Method for recovering aluminium trihydroxide and gypsum as byproducts of sulphate-bearing wastewater treatment

Abstract

A method for treating sulphate-bearing wastewater that recovers and recycles aluminium trihydroxide used in ettringite precipitation. The method selectively dissolves a portion of ettringite to generate aluminium sulphate, which is used to regenerate aluminium trihydroxide from remaining ettringite. Gypsum containing entrained aluminium trihydroxide is recycled for further aluminium recovery.

Description

BACKGROUND OF THE INVENTION

[0001] This invention relates to a method of removing calcium and / or sulphate from wastewater streams such as Mine Impacted Wastewater (MIW).

[0002] International patent application no. PCT / GB98 / 01610 describes a process, generally referred to as "the SAVMIN process" which includes a method of treating sulphate bearing waste or effluent waters to reduce sulphate and calcium concentrations to levels suitable for potable water usage or environmentally safe disposal. These effluent waters also typically contain base metals and heavy metals which, in an initial step, are removed in a metal-hydroxide precipitation step conducted at an elevated pH (above 11). This initial treatment produces an aqueous decant that is saturated with calcium and sulphate (CaSO4), and contains approximately 1.5 g / L sulphate (SO4) i.e. a concentration too high for compliant disposal.

[0003] The SAVMIN process is fully described in the aforementioned patent application, and the contents of which are hereby incorporated by reference into this specification. The process allows for the effective removal of CaSO4 from aqueous decant with the use of amorphous aluminium trihydroxide (Al(OH)3), yielding mineral ettringite (Ca6Al2(SO4)3(OH)12.26H2O) based on the following reaction:2Al(OH)3 + 3CaSO4 + 3Ca(OH)2 + 26H2O → Ca6Al2(SO4)3(OH)2.26H2O

[0004] Studies have shown, however, that commercially available Al(OH)3 is chemically unreactive in the precipitation of CaSO4. As a result, it is necessary to generate amorphous Al(OH)3 in-situ. International patent application no. PCT / ZA2020 / 050049 describes a method of generating Al(OH)3 by contacting aluminium sulphate salt (Al2(SO4)3.18H2O) with a neutralizing reagent. Hydrated lime (Ca(OH)2) is typically used as neutralising agent, due to its compatibility with the process. However, the use of hydrated lime introduces a drawback: the low solubility of CaSO4, which leads to the formation of gypsum (CaSO4.2H2O) as shown in the following reaction:Al2(SO4)3.18H2O + 3Ca(OH)2 → 2Al(OH)3 + 3CaSO4.2H2O + 15H2O

[0005] To sustain the ettringite formation process the generated Al(OH)3 must be separated from gypsum using solid-solid separation to yield a product containing sufficient Al(OH)3 to precipitate incoming CaSO4 bearing effluent water into ettringite. The solid-solid separation, especially of the precipitate, is inherently imperfect, as its efficacy relies on the separation equipment employed and the particle size difference between Al(OH)3 and gypsum to effect successful separation. As a result, some gypsum is entrained in the recovered Al(OH)3 and some Al(OH)3 is lost with the rejected gypsum. The entrainment of gypsum in recovered Al(OH)3 results in in large stream of recovered Al(OH)3 being recycled to ettringite formation step, thereby requiring additional hydrated lime to effect CaSO4 precipitation. The loss of Al(OH)3 with the rejected gypsum, increases the demand of fresh Al2(SO4)3.18H2O salt to sustain the process. Both effects contribute to increased reagent consumption.

[0006] To sustain the process with amorphous Al(OH)3, the SAVMIN process includes an Al(OH)3 recovery step, wherein ettringite is decomposed using an acid to yield Al(OH)3 and gypsum, followed by solid-solid separation. Sulfuric acid (H2SO4) is typically selected due to its affordability, accessibility and its compatibility with the process. However, the use of H2SO4 to decompose ettringite introduces additional sulphate to the process, further driving the formation of gypsum, as shown in the following reaction:3H2SO4 + Ca6Al2(SO4)3 (OH)2. 26H2O → 2Al(OH)3 + 6CaSO4.2H2O + 26H2O

[0007] The increased gypsum formation exacerbates the challenges associated with solid-solid separation leading to greater entrainment of gypsum in recovered Al(OH)3. This, in turn, necessitates recycling a larger volume of Al(OH)3 to maintain the desired Al(OH)3 ratio for re-use in the CaSO4 precipitation step. This result in excessive reagent consumption in the ettringite formation and decomposition stages of the SAVMIN process, rendering the process economically unviable.

