HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Ritvo, G. Articles by Dixon, J. B. Search for Related Content PubMed Articles by Ritvo, G. Articles by Dixon, J. B. GeoRef GeoRef Citation Agricola Articles by Ritvo, G. Articles by Dixon, J. B. Related Collections Wetlands and Aquatic Processes Colloids Geochemical Processes Soil Chemistry

DIVISION S-9—SOIL MINERALOGY

A New Iron Sulfide Precipitated from Saline Solutions

^a Technion-Israel Institute of Technology, Faculty of Agricultural Engineering, IL-32000 Haifa, Israel
^b Dept of Soil and Crop Sciences, College of Agriculture and Life Sciences, Texas A&M University, College Station, TX 77843-2474

^* Corresponding author (j-dixon{at}tamu.edu)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The production of shrimp in ponds is a growing industry and shrimp live in the soil–water interface zone. These and other marine organisms are subject to sulfide toxicity. These investigations were conducted to determine the fate of sulfide (S^2-) in a saline environment like that in which shrimp are grown. A previously unrecognized iron sulfide, which we gave a provisional (unapproved) mineral name, dorite formed during the initial reaction of Fe and S and it may influence sulfide concentration in ponds and transitional marine environments. As the reaction proceeded the 1-nm dorite component converted to mackinawite. We propose a structure for dorite composed of alternating FeS tetrahedral sheets, like mackinawite and defect sheets of the same configuration in which many of the Fe and S atoms are missing preserving electrical neutrality.

Abbreviations: XRD, x-ray diffraction • TEM, transmission electron micrograph. EDS, energy dispersive spectra

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
NUMEROUS LABORATORY studies have examined the synthesis of iron sulfide minerals by the reaction between Fe and S anions at different temperatures, pressures, and reagent concentrations (Berner, 1964; Rickard, 1969; Wilkin and Barnes, 1996). Iron sulfide formation by sulfidation of Fe hydroxides has been studied by several researchers (Berner, 1964; Kumai, 1960; Rickard, 1974). Kumai (1960) proposed a two-stage mechanism for the sulfidation of ferric hydroxide. The first stage involved the diffusion of H₂S to the solid surface and the second was the reaction between H₂S and the dissolved Fe³⁺. No description of the product phase was given. Rickard (1974) hypothesized that goethite sulfidation involved redox and substitution processes with the production of poorly crystalline metastable Fe sulfides and S. He studied the kinetics and mechanism of goethite sulfidation and suggested that the reaction consisted of the formation of Fe³⁺ by acidic dissolution of goethite followed by the reaction between Fe³⁺ and dissolved S^2- with FeS and elemental S as end products. Berner (1964) studied the reaction of sulfide ions with goethite and found that the end product depended on the reaction pH. At pH of about 4, tetragonal FeS and pyrite formed, whereas at pH range 6 to 9, tetragonal FeS and elemental S were produced.

In all the above studies, the reactions were studied in distilled water with relatively low concentrations of ions such as Na⁺ and Cl^- originating from the starting materials. In this study, our objective was to determine the early phases of iron sulfide forming under ionic conditions approximating a marine environment. Therefore, artificial seawater or NaCl was used as the background electrolyte.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Synthesis of Sulfide Phases
The sulfide minerals were synthesized in saline water (except for the check that contained no salt) by a two-step process using only deionized deoxygenated water. The saline solutions were prepared first and then the H₂S gas was prepared in a simple reactor that was subsequently linked to the saline solutions containing sources of Fe. Water containing 26 g L^-1 NaCl or 35 g L^-1 artificial seawater-mix (Forty Fathoms Crystal Sea, Marine Enterprises International, Baltimore, MD) was deoxygenated by bubbling N₂ for 1 h into a sealed glass bottle containing 1 L of the solution. The N₂ gas was added through a diffuser tube and the excess gas removed through an outlet tube. Seventeen milliliters of the solution was delivered by gravity through the outlet tube into each 150 by 16 mm glass test tube, and the test tube was promptly sealed with a rubber septum.

