192 MUIR ON CERTAIX BISMUTH COMPOUNDS. By M. M. PATTISON MUIR, F.R.S.E., Praelector in Chemistry, Caius College, Cambridge. 1. I n the sixth part of these investigations I have pointed out one or two points of analogy and difference between the trichlorides of bismuth and phosphorus (Chem. Xoc. Jour., 1877, ii, 136). The former is less completely oxidised than the latter, when acted upon by air or by nitrogen trioxide, the chloride being in the state of fusion. Michaelis (Jenaische Zeitschr., vi, 239, vii, 110) has shown that a t high temperatures phosphorus trichloride is capable of decomposingMUIR ON CERTAIN BISMUTH COJIPOUXDS. 193 very stable compounds ; in these decompositions the trichloride acts as a reducing agent. I have examined the action of bismuth tri- chloride upon a few of those substances, the action of which upon the corresponding phosphorus chloride was investigated by M i c h a e lis.Sulphur chloride and phosphorus trichloride, when heated in sealed tubes to 160°, are decomposed in accordance with the reaction, szc12 + 3Pc13 = Pc1, + 2Psc13. A quantity of sulphur chloride was heated in a sealed tube with excess of bismuth trichloride for 10 hours at 170 to 175". The tube then contained crystals of separated sulphur ; the bismuth trichloride remained altogether unacted upon. AccordingtoMichaelis,the reaction 3PC1, + SO,= 2POC1, + PSCI, takes place when sulphur dioxide and phosphorus trichloride are passed through a red-hot tube. Bismuth trichloride was heated over a Bunsen lamp in a current of dry sulphur dioxide ; the greater part of the trichloride was quickly volatilised and deposited again in the cold portion of the tube; the sublimate consisted of pure bismuth frichloride; the residue in the boat presented the appearance and possessed the properties of the oxychloride described in Part VI of these papers (Chem. Soc.Jouvn., 3877, ii, 136). Estimation of chlorine showed that the residue, the amount of which was small compared with the weight of trichloride used, really consisted of this oxychloride. An experiment, in which sulphur dioxide and vapour of bismuth trichloride were passed through a red-hot tube, gave the same results. Chromyl dichloride and phosphorus trichloride react very energeti- cally upon one another. Michaelis gives the equation 4Cr02C12 + GPCl, = 2CrzC16 + PC1, + 3Poc1, + Pz05, as expressing the results of this reaction.When bismuth trichloride and chromyl dichloride are mixed, no action takes place; on warming the mixture, the bismuth salt gradually dissolves. When the dark-red liquid is left for some days over sulphuric acid, the chromyl dichloride is gradually absorbed by the acid, and a dark-red semi-solid mass is produced, which continues t o evolve vapours of the dichloride, t o a greater or less extent, for a long time. If exposed t o the air, this dark-coloured mass quickly deliquesces, with production of bismuthyl chloride (BiOC1) and a solution containing chromic and hydrochloric acids, but entirely free from bismuth. The same decomposition is more quickly brought about by the addition of water to the semi-solid mass.Ether dis- solves out a small quantity of bismuth trichloride from this ma= ; the solution is free from chromium. Absolute alcohol causes the pro- duction of a green liquid containing chromium and chlorine, but free194 NUIR ON CERTAIN BISMUTH COMPOUNDS. from bismuth. These reactions, I think, show that no chemical action has taken place between the bismut,h and chromyl chlorides, but that the substance obtained is merely a solution of the bismuth salt in chromyl dichloride. 2. These reactions are in keeping with the facts that bismuthous chloride is less completely oxidised, under similar conditions, than the corresponding phosphorus compound ; and that no bismuthic chloride analogous to phosphoric chloride (PCI,) exists. 3.In a paper lately communicated to the Society (p. 170 of this volume) I have described a process for det'ermining bismuth volu- metrically, based upon the production of an ox a1 a t e having the formula, Bi,CIO,. This oxalate is produced by boiling the normal oxalate, Bi2C60L2, with water. Sonchay and Lenssen (Awn,. Chem. Pharm., cv, 245) have described normal bismuth oxalate as a salt containing 7 i molecules of water, 64 of which are given up by heating to 100." In Watts's Dictionaq (vol. iv, p. 253), the formula of this salt is given as Bi2C6012.15Hz0: this is evidently a mistake. In the old notation the formula was Bi2C12021.15H0, which in the new becomes Bi2C6012. 