NO115566B - - Google Patents

Description

Procedure for redistributing the constituents of a fusible material consisting of at least two constituents or substances.

The present invention relates to a method for redistributing the components of a fusible material consisting of at least two components or substances in a column, where alternating, separate, relatively hot and cold areas are made to move in a direction from one end of column to the other, where the fusible material is melted in the hot areas and solidifies in the cold areas. The method according to the invention utilizes the variations in the solubility of the dissolved substance (component) in solid and liquid phases that come into contact with each other in the material being treated, in order to separate dissolved substances (components), and the invention can be used in connection with systems of metals and their alloys, salts and salt solutions, both organic and inorganic, and other systems consisting of solid and soluble substances that can be subjected to a transformation from solid to liquid phase and vice versa. All the processes that the invention encompasses are continuous in the sense that the material can be supplied because the products are taken out continuously.

For practical reasons, the method according to the invention shall be described in connection with a binary system consisting of solvent and solute (component), it being assumed that the solute is the contaminant to be removed and that k, the distribution coefficient which is defined as the ratio between the concentration of solid solidified out of a molten zone, and the same in the liquid part of the zone, is constant and less than one. It is obvious that the dissolved substance can just as well be considered the product that one wants to extract, that the method works equally well with systems where k is greater than 1 and furthermore that the invention is not limited to use in connection with binary systems. It is noted that the constant k is identical to and here used instead of y (gamma) which is the usual symbol.

In the event that the system being treated is a semi-conducting material, e.g. silicon or germanium alloyed with small quantities of one or more dissolved substances which give the materials the so-called independent semi-conducting properties, these dissolved substances are known as "noticeable impurities" or "noticeable amounts of dissolved substance". As a result of the high degree of purity required in the production of semiconducting materials of the above-mentioned type, and as a result of the favorable separation constants that distinguish these systems, the method according to the invention is particularly well suited when it comes to achieving an increased purity of such materials for use in connection with rectifiers, transistors and other semi-conducting converter devices.

The method according to the invention utilizes the principle that will be referred to in the following as "rate zone-refining" as described in the article "Principles of Zone-Refining", Journal of Metals, volume 4, page 747, 1952, see also Norwegian patent 84 977. In in one of its forms, the batch-zone refining consists in a series of molten zones being moved in the longitudinal direction of an elongated, solid mold or charge containing impurities, where each molten zone in a system with k less than 1 results in. a transfer of contaminants towards the end of the charge in the zone's direction of travel. It has been shown that a very high degree of purity can be achieved in this way. With rate-zone refining, one can e.g. achieve a concentration of noticeable donor impurities in germanium of less than 2.10i 2 atoms per cm" of germanium which, in this example, was less than one donor atom per 10if germanium atoms (see W. G. Pfann and K. M. Olsen, "Physical Review", volume 89, page 322, 1953).

It is known in connection with separation methods that continuous processes offer certain advantages over batch processes. One of the purposes of this invention is to achieve zone refining with the same advantages on a continuous basis and, in addition, to increase the scope and application of the zone refining method.

A particular case of zone repair, where the use of a continuous method increases the area of application of the zone process, is the following. During crystallization work, it is common practice to remove a contaminant B from the crystalline substance A by dissolving both A and B in a solvent C. After a solid substance has been crystallized from the solutions consisting of A and B in C, is the separation between A and B greater than could be achieved by crystallization in a binary system of AB. If one wishes to use a common solvent C to improve the separation generally obtainable by s^ts-zone refinement of the binary system AB, it would generally be necessary to add a fresh amount of solvent C each time a heated area entered at one end of the charge, so that the process would become more complicated. In a subsequent process, however, the solvent can be continuously added as part of the added material, which solvent can be continuously removed together with the top and bottom products. As far as the refining process is concerned, this is thus as simple as in the case of a binary system, even if it concerns ternary systems or systems of a higher order.

As the method according to the invention in the area of crystallization is to be considered the counterpart of the continuous rectification column in the destination area, it is appropriate to refer to the terminology and methods used in connection with continuous fractional distillation processes. In this connection, however, it should be noted that although there is a certain similarity between the distillation methods and the zone refining methods, there are fundamental differences between the two. This will be clear from the following description.

In batch zone refining, molten zones move slowly through a solid charge or mold and carry impurities with them (in a system with k less than 1) and thereby cause a separation of solids and solvent. Generally speaking, no material will flow into or out of the apparatus containing the charge during the separation process.

The method according to the invention is particularly important in connection with the problem of separating solid from solvent, e.g. in rate-zone refining, but also brings with it an improvement with regard to feed, waste and product. The container can be regarded as a column with an enriching section and a draining section, where the feeding takes place between the sections and the product and the waste or first and second product are removed at each end. In such a process, the resulting or desired material moves away from the feed point in both sections. It will be shown below that the movement of the molten zones forms a backflow and that the backflow conditions can be defined and regulated during the process.

In the process according to the invention, a current is obtained both in forward and backward direction without the need to use any mechanical aids to achieve this, when one disregards the movement of the hot surfaces which must take place in a single one-channel column.

Although the continuous rectification column contains a rectification section, a stripping section and means of feeding between these two sections, and although product and waste are continuously withdrawn from each end of the column, it is obvious that the principles of continuous, fractional distillation are not without can further be transferred to the zone melting process. In processes where liquid is transferred in vapor form, as is the case with fractional distillation, the feed can take place at any point in the column as a result of the compressibility of the vapor phase and as a result of the liquid not adhering to the walls of the column. The same cannot be achieved as easily in transformation processes where the two states are liquid and solid state, where the liquid state is not compressible to any significant degree and the solid material, as a result of adhesion to the walls of the column, cannot be made to move without difficulty move away from the feed point to make room for the supply of new material. The method according to the invention also aims to avoid this difficulty.

The method according to the invention basically consists in providing "cavities", i.e. areas in which there are considerably smaller quantities of material under treatment, which cavities are made to travel through the apparatus. The cavities are provided by removing a quantity of substance in liquid form from one end of the column, where this quantity is smaller than the total quantity contained in a molten area. The cavities may constitute gas-filled or completely evacuated areas or may be filled with displaceable and essentially immiscible liquids. The cavities are generally assigned to the molten zones, so that one cavity is obtained for each molten zone, thereby making it possible to supply new material in the same amount as the corresponding cavities. If the cavities are completely evacuated, the added material will simply be drawn in and fill the entire cavity. If the cavities are in fact filled with liquid, the liquid will be displaced by the added material. With the method according to the invention, this displacement is achieved as a result of a difference in specific weight between the molten material in the feed and the liquid in the cavity.

For the sake of simplicity, in connection with the method according to the invention, it must be assumed that the molten material that is supplied has a greater specific weight (density) than the material found in the cavities, so that the liquid in the cavity will bubble through and consequently be displaced by the molten , added material. However, the method works just as well if the added material is made to rise and thereby displace a heavier, immiscible liquid that is in the cavity.

In the following description, the term "cavity" shall include areas where there are no significant quantities of the material being processed. Consequently, voids may include evacuated areas or areas that contain material in liquid or gaseous form at the equipment's working temperature. A general requirement is that the material in the cavities does not mix to a significant extent with the material being treated. If it is thus a liquid, this must not mix with the material being treated, even if it is in a liquid state. In cases where it is desirable for the liquid to have a certain effect on the system, e.g. impregnation, the liquid may, however, be partially miscible or soluble in the material being treated.

The invention shall be described in more detail with reference to the drawings, where fig. IA, IB, 1C, ID, 1E and 1F show in section the bottom part of a rectification or enrichment section of one type for carrying out the method according to the invention, fig. 2A, 2B, 2C, 2D, 2E and 2F show similar sections of another embodiment of the apparatus, fig. 3 a longitudinal section of a third embodiment, fig. 4A and 4B a column seen from the front acc. from the side, fig. 5 a modified version where the heating devices move along a circle, fig. 6A, 6B, 6C and 6D sections of various columns and heating devices, fig. 7 in perspective of a column zone with strip-shaped heating devices and fig. 8A, 8B, 8C, 8D, 8E, 8F, 8G and 8H cross sections of different types of cavity generators. Fig. 9A schematically shows a helical cavity refining device containing both enrichment and stripping zones, Fig. 9B same seen from one end, fig. 10 a modified embodiment of the apparatus according to fig. 9, fig. 11A, 11B and 11C a stripping section of a helical cavity refining device, Figs. 12 schematically shows a column process where the apparatus allows varying backflow conditions as the material moves through the column and fig. 13 schematically shows the apparatus for an enrichment and stripping column process similar to that in fig. 12 with devices to vary the return flow ratio.

