FLUIDIZED COKE-BED CHLORINATION OF ILMENITES by H. M. Harris, 1 A. W. Henderson, 2 and T. T. Campbell 3 ABSTRACT The Bureau of Mines investigated chlorination of domestic ilmenites in a fluidized coke bed at temperatures of 950° to 1,150° C. Objectives were to develop an effective process for producing TiCl4 suitable for TiO2 pigment and to reduce pollution by collecting and treating waste byproduct chlorides. The reactor for these tests was a 10.2-cm-diam quartz tube heated externally by electrical resistance heaters. It contained coke beds of from 35.6 to 82.5 cm in depth that were fluidized by 13 to 25 1/min of chlorine. Ilmenite concentrate powders were introduced at the bottom of the fluid-bed reactor, using the chlorine gas stream as a carrier. Test results showed that greater than 90 pct of the ilmenite reacted at optimum conditions and recovery of Tic14 ranged from 95 to 99 pct of stoichiometric. A temperature-controlled cyclone was used to remove high-boiling-point chlorides such as MnCl2, MgCl2, and FeCl2, and also collected fine unreacted ilmenite and coke dust. Impure TiCl4 and FeCl3 were separated by collecting the iron chloride in a condenser controlled at 200° C. Titanium tetrachloride vapor was cleaned by passage through a salt-column filter followed by condensation in water-cooled and refrigerated condensers. Liquid TiCl4 was purified further by well-known industrial methods to meet specifications for pigment-grade chloride. Inte grated tests were conducted in which chlorine, produced by dechlorination of FeCl3 with oxygen, was recycled and used directly in a fluidized-bed chlorinator. INTRODUCTION An effective economic process is needed to make titanium dioxide pigment from ilmenite (FeTi03) because of a pending worldwide shortage of rutile (TiO2), and also because of the pollution and waste generated from current industrial processes. Production of titania pigment from ilmenite by the sulfuric acid process generates waste sulfuric acid containing iron and other sulfates, which causes a water pollution problem when disposed of on land or in the sea (8-10).+ Waste iron chlorides must be disposed of when titania Supervisory chemical research engineer. *Underlined numbers in parentheses refer to items in the list of references at the end of this report. pigment is produced by the chlorination process. Processes that make rutile substitutes are of interest for resources far distant from a market and for the supply of existing chlorination plants (3, 7, 11-12, 15). These methods add to costs and may not eliminate pollution problems. A direct conversion process for ilmenite to pigment is needed that recovers and recycles the major chemical reagents with little soluble waste and with usable byproducts. Previous research showed the feasibility of dechlorinating iron chlorides at 500° C using a fluidized bed treated with a NaCl catalyst to recover chlorine for recycling and to generate a nonpolluting iron oxide byproduct (6), thus enhancing the economics of the direct chlorination of ilmenites where iron chlorides are a major waste byproduct. In the United States, where supplies of ilmenite are available close to markets, the direct chlorination of ilmenite should provide an economic process for producing TiO2 pigment. when coupled with recycling Cl2 from iron chlorides. Existing methods used to convert rutile to TiO2 pigment are probably not usable with the more complex ilmenite minerals (10, 23). A variety of methods have been used to chlorinate ilmenite, but none of these methods have been fully evaluated (1, 14-15, 17, 21-22). Data on kinetics (4) and thermodynamics of reactions involved in ilmenite chlorination indicate that the use of C and Cl2 together is an effective means for converting oxides to chlorides (13, 20). Industrial success with use of C and Cl2 for chlorination of rutile (10, 23) fortifies this conclusion, as does the production of pigment-grade TiO2 from enriched ilmenite by E. I. du Pont de Nemours & Co. as disclosed in various patents (25). The Du Pont process is economic because it produces a significant portion of the TiO2 pigment consumed even though it is an incomplete process without recovery of chlorine from iron chloride wastes. A pilot plant dechlorinator is reportedly under construction in Tennessee (2, 16, 24). Fluidized-bed chlorination of ilmenite is difficult because the highboiling-point chlorides produced from iron and impurities tend to clog reactor systems. Equation 1 is a general formula for the chlorination of ilmenite. The indefinite moles of C and Cl2 denoted by (x) and (y) result from the wide differences in ilmenite ore compositions (33 to 66 pct Ti02), and variations in FeCl3 -FeCl2 and CO-CO2 ratios with process conditions. FeO TiO2 + (x) Cl2 + (y) C-FeCl3/FeCl2 + CO/CO2+ TiCl4. (1) Most ilmenite ores differ from the idealized mineral formula, and table 1 gives melting and boiling points of most of the compounds formed during chlorination. Most of the chlorine used for chlorination of rutile or enriched ilmenite is recovered for recycling by oxidation of TiCl4 to form titania pigment as shown in equation 2. 