326 WOOD AND CLARK: THE: DETERMINATION OF TUNGSTEN IN [Vol. 83 The Determination of Tiingsten in Titanium, Zirconium and their Alloys BY D. F. WOOD AND R. T. CLARK (Research Defiartment, Imfieria,! Chemical Industries Ltd., Metals Division, Kynoch Works, Witton, Birmingham, 6 ) To fulfil the need for satismfactory methods for the determination of tungsten over the range 0.002 to about 1 per cent. in titanium and its alloys, two absorptiometric procedures have been developed, and subsequently extended to the examination of zirconium and its alloys. A direct procedure, based on the yellow colour produced by the reaction of reduced tungsten with thiocya.nate ions, is suitable for the determination of tungsten from 0.05 to 1.6 per cent. in titanium and many of its alloys, and from 0.01 to 1.6 per cent.in zirconium and its alloys; larger amounts can be determined by using a smaller sample weight. The procedure is simple and rapid, and is also suitable for application on a routine basis. Vanadium and molybdenum interfere, and, when these metals exceed specified limits, the toluene-3 : 4-dithiol pmcedure is recommended. A procedure, which depends on the formation of a bluish green complex with toluene-3 : 4-dithiol, is applicable to the determination of tungsten in both titanium and zirconium-bearing materials over the range 0.002 to 0.8 per cent. Vanadium does not interfere, and interference by molybdenum is overcome by means of a preliminary sulphide separation. TUNGSTEN may be introduced into titanium, zirconium and their alloys either from the parent ores or by contamination from the tungsten electrodes that are frequently used in the arc-melting of evaluation buttons.Its presence is undesirable because of its considerable hardening effect ; further, in zirconium-bearing materials used in nuclear reactors, only small amounts of tungsten (below about 0.005 per cent,) can be tolerated because of its high neutron capture cross-section. The objective was the determination of tungsten in these materials over the range 0.002 to about 1 per cent. Colorimetric procedures based on the use of thiocyanatels2*s and toluene-3 : 4-dithiol (dithiol),4 commonly used for the determination of tungsten in steel, have been applied to the determination of tungsten in titanium metal,516 and these procedures are at present used in this laboratory.’ In the thiocyanate pr~cedure,~ the sample is dissolved in hydrochloric acid and, in order to obtain reproducible blank values, titaniium is completely reduced by boiling the solution with stannous chloride before the addition of thiocyanate.In this procedure, the small absorption of the yellow complex formed between reduced tungsten and thiocyanate ions is measured in the presence of a considerable absorption due to the deep violet colour of titanous chloride, and this limits its application to the determination of tungsten in amounts above about 0.1 per cent. Further, during dissolution in hydrochloric acid, which usually takes several hours, part hydrolysis of titanium sometimes occurs, which results in inconsistent blank values.In order to extend the application of the thiocyanate procedure, modifications to the published method5 were examined. Thew included the use of fluoroboric acid as a sol- vent ; it forms a green-coloured complex fluorotitanate with titanium, and thereby inhibits hydrolysis and minimises blank values, thus allowing the range of the determination to be extended below 0.1 per cent. It was expected that boiling the solution with stannous chloride would be unnecessary because tungsten would be effectively reduced by titanous ions formed during dissolution of the sample. Application of this modified procedure was extended to titanium alloys and, subsequently, to zirconium and its alloys. The dithiol procedure depends on the formation of a bluish green complex of tungsten with dithiol and extraction of the complex: with isoamyl acetate.Although more lengthy than the thiocyanate procedure, it is suitable for the determination of tungsten down to about 0.001 per cent. Examination of titanium - vanadium alloys showed that vanadium This acid ,also expedites dissolution of the metal.June, 19581 TITANIUM, ZIRCONIUM AND THEIR ALLOYS 327 interferes in the dithiol procedures as applied to titanium metal, whereas, in the procedure as applied to steel! it is claimed that vanadium does not interfere. In