MICROELECTRODE AND METHOD FOR PRODUCING SAME

Abstract

The invention disclosed herein generally contemplates novel microelectrodes and methods of preparing same.

Claims

1-43. (canceled)

44. An electrically conductive capsule comprising an electrically conducting network core comprising a plurality of fibers exhibiting open hierarchical porosity, the capsule having a porosity enabling flow of liquid and gases therethrough.

45. The capsule according to claim 44, the capsule being an electrically conductive porous capsule encapsulating an electrically conductive porous network of conducting fibers, each having an open hierarchical porosity, wherein the conductive porous capsule conducts electricity to the fibers contained therein and permits liquid and gaseous communication therethrough.

46. An electrically conductive porous capsule encapsulating an electrically conductive porous network of conducting fibers, each having an open hierarchical porosity, wherein the conductive porous capsule conducts electricity to the fibers contained therein and permits liquid and gaseous communication therethrough.

47. The capsule according to claim 44, wherein the conducting fibers are composed of a high surface-area conductive material having open hierarchical porosity.

48. The capsule according to claim 44, wherein the conductive fibers are composed of a conductive material being a metallic material, a carbonaceous material or a polymeric material.

49. The capsule according to claim 48, wherein the conductive fibers comprise a metallic material, optionally selected from nickel, zinc, aluminum and magnesium.

50. The capsule according to claim 49, wherein the metallic material is or comprises nickel.

51. The capsule according to claim 49, wherein the metallic material is or comprises Ni(OH).sub.2 or Ni/Ni(OH).sub.2.

52. The capsule according to claim 51, wherein the conductive fibers are core-shell structures comprising each a Ni core and a Ni(OH).sub.2 shell.

53. The capsule according to claim 49, wherein the conductive fibers comprise nickel, optionally in combination with at least one other metal.

54. The capsule according to claim 44, wherein the conductive fibers are manufactured by a process comprising treating a composite of a conductive material and a polymeric binder under conditions sufficient to remove the polymeric binder to obtain the conductive fibers.

55. The capsule according to claim 54, wherein the conductive material is in a form of electrospun conductive fibers.

56. The capsule according to claim 54, wherein the process comprises treating a composite of a conductive material and at least one polymeric binder under conditions sufficient to remove the polymeric binder and obtain conductive fibers having open hierarchical porosity.

57. The capsule according to claim 54, wherein the process comprises treating conductive fibers with at least one polymeric binder to form a composite of the conductive fibers and the at least one polymer.

58. The capsule according to claim 54, wherein the conductive material is a metal.

59. The capsule according to claim 54, wherein the conductive material is a metallic conductive material provided in a form of a mat or foam of fibers, and wherein process comprises: forming a metal fiber mat or foam; grinding the foam to obtain fragments of said mat or foam; treating said fragments with a polymeric binder to obtain a metal-polymer composite; treating the composite under conditions sufficient to remove the polymeric binder and obtain a plurality of conductive fibers having each open hierarchical porosity.

60. The capsule according to claim 54, wherein the conductive fibers are prepared by a process comprising: electrospinning a solution comprising at least one nickel precursor and optionally at least one another metal precursor to provide a fiber mat or foam comprising the at least one nickel precursor and optionally the at least one another metal precursor; transforming the nickel precursor in said mat or foam to nickel oxide (NiO) to provide a NiO mat or foam; grinding the NiO mat or foam into a powder to obtain fragments of the NiO fibers and mixing with a polymer binder to form a composite; transforming the NiO in said composite into nickel metal (NiO) under conditions permitting removal of the polymeric binder to provide porous nickel conductive fibers; and transforming at least some of the nickel metal (NiO) in the fibers to Ni(OH).sub.2 to thereby afford the Ni/Ni(OH).sub.2 fibers.

61. A microelectrode in a form of a capsule according to claim 44.

62. A system for generation of hydrogen and/or oxygen gas, the system comprising a plurality of microelectrodes according to claim 61 dispersed in a medium.

63. The system according to claim 62, comprising an electrochemical device configured for generating the gases.

64. The system according to claim 63, wherein the electrochemical device is an electrochemical thermally activated chemical cell (E-TAC).

65. An electrochemical device for generation of a gas by utilizing microelectrodes according to claim 61, the device being adapted and operable to convert the microelectrodes from an oxidized form to a reduced form.

66. A medium of an electrochemical cell comprising a plurality of capsules according to claim 44.

67. An anode electrode in a form of a microelectrode according to claim 61.

68. A device comprising a microelectrode according to claim 61, the device being selected from electric cells, electric furnaces, thermionic tubes, gas-discharge devices, semiconductor devices and electrochemical cells.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to better understand the subject matter that is disclosed herein and to exemplify how it may be carried out in practice, embodiments will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which:

[0084] FIG. 1 provides a schematic illustration of the E-TAC water splitting cycle. The electrochemical hydrogen generation step (Step 1, left) is carried out in a cold (25° C.) alkaline solution. Then, the chemical oxygen generation step (Step II, right) is carried out in a hot (95° C.) alkaline solution at open circuit.

