118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No.13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions.Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13. The Fwaday Society is not respoirsible for opittiotzs expressed before it by Authors or Speakers. Cransactione OF Cbe TO PROMOTE THE FOUNDED 1903. 8TUDY OF ELECTROCHEMISTRY, ELECTROMETALLURQY, CHEMICAL PHYSICS, METALLOQRAPHY.AND KINDRED 8UWECTS. VOL. XIX. MARCH, 1924. PART 3. THE ELECTRODEPOSITION OF MANGANESE.-PART I.. BY A. J. ALLMAND AND A. N. CAMPBELL. (A paper read befare THE FARADAY SOCIETY, Monday, November I 2th, I 923, SIR ROBERT ROBERTSON, K.B.E., F.R.S., PRESIDENT i~ the chi^.) ReceivedJuZy 3rd, I 92 3. I. h?tYodwto?y. The problem of the cathodic deposition of manganese from aqueous solution has received little attention in the past. Bunsen states that he obtained the metal by electrolysis of aqueous manganous chloride solutions, using the apparatus employed in the deposition of chromium. The metal was deposited as sheets I cm.2 in area; it was metallic looking, and oxidised almost as rapidly as potassium. If the current density were reduced, mangano-manganic oxide came down.He gives no further details. Moore2 states that metallic manganese may be deposited as such from a neqtral solution containing a large excess of ammonium sulphocyanide ; a pderful current is necessary. Smith and Frankel find that, if an excess of potassium sulphocyanide be present, the metal comes down in greyish- white compact form. Under the conditions specified, the current must be low. Van Arsdale and Meier 4 give results of some experiments conducted on the electrolysis of manganese sulphate solutions. These will be referred to in more detail below, as they were to a large extent repeated by the present authors, t o whom the paper was unknown when the work was started. Finally, The metallic deposit is inclined to oxidise rapidly.Pogg. Ann., 1854.91, 621. Chem. News., 1886,153, ZOQ. 9 your. Aturlytical Chem., 1889, 3, 386. 4 Trans. Am#. Electrochem. Soc., 1918, 3, 109. 5595 60 THE ELECTRODEPOSITION OF MANGANESE-PART I. Foerster quotes some results obtained by Grube, who states that very pure manganese can be obtained by electrolysis of a 6-7 N. MnCl, solution, separated by a diaphragm from the anode. The electrolyte is also 1.5 N. with respect to NH,C1 and 0.1 N. with respect to HCl. The current density is 20 amps./dm.2 The electrolyte is strongly stirred, and its temperature 30’. The current efficiency is said to be between 5 0 and 60 per cent., and the purity of the deposit 99.9 to IOO per cent. “There is a marked tendency for the deposit to sprout at the edges of the cathode, as a result of the high current densities used.This can, however, be over- come by arranging that the cathode fills the whole cross-section of the cell, and manganese is then obtained as a smooth microcrystalline deposit, which can be removed from the copper cathode in the form of coherent sheet.” In later papers Grube and Metzger2 refer to sheets of metal I mm. thick made by this method. No detailed account of Grube’s work has yet been published. For this reason, and apart from the interest of our own experiments, we have decided to publish the results of our work, though it may be stated at once that we have not yet succeeded in obtain- ing coherent metal sheet as Grube claims to have done. The problem then was the working out of the conditions for the successful electrodeposition of a highly electropositive metal from aqueous solution.We decided in the first instance to use as electrolytes the simple salts of manganese (sulphate, chloride). I t could be predicted in advance that the electrolysis would be favoured by a high metallic salt concentration and by a low hydrogen ion concentration. A high current density was also likely to be advantageous. With regard to temperature, the matter was more complex. To take extreme cases, if manganese were a metal like zinc, with high hydrogen over-voltage and small irreversible resistance to cathodic deposition, a low temperature would be better; if like iron, with small hydrogen over-voltage and large irreversible cathodic effects, then high temperatures would be better.We commenced on the supposition that it would behave more like iron than like zinc, a view which turned out to be only partly correct, as both high hydrogen over-voltage and high irreversible resistance to manganous ion discharge were found. 