Analysis of environmental behavior and potential toxic metal(loid)s barrier countermeasures for lithium leach residue stockpiling

2.2.1. Determination of basic properties

The samples were shaken for 5 min at a liquid-solid ratio of 2.5:1 and then left for 20 min, and the pH value was determined by a pH meter (Starter 3100C, China). Cation exchange capacity (CEC) was measured according to the Chinese standard HJ 889–2017, with a 10% parallel sample rate per batch and a standard curve R² = 0.99912. Moisture content (MC) was determined based on the Chinese standard HJ 613–2011. The resulting LLR exhibited a pH of 8.2±0.013, an MC of 2.51%±0.01, and a CEC of 0.08±0.002 cmol/kg. Furthermore, the elemental distribution of PTMs within the samples was characterized by combining a cold field emission scanning electron microscope (SEM, Hitachi 8010, Japan) with an energy dispersive spectrometer (EDS, Bruker Q200).

2.2.2. Leaching release experiment of PTMs in LLR

The leaching of PTMs was analyzed according to the Chinese standard HJ/T299-2007 by utilizing the pH gradient leaching test. Considering that Tl⁺ is soluble and more soluble in the pH range of 0–13, TlOH precipitation will be formed at pH > 13 [16]. Therefore, the acidic leachate with pH 1, 3.21, and 5 was prepared by the H₂SO₄-HNO₃ method, the neutral and natural weakly acidic acid rain conditions with pH 7 and 6 were simulated by applying ultrapure and carbonated water, and the alkaline leachate with pH 12 and 13 was prepared by the NaOH method. After leaching, the content of PTMs was determined through inductively coupled plasma-mass spectrometry (ICP-MS, PinAAcle 900H). Details of the leachate configuration and leaching method can be found in the Supplementary Materials A.

2.2.3. Determination of metal content and sequential extraction procedure

The total amount of each metal element was detected in accordance with the Chinese standard HJ1315-2023. The concentration of PTMs was analyzed with an ICP-MS (PinAAcle 900H). The presence of Tl was assessed via an improved Bureau Community of Rearence (BCR) three-step extraction method [17]. This method separates heavy metals into different components, covering exchangeable components (Tl_Exc.), reducible components (Tl_Red.), oxidized components (Tl_Oxi.), and residual components (Tl_Res.). The concentrations in the extraction solutions were quantified with the ICP-MS. Detailed procedures have been presented in the Supplementary Materials B.

3.2.1. Evaluation of soil contamination

The pH value, CEC, MC, and analysis results for each sample have been detailed in the SI. The statistical data for PTMs at sampling points Z1 and Z2 have been shown in Table 1. The maximum values were compared with the local soil background values, and the excess multiples of PTMs were obtained in the following order: Tl (3.66)>Hg (2.18)>Be (1.36)>Cd (1.32)>Co (1.05)>Pb (0.63)>Zn (0.21)>As (0.17) >Cu (0.11) >Ni (0.1) >Cr (0.05). All PTM concentrations from Zone Z3 (Table 2) were lower than those in Zones Z1 and Z2, indicating that the migration risk in rainwater was lower than that in soil.

Additionally, the mean values (n=3) of each point in the three plots were compared with the corresponding risk screening values to obtain the exceedance rate P (Figure 4). The results showed that only the Tl concentrations at sampling points Z11 and Z15 exceeded the standard, indicating that the concentration of Tl in the soil decreases with increasing distance, with a maximum exceedance rate of 2.84. Furthermore, all other PTMs were lower than the standard level. According to the I_geo assessment classification standards, the soil contamination level of Tl was classified as Level 2 and that of Hg as Level 1, indicating that the PTMs contamination risk around the LLR storage site was relatively low.

Close

Figure 4.

Plot of exceedances of metal elements in soil at sampling points (taking the maximum concentration), where P < 1 represents non-exceedance points and P ≥ 1 represents exceedance points.

Export to PPT

3.2.2. Source analysis of PTMs in soil

The release pathways of PTMs from smelting slag piles [20,24] are typically categorized into the following aspects. Firstly, unstable minerals attached to the surface of the slag are immediately released under the shear action of rainwater when coated with salts and enter the soil with the rainwater. Secondly, as soluble minerals on the surface are eroded, the slag contacts rainwater, PTMs are released through surface oxidation, and migrate into the soil. To further analyze the migration mechanisms of PTMs in lithium leaching slag, multivariate statistical analysis was employed to identify the sources of heavy metal contamination in the soil. Pearson correlation analysis was performed with SPSS software (Table 3), and the original data were validated through Kaiser-Meyer Olkin (KMO) and Bartlett's Sphericity (BS) tests. The results showed that the KMO value was 0.533, and the significance level of the BS test was <0.01, demonstrating that the data were suitable for principal component analysis. Based on the criterion of eigenvalues > 1, three principal components were extracted, explaining 87.82% of the total variance (Table 4).

