Gypsum solubility in seawater, and its application to bauxite residue amelioration.
Approximately 30Mt of highly alkaline (pH 12) and sodic (ESP 100%) residues are produced throughout the world annually from the refining of bauxite (Oeberg and Steinlechner 1996). Bauxite residues are generally separated into 2 fractions: a coarse fraction (residue sand, 96% particle diameter 0.02-2.0mm) and a fine fraction (red mud, 90% particle diameter <0.02 mm). Due to its high hydraulic conductivity and comparatively low buffering capacity, residue sand is easily leached to remove excess soluble salts. However, even following neutralisation with seawater and leaching with freshwater, pH (9.5) and sodicity (ESP 50%) remain problematic for the revegetation of the residue. Neutralisation of residue sand with seawater is most economically achieved by pumping the residue sand in a seawater slurry from the refinery to the residue disposal area, with neutralisation occurring during transit.
Plant growth in residue sand is severely limited by nutrient deficiencies, in particular Ca and Mn. The use of gypsum (CaS[O.sub.4].2[H.sub.2]O) as a residue sand ameliorant increases available Ca while also reducing pH (Wong and Ho 1993). The addition of gypsum to the residue sand/seawater slurry during pipelining, rather than surface application of gypsum on the residue sand after it is placed in the field, is not only economically attractive but would also ensure a uniform distribution of gypsum throughout the soil profile. However, the dissolution of gypsum in seawater during pumping is potentially problematic, because once at the residue disposal area, the seawater is separated from the residue sand before being treated and discharged. The residue sand and solid phase gypsum are then stored and prepared for revegetation. Hence, unless the Ca is adsorbed from the seawater by the residue sand, any gypsum which has dissolved during pipelining will be lost from the system. The presence of supernatant liquor (SNL), a highly saline (electrical conductivity (EC) 35 dS/m) and alkaline (pH 11.4) waste product, in the residue sand/seawater slurry may also affect gypsum dissolution.
In the experiments reported here, we aimed to determine the rate and extent of dissolution of various gypsum sources and size fractions in seawater. Three gypsum types from different sources and of differing purity were considered. The degree to which Ca is removed from gypsum-treated seawater by residue sand was also investigated, and the effect of highly alkaline SNL contamination on gypsum dissolution during pipelining examined.
Materials and methods
Gypsum samples were separated into 8 treatments, consisting of analytical grade reagent (AR) gypsum (BDH Analar CaS[O.sub.4].2[H.sub.2]O), phosphogypsum (PG) (produced as a byproduct of the phosphoric acid industry), and 6 mined gypsum (MG) size fractions (>2.0, 1.0-2.0, 0.5-1.0, 0.25-0.5, 0.125-0.25, <0.125mm). The MG was separated into size fractions by wet sieving in a saturated gypsum solution. The presence of impurities was determined using X-ray diffraction (XRD) (Philips PW1800, 0.05[degrees] 2 theta steps with 3.0 s counting per step, quantitative analysis using SIROQUANT (Sietronics Pty Ltd)), surface area using the BET method (Quantachrome Nova 1200), and solution composition using inductively coupled plasma-atomic emission spectroscopy (ICPAES) (Spectro Analytical Instruments) (Table 1). Solution pH (TPS 901-CP) and EC (Radiometer CDM210) were also determined. Particle diameter was determined for each MG size fraction (Malvern Mastersizer 2000) and external surface area estimated (assuming spherical particles and a specific gravity of 2.32g/[cm.sup.3]; Jackson 1958).
The compositions of seawater, SNL, and the seawater/SNL mixture (9 : 1) were determined using ICPAES after 0.025-[micro]m filtration (Millipore VSWP) (Table 2). Total alkalinity was determined for each solution using standardised HCl (Greenberg et al. 1992), and pH and EC were also measured. The precipitate formed during the mixing of the seawater/SNL mixture was analysed using XRD and found to consist predominantly of monohydrocalcite (CaC[O.sub.3].[H.sub.2]O) (76%).