Hydrogen sulfide was produced in a reactor (100-mL glass test tube equipped with a rubber stopper with two outlets), by the acidification of a sodium sulfide solution (about 2 g Na₂S·9H₂O/25 mL H₂O). Acidification was achieved by drop-wise delivery (to prevent aggressive frothing) of approximately 1 mL of 12 M HCl solution with a syringe that was mounted on one of the two openings via a hypodermic needle inserted in the stopper. The gas exited through the second opening in the reactor stopper via a 3-mm i.d. glass tube during several seconds. The other end of the 3-mm i.d. glass tube was connected with a 4-mm i.d. Tygon tube to another double-outlet rubber stopper. This stopper was equipped with a 140 mm-long by 3-mm i.d. inlet glass tube and a hypodermic needle acting as a vent for excess gas. Once the H₂S gas started to evolve from the reactor, the stopper with the 140 mm-long glass tube was mounted on a test tube containing the abovementioned deoxygenated solutions, and gas bubbled for a few seconds, producing sulfide-enriched-solutions with a concentration range of 600 to 1400 mg of total S^2- per liter; and solution pH was between approximately 4.5 and 6. At that point, the test tube was closed with the rubber septum and was ready for the addition of an Fe source. The sulfide enrichment process was performed under a vent hood to remove any H₂S gas that escaped.

The sulfidized solution pH was determined with a glass electrode by removing the rubber septum and a sample was removed for sulfide determination. Thereafter, test tube headspace was flushed with N₂ for 30 s. Flushing of the test tube headspace was done by N₂ injection through a hypodermic needle inserted into the stopper, where excess N₂ was removed via a second hole made by passing a second hypodermic needle through the stopper. After 30 s both needles were removed simultaneously. Sulfide concentrations were determined by the methylene blue method (Cline, 1969).

Mineral Preparation
Pure goethite was prepared synthetically using slow oxidation of a mixture of FeCl₂·4H₂O and NaHCO₃ (Schwertmann and Cornell, 1991). The ferrihydrite used was 2-line ferrihydrite, which was prepared by the addition of KOH to Fe(NO₃)₃·9H₂O (Schwertmann and Cornell, 1991). The excess salt was removed by centrifugation using several washing with deionized water. Part of each Fe oxide was then stored in the deionized solution and the other part was freeze-dried. The hematite used in this experiment was a fine clay fraction from commercial reagent grade Fe₂O₃. The purity of the Fe oxides was confirmed with x-ray diffraction (XRD) analysis (Joint Committee on Powder Diffraction Standards [JCPDS], 1980).

A 1-mL aliquot of deionized deoxygenated water with different concentrations (ranging between 2.1 x 10^-3 and 0.04 g) of goethite, ferrihydrite, hematite or FeSO₄·7H₂O was delivered to the test tube through a pipette with the stopper removed for about 5 s. The test tube headspace was then flushed with N₂ for 30 s.

X-Ray Diffraction and Transmission Electron Microscopy Methods
The material precipitated during the reaction of the Fe source and the S^2- was transferred from the test tube with a pipette, to a polished quartz slide (The Gem Dugout, State College, PA) mounted in a conventional Al box mount provided with the x-ray diffractometer and dried in approximately 3-min in a desiccator under vacuum (Vacuum pump model 1405M Duoseal, Sargent-Welch Scientific Co. Skokie, IL). To remove soluble salts, the slide was then washed by emersion in deionized water for approximately 1 min and redried in a 6-L desiccator under vacuum. The sample in the desiccator changed from glossy to lusterless appearance within 60 to 90 s indicating wet and dry conditions, respectively. Past experience with framboidal pyrite has indicated that keeping Fe sulfide samples dry is essential to maintaining their stability. The slide was then analyzed by XRD from 2 to 60° 2 using a step size of 0.05° 2 and a count time of 5 s. The XRD analysis was done on a Norelco diffractometer (Philips, Mahwah, NJ) with a graphite monochromator, using a theta compensating slit and Cu target x-ray tube operated at 30 kV and 18 mA.

