54 Analyst, January, 1967, Vol. 92, PP. 54-56 Determination of Sulphur in Carbons and Cokes by Gas Chromatography BY F. M. W. OLDS,* J. W. PATRICK AND F. H. SHAW (The British Coke Research Association, Chesterfield, Derbyslzive) A combined combustion - chromatographic method for determining sulphur in amorphous carbons and cokes is described. The method consists essentially in burning the sample in a stream of oxygen, separating the sulphur dioxide produced by using silica gel as adsorbent, and determining the concentration of sulphur dioxide with a thermal conductivity detector. The method is rapid, is eminently suitable for small samples and covers a wide range of sulphur contents (0.2 to over 20 per cent.). The reproducibility of the method is comparable with the accepted tolerances of the British Standard methods for determining sulphur in coke. The difficulties experienced with cokes of high ash content are discussed, and suggestions made for eliminating these difficulties.IN the course of studies of the formation and properties of carbon - sulphur complexes,l the need arose for a satisfactory method for determining 0.2 per cent. to some 20 per cent. of sulphur in small samples of sulphurised carbons. The present paper describes the experi- mental basis of the method, which may be of general interest, because of the ease and rapidity with which sulphur contents of small samples can be determined. The method depends on the chromatographic measurement of the sulphur dioxide produced on combustion of the sample. This was a novel application when the work was commenced, but a gas - liquid chromatographic method has since been described.2 Preliminary experiments in which various adsorbents were used, such as silica gel, alumina, active charcoal and molecular sieves, showed that the most satisfactory results were obtained with a silica-gel column at about 100" C.In the early work, in which nitrogen was used as a carrier gas, the sample was burnt in a separate combustion tube and admitted to the column either as a whole or as samples extracted by a hypodermic syringe and injected into the carrier-gas stream through a serum cap. For ease in routine analysis this method was replaced by a flow system in which oxygen is used as carrier gas and the sample is burnt in a combustion tube placed in the gas line before the column.METHOD APPARATUS- The chromatographic column is made of $-inch PTFE tubing wound into a coil, of 2-inches diameter, around an electrically heated aluminium block, and mounted inside a container filled with vermiculite and lagged with asbestos string. In the present work a column length of 6 inches gives adequate separation and the column temperature is controlled, by the use of a Variac transformer, at 100" C. The silica-gel adsorbent (chromatographic-analysis grade 25 to 52 B.S. mesh size) is pre-dried at 120" C in an oven for 2 hours. The column is packed by vibrating it until there is no further diminution in the volume occupied by the adsorbent, which is finally dried at 120" C in situ, by using a 50 ml per minute stream of oxygen, dried by passage through magnesium perchlorate.A differential thermal conductivity detector comprising two tungsten filaments (Gow- mac W2) in a stainless-steel cell is used. It is enclosed in a thermostatically controlled box (30" 0.1" C), together with the components of the associated Wheatstone bridge. The current (140 mA) for the Wheatstone bridge circuit is supplied by a constant-voltage supply unit, and a pen recorder (1-mV full-scale deflection) measures the out-of-balance e.m.f. of the bridge. The flow-rate of the oxygen carrier gas is 50 ml per minute. The sampling system consists of a transparent silica combustion tube (15 cm long x 1 cm diameter) with a B7 socket at one end. This tube is placed in the gas line immediately before a phosphorus pentoxide moisture trap and the silica-gel column.* Present address : Department of Science, College of Technology, Southend-on-Sea, Essex.OLDS, PATRICK AND SHAW 55 PROCEDURE- Weigh (to an accuracy of kO.005 mg) between 1 to 10 mg of the sample, depending upon the sulphur content, into a small silica boat, and place in the centre of the combustion tube. Steady conditions, as represented by a good recorder base-line, are attained after about 3 minutes, when the sample is rapidly burnt with a Teclu burner. The sulphur dioxide is eluted from the column, after the carbon dioxide, in about 5 minutes, and the time taken for a complete determination is about 10 to 12 minutes. The system is calibrated by using pure carbon - sulphur mixtures together with sulphurous carbons and cokes (of low ash content) in which the sulphur has been determined previously