EXCHANGE 
 
UNIVERSITY OF PENNSYLVANIA 
 
 I. TUNGSTEN HEXABROMIDE, 
 II. TUNGSTEN COMPLEXES. 
 
 A THESIS 
 
 PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE 
 
 UNIVERSITY OF PENNSYLVANIA IN PARTIAL FULFILLMENT 
 
 OF THE REQUIREMENTS FOR THE DEGREE OF 
 
 DOCTOR OF PHILOSOPHY 
 
 BY 
 
 HARRY BECKERMAN 
 
 PHILADELPHIA, PA. 
 1918 
 
UNIVERSITY OF PENNSYLVANIA 
 
 I. TUNGSTEN HEXABROMIDE. 
 II. TUNGSTEN COMPLEXES. 
 
 A THESIS 
 
 PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE 
 
 UNIVERSITY OF PENNSYLVANIA IN PARTIAL FULFILLMENT 
 
 OF THE REQUIREMENTS FOR THE DEGREE OF 
 
 DOCTOR OF PHILOSOPHY 
 
 BY 
 
 HARRY BECKERMAN 
 
 PHILADELPHIA, PA. 
 1918 
 

 ACKNOWLEDGMENT. 
 
 The author desires to express his sincere thanks to DR. 
 EDGAR F. SMITH, under whose direction and kindly encourage- 
 ment this work was carried out. 
 
I. TUNGSTEN HEXABROMIDE. 
 
 INTRODUCTION. 
 
 In 1851, v. Borch 1 passed bromine over highly heated tung- 
 sten and obtained tungsten pentabromide. In 1857, Riche*, 2 
 employing the same reaction, obtained what he believed to be 
 tungsten hexabromide, submitting, however, no analyses to 
 support his belief. In 1861, Bloomstrand 3 repeated the work 
 of Riche and showed that the conditions outlined by the latter 
 yielded pentabromide and not hexabromide. In 1872, Roscoe 4 , 
 in an endeavor to prepare tungsten hexabromide, acted upon 
 metallic tungsten with bromine at a red heat, in an atmosphere 
 of carbon dioxide, and obtained tungsten pentabromide. Tung- 
 sten hexabromide was first prepared in this laboratory, in 1896, 
 by Schaffef and Smith, 5 by allowing bromine to act on tungsten 
 metal at a gentle heat, in an atmosphere of nitrogen. These 
 investigators attributed their success in the preparation, to the 
 low temperatures at which they operated, the high temperatures, 
 previously employed, tending to prevent the formation of hexa- 
 bromide. They also replaced the carbon dioxide by nitrogen, 
 fearing that the former would be deoxidized in the reaction with 
 the formation of oxybromides. The amount of tungsten hexa- 
 bromide obtained, however, was small and the two complete 
 analyses of the product were not very concordant. As a conse- 
 quence, a confirmation of this work seemed highly desirable. 
 The purpose of the present investigation was to confirm the 
 existence of tungsten hexabromide as well as to determine more 
 exactly the conditions under which it is formed. 
 
 In 1916, Hill, 6 in this laboratory, noticed that the action of 
 chlorine on tungsten metal to form tungsten hexachloride is 
 
 'Oefvers. af. k. Vetensk. Akad. Forh. 1851, 150. Handbuch der 
 anorganischen Chemie, 1912 (Gmelin-Kraut) 111, I, 762. 
 
 2 Ann. Chim. Phys. (3) 50 (1857) 24. 
 
 3 Jour. pr. ch. 82, 429. 
 
 4 Ann. der. chem. 162, 362. 
 
 5 Am. Ch. Jour. 18 (1896), 1098. 
 
 6 Jour. Am. Chem. Soc. 38 (1916), 2383. 
 
 (3) 
 

 catalyzed to a remarkable degree by the presence of a small 
 amount of platinum black. The temperature of formation of 
 tungsten hexachloride was reduced and the speed of reaction 
 was increased. The idea presented itself that this method applied 
 to the preparation of tungsten hexabromide might so far facilitate 
 its formation that a good yield of this body might be obtained. 
 
 EXPERIMENTAL PART. 
 
 MATERIALS. 
 
 The best results were obtained with a finely divided tung- 
 sten, formed by heating tungsten trioxide in hydrogen for a long 
 time, at as low a temperature as possible, care being taken that 
 the metal was entirely reduced. The tungsten formed in this 
 way is black in color and is more finely divided than the grey 
 variety. The metal was not permitted to remain in contact 
 with the air because it absorbs both oxygen and moisture. It 
 was preserved in dry hydrogen. 
 
