Lewis-base ligand-reshaped interfacial hydrogen-bond network boosts CO2 electrolysis

ABSTRACT Both the catalyst and electrolyte strongly impact the performance of CO2 electrolysis. Despite substantial progress in catalysts, it remains highly challenging to tailor electrolyte compositions and understand their functions at the catalyst interface. Here, we report that the ethylenediaminetetraacetic acid (EDTA) and its analogs, featuring strong Lewis acid-base interaction with metal cations, are selected as electrolyte additives to reshape the catalyst-electrolyte interface for promoting CO2 electrolysis. Mechanistic studies reveal that EDTA molecules are dynamically assembled toward interface regions in response to bias potential due to strong Lewis acid-base interaction of EDTA4–-K+. As a result, the original hydrogen-bond network among interfacial H2O is disrupted, and a hydrogen-bond gap layer at the electrified interface is established. The EDTA-reshaped K+ solvation structure promotes the protonation of *CO2 to *COOH and suppressing *H2O dissociation to *H, thereby boosting the co-electrolysis of CO2 and H2O toward carbon-based products. In particular, when 5 mM of EDTA is added into the electrolytes, the Faradaic efficiency of CO on the commercial Ag nanoparticle catalyst is increased from 57.0% to 90.0% at an industry-relevant current density of 500 mA cm−2. More importantly, the Lewis-base ligand-reshaped interface allows a range of catalysts (Ag, Zn, Pd, Bi, Sn, and Cu) to deliver substantially increased selectivity of carbon-based products in both H-type and flow-type electrolysis cells.


INTRODUCTION
Electrocatalysis, powered by electricity produced from renewable energy sources, is central to enable the conversion of earth-abundant molecules (typically H 2 O, O 2 , N 2 and CO 2 ) into carbon-neutral fuels and chemicals [1 -5 ].The electrode-electrolyte interface where electrocatalytic reactions occur, buried between solid-catalysts and electrolytes, involves complicated processes of electron transfer and mass diffusion under an applied electric field [6 ,7 ].Understanding the interfacial organization and possible interfacial interactions [8 -10 ], such as those between the electrocatalysts and electrolytes or among electrolyte components, is essential for improving electrochemical performance via the co-optimisation of electrocatalysts and electrolytes.
CO 2 electrolysis has emerged as a promising approach to upgrade rich carbon resources into value-added chemical feedstocks [11 ,12 ].Currently, most of the reported studies focus on developing catalysts with well-designed structure, morphology and composition in order to enhance their intrinsic activities and/or increase the number of active sites [13 ,14 ].However, the role of electrolyte components is equally crucial but extraordinarily under-researched [15 ,16 ].The employment of electrode modification and electrolyte additives at the electrode-electrolyte interface has been shown to enable improved selectivity toward carbon-based products [17 -28 ].These studies highlight new opportunities to harness electrolytes to steer activit y and selectivit y.Even so, many open questions remain regarding the molecule-level picture of interfacial organization and dynamic evolution, intrinsic interfacial interactions, and their functions toward electrocatalytic processes [7 ].
For CO 2 electrolysis in aqueous electrolytes, H 2 O molecules acting as proton sources are usually involved in the proton-coupled electron transfer (PCET) process, but inevitably lead to an undesired hydrogen evolution reaction (HER) [29 ,30 ].The interfacial hydrogen-bond (H-bond) network is the highway for proton transfer from bulk to electrode surface and therefore greatly influences hydrogenrelated reaction kinetics at the electrified interface [26 ,31 -35 ].Several recent studies have highlighted the important role of the H-bond network on reaction kinetics for various electrochemical processes.For instance, Li et al. revealed the evolution of interfacial H-bonds with changing bias potential, and found that an ordered interfacial water achieved high-efficiency electron transfer for HER [34 ].Chen and co-workers demonstrated that the discontinuity of the H-bond network in the alkaline electric double layers (EDLs) resulted in a slow kinetics of hydrogen electrocatalytic reaction [32 ].Jia et al. reported that the introduction of N-methylimidazoles into electrolytes can restore the interfacial H-bond network to improve hydrogen reaction rates [33 ].Nevertheless, the complicated interfacial environment of the co-electrolysis of CO 2 and H 2 O, at the molecular level, makes understanding the role of the interfacial H-bond network and developing corresponding modulation strategies a great challenge.
In this work, ethylenediaminetetraacetic acid (EDTA) and its analogs were introduced into an aqueous KHCO 3 electrolyte to reshape the electrified interface, enabling a prominent enhancement of CO 2 electrolysis to carbon-based products.Nuclear magnetic resonance (NMR) spectroscopy analysis and molecular dynamic (MD) simulations demonstrated the interactions among electrolyte compositions.In situ surface-enhanced infrared and ab initio molecular dynamics (AIMD) simulations were performed to elucidate the nature of interfacial environment.It was revealed that, when the EDTA molecules were assembled toward the EDL region in response to bias potential, it was able to reconstruct the original interfacial H-bond network via the Lewis acid-base interaction of EDTA 4--K + .A H-bond gap layer in the EDL region was established, significantly affecting proton transfer kinetics.AIMD simulation provided the information regarding the energy barrier of hydrogen-related reaction kinetics.The EDTA molecules acted on the cationic hydration shell to promote the protonation of *CO 2 to *COOH and inhibit the rate-limiting Volmer step of HER.We founded that the EDTA-containing electrolyte afforded a kinetically favourable local en-vironment for CO 2 conversion.Consequently, CO Faradaic efficiency of 90% at 500 mA cm −2 was delivered on a commercial Ag catalyst in a flow-type cell.This strategy was universal for improving the carbon-based product selectivity of other catalysts, such as palladium (Pd), zinc (Zn), bismuth (Bi), tin (Sn) and copper (Cu).