Fig. IA-1F shows the lower or enrichment section of a column apparatus. As indicated by an arrow, the heating devices 1 move continuously upwards along the column 2 to provide molten zones 3 in the apparatus. The material in the areas 4 located between the heating devices 1 is solid.

Also in fig. 2A—2F as well as in fig. 3 and fig.

13, the heating devices are only indicated schematically as rails located directly outside the liquid zones. Physically speaking, the heating devices enclose the column, and are annular in the case of a column with a circular cross-section. The correspondence between the liquid zones in the column and the heating devices outside will not be exact as stated, as there will always be a certain displacement present. The stated heating devices can therefore be regarded as the effective heating elements viewed from the inside of the column. The regular planar colliding surfaces are also to be considered theoretically and without significance.

The smooth heating elements can consist of ordinary heating devices, such as resistance or high-frequency windings or burners. If it is an element that normally remains in liquid form, the intermediate parts will be occupied by cooling elements.

With the aid of the retracted outlet pipe 5, liquid material can be taken out until the upper edge of the lower heating element 1 is located above the upper edge of the pipe 5. As the heating elements 1 are longer than the pipe 5, the material in liquid form there will - after flowing out until the lower edge of the lowermost heating element lies directly outside the lower edge of column 2. The further travel causes the lower end of the outlet pipe 5 to cool and prevents a further flow until the next cycle begins. If the tube 5 had not been retracted, all the molten material contained within a heating element would flow out of the column 2 at the bottom of this until the lower edge of each heating element was level with the lower end of the column on otherwise become blocked by solid material. This would cause a number of voids to migrate through the rectification section, but as there would be no molten material present in the column, no interphase separation of solute would take place here. The result would be a flow of material from the intake opening as a result of the displacement of the material within each cavity. Fig. IA shows the apparatus just after it has been started and the lowermost heating element has moved sufficiently to produce a small quantity of liquid material in the annular space at the bottom around the tube 5, but has not yet reached so high that some of the molten material in the lower molten zone 3 may flow out of the outlet pipe. Fig. IB shows the situation a little later. The first heating element 1 has reached sufficiently high that its lower edge is located directly in front of the lower end of the column 2. As will be seen, part of the molten material has already found its way to this lower heating -element, cleanly out through tube 5 onwards brought a first cavity 6 whose height is equal to the height of the heating element 1 minus the length of the tube 5. As already mentioned, this cavity represents either an evacuated area or one that contains material that is liquid at the working temperature of the apparatus and has a lower specific gravity than the molten material contained in the molten zones 3.

By fig. 1C, the situation is that the hot element 1 has moved a little further upwards along the column, so that part of the material in the drawn-in tube 5 has solidified and thereby prevented further molten material in the lower molten zone 3 from flowing out.

In fig. ID is a new hot item about to reach the bottom end of the column. As will be seen, the movement of the molten area 3 and of the cavity 6 coincides with the movement of the heating element 1, so that both areas are within the relevant heating element's area. Consequently, there will always be colliding surfaces 7 between the lower ends of the molten zones 3 and the upper ends of the solid zones 4, which colliding surfaces are those at which the transfer from liquid to solid form takes place. As the hot elements 1 move upwards, molten material from the solid zones 4 will drip through the cavities 6 and into the molten zones 3. This material mixes with the material already present in the molten zones 3, which material is solidified at the impact surface 7, so that the requirements for zone refining are satisfied.

In fig. 1E is a new heating element

1 in the same position as shown in fig. IA.

In fig. 1F has its heating element moved so far upwards that its lower edge is at the same height as the lower edge of the column 2, so that a new cavity 6 is created. The liquid material which flows out through the tube 5 in fig. IB and 1F constitute the material which, in a system with k less than 1, is richer in solute than the material which is in the upper part of the column.

While molten zones in the enrichment section migrate towards and the enriched product away from the feed opening, see fig. IA-1F, the movement of the molten zones and of the purified product takes place in the stripping zone in the direction of the outlet opening. Cavities in the stripping section can be provided by means of a "cavity generator" similar to an inner tube 5 as shown in fig. IA—1F. Here the relative movements of voids and molten zones are different, in that the molten zones are caused to migrate in a downward direction while the voids bubble up through the molten zones as they come into contact with each other. The volume of the cavities is determined by the relative length of the heating devices and the inner tube. The volume of the cavities in the removal and in the rectification sections can be regulated independently of each other.

The creation and migration of voids in the removal section will appear more clearly from fig. 2A—2F. Also these figures only show the lower part of a stripping column 7 during different stages. The first two figures show how the cavities are produced when the apparatus is put into operation. As shown, the heating devices 8 here move in a downward direction. The cavities are produced by means of an outer outlet tube 9 which constitutes another type of cavity generator. Such cavity generators will be discussed in more detail in connection with fig. 8A—8H.

According to fig. 2A, the column 7 contains only fixed and floating zones 10 respectively. 11 when the apparatus is started, and thus for the time being no cavities. The molten material in the lowermost zone 11 is held in place by the solid material located at the lower end of the tube 9. Fig. 2B shows the heating devices after they have moved a short distance. The lowermost heating device is located directly outside the pipe 9 and as a result causes part of the molten material to flow out and create a first cavity 12. Fig. 2C shows the heating devices 8 after a further movement. The only change is that the zones 11 have moved further down and the height of the lowermost fixed zone 10 has been reduced. The cavity 12 cannot yet move through the fixed zone and remains in place.

In fig. 2D, the bridge over the cavity 12 has melted so that the cavity has bubbled through the molten zone 11 and is now immediately above the molten material in the lower zone 11.

In the event of a further movement of the heating elements, the cavity 12 will be enclosed in the fixed zone, see fig. 2E. The lowermost molten zone 11 has continued to move downward.

Fig. 2F shows the same situation as fig. 2B. Part of the molten material has flowed out through the tube 9 and created a new cavity 12. The further movement of the heating elements 8 causes the molten

zones move further downwards and produce further cavities. Figs 1 and 2 thus show how material rich in solvent leaves the enriched section and material rich in solute leaves the stripping section, in both cases in a system where k is less than 1, i.e. where the concentration of solute is greater than the melted phase than in the solid phase on the collision surfaces at equilibrium. Fig. 3 schematically shows an enrichment or stripping section depending on the direction in which the heating elements move. The figure also shows an intake pipe for added material. In the processes described here, there must be a flow of treated material from the food nirig intake to each of the enrichment and removal outlets. It shall be assumed that it is desirable to extract purified solvent, so that the outlet in the enrichment section is to be regarded as product outlet and the outlet in the stripping section as waste outlet. It is clear that the roles of the outlets can be reversed or that they can both be considered product outlets. The material in a liquid state is made to move or flow by providing voids in the column at the waste and product ends and by causing these voids to move towards the feed end where they are filled with new material. As explained in connection with fig. 1 and 2, the cavities are produced by allowing some material in a liquid state to flow out of the apparatus. The cycle which constitutes the generation of a cavity which is allowed to move towards the intake end where it is filled with new material causes a product flow through the column in the direction of the cavity generator. The spreading rate is determined by the size, number and travel speed of the voids.

The one in fig. 3 shown column 13 serves as an enrichment section if the heating devices 14 move in the direction from below upwards. If the length of the heating devices in the direction of movement is h and a sufficient amount of heat is exchanged, so that all the solid material inside the heating device is melted and all material outside the heating device is solid, it will be seen that a movement of the heating devices in the upward direction will produce molten zones 14a, solid zones 15 and cavities 16, the latter having a height equal to h—-1, where 1 is the length of a molten zone. If the inner tube 121 has a cross-section that can be ignored compared to the entire cross-section of the column, the length of this tube can be considered equal to the height of a molten zone. This simplified consideration shall be applied throughout this description, so that 1 will indicate both the length of the molten zone and the length of the cavity generator tube. As a cavity 16 moves into the area of the intake pipe 17, new material will flow to and fill this cavity. Any material in the cavity that has a lower specific gravity than the added material will bubble out through the intake pipe.

