THE EXTRACTION OF TIN FROM ITS ORES AND THE PREPARATION AND BEHAVIOR OF CERTAIN PME SALTS OF TIN. by Harry Borden Marshall A Thesis submitted for the Degree of MASTER OF ARTS i n the Department CHEMISTRY The University of B r i t i s h Columbia A p r i l - 1931. TABLE OF CONTENTS INTRODUCTION. I. EXTRACTION OF TIN FROM ITS ORES. 1. Occurrence of Samples. 2. Disintegration of G a s s i t e r i t e . 3. Disintegration of Stannite. 4. P u r i f i c a t i o n of C a s s i t e r i t e Tin. 5. P u r i f i c a t i o n of Stannite Tin. I I . THE STABILITY OF STANNIC OXIDE. I I I . THE DETERMINATION OF TIN AS CAESIUM CHLOROSTANNATE. 1. Preparation of Stannic Chloride. 2. Preparation of Caesiian r0hloride. 3. Method. IV. SOLUBILITY OF CAESIUM CHLOROSTANNATE. ACKNOWLEDGMENT I wish to express my greatest appreciation and thanks to Dr. E.H. Archibald, who has not only provided many materials necessary i n this investigation but has also gladly given every possible assistance and advice. THE EXTRACTION OF TIN FROM ITS ORES AND THE PREPARATION AND BEHAVIOR OF CERTAIN PURE SALTS OF TIN. INTRODUCTION. The existence of isotopes has explained a great many-discrepancies i n the theory of atomic structure. I t has also d e f i n i t e l y s e t t l e d the v a l i d i t y of the whole number rule for the atomic weights of the elements. Henee any experimental work whieh w i l l corroborate the existence of isotopes, w i l l be of great importance i n strengthening the proof of the atomic Theory. 1 F. W. Aston, by using the p o s i t i v e ray spectograph, and pla t e s of increased s e n s i t i v i t y to po s i t i v e rays, has found i t possible to d e f i n i t e l y prove that a group of eight l i n e s , corresponding approximately to atomic weights of 116, 117, 118, 119, 120, 121, 122, 124, was due to Tin. I t i s extremely l i k e l y that these eight isotopes do exist i n the case of t i n , but as yet no dir e c t chemical evidence has proven the i r existence. One method of attacking t h i s problem i s to determine the atomic weight of t i n i n i t s d i f f e r e n t ores. Due to extremely diffe r e n t geological formation there may be wide variations i n the concentration of the isotopes i n these different ores. Such a v a r i a t i o n might be detected by separate atomic weight determinations of the t i n i n the different samples. This was the o r i g i n a l problem attempted. However, the extraction of t i n from i t s ores proved to constitute a problem i n i t s e l f , and t h i s thesis w i l l deal l a r g e l y with t h i s part of the work, together with the preparation of pure t i n from the material extracted. Considerable work has been carried out towards evolving a convenient and generally applicable method of analysing t i n ores. This search has brought f o r t h many methods of di s i n t e g r a t i o n and may be b r i e f l y reviewed under two headings: A. Reduction methods i n which the oxide i s reduced to m e t a l l i c t i n . B. S a l i f i c a t i o n methods i n whieh advantage i s taken of the acid properties of stannie oxide i n the formation of a l k a l i stannates. A. Reduction Methods: Cornouailles f i r s t used carbon as a reducing agent f o r 2 t i n ores. The process has been studied by l e v o l and Moissenet, 4 but i t i s best described by Beringer. He also suggests the addition of a l i t t l e powdered f l u o r spar to assist the fusion of refr a c t o r y slags. The method of reduction i n hydrogen or ill u m i n a t i n g gas 5 i s due to Hampe, but has been described i n some d e t a i l by 6 ,7 Parry. Fawns considers i t the most p r a c t i c a l method of reduction. Reduction with potassium cyanide has been used by 6 8 assayists f o r some time. Parry and Bayerlein-Essen have outlined t h e i r methods of analysing t i n ore with t h i s reagent. 3 9 a. M. Henderson has found that t h i s method of fusion i s out 10 of the question f o r low grade ores. 0. Boy ascribes t h i s i n e f f i c i e n c y f o r low grade ores to the presence of sulphur or s i l i e a , and modifies the procedure by f i r s t roasting to remove sulphur, and then b o i l i n g with hydrofluoric acid to remove s i l i e a . B. S a l i f i c a t i o n Methods: Fusion with sodium carbonate and sulphur to form soluble 11 12 sulpho-stannates, has been worked out by Golick and C o r t i . 