7 $H,O. According to Souchay and Lenssen, the salt loses 14.3 per cent. of its weight when heated to 100".A quantity of bismuth oxalate was prepared by precipitating a slightly acid solution of bismuth nitrate with excess of a saturated solution of oxalic acid. The precipitate was washed with a very dilute solution of oxalic acid, dried by long-continued pressure between porous paper, and analysed. The oxalic acid was determined by titration with standard perman- ganate ; water was determined by heating in an air-bath, and bismuth by difference. (1.) 0,471 gram required 67.2 C.C. permanganate = 0.1565 gram (2.) 0.659 gram required 95.1 C.C. permanganate = 0.2215 gram (3.) 1.363 grams lost 0.1745 gram after 5$ hours' drying a t 100'. c20,. C?O& 1 C.C. permanganate = 2.329 mgm. C204. Calculated for Found. Bi23C,0j.6H20. Bi23C2O4.'i'+H20. I. 11. 111. c200...... .. 33-33 32.23 33-23 33.61 - Bismuth. . . . 53.03 51.29 53.77 c Water.. . . . . 13.64 16.48 - - 12.81 When the oxalate was heated to 130" for an hour, the coloupNUIR ON CERTAIN BISMUTH COMPOUNDS. 195 changed to reddish-violet, the total loss amounting to 14.09 per cent. After 3+ hours at the same temperature, the loss amounted to 14.38 per cent., the violet colour having become considerably deeper. At 150" decomposition proceeded with greater rapidity. These results are more in keeping with the formula contahing 6 molecules of water than with that given by S ouch ay and L ens s e u, which requires 'i'i molecules, The water is evidently driven off at a temperature of 110" or so ; and at a point not much above this, if one may judge from the change in the colour of the salt, decomposition begins. The chemists already quoted state that bismuth oxalate is decomposed at temperatures above loo", with production of an acid oxalate and hypobismuthous oxide (Bi,Oz).I heated a quantity of the oxalate, prepared as already described, in a closed crucible over a low Bunsen flame ; the salt became black, and then began to assume a yellow colour; the heating was stopped; the residue washed with hydrochloric acid and exposed in a moist state to the air for several hours ; no trace of the white hydrate which is produced by the oxida- tion, under similar conditions, of hypo bismuthous oxide, was obtained. Examined under the lens, the residue was seen to consist largely of metallic bismuth ; on heating, it was slowly converted into bismuthous oxide.The hydrochloric acid washings showed no traces of oxalic acid. It would thus appear that at a temperature considerably below redness bismuth oxalate is decomposed in a closed crucible, the oxalic acid being destroyed, and metallic bismuth being produced by the re- ducing action of the carbon particles, and perhaps also by carbon monoxide given off from the decomposing salt. 4. Sonchay and Lenssen (Zoc. cit.) have described a basic oxalate of bismuth produced by boiling the normal oxalate with water until the supernatant liquid ceases to exhibit an acid reaction. To the oxalate thus obtained these chemists have assigned a formula which, translated into the new notation, becomes BizCa09.H20. Hein t z (Pogg. Ann., lxiii, 90) has examined the same salt and assigned to it the formula Bi2C40,.1+H20. According to Souchay and LensPen this oxalate loses no water at loo", but begins to decompose at 132".According to Heintz, decomposition begins at 200" to 210°, and is attended with evolution of carbon dioxide only. I prepared a quantity of the salt in accordance with the method already quoted. The analysis was conducted by dissolving in dilute hydrochloric acid, and titrating with standard permanganate solution. 1. 0.833 gram required 104.5 C.C. permanganate = 0,24338 gram C204 = 0.5808 gram Bi. 2. 0.6805 gram required 84.4 C.C. permanganate = 0.19657 gram C204 = 0.41691 gram Bi.196 MUIR ON CERTAIN BISXUTH COJIPOUSDS. 1 C.C. permanganate = 2.329 mgm. C,04. Bi2C409. Bi2C,0g.H20. Bi,C,OJkH,O. I.11. Mean. Calculated for Found. C20,. . . . . . 28.76 27.94 27.55 29-22 28.89 2905 Bismuth . . 68.63 66.66 65.73 69.72 68.93 69-32 Water . . . . - 2.86 4.23 - - - In the foregoing analyses the bismuth was determined from the re- sults of the titration with permanganate ; whatever be the amount of wat'er in the salt, the relation between C204 and Bi remains unaltered. Inasmuch as C,Oa corresponds with Bi, it is easy to calculate the per- manganate in terms of bismuth after it has been standardised against oxalic acid or an oxalate. 0.849 gram was dried at 100" for some hours, but without suffer- ing any diminution in weight. Dried for 1+ hours at 150") the salt was slightly blackened, and lost 0.007 gram = 0.83 per cent. After a second hour the loss amounted t o 0.008 gram = 0.94 per cent., the salt being rather more blackened than before. After an additional hour at 185-190", the salt became dark brown throughout, and lost 0.019 gram = 2.24 per cent.When heated over a Bunsen lamp, the oxalate was gradually converted into bismuthous oxide. These num- bers certainly agree best with the supposition that this oxalate con- tains no water of crystallisation. At that temperature at which it begins to lose weight, it also begins to undergo decomposition. 