124 On Alternating Current Electrolysis .
By S. Mahsh , B.Sc. , Ph. D. , formerly Fellow of the University of Wales , Head of Department of Physics , Battersea Polytechnic , London .
( Communicated by E. H. Griffiths , Sc. D. , F.R.S. Received October 21 , 1919 .
) In the summer of 1915 , Mr. E. Meigh , M.Sc .
, and the author , were engaged in some research dealing with the behaviour of bacteria under the action of electric fields .
No definite conclusions were formed concerning the action of alternating electrostatic fields , and it was decided , therefore , to examine the effect of an alternating electric current on bacteria .
It was noticed , however , that small bubbles of gas were formed at the electrodes , and , although these bubbles were not in the field of view of the microscope , nevertheless , they gave rise to a displacement of the volume of liquid under examination , this displacement masking completely any changes that might have occurred in the movement of the bacteria .
This led to an investigation of the conditions under which such bubbles were formed , and , in the end , to a detailed examination by the author of some of the phenomena of alternating current electrolysis .
S 1 .
Introductory .
The subject of alternating current electrolysis has been investigated by Le Blanc and Schick , * Ruer/ f* Brocket and Petit , ] : Maneuvrier and Chappuis , S among others .
Most of the investigations deal with the passage of alternating current through solutions of cyanides , chlorides , etc. , electrodes of different metals being employed , and they are concerned principally with the question of the solution of the electrodes and the chemical changes occurring in the electrolytes .
The experiments of Ruer , Brochet and Petit , and Maneuvrier and Chappuis , deserve further notice .
Ruer ( loc. cit. ) used platinum electrodes in dilute sulphuric acid .
Three electrodes were employed , A , B , C ; alternating current was passed between A and B , and , simultaneously , a direct current was passed between C and A and B , the two latter connected together and serving as anode .
He stated that platinum was not dissolved by .alternating current alone , but dissolved when the electrodes between which the alternating current is passing are subjected to anodic polarisation .
He suggested that , if the platinum anode * Le Blanc and Schick , ' Journ. Ckem .
Soc./ vol. 2 , p. 229 ( 1904 ) .
t Ruer , ' Journ. Chem. Soc./ vol. 2 , p. 407 ( 1903 ) .
| Brochet and Petit , ' Ann. Ohiin .
Phys./ vol. 3 , p. 453 ( 1904 ) .
S Maneuvrier and Chappuis , 'Comptes Rendus/ vol. 106 , pp. 499 , 1719 ; vol. 107 , p. 31 ( 1888 ) .
i On Alternating Current Electrolysis .
125 is covered with a thin film of peroxide , then the hydrogen evolved in one half of a period would reduce the peroxide to the oxide , which would be readily acted upon by the acid .
This view was contested by Brochet and Petit ( loc. cit. ) , who maintained that the solution of the platinum is due to variation in the current density , and not to any specific action of the alternating current .
According to them , a sinuoidal alternating current from a special valve gives the same effect as the combination of alternating and direct currents .
Ruer criticised these experiments , and offered further evidence in support of this theory .
Maneuvrier and Chappuis made qualitative investigations on alternating current electrolysis , dealing more especially with the effect of variation of current density and current frequency upon the production of visible electrolysis .
Their results will be referred to in a later section of this paper .
S 2 .
Earlier Experiments Meigli and Marsh .
The electrolyte was contained in a wide tube 5 cm .
diameter , fitted with three side tubes ( fig. 1 ) .
Of these , X and Y were graduated burettes , EF- 36cm .
AD*20cm .
A B = lieu .
( approx ) C D * 0*5 c m.(apprcx ) .
while Z was a feed tube , through which electrolyte was poured into OF .
the ends E and i were closed with rubber stoppers , each pierced with three holes , through which passed tubes holding the electrodes A and I ) , Dr. S. Marsh .
explorers B and C , and also thermometers ( not shown ) .
The electrodes , consisted of thin discs of platinum approximately 1 cm .
diameter , which were attached to short pieces of platinum wire sealed through the ends of the tubes .
These tubes were filled with mercury , and the ends of the supply leads placed in them .
In order to measure the potential drop in the neighbourhood of the electrodes during the experiment , explorers were introduced .
Each consisted of a very line platinum wire , which projected about a millimetre from a fine glass tube , enclosing and thus insulating the wire ( with the exception of the free end ) from the liquid .
Contact was made by means of mercury , as before .
