By Dr. Otto Steiner
Electrochemical methods are generally employed in chemical industries for the sake of simplifying the operation. Purely chemical processes, like the LeBlanc, McDonald, Weldon, Deacon and Solvay processes, involve a whole series of different reactions in order to get the desired end products from common salt. The electric current, on the other hand, decomposes the salt in one single apparatus and in one single operation. The immediate products of electrolysis are chlorine gas at the anode and caustic soda at the cathode. It is only necessary to so construct the apparatus as to prevent die reunion of these two products.
For this purpose it is necessary to separate the anodic liquid, in which some chlorine gas is dissolved, from the cathode liquid which contains the caustic alkali. While the chlorine set free at the anode mostly escapes 'as gas, yet some of it dissolves in the liquid. If this anode liquid containing the chlorine then mixes with cathode liquid containing hydroxide hypochlorite will be formed. By application of a diaphragm, it is possible to separate the anolyte from the catholyte so as to prevent the mechanical mixing of the two solutions, without preventing the passage of the electric current.
But the action between the two solutions is due not only to mechanical mixing, but is also electrochemical in nature, since the caustic alkali in the catholyte participates in the conduction of the current. The OH ions of the catholyte will travel through the diaphragm into the anolyte. In this way caustic alkali is transported towards the anode in spite of the diaphragm. In consequence of this, it is impossible on a large commercial scale to manufacture highly concentrated caustic solutions by the diaphragm process without running the risk of greatly reducing the ampere-hour efficiency, together with other troublesome disadvantages.
Various methods have been suggested to prevent the electro-chemical reaction between the two liquids. There are some hundred different processes, aiming at this end, but only those of special simplicity have proven industrially successful. This is especially the case for the Aussig "Glocken" (bell) process of the "Oesterreichische Verein fur Chemische and Metallurgische Production in Aussig" (Bohemia). This process is now used on a very large scale (about 6,000 h.p.) in Austria as well as in Germany.
The general principle of the process is shown in Fig. 1. An earthenware bell (Glocke) is suspended in the salt solution in such a way that it does not reach down to the bottom. The anode is inside of the bell, the cathode outside.
The anode liquid, which has a green color on account of its content of dissolved chlorine gas, is of lower specific gravity and fills the upper part of the bell. The colorless cathode liquid fills the part underneath as well as the whole outside of the bell.
Between both solutions is a layer of neutral solution, which remains at the same place, if the current, the content of hydroxide in the catholyte and the content of alkali chloride in the anolyte are kept constant. This can be accomplished in continuous operation by introducing the fresh salt solution into the bell above the anode, so as to distribute-it very uniformly over the whole anode surface, and by continually removing the formed caustic solution from the cathode compartment by simply overflow.
The anolyte, containing chlorine, moves inside of the bell downwards in consequence of the continuous supply of fresh electrolyte. It finally passes outside of the bell, and the hydroxide of the catholyte will then react with the dissolved chlorine and form hypochlorite, which will be reduced to alkali chloride at the cathode. In practice this reaction between chlorine and hydroxide will cause a loss of 6 per cent in the ampere-hour efficiency.
The idea of permitting the anode liquid with the chlorine to pass over directly to the cathode is rather audacious and is almost stunning at first sight, because this is exactly the thing which ought to be prevented— namely, the mixing of the anolyte with the catholyte. But as a strong salt solution does not dissolve more chlorine than 0.4% of its weight, the loss in ampere-hour efficiency is very small, it amounts on a large scale to only 6%. On the other hand, this arrangement has other very great advantages. The liquid surrounding the anode is constantly maintained saturated with chloride and the OH ions, which migrate away from the cathode towards the anode, are thus prevented to reach the anode.
Figure.1 Principle of the Glocken Process
The reason is this: Both the OH ions and the CI ions are migrating from the cathode to the anode, and the part which these two kinds of ions take in the electric conduction depends, according to Hittorf, on their number and their speed of migration. If, for example, the solution at the cathode contains 12 per cent KOH and 12% KC1, it will be found that the negative ions traveling away from the cathode are almost completely OH ions.
But on their further advance towards the anode they will come into layers, which contain more KC1, and therefore more CI ions, on account of the continuous supply of fresh salt solution at the anode. Therefore, the OH ions will gradually remain behind and the CI ions will chiefly undertake the negative transmission of the current. At the anode itself the KC1 concentration is maintained constantly at 20 per cent, so that the only negative ions moving at this place will be the CI ions and no OH ions will reach the anode.
In expositions of the theory of the Glocken process the wrong idea is often found that the electrolyte moves inside of the bell downwards with the same speed with which the OH ions move upwards, and that it is necessary to regulate the rate of supply of fresh alkali-chloride solution in such a way that the neutral layer which forms between the anolyte containing chlorine and the caustic catholyte remains on the same place. Even the German patent 141,187 of the Aussig Company uses this explanation. It is said there: "2. The continuous admission of fresh electrolyte at the anode is so regulated, that the speed with which the anode liquid passes downward, is equal to that with which the caustic alkali moves upwards on account of ionic migration, so that the separating layer between anolyte and catholyte is maintained constantly at the same place." It would probably be more exact to say that the separating layer is maintained at the same place by keeping the alkali chloride concentration constant in the surroundings of the anode.