An experiment was undertaken to determine the extent and rate at which gypsum dissolution occurs in triple de-ionised water (TDI) and seawater for the various sources and size fractions. The dissolution rates of the 8 gypsum treatments were measured by dissolving 1 g in 20 mL of TDI or seawater, both with 3 replicates. For each of TDI and seawater, the treatments were selected in random order from the 24 solutions. Solution EC was recorded every 5 s for 60 min, except in the AR, PG and MG <0.125 mm treatments, in which EC was recorded every 1 s due to their higher dissolution rates. Solutions were stirred using a 15-mm magnetic stirring bar at 100 r.p.m, and temperature maintained at 25 [+ or -] 1[degrees]C. Upon equilibration, samples were taken from each solution, filtered using a 0.025-[micro]m filter, and analysed for Ca and S by ICPAES. Using the initial and saturation Ca concentrations measured by ICPAES, the observed changes in solution EC during the dissolution studies were converted to solution Ca concentrations. The solubility of gypsum was calculated from the average saturation concentrations of Ca and S using the relative formula mass of CaS[O.sub.4].2[H.sub.2]O (172.17 g/mol).
Using GENSTAT 6 (GENSTAT 2002), a 1-way analysis of variance (completely randomised design) of the equilibrium solubilities for each of TDI and seawater was performed. Comparisons between means were made using Fisher's protected least significant difference (1.s.d.) test. A linear regression was performed to examine the relationship between first-order rate constant and surface area for the 6 MG size fractions in both TDI and seawater.
Removal of solution Ca by residue sand
While gypsum source and size fraction influence the dissolution rate (and hence rate of increase in solution Ca), solution Ca concentrations are also affected by the removal of Ca from the solution by exchange for Na on the cation exchange capacity of the residue sand. The rate and extent at which Ca is removed from solution by residue sand were therefore investigated.
Six seawater treatments were prepared by the addition of AR gypsum to seawater in various quantities up to saturation (g/L): 0.00, 0.25, 0.50, 1.25, 2.50, 3.75. Fresh residue sand (2 g) was added to a 50-mL tube and 40 mL of the appropriate seawater solution incorporated. Tubes were then shaken, with 6 equilibration times (0 (prior to residue sand addition), 5, 15, 30, 60min, and 12 h), and 2 replicates, giving a total of 72 solutions. After shaking, the solutions were filtered (0.025 [micro]m) and analysed by ICPAES. The pH values of the initial seawater/gypsum solutions were measured, and suspension pH also measured after 12 h.
Effect of SNL on gypsum dissolution
An experiment was conducted to determine whether the presence of highly alkaline SNL as a contaminant during the seawater neutralisation of residue sand affects gypsum dissolution in the pipeline.
A seawater--SNL (9 : 1) mixture was prepared; 1 g of each of the 8 gypsum treatments was placed in 20mL of the mixture and dissolution recorded as described above. Treatments were sampled after 60 min, filtered (0.025 [micro]m), and analysed for Ca and S using ICPAES. Gypsum particles (MG >2.0) were removed from the solution after 60 min, air-dried, embedded in 'Daystar' polyester resin, polished to cross-section with silicon carbide paper and diamond caps, and platinum coated. The particles were analysed using low vacuum analytical scanning electron microscopy (SEM) (JEOL JSM-6460 LA, JEOL JSM silicon energy dispersive X-ray spectrometer (EDS) at 20 kV accelerating voltage).
Results and discussion
The solubility of gypsum in TDI (approximately 2.9 g/L) was found to be lower than that in seawater (approximately 3.8 g/L) (Table 3). The higher gypsum solubility in seawater is attributable to the effect of ionic strength, with gypsum solubility increasing with increasing salt concentrations (Tanji 1969). However, the observed solubility of gypsum in seawater (ionic strength (I) = 0.7) was lower than the 6 g/L reported for a 650 mM NaCl solution (I = 0.7) (Bock 1961; Marshall et al. 1964). This solubility reduction observed in seawater when compared to the NaCl solution is due to common ion effects (Kemper et al. 1975), with [Ca.sup.2+] (10.6 mM) and S[O.sub.4.sup.2-] (26.8 mM) both present in seawater. The observed gypsum solubilities were similar to reported values for both TDI [(g/L): 2.72, McElnea 2003; 2.7, Noble and Randall 1998; 2.63, Bennett and Adams 1972; 2.6, Bolan et al. 1991; 2.6, Keren and Shainberg 1981; and 2.6, Gobran and Miyamoto 1985] and seawater (3.6 g/L, Shaffer (1967).
