GB2035377A - Electrolytic production of alkali metal carbonates - Google Patents

Abstract

Alkali metal chlorides are electrolyzed to the corresponding carbonates by employing a cell cathode which is continugous to a permselective cation exchange membrane which divides the anolyte compartment from the catholyte compartment, while introducing carbon dioxide into the catholyte, either in the cell or while it is recirculating outside the cell. Suitable membranes are those containing hydropholic ion exchange groups such as sulphonic groups, carboxylic groups or sulphonamide groups.

Description

SPECIFICATION Electrolytic production of alkali metal carbonates bescription This invention relates generally to a process for electrolytically producing an alkali metal carbonate. More particularly, it relates to an improved method for electrolytically producing an alkali metal carbonate employing a membrane cell with a particular membrane-cathode configuration.

It is known that alkali metal carbonates can be produced electrolytically from alkali metal chlorides in diaphragm and membrane cells, by introducing carbon dioxide into the catholyte compartment or into recirculating catholyte outside the cell, as shown, for example, in U.S. Patent Specifications 552,895, 2,967,807,3,179,579,3,374,164 and 4,080,270. In all but U.S. Patent Specification 552,895, the diaphragm or membrane separating the electrolytic cell into anode and cathode compartments is spaced from the cathode, so as to permit the C02 gas or HCOB- ions (formed by the reaction of CO2 with OH- ions) introduced to reach more readily and react with the OH ions generated at the membrane side of the cathode.This minimizes back-migration of the OH ions through the separator into the anode compartment, which is the primary cause of poor current efficiency in the electrolysis process. Typically, carbonate salt-production processes having the cathode separated from the membrane achieve current efficiencies (measured on catholyte products) of at least 95% and often approach theoretical. Further, U.S. Patent Specification 3,374,164 shows that the diaphragm cell process is significantly improved when the diaphragm is separated from the cathode; the quantity of Na+ ions flowing through the diaphragm and being converted to Na2CO3 increases to 80% as compared with only 60% when the two are contiguous, as shown in U.S.

Patent Specification 552,895.

It is known that the electrolyzing voltage is reduced by laminating or juxtaposing the membrane to the cathode as shown, for example, in U.S. Patent Specifications 2,967,807,3,057,794 and 4,101,395 and German published Patent Specifications 2,704,213 and 2,741,956. However, these expedients have the drawback that the OH ions generated at or near the membrane-cathode interface can more readily escape into the membrane and so migrate into the anolyte compartment, causing a loss of current efficiency as previously described. Because of this, and the nearly theoretical efficiencies achieved when the cathode is separated from the membrane, the separated cathode-membrane configuration has been employed by the prior art for producing alkali metal carbonates in membrane cells.

Having regard to the state of the art, this invention provides a membrane-cell method for the production of alkali metal carbonates, employing a low electrolyzing voltage and yielding a high current efficiency, so that the method is more efficient than those now known.

According to this invention, a method is provided for producing an alkali metal carbonate, which comprises electrolyzing an alkali metal chloride in an electrolytic cell having an anode and a cathode in anolyte and catholyte compartments which are separated by a permselective cation-exchange membrane impervious to hydraulic flow, the cathode being juxtaposed contiguous to the membrane, introducing carbon dioxide into catholyte in the cell or into catholyte being recirculated outside the cell so as to convert substantially all of the alkali metal hydroxide produced in the catholyte compartment to alkali metal carbonate, and removing alkali metal carbonate from the catholyte compartment or from the recirculating catholyte.

The preferred features and the various advantages of the invention will become apparent from the following description. It has been ascertained that, by electrolyzing an alkali metal chloride in an electrolytic cell according to the method of the invention, current efficiencies of 95%-100% are surprisingly attained, even at high catholyte total solids and with low electrolyzing voltages.

The electrolytic cell, having the contiguous cathode-membrane configuration used in the invention, may be a single cell or a plurality of cells combined together in a single electrolyzing unit either in series using bipolar electrodes or in parallel using monopolar electrodes. The cells are generally conventional having a housing resistant to the electrolytes, and being separated by the membrane into anolyte and catholyte c,ompartmentsthe anolyte compartment having an inlet and outlet for the alkali metal chloride brine and outlet for chlorine gas; and the catholyte compartment having an inlet(s) for water and/or recirculated catholyte and outlets for product catholyte and hydrogen, and a CO2 inlet, preferably at or near the bottom of the cell, if CO2 is to be introduced into the catholyte in the cell.