When the heating devices 14 in fig. 3 is moved in a downward direction, the apparatus will act as a drainage section and the cavities 16 will behave in the manner described above in connection with fig. 2A— 2F, as they bubble through the molten zones 14a in an intermittent manner to finally reach the height of the intake opening 17 where they are gradually filled with new molten material which is fed in.

Figs. 4, 5 and 7 show three different designs of column-cavity-zone refining devices. Before these designs are described in detail, certain general requirements for such zone-cavity devices must be briefly mentioned. As the flow of material from the intake into the cavities will usually take place under the influence of gravity, the column is carried out either vertically or at an angle. If the cavity is evacuated or contains a material with a lower specific weight than the molten material to be processed, the intake opening must generally be higher than the product and waste openings. If, on the other hand, the specific weight of the material in the cavities at the working temperature is greater than the specific weight of the molten material to be processed, the intake opening must be located lower than the two outlet openings. The added material will then rise through the liquid in the cavities.

In the construction of the apparatus, it is necessary that the present flows of the molten material are limited in such a way that molten zones can be maintained in the column at all times. Ways to achieve this shall be discussed in more detail with reference to fig. 8.

Since, in the method according to the invention, it is generally desirable to keep the material being treated in a molten state within the individual heating elements and the material outside the heating elements in a solid state, the walls of the column are made of constant thickness with the heating elements arranged close to the walls. In some cases, it proves appropriate to cool the surfaces of the column that are outside the heating elements if columns with a large diameter are used and the melting point of the material being processed is relatively close to the ambient temperature, as well as where the heat transfer is not sufficient for other reasons .

In the description of the different types of apparatus, it is not only assumed that the material being treated is melted when it is inside the heating elements and that the other material in the column is solid, but also that the material that is supplied . the intake opening, is at all times in a molten state or in any case in a molten state for a sufficiently long period for the material to flow into each cavity as it passes past the intake opening. It is also necessary that the various zone-cavity apparatuses are designed so that the walls of the outlet pipe are not kept permanently cold so that a plug can form which prevents the necessary flow.

It should be noted here that it is not necessary for a cavity to occupy the entire cross-section of the column as indicated in fig.

1, 2 and 3. In the case of a vertical column with a circular cross-section, a small cavity will generally take the shape of a ring. If the column is tilted, the largest part of the cavity containing material lighter than the molten material on which it "floats" in the enrichment section and through which it bubbles into the stripping section will tend to migrate towards the part of the column wall that lies at the top when the displacement takes place from the upper part of a molten zone.

Fig. 4A and 4B show a simple refining device of the zone-cavity-column type seen from the front according to. from the side. It contains a rectification section 18, a stripping section 19, a feed container 122 with associated heating coils 123, a product outlet 20, a waste outlet 21, heating devices 22 and means for moving these heating devices. As can be seen from fig. 4A, the apparatus contains two sets of heating devices 22 which are attached to plate-shaped carriers 23 and 24. During operation, the carrier plate 23 is moved slowly upwards and the carrier plate 24 slowly downwards by means of a line 25 which is laid over line wheels 26 and is driven by means of a motor 27 and a drive wheel 28. During operation, the heating elements 22 will therefore move upwards in the enrichment section and downwards in the removal section.

After the heating devices have moved a whole number of intervals where one interval corresponds to the mutual distance between corresponding points on the heating devices, an arm 29 which is seen in fig. 4B and which is attached to the carrier plate 23, affect the coupling contact 30 which releases the magnetic coupling 31. The weight 32 causes a rapid backward movement of the heating elements until the arm 29 affects the switch 33 which causes a new engagement of the coupling 31 so that the described cycle is repeated. In order for the apparatus according to fig. 4 is to work, it is necessary that there is molten material in the feed container 122 and that the upper part of the upper heating device 22 in each of the columns 18 and 19 is located sufficiently close to the container 122, so that a free exchange of molten material within this heating element and the material in the container, and so that the cavities within these heating elements can escape up through the container. The container is heated by means of heating coils 123 and a power source not shown or by other suitable means. If the material to be treated is slightly contaminated and if the container is open, a protective liquid layer can be provided on top of the molten feed material. During treatment of e.g. tin, it has proved beneficial to use such a protective layer of lamp soot.

By the one in fig. 4, the mutual distance between corresponding points on the heating elements is constant. The elements are moved a total distance equal to a whole number of these intervals. After the hot elements have moved a whole number of times, e.g. one, they are quickly moved back the same distance, so that each hot element will be outside the molten zone that was previously inside the element, after which the movement is repeated. The hot elements must be moved back quickly enough to avoid any appreciable amount of the material in the molten zones solidifying.

The smallest number of heating elements that can be used for an entire column according to fig. 4 are two, one for the enrichment section and one for the stripping section. As will be seen, with the help of this very simple equipment with only two heating elements, it is possible to achieve the effect of a large number of separation steps. A time saving is achieved, however, by using a maximum number of heating elements as close together as possible, as the mutual distance will depend on the efficiency of the heat transfer in the column. In the enrichment section, this minimum distance is determined by the requirement that a bridge of solid material is controllably maintained between the molten zones corresponding to the adjacent hot elements.However, in the drift zone, it is necessary that two solid bridges with an intermediate cavity be maintained between successive molten zones during part of the cycle, so that the minimum heater spacing in this section of the apparatus becomes larger than for the enrichment section, more precisely of the order of magnitude twice a cavity height.

The practical number of heating elements will primarily be determined from purely economic points of view and will usually be a compromise between the shortest separation time and the lowest equipment costs.

When reciprocating heating elements are used, these must be found in a sufficient number so that the extreme positions of the first and last heating element in each section at a given distance contain at least the length of the section.

An advantage of this reciprocating type is that the device construction is somewhat simpler than the construction where the heating elements move continuously. The heating elements can be firmly and permanently connected with a heating support plate, as shown in fig. 4 and can form closed rings as they do not need to pass past the intake opening.

Fig. 5 shows a refining device of the zone-cavity-column type where the heating elements move along a circular arc. In this case, the heating elements 34 are attached to a rotating part 35 and enclose a column 36 of inverted Y shape. The heating elements are designed in such a way that they can pass unobstructed past the intake pipe 37. The cavity generators 38 and 39 work in a manner as already described. The material 40 to be supplied is kept in a molten state by means of heating coils 41 which are placed around the container 42. In the case of a system of solid substance and solvent with k less than 1, the desired product will flow out of the pipe 38 and waste product out of the tube 39 when the heating elements are moved anti-clockwise.

In the embodiment shown in fig. 5, the Y-shaped tube 36 is bent or shaped into a circular arc comprising approx. 180°. The hot elements 34 are moved continuously. If, however, one were to prefer a reciprocating movement as in fig.

4, the heating elements 34 do not need to pass past the intake pipe 37, so that they can be of the ring type. With a continuous movement of the heating elements as in fig. 5, rectification and stripping can be achieved using a minimum of one heating element.

In all of the column devices described above, the material in the column is selected in accordance with the substances to be treated. The construction material can be glass, quartz, plastic, ceramic, graphite, metal or other materials, or the column can be made of two materials such as mainly metal or similar and partly of glass, quartz, mica or other transparent materials in order to be able to see the process. Such constructions of two materials can also be useful in order to avoid possible difficulties due to expansion of the molten or solidified material which is being treated, although by taking sufficient precautions, the column can also be made entirely of a brittle material. In order to prevent the fixed zones from sliding downwards, especially where it is a matter of material that contracts during solidification, the inner wall of the column can be made rough, grooved or with small projections, support rods or incisions with which the solid material forms an engagement. Optionally, the solid material can be held up by a chain, string, a metal rope or a pipe suspended in the column. If a tubular carrier made of a relatively deformable material is used, this will absorb tension as a result of expansion during the hardening and thereby enable the use of brittle material in the column wall.