 The bromine used was prepared by treating ordinary bromine 
 with potassium bromide and manganese dioxide, to remove 
 chlorine and organic matter. This mixture was allowed to stand 
 over night, and the bromine was then distilled out. It was 
 washed three times with distilled water and then treated with 
 fused CaCl2 over night and distilled again. The treatment 
 with fused calcium chloride and the distillation were repeated 
 twice. The distillation of the bromine took place entirely in 
 glass. The bromine was preserved in well stoppered bottles 
 in a desiccator over concentrated sulphuric acid. 
 
 Since platinum black absorbs oxygen from the air, it was 
 mixed with the tungsten trioxide before the latter was reduced in 
 hydrogen. The tungsten formed was thus intimately mixed 
 with the platinum and was entirely free of oxygen. 
 
 PROCEDURE. 
 
 The preparation of tungsten hexabromide is very difficult. 
 Unless the greatest possible attention is paid to every minute 
 detail, no hope for success can be entertained. The greatest 
 
difficulty experienced was the elimination of small traces of 
 oxygen which leaked into the apparatus, in spite of precautions 
 taken to prevent it. This trace of oxygen was removed by plac- 
 ing in the tube A, before the boat E, a long layer of sugar carbon. 
 This was heated to redness. The carbon dioxide formed, being 
 diluted with a large quantity of nitrogen, passed on without 
 entering the reaction. Sugar carbon made by heating pure 
 cane sugar contains oxygenated organic bodies, even when the 
 carbon is heated to redness in a blast lamp. These oxygenated 
 organic bodies were removed by heating the carbon to a high 
 temperature in bromine vapor until moisture and liquid brominated 
 bodies were no longer given off. The sugar carbon was then 
 heated in dry hydrogen and preserved in that gas. 
 
 The character of the tungsten used is vital to the success 
 of the preparation. Black tungsten proved successful, the grey 
 variety giving mostly tungsten pentabromide mixed with a 
 little tungsten hexabromide. 
 
 The air was displaced from the apparatus by means of a 
 stream of nitrogen which was made as follows. Air was forced 
 by means of a water blast through potassium hydroxide solution, 
 to remove carbon dioxide, then through two wash bottles con- 
 taining concentrated ammonia water. The air took up ammonia 
 gas and then passed into a long Jena tube containing red-hot 
 copper. The hydrogen of the ammonia combined with the 
 oxygen of the air to form water. The latter was condensed in an 
 empty cooled flask. The nitrogen from the air and that from 
 the ammonia passed into another Jena tube containing a red- 
 hot, reduced copper spiral, to remove the last traces of oxygen. 
 The gas was then washed through two bottles of dilute sulphuric 
 acid to remove traces of ammonia which might still be retained 
 by the nitrogen. The rest of the apparatus is as pictured in 
 Fig. 1 . Moist chromous acetate mixed with glass beads removed 
 the last traces of oxygen. From here on, the apparatus was 
 made entirely of glass, no rubber connections or stoppers being 
 used. 
 
 The nitrogen was thoroughly dried by passage through 
 boiled sulphuric acid, solid potassium hydroxide and phosphorus 
 pentoxide. The gas then passed through the U-tubes, F and G, 
 and then into the tube A. After the tube A was placed an empty 
 
i 
 
 0* 
 CD 
 
 CQ 
 
 D 
 
7 
 
 tower to condense the excess of bromine, a tower with fused 
 CaCl2 and then a vessel containing soda lime to prevent the 
 bromine vapor from reaching the air. 
 
 A porcelain boat containing the tungsten, mixed with a 
 trace of platinum, was removed from the tube in which it was 
 reduced in hydrogen. It was quickly inserted into the tube A 
 through the ground-glass joint B. A plug of dried glass wool 
 was inserted and then the tube was filled from C to D with sugar 
 carbon. The ground glass connection B was shut and tightly 
 clamped by means of rubber bands. It was coated with paraffine 
 on the outside. The current of dry nitrogen was continued for 
 eight hours more to remove the air and thoroughly dry the 
 apparatus. 
 
 The sugar carbon was heated to redness. Bromine was 
 then allowed to drop from the separatory funnel H into the 
 U-tube F, from which it passed into G. G contained glass wool 
 moistened with boiled sulphuric acid to remove traces of moisture 
 which might be introduced by the bromine. F and G were 
 immersed in a beaker of hot water. When an excess of bromine 
 had reached the porcelain boat, E, the latter was gently heated. 
 An excess of bromine was maintained at the moment of formation, 
 otherwise tungsten pentabromide would have been formed. 
 The latter does not combine with more bromine and stays as 
 tungsten pentabromide. The temperature of the tube at E 
 was ascertained by placing a thermometer in contact with it. 
 The reaction started at 200 C. At 230 C. it was fairly rapid. 
 The tungsten hexabromide rose from the boat in the form of a 
 heavy reddish brown vapor which condensed in the colder part 
 of the tube to a reddish brown liquid. The latter on cooling 
 solidified with a crackling sound, forming a crystalline solid. 
 The first portion was driven beyond the point at which it was 
 intended to collect the main part of the preparation. The 
 remainder was fractionally redistilled in an excess of bromine 
 vapor. A non-volatile black residue was left behind. The 
 middle fraction was collected between two constrictions. After 
 cooling, the excess bromine was removed by means of the stream 
 of nitrogen. The bulb containing the preparation was then 
 sealed off. 
 