The membrane dividing the cell housing into anolyte and catholyte compartments may be, in general, any hydraulically impermeable cation-exchange membrane electrolytically conductive in the hydrated state obtaining under cell operating conditions and useful for electrolyzing alkali metal chloride brines. These membranes comprise a film of a polymer, chemically resistant to the anolyte and catholyte, containing hydrophylic, ion-exchange groups such as sulfonic groups, carboxylic groups and/or sulfonamide groups.

Membranes made from polymers containing sulfonic and/or carboxylic groups have been found to have good selectivity (that is, they transport virtually only alkali metal ions) and low-voltage characteristics for the production of both sodium and potassium carbonates, while membranes containing sulfonamide groups appear to be useful for sodium carbonate production but require a somewhat higher electrolyzing voltage.

Typically, these membrane polymers have an ion-exchange group equivalent weight of about 800-1500 and the capacity to absorb, on a dry basis, in excess of 5 weight percent gel water.

The cation of the ion-exchange group

and the like) in the membrane will mostly be the same alkali metal as present in the chloride salt being electrolyzed to the carbonate salt. While the acid or other alkali metal salt form can be employed at start-up, it will be appreciated that the membrane will exchange virtually all of these cations for the cation of the salt being electrolyzed within a relatively short period of cell operation. Polymers having all of its carbon hydrogens replaced with fluorine atoms or the majority with fluorine atoms and the balance with chlorine atoms, and having the ion-exchange groups attached to a carbon atom having at least one fluorine atom connected thereto, are particularly preferred for maximum chemical resistance to the anolyte.To minimize electrolyzing voltage, the membrane preferably has a thickness in the range of about 3 to 10 mils, with thicker membranes in this range being used for better durability and selectivity. Because of the large cross-sectional areas of commercial cells, when the membrane is not supported on both sides by contiguous electrodes, it typically will be laminated to and impregnated into a hydraulically permeable, electrolytically nonconductive, inert reinforcing member such as a woven or nonwoven fabric made from fibers of asbestos, glass, TEFLON and the like. In film-fabric composite membranes, it is preferred that the laminate have an unbroken surface of the film resin on both sides of the fabric to prevent leakage through the membrane caused by seepage along the fabric yarns. Such composites and methods for their manufacture are disclosed in U.S. 3,770,567.Alternatively, films of the membrane polymer may be laminated to each side of the fabric.

Suitable membranes are available from the E. I. duPont de Nemours & Co. under the trademark NAFION.

The preparation and description of suitable NAFION and other types of membranes is provided, among others, in British Patent 1,184,321, German Patent Publication 1,941,847, U.S. Patent Nos. 3,041,317, 3,282,875,3,624,053, 3,784,399, 3,849,243, 3,909,378,4,025,405, 4,080,270,4,101,395, and U.S.S.N. 817,007.

The cathode used in the electrolysis cell of the invention process, may be any conventional electrically conductive material resistant to the catholyte, such as iron, mild steel, stainless steel, nickel, and the like. The cathode is foraminous and gas permeable, preferably having at least 25% of its surface area open to facilitate the generation, flow and removal of hydrogen gas in the catholyte compartment and the circulation of carbon dioxide and/or bicarbonate ions to the cathode-membrane interface.To reduce the electrolyzing voltage, all or part of the surface of the cathode may bear a coating or layer of a material lowering the hydrogen over-voltage of the cathode, such as are disclosed in U.S. 4,024,044 (melt-sprayed and leached coating of particulate nickel and aluminum), U.S. 4,104,133 (electrodeposited coating of a nickel-zinc alloy), and U.S. 3,350,294 (coating of molybdenum and tungsten and cobalt, nickel or iron). When some of the cathode surface is devoid of such coating or layer, it typically will be the area of the cathode juxtaposed or laminated to the membrane. Suitable cathodes can be made from, for example, expanded mesh sheet, woven wire screen or perforated plates. Especially preferred are cathodes having an opening (void) area of at least about 50% and good gas-release characteristics, such as the parallel-plate electrodes described in S.

African Patent No.73/8433 (and U.S.S.N. 303,082), the disclosure of which is incorporated herein for more explicit teaching of these types of electrodes. This type of cathode is particularly effective for concentrated catholyte solutions (e.g. 80% to 100% saturated), especially K2C03 solutions, in which higherviscosities and other solution hydrodynamic effects impede the formation, flow and escape of hydrogen gas. Retained hydrogen causes "gas blinding or blanketing" of the cathode, thus increasing the electrolyzing voltage.

