ARTICLE

J. Cent. South Univ. (2019) 26: 361-372

DOI: https://doi.org/10.1007/s11771-019-4008-8

Variation of alkaline characteristics in bauxite residue under phosphogypsum amendment

LI Yi-wei(李义伟)^{1, 2}, LUO Xing-hua(罗兴华)¹, LI Chu-xuan(李楚璇)¹,Graeme J. MILLAR³, JIANG Jun(江钧)¹, XUE Sheng-guo(薛生国)¹

1. School of Metallurgy and Environment, Central South University, Changsha 410083, China;

2. Hunan Rare Earth Metal Material Research Institute, Changsha 410126, China;

3. Institute for Future Environments, Science and Engineering Faculty, Queensland University of Technology (QUT), Brisbane Qld 4000, Australia

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2019

Abstract: Aiming at alkaline problem of bauxite residue, this work focused variation of alkaline characteristics in bauxite residue through phosphogypsum treatment. The results demonstrated that the pH of bauxite residue reduced from initial 10.83 to 8.70 when 1.50 wt% phosphogypsum was added for 91 d. The removal rates of free alkali and exchangeable sodium were 97.94% and 75.87%, respectively. Meanwhile, significant positive correlations (P<0.05) existed between pH and free alkali, exchangeable sodium. The effect of free alkali composition was CO₃^2–>OH^–> AlO₂^–>HCO₃^–. In addition, alkaline phase decreased from 52.81% to 48.58% and gypsum stably presented in bauxite residue which continuously provided Ca²⁺ to inhibit dissolution of combined alkali. Furthermore, phosphogypsum promoted formation of macroaggregate structure, increased Ca²⁺, decreased Na⁺ and Al³⁺ on the surface of bauxite residue significantly, ultimately promoting soil formation in bauxite residue.

Key words: bauxite residue; alkaline regulation; free alkali; phosphogypsum amendment; soil formation in bauxite residue

Cite this article as: LI Yi-wei, LUO Xing-hua, LI Chu-xuan, Graeme J. MILLAR, JIANG Jun, XUE Sheng-guo. Variation of alkaline characteristics in bauxite residue under phosphogypsum amendment [J]. Journal of Central South University, 2019, 26(2): 361–372. DOI: https://doi.org/10.1007/s11771-019-4008-8.

1 Introduction

Alumina produced from bauxite ore via the Bayer process is an important raw material for production of strategic aluminum metal. However, limited by effective technical methods to recovery comprehensively, bauxite residue generated during alumina production process continues to be a problem in terms of environmental impact. Traditionally, bauxite residue is only deposited in bauxite residue disposal areas (BRDAs) [1]. The global accumulative bauxite residue inventory reached approximately 4.60×10⁹ t in 2018 with an annual production rate of approximately 2.00×10⁸ t, which become the largest emission of industrial waste in non-ferrous metal smelting industry [2, 3]. As such, BRDAs occupied large areas of land, seriously polluting the surrounding soil and water (especially if dam-failure occurs) [4–6]. According to reports, China, Ajka, Odishaand Bc Kan suffered many serious pollution incidents due to dam-failure, which seriously threaten the surrounding environment safety and residents’lives [7, 8].

Bauxite residue exhibited high alkaline and complex composition feature. Due to the inherent challenges of using such a corrosive material, global utilization rate is presently less than 10% [8]. Alkaline substances in bauxite residue are mainly classified as soluble alkali (free alkali) and insoluble alkali (combined alkali) types [9, 10]. Soluble alkali is formed during the dissolution of bauxite in high concentration caustic solution which could be removed by washing and neutralization. Insoluble alkali is generated in pre-desiliconization, high-pressure dissolution and sedimentation separation with the characteristics of low solubility, high content, strong buffering capacity, complicated chemical structure and difficulty to be removed, which were discharged by precipitation and mineralization [11, 12]. Insoluble alkali may in theory continuously replenish free alkali by slowly dissolving, thus maintaining bauxite residue in high pH environment for a relatively long time [13]. The problem of how to mitigate bauxite residue alkalinity has not been resolved reasonably and seriously restricts the sustainable development of alumina industry. Thus, if the alkaline of bauxite residue could be solved, then new strategy may promote soil formation and ecological disposal of bauxite residue in BRDAs.

