Electrode materials that include an active composition of the formula MgzMxOy for group II cation-based batteries

Abstract

Preparation, characterization, and an electrochemical study of Mg.sub.0.1V.sub.2O.sub.5 prepared by a novel sol-gel method with no high-temperature post-processing are disclosed. Cyclic voltammetry showed the material to be quasi-reversible, with improved kinetics in an acetonitrile-, relative to a carbonate-, based electrolyte. Galvanostatic test data under a C/10 discharge showed a delivered capacity >250 mAh/g over several cycles. Based on these results, a magnesium anode battery, as disclosed, would yield an average operating voltage ˜3.2 Volts with an energy density ˜800 mWh/g for the cathode material, making the newly synthesized material a viable cathode material for secondary magnesium batteries.

Claims

1. A composition of matter comprising a compound according to the formula: Mg.sub.z,M.sub.xO.sub.y, wherein M is vanadium (V) metal, a mole fraction of Mg in the compound is z/(x+y+z), a mole fraction of metal M in the compound is x(x+y+z) and a mole fraction of O in the compound is y/(x+y+z), wherein x=2, y=5 and z=0.1, and wherein the composition of further comprises: an internal water content of 1.8 equivalents of H.sub.2O per formula unit at room temperature, resulting in a compound formula of Mg.sub.0.1V.sub.2O.sub.5*1.8H.sub.2O.

2. The composition of claim 1, wherein: the compound exhibits a crystallographic structure that forms ion channels through which magnesium ions in the +2 valence state (Mg.sup.2+) flow.

3. The composition of claim 2, wherein: a rate of Mg.sup.2+ ion flow is affected by the crystallographic structure.

4. The composition of claim 2, wherein: a rate of Mg.sup.2+ ion flow is affected by a bonding character within the ion channels.

5. The composition of claim 4, wherein: the bonding character within the ion channels is a function of a water content within the compound.

6. The composition of 2, wherein: the compound comprises a plurality of crystallites, each crystallite having a characteristic size.

7. The composition of claim 6, wherein the characteristic size of the crystallites is less than one micrometer (1 μm).

8. The composition of claim 6, wherein: a rate of Mg.sup.2+ ion flow is affected by the characteristic size of the plurality of crystallites.

9. The composition of claim 8, wherein: the rate of Mg.sup.2+ ion flow increases with a decrease in the characteristic size of the plurality of crystallites.

10. An electrode material comprising: an active material, the active material comprising a composition of the formula Mg.sub.zM.sub.xO.sub.y, and an internal water content of 1.8 equivalents of H.sub.2O per formula unit at room temperature, resulting in a compound formula of Mg.sub.0.1V.sub.2O.sub.5*1.8H.sub.2O; wherein M is vanadium (V) metal, z/(x+y+z) is the mole fraction of Mg, x/(x+y+z) is the mole fraction of metal M, and y/(x+y+z) is the mole fraction of O in the active material; and wherein x is equal to 2, y is equal to 5 and z is equal to 0.1.

11. The electrode material of claim 10, wherein: the active material is incorporated into a cathode; and a response of the cathode to an applied voltage is reversible when the cathode is immersed in an electrolyte containing Mg.sup.2+ ions.

12. A battery system comprising: a battery comprising a cathode, an anode, and an electrolyte; wherein the cathode comprises an active material having the formula Mg.sub.gM.sub.xO.sub.y, and an internal water content of 1.8 equivalents of H.sub.2O per formula unit at room temperature, resulting in a compound formula of Mg.sub.0.1V.sub.2O.sub.5*1.8H.sub.2O; wherein M is vanadium (V) metal, z/(x+y+z) is the mole fraction of Mg, x/(x+y+z) is the mole fraction of the metal V, and y/(x+y+z) is the mole fraction of O in the compound; wherein x is equal to 2, y is equal to 5 and z is equal to 0.1; wherein the anode comprises a Mg-containing material; and wherein the electrolyte comprises an ionic solution containing Mg.sup.2+ ions, Li.sup.1+ ions, or both.