order to provide a method for the determination of tungsten below about 0.05 per cent. in titanium alloys, including those containing vanadium, and also in zirconium and its alloys, the dithiol procedure, as applied to steel! was selected as a basis for further investigation.EXPERIMENTAL THIOCYANATE METHOD FOR TUNGSTEN IN TITANIUM PRELIMINARY EXPERIMENTS- Initial experiments were made to establish the wavelength at which maximum absorption of the tungsten - thiocyanate complex occurs and then, by using this wavelength, to prepare a calibration graph. The concentrations of hydrochloric acid and ammonium thiocyanate used in these preliminary experiments were those recommended by Bacon.5 Absorption curve-A 0.2-g sample of tungsten-free titanium was dissolved in a mixture of 20 ml of hydrochloric acid, sp.gr. 1.18, and 1 ml of fluoroboric acid; 20 ml of sodium tungstate solution (1 ml 3 0.1 mg of tungsten) were added and the solution was diluted to 100 ml in a calibrated flask. To a 25-ml aliquot in a 100-ml calibrated flask were added 20 ml (making 25 ml in all) of hydrochloric acid and then 10 ml of 15 per cent.ammonium thiocyanate solution. The solution was diluted to the mark and set aside for 5 minutes at 20" C. Optical-density measurements were then made in 2-cm cells with a Unicam SP600 spectrophotometer. Wavelengths between 3700 and 4500 A were used, and it was found that the maximum absorption is at 4 0 0 0 ~ . This procedure was repeated with a solution containing titanium only, and at 4 0 0 0 ~ the absorption, due mainly to fluorotitanate ions, was found to be at a minimum (optical density 0-06). Calibration graph-Solutions containing 0.2 g of tungsten-free titanium and amounts of sodium tungstate solution (1 ml = 0.1 mg of tungsten) ranging from 0 to 32.0 ml were prepared as before.These were diluted to 100 ml and 25-ml aliquots were taken. The reagents were added as before, the solutions were diluted to 100 ml and the optical densities were measured at 4000 A in 2-cm cells. The prepared calibration graph was linear and was suitable for the determination of tungsten in the range 0.05 to 1.6 per cent. The blank value (optical density 0.06) due to the green-coloured fluorotitanate ions was equivalent to only half of that obtained in the method described by Bacon5 The tendency for titanium to hydrolyse was completely prevented, and tungsten was effectively reduced by titanous ions produced during dissolution of the sample. In addition, the sample dissolved in 10 to 20 minutes, compared with about 2 hours when hydrochloric acid is used alone as recommended> EFFECT OF VARIOUS FACTORS- Because the optical density of the complex decreased over the range 18" to 35" C by 0.0015 per "C rise in temperature, solution temperatures were controlled at 20" 1" C while the optical-density measurements were made.Variation in blank values over the same range of temperatures was negligible. The colour was fully developed within 5 minutes of the addition of ammonium thiocyanate solution, and it was stable for about 30 minutes, after which the optical densities slowly decreased, the blank values remaining constant up to 90 minutes after addition of the thiocyanate. Different amounts of hydrochloric acid between 23 and 40 ml had no effect on the optical density.Below 23 ml, optical densities decreased sharply with decrease in acid concen- tration ; the blank value increased slightly with increase in acid concentration. Different amounts of fluoroboric acid between 0.5 and 2 ml had no effect on the optical density of the complex or on blank values. Over the range 8 to 20 ml, the amount of 15 per cent. ammonium thiocyanate solution had no significant effect on the optical density, but, below 8 ml, optical densities decreased sharply with decrease in thiocyanate concentration. There was a marked increase in blank values with increase in the amount of thiocyanate from 5 to 20ml. To ensure full colour development in conjunction with a consistent blank value, 10 ml of ammonium thiocyanate, controlled to within +Om5 ml, was adequate.328 When the procedure was applied to solutions containing titanium in the absence of tungsten, optical densities increased from 0.001 to 0.076 with increase in titanium content over the range 0 to 60 mg, as shown by the following results- WOOD AXD CLARK: THE DETERMINATION OF TUNGSTEX I N [Vol.83 Titanium present, mg . . . . Nil 10 20 30 40 60 ti0 Optical density . . . . . . 