ESTIMATION OF SULPHIDE AND SULPHATE SULPHUR 275 THE ESTIMATION OF SULPHIDE AND SULPHATE SULPHUR, AND THE ACTION OF SOLVENTS ON VULCANISED RUBBER. BY HENRY P. STEVENS, M.A., PH.D., F.I.C. (Read at the Meeting, M a d 30, 1915.) PART ESTIMATION OF SULPHIDE SULPHUR. MANY technical rubber articles contain metallic sulphides and sulphates the presence of which complicates the estimation of the ‘‘ combined ” sulphur or ‘( coefficient of vulcanisation.” I n the ordinary analytical process the free sulphur and that present a8 substitute are extracted with acetone and alcoholic potash respectively.In the absence of sulphides and sulphates an estimation of sulphur in the residue gives the percentage in combination with the rubber. In the presence of sulphides and sulphates the usual process consists in heating a portion of the residue with high- boiling solvents to destroy the rubber and render it soluble.The sulphur is then estimated in the washed mineral residue. The sulphur is also estimated in another portion of the extracted rubber, and the sulphur combined with the rubber estimated by difference. The method is tedious and unsatisfactory. It has two disadvantages. I n the first place, many vulcanised rubbers are difficult to decompoRe ; they appear to carbonise and cake when heated even at carefully regulated temperatures.Con- sequently the residue, after washing with benzene, contains undissolved organic matter which protects the mineral sulphides . from deoomposition by acid. The heating process can be repeated, but if the first operation is not successful, reheating is usually without effect.Secondly, the method assumes that the vulcanised rubber does not react with basic substances, such as litharge or magnesia, present in the mixings during heating, with the formation of metallic sulphides, although vulcanising temperatures are employed. The method to be described is applicable to those sulphides which are decom- posed by heating with acids.I t is therefore suitable for the estimation of the sulphides of zinc and lead, whether as original constituents of the mixing or formed during vulcanisation. I t thus embraces one of the commonest types of rubber- mixings--namely, that in which litharge is a constituent. In the ordinary way the metallic sulphides present in either vulcanised or unvulcanised rubbers are so pro- tected by the rubber surrounding the mineral particles that the surface only is attacked by prolonged boiling with strong acids, such as hydrochloric acid solution.276 STEVENS : ESTIMATION OF SULPHIDE AND SULPHATE SCTLPHUR, I find, however, that if the vulcanised rubber be first swollen in a suitable solvent in which the aqueous acid is to some extent soluble, the metallic sulphides.such as those of zinc and lead, are easily and completely decomposed. Various solvents have been tried, of which ordinary methylated ether has been found the most suit- able, no doubt partly because vulcanised rubber swells rapidly in this solvent, and because it is miscible to a considerable extent with aqueous hydrochloric acid.Nevertheless, if preferred, benzene or one of the chlorinated hydrocarbons, such as dichlorethylene, etc., can be employed. The hydrogen sulphide liberated is estimated, and from it the percentage of sulphide sulphur can be calculated. The usual method for the estimation of hydrogen sulphide, such as oxidation to sulphuric acid, I have not found satisfactory. By precipitation as zinc sulphide in an excess of a solution of zinc acetate better results were obtained, but I prefer to use a solution of lead acetate.The absorption is very complete, barely a trace of lead sulphide being precipitated in the second wash-bottle. The freshly precipitated and washed sulphide is decomposed by shaking with iodine solution. The apparatus employed is simple, and consists of a Kipp’s apparatus for the production of carbonic acid, connected to a Voigt’s flask of 250 C.C.capacity, The latter consists of a stoppered flask with outlet tube connected to the ground-in stopper. The inlet tube is fused into the side of the flask, nearly reaching to the bottom. In this flask the rubber is decomposed, and the liborated hydrogen sulphide passes directly into two 100 C.C.stoppered absorption bottles containing a solution of lead acetate. Any suitable apparatus, such as a, Corleis flask (“Indiarubber Laboratory Practice,” Caspari, p. 151), can be used in the place of the Voigt’s flask above referred to, but the latter is much simpler, cheaper, and quite as serviceable. The apparatus is worked as follows: Ten to twenty C.C.pure concentrated hydrochloric acid are poured into the Voigt’s flask and covered with a layer of ether, 20 to 30 C.C. being required.