2. Experiments wifh &@hate Solutions. The electrolyte was prepared from crystallised manganese sulphate, which contained a very slight trace of iron as its only impurity. Unless otherwise stated, the solution used always contained 400 gms. per litre of the tetra- hydrate. SERIES I. The electrolysis bath was an ordinary beaker, lagged with asbestos cord, and kept at the required temperature by a gas burner. The cathode was of aluminium sheet, rectangular in shape, with an immersed area of 2 0 cm.2 On either side of it was hung a platinum foil anode wrapped in parchment paper.This precaution was taken to prevent the manganese dioxide formed in accordance with the equation Mn” + zH,O + 2 @ += MnO, + 4H’ Elektrochemie wdsseriger Losilngen, 1922, page 560. This information only became available when the greater part of the work had been completed. In the 1915 edition of the same book (p. 317), experiments are mentioned according to which the metal can be got by electrolysis of a very concentrated hot chloride solution with a current density of 10-20 amps./dm.z Zeitsch. Elektrochem., 1923, 29, 17, 100.THE ELECTRODEPOSITION OF MANGANESE-PART I. 56 I fouling the bath. No attempt was made at this stage to control the acidity of the bath, and no acid was present initially.Current was supplied by a battery of eight accumulators. The results of these preliminary experiments are summarised in Table I. The deposit from these experiments was very unsatisfactory, consisting largely of basic material. The ready disengagement of hydrogen on treat- ment with very dilute acid showed, however, metal to be present in every case but (8), and in experiments (10) and (11) this amount was estimated by collecting the hydrogen in a nitrometer. The '' current efficiencies " were determined, as throughout this paper, by drying the cathode with filter paper and weighing. Any value they have here is merely comparative, as, apart from the basic nature of the deposits, these also contained much water which could not be removed by the simple treatment described.In experiments (3) to (s), it will be noticed that the current continually falls off, owing to the increasing resistance produced by the basic film. In ex- periments (6) to ( I I), on the other hand, current density was kept constant, and the experiment stopped when the bath voltage had reached the limit available for the electrolysis. In (S), the manganese dioxide precipitated at the anodes was collected, washed, ignited to Mn,O,, and weighed. I t cor- responded to a practically IOO per cent. current efficiency for the process Mn" + 2H,O + 2 @ + MnO, + 4H' A single experiment was done under similar conditions, using a cathode of copper. Only basic deposit was obtained, probably owing to the low hydiogen over-voltage at this metal. TABLE I. 4'4 11-5 9'1 +'4'5 11.8 3 5'5 10 3 1.4 1 '4 I0 I 0 I0 I0 I0 - - I - 5'6 3 7'8 2-8 3 7'7 4.7 + 16 5.2 4 16 4'4 --3 16 5*2+ 16 5.8 4 16 N.- - - - 0.07 0.025 0.018 0'071 0.078 0.032 0'172 -- Nature of Deposit. Dull black ; soluble in HCI, giving hydrogen. Heavier than (I) ; very much fouled. Much basic deposit. Some basic deposit. Deposit badly fouled. Deposit impure. Metallic in places-fouled elsewhere. Entirely basic material. Deposit badly fouled. Contained 4.2 per cent. Contained 15.8 per cent. metal. metal. SERIES II. The above experiments having shown that the avoidance of formation of a basic precipitate at the cathode was the first thing to look to, we modified the conditions (u) by giving the bath a certain acidity in advance (8) by stirring the electrolyte vigorously with a glass stirrer.As further, experiments (I) and ( I I) had indicated that 55' was the most favourable temperature of those worked at, this was adhered to throughout, and the current density kept constant at 10 amps./dm*. VOL. XIX-T22562 THE ELECTRODEPOSITION OF MANGANESE-PART I. 22 23 24 Number of Experiment, O c. 73 55 95 30 113 27'5 I2 I3 I4 '5 16 17 18 I9 20 21 TABLE 11. Duration of Electrolysis in Minutes. 35 37 I7 32 31 47 34 33 87 182 Initial Acidity. Neutral Neutral o.oo6o N. 0'0065 0'01 0'01 0'013 0.05 0.05 0.05 Vol tnge , 5'1 + I2 4-8 + 11.4 5'0 9 6.5 5'8 5'3 -3 I2 5'1 3 9 4-6 --* 8.6 4'6 3 5'3 5'2 --3 5'9 4'6 - 9'7 True Current Efficiency. Per cent. ? 