LOW TEMPERATURE, HIGH YIELD SYNTHESIS OF HYDROGEN TERMINATED HIGHLY POROUS AMORPHOUS SILICON, AND NANOMATERIALS AND COMPOSITES FROM ZINTL PHASES

Abstract

The present disclosure relates to an improved process for the synthesis of hydrogen terminated silicon from Zintl phases. The hydrogen terminated silicon is useful, for example, as explosives, chemical and biochemical sensors, optoelectronic materials and Li-ion battery anode ion storage materials. The present disclosure also relates to an improved process for the synthesis of nanomaterials and composites from Zintl phases. The nanomaterials and composites are useful, for example, as ion storage materials.

Claims

1. A process for preparing hydrogen terminated silicon, the process comprising: a) reacting a silicon containing Zintl compound with an aluminum compound, optionally in the presence of a support material b) optionally isolating the resulting product; c) adding a hydroxyl containing compound to produce the hydrogen terminated silicon; and d) optionally isolating they hydrogen terminated silicon, wherein step a) is conducted: i) by mechanical milling; ii) in a solvent; or iii) in a molten salt comprising the aluminum compound in combination with one or more alkali metal halides, alkali earth metal halides, or any combination thereof.

2. The process according to any one of claim 1, wherein the hydrogen terminated silicon is substantially amorphous.

3. The process according to claim 1, wherein the hydrogen terminated silicon has a surface area of about 100 to about 1500 m.sup.2/g.

4. The process according to claim 1, wherein the hydrogen terminated silicon has a porosity of about 40 to about 90%.

5. The process according to claim 1, wherein the hydrogen terminated silicon has a BM desorption cumulative pore volume of about 0.4 to about 1.1 cm.sup.3/g.

6. The process according to claim 1, wherein the hydrogen terminated silicon has an average BJH desorption pore width of about 1 to about 40 am.

7. The process according to claim 1, wherein the hydroxyl containing compound is an alcohol, a hydroxyl containing polymer, a carboxylic acid, or any combination thereof.

8. The process according to claim 1, wherein the hydroxyl containing compound is an alcohol.

9. The process according to claim 1, wherein the hydroxyl containing compound is methanol, ethanol, propanol, isopropanol, pentanol, hexanol, or any combination thereof.

10. The process according to claim 1, wherein step c) further comprises adding an inorganic acid.

11. The process according to claim 1, wherein step c) further comprises adding hydrochloric acid.

12. The process according to claim 1, wherein the Zintl compound is a compound comprising a) a Group 1 metal, a Group 2 metal, or any combination thereof, and b) silicon.

13. The process according to claim 1, wherein the Group 1 and/or Group 2 metal is selected from Li, Na, K, Rh, Cs, Be, Mg, Cn, Sr, Ba, and any combination thereof.

14. The process according to claim 1, wherein the Group 1 and/or Group 2 metal is Mg.

15. The process according to claim 1, wherein the Zintl compound is Mg.sub.2Si.

16. The process according to claim 1, wherein the aluminum compound is selected from halides, oxides, alkoxides, hydroxides, sulfates, sulfites, nitrates, nitrites, phosphates, and phosphites of aluminum, hydrates of any of the foregoing, and any combination thereof.

17. The process according to claim 1, wherein the aluminum compound is an aluminum halide.

18. The process according to claim 1, wherein the aluminum compound is AlCl.sub.3, AlBr.sub.3, AlI.sub.3, or any combination thereof.

19. The process according to claim 1 wherein the support material, if present, is selected from carbonaceous materials, metals, glasses, oxides, nitrides, borides, and any combination thereof.

20. The process according to claim 1, wherein the support material, if present, is a carbonaceous material selected from carbon spheres, graphite, carbon nanotubes, carbon nanofibers, carbon fibers, hard and soft carbons, activated carbons, carbon blacks, amorphous carbons, soots, graphene, fullerenes, coal, coke, fossil fuel derived materials, biomass, biochar, charcoal, chars, and any combination thereof.

21. The process according to claim 1, wherein the support material, if present, is selected from a multiwall graphene nanoshell (MGNS), graphite, and any combination thereof.

22. The process according to claim 1, wherein step a) is conducted in a solvent, in the solid state, by molten salt or by solid/solid reaction.

23. The process according to claim 1, wherein step a) is conducted in a molten salt comprising an aluminum compound, optionally in combination with one or more alkali metal halides, alkali earth metal halides, or any combination thereof.

24. The process according to claim 1, wherein step a) is conducted in a solvent selected from aromatic hydrocarbons, nonaromatic hydrocarbons, amities, aprotic solvents, and any combination thereof.

25. The process according to claim 1, wherein step a) is conducted in a solvent selected from cyclohexane, toluene, and any combination thereof.

26. The process according to claim 1, wherein step a) is conducted at a temperature ranging between about 50° C. and about 600° C., between about 50° C. and about 200° C., between about 100° C. and about 200° C., between about 120° C. and about 180° C., at about 125° C., or at about 180° C.

27. The process according to claim 1, wherein the hydrogen terminated silicon comprises between 0 and about 20 wt. % Al, between 0 and about 15 wt. % Al, between 0 and about 10 wt. % Al, between 0 and about 5 wt. % Al, between 0 and about 4 wt. % Al, between about 0 and about 3 wt. % Al, between about 0 and about 2 wt. % Al, between about 0 and about 1 wt. % Al or between about 1 and about 3 wt. % Al.

28. Hydrogen terminated silicon prepared by a process according to claim 1.

29. An explosive comprising hydrogen terminated silicon prepared by a process according to claim 1.

30. An ion storage material comprising hydrogen terminated silicon prepared by a process according to claim 1.

31. An electrode comprising an ion storage material of claim 30.

32. A secondary ion battery comprising the electrode of claim 31.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0076] FIG. 1(a) is a powder X-ray diffraction pattern of a crude Si/C nanomaterial prepared in toluene at 125° C.

[0077] FIG. 1(b) is a powder X-ray diffraction pattern of a Si/C nanomaterial prepared in toluene at 1255° C. following washing with HCl.

[0078] FIG. 2(a) is a powder X-ray diffraction pattern of a crude Si/C nanomaterial prepared in cyclohexane at 180° C.

[0079] FIG. 2(b) is a powder X-ray diffraction pattern of a Si/C nanomaterial prepared in cyclohexane at 180° C. following washing with HCl.

[0080] FIG. 3(a) is a powder X-ray diffraction pattern of a Si/MGNS nanomaterial composite.

[0081] FIG. 3(b) is a powder X-ray diffraction pattern of MGNS.

[0082] FIG. 4(a) shows nitrogen isotherms for MGNS.

[0083] FIG. 4(b) shows nitrogen isotherms for Si/MGNS.

[0084] FIG. 4(c) shows nitrogen isotherms for Si.sub.nano.

[0085] FIG. 5(a) shows a TEM micrograph of purified agglomerates of Si nanocrystals.

[0086] FIG. 5(b) shows a High-Resolution TEM (HRTEM) micrograph of purified agglomerates of Si nanocrystals showing lattice spacing characteristic of crystalline Si.

[0087] FIG. 6 shows a HRTEM micrograph of a Si particle, showing it to be an agglomerate of sintered nanocrystals.

