LOW TEMPERATURE ELECTROCHEMICAL PRODUCTION OF SILICON

Abstract

A method for the electrochemical production silicon comprises applying an electrical potential across an anode and a cathode to provide electrons at the cathode. The anode and the cathode are in contact with an electrolyte melt at a reaction temperature. The electrolyte melt comprises a molten salt or a mixture of molten salts; a silicon-containing precursor at least partially dissolved in the electrolyte melt to provide soluble silicon-containing ions in the electrolyte melt; and a supporting electrolyte at least partially dissolved in the electrolyte melt to provide O2− ions in the electrolyte melt. The soluble silicon-containing ions at the cathode undergo reduction reactions with the electrons to release O2− ions and deposit silicon on the cathode.

Description

BACKGROUND

Silicon is not only the foundational material for microelectronics, but also a key material in many renewable energy technology and chemical and metallurgical applications¹. The demand for silicon is increasing rapidly and for the solar energy industry alone, there is an estimated 30% increase in demand every year². As the second most abundant element in the earth's crust, silicon is naturally found in silica and silicates, which are the primary components of most rocks and sand³. Current industrial production of silicon is accomplished by the carbothermic reduction of SiO₂ by coke (elemental carbon or charcoal), which requires high operating temperatures (1700° C. or higher) in an electric furnace. The net reaction is:

SiO₂(s)+C (s)→Si (l)+CO₂(g)  (1)

While this approach is scalable, it has high energy consumption (over 20 kWh per kg silicon), very poor energy efficiency (less than 30%)⁴, and produces significant carbon emissions both from the direct reaction product of CO₂ as well as indirectly from CO₂ emissions produced by the generation of electricity required to power high temperature furnaces⁵. Additionally, subsequent purification steps based on hydrosilane and distillation via the Siemens process are needed to make high-purity silicon for applications⁶. Therefore, the primary reason that Si is expensive is the high cost and large energy consumption associated with its production. Thermal reduction with reactive metals such as magnesium (magnesiothermic reduction) can also reduce silica to silicon^7,8 but the active metal precursors used have to be produced by electrolysis processes as well.

SUMMARY

Provided are electrochemical methods for producing silicon. In one embodiment, a method for the electrochemical production silicon comprises applying an electrical potential across an anode and a cathode to provide electrons at the cathode. The anode and the cathode are in contact with an electrolyte melt at a reaction temperature. The electrolyte melt comprises a molten salt or a mixture of molten salts; a silicon-containing precursor at least partially dissolved in the electrolyte melt to provide soluble silicon-containing ions in the electrolyte melt; and a supporting electrolyte at least partially dissolved in the electrolyte melt to provide O²⁻ ions in the electrolyte melt. The soluble silicon-containing ions at the cathode undergo reduction reactions with the electrons to release O²⁻ ions and deposit silicon on the cathode.

Other principal features and advantages of the invention will become apparent to those skilled in the art upon review of the following drawings, the detailed description, and the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Illustrative embodiments of the invention will hereafter be described with reference to the accompanying drawings.

FIG. 1A is a schematic of an electrochemical cell for producing silicon according to an illustrative embodiment of the disclosed method. FIG. 1B is a schematic which illustrates the progression of experiments conducted in developing an illustrative embodiment of the disclosed method.

FIG. 2A shows cyclic voltammetric (CV) curves before the electrochemical reduction process from −2.0 to 2.0 V with different amounts of CaO added to molten CaCl₂. FIG. 2B shows current-time curves at a constant voltage of −1.6 V with different amounts of CaO.

FIG. 3A shows a photograph of the Si product formed from electrolysis in molten CaCl₂—NaCl—MgCl₂. FIG. 3B shows the CV curves before the electrochemical reduction process from −2.0 to 2.0 V when different molten salts of CaCl₂, CaCl₂—NaCl, CaCl₂—MgCl₂ were used. FIG. 3C shows current-time curves at a constant voltage of −1.6 V when different electrolytes were used.

FIGS. 4A-4G show the details of the electrochemical reduction synthesis of Si using ternary molten salts with the mass ratio of NaCl:CaCl₂:MgCl₂ at 2:4:1 and characterization of the silicon nanowire (NW) products. FIG. 4A shows CV graphs of the electrochemical reduction process from −2.0 V to 2.0 V at different times. FIG. 4B shows current curves during the electrochemical reduction process under a constant voltage of −1.5, −1.6, −1.8 and −2.0 V. FIG. 4C shows x-ray diffraction (XRD) patterns of the Si NWs synthesized using CaCl₂—NaCl—MgCl₂ melts at different voltages of −1.6, −1.8 and −2.0 V in comparison with the standard peaks from JCPDS 00-027-1402. FIGS. 4D and 4E show scanning electron microscope (SEM) images of the silicon NW product. FIG. 4F shows energy dispersive spectroscopy (EDS) analysis of a NW, confirming its predominant silicon content. FIG. 4G shows a high resolution transmission electron microscopy (HRTEM) image with a fast Fourier transform (FFT) inset of the Si NWs produced by using the optimized CaCl₂—NaCl—MgCl₂ molten salts at 650° C. at −1.6 V.

FIG. 5 shows a SEM image of the final product from a comparative experiment using SiO₂ powder as the precursor in CaCl₂ salt after a 3 h electrolysis under 1.6 V at 850° C. (with no supporting electrolyte).

FIG. 6 is a flow chart showing the steps of an illustrative embodiment of the disclosed method.

FIG. 7 shows a phase diagram of the ternary molten salt system (NaCl—CaCl₂—MgCl₂) (reprinted from Acta Chemcia Scandinavica 1972, 26, 1751).

DETAILED DESCRIPTION

Provided are methods for the electrochemical production of silicon. In a basic embodiment, a method for producing silicon comprises applying an electrical potential across an anode and a cathode, the anode and cathode in contact with an electrolyte melt. The electrolyte melt comprises a molten salt or a mixture of molten salts, a silicon-containing precursor and a supporting electrolyte. The applied electrical potential provides a flow of electrons from the anode to the cathode. These electrons reduce soluble silicon-containing ions in the electrolyte melt which originate from the silicon-containing precursor to release O²⁻ ions and deposit silicon on the cathode.

The disclosed method involves the direct reduction of the soluble silicon-containing ions, rather than the solid silicon-containing precursor compound itself. This is by contrast to conventional methods for the production of silicon involving the direct reduction of solid SiO₂ in a molten salt at 850° C. (SiO₂ (s)+4 e−→Si+2 O²⁻). (See Nohira, T., et al., “Pinpoint and bulk electrochemical reduction of insulating silicon dioxide to silicon,” Nature Materials, Vol. 2, June 2003, pages 397-401.) Although this conventional method allows for reduced reaction temperatures and reduced CO₂ emission, the complex electrolysis mechanism occurs at a three-phase interface between a working electrode, insulating silica, and a molten salt. This makes it difficult to achieve complete reduction of the silica and results in low yields of silicon. At least some embodiments of the disclosed method offer the advantages of low CO₂ emission, even lower reaction temperatures (e.g., 650° C. or less) and much higher yields of silicon (e.g., 20% or more) for a selected reaction time (e.g., about 3 hours). In addition, the disclosed method provides greater flexibility by allowing for a variety of silicon-containing precursors besides solid silica, thereby providing a practical, cost-effective pathway for the production of silicon from industrial waste products, including recycled glass, coal ash, and cinder ash.

