DE2103255A1 - Reductn of reactive metal oxides - with carbon without formation of c using at solvent metal - Google Patents

Abstract

Carbothermic reduction of reactive metal oxides comprises reacting the oxide with a stoichiometric amt. of carbon at temps. above 600 degrees K in a molten metal solvent which reduces the activity of the metal formed during the reduction reaction to a level below that required for the metal to form a carbide under the reaction conditions in the system, and maintaining the partial pressure of the carbon monoxide present in the system at a level below the theoretical equilibrium value for the system as the reduction of the metal oxide takes place.

Description

Carbothermal process for reducing an oxide to a reactive one Metal.

The invention relates to an earbothermal method for Reduction of an oxide from a reactive metal. Such procedures are already known.

However, the practical importance of the known processes is proportionate low as the product received; a mixture of the free metal with metal oxides and carbides, and it is uneconomical, the metal component of such Burn off mixtures from the residues.

The invention is the task of train @ unde, a landlord - @@ haftll @ h @ interesting carbothermal reduction process of the @rt in question indicate the one enables good yield in the conversion of metal oxides into the form @es Met @ lls, without at the same time metal carbides, oxycarbides and other undesirable products in link thus arise, and that the extraction of these metals made possible in substantially pure form from the resulting reaction mixture.

The new method according to the invention is mainly thereby characterized that this oxide with a stoichiometric amount of carbon at temperatures silver about 600 ° Kelvin in a molten metal Solvent is brought to reaction that the activity of the during the reducing Reaction of formed metal reduced to a level below that which for the metal would be necessary under the reaction conditions prevailing in the system to form a carbide, and that the partial pressure of the carbon monoxide present in the system is maintained at a level below the theoretical equilibrium value for the system while the reduction of the metal oxide takes place.

It has been found that the disadvantages of the already known oarbothermlscheri reduction processes can be overcome in that the Reaction between carbon and oxides of reactive metals (see Equation 1) carried out in a bath of a molten metal as a solvent that has the ability to monitor the activity of what is shown during the reaction and dissolved in the molten solvent existing metal to reduce a level below that at which the metal is noticeable reacts with carbon to form a carbide. Tin, copper and iron are those preferred metals which come into consideration as solvents for this purpose. Each of them is especially good for using with a specific educated Metal or with a group of such metals. It can also be different other metals can be used either as such or in the form of alloys. In particular, the system used should be of such a type that the formed Metal present in solution under the prevailing conditions of Temperature and concentration has an activity that is below the activity level in which this metal is present in the pure state with carbon reacts to these temperatures to form corresponding metal carbides.

Further, the reaction is carried out under such conditions that the partial pressure of carbon monoxide above the molten bath is less than the theoretical equilibrium level as derived from the constant values of the free Energy and equilibrium are calculated for the particular reaction under consideration so that the reaction can be brought to completion. The CO partial temperature Can be done by common methods such as drawing a vacuum on the system or brushing the Solution can be maintained with an inert gas.

A gas such as argon is preferably used for this purpose, which has a low thermal conductivity. At such relatively high temperatures where Nitrides of the formed metal can decompose, nitrogen gas can also for the purpose serve for CO coating.

According to another embodiment of the process, the reaction can according to the invention also by inserting a metal nitride-forming stage in the A sequence of procedural steps can be carried out. When working on this In this way, the desired partial pressure of nitrogen is maintained over the system, while the metal oxide reacts with carbon in the molten solvent. The overall equation for the reaction is given below as equation (2). A Nitrogen gas applied in this manner can also be used, either alone or in conjunction with an inert gas like argon, serve to remove the CO as quickly as it is formed and in this way to produce the desired CO pressure. These The solvent assisted reaction results in the formation of a solid nitrate product from the: oxide, which is practically free from any noticeable pollution by oxides, Carbide or other foreign peat is. If desired, this solid product can be used al such from the molten metal solvent and thus from all other materials (such as Fridstoffen in the used Metal oxide feed) present in the reacting mixture could be. Such separated nitride products are, for example, used as abrasives, can be used as steel alloy additives as well as for other purposes and can therefore can be used accordingly. However, the nitride-forming reaction is easily reversible, when carried out in the molten metal solvent in the sense that by reducing the partial pressure of nitrogen above the system to corresponding low levels the nitride compound is decomposable, resulting in development of nitrogen gas and the formation of a solution of the residual metal in the tin or the other molten solvent metal (see equation (3) below).

This reduction in the partial pressure of nitrogen above the system can either by vacuum pulling or by brushing the space with an inert gas happen.

In addition to creating a process by which the metal solution in the solvent made of metal (by separating the solid nitride compounds from the reaction mixture and then adding the nitride to a pure one Bath of molten metal solvent, which process the Nitride dissolution and formation is, if desired, repeatable) of impurities can be freed, the introduction of the nitride-forming process stage in the Process still have other advantages with certain ones made of metal oxide Raw materials. Thus, the use of nitrogen increases with the result of a Formation of the nitride the thermodynamics of the carbothermal reduction, because through doing this will free the deoxidizing reactions with great positive Energies changed to those that have negative free energy values. The metal oxides, which have a particularly beneficial effect in this regard are discussed below with regard to discussed in more detail on their suitability as starting materials.

However, if the solution of the desired metal in the molten Solvent is formed by the application of the metal known as such Trennverrahren can be recovered. After such a process, the tin can or the metal otherwise used as a solvent from the solution by heating be distilled off by means of an electron beam, so that, for example, the desired Metal as RUcksXand remains when the electron beam heating method is used, this operation is performed in a vacuum.

Furthermore, liquid phase separation methods can be used and other methods can also be specified.

When the method according to the invention in the bath of a closed Solvent below one atmosphere, which is free of nitrogen can be carried out in the molten metal solution Reaction can be expressed roughly as follows: MO (s) + 2C (8) = 2 CO (g) + M (1) or, more generally, as follows MxOy (s) + yC (s) = xM + y CO (g) (ia) In the following and other equations given here, M means that to be recovered Metal, M means the same metal in a dissolved state, and (s) and (g) mean solid or gaseous components.

The letter (1) means one liquid component in each case.

When the reaction occurs in the presence of a nitrogenous atmosphere is carried out, the overall reaction that takes place in the solution can be express as follows: M02 (s) + 2C (s) + 1 / 2N2 (g) MN (s) + 2cO (g) (2) In the corresponding This reaction can generally be represented, if desired, in the form (Ia).

It should also be noted that in addition to that according to equations (1), (la) and (2) produced CO still small amounts of CO2 with different Ne tafloxl that can be generated. However, this does not affect the overall reaction, with which the invention is concerned.

