Silicon manufacturing method and system

Abstract

Apparatus (100, 200) for processing metal or metalloid compounds, their oxides and / or sulfides, Chambers (110, 120, 210, 120) equipped with exhaust gas outlets located around the periphery, A plasma generator (160, 260) configured to discharge a plasma stream (163, 263) containing a compound, its oxides and / or sulfides into a chamber (110, 120, 210, 120), A first material supply means (130) configured to supply a material (132) containing a compound, its oxide and / or its sulfide to a plasma generator (160, 260), A gas supply means (170) configured to supply an inert gas (171) and / or a reducing gas (172) to the chambers (110, 120, 210, 120), Apparatus (100, 200) comprising a container (190, 290) through which a plasma flow (163, 263) is directed, causing the compound to be deposited directly onto a molten material (191, 291) placed inside the container (190, 290) without first solidifying. The present invention also relates to a method.

Description

[Technical Field]

[0001] The present invention relates to a method and system for producing high-purity silicon or other metals or metalloids. [Background technology]

[0002] For example, the production of high-purity silicon used in solar panels currently generates undesirable carbon dioxide emissions. Therefore, there is a need to switch to non-carbon reducing agents to reduce SiO2 to Si, thereby producing Si of sufficient purity to avoid further purification processes.

[0003] Hydrogen is one such reducing agent. However, a hydrogen-silica gas system requires a high temperature of 3000°C to operate efficiently.

[0004] The use of plasma is one way to circumvent this unfavorable thermodynamics, allowing refractory materials to continue functioning properly by lowering the reactor temperature to below 2000°C instead of relying on dynamics. Thus, plasma can be a mechanism to supply active species for reduction and lower the temperature required for the reaction. The localized, very high enthalpy change caused by the rapid cooling of the plasma itself allows for a low average temperature while controlling where and how large an enthalpy change occurs in the oxide. However, although the plasma itself operates at very high temperatures, the low heat capacity of the molecules involved contributes to the temperature rise, leading to side reactions and re-oxidation of silicon.

[0005] There are several methods for forming a plasma state and applying it to reduce materials as an electrical discharge. Plasma is simply a gas in which most of the constituent atoms are ionized, but it can be formed by applying power to a gas with direct current (DC) or alternating current (AC) to ionize it. DC plasmas include transitional arc plasmas (TAP) and non-transitional arc plasmas (NTAP). AC or radio frequency (RF) plasmas include inductively coupled plasmas (ICP), capacitively coupled plasmas (CCP), and microwave plasmas. The gas is supplied between electrodes in DC configurations and within the electric field in RF and microwave plasmas. As energy flows into the ionized gas through the arc current within the confined space, the plasma flow accelerates. In NTAP, this flow can approach the speed of sound. The specific energy of a plasma is defined as the ratio of the supplied power to the plasma gas flow rate: P = W / Q.

[0006] DC discharges are classified into TAP and NTAP depending on the anode configuration. The cathode is usually located upstream of the gas inlet relative to the anode. When the anode is located separately from the molten target or workpiece in the electrical circuit configuration, it is an NTAP configuration, and the anode can be the nozzle of the plasma torch device. When the anode is the workpiece, molten target, or the slag itself, it is known as TAP, and the arc transitions to the workpiece. The temperature of DC discharge plasma (high-temperature plasma) can reach up to 28,000K, but current methods limit the maximum temperature to around 15,000K. The pressure (several millitorres to several millibars) and gas flow rate are around 100-102 liters per minute, and are typically 10 1 ~10 2 It is lower than that of AC plasma with high currents of around amperes.

[0007] AC plasma can be generated by supplying the necessary current to the input gas flow via electromagnetic induction. The gas itself can conduct electrons through a small number of atoms ionized by thermal collisions or gamma rays. The gas is exposed to a magnetic field that varies with an AC power source via a magnetron, typically operating at 13.56 MHz (see U.S. Patent No. 4626767) in ICP RF plasmas and at 2.45 GHz or 915 MHz in microwave CCPs. Other frequencies are also available depending on the wavelength required for the generated radiation. In ICPs, this varying magnetic field can be supplied by an AC current through a planar coil, cylindrical coil, or toroidal coil. In CCPs, the magnetic field is generated by applying AC to a capacitor in an electrical circuit. In this circuit, the gas is present between the plates of the capacitor or between the cylindrical cathode in a coaxial anode. The electromotive force generated in the gas is proportional to the rate of change of the magnetic field over time, creating an electric field that gives motion and incident ionization to electrons. Due to the low specific heat of electrons, AC plasmas operate at lower temperatures (e.g., 5500-6000°C) than DC plasmas. The pressure is 10 1 ~10 3 Militol, gas flow rate is 10 1 ~10 3 It is about liters / minute, and typically the current is 10 0 ~10 1 This is higher than a DC plasma current of around amperes.

[0008] Glow discharge plasma, a type of CCP, has been widely known since the 1960s to provide coatings that deposit amorphous silicon, silicon nitride, and silicon dioxide via hydrogen reduction using plasma-enhanced chemical vapor deposition (PECVD) (see FR 1442502 A, GB 1104935 A). Such high-frequency plasmas often use a frequency of 13.56 MHz in industrial applications (see also US 4626767 A). Hydrogen reduction ICP is also used to reduce SF6 at an H2 / SF6 flow rate ratio of 6, achieving yields of up to 70% (see Mochalov et al., "Investigation of the Process of Hydrogen Reduction of 32S from 32SF6 via RF Capacitive Plasma Discharge"). It is also used for depositing carbon-tungsten layers (see Katayama et al., "Hydrogen Retention in Carbon-Tungsten Co-Deposition Layer Formed by Hydrogen RF Plasma"). Therefore, hydrogen plasma is also a candidate for silica reduction and silicon deposition, but it may be limited by its low yield.

[0009] One possibility is to modify an existing electric arc furnace used for carbon thermal reduction in DC transition arc (TAP) mode to achieve the following overall reaction. SiO2 + 2H2 → Si + 2H2O (1)

[0010] This would involve, for example, the use of a vacuum electric arc furnace (VEAF) similar to the transitional arc technology described in U.S. Patent Application Publication 2018 / 0237306 A1. In a transitional arc apparatus (see Figure 1 in U.S. Patent Application Publication 2018 / 0237306 A1), a short arc distance (labeled 6 in Figure 1 of this publication) is used between the molten material (7) and the graphite cathode (4). Conventional electric arc furnaces (EAFs) in silicon production typically immerse the electrodes in the raw material or molten material and do not use a vacuum.

[0011] However, the configuration of this vacuum electric arc furnace (VEAF) has been found to cause several problems when switching to hydrogen gas or plasma as a reducing agent instead of a carbon-based reducing medium, making the switch difficult.

[0012] As will be described later, the inventors have devised a number of experiments to improve the existing technology in order to produce high purity silicon while avoiding one or more of the above drawbacks. The corresponding technology has been found to be useful also for the remelting of solid Si.

[0013] The prior art also includes WO 2007 / 102745 A1 and US 2012 / 020461 A1.

[0014] One or more of the problems described above also exist in the production of high purity of other metals or metalloids such as beryllium, gallium, germanium, lithium, niobium, tantalum, titanium, tungsten, vanadium, aluminum, copper, indium, lead, tellurium, zinc, nickel, cobalt. SUMMARY OF THE INVENTION

[0015] Therefore, the present invention relates to an apparatus for treating a compound, an oxide and / or a sulfide thereof which is a metal or a metalloid, the apparatus comprising a reaction chamber and an expansion chamber, and a plasma generator configured to emit a plasma stream containing the compound, an oxide and / or a sulfide thereof into the reaction chamber, a first material supply means configured to supply a material containing the compound, an oxide and / or a sulfide thereof to the plasma generator, a gas supply means configured to supply an inert gas and / or a reducing gas to the reaction chamber, The Draval nozzle is configured to transport gas from the reaction chamber to the expansion chamber, and to lower the temperature of the gas passing through the Draval nozzle to below the condensation temperature of the compound, its oxides and / or its sulfides, and the Draval nozzle is configured to condense the compound, its oxides and / or its sulfides, which have entered the Draval nozzle in a gaseous state as a result of passing through the Draval nozzle, into a liquid or solid phase. The system comprises a container through which the condensed compound is guided by a Draval nozzle.

[0016] The present invention also relates to an apparatus for processing compounds, their oxides and / or sulfides, the apparatus is A chamber equipped with exhaust gas outlets located around its periphery, A plasma generator configured to discharge a plasma stream containing the compound, its oxide and / or sulfide into the chamber, A first material supply means configured to supply a material containing the compound, its oxide and / or sulfide to the plasma generator, A gas supply means configured to supply an inert gas and / or reducing gas to the chamber, The system comprises a container through which the plasma flow is directed, so that the compound is deposited directly onto the molten compound placed inside the container without first solidifying.

[0017] In one embodiment, the plasma generator is configured to supply a plasma flow to a container at a certain velocity. This velocity is sufficient to move the H2O in the plasma flow radially outward along the direction of the plasma flow due to Safman lift when the effective Stokes number exceeds a certain threshold (e.g., 2.8), and to move the H2O radially outward toward the exhaust gas outlet due to turboforesis when the effective Stokes number is below the threshold.

[0018] In one embodiment, the plasma generator is a non-transition type arc device that operates using a DC voltage, an AC voltage device such as a radio frequency (RF) or inductively coupled plasma (ICP) device, a microwave plasma device, or a capacitively coupled plasma (CCP) device.

[0019] In one embodiment, the gas supply means is configured to supply hydrogen in an amount exceeding the stoichiometric amount in order to completely reduce the oxide of the compound present in the reaction chamber.

[0020] In one embodiment, the first material supply means is configured to supply the material circumferentially around the anode of the plasma generator or through the anode.

[0021] In one embodiment, the first material supply means is configured to supply the compound, its oxide and / or sulfide as a solid compound, oxide and / or sulfide with an average particle size in the range of 20 to 200 μm, for example, 50 to 100 μm, preferably more than 99% of the particles being in this range.

[0022] In one embodiment, the compound, its oxide and / or sulfide supplied to the plasma generator is heated in the plasma stream so that the compound atoms in the plasma stream reach temperatures above the evaporation temperatures of B, Ti, Al, Fe, Ca, Na, Ni, P, and / or W under the associated operating pressure at the gas-liquid interface.

[0023] In one embodiment, the first material supply means is configured to supply the plasma generator with at least one of hydrogen and a chemically unreactive species (e.g., an inert gas), in addition to the compound, its oxide and / or sulfide, the chemically unreactive species may be Ar, He, or N2.

[0024] In one embodiment, the chamber comprises a reaction chamber located upstream of the expansion chamber, the reaction chamber being axially separated from the expansion chamber by a throttling section.

[0025] In one embodiment, the apparatus further comprises a second material supply means configured to supply at least one of hydrogen and a chemically unreactive species (e.g., an inert gas) into the reaction chamber around the plasma flow, wherein the chemically unreactive species may be Ar, He, or N2.

[0026] In one embodiment, the second material supply means is configured to supply hydrogen and / or the chemically unreactive species to induce laminar flow in the radially outer sheath of the plasma reaction region within the reaction chamber.

[0027] In one embodiment, the second material supply means is configured to supply hydrogen and / or the chemically unreactive species, and as the supplied hydrogen and / or the chemically unreactive species flow into the throttling section, it generates a shroud (envelopment) of the hydrogen and / or the chemically unreactive species, the shroud being located at the entrance around the material reaching the throttling section from the plasma flow.