8 
 
 Tungsten hexabromide is a blue-black crystalline solid 
 which can, with care, be sublimed in bromine vapor in the form 
 of blue-black shining needles. It appears to be soluble in liquid 
 bromine. It fumes in the air. Water decomposes it giving a 
 deep blue oxide. Ammonium hydroxide dissolves it completely. 
 Concentrated sulphuric acid and concentrated nitric acid form 
 tungstic oxide. On heating it evolves bromine. 
 
 METHOD OF ANALYSIS. 
 
 TUNGSTEN CONTENT. 
 
 A porcelain crucible containing a weighed portion of the 
 preparation was placed in a covered beaker containing a little 
 warm water. The water vapor acted rapidly on the preparation, 
 decomposing it, without causing it to spatter. The mass liquified. 
 The crucible was then placed on a stove plate and the contents 
 carefully evaporated to dryness. Dilute nitric acid was added 
 and again evaporated to dryness. Moderately concentrated nitric 
 acid was added; then finally concentrated nitric acid was added 
 repeatedly, evaporating to dryness each time. The tungsten 
 trioxide so formed was ignited and weighed. 
 
 BROMINE CONTENT. 
 
 A weighed sample was placed in a Bunsen apparatus. Nitric 
 acid was added by means of a separatory funnel fused into the 
 delivery tube of the flask. The mixture was distilled into silver 
 nitrate solution. The distillate, containing silver bromide, 
 silver bromate and excess of silver nitrate, was transferred to a 
 porcelain dish and evaporated to dryness. The residue was 
 ignited to change the silver bromate to silver bromide. The 
 dish was allowed to cool. Nitric acid was added and then hot 
 distilled water containing a little nitric acid. The silver bromide 
 was filtered and washed. It was weighed as silver bromide. 
 The results are given in Table I. 
 
TABLE I. 
 
 Preparation No. 9. 
 
 Wt. of Sample. 
 
 Wt. of W0 3 . 
 
 %w. 
 
 0.1023 gr. 
 
 0.0366gr. WO 3 
 
 28.35 W 
 
 0.4126 " 
 
 0.1492 " " 
 
 28.67 " 
 
 0.2156 " 
 
 0.0774 " " 
 
 28.43 " 
 
 0.4249 " 
 
 0.1528 " " 
 
 28.53 " 
 
 0.5236 
 
 0.8834 " AgBr 
 
 Preparation No. 12. 
 
 Wt. of Sample. Wt. of WO 3 . 
 
 0. 1575 gr. 0.0554gr. WO 3 
 
 0.8152 ' 
 
 0.0698 " 
 
 0.6354 " 
 
 0.4582 " 
 
 0.2882 " 
 
 0.0243 " " 
 
 /0. 2243 " " 
 
 { 1.0735 " AgBr 
 
 0.1623 " WO 3 
 
 0.7730 " AgBr 
 
 27.87 
 28.04 
 27.65 
 
 28.00 
 
 28.09 
 
 Theoretical Percentages. 
 
 %Br. 
 
 71.79 Br 
 
 %Br. 
 
 WBr 5 . 
 31.52 W 
 68.48 Br 
 
 WBr. 
 27.72 W 
 72.28 Br 
 
 71.89 
 
 71.80 
 
 THE ACTION OF CHLORINE AND BROMINE ON 
 MOLYBDENUM MIXED WITH PLATINUM BLACK. 
 
 The success attained with the preparation of tungsten 
 hexabromide raised the hope that platinum black would cata- 
 lyze the reaction between chlorine and molybdenum. Chlorine 
 acts on molybdenum metal at a dull red heat to form molyb- 
 denum pentachloride. It was thought that the temperature of 
 formation would be reduced so that molybdenum hexachloride 
 might be formed. The latter has never been obtained. 
 
 Finely divided molybdenum was mixed with platinum 
 black and was subjected to the action of chlorine in an atmosphere 
 of nitrogen. The action started at 40 C. Unless the chlorine 
 was diluted with a large quantity of nitrogen, the reaction became 
 so violent that the metal caught fire in the chlorine. The action 
 was allowed to proceed slowly at 50 to 60 C. Large, black, 
 highly lustrous crystals formed in the boat. These were slowly 
 sublimed out, collected between two constrictions and sealed 
 up in glass. They were analyzed by the method described under 
 
id 
 
 tungsten hexabromide. Care was taken to prevent volatilization 
 of the molybdenum trioxide, by careful ignition. The results 
 are given in Table II. 
 