The terms “melt” and “molten” are used herein in reference to materials or compositions which are at a temperature which is higher than their melting point such that they are in a liquid state. The temperature of the electrolyte melt may be referred to as the reaction temperature, i.e., the temperature at which the reduction reactions occur in the electrolyte melt.

A variety of molten salts, molten salt mixtures, silicon-containing precursors and supporting electrolytes may be used in the electrolyte melt. However, the choice of materials and relative amounts of these materials is selected, at least in part, to ensure that both the silicon-containing precursor and the supporting electrolyte are at least partially dissolved in the electrolyte melt at the selected reaction temperature. In embodiments, this reaction temperature is no more than about 900° C. This includes embodiments in which the reaction temperature is no more than about 850° C., no more than about 800° C., no more than about 750° C., no more than about 700° C., no more than about 650° C., no more than about 600° C., no more than about 550° C., no more than about 500° C., no more than about 450° C., no more than about 400° C., no more than about 350° C., or no more than about 300° C. This further includes embodiments in which the reaction temperature is in the range of from about 300° C. to about 700° C., from about 450° C. to about 700° C., or from about 500° C. to about 700° C.

The molten salt or molten salt mixture provides the molten electrolyte medium in which the electrochemical reduction of soluble silicon-containing ions occurs. The molten salt or molten salt mixture has a melting point which is lower than the reaction temperature to be used in the method. In embodiments, the molten salt or molten salt mixture has a melting point of no more than about 700° C., no more than about 650° C., no more than about 600° C., no more than about 550° C., no more than about 500° C., no more than about 450° C., no more than about 400° C., no more than about 350° C., no more than about 300° C., no more than about 250° C., or no more than about 200° C. This further includes embodiments in which the reaction temperature is in the range of from about 200° C. to about 700° C., from about 300° C. to about 600° C., or from about 400° C. to about 550° C. To facilitate dissolution of the silicon-containing precursor in the electrolyte melt, the molten salt or molten salt mixture may be selected to achieve a particular solubility (e.g., maximum solubility) of the silicon-containing precursor in the molten salt/molten salt mixture at the selected reaction temperature.

Selection of the molten salt/molten salt mixture having a certain melting point may be accomplished by consulting phase diagrams of various molten salts/molten salt mixtures or compiled lists of molten salt mixtures and their melting point data (e.g., Janz, G. K., et al., “Physical Properties Data Complications Relevant to Energy Storage. I. Molten Salts: Eutectic Data,” Nat. Stand. Ref. Data Ser., Nat. Bu. Stand. (U.S.), 61, Part I, 244 pages (Mar. 1978), which is hereby incorporated by reference in its entirety). Selection of the molten salt/molten salt mixture to achieve a certain solubility of silicon-containing precursor may be accomplished and measuring the solubility of the silicon-containing precursor in the molten salts/molten salt mixtures versus temperature as described in the Example, below.

Illustrative molten salts include metal halides and mixtures thereof. Specific, illustrative mixtures of metal halides include mixtures of two or more (i.e., binary or ternary mixtures) of CaCl₂, NaCl, and MgCl₂. As shown in FIG. 7, the ratios of these three metal halides may be adjusted to provide an eutectic mixture having a desired melting point. Table 1 in the Example, below, includes specific, illustrative eutectic mixtures based on different combinations of CaCl₂, NaCl, and MgCl₂. Other molten salt mixtures described in Janz, G. K., et al. (referenced above) may also be used. In embodiments, the molten salt mixture comprises two or more of the following metal halides: CaCl₂, BaCl₂, MgCl₂, SrCl₂, NaCl, KCl and LiCl. In embodiments, the molten salt mixture comprises at least one of CaCl₂, BaCl₂, MgCl₂, and SrCl₂ and one or more of another metal halide. In embodiments, the molten salt mixture comprises CaCl₂ and one or more of another metal halide. In embodiments, the molten salt mixture is selected from CaCl₂, KCl, and MgCl₂; CaCl₂, LiCl, and MgCl₂; CaCl₂, NaCl, and BaCl₂; CaCl₂, KCl, and BaCl₂; CaCl₂, LiCl, and BaCl₂; CaCl₂, NaCl, and LiCl; CaCl₂, NaCl, and KCl; CaCl₂, LiCl, and KCl; BaCl₂, NaCl, and LiCl; BaCl₂, NaCl, and KCl; BaCl₂, LiCl, and KCl. In each of these mixtures, the ratios of the metal halides may be adjusted to provide the desired melting point.

The electrolyte melt also comprises the silicon-containing precursor. The silicon-containing precursor is a compound which comprises silicon and oxygen and has sufficient solubility in the electrolyte melt at the reaction temperature such that it at least partially dissolves to provide the soluble silicon-containing ions. The silicon-containing precursor may be a silicate. The silicate may be a mineral. Many minerals are abundant, easily found and readily extractable, which provides a low cost and convenient silicon-containing precursor. The type of silicate is not particularly limited. However, in embodiments, the silicate is an orthosilicate, an inosilicate, or a phyllosilicate. Combinations of different silicates may be used.

In embodiments, the silicate has the formula MSiO₃ or M₂SiO₄, wherein M is a metal or a combination of different metals. In embodiments, the metal may be selected from alkali metals (e.g., Li, Na, K, Rb, Cs) and alkaline earth metals (e.g., Be, Mg, Ca, Sr, Ba). The dissolution of MSiO₃ silicates provides metal cations and soluble SiO₃²⁻ ions in the electrolyte melt. The dissolution of M₂SiO₄ silicates provides metal cations and soluble SiO₄⁴⁻ ions in the electrolyte melt. The Example below describes the use of the illustrative silicate CaSiO₃ which dissolves to provide Ca⁺ cations and soluble SiO₃²⁻ ions in an electrolyte melt. In embodiments, the silicate is an aluminosilicate. The aluminosilicate may be a clay mineral such as Kaolinite, Al₂Si₂O₅(OH)₄. Feldspars may be used, e.g., KAlSi₃O₈—NaAlSi₃O₈ CaAl₂Si₂O₈.

In other embodiments, the silicon-containing precursor is silica, SiO₂. As further described below, in the presence of the appropriate type and amount of supporting electrolyte, silica can effectively react with the supporting electrolyte and then dissolve in the electrolyte melt to provide the soluble silicon-containing ion. In other words, the soluble silicon-containing ion is created in situ via the formation of a corresponding silicate compound. Various amounts of the silicon-containing precursor may be used in the electrolyte melt. In embodiments, the amount is the maximum as determined by the solubility of the silicon-containing precursor in the molten salt/molten salt mixture at the selected reaction temperature.

