Analyst, March, 1974, Vol. 99, @. 163-165 163 A Combustion Method for the Determination of Total Sulphur in Limestones BY L. M. RUNDLE (Geochemical Division, Institute of Geological Sciences, 64-78 Gray's Inn Road, London, WClX SNG) A method is described for the determination of total sulphur in limestones. The sample is mixed with tungsten trioxide and quartz flour and placed in a boat of refractory material. The boat and contents are inserted into an aluminous porcelain combustion tube maintained at a temperature of about 1200 "C in a resistance-type tube furnace. The combustion tube carries a coil of copper gauze at the outlet end to reduce sulphur trioxide. Nitrogen is used as carrier gas and the sulphur dioxide evolved is absorbed in dilute hydrochloric acid containing potassium iodide and starch as indicator, titra- tion with standard potassium iodate solution being carried out as the analysis proceeds.SINCE this Institute undertook a feasibility study for a survey of resources of limestones, the need for a rapid method for the determination of sulphur in such material became apparent. The classical method of precipitating sulphur as barium sulphate is time consuming and susceptible to errors. Many combustion method~l-~ have been used, all of which involved somewhat different conditions. That proposed by Coller and Leiningerl requires a high tem- perature and equipment that is not readily available in this laboratory. The method of Sen Gupta2 in which vanadium pentoxide was used as a flux was found to give low recoveries when applied to the analysis of limestones, and tests made on the residues from the com- bustion boats in this method showed that some sulphur remained in them. Results improved when the heating time was increased from 30 to 50 minutes or longer, but in view of the number of samples to be analysed, this heating time was not practicable.In an attempt to overcome these problems, a flux composed of tungsten trioxide mixed with quartz flour, and a furnace temperature of 1200 "C were used. Results were slightly improved but still tended to be low. Examination of the residue showed sulphur to be absent, so it was assumed that part of the sulphur had been lost as sulphur trioxide. Intro- duction of a piece of copper gauze, about 40 x 120 mm, in the form of a 40-mm long coil into the outlet end of the furnace tube where the temperature lay in the range 750 to 850 "C overcame the losses.Results obtained with the modified method compared favourably with certificate values and gravimetric determinations. Samples other than limestones were also analysed and the results indicated that the method could be used for a wide variety of geo- logical materials regardless of the form in which sulphur was present. The presence of the copper gauze had no detrimental effect if sulphur were present as sulphide. EXPERIMENTAL APPARATUS- The apparatus used is shown in Fig. 1 and consists of the following: a furnace capable of maintaining a temperature of 1200 "C; an aluminous porcelain combustion tube, 0.6 mm long and of 20 mm i.d.; combustion boats of high-alumina fireclay with over-all dimensions of length 72 mrn, width 16 mm and depth 10 mm; oxygen-free (white-spot) nitrogen from a cylinder (further purification is unnecessary) ; a 100-ml tall-form measuring cylinder; and a multi-hole bubbler that reaches the bottom of the measuring cylinder.REAGENTS- Tungsten trioxide. Dilute hydrochloric acid (1.5 + 98.5). Potassium iodide solution, 3.0 per cent. m/V. Quartz flour, 240 mesh B.S.S. Starch solution, 2.0 per cent. m/V. @ SAC and the author.164, RUNDLE A COMBUSTION METHOD FOR THE DETERMINATION [Analyst, VOl. 99 Standard potassium iodate solution, 0.006 25 N-Disso1ve 0.223 g of potassium iodate (previously dried for 1 hour at 110 "C) in distilled water, dilute the solution to 1 litre in a calibrated flask and mix thoroughly. 