MaterialsThe biochar (Wakefield Biochar) was purchased from Amazon. Before mixing with the soil, ~300 mg of biochar in a batch was pretreated by rapid electrothermal process for 1 s with input voltage of 60 V to fulfill the required conductivity as the conductive additive. The equipment for this process was shown in Supplementary Fig. 1. The average size of the pretreated biochar was ~150 μm, and its morphology and size distribution are shown in Supplementary Fig. 33. Carbon black (Cabot, Black Pearls 2000, average diameter ~10 nm) or metallurgical coke (metcoke, SunCoke Energy) were also used as the conductive additives. The used PFAS include perfluorooctanoic acid (PFOA, 95%, Millipore-Sigma), heptadecafluorooctanesulfonic acid tetraethylammonium salt (PFOS, 98%, Millipore-Sigma), tridecafluorohexane-1-sulfonic acid potassium salt (PFHxS, 98%, Millipore-Sigma), nonafluorobutane-1-sulfonic acid potassium salt (PFBS, 98%, Millipore-Sigma), and polytetrafluoroethylene (PTFE, AF2400, Runaway Bike). Raw clean soil was obtained from the Rice University campus. The as-collected soil was crushed by a hammer grinder (Wenling LINDA machinery Corporation, DF-15) and then dried in an oven for 2 h at 100 °C to remove the residual moisture.PFAS mineralization by REM process~1.0 mg PFAS were dissolved in 10.00 g of ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis). Then, 10.00 g of raw soil was added, followed by shaking on a horizontal shaker (Burrell Scientific Wrist Action, Model 75) for 24 h and then dried overnight. The specific mixed PFAS concentration was tested by LC-MS and listed in Supplementary Table 1.The electrical diagram and picture of the REM system are presented in Supplementary Fig. 1. During the REM process, a mixture of PFAS-contaminated soil (~200 mg) and carbon conductive additives (~100 mg) with a total mass of ~300 mg was loaded into a quartz tube with inner diameter (ID) of 8 mm and outer diameter (OD) of 12 mm after hand-milling for 2 min. Two graphite rods were fixed on each side of the quartz tube as the separators to avoid contamination from the metal electrodes. The tube was loaded on a customized reaction jig and connected to the external REM power system. The two brass electrodes with O-rings were applied to compress and seal the sample inside the tube to prevent the gas emission (Supplementary Fig. 1c). A spring coiled on the surface of the tube was used to avoid the accumulated pressure-induced breaking of the quartz tube during the REM process. The jig was put into a vacuum desiccator under the vacuum of ~10 mm Hg and then connected to the REM system. The capacitor bank (60 mF) was charged by an AC supply and output a DC pulse. The maximal voltage of the capacitor bank can reach 400 V. The relay with programmable delay time with millisecond controllability was applied to control the discharging time. The input voltage was modulated from 0 to 150 V and the discharging time was regularly set as 1 s. The REM temperature was recorded using two IR thermometers (Micro-Epsilon), the detecting range of which are 200–1500 °C and 1000–3000 °C, respectively. These thermometers are connected to LabView using a Multifunction I/O (NI USB-6009) for real-time temperature recording with time resolution of 0.1 ms. After REM, the samples rapidly cooled to room temperature and were collected for further analysis.To investigate the cation influence on the PFAS mineralization, different alkaline and alkaline earth carbonates, including CaCO3, MgCO3, and Na2CO3, were mixed with PFAS. The metal counterion content is 1.2 mole equivalent compared with the F mole content in PFAS to ensure complete PFAS mineralization. Metcoke was used as the carbon additives and the total sample mass per batch was set as 300 mg. During REM, the input voltage was set as 100 V, and the discharging time was set as 1 s (See details in Supplementary Table 4).For the enlarged sample, a mixture of soil (~7 g) and biochar (~4 g) was loaded into a quartz tube with ID of 16 mm and OD of 20 mm (Supplementary Fig. 54b). A second-generation REM system with a larger capacitance