3.2. WMA 2
Acidic wastewater produced after one acid leaching purification cycle of GRPW at 90 °C for 1 hour using a 5 wt% H
2SO
4 solution prepared with municipal water and stirring rate of 50 rpm was treated with Ca(OH)
2 to pH 5.5, pH 8.0 and pH 10.5. The precipitates were filtered and characterized through XRF (
Figure 3). At pH 5.5, the precipitate was mainly composed of CaSO
4 (63 wt%) and Fe
2O
3 (20 wt%). At pH 8.0, the amount of CaSO
4 in the precipitate increased to 75 wt% and the amount of Fe
2O
3 decreased to 10 wt%. The increase in pH from 5.5 to 8.0 also decreased the SiO
2 content from 10 wt% to 5 wt% and increased the MgO content from 1 wt% to 4 wt%. At pH 10.5, the precipitate was mainly constituted by MgO (47 wt%) and CaSO
4 (38 wt%).
In order to get a better understanding of the required pH for the complete precipitation of MgO, the precipitates obtained at pH 5.5, 8.0 and 10.5 were characterized through XRD.
Table 2 shows that the gypsum content of the precipitate was fairly similar when the pH was 5.5 and 8.0 (94.1-95.8 wt%) but decreased to 86.3 wt% when the pH was 10.5. The reduction in gypsum content with the increase in the pH value was accompanied by an increase in brucite, kieserite and portlandite. Brucite and kieserite were present in significant amounts at pH 10.5 (3.8 wt% and 3.5 wt%, respectively).
In previous studies [
27,
28,
29], MgSO
4 was reacted with Ca(OH)
2 to precipitate Mg(OH)
2 and CaSO
4.2H
2O according to the following equation:
The solubility of Mg(OH)
2 was found to be significantly lower than that of CaSO
4. However, all these studies have disagreed on the required pH to achieve the complete precipitation of Mg(OH)
2. For instance, Semerjian and Ayoub [
27] found that Mg started to precipitate as Mg(OH)
2 at approximately pH 9.5 and significant precipitation occurred at pH 10.5. On the other hand, Xiong et al. [
28] indicated that the precipitation rate began to increase with pH ranging from 8.0 to 10.5, with most of the Mg precipitated at pH 10.0. These authors also suggested that Mg(OH)
2 precipitation could be accompanied by co-precipitation of iron hydroxides, as found here at pH 5.5 (
Figure 3). However, Tolonen et al. [
29] found that pH 9.6 was too low for Mg(OH)
2 precipitation since 43 wt% of Mg was present as soluble MgSO
4 at pH 9.6. On the other hand, all Mg was precipitated as MgSO4 at pH 12.5. The fact that MgSO
4 was found to be highly soluble in water at pH 9.6 might suggest that MgSO
4.H
2O molecules could be trapped inside Mg(OH)
2 crystals formed at pH 10.5.
A comparison of the chemical composition of the precipitates obtained with Ca(OH)
2 at pH 10.5 from the acidic wastewater generated after acid leaching purification of GRPW and GDPW at 90 °C for 1 hour using a 5 wt% H
2SO
4 solution is presented in
Figure 4. The precipitate from GRPW contained similar CaSO
4 content that the precipitate from GDPW (38.2 wt% and 40.3 wt%, respectively). However, the MgO content in the precipitate from GDPW was around 4.5 wt% higher than that in the precipitate from GRPW. The Na
2O content in the precipitate from GDPW (4.8 wt%) was twice as high as that in the precipitate from GRPW (2.4 wt%). On the other hand, the precipitate from GRPW contained between 3-4 wt% of SiO
2, Al
2O
3 and Fe
2O
3, representing around 10.5 wt% of the precipitate, whereas the contents of these compounds were below 1 wt% in the precipitate from GDPW. As a result, the combined CaSO
4 and MgO contents in the precipitates from GRPW and GDPW were respectively around 85 wt% and 92 wt%. The ratio of MgO content to CaSO
4 content in both precipitates was close to 1.25.