The electrolyte melt also comprises the supporting electrolyte. The supporting electrolyte is a compound which comprises oxygen and has sufficient solubility in the electrolyte melt such that it at least partially dissolves to provide O²⁻ ions. The supporting electrolyte increases the ionic flux of O²⁻ ions in the electrolyte melt and improves the kinetics of the reduction reactions. A variety of alkali metal (e.g., Li, Na, K, Rb, Cs) oxides and alkaline earth metal (e.g., Be, Mg, Ca, Sr, Ba) oxides may be used as the supporting electrolyte. Combinations of different supporting electrolytes may be used. The Example below describes the use of the illustrative supporting electrolytes CaO and Na₂O in production of Si. Various amounts of the supporting electrolyte may be used in the electrolyte melt. The amount may be the amount at which the electrolyte melt is saturated with the supporting electrolyte(s) which provides the maximum concentration of O²⁻ ions at the selected reaction temperature. In embodiments, the electrolyte melt comprises at least 1 weight (wt) % supporting electrolyte, at least 1.5 wt %, at least 2 wt %, or at least 2.5 wt %. This includes embodiments in which the electrolyte melt comprises from about 0.1 wt % to about 15 wt % supporting electrolyte.

As noted above, the supporting electrolyte may also facilitate the dissolution of the silicon-containing precursor to provide the soluble silicon-containing ions. By way of illustration, at a sufficient concentration of CaO in an electrolyte melt comprising silica, the solid SiO₂ can effectively be dissolved to provide soluble SiO₃²⁻ ions in situ via the formation of CaSiO₃.

In embodiments, the silicon-containing precursor (or a portion thereof) and the supporting electrolyte (or a portion thereof) may be provided by a material which itself comprises these components. Such a material includes molten glass, molten coal ash, and molten cinder ash. By way of illustration, common glass is composed of SiO₂ (a suitable silicon-containing precursor) and CaO and Na₂O (suitable supporting electrolytes). Such a glass may be heated to above its melting temperature to provide molten glass in the electrolyte melt and if necessary, additional SiO₂ and/or supporting electrolyte may be added to facilitate dissolution of SiO₂ to provide soluble SiO3²⁻ ions in the electrolyte melt. Similarly, coal ash and cinder ash are also primarily composed of SiO₂, CaO, and in some cases, Na₂O. These embodiments demonstrate that the disclosed method can provide a sustainable, cost-effective, value-added pathway to the production of silicon from industrial and environmental wastes.

The electrolyte melts may be formed by combining the various ingredients (e.g., the molten salt/molten salt mixture and the silicon-containing precursor and the supporting electrolyte (either as individual components or as a waste product or both) and heating to liquefy the mixture. The electrolysis may be carried out in an electrochemical cell. A schematic of an illustrative electrochemical cell is shown in FIG. 1A. The material of the anode and the cathode is graphite. However, a variety of conducting materials may be used for the anode and cathode, including other carbon-containing materials. Also illustrated are the various chemical reactions relevant for an electrolyte melt based on a CaCl₂—NaCl—MgCl₂ molten salt mixture, a CaSiO₃ silicon-containing precursor, and a CaO supporting electrolyte. As described in the Example below, the applied electrical potential between the anode and the cathode may be adjusted to provide a desired yield of Si (e.g., maximum yield) and/or purity of Si (e.g., maximum purity). Similarly, the reaction time may be adjusted to achieve a desired yield of Si.

The disclosed method may be characterized by its yield of silicon. Yield may be reported as a percentage, i.e., (amount of recovered silicon)/(amount of silicon-containing precursor)*100. In embodiments, the yield is at least about 10%, at least about 20%, at least about 30%, at least about 40%, or at least about 50%. These yields may be reported at a particular reaction temperature and a particular reaction time, e.g., a yield of at least about 20% at a reaction temperature of about 650° C. and a reaction time of about 3 hours. These yields are significantly higher than those obtained using conventional methods such as those described in Nohira et al. (referenced above). This is shown via the comparative experiment using SiO₂ powder described in the Example below, which provided a yield of 6.9% at a reaction temperature of about 850° C. and a reaction time of about 3 hours. (See FIG. 5.)

The silicon deposited on the cathode may be removed from the electrolyte melt and cleaned using standard methods, including those described in the Example, below. The silicon produced by the disclosed method may be characterized by its crystallinity, its purity, and its morphology. The silicon is crystalline. The silicon product may include impurity crystal phases, but such phases are present in such a small amount that the silicon would be considered to phase pure. (See FIGS. 4C and 4G.) The silicon is of high purity. (See FIG. 4F.) In embodiments, the silicon is greater than 90 atomic % Si, greater than 95 atomic % Si, or greater than 98 atomic % Si. The silicon may be nanostructured or may comprise nanostructures, including nanostructures in the form of nanowires. (See FIGS. 4D and 4E). The average diameter of the nanowires may be in the range from about 10 nm to about 500 nm. The average length of the nanowires may be in the range of from about 1 μm to about 200 μm. The nanowire morphology is by contrast to silicon produced using conventional methods such as those described in Nohira et al. (referenced above). This is shown via the comparative experiment using SiO₂ powder described in the Example below, which produced micrometer sized particles. (See FIG. 5.)

Silicon produced using the disclosed method may be used in a variety of applications. The silicon nanowires may be used as materials in various energy storage and conversion devices such as batteries and photovoltaic cells. The Example below describes the use of the silicon nanowires as an anode in a lithium ion battery. The silicon nanowires may also be used as raw materials in a variety of metallurgical applications (e.g., steel production).

EXAMPLE

This Example describes the low temperature electrochemical production of silicon in bulk quantity via the reduction of soluble CaSiO₃ precursor by using a low-temperature melt, consisting of the ternary eutectic CaCl₂—MgCl₂—NaCl salts, and CaO as a supporting electrolyte. The engineered ternary molten salts lower the operating temperature, while maintaining the solubility of CaSiO₃, and CaO provides higher concentration of O²⁻ ions in the melt, facilitating the reaction kinetics. The synergistic effect of these advances enabled the high yield electrolytic synthesis of silicon nanowires (NWs) at a low reaction temperature of 650° C. As a demonstration, the produced silicon NWs were used as Li-ion battery anodes and showed excellent capacity and cycling performance.

Materials and Methods

All chemicals and reagents were purchased from Sigma-Aldrich and used as received unless noted otherwise.

