This is the submitted version of the journal article: Wang, Xiang; Han, Xu; Du, Ruifeng; [et al.]. «Unveiling the role of counter- anions in amorphous transition metal-based oxygen evolution electrocatalysts». Applied catalysis. B, Environmental, Vol. 320 (Jan. 2023), art. 121988. DOI 10.1016/j.apcatb.2022.121988 This version is available at https://ddd.uab.cat/record/271372 under the terms of the license 0DQXVFULSW &OLFNKHUHWRYLHZOLQNHG5HIHUHQFHV 1 Unveiling the Role of Counter-Anions in Transition Metal-Based Oxygen 2 3 4 Evolution Electrocatalysts 5 6 Xiang Wang,[a,b] Xu Han,[c] Ruifeng Du,[a,b] Zhifu Liang,[a,c] Yong Zuo,[a,d] Pablo Guardia,[a] Junshan 7 8 Li,[e] Jordi Llorca,[f] Jordi Arbiol,[c,g] Renji Zheng,*[h,i] Andreu Cabot*[a,g] 9 10 11 a 12 Catalonia Institute for Energy Research (IREC), Sant Adrià de Besòs, 08930 Barcelona, Spain 13 b 14 Departament d’Enginyeria Electrònica i Biomèdica, Universitat de Barcelona, 08028, 15 16 Barcelona, Catalonia, Spain 17 c Catalan Institute of Nanoscience and Nanotechnology (ICN2), Campus UAB, Bellaterra, 08193 18 19 20 Barcelona, Catalonia, Spain 21 d Istituto Italiano di Tecnologia, Via Morego 30, 16163, Genova, Italy 22 23 e Institute of Advanced Study, Chengdu University, Chengdu 610106, China 24 25 26 f Institute of Energy Technologies, Department of Chemical Engineering and Barcelona 27 28 Research Center in Multiscale Science and Engineering, Universitat Politècnica de Catalunya, 29 30 EEBE, 08019, Barcelona, Catalonia, Spain 31 32 g ICREA, Pg. Lluis Companys, 08010 Barcelona, Catalonia, Spain 33 34 h School of Minerals Processing and Bioengineering, Central South University, Changsha 35 36 410083, China 37 38 i Key Laboratory of Hunan Province for Clean and Efficient Utilization of Strategic Calcium- 39 40 containing Mineral Resources, Central South University, Changsha 410083, China 41 42 43 * E-mails: zhengrj@csu.edu.cn; A. Cabot: acabot@irec.cat; 44 45 46 Abstract 47 48 At the initial stage of the oxygen evolution reaction (OER) most electrocatalysts undergo a 49 50 51 structural and chemical surface reconstruction. While this reconstruction strongly influences 52 53 their performance, it is frequently overlooked. Herein, we analyze the role of the oxidized 54 55 anions, which is particularly neglected in most previous works. We introduce a range of 56 57 different anionic groups (Cl-, CH3COO-, NO3-, SO42-) on the surface of an amorphous ZnCoxNiyOz 58 59 catalyst by a facile proton etching and ion exchange method from a ZIF-8 self-sacrificial 60 61 62  63 64 65 template. The structural and chemical properties of the obtained set of materials are 1 2 thoroughly analysed and correlated with their electrocatalytic performance to study the effect 3 4 of surface anionic groups, phase transition, metal leaching and defect generation on OER 5 6 7 activity. Exploiting the control possibilities provided by the synthesis method here described 8 9 and taking into account the uncovered property-performance correlations, the electrocatalyst 10 11 is optimized. As a result, we produce ZnCo1.25Ni0.73Ox-SO4 catalysts with outstanding OER 12 13 performances, including a low overpotential of 252 mV at 10 mA cm−2 with a reduced Tafel 14 15 slope of 41.6 mV dec−1. Furthermore, this catalyst exhibits remarkable stability with negligible 16 17 overpotential variation for 100 h. The excellent catalytic properties are rationalized using 18 19 density functional theory calculations, showing that the surface-adsorbed anions, particularly 20 21 SO42−, can stabilize the OOH* intermediate, thus enhancing the OER activity. 22 23 24 Keywords: amorphous oxide, anion adsorption, structural reconstruction, leaching, oxygen 25 26 evolution reaction. 27 28 29 1. Introduction 30 31 32 The electrochemical oxygen evolution reaction (OER) is a vital half-reaction in several 33 34 renewable energy conversion and storage technologies. OER is coupled to hydrogen evolution 35 36 in electrolyzers, a metal reduction in rechargeable metal-air batteries, nitrogen reduction to 37 38 ammonia, and CO2 reduction to methane or methanol for instance.[1-3] However, the slow 39 40 kinetics of the involved multi-electron transfer processes severely reduce the OER activity.