[0028] In one embodiment, the second material supply means is configured to supply hydrogen and / or the chemically unreactive species at a temperature of up to 1500°C.

[0029] In one embodiment, the apparatus is configured to supply all excess H2 in the reaction chamber relative to the stoichiometric conditions via the second material supply means.

[0030] In one embodiment, the apparatus is configured to supply an excess of at least 50%, for example, at least 100%, of H2 relative to the stoichiometric conditions in the reaction chamber.

[0031] In one embodiment, the reaction chambers (110, 210) are vacuum chambers having a pressure of up to 800 mbar, for example, up to 100 mbar, for example, between 1 mbar and 50 mbar, for example, between 5 mbar and 20 mbar.

[0032] In one embodiment, the throttling section is positioned such that the condensed compound exiting the throttling section is guided toward a molten compound supplied into the container below the throttling section and deposited directly into the molten compound.

[0033] In one embodiment, the axial length of the expansion chamber is at least 10 times, for example, 20 to 40 times, the axial length of the reaction chamber.

[0034] In one embodiment, the first material supply means is configured such that the plasma flow contains 30-70% carrier gas and 30-70% H2 in the form of chemically nonreactive species.

[0035] In one embodiment, the carrier gas in the plasma flow has a temperature of 5 to 15 kK due to energy applied to the gas via a DC or AC circuit.

[0036] In one embodiment, the plasma generator is configured to heat the H2 in the plasma stream to at least 2kK.

[0037] In one embodiment, the plasma generator is configured such that the temperature of the plasma flow immediately after the outlet of the plasma generator is 5 to 15 kK.

[0038] In one embodiment, the reaction chamber has a cross-sectional radius ranging from 120% to 300% of the outlet diameter of the plasma generator.

[0039] In one embodiment, the upstream convergence portion of the Draval nozzle has a radius of curvature that is within 50% of the cross-sectional radius of the reaction chamber.

[0040] In one embodiment, the Dragal nozzle is configured to generate a turbulent gas flow within the expansion chamber.

[0041] In one embodiment, the Dragal nozzle is designed such that the outlet velocity of the gas passing through the Dragal nozzle is below the melting point of the oxide of the compound.

[0042] In one embodiment, the Dragal nozzle is designed such that the outlet velocity of the gas passing through the Dragal nozzle is between the melting point of the compound and the melting point of its oxide.

[0043] In one embodiment, the Dragal nozzle is configured such that the condensed compound discharged from the Dragal nozzle is guided toward and directly deposited upon a molten compound located in a container positioned below the outlet of the Dragal nozzle.

[0044] In one embodiment, the apparatus further comprises a third material supply means configured to supply at least one of hydrogen and a chemically unreactive species (such as an inert gas) to the chamber, such as the expansion chamber. The chemically unreactive species may be Ar, He, or N2.

[0045] In one embodiment, the third material supply means comprises a non-migration plasma torch that supplies hydrogen and / or the chemically nonreactive species, the plasma flow being positioned to directly impact the surface of the molten compound in the container.

[0046] In one embodiment, the third material supply means comprises a non-migration plasma torch that supplies hydrogen and / or the chemically nonreactive species, the plasma flow configured below the surface of the molten compound in the container.

[0047] In one embodiment, the third material supply means is configured to operate the non-migration plasma torch intermittently, for example, with a duty cycle of up to 50%.

[0048] In one embodiment, the third material supply means is configured to supply hydrogen and / or the chemically unreactive species by applying an electric plasma torch power of 0 to 20% compared to the electric plasma torch power applied to generate the plasma flow of the plasma generator.

[0049] In one embodiment, the container comprises an upper part configured to contain the upper surface of the liquid compound deposited from the plasma flow, and a lower part located below the upper part and configured to contain the compound arriving from the upper part.

[0050] In one embodiment, the lower part includes a heating element configured to control the solid-liquid interface of the compound within the compound molten body.

[0051] In one embodiment, the chamber comprises an inner wall, which is coaxially arranged along the main axis of the chamber and comprises one or more perforations penetrating the inner wall, thereby allowing gas to flow out from a radially central region within the chamber located inside the inner wall to a radially peripheral region located circumferentially outside the inner wall.

[0052] In one embodiment, the peripheral region is connected to a peripheral exhaust gas outlet located within the chamber.

[0053] In one embodiment, the peripheral region is open at the upper end of the inner wall, allowing gas to flow from the peripheral region to the central region.

[0054] In one embodiment, the apparatus further comprises a fourth material supply means configured to supply hydrogen gas to the central region.

[0055] In one embodiment, the container includes a liquid impurity extraction line located below the liquid surface of the liquid compound deposited from the plasma flow and present in the container.

[0056] The present invention also relates to a method for operating such an apparatus to process a compound that is a metal or metalloid, its oxides and / or sulfides, the method being: The plasma generator of the apparatus releases a plasma stream containing the compound, its oxide and / or its sulfide into the reaction chamber of the apparatus. The first material supply means of the apparatus is used to supply the material containing the compound, its oxide and / or its sulfide to the plasma generator, The gas supply means of the apparatus is used to supply an inert gas and / or a reducing gas to the reaction chamber, H2O is discharged from the chamber through an exhaust outlet located at the periphery of the apparatus, The method comprises introducing the condensed compound into the container of the apparatus without first solidifying it, and recovering the condensed compound in the container.

[0057] The present invention also relates to a method for operating such an apparatus to process a compound that is a metal or metalloid, its oxides and / or sulfides, the method being: The plasma generator of the apparatus releases a plasma stream containing the compound, its oxide and / or its sulfide into the reaction chamber of the apparatus. The first material supply means of the apparatus is used to supply the material containing the compound, its oxide and / or its sulfide to the plasma generator, The gas supply means of the apparatus is used to supply an inert gas and / or a reducing gas to the reaction chamber, The Dragal nozzle of the apparatus transports gas from the reaction chamber to the expansion chamber, lowers the temperature of the gas passing through the Dragal nozzle to below the condensation temperature of the compound, its oxides and / or sulfides, and condenses the compound, its oxides and / or sulfides that have entered the Dragal nozzle in a gaseous state as a result of passing through the Dragal nozzle into the liquid phase. The apparatus includes guiding the condensed compound into the container of the apparatus and recovering the condensed compound in the container.

[0058] The present invention will now be described in detail with reference to embodiments and accompanying drawings. [Brief explanation of the drawing]

[0059] [Figure 1a] Figure 1a is a schematic diagram of hydrogen shrouding (encirclement) and inert gas jacketing (covering) of the plasma flow. [Figure 1b] Figure 1b is a schematic diagram of hydrogen supply near the surface of the molten silicon. [Figure 2] Figure 2 is a schematic diagram of the Draval nozzle. [Figure 3] Figure 3 shows the axial flow velocity diagram of the turboforesis system. [Figure 4] Figure 4 shows the Safman lift acting on a fine particle. [Figure 5] Figure 5 is a schematic flow diagram illustrating the various physical effects that influence the flow inside the reactor. [Figure 6] Figure 6 is a simplified schematic diagram showing the first embodiment. [Figure 7] Figure 7 shows the phase diagram of the molybdenum-silicon system. [Figure 8] Figure 8 shows the temperature drop in the convergence / diffusion nozzle against the Mach number of the expanding gas. [Figure 9] Figure 9 is a graph showing the gas velocity as a function of temperature in a supersonic flow through a Drabal nozzle. [Figure 10]Figure 10 is a phase diagram of the Ta-Si system. [Figure 11] Figure 11 is a simplified schematic diagram showing a second embodiment. [Figure 12] Figure 12 is a flowchart of the first method. [Figure 13] Figure 13 is a flowchart of the second method. [Modes for carrying out the invention]

[0060] In the drawings, the last digit of the reference number is standardized for identical or corresponding parts. In addition, for clarity, the first digit of the reference number belonging to the two main embodiments presented is either "1" or "2," depending on the embodiment.

[0061] Hereinafter, the compound will be described as silicon (Si) and its oxide as SiO2. However, it will be understood that this compound may also be other metals or metalloids such as beryllium, gallium, germanium, lithium, niobium, tantalum, titanium, tungsten, vanadium, aluminum, copper, indium, lead, tellurium, zinc, nickel, and cobalt. The same applies to the corresponding oxides. In particular, the present invention and its principles are applicable to the group of compounds including silicon, titanium, germanium, and nickel. Among these possible compounds, silicon is a particularly noteworthy candidate. It will also be recognized that sulfides of the aforementioned metals or metalloids may, if necessary, be treated in the manner corresponding to the treatment of oxides described herein.

[0062] Returning to the description of the possible configurations of the VEAF apparatus described above, the inventors have identified that in one embodiment, the use of a low-conductivity graphite crucible and a hollow graphite cathode may be a potential problem in achieving this objective.

[0063] Specifically, it is desirable to achieve reduction of SiO2 in flight, control or suppress SiO-related side reactions (SiO disproportionation reactions), control the H2O-driven reverse reaction (reoxidation), and further stir or agitate the resulting Si melt to improve the uniformity of temperature and composition.

[0064] The ability to achieve reduction in flight in the transitional arc mode is considered unfavorable due to the rapid voltage changes associated with particles passing through the arc-generating electrode at various distances. Since the voltage is a function of distance, the arc-generating cathode is selected by the least resistance path, i.e., the lowest voltage. Particles extremely close to the anode represent this lowest voltage, and the power applied through the arc fluctuates significantly as new particles are selected during flight, or short-circuits occur through these particles. This leads to chaotic system fluctuations, increasing the difficulty of system control. Therefore, in the non-transitional mode, this challenge can be avoided by controlling the arc so that it occurs upstream of the solid particle injection.

[0065] In embodiments of the present invention, the carrier gas and discharge material are supplied around the cathode (before the anode) in a manner similar to a thermal spraying process, and / or the discharge material is supplied around the anode outlet after plasma formation.

[0066] As part of our research, experiments using a transitional arc apparatus investigated a method for silicon production under a hydrogen reduction atmosphere, similar to our research on magnetite reduction in the presence of hydrogen plasma for sustainable iron ore production for the steel industry. This research is described in Souza Filho et al., "Sustainable Steel through Hydrogen Plasma Reduction of Iron Ore".

[0067] For example, we applied Filho et al.'s experimental setup to silicon reduction experiments. However, while the application of a transitional arc enabled reduction by hydrogen plasma in a hydrogen reduction atmosphere, the advantages of hydrogen plasma alone in terms of reducing emissions and energy footprint were not realized in silicon production.

[0068] These experiments suggested that SiO and Si undergo a "Duvet effect," shielding the target SiO2 for reduction. This led to the easy disproportionation and decomposition of SiO into Si and SiO2, and that the decomposed SiO then reformed to form SiO again. This created a reaction loop of SiO-SiO2 formation, making the formation of Si highly unfavorable.

[0069] The inventors identified the reduction of SiO2 in flight as a method to mitigate the so-called "Duvet effect" by (i) preventing prolonged contact of Si with SiO2 for SiO generation, and (ii) achieving complete dissociation and reduction of SiO2. This reduction effect in flight allows for the rapid generation of Si from the complete dissociation and pseudo-reduction of SiO2 in the presence of monatomic hydrogen atoms derived from a hydrogen plasma.

[0070] During reduction in flight, SiO2 and H2 dissociate into Si, H, and O, which then recombine into Si and H2O. However, this leads to the problem of Si reacting with free O or H. To avoid this, the inventors discovered that (i) shrouding with hydrogen (and possibly an inert gas), (ii) quenching, and / or (iii) extraction of by-products are available. This will be discussed later.