* The air is then swept out of the apparatus by means of a current of GO,. This should be &one before joining up the absorptiou flasks, so as to avoid a heavy precipitation of lead carbonate. The stopper is then removed from the flask for a couple of seconds while a weighed quantity of the rubber under examination is dropped in.The quantity taken may be 0.1 grm. to 1 grm., according to the proportion of sulphides the rubber is likely to contain. The absorption flasks are now joined up, and a little more carbonic acid passed through to drive out any traces of air. The whole is then allowed to stand, with occasional shaking, fifteen to thirty minutes for ordinary soft, rubbers, or longer in the case of hardened or very fully cured specimens.During this period the rubber swells and is gradually decom- posed by the acid, with the evolution of hydrogen sulphide. The ether, with the hydrogen sulphide, ie now driven over by gently heating the flask, and the decom- position completed by boiling for a few minutes.Rubbers compounded with litharge or containing sulphides of lead turn white, and heating should not be commenced until all black specks or patches have dis- appeared. After heating, a gentle current of carbonic acid is passed to wash out * If the rubber teiids t o sink in the concentrated aqueous hydrochloric acid, it should first be swollen in ether alone, as when immersed in the acid the swelling action of the ether is relatively slow ; or the specimen may bc wrapped in filter-paper, to keep it afloat in the ether.AND THE ACTION OF SOLVENTS ON VULCANISED RUBBER 277 any remaining traces of hydrogen sulphide from the flask, while the whole is allowed to cool.The contents of the absorption vessels are acidifiea with acetic acid to decompose any carbonate present, and the lead sulphide filtered off cold and well washed on the filter-paper till free from lead salts.I t is transferred while still wet fo ib stoppered flask, and allowed to stand;with occasional shaking,in contact with n, known volume of standard iodine solution. The excess of iodine is then titrated back with thiosulphste, or any other suitable method may be employed for the estimation of the lead sulphide formed.It is preferable to take a portion of the acetone-extracted specimen for this estimation, as it swells faster in the ether, but a portion of the original rubber may be used. As examples, the following may be quoted : 1. Zinc sulphide mixing, suitably vulcanised, of the followiug composition : Rubber (Para), 33 per cent.; sulphur, 10 per cent. ; zinc sulphide, 57 per cect. 0.1 grm. of rubber yielded lead sulphide, which required 11.6 C.C. of & iodine Sulphide sulphur found 18.6 per cent. ; theory, 18-8 per cent. 2. Lead sulphide mixing, vulcanised of the composition : solution. Rubber, 100 parts ; sulphur, 5 parts ; lead sulphide, 7 parts. One grm. of rubber yielded sulphide which required 5-1 C.C.of & iodine Sulphide sulphur found 0.82 per cent. ; theory, 0.84 per cent. Zinc and lead sulphides are only occasionally added as such in the manufacture of rubber goods. It much more frequently happens that they are formed during vnlcanisation by reaction between the sulphur present and the corresponding oxides. This is particularly the case with lead sulphide, as litharge is a very common ingredient of rubber goods, acting as an accelerator and toughening the rubber.It is in the analyses of such'vulcanised rubber goods that the above process is of particular value. The following figures, to which I shall have occasion to refer in a subsequent paper (J. SOC. Chem. Ind.), give duplicate results for a series of rubbers compounded with sulphur and litharge : Series I.First determination ... 0-61 0.70 1-15 1.22 1.28 1-47 1.60 1.70 1.50 Second ,, ... 0.61 0.75 1.19 1.28 1.34 1.44 1-63 1-76 1.44 Mean ... ... ... 0.61 0.73 1.17 1.25 1.31 1-46 1.62 1.73 1.47 solution. Sulphide Sulphur per Cent. 2. 3. 4 . 5 . 6. 7. 8 . 9 . 1 0 . Series II. Sulphide Snlphur per Cent. 2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 . 1 0 . First determination ...0.52 0.67 - - 1-12 1-41 1.70 1.89 1.76 Second ,, ... 