[0008] The invention aims to address, at least in part, the aforementioned limitations to improve the economic feasibility of calcium and sulphate removal from wastewater.SUMMARY OF INVENTION

[0009] The invention provides a method of treating a sulphate and calcium bearing wastewater stream, the method comprising:a) a calcium sulphate precipitation step, wherein CaSO4 containing effluent water is contacted with Al(OH)3 and a neutralising agent to precipitate ettringite;b) an Al(OH)3 dissolution step, wherein at least a portion of the precipitated ettringite is contacted with acid, primarily H2SO4, to produce aqueous aluminium sulphate (Al2(SO4)3) and waste gypsum;c) an ettringite decomposition step, wherein a remaining portion of ettringite is contacted with the produced Al2(SO4)3 recovered from waste gypsum, yielding Al(OH)3 and gypsum;d) a solid-solid separation step, wherein Al(OH)3 is separated from gypsum using suitable physical separation apparatus;e) recycling the rejected gypsum from the solid-solid separation step to the Al(OH)3 dissolution step wherein the gypsum is contacted with acid to dissolve entrained Al(OH)3 into Al2(SO4)3 solution for reuse in the ettringite decomposition step.

[0010] The recovered Al(OH)3 from the decomposition step (c) may be recycled for use in the calcium sulphate precipitation step (a).

[0011] The neutralising agent in the calcium sulphate precipitation step (a) may be hydrated lime.

[0012] The aluminium dissolution step (b) may be carried out according to the following reaction:6H2SO4 + Ca6Al2(SO4)3(OH)2.26H2O → Al2(SO4)3(aq) + 6CaSO4.2H2O + 26H2O

[0013] The Al(OH)3 dissolution step (b) may be controlled such that the produced Al2(SO4)3 solution is stoichiometrically sufficient to decompose the remaining portion of ettringite.

[0014] The solid-solid separation step (d) may be controlled such that, the Al(OH)3 entrained in rejected gypsum is stoichiometrically sufficient to generate enough Al2(SO4)3 solution required to decompose the precipitated ettringite in step (a).

[0015] The step (e) may be carried out using H2SO4 according to the following reaction:3H2SO4 + 2Al(OH)2+xCaSO4.2H2O + → Al2(SO4)3(aq) + xCaSO4.2H2O

[0016] The Al2(SO4)3 solution and waste gypsum produced in (e) may be separated by way of solid-liquid separation. The waste gypsum may be designated to discard.

[0017] The ettringite decomposition step (c) may proceed according to the following reaction:Al2(SO4)3 + Ca6Al2(SO4)3(OH)2.26H2O → 4Al(OH)3 + 6CaS) 4.2H2O + 26H2O

[0018] The portion of ettringite used in the Al(OH)3 dissolution step is dependent on the efficiency of the solid-solid separation of Al(OH)3 and gypsum produced in the ettringite decomposition step. If the recovery of Al(OH)3 is less than or equal to 50%, the gypsum (containing entrained Al(OH)3) recycled to the aluminium dissolution step, is sufficient to produce the amount of Al2(SO4)3 required for the decomposition of ettringite in the ettringite decomposition step (c).

[0019] Preferably the ettringite decomposition step is carried out at a pH of between 2.5 and 4.0. This ensures that the pH is low enough to fully dissolve Al(OH)3 into Al2(SO3)4 solution, and high enough to minimise free sulphate within the solution.

[0020] The method may be carried out at ambient temperature and pressure.DESCRIPTION OF THE DRAWINGS

[0021] The invention is further described by way of example with reference to the accompanying drawings wherein:Figure 1 is a flowsheet representation of a method according to the invention;Figures 2A is a Scanning Electron Microscopy (SEM) image of decomposed ettringite;Figure 2B is a SEM image of cyclone overflow products; andFigure 3 is a SEM image of a gypsum discard product.DESCRIPTION OF PREFERRED EMBODIMENT

[0022] A method 10 according to the invention aims to achieve the recovery and reuse of Al(OH)3 for use in the SAVMIN process, while reducing the sulphate load associated with sulfuric acid addition, thereby minimising the amount of gypsum recycled with recovered Al(OH)3. This ensures that less Al(OH)3 is required to react with incoming and entrained CaSO4.