[4, 41 42 43 5] To address this crucial concern, a wide range of potential OER electrocatalysts have been 44 45 developed. Among them, noble metal-free transition metal oxides,[6, 7] nitrides,[8, 9] 46 47 phosphides,[10-12] and dichalcogenides[13, 14] are particularly interesting from a cost- 48 49 effectiveness point of view. However, in-situ electrochemical analyses have frequently 50 51 revealed that these compounds are not the real active material in the OER process, but they 52 53 are inevitably oxidized and reconstructed into transition metal hydroxides (M(OH)x) and 54 55 oxyhydroxides (MOOH) during the first steps of the electrochemical reaction.[15-17] While 56 57 several pieces of evidence of this transformation have been published, the fate and role within 58 59 the electrochemical cell of the anions after the pre-catalyst is oxidized is rarely reported. 60 61 62  63 64 65 Besides, the extent of the transition metal transformation and the final metal reorganization 1 2 and composition is hardly investigated. In particular, during the compound oxidation, it is very 3 4 likely that some metal leaching occurs, which can also strongly affect the OER process. 5 6 7 During reconstruction, the surface anions are probably oxidized to nitrates (NO3-), phosphates 8 9 (PO43−), sulfates (SO42-), or selenates (SeO42−). In a few recent works, we and others have 10 11 demonstrated the presence of such nonmetallic anionic groups to play a key role in the OER 12 13 activity.[18-23]Accordingly, the rational design and engineering of catalysts containing 14 15 controlled amounts of such anionic groups could be an effective strategy to improve their 16 17 performance. In this direction, the introduction of anionic groups a posteriori, through surface 18 19 adsorption, generally results in a weak chemical interaction that limits the effectiveness and 20 21 stability of the resultant catalyst.[21] Thus, approaches for the direct incorporation of 22 23 controlled amounts of anionic groups that enable a strong bonding between anions and metal 24 25 atoms need to be developed. 26 27 On the other hand, amorphous catalysts generally display improved OER performance over 28 29 their crystalline counterpart.[24, 25] Such catalysts present highly tunable atomic 30 31 arrangement, compositions, and easier doping with a high concentration of extrinsic 32 33 dopants.[26] Besides, amorphous oxides, also enable a larger range of metal oxidation states 34 35 and result in higher concentrations of unsaturated coordination spheres.[27] Additionally, 36 37 38 compared with crystalline materials, amorphous catalysts usually form less strong bonds with 39 40 oxygen, which can facilitate an efficient lattice oxygen-mediated OER mechanism.[28, 29] 41 42 Beyond the abundant structural defects found in amorphous materials, the leaching of 43 44 controlled amounts of the composing transition metal ions can in situ generate additional 45 46 defects that may translate into active OER sites.[30-34] 47 48 Herein, we introduce different anionic groups (Cl-, CH3COO-, NO3-, SO42-) in ZnCoxNiyOz 49 50 nanosheets by a facile proton etching and ion exchange method. We test the resulting 51 52 materials for OER and study the material transformation during the reaction, including metal 53 54 ion leaching. We thoroughly analyze the structural and chemical properties of the initial and 55 56 evolved materials and correlate them with their electrocatalytic performance. Exploiting the 57 58 control possibilities provided by the synthesis method and taking into account the uncovered 59 60 property-performance correlations, the electrocatalyst is optimized to reach outstanding OER 61 62  63 64 65 performances. Additionally, the optimized catalyst is further studied using density functional 1 2 theory (DFT) calculations, which allows for determining the influence of particular surface- 3 4 adsorbed groups on the OER activity. 