0.54 0.70 0.86 1-00 1.15 1-44 1.68 1-83 1.71 Mean ... . . . ... 0.53 0.69 - - 1.13 1-43 1.70 1.86 1-74278 STEVENS : ESTIMATION OF SULPHIDE AND SULPHATE SULPHUR, PART II.-ESTIMATION OF SULPHATE SULPHUR. The reaction between the litharge and sulphur is usually expressed by the following equation : That is to say, for every three parts of sulphide sulphur there is also formed one part of sulphate sulphur.The lead sulphate will be retained in the Voigt’s flask at the end of the reaction partly dissolved in the hydrochloric acid. A number of experi- ments were made by repeatedly extracting the residual rubber with hydrochloric acid, and uniting the extracts and original hydrochloric acid solution.The sulphate was then estimated as barium sulphate. In no instance has the theoretical proportion of sulphate been obtained. Generally the amount of sulphate sulphur did not exceed one-quarter to one-fifth that of the sulphide sulphur present. Although the equation given above represents tbe reaction which takes place between litharge and sulphur in an inert medium, the reaction may be modified in the presence of the caoutchouc hydro- carbon SO that the latter is partially oxidised, with the formation of a larger propor- tion of lead sulphide and a smaller proportion of lead sulphate.This would explain the deficiency in the sulphate formed as shown in the numerous estimations cited. To study the matter as closely as possible, specimens of vulcanised rubber were subjected to repeated extractions with a mixture of hydrochloric acid and ether.When rubber is swollen under these conditions, a small part of the vulcanised rubber passes into solution with the sulphur, with which it is associated or in combination. Figures are given for repeated extractions carried out on two vulcanised rubber mix- ings. At each operation the percentage of vulcanised rubber dissolved in the ether was estimated, as also its sulphur content and the amount of sulphates in the aqueous hydrochloric acid solution.After extraction every few days extending over a period of forty to fifty days, the rubber was allowed to stand with the ether and hydrochloric acid for several months, when the experiments were brought to a conclusion by deter- mining the percentage extract, sulphur as sulphate as above, and also the sulphur content of the residual rubber. The following figures were obtained : 4Pb0 + 2S, = 3PbS + PbSO, Repeated Estraction of Vulcanised Rubber with Ether and Aqueous Hydrochloric Acid.-Table I.Formula of mixture : Rubber . . . ... 89.0 per cent. Sulphur . . . i.. 5.5 ,, Litharge . . . ... 5.5 ,, Vulcanised in steam between cloth.The specimen was first extracted with acetone. Number of extractions ... 1 2 3 4 5 6 7 8 9 Periodof extraction: days 2 3 5 5 7 7 7 14 3months. Etherealextractper cent. 6.9 1.8 4.4 3.8 5.0 3.2 5.0 7.6 23.8 61.5 Sulphur in acid ethereal extract per cent. ... lost 0.03 0.06 0.06 0.08 0.06 0.06 0.15 0.38 0.88 Sulphur in aqueous hydro- chloric acid extract per cent.... ... ... 0.10 trace nil trace trace nil nil nil nil 0.10 Totals.AND THE ACTION OF SOLVENTS ON VULCANISED RUBBER 27‘3 The residual rubber left after extractions (38.5 per cent.) gave 0.80 per cent. combined sulphur, which, together with 0.88 per cent. present in the extracts, gives 1.68 per cent. sulphur in combination with the rubber, and a coefficient of vulcanisa- eion of 1.87.All percentages are calculated on the weight of original vulcanised rubber taken. The figures giving the total extracts, etc., are not of great value, as there is a cumulative error of seven to nine separate determinations. The results are of interest as illustrating the gradual solution of the rubber, together with its proportion of sulphur of vulcanisation.If, after estimation of the sulphide and sulphate sulphur in the manner described, the residual rubber be oxidised, the sulphur in combination with the rubber may be estimated. I n this manner the whole of the.sulphur in a vulcanised compound containing lead or zinc sulphides may be accounted for as follows : 1. The free sulphur by oxidation of the acetone extract. 2. The sulphide sulphur by decomposition with ether and hydrochloric acid, and 3.The sulphate sulphur present in the hydrochloric acid solution. 4. The sulphur combined with rubber by oxidation of the residue. I n illustration, the following analyses of vulcanised litharge mixings may be estimating the hydrogen sulphide evolved. quoted : 1. Free sulphur . . . 