ANALYTICAL EDITION
82
Oxides of nitrogen and free bromine, if present in the solution when potassium iodide is added, will cause results to run high and may cause a return of the blue after the end point has been reached. The precipitate of ferric hydroxide should dissolve completely when shaken with ammonium bifluoride. If it does not, either too much ammonium hydroxide or too little ammonium bifluoride was added. The results given in Table VI1 were obtained by treating a copper sulfate solution of 0.2 gram of copper containing the indicated impurities by the above method. The pH values of the solutions at the end point were determined by means of the quinhydrone electrode. From an inspection of the results given in Table VI1 it is apparent that neither iron nor arsenic affect the copper value, as determined by this method, except in the presence of manganese. Manganese probably acts as catalyst for the reaction between ferric iron and iodide and speeds it up to such an extent that the fluoride present is not capable of completely preventing it. Fortunately manganese is not a common impurity in copper ores. When it is present, iron must be removed before the addition of ammonium hydroxide by one of the well-known methods.
Vol. 3, No. 1 Literature Cited
Bassett, “Theory of Quantitative Analysis,” p. 265, Rutledge, 1925. Blanc, J . chim. phys., 18, 28 (1920). Bray, J. Phys. Chem., 6, 365‘(1902). Clark, U. S. Pub. Health Service, Pub. Health Repls. 38, 443-65 (1923). (5) Dhar, Z . anorg. allgem. Chem., 153,330 (1926). (6) Fleury, Bull. soc. chim., 27, 490 (1920). (7) Foerster and Pressprich, 2.Elektrochem., 33, 176 (1927). (8) Goard and Rideal, Proc. Roy. Soc. (London),105A, 135 (1924). (9) Hughes, J. Chem. Soc., 491 (1928). (10) Jones and Kaplan, J . Am. Chem. S O L ,50, 2066 (1928). (11) Kolthoff, Pharm. WeekbZad, 56, 621 (1919). (12) Kolthoff, Z. anal. Chem., 60, 393 (1921). (13) Kolthoff and Furman, “Volumetric Analysis,” Vol. 11, p. 350, Wiley, 1929. (14) Low, “Technical Methods of Ore Analysis for Chemists and Colleges,” Wiley, 1927. (15) Mott, Chemist-Analyst, 1912, 5-7. (16) Pedersen and BjergaaFd, Dansk Tids. Farm., 2, 1-7 (1928); C. A., 22, 1116 (1928). (17) Roebuck, J . Phys. Chem., 6, 365 (1902). (18) Roebuck, Ibid., 9, 727 (1905). (19) Rosenthaler, Z. anal. Chem., 45, 596 (1906). (20) Thiel and Meyer, Ibid., 66, 177 (1916). (21) Washburn, J. Am. Chem. Soc., 30, 31 (1908). (22) Washburn and Strachan, Ibid., 35, 681 (1913). (23) Washburn and Strachan, Ibid., 35, 694 (1913). (24) Wood, J. Chem. Soc., 93, 411 (1908). (25) Zswidzky, Ber., 36, 1427 (1903).
(1) (2) (3) (4)
Potentiometric Determination of Acidity in Insulating Oils’ R. N. Evans and J. E. Davenport RESEARCH BUREAU,BROOKLYN EDISON COMPANY, INC.,BROOKLYN, N. Y.
N THE past the deter-
The estimation of acidity in oils employing the alkali buret was j CC. in capacity mination of acidity in blue procedure has been found to give more satisfactory and graduated to 0.02 cc, and oils has been accomresults than the A. S. T. M. procedure. In the latter was permanently connected plished generally by a procemethod, methyl alcohol should not be substituted for to the source of alkali. All ethyl alcohol. dure similar to that outlined precaution was taken to exin the A. 8. T. M. Tentative Electrometric titrations have been carried Out O n clude carbon dioxide. The type organic acids successfulb‘. Standards 1929, page 397. solution was stirred with comThe silver-silver chloride electrode can be used pressed tank nitrogen except Recently Seltz and McKinney ( E ) described a quinhydronesatisfactorily in routine determination Of acidity. when a volatile acid was tipotentiometric method emThe most important disadvantage of the electrotrated. I n this case, the metric procedure is the time required. It is recomsolution was stirred mechaniplaying amyl alcohol as a solvent, lithium chloride as mended that a preliminary approximate titration be cally, or approximately ninethe conducting salt, and a carried Out. tenths of the alkali was added and the remainder of the tiboundary of agar-agar containing lithium chloride to separate the reference electrode tration carried out with nitrogen stirring. When using the from the solution under investigation. Later, Seltz and Silver- agar-agar boundary, the reference solution was n-butyl alcohol man (6) eliminated the agar-agar boundary and substituted saturated with respect to potassium chloride and benzoic acid. the silver-silver chloride electrode as the reference electrode. The solvent was n-butyl alcohol saturated with potassium A comprehensive survey of the past work on potentiometric chloride. The alkali was 0.05 N and was standardized potentiometrically against B. S. benzoic acid. It was prepared by titration may be had by reference to a paper by Furman (2). The following paper deals with the comparison of the elec- dissolving c. P. stick potassium hydroxide in n-butyl alcohol trometric method and several other procedures, empIoying and the insoluble carbonate removed by filtration. The indicators in which known acids were dissolved in a trans- quinhydrone was crystallized from n-butyl alcohol and melted former oil. Many factors influencing the electrometric sharply at 170”C. titration have also been experimentally studied.
I
Experiments
Apparatus
The diagrammatic arrangement of the electrometric aPparatus may be Obtained by reference to the previously mentioned paper by Seltz and McKinney. The Potentiometer was a Biddle high resistance instrument, accurate to 0.1 millivolt, used with an external galvanometer. The 1
Received September 4, 1930.
Known acids were dissolved in oil with subsequent determination of neutralization number in order to compare the results obtained employing four different procedures-namely, A, 8. T. M. using methyl alcohol as a solvent, A. S. T. M. using ethyl alcohol as a solvent, alkali blue, and electrometric. The latter method was carried out with the two previously described reference electrodes (5, 6). The concentration
January 15, 1931
INDUSTRIAL AND ENGINEERING CHEMXTRY
of acids was chosen a t random and in the case of the higher concentration of stearic acid in oil, a supersaturated solution was obtained. On long standing, stearic acid crystals separated and a more dilute solution was prepared. In each procedure, except the electrometric, the blank on the solvent was obtained by titration to a definite shade, and the solvent then added to the oil sample, whereupon the titration was carried out as near as possible to the same shade. In the case of the A. S. T. M. procedure using methyl alcohol, the point was difficult to obtain. A faint pink color is soon
400 600
300 500 200 400 >L;
il4 100 300 zoo
Cc. of Alkali Figure 1-Electrometric Titration I, Stearic acid; 11, Abietic acid
obtained in the alcohol-water phase which does not appear to increase in depth until the end point recorded in the table was reached. Since the color change should be observed in the alcohol-water phase, care should be taken to allow the phases to separate even though working with an almost colorless oil. Table I is a tabulation of neutralization numbers obtained by titrating oil samples employing different methods. Table I-Neutralization Numbers of Acids Dissolved in Transformer Oil NAPHTHENIC STEARICSTEARIC CYCLOHEXANE ROSIN METHOD BENZOIC ACETIC (I) (11) CARBOXYLIC TECHNICAL CHsOH, 50 cc. Ha0,50cc. 2.23 0.50 0.26 0.64 2.14 0.99 0.92 Phenolphthalein (l%).lcc. CzHsOH, 50 cc. MaO,50 cc. 2.29 0.52 0.37 0.91 2.15 1 . 0 3 0.97 Phenolphthalein, 1 cc. CHaOH, 75 cc. Benzene 75cc. 2.29 0.51 0.42 1.00 2.18 1 . 0 2 0 97 Alkali b