0.001 0.010 0.024 0.038 0.062 0.064 0,076 In blank determinations, therefore, it is essential to use a weight of tungsten-free titanium corresponding to the weight of titanium present in the sample. Tests on solutions containing 50 mg of titanium and the equivalent of 0.10 and 0.80 per cent. of tungsten, to determine the effect of common alloying constituents and likely impurities, established that no interference was caused by amounts up to at least 20 per cent.of tin or manganese, 10 per cent. of aluminium, 5 per cent. of iron or chromium and 1 per cent. of copper. Molybdenum above 0.25 per cent, caused high results, but interference by molybdenum up to 2 per cent. can be compensated by adding an equivalent amount of molybdenum to the blank solution. Vanadium introduced a positive error, which became significant above about 0.05 per cent. PRELIMINARY EXPERIMENTS- Initial experiments were made to determine the smallest amount of titanium necessary to effect complete reduction of tungsten and then to prepare a calibration graph in the presence of major amounts of zirconium. Effect of titanium-Samples of tungsten-free zirconium (0.2 g) were dissolved in a mixture of 25 ml of hydrochloric acid, sp.gr.l.lS, and 1 ml of fluoroboric acid. Ten millilitres of sodium tungstate solution (1 ml = 0.1 mg of tungsten) were added, and the solutions were transferred to 100-ml calibrated flasks. Amounts of titanous chloride solution (1 ml = 10 mg of titanium) ranging from 0 to 3 ml were added, followed by 10 ml of 15 per cent. ammonium thiocyanate solution. After dilution to the mark, optical densities were measured at 4000 A in 1-cm cells. Similar tests were made in the absence of tungsten in order to establish blank values. Appropriate blank values were deducted and the results showed that at least 5 mg of titanium111 were required for complete reduction of 1 mg of tungsten (see Table I). To THIOCYANATE METHOD FOR TUNGSTEN IN ZIRCONIUM 'rABLE 1 EFFECT OF TITANOUS CHLORLDE SOLUTION ON THE THIOCYANATE Each solution contained 0.2 g of zirconium PROCEDURE FOR DETERMINING TUNGSTEN I N ZIRCOXIUM Titanous chloride Optical-density Tungsten added, solution added Optical density difference mg ml Nil 0.0 0.002 1.0 0.0 0.031 0.029 Nil 0.26 0.003 1.0 _ .Xi1 1.0 Nil 1.0 Xi1 1.0 Nil 1.0 0.25 0.5 0.5 1.0 1.0 2.0 ~. 2.0 3.0 3.0 0.433 0.007 0.621 0.010 0.630 0.018 0.636 0.025 0.644 0.430 0.614 0.620 0.618 0.619 provide a reasonable safety margin, 1 ml of this reagent was used in all later experiments; this contributes about 0.01 to the optical density. Calibration gma@h-Sodium tungstate solution (1 ml E 0.1 mg of tungsten) in amounts ranging from 0 to 16.0 ml was added to solutions containing 0.2 g of tungsten-free zirconium prepared as before.The reagents were added and the optical densities were measured as before. A satisfactory calibration graph suitable for the determination of tungsten in the range 0.01 to 0.8 per cent. was obtained. EFFECT OF VARIOUS FACTORS- The effects of temperature, acidity, ammonium thiocyanate concentration and stability of the tungsten - thiocyanate complex were studied in the presence of zirconium, and resultsJune, 19581 TITANIUM, ZIRCONIUM AND THEIR ALLOYS 329 agreed with those from similar experiments relating to the determination of tungsten in titanium. Tests on solutions containing 0.2 g of zirconium and the equivalent of 0.10 and 0.80 per cent. of tungsten established that amounts up to at least 20 per cent.of tin, 5 per cent. of nickel or magnesium and 2.5 per cent. of chromium did not interfere. Copper in excess of 0.5 per cent. interfered, owing to precipitation of cuprous thiocyanate. Iron above about 2.5 per cent. interfered, owing to the yellow colour of ferric ions formed during dissolution of the sample, but interference (up to 5 per cent. of iron) was overcome by increasing the amount of titanous chloride to 2 ml. Molybdenum above 0.05 per cent. caused high results, but interference (up to 0-5 per cent. of molybdenum) was overcome by adding an equivalent amount of molybdenum to the blank solution. Vanadium above 0.01 per cent. interfered. EXTRACTION OF THE TUNGSTEN - THIOCYANATE COMPLEX- As a possible means of overcoming interference from alloying amounts of vanadium and molybdenum, and at the same time increasing the sensitivity of the method, attempts were made to