2. . Sulphide sulphur 3. Sulphate sulphur 4.Combined sulphur Total ... ... Percentage added PART III.-EXTRACTION Table II. 1. 2. 3. ... ... 1.21 0.48 0.04 ... ... 0.69 1 -00 1.86 ... ... 0.10 0.13 0.34 ... ... 2.11 2.33 0.67 ..- ... 4.11 2-94 2 *91 ... ... 4-20 3-97 3.11 - - OF VULCANISED RUBBER WITH SOLVENTS. I t is generally held that fully vulcauised rubber is not soluble in common solvents, although these have a marked swelling effect.The tables above show that i n the presence of an acid, such as hydrochloric acid, the vulcanised rubber is gradually dissolved even in the cold, and in these cases the rubber dissolved contains about 1.5 per cent. of sulphur. This applies not only to mixings containing litharge, but also to rubber vulcanised with sulphur only, as well as with sulphur and other mineral ingredients.The rubber cannot be recovered physically unchanged from the solution, and the product tends to be ‘ 6 tacky” and viscous. Tables 111. and IV. give the figures obtained in a series of extractions with ether and hydrochloric acid carried out in the same manner as in Table I,, but with a, rubber vulcanised with 4-75 and 7 per cent. of sulphur respectively.280 ESTIMATION OF SULPHIDE AND SULPHATE SULPHUR Table III.Rubber . . . . . 100 parts-95*25 per cent. Sulphur . . . ... 5 parts- 4.75 ,, Formula of mixture : The specimen was first extracted with acetone. No. of extractions ... 1 2 3 4 5 6 7 8 9 Periodofextraction: days 2 3 5 5 7 7 7 14 3months. Totals. Extract per cent. ... 10.3 4.5 6.2 7.1 12.0 12.2 11.1 14.3 18.6 96.3 S in extract per cent.... 0-13 nil 0-07 0.09 0.16 0.12 0.21 0.24 0.36 1-38 The residual rubber left after extractions (3.7 per cent.) gave 0-10 per cent. combined sulphur, which, together with 1.38 pe; cent. present in the extracts, gives 1-48 per cent. sulphur in combination with the rubber, and a coefficient of vulcanisa- tion of 1-55. Table IV. Rubber , . . . . . 93 per cent. Formula of mixture : Sulphur .. . ... 7 ,, The specimen was first extracted with acetone. Number of extractions 1 2 3 4 5 6 7 8 9 10 Periodofextraction: days 2 2 3 6 6 5 7 7 7 14 Totals. Extract per cent. ... 1.20 lost 2.5 6.5 2.5 6.5 6.5 6.8 4-7 16.2 64.2 S in extract per cent. 0.14 lost nil 0.11 0.6 0.11 0-12 0.12 0.7 0.4 1.14 These figures ehow a progressive solvent action, although the flasks were not heated and were kept in a dark.cupboard.A similar effect can be obtained with a mixture of benzene and strong aqueous hydrochloric acid. A sample of rubber vulcanised with sulphur only was extracted with acetone and then treated in the cold for varying periode with fresh quantities of this mixture, with the following results : Per Cent. Bnlphnr Original RnLber.Extract. Per Cent. Sulphur E ~ ~ ~ $ ~ ~ s f . Per Cent. Extract. calculated on calculated on 1 24-25 0.42 1.72 2 37-40 0.80 2.14 3 23.45 0.52 2.23 4 13.80 0.30 2 *18 Total 98-90 2.04 Mean 2.07 -- - Four treatments over a period of a few weeks sufficed to dissolve practically the whole of the rubber. It is noteworthy that the first extract is poorer in sulphur than the later ones.These latter give the same figure of 2.2 per cent., which does not vary appreciably from the average of 2.07 per cent. The solvent action after the first extraction is not selective, and may be taken as an indication that the vulcanised rubber, regarded as a chemical compound, is practically homogeneous.VERSFELD : BROMINE METHOD OF ESTIMATING PHENOL 281 Chlorhydrocarbons act similarly to the mixture of solvents and hydrochloric acid. Dichlorethylene causes the rubber to swell quickly, and on standing for a lengthened period, the rubber partially dissolves. This solvent action may be due to the presence of acid decomposition products in the chlorhydrocarbon.* As a whole, however, the action of the chlorhydrocarbons does not appear to be more rapid than that of a mixture of benzene and acids. Further experiments are in progress to determine the best solvent or combination of solvents for treating the vulcanised rubber. If it is found possible to dissolve the rubber fairly rapidly, a new method of analysis will be opened up for soft rubbers, as the mineral matters will then be separable from the rubber hydrocarbons. Some preliminary experiments with soft rubbers containing a single mineral ingredient have given satisfactory results.