Foote and Langley — Colunibium and Tantalum. 393 



Art. XLIII. — On an Indirect Method for Determining 

 Columhinm and Tantalum / by H. W. Foote and R. W. 

 Langley. 



Introduction. 



The determination of tantalum and columbium in mixtures 

 of their oxides has always been one of the difficult operations 

 in analytical chemistry. The method for separating the two 

 metals was devised by Marignac* and consisted in separating 

 by crystallization the difficultly soluble potassium-tantalum 

 fluoride from the more soluble columbium salt. This method, 

 with various minor modifications, is still the one commonly 

 used. The operations involved are tedious and the results 

 are only approximate. Judging by our own experience, the 

 method requires some practice before even approximate results 

 can be obtained. The reason why the results are inaccurate 

 is first, because the tantalum salt obtained is not quantitatively 

 insoluble, so that tantalum is left in the filtrate with the colum- 

 bium, and second, because of the tendency of columbium to 

 crystallize to a limited extent with the tantalum double fluoride. 



Two volumetric methods have been proposed, both depend- 

 ing on a preliminary reduction of columbium to a lower oxide 

 and titration with potassium permanganate. By the first of 

 these methodsf columbium is reduced by zinc in strong hydro- 

 chloric acid solution, and titrated with potassium permanganate 

 which has been standardized by means of a pure columbium 

 salt. In the second method :}:, columbium in a solution contain- 

 ing succinic acid is reduced by amalgamated zinc. Under 

 fixed conditions, by this method, columbic oxide is reduced to 

 the empiri caloxide Cbj^Ogi, and can be titrated with perman- 

 ganate. 



The method of separation proposed by "Weiss and Landecker§ 

 will be discussed in the article immediately following this. 



The low density of columbic oxide (4*552) as compared with 

 the density of tantalic oxide (8"Y16) suggested that the compo- 

 sition of any mixture of the two could be deduced from its 

 density, if the density of mixtures of known composition was 

 first determined. The principle has been applied by Fenfield 

 and Ford II to the estimation of the proportions of tantalic and 

 columbic oxides in stibiotantalite, assuming that the density of 

 mixtures of the oxides is a linear function of the composition. 



* Archiv. des Sci. Phys. et Nat., 1866. 



f Osborne, this Journal [3], xxx, 339, 1885. 



X Metzger & Taylor, Zeitschr. anorg. Chem., Ixii, 383, 3909. 



§ Zeitschr. anorg. Chem., Ixiv, 65, 1909. 



II This Journal [4], xxii, 61, 1906. 



Am. Jour. Sci, — Fourth Series, Vol, XXX, No. 180. — December, 1910. 

 26 