At present, one option for reduction of the alkaline of bauxite residue is gypsum (CaSO₄) addition. BARROW [14] revealed that Ca²⁺ preferentially precipitated with alkaline carbonate when gypsum was added into bauxite residue. Subsequently, calcium–sodium replacement with combined alkali occurred. Gypsum could also precipitate with free hydroxide, carbonate and aluminate in liquid phase to form calcium hydroxide, calcium carbonate, tricalcium aluminate and hydrocalumite. As a result, significant reductions in the pH, electrical conductivity (EC), and contents of Na and Al were achieved. The accompanying continuous supply of Ca²⁺ for the liquid phase system decreased exchangeable sodium and provided favorable conditions for plant growth [13, 15]. Furthermore, SHI et al [16] demonstrated that gypsum could significantly reduce hydroxyl free alkali in bauxite residue and thus plant growth was further promoted. XUE et al [17] established that the pH of bauxite residue sand system could be reduced to 8.0 and the measured clover biomass was high when gypsum addition was 3 wt%. Gypsum could not only effectively regulate alkaline and reduce pH, but also increase contents of Ca and Mg in bauxite residue, improving physicochemical properties, which accelerated plant growth. COURTNEY et al [18] used gypsum and organic matter to ameliorate bauxite residue, reducing the pH from 12.5 to 8.1 and the effective Al content from 26.0 to 1.3 mg/kg. Gypsum could also inhibit the dissolution of tricalcium aluminate (TCA) and reduce the content of effective Al in the liquid phase. In summary, gypsum improvement relates to factors including:1) low solubility and mobility of gypsum; 2) continuous provision of Ca²⁺ in solution which precipitates with alkaline anions and additionally exchanges sodium with combined alkali [8]:

Ca²⁺+2OH^–=Ca(OH)₂                                  (1)

Ca²⁺+CO₃^2–=CaCO₃                                    (2)

3Ca²⁺+4OH^–+2Al(OH)₄^–=Ca₃Al₂(OH)₁₂           (3)

4Ca²⁺+4OH^–+2Al(OH)₄^–+CO₃^2–=Ca₃Al₂(OH)₁₂·CaCO₃·5H₂O             (4)

6Ca²⁺+4OH^–+2Al(OH)₄^–+3SO₄^2–=Ca₃Al₂(OH)₁₂·3CaSO₄·26H₂O           (5)

Actually, ideally treatment of bauxite residue should best be conducted by a low cost approach due to the vast amount of material present. Consequently, application of an acidic waste material to mitigate the negative effects of bauxite residue is sensible. Phosphogypsum is an acidic industrial waste generated in phosphoric acid production process and the main component is CaSO₄·2H₂O [19–21]. At present, the annual global emission of phosphogypsum reaches 2.8×10⁸ t, while the utilization rate is only about 20% [22]. Hence, phosphogypsum disposal also represents an environmental liability and economic cost. Therefore, use of phosphogypsum instead of gypsum for alkaline regulation of bauxite residue is attractive as it is not only abundant and inexpensive but also reduces the environmental impact of two waste products.

However, as indicated above, insoluble alkali can slowly dissolve to replenish soluble alkali. Consequently, there is a need to explore the ability of phosphogypsum to regulate bauxite residue alkalinity under phosphogypsum treatment for longer periods. Therefore, the aim of this study was to examine the control of alkalinity in bauxite residue using phosphogypsum based on soil culture experiments. The objectives of the present study were to investigate: 1) the variation in characteristics of various alkaline indicators relating to efficient stabilization of free alkali; 2) the transformation of bauxite residue phase with time including the characteristics of mineral composition, phase microscopic morphology, and the transformation relationship between combined alkali and soluble alkali.

2 Materials and methods

2.1 Bauxite residue and phosphogypsum preparation

Raw bauxite residue used in this study was collected from Guangxi Pingguo Aluminum Refinery, China. The sample was air dried and sieved (830 μm) prior to use. The main chemical composition of bauxite residue is presented in Table 1.

Table 1 Chemical composition of bauxite residue (mg·kg^–1)

Raw phosphogypsum was collected from a phosphogypsum yard in Guizhou, China. The sample was uniformly mixed, dried at 105 °C for 48 h and sieved (150 μm). The chemical composition and identity of the crystalline phases are shown in Table 2 and Figure 1, respectively, indicating that the main chemical composition of phosphogypsum is CaO, SO₃ and the content of CaSO₄·2H₂O in phosphogypsum is over 90%.

Table 2 Chemical composition of phosphogypsum (wt%)

2.2 Experimental design

Based on soil culture experiment, 500 g of bauxite residue sample was placed in a polypropylene plastic bowl, to which were added 0,0.50 wt%, 1.00 wt%, 1.50 wt%, 2.00 wt%,2.50 wt%, 3.00 wt%, 3.50 wt% and 4.00 wt% phosphogypsum. Then these materials were thoroughly mixed to ensure homogenization of the phosphogypsum and bauxite residue. After adding 150 mL of distilled water, the mixture was agitated with a glass rod and maintained at room temperature. Deionized water was added periodically as required in order to ensure a constant water content. Samples were collected for analysis after 1, 2, 3, 4, 5, 6, 7, 14, 28, 42, 56, 70, and 91 d.