 TABLE II. 
 
 Theoretical Percentages. 
 Wt. of Sample. Wt. of MoOa. % Mo. % Cl. MoCU. MoCl. 
 
 0.2065gr. 0.1083gr. 34.96 35. 13 Mo 31. 09 Mo 
 
 0.1843 " 0.0976 " 35.33 64.87 Cl 68.91 Cl 
 
 0.5168 " 0.2734 " 35.26 
 
 0.1519 " 0.0799 " 35.02 
 
 fo.3875 " 35.32 
 
 \ 1.9445 " AgCl 65.75 
 
 Average Mo 35. 18 
 The compound, therefore, was molybdenum pentachloride. 
 
 Molybdenum pentabromide has never been made. Indeed, 
 molybdenum tetrabromide is difficult to prepare, it being formed 
 in small traces when bromine is passed over molybdenum, the 
 chief product being molybdenum tribromide. Attention was 
 next directed to the preparation of molybdenum pentabromide. 
 
 Molybdenum metal, mixed with platinum black, was heated 
 in Br vapor. At 130 to 140 C., the reaction proceeded rapidly. 
 Long, sharp, hairy needles formed above the boat. On resublim- 
 ing they decomposed into yellow Mo Br 2 . Therefore the boat 
 was removed from the tube, the crystals were quickly detached 
 and dropped into a weighing bottle. The substance was very 
 voluminous, a large bulk of the hairy crystals weighing very little. 
 
 TABLE III. 
 
 Theoretical Percentages. 
 
 Wt. of Sample. Wt. of MoOs. % Mo. MoBn. MoBrs. 
 
 0.0431gr. 0.0155gr. 23.89 23. 08 Mo 19. 36 Mo 
 
 The substance, therefore, was MoBr 4 . 
 
II. TUNGSTEN COMPLEXES. 
 
 INTRODUCTION. 
 
 Inorganic complexes play a more important role than is 
 generally believed to be the case. The difficulty which is fre- 
 quently met with, in the purification of substances can, often, 
 be directly attributed to the formation of complexes. For 
 example, Smith and Exner 1 in an exhaustive paper on the 
 atomic weight of tungsten show conclusively that the difficulty 
 of preparing pure tungstic acid is due to the ease with which 
 the latter substance forms complexes. They found that on 
 digesting tungstic acid, which was free of iron, with hydrochloric 
 acid or nitric acid, in which iron was present, the latter would 
 enter the tungstic acid. Iron, manganese, phosphorus and 
 vanadium were extremely difficult to remove. Indeed an 
 ammonium salt of a complex, containing the oxides of the above 
 elements, was isolated. Similarly, the slimy, greenish or bluish 
 white residue remaining when tungstic acid dissolved in ammonium 
 hydroxide, was found by them to contain ammonium chloride. 
 It was probably an ammonium chlorinated tungstic acid derivative. 
 
 Again, Rogers and Smith 2 have shown that the introduction 
 of only a few tenths of one per cent of various dioxides, like 
 TiO 2 and ZrO2, into ammonium vanadico-phospho-tungstate 
 changed its properties and reactions entirely. The authors 
 point out that in mineral analysis the influence of such minute 
 amounts of supposedly foreign constituents is too often dis- 
 regarded and not considered in the deduction of formulas. 
 
 As a consequence an intimate knowledge of complexes, 
 the conditions under which they are formed, their analyses, 
 are subjects of importance. The purpose of the present investi- 
 gation is to continue the study of complexes which has been 
 carried on in this laboratory for a number of years. 
 
 In 1895, Wolcott Gibbs, 3 the pioneer in this field of research, 
 
 1 Proc. Am. Phil. Soc. 43 (1904), 123. 
 
 2 Jour. Am. Chem. Soc. 25 (1903), 1223. 
 
 3 Am. Chem. Jour. 17 (1895), 173. 
 
 (11) 
 
12 
 
 prepared complexes containing molybdenum dioxide by boiling 
 hydrated MoO 2 with sodium para tungstate. From the deep 
 orange colored solution, potassium bromide gave a buff colored 
 precipitate which was recrystallized from water in small pale 
 brown scales. Gibbs ascribed the formula 5 K 2 O.MoO2. 12 WCV 
 16 H 2 O to the compound. It was thought that corresponding 
 complexes containing tungsten dioxide might exist. 
 
 EXPERIMENTAL. 
 