Sections of different column tubes and heating devices can be seen in fig. 6A, 6B, 6C and 6D, the former showing a tubular heating element 43 with windings and devices at intervals (the windings are not shown). This heating element moves within a hollow cylindrical column 44 which contains the material to be treated. Externally, the column 44 can be provided with a thermal insulation 45. Fig. 6B shows a section of a more conventional type of a tube-shaped refining apparatus where the material moves in the tube 46 and the heating element 47 of the ring type moves outside the tube. Fig. 6C shows a section of a U-type heating element 48 which encloses a column 49 with a rectangular cross-section. Such a form is applicable where the heating elements must be able to pass past the feed opening or support means. Fig. 6D shows a column 50 which has a rectangular cross-section and which is heated by means of a plate-shaped heating element 51. The heating elements can be constituted by resistance windings, gas flames, induction coils or other known means.

A refining device of the zone-cavity type with a large cross-section appears in fig. 7. The apparatus essentially consists of two concentric half-cylinders 52 and 53 with strip-shaped heating elements

54 which is mounted on a rotating drum or frame not shown. The feeding takes place through the intake 55, and at a

efficient, continuous movement of the heating elements in a clockwise direction, provided either by a constant movement of the heating elements in a clockwise direction or a reciprocating movement with a controlled refining speed in a clockwise direction and rapid movement back a whole number of heater distances in the opposite direction, the product is withdrawn from outlet 56 and the waste from outlet 57, which two outlets at the same time

serve as cavity generators. By keeping the thickness of the column, i.e. the distance between the two half-cylinders 52 and 53, small, the zone lengths measured in the direction of travel can be kept small to thus allow a large number of simultaneous zones or steps. As will be described in more detail later, the degree of separation achievable increases with the length of the column expressed in lengths of molten zones.

In the methods mentioned above, it may be appropriate to provide cooling between the zones, particularly if it concerns substances with a low melting point or a high heat conduction coefficient. This can e.g. take place by blowing cooling air or other gas towards the areas between the heating elements, immersing the device in a cooling liquid or arranging tube-shaped heat exchangers around the column. If the latter method is used, some of the heat can be returned to the system in ways well known in chemical processes.

It has been mentioned previously that the cavity generator described in connection with fig. IA— 1F, is only to be considered as an example. Other types of cavity generators are shown in fig. 8A—8H. The main requirement for the cavity generator is that it limits the flow of molten material and generally replaces the part of the molten zone that has cleared out with some material that is liquid at the device's working temperature. If a cavity liquid with a lower specific gravity than the molten material being treated is used in the system, this can be achieved by keeping the cavity generator immersed in this liquid, or if the cavities are to contain material in gaseous form, the entire apparatus can work in such a gas atmosphere -fear. An example of the latter is the refining of germanium or silicon, where the entire process usually takes place in a protective atmosphere of nitrogen or hydrogen.

As already mentioned, for the sake of simplicity, above we are talking about cavity liquid with a specific weight that is less than the treated material when this is in a molten state.

It is necessary that the path of the heating devices is chosen so that the material at the end of the outlet pipe is kept molten during part of each cycle so that it can flow out, and furthermore that the material in the outlet pipe is kept solidified during part of each cycle to prevent it from leaking out. In certain cases, it may be appropriate to provide for a cooling of the outlet pipe during certain intervals of the treatment.

Fig. 8A shows a longitudinal section of one type of the cavity generator where the tube faces out. It appears that the pipe 58 has cooling fins 59 and is glued to the lower end of the column 61 by means of a binder 60. The purpose of the cooling fins is to speed up the heating and cooling in the transverse direction so that the outlet pipe will quickly be exposed to the effect of the heating device and so that the cooling will also take place quickly. The walls of the tube 58 are suitably thin.

The cavity generators according to fig. 8B and 8C work identically with the generator of fig. 8A, as there are only small deviations. For fig. 8B's case, the deviation lies in the placement of the cooling ribs and in fig. 8C in the form of the outlet pipe. The embodiment according to fig. 8B consists of a collar 62, an externally directed tube 63 and cooling fins 64. The generator according to fig. 8C also has a collar 65, a conical outlet pipe 66 and cooling ribs 67. In both cases, the connection to the column 61 takes place by means of an adhesive 60.

Fig. 8D shows a cavity generator in its simplest form consisting of a column 68 and a narrowed outlet tube 69.

In the embodiment according to fig. 8E and 8F have the cavity generators ventilation pipe 72 respectively. 73 which acts to convey a stream of molten material out through the pipes 71 acc. 75. These ventilation pipes can lead out to the free air, to a protective atmosphere or be connected to a gas or liquid source under pressure. A silk ventilation may be desirable near the treated material has low specific gravity or high surface tension. The generator according to fig. 8E consists of a pipe which is directed inwards into the column and sits on an end piece which is attached to the column 61 by means of a collar 70. In the embodiment according to fig. 8F is a ventilation pipe 73 led out from the side wall of the column 74 and bent upwards so that the distance between the point where it leads into the column and the opening to the atmosphere or other gas or liquid is greater than the largest vertical dimension of the molten zone. The cavities themselves are also produced here with the help of the inwardly directed tube 75.

In the embodiment according to fig. 8G and 8H, the outlet pipes 77 and 79 are connected to the columns 76 respectively. 78 and otherwise works in the same way as an inwardly directed pipe.

The only requirement for a cavity generator is that it must remove a controlled amount of the material in each successive molten zone, the cavity being generally filled with a gas or liquid whose specific gravity is different from the specific gravity of the material being processed. This can be achieved in different ways, so you can e.g. instead of the cavity generators shown in fig. IA or 8A to 8F, use valves or plugs which are affected by the moving heating devices and only allow part of the material in the molten zone to escape.

In order to obtain small values of the ratio between the cavity volume and the volume of the molten zone, it may be desirable to reduce the diameter of a part of the column closest to the cavity generator. Furthermore, it is not necessary for the material to be removed and the cavity formed at the point in time when the heating device is located at the end of the column, although this would be the most appropriate in the closed case. The cavity can be formed when the heating device is located at a certain distance from the cavity generator, which in that case may consist of a separate, stationary heating coil located at the end of the column and which is affected by the position of the heating device.

An important factor must be taken into account when choosing the cavity generator type for refining equipment as described here. As will be shown, the degree of separation that can be achieved is proportional to the column length expressed in zone lengths. Consequently, it will generally be preferable to use a generator that draws molten material out from the end of the column so that the entire column length can be utilized. The generators shown in fig. 8A—8D, satisfies this requirement. In the case of zone generators that consist of an inwardly directed pipe, such as e.g. in fig. 8E-8H, the effective length of the column will decrease with the length of the pipe.

For the column's drift section, this factor is of particular importance. Here, there is no supply of fresh material to the molten zone, and as the material reaches the waste outlet, what is in the last zone length will be subjected to solidification by normal cooling, so that the last part that solidifies in this zone will have a greater concentration of solute than anywhere else. If one pulls this last part out using an external tube. you will get a more efficient removal. As the greater part of the drift zone in certain systems depends on the separation provided by the normal solidification in the last zone in the direction of the molten zone's movement, the use of an inwardly directed tube will in reality, as will be shown later, significantly reduce efficiency.

Some of the advantages of zone-cavity refining devices implemented in accordance with FIG. 4A and 4B as well as 5 and 7 will be self-evident. The equipment is very simple, as the column in its simplest form is made as a U-tube or semi-circular tube with an intake opening. The column itself contains no moving parts and no solid material moves in the column, where all movement is due to gravity and only takes place in the liquid phase. There is no need for valves or control of the flows because the fixed zones block the liquid flow and the flow rates and recirculation conditions are easily varied by varying the speed and capacity of the heating devices and by varying the dimensions of the cavity generator.

