Labs., WAPD-CTA (GLA)-203 (April 24. 1956). (5) Hague; J. L., Machlan, L. A,, J . Res. Natl. Bur. Std. 62,53-7 (1959). (6) Huber, F. E., Chase, D. L., Chemist Analpst 50,71(1961). ( 7 ) Laux, P., General Electric Co., Evendale, Ohio, private communication, Dec. 3, 1959. (8) McKaveney, J. P., Vassilaros, G. L., ANAL. CHEM.34,384-8 (1962). (9) Mallett, M. W., “Determination of Oxygen, Nitrogen, Hydrogen, and Car-
bon in Molybdenum, Tungsten, Columbium, and Tantalum,” Memorandum 49, Defense Metals Information Center, Battelle Memorial Institute, March 31,1960; Talanta 9,133-44 (1962). (10) Mitchell, B. J., ANAL. CHEW 30, 1894 (1958). (11) Porter, G., Union Carbide Metals Co., Niagara Fa&, N. Y., private communication, Jan. 8, 1960. (12) Powell, A. R., Schoeller, W. R.,’ Jahn, C., Analyst 60,506-14 (1935). (13) Raber, W. J., “Spectrophotometric ’
Method for Determination of Zirconium in Iron, Nickel, and Cobalt-Base Al-
loys,” Paper 74, Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, March 6, 1962. (14) Reed, J. F., Westinghouse Electric Corp., Pittsburgh 35, Pa., private communication, March 15,1960. RECEIVEDfor review May 27, 1963. Accepted August 21, 1963. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., March 6, 1963.
Method and Apparatus for Determination of Small isotopic Oxygen Variations in Beryllium Oxide R. A. MEYER, S. B. AUSTERMAN, and D. G. SWARTHOUT North American Aviation Science Center, North American Aviation, Inc., Canoga Park, Calif. An analytical method has been developed for detecting small changes in The the 0 I 8 abundance of BeO. method has been tested for 1 to 5-mg. samples of finely ground B e 0 containing 0.2 to 2.0 atom per cent of the A vacuum fusion oxygen as 0 I s . apparatus using a platinum flux contained in a graphite crucible at 2200” C. was used to release the oxygen in B e 0 as CO. The resultant CO was analyzed for C01*-C016 ratio with a modified Consolidated 2 1-620 mass spectrometer. The standard deviation among 19 successive samples was k.002 when the 0l8abundance was approximately 0.2%;.
-
F
OR A N E x P E R I M m i - to determine diffusion coefficients for oxygen in beryllium oxide, a method combining mass spectrometer and vacuum fusion techniques was developed for measuring ratio in finely ground BeO. the 018-01e During the experiment, one face of each B e 0 diffusion block had been plated with and the 0’8 forced to diffuse into the block by annealing in an argon atmosphere at various elevated temperatures. Thus, changes in the 01*0 1 6 ratio of successive serially ground sections could be used to follow the oxygen diffusion into the block. Due to mass spectrometer accuracy problems, all data were referred to the COI8 content of a single cylinder of CO run as a standard interspersed every fifth sample within a series. Since only small changes in 0l8abundance were of interest, no effort was made to determine the absolute 0’8 abundance in the “standard” CO. Methods for the determination of low
2144
ANALYTICAL CHEMISTRY
concentrations of oxygen in beryllium have been reviewed by 13radshaw ( 2 ) . Of the established procedures such as volatilization, differential solubility, inert gas fusion, radio isotope formation, and isotopic dilution, vacuum fusion seemed most applicable to the present problem of liberating the oxygen in B e 0 for isotopic oxygen ratio determination. Gregory and Mapper ( 5 ) , using B e 0 as a standard during the development of a micro method for the determination of total oxygen in beryllium metal, showed that the oypgrn in B e 0 was quantitatively liberated as CO within 4 minutes when reacted with carbon dissolved in a platinum flux a t 1900” C. Their recommendation of a maximum Be concentration in the platinum flux of one part in fifty was based on the determination of small amounts of oxygen in beryllium. Others ( 2 ) have worked on the
problem of quantitative recovery of oxygen from beryllium metal in the range of parts per million to as high as 1%. Brewer (S), Evans and Kubaschewski ( d ) , and Sloman, Harvey, and Kubaschewski (7) in their independent discussions of vacuum fusion fundamentals and thermodynamic considerations generally agree that a mobile liquid bath containing an adequate amount of carbon is a prerequisite for efficient operation of the method. The choice of metals for the bath material is wide, but platinum seems most applicable to the work at hand. Development of the method was simplified hecause the weight of Be added to the system was quite small and thus did not lead to the gettering problems reported by Gregory and Mapper (6). Slso, a reasonable assumption could be made that any isotope effect upon the reaction was small and could be neglected.