Figure 1 XRD patterns collected from phosphogypsum

2.3 Sample analysis

The collected samples were slightly broken apart, then naturally dried for one week. Finally, they were ground, sieved (150 μm). Then leachate was prepared with a ratio of bauxite residue to water of 1:5. The pH of leachate was measured by a PHS-3C pH meter and the OH^– concentration was thus calculated. The electrical conductivity (EC) was measured by a conductivity meter (Ray- magnetic DDS-307). Carbonate (CO₃^2–) and bicarbonate (HCO₃^–) were determined by double indicator-neutralization titration [8]. The concentration of Al³⁺ was determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, Optima 5300DV, Perkin Elmer, USA) and converted to AlO₂^– content. The exchangeable sodium content was determined by NH₄OAc–NH₄OH flame photometry [30]. The solid phase component and surface properties of bauxite residue were analyzed by X-ray diffraction (XRD), scanning electron microscopy (SEM) and energy dispersive spectrometry (EDS) [8].

2.4 Data analysis

All data shown in the tables and figures were analyzed in Microsoft Office Excel 2010 and Origin 9.0. Data processing was operated using SPSS Statistics 20. Different letters were verified using the ANOVA tests by least-significant difference (LSD) (P<0.05).

3 Results and discussion

3.1 Effect of phosphogypsum addition on pH and EC

The pH change of bauxite residue with different phosphogypsum additions (0–4.00 wt%) is shown in Figure 2. In general, the greater the quantity of phosphogypsum added, the lower the pH of bauxite residue system. As the addition of phosphogypsum increasing, the pH decreased continuously. When phosphogypsum addition was 1.50%, the pH of bauxite residue decreased from 10.83 to 9.29 on the first day, and then the change was slow and flat. When phosphogypsum addition was 4.00%, the pH dropped to 8.80. With the increase of time, the pH showed a downward trend as a whole, but the change was not obvious. Finally, the pH of raw bauxite residue, bauxite residue with phosphogypsum addition of 1.50%, bauxite residue with phosphogypsum addition of 4.00% maintained at about 11.00, 8.70 and 8.55, indicating that the phosphogypsum could significantly reduce the pH of bauxite residue with long-term stability and phosphogypsum addition of 1.50% was the optimal amount for alkaline regulation of bauxite residue.

The corresponding EC change of bauxite residue with different phosphogypsum additions (0–4.00 wt%) is shown in Figure 3, which was used to measure the concentration of soluble salts in bauxite residue. As the addition of phosphogypsum increasing, the EC of bauxite residue generally increased. This effect was ascribed to Ca²⁺ and SO₄^2– increasing the soluble salt concentration in bauxite residue. Moreover, the presence of phosphogypsum was expected to improve the physical properties and increased the permeability of bauxite residue, which was also beneficial with respect to the dissolution of sodium, aluminum and other soluble salts [18, 23]. Since the phosphogypsum was slightly soluble, solid phase continuously dissolved to promote the unreacted phosphogypsum reacting with bauxite residue. Finally, bauxite residue and phosphogypsum reached a reaction equilibrium state in liquid phase, and the EC remained stable.

Figure 2 pH for supernatants of bauxite residue transformed by phosphogypsum

Figure 3 EC for supernatants of bauxite residue transformed by phosphogypsum

3.2 Effect of phosphogypsum addition on alkaline anions

The changes of CO₃^2–, HCO₃^–, AlO₂^–, and OH^– contents in bauxite residue with different phosphogypsum additions are shown in Table 3. The contents of CO₃^2–, HCO₃^–, AlO₂^– and OH^– in bauxite residue without phosphogypsum treatment were 141.32, 38.28, 88.03 and 3.38 mmol/kg in the first sampling. When phosphogypsum addition was 1.50%, alkaline anions decreased to 3.98, 3.67, 1.75, 0.12 mmol/kg, respectively. Alkaline anions in bauxite residue treated by phosphogypsum gradually decreased and finally stabilized with the increase of time. On the 91st day, alkaline anions were stable at 0, 3.75, 1.75 and 0.02 mmol/kg. CO₃^2– reacted completely and HCO₃^–, AlO₂^– and OH^– also decreased significantly with phosphogypsum addition of 1.50%. Phosphogypsum had a significant influence on the distribution of alkaline anion components in bauxite residue, which could promote a series of ion conversions and precipitation reactions in the liquid phase.