[0085] FIG. 2 provides a schematic depiction of an exemplary process for manufacturing a nickel electrode material according to the invention.

[0086] FIGS. 3A-C provide pelletized nickel hydroxide electrodes. (FIG. 3A) the conventional nickel foam substrate (1) is broken down into small discrete units (pellets, 2). (FIG. 3B) zoom-in on the pellet, showing a porous internal conductive matrix (3), encapsulated by a conductive perforated protective metal sheet (4). (FIG. 3C) zoom-in on a single porous nickel nanofiber, a possible type of morphology for the inner matrix. The metallic fiber can be oxidized electrochemically in an alkaline solution, forming a core/shell Ni/Ni(OH).sub.2 structure.

[0087] FIGS. 4A-4G present the electrode synthesis process according to the invention. FIG. 4A) Electrospun NiAc-PVP “green” fiber sheet. FIG. 4B) Firing schedule of the mat. FIG. 4C) NiO fiber mats after calcination. FIG. 4D) NiO nanofiber coarse powder. FIG. 4E) NiO nanofiber powder and PPC paste. FIG. 4F) Metallic nickel porous nanofiber substrate. FIG. 4G) Electrodeposited Ni(OH).sub.2 electrode on a nickel foam substrate.

[0088] FIGS. 5A-C are HRSEM micrographs of NiAc-PVP “green” nanofibers under various magnifications. FIG. 5A) ×1,000, FIG. 5B) ×10,000, and FIG. 5C) ×50,000.

[0089] FIGS. 6A-B are HRSEM micrographs of nickel oxide nanofibers after calcinations. FIG. 6A) low-, and FIG. 6B) high-magnification.

[0090] FIGS. 7A-B are HRSEM micrographs of nickel oxide nanofibers after grinding and binder addition. FIG. 7A) low-, and FIG. 7B) high-magnification.

[0091] FIGS. 8A-D are HRSEM micrographs of the pressed and reduced nickel substrate and nanofibers. FIG. 8A) low-, FIG. 8B) and FIG. 8C) medium-, and FIG. 8D) high-magnification.

[0092] FIGS. 9A-B depict Ni(OH).sub.2 layer growth (FIG. 9A) and the corresponding change in electrode charge capacity (FIG. 9B) with layer formation cycles. The charge capacity was estimated using potentiostatic charge/discharge cycles at 1.48 V.sub.RHE/1.23 V.sub.RHE for charging/discharging durations of 100, 200, 400, 800, 1600, and 3200s in 5M KOH electrolyte at ambient temperature.

[0093] FIGS. 10A-B provide (FIG. 10A) HRSEM image of an undoped Ni/Ni(OH).sub.2 NFA after layer growth by E-TAC cycling. (FIG. 10B) diffractograms of an undoped Ni/Ni(OH).sub.2 NFA formed in situ by E-TAC cycles after different preparation stages: as-prepared NF nickel substrate (red line), after aging for 2 h in 5M KOH at 90° C. (green line), after 100 E-TAC cycles of 100s charge/100s regeneration (blue line), and after final characterization in a full E-TAC test (black line).

[0094] FIG. 11 provides X-ray diffractograms of a Ni/Ni(OH).sub.2 nanofiber anodes. Diffractograms of a Ni/Ni(OH).sub.2 NFA formed by galvanostatic cycling after different preparation stages: as-prepared NF nickel substrate (red line); after aging for 2 h in 5M KOH at 90° C. (green line); after active layer growth by galvanostatic cycling at 55° C. (blue line); and after final characterization in a full E-TAC water splitting test (black line). The dashed vertical lines indicate the position and relative intensities of Bragg reflections of metallic nickel (JCPDS 00-004-0850, green), NiO (JCPDS 01-071-4750, red) and β-Ni(OH).sub.2 (JCPDS 00-059-0462, black).

[0095] FIGS. 12A-C are HRSEM images of undoped Ni/Ni(OH).sub.2 nanofiber anodes. (FIG. 12A) As-prepared Ni nanofiber substrate. (FIG. 12B) Ni/Ni(OH).sub.2 nanofiber anode after layer growth by galvanostatic cycling. (FIG. 12C) Ni/Ni(OH).sub.2 anode after E-TAC water splitting tests.

[0096] FIGS. 13A-D depicts elemental distribution in a cobalt-doped nanofiber anode. HRSEM micrograph (FIG. 13A) and the corresponding qualitative EDS mapping (FIGS. 13B-C) of a cobalt-doped Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 nanofiber anode prepared by galvanostatic cycling.