Significant differences in solubilities were found between gypsum treatments in both TDI (P < 0.001, 1.s.d. (P = 0.05) 0.07) and seawater (P < 0.001, 1.s.d. (P = 0.05) 0.06) (Table 3). These differences in solubility can be attributed to differences in purity between the different sources and size fractions (Table 1). Studying AR gypsum, PG, MG, and flue-gas desulfurisation gypsum, Bolan et al. (1991) also found the solubilities of the various gypsum sources to be similar, whereas Noble and Randall (1998) found the solubility of MG (2.36 g/L) to be less than that of PG (2.67 g/L), with AR having a solubility of 2.51 g/L.
The equilibrium Ca concentrations were used to fit the dissolution data to the first-order (Eqn 1) and second-order (Eqn 2) kinetic equations:
(1) dC/dt = [K.sub.1]([C.sub.s] - [C.sub.t])
(2) dC/dt = [K.sub.2][([C.sub.s] - [C.sub.t]).sup.2]
where [C.sub.t] is the concentration of Ca in solution at time t, [C.sub.s] is the saturation Ca concentration, and [K.sub.1] and [K.sub.2] are the first- and second-order dissolution rate constants. With the initial condition of [C.sub.t] = [C.sub.0] at t = 0, the integral of Eqn 1 yields:
(3) ln[([C.sub.s] - [C.sub.0])/([C.sub.s] - [C.sub.t])] = [K.sub.1]t
with the integral of Eqn 2 yielding:
(4) [1/([C.sub.s] - [C.sub.t]) - 1/([C.sub.s] - [C.sub.0])] = [K.sub.2]t
Gypsum dissolution in both TDI and seawater was found to fit the first-order equation (Eqn 1), with the plot of ln[([C.sub.s] - [C.sub.0])/([C.sub.s] - [C.sub.t])] v. t producing a straight line (Fig. 1).
[FIGURE 1 OMITTED]
The first-order equation has been widely used to describe gypsum dissolution (Liu and Noncollas 1971; Kemper et al. 1975; Keren and Shainberg 1981; Keren and O'Connor 1982; Noble and Randall 1998). Recently, however, gypsum dissolution in TDI has been reported to follow the second-order reaction equation, with the first-order equation deviating from the linear relationship at a [C.sub.t]/[C.sub.s] of approximately 0.3 (Gobran and Miyamoto 1985; Frenkel et al. 1989; Bolan et al. 1991). Frenkel et al. (1989) concluded that the reaction order is dependent upon the increase in [Ca.sup.2+] and S[O.sub.4.sup.2-] concentrations in the solution. In solutions where [Ca.sup.2+] and S[O.sub.4.sup.2-] are continually removed (e.g. in the presence of soil or an ion exchange resin), gypsum dissolution follows first-order kinetics. In comparison, in solutions in which [Ca.sup.2+] and S[O.sub.4.sup.2-] are not removed, dissolution follows second-order kinetics.
Gypsum dissolution is a transport-controlled process, with [Ca.sup.2+] and S[O.sub.4.sup.2-] moving into solution at a rate faster than that at which they move from this solution surface layer into the bulk solution (Berner 1981). The rate of gypsum dissolution is therefore a function of the rate of solution mixing. In this study, it is thought that the high mixing rate was sufficient to prevent the build-up of [Ca.sup.2+] and S[O.sub.4.sup.2-] at the gypsum surface. Hence, gypsum dissolution was not limited by the surface solution concentrations of [Ca.sup.2+] and S[O.sub.4.sup.2-] and followed first-order, rather than second-order, kinetics (Noble and Randall 1998).