As stated above, the system using the solid MN product contains, processed by one of several methods so that the solid MN product converted into dissolved components of metal and gaseous nitrogen will. This conversion can be carried out using either the original reaction mixture or such a mixture can be effected by separating out the solid MN product and adding it to a fresh, pure bath of the molten metal existing solvent is obtained. The reaction that takes place can be like represent as follows: MN (s) = M + 1 / 2N (g) (3) Another phase separation method for Elimination of the reactive metal from the solvent, which is especially suitable for a system in which titanium is the reactive product is suitable includes the hydride formation. TiH becomes hydride stable under a hydrogen atmosphere Temperature formed. The TiH floats on the solvent and can be skimmed off will.

Another phase separation process that is used for a system with aluminum as a reactive metal and tin as a solvent only requires cooling the melt down to a temperature at which the aluminum is precipitated, while the tin is still is in liquid form.

The reaction of the present invention is carried out at elevated temperatures carried out in the range from about 6000 to 2300 K, with the temperature in individual cases is chosen in each case so that the metal serving as the solvent is in the desired liquid state is obtained and the desired concentration of the oxide metal formed during the reaction is created in the solution by about 1% to 50 ß or more. It is also known to affect the temperature of the solution, the activity of the metal and its nitride and carbide-forming properties. The selection of the metal concentration and at least to some extent the working temperatures should be based on economic factors. This is how the preferred work levels fall for the metal concentration in the molten solvent in a range from about 3 to 40 wt.%, while a preferred temperature range is in the range of about 1000 ° to 20,000 K.

Dis heat that is required to complete the various stages of the process and the individual parts of the apparatus used in connection therewith on the desired Maintaining temperatures can be supplied in the usual way, for example by using electrical Induction heaters. If the latter are applied to graphite tubes or reactor vessels (which are conductive and therefore getting hot on its own), precautions should be taken to avoid it to protect the outer graphite surfaces against contact with the atmosphere. For this the graphite apparatus can be coated with a fire-resistant barrier material such as mica or aluminum, which are preferably non-conductive intended to prevent a short circuit of the electric current when the barrier material comes into contact with any work coils. The coils or windings of the Heating device can also be externally covered with a thermally insulating material be wrapped so that the heat is retained and the operators against burns protected by contact with the heating device.

A molten bath is used to carry out the method according to the invention of a metal e.g. tin is used as a solvent, which the metal oxides in addition to Carbon can be supplied in such stoichiometric amounts that at least a reaction takes place with the oxygen components of the oxides to convert them into carbon monoxide to convert. Preferably the carbon is used in a moderate excess of, for example 0.1 to 10 ß above the atoichiometric requirement added. For small-scale operations or when working with extreme valuable metal oxides, it is economically feasible to provide the required carbon in such a way that a graphite crucible to receive the reaction mixture which is consumed as the reaction proceeds. However, the more the process requires relatively large amounts of carbon (e.g.

0.1 weight unit per weight unit processed uranium and 0.5 Weight unit per weight unit titanium), so require most of the operations for commercial purposes that the carbon is added separately.

Any form of carbon can be used for this purpose. For example, oil black, gas black, acetylene black, lampel black charcoal, Coke and usually graphite particles. Carbon source materials are also suitable such as petroleum gases, wood products and the like. It can also be beneficial to reduce the carbon add in the form of a graphite rod or an open graphite cylinder, the only needs to be sunk into the reaction bath. Such an approach is permitted ee, an excess of carbon without accepting pollution providing the mixture with an excess of carbon particles; also results the addition of In this way, carbon has a faster reaction rate.

To find out the various solutions in molten metal, according to of the invention can be practically used to hold together is called material preferably graphite used, especially when working with a uranium oxide charge and as the metal solution in this case, e.g.

Tin, copper or iron can be used. But others can too inert refractory materials used for the manufacture of the containers or vessels as well as for the various pipes, fittings and other individual parts of the associated equipment. For example, with tin as a solvent, refractory materials can also be used such as boron nitride and beryllium nitride can be used. In the case of copper as a metal solution Tungsten can be used for the container. Instead of an extremely fireproof container system as described here, it is also possible to use the melted Solute metal in a so-called.

"skull" to hold together, i.e. in a kind of shell, which by constant Cooling the outer parts of the molten metal system to below its solidification temperature while maintaining the internal parts of the system in the desired melting state is formed.

Heat is added to the unit as indicated above as needed, where induction, heating by a Electron beam or others Heat methods can be used so that both the reactor vessel and any stills are kept at the required temperatures, to drive off the solute metal. As the reaction progresses, it is ensured that that the specific CO pressure is kept below the equilibrium value, wherein the reaction continues and is brought to completion; this can be done by pulling a vacuum over the reactive mixture or by painting the space with an inert gas (or in the case of a nitride-forming reaction, with nitrogen or nitrogen + an inert gas).

Creating a vacuum across the system is also useful during the initial recovery stage if the tin or other dissolving metal is distilled off from the dissolved metal. The process can be done in batches or be carried out continuously. In the latter case, parts of the liquid reactive mixture is constantly removed from the system to the solvent and separating the dissolved metal components, the separated being used as a solvent Rolling metal is returned to the reactor when new metal oxide additions and carbon advances occur. In the case of a shuffle-free way of working the separation of the dissolved metal from the metal used as solvent either be effected in the reactor vessel itself or in another vessel.

Depending on the metal used as the solvent, the method is useful in accordance with the invention with a wide variety of oxides being more reactive Metals, the term "reactive" being used herein to mean those To designate metals that are specifically specified here, which lightly with carbon react to form the corresponding metal carbides, i.e. tin as a solvent is used, the invention can be implemented with good success in the manner that the oxides of the reactive metals niobium, Tantalum, plutonium, uranium, zirconium, hafnium, titanium and dor, magnesium, chromium, manganese, Vanadium, silicon, aluminum and iron can be used. When using an oxide from a given metal contaminated by other metal compounds, it is proposed to cause its reaction through the nitride-forming stage, since this embodiment of the method in most cases offers the opportunity a nitride - to separate the desired metal that is relatively free of other components is. So look for foreign substances that are present, but in a small @ Percentage in which the molten solvent is in solution remain, while the desired metal, which is present in much larger amounts, is a nitride that either sinks, such as uranium nitride or hafnium nitride, or on the Melt floats. Typical floating nitrides are those of titanium, aluminum and zirconium.

These differences in specific gravity between the nitrides allow another convenient method of separating the different metals, which may be present in the oxide charge.