Because the vertical parallel configuration of the electrode elements minimizes gas holdup and hence gas blinding of the electrode, electrolyzing voltage is minimized. While parallel plates are described and iliustrated in the South African Patent, it is evident that other elongated electrode elements having other cross-sectional shapes, such as round, elipsoid, triangular, diamond, and square, can be utilized for these preferred cathodes, so long as they are disposed in substantially vertical alignment and with sufficient spacing between adjacent elements to provide good electrolyte circulation and unimpeded flow and release of gas in the catholyte compartment.

In the invention process, the cathode and membrane are juxtaposed such that at least a major portion of the cathode-membrane interface is contiguous. While as little as 50% touching is advantageous, lowest IR' drops will be achieved when 90 or 100% of their interface is contiguous. Means for achieving this are varied and well-known: as for example, employing a greater anolyte hydrostatic pressure to force the membrane against the cathode as shown in U.S. 3,057,794; sandwiching the membrane between the anode and cathode with zero clearance at their interfaces; compressing the membrane between the cathode and anode with suitable resilient compressing means such as shown in U.S. 3,873,437; forming the membrane in situ upon the surface of the cathode (by means such as coating, dipping, spraying, polymerizing or fusing together suitable polymer precursors, solutions, dispersions, powders or fibers) as shown in U.S. 4,036,728 and 4,101,395; or laminating a membrane film to the cathode using heat and pressure as shown in U.S.

4,101,395. In any of these methods, an intermediate layer of suitable inert, nonconducting polymeric material or polymeric precursor (nonhydrophylic and substantially free of ion exchange groups) may first be applied to the cathode (by spraying, brushing, dipping and the like) to inactivate cathode surfaces to be abutted against the membrane and/or to improve interfacial adhesion when the membrane is formed in situ upon or is laminated to the cathode. Considering the foregoing it is apparent that the expressions "juxtaposing, abutting and contiguous," as used in the specification and in the following claims, mean and are meant to encompass, unless otherwise indicated, not only a touching of the membrane and cathode at their interface, but also configurations in which the membrane is formed upon or is laminated to the cathode.

The anode used in the electrolysis cell of the invention process, similarly, may be any conventional, electrically-conductive, electrocatalytically active material resistant to the anolyte such as graphite or, more preferably, a valve metal such as titanium, tantalum or alloys thereof bearing on its surface a noble metal, a noble metal oxide (either alone or in combination with a valve metal oxide), or other electrocatalytically active, corrosion-resistant material. Anodes of this preferred class are called dimensionally stable anodes and are well-known and widely used in industry. See, for example, U.S. Patents 3,117,023,3,632,498, 3,840,443 and 3,846,273.While solid anodes may be used when the anode is spaced apart from the membrane, foraminous anodes having about 25% or more of their surface area open, such as an expanded mesh sheet, woven mesh screen, or perforated plate, are preferred since they have greater electrocatalytic surface area and facilitate the formation, flow and removal of the chlorine gas in the anolyte compartment.

As previously described, good gas-release electrodes having 50% or more open area, such as disclosed in South African Patent No. 73/8433, and discussed hereinbefore, may especially be preferred when the anode also is juxtaposed contiguous to the anode and/or when nearly saturated brines are used.

With respect to the spacing of the anode from the membrane, this distance ideally is the minimum that maintains high current efficiency with respect to chlorine generation, and minimizes voltage. Usually, minimum voltage is achieved when the anode is contiguous to the membrane or the membrane is laminated to the anode.

The invention process can be used to produce any alkali metal carbonate starting with the corresponding alkali metal chloride. Thus, sodium, potassium and lithium carbonates are made from sodium, potassium and lithium chlorides respectively.

As is the conventional electrolysis of alkali metal halides to form chlorine and alkali metal hydroxide and hydrogens, the alkali metal chloride is charged to the anode compartment to become the cell anolyte as an aqueous solution commonly referred to as "brine". The brine typically is acidified with an acid, such as hydrochloric acid, to a pH of about 4 or less to minimize oxygen evolution at the anode and to minimize the formation of insoluble precipitates on or in the membrane from the polyvalent cation salts, such as calcium or magnesium chlorides, present in the brine.

