As the demand for renewable energy technologies accelerates, lithium has become a cornerstone of battery-driven decarbonization strategies. Lithium-ion batteries power electric vehicles, enable grid-level energy storage, and underpin the global shift away from fossil fuels. Global lithium mine production reached 204 kilotons in 2023. By 2050, demand is projected to rise significantly, reaching approximately 1600 to 1700 kilotons under the Announced Pledges Scenario (APS) and the Net Zero Emissions by 2050 (NZE) Scenario [1–4]. While much attention has focused on lithium availability, carbon emissions, and energy consumption during extraction, an equally important yet underexplored issue is the environmental risk posed by toxic trace elements co-extracted during lithium processing.

Lithium is primarily sourced from hard-rock ores and brine deposits [5]. Spodumene and lepidolite ores are mined in countries such as Australia, China, Zimbabwe, and Brazil, while brine resources are concentrated in Chile, Argentina, and Bolivia. In China, over 60% of lithium carbonate production relies on ore-based extraction, making it the dominant route. This pathway—typically involving flotation, roasting, and leaching—is energy-intensive and generates significant volumes of solid waste.

According to the International Energy Agency (IEA), by 2050, ore-based lithium extraction could emit around 60 million tons of CO₂ annually, comparable to the electricity consumption of approximately 39 million households [6]. Furthermore, producing 1 ton of lithium carbonate generates 8–10 tons of waste from spodumene and 30–50 tons from lepidolite. China alone produces over 12 million tons of lithium-processing waste each year. Of growing concern is the release of highly toxic elements, such as beryllium (Be) and thallium (Tl), from these wastes. Both elements have been detected not only in solid residues but also in process solutions, the final lithium carbonate product, and even rechargeable lithium batteries [7].

Be is a potent carcinogen that can damage mucous membranes and the cardiovascular system. Tl is an extremely toxic heavy metal that targets the nervous system, liver, and kidneys [8]. Both elements naturally occur in lithium-bearing minerals such as lepidolite and spodumene. Be and Tl remain relatively stable in lepidolite prior to metallurgical lithium extraction. However, after undergoing high-temperature calcination and leaching, these elements become mobilized in the resulting lithium slags. For instance, leachate analyses of lithium slags from Yichun area of China, conducted using the WET and TCLP protocols, have revealed concentrations as high as 789 ppb (Be) and 33 ppb (Tl) under WET, and 404 ppb (Be) and 3 ppb (Tl) under TCLP—exceeding the U.S. EPA maximum contaminant levels for drinking water (Be: 4 ppb [9]; Tl: 2 ppb [10]) by severalfold to several hundred times. Recent pollution incidents in lithium-refining regions of China have drawn attention to the ease with which Be and Tl migrate into surrounding soil and water. Thallium’s lithophilic nature presents similar environmental risks for other lithium-producing countries such as Canada and Australia [11]. Despite increasing awareness, efforts to mitigate Be and Tl pollution remain constrained by a limited understanding of their geochemical occurrence and transformation mechanisms.

The key challenge lies in the ultra-low concentrations and heterogeneous distribution of Be and Tl within ore and waste slag matrices. For example, in the Yichun region of China, the typical Be concentrations ranged 70–330 ppm in lepidolite. After metallurgical processing, the concentrations in the resulting slags were typically 60–100 ppm. Similarly, Tl concentrations were 30–70 ppm in lepidolite and were reduced to around 20–30 ppm in the slags. Without nanoscale insight into how these elements are structurally hosted—whether in individual mineral particles, within silicate frameworks, or adsorbed at interfaces—designing targeted strategies for immobilization or removal remains elusive. This knowledge gap limits our ability to optimize extraction processes, design effective stabilization treatments, and develop regulatory frameworks.

Two fundamental questions remain unresolved.

(1) Do Be and Tl occur as individual mineral grains (e.g., beryl, bertrandite, crookesite, lorandite)? Spodumene and beryl are both aluminosilicate minerals that commonly coexist in the same pegmatite deposits. As both minerals are non-magnetic and have similar specific gravities—comparable as well to those of gangue minerals—they are difficult to separate. Thallium, on the other hand, rarely forms independent mineral phases; crookesite and lorandite are rare exceptions. More commonly, Tl occurs in trace amounts within sulfide or selenide matrices.

(2) If not present as individual minerals, which host phases accommodate them, and by what mechanisms—lattice substitution, surface adsorption, or nanoparticle segregation? At ppm concentrations, Be and Tl are typically dispersed within fine-grained, compositionally complex matrices, posing significant analytical challenges using conventional techniques. Powder X-ray diffraction (XRD) and inductively coupled plasma mass spectrometry (ICP-MS) provide bulk data but lack structural resolution. Combined scanning electron microscopy and energy-dispersive X-ray spectroscopy (SEM-EDS) and electron probe microanalysis (EPMA) offer morphological and compositional insights but fall short in detecting trace-level species or elucidating crystal structure.

To address these, we propose an integrated structural characterization framework based on advanced, spatially resolved techniques.

(1) Three-dimensional electron diffraction (3D-ED). When Be and Tl occur in nanocrystalline minerals, 3D-ED is uniquely suited to resolve their structures. Many such phases exhibit polytypism, disorder, or twinning, escaping detection by conventional powder XRD. 3D-ED provides single-crystal data of structure solution and refinement quality, which enables structure solution and refinement from crystallites smaller than 100 nm, providing definitive atomic-scale information [12].