[0088] FIGS. 7(a)-7(d) are energy dispersive X-ray HyperMaps of Si.sub.nano: HAADF image FIG. 7(a), silicon FIG. 7(b), aluminum FIG. 7(c), and oxygen FIG. 7(d).

[0089] FIG. 8 is a TEM micrograph of Si/MGNS composite, showing the MGNS intimately mixed with the Si nanocrystals. The inset is an electron diffraction pattern of the area shown in the figure showing reflection rings for MGNS (labeled “C”) and Si.

[0090] FIG. 9 is a HRTEM micrograph of the Si/MGNS composite.

[0091] FIGS. 10(a)-10(d) are energy dispersive X-ray HyperMaps of Si/MGNS: HAADF image FIG. 10(a), silicon FIG. 10(b), oxygen FIG. 10(c), and carbon FIG. 10(d).

[0092] FIG. 11(a) is a galvanostatic cycling profile for Si/MGNS for the formation cycle at C/20.

[0093] FIG. 11(b) is the differential capacity plot of the galvanostatic cycling profile shown in FIG. 11(a).

[0094] FIG. 12 is a differential capacity plot for MGNS cycled at C/5 rate.

[0095] FIG. 13 is a galvanostatic charge/discharge plot for a MGNS cell cycled at C/5 rate.

[0096] FIG. 14(a) is a galvanostatic charge/discharge plot for Si/MGNS cycled at C/5 rate for cell cycles 2 (solid line), 25 (dots), 50 (dashes) and 75 (dots and dashes).

[0097] FIG. 14(b) is a differential capacity plot Si/MGNS cycled at C/5 rate for cell cycles 2 (circles), 25 (squares), 50 (triangles) and 75 (diamonds).

[0098] FIG. 15(a) is a galvanostatic charge/discharge plot for Si/MGNS cycled at C/2 rate for cell cycles 2 (solid line), 25 (dots), 50 (dashes) and 75 (dots and dashes).

[0099] FIG. 15(b) is a differential capacity plot Si/MGNS cycled at C/52 rate for cell cycles 2 (circles), 25 (squares), 50 (triangles) and 75 (diamonds).

[0100] FIG. 16 compares the capacity of the Si/MGNS electrode cycled at C/5 rate (top line, circles), C/2 rate (middle line, squares), with mechanically mixed Si and MGNS at C/2 rate (bottom line, triangles), plotted as a function of cycle number. The first cycle of each electrode was performed at C/20.

[0101] FIG. 17 compares the coulombic efficiencies of Si/MGNS cells cycled at C/5 rate (top line, circles), C/2 rate (middle line, squares), with mechanically mixed Si and MGNS at C/2 rate (bottom line, triangles). The first cycle was performed at C/20. The inset shows data for cycles 60-75, rescaled for clarity.

[0102] FIG. 18 shows the cycle life of Si/MGNS cells with 0.7 mg/cm.sup.2 loading (top line, circles) and 1.25 mg/cm′ loading (bottom line, crosses).

[0103] FIG. 19 shows the coulombic efficiencies of Si/MGNS cells with 0.7 mg/cm.sup.2 loading (top line, circles) and 1.25 mg/cm.sup.2 loading (bottom line, crosses).

[0104] FIG. 20(a) is a galvanostatic cycling profile for Si MGNS (mechanically milled) for the formation cycle at C/20.

[0105] FIG. 20(h) is the differential capacity plot of the galvanostatic cycling profile shown in FIG. 20(a).

[0106] FIG. 21(a) is a galvanostatic charge/discharge plot for Si+MGNS (mechanically milled) cycled at C/2 rate for cell cycles 2 (solid line), 25 (dots), 50 (dashes) and 75 (dots and dashes).

[0107] FIG. 21(b) is a differential capacity plot Si+MGNS (mechanically milled) cycled at C/2 rate for cell cycles 2 (circles), 25 (squares), 50 (triangles) and 75 (diamonds).

[0108] FIG. 22 is a EIS spectrum of Si/MGNS (circles) and Si MGNS (mechanically milled) anodes (squares) obtained after loading Li+ to 0.2 V and their corresponding fits (solid lines). The Si+ MGNS shows higher impedance.

[0109] FIG. 23 depicts an exemplary equivalent circuit of the EIS model.

[0110] FIG. 24 is a powder X-ray diffraction pattern of a purified (HCl washed) Si nanomaterial prepared by mechanical milling of Mg.sub.2Si and AlCl.sub.3.

[0111] FIG. 25 depicts a first cycle galvanostatic plot of a purified (HCl washed) Si nanomaterial prepared by mechanical milling of Mg.sub.2Si and AlCl.sub.3.

[0112] FIG. 26 shows the cycle life of the purified (HCl washed) Si nanomaterial prepared by mechanical milling of Mg.sub.2Si and

[0113] FIG. 27 shows the XRD pattern for the synthesize Si/graphite product, with the Si mass fractions of each composite indicated on the figure.

[0114] FIG. 28 shows the galvanostatic cycling profile of Si/graphite product in which the mass fraction of Si in the composite is 10%.

[0115] FIG. 29 shows the cycle life (circles) and Coulombic efficiency (triangles) of Si/graphite product in which the mass fraction of Si in the composite is 10%.

[0116] FIG. 30 shows the galvanostatic cycling profile of Si/graphite product in which the mass fraction of Si in the composite is 30%.

[0117] FIG. 31 shows the cycle life (bottom line, circles) and coulombic efficiency (top line, triangles) for Si/graphite product in which the mass fraction of Si in the composite is 30%.

[0118] FIG. 32 shows the XRD pattern for the nanoscale Si produced from Mg.sub.2Si in a molten salt matrix of NaCl, KCl and AlCl.sub.3 at 150° C.

[0119] FIG. 33 shows a TEM image of the purified crystalline Si product produced from Mg.sub.2Si in a molten salt matrix of NaCl, KCl and AlCl.sub.3 at 150° C.

[0120] FIG. 34 shows a TEM image of the purified product as shown in FIG. 33 but at higher resolution.

[0121] FIG. 35 shows a HR-TEM image of the purified product as shown in FIGS. 33 and 34 but at higher resolution.

[0122] FIG. 36 shows an EDS hypermap of the purified product highlighting the location of Si (upper right), O (lower left), and Al (lower right) in the TEM image (upper left).

[0123] FIG. 37 shows the XRD pattern for the nanoscale Si produced from AlBr.sub.3 and Mg.sub.2Si in refluxing toluene.

[0124] FIG. 38 shows the adsorption (solid circles) and desorption (hollow circles) nitrogen isotherm of the hydrogen terminated silicon prepared in Example 6.

[0125] FIG. 39 shows the BM pore size distribution of the hydrogen terminated silicon prepared in Example 6.

[0126] FIG. 40 shows the powder XRPD pattern of the hydrogen terminated silicon prepared in Example 6.

[0127] FIG. 41 shows the Infrared (IR) ATR spectrum of the hydrogen terminated silicon prepared in Example 6.

[0128] FIG. 42 is an SEM image of the hydrogen terminated silicon prepared in Example 6.

[0129] FIG. 43 is a focused ion beam (FIB) cross section of the purified hydrogen terminated silicon in Example 6.

