V. D. Kuznetsov

Electrical Conductivity of Pure Toluene

(with 1 page of drawings accompanying the article).

To trace the dependence of the current strength on the potential difference for a purified liquid dielectric of definite chemical composition is of great interest, since from this dependence important conclusions can be obtained.

Toluene was chosen as the object of investigation solely because this liquid was available in sufficient quantity and could be obtained in pure form.

The method of purifying the toluene was as follows. First of all, the toluene was washed in a separatory funnel with a ground-glass stopper with distilled water, the water being changed many times. After this, in the same funnel the toluene was shaken for a long time with mercury and again washed with water. After the appearance of the dirty precipitate, which was continuously removed, had ceased, the liquid was dried with ignited calcium chloride and subjected to double distillation.

The distillation apparatus consisted of a distillation flask, a three-bulb dephlegmator, a thermometer, a condenser, and a forerunner receiver with a branch, into which calcium chloride was placed between two cotton plugs. All glass parts were washed with nitric acid with potassium dichromate, distilled water, absolute alcohol, and ether. In addition, all parts were treated with steam from distilled water. The liquid was collected in a jar with a ground-glass stopper and ground-glass contact. The distillation flask and the jar were rinsed with pure toluene before each distillation. The first portions during distillation, as well as part of the liquid used for rinsing, were discarded. The apparatus was assembled on cork stoppers, and in such a way that the stoppers were washed only by toluene vapors, and not by the liquid itself. The distillation flask was placed in an oil bath with electric heating.

In Fig. 1 is shown the measuring condenser constructed by me, used for investigations with toluene. This condenser, among other things, was intended for the study of the ionization of liquid dielectrics by rays of radioactive substances. At the bottom of a glass thick-walled crystallizer \(K\) three recesses were made, into which the screw feet of a brass disk \(DD\) were inserted (\(D = 90.5\) mm, thickness \(= 2.5\) mm). This disk was connected by a brass wire with the clamp \(Z_1\) and served as the lower plate of the condenser. The brass cover \(LK\) of the condenser had, at the center, a round aperture through which there passed with friction a brass tube (external \(D = 44.0\) mm, thickness \(= 1.2\) mm), provided at its lower end with a screw-on guard ring \(OK\) (external \(D = 90.5\) mm, thickness \(= 2.5\) mm). Inside this tube another brass tube was placed (external \(D = 37.0\) mm, thickness \(= 0.7\) mm), insulated from the first by an ebonite washer \(ESh\). The inner tube had on the lower—

end a ring with a thread and a conical gap. Into the opening of the ring a brass disk with a beveled edge was inserted; it fitted tightly into the gap and was pressed by a ring-shaped nut. The lower surfaces of the guard ring and of the bottom of the inner tube lay in one plane¹. The guard ring was connected to earth, and the inner tube to the electrometer. Between the shoulders of the outer tube and the lid there was placed either a metal ring MK, cut into two halves and having a width equal to the distance between the electrodes, or three pieces of glass of identical thickness. The setting screws were adjusted so that, without the MK ring, the upper and lower plates were exactly in contact with their planes. All metal parts of the apparatus were nickel-plated. From above the apparatus was covered with a metal cap with a removable cover.

The current strength was measured by Bronson’s method. As the large resistance a graphite line on an ebonite plate was used. By thickening the line with a pencil or erasing it with rubber, it was possible to make the deflection of the electrometer needle conveniently large for observation. The distance between the electrodes was 10 mm.

Experiments 8·3·19. The condenser was washed with pure toluene, distilled twice, and then filled with this dielectric.

The experiment consisted in the following: the lower plate of the condenser was given a potential from the positive pole of a battery of small accumulators, the other pole of which was led to earth. Two minutes after the field had been created, a reading was taken on the scale—the position of the light signal on the translucent scale; then, after one minute, the field was destroyed; a new field was created one minute after removal of the preceding one; after 2 minutes a reading was taken, and so on. The results of the observations are given in Table I.

Table I.

As is seen from the table, the current strength \(i\), proportional to the deflection \(a\), tends to lag behind \(V\), but current saturation is not obtained; on the contrary, at large \(V\) the current strength obeys Ohm’s law. In Fig. 2 (curve I) the relation between \(i\) and \(V\) is represented graphically.

Experiments 12·3·19. The condenser with toluene remained at rest after the experiments of 8·3·19. The field was created for 5 minutes; then, during the following 5 minutes, the field was destroyed; thereafter a new field was applied, and so on. Table II and Fig. 2 (curves II and III) present the results of the observations. At the first two potentials (22·5 and 54 volts) the zero did not remain constant, and therefore the first two readings are doubtful; the following readings did not have this defect.

Table II.

12·3·19. \(V>0\)

Continuation of Table II.

In order to express the current strength in absolute measure, immediately after the end of the experiments the electrometer was calibrated by means of a Weston standard cell and accumulators, the electromotive force of which was measured with a Weston voltmeter. It turned out that, for deflections up to 20 mm, the dependence of the deflection on the voltage may be regarded as linear, with 2.05 volts producing a deflection of 14.6 mm.

