BIPOLAR ELEMENT FOR HYDROCHLORIC ACID ELECTROLYSIS
DESCRIPTION OF THE INVENTION
The main electrochemical process of industrial interest is represented by chlor-alkali electrolysis, the overall reaction of which is
2 NaCI + 2 H20 → Cl2 + H2 + 2 NaOH
Of the three products, chlorine [ Cl2 ] has the greatest importance and is used for the production of vinyl chloride by its addition to the double bond of ethylene with the formation of dicloroethane, which is then subjected to pyrolysis.
Vinyl chloride is then polymerised through different techniques to produce the corresponding polymer, PVC, which, as well-known, has a large use especially in the building field, for example for piping, frames, furniture of various types.
During pyrolysis, in addition to vinyl chloride, also hydrochloric acid is formed, as a by-product. Hydrochloric acid in turn is reacted in a special unit, known as oxychlorination unit, together with ethylene and oxygen, with the formation of additional dichloroethane, which is sent back to pyrolysis.
The oxychlorination unit solves the problem of utilising the hydrochloric acid, but poses safety problems connected with the presence of oxygen and highly toxic chlorinated by-products which must be disposed of in special incinerators.
An alternative simple method for converting hydrochloric acid to chlorine to be sent to dichloroethane production units would be therefore much more appreciated at industrial level.
Further, chlorine is a key reactant, together with carbon oxide and amines, to obtain isocyanates which, by reaction with glycoles, permit to obtain the family of polyurethanes, increasingly employed for the production of high added value paints and of expanded cell foams appreciated for thermal insulation systems, such as insulation for refrigerators and walls of buildings.
Also in the case of processes for producing isocyanates, hydrochloric acid is formed as a by-product, even in higher quantities with respect to those produced in the vinyl chloride plants. However, this hydrochloric acid cannot be recycled to the process as in this case there is no unit having the previously mentioned characteristic of oxychlorination. The managers of isocyanate plants have therefore two choices, as a first choice sending the by-product acid to a vinyl chloride plant, and more particularly to its oxychlorination unit, if this plant is at a reasonable distance, or as a second solution, commercialising the acid which has various uses in the industrial field, for example in metallurgical applications. Neither of these solutions is satisfactory, as in the first case the operation of the isocyanate plant is clearly subject to the operation and shutdown procedures, scheduled or not, of the vinyl chloride plant, while in the second case commercialisation is penalised by the degree of purity often required and by the fact that not necessarily the geographical areas where the acid is used coincide with those of the isocyanate plants locations.
It is evident therefore that also in this case the availability of a simple process for the conversion of acid to chorine to be used in the reaction stage would certainly attain a great industrial interest. It must be noted that
the vinyl chloride and isocyanate processes are only the most significant examples among the many existing cases where the use of chlorine generates hydrochloric acid as a by-product. The electrochemical conversion of hydrochloric acid to chlorine
2 HCI → Cl2 + H2 is a well-known reaction; however, it is characterised by a scarce industrial application. The reason for its lack of success, certainly surprising considering the above mentioned problems of the existing technologies, is connected to the high investment required, by a large use of graphite as the construction material, the high maintenance costs connected to the brittleness of graphite itself, and finally the high energy consumption. A remarkable improvement is described in US patent 5,770,035 where electrolysis is carried out using an oxygen depolarised cathode and this characteristic, added to others relating to certain process parameters, permits to produce electrolysis cells completely made of titanium (or alloys thereof). The use of titanium solves the problem of the maintenance costs (titanium is a resistant material and does not cause mechanical defects during assembly and operation, in particular start-up and shut-down thermal transients). Further the energy costs are about 30%reduced. For a better comprehension of the present invention, which will be illustrated in the following paragraphs, it must be remembered that industrial electrolysis is carried out with electrolysers (the equivalents of conventional chemical reactors) made of an assembly of elementary cells pressed together to form a stack (the electrolyser in fact) by means of tie-rods or hydraulic jacks. The cells, each one provided with electrolyte and electric
current feed, are in practice mini-reactors, the production of each one summing up to form the overall production capacity of the electrolyser. From the construction standpoint, the cells may be produced as independent units each one made of a pair of half shells, respectively cathodic (negative polarity) and anodic (positive polarity), as illustrated for example in patent application no. DE 19816334 A1. Each pair of half shells is pre-assembled to form a single cell: when the single cells are in turn assembled to form the electrolyser, the separation between the electrolyte contained in one cell and the electrolyte contained in the adjacent cell is provided by the walls of the two shells which are put into contact (3 and 4 in figure 1 of the above patent application).