The dissolution rate constants were found to be dependent upon both gypsum source and size fraction (Table 4). The dissolution rate constants decreased in the order AR > PG > MG, with larger constants indicating a faster rate of dissolution. Each of the gypsum treatments generally reached saturation within 15s (AR) to 30 min (MG > 2.0 mm). Bolan et al. (1991) and Noble and Randall (1998) also reported similar trends in dissolution rate constants between the various AR, PG and MG gypsum sources. The size fraction of the MG was also found to affect the dissolution rate constant, decreasing with increasing particle size (Table 4).
The high surface areas of the MG treatments compared with the AR and PG treatments (Table 4) are attributed mainly to the high internal surface areas of the MG, as indicated by SEM. Although the surface areas of the AR and PG treatments were lower than the MG treatments, dissolution rate constants were observed to be higher (Table 4). These differences in rates between gypsum sources are attributable to differences in reactivities (Barton and Wilde 1971), with the presence of impurities (in particular CaC[O.sub.3]) in MG decreasing dissolution rates (Keren and Kauschansky 1981). Dissolution rate constant has previously been reported to increase linearly with BET surface area (Bolan et al. 1991). Despite the high [r.sup.2] values obtained in this study for the linear regression of BET surface area and dissolution rate constant [0.861 (TDI) and 0.963 (seawater)], the linear curve was a poor predictor of this relationship (Fig. 2). It is believed that this can be attributed to the high internal surface area of the MG particles, which, although included in the BET surface area measurement, would have contributed little to the rate of dissolution. As dissolution is controlled by the rate of movement of [Ca.sup.2+] and S[O.sub.4.sup.2-] from the surface layer into the bulk solution (see earlier discussion), dissolution from internal surfaces is limited by the rate at which ions move from these internal spaces into the bulk external solution (which is anticipated to be relatively slow even in a well-stirred solution).
[FIGURE 2 OMITTED]
When related to external surface area (as estimated from particle size distribution, a specific gravity of 2.32 g/[cm.sup.3], and spherical particles), the dissolution rate constant was found to increase linearly with increasing external surface area (Fig. 2). Although the <0.125 mmMG size fraction was found to have high leverage (external surface area of 1973 [cm.sup.2]/g, Fig. 2), comparison of the TDI and seawater regressions fitted with and without this point indicated that there was no significant difference between either the slopes (P = 0.155 for TDI and P = 0.145 for seawater) or the intercepts (P = 0.956 for TDI and P = 0.955 for seawater) for these 2 curves.
Although dissolution constants are dependent upon many factors (including mixing/flow rate, impurities, particle size, and solution composition), those observed in this study in TDI ([K.sub.1] = 0.09-12/min) are similar to those reported in other studies (1.6-3/min, Kemper et al. 1975; 0.012-0.348/min, Keren and Shainberg 1981).
Removal of solution Ca by residue sand
Adsorption of solution Ca by residue sand was found to occur rapidly, with approximately 75% of the equilibrium absorbance occurring by 5 min (Fig. 3). The amount of Ca adsorbed was found to increase with increasing solution Ca concentration, with approximately 1 cmol Ca/kg residue sand adsorbed from the seawater control (0.00 g/L) and 2.0 cmol Ca/kg residue sand from the 3.75g/L treatment (Fig. 3).
[FIGURE 3 OMITTED]
Following the addition of residue sand to the seawater solutions, mean solution pH increased an average of 0.07 pH units across all treatments (Table 5). This modest increase in pH is due to the precipitation of Ca and Mg carbonates resulting from the addition of the highly alkaline (pH 12), fresh residue sand. Although the addition of residue sand resulted in the precipitation of CaC[O.sub.3], it is thought that the observed decreases in solution Ca concentrations were, for the most part, due to the adsorption of Ca by the residue sand. The decrease in Ca concentration in the control (1.0 cmol Ca/kg residue sand) was found to be similar to the exchangeable Ca of residue sand following seawater neutralisation (0.75 cmol/kg residue sand). Modelling using PhreeqcI 2.8 (Parkhurst 2003) also indicated that only a slight reduction in solution Ca (approximately 5 [micro]M) was due to CaC[O.sub.3] precipitation. The observed decrease in solution Ca is therefore not thought to be a solution pH/CaC[O.sub.3] precipitation effect.