Samples of the early products of the reaction between H₂S and either ferrihydrite or goethite (i.e., 5–12 h) in the NaCl solutions were examined promptly by XRD and transmission electron microscopy. The transmission electron micrograph (TEM) analyses were performed using a JEOL 2010 microscope (JEOL [U.S.A.] Inc., Peabody, MA) operated at 200 kV and energy dispersive spectra (EDS) by an attached Oxford EXL unit. Precipitated samples were washed with deionized and distilled water and sonicated before drop-mounting on a holey carbon film. The samples were maintained in an oxygen-free environment before analysis.

The crystallographic and XRD characteristics of proposed structures for the 1.0-nm phase were modeled using the Atoms software package (Dowty, 1999). Without thermal factors for the atoms in the proposed structures, the program can only approximate intensities for XRD peaks but the approximations provided trends on peak intensities as a function of filling in the disordered layer, and allowed determination of the approximate XRD patterns of potential structures.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
A broad XRD peak was observed at about 1.0 nm in the initial stages of all treatments in salt solutions and it was not observed in experiments where salt was excluded. The phase producing this peak was found to be a previously unidentified Fe sulfide mineral, which we gave a provisional mineral name, dorite. Employing a separate name for a regularly interstratified phase is consistent with the naming practice of layer silicates (Reynolds, 1980).

Goethite and sulfidic solution reactions gave the strongest evidence for dorite by the relative equivalence of 1- and 0.5-nm peaks. Dorite also contributes to the 0.5-nm peak that is mostly attributed to poorly crystalline mackinawite (Fig. 1A). Both XRD peaks maintained about equal intensities through the first 31-h reaction time. The dorite first-order peak waned within 96 h. In addition, pyrite and rhombic sulfur were present after 96 h. Dried goethite did not form dorite or mackinawite as discussed later (Table 1).

View larger version (22K): [in this window] [in a new window]   Fig. 1. X-ray diffraction curves of precipitated iron sulfides from the reactions between sulfide ions and three Fe sources in saline environments at different times. Goethite in NaCl solution (A). Ferrihydrite in NaCl solution (B). Ferrihydrite in artificial seawater (C). S = rhombic sulfur; P = pyrite, N = NaCl; peak labels are in nanometers.

The ferrihydrite in artificial seawater also produced dorite and mackinawite early in the reaction series (Fig. 1C). The dorite peak weakened and mackinawite was dominant after 120 h. Further reaction time produced pyrite and little mackinawite remained. The narrow pyrite and rhombic sulfur peaks produced in both ferrihydrite reaction series depicted a sharp contrast in the crystal size of dorite compared with pyrite and rhombic sulfur.

Hematite, represented by previously dried reagent Fe oxide, yielded a 1-nm peak, attributed to dorite, and rhombic sulfur. The 1-nm dorite peak disappeared before the 21-h XRD curve was obtained and mackinawite and rhombic sulfur were present. After 65 h, mackinawite and pyrite were observed.

The XRD data from the reaction between FeSO₄·7H₂O and sulfide indicate that mackinawite and rhombic sulfur precipitate in the first 4.5 h. After 30 h, pyrite was detected in addition to the other two minerals (Table 1).

Ferrihydrite and goethite samples that were dried before the reaction with sulfide in the NaCl solutions did not yield dorite (Table 1).

Transmission Electron Microscopy Observations
Electron microscopy observations indicated good electron beam penetration was limited to edges of the particles because they tended to be thick in the middle and were composed of high atomic number elements. Long sets of fringes with dimensions of approximately 1-nm and only a few per set were present in the sample in few places (Fig. 2A) indicating a poorly developed crystalline phase. Some fringes were long yet occurred in a very disorderly arrangement (Fig. 2B). Also, short fringes at approximately 1 nm, in various orientations, were visible in many places (e.g., Fig. 2C). The infrequency of coarse fringes and their presence in pairs or a few repeats indicates that the individual crystals were very thin. The goethite used as a source of Fe in some experiments produced very orderly fringes at 0.27 nm that were useful as an internal standard yet they were too closely spaced to interfere with the fringes attributed to dorite.

