NO138256B - PROCEDURES FOR ELECTROLYZING A WATER SOLUTION CONTAINING SODIUM AND / OR POTASSIUM IONS AND ELECTROLYSIS CELLS FOR CARRYING OUT IT - Google Patents

Description

Several industrial processes are based on the electrolysis of

aqueous salt solutions. An important process consists in the electrolysis of a solution of sodium chloride for the production of sodium

hydroxide and chlorine. In this electrolysis, in the same way as in other electrolysis processes for aqueous salt solutions,

according to one method, a separation of the anolyte and catholyte compartments in the cell by means of a porous partition wall. The term "diaphragm"

will hereafter be used to denote a porous partition wall which enables electrolyte to pass through the partition wall without significant changes in the composition of the electrolyte.

This method produces hydrogen and caustic soda from wood

the cathode in the cell's catholyte compartment, while chlorine is produced at the anode

in the anolyte department. The salt solution flows through the diaphragm from the anolyte compartment to the catholyte compartment. The cathodes usually consist of iron wire mesh, while the anodes consist of graphite or platinum-coated titanium. The diaphragm usually consists of asbestos.

In diaphragm cells, it is necessary that the flow of solution is sufficient to ensure that back-diffusion of sodium hydroxide to the anolyte is avoided or reduced to a minimum. This is necessary to avoid chlorate formation in the anolyte and reduced stromal yield. When the lowest flow of salt solution in the diaphragm is used to avoid chlorate formation, only approx. half of the sodium chloride. The spent anolyte must then be evaporated to concentrate the caustic soda and to crystallize out the salt. The salt is finally filtered or centrifuged from the caustic solution. It is common practice to clean it during the process by adding saline solution to reduce its content of contaminants that clog the diaphragms. It is also common to renew the diaphragms at regular intervals.

Attempts have been made to construct diaphragm cells using permeation-selective membranes instead of the diaphragms. This has not contributed/to a resolution of the main problem. All membranes that have a reasonably high electrical conductivity and chemical resistance are also subject to considerable back diffusion and electromigration of hydroxyl ions at a rate of increases with the catholyte's concentration and temperature.

Another type of cell for the production of chlorine and caustic soda is the mercury solar cell, in which a salt solution is electrolysed and a sodium amalgam is formed at the cathode and chlorine at the anode. The amalgam is reacted with water to form a salt-free solution of caustic soda together with hydrogen. With this method, a diaphragm is not necessary because the solution of caustic soda is formed in a separate part of the apparatus or in a compartment that is separated from the compartment containing saline solution and chlorine.

In one embodiment of the mercury solar cell, purified sodium chloride solution is supplied to a slightly inclined, horizontal trough on the bottom of which the mercury cathode flows in the same direction as the salt solution. Above the mercury and inside the salt solution, horizontal anodes of graphite or of titanium coated with platinum or a platinum group metal are arranged. These anodes are suspended from gas-tight lids above the trough. Electric current is supplied to the anodes via rods which are suspended and sealed in holes in them

cover.

The salt solution is held in place in the trough by means of quick-sol siphons at the ends of the trough. The concentration in the cell is typically reduced from 315 to 275 g/l NaCl, and the salt solution is removed from the trough via an overflow. The salt solution is usually supplied to the trough via a valve and a rotameter. After it has left the cell, chlorine is removed from the salt solution by a combination of addition of HCl and removal under vacuum and in air, and the salt solution thus treated is resaturated, purified and returned to the cell at a pH of about. 7.0.

Electric current is led into the quicksilver via current rails which are connected to the bottom of the steel troughs. The anode rods which project above the lid and which must normally be vertically adjustable, are connected to the main busbar by means of "pigtails", clamps, soldered rods or similar fittings. In order to stop the operation of each trough, short-circuit switches are usually used which connect the anode current rail to the cathode current rail.

IN

The trough's lid, side walls, end boxes, seals and sometimes most of its bottom are covered with a corrosion-resistant material, usually hard rubber. The lifespan before repairs or replacement are necessary is rarely more than 5 years and often much lower. This assembly of constituents is usually referred to as the primary cell.

The mercury flowing out of the primary cell contains sodium plus impurities, such as calcium, magnesium and iron, which have not been completely removed by the previous purification of the salt solution. To remove the sodium and thus produce caustic soda, the quicksilver is washed with distilled water. This washing is carried out in contact with graphite, which causes the formation of a caustic solution, hydrogen gas and a relatively sodium-free mercury solution which is pumped back to the primary cell's intake box. The apparatus for this is usually referred to as the secondary cell or splitter ("denuder"). For the currently used practice, this is done by a horizontal trough with graphite grates or by a short tower with graphite as filling material. If the apparatus is made of a trough, this is arranged along the side of or below the primary cell, and if it is made of a tower, this is usually arranged on the primary cell's outlet with the pump under the tower and a long pipe line under the primary cell to return the mercury.

To understand how the mercury vapor cathode works in a mercury solution cell for the production of chlorine and caustic soda, it may be useful. to consider the electrolysis of a sodium chloride solution between a steel cathode and a graphite anode. Chlorine is released at the anode and hydrogen at the cathode, while caustic soda is simultaneously formed at the cathode in accordance with the following reaction equation:

If the anolyte and catholyte are not kept separate, the following secondary reactions will occur:

It is clear that such a cell cannot be used for the production of caustic soda and chlorine. On the other hand, in a quicksilver cell in which a quicksilver cathode is used, hydrogen will not be emitted preferentially at the cathode in relation to the discharge of sodium if the quicksilver is relatively pure, and the quicksilver will be transferred to sodium amalgam in accordance with the following reaction equation:

because the hydrogen overvoltage on a mercury-sol surface is higher than the necessary voltage for the deposition of sodium on such a surface. Above the voltage is the electrode voltage that is higher than the theoretically necessary voltage for the discharge of a gas on the electrode's surface.

If the mercury contains more than a few tenths of a percent of sodium, or traces of magnesium, nickel, or similar metals with a low hydrogen overvoltage, hydrogen and caustic soda will be formed to a greater or lesser extent than sodium amalgam. When this occurs, the current yield will decrease, the graphite consumption will increase and the chlorine gas in the cell, or in the non-condensable gas after the main amount of chlorine has been condensed to liquid, will become explosive due to mixing with hydrogen.

The reverse effect is desired when producing caustic soda and hydrogen from amalgam. If pure sodium amalgam is placed in a beaker and water or a caustic solution is poured over it, little or no reaction will take place because hydrogen is not easily given off from a mercury-sol surface. If a piece of graphite is partially immersed in the mercury, it can be observed that hydrogen bubbles rise from the graphite very close to the surface of the mercury, the water or caustic solution becomes more alkaline, and the amalgam is deprived of its sodium content. Under working conditions, it is the secondary cell. thus a short-circuited battery in which the amalgam forms the anode and the graphite the cathode.