EDTA-reshaped electrolyte structure
EDTA, a typical Lewis-base ligand, is preferentially selected as an electrolyte additive together with Lewis acid cations to build a composite solvation structure [36 ].First, 13 C NMR spectroscopy was used to probe the structural form of EDTA in the electrolyte.No carboxylate ( −COOH) signal is observed in the 13 C NMR spectrum of the KHCO 3 -EDTA electrolyte ( Fig. S1), suggesting that EDTA molecules are completely deprotonated and exist as EDTA 4-ions [37 ].The deprotonated carboxyl group of EDTA 4-is expected to form a strong Lewis acid-base interaction with K + , and H-bond interaction with H 2 O. Deuterium ( 2 H, D) NMR spectroscopy was performed to investigate EDTA-water interaction in the electrolyte.The 2 H peak presents a high-field shift and broadening phenomena after the introduction of EDTA to KHCO 3 electrolyte (Fig. 1 a and Fig. S2a), suggesting that the electron density of water is increased by EDTA molecules [38 ].An upfield chemical shift of water indicates disruption of the original H-bond network among water molecules by EDTA-water H-bond interaction [39 ,40 ].To further confirm the H-bond interaction of EDTA-water, solutions without KHCO 3 were also prepared and tested ( Fig. S2b), in which the same trend was also observed.The Raman shift signal variations of the O-H stretching vibration ( Fig. S3) also evidence this inference.
To validate the Lewis acid-base interaction of EDTA 4--K + in the electrolyte, the electrolyte was studied using 39 K NMR spectroscopy.After the introduction of EDTA, a narrow down-field shift of 0.21 ppm is observed for the electrolyte with EDTA (Fig. 1 b), indicating a strong de-shielding effect on the K nucleus [38 ].This decreases the content of cation-coordinated water.In other words, the K + solvation structure is changed due to the addition of EDTA.UV-vis spectrum of the electrolyte with EDTA displays an enhanced absorption peak starting from ∼265 nm (Fig. 1 c), implying the presence of EDTA 4--K + complexes [41 ,42 ].Furthermore, MD simulation and quantum chemical calculation were applied to understand the coordination environment of K + ions in the electrolytes.We performed the MD simulation using the Large-scale