A typical small refining device of one of the above-mentioned types according to fig. 4, 5 and 7, which are suitable for work in the laboratory, can have a column with a diameter of approx. 12.7 mm or more and an outlet pipe with a diameter of 1.5875-3.175 mm or more. The zone lengths can be approx. 25 mm with the mutual distance between the zones of the same order of magnitude.

The processes according to the invention can in certain cases be carried out more quickly with spiral or helical devices (hereinafter for brevity referred to as helical devices) as shown in fig. 9A. The advantage of this device design lies primarily in the fact that only one heating direction is required for all zones, so that the heating problem is substantially simplified while a large number of molten zones can be achieved within a small space. The movement is then reduced to a simple rotational movement around the helix's central axis. Heating and cooling can take place by immersing the lower part of the apparatus in a bath, although in certain cases it is necessary to use other heating methods. The use of screw line apparatus for conversion from solid to vapor form is described by A. F. Reid, in "Industrial and Engineering Chemistry" vol. 43, 1951, page 2151.

Screw line devices can be made of a coiled tube, or of curved sections between inner and outer concentric cylinders, if larger cross-sectional areas are required. Such cylindrical devices need only contain a single screw-shaped wall that separates the successive steps or can have a double wall, so that the hot or cooling liquid can be brought into contact between the windings and thereby improve the heat transfer.

Screw-shaped refining devices can work with the axis in any position, i.e. horizontal, vertical or inclined, the most important requirement being that the part of the helical tube with which the molten material and the cavity liquid are in contact assumes such an inclination in relation to to the horizontal that the material can move.

Cavities can be generated in the helical tube in the same way as in a column by providing a controlled portion of a molten zone at the end of the tube, which can flow out of the outlet tube containing a cavity generator. In the same way as in column refining, the cavity generator can work according to the principle that the outlet pipe is blocked by solid material when a desired part of the molten zone has cleared out. This principle can be applied in any position of the helical tube. Another principle for controlled cavity generation which can be used with horizontal or inclined screw line pipes and is particularly suitable for screw line refining devices of the cylindrical type discussed in more detail below with reference to fig. 10, makes use of an outlet pipe which is open at all times (being sufficiently hot to keep the material in a liquid state) and which has such a shape or position as to allow a portion of the molten material in a molten zone to release during each revolution.

The apparatus according to fig. 9A and 9B consists of an enrichment section 81, a stripping section 80 and a feed section 82. Cavity generators located at the outer ends of sections 80 and 81 are not shown.

In the zone-cavity refining device of the helix type, which has both enrichment and stripping sections, as with the column type, it is necessary to generate voids in the waste and product outlets and cause the voids in both sections to migrate towards the common intake opening located between the two sections. In the same way as with the column equipment, the voids and the molten zones in the enrichment section must travel in the same direction, but in the opposite direction in the stripping section.

The directions in which the cavities and zones will travel in the helical tubes

depends on whether the spiral line is clockwise or counterclockwise and whether the direction of rotation is with or against the sun. By

helical tubes whose axis lies horizontally, the location and extent of the annular sector forming the heated area is of importance. In the case of helical pipes whose axis is vertical, the heating devices can be arranged anywhere around the periphery of the windings.

In the construction of helical refining devices, different combinations will. of winding direction (right or left), direction of rotation (with or against the sun) and the location of the heating device work satisfactorily. This is evident from the following table, which also shows the location of the heating device by horizontal screw lines.

When the device according to fig. 9A and 9B rotate with the sun as seen from the right side, a product rich in solvent will flow out on the right side and a waste rich in solute will flow out on the left side. The intake pipe 82 for the material to be treated takes a fixed vertical position and is kept heated so that the material here will always be in a molten state. The connections between the intake 82 and the removal and enrichment sections 80 and 81 take place by means of rotatable connecting joints 83 according to 84. Although the heating in the enrichment section can take place with the help of a hot bath in which the lower part of the helical tube is kept submerged, the heating in the stripping section must take place with the help of other means for systems where k is less than 1. The heating device 85 which best seen in fig. 9B, produces the desired effect in the stripping section. The heating device 11

85a is shown in the correct position for the enrichment section. This will be discussed in more detail in connection with fig. 11.

Fig. 10 shows a helical drum cavity generator in end view and suitable in connection with the enrichment section of a helical zone cavity refining apparatus. In the annular space between the concentric cylinders 92 and 93 are located

there is a molten region 87 provided by means of a heating bath 88, and a solid region 89. When the cylinder

is in the position shown, has a portion of the molten material in the area 87

cleanly out through the outlet pipe 90 which is stationary in relation to the drum and causes the creation of a cavity 91. As the cylinder rotates towards the sun, solid material 89 will melt and fall into the molten area 87, while molten material at the separation surface 94 solidifies as the drum revolves. For each full revolution

ning of the cylinder is produced where a new cavity 91, which cavities are caused to migrate towards the molten zones 87 inward direction until they reach an intake which has the same effect as the intake 82 in fig. 9A. In this intake, which is not shown, new material will displace the "liquid" located in the cavity, which movement corresponds to that which takes place in the enrichment section of a column device described with reference to fig. 1.

Cavity generators used in connection with equipment according to fig. 10, can be led directly radially out through a side wall as shown, or be led axially out through the end of the drum.

In the drift section, it is necessary that the cavity moves opposite to the molten zones. This is achieved in a column device by the heating devices and consequently the molten zones moving downwards, while the lighter cavity material will bubble up through the molten material. The same cannot be achieved with a horizontally lying screw-line device with the melted area having a position and direction of rotation as shown in fig. 10, as the cavity fluid, which is lighter than the material in the molten state, will continue to flow on the surface of the molten zone.

Figs 11A, 11B and 11C schematically show a way to achieve drift. A molten zone 95, fig. 11A, is produced by means of a vertical heating device 96. It is assumed that the cavity 97 has already been generated in one or other of the ways described above. The upper visible zones are the fixed zone located at the front end of the screw winding, and the fixed zone 99 which is visible through a broken part of the screw winding and is located in the only turn behind the plane of the paper. As the screw line rotates, the direction now being with the sun as the refining device is viewed from the waste end, cavity 97 will be enclosed in the solid material 99 as shown in fig. 11B. The molten zone 100 is located in the turn of the helical tube which is closest to the turn containing the molten zone 95, and the solid zone 101 is that found in the third turn of the tube. In the position according to fig. 11C, the cavity 97 has entered the next molten zone 100 and has bubbled up through this to the solid zone 101, thereby causing the molten zone 100 to travel in the opposite direction to the cavity 97. As in fig. 10, the cavity 97 continues to travel in an inward direction relative to the plane of the paper until it reaches

the intake, while new cavities are created for each full revolution of the tube.

The effect in the enrichment section of a zone moving vertically is similar to that in fig. 10, where the rotation now takes place towards the sun viewed from the enrichment end, with the solid and molten zones in the same relative positions as shown in fig. 11A, 11B and 11C. Here, the solid material, as it melts, will fall down through the voids and into the molten areas, so that the voids will never be closed into solid material and so that voids and molten zones move in the same direction.

It is known in connection with separation methods that the residual flow through the various stages must be different when you want to achieve an effective effect, where a desired component is exposed to a high percentage of enrichment, see e.g. H. D. Smyth, "Atomic Energy for Military Purposes", Princeton University Press, 1945, page 167, expressed in another way, the downward current should vary and be greater for successive steps as the intake is approached. The effect of the variable downward flow can be easily achieved by a zone-cavity process by varying the volume of the molten zones as they migrate. This can either happen by varying the cross-sectional area of the zones and keeping the zone length constant as in fig. 12, or by varying the zone length in a column with a constant cross-section, as in fig. 13, or by a combination of both of these methods. The zone lengths can be varied by varying the amount of heat that is transferred from the heating devices to the material, such as by varying the surface area of the heating devices, see fig. 13, by varying the temperature of the heating devices, etc. The apparatus according to fig. 12 operates in a manner identical to that shown in FIG. 4A and 4B, as supplied material is introduced into the intake 112 and product is taken out by the cavity generator 102 and waste is taken out through the cavity generator 103. In the embodiment shown, the heating devices 104 move back and forth, moving gradually upwards in the enrichment section 105 and downwards in the stripping section 106 and then quickly back to make another cycle. The heating devices 104 can be moved in the same way as shown in fig. 4A and 4B.

