INDUSTRIAL AND ENGINEERING CHEMISTRY
350
cup to avoid contamination of the sampIe by the hold-up in the reflux. For very careful work, the sample should be removed by applying vacuum to the receiver, rather than pressure to the flask. TABLEI. LIQUID-VAPOR EQUILIBRIUM DAT.4 ALCOHOL-WATER
11 .~
12 13 15 17 18 19 20 21 22 25 26 27 28
0.8083 0.8329 0.8629 0,8839 0,8983 0.9166 0.9169 0.8348 0.8879 0.9301 0.9811 0.9777 0.9636 0.9890
ETHYL
Mole Per cent of Ethyl Alcohol Liquid Vapor
Density 259/25a Liquid Vapor
Run
FOR
81.8 66.5 49.3 39.3 33.4 27.0 27.0 65.3 37.7 22.7 4.8 6.0 11.0 2.5
0.8077 0.8231 0.8365 0.8442 0.8480 0,8516 0.8517 0.8230 0.8437 0.8547 0.8867 0,8820 0,8693 0.9230
82.5 72.8 64.0 59.0 57.0 55.0 55.0 73.5 59.5 53.5 35.7 40.0 46.0 25.0
VOL. 7 , NO. 5
AS a check on the apparatus, the curve shown in Figure 2 , was obtained, using ethyl alcohol and water. Samples were analyzed by the method of specific gravity, using a 10-ml. pycnometer in a constant-temperature bath a t 25" C., with the tables of the U. S. Bureau of Standards. Samples were drawn by suction on the receiver. Pressure was atniospheric (29.690 inches of mercury). The data of Carey and Lewis (2) are plotted with the authors' in Figure 2 and show good agreement, the authors points being slightly lower than theirs in the middle range, and slightly higher in the lower range.
Literature Cited (1) Brown, E.H., J. Chern. Education, 9, 1114 (1932). (2) Carey, J. S., and Lewis, W. K., IND. ENG.CHEM.,24, 882-3 (1932).
(3) Othmer, D.F., Zbid., 20, 743 (1928). RECEIVED September 5, 1934.
The Germanium-Germanium Dioxide Electrode M. L. NICHOLS
AND
S. R. COOPER, Department of Chemistry, Cornell University, Ithaca, N. Y.
SI"
TCE the introduction of the hydrogen electrode by Bijttger (1) in 1897 and its popularization by Hildebrand (6) in 1913, potentiometric measurements and titrations have attained great importance. Because of the difficulties of obtaining the highest degree of accuracy with this electrode and its limitations, several electrodes, such as the quinhydrone and many metal-metal oxide electrodes, have been investigated as substitutes (2, 4, 6). In order for a metal to function as a metal-metal oxide electrode it should be slightly soluble, cathodic to hydrogen, and have a slightly soluble oxide. Germanium is a slightly soluble metal, whose dioxide is polymorphic in that it exists in a soluble and insoluble form. Laubengayer and Morton (8) state that the insoluble form of the dioxide is not appreciably soluble in 25 N hydrofluoric acid, 12 N hydrochloric acid, or 36 N sulfuric acid, but is 16.8 to 17.8 per cent soluble in 5 N sodium hydroxide solution. Since it is also cathodic to hydrogen, this electrode was studied to determine its application to pH determinations and electrometric titrations. For the electrode Ge/GeOz, H f , H20 the anodic reaction can be represented as Ge(s)2H20 --+
Ge02(s)
+ 4H+ + 4F
and therefore
Then for the cell Ge/GeOz, HA/salt bridge/KCl, HgzClz/Hg E = E,, - ECaI.and EZ5' = E;;L,25" - E?;, - 0.0591 log [H+] or E260 = E&Z50 - E26" cat, 0.0591 PH
+
Therefore the germanium-germanium dioxide should function as a metal-metal oxide electrode with the observed electromotive force for the cell changing 0.0591 volt a t 25" C. for each unit change in pH of the solution.
Experiment a1 The insoluble germanium dioxide was prepared according to the method of Laubengayer and Morton (8), and showed practically no solubility in a 48 per cent solution of hydrofluoric acid. The metallic germanium used had been prepared according to the method of Tressler and Dennis (IO). The germanium prepared as above was cast into a rod approximately 5 X 5 x 50 mm. A clay boat was filled with an alundum paste and a piece of wood the size of the ingot desired was pressed into the moist paste and the whole dried at 110" C. When the paste had dried, the wood was removed and the depression was filled with powdered germanium. The boat and its contents were heated in an electric combustion furnace at 950" to 1000" C.in an atmosphere of nitrogen, until the metal had melted. The resulting bar of germanium was smoothed with sandpaper. The electrode was mounted in a glass tube with Kronig's cement. The excess of cement was removed with benzene and the electrode was then washed with alcohol and distilled water. The glass tube was filled with mercury to make electrical contact with the metal. An electrode prepared in this manner could be immersed completely in any solution in which it was used. Pure calomel was prepared from mercury obtained by the electrolysis of mercurous perchlorate (9). A saturated calomel electrode was prepared from this pure mercury and calomel. The potassium chloride used was Kahlbaum's "for analysis" which was recrystallized twice from distilled water. The measurement of the voltage was made with a Leeds and Northrup Type K potentiometer and Type R No. E. galvanometer, shielded with a grounded copper case. The potentiometer was mounted on four rubber stoppers a n d all wires which were connected to it were either heavily insulated or covered with rubber tubing. A three-cell two-volt storage battery and a Weston standard cell supplied the voltage necessary. The Weston cell was calibrated by the Physics Department of Cornell University and found to give a reading of 1.0184 volts, the value of the last decimal place being uncertain. The solution which was measured and the saturated calomel electrode along with the intermediate vessel were immersed in a thermostat which was regulated a t 25' * 0.05" C. The electrode was first studied to ascertain if it gave a definite reproducible electromotive force and, if so, the time required for the electrode to come to equilibrium.
ANALYTICAL EDITION
SEPTEMBER 15, 1935
351
TABLEI. TITRATIONS
NaOH
Hydrogen Electrode PH E . m . f.
311.
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 22.0 24.0 26.0 28.0
VOll
1.98 2.27 2.75 3.69 4.39 5.64 6.70 7.40 8.56 9.26 10.17 11.14 11.55 11.78 11.94
0.3629 0.3803 0.4081 0.4636 0.5053 0.5782 0.6412 0,6826 0,7509 0.7927 0.8461 0.9070 0,9277 0.9412 0.9503
Nechanical Stirring Time Read of after E . m. f. stir. stir. Volt Min. Min. 0.285 30 20 0.293 7 18 0.304 7 12 0.327 7 15 0.353 7 27 0.401 7 12 0.462 .. 22 0.512 7 12 0,594 6 19 0.645 8 14 0.695 7 12 0.735 6 12 7 11 0.742 0.749 8 20
...
..
..
Germanium Electrode Oxygen Stirring Time Read of after E. m. f . stir. stir. .Win. Xin. 0.164 10 10 0.181 5 7 0.198 5 9
0:295 0.346 0.404 0.455 0,528 0.575
0.629 0,683 0.692 0.704
...
..
8 4 3 4 3 3 4 4 5 3
..
Air Stirring Time
E. m. E. 0.246 0.262 0.279 0,306 0.329 0.380 0.440 0.492 0.572 0.628 0.685 0.714 0.726 0.740
6 10 10 13
6 7 16 8
8
4
..
...
of
stir.
Read after stir.
Min.
.Win.
15 6 5 5 5 5 6 5 6 5 5 6 6 6
19 18 7 10 12 13 18 12 14 9 22 24 19 14
..
-
Fifty milliliters of a Prideaux-Ward universal buffer solution (3) were placed in a small beaker and an excess of insoluble ger-
manium dioxide was added. The stick electrode was placed in the solution, the saturated calomel electrode was connected by means of a saturated potassium chloride bridge, and an intermediate vessel filled with a saturated solution of potassium chloride. The solution with the electrodes in contact was stirred mechanically and the initial reading taken during the stirring; the stirring was stopped and a second reading taken. This procedure was repeated at intervals throughout each observation.
Three experiments of this type showed that the electrode did not come to equilibrium in the course of a day, but that with constant stirring during the last 20 minutes of the observation, the value did not change greatly, thus giving the appearance that a constant value had been reached. Duplicate values could not be secured and the values obtained while the solution was being stirred were lower than those when the solution was not stirred. This observed drifting of the electromotive force might be due to the solution in which the readings were made, a characteristic of the electrode itself which could be corrected only by proper pretreatment, or the particular gas which is present in the solution being measured. In order to ascertain if the drifting was due to the type of Solution used, experiments similar to the above were made using a 0.05 M potassium hydrogen phthalate solution and McIlvaine’s buffer solution of pH 2.2. The readings were made both with no stirring and no insoluble germanium dioxide in the solution and with the insoluble germanium dioxide present and intermittent stirring. I n all these experiments no reproducible constant value was obtained and any apparent equilibrium established was always destroyed by allowing the electrode to stand for a longer period thereafter. Apparently neither the presence nor absence of the insoluble germanium dioxide, nor the character of the solution in which the measurements are made is entirely responsible for the drifting shown by the electromotive force. Attention was next directed to different methods of treating the surface of the electrode. It was hoped that such treatment might, as found with the antimony electrode, clean the electrode or coat it with a stabilizing f l m so that it would rapidly assume a reproducible equilibrium value in the solution. Treatment with nitric acid (1 to l),plating with germanium, and treatment with hydrogen peroxide (perhydrol) were tried. For the treatment with nitric acid, the electrode was immersed in a solution of nitric acid (1 to 1) for varying intervals of time. After each treatment with the acid, it was washed thoroughly with distilled water and then immersed in distilled water until it was used. This method of treatment did not yield an electrode which reached a constant reproducible electromotive force within a reasonable length of time. The
values obtained varied about 40 millivolts over a period of one hour and were always higher when there was no insoluble dioxide present. The effect of plating the electrode with germanium from a solution of germanium sulfate was next investigated. Measurements made with this electrode, in a manner similar to that used with the nitric acid-treated electrode, also showed large variations in the electromotive force over periods of 1 to 4 hours.
PH
FIGURE1
To investigate the effect of treatment with hydrogen peroxide, an electrode was immersed for 5 to 10 minutes in hydrogen peroxide, washed with distilled water, and then placed in the Prideaux-Ward universal buffer solution in which the measurements were made. When the electrode was not employed it was kept either in distilled water or in 3 per cent hydrogen peroxide. From a total of thirty-four experiments performed with a hydrogen peroxide-treated electrode it was found that a t first there was a great change in the electromotive force but within an hour the value became fairly constant. However, this value is not reproducible and it will change if the electrode is removed and retreated or washed with distilled water.
352
INDUSTRIAL AND ENGINEERING CHEMISTRY
The effect of stirring the solution with different gases was next investigated. Some McIlvaine buffer solution of p H 2.2 containing insoluble germanium dioxide was placed in a rubber-stoppered 150-ml. cylinder. The rubber stopper contained four holes, one mas for the insertion of the germanium electrode and a second for the salt bridge connecting the calomel electrode with the solution. The gas was bubbled through a tube inserted in the third hole and reaching to the bottom of the solution. The excess gas found exit through a stopcock placed in the fourth hole, and opened only when gas was being passed through the solution. Readings of the electromotive force were taken while the solution was being stirred and after the stirring was stopped. Similar experiments were performed using purified air, nitrogen, hydrogen, and oxygen. Stirring with nitrogen appeared to give a fairly constant value of 0.273 volt on the first day at the end of about 2 hours, but on the second day it gave a value of 0.293 volt after 100 minutes. Hydrogen also failed to give a reproducible value. Air gave an initial value of 0.252 volt which changed to 0.272 volt over a period of 0.75 hour. Oxygen gave an initial value of 0.235 volt which varied between this value and 0.237 volt over a period of 1 hour. Of these four gases, nitrogen and hydrogen gave widely varying results, air gave more constant values, and the values with oxygen were fairly constant throughout the observation. Although it was impossible to secure constant and reproducible electromotive force measurements with the germanium-germanium dioxide electrode in any of the buffer solutions used, it was deemed advisable to determine the relationship between p H and the electromotive force of this electrode. To obtain a variable pH, the Prideaux-Ward universal buffer solution was used and titrated with 0.2 N sodium hydroxide solution. The relationship between the volume of sodium hydroxide used and the p H of the buffer solution was determined by measurements with a Bunker type of hydrogen electrode. The titration vessel, salt bridge, and saturated calomel electrode wereallimmersedin a thermostat, kept a t 25’ * 0.05’ C. Three titrations were made and the average results and the calculated p H values are given in Table I and Figure 1, curve A. The titration of the Prideaux-Ward buffer solution was repeated with the germanium-germanium dioxide electrode in combination with a saturated calomel electrode. Fifty milliliters of the buffer solution were placed in the titration cell, an excess of germanium dioxide was added, and the electrode inserted. Three different titrations were made, stirring the solution mechanically (curve B) and with oxygen (curve C) and air (curve D), both of the gases being first passed through potassium hydroxide solution and water. The solutions were first stirred and after the stirring was stopped the electromotive force was measured a t intervals until it did not show much change in value. Following this some sodium hydroxide was added, the solution stirred for several minutes, and the intermittent reading repeated until an apparent constant value was obtained. This general procedure was used in all the titrations. The same apparatus was used for the gas stirring as in the previous experiments, except that there was one additional hole in the stopper for the buret. The results of these titrations with the time of stirring and the time after stirring when the apparent constant value was obtained, are also given in Table I and Figure 1.