4.4 34'0 65'4 12.9 4'8 35'5 27'3 33'3 2'1 Percentage of Metal in Deposit. Not determined 3 3 33'6 68.6 13.0 3'5 51.0 79'1 73'6 4'1 It should be noted that the current efficiencies in this table, as subse- quently, are true current efficiencies reckoned on metallic manganese, and determined by estimating the volume of hydrogen evolved on treatment of the cathodic deposit with very dilute mineral acid.I t is clear that the experimental inodifications introduced have improved the purity of the deposit. The results are very fluctuating, but on the whole the higher initial acidities correspond to the better deposits. These were black and dense, resembling gas carbon, and showed a metallic lustre when cut with a knife. The irreproducible nature of the results is accounted for by the fact of the production of acid at the anode, together with the difficulty of securing stirring of the same efficiency in the different experiments.In (21) an attempt was made to keep the acidity of the electrolyte constant by running in caustic soda solution during the experiment. I t was, however, unsuccessful, the bath becoming fouled with rapidly oxidising manganous hydroxide. In the other experiments, the acidity was determined at fifteen- minute intervals, the calculated volume of the electrolyte removed, and re- placed by neutral manganous sulphate solution. This is clearly not a perfect solution of the difficulty. SERIES 111. To improve the efficiency of stirring in the immediate neighbourhood of the cathode, a rotating cathode was introduced, and used in subsequent experiments. I t consisted of an aluminium rod, the lower part of which was split to admit of the insertion of a glass fin.The separate glass stirrer was dispensed with. At the same time, the effect of lower tempera- tures was tried. Current density was as in Series II., the acidity being regulated near 0.05 N. as described above. TABLE 111. Number of Duration in Temperature. Current Purity of Experiment. I Minutes. I 1 1 Efficiency. 1 Metal. _.____ 4'9 5'6 5'7 Per cent, 5'3 17.1 I4.3 Per cent. 88-3 98'7 98.6 A comparison of the results of (19) and ( 2 0 ) with those of (22) indicates that the introduction of the rotating cathode has improved the purity of theTHE ELECTRODEPOSITION OF MANGANESE-PART I. 563 Current Efficiency. deposit. I t will be noticed also that, as a result of the smaller production of basic material, it is now possible to keep the voltage practically constant throughout the run.Equally striking is the effect of dropping the tempera- ture further-there is a marked increase in the purity of the metal, and the current efficiency, very small in ( 2 2 ) , is improved. The deposits from (23) and (24) were dead black in appearance on removal from the bath, showed glistening white streaks on being scraped, and could be burnished with emery paper to give an appearance like that ofipolished iron. The differences in current efficiency in the two experiments are attributable to lack of exact control over the acidity factor. SERIES IV. Using the same apparatus, the effects of varying current density and initial acidity, and of a further lowering of temperature were tried suc- cessively.Purity of Metal. TABLE IV. Pa cent. 17.6 23'3 22'9 13.2 4'1 15'5 22.0 27.4 26'3 52'5 57'8 2'2 Number of Experiment 25 26 27 28 29 30 31 32 33 34 35 36 Pa cent. 53'6 87'6 94'3 91.2 81'5 1'00 85.5 60 98'3 98.4 9*3 98-3 Temperature. 30'5' C. 26'0 29.0 263 26'0 29.0 25'5 23.0 23 '5 5 '0 5 '0 22'0 Initial Acidity. 0.05 N 0'05 0.05 0 '05 005 0'10 0.075 0.025 0.025 0.025 0.025 0.025 Current densit] in amps./dmz. 20 15 12.5 7'5 5 20 20 I0 7'5 5 7'5 7'5 Duration in Minutes. Voltage. 8.0 6.6 5'9 5'0 4'5 7 '0 6.9 5'9 4'9 4'4 5'9 6.1 I- Small differences between the figures in this table have no significance, owing to the ill-regulated acidity. The following main points, however, emerge. An increase in current density will reduce the purity of the deposit, unless the acidity is correspondingly increased. A decrease in current density will lower the yield, unless the acidity is proportionately reduced.To high acidities, therefore, correspond good deposits and poor current efficiencies. Finally, a lowering of temperature allows of much better current efficiencies. The deposits from (35) and (36) were of excellent quality, and metallic in appearance. SERIES V. Hitherto not more than one gram of manganese had been produced