Deactivation of Pyrophoric Iron Sulfides - Industrial & Engineering

Sep 2, 1997 - The pyrophoric nature of iron sulfides, which could account for the minor .... series of endothermic and exothermic peaks was observed, ...
25 downloads 0 Views 186KB Size
3662

Ind. Eng. Chem. Res. 1997, 36, 3662-3667

Deactivation of Pyrophoric Iron Sulfides Robert Walker* Department of Materials Science and Engineering, University of Surrey, Guildford GU2 5XH, U.K.

Alan D. Steele and David T. B. Morgan Shell Research and Technology Centre, Thornton, P.O. Box 1, Chester CH1 3SH, U.K.

Rust formed by corrosion on the inner surfaces of oil tanks may react with hydrogen sulfide gas in the vapor above crude oil. This reaction produces mixtures of iron sulfides which are pyrophoric, and they cause explosions when they are exposed to air during the unloading of the tankers. This paper shows that it is possible, by exposure to an atmosphere of nitrogen containing 2 or 4 vol % oxygen, to deactivate these sulfides with mackinawite, FeS, pyrite, FeS2, and greigite, Fe3S4. Mackinawite undergoes slow oxidation to give goethite, FeO(OH), which has a lower reactivity. Pyrite and greigite are pyrophoric, present in the mixture in a finely divided state and produce sparks on exposure to air. During deactivation they do not oxidize but transform to a framboidal morphology which has a lower surface area and is less reactive. following equations:

1. Introduction The pyrophoric nature of iron sulfides, which could account for the minor explosions in crude oil tankers in Thailand, has been described in an earlier paper (Walker et al., 1996). It is considered that the Qatar oil being carried had a high vapor pressure and contained up to 5 vol % hydrogen sulfide which was present in the vapor and reacted with rust in the tanks to give these dangerous sulfides. It is possible to overcome this danger on ships by doing the following: (1) equipping the vessel with an inert gas plant, but this is expensive; (2) spraying the deposits with water but, although this has been shown to be effective in the laboratory, it is difficult to achieve under practical conditions; (3) compaction and oil saturation of iron sulfides to prevent pyrophoric oxidation. Unfortunately, in practice, cargo pressing is likely to create both safety and environmental problems which could arise from the release of hydrocarbon vapors above deck and the possibility of minor deck oil spillage; (4) the preoxidation of any iron sulfide deposits which is a real possibility, and so it is reported in this paper. 2. Literature Review The principal constituent of the rust scale on ships is goethite {RFe0(0H)} (Hughes et al., 1976): this is also found on carbon steels in humid tropical climates together with lepidocrocite {γFeO(OH)} (Haces et al., 1981) which gradually changes to goethite with increasing exposure to humid rural atmospheres (Ases et al., 1982). The corrosion of plain steel in refineries has also been reported to give iron oxyhydroxides and pyrite (Peev et al., 1992). Goethite has been shown to react with hydrogen sulfide to form different iron sulfides depending upon the environmental conditions (Walker et al., 1987). Mackinawite, FeS, is favored at low temperatures with relatively little hydrogen sulfide. At higher temperatures greigite, Fe3S4, and pyrite, FeS2, are formed with pyrite being favored at the elevated temperatures and with more gas; the reactions may be represented by the * Author to whom all correspondence should be addressed. Telephone: +44 (0)1483 300800, ext. 2408. Fax: +44 (0)1483 259508. E-mail: [email protected]. S0888-5885(96)00575-1 CCC: $14.00

2FeO(OH) + 3H2S f 2FeS + S + 4H2O

(1)

3FeS + S f Fe3S4

(2)

FeS + S f FeS2

(3)