5 6 7 2. Material and methods 8 9 2.1. Chemicals 10 11 12 Zinc nitrate hexahydrate (Zn(NO3)2·6H2O), nickel nitrate hexahydrate (Ni(NO3)2·6H2O), nickel 13 14 chloride hexahydrate (NiCl2·6H2O, 98%), cobalt nitrate hexahydrate (Co(NO3)2·6H2O, 99.9%), 15 16 cobalt chloride hexahydrate (CoCl2·6H2O, 98%), cobalt acetate tetrahydrate (Co(CH3COO)2, 17 18 98%), nickel(II) acetate tetrahydrate (Ni(CH3COO)2, 98%), cobalt sulfate hexahydrate 19 20 (CoSO4·6H2O,99%), nickel sulfate hexahydrate (NiSO4·6H2O, 99%), potassium hydroxide (KOH, 21 22 85%), and 2-methylimidazole (C4H6N2, 99%) were purchased from Acros Organics. Iridium(IV) 23 24 oxide (IrO2, 99.9% metal basis) and Nafion (5 wt% in a mixture of low aliphatic alcohols and 25 26 water) were obtained from Sigma-Aldrich. Methanol, ethanol and isopropanol were of 27 28 analytical grade and obtained from various sources. Milli-Q water was obtained from a Purelab 29 30 flex from Elga. All chemicals were used as received, without further purification. 31 32 33 2.2. Zeolitic imidazolate framework (ZIF-8) MOF 34 35 36 Among the different MOFs, we selected the ZIF-8 due to its high versatility and ease of 37 38 synthesis. The synthesis of ZIF-8 was based on a previous procedure with modifications.[37] 39 40 Typically, 0.87 g Zn(NO3)2·6H2O was dissolved in 30 mL methanol. This solution was 41 42 incorporated into a second methanol solution having 1.97 g 2-methylimidazole. The final 43 44 mixture was kept at room temperature for 24 h. White precipitates were collected using a 45 46 centrifuge, washed with methanol, and dried overnight at 60 . 47 48 49 2.3. Synthesis of ZnCoxNiyOz. 50 51 100 mg of ZIF-8 was sonicated in 20 mL of ethanol for redispersion. The obtained dispersion 52 53 was added into a 100 mL water solution having 300 mg CoSO4·6H2O and 300 mg NiSO4·6H2O. 54 55 The final mixture was vigorously stirred at room temperature for 12 h. The product was 56 57 centrifuged, washed with water, and freeze-dried. The obtained products were calcined in the 58 59 muffle furnace at 350 for 2 hours under an air atmosphere. The final products were obtained 60 61 62  63 64 65 after natural cooling to room temperature and denoted as ZnCoxNiyOz-SO4. Besides, 1 2 ZnCoxNiyOz-NO3, ZnCoxNiyOz-OAc, and ZnCoxNiyOz-Cl were obtained using the same method, 3 4 but substituting sulfates with equal amounts of chlorides, acetates, and nitrates, respectively. 5 6 7 As a reference material, ZnO nanoparticles were obtained by the direct calcination of ZIF-8. 8 9 Besides, ZnCoxNiyOz-SO4 samples with different metal contents were obtained by adjusting the 10 11 amount of cobalt and nickel salts added. 12 13 3. Results and discussion 14 15 3.1. Synthesis and characterizations of catalysts 16 17 18 Figure 1 illustrates the two-step approach used to produce ZnCoxNiyOz from a self-sacrificial 19 20 ZIF-8 template (see details in the experimental section). In the first step, the Zn-based ZIF-8 21 22 MOF is reacted with Ni and Co salts (sulfates, nitrates, chlorides or acetates). In this process, 23 24 the protons produced by the metal cation (Co2+, Ni2+) hydrolysis (M2+ + 2H2O = M(OH)2 + 2H+) 25 26 etch the ZIF-8, dissociating its polyhedron structure.[38, 39] At the same time, the cobalt and 27 28 nickel ions partially replace zinc within the structure. Meanwhile, the salt anions (SO42-, NO3-, 29 30 Cl-, OAC-) are anchored to the material surface by exchange with the MOF organic ligands. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Figure 1. schematic illustration for the two-step preparation of ZnCoxNiyOz nanosheets from ZIF-8 MOF. 45 46 47 Figure 2a displays a representative SEM micrograph of the ZIF-8 particles, which were 48 49 characterized by a dodecahedron shape and an average size of ca. 100 nm. Upon reaction with 50 51 the transition metal sulfate, a nanosheet-like structure progressively expands around the 52 53 gradually disappearing ZIF-8 template (Figure 2b). Eventually, the dodecahedron template has 54 55 vanished and just the porous nanosheet structure is observed. 56 57 In the second step, the material is annealed under air to form an amorphous oxide. Figures 2c- 58 59 e and S1 show SEM, TEM and STEM images of the ultrathin nanosheets with wrinkled 60 61 62  63 64 65 structures obtained after an annealing step at 350 °C in an air atmosphere. HRTEM analyses 1 2 confirmed that the final ZnCoxNiyOz-SO4 presented an amorphous structure with no 3 4 crystallographic ordering (Figure 2f,g). Meanwhile, HAADF-STEM and EELS compositional maps 5 6 7 showed C, N, O, S, Zn, Co and Ni homogeneously dispersed within each nanosheet (Figures 2h 8 9 and S2). 