The first principal component (PC1) explained 54.89% of the total variance, primarily reflecting information on components such as Be (0.702), Co (0.916), Cr (0.955), Ni (0.925), and Cu (0.801), with all component loadings exceeding 0.7. Pearson correlation analysis revealed highly significant positive correlations (P < 0.01) between Be and Co, Cr, Ni, and Cu, suggesting that they have the same origin. Moreover, SEM-EDS also confirmed that Co, Ni, and other elements were not distributed on the surface of LLR, while the distribution of Cr and Ni in the soil was mainly influenced by the parent material and soil formation process [25]. Co is generally considered to originate from rock weathering [26]. Therefore, it can be inferred that PC1 (including Be, Co, Cr, Ni, and Cu) is the second stage of slag release, representing natural source.

The second principal component (PC2) accounted for 18.58% of the total variance, mainly reflecting the compositional information of Cd (0.808), Pb (0.717), and Zn (l0.662). The CV% of such type metal(loid)s (Table 1) is greater than that of PC1, confirming that it is more affected by anthropogenic activities than PC1. Therefore, PC2 (including Cd, Pb, and Zn) migration has already initiated, and both natural and pollutant sources exist simultaneously.

The third principal component (PC3) explained 14.35% of the total variance, mainly representing the correlated components of Tl (0.132) and Hg (0.367). There was a highly significant positive correlation (P < 0.01) between Tl and Hg, suggesting that the two can be considered synchronously in the management. The CV% of Tl was observed to be 92.96% (Table 1), highlighting that Tl is the most heavily polluted element in the LLR inventory. Consequently, PC3 (including Tl and Hg) can be identified as the first stage of LLR release, and has become a source of pollution in the soil.

3.3.1. Influence of soil microbial community characterization

Microorganisms are important components of soil ecosystems, which respond rapidly and sensitively to environmental changes and disturbances, particularly in cases of high concentrations of heavy metal pollutants [27]. Moreover, the sensitive populations in the soil were gradually decreased during heavy metal exposure, and heavy metal-tolerant strains ultimately dominated the core position [28], thereby inhibiting the colonization of other sensitive bacteria. Regarding classification at the bacterial phylum level (Figure 5), the community structure of Z11 was homogeneous. This is due to the poor structure of the surrounding soil, low pH, lack of organic matter and nutrients, high concentration of metallic elements, and loss of biodiversity. Secondly, soil bacteria primarily belong to Proteobacteria, Gemmatimonadota, Actinobacteriota, Acidobacteriota, Bacteroidota, Chloroflexota. Among them, Proteobacteria, Acidobacteriota, Actinobacteriota, and Gemmatimonadota accounted for an average of 76.12% of the total microorganisms, constituting the dominant bacterial communities in the soil surrounding the LLR yard. These findings are consistent with those observed in granite tailings [29] (duration > 10 years) and gold tailings [30] (duration > 5 years), demonstrating that long-term interference from metallic elements has an impact on the uniformity of bacterial community composition in the soil environment of various types of tailings.

Close

Figure 5.

Relative abundance of the major bacterial phyla at different sampling sites. The bacterial phyla with low frequencies (< 0.1%) in all the samples were pooled and are denoted “others”.

Export to PPT

In addition, several studies found that the flora of Proteobacteria, Acidobacteriota, Firmicutes, and Chloroflexota could tolerate the concentration of high concentrations of metal, thereby becoming dominant flora during long-term accumulation [31,32]. Li et al.[33] discovered that the relative abundance of Proteobacteria in the soil around the rare-earth mining area increased under the action of pioneer plants, and it exhibited the highest sensitivity to soil metal(loid) s concentrations. These results revealed that the transformation of metalloids by Proteobacteria is a gradual process influenced by the presence of metals in the tailings area. Although the storage behavior of LLR is relatively short-term (<3 years), the surrounding soil microorganisms have gradually developed the Proteobacteria community structure with tolerance to metal(loid)s concentrations. The migration disturbance of metals in LLR has affected the microbial community structure of the surrounding soil environment.