Assume, however, that a metal oxide feed becomes more acceptable If purity is used, the process is preferably carried out without the use of a nitride-forming agent Stage, while oxides of the metals boron, titanium, chromium, manganese, vanadium, Silicon, aluminum or iron are used, all the more so than the deoxidizing ones Reactions (reaction (l) see above) with oxides of these metals are acceptable in accordance with the standards Have free energy values such that the reaction takes place at reasonable temperatures and pushing forward.

On the other hand, in the case of deoxidizing reactions, the process is with large positive free energies such as those containing the oxides of niobium, tantalum, Plutonium, uranium, zirconium and hafnium concern, preferably in presence feasible by nitrogen gas to produce the corresponding metal nitrides either as End product or to form as an intermediate. In this reaction (this is reaction (2), see above) the free ones to be considered Energies when using these metal oxides lead to small positive values or too negative Values changed. The nitride formation shifts the free energy of the reaction, see above that at reasonable temperatures and pressures, the reaction would be propelled again can be.

A requirement for the melt of the dissolving metal to be used here or an alloy thereof is that the metal has good solubility for the Uranium, titanium or other metal must have added in the form of an oxide will. It is also important, even more so than the use of a nitrogen atmosphere part of the invention in one embodiment of this forms that the solution metal itself does not easily form nitrides. In general, for very stable oxides it is desirable that the solution metal has a tendency to form stable intermetallic compounds at lower temperatures with the metal derived from the oxide present in solution to form with it the activity of this metal when in solution to a relatively low level Level can be lowered so that it does not react with the carbon to form carbides under the conditions given when carrying out the process.

This ability of a given solution metal to such intermetallic compounds Forming at low temperatures appears to be consistent with ability of the metal in solution, the chemical activity of the metal itself dissolved in it reduce at higher temperatures. Among the metals that have these properties to a greater or lesser extent include lead, zinc, bismuth, cadmium, Silver, copper, iron and tin and various alloys of these metals, in particular Tin-lead, tin-bismuth, tin-lead-bismuth, tin-cadmium and cadmium-lead. However Among these metals, tin (or a tin alloy3 is often preferable, since For such systems it best all of the solution metal requirements as previously discussed were met.

Furthermore, tin has the added advantage of being extremely broad The range from about 505 to 29600 K remains molten and thus the application of the Process even at relatively high temperatures without damaging the system needs to be pressurized or vented to prevent steam loss. In the same way, a system can operate at relatively high temperatures or at low Pressure can run without loss of tin, although the distillation of this metal will run out the system is easily possible at correspondingly high temperatures, especially if it is carried out in a vacuum. So the distillation temperature becomes, while the Tin distilled at about 2960 ° K and at atmospheric pressure, to about 1700 ° K decreased at 0.115 mm Hg.

Due to the favorable properties of tin - with which mainly practically pure tin together with its alloys containing at least 50% by weight of this Contain metal - as well as for the sake of simplicity and clarity of expression, is described below - the invention in its preferred embodiments, using essentially pure tin as the molten metal in solution.

When carrying out the reaction so that the carbon with the metal oxide reacts to form the pure dissolved metal, or the metal nitride, becomes the Reaction forced in the desired direction in which the CO partial pressure is below the Equilibrium levels prevailing for the system entered. in the in general, this CO pressure threshold becomes higher (i.e. less restrictive) at higher Solution temperatures and vice versa. Furthermore, the greater the chemical activity of the metal present in solution, the lower the CO partial pressure threshold; with In other words, the partial pressure of carbon monoxide across a system must be increasing low values are maintained when the chemical activity of the metal gets bigger. The procedure for determining this CO threshold for any given The system is known to the knowledgeable specialist, and the special features are known the determination is made below in conjunction with the discussion of metal activities specified ts is essential that the chemical activity of the off derived from the oxide and in solution in the molten tin or other dissolving metal metal present at any given temperature and concentration has a value which is below the chemical activity at which the corresponding metal seeks to react with the carbon present in the system to form a carbide to build. This latter value is a function of the temperature, it increases with this too, although it is not a function of the concentration of the metal in solution is. On the other hand, the activity of the dissolved metal decreases below that of the carbide-forming Activity threshold must be maintained, both with the temperature and with the Concentration too. The impossibility of keeping this metal activity low while is still being operated in an effective and economical manner (i.e. at relatively high concentrations and without carbide formation), is the main obstacle the previously available carbothermal reduction process -been. Accordingly the present invention is based primarily on the recognition that, by implementation the reaction in molten tin or another suitable molten solvent, that metal going into solution in the bath has a low activity, that below the carbide-forming activity threshold, even at relatively high concentrations of the metal in the metal molten bath and at high temperatures. At the same time, the advantage is achieved that the carbon monoxide partial pressure threshold can be raised to acceptable working levels, even if this solution method is used to reduce metal activity.

Determination of metal activity. A method of determining the chemical activity of the liquid metal produced during the process according to the Invention is formed and which is present in solution in the molten bath, leaves explain themselves based on the metal oxide-forming reaction. The so determined activity is characteristic of a given system, regardless of whether it is a certain amount the free metal formed during the carbothermal reduction step, is simultaneously converted into insoluble metal nitride, all the more so than a certain one Equilibrium percentage of metal formed normally in the molten tin bath remains in solution. The equation for the metal oxide forming reaction is at a representative embodiment: M (l) + O2 (g) = M02 (s) (4) where (1), (g) and (s) mean components which are in the liquid1 gaseous and solid states, respectively available. The equilibrium constant tK) for reaction (4) finds its own Expression in the following equation: ZG ° = -RTlnK @ or otherwise expressed as log K4 = - # G ° 4.575T where AG ° = net change in free energy in the standardized oxide-forming equation (oxygen at 1 atmosphere) for one determined temperature, R = the gas constant 1987, and T = the temperature in OK All have the logarithms used here with the exception of the natural logarithm ln the base 10.

Typical values of aGo of oxides are in the literature for each of the metals with which this invention is concerned and from these data the value of K4 can be determined using the above equation (see Thermodynamic Properties of 65 Elements -Their Oxides, Halides, Carbides and Nitrids. Wicks and Block, U.S. Bureau of Mines Bulletin No. 605 (1963).

U.S. Government Printing Office, # 1.00). It's experimental data available, by which one determines the 02-Gasteildruck (in at), which over a system exists that a solution of the metal (M) of a given concentration in one It contains metal as a solvent when the oxide is just beginning to turn out at temperature to be formed by the # G ° value obtained from the literature.

This particular O2 partial pressure is then compared to that which exists above the pure liquid metal (M (l)) at the same temperature as when it is being converted to the oxide form. This latter O2 partial pressure for pure metal is determined by the equation where po is the equilibrium partial pressure of oxygen in 2 at above M (l). Since the chemical activity of the MO2 compound (aMO2) and that of the pure liquid metal (a (1)) both have the value 1, K4 = and since the value of K4 has already been determined above, 2 is therefore the value of PO made evident. The 2 chemical activity of the dissolved metal (aM) at the particular concentration and temperature used can then be determined by solving the following equation this gives the desired level of chemical activity.

