642
.
INDUSTRIAL AND ENGIKEERING CHEMISTRY
to small volume three times t o drive off the excess nitric acid. This was diluted with water to about 250 ml. and sufficient c. P. potassium cyanide was dissolved in it to su ply the free cyanide content of the average gold plating bat’h Pabout 10 grams per liter, 3) upon dilution t o 1 liter. This solution was now carefully transferred to a I-liter volumetric flask and carefully diluted t o the mark with distilled water. T.4BLE I. ACCERACY O F blETHoDS Evaporation Sulfuric Sample Weight Method Difference Method G./1. G./1. MQ. G./1. 0.1048 -0.2 1 0,1050 0.1051 0.4999 -0 2 0.4998 2 0.5001 5.0044 -0.1 5.0044 3 5.0046 20.0099 -0.1 20.009s 4 20.0100
Av. -0.15
Difference Mg. f0.1 -0.3 -0.1 -0.2 A V . -0 13
Four samples prepared this may were analyzed by both
methods with the results given in Table I. This comparison shows that the sulfuric acid method is accurate for routine determinations of gold in cyanide plating solutions. I n actual laboratory practice the author has used this method with great success during the past year, his evaluation analyses closely checking with those of the chemists of submitting firms.
VOL. 10, NO. 11
The method will be found very convenient in control analyses, as from one sample the following determinations may be made: free cyanide, gold, silver if any (by precipitating silver in filtrate with hydrochloric acid and subtracting the known amount of silver added), and other base metals such as copper, nickel, and zinc, by analyzing the filtrate according t o standard methods.
Literature Cited (1) Bamford and Harris, “The Metallurgist’sManual,” p . 2 5 , London, Chapman and Hall, 1927. (2) Blum and Hogaboom, “Principles of Electroplating and Electroforming,” 2nd ed., p . 371, New York, McGraw-Hill Book Co., 1930. (3) Ibid., pp. 3656. ( 4 ) Fulton and Sharwood, “Manual of Fire Assaying,” 3rd ed., pp. 191-2, Kew York, McGraw-Hill Book Co., 1929. ( 5 ) Scott, TT. W., “Technical Methods of Metallurgical Analysis,” pp. 718-19, New York, D. Van Nostrand Co., 1933. (6) Treadwell and Hall, “Analytical Chemistry,” 5’01. 2, 7th ed., p. 606, New York, John Wiley & Sons Co., 1930. (7) Weisberg, L., private communication, 1934, RECEIVED August 6, 1938.
p-Hpdroxvphenylarsonic Acid as a Reagent for Titanium and Zirconium J
C. T. SI1IIPSOR’’ ASD G. C. CHANDLEE, The Pennsylvania State College, State College, Pa.
S
EVERAL investigators have reported the value of certain arsonic acids as reagents for the gravimetric determination of zirconium (2, 4, 6 ) , thorium (4, tin (S), and iron (1). The need for a reagent that would give a convenient and satisfactory separation of titanium from the common elements, and particularly from iron, led to a further investigation of this field. p-Hydroxyphenylarsonic acid was found to be a very favorable reagent for titanium. I n the presence of dilute mineral acid it gives a n effective separation of titanium from iron and most other commonly occurring elements in one precipitation. It is also a n excellent reagent for zirconium. I n mineral acid solutions not stronger than 0.5 K it also gives a quantitative precipitation of tin, but the possibility of using it as a reagent for this element has not been investigated in detail.
tate filtering. After cooling to room temperature filter off the precipitate. Kith a good paper (No. 42 Whatman) and a filter cone one may employ suction advantageously. Wash the precipitate about five or six times with a wash liquor of dilute (0.25 A‘) hydrochloric or sulfuric acid containing about 0.5 gram of reagent per 100 cc. Rhen iron is present 1 or 2 grams of ammonium thiocyanate should also be added to each 100 cc. of this liquor. Finally wmh the precipitate two or three times with a dilute (2 per cent) aqueous solution of ammonium nitrate, and then ignite it in a porcelain crucible (with propped lid) at low temperature until all the carbon is burned off, then at the full heat of a Bunsen or Fisher burner until constant weight is attained, leaving a residue of titanium dioxide. The ignition must be carried out in an efficient fume hood.