Alternating current was supplied from a small Westinghouse motor generator set , giving 105 volts at 40 cycles .
The tube OF was immersed in a trough , through which cold water circulated , and in this way the temperature was maintained fairly constant .
It was soon observed that the rate of gas evolution was by no means uniform , but fell off rapidly , and finally ceased altogether .
Simultaneously , the platinum changed from its usual appearance to a dull black or greyish black .
It appeared that this was the normal procedure with all specimens of platinum used under the conditions of the experiment .
Experiments were made with a view to the restoration of such used platinum to its original state as regards gas evolution .
To this end a third electrode of platinum was introduced into OF down the tube Z , A and D were connected together , and a direct current was passed between A and Dr acting as anode , and the middle electrode , which served as cathode .
This direct current was allowed to pass for different times , and , after each such passage , alternating current was passed through until no further evolution of gas took place .
The volume of gas given off by the alternating current was to some extent an indication of the degree of recovery of the electrode due to the passing of direct current .
During the passage of direct current , the electrodes A and D were surrounded by nascent oxygen , and it was observed that the degree of recovery of the electrodes depended in some measure upon the duration of the direct current .
The actual details , and also the values of the potential distribution in the electrolyte , will be reserved for a later paragraph .
S 3 .
Later Experiments by Marsh .
More systematic and detailed experiments have been carried out by the writer .
It seemed highly probable that the phenomena would depend on the frequency of alternation , and , in order to investigate this point , a Crompton machine was used , in which the current frequency could be varied between On Alternating Current Electrolysis .
127 25 and 80 per second .
The frequency was read on Frahm and Hartman and Braun meters .
The current was maintained at 1 ampere throughout the series .
( Transformers were used when the machine was running at low speeds , in order to supply the voltage necessary to drive a current of 1 ampere through the electrolyte .
) Several series of experiments have been carried out with apparatus similar in type to that already described ; thick platinum foil , thin platinum foil , gold foil , nickel foil being used as electrodes , and dilute sulphuric acid and barium hydrate solution as electrolytes .
The electrodes had practically the same area in all cases .
In carrying out a series of experiments at different frequencies , sufficient electrolyte was made for the complete set .
In the case of the acid , the strength of the solution was approximately normal .
The left electrode ( consisting of thin platinum ) was used throughout , and no evolution of gas occurred there throughout the series .
Fresh acid and a fresh right electrode were used for each frequency test .
The tube OF was cooled by immersion in a trough of cold water .
The volumes of gas given off were reduced to 0 ' C. and 76 cm .
of mercury pressure .
S 4 .
Experiments with Thick and Thin Platinum Foil in Dilute Sulphuric Acid .
The curves showing the relation between the volume of gas given off and the time for different frequencies are given in fig. 2 .
Their striking resemblance to the saturation curves in radioactivity will be noticed , as also the outstanding effect of variation of frequency upon the total volume evolved .
Curves of similar type were obtained with thin platinum foil , the volumes given off , however , being less than in the corresponding cases with thick foil .
The thicknesses of the thick and thin foils were 021 mm. and Oil mm. respectively .
The volumes given off from the thin foil were approximately half those from the corresponding experiments with thick foil .
Later experiments , with different thick foil , showed that although more is evolved , in general , from thick than thin foil , yet the relations between the volumes given off were not of such a simple nature as the above .
In fig. 3 are shown the relations between the total volumes liberated and the corresponding frequencies for both thick and thin foil .
On plotting log volume against log frequency we get two approximately straight and parallel lines .
This would represent a relation of the form In the values of ( 3 and C for thick foil are 2-6 and 1-406 x 105 respectively ; for thin foil 2"6 and 7-889 x 104 .
The following Table shows the good agreement between the observed volumes and those calculated from the above formulae .
VOL. xcvii.\#151 ; a. K Dr. S. Marsh , 34 " THICK PLATINUM FOIL .
DILUTE ACID .
n =25 20- uj Fig 2 .
n=48 n=-60 TIME\#151 ; ^MINUTES .
Table I. Thick Foil .
Thin Foil .
Y " .
M2-6 = 1 -405 X 105 .
V , " .
\#187 ; 2'6 = 7 -889 x 104 .
Frequency n. Calculated volume .
Observed volume .
n. Calculated volume .
Observed volume .
25 c.c. 32 -6 c.c. 32-0 25 c.c. 18-28 c.c. 18-1 30 5-31 18-66 18-4 30-5 10-91 10 -9 39 5 9-89 9 -7 39 5 -75 5-6 48 5-98 7 1 48 3-36 3-6 60 3-37 3-54 60 1 -89 1 -8 80 1 -58 1 -1 On Alternating Current Electrolysis .