Plant growth trials have shown that in order to avoid limitations caused by Ca deficiency, following the seawater neutralisation and fresh water leaching of residue sand, exchangeable Ca concentrations need to be raised to 2 cmol Ca/kg by the application of gypsum at 10 t/ha (assuming a bulk density of 1.3 g/[cm.sup.3] and an incorporation depth of 150 mm). To achieve this critical exchangeable Ca concentration by the application of gypsum to the residue sand/seawater slurry (rather than directly to the residue sand), a gypsum application of 3.75 g/L would be required (assuming 1 g residue sand/20 mL seawater) (Fig. 3), yielding an equivalent gypsum application rate of 150 t/ha (assuming a bulk density of 1.3 g/[cm.sup.3] and an incorporation depth of 150 mm). Therefore, due to the long pumping times (20 min) and comparatively rapid dissolution rates (reaching saturation within 15 s (AR) to 30 min (MG > 2.0mm)) the application of gypsum to the residue sand/seawater slurry is an inefficient method of improving the Ca status of residue sand with only 5% of the total Ca in solution adsorbed and the remaining Ca (95%) lost from the system upon seawater discharge.
Effect of SNL on gypsum dissolution
The addition of SNL to the seawater increased total alkalinity from 150 to 1375 mg CaC[O.sub.3]/L, with pH also increasing from 8.2 to 10.0 (Table 2). Although gypsum solubility is not affected by pH, solution Ca concentrations are often reduced with increasing pH due to the precipitation of CaC[O.sub.3] (Lindsay 1979). Due to this CaC[O.sub.3] precipitation, solution EC changes were an unreliable measure of gypsum dissolution in the highly alkaline seawater/SNL solution, with ECs of the coarser MG size fractions even decreasing with time (Fig. 4). Increases in solution S concentrations were used as a more reliable measure of gypsum dissolution. Following 60 min equilibration, gypsum had not reached saturation even in the AR treatment (assuming a gypsum solubility similar to that in seawater, Table 3), with solution S concentrations increasing by only 10 mM (Fig. 4).
[FIGURE 4 OMITTED]
It was hypothesised that this observed reduction in the gypsum dissolution rate was caused by the formation of a sparingly soluble CaC[O.sub.3] coating around the gypsum particles. Analysis of a >2.0 mm MG particle after 60 min equilibration in the seawater/SNL mixture with SEM showed the formation of both nodules and a coating surrounding the gypsum particle (Fig. 5). These nodules (Fig. 5, points 4 and 10) and the particle coating (Fig. 5, points 2, 8, and 9) were analysed using EDS and found to consist largely of CaC[O.sub.3], while the bulk particle consisted of CaS[O.sub.4] (Fig. 5, points 1, 5, 6, and 7). As this CaC[O.sub.3] coating was examined only using elemental analysis (rather than mineralogical analysis), it was not possible to determine the degree of hydration of the CaC[O.sub.3]. An aluminosilicate material (perhaps a feldspar) was also found to be present (Fig. 5, point 3), possibly formed during equilibration with the seawater/SNL mixture due to the SNLs high Al content (Table 2). Other randomly selected locations on the edge of other gypsum particles were selected, and all were found to show a CaC[O.sub.3] coating. Due to the low solubility of CaC[O.sub.3] at pH 10, it is thought that the presence of these CaC[O.sub.3] nodules and coatings would substantially reduce the rate of gypsum dissolution by reducing the gypsum surface area in contact with the solution.
[FIGURE 5 OMITTED]
A CaC[O.sub.3] coating surrounding gypsum particles has also been reported by Keren and Kauschansky (1981) using the electron microprobe technique, although the effect of this coating on dissolution rate was not examined.