10 11 Figure 3a shows an AFM image of the ZnCo1.26Ni0.73Ox-SO4 nanosheets. From the height line 12 13 profiles of some of the nanosheets shown in Figure 3a, their average thickness was estimated 14 15 at 1.7±0.2 nm (Figure 3b). From the N2 adsorption-desorption isotherms displayed in Figure 16 17 3c, the Brunauer-Emmett-Teller (BET) surface area of ZnCoxNiyOz-SO4 was estimated at 93 m2 18 19 g-1, almost twice that of ZnO particles obtained from the annealing of ZIF-8 (49 m2 g-1). 20 21 22 Additionally, the Barrett-Joyner-Halenda (BJH) average pore size and the integrated pore 23 24 volume were 19.1 nm and 0.47 cm3 g−1 for ZnCoxNiyOz, and 18.1 nm and 0.16 cm3 g-1 for ZnO 25 26 particles. Such nanosheet morphology and mesoporous structure dramatically increase the 27 28 percentage of surface unsaturated atoms, providing a high potential electroactive surface 29 30 area, while facilitating the rapid diffusion of reactants and reaction products. 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 61 62  63 64 65 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 26 27 28 29 30 31 32 33 34 35 36 Figure 2. (a-c) SEM images of a) ZIF-8 MOF particles, b) ZnCoxNiyOz-SO4 precursor, and c) ZnCoxNiyOz-SO4 37 nanosheets. d) TEM image, e) HAADF STEM image, f) HRTEM image, g) corresponding power spectrum (FFT) and 38 h) HAADF STEM images and EELS chemical composition maps of ZnCoxNiyOz-SO4 nanosheets. The EELS maps were 39 40 obtained from the red squared area on the left HAADF STEM micrograph (left). Individual Zn L2,3-edges at 1020 41 eV (red), Co L2,3-edges at 779 eV (green), Ni L2,3-edges at 855 eV (blue), S L2,3-edges at 165 eV (orange), O K-edges 42 43 at 532 eV (pink) and C K-edges at 285 eV (grey) and composites of Zn-Co-Ni, S-O-C and Zn-Co-Ni-S-O-C. 44 45 Figure 3d displays the FTIR spectra of ZnCoxNiyOz samples with different anions. ZnCoxNiyOz- 46 47 SO4 displayed an evident FTIR absorption peak at about 600 cm−1, which is a fingerprint of the 48 49 presence of SO42- ions.[40] The peaks at about 400, 550, and 640 cm-1 can be assigned to 50 51 nitrate, chloride, and acetate, respectively.[41] The anionic groups played an important role 52 53 in determining the crystallinity, morphology and composition of the obtained material, as 54 55 evidenced by XRD, SEM and EDX analysis of ZnCoxNiyOz obtained with the different precursors 56 57 (Figures 3e and S3-6). Because protons react with the surface of ZIF-8, breaking the 58 59 coordinative bond between Zn2+ and 2-methylimidazole linkers in the ZIF-8, thus etching the 60 61 62  63 64 65 MOF and releasing Zn2+ ions into the solution,[37] we hypothesize that the crystallinity and 1 2 architecture dependence on the used anion may be related to a different pH resulting from 3 4 the hydrolysis of the different anions. To confirm this hypothesis, the pH of the solutions 5 6 7 obtained from the dissolution of the same amount of different anionic metal salts was 8 9 determined. The smallest pH value was obtained with the SO42- (4.36) and increased in the 10 11 following order: NO3-(5.75), Cl-(7.02), and OAC-(7.82). Such different pH values strongly 12 13 determine the extent of MOF etching, thus affecting the crystallinity, morphology and also 14 15 composition of the products. We observed that the lower the pH value, the weaker the 16 17 crystallinity, the more favorable formation of nanosheets, and the lower Zn content on the 18 19 final material. 20 21 22 In the same direction, the salt concentration also played an important role. Figures 3f and S7- 23 24 9 show the XRD patterns, SEM images and EDX spectra of ZnCoxNiyOz-SO4 obtained using 25 26 different concentrations of Ni and Co sulfates. When increasing the metal salt concentration, 27 28 deeper hydrolysis and etching take place, which translates into a more amorphous structure, 29 30 a more favorable nanosheet growth and a lower Zn content. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 Figure 3. a) AFM images and b) the corresponding thickness measurement data of ZnCo 1.26Ni0.73Ox-SO4 55 Nanosheets. c) N2 adsorption-desorption isotherms for ZnCo1.26Ni0.73Ox-SO4 and ZnO. d) FTIR spectra and e) XRD 56 57 patterns of ZnCoxNiyOz obtained from different anions. f) XRD pattern of ZnCoxNiyOz-SO4 with different metal 58 content. 