Figure 1. a) In situ characterization and surface reconstruction processes of transition metal phosphides. b) HRTEM-HAADF and element distribution of the Ni-CoP. c) HRTEM images of Ni-CoP. d) High resolution XPS-Co 2p and e) soft XAS-Co L-edge of CoP and Ni-CoP.
2.2. Alkaline water splitting properties
An important limiting factor in the water splitting reaction rate and energy consumption is the anode OER[31-33]. Therefore, anode catalysts with excellent performance and stability are required to achieve efficient green hydrogen production. Covalent-like bonding mode between transition metal sites and non-metal sites in CoP-based catalyst can lead to excellent thermodynamic behavior and conductivity. The electrochemical OER properties of CoP-based catalysts were studied in the electrolyte solution 1 M KOH, which exhibits lower overpotentials of 236, 287, and 291 mV to reach 10, 150 and 300 mA cm-2, respectively (Table S1, Figure 2a and Figure S7a,b). Such overpotential is competitive among the other Co-based phosphide or noble metal catalysts reported in the related reference (Table S2 and Figure S7c). The Tafel slope and Rct of Ni-CoP nanowire are fitted as 52.55 mV dec-1 and 1.80 Ω, which is much lower than that of other M-CoP catalyst and 6 mg-commercial catalysts (Figure 2b, Figure S7d-f). The above results indicate that Ni-CoP nanowire has excellent reaction kinetics and charge transport behavior. In addition, the double layer capacitor (Cdl) of sample are calculated as 698.84 (CoP), 680.51 (Cr-CoP), 621.43 (Fe-CoP) and 538 mF cm-2 (Ni-CoP) (Figure S8a-d). The linear scanning voltammetry (LSV) curves of all samples are normalized by Cdl, which is purposed to study the intrinsic catalytic activity. The normalized LSV curves exhibit that OER performance of Ni-CoP nanowires is still better than that of CoP and other M-CoP nanowires (Figure S8e,f). To explore the relationship between heteroatom Ni content and catalytic performance, the heteroatom Ni doping content is controlled by the molar ratio between nickel nitrate and cobalt nitrate, which has been described in the experimental part. Similar to Ni-CoP-0.1 mmol, CoP nanowires with different heteroatom Ni content (0.15 mmol and 0.2 mmol) are successfully prepared in CC, which was confirmed in SEM and XRD (Figure S9). Compared with Ni-CoP-0.15 mmol and 0.2 mmol, Ni-CoP-0.1 mmol nanowire still exhibit the smallest overpotential at 10, 150 and 300 mA cm-2, which indicate most excellent OER performance (Table S3 and Figure S10).
Similarly, three-electrode system was used to analyze the hydrogen evolution reaction (HER) activity of CoP-based catalysts in 1 M KOH. The Ni-CoP nanowire with superior HER performance still exhibits lower overpotential at 100, 200 and 400 mA cm-2, which is 99, 163 and 298 mV lower than that of bare CoP, respectively (Table S4, Figure 2c and Figure S11b). Specific current density can be achieved by Ni-CoP at lower applied potentials, which are significantly better than bare CC, Co(OH)F, CoP, other M-CoP nanowires, commercial Pt/C and other TMPs-based catalysts (Table S5, Figure S11a,c). The Tafel slope of Ni-CoP (40.54 mV dec-1) is much smaller than CoP (63.17 mV dec-1), commercial Pt/C (67.82 mV dec-1) and other M-CoP nanowires, revealing that the HER kinetic are accelerated by heteroatom Ni (Figure S11d). The Ni-CoP nanowires with lower Tafel slope and charge transfer resistance (Rct = 1.837 Ω) exhibit ultrafast reaction kinetics based on Volmer-Heyrovsky mechanism during HER (Figure S11e,f). The Cdl of CoP, Cr-CoP, Fe-CoP and Ni-CoP nanowires are calculated to be 35.34 mF cm-2, 29.38 mF cm-2, 20.39 mF cm-2 and 19.94 mF cm-2, respectively (Figure S12a-d). Besides, the normalized LSV curves of Ni-CoP nanowire has the most excellent catalytic activity, which originates from the regulation behavior of heteroatom Ni on the electronic structure around Co sites (Figure S12e,f). For relationship between different Ni doping amounts and HER performance, Ni-CoP-0.1 mmol nanowire exhibits the smallest overpotential (at 100, 200, 300 mA cm-2), which imply excellent HER performance compared with other heteroatom contents (Table S6 and Figure S13).