[0130] FIG. 44 is a TEM image of the hydrogen terminated silicon in Example 6,

[0131] FIG. 45 is HRTEM image of the hydrogen terminated silicon prepared in Example 6, showing its amorphous structure and porosity.

[0132] FIG. 46(a) shows a sample of the hydrogen terminated silicon prepared in Example 6, prior to oxidation and ignition.

[0133] FIG. 46(b) shows a sample of the hydrogen terminated silicon prepared in Example 6, after oxidation and ignition.

[0134] FIG. 47(a) shows a sample of the hydrogen terminated silicon prepared in Example 6, prior to oxidation and ignition.

[0135] FIG. 47(b) shows a sample of the hydrogen terminated silicon prepared in Example 6, after oxidation and ignition.

DETAILED DESCRIPTION OF THE INVENTION

[0136] U.S. Publication No 2016/0351894, which is hereby incorporated by reference in its entirety, describes hollow nanosphere composite based secondary cell electrodes.

[0137] As used herein, the term “substantially amorphous” when applied to hydrogen terminated silicon, means that the hydrogen terminated silicon is greater than about 90% amorphous, such as greater than about 91% amorphous, greater than about 92% amorphous, greater than about 93% amorphous, greater than about 94% amorphous, greater than about 95% amorphous, greater than about 96% amorphous, greater than about 97% amorphous, greater than about 98% amorphous, greater than about 99% amorphous, greater than about 99.5% amorphous, or greater than about 99.9% amorphous. The term “substantially amorphous” can also mean that the hydrogen terminated silicon exhibits less than about 10% crystallinity, such as less than about 9% crystallinity, less than about 8% crystallinity, less than about 7% crystallinity, less than about 6% crystallinity, less than about 5% crystallinity, less than about 4% crystallinity, less than about 3% crystallinity, less than about 2% crystallinity, less than about 1% crystallinity, less than about 0.5% crystallinity, or less than about 0.1% crystallinity.

[0138] As used herein, the term “highly porous” when applied to hydrogen terminated silicon, means, for example, that the hydrogen terminated silicon has a porosity of greater than about 40%, such as greater than about 50%, greater than about 60%, greater than about 70%, greater than about 80%, greater than about 90%, greater than about 95%, greater than about 99% or greater than about 99.5%.

[0139] One non-limiting example for step a) according to any of the processes described herein is shown by the following equation:

3 Mg.sub.2Si+4 AlBr.sub.3.fwdarw.3 Si+4 Al+6 MgBr.sub.2

[0140] Mg.sub.2Si is readily available in high purity, is low-cost and both magnesium and silicon are earth abundant. When using Mg.sub.2Si as a silicon source for Li-ion battery anodes, high product yield is desirable not only for economic reasons, but also because unreacted Mg.sub.2Si dissolution during purification can cause silane oxidation that leads to silica formation. Silica is non-conductive and as such, should be minimized so that sufficiently low impedance anodes can be realized. In addition, surface passivation of silicon nanoparticles with silica results in an increase in irreversible capacity, an effect that becomes increasingly important as particle size is reduced (see, e.g., Jaumann et al., Phys. Chem. Chem. Phys., 17, 24956, 2015). Etching silicon nanomaterials with hydrofluoric acid (HF) can mitigate the deleterious effects of silica formation to some extent. However, the processes described herein eliminate the need for this extremely hazardous agent. As described herein, nanoscale silicon (Si.sub.nano) prepared by Mg.sub.2Si oxidation in organic solvents can be worked up without concern for air or water oxidation, nor HF use, to produce highly reversible Si anodes.

[0141] Mg.sub.2Si has no appreciable solubility in any solvent; therefore, reactivity is limited by temperature, particle size, and the oxidizing agent used. Aluminum bromide may be selected as the oxidant because it is highly soluble in hydrocarbons and is unlikely to produce an alloy or intermetallic of Al with Si at low-temperatures. In addition, both aluminum and bromine are earth abundant. Aluminum is the most abundant metal and the third most abundant element in the Earth's crust and bromine is readily extracted from brine reservoirs and ocean water, which are estimated to hold 100 trillion tons (see, e.g., Ober, Mineral Commodity Summaries 2015, in US. Geological Survey, p. 34. U.S. Geological Survey, 2015). Additionally, AlBr.sub.3 can form complexes with alkali and alkaline earth metal salts, potentially improving reaction kinetics by inhibiting Mg salt precipitation during the reaction (see, e.g., U.S. Pat. No. 4,071,415). Following the reaction, aluminum and MgBr.sub.2 can be readily and safely removed by washing with hydrochloric acid (HCl), resulting in crystalline silicon (c-Si) in high yield. Thus, the processes described herein can be used to create Si/C composites with high reversible lithium electrochemical storage capacity.

[0142] The ion storage materials (e.g., Si nanoparticles) described herein may also be prepared in a molten salt comprising an aluminum compound (such as AlCl.sub.3, AlBr.sub.3, AlI.sub.3, or any combination thereof, such as, in one embodiment, AlCl.sub.3), optionally in combination with one or more alkali or alkali earth metal salts (e.g., in the absence of a solvent). Addition of the alkali or alkali earth metal salt (or combination of such salts) may lower the melting temperature of the aluminum compound, thereby lowering the cost of the process and easing the difficulty in the use of aluminum salts that melt at temperature higher than their sublimation temperature (such as, in one embodiment, AlCl.sub.3). Eliminating the solvent in the reaction provides an even more environmentally benign and potentially less expensive route to silicon nanoparticle production. For example, the ion storage materials (e.g., Si nanoparticles) described herein may be prepared by reacting Mg.sub.2Si in molten AlBr.sub.3 in the range of about 97 to about 263° C. (see, e.g., CRC handbook 73.sup.rd Edition, page 4-36, Physical Constants of Inorganic Compounds). A molten solution of AlCl.sub.3 may be advantageous over AlBr.sub.3 due to higher abundance of chlorine (further reducing cost), however, AlCl.sub.3 sublimes at 177.8° C. (1 atm), lower than its melting point at 190° C. (CRC handbook 73.sup.rd Edition, page 4-36), Inexpensive alkali metal chlorides, such as NaCl and KCl, form congruent melts with AlCl.sub.3 at 108.7° C. (for molar ratios 0.388 NaCl and 0.612 AlCl.sub.3) and at 156.7° C. (for molar ratios 0.5 NaCl to 0.5 AlCl.sub.3) (see e.g., Robelin et al J. diem. Thermodynamics, 36, 683-699). At the temperature of 156.7° C. and molar ratio of 0.388 NaCl and 0.612 AlCl.sub.3, a designated quantity of AlCl.sub.3 (0.112 mol %) can oxidize Mg.sub.2Si while a molten reaction medium is maintained. Up to approximately 0.18 mol. % could be used in theory as the salts are still liquid near approximately 0.68 AlCl.sub.3 and approximately 0.32 NaCl. The addition of KCl can reduce the NaCl and AlCl.sub.3 melt temperatures to 91.4° C. and 132.1° C., from 108.7° C. and 156.7° C., respectively. This can provide a lower reaction temperature for the production of nanomaterials and composites from the oxidation of a Zintl compound.

[0143] The description of the present embodiments of the invention has been presented for purposes of illustration, but is not intended to be exhaustive or to limit the invention to the form disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art. As such, while the present invention has been disclosed in connection with an embodiment thereof, it should be understood that other embodiments may fall within the spirit and scope of the invention. Patents and publications cited herein are incorporated by reference in their entirety.