extract the complex with organic solvents.Tests with isoamyl alcohol, isoamyl acetate, hexyl acetate, diethyl ether, chloroform and carbon tetrachloride indicated that only isoamyl alcohol extracted the complex to any appreciable extent, but even after four extrac- tions with this solvent it was not completely extracted, and so this possible extension of the procedure was discontinued. APPLICATION OF THE THIOCYANATE PROCEDURES The proposed thiocyanate procedures, see p. 331, were applied to samples of titanium, zirconium and their alloys, and, as shown in Table 11, reproducible results were obtained. TABLE I1 DETERMINATION OF TUNGSTEN IN SAMPLES OF TITANIUM, ZIRCOFIUM AND THEIR ALLOYS Sominal composition Titanium .. . . . . . . .. .. . . Titanium + 2 per cent. of aluminium + 2 per cent. of manganese . . . . . . .. .. .. Zirconium . . . . * . .. .. .. .. Zirconium + 1.5 per cent. of tin + 0.1 per cent. of iron + 0.1 per cent. of chromium + 0.05 per cent. of nickel . . . . .. .. .. .. .. Titanium* . . . . . . .. .. .. .. Titanium + 5 per cent. of aluminium + 2.5 per cent. of tin* . . .. .. . . .. .. .. Titanium + 2.5 per cent. of aluminium + 13 per cent. of tin* . . * . .. .. .. .. .. Titanium + 6 per cent. of aluminium + 4 per cent. of vanadium* .. . . . . .. * . .. Zirconium + 1.5 per cent. of tin + 0.1 per cent. of iron + 0.1 per cent. of chromium + 0.05 per cent. of nickel* . . .. .. .. .. .. Zirconium + 2.5 per cent.of tin* .. * . .. Tungsten found by thiocyanate method, % 0.44, 0.44, 0.44, 0.445, 0.445, 0.44 0.16, 0.16 0.47, 0.46 - 0.07, 0.07 0.40, 0.40, 0.405, 0.40, 0,395, 0.40 Tungsten found by dithiol method, 0.0015. 0.0015 % 0.002, 0.002 0.002, 0.002 0.002, 0.002 0.434, 0.430, 0.435, 0.428, 0.430, 0.430 0.170. 0.170 0.470, 0.475 0.425, 0.420 0.069, 0.070 0.402, 0.404, 0.404, 0.405, 0.400, 0.395 Standard deviation of thiocyanate method at 0.4 per cent. of tungsten = & 0.003 per cent. Standard deviation of dithiol method a t 0.4 per cent. of tungsten = 2 0.003 per cent. * Prepared by an arc-melting process with tungsten electrodes. DITHIOL METHOD FOR TUNGSTEN I N TITANIUM A mixture of sulphuric and phosphoric acids, such as that used for dissolving steel, is not suitable for dissolution of titanium because of the insolubility of titanium phosphate ; a mixture of sulphuric and fluoroboric acids was therefore used in the experimental work.330 [Vol.83 PRELIMINARY EXPERIMENTS- Preliminary experiments were made to establish the wavelength at which maximum absorption of the tungsten - dithiol complex occurs, and then, by using this wavelength, to prepare a calibration graph. Absorption curve-A 0.25-g sample (of tungsten-free titanium was dissolved in a mixture of 50 ml of dilute sulphuric acid (1 + 4) and 0.5 ml of fluoroboric acid, and the solution was oxidised with nitric acid, sp.gr. 1.412. Fifteen millilitres of sodium tungstate solution (1 ml = 0.1 mg of tungsten) were added and, after dilution to 100 ml, a 5-ml aliquot was taken and evaporated to fumes of sulphur trioxide.The solution was cooled, 5 mlof stannous chloride (10 per cent. in hydrochloric acid, sp.gr. 1.18) were added and the solution was heated on a boiling-water bath for 4 minutes. A 10-ml portion of dithiol solution (1 per cent. in isoamyl acetate) was added and heating was continued, with frequent shaking, for 10 minutes. The solution was then cooled, transferred to a separating funnel and the aqueous layer removed. The solvent layer was washed twice with diluted hydrochloric acid (4 + 1) and diluted to 50 ml with isoamyl acetate in a calibrated flask. The optical density of the solution was determined in a 4-cm cell with a Unicam SP600 spectrophotometer at wavelengths between 5000 and 6600 A ; the maximum absorption was found to be at 6300 A.Calibration graph-Solutions containing 0.25 g of tungsten-free titanium and amounts of sodium tungstate solution (1 ml E 0.1 mg of tungsten) ranging from 0 to 20 ml were diluted to 100 ml. Aliquots of 5 ml were taken and examined as described above, but the optical densities at 6300 A were not proportional to the amounts of tungsten added and colour development was not complete. Further tests with solutions containing tungsten only, and iron in addition to tungsten, indicated that the presence of iron is necessary in order to obtain quantitative values for tungsten To investigate the effect of the presclnce of iron, solutions