View larger version (144K): [in this window] [in a new window]   Fig. 2. Transmission electron microscope image of FeS with long lattice fringes spaced at approximately 1 nm trending downward to the right (A); long lattice fringes at various orientations (B); and short lattice fringes some parallel and others converging in almost horizontal orientation (C). The scale is in nanometers.

Dorite has a unit cell with at least one dimension of 1 nm or greater. Dorite was formed only in the saline environments where the major ions were Cl^- and Na⁺. Furthermore, dorite was formed in the goethite and ferrihydrite treatments only if those minerals where left hydrated after synthesis. The surface area of a freshly precipitated 2-line ferrihydrite as used in this study was 200 to 300 m² g^-1 (Schwertmann and Cornell, 1991). Where ferrihydrite was dried before the sulfidization to make Fe sulfides the surface area was <10 m²g^-1 (Wang, 1995). These results indicate that the ferrihydrite surface area is greatly reduced on drying. It is suggested that the formation of the new 1-nm phase is surface area dependent and forms only when the oxides have a large surface area. The formation of dorite from reagent quality hematite that had been dried, of course, is an exception to the behavior of the other oxides that were dried. The hematite behavior is analogous to the reactivity of pyrolusite from a chemical reagent observed by Kim et al. (2002). They attributed the higher than expected reactivity to heating during the manufacture of the reagent. Thus the reagent quality oxides tend to be far from equilibrium expected for a fully crystallized mineral.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Studying the formation of iron sulfide minerals using XRD, several researchers (Herbert et al., 1997; Wang and Morse, 1996) started their XRD scans at 10° 2 (0.884 nm). Our results suggest that the newly observed phase may not have been observed, if present, by the earlier studies because of their choice of the starting angle for obtaining the XRD data. In any run where we produced dorite indicated by XRD, a prompt analysis of it in the TEM revealed very weak fringes approximately 1-nm d-spacing.

The lattice fringes of dorite resemble thin smectite particles that fold and often show a few lattice fringes with variable spacings near the edge of the particle (Reid-Soukup and Ulery, 2002). Many small sets of short dorite lattice fringes (Fig. 2C), correlate with the compositional asymmetry and weak interlayer bonding of the layers suggested by structural interpretations. The short fringes (Fig. 2 area C) are identified as lattice fringes because of the spacing (about 1 nm), uniform level of contrast, and wide distribution. Alternatively they could conceivably be moiré fringes.

The thin sheets of dorite that are approximately eight layers thick (as measured from XRD peak broadening by the Scherrer equation) probably are very chemically and physically reactive like smectite. The fact that dorite contains reduced Fe and exists in small particles makes it very vulnerable to oxidation, Fe oxide precipitation, and acid formation.

If natural environments with similar saline and pH conditions to those in our study were carefully sampled, the dorite might be observed. An example of a possible environment where this phase may be observed would be in coastal environments where sediments containing high-surface-area Fe oxides enter sulfate-containing saline water. If a storm resulted in nutrient-rich runoff, a sudden bloom of phytoplankton might result. The bloom crash might then cause accelerated microbial activity in the sediments resulting in the production of sulfide that can react with the Fe oxide. Another possible environment would be near a hydrothermal vent on the ocean floor with Fe oxides from sedimentation where high sulfide concentrations (up to 110 mmol kg^-1 according to Von Damm et al., 1995) are present.

The formation of dorite may serve an important function in the pond-grown shrimp and fish industry by keeping excessive S^2- in the solid phase for a short time, thus protecting shrimp and other marine organisms from toxic levels of the ion. Dorite may be viewed as a short-term buffer against H₂S in solution until mackinawite or other sulfide phase forms.