Fig. 13 shows a section of a column 107 with heating devices 108 whose surface area varies in the direction of the column. The column contains solid zones 109, molten zones 110 and cavities 111 which are produced by means of cavity generators not shown.

With this device, the heating devices 108 move back and forth. If the movement takes place slowly in an upward direction, the section shown is the enrichment column, but if the slow movement takes place in a downward direction, fig. 13 the drift section for a system with k less than 1.

In what follows, the method according to the invention and the operation of the described apparatus will be treated more theoretically. The basic characteristics of the zone-cavity processes must be treated in analogy with the known multi-stage processes used in distillation. However, it will be clear that there are also fundamental deviations between these two processes. The mathematical equations must be given for the initial concentration expressed on the basis of the concentration of the added or untreated material and significant parameters of the equipment. There must also be a calculation and a general description of the most important aspects of the construction of the equipment.

First, a distillation column must be considered that works with total reflux, i.e. under the condition that no waste is removed. When the steady state is reached, there is a rising flow of steam and a falling flow of liquid in the apparatus. The number of steps that such a column must contain to separate added material into top fractions with the composition yD and bottom fractions with the composition x,, is given by Fenske's equation:

where x = fraction in moles of the lighter component in the liquid phase and y = fraction in moles of the lighter component in the vapor phase and where both values x and y apply to a given step. The number of stages given by Fenske's equation (1) is the smallest number that can give the specified separation when the column works with total reflux. As the flow of added substance, waste and product must in practice be constant, the process must work with partial reflux, which gives a slightly smaller separation per steps. The minimum number of steps for partial return will in practice be in the range of 1.5-3 times the minimum number of steps according to Fenske's equation.

In a distillation column, the number of stages is proportional to the number of plates in the column of bubble-through plates or proportional to the height of a column of packed material.

The above discussion of the theory of distillation and much of the general theory which shall not be discussed further here appears in the article "Multistage Separation Processes" by Manson Benedict in "Transactions of the American Institute of Chemical Engineers", volume 43, February 1947, pages 41-60.

As Fenske's equation in principle gives the number of factors for a, the separation factor (equation 2) which is contained in equation (1) and which is expressed by the ratio

between the concentrations in the outlet, is written

in the following form

where Nni = the smallest number of steps.

As will be seen, the separation that can be achieved by zone-cavity refining under conditions corresponding to the total reflux can be expressed by an equation similar to equation (3).

Benedict also gives equations for the smallest amount of reflux, Lllin at which a cascade is just capable of separating, and highlights the following features of 1) It is inversely proportional to the enrichment factor a—1, so that Lmin must increase over time as a approaches 1. 2) It increases as the difference in concentration between the material in the step and the product increases. For this reason, the return flow at each end of the apparatus can be reduced. Examples of how this can be achieved are shown in fig. 12 and 13.

The minimum amount of reflux can also be expressed in terms of a minimal reflux ratio L/D where L is the amount of reflux and D is the amount of the top fraction. Regarding this expression, e.g. refer to Badger and McCabe "Elements of Chemical Engineering", 2nd edition, 1936, page 351. It will be shown in the following that the relationship between the molten zone length and the cavity length in the zone-cavity refining device is analogous to this to-reflux ratio.

Certain analogies between zone-cavity separation and distillation and other known separation processes will be apparent from the following:

Relationship between a and k.

The relationship between the distribution coefficient k for zone refining and the separation factor a in connection with distillation will appear from the following: In binary systems that include liquid-solid transfer, it has become common to regard one component as the solute and the other as the solvent and express the ratio between the concentration of solute in the solidified phase and in the liquid phase from which the substance solidifies, by the above-mentioned distribution coefficient k. Procedures by means of which the equilibrium values for k can be calculated with reference to the state diagram of this system as well as methods by means of which these values can be found based on thermodynamic considerations, are well known, see e.g. A. Hayes and J. ChirJman, "Transactions A.I.M.E.", volume 135, page 85. Effective values of k in the solidified state differ from the values in the equilibrium state. The effective value has, where it deviates, the effect of reducing the value of the quantity 1-k. Expressed briefly, it can be said that the deviation from the value under the equilibrium state is due to poor diffusion of the solute in a molten region or an existing diffusion of solute in the solidified surrounding

advise. Procedures which effect that

the effective value of k approaches the value under the equilibrium condition and consequently a

improving the separation efficiency consists in stirring in the molten area, maintaining a sufficiently low crystallization ratio so that dissolved matter will only be held to a small extent in the solidified zone, and increasing the temperature gradient at the transition surface between liquid and solid form.

The effective k-values for different binary systems are discussed in the literature,

see e.g. above reference Hayes and

Chipmanfor k-values for different solutions in iron; W. G. Pf ann «Journal of

Metals", vol. 4, page 861, and R. N. Hall "Physical Review", vol. 88, page 139, for germanium solutes, and R. H. McFee "Journal of Chemical Physics", vol.

15, page 856 for dissolved substances in sodium chloride.

The relationship between k and « can be derived by the following illustrative calculation:

Calculation 1.

It is assumed that germanium containing arsenic as an undesirable solute must be purified. A k-value of 0.1 is assumed. Purified germanium is assumed to be withdrawn at the top of the column. Consequently, y„ = fraction in moles of germanium at the top and x„ = fraction in moles of germanium at the bottom, where y applies to the solidified solid and x to the liquid solid. A very low arsenic concentration of less than . than 1 percent, such materials have a wide application for the manufacture of transistors, yD, <y>x and x, become; all very close to 1.

As mentioned above is

According to the above equation (2) is:

Consequently, in the case of zone-cavity-refiner-irig, in general a = l/k.

In zone-cavity processes, the direction in which the components migrate is as follows: 1. In a system where the k-value of a solute in a solvent is less than 1, i.e. a system where the presence of solute causes a lowering of the melting point, the concentration of solute is greater in the molten zone than in the solidified zone on the transition surface, and consequently the solute migrates with the molten zones while the solvent migrates in the opposite direction. 2. In a system where k is greater than 1, i.e. where the melting point of the mixture is higher than that of the pure solvent, the solvent migrates with the melted zones and the dissolved substance in the opposite direction.

As the refining process above has been described as a process for extracting purified solvent from a solution where the solute in question has a k-value less than 1, the molten zones enter the column at the end where the product is taken out (the "top" of the column).

The ratio between the length of the distillation column and the zone refining column. In zone refining, the state which corresponds to the equilibrium state in total reflux in distillation is an equilibrium state which has been reached after a large number (theoretically an infinite number) of zones have passed along the mold. If there is no supply of new material or removal of product, this equilibrium state in zone-cavity refining for total reflux will correspond to the conditions in batch-zone refining. The resulting distribution as explained in the above article "Principles of Zone Refining", gives the solute concentration C as a function of the distance x from the exit end of a mold or charge as follows:

where 02 where 1 = the zone length, L = the mold or column length and Cn = the mean solute-solute concentration. It appears from equation (4) that the constant A for x = o (the end of the column where the molten zones enter) is equal to the solute concentration. In zone-cavity refining, the ratio A/C(), which is a purity ratio, can therefore be written Cp/Cf, i.e. the ratio between the solute f concentration in the product and in the added material in the enrichment section of the refining device. By solving equation (6) for a wide range of values of k, 1 and L, it has been found that A/C„ can be empirically expressed

where f is a factor indicated in table II below and N =p= L/l = the number of molten zone lengths in the mold or the mold section. Equation (7) has the same form as the rewritten Fenske's equation (3) and expresses that fN is the number of separation steps that can be produced by zone-cavity refining with total reflux in a column section of length N.

Values of the factor f in equation (7) corresponding to four different numbers of melts, zone lengths N and five different values of the distribution coefficient k, appear in the following table:

The flow of substances in a multi-stage continuous separation process can be divided into two parts, namely the flow of added material, waste and product and the flows that form counterflow or the return flows in the process. For example in ordinary distillation, these streams can easily be distinguished, as noticeable amounts of substance move up and down the column, where the flow rates can easily be measured.