According to an alternative embodiment, the cells of the electrolyser are not independent units but are made by pressing together suitable construction elements when assembling the electrolyser, known in this case as filter- press electrolyser, as shown for example in figure 3 of US patent no. 4,767,519. In the case of construction elements of the bipolar type, the element comprises a pair of half-shells mechanically and electrically connected in correspondence of the walls, wherein a half shell is directed to operate as cathodic half shell of a cell and the other as anodic half shell of an adjacent cell, where said cells, as aforesaid, are formed only when assembling the electrolyser. Conversely, in the case of elements of the monopolar type, each construction element comprises a pair of half shells both directed to operate either as cathodic or anodic half shells of two adjacent cells: the construction elements in this case are defined respectively as monopolar cathodic or monopolar anodic elements.
Reverting to the case of greater industrial interest of the bipolar construction elements, applied to the chlor-alkali electrolysis process, the cathodic half shell, which will be in contact with concentrated hot caustic soda during operation, is made of a suitably shaped nickel sheet, while the other half shell of the same bipolar element, the anodic half shell, which will be in contact with an acidic and hot solution of sodium chloride and chlorine, is made of a suitably shaped titanium sheet. When the electrolyser is assembled and the single cells are formed as a consequence of the connection of the various elements, the separation between the electrolytes contained in two subsequent cells is provided by the nickel and titanium walls connected to each other in each element.
In the case of monopolar elements, as the material for each pair of half shells is the same for each element, it is possible to achieve an appreciable saving by producing the two half shells in a single sheet provided with means for fixing the electrodes on both sides. This possibility is used in US patents nos. 4,464,242 and 5,013,418, where the fixing means for the electrodes consist in projections obtained on both sides of a single sheet by pressing. This advantage however is largely counterbalanced by the higher cost of the electric equipment, as aforementioned, and for this reason normally the electrolysers assembled with bipolar construction elements are preferred.
Conversely, when the electrolyser is directed to hydrochloric acid electrolysis carried out according to the teachings of US patent no. 5,770,035, it is the design of the bipolar elements which can certainly be modified into a form, which so far had been possible only with monopolar
elements.
In fact, as aforementioned, the process disclosed in US patent no. 5,770,035 permits to use titanium for the construction of both the cathodic and the anodic half shells. Taking into consideration this aspect, the production of bipolar elements is considerably simplified as a single, suitably shaped, titanium sheet can be utilised, which acts both as cathodic and anodic half shell at the same time. This simplification is not possible with the cell design of the previously mentioned patent application no. DE 19816334 A1, as in this case each cell is pre-formed and therefore exists as an independent unit which is subsequently assembled together with other cells, likewise pre-formed, to form the electrolyser.
The possibility of simplifying the design of the bipolar elements to be assembled in an electrolyser suitable for the process disclosed in US patent no. 5.770.035 using a single sheet to form the two anodic and cathodic half shells is taken into account in general terms in Italian patent application No. MI2001A 000401. However, this document is essentially directed to describe the most suitable geometries for the peripheral flange fixed along the edges of the single sheet. Said patent application mentions the fact that the electrodes must be connected to the single sheet by suitable titanium supports, but does not analyse in particular how this connection may be obtained in practice.
