Physicochemical properties. The typical Ni2Fe(CN)6 catalyst was prepared on nickel foam through a self-assembly method by immersing cleaned Ni foams in the mixed solution of polyvinylpyrrolidone, C6H5Na3O7·2H2O, NiCl2·6H2O and K3[Fe(CN)]6 using different aging times and salt concentrations. Scanning electron microscope (SEM) image shows that the as-synthesized product is composed of nanocubes assembled on the surface of nickel foam with a single layer coating (Fig. 1a). Transmission electron microscope (TEM) image shows a highly crystalline cubic nanoparticle with an edge length of 180 ± 10 nm (Fig. 1a inset). Different reaction conditions can be used to tune the nanocube size and coverage on the nickel foam (Supplementary Fig. 1, 2). Analysis of the X-ray powder diffraction (XRD) pattern obtained for Ni2Fe(CN)6 powder (Fig. 1b) allowed for identification of a single cubic phase where Fe atom coordinates with carbon atoms in CN− species and Ni atom has two coordination forms: one coordinates with N atoms in CN− species and the other is situated in the center alone.21,22 Interestingly, both Fe2+ and Ni2+ species were not oxidized during UOR process (Fig. 1c, d), which differs from the mechanism reported for other Ni-based catalysts. 9, 15
UOR performance evaluation. Since human urine contains 2-2.5 wt. % of urea (equals to a molar concentration of ~ 0.33 M), 0.33 M urea was chosen in the electrolysis.2, 12 As shown by the linear sweep voltammetry (LSV), UOR on an optimized Ni2Fe(CN)6 catalyst exhibits a more negative onset potential than that of OER (Fig. 2a, Supplementary Fig. 3). To obtain a current density of 100 mA cm− 2, UOR needs a potential of 1.35 V, much smaller than that for OER (1.68 V). More importantly, as shown in Fig. 2a, the urea oxidation on Ni2Fe(CN)6 proceeds before the self-oxidation of Ni2+ to Ni3+. This is very different from the mechanism reported for other reported Ni-based catalysts, in which a NiOOH phase is formed before UOR and serves as the active sites.15 Specifically, for Ni2Fe(CN)6 catalyst a very small apparent UOR activation energy of 14.0 kJ/mol (at a potential of 1.39 V) is observed (Supplementary Fig. 4). Importantly, the potential of 1.35 V to achieve an UOR current density of 100 mA·cm− 2 on Ni2Fe(CN)6 is the lowest one among all known UOR electrocatalysts reported including nickel hydroxides, metals, phosphides, etc.10,13,16,23−33 In addition, the current density at a certain potential (e.g. 1.40 V) on Ni2Fe(CN)6 is 2.5 times higher than that obtained for the state-of-the-art nickel-based supported electrocatalysts (Supplementary Fig. 5, Table 2).14 Interestingly, it was also found that two other cyanides, ferric ferrocyanide (Fe4[Fe(CN)6]3) and nickel cobaltcyanide (Ni3[Co(CN)6]2), showed worse activity compared to Ni2Fe(CN)6 (Fig. 2b, Supplementary Fig. 6).34,35 Therefore, we speculate that the observed high performance of Ni2Fe(CN)6 is caused by cooperative action of two active sites of Ni and Fe in Ni2Fe(CN)6 catalyst.
The kinetic study shows that the UOR process on Ni2Fe(CN)6 is independent of urea concentration (Fig. 2c), which agrees with the reported so far Ni-based electrocatalysts.36 As can be seen in Fig. 2d, UOR shows a strong dependence on the amount of KOH with a reaction order of 1.10 to OH− concentration, which is different from that of about 2.00 for other Ni-based electrocatalysts.36 For example, a reaction order of 1.75 was obtained on a conventional UOR catalyst of NiC2O4 (Supplementary Fig. 7,8).37 It was revealed that the UOR’s RDS on NiOOH intermediated electrocatalysts is desorption of CO2 from NiOOH active sites, which requires a 2 M OH− to form CO32− in the electrolyte. 15, 38–40 Therefore, the reaction order of 1.10 on the newly developed Ni2Fe(CN)6 catalyst indicates an alternative reaction pathway and RDS.
Comparison to the conventional NiC 2 O 4 catalyst. The above statement was validated by in-situ Raman spectroscopy studies of two types of catalysts, Ni2Fe(CN)6 and conventional NiC2O4 (Supplementary Fig. 9). As shown in Fig. 3a, two strong peaks at 2100 and 2140 cm− 1 are characteristic of cyanide stretching in Ni2Fe(CN)6 complex,20 while the peaks at 250 and 348 cm− 1 belong to the Ni-N stretching vibration and the peak at 510 cm− 1 belongs to the Fe-C stretching vibration, respectively.22 Importantly, at various potentials and different reaction times (Fig. 3a, Supplementary Fig. 10a), the Raman spectra of Ni2Fe(CN)6 are similar to that at the open circuit potential (OCP). However, for NiC2O4, the NiOOH doublet peaks at 473 and 560 cm− 1 appear in a very short time at a high potential during UOR process (Fig. 3b and Supplementary Fig. 10b), indicating it is partially reconstructed to NiOOH, which serves as the active phase for UOR and agrees with the existing literature.38,39 As expected, Ni2Fe(CN)6 without NiOOH species showed a significantly enhanced apparent UOR activity and stability compared to NiC2O4 catalyst (Fig. 3c). About 90% of the current density remained after 50 hour operation on Ni2Fe(CN)6, which is a record for UOR catalysts.