59 60 61 62  63 64 65 3.2. OER performance 1 2 3 The OER activity of ZnCoxNiyOz modified with different anions was initially assessed using LSV 4 5 with a conventional three-electrode set-up in a 1 M KOH electrolyte. A commercial IrO2 6 7 electrocatalyst was also tested as a reference. As shown in Figure 4a, the iR-compensated LSV 8 9 curve of ZnCoxNiyOz-SO4 showed a very small overpotential at 10 mA cm-2 (η10), just 252 mV, 10 11 well below that of ZnCoxNiyOz-NO3 (287 mV), ZnCoxNiyOz-Cl (314 mV), ZnCoxNiyOz-OAc (348 mV) 12 13 and IrO2 (323 mV). The significant differences in overpotential demonstrate the important role 14 15 played by the metal precursor counterions. Besides, by modifying the cobalt and nickel content, 16 17 we determined the best OER performance to be obtained with a ZnCo1.26Ni0.73Ox-SO4 18 19 composition (Figure 4b). In contrast, the ZnO produced from the annealing of the ZIF-8 was 20 21 22 characterized by a much poorer performance with a η10 of 407 mV at 10 mA cm−2. 23 24 Figure 4c displays the Tafel plots of the different catalysts. The ZnCo1.26Ni0.73Ox-SO4 exhibited 25 26 the smallest Tafel slope (41.6 mV dec−1) among all the tested catalysts, including commercial 27 28 IrO2, thus demonstrating the fastest reaction kinetics. ZnCo1.26Ni0.73Ox-SO4 also exhibited 29 30 outstanding OER activity and kinetics when compared to some previously reported oxide- 31 32 based OER electrocatalysts (Figure 4d and Table S4). 33 34 35 The ECSA was estimated from the double-layer capacitance (Cdl) measured by CV at different 36 37 scanning rates (Figure S9 a-e). Cdl was calculated from the slope of the linear fitting of the 38 39 charge current versus the scan rate (Figure 4e, Table S3).[27, 42] Using a specific capacity of 40 41 0.04 mF cm−2, ECSA was estimated at 288 cm2 for ZnCo1.26Ni0.73Ox-SO4, well above that of the 42 43 other catalysts evaluated (Table S2).[43] Figure 5f displays the specific activity obtained by 44 45 normalizing the current to the ECSA. Despite its much larger ECSA, ZnCo1.26Ni0.73Ox-SO4 also 46 47 displayed the highest specific activity. We hypothesize this excellent activity to be related to a 48 49 combination of different parameters, including a proper Zn/Co/Ni ratio, the presence of sulfate 50 51 52 ions at the catalyst surface, its amorphous structure and its highly porous nanosheet-based 53 54 architecture. 55 56 EIS was used to measure the interface charge transfer rate.[44] The charge transfer resistance 57 58 (Rct) of ZnCo1.26Ni0.73Ox-SO4 was lower than those of the other samples (Fig. 4g), pointing at a 59 60 faster charge transfer rate, in good agreement with the smallest Tafel slope. 61 62  63 64 65 Figure 4h shows the LSV curves of ZnCoxNiyOz-SO4 after 1000 and 5000 CV cycles. The 1 2 overpotential needed to catalyze water oxidation at 10 mA cm−2 after 1000 and 5000 cycles 3 4 test was virtually the same. 5 6 7 Figure 4h shows the initial LSV curve and the LSV profiles of ZnCoxNiyOz-SO4 after 1000 and 8 9 5000 CV cycles. The overpotential needed to catalyze water oxidation at 10 mA cm−2 after 1000 10 11 and 5000 cycles test was virtually the same. This overpotential was much lower than that 12 13 displayed on the first LSV curve. The chronoamperometry (i-t) durability test for ZnCoxNiyOz- 14 15 SO4 at the initial overpotential, showed a relatively stable and even increasing current density 16 17 during 100 h of continuous operation (Figure 4i), indicating that the ZnCo1.26Ni0.73Oz-SO4 18 19 electrocatalysts have excellent stability under the electrochemical conditions of water 20 21 oxidation. 22 23 24 We hypothesize the overpotential decrease during the first cycles and the gradual increase of 25 26 current density measured in the chronoamperometry test to be associated with progressive 27 28 leaching of zinc ions. To test this hypothesis, the dissolution rate of different elements was 29 30 monitored by analyzing the metal content within the electrolyte using ICP-MS. As shown in 31 32 Figure S11f, negligible amounts of Co and Ni were dissolved during the 18 h 33 34 chronoamperometric test. In contrast, a significant increase in the Zn concentration was 35 36 detected within the electrolyte during the first 12 h of the reaction. The leaching of Zn initially 37 38 generates metal vacancies and active sites within the reconstituted electrocatalyst, which can 39 40 promote the OER performance.