In conclusion, gypsum solubility was found to be greater in seawater (3.8 g/L) than TDI (2.9 g/L) due to an ionic strength effect. Dissolution in both TDI and seawater was found to follow first-order kinetics. Dissolution rate constants varied with gypsum source (AR > PG > MG) due to reactivity and surface area differences, generally reaching saturation within 15 s (AR) to 30 min (MG >2.0 mm). Although bauxite residue was observed to adsorb Ca from saturated solutions, the quantity of the total adsorbed was small (5%). These low rates of solution Ca adsorption, combined with comparatively rapid dissolution rates, preclude the application of gypsum to the residue sand/seawater slurry as a method for residue amelioration. In addition, the rate of gypsum dissolution in highly alkaline seawater/SNL solutions was observed to be reduced due to the formation of a sparingly soluble CaC[O.sub.3] coating around the gypsum particles, limiting gypsum dissolution in the field.
Table 1. Properties of saturated solutions of various gypsum types and size fractions in triple deionised (TDI) water AR, Analytical grade reagent; PG, phosphogypsum; MG, mined gypsum Gypsum Size Purity pH EC Ca Mg Na P fraction (%) (dS/ (mM) (mM) (mM) (mM) (mm) m) AR NA 99.7 7.1 2.5 16.6 0.04 0.10 n.d. PG NA 98.4 4.4 2.5 16.3 0.02 0.06 0.10 MG <0.125 96.6 7.0 2.7 17.5 0.33 0.82 n.d. MG 0.125-0.25 97.6 7.0 2.6 16.9 0.04 0.08 n.d. MG 0.25-0.5 96.8 7.0 2.6 16.0 0.02 0.06 n.d. MG 0.5-1.0 97.2 6.8 2.6 17.0 0.05 0.10 n.d. MG 1.0-2.0 95.9 6.7 2.6 18.5 0.06 0.19 n.d. MG >2.0 97.8 6.8 2.6 16.8 0.07 0.15 n.d. NA, Not applicable; n.d., not detected (<0.01 mM). Table 2. Properties of the seawater, SNL (supernatant liquor), and seawater: SNL mixture (9 : 1) in which the gypsum dissolution rates were determined pH EC Alkalinity Ca (dS/m) (mg CaC[O.sub.3]/L) (mM) Seawater 8.18 53.2 150 10.6 SNL 11.44 34.5 23 500 0.08 Seawater: SNL 10.01 51.8 1 375 6.45 Mg K Na S Al P (mM) (mM) (mM) (mM) (mM) (mM) Seawater 57.7 9.89 440 26.8 n.d. 0.08 SNL 0.12 1.57 436 4.67 1.68 0.22 Seawater: SNL 53.4 9.00 421 32.0 0.02 0.03 n.d., Not detected (<0.01 mM). Table 3. Increase in concentrations (mM) of calcium and sulfate and calculated gypsum solubilities in triple deionised water (TDI) and seawater for various gypsum sources and size fractions Within a column, means with the same letter are not significantly different at P=0.05. Results are the arithmetic mean of 3 replications TDI Gypsum Size Change Change in Solubility fraction in [[Ca. [S[O.sub.4. (g/L) (mm) sup.2+]] sup.2-]] AR NA 16.4 16.0 2.83a PG NA 16.3 15.6 2.80a MG <0.125 17.6 17.4 3.03b MG 0.125-0.25 16.9 16.4 2.90c MG 0.25-0.5 16.1 15.5 2.78a MG 0.5-1.0 17.0 16.5 2.92c MG 1.0-2.0 16.5 16.4 3.18d MG >2.0 17.0 16.9 2.93c l.s.d. (P = 0.05) 0.07 Seawater Gypsum Change in Change in Solubility [[Ca.sup.2+]] [S[O.sub.4.sup.2-]] (g/L) AR 23.1 21.9 3.91a PG 22.7 21.3 3.90a MG 22.2 20.6 3.81bc MG 22.5 20.6 3.83b MG 22.0 19.8 3.76c MG 22.1 20.2 3.79bc MG 21.9 19.9 3.78bc MG 21.8 19.7 3.69d l.s.d. (P = 0.05) 0.06 NA, Not applicable. Table 4. First-order dissolution rate constants ([K.sub.1], per