[0071] Furthermore, the inventors have recognized that reduction in flight and direct, immediate deposition onto the Si molten material are effective in retaining and capturing the thermal energy required for the remelting of solid silicon. This enables an integrated reduction and purification method in which purification can be performed both before and during solidification while the silicon is still in a molten state due to reduction.

[0072] In the dissociation reaction shown in equations (2)-(3) below, the dissociated oxygen and hydrogen escape from the plasma flow and then recombine to form vapor (4). This forms the background for the overall reaction in equations (5) and (6), but the thermodynamic considerations also take into account the dissociation before recombination occurs. See Miao and Grishin, "Gas Dynamic-Thermal-Concentration Fields and Evaporation Process of Quartz Particles in Ar-H2 Inductively Coupled Plasma".

[0073] SiO2 → SiO + O → Si + 2O (2) H2 → 2H (3) O + 2H → H2O (4) Si + H2O → SiO + H2(5) SiO + H2O → SiO2 + H2(6)

[0074] Complete dissociation has the added advantage of avoiding SiO2 aggregation. However, if the SiO2 particle size is sufficiently small, complete melting dissociation occurs with high gas flow rates and very short residence times, so that SiO2 and SiO species exist only in the plasma flow where these high enthalpy changes persist, and liquid silicon exists only outside the plasma flow. The latter is further facilitated by preventing reoxidation through shrouding, rapid cooling, and / or rapid removal of the by-product H2O.

[0075] One advantage of this strategy is to reduce the exposure of Si to H2O or SiO2 at high temperatures, thereby preventing the re-oxidation of silicon and the formation of SiO. In the outer jacket (coating) of the plasma jet where the temperature has dropped and only partial dissociation and reduction may have occurred, this can be facilitated by using additional inert gas and hydrogen shrouding (enclosure). Refer to Figure 1a. Figure 1a shows the hydrogen shrouding (solid line) and the Ar jacket (dashed line) to maintain the gas mixing state and laminar flow within the reaction chamber. Further hydrogen shrouding at the melt surface (Figure 1b) prevents the reverse reaction of Si to SiO2. Refer to Figure 1b. Figure 1b shows supplying a flow of H2 to create a highly reducing atmosphere at the melt surface and reduce the local H2O concentration. Si is reactive at the high temperature in the dissociation reaction region, but re-oxidation does not occur unless the temperature is below T = 6 kK. Cooling in the T = 2 - 6 kK region, even if carried out at a rate of 10 5 ~10 6 K·s -1 causes re-oxidation due to the reactivity of Si with available O (recombination in the T = 5 - 6 kK region) and the H2O reverse reaction (re-oxidation) in the T = 2 - 5 kK region.

[0076] Rapid cooling can be achieved by directing the flow into a convergent-divergent nozzle called a de Laval nozzle. The gas flowing into the convergent part of the nozzle at a relatively high pressure and relatively low subsonic speed is accelerated in the process of converting pressure and thermal energy into kinetic energy due to the continuity of the mass flow. This acceleration continues until the gas velocity reaches the Mach number at the throat of the nozzle. After passing through the throat, the flow continues to accelerate in the divergent part and reaches supersonic speed. This increase in kinetic energy is brought about by the decrease in the pressure and thermal energy of the flow in the divergent part.

[0077] Figure 2 shows a Draval nozzle whose radius of curvature r is equal to, or at least close to, the radius R of the reaction chamber. This allows the flow to converge and diffuse rapidly with small axial displacement. Thermal energy is thus converted into kinetic energy, allowing the nozzle to be designed to achieve the target temperature (T=1500°C) at which the silicon flow condenses into a liquid. In Figure 2, the argon shrouding gas is shown by a dashed line, and the flow of active hydrogen is shown by a solid arrow.

[0078] The Reynolds number (Re) is a dimensionless number used to describe the characteristics of fluid flow. It is defined as the ratio of inertial forces to viscous forces in a fluid: Re = ρuL / μ. Apart from terms such as density (ρ), velocity (u), characteristic length (L), and viscosity (μ), density is related to pressure, and an increase in pressure leads to an increase in Re, while an increase in temperature increases viscosity and decreases Re.

[0079] Debris or corrosion inside the nozzle can cause flow turbulence, increasing the likelihood of turbulence generation. Temperature changes also cause flow turbulence and induce turbulence. Furthermore, to minimize turbulence and pressure loss, the nozzle requires an appropriate shape design, such as a smooth transition zone or a circular cross-section.

[0080] The inventors discovered that by making greater use of this shock wave effect amplified by the Draval nozzle, the purity of the deposited layer can be improved. The shock wave creates a region of intense turbulence. This shock wave and its associated turboforetic effect are thought to work to remove lighter impurities from the Si deposited region.

[0081] Turboforesis is the tendency for particles to move in the direction of decreasing turbulence. This is not a minor correction, but governs the dynamic behavior of particles in the diffusion shock and inertial relaxation regions (for example, this tendency is prominent in the embodiment of the present invention that is effective for H2O removal). When a gas-liquid mixed flow collides with a silicon melt, turboforetic forces act on the particles. Light particles with low momentum, such as H2O, are pushed towards the walls of the deposition and extraction region. On the other hand, particles with high momentum, such as relatively dense liquid Si, require a greater force and are not pushed out of the melt. Figure 3 illustrates this phenomenon using a velocity diagram of the turboforesis axial flow that allows for overflow of light particles (non-Si) in an embodiment in which a gap is provided between the crucible and the reactor wall.

[0082] Similar effects to those in Figure 3 are also observed in the embodiments shown in Figures 6 and 11 (described later), but in this case, H2O is pushed towards the walls. These walls obstruct the flow, generating shock waves. A highly turbulent region is formed here, while a less turbulent state persists in the outer region of the expanded chamber where the flow velocity is relatively lower. Lighter particles, mainly light diatomic particles and triatomic particles such as H2O, are subjected to this turboforesis force towards the outlet.

[0083] Furthermore, Safman lift is a force acting in shear flow, and it acts on particles in the direction of the fastest flow. In the embodiment of the present invention, the flow velocity is maximum at the axial centerline and outlet, separating by-products and impurities. Figure 4 illustrates this concept, showing fine particles in a medium with changing flow velocity. In this embodiment, the effect of Safman lift is usually not dominant compared to the turboforetic effect.

[0084] Thermophoresis also has a minor effect on particle flow in the silicon melt. First, lighter particles move to the outer layer of the plasma sheath, and then, as low-momentum particles cross the turbulent boundary layer towards the reactor walls and exhaust outlet, the lighter particles move towards the walls of the water-cooled reactor.

[0085] Shock-induced evaporation on the surface of the silicon melt also contributes to the vaporization of heavier metal oxides and metal complex impurities (e.g., metal and boron oxides and hydroxides), which is effective against particles other than silicon droplets in the gas stream that should be removed from the workpiece or melt. See Smolders and van Dongen, "Shock Wave Structure in a Mixture of Gas, Vapour and Droplets." This effect has been observed in experiments by the inventors to lead to an improvement in the silicon deposition rate.

[0086] Figure 5 shows an overall flowchart illustrating the effects described above: Safman lift, turbophoresis, shock wave-induced evaporation, and thermophoresis. Note that these effects work in coordination to prevent impurities from reaching the silicon melt.

[0087] At high temperatures, Si is relatively reactive in the presence of the byproduct H2O (equation (4)) (see equations (5)-(6)), so the problem of reoxidation becomes significant. This problem can be mitigated by a protective method using a shroud (encircle) of argon (Ar) or other inert gas, as shown in Figure 2. Such inert gas shrouding (encirclement) is H + Ions and Si + This prevents ions from reacting with the nozzle material.

[0088] In one embodiment, the temperature of the flowing inert gas is significantly lower than the centerline temperature of the plasma flow. This can be injected into the edge of the fluid, exposing the nozzle to the low temperature and rapidly cooling the temperature of the reactive flow.

[0089] The process's scalability is achieved by reducing SiO2 in flight and directly depositing the reduced Si into the silicon melt. That is, in silicon purification, reheating for silicon recrystallization typically consumes a considerable amount of energy. A significant reduction in speed (approximately Mach 3.15 in the inventors' experiments) leads to the recovery of thermal energy, helping to maintain the melt at a high temperature and minimizing the need for reheating. The useful speed range includes a range above Mach 2. In some embodiments of the present invention, the range used is less than Mach 5 or less than Mach 4.

[0090] In one embodiment, surface plasma purification is used, in which localized heating is applied to volatilize boron and / or other metallic elemental impurities. This is useful for purifying silicon of metallic grade or higher to solar cell grade.

[0091] The temperatures required to evaporate typical metals found in quartz or other silicon raw materials range from 1457°C for titanium to 3527°C for boron. The table below shows the evaporation temperatures (evaporation from molten silicon) for various metals.

[0092] [Table 1]

[0093] This overheating normally occurs during the silicon reduction process in flight, but for impurities that are not properly vaporized and extracted, this localized overheating of the melt allows for the vaporization and extraction of such residual impurities, while also providing additional thermal energy to maintain the melt in a liquid state.

[0094] Such superheating can be applied intermittently or continuously, with intermittent application allowing for reduced energy consumption. Furthermore, boron and / or other impurities can be removed from the melt by injecting a purge gas such as H2O to promote boron volatilization. H2 and O2 can also be used as purge gases. Applying superheating locally is sufficient to avoid re-oxidation. In other words, superheating does not need to be applied to the entire surface of the silicon melt; it only needs to be applied locally, such as to one localized area of ​​the surface at a time, or to the same localized area of ​​the surface consistently.

[0095] Superheating of the surface can be achieved by directing a plasma flow of high-temperature hydrogen from above toward the surface. Hydrogen applied in this manner can also contribute to the formation of a reducing atmosphere within the reactor.

[0096] Alternatively, a plasmatron or similar device can be used to apply hydrogen plasma or another localized superheating medium below the surface of the silicon melt, resulting in a mixing effect along with the effects described above. In fact, in one embodiment, this may eliminate the need for stirring (typically electromagnetic stirring) and allow for a higher cooling rate, potentially saving energy in a directional solidification process. The increased cooling rate reduces the energy required to continuously heat the melt to maintain a controlled crystallization front.

[0097] Instead of using an externally supplied superheating medium designed to act directly or internally on the silicon melt, hydrogen gas can be used as a shrouding gas near the surface of the silicon melt. In this case, the shrouding hydrogen is supplied via a gas injection lance directed towards the melt surface (see Figure 6).

[0098] When purifying metals or metalloids other than silicon, the temperature of the molten compound is not expected to exceed the melting point of the compound.

[0099] A communication channel can be provided from the silicon molten material to a communication chamber for metallurgical purification of the molten silicon. This communication channel can be configured to be directly integrated into metallurgical purification techniques such as vacuum purification, float zone purification, directional solidification, and / or continuous Czochralski process.

[0100] Referring to Figure 6, a typical apparatus 100 for processing Si and / or SiO2 is shown. That is, apparatus 100 can be designed to be suitable for processing SiO2, producing reduced Si in the molten material 191. Apparatus 100 can also be designed to be suitable for processing Si, in which case Si is also produced in the molten material 191. In the latter case, the process is a pure remelting of the Si being introduced. In other embodiments, apparatus 100 can also be designed to be suitable for processing a mixture of Si and SiO2.

[0101] Depending on the type of material to be processed, the embodiments described herein can be adapted to such processing in various ways. For example, the existence and design of the various mechanisms described herein for removing waste can be modified. That is, the use and design of a Draval nozzle or other types of throttling can be modified. Also, various gas pressures, gas velocities, and gas temperatures can be modified, etc.