[0023] The method 10 includes a calcium sulphate precipitation step (ettringite formation step) 12, wherein a calcium-sulphate containing wastewater stream 14 is contacted with Al(OH)3 16 and a neutralizing agent i.e. hydrated lime 18 to precipitate a low-sulphate solution 20 containing ettringite. Ettringite 22 is separated from the low-sulphate solution 20 using a suitable solid-liquid separation unit 24.

[0024] In a deviation from the SAVMIN process, a portion 22A of the separated ettringite 22 is reacted with sulfuric acid (H2SO4) 26 in an Al(OH)3 dissolution step 28 to provide aqueous aluminium sulphate (Al2(SO4)3) 30 and gypsum 32. The Al(OH)3 dissolution step 28 proceeds according to the following reaction:6H2SO4 + Ca6Al2(SO4)3(OH)2. 26H2O → Al2(SO4)3(aq) + 6CaSO4.2H2O + 26H2O

[0025] The gypsum 32 is separated from the Al2(SO4)3 30 through solid-liquid separation 34 and is discarded, which allows the complete recovery of aluminium solution.

[0026] In contrast to the SAVMIN process, wherein Al(OH)3 is recovered for use in the process using sulfuric acid, the recovered Al2(SO4)3 30 is used in an ettringite decomposition step 36, wherein the remaining portion of ettringite 22 is decomposed to yield Al(OH)3 16 and gypsum 38 according to the following reaction:Al2(SO4)3 + Ca6Al2(SO4)3(OH)2. 26H2O → 4Al(OH)3 + 6CaSO4.2H2O + 26H2O

[0027] The Al(OH)3 16 and gypsum 38 are separated from each other using a suitable solid-solid separation unit 40. The Al(OH)3 16 is recycled to the ettringite formation step 12, to precipitate incoming sulphate bearing wastewater 14.

[0028] The recovered gypsum 38 is sent to the Al(OH)3 dissolution step 28 to recover entrained Al(OH)3 as Al2(SO4)3 30.

[0029] The split ratio of the portion 22A of ettringite 22 which is used in the Al(OH)3 dissolution step 28 is dependent on the efficiency of the solid-solid separation 40 of Al(OH)3 16 and gypsum 38 in the ettringite decomposition step 36.

[0030] If an aluminium recovery of less than or equal to 50% is achieved, the use of a portion 22A of ettringite 22 in the Al(OH)3 dissolution step 28 is unnecessary and the use of recovered gypsum 38 in the Al(OH)3 dissolution step 28 is sufficient to provide the aluminium required for the process.

[0031] The method of the invention employs solid-liquid separation in the Al(OH)3 dissolution step to recover Al2(SO4)3 solution, which obviates the loss of Al(OH)3 through solid-solid separation. Additionally, the use of Al2(SO4)3 solution which is recovered as a byproduct, to decompose ettringite, yields gypsum and Al(OH)3 which constitute the exact product required for the formation of ettringite, thus allowing for the recovery of a much purer Al(OH)3 stream in solid-solid separation step.

[0032] The method of the invention can also easily be integrated with existing acid neutralisation and metal precipitation processes. As the gypsum produced as a waste is relatively pure, it can be utilised in other commercial applications.EXPERIMENTAL RESULTS

[0033] The performance of the method of the invention was tested by the applicant using Mine Impacted Wastewater (MIW) sourced at Witbank coal mine in Mpumalanga Province of South Africa to validate efficacy of the invention. The following diagram shows the experimental design followed to generate the test results:

[0034] To initiate the test work, 82 L of MIW was pre-conditioned by neutralising the MIW with hydrated lime to a target pH of 11. The supernatant of neutralised MIW was decanted and used in subsequent test work, whilst the sludge was drained and disposed of safely.Generation of Al(OH)3

[0035] To start the CaSO4 precipitation test work, Al(OH)3 was prepared by contacting Al2(SO4)3 solution containing 5 g / L Aluminium (Al) with hydrated lime to achieve a target pH of 8.2. The Al2(SO4)3 solution was prepared by dissolving 1.8 kg of Al2(SO4)3.18H2O salt in 30 L of MIW decant. A lime slurry was prepared by mixing 1.2 kg of hydrated lime with 5 L of MIW decant.