13. The battery system of claim 12, wherein: the battery is a secondary battery.

14. The battery system of claim 13, wherein: the cathode exhibits electrochemically relevant reversibility when immersed in the electrolyte and subjected to a voltage differential with respect to the anode.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a cyclic voltammogram showing the current at the Mg.sub.zV.sub.xO.sub.y electrode as a function of applied voltage in a lithium ion-based electrolyte.

(2) FIG. 2 is a cyclic voltammogram showing the current at the Mg.sub.zV.sub.xO.sub.y electrode as a function of applied voltage in a magnesium ion-based electrolyte.

(3) FIG. 3 is a cyclic voltammogram showing the current at the Mg.sub.zMn.sub.xO.sub.y electrode as a function of applied voltage in a lithium ion-based electrolyte.

(4) FIG. 4 is a cyclic voltammogram showing the current at the Mg.sub.zMn.sub.xO.sub.y electrode as a function of applied voltage in a magnesium ion-based electrolyte.

(5) FIG. 5 is an X-ray powder diffraction (XRD) pattern of Mg.sub.0.1V.sub.2O.sub.5 with the inset providing a schematic of the Mg.sub.xV.sub.2O.sub.5 structure.

(6) FIGS. 6A and 6B are scanning electron micrographs of Mg.sub.0.1V.sub.2O.sub.5 at magnifications of 20,000× and 1,000×, respectively.

(7) FIGS. 7A and 7B are cyclic voltammograms obtained by slow scan voltammetry at 10.sup.−4 Volts/s in which the working electrode is Mg.sub.0.1V.sub.2O.sub.5, the reference electrode is Ag/Ag.sup.+, and the auxiliary electrode is Pt. The results shown in FIG. 7A were obtained using an electrolyte containing 0.1M Mg(ClO.sub.4).sub.2 in CH.sub.3CN, while those in FIG. 7B were obtained using an electrolyte containing 0.1M Mg(TFSI).sub.2 in EC:DMC (30:70).

(8) FIGS. 8A and 8B are graphs that show the results of a galvanostatic cycle test at C/10 of Mg.sub.0.1V.sub.2O.sub.5 in 0.5 M Mg(ClO.sub.4).sub.2 CH.sub.3CN electrolyte versus Ag/Ag.sup.+. FIG. 8A shows a plot of voltage versus capacity for cycles number 1, 2, and 5, while FIG. 8B charts the capacity versus cycle number for cycles 1-7.

(9) FIG. 9 is a composite of three XRD patterns, offset vertically for clarity, showing the patterns for electrodes in the a) discharged, b) charged, and c) as-prepared states.

DETAILED DESCRIPTION

(10) Disclosed is a new battery system that uses naturally abundant, low cost materials with minimal environmental impact. The disclosed batteries are based on alternative technologies to lithium ion batteries which are currently in widespread use. The composition of the battery materials becomes increasingly significant as the power source installations increase in size. The energy density of the final battery is somewhat lower than lithium ion batteries, but is higher than lead acid batteries. However, the new battery technology will have significantly lower environmental impact than the lead-based systems.

(11) Lithium ion batteries and lead acid batteries are used as benchmark systems. Variation in the cost of specific batteries is dependent on many factors including design, manufacturing process, and the number of units manufactured. The cost comparisons prepared here take into account the cost of the metals that form the basis of the battery system materials. Comparisons among lithium, lead, and magnesium metals and anodes formed from those metals are provided in Tables 1 and 2. Several comparisons provide insight into the projected impact of this invention. The cost of magnesium per pound is 25 times lower than that of lithium metal and approximately equal to that of lead. However, the metal's cost needs to be considered in light of its electrochemical value. The cost per 1000 Ah is dramatically different where magnesium is 12 and 14 times lower than lead and lithium, respectively. When this is translated to $/Wh, there is a 12 and 23 times lower cost for magnesium. Thus, use of a magnesium-based system paves the way toward more than an order of magnitude cost reduction when compared to lead- and lithium-based systems.