in any one experiment. In order to make larger quantities, and at the same time to amve at a better solution of the fundamental question of regulation of acidity at the cathode, we proceeded to the design and trial of various fornis of apparatus in which the electrolyte of slightly acidified manganous sulphate was caused to flow continuously through the cell, entering at the cathode and leaving at the anode, being passed once more through the cell after its acidity had been corrected.The first forms used need not be described-they tailed because the rate of flow of electrolyte was too low, and basic material became precipitated on the cathode. A third modifica- tion allowed of far more rapid flow, and gave comparatively satisfactory results, a IOO per cent. pure metal being produced at 56.5 per cent. current564 THE ELECTRODEPOSITION OF MANGANESE-PART I. Current Density in amps./dm2. _~ I0 I0 efficiency, using an initial acidity of nearly o*IN, a temperature of 8" C., and a cathodic current density of KO amps./dm2. Unfortunately the rate of flow was so excessive that it was impracticable on a laboratory scale to test the acidity of the electrolyte and to correct the same by regulated additions of alkali in time for its return to the supply reservoir.Eventually a design of cell was adopted in which anolyte and catholyte were kept apart by a diaphragm, each flowing separately through the cell. The catholyte as before was a manganous sulphate soiution, with an added acidity of o*IN. The anolyte was the same manganous sulphate solution, but with an acidity of 0'2N. This increased acidity was used in the hope that hydrogen ion migration from anolyte to catholyte during electrolysis would compensate for hydrogen ion discharge at the cathode, and thus keep the acidity in the catholyte substantially constant, doing away with the necessity of correcting it by the addition of alkali. This hope was realised, it being found possible to carry out runs of fair duration with the minimum of alteration to the composition of the electrolyte. The cathode compartment consisted of a Soxhlet extraction thimble (7.5 cm. high and 3.3 cm.in diameter), the bottom of which fitted in a cup-shaped glass vessel, into which it was cemented by sodium silicate. A tube descended vertically from the centre of the base of the cup, and was provided with a tap. The extraction thimble was pierced at the bottom with a hole 5 mm. in diameter. Electrolyte introduced at the top of the cathode chamber would thus pass freely through at a rate essentially determined by the tap. A glass tube (4.5 cm. in diameter) was sealed on outside the vertical exit tube, a little distance below the base of the '' cup," so as to surround concentrically the extraction thimble. The annular space between this tube and the thimble constituted the anode chamber.The whole was surrounded by an inverted bottle with base removed, the outer annular space serving as a cooling bath. The rotating aluminium cathode and platinum sheet anodes were as before. The catholyte, introduced at the top of the thimble, passed through the cell at nine litres per hour. The anolyte was introduced by a narrow tube, passing down to the bottom of the anode compartment, and left by an overflow at the top. The results of two typical experiments only are recorded here. Its rate of passage was about one litre per hour.hrs. 12 2* TABLE V. per cent. I 6.3 47% IOO per cent. pure; but 6-3 26'5 1 xoopercent. pure; flaked 1 very loose. , excessively. O c. 37 1 6 Duration. Voltage. EkE$. l i ! Nature of Deposit. Although the electrolysis was easy to carry out, and furnished a pure metal, the main object of the experiments, i.e. the production of heavy coherent manganese deposits, was defeated. The drop in current efficiency in (38) compared with (37) is simply due to the fact that the yield was always estimated from the increase in weight of the cathode, and in (38) consider- able flaking took place after the lapse of about one hour, strips of metallic manganese being noticed floating in the electrolyte. This behaviour wasTHE ELECTRODEPOSITION OF MANGANESE-PART I. 565 invariably observed.Deposits up to a calculated thickness of 0.03 mm. were very smooth and adherent. At about 0-1 mm. they were already very loose, and a continuance of the electrolysis led to flaking. The same phenomenon was observed in a number