WO2013007923A1 - Electrochemically-assisted method for preparing monosilane - Google Patents

Abstract

Method for producing monosilane (SiH₄) comprising at least one step a) of reaction in a reactor (1) of at least one silicon alloy of formula M1_xM2_ySi_z, in which M1 is a reducing metal, M2 an alkali or alkaline-earth metal, x, y and z vary from 0 to 1, z being other than 0 and the sum x + y being other than 0, with an acid solution, said reaction being electrochemically assisted by cathodic polarization of the alloy in the acid solution by application of a current density or a voltage such that the molar degree of conversion of silicon contained in said alloy to monosilane is greater than at least 80%.

Description

 PROCESS FOR THE PREPARATION OF ELECTROCHEMICALLY MONOSILANE

 ASSISTED

The present invention relates to a method of generating monosilane (SiH) having a very high efficiency assisted by the application of a current or a voltage to an alloy containing silicon in an electrochemical reactor containing an acid solution.

Monosilane, or silicon tetrahydride (SiH), is used as a silicon vector in deposition techniques of amorphous silicon, polycrystalline silicon, nanocrystalline silicon or microcrystalline silicon also called nano or micromorph, silica, silicon nitride, or other compound of silicon for example in vapor deposition techniques.

 Thin film deposition of amorphous silicon, and microcrystalline silicon obtained from silane, make it possible to manufacture solar cells.

 It is also possible to obtain coatings resistant to acid corrosion, by cracking silane and to manufacture compounds such as silicon carbide.

 Finally, the silane is capable of adding to the single or multiple bonds of the unsaturated hydrocarbons to give organosilanes.

 The monosilane market will experience a very strong expansion both for the manufacture of integrated semiconductors, the manufacture of solar cells (photovoltaic), semiconductor components and the manufacture of flat screens.

Several types of processes described below have been used until now. First, the reduction of SiCI ₄ by LiH in a KCl / LiCl bath at temperatures between 450 ° C and 550 ° C is known. The efficiency of the reaction is interesting, but the process relies, firstly, on the availability of LiH while the lithium resources are very limited, and secondly, on the possibility of recycling the lithium metal by electrolysis. The reaction medium is very corrosive and uses particular materials. This process has been used to produce small amounts of silane. The reduction of SiF by NaAIH in organic solvent medium is another example. This process is industrially viable only when there is SiF, a by-product of another chemical production and sodium to make sodium aluminum hydride. This method is not easily usable, especially for these two reasons.

Another known reaction is the acid attack in a liquid NH ₃ medium of a stoichiometric alloy Mg ₂ Si. The reaction is as follows:

Mg ₂ Si + 4 HCI SiH ₄ + 2 MgCl ₂

 NH3 liq.

This process is carried out at a temperature close to ambient temperature at atmospheric pressure. Magnesium silicide (Mg ₂ Si) has been extensively tested in aqueous media and ammonia. Although sufficiently acceptable to lead to industrial production units, the process has the following major drawbacks:

 · Industrial magnesium silicide due to the volatility of magnesium contains only 70% to 80% of the stoichiometric compound. The conditions of manufacture of the stoichiometric compound make it a product too expensive for this industry.

 In parallel with the production of monosilane by this route, many higher silanes including polychlorosilanes, siloxanes and silicone gums are manufactured, making the balance of monosilane material unattractive and inducing significant difficulties in the management of the process.

 This process is not satisfactory because of the difficulty of controlling the process and the implementation of highly regulated liquid ammonia.

Another known reaction is the disproportionation of SiHCl 3 on resins containing grafted amino groups or the like. The complete process can be described as follows: a) 4 Si Métai. + 12 HCl -> 4 S1HCI3 + 4 H ₂ (temperature between about 300 ° C and about 1000 ° C)

b) 4 SiHCl3 <-> SiH ₄ + 3 SiCl ₄ (temperature close to ambient)

c) 3 SiCl ₄ + 3H ₂ -> 3 SiHCI ₃ + 3HCl (temperature of about 1000 ° C.),

 the following reaction report:

4 Si Metal + 9 HCI SiH ₄ + 3 SiHCI ₃ + H ₂

A variant of the above reaction is described as follows: a) 4 Si Metai + 6 HCl -> 4 SiCl ₄ + 8 H ₂ (temperature between about 1000 ° C and about 1100 ° C)

b) 4 SiCl ₄ + 4H ₂ -> 4 SiHCI ₃ + 4HCl (temperature around 1000 ° C)

4 SiHCIs SiH ₄ + 3 SiCl ₄ ,

 the following reaction report:

4 If _m + 12 HCI ^ SiH ₄ + 3 SiCl ₄ + 4H ₂

This process requires high temperatures in an extremely corrosive environment and consumes a lot of energy (about 50 kWh / kg for step b)). To achieve maximum efficiency, step b) requires many recirculation loops of chlorosilane mixtures. In addition to the use of highly corrosive, toxic and flammable products, such a type of process is very energy-intensive and presents a lot of industrial risks.