(12) Comparisons of energy density in Wh/kg are also important to consider. The energy density comparisons are based on an anode-focused analysis. The cathode energy density for magnesium-based batteries is likely to be similar to lithium ion systems based on metal oxides and higher than lead, thus, the relative comparisons here are reasonable. Lithium metal provides high energy density; however, lithium ion batteries do not currently use lithium metal anodes and have an energy density 10-fold less. Thus, the anode energy density of lithium ion, lead, and magnesium are 1390, 570 and 4850, respectively. Therefore, the energy density of the magnesium-based batteries proves to be 3.4× higher than lithium ion and 8.5× higher than lead batteries.

(13) Through our initial research we identified materials suitable for use in a new battery system using naturally abundant, low cost materials with minimal environmental impact. Anode materials may include Mg metal, materials alloying with Mg, or other Mg-containing materials. Cathode materials, which are the subject of this disclosure, are based on metal oxides Mg.sub.zM.sub.xO.sub.y (M=Fe, Mn, V), where rates of magnesium ion (Mg′) transport are facilitated by small crystallite size and tuning the crystallographic structure and bonding character within the ion channels. We have demonstrated the ability to control composition, crystallite size and interior water content of several metal oxide systems by direct low-temperature syntheses without secondary processing or constraining media. Further, we have demonstrated significant favorable crystallite size and water content effects on capacity and capacity retention during cycling in lithium based cells. Similar favorable impact in magnesium-based cells is also achieved.

(14) Successful utilization of the disclosed synthesis of new electrode materials will lead to a new class of secondary batteries based on magnesium ions. The criteria for the composition of materials selected are based on high natural earth abundance, low environmental impact, and opportunity for low cost. These considerations become increasingly significant as the power source installations increase in size. Magnesium is ˜1000 times more abundant than lithium and is air stable, both beneficial criteria for large power systems. Magnesium would have much lower environmental impact than lead-, nickel-, or cadmium-based batteries.

(15) The disclosed invention provides for the development of a new battery system using naturally abundant, low-cost materials with minimal environmental impact. The cost of magnesium is ˜$1.11/lb compared to $28.24/lb for lithium, providing an opportunity for substantial cost savings. A crystalline MgV.sub.2O.sub.6 precursor was used to demonstrate feasibility; however, this precursor was treated by ion exchange to form a magnesium vanadium oxide gel, with variable Mg/V ratios being possible in the target product. This synthetic approach avoided the need for extensive high temperature post synthesis treatment. The magnesium content was controlled and, in this case, maintained as x=0.1 in Mg.sub.xV.sub.2O.sub.5, yielding a layered material. This low temperature sol-gel-derived magnesium vanadium oxide was then evaluated as a possible cathode material in magnesium ion-based electrolytes, where both solvent and salt variations of the electrolytes were explored. This work demonstrates the utility of a low-temperature, aqueous-based synthesis of a Mg.sub.xV.sub.2O.sub.5 material and the promise of Mg.sub.xV.sub.2O.sub.5 materials as cathode materials in magnesium ion battery systems.

(16) A novel low-temperature preparation of a sol-gel-based magnesium vanadium oxide material, Mg.sub.xV.sub.2O.sub.5, is disclosed. X-ray diffraction showed 00/turbostratic ordering, with an interlayer spacing of 12.3 Å. Inductively-coupled plasma optical emission spectroscopy and thermogravimetric analysis are consistent with the composition of Mg.sub.0.1N.sub.2O.sub.5.1.8H.sub.2O. Cyclic voltammetry demonstrated quasi-reversible behavior, with improved current per gram for an acetonitrile-based electrolyte relative to carbonate-based electrolyte. Cycle type testing was conducted at a C/10 rate where the material displayed a sloping discharge curve, delivering ˜250 mAh/g over multiple discharge-charge cycles consistent with the insertion of one equivalent of Mg.sup.2+ (two electron equivalents) per formula unit.