EXAMPLES

Powder X-Ray Diffraction

[0144] Powder X-ray diffraction (XRD) patterns were obtained using a Rigaku Miniflex+ diffractometer with Cu Kα radiation. Electrochemical galvanostatic cycling was performed using an Arbin BT-2000. Ail mass measurements were recorded with Ohaus Analytical Plus balance (AP250D, 0.01 mg resolution).

Brunauer-Emmett-Teller (BET) Surface Area

[0145] BET surface area was determined from nitrogen adsorption isotherms obtained with a Micrometrics Tri-Star 3000. One of the three sample measurement ports of the Tri-Star was equipped with an empty sample tube with which the saturation vapor pressure (P.sub.0) of nitrogen (N.sub.2) was measured concurrently with each measurement of the equilibrium vapor pressure (P) over the sample. Isotherm adsorption data was recorded from 0.05 to 0.3 P.sub.0/P.

Barret-Joyner-Halenda (BJH) Pore Size and Pore Volume were Calculated from the Isotherm Adsorption Data.

Infrared (IR) Spectroscopy

[0146] IR spectra were collected on a Perkin Elmer Spectrum One Fourier Transform Infrared (FT-IR) spectrometer using attenuated total reflectance

Focused Ion Beam (FIB) Cross Section

[0147] Images were obtained on a FEI Helios Focused ion Beam (FIB) Scanning Electron Microscope (SEM) before and after etching to create a cross section using a gallium ion source. 130

Transmission Electron Microscopy (TEM)

[0148] TEM micrographs were obtained on a FEI Talos 200X and elemental analysis was obtained in scanning (STEM) operational mode with its Super-X Energy Dispersive Spectroscopy (EDS) detector.

Electrode Preparation

[0149] Anode material was formulated by combining 75 mg of a Si/MGNS composite material, 5 mg of carbon black (Super C45, TIMCAL America Inc.), 50 μL of ethanol (200 proof, Pharmco-Aaper) and 400 μL of 5% Li-polyacrylate binder solution prepared by dissolving poly(acrylic acid) (1000 kDa, Poly sciences) in deionized water and neutralizing with LiOH (95%, Strem) (see, e.g., U.S. Pat. No. 7,875,388). Mixing was performed with a Fritsch Pulverisette 23 MiniMill using a stainless steel cup (10 mL) and six stainless steel balls (5 mm diameter, 30:1 ball to powder mass ratio) at 50 Hz for 15 minutes to form a slurry. The slurry was cast onto copper foil (0.127 mm, 99.9%, Alfa Aesar) and dried under vacuum at 150° C. for 2 hours. Round electrodes (16 mm diameter) were cut from the resulting sheet using a die cutting press (MSK-T-07 Precision Disc Cutter, MTI Inc.). Electrodes were made tested at two different mass loadings: 0.7 mg/cm.sup.2 (13 μm thick) and 1.35 mg/cm.sup.2 (20 μm thick).

[0150] Mechanically mixed Si with MGNS was prepared for comparison by milling Si.sub.nano with MGNS (40% wt Si) for 30 minutes with a Fritsch Pulverisette 23 MiniMill using a stainless steel cup (10 mL) and six stainless steel balls (5 mm diameter) at 50 Hz. Electrodes for this active material and the Si/MGNS composite were prepared in an identical manner with a mass loading of 0.75 mg/cm.sup.2.

Electrochemical Testing Coin cells (CR2016, MTI Inc.) containing Si/MGNS and Li metal (2 cm.sup.2 disk, 0.05 g. 99.9%, MTI Inc.) electrodes separated by a polypropylene porous membrane (Celgard 3401) were assembled in an argon filled dry box (<0.1 ppm O.sub.2 and H.sub.2O). The electrolyte used was 1 M LiPF.sub.6 in EC:DMC (1:1 v/v, 40 μL) mixture (battery grade, <15 ppm H.sub.2O content, Sigma Aldrich) with 10% FEC (fluoroethylene carbonate) (>99%, Solvay) by volume. Cells were cycled at constant current (C/20) for one cycle between open circuit voltage to a cutoff potential then under reverse current to 1.5 V vs. Li/Li.sup.+. Subsequent cycling was performed at a C/5 or C/2 rate, with cutoff potentials of 10 mV and 20 mV vs. Li/Li.sup.+, respectively. A higher potential was chosen for C/2 to reduce the possibility of Li plating at high current density. The cells were rested for 15 minutes between discharge and charge. Impedance testing was performed using a Solartron Analytical 1260 impedance/gain-phase analyzer with 1287A potentiostat/galvanostat using an excitation voltage of 5 mV and a logarithmic sweep from 100 kHz to 100 mHz. Spectra were modeled to an equivalent circuit using ZView software (Scribner Associates, Inc.)

Results and Discussion

Example 1

MGNS Synthesis

[0151] MGNS consists of highly graphitic, nested, graphene layers that form the walls of “spheroidal” shells (they can appear significantly distorted from spherical), each surrounding a void left by removal of a nanoparticulate metal catalyst around which it formed (see, e.g., Wagner et al, Hollow Carbon Nanosphere Based Secondary Cell Electrodes, 2009). MGNS is a CO.sub.2-negative carbon material synthesized from biomass char, the waste product of pyrolytic bio-oil synthesis (see, e.g., Woolf et al, Environ. Sci. Technol, 48, 6492, 2014 and Xiu et al., Renew. Sustainable Energy Rev., 16, 4406, 2012).

[0152] MGNS was prepared by mixing 10 g of microcrystalline cellulose (Avicel PH-105 NF, FMC BioPolymer) and 4.5 g of CoCl.sub.2.6H.sub.2O (99.9%, Alfa Aesar) in a hardened steel cup (80 mL) with six hardened steel balls (1 cm diameter, 10:1 ball to powder mass ratio) at 300 rpm for 30 min using a Fritsch Pulverisette 6. The resulting powder was pressed at 10,000 psi (Carver 3851 benchtop laboratory press) into a 25.4 mm diameter pellet and heated under N.sub.2 gas from room temperature at a 75° C. min.sup.−1 ramp rate to 375° C. and held for a total heating time of 30 minutes. After cooling, the charred pellet was placed in a chamber with a 0.5 torr He atmosphere and irradiated by a 2 mm diameter 10.4 μm laser beam (Firestar t60, Synrad Inc.) while rotating at a linear velocity of 1.63 mm s.sup.−1 (1.2 rev min.sup.−1) for one full rotation. The surface of the pellet exposed to the laser was removed by gentle abrasion and purified by reflux in nitric acid (HNO.sub.3) for 4 hours then washed with deionized water and collected by vacuum filtration (P8 cellulose fiber, Fisher Scientific).

