Electrosynthesis of high-entropy metallic glass nanoparticles for designer, multi-functional electrocatalysis

Creative approaches to the design of catalytic nanomaterials are necessary in achieving environmentally sustainable energy sources. Integrating dissimilar metals into a single nanoparticle (NP) offers a unique avenue for customizing catalytic activity and maximizing surface area. Alloys containing five or more equimolar components with a disordered, amorphous microstructure, referred to as High-Entropy Metallic Glasses (HEMGs), provide tunable catalytic performance based on the individual properties of incorporated metals. Here, we present a generalized strategy to electrosynthesize HEMG-NPs with up to eight equimolar components by confining multiple metal salt precursors to water nanodroplets emulsified in dichloroethane. Upon collision with an electrode, alloy NPs are electrodeposited into a disordered microstructure, where dissimilar metal atoms are proximally arranged. We also demonstrate precise control over metal stoichiometry by tuning the concentration of metal salt dissolved in the nanodroplet. The application of HEMG-NPs to energy conversion is highlighted with electrocatalytic water splitting on CoFeLaNiPt HEMG-NPs.

scan for a high coverage CoFeLaMnNi is presented demonstrating characteristic carbon and oxygen regions in addition to distinct XPS peaks corresponding to individual metallic species. (b) Fe 2p XPS region demonstrates mixed FeO and Fe2O3 species at 2p3/2 binding energies of 710.7 eV (∆E = 13.5 eV) and 716.0 eV (∆E = 13.6 eV) respectively. (c) Co 2p XPS region with characteristic mixed CoO and Co3O4 XPS peaks at a binding energy of 780.2 eV (∆E = 15.4 eV) with a characteristic satellite peak at a binding energy of 787.7 eV (∆E = 15.5 eV) and a distinct CoO satellite peak at a binding energy of 782.0 eV (∆E = 15.5 eV). (d) Ni 2p XPS region with the presence of a peak corresponding to NiO at a binding energy of 855.3 eV (∆E = 17.3 eV) and a satellite peak at binding energy 861.8 eV (∆E = 17.1 eV). (e) La 3d XPS region with two peaks at binding energies of 834.2 eV (∆E = 16.8 eV) and 837.8 eV (∆E = 16.8 eV) potentially corresponding to convoluted La2O3 and La(OH)3 XPS signatures. (f) Mn 2p XPS region with the presence of a mixed MnO, MnO2, and Mn2O3 peak at a binding energy of 641.8 eV (∆E = 11.1 eV) with a satellite XPS peak at a binding energy of 645.1 eV (∆E = 11.2 eV). \ Supplementary Figure 4 | XRD of as-deposited HEMG-NPs. X-ray diffraction spectra of CoFeLaMnNi HEMG-NP high coverage sample on amorphous glassy carbon substrate electrode in a 2-Theta grazing angle orientation indicating characteristic glassy carbon broad amorphous peaks and no significant sharp peaks corresponding to metallic crystal facets or crystalline alloys.
Cyclic voltammetry traces for 5mM of each individual metal salt precursor reduction on HOPG with an overlay of the background (red) showed the oxidation and reduction potentials of each individual metal and showed that activity at the electrode observed was from the metal deposition. Scan rate was 50mV/s with a HOPG (WE), Ag/AgCl (RE) and Pt-wire (CE).

Supplementary Figure 5 | Cyclic voltammetry traces of individual metal chlorides.
CV scans of metal chloride precursor on a 3 mm diameter HOPG electrode for electrodeposition with 250 mM KCl and 137 mM phosphate buffered saline (1X PBS) with a glassy carbon rod counter electrode and a double-junction Ag/AgCl reference electrode. Cathodic scans from 0.5 V vs. Ag/AgCl reveal nucleation behavior at varying onset potentials for (a) CoCl2, (b) CrCl3, (c) GdCl3, (d) InCl3, e) MnCl2, (f) NiCl2, (g) SnCl2, (h) VCl3, (i) ZnCl2. Under these experimental conditions, metals contained in droplets may initially electrodeposit on the carbon substrate, and eventually a newly formed metal phase, making a kinetic evaluation of electrodeposition difficult. This evaluation is difficult since the nucleation and growth of metals is an innersphere process, and the reduction of each metal species on corresponding dissimilar metal electrodes has not been studied in detail. To further this complication, the potential at which single atoms can be oxidized is different from bulk, polycrystalline metal, implying that differences in heterogeneous kinetics may manifest from very small agglomerates of metal atoms. Finally, the stabilization of deposited metal atoms on the growing amorphous structure is not well understood. RHE, corresponding to an overpotential of 201 mV. The ECSA of the electrode was determined by EIS to be 0.0993 cm 2 assuming 20 μFꞏcm -2 as the specific capacitance. A Ag/AgCl reference electrode separated from the cell by a salt bridge and a glassy carbon rod acted as the reference and counter electrode, respectively. Scan rate was 10 mV/s. Figure 19 | HEMG-NP stability for OER/HER. (a) Stability of the CoFeLaNiPt HEMG-NP electrocatalyst over 3600 seconds at an overpotential of 377 mV (corresponding to an initial current density j = 10 mAꞏcm -2 ) for the OER at 25° C and under conditions of O2 saturation revealing minimal drift in current density. (b) Stability of CoFeLaNiPt HEMG-NP electrocatalyst over 3600 seconds at an overpotential of 557 mV (corresponding to an initial current density j = -10 mAꞏcm -2 ) for the HER at 25° C and under conditions of H2 saturation revealing minimal drift in current density.

Supplementary Figure 20 | Determination of metal salt leakage into organic phase by UV-vis.
Metal salt precursor leakage from the aqueous nanodroplet phase to the organic DCE phase was investigated by UV-vis spectroscopy. In brief, 10 µL of each individual metal salt precursor was mixed with 20 µL ultrapure water and suspended in a 5 mL DCE continuous phase with 0.1 M TBAP and an emulsion was prepared as outlined in the Materials and Methods section. Following sonication, the emulsion mixture was centrifuged in an Eppendorf Centrifuge 5804 R for 10 minutes at 4200 rpm to separate the aqueous and organic phases. A 2 mL aliquot of the DCE phase was extracted and absorbance spectra were collected with a JASCO V-650 Spectrophotometer in the 200-800 nm wavelength range. (a) UV-vis spectra of the DCE fraction containing CoCl2,CrCl3,GdCl3,InCl3,MnCl2,NiCl2,SnCl2,VCl3,and ZnCl2. (b) Theoretical maximum metal salt concentrations in DCE continuous phase following emulsion/centrifugation procedure calculated from Beer-Lambert Law (path length = 1 cm), suggesting minimal metal salt precursor leakage in the micromolar regime.

Supplementary Figure 21 | Evaluation of the Uncompensated
Resistance. EIS plot of Impendence and phase angle vs. frequency for Co NPs electrodeposited on a HOPG substrate allows the determination of uncompensated solution resistance (Ru). At high frequencies the double layer capacitance of the substrate, represented by a capacitor in the Randles cell, behaves as a short, signifying the observed impedance stems from the solution resistance. This model is verified by evaluating the phase angle, which approaches zero when the system consists of only resistive components. Therefore, this pre-programed evaluation included on the Pine instrumentation determines the maximum phase angle and evaluates the solution resistance at that point to give the Ru. Using this evaluation, the uncompensated resistance was determined to be 217 ± 29 Ω over 13 experiments.

Alloy
Co