min) in triple-deionised water (TDI) and seawater for various gypsum sources and particle sizes Gypsum Size Surface Mean particle TDI Seawater fraction area diameter (mm) ([m.sup.2]/g) ([micro]m) AR NA 0.548 12 14 PG NA 2.65 6.5 5.0 MG <0.125 7.79 0.056 3.4 1.9 MG 0.125-0.25 5.52 0.20 0.28 0.19 MG 0.25-0.5 5.11 0.39 0.14 0.18 MG 0.5-1.0 4.80 0.77 0.13 0.18 MG 1.0-2.0 4.18 1.4 0.10 0.15 MG >2.0 4.00 2.3 0.09 0.11 NA, Not applicable. Table 5. Effect of the addition of various quantities of gypsum to seawater on solution Ca concentration and on solution pH before and after 12 h equilibration with residue sand Gypsum Initial Final Ca Initial Suspension treat- Ca conc. conc. solution pH (after 12 h) ment (mM) ([+ or -] s.e.) pH ([+ or -] s.e.) (g/L) (mM) 0.00 10.1 9.5 [+ or -] 0.04 8.18 8.25 [+ or -] 0.03 0.25 11.4 10.9 [+ or -] 0.04 8.08 8.20 [+ or -] 0.02 0.50 12.8 12.3 [+ or -] 0.06 8.15 8.22 [+ or -] 0.03 1.25 17.2 16.6 [+ or -] 0.07 8.12 8.21 [+ or -] 0.01 2.50 24.6 23.7 [+ or -] 0.09 8.01 8.06 [+ or -] 0.02 3.75 30.8 29.8 [+ or -] 0.04 7.90 7.94 [+ or -] 0.02
The authors acknowledge the assistance of Graham Kerven and David Appleton with chemical analyses and Bernhard Wehr for his suggestions, ideas, and comments. The statistical assistance provided by Rosemary Kopittke is also gratefully acknowledged as is the assistance of Ron Rasch and Kim Sewell (Centre for Microscopy and Microanalysis, The University of Queensland) with the use of the SEM. The support of Ian Fulton is also gratefully received as is his assistance in placing the research in the context of the alumina industry.
This work was conducted as part of an environmental research program funded by Alcan Gove Pty Ltd.
Abbreviations used: AR, analytical reagent; EDS, energy dispersive X-ray spectrometer; I, ionic strength; ICPAES, inductively coupled plasma-atomic emission spectroscopy; MG, mined gypsum; PG, phosphogypsum; SEM, scanning electron microscopy; SNL, supernatant liquor; XRD, X-ray diffraction
Barton FM, Wilde NM (1971) Dissolution rates of polycrystalline samples of gypsum and orthorhombic forms of calcium sulphate by rotating disc method. Transactions of the Faraday Society 67, 3590-3597. doi: 10.1039/TF9716703590
Bennett AC, Adams F (1972) Solubility and solubility product of gypsum in soil solutions and other aqueous solutions. Soil Science Society of America Journal 36, 1972.
Berner RA (1981) Kinetics of weathering and diagenesis. In 'Reviews in mineralogy'. (Eds AC Lasaga, RJ Kirkpatrick) pp. 111-134. (Mineralogical Society of America: Washington, DC)
Bock E (1961) On the solubility of anhydrous calcium sulphate and of gypsum in concentrated solutions of sodium chloride at 25[degrees]C, 30[degrees]C, 40[degrees]C, and 50[degrees]C. Canadian Journal of Chemistry 39, 1746-1751.
Bolan NS, Syers JK, Sumner ME (1991) Dissolution of various sources of gypsum in aqueous solutions and in soil. Journal of the Science of Food and Agriculture 57, 527-541.
Frenkel H, Gerstl Z, Alperovitch N (1989) Exchange-induced dissolution of gypsum and the reclamation of sodic soils. Journal of Soil Science 40, 599-611.