13 J. Gray has found t h i s to be an accurate method, but describes i t as " d i r t y , tedious, and slow." This conclusion has been affirmed by S. Fawns who also adds that f o r t h i s reason the method i s never used. 14 E. ?. Pearce suggested the use of sodium hydroxide for the d i s i n t e g r a t i o n of t i n ores and worked out a scheme of analysis based on t h i s method of fusion. The present accepted method i s a s l i g h t modification of Pearce's o r i g i n a l procedure 15 16 and i s c a l l e d the Pearce-low method. R.J. Morgan also uses the a l k a l i fusion method for dis s o l v i n g t i n ores and finds that consistent r e s u l t s are obtainable using t h i s process. J. Gray has obtained accurate r e s u l t s f or low grade ores with sodium hydroxide whereas potassium cyanide i s e n t i r e l y unreliable. 9 G. M. Henderson used sodium peroxide instead of the hydroxide and found that the speed of the analysis was 13 greatly increased. The peroxide was also discussed by J. Gray who claims i t s advantage over the hydroxide i s only i n special cases, as when chromite i s present. 4 EXTRACTION OF TIN FROM ITS ORES. 1. Occurrence of Samples. Tin occurs i n Nature i n three forms: (1) as Cas s i t e r i t e or stannic oxide, (2) as stannite, the combined sulphides of Iron, Copper and Tin and (S) as native Tin. By f a r the most abundant source of t h i s element i s C a s s i t e r i t e and hence t h i s mineral has been studied i n some d e t a i l . Stannite occurs usually i n much smaller quantities and often i n small proportions along with the oxide. I t i s very seldom used as a souree of Tin not only because of i t s r a r i t y but also because of i t s low t i n content. Hence i t has not been studied i n the same d e t a i l . Hence i n eff e c t i n g the disintegration of the mineral, the methods already i n use i n connection with C a s s i t e r i t e were t r i e d , and s i m i l a r ones i n the case of the Stannite. The sample of C a s s i t e r i t e referred to was obtained from the Consolidated Mining and Smelting Company, from t h e i r S u l l i v a n mine, at Kimberley, B. C. It had bein concentrated from the i r t a i l i n g s and was a f a i r l y r i c h ore. A sample of Stannite was obtained from the Snowflake mine, situated twenty miles north of Revelstoke. I t was taken from a quartz vein and was of low quality. These two ores seemed to f u l f i l l the conditions l a i d down above as to difference i n conditions of formation. 5 2« The Disintegration of Oasaiterite Sample. The methods of tre a t i n g C a s s i t e r i t e are varied and have 17 13 been aiseussed by H. Milon and R. Fouret, and J. Gray. The t i n oxide may be reduced by carbon, hydrogen, i l l u m i n a t i n g gas, zinc or potassium cyanide. Of.these carbon and cyanide were t r i e d . The f i n e l y ground ore was wel l mixed with powdered charcoal and placed i n a porcelain crucible w e l l l i n e d w ith the reducing agent. The dross was heated t i l l redness and kept at t h i s temperature for ten minutes. On cooling the reduced t i n was found to exist i n very small p a r t i c l e s which had to be separated by s l u i c i n g . The cyanide method was next t r i e d and since the result s obtained i n t h i s ease were most satisfactory, the pursuit of a better method was discontinued. A 25 gram sample was w e l l mixed with an equal quantity of Potassium Cyanide and placed i n a 3 inch porcelain evaporating dish. On both top and bottom of the mix was placed a layer of pure cyanide. The crucible was heated, slowly at f i r s t and then more rapi d l y , u n t i l the whole mass had fused. The temperature was f i n a l l y raised to a bright red heat and then the crucible and i t s contents were allowed to cool gradually. The whole operation required about t h i r t y minutes. The melt was then leached out with water and the metallic t i n button found on the bottom of the c r u c i b l e . I t has been noted above that the Pearce Low method of fusion with Sodium Hydroxide gives the most accurate r e s u l t s , e specially with low grade ores, and seems to be the best method thus f a r suggested. While the cyanide fusion i s not an accurate a n a l y t i c a l method, i t i s r e l a t i v e l y most effective i n the separation of the t i n from other constituents and for t h i s reason was adopted. 