[0097] FIGS. 14A-B show surface area and pore size distribution of a Ni/Ni(OH).sub.2 nanofiber anode. (FIG. 14A) B.E.T N.sub.2-absorption isotherms before and after Ni(OH).sub.2 active layer growth on a nanofiber nickel substrate (red and blue curves, respectively). (FIG. 14B) B.J.H N.sub.2-desorption pore volume distribution before and after Ni(OH).sub.2 active layer growth by galvanostatic cycling. The inset shows the B.J.H plot before layer growth

[0098] FIG. 15 provides cyclic voltammograms of Ni/Ni(OH).sub.2 nanofiber anodes with and without cobalt. CVs of an undoped Ni/Ni(OH).sub.2 NFA (red) and a cobalt-doped Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 NFA (blue) measured at a scan rate of 0.05 mV/s. The red and blue circle data series represent the steady-state Tafel plots for the undoped and doped NFAs respectively. All tests were carried out in an aqueous solution of 5M KOH at ambient temperature. The inset shows a magnification of the steady-state Tafel area.

[0099] FIGS. 16A-D provide comparison of the regenerated charge and current density of an undoped Ni/Ni(OH).sub.2 nanofiber anode (red lines), a doped Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 nanofiber anode (blue lines), and a doped Ni.sub.0.9Co.sub.0.1(OH).sub.2 electrodeposited anode (green lines) in E-TAC water splitting tests. Charging was carried out by applying a constant potential of 1.48 V.sub.RHE (full lines) or 1.43 V.sub.RHE (dashed lines) in 5M KOH aqueous solution at ambient temperature for a set duration, and regeneration was carried out by dipping the charged anode in 5M KOH aqueous solution at 95° C. for the same duration. The comparison is shown on the basis of the anode's volume, geometric area, total weight and active layer weight for E-TAC water splitting tests with charging/regeneration durations of 100s (FIG. 16A), 200s (FIG. 16B), 400s (FIG. 16C) and 800s (FIG. 16D). In each plot, all values are normalized by the highest values in that category, which are listed in Table 2. The center and circumference of each square represent relative values of 0% and 100%, respectively, and the inner gridlines are spaced at 20% intervals

[0100] FIGS. 17A-D show the effect of temperatures parameters on NFE microstructure. HRSEM images of nickel nanofiber anodes prepared at an ES voltage of 14 kVs with various maximum sintering temperatures of 500° C. (FIG. 17A), 600° C. (FIG. 17B) and 700° C. (FIG. 17C), and at 500° C. with a higher ES voltage of 26 kVs (FIG. 17D).

DETAILED DESCRIPTION OF EMBODIMENTS

Experimental

[0101] Electrospun Ni and NiCo Substrate Synthesis

[0102] Nickel acetate tetrahydrate with and without cobalt acetate (Ni:Co=0.9:0.1) was dissolved (10.6 wt. %) in acetic acid and ethanol (6:4 by wt.). The solution was stirred and filtered (0.22 μm), resulting in a light-green clear solution.

[0103] Next, polyvinyl pyrrolidone was added (8.8 wt %) and stirred at ambient temperature to achieve a final viscosity of 500-650 cPs.

[0104] The solution was electrospun (21-gauge needle, 17 kV, 1.25 ml/h, 15 cm WD), resulting in green NiAc-PVP fiber mats (FIG. 4a). The green fiber mats were then thermally treated in air (FIG. 4b) and the Ni was oxidized to NiO. The NiO mats (FIG. 4c) were ground to a powder (FIG. 4d), and mixed with a solution of Poly(propylene carbonate) 10.5 wt. % to form a paste (FIG. 4e). The powder-binder composite paste was then pressed into a mold at˜3000 kgf to form a pellet having a diameter of ˜1.1 cm (FIG. 4F). Each pellet was heated in a tubular furnace under reducing atmosphere (5 vol. % H.sub.2/N.sub.2, 45 cc/min), to decompose the polymer and reduce the NiO to metallic Ni, and to sinter the fibers.

[0105] SEM images of the green fibers, the calcined fibers, the ground fibers and the final fibers after reduction are presented in FIGS. 5A-C, FIGS. 6A-B, FIGS. 7A-B, and FIGS. 8A-D, respectively. Qualitative mapping and quantitative elemental analysis was carried out by EDS (Oxford SDD EDS), and are summarized for the doped and undoped electrodes in Tables 1 and 2, respectively. The total porosity of the as-prepared Ni substrates was calculated as 74-75%.

