N e w Fire Assay Methods for the Analysis of Iridosmines J. G. SEN GUPTA and F. E. BEAMISH Department of Chemistry, University of Toronto, Toronto, Ontario, Canada Tasmanian iridosmines in artificial mixtures of iron, copper, and nickel oxides were completely attacked at elevated temperatures by fire assay with sodium carbonate and graphite. The iron-copper-nickel alloy obtained from the fire treatment retained quantitatively all six of the platinum metals. A new analytical scheme for the determination of each platinum metal in the base metal alloy is presented. The results agree with those obtained by a direct chlorination of the iridosmines.
he term “iridosmine” or “osmiridT i u m , , has been used to indicate a variety of minerals. Kestland and Beaniish (14)suggested that much of the early literature dealing with the compositions of these minerals is unacceptable, as no specific analytical procedure accompanied the data and the contemporary methods were inapplicable for the separation and accurate determination of the rarer platinum metals. Further analytical complications arose from the uncritical use of the term iridosmine to indicate all or a part of the final “insoluble” resulting from fire a w y treatments, etc. Generally no eiidence is presented to prove the exktenre of iridosmines in the reported insoluhles. However. it is not unlikely that the rather severe fire treatment to which platinum metals are subjected in a fire assay may encourage the formation of insoluble metal mixtures containing the constituents found in natural iridosmines. It is also not improbable that some of the platinum metals deposits contain proportions of natural iridosmines. In general, classiral fire assays do not recognize the final ineolubles as a potential source of each of the six platinum metals. From a practical point of view the proportions of “insolubles” are sometimes very small and may be ignored as a significant source of any of the six metals. I t was desirable to provide a method of fire assay for platinum metals which would reveal the complete content of each of the six metals, irrespective of the identity of the mineral source and thus to avoid any question of the composition or value of the “insoluble.” T o accomplish this purpose it seemed rcwonnlile t o test the : ~ e a ymethod by
applying, i t to the most resistant of known iridosmines-e.g., the Tasmanian variety. These were formerly used for pen nibs and their high resistance to corrosion was demonstrated by Hill and Beamish (6). These authors succeeded in effecting complete dissolution through the use of chlorination which, for at least one iridosmine, proved superior to other recorded methods. A series of reports has shown that fusions of platinum metals-bearing materials containing iron, copper, and nickel with sodium carbonate and graphite will produce base metal alloys which will collect each of the six platinum metals quantitatively. These fusions with iridosmines resulted in the formation of similar alloys, and dissolution of these alloys or buttons in concentrated hydrochloric acid, followed by filtration and washing of the residue and subsequent examination of the silica residue on the filter paper under a microscope, showed complete absence of any unattacked iridosmine. Osmium and ruthenium were recovered by treating t h e button with perchloric acid in a distillation apparatus (8). A new method mas developed for the recovery of platinum, palladium, rhodium, and iridium from the perchloric acid pot liquid. The perchloric acid pot liquid was transferred t o a beaker, mixed with some concentrated nitric acid, and evaporated to dryness. After being converted to chlorides by repeated evaporations with hydrochloric acid, the base metals were separated by Dowex 50-X8 cation exchange resin a t p H 1.3 + 0.1. The four platinum metals mere recovered from the effluents and determined by a combination of solvent extraction (17) and copper powder reduction techniques (13). The method was developed by adding known amounts of the six platinum metals to roasted artificial concentrates of the base metals as reported by Plummer and Beamish (IO) and recovering osmium and ruthenium from the synthetic button by distilling with perchloric acid and subsequently recovering the four remaining precious metals from the perchlorates pot liquid. The details of the method and analytical results, which showed excellent recoveries, are given below. The Tasmanian iridosmines were also
subjected to a chlorination a t i O O o C. in the presence of sodium chloride and the platinum metals were then determined as described below. The values were compared with those obtained from the synthetic button salted with these iridosmines and treated by the above-mentioned perchloric acid method. They were in good agreement with each other. 9 qualitative analysis of the two Tasmanian iridosmines bv a spectrographic method showed the presence of the elements given in Table I. Their relative abundance is also indicated. EXPERIMENTAL
Apparatus, reagents, and standard solutions were the same as reported in a previous paper (16). Procedures. METHOD 1. A synthetic mixture of base metals approximating the composition of Sudbury nickel ore was roasted and mixed with fluxes containing 11 grams of graphite, by the method of Sant and Beamish (11). A part of this mixture was placed in a 30-gram assay crucible and an accurately xeighed quantity of Tasmanian iridosmine was added. The mivture was covered by the remaining part of the above charge and placed in a gas-air furnace previously heated to 1260’ C. for 45 minutes. The temTable I. Qualitative Spectrographic Analysis of Two Tasmanian Iridosmines
Element present
Cr
Fe
Relative abundance in iridosmine sample Sample 2 Sample 1
S S
S
R
VS very strong; S strong; M moderate; W weak; ? doubtful.