3.3.2. Influence of soil microbial diversity and community structure

In response to the transport perturbation of PTMs in LLR, the microbial community of the soil environment around the yard has gradually formed a characteristic structure with Proteobacteria as the dominant group. From an ecotoxicological perspective, metal toxicity inhibits microbial activity, leading to changes in microbial diversity and community structure [34]. Figure 6(a) shows that Z3 has the highest Chao1 index, indicating higher community diversity than Z1 and Z2. This is related to two situations: (1) The previously analyzed soil migration risk exceeds that of rainwater leaching, resulting in reduced LLR perturbation in Z3; (2) Z3 is an agricultural plot where long-term cultivation has resulted in high microbial abundance; whereas, the relative abundance of Proteobacteria in the Z1 contaminated area was low, except at sampling sites Z11 (31.41–79.1%) and Z2 (28.6%–70.6%), This indicated that the transport perturbation of metals had a direct inhibitory effect on soil microbial abundance.

Close

Figure 6.

Microbial community α-diversity of different soil samples from three plots: (a) Chao index, OTU level of (b) Shannon index; RDA of correlation of environmental factors (c) correlation of regional plots with environmental factors. (d) VPA analysis of soil microbial communities by E1 (Tl, Hg), E2 (other metal(loid)s), and E3 (soil physicochemical properties).

Export to PPT

The Shannon indices (Figure 6b) of the three plots were comparable, with Z1 slightly higher than the Z2 and Z3 plots, indicating that microbial diversity gradually increases in soils affected by metal stress in the LLR. Its mean Shannon index was 8.87, comparable to the mean of 7.62 in sedimentary samples from abandoned Pb-Zn mining areas [35]. We found that the relationship between PTMs-induced disturbance effects on soil in LLR and microorganisms is complex, with context-dependent outcomes. Overall enrichment of PTMs inhibits microbial activity, leading to lower microbial abundance. However, since different microorganisms have different tolerances to the toxicity of PTMs, the diversity of the contaminated plots gradually increased, while overall microbial abundance decreased. For example, Proteobacteria demonstrate excellent tolerance to Cd and Pb [36], Chlamydiae and Proteobacteria have high relative abundance at high heavy metal contamination levels, and Proteobacteria show high response to Pb levels in different ecosystems [15].

The migratory enrichment of metals in LLR stockpiling behavior not only limits the soil microbial community but also interacts with the distinctive soil physicochemical properties of the site, which further affect metal toxicity; these three elements are correlated [37]. Therefore, the effects of environmental factors (three soil physicochemical properties: pH, CEC, and MC; and PTMs: Cr, Hg, As, Pb, Cu, Ni, Zn, Be, Tl, Cd, and Co) on the microbial community were analyzed using Redundancy Analysis (RDA) based on the abundance of sequences in the Operational Taxonomic Units (OTUs) (Figure 6c). When considering single factors, Tl was the most dominant factor influencing the structure of the biological community under the LLR stockpile. It was positively correlated with the Z1 and Z2 sampling sites, in which Tl contributed the most to the contamination, whereas Z3 was more affected by Zn, Hg, and Ni. These findings are consistent with the previously analyzed results identifying Tl as the main environmental risk factor of LLR. Meanwhile, pH, MC, and CEC were negatively correlated with Tl, indicating that soil physicochemical properties were not the main influencing factors for the migration and enrichment of Tl. In contrast, CEC, pH, and MC showed positive correlations with Hg, Zn, and other metals. It can be attributed to the fact that at higher pH values, many metal(loid)s form solid phases through complexation with organic matter or precipitation as carbonates (e.g., Zn/Cd), while these compounds are dissolved at lower pH values, thereby enhancing metal mobility [38].

The microbial community structure (relative abundance and diversity) served as the response variable, and the VPA quantified the relative contributions of environmental risk factors and soil physicochemical properties (Figure 6d), and all the environmental factors collectively explained 43.53% of the variation in the soil microbial community, with the independent contributions being approximately 7.29% from primary and secondary environmental risk factors, 5.97% from other environmental risk factors, and 6.89% from soil physicochemical properties. In addition, environmental factors exhibited interactions, with PTMs accounting for 36.63% of the total variation and soil physicochemical properties explaining 24.54%. The effect of E1 alone slightly exceeded that of E2 and E3. The results showed that the microbial community structure of the surrounding soil was influenced by multiple factors from LLR stockpiling activities, among which E1 (Tl, Hg) dominated over soil physicochemical properties and should be identified as the primary factor affecting microbial communities.

3.4.1. Identification and analysis of priority pollutants

PTMs in LLR storage undergo leaching and soil migration, adversely impacting ecosystems and microbial communities. Leaching processes primarily pose environmental risks from Tl, Be, Ni, Co, and Hg, while soil migration is dominated by Tl, Hg, Be, Cd, and Co. The combination classified their environmental risks into four levels (Figure 7a), identifying Tl as the priority pollutant. Therefore, Tl migration in LLR stockpiles represents a major environmental risk requiring primary management attention.