Si and Si/MGNS Synthesis

[0153] Mg.sub.2Si (99.5%, Alfa Aesar) was milled with a Fritsch Pulverisette 6 under a nitrogen atmosphere at 300 rpm for 5 minutes in an 80 mL WC cup with four 1 cm WC balls (10:1 ball to powder mass ratio). The milled Mg.sub.2Si (0.5 g) was then combined with 10 mL of cyclohexane or toluene (99.8% HPLC grade previously dried over NaK alloy and distilled. Acres Organics) and 2.6 g of AlBr.sub.3 (13% excess, 99.999% lump form, Alfa) in a dry 23 mL Teflon-lined digestion bomb in a nitrogen filled dry box (<0.1 ppm O.sub.2 and H.sub.2O). The bomb was sealed and placed into a muffle furnace pre-heated to 180° C. or 125° C., for cyclohexane and toluene, respectively. After 24 hours, the bomb was allowed to cool to room temperature and opened in air. The contents were added to acetone (100 ml, 99.7%, Fischer Scientific) and the product was recovered by centrifuge. The raw product was dispersed in 50% HCl solution (0.100 ml, ACS grade 36.5 to 38%, VWR Analytical) by sonic agitation and then magnetically stirred overnight. The solid product was isolated by vacuum filtration (polyester membrane with 1 μm pore size, Maine Manufacturing LLC) and rinsed with deionized water and acetone. A light-brown powder (178 mg Si.sub.nano) WES recovered after drying under vacuum. The product yield was 97% in cyclohexane and 93% in toluene based on initial mass of Mg.sub.2Si (this mass also includes any surface oxide that may have formed during purification).

[0154] A Si/MGNS composite was prepared in an analogous manner to the reaction in cyclohexane by adding 300 mg of MGNS to the reaction vessel prior to heating. 458 mg total mass of product was recovered after purification and drying (95% yield).

XRD Characterization

[0155] Prior to rinsing with acetone and HCl, the raw product of the reaction to produce Si.sub.nano in toluene at 125° C. was analyzed by XRD. The XRD pattern for the raw product and the product after HQ wash are shown in FIG. 1(a) and FIG. 1(b), respectively. As seen in FIG. 1(a), the highest intensity peaks match those expected for MgBr.sub.2. A small amount of Mg.sub.2Si still remained in the product, as indicated by the presence of one low intensity peak at 40.1 degrees (220). However, this minor impurity is easily removed during the HCl washing (FIG. 1(b)), and has little impact on the product yield (93%).

[0156] This reaction was also performed in cyclohexane but required a higher temperature of 180° C. for 24 hours. The XRD pattern for the raw product and the product after vacuum filtration and HCl wash are shown in FIG. 2(a) and FIG. 2(b), respectively. No peaks for Mg.sub.2Si are present indicating that the reaction had proceeded to completion (FIG. 2(a)). Other identifiable phases are Al with peaks at 38.4 (ill) and 44.7 degrees (200). Addition of HCl to the raw product results in the slow evolution of H.sub.2 gas, presumably from reaction with Al and its dissolution, with the solution color changing from black to brown over a period of a few hours. Is should he noted that the addition of HCl to raw products containing a significant fraction of residual Mg.sub.2Si results in a violent reaction, presumably from pyrophoric SiH.sub.4 formation and combustion. The XRD diffraction pattern for the purified product after vacuum filtration and washing is shown in FIG. 2(b). Peaks at 28.4, 47.2, and 56.4 degrees match the Si (111), (220), and (311) reflections, respectively. The average crystallite size, estimated using Scherrer's equation, is 35±3.5 nm (see, e.g., Scherrer, Gottingen, 2, 98, 1918).

[0157] Powder XRD patterns of Si/MGNS material are essentially identical to the sum of the patterns of the MGNS and separately synthesized Si, indicating that the synthesis was unaffected by the presence of the MGNS. The powder XRD pattern for the MGNS material and Si/MGNS composite material are shown in FIG. 3. The product yield was 97% based on initial mass of Mg.sub.2Si (this mass also includes any surface oxide that may have formed during purification).

BET Surface Area

[0158] The BET surface area was measured by N.sub.2 adsorption and found to be 115 m.sup.2 g.sup.−1, 60 m.sup.2 g.sup.−1, and 70 m.sup.2 g.sup.−1 for the Si.sub.nano, the MGNS and the Si/MGNS composite respectively. Each of the materials display type II isotherms (FIG. 4). Hysteresis is present in those of MGNS and Si/MGNS, indicating capillary condensation by the hollow MGNS, but absent in the solid Si.sub.nano alone. The surface area of the Si/MGNS is somewhat less than the summation of the surface areas of the Si.sub.nano and MGNS multiplied by their mass fractions (82 m.sup.2 g.sup.−1). This is consistent with the high surface area Si.sub.nano being in intimate contact with the MGNS, reducing N.sub.2 available surface area.

TEM and Energy Dispersive X-Ray Spectroscopy (EDX) Characterization

[0159] FIG. 5(a) shows the TEM micrograph of purified agglomerates of silicon nanocrystals (Si.sub.nano). FIG. 5(b) is a high-resolution TEM micrograph of a silicon nanocrystal showing lattice spacing characteristic of crystalline Si.

[0160] The TEM observations showed that the purified Si.sub.nano sample consists of agglomerates of spherical and irregularly shaped particles. The size of the particles ranges mostly from 30 to 60 nm in their largest dimension with some larger (˜100 nm) particles present (FIG. 5(a)). The image shows lattice fringes with 0.31 nm spacing matching the (111) planes of Si (FIG. 5(b)).

[0161] FIG. 6 is a high resolution TEM micrograph of a silicon nanoparticle. Higher magnification reveals that the particles consist of multiple nanocrystallites. 10 to 30 nm in diameter, that appear to be sintered (FIG. 6), FIG. 7 is an EDX HyperMap of Si.sub.nano. The composition was determined by EDX to be 91% Si, 7.5% O and 1.5% Al.

[0162] Without wishing to be bound by theory, the oxygen content found is an upper limit estimate, as it does not include any correction for adsorbed species (FIG. 7). The actual oxygen (O) content of the nanocrystals may likely be much less that 7.5%, as the lattice fringes of the silicon nanocrystals extend essentially to, or within a fraction of a nm of, the edge of the nanocrystals with no evidence of significant oxidation (FIG. 5(b)). Even so, the lower limit of the Si.sub.nano reaction yield, taking into account EDX oxygen content results, is 90%. The Al content may explain the lack of significant surface oxidation on the Si nanocrystals despite not using HF to remove SiO.sub.2 as is common in other studies and their repeated, and extensive, exposure to air. Al is not soluble in Si at the low temperatures used for the synthesis (see, e.g., Murray et al., Bulletin of Alloy Phase Diagrams, 5, 74, 1984). Thus, the Al is likely to have phase segregated to the surface of the Si nanocrystals and oxidized during air exposure and HCl washing, providing an effective passivation layer that prevents significant oxidation of the underlying Si, despite further exposure to air.

[0163] FIG. 8 and FIG. 9 shows the TEM and STEM micrographs, respectively, of the Si/MGNS composite. The Si/MGNS composite consists of intimately mixed agglomerates of silicon nanocrystals and MGNS. The MGNS are of irregular shape, appearing to readily distort from their idealized spherical shape. These shells were selected as the conductive carbon material for the Si/C composites in part due to the ability of their walls to flex, to act as “pillows”, relieving mechanical stress caused by the large volume expansion of Si during lithiation, maintaining contact and high conductivity.