of experiments (unrecorded) carried out at and above room temperature-it does not appear to be connected with the low temperatures used in (37) and (38). SERIES VI. At this stage in the work, we again modified the conditions of electrolysis. (I) We had no pump available for circulating the catholyte, and the cell just described could not therefore be left to itself for a long run. (2) I t seemed dsirable to avoid the use of manganese salt in the anode compartment, and thus the consequent formation of MnO,. (3) So-called " conducting salts " ( e g .sulphates of sodium, magnesium, and ammonium) are known to improve the nature of the deposit in a nickel-plating bath. Preliminary experiments with sodium sulphate had led to very bad deposits-we therefore settled on ammonium sulphate. The bath employed worked with a stationary electrolyte. I t consisted of a rectangular glass cell, containing the catholyte, in which were stood two porous pots containing the platinum anodes. The same rotating cathode was used as before. The catholyte contained per litre 300 grams of MnSO,, 4H20, IOO grams of (NH,),SO,(I.~N), and 2 - 5 grams H2SOl (o*ogN. ). By the regulated addition of strong H2S04, the acidity of the catholyte was kept as constant as possible. The results of two preliminary experiments with this apparatus are given in Table VI.Our reasons were as follows. The anolyte was 1a5N. (NH,),SO, + o*ogN. HTSO,. TABLE VI. Current hrs. I per cent. 39 14'0 I 0 4 15'8 100 per cent. pure; crys- talline. 40 13'5 I0 I 9 9 i The deposits, besides being pure, were adherent and markedly crystalline. At the commencement of (39) the catholyte was found to have become neutral before the first addition of sulphuric acid was made. No basic material, however, came down on the cathode. In view of this fact, and of the low current efficiency in (qo), we were induced to try experiments with the electrolyte as above, but without the addition of any acid Somewhat to our surprise, the results were satisfactory. The yield of metal rose, and the deposit remained pure, no precipitation of basic material taking place either in the electrolyte or on the cathode, although the liquid smelt of ammonia at the end of the run.The cause is of course the well-known action of ammonium salts in suppressing the ionisation of ammonium hydroxide. It is nevertheless somewhat remarkable that, using such a high current density, the metal deposit should remain pure. Experiments were then directed towards finding the optimum conditions for current density and temperature, the essential results being contained in Table VII.566 THE ELECTRODEPOSITION OF MANGANESE-PART I. Number of gxperiment. 41 42 43 44 45 46 48 47 49 remperature. O c. '4 15 '4 '3 30 30 30 51 7 TABLE VII. Current Densit in .IllPS./lL2. I 0 I0 'LO to 15 ' 5 20 10 20 Duration. Current 3tliciency 22'6 20% 25.8 40'7 35'6 6.6 40.1 5'0 - Nature of Deposit.100 per cent. pure ; crys- talline. 9 9 9 9 Much basic deposit. 100 per cent, pure ; crys- talline. 99 9 ) Some basic deposit, 100 per cent. pure ; loose and fine-grained. Save (44),* (48), and (49), all these runs yielded metal of excellent purity, and quite adherent. Thus (41) gave 5.1 grams of highly crystalline, very With regard to other points, much ozone is liberated at the anode when the electrolysis is carried out at room temperature and the small amount of manganese salt diffusing through the porous pots is oxidised to permanganate. At higher temperatures, ozone formation is slight, the diffusion of manganese sulphate increases, and there is considerable anodic MnOa formation. Some of the results given in Table VII.are rather puzzling and need further work. The best conditions of those investigated would, however, definitely appear to be a temperature of 30° and a cathodic current density of 10 amps. /dm.2. Using these optimum conditions, attempts were made to get heavier deposits of manganese, but without success. After a certain point had been reached, the deposits began to become loose and deterio- rated, just as was the case in the experiments of Series V. In view of the marked difference in the physical nature of the original deposit in the two cases, the reason for this behaviour is by no means clear. In order to investigate the cathodic potential during the manganese deposition, a few experiments were carried out using a stationary