B' while the rubber tube between J and H is pinched. Cock E is turned so that E and H communicate and liquid flows into H to form a liquid junction with the saturated potassium chloride salt bridge.

Measurement is made with the potentiometer to give a combined emf for the entire chain which, let us assume, is 0.645v at 25° C. The emf of the saturated calomel half-cell and potassium chloride salt bridge at 25° C, as given by Clark, is 0.2458v. Then

the barometric correction being ignored.

Simpler forms of hydrogen electrodes that are satisfactory are illustrated in figure 58. Electrodes A and C are made of platinum

FIGURE 58. Various types of hydrogen electrodes.
A, B, C, D, dipping electrodes; E, rocking electrode.
Courtesy of Leeds & Northrup Co.

wire and foil, respectively, sealed into glass tubing, contact with the wire leading to the potentiometer being made through mercury contained in the tubing. At B and D are shown the electrodes in glass jackets which are connected with the supply of hydrogen gas by rubber tubing fitted to the side arm. Hydrogen under slight pressure bubbles from the bottom openings of the jacket, alternately bathing the electrode with hydrogen and with solution. When the solution has become saturated with hydrogen, the flow of gas is stopped and the emf is measured.

For the deposition of platinum or palladium black on bright platinum, Clark electrolyzes a 3-percent solution of the chloride of either metal, acidified with hydrochloric acid. The current from a 4-volt storage battery is allowed to produce a vigorous evolution of gas until the coating of black conceals the glint of the metal. The coated electrode as negative, is then placed in a dilute solution of sulfuric acid and electrolyzed to charge with hydrogen. The bubbles should come off evenly and the electrode should be evenly coated. The deposit should adhere under a vigorous stream of water and the electrodes should not be allowed to become dry. In use, electrodes frequently become clogged, poisoned, or otherwise unfitted for further use. The black should then be removed and the metal recoated. To remove the deposit of black, the electrode is connected to the positive pole of the battery and immersed in 1:1 hydrochloric acid which is electrolyzed. Palladium is more easily removed by electrolysis than

is platinum. The latter, however, is easily removed if the platinum electrode is plated with gold before the black is deposited.

(d) QUINHYDRONE ELECTRODE

The quinhydrone electrode (or half-cell), because of its simplicity and ease of manipulation, has been widely used for pH measurement. To make a measurement, the platinum or gold electrode is immersed in the test solution and a small amount of quinhydrone, which is sparingly soluble, is added and the potential is measured against a saturated calomel half-cell in liquid junction with the solution.

To calculate the pH of the solution from the observed potential, Biilmann [5], whose investigations contributed largely to our knowledge of the quinhydrone electrode, used the following equation:

pH=

-E+Eq-Ec
0.05912

at 25° C,

(87)

where E is the observed potential, Eq is the potential of the quinhydrone-platinum electrode when (H+)=1, and Ec is the potential of the saturated calomel half-cell when (H+)=1. The numerical values of Eq and Ec, as given by Clark [3, p. 672], are, respectively, 0.6992v and 0.2458v, so that eq 87 becomes

the sign of E being negative over most of the useful range of the electrode. Tables or charts are furnished with quinhydrone pH instruments, giving pH values corresponding to observed potentials for several temperatures.

For accurate results, the platinum or gold electrodes should be cleaned after use, and with careful manipulation the limit of error is about 0.01 pH. The quinhydrone electrode is inapplicable in the presence of oxidizing and reducing substances and in alkaline solutions with pH above 9.0. The arrangement of apparatus for pH measurements is shown in figure 59 (A).

(e) GLASS ELECTRODE

Haber and Klemenziewicz [6] appear to have been the first to demonstrate the usefulness of the glass electrode. Their electrode vessel was made by blowing a thin bulb on the end of a glass tube and contained an electrolyte in which dipped a platinum wire connected with an electrometer. The bulb was immersed in the test solution and the concentration chain was completed by means of a normal potassium chloride-calomel half-cell. The apparatus was used for electrometric titration of acids and bases, and the authors considered that the potential of the cell was determined in part by the concentration of hydrogen ions. They used a soft glass in preference to others that were tried, and directed that after the bulb was formed it was to be steamed inside and out for an hour and kept in pure water until used. Hughes [7], using apparatus similar to the above with the addition of a potentiometer, showed that the potential varies quantitatively with pH over a wide range of values and that the glass electrode may

be used in the presence of strong oxidizing or reducing agents when a hydrogen electrode is useless. He also showed that the linear relationship of glass-electrode potential and pH is affected by strong concentrations of certain salts. Kerridge [8] found the glass electrode useful in the presence of substances of biological origin that poison a hydrogen electrode.