In order to determine the impact of treated water reuse, acid leaching purification of GDPW was carried out at 90 °C for 2 hours using a 3 wt% H
2SO
4 solution prepared with purified water and stirring rate of 50 rpm. Under these acidic leaching conditions, the chemical purity of the resulting gypsum was above 96 wt%. However, these conditions are not optimum because the purity level after purification of different batches might not be consistent, as suggested by the lower chemical purity level of the samples in our previous work [
2]. The recycled GDPW was filtered, and the acidic wastewater was treated with Ca(OH)
2 to pH 10.5. The resulting precipitate was filtered, and the recovered treated water (treated water 1) was reused to prepare the 3 wt% H
2SO
4 solution for the next acid leaching purification test. This methodology was repeated once again to produce recycled GDPW using treated water 2.
Figure 5 presents the chemical purity of recycled GDPW when using purified water, treated water 1 and treated water 2. The results show that there was no variation in the chemical purity of the samples within experimental error. The CaSO
4 content in recycled gypsum dropped by 1.3 wt% when treated water 1 was used. However, the CaSO
4 content in recycled gypsum when treated water 2 was used (94.6 wt%) was similar to that when purified water was used (94.3 wt%).
The chemical composition of the precipitates from GDPW was determined through XRF (
Figure S2 in
Supplementary Materials). The MgO content in the precipitate increased from 18.0 wt% with purified water to 21.8 wt% with treated water 1 and to 26.2 wt% with treated water 2. In addition, the sum of CaSO
4 and MgO contents in the precipitates decreased from approximately 95 wt% when using purified water to 93 wt% when using treated water 1 and to 91 wt% when using treated water 2. There was also a simultaneous increase in Fe and Si compounds in the precipitate. XRD results in
Table 2 indicate that the gypsum content in the precipitates increased from 79.0 wt% with purified water to 81.7 wt% with treated water 1 and to 87.7 wt% with purified water 2. There were significant amounts of kieserite (6.1-9.6 wt%) and brucite (5.2-8.9 wt%) in the precipitates.
In further acid leaching gypsum purification tests, municipal water was used instead of purified water to determine its impact on process efficiency and precipitate composition. Acid leaching purification of GRPW was carried out at 90 °C for 2 hours using a 3 wt% H
2SO
4 solution prepared with municipal water and stirring rate of 50 rpm. Initially, the gypsum slurry was neutralized with Ca(OH)
2 to pH 5.5 and the resulting wastewater was treated with Ca(OH)
2 to pH 10.5 to recover the precipitate and treated water 1. Treated water 1 was used to prepare the 3 wt% H
2SO
4 solution of the second acid leaching purification test. In total, 6 cycles were performed with treated water. The chemical purity of the recycled GRPW samples obtained when using municipal water and reused treated waters was determined (
Figure 5). As for GDPW, the chemical purity of the samples was not affected by the reuse of the treated water, although the CaSO
4 content increased in the first 4 cycles. The low CaSO
4 content in the recycled GRPW when using treated water 5 could be due to the higher pH achieved in the gypsum slurry neutralization stage (pH 7.0 rather than pH 5.5). Nevertheless, the consistent chemical purity of the samples suggest that treated water obtained at pH 10.5 could be reused for at least 6 cycles without impacting the acid leaching gypsum purification process efficiency. In addition, the use of municipal water rather than purified water should not cause a reduction in the chemical purity of the recycled gypsum.
Table S1 in
Supplementary Materials presents the chemical composition of the precipitates obtained at pH 10.5 when using municipal water or treated water obtained in six acid leaching cycles. The main oxides in the precipitates were MgO (> 20 wt%), SO
3 (> 20 wt%) and CaO (> 12 wt%). Na
2O and Fe
2O
3 were also present in significant amounts in most precipitates (typically between 3-11 wt%), being higher than those found in the precipitates from GDPW (
Figure S2 in
Supplementary Materials).
Table 1 shows that the Na
2O content in GRPW (0.5 wt%) is higher than the Na
2O content in GDPW (< 0.1 wt%). On the other hand, the Fe
2O
3 content in GRPW and GDPW is similar (0.2-0.4 wt%). It must be noted that the acid leaching purification tests with GRPW were performed with municipal water, which may contain traces of Na and Fe salts. Therefore, the higher Na
2O and Fe
2O
3 contents in the precipitates from GRPW could be due to the different contents of Na and Fe compounds in GRPW and GDPW and to possible presence of these compounds in municipal water. SiO
2, Al
2O
3, and P
2O
5 contents were ≤ 4 wt%, Cl content was between 1-2 wt%, and K
2O and MnO contents were < 1 wt%. The MgO content in the precipitate increased in the first 2 cycles, similarly to the precipitates obtained from GDPW (
Figure S2 in
Supplementary Materials). However, the MgO content in the precipitates from subsequent cycles did not follow a pattern.