5. The propertiesof the two oxalates of bismuth, so far as the action upon them of acids and other reagents is concerned, have been de- tailed by Souchay and Lenssen with tolerable fulness. I find that, by mere contact with cold water, the normal is slowly converted into the basic oxalate. 6.The two oxalates of bismuth may be regarded either as derived from oxalic acid or from two of the hydrates of bismuth. From the former point of view the salt described in par. 3 is regarded as three molecules of oxalic acid, in which the whole of the bydrogen has been replaced by bismuth- The basic salt is derived from two molecules of oxalic acid in which two hydroxyl groups are replaced by oxygen, that is to say, from an acid which bears exactly the same relation to oxalic acid that dichromic doe8 to chromic acid, or pyrosnlphuric to sulphuric acid. By repla-MUIR ON CERTAIN BISMUTH COMPOUNDS. 197 cing the whole of the hydrogen in this dioxaZic acid by bismuthyl (BiO) we shall have the bismuth salt (BiO)zC407- If the oxalates be regarded as derivatives of bismuth hydrates, the normal salt will be obtained from the hydrate Biz(OH)6 by replacing the hydrogen entirely by the group C202, thus- BiTO i0>C,O, while the basic salt will be obtained from the hydrate Bi,O(OH), by a similar replmement, thus- From whichever point of view these oxalatm are regarded, the basic salt is completely analogous to Lowe's dichromate of bismuth (see Chem. Xoc.Jour., 1876, ii, 19; and 1877, i, 649-651). The normal salt must obviously be called bismuth oxalate; for the basic salt I would propose the name of bismuthyl dioxalate. The cor- responding antimony1 dioxalate has been described by Souchay and Lenssen. 7. I have already described several attempts which I have made to prepare salts of the so-called bismuthic acid (Bi,O,.H,O).In one of the oldest papers upon bismuth salts, the preparation of three so-called bismuthates of potassium is detailed. Jacquelain (J. pr. Chem., xiv, l), in the year 1838, described experiments which resulted in the production of three salts, to which he assigned the for- mulze (new notation) BBi407.K20 ; 4Biz04.Kz0 ; and 7Bi20,.2Kz0 respectively. I have again carried out two of the experiments made by Jacquelain exactly in the manner described by him. A quantity of caustic potash was fused in a silver crucible, and bis- muthous oxide was thrown in small successive quantities into the molten mass. The oxide dissolved very readily, the liquid becoming first greenish-yellow, then yellow-brown, and finally dark red-brown. The latter colour was attained when a considerable amount of bis- muthous oxide had been added, and when the whole had been strongly VOL.XHXIIT. Q198 MUIR ON CERTAIN BISMUTH COMPOUNDS. heated for 10 or 15 minutes. Tbese observations are entirely in keeping with those made by Jacquelsin. On allowing the fused mass to cool, a black solid was obtained, which appeared under a lens as a non-homogeneous mass containing grey or greyish specks. J a c que 1 ai n describes the fused mass as presenting the appearance of aventurin, and as containing quantities of crystals. I have re- peated the experiment several times, and each time have obtained a similar result. The black mass is very deliquescent; it is readily acted upon by water, with production of a strongly alkaline solution containing no bismuth, and a brownish-red residue, which, when washed with cold water until the washings cease to turn red litmus blue, is entirely free from potassium.The black mass is slowly decomposed by alcohol, with production of a brownish residue and an alkaline liquid free from bismuth. When thrown into fused potash, the black solid is very quickly dissolved. Jacquelain says that it is much more quickly dissolved than bismuthous oxide. I have not observed much difference be- tween the solubilities of the two substances in fused potash, so far as can be judged by the eye alone. The formula which Jacquelain assigns to the fiubstance which he obtained by adding bismuthous oxide to fused potash is 2Bi407.K20. The numbers obtained by Jacquelain agree tolerably well with this formula ; they are as follow :- Calculated. Found.Bismuth = 84.07 86-16 Oxygen = 11.21 11.56 Potash = 4-71 2.38 There is a considerable discrepancy between the amount of potash found and the amount calculated. Jacquelain’s formula, moreover, does not appear at all a probable one. If it be adopted, we must regard the substance as composed of bismuthous and hypobismnthic oxides combined with potash ( Biz03.Biz04)z.Kz0. The substance ob- tained as already described comports itself towards reagents exactly as a mixture of bismuthous oxide with one of the higher oxides would be expected to do. The two salts to which Jacquelain assigns the formulae 4Bi2O4.K20 and 7BizOa.2K20 respectively, are prepared, according to him, by the action of chlorine upon caustic potash-solution holding bismuthous oxide in suspension. The second salt is produced when a strong pot.