This aspect conversely is the core of the discussion contained in the above mentioned US patents 4,464,242 and 5,013,418 relating to the structure of construction elements of the monopolar type. In both documents the construction element comprises a single sheet which is cold-formed in
order to obtain on both sides projections, for example cylindrical or prysmatic in US 4,464,242 and frusto-conical in US 5,013,418, on which the catalysed electrodes, which are made of perforated or expanded sheets or interwoven wire meshes, are fixed, preferably by welding. US 4,464,242 overlooks the consequences created by the particular area where the electrodes and the planar terminal portion of the projection overlap in the case an ion exchange membrane is used and is maintained in contact with the electrodes. In fact these areas lack a free exchange between the electrotyte and the gas generated by the reaction, with the possibility of forming zones with stagnant diluted liquid and/or with entrapped gas practically free from liquid. Consequently these areas become inactive towards electrolysis and therefore the current flow is forcibly concentrated on the residual portion of electrodic surface with an increase of its effective density and of the cell voltage, which results in a higher electric energy consumption. It is further possible that in the crevice zones corresponding to the contact surface with the projections of the sheets, the presence of gas or diluted electrolyte pockets may give rise to totally uncontrollable contractions or expansions of the ion exchange membranes with negative effects onto the mechanical properties of the membranes and the consequent high risk of damages. In the case of US Patent 5,013,418 the above discussed problems are decreased by the presence of suitable spacers which maintain the membrane at a predetermined distance from both the cathodic and the anodic electrodes of each single cell. The distance between the membrane and the electrodes permits a certain exchange of liquid and gas in the
areas of contact between the electrodes and the projections and therefore protects the membrane from deterioration. Also this construction solution, however, is far from being satisfactory for at least two reasons. The first is connected to the ohmic resistance of the electrolytes present in the gap between membrane and electrodes, which results in a higher operation voltage. The second derives from the position of the membrane which is fixed to the spacers by contact points; the membrane is not adequately supported and any pressure variation in the compartments containing the electrodes causes dangerous vibrations. It must be noted also that the design illustrated in US 5,013,418 cannot be used for the cell of US 5,770,035, or better it could be used only on the anode side but certainly not on the oxygen cathode side which must necessarily be in direct and intimate contact with the membrane. Another solution to the problem of the planar contact area between the electrodes and the projections is described in US 5,013,414. In this case the electrode is cold-formed to obtain depressions which have the same arrangement as the projections of the sheet which delimits each cell. Each depression of the electrodes is fixed to the corresponding projection of the sheet and a thin additional perforated or expanded sheet or a mesh of interwoven wire is applied to the surface of the electrodes. This solution, although very efficient to prevent stagnation areas, poses in fact problems, especially from the economical stand-point as the for the electrodes, made as aforementioned of two superimposed sheets, clearly a larger use of material is involved and in the second instance, from a mechanical point of view, the electrode sheets, provided with depressions, are substantially less rigid with a consequently high
flexibility of the overall electrodes-construction element structure. This flexibility may be dangerous during operation when gaseous products are evolved, such as for example chlorine on the anode side, producing internal pressure variations and therefore vibrations capable of causing mechanical damages with time, in particular on the fixing points of the electrodes onto the projections of the sheets and also on the membranes when the vibrations cause rubbing against the electrode surfaces. In the case of hydrochloric acid electrolysis according to US 5,770,035 in addition to the above mentioned inconveniences consisting in the possible presence of crevice areas and vibration of the membranes, a further problem arises by the use of the oxygen cathode. The oxygen cathode performs regularly only when all the points of its catalytic surface are adequately fed with oxygen. If there are blinded areas, as it certainly occurs in correspondence of the planar areas of the cold-formed projections onto the sheets forming the walls of the cells, in these areas the diffusion may be no more sufficient to supply the required quantity of oxygen at the current densities of industrial interest (typically 4000 - 5000 Ampere/m2). When this occurs, the cathode produces variable quantities of hydrogen which mixes with oxygen involving a risk of explosion.
It is therefore the object of the present invention to provide a design of a bipolar construction element suitable for being assembled to form an electrolyser particularly useful for processes of membrane electrolysis of hydrochloric acid with oxygen cathodes, where said design foresees the use of a single sheet of titanium (or alloys thereof) provided with projections on both sides and it is directed to overcome the inconveniences negatively
affecting the construction solutions suggested by the prior art. In particular the construction element of the invention is characterised by rigidity, absence of crevice areas in the contact points between electrodes and projections and capability of maintaining the membrane in a safe position without any risk of dangerous vibrations. .
The construction element of the invention, suitable for being assembled in an electrolyser for the membrane electrolysis of hydrochloric acid with oxygen cathodes, for example as described in US patent no. 5,770,035, is made of a single sheet of titanium (or alloys thereof) provided along its periphery with a suitable perimetral frame capable of ensuring acid and gas sealing.
Oblong projections with the longer dimensions vertically oriented are formed in the single sheet. The projections altematively extend on the two sides of the single sheet.
The projections on one side of the sheet may the identical or even have different dimensions with respect to the ones of the other side. They are preferably characterised by a triangular section with the acute vertex representing the most prominent part. The vertex is necessarily rounded to avoid ruptures of the material under the forming operation, but in any case it is free from appreciable planar surfaces to avoid the formation of blinded areas when the electrodes are fixed onto the projections on both sides of the sheet.