1 ml of solution = 0-10 mg of sulphur.Copper gauze, 40 mesh. PROCEDURE- Mix, in a small bottle, 1 g of tungsten trioxide, 0.5 g of quartz flour and a known suitable amount of sample. To a measuring cylinder containing 80 ml of dilute hydrochloric acid (1.5 + 98.5) add 1 ml of potassium iodide solution and 1 ml of starch indicator. Place the measuring cylinder in position at the end of the combustion train. Pass nitrogen through the system at a steady rate (approximately 150 ml min-l) and titrate the absorbing solution with the standard potassium iodate solution to a faint blue colour. (The intensity of the blue colour can be adjusted to suit individual needs.) Resistance furnace I . :; r'..' . . .. . . ,,:, the red zone) Soda- Magnesium asbestos perch lorate Fig.1. Apparatus for the determination of total sulphur Transfer the pre-mixed sample into a combustion boat (which has been ignited previously and then cooled) made of refractory material and insert the boat quickly into the hot zone of the furnace (1200 "C). Continue to pass nitrogen through the system and titrate the absorbing solution so as to maintain the blue colour. Continue passing the nitrogen for about 15 minutes or until the colour of the absorbent remains unchanged for 2 to 3 minutes. Wash the inside of the bubbler by closing the screw-tap and again opening it. Any condensa- tion appearing on the glass tube should be vaporised either by increasing the nitrogen flow-rate at the end of the run or by heating the tube gently with a burner.If the colour of the absorbent fades, add another drop of titrant and ensure that the colour remains unchanged after a further 2 minutes. Carry out a blank test by using tungsten trioxide and quartz flour and make any necessary changes to the volume of titrant used. Sulphur, per cent. = - 100 x w Note the volume of potassium iodate solution required. 7) where v ml is the volume of potassium iodate solution used after correcting for the blank and w g the amount of sample taken. RESULTS Results obtained on standard materials by using the above method are shown in Table I. Selected limestone samples covering the whole range of sulphur contents encountered in the feasibility study were analysed by both the combustion method and the barium sulphate gravimetric method, which was applied after decomposition of the sample by fusion with sodium carbonate.The results of these determinations are shown in Table 11.March, 19741 OF TOTAL SULPHUR I N LIMESTONES TABLE I RESULTS OBTAINED ON STANDARD MATERIALS Total sulphur, per cent. 165 r 1 Value by Sample Certificate value combustion meth3d Argillaceous limestone N.B.S. No. la 0.27 0.28 Portland cement N.B.S. No. 177 . . 0.637 0.623 B.C.S. iron ore “A” . . .. .. 0.063 0.055 U.S. Geological Survey: Standard rock BCR-1 . . .. 0-04 0.040 Standard rock GSP-1 . . .. 0.03 0.033 The coefficients of variation obtained for the argillaceous limestone and Derbyshire limestones Nos. 4 and 11 were 6-07, 6-02 and 3.15 per cent., respectively, based on twelve determinations in each instance.The procedure described is satisfactory for all types of samples tested, although the heating period can be varied slightly, depending upon the form in which the sulphur is present. For example, if the sulphur is present entirely as sulphide the heating period can be reduced to less than 15 minutes. The copper gauze turns black after three or four runs and should be replaced; it can, however, be re-used after the oxide layer has been removed. In order not to melt the copper, it is important to place the gauze just outside the red zone, otherwise its purpose is not served. TABLE I1 RESULTS OF ANALYSIS OF LIMESTONE SAMPLES Total sulphur, per cent, by gravimetric combustion A I ’I Sample method method Derbyshire limestone No. 4 .. 0.05 0.04 No. 11 .. 0.50 0.45 No. 12 . . 1.95 1-92 Calcareous marl . . .. .. 0.22 0.22 CONCLUSIONS A rapid combustion - titrimetric method has been developed for the determination of sulphur in limestones. By using a flux composed of a mixture of tungsten trioxide and quartz flour, good sulphur recoveries are obtained over a wide range of sample compositions. Inclusion of a coil of copper gauze in the cooler part of the furnace ensures that any sulphur trioxide evolved is reduced to dioxide. The method is comparable in accuracy with the barium sulphate gravimetric method but has the advantage of greater speed. The author thanks Mr. P. J. Moore for helpful discussion with this paper, and the Director of the Institute of Geological Sciences for permission to publish this work. REFERENCES 1. 2. 3. Coller, M, E., and Leininger, R. K., Anulyt. Chem., 1955, 27, 949. Sen Gupta, J. G., Ibid., 1963, 35, 1971. - , Analytica Chim. Acta, 1970, 49, 519. Received July 9th, 1973 Accepted October 16t12, 1973