3. The Disintegration of Stannite. Very l i t t l e study has been made of the disintegration of stannite ores. Accordingly those methods which are successful i n fusing C a s s i t e r i t e , were also t r i e d with stannite with v a r i e d r e s u l t s . A small sample of f i n e l y ground ore was boiled with aqua regia f o r sometime but was not attacked. Potassium n i t r a t e i s a powerful oxidizing agent and i s often used successfully for t h i s purpose. In order to test out the method, f i f t e e n grams of ore and an equal quantity of the n i t r a t e were w e l l mixed and placed i n a f i r e clay crucible of two hundred c c . capacity. An upper and lower layer of pure potassium n i t r a t e was used and the crucible heated gently i n a muffle furnace u n t i l fusion. The temper-ature was then raised and maintained at red heat for ten minutes longer. On eooling the charge was found to have changed somewhat i n color and appearance, and had become extremely hard and b r i t t l e . It was found to be very d i f f i c u l t to remove by mechanical means, i . e . , with hammer and c h i s e l , and could not be dissolved by any of the ordinary acids or aqua regia. Therefore t h i s attempt was considered unsuccessful. 7 The Cyanide fusion was used i n a manner similar to that outlined above. When the cyanide was leached out with water i n the customary manner, the f i n e l y divided ore came out as w e l l , s t i l l i n the form of a fine black powder. The ore had not been attacked. The duration of the heating period after fusing did not seem to a s s i s t the process of reduction. It has been already noted that the sodium hydroxide fusion i s the standard method of disintegrating and analysing C a s s i t e r i t e . About t h i r t y grams of broken Sodium Hydroxide was mixed with f i v e grams of f i n e l y divided ore and the mixture fused for h a l f an hour i n a porcelain erucible. The melt was f i r s t softened with water and then with HC1. The ore did not dissolve even on b o i l i n g . 18 J.H. Walton and H.A. Scholz found that many refractory substances could be e a s i l y decomposed by fusion with Sodium Peroxide. Their experiments included a Titanium Tin ore. This method as applied to C a s s i t e r i t e has been highly 13 recommended by Gray. In general, even where the sodium hydroxide fusion i s successful, the peroxide has been found to produce a surer, safer, and more complete fusion, for not only i s a better mix possible, but the combined effect of an oxidizing and s a l i f i c a t i o n agent i s obtained. Accordingly twenty-five grams of ore were well mixed with an equal quantity of Sodium Peroxide and the mixture transferred to a f i r e clay crucible. I g n i t i o n of the mix was obtained by heating gently i n a muffle furnace for three 8 or four minutes. The molten mass was allowed to cool somewhat and was then poured c a r e f u l l y into about s i x l i t r e s of water. Violent e b u l l i t i o n took plaee at f i r s t , but soon subsided. The remainder of the melt was e a s i l y chipped from the crucible and added to the solution. This mixture s t i l l appeared p a r t i a l l y undecomposed but on a c i d i f y i n g with HC1 and warming, complete solution was obtained, y i e l d i n g a clear yellow l i q u i d . To remove s i l i e a t h i s solution was evaporated u n t i l the large Jfcl which precipitated, became quite viscous. The mixture was then d i l u t e d , and the s i l i c a separated by decantation. The s i l i e a was then further extracted with d i l u t e Hydrochloric acid u n t i l the solution obtained was no longer yellow. This method of fusion has proven to be quite satisfactory fo r the d i s i n t e g r a t i o n of stannite ore. I t i s exceedingly rapid and requires only a r e l a t i v e l y small amount of fusing material. Since the material i s to be l a t e r p u r i f i e d , another important feature i s to be noted. No foreign material which would be d i f f i c u l t to remove l a t e r on, i s introduced. 4. P u r i f i c a t i o n of C a s s i t e r i t e Tin. The metal i s obtained from the cyanide fusion i n the form of a shiny button. A l l the cyanide i s f i r s t carefully dissolved out with water and the metal then dissolved by warming i n HC1. The t i n i s precipitated from t h i s solution as the sulphide by passing i n hydrogen sulphide gas, i r o n and other group (3) metals remaining i n solution. The 9 pr e c i p i t a t e was f i l t e r e d , washed, and dissolved by warming with ammonium polysulphide. When copper i s present i n large quantities, i t i s dissolved to a small extent by t h i s reagent, but there was no i n d i c a t i o n of i t s presence i n the sulphide as i t appeared , bright yellow and perfectly homogeneous. The ammonium sulpho stannate was then f i l t e r e d and the stannic sulphide