The zone-cavity process is the net flow of added material, waste and product, denoted F, W and P, determined by the volume flows of cavities and stands in the following relation to each other:

In the enrichment section, the net volume flow P from the feed inlet to the product outlet is equal and opposite to the volume flow rate of cavities from the product outlet to the feed inlet according to the following equation:

where A = the cross-sectional area of the column, v = the linear velocity of the cavity movement, (h—l) = the length of a cavity and de = the length between successive cavity fronts.

A similar equation can be derived for the drift section:

In equation (9), the foot index e applies to the enrichment section and in equation (10) the foot index s to the drift section. These parameters can be controlled independently of each other in each of the two sections. The streams of waste and product can also be considered in another way, remembering that a cavity travels through a column section every time a heating device passes past the end of the section.

In normal distillation, there is a countercurrent, namely rising steam and sinking liquid, and calculations for the construction of the column include these currents in the form of backflow conditions. In zone-cavity processes, the counter-flow between solid and liquid phases is provided by means of the migrating molten zones. While the liquid in the zones does not flow in the same direction as the liquid in the distillation column, the effect of the solute distribution as a molten zone migrates down the column is in principle the same as if the same volume of liquid actually flowed down the column for itself to form equilibrium with the solid substance during the movement. In zone-cavity refining, the quantity that is analogous to the downward flow during distillation is, in principle, the volume velocity of the flow of molten zones downwards in the column, i.e. in the direction towards the waste outlet.

When the refinery is in normal operation, there will be one cavity for each molten zone, so that the ratio between the molten volume and the volume of the cavity within a heating device will be analogous to a reflux ratio determined by the effective length h of a heating device, by the temperature and by the effective length 1 of the cavity generator outlet pipe. Consequently, l/(h—1) can be considered as a backflow ratio for understanding the process. The working equations given below do not use the reflux ratio as such.

Another feature of the molten zones in relation to the media that flows back is the degree of contact between the counter-flow basins. In distillation, there is a limit to how close one can get to the ideal state of equilibrium, as this is limited by the degree of contact, i.e. the colliding surfaces between solid and vapor, and by the diffusion of components in the liquid phase. In molten-zone-reflow or the zonal flow-back, the degree of contact can be considered perfect, as the solidified phase melts completely and passes into the liquid phase over the entire cross-section of the column. Diffusion in the liquid phase remains within a limit, although this effect can be reduced by using a low travel speed or by using stirring agents. An important feature of the degree of contact achieved by zonal reflux is that the column construction is greatly simplified, as no bubble jackets or packing agents are needed.

The volume flows of added material, waste and product in the rectification and stripping sections shall now be discussed, after which the concentration of solute in the outlet of the zone-cavity refining facility shall be expressed on the basis of the added material's concentration or other important parameters of the process. Separate equations will also be given for the enrichment and drift sections.

The enrichment and stripping sections are fed at the intake, for added material and have a common concentration Cf of added material. Added material that enters the system can be considered as divided into two parts, which travel through each section. The effect of the molten zones in the enrichment section is to remove dissolved matter from the part of the feed material that travels through this section and transfers this matter to the feed container. The effect of the stripping section is to remove solvent from the part of the supplied material which travels through this section and transfers the solvent to the feed tank. In order to keep the contents of the feed tank at a concentration Cf, it is therefore necessary to balance the quantities of waste W taken out and of product P so that Cf remains unchanged. The relationship between P and W will depend on the values of Cp and C„. as follows:

where a and (3 are constants.

The general material balance for refining facilities is expressed by the equation

where F = the feed rates of added material.

The general balance for solute is: FC, = WC„. -[- PC,, or if you insert the values for C and Cw. from equations (11) and (12):

If therefore the concentration of solute in the added material is known, and it is desirable to achieve certain concentrations of solute in the waste and the product, the values for a and (3) can be determined by inserting into equations (11) and ( 12). These equations can then in turn be inserted into equation (15) to give the desired ratio between the product and waste (volume) flows. This ratio can be set in different ways, for example by varying the cross sections of the two sections of the column, by varying the cavity lengths, by varying the distance between the heating devices or in some other way the speeds at which the material is passed through the molten zones.

The equations for the enrichment section.

The following takes place as a molten zone travels through the enrichment section from the product outlet to the feed inlet: (1) The voids are formed by withdrawing a length (h-1) of molten material in

the producing end of the column.

(2) As the cavity travels through the section, the total effect will be to change the solute distribution in the section a distance (h-1) or a cavity length in the direction

towards the outlet.

(3) As the molten zone moves through the section, it carries solute towards the feed intake or in a direction opposite to the change effected as a result of the void concentration. (4) When the molten zone arrives at the feed inlet or material container, the liquid material in the zone mixes with the liquid feed material and assumes the constant concentration Cf. In the system considered, i.e. one where the k-value for solute in solvent is less than 1, the concentration from the molten zone just before it reaches the feed inlet is greater than Cf, so that the effect is a transfer of solute from the product outlet to the feed intake.

Once the equilibrium state is reached, the completion of steps (1) to (4) will leave the distribution of solute in the column section unchanged. As a result of the discontinuity in the concentration at the feed intake, the equilibrium state distribution will in reality fluctuate around a mean value when the column is in operation. However, such a variation will be of no importance as far as the following equations are concerned, except when the section length is very small. It can be shown that the section length Le which

is necessary to achieve a separation ratio a in the enrichment section, is given by

the equation:

where Be, co and XY are constants that depend on h, 1 and k and are given by the equations:

The length Le of the enrichment section according to equation (16), where the foot index e applies to the enrichment section, is defined as the distance from the outlet end to the point where the concentration of solute in the solid is kCf. In practice, this length will usually be approximately equal to the length of one heating device minus the actual section length.

Calculation 2.

For the design of the enrichment section, the correct length of the column section is determined using equations (16) to (19) based on the following calculations:

Given a — C(1/Cf = 0.01, k = 0.5, h = 1; 1 = 0.8. Le is sought. Inserting into equation (17) yields:

I By inserting into equation (18) we get:

By addition in equation (15):

When inserting into equation (19):

These equations give certain limitations with regard to the ratio l/h which is to a certain extent analogous to the concept of the minimal reflux ratio in distillation. Thus, for k values less than 1 there is no lower limit for the achievable value of a, provided that l/h is greater than k. If k is less than 1 and l/h is less than k, there will be a minimum value for a even for a column section of infinite length. If k is greater than 1 (in which case a values greater than 1 are desirable) there is no upper limit for a.

Equations for the drift section.

The following takes place when a molten zone travels through the stripping section from the feed inlet to the waste outlet in the case that the cavity generator is of the outlet type. (1) The heating device is in the starting position and affects the liquid supplied material, x = O, and melts. ter solid substance from O to h, where h is the length of the heating device. The molten part mixes with the liquid feed material. The concentration of dissolved substance in the liquid material that is inside the heating device is therefore identical to the concentration of dissolved substance in the liquid material (Cf). This leads to solute being fed into the stripping section as Cf is greater than the average concentration that was replaced. (2) The molten zone migrates along the drift section, and the first concentration that solidifies at position x

= O, is kCf.

(3) The material in the final length of the molten zone in its outermost position in the stripping section solidifies by normal solidification. This final zone length is defined as Ls-h, where

LA is the length of the exhaust section.

(4) When the collision surface between the solid and liquid zone has migrated off-

the position L.-h + 1, where 1 is the length of the cavity generator's outlet pipe, the remaining liquid part will flow out as waste. This liquid part contains dissolved matter in strong concentration as a result of the drift effect in the section and the normal stiffening effect over the last zone length. The concentration will be calculated below.

(5) The cavity formed each time the heater passes the discharge tube travels through the stripping section of the feed vessel, each time rising through each molten zone as it encounters them. The average travel speed of the cavities depends on the length of the heating device and the mutual distance between the individual heating devices. (6) During its migration, the molten zone hits a number of cavities. For each such encounter, the zone advances one cavity length (h-1). A meeting takes place for each heating device before the heating device that is located at the feed container, because each such heating device in reality produces a cavity which must eventually pass through the relevant heating device. These movements shorten the effective length of the section with respect to the refinement. As in the enrichment section, a varying state of equilibrium is reached. It can be shown that the section length Ls required to produce a separation p = Cw/Cp is given by the equation where Bs, f> and y are constants that depend on h, 1, k and p and are given by the following equations:

The length L. of the drift section in equation (20) is defined by the equation:

where L'K is the real section length and nw is the number of encounters with voids.