Ca²⁺+CO₃^2–=CaCO₃                           (6)

H⁺+OH^–=H₂O                            (7)

3Ca²⁺+4OH^–+2Al(OH)₄^–=Ca₃Al₂(OH)₁₂           (8)

6Na⁺+2Ca²⁺+6H₄SiO₄+6Al(OH)₃+2HCO₃^–+8OH^–=[Na₆Al₆Si₆O₂₄]·2[CaCO₃]+26H₂O        (9)

3Ca²⁺+(6–4x)OH^–+xH₄SiO₄+2Al(OH)₃=Ca₃Al₂(SiO₄)_x(OH)_12–4x                         (10)

Phosphogypsum introduced a large amount of Ca²⁺, which could react with CO₃^2– to form CaCO₃ and rapidly form calcite to stably present in bauxite residue. When phosphogypsum addition was more than 1.50%, excess Ca²⁺ reacted with CO₃^2– to completely consume CO₃^2–. Due to the conversion process between HCO₃^– and CO₃^2–: HCO₃^–+OH^–CO₃^2–+H₂O, HCO₃^– continuously supplemented the consumed CO₃^2–. However, the continuous dissolution of Ca²⁺ promoted the conversion of HCO₃^– to CO₃^2– due to the slight dissolution of phosphogypsum, which resulted in decreases in HCO₃^–, OH^– and pH of bauxite residue. Simultaneously, excessive Ca²⁺ reacted with AlO₂^– to form Ca₃Al₂(OH)₁₂, which formed a mineral with low solubility of tricalcium aluminate and reduced AlO₂^– [24]. When the addition of phosphogypsum exceeded 1.50%, alkaline anions in bauxite residue changed indistinctively. Phosphogypsum reached chemical equilibrium in solid phase system of bauxite residue, consumed a large number of alkaline anions. Alkaline anions decreased with the increase of time during the continuous cultivation process in 91 d, indicating that phosphogypsum reduced alkaline anion in bauxite residue with long- term stability.

3.3 Effect of phosphogypsum addition on free alkali and exchangeable sodium

Correlations between the pH of bauxite residue with not only free alkali (Free alkali=c(HCO₃^–)+ c(CO₃^2–)+c(AlO₂^–)+c(OH^–)) but also exchangeable Na are shown in Figure 4. The relatively positive correlation between free pH and alkali (R²=0.7833, P<0.05), and also pH and exchangeable sodium (R²=0.7030, P<0.05) indicated that both free alkali and exchangeable sodium could objectively characterize alkalinity of bauxite residue and quantitatively describe the strength of bauxite residue alkaline which provided new indicators for alkaline regulation of bauxite residue.

Table 3 Alkaline anions for supernatants of bauxite residue transformed by phosphogypsum (mmol·kg^–1)

Figure 4 Relationship between pH with free alkali (a) and exchangeable Na of bauxite residues (b)

In addition, the correlations between free alkali and each alkaline anion in bauxite residue are shown in Figure 5. HCO₃^– and pH had a significant positive correlation (R²=0.8015, P<0.05); CO₃^2– and pH showed had a significant positive correlation (R²=0.9890, P<0.05); AlO₂^– and pH had a significant positive correlation (R²=0.9479, P<0.05); OH^– and pH had a significant positive correlation (R²=0.9481, P<0.05). Based upon the R² values for goodness of fit, it was determined that the importance of factors affecting free alkali was CO₃^2–>OH^–>AlO₂^–>HCO₃^–, of which CO₃^2– dominated.

The change of free alkali in bauxite residue with different phosphogypsum additions is shown in Table 4. As phosphogypsum addition increasing, the effect of reducing free alkali in bauxite residue was obvious. At the first sampling, free alkali in bauxite residue without phosphogypsum treatment was 267.28 mmol/kg. When phosphogypsum addition was 1.50%, free alkali decreased to 11.42 mmol/kg, which significantly reduced free alkali in bauxite residue. The raw bauxite residue promoted the dissolution of solid phase alkali as the time increasing, due to the continuous addition of water during soil culture process, resulting in an increase in free alkali [12]. When phosphogypsum addition was more than 1.50%, free alkali gradually decreased with the increase of time mainly due to the slight dissolution of phosphogypsum, which continuously precipitated with alkaline substances dissolved in solid phase of bauxite residue and inhibited the dissolution of solid phase alkali. Finally, free alkali reduced to 5.50 mmol/kg and the removal rate of free alkali was 97.94% when phosphogypsum addition of 1.50% for 91 d, which effectively regulated free alkali in bauxite residue.

