Lithium-ion battery components are at the nexus of sustainable energy and environmental release of per- and polyfluoroalkyl substances

Abstract

Lithium-ion batteries (LiBs) are used globally as a key component of clean and sustainable energy infrastructure, and emerging LiB technologies have incorporated a class of per- and polyfluoroalkyl substances (PFAS) known as bis-perfluoroalkyl sulfonimides (bis-FASIs). PFAS are recognized internationally as recalcitrant contaminants, a subset of which are known to be mobile and toxic, but little is known about environmental impacts of bis-FASIs released during LiB manufacture, use, and disposal. Here we demonstrate that environmental concentrations proximal to manufacturers, ecotoxicity, and treatability of bis-FASIs are comparable to PFAS such as perfluorooctanoic acid that are now prohibited and highly regulated worldwide, and we confirm the clean energy sector as an unrecognized and potentially growing source of international PFAS release. Results underscore that environmental impacts of clean energy infrastructure merit scrutiny to ensure that reduced CO₂ emissions are not achieved at the expense of increasing global releases of persistent organic pollutants.

Fig. 1. Minnesota field data.

Concentrations of bis(trifluoromethylsulfonyl)imide (bisFMeSI) linked to atmospheric deposition in snow (January; ng L⁻¹) and soil (June; ng kg⁻¹) 2022 (a); concentrations of bisFMeSI in surface water (June; ng L⁻¹) and sediment (ng kg⁻¹) in the northern (b) and southern (c) sampling regions in June 2022. The concentric circle reflects concentration at MN 22, which is immediately downstream of MN 4. Blue arrows are the primary wind direction based on local windrose data, and red markers denote the location of 3M Cottage Grove. Base maps are from the United States Geological Survey. ND non-detect. Source data can be found in Supplementary Data 1–3.

Fig. 2. D. Magna swimming track density and metrics.

One minute of representative swimming track density for each of 10 D. magna replicates exposed to 0 (a) and 5000 (b) ng L⁻¹ bis(trifluoromethylsulfonyl)imide (bis-FMeSI). Swimming parameters of distance and percent time swimming (c), and swimming velocity (d). In box and whisker plots, the lower and upper extent of each box represent the 25th and 75th percentiles, and the whisker represents the 1.5 interquartile range (1.5 IQR). Data beyond the 1.5 IQR are considered outliers. Letters reflect statistical differences in the variance of treatments vs. the control based on $X^2$ tests (p ≤ 0.002). Note that n = 10 D. magna per dose except 5 ng L⁻¹ (n = 7), 1000 and 5000 ng L⁻¹ (n = 9) where organisms were immobilized or died prior to data collection. Source data can be found in Supplementary Data 8 and in the zip file Track Density Raw Data Full.zip.

Fig. 3. Zebrafish larval locomotion.

Mean distance traveled (mm) during four, 10 min phases of the assay (two light and two dark) by larvae exposed to 25–250,000 ng L⁻¹ bis(trifluoromethylsulfonyl)imide (bis-FMeSI) (a) and field-collected aqueous samples MN 4 (including duplicate), MN 22, and MN 15 (b). Error bars represent means ± standard error of the mean. Different letters represent statistical differences within each parameter (one-way ANOVA or Kruskal–Wallis p < 0.05; n = 30 zebrafish per dose). Individual p values are provided in Supplementary Data 25 (a) and Data  26 (b). A version of these plots overlaid with individual data points is in Supplementary Fig. 20. Source data can be found in Supplementary Data 21–29.

Fig. 4. Treatability of bis-FMeSI by GAC and IX.

Normalized breakthrough (C/C₀) curves of bis(trifluoromethylsulfonyl)imide (bis-FMeSI), perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) obtained with granular activated carbon (GAC) in coagulated, settled surface water with a TOC concentration of 2.3 mg L⁻¹ (a) and ion exchange resin (IX) in groundwater with a TOC concentration of 4.6 mg L⁻¹ (b). The IX rapid small-scale column test (RSSCT) was also conducted with the coagulated, settled surface water, but no meaningful (<10%) bis-FMeSI breakthrough was observed. Breakthrough curves of all 21 PFAS are shown in Supplementary Figs. 21 and 22. Source data can be found in Supplementary Data 29. TOC total organic.