This process produces a pure, concentrated caustic solution with a concentration of usually approx. 50%, while the concentration in the diaphragm cell is approx. 11%. The circulation of the mercury, the exposure of the surface of the mercury to salt solution so that it is always depleted to a greater or lesser extent, and other problems with which this technology is beset, have always forced the construction of extensive, expensive and complicated equipment that requires large buildings and inevitably leads to contamination with mercury due to foam removal etc. Furthermore, mercury cells are very sensitive to contamination in the sodium chloride solution as these increase the depletion of the amalgam that takes place during the electrolysis and lead to", a high and often explosive hydrogen content in the chlorine.

The problems associated with circulating the mercury can be avoided if the equipment is designed in such a way that the mercury is used as a membrane, one surface of which forms the cathode in the salt solution that is electrolysed, and the other surface forms the anode that is in contact with the caustic soda solution . In this way, deposition of sodium in the amalgam occurs at the same time as the amalgam is depleted on the opposite surface.

Attempts to achieve this inevitably lead to a construction in which the mercury is supported in siphon-like overhead channels or on porous or woven materials, essentially on a diaphragm. When the mercury is supported in siphons, the long path that the metallic sodium has to travel through the mercury will lead to an overconcentration of sodium at the cathode surface followed by a release of hydrogen on the cathode surface and into the chlorine, and if the mercury is supported by a diaphragm, the final the result has always been that the electrical resistance increases due to

that gas bubbles penetrate into the pores. In addition, a deposit of metallic contaminants, such as iron, in the pores causes these contaminants to be wetted by mercury, and this leads to mercury leaking through the diaphragm and being lost.

Another embodiment of the mercury cathode cell is described in US Patent No. 2749301. The mercury cathode is carried on a porous diaphragm of woven plastic or asbestos cloth. The salt solution flows under the cathode on the anode surface. A very strong and thus uneconomical stream of salt solution must be pumped through the space between the anode and the diaphragm to avoid the diaphragm being covered with a blanket of gas bubbles. However, even when a strong current is maintained, bubbles of chlorine and also hydrogen from the mercury layer above the diaphragm will slowly be taken up in the diaphragm and reduce the yield of the process.

From Norwegian patent application no. 4781/72, a method and apparatus are known for the electrolysis of aqueous solutions, where a permeation-selective diaphragm is used between the anode and the cathode, which is essentially impermeable to liquids and gases and which essentially consists of a hydrolyzed copolymer of tetrafluoroethylene and a sulfonated perfluorovinyl ether of the formula

FS02CF2CF2OCF (CF3)CF2OCF=CF2,

the copolymer having an equivalent weight of 900-1600. In the patent application made available, the electrolysis of a saturated solution of sodium chloride is described more specifically, where chlorine is discharged at the anode and removed as gas from the anode compartment.

The invention aims to provide a new method and a new apparatus for the electrolysis of alkali metal solutions. The invention also aims to overcome certain disadvantages with which the currently used electrolysis cells and electrolysis processes are affected.

The invention thus relates to a method and an electrolysis cell for the electrolysis of aqueous solutions containing alkali metal ions, such as e.g. by electrolysis of sodium chloride solutions to produce chlorine, sodium hydroxide and hydrogen. The alkali metal ions will be present in solution together with anions of the mineral acids and/or hydroxyl ions and/or anions of organic acids. Examples of such aqueous solutions are solutions of the chlorides, bromides, sulphates, sulphites, phosphates, acetates and hydroxides of sodium and potassium.

The invention thus relates to a method by electrolysis of aqueous solutions, preferably sodium chloride solutions for the production of chlorine and caustic soda, containing dissolved sodium and/or potassium ions together with dissolved anions of mineral acids and/or organic acids and/or hydroxyl ions, where an electric current is conducted through the solution held between an anode and a membrane, and the method is characterized by the fact that the electrolysis is carried out using a composite membrane consisting of a polymer membrane facing the anode, and of a metal layer through which alkali metal can penetrate and which is in intimate contact with the polymer membrane on the surface thereof facing away from the anode, at superatmospheric pressure under such conditions where the anode product is obtained in a substantially liquid condensed state or in a state where it is dissolved in the solution.

The invention also relates to an electrolysis cell with an anode and which is characterized by the cathode having the form of a composite membrane consisting of a polymer membrane and a cathodic metal layer through which alkali metal ions can penetrate and which is in intimate contact with the membrane.

The term "membrane" is used herein for a material which can be easily penetrated by ions and which the anolyte can penetrate with relative difficulty. According to the invention, it is avoided that a gas phase occurs in the electrolyte near the composite membrane, and this is achieved by using above-atmospheric pressure under such conditions where normally produced gaseous products will condense into liquid or dissolve in the electrolyte. The anode can be kept covered with a diaphragm that can be penetrated by an electrolyte, or with a membrane that can be penetrated by anions and the anode products are removed through the anode. Electrolyte compartments and secondary cell compartments can be kept separate by means of composite membranes and stacked one above the other so as to provide a distributed series flow of electrical current from cell to cell, while fluid flows uniformly to and from these cell compartments in electrically non-conductive channels, tubes or pipe lines.

It is a distinctive feature of the composite membrane that when it is used for the electrolysis of an aqueous alkali metal salt solution, alkali metal ions will migrate from the aqueous solution through the membrane into the metal layer and then out through the metal layer. In order to achieve the greatest advantages of the present invention, it is clear that the electrical resistance to electromigration of alkali metal ions in the membrane should be low and that the transport capacity of the metal layer for alkali metal should be high so that a maximum capacity per unity of the composite membrane and the lowest possible need for electrical energy. Furthermore, the composite membrane is constructed in such a way that it enables a direct interface contact between the polymeric membrane and the metal layer so that the alkali metal ions can be discharged directly from the polymeric membrane into the metal layer and be conducted through this in the form of alkali metal atoms, or as alkali metal ions together with free electrons.

A preferred metal layer consists of liquid mercury

which form liquid amalgams with the alkali metals. This special feature is of course well known in the field of technology where mercury cells are widely used for the electrolytic production of chlorine and caustic soda. In the following description, mercury will be used to describe the composite membrane's metal layer.

The interface between the polymeric membrane and the mercury layer, which together constitute the composite membrane for the electrolysis cell according to the invention, can be increased beyond the area obtained with a flat interface, by corrugating the membrane or by forming bulges in it that protrude into the mercury layer. It is also advantageous to increase this interface by depositing mercury in the polymeric membrane, e.g. by means of a prior electrolysis of a mercury-salt solution or otherwise depositing mercury in the membrane. It has also been shown that the resistance can be reduced by applying a swelling agent, usually a polar solvent such as ethanol or glycol, to the polymeric membrane alone or together with the mercury impregnating agent.

The polymeric part of the composite membrane should have low electrical resistance and high chemical resistance to chlorine and salt solution under the working conditions used.