Exchangeable sodium could also potentially be used to indicate the degree of bauxite residue alkalization, which is an important indicator to measure alkalinity of bauxite residue with the characteristics of the higher exchangeable sodium. The worse the dispersion of bauxite residue colloidal particles, the worse the particle structure [25]. The change of exchangeable sodium in bauxite residue with phosphogypsum addition is shown in Table 4. The initial exchangeable sodium of raw bauxite residue was 391.33 mg/kg. With the increase of time, the continuous addition of water improved physical properties of bauxite residue and the pH decreased, which promoted the dissolution of a small amount of minerals such as calcite and sodalite. The increased Ca²⁺ exchanged with Na⁺ adsorbed on bauxite residue, causing exchangeable sodium to decrease but not obvious. Phosphogypsum introduced a large amount of Ca²⁺, which replaced Na⁺ adsorbed on the surface of bauxite residue, promoted the exchange of a large amount of exchangeable sodium into the liquid phase and reduced exchangeable sodium in bauxite residue [26]. On the first sampling, exchangeable sodium decreased significantly as phosphogypsum addition increasing. When phosphogypsum addition was 1.50%, exchangeable sodium decreased from 391.33 to 175.91 mg/kg. With the increase of time, Ca²⁺ in phosphogypsum continuously released to react with the adsorbed sodium and exchangeable sodium was continuously replaced. After 28 d, the effect of exchangeable sodium was no longer obvious and remained at a low level. Finally, exchangeable sodium maintained at about 94.42 mg/kg with the removal rate of 75.87% when the phosphogypsum addition is of 1.50% for 91 d, which effectively reduced the alkalization degree of bauxite residue.

Figure 5 Relationship between free alkali with alkaline anion of bauxite residues

Table 4 Free alkali and exchangeable sodium for supernatants of bauxite residue transformed by phosphogypsum (mmol·kg^–1)

3.4 Mineral and morphology characteristics

The mineral phase of bauxite residue revealed by XRD pattern under different phosphogypsum additions for 91 d is shown in Figure 6. The alkaline substance present in bauxite residue without phosphogypsum treatment included calcite, calcium nepheline, hydrated garnet, sodalite, and tricalcium aluminate. The contribution of these highly buffering minerals to alkalinity in bauxite residue was mainly due to their ability to continuously dissolve and release alkaline anions (Eqs. (11)–(14)).

Na₈Al₆Si₆O₂₄(CO₃)(H₂O)₂+22H₂O=8Na⁺+6Al(OH)₃+6H₄SiO₄+6OH^–+CO₃^2–           (11)

Ca₃Al₂(SiO₄)_x(OH)_12–4x+4xH₂O=3Ca²⁺+2Al(OH)₃+xH₄SiO₄+6OH^–                 (12)

Na₈Al₆Si₆O₂₄Cl₂+24H₂O=8Na⁺+6Al(OH)₃+6H₄SiO₄+6OH^–+2Cl^–                    (13)

Ca₃Al₂(OH)₁₂=3Ca²⁺+2Al(OH)₃+6OH^–           (14)

Phosphogypsum had a significant effect on combined alkali in bauxite residue. When the diffraction angle was about 13° and phosphogypsum addition was 1.50%, a gypsum peak appeared in the XRD pattern, indicating that a small amount of gypsum was present in bauxite residue with a relatively stable form after bauxite residue stabilization reaction, which could continuously provide Ca²⁺ for bauxite residue liquid phase system and played an important role in inhibiting the dissolution of combined alkali in bauxite residue [27].

Figure 6 XRD patterns from bauxite residue transformed by phosphogypsum

The results of quantitative analysis of XRD phase (Table 5) showed that the total content of alkaline mineral phase in bauxite residue was 52.81%. The type and content of these alkaline minerals mainly depended on alumina extraction process (Bayer process, sintering method and combined method), source of bauxite (monohydrate, boehmite, gibbsite), dissolution conditions, etc [28]. The minerals of calcite, cancrinite, hydrated garnet, sodalite and tricalcium aluminate in bauxite residue played a decisive role in the strong alkalinity of bauxite residue, which could continuously dissolve and release alkaline anion to determine the strength of alkaline in bauxite residue. The content of alkaline phase after treatment with phosphogypsum addition of 1.50% and 4.00% decreased to 48.58% and 46.22%, respectively.

Hydrated garnet was the main alkaline mineral in bauxite residue. A small amount of Ca²⁺ reacted with H₄SiO₄ and Al(OH)₃ to form hydrated garnet after phosphogypsum addition. Hydrated garnet could dissolve and convert into insoluble gibbsite during natural storage, which was stably present in bauxite residue. Similarly, tricalcium aluminate slightly increased in prevalence mainly due to the small amount of tricalcium aluminate formed during the stabilization of free alkali and the increase in tricalcium aluminate was accompanied by the dissolution of the tricalcium aluminate. Sodalite and cancrinite decreased with phosphogypsum addition, eventually producing more insoluble minerals due to the calcium and sodium substitution (Eqs. (6), (15)–(19)) [29]. In addition, free alkali was consumed, sodalite and cancrinite were dissolved, which reduced by 3.08% and 3.76% with the addition of 1.50%. Ca²⁺ and CO₃^2– generated CaCO₃ and rapidly formed calcite (Eqs. (6), (15)–(19)), which increased by 1.34%. When the phosphogypsum addition increased to 4%, sodalite and cancrinite decreased by 5.13% and 6.63% respectively, calcite increased by 2.86%. Free alkali was fixed and the dissolution of combined alkali was inhibited, thereby reducing the alkalinity of bauxite residue.