Taking into account the requirement that the chemical activity of M must be smaller than those which lead to metal carbide formation (MC) would if MO2 and C are used together in the molten solvent (with or without a nitrogen atmosphere) the activity level of M must be for the following equation can be determined: M + C (s) = MC (s) (5) To for this purpose the equation G ° = -RTlnK5 is used.

The carbide value of # G ° at a particular temperature is available in the aforementioned reference for each of the metals with which the invention is concerned, and thus the value of K5 can be determined. One then moves on to the expression The more that the activity of the metal carbide (aMC) and that of carbon (ac) both have the value 1, K5 = 1 / aM and the desired value of am at a given temperature dn of the carbide-forming reaction is easily obtained. Should it now be found that the previously determined activity level of M falls above or too close to the carbide-forming activity threshold, as determined above, one can go back to correspondingly lower concentrations of M in order to reduce its activity. down to a satisfactory working level. In addition, changing reaction temperatures are also beneficial in changing the level of activity of M to the desired extent, although decreasing the temperature also decreases the level of carbide-forming activity to some extent.

The carbothermal reduction reaction according to the invention requires it does not matter whether it is carried out in the presence of a nitrogen atmosphere or not that the CO partial pressure is as varied as it is by one or the other of the following Reactions generated is M02 (s) + 2C (s) = M + 2CO (g) or (1) M02 (s) + 2C (s) + ½ N2 (g) = MN (s) + 2CO (g) (2) below the theoretical equilibrium level as it is for a given system is calculated is held.

When calculating this partial pressure and knowing the value the equilibrium constants (K1 or K2) and the chemical activity of the dissolved Metal can be the CO partial pressure from one or other of the following relationships can be calculated.

Ki M aM. P002 or where PCO means the partial pressure of CO in the atmosphere in one or the other system.

In order to specify representative CO partial pressures and aM values that were used for Uses suitable in practicing the invention are provided below both for uranium oxide and titanium oxide systems.

System UO2 in Molten Sn Table I below contains Information on the values for the equilibrium constants and the like in the reaction U02 (s) + 2C (s) = U (l) + 2CO (g) (6) Table I also contains values for # G ° and the activity of uranium (au) in the carbide-forming reactions U (l) + C (s) = UC (s) (7a) U (1) + 2C (s) = UC (s) (7b) 2 These values, referred to above becomes 18000 K and 19000 K for the three representative working temperatures and 20000 K.

Table 1 # G ° from U (1) K6 = aU # PCO2 # G ° from UC aU for UC temp. Response (6) Reaction (7) Formation 1800 ° K + 55,000 cal 2.00 x 10-7 -18 100 cal 6.3 x 10-3 1900 + 47 500 3.40 x 10-6 -17 900 9.1 x 10-3 2000 + 40 000 4.27 x 10-5 -17 700 1.2 x 10-2 The following table II gives values for the activity (au) of the in the Dissolution of dissolved uranium and the partial equilibrium pressure by CO over the system for these same temperatures at uranium concentrations of 1, 2, 5, 9 and 18% by weight in the molten tin solution.

Table II Temperature Uranium concentration in tin (% by weight) 1% 2% 5% 18000K PcO o, o45 at 0.037 at 0.022 at aU 1.0 x 10-4 1.96 x 10-4 4.14 x 10-4 1900 ° pCO 0.137 at 0.105 at 0.072 at aU 1.82 x 10-4 3.1 x 10-4 6.55 x 10-4 2000 ° pCO 0.394 at 0.302 at 0.208 at aU 2.75 x 10-4 4.68 x 10-4 9.9 x 10-4 9% 18% 18000 pCO 0.019 at o, o18 at aU 5.4 x 10-4 5.87 x 10-4 1900 ° pCO 0.063 at 0.060 at aU 8.55 x 10-4 9.3 x 10-4 2000 ° pCO 0.182 at 0.175 at aU 12.9 x 10-4 14, o x 10-4 the end the data in Table II shows that the activity of the dissolved uranium in each case is less than that given in Table I for the carbide-forming Activity is indicated. In any case, this applies to temperatures of 18000K Tin solutions with a concentration of uranium that is at least about 25%.

In addition, the limiting CO partial pressures given above are never less than about o, ol5 at, and pressures below that level can be easily over the system can be maintained by conventional methods.

System TiO2 in molten Sn In the following Tables III and IV are data for the titanium systems similar to those in the tables I and II for uranium were given, with the corresponding thermal reduction and the carbide-forming reactions are the following: TiO2 (s) + 2C (s) = Ti (l) + 2CO (g) and (8) Ti (l) + C (s) = TiC (s) (9) Therein Ti (l) likes melted or supercooled Mean titanium liquid. It should be noted that the data given below approximate calculation values for model systems are used to demonstrate the principle should serve. However, foreign substances are used in the feed present, the actual values of the various numbers may vary to some extent Change extent.

Table III # G ° of Ti (1) K8 - aTi # pCO2 # G ° of TiC aTi for TiC Temp. Reaction (8) Reaction (9) Formation 1600 ° K +36 050 cal 1.1 x 10-5 -39 850 cal 0.4 x 10-5 18000 +20 o50 3.7 x 10-3 -39 400 2, o x 10-5 Table IV Temperature titanium concentration in tin (wt.%) 40% 50% 16000K pCO 26.8 at aTi 1.3 x 10-8 3.2 x 10-5 1800 ° pCO 721-aTi 7.0 x 10-9 2.3 x 10-5 From the above data it can be seen that with this System the CO partial pressures are extremely high at equilibrium and that in some Cases from the point of view of the CO pressure only required a pressure reducing valve to vent the system to atmospheric pressure. However, these show Data that the system with use of titanium concentrations in tin below of about 50 wt. Must be treated if the chemical activity of dissolved titanium should be kept below that of titanium forms a carbide. In particular at 16000K the limit value for the concentration is of the titanium in solution about 47% by weight, while the limit value is 18000K is about 49 wt.%.

