SOLUBLE CHLORIDES AND TOTAL CHLORINE IN SOME ENGLISH COKES 143 THE SOLUBLE CHLORXDES AND TOTAL CHLORINE IN SOME ENGLISH COKES. BY STANLEY W. BRIDGE, B.Sc., F.I.C. (Read at the Meeting, Nawh 3, 1915.) DURING some experimental work carried out on the purification of water by percola- tion through coke filter-beds, the chlorine content of the water was found in one instance to have increased from 4 parts per 100,000 to 18 parts per 100,000, and the author was asked if he could supply any figures relating to the amount of soluble chlorides generally present in gas coke.A considerable number of figures have been publishedas to coal, the amount of soluble chlorides and total chlorine in which is a matter of coneiderable importance where the coal is carbonised in gas retorts or coke ovens, inasmuch as the presence of sodium chloride in the coal in material amount has a destructive action on the fireclay material used for the construction of the retorts or oven walls, especially where the coal is carbonised in a moist condition and at high temperatures.In addition, as the chlorine is largely driven off with the rest of the volatile matter in the form of hydrochloric acid, a large proportion of chlorine in the coal brings about difficulties in the removal of the tar and ammonia, owing to the formation of solid ammonium chloride in the condensing plant.Where the coal is burnt in apparatus involving contact of the waste gases with copper alloys, as in the copper fireboxes and brass tubes of locomotive boilers, considerable trouble is experienced with coals high in chlorides from corrosion of the copper-containing materids by the hydro- chloric acid evolved.Search of the literature failed, however, to reveal any information with regard to the chlorine content of coke except in one instance-namely, in a paper by Schreiber (Stahl and Eisen, 1910, 30, 1839)-where an analysis is given of the soluble salts extracted from a single coke sample.In view of the fact mentioned above, that in the case of water purification, when employing fresh coke on filter- beds, erroneous conclusions might be drawn if account were not taken of the144 STANLEY W. BRIDGE: THE SOLUBLE CHLORIDES chlorides likely to be extracted from the coke by the water, it was thought worth while to examine the soluble and total chlorine contents of a number of samples of coke obtained in ordinary practice in gasworks, and in addition a sample of metallurgical coke and a cannel coke prepared at low temperature were examined.The following table gives the results obtained, the chlorine in all cases being calculated as NaCl, and given as percentage of this salt on the original air-dried coke : TABLE I.Description. Gas-coke ... ... ... ... ... ... ... ... ,? 9 3 # ? ? ? 9 ) ? ? I ? 9 , ? ? ... 9 , I , I ? ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... . I . ... ... ... ... ... ... ... ... ... ... ... .. Metallurgick cok'e' . . . . ... Low temperature cannel residue District of Origin of Coal. Soluble Chlorides as NaC1.Derbyshire ,I 9 , I2 Durham Y Y 9 , 9 , $ 9 Lancashire Yorkshire Durgam Flint ,9 Per Cent. 0.262 0.191 0.045 0.040 0.122 0.087 0.055 0.037 0.013 0.056 0.080 0.029 00016 0.007 0.357 Total Chlorine aa NaCl. Per Cent. 0.499 0.341 0.134 0.137 0.344 0.119 0.102 0.086 0.039 0.076 0.144 0.099 0.086 0.071 0.918 The results show the considerable variation in the chlorine content of the coke, and generally the variations in the chlorine in the cokes are much the same as has been found by others in that contained in coal.I t is well known among coke-oven managers that many of the Derbyshire coals are exceptionally troublesome in respect of chlorides, and the average amount found in the Derbyshire cokes is higher than from any other district. The low-temperature cannel residue was very high in both soluble and total chloride, and the same has been found true in the case of several other cannel coals and cokes which have been partly examined.It will be noticed that, in the case of the metallurgical coke, the percentage of soluble chlorides and the ratio of soluble to total chloride ie very low, which would indicate that a greater proportion of the chlorine is driven off from the coal with the volatile matter when the coal is carbonised under coke-oven conditions--i.e., in a moist state, and allowed to remain in the oven till the percentage of volatile matter is very low.In order to obtain some idea as to extent to which the chlorine is driven off from the coal, a few experiments were carried out on the laboratory scale on a