1-8 LOG FREQUENCY VOLUME FREQUENCY CURVES XTHICK PLATINUM OTHIN PLATINUM LOG VOLUME , LOG FREQUENCY CURVES 0 THICK PLATINUM \#174 ; THIN PLATINUM Fig 3 .
S 5 .
Effect of Variation of Current Density upon Evolution of Gas .
The machine was run at 305 cycles per second , and a series of experiments made with electrodes of thick platinum foil of different areas , the current being 1 ampere in each case .
Four electrodes were employed , their areas decreasing approximately in geometric progression .
The largest electrode was that already used at 30*5 cycles in the experiments described in S 4 .
As was expected , there was a much greater evolution of gas with the smaller electrodes .
The following Table shows the values for the different electrodes:\#151 ; Dr. S. Marsh .
Table II .
Electrode .
Area ( one face only ) .
Gas evolved ( N.T.P. ) .
sq .
cm .
c.c. I 1-298 18 -4 II 0-608 58 -4 III 0 323 126 -2 IV 0-162 341 *0 approx. It should be noted that gas is still coming off , although in minute quantities , even although the current has passed for a long time , e.g.:\#151 ; With electrode II the gas evolution is O'Ol c.c. per minute after t = 120 ' .
HI " 0 02 " II MO \#163 ; \#163 ; 170 ' .
IV " 0-05 " II 470 ' , .79 --69 -59 \#151 ; 49 -39 \#151 ; 29 --I9 -.09 + \#187 ; 01 -MI LOG AREA THICK PLATINUM FOIL .
DILUTE ACID X VOLUME-AREA CURVE O LOG .
VOLUME-LOG AREA CURVE .
FREQUENCY CONSTANT 30-5-31 .
Fig 4 ELECTROOE AREA ( ONE FACE ONLY)+SQ CM .
I On Alternating Current Electrolysis .
It will be seen from reasons advanced in S 11 that it is doubtful whether* in , say , case IV , the rate of gas evolution would ever be reduced to zero , however long the current be passed .
In fig. 4 the total volume given off is plotted against the area of the electrodes , and it will be seen that the points lie very well on a smooth curve .
On the same diagram , log volume is plotted against log area , and the result is approximately a straight line .
The relation between volume and area is of the form V.S " = b , where S = area ; \#171 ; = 1*403 , b = 26*55 .
An experiment was made with the same current density , but with a smaller THICK PLATINUM .
n= 30*5 CURRENT DENSITY SAME IN BOTH CASES .
I.AREA 1-298 sq.c.m .
CURRENT5 !
AMPERE VAREA -6401 sq.c.m .
CURRENT=493AMPERE THE RATIO V IS VERY NEARLY CONSTANT FOR ALL VALUES OfY TIME\#151 ; ^-MINUTES .
Dr. S. Marsh .
electrode , the frequency being 30*5 .
Full details are given in the following Table and the curves shown in fig. 5:\#151 ; Table III .
Electrode .
Area ( on face ) .
Relative areas .
Current .
Volume of gas evolved .
sq .
cm .
ampere .
c.c. I 1*298 i 1 18 *4 V 0*6401 0-4932 0 *4932 9*69 The ratio between the total volumes as also between the volumes evolved at intermediate times is practically constant and approximately equal to the current ratio .
Hence we infer that the actions taking place at I are occurring in the same degree and manner at V , and that for equal current density the total volume of gas liberated by an electrode is proportional to its .
area , frequency being constant throughout .
It should be observed that the current density is not constant over the electrode , but is greatest at the edges .
Hence halving , say , electrode area and current does not give exactly the same current density as before .
S 6 .
Experiments with Gold and Sulphuric Acid .
The results of these experiments are shown in fig. 6 .
The rate of evolution falls off rapidly with time , but for frequencies less than 48 gas is still given off at the electrodes for a much longer time than with platinum .
The volumes of gas given off are considerably smaller , in general , than in the case of platinum .
The surface of the gold becomes curiously speckled or mottled by the passage of the current .
The current density was also varied by using electrodes of different areas , the current being maintained at 1 ampere .
The results are given in the following Table:\#151 ; Table IV .
I Frequency 39 *5 .
Full size electrode .
Half size electrode .
Quarter size electrode .