[0036] The resulting precipitate, a mixture of Al(OH)3 and gypsum, was subject to solid-solid separation using a hydro-cyclone (Cycle 1).

[0037] The cyclone underflow, presumed to contain mostly gypsum, was used in Cycle 1 Al2(SO4)3 regeneration, while the overflow, presumed to contain mostly Al(OH)3, was used in Cycle 1 CaSO4 precipitation.Aluminium Sulphate Regeneration (Cycle 1)

[0038] To regenerate Al2(SO4)3, the cyclone underflow was contacted with H2SO4 to lower the pH to 2.8. The resulting slurry was subjected to solid-liquid separation using a pressure filter. The filtrate containing Al2(SO4)3 was recovered for use in the ettringite decomposition, while the filtered cake, presumed to be gypsum, was designated for discard.Calcium Sulphate Precipitation (Cycle 1)

[0039] The pH of the cyclone overflow slurry containing Al(OH)3 was adjusted and maintained to a pH of 11.8 using hydrated lime. Neutralised MIW decant was added to react with excess Al(OH)3 with the pH maintained at 11.8 through the addition of hydrated lime. The resulting precipitate was filtered using a vacuum filter. The filtrate, containing reduced sulphate was collected as Cycle 1 CaSO4 free water, while the filtered cake, presumed to be ettringite, was used as a feed to Cycle 1 ettringite decomposition.Ettringite Decomposition (Cycle 1 and 2)

[0040] The aluminium sulphate solution and the ettringite cake generated in Cycle 1 were used in the Cycle 1 ettringite decomposition. To decompose the ettringite, the ettringite cake was pulped with a portion of Al2(SO4)3 solution and an additional Al2(SO4)3 solution was added to adjust the pH to 8.2. The resultant precipitate was passed through the cyclone (Cycle 2 solid-solid separation). As with Cycle 1, the cyclone underflow, presumed to contain mostly gypsum, was allocated to Cycle 2 Al2(SO4)3 regeneration, while the overflow, presumed to contain mostly Al(OH)3, was used in Cycle 2 CaSO4 precipitation.Aluminium Sulphate Precipitation (Cycle 2)

[0041] To precipitate the CaSO4 in Cycle 2, hydrated lime was added incrementally into 10 L of neutralised MIW decant at target hydrated lime additions of 2.50 g / L, 3.50 g / L, 4.50 g / L, 5.00 g / L and 7.00 g / L. Cycle 2 cyclone overflow slurry was added to adjust the pH to 11.8 in each lime increment, with solution sample been taken after each lime increment.Aluminium Sulphate Regeneration (Cycle 2)

[0042] Five Al2(SO4)3 regeneration tests were conducted using Cycle 2 cyclone underflow slurry. The volumes of cyclone underflow slurry used in each test were based on the cyclone split volumes and the volume of Cycle 2 cyclone overflow slurry used to adjust the pH in each lime increment. To regenerate Al2(SO4)3, each cyclone underflow was contacted with H2SO4 to lower the pH to 2.8, followed by solid-liquid separation using pressure filter. The filtered cake of each test, presumed to be gypsum, were safely discarded, while the filtrates, presumed to be Al2(SO4)3 solution, were used in subsequent ettringite decomposition tests.Results

[0043] The reported results are for MIW neutralisation and Cycle 2 test work. All reported samples were analysed using Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) with a detection limit of 0.05% and 2.0 mg / L for solid and solution, respectively. Sulphur was analysed through combustion (“LECO").

[0044] Presented in Table 1 are the conditions and analytical results of MIW neutralisation test:Table 1. Mine Impacted Wastewater neutralisation test resultsConditionsAnalytical resultsinitial pH, Ag / AgCl4.07AICaFeMgMnSSiFinal pH, Ag / AgCl11.0MIW, mg / L10.5475<2.01606.428248.39Lime addition, g / L1.20MIW decant, mg / L<2.0701<2.0<2.0<2.05423.95Precipitate, g / L2.02

[0045] The major elements detected in the MIW were Calcium (Ca), Magnesium (Mg) and Sulphur (S) at a concentration of 475 mg / L, 160 mg / L and 824 mg / L, respectively. To increase the pH of MIW from an initial pH of 4.07 to pH of 11.0, the MIW required 1.20 g of hydrated lime per litre of MIW treated. The concentrations of most elements in the neutralised MIW decant were below ICP-OES detection limit of 2 mg / L, except for Ca, S and Silicon (S) which reported concentrations of 701 mg / L, 542mg / L and 3.95 mg / L, respectively.