containing 0.25 g of titanium and the equivalent of 1.5 mg of tungsten were diluted to 100 ml.Five-millilitre aliquots of these solutions, to which were added amounts of ferrous sulphate solution (1 ml = 5 mg of iron) ranging from 0 to 5.0 ml, were examined as before. The optical densities were measured at 6300 A. Increase in the amount of iron over the range 0 to 12.5 mg resulted in an increase in optical density of the tungsten - dithiol complex, but from 12.5 to 25.0 mg of iron there was no significant change in optical density and colour development was complete, as shown by the following results- 11 OOD AiXD CLARK: TH 5 DETERMINATION OF TUNGSTEN I N Iron present, mg . . Nil 1.0 5-0 10.0 12.5 15.0 17.5 20.0 25.0 Optical density . . 0.688 0.701 0 712 0.735 0.742 0.748 0.747 0.748 0,749 In all later experiments, 1 ml of ferrous snlphate solution (equivalent to 15 mg of iron) was added to the solutions immediately before evaporation to fumes of sulphur trioxide.By using this modification, satisfactory calibration graphs were prepared covering the ranges 0.05 to 0.8 per cent. and 0.002 to 0.08 per cent. of tungsten. EFFECT OF VARIOUS FACTORS- Tests showed that no significant variation in optical density of the complex occurred over the range 18" to 35" C and strict control of solution temperature during optical-density measurement was therefore not essential. The complex was stable for a period up to 1 hour, after which a decrease in optical density gradually occurred. In further tests, different amounts of sulphuric acid, sp.gr. 1.84, from 0 to 5 ml had no significant effect on the optical density of the final solution.Increase in the amount of 10 per cent. stannous chloride solution over the range 2 to 10 ml had, likewise, a negligible effect. Below 2 ml, the optical density decreased with decrease in the stannous chloride concentration, owing to incomplete reduction of the tungsten. Different times of boiling with stannous chloride from 1 to 6 minutes had no effect on the optical density of the complex. At least 7.5 ml of dithiol solution were required in order to obtain full development of the colour, and no change in the optical density occurred over the range 7.5 to 15 ml. Vari- ation of the time of boiling with this reagent from 3 to 15 minutes had no significant effect. Tests by the proposed dithiol procedure (see p. 332) established that amounts up to about 20 per cent.of tin, manganese or varladium, 10 per cent. of aluminium and 5 per cent. of copper, iron, chromium or nickel did not interfere in the determination of 0-02 and 0.60 per cent. of tungsten.June, 19581 TITANIUM, ZIRCONIUM AND THEIR ALLOYS 331 Molybdenum forms a green complex with dithiol under test conditions and this caused serious interference, Molybdenum has not, so far, occurred as a significant impurity in titanium, but it is used as an alloying constituent and the presence of a large amount of this metal had, therefore, to be considered. Attempts to overcome molybdenum interference in the presence of titanium by extraction of the dithiol complex with isoamyl acetate before reduction of tungsten, in accordance with a published method for the examination of steel,4 were unsuccessful and the results were erratic.Results were also erratic when the method of Short6 was used. High results were attributed to incomplete extraction of molybdenum and low results to extraction of some tungsten during the preliminary separation of molybdenum. Interference was overcome, however, by precipitation of molybdenum sulphide in acid - tartrate solution. In this method, prior oxidation of the solution with potassium permanganate is necessary, otherwise the results are high, due, presumably, to reduction of molybdenum (by titanous ions) to a lower valency state, which results in incomplete precipitation. The results in Table I11 show that this procedure can be satisfactorily applied to solutions containing from 0.1 to 10 per cent.of molybdenum, 0.02 to 0.6 per cent. of tungsten and major amounts of titanium. TABLE I11 DETERMINATION OF TUNGSTEN BY THE MODIFIED DITHIOL PROCEDURE FOR Each solution contained 0.25 g of titanium Molybdenum added, Tungsten added, Tungsten found, Yo % % USE IN THE PRESENCE OF MOLYBDENUM 0.1 1.0 10.0 0.1 1.0 10.0 0.1 1.0 10.0 0.02 0.02 0.02 0.05 0.05 0.05 0.60 0.60 0.60 0.021 0.022 0.023 0.050 0.051 0.051 0.61 0.612 0.611 DITHIOL METHOD FOR TUNGSTEN