Q-as evolved C.C. 5 -11 0 *01 after t = 50 ' c.c. 16*9 0 *065 after t = 70 ' c.c. 82 *4 0 *13 after t = 156 ' j Rate of evolution of gas in c.c. per minute I On Alternating Current Electrolysis .
n = 25 Fig 6 .
TIME\#151 ; ^ MINUTES S 7 .
Experiments with Nickel and Dilute Acid .
Area 1'43 sq .
cm .
, frequency 39*5 .
A copious evolution of gas took place , the ( uncorrected ) rate per minute being 3T c.c. for the first 20 minutes , after which the electrode fell off owing to dissolving in the acid .
The evolved gas was passed into the burette for absorption of oxygen , but after several days the volume had only diminished from 69-8 c.c. to 69-4 c.c. showing that there was practically no oxygen present and the gas entirely hydrogen .
The theoretical rate of production of hydrogen is 3T4 c.c. per minute , which agrees remarkably well with the rate observed above .
S 8 .
Experiments with Barium Hydrate as Electrolyte .
As somewhat complicated reactions occur when sulphuric acid is used as the electrolyte , it was thought desirable to use an electrolyte of entirely different nature .
A solution of barium hydrate of strength approximately N/ 10 was made up .
Owing to the high specific resistance of the solution , it was necessary to use a two step-up transformer at the lower frequencies .
There was considerable Dr. S. Marsh .
Rate of evolr .
43 c.c.jmin .
THrCK PLATINUM BARIUM HYDRATE n * 31*5 Rate ofevol " .125 c.c./ m\lt ; n Rate ofevol " .
044 c.c.jmin TIME\#151 ; ^MINUTES .
heating in the tube , and this was the more marked in the case of the large machine , owing to the impossibility of maintaining a constant flow of cold water round the tube ; ( with the small machine it was possible to circulate water round the tube ) .
The solution was boiled for some minutes before using , in order to drive out as much as possible of the dissolved air .
Experiments , similar in nature to those carried out with acid , were made using electrodes of gold , platinum and nickel foil .
The results are shown in figs. 7 and 8 .
It will be noticed that there are no signs of the flat saturation type of curve obtained when acid is used .
On the contrary , the curves seem to show that , after a long period of time , gas is evolved at a steady rate , the rate decreasing with increasing frequency .
The volumes liberated in the same time are much greater with barium hydrate as electrolyte than with acid .
The curves for platinum and gold are of a similar type .
With nickel electrodes at a frequency of 60 cycles only the minutest bubbles were given off .
On reducing the speed to 40 cycles a lieu more bubbles were given off , but the volume was far too small to measure .
VOLUME\#151 ; ?C.C .
210- \#187 ; 0 eo so On Alternating Current Electrolysis .
gold foil , barium hydrate .
Ral-e of evol " i-35c.c./ mm. Rate of evoln .74c.c/ min. Rate of evoln '65 c.c/ min. Rate of avoloHon *048 c.c./ min. TIME\#151 ; *-MINUTES S 9 .
Critical Discussion of the Results with Theoretical Suggestions .
The effects previously described are brought about by two actions : ( a ) changes occurring at and inside the electrode , ( 6 ) changes taking place in the electrolyte .
It is fairly easy to account in general terms for the various phenomena .
There is a steady production of gas at each electrode of 471 c.c. per minute due to the passage of the current .
We can imagine that some of the gas liberated in any single half period is kept at the electrode and is then attacked by gas liberated in the succeeding half period , water being re-formed .
Hence the total volume actually liberated per minute is considerably less than the theoretical volume .
The experiments show that in all cases- the activity or power of the electrode in promoting re-combination increases with time for a definite frequency , and increases very rapidly with frequency .
Now the process , in the case of platinum , may be regarded from two different standpoints:\#151 ; i. Hydrogen is easily absorbed by platinum .
Hence , of the hydrogen produced in one half period at one electrode , some will escape and some will be absorbed by the electrode .
Of the oxygen produced at this electrode Dr. S. Marsh .
during the succeeding half period , some will escape , some will attack the hydrogen in the electrode , re-forming water , some undoubtedly will react with the electrolyte .
The re-combination may be such as to remove all the absorbed hydrogen from the electrode , although this does not appear to be very likely .
In any case , the result would be a change in the quality of the surface , which would alter from a coherent compact nature to one of a porous type due to absorption and re-combination .
This , in effect , would result in increased area of surface of the electrode , which would allow of increased activity both of absorption and re-combination .
