V O L U M E 2 6 , NO. 12, D E C E M B E R 1 9 5 4 (9) Gray, E. LeB.. AIacXamee, J. K., and Goldberg, S. B.. Srch. I n d . Hug. and Occupational Med., 6, 20 (1952). Haagen-Smit. A. J., I n d . Eng. Chem., 44, 1342 (1952). (11) Holler, A. C.. and Huch, R. V., ANAL.CHEM.,21, 1385 (1949). (12) Jacobs, AI. B., “The Analytical Chemistry of Industrial Poisons, Hazards, and Solvents,” 2nd ed., p. 36S, New York, Interscience Publishers, 1949. (13) Johnston, H. S.,and Yost, D. M., J . Chem. Phys., 17,386 (1949). . 16, 766 (1944). (14) Kieselbach, R., ISD. ENG.CHEM., d s a ~ED., (16) LaTowsky, L. JV., ct al., J . Ind. Hug. Tosicol., 23, 129-47 (1941). (16) Patty, F. A , and Petty, G. 31., Ibid., 25, 301 (1943). (17) Reindollar, W.F., ISD. EXG.CHEM.,;ISAL. ED., 12, 326 (1940). (18) Rider, B. F., with IIellon, 31. G., Ibid., 18, 90 (1940). (10)
1955 (19) Shinn, 11. B., Ibid.. 13.33 (1941). (20) Shnidman, L., and Yeaw, J. S., A7n. Gas Assoc. Proc., 24 (1942), 277. (21) Stanford Research Institute, “Third Interim Report on the Smog Problem in Los Angeles County,” 1950. ( 2 2 ) U. S. Public Health Service. Public Health Bull., KO. 272, 1941. (23) Usher, F. L., and Rao, B. S., J . Chem. Soc., 1 1 1 , 799 (1917). Arch. Ind. (24) Wade. H. A, Elkins, H. B., and Ruotolo, B. P. W., Hyg. and Occupatzonal N e d . , 1, 81 (1960), (26) Yagoda, H., and Goldman, F. H., J . Ind. Hug. Tosicol., 25, 440 (1943). RECEIVED for review J u n e 2 1 , 1954. Accepted September
1, 1954.
Extension of Isotopic Method for Determining Oxygen in Metals To Copper Containing 0.01 to 0.1 Weight yo of Oxygen A. D. KIRSHENBAUM and A. V. GROSSE Research lnstitute o f Temple University, Philadelphia, Pa. The isotopic method. originally developed for the determination of oxygen in organic compounds and fluorocarbons and recently applied to metals and metal-oxygen alloys in the oxygen range of 0.2 to 30 weight yo, has now been extended to the oxygen range of 0.01 to 0.1 weight 70 in copper samples. Although the method, niaking use of oxygen-18, does not require quantitative separation or recovery of the oxygen, it gives accurate results rapidly. The only requirement of this method is that all of the oxygen atoms in the system be exchanged at the temperatures and under the conditions used.
I
N T H E pa.st the oxygen content of metals has been determined directly by several different methods-the
wire 0.23 mm. in diameter, G, into the platinum tube centered in t h e induction furnace, F , 10 em. long, 64 mm. in internal diameter, with the platinum reflector 0.07 mni. thick centered 1.3 em. outside the furnace. The crucibles can be made out of graphite, hydrogen-reduced molybdenum or platinum, the latter tlvo being preferred. The graphite crucibles 1.6 cm. in outside diameter, 3 em. long with a wall thickness of 0.4 to 0.5 em., weighing 1.5 to 1.8 grams, were made from Dixon E-821 grade graphite rods. T h e y had a n oxygen content of 0.085370 by weight. T h e h?-drogen-reduced molybdenum and platinum crucibles had a wall thickness of 0.1 mm., I\-cighed 1.5 to 2 grams, and had a n oxygen content of 0.000170. EXPERIMEKTAL PROCEDURE
I n determining the oxygen content of the copper samples, a known weight of t h e copper t,o be analvzed (8 t o 10 grams) was mixed with a known weight (about 100 mg.), of oxygen-18 labeled
vacuum-fusion
(1, 14-17, 19, 80, ai?), carbon reduction (12, 18), hydrogen reduction (2, 2 1 ) , and chemical methods (S-5> I S ) , and indirectly by difference (13). The vacuum-fusion and hydrogen reduction are
a t present the most popular methods for the analysis of copper samples. Both methods, however, require quantitative removal a n d recovery of a11 the oxygen in the metal. T h e “isotopic method” (6-8, l l ) , applied recently t o metals and metal oxides (9, I O ) , does not incorporate this difficulty because quantitative removal and recovery of the oxygen are not necessary.
:% A ;I
L
APPARATUS
rswntially of :t 6-ku. COI Tht, apparatus con high-fwquency induc furnitce, and a vacuum separating and sampling gases. High-Frequency Induction Furnace. T h e furnace used is a n Ajas-Sorthrup 6-liw. ronverter-type high-frequency induction furnacr. manufactured hy the .Ijax Electrothermic Coqi., Trenton, h-,J. Glass Vacuum System. T h e apparatus as shown in Figure 1 consisted (wentially of it vei,tic:tl platinum or molybdmum reacrl :tnd a Imrosilicate glass vacuum system. Th! reacVI (:onsirttd of a platinum test tube, H , 32 mm. in outcstf’r, :30 em. long, with a all thickness of 0.4 mni., centered in :tn rv;tcuntcd cju;wtz tuhe, E , 64 mm. in outside diameter, nd sealed with de Khotinsliy cement t o a braes I, B , 5 cni. long. T h e platinum tube was I f a i.5-cni. long water jacket, C, 2.5 cm. below thc d ( ~Iihotinsky seal outside the induction furnare, F . IKW h m d \ v i t ~rc4ed Tvith de Khotinsky cement to a boror-xiij. vacuum stopcock which in turn was connected to i t Tocsplei. pump, J , 3 mm. in outsidc diameter sampling tuhes. K . itnd t h r vit(~uunisystpm. .4n optically clear sight-ghss \vindn\y, 9,WLP ~ ~ a l teodthe top of the brass watercooled head for temperaturp readings. A metal hook on which a crucible, I , ant1 :I platinum reflector, D , were hung, was centered in the hcad. T h e crurible n - a p hung b y means of moll-bdenum
/
Figure 1. A. B. C. D.
Sight glass ’Water-cooled head Water jacket Platinum reflector
K
Glass Vacuum System
E. Quartz tuhe 1. Graphite crucible F. Induction furnace J . Toepler pump 6. Molybdenum wire K . 3-mm. sampling H . Platinum tuhe tubes
1956
ANALYTICAL CHEMISTRY
copper-oxygen “master alloy” containing a given amount of oxygen of known oxygen-18 concentration and placed in a hydrogen-reduced molybdenum, platinum, or baked graphite crucible (baked a t 1200” C. for 2 hours while pumping a t
