Volatilization of Elements from Perchloric and Hydrofluoric Acid

FRANCIS W. CHAPMAN, JR.\ GEORGE G. MARVIN2, AND S. YOUNG TYREE, JR.' University of North Carolina, Chapel Hill, N, C,, and Massachusetts Institute ...
1 downloads 0 Views 2MB Size
of Elements from Perchloric and FRANCIS W. CHAPMAN, JR.', GEORGE G. MARVIN', AND S. YOUNG TYREE, JR.' University of North Carolina, Chapel Hill, N . C., and Massachusetts Institute of Technology, Cambridge, Mass. Mixed perchloric and hydrofluoric acid solutions containing eonipoicnds of 37 elements were evaporated at 200" C. Analysis of the residues shou-ed that appreciable quantities of boron, silioon, gcrrnanium, arsenic, antimony, ohromium, selenium, manganese, and rhenium are lost during sueb treatment. The loss is due, in most eases, to the volatilization of the fluorides of the elcrnents. No loas was observed on similar treatment of compounds of sodium, potassiom, copper, silver, gold, beryllium, magnesium, calcium, strontium, barium, zinc, cadmium, mercury, lanthanum, cerium, titanium, thorium, tin, lead, "anadiitm, bismuth, molybdenum, tungsten, uranium, iron, cobalt, and niokcl.

T

HE preparation of a solution of a sample for analysis is not always handled so carefully as are the succeeding analyticsl operations. Lundell and Hoffman (5)have pointed out many solvents and fluxes for preparation of solut,ion, and discussed the difficulties oft,en encountered in this operation. Theistate bhat, depending on the solvent, many elements are volatiliaed during the preparation of solution. Mixed sulfuric and hvdroRuoric used ns a salvent,

difference from the known weight of sample talien. case, blanks were run without hydrofluoric xeid.

In every

PROCEDURE

weight a high purity compound of the element was dissolved in a solution (usually perchloric seid) which introduced no interfering constituents into the subsequent analysis. This solution was fiitered and diluted to 500 ml. and 25-ml. portions were taken for analysis. The portions were evaporated to 3 to 5 ml. in platinum dishes, and 10 to 15 ml. of 70% perchloric aoid and 8 to 10 ml. of 30% hydrofluoric acid were added. The solutions ere evaporated to atrong fumes of perchloric seid by surface evaporation (Figure 1). The overhead heating clement urns adjusted to obtain a temperature af 200" C. a t the surface of the liquid, without the air blast in operation. I n operation, this easily assembled apparatus gave rapid evaporation with no determinable losses due to spattering. Following t.he first fuming. 5 to 10 ml. more of 30% hydrofluoric acid !;:ere added, and the solution was again e v a p o r k d to strong fumes of perchloric acid. The residue was analyzed for t h e dement, nnder consideration. TJmnirc methods of analvsic ~~.~~~~~~~ ( 2 ) were used~inall cases It NBS necessary to use gravimetric sampling procedures for titanium, germanium, and tin. ~~

RESULTS

The above procedure as used with compounds 0 , u, rlrjllnllY_ Analyses of the resulting solutions gave low results in the case of nine elements (Table I j. The figures in Table I stating the amounts lost are approximate only, and represent the actual losses determined in the course of the standardized procedure. More prolonged fuming with the mixed acid solution did not affect the elements showing no loss, hut increased the losses in the other nine elements. For example, i t was possible to volatilize all the chromium from solutions by

re

elemems zesten (excepc manganese ana ieaoi xu m e r I I I ~ X I I I I U L I I valence states. Therefore the use of only one valence state of each element was justified. Lundell and Hoffman (6) give an excellent discussion of the effect of digestion with perchloric acid on the several elements. The volatility of some compound of aluminurn cannot be the cause of IONresults i n ' t h e mse of aluminum (not, included in Table I). According to Zerniake (le), aluminum fluoride sublimes at 1290" C. Marvin snd Woolaver (7) have shown that aluminum perchlorate decomposes to yield the nonvolatile oxide of aluminum a t 200' to 300" C. The procedure used in the

30

V O L U M E 21, NO. 6, J U N E 1 9 4 9 Table 1.

Effect of Treatment with Perchloric and Hydrofluoric Acids 1-0

Loss

Aodiom, potassium Copper. silver, gold Beryllium, magnesium Calcium strontium barium z i n c , oahrnium, meicury L a n t h a n u m , cerium Titaniurn, thorium T i n , lead Vanadium Risiiiuth .\lolybdenuin, tungsten, uranium I r o n , cobalt, nickel