IN ZIRCONIUM Calibration graphs over the range 0.05 to 0.8 per cent. and 0.002 to 0.08 per cent. of tungsten, prepared from solutions also containing 0.25 g of zirconium, were identical with those prepared in experimental work relating to tungsten in titanium. APPLICATION OF THE DITHIOL PROCEDURE The dithiol procedure for 0.002 to 0.3 per cent.of tungsten was applied to samples of titanium, zirconium and their alloys and, as shown in Table 11, results were reproducible and agreed with those found by the thiocyanate procedures. METHODS THIOCYANATE METHOD FOR THE DETERMINATION OF TUNGSTEX IN TITANUM AND ITS ALLOYS REAGENTS- Hydrochloric acid, sp.gr. 1.18. Fluoroboric acid-To 280 ml of hydrofluoric acid maintained at 10" C add, in small Ammonium thiocyanate solution, 15 per cent.-Dissolve 75 g of ammonium thiocyanate Standard tungsten solution-Dissolve 0,1794 g of sodium tungstate, Na,W04.2H,0, in amounts, 130 g of boric acid and stir well. in about 250 ml of water and dilute to 500 ml. water and dilute to 1 litre. PREPARATION OF CALIBRATION GRAPH FOR 0.05 TO 1.6 PER CENT.OF TUNGSTEN- Transfer 0.2-g portions of tungsten-free titanium to each of seven beakers and dissolve each portion in 20 ml of hydrochloric acid and 1 ml of fluoroboric acid; warm gently to assist Store in a polythene bottle. 1 ml = 0.1 mg of tungsten.332 [Vol. 83 dissolution. Cool and add, separately, 4.0, 12.0, 16.0, 20.0, 24.0 and 32.0ml of standard tungsten solution. Dilute each solution to 100ml in a calibrated flask and transfer 25-ml aliquots to separate 100-ml calibrated flasks. Add 20 ml of hydrochloric acid and 10 ml of ammonium thiocyanate solution to each. Dilute to the mark and set them aside for 5 minutes at 20" i: 1" C. Measure the optical density of each at a wavelength of 4000 A in 2-cm cells. PROCEDURE- Determine a blank value by using a weight of tungsten-free titanium corresponding approximately to the weight of titanium in the sample.Dissolve a 0.2-g sample in 20 ml of hydrochloric acid and 1 ml of fluoroboric acid; warm gently to assist dissolution. (For amounts of tungsten in excess of 1.6 per cent. up to 3.2 per cent., use a smaller weight of sample.) Cool, dilute to 100ml in a calibrated flask and proceed as described for the preparation of the calibration graph. Calculate the tungsten content of the sample from the calibration graph. Note that vanadium above about 0.05 per cent. and molybdenum above about 0.25 per cent. introduce significant positive errors. Molybdenum (up to about 2 per cent.) can be corrected for by adding an equivalent amount of molybdenum to the blank solution.For samples containing vanadium or molybdenum in excess of these limiting amounts, the dithiol method is recommended. THIOCYANATE METHOD FOR THE DETERMINATION OF TUNGSTEN IN WOOD AND CLARK: THE DETERMINATION OF TUNGSTEN I N Use the remaining solution as a blank. ZIRCONIUM AND ITS ALLOYS REAGENTS- The reagents used are the same as for the determination of tungsten in titanium, plus- Titanous chloride solution-Dissolve 1.0 g of titanium in about 50 ml of hydrochloric acid, sp.gr. 1-18. Cool and dilute with hydrochloric acid to 100ml. This reagent must be freshly prepared. Transfer 0.2-g portions of tungsten-free zirconium to each of seven beakers and dissolve each portion in 25 ml of hydrochloric acid ;and 1 ml of fluoroboric acid; warm gently to assist dissolution.Cool and add, separately, 2.0,4*0, 6-0, 8.0, 12.0 and 16.0 ml of standard tungsten solution. Transfer each solution to a 100-ml calibrated flask, washing with about 25 ml of water. Add 1 ml of titanous chloride solution and 10 ml of ammonium thiocyanate solution to each Dilute to the mark and set them aside for 5 minutes at 20" Measure the optical density of each at a wavelength of 4 0 0 0 ~ in 1-cm cells. PROCEDURE- Determine a blank value on the reagents with each batch of samples. Dissolve a 0.2-g sample in 25 ml of hydrochloric acid and 1 ml of fluoroboric acid; warm gently to assist dissolution. (For amounts of tungsten in excess of 0.8 per cent. up to 3.2 per cent., use a smaller weight of sample. When the tungsten content is below 0.05 per cent., use 4-cm cells.) Cool and transfer the solution to a 100-ml calibrated flask with about 25 ml of water, and then continue as described for the preparation of the calibration graph.Calculate the tungsten content of tht: sample from the calibration graph. Note