Concentration of Hafnium. Preparation of Hafnium-Free Zirconia

---

INDUSTRIAL AND ENGINEERING CHEMISTRY

512

of silt-bearing waters and of those of varying organic content for separate treatment. Since the hydrochloric acid used to dissolve suspended iron is not removed by evaporation, no fume hood is required. Hydrochloric acid is used rather than nitric acid, since the latter decomposes on storage unless protected from light. The deep color obtained by adding a large amount of thiocyanate permits the use of a comparatively small sample. When relatively large amounts of thiocyanate and acid are used, the hue and intensity of color produced with the iron are not affected by moderate variations in concentration. The 2-ounce French square bottles used for the test are inexpensive, may be stored compactly, and are subject to relatively slight loss from breakage. The attention required when individual samples are boiled with acid or evaporated and redissolved in acid is avoided by heating in a water bath under conditions that involve no change in volume or acidity. Interference by copper is approximately the same for both methods, but interference due to sodium hexametaphosphate or sodium pyrophosphate is negligible. Compared with a number of other color reac-

Vol. 15, No. 8

tions for iron, the one described is simpler, since it is applicable in strong acid solution. Summary The simple coloriinetric thiocyanate method described for the determination of iron in water is applicable in the presence of amounts of organic matter, sodium hexametaphosphate, or sodium pyrophosphate greater than are likely to be encountered. The test is not applicable in the presence of copper; in such instances it constitutes a qualitative test for this material. The color standards are stable for a t least 6 months. Literature Cited (1) Am. Public Health Assoc. and Am. Water Works Assoc., “Standard Methods for Examination of Water and Sewage”, 8th ed., New York, 1936. (2) Swank, H. W., with Mellon, M. G., 1x11.ENQ.CHEM., ANAL.ED., 10, 7-9 (1938). (3) Woods, J. T., with Zvfellon, M. G., Ibid., 13,551-4 (1941). PREBENTED before the Division of Water, Sewage, and Sanitation Chemiatry at the 105th Meeting of the AMERICAN CHEMICAL SOCIETY,Detroit, Mich.

Concentration of Hafnium Preparation of Hafnium-Free Zirconia EDWIN M. LARSEN’, W. CONARD FERNELIUS2, AND LAURENCE L. QUILL3 Ohio State University, Columbus, Ohio

T

HE development of the chemistry of hafnium since its discovery in 1923 has not been particularly extensive.

The lack of a convenient method for the separation of hafnium from zirconium has undoubtedly been the principal deterring factor. Previous methods have involved fractional crystallization of the hexafluorides ( 2 ) or oxychlorides (4) and fractional decomposition of complex ions of zirconium-hafnium formed with sulfuric acid, phosphoric acid, hydrofluoric acid, dicarboxylic acids, alpha-hydroxycarboxylic acids, and polyhydroxy alcohols (1) to yield the oxides, phosphates, or ferrocyanides (10) as precipitates. These methods require a large number of fractional separations, entail the handling of highly corrosive solutions, or utilize expensive reagents, all of which make large-scale operations impractical. The direct precipitation of the phosphate in dilute acid solution has never been carried through for the fractionation of zirconium-hafnium compounds, although it is known that hafnium phosphate is less soluble than zirconium phosphate (5, 5, 8). The utilization of the phosphates is handicapped by the gelatinous character of the usual phosphate precipitate, which results in slow filtration and inefficient washing, and by the lack of a convenient method for converting the insoluble phosphate into soluble compounds for reprecipitation. Both these difficulties have been overcome and a more satisfactory method for the concentrabion of hafnium oxide has been developed. 1 2

3

Present address, University of Wisconsin, Madison, Wis. Present address, Purdue University, Lsfayette, Ind. Present address, University of Kentucky, Lexington, K y .