The polymeric part of the composite membrane can be constituted

of a solid perfluorocarbon polymer with attached sulfonic acid

or sulfonate groups or sulfonic acid and sulfonate groups. (The term "sulfone groups" is used to generally denote sulfonic acid and/or sulfonate groups). This perfluorocarbon polymer has the pendant groups attached directly to the main chain of the polymer or to perfluorocarbon side chains which are attached to the main chain of the polymer. One or both of the polymer's main chains and any side chain may contain oxygen atom bonds (ie, ether bonds). The perfluorocarbon polymer from which the polymeric part of the composite membrane is made includes perfluorocarbon polymers with the attached groups and also perfluorocarbon polymers with a mixture of chlorine and fluorine substituents, the total number of chlorine atoms not exceeding approx. 25% of the total number of chlorine and fluorine atoms, with the attached groups. The polymeric part can optionally be reinforced, e.g. by using

a wire mesh of a suitable metal or a polytetrafluoroethylene cloth or another reinforcing material, as described in US patent no. 3770567. The perfluorocarbon polymers used for the membrane part of the composite membrane can be produced as described in US patent documents no. 3041317, no. 3282875 and no. 3624053.

The preferred perfluorocarbon polymers are prepared by copolymerization of a vinyl ether of the formula FS02CF2CF2OCF (CF2)CF2OCF=CF2 and tetrafluoroethylene, followed by a conversion of the SO-jF group to

-SO^H or sulfonate (eg, alkali metal sulfonates) or both.

The equivalent weight of the preferred copolymers is 950-1350, the equivalent weight being defined as the average molecular weight per sulfonyl group. The preferred thickness of the membrane part is 0.025-0.254 mm.

The cells according to the invention thus comprise an anode and a composite membrane with a polymer part which can be made of perfluorocarbon polymers with attached sulphonic acid and/or sulphonate groups, and a cathodic layer of a metal which can be penetrated by alkali metal ions and which is in intimate contact with the polymeric part.

Although the metal layer in the composite membrane has been described as consisting of mercury, other metals can be used instead. The type of metal chosen depends on the cation in the electrolyte, how easily the metal layer can be penetrated by the cation, and the interaction between the metal layer and the cation. Thus, e.g. thin films of silver and/or lead or combinations of these metals together with mercury suitable for this purpose. As the electrolysis cell according to the invention can be used under high pressure and at elevated temperature, under normal conditions solid metals and alloys can be used in a molten state. A very thin layer of a solid metal with sodium diffusion properties makes it possible to use a solid metal component in the composite membrane. This technique will also be able to be used in connection with a caking metal. These and other designs would facilitate an arrangement where the composite membrane is used in a position other than horizontal.

The anodes in the electrolysis cell can be made of any material suitable for the intended electrolysis process, such as the platinum group metals and their oxides alone or in the form of a coating of titanium or tantalum. The anode can be of any shape, such as plates, expanded or perforated metal, small sections of such shapes, or any other shape that will not lead to damming or entrapment of the anode product.

The short-circuited electrodes used in the secondary cell consist of

of graphite or a similar material with a relatively low hydrogen overvoltage. These electrodes are distributed in a pattern over the surface of the mercury in the secondary cell compartments and are partially submerged in the mercury. Bubbles of hydrogen gas are emitted on the graphite surface, while hydroxyl ions are formed as sodium ions pass into the aqueous solution from the mercury layer.

The short-circuited graphite electrodes can also serve as conductors

of cathode strbm to the mercury, but it is preferred to use metallic conductors from the anode in the closest overlying cell. It has been shown that graphite cloth, which is twisted between and around these metallic conductors, effectively causes the amalgam to rapidly and thoroughly deplete, while at the same time enabling a tight

arrangement of metal eaere-..

The invention will be described in more detail in connection with the drawings. Of these shows

Fig. 1 a section through a single element of a cell with

a composite membrane, a plate-like anode and a graphite cloth in the secondary cell and intended for the production of chlorine under pressure in the form of a dissolved gas. Fig. 2 shows a partial section through a single element of a cell, with a composite membrane, an anode (shown alternatively as a through-hole anode and a button anode) and a graphite cloth in the secondary cell and intended for the production of liquid chlorine under pressure.

Fig. 3 shows a partial section through a single element of a

cell with a composite membrane and an anode compartment supplied with avlbp with a wire mesh anode, and a graphite cloth in the secondary cell and intended for the electrolysis of sodium sulphate.

Fig. h shows a section through a stack of cell elements i

a pressurized enclosure and intended for the production of chlorine under pressure.

Fig. 5 shows a flow diagram for the salt dissolution and chlorine system in a stack of cells with composite membranes and intended for the production of liquid chlorine under pressure. Fig. 6 shows a flow chart for the salt solution and chlorine system and a stack of cells with composite membranes and designed for the production of chlorine dissolved in salt solution under pressure. Fig. 7 shows a flow chart for the systems of water, caustic solution and hydrogen for a stack of cells in which the caustic solution is cooled by means of recirculation. Fig. 1, 2, 3 and h apply to a stack of cells in which the individual parts occupy an essentially annular space, the electrolyte and water being supplied from the outside of the annular space, while the cylindrical core at the center of the annular space is used to insulate lines for used electrolyte. Other geometric designs, such as rectangular parts, can also be used. For each of Figs 1, 2 and 3, reference in the following description is given to the upper part of the cell which is identical to the upper part of the nearest underlying cell, as shown in the drawings. According to Fig 1, 1 denotes the outer anodizing, 2 the outer casing for the secondary cell, 3 the water supply line, h the anolyte, 5 the outer flow channel for the anolyte, 6 a regulation device for the anolyte, 7 the supply line for the anolyte, 8 the anode, 9 the room for hydrogen, 10 the chamber for caustic solution, 11 the graphite cloth band, 12 a conductor for electric current, 13 the inner anodization, lh the inner casing for the secondary cell, 15 the outlet line for the caustic solution and the hydrogen, 16 the inner circulation channel for the anolyte, 17 the inner membrane ring , 18 the overflow line for the anolyte and 19 the outer membrane ring. The layers A and B together form the composite membrane, with A being the polymeric membrane layer and B the metal layer, i.e. the mercury layer.

The electrolyzer part is bounded by the anode 8, the outer anodizing 1, the inner anodizing 13, the outer casing 2 for the secondary cell, the inner casing 1^f for the secondary cell and the bottom of the next anode 8'. In the electrolysis cell parts, the composite membrane AB, the graphite band 11 and the conductors .12 for electric current are arranged.

The anode 8 has the shape of a plate and can be made of steel or nickel and have a thin titanium layer bonded with good electrical contact to the surface of the plate. The upper surface of this titanium layer is in turn coated with a metal from the platinum group or with an oxide of such a metal, such as e.g. ruthenium oxide. On the underside of the anode, the electrical conductors 12 are attached. These must again be attached in such a way that an intimate electrical contact is obtained. The managers 12 can e.g. is made of nickel wires that have been attached to the anode plate by means of electron beam welding. These conductors are not necessarily made of straight wires, but can be designed as hairpin loops or have another suitable design that will enable electric current to be conducted between the mercury and the nearest overlying anode and provide good electrical contact between them without a tendency to carpet formation in the space between the mercury surface or the interface between the mercury B and the membrane A. An expert will naturally be able to calculate the electrical conductivity of these parts. A very good distribution of the electric current in the mercury solution is necessary. Regardless of whether this is achieved by means of a large number of thin conductors or by means of a smaller number of thicker conductors provided with thinner distribution conductors in the mercury solution, such as a mesh cloth, this is of no importance as long as the above-mentioned basic requirements are satisfied.