reprecipitated by a c i d i f y i n g the f i l t r a t e with d i l u t e hydrochloric acid. The large yellow precipitate obtained was washed free of acid and then treated i n a beaker with eonc. N i t r i c Acid. Small pieces of yellow amorphous sulphur were l e f t undissolved even on warming. These were f i l t e r e d o f f , washed and rejected, the wash water being added to the main solution. The solution of the t i n i n n i t r i c acid was then evaporated almost to dryness, a clear transparent j e l being obtained. This j e l gave a fine white flocculent precipitate of metastannio acid when agitated with a large excess of water. Dilute n i t r i e acid was added to increase the s o l u b i l i t y of any impurities and the pre c i p i t a t e most conveniently f i l t e r e d by decantation. Antimony forms a corresponding pentoxide which however i s insoluble only a f t e r i g n i t i o n . Hence t h i s procedure should remove the l a s t possible traces of any impurities. To obtain the t i n i n a metallic form, the precipitate (HoSn0„) was dissolved i n concentrated hydrochloric acid, 2 3 di l u t e d , and electrolysed. As Platinum i s dissolved somewhat by nascent chlorine, i t was necessary to substitute some other r e s i s t i v e material which i s not attacked, for an anode. Tantalum metal i s not attacked, by aqua regia or chlorine, hence would be quite useful for the purpose. A th i n s t r i p , one mm. i n width and ten cm. long was obtained, and was substituted for the anode, and a fine platinum rod used as cathode. The tantalum however, on imposing an E.M.F., was found to take on an iridescent blue f i l m of the oxide, which was quite impervious to the passage of the current. I t i s remarkable that the current was found to pass i n the opposite d i r e c t i o n with the tantalum serving as a cathode. A spectroscopically pure carbon electrode, f i f t e e n cm. long and one h a l f cm. i n diameter was obtained from Adam Hilger Ltd., and substituted for the anode. I t was found to be very s a t i s f a c t o r y and r e s i s t e d the action of the chlorine quite w e l l . The platinum rod was again used as cathode. The t i n deposited i n shining l e a f l e t s which were i o l l e c t e d d a i l y and stored i n d i s t i l l e d water. To obtain i n a more compact form the cry s t a l s were placed i n a porcelain boat and fused i n a current of pure dry hydrogen. A shiny m e t a l l i c button was obtained. This completed the p u r i f i c a t i o n . The hydrogen f o r t h i s purpose was prepared by the action of zine on sulphuric acid and was p u r i f i e d by passing through a series of wash bottles containing concentrated a l k a l i , and over hot copper gauze. 11 5. The P u r i f i c a t i o n of Stannite Tin. The percentage of t i n i n t h i s ore i s very much lower than i n the C a s s i t e r i t e . Hence the procedure had to be modified considerably, as certain separations used i n the preparation of pure t i n from C a s s i t e r i t e can not be used where there i s only a small quantity of t i n , and much larger quantities of other elements. From the hydrochloric acid solution of the ore, the t i n was p r e c i p i t a t e d as the sulphide along with other group (8) metals, the most serious of these being antimony and copper. This p r e c i p i t a t e was f i l t e r e d o f f , washed, and then warmed with ammonium polysulphide. Instead of leaving the copper behind as expected, quite a large quantity of i t passed into s o l u t i o n . This was indicated by the dark color of the sulphide p r e c i p i t a t e obtained by a c i d i f y i n g the solution of ammonium sulpho-stannate with d i l u t e hydrochloric acid. The process of d i s s o l v i n g i n ammonium polysulphide and p r e c i p i t a t i n g with hydrochloric acid was repeated, but there was s t i l l considerable copper present. Consequently the precipitate was dissolved by warming i n d i l u t e n i t r i c a c id, the sulphur residue f i l t e r e d o f f , and the solution evaporated to dryness. On d i l u t i n g and agi t a t i n g the residue with water, a white floceulent precipitate of metastannie acid appeared. By evaporating to dryness, i t was found that some of the antimony was rendered insoluble by being converted to the pentoxide. The evaporation had to be carried that f a r because of the small quantity of t i n j I E present. However, by taking up the residue i n a i l u t e n i t r i c a cid, the copper and other metals form soluble n i t r a t e s , while