Alternatively or in addition to the aforedescribed control of pH, the deleterious effect of polyvalent cation salts may be minimized by adding to the brine a compound capable of forming with the polyvalent cation salts at a pH of greater than 5.5 an insoluble gel at the anolyte-membrane interface, the gel being reversible at a pH of less than 3.0, as disclosed in U.S. Patent 3,793,163. Illustrative of such gel-forming compounds, which can be used in the present invention, are alkali metal phosphates, orthophosphates, and metaphosphates (preferably having the same alkali metal as the charged brine), or the free acid form of these phosphates.

Typically, the brine is charged at or close to saturation in order to maximize the anolyte concentration, and hence minimize the voltage requirements of the cell. Also affecting anolyte concentration are the rate of charging the brine and the current density of the cell. More rapid brine-charging rates increase anolyte solids, while higher current densities, conversely, deplete anolyte solids more rapidly. Ideally, these three interrelated parameters are chosen and controlled so that the anolyte will have a solids concentration of about 75% or greater of saturation to minimize voltage requirements. Anolyte concentrations of less than 75% of saturation, of course, are equally suitable when higher cell voltages are acceptable.

In the cathode compartment, electrolyte is charged at the start-up of the process to provide initial catholyte. Typically, this electrolyte will have the same alkali metal as the brine and will be a carbonate salt to facilitate rapid equilibrium. After start-up, the catholyte is continuously replenished during electrolysis by the alkali metal ion of the charged brine migrating through the membrane; and the catholyte solids are adjusted to the desired concentration by adding water to the catholyte. If it is desired to minimize the energy required for drying the carbonate salt product, the catholyte concentration will be maintained at or near the -saturation point of the carbonate salt, e.g. 75-100% of the saturation concentration.Conversely, if lower electrolyzing voltages are the paramount consideration, then lower catholyte concentrations will be used ~with the optimum concentration being determined by the cell and cathode design, the type of carbonate salt, and the catholyte temperature. For example, when a small mesh 316 stainess steel cathode (having diamond-shaped openings 0.25 X 0.5 inches) was used to produce K2CO3 at 2 asi, voltage dropped from 4.77 volts @ 640 g/l to 4.2 volts @ 490 g/l or a decrease of 0.375 volts per 100 g/l decrease in catholyte solids.Using a nickel parallel-plate cathode, such as described in the examples and having better gas release characteristics, it was observed that voltage dropped from 3.95 volts @ 640 g/l to 3.7 volts @ 490 g/l or a decrease of 0.167 volts per 100 g/l decrease in catholyte solids.

In the invention process, carbon dioxide gas is introduced into the catholyte so that it and/or bicarbonate ions, resulting from the reaction of the carbon dioxide with OH ions, react with the alkali metal hydroxide produced in the catholyte compartment from the alkali metal ions migrating through the membrane and the hydroxyl ions generated at the cathode. This may be accomplished by directly injecting carbon dioxide into the catholyte compartment usually at or near the bottom to provide maximum mixing and contact time of the carbon dioxide with the catholyte, and preferably with sufficient exit velocity to minimize plugging of the carbon dioxide inlet port(s) and provide good mixing. Alternatively, carbon dioxide can be passed in a carbonator into recirculating catholyte outside the cell.This mode of operation and typical carbonators are shown in U.S. 552,895, 3,179,579 or 3,819,813. The recirculation rate preferable is at a rate sufficient to ensure that the catholyte solids in the cell contain at least 90% by weight of the carbonate salt.

The quantity of carbon dioxide introduced into the catholyte should be sufficient to give catholyte solidscontaining at least about 90% by weight of the desired carbonate salt if high current efficiencies, i.e. on the order of about 90% or greater, are to be attained. More preferred, however, is the use of carbon dioxide in the quantity producing about 95% by weight or more of alkali metal carbonate in the catholyte solids, since current efficiencies (with respect to catholyte product solids) are maximized in this range, generally exceeding 95%. For this reason, a quantity of carbon dioxide producing substantially only carbonate salt is ideally and most preferably used. When less is used, the carbonate product will contain minor amounts of the alkali metal hydroxide, while more may give carbonate product containing a minor quantity of the bicarbonate salt.

The carbon dioxide employed in the invention process may be essentially 100% pure or may be admixed with other gases such as nitrogen and oxygen, as for example when flue gases resulting from the combustion of coal, gas, oil and the like are used as the source of the carbon dioxide. However, flue-gas carbon dioxide will not normally be used when high-purity hydrogen gas is desired.

The temperatures of the anolyte and catholyte in the invention process are not especially critical with respect to achieving high current efficiency. However, because voltage diminishes as the temperature increases,temperatures of about 90"C or more are preferably utilized when it is desired to minimize power consumption.

In the invention process, typically a magnitude of current density in excess of one ampere per square inch (asi) of membrane area is utilized to reduce the alkali metal chloride level in the catholyte solids to less than 400 parts per million (ppm) as described in U.S. 4,080,270. The magnitude of current density required to achieve this low level of salt impurity will vary depending upon the thickness, ion-exchange groups and equivalent weight of the membrane utilized and can be readily ascertained. If higher chloride salt impurity levels are acceptable then lower current density levels may be used. Typical current densities that may be used are 1 to 4 asi (15.5 to 62 asdm).

Typically, the catholyte is discharged from the cathode compartment or drawn off from the recirculating catholyte at a rate proportional to the rate of transport of the hydrated alkali metal ions through the membrane (proportional to current density) and the rate of any external water added to the catholyte so as to maintain an essentially constant catholyte volume. After being discharged, the catholyte typically is transported to a holding tank prior to further processing such as concentrating, drying or packaging for shipment. At this point any residual by-product hydroxide or bicarbonate can be chemically removed if deemed undesirable in the final product.

Alkali metal carbonates, and particularly the sodium and potassium carbonates, are well-known large volume industrial chemicals. Like the products of the prior art, the alkali metal carbonate produced by the invention process can be marketed either as liquors or as anhydrous or hydrated solid materials and are produced from the discharged catholyte by means conventional to the industry such as concentrating, drying and the like. Similarly, they can be used for like end uses such as: in the manufacture of glass, alumina, paper and detergents; as the precursor of other alkali metal compounds; and as regenerable absorbents for carbon dioxide and hydrogen sulfide.

While the preceding description and following examples are directed, for clarity, to single cells, it will be obvious that in commercial operation a plurality of such cells will usually be combined in a single electrolyzing unit either in a series arrangement using bipolar electrodes or in a parallel configuration using monopolar electrodes.

Examples 1-7 In the examples, a cylindrical laboratory electrolysis cell separated by a cation-exchange membrane into anolyte and catholyte compartments and having an inside diameter of 2 inches (50.8 mm) was used. The anode and cathode, both slightly smaller in diameter, were positioned contiguous to the membrane, except in Examples 5 to 7 where the anode was spaced 0.125 inch (3.2 mm) from the membrane. The anode in air the examples was an expanded mesh of titanium metal bearing a 2TiO2:RuO2 coating.The cathode in Examples 1-4 was a 0.063 inch (1.6 mm) thick low-carbon steel plate having 0.25 inch (6.4 mm) diameter holes therethrough spaced apart from each other 0.62 inch (15.8 mm) center to center (58% open area), while Examples 5 to 7 employed an array of vertical nickel plates 0.062 inch (1.6 mm) thick and 0.5 inch (12.7 mm) wide disposed vertically and perpendicularly to the membrane and spaced about 0.19 inch (4.8 mm) apart from each other. In all the examples, the hydrostatic pressure of the anolyte upon the membrane exceeded that of the catholyte. Except for Example 6 run as a comparison experiment, carbon dioxide was introduced into the rear of the cathode compartment at the bottom of the cell and in a quantity theoretically sufficient to convert all the alkali metal hydroxide electrolytically formed to carbonate salt.

In Examples 1 to 4, the membrane utilized was a supported film (T-12 square-woven TEFLON fabric), having an average thickness of about 3.5 mils (0.09 mm) in Examples 1-3 and 5 mil (0.13 mm) in Example 4, of a copolymer having recurring units of: - CF2-CF2and

and an -SO3H equivalent weight of 1100. The membrane of Examples 5 and 6 also utilized the same copolymer and had a thickness of 5 mils (0.13 mm), but was reinforced with the more open T-900 G square-woven TEFLON fabric.

In Example 7, the membrane used (NAFION-415) was a 7.0 mil (0.18 mm) film comprised of two integral layers of different copolymers laminated to the T-900 G fabric. The layer laminated to the fabric had a thickness of about 6.1 mils (0.155 mm) and comprised a copolymer having recurring units of: CF2-CF2- and

and an -SO3H equivalent weight of 1100. The second layer had a thickness of about 0.9 mils (0.023 mm) and comprised a copolymer having recurring units of: - CF2-CF2and

and an -COOH equivalent weight of 1014. In the cell, the layer containing carboxyl groups faced the cathode.