In order to fix the electrodes onto the projection, a very precise positioning is necessary and can be obtained by means of suitable jigs, followed by electric resistance, arc or laser welding. This last system is preferred as it
permits to obtain high quality linear welding capable of ensuring an optimum electrical contact between projections and electrodes. As the laser beam power may be regulated accurately, it is further possible to avoid the formation of defects such as porosity crossing the material thickness in correspondence of the projection vertexes and maintain an adequate planarity of the construction element. Finally, the laser procedure is characterised by high speed of execution, in the range of a meter per second, with the consequent high capacity of production of the construction elements.
Figure 1 represents a front view of one possible embodiment of the construction element of the invention wherein the projections on the two sides are symmetrical.
Figure 2 shows two transversal cross-sections, in the vertical and horizontal directions, of the construction element area defined by the circle in figure
1.
Figure 3 schematises the cross-section along line A of figure 1 of the electrolyser formed by the assembly of a plurality of construction elements of the type illustrated in figures 1 and 2 and suitable for hydrochloric acid electrolysis with oxygen cathodes.
Figure 4 shows a magnified detail indicated by a circle in figure 3 and representing the lower section of the electrolyser in the area of the perimetral gaskets.
The present invention concerns an innovative design of a bipolar construction element suitable for assembling electrolysers directed to hydrochloric acid electrolysis according to the teachings of US 5,770,035.
Said construction element consists of a single sheet of titanium (or alloys thereof) provided with a suitable perimetral frame capable of preventing acid and/or gas leaks according to the design illustrated in Italian patent application no. MI2001 A 000401.
Oblong projections arranged in parallel with the larger dimension vertically oriented are formed on the single titanium sheet by cold, or preferably hot- forming when the desired dimensions for the projections would significantly stress the material during the working procedure.
The projections alternatively extend on the two sides with respect to the original surface of the single sheet and are preferably characterised by a triangular cross-section with the base coinciding with the original surface of the single sheet and the vertex representing the most prominent part of the projections themselves.
Figure 1 represents a frontal view of a possible embodiment of the construction element of the present invention, with projections on the two sides substantially identical but necessarily off-set. Bipolar elements with asymetrical projections on the two sides are a feasible alternative as well. The vertical alignment of the projections of each side is particularly relevant as it permits to obtain during operation a mixing of the acid flowing upwards thanks to both a pump forced feeding and to the uprising action of the gaseous chlorine bubbles formed by electrolysis. Mixing, identified by arrows in the figure, is important as it is essential avoiding local dilutions of the acid in view of the fact that below certain critical concentration values the anodes generate oxygen. Chlorine containing high quantities of oxygen cannot be directly utilised as such by the electrolyser but must be subjected
to expensive purification processes, based on its liquefaction and re- evaporation.
The vertical orienting of the projections further permits to obtain an improved current distribution inside the electrodes, in fact the electrode- projection welding lines are also vertical and parallel with each other and as a consequence the path that electric current must cross results lower when electrodes made of expanded metal are used, as is normal practice. This type of sheet is typically anisotropic and presents a better conductivity in the perpendicular direction with respect to the expansion direction, which is made coincident with the vertical direction in the production of the construction elements. This production procedure is preferred as actually permits a better gas-stripping of the chlorine bubbles. The alternate vertical alignment is also important on the cathodic side where oxygen or an oxygen containing gas, such as air, flows. The alternate alignment causes a certain turbulence of the flow, the linear speed of which is not particularly high, and consequently favours a better refurnishing at the cathode, which in case of insufficient oxygen feed may generate hydrogen inducing a dangerous situation. For this reason the alternate alignment of the projections is preferred over the other type of design where each projection, always vertically disposed, has the same length as the height of the whole bipolar element. In this last case the continuous space among the adjacent projections would cause a channelled motion both for the acid and for oxygen or air, characterised by low lateral mixing with consequent mass transfer limitations. In figure 1 reference numeral 1 indicates the projections on sheet 7 which
protrude towards the observer and 2 indicates the projections which conversely extend towards the opposite direction (dotted lines for a better identification). The perimetral frame 3, aided by suitable gaskets not shown in the figure, permits to prevent add and/or gas leaks. On part of the element surface a reticulated area 4 is schematised to represent an electrode, for example the anode, which is fixed to the projections with a linear welding 21 made along the whole length of each projection. As will be seen, the various welding lines are parallel and with a close pitch corresponding to the lateral pitch of the projections: as above discussed, in this way a better distribution of the electric current is obtained, the flow of which is represented by the arrows in figure 2. Figure 2 shows two transversal cross-sections, in the horizontal and vertical directions, of the bipolar element of figure 1 in correspondence of the area indicated by a circle: reference numerals 1 and 2 indicate the projections which alternate on the opposite sides, 5 and 6 identify the analogous projections arranged in parallel with 1 and 2 and represented by a thin line to facilitate their identification.