(17) Translating these results to a magnesium anode battery based on standard potentials in aqueous solution (G. G. Perrault in Chapter 22 “Beryllium, Magnesium, Calcium, Strontium, Barium, and Radium,” of A. J. Bard, et al., Standard Potentials in Aqueous Solution, New York: Marcel Dekker Inc., 1985, pp. 687-699, International Union of Pure and Applied Chemistry, CRC Press New York (1985), which is incorporated by reference) would yield an average operating voltage of approximately 3.2 Volts with an energy density of approximately 800 mWh/g for the cathode material. Consistent interlayer spacing was observed upon Mg.sup.2+ insertion and removal, demonstrating promise for improved cathode material stability over multiple long-term discharge-charge cycling. Thus, the electrochemistry of the sol-gel prepared Mg.sub.0.1N.sub.2O.sub.5.1.8H.sub.2O material demonstrates that Mg.sub.xV.sub.2O.sub.5 materials prepared by low-temperature sol-gel methods are useful cathode materials for magnesium-based batteries.

(18) Cathode examples. Manganese (Mn), vanadium (V), and iron (Fe) have been selected to form the basis of the material framework structures for the proposed oxides (Mg.sub.zM.sub.xO.sub.y). Manganese (Mn) is advantageous due to its environmental sustainability and low cost. Structural diversity is another advantage, where manganese oxides can be tuned to dimensions suited for ion transport. Vanadium (V) offers the greatest synthetic diversity, allowing for many different types of layered structures. Iron oxides provide the advantage of low cost and earth abundance. While the Fe(III)/Fe(II) couple would be expected to have lower potential than the vanadium (V) compound, the Fe(III)/Fe(IV) couple would be expected to have higher voltage if accessible.

(19) Specific examples for Mg.sub.xMn.sub.xO.sub.y and Mg.sub.zV.sub.xO.sub.y are disclosed. FIG. 1 shows the voltammetry of Mg.sub.zV.sub.xO.sub.y in lithium ion-containing electrolyte. FIG. 2 shows voltammetry of Mg.sub.zV.sub.xO.sub.y in magnesium ion-containing electrolyte. FIG. 3 illustrates an example of the voltammetry of Mg.sub.zMn.sub.xO.sub.y in lithium ion-based electrolyte. FIG. 4 shows the voltammetry of Mg.sub.zMn.sub.xO.sub.y in magnesium-based electrolyte. Notably, the cathodes show reversibility in the magnesium-based electrolyte, illustrating their suitability for use in a magnesium ion battery system. Thus, electrochemically relevant reversibility of the materials is demonstrated in Mg′ ion-containing electrolytes. This initial data demonstrates the utility of these materials for use in magnesium-based batteries. The Fe(III)/Fe(II) couple in iron oxides would have lower potential than the vanadium (V) compound; however, if accessible, the Fe(III)/Fe(IV) couple would be expected to have higher voltage.

(20) The synthesis of the active materials, Mg.sub.zMn.sub.xO.sub.y and Mg.sub.zV.sub.xO.sub.y, can be accomplished by various means including coprecipitation, ion exchange, sol-gel synthesis, high temperature reactions, and hydrothermal synthesis. High temperature reaction conditions as described here are those >300° C. Low temperature conditions are those that are below the reflux point of water. In this case, the sol-gel reaction is conducted at ambient (room) temperature. Iron-based materials may be prepared by coprecipitation methods. Fe(II) salts that are soluble, such as iron sulfate and iron nitrate, may be used as starting materials for the reaction. The coprecipitation reactions may be carried out at ambient temperature.

