Sulfur in Ores, Concentrates, and Other Metallurgical Samples Routine Determination by the Combustion Method W. GEOFFREY RICE-JONES Consolidated .')lining und Smelting Co. of Canada, L t d . , Tmil, R . C . , Cunadu A rapid and accitrate method of sulfur determination was sought, which would be adaptable to wide variation in the types of samples handled and woulcl not require additional laboratory space. The conibustion method of sulfur determination permits conipletion of 60 determinations per da: h! one operator writ11 accuraq within f0.27~. I t is reliable for a wide variety of nietallurgical samples with a sulfur content from a trace up to XCl, o r more. Even samples containing a high percentage of barium may be handled. The field of application of the combustion method has been extended, aided by inclusion of the hack-titration feature. Depending on the type of sample handled, this method is comparable to, or better than, pre\ioizs methods from the cost viewpoint.
T
HE procedure dcscrihd hctre has heen a ( h p t d f r o m mcthocls in common use for the deterniiiiation of total sulfur in steel ( 1 ) to provide a rapid routine method for determining total sulfur from a trace up to 35% in metallurgical samples. This combustion method avoids the difficulties encountered in classical methods ( 2 )antl it is simple and fast. REAGENTS
Standard Potassium Iodate. Under conditions in this procedure the salt is dissolved in wat'er to give a concentration of 1.0698 grams per liter, so that 1 ml. is empirically e uivalent t o about 0.000500 gram of sulfur. An exact empiric3 factor for each class of sample is determined following the general procedure outlined below. Hydrochloric Acid, 0.2 N . Starch Solution is made by dissolving 4.5 grams of starch and 45 grams of potassium iodide in 1 liter of water. Sodium Thiosulfate. This solution is made up so that 0.05 ml. is equivalent t o 0.01 ml. of the potassium iodate solution. To standardize, a 10-ml. aliquot of the thiosulfate is added t o t,lie hydrochloric acid-potassium iodide-starch absorbing solution used in the general procedure, and the oxygen is allowed to flon- to keep the solution stiri,etl while titrating with the standard iodate. APPARATUS
Oxygen Supply. A flow of 1100 zk 100 ml. per minute is maintained using a Bow meter. .1 mercury manometer is connected to this supply to check the back pressure and to check for leaks i n the circuit. The Combustion Furnace is a Dietert Hit,enip furnace heated
by silicon carbide elements and equipped nit9h an automatic electronic temperature control t o maintain a constant temperature up to 2800' F. Combustion Tube and Accessories. A 27-inch zirconium oxide tube with a 10-inch restriction is oreferable. The inside diameters are 1 1 / 1 6 inches in the open poition and 1/8 inch in the restricted portion. These tubes last about 3 weeks with continuous daily use. The greatest care must be exercised in removing or replacing the stopper t o prevent any strain on the combustion tube, as t'hey are easily broken a t the operat'ing temperature of 2600" F. It is preferable to leave the clamps on either end slightly loose. A perforated refractory disk is placed inside the combustion chamber against the restrict'ed portion and 10 to 15 grams of a coarse refractory-type filter material (H. W. Dietert Co.) is pushed against this. The tube is placed in the furnace so that the filter is in a temperature zone of about 2000" to 2200" F. when the hot zone is set at 2600' F. The combustion boats have the following inside dimensions: length 31/8 inches, width 5/16 inch, height g/32 inch. It is necessary to ignite them for a few minutes before they can be used, as they contain a trace of sulfur. They are stored in clowd containers, since they will adsorb sulfurous gases from the air. Their rost is small and they can be used several times. The Bubbler is a glass tube open a t the entrance and perforated a t the exit with sis or eight small holes. The Absorption Vessel is a cylindrical vessel with a diameter of 1.25 inches and height of about 7.5 inches. The capacity is approximately 200 ml. Titration Assembly is made from a long 15-nil. buret graduated to 0.05 ml. for the potassium iodate solution antl a 2-nil. buret for the sodium thiosulfate. Balance. lccuracy of about i 0 . 0 5 mg. is required. PROCEDURE
The furnace temperature is set a t 2600" F. A 2OO-mesh sample of 20 to 100 mg., depending on the sulfur content,, is spread evenljin the boat, using no bedding. If the sulfur is over 20%, 20 nlg. of sample is used; if betmen 3 and 20%, 40 mg. is used; and if under 3%, 100 mg. is used. Then 65 ml. of 0.2 K hydrochloric acid is dispensed into the receiving flask and 4 ml. of the potassium iodide-starch solution is added. The oxygen flow is adjusted to 1100 ml. per minute and a little iodate is added to the solution to produce a faint blue. About 50% of the expected titer is now added. The sample is then pushed int'o the hot, zone with no preheating and the stopper is quickly replaced. The standard iodate is added quickly enough so that the starch remains a fairly deep blue until the end point is near. The sulfur dioxide comes off very quickly from some samples, and in order to avoid low results an excess of iodine must be niaint'ained. When the reduction of the excess iodine has ceased, the thiosulfate solution is used to back-titrate to a faint blue color. Only a drop or two is necessary if the addition of iodate is carefully controlled. The oxygen flow is sufficient to keep t,he solution adequately stirred.
TYGON TUBING OUTLET TO BUBBLER
-OXYGEN
SUPPLY
ERFORATED DISC
MEAT DEFLECTOR
Figure 1.
Conibristion Tithe and iccessories
1383
The same absorbing solution can be used for two or three samples w e n of high sulfur content without decreasing the accuracy. This rather unorthodox procedure saves considerable time in routine determinations and if the back-titration is taken to the point where one more drop will decolorize the solution, it is not
1384
ANALYTICAL CHEMISTRY
Table I.
Determination of Sulfur in Samples Containing Barium Sulfur, %
Sample Jig concentrate
Approx. Barium Content, 70
Ore
Combustion method 18.9 18.9 19.0 17.9 17.8 17.9 16.3 16.2 16.4 15.7 15,s 15.6 15.4 15.5 18.7 18.8 18.8
25
15
Lead concentrate
1
Lead concentrate
1
Lead concentrate
1
Ore
2
Fusion method 18 7
17.9 16 4 15.7 15 4
19 0
difficult after a little practice to detect the end point Kith varying volumes of absorbing solution. As residue accumulates in the cooler portion of the combustion tube, the yield of sulfur dioxide is reduced and the empirical iodate factor must be adjusted accordingly. It is advantageous to burn out this accumulation every morning if the equipment is in continuous daily use. The delivery tube is disconnected and the cooler portion of the combustion tube is moved to the hot zone foi 15 or 20 minutes while the oxygen is flowing. After this operation, three or four standard samples are ignited until the yield of sulfur dioxide is again stable (see discussion). The filter material is changed after about 400 determinations. Samples to be tcstrd are grouped according to sulfur content and the factor on thc iodate solution is checked for each rlass of sample before they are run.
with 6 inches of filter material. With this type of tube the yield of sulfur dioxide is about 5% less and the empirical iodate factor was adjusted accordingly. DISCUSSION
In order to measure sulfur in the sample i t is necessary to convert i t to sulfur dioxide. There are two types of reactions to be considered: (1) the oxidation of sulfides or of elemental sulfur; and (2) the decomposition of the sulfates, sulfites, etc., in the Eample. Some sulfides may oxidize to a mixture of oxides and sulfates during the combustion, and the sulfates will then have to be decomposed to liberate sulfur dioxide. During the combustion of galena, the evolution of sulfur dioxide is rapid a t first but then proceeds a t a much slower rate, indicating the formation of basic sulfates. Sphalerite concentrates oxidize rapidly and the end point is more quickly reached.
Table 111. Determination of Sulfur in Miscellaneous Flue Dusts Sulfur, % Sample 1
Combustion method 3.5 3.6 3.6
Wet method 3.7 3.7
2
9.4 9.4 9.5
9.4 9.2
3
13.2 13.2 13.1
13.0 13.0
4
8.2 8.2
8.3 8.3
The decomposition temperatures of most sulfates are below RESULTS
2600" F. (S), and the presence of metals whose sulfates decompose
Table I shows a few typical results obtained from samples containing barium. The results of the fusion method were obtained by treating 0.25 gram of the sample as outlined by Low, Wcinig, and Schoder (6). Decomposition by wet oxidation (6) was used to obtain the values of the wet method in Tables I1 and 111. The results in Table I11 listed under combustion method were obtained by using a combustion tube with a short restriction packed
a t relatively low temperatures will help decompose the more refractory sulfates (a). The evolution of sulfur dioxide is usually complete after about 3 minutes, but metallurgical samples containing barium sometimes require 10 to 15 minutes (Table 111). Because sulfur trioxide is not measured by the potassium iodate, it is necessary to set up the physical conditions in such a way as to favor the production of sulfur dioxide. Factors to be considered are temperature, type of tube, oxygen flow, rate of cooling of the evolved gases, size of sample, etc. Sulfur dioxide oxidizes to sulfur trioxide in the presence of oxygen:
Table 11.
Determination of Sulfur in Ores, Mill Samples, and Concentrates Sulfur, %
Sample
Combustion method 16.0 16.0 15.9
Wet method
Zinc sulfide concentrate
29 9 29 8 29 9
29.7 29.5
Zinc sulfide concentrate
31.6 31.5 31 5
31.5 31.2
Galena Concentrate
12 2 12 2 12 3
12.4 12 2
Galena concentrate
16.7 16.8 16 8
16.9 16.9
Mill feed
15.0 14.7 15.0
15.0
1 9 I .8 1.8
1.9
Ore
Flotation tailings
15.7 15.8
so2 (g) f
'/zOz( 9 ) = sos (g)
As the mole fraction of oxygen in the combustion tube is near unity, the equilibrium equation of this reaction can be reduced to [sox' - - K . The theoretical equilibrium constants a t various [SO21 temperatures can be calculated from free energy data (a), and the above relation can then be used to estimate the approximate theoretical composition a t various temperatures (Figure 2). The theoretical yield of sulfur dioxide a t 2600" F. is about 98%, compared with 96% obtained in practice. An operating temperature in the horizontal portion of the curve in Figure 2 is necessary to obtain consistent results. I n order to effect a rapid decomposition of relatively refractory sulfates often found in smelter samples, a temperature of 2600 O F. has been adopted. If the gases are evolved very quickly, they will tend to form a pocket of low oxygen content in the gas stream. I n this case the production of sulfur dioxide is favored and the theoretical yield is increased. It has been found that 40 mg. of sphalerite concentrate has a slightly higher percentage yield of sulfur dioxide than 20 mg of the Same sample. The rate of evolution of sulfur diox-
V O L U M E 25, NO. 9, S E P T E M B E R 1 9 5 3 ide is considerably higher from the larger sample. However, the difference in yield is not great enough to interfere with the reliability of the method. Using the size of sample recommended, the iodate factors are nearly the same for various types of metallurgical samples containing up to about 35% combined sulfur. A 5mg. gample of pure sulfur burns very rapidly and will yield more sulfur dioxide than 20 mg. of a sample containing 25% combined sulfur, for example. Samples containing about 70% free sulfur have been tried and results were consistent enough to indicate that the combustion mzthod could be adopted for samples containing free sulfur. Addition of material such as preburned alundum to slow the combustion is advantageous. Ammonium sulfate decomposes rapidly in the combustion furnace and the percentage yield of sulfur dioxide is slightly higher than from metallurgical samples containing combined sulfur. However, it does provide a stable standard to check oxidation or other deterioration of the standard samples. Surface adsorption of sulfur dioxide inside the evolution apparatus has an important effect and it must be fairly closely controlled to produce accurate results. There is a very large potential adsorbing surface area in the combustion tube, filter material, and the delivery tube, and this surface must be saturated with sulfur dioxide prior to making any determinations. Metallurgical samples ignited in the manner described evolve a dust which tends to adhere to exposed surfaces, and a slaglike material br hich condenses in the cooler restricted portion of the combustion tube This unwanted accumulation provides a large surfase area on which sulfur dioxide will oxidize to sulfur trioxide and react nith metallic oxide to form sulfates. lfost of this loss of sulfur dioxide occurs in the restriction, where the temperature is high enough to promote a reaction and where the condensed slag continually catches more particles to form new surfaces. JVhen the filter and combustion tubes are new, the yield of sulfur dioxide remains fairly constant until residue begins to accumulate in the tube. If a residue has accumulated in the restriction, the yield drops after the furnace has been out of use a feTv hours. After the accumulation is burned out as described in the general procedure, the yield recovers to near the original value. As a considerable quantity of sulfur dioxide is evolved when the cooler portion of the tube is moved into the hot zone, it is evident that it is necessary to prevent any change in the relative position of the tube during a series of determinations. Khen the furnace is in continuous operation the yield will re-