The nitride-forming reaction stage becomes the carbothermal reaction Reduction according to the invention at a suitable temperature and partial pressure carried out by nitrogen, so that the reaction taking place by the following Express the equation: M02 (s) + 2C (s) + 1 / 2N2 (g) = MN (s) + 2CO (g) (2) This reaction only reaches completion when the carbon monoxide partial pressures are below the Equilibrium values are maintained. This can be determined from the relationship K2 = pCO2 / pN21 / 2 Here, K2 is calculated from the known # G ° values as they are are available in the literature for this reaction. If it is assumed that the partial nitrogen pressure, which are in a range of about 0.1 to 1.0 at, what is usually appropriate around the desired nitride-forming Stage, so the carbon monoxide equilibrium partial pressure can something be greater or lesser than the one who performed the response in the absence of nitrogen. While both reactions are temperature sensitive - the CO partial pressure increases with the temperature - the CO partial pressure becomes during the nitride reaction not by the concentration of the metal in the molten metal as a solvent impaired.

For example, in a system in which uranium oxide a molten tin solution is added at the specified temperatures and the Partial pressure of nitrogen is kept constant at 1 at, the following relationships : U02 (s) + 2 C (s) + 1 / 2N2 (g) = UN (s) + 2C0 (g) (2a) This reaction only leads to Completion when the carbon monoxide partial pressures are below those in the below last column 'of Table V must be maintained.

Table V # G ° from UN K2 = pCO2 / pN21 / 2 pCO (N2 = 1 at) Temp. Reaction (2a) 2 18000K +23 400 cal 1.45 x 10-3 3 3.8 x 10-2 1900 +17 200 1, o5 x 10-2 1, o x 10-2 2000 +11 800 5.14 x 10-2 2.2 x 10-1 The approximately corresponding thermodynamic data for the equation TiO2 (s) + 2C (s) + 1 / 2N2 (g) = TiN (s) + 2CO (g) (2b) are given in Table VI below.

Table VI # G ° of TiN K2 = pCO2 / PN21 / 2 pCO (N2 = 1 at) Temp. Reaction (2b) 2 15000K -2 850 cal 2.6 1.6 1600 --8 900 16.2 4, o 1700 -15 ooo 85 9.2 1800 -20 ooo g16 17.8 The exact conditions under which a given metal is a solid Nitride product forms in a given molten solute metal such as tin not available in the literature, but when working with relatively moderate nitrogen partial pressures (from, for example, 0.1 to 1.0 at) in a temperature range of about 1000 ° to 20000K, in which the respective metal nitride is stable, and with knowledge of the CO equilibrium partial pressure value for a given metal at the temperature used, one skilled in the art is at that Location, through routine trial and error, the optimal nitride-forming properties of any given system as well as the properties under which the nitride is with accompanying Solution of the metal decomposes when the partial pressures of nitrogen be progressively decreased across the system. For example decomposed at a nitrogen pressure of 1 atm and using tin as the dissolving metal the highly stable uranium nitride compound does not, unless at temperatures above from 20000K.

On the other hand, other nitrides decompose at lower temperatures e.g. aluminum nitride at about 17,000 The data given below, which are refer to the conditions under which uranium nitride components in a molten Tin solution can be formed or decomposed are typical of those who can be easily developed.

Uranium when it is in molten tin (or some other suitable Metal) under an atmosphere with a partial pressure of nitrogen in the order of magnitude of at least about 0.02 at can be partially or practically dissolved in the system exist entirely either as UN or as U2N5, depending on the temperature and uranium concentration in the system. In general, the solid U2N3 product is that which is at temperatures from about 575 to 1753 ° K is formed. The UN product is the one that works at temperatures above about 17580K is formed and it is capable of operating in the system at temperatures of 2275 ° K or higher temperatures under correspondingly high pressures and solution concentrations to remain. The equations for these two reactions are the following: 2U +) / 2N2 (g) = U2N3 (s) (lo) U- + 1 / 2N2 (g) = UN (s) (11) The two systems do not speak in the same way to changes the nitrogen pressure. So the UN system is reversible and it quickly speaks up Pressure changes. The U2N2 system, on the other hand, is irreversible - except likely when it is sustained for long periods of time; particular is while an increase in the partial pressure of nitrogen produces a greater precipitate of the dissolved Uranium as U2N3 results in a lowering of the nitrogen partial pressure without noticeable Effect until the pressures are then reduced to a very low level Nitrogen gas is evolved and the product is converted into the form UN. If a U2N3 system is brought above a level of about 17580K for some reason it is nitrogen gas and the existing U2N3 is converted to the UN product. on the other hand becomes the existing UN material when a UN system cools below 17580K is converted to the U2N3 product. Should not have enough nitrogen be to effect this transition, part of the UN is converted to tJ2N3, while the rest dissociates and uranium into the molten tin solution supplies. So it is possible to have either a solid UN product or a solid U2N3 product to form, which is then separated from the molten tin metal in any way can be. However, if the solid product is to be decomposed, about the uranium To bring back into the solution, then preferably conditions are chosen that bring the uranium into the UN form.

As stated above are the equilibrium properties of the U-Sn-UN system (Equation ii) such that there is a slight shift in either the UN or Allow the U-direction when nitrogen pressures temperatures or solution concentration to be changed. So brings for any given concentration of solute Uranium (U) in the molten tin or other molten metal in solution Raising the nitrogen pressure or lowering the temperature increases the tendency to increase the relative amount of the dejected UN that exists.

Likewise, the greater the concentration of U in the molten one Metal, the greater the mass of UN. The importance of the factors of pressure and the U concentration in the equilibrium state is exemplified by the in The table below gives data for an 8yse1n that was held at 18230K wind.

Table VII Nitrogen pressure Kg uranium, deposited as UN at 1823 ° K (in at) from 100 kg of melt solution with the initial percentage given in% by weight of uranium in tin 18% 9% 5% 2% 1, o 15.3 8.7 1.1 0.9 14.8 8.6 0.4 o. 6 14.4 8.4 0.5 14.1 8.1 o, 4 15.3 7.7 o, 3 10.4 6.5 2.2 o, 2 6.4 3.8 o, o7 o, 1 l, o o, o4 In similar Way is the influence of temperature on the equilibrium of a typical U-Sn-UN system at various nitrogen pressure levels by those given in Table VIII Data illustrated.

Table VIII Temperature Kg uranium, precipitated as UN from 100 kg (° K) of a molten solution of 9% uranium in Sn at the specified nitrogen pressure values o, o9 at 0.122 at o.16 at 0.202 at o.25 at 1825 0.58 2.05 3.46 4.80 6.21 1840 1.05 2.48 3.92 5.46 1865 1.38 2.82 4.21 1883 1.53 2.98 The foregoing Tables provide explanatory data for specific systems. Against this is the following Equation (which in contrast to Equation 11, in which the uranium is in a tin solution is present, one in which the system is only pure liquid uranium and Contains nitrogen) can be used to determine any of the variables that make up the Affect equilibrium, i.e.

the point at which the UN is just knocking out a given system begins while the other variables are held constant. Here is the equation to be considered: UN (s) = U (l) + 1 / 2N2 (g) (12) Log K12 = -15 60o / T + 4.7 where the Equilibrium constant K12 = aU.PN 1/2.