of C = 0.624 F was applied for REM energy input (Supplementary Fig. 54a).For the scale-up, 2 kg of soil mixed with 500 g of metcoke inside a clay pot with the outer diameter of 25.4 cm and the soil depth of ~7 cm. Four graphite rods (30 cm in length and 4 mm in diameter) were inserted into the soil as the electrodes. The resistance of each two adjacent electrodes is ~50 Ω. Each two adjacent electrodes were connected to the external power system (Supplementary Fig. 55) sequentially for the uniform treatment of the soil, with 10 s duration time and 4 times REM with each two adjacent electrodes were conducted in total.PFAS measurement by HPLC and QQQ LC-MS200 mg of PFOA-contaminated soil and REM soil were separately added into 2 mL of ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis), with the soil-to-water mass ratio of 1:10. The mixture was shaken on a horizontal shaker (Burrell Scientific Wrist Action, Model 75) for 2 h for complete extraction. Then, the sample was centrifuged (Adams Analytical Centrifuge) with the speed of 580 g for 2 min, followed by filtration using polyethersulfone (PES) membrane (0.22 µm, Millipore-Sigma). The PES filter had a negligible influence on PFAS detection (Supplementary Fig. 66). Afterwards, the concentration of PFOA in the extractant was determined by HPLC-DAD (1260 Infinity II Agilent) and a WPH C18 column (4.6 mm × 250 mm, 5 μm), where the mobile phase was 50% acetonitrile and 50% of 5 mmol L−1 Na2HPO4 with a flow rate of 0.8 mL min−1 and 50 μL injection volume for each sample10. The calibration curve was prepared by dissolving a known amount of PFOA in ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis) in the range from 1 to 100 ppm (Supplementary Fig. 5).In order to detect the trace amount of residual PFOA after REM (<1 ppm), a QQQ-LC/MS system (6740B LC/TQ, Agilent) using dynamic multiple reaction monitoring (DMRM) was applied. Here, the chromatographic separation was performed on a C18 analytical column (Zorbax Eclipse Pluse C18 Rapid Resolution HT, 2.1 × 50 mm 1.8-µ column, Agilent) with an ultra-high-performance LC (UHPLC) system (1290 Infinity II, Agilent). The aqueous phase consisted of 20 mM ammonium acetate solution, and the organic phase of methanol. The column was operated at a temperature of 40 °C and 40 μL sample was injected each time. The mobile phase flow rate was maintained at 0.4 mL min−1 throughout the run. The column was equilibrated at initial conditions for 3 min before the next injection. The LC-MS system was interfaced to the MS system through an Agilent Jet Spray (AJS) electrospray ionization (ESI) source that was operated in the negative ionization mode. The sample preparation procedure is the same with the HPLC-DAD test, and the extractant needs to be further diluted to keep PFOA concentration in the range of 0.5 to 100 ppb. The calibration curve was prepared by dissolving a known amount of PFOA in ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis) in the range of 0.5 to 100 ppb (Supplementary Fig. 6).For detecting other PFAS (PFOS, PFHxS, and PFBS) in the soil by LC-MS, the PFAS extractants were diluted using 90 vol% ultrapure water and 10 vol% methanol (Millipore-Sigma, HPLC standard, >99.9%) to a concentration within the detecting range of 0.5–100 ppb.The removal efficiency (E) of PFAS was calculated according to Eq. (1),$$E=\,\frac{c\left({{{{\rm{Residue\; PFAS}}}}}\right)\,{\times D}_{1}}{c\left({{{{\rm{Original\; PFAS}}}}}\right)\,\times \,{D}_{2}}\times 100\%$$
(1)