A study of the composition of glasses was made by MacInnes and Dole [9], who consider as most suitable a glass composed of 22 percent of sodium oxide, 6 percent of calcium oxide, and 72 percent of silica. Glass of this composition, known as Corning 015, is made by Corning Glass Works, Corning, N. Y. The same investigators [10] produced membranes as thin as 0.025 mm by first blowing a bubble on the end of a glass tube until red and blue interference colors appeared. The end of a second tube, heated to dull redness, is placed against the bubble and the film fused to it. These membranes are fragile but have comparatively low resistance. Robertson [11] produced thinwalled bulbs by drawing 10-mm tubing to a tapering point and forming on this a small lump of glass by heating in a pointed flame. The bulb was then blown to a volume of 8 or 10 ml. Thompson [12] devised a metal-connected glass electrode of bulb or test-tube form by silvering the outer surface. The silver film was protected by lightly copperplating and the metal coating was then wired to a potentiometer. The test solution was added to the vessel and the chain was completed by means of a saturated calomel half-cell. No standard electrolyte is used in actual pH measurement, but each electrode must be calibrated before being used. This may be done with the help of standard buffer solutions, and the cell constant is determined.

Accounts of the evolution of the glass electrode are given by Perley [13], MacInnes and Longsworth [14], and by Clark [3], while theories of the action of the electrode have been advanced by Haber and Klemenziewicz [6], Horowitz [15], Michaelis [16], Dole [17], and MacInnes and Belcher [18].

The Haber type of vessel, consisting of a thin bulb blown on the end of a glass tube, is now commonly employed. Various electrolytes have been used inside the electrode but a 0.1 N solution of hydrochloric acid (pH=1.0) saturated with quinhydrone is generally favored for routine measurements. The electrode tube is half-filled with the solution, connection with the potentiometer being made with a short length of platinum wire sealed in a narrow glass tube containing mercury into which the potentiometer lead extends. The latter tube is supported in the rubber stopper of the electrode vessel. The bulb of the glass electrode is immersed in the test solution and a calomel half-cell is used as a reference electrode. The concentration chain may be represented thus:

Hg HgCl, KCl (sat.) solution 0.1 N HCl
Hg|HgCl, (sat.)|solution|

quinhydrone'

glass membrane

Pt.

The pH of an unknown solution is calculated from potential reading, E, as with the quinhydrone electrode, as described under (d), p. 289. That is, at 25° C.

Tables giving pH values corresponding to potentiometer readings at various temperatures are supplied with some instruments, whereas others are calibrated to read directly in pH values.

Since the resistance of the glass electrode is high, a sensitive detecting instrument must be used. Some workers prefer an electrometer, but with improved electrodes, a lamp-and-scale galvanometer with a sensitivity of 0.0005 μa per mm at 1 m is satisfactory, and an inexpensive potentiometer of the usual type may be used (fig. 59, A). The use of grounded metal supports for the electrodes prevents leakage currents from reaching the galvanometer.

Various methods of thermionic amplification of the weak currents of the glass electrode have been reported by Goode [19], Elder and Wright [20], Partridge [21], Müller [22], DuBridge [23], Morton [24], Gilbert and Cobb [25], and others. In figure 55 is shown a potentiometerelectrometer in which the amplification is incorporated in the potentiometer housing.

In some apparatus the entire assembly, consisting of potentiometer, vacuum-tube amplifier, dry cells, galvanometer, pH electrodes, etc., is housed in a single portable unit. One of these units is illustrated in figure 59 (B). With amplification, the employment of the glass electrode in automatic pH apparatus becomes possible. Longsworth and MacInnes [26] have described such a device for the automatic control of the addition of alkali solution to a growing culture of acidforming bacteria. A pH recorder for sugar juices, in which a glass electrode is employed, is described by Crites [27].

Electrodes made of the usual glass are considered inaccurate in the presence of sodium salts at ranges above pH 9.5; however, a calibration curve may be obtained for the range 9.6 to 12.5 which is said to be reproducible within 0.1 pH. Very recently the use of a different glass suitable for high alkaline ranges, has been reported [42].

(f) SILVER-SILVER CHLORIDE ELECTRODE

The silver-silver chloride half-cell, represented by Ag|AgCl KCl(m), where (m) refers to molar concentration, is analagous to the calomel half-cell for which it has been substituted as a secondary reference electrode, particularly in the study of reactions in chloride solutions. The electrode consists of a coating of silver chloride, intimately mixed with silver, on platinum or silver-plated platinum. The platinum support, which may be in the form of a small rectangle of foil or gauze or a wire coil, is sealed by means of a short lead of platinum wire to a convenient length of glass tubing, as in the construction of hydrogen electrodes. The silver-silver chloride coating may be produced in various ways, either electrolytically [37], thermoelectrolytically [38], or thermally [39]. The last-named process is simple and produces satisfactory electrodes. A coil of platinum wire is sealed into a glass tube and covered with a paste composed of 7 parts of silver oxide (precipitated and washed) and 1 part of silver chlorate and heated to decomposition in an electric furnace. After coating the electrodes, regardless of the process used, they are washed in many changes of distilled water to remove contaminants, and finally washed in the solution in which they are to be used. The period of washing and aging to stability may extend to as much as 20 days.

Smith and Taylor [40] studied the reproducibility and stability of silver chloride electrodes and found an average agreement of 0.02 mv.