TABLE-US-00001 TABLE 1 Compositional EDS analysis of cobalt-doped nickel nanofibers after calcination and reduction. Sample Composition (wt. %) Composition (at. %) Cobalt-doped Ni fibers Ni: 87.46% Ni: 83.21% (reduced) Co: 10.6% Co: 10.05% O: 1.93% O: 6.74% Cobalt-doped Ni fibers Ni: 89.31% Ni: 86.23% (reduced) Co: 9.33% Co: 8.98% O: 1.35% O: 4.80%

TABLE-US-00002 TABLE 2 Compositional EDS analysis of nickel nanofibers after calcination and after reduction. Sample Composition (wt. %) Composition (at. %) Calcinated NiO fibers Ni: 78.49% Ni: 49.86% O: 21.51% O: 50.14% Reduced Ni fibers Ni: 98.67% Ni: 95.29% O: 1.33% O: 4.71%

[0106] Ni(OH).sub.2 Layer Growth by Oxidation/Reduction Cycles

[0107] In order to grow the active Ni(OH).sub.2, the Ni pellet was connected as the working electrode (WE) to a potentiostat in in 5M KOH aqueous electrolyte solution (˜50° C.), and subjected to 60 chronopotentiometric oxidation-reduction cycles at ±150 mA/cm.sup.2 up to ±200 C/cm.sup.2 or up to 1.55 V (during oxidation) and 1 V.sub.RHE (during reduction).

[0108] Every few cycles, the electrode was dried and weighed to calculate the percent of nickel atoms oxidized to Ni(OH).sub.2 (eq. 1):

[00001]$\begin{array}{cc}\%\mathrm{Ni}\mathrm{oxidized}\mathrm{to}\mathrm{Ni}{\left(\mathrm{OH}\right)}_2=\frac{\frac{m_i}{M{w}_{\mathrm{Ni}}}-\frac{\Delta m}{{\mathrm{Mw}}_{{\left(OH\right)}_2}}}{\frac{m_i}{M{w}_{\mathrm{Ni}}}}& \left(1\right)\end{array}$

[0109] where m.sub.i is the initial pellet mass, m.sub.f is the final pellet mass, and Δm is the weight gain.

[0110] To characterize the added charge capacity, the pellet was also subjected to chronoamperometric redox cycles. FIG. 9B shows the change in the electrode's charge capacity with layer growth cycles. By the end of the process, the electrode's weight was increased by 18% relative to the initial substrate weight, corresponding to a conversion of 30% of the Ni atoms to Ni(OH).sub.2.

[0111] In-Situ Layer Growth

[0112] The Ni(OH).sub.2 can also be grown in-situ during E-TAC water electrolysis without preliminary layer growth. An as-prepared Ni pellet was connected as the WE to a potentiostat in a three electrode configuration (Hg/HgO reference electrode) in 5M KOH aqueous electrolyte at ambient temperature. In this setup, the pellet was subjected to repeated E-TAC cycles with charging at 1.48 V.sub.RHE for 100s and regeneration in a hot (95° C.) 5M KOH solution for 100s. X-ray diffractograms and SEM images of the resulting Ni(OH).sub.2 layer are presented in FIGS. 10A-B.

[0113] Electrochemical Characterization and E-TAC Water Splitting Tests

[0114] Electrochemical tests were carried out in a three-electrode system, with the either the nanofiber anode (NFA) or an electrochemically deposited anode (EDA) as the working electrode, with a platinum foil (1 cm.sup.2) counter electrode, and an Hg/HgO/1M NaOH reference electrode in 5M KOH. Electrodeposited anodes were prepared as described elsewhere.

[0115] E-TAC water splitting tests were carried out by first charging the pellet anode at 1.48/1.43 V.sub.RHE in 5M KOH solution at ambient temperature (for 100, 200, 400 or 800s), and then regenerating the pellet anode by dipping it in a hot (95° C.) 5M KOH solution for the same duration. Stability tests were carried out with 100s charge/regeneration cycles. In addition, the anode was also subjected to E-TAC cycles with short regeneration times, wherein the charging duration varied between 100 and 800s, but the regeneration duration was kept constant at 100s.

Results and Discussion

[0116] Phase and Morphology Characterization

[0117] FIG. 11 shows X-ray diffractograms of the NFA in different stages of preparation displaying Bragg reflections of metallic nickel for the as prepared pellet, and Bragg reflections of β-Ni(OH).sub.2 after layer growth and E-TAC tests.

[0118] FIG. 12 presents high-resolution SEM micrographs of the Ni/Ni(OH).sub.2 nanofibers before and after active layer growth. The as-prepared pellet comprised of short (up to ˜1 μm) highly porous nanofibers with an average diameter of ˜250 nm (overall porosity=˜75%). After layer growth the pellet comprised of fibers with a highly porous “coral” structure comprising of thin sheets (20-50 nm). In the cobalt-doped anodes, cobalt was distributed homogenously along the fibers, as shown in FIG. 13.