GENSTAT (2002) 'GenStat for Windows. Release 6.1.' 6th edn (VSN International Ltd: Oxford)
Gobran GR, Miyamoto S (1985) Dissolution rate of gypsum in aqueous salt solutions. Soil Science 140, 89-93.
Greenberg AE, Clesceri LS, Eaton AD (1992) 'Standard methods for the examination of water and wastewater.' (American Public Health Association: Washington, DC)
Jackson ML (1958) 'Soil chemical analysis.' (Constable: London)
Kemper WD, Olsen J, DeMooy CJ (1975) Dissolution rate of gypsum in flowing water. Soil Science Society of America Proceedings 39, 458-463.
Keren R, Kauschansky P (1981) Coating of calicum carbonate on gypsum particle surfaces. Soil Science Society of America Journal 45, 1242-1244.
Keren R, O'Connor GA (1982) Gypsum dissolution and sodic soil reclamation as affected by water flow velocity. Soil Science Society of America Journal 46, 726-732.
Keren R, Shainberg I (1981) Effect of dissolution rate on the efficiency of industrial and mined gypsum in improving infiltration of a sodic soil. Soil Science Society of America Journal 45, 103-107.
Lindsay WL (1979) 'Chemical equilibria in soils.' (John Wiley & Sons: New York)
Liu ST, Noncollas GH (1971) The kinetics of dissolution of calcium sulphate dihydrate. Journal of Inorganic and Nuclear Chemistry 33, 2311-2316. doi: 10.1016/0022-1902(71)80205-1
Marshall WL, Slusher R, Jones EV (1964) Aqueous systems at high temperature XIV. Solubility and thermodynamic relationships for CaS[O.sub.4] in NaCl-[H.sub.2]O solutions from 40 to 200[degrees]C, 0 to 4 molal NaCl. Journal of Chemical and Engineering Data 9, 187-191.
McElnea AE (2003) Assessing the ability of acid sulfate soil laboratory tests to predict environmental risk and lime amelioration. PhD thesis, University of Queensland.
Noble AD, Randall PJ (1998) Dissolution rates of selected Australian gypsum sources. Australian Journal of Experimental Agriculture 38, 273-277. doi: 10.1071/EA97162
Oeberg NCR, Steinlechner EH (1996) Red mud and sands handling: New thoughts on an old problem. In 'Light metals. 1996 TMS Annual Meeting.' Anaheim, Ca. pp. 67-73 (TMS)
Parkhurst D (2003) PhreeqcI. United States Geological Survey. http://water.usgs.gov/owq/software.html (Accessed Jan 2004)
Shaffer LH (1967) Solubility of gypsum in sea water and sea water concentrates at temperatures from ambient to 65[degrees]C. Journal of Chemical and Engineering Data 12, 183-189.
Tanji KK (1969) Solubility of gypsum in aqueous electrolytes as affected by ion association and ionic strengths up to 0.15 M and at 25[degrees]C. Environmental Science and Technology 3, 656-661.
Wong JWC, Ho GE (1993) Use of waste gypsum in the revegetation on red mud deposits: A greenhouse study. Waste Management and Research 11, 249-256. doi: 10.1006/WMRE.1993.1024
Manuscript received 12 March 2004, accepted 23 July 2004
P. M. Kopittke (A,C), N. W. Menzies (A), and I. M. Fulton (B)
(A) Centre for Mined Land Rehabilitation, University of Queensland, St Lucia, Qld 4072, Australia.
(B) EHS Department, Alcan Gove Pty Ltd, PO Box 21, Nhulunbuy, NT 0881, Australia.
(C) Corresponding author. Email: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Kopittke, P.M.; Menzies, N.W.; Fulton, I.M.|
|Publication:||Australian Journal of Soil Research|
|Date:||Dec 15, 2004|
|Previous Article:||Simulation of water and solute transport with MACRO model in Cecil loamy sand soil.|
|Next Article:||Absorption of artificial piggery effluent by soil: a laboratory study.|