(21) TABLE-US-00001 TABLE 1 Properties of battery anode materials Electron Specific Atomic weight equivalents/ capacity Average Specific energy Metal (g/mol) formula unit (Ah/kg) voltage (Volts) (Wh/kg) Lithium 6.94 1 3,862 3.6 1,390 Lead 207 2 259 2.2 570 Magnesium 24.3 2 2,205 2.2 4,850

(22) TABLE-US-00002 TABLE 2 Cost analysis of battery anode materials Cost per Cost per Cost per mass unit capacity unit energy unit Metal ($/kg) ($/1000 Ah) ($/1000 Wh) Lithium $62.20 $16.11 $57.98 Lead  $3.70 $14.29 $31.44 Magnesium  $2.46  $1.11  $2.45

Experimental Protocol

(23) Magnesium vanadium oxide (Mg.sub.xV.sub.2O.sub.5) was synthesized via a novel sol-gel based process, inspired by the sol-gel preparation of sodium vanadium oxide (Na.sub.xV.sub.2O.sub.5) (C.-Y. Lee, et al., “Synthesis and characterization of sodium vanadium oxide gels: the effects of water (n) and sodium (x) content on the electrochemistry of Na.sub.xV.sub.2O.sub.5.nH.sub.2O,”Physical Chemistry Chemical Physics, 13, 18047 (2011), which is incorporated by reference in its entirety). First, magnesium vanadate (MgV.sub.2O.sub.6) was prepared and isolated as a crystalline material, as described previously (Sun, 2011). The crystalline material was dissolved in water at ambient temperature (20° C. to 25° C.) or up to 60° C. with stirring, and then treated by ion exchange to form the magnesium-deficient Mg.sub.xV.sub.2O.sub.5 sol. When an ion exchange resin was used, the solution interaction with the ion exchange resin was carried out in two ways; each was effective. The resin was formed into a column and the solution was passed through the column. Alternatively, the ion exchange resin was added to the solution and gently swirled for several minutes followed by filtration to remove the resin. After gelation at room temperature, the material was recovered and characterized by x-ray powder diffraction (XRD) using a Rigaku SmartLab XRD (sold by Rigaku Americas Corporation, 9009 New Trails Dr., The Woodlands, Tex. 77381-5209), with Cu K.sub.α radiation and Bragg-Brentano focusing geometry. The gelation time could be varied from 6 hours to 1 week, but was typically 1 to 3 days. Elemental composition was determined via inductively coupled plasma-optical emission spectrometry (ICP-OES) with a Thermo Scientific ICAP ICP-OES (sold by Thermo Fisher Scientific Inc., 81 Wyman Street, Waltham, Mass. 02451). Scanning electron microscopy (SEM) was performed using a Hitachi 4800 operating at 10 kV. Simultaneous thermogravimetric analysis/differential scanning calorimetry (TGA/DSC) was performed using a TA Instruments Q600 (sold by TA Instruments, 159 Lukens Drive, New Castle, Del. 19720). Brunauer-Emmett-Teller surface area was measured by nitrogen adsorption, with a Quantachrome Nova e Series instrument (sold by Quantachrome Instruments, 1900 Corporate Dr, Boynton Beach, Fla. 33426.)

(24) Electrochemical testing was conducted at room temperature between 20° C. and 27° C. A three-electrode assembly was used, with silver/silver ion (Ag/Ag.sup.+) reference and platinum auxiliary electrodes. For the electrolyte, 0.1 M or 0.5 M magnesium perchlorate (Mg(ClO.sub.4).sub.2) or magnesium bis(trifluoromethylsulfonylimide) (Mg(TFSI).sub.2) was used, in either acetonitrile (CH.sub.3CN) or 30:70 ethylene carbonate:dimethyl carbonate (EC:DMC) solvent. Each combination of salt and solvent was tested. Cyclic voltammetry data was collected between voltage limits of −1.0 Volt and +1.2 Volts using a scan rate of 0.1 mV/s. Galvanostatic data used a C/10 rate for both discharge and charge, between voltage limits of −1.0 Volt and +1.0 Volt. A 1C rate is defined as the full capacity discharging in 1 hour (1C). A C/2 rate would be full capacity discharging over the course of 2 hours and a 2C rate would be full discharge in ½ hour.