where c(Original PFAS) and c(Residue PFAS) are the concentration of PFAS measured by LC-MS before and after REM, D1 and D2 are the dilution factors of PFAS in the raw soil and REM soil, respectively.Total fluorine content test for CICThe soil sample (~10 mg) was loaded into a combustion furnace (AQF-2100H, NITTOSEIKO ANALYTECH) with the temperature of 1100 °C under 400 mL min−1 oxygen flow. The combusted anions were absorbed by 100 mL min−1 water-saturable Ar and 200 mL min−1 Ar, and then flowed into a gas absorption unit (GA 211, Mitsubishi Chemical Analytech). Afterwards, total F content was analyzed by an IC system (Dionex ICS-2100, Thermo Scientific).Inorganic fluoride measurement by IC200 mg of raw soil and REM soil were separately added into 4 mL of ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis), with the soil to water mass ratio of 1:20. Then, the mixture was immersed into an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min for the extraction, followed by centrifugation (Adams Analytical Centrifuge) with the speed of 580 g for 2 min, and filtration using PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. The concentration of mineralized fluorine ion in the sample was measured by IC (Dionex Aquion, 4 × 250 mm IonPac AS23, AERS 500 Carbonate Suppressor). The calibration curve was prepared by dissolving a known amount of sodium fluoride in ultrapure water (Millipore-Sigma, ACS reagent for ultratrace analysis) in the range of 1 to 5 ppm (Supplementary Fig. 7).The mineralization ratio (R) of PFAS is calculated according to Eq. (2),$$R=\,\frac{c\left({{{{\rm{F}}}}}-{{{{\rm{ion}}}}}\right)\,\times \,r\,{\times D}_{1}}{c\left({{{{\rm{PFAS}}}}}\right)\,\times \,{D}_{2}}\times 100\%$$
(2)
where c(PFAS) is the concentration of PFAS measured by LC-MS, c(F-ion) is the concentration of fluorine ions measured by IC, r is the mass ratio of fluorine atom in a certain PFAS molecular (listed in Supplementary Table 1), and D1 and D2 are the dilution factors of PFAS and fluorine ions, respectively. Note for the PTFE, c(PTFE) was calculated by dividing initial F content from CIC data by r(PTFE)NMR test1 g of PFAS-contaminated soil and REM soil were separately added into 2 mL of deuterium oxide (D2O, 99.9%, Millipore-Sigma). The mixture was immersed into an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min for the extraction, followed by centrifugation (Adams Analytical Centrifuge) with the speed of 580 g for 2 min, and filtration using a PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. The solution was then added into the NMR tube and the chemical shift was tested by NMR spectrometer (600 MHz Bruker NEO Digital NMR Spectrometer).GC-MS testTo study the evolved gases, REM was carried out in a home-designed jig. The evolved gas can vent from the quartz reaction tube through a hollow electrode into a sealed tube with pressure gauge (Supplementary Fig. 11). The REM parameters were kept same as mentioned above. The system was purged with argon gas, and evacuated to ‒75 kPa before REM. After REM, the evolved gases were injected into the GC-MS using a gas-tight syringe. The GC-MS instrument used here was an Agilent 8890 GC system equipped with an Agilent HP-5ms low-bleed column (30 m, 0.25 mm internal diameter, 0.25 µm film) with helium as the carrier gas for liquid and headspace sampling. A tandem Agilent 5977B mass selective detector was used for liquid and headspace gas analysis. For the gas detection, the injector and the transfer line were set with the temperature of 120 and 200 °C, respectively. The temperature program was initiated at 48 °C for 3 min, and then increased to 80 °C at 8 °C min−1. The carrier gas was helium at a flow rate of 0.5 mL min−1. For the F-containing residue detection, ~200 mg REM treated soil samples were added into 5 mL extractant solvent (mixture of hexane, acetone, and toluene with volume ratio of 10:5:1). Then, the mixture was immersed into an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min for the extraction, followed by centrifugation (Adams Analytical Centrifuge) with the speed of 580 g for 2 min, and filtration using PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. The filtered solution was loaded onto a GC autosampler. During the test, the injector and the transfer line temperature were set to 280 and 300 °C, respectively. The temperature program was initiated with 75 °C for 1 min, increased to 230 °C at 10 °C min−1 held for 7 min, then to 280 °C at 20 °C min−1, and held for 15 min. The injection volume was 1 μL each time in a splitless mode, and solvent delay was 5 min to prevent filament damage. The carrier gas was helium at a flow rate of 1.2 mL min−1.Biochar recycling by