1 Chemical Rubber Co. Handbook of Chemistry and Physics. When chlorinating ilmenite, recovery of chlorine for recycling from iron chloride and other impurities as well as TiCl4 is an economic and ecological necessity (10). A possible method for Cl2 recycling, investigated by Henderson (6) and Paige (19) in related Bureau of Mines research, involved the dechlorination of FeCl3 in a heated fluidized bed as shown in reaction 3. 2 FeCl3 + 3/2 02 Fе203 + 3 C12. (3) Reaction 1 was evaluated by estimating the mole-fractions of products at conditions of interest. Equilibrium calculations based on thermodynamic and phase data were made for the reaction as a function of reaction temperature. Results for two ilmenitechlorine ratios are shown in figures 1 and 2. A carbon-ilmenite mole ratio of 3.8 was used to provide carbon in excess of that needed for all possible reactions in a coke-bed reactor. The moles of carbon remaining are not shown. Although equilibrium was not reached in the fluidized-bed chlorinator, it was closely approached when finely powdered ilmenite and coke were reacted with chlorine at high temperatures. tion with the higher Opera ilmenite-Cl2 ratio of 0.7 as shown in figure 1 would impose the problem of freezing the bed with liquid ferrous chloride FeCl2 (1) or condensed solids (1,025° C boiling point) unless very high offgas temperatures and gas velocities were maintained. High-boiling-point chloride impurities also add to these difficulties. Operation at the lower ilmenitechlorine ratio of 0.6 as shown in figure 2 is preferable with some excess Cl2 in the offgas because this would produce a high concentration of FeCl3, which has a higher vapor pressure (boiling point 315° C) than FeCl2. Also important is the high concentration of carbon monoxide reductant present in the temperature range of operation (900° to 1,100° C). Carbon monoxide allows rapid gas-solid reduction rather than the slower solid-solid reaction of carbon and ilmenite. Stoichiometric calculations for converting rutile and domestic ilmenites to TiCl4 and FeCl3 are shown in table 2. These data illustrate the increased chlorine required for ilmenite chlorination. The low cost of ilmenite compared with that of rutile (approximately 1 to 10) is a favorable economic factor for the direct chlorination process. TABLE 2. - Ore and reagents required to produce 1 kilogram of TiO2 pigment and byproducts from rutile and ilmenites1 Ilmenites. 2 1.9 1.7-3.2 2.7-4.9 .3 -0.6 1.00-3.9 0.02 .70-2.6 With stoichiometric reaction of all elements in the ore to form chlorides and of all iron to form FeCl3 from Idaho, Florida, New York, and New Jersey ilmenites and Australian rutile. Coke reacted for an offgas of equal CO and CO2 (volume-percent). A flowsheet of the coke-bed chlorination process being investigated by the Bureau of Mines is given in figure 3. As shown, the main products are Fe2O3 and TiO2 pigment with nearly complete recycling of chlorine from iron and titanium chlorides. Features of the new process (5) that provide ease of operation and clean byproducts are conditions that limit troublesome buildup of ilmenite or chlorides in the bed; for example, controlled feed of reactive finely ground ilmenite, and use of gas velocities that carry high-boiling-point chloride dusts out of the bed into a collecting cyclone. Maintaining a clean bed is aided by using self-cleaning coke particles (18) rather than ore, and the cyclone cleans the chloride vapors before condensation. A ferromanganese byproduct, ilmenite, and coke for recycling might be recovered by further treatment of the cyclone dusts. Other chlorination methods were considered, but they lacked the advantages of simplicity, ease of control, and superior reaction afforded by a bottom-fed, fluidized coke bed. Research on the dechlorina tion of iron chlorides is covered in previous reports (6, 19) and is not included here except for the discussion of integrated chlorination-dechlorination tests. Some advantages of a bottom-fed, fluidized coke bed were shown by the ease of operation and the high product recoveries at 950° C in previous chlorination tests in a 5.1-cm reactor (5). Other advantages are maximum reduction efficiency and accurate control of the ilmenite-chlorine ratio, and therefore of the FeCl3 -FeCl2 ratio. This is not always possible with bed mixtures containing ilmenite in excess of reaction needs. The use of finely ground ilmenite and continuous monitoring of chlorine is necessary to prevent overfeeding and loss of control of the FeCl3 -FeCl2 ratio. Also, finely ground ilmenite should provide finely divided high-boiling-point chlorides or oxides that carry over into cyclones, thus preventing clogging of the bed or crossResearch reported herein was done to determine results in a larger reactor system, at temperatures higher than 950° C, and with cyclones to clean over. chloride vapors. ACKNOWLEDGMENT Professor John R. Riter of the University of Denver, Denver, Colo., under à cooperative program, determined the equilibrium products for the reaction of ilmenite, chlorine, and carbon. equilibrium values based on thermodynamic calculations and phase equilibria. A computer program was used to determine |