2 E.g. the equilibrium pressure of nitrogen that is required is to just the UN formation in a 12% uranium solution in tin at 1873 ° K can be determined as follows: At 1873 ° K, the following applies: log K12 = -3.63; therefore K12 = 2.34 x at 12% U level (mole fraction = o, o64) applies aU = o, ooo8.

Finally, 2.34 x 10-4 = = 0, ooo8 PN 1/2; therefore 2 pN2 = 0.085.

Generally speaking, after determining the partial pressure of nitrogen that must be maintained over a given solution of a metal in molten tin or other solute metal to form the desired nitride product, one can substitute the partial pressure value so obtained in the following equation: which relates to reaction (2) to obtain the corresponding equilibrium partial pressure of carbon monoxide: In order to bring reaction (2) to completion, the CO partial pressures must be kept below the equilibrium values determined in this way.

When carrying out the method according to the invention, a large Variety of refractory building materials used to make the apparatus manufacture or line them. The nature of the materials chosen depends somewhat on the molten metal used as a solvent that will be used target. While graphite can be used with virtually any such metal, know boron nitride and beryllium nitride in connection with a tin bath and tungsten used in conjunction with copper or iron. The "skull" technique mentioned above can also be used, with the "skull" shell the molten metal can be kept in a cooling socket. A preferred one Material, especially when working with uranium oxides, is graphite or a refractory Material such as graphite, shielded with aluminum. This results in a good one Total strength, and the graphite does not undergo chemical reactions and withstands easily temperatures in the range of 600 to 2300 ° K or even more, in particular in the absence of oxygen. It is also possible to use a larger part of the apparatus Manufacture from graphite, a method that is particularly well suited to use when carried out on a smaller scale, the reaction vessel to some extent is consumed to complete the reducing reaction with the required amount of carbon to complete.

In the following the invention is exemplified with reference to the drawings explained. 1 shows a general schematic representation of an apparatus for carrying out a carbothermal reduction of metal oxides without application an intermediate nitride-forming stage, wherein the reduced metal is in a molten tin bath is dissolved, from which the tin can then be distilled off; Fig. 2 is also a schematic representation of a system in which the carbothermal reduction step is accompanied by the simultaneous formation of a solid metal nitride product, that can either be recovered as such or in the presence of a molten one Metal as a solvent can be converted into nitrogen gas and the pure metal that is present in solution in the molten metal and for subsequent recovery is available.

In Fig. 1, 10 is a reactor vessel which has a molten bath 11 made of tin contains, in which there is a stirring device 12 and that on the desired Temperature is maintained while the tin is heated by an electric induction heater 15 is put into circulation. The metal oxide feed component is added to the reactor vessel fed through a line 16, while the carbon that is required to react with the oxygen of the oxide and form carbon monoxide, through line 17 is added. To reduce the partial pressure of carbon monoxide that in the reducing reaction is formed below the equilibrium level too hold, an inert gas like argon or helium gets into the body of the melted Tin is injected via line 18 as it serves to remove the carbon monoxide from the System to be discharged via line 19 and at the same time the molten metal solution to set in motion. According to a modified method (not shown) can be the desired CO partial pressure by using a vacuum created and maintained. The outlet gas line 19 runs to a Gas cleaning zone 20, which is able to remove the carbon monoxide and To leave the inert gas, which is then via line 21 together with a required Make-up gas as supplied through line 18 is returned to the unit will. The elimination of the carbon monoxide in zone 20 can be known per se Manner, for example by oxidizing the gas to carbon dioxide, which is then absorbed or condensed to escape via line 22. 23 is a tin processing line.

Who is the amount of reduced metal that is in solution in the Tin bath 11 is present up to the desired level, e.g. 20-30 % By weight, the further addition of metal oxide and carbon is interrupted and the solution present in the vessel lo then becomes one through the line 25 Induction heating device 26 and passed from there to a distillation column 30, from where the tin is distilled off.

The column is provided with a heater 31 heated by induction and is operated at reduced pressures. The heat can also be through use an electron beam heating device. The tin swamps will from the top of the column 30 through a Line 32 removed and fed to a heat exchanger 33, where the vapors condensed in the form of a liquid will. The liquid tin then runs to a sump 34, which is connected to a vacuum line 35 is provided so that the desired pressure in the column 30 of the line 32 and the condenser 33 and above the condensed liquid in the container 34 can be produced. The molten tin then flows out of the latter Container via line 37 and pressure reducing valve 38 into the vessel lo. After this the tin has been distilled off, the desired residual metal can be removed from the column 3o are withdrawn via a line 39.

In Fig. 2, a reaction vessel 40 is shown which has a molten Body 41 made of tin, which is kept in motion by a stirring device 42 can be.

The metal oxide charge is fed to the vessel 4o via the output 43, while carbon is supplied via line 44. In some cases a natural gas or organic waste instead of a source of pure carbon to be used. The molten tin is, if necessary, over the pipe 45 is fed to a storage container 46, from which the tin is fed into the Vessel 40 can be transferred. The tin held in the reservoir 46 can be kept at the desired temperature as the molten metal passes through an induction heater 48 is diverted.

Likewise, the vessel 40 is provided with a circulating induction heater 49 provided.

When the metal oxide and carbon are added to the vessel 4o, nitrogen gas is introduced into the tin melt via line 55. If so desired the nitrogen stream entering the melt can with the addition of an inert Gas such as argon, helium or the like. Be added if this is the nitrogen line through lines 56 and 57 is supplied. Non-reacting nitrogen gases will be besides the carbon monoxide formed during the reducing reaction and any applied inert gases withdrawn from the vessel 4o via line 60. At the reaction taking place in the vessel 4o reacts the metal oxide and forms a insoluble metal nitride product. If this nitride product is less dense than the tin melt, it floats on this as indicated at 65. On the other hand, it sinks a nitride product that is denser than the melt to its bottom as indicated at 66. Occasionally the system includes both a layer 65 and a layer 66, a case that occurs, for example, when the ore or other oxide feed Contains oxides of various metals, one of which floats on the tin, when it is in the form of one nitride while the other sinks into it. Also like that Metal oxide in turn contain contaminating metals with nitrogen react to some extent to form one nitride product that resides in another Layer occurs as the main metal nitride product.

Once the nitride formation stage is complete, it can solid nitride product or the solid nitride products from the tin melt 41 and the vessel 40 can be removed in any manner. Optionally, these Nitrides in situ in the vessel 4b in the presence of the molten tin 41 to nitrogen gas and converted to the essentially pure metal contained in the tin in Solution works.