Knowledge of these sulfides, however, is relatively limited. Mackinawite was first identified in 1964 from the reaction of goethite and aqueous sodium sulfide followed by aging which gave a black amorphous product, probably poorly crystalline mackinawite (Berner, 1964). The earliest reference to greigite was also in 1964 when it was found as tiny grains and crystals in certain clay layers in California (Skinner et al., 1964). These sulfides can be pyrophoric under certain climatic conditions, particularly in warm moist atmospheres. The tendency to oxidize increases considerably as the particle size is decreased and the surface area increased. Iron sulfide species have been found in gas lines and chemical plant and distillation units (Hughes et al., 1974). They are also produced by the corrosion of mild steel in wet elemental sulfur and can cause spontaneous combustion of the surrounding sulfur (McDonald et al., 1978). Popa et al. (1992) used X-ray diffraction to prove that mackinawite is formed in the Girdler sulfide process when alloy steels react with hydrogen sulfide: this tended to transform into other phases and they found that it was desirable to convert the mackinawite on the surface of the steel into pyrite. The reactions of sulfides in coal have been studied (Krs et al., 1992; Ge et al., 1992) and are known to promote the ignition of coal deposits (Lomax, 1925). Pyrophoric oxidation can occur when the sulfide is sufficiently finely divided and stored in bulk to which air has access, so Bowes (1954) suggested the introduction of an inert gas, carbon dioxide, into storage hoppers to reduce the fire hazard. Water plays an important role in these processes involving mackinawite, and it has been suggested that ferrous sulfides are capable of undergoing pyrophoric oxidation only in a moist atmosphere (Mellor, 1942). At a relative humidity below 50%, 3 g of the dried mackinawite-sulfur mixture produced by the sulfidation of goethite, according to reaction 1, oxidized at 20 °C in a slow controlled manner but at a value above 50% it was © 1997 American Chemical Society

Ind. Eng. Chem. Res., Vol. 36, No. 9, 1997 3663

pyrophoric (Walker et al., 1988): larger samples always showed pyrophoricity. Mackinawite with sulfur is considered to undergo pyrophoric oxidation according to the following reactions:

4FeS + 4S f 4FeS2

(3)

4FeS2 + 11O2 f 2Fe2O3 + 8SO2 4FeS + 2H2O + 7O2 f 4FeO(OH) + 4SO2 and the overall reaction is

8FeS + 4S + 2H2O + 18O2 f 4FeO(OH) + 2Fe2O3 + 12SO2 (4) For small (0.5 g) samples of mackinawite, however, the heat dissipated to the surroundings is far greater than can be supplied by the oxidation reaction. Hence, regardless of the atmosphere humidity, the process was controlled, nonpyrophoric, and appears to follow the route

4FeS + 2H2O (atm) + 3O2 f 4FeO(OH) + 4S The elemental sulfur produced was shown to be orthorhomic because it was soluble in carbon disulfide. The decomposition of pyrite has also been investigated. The surface oxidation of coal-derived pyrite was influenced to a greater extent by the relative humidity than by the partial pressure of oxygen (Baltrus and Diehl, 1994). Electrochemical studies have shown that the oxidation of pyrite involved the chemisorption of the hydroxyl group from water and, in contact with oxygen, the process was rapid (Wang et al., 1992). Similarly the relative humidity had an effect on the weathering of pyrites to give ferrous sulfate (Borek, 1994). Pyrolytic decomposition of pyrite has been reported to form very porous pyrrhotite (Dunn and McKay, 1993) while Shukla and Singh (1992) found that the oxidation rate of sulfur in pyrite decreased as the particle size increased at 2627 °C. Hence it may be concluded that the rate and nature of the oxidative reaction of iron sulfides depend upon the amount of water present, the concentration of oxygen and the size and mass of the sample. The stability of finely divided mackinawite is indicated by the experimentally determined heat of oxidation of -7.45 kJ g-1: the value, however, is lower for samples with a larger particle size and corresponding smaller surface area due to incomplete oxidation (Walker et al., 1996). The aim of this paper is to investigate the reactions involved in the controlled oxidation and deactivation of the different, pyrophorie iron sulfides. 3. Experimental Procedure The rust scales collected from the insides of oil tanks were heterogeneous, but the main constituent was always goethite, RFeO (OH), usually with elemental sulfur and occasionally magnetite, Fe3O4, ferric oxide, RFe2O3, and sometimes hydrated ferrous sulfate, FeSO4‚ 7H2O. Because the major constituent was goethite, it was logical to use a pure form of this compound as a starting component to examine the reactivity toward hydrogen sulfide and the chemical species formed from this reaction. The standardized pure form of goethite, the Riedel de Haen product of approximately spherical particles