VOL 34, NO. 13, DECEMBER 1962
1761
perature was slowly raised to 1480" C. and maintained for 5 minutes, after which the crucible was removed. The buttons were cooled and separated by breaking the crucible, and the adhering slag was removed by gentle tapping. The buttons weighed betm-een 22 and 25 grams. The button was transferred to a distillation flask and the octavalent oxides of osmium and ruthenium were distilled from perchloric acid by the method of Kavanagh and Beamish (8), except that ruthenium was collected in 1 to 2 hydrochloric acid containing about 3% hydrogen peroxide. Osmium was determined gravimetrically from hydrobromic acid solution by the thionalide method (7) and ruthenium spectrophotometrically by the anthranilic acid method (12). The perchloric acid pot liquid was transferred to a 600-ml. borosilicate glass beaker by washing the flask and the connecting tube with mater. Ten milliliters of concentrated nitric acid were added and the beaker was covered with a raised cover glass. The solution was evaporated to dryness by a low flame in a specially constructed all-steel fume hood with a good suction and equipped with safety glass. When the perchloric acid fumes were expelled, the mass was allowed to cool, treated with 30 to 50 ml. of concentrated hydrochloric acid, and evaporated on the steam bath. The beaker was heated on a hot plate to expel hydrochloric acid and any residual perchloric acid. The treatment with concentrated hydrochloric acid and evaporation on the steam bath and hot plate were repeated three times. The mass was then treated with 30 to 50 ml. of concentrated hydrochloric acid, heated on the steam bath until all salts dissolved and the solution became clear, then evaporated to a moist residue, and diluted to 150 ml. with water. Any undissolved solid separating out was brought into solution by heating with an additional small amount of concentrated hydrochloric acid. From a total volume of 150 ml., a 10-ml. portion was extracted, diluted to 100 ml., and tested to ensure a p H of 1.3 zt 0.1. This solution was returned to the main solution, which was filtered through a 9-cm. Whatman No. 42 filter paper and washed well with water. The residue (A) was set aside and treated as described below. The filtrate (A) was diluted to 1500 ml. with water to attain a p H 1.3 =t 0.1 and passed through a Dowex 50-X8, 20to 50-mesh cation exchange resin, 70 cm. deep and 4 cm. in diameter. The effluent was collected in a 4-liter beaker. The column was then washed with 2.5 liters of water adjusted to pH 1.5 with hydrochloric acid. The effluent was evaporated on a hot plate to about 10 ml. and transferred by thorough washing with mater t o a 600-ml. beaker in the case of milligram amounts and to a 150-ml. beaker in the case of microgram amounts. The resin bed retained traces of platinum metals when filtrate A was passed through it. To recover this the column was eluted with 4 liters of 3N
-
1762
ANALYTICAL CHEMISTRY
hydrochloric acid and the washing was collected in a 4-liter beaker, evaporated to a small volume (10 to 20 ml.), and transferred to a 600-ml. beaker. It was evaporated to a moist residue and dissolved in 150 ml. of water by warming and addition of a little hydrochloric acid, if necessary. After testing the pH, which should be 1.3 ==! 0.1 when diluted to 1500 ml., the base metals mere again separated by passing through a neutral Dowex 50-X8 ion exchange column as used before and then washed with 2.5 liters of hydrochloric-acidulated water of p H 1.5. The second effluent, collected in a 6-liter beaker, was evaporated to 10 ml. and transferred to the beaker containing the first solution. To this combined solution 10 or 20 ml. of 2% sodium chloride solution were added, depending on whether milligram or microgram amounts of the platinum metals were present, and the solution was evaporated to dryness on a steam bath. The organic matter from the resin was destroyed by treating two to three times with 2 ml. of concentrated nitric acid and a few drops of 30% hydrogen peroxide. The solution was then treated with concentrated hydrochloric acid, evaporated to dryness on the steam bath, and again heated on a lowheat hot plate to remove perchloric acid. The residue was allowed to cool and after being treated