An EXPERIMENT to ILLUSTRATE the MECHANISM of the CONDUCTANCE

PROCESS in IONIC SOLUTIONS LEWIS G. LONGSWORTH

The Laboratories of The Rockefeller Institute for Medical Research, New York Cily

A simplified experiment for the accurate measurement of taneous growth of a visible diffusion layer illustrates the transference data by the method of mooing boundaries i s mechanism of electrical transport at a metal-solution described. In this method the conductance process i n the interface. The experiment i s suitable for laboratory main body of the solution may be followed by obsernring the courses i n physical or electro-chemistry and i s capable of motion of the boundary while at the electrode the simul- yielding precise results.

T HE LABORATORY experiment usually em- ployed for the illustration of electrical transference in ionic solutions and the phenomena occurring

a t a solution-metal interface is based upon the classic researches of Hittorf and depends upon the analysis of those portions of the solution which surround the elec- trodes before and after the passage of a measured quantity of electricity. In this Hittorf or gravimetric method the electrolysis is seldom accompanied by visually observable changes in the solution and most of the time is spent in analytical work--operations which teach little concerning the mechanism of the conductance process. The laboratory experiment which is the subject of this paper overcomes some of these diEculties and requires only apparatus which is usually available for undergraduate laboratory instmc- tion in electrochemistry. The experiment is a simpli- fied form of the moving houndary method for the determination of transference numbers and possesses the following advantages over the gravimetric method.

occurs a t the anode. Thus the solution for trans- ference-number determination is placed in the tube A, Figure 1, in the bottom of which is sealed a cylinder of some metal which forms a soluble salt in combination with the anion of this solution. As an example, the solution in A may be 0.1 N potassium chloride and the metal cadmium. If now current is passed in the direc- tion indicated by the arrow, cadmium chloride will form at the electrode, and since the cadmium ion

constituent is slower than the potassium ion const i tuent a boundary between the solutions of cadmium chloride and potas- sium chloride will leave the face of the electrode and move up the tube as indicated in Fig-

Flcrrne 1 mcuan 2 ure 2.

SrMpuaIeD METHOD Of course, as the boundary

p o n ~ l ~ ~ THG BOUND. moves upward, the common- ALRY ion constituent, which is the

chloride ion in this example, 1. The actual migration of an ion constituent may he moves downward the and eventually

observed. accumulates around the anode. Electrical neutrality 2. The concentration change which occurs around one is preserved in the solution near the electrode due

of the electrodes is visible and furnishes an accurate to the fact that only a fraction of the ions conception of how the electrode processes differ which are formed by the electrochemical oxidation of from those in the main body of the solution. the anode are necessary for the maintenance of the

3. The method is capable of yielding precise results in growing column of indicator solution at the adjusted a short time and it is possible to check the progress concentration and consequently only this fraction is of the experiment almost from the beginning. carried away from the anode by electrical mimation.

DESCRIPTION OF THE EXPERIMENT A concentrated solution of cadmium chlorize thus forms just above the anode and slowly diffuses upward

The experiment described below is a simplification, into the more dilute indicator solution. As the experi- due to Franklin and Cady (1) and Cady and Longs- ment proceeds the thickness of this "diffusion layer" worth (Z), of the general moving-boundary method in gradually increases, the layer being quite visible be- that no special mechanism is required for the initial cause of ,+e variation of the refractive index in formation of the boundary by the superposition of two this region. The diffusion layer is indicated by the solutions. The "indicator" or following solution is shading above the electrode in Figure 2. If , at the formed as a result of the electrochemical process which conclusion of the experiment, the tube i s tip#ed, this

420

hemy solution of cadmium chloride may be seen falling d m the tube.

ELEMENTARY THEORY

In accordance with the definition of the transference number, the passage of 1 Faraday, or 96,500 coulombs, of electricity through a solution of 0.1 N potassium chloride, for example, causes T+ equivalents of potas- sium ion to pass a fixed point in the solution. T+ is the cation transference number and this number of equiva- lents is contained in a volume of solution, V, in ml.,

equal to 'OoO '+, where C is the concentration in C

equivalents per liter of solution. Of course, the quantity of electricity, f , necessary to cause the bound- ary to move through a smaller volume, v, will be pro; portionately smaller than F, so that

Since the amount of electricity involved in a de- termination is less than can be measured accurately with a conlometer i t is necessary to determine the number of coulombs by passing a constant current, i, through the tube and multiplying this by the t i e , t , in seconds, required for the boundary to move through the given volume, v. Thus f becomes equal to it and equation (1) may be written:

the equation which is used in computing the transfer- ence number from the experimental data.