The generation of monosilane and higher silanes has been described in the Handbook of Inorganic Chemistry Gmelin Si-Silicon, by reacting in the aqueous phase, silicon silicides and alloys in acidic or basic medium. In the patent applications EP146456 and WO2006 / 041272, the synthesis of monosilane in aqueous phase by dropping a powder of Al _x Si _y Ca _z , x, y and z representing the percentages respectively of aluminum, silicon and calcium, in a solution of HCI, is described. By-products resulting from such a reaction are produced in large quantities and are harmful to the environment (in particular chlorides). Another disadvantage of such a process is the formation in abundance of a foam in the reaction medium, which decreases the yield of the reaction and requires the presence of an anti-foaming agent. Such a reaction is very exothermic and temperatures above 100 ° C are reached fairly quickly if the rate of introduction of the alloy powder is not significantly reduced.

In patent US 5510007, it is proposed to generate silane, from the attack of an Mg ₂ Si alloy by electrochemically generated protons according to the reaction H ₂ → 2 H ⁺ + 2 e ^" . the generation of protons, later exploited to form silane.

All these works described above do not guarantee the conditions necessary for the realization of a profitable process for an industrial development. The development of processes involving less difficult reaction conditions and or allowing to be used for small and medium units in almost all environments and close to the use of monosilane is a major challenge for the industries mentioned above. The silane generation process by chemical etching of a silicon alloy is an interesting process but has at least three major disadvantages: the silane yields are low, the H ₂ / SiH ratio in the gas mixture is not favorable and generates significant amounts of solid residues. In addition, the silane produced in this way must be transported from the silane manufacturing site to the packaging plant and finally to its consumer location. Since the silane is highly pyrophoric and toxic, many safety measures must be taken to transport the silane to its final destination. For users, the silane is provided in storage means to be disposed at a protected location. Thus, silane transport is an expensive and dangerous process. Attempts have been made to avoid the risks of silane transport and storage by producing silane at the site of its use. However, such silane generators have been limited to the formation of species such as those produced by the excitation of hydrogen plasma, which react with a source of silicon to create the silane. Such techniques produce too little silane and are generally too complex to meet the needs of a large scale production facility. Thus, there is a need for silane production at the very location of the point of use of the silane.

An object of the present invention is to overcome all or part of the disadvantages of the prior art noted above.

To this end, the present invention relates to a process for producing monosilane (SiH) comprising at least one reaction step a) in a reactor of at least one silicon alloy of formula M ¹ _× M ² y Si _z , in which M ¹ is a reducing metal, M ^{2 is} an alkali metal or alkaline earth metal, x, y and z vary from 0 to 1, z being different from 0 and the sum x + y different from 0, with an acid solution, said reaction being electrochemically assisted by cathodic polarization of the alloy in the acid solution by application of a current density or a voltage such that the molar conversion rate of silicon contained in said alloy to monosilane is greater than at least 80%.

 Preferably, the voltage or current density applied to the cathode is such that the molar conversion rate of silicon contained in said alloy to monosilane is greater than at least 90%.

Furthermore, embodiments of the invention may include one or more of the following features: A method as described above, characterized in that said reactor comprises: a cathode compartment in which is inserted a working electrode a material containing said silicon alloy; an anode compartment in which a counter electrode is inserted; both electrodes being immersed in said acidic solution and connected to a voltage or current source.

A method as described above, characterized in that the reactor further comprises a reference electrode, the three electrodes being immersed in said acid solution and connected to a voltage or current source. A process as described above, characterized in that M ¹ is aluminum and

M ² is calcium or magnesium.

A process as described above, characterized in that the silicon alloy has the formula CaAl ₂ Si ₂ .

A process as described above, characterized in that the acidic solution is hydrochloric acid.

A process as described above, characterized in that the concentration of the acid solution is between 0.6 mol / L and 3.7 mol / L and preferably between 2.5 mol / L and 3.7 mol / L.

A process as described above, characterized in that the temperature of the reaction is kept constant at a value between 30 ° C and 85 ° C. Of preferably the temperature is between 50 ° C and 85 ° C. The optimum temperature for reaction a) of the process which is the subject of the present invention is between 70 ° C. and 85 ° C. A method as described above, wherein said reaction a) is electrochemically assisted by cathodic polarization of the alloy in the acid solution by application of a current density, said current density applied to the cathode is between - 10 mA / cm ² and -250 mA / cm ² and preferably between -100 mA / cm ² and-250 mA / cm ² .