Molten salt electrolysis. Different ratios of CaSiO₃, CaO (Na₂O), CaCl₂ (NaCl, MgCl₂) were added into a cylindrical alumina crucible (inner diameter: 4 cm, height: 15 cm) and heated to 200° C. for 24 h in a vacuum oven to remove moisture. Then, the alumina tube was transferred to a sealed fused silica reactor, which was heated by a tube furnace and under a continuous argon flow to the desired temperatures (850° C., 700° C., or 650° C.) for electrolysis. A schematic of the electrochemical cell is shown in FIG. 1. Two graphite rods (d=3 mm) serve as the anode and cathode, respectively. Constant-voltage electrolysis at 850, 700, 650° C. in various molten salt and CaO (Na₂O) compositions (see Table 1) was then carried out at −1.6 V for 3 h. As a comparison, constant-voltage electrolysis was also carried out under −1.8, −2.0 V in some molten salts at their respective temperatures. After electrolysis, the working electrode was pulled out of the melt and allowed to cool down to room temperature in Ar atmosphere. The product was washed with buffered HF (2 M), HCl (2 M), and the deionized water three times to remove the solidified salts and other impurities, and dried in an oven at 80° C. for 12 h. The complete process is illustrated in a flow chart in FIG. 6. All of the melts investigated as well as the reaction temperatures, reagent ratios, and product yields are summarized in Table 1.

Electrochemical reduction synthesis of Si from SiO₂ powder. An alumina tube crucible (inner diameter: 4 cm, height: 15 cm) containing 1.5 g of SiO₂ and 40 g of CaCl₂ was heated to 200° C. for 24 h in a vacuum oven to remove moisture. Then the alumina tube was setup in the same reactor described above and heated to the desired temperature 850° C. for constant-voltage electrolysis at −1.6 V for 3 h.

Materials characterization. The PXRD data were collected on as-synthesized samples on glass substrates using a Bruker D8 Advance powder X-ray diffractometer with Cu Ka radiation. The scanning electron microscopy (SEM) images were collected on a LEO SUPRA 55 VP field-emission scanning electron microscope operated at 5.0 kV. Energy-dispersive X-ray spectroscopy (EDX) was performed on single nanowires (NWs) transferred onto an aluminum foil using the same SEM equipped with an EDX detector operating at 10.0 kV. The transmission electron microscope (TEM) characterizations were performed using a Tecnai TF-30 microscope operating at an accelerating voltage of 300.0 kV.

Solubility tests. The solubility of CaSiO₃ in molten CaCl₂—NaCl, MgCl₂—NaCl, and CaCl₂—MgCl₂—NaCl salts (40 g in total) with 0.5 g of CaO as the supporting electrolyte at different temperatures was measured by a series of static dissolution experiments at 450, 550, 650, and 750° C. CaSiO₃ powder was pressed into pellets. Each pellet was weighed and then wrapped by foamed copper foil and Ni wire. The initial masses of the CaSiO₃ pellets were denoted as A. The wrapped CaSiO₃ pellets were immersed in the molten salts for 2 h at different temperatures and then taken out from the melt. After being cleaned by distilled water, the residual CaSiO₃ was weighed again, and the masses were denoted as B. Correspondingly, the solubility (S) of CaSiO₃ in the salt at different temperatures was calculated by the following equation:

$S=\frac{B-A}{40g}\times 100\%$

Lithium-ion battery assembly and measurements. The silicon NWs were used as active materials for lithium batteries and the electrodes were prepared as follows: A 6.5 wt % solution of polyvinylidene difluoride (PVDF) binder in N-Methyl-2-pyrrolidone (NMP) was prepared. A slurry consisting of Si NWs, carbon black, and PVDF in a ratio of 5:3:2 wt % was pasted on a smooth Cu foil (18 μm thickness) and dried at 80° C. in an vaccum oven overnight. The mass loading of the electrode was 1.1 mg cm⁻². The electrodes were packed into CR2016-type coin cells in an argon-filled glovebox with Li metal as the counter electrode, 1 M LiPF₆ in EC/DMC (1/1 by volume, BASF) as the electrolyte, and Celgard membrane as the separator. Galvanostatic cycling experiments were performed using a Biologic SP-200 Potentiostat controlled by EC-Lab software.

Results and Discussion

As noted above, the electrochemical reduction reactions were carried out using the apparatus shown in FIG. 1A. A symmetrical two electrode configuration was used with individual graphite rods acting as the cathode and anode. As a comparison, the electrochemical reduction of SiO₂ powder in a CaCl₂ melt was conducted at 850° C. (SiO₂ +4 e⁻→Si+2 O²) at a constant voltage of −1.6 V following earlier reports⁴. However, the reaction produced only a small quantity (˜25 mg) of micrometer silicon particles after electrolysis for 3 h (see FIG. 5).

FIG. 1B illustrates the progression of subsequent experiments. First, soluble calcium silicate (CaSiO₃) was used instead of the sparingly soluble SiO₂ powder as the silicon source. CaSiO₃ is an abundant and inexpensive silicate mineral³. The reaction preparation process and the process flow to separate the electrolysis product formed on the graphite cathode from the molten salts after the reaction is described in detail above and is illustrated in FIG. 6. After an identical electrolysis reaction at 850° C. for 3 h at −1.6 V, the total mass of the deposited Si products from CaSiO₃ on the graphite rod doubled (55 mg). Unlike previous reports on solid-solid electrochemical reduction of SiO₂^12,13,20,21, in the melt, CaSiO₃ dissolves to generate Ca²⁺ and SiO₃²⁻ ions (CaSiO₃(s)→Ca²⁺+SiO₃²). The relevant reactions are shown in FIG. 1A. On the surface of cathode, SiO₃²⁻ is reduced to Si under constant influx of electrons and generates O²⁻ ions through the following reaction:

SiO₃²⁻+4 e⁻→Si(s)+3 O²⁻ (cathode)

The ionized oxygen then diffuses across the melt driven by the constant potential to the graphite anode, where the oxidation happens as follows:

C(s)+O²⁻→CO₂+4 e⁻ (anode)

Overall, O²⁻ anions are the limiting charge carriers in the cell and drive the electrochemical reaction from cathode to anode all the way throughout reaction. It is thought that the diffusion of oxygen ions and the solubility of CaSiO₃ in the molten salts are two important factors that govern the reduction rate of the electrolysis.

Next, the concentration of the dissolved oxygen ions in molten electrolytes was increased by adding CaO as a supporting electrolyte (FIG. 1A). In the molten salt, CaO will dissociate according to the following reaction:

CaO→Ca²⁺+O²⁻

The dissociation of CaO creates more O²⁻ anions. Although CaO is not directly involved in the net electrochemical reduction of SiO₃²⁻, it is hypothesized that a higher concentration of CaO could increase the ionic flux of the O²⁻ ions in the melt and improve the overall kinetics of the reaction. Initially, 0.5 g of CaO was added to the CaCl₂ melt (40 g) which means that CaO is 1.25% by weight in the electrolyte. This CaCl₂—CaO electrolyte system yielded significantly more Si product from 55 to 135 mg after a 3 h electrolysis at 850° C. The impact of CaO is also apparent in the cyclic voltammetric (CV) curves in the various CaCl₂ melts with or without CaO (FIG. 2A). For the CaCl₂ melt without CaO, two pairs of pronounced CV peaks are obvserved at aproxmately −1.56/1.48 V (denoted as A1/C1) and −0.35/0.25 (denoted as A2/C2). As the reaction is configured as a symmetrical two-electrode system, the anodic and cathodic peaks in each pair of CV peaks are based on the same redox reactions. The inital peaks, A1 and C1, are attributed to the reduction of the dissolved silicates to form silicon. The second pair, A2 and C2, are attributed to the formation and dissolution of Ca²². The addition of CaO clearly significantly increased the A1/C1 peak area while the potential difference between two peaks slightly decreased. This indicates that the addition of CaO also makes the electrochemical reduction of silicate slightly more energetically favorable. However, when the amount of CaO was increased to 1.0 g (2.50% by weight), the product yield barely increased further (to 137 mg). The CV curve was nearly identical to that with 1.25% CaO with only a small increase in A2/C2 peak area. This suggests that the O²⁻ anions were likely saturated in molten CaCl₂. This value is lower than the reported solubility of CaO in melt CaCl₂ (13% by weight) at 850° C.²³, because the dissolved CaSiO₃ can decrease the solubility of CaO in the same molten CaCl₂ solution. Therefore, any additional CaO would result in little change in the rate of the reaction. This is also confirmed by the current-time curves at constant voltage of −1.6 V (FIG. 2B). When 1.25% and 2.5% CaO were added, the current remained saturated at around −20 mA, while the reaction without CaO stablilized at a current of around −10 mA.

After the important role of CaO was revealed, the electrolysis temperature was lowered by finding suitable eutectic molten salts with lower melting temperatures than the melting point of pure CaCl₂ (782° C.) while fixing the supporting electrolyte CaO at 0.5 g. NaCl is a common salt that can decrease the melting temperature as well as the viscosity of electrolyte. The lowest melting point of CaCl₂—NaCl eutectic system reaches 601° C. Indeed, the reaction temperature could be lowered to 700° C. for the CaCl₂—NaCl eutectic systems (see Table 1 for details of the molten salt compositions). However, after a 3 h electrolysis at −1.6 V, the best product yield was significantly lower at merely 15 mg. Thus, MgCl₂ was selected. This salt can also decrease the overall eutectic temperature based on the phase diagram but has similar properties to CaCl₂²³. In the CaCl₂—MgCl₂ melts, at the same reaction temperature of 700° C., the product yield reached 60 mg. It was hypothesized that the product yield was correlated with the solubility of CaSiO₃ in the molten salts, with higher solubility providing higher yield. Thus, solubility tests in CaCl₂—NaCl and CaCl₂—MgCl₂ were carried out. The solubility of CaSiO₃ reaches up to 0.8 wt % at 700° C. in CaCl₂—MgCl₂ mixture, however, it remains marginal (below 0.2 wt %) in CaCl₂—NaCl. Thus, although NaCl can lower the melting temperature, the solubility of CaSiO₃ is too low to realize low-temperature electrolysis.

In order to achieve both low melting point and high CaSiO₃ solubility to realize low-temperature electrolysis in high yield, the ternary eutectic melts consisting of CaCl₂—MgCl₂—NaC were investigated. A previous report²⁴ has shown that different salt ratios vary the chemical potentials of the molten salts, thus creating a molten salt eutectic and overall lowering the melting temperature of the system. The melting point of the eutectic mixture with the mass ratio of NaCl:CaCl₂:MgCl₂ at 2:4:1 decreases to 424° C. according to the phase diagram in Ref 24. The solubility of CaSiO₃ in this melt was also evaluated. Solubility increases with temperature between 450 to 750° C. Although the solubility remains small (less than 0.2 wt %) below 550° C., the solubility increases sharply between 550 to 650° C., at which point the solubility is about 1.0 wt % and then further rises to 1.2 wt % at 750° C. Therefore, a reaction temperature of 650° C. was selected, which is lower than that of the binary melts yet maintains sufficient CaSiO₃ solubility, thus allowing for fast electrolysis rate and a good yield. After further experiments (summarized in Table 1), the electrolysis in the most optimized CaCl₂—NaCl—MgCl₂ tenary melt (with mass ratios of 4:2:1) at 650° C. provided a Si yield of 75 mg (FIG. 3A), the highest yield among those low-temperature electrolysis reactions.

TABLE S1
A summary of the details of the reaction conditions (reaction temperature, electrolysis voltage)
using different molten salts (CaCl2, NaCl, MgCl2) and supporting electrolytes (CaO, Na2O) in
various ratios and the yield of the silicon product after a 3 h electrolysis reaction.
	CaSiO3	CaO	CaCl2	NaCl	MgCl2	Na2O	Temp	Voltage	Product
	(g)	(g)	(g)	(g)	(g)	(g)	(° C.)	(V)	Mass (mg)	Salts
1	1.5	—*	40	—	—	—	900	−1.6	55	CaCl2
2	1.5	0.5	40	—	—	—	850	−1.6	135	CaO+
3	1.5	0.5	40	—	—	—	850	−1.8	80	CaCl2
4	1.5	0.5	40	—	—	—	850	−2.0	50
5	1.5	0.5	35	5	—	—	800	−1.6	5	CaO+
6	1.5	0.5	30	10	—	—	700	−1.6	10	CaCl2 + NaCl
7	1.5	0.5	30	10	—	—	700	−1.8	10
8	1.5	0.5	30	10	—	—	700	−2.0	10
9	1.5	0.5	30	10	—	—	700	−2.2	15
10	1.5	0.5	25	15	—	—	700	−1.6	10
11	1.5	0.5	20	20	—	—	700	−1.6	—
12	1.5	0.5	15	25	—	—	700	−1.6	—	CaO+
13	1.5	0.5	10	30	—	—	650	−1.6	—	CaCl2 + MgCl2
14	1.5	0.5	—	40	—	—	600	−1.6	—
15	1.5	0.5	35	—	5	—	750	−1.6	40
16	1.5	0.5	30	—	10	—	700	−1.6	60
17	1.5	0.5	25	—	15	—	700	−1.6	53
18	1.5	0.5	20	—	20	—	700	−1.6	10
19	1.5	0.5	15	—	25	—	700	−1.6	—
20	1.5	0.5	10	—	30	—	700	−1.6	—
21	1.5	0.5	—	—	40	—	700	−1.6	—
22	1.5	0.5	15	15	15	—	650	−1.6	—	CaO+
23	1.5	0.5	20	10	10	—	650	−1.6	35	CaCl2 + NaCl+
24	1.5	0.5	24	8	8	—	650	−1.6	40	MgCl2
25	1.5	0.5	16	16	8	—	650	−1.6	—
26	1.5	0.5	24	16	8	—	650	−1.6	55
27	1.5	0.5	24	16	6	—	650	−1.6	20
28	1.5	0.5	24	11	6	—	650	−1.6	35
29	1.5	0.5	24	12	6	—	650	−1.6	75
30	1.5	0.5	24	13	6	—	650	−1.6	45
31	1.5	0.5	24	12	5	—	650	−1.6	15
32	1.5	0.5	24	12	4	—	650	−1.6	14
33	1.5	0.5	24	12	6	—	650	−1.8	45
34	1.5	0.5	24	12	6	—	650	−2.0	20
35	1.5	0.5	24	12	6	—	650	−1.5	12
36	1.5	0.5	24	12	8	—	650	−1.6	12
37	1.5	0.5	16	8	16	—	650	−1.6	—
38	1.5	0.5	18	6	12	—	650	−1.6	—
39	1.5	0.5	24	6	12	—	650	−1.6	—
40	1.5	0.5	10	10	20	—	650	−1.6	—
41	1.5	0.5	10	20	10	—	650	−1.6	—
42	1.5	0.4	24	12	6	0.11	650	−1.6	80	CaO+
43	1.5	0.3	24	12	6	0.22	650	−1.6	75	Na2O + CaCl2+
44	1.5	0.2	24	12	6	0.33	650	−1.6	45	NaCl + MgCl2
45	1.5	0.1	24	12	6	0.44	650	−1.6	23
46	1.5	—	24	12	6	0.55	650	−1.6	15
*means that the salt is not added or no silicon is produced.