by British Standard methods.Well defined linear relationships between the sulphur content and either the area under the chromatogram or the peak height are obtained. For ease of measurement in routine procedure the peak height is used and the sulphur content in the sample being analysed is computed from an appropriate calibration graph. No evidence was found of any sulphur trioxide formation during combustion. RESULTS AND DISCUSSION The results in Table I for a variety of sulphur - carbons, e.g., sulphurised cellulose or poly(viny1idene chloride) carbons, show that the repeatability of the method is satisfactorily comparable with the accepted tolerances of the British Standard methods for determining sulphur in coke.TABLE I REPEATABILITY TESTS WITH SULPHUROUS CARBONS Sample Percentage of sulphur No. Determination Determination (1) ( 2 ) 1 0.55 0.53 2 0.73 0.76 3 0.98 0-92 4 0.54 0.56 5 1-17 1-19 6 0-83 0.83 7 0.40 0.40 8 0.69 0-74 9 1.15 1.17 10 0.93 0.98 11 1-58 1.65 12 16.04 16-02 13 5.19 5.13 14 8.83 8.89 Difference 0.02 - 0.03 0.06 - 0.02 - 0.02 0.00 0.00 - 0.05 - 0.02 - 0.06 - 0.07 0.02 0.06 - 0.06 (1-2) Good agreement is obtained between the chromatographic and standard methods with sulphurous amorphous carbons, for which this method was devised, but with coke samples the present method tends to give low results compared with the standard high temperature method (Table 11), apparently because of the influence of mineral constituents and fixation of sulphur in the residue as indicated by the results in Table 111.Cokes A, B, C, D and E were prepared from one parent coal, washed and cleaned to different ash contents. With cokes of low ash content the agreement between the present method and a standard method (high temperature method, British Standard 1016 : Part 7 : 1959) is good, but with cokes of high ash content the chromatographic method gives low results, the discrepancy increasing with increasing ash content. I t should be emphasised that the samples must be finely ground (-240 B.S. mesh), both from the point of view of reducing sampling errors in weighing small amounts, and of ease of obtaining complete combustion. I t is apparent that for cokes a combustion temperature, higher than that attainable with the Tech burner used in the present work, is necessary to obtain complete conversion of the total sulphur to sulphur dioxide and, where desirable, a small electric furnace may be suitable for this purpose.In this context a layer of finely divided alumina covering the sample should also prove to be beneficial.56 OLDS, PATRICK AND SHAW TABLE I1 COMPARISON WITH STANDARD METHODS Sample Sulphur carbon No. 1 No. 3 No. 4 No. 7 Coke No. 1 . . .. No. 2 . . .. No. 3 . . . . No. 4 .. . . No. 5 . . . . No. 6 . . . . No. 7 . . . . No. 8 . . . . No. 9 . . . . No. 10 . . No. 11 . . 2-Mercaptobenzothiazole S-Trithiane . . . . 1,2-Benzodiphenylene Sulphanilic acid? . . . . . . . . . . . . . . . . . . . . .. . . . . .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Percentage of sulphur found Chromatographic Standard method method* 3.35 3.47 3.72 3.86 3-72 3.49 3.88 3.57 0.53 0.65 0.73 0.86 0-95 1.04 0-57 0.54 1.18 1.31 0.83 1.02 0.40 0.45 0.72 1 -00 1.16 1.34 0.96 1-07 1.62 1.58 37-8 38.3: 65.6 69.6: 14.6 13.9; 18-1 18.5: ( A ) (B) Difference - 0.12 -0.14 0.23 0.3 1 -0.12 -0.13 - 0.09 0.03 - 0.13 - 0.19 - 0.05 - 0.28 - 0.18 -0.11 0.04 - 0.5 - 4.0 ( A 4 0.5 - 0.4 * High temperature method, British Standard 1016 : Part 7 : 1959. t AnalaR quality. Other organic compounds were not of AnalaR quality. $ Theoretical value. TABLE I11 EFFECT OF ASH CONTENT OF COKE Coke sample Ash, per cent. 0.60 0.96 2-54 4.52 8.88 6.87 7.58 Percentage of sulphur found Chromatographic Standard method method 0.71 0.65 0.64 0.65 0.80 0.86 1.03 1-16 1-04 1.24 0.84 1.14 0.83 1-12 (1) (2) Difference + 0.06 - 0.01 - 0.06 -0.13 - 0.20 - 0.30 - 0.29 (1-2) Organic compounds give reasonable agreement with the theoretical sulphur content, but care is needed in the heating procedure, in particular in the rate of heating, in order to avoid premature volatilisation and ensure complete combustion ; some compounds being more difficult than others in this respect. Controlled electric heating and possibly a lower temperature of combustion may overcome this difficulty. The work described formed part of the research programme of the British Coke Research Association, and the authors are indebted to Dr. H. E. Blayden for his helpful interest, and to the Director of the Association €or permission to publish this paper. REFERENCES 1. 2. Blayden, H. E., and Patrick, J. W., J . Chiw. Phys., 1960, 57, 931. Beuerman, D. R., and Meloan, C. E., Analyt. Chew., 1962, 34, 319. Received March 28th, 1966