sample of coal well known for its high chlorine content, and on two samples of cannel also high in ohlorine.Eaoh of these coals was carbonised (1)jin a fireclay crucible at aAND TOTAL CHLORINE IN SOME ENGLISH COKES 145 Soluble Chlorides as NaC1. temperature of about 400’ to 500’ C., and (2) in a platinum crucible for fifteen minutes over a powerful Meker burner.The soluble and total chlorine were deter- mined in each coal or cannel, and in the low-temperature and high-temperature residue, the chlorine as before being expressed as NaCl; and in the case of the residues this is calculated back to percentage on the original coal or cannel, so that the figures show directly the losses which occur on heating. The results are given in Table 11. : TABLE 11.Total Chlorine as NaC1. Description. ... ... Crtnnel ... ... ... ... Low - temperature residue (volatile matter = 10.77 per cent.) ... ... ... ... --- --- ... Camel ... ... ... ... ... 1 0.077 Low - temperature residue (volatile matter ... = 7-79 per cent.) ... ... ... 0.192 High-temperature residue ... ... 0.068 ... --- 1.141 0.498 0.288 Per Cent. 0.068 ! 0.525 0.357 0.123 Per Cent.0.987 0.918 0.314 ... Coking-coal ... ... ... ... 1 I ::: I ... 1 Low-temperature coke (volatile matter = 13-08 ... ... per cent.) ... ... High-temperature coke (15 minutes) ... High-temperature coke (90 minutes) . , . ‘ 0.092 0.140 0.048 0.048 0.237 0.094 0.091 It will be seen that, in accordance with expectation, the proportion of total chlorine driven off at high temperatures is much greater than at low temperatures.On the other hand, in every case the amount of (‘soluble ” chloride is greater in the low-temperature residue than in the original coal or cannel. This doubtless arises from the fact that the soluble chloride cannot be completely extracted from the original coal when it is well compacted with bituminous matter, whereas after the bituminous matter has been destroyed by carbonisation and the whole mass rendered porous, the soluble chlorides can be much more completely extracted.In the cam of the coal sample, a further test was made to see what effect a more prolonged heating would have on the chlorine content, the sample in this case being heated over the Meker burner for ninety instead of fifteen minutes.As will be seen from the table, the prolongation of the heating does not affect any material change. The analytical methods employed were as follows : Solable ChZorzdes.-Fifteen to 30 grms. of a,n average sample crushed to pass a ten-mesh sieve were boiled with water for one hour in a, flask connected to a reflux condenser, the coke then filtered and washed, and the filtrate on cooling acidified with nitric; acid, followed by slow addition of dilute permanganate till the solution146 SOLUBLE CHLORIDES AND TOTAL CHLORINE IN SOME ENGLISH COKES was just permanently pink.Standard silver nitrate solution (1 C.C. = 0.001 grm. C1) was then added, the silver chloride filtered off, and the excess of silver determined by titration with standard ammonium thiocyanate according to the Volhard method. The addition of permanganate was found necessary, as on the direct addition of silver nitrate to the acidified solution some blackening always occurred. This was a t first thought to be due to soluble sulphides from the coke, but when lead carbonate was added to the coke suspension before filtering the blackening still took place, and appeared to be due to some reducing agent extracted from the coke.The addition of permanganate to the acidified solution entirely obviated the difficulty. TotaE Chtlorim.-2-5 to 5 grms. of the coke were gently ignited with 3 to 6 grms. of a mixture of 1 part pure sodium carbonate and 2 parts pure lime until the carbon was completely burnt, the mass being then boiled with water to which a little lead carbonate was added to remove sulphides. After filtering and washing, the filtrate- concentrated if required-was acidified with nitric acid, and the chlorine determined in the manner described above. To ascertain whether any loss of chlorine took place during the ignition for burning off the carbon, pure sodium chloride was mixed with sugar, charcoal, and the above lime-sodium carbonate mixture, and strongly ignited to burn off the carbon, and 99.8 per cent. of the sodium chloride was recovered in the filtrate. Resublimed monochloracetanilide was also gently ignited with the mixture, and found to contain 20.11 per cent. of C1 instead of the theoretical 20.92 per cent.-an immaterial error for the purpose of the present investigation.