the t i n and antimony remain quite insoluble. The antimony was then separated by taking advantage of the fact that t i n can not be precipitated i n a solution of a certain acid concentration i n which the antimony i s r e a d i l y precipitated. The insoluble residue obtained on evaporation with n i t r i c a cid was dissolved i n concentrated hydrochloric a d d , and the solution d i l u t e d with f i v e times i t s volume of water. The antimony was p r e c i p i t a t e d by saturating the hot solution with hydrogen sulphide. The t i n was then obtained as yellow st stannic sulphide by almost n e u t r a l i z i n g the f i l t r a t e with ammonium hydroxide, d i l u t i n g , ana saturating the cold solution with hydrogen sulphide. II. THE STABILITY OF STANNIC OXIDE. The atomic weight of t i n has been determined by several 19 Investigators by means of the r a t i o Sn: SnOgjwithout exception the values obtained by them have been too low, due to absorption of a i r on so large a surface. In order to determine the accuracy of the method as a means of estimating t i n , the temperature of drying and s t a b i l i t y of the oxide to heat treatment was investigated. A small button of C P . t i n , weighing approximately one gram was f i l e d to remove surface oxide and etched with d i l u t e hydrochloric acid. I t was then washed with d i s t i l l e d water, dried, and c a r e f u l l y weighed. The metal was converted to the oxide "by warming i n cone, n i t r i c acid, the excess of acid being c a r e f u l l y evaporated when the metal no longer remained. The residue when dried at different temperatures gave the following r e s u l t s . Original weight of Sn button* 0.9923 grams. After drying i n the oven for two hours, at 150 degrees. After drying i n the oven f o r twenty-four hours at 150 degrees. After heating with a blast lamp for h a l f an hour After heating with a blast lamp f o r another h a l f an hour. After heating with a blas t lamp for another half an hour. Weight of Sn0 2 1.2743 1.2643 1.2632 1.2623 1.2619 Atomic Weight Calculated. 11.6 116.73 117.21 117.60 117.78 I t can thus be seen from these measurements that prolonged heating tends to decrease the amount of a i r absorbed, but i t i s doubtful whether a perfect removal of the absorbed gases can be attained. I I I . THE DETERMINATION OF TIN AS CAESIUM CHLORO STANNATE. Tin forms several complex chlorides with the a l k a l i metals, of which the caesium chloro stannate i s insoluble. As the o r i g i n a l object of the investigation was to determine the atomic weight of t i n i n i t s different ores, the p o s s i b i l i t y of using the formation of t h i s s a l t as a method was considered. The method was investigated i n the following manner. Preparation of Stannic Chloride. A small button of t i n corresponding to about one gram was f i l e d u n t i l a l l the surface layer had been removed. I t was then etched i n d i l u t e hydrochloric acid and washed i n d i s t i l l e d water. After drying and dessieating, the button was accurately weighed. I t was then dissolved i n a covered beaker i n concentrated hydrochloric acid with gentle warming, and the stannous chloride oxidized by passing i n chlorine, the temperature of the l i q u i d being kept as low as possible by running water. Previous tests have shown that a corresp-onding quantity of t i n can be oxidized i n t h i s manner i n h a l f an hour, so i n the present case one hour was considered as s u f f i c i e n t . Preparation of Caesium Chloride. The caesium used f o r the purpose was obtained from Dr. E.H. Archibald i n the form of caesium sulphate. It had o r i g i n a l l y been used for the determination of the atomie weight of caesium. To convert t h i s material to caesium chloride the calculated weight of r e c r y s t a l l l z e d barium chloride necessary was added to i t . The p r e c i p i t a t i o n was carried out i n a t o t a l volume of two l i t r e s and both solutions were kept hot during the mixing. After s t i r r i n g and digesting the p r e c i p i t a t e f or sometime, i t was f i l t e r e d off and the f i l t r a t e of caesium chloride evaporated to a small volume. To remove possible traces of sulphate, the caesium chloride was reprecipitated from the solution by saturating i t with HC1 gas. The c r y s t a l s , after washing and drying were ready for use. Method. A s l i g h t excess of the calculated amount of caesium chloride was weighed up and dissolved i n twenty-five c c . of water and added a l i t