If the latter conversion is to be done in situ, then the nitrogen flow and the other feed flows to vessel 4o are shut off and the component pressure above the melt 41 is reduced.

This can be done either by means of the line 60 Vacuum is drawn, with the evolved nitrogen gas being withdrawn or only by passing an inert gas through the melt and thence into the conduit 60, whereby the evolved nitrogen gas is discharged without affecting the entire system pressure needs to be reduced. Once the conversion is complete, it can the dissolution of the metal in the molten tin through the filter 70 via the line 71 containing an induction heater 26 'is withdrawn to a distillation column 50' be that with a Circulating induction heater 31 'is provided. The evaporated tin is discharged from the column 3o 'at the upper end through a line 32' withdrawn and fed to a capacitor 33 ', while the remaining metal product can be derived via the line 39 '. The molten tin that is in forms the condenser, then runs to a liquid collecting tank) 4 ', the is provided with a vacuum line 35t so that the desired reduced pressure can be made in the column 3o1 of the line 32 and the container 34 '. That Tin can be obtained from the latter container either via line 37 'which is a pressure reducing valve zi8 'to the vessel 80 or via lines 37' and 81 in the form of a kind of return flow flow back to the container 46.

If desired, the solid nitride product layer 65 from the To remove vessel 40 instead of converting it there in situ - as described above, this can be done by discharging this product through a line 83, this discharged material either into a fresh bath 85 of a tin melt withdrawn in the vessel 80 or via line 86 for recovery in nitride form will. In another way of working, the nitride layer 65 can go down into the for the layer 66 shown position are brought by the liquid tin from the Vessel 40 is withdrawn through the filter 70, this tin then either in the container 46 via the lines 71 and 86 'or via the line 71 is passed into the column 30 for distillation. In the latter way of working can be those present in the tin when introduced with the metal oxide charge Impurities are withdrawn from the column 3o 'via the line 39'.

The information given above about available working methods at Stripping the solid nitride product 6s from the vessel 40 is also suitable for use on a large scale if the nitride layer to be extracted is the relatively denser one 66 solid shown here is this solid or this solid pure Substance separated by mechanical means (not shown) and either taken over into the vessel 80 via the line 88 or from the system via the line 9o can be derived.

This separation of the product 66 can thereby be facilitated if desired be that first the molten tin from the vessel 80 through the filter 70 and line 71 is withdrawn so that only the solid or pure product is in the vessel remains for transport either to the vessel 80 or out of the system.

If one or the other of the solid nitride products 65 or 66 in the pure molten bath of tin 85 is set up in the vessel Ao, the transformation karm of the nitride in nitrogen gas and metal, which is then dissolved in the tin is, in the vessel either by causing a vacuum over the molten tin drawn through line 95 or an inert gas from line 56 through the melt is sent, then this gas through line 95 together with the developed Nitrogen gases is given off. The resulting metal solution in the melted Tin can be discharged through lines 96 and 71 to distillation column 30 ', to in the generally similar manner - as described above - in. Connection with the same current as is formed in situ in the vessel 4o and then to be delivered through line 71. However, is located at of the present procedure in which a substantially pure nitride solid is introduced from the vessel 40 into the pure tin melt 85, the stream of metal product, which is taken from the column 3o 'via the line 39, in a high state Purity; an even higher purification can, if necessary, be achieved by that a second metal nitride-forming stage is used, which is so formed relatively even purer solid nitride then recovered and in another Another bath of pure molten tin is transformed into the form of a metal.

When discussing the separation of the pure after the present Process from a bath of molten Tin formed metal, in which the metal thus formed is dissolved, there is the example given above Method is to distill the tin from the solution, preferably under high Vacuum. On the other hand, certain metals can also be used as solution components themselves serve, which distilled off from the molten metal serving as solvent will.

Other phase separation processes can also be used. A good liquid-liquid separation method such as the one used in conjunction can be applied with a solution of uranium in molten tin. The addition of magnesium metal to this solution results in the formation of a molten tin-magnesium phase, which are then physically separated from the molten uranium phase can.

Another phase separation process that is based on a system in which Titanium is the reactive metal because of its stable hydride-forming properties of titanium is applicable, is that the mixture; according to the above reducing acting reaction to a temperature below the-Jenigen.gekühlt is, at which Till decomposes (e.g.

6000K) and that a hydrogen-containing atmosphere above the melt is created at a hydrogen pressure sufficient to keep the metal formed to cause it to combine with the hydrogen. That educated till that on iem solvent metal floats, can be skimmed off and heated by heating of Hydrids are converted to titanium metal powder in a vacuum at temperatures above about 14000K will.

Yet another phase separation process as shown in a final diagram is based on the solubility properties inherent in the system. E.g. the temperature in a system in which aluminum is the reactive Metal and tin serve as the dissolving melt, are lowered to precipitate the aluminum, while the tin remains in solution. A suitable temperature range for this is in the order of magnitude of 510 - 600 ° K.

It should be noted that the chemical activity of aluminum in Tin is low enough to avoid significant carbide formation without Intermetallic compounds are generated. This is in marked contrast to that previously discussed general case of stable oxide systems in which intermetallic compounds be formed in order to avoid carbide formation.

In the interests of simplicity and clarity of presentation, various common details of the apparatus such as pumps, instruments, mechanical Accessories and various valves and the like. From the essentially symbolic in The flow diagrams shown in the drawings are omitted.

The invention allows numerous modifications to the method the processing of special oxide charge batches and ores with varying degrees of relieving force to fit. It is also possible to use not only oxides, but also to use metals that are contaminated with oxides or with dissolved oxygen and which also contain various metallic impurities or foreign matter can. For example, titanium waste, which, in addition to titanium, can also contain titanium oxide and steel particles contained, as a charge for the molten tin bath in addition to carbon in a sufficient amount Amount used that it oxidizes with the oxygen present in the waste, to form carbon monoxide. When handling feed batches this Type in which the added metal contains small amounts of other metals the processing preferably via a nitride-forming stage, the so from the The solid nitride product formed in the main metal is then separated and converted into pure metal a pure tin bath. The one present in the original reactor bath Tin can be put into circulation again until the foreign substances it contains have built up to a level where the tin will distill off therefrom seems appropriate.

The nitride-forming metal separation process described in the previous paragraph and Also discussed earlier, one peculiarity of the invention is broader Applicability. E.g.

this method can be used »even if the compositions of the loading contain a majority of reactive metaste, or if the Charging one or more reationsf "ihi £ c 'metals in addition to other metals such as e.g. gold, Contains silver or platinum; which are not easily stable Form carbides or nitrides and therefore cannot be classified here as "reactive" metals are. When used herein in the description and in the claims of "metal" as a description is spoken, so are both free metals and such meant in which the metal occurs in the form of a compound or alloy.