Effect of EDTA additive on electrocatalytic performance
In order to investigate the effect of Lewis-base EDTA molecules on electrocatalytic reactions, we conducted cyclic voltammograms (CVs) on rotating disk electrodes (RDEs) for Ag electrodes in N 2and CO 2 -saturated KHCO 3 (0.1 M) electrolytes w ith and w ithout EDTA.A s shown in Fig. 2 a, in N 2 -saturated electrolytes, the EDTA-containing system shows decreased current density compared with pure KHCO 3 electrolytes, demonstrating the decreased activity of H 2 O dissociation and the suppression effect of HER in the presence of EDTA.Combined with the above NMR spectroscopy, the drop in HER performance may be attributed to the H-bond formation between EDTA and H 2 O, which leads to a limited proton source for HER [26 ].In CO 2 -saturated electrolytes, the current density is higher than that in N 2 -saturated electrolytes with EDTA (Fig. 2 a and Fig. S7), implying that the interfacial CO 2 conversion is indeed promoted after the introduction of EDTA [45 ,46 ].Furthermore, we used electrochemical impedance spectroscopy (EIS) as an interface analysis technique that is capable of probing potential-dependent reaction kinetics, from which charge transfer resistance (R ct ) and constant-phase elements (CPE) can be extracted and linked to charge transfer kinetics and interfacial ion organization [47 ,48 ].We performed EIS on the RDE electrode near the onset potential (from −0.6 to −0.9 V RHE , reversible hydrogen electrode, RHE) to ensure that tests were kinetically controlled, and to avoid as far as possible the effects of bubble disturbance ( Fig. S8).A modified Randle's circuit model that includes two possible electrochemical interfaces with different timescales was used (Fig. 2 b) [49 ].High-frequency intervals (1st semicircle) reflect the kinetic characteristics of electrochemical reactions [50 ].Compared with two systems in the high-frequency region, we observe that the EDTAcontaining s ystem exhib its a lower R ct (that is R 1 , Fig. 2 c), indicating that the electron transfer process is more facile during CO 2 electroreduction.
The capacitive terms (CPE values) relate to the accumulated ions at or near the electrode-electrolyte interface [48 ,51 ].From -0.6 to -0.75 V RHE , the EDTA system exhibits higher CPE 1 values (Fig. 2 c and Tables S1, S2), which is attributed to the assembly of EDTA 4-ions bonded with K + ions toward the interface region in response to bias potential, evidencing the existence of Lewis acid-base interaction of EDTA 4--K + .Particularly, a decrease in CPE 1 values is observed after the bias potential of -0.75 V RHE , which is attributed to the loose alignment that occurs in the outer Helmholtz plane (OHP, a closely packed layer of cations) [52 ].The loose alignment of K + ions is caused by repulsive interaction of the electric field toward EDTA 4-ions under a higher reduction potential, further suggesting the strong interaction of EDTA 4--K + .
Electrocatalytic performance was measured at −0.8 to −1.2 V RHE in CO 2 -saturated 0.5 M KHCO 3 electrolytes with and without EDTA in H-type cells (Fig. 2 d and Figs S9, S10).In the electrolyte with 1 mM EDTA, a commercial Ag foil electrode delivered a much higher CO Faradaic efficiency (FE) and partial current density (92% and 11.5 mA cm −2 ) than those (34% and 3 mA cm −2 ) in the EDTA-free electrolyte at −1.0 V RHE , highlighting the key role of EDTA additives on boosting the selectivity of CO 2 reduction to CO.The Staircase voltammetry (SCV) tests of both electrolyte systems also clearly i l lustrate the inhibition of HER and acceleration of CO 2 RR by EDTA ( Fig. S11).The commercial Ag nanoparticle (NP)-supported carbon paper electrode shows similar performance trends ( Fig. S12).The Ag foil electrode in the EDTA-containing electrolyte also exhibits decent stability for CO 2 electroreduction to CO ( Fig. S13).In addition, we fixed the same Ag electrode and tested it successively in different electrolytes.After replacing the new KHCO 3 electrolyte, the Ag electrode recovers its original performance (Fig. 2 e).This suggests that the elevated catalytic activity is not due to irreversible adsorption [19 -23 ], but may originate from the change of interfacial microenvironment through a non-adsorbed form of EDTA.
To verify the universality of EDTA additives on CO 2 electrolysis performance, other metal catalysts are also studied.Pd and Zn have been demonstrated to possess decent activities for CO 2 conversion to CO.Both the metal catalysts display obvious FE CO enhancement (Fig. 2 f and Figs S14, S15), wherein a Pd/C catalyst achieves 88% FE CO at −1 V RHE in the EDTA-containing electrolyte.We also explored whether EDTA has the same effect on catalysts toward formate productuction.Both Bi and Sn, common catalysts for the preparation of formate, show significant FE formate enhancement in the EDTA-containing electrolyte (Fig. 2 f and Figs S16, S17), whereby the Sn foil catalyst achieves 80% FE formate at −1.0 V RHE .Furthermore, we followed on to explore its potential for multi-carbon products (C 2 + ) formation on Cu catalysts ( Figs S18 and S19).The commercial Cu foil shows a decreased FE of H 2 , and the FE of carbon-based products increased from 33% to 57% at −1.0 V RHE (Fig. 2 f and Fig. S18).
To exclude the change of catalyst structure that may impact performance [53 ,54 ], a series of characterizations were carried out, showing that the EDTA in the electrolyte does not induce the change of morphology, crystallinity, and oxidation states of the Ag electrode ( Figs S20-S22).We also evaluated the stability of EDTA under operational conditions by NMR analysis.The chemical structure of EDTA does not change after electrolysis ( Fig. S23).