Table 5 Alkaline mineral composition of bauxite residue transformed by phosphogypsum (mass fraction, %)

3Ca²⁺+4OH^–+2Al(OH)₄^–=Ca₃Al₂(OH)₁₂          (15)

6Na⁺+2Ca²⁺+6H₄SiO₄+6Al(OH)₃+2HCO₃^–+8OH^–=[Na₆Al₆Si₆O₂₄]·2[CaCO₃]+26H₂O       (16)

3Ca²⁺+(6–4x)OH^–+xH₄SiO₄+2Al(OH)₃=Ca₃Al₂(SiO₄)_x(OH)_12–4x                         (17)

Ca²⁺+Na₈Al₆Si₆O₂₄(CO₃)(H₂O)₂=Na₆Al₆Si₆O₂₄·CaCO₃·(H₂O)₂+2Na⁺         (18)

9Ca²⁺+[Na₈Al₆Si₆O₂₄][Na₂CO₃]=Ca₈Al₆Si₆O₂₄+CaCO₃+10Na⁺                                   (19)

SEM images of bauxite residue surface treated by different phosphogypsum additions are shown in Figure 7. The microstructure of bauxite residue without phosphogypsum addition was relatively loose and contained fine particulate matter and debris. With the increase of phosphogypsum addition, alkaline particles were effectively stabilized to form a poorly soluble substance attached to the surface of bauxite residue. Phosphogypsum also contributed to the bonding of bauxite residue particles and promoted the formation of large agglomerates. Therefore, the fine particles and debris in bauxite residue gradually reduced, bulk materials gradually increased. When the addition reached 4.00%, a strip-like bulk structure appeared, indicating that phosphogypsum promoted the formation of mineral structures.

Corresponding EDS examination of bauxite residue treated by different phosphogypsum additions to reveal the surface composition and relative is shown in Figure 7. As the increase of phosphogypsum addition revealed major changes in the amount of Ca detected (1.55% for bauxite residue and 10.04%, 21.58% for the phosphogypsum addition of 1.50% and 4.00%, respectively), while the content of Na decreased from 3.32% to 2.43%. Similarly, the content of Al on bauxite residue surface reduced from 23.13% to 8.85%, which again reflected the fact that phosphogypsum could effectively regulate alkaline of bauxite residue and stabilize Al to prevent aluminum from poisoning. The content of P and S slightly increased but not significant.

4 Conclusions

Based on soil culture experiment, the phosphogypsum optimum addition was 1.50 wt%. After 91 d treatment, the pH of bauxite residue decreased from 10.83 to 8.70, the removal rates of free alkali and exchangeable sodium were 97.94% and 75.87%, respectively. Moreover, the pH and free alkali, exchangeable sodium had significant positive correlations (P<0.05), providing new indicators for alkaline regulation. The effect in the factors affecting free alkali was CO₃^2–>OH^–> AlO₂^–>HCO₃^–. Furthermore, a small amount of gypsum was present in bauxite residue with a relatively stable form, which continuously provided Ca²⁺ to liquid phase system and inhibited the dissolution of combined alkali. The content of alkaline mineral phase in raw bauxite residue was 52.81% and reduced to 48.58% with 1.50% phosphogypsum treatment. In addition, phosphogypsum significantly promoted the formation of large agglomerates and increased Ca content on the surface of bauxite residue, while the Na and Al contents were reduced.

Figure 7 Morphological structure and energy-dispersive X-ray analysis spectrum obtained from bauxite residue treated by phosphogypsum:

References

[1] ZHU Feng, CHENG Qing-yu, XUE Sheng-guo, LI Chu-xuan, HARTLEY W, WU Chuan, TIAN Tao. Influence of natural regeneration on fractal features of residue microaggregates in bauxite residue disposal areas [J]. Land Degradation and Development, 2018, 29(1): 138–149. DOI: 10.1002/ldr.2848.

[2] XUE Sheng-guo, ZHU Feng, KONG Xiang-feng, WU Chuan, HUANG Ling, HUANG Nan, HARTLEY W. A review of the characterization and revegetation of bauxite residues (red mud) [J]. Environmental Science and Pollution Research, 2016, 23(2): 1120–1132. DOI: 10.1007/s11356- 015-4558-8.

[3] ZHU Feng, XUE Sheng-guo, HARTLEY W, HUANG Ling, WU Chuan, LI Xiao-fei. Novel predictors of soil genesis following natural weathering processes of bauxite residues [J]. Environmental Science and Pollution Research, 2016, 23(3): 2856–2863. DOI: 10.1007/s11356-015-5537-9.