VOL. 7, NO. 5
From the preceding titrations it appears that the germanium-germanium dioxide electrode shows a potential difference against the saturated calomel electrode which is somewhat dependent upon the p H value of the solution. Mechanical stirring or stirring with purified air or oxygen did not give a straight-line relationship for the titration of the Prideaux-Ward buffer solution, although the titration performed by stirring with oxygen approached it fairly closely. In this latter case the variation approaches the relationship E = 0.022 0.059 p H between a p H of 3 and 11. Equilibrium between the germanium electrode and the solution in which it is immersed cannot be obtained by long standing, stirring with purified air, nitrogen, hydrogen, or oxygen, or by special treatment of the electrode surface. This characteristic of the electrode is shown by the fact that it is impossible to obtain reproducible measurements in different samples of the same solution whether they are made on the same day or different days. In the study of the antimony electrode the presence of the trioxide made it possible to determine p H (7) from the electromotive force readings, but the presence of the insoluble germanium dioxide along with the metal electrode does not have a similar effect. If a stabilizing film does exist, it is different from any which has been employed here. It is apparent that the electrode cannot, in its present condition, be used for accurate determinations of p H values but it may be of value in potentiometric titrations in which only the equivalence point of the reaction is sought. The results of the investigation of this use of the germanium-germanium dioxide electrode will appear in a subsequent paper.
+
Summary A new germanium-germanium dioxide electrode, described above, shows a potential difference against a calomel electrode which is somewhat dependent upon the p H of the solution. The electrode was tested for p H determinations and found to be unsatisfactory. Treatment of the electrode surface with nitric acid (1 to l), or hydrogen peroxide, or plating it with germanium failed to stabilize it. Stirring with purified oxygen, nitrogen, hydrogen, or air failed to stop the drifting of the electromotive force.
Literature Cited (1) Bottger, 2. p h y s i k . Chem., 24, 253 (1897). (2) Britton, H. T. S., “Hydrogen Ions,” pp. 57-88, London, Chapman and Hall, 1929. (3) Britton and Robinson, J . Chem. Soc., 125, 426 (1924). (4) Clark, W. M., “Determination of Hydrogen Ions,” _DD. _ 418-29. Baltimore, Williams & Wilkins Co., 1928. (5) Hildebrand, J . Am. Cheq. Soc., 35, 847 (1913). (6) Kolthoff, I. M., and Furman, N. H., “Potentiometric Titrations,” 2nd ed., pp. 215-47, New York, John Wiley & Sons, 1931. (7) Kolthoff and Hartong, Rec. trav. chim., 44, 113 (1925). (8) Laubengayer and Morton, J. Am. Chem. Soc., 54, 2318 (1932). (9) Newberry, Ind. Chemist, 5, 289 (1929). (10) Tressler and Dennis, J . Phys. Chem., 31, 1429 (1927).
RBCEIVBD April 20, 1935. Based upon the thesis presented to the Faculty of the Graduate School of Cornell University by S. R. Cooper in partial fulfillment of the requirements for the degree of doctor of philosophy.
Microchemical Methods in Toxicology A. 0. GETTLER, of New York University and Bellevue Medical College, discussed the subject of ‘LMicrochemical Methods in Toxicology” at the Symposium on Recent Advances in Microanalysis, held April 26, 1935, at the New York Meeting of the American Chemical Society, and gave interesting accounts of
many of his experiences in medico-legal work as toxicologist to the City of Sew York. He also described his work in proving that a small quantity of ethyl alcohol is normally present in the human brain, in which research he employed microtechnic extensively.