In-situ probing of the electrified interface
To gain insights into the electrified interfacial Hbond network, in situ attenuated total reflection surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) was applied ( Fig. S24).Figure 3 a and Fig. S24 show in situ ATR-SEIRAS spectra on Ag at various potentials in CO 2 -saturated 0.5 M KHCO 3 with and without 1 mM EDTA, respectively.For two electrolyte systems, several vibrational bands were found to appear at 3650-30 0 0, 2950-2800, 2000-1850 and 1650-1610 cm −1 , which can be assigned to O-H stretching mode ( ν-OH) of H 2 O [55 ,56 ], C-H stretching mode ( ν-CH) of EDTA, C-O stretching mode ( ν-CO) of CO adsorbed on the top or bridge sites of Ag surface [57 ,58 ], and H-O-H bending mode ( δ-HOH) of H 2 O, respectively.First, we analyzed the H-bond environment of the interfacial water.After the introduction of EDTA, the vibration of δ-HOH shifts from 1647 to 1618 cm −1 (Fig. 3 a), indicating the formation of a weak H-bond network among the interfacial water molecules [56 ].The red shift of ν-OH can be attributed to the presence of a stronger H-bond interaction among water molecules or from other electrolyte components [59 ].Considering the weak Hbonds among interfacial water, we attribute the red shift of ν-OH in the EDTA-containing system to the strong H-bonds between EDTA and H 2 O.In addition, the ν-OH peak in two electrolyte systems can be deconvoluted into three distinct components through Gaussian fitting [54 ,60 ] (Fig. 3 b, c and  Figs S26, S27), corresponding to a strong H-bond at ∼3250 cm −1 , medium H-bond at ∼3450 cm −1 , and weak H-bond at ∼3600 cm −1 , respectively.Specifically, the ν-OH is dominated by a strong H-bond in the KHCO 3 system (Fig. 3 b).But the medium Hbond is dominant after introducing EDTA (Fig. 3 c).This is attributed to the spatial effect of EDTA molecules and the H-bond effect on water disrupting the initial interfacial H-bond network among water.Therefore, the original H-bond network among interfacial H 2 O is reconstructed by EDTA, forming a weakened H-bond network at the electrified interface.The weakened H-bond network can result in an increased proton transfer barrier in the interfacial region [32 ].
To further explore the interfacial interaction among electrolyte components, we used the absorbance intensities of ν-OH and ν-CH to identify the interfacial content of H 2 O and EDTA in response to bias potential, respectively (Fig. 3 d and e).For the pure KHCO 3 system, the intensity of ν-OH enhanced with increasing bias potential, suggesting that water molecules are gradually adsorbed on the electrode surface from open circuit potential (OCP) to −1.2 V RHE .In contrast, for the EDTA system, the intensity of ν-OH appears to first decrease and then increase with increasing bias potential, exhibiting a minimum value at −0.3 V RHE .Notably, from OCP to −0.3 V RHE , a gradual increase of EDTA at the interface is accompanied by a decrease in the adsorbed water content.Along with the electrostatic effects of K + cations with the electrode, EDTA molecules are adsorbed together by strong Lewis acid-base interactions.At the same time, water molecules are difficult to be quickly replenished to the catalyst surface due to the spatial effect of EDTA.From −0.3 to −1.2 V RHE , the amount of EDTA adsorbed at the interface gradually decreases, which is attributed to increased electrostatic repulsion between the electrode surface and the EDTA 4-ions.However, the amount of water increases with increasing bias potential, which is at-tributed to enhanced interfacial electrostatic adsorption between water and electrode [58 ].The Raman shift signal variations of the C-H stretching vibration ( Fig. S28) also confirm this feature of EDTA distribution under external electric field.From these in situ spectra results in conjunction with above EIS analysis, it is revealed that EDTA molecules are assembled toward the interface region in response to bias potential, leading to weakening of the interfacial H-bond network.