 The reduction of hydrated, precipitated tungsten trioxide 
 to hydrated tungsten dioxide by zinc and hydrochloric acid 
 was found to be a difficult task. After five days' action, the 
 reduction had not reached the tungsten dioxide stage. Several 
 known methods were tried and the best was found to be the 
 following: A large excess of tungsten trioxide was boiled up 
 with sodium hydroxide solution. A small test portion of the 
 solution was treated from time to time with dilute hydrochloric 
 acid until a precipitate of tungstic acid no longer formed. This 
 indicated the formation of a meta tungstate, Na 2 O.4WO 3 . 
 The solution was filtered. Zinc dust, added to this solution, 
 turned it intense indigo blue. Dilute hydrochloric acid was 
 then added. In a few minutes copper-red, hydrated tungsten 
 dioxide formed. The increased speed of the reduction was due 
 to the fact that the action took place in solution instead of in 
 the solid phase. The tungsten dioxide was filtered and washed 
 on the suction pump in an atmosphere of hydrogen. 
 
 I. Ammonium para tungstate, dissolved in water, was 
 boiled up with the hydrated tungsten dioxide. A deep blue 
 solution formed. The excess tungsten dioxide was filtered out 
 carefully. On standing the solution turned bright green and 
 then deep orange colored. The same changes in color took place 
 in an atmosphere of hydrogen, so that the change was not due to 
 oxidation. On heating, the orange colored solution changed 
 back to green and then to deep blue. Potassium iodide, added 
 in large excess, gave a buff colored precipitate. This was filtered 
 out. It was soluble in hot water, difficultly soluble in cold. 
 
13 
 
 It was recrystallized from water, coming down in beautiful light 
 brown crystals. Sodium bromide, added to the orange colored 
 solution, gave the light brown sodium salt. The similarity to 
 the MoO 2 compounds obtained by Gibbs is marked. Lead 
 acetate gave a heavy yellowish white crystalline precipitate. 
 Silver nitrate gave a red precipitate, somewhat soluble in hot 
 water giving a red solution. Mercurous nitrate gave a reddish 
 brown precipitate. Alcohol threw down the brown alkali salt. 
 Manganous sulphate gave a white precipitate. The cobalt salt 
 was rose colored, while a copper compound, precipitated by 
 copper sulphate, was greenish yellow. Ferric chloride gave no 
 precipitate. The barium salt was white. 
 
 ANALYSIS OF THE POTASSIUM SALT. 
 
 WO2 CONTENT. 
 
 A weighed portion was dissolved in water. Dilute sulphuric 
 acid was added and then an excess of standard potassium per- 
 manganate solution. The solution was warmed on the water 
 bath. It was then cooled and an excess of ferrous ammonium 
 sulphate, dissolved in water, was added. The excess ferrous 
 iron was titrated back with potassium permanganate solution. 
 The tungsten dioxide content was calculated. 
 
 AND KzO CONTENT. 
 
 The complex was found to be only slowly decomposed by 
 concentrated nitric acid or aqua regia. . The following method 
 proved satisfactory. A weighed sample was dissolved in hot 
 water. A few cubic centimeters of dilute sulphuric acid and a 
 little nitric acid were added. The solution was evaporated, in 
 an air bath, to strong fumes. After cooling, water was added 
 and the evaporation to fumes repeated. Nitric acid and water 
 were added and the mixture was digested for one hour. The 
 tungsten trioxide was filtered out and washed with hot water 
 containing nitric acid. It was then ignited and weighed. The 
 filtrate from the WOa was evaporated to small bulk, transfered 
 to a weighed crucible. The evaporation was continued, the sul- 
 phuric acid was driven off. The potassium sulphate was ignited 
 
14 
 
 and weighed. The water was obtained by difference. The 
 results are contained in Table 4. 
 
 TABLE IV. 
 
 Theoretical % calculated for 
 Wt. of Complex. Wt. of Constituent Found. % 5 KzO. WO 2 . 12 WOs. 13 HiO. 
 
 0.3072gm. 0.0184gr. WO 2 5.99WO 2 5.83WO 2 
 
 0.4664 " 0.0271 " " 5.81 " 
 
 0.3282 " 0.2465 " WO 3 75 . 1 1 WO 3 75.16WO 3 
 
 0.3236 " 0.2417 " " 74.69 " 
 
 0.3282 " 0.0762 " K 2 SO 4 12.55 K 2 O 12.69 K 2 O 
 
 0.3236 " 0.0768 " ' : 12.82 " 
 
 by difference 6.52 H 2 O 6.32H 2 O 
 
 The above analysis shows the complex to be 
 5 K 2 O. WO 2 . 12 WO 3 . 13 H 2 O. 
 
 The analysis of the sodium derivative is given in Table V. 
 TABLE V. 
 