OPTICAL
,TO T O E P L E R P U M P 8 -
1
STOPCOCKJO M M RIGHT ANGLE HIGH VACUUM, HOLLOW PLUG
Figure 1 .
Vacuum fusion apparatus
Therefore, the C01s-C016 ratio would not be influenced by nonquantitative recovery of the CO from the B e 0 being analyzed. EXPERIMENIAL
Apparatus. T h e reaction furnace shown in Figure 1 follows accepted vacuum fusion techniipes (6). Samples were contained in a series of loading galleys whose total capacity was 30 samples An external magnet operated the pusher to drop the samples into the crucible. The induction furnace power source was a LePe1 2.5-kw. high frequency generator. Maximum temperature attained was 2400' C. as measured with a n optical pyrometer. Eighteen grams of Pt in a graphite crucible provided the vacuum fusion -4s a purelv precautionary matrix. measure, the Pt and crucible nere changed after every 2 sptq of samples The ~ a c u u msystem was standard and consisted of a rotary mechanical forepump, a n oil diffusion pump, and a liquid nitrogen-cooled trap large capacity Toepler pump wa- uqed to transfer the gaseous r3action products into sample ressels fcr ma+ ipcctrometric analy-is. The niass spectrometer was a Consolidated 21-620 modified t o include a 21-103 type giis in1 Procedure. Samples $11 anallsis were I- to 5 mg portions of finely divided B e 0 ~ r h i c hliad heen serially ground in parillel lavers from t h e face of each diffusion block and collected in carbon tetrachloride. (The collection of c u t > from a given block is referred to hpre as a set.) The B e 0 suspension was concentrated by centrifuging, the iixcess liquid removed, and the B e 0 rempended in 0.1ml. carbon tetrachloride. The resultant suspenqion was transferred t o a 2-sq. em. piece of Pt foil weighing about 0.15 gram. sifter evaporation of the CC14, the edger of the foil were folded over the B e 0 powder t o protect it from mechanical scattering during the loading and dropping processes. The 20 to 25 samples of a given set were loaded into individual depressions iii the qeveral sample galleries and the system was e l ncuated. The system was outgassed at 2400' C. until a pre.mre of 1 X 10-5 mm Hg h a s reached. With a new crucible and Pt bath, 18 grams of platinum spheres were dropped into the crucible from one of the sample galleries and outgassing continued until a pressure of 1 X 10-6 rim. was reached. If the crucible and bath were not changed, the initial oui gassing was continued to a pressure of 1 X 10-6mm After preparatory outgassing was completed, the f u r n x e power was turned off for 5 minutes to allow the temperature of the crucible to drop to about 1550' C. A Pt foil sample container 1%-aspushed from its sample gallery depression w t h an external magnet and its assxiated internal pusher and dropped i ito the crucible. With the system still open to the vacuum pumps, the temperature of the crucible was held at 1550' C. for outgassing of the sample until a pressure of
2 X mrn was reached. This outgassing operation normally consumed 1 minute from the time of sample drop. The fusion apparatus was isolated from the vacuum system, full furnace power applied and Toepler pumping started to transfer the gaseous reaction products into a small sample container made from a stopcock (Figure 1 insert). In 5 minutes the Pt flux reached 2200' C. and mas held at that temperature for an additional .iminutes. The Toepler pump wab then isolated, the gaseous sample container removed, and the fusion apparatua opened to the vacuum pumps for outgassing until a pressure of 2 X 10-1 was reached .It this pressure, the furnace pol? er was turned off, When the fluu temperature had fallen to 1530' C , the neut