Vanadium-Incorporated CoP 2 with Lattice Expansion for Highly Efficient Acidic Overall Water Splitting

Proton exchange membrane water electrolyzer (PEMWE) in acidic medium is a hopeful scenario for hydrogen production by using renewable energy, but the grand challenge  lies in substituting noble-metal catalysts. Herein, a robust electrocatalyst of V-CoP2 porous nanowires arranged on carbon cloth is successfully fabricated via  incorporating vanadium into CoP2 lattice. Structural characterizations and theoretical analysis indicate that lattice expansion of CoP2 caused by V incorporation results in the upshift of d-band center, which is conducive to hydrogen adsorption for boosting hydrogen evolution reaction (HER). Besides, V promotes surface  reconstruction to generate a thicker Co3O4  layer with oxygen vacancy that enhances acid-corrosion resistance and optimizes the adsorption of water and  oxygen-containing species, thus improving activity and stability toward oxygen evolution reaction (OER). Accordingly, it presents a superior acidic overall  water splitting activity (1.47 V@10 mA cm-2) to Pt-C/CC||RuO 2 /CC (1.59 V@10 mA cm-2), and remarkable stability. This work proposes a new route to  design efficient non-noble metal electrocatalysts for PEMWE.


Supporting Information
The content of SI 1. Experimental section 2. Figure S1. SEM images of (a, b) bare carbon cloth.
17. Figure S16. Calculation of exchange current density of CoP2/CC, V-CoP2/CC and Pt-C/CC in 0.5 M H2SO4 for HER.
18. Figure S17. XRD patterns and SEM images of (a, b) V-CoP2/CC before HER and (c, d) V-CoP2/CC after HER.
Synthesis of V-CoP2 porous nanowire. In a typical synthesis, 0.05 mmol of NH4VO3, 0.95 mmol of Co(NO3)2·6H2O, 5 mmol of CO(NH2)2, and 2 mmol of NH4F were dissolved in 25 mL of deionized water and stirred for 2h to form a homogeneous solution. Afterward, the mixture was transferred to a 50-mL Teflon-lined stainless steel autoclave containing CC (3 × 3 cm -2 ) and hydrothermally treated at 120 o C for 4h. After cooling to the room temperature, the V-CoOOH precursors grown on the CC (V-CoOOH/CC) were ultrasonically cleaned with deionized water and ethanol for several times and dried in an oven at 60 o C overnight. Finally, 1.2 g of NaH2PO2 powder and V-CoOOH/CC were placed on the upstream and downstream of the quartz tube, respectively. Before phosphorization, N2 was purged for 20 min to remove the air. The V-CoOOH/CC was heated at 600 o C with a ramp rate of 2 o C min -1 and maintained at 600 o C for 2h. After cooling to room temperature, the V-CoP2/CC was harvested, in which the V incorporation mass ratio is 5%. For comparison, CoP2/CC were prepared with similar process to that employed for the fabrication of V-CoP2/CC, except that only adding Co source.
Additionally, a series of control experiments were performed. Specifically, the corresponding phosphorization products with other hydrothermal times of 2 and 6h were prepared (labeled as V-CoP2/CC-2 and V-CoP2/CC-6). The V-CoP2/CC samples with other V incorporation mass ratios of 2.5% and 10% were prepared (named as V-CoP2/CC-2.5% and V-CoP2/CC-10%).
The V-CoOOH/CC was treated at the phosphorization temperatures of 500 and 700°C and labeled as V-CoP2/CC-T500 and V-CoP2/CC-T700. The V-CoOOH/CC was phosphatized for 1 and 3h and tagged as V-CoP2/CC-T1h and V-CoP2/CC-T3h. Co3O4 grown on the CC (named as Co3O4/CC) was prepared by calcinating V-CoOOH/CC in an air atmosphere at 350°C for 2h.
In addition, the commercial Pt-C and RuO2 were coated on CC substrate (1.5 mg cm -2 ), respectively, and labeled as Pt-C/CC and RuO2/CC. Raman was conducted with a Jobin Yvon HR 800 micro-Raman spectrometer at 457.9 nm. Xray photoelectron spectroscopy (XPS) characterization was conducted on a VG ESCALAB MK II with the excitation source of Mg Kα (1253.6 eV) achromatic X-ray radiation. The N2 adsorption-desorption isotherms of the samples were performed by Micromeritics Tristar II.