[34] 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62  63 64 65 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 26 27 28 29 30 31 32 33 Figure 4. Electrocatalytic OER performance of the various catalysts. a-b) LSV curves of ZnCoxNiyOz-M and 34 ZnCoxNiyOz-SO4. c) Tafel plots. d) Comparison of overpotential at 10 mA cm-2 and Tafel slopes for ZnCo1.26Ni0.73Ox– 35 36 SO4 and some cobalt-based OER catalysts (Table S6). e) Capacitive current density measured at 1.17 V vs. RHE 37 plotted as a function of scan rate. f) ECSA-normalized OER polarization curves. g) Nyquist plots at the potential 38 of 1.52 V vs. RHE. h) LSV curves of ZnCo1.26Ni0.73Ox–SO4 after 1000 and 5000 cycles. i) chronoamperometry curve 39 40 of ZnCo1.26Ni0.73Ox–SO4. 41 42 43 3.3. Structural and chemical evolution 44 45 To further reveal the parameters behind the excellent OER catalytic activity, the surface 46 47 composition and chemical states of ZnCo1.26Ni0.73Oz-SO4 before and after the OER test were 48 49 characterized by XPS (Figure 5). The high-resolution O 1s XPS spectra of ZnCo1.26Ni0.73Oz-SO4 50 51 52 showed the OER test to result in a negative shift of the lattice oxygen (OL) signal located at 53 54 529.7 eV (Figure 5a). We associate this shift with the formation of surface metal (oxy)hydroxide 55 56 that increases the electron density around the lattice O.[45, 46] [47] In addition, the O 1s peak 57 58 at 531.5 eV, associated with surface hydroxide (OHL), negatively shifted by 0.35 eV after the 59 60 OER test. The OHL/OL ratio in the post-OER catalyst (2.50) was significantly higher than that in 61 62  63 64 65 the pristine ZnCo1.26Ni0.73Oz-SO4 (2.03).[16] Interestingly, a clear S 2p XPS signal associated 1 2 with sulfur within a sulfate chemical environment (S 2p3/2 = 168.7 eV) was measured both 3 4 before and after the OER test. Besides the satellite peaks, two Co components were observed 5 6 7 in the Co 2p spectra, and both also shifted to lower binding energies after the OER, from 780.7 8 9 eV to 780.5 eV for Co2+ ( Co 2p3/2) and from 782.4 eV to 782.1 eV for the Co3+ (Co 2p3/2) 10 11 component (Figure 5c). Additionally, the ratio Co3+/Co2+ increased after the OER, from 1.1 to 12 13 1.5, indicating the formation of additional CoOOH species on the surface.[32] In the Ni 2p 14 15 spectrum in Figure 5d, the Ni 2p3/2 peak was deconvoluted into two main peaks at 855.7 and 16 17 857.6 eV corresponding to Ni2+ and Ni3+.[40] After OER, the Ni 2p3/2 peaks had shifted about 18 19 0.6 eV. Besides, the proportion of Ni3+ on the surface greatly increased from 36% to 49%. This 20 21 result implies the transition of nickel species from oxides to (oxy)hydroxide.[48] Notably, the 22 23 XPS signals of Zn became weak with the OER, consistent with the Zn leaching in an alkaline 24 25 medium observed by ICP-MS.[48] This Zn leaching can favor the conversion of cobalt and Ni 26 27 oxide on the surface of ZnCoxNiyOz to CoOx(OH)y and NiOx(OH)y species.[49, 50] Overall, XPS 28 29 results confirmed the surface reconstruction of ZnCo1.26Ni0.73Oz-SO4 during OER, chemically 30 31 evolving from a surface metal oxide to metal (oxy)hydroxide species and loosing Zn during the 32 33 process.[51] 34 35 The EPR spectra of ZnCo1.23Ni0.73Ox-SO4 before and after OER were measured to explore 36 37 38 species with unpaired electrons (Figure 5f).[52, 53] We observed the signal at g = 2.12, 39 40 associated with the presence of oxygen vacancies Vo, to be strongly enhanced after OER test. 41 42 Thus, a large density of oxygen vacancies was generated during OER, which is potentially 43 44 associated with the zinc leaching, and which can favor the OER catalytic activity by introducing 45 46 higher energy dangling bonds at the catalyst surface.[54, 55] 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62  63 64 65 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 5. a-e) O 1s, S 2p, Co 2p, Ni 2p and Zn 2p high resolution XPS spectra. f) EPR spectrum of ZnCo1.26Ni0.73Ox- 26 27 SO4 before and after OER. 28 29 30 After OER, XRD and HRTEM analysis of ZnCo1.26Ni0.73Oz-SO4 displayed the presence of the 31 32 CoOOH crystal phase, formed during OER (Figure 6a-c). In contrast, electron microscopy 33 34 images showed the nanosheet morphology to be preserved during the OER test. EELS 35 36 elemental maps and HAADF-STEM images (Fig. 6d-e) showed a homogeneous dispersion of O, 37 38 S, Zn, Co, and Ni in the cycled material. Besides, the comparison of the ZnCo1.26Ni0.73Oz-SO4 39 40 composition before and after OER, as determined by EDX, ICP-MS and XPS analyses, showed a 41 42 strong decrease in the amount of Zinc (Table S3). 