(25) Material Characterization

(26) Vanadium oxide xerogels (V.sub.2O.sub.5.nH.sub.2O) are comprised of vanadium oxygen layers formed from square pyramidal VO.sub.5 polyhedra, with water molecules present in the interlayer positions. In addition to water, a variety of metal ions can be positioned within the interlayer positions through ion exchange. Incorporation of sodium ions into layered vanadium oxides by introducing a Na.sup.+ source such as sodium hydroxide, sodium nitrate, sodium sulfate, or sodium chloride to a vanadate gel precursor has been previously described (M. Millet, et al., “A new hydrated sodium vanadium bronze as Li insertion compound,” Solid State Ionics, 112, 319 (1998); E. M. Sabbar, et al., “Synthetic Pathways to New Hydrated Sodium and Lithium Vanadium Bronzes,” Journal of Solid State Chemistry, 149, 443 (2000); L. Znaidi, et al., “Kinetics of the H.sup.+/M.sup.+ ion exchange in V.sub.2O.sub.5 xerogel,” Solid State Ionics, 28-30, 1750 (1988); and O. Durupthy, et al., “Influence of pH and ionic strength on vanadium (V) oxides formation. From V.sub.2O.sub.5.nH.sub.2O gels to crystalline NaV.sub.3O.sub.8.1.5H.sub.2O,” Journal of Materials Chemistry, 15, 1090 (2005), each of which is incorporated by reference in its entirety).

(27) We developed a streamlined synthetic approach allowing for direct incorporation of sodium during the gel formation step (Lee, 2011). In this methodology, a divalent cation (Mg.sup.2+) is incorporated via sol-gel methodology, resulting in the direct synthesis of a new magnesium vanadium oxide material, Mg.sub.zV.sub.2O.sub.5. Both the preparation and resulting material are new. Incorporation of divalent cations can be more difficult than that of monovalent cations due to their lower solubility and lower mobility in the solid state, requiring adaptation of the synthesis method to the alternate metal types used here.

(28) X-ray powder diffraction (XRD) data was collected on the as-prepared material (FIG. 5). This data showed pronounced 001 reflections consistent with lamellar turbostratic ordering, not seen before in Mg.sub.zV.sub.2O.sub.5-type compounds, (cf. V. Petkov, et al., “Structure of V.sub.2O.sub.5.nH.sub.2O Xerogel Solved by the Atomic Pair Distribution Function Technique,” J Am Chem Soc, 124, 10157 (2002), which is incorporated by reference in its entirety), and a pattern similar to previously reported sol-gel based sodium vanadium oxide materials (Lee, 2011). The observed XRD pattern was similar to previously indexed vanadium oxide (Na.sub.0.3V.sub.2O.sub.5.1.5H.sub.2O) prepared via ion exchange of a V.sub.2O.sub.5 xerogel (Durupthy, 2005). Based on this report, the observed pattern could be indexed with the major peaks at positions of 7°, 25°, and 50° two theta corresponding to the 001, 310, 020 reflections, respectively. The 2-0 position of the 001 peak was used to calculate an interlayer spacing of 12.3 Å.

(29) Previous findings on sodium-based vanadium oxides generated by a sol-gel method noted significant influence of the water content on the interlayer spacing which ranged from 11.1 Å to 11.9 Å at ambient temperature with water content (n) ranging from 0.75 to 1.38 (Lee, 2011). The level of hydration of the magnesium vanadium oxide was determined using thermogravimetric analysis (TGA), which showed the material to have 1.8 (ranging from 1.0 to 3.0) equivalents of water per formula unit at room (ambient) temperature (20° C. to 25° C.). The somewhat larger interlayer spacing observed with the magnesium vanadium oxide material is consistent with a higher water content as one interfoliar water layer has been reported to contribute a thickness of 2.8-3.0 Å (O. Pelletier, et al., “The effect of attractive interactions on the nematic order of V.sub.2O.sub.5 gels,” Europhysics Letters, 48, 53 (1999) and P. Aldebert, et al., “Vanadium pentoxide gels: Ill. X-ray and neutron diffraction study of highly concentrated systems: One-dimensional swelling,” J. Colloid Interface Sci., 98, 478 (1984), each of which is incorporated by reference in its entirety).