As a result , the rate of gas evolution would fall off very rapidly .
As a variation of the above process , it might be supposed that that portion of the hydrogen which remains at the electrode does not remain at the surface to be wholly and entirely removed by the oxygen of the succeeding half period , but diffuses into the electrode in accordance with the laws of diffusion .
The curve AB in the accompanying figure , fig. 9 , represents Fig 9 .
Depth into electrode .
qualitatively the distribution of the hydrogen in the electrode .
The oxygen is able to combine with a portion only of this hydrogen { e.g. , that to the left of CD , since the oxygen is not able to diffuse as quickly as hydrogen ) , and as a result , the electrode accumulates a store of hydrogen .
The hydrogen of succeeding half periods is not able to diffuse so rapidly into the electrode , owing to the gas already present there .
Hence , more and more oxygen is removed On Alternating Current Electrolysis .
137 by re-combination .
According to this view , it would appear that oxygen would be in excess in the evolved gases .
Careful tests of the composition of those gases indicate that oxygen is , if anything , in deficit .
It must be pointed out , however , that the effect of the electrolyte has been ignored up to the present .
In this connection it is of interest to note the results of experiments which were carried out by Meigh and Marsh with a view to the restoration of the " used " platinum electrodes to their original state as regards evolution of gases when again employed with alternating current .
These " used " electrodes were made to act as anodes with direct current for different times , and they were brought into intimate contact with active oxygen .
When used subsequently as electrodes with alternating current , they permitted the evolution of gas as with new electrodes , but never to the same extent .
On the other hand , such electrodes , when made to act as cathodes with direct current , do not regain this power , and no evolution of gas takes place at them when they are used subsequently with alternating current .
These results seem to lend support to the theory that the inability of the electrodes to evolve gases with alternating current is due to the hydrogen which they have adsorbed and absorbed .
Such hydrogen would be removed , in part , at any rate , when the electrode is made to act as anode with direct current .
ii .
The process may , however , be regarded as taking place in a different manner .
We may conceive of chemical changes occurring at the electrodes during the passage of the alternating current .
The electrode may get oxidised to some oxide of platinum by the oxygen of one-half period , this oxide being reduced to a lower oxide or to the metal itself by the hydrogen of the succeeding half-period .
Such oxidation and reduction would alter the constitution of the surface , making it of a more porous nature , thus exposing a greater area to the subsequent oxidising and reducing actions .
The experiments on gold and nickel may throw light upon this point .
When gold electrodes ( which have been used already with alternating current ) are made to serve as anodes with direct current , the electrolyte being dilute acid , they recover ( partially ) their ability to evolve gases with alternating current .
During the passage of the direct current the cathode became brighter , while a dark stain , with traces of red , appeared round the edges of the anode .
( Both electrodes initially had the mottled appearance already remarked upon .
) When this direct current is suddenly reversed in direction , bubbles are given off at once at that electrode which had previously acted as cathode , while at the other electrode ( viz. , that serving previously as the anode ) no evolution of gas took place for several seconds\#151 ; in one case 25 seconds , in another 8^ seconds .
Colour changes also took Dr. S. Marsh .
place .
The electrode acting now as anode became brown , while the other changed from reddish brown to a lighter , almost greenish , shade .
These alterations of colour are probably indicative of chemical changes at the surface of the electrode .
The metal is oxidised at the anode , and , on subsequent reversal of the direct current , this oxide is reduced to a lower oxide or to the metal , by the hydrogen .
Until this action is complete , no gas is evolved at the electrode .
With nickel , the results are of a different nature .
When nickel electrodes are used in dilute acid with direct current , then , if the current is small enough , no evolution of gas occurs at the anode , while there is a steady production of gas at the cathode .
The latter remains bright , while the former becomes dull .
On reversal of the direct current , gas is evolved at once at the new cathode .
In this case the nickel oxidises , but the oxide formed is very soluble in acid , so that , on reversal , there is no oxide remaining for reduction by the hydrogen .
This explains at once the effect observed when nickel is used with alternating current in dilute acid .
The oxygen of one-half period oxidises the nickel , and the oxide thus formed dissolves in the acid .
Hence little , if any , of the hydrogen is used for reducing purposes , and the evolution of the hydrogen is quite regular , and agrees with the theoretical value .
The gold oxide formed is not soluble in the dilute acid , and hence reduction takes place .
Striking changes occur when nickel electrodes are used in barium hydrate solution with direct current .