Dorite forms first and eventually is converted to mackinawite indicated by the relative prominence of the 0.5-nm peaks in 96 to 240 h (Fig. 1A, B, & C). Both dorite and mackinawite were very fine grained indicated by broad XRD peaks at 1 and 0.5 nm that formed first, followed by weakening of the 1-nm dorite peak and strengthening of the mackinawite part of the 0.5-nm peak. As crystallization continued increasing Fe in the defect tetrahedral sheet, the intensity of the 1-nm peak decreased as the structure gradually changed from dorite to poorly ordered mackinawite. The increased d-spacing from 1 nm (twice the ideal mackinawite spacing) to 1.1 nm suggests that the defect tetrahedral sheet was thicker than the normal FeS sheet of mackinawite possibly because of the inclusion of water molecules.

The roles of Na and Cl ions in the formation of dorite are speculative yet the salt was required to form dorite in our experiments. Sodium is a highly hydrated ion that is prone to disperse silicate systems and may contribute to the dispersion of dorite particles thus preventing mackinawite formation initially. This inference is analogous to the complete separation of smectite layers in the presence of Na in dilute solution.

We hypothesize that an Fe-Cl complex serves as an intermediary that contributes to the formation of dorite. Although the bond between Fe²⁺ and S^2- is thermodynamically much more stable than the one between Fe²⁺ and Cl^-, the large excess of chloride ions (i.e., 17 000 mg L^-1 Cl^- vs. 1400 mg L^-1, at the most, of total S^2-) might result in an initial fast complexation reaction between Fe²⁺ and Cl^-, which would influence the sulfide reaction due to mass action.

The proposed iron chloride complex probably is octahedral because most complexes of ferrous Fe are octahedral (Cotton and Wilkinson, 1980). Tetrahedral coordination of ferrous Fe does occur in contrasting systems e.g., in proteins (Lauher and Ibers, 1975) and has been suggested for brines at high temperatures and pressures (D.C. McPhail, personal communication, 2002). Thus, the following conceptual roles of Na⁺ and Cl^- in dorite are suggested:

1. 3NaCl + 3H₂O + Fe²⁺ FeCl₃·3H₂O^-_{(octahedral complex)} + 3Na⁺
   
2. H₂S HS^- + H⁺
   
3. FeCl₃·3H₂O^-_{(octahedral complex)} + Na⁺ + HS^- + H⁺ FeS_(colloid) + Na⁺ + 3Cl^- + 3H₂O + 2H⁺
   
4. It is not surprising that ions required to form crystals are not present in the final phase. Ferrous Fe is generally recognized to be required for the formation of lepidocrocite yet it is not considered to be a part of the structure (Schwertmann and Taylor, 1989). In the synthesis of todorokite hydrated Mg ions were considered to be necessary to form tunnels of the desired size (Golden et al., 1986) yet the hydrated Mg is not essential to the structure. And of course the ions held between the layers of smectites and vermiculites are not considered as structural constituents.
   
5. Sodium and Cl ions were diluted by several orders of magnitude during the preparation of grids where separate dorite particles were investigated by TEM. Precipitated salts, which rarely occur on TEM grids as we prepare them, tend to form local particles during drying of the very slightly turbid drop of suspension placed on the grid and they might not be encountered during examination of mineral particles.
   

Proposed Structure for Dorite
The structure proposed for dorite is based on that of mackinawite, which has tetrahedral sheets of FeS where each Fe is shared by four S atoms (Lennie et al., 1995). The individual FeS-tetrahedra are joined in the sheet by sharing edges. The proposed structure for dorite has two FeS tetrahedral sheets, one sheet having an ideal mackinawite composition and a defect sheet containing less Fe, correspondingly less S to balance the charge, and water molecules to complete the tetrahedral coordination of Fe (Fig. 3). Lower Fe occupancy in the second tetrahedral sheet is required to produce a 1-nm peak. As the proposed structure was modeled with the Fe occupancy of the two sheets approaching equality, the intensity of the 1-nm peak decreased relative to the 0.5-nm peak until it was not detectable (Dowty, 1999).