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62  63 64 65 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Figure 6. a) XRD pattern, b) TEM images, c) HRTEM and corresponding power spectrum (FFT), d) HAADF STEM 25 image and e) HADDF EELS image together with the corresponding EELS chemical composition maps of 26 27 ZnCo1.26Ni0.73Ox–SO4 after OER. The EELS maps were obtained from the red squared area on the left HAADF STEM 28 micrograph. Individual Zn L2,3-edges at 1020 eV (red), Co L2,3-edges at 779 eV (green), Ni L2,3-edges at 855 eV (blue), 29 S L2,3-edges at 165 eV (orange), O K-edges at 532 eV (pink) and C K-edges at 284 eV (grey) and composites of Zn- 30 31 Co-Ni, S-O-C and Zn-Co-Ni-S-O-C. 32 33 34 3.4. DFT calculations 35 36 DFT calculations were carried out to better understand the role of surface SO42− on the 37 38 enhancement of the OER activity. Owing to the complexity of the ZnCo1.23Ni0.73Ox-SO42− 39 40 amorphous structure, we used a cobalt oxohydroxide surface as a model system. First, we 41 42 determined the CoOOH and CoOOH-SO4 OER active sites. Besides the charge density and 43 44 density of state (DOS) were calculated. Figures 7a and S11a show the optimized structural 45 46 47 models of CoOOH-SO4 and CoOOH for different OER steps. As displayed in Figure 7b, with the 48 49 adsorption of the SO42−, the contribution of the Co 3d orbital to the DOS at the Fermi level 50 51 strongly increases.[20, 56] 52 53 The Gibbs free energies of OER pathways on CoOOH-SO4 and CoOOH are displayed in Figures 54 55 6b and S11b,c. The formation of OOH* was determined to be the rate-determining step (RDS) 56 57 on CoOOH. In the presence of the sulfate ion, CoOOH-SO4 displayed a much lower free energy 58 59 for OOH* generation when compared with CoOOH, which is consistent with the higher OER 60 61 62  63 64 65 activity experimentally measured for this material.[18, 57] 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 26 27 28 29 30 31 32 Figure 7. a) Optimized structural models of CoOOH-SO4 for different OER steps. White balls represent hydrogen 33 34 atoms, blue balls is cobalt atoms, and yellow balls stands for sulfur atoms, red ball represent the oxygen in the 35 catalyst structure; green ball is the oxygen adsorbed on the intermediate; pink ball is the oxygen on the sulfate 36 group. b) DOS of CoOOH and CoOOH-SO4 regarding the Co 3d orbitals. c) Gibbs free energy diagrams of OER 37 38 intermediates over CoOOH and CoOOH-SO4. 39 40 41 4. Conclusions 42 43 In summary, the influence of several counterions on the structural, chemical and functional 44 45 properties of ZnCo1.23Ni0.73Ox electrocatalyst was analyzed. Materials were produced via a 46 47 48 facile etching and annealing methods using ZIF-8 as a self-sacrificial template. ZnCo1.23Ni0.73Ox- 49 50 SO4 displayed excellent OER performances, including a low Tafel slope of 41.6 mV dec−1, a very 51 52 small overpotential of 252 mV for 10 mA cm−2, and high stability during 100 h continuous 53 54 operation. The high OER activity is associated with several factors, including the presence of 55 56 oxygen vacancies, zinc leaching and the presence of sulfate anions. CV, I-T, XPS and EPR results 57 58 indicate the simultaneous leaching of Zn, formation of oxygen vacancies and electrochemical 59 60 surface reconstruction to CoOOH active species. DFT calculations proved that the decorated 61 62  63 64 65 SO42- facilitates the formation of OOH* on CoOOH active sites, thus accelerating the OER 1 2 process. 