The parts in contact with the anolyte and the chlorine, the anode rings 1 and 13 and the lower parts of the ring covers 2 and 1k for the secondary cell, the membrane rings 17 and 19 and the necessary gaskets, which for simplicity are not shown, must be made of fit . such material and so constructed that they will withstand the corrosive environment. Furthermore, the construction and bolting must be such that short circuits will not occur even if the protective material were to fail. It is therefore preferred that all these parts are made of non-conductive materials instead of coated metals. The fluorocarbons, high density polyolefins, some polyesters and,

for materials that only face the secondary cell, the epoxy compounds are suitable materials. The electric current conductors 12, the inner lip of the casing 2 and the outer lip of the casing lh must be metallic and can be moistened with amalgam, and are preferably made of iron or nickel. This is. desirable to prevent solution seeping around the mercury due to the annular enclosures for secondary

the cell and poor electrical contact for the conductors 12.

The graphite cloth band 11 serves as a very effective splitting means, but it would also have been suitable with graphite tubes arranged

around the conductors 12 or other shaped parts of the graphite placed between these electrical conductors. However, it is preferred to use the graphite cloth. wrapped around the conductors 12 in such a way that flow channels will be formed to thereby facilitate the flow and mixing of the water and the caustic solution and thereby avoid stratification and poor depletion. The supply line 3 for water and caustic solution and the hydrogen supply line 15 are simply ordinary pipelines or pipes which are made of ordinary insulating materials or which are provided with suitable insulating joints.

The supply line 7 for salt solution is provided with a regulation device 6 to ensure that the correct amount of electrolyte reaches each electrolysis cell part. This regulating device 6 for the anolyte can simply be made of an opening with a suitable valve or of a regulating valve together with a discharge device. If a stack of these parts is supplied with materials from a common distribution line, it is preferable to regulate the current and,

if desired, to measure this by means of a flow regulating system calculated for the supply of a constant proportional amount of the total flow to a number of vertically arranged containers.

The current of the separated electrolysis solutions must be uniform

despite the different hydrostatic pressures that occur up-up through the stack of cells. This can be achieved with the help of different distribution devices such as separate supply vessels which are supplied with materials in accordance with a timetable and used together with flow regulating openings.

A preferred method of maintaining a regulated and substantially equal volume flow of brine into each of the brine streams is, despite the different hydraulic pressures, to deliver each stream vertically upward through a taper bore tube. A rotor or a float is suspended in the current kept in motion in this rudder. The force of gravity acting on the rotor (minus the buoyancy) is then counteracted by an equally large and oppositely directed force as by the current held in motion outwards towards the rotor. This force is independent of the flow rate and corresponds to the pressure difference multiplied by the largest cross-sectional area of the rotor. The pressure difference is therefore also independent of the flow rate.

The float is made up of an axially symmetrical body with a spherical shape or preferably with a center of gravity that lies substantially on the underside of the cross-section of the largest area. It is constructed in such a way that it is self-centering in the current kept in motion and can be described as plumb-shaped. Its vertical position in the tapered rudder will vary with the flow rate. When this position is measured on a linear scale as a measure of this flow, the device is known as a rotameter.

The parameters for this device have the following dependencies:

where AP represents the pressure difference, S^. the largest cross-sectional area of the float, vf the volume of the float, df the specific weight of the float, d the specific weight of the fluid, C an opening constant, g the gravitational constant, S the area of the annular space between the float and the rudder wall at the largest cross-section and q the volume velocity of the flow..

These equations suggest that the pressure difference ,AP, which occurs along the flow, can be regulated by choosing the float parameter, 'vf (df-d)/Sf, i.e. by choosing the appropriate design of the float and the apparent specific gravity, which together is referred to as the specific weight .

The invention is used to maintain an equal flow in the different flows as follows: The different flows can be denoted by the numbers 1, 2, 3,...n and the hydrostatic pressure by

■the signs p-^ p2, p^, pn, the indices indicating the current numbers. As

stated above, the float parameter for the float in each stream is adjusted so that the pressure differences AP-^, A<P>2, AP^, APn will satisfy the condition:

When the float parameters are regulated as indicated above, the current distribution will be kept equal regardless of the total flow of saline solution supplied to the system.

In the above description of this part of the invention, the flow of salt solution is equalized in the different streams, and the current in the different rudders is directed upwards against the downward force on a float that is heavier than the fluid. However, it is possible to utilize this feature according to the invention to maintain the proportionality between the amount of flow in streams that are not equal, by using suitably designed and weight-matched floats. For regulating the flow ratio between the different current

more, a device such as a valve in each stream can be used together with the regulating float, whereby the flow ratio between the different streams (i.e. as regulated by means of a valve) will be maintained regardless of the total amount of flow from the collection tank.

According to the invention, it is also aimed to use a float which

has a lower average specific gravity than the fluid. For this purpose, the fluid flow in the tapered rudder directs downwards,

as the rudder is extended downwards and the center of gravity of the float lies above the cross-section at the largest area.

This flow regulating system offers the further

share that the vertical position of the float is in relation to the annular area S and the flow volume q, assuming that all parameters on the right-hand side of equation (2) are kept constant. By varying the upstream pressure P-^ at the common distribution line, it is possible to maintain a proportional change in the flow rate of the liquid stream to each cell, while maintaining the hydrostatic pressure differences. It offers great advantages when producing chlorine and caustic soda proportionally to be able to regulate the flow of solution to the individual cells by means of a regulating device in a common distribution line or manifold. This is because it is of significant importance that production can be made flexible depending on varying demand without having to have large storage options. If desired, these current balancing devices, when connected to a position sensing device of the capacitor type or the inductive type, can be used as sensors for

information regarding the volume drum.

This flow regulating system is also of value for systems other than the chlorine cell described here. This type of flow regulation can be used in general in connection with any system where a series of supplies from a collection tank is used and where it is desirable to maintain equal or proportional volume flows between the series of supply streams despite different upstream pressures.

The anolyte overflow line 18 must be sufficiently high that the hydrostatic pressure due to its height will compensate for the weight of the mercury solution, of the membranes and of the caustic solution and provide the necessary force so that the membrane A will be kept firmly pressed against the bottom of the conductors 12. The membrane A is shown as a flat film, but it is preferred that it is provided with a surface that has been made larger by means of corrugations or bulges that protrude into the mercury so that the interface between the membrane and the mercury increases. This reduces the electrical resistance across the interface and the membrane itself, and the amount of mercury in the system. The hydrostatic pressure which is provided by means of the line 18 should preferably be at least 5.08 cm water column above the pressure which is necessary to achieve hydraulic balance. The anolyte lines must also be electrically insulated. Of course, some or all wires attached to the cell could be replaced by channels inside the cell.

In Fig. 2, 20 denotes the cell box, 21 an anode and 22 and 23 gaskets. All other parts of the cell described in connection with Fig. 2 serve the same purpose as the similar elements in Fig. 1. Together with the current conductors, the anode 21 fulfills the same function as the anode 8 and the current conductors 12 in Fig. 1, but it facilitates the drainage of liquid chlorine.