Theoretical investigation into EDTA-reshaped interface
To further provide in-depth insights into the EDTAreconstructed interface, we conducted AIMD simulation to study the EDL structures of K-H 2 O and K-H 2 O-EDTA interfaces on an Ag(111) supercell ( Fig. S29).Figure 4 a and b shows the representative AIMD snapshots for the EDL structures of K-H 2 O and K-H 2 O-EDTA configurations, respectively.The planes consisting of the cations closest to the electrode surface are defined as the closest ion planes (CIPs), which are at distances of ∼3.27 Å and ∼3.72 Å away from the electrode surface for K-H 2 O and K-H 2 O-EDTA systems, respectively.The cations at the K-H 2 O-EDTA system lose some solvated molecules (Fig. 4 b), which is due to the participation of EDTA molecules in K + solvation structure.The CIP shift away from surface normal direction is attributed to the Lewis acid-base interaction between EDTA 4-and K + .The cation and water distributions in the two systems suggest that the CIPs correspond to the outer Helmholtz plane at the interface, which agrees w ith prev ious work about the alkaline interface [32 ].The interfacial oxygen concentration profiles along the surface normal direction were studied for the distribution of water molecules and interaction of EDTA-H 2 O (Fig. 4 c).Since oxygen atoms are mainly contributed by water molecules, the significantly lower oxygen concentration in the EDTA system indicates that the introduction of EDTA brings about a low content of interfacial water, which matches the results of the above in situ experiments.The decrease in the concentration of water molecules would reduce connectivity of the H-bond network in the EDL region, which is verified by the statistical distribution of H-bonds along the surface normal direction (Fig. 4 d).The statistics for H-bond number in the K-H 2 O system are similar to previous reports [32 ].In the K-H 2 O-EDTA system, the H-bond number is the sum of the H 2 O-H 2 O H-bonds and the EDTA-H 2 O H-bonds.Over a large spatial range from 3.0 to 7.8 Å, the H 2 O-H 2 O H-bond number is significantly lower than that of the K-H 2 O system, forming a H-bond gap layer in the EDL region (Fig. 4 e and f).The H-bond gap layer decreases the connectivity of the H-bond network and can affect hydrogen-related reaction kinetics [32 ].
The EDTA has a direct impact on the firstsolvated shell of cations at CIPs, and may alter the reaction kinetics of hydrogen-related electrocatalytic processes.After introducing a CO 2 molecule into the K-H 2 O and K-H 2 O-EDTA systems ( Figs S30 and S31), we further performed constrained AIMD (cAIMD) simulations to investigate the effect of EDTA on the free energy of the proton transfer processes.The cAIMD method was performed to evaluate the cation-coordinated CO 2 electroreduction on Ag-water interfaces [61 -63 ].Based on the obtained interface pictures (Fig. 4 a and b), the CO 2 and H 2 O molecules in the cationic coordination environment were selected for cAIMD calculations in the two systems with and without EDTA.We first simulated the *CO 2 to *COOH conversion process in both EDTA and EDTA-free systems, where the water was the proton donor (Fig. 5 a and Fig. S32a).Afterward, for the rate-determining Volmer step of HER (*H 2 O to *H), both EDTA and EDTA-free systems were also considered (Fig. 5 b and Fig. S32b).According to these snapshots of species evolution, it is found that the EDTA undergoes a cation-mediated pathway to affect the proton transfer steps, when the interplay between the EDTA and the cationic co-ordination environment occurs in the H-bond network (Fig. 5 c ) .For the K-H 2 O-EDTA-CO 2 system, the free energy of *CO 2 to *COOH step (0.94 eV) is lower than the EDTA-free system (1.01 eV), indicating that the EDTA promotes this cation-mediated proton transfer process (Fig. 5 d and Fig. S33a, b).With K-H 2 O-EDTA system, the kinetic barrier of *H 2 O to *H is 1.78 eV, higher than the 1.67 eV of the EDTA-free system (Fig. 5 e and Fig. S33c, d), demonstrating that EDTA-containing cation shell leads to poor HER activ ity v ia suppressing water dissociation.Eventually, by regulating the proton transfer kinetics of cation-coordinated water, in EDTAcontaining s ystems, CO 2 reduction is facile with the lower energy barrier for the protonation of *CO 2 to *COOH.Meanwhile, HER is suppressed, showing a higher energy barrier in the rate-limiting Volmer step.