cathode, but stirring the catholyte vigorously.Provided the stirring were good, and the volume of the catholyte sufficiently large, a pure bright and coherent (but thin) deposit of metal was obtained, using 10 amps./dmz at the cathode. The current efficiencies were low. The cathode potentials were measured during and after deposition, using a normal calomel electrode and a Luggin capillary. w a q " deposit with a lustre like freshly cut bismuth. The following were the results obtained. (a) Temperature 14' C. (6) Temperature 18" C. E, = - 1.210 volt during deposition. Readings taken hourly during deposition up to a period of three hours. E, = - I -220 volt (remarkably constant). On cutting off current, E, fell to - 1.096 volt, from which value it altered very slowly.(c) A cathode covered with manganese some time previously was placed in the bath and gave a value of E, = - 0.796 volt. Current was passed for five minutes, when E, was found to be - 1.266 volt. On cutting off current, the potential fell to - 0.850 volt and remained sub- stantially unchanged for thirty minutes. Current passed for thirty minutes, when E, was - 1-322 volt. Current cut off, when the potential fell quickly to (6) Temperature 31" C.THE ELECTRODEPOSITION OF MANGANESE-PART I. 567 - 0.803 volt, and, on further stirring, to - 0,771 volt. Current was then passed for another thirty minutes, the value of E, reached being - 1.373 volt. On stopping the current, there was a rapid drop to - 0.858 volt, followed by a slower fall to - 0.832 volt, after which the rate of change became very slow.In all these cases, the potential of a manganese electrode prepared some time previously was practically the same in the compound electrolyte as in a solution simply containing the same amount of manganese sulphate, ie. E, = - 0.796 to - 0'798 volt. The figures thus afford clear evidence that a very considerable excess polarisation above the equilibrium value is necessary for the cathodic deposition of manganese, the metal behaving in this respect very like nickel. Increased hydrolysis makes it impossible to overcome this factor by increasing the temperature, as can be done in the case of the ferrous metals. The fact which renders possible the deposition of this very electropositiie metal is the high over-voltage it presents to hydrogen ion discharge.At a current density of 10 amps. /cm4. this amounts to 1.044 volt at 16" C., as determined in an electrolyte consisting of N/,oNaOH + N/l NaaOkl The potential of the reversible hydrogen electrode in the solution used for electrolysing out the manganese was found to be E, = - 0.166 volt at 18" C., and the cathodic potential necessary for hydrogen discharge under these conditions consequently - 1.210 volt. This is very close to the second of the values observed during the manganese electrodeposition, and is actually equal to the first of them. The reason for the simultaneous deposition of hydrogen and of manganese is thus clear. SERIES VII. We referred at the beginning of this paper to the work of van Arsdale and Meier.z These authors electrolysed a neutral molar manganous sul- phate solution at 23O C., varying the cathodic current density between about g to 43 amps./foot2 (I to 4-6 amps./dm2), and claimed current efficiencies of 73 - 89 per cent., the maximum figure being obtained at about 2 amps./dm2.In a second series of experiments they added in- creasing initial amounts of acid, working at a current density of 8 amps./ foot2, and found that the current efficiencies fell off rapidly and almost linearly, becoming zero at about 0.36 per cent. of H2S04 in the electrolyte. They state their deposits to have been dark grey and powdery in character, very readily oxidising in the air, but do not say how the degree of purity was established. The current efficiencies were determined by dissolving the cathodic deposit in dilute acid, and titrating.In view of our own results, it seemed highly improbable that anything but very impure metal was obtained under these conditions. We therefore repeated their work, using a stationary copper cathode and platinum anodes. The results are contained in Tables VIII. and IX. The current efficiencies are "ap- parent" values only, as in Table 1. The deposits were in all cases highly impure, as analysis figures and the high '' current efficiencies " in Table VIII. show. The effect of variations