[4] LIAO Jia-xin, JIANG Jun, XUE Sheng-guo, CHENG Qing-yu, WU Hao, MANIKANDANA R, HARTLEY W, HUANG Long-bin. A novel acid-producing fungus isolated from bauxite residue: The potential to reduce the alkalinity [J]. Geomicrobiology Journal, 2018, 35(10): 840–847. DOI: 10.1080/01490451.2018.1479807.

[5] ZHU Feng, LIAO Jia-xin, XUE Sheng-guo, HARTLEY W, ZOU Qi, WU Hao. Evaluation of aggregate microstructures following natural regeneration in bauxite residue as characterized by synchrotron-based X-ray micro-computed tomography [J]. Science of Total Environmental, 2016, 573: 155–163. DOI: 10.1016/j.scitotenv.2016.08.108.

[6] ZHU Feng, LI Xiao-fei, XUE Sheng-guo, HARTLEY W, WU Chuan, HAN Fu-song. Natural plant colonization improves the physical condition of bauxite residue over time [J]. Environmental Science and Pollution Research, 2016, 23(22): 22897–22905. DOI: 10.1007/s11356-016-7508-1.

[7] XUE Sheng-guo, YE Yu-zhen, ZHU Feng, WANHG Qiong-li, JIANG Jun, HARTLEY W. Changes in distribution and microstructure of bauxite residue aggregates following amendments addition [J]. Journal of Environmental Sciences, 2019, 78: 276–286. DOI: 10.1016/j.jes.2018.10.0 10.

[8] LI Yi-wei, JIANG Jun, XUE Sheng-guo, MILLAR G J, KONG Xiang-feng, LI Xiao-fei, LI Meng, LI Chu-xuan. Effect of ammonium chloride on leaching behavior of alkaline anion and sodium ion in bauxite residue [J]. Transactions of Nonferrous Metals Society of China, 2018, 28(10): 2125–2134. DOI: 10.1016/S1003-6326(18)64857-5.

[9] PAN Xiao-lin, YU Hai-yan, TU Gan-feng. Reduction of alkalinity in bauxite residue during bayer digestion in high-ferrite diasporic bauxite [J]. Hydrometallurgy, 2015, 151(1): 98–106. DOI: 10.1016/j.hydromet.2014.11.015.

[10] PARADIS M, DUCHESNE J, LAMONTAGNE A, ISABEL D. Long-term neutralisation potential of red mud bauxite with brine amendment for the neutralisation of acidic mine tailings [J]. Applied Geochemistry, 2007, 22(11): 2326– 2333. DOI: 10.1016/j.apgeochem.2007.04.021.

[11] WHITTINGTON B I, FLETCHER B L, TALBOT C. The effect of reaction conditions on the composition of desilication product (DSP) formed under simulated Bayer conditions [J]. Hydrometallurgy, 1998, 49(1, 2): 1–22. DOI: 10.1016/S0304-386X(98)00021-8.

[12] ZHU Xiao-bo, LI Wang, GUAN Xue-mao. An active dealkalization of red mud with roasting and water leaching [J]. Journal of Hazardous Materials, 2015, 286: 85–91. DOI: 10.1016/j.jhazmat.2014.12.048.

[13] GRAFE M, POWER G, KLAUBER C. Bauxite residue issues: III. Alkalinity and associated chemistry [J]. Hydrometallurgy, 2011, 108(1): 60–79. DOI: 10.1016/j.hyd romet.2011.02.004.

[14] BARROW N J. Possibility of using caustic residue from bauxite for improving the chemical and physical properties of sandy soils [J]. Crop & Pasture Science, 1982, 33(2): 275–285. DOI: 10.1071/AR9820275.

[15] MAYES W M, YOUNGER P L, AUMONIER J. Buffering of alkaline steel slag leachate across a natural wetland [J]. Environmental Science & Technology, 2006, 40(4): 1237– 1243. DOI: 10.1021/es051304u.

[16] SHI Ben, QU Yang, LI Hui. Gypsum alleviated hydroxyl radical-mediated oxidative damages caused by alkaline bauxite residue in leaves of atriplex canescens [J]. Ecological Engineering, 2017, 98: 166–171. DOI: 10.1016/j. ecoleng.2016.10.008.

[17] XUE Sheng-guo, WU Yu-jun, LI Yi-wei, KONG Xiang-feng, ZHU Feng, WILLIAM H, LI Xiao-fei, YE Yu-zhen. Industrial wastes applications for alkalinity regulation in bauxite residue: A comprehensive review [J]. Journal of Central South University, 2019, 26(2): 268–288.

[18] COURTNEY R, KIRWAN L. Gypsum amendment of alkaline bauxite residue-plant available aluminium and implications for grassland restoration [J]. Ecological Engineering, 2012, 42(5): 279–282. DOI: 10.1016/j.ecolen g.2012.02.025.