A method as described above, wherein said reaction a) is electrochemically assisted by cathodic polarization of the alloy in the acid solution by applying a voltage, said voltage applied to the cathode is between -0.55 V / ENH and -2.5 V / ENH (-0.795 V / ECS and -2.745 V / ECS), more preferably between -1 V / ENH and -2 V / ENH (-1, 245 V / ECS and -2.245 V / ECS).

A method as described above wherein the anode material is of the same nature as the cathode material. A method as described above, comprising a step prior to step a) during which a purge and an inerting of the reactor are established before any contact between the acid solution and the silicon alloy.

Purge consists of performing compression-expansion cycles to flush the air from the pipes and the reactor. Inerting aims to "drown" the silane in nitrogen, for example, to prevent it from exploding, since it is pyrophoric.

The silicon alloy is selected from CaAl ₂ Si _2, Sio.sMg, Sio.sCa, AlSiCa, CaSi, Ca _{0, 5} Si, MgSi, AISiNa, AISiMg, SiNa, AISiLi, SiK, Ca _{0, 5} AISio, 33 and Cao, AlSi _{5 0,} 75, or a mixture thereof, preferably Sio.sMg, AISiNa, SiNa, Sio ₂ SLI Sio ₂ ANS Sio ₂ sK, or SiK. Other silicon alloys that are suitable for the present invention are ferosilicon-type alloys, for example FeSi, FeSiMg, FeSiCa. Preferably, the alloy used in the process that is the subject of the present invention is the CaAl ₂ Si ₂ composition, which is the most active phase giving the best yields.

 The proposed method for silane generation using an electrochemical-assisted process improves yields by limiting spurious reactions from taking place, thus leading to a considerable lowering of production costs. In fact, with the chemical process known from the state of the art, the molar conversion rate of silicon to silane does not exceed 15% of the total silicon of the alloy. In fact, 85% of the silicon contained in the alloy gives rise to parasitic reactions or does not react. The present invention proposes to increase, by means of the application of a current or a voltage, the amount of silicon converted to silane and to limit the quantities which serve as parasitic reactions.

One of the parasitic reactions is the oxidation of silicon, contained in a CaAl ₂ Si ₂ or Mg ₂ Si alloy, in SiO 2, or in higher silanes (Si ₂ H ₆ or Si ₃ H ₈ ) to the detriment of monosilane. The method according to the present invention avoids these reactions by applying a current or a potential whereby the formation of S1O2, Si2H ₆ and S13H8 is thermodynamically impossible.

The alloys used for the synthesis of silane by a chemical process are generally binary alloys of magnesium and silicon, preferably of Mg ₂ Si composition, or ternary of aluminum, calcium and silicon, preferably stoichiometry CaAl ₂ Si ₂ . Whatever the alloy used, larger or smaller amounts of hydrogen are generated. However, the formation of hydrogen on the surface of a metal or a metal alloy (or any other conductive material), in an acid medium, is the result of the reduction reaction of the protons present in this medium, depending on the reaction. : o 2 H ⁺ + 2 e ^" → H ₂ (g) (1)

Protons form hydrogen on the surface of metals, a source of electrons. It is therefore obvious that the generation of silane by acid etching of a silicon alloy involves electrochemical reactions. The process according to the present invention makes it possible to control the kinetics of the reactions produced during the silane generation process and to improve their yield by means of a device making it possible to apply a current or voltage to a material containing silicon, preferably Mg ₂ Si or CaAl ₂ Si ₂ .

Analysis of the reaction system indicates that certain reactions occurring during silane generation involve electrons as summarized in the diagram below: CaAl ₂ Si ₂ + 8H ⁺ → 2 SiH ₄ + 2 Al ³⁺ + Ca ²⁺ (2)

o CaAl ₂ Si ₂ + 6 ^÷ H → Si ₂ H ₆ + 2 Al ³⁺ + Ca ²⁺ + 2e ^"(3)

o 3 CaAl ₂ Si ₂ + 16 H ⁺ → 2 Si ₃ H ₈ + 6 Al ³⁺ + 3 Ca ²⁺ + 8 e ^"(4)

o CaAl ₂ Si ₂ + 4H ₂ O → 2 SiO ₂ + 2 Al ³⁺ + Ca ²⁺ + 16 e ^" + 8 H ⁺ (5) Reactions (3) to (5) have the double disadvantage of consuming silicon to the detriment of the production of the silane and to provide the electrons necessary for the formation of the hydrogen according to the reaction (1). Their blocking, or slowing down, by an adequate cathodic polarization of the alloy in the acidic solution, has for its consequence of maintaining all the silicon of the alloy available for the generation of monosilane according to reaction (2), purely chemical.