3 4 5 Acknowledgements 6 7 8 The authors thank the support from the projects COMBENERGY (PID2019-105490RB-C32) 9 10 from the Spanish Ministerio de Ciencia e Innovación, ENE2016-77798-C4-3-R and NANOGEN 11 12 (PID2020-116093RB-C43) funded by MCIN/AEI/10.13039/501100011033/ and by “ERDF A way 13 14 of making Europe”, by the “European Union”. X. W., X. H. and R. D. thank the China Scholarship 15 16 17 Council (CSC) for the scholarship support. The authors acknowledge funding from Generalitat 18 19 de Catalunya 2017 SGR 327 and 2017 SGR 1246. ICN2 acknowledges the Severo Ochoa 20 21 program from Spanish MINECO (Grant No. SEV-2017-0706). IREC and ICN2 are funded by the 22 23 CERCA Programme /Generalitat de Catalunya. Z. L. acknowledges funding from MINECO SO- 24 25 FPT PhD grant (SEV-2013-0295-17-1). Li is grateful for the project No.22NSFSC0966 provided 26 27 by the Science & Technology Department of Sichuan Province. J.L. is a Serra Húnter Fellow and 28 29 is grateful to MICINN/FEDER RTI2018-093996-B-C31, GC 2017 SGR 128 and ICREA Academia 30 31 program. Part of the present work has been performed in the frameworks of Universitat de 32 33 Barcelona Nanoscience PhD program and Universitat Autònoma de Barcelona Materials 34 35 Science PhD program. 36 37 38 39 References 40 41 42 >@ - 6RQJ & :HL =) +XDQJ & /LX / =HQJ ; :DQJ =- ;X $ UHYLHZ RQ IXQGDPHQWDOV IRU 43 44 GHVLJQLQJR[\JHQHYROXWLRQHOHFWURFDWDO\VWV&KHP6RF5HY   45 46 47 >@-'X)/L/6XQ0HWDORUJDQLFIUDPHZRUNVDQGWKHLUGHULYDWLYHVDVHOHFWURFDWDO\VWVIRUWKHR[\JHQ 48 49 50 HYROXWLRQUHDFWLRQ&KHP6RF5HY   51 52 53 >@.:DQJ;:DQJ=/L%@/*DR;&XL&'6HZHOO-/L=/LQ5HFHQWDGYDQFHVLQDFWLYDWLQJVXUIDFHUHFRQVWUXFWLRQIRU 2 3 WKHKLJKHIILFLHQF\R[\JHQHYROXWLRQUHDFWLRQ&KHP6RF5HY   4 5 6 >@=3:X;)/X64=DQJ;:/RX1RQಣ1REOHಣ0HWDOಣ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ಣ 32 33 34 ,URQ R[\ +\GUR[LGHZLWK(QKDQFHG2[\JHQಣ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ದ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ಣ3URPRWHG6XUIDFH5HFRQVWUXFWLRQ 13 14 LQ3RO\PRUSKLF&REDOWIRU,QLWLDWLQJ5DSLG2[\JHQ(YROXWLRQ$GY(QHUJ\0DWHU   15 16 17 >@-=KDQJ74XDVW:+H6'LHFNK¸IHU-5-XQTXHLUD'˜KO3:LOGH'-DPEUHF<7&KHQ 18 19 20 :6FKXKPDQQ,QಣVLWX&DUERQ&RUURVLRQDQG&X/HDFKLQJDVD6WUDWHJ\IRU%RRVWLQJ2[\JHQ(YROXWLRQ 21 22 23 5HDFWLRQLQPXOWLಣPHWDO(OHFWURFDWDO\VWV$GY0DWHU   24 25 >@6:DKO60(Oಣ5HIDHL$*%X]DQLFK3$PVDOHP.V/HH1.RFK0/'RXEOHW13LQQD 26 27 28 =Q &R 2ದ$ VWDEOH DQG KLJKO\ DFWLYH R[\JHQ HYROXWLRQ FDWDO\VW IRUPHG E\ ]LQF OHDFKLQJ DQG 29 30 31 WHWUDKHGUDOFRRUGLQDWHGFREDOWLQZXUW]LWHVWUXFWXUH$GY(QHUJ\0DWHU   32 33 34 >@;)/X<)DQJ'/XDQ;:'/RX0HWDOದRUJDQLFIUDPHZRUNVGHULYHGIXQFWLRQDOPDWHULDOVIRU 35 36 HOHFWURFKHPLFDOHQHUJ\VWRUDJHDQGFRQYHUVLRQDPLQLUHYLHZ1DQROHWWHUV   37 38 39 >@ 0 @::DQJ+@ : &KHQJ =3 :X ' /XDQ 64 =DQJ ;: /RX 6\QHUJHWLF &REDOWಣ&RSSHUಣ%DVHG %LPHWDOದ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ದ=Q R[\K\GUR[LGH R[\JHQ HYROXWLRQ 43 44 45 HOHFWURFDWDO\VWV1DW(QHUJ\   46 47 >@0)DQJ'+DQ:%;X<6KHQ</X3&DR6+DQ:;X'=KX:/LX6XUIDFHಣJXLGHG 48 49 50 IRUPDWLRQRIDPRUSKRXVPL[HGಣ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ಣ6SLQ6WDWH&REDOW&DWLRQVLQ6SLQHO 18 19 20 =LQF &REDOW 2[LGH IRU 6SLQ &KDQQHO 3URSDJDWLRQ DQG $FWLYH 6LWH (QKDQFHPHQW LQ :DWHU 2[LGDWLRQ 21 22 23 $QJHZ&KHP,QW(G   24 25 >@$%HUJPDQQ7(-RQHV(0DUWLQH]0RUHQR'7HVFKQHU3&KHUQHY0*OLHFK75HLHU+ 26 27 28 'DX36WUDVVHU8QLILHGVWUXFWXUDOPRWLIVRIWKHFDWDO\WLFDOO\DFWLYHVWDWHRI&R R[\K\GU R[LGHVGXULQJ 29 30 31 WKHHOHFWURFKHPLFDOR[\JHQHYROXWLRQUHDFWLRQ1DW&DWDO   32 33 34 >@ - +H ; =KRX 3 ;X - 6XQ 3URPRWLQJ HOHFWURFDWDO\WLF ZDWHU R[LGDWLRQ WKURXJK WXQJVWHQ 35 36 PRGXODWHGR[\JHQYDFDQFLHVRQKLHUDUFKLFDO)H1LOD\HUHGGRXEOHK\GUR[LGH1DQR(QHUJ\   37 38 39  40 41 42 >@/=KDQJ&/X)@'*XDQ*5\X=+X-=KRX&/'RQJ<&+XDQJ.=KDQJ<=KRQJ$&.RPDUHN0 51 52 53 =KX;:X&:3DR&.&KDQJ+-/LQ&7&KHQ:=KRX=6KDR8WLOL]LQJLRQOHDFKLQJHIIHFWV 54 55 56 IRUDFKLHYLQJKLJKR[\JHQHYROYLQJSHUIRUPDQFHRQK\EULGQDQRFRPSRVLWHZLWKVHOIRSWLPL]HGEHKDYLRUV 57 58 1DW&RPPXQ   59 60 61 62  63 64 65 1 >@5,6D\OHU%0+XQWHU:)X+%*UD\5'%ULWW(35VSHFWURVFRS\RILURQDQGQLFNHOGRSHG 2 3 >=Q$O@OD\HUHGGRXEOHK\GUR[LGHV0RGHOLQJDFWLYHVLWHVLQKHWHURJHQHRXVZDWHUR[LGDWLRQFDWDO\VWV- 4 5 6 $P&KHP6RF   7 8 9 >@-:DQJ-/LX%=KDQJ)&KHQJ<5XDQ;-L.;X&&KHQ/0LDR--LDQJ6WDELOL]LQJ 10 11 12 WKHR[\JHQYDFDQFLHVDQGSURPRWLQJZDWHUR[LGDWLRQNLQHWLFVLQFREDOWR[LGHVE\ORZHUYDOHQFHVWDWH 13 14 GRSLQJ1DQR(QHUJ\   15 16 17 >@:;X:*DR/0HQJ:7LDQ//L,QFRUSRUDWLRQRI6XOIDWH$QLRQVDQG6XOIXU9DFDQFLHVLQ 18 19 20 =Q,Q63KRWRDQRGHIRU(QKDQFHG3KRWRHOHFWURFKHPLFDO:DWHU6SOLWWLQJ$GY(QHUJ\0DWHU   21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62  63 64 65 *UDSKLFDO$EVWUDFW Graphical Abstract Different anionic groups decorated ZnCoxNiyOz nanosheets are produced by a facile proton etching and ion exchange method used ZIF-8 as templates. Unveiling the role of counter-anions on electrocatalyst performance. Oxygen defects caused by zinc leaching and the presence of sulfate anions improved the OER activity.