in current density and in acidity are as would be anticipated, and it would appear that van Arsdale and Meier could never have got anything but very impure material. Their conclusion that, at an acidity of about 0.36 per cent. H2S04, cathodic deposition entirely ceases is confirmed.Campbell, Tram. Cltent. SOL, 1923, 123, 0323. ' L O C . cit.568 THE ELECTRODEPOSITION OF MANGANESE-PART I- ____- 0.56 1-7 2.8 1'1 2'2 55 57 56 I& I I* 1 I TABLE VIII. e; a L? CI i? G O c. 17 17.5 IS 17'5 IS J M CI d P 2'6 3'3 3'7 4'1 4'9 TABLE IX. Per cent. H&04 0.1 13 0.196 0.319 Per cent. H2S04 0.132 0.270 0'407 Undetermined 0'0095 N 0'011 0.0245 0'0295 Y ." v) P 6 w >. .d ._ 2 a - Undetermined. 16'4 per cent. metal. 5-4 per cent. metal. 3'6 per cent. metal. Highly impure. I I 4 Y .- 0 a. a" c 0 x .- a a4 iPer cent./ 3. Exjerimen fs with Chloride SoZzitions. Unless otherwise stated, all experiments were carried out with solutions The salt containing 350 grams of the crystallised tetrahydrate per litre.was found to be free from detectable traces of iron. SERIES VIII. Preliminary experiments were carried out at an early stage in this work, before the disturbing effects of high temperatures were realised, under similar conditions to those described under Series I. for sulphate electrolysis. The results obtained were in many respects like those already given in Table I, and need not be detailed here. There were, however, certain striking differences. The deposits, although impure, were generally denser and more adherent that those from sulphate solutions, and showed no sign of white basic material, being, on the contrary, dead black in appearance. Further, the voltage, instead of rising gradually to very high figures owing to increased resistance at the cathode, remained constant throughout the runs at values, depending on current density and on temperatu're, varying between 2-6 and 4-2 volts, figures some 2 to 2-5 volts lower than the lowest rewrded with sulphate solutions under similar conditions.The fact that the impurity, whatever its nature, was conducting metallically, together with its colour, pointed to its being manganese dioxide. Subsequent work, already described,' made it clear that manganous chloride can readily be anodically oxidised at platinum anodes to manganese tetrachloride. This Campbell, Tratrs. C h m . Sot., 1923, 123, 892.THE ELECTRODEPOSITION OF MANGANESE-PART I. 569 being the case, the mechanism of manganese dioxide precipitation at the cathode is an obvious one. Although with the more dilute manganous chloride solutions used in the experiments here described, a large amount of MnO, is precipitated at the anodes as in the sulphate experiments, enough of the tetrachloride nevertheless is carried over to the cathode to be precipitated there in considerable quantity, as a consequence of the local impoverishment of hydrogen ions.SERIES IX. Some experiments were carried out with the continuous Aow apparatus described under Series V. Using the same acidities as with the sulphate solution, viz., catholyte 0-1 N and anolyte 0.2 N, and keeping the level of the catholyte above that of the anolyte to counteract any tendency for per- chloride to diffuse into the former, it was found that no deposit of any kind was obtained. The acidities in both the cell liquids were accordingly halved, and over short runs, smooth and brilliant deposits were got, at current efficiencies of almost 70 per cent.These deposits were, however, not quite pure, averaging 97-5 per cent. of metal. The apparatus was then somewhat modified, by substituting a porous pot with a rubber stopper in the bottom for the Soxhlet thimble cemented in its glass base, and doing away with the external water bath. It was hoped that the increased dia- phragm resistance would in large measure prevent the Mn"" ions passing into the catholyte, and a run of 24 hours was carried out. The results were disappointing. A pure metal was certainly got, but the current effi- ciency was low, and there was a very marked tendency to flaking after half an hour. SERIES X.We have carried out a few experiments on the lines indicated by him, in order better to compare his results with ours. We invariably found that when we used manganous chloride in our anode compartment as he does, tetrachloride of manganese is formed, resulting in, sooner or later, the deposition of manganese