The different melts were further investigated through electrochemical characterization. FIG. 3B shows the CV curves of CaSiO₃ electrolysis in different salt systems. Initially, both CV curves of CaCl₂—NaCl and CaCl₂—MgCl₂ systems at 700° C. exhibited small A1/C1 redox peaks and sharp A2/C2 peaks, which suggest that SiO₃²⁻ reduction is mostly prohibited and the formation and re-dissolution of Ca dominates in these systems. The low solubility of CaSiO₃ likely leads to slow reaction rate and more side reaction of Ca formation, which results in a liquid Ca layer deposited on the graphite cathode detrimental to the electrochemical reduction of SiO₃²². However, the optimized tenary molten salts display CV curves with much higher A1/C1 peak current of over 120 mA, which is almost 3 times higher than that of other two systems, indicating the most favorable reaction kinetics. This is also consistent with the results of current-time curves at a constant voltage of −1.6 V when different electrolytes were used (FIG. 3C).

The rection mechansims during electrolysis in the ternary molten salts were further investigated. The CV curves at different reaction times were acquired (FIG. 4A). After electrolysis for 1 h, the A1/C1 peaks became much more pronouced, compared to the peaks before electrolysis. In addition, larger overpotential was also observed, which favors the reduction of SiO₃²⁻ at the cathode (SiO₃²⁻+4 e⁻→Si(s)+3 O²) and faciliates the dissolution of CaO and CaSiO₃. This is believed to originate from the activation process, during which more ionized O²⁻ was generated in the melt due to the high cathodic overpotential. After electrolysis for 3 h, the CV curve barely changed from that after 1 h, suggesting that the electrolysis process has been stablized after the initial activation step and steady state O²⁻ migration in the melt was achieved.

To determine the proper voltage required for the electrochemical reduction at the highest rate with the purest product, the reaction was studied under various applied constant voltages (FIG. 4B). Under −1.6 V, despite the initial polarization, the galvanostatic current was the most stable and remains the highest after 3 h, indicating the most stable Si deposition. When −1.8 V was applied, the initial current trend was almost the same as that under −1.6 V. However, the current decreased more rapidly after the intial drop. Under a voltage of −2.0 V, the current was higher for the initial 0.5 h, but then slowly approached ˜−4 mA, indicating the reaction has stopped. When electrolysis took place at a lower voltage of −1.5 V, the current also quickly droped to ˜−2 mA after the initial polarization. These results showed that −1.6 V is the optimal voltage for the stable electrodeposition of Si in CaCl₂—NaCl—MgCl₂ ternary molten salts with the supporting electrolyte of CaO at 650° C.

The Si produced on the graphite working electrode (cathode) was then separated from the solidified molten salts and characterized (see Methods for details) to study its structure and morphology. The powder X-ray diffraction (PXRD) patterns of the samples produced under different applied voltages of −1.6, −1.8, and −2.0 V (FIG. 4C) were almost identical and the three pronounced peaks could be indexed to the cubic crystal structure of silicon (space group Fd-3m, a=0.54305 nm, JCPDS: 00-027-1402). The sharp diffraction peaks indicated the high crystallinity of the products. The products from other reactions in different electrolytes were also well-crystallized Si (data not shown). Scanning electron microscopy (SEM) images revealed nanowire morphology (see FIGS. 4D and 4E). SEM images of the Si NWs synthesized at the voltage −1.8 and −2.0 V were also obtained (data not shown). Energy dispersive spectroscopy (EDS) analysis (FIG. 4F) confirmed that the NWs consisted mainly of silicon. The oxygen was likely from the formation of a thin native oxide layer on the surface. High resolution transmission electron microscopy (HRTEM) image (FIG. 4G) highlights the single crystalline nature of the Si NWs. The cubic Si structure was confirmed by indexing the fast Fourier transform (FFT) of the HRTEM image (FIG. 4G, inset). The observed plane spacings of 0.316 and 0.274 nm correspond well to the expected values for the (111) and (200) lattice planes of Si, respectively. The forbidden (200) lattice plane absent in the PXRD patterns was still observed in the TEM due to dynamic double diffraction of the incident electron beam. It is interesting that the reduced silicon products are mostly of the NW morphology. Without wishing to be bound to any particular theory, it is hypothesized that the NW morphology may be due to metal impurities that could act as catalytic sites essential for the continuous anisotropic growth of silicon²¹. The mechansim of this electrolysis reaction driven by anion diffusion between electrodes could also promote the anistropic growth.

Other soluble Si precursors may be used for the low temperature electrolytic systhesis of silicon. For example, common glass is primarily composed of SiO₂, CaO, and Na₂O, and therefore, recyled glass waste can be used to provide the Si precursor. Therefore, electrolysis reactions were carried out with the addition of Na₂O under the same reaction temperature of 650° C. (details of the reaction conditions summarized in Table 1). The morphology of the product was Si NWs and PXRD confirmed that the phase was also mainly Si although some impurities were present (data not shown). In fact, soluble Na₂O could also provide O²⁻ ions in the molten salts just as CaO does, therefore, replacing up to about 40 wt % of CaO with Na₂O in the melt mixture did not seem to affect the reactions (see Table 1). The highest product yield was 80 mg when an optimal combination of 0.40 g CaO and 0.11 g Na₂O was used, which is even slightly higher than the yield from the optimized molten salt mixtures with just CaO supporting electrolyte. These results establish the feasibility for recycling glass waste, or even coal ash (whose major components are also SiO₂ and CaO²⁵), to produce value-added silicon nanomaterials using this molten salt electrolysis method.