3.3. WMA 3
The addition of Ca(OH)
2 to neutralize the spent H
2SO
4 solution prior filtration would avoid the use of expensive corrosive-resistant pumps and filtration equipment. However, soluble impurities could precipitate during neutralization, reducing the chemical purity of the recycled gypsum. Two acid leaching purification tests were performed with GRPW at 90 °C for 1 hour using 5 wt% H
2SO
4 solutions prepared with purified water. In the first acid leaching purification test, the purified gypsum slurry was filtered and washed with purified water as per WMA 2. In the second acid leaching purification test, the purified gypsum slurry was neutralized with Ca(OH)
2 to pH 5.5 and then filtered without washing. Neutralization only caused reductions of 0.3 wt% in chemical purity and 0.8 wt% in CaSO
4 content compared to washing (
Figure 6). Therefore, neutralization seems to be a cheaper alternative to washing (high water consumption and wastewater production) to preserve the chemical purity of the gypsum product.
Acid leaching purification tests were then carried out in triplicate with GRPW at 90 °C for 2 hours using 3 wt% H
2SO
4 solutions, followed by neutralization with Ca(OH)
2 to pH 5.5 to verify the reproducibility of the results (
Figure 7).
The chemical purity of the gypsum samples showed good reproducibility (around 96.8 wt%). However, the CaSO4 content varied between 94.1 wt% and 94.7 wt%. Overall, it could be argued that the acid leaching-neutralization-filtration-precipitation approach (WMA 3) could be a viable industrial-scale process configuration for acid leaching purification of gypsum from post-consumer plasterboard wastes.
3.5. Comparison of WMA 3 with other acidic wastewater technologies
Several stepwise processes have been proposed to treat acidic wastewater. A three-step process was proposed by Zhang et al. [
30] to treat highly acidic wastewater derived from TiO
2 production. In the first step, high quality gypsum was produced after neutralization with CaCO
3 to pH 2. In the second step, schwertmannite (Fe
8O
8(OH)
8‒2x(SO
4)
x, x = 1-1.75) was formed by the reaction of FeSO
4 with H
2O
2 with stirring for 24 hours. In the third step, a NaOH solution was added to adjust the pH to 7.5, which resulted in the precipitation of metals as hydroxides after stirring for 24 hours. In another study, Salo et al. [
31] applied biological sulfate reduction to leachates obtained after acid leaching of phosphogypsum. This biological treatment produced a precipitate concentrating the rare earth elements present in phosphogypsum and converted SO
42− into S
2− in the liquid phase. However, the efficiency of the bioreactor was highly dependent on the acidity of the wastewater, which would limit process control and scalability, and a hydraulic retention time of 38 hours was required under optimum bioreactor conditions. Xiong et al. [
28] also developed a laboratory-scale process consisting of precipitation, acid leaching and oxidation steps to recover Mg(OH)
2 from a leachate of dolomitic phosphate ore. The leachate liquor was neutralized with Ca(OH)
2 to pH 7 to precipitate Fe(OH)
3. Then. filtered leachate was neutralized with Ca(OH)
2 to pH 10.0 to precipitate Mg(OH)
2. The Mg(OH)
2 product was further purified through acid leaching, oxidation, and precipitation steps. However, numerous chemicals [H
2SO
4, NH
4OH, Mg(OH)
2, (NH
4)
2S
2O
8 and NH
3] were needed to achieve high Mg(OH)
2 recovery yields, which would significantly increase material costs. In comparison, only H
2SO
4 and Ca(OH)
2 were required in WMA 3. Therefore, the novelty of the in-house wastewater treatment based on WMA 3 proposed in this work are the minimum economic impacts on the acid leaching gypsum purification plant whilst preserving the high purity (> 96 wt%) of the gypsum product, the valorization of the Mg-rich gypsum by-product, and the reduction of the environmental impact by reusing the treated water. Material costs could be reduced further by replacing commercial Ca(OH)
2 with by-products from quicklime manufacturing [
32]. The reuse of the treated water would adhere to the EU’s new Circular Economy Action Plan [
7], which promotes water reuse and efficiency in industrial processes.