For, in reality, the explanation of the patent is not admissible. The position of the neutral layer depends chiefly on the concentration of caustic alkali at the cathode and on the KCI concentration at the anode and not on the speed with which the solution moves. The speed with which the electrolyte inside of the bell moves downward (in practice 1 cm. per hour) is almost insignificant, since the OH ions travel with a speed of 6 cm. per hour upwards.
According to the patent these two speeds should be equal. But that is impossible in practice. To bring the downward speed of the<solution up to 6 cm. per hour it would either be necessary to accelerate the supply of fresh electrolyte so much that with a current density of 2 amperes per square decimeter of horizontal cross section of the bell (which is the limit on account of the disintegration of the anode material) a very weak caustic alkali solution would be obtained. Or it would be necessary to make the cross section of the bell below the anode so small that the voltage, and therefore the expenses for electric energy would be considerably increased. But all these things are unnecessary. If the liquid around the anode is continuously kept saturated with KCI or NaCl by careful subdivision and mixing of the fresh electrolyte, which is introduced into the bell, it is possible to prevent the OH ions from reaching the anode, provided there is no disturbing circulation of liquid within the bell.
For other reasons it would also be evidently impossible to regulate the admission of electrolyte so that the speed with which the anode liquid passes downwards is equal to that with which the OH ions move upwards.
Figure.2 Later Construction of Cell
On that theory the neutral separating layer would then only be stationary in the one case that both speeds are exactly equal. This can, of course, never be accomplished on a commercial scale, so that whenever the two speeds are not absolutely equal, the neutral layer would travel either upwards or downwards. But, as a matter of fact, I have proved by experiments that the neutral layer always remains at the same place, whatever the speed of the electrolyte may be, if only the current and the supply of fresh electrolyte are maintained constant. My explanation is therefore the only one admissible-namely, that the passage of the caustic alkali right to the anode is prevented in the Glocken process "by keeping the liquid surrounding the anode as much as possible saturated with KCI, so that near the anode no OH ions participate in the negative transmission of the current.
The employment of a concentrated alkali chloride solution and a good distribution and mixing of the new with the old solution at the anode may also be used to advantage in the diaphragm and mercury processes where similar conditions exist.
The technical arrangement of the Aussig Glocken process requires a somewhat different construction, somewhat as shown in the adjoining Fig. 2. The even uniform distribution of the fresh solution, over the whole anode surface and its thorough mixing with the solution, which already surrounds the anode, and which is weaker and therefore of lower specific gravity, are 01 greatest importance, that for this purpose the anode should be made so large that it fills almost completely the whole horizontal cross-section of the cell. Only a small space of a few millimeter is left between the anode and the cell wall, and in this narrow space the rising chlorine bubbles cause an intimate mixing of the solutions.
I have proved experiments, that this condition alone is not sufficient, but that it is also necessary that the upper surface of the anode, or at least the exterior rim of it, should be absolutely level and horizontal. If this surface is only a little inclined, all the fresh solution, being of higher specific gravity, will flow down in one stream ("schliere") from the lowest point of the anode surface without being distributed over the whole anode surface and without getting thoroughly mixed with all the old solution. As a result the anode liquid will get weaker in alkali chloride. OH ions will come near to the anode, oxygen will be set free, disintegration of the anode material will occur, in short, we have all troubles which it is of greatest importance to avoid.
Not only the flat form but also the absolutely horizontal position of the anodes is therefore of greatest importance and every bell must be an apparatus of precision if it shall operate properly. By careful observation of this precaution Acheson graphite will prove such an excellent anode material that it will last for five years.
To sum up the advantages of the Aussig bell process, first of all we have the simplicity of its apparatus, then the high ampere-hour efficiency, the high concentration of the caustic alkali produced and the long life of the anodes. Its disadvantage is that for a large scale production a very great number of small apparatus are necessary, since the bells cannot be made too large. The diaphragm and mercury processes have also their advantages and disadvantages and it depends on local conditions which process one shall select. With respect to simplicity the Aussig Glocken process will not be excelled by any other electrolytic alkali-chloride process.
The "Oesterrenchischer Verein fur Chemische und Metallurgische Production in Aussig" has not only invented the Glocken process but has also worked out all the details with so much success that the process is now used not only in Austria but in several places in Germany, in successful competition with the other older electrolytic processes.
The experiments on which the results given in this article are based were carried out under the direction of Prof. Dieffenbach at the Electrochemical Laboratory of the Institute of Technology of Darmstadt. The author wishes to express here his thanks to Prof. Dieffenbach for the assistance given him during the investigation.
Image Credit: Miami University Libraries