that vanadium above about 0.01 per cent. and molybdenum above about 0.05 per cent. introduce significant positive errors. Molybdenum (up to about 0.5 per cent.) can be corrected for by adding an equivalent amount of molybdenum to the blank solution. For samples containing vanadium or molybdenum in excess of these limiting amounts, the dithiol method is recommended. DITHIOL METHOD FOR THE DETERMINATION OF TUNGSTEN IN TITANIUM, PREPARATION OF CALIBRATION GRAPH FOR. 0.01 TO 0.8 PER CENT. OF TUNGSTEN- Use the remaining solution as a blank.1" C. ZIRCONIUM AND THEIR ALLOYS REAGENTS- 1.84. Mix well and cool. Sulphuric acid, dilute (1 + 4)-To 400 ml of water, add 100 ml of sulphuric acid, sp.gr. Fluoroboric acid-Prepare as described on p. 331.June, 19581 TITANIUM, ZIRCONIUM AND THEIR ALLOYS 333 Nitric acid, sp.gr. 1.42. Iron soZutioH-Dissolve 1.5 g of electrolytic iron in 25 ml of dilute sulphuric acid (1 + 4) and dilute to 100ml. 1 ml z 0.015 g of iron. SnC1,.2H,O, in 100ml of hydrochloric acid, sp.gr. 1.18 100ml of water. water and dilute to 1 litre. For use, dilute 100ml of this solution to 1 litre. PREPARATION OF CALIBRATION GRAPHS- Dissolve 0.625g of tungsten-free metal (titanium or zirconium) in 125ml of dilute sulphuric acid (1 + 4) and 1 ml of fluoroboric acid; warm gently to assist dissolution and then oxidise with a few drops of nitric acid.Boil to remove nitrous fumes, cool and dilute to 250 ml in a calibrated flask. Calibration graph for 0.05 to 0.8 per cent. of tungsten-Transfer five 5-ml aliquots of the solution (titanium or zirconium) to each of five 100-ml conical flasks and add, separately, 2.5, 5-0, 7.5 and 10.0ml of standard tungsten solution. Use the remaining solution as a blank. Cool, add 5ml of stannous chloride solution to each, place them on a boiling-water bath and agitate at frequent intervals over 4 minutes. Add 10ml of dithiol solution and continue heating on the water bath, with frequent agitation, for 10 minutes. Cool to about 30" C and transfer each solution to a separating funnel, rinsing with three 2-ml portions of isoamyl acetate.Shake and allow the layers to separate; run off and discard each lower acid layer. Wash the solvent layers twice with 10-ml portions of diluted hydrochloric acid (4 + 1) and each time discard the lower acid layers. Remove the upper solvent layers containing the tungsten, and transfer them to 50-ml calibrated flasks that have previously been washed free from water with ethanol and then with isoamyl acetate. Dilute each solution to the mark with isoamyl acetate and mix well. Measure the optical density of each solution at a wavelength of 6 3 0 0 ~ in 4-cm cells. Calibration graph for 0.002 to 0.08 per cent. of tungsten-Transfer five 25-ml aliquots of the solution (titanium or zirconium) to each of five 100-ml conical flasks and add, separately, 1-25, 2.5, 3.75 and 5.0 ml of standard tungsten solution.Use the remaining solution as a blank. Add 1 ml of the iron solution to each and continue as described for the preparation of the calibration graph for 0.05 to 0.8 per cent. of tungsten, but transfer the final isoamyl acetate layers containing the tungsten to 25-ml calibrated flasks washed free from water. Dilute each to the mark with isoamyl acetate and continue as previously described. PROCEDURE FOR 0.002 TO 0.8 PER CENT. OF TUNGSTEN- Dissolve a 0-25-g sample in 50 ml of dilute sulphuric acid (1 + 4) and 0.5 ml of fluoroboric acid; warm gently to assist dissolution and oxidise with a few drops of nitric acid. Boil to remove nitrous fumes, cool and transfer to a 100-ml calibrated flask and dilute to the mark.Transfer 5 ml of the solution to a 100-ml conical flask, add 1 ml of the iron solution and con- tinue as described for the preparation of calibration graph for 0.05 to 0.8 per cent. of tungsten. Calculate the tungsten content of the sample from the calibration graph. If the tungsten content of the sample is below 0.05 per cent., transfer a 25-ml aliquot from the remainder of the 100 ml of solution to a 100-ml conical flask, add 1 ml of the iron solution and continue as described for the preparation of the calibration graph for 0.002 to 0.08 per cent. of tungsten. PROCEDURE FOR TUNGSTEN IN SAMPLES CONTAINING MOLYBDENUM- Dissolve a 0.25-g sample in 25 ml of dilute sulphuric acid (1 + 4) and 0.5 ml of fluoro- boric acid; warm gently to assist dissolution and oxidise by dropwise