View larger version (93K): [in this window] [in a new window]   Fig. 3. Proposed structure for dorite is composed of two sheets of edge-linked tetrahedra, one orderly like mackinawite and the other containing many defects. The two sheets with only light tetrahedra have the mackinawite structure with Fe in all tetrahedral positions and sulfur atoms at the vertices. The other sheets do not have Fe atoms in all the tetrahedral sites with the larger, darker tetrahedra having water molecules filling out the Fe coordination shell.

The potential choices for atoms filling out the hydration shell for the Fe²⁺ in the incomplete defect tetrahedral sheet include OH^-, Cl^-, and OH₂. The second sheet must be uncharged and contain no Cl or exchangeable cations (Cl was not detected by EDS). Also, Cl bonds with Fe²⁺ would be unlikely; they have high covalent character and would not result in longer bond lengths. Bond lengths for Fe²⁺–OH bonds are ideally 0.176 and 0.201 nm, for hydroxyls bonded to one and two Fe²⁺ atoms, respectively (Brown and Shannon, 1973). These bond lengths are shorter than Fe-S bond lengths and would therefore not increase the thickness of the defect sheet. A bond between Fe²⁺ and H₂O filling out the hydration shell would be much longer, about 0.250 nm (Donnay and Allmann, 1970) resulting in a thicker defect sheet. These bonds would increase the thickness of the sheet by approximately 0.04 nm if present on one side of the sheet only or 0.08 nm if present on both sides of the sheet. The sheet thickness obtained from having Fe-OH₂ bonds on both sides of the defect sheet compares well with the increase in thickness of the two-sheet structure as compared with two ideal mackinawite sheets.

The ideal symmetry for mackinawite is P4/nmm. The symmetry was reduced to P1 and the unit cell a-dimension increased to describe the defect structure and allow for multiple Fe, S, and H₂O sites within a single plane of atoms because the ideal structure contains only one Fe and one S position. The relative atom locations were retained from the mackinawite structure other than an increase in distance from the Fe atoms to the H₂O molecules, which replaced some S atoms in the defect tetrahedral sheet.

Where adequate Fe was provided in the presence of sulfide in saline solutions, a new mineral, dorite, formed quickly as a precursor to mackinawite, which in turn was a precursor to pyrite. The crystals for this newly observed mineral were small as evidenced by broad x-ray diffraction peaks and the sets of two to four lattice fringes commonly observed in TEM. It is hypothesized that this mineral forms in the early reduction of Fe in saline environments. The mineral has not been observed earlier because of the short period of persistence in the environment and because of the experimental conditions used by earlier researchers to characterize their samples.

Dorite is colloidal in size and is likely to be chemically and physically important in saline environments where marine organisms grow naturally or in aquiculture. More research is needed to determine the role of NaCl in the formation of dorite, the range of properties of dorite, and its natural occurrence.

	ACKNOWLEDGMENTS

 
We thank A.L. Lawrence for inspiration to pursue this research and recognize Texas Agricultural Experiment Station projects H-7801 and H-8158 for support and the Microscopy and Imaging Center, Texas A&M University for our use of their facilities.

Received for publication June 12, 2002.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