Considering the excellent activity and reaction kinetics exhibited by Ni-CoP nanowires (even better than 6 mg-commercial Pt/C and RuO2 catalysts), the Ni-CoP nanowire grown on CC were used as cathode/anode in the electrolytic cell (two-electrode), which can directly evaluate the feasibility of the electrolytic water splitting system (Ni-CoP || Ni-CoP) in practical applications. Compared with commercial Pt/C||RuO2 catalytic system (1.62 V), the Ni-CoP || Ni-CoP electrodes exhibit an excellent overall water splitting performance, which can reach a current density of 10 mA cm-2 at 1.59 V (Figure 2d). For other recently reported Co-based bifunctional catalysts, the excellent water-splitting performance exhibited by the Ni-CoP || Ni-CoP catalyst with lower overpotential is still competitive (Table S7 and Figure S14). The chronoamperometric curve collected at current density of 10 mA cm-2 demonstrate that the corresponding potential of Ni-CoP || Ni-CoP catalyst failed to show noticeable change for at least 24 h (Figure S15a). In addition, compared with the LSV curves of Ni-CoP nanowires before the stability test, the Ni-CoP nanowire demonstrate superior durability with no obvious attenuation after continuous operation 24 h (Figure S15b). Notably, the SEM images exhibit that the Ni-CoP electrode retains the original nanowire structure even after the HER test. It is worth noting that the Ni-CoP electrode still maintained the pristine nanowire structure after the long-time HER test, which is obviously different from the Ni-CoP electrode after the long-time OER test (inset in Figure 2e). Considering the deep reconstruction phenomenon under long-term working conditions, high-resolution XPS were used to analyze the elemental composition and chemical valence state after the OER stability test (current density = 10 mA cm-2, acquisition time = 24 h). The high-resolution XPS Co 2p peak signals of the double-transition metal active sites in Ni-CoP nanowire are not annihilated by the long-term alkaline oxygen evolution environment, which reveal the fundamental source of the excellent stability (Figure S16a). Subsequently, it can be clearly seen that Co 2p in the high-resolution XPS image is composed of spin-splitting signals introduced by Co3+ and Co4+, including Co 2p3/2, Co 2p1/2 and satellite peaks. The high-resolution P 2p spectrum exhibits a distinct typical peak introduced by the P-O bond, which is attributed to the re-adsorption behavior of PO43- on the catalyst surface (Figure S16b). However, the signals introduced by Co-P bonds in pristine Ni-CoP could not be observed, corresponding to the dissolution process of Ni-CoP at low positive potential. For alkaline OER, OH- adsorption, proton desorption and electro-oxidation processes are index to accompanying phenomena of the surface reconstruction behavior. Compared with pristine Ni-CoP, the high-resolution O 1s of Ni-CoP nanowires after the stability test exhibit obvious peaks at 529.59 eV, 530.86 eV, 531.53 eV, and 533.37 eV, which are indexed by lattice oxygen (Co-O), O22- species, OH-group (or Vo) and adsorbed H2O molecules (Figure S16c). In particular, the signal peak provided by the O22- and OH- group represents the incomplete evolution stage of CoOOH to CoO2 species, which implies that the OER active phase of Ni-CoP is identified as Ni-CoOOH/CoO2 species after stability test. Besides, chronopotentiometry curves under multicurrent step and large current density were used to further study the stability of Ni-CoP in water electrolysis reactions. The multicurrent chronopotentiometry curve of Ni-CoP nanowire with the current increasing from 10 to 190 mA cm-2 (increase at a rate of 10 mA cm-2 per 10 min) (Figure S15c). The stable overpotential can be clearly observed, which undoubtedly further confirms the excellent stability and mass transfer properties (Figure S15d). The long-term stability of Ni-CoP || Ni-CoP system was examined by a constant current chronopotentiometry at the current density of 10 and 100 mA cm-2 for 100 h. Compared with previously reported noble metal-based catalysts, the overpotential of the Ni-CoP system only decreased by 0.61% and 2.62% after water splitting reaction for 100 h, which confirms the reference value of the Ni-CoP system in the green hydrogen production process (Figure 2e).