[19] TAYIBI H, CHOURA M, LOPEZ F A, ALGUACIL F J, LOPEZDELGADO A. Environmental impact and management of phosphogypsum [J]. Journal of Environmental Management, 2009, 90(8): 2377–2386. DOI: 10.1016/j.jenvman.2009.03.007.

[20] TAHER M A. Influence of thermally treated phosphogypsum on the properties of portland slag cement [J]. Resources Conservation & Recycling, 2008, 52(1): 28–38. DOI: 10.1016/j.resconrec.2007.01.008.

[21] PAPAGEORGIOU F, GODELITSAS A, MERTZIMEKIS T J, XANTHOS S, VOULGARIS N, KATSANTONIS G. Environmental impact of phosphogypsum stockpile in remediated schistos waste site (Piraeus, Greece) using a combination of γ-ray spectrometry with geographic information systems [J]. Environmental Monitoring & Assessment, 2016, 188(3): 1–14. DOI: 10.1007/s10661-01 6-5136-3.

[22] ZHANG Ru, LI Yan-jun, LIU Jie, ZHAO Yue-yang, JIANG Yu-da. Utilization of phosphogypsum and treatment of the impure elements [J]. Conservation and Utilization of Mineral Resources, 2015(2): 50–54. (in Chinese)

[23] CHANDARA C, AZIZLI K A, AHMAD Z A, SAKAI E. Use of waste gypsum to replace natural gypsum as set retarders in portland cement [J]. Waste Management, 2009, 29(5): 1675–1679. DOI: 10.1016/j.wasman.2008.11.014.

[24] SANTINI T C, MARTIN V F, ROBERT J G. Experimental simulation of long term weathering in alkaline bauxite residue tailings [J]. Metals, 2015, 5(3): 1241–1261. DOI: info:doi/10.3390/met5031241.

[25] LIU Jin-wei, HU Hui-ping, WANG Meng, CHEN Xiang-pan, CHEN Qi-yuan, DING Zhi-ying. Synthesis of modified polyacrylamide with high content of hydroxamate groups and settling performance of red mud [J]. Journal of Central South University, 2015, 22(6): 2073–2080. DOI: 10.1007/s11771-015-2731-3.

[26] YU Yang-hua, WU Yong-gui, YU Li-fei, SHEN Wan-lun. Effect and mechanism of phosphogypsum and CaCO₃ on dealkalization of red mud [J]. Inorganic Chemicals Industry, 2014, 46(10): 58–61. (in Chinese)

[27] SUN Dao-xing. Dealkalization and reclaiming valuable Ti and Sc from red mud [J]. Inorganic Chemicals Industry, 2008, 40(10): 49–52. (in Chinese)

[28] LIAO C Z, ZENG L, SHIH K. Quantitative X-ray diffraction (QXRD) analysis for revealing thermal transformations of red mud [J]. Chemosphere, 2015, 131: 171–177. DOI: 10.1016/j.chemosphere.2015.03.034.

[29] GATTA G D, LOTTI P, KAHLENBERG V, HAEFEKER U. The low-temperature behaviour of cancrinite: An in situ single-crystal X-ray diffraction study [J]. Mineralogical Magazine, 2012, 76(4): 933–948. DOI: 10.1180/minmag.2 012.076.4.10.

(Edited by YANG Hua)

中文导读

磷石膏改良作用下赤泥的碱性变化

摘要：赤泥碱性转化是氧化铝工业处置固体废物面临的世界性难题。本文开展工业废弃物对赤泥碱性改良研究，结果表明：磷石膏最优添加量为1.50 wt%，处理91天后，赤泥pH由10.83降至8.70；自由碱去除率为97.94%，可交换性钠去除率为75.87%，自由碱、可交换性钠和赤泥pH存在显著正相关关系 (P<0.05)；自由碱的作用效果为CO₃^2–＞OH^–＞AlO₂^–＞HCO₃^–；化学结合碱含量从52.81%降至48.58%，固相中稳定性强的石膏为液相体系持续提供Ca²⁺，抑制结合碱的溶解；磷石膏促进团聚体形成，表面Ca²⁺含量增加， Na⁺、Al³⁺含量降低，有利于加快赤泥土壤化进程。

关键词：赤泥；碱性调控；自由碱；磷石膏改良；赤泥土壤化

Foundation item: Projects(41877511, 41842020) supported by the National Natural Science Foundation of China

Received date: 2018-10-15; Accepted date: 2018-11-27

Corresponding author: XUE Sheng-guo, PhD, Professor; Tel: +86-13787148441; E-mail: sgxue70@hotmail.com, sgxue@csu.edu.cn; ORCID: 0000-0002-4163-9383