• Berner, R.A. 1964. Iron sulfides formed from aqueous solution at low temperatures and atmospheric pressure. J. Geol. 72:293–306.
• Brown, I.D., and R.D. Shannon. 1973. Empirical bond-strength-bond-length curves for oxides. Acta Crystallogr. A29:266–282.
• Cline, J.D. 1969. Spectrophotometric determination of hydrogen sulfide in natural waters. Limnol. Oceanogr. 14:454–458.
• Cotton, F.A., and G. Wilkinson. 1980. Advanced inorganic chemistry: A comprehensive text. p.689–821. John Wiley & Sons. New York.
• Donnay, G., and R. Allmann. 1970. How to recognize O^2-, OH^-, and H₂O in crystal structures determined by X-rays. Am. Mineral. 55:1003–1015.
• Dowty, E. 1999. Atoms for Windows. Version 5.1b. Shape Software, Kingsport, TN.
• Golden, D.C., C.C. Chen, and J.B. Dixon. 1986. Synthesis of todorokite. Science. 231:717–719.[Abstract/Free Full Text]
• Herbert, R.B., S.G. Benner, A.R. Pratt, and D.W. Blowes. 1997. Surface chemistry and morphology of poorly crystalline iron sulfide precipitated in media containing sulfate-reducing bacteria. Chem. Geol. 144:87–97.
• Joint Committee on Powder Diffraction Standards (JCPDS). 1980. Mineral powder diffraction file. International Center for Diffraction Data, Swarthmore, PA.
• Kim, J.G., J.B. Dixon, C.C. Chusuei, and Y. Deng. 2002. Oxidation of chromium(III) to (VI) by manganese oxides. Soil Sci. Soc. Am. J. 66:306–315.[Abstract/Free Full Text]
• Kumai, T. 1960. Reaction between ferric hydroxide and hydrogen sulfide. Nippon Kagaku Zasshi 81:1687–1695.
• Lauher, J.W., and J.A. Ibers. 1975. [FeCl₄] Structure of tetramethylammonium tetrachloroferrate(II), [N(CH₃)₄]₂. Inorg. Chem. 14:348–352.
• Lennie, A.R., S.A.T. Redfern, P.F. Schofield, and D.J. Vaughan. 1995. Synthesis and Rietveld crystal structure refinement of synthetic mackinawite, tetragonal FeS. Mineral. Mag. 59:677–683.[Abstract]
• Reid-Soukup, D.A., and A.L. Ulery. 2002. Smectites. p.467–499. In J.B. Dixon and D. G. Schulze, (ed.) Soil mineralogy with environmental applications. SSSA Book Ser. 7. SSSA, Madison, WI.
• Reynolds, R.C. 1980. Interstratified clay minerals. p.249–303. In G.W. Brindley and G. Brown (ed.) Crystal structures of clay minerals and their x-ray identification. Mineralogical Soc. Monogr. 5. Mineralogical Society, London.
• Rickard, D.T. 1969. The chemistry of iron sulfide formation at low temperatures. Stockholm Contrib.Geology 20:67–95.
• Rickard, D.T. 1974. Kinetics and mechanism of the sulfidation of goethite. Am. J. Sci. 274:941–952.[Abstract]
• Schwertmann, U., and R.M. Cornell. 1991. Iron oxides in the laboratory. p.53–94. VCH, Weinheim, FRG.
• Schwertmann, U., and R.M. Taylor. 1989. Iron Oxides. p. 379–438. In J.B. Dixon and S.B. Weed (ed.) Minerals in soil environments. SSSA Book Ser. 1. SSSA, Madison, WI.
• Von Damm, K.L., S.E. Oosting, R. Kozlowski, L.G. Buttermore, D.C. Colonder, H.N. Edmons, J.M. Edmond, and J.M. Grebmeier. 1995. Evolution of east Pacific rise hydrothermal vent fluids following a volcanic eruption. Nature 375:47–50.
• Wang, Q.W. 1995. Experimental investigation of pyrite formation under conditions approximating those found in anoxic marine sediments. Ph.D. dis. Texas A&M University, College Station, TX. (Dis. Abstr. AAT 9615917).
• Wang, Q.W., and J.W. Morse. 1996. Pyrite formation under conditions approximating those in anoxic sediments. 1. Pathway and morphology. Mar. Chem. 52:99–121.[ISI]
• Wilkin, R.T., and H.L. Barnes. 1996. Pyrite formation by the reaction of iron monosulfides with dissolved inorganic and organic sulfur species. Geochim. Cosmochim. Acta 60:4167–4179.

This article has been cited by other articles:

				M. E. Fleet Phase Equilibria at High Temperatures Reviews in Mineralogy and Geochemistry, January 1, 2006; 61(1): 365 - 419. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Ritvo, G. Articles by Dixon, J. B. Search for Related Content PubMed Articles by Ritvo, G. Articles by Dixon, J. B. GeoRef GeoRef Citation Agricola Articles by Ritvo, G. Articles by Dixon, J. B. Related Collections Wetlands and Aquatic Processes Colloids Geochemical Processes Soil Chemistry