[0102] The apparatus 100 may include a reaction chamber 110 and an expansion chamber 120 located downstream of the reaction chamber 110. Si and / or SiO2 are directly introduced into the reaction chamber 110. Depending on the embodiment, the reaction chamber 110 and the expansion chamber 120 may be the same chamber, or they may be two different parts within the same chamber, and different types of throttling sections 180 may or may not be provided between such sub-chambers.

[0103] The apparatus 100 further comprises a plasma generator 160, which is configured to release a plasma stream 163 into the reaction chamber 110, the plasma stream 163 containing introduced Si and / or SiO2. The plasma generator 160 may be made of a heat-resistant (fire-resistant) material such as tantalum and / or water-cooled steel.

[0104] In the illustrated embodiment, the apparatus 100 is generally configured to transport silicon material vertically downward from the plasma generator 160, through chambers 110 and 120, and finally to the silicon collection container 190 (described later). In this and other cases, the apparatus 100 can be generally cylindrical with a common vertical axis through which the silicon material is transported downward. In this and other cases, the reaction chamber 110 is associated with a corresponding central axis (which may be vertical or at least nearly vertical), and the expansion chamber 120 is also associated with a corresponding central axis (which may be vertical or at least nearly vertical), which may be identical or at least nearly identical to the central axis of the reaction chamber 110. If there is only one chamber, this chamber has a corresponding vertical or at least nearly vertical central axis, and the silicon material flows downward along this axis. The apparatus 100 will be described below using polar coordinates. In polar coordinates, there is an axial direction downward, a radial direction outward from the central axis, and an angular direction.

[0105] The apparatus 100 further comprises a first material supply means 130. The first material supply means 130 is configured to supply a material 132 containing Si and / or SiO2 to be supplied to the plasma generator 160. The first material supply means 130 may itself be a conventional material supply device, such as a screw feeder or conveyor belt, and may supply silicon and / or silicon oxide in a solid state (granular or powdery). Alternatively, the first material supply means 130 may be a fluidized powder supply system. The first material supply means 130 may be configured to supply solid Si and / or SiO2 with an average particle size in the range of 20 to 200 μm, for example, 50 to 100 μm, preferably more than 99% of the particle size being in this range. When processing high-purity Si wafer waste (known as kerf), the average particle size may be as small as 0.1 μm. The particle size is generally selected so that the Si and / or SiO2 flows properly as a pseudo-continuum and the Si is completely dissociated as described above. This dissociation may require particle sizes smaller than 500 μm.

[0106] In one embodiment, SiO and / or SiO2 are supplied at a rate of 10 to 1000 g / min (e.g., 15 to 500 g / min, about 30 g / min).

[0107] The apparatus 100 further includes a gas supply means 170 configured to supply at least one of an inert gas 171 and a reducing gas 172 to the reaction chamber 110. As shown in Figure 6, the gas supply means 170 may be connected to a plasma generator 160 and configured to supply at least a portion of the inert gas 171 and / or the reducing gas 172. This means that the inert gas 171 and / or the reducing gas 172, along with the supplied Si and / or SiO2, are directly supplied to the plasma flow in the plasma generator 160. This means that the material 132 supplied by the first material supply means 130 may include at least a portion of the inert gas 171 and / or the reducing gas 172 supplied to the chamber, in particular the reaction chamber 110.

[0108] The material can be supplied coaxially with the gas flow, for example, radially outward from the gas flow supplied by the plasma generator 160, or behind the anode outlet. In other words, the material can be supplied into a process gas (hydrogen and / or inert gas), which is further supplied radially outward from the main axial gas flow. The material may also be supplied after the process gas has been plasma-converted.

[0109] The apparatus 100 may further include a throttling section 180 arranged to transport gas from the reaction chamber 110 to the expansion chamber 120. The throttling section 180 is or comprises a Draval nozzle. In particular, if the throttling section 180 is a Draval nozzle (but not limited thereto), the throttling section 180 is configured to reduce the temperature and pressure of the gas passing through it. In one embodiment, such a reduction in the temperature of the gas is below the condensation temperature of Si, SiO, SiO2, SiS, and / or SiS2. In other words, gaseous Si passing through the throttling section, and especially gaseous Si passing through a Draval nozzle, condenses into a liquid, or possibly solid, form.

[0110] The walls of the constricted section 180 may be made of refractory material such as water-cooled molybdenum or high-temperature steel. The walls of the constricted section 180 can also be water-cooled. In the case of steel, the steel can be protected by a jacket / shroud flow of an inert gas such as argon.

[0111] The chambers 110, 120, and in the illustrated embodiment at least the expansion chamber 120, may be provided with an exhaust gas outlet 126. The exhaust gas outlet 126 may be located on the periphery of the chamber. That is, the exhaust gas outlet 126 has a hole provided in the periphery wall of the chamber at a distance from the central axis of the chamber, allowing exhaust gas 127 to be discharged from the chamber through the periphery hole.

[0112] The apparatus 100 further comprises a container 190 into which condensed Si is deposited as a result of flowing through the apparatus 100, particularly through the chambers 110, 120 (for example, along the central axis). In one embodiment, the condensed Si may be directed to be deposited directly into the container 190 as a result of the flow after passing through the throttling section / Draval nozzle 180. The flow of condensed Si may be directed to be deposited directly from the chambers 110, 120 to the container 190 as a result of the relative geometric arrangement of the plasma generator 160, the chambers 110, 120, and / or the throttling section 180, without physical contact with any part of the apparatus 100.

[0113] The container 190 may contain a molten Si 191 into which condensed Si is deposited directly in the manner described above without first solidifying. For example, the container 190 may be a bowl or similar open container formed from a suitable material such as refractory material for holding the molten Si. The surface 191a of the molten Si 191 is open and visible from above, allowing Si arriving from above to simply fall and be deposited directly into the molten Si 191.

[0114] The Si molten material 191 can be preheated to dissolve the deposited solid phase material, particularly the condensed solid phase Si.

[0115] The illustrated general vertical material flow design is provided for illustrative purposes only, and it should be understood that horizontal or inclined configurations are also possible. Furthermore, while it is preferable that the general flow direction of the Si material is the same from the plasma generator 160 to the vessel 190, the flow direction of Si may change as a result of the Si passing through the throttling section 180.

[0116] In principle, the throttling section 180 can be configured such that the condensed silicon discharged from the throttling section 180 is guided toward the molten material 191 and deposited directly onto the molten material 191, and the molten material 191 can be located below the throttling section 180.

[0117] The plasma generator 160 can be configured to supply the plasma flow 163 to the container 190 at a certain velocity. This velocity is sufficient to move the H2O in the plasma flow 163 radially outward along the direction of movement of the plasma flow 163 due to the Safman lift described above, and further to move the H2O radially outward toward the exhaust gas outlet 126 due to the turboforesis described above. The velocity at which this is achieved will vary depending on the overall design of the device 100, but in a preferred embodiment, at least 50% of the H2O mixed in with the material flow exiting the throttling section 180 is diverted in this manner through the exhaust gas outlet 126.

[0118] Particle aggregation occurs along streamlines at moderate Stokes numbers (Stk > 1, Stk < 2.8). The Stokes number Stk is a dimensionless number defined as the ratio of the particle characteristic time to the flow characteristic time. The particle characteristic time is given by t0 = (ρ p d 2 p ) / (18μ g ) can be expressed as ρ p d is particle density, 2 p μ is the particle diameter. gis the kinematic viscosity of the fluid. The effect of Stokes number on particle-fluid interaction is widely known. Clustering in this two-phase flow occurs specifically when the Stokes number is approximately 1 (Stk≈1). Lau and Nathan, "The Effect of Stokes Number on Particle Velocity and Concentration Distributions in a Well-Characterized, Turbulent, Co-Flowing Two-Phase Jet", 2016, report that in compressed air alumina particle-containing flows, when the Stokes number exceeds 2.8, the compressed air containing particles clusters along the axis, while below 2.8, clustering occurs away from the axis. Turbophoresis is dominant at low Stokes numbers, while Saffman lift becomes dominant at high Stokes numbers, causing particles with specific characteristics to aggregate axially. These properties that promote axial aggregation are explained by particle density, and in the plasma spray applications described in this embodiment, a Stokes number greater than 10, and even greater than 20, is advantageous for axial aggregation of particle-containing multiphase flows. This is as described in the study of cold spray Drabal nozzles by Meyer, Caruso, and Lupoi, "Particle Velocity and Dispersion of High Stokes Number Particles by PTV Measurements inside a Transparent Supersonic Cold Spray Nozzle", 2018. Meyer et al. also explain the importance of mass loading, where it was found that the injection rate of titanium increases with increasing injection rate, and the optimal Stokes number is about 6. In embodiments of the present invention, turboforetic and Safman flow characteristics are optimized using Stokes numbers in the range of 1 to 10 to optimize axial particle concentration and velocity.

[0119] Therefore, the radial outward movement of H2O within the plasma flow 163 follows an efficient variation in Stokes number throughout the flow. As the fluid velocity slows near the wall, turboforesis becomes dominant again, but condensation becomes non-uniform throughout the flow because Safman lift becomes dominant in the fast-flowing, hotter central part of the flow. Precise control of the resulting H2O movement depends on the empirical parametric drive design of the gas flow rates (main gas flow rate, ambient gas flow rate, shrouding gas flow rate) and the quenching characteristics.

[0120] The plasma generator 160 is shown to include an anode and a cathode. Both are located upstream of the (reaction) chamber 110 and cooperate to generate a plasma flow 163.

[0121] Generally, the plasma generator 160 may be a non-transitioning arc device operating with a DC voltage (as shown in Figure 6), an AC voltage-using device such as a radio frequency (RF) or inductively coupled plasma (ICP) device, a microwave plasma device, or a capacitively coupled plasma (CCP) device. In other embodiments, the plasma generator 160 is instead a transitioning arc device having a plasma generator cathode or a nozzle constituting a plasma generator cathode. In the latter case, the anode may be located, for example, below the vessel 190 and connected to the vessel 190.

[0122] The cathode can be formed from tungsten or tungsten oxide. The anode can be formed from copper (for example, a copper anode with a diameter of 50 mm is used). The plasma generator 160 can operate in a power range higher than approximately 50 kW, for example, with a nominal power rating of up to approximately 1 MW.

[0123] The reducing gas supplied by the gas supply means may be, at least partially, or entirely, hydrogen gas. The hydrogen gas may be at least partially composed of hydrogen gas that is propelled by the plasma generator 160 to generate the plasma flow 163.

[0124] In one embodiment, the gas supply means 170 is configured to supply hydrogen gas to achieve a stoichiometric amount of hydrogen in order to completely reduce the SiO2 present in the reaction chamber 110.

[0125] In one embodiment, the material supplied to the plasma generator 160 to form the plasma flow 163 includes, separately from the silicon material to be processed, a mixed gas of 30% to 70% reducing gas and the remainder being a chemically unreactive gas. For example, the amount of reducing gas is 40% to 60% or 45% to 55% (e.g., about 50%) by volume, with the remainder being a chemically unreactive gas. In other words, the first material supply means 130 can be configured so that the plasma flow 163 contains 30-70% carrier gas as a chemically unreactive species and 30-70% H2.