Source of Material The ore used in this investigation was cyrtolite, an altered zircon which conforms approximately to the formula RsYI(Zr,Hf)(SiO& It was obtained from deposits near Bedford, N. Y and Hybla, Ontario. Orgnarily silicates are not easily attacked by acid treatment, and must be handled by some fusion method. Urbain (9), however, reported successful extraction by sulfuric acid at 65’ C. of malacon, an altered zircon. Recently Schumb and Pittman (10) noted that cyrtolite yielded to sulfuric acid treatment, but gave no details of the procedure. Van Osdall (11) made preliminary investigations on the acid “cracking” of cyrtolite, which resulted in satisfactory yields.

To be susceptible to acid attack, the ore must be in a finely divided state. When material of particle size larger than that separated by 100- or 200-mesh screens is treated with concentrated sulfuric acid, the yields are not satisfactory. After an exhaustive series of quantitative experiments to determine the optimum conditions for the acid cracking of the ore, the following procedure was found to give nearly 100 per cent extraction of the zirconium-hafnium content in a minimum of time. One hundred grams of 200-mesh ore and 200 grams of concentrated sulfuric acid are mixed in an evaporating dish and placed on a hot sand bath. After about 20 minutes the temperature reaches a maximum of 210” to 220” C. and the mixture begins to thicken. The heating process continues for about 10 minutes, or until a stiff mud results. A t this point heating is discontinued, and the “cracked” ore is cooled and added with stirring to 500 ml. of rvater. The addition of 10 ml. of 10 per cent glue solution aids in coagulation and filtration of the insoluble material. The solution is filtered and the residue on the funnel washed with water. The filtrates are combined. Treatment of the residue viith sulfuric acid under the same conditions does not result in any further loss of weight.

ANALYTICAL EDITION

August 15, 1943

TAB LE

I. DETERitIIS.\TIOS

HAFSIUMO X I D E

OF

HfOn b y SelenitPs

HfOn by Density

70

5%

16.8

17.1 23.6 51.0 93.3 9i.7

23.2 50.4 87.8 89.0

Method of Analysis Because of the large number of zirconium-hafnium saniples to be analyzed, a rapid method of analysis rvas desired. Of the methods available a t the time this work was started. the density determination .of the ignited osides seemed most satisfactory. Difficulties mith adsorbed gases and the preparation of the samples weye oyercome by the following procedure :

513

for the density of hafnium oxide would also result from this method of preparation. Thus the value of 9.67 as previously reported by von Hevesy ( 7 ) \vas selected and used in a11 calculations, with 5.62 for zirconium oxide. In order to have an independent check upon the per cent of hafnium oxide present in any sample, the method of analysis described by Schumb and Pittman (10) was used. This invoh-es precipitation of the basic zirconium-hafnium selenites, digestion to the normal selenites, (Zr, Hf) (SeOJ2. and ignition t o the oxides of a washed and dried sample of the mixed selenites. In all cases the values obtained by the selenite method (Table I) are higher, but fall within the accuracy of the methods except in the hafnium oxide-rich region. (The relation of time and temperature of ignition to density, crystal structure, and composition of various zirconium oxide-hafnium oxide mixtures is now under investigat ion .)

Sufficiently purified zirconium-hafnium solution is taken to provide 1. to 1.5 grams of the ignited oxide. (The sample is purified by the solution of the hydrated oxides in concentrated hpdrochlorir acid and subsequent crystallization of the oxychlorides. The iron is removed by washing the oxychlorides with ether.) The hvdrated oxide is precipitated from solution with ammonium hydroxide, washed free of sulfate ion. and ignited over a Fisher burner (ca. 900" C.) t o constant \%-eight. The oxide is ignited onre more, transferred to a weighed pycnometer, and the weight of the The oxide is rovered with about 1 ml. of led water and the pycnometer is then placed ator where it remains until gas bubbles no longer rise from the surface of the oxide. It is assumed that the oxide has been degassed when bubbles cease to be eliminated. The pycnometer is then filled with water and the thermometer inserted. The apparatus is placed in a constant-temperature bath a t 25.00 * 0.01"