[0119] FIG. 13 shows EDS maps of the as-prepared NiCo fibers (10 wt. % cobalt, Table 2), displaying a homogeneous distribution of cobalt within the nickel fibers.

[0120] The surface area of the pellets was measured using BET (FIG. 14). The BET surface area of the as-prepared pellet was 4.5 m.sup.2/g (FIG. 14A), two orders of magnitude higher than the surface area of the nickel foam used for the electrochemically deposited anodes (˜0.01 m.sup.2/g). The surface area increased to 22.9 m.sup.2/g following layer growth. A bimodal mesopore size distribution was observed as shown in FIG. 14B, with two peaks at pore diameters of approximately 3 and 30 nm before layer growth, and 3 and 16 nm after layer growth. Additionally, the total pore volume of small pores (with diameter <100 nm) increased by a factor of 6.5 after Ni(OH).sub.2 layer growth. This can be attributed to the growth of the porous Ni(OH).sub.2 layer, as shown in FIG. 12. Overall, a hierarchical meso/macro porous structure with a high surface area was obtained.

[0121] Electrochemical Characterization of Ni/Ni(OH).sub.2 and Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 Nanofiber Anodes

[0122] Core-shell electrospun nickel hydroxide electrodes were prepared to serve as redox anodes for decoupled E-TAC water splitting (FIG. 1). To characterize their electrochemical properties, the anodes were subjected to cyclic voltammetry (CV) scans. The CV scan of a Ni/Ni(OH).sub.2 NFA (FIG. 15) displayed a single redox wave centered around 1.34 V.sub.RHE, which provides a good approximation for the anode's reversible redox potential)(E.sup.0). The reversible potential is an important electrochemical property of Ni(OH).sub.2 anodes for E-TAC water splitting, as the cell voltage during the first (hydrogen generation) step is directly related to the Ni(OH).sub.2 oxidation potential according to: V.sub.cell=E.sub.anode−E.sub.cathode+ΣiR=(E.sup.0.sub.anode rxn+η.sub.anode)−(E.sup.0.sub.cathode rxn−η.sub.cathode)+ΣiR (eq. 1); where V.sub.cell is the cell voltage, E is the applied potential, E.sup.0 is the reversible potential, η=E−E.sup.0 is the overpotential and ΣiR is the sum of all the series resistance Ohmic losses. Thus, a low anodic redox potential alongside a minimal anodic overpotential contribute to a high voltage efficiency (1.48/V.sub.cell).

[0123] Additionally, since parasitic oxygen evolution may take place upon charging the anode to a high state of charge (SOC), a lower redox potential shifts the anode charging reaction away from the OER and enables charging the anode to higher SOCs without parasitic oxygen evolution. To evaluate the OER rate at the NFAs, the anodes were subjected to chronoamperometric measurements between 1.43 and 1.48 V.sub.RHE By allowing the current to stabilize, a distinction could be made between the pseudo-capacitive Ni(OH).sub.2 oxidation current and the steady-state OER Faradaic current. The corresponding steady-state Tafel plot for the OER Faradaic current at the Ni/Ni(OH).sub.2 NFA is plotted in FIG. 15. As can be seen, the onset potential for the OER at the Ni/Ni(OH).sub.2 NFA (taken at j.sub.OER=1 mA/cm.sup.2 from the exponential fit) was 1.47 V.sub.RHE, about 130 mV above the reversible redox potential of the Ni(OH).sub.2/NiOOH couple (1.34 V.sub.RHE), and 30 mV above the peak of its oxidation wave (at a potential of 1.44 V.sub.RHE) Thus, this anode can be oxidized at ambient conditions without concurrent oxygen evolution.

[0124] The reversible redox potential, as well as the charging and discharging overpotentials, are also affected by incorporation of various additives and dopants into the Ni(OH).sub.2 anode. Specifically, cobalt is known to shift the Ni(OH).sub.2 redox potentials cathodically and improve the electron and proton conductivity of Ni(OH).sub.2, allowing the anode to reach greater SOCs. Here, cobalt was embedded into the electrospun fibers, and evenly dispersed along the fibers, as shown in FIG. 13. The addition of cobalt resulted in a cathodic shift of the Ni(OH).sub.2/NiOOH reversible redox potential by 53 mV (FIG. 15). However, it also catalyzed the OER compared to the undoped anode, as shown by the steady-state Tafel plot in FIG. 15. The OER onset potential at the cobalt doped NFA was around 1.46 V.sub.RHE, lower than for the undoped anode. Error! Reference source not found. summarizes the potential values derived from the CV scans and steady-state polarization tests of the doped and undoped anodes, including: Ni(OH).sub.2/NiOOH redox, oxidation and reduction potentials (E.sup.0, E.sub.ox and E.sub.red, respectively), the OER onset potential (E.sub.onset) and the potential difference between E.sub.ox and E.sub.onset.