of size 0.95 µm was chosen as the initial material. About 30 g of this geothite was pressed into a cylindical plug and put into a U tube of diameter 2.5 cm with glass wool plugs and taps at both ends. This was fitted into a specially designed apparatus, flushed with pure nitrogen to remove the air, and then filled with hydrogen sulfide gas. Sulfidation continued, and gas was passed through the tube and plug at a rate of 0.5 cm3 s-1 at ambient temperature. The resulting mixture has been shown to consist of mackinawite, sulfur, and water according to the following equation:

2FeO(OH) + 3H2S f 2FeS + S0 + 4H2O This procedure was repeated at 100 °C and gave a different product, viz. pyrite (approximately 40 wt %), greigite, sulfur (approximately 5 wt %), and watersthis combination was extremely reactive (Walker et al., 1987). These mixtures, produced at different temperatures, were separately exposed to an atmosphere of either 2 or 4 vol % oxygen in nitrogen for a time period of 6 h. It had previously been suggested that this procedure would give controlled oxidation i.e., conditions which enable the sulfide to oxidize in a nonpyrophoric manner (Hughes et al., 1976). Because of the high reactivity of these chemicals, many precautions were necessary. The experiments were carried out in a glovebox fitted with both a nitrogen and an air supply, the atmosphere being maintained by adjusting the flowmeter: the oxygen content was monitored using a Taylor Servomex oxygen analyzer. The temperature of a sample (approximately 20 g) of the mixture was measured by a thermocouple inserted into the bulk sulfide: the relative humidity was less than 20%. In order to find out which reactions were occurring, it was necessary to determine the exact nature of the substances present before and after reaction. Hence it was necessary to measure quantitatively the concentrations of iron and sulfur as well as qualitatively the compounds and oxidation states. This was achieved using X-ray diffraction (XRD) to ascertain the presence of compounds. The Debye-Scherrer powder technique was employed with cobalt K radiation and an iron filter. The identification was carried out by a comparison of the XRD spacings of the mixture with those obtained for the separate sulfides which were produced by the methods of Berner (1967). X-ray photoelectron spectroscopy was used to identify the oxidation states of the iron, sulfur, and oxygen species. The concentation of iron was measured following the ASTM Designation E277-69 “Total Iron in Iron Ores by Stannous Chloride Reduction and Dichromate Titration” and the sulfur with Vogel’s Quantitative Inorganic Analysis as “The Determination of Sulfur from Iron Ores”. It is fortunate that pyrite can be distinguished from mackinawite and greigite because it is insoluble in hydrochloric acid. The form of sulfur could be ascertained because the orthorhombic form is soluble in carbon disulphide, but the amorphous type is not. Differential thermal analysis (DTA) of the deactivated product was carried out with a Stanton Redcroft Model 674 to help in the understanding of the results. Because the plot consisted of several exothermic and endothermic peaks it was necessary to identify the different chemical processes involved. Hence DTA was performed on the separate chemicals involved including elemental orthorhombic sulfur, water, goethite, etc., and the peaks were used for comparison.

3664 Ind. Eng. Chem. Res., Vol. 36, No. 9, 1997 Table 1. Changes Observed When Sulfide Mixtures Were Exposed to N2/O2 Atmospheres vol % oxygen

rise in temp (°C)

temp maintained

complete deactivation

other

mackinawite

mixture

2

30-35

3

x

mackinawite pyrite/greigite

4 2

40-45 30

2 1

x x

black color f dark brown moisture content 20 f 8 wt %, in 6 h no change in composition

Figure 2. Differential thermal analyses of the deactivated mackinawite mixture.