3.6. By-product applications
The EU’s Directive 2008/98/EC31 classifies the acidic wastewater obtained after acid leaching purification of gypsum from consumer plasterboard wastes as a hazardous by-product because it is highly corrosive (pH < 1). Therefore, acidic wastewater must be neutralized to around pH 6 before it can be considered for commercial application. The neutralized wastewater, which would mainly contain magnesium and calcium sulfates, could be used as a liquid fertilizer but there are two major issues for its commercial exploitation. Firstly, the gypsum/solution ratio of 1:3 wt/wt used during acid leaching would imply that around 3000 liters of liquid fertilizer would be produced per ton of recycled gypsum. As a result, the water consumption in the acid leaching process would be extremely high, which would lead to high operating costs at the acid leaching purification plant. Secondly, this liquid fertilizer might not have enough demand from the agricultural industry because of the high volumes produced.
On the other hand, the precipitation of soluble impurities in wastewater at pH 10.5 not only offers the advantage of producing much lower amounts of solid fertilizer, but also treated water that can be reused without impacting the acid leaching gypsum purification process. The main disadvantages of precipitating the soluble impurities in the acidic wastewater at pH 10.5 are that around an extra 30 wt% of Ca(OH)
2 would be needed compared to acidic wastewater neutralization to pH 5-6, and additional equipment such as precipitation and storage tanks, filter press and pumps would be required. As shown in
Table 2, the precipitate obtained at pH 10.5 is a Mg-rich gypsum constituted by 79.0-87.7 wt% CaSO
4.2H
2O, 6.1-9.6 wt% MgSO
4.H
2O, 5.2-8.9 wt% Mg(OH)
2 and 1.0-4.0 wt% Ca(OH)
2. The Mg-rich gypsum can be classified as an inorganic macronutrient fertilizer [
33,
34] as it contains more than 1.5 wt% MgO, more than 1.5 wt% CaO and more than 1.5 wt% SO
3. Furthermore, Ca and Mg compounds in the precipitate are considered secondary nutrient fertilizers [
35]. Ritchey and Snuffer [
36] indicated that abandoned pasture soils are particularly likely to be low in Ca and Mg species. These authors used gypsum supplemented with 5-6 wt% Mg(OH)
2 to maintain adequate Mg levels in the soil of an abandoned Appalachian pasture. In another study, Ayanda et al. [
37] found that a Mg-rich gypsum with pH 8.8 that contained 45 wt% CaSO
4.2H
2O, 17.1 wt% Mg(OH)
2, 4.3 wt% Ca(OH)
2 and 2.3 wt% CaCO
3 was an effective source of Ca and Mg for oil palm growth and a good soil ameliorant. Oil palm is one of the world’s most important oil crops because it can produce more vegetable oil per unit of land than any other crop (e.g., soybean, rapeseed, sunflower) and it is currently being used as biofuel and as ingredient in many processed foods, cosmetics, pharmaceuticals, etcetera [
38]. The global area utilized for oil palm growth increased from less than 4 million hectares in 1980 to 20 million hectares in 2018, and future global demand for palm oil is expected to increase [
38]. The chemical composition of the Mg-rich gypsum used by Ayanda et al. [
37] is comparable to that of the precipitate recovered from the acidic wastewater after acid leaching gypsum purification. In addition, a mass balance of the laboratory-scale acid leaching purification process indicated that around 30 kg of precipitate would be produced per ton of gypsum waste processed, which is 100 times lower than the amount of liquid fertilizer that would be produced after neutralization of the acidic wastewater. Therefore, the commercialization of the Mg-rich gypsum as fertilizer for oil palm soils and reuse of the treated water in the acid leaching process are proposed as the most compelling sustainable solutions to preserve recycled gypsum quality and minimize waste production and disposal costs. An alternative commercial use for the alkaline Mg-rich gypsum obtained at pH 10.5 could be acidic soil ameliorant.