Table II and curve II (Fig. 2) show that, as the potential difference is raised, at about 400 volts a saturation current is obtained; with a further increase of the voltage, beginning with 475 volts, an increase is again obtained, and the current passes from steady to unsteady. Beginning at 700 volts the current begins to diminish and then becomes constant. This shows that the current in toluene reaches saturation; but the maintenance of its constancy as the voltage is increased is apparently hindered by foreign particles that have settled on the lower plate, which at high voltage break away from the electrode and distort the phenomenon.

It is important for us to calculate the order of magnitude for the sum of the mobilities of the ions giving the saturation current. We shall do this.

The current strength according to Bronson’s method is expressed by the formula \(i = \dfrac{V}{R}\), where \(V\) is the potential at the end of the large resistance \(R\). At the saturation current, an electrometer deflection of 14.5 mm was obtained, which corresponds to a potential \(V = 2.6\) volts. To determine the resistance \(R\) of the graphite line I used a known resistance \(R_0 = 4.3 \cdot 10^9\) ohms (glass tube No. 2 with platinum electrodes, coated inside with a layer of soot). In Fig. 3 a circuit for determining \(R\) is presented. The current from two accumulators \(A\) passed successively through \(R_0\) and \(R\), one pole of the battery being connected to earth. Denoting by \(V_A\), \(V_B\), and \(V_C = 0\) the potentials at the points \(A\), \(B\), and \(C\), we have

\[ i = \frac{V_A - V_B}{R_0} = \frac{V_B}{R}, \]

whence

\[ R = R_0 \frac{V_B}{V_A - V_B}. \]

The measurements gave: \(V_A = 4.0\) volts (26.6 mm) and \(V_B = 2.1\) volts (15.0 mm), hence \(R = 4.8 \cdot 10^9\) ohms. On the basis of this saturation current

\[ J=\frac{2.0}{4.8}\cdot 10^{-9}=0.42\cdot 10^{-9}\ \text{amp.} =0.42\cdot 10^{-9}\cdot 3\cdot 10^9=1.26\ \text{el.-stat. units}. \]

To determine the sum of the mobilities of the ions \(K_1+K_2\) I used the formula\(^1\):

\[ K_1+K_2=\frac{4\pi S d \lambda_0^2}{J K}, \]

where \(S\) is the area of the upper electrode, \(d\) the distance between the electrodes, \(\lambda_0\) the specific electrical conductivity at weak voltages, i.e. when the current strength obeys Ohm’s law, \(J\) the saturation current, and \(K\) the dielectric constant. We have \(S=10.8\ \text{cm}^2\), since the diameter of the inner cylinder of the capacitor \(D=3.70\ \text{cm}\), \(d=1.0\ \text{cm}\), and \(K=2.3\). Using the table for calibrating the electrometer, we calculate the value of the current density at three voltages, 22.5, 54.0, and 105 volts, and from these values obtain respectively \(\lambda_0=1.48,\ 1.41,\) and \(1.41\), with the mean \(\lambda_0=1.43\) electrostatic units. On the basis of these data we calculate:

\[ K_1+K_2=9.55\ \text{el.-stat. units} =\frac{9.55}{300} =3.2\cdot 10^{-2}\ \frac{\text{cm}}{\text{sec.}} \bigg/ \frac{\text{volt}}{\text{cm}} . \]

The value obtained for the sum of the mobilities is approximately 100 times greater than the number

\[ \left[\,3.8\cdot 10^{-4}\ \frac{\text{cm}}{\text{sec.}} \bigg/ \frac{\text{volt}}{\text{cm}}\,\right], \]

obtained by C. Böhm-Wendt and E. v. Schweidler for petroleum ether. It also considerably exceeds all the other values of \(K_1+K_2\) obtained by some authors for other liquids. There is no error in my calculations, and consequently it is necessary to assume that in pure toluene, during the first moments after the application of a potential difference, there are present very easily mobile ions which can give a saturation current. These ions are gradually removed by the current, and less mobile ions remain in the toluene. It would be interesting to continue investigations in this direction, but this will be possible only when chemically pure toluene is prepared or obtained.

In a separate work I have established that platinum electrodes, when a potential difference is applied, liberate foreign substances which form heavy ions, and that it is impossible to obtain a saturation current with these electrodes. It could be expected that for pure toluene, with platinum electrodes, the dependence of the current strength \(i\) on the potential difference \(V\) is expressed by the formula

\[ i=f(V)+cV \]

where \(c\) is a constant multiplier. This means that the current consists of two parts. The first part \(f(V)\), as \(V\) increases, asymptotically approaches a certain constant value \(a\), reaching saturation, as occurs in gases, while the second part, \(cV\), expresses a current obeying Ohm’s law.

It is known that the current through a liquid dielectric decreases with time. It is very important to determine which part of the current, \(f(V)\) or \(cV\), decreases.

\(^1\) C. Böhm-Wendt und E. v. Schweidler: Phys. ZS. 10, 379 (1909).