The new nitride-forming metal separation process is advantageously useful both in connection with .feed compositions in which the in a nitride reactive metal to be converted is present in oxide form (in which case Carbon is also present as a reactant), as well as in connection with such Feeds that are substantially or even completely free of reactive metal oxide components in which case little, if any, carbon is added to the system needs to become. A typical example of a separation method of the latter type is one in which the feed composition consists of silver or other precious metal, contaminated with small amounts of actinic or lanthanic metals. Here the composition can either be, for example, molten tin as the solvent or if the composition is particularly rich in silver, the silver component may of it itself as a solution metal melt of the system, alone or mixed with Zi-nn, -serve. In any case, there will be a nitrogenous atmosphere manufactured over the melt to separate the actinic or lanthanic To induce components when they are converted into insoluble nitrides The noble metal can separate these nitrides in one of actinic-lanthanic Impurities free form can be obtained. That or the separated nitride product can either be viewed as waste and disposed of or a fresh one made from tin or any other molten metal existing solvent and can then be converted into nitrogen and free metal present in solution when the pressure above the system is released.

The invention also provides a method according to which the charge material of the melt formed by a metal in the form of a carbide instead of adding an oxide. Carbides of this type can easily be outside of the system can be produced by one or more known methods, such as for example by pre-reduction of a reactive metal oxide (e.g.

Al2O3) with carbon. With this working method and in a system at which the chemical activity of the added metal is at a level below that is diminished in which the metal can exist as a carbide, it was found that the metal carbide, the molten tin or other as a solvent The metal used is added easily decomposed.

The carbon component of the compound or mixture is thereby released in particulate form, while the metal component of the added carbide goes into solution in the molten solvent.

Limit the possibilities for using and carrying out the invention does not apply to the details described and illustrated here. Further modifications Rather, these result for the person skilled in the art on the basis of the disclosure in the documents just like that.

Claims (19)

P a t e n t a n s p r u c h e 1. Carbothermal process for reducing an oxide of a reactive one Metal, characterized in that this oxide in a stoichiometric amount of carbon at temperatures above about 6000 Kelvin in one made from molten Metal existing solvent is made to react, which reduces the activity of the metal formed during the reducing reaction to a level below of that which would be required for the metal to be among those in the system prevailing reaction conditions to form a carbide, and that the partial pressure of the carbon monoxide S present in the system is maintained at a level below of the theoretical equilibrium value for the system is during the reduction of the metal oxide takes place. 2. The method according to claim 1, characterized in that the from molten metal solvent consisting of one of tin, copper and iron existing group is selected 3. The method according to claim 2, characterized in that that tin is used as a solvent in the form of a molten metal. 4. The method-according to claim 1, characterized in that the metal, that is formed during the reducing reaction and that of molten Metal existing solvent is present by separation from the solvent regained will. 5. The method according to claim 1, characterized in that the reactive Metal from the molten solvent through one. Phase separation process removed will. 6. Verfanren according to claim 5, characterized in that the phase separation is carried out in such a way that over the existing molten metal Solvent an atmosphere containing nitrogen with a partial pressure of nitrogen sufficient to maintain this during the reducing reaction formed metal to combine with the nitrogen to cause a to form insoluble nitride product which is precipitated when the reduction of the metal oxide takes place. 7. The method according to claim 6, characterized in that the formed insoluble nitride is a uranium nitride and that consists of molten metal Solvent is tin. 8. The method according to claim 6, characterized in that the formed insoluble nitride is a titanium nitride and that that consists of molten metal Solvent is tin. 9. The method according to claim 6, characterized in that the during insoluble metal nitride formed during the reaction in the presence of a molten one Metal existing solvent is converted into nitrogen gas and metal, which is dissolved in the solvent metal, and that the conversion is effected by that the partial pressure of nitrogen in the system is below equilibrium levels is held when the nitride is progressively converted, and that that in the Molten metal dissolved metal through its solvent Separation from the solvent is recovered. lo. Method according to claim 9, characterized in that tin is used as Solvent metal is used that the insoluble metal nitride is UN and that the recovered metal is uranium. 11. The method according to claim 9, characterized in that the solution metal Tin, the insoluble metal nitride titanium nitride and the recovered metal titanium is. 12. The method according to claim 6, characterized in that the insoluble Physically separated nitride product from the remaining liquid parts of the system that the separated nitride is then transferred to a fresh, relatively pure bath of a made of a molten metal Solvent added becomes that the nitride added in this way is converted into nitrogen gas and metal that dissolves in the solvent made of metal and that the dissolved Metal by distilling off the solvent consisting of a metal the solution is recovered. 13. The method according to claim 6, characterized in that the phase separation after the formation of the metal in the reducing reaction carried out in the manner becomes that a hydrogen-containing atmosphere above that consisting of molten metal Solvent is provided which has a nitrogen partial pressure that is sufficient is to get the reduced metal to combine with the hydrogen at a temperature below that at which the reactive metal hydride decomposes, to form an insoluble hydride product that floats on the melt. 14. The method according to claim 13, characterized in that the formed insoluble hydride titanium hydride and the molten metal solvent Tin is. 15. The method according to claim 6, characterized in that the separation by causing the melt to cool until it is reactive Metal precipitates while the solvent remains in solution. 16. The method according to claim 15, characterized in that the reactive Metal aluminum and the solvent tin, which consists of molten metal is. 17. The method according to any one of the preceding claims, characterized in that that the separation of a reactive metal from a multiple metals in the form pure metal or in the form of a compound-containing feed alloy is made in such a way that a melt of a molten metal existing solvent is formed that will put the feed alloy together containing any carbon that may be required to deal with any to react metal oxide present in the alloy, the solvent being a that does not easily form nitrides that a nitrogenous atmosphere over the melt is maintained, which has a nitrogen partial pressure that is sufficient to cause the reactive metal present in the melt, combine with the nitrogen to form a titride product that is insoluble in the melt to form and that this nitride product is separated from the melt. 18. The method according to claim 17, characterized in that the melt contains molten silver mixed with foreign substances selected from the group the actinl see and lanthanic metals, which are insoluble Form nitrides, which are then separated from the equilibrium of the melt. 19. The method according to claim 17, characterized in that .that tin is used as the molten metal solvent and the feed alloy is at least contains a precious metal that has been mixed with foreign substances that belong to the group of actinic and lanthanic metals. which form insoluble nitrides which are then out of balance the melt are separated, and that the noble metal component from the remaining Molten tin is recovered by distilling off the tin. L e r s e i t e