When a small current is employed , bubbles are given off ' at the cathode only .
Dark streaks and dark-coloured edges appear at the anode , and , with time , a greenish deposit forms on the anode .
On reversing the current , the dark stain appears very quickly on that electrode previously serving as cathode , and is soon followed by the green deposit .
It should be remarked that the dark stain spreads much more rapidly over the electrode when this has just previously been acting as cathode .
The dark stain on the nickel dissolves at once in dilute sulphuric acid ; the stain on the gold , when used with barium hydrate , does not dissolve in this acid .
It would appear from these facts that , in the case of gold and nickel , there is little doubt but that oxidation occurs in the process of electrolysis .
Further evidence in support of the " oxidation and reduction " theory is obtained from experiments made by Maneuvrier and Chappuis .
They observed that , when copper electrodes were used with dilute sulphuric acid , and alternating current passed , no visible electrolysis set in .
The electrodes became covered with a coating of red reduced copper , caused by On Alternating Current Electrolysis .
oxidation by the ozonised oxygen , coupled with subsequent reduction by hydrogen .
Using mercury electrodes , of large surface , they discovered that a small volume only of gas is evolved ( nearly all hydrogen ) , and that a cloud of white matter is produced , which spreads gradually through the electrolyte .
They showed this to be the sulphate of the protoxide .
We have next to consider whether the oxidation theory is capable of explaining ( a ) the different volumes of gases liberated at different frequencies ; ( b ) the partial recovery of the electrode when made to act as anode in direct current electrolysis .
In the latter case , we may suppose that the surface of the metal gets oxidised , and , further , that this oxidised film becomes more and more coherent the greater the duration of this electrolysis .
The speed of the reducing action of the hydrogen on this film will depend very much on the surface state of the latter , being greater when this film is of a porous nature .
Hence , with subsequent alternating current , the efficiency of the reducing action is small at the start , and for some time evolution of gases occurs at the electrode .
An explanation of the frequency effect can be offered on the same general grounds .
At high frequency , the oxidising action in one-half period persists for a short time only .
The film formed is not of a close texture , and is readily reduced by the hydrogen in the succeeding halfperiod .
With low frequencies , on the other hand , the oxidation period is longer , the film formed is more coherent and massive , and the reducing action not so speedy .
A study of the curves for gold and platinum shows that the same general tendencies are present in both cases .
One would be led to infer that oxidation must play an important part in the case of platinum .
Again , if the used electrodes , gold and platinum , be strongly heated , they recover partially their ability to give off gases when used with alternating current .
The gold electrodes , on heating , became quite bright , while the platinum changed from greyish black to a dull silver colour .
The films of oxide would break up on heating , giving a pure metal surface .
With regard to the oxidation of platinum , several investigations may be quoted .
The experiments of Ruer ( loc. cit. ) have already been alluded to .
The experiments of Mond , Ramsay , and Shields , * on the absorption of oxygen ( and hydrogen ) by platinum black , point to the formation of Pt(OH)2 , since the heat of occlusion of the oxygen is the same as the heat of formation of Pt(OH)2 .
* Mond , Ramsay and Shields , 'Roy .
Soc. Proc. , ' vol. 58 , *p .
242 ( 1895 ) ; vol. 62 , p. 50 Dr. S. Marsh .
Newberry , * in a series of measurements of cathodic and anodic overvoltages of different metals in acid and alkaline electrolytes , suggests that anodic over-voltage is due to the formation of higher oxides on the surface of the electrodes , such oxides being only stable under the influence of the great pressure set up in the electrolysis .
It is not at all improbable that the tw 'processes of occlusion and , which have been outlined in the preceding pages , may be taking place simultaneously in the experiments which have been described .
In fact , Mond , Ramsay , and Shields showed that platinum black , which had absorbed 100 volumes of oxygen , was able to absorb 310 volumes of hydrogen , 200 of which combined with the oxygen , the remainder being occluded .
They concluded that , in platinum black , a certain amount of occluded oxygen could co-exist with occluded hydrogen .
It would not be unreasonable to suppose that occlusion would play a more prominent part in the case of platinum than with gold or nickel , but one is led to believe that oxidation also plays an important role in these experiments .
S10 .
Influence of the Electrolyte upon the Phenomena .
It is evident that the nature of the electrolyte plays an important part in these experiments .
With the current densities employed there was no trace , when barium hydrate was used as the electrolyte , of the flat " saturation type " of curve obtained when sulphuric acid served as the electrolyte .