Lewis-base ligands enabling the boosting CO 2 electrolysis in flow-type cells
To evaluate the benefits of EDTA additives for high-rate CO 2 electrolysis at commercially relevant performance metrics, we introduced the EDTA to the electrolytes in flow-type cells equipped with gas diffusion electrodes (GDE) ( Figs S34, S35).Over full current density range (10 0 −50 0 mA cm −2 ), the EDTA-containing system greatly improves FE CO (Fig. 6 ) and also delivers good stability ( Fig. S36).These results suggest that a high FE CO is sti l l achieved in flow cells with the EDTA additive.The commercial Cu NPs were also tested in flow-type cel ls, and simi lar results of the boosted FE of carbon-based products were shown ( Fig. S37).
In response to bias potential, EDTA molecules are assembled toward the EDL region via the Lewis acid-base interaction between deprotonated carboxyl groups and K + ions [36 ].This leads to the formation of a favourable interface environment for CO 2 conversion.To further expand this concept, we also employed EDTA analogs as electrolyte additives, including EDDA, NTA and DTPA (where EDDA is ethylenediamine-N , N -diacetic acid, NTA is nitrilotriacetic acid, DTPA is diethylenetriaminepentaacetic acid).The number of carboxyl groups for EDDA, NTA, EDTA and DTPA is 2, 3, 4 and 5, respectively (Fig. 6 a).All these ligand systems compared to the pure KHCO 3 system show an increased FE CO in flow-type cells with the cathode area of 1 and 10 cm 2 (Fig. 6 d and Figs S38-S41).The FE CO increases with increasing number of carboxyl groups.This is because more carboxyl groups result in a more significant disruption of the H-bond network among H 2 O molecules, as evidenced by 1 H-NMR spectra of the different additive molecules ( Fig. S42).Notably, the shift degree of the water peak ( Fig. S41) toward high field is consistent with the trend of FE CO (Fig. 6 d), highlighting that modulation of the Hbond network is closely associated with the delivered FE CO .The DTPA system compared to the EDTA exhibits slightly higher FE CO (Fig. 6 d).It should be pointed out that the DTPA compared to the EDTA exhibits lower water solubility and higher cost, implying that the EDTA may be optimal among these Lewis-base ligands.

CONCLUSION
In summary, by virtue of comprehensive in situ vibrational spectroscopic characterizations and AIMD simulations, we have provided molecule-level insights into the Lewis-base ligand-reshaped interfacial organization for promoting CO 2 electrolysis.EDTA, a typical Lewis-base ligand, has been introduced to the electrolyte to construct a H-bond gap layer in the EDL region, resulting in a lower proton source and a decreased activity of undesired HER.The EDTA enters the K + solvation structure via the strong Lewis acid-base interaction of EDTA

Figure 1 .
Figure 1.Investigation into electrolyte structure.(a) 2 H-NMR spectra of the pure D 2 O, KHCO 3 electrolytes without and with 5 mM EDTA.The electrolyte is prepared by using D 2 O as solvent.(b) 39 K-NMR spectra of KHCO 3 electrolytes without and with 5 mM EDTA.(c) UV-vis spectrum of KHCO 3 electrolytes without and with 5 mM EDTA.(d) Optimized solvation structure of the electrolyte before and after the introduction of EDTA by MD simulation.The O, H, C, N and K are colored in red, white, cyan, blue and purple, respectively.(e) The LUMO and HOMO energies of K + -7H 2 O and 2(K + -5H 2 O)-EDTA 4 − .

Figure 2 .
Figure 2. CO 2 electrolysis performance in KHCO 3 and KHCO 3 -EDTA electrolytes.(a) CV tests in KHCO 3 electrolytes with and without 5 mM EDTA at a scan rate of 10 mV s −1 under N 2 -and CO 2 -saturated conditions.(b) Nyquist plots of two systems measured at −0.65 V RHE .The inset equivalent circuit was used for EIS spectra fitting.(c) Numerical data fitting results of EIS spectra measured at different potentials: R 1 and CPE 1 .The concentration of KHCO 3 in the EIS test was 0.5 M, and the added EDTA concentration was 5 mM.(d) FE of CO tested with Ag foil as the electrode in 0.5 M KHCO 3 electrolytes with and without 5 mM EDTA.(e) FE of CO tested with Ag foil as the electrode in 0.5 M KHCO 3 electrolytes with and without 5 mM EDTA, and new KHCO 3 electrolyte at −1.0 V RHE .(f) FE of carbon-based products with Zn foil, Pd/C, Bi NPs, Sn foil and Cu foil as the electrode in 0.5 M KHCO 3 electrolytes with and without 1 mM EDTA at −1.0 V RHE .The error bars represent three independent tests.