t l e at a time to the stannic chloride. The caesium chloro-stannate came down as a fine white p r e c i p i t a t e . It was f i l t e r e d off and washed with pure water i n a tared Sooch crucible, which had been prepared with an asbestos mat. The crucible and i t s contents were then dried i n an oven f o r several hours at 150°G, and weighed. The f i r s t r e s u l t s calculated from the r a t i o Sn:0s oSnGl f l 2 o gave a value f o r the atomic weight of t i n whieh was much too high, i n d i c a t i n g that the weight of the precipitate was correspondingly low. Moreover, on evaporating the f i l t r a t e and wash water a number of c r y s t a l s s e t t l e d out. As the f i l t r a t e should contain nothing but a s l i g h t concentration of caesium chloride, i t was concluded that the low r e s u l t s could be attributed to the s o l u b i l i t y of caesium ehlorostannate. Two more determinations were made taking greater precaution to decrease an error due to s o l u b i l i t y and more accurate values were obtained. This j u s t i f i e s the assumption that the error i s due at least i n part to s o l u b i l i t y . As very l i t t l e work has been done on t h i s compound i t was considered necessary to obtain more d e f i n i t e information on i t s s o l u b i l i t y i n various concentrations of hydrochloric acid and caesium chloride. Such information would enable us to state more d e f i n i t e l y whether t h i s method i s suitable or not f o r atomic weight determinations, and i f so, the most suitable concentration f o r p r e c i p i t a t i o n . Accordingly, t h i s has constituted another phase of the investigation. IV. THE DETERMINATION OF THE SOLUBILITY OF CAESIUM CHLORO ST ANNATE. Apparatus. In order to agitate the solutions at a constant temperature u n t i l the equilibrium of saturation was obtained, a granite k e t t l e of f o r t y l i t r e s capacity was nearly f i l l e d with water and covered with a thick layer of f e l t to prevent ra d i a t i o n . The water was maintained at constant temperature by a tungsten lamp which was controlled by a thermostat and o o relay. The temperature remained within .05 of SO with t h i s apparatus. Pyrex test tubes of twenty-five c c . capacity were used and rubber stoppers which had been previously boiled for several hours with ammonium hydroxide. Method of Determination. To the tubes were added twenty-five c c . of solution, and somewhat more of the caesium chlorostannate than would dissolve i n that p a r t i c u l a r concentration. The sal t was prepared as i n the determinations above. The rubber stoppers were then wired i n and the tubes attached to the s t i r r i n g apparatus and rotated for twenty-four hours. In order to determine how much of the s a l t had dissolved, the test tube was placed i n the bath so that the top just stood above the water. After the excess s o l i d had settled to the bottom, the l i q u i d was drawn up into the calibrated pipette and two ten c.e. portions withdrawn. The t i p of the pipette was prepared by tying a f i l t e r paper around i t with a small piece of s i l k thread. In t h i s manner the p o s s i b i l i t y of s o l i d p a r t i c l e s being drawn off was eliminated. The twenty c.e. portions of l i q u i d were c a r e f u l l y evaporated i n small porcelain crucibles, and the amount dissolved taken as the weight of the residue obtained on evaporation. The r e s u l t s as f a r as they have been carried at the present time are given i n the following table. BIBLIOGRAPHY 1. F. W. Aston, 2. Levol, 3. Moissenet, 4. Beringer, 5. Hampe, 6. Parry, 7. S. Fawns, 8. Bayerlein-Essen, 9. G.M. Henderson, 10. C. Boy, 11. T. GolieJc, 12. C o r t l , 13. J. Gray, 14. E. 7. Pearee, 15. Low, 16. R.J. Morgan, 17. H. Milou and R. Fouret. 18. J. H. Walton and H.A. Scholz. 19. Briscoe, Uature, 109, 813, (1922). Annales de ehimie et de physique (3, t XLIX), page 87. Comptes rendus t LI , page 205. Text Book of Analysis, page 278. Chemixer Zeitung 1887, 11-19. The Assay of Tin and Antimony, p.34. Tin Deposits of the World, page 215. Z Agnew Chem. 23, 969. Eng. Min. Journal, 103, 267, 1917. Meta l l m Erz 20 210-7 (1923). Eng. Min. Journal, 102, 827, 1916. Anal Soe. quim Argentina 9, 44-53 (1921). Chem. Ab. 15, 3801. J. Chem. Met. S o c , S. A f r i c a , 10, 402-3. Tin Deposits of the World, page 207. Technical Methods of Ore Analysis, page 246. Chem. Eng. 14 289-91. Orig. Com. 8th Intern. Cong. Appl. Chem., 1, 373. Chem. Abs. 2, 2530. Journal Chem. Soc. 107, 63, 1915.