Two alternative designs of the anode are shown in Fig. 2. On the left side of the drawing, the top of the current conductors 21' is designed as large buttons. This leads to an anode which consists of a number of small pieces with channels between them for the drainage of the liquid chlorine. It is self-evident that more than one conductor can be connected to a single button in this way and that the conductor could be fed through the top of the cell box and that the button could be completely placed on the top of the cell box instead of being fast through the top of the cell box. This embodiment offers the advantage that it can be produced by means of forming techniques and metalworking on thread and screw turning machines and enables the use of sproe anode materials which cannot be easily produced so that they will cover a large surface in a thin section. In this way, a choice of materials is possible, and the current conductor and the button can be made of the same material or of two completely different materials.

The one on the right side of the drawing showed anode design with

the anode as a perforated plate 21'' can be produced using other manufacturing techniques. Perforated plates, sieves or expanded metal can be electrically connected to current conductors by feeding each current conductor through the top of the cell box or by connecting several current conductors to each other in the secondary cell compartment and only a single connection part from such a bundle of current conductors is brought to the anode plate. If the anode plate is arranged slightly above the top of the cell box, a space for liquid chlorine is obtained.

The top surface of the buttons or anode plate must be so designed

that liquid chlorine will not form a blanket, but will run down the top of the cell box. Should this be desirable in order to increase corrosion resistance, the top of the cell box can be protected with

a fluorocarbon layer held in place by collars on the current conductors or buttons.

By means of the seals 22 and 23, the polymeric membrane of the cell boxes and neighboring cell boxes is sealed to each other. When perfluoro-sulfonic acid polymers are used as a membrane, it is possible to seal this material to other fluorocarbons or to let the edge of the membrane retain the sulfonyl fluoride form and build up packing designs on

this way0 This material or the sulfonyl fluoride form of it can be used in a similar way as such or attached to fluorocarbon linings for other cell parts, thereby enabling a number of different attachment devices and good corrosion protection.

The ring line construction according to Fig. 2 is somewhat different

from the ring line construction according to Fig. 1. This is because the cell according to Fig. 1 is intended for the production of chlorine that is dissolved under pressure, while the cell according to Fig. 2 is intended for the production of liquid chlorine which must be able to flow freely through holes or channels in the anode along the top of the cell box and into the interior of the ring line.

In Fig. 3, 2h denotes a diaphragm, 25 an anode sieve, 26 an anode trough and 27 an outlet for anolyte and anode gas.

The cell according to Fig. 3 is essentially the same as the cell according to Fig. 1, but with a fundamentally different anode construction. If e.g. a sodium sulfate solution is electrolysed to produce caustic soda, sulfuric acid, hydrogen and oxygen, it is clear that the anode gas, i.e. the oxygen, is neither soluble nor condensable to liquid under any of the practical conditions that exist in a cell. In order to carry out such an electrolysis, the disadvantages associated with gas bubbles in the anolyte must therefore be accepted, but this requires a large volume of flowing anolyte in order for the gas bubbles to be carried out of the cell and still leads to a higher cell resistance, or it can a cell is used with a composite membrane at the cathode and an anode intended for drainage which is protected by a diaphragm, so that the electrolyte is kept essentially free of gas bubbles. The electrolyte can percolate completely through the anode diaphragm or be partially circulated through the cell, while part of the electrolyte is allowed to flow through the diaphragm. This can be regulated by choosing a diaphragm with the right resistance to flow and by regulating the pressure difference between the main flow of electrolyte and the gas phase in the anode trough in a well-known way.

The diaphragm can be replaced with a membrane which can be penetrated by anions or with a membrane through which anions can selectively penetrate, so that the anolyte during electrolysis of a sodium sulphate solution will mostly contain sulfuric acid and a minimum of sodium sulphate which would have cleaned off at the anode, instead of a mixture of sulfuric acid and sodium sulfate. The circulating stream of sodium sulphate can then simply be saturated again without it being necessary to carry out external crystallization to separate the sodium sulphate from the sulfuric acid. As one of the main uses for the electrolysis of sodium sulphate is the recovery of spinning bath for rayon, such crystallization can be carried out by circulating the reading from the cell through the spinning bath to be recovered.

A cell of this type can also be used for the electrolysis of sodium chloride by choosing a suitable diaphragm and anode materials, and the cell can then be operated close to atmospheric pressure.

Although an anode sieve is shown in Fig. 3, other anodes which enable runoff or draining anodes can be used, and this enables the use of such materials as lead/sol alloys, magnetite and other materials.

In Fig. h, 1 represents a pressure jacket, 2 a distribution device for caustic solution or water, 3 a supply device for caustic solution or water, h a flexible cathode rail, 5 a supply pipe for caustic solution or water, 6 an insulating pressure fluid, 7 a cathode end cap, 8 a combined supply pipe for anolyte, 9 a regulating device for anolyte, 10

an interrupter core for anolyte, 11 a cell element, 12 an overflow line for anolyte, 13 an insulating ring, 1^ a supply device for anolyte, 15 an anode end cap, 16 an outlet for caustic solution and hydrogen, 17 an outlet for anolyte and chlorine , 18 a cathodic connection rail from the rectifier and 19 an anodic connection

shine from the rectifier.

The drawing schematically shows a section through a stack of cell elements for producing chlorine in liquid form or as a dissolved gas. The bottom of the stack is made up of the end anode, which is also connected to most or all of the rudder lines, as the supply device 3 for water or caustic solution can also be supported at the bottom of the stack. This bottom of the stack is also connected to the positive side of the rectifier. Above the stack is the pressure mantle 1 which is pressure-tightly attached to the bottom, and the space between the pressure mantle and the cell stack is filled with an insulating pressure fluid 6. The cathode cap - 7 is placed on top of the cell stack. This end cap is via a flexible current rail connected to the mantle 1 which serves as the vertical current rail which conducts the negative current from the rectifier which is connected at its bottom to the top of the cell stack. The distribution device 2 for the water or the caustic solution and which supplies these to the secondary part of each element is also arranged on top of the cell stack. The liquid volume in each secondary part is relatively large so that a discontinuous supply can be accepted, provided that the supplied average volume is accurate. It is therefore suitable to measure the total volume of liquid supplied from the supply device 3 to the stack in connection with an accurate distribution by means of such a device as a rotary valve or small cooperating pressure pumps

etc. The distribution device must be arranged at the top of the stack so that the individual supply pipes can be emptied into the secondary cells to thereby prevent current leakage through the set of supply pipes 5.

If the insulating fluid 6 provides perfect insulation, the pressure jacket, with the exception of the insulating ring 13, can consist of pure metal. However, the danger of leakage from the cell stack is always present and a severe short circuit can occur despite the pressure balance controls between the cell stack and the fluid 6. It is therefore preferred that the inner surface of the pressure jacket 11 is lined with a suitable insulating material which can consist of rubber or any other plastic material that is compatible with the fluid 6 and the working temperature. In addition, the area at the bottom of the mantle and at the top of the anode end cap 15 must be leak-proof combined with the ring 13 as the total voltage difference across the stack exists at this point.

The cell construction described above assumes that the main output from the rectifier or another direct current source is located at or near the ground level. If this is not the case, the insulating joint in the jacket can be located at any point along the height of the jacket, and the current connections will then have to be placed on each side of the insulating joint.

In Fig. 5, 1 denotes a cooling or heat recovery unit,

2 a cell stack, 3 a circulation pump for anolyte, k- a cooling device for used anolyte, 5 a device for saturating the anolyte again, 6 a decanting device for chlorine, 7 a dechlorination system, 8 a drying system for chlorine, 9 a slurry tank, 10 a slurry pump , 11 a supply device for salt, 12 the chlorine product, 13 a branched anolyte stream and 1^- a purified, branched anolyte stream for return.

Fig. 5 shows a flowchart for the salt dissolution and chlorine systems that surround one or more cell stacks in which liquid chlorine is produced directly in the cells. A person skilled in the art will easily understand how the process indicated on the flowchart is to be carried out. It briefly comprises the following process steps. Used anolyte is led from cell 2 into the heat exchanger h in which the temperature of the anolyte is lowered. The cooled anolyte is then transferred to the apparatus 5 to re-saturate the anolyte by increasing its salt content. The anolyte then flows into the decanting device 6 for chlorine from which the majority of the anolyte via the pump, 3 is recycled back to the cell 2. Part of the anolyte from the decanting device 6 for chlorine flows into the dechlorination device 7 from where it is removed as a stream 13. The purified anolyte stream lh and salt 11 is supplied to a slurry tank 9 and transferred by means of the slurry pump 10 to the apparatus 5 for saturating the anolyte. The chlorine removed from the decanting device 6 and the dechlorinating device 7 is led through the chlorine drying device 8 and removed as a stream 12. The heat transfer fluid from the cell 2 passes through the heat exchanger 1 and the cooling device h for used anolyte, from where it is fed to

bake to the cell stack 2.

The theoretical splitting voltage for sodium chloride is approx. 2.3, and a cell stack is operated on a technical scale at a voltage per element of approx. 2 volts above this theoretical voltage. This over-voltage is converted into heat and corresponds to an effect of approx. 60 kW per tonnes of chlorine produced per day. Most of this heat will manifest itself in a temperature rise in the circulating electrolyte and in the produced caustic solution, and part of the heat will lead to a temperature rise in the insulating fluid 6 according to

Fig. i+. This heat must be removed. This can be achieved simply by heat exchange and by the heat being released to air or water. If, however, there is a sufficient temperature difference between the working temperature and the heat source, a significant amount of the electrical energy beyond the theoretically necessary can be recovered for power production, process heat or desalination of water etc. For power production it is advantageous to use such a fluid as "Freon" which has a boiling point and pressure properties that are close to the conditions in the cell stack. Thereby, the cooling devices and the cell stack are converted into a boiler. The boiling, insulating fluid can be used to operate a turbine, after which it can be condensed and returned to the cooling devices h. Otherwise, distilled water is also completely satisfactory as an insulating fluid.

Liquid chlorine has an exceptionally high thermal expansion coefficient.

At room temperature, liquid chlorine is far heavier than the anolyte,

but at higher temperatures the specific gravity of the chlorine approaches or becomes less than the specific gravity of the anolyte.

The chlorine must therefore be separated from the anolyte by decanting below the anolyte or removed at the top of the anolyte. The temperature conditions in the decanting apparatus must, if necessary by means of regulation, be such that a sufficient weight difference is ensured for the separation to take place. Moreover, chlorine's special properties suggest that the cell design and its operation must be such that a chlorine blanket is avoided over the membrane when the chlorine is lighter than the electrolyte, and so that chlorine is prevented from flooding the anode when the chlorine is heavier than the electrolyte.

The dechlorination system 7 shown in Fig. 5 enables branching

of a relatively small portion of the recirculating anolyte for two reasons, i.e. to keep the amount of contaminants at an acceptable level and to slurry the salt fed to the system.

The circulation speed for the anolyte can be approx. 18.9-37.9 liters per minute per tonnes of chlorine produced per day, while the branched stream 13 can be approx. 1.1-1.9 liters per minute. The dechlorination under pressure is achieved by heating the branched stream using a suitable heat exchange device for extracting heat. If it is necessary to use a small amount of gaseous chlorine, such as for the production of hypochlorite or hydrochloric acid, the dechlorination can be accomplished simply by suddenly depressurizing the branch stream to atmospheric pressure and then carrying out the dechlorination by means of an air blast or under vacuum in a regular way.