[0046] Presented in Table 2 are the conditions and analytical results of Cycle 2 ettringite decomposition test.Table 2. Cycle 2 ettringite decomposition test resultsConditionsMass of ettringite sludge5.96kgEttringite sludge solid concentration12.5%(m / m)Volume of Al2(SO4)3 solution used17.0LFinal pH8.18AglAgCIAnalytical resultsAICaFeMgMnSSiEttringite solid, %4.0220.30.320.76<0.058.040.54Al2(SO4)3 solution, mg / L586051584.4173<212200114Decomposed ettringite solid, %7.1416.10.360.67<0.0513.20.93

[0047] Cycle 2 ettringite decomposition was conducted using ettringite sludge and the Al2(SO4)3 solution generated in the Cycle 1 test. The ettringite sludge required 17.0 L of Al2(SO4)3 solution to attain a pH of 8.18. The use of Al2(SO4)3 to decompose the ettringite, increased the grade of Al from 4.02% reported in the ettringite solid, to 7.14% reported in the decomposed ettringite. The grade of S was significantly reduced from 20.3%, reported in the ettringite to 16.1%, in decomposed ettringite solid. The stoichiometric grade of Al and S required to form ettringite is 8.04% and 14.3%, respectively. The slightly lower than expected grade of Al and higher grade of S reported in decomposed ettringite can be attributed to excess S in the Al2(SO4)3 solution, which reported S concentration of 12.2 g / L. The excepted stoichiometric concentration of S at 5.86 g / L Al is 10.5 g / L. These results assert the efficacy of using regenerated Al2(SO4)3 solution to decompose ettringite, yielding an Al and S product that constitutes the required stoichiometric ratio to form ettringite.

[0048] Presented in Table 3 are the conditions and analytical results of Cycle 2 solid-solid separation test work. The solid-solid separation test was conducted using FCY25 cyclone operating at 125 kPa feed inlet pressure with 6 mm spigot and 8 mm vortex finder.Table 3. Cycle 2 solid-solid separation test resultsConditionsVolume of cyclone overflow, L13.0LCyclone overflow solid mass pull, %29.2%(m / m)Volume of cyclone underflow, L8.50LCyclone underflow solid mass pull, %70.8%(m / m)Analytical resultsAlCaFeMgSSiCyclone overflow solids, %13.28.760.430.2910.01.05Cyclone underflow solids, %6.1915.60.230.1914.20.79Mass balance resultsAICaFeMgSSiCyclone overflow recovery, %46.618.743.338.422.335.2Cyclone underflow recovery, %53.481.356.761.677.764.8

[0049] The solid-solid separation test increased the grade of Al from 7.14% reported in the Cycle 2 ettringite decomposition solid to 13.2% reported in the Cycle 2 cyclone overflow. The grade of Ca and S decreased from 16.1% and 13.2 %, respectively, to 8.76% and 10%. These results assert the efficacy of separating Al(OH)3 from gypsum, yielding Al(OH)3 product (cyclone overflow) that contain sufficient Al(OH)3 to precipitate entrained gypsum and incoming CaSO4 solution into ettringite.

[0050] Figure 4A and Figure 4B, respectively shows the Scanning Electron Microscopy (SEM) images of decomposed ettringite and cyclone overflow products. The pink colour shows the gypsum phases, and the green colour shows the Al(OH)3 phases.

[0051] Data presented in Table 4 shows the conditions and analytical results of the Cycle 2 CaSO4 precipitation test at varying lime addition. The test was conducted using the Cycle 2 cyclone overflow slurry containing 25 g / L solid. The actual lime increments of 2.26 g / L, 3.09 g / L, 3.78 g / L, 4.04 g / L and 5.46 g / L were calculated based on the initial WIM decant volume of 10 L used and the volume of cyclone overflow slurry added.Table 4. Calcium sulphate precipitation test resultsConditionsLime increment, g / LActual lime increment, g / LVolume of cyclone overflow added, L2.502.261.403.503.092.004.503.782.705.004.043.407.005.464.90Analytical resultsLime addition, g / LAl, mg / LCa, mg / LS, mg / LSi, mg / L0.0<27015423.952.26<22061703.843.09<212697.42.643.786.7190.761.37.224.04<239.254.152.645.4635.956.717.23.41