The quantity 5 in equation (20) represents the contribution to the concentration of solute that occurs by a normal solidification action when the molten zone reaches the end of the section. The value is the ratio between the concentration in the solid material at the beginning of the normal solidification effect, and the concentration in the liquid part which is ejected as waste somewhat later during the normal solidification process. Under certain conditions, the normal solidification effect represents the largest part of the concentration effect with respect to solute in the stripping section. In such cases, in order to simplify the apparatus, it may be desirable to work with a very short removal section, one heter length long, where the normal stiffening makes up the entire effect.

Calculation 3:

Given h = 1, 1 = 0.9, k = 0.5, |3 = Cw/Cf = 3.5 and L's are to be determined. Inserting into equation (21) yields:

By insertion in (22):

By insertion in (23):

By insertion in (20):

By insertion in (24):

(The section length is so short that you can only have one heating device, consequently nw

= O).

In general, a limited range of p-values is possible for certain ranges of h, 1 and k as L varies from zero to infinity. For k-values less than 1, the given average results show that |5 will lie between the values

as L. varies from zero to infinity.

Above, it has been shown how to calculate the lengths of the enrichment and stripping sections for given values of C, Cp, CM„ h and L. It is assumed that the zone length and number, mutual distance and movement speed of the heating devices are the same in each section , the relative cross-sectional areas of the sections are determined by P/W in equation (15). In the examples given above which are interconnected, P/W was found equal to 2.52 and (h-l)£ equal to 2(h-l)s, which means that the cross-sectional area of the enrichment section for the given example is five times the cross-sectional area of the stripping section.

In the following, certain practical considerations must be made.

Since the degree of separation for given column dimensions only depends on the number of zone lengths in the column section, any degree of separation as indicated by the above equations can be achieved with a minimum of one heating device for each section. As discussed earlier, however, you will achieve a time saving if the heating devices are located as close to each other as possible. A limitation with regard to the smallest mutual distance

between the heating devices in addition to that

as stated earlier, occurs when the diffusion rate of solute in the solidified phase is high, in the extreme case comparable to the diffusion rate in the liquid phase. When such systems undergo treatment, it may be advisable to cool the solid between the heating devices and increase the length of the solidified zones.

A shorter mutual distance between the heating devices brings less 'delay'

love and. a shorter start-up time, i.e.

it will take less time to reach the stable

condition. A small zone length 1 will, as stated above, allow more rectification steps for a given column length and also reduce the start-up time.

The apparatus will approach faster

the stable operating state, the greater the separation constant k deviates from 1. The analogous situation also applies to other continuous

similar processes, such as distillation. This reinforces the value of a continuous zone refining method, because when the steady state is reached, all the material included in the process will be subjected to the maximum separation, and the time required to carry a quantity of feed

rial through the refining device, is relatively short compared to the time required to reach the steady state.

In the same way as with rate zone-

refining, zone-cavity refining is particularly applicable in connection with crystallization from a molten substance and in systems where a quantity of solids can be easily solidified by a uniform collision

surface between the liquid and solid phase and without liquid substance being trapped inside. With regard to crystallization from solvent

means, the zone-cavity process has a significant advantage over rate methods,

as new solvent can be continuously added and removed. For systems be-

standing of solute in solvent

part where the tendency is present to freeze out fine crystals, both progress-

ways have their advantages even if the conditions in certain cases may be less ideal.

As a result of the high degree of con-

trolled purity required in the manufacture of semi-conducting devices, -is it be-

written procedure is particularly applicable in precisely this connection. These feed-

rials for the production of semiconductors consist of, e.g. silicon and germanium with com-

ponents that have similar properties,

e.g. aluminum phosphide, aluminum arsenide, aluminum antimonide, gallium phosphide, gallium

unarsenide, gallium antimonide, indium phos-

fid, indium arsenide and indium antimonide.

The invention has above neces-

show had to be described with reference to special designs, but it is clear that a number of modifications are conceivable, thus it is • clear that refining devices with only one section, either rectification

or stripping section, where feed-

the intake is located near some outlet opening, it can also • ar-

both according to the zone-cavity principle

pet. It is also clear that a back and forth

backward movement of heating equipment

genes can be used in connection with helical zone columns and that this in

bonding with special devices can be advantageous from constructive considerations

nings.

When the above description and in

the subsequent claims speak of "co-

lonne', it is clear that this expression covers both horizontal and vertical columns as well as columns which form some angle with the horizontal, as well as those which are shaped according to a spiral or helical line and all types of chambers with openings at one end or the other and with some usable cross-sectional shape.

Claims (15)

1. Procedure for redistributing the components of a fusible material ale consisting of at least two constituents or substances, in a column where alternating, relatively hot and cold areas are made to travel in one direction from one end of the column to the other, and where the fusible material is melted in the hot areas and solidified in the cold areas, characterized in that material is removed from one end of the column in an amount that is smaller than the amount contained in a hot area, each time the material at this end of the column is melted, and that in another point along the column, fusible material of the system to be treated is added in an amount equal to the amount taken out. 2. Method according to claim 1, characterized in that material is withdrawn from each end of the column and that the supplied amount of the fusible material is equal to the total amount of withdrawn material. 3. Method according to claim 1 or 2, where the column forms an angle of more than 0 degrees with the horizontal, characterized in that the areas are made to move upwards from one end of the column to the other and that the material at the top of the column is taken out as a product. 4. Method according to claim 1 or 2, where the column forms an angle of more than 0 degrees with the horizontal, characterized in that the areas are made to move downwards from the top of the column. 5. Method according to claim 2 or 4, characterized in that material of the system being treated is added at a point between the ends of the column. 6. Method according to claim 2, where the column is U-shaped, characterized in that the material is removed from each hot area in its starting position and that the material is removed from each hot area in its end position of the travel. 7. Method according to claim 5, characterized in that the feed inlet is located at a different height than each of the outlets and that all other parts of the column are located at intermediate heights. 8. Method according to claim 5, where the column has the shape of a helix, characterized in that the movement of alternating hot and cold areas is effected by the helix-shaped column rotating about its helix axis by stationary heat exchangers. 9. Method according to one or more of claims 1-3 and 5-8, characterized in that a cavity is created in the column every time a molten zone is in its starting position for migration at one end of the column, by removing molten material. 10. Method according to claim 4 or 6, characterized in that a cavity is created in the column every time a molten zone is in its final position of travel through the column, by molten material being removed. 11. Method according to one or more of the preceding claims, characterized in that the dimensions of and the , mutual distance between each melted area is chosen so that they encompass the entire cross-section of the column and so that they are connected at all times by material in solidified form. 12. Method according to claim 9, where the column is located in a container open at the top, characterized in that the material that is removed from the column in the initial position is replaced with a cavity material which is essentially immiscible with the material being treated, and that this cavity material is made to travel through the container in the same direction and at the same speed as the hot area. 13. Method according to claim 10, where the column is located in an open top container characterized in that the material that is removed from the column in said position is replaced with a cavity material that is essentially immiscible with the material being treated, and that this cavity material is made to travel through the container in a direction opposite to the direction of travel of the hot region. 14. Method according to claim 12 or 13, characterized in that fusible material is added to the column at a point located at a height that is different from the height of the beginning and end positions, and that cavities are caused to migrate by the action of gravity. 15. Method according to claim 1, characterized in that the movement of the hot area is produced by a corresponding movement of heat sources that move in the direction of travel of the hot and cold zones a distance equal to a whole number of cold areas, after which the heat sources are brought to move in the opposite direction the same number of lengths of cold regions and at such a rate as to allow the material in the hot regions to remain molten but not to allow any appreciable amount of additional material to melt during this backward motion, and that the said series of motions step is repeated at least once.