In Fig. 6, 1 denotes a cooling device or a heat recovery unit, 2 a cell stack, 3 a cooling device for anolyte, h a device for re-saturating the anolyte, 5 an expansion motor, 6 a flash separator for chlorine, 7 a circulation pump for anolyte, 8 a dechlorination system, 9 a condenser for chlorine, 10 a drying system for chlorine, 11 a slurry pump, 12 a slurry tank, 13 salt supply, 1<*>+ a purified, branched anolyte stream for return, ;15 the chlorine product and 16 a branched anolyte stream. The process shown by means of the flowchart in Fig. 6 is carried out in summary as follows. The used anolyte flows into the apparatus h for renewed saturation of the anolyte and then into the expansion motor 5 which is driven by the anolyte as it flows to the flash separator 6 for chlorine. The main part of the anolyte flows from the separator 6 into the anolyte recirculation pump 7 which pumps the anolyte into the anolyte dresser 3 and back into the cell stack 2. A smaller part of the separated anolyte flows into the dechlorination system 8, from where it is branched off from the system as a stream 16 The chlorine from the flash separator 6 and from the dechlorination system 8 is led into a condenser 9 for chlorine and a drying system 10 for chlorine and is then removed as a product stream 15. Purified anolyte 1^ and salt 13 are fed to a slurry tank 12 and transferred by means of a immersion pump II to the apparatus h for renewed saturation of anolyte. An electrically insulating heat transfer fluid is pumped through the cell stack 2 and into a heat transfer unit 1, from where it is passed through the anolyte jacket apparatus and back into the cell stack. This flow chart is therefore similar to the flow chart shown in Fig. 5, but with the exception that it is calculated for a cell from which the entire amount of produced chlorine is removed as a dissolved gas. For the production of the chlorine product, the pressure of the used anolyte is lowered from the cell working pressure which leads to the emission of gaseous chlorine, relatively directly depending on the relationship between the working pressure and the pressure maintained in the flash separator 6 for chlorine. As the chlorine separated in the flash separator 6 is hot and moist and has a certain pressure, it can be condensed by means of cooling. It is preferred to carry out the flash separation at pressures above 7 kp/cm 2 in order to take advantage of the usually available cooling water. As very large quantities of anolyte must be pumped from flash pressure to working pressure and this entails a considerable consumption of energy, it is preferred to recover a large part of this energy by sudden expansion of the used anolyte and the produced chlorine in an expansion ;sion engine 5. ;The flow charts on Fig. 5 and 6 relate to the production of moist, ;liquid chlorine. If necessary, this can be washed free of adhering salts and then dried. In normal implementation of the process 1 technical scale, it is of essential importance that the chlorine is dried so that its moisture content will be in equilibrium with approx. 95% sulfuric acid at 15.6°C as the steel in the equipment in which the liquid chlorine is normally transported and stored will otherwise corrode. The salt supplied to the system is usually purified and produced on site. The added salt should preferably be relatively finely divided to ensure that it will dissolve quickly, as salt crystals can otherwise enter the cell and erode the membranes. In Fig. 7, 1 denotes a cell stack, 2 a cooling device for caustic dissolution possibly combined with a heat recovery system (not shown), 3 a separator for caustic dissolution, h a circulation pump for caustic dissolution, 5 a flash separator tank for caustic solution, 6 a cooling apparatus for hydrogen, 7 a system for the removal of quicksilver, 8 a vacuum degassing vessel for caustic solution, 9 a vacuum pump, 10 a hydrogen stream, 11 a recovered quicksilver stream, 12 a stream of caustic solution and 13 a water stream. ;Hydrogen and caustic solution are removed together from the cell stack and separated in apparatus 3. As hydrogen is only very slightly soluble in a caustic solution, the main amount of hydrogen will be separated in the form of a hot gas under pressure. After cooling this hot gas in the cooling device 6, the mercury content in the cold gas is lowered further, as this content is already very low due to the fact that the hydrogen is under pressure and the condensed mercury can be recovered. The very small amount of mercury vapor that is present in the cold hydrogen that is under pressure can be further removed in the system 7 using known techniques, such as washing with chlorine water or adsorption etc. As the system requires a completely enclosed apparatus in counter- addition to an ordinary mercury cell and because the mercury-sol content of the hydrogen is inversely proportional to the pressure at which the hydrogen is formed, the pollution that the mercury brings with it or that must be combated is reduced by a factor of 10 or more. The hydrogen that flows out at 10 will then no longer have ecological consequences because the mercury has been effectively removed. It should also be noted that the chlorine is produced without the presence of a non-condensable gas, and no unpleasant-smelling gas must therefore be released or treated. ;After the separation in the separator 3, the majority of the caustic solution is recycled to the apparatus 2 for cooling. This cooling system together with the cooling system shown in Fig. 5 and 6 can be operated in different ways depending on the desired results. It can be used simply for cooling the caustic solution to thus remove the heat content in the stream of caustic solution, or the heat can be removed from the caustic solution by keeping the stream of this at a higher temperature than the stream of salt solution, thereby improving the possibilities for heat recovery instead of a fairly simple cooling down. The added net amount of caustic solution in the system, i.e. the net production of caustic solution, can be led from the separator 3 "to the flash separator tank 5 where it will expand to near atmospheric pressure. Dissolved hydrogen; and furthermore small hydrogen bubbles will at this point be removed from the caustic solution. Depending on the design of the system, the hydrogen that is removed in the flash separator tank 5 can be free of quick _y----~s5fv'"~'and it does not then need to be subjected to any other treatment than to remove small droplets of caustic solution, or it may contain some mercury impurities, and it can then be treated in parallel with the main flow of hydrogen from the separator 3. Hydrogen tends to stay suspended as very small bubbles in a caustic solution when the hydrogen pressure is lifted, and it can therefore, it is desirable to quickly expand the remaining hydrogen bubbles under vacuum from the caustic solution in the degassing vessel 8C The small amount of hydrogen that e point will be pumped from the caustic solution, can be combined with the hydrogen stream from the flash separator tank 5 and treated as needed. The caustic solution from the degassing container 8 can be transferred directly for storage without fear that hydrogen explosions may occur in the storage tanks. The system can be used within a wide range of temperature and pressure which will lead to the production of liquid or dissolved chlorine in accordance with the physical properties of these products. Within the range of applicable conditions, the choice of a particular pressure and a particular temperature is determined by a balancing of the economic factors by means of well-known optimization techniques. As examples of temperature and pressure ranges for the production of liquid chlorine, an absolute pressure of cac 7.0-70.3 kp/cm 2 can be mentioned at a temperature of approx. 15.5-132.2°C or an absolute pressure of 21.0-^5.7 kp/cm^ at a temperature of 65.5-107.2°C„ ;Based on economic considerations, the process will generally be carried out at a temperature above the temperature of an available heat source. In normal industrial practice, this temperature will be 26.7°C. This means that a cell in which liquid chlorine is produced will be kept at a pressure above 7.0 kp/cm and that the flash pressure for a system in which chlorine is produced completely in the form of a dissolved gas is not lower than 7.0 kp/cm cm 2 so that the chlorine can be condensed into a liquid without having to resort to mechanical cooling"; The upper limits for the pressure and temperature are dictated again essentially based on economic considerations" The upper limit for liquid chlorine is of course its critical temperature. At very high pressures, the necessary equipment becomes very expensive, and the method will therefore not usually be carried out at a pressure above 70.3 kp/cm 2 c. The working temperatures that are possible for these pressure ranges can be determined on the basis of well-known data. The possibilities for recovering heat from the system are improved at high temperatures so that there is a temperature difference between the salt solution and the caustic solution on the one hand and the temperature in the heat source on the other. When producing liquid chlorine directly in the cell, considerable weight must be attached to the boiling point curve for liquid chlorine. When dissolved chlorine is produced in the cell, the solubility of chlorine will decrease with increasing salt concentration and temperature and increase with increasing pressure. The greater the pressure difference between the cell and the flash separator, the more chlorine will be formed per volume unit salt dissolution flow per review. The higher the salt concentration, the lower the chlorine production per review of salt solution become. The higher the cell working temperature, the less chlorine will be dissolved in the salt solution, but the more heat can be recovered from it. To determine working conditions for any particular case, standard optimization techniques together with known data regarding chlorine's boiling point and solubility can be used. ;All normal chlorine cells must be located in a building as they cannot in practice be operated outdoors. The present cells can be used outdoors in practically any climate, thereby avoiding costs for erecting buildings. With the construction that the cells according to the invention have, there is also no need for a busbar system from part to part or a busbar distribution system for each part, and this means that there are no copper or aluminum busbars around the cells. This not only saves capital costs, but also the main maintenance costs for the overall chlorine plant. The invention has been described above in connection with the electrolysis of sodium chloride and sodium sulfate for the production of caustic soda and hydrogen in each case. However, the invention is not limited to the electrolysis of these solutions. Solutions containing other ions, especially the corresponding potassium ion-containing solutions, can be electrolysed in a similar manner. The present invention has general applicability to compounds which can be electrolysed in ordinary mercury solar cathode cells. The present invention is also beneficial for carrying out other known electrolysis processes for solutions containing alkali metal ions. Obviously, anolytes (or contaminants or adversely affecting ions in the anolytes) or materials for the secondary part of the process and which are known to adversely affect the performance of the electrolysis process should be avoided. In addition, the secondary cell can be used, occasionally without graphite as a depleting agent, for the production of sodium sulphide using sodium polysulphide as a depleting agent, for the production of sodium hydrosulphite using a sulfur dioxide solution as a depleting agent, for the production of alcoholates, for various organic reductions, for dimerizations or similar reactions. ;By accurately setting the strb exchange conditions in order to thereby avoid oxidation of the metal component in the composite membrane, it is also possible | use this metal component as a bi-polar electrode and thus carry out a further reaction step between a final electrode in the secondary chamber and the anionic side of the now bipolar metal component in the composite membrane. As an example of such a reaction, the production of sodium can be mentioned. Moreover, the description of the mercury vapor layer as stationary does not exclude that the mercury solution can alternatively be circulated from the composite membrane. In the examples below, cells were used which contained a platinum-coated titanium anode arranged opposite a composite membrane, with an essentially saturated sodium chloride solution flowing between the anode and the composite membrane. The produced chlorine was removed from the cell together with the salt dissolution stream. The composite membrane was a membrane of perfluorosulfonic acid which is marketed under the trademark "Nafion" and which was prepared from a copolymer of tetrafluoroethylene and a vinyl ether of the formula FSC^CT^CF20CF(CF2)CF20CF=CF2 by means of the usual technique for the preparation of thermoplastics, followed by a conversion of the attached sulfonyl fluoride group to the corresponding acid from the obtained copolymer with an equivalent weight of 950-1350, and with a mercury film over the membrane. The amount of mercury was sufficient to cover the polymeric membrane completely, care being taken so that any corrugations or bulges in the polymeric membrane were also covered. The mercury for the composite membrane was in contact with graphite elements, and water was passed over the mercury to produce caustic soda and hydrogen over the composite membrane. In all examples, the surface of the membrane extended into the mercury as a result of swelling of the membrane due to a solvent or electrolyte and because the pressure against the salt solution was greater than the weight of the mercury and the caustic solution above the membrane. The gross cell surface was approx. 1 dc p (approx. 5.1 x 20.3 cm). The flow of water through the depletion cell or secondary cell was varied so that different concentrations of caustic solution containing from less than 10% to more than 50% caustic soda were produced, and no effect on the operation of the cell could be observed. ;Example 1 ;A cell was operated at atmospheric pressure and with a polymeric membrane having a nominal thickness of 0.051 mm. The flow rate of the salt solution was approx. 15.9 liters per minute. The cell voltage was ^.9 V at 50 A and 6.6 V at 80 A. The flow rate of the salt solution was increased to 2^.2 liters per minute to improve the removal of gas from the cell, and the voltage dropped to 6.0 V at 80 A. The temperature was 87.8-96.1°C. Eczema! 2 ;The same cell was used with a membrane which had first been immersed in hydrochloric acid and then in a saturated aqueous solution of mercuric chloride for 2 hours at 100°C. The mercury in the membrane was then reduced to metallic mercury in situ by means of hydroxylamine, and the cell was operated under the same conditions as before using a brine flow rate of 2^.2 liters per second. minute. The cell voltage was 5.6 V at 80 A and <>>+.6 V at 50 A. ;in Example 3 ;The same cell was used under the same conditions, but the membrane had been swollen with glycol before it was impregnated with quicksilver as described in Example 2. The voltage across the cell was 0.2-0.3 V lower at 80A than the voltage without the use of a glycol-treated membrane. ;Example h ;A cell with a polymer membrane with a nominal thickness of 0.089 mm was operated at a temperature of 23.9°C and a pressure of 31.6 kp/cm 2 and with a flow rate for the salt solution of 50 cm0^ per . minute. A groove had been cut out in the anode so that the liquid chlorine would run off from it. The cell parts above the mercury consisted of nickel pins with a diameter of 1.59 mm on 6.35 mm centers and with a graphite cloth around and between the pins. The graphite and nickel were in contact with the mercury. At approx. 10 A and a cell voltage of 3.6<*>+ V could initially be observed with chlorine dissolved in salt solution through an observation glass at the cell outlet. As the amperage was increased to 50 A, the cell voltage rose and then fell as liquid chlorine was formed, and the cell voltage stabilized at 5.1 V. The liquid chlorine could be observed as a separate phase through the observation glass together with salt solution containing dissolved chlorine. Liquid chlorine came into view as the current strength was increased because the chlorine production then exceeded the chlorine's solubility in the salt solution. At a higher temperature and with a thinner polymer membrane, a lower cell voltage could be obtained. The observation glass was arranged in such a way that the salt solution and liquid chlorine entered through a submerged rudder at the top. The observation glass was also provided with an overflow over the bottom edge of the submerged tube. At these working temperatures, this arrangement led to the liquid chlorine being deposited on the bottom of the observation glass and the salt solution overflowing from the top of the observation glass. This therefore acted as a decanting device. From the overflow, the salt solution containing dissolved chlorine flowed to a receiving container. Then the pressure in the receiver vessel was reduced, and the dissolved chlorine bubbled out of the salt solution. Hydrogen and caustic solution flowed from the cell into a container in which the hydrogen and the caustic solution were separated. The entire system was initially filled with nitrogen, and the hydrogen was then released into the atmosphere together with the filled nitrogen.