(30) Differential scanning calorimetry (DSC) showed a broad exotherm between 360° C. and 400° C. that could be attributed to crystallization of the amorphous vanadium oxide to a more ordered phase (P. Aldebert, et al., “Layered structure of vanadium pentoxide gels,” Materials Research Bulletin, 16, 669 (1981), which is incorporated by reference in its entirety). Although poorly crystalline, our as-prepared material was not fully amorphous. The Mg/V ratio in our product could be controlled via the synthetic approach and a range of compositions from 0.01 Mg/2 V to 1 Mg/2 V (Atom/Atom) was prepared and explored. A typical range of Mg used for this set of experiments was between 0.08 and 0.25 for each 2 V. For the experiments described here, the ratio was held at 0.1 Mg/2.0 V, as determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES). Therefore, based on the ICP-OES and TGA data, a formula of Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O was assigned. Scanning electron microscopy showed a granular morphology consisting of agglomerates of sub-micron sized particles (FIGS. 6A and 6B). The granular morphology was consistent with the low measured surface area of 4 m.sup.2/g. The surface area range was typically 2 m.sup.2/g to 6 m.sup.2/g.

(31) Electrochemical Evaluation

(32) The electrochemistry of the Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O material was initially assessed by slow scan rate voltammetry, using Ag/Ag.sup.+ as reference (FIGS. 7A and 7B). Two solvent systems (acetonitrile (CH.sub.3CN) or ethylene carbonate:dimethyl carbonate (EC:DMC)) and two electrolyte salts (magnesium perchlorate (Mg(ClO.sub.4).sub.2) or magnesium bis(trifluoromethylsulfonylimide) (Mg(TFSI).sub.2)) were evaluated. In CH.sub.3CN, a cathodic peak is noted at ˜0.0 Volts with an anodic peak at ˜0.5 Volts for the Mg(ClO.sub.4).sub.2 electrolyte (FIG. 7A).

(33) In CH.sub.3CN, a cathodic peak is noted at ˜0.0 Volts with an anodic peak at ˜0.2 Volts for the Mg(TFSI).sub.2 electrolyte. While the TFSI based-electrolyte shows slightly improved reversibility, the perchlorate based-electrolyte showed evidence of increased current per gram of active material. In EC:DMC, a cathodic peak is noted at ˜0.5 Volts for the Mg(ClO.sub.4).sub.2 electrolyte, with a much smaller cathodic peak at ˜0.0 Volts for the Mg(TFSI).sub.2 electrolyte (FIG. 7B). In the EC:DMC based solvent system, the system does not show an anodic peak in either electrolyte. In all cases, the peak positions and the current per gram do not change substantially from cycles 2 to 3. The peak potentials measured through the experiments could shift by ±100 mV.

(34) Recent theoretical analysis of the de-solvation energy for Mg′ ions in organic solvents provides a basis for understanding the experimental observations of significantly improved current per gram of active material for the CH.sub.3CN-based electrolyte relative to the EC:DMC electrolyte (M. Okoshi, et al., “Theoretical Analysis on De-Solvation of Lithium, Sodium, and Magnesium Cations to Organic Electrolyte Solvents,” J. Electrochem. Soc., 160, A2160 (2013), which is incorporated by reference in its entirety). The de-solvation energy of Mg.sup.2+ in CH.sub.3CN was calculated to be 490.8 kJ/mol, while that of EC is reported as 552.9 kJ/mol. While DMC was not reported, diethyl carbonate (DEC) was reported to be 623.0 kJ/mol. Thus, based on the prior theoretical analysis, the kinetics of Mg.sup.2+ ion insertion would be more favorable in CH.sub.3CN relative to the carbonate-based solvents, consistent with the experimental observations.