An average deviation of 0 . i part per thousand was found in analyzing pure standard solutions of titanium by this method.
Separation of Titanium from Mixtures Determination of Titanium MATERIALS.The p-hydroxyphenylarsonic acid used in this investigation was supplied by the Mallinckrodt Chemical Works and was suitable for use without further purification. A 4 er cent aqueous solution was found to be convenient for use. ure standard solutions of titanium sulfate were prepared and standardized by qccepted methods. All other reagents were of c. P. or equivalent grade. Dissolve the sample (containing not more than PROCEDURE. about 0.06 gram of titanium dioxide) in hydrochloric or sulfuric acid solution and remove interfering elements by appropriate means. The amount of acid present should be such that the solution will be approximately but not more than 0.60 N in hydrochloric or 1.80 N in sulfuric acid after the reagents have been added and the precipitation is complete. After adjusting the volume to about 200 cc., heat the solution to boiling and (after the addition of 2 to 3 grams of ammonium thiocyanate when iron is present) add 100 cc. of a 4 per cent aqueous solution of phydroxyphenylarsonic acid. Continue boiling gently for at least 15 minutes to coagulate the precipitate and thus facili-
-r:
*
Present address, Rahm a n d H a a s Company, Bristol, Pa.
IKON A K D PHOSPHATE. A synthetic sample (containing 0.0636 ) gram of TiOz, 0.1392 gram of FezO3, and 0.071 gram of P z O ~~v-as
analyzed by the above procedure for titanium, with a recovery of 0.0636 gram of titanium dioxide. Another sample, containing the same amount of titanium and iron but no phosphate, gave the same result.
ALUMINUM, ZISC,
COBaLT, KICKEL, BERYLLIUAf, AND BASIC
CHRonirunr AKD MASGANESE. From a composite sample (containing 0.0521 gram Ti02,0.051 gram of A120310.063 gram of CrZO3, 0.079 gram of MnO. 0.081 gram of ZnO, 0.075 gram of COO,0.075 gram of K O , and 0.030 gram of BeO) 0.0521 gram of titanium dioxide was easilyseparated. From another such mixture containing 0.0636 gram of titanium dioxide this procedure returned 0.0636 gram. C a L c I u h i AND >1.4GNESI~>l. TKO samples containing, respectively, 0.0521 and 0.0636 gram of titanium dioxide with a mixture of 0.056 gram of calcium oxide and 0.040 gram of magnesium oxide were analyzed. The results were 0.0521 and 0.0636 gram of titanium dioxide. DICHROMATE, PERUANGANATE, L-RANYL, AKD V A N A D Y L IONS.
A composite sample (containing 0.0636 gram of TiOa,0.052 gram of Cr20,, 0.071 gram of MnO, 0.180 gram of V ~ O Sand , 0.286
NO\-EMBER 15, 1938
643
ANALYTICAL EDITION
gram of LO,) was used. It was necessary to use 5 rather than the usual 4 grams of reagent to secure a quantitative precipitation. Otherwise the separation was normal, 0.0636 and 0.0637 gram being found in duplicate analyses. R~OLYBDATE. No difficulty was experienced in separating 0.0636 gram of titanium dioxide from 0.071 gram of molybdenum oxide present as ammonium molybdate. Duplicate analyses gave 0.0637 and 0.0639 gram of titanium dioxide. THALLIUM ASD CERIchf (CEROUS). The separation of 0.0636 and 0.0521 gram of titanium dioxide, respectively, from 0.213 gram of cerous oxide (present as cerous nitrate) mixed with 0.076 gram of thallic oxide was satisfactory, yielding 0.0637 and 0.0521 gram of titanium dioxide. However, cerium in the ceric form cannot be separated from titanium by this method. THORIUM. Some difficulty was encountered here. Apparently thorium ties up sonie of the arsonic acid as an undissociated soluble complex, so that the regular procedure gives low results for titanium. It Tyas found best to use sulfuric rather than hydrochloric acid, and also to use considerable excess of reagent. Gsing 6 grams of parahydroxyphenylanonic acid and sulfuric acid solution, 0.0636 gram of titanium dioxide was easily separated from 0.0516 gram of thorium oxide by one