[0126] In one embodiment, the total volumetric gas flow rate of reducing gas and non-reactive gas supplied through the plasma generator 160 is at least 10 slpm (standard liters per minute), e.g., at least 50 slpm, e.g., at least 100 slpm, and / or up to 1000 slpm, e.g., up to 500 slpm, e.g., up to 300 slpm. In one embodiment, the total volumetric gas flow rate is approximately 160 slpm. These ratios are maintained in several embodiments as a function of input power. These volumetric flow rates were set to match a nominal plasma generator power of 100 kW.

[0127] The first material supply means 130 can be configured to supply material 132 circumferentially around the anode of the plasma generator 160. Alternatively, the material 132 can be supplied through the anode by the first material supply means 130. For example, the material 132 can be supplied through a hole provided on the side of the anode into a cavity formed by the anode, and can merge with and be drawn into a gaseous material propelled by the electric field of the plasma generator 160 inside the cavity, thereby forming part of the plasma flow 163.

[0128] In these embodiments and other embodiments, the Si and / or SiO2 supplied to the plasma generator 160 are heated in the plasma stream 163. As a result, the Si atoms in the plasma stream 163 reach temperatures above the evaporation temperatures of one or more of B, Ti, Al, Fe, Ca, Na, Ni, P, and / or W under the associated operating pressure at the gas-liquid interface.

[0129] As described above, the first material supply means 130 may be configured to supply to the plasma generator 160 a chemically unreactive species in addition to the supplied Si and / or SiO2, and further in addition to a reducing gas which may itself be hydrogen gas. The chemically unreactive species may be any inert gas. In principle, it is also possible to use a gas such as nitrogen, but the inventors have achieved good results by using inert gases other than nitrogen, such as noble gases, in particular inert gases containing at least 50% Ar.

[0130] Therefore, plasma flow 163 may contain Si, O, Ar, and H components.

[0131] The reaction chamber 110 can be a vacuum chamber in the sense that a pressure reduced relative to atmospheric pressure is distributed throughout. In one embodiment, the pressure in the reaction chamber 110 can be up to 800 mbar, for example, up to 100 mbar, or even up to 50 mbar. In a particular embodiment, the pressure in the reaction chamber is 1 to 50 mbar, for example, 5 to 20 mbar. The expansion chamber 120 typically has a lower average pressure than the reaction chamber 110 due to pressure loss in the Draval nozzle. If there is only one chamber, this chamber may have a pressure similar to that described for the reaction chamber 110.

[0132] The apparatus 100 may further include a second material supply means 140 configured to supply at least one of hydrogen 141 and a chemically unreactive species 142 (such as an inert gas) to the reaction chamber 110. The unreactive species 142 may be the same as or different from the unreactive species described above. The second material supply means 140 may be configured to supply hydrogen and / or the unreactive species 142 around the plasma flow 163. In certain embodiments, the unreactive species 142 is supplied only by the second material supply means 140 and not by the plasma generator 160. In other embodiments, the unreactive species 142 is supplied partly by a direct connection to the plasma generator 160 and partly by the second material supply means 140. It should be noted that both the plasma generator 160 and the second material supply means 140 open into the (reaction) chamber, thereby allowing both the reducing (hydrogen) gas and the inert gas supplied by the first material supply means 130 (directly connected to the plasma generator 160) and the second material supply means 140 to contribute to the chemical equilibrium within the chamber.

[0133] In one embodiment, the second material supply means 140 is configured to supply hydrogen 141 and / or the chemically unreactive species 142 to induce laminar flow in the reaction chamber 110, particularly in the radially outer sheath of the plasma reaction region within the reaction chamber 110. This laminar flow may be oriented along the central axis as described above. Laminar flow can be achieved by supplying hydrogen 141 and / or the chemically unreactive species 142 through one or more inlet lances positioned at an angle to supply them obliquely to the flow direction of the plasma flow 163. To achieve such laminar flow, the flow velocity of hydrogen 141 and / or the chemically unreactive species 142 can also be adjusted with respect to the flow velocity of the plasma flow 163, the internal geometry of the reaction chamber 110, the internal dimensions of the reaction chamber, etc. This limits mass transfer between the plasma turbulence and the surrounding shrouding flow of the laminar flow, making mass transfer by eddy diffusion and molecular diffusion roughly equal, and vortices in most of the fluid do not promote significant mixing as in the turbulent region.

[0134] Hydrogen 141 and / or the chemically unreactive species 142 can be supplied, for example, via a quartz tube.

[0135] In one embodiment, the second material supply means 140 is configured to supply hydrogen 141 and / or chemically unreactive species 142 to form a shroud 143 around the plasma flow 163. This shroud is formed by the supplied hydrogen 141 and / or chemically unreactive species 142. The shroud 143 can be located at any position along the path of the plasma flow 163 through the reaction chamber 110. However, the inventors have found it advantageous to position the shroud 143 so as to surround the plasma flow 163 in a cross section perpendicular to the direction of movement of the plasma flow 163, at the entrance to the throttling section 180 on the path of movement of the plasma flow 163. Thus, when using a Draval nozzle, the shroud 143 is positioned at the entrance around the silicon-containing material that travels from the plasma flow 163 to the Draval nozzle. In this way, the shrouding inert / reducing gas forms a protective jacket around the silicon-containing plasma stream 163 as it passes through the throat 182 of the nozzle.

[0136] The second material supply means 140 may be configured to supply hydrogen 141 and / or the chemically unreactive species 142 at a temperature of up to 80°C.

[0137] The apparatus 100 may be configured to supply an excess of H2 to stoichiometric conditions within the reaction chamber 110. Here, the term “stoichiometric” refers to a stoichiometric volumetric flow rate just sufficient to reduce all the Si and / or SiO2 supplied to the reaction chamber 110. This excess H2 may be supplied entirely via a second material supply means 140. This ensures that the volumetric flow rate of H2 supplied via the first material supply means 130 is at most a stoichiometric flow rate.

[0138] In either case, the apparatus 100 can be configured to supply at least 50% (e.g., at least 100%) of excess H2 to the stoichiometric conditions in the reaction chamber 110.

[0139] It is understood that the above description regarding the volumetric flow rate balance of H2 to Si and / or SiO2 is applicable to the single chamber used, rather than a specific reaction chamber 110.

[0140] As mentioned above, silica may partially melt and form a duvet that insulates the remaining powder during discharge. Furthermore, silica may react with hydrogen at temperatures above 2000°C, generating SiO2 via the reaction SiO2 + H2 → SiO2 + H2O. The generated SiO then deposits on the substrate used as the anode target in conventional transition arc anode systems.

[0141] In contrast, according to embodiments of the present invention, silica particles are atomized by evaporating them using a plasma flow 163 at a temperature of at least 3kK, for example, at least 5kK, for example, 5-11kK, or even close to 15kK. This solves the aforementioned "Duvet" problem when using a transitional arc as a melting and reduction mechanism in silicon production. Sufficiently fine atoms can be easily atomized without the shielding effect of semi-molten silica. Generally, the carrier gas in the plasma flow 163 may have a temperature of at least 3kK, for example, at least 5kK, for example, between 5kK and 15kK, or between 5kK and 11kK. More generally, the plasma generator 160 may be configured so that the temperature of the plasma flow 163 immediately after the outlet 164 of the plasma generator 160 reaches at least 3kK, for example, at least 5kK, for example, between 5kK and 11kK, or as a maximum of 153kK or more.

[0142] When a noble gas such as Ar is used as a carrier gas, a portion of the noble gas may be ionized at temperatures up to at least 11 kK, especially in the presence of a non-migrating arc plasma. The ionized Ar heats the residual gas in the H2-containing plasma to at least 2 kK, e.g., at least 3 kK, and in some embodiments to nearly 10 kK as a maximum temperature. By measuring or estimating this temperature in the plasma flow 163 within the reaction chamber 110, the temperature of the plasma gas discharged from the reaction chamber can be calculated, thereby achieving the target temperature as a result of rapid quenching and passing through the throttling section 180. The quenching mechanism is relatively fixed according to the rocket nozzle design equations, and the dimensional design of the nozzle at the throttling section 180 determines the temperature drop over the entire length of the nozzle throat 182. The temperature of the gas reaching the throttling section 180 within the reaction chamber 110 can be determined, for example, by empirical inverse calculation using CFD / MHD (Computational Fluid Dynamics / Magnetohydrodynamics) or plasma multiphysics modeling.

[0143] The temperature of the ionized gas in non-migration arc plasmas and AC plasmas can be modeled, and the inventors experimentally determined the following specific example. According to this model, the average temperature at a plasma input power of 100 kW was approximately 6000 K at a distance of 0.15 m from the outlet 164 of the plasma generator 160. This temperature-length configuration varies depending on the input current and the degree of vacuum in the reaction chamber 110. The axial length 112 of the reaction chamber 110 can be set according to the temperature drop that occurs along the axial length of the plasma flow 163 before the converging / diffusing nozzle of the throttling section 180. In this particular test example, a reaction chamber length of 1.5 m was found to be optimal. However, the temperature value at 3.0 m depends on the plasmatron and volumetric gas flow rate used.

[0144] Generally, the axial length 112 of the reaction chamber 110 can be in the range of 5 to 30 times the outlet diameter of the plasma generator 160, for example, in the range of 8 to 20 times.

[0145] The inventors discovered that by using the non-migration arc plasma apparatus 160, powdered Si and / or SiO2 can be added in the gas inlet, and the formation of SiO can be avoided by the complete dissociation of the added Si and / or SiO2. They also found that the electrode wear rate is reduced compared to when using other types of plasma generators. Furthermore, reducing H2 gas can be supplied by shrouding and / or flushing as described above. Also, as will be described later, rapid quenching can be achieved by passing the material in the plasma flow 163 through the Draval nozzle of the throttling section 180 at supersonic speed.

[0146] Silica first forms SiO in the intermediate reaction of silicon production. The inventors discovered that the reaction SiO2 + H2 → SiO + H2O is preferable to the reaction SiO + H2 → Si + H2O. This has caused several problems in the development of hydrogen-silicon production routes. Specifically, the formation of SiO so far has led to disproportionation of Si to SiO2 due to the increasing presence of SiO, removal as an exhaust gas, or the formation of SiO x We have shown a loss mechanism that leads to three undesirable effects: condensation into [a certain substance].

[0147] Si is produced very rapidly due to its favorable thermodynamic properties. The produced SiO can condense and deposit on water-cooled surfaces such as gas outlets and water-cooled crucibles in transitional arc plasma systems. To overcome this, a complete dissociation reaction can be employed. A thermodynamically less optimal reaction pathway is O 2- Ions are H +This is overridden by reaction kinetics that ensure the possibility of recombination. This can be achieved by supplying hydrogen at a shrouding and / or flushing rate in the range of 100% to 200%, e.g., at least 50%, e.g., at least 100%, e.g., at least 150%, e.g., 100% to 200%, relative to the reducing hydrogen flow rate (stoichiometric, see above) in the reaction chamber 110. As described above, the hydrogen gas for shrouding and / or flushing can be supplied through a second material supply means 140. This means may take the form of an auxiliary chuck located in the reaction chamber 110, positioned radially away from the plasma flow 163, and may be set at an angle to the overall flow direction of the plasma flow 163.

[0148] The reaction chamber 110, particularly its walls, can be made of zirconia, freeze-lined quartz, and / or steel, and in the case of steel, it can be water-cooled. In addition, or alternatively, it can be made of molybdenum, with or without water cooling. Molybdenum has been confirmed to withstand the thermal shock associated with the startup of the plasma generator 160. The resulting molybdenum-silicon system maintains the maximum atomic percent of solid molybdenum along the 2023°C isotherm (see Figure 7; the temperature range of pure molybdenum is up to 2623°C, showing maximum stability at the 2023°C isotherm). Other materials may experience problems due to thermal shock. This phenomenon has been observed in alumina, for example. The same can be said for the expansion chamber 120.