- ash-lye is employed.If the percentage composition of the two salts be calculated, it is found that the numbers are very nearly the same for both. Jacquelain’s actual results for the second salt agree quite as well (indeed rather better) with the first, as with the second formula.MUIR ON CERTAIN BISNUTH COMPOUNDS. 199 I have attempted to prepare the salt 4Bi20,.K20 by following exactly the directions given by J a c qu el a i n, only employing bromine as oxidiser in place of chlorine, as the former is much easier to mani- pulate. The result was a dark puce-coloured heavy powder, which, after washing with cold water until the washings were no longer alka- line to test.paper, contained small quantities of potassium. Ou boil- ing this substance with water, the water again acquired an alkaline reaction, which reaction became more apparent the longer the boiling was continued. After being boiled with successive quantities of water until the latter ceased to exhibit an alkaline reaction, the substance was found to be perfectly free from potassium. On treating it with a little strong nitric acid, red biamuthic hydrate was produced. These results, taken along with the fact-insisted on by Jacque- 1 ain-tbat in the fusion cif bismuthons oxide with potash, an oxide higher than Bi,Os is certainly produced, and continues to exist at tempera- tures above that at which either Bi20a or BizO, is decomposed, can be best explained, it seems to me, by supposing that a loose combination of potash with Bi,Oa or Bi,O, is actually produced, but that this com- bination is very readily decompwed.In the first instance washing with cold water seems to effect decomposition ; in the second boiling water is necessary. I am aware that the expression " loose combina- tion " is a bad one on account of its vagueness, but I do not know of a better. The general result appears then to be that the higher hydrates of bismuth exhibit exceedingly feeble acid properties, so feeble as cer- tainly not to entitle them to the name of acids. These oxides do not combine with acids to form salts: hence they occupy a neutral posi- tion between the more marked positive oxides on the on8 hand, and the negative oxides on the other.8. I thought it might be possible to form sulphobismuthates. With this end in view, several experiments were carried out, but they have all led to negative results. The experiments were briefly as follows :- An attempt was made to produce a sulphide of bismuth higher than Bi2S, by passing sulphuretted hydrogen through water holding bis- muthic oxide (Bi,O,) in suspension. A dark red powder was pro- duced, which contained only 1 per cent. of sulphur. Treatment with alkali removed the sulphur, leaving bismuthic oxide. Another attempt to prepare a sulphide higher than bismuthous sul- phide was made by fusing carbonate of potassium, sulphur, carbon, and bismuthous sulphide together. The result was potassium sulphide in addition to the material employed in the fusion.Lastly, an attempt to prepare a salt of bismuthous sulphide by fusing metallic bismuth, potassium carbonate, and sulphur led only to the for- mation of bismuthous sulphide and potassium sulphide. The fused Q 2200 XUIR ON CERTAIN BISMUTH COMPOUSDY. mass, treated with water, yielded a solution free from bismuth, the residue, after washing with water, yielded a solution free from potas- sium. Schneider (Zeitsch. f. Chem. [ a ] , v, 630) stat8es that a soluble double sulphide, K2S.Bi2S3, is formed by the foregoing process. I have altogether failed to produce such a compound. 