precipitation, duplicate samples giving 0.0637 and 0.0638 gram of titanium dioxide. CHROME-VASADIUhf STEEL. A 2-gram Sample O f Bureau O f Standards chrome-vanadium steel No. 30-C (containing 0.76 per cent sio2, 1.43 per cent C1-203, 0.64 per cent h h o , 0.10 per cent NiO, 0.045 per cent P20s,0.19 per cent Mo03, and 0.125 per cent CuO) was dissolved in dilute hydrochloric acid and the silica was removed in the usual way. Then, since the sample contained no titanium, 25 cc. of a solution of titanium sulfate containing 0.0621 gram of titanium dioxide were added, followed by 5 grams of ammonium thiocyanate. The titanium was then precipitated and determined as usual. The ignited residue was white, containing no trace of iron oxide. Duplicate samples each gave a recovery of 0.0521 gram of titanium dioxide. IROXORE. Four separate 5-gram samples of Bureau of Standards iron ore No. 29 (containing 12.02 per cent SiOl, 1.91 per cent &03,0.08 per cent V20!, 0.09 per cent h h o , 2.90 per cent CaO, 2.01 per cent MgO, 0.31 per cent KzO, 0.45 per cent Ka20, and 0.99 per cent TiOl) were taken up with 100 cc. of dilute hydrochloric acid and filtered, and the undissolved residue was fused with sodium carbonate. The fusion mas wished into the main filtrate and after removing the silica in the usual manner, 5 grams of ammonium thiocyanate were added, the solution was diluted to 300 cc., and the titanium was precipitated with 4 grams of p-hydroxyphenylarsonic acid dissolved in 100 cc. of water. The analysis as finished in the usual way. The titanium dioxide results were 0.964, 0.970, 0.970, and 0.976 per cent, respectively. BURNTREFRACTORY. Bureau of Standards burnt refractory No. 78 (20.69 per cent SiOl, 59.97 per cent A1203, 0.79 per cent Fe203,0.51 per cent MgO, 0.38 per cent CaO, 2.83 per cent K,O, 0.53 per cent Sa,O, 0.62 per cent P~OS,0.047 per cent V Z O S , 0.12 per cent Zi-02, and 3.37 per cent Ti02) was used. Four 2gram sample; n-ere run for titanium by the following procedure: The sample was fused with 15 grams of potassium pyrosulfate and the fusion was taken up with 100 cc. of viater containing 9 cr. of concentrated sulfuric acid. The residue TTas filtered off, and the .silica was volatilized with a mixture of hydrofluoric and sulfuric acids. The small remaining residue !vas again fused nith a small amount of potassium pyrosulfate and then dissolved in the original filtrate. After filtration, the solution was diluted to 225 cc., 2 grams of ammonium thiocj-anate were added, and the analysis for titanium was finished as usual with p-hydroxyphenylarsonic acid. The final titanium oxide figures viere obtained by subtracting the Bureau of Standards figure for zirconium oxide from the weight of the mixed titanium and zirconium oxide, found in the preFent work. The four results were 3.33, 3.31, 3.35, and 3.35 per cent of titanium dioxide, respectively. Bureau of Standard5 figures varied from 3.18 to 3.68 per rent of titanium dioxide, with an average of 3.37 per cent. PLasrIc C L ~ Y Bureau . of Standards plastic clay No. 98 [59.11 per cent Sio?, 25.54 per cent & 0 3 , 2.05 per cent Fe2O3,0.08 per cent PlOs, 0.025 per cent V20s, 0.021 per cent ( 3 2 0 3 , 0.21 per cent CaO, 0.72 per cent MgO, 3.17 per cent K 2 0 ,0.28 per cent Ka20, 0.07 per cent SOa, 0.005 per cent MnO, 0.009 per cent CuO, 0.041 per cent Zr02, and 1.43 per cent Ti02 (average of several analyses varying from 1.35 to 1.50 per cent)] was analyzed for titanium. Triplicate 2-gram samples were fused with sodium carbonate, and taken up in dilute hydrochloric acid, and evaporated to dryness on asteam bath. The residue was extracted with dilute hydrochloric acid and the insoluble material was then filtered off and evaporated n-ith a mixture of sulfuric and hydrofluoric acids to dryness, finally being fused again with a small amount of sodium carbonate. The fusion was dissolved in the original filtrate and ammonia was added until the solution re-