[0052] The results show a progressive decrease in Ca and S concentration with increase in lime and cyclone overflow addition. A significant decrease was reported at 2.26 g / L lime and 3.07 g / L Al(OH)3 additions. These results assert the efficacy of using recovered Al(OH)3 to precipitate CaSO4, with the concentration of S reaching 17.2 mg / L with lime and Al(OH)3 addition of 5.46 g / L and 8.22 g / L, respectively.

[0053] Presented in Table 5 are the conditions and analytical results of Cycle 2 Al2(SO4)3 regeneration test work. The test was conducted using Cycle 2 cyclone underflow slurry containing 58 g / L solids. The acid consumption was calculated based on the total volume of neutralised MIW decant treated, which included the volume of cyclone overflow slurry used to precipitate the CaSO4 in each lime increment. The reported analytical results are for test work conducted using the cyclone underflow slurry calculated based on 5.46 g / L actual lime increment.Table 5. Aluminium sulphate regeneration test resultsConditionsCyclone underflow volume used, LAcid consumption g / L0.953.371.354.611.835.562.305.923.327.88Analytical resultsUnitAICaFeMgSSiAl2(SO4)3 solution @ 8.0 g / L acid consumptionmg / L575055654443812.7221Gypsum @ 8.0 g / L acid consumption%0.5821.90.220.1417.20.39

[0054] These results assert the notion of recovering entrained Al(OH)3 in rejected gypsum (cyclone underflow) as Al2(SO4)3 solution by dissolving it with H2SO4, thereby minimising the loss of aluminium with gypsum discard. Figure 3 shows the SEM image of gypsum, containing 0.58% Al, which is significantly lower than 6.19% Al grade reported in Cycle 2 cyclone underflow solid.

Claims

1. A method of treating a sulphate and calcium bearing wastewater stream to reduce aluminium loss with gypsum discard and control gypsum generation during ettringite decomposition, the method comprising:(a) a calcium sulphate precipitation step, wherein CaSO4 containing effluent is contacted with Al(OH)3 and a neutralising agent to precipitate ettringite;(b) an Al(OH)3 dissolution step, wherein at least a portion of the precipitated ettringite is contacted with sulphuric acid (H2SO4) to produce aqueous aluminium sulphate (Al2(SO4)3) and gypsum;(c) an ettringite decomposition step, wherein a remaining portion of the ettringite is contacted with the Al2(SO4)3 produced in step (b), to yield Al(OH)3 and gypsum;(d) a solid-solid separation step, wherein Al(OH)3 is separated from gypsum; and(e) recycling the gypsum rejected in step (d), comprising entrained Al(OH)3, to step (b), wherein it is contacted with acid to dissolve Al(OH)3 and form additional Al2(SO4)3 for reuse in step (c).

2. A method according to claim 1 wherein the Al(OH)3 recovered in step (c) is recycled to step (a).

3. A method according to claim 1 wherein the neutralising agent in step (a) comprises hydrated lime (Ca(OH)2).

4. A method according to any one of claims 1 to 3 further comprising a solid-liquid separation step following step (e), wherein the Al2(SO4)3 solution is separated from the gypsum, and the gypsum is discarded.

5. A method according to any one of claims 1 to 4 wherein the amount of Al2(SO4)3 produced in step (b), including that from recycled gypsum in step (e), is stoichiometrically sufficient to decompose the remaining portion of ettringite in step (c).

6. A method according to any one of claims 1 to 5, wherein the amount of ettringite subjected to step (b) is determined based on the recovery efficiency of Al(OH)3 in step (d).

7. A method according to claim 6 wherein if the recovery efficiency of Al(OH)3 in step (d) is less than or equal to 50%, the Al(OH)3 entrained in the gypsum recycled in step (e) is sufficient to produce the required Al2(SO4)3 to decompose the remaining ettringite in step (c).

8. A method according to any one of claims 1 to 7 wherein the pH of the ettringite decomposition step (c) is maintained between 2.5 and 4.0.

9. A method according to any one of claims 1 to 8 wherein all steps are carried out at ambient temperature and pressure.

Images