EXPERIMENTAL DETAILS

The Moving Boundary Cell. An easily constructed moving boundary cell which is suitable for measure- ments that can be made by the simplified method out- lined above is shown in Figure 3. In this method the cathode must be sufficiently far removed from the tube in which the boundary is observed to prevent the concentration changes which occur there from pene- trating, either by diffusion or convection, into the neighborhood of the boundary. The cathode chamber e, Figure 3, is consequently of the type described by MacInnes and Brighton (3), slightly modified, since this arrangement satisfies the preceding requirement and is quite compact. The inner tube c is closed a t the lower end, thus retaining the heavy solution of potassium chloride which is developed around the silver-silver chloride cathode f , and is supported by the projections d. This tube may be slipped out of the cell when the rubber stopper b, wbich carries the cathode connection and the vent a, is not in place and i t is with this inner tube removed that the cell may be most conveniently filled.

The author is indebted to Dr. D. A. MacInnes for the very practical suggestion that a 1-ml. graduated pipet, on which the graduations do not extend to the tip and from which the tip has been removed, may be used for

the tube, g, in which the boundary is observed. The upper end of this tube is sealed to the cathode chamber while a cylinder of cadmium, h, to the base of which a connecting wire, k, is soldered, is sealed into the lower end of the tube by means of the piscein plug, i. The tube, 1, leads the anode connection out of the water bath, care being taken that the rubber connection, j, does not obscure the solution-metal interface where the boundary is first detected and where the diffusion layer is observed to form. This interface should be a few mm. below the first graduation on the tube.

The introduction of the 0.1 N potassium chloride into the cell, which has been pre- viously cleaned and thoroughly

h dried, may be accomplished as follows. With the cathode and inner tube c removed, a few ml. of solution are poured into the cell. A solid glass rod, with a flat end and a diameter

e slightly less than that of the graduated tube, is used as a pis- ton in the latter until it is en-

f tirely filled with solution. d Care should be taken that the

bubble of air wbich clings te- naciously to the face of the

g anode is forced out of the tube. After the graduated tube has been filled with solution the in- ner tube c is placed in position and the, cathode chamber filled. The silver-silver chloride cath- ode which is finally inserted

1 is of the form described by j Smith and MacInnes (4). I t k .

is prepared by heavily plating I platinum gauze with silver and FIGUKE 3.-~ne MOT- then thoroughly chloridizing

'NG BOUNDARY CELL the silver deposit by electrolyz- ing as anode in a solution of

hydrogen chloride. Since this electrode must carry considerable quantities of electricity with no evolution of gas the area of the electrode should be a t least one square centimeter.

It is necessary to immerse the tube in which the boundary moves in a water bath (T, Figure 4) in order to dissipate the heat which is developed in the tube by the passage of the electric current. Otherwise convec- tion currents in the tube will distort and even destroy the boundary. The author has used a cylindrical glass jar, 9 an. in diameter and 24 em. high, for the water bath. Only rough temperature control is required, and it is unnecessary to stir the water in the bath.

The Electrical Circuit. The measurement of the current and the maintenance of this current a t a con- stant value may be accomplished as follows. The fixed resistance R, Figure 4, is placed in series with the moving boundary cell and the current through the latter is determined by measuring, on the potenti-

ometer P, the potential drop across this accurately calibrated resistance. A portion of the battery, B, is shunted across the rheostat R' which has a sliding contact F and by adjustment of this contact a constant current through the cell may be maintained. With the potentiometer set a t a given value the observer con- tinually adjusts the contact F so as to keep the needle of the galvanometer G near its zero position. As the boundary ascends the tube the length of the column of indicator electrolyte increases correspondingly and,

- - FIGURE 4.-THE COMPLETE ASSEMBLY

Electrical Apparatus Used in a Typical Determination.

P-Students' type potentiometer &Portable type galvanometer R-3M)A resistancefrom a dial decade box R'-7000n adjustable rheostat A-90 volts "B" battery B-150 volts "B" battery

since this solution is a poorer conductor than the one which it replaces, the total resistance of the cell con- tinuously increases. The potential applied to the cell must therefore be continually increased in order to maintain a constant current.

The time, to the nearest second, a t which the bound- ary crosses successive graduations on the tube is obtained with the aid of a watch. When the current is constant the time intervals necessary for the bound- ary to sweep through successive 0.1-ml. portions of the tube should be identical and the constancy of these intervals furnishes an important check upon the progress of the experiment.