These results showed that the low temperature electrochemical production of Si in high yield is enabled by three key advances: i) the use of the soluble CaSiO₃ as the silicon precursor to enable the production of silicon in bulk quantity, ii) the enhanced diffusion of O²⁻ ions due to the introduction of the CaO (or Na₂O) supporting electrolyte and thus higher concentration of O²⁻ ions in the molten salts, and iii), the design of the low melting point ternary CaCl₂—NaCl—MgCl₂ eutectic molten salts that retain sufficiently high solubility of CaSiO₃ precursor. In addition, if steady state concentration of CaO in the molten salts is maintained, SiO₂ can also be dissolved into the melts through the formation of CaSiO₃. This successful high yield electrolysis synthesis of silicon from the abundant and inexpensive CaSiO₃ precursor at a much lower temperature of 650° C. is more sustainable and practically significant for several reasons. First, the low reaction temperature of 650° C. can enable the recycling of residual medium-grade waste heat energy in various industrial and energy technologies and thus improve the overall cycle and power plant efficiencies¹⁹. Second, because metal halide molten salts in the temperature range of 600-700° C. have been successfully ultized in the molten salt nuclear ractors and concentrated solar thermal energy conversion technologies as heat exchange medium^18,26, the material compatibility and other practical issues have already been addressed at the industrial scale. Furthermore, this also suggests a sustainable pathway to recycle glass waste or coal ash to produce value-added silicon nanomaterials using molten salt electrolysis.

Finally, the electrolytically produced silicon NWs in bulk quantity are useful materials for high-capacity lithium battery anodes. Silicon has a very large theoretical energy storage capacity (˜4200 mAh g⁻¹) as a Li-ion anode material, but it suffers from the huge volume expansion and pulverization during lithiation and delithiation. Nanostructured silicon materials, such as Si NWs, can better accommodate the volume expansion involved in the Li battery cycling than bulk silicon^27,28. Therefore, using the Si NWs produced directly by this electrochemical reduction method in Li batteries could improve the total capacity of the anode and the cycle lifetime. The electrochemical performances of Si NWs were investigated in CR2016 coin cells with lithium plate counter electrode (see Methods for details of battery assembly and test). Note that bulk quantity of Si NWs was used at a mass loading of 1.1 mg cm⁻² and the electrode coating was estimated to be about 100 μm thick. The charge-discharge profiles of Si NWs at current densities from C/40 to 1 C over a potential window of 0.01 to 2.0 V for the first cycle were obtained (data not shown). The specific discharge capacity at C/40 was 3333 mAh g⁻¹, which is close to the theoretical capacity of ˜4200 mAh g⁻¹ for silicon. The discharge curve exhibited an obvious voltage plateau consistent with previous studies corresponding to Si alloying with Li to form Li_4.4Si²⁹. There was an obvious drop in the capacity during the 2^nd cycle to 2468 mAh g⁻¹, which is mainly attributed to the formation of solid electrolyte interface and other side reactions. The cycling performance of these Si NWs was also plotted (data not shown). Under C/2 rate, the initial discharge capacity was 2865 mAh g⁻¹. After initial loss through the first 100 cycles, the discharge capacity began to stabilize. After 500 cycles, the discharge capacity still reached 714 mAh g⁻¹, retaining 76% of the capacity at the 100^th cycle, indicating stable cycling performance following the 100^th cycle. When evaluated at progressively increasing current densities (from C/40 to 1 C), the Si NWs exhibited reasonably good cycling performance at various current rates (data not shown). After 10 cycles at C/40, the capacity was 2310 mAh g⁻¹. Remarkably, even at a high current density of 1 C, the capacity still reached 521 mAh g⁻¹, which is still higher than the commercial graphite anode (372 mAh g⁻¹). When the current density returned to C/40, the Si NWs could achieve a high capacity of 2307 mAh g⁻¹ again, almost equal to the initial capacity at C/40 before high rate cycling. These results demonstrated the excellent performance of these electrochemically synthesized Si NWs as high-capacity Li-ion battery anode materials without further elaborate engineering³⁰.

Conclusions

A novel molten salt electrolysis method was developed to produce silicon NWs from inexpensive soluble CaSiO₃ precursor at a low temperature of 650° C. by designing a ternary molten salt system to decrease operating temperature and introducing a CaO supporting electrolyte to provide a high concentration and facilitate transport of the oxygen ions. The enhanced kinetics, together with an optimized ternary CaCl₂—NaCl—MgCl₂ molten salt mixture with low eutectic melting point but retaining high solubility of CaSiO₃, enabled the low temperature electrochemical reduction at 650° C. for effective Si production in high yield. The as-synthesized Si NWs demonstrate excellent electrochemical performance as Li-ion battery anode materials. The method provides a new route for more practical and sustainable direct production of silicon at an industrial scale that is compatible with existing industrial processes and can be integrated with medium waste heat recycling. It also provides general new insights to molten salt electrolysis reactions that could help to improve other important electrolytic metal extraction processes¹¹, and further supports the use of recycling glass waste and coal ash to produce value-added materials in an environmentally friendly way. Especially as the electricity from renewable energy sources (solar, wind, hydroelectric, etc.) become more available and less expensive, using electricity to produce silicon at the industrial scale will become more attractive than the existing carbothermic process, not only because of the lower cost but also the reduced CO₂ emission.

REFERENCES

1. Francesco Priolo, Tom Gregorkiewicz, Matteo Galli & Krauss, T. F. Silicon nanostructures for photonics and photovoltaics. Nature Nanotechnology 9, 19, (2014).2. Yue, D., You, F. & Darling, S. B. Domestic and overseas manufacturing scenarios of silicon-based photovoltaics: Life cycle energy and environmental comparative analysis. Solar Energy 105, 669-678, (2014).

3. N. N. Greenwood & Earnshaw, A. Chemistry of the Elements. (Butterworth-Heinemann, December 9, 1997).

4. Xiao, W., Jin, X., Deng, Y., Wang, D. & Chen, G. Z. Rationalisation and optimisation of solid state electro-reduction of SiO₂ to Si in molten CaCl₂ in accordance with dynamic three-phase interlines based voltammetry. Journal of Electroanalytical Chemistry 639, 130-140, (2010).5. Yin, H. et al. Capture and electrochemical conversion of CO₂ to value-added carbon and oxygen by molten salt electrolysis. Energy & Environmental Science 6, 1538, (2013).6. Yasuda, K., Morita, K. & Okabe, T. H. Processes for Production of Solar-Grade Silicon Using Hydrogen Reduction and/or Thermal Decomposition. Energy Technology 2, 141-154, (2014).7. Bao, Z. et al. Chemical reduction of three-dimensional silica micro-assemblies into microporous silicon replicas. Nature 446, 172-175, (2007).8. Luo, W. et al. Efficient Fabrication of Nanoporous Si and Si/Ge Enabled by a Heat Scavenger in Magnesiothermic Reactions. Scientific Reports 3, (2013).9. Norgate, T. E., Jahanshahi, S. & Rankin, W. J. Assessing the environmental impact of metal production processes. Journal of Cleaner Production 15, 838-848, (2007).10. Chen, G. Z., Fray, D. J. & Farthing, T. W. Direct electrochemical reduction of titanium dioxide to titanium in molten calcium chloride. Nature 407, 361-364, (2000).11. Abdelkader, A. M., Kilby, K. T., Cox, A. & Fray, D. J. DC Voltammetry of Electro-deoxidation of Solid Oxides. Chemical Reviews 113, 2863-2886, (2013).12. Nohira, T., Yasuda, K. & Ito, Y. Pinpoint and Bulk Electrochemical Reduction of Insulating Silicon Dioxide to Silicon. Nature Materials 2, 397-401, (2003).13. Jin, X., Gao, P., Wang, D., Hu, X. & Chen, G. Z. Electrochemical Preparation of Silicon and Its Alloys from Solid Oxides in Molten Calcium Chloride. Angewandte Chemie 116, 751-754, (2004).14. Yin, H., Chung, B. & Sadoway, D. R. Electrolysis of a Molten Semiconductor. Nature Communications 7, 12584, (2016).15. Allanore, A., Yin, L. & Sadoway, D. R. A New Anode Material For Oxygen Evolution In Molten Oxide Electrolysis. Nature 497, 353-356, (2013).16. Wang, D., Jin, X. & Chen, G. Z. Solid state reactions: an electrochemical approach in molten salts. Annual Reports Section “C” (Physical Chemistry) 104, 189, (2008).17. Xiao, W. et al. Verification and implications of the dissolution-electrodeposition process during the electro-reduction of solid silica in molten CaCl₂. RSC Advances 2, 7588, (2012).