with 5 to 10 ml. of concentrated hydrochloric acid, evaporated on the steam bath, and finally heated on the hot plate. The treatment with concentrated hydrochloric acid and evaporation on the steam bath and hot plate were repeated three times. The residue was treated with 1 ml. of concentrated hydrochloric acid, warmed, diluted to 20 ml. with water, filtered through a 7-cm. Whatman No. 42 filter paper, and washed thoroughly with water. The filtrate (B) was reserved and the residue (B) was combined with residue A. Treatment of Residues A and B. The filter papers containing residues A and B were placed in a porcelain crucible and heated slowly in an oven a t 350" C. until all carbon was burned. The temperature was then raised to 600" C. and the residue heated for 1 hour. It was then cooled, inserted in a Vycor tube, hydrogenated for 2 hours at 6OO0C., cooled in nitrogen, and removed. The crucible contents were covered with finely powdered sodium chloride and the mixture was chlorinated in a Vycor tube a t 700" C. for 8 hours. After being cooled in a current of chlorine, the contents of the crucible and the tube were mashed into a 100-ml. beaker with 0.1N hydrochloric acid and filtered through a 3.7-cm. glass fiber filter, and the filtrate was collected in filtrate B. The residue, consistently found free of platinum metal, was discarded. The combined filtrates were evaporated to dryness on a steam bath, moistened with 2 to 3 drops of concentrated hydrochloric acid, and diluted with water, and the p H was adjusted to 1.5. This solution was then passed through a small cation exchange resin (Dowex 50-X8), 20 cm. indepth
and 2 cm. in diameter, to separate traces of base metals and washed with 200 ml. of acidulated water of pH 1.5. Ten milliliters (for milligram amounts) or 2 ml. (for microgram amounts) of 2% sodium chloride were added to the effluent and the solution was evaporated to dryness on the steam bath to produce residue C. Separation and Determination of Micro Amounts of Platinum, Palladium, Rhodium, and Iridium. Microgram amounts of platinum, palladium, rhodium, and iridium were separated from each other by a combination of solvent extraction (17) and copper powder reduction (13) methods. Platinum and palladium were first separated from rhodium and iridium by solvent extraction of diethyldithiocarbamate complexes (17 ) . From the data given by Yoe and Kirkland (17) it is evident that platinum and palladium can be quantitatively separated from microgram as well as milligram amounts of rhodium and iridium by chloroform extraction of the diethyldithiocarbamate complexes of platinum and palladium. However, these authors made no attempt to recover rhodium and iridium from the aqueous phase. The method was examined by applying it to a standard solution containing four platinum metals. Subsequent to the extraction rhodium and iridium were recovered by evaporating the aqueous phase, oxidizing the organic matter by fuming nitric acid and hydrogen peroxide, converting the metals to chlorides with hydrochloric acid, and separating and determining them by the method re. ported by Tertipis and Beamish ( I S ) . The details of the procedure for the determination of the four platinum metals in the evaporated residue (C) are given below. Evaporated residue C was transferred to a 50-ml. separatory funnel by washing first with 5 ml. of concentrated hydrochloric acid and then with 10 ml. of water. After reduction of platinum by potassium iodide, both platinum and palladium were separated from rhodium and iridium by extraction of diethyldithiocarbamate complexes with chloroform (17). The chloroform extract of platinum and palladium was evaporated to dryness on the steam bath in the presence of 2 ml. of 2% sodium chloride and the organic matter was destroyed by repeated addition of fuming nitric acid and hydrogen peroxide. The mouth of the beaker was kept tightly closed with a watch glass, while the oxidation was carried out. The watch glass was then raised with hooks and the solution was evaporated to dryness and converted to chloride by evaporation three times with concentrated hydrochloric acid. The residue was dissolved in 5 ml. of acidulated water of p H 2 and transferred to a separatory funnel by washing with another 5 ml.