1385 main constant. A series of ten determinations was made on a sample of zinc sulfide concentrate containing 31.9% sulfur, using a combustion tube and filter which had been used for several hundred determinations. The mean titration was 12.70 ml., the mean deviation was ~ t 0 . 0 2 ,and the maximum deviation was 3~0.04. The iodate factors determined from standard samples a t hourly intervals during daily routine determination show the same order of reproducibility. Because of this precision that can be attained, it is possible to keep an accurate check on the condition of the furnace. The long restriction in the combustion tube has the advantage of offering a minimum surface area exposed to the sulfur dioxide and of providing a sharp temperature gradient in the gas stream, thus quickly cooling the gases to a temperature where little reaction will occur. The restricted portion of the combustion tube can IF rrater-cooled, using a spiral of copper tubing ( 1 ) . .-1 combustion tube restricted just at the exit has been used .$bout 6 inches of this tube was packed with filter material, so that the dust would be caught in the hot zone. In this case consistent results were obtained, but the yield of sulfur dioxide was about 5% less than obtained n-ith the tube with the long restriction (Table 111). Experiments were conducted to determine the loss of iodine from the absorbing solution. -4second abporption vessel containing a slight excess of standard thiosulfate was connected in series with the first. Some iodine loss wcurs from the first s o h tion when a large excess of iodate is added. K i t h 5-ml. excess of iodate in the solution and an ovygen flow of llC0 ml. per minute, the loss of iodine is equivalent to 0.25 ml. of the standard iodate or 0.00012 gram of sulfur in 5 minutes, and if the oxygen flow is increased to 2000 ml. per minute the loss is increased threefold. In practice only about one half the expected titer is added and most of the excess iodine is reduced in a few seconds, as the initial evolution of sulfur dioxide is rapid. I n this case the loss of iodine is slight and thP empirical factor for the iodate will correct for it The oxygen flow was varied to determine the optimum rate The combustion of the sample is appreciahly slower at 500 nil per minute, and no compensating advantage was observed. At 2 liters perminute, the evolved gases are a t a temperature of about 500’ F when they reach the plastic delivery tube, and the loss of iodine from the absorbing solution is appreciable. At 1100 ml. per minute, the temperature of the gas entering the plastic delivery tube is only 200” F., the loss of iodine from the absorbing solution is not serious, and the faint blue end point is relatively permanent. The iodide slcfwly oxidizes to free iodine and the blue color will deepen slightly in 15 or 20 minutes. Reports indicate the rate of oxygen flow on steel samples ( 1 ) to be much more critical than on smelter samples. ACKNOWLEDGblERT
The author wishes to thank The Consolidated Mining and Smelting Co. of Canada, Ltd., for permission to publish this report, Joe Morris, supervisor, Analytical Laboratories, for his encouragement and assistance, and members of the laboratory for their cooperation in the wet dptprminations. LITER4TURE CITED
(1) Division of Analytical Chemistry, AMERIC~S CHEMICAL SOCIETY, ASAL. CHEM.,24,203 (1952). (2) Hillebrand, W. F., and Lundell, G . E. F., “dpplied Inorganic bnalyses,” New York, John W-iley & Sons, 1929. (3) Hodgman, C. D., “Handbook of Chemistry and Physics,” 30th
ed., Cleveland, Ohio, Chemical Rubber Publishing Co., 1947.
TEMPERATURE (ABSOLUTE)
Figure 2. Approximate Theoretical Yield of Sulfur Dioxide at Equilibrium with Sulfur Trioxide in a Large Excess of Oxygen
(4) Hougen, 0. A,, and Watson, K. &I., “Chemical Process Principles,” Part 11, “Thermodynamics,” Chap. XVI, Fig. 156, New York, John Wiley &Sons, 1943. ( 5 ) Low, 4.H., Weinig, A. T., and Schoder, R. P., “Technical Methods of Ore Analysis.” New York, John Wiley & Sons, 1939. RECEIVEDfor review November 10, 1952.
Accepted J u l y 1 , 1953.