counter electrodes, respectively. Electrochemical data were recorded on a CHI660E 18 electrochemical workstation. Before each experiment, CO 2 (99.999%) was continuously bubbled 19 into electrolyte for 30 minutes to eliminate O 2 , and saturate electrolyte with CO 2 . Electrochemical 20 experiments were measured at room temperature (25±3 °C), and all potentials reported in this paper 21 are referenced to reversible hydrogen electrode (RHE, E RHE =E Ag/AgCl +0.197+0.0591×pH), and an 22 automatic iR compensation (85%) was used. The pH of N 2 and CO 2 -saturated 0.5 M KHCO 3 23 electrolyte in this study is 8.56 and 7.33, respectively. Linear sweep voltammetry (LSV) was carried 24 out in CO 2 -saturated or N 2 -saturated 0.5 M KHCO 3 electrolyte with a scan rate of 50 mV s -1 . 25 Electrochemical impedance spectroscopy (EIS) was recorded at -1.0 V in CO 2 -saturated 0.5 M 26 KHCO 3 electrolyte with an amplitude of 5 mV, and the frequency range is from 0.1 Hz to 10000 27 Hz. Tafel plot (overpotential versus log j CO ) was derived from the controlled potential electrolysis 28 results. Electrocatalytic reduction of CO 2 was investigated for 30 min at each applied potential by 29 controlled potential electrolysis method. Prior to each new electrolysis , 50 cycles of cyclic 30 voltammetry (CV) were used to activate the catalysts. Stirring (1000 rpm) was applied during CO 2 31 electrocatalytic reduction. method, and an automatic iR compensation (85%) was used. Catalyst-supported GDE, Ag/AgCl (saturated KCl) electrode equipped with a salt bridge, and squashed nickel foam (0.2 mm thickness, 1 400 mesh) were used as cathode (for CO 2 reduction), reference electrode and anode (for O 2 2 evolution), respectively. Cathode and anode are respectively connected with copper tape (current 3 collector).Three chambers of catholyte, anolyte and CO 2 gas diffusion were made of 4 polytetrafluoroethylene (PTFE), and the depth of each chamber is 0.3 cm. Each chamber had an 5 inlet and an outlet for electrolyte or CO 2 gas, and reference electrode was fixed in catholyte chamber. 6 Cathode were placed between CO 2 gas diffusion chamber and electrolyte chamber, and an anionic 7 exchange membrane (Fumasep FAB-PK-130) was interposed between anolyte and catholyte 8 chamber. Cathode area was controlled to 1 cm 2 (2.0×0.5 cm) when assembling the cell, and a 9 silicone gasket was placed between each GDE, anionic exchange membrane and electrolyte 10 chamber for sealing. 300 mL of catholyte (1 M KOH) was circulated in cathode chamber by means 11 of a peristaltic pump at a constant flow 35 mL min -1 . It should be emphasized that anolyte (1 M 12 KOH) was circulated through anode chamber by using a specially-made gas-liquid mixed flow 13 pump instead of conventional peristaltic pump, which can effectively remove O 2 produced in anode 14 chamber in time, and this is one of key conditions for obtaining ultra-high current density in this 15 flow cell configuration. High purity CO 2 was purged in the back chamber of catholyte chamber at a 16 constant flow 30 sccm by mean of a digital mass flow controller, and CO 2 output was connected to 17 a GC system. Due to the limitation of current range of electrochemical workstation, the actual area 18 of GDEs was controlled to 0.5 cm 2 (1.0×0.5 cm) when catalytic activity was tested at potentials of 19 -0.9 and -1.0 V. 20 copper-based bimetallic materials for selective CO 2 electroreduction. Chem 4, 1809-1831