 Theoretical % calculated for 
 Wt. of Complex. Wt. of Constituent Found. % 5 Na2O.WO z . 12 WOs.24 H 2 O. 
 
 0.2293gr. 0.0135gr. WO 2 5.89WO 2 5.77WO 2 
 
 0.5062 " 0.0287 gr. " 5.67 " 
 
 0.2646 " 0.1969 " WO 3 74.41 WO 3 74.40 WO 3 
 
 0.2035 " 0.1508 " " 74.10 " 
 
 0.2646 " 0.0508 " Na 2 SO 4 8.39 Na 2 O 8.29 Na 2 O 
 
 0.2035 " 0.0382 " " 8.21 " 
 
 by difference 11.67H 2 O 11.54H 2 O 
 
 The complex, therefore, is 
 
 5 Na 2 O.WO 2 .12 WOs-24 H 2 O. 
 
 II. In an attempt to repeat the preparation of the potassium 
 complex, the ammonium para tungstate solution was boiled 
 up with the hydrated tungsten dioxide for a much longer time 
 than in the first preparation. Potassium iodide, added to the 
 solution, gave a deep brown precipitate, instead of the buff 
 colored precipitate obtained previously. On recrystallizing from 
 water, the difficultly soluble light brown complex, obtained 
 before, crystallized out first. This was followed by dark brown 
 
15 
 
 crystals. The latter were much more soluble in water and could 
 easily be separated from the light brown crystals by several 
 recrystallizations. The mother liquor from the first preparation 
 of the light brown compound was deep brown in color and 
 probably contained this new body. The dark brown crystals 
 were analyzed as outlined before. 
 
 TABLE VI. 
 
 Theoretical % calculated for 
 Wt. of Complex. Wt. of Constituent Found. % 5 K 2 O.2 WO 2 . 12 WOs. 11 HjO. 
 
 0.8765gr. 0.0995gr. WO 2 11.35WO 2 11.12WO 2 
 
 0.4966 " 0.0552 " " 11.12 " 
 
 0.2111 " 0.1514 " WO 3 71.72 WO 3 71.68WO 3 
 
 a.5199 " 0.3716 " " 71.48 " 
 
 0.2111 " 0.0471 " K 2 SO 4 12.03 K 2 O 12.10K 2 O 
 
 0.5199 " 0.1164 " " 12.09 " 
 
 by difference 5.11 H 2 O 5.10 H 2 O 
 
 The formula of the dark brown crystals as deduced from the 
 above analysis, therefore, is 
 
 5 K 2 O.2 WO 2 .12 WOs.ll H 2 O. 
 
 Lead acetate added to a solution of the potassium salt gave 
 a heavy white precipitate. Mercurous nitrate gave a green 
 precipitate. The ferric salt was yellow. Silver nitrate gave a 
 chocolate colored precipitate. On boiling, the latter darkened. 
 This was soluble in ammonium hydroxide leaving metallic silver. 
 The cobalt salt was pink. Copper sulphate gave a greenish white 
 precipitate. 
 
 III. Sodium metatungstate, Na 2 O.4 WO 3 . 10 H 2 O, was dis- 
 solved in water. To this solution hydrated tungsten dioxide 
 was added and the solution was boiled for some time. The 
 excess tungsten dioxide was filtered from the deep blue solution. 
 On standing the solution turned green and then orange colored. 
 The compound would not crystallize out, on concentrating and 
 cooling the solution. It was isolated by adding alcohol, when 
 it settled as a brown oil. The latter was separated and warmed 
 gently in hydrogen to dry it. It soon solidified as a white crystal- 
 
16 
 
 line solid. It was soluble in water giving an orange colored 
 solution. The analysis of the sodium compound follows: 
 
 Wt. of Complex. 
 
 Wt. of Constituent Found. % 
 
 0.3654gr. 
 
 0.0141 g 
 
 r. WO 2 
 
 3.86 WO 2 
 
 0.3087 " 
 
 0.0122 ' 
 
 ( t 
 
 3.95 " 
 
 0.7645 " 
 
 0.0286 ' 
 
 < 
 
 3.74 " 
 
 0.2760 " 
 
 0.2297 ' 
 
 ' WO 3 
 
 83.22 W0 3 
 
 0.2472 " 
 
 0.2052 ' 
 
 
 
 83.01 " 
 
 0.4334 " 
 
 0.3615 ' 
 
 ' " 
 
 83.41 " 
 
 0.2760 " 
 
 0.0357 ' 
 
 ' Na 2 S0 4 
 
 5.65 Na 2 O 
 
 0.2472 " 
 
 0.0321 ' 
 
 
 
 5.67 " 
 
 0.4334 " 
 
 0.0543 ' 
 
 4 l 
 
 5.47 " 
 
 
 by 
 
 difference 
 
 7.38 H 2 O 
 
 TABLE VII. 
 
