q
handling steps combined with less required floor space. The over-all yield of pure TaClj extracted from the ore is 60 to 70% for the batch distillation approach.
5 1 W
J
-
I-!
1000
L
0 0
40
20
60
80
100
I20
P E R C E N T DlSTl LLED Figure
9.
Conclusions
A mixture of metal chlorides obtained by chlorination of tantalite ore can be distilled to give a pure TaC16 fraction. The molten chlorides behave like other liquids during distillation; the only unusual requirement necessary for distillation is that the entire still be heated to prevent solidification of the halides of moderately high melting point. Theoretical curves based on the Smoker and Rayleigh equations for batch distillation give good approximations for actual experimental values. Actual distillation results were better than predicted ; however, the calculated values were based on several assumptions which probably were not entirely accurate. If FeCI3 decomposition is not a problem, our calculations indicate the feasibility of performing a continuous distillation. However, no experimental work was done to confirm this. literature Cited
Purity of TaClj distillate
to total reflux. Chlorides with boiling points higher than that of TaC16, such as niobium and iron, tended to remain in the charge flask until most of the TaC15 had been removed. Niobium began to appear in the distillate only after a large portion of the initial charge had been distilled. T h e emission spectrographic analysis showed less than 10 p.p.m. of metallic impurities in some samples of TaC16. T h e pure T a C l j fraction was readily reduced to metallic tantalum by heating TaClj vapors in a stream of Hz a t 1200’ C. No further metallic contamination was encountered during this reduction step, as witnessed by