Apparent Lo.. Boron. 100%

Selenium, varies greatly Manganese, u p t o 3% Rhenium, varies greatly

for aluminum was gravimetric, and the appareiit loss was due to the inability of fuming perchloric acid t o remove all fluoride associated with the aluminum as a complex ion. This complex prevents the precipitation of hydrous aluminum oxide in the usual manner (3). XOattempt is made to explain the loss of niaiigaiiese; but Hoffman and Lundell (4)observed that mang:inese conipounds ivere slightly volntile from mixed perchloric and hydrochloric acid solutions. Tlie eight other elements which shon-ed loes on treatment with the acid mixture were expected to do so. Boron and silicon fluorides are very volatile (8). Selenium forms at least two volatile fluorides (14). iintimony and arsenic form volatile fluorides ( l l ) ,as do chromium ( I O ) , germanium ( I ) , and rhenium (12). According to Ruff (II), molybdenum and tungsten form volatile fluoi,ides and oxyfluorides. T o such fluorides are formed, however, when salts of these elements are fumed with the acid mixture, despite the vigorous dehydrating action of concentrated perchloric acid. I n contrast chromium is easily volatilized as the osyfluoride from the acid mixture. This indicates that it is virtuslly impossible t o predict the volatility of many elements from mixed acid solutions. Selenium and rhenium are appreciably volatile from fuming perchloric acid solutions alone, as might be expected from the physical properties of the oxides of these acid-forming elements (9, 15). The volatilit,y of rhenium, under these conditions, has been suggested (6).

701 The results of this investigation show that nine clcments, if present in a material that is subjected to routine analyticnl treatment ri-ith perchloric and hydrofluoric acids, will t i c lost in varying amounts. Consequently, determination of any of thebe elements in the resulting solution would give low v:ilucs bnsed on the original material. -\CKSOI’LEDGhlENT

Tlie authors are iridehted to the CLinadian (’rili!)c’i’ Il(jfirir.rs Limited of >fontreal for generosity in supplying high-purity elementary selenium, to The Eagle-Pic1ir.r Compn~iyRcsearch Laboratories for generosity in supplying high-purit y gcmiariiuni dioxide, and to A , D. Jfelaven, University of Triincwv, for promptly supplying high-purity potwsium rhcn:itc (\-I1 ). LITERATURE CITED

Fischer, W., and Weidemann, W., 2.anorg. allgt,rri. L’hern., 213, 106-14 (1933). Hillebrand, W.F., and Lundell, G. E. F., “Apj)licd Inoi.g:tiiic .4nalysis,” New York, John Wiley R: Sons. 1929. Ibid., p. 390, line 5 ; p. 730, footnote 89. Hoffman, J. I., and Lundell, G . E. F., J . Research .Vutl. B u r . Standards, 22, 465-70 (1939). Lundell, G . E. F., and Hoffman, J. I., “Outlines of LMetliodj of Chemical Analysis,” pp. 24-9, New York, John Wiley & Sons, 1929. Ibid., pp. 46-7. Marvin, G . G.. and Woolaver, L. B., I s n . ENG.CHEM.,ASAL. ED.,17, 474 (1945). Noyes, A. A , and Bray, W. C., “System of Qualitative Analysis for the Rare Elements,” pp. 35-7, New York, Jlacmillan Co., 1927. Ogawa, E., Bull. Chem. SOC.J a p a n , 7, 265-73 (19323. Oliveri, diV., Gazz. chim. ital., 16, 218 (1886). Ruff, O., 2 . angcw. Chem., 20,1217 (1907). Ruff, O., and Kwasnik, W., Ibid., 47, 480 (1934). Smith, G. F., “Perchloric Acid,” pamphlet, 4th ed., p. 15, Columbus, Ohio, G. Frederick Smith Chemical Co., 1940. Yost, D. M., and Russell, H. J., “Systematic Inorganic Chemistry,” p. 299, New York, Prentice-Hall, 1944. Ibid., p. 318. Zernicke, J., Chem. Weekblad, 36, 748-50 (1939). RECEIVED August 7, 1948.

Determination of Alumina in Steel A Spectrochemical Method R. H. COLIS k Y D D. A. GARDNER Carnegie-Illinois Steel Corporation, Gary Steel Works, Gary, I n d .

T

H E extensive use of aluminum in the steel-making industry as a deoxidizer, for control of grain size, etc., has created H need for an accurate ant1 rapid means of determining aluminum iii steel. The aluminum may be present in steel in solid solution and as compounds such as alumina, *112Os, and aluminum nitride, Al2N2. Aluminum present in solid solution is soluble in hydrochloric acid. The aluminum compounds are insoluble. Khen alumina is determined in steel, the insoluble rpsidue is Fashed with a solution of sodium carbonate to ensure the renioval of any nitrides left undissolved by washing with dilute acid (1). In this paper the term “alumina” designates the combined aluminum that remains in the residue after the sodium rnrhonate wash. Gravimetric and volumetric methods of analysis for alunlina, because of the many separations and reprecipitations involved, have proved t,oo tedious and time-consuming when a relLttively

large number of samples are to be analyzed. Coloriiiictric arid photometric procedures also are too involved, because they require the complete removal of iron and other int,erfering elements before comparison can be made. S o particular problems were encountered in the spectrographic determination of the total aluminum which is carried out in this laboratory by means of pressed pellets made from drillings or millings of the steel using a high voltage condensed spark for excitation. Total aluminum results obtained in this manner are, for practical purposes, independent of the concentration of alumina. Results on several samples analyzed spectrogr~bphically for total aluminum containing alumina in concentrations as high as 0.06% were in good agreement with the chemical results obtained on the same samples. Because spectrographic analysis does not provide a direct meaiis for distinguishing between acid-insoluble and acid-