Other particularities and advantages will appear on reading the following description, made with reference to FIG.

 FIG. 1 represents the diagram of an electrochemically assisted silane generation reactor in which the current (or voltage) source is a potentiostat.

In the present invention, the silane generation requires the use of an electrochemical reactor for example as that described in FIG. Said reactor 1 comprises: a cathode compartment 2 in which is inserted a working electrode 3 of material containing said silicon alloy as described above according to the present invention; an anode compartment 4 in which is inserted a counter electrode 5. The reactor further comprises a reference electrode 6. The three electrodes are immersed in an acid solution, source of protons, and connected to a potentiostat 7 (a voltage source / or current). Preferably the acidic solution is hydrochloric acid (HCl). According to the present invention, the cathode is a silicon-containing alloy, preferably a ternary alloy of calcium, aluminum and silicon, more preferably an alloy containing a CaAl ₂ Si ₂ , CaSi and / or Ca ₂ Si phase, or a binary alloy of magnesium and silicon, preferably containing a Mg ₂ Si phase.

In order to increase the exchange surface, according to a particular embodiment of the invention, the cathode and anode compartments are separated by a diaphragm or an ion exchange membrane. On the other hand, the use of a fixed bed cathode of particles of the selected alloy (CaAl ₂ Si ₂ , Mg ₂ Si ...) can be implemented. The anode material may be of conductive material, slightly corrosive in acidic media such as titanium or one of its alloys, graphite or noble metals. Advantageously, the material placed in the anode may be the same as that of the cathode so that it dissolves before the oxidation of the water (formation of O ₂ from the reaction: 2 H ₂ O → O ₂ + 4 H ⁺ + 4 e ^" (6)) and gives rise to the formation of the same ions in solution The use of anode material of the same nature as the cathode material is beneficial for several reasons: a) - To avoid the formation of oxygen on its surface (safer reactor) b) - Homogeneity of ions in solution in the two compartments of the reactor c) - Production of gaseous by-products with high added value at the anode ( SiH ₄ ...) Any reference electrode can be used, preferably a calomel electrode saturated with KCl (E = 0.245 V / ENH), to avoid contaminating the solution with anions other than chlorides. use of an electrolytic bridge, containing the same acid as that contained in the co compartments 2 and 4 may be recommended. The use of a mercurous sulphate reference electrode may be advantageous if the sulfuric acid is used as the electrolyte in the two compartments.

Any acidic solution, or other source of protons, may be suitable for this application; in particular the mineral acids of H ₂ SO ₄ or HCl type, preferentially HCI. The concentration of this acid must be greater than 0.6 mol / L or at a pH below 0.3, preferably between 0.6 and 3.7 mol / L and more preferably between 2.5 and 3.7 mol. / L.

The temperature of the electrolytic bath (acid solution) must be kept constant using a thermostatic bath, connected to the two compartments 2 and 4 (anode and cathode). The temperature of the bath is set at a value greater than 30 ° C., preferably between 50 ° C. and 85 ° C. and more preferably between 70 ° C. and 85 ° C.

A hot fluid is injected via an inlet 9 into the compartments 2 and 4. The water may be used as a reactor heating fluid, however any other mineral oil type fluid may be employed. The temperature of the thermostat should be adjusted so that the electrolyte inside the reactor is at the temperature recommended above. The cold fluid outlet 8 indicates that to heat the acid volume necessary for the reaction, the hot fluid leaves calories. Several sources of voltage (or current) can be used in the context of the method that is the subject of the present invention.

For example, the voltage source may be a DC voltage generator on which the positive (+) pole is connected to the anode material and the negative (-) pole to the cathode material (Si-containing alloy). The DC voltage delivered by the latter must be between 0 and 4.5 V, preferably between 0.5 and 1.5 V and more preferably between 0.8 and 1, 2 V. A battery, a battery or any other source voltage that can deliver an electromotive force (emf) equivalent to the potentials indicated above, may also be used.