 Theoretical % calculated for 
 5 NazO.WOt.20 WO3.23 H 2 O. 
 
 3.85 WOj 
 83.15 WO, 
 5.55 Na 2 O 
 
 7.45 H,O 
 
 The analysis indicates the formula to be 
 
 5 Na 2 O.W0 2 .20 WO 3 -23 H 2 O. 
 
 Lead acetate formed a heavy white precipitate. Mercurous 
 nitrate gave a yellow- white precipitate, while that formed by 
 silver nitrate was purple colored, insoluble in hot or cold water. 
 Ferric chloride gave an orange colored precipitate. Potassium 
 iodide added to a cooled solution of the sodium salt, gave the 
 less soluble potassium salt of the complex in long colorless crystals. 
 
 IV. Pechard 1 has shown that when sulphur dioxide is passed 
 into a solution of ordinary ammonium molybdate, compounds 
 are formed which conform to the formula, 4(NH 4 )2O.3 SO 2 . 
 10MoO3-6H 2 0. It was thought that corresponding tungsten 
 complexes might exist. Accordingly ammonium paratungstate 
 was dissolved in water and sulphur dioxide was passed in. The 
 gas was rapidly absorbed. The solution was concentrated by 
 evaporating on the water bath, sulphur dioxide being passed 
 through the solution, during the evaporation, to keep it saturated 
 with the gas. On cooling in ice water, large colorless octahedrons 
 formed. Some of the crystals were fully three-quarters of an 
 inch across. They were extremely soluble in water. The com- 
 
 1 Compte Rendus 116, 1441 (1893). 
 
17 
 
 pound was recrystallized several times from water. Analysis 
 showed no sulphur dioxide to be present in the compound. It 
 proved to be ammonium metatungstate, (NH 4 )2O.4 WOa.8 H 2 O. 
 
 TABLE VIII. 
 
 Theoretical % calculated for 
 Wt. of Compound. Wt. of Constituent Found. % (NH4)zO.4 WOa.8 HjO. 
 
 0.3014gr. 0.0166gr. (NH 4 ) 2 O 5.52(NH 4 ) 2 O 4.63 (NH 4 ) 2 O 
 
 0.5752 " 0.0315 " " 5.47 " 
 
 0.3424 " 0.2858 " WO 3 83.47WO 3 82.56WO 3 
 
 0.4598 " 0.3836 " " 83.43 " 
 
 by difference 11.15 H 2 O 12.81 H 2 O 
 
 This reaction constitutes a rather quick and easy method 
 for the preparation of ammonium metatungstate. 
 
 V. Titanium hydroxide, made by adding redistilled 
 ammonium hydroxide to a solution of potassium titanium 
 fluoride, and zirconium hydroxide, made by adding ammonium 
 hydroxide to a solution of zirconium sulphate, were boiled up 
 separately with a solution of potassium paratungstate for several 
 hours. The excess titanium and zirconium hydroxides were 
 filtered out in each case and the complexes were crystallized. 
 They were analyzed by precipitating titanium and zirconium 
 hydroxides with ammonium hydroxide and weighing them as 
 dioxides. The tungsten trioxide and potassium sulphate were 
 formed as outlined before. 
 
 TABLE IX. 
 
 Theoretical % calculated for 
 Wt. of Complex. Wt. of Constituent Found. % 22 K 2 O.TiO 2 .57 WO 3 .39 HiO. 
 
 0.3746gr. 
 
 0.0019gr. TiO 2 
 
 0.51 TiO 2 
 
 0.50 TiO 2 
 
 0.4282 " 
 
 0.0021 " " 
 
 0.49 " 
 
 
 0.3746 " 
 
 0.3072 " WO 3 
 
 82.01 WO 3 
 
 82.27 WO 3 
 
 0.4282 " 
 
 0.3524 " " 
 
 82.30 " 
 
 
 0.3746 " 
 
 0.0901 " K 2 SO 4 
 
 13.00 K 2 O 
 
 12.87 K 2 O 
 
 0.4282 " 
 
 0.1016 " " 
 
 12.82 " 
 
 
 by difference 4.43 H 2 O 4.36 H 2 O 
 
 22 K 2 O.TiO 2 .57 WO 3 .39 H 2 O. 
 
18 
 TABLE X. 
 
 Theoretical % calculated for 
 
 Wt. of Complex. 
 
 Wt. of Constituent Found. 
 
 % 
 
 29K2O.Zr0 2 .70W0 3 
 
 . 73 H 2 O. 
 
 0.3446gr. 
 
