Synergistic Effect of Co–Ni Hybrid Phosphide Nanocages for Ultrahigh Capacity Fast Energy Storage

Abstract Rational design of metal compounds in terms of the structure/morphology and chemical composition is essential to achieve desirable electrochemical performances for fast energy storage because of the synergistic effect between different elements and the structure effect. Here, an approach is presented to facilely fabricate mixed‐metal compounds including hydroxides, phosphides, sulfides, oxides, and selenides with well‐defined hollow nanocage structure using metal–organic framework nanocrystals as sacrificial precursors. Among the as‐synthesized samples, the porous nanocage structure, synergistic effect of mixed metals, and unique phosphide composition endow nickel cobalt bimetallic phosphide (NiCo‐P) nanocages with outstanding performance as a battery‐type Faradaic electrode material for fast energy storage, with ultrahigh specific capacity of 894 C g−1 at 1 A g−1 and excellent rate capability, surpassing most of the reported metal compounds. Control experiments and theoretical calculations based on density functional theory reveal that the synergistic effect between Ni and Co in NiCo‐P can greatly increase the OH− adsorption energy, while the hollow porous structure facilitates the fast mass/electron transport. The presented work not only provides a promising electrode material for fast energy storage, but also opens a new route toward structural and compositional design of electrode materials for energy storage and conversion.

The resultant ZIF-67 nanocrystals were separated by centrifugation and washed with ethanol for 3 times, followed by drying at 80 °C for 12 h.
Synthesis of ZIF-67 nanocrystals with size of ~100 nm: ZIF-67 nanocrystals with size of ~100 nm were synthesized for comparison. Typically, 1.176 g of Co(NO 3  without well-defined nanocage hollow structure was synthesized using ZIF-67 nanocrystals with size of ~100 nm as precursors. Typically, 50 mg of the as-prepared ZIF-67 nanocrystals with size of ~100 nm were dispersed in 10 mL of ethanol by sonication. Then 5 mL of ethanol containing 200 mg of Ni(NO 3 ) 2 ·6H 2 O was injected into the ZIF-67 suspension. The resultant mixture was transferred in a 20 mL Teflon-lined autoclave and was heated at 120 °C for 4 h. d-NiCo-LDH was then separated by centrifugation and washed with ethanol for 3 times, followed by drying at 80 °C for 12 h.
Synthesis of NiCo-P and d-NiCo-P: Typically, 10 mg of the as-prepared NiCo-LDH was placed in a small crucible, and 200 mg of NaH 2 PO 2 was placed in a bigger crucible. The small crucible containing NiCo-LDH was place in the bigger one containing NaH 2 PO 2 which was then transferred into a tube furnace and heated at 350 °C for 2 h under Ar atmosphere with a temperature ramping rate of 2 °C/min, and NiCo-P was obtained. Disordered NiCo-P (d-NiCo-P) without well-defined nanocage hollow structure was obtained following the same method, except that d-NiCo-LDH was used as precursor. suspension. The resultant mixture was transferred in a Teflon-lined autoclave and was heated at 120 °C for 4 h. Co-LDH or Ni-LDH was then separated by centrifugation and washed with ethanol for 3 times, followed by drying at 80 °C for 12 h. Co-P and Ni-P were obtained following the same method used for synthesis of NiCo-P, except that Co-LDH and Ni-LDH were used as precursor, respectively.
Characterization: XRD data was measured by Rigaku Corporation SmartLab 9 kW at 45 kV and 200 mA using Cu K radiation. SEM images were measured on a Hitachi S-4800 microscope. TEM images were measured using a transmission electron microscope (Hitachi, H-9000NAR). HRTEM were measured using a FEI Tecnai F30 microscope equipped with an energy-dispersive X-ray (EDX) detector. XPS was measured by X-ray photoelectron spectroscopy (XPS, Axis Ultra, Kratos Analytical, Japan) with monochromatic aluminum K as source of X-ray. The nitrogen adsorption/desorption isotherms of the materials were measured within the pressure range 0-1 atm at 77K using a Quadrasorb system from a Quantachrome Autosorb-IQ gas adsorption analyzer. Non-local density functional theory (NL-DFT) method was used to measure the pore size distribution.
Electrochemical measurements: Except that EIS measurements were conducted in a Zahner Zennium electrochemical workstation, all the other electrochemical tests were carried out in a CHI 760E electrochemical workstation. For three-electrode configuration tests, an Ag/AgCl (saturated KCl) electrode was used as reference electrode, a Pt foil electrode was used as counter electrode, and 2 M KOH was used as electrolyte. The working electrode was prepared using the following method: a mixture slurry containing of 80 wt% active material, 10 wt% Super P, and 10 wt% PTFE binder was made, which was then rolled with the assistance of ethanol to make a piece of uniform film with a typical areal mass of ~2.5 mg cm -2 . The film was dried at 60 ˚C for 12 h under vacuum, and was then pressed between two nickel foam where q is the charge, m represents the mass of the active material, and is the integral area of the CV curves.
The energy density E (Wh/kg) and power density P (W/kg) in Ragone plot were calculated with the following equations, Where is the integral current area, =I/m is the current density, where I is the current and m is the mass of active materials, and t is the discharge time (s).

Theoretical calculation method:
All calculations were performed based on DFT as implemented in Vienna Ab-initio Simulation Package (VASP). [1][2] Electron wave functions were expanded by using the projector augmented wave (PAW) method with a kinetic energy cutoff of 400 eV. The Perdew-Burke-Ernzerhof (PBE) functional [3] for the generalized gradient approximation (GGA) was used to treat the electron exchange-correlation interaction.
For bulk NiCo-P and 2D NiCo-LDH, Monkhorst-Pack sampling [4] of 6x6x10 and 10x10x1 were used and all atoms are fully relaxed, We have taken into account the Coulomb correction for Co 3d and Ni 3d electrons by using LSDA+U method with the Hubbard U values of 5.1 and 6.4 for Co and Ni, respectively. which have been used in some similar systems [5] To calculate the adsorption energy of OHon Ni 2 P, CoP, and NiCoP surfaces, we used supercell approach where Ni 2 P, CoP, and NiCoP were modeled by using (2x2) six-layer slab supercells having the (111)  Monkhorst-Pack grid, respectively, for the structure optimization and for the total energy calculations were used. In all the calculations, self-consistency was achieved by allowing the total energy to converge with 0.0001 eV. The Hellman-Feynman force components on each ion in the slab supercells are converged to 0.01 eV/Å.
The adsorption energy of OHwas calculated using the formula below: where , , and are the total energies of the slab with OHadsorbed on the Ni site, the slab supercell, and an isolated OH -, respectively. Meanwhile, Dipole corrections are also considered in those calculations of adsorption energy. [6] Physical and electrochemical characterizations Figure S1.    where q is the charge, m represents the mass of the active material, and is the integral area of the CV curves.
To achieve charge balance, , therefore, Specifically, the mass ratio of negative to positive materials (m -:m + ) was calculated to be ~2.6. For CV curves measured within 0-1 V, 0-1.2 V, and 0-1.4V, the voltage window did not fully cover the potentials of the redox reaction while for an extended voltage window (0-1.8 V), a rapidly increasing current density due to the oxygen evolution reaction (OER) was observed. Therefore, voltage window of 0-1.6 V was selected for the electrochemical measurements of NiCo-P//PANI/rGO. Besides CV measurements,