TABLE-US-00003 TABLE 3 Summary of key electrochemical parameters. Electrochemical parameters extracted from the cyclic voltammograms and steady-state polarization tests of Ni/Ni(OH).sub.2 and Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 NF. E.sub.ox E.sub.red E.sup.0 E.sub.onset E.sub.onset − [V.sub.RHE] [V.sub.RHE] [V.sub.RHE] [V.sub.RHE] E.sub.ox [mV] Ni/Ni(OH).sub.2 1.45 1.24 1.345 1.5 50 NFA Ni/Ni.sub.0.9Co.sub.0.1(OH).sub.2 1.39 1.2 1.295 1.46 70 NFA

[0125] Performance of Cobalt-Doped and Undoped NF Anodes in E-TAC Water Splitting

[0126] There are several key performance metrics in the selection and optimization of Ni(OH).sub.2 anodes for E-TAC water splitting. As discussed in the previous section, the first critical metric is the Ni(OH).sub.2 redox potential, which directly influences the cell voltage (see eq. 1) and therefore the process' voltage efficiency. Here, the anodes' charging potential was controlled by applying a constant polarization potential of either 1.48 or 1.43 V.sub.RHE during the hydrogen generation step. The highest value, 1.48 V.sub.RHE, was chosen since it is the thermoneutral water splitting voltage, i.e., this is the anode potential required in order to achieve a voltage efficiency of 100%.sub.HHV during the hydrogen generation step, neglecting all other losses (see eq. 1).

[0127] The second key metric is the anode's charge capacity, as anodes with higher charge capacities can sustain longer cycles, reducing the cycling frequency. In this respect, Ni(OH).sub.2 anodes for E-TAC water splitting are similar to battery electrodes, which also benefit from high charge capacities that extend their operation duration between recharges. Thus, similarly to battery electrodes, the Ni(OH).sub.2 anodes are expected to exhibit high energy densities normalized by the anode's volume and mass. To achieve this, a sufficiently high mass loading is necessary, along with a high utilization efficiency of the active mass. It should be noted that, unlike battery electrodes, the Ni(OH).sub.2 anodes in E-TAC water splitting are regenerated in a chemical process, rather than an electrochemical one (discharge), and therefore the rate of their regeneration cannot be directly controlled. As a result, the charge that can be chemically regenerated, Q.sub.regen, is often lower than the charge that can be extracted by electrochemical discharge; in electrochemical discharging, a preset electrochemical driving force or reaction rate is imposed by an external power source, whereas in chemical discharging the driving force and reaction rate decrease progressively as the NiOOH anode is discharged. In addition, the chemical discharging (regeneration) reaction takes place at the electrode-electrolyte interface and is limited by the diffusion of OH.sup.− ions into the active material. Thus, while a high mass loading may improve the electrochemical charge capacity, it may not contribute to, and perhaps even reduce, the chemically-regenerated capacity, especially if it is achieved by increasing the bulk volume at the expense of surface area.

[0128] The third key metric is the current density during the hydrogen generation step, which is proportional to the hydrogen production rate. In this respect, Ni(OH).sub.2 anodes should support operation under high current densities normalized by the geometric area, similarly to alkaline electrolyzers, which typically operate at 200-400 mA/cm.sup.2. However, in contrast to electrolyzer anodes wherein the OER takes place at the surface of the otherwise inert electrocatalyst, the charging reaction in the E-TAC water splitting process involves an electrochemical transformation of the anode itself from Ni(OH).sub.2 to NiOOH. Since the Ni(OH).sub.2 mass loading must be high enough to sustain high charge capacities, as discussed above, there is a trade-off between rate capabilities and charge capacity, similar to the trade-off observed in electrodes for supercapacitors.

[0129] In each test, the average regenerated charge was calculated based on the current×time product, according to: Q.sub.regen=∫i(t).Math.dt=<i>.Math.t, (eq. 2); where t is the time, i(t) is the current as a function of time, and <i> is the average current over the entire test duration. Accordingly, for the tests carried out at 100, 200, 400 and 800s, the ratio between the average current (in A) and the regenerated charge (in A.Math.s) is 100, 200, 400 and 800, respectively. FIG. 16 shows the regenerated charge capacities and current densities of the undoped and cobalt-doped NFAs compared to a cobalt-doped EDA. The values are normalized by all the key metrics, namely, the anode's volume, geometric area, total weight and active layer weight. The results are shown for tests carried out with equal durations of charging and regeneration of 100s (FIG. 16A), 200s (FIG. 16B), 400s (FIG. 16C) and 800s (FIG. 16D). The maximum values in each category are summarized in Table 2. Cobalt doping of the NFA resulted in enhanced performance in all metrics, with a higher charge and current density even at the low charging potential of 1.43 V.sub.RHE compared to both the undoped NFA and the cobalt-doped EDA.