centrifugationIn all, 500 mg of REM soil mixed with biochar was dispersed into 5 mL of water (Millipore-Sigma, ACS reagent for ultratrace analysis), followed by shaking on horizontal shaker (Burrell Scientific Wrist Action, Model 75) for 15 min. The mixture was then centrifugated (Adams Analytical Centrifuge) with a speed of 580 g for 2 min. After centrifugation, the lightweight biochar floated on the water, while the dense soil particles sank (Supplementary Fig. 19). The floating biochar was then poured and filtered using a sand core funnel (Class F). The separated soil and biochar were then separately dried in an oven at 100 °C for 2 h to remove the residual moisture.Soil calcinationIn total, ~10 g of PFOA-contaminated soil placed in an alumina crucible was heated in a muffle furnace (Carbolite RHF 1500). The sample temperature increases to 900 °C with the heating rate of 20 °C min–1 and maintained at 900 °C for 2 h in the air. Afterwards, the sample naturally cools to room temperature.Infiltration rate testA quartz tube with an ID of 16 mm was used as the container with a sponge to hold the soil samples, enabling fast water penetration. Water can penetrate the sponge rapidly thus does not affect the infiltration rate. The raw soil, REM soil, and calcined soil with the same volume were separately placed on top of the sponges, and 2 cm of water was then gently added atop the soil. The liquid level was measured at different times, and the water infiltration rate was calculated by Eq. (3),$${{{{\rm{infiltration\; rate}}}}}=H/t$$
(3)
where H is the liquid level in cm and t is the time in min (Supplementary Figs. 45 and 46).Particle size measurementTo prepare the samples, we separately added 1.0 g of raw soil, REM soil, calcined soil, or biochar into 10.0 mL of 0.1 M HCl solutions. The carbonate inside the soil was removed by reacting with HCl under an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min. Then, the samples were centrifuged (Adams Analytical Centrifuge, 580 g, 5 min) and washed three times with ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis). Next, 2.0 mL of H2O2 solution (Millipore Sigma, ~35 wt%) was mixed with the soil in a 90 °C water bath for 45 min to remove the soil organic matter55. After another round of centrifugation and water washing three times, 1 g of soil particles were dispersed in 5 mL of water solution with 3.3 wt% sodium hexametaphosphate and 0.7 wt% sodium carbonate. Afterwards, it was injected into a laser particle size analyzer (ZEN 3600 Zetasizer Nano, Malvern, Worcestershire, UK) for particle size measurement. According to the measured data, we further counted the ratio of clay (<2 μm), silt (2–50 μm), and sand (>50 μm) in the soil based on the particle size distribution.Soil carbon content measurementThe soil carbon content was measured using an ECS 4010 – CHNS-O Elemental Combustion System. Before the measurement, 1.0 g of soil sample was dispersed into 10.0 mL of 0.1 M HCl in an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min to remove carbonate. Then, the samples were washed three times with ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis). Afterwards, the sample dried at 105 °C. Acetanilide was used as the standard material for calibration. Raw soil, REM soil, and calcined soil were subjected to carbon content measurement. Each sample was tested in triplicate to obtain the standard deviations.Exchangeable nutrients measurementThe exchangeable P, Ca, K, Mg, Mn, and Fe in raw soil, REM soil, and calcined soil were extracted using the Mehlich-3 reagent56. The extractant is composed of 0.2 M CH3COOH, 0.25 M NH4NO3, 0.015 M NH4F, 0.013 M HNO3, and 0.001 M ethylenediaminetetraacetic acid (EDTA). 1 g of soil sample was added to 10 g of the extract with a soil to solution ratio mass of 1:10. The mixture was shaken immediately on a horizontal shaker (Burrell Scientific Wrist Action, Model 75) for 15 min. Then, the sample was centrifuged (Adams Analytical Centrifuge, 580 g, 5 min), followed by filtration using a PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. The filtrate was diluