A N A L Y TI C A L E D I3‘16N ’
580
Before titrating, 3 CC. of 10 per cent potassium iodide are added and the customary 3 minutes allowed t o elapse. The titration is then carried out as above, using 0.03 N sodium thiosulfatei It will not be necessary for all operators to use starch in detecting this end point. The 0.01 N sodium thiosulfate solution is standardized by means of pure arsenious oxide, treating from 2.5 to 3.5 mg. A blank should be run and deducted. In order to check the arsenic factor thus obtained, the sodium thiosulfate was also titrated against resublimed iodine and against the iodine liberated from potassium iodide by potassium iodate. The three factors obtained checked within experimental error.
SUMMARY The arsenic content of organic compounds may be determined by means of a sodium peroxide-sugar explosion in a
Vol. 6, No. 5
nickel microbomb mu& more rapidly than by existing microme degree of accuracy. A duplicate e weighings and final titrations, de may be made in 50 minutes.
ACKNOWLEDGMENT The authors are indebted to H. E. Woodward and F. Bremner who checked the analyses using the standard micromethod of Pregl ( 3 ) .
LITERATURE CITED (1) Beamish, IND. ENG.CHEM.,Anal. Ed., 5, 348 (1933).
(2) Parr, J . Am. Chem. SOC.,30, 746 (1908). (3) Preg1;- “Quantitative Organic Microanalysis,” 2nd ed., pp. 160-3, J. & A. Churchill, London, 1930. (4) Pringsheim, Am. Chem. J . , 31, 393 (1904). (5) St. Warunis, Chem.-Ztg., 36, 1205-6 (1912). RBCBIVED June 8, 1034.
Removal of Samples of Oil from OilImpregnated Paper JOHND. PIPER,The Detroit Edison Company, Detroit, Mich.
W
HEN oil and paper high-voltage cable insulation is has not been investigated thoroughly, the work done here subjected to conditions of service or to laboratory indicates that the magnitude of such errors is small as comtreatments simulating such conditions, certain pared with those due to the use of a solvent. I n the first definite changes frequently take place in the oil. In order two mechanical methods tried, both of which involved pressing to investigate these changes, it is necessary to remove the oil the oil from the paper, difficulty was experienced in obtaining from the paper without contaminating the oil. Extraction the oil free from fine paper fibers which interfere with vismethods have been found to interfere with the viscosity and cosity measurements and in obtaining sufficiently large used by this company. As is well known, samples of oil from tapes of relatively low oil content. hydrophil tests (W,4) A centrifuge method, that is rapid as well as effective, has the removal of the last traces of solvent from a heavy oil is a difficult and uncertain process. Since the viscosity of a been found to eliminate the difficulties enumerated. The apparatus is shown in Figure 1. Except as otherwise system of two components of widely differing viscosities is greatly changed by small variations in the concentration of indicated, all measurements are outside dimensions. I n use, a tape from which oil is to be removed is wound on the component of lower viscosity (I), efforts to follow chemical changes in oils by means of viscosity measurements are the reel, A . After the reel has been reversed a few times to partially invalidated by the presence of variable traces of loosen the winding, the roll is pushed off the shaft by means of solvent. Another objection to the use of a solvent is that in the sliding guide and is placed in the funnel with the axis of addition to removing the oil from the paper, it may remove the roll parallel to that of the funnel, as shown. The wide certain substances which are not soluble in the oil and which end of the funnel is then covered with a rubber cap to prevent air from circulating through are not wanted in the specithe apparatus d u r i n g t h e men obtained, as, for incentrifugalizing process, stance, natural resins. In The apparatus is assembled such cases the values obby dropping into the centritained by using the hydrofuge tube, in the order given, phi1 test to measure the exthe lower mounting oomtent of oxidation are too plete with vial, the upper high. mounting, and finally the In order to avoid these loaded funnel, the stem of errors, several mechanical which projects just inside methods, other than those FIGURE1. APPARATUSFOR REMOVING OIL FROM PAPERBY the mouth of the vial. requiring the use of solvents, CENTRIFUGALIZING The length of time and h a v e been t r i e d i n t h i s A . Reel with slotted shaft and sliding guide the speed required to centricompany. It was recogE. Pyrex funnel. Body, i y l u d i n g taper: 90 mm. long X 38 mm. diameter, stem angle 58 , Stem 30 mm, long, 7 mm. diameter fugalize most of the oil from nized that the oil obtained C. Platinum filtering cone, angle 60’ the tapes vary, of course, by such methods might not D. Rubber rap E . Upper mounting: aluminum cylinder 39 mm. long X 38 mm. diw i t h t h e n a t u r e of t h e be entirely representative amet,er, 6 mm. wall. Top 12 mm. beveled on inside to 1mm. thickness t o support rubber ring which fits both cylinder and shoulder of funnel samples. Often a 20-minute of the total oil in the paper, F . Weighing vial with glasa stopper period at about 2000 r. p. m. owing to the selective adG. Lower mounting: aluminum cylinder 52 mm. long X 38 mm. diameter X 3 mm. wall. Cylinder holds two g u m rubber stoppers, the upper suffices. By continuing the sorption by the paper of bored to fit vial F H . Assembly complete in centrifuge tube 137 mm. long X 39 mm. inside process, so much oil may be some of t h e i m p u r i t i e s diameter removed that the paper is. present. While the matter Roll of impregnated sample in tube I.
*I I
September 15,1934
INDUSTRIAL AND ENGINEERING CHEMISTRY
no longer translucent but is opaque. Even when a longer period is required, the time?-consumed is not considered excessive. Unless fine suspended matter is present in the oil in the paper, the sample obtained by the method is clear and free from sediment. The use of the filtering cone is recommended for straining out cable wax or bits of torn paper. Paper fibers do not seem to be dislodged during centrifugalizing. Not only is the method satisfactory for ordinary samples, but it has also been found to be applicable to samples containing abnormally low quantities of very viscous oils. Since the oil is deposited in a small weighing vial, it can be weighed for test and used without loss. As has been intimated, one of the reasons the device was designed was to obtain oil for viscosity tests. That the vis-
381
cosity of the oil taken from radial samples of aged cable varies was observed in the investigations in this company. The possible significance of this phenomenon as regards electric transmission cable life in service is reported in the literature (8). LITERATURE CITED (1) $endall, ,James, and Monroe, K. P., J. Am. Chem. SOC.,39, 1787
(1917). ( 2 ) Shanklin, G . B., and McKaye, G. M. J., Trans., Am. Inst, Elec. Engrs., 48, 338 (1929). (3) Wyatt, K. S. Ibid., 52, 1025 (1933).
(4) Wyatt, K. S., Spring, E. W., and Fellows, C H., Ibid., 52, 1035 (1933). RECEIVED May 14, 1934. This work forms part of a general research on the deterioration of high tenaion underground cable being undertaken by The Detroit Edison Company
ter Meulen Method for Direct Determination of Oxygen in Organic Compounds Containing Nitrogen W. WALKERRUSSELLAND MAURICEE. MARKS,Metcalf Laboratory, Brown University, Providence, R. I.
I
T HAS been shown recently (3) that when certain modifications are employed, the ter Meulen method for the direct determination of oxygen in organic combination yields very satisfactory results in the case of compounds containing carbon, hydrogen, and oxygen. By using a very active, thoria-promoted nickel catalyst, all the oxygen in the organic compound was converted to water which was completely retained by the first calcium chloride tube of the absorption train, since no oxides of carbon escaped methanation. Under such conditions the analysis of organic compounds also containing nitrogen should yield hydrogen, hydrocarbons, nitrogen, water, and ammonia. Since calcium chloride absorbs ammonia to a certain extent, ter Meulen (1) has proposed the use of a special tube carrying a charge of standard acid and in a second compartment calcium chloride. Any ammonia evolved during analysis is retained by the acid which is back-titrated a t the end of the run. Thus the gain in weight of the tube due to ammonia can be evaluated and the water absorbed ascertained by difference. It appeared to the authors that in the absence of oxides of carbon in the effluent gases, procedure might be considerably simplified if an efficient water absorbent could be found which would not absorb ammonia. Reagent-grade sodium hydroxide in pellet form proved to meet these requirements. Pellets are to be preferred to a more finely divided form because they offer little chance for mechanical retention of ammonia. APPARATUS AND METHOD The apparatus previously described elsewhere (3) can be used without modification other than the replacement of all calcium chloride and ascarite with pellets of reagent-grade sodium hydroxide. A 5-inch (12.5-em.) Schwartz tube charged with this sodium hydroxide served to retain completely the water corresponding to the oxygen in the sample. A second and third tube similarly charged, t o act as guard tubes, complete the absorption train. The method of analysis is identical with that already described (S), except that for compounds containing much nitrogen the time for an analysis may well be extended to 1.25 or 1.5 hours, in order to allow ample time for sweeping am-
monia from the system. A relatively large amount of catalyst-5 to 10 grams-is desirable to prevent the appearance of oxides of carbon in the effluent gases to be absorbed by the sodium hydroxide. Such an amount of active catalyst gives persistent blanks. Thus the blank per half hour may be as high as 2 mg. with a freshly prepared catalyst. The initial blank gradually decreases with continued use of the catalyst to a few tenths of a milligram, owing to the slow reduction of the last traces of nickel oxide. Therefore, the values of these blanks must be carefully ascertained and used to correct the results obtained. In case there is a difference between the blanks determined before and after an analysis, the average blank is best employed. TABLEI.
RESULTS OF ANALYSESBY MODIFIEDTER MEULEN METHOD TEMPERATURE OF --OXYOENCATACalmSAMPLE LYST Found lated Gram ' C. % % 0.20685 350 27.14 27.10 0.1910 350 27.14
WEIGHT OF
SUBSTANCE Acetamide
NITROQEN
CON-
VERTED
TONHX % 85 58
Dimethylglyoxime
0.18615 0.2117
350 350
27.50 27.61
27.57
45 53
Urea
0.1939 0.1497
300 350
26.59 26.71
26.65
33
3,5-Dinitrobenzoic acid
0.1596 0.2010
350 350
45.11 45.22
45.27
60
p-Nitroacetanilide
0,20745 0.1946
350 350
26.70 26.88
26.66
58
p-Nitrobenzeneazoresorcinol
0,2295 0.2337
350 350
24.53 24.61
24.70
54
RESULTB The results of analyses made upon high-grade organic chemicals containing nitrogen, with sodium hydroxide employed as absorbent, are given in Table I. In addition to the oxygen content found, the percentage of the calculated nitrogen appearing in the exit gases as ammonia is also given. Although over 50 per cent of the nitrogen frequently appears as ammonia, it is improbable that a simultaneous de-