additions of a saturated solution of potassium permanganate, until a permanent brown precipitate is present.Add Stannous chloride solution, 10 per cent. w/v-Dissolve 10 g of stannous chloride, Dithiol solution-Dissolve 1 g of toluene-3 : 4-dithiol in 100 ml of isoamyl acetate. Hydrochloric acid, diluted (4 + 1)-Dilute 400 ml of hydrochloric acid, sp.gr. 1-18, with Standard tungsten solution-Dissolve 0.1794 g of sodium tungstate, Na,W0,.2H,O, in 1 ml = 0.01 mg of tungsten. Add 1 ml of the iron solution to each and evaporate to fumes of sulphur trioxide.334 WOOD AND CLARK [Vol. 83 sulphurous acid dropwise until the precipitate has dissolved and then boil for 2 minutes to remove excess of sulphur dioxide. Add 15 ml of 50 per cent.tartaric acid solution, 250 ml of water and warm to 80” C. Pass a rapid stream of hydrogen sulphide through the solution €or 30 minutes and then allow to cool to room temperature. Transfer the solution and precipitate to a 500-ml calibrated flask and dilute to the mark. Filter an approximately 50-ml portion through a Whatman No. 42 filter-paper, transfer 25 ml of the filtrate to a 100-ml conical flask, add 1.0ml of dilute sulphuric acid (1 + 4) and 1 ml of the iron solution and then evaporate to fumes of sulphur trioxide. Destroy carbonaceous matter by adding nitric acid, sp.gr. 1.42, in 2-ml portions to the hot residue. Continue as described for the preparation of the calibration graph for 0.05 to 0.8 per cent. of tungsten. Calculate the tungsten content of the sample from the calibration graph. If the tungsten content of the sample is below 0.05 per cent., transfer a 100-ml aliquot of the filtrate to a 250-ml conical flask, add 1.5 ml of sulphuric acid, sp.gr.1-84, and 1 ml of the iron solution and then evaporate to fumes of sulphur trioxide. Destroy carbonaceous matter and continue as described for the preparation of the calibration graph for 0-002 to 0.08 per cent. of tungsten. CONCLUSIONS The proposed absorptiometric thiocyariate procedure is satisfactory for the determination of 0.05 to 1.6 per cent. of tungsten in titanium and many of its alloys. Vanadium above 0.05 per cent. and molybdenum above 0.5 per cent. interfere. Interference by molybdenum, up to about 2 per cent., can be overcome by adding an equivalent amount of molybdenum to the blank solution. The proposed thiocyanate procedure for the examination of zirconium and its alloys is applicable over the range 0.01 to 1.6 per cent.of tungsten. Vanadium above 0.01 per cent. and molybdenum above 0.05 per cent. interfere. Interference by molybdenum, up to 0.5 per cent., can be compensated for. Both procedures can be extended to the deter- mination of tungsten up to about 3.2 per cent. by using a smaller weight of sample. The procedures are simple, rapid and particula.rly suitable for control analysis. The dithiol procedure is suitable for the determination of tungsten in the range 0.002 to 0.8 per cent. in titanium, zirconium and their alloys. By using smaller absorption cells, it can be extended to the determination of am.ounts of tungsten up to about 3.2 per cent. The presence of iron is necessary in order to obtain quantitative results. Vanadium does not interfere, and interference by molybdenum can be overcome by incorporating a preliminary precipitation of molybdenum sulphide. The dithiol method is more time-consuming than the thiocyanate method and is, therefore, only recommended in the examination of samples containing amlsunts of vanadium or molybdenum sufficient to cause interference with the thiocyanate procedure, and for general application to materials containing less than about 0.05 per cent. of tungsten. We thank Mr. W. T. Elwell, Division Chief Analyst, for helpful suggestions and assistance in preparation of this paper. REFERENCES 1. 2. 3. 4. 5. 6. 7. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Second Edition, Interscience Gentry, C. H. R., and Sherrington, L. G., Analyst, 1948, 73, 57. Freund, H., Wright, M. L., and Brookshier, R. K., Anal. Chew., 1951, 23, 781. Bagshawe, B., and Trueman, R. J.. Analy.ct, 1947, 72, 189. Bacon, A., Royal Aircraft Establishment Technical Note N o . MET.119, 1950. Short, H. G., Analyst, 1951, 76, 710. “The Analysis of Titanium and its Alloys,” Imperial Chemical Industries Ltd., First Edition, Received December 12th, 1957 Publishers Inc., h’ew York and London, 1950, Volume 111, p. 584. London, 1956.