To clarify the role of the platinum electrodes and to resolve the question just raised, a series of experiments was carried out with the very same toluene. I took a new Kohlrausch measuring vessel (Fig. 4), never previously used, with platinum electrodes. The vessel was washed with strong sulfuric acid, with a solution of potassium dichromate, and with distilled water. In addition, it was treated with steam. After drying over a flame, the vessel was washed repeatedly with pure toluene, twice distilled, and was filled with this liquid.

Experiments of 16·3·19. Bronson’s method. High resistance—a glass tube with soot. Weight of the empty vessel 55.598 g, weight of the vessel with toluene 78.270 g. The determination of the resistance \(R\) of the tube with soot was carried out by Bronson’s method, a graphite line with a resistance of \(4.8 \cdot 10^{9}\) ohms being taken as the standard resistance. From four separate measurements at 106, 209, 315, and 418 volts, the mean value obtained was \(R = 2.8 \cdot 10^{11}\) ohms. The electrometer had first been calibrated. The results of the observations are presented in Table III. The current was closed for 5 minutes; then, for 5 minutes, the field was absent, after which a new voltage was applied.

Table III.

16·3·19.

Experiments of 19·3·19. The toluene remained after the experiments of 16·3·19 in the Kohlrausch vessel. In the evening at 6 o’clock on 16·3·19, after the experiments, a current was passed through the toluene, which continued to flow until 6 o’clock in the evening on 18·3·19, i.e. for exactly two days. A new investigation began on the morning of 19·3·19.

The method and apparatus were the same. The current was not interrupted, but the voltage was changed every 30 minutes. The results of the observations are contained in Table IV and are represented graphically, together with the results of the preceding experiments, in Fig. 5.

Table IV.

19·3·19

Discussion of the results.

1. The tables and diagrams show that a saturation current is not obtained with platinum electrodes, whereas the very same toluene with brass nickel-plated electrodes gave a saturation current. This compels us to attribute the reason for the failure to obtain a saturation current to the platinum electrodes.

¹) Current closed.
²) Null reset.
³) Oscillations.
⁴) Current opened.

Jaffé¹ very easily, already at a field strength of 200 volts/cm, obtained a saturation current for pure hexane enclosed in a cylindrical brass condenser, whereas Schröder² could not obtain a saturation current for ethyl ether even with carefully heated electrodes (with unheated platinum electrodes the current increased faster than the potential difference).

All the lack of success of Schröder’s experiments should, it seems to me, be attributed exclusively to the fact that he used platinum, since he carried out the purification of the ether very carefully.

1. With platinum electrodes the current strength in all cases can be expressed by the formula proposed by Jaffé:

\[ i = f(V) + cV, \]

where the term \(f(V)\) becomes constant beginning with some value of \(V\).

1. Drawing straight lines through the last points of the curves to their intersection with the ordinate axis, we obtain intersection points characterizing the saturation currents at various moments from the beginning of the passage of the current. For curve I, marked with dots and expressing the current two minutes after closing, the deflection corresponding to the saturation current is 110 mm, and for curve II, marked with crosses (after 4 minutes), the deflection is about 90 mm. Both straight lines are parallel to one another, which can be the case only when the decrease of the current with time belongs to the part \(f(V)\), i.e. to the saturation current, since if the decrease belonged to the part \(cV\), the slopes of the straight lines would be different, and the straight line corresponding to the larger interval of time from the beginning of the current would have the smaller angular coefficient. Everything said holds in the case when the increase of voltage is accompanied by interruptions.

2. Passing to curve III, we see that the straight line passing through its last points intersects the ordinate axis at the point 110 mm, despite the fact that it lies below curve I. Consequently, the saturation current obtained from readings two minutes after increasing the voltage is the same, whereas the current obeying Ohm’s law in toluene purified by prolonged action of the electric field proves to be weaker than the current in toluene not subjected to preliminary action of the field.

3. Curve IV, marked with small circles, coincides with curve II; both of them give, for the saturation current, deflections of 90 mm. The slope of straight line IV is greater than that of straight line III. This shows that the saturation current approximately after 4 minutes reaches its constant value and does not decrease further with time. The current obeying Ohm’s law may have different values depending on the electrical history of the toluene.

Fig. 1.

Fig. 2.

Graph labels:

• Vertical axis: deflection
• Horizontal axis: volts
• Toluene
• 8-3-19 and 12-3-19
• Curves: I, II, III

Fig. 3.

Circuit label:

• to the electrometer

To the article by V. D. Kuznetsov:

“Electrical Conductivity of Pure Toluene”

Fig. 4.

Scale label:

• 1 cm

Electrode labels:

• A
• B

Level marks:

• 0
• 1
• 2
• 3
• 4

Fig. 5.

Graph labels:

• Toluene.
• Kondrashov vessel.
• Curves: I, II, III, (IV)

Legend:

• 8-3-19
• • — after 2 minutes
• × — after 4 minutes
• 10-3-19
• ▽ — after 2 minutes
• ○ — after 28 minutes

Notes on the graph:

• After closing for 5 minutes — rest 5 minutes.
• Change of voltage without interruptions.

Formula on the graph:

\[ i = f(v) + c v . \]

1. G. Jaffé. Ann. d. Phys. 28, 326 (1909) ↩↩

2. Schröder. Ann. d. Phys. 29, 125 (1909) ↩