The activity of the electrode is much less in barium hydrate than in acid .
One noteworthy fact deserves mention .
If the mixed gases produced by electrolysis of the acid are allowed to stand for some time in the tubes over the acid , it is found that the volume diminishes by a considerable amount .
With barium hydrate , on the other hand , there is little or no change of volume .
It will be seen that when Ba ( OH)2 is the electrolyte , there is a much greater evolution of gas with gold electrodes than with platinum electrodes .
With acid , on the other hand , the opposite is the case .
Further , a variation in the current density had a very marked effect on the volume of gas liberated .
Gold electrode and Ba(OH)2 .
Current frequency 60 .
Current 1*0 ampere , rate of evolution per minute 0-80 c.c. " 1*5 " " " 175 c.c. " 1*8 j , , , 2'5 c.c. * Newberry , ' Journ. Chem. Soc. Trans. , ' p. 1420 ( 1914 ) ; p. 1051 ( 1916 ) .
On Alternating Current Electrolysis .
S 11 .
Activity of the Electrode .
It has been noted earlier that in no case is the theoretical quantity of gas evolved at an electrode , even at the commencement of electrolysis .
Each electrode has , per unit area , a certain power of promoting re-combination of the gases t = o ' , this power increasing with the frequency of alternation .
Furthermore , this power increases with the time of passage of the current until finally , with platinum and acid , and in some cases with gold and acid , the activity of the electrode becomes so great that no further evolution of gas occurs .
It appears from an analysis of the results that this activity can increase to a maximum per unit of area , the value of the maximum increasing with frequency .
If the current density is less than that which would call forth the maximum activity of the electrode , then sooner or later the evolution of gas will cease .
If , on the other hand , the current density is higher than that corresponding to maximum activity , then the gas after a considerable time will continue to come off , but at a uniform rate .
Figs. 2 and 7 show these differences very clearly .
The current density with platinum and sulphuric acid was less than that which would call forth the -maximum activity , and so the curves soon run horizontally ; but the current density with platinum and barium hydrate was greater than that representing the maximum activity , and as a result the curves become straight lines whose slopes decrease with increasing frequencies .
Maneuvrier and Chappuis made qualitative experiments on alternating current electrolysis , and came to somewhat similar conclusions regarding the activity of the electrodes .
They did not observe , however , the important fact that for a given metal and frequency the activity is a function of the time .
S 12 .
Initial Activity of Different Electrodes .
A few experiments were made on the initial activity of different electrodes .
The machine was run at a definite speed and the current gradually increased until visible electrolysis set in .
Frequency 41-5.\#151 ; With two pieces of new thin platinum foil of the same area as that previously employed , the critical initial currents for acid and Ba(OH)2 respectively were 0*29 ampere and OT ampere , indicating a much higher activity on the part of the electrode when in acid than when in hydrate .
With gold ( heated after use with alternating current ) in Ba(OH)2 , bubbles were evolved with a current of 0T8 ampere ( frequency 40'5 , area same as platinum ) .
Nickel and Barium Hydrate.\#151 ; Fine nickel wire was employed .
Bubbles were given off freely when little of the wire was inserted in the hydrate .
The Dr. S. Marsh .
bubbles ceased when 16 cm .
of wire ( diameter 021 mm. ) were immersed ; current 0*16 ampere .
In another experiment the current was kept at 0*10 ampere and bubbling ceased when 10 cm .
of wire were immersed .
Both these give an initial activity corresponding to 0T5 ampere per square centimetre .
With nickel foil and barium hydrate a current 0'24 ampere was just able to give bubbles with foil 2*7 cm .
long and 2 mm. wide ; in another experiment 036 ampere was the value for foil 4 cm .
long and 2 mm. wide .
Both these give an initial activity of about 0'44 ampere per square centimetre .
S 13 .
Effect of Thickness of Electrodes upon the Phenomena .
Attention has been drawn ( p. 127 , S 4 ) to the striking results obtained with thick and thin platinum in dilute acid .
At all the frequencies used , approximately twice as much gas was evolved from the thick as from the thin platinum , the ratio of the thicknesses also being as 2 : 1 .
In general , more gas is evolved from a thick eleetrode than from a thin one .
It would be highly interesting to determine the relation between the volume of gas given off and the thickness of the electrode ( particularly with very thin films ) and the author intends to investigate this point by depositing such thin films of varying thickness upon glass .
It would appear that the activity of a thin film is greater than for a thick film of the same area .