(2) Micro-/nano-XRD and synchrotron diffraction. If Be and Tl are structurally incorporated within common ore minerals, high-resolution and micro-/nano-area diffraction can identify subtle lattice distortions or secondary phases. When interpreted in the context of ionic radii, oxidation states, and coordination environments of Be and Tl, these data help pinpoint candidate host structures. For phase identification, the detection limit of XRD for well-crystallized phases is typically around 1–3 wt%. Micro-/nano-XRD and synchrotron-based diffraction techniques can focus on specific regions with micrometer-scale resolution, thereby enhancing the effective detection limit when the target phase is locally enriched. This spatial selectivity allows for the identification of phases whose concentrations may be significantly higher within microdomains than in the bulk materials.

(3) Correlative elemental mapping. Combining principal-components mapping (SEM-EDS or EPMA) with trace-element techniques such as laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) or time-of-flight secondary ion mass spectrometry (TOF-SIMS) enables co-localization of Be and Tl with specific mineral domains. This spatial correlation is critical for inferring structural associations within heterogeneous matrices. The typical detection limit for EDS is approximately 0.1–1 wt% (1000–10000 ppm), whereas EPMA offers higher sensitivity, with detection limits typically ranging from 0.001 to 0.01 wt% (10–100 ppm). Despite its ability to detect elements at the ppm level, EPMA is commonly used for principal components analysis. For trace element determination, techniques such as LA-ICP-MS and TOF-SIMS are more suitable, offering detection limits down to the ppm to sub-ppm range [13].

Together, these tools offer a path toward resolving the structural environments of Be and Tl, laying the foundation for risk-informed mitigation strategies. With structural insights in hand, several targeted approaches can be developed.

(1) Process innovation for source control. Extraction pathways can be redesigned to prevent the transformation of Be and Tl into mobile forms. Mechanochemical activation, electrochemical methods, or lower-temperature calcination may help preserve stable host structures. Flux additives could be used to promote the formation of inert silicates or phosphates that sequester Be and Tl.

(2) Stabilization of legacy waste. For the vast existing stockpiles of Be/Tl-bearing lithium slag, leverage knowledge of their specific occurrence forms to develop tailored stabilization or solidification technologies. This could involve engineered additives that form insoluble compounds with Be/Tl (e.g., specific phosphates, sulfides, or engineered ceramics) or encapsulate reactive phases, creating robust chemical barriers to leaching.

(3) Beneficial utilization through phase engineering. Leverage the microstructural associations between trace beryllium/thallium and principal components (Si, Al, Ca, etc.) in lithium slag to guide the development of cost-effective phase-regulation strategies. Techniques such as controlled thermal treatment and targeted flux additions can be employed to induce the formation of stable crystalline phases—such as feldspars or pyroxenes—that effectively incorporate Be and Tl into inert structures or facilitate their selective separation. These approaches enable the detoxified bulk residue to be safely repurposed as supplementary cementitious materials or construction aggregates, simultaneously reducing hazardous waste and alleviating raw material scarcity. This strategy supports circular economy goals by transforming hazardous by-products into valuable resources.

As lithium production rapidly expands to meet global energy demands, the environmental risks associated with co-occurring toxic elements must receive equal attention. Among these, beryllium and thallium pose a particularly significant yet underrecognized threat in hard-rock lithium extraction (Figure 1). Their behavior at the nanoscale governs both their environmental mobility and long-term persistence, yet the structural mechanisms underlying their occurrence and transformation remain poorly understood.

Figure 1 A balance must be maintained between lithium extraction and its utilization.

Addressing this challenge requires more than empirical observation—it demands a fundamental shift toward a structural understanding. Structural insights are not merely beneficial; they are essential for the development of effective mitigation strategies. We therefore advocate for a structure-based paradigm for managing Be and Tl risks, grounded in advanced characterization techniques and informed by interdisciplinary collaboration across crystallography, mineralogy, process engineering, and environmental science. By elucidating how these elements are incorporated at the atomic level—and by linking structure to reactivity—we can enable the rational design of extraction processes, targeted waste stabilization, and robust containment solutions. For example, elucidating the structural and chemical transformations that occur during high-temperature calcination processing of lithium ores is essential to understanding the mechanisms by which Be and Tl transition into leachable forms. This mechanistic insight enables the rational design of upstream metallurgical processes to control pollution at its origin, forming a critical complement to downstream waste management strategies.

This shift is urgently needed. While the rapid growth of the new energy sector is both necessary and inevitable, it must not proceed at the expense of environmental integrity or human health. At present, the effective treatment of lithium-processing waste and co-existing toxic elements like Be and Tl has become a critical bottleneck in achieving sustainable development across the lithium value chain. Overcoming this barrier will require coordinated efforts from structural chemists, mineralogists, and environmental scientists to deliver actionable technological breakthroughs. Building upon the insights gained for Be and Tl, other co-occurring elements—such as Cr, Mn, Pb, and even Li itself—also warrant deeper investigation due to their emerging environmental impacts, while recent studies have begun to highlight the environmental risks associated with lithium loss and dispersion [14–16], a more systematic understanding and mitigation framework remains urgently needed. Ultimately, the success of the clean energy transition depends not only on securing sufficient lithium resources but on ensuring that lithium is extracted without leaving behind a legacy of toxic contamination.

Funding

This work was supported by the National Natural Science Foundation of China (22336006, 22494684), the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (52121004), the National Key Research and Development Project (2024YFC3909701), and the Strategic Priority Research Program of the Chinese Academy of Sciences (XDC0230101).

Conflict of interest

The authors declare no conflict of interest.

References

© The Author(s) 2025. Published by Science Press and EDP Sciences.

All Figures

	Figure 1 A balance must be maintained between lithium extraction and its utilization.
In the text