dioxide on the cathode. I t may be remarked that the nature of Grube's anodes is not stated. Ours, as always, were of platinum. If, on the other hand, we used an apparatus in which the catholyte was of comparatively large bulk, and of composition 6N. MnCl, + 1.5N. NH,CI, whilst the anodes were contained in small porous pots dipping in the catholyte, the anolyte being simply I . ~ N . NH,Cl, we were able to obtain pure metallic manganese, the conditions of temperature, current density, and stirring being as indicated by Grube.Thus our first experiment on these lines lasted for five and a half hours, and gave us IOO per cent. metal at a 55 per cent. current efficiency. The physical character of the deposit was, however, if anything, worse than that of the manganese prepared from sulphate solutions-nodular and very loose. I t may be noted in connection with these experiments that we did not employ the addition of HC1 recommended by Grube-our experiments under Series VI. appeared to us to show that it would serve no useful purpose. Further, we lay considerable stress on the necessity for a large bulk of catholyte or small cathodic current concentration. If this is not arranged for, it is quite likely that manganous hydroxide will be precipitated in the catholyte during the course of a long run.In the introductory section, the work of Grube is referred to.5 70 THE ELECTRODEPOSITION OF MANGANESE-PART I. 4. Attempts to improve ihe naiure of ih dejosii. A large number of experiments have been carried out with this aim in view, without, however, any striking success. (a) A variety of addition agents were tried, using the standard manganese sulphate and ammonium sulphate electrolyte. None produced any marked improvement, and in several cases (e.6 gum arabic, dextrine, gelatin) the deposit was made much less pure. (6) With the same electrolyte, an experiment was done, making use of a burnisher (an ebonite strip) pressed against the rotating cathode. The deposit obtained was smooth and dense, but the current efficiency was reduced to 8.6 per cent. (c) Other electrolytes were tried. The use of sulphocyanide (see Moore I and Smith and Franke12) gave a more coherent but less pure deposit. The use of a mixture of manganese and ammonium perchlorates also led to nothing, as the whole catholyte hydrolysed with great rapidity, becoming filled with manganous hydroxide. (d) I t was thought that the flaking described in Series V. might have been due to the presence of the trace of iron mentioned as being present in the manganous sulphate used. This was removed by the addition of ammonia (its absence being subsequently shown by the ferrocyanide test), but no improvement in the nature of the deposit was subsequently noted. (e) I t may be mentioned in conclusion that it is a matter of common knowledge that the simultaneous evolution of large amounts of hydrogen renders it very difficult to get good cathodic deposits of any metal. The work of Kohlschutter and has pupils6 connects this difficulty with the strains caused in the electrodeposited layers by the presence or tne hydrogen. Attempts were made to see whether such depolarisers as were found by Stager to be successful in the case of nickel could also be used in manganese electrodeposition. Hydrogen peroxide, potassium chlorate, nitrobenzene and cinnamic acid were all employed, but with negative results. 5. Summary. (I) The electrodeposition of manganese from aqueous solutions of its (2) The effects of changes in composition of electrolyte, current (3) Pure manganese in coherent form can be prepared in small quantity (4) Attempts to prepare larger amounts in coherent form were sulphate and chloride has been studied. density, temperature, and type of cell have been investigated. with a current efficiency of 40-50 per cent. unsuccessful. This work was commenced in the summer of I 92 I and finished about Further experiments are being carried out on some of Christmas, 1922. the points mentioned in the paper. University of London, King’s Co Zlege, June, 1923. Loc. cit. * LOC. cit. J Tried in view of the favourable results obtained by Mathers in depositing lead and Cf. Engemann. Kohlschiitter and VuilIeumier. Zeitsch. Elekfrocitem., 1918, 3, 300. Stager. other metals. Zeitsch. Elektrochem., 19x1, 17, gro. Kohlschiitter. Helv. China. Acta, 1920, 3, 584. Helv. Chim. Acta, 1920,3, 614.