18. Ramesh K. Shah & Sekulic, D. P. Fundamentals of Heat Exchanger Design. (John Wiley & Sons, 2003).

19. Phillip, W. A. Thermal-energy conversion: Under pressure. Nature Energy 1, 16101, (2016).20. Xiao, W., Zhou, J., Yu, L., Wang, D. & Lou, X. W. D. Electrolytic Formation of Crystalline Silicon/Germanium Alloy Nanotubes and Hollow Particles with Enhanced Lithium-Storage Properties. Angewandte Chemie International Edition 55, 7427-7431, (2016).21. Cho, S. K., Fan, F. R. F. & Bard, A. J. Electrodeposition of Crystalline and Photoactive Silicon Directly from Silicon Dioxide Nanoparticles in Molten CaCl₂. Angewandte Chemie International Edition 51, 12740-12744, (2012).22. Xiao, W. & Wang, D. The electrochemical reduction processes of solid compounds in high temperature molten salts. Chemical Society Reviews 43, 3215, (2014).23. Wang, S., Zhang, F., Liu, X. & Zhang, L. CaO solubility and activity coefficient in molten salts CaCl_2−x (x=0, NaCl, KCl, SrCl₂, BaCl₂ and LiCl). Thermochimica Acta 470, 105-107, (2008).24. Ostvold, T. Determination of the Change in Chemical Potential of Sodium Chloride of mixing in the Ternary System MgCl₂—NaCl—CaCl₂ by EMF Measurements. Acta Chemcia Scandinavica 26, 1751-1762, (1972).25. Ritter, S. K. A New Life For Coal Ash. Chemical & Engineering News 94, 10, (2016).26. Serp, J. et al. The molten salt reactor (MSR) in generation IV: Overview and perspectives. Progress in Nuclear Energy 77, 308-319, (2014).27. Chan, C. K. et al. High-performance lithium battery anodes using silicon nanowires. Nature Nanotechnology 3, 31-35, (2007).28. Szczech, J. R. & Jin, S. Nanostructured silicon for high capacity lithium battery anodes. Energy Environ. Sci. 4, 56-72, (2011).29. Lee, S. W. et al. Kinetics and fracture resistance of lithiated silicon nanostructure pairs controlled by their mechanical interaction. Nature Communications 6, 7533, (2015).30. Nian Liu et al. A pomegranate-inspired nanoscale design for large-volume-change lithium battery anodes. Nature Nanotechnology 9, 187, (2014).

The word “illustrative” is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “illustrative” is not necessarily to be construed as preferred or advantageous over other aspects or designs. Further, for the purposes of this disclosure and unless otherwise specified, “a” or “an” means “one or more”.

The foregoing description of illustrative embodiments of the invention has been presented for purposes of illustration and of description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. The embodiments were chosen and described in order to explain the principles of the invention and as practical applications of the invention to enable one skilled in the art to utilize the invention in various embodiments and with various modifications as suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims appended hereto and their equivalents.

Claims

1. A method for the electrochemical production silicon, the method comprising applying an electrical potential across an anode and a cathode to provide electrons at the cathode, the anode and the cathode in contact with an electrolyte melt at a reaction temperature, wherein the electrolyte melt comprises a molten salt or a mixture of molten salts;a silicon-containing precursor at least partially dissolved in the electrolyte melt to provide soluble silicon-containing ions in the electrolyte melt; anda supporting electrolyte at least partially dissolved in the electrolyte melt to provide O2− ions in the electrolyte melt;and further wherein the soluble silicon-containing ions at the cathode undergo reduction reactions with the electrons to release O2− ions and deposit silicon on the cathode.

2. The method of claim 1, wherein the reaction temperature is no more than about 900° C.

3. The method of claim 1, wherein the electrolyte melt comprises an eutectic mixture of metal halides.

4. The method of claim 1, wherein the soluble silicon-containing ions comprise SiO32− ions, SiO44− ions, or both.

5. The method of claim 1, wherein the silicon-containing precursor is a silicate.

6. The method of claim 5, wherein the silicate is an orthosilicate, an inosilicate, a phyllosilicate, an aluminosilicate or combinations thereof.

7. The method of claim 5, wherein the silicate is selected from MSiO3, M2SiO4, or combinations thereof, wherein M is selected from alkali metals, alkaline earth metals or combinations thereof.

8. The method of claim 5, wherein the silicate is CaSiO3.

9. The method of claim 1, wherein the silicon-containing precursor is SiO2.

10. The method of claim 1, wherein the supporting electrolyte is selected from alkali metal oxides, alkaline earth metal oxides, and combinations thereof.

11. The method of claim 10, wherein the supporting electrolyte is selected from CaO, Na2O, or combinations thereof.

12. The method of claim 1, wherein the amount of supporting electrolyte saturates the electrolyte melt with O2− ions.

13. The method of claim 10, wherein the silicon-containing precursor is SiO2 and the supporting electrolyte is of a type and at an amount to dissolve the SiO2 to provide the soluble silicon-containing ions in situ via the formation of a silicate.

14. The method of claim 13, wherein the supporting electrolyte is CaO.

15. The method of claim 1, wherein at least a portion of the silicon-containing precursor and at least a portion of the supporting electrolyte are provided as molten glass, molten coal ash, molten cinder ash, or combinations thereof.

16. The method of claim 15, wherein additional SiO2, additional CaO, additional Na2O, or combinations thereof are added to the electrolyte melt.

17. The method of claim 1, wherein the silicon produced is crystalline.

18. The method of claim 17, wherein the crystalline silicon has a purity of at least about 95 atomic %.

19. The method of claim 1, wherein the silicon produced is in the form of nanowires.

20. The method of claim 1, wherein the method provides a yield of silicon of at least about 20% at the temperature of about 650° C. and a reaction time of about 3 hours.