of acid water of p H 2. Palladium was separated from platinum b y Yoe and Kirkland's method (17) and after destruction of the organic matter the platinum and palladium were determined spectrophotometrically by p-nitrosodimethylaniline ( 1'7). Determination of Microgram Amounts of Rhodium and Iridium. After separation of platinum and palladium by solvent extraction of diethyldithiocarbamate compleues, t h e colution from t h e separating funnel n-as transferred to a 150-ml. beaker, 2 ml. of 2y0 sodium chloride solution were added, and t h e solution was evaporated t o dryness o n t h e steam bath. T h e organic matter was destroyed by the addition of fuming nitric acid and hydrogen peroxide (three times) and heating on a steam bath: the beaker mas kcpt tightly covered with a n-atch glass, which was raised b y glass hooks onlv during evaporation. 'I'he inass waq then converted t o chlorides hv evaporation with hvdrochloric acid (three times) and then dissolved in 20 m! of 1S hydrochloric acid. The separation of rhodium from iridium bv precipitation with copper powder and its subsequent determination by the stannous chloride method were carried out hv the method of Tertipis and Beamiqh ( I S ) . The filtrate containing iridium and some dissolved copper was evaporated to dryness, iridium was osidized bv adding 1 t o 2 ml. of concentrated nitric acid, and the salts were then convcrted t o chloride by repeated waporation with Concentrated hydrochloric acid. Failure t o oxidize iridium before passing through the ion eschange column for the separation of ropper results in the loss of iridium on the resin bed. The evaporated mass was treated with 2 drops of concentrated hydrochloric acid, diluted with water, and after the p H had been adjusted to 1.5 it was passed through the ion eschange column and washed with 200 ml. of hydrochloric-acidulated water of p H 1.5. The effluent was evaporated to dryness in the presence of 2 ml. of 2% sodium chloride, and the organic matter n.as destroyed by nitric acid and hydrogen peroxide. The metals were converted to chlorides and the solution was transferred t o a 30-ml. heaker by washing with 5 m!. of concentrated hydrobromic acid. The
Table II.
Iridosmine Sample taken, mg.
a
Osmium Mg.
beaker was tightly covered with a small watch glass and the solution was boiled to 1 ml. The solution was transferred t o a 25-ml. flask by washing with 5 ml. of concentrated hydrobromic acid and iridium was determined spectrophotometrically b y a modification (9) of the Berman-McBryde (3) method. The value was compared with a known iridium value obtained by taking a n aliquot from freshly diluted stock solution along with 0.12N hydrochloric acid, boiling t o 1 ml. with 5 ml. of hydrobromic acid and 2 ml. of 2% sodium chloride, and developing the color simultaneously with the unknown.
diluted, filtered, and washed. The filtrate was diluted to 200 ml. and the platinum was precipitated and determined by the thiophenol method (4).
DETERMINATION OF
PALLADIUM.
After the hydrated oxides of palladium, rhodium, and iridium had been dissolved in hydrochloric acid, the palladium was precipitated and determined by dimethylglyoxime. Recovery of Platinum and Palladium from Rhodium-Iridium Fraction and Determination of Rhodium and Iridium. The filtrate from t h e separation of palladium was evaporated t o dryness on the steam b a t h and after being tightly covered with a watch glass, which was raised only during evaporation, t h e organic matter was destroyed by adding fuming nitric acid and hydrogen peroxide (three times), after which the nitrates were converted to chlorides by thrce evaporations on the steam bath with hydrochloric acid. The evaporated residue was treated with 0.5 ml. of hydrochloric acid and transferred to a 50-ml. flask and the volume was made up n i t h mater. A 10-ml. aliquot was transferred t o a separatory funnel, and palladium and platinum were separated from rhodium and iridium and then from each other by the method of Yoe and Kirkland (17). Each metal was then determined spectrophotometrically by the p-nitrosodimethylaniline method (17 ) . The solutions containing rhodium and iridium were treated exactly as when microgram amounts were determined, escept that they were finally determined gravimetrically by 2-thiobarbituric acid (4) and 2-mercaptobenzothiazole ( I ) , respectively. The results are included in Table 11. METHOD2. The Tasmanian iridosmine mas weighed in a very small Vycor glass boat, covered with finely ground sodium chloride, inserted into a specially designed Vycor glass tube, and chlorinated by the method of Westland and Beamish (14, 16). The exit gases were passed through sulfurous acid first and then into three receivers containing 6.Ir hydrochloric acid kept saturated with sulfur dioxide. The hydrochloric acid solutions were evaporated t o a few milliliters on the steam bath and then combined with the main solution obtained by mashing the chlorination
Separation of Milligram Amounts of Platinum, Palladium, Rhodium, and Iridium. Gilchrist and JJ7ichers' hydrolytic yrccipitation (5) and the dimethylglyovime method were rmployed for separating platinum and palladium. Because of fuming with perchloric acid in the previous treatment of these metals, some platinum was coprecipitated with palladium, rhodium, and iridium and could not be recovered even by double precipitation. A similar difficulty was encountered by Beamish and Scott ( 2 ) with platinum solutions which were fumed rvith sulfuric acid. Moreover, traces of palladium escaped precipitation by dimethylglyoxime and accompanied rhodium and iridium. The present authors were able t o rwover the lost platinum and palladium by solvent extraction of diethyldithiocarbamate complexes (17) of these two metals from an aliquot of the rhodium-iridium fraction.