Figure 3 .
Figure 3. In-situ probing electrified interface.(a) In situ ATR-SEIRAS spectra under various potentials for the KHCO 3 electrolyte with and without 1 mM EDTA.(b) Deconvolution of the ν-OH peak in 0.5 M KHCO 3 electrolyte at −0.7 V RHE .(c) Deconvolution of the ν-OH peak in 0.5 M KHCO 3 electrolyte with 1 mM EDTA at −0.7 V RHE .(d) Intensity of ν-OH peaks of KHCO 3 and EDTAcontaining electrolytes under various potentials.(e) Intensity of ν-CH peaks of EDTA-containing electrolytes under various potentials.

Figure 4 .
Figure 4. EDL structures of K-H 2 O and K-H 2 O-EDTA interfaces.Representative snapshots of EDL structures on the Ag(111) electrode surface for K-H 2 O system (a) and K-H 2 O-EDTA system (b).The Ag, O, H, C, N and K are colored in grey, red, white, blue, green and purple, respectively.The blue dashed lines represent the H-bonds.(c) Concentration distribution profiles of O atoms in water and EDTA along the surface normal direction.The vertical dashed lines represent the CIPs.The area marked in green is the location of the EDTA molecule.The horizontal black dashed lines represent the bulk water concentration (0.056 mol cm -3 ).(d) Statistical distribution of H-bond number along the surface normal direction for K-H 2 O and K-H 2 O-EDTA systems.(e) Schematic of interfacial H-bond network for EDTA-free electrolyte.(f) Schematic of interfacial H-bond network and the proposed regulation mechanism for EDTA-containing electrolyte.

Figure 5 .
Figure 5. Theoretical insights into hydrogen-related electrocatalytic processes at K-H 2 O and K-H 2 O-EDTA interfaces.(a) Representative snapshots of the *CO 2 to *COOH step at K-H 2 O-EDTA interface.(b) Representative snapshots of the Volmer step at K-H 2 O-EDTA interface.The Ag, O, H, C, N and K are colored in grey, red, white, cyan, blue and purple, respectively.The green hydrogen atoms represent those involved in hydrogen-related electrocatalytic processes.(c) Schematic illustration of the EDTA effect on the cationic coordination environment, where in the H-bond network the processes of CO 2 protonation and H 2 O dissociation occurs.The Ag, O, H, C, N and K are colored in grey, red, white, sky blue, green and purple, respectively.(d) Energy barrier of the *CO 2 to *COOH step at K-H 2 O and K-H 2 O-EDTA interfaces.(e) Energy barrier of the Volmer step at K-H 2 O and K-H 2 O-EDTA interfaces.

Figure 6 .
Figure 6.CO 2 electrolysis performance in a flow-type cell.(a) Structural formulae of EDDA, NTA, EDTA and DTPA.(b) Faradaic efficiency of H 2 and CO at various current densities in 1 M KHCO 3 electrolytes.(c) Faradaic efficiency of H 2 and CO at various current densities in 1 M KHCO 3 electrolytes with 5 mM EDTA.(d) Faradaic efficiency of H 2 and CO at 500 mA cm -2 in 1 M KHCO 3 electrolytes without and with different Lewis base molecules.The error bars represent three independent tests.
4--K + .This facilitates the protonation step of *CO 2 to *COOH while suppressing *H 2 O dissociation to *H.The reconstructed interface in flow cells has enabled commercial Ag nanoparticle catalyst to deliver over 99% and 90% FE CO for CO 2 electrolysis at −200 and −500 mA cm −2 , respectively.This electrolyte regulation strategy is universally applicable to different metal catalysts (such as Ag, Zn, Pd, Bi, Sn and Cu) and other similar Lewis-base ligands (such as EDDA, NTA and DTPA).The findings established here propose future directions that electrocatalytic performance of various electrochemical reactions can effectively be steered by tailoring the interactions among ions and/or molecules in electrolytes.