Figure 1. X-ray photoelectron spectroscopy of the deactivated mackinawite mixture.

Photographs of the powders were taken on a scanning electron microscope modified to minimize any exposure of the sample to oxygen.

Differential thermal analysis of the deactivated product under an oxygen atmosphere confirmed that some mackinawite, identified by X-ray diffraction and chemical analysis, was still present. For a 0.123 g sample of the deactivated sulfide product the heat of oxidation was determined by differential thermal analysis as -423 J, i.e., -3.44 kJ g-1. The heat of oxidation of the sulfide prepared from the Riedel de Haen goethite has previously been measured (Walker et al., 1996) as -5.37 kJ g-1. Hence, as expected, there is a significant reduction in the heat of oxidation during the deactivation process. It would appear from the analysis that the deactivation process or controlled oxidation occurred according to the following equation:

4. Results and Discussion The results are shown in Table 1. Mackinawite Mixture in 2 Vol% O2 for 6 h. This experiment produced complete deactivation because, when the product was exposed to air, there was no detectable increase in bulk temperature. The position of the lines on the X-ray diffraction photograph of the product were compared with published values: these indicated the presence of goethite, orthorhombic sulfur, and some remaining mackinawite (Berner, 1964). X-ray photoelectron spectroscopy showed that there were two forms of iron, and the bonding energies of the species proved that it was mainly Fe3+ and only a slight indication of Fe2+ (Figure 1). Two forms of sulfur were identified as sulfur S0 and sulfide S2- and oxygen, characteristic of RFeO(OH), was also found (Figure 1). Chemical analysis of the deactivated product gave a total sulfur/iron molar ratio of 1.51/1: after extraction of the soluble orthorhombic sulfur in carbon disulfide this ratio became 0.67/1.

4FeS + 2H2O + 3O2 f 4FeO(OH) + 4S Analysis indicated that the amorphous elemental sulfur formed initially was apparently slowly transformed into the solvent soluble orthorhombic form. The deactivated product was also studied using the differential thermal analysis equipment with an atmosphere of nitrogen. An interesting series of endothermic and exothermic peaks was observed, so further work was carried out to identify the reactions. The following differential thermal analysis peaks (Figure 2) were found. (1) 82 °C: endothermic peak due to the latent heat of evaporation of sample water. (2) 116 °C: endothermic peak due to the latent heat of fusion of orthorhombic sulfur. (3) 170 °C: exothermic peak due to the reaction between mackinawite and amorphous elemental sulfur to form pyrite,

Ind. Eng. Chem. Res., Vol. 36, No. 9, 1997 3665

FeS + S0 f FeS2

(3)

(4) 235 °C: endothermic peak due to the dehydration of goethite to give RFe2O3,

2FeO(OH) f Fe2O3 + H2O (5) 270 °C: exothermic peak due to the reaction between mackinawite and orthorhombic elemental sulfur to form pyrite. The peak at 170 °C and also at 270 °C are both attributed to the formation of pyrite from mackinawite. Therefore another DTA experiment was carried out with the original product to which excess orthorhombic elemental sulfur had been added, and this was heated under a nitrogen atmosphere. X-ray diffraction of the products formed at 170 °C identified pyrite, mackinawite, and orthorhombic sulfur while above 270 °C only pyrite and orthorhombic sulfur were present. It can be concluded that the controlled oxidation process decreased the enthalpy of oxidation and produced relatively stable goethite, orthorhombic sulfur, and some remaining mackinawite and water. Mackinawite Mixture in 4 Vol % O2 for 6 h. Similar conditions as above were used, but with an increase in the oxygen concentration to 4 vol %. Again this experiment produced complete deactivation, Table 1. X-ray diffraction of the deactivated product showed the presence of pyrite which was confirmed by chemical analysis indicating about 12 wt % present. Chemical analysis, after solvent extraction for orthorhombic sulfur, gave a sulfur/iron molar ratio of 0.63/1, which confirms that substantial oxidation of the sulfide had occurred. The presence of the pyrite is interesting. It suggests that a minor reaction took place between the mackinawite and the elemental sulfur:

FeS + S0 f FeS2

(3)

This was probably due to the heating effect during the reaction, which increased the temperature of the bulk mix to 40-45 °C. The observation that the sulfide/oxide mixture containing pyrite was completely unreactive in air is surprising since mixtures with pyrite can be extremely reactive. Hence, the effect of controlled oxidation on a much more reactive initial product of pyrite/greigite was investigated. Reactive Pyrite/Greigite Mixture in 2 Vol % O2. The product of the sulfidation of the Riedel de Haen goethite at 100 °C was used because it gave a very reactive combination of pyrite and greigite which produced sparks instantaneously on exposure to air. As soon as it was prepared, X-ray diffraction was used to confirm the presence of pyrite and greigite: chemical analysis showed that the product contained 41 wt % pyrite. The sulfur/iron molar ratio, after solvent extraction, was 1.63/1. The mixture was exposed to an atmosphere of 2 vol % oxygen for 6 h in exactly the same manner as before. This produced complete deactivation toward air (Table 1). Chemical analysis showed that neither the greigite nor the pyrite underwent any oxidation during deactivation, and the pyrite content was 40.4 wt % with a molar ratio of 1.61:1. This was surprising because these values are very similar to the initial reaction mixture,

so the experiment was repeated and gave similar results with 38.8 wt % pyrite and a molar ratio of 1.62/1. In order to test the stability, the deactivated mixture was exposed to air and the pyrite content analyzed over a period of time. The initial value of 38.6 wt % became 39.2% after 24 h, 37.9% after 45 h, and 36.0% after 69 h. Hence it can be concluded that upon exposure to air the pyrite underwent slow controlled deactivation. During pre-exposure, oxygen was absorbed, and regular infusions of air were required to maintain the oxygen level at 2 vol %. This also occurred during the deactivation treatment and could be due to surface adsorption of oxygen. It can therefore be concluded that the controlled deactivation of this reactive sulfide mix gives no appreciable change in composition. Hence the process must be completely different from that of the controlled oxidation of the mackinawite product. One explanation for this reduced reactivity could be due to a decrease in the surface area produced by the formation of framboidal pyrite. Rickard (1975) suggested that this species is produced by a pathway involving an oxidizing species such as atmospheric air: this idea is supported by Sweeney and Kaplan (1973) who showed that pyrite framboids formed when mackinawite reacted with limited oxygen present in the reaction vessels. Framboidal pyrite has been found to consist of microspheres of diameter about 1 µm by Berner (1969) and as spheroidal aggregates of microcrystallites typically of the order of 10 µm in diameter (Rickard, 1975). These might be expected to be less reactive than non-framboidal pyrite due to the reduced surface area: similar observations have been made by Shukla and Singh (1992) and Gupta and Singh (1984). The deactivated mixture was examined in a scanning electron microscope (SEM) (Figure 3), and appears to be in the form of spheroidal aggregates with no sign of individual particles. Due to the extremely high oxidative reactivity of the original pyrite/greigite mixture, it was very difficult to avoid exposure to air when using the SEM. This was overcome by modifying both the SEM and the gold-coating apparatus to minimize the possibility of sample oxidation by attaching a nitrogen supply to permit a high gas velocity in the vicinity of the sample-loading aperture. The transfer from coating apparatus to the SEM was effected using an airtight bottle flushed out with nitrogen. The SEM micrographs of the powder mixture before deactivation (Figure 3) did show an abundance of individual particles. There were some isolated areas where the sulfide appeared to be aggregated which could possibly be due to sample oxidation. The photographs do show that there is a considerable decrease in the surface area of the powder produced by the deactivation treatment. 5. Significance for Treatment of Oil Tanks in Tankers From the above work it can be concluded that there is the possibility of explosions during the unloading of crude oil known to be sour to some degree, ie containing hydrogen sulfide. The problem can be overcome if the vessel is equipped with an inert gas system (see Introduction), but this involves considerable expense. There is also a risk if the system fails! There is little risk caused by the surfaces below the oil level during discharge even if they are covered with sulfided rust. The reason for this is that one method of deactivating the reactive iron sulfides is oil wetting (see