Obsemtion of the Boundarv. The boundarv is

of the two solutions. Although a trained observer can locate the average boundary unaided, the simple optical system shown in Figure 4 greatly helps in the detection of the boundary. A source of diffused light, CD, should be placed behind the graduated tube. An opaque screen, S, such as a rectangular piece of card- board, is held in one hand and placed between the light source and the tube while the boundary is viewed with a simple reading lens, L, held in the other hand. When the upper edge of the opaque screen is nearly in line with the boundary, the lens, and the eye, this edge appears to be distorted. Keeping the lens and the eye approximately on a horizontal plane with the upper edge of the opaque screen, and in line with the tube, these are moved slowly up or down until a distortion of the edge of the screen indicates the position of the boundary. As the boundary leaves the face of the anode immediately after the electrical circuit is closed i t is quite easily located and the observer should familiarize himself with its detection in this interval before it has reached the first graduation on the tube.*

RESULTS OF A TYPICAL EXPERIMENT

The results of an experiment which was performed in the manner outlined are given in Table 1. In this table, which also illustrates a method of recording the data, the time recorded in the second column is that a t which the boundary passed the graduation corre- sponding to the volume recorded in the first column. The differences between successive time observations, reduced to seconds, are recorded in the third column and it is the approximate constancy of these values which indicates that the boundary has moved at a uniform rate.

TABLE 1 Dsmsm~rroa OP TBB CATSON TEANSPBBBNCB NOXB&R or 0.1 N KC1

Repistmce (R, Fig. 4) = 3003 Potentiometer setting - 1.2188 volts

Current - 0.004082 ampere Vdumc, Time Al,

d. rrconds

Total 4885

DISCUSSION OF TBE RESULTS

In the computation of T+ from the data of Table 1 by means of equation (2) i t is evident that u = 1 ml., C = 0.1, F = 96,500, i = 0.004062 and t = 4885. The value of T+ computed in this manner, 0.4868, is subject to a correction of 4-0.2 per cent. for volume changes and thus becomes 0.487~. This relatively small correction unfortunately involves some rather d i c u l t computa-

* Since it is necessary to adjust the potential across the cell very frequently in order to maintain a constant current, it may he desirable to have two observers. one followine the boundam

visible because of the difference in the refractive indices while the other is regulating the potential. -

tions but is completely treated in reference (5). page 203. The value, 0.4878, is 0.5 per cent lower than the value 0.4898, which has been obtained in the most recent determinations (6) and i t was observed later that this discrepancy between the two values was almost entirely due to the manufacturer's calibration of the 1- ml. pipet which was used in the construction of the moving boundary cell. An independent calibration of this volume gave the value, 1.0046 ml. Using this figure for the volume, the data of Table 1 give a cor- rected transference number of 0.4895, a value which agrees well with the value previously mentioned. This value, which has been obtained by the moving bound- ary method, is in excellent agreement with the most recent determinations of this constant by the Hittorf method (7).

It is probable that the l-ml. pipet was calibrated by the maker "to deliver" instead of "to contain." If the measurements of time and current are of s&cient accuracy to warrant it, an independent calibration of the tube volume should be made.

The particular tube which was used in the preceding determination was graduated for a length of 16 cm. Since this is somewhat longer than the tubes which are ordimarily used in precise measurements by the moving boundary method, a large battery is required. Thus in the experiment described above a total of about 240 volts in the form of "B" batteries was required. The battery requirements may be decreased by either shortening the tube or using a lower current or both. The latter expedient is practical only within certain limits, however, since the boundary is diicult to locate

if the current density is too low. It is also practicable to decrease the drain on the battery, B, Figure 4, by regulating the current with a variable resistance in series with this battery instead of the shunted resist- ance, R'.

In this brief paper i t is not possible to discuss such subjects as the persistence and visibility of the bound- ary and the concentration relation between the leading and indicator solutions. The entire subject of the moving boundary method, however, is fully treated in a recent review by MacInnes and Longsworth (5), a paper which also contains a complete bibliography of the subject up to 1932.

The author wishes to express his thanks to Dr. D. A. MacInnes for his kind suggestions in connection with this work.

LITERATURE CITED

(1) FRANKLIN AND CADY, "Ionic velocities in liquid ammonia." 3. Am. Chcrn. Soc., 26,499-530 (1904).

(2) CADY AND LoNGsWoRTn. "A modification of the movin~ boundary method for the determination of transference numbers," ibid., 51, 1656-64 (1929).

MACINNES AND BRIGHTON, "Moving boundary method for the detumination of transference numbers. 111. A novel form of apparatus," ibid., 47, 994-9 (1925).

S m ~ n AND MACINNES, "The moving boundary method of determining transference numbers. 11." ;bid.. 46, 1398- 403 (1924).

M A C I ~ N E S AND LONGSWORTH, "Transference numbers by the method of moving bbundaries." Chem. Rm., 11, 171-230 1 1 9.17) ,----,.

LONGSWORTH, "An application of moving boundaries to a study of aqueous miAures of hydrogen chloride and potas- sium chloride," I. Am. Chern. Soc., 52, 1897-910 (1930).

MACINNES AND DOLE, !'The transference numbers of patas- sium chloride," ibid., 53,1357-64 (1931).