9. I have carried out a few experiments with bismuthous iodide, which it may be well to record in this place. The iodide was pre- pared by the wet method (Rammelsberg, Pogg. Ann., xlviii, 166, and Arppe, Pogg.Ann., lxiv, 248), also by Schneider's dry method (Pogy. An%., xcix, 470), which consists in heating 'together a mixture of bis- muthons sulphide and iodine. Rammelsberg says that a solution of potassium iodide converts bismuthyl chloride (BiOC1) into bismuthous iodide, and that in the preparation of the latter s+lt it is therefore immaterial whether the solution of bismuth contain precipitated oxychloride or not. The second part of this remark I find to be perfectly correct, but there is an error in the first part. Pure bismuthyl chloride (BiOCl) was digested at the ordinary tem- perature for several hours with a large excess of a solution of potassium iodide, but not a trace of bismuthous iodide was produced. The addition of a few drops of hydrochloric or nitric acid caused the forma- tion of a yellow liquid, but still no solid bismuthous iodide was formed.A solution of bismuth in excess of nitric acid was added to another portion of the oxychloride ; the first few drops caused thc production of ;t yellow liquid, the addition of a larger quantity produced a preci- pit(ate of bismuthous iodide. Hence it is evident that bismuthyl chlo- ride is not decomposed by potassiuni iodide solution. The addition of excess of potassium iodide solution to a mixture of bismuth chloride, bismuth oxychloride, and hydrochloric acid probably first causes the conversion of the dissolved bismuth into iodide, which is at once for the most part precipitated, with simultaneous production of potassium chloride. As potassium chloride is an easily soluble salt, a portion of the hydrochloric acid previously employed in holding the bismuth in solution -ill be free to dissolve fresh portions of bismuth by acting on the oxychloride present ; the bismuth thus dissolved will be a t once converted into iodide, and as such removed from the sphere of action ; and these processes will continue until the whole of the bismuth is precipitated in the form of iodide.In keeping with this explanation is the fact which I have noticed, that if a small quantity of potassium iodide solation be added to a somewhat large quantity of bismuthyl chloride, and a few drops of hydrochloric acid be then poured on the mixture, the bismuth is slowly converted into iodide. There appear to be some discrepancies in the accounts given of the action of water uponbismuthous iodide.I find that the iodide producedMUIR ON CERTAIN BISJiUTH COMPOUNDS. 201 by precipitation is decomposed with tolerable facility by cold water with production of the red oxyiodide, BiOI. The complete decom- position requires the agitation for some minutes of the precipitated iodide with three or four successive quantities of water. If hot water be employed, the decomposition is very much hastened. The iodide produced by subliming a mixture of bismuthous sulphide and iodine is decomposed by cold water only very partially, and only after pro- longed contact; even when boiled with water the decomposition of the iodide is very slow. Bismuthous iodide is therefore a much more stable salt than either the corresponding chloride or bromide.When heated in a closed crucible, the iodide is for the most part sublimed unchanged, but a small quantity of the oxyiodide, BiOI, is produced, as noticed by Schneider (J. pr. Chew,., lxxiv, 424). I n this re- spect the iodide differs from the other haloid bismuth compounds. In a paper communicated to the Society (" On Certain Bismuth Com- pounds," Part TI, Chem. SOC. J., 1877, ii, p 137) I have compared the oxidising action of air upon fused bismuthous chloride with the action upon fused bismuthous bromide. I n the case of the latter salt the oxidation is carried further than in the case of the former. The iodide is, however, oxidised much more slowly than either the chloride or bromide. Hydrochloric acid, i t is stated, dissolves bismuthyl iodide (BiOI) readily ; nitric acid causes the production of free iodine, while the bis- muth goes into solution. I find that the addition of a very small quantity of hydrochloric acid to bismuthyl iodide brings about the production of bismuthous iodide, which is for the most part precipi- tated, and bismuthous chloride, which goes into solution. The addi- tion of a few drops of sulphuric acid also causes the production of bismuthous iodide and a liquid holding bismuth in solution. Nitric acid when added in very small quantity exerts a Rimilar reaction, but the bismuthous iodide produced is very quickly decomposed, with solu- tion of bismuth and separation of iodine. I also find that hydriodic acid acts slowly upon bismuthyl chloride, with production of bismuthous iodide. The two following equations in all probability represent the production of bismuthous iodide from the oxyiodide and oxychloride respectively :- 3BiOI + 6HCl = 2BiC1, + BiI, + 3H20. 3BiOCl + 6HJ = 2Bi13 + BiC1, + 3H20. Just as potassium iodide solution is without action upon bismuthyl chloride, so is potassium chloride solution without action upon bismu- thy1 iodide.