Scanning Kelvin Probe (SKP5050 system, Scotland) were performed in ambient atmosphere with a gold electrode as the reference electrode. The amounts of gases were quantified by gas chromatography (Aglient, 7820A).
Electrochemical measurements. Electrochemical measurements were performed with a CHI660E electrochemical workstation. A three-electrode configuration at room temperature was used, where self-supported electrocatalytic materials (1×1.5 cm -2 ), graphite rod and saturated calomel electrode (SCE) were used as the work electrode, the auxiliary electrode and the reference electrode, respectively. All the potentials were recorded with respect to the reversible hydrogen electrode (RHE) and 95% iR-corrected according to the equation ERHE= ESCE + 0.244 V+0.059pH-iR. Linear sweep voltammetry was collected in 0.5 M H2SO4 with the scan rate of 5 mV s -1 . The potential cycling stability test was examined by taking continuous cyclic voltemmgrams (CVs) at a scan rate of 100 mV s -1 . The I-t chronoamperometric test was tested for 12h at the potential required 10 mA cm -2 . Electrochemical impedance spectroscopy (EIS) was performed with frequency range from 1 Hz to 100 kHz. CV curves were conducted with different rates from 20 to 100 mV s -1 to determine double-layer capacitance. The voltage range for overall water splitting was tested from 1 to 2 V. The Faradaic efficiencies for HER and OER were measured by comparing the experimental and theoretical amounts of H2 and O2. In test, chronoamperometric was performed at 100 mA cm -2 for 60 minutes in the airtight H-type electrolytic cell. The amount of gas generated was analyzed by gas chromatography (GC).
Computational details. The projector augmented-wave (PAW) technique with the set planewave energy cutoff of 300 eV was conducted [1] . Perdew-Burke-Ernzerhof (PBE) functional was employed to settle the correlation-exchange energies of the systems [2] . The sampling over Brillouin zone was treated by a (2×2×1) Monkhorst-Pack grid, and a vacuum slab with the length of 10 Å was placed along z axis on each slab to avoid the pseudo interactions between periodic images. Geometry optimization was repeated until the total energy tolerance was converged to 2×10 -5 eV and the changes of the force on the atoms less than 0.03 eV/Å. The model of CoP2 was established according to literature [3] , where each Co is coordinated with six P atoms to form a slightly distorted octahedron. The V-CoP2 was modeled based on CoP2 and    From Figure S4, we can see that V-CoP2/CC with low V incorporation ratio of 2.5% shows the similar nanowires with V-CoP2/CC that the V incorporation ratio is 5%. When the V incorporation ratio further increases to 10%, the nanowires show a certain aggregation. The result indicates that the V incorporation ratio has a certain influence on the morphology of the formed nanowires. The phosphatizing products prepared by other hydrothermal times of 2 and 6h were also studied. As shown, by extending the hydrothermal time, the shape of the sample changed from individual nanowires with distinct roots to slightly bonded nanowires, demonstrating the significant effect of hydrothermal time on the catalyst structure. When the phosphrization temperature is 500 o C, the morphology of V-CoP2/CC-T500 sample is the irregular bundles of nanowire, while when the phosphating temperature reaches 700 o C, the morphology of V-CoP2/CC-T700 sample begins to gather and forms a flower-like morphology composed of radiated nanowires. Thus, we infer that too low or too high phosphrization temperature is not conducive to maintain the morphology of the nanowires.   