There is no doubt that the difference in the activity of the two specimens of foil depends , in part , upon the nature of their surfaces , but nevertheless one must infer that the activity is a function of the thickness also .
One may picture the process as taking place somewhat as follows:\#151 ; Supposing that one of the causes of the activity is the presence of gas in the external portions of the electrode , then the greater the amount of gas present A \a ) C ( b ) C ----Electrode hhickness-------\gt ; D \lt ; \#166 ; ----Electrode-----\gt ; 0 Thickness Fig 10 .
On Alternating Cur Electrolysis .
at and near the face of the electrode , the greater will be the activity .
Now there will be a tendency for the gases to diffuse inwards , and the rate of such diffusion will be the greater when the amount of gas present in the interior is small .
In fig. 10ft we have the case of a thick electrode where the diffusion of gas into the electrode from the face AB is unaffected by that from CD .
In fig. 106 , on the other hand , the diffusion through one face is evidently affected by the presence of gas in the interior of the electrode , due to diffusion through the opposite face .
The gas density in the interior of the thin film is higher than that in the thick film , hence the rate of diffusion is slower in the case of the thin film than in that of the thick .
In other words , there will be a greater surface activity with thin films than with thick .
S 14 .
Gas Tests .
In most of the later experiments , the constitution of the evolved gases was determined .
In some cases the oxygen was absorbed by pyrogallic acid , in others by a solution of sodium hydro-sulphite and sodium hydrate .
The hydrogen present was determined by admixture with air , followed by explosion in an explosion chamber .
The general result was that the hydrogen was slightly in excess of the volume necessary to combine with the oxygen .
In every case the sum of the separate volumes was slightly less ( 1 c.c. or so ) than the volume of mixed gas .
It will be advisable in future work to carry out gas tests during the progress of an experiment , and not alone at the end of the experiment .
The difficulty will be that of dealing with the small quantities of gas evolved .
S 15 .
Partial Restoration of Electrodes to their State as Regards Gas Evolution .
If the platinum electrodes used in the acid electrolyte with alternating current are made to act as anodes with direct current they recover their ability to evolve gases when used again with alternating current .
During the passage of the direct current , the surface of the electrodes changes from greyish-black to dull grey .
The restoration is never perfect .
The longer the time of passage of direct current , the greater is the volume of gas evolved subsequently by alternating current , but there appears to be no simple relation between the two quantities .
If the electrodes after use with alternating current are strongly heated , they recover , in part , this power of evolving gases .
Similar effects have been observed with gold .
It may be restored partially by ( a ) making it serve as the anode with direct current , ( strong heating , VOL. xcvii.\#151 ; a , On Alternating Current Electrolysis .
( c ) immersion for a short time in aqua regia .
In the two latter operations , the mottled appearance of the gold changes immediately to the bright yellow colour of the unused foil .
In the case of barium hydrate it was observed that the ability to evolve gases was restored ( in part ) by giving the electrodes a rest while in the electrolyte .
With sulphuric acid this was not the case\#151 ; merely leaving the electrodes for some time was not efficacious in restoring them .
There are marked differences in the surface changes produced in the electrodes when in different electrolytes .
On platinum in acid a greyish-black stain was produced , which could be easily wiped off .
With barium hydrate , on the other hand , the stain was much blacker , particularly round the edges , and was much more adherent .
Strong heating had little effect upon it .
Conclusion .
In the preceding pages the behaviour of platinum , gold and nickel electrodes during the passage of alternating current through different electrolytes has been examined .
Two suggestions have been put forward to explain the phenomena : ( a)the oxidation of the electrode , coupled with subsequent reduction by the hydrogen of the next half-period , ( b ) the absorption and adsorption of hydrogen by the electrode coupled with re-combination with oxygen of the next half-period .
These processes are not mutually exclusive , and it is possible that both agencies are operative in degrees varying with the metal employed .
Further experiments with other electrodes and electrolytes ( in which also the concentration is varied ) are desirable .
The author hopes to resume work in some of the points outlined in the paper at an early date .
The preceding research has been carried out in the Physics Department of Battersea Polytechnic , and the author desires to take this opportunity of expressing his thanks and indebtedness to his colleagues , Mr. Wilson and Dr. White of the Chemical Department , and Mr. Thomson , late head of Physics Department , for their interest and valued criticisms during its progress .
He is also indebted to Dr. E. H. Griffiths , F.R.S. , for his valuable advice and suggestions during the writing of the paper ,