[0130] At a charging potential of 1.48 V.sub.RHE (full line data series in FIG. 16), the regenerated charge capacity was higher for both the doped and undoped NFAs compared to the EDA in all categories, with the exception of a higher gravimetric charge capacity per unit mass of the active layer for the shortest test duration. However, as the test duration increased, the gravimetric charge capacity of the EDA fell short of the NFAs, suggesting a mass and/or charge transport limitation into the depth of the electrodeposited Ni(OH).sub.2 active layer. An opposite trend was observed for the NFAs, wherein the regenerated charge increased at longer test durations. This could be attributed to the high porosity of the Ni(OH).sub.2 layer, which enables facile mass transport and therefore enhances the rate of the diffusion-limited regeneration reaction.

[0131] Both the cobalt-doped and undoped NFAs displayed similar regenerated charge capacities at a charging potential of 1.48 V.sub.RHE It therefore appears that under these experimental conditions, the benefits of cobalt doping were lost. Furthermore, the extent of the OER is greater at this potential for the cobalt-doped NFA compared to the undoped NFA. Indeed, 1.48 V.sub.RHE lies anodicaly of the oxidation waves of both samples, and above the OER onset potential of the cobalt-doped NFA (FIG. 15). However, we hypothesize that the lower oxidation potential of the cobalt-doped NFA compared to the undoped NFA could enable extraction of a higher charge at lower potentials, increasing the voltage efficiency without compromising the Faradaic efficiency of the charging reaction. This was demonstrated by repeating the same tests for both samples at a charging potential of 1.43 V.sub.RHE.

[0132] As expected, the charge that could be extracted from both NFAs at 1.43 V.sub.RHE was lower than at 1.48 V.sub.RHE. Nevertheless, while the undoped NFA's retained only 34-49% (depending on the test duration) of its charge compared to that at 1.48 V.sub.RHE, the cobalt-doped NFA retained 54-57% of its charge under the same conditions. Moreover, the regenerated charge for the cobalt-doped NFA, even at the low charging potential of 1.43 V.sub.RHE, surpassed that of the EDA when charged at 1.48 V.sub.RHE (FIG. 16A-D, green lines). This is a significant improvement in anode performance, as the cobalt-doped NFA enables more charge to pass during the hydrogen production step, i.e., increased hydrogen production, at lower potentials that give rise to a higher voltage efficiency. In addition, as opposed to 1.48 V.sub.RHE, charging at 1.43 V.sub.RHE improves the charge acceptance and lowers the risk of parasitic oxygen evolution during anode charging, as the OER reaction rate at this anode is negligible at 1.43 V.sub.RHE (FIG. 6). This was also demonstrated by dissolved oxygen measurements, as discussed below.

[0133] The high surface area of the hierarchical meso/macro porous structure of the NFAs resulted in a significant increase in the obtainable current density at each potential. Compared to the EDA, the average current densities of the undoped and cobalt-doped NFAs at 1.48 V.sub.RHE were higher by up to 6.5 and 7.6 times, respectively. Furthermore, even at 1.43 V.sub.RHE, the current densities of the NFAs were higher compared to the EDA when charged at 1.48 V.sub.RHE (by up to 2.2 and 4 times for the undoped and doped NFAs, respectively). At both 1.48 and 1.43 V.sub.RHE charging potentials, the cobalt-doped NFA achieved higher current densities compared to its undoped counterpart, which is attributed to the larger driving force, i.e., a larger difference between the applied potential (1.48 V.sub.RHE) and the anode's redox potential.

[0134] The voltage efficiency of the E-TAC water splitting process using our cobalt-doped NFA in a two-electrode configuration was also compared to our previously reported anode. At a current density of 50 mA/cm.sup.2 (normalized by the geometric area), the low charging potential of the cobalt-doped NFA resulted in an overall lower cell voltage, with an increase in average voltage efficiency from 98.7% to 100%.