Close

Figure 7.

(a) Classification of potential ecological and environmental risks in the LLR and to the surrounding soil potential ecological and environmental risks, summarizing the major environmental risk factors and minor risk factors. Among them, level 1 risk is the largest, and level 4 is the smallest. (b) Percentage of each extracted state in the sequence extraction results of Tl in LLR and each soil sampling site, (c) Concentration of each extracted state, (d) Pearson correlation analysis of each extracted state with pH.

Export to PPT

In LLR, Tl can enter sulfides (e.g., ZnS, PbS, FeS) through isomorphous substitution, replace K, Rb, and other elements within the crystal lattice of silicate minerals, and adsorb onto mineral colloid surfaces. This results in diverse chemical species across different storage patterns, where the toxicity and stability of metals depend not only on their concentration but also on their speciation [24]. Based on this, Tl concentrations in LLR and surrounding sampling points were further analyzed via BCR extraction (Figures 7b,c). The results indicate that the highest concentrations of Tl_Exc. were found in LLR and the surrounding soil. The concentrations of each form, ranked by abundance, are as follows: Tl_Exc. (0.089–6.12 mg/kg) > Tl_Red. (0.11–2.089mg/kg) > Tl_Oxi. (0.067–0.396 mg/kg) > Tl_Res. (0.006–0.226 mg/kg). This suggests that Tl in LLR primarily binds to tailings particles through carbonate co-precipitation and electrostatic adsorption, a mechanism consistent with prior studies [39,40]. At the same time, Tl_Exc. is the main storage form in the soil surrounding LLR stockpiles, which suggests that the mobility of Tl in LLR is very high and has a great impact on the surrounding environment. While Tl_Red. and Tl_Oxi. dominated topsoil speciation, Tl_Oxi. exhibited significantly lower mobility than Tl_Exc. along vertical soil profiles [41]. This implies that converting Tl_Exc. to Tl_Red. or Tl_Oxi. in LLR-affected zones would effectively reduce Tl migration through soil profiles. Meanwhile, soil Tl_Exc. and Tl_Red. were found to have a high correlation with pH (Figure 7d). In alkaline soils, released Tl_Exc. is adsorbed by Fe-Mn oxides and oxidized to T_Red. [42]. This redox transformation forms the basis for PTMs immobilization strategies.

3.4.2. Blocking or repairing pathways

Based on the blocking path approach, a remediation mechanism was developed (Figure 8): (i) Physical method: Incorporate lime and alkaline stabilizers to suppress Tl_Exc. mobility and promote its conversion to more stable Tl_Red. Subsequently, apply inorganic-organic composites to reduce Tl reactivity, thereby controlling further contamination. (ii) Chemical method: Leverage Tl's dual valence states (Tl⁺/Tl³⁺). Oxidize Tl⁺ to Tl³⁺ to form insoluble Tl₂O₃ with lower mobility and toxicity [43], Fe-Mn oxides can convert Tl_Exc. to Tl_Red. [42], reducing its mobility and toxicity. Alternatively, manganese oxide-based adsorbents, such as activated carbon-supported MnO₂(Mn-AC) or metal-organic framework- loaded MnO₂ (Mn-MOF) [16,43], enable the formation of insoluble Tl₂O₃ or Tl(OH)₃ co-precipitation of Tl with PTMs (Al, Mg, Zr) in LLR on the surface of Mn oxides [27]. Or enhancing soil organic matter (SOM) or S content. Metal sulfides were employed to further convert Tl_Red. to Tl_Oxi. [23] and improve long-term stability. (iii) Plant and biological methods: The ionic radii of Tl⁺ and K⁺ are similar [11], so plants highly enriched in K⁺ can be preferred as pioneer plants. For example, cinnamon violet [44], kale, foxtail algae, and ferns exhibit significant Tl accumulation. Specifically, kale shows a Tl enrichment coefficient up to 10.7, with higher efficiency for Tl⁺ than Tl³⁺, the latter primarily retained in roots. In microbial remediation, while microorganisms cannot directly degrade heavy metals, they alter metal speciation and bioavailability by modifying physicochemical properties (e.g., redox potential, pH), thereby reducing environmental impacts [39]. Acidovorax can oxidize Fe²⁺, sulfides, and sulfide minerals, whereas Pseudomonas fluorescens has a strong affinity for Fe²⁺. Two species have been shown to convert Tl_Exc. [45].

Close

Figure 8.

Diagram of Tl contamination barrier pathways and mechanisms in LLR stockpiles.

Export to PPT