The new reaction pathway of UOR for Ni2Fe(CN)6 was further investigated by the real-time ammonia detection in the electrolyte using an ion ammonia-selective electrode (Orion™ High-Performance Ammonia Electrode 9512HPBNWP, Supplementary Fig. 11).41− 43 As shown in Fig. 3d, a large amount of ammonia was produced and detected on Ni2Fe(CN)6 under a potential of 1.36 V (current density of 70.0 mA cm− 2), while very little was detected on NiC2O4 at similar conditions (current density of 72.0 mA cm− 2 at a potential of 1.45 V). It was further found that Ni2Fe(CN)6 also can efficiently catalyze the oxidation of ammonia (Supplementary Fig. 12). Therefore, we propose that the UOR process on Ni2Fe(CN)6 contains *NH3 as a key reaction intermediate, which is very different from the currently reported mechanisms involving *NCO, *HN-CO, or *COO as the intermediates.38,39
DFT computations. Based on the above analysis of kinetics and real-time ammonia detection, we propose a new mechanism for UOR, which comprises of two steps, namely production of NH3 and oxidation of NH3 (the detailed pathways can be found in the experimental section). The Gibbs free energy of each reaction intermediate was computed by density functional theory (DFT, Supplementary Table 1 and Supplementary Figs. 13–16). For the first step (Fig. 4a), Ni sites were proven to be the primary active sites because the free energy value of RDS ([M·OCO2]ads + 2e− → M + CO32−) on Ni site is smaller than that on Fe site. For the second step (Fig. 4b), the biggest difference is in the formation of *NH·NH2 intermediates, which reveals that the dehydrogenation reaction ([M·NH2·NH2]ads + OH− → [M·NH·NH2]ads + H2O + e−) is harder to occur on Ni sites than Fe sites (free energy of 1.23 vs 0.94 eV). For the overall reaction pathways, the calculated RDS is the formation of *NH + * NH2 (IMFe9 to IMFe10 as shown in Fig. 4b), in which 1 M *NH2 + *NH2 react with 1 M OH−, which yields the reaction order of 1.0 with respect to the OH− concentration. This result agrees well with the experimentally determined reaction order of 1.10 on Ni2Fe(CN)6 catalyst (Fig. 2d). Overall, it is proposed that Ni is responsible for conversion of urea into ammonia and carbonate, while Fe is responsible for transformation of ammonia into nitrogen (Fig. 4c). This synergistic catalysis between Ni and Fe sites revealed by DFT computation agrees well with experimental observation that Ni2Fe(CN)6 exhibits remarkably higher activity than other two types of Fe4[Fe(CN)6]3 and Ni3[Co(CN)6]2 catalysts (Fig. 2b).
Identification of intermediates. We performed in-situ synchrotron radiation-Fourier transform infrared spectroscopy (SR-FTIR) analysis to identify the critical reaction intermediates proposed in the DFT computation and real-time ammonia detection. As shown in Fig. 5a, compared with the spectrum at OCP, two obvious absorption bands appear at 2925 cm− 1 and 1203 cm− 1 under the UOR working potential (e.g. 1.35–1.65 V), which can be assigned to the N-H stretching vibration of *N = NH2+ and C-O stretching vibration of *OCONH2 species, respectively.44, 45 In addition, with increasing potential and operating time, these two characteristic peaks become stronger (Fig. 5b-c and Supplementary Fig. 17). The simulated results of harmonic vibrational frequencies (Fig. 5d) also indicate that two peaks can be attributed to *N = NH2+ and *OCONH2 species. This clearly indicates that these two intermediates are produced in the UOR process, which supports the DFT computation with IMFe12 and IMNi3 intermediates. At this stage, the combination of kinetics analysis, DFT computation and in-situ SR-FTIR spectroscopy data confirms a new UOR mechanism on Ni2Fe(CN)6 without NiOOH generation, namely two-step processes of NH3 production and oxidation at two different active sites.
Energy-saving systems driven by UOR for replacement of OER. To establish an energy-saving system benefiting from the low overpotential of UOR on Ni2Fe(CN)6 catalyst, we assembled an UOR//HER electrolyzer using Ni2Fe(CN)6 as the anode in an electrolyte containing 1 M KOH and 0.33 M urea. For comparison, an OER//HER electrolyzer with RuO2 as the anode for OER was performed in 1.0 M KOH (Supplementary Fig. 18). To obtain a current density of 10 and 100 mA cm− 2, urea electrolysis needs a cell voltage of 1.38 and 1.50 V, respectively, whereas water electrolysis needs the value of 1.56 and 1.85 V (Fig. 6a). This clearly indicates the energy-saving advantage of the UOR process on Ni2Fe(CN)6 electrocatalyst to replace OER. In addition, the H2 production in this UOR//HER cell was very stable with a Faradaic efficiency higher than 90% (Fig. 6b, Supplementary Fig. 19). Besides H2 production, the energy-saving system can also be applied in a UOR//2e−ORR flow cell composed of Ni2Fe(CN)6 anode and mesoporous carbon (CMK-3) cathode 46 (Supplementary Fig. 20). Nowadays, in-situ electrochemical production of H2O2 via 2e− ORR has become a promising way because it can reduce the danger and costs of the transportation of H2O2.46, 47 As expected, the urea electrolysis needs a smaller energy input than water electrolysis for urea elimination and H2O2 generation (Supplementary Fig. 21). Specifically, a H2O2 production rate of 225.3 g m− 2h− 1 (with a Faradaic efficiency of 82.3%) and an urea elimination rate of 140.1 g m− 2h− 1 (with a Faradaic efficiency of 94.9%) were achieved at the cell voltage of only 0.63 V (Fig. 6c, Supplementary Fig. 22). In addition, the urea elimination and H2O2 generation were very efficient for urea concentrations varying from 0.0033 to 0.33 M (Fig. 6d, Supplementary Fig. 23), which is in the range of industrial urea-containing wastewater.