76
INDUSTRIAL A S D ESGINEERING CHEMISTRY
other anions which can be oxidized by bromate in acid solution. However, the selenium must be in the colloidal condition, or be oxidized directly to selenious acid, in order that the Norris and Fay method for selenious acid may be subsequently applied.
Summary A rapid and precise volumetric method for the determination of sulfite and selenium simultaneously in the same sample of solution has been developed. The results of the method agree within 4 parts per 1000 for selenium and 2 parts per 1000 for sulfite with the best gravimetric results obtained. The
VOL. 12, NO. 2
principle of the method suggests other procedures for the application of the bromate titration of colloidal selenium.
Literature Cited (1) Coleman, W. C., and McCrosky, C. R., IND. ENQ.CHEM.,Anal. Ed., 9, 431 (1937). (2) Coleman, W. C., and McCrosky, C. R., J . Am. Chem. Soc., 59, 1458 (1937). (3) Latimer, W. M.. "Oxidation Potentials", New Tork, PrenticeHall, Inc., 1988. (4) Lenher, V., and Kao, C. H., J . Am. Chem. SOC.,47,2454 (1925). ( 5 ) Norris, J. F., and Fay, H., Am. Chem. J . , 18, 704 (1896). (6) Sullivan, V. R., and Smith, G. F., J . SOC.Chem. Ind., 56, 104T (1937).
Analysis of Naphthenic Acids J. R . RI. KLOTZ
AND
EDWILY R. LITTRIANN, Stanco Incorporated, Elizabeth, N. J.
D
URIXG the course of an investigation on naphthenic acids it became necessary to analyze several acids for total acid content and unsaponifiable matter. The customary procedure of extracting the unsaponifiable material with ethyl or petroleum ether from a strongly alkaline solution of sodium naphthenate possesses several disadvantages. The unsaponifiable matter recovered from acids R-hich were extracted from crude oil or gas oil may be too low-boiling to assure a n accurate weight after evaporation of the solvent. Any phenols occurring in the acid would be retained by the alkaline solution and be included in the figure for total acid content. Finally, the customary unsaponifiable matter determination gives no information on the actual acids present in the sample under examination. The present analytical procedure eliminates these difficulties. When a sample of naphthenic acid is exactly neutralized to phenolphthalein with 0.5 N sodium hydroxide, any phenols and nonacidic materials should be removed by shaking out the solution with petroleum ether. The purified naphthenic acids are then recovered from the aqueous solution by acidification, weighed, and titrated. I n the absence of mechanical losses the original and final titrations should be the same and the weight of acid recovered would represent the actual acid present in the original sample. Since mechanical losses are inevitable the weight of acid obtained must be corrected. If the product finally weighed has been freed from nonacidic material, the mechanical losses are proportional to the difference between the original and final titrations and the total acid content is given by the following equation:
Procedure Titrate a 2- to 3-gram sample of naphthenic acid, weighed to the closest milligram and dissolved in 25 cc. of alcohol, with 0.5 N sodium hydroxide, using phenolphthalein as indicator. Dilute a-it,h25 cc. of water, transfer to a 250-cc. separatory funnel, and shake out with two 15-cc. portions of petroleum ether. Return the aqueous layer to the separatory funnel (previously cleaned with alcohol), acidify with 15 cc. of 10 per cent sulfuric acid, and again shake out with 25 cc. of petroleum ether. Discard the lower aqueous layer and filter the petroleum ether layer through a dry paper into a tared, glass-stoppered, 250-cc. Erlenmeyer flask. Rinse the separatory funnel with 10 cc. of petroleum ether which is also used to wash the filter paper. Evaporate the combined petroleum ether solutions on the steam bath and finally to constant weight on a sand bath in which the sand is maintained at 110" to 115" C. After 2 hours' heating weigh the flask, reheat for 30 minutes, and again weigh. Continue the heating until the consecutive readings check within 1 mg. If after 3.5 hours constant weight has not been obtained, use the veight taken at that time, since experience has shown all the petroleum ether will have been removed-. After the final weighing dissolve the acid in 25 cc. of alcohol and again titrate with sodium hvdroxide of the same normality as previously used.
Per cent of acid (regenerated) X original titration = final titration per cent of total acids
I n order to determine the accuracy of the analytical method it was checked against a sample of naphthenic acid which contained 96.0 per cent of total acid, and which had been diluted with Sujol to give a theoretical total acid content of 84.6 per cent and an acid number of 206. The results of the check analyses are given in Table 11.
From the weight of regenerated acid and the final titration the acid number of the regenerated acid may be obtained. The fact that the acid numbers of the regenerated acids in any particular sample check each other closely, despite differences in the percentage of such acids actually recovered, makes i t probable that such acid numbers are characteristic of the acids present in the original sample. The analysis of a check sample prepared by diluting a sample of naphthenic acid with Nujol showed an agreement within 0.55 per cent between the calculated and actual acid content and 0.5 unit between the acid numbers of the recovered acids from the original and diluted sample. The analyses of several commercially available naphthenic acids by the above procedure are given in Table I. The unsaponifiable matter was obtained by difference.
TABLE I. Acid
A4S.4LYSES O F COMMERCIAL XAPHTHESIC
Acid No.
Aruba Romanian Calif. 160 Calif. 250 hlexican Austrian B. R. R.
205 280 I57 255 285 248 227
TABLE
Acid No. of Residual Acid Total Acid 238 295 212 280 302 277 246
11. AKALYSISOF
Sample, grams NaOH (0.590). initial, cc. Acid Xu. Regenerated acid, grams Regenerated acid, % XaOH (0.590). final, cc. Total acid, % Acid No. of regenerated acid
2,569
16.6 206 1.988 77.4 15.1 85.0 243
ACIDS
Unsaponifiable
%
%
86.5 54.7 74.5 51.1 54 1 85.3 92.1
13.5
5,3 25.5 8.5
5.9 12.7 7.5
CHECK SAMPLE
3.905 255 207 3.184 81.5 24.4 85.3 245
2 18 207 2 82
809
2 325 7
17 7
85 0 244
2.816 18.3 20s 2,269 80.6 17.3 85.3 244
On the basis of these analyses the difference between the calculated and actual analysis for total acid is 0.55 per cent and between the calculated and actual acid number is 1.0. The acid number of the regenerated acid in the undiluted
ANALYTICAL EDITION
FEBRUARY 15, 1940
OF COMMERCIAL SAPHTHENIC ACIDS TABLE 111. ANALYSES
Acid Aruba Romanian Calif. 160 Calif. 250 Mexican lustrian B. R . R . Flask and sample, 54 922 32.152 68.695 43.137 53.331 43.351 55 290 32.350 68.654 43 329 grams 68.472 43.058 77.911 8 8 . 0 8 3 65.796 40.432 52.248 29.305 65.797 40.433 50 442 40.398 52.248 29.305 65.796 40 432 Flask, grama 73.004 83.409 2.899 2,705 2.674 2.675 2.625 2.847 2.889 2.953 3.042 3.045 2 . 8 5 8 2.897 4.907 4.674 Sample, grams N a O H (0.5901, cc. 17.2 16.9 ... ... ... ... ... 27.0 27.1 ... .., 35.0 33.2 Na,OH (0.562), cc. ... ,.. 25.7 241 13:4 14.2 23.4 23.9 ... 22.4 22.9 ... ... Acid No. 205 206 280 281 158 157 255 255 285 285 247 249 228 227 Flaek and residual acid, grams 68.009 63.623 68.202 42.416 53.907 30.985 52.668 42.670 54.770 31.791 67.471 42.433 77.304 87.554 Flask. grams 65.797 61.445 65.796 40.432 52.248 29.305 50.442 40.398 52.248 29.305 65.796 40.432 73.004 83.409 R e s i d i a l acid, 1.659 1.680 2.226 2.272 2 , 5 2 2 2 486 2.212 2.178 2.406 1.984 1.675 2.001 4.300 4.145 grama 16.5 16.2 ... ... ... 23.8 23.5 ... 33.2 32.0 N a O H (0.590), 00. ii:ij ii:i5 19.8 20.2 ,.. 22.6 18.6 ... 14 8 17 7 ... ... NsOH (0.562), cc. 62.0 59.0 77.0 82.9 83.2 83.0 73.4 77.0 82:Q 81.6 58 7 69 1 Residual acid, yo 87.6 88.6 74.6 74.4 91.0 86.3 86.8 94.4 95.0 91.0 94.1 94.1 89 2 89.4 92.3 92.0 T o t a l acid, % Acid No. of residual 212 211 238 238 296 295 280 280 302 303 245 247 277 278 acid
...
...
...
sample was 243.5 (average), while that of the check sample was 244 (average). The method therefore appears sufficiently accurate for technical work. The details of the analyses on the several samples of commercial naphthenic acids are given in Table 111.
Discussion The choice of weighing procedure given above depends mainly upon the naphthenic acid being investigated. Since commercial naphthenic acids are complex mixtures, some samples will contain acids which are appreciably volatile a t
the evaporation temperature required to remove the petroleum ether and will be lost during arid after the evaporation of the solvent. However, if the unsaponifiable matter has been completely extracted from the sample and only acid is lost during the evaporation procedure this loss may be quantitatively corrected for by the final titration. The quantitative removal of petroleum ether under the conditions used is indicated by the concordance of the acid numbers of the residual regenerated acids. PRESENTED before the Division of Paint rtnd Varnish Chemistry a t the 98th Meeting of the American Chemical Sopiety, Boston, Mass.
Continuous Laboratory Method for Bodying Oiticica Oil VINCENT RIARCHESE, Columbia University, New York, N. Y., JOSEPH M,ITTIELLO, Hilo Varnish Corp., Brooklyn, N. Y., AND LINCOLN T. WORK, Columbia University, New York, N. Y. Oiticica oil was heat-bodied by a continuous method to viscosities from 3.8 to 22.7 poises employing temperatures of 540' F. (282 ' C.), 590' F. (310 ' C.), and 662" F. (350 C.), and properties of this bodied oil were measured. The apparatus consisted of a RZonel metal heating coil in a lead-antimony bath with thermocouple indication of the metal bath temperature and the outlet temperature. The oil was given a single pass through the coil, the velocity being controlled by a recirculating by-pass a t the pump; and it was quickly cooled after it had completed passage through the heating coil. Viscosities were high for low rates of flow and became substantially constant for each bodying temperature a t the higher flow rates. Acid number increased
S
IKCE oiticica oil was first introduced to the protective
coating industry in this country, its distinctive properties have been studied in comparison with other oils used in the paint and varnish industry (1-4, 6). Sorenson, Schumann, Schumann, and Mattiello ( 7 ) briefly summarized its properties in the introduction to their paper and showed its kettle-bodying characteristics both in air and in an atmosphere of carbon dioxide. I n this investigation it was found desirable to develop a continuous method of bodying for laboratory testing, and to body samples at several temperatures, the resulting oil to be subjected to tests of physical characteristics, gelation time, and livering characteristics.
slightly with the increasing viscosity but w-as generally around 5.5 to 7. Iodine number decreased with increasing viscosity from a Falue of 170 to as low as 140. A curve of gelation time us. viscosity bears some relation to Sorenson. Schumann, Schumann, and Mattiello's curve of gelation time us. temperature. Pastes made with zinc oxide and bodied oiticica oil were subjected to the accelerated livering test and flow characteristics were defined. All samples were in the nonlivering area and showed generally good to excellent flow. The acid number-viscosity plot to differentiate livering areas from nonlivering areas is presented with data on a number of bodied oils to indicate its potential value as a criterion of livering tendencies for oils bodied under normal procedures.
-Apparatus A diagram of the apparatus is presented in Figure 1. It consisted of a container, B , which served as a reservoir for the oil, a pump, A , t o circulate the oil, valves D and C to regulate the amount of oil flowing through the coiled tube, E , an electric furnace, F , containing a lead-antimony bath in which the coil was immersed, and a container, I , cooled by water to receive the bodied oil. The heating coil was of Monel metal 0.3 em. (0.125 inch) in inside diameter and 4 meters (12 feet) long. Two thermocouples, G and H , were used to measure the temperature, the former being in the metal bath to meamre the temperature of the coil, and the latter being a t the outlet to ensure that the temperature had been obtained. Both thermocouples read the same temperature while runs mere in progress.