 0.0019gr. ZrO 2 
 
 0.55 ZrO 2 
 
 0.60ZrO 2 
 
 
 0.3767 " 
 
 0.0024 " " 
 
 0.64 " 
 
 
 
 0.3446 " 
 
 0.2743 " WO 3 
 
 79.60 WO 3 
 
 79.60 WO 3 
 
 
 0.3767 " 
 
 0.2982 " " 
 
 79.17 " 
 
 
 
 0.3446 " 
 
 0.0834 " K 2 SO 4 
 
 13.09K 2 O 
 
 13.36K 2 O , 
 
 
 0.3767 " 
 
 0.0975 " " 
 
 13.99 " 
 
 
 
 by difference 6 . 48 H 2 O 6 . 44 H 2 O 
 
 29 K 2 O . ZrO 2 . 70 WO 3 . 73 H 2 O. 
 
 Had the zirconium dioxide content been only 0.13% higher, 
 then the formula of the zirconium dioxide complex would have 
 corresponded exactly to the formula of the titanium dioxide 
 complex. Both the ZrOz and the TiC>2 complexes were white 
 crystalline solids, soluble in hot water. 
 
 VI. A solution of potassium paratungstate was boiled up 
 with freshly precipitated lead dioxide. After filtering out the 
 excess of lead dioxide, the complex was crystallized. It is a 
 crystalline solid with a faint orange tint, soluble in water. The 
 compound was analyzed by adding to the solution of a weighed 
 portion, a few drops of alcohol and concentrated nitric acid. 
 This converted the lead dioxide to lead nitrate. The solution 
 was evaporated to dryness, the residue was baked and taken up 
 in 1 : 1 nitric acid. The evaporation and baking were repeated 
 twice. Nitric acid was added, then hot water containing nitric 
 acid. The tungsten trioxide was filtered out and weighed. The 
 filtrate containing lead nitrate and potassium nitrate was evapo- 
 rated to fumes with sulphuric acid. Water and alcohol were 
 added. The lead sulphate was filtered out and weighed. The 
 filtrate was evaporated to dryness and the potassium sulphate 
 was weighed. 
 
19 
 
 TABLE XI. 
 
 Theoretical % calculated for 
 
 Wt. of Complex. Wt. of Constituent Found. % 64 K 2 O.PbC>2 171 WO 3 . 1 19 HuO. 
 
 0.3341 gr. . 0022 gr. PbSO 4 . 5 1 Pb O 2 0.50 PbO 2 
 
 0.3294 " 0.0020 " " 0.49 " 
 
 0.3341 " 0.2763 " WO 3 82.70WO 3 82.53WO 3 
 
 0.3294 " 0.2712 " " 82.33 " 
 
 0.3341 " 0.0772 " K 2 SO 4 12.5lK 2 O 12.52K 2 O 
 
 0.3294 " 0.0767 " " 12.57 " 
 
 by difference 4 . 44 H 2 O 4 . 45 H 2 O 
 
 64 K 2 O.Pb0 2 . 171 W0 3 . 119 H 2 O. 
 
 It will be noticed that this formula is just three times the 
 formula of the titanium dioxide complex. 
 
 SUMMARY. 
 
 1. The existence of tungsten hexabromide is confirmed. 
 Exact details are given for its preparation in larger quantities. 
 
 2. Platinum black will catalyze the reaction between chlorine 
 and molybdenum and between bromine and molybdenum so 
 that the formation of molybdenum pentachloride and molyb- 
 denum tetrabromide is facilitated. 
 
 3. The formulas of the following complexes were derived 
 from their analyses : 
 
 5 K 2 O . WO 2 . 1 2 WO 3 . 1 3 H 2 O \ analogous to 5K 2 O . MoO 2 . 1 2 WO 3 . 1 6 H 2 O 
 5 Na 2 O . WO 2 . 1 2 WO 3 . 24 H 2 O / obtained by Wolcott Gibbs. 
 5 K 2 O.2 WO 2 .12 WOa.ll H 2 O 
 5 -Na 2 . WO 2 . 20 WO 3 . 23 H 2 O 
 
 4. The following ratios were also obtained: 
 
 22 K 2 O.TiO 2 .57 WO 3 .39 H 2 O 
 29 K 2 O . Zr0 2 . 70 WO 3 . 73 H 2 O 
 64 K 2 O . PbO 2 .171 WO 3 .119 H 2 O 
 
 5. Sulphur dioxide acts upon ammonium paratungstate 
 removing part of the volatile alkali and forming ammonium 
 metatungstate. This forms a convenient method for the prepara- 
 tion of the latter compound. 
 
-Wot - *. T oo S0 ^r s V 4E2 
 
 -"fiSTs* 
 
 *<J>2I_ 
 
 10 Om-R >. 
 
YD 04897 
 
 UNIVERSITY OF CALIFORNIA LIBRARY