3666 Ind. Eng. Chem. Res., Vol. 36, No. 9, 1997

to evaporation, but it is likely that the water also participates in the oxidation reaction. The sulfur formed was initially amorphous and transformed into the orthorhombic form that could be solvent extracted. Some pyrite was formed when the higher oxygen concentration was used. The product of the deactivation or controlled oxidation process was unreactive upon exposure to air. The much more pyrophoric mixture of pyrite/greigite which was produced by sulfidation at 100 °C also underwent deactivation during exposure to low concentrations of oxygen. This, however, was by a completely different process from that observed for the mackinawite. The deactivation was accomplished far more quickly, but surprisingly there was no detectable oxidation of either the greigite or pyrite species. It is considered that the loss of oxidative reactivity could be explained by the formation, from a finely divided powder, of framboidal pyrite and greigite with a much reduced surface area. Literature Cited

Figure 3. Scanning electron microscopy of the pyrite/greigite mixture after (a) and before (b) deactivation.

Introduction). The ullage areas, above the oil level, do, however, pose a problem. The recommended procedure before cargo discharge of sour crude oils, based on this and other work, is to measure the oxygen content in these ullage areas of the tanks. Air is then pumped in through gas diffusers, usually up to about 8 vol % oxygen, and then a time period of 2-4 h is allowed for deactivation to occur. The cargo can then be safely discharged. The tanks can be treated after draining by cleaning with sea water to remove any loose scale, and they are then ballasted with sea water for the journey back to the crude oil production location. In addition all the cargo tanks are periodically cleaned with high-pressure equipment to remove most of the rust scale. After several trips with cargo/ballast, the whole cycle of fresh rust generation starts again. 6. Conclusions It has been shown that pyrophoric iron sulfides can be deactivated by exposure to low concentrations of oxygen in a nitrogen atmosphere. Mackinawite underwent slow controlled oxidation according to the following equation:

4FeS + 2H2O + 3O2 f 4FeO(OH) + 4S An increase in temperature accompanied this reaction and lasted for 2 or 3 h. A significant loss of bulk water occurred during this reaction: this could be partly due