Claims (11)

Process for the electrolysis of aqueous solutions, preferably sodium chloride solutions for the production of chlorine and caustic soda, containing dissolved sodium and/or potassium ions together with dissolved anions of mineral acids and/or organic acids and/or hydroxyl ions, where an electric current is passed through the solution held between an anode and a membrane, characterized in that the electrolysis is carried out using a composite membrane consisting of a polymer membrane facing the anode, and of a metal layer through which alkali metal can penetrate and which is in intimate contact with the polymer membrane on that surface of this facing away from the anode, at superatmospheric pressure under such conditions where the anode product is obtained in a substantially liquid condensed state or in a state where it is dissolved in the solution. 2. Method according to claim 1, characterized in that the electrolysis is carried out at a pressure of 7-70 kg/cm <2> and a temperature of 15-132°C. 3. Method according to claim 1 or 2, characterized in that a composite membrane is used, the metal layer of which consists of mercury. 4. Method according to claims 1-3, characterized in that a polymer membrane is used which consists of a per- fluorocarbon polymer with attached sulfonic acid and/or sulfonate groups. 5. Electrolysis cell for carrying out the method according to claims 1-4, characterized in that the cathode has the form of a composite membrane (AB) which consists of a polymer membrane (A) and of a cathodic metal layer (B) through which alkali metal ions can penetrate and which is themselves in intimate contact with the membrane. 6. Electrolysis cell according to claim 5, characterized in that the polymer membrane (A) consists of a perfluorocarbon polymer with attached sulfonic acid and/or sulfonate groups. 7. Electrolysis cell according to claim 5 or 6, characterized in that the metal (B) is mercury. 8. Electrolysis cell according to claims 5-7, characterized in that metallic mercury is deposited in the polymer membrane (A). 9. Electrolysis cell according to claims 5-8, characterized in that the polymer has been treated with a swelling agent. 10. Electrolysis cell according to claims 5-9, characterized in that a number of electrical cathode lines (11) are in contact with the metal (B). 11. Electrolysis cell according to claims 5-10, characterized in that the polymer membrane (A) consists of a copolymer of FS02CF2CF2OCF (CF3)CF2OCF=!CF2 with F2C=CF2 which has been treated to convert the -SO2F groups into sulfonic acid and/or sulfonate groups.