Further 4 and 11 identify the two electrodes, respectively the anode and the cathode, or better the cathodic current collector as will discussed here below.
In its simplest embodiment, the bipolar element of the invention comprises two electrodes as shown in figure 2, each one made of a perforated or expanded sheet or a mesh of interwoven wire. This solution, however, is not completely satisfactory as in the case of deactivation the substitution of exhausted electrodes for new electrodes is very difficult in view of the
stability of the linear electric arc or laser welding. Further, the electrodes offer the best electrochemical performances when the perforations or the meshes have small dimensions: this would require a small thickness which would make the construction elements excessively flexible. To achieve both results, rigidity of the assembly (thick electrodes with perforations or meshes of large dimensions) and optimum electrochemical performances (thin electrodes with small dimension perforations or meshes) it is necessary to resort to a composite structure, not indicated in figure 2, wherein the electrodes are each one made by superimposing two perforated or expanded sheets or meshes, one of which is coarser and directed to grant the rigidity of the element, and the other is finer, fixed to the coarser one, directed to obtain the best electrochemical performances. Figure 2 permits to appredate the importance of the geometry of the contact area between the electrodes 4 and 11 and the projections 1 and 2. This area, when exceedingly large, results affected on the anodic side by scarce add diffusion with oxygen formation as well as possible stagnation of entrapped chlorine bubbles. The presence of these stationary bubbles generates an increased chlorine flow through the membrane, in particular when the membrane is in contact with the anode. The consequent chlorine penetration in the cathodic area causes corrosion of the oxygen cathode. The problem is less severe when the membrane is maintained dose to but not in direct contact with the anode, for example at a distance of 1 - 3 mm (see figure 3).
An excessively large contact area on the cathodic side easily causes a problem of oxygen flow on the corresponding portion of gas cathode. As a
consequence, as already said, hydrogen may be generated with a considerable risk of forming explosive mixtures. For this reason, as schematised in figure 2, the vertex of the projections is maintained sharp and only sufficiently rounded to avoid damages to the material when the original sheet is pressed.
Obviously, a sharp vertex of the projections poses problems of positioning and contact with the electrode and then the subsequent welding. For this reason a sheet provided with projections and perimetral frame is positioned together with the two electrodes in a suitable jig where the three components are pressed and maintained in a definitely planar alignment. To favour the achievement of this planarity, the bipolar element, after printing of the projections and before assembling in the jig may be subjected to compression in a suitable press, with the aim of eliminating distortions due to the printing procedure. This compression involves a certain deformation of the vertexes of the projections, with the formation of planar areas the width of which is anyway maintained within moderate limits, for example below 1 mm, to avoid the diffusion problems during operation, already illustrated above. The assembly jig-sheet-electrodes is then subjected to welding which may be carried out by different methods, such as resistance, electric arc and laser welding. The latter is certainly preferred as it permits to obtain linear welding at high speed (in the range of 1 meter/second), and ensures a high production capacity of the bipolar element described herein. Further, as the laser beam power may be accurately adjusted, the electrode-projection welding does not penetrate completely through the thickness of the material in the vertex area of the
projections, where contact with the electrode is established, thus avoiding with a considerable reliability the risk of producing pass-through porosity, capable of putting in communication the anodic and cathodic compartments, with consequent severe operation problems. This possibility, conversely, is a physiologic characteristic of the electric arc welding which affects thoroughly the material of the projection vertex characterised by a rather reduced thickness, as will be illustrated here below. The resistance welding, if carried out with suitable parameters, is likewise free from pass- through porosity, however, with this procedure the welding line is discontinuous rather than continuously linear, with a less efficient distribution of electric current. Further, this procedure requires high contact pressures among the various pieces, with possible deformation of the assembly.