In order to study the effect of the V incorporation ratio on the activity of V-CoP2/CC, we evaluated the HER activity of V-CoP2/CC with different V incorporation ratio (2.5%, 5% and 10%). The results show that V-CoP2/CC with V incorporation ratio of 5% has the best HER activity with the lower overpotential of 50 mV at 10 mA cm -2 and Tefel slope of 32 mV dec -1 compared with those of V-CoP2/CC-2.5% (112 mV@10 mA cm -2 and 77 mV dec -1 ) and V-CoP2/CC-10% (106 mV@10 mA cm -2 and 62 mV dec -1 ). Too low or too high incorporation ratio is not conducive to improving the HER performance of V-CoP2/CC. The effect of hydrothermal time on HER activity was studied. The V-CoP2/CC catalyst (obtained by hydrothermal time for 4h) shows the best HER electrocatalytic activity with the lower overpotential of 50 mV at 10 mA cm -2 and Tefel slope of 32 mV dec -1 relative to those of V-CoP2/CC-2 (92 mV@10 mA cm -2 and 47 mV dec -1 ) and V-CoP2/CC-6 (122 mV@10 mA cm -2 and 54 mV dec -1 ). The hydrothermal time is too short or too long, a high HER activity catalyst cannot be obtained. V-CoP2/CC with the phosphrization temperature of 600 o C also shows the better catalytic HER activity with the lower overpotential of 50 mV at 10 mA cm -2 and Tefel slope of 32 mV dec -1 relative to V-CoP2/CC-T500 (117 mV@ 10 mA cm -2 and 65 mV dec -1 ) and V-CoP2/CC-T500 (112 mV@ 10 mA cm -2 and 71 mV dec -1 ). It can be seen intuitively that too low or too high temperature is not conducive to the catalytic reaction. As shown in Figure S12, the V-CoP2/CC catalyst with a phosphrization time of 2h showed the best catalytic activity with the lower overpotential of 50 mV at 10 mA cm -2 and Tefel slope of 32 mV dec -1 in comparison with V-CoP2/CC-T1h (108 mV@10 mA cm -2 and 66 mV dec -1 ) and V-CoP2/CC-T3h (95 mV@10 mA cm -2 and 56 mV dec -1 ). It can be clearly seen that the appropriate phosphorous time is very important to improve the catalytic activity of HER.  The ECSA is the electrochemical active surface area, which can be calculated from the following formula: ECSA =specific capacitance/40 μF cm -2 per cm -2 where specific capacitance is Cdl, and 40 μF cm -2 is a constant to convert capacitance to ECSA. The specific capacitance can be converted into an ECSA using the specific capacitance value for a flat standard with 1 cm 2 of real surface area. As depicted, V-CoP2/CC still shows the better specific catalytic activity toward HER.  where I represent the current density for different samples during the LSV measurement in 0.5 M H2SO4, F is the Faraday constant (C/mol), and n is the number of the active sites (mol) for different samples. At an overpotential of 100 mV, the TOF of V-CoP2/CC is calculated to be 0.379 s -1 ( Figure S15b), which is superior to CoP2/CC (0.137 s -1 ). The high TOF of V-CoP2/CC is corresponded with its superb activity toward HER. The exchange current (j0) of the catalyst is calculated by extrapolating the Tafel diagram.
This finding indicates the enhanced intrinsic catalytic activity for HER by V incorporation. Further XRD and SEM analyses for V-CoP2/CC after HER test were performed. It is shown that the structure of V-CoP2/CC after the HER test did not change and V-CoP2/CC preserved its porous nanowires.  We also tested the OER performance of catalysts with different V incorporation ratio (2.5%, 5% and 10%). The results show that V-CoP2/CC with V incorporation ratio of 5% exhibits the best catalytic OER activity with the lower overpotential of 91 mV at 10 mA cm -2 and Tefel slope of 40 mV dec -1 compared with V-CoP2/CC-2.5% (160 mV@10 mA cm -2 and 81 mV dec -1 ) and V-CoP2/CC-10% (177 mV@10 mA cm -2 and 99 mV dec -1 ).      V-CoP2/CC also exhibits the higher TOF value of 0.207 s -1 at the overpotential of 240 mV than CoP2/CC (0.096 s -1 ) for OER. Figure S26. The i-t chronoamperometric curve for CoP2/CC during OER process for 6h.
As shown in Figure S26, the OER stability of CoP2 is worse. The performance during the period deteriorates and improves at the initial of 2h. The improved stability may be caused by the reconstruction of the Co3O4 layer. After 2h, the current density of CoP2 substantially decays to 0 after 6h. It is not difficult to see that the stability of CoP2 cannot be maintained for a long time. Figure S27. XRD pattern of V-CoP2/CC for OER after 5000 cycles.