To the mixture of milligram amounts of platinum, palladium, rhodium, and iridium as obtained above, 2 ml. of concentrated hydrochloric acid mere added and after dilution t o 300 ml. palladium, rhodium, and iridium mere twice precipitated hydrolytically by the Gilchrist and Wichers method (6). DETERMINATION OF PLATINUM. Excess bromate was destroyed in the combined filtrates of platinum b y repeated evaporation on the steam bath with concentrated hydrochloric acid. The evaporated residue was acidified with 4 drops of concentrated hydrochloric acid and the solution was
Recovery of Platinum Metals from Tasmanian Iridosmines
RutheniumPg.
c7 IC
Palladium !4.
Platinum
Rhodium
Iridium
70
!-l%
70
!Jg.
5%
LIg.
70
I
1 1 2 2
5.01 3.86 3.51 2.15
2.355 1.775 1.264 0.765
47.0 46.0 36.0 35.6
200 158 49.8 29
3.99 4.10 1.42 1.35
Method 11.7 9.3 42.1 28
0.23 0.24 1.20 1.30
205.4 150.5 330 200
4.1 3.9 9.4 9.3
30 27 61.4 38.7
0.60 0.70 1.75 1.80
1.683 1.293 1.45 0.886
33.6 33.2 41.3 41.2
1 1 2 2
3.865 3.016" 3.21 4.052"
1.76 1.40 1.15 1.48
45.5 46.5 35.8 36.5
150.7 120.64 44.3 56.7
3.90 4.00 1.38 1.40
Method I1 9.1 0.23 6.9 0.23 40.1 1.25 49.4 1.22
154.6 126.7 29.5 378.9
4.0 4.2 9.2 9.3
25 21.7 54.6 70.5
0.65 0.72 1.70 1.74
1.279 1.004 1.323 1.673
33.1 33.3 41.2 41.3
Analyzed by perchloric acid method.
VOL. 34, NO. 13, DECEMBER 1962
1763
product from the tube with 0.12N hydrochloric acid. The combined solution was transferred to a distillation unit (15)and sulfur dioxide was removed by boiling under reduced pressure, using a good water suction. Two methods were used for separating the six platinum metals from this solution. Osmium and ruthenium were distilled from one sample for 2 hours by 200 ml. of perchloric acid and collected in 3% cold hydrogen peroxide. The trap was boiled for 30 minutes with 15 ml. of perchloric acid. Osmium was separated from ruthenium by the method of Westland and Beamish (15) and determined gravimetrically by thionalide ( 7 ) . Ruthenium was distilled by oxidation in 10 ml. of concentrated sulfuric acid and 20 ml. of 10% sodium bromate and collected and determined spectrophotometrically by a method previously reported (1%'). The method of analyzing the perchloric acid pot liquid is described above. By the second process the chlorination product of Tasmanian iridosmine as obtained above was treated by the method of Westland and Beamish (16) for the separation of osmium, rutheniurn, and the base metals from platinum, palladium, rhodium, and iridium. However, osmium was collected in hydrobromic acid ( l a ) and determined gravimetrically by thionalide ( 7 ) and ruthenium was collected in a 1 t o 2 hydrochloric acid solution containing 3%
hydrogen peroxide and determined spectrophotometrically by the anthranilic acid method (la). The solution containing platinum, palladium, rhodium, and iridium, free from any base metals, was evaporated t o dryness in the presence of 2 ml. of 2% sodium chloride. Organic matter from the ion exchange column was destroyed by nitric acid and hydrogen peroxide. The small amount of sulfuric acid remaining was removed by first fuming and then evaporating to dryness several times on a hot plate with concentrated hydrochloric acid. The chloride residue was then transferred to a 50-ml. separatory funnel by washing with 5 ml. of concentrated hydrochloric acid and 10 ml. of water and the separations and determinations of platinum, palladium, rhodium, and iridium were carried out by solvent extraction and copper powder reduction techniques as described above. The results are included in Table 11. ACKNOWLEDGMENT
The authors are grateful to the National Research Council for financial support and t o C. L. Lewis, Falconbridge Kickel Mines, Richvale, Ontario, for carrying out the qualitative spectrographic analysis of the two iridosmine samples.