[0149] The aperture section 180 will now be described. As mentioned above, in one embodiment, the aperture section 180 is composed of a convergence / diffusion nozzle such as a Draval nozzle.

[0150] To reduce the risk of the Si vapor in the plasma flow 163 reacting in the opposite direction with the by-product H2O as described above, the Si vapor can be rapidly quenched using the throttling section 180, particularly the nozzle of the throttling section 180 through which the material passes between the reaction chamber 110 and the expansion chamber 120. In this way, the Si vapor is cooled to form liquid Si, which can be deposited directly into the molten material 191. This overcomes the challenge of silica reduction at a sufficient conversion rate. Furthermore, the high-temperature, high-enthalpy plasma flow expands and is aerodynamically quenched to become a more uniform, low-temperature flow within a temperature range that typical refractory materials can withstand, thereby mitigating some of the high-temperature risks of the reactor.

[0151] To achieve appropriate rapid cooling and prevent recirculation to the reaction chamber 110, one or more parameters can be selected from among the radius 111 of the inner wall of the reaction chamber 110, the radius of curvature 181a of the upstream convergence portion 181 of the convergence / diffusion nozzle, and the minimum radius 182a of the throat portion 182 of the convergence / diffusion nozzle.

[0152] For example, in one embodiment, the cross-sectional radius 111 of the reaction chamber 110 is between 120% and 300% of the outlet diameter of the plasma generator 160 in the case of a confined flow with insufficient expansion. In another embodiment, the reaction chamber radius is 3500% of the anode outlet of the plasma generator in the case of an over-expanded flow.

[0153] In another example, the radius of curvature 181a of the upstream convergence section 181 of the nozzle 180 is within 50% of the cross-sectional radius 111 of the reaction chamber 110. The upstream convergence section 181 may have a constant radius of curvature 181a, or it may have a radius of curvature within 50% of the cross-sectional radius 111 over at least 50% of its axial length.

[0154] The diameter of the throat has the following relationship to the reaction chamber:

number

[0155] Similarly, the relationship between the gas Mach number and the desired expansion gas temperature is as follows:

number

[0156] isentropic expansion coefficient γ p =1.48, constituent gas γ Ar =1.66 and γ H2 Assuming a weighted average of =1.30, as shown in Figure 8, a Mach number Ma = 3.15 is required for the desired temperature drop from T = 6kK. Figure 8 shows the temperature drop across the entire convergence and diffusion nozzle with respect to the Mach number of the expanding gas. T0 is the temperature before the nozzle inlet, and T is the outlet temperature of the diffusion section. The ratio of the radius of curvature r of the reaction chamber wall surface in the convergence and diffusion section of the reaction chamber 110 to the radius r* of the nozzle throat 182: the greater the decrease in radius, the greater the temperature drop.

[0157] In this example, the exit velocity is v e =2770ms -1 The expansion temperature is T e This is 1500°C, which is higher than the melting point of silicon (1410°C) and lower than the melting point of silica (1710°C).

[0158] Generally, the throttling section 180 can be designed such that the outlet velocity of the gas passing through the throttling section 180, particularly the gas flowing out of the throttling section 180, is less than the melting point of SiO2, i.e., between the melting point of Si and the melting point of SiO2.

[0159] Figure 9 shows how the outlet velocity changes with inlet temperature in a supersonic flow through a Draval nozzle, according to the law of conservation of mass.

[0160] Continuing with the same example, the range of the throat radius 182 of a suitable nozzle to obtain the desired temperature range 1410-1710°C is rchamber / r throat The ratio becomes 1.85 - 2.15. If we select the ratio 2.06, T e A temperature of approximately 1500 °C can be obtained. In this particular example, this corresponds to the throat radius r throat =160%·mm / 1.67=0.92·r c This corresponds to millimeters.

[0161] The throttling section 180 itself is exposed to very high temperatures. As mentioned above, flushing with an inert gas such as Ar protects the nozzle from hydrogen corrosion at high temperatures. This is because the inert gas surrounds the Si as a jacket (coating), protecting the nozzle surface. As an example, Ar flushing is performed at a volume flow rate of approximately 200% of the initial carrier gas supply amount (in this case) This was performed at 100 slpm. Such flushing also served to maintain the gas composition across the nozzle at a volume ratio of 50 / 50 Ar / H2. The nozzle dimensions were designed according to this volume gas fraction.

[0162] As described above, the reaction chamber 110 can be manufactured from molybdenum, and it can also be manufactured from tantalum. In one embodiment incorporating over-expansion, it may be manufactured from austenitic steel. The same applies to the nozzle material. As shown in Figure 10 (showing the phase diagram of the Ta-Si system), tantalum maintains a pure solid state along the 2260°C isotherm in the Ta-Si system.

[0163] In particular, by using a non-migration arc plasma, silica is completely dissociated, O - Ions are H + It reacts directly with ions. SiO x The complete dissociation of SiO means that Si does not inhibit further reduction. The SiO partially reduced in the radially outer portion of the plasma flow 163 is then subjected to hydrogen shrouding and confinement at the nozzle's convergence point, resulting in H +This leads to the rapid cooling by the throttling section 180, which prevents re-oxidation of Si. As a further advantage, the evaporated boron complex easily exceeds its evaporation temperature (e.g., 427°C) during flight, making it separable from the exhaust gas outlet 126 in the radial periphery. In other words, there is no need to achieve this temperature later in the melting process.

[0164] As described above, the first phase in the plasma flow 163 immediately after the outlet 164 of the plasma generator 160 has a high outlet temperature, for example, T = 5 to 11 kK. Silicon dissociation occurs in this region. The dissociated hydrogen and oxygen atoms in the plasma flow 163 recombine in the reaction chamber 110 in the temperature range T = 2 to 5 kK to form water vapor. Re-oxidation of Si and SiO in this temperature range is avoided by a non-equilibrium state. Subsequently, the silicon vapor recombines as vapor at T < 2 kK. t Rapidly quenching the gas flow from near a threshold temperature of 5kK helps to efficiently suppress the reoxidation of Si and SiO in the 2-5kK range.

[0165] The SiO reaction method 2SiO → Si + SiO2 should be avoided below 2.2 kK and over this temperature range (e.g., 1.8–5.0 kK) through rapid gas quenching by shrouding. Undesirable thermodynamic properties associated with these reactions can also be avoided below 2 kK, close to the melting point of Si (1410°C), by rapid aerodynamic quenching using a Draval nozzle.

[0166] The sealed structure of the reaction chamber 110 reduces mixing between the supply gases (at least between the supplied hydrogen and the inert gas). This is because, in the laminar flow state, all the gases move in almost the same axial direction within the reaction chamber 110. This results in the beneficial effect of radial non-uniformity of the gas composition. The low-temperature jacket (coating) of the plasma flow 163, where SiO is most abundant, is also the region where the concentration of single-atom hydrogen is highest due to flushing through the input lance.

[0167] Generally, the plasma generator 160, the first material supply means 130, the second material supply means 140, and the reaction chamber 110 are designed to realize the laminar flow of the type described above within and through the reaction chamber 110, for example, toward the throttling section 180, and, where applicable, toward the Draval nozzle. This may involve selections of flow rate, velocity, pressure, shape, etc.

[0168] In contrast, the throttling section 180, and in particular the Dragal nozzle used, can be configured to generate turbulent gas flow within the expansion chamber 120 and / or reaction chamber 110. Such turbulence may occur at the outlet of the throttling section 180 and its associated portions, and possibly throughout the entire expansion chamber 120.

[0169] As shown in Figure 6, the upstream portion of the expansion chamber 120 may be defined by a wall that extends axially (for example, downward) to the throttling portion 180. As a result, the diffusion portion 183 located downstream of the throttling portion 180 gradually moves into the expansion chamber 120.

[0170] In one embodiment, the diffusion half-angle 183a of the downstream diffusion section 183 is between 5° and 45°, for example between 6° and 34°, for example 19°. In any case, the diffusion half-angle 183a can be at least 10°, for example at least 15°, and / or the diffusion half-angle 183a can be up to 30°, for example up to 25°.

[0171] The expansion chamber 120 located downstream of the throttling section 180 is generally longer in the axial direction than the reaction chamber 110 located upstream of the throttling section 180. More specifically, the axial length 122 of the expansion chamber 120 may be at least 10 times, for example, 20 to 40 times, the axial length 112 of the reaction chamber 110. In the over-expansion flow embodiment, the expansion chamber 120 may be at least twice as long as the axial length 112 of the reaction chamber 110. Generally, the axial length of the expansion chamber 120 relative to the axial length of the reaction chamber 110 can be set such that the ratio of the volume of the expansion chamber to the volume of the reaction chamber exceeds the ratio of the pressure drop before and after the throttling section. This pressure drop is calculated using a well-known nozzle flow relation, where, for example, a 5-fold decrease in temperature results in a 146-fold decrease in pressure. For example, if the operating pressure of the reaction chamber is 100 mbar, the pressure in the expansion chamber will be 0.685 mbar (68.5 Pa).

[0172] As described above, the throttling section 180, and in particular the Dragal nozzle used, can be configured so that the condensed Si discharged from the throttling section 180 is guided toward and directly deposited upon the silicon molten material 191 provided in the container 190 below the outlet of the throttling section 180.

[0173] The kinetic energy of the material flow (preferably a supersonic flow) flowing out of the throttling section 180 is converted into thermal energy upon impact with the molten material 191. The amount of energy converted is proportional to the square of the average gas velocity minus the loss term due to entropy generation. As an example, in one embodiment, the target temperature is 1500°C, which is achieved from an inlet temperature of 6000K in the throttling section 180. In this case, the corresponding impact velocity is v e This becomes 3027 m / s, and assuming the isentropic expansion coefficient is constant as an approximation, if the hydrogen gas in argon is 50%, it corresponds to Mach 3.15.

[0174] Generally, the temperature of the material flowing out of the throttling section 180 is higher than the melting point of silicon, so the silicon particles within it are in a liquid state.

[0175] However, it is also possible to target the temperature of the material flowing out of the throttling section 180 to a temperature below 1410°C, the temperature at which silicon becomes solid. In this case, when high-temperature silicon particles collide with the surface of the molten material, the temperature rises and it becomes liquid again.

[0176] The dimensions of the throttling section 180 itself are fixed. However, by changing the flow rate of the unreactive seeds 171 and / or hydrogen gas 172 supplied to the reaction chamber 110 by the first material supply means 140, the unreactive seeds 171 and / or hydrogen gas 172 flow around the silicon material at a relatively high pressure through the throat 182 of the throttling section 180, thereby forming a quasi-virtual or fully virtual converging / diffusing nozzle. This efficiently compresses the silicon material radially inward through the throat 182. A virtual converging / diffusing nozzle refers to a state where no physical nozzle exists, and the throttling section is created by narrowing the flow of a particle-containing gas with a high-pressure, high-speed gas. Correspondingly, a quasi-virtual converging / diffusing nozzle amplifies the nozzle action by using such a converging gas. This allows the outlet velocity (and therefore temperature) of the material flow from the throttling section 180 to be controlled during operation.

[0177] In one embodiment, the apparatus 100 may include a third material supply means 150 configured to supply at least one of hydrogen 153 and a chemically unreactive species 154 to the expansion chamber 120. The chemically unreactive species 154 may be the same as the chemically unreactive species 142. In one embodiment, this is an inert gas, such as a noble gas such as Ar.

[0178] A third material supply means 150 may be configured to supply additional thermal energy to the expansion chamber 120, for example, directly to the molten material 191. For example, the third material supply means 150 may include a plasma torch 151, such as a non-migration plasma torch 151, that supplies hydrogen 153 and / or the chemically nonreactive species 154. In this case, the plasma flow 152 generated by the plasma torch 151 can be positioned to directly impact the surface 191a of the molten material 191 in the container 190. It should be noted that when using this term, the container 190 (or at least the molten material 191, or at least the surface 191a) is considered to be inside the expansion chamber 120. The plasma flow 152 on the surface 191a may be supplied at a position about 30 to 80 mm from the surface 191a and / or at an angle of 20° to 40° (e.g., 30°) with respect to the horizontal plane.

[0179] Alternatively, the third material supply means 150 may further include a plasma torch 151, such as a non-migration plasma torch. The plasma torch 151 supplies hydrogen 153 and / or the chemically nonreactive species 154 so that its plasma flow 152 is located below the surface 191a. The plasma flow 152 below the surface 191a may be supplied to a depth of 500 mm to 1000 mm from the surface 191a.

[0180] In one embodiment, the third material supply means 150 is configured to intermittently operate the plasma torch 151 using a duty cycle of, for example, up to 50% (e.g., between 20% and 50%).

[0181] In one embodiment, the third material supply means 150 is configured to supply hydrogen 153 and / or the chemically unreactive species 154 by applying electric plasma torch power of 0 to 20% compared to the electric plasma torch power applied to generate the plasma flow 163 of the plasma generator 160.

[0182] The third material supply means 150 may be configured to supply a mixture consisting of 0-10% hydrogen, 0-10% water vapor, and the remainder being Ar.

[0183] In addition to heating the molten material 191, the third material supply means 150 removes impurities from the molten silicone as described above.

[0184] As shown in Figure 6, the container 150 may be equipped with an impurity accumulation line 195 through which liquid impurities are discharged. In this case, the extraction line 195 is located below the surface 191a. Separation occurs by preferential separation, through which impurities in the silicon molten material remain in the liquid phase. The solid silicon is purified, but the impurities accumulate in the liquid phase. In one embodiment, this liquid phase is periodically removed.

[0185] The container may have an upper part 192 and a lower part 193 arranged to accommodate the surface 191a. The lower part 193 is located below the upper part 192 and is configured to receive Si arriving from the upper part 192 as a downward flow of molten Si. The extraction line 195 is typically located in the upper part 192 while the silicon is still in the liquid phase.

[0186] In one embodiment, the lower part 193 includes a heating element 194 positioned to control the height of the solid-liquid interface 191b of Si in the Si molten material 191. The solid-liquid interface 191b is typically located in the lower part 193. The heating element 194 may be an induction heating element or other conventional heating element.

[0187] The rate at this solid-liquid interface 191b can be controlled by thermocouple feedback control to a heating element 194 in the container 190. A silicon product with a purity of 99.9999% or higher, less than 1 ppba of boron, a phosphorus / boron ratio of 0.1-0.5, and a metal content of 0.5-30 ppba is envisioned. For example, the manufactured silicon product can be used as solar cell grade silicon or electronic component grade silicon, and can be divided into ultra-high purity polysilicon blocks and sold to Czochralski process refiners, etc.

[0188] As described above, this device may be equipped with one or more exhaust gas outlets 126, such as exhaust gas lances. The exhaust gas 127 discharged from the expansion chamber 120 through one or more exhaust gas outlets 126 contains OH, H2O, and metal oxides M. O、 The mixture may contain chemically unreactive species (such as Ar) and H2. These impurities are removed based on turbophoresis, Safman lift, and / or thermophoresis within the expansion chamber 120, and optionally by the shock-induced evaporation described above.

[0189] Figure 11 shows apparatus 200 according to another embodiment for producing Si and / or SiO2. Apparatus 200 is similar in many respects to apparatus 100, and reference numerals 210, 220, 226, 227, 240, 241, 242, 260, 263, 265, 280, 290, 291, 291a, and 295 have meanings corresponding to the reference numerals in Figure 6 above, with "1" as the first digit.

[0190] As shown in Figure 11, the chamber, in particular the expansion chamber 220, is provided with an inner wall 225, which is coaxially positioned along the main axis of the expansion chamber 220 as described above. Thus, the inner wall 225 is at least partially vertical and can be completely enclosed by the expansion chamber 220.

[0191] The inner wall 225 may be provided with one or more perforations 225a that penetrate the inner wall 225, thereby allowing gaseous material to flow out from the radial central region 223 of the expansion chamber 210 located inside the inner wall 225 to the radial outer region 224 of the expansion chamber 220 located on the outer periphery of the inner wall 225.

[0192] The inner wall 225 may be open at its upper end 225b. The inner wall 225 may be sealed at its lower end to the bottom of the expansion chamber 220 or to the container 290. The height of the inner wall 225 is in the range of 50% to 90% of the axial length of the expansion chamber 220, and the diameter perpendicular to the approximate flow direction within the expansion chamber 220 is greater than the corresponding diameter of the throat of the throttling section 280.

[0193] Furthermore, the peripheral region 224 may be connected to one or more peripheral exhaust gas outlets 226 located within the expansion chamber 220 as described above, enabling gas communication. The peripheral region 224 may be open at its upper end 225b, thereby allowing gas to flow from the peripheral region 224 to the central region 223.

[0194] In this way, the inner wall 225 forms a barrier that creates a pressure difference, maintaining a higher pressure in the central region 223 compared to the pressure in the surrounding region 224. High-speed material arriving from the constricted section 280 flows mainly directly into the central region 223 and towards the molten material 291. Si then deposits in the molten material 291. Impurities are discharged from the exhaust gas outlet 226 via the perforation 225a by the mechanism described above. Recirculation returns to the central region 223 via the upper end 225b. This makes the separation process of impurities other than deposited Si, and impurities discharged from the molten material 291 as a result of the Si flow colliding with the molten material 291, more efficient.

[0195] The apparatus 200 may also include a fourth material supply means 255 for supplying hydrogen gas 256 to the central zone 223 to obtain an additional reduction potential. In one embodiment, it is supplied at a volumetric flow rate of 10–50% of the main gas supply: that is, if 50 slpm of hydrogen is used as the plasma gas, 5–25 slpm of auxiliary hydrogen gas is required to reduce the product concentration and decrease the tendency of the system (silicon) to undergo a reverse reaction with H, in order to reduce the tendency to undergo a reverse reaction with 2O. In one embodiment, this hydrogen 256 (which may be supplied in addition to the hydrogen already supplied via the first, second 2240 and / or third 2250 material supply means in the general manner outlined above) is supplied directly to the central zone 223 as a gas stream, not through a plasmatron, but via a conventional lance or the like.

[0196] The container 290 may be in liquid communication with another chamber for various post-treatments of the generated liquid-phase silicon.

[0197] Figure 12 shows a first method for processing Si and / or SiO2 using the general types of apparatus 100, 200 described above.

[0198] In the first step, the method is initiated.

[0199] In the next step, plasma generators 160, 260 are operated to release plasma streams 163, 263 into reaction chambers 110, 210. Plasma streams 163, 263 contain Si and / or SiO2. In one embodiment, plasma streams 163, 263 contain no Si by mass ratio or up to 20%. In another embodiment, plasma streams 163, 263 contain no SiO2 by mass ratio or up to 20%. In the former case, the process is for reducing SiO2 to Si. In the latter case, the process is for remelting solid Si. In other cases, plasma streams 163, 263 contain both Si and SiO2, and the process is for both reduction and remelting.

[0200] In another step, the first material supply means 130 is activated to supply material 132 containing Si and / or SiO2 to the plasma generators 160, 260.

[0201] In another step, the gas supply means 170 is activated to supply the inert gas 171 and / or reducing gas 172 to the reaction chambers 110, 210 as described above.

[0202] In another step, the throttling sections 180 and 280 are operated to transport the gas from the reaction chambers 110 and 210 to the expansion chambers 110 and 210, thereby lowering the temperature of the gas passing through the throttling sections 180 and 280 below the condensation temperature of Si, SiO, SiO2, SiS, and / or SiS2. This causes the gaseous Si passing through the throttling sections 180 and 280 to condense into a liquid or solid phase.

[0203] In another step, the condensed Si (and / or its oxides and / or sulfides) is first introduced into containers 190, 290 without solidifying, where it is collected.

[0204] This method will be terminated in the following steps.

[0205] It should be noted that these steps are usually performed simultaneously in a continuous process. The resulting molten silicon may, if applicable, be moved downward to the lower part 193 of the container 190 to solidify there, and / or transferred to the chamber 297 for further processing.

[0206] Figure 13 illustrates a method for processing Si and / or SiO2. Since the embodiment in Figure 13 is similar to the embodiment in Figure 12, the following description will focus on the differences.

[0207] As the first step, this method will be initiated.

[0208] In the next step, the plasma generators 160 and 260 are operated to release plasma streams 163 and 263 into chambers 110, 120, 210, and 120 (for example, reaction chambers 110 and 210).

[0209] In another step, the first material supply means 130 is activated to supply material 132 containing Si and / or SiO2 to the plasma generators 160, 260.

[0210] In another step, the gas supply means 170 is activated to supply inert gas 171 and / or reducing gas 172 to chambers 110, 120, 210, and 120.

[0211] In the subsequent step, H2O is discharged from chambers 110, 120, 210, and 120 through exhaust outlets 126 and 226 located at the periphery. This can be achieved by utilizing turboforesis, Safman lift, thermophoresis, and / or shock-induced evaporation, as described above.

[0212] In another step, the condensed Si is first introduced into containers 190 and 290 without solidifying, where it is recovered.

[0213] This method will be terminated in the following steps.

[0214] It should be noted that the methods described herein are generally performed automatically or semi-automatically to control the process using control devices connected to various sensors (pressure sensors, temperature sensors, flow sensors, etc.) and actuators (flow control, temperature control, pressure control, etc.). Such control devices, sensors, and actuators may be known in themselves and are therefore not described in detail here.

[0215] Preferred embodiments have been described above. However, as will be apparent to those skilled in the art, many modifications can be made to the disclosed embodiments without departing from the basic concept of the invention.

[0216] For example, apparatus 100, 200 may have many further components not described herein. For example, various pre-treatment and post-treatment steps may be employed with respect to the silicon material to be processed. Furthermore, the various gas and material supply means 130, 140, 150, 170 may themselves be conventional supply means, with gases being supplied from a pressurized gas source via gas conduits and controllable valves, while solid materials may be supplied using screw feeders, belt feeders, pressurized air feeders, or any other conventional method.

[0217] In general, the information provided for any apparatus and method described herein can be freely combined, except where they are incompatible.

[0218] Therefore, the present invention is not limited to the embodiments described and can be modified within the scope of the appended claims.

[0219] As described above, the principles presented herein can also be used to produce compounds other than silicon in high purity.

[0220] In the application of the present invention to materials other than silicon, oxides and / or sulfides are vaporized above their dissociation temperature point (e.g., above 5 kK). Subsequently, gas quenching can be used to (a) protect the refractory materials used in the reactor for the reaction chamber and convergence / diffusion nozzle, and (b) create conditions for nucleating only the desired compound while removing H2O and other by-products. Subsequently, aerodynamic quenching on the convergence / diffusion nozzle isolates the compound from the reaction along its reverse and side reaction pathways. A table showing materials and their target quenching temperatures is shown below. Certain materials can be quenched to temperatures below their melting point, producing solid powder in the gas flow within the target vessel. The axial / auxiliary gas flow rate with respect to the plasma input power can be selected according to the Gibbs free energy of the oxide reduction reaction and the reactivity of the related compounds, and the ratio of the nozzle throat area A* to the reaction chamber area A is selected according to the desired target temperature range.

[0221] [Table 2]

[0222] For sulfides where roasting and reduction occur simultaneously with hydrogen, the following input values ​​can be used. [Table 3]

Claims

1. Apparatus (100, 200) for processing a metal or metalloid substance, its oxides and / or sulfides, Chambers (110, 120, 210, 120) equipped with exhaust gas outlets located around the periphery, A plasma generator (160, 260) configured to discharge a plasma stream (163, 263) containing the aforementioned substance, its oxides and / or sulfides into the chambers (110, 120, 210, 120), A first material supply means (130) configured to supply a material (132) containing the aforementioned substance, its oxide and / or sulfide to the plasma generator (160, 260), A gas supply means (170) configured to supply an inert gas (171) and / or a reducing gas (172) to the chambers (110, 120, 210, 120), The plasma flow (163, 263) is directed to a container (190, 290) which is placed inside the container (190, 290) and is configured such that the material does not first solidify but is deposited directly onto the molten material. Devices possessing (100, 200).

2. The plasma generator (160, 260) is configured to supply the plasma flow (163, 263) to the container (190, 290) at a certain velocity, and when this velocity exceeds a certain threshold effective Stokes number, such as 2.8, the Saffman lift causes H in the plasma flow (163, 263) 2 O moves radially outward along the direction of movement of the plasma flow (163, 263), and if the effective Stokes number is less than the threshold, H is controlled by turboforesis. 2 The velocity is sufficient to move O radially outward toward the exhaust gas outlet (126, 226). The apparatus (100, 200) according to claim 1.

3. The plasma generators (160, 260) are non-transition type arc devices that operate using DC voltage, AC voltage devices such as radio frequency (RF) or inductively coupled plasma (ICP), microwave plasma devices, or capacitively coupled plasma (CCP) devices. The apparatus (100, 200) according to claim 1 or 2.

4. The first material supply means (130) is configured to supply the material (132) circumferentially around the anode of the plasma generator (160, 260) or through the anode. The apparatus (100, 200) according to any one of claims 1 to 3.

5. The first material supply means (130) is configured to supply the substance, its oxides and / or sulfides as solid substances, oxides and / or sulfides, with an average particle size in the range of 20 to 200 μm, for example, 50 to 100 μm, preferably more than 99% of the particles being in this range. The apparatus (100, 200) according to any one of claims 1 to 4.

6. The substance, its oxides and / or sulfides supplied to the plasma generator (160, 260) are heated in the plasma stream (163, 263) so that the substance atoms in the plasma stream (163, 263) reach temperatures above the evaporation temperatures of B, Ti, Al, Fe, Ca, Na, Ni, P, and / or W under the associated operating pressure at the gas-liquid interface. The apparatus (100, 200) according to any one of claims 1 to 5.

7. The first material supply means (130) is configured to supply the plasma generators (160, 260) with at least one of the substance, its oxides and / or sulfides, as well as hydrogen and a chemically unreactive species (e.g., an inert gas), wherein the chemically unreactive species is Ar, He, or N 2 It could be The apparatus (100, 200) according to any one of claims 1 to 6.

8. The chambers (110, 120, 210, 120) comprise reaction chambers (110, 210) located upstream of the expansion chambers (110, 210), and the reaction chambers (110, 210) are axially separated from the expansion chambers (110, 210) by a throttling section. The apparatus (100, 200) according to any one of claims 1 to 7.

9. The gas supply means (170) is configured to supply a stoichiometric amount of hydrogen in order to completely reduce the oxides present in the reaction chambers (110, 210). The apparatus (100, 200) according to claim 8.

10. The system further comprises a second material supply means (140, 240) configured to supply at least one of hydrogen (141, 241) and a chemically unreactive species (142, 242) (e.g., an inert gas) into the reaction chamber (110, 210) around the plasma flow (163, 263), wherein the chemically unreactive species (142, 242) is Ar, He, or N 2 It could be The apparatus according to claim 8 or 9 (100, 200).

11. The second material supply means (140, 240) is configured to supply hydrogen (141, 241) and / or the chemically unreactive species (142, 242) to induce laminar flow in the radially outer sheath of the plasma reaction region within the reaction chamber (110, 210). The apparatus (100, 200) according to claim 10.

12. The second material supply means (140, 240) is configured to supply hydrogen (141, 241) and / or the chemically unreactive species (142, 242), As the supplied hydrogen (141, 241) and / or the chemically unreactive species (142, 242) flow into the throttling section (180, 280), a shroud (envelopment) (143) of the hydrogen (141, 241) and / or the chemically unreactive species (142, 242) is generated, and the shroud (143) is provided at the inlet around the material that reaches the throttling section (180, 280) from the plasma flow (163, 263). The apparatus (100, 200) according to claim 11.

13. The second material supply means (140, 240) is configured to supply hydrogen (141, 241) and / or the chemically unreactive species (142, 242) at a maximum temperature of 1500°C. The apparatus according to any one of claims 10 to 12 (100, 200).

14. The apparatus (100, 200) generates excess H in the reaction chamber (110, 210) under stoichiometric conditions. 2 It is configured to supply all of these materials via the second material supply means (140, 240). The apparatus according to any one of claims 10 to 13 (100, 200).

15. The apparatus (100, 200) provides at least 100% excess H to the stoichiometric conditions within the reaction chamber (110, 210) 2 Configured to supply The apparatus (100, 200) according to any one of claims 8 to 14.

16. The reaction chambers (110, 210) are vacuum chambers having a pressure of up to 800 mbar, for example, up to 100 mbar, for example, between 1 mbar and 50 mbar, for example, between 5 mbar and 20 mbar. The apparatus (100, 200) according to any one of claims 8 to 15.

17. The aforementioned throttling section is configured such that the condensed material flowing out of the throttling section is guided toward the molten material (191, 291) of the material placed in the container (190, 290) below the throttling section, and deposited directly onto the molten material. The apparatus (100, 200) according to any one of claims 8 to 16.

18. The axial length (122) of the expansion chamber (110, 210) is at least 10 times, for example, 20 to 40 times, the axial length (112) of the reaction chamber (110, 210). The apparatus (100, 200) according to any one of claims 8 to 17.

19. The first material supply means (130) provides the plasma flow (163, 263) with 30-70% carrier gas and 30-70% H in the form of chemically nonreactive species. 2 It is configured to include The apparatus (100, 200) according to any one of claims 1 to 18.

20. The carrier gas in the plasma flow (163, 263) has a temperature of 5 to 15 kK due to the energy applied to the gas via a DC or AC circuit. The apparatus (100, 200) according to claim 19.

21. The plasma generator (160, 260) uses H in the plasma flow (163, 263) 2 It is configured to heat up to at least 2kK. The apparatus (100, 200) according to claim 20.

22. The plasma generators (160, 260) are configured such that the temperature of the plasma flow (163, 263) immediately after the outlet (164) of the plasma generators (160, 260) is 5 to 15 kK. The apparatus (100, 200) according to any one of claims 19 to 21.

23. The system further comprises third material supply means (150, 250) configured to supply at least one of hydrogen (153) and a chemically unreactive species (154) (e.g., an inert gas) to the chambers (110, 120, 210, 120), The chemically non-reactive species (154) is Ar, He, or N 2 is The apparatus (100, 200) according to any one of claims 1 to 22.

24. The third material supply means (150, 250) comprises a non-migration plasma torch (151, 251) that supplies hydrogen (153) and / or the chemically non-reactive species (154), and the plasma stream (152, 252) is arranged to directly impact the surface (191a, 291a) of the molten material (191, 291) in the container (190, 290). The apparatus (100, 200) according to claim 23.

25. The third material supply means (150, 250) comprises a non-migration plasma torch (151, 251) that supplies hydrogen (153) and / or the chemically non-reactive species (154), and the plasma flow (152, 252) is configured below the surface (191a, 291a) of the molten material (191, 291) in the container (190, 290). The apparatus (100, 200) according to claim 23 or 24.

26. The third material supply means (150, 250) is configured to operate the non-migration plasma torches (151, 251) intermittently, for example, using a duty cycle of up to 50%. The apparatus (100, 200) according to claim 25.

27. The third material supply means (150, 250) is configured to supply hydrogen (153) and / or the chemically unreactive species (154) by applying electric plasma torch power of 0 to 20% compared to the electric plasma torch power applied to generate the plasma flow (163, 263) of the plasma generator (160, 260). The apparatus (100, 200) according to any one of claims 28 to 30.

28. The chambers (210, 220) are provided with an inner wall (225), the inner wall (225) is coaxially arranged along the main axis of the chambers (210, 220) and is provided with one or more perforations (225a) penetrating the inner wall (225), thereby enabling gas to be discharged from the radial central region (223) within the chambers (210, 220) located inside the inner wall (225) to the radial peripheral region (224) within the chambers (210, 220) located circumferentially outside the inner wall (225). The peripheral region (224) is connected to a peripheral exhaust gas outlet (226) located within the chamber (210, 220). The peripheral region (224) is open at the upper end (225b) of the inner wall (225), allowing gas to flow from the peripheral region (224) to the central region (223). The apparatus (100, 200) according to any one of claims 1 to 27.

29. The system includes a fourth material supply means (255) arranged to supply hydrogen gas (256) to the central region (223). The apparatus (100, 200) according to claim 28.

30. The container (190, 290) comprises an upper part (192) configured to contain the upper surface (191a, 291a) of the liquid material deposited from the plasma flow (163, 263), and a lower part (193) located below the upper part (192) and configured to contain the material arriving from the upper part (192), The lower part (193) includes a heating element (194) configured to control the solid-liquid interface (191b) of the substance within the molten substance (191, 291). The apparatus (100, 200) according to any one of claims 1 to 29.

31. The containers (190, 290) are equipped with liquid impurity removal lines (195, 295) located below the liquid level (191a, 291a) of the liquid substance deposited from the plasma flow (163, 263) and present in the containers (190, 290). The apparatus according to any one of claims 1 to 30 (100, 200).

32. A method for operating the apparatus (100, 200) according to any one of claims 1 to 31 for processing a substance that is a metal or metalloid, its oxides and / or sulfides thereof, The plasma generator (160, 260) of the apparatus (100, 200) releases a plasma stream (163, 263) containing the substance, its oxides and / or sulfides into the chambers (110, 120, 210, 120) of the apparatus (100, 200), The first material supply means (130) of the apparatus (100, 200) is used to supply a material (132) containing the substance, its oxide and / or its sulfide to the plasma generator (160, 260), The gas supply means (170) of the apparatus (100, 200) is used to supply inert gas (171) and / or reducing gas (172) to the chambers (110, 120, 210, 120), H 2 To discharge O from the chambers (110, 120, 210, 120), The method comprises introducing the condensable substance into the containers (190, 290) of the apparatus (100, 200) without first solidifying it in a condensed form, and recovering the condensable substance into the containers (190, 290). method.

33. The substance is selected from the group including silicon, titanium, germanium, and nickel. The apparatus (100, 200) or method described in any one of the preceding paragraphs.

34. The substance is Si, and the oxide is SiO 2 That is The apparatus (100, 200) or method according to claim 34.

Images