The cross-section of figure 2 dearly shows also the electric current path inside the structure during operation: like the current distribution inside the electrodes (continuous lines) also this path is identified by arrows (dotted lines for an easier identification). To obtain the best performance of the electrolyser made of the present bipolar elements it is necessary that the ohmic drop connected with the current flow be minimised, for example reduced to the maximum value of some millivolts. This is obtained by suitably dimensioning both the thickness of the sheet (7 in figure 1) used for the construction of the bipolar element as well as the distance between each pair of projections located on the two opposite sides of the elements. It has been found that with a current density in the range of 4000 - 5000 Ampere/m2, sheets with a thickness of 1 - 2 mm are quite adequate. As
regards the distance between each pair of projections located on the opposite sides of the sheet, which defines the lateral pitch of the bipolar element (8 in figure 2), it has been observed that the most important factor to be taken into account is the space required by the anodic compartment (distance between anode and sheet) which contains the hydrochloric acid solution and the produced chlorine. A circulation suffidently homogeneous and regular with time may be obtained only when said space has a depth comprised between 2 and 4 cm.
Figure 3 shows the cross-section of a portion of electrolyser obtained by assembling the bipolar elements of the invention. The cross-section refers to the case of two subsequent elements assembled so that projections 1 , 2 may face off-set the same membrane 9. In this way it may be avoided that possible concentrations of current are formed, as may occur when the projections facing the same membrane are exactly one against the other. However, also this solution may be resorted to.
The cross-section of figure 3 refers to the design of an electrolyser wherein the anode 4 is maintained a certain distance spaced apart from the membrane 9. In this arrangement the pressure of the anodic compartment, wherein the anode 4 is in contact with the hydrochloric acid and chlorine solution, is kept at a level higher than that characterising the oxygen fed in the cathodic compartment. The pressure differential maintains the membrane pressed against the oxygen cathode 10, which in turn is pressed against the electrode 11 , which should be more suitably defined as current collector for the cathode, where said electrode 11 or current collector is welded as above mentioned onto the projections of the
cathodic side of the adjacent bipolar element. Also the design of an electrolyser with the anode 4 in contact with the assembly membrane 9 - oxygen cathode 10 - electrode or current collector 11 may be used in practice: this solution permits to obtain in fact better cell voltages, but it is much more critical from a point of view of mechanical construction and assembling tolerances. The risk in fact is that the contact pressure between anode 4 and the assembly membrane - oxygen cathode - current colledor be high and the polymeric membrane be irreversibly damaged. In figure 3 the hydrochloric acid solution inlet and the outlet for the exhaust solution containing chlorine are identified respectively with 12 and 13, the oxygen inlet and the outlet for the exceeding amount containing also the water formed by reaction onto the oxygen cathode are identified respectively by 14 and 15. Further, the anodic and cathodic perimetral gaskets are identified respectively by 16 and 17 and the frames of the bipolar elements are identified by 3.
Figure 4 is a magnification of the detail indicated by a circle in figure 3. The detail represents the lower part of an electrolysis cell, comprising the frames 3 of two adjacent bipolar elements, the peripheral cathodic and anodic gaskets 16 and 17, the membrane 9, the anode 4, the oxygen cathode 10, the electrode or current collector 11. The oxygen cathode is made of carbon doth containing catalyst, polymeric material and ionomeric material as known to the expert in the art. The cathode is therefore more flexible and in order to facilitate its installation on the cell it is a normal procedure to prolong the same until it goes under the peripheral gasket. This cathode portion is identified by 18. As it is not possible in practice to
superimpose the gaskets perfectly, it generally happens that one of the two is not compressed on the internal portion in contact with the process media. This situation is represented in figure 4 for the cathodic gasket. A crevice is formed, identified by 19, which is filled with part of the water formed by the reaction on the oxygen cathode and dropping out through outlet 15. The crevice area, thus flooded, is therefore scarcely accessible to oxygen and therefore gives rise to the dangerous situation already discussed in connection with the possible blinded areas of the electrode or current collector 11 caused by projections with a planar and large vertex. If it is desirable to maintain the installation technique of the oxygen cathode, which results extremely convenient, it is necessary either to avoid the formation of the crevice 19, which would be rather problematic, or deactivate the crevice itself. This result is efficiently and simply obtained by applying a non conductive adhesive tape or a non conductive polymeric paint 20 onto the portion 18 of the oxygen cathode directed to being positioned under the gasket. For a more reliable result of this measure, it is obviously safer apply the tape or paint on a portion larger than the part which has to be covered by the gasket, to compensate for the uncertainty of the dimensions of said portion, depending on the dimensional tolerances of the elements, of the gaskets and by the imprecision typical of industrial assembling. In this way it is avoided that the part of cathode contained in the crevice 19 be affected by current and thus hydrogen generation is completely avoided.