(35) Discharge-charge type cycle tests were conducted under galvanostatic control at a C/10 rate to assess behavior in secondary batteries, using the CH.sub.3CN-based electrolyte (FIGS. 8A and 8B). A sloping voltage profile was noted with small evidence of a voltage plateau near 0.0 Volts versus Ag/Ag.sup.+. The first discharge delivered 300 mAh/g, while the second discharge delivered 230 mAh/g indicating capacity loss between the first and second cycles. As cycling continued, there was a small increase in delivered capacity noted where the delivered capacity was 280 mAh/g at cycle 7, consistent with exchange of approximately one Mg′ ion (two electron equivalents) per formula unit. Although the range of Mg′ insertion was rate dependent, under similar conditions a typical range was 0.8 to 1.0.

(36) An increase in discharge capacity with successive cycling has been previously reported for Mg′ insertion into V.sub.2O.sub.5, attributed to a gradual wetting of electrode with electrolyte or an increase of electronic conductivity of the electrode (Imamura, J. Electrochem. Soc., 2003). Notably, this capacity was achieved with conventional electrode fabrication and processing, demonstrating that continuous cycling of the sol-gel prepared Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O material is feasible. Alternative electrode fabrication methods resulting in electrodes with higher porosity may further facilitate the ion insertion and extraction.

(37) As noted above, an interlayer spacing of 12.3 Å was determined for the as-synthesized Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O material. XRD scans were recorded for the magnesium vanadium oxide electrodes as prepared, after discharge, and after charge. (See FIG. 9, in which the observed narrow peak at 26° 2 ⊖ is due to graphite.) The material does have some crystalline nature, as it can be characterized by XRD. Nevertheless, the crystallinity is so low that detailed structural analysis from the diffraction pattern is not possible. However, no change in the (001) peak position at ˜7° 2 ⊖ was noted, indicating no change in the interlayer spacing of the vanadium-oxygen layers.

(38) This is in contrast to previous reports on insertion of Mg.sup.2+ into V.sub.2O.sub.5, where a change from ˜14.1 Å for V.sub.2O.sub.5 to ˜12.3 Å for Mg.sub.1.0V.sub.2O.sub.5 was reported on discharge, and on charge the spacing increased to ˜13.7 Å for Mg.sub.0.1V.sub.2O.sub.5 (D. Imamura and M. Miyayama, “Characterization of magnesium-intercalated V.sub.2O.sub.5/carbon composites,” Solid State Ionics, 161, 173 (2003), which is incorporated by reference in its entirety). This observation can be understood due to the differences in the two materials. As the water content was not identified in the prior report, it is possible that differences in the water content could also contribute to the differences in the interlayer spacing.

(39) Further, in this disclosure, Mg.sup.2+ was present during formation of the vanadium oxide layers, resulting in reduced interlayer repulsion of the oxide layers and a smaller interlayer spacing for the as-synthesized Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O material relative to the previously reported V.sub.2O.sub.5 material. Consistent d-spacing upon Mg.sup.2+ insertion and removal from the cathode active material should promote improved stability over multiple long-term discharge-charge cycling. In these experiments good capacity was retained after 10 discharge-charge cycles. Notably, the d-spacing of sol-gel based layered vanadium oxide materials is directly related to water content where the d-spacing decreases as the water level in the interlayer spacing decreases (Lee, 2011). One interfoliar water layer has been associated with 2.8-3.0 Å thickness (Aldebert, 1984 and Aldebert, 1981). Thus, the observation of no change in d-spacing for the Mg.sub.0.1V.sub.2O.sub.5.1.8H.sub.2O indicates that the interlayer water remained within the structure during the discharge and charge processes.

(40) While the above is a description of what are presently believed to be the preferred embodiments of the invention, various alternatives, modifications, and equivalents may be used. Those skilled in the art will realize that other and farther embodiments can be made without departing from the spirit of the invention, and it is intended to include all such further modifications and changes as come within the true scope of the following claims. Therefore, the above description should not be taken as limiting the scope of the invention, which is defined solely by the claims.