mained just slightly acid. Tn-o grams o f ammonium thiocyanate were added and the titanium was determined with p-hydroxyphenylarsonic acid in the usual way. After subtracting 0.041 per cent of zirconium oxide 1.41, 1.41, and 1.41 per cent of titanium dioxide was found. ZIRCONIUM. The method already outlined for titanium precipitates zirconium as well, so t,hat ivhen both metals are present in the sample separate analyses must be made for each. The zirconium alone may be precipitated in the presence of excess hydrogen peroxide, while in a second sample the zirconium and titanium are both precipitated (in the absence of hydrogen peroxide) and determined as the mixed oxides. This is preferable to trying to determine the tiianium in the filtrate containing hvdrogen peroxide. If it is desirable to run both zirconium and titanium on the same sample, the best procedure for determining the titanium in the filtrate from the zirconium is to evaporate to fumes, add nitric acid, and evaporate again to fumes. After taking up in water, the acidity is adjusted to the desired concentration with ammonia and the titanium is precipitated and determined by t,he procedure given above for titanium. Attempts to eliminate hydrogen peroxide without evaporating d0n.n to white fumes proved impractical. The separation of zirconium from titanium is best effected in 2.5 to 3.0 N sulfuric acid solution. A large excess of hydrogen peroxide (15 to 20 cc. of 30 per cent solution) must be present to prevent titanium from being precipitated, and hydrogen peroxide should also be added to the mash liquor. Moreover, several times the usual quantity of reagent must be used. In separating 0.10 gram of zirconium oxide from 0.05 gram of titanium dioxide it was found necessary to employ 3.0 grams of reagent, instead of the usual 1.0 gram required to give a quantitative separation of this amount of zirconium from most of the common elements other than titanium. The general procedure for determining zirconium is essentialIy the same as that outlined for titanium, except that a mineral acid concentration of 2.5 to 3.0 is employed and less reagent is necessary. This method is particularly applicable to the determination of zirconium in the presence of a large amount of iron. Quantities of zirconium oxide ranging from 0.004 to 0.10 gram were quantitatively separated from 10 grams of the aforementioned chrome-vanadium steel in one precipitation. If present in more than very m a l l amounts, phosphate interferes with the determination of zirconium by this method.
Summary p-Hydroxyphenylarsonic acid has been found a n advantageous reagent for separating titanium from the following ions : ferric, ferrous, aluminum, zinc, cobalt, nickel, beryllium, chromic, manganous, calcium, magnesium, thallium, cerous, thorium, sodium, potassium, ammonium, as well as phosphate, molybdate, chromate, vanadate, permanganate, uranyl, and vanadyl. Interfering ions are zirconium, cerium (ic), and tin. Hydrogen peroxide also must be absent. The reagent may be used to determine zirconium in the presence of the above ions. Phosphate in more than very small amounts interferes. Zirconium may be separated from titanium if hydrogen peroxide is present. Cerium (ic) and tin interfere.
Acknowledgrrient The authors are grateful to F. C. Whitmore, D. 11.Jones, and the hIallinckrodt Chemical Workq for making available certain valuable reagents ubed in this investigation.
Literature Cited (1 Craig, K. A , and Chandlee, G. C , J . -4m. Chem. SOC.,56, 1278 11934) (2) Geist, H. H., and Chandlee, G. C., IXD. ENG.C H m f . , Anal. Ed., 9. 169 (1937). ~~
~
( 3 ) Knapper, J. S., Craig, K. .A , and Chandlee, G. C., J Am. Chem. SOC.,55, 3945 (1933). (4) Rice, A. C., Fogg, H. C., and James, C., Ihid., 48, 895 (1926). (5) Stambaugh, 0. F., M.S.thesis, The Pennsylvania State College,
1933. RECEIVED March 3 , 1938.
.