In all the examples, KCI or NaCI brine, containing about 350 parts per million of H3P04 and acidified with HCI to a pH of about 2.0, was charged to the anolyte compartment at a rate of 0.75-1.0 ml per ampere-minute, while aqueous carbonate salt was charged to the catholyte compartment to provide initial catholyte. During electrolysis water was added to the catholyte compartment at a rate sufficient to provide the desired catholyte concentration. The temperature of the anolyte was controlled at about 90"C, while catholyte temperatures varied between 70 to 90"C depending upon the current density used. Other process parameters and the results obtained were as shown in Table 1.

TABLE 1 Data for Examples 1-5 K2CO3 KOH KCl Brine or or or Current Current Kind Concentration Na2CO3 NaOH NaCl Density Electrolyzing Efficiency2 Example (g/l) (g/l) (g/l) (g/l) (asi) Voltage (%) 1 KCl 210 247 8.4 1.0 2 3.2 --2 KCl 205 434 15 0.15 2 3.7 100 3 KCl 205 486 10 0.04 4 5.6 97 4 NaCl 285 348 12 0.20 2 3.9 104 5 KCl 280 712 11 0.053 2 4.1 98 6 KCl 280 --- 600 0.053 2 4.6 42 7 NaCl 300 266 5 0.24 2 4.1 100 1 Rounded out to the nearest one-tenth volt.

2 Rounded out to the nearest one percent. Current efficiency not determined for Example 1; believed to be greater than 95%. In Example 4, current efficiency exceeding 100% due to experimental error in the sampling and/or analysis procedure utilized.

3 Not determined. Typical value for similar runs.

From the data in Table 1, it can be seen that the invention process provides nearly theoretical (100%) current efficiency with respect to catholyte product and operates at low electrolyzing voltages. In this connection, the advantage of having the membrane contiguous to the cathode was demonstrated by backing the cathode away from the membrane in increments at about one minute intervals in Example 1, and noting the effect upon electrolyzing voltage. The results of this experiment are summarized in Table 2, which shows the considerable voltage savings that are realized when the invention process is used.

TABLE 2 Effect of Cathode-Membrane Gap in Example 1 Cathode-Membrane Gap (inches) Electrolyzing Voltage o (start) 3.18 0.03 3.20 0.12 3.34 0.25 3.57 0.38 3.77 0.50 4.01 0.62 4.23 0 (return) 3.18 Example 5 shows that at higher catholyte concentrations (50-100% of saturation - particularly for potassium carbonate, which can give 50% total solids catholyte) lowest electrolyzing voltages are obtained with a more-open, good-gas-release cathode. Lastly, Example 6, illustrating the prior art, demonstrates the poor current efficiencies which characterize membrane-cell electrolysis process for caustic run at high catholyte concentrations and with the membrane contiguous to the cathode.

Claims (9)

1. A method for producing an alkali metal carbonate, which comprises electrolyzing an alkali metal chloride in an electrolytic cell having an anode and a cathode in anolyte and catholyte compartments which are separated by a permselective cation-exchange membrane impervious to hydraulic flow, the cathode being juxtaposed contiguous to the membrane, introducing carbon dioxide into catholyte in the cell or into catholyte being recirculated outside the cell so as to convert substantially all of the alkali metal hydroxide produced in the catholyte compartment to alkali metal carbonate, and removing alkali metal carbonate from the catholyte compartment or from the recirculating catholyte.

2. A method according to claim 1, wherein the membrane is laminated to the cathode.

3. A method according to claim 1, wherein the membrane is formed upon the cathode.

4. A method according to any preceding claim, wherein the cathode comprises a plurality of substantially vertical electrode elements spaced from one another a distance sufficient to facilitate circulation of catholyte to the interface of the cathode and the membrane and to promote gas release from the catholyte and minimise gas blanketing of the cathode.

5. A method according to any preceding claim, wherein the alkali metal chloride electrolyzed is potassium chloride.

6. A method according to claim 5, wherein the catholyte solids contain 75% to 100% potassium carbonate and the concentration of potassium carbonate in the catholyte is 60% to 100% of its saturation concentration.

7. A method according to any of claims 1 to 4, wherein the alkali metal chloride electrolyzed is sodium chloride.

8. A method according to claim 1, substantially as hereinbefore described.

9. An alkali metal carbonate, when produced by a method according to any preceding claim.