Ases, D. K.; Corvo, F.; Peres, R. A.; Macorra, A. H. Steel Corrosion in a Humid Tropical Climate. Zashch. Met. 1982, 18 (4), 58992. Baltrus, J. P.; Diehl, J. B. An Investigation of the Weathering Behaviour of Coal-Derived Pyrite Surfaces by X-Ray Photoelectron Spectroscopy. Fuel 1994, 73, 229-35. Berner, R. A. Iron Sulfides Formed from Aqueous Solution at Low Temperatures and Atmospheric Pressure. J. Geol. 1964, 72, 293-306. Berner, R. A. Thermodynamic Stability of Sedimentary Iron Sulfides. Am. J. Sci. 1967, 265 (9), 773-85. Berner, R. A. Synthesis of Framboidal Pyrite. Econ. Geol. 1969, 64 (4), 383-4. Borek, S. L. Effect of Humidity on Pyrite Oxidation. Environmental Geochemistry of Sulfide Oxidation. ACS Symp. Ser. 1994, 550, 31-44. Bowes, P. C. Spontaneous Heating the Ignition in Iron Pyrites. Ind. Chemist. 1954, 30, 12-14. Dunn, J. G.; MacKay, L. C. T6-DTA Studies of Pyrite, Pyrrhotite, Pentlandite and Violarite Under Ignition Conditions. J. Therm. Anal. 1993, 39, 1255-71. Ge, Y.; Hsieh, K. C.; Tseng, B. H.; Wert, C. A. Pyrite in Coal and Organic Sulfur in its Vicinity. Ranliao Huaxue Xuebao 1992, 20 (1), 90-95. Gupta, V. P.; Singh, O. P. Weathering Effect on Pyrite (FeS2). Indian J. Pure Appl. Phys. 1984, 22 (10), 626. Haces, C.; Corvo, F.; Reyes, A. P. Study of Corrosion Products Forming on Carbon Steels in a Humid Tropical Climate, Part 1. Rev. Cienc. Quim. 1981, 12 (1), 55-65. Hughes, R. I.; Morgan, T. D. B.; Wilson, R. W. Is Pyrophoric Iron Sulfide a Possible Source of Ignition? Nature 1974, 248, 670. Hughes, R. I.; Morgan, T. D. B.; Wilson, R. W. The Generation of Pyrophoric Material in the Cargo Tanks of Crude Oil Carriers. Trans. Inst. Mar. Eng. 1976, 88, 1-7. Krs, M.; Novak, F.; Krsova, M.; Pruner, P.; Kouk-likova, L.; Jansa, J. Magnetic Properties and Metastability of Greigite-Smythite Mineralization in Brown-Coal Basins of the Krusne Lory Piedmont, Bohemia. Phys. Earth Planet. Inter. 1992, 70 (3, 4), 273-87. Lomax, J. Further Researches on the Various Types of Pyrites in Coal, Especially in Relation to Spontaneous Combustion. Colliery Guardian 1925, 129, 1317-8. McDonald, D.; Roberts, B.; Hyne, J. The Corrosion of Carbon Steel by Wet Elemental Sulfur. Corros. Sci. 1978, 18, 411-25. Mellor, J. W. A Comprehensive Treatise on Inorganic and Theoretical Chemistry; Longmans: London, 1942; Vol. 14, pp 157-8. Peev, T.; Taseva, V.; Akala, A. Study of the Corrosion Processes in Primary Oil Refining by Moessbauer and X-Ray Investigation, Part 1, Atmospheric Distillation Columns. Werkst. Korros. 1992, 43 (11), 527-31. Popa, M. V.; Bizadea, G.; Roman, E.; Radovici, D. The Iron Sulfides and Their Role in the Gridler Sulfide Process. Rev. Roum. Chim. 1992, 37 (8), 637-47. Rickard, D. T. Kinetics and Mechanisms of Pyrite Formation at Low Temperatures. Am. J. Sci. 1975, 275 (6), 636-52.

Ind. Eng. Chem. Res., Vol. 36, No. 9, 1997 3667 Shukla, A. K.; Singh, R. S. Oxidation of Sulfur in Pyrites in Relation to Soil and Water Regime. J. Indian Soc. Soil Sci. 1992, 40 (4), 848-50. Skinner, B. J.; Erd, R. C.; Grimaldi, F. S. Greigite, the Thiospinel of Iron: a New Mineral. Am. Mineral. 1964, 49, 543-55. Sweeney, R. E.; Kaplan, I. R. Pyrite Framboid Formation. Laboratory Synthesis and Marine Sediments. Econ. Geol. 1973, 68 (5), 618-34. Walker, R.; Steele, A. D.; Morgan, T. D. B. The Formation of Pyrophoric Iron Sulfide from Rust. Surf. Coat. Technol. 1987, 31, 183-97. Walker, R.; Steele, A. D.; Morgan, T. D. B. Pyrophoric Oxidation of Iron Sulfide. Surf. Coat. Technol. 1988, 34, 163-175. Walker, R.; Steele, A. D.; Morgan, T. D. B. Pyrophoric Nature of Iron Sulfides. Ind. Eng. Chem. Res. 1996, 35, 1747-1752.

Wang, X. H.; Jiang, C. L.; Raichier, A. M.; Parekh, B. K.; Leonard, J. W. Comparative Studies of Surface Properties of Pyrite from Coal and Ore Sources. Proc. Electrochem. Soc. 1992, 92-117, 410-32.

Received for review September 20, 1996 Revised manuscript received May 19, 1997 Accepted May 27, 1997X IE960575Y

X Abstract published in Advance ACS Abstracts, August 1, 1997.