[0135] The fourth key metric in E-TAC water splitting is the Faradaic efficiency of the Ni(OH).sub.2 charging reaction, (1−Q.sub.OER/Q.sub.total, where Q.sub.OER is the parasitic charge of the OER during the hydrogen generation step, and Q.sub.total is the total charge passed during the hydrogen generation step). The Faradaic efficiency is influenced by the cycle duration, which is linked to the anode's SOC, as well as by the anode's charging potential and its relation to the OER onset potential at the same anode. Q.sub.OER was calculated at each E-TAC water splitting test from the measured amount of dissolved oxygen. At 1.48 V.sub.RHE charging potential, the Faradaic efficiency of the cobalt-doped NFA was high for the short tests (99%±1% for the 100s test and 98.5%±0.7% for the 200s test), but it decreased with increasing test duration, i.e., upon charging to higher SOCs, reaching 94%±3% for the 400s test, and only 85%±2% for the 800s test. However, by lowering the charging potential to 1.43 V.sub.RHE, the Faradaic efficiency for the cobalt-doped NFA increased, reaching 98%±2% at the longest (800s) test. This result was expected, as the oxygen evolution rate is negligible at this potential. In fact, the ratio between the steady-state OER rate (FIG. 15) and the overall reaction rate during charging can be used to estimate the lower limit of the Faradaic efficiency. Thus, the combination of increased surface area, porosity and cobalt doping significantly improves the anode's performance and allows operation at lower potentials with higher charge capacities and better rate capabilities. For example, a cobalt-doped NFA can withstand long cycles of 800s charging and regeneration, at a potential lower by 50 mV (1.43 V.sub.RHE) while supplying charge and current densities that are up to 4 times greater (depending on the normalization basis) compared to the cobalt-doped EDA, at a Faradaic efficiency of ˜100%.

[0136] In addition to the current density during the hydrogen generation step of the E-TAC water splitting cycle, it is advantageous for the oxygen generation step to be faster than the hydrogen generation step. This way the overall hydrogen production throughput is maximized. The fifth key metric is therefore the regeneration reaction rate, or the charge that can be regenerated at a given regeneration time, which is correlated to the hydrogen production throughput. To examine the influence of the anode preparation method on the regeneration rate, the E-TAC water splitting tests were repeated with shorter regeneration steps of only 100s. For the EDA, decreasing the regeneration time resulted in a capacity drop of up to 31±14%. However, for the undoped and cobalt-doped NFAs, the capacity dropped by only 12±4% and 13±4%, respectively. The enhanced surface area and availability of active sites of the NFAs thus also facilitate higher chemical reaction rates, increasing the hydrogen production throughput without a substantial loss in capacity. For example, a two-fold increase in hydrogen production duration (from 100s to 200s) with the cobalt-doped NFA corresponds to an overall hydrogen production throughput increase of 33%, with a capacity loss of less than 4%.

[0137] Finally, a highly important performance metric is the anode's cycling stability. Ni(OH).sub.2 battery electrodes can withstand thousands of electrochemical charge-discharge cycles by avoiding overcharging, as discussed above. Moreover, we have previously shown that the electrochemical discharge (reduction) of NiOOH can be seamlessly replaced by chemical regeneration without damage or degradation of the anode's microstructure, phase composition or chemical composition. Here, the cycling stability of an undoped NFA was demonstrated by subjecting it to a series of 100 E-TAC water splitting cycles. After about 20 cycles, both anodes displayed high stability up the 100th cycle (Q=5.2±0.3C and Q=1.8±0.2 C for the EDA and NFA, respectively), with similar X-ray diffraction pattern and morphology to those observed after layer growth by galvanic cycling. This is a preliminary positive indication of the anode's cycling stability in E-TAC water splitting.

[0138] Anode Optimization: Effect of Fiber Morphology

[0139] The electrochemical properties of a core-shell Ni/Ni(OH).sub.2 NFA depend on the microstructure and surface area of the nickel substrate. Tuning the anode's microstructure and consequent electrochemical performance is possible by controlling the various synthesis process parameters. To demonstrate this, we explored the influence of the fiber microporosity and diameter on the charge density of the resulting anodes. SEM images of three nickel nanofiber substrates prepared with maximum sintering temperatures of 500, 600, and 700° C. are shown in FIGS. 17A, B and C, respectively. The sintering temperature affects the fiber micro-structure and porosity, with almost complete loss of microporosity after sintering at 700° C., leaving only a very coarse granular microstructure. Preliminary electrochemical performance was then evaluated by electrochemical charge-discharge cycles, showing a three-fold increase in volumetric charge capacity for the anode sintered at 500° C. compared to 700° C., presenting a clear advantage for maintaining a fine porous structure by controlling the sintering temperature.

[0140] Next, we looked at the fiber diameter, and its effect on the anode's performance. We postulated that reducing the fiber diameter would contribute to an enhanced performance both by increasing the active surface area, and by increasing the weight ratio of the active shell layer and the nickel core. Thinning the fiber diameter was achieved by diluting the original precursor solution, so that its viscosity decreased twofold, and by raising the voltage during the ES process, which resulted in extra stretching of the fibers during electrospinning. FIG. 17D shows the SEM images of a nickel nanofiber substrate having a fiber diameter of 130 nm. The preliminary electrochemical performance demonstrates that reducing the fiber diameter by half results in almost doubling of the volumetric charge capacity.