LITERATURE CITED
(1) Barefoot, R. R., McDonnell, W. J., Beamish, F. E., ANAL.CHEM.23, 514 (lW51). (2) Beamish, F. E., Scott, M., IXD.ENC. CHEM., ,4NAL. ED. 9, 460 (1937). (3) Berman. S. S., McBrvde,' JT7. $. E.. . Anaiyst si, 566 (i956). " (4) Currah, J. E., McBryde, W. A. E., Cruickshank, 8. J., Beamish, F. E., ISD. ESG. CHEM., ANAL.ED. 18, 120 (1946). (5) Gilchrist, R., Wichers, E., J . ,477~. Chem. SOC.57,2565 (1935). (61 ~, Hill. M. A.. Beamish. F. E.. ASAL. CHEM.22,590'( 1950). ' \ - - - - I
(i) Hoffman, I., Schweitzer, J. E., Ryan, D. E., Beamish, F. E., Ibid., 25, 1091 (1953). (8) Iiavanagh, J. M., Beamish, F. E., Ibid., 32, 490 (1960). (9) Marks, A. G., Beamish, F. E., Ibid., 30,1464 (1958). ( I O ) Plummer, M. E. V., Beamish, F. E., Ibid., 31, 1141 (1959). (111 Sant. B. R.. Beamish. F. E.. Zbid.. ' 33,30411961).' (12) Sen Gupta, J. G., Beamish, F. E., communicated t o Am. Mineralvgzst for publication. (13) Tertipis, G. G., Beamish, F. E., ANAL. CHEM. 32,486 (1960). (141 Westland. A. D.. Beamish. F. E.. ' Am. Minerahgist 43,' 503 (1958). (15) Westland, A. D., Beamish, F. E., ASAL.CHEW26,739 (1984). (16) Ibid., 30, 414 (1958). (17) Yoe, J. H., Kirkland, J. J., Ibid., 26,1335, 1340 (1954). RECEIVEDfor review August 15, 1962. Bccepted October 4, 1962.
A Complete Separation of a Mixture of Zi rco nium(IV), Copper(I I), Molybdenum(VI), Titu nium(IV), Vu nud ium(IV), and Magnesium(l1) by Ion Exchange Chromatography CARL MICHAELIS, SYLVESTER EVESLAGE, PAUL COULTER,' and JOHN FORTMAN2 Chemisfry Department, University of Dayton, Dayton, Ohio
b A detailed method of separation of a mixture of Zr(lV), Cu(ll), Mo(VI), Ti(lV), V(IV), and Mg(ll) on Dowex cation and anion exchange resins using hydrochloric acid elution is reported. The titanium, vanadium, and magnesium were not retained by the anion exchange resin, but the zirconium, copper, and molybdenum which form the anionic chloro complexes were. After the titanium, vanadium, and magnesium were washed from the anion exchange column with 9M hydrochloric acid, this mixture was made 11M in acid and passed through the same column. The titanium was now retained while the other two passed through. The vanadium and magnesium were then separated on the cation exchange column. The zirconium, copper, and molybdenum were successively eluted from the anion exchange column by 1764
ANALYTICAL CHEMISTRY
selecting the proper concentration of eluting acid. Elution curves were prepared for zirconium, copper, molybdenum, magnesium, and vanadium.
Mu::
has been done recently ion exchange separation of metals (8), especially those which make up complex alloy steels and high temperature alloys which are useful for supersonic air craft, guided missiles, rockets, etc. These alloys are made up of titanium, molybdenum, zirconium, vanadium, copper, and magnesium among other metals. The problems of separation for analyses are well known. Frequently ion exchange chromatography affords a convenient method of separation. Many complexing agents such as HF ( I S ) , oxalates ( l a ) ,etc., have been used WORK
in separating these metals. The procedure described here uses only hydrochloric acid to separate all six metals consecutively. It has been shown that Zr(IV), Cu(II), and Mo(V1) in concentrated hydrochloric acid solutions are strongly adsorbed by anion exchange resins (4, 6, 9). Magnesium(i1) and V(IV) do not form stable anionic chloro complexes, and Kraus and coworkers (7) have sh0n.n that Ti(1V) forms a stable anionic chloro complex in hydrochloric acid concentrations of 1OM or greater. Therefore by passing a mixture of the 1 Present address, Chemistry Department, University of Kansas, Lawrence, Kan. 2 Present address, Chemistry Department, University of Notre Dame, South Bend, Ind.