As shown, the XRD pattern of the sample after OER test matches well with CoP2(PDF#77-0263) and Co3O4(PDF#43-1003). It proves that a Co3O4 layer was formed on the surface of the sample during the OER test. It can be clearly seen from the Figure S28a that the CoP2/CC burned with air in the muffle furnace is completely converted into Co3O4/CC, and there is no major change in the morphology ( Figure S28b). As shown in Figure S28c, Co3O4/CC shows the OER catalytic activity of 359, 494 and 537 mV at 10 mA cm -2 , 50 mA cm -2 and 100 mA cm -2 , and the attenuation is not obvious in the 20h stability test ( Figure S28d). As depicted, V-CoP2/CC after OER possesses a smaller charge transfer resistance (Rct) of 6 Ω than Co3O4/CC (~16 Ω), manifesting the better electronic conductivity. For Co3O4/CC, although the activity of Co3O4 is not satisfactory caused by the poor conductivity, it can enhance corrosion resistance in acidic OER validated in our manuscript. As for V-CoP2/CC after OER, the inner V-CoP2 provides a superior conductive platform to remedy the insufficient conductivity of the generated Co3O4. Therefore, the excellent activity and exceptional stability towards OER of V-CoP2/CC give concurrently credit to the synergism of inner V-CoP2 and outer Co3O4.  The adsorption free energy of H (ΔGH*) on catalyst can serve as a parameter to evaluate the HER activity. The large and negative values of ΔGH* indicates the strong chemical adsorption (difficult to be desorbed), while the large and positive ΔGH* represents the weak H* adsorption (difficult to be adsorbed). More positive or more negative are not favorable for HER. A catalyst that gives absolute values of ΔGH* ≈ 0 is considered a good candidate for HER, whatever it is positive or negative. To further realize the active centers, therefore, we calculate the free energy of H* (ΔGH*) on different systems. In the calculation, the absolute values of ΔGH* on different catalyst were compared.
The adsorption free energy of H* (GH*) on different surfaces are calculated as:

ΔGH* = ΔE H* + ΔZPE -TΔS
Where EH* is the adsorption energy of H species. E, ZPE and S are the energy change in zero point energy and entropy, respectively.
V-CoP2 holds a much smaller |ΔGH*| than CoP2 (Supplementary Table 5), indicating that V-CoP2 is highly active for HER. Moreover, the smallest |ΔGH*| at P site (0.159) in V-CoP2, as compared with that of Co site (0.436) and V site (0.510) indicates that P site is the active site for HER. This can be ascribed to the high electron density at P site induced by V incorporation, as verified by XPS analysis, which is conducive to adsorbing more hydrogen protons.     OER activity and stability. However, we found that the thermodynamic overpotential of V-CoP2-Co3O4 is 170 mV from the scaling relation, which is large the experimental results (75 mV). It is reported that the presence of defects or oxygen vacancy could tune the adsorption/desorption behaviors of oxygen intermediates to overcome the limitation of RDS, reducing the overpotential ceiling. Thus we also consider the model of Co3O4 with oxygen vacancy on the surface of V-CoP2 (labeled as V-CoP2-Co3O4-x) ( Figure S34, S36). As displayed, the thermodynamic overpotential of V-CoP2-Co3O4-x is decreased to 90 mV, much closer to the experimental result. But there is still small existed discrepancy between them, which may be caused by other inevitable factors, such as voltage, interface impedance, pH etc. [1,2]         The result of Faradic efficiency was obtained by taking the average of three sets of parallel tests. The amounts of H2 and O2 gases practically collected were listed in Table S7. In the two-electrode cell, the hydrogen generated was collected via a simple drainage method, and the gas diffusion was prevented by using the Nafion membrane. The gas collection is carried out -at 100 mA cm -2 for 60 minutes. The theoretical H2 and O2 amounts was calculated as follows: ngas = Q/zF where ngas, z and Q represent the number of moles of gas produced, the number of electron transfer (for H2 is 2 and O2 is 4), and the charge passed through the electrodes, respectively. F is the Faraday constant (96485 C mol -1 ).