A potentiostat will be preferred to a conventional generator because it allows, on the one hand, a precise control and monitoring of the potential (or current) applied to the electrode (alloy cathode containing Si) and the current response (or in voltage), as well as access to data such as the amount of filler consumed, on which the amount of silane depends, or any other byproduct generated by the electrochemical reactions. On the other hand, several forms of the current signal (or input voltage) can be applied to the system using this device. In potentiostatics, a potential lower than -0.55V / ENH (normal hydrogen electrode), corresponding to -0.795V / ECS (calomel electrode saturated with KCl), preferably between -0.55 and -2.5 V / ENH (-0.795 and -2.745 V / ECS), more preferably between -1 and -2 V / ENH (-1, 245 and -2.245 V / ECS) can be applied to the silicon alloy cathode.

In intensiostatic, a cathodic current density of less than -10 mA / cm ² , preferably between -10 and-250 mA / cm ² , more preferably between -100 and -200 mA / cm ² .

A voltage or a pulsed current (as illustrated in FIG. 2) may also be applied, making it possible to have a better renewal of the electrolyte on certain parts of the alloy electrode 3, a good distribution of the current lines and a uniform dissolution of the alloy. This will also make it possible to exploit intermittently the advantages of each of the potentials, or intensity of current, applied (limits of the intervals previously defined). For example, a voltage pulse in which Ei = E ₂ = -0.55 and -2.5 V / ENH and preferably Ei = E ₂ = -1 and -2 V / ENH, may be applied to ensure a potential range where the formation of SiH from Si is thermodynamically possible (well below the potential at which silane is formed (E si / siH4 -0.14 V / ENH)), without generating too much Η ₂ , then impose a potential much lower than the formation potential of S1O2, thus regenerating the Si which can be transformed into S1O2, if the potential of E1 is not sufficient to prevent S1O2 from forming. The durations of application of E 1 and E ₂ voltages can be 10% of the total pulse duration (τ) for ti and 90% of τ for t _2.

In any case, because of the high reactivity of the silane, and its pyrophoric character, it is preferable that its generation using a reactor such as those described above is done in an inert environment. Purge and inerting (nitrogen for example) compartments is to operate before any contact between the acid and the alloy.

Claims

1. A process for the production of monosilane (SiH) comprising at least one step a) of reaction in a reactor (1) of at least one silicon alloy of formula M ¹ _× M ² y Si _z , in which M ¹ is a reducing metal, M ² an alkali metal or alkaline earth metal, x, y and z vary from 0 to 1, z being different from 0 and the sum x + y different from 0, with an acid solution, said reaction being electrochemically assisted by cathodic polarization of the alloy in the acid solution by applying a current density or a voltage such that the molar conversion rate of silicon contained in said alloy to monosilane is greater than at least 80%. 2. Method according to claim 1, characterized in that said reactor (1) comprises: a cathode compartment (2) in which is inserted a working electrode (3) of material containing said silicon alloy; an anode compartment (4) in which a counter electrode (5) is inserted; both electrodes being immersed in said acid solution and connected to a voltage or current source (7). 3. Method according to claim 2, characterized in that the reactor (1) further comprises a reference electrode (6), the three electrodes being immersed in said acid solution and connected to a source of voltage or current (7) . 4. Method according to one of the preceding claims, characterized in that M ¹ is aluminum and M ² is calcium or magnesium. 5. Method according to one of the preceding claims, characterized in that the silicon alloy has the formula CaAl ₂ Si ₂ . 6. Method according to one of the preceding claims, characterized in that the acidic solution is hydrochloric acid. 7. Method according to one of the preceding claims, characterized in that the concentration of the acid solution is between 0.6 mol / L and 3.7 mol / L and preferably between 2.5 mol / L and 3, 7 mol / L. 8. Method according to one of the preceding claims, characterized in that the temperature of the reaction is kept constant at a value between 30 ° C and 85 ° C, preferably between 50 ° C and 85 ° C and more preferably between 70 ° C and 85 ° C. 9. Method according to one of the preceding claims, wherein said reaction a) is electrochemically assisted by cathodic polarization of the alloy in the acid solution by applying a current density, said current density applied to the cathode is included between -10 mA / cm ² and -250 mA / cm ² and preferably between -100 mA / cm ² and -250 mA / cm ² . The method according to any one of claims 1 to 8, wherein said reaction a) is electrochemically assisted by cathodic polarization of the alloy in the acid solution by applying a voltage, said voltage applied to the cathode is between -0.55 V / ENH and -2.5 V / ENH (-0.795 V / ECS and -2.745 V / ECS), more preferably between -1 V / ENH and -2 V / ENH (-1, 245 V / ECS and -2,245 V / ECS). 1 1. Method according to one of claims 2 to 10, wherein the anode material is of the same nature as the cathode material. 12. Method according to one of the preceding claims, comprising a step prior to step a) during which purge and inerting of the reactor (1) are established before any contact between the acid solution and the silicon alloy. .