[0164] FIG. 10 is an EDX HyperMap of the Si/MGNS composite, showing silicon localized on and around the MGNS, identified by correlation of the C intensity beyond that due to the carbon grid with the high-angle annular dark field (HAADF) image. The silicon to carbon ratio of the sample area was measured by EDX and found to be 1:1.375. This is in good agreement with the synthetic reagent stoichiometry and the mass ratio of silicon to 2.0 MGNS (1:1.5) determined by etching a known mass of the composite with 4 M KOH overnight, removing the silicon and leaving MGNS which was weighed after washing with water and drying, the difference between the pre and post etching mass being the mass of silicon.

Electrochemical Characterization

First Cycle Characterization

[0165] The first load/unload cycles of half-cells made with Si/MGNS composite electrodes were conducted at a low rate, 70 mA/g (179 mA/g Si, C/20), in attempt to “activate” all the c-Si before reaching the cutoff potential, as high current density can result in significant polarization.

[0166] Electrochemical cycling of c-Si is characterized by lithiation to form Li.sub.15Si.sub.4 below 50 mV and de-lithiation to amorphous Si (a-Si) above ˜420 mV, but these potentials depend on the applied current density (see, e.g., Obrovac et al., J. Electrochem, Soc., 154, A103, 2007). The onset of charge transfer to Si/MGNS electrodes occurred at ˜1.2 V due to the reduction of carbonated solvents forming a solid electrolyte interface (SEI) layer (FIG. 11(a)) (see, e.g., Aurbach et al., Electrochim. Acta, 50, 247, 2004).

[0167] Below 140 mV, a plateau in the galvanostatic discharge curve accompanied by a broad reductive peak on the differential capacity plot is observed due to lithiation of c-Si to form crystalline Li.sub.15Si.sub.4 (c-Li.sub.15Si.sub.4).

[0168] The voltage profile, in agreement with XRD and TEM data, indicates little or no amorphous silicon (a-Si) is initially present; a-Si lithiates over two sloping plateaus at higher potential (see, e.g., Obravac, 2007). MGNS is also lithiated; two reduction peaks at 80 and 25 mV, and reverse current complementary oxidation peaks at 100 and 120 mV associated with electrochemical Li storage in the MGNS are visible in the differential capacity curve (FIG. 11(b)).

[0169] The MGNS material, when prepared in a separate cell, has lithiation peaks at 90 mV and 60 mV and delithiation peaks at 100 mV and 120 mV, respectively (FIG. 12). The differences in the lithiation potentials when prepared in the composite may be a result of cell polarization. The majority of the charge transfer during the unloading step is observed at ˜420 mV, indicating that the c-Li.sub.15Si.sub.4 de-lithiated to a-Si rather than c-Si (see. e.g., Obravac, 2007).

[0170] The reversible capacity of the first cycle is 1382 mAh g.sup.−1 (for the Si/MNGS composite) (3454 mAh g.sup.−1 Si). This is 3.7 times the capacity of the theoretical capacity of graphite anodes. The difference in the unload potential of MGNS and silicon allows the contribution of the former to be estimated to be ˜100 mAh g.sup.−1 of the composite material by noting the fractions of full capacity at which the MGNS staging occurs (FIG. 13) and dividing the capacity at which they occur in Si/MGNS by those fractions. Taking the MGNS capacity into account, the capacity due to the Si.sub.nano is 3354 mAh g.sup.−1, ˜ 94% of the theoretical capacity of silicon, indicating a very high degree of utilization and low level of oxidation.

Subsequent Cycle Characterization

[0171] For cycles subsequent to the first, the electrodes were loaded/unloaded at either 286 mA g.sup.−1 (715 mA g.sup.−1 Si, C/5) or 715 mA g.sup.−1 (1.79 Ag.sup.−1 by Si, C/2). Increasing the rate from C/20 (1.sup.st cycle) to C/5 resulted in a modest (<6%) decrease in reversible capacity. Increasing the rate further to C/2 resulted in a larger (19%) decrease. Second cycle capacities at C/5 and C/2 were 1318 and 1114 mAh g.sup.−1 composite (3295 and 2785 mAh g.sup.−1 Si), respectively (see FIG. 14(a) and FIG. 12(a)). Lithiation at a C/5 rate during these cycles occurred at higher potentials than observed during the first, slow, lithiation, displaying the two sloping plateaus expected for a-Si (FIG. 14). Despite cycling below the Li.sub.15Si.sub.4 formation potential of ˜50 mV, charge transfer on de-lithiation indicates the presence of both a-Li.sub.xSi and Li.sub.15Si.sub.4, with two broad peaks from 200 to 600 mV in addition to the 420 mV peak. Polarization due to the high current density is likely the cause of the incomplete conversion to crystalline Li.sub.15Si.sub.4. Cycling performed at C/2 rate resulted in loading curves that are similar to those performed at C/5, indicative of a-Si lithiation (FIG. 15). Notably absent was a delithiation peak for Li.sub.15Si.sub.4, and only one delithiation peak for MGNS was observed as a result of greater polarization at this current density. The lithiation potentials progressively decrease with increasing cycle number, the effect being more acute at C/2, consistent with the gradual buildup of the SEI layer.

Capacity Retention and Coulombic Efficiency

[0172] As can be seen from FIG. 16, reversible capacity is very stable for the first 20 cycles but fades gradually upon subsequent cycling, achieving 1115 and 946 mAh g.sup.−1 composite (2788 and 2366 mAh g.sup.−1 Si) after 75 cycles for cells cycled at C/5 and C/2, respectively. The first cycle Coulombic efficiency of the cells was 78% and 75% (FIG. 17). Cells cycled at C/5 and C/2 achieved and maintained greater than 99% Coulombic efficiency. Less than 100% Coulombic efficiency is typical of Li alloying materials due to physical degradation of the SEI layer during expansion and contraction of the lithiation/delithiation cycle and is likely the cause of the capacity fade for these cells (see, e.g., Liu et al., Nano Lett., 12, 3315, 2012).

[0173] Cells prepared at higher mass loading (1.35 mg/cm.sup.2) showed similar electrochemical performance with only 5% lower gravimetric capacity at C/20, 7.5% lower capacity at C/5 rate and 6% higher capacity fade over 75 cycles (FIG. 18). The higher mass loaded cell had a higher first cycle Coulombic efficiency (81% vs 75%) and a similar long-term Coulombic efficiency of greater than 99% at 75 cycles (FIG. 19).

Comparative Cell by Mechanical Milling

[0174] Electrode active material prepared by mechanically milling Si.sub.nano with MGNS (Si+MGNS) performed significantly worse than electrode active material made by synthesizing Si.sub.nano in the presence of MGNS (Si/MGNS).

[0175] The initial reversible capacity at C/20 rate was 1050 mAh g.sup.−1 composite (2625 mAh g.sup.−1 by Si), 332 mAh g.sup.−1 less than displayed by Si/MGNS (FIG. 20(a) and FIG. 11(a)). The onset of c-Si+MGNS lithiation occurred at a potential closer to Li/Li.sup.+ (˜90 mV) than with Si/MGNS (˜120 mV). This potential decrease is likely due to inferior homogeneity and electrical contact of the Si material with the MGNS by mechanical milling, reducing conductivity within the anode. As a consequence, incomplete conversion to Li is Sty may have occurred, indicated by a less prominent plateau in the galvanostatic unload curve at 420 mV as well, as the presence of a broad peak from 200 to 400 mV consistent with delithiation of a-Li.sub.xsi (FIG. 11(b) and FIG. 20(b)).

[0176] Cycle 2 at a C/2 rate had a reversible capacity of 741 mAh g.sup.−1 (1854 mAh g.sup.−1 Si) that faded further until cycle 7, after which it began to gain capacity, possibly due to achieving better electrical contact and activation of additional Si as the electrode expanded and contracted (FIG. 16). Improvement in the electrical contact of the Si is indicated by the decrease in delithiation potential with increasing cycle number (FIG. 21). While the de-lithiation potential at cycle 50 and 75 occurred at similar values, the lithiation potential decreased, becoming closer to Li/Li.sup.+, consistent with an increase in impedance. Comparing the lithiation/delithiation curves of Si+MGNS (FIG. 21) to those of Si/MGNS (FIG. 14) indicates that the mixed material has inferior load/unload potentials with significantly larger hysteresis over all cycles. This is not surprising considering the Coulombic efficiency of Si with MGNS never exceeded 95.6%, consistent with a far greater degree of solid electrolyte interface (SEI) growth occurring each cycle than with Si/MGNS.

Impedance

[0177] Impedance spectra of the Si/MGNS and Si+MGNS (FIG. 22) were modeled to a modified Randles equivalent circuit composed of an element for the resistance of the electrolyte and cell components (R.sub.cell), RC elements composed of SEI resistance (R.sub.SEI), charge-transfer resistance (R.sub.CT) and constant phase elements (C.sub.SEI and C.sub.DL) for the SEI and charge transfer processes and a Warburg element (W.sub.diff) for solid state diffusion (FIG. 23).

[0178] Analysis of the spectra show that, following the initial cycle, R.sub.SEI and R.sub.CT of Si/MGNS are nearly identical during loading and unloading, and show little growth with extended cycling, consistent with the high Coulombic efficiency observed and indicating the formation of a stable SEI layer. In contrast, R.sub.SEI of Si+MGNS increases upon loading to a value that is approximately the same as for Si/MGNS, but then decreases significantly on fully unloading. Additionally, R.sub.CT of Si+MGNS increases during loading and retains its magnitude upon unloading, growing from cycle to cycle. This behavior and the low Coulombic efficiency observed is consistent with the formation of an unstable SEI, that grows during loading to become similar to that on Si/MGNS, but then is significantly disrupted during unload, providing more direct contact between the silicon and the electrolyte and thus lower R.sub.SEL but increasing Rex as the remnants of the SEI that have broken free of the Si restrict the porosity of the electrode. The difference in the behavior of the Si/MGNS and the Si+MGNS in this respect may be due to the better dispersion of the silicon in the former as observed in SEM, while larger agglomerates in the latter experience more mechanical degradation during the load/unload cycles.

Example 2

Synthesis of Si Nanomaterial by Mechanical Milling of Silicide and Aluminum Salt

[0179] In a nitrogen filled glove box (<0.1 ppm O.sub.2 and H.sub.2O) 1.5 mg of Mg.sub.2Si (99.9% Alfa Aesar) was combined with 4.0 g of anhydrous AlCl.sub.3 (99.9% Sigma Aldrich) in a 80 mL tungsten carbide cup with 4 tungsten carbide 1.5 cm balls. The cup was capped with a lid and a Viton Q-ring. Insulating tape was placed around the seal gap to provide additional protection from oxygen. The reactants were milled at 450 rpm for 18 hours. The cup was transferred back into the nitrogen filled glove box and the contents were transferred to a round bottom flask containing a stir bar and capped with a rubber septum. Under nitrogen gas purge, deionized water (purged with argon) was transferred to the round bottom flask via cannula to dissolve the unwanted salts. A 10% HCl solution was added to dissolve Al and any unreacted Mg.sub.2Si. The solution was stirred overnight, the Si product washed with distilled water until a neutral pH was obtained, then further dried under vacuum.

[0180] FIG. 24 shows the XRD pattern of the purified Si product obtained after the HCl wash, showing evidence of crystalline Si with no apparent impurities.

Electrode Preparation

[0181] 150 mg of the Si product were combined with 50 mg of Super C45 carbon in a 80 mL tempered steel cup and 6 1.5 cm tempered steel balls and mechanically mixed with a Fritsch Pulverisette 6 at 200 RPM for 30 minutes. The resulting powder was combined with 50 mg of a 1:1 wt. ratio of carboxymethyl cellulose and poly-acrylic acid then stirred in a 10% ethanol solution until a viscous slurry was achieved. The slurry was cast onto copper electrodes with a doctor blade and dried under vacuum at 150° C. for 2 hours. The electrode mass was 0.7 mg/cm.sup.2.

Electrochemical Cycling

[0182] Coin cells (CR2016, MTI Inc.) containing Si/C and Li metal (99.9%, MTI Inc.) electrodes separated by a polypropylene porous membrane (Celgard 3401) were assembled in an argon filled dry box (<0.1 ppm O.sub.2 and H.sub.2O). The electrolyte used was 1 M LiPF.sub.6 in EC:DMC (ethylene carbonate: dimethyl carbonate, 1:1 v/v) mixture (battery grade, <15 ppm H.sub.2O content, Sigma Aldrich) with 10% FEC (>99%, Solvay) by volume. Cells were cycled at constant, current (C/20) for one cycle between open circuit voltage to a cutoff potential then under reverse current to 1.2 V vs. Li/Li+. Subsequent cycling was performed at a C/4 rate, with cutoff potentials of 10 mV vs. Li/Li+.

[0183] FIG. 25 show's the first cycle galvanostatic plot for the Si/C composite.

[0184] FIG. 26 shows the cycle life of the Si/C composite cycle at C/4 rate.

Example 3

Synthesis of Si in the Presence of Graphite as a Support Material to Form a Si/Graphite Composite Material

[0185] A Si/graphite composite was prepared according to the procedure described in Example 1, with the exception that Timcal SFG-6™ graphite was used as the support instead of carbon nanospheres. The mass fraction of the Si in the composites was varied (10, 20 and 30 w/w % Si in the Si/graphite composites).

[0186] FIG. 27 shows the XRD pattern for the synthesize Si/graphite product, with the Si mass fractions of each composite indicated on the figure. FIG. 27 shows evidence of crystalline Si and graphite.

[0187] FIG. 28 shows the galvanostatic cycling profile of a Si/graphite product in which the mass fraction of Si in the composite is 10%. FIG. 28 shows voltage plateaus consistent with the insertion and de-insertion of Li from Si and graphite.

[0188] FIG. 29 shows the cycle life (black circles) and Coulombic efficiency (blue triangles) of Si/graphite product in which the mass traction of Si in the composite is 10%.

[0189] FIG. 30 shows the galvanostatic cycling profile of Si/graphite product in which the mass fraction of Si in the composite is 30%.

[0190] FIG. 31 show's the cycle life (black circles) and Coulombic efficiency (blue triangles) of Si/graphite product in which the mass fraction of Si in the composite is 30%.

Example 4

Synthesis of Si Nanoparticles in a Molten Salt.

[0191] Sodium chloride (2.0 g, ACS certified, Fisher Scientific) and potassium chloride (1 g, ACS certified, Fisher Scientific) were dried overnight at 300° C. and then added with aluminum chloride (7.6 g, anhydrous 99.985%, Alfa Aesar), and Mg.sub.2Si (0.24 g, 99.5% Alfa Aesar) to a 50 mL Pyrex round bottom flask containing a polytetrafluoroethylene coated stir bar. The flask was then heated to 150° C. resulting a molten salt mixture. The mixture was stirred for 2 hours and then allowed to cool and solidify, followed by the addition of hydrochloric acid (HCl, 100 ml, ACS grade 36.5 to 38%, VWR Analytical). Following a 3 hour treatment with HCl, the product (Si nanocrystals) was isolated by vacuum filtration and washed with deionized water.

[0192] FIG. 32 shows the XRD pattern for the nanoscale Si produced from Mg.sub.2Si in a molten salt matrix of NaCl, KCl and AlCl.sub.3 at 150° C., showing evidence of crystalline Si with no apparent impurities.

[0193] FIG. 33 shows a TEM image of the purified crystalline Si product produced from Mg.sub.2Si in a molten salt matrix of NaCl, KCl and AlCl.sub.3 at 150° C., showing the nanoscale dimensions of the particles.

[0194] FIG. 34 shows a TEM image of the purified product as in FIG. 33 but at higher resolution.

[0195] FIG. 35 show's a HR-TEM image of the purified product as in FIG. 33 and FIG. 34 at higher resolution, displaying visible lattice fringes consistent with crystalline Si with little or no oxidation despite its previous exposure to air and water.

[0196] FIG. 36 show's an EDS hypermap of the purified product highlighting the location of Si (upper right), O (lower left), and Al (lower right) in the TEM image (upper left). Spectral analysis found the overall composition to be 87% Si, 10% O, and 3% Al.

Example 5

Synthesis of Si Nanoparticles in Solution

[0197] In an inert atmosphere (N.sub.2) dry box, AlBr.sub.3 (2.5 g) was combined with Mg.sub.2Si (0.523 g), toluene (10 mL) and a polytetrafluoroethylene coated magnetic stir bar in a 50 mL dual neck round bottom flask. One of the flasks necks was then capped with a rubber septum and the other with a glass stopper. The flask was removed from the dry box and connected to a water cooled condenser attached to a Schlenk line, while purging the headspace of the flask with nitrogen using a needle inserted through the septum. The solution was brought to and held at reflux for 24 hours, then allowed to cool to room temperature. The resulting solid product was separated from the solution by centrifuge. 40 mL of acetone was then added to the solid and the product again separated by centrifuge. The solid product was combined with HCl solution (300 mL, 50% v/v) and left to sit at room temperature for 5 hours. The purified Si nanoparticle product was then isolated by vacuum filtration and washed with deionized water until a neutral pH was obtained.

[0198] FIG. 37 shows the XRD pattern for the nanoscale Si produced from AlBr.sub.3 and Mg.sub.2Si in refluxing toluene, showing evidence of crystalline Si with no apparent impurities.

Example 6

[0199] Synthesis of Hydrogen Terminated Highly Porous Amorphous Silicon Prior to mixing, NaCl (99.5%, Acros Organics) was dried at 300° C. for 12 hours and Mg.sub.2Si (99.99%, Alfa Aesar) was hand ground into a fine powder under an argon atmosphere. In an argon filled dry box (<1 ppm H.sub.2O and <1 ppm O.sub.2 content), 10.85 g AlCl.sub.3 (anhydrous 99.985%, Alfa Aesar), 0.367 g Mg.sub.2Si and 2.38 g NaCl were combined in an 80 mL tempered steel ball mill with ten 1 cm tempered steel balls (10:1 ball to powder mass ratio) and mixed at 150 rpm for 60 minutes.

[0200] In the argon filled dry box, the mixed powder was transferred to a 250 mL round bottom flask and equipped with a Teflon coated stir bar and capped with a rubber septum to prevent air exposure. Under flowing argon gas, the round bottom flask was partially submerged into a pre-heated oil bath at 125° C. Upon complete melting of the salt, the solution was stirred for an additional 10 minutes. The round bottom flask was removed from the oil bath and allowed to cool to room temperature. The flask was then submerged in an ice bath and 200 mL of degassed methanol was added and stirred until gas ceased to evolve and the solution turned dark red. 10 mL HCl (ACS grade 36.5 to 38%, VWR Analytical) was added and the mixture stirred for 15 minutes. The resulting product (hydrogen terminated highly porous amorphous silicon) was isolated by vacuum filtration, dispersed into 200 mL deionized water, agitated by sonication for 1 minute, isolated again by vacuum filtration and washed with 200 mL deionized water. The product was immediately dried under dynamic vacuum and transferred to an argon filled dry box.

[0201] The BET surface area of the resulting product was 769±3.5 m.sup.2/g.

[0202] The Barret-Joyner-Halenda (BJH) desorption cumulative pore volume of the resulting product was 0.66 cm.sup.3/g.

[0203] The average BJH desorption pore width of the resulting product was 4.3 nm.

[0204] FIG. 38 shows the adsorption (solid circles) and desorption (hollow circles) nitrogen isotherm of the hydrogen terminated silicon product.

[0205] FIG. 39 shows the BJH pore size distribution of the hydrogen terminated silicon product.

[0206] FIG. 40 show's the powder XRPD pattern of the hydrogen terminated silicon product. Three peaks due to the presence of a low percentage (<1%) of residual silicon impurity present in the Mg.sub.2Si starting material can be observed.

[0207] FIG. 41 shows the Infrared (IR) ATR spectrum of the hydrogen terminated silicon product. Peaks observed at approximately 2100 cm.sup.−1 (Si—H stretch); 1060 cm.sup.−1 (Q-Si-Q stretch), 910 cm.sup.−1 and a shoulder 660 cm.sup.−1 (Si—H.sub.x stretch), 620 cm.sup.−1 (Si—Si stretch) are all in agreement with literature results for hydrogen terminated highly porous amorphous silicon. See, e.g., FIG. 4.5 at page 83 of Becker, “Galvanic Porous Silicon: Processing and Characterization for Nanoenergetics,” Mechanical Engineering Graduate Theses & Dissertations, University of Colorado at Boulder, 2010.

[0208] FIG. 42 is an SEM image of the hydrogen terminated silicon product.

[0209] FIG. 43 is a focused ion beam (FIB) cross section of the purified hydrogen terminated silicon showing internal porosity (the lighter, upper half of image). The dark middle image is the bottom of ion etched well, and the light foreground is a surface of sample in which the well was etched.

[0210] FIG. 44 is a TEM image of the hydrogen terminated silicon product.

[0211] FIG. 45 is HRTEM image of the hydrogen terminated silicon product, showing its amorphous structure and porosity.

[0212] FIG. 46(a) shows a sample of the hydrogen terminated silicon product prior to oxidation and ignition.

[0213] FIG. 46(b) shows a sample of the hydrogen terminated highly porous amorphous silicon product after oxidation and ignition following addition of concentrated nitric acid.

[0214] FIG. 47(a) shows a sample of the hydrogen terminated highly porous amorphous silicon product prior to oxidation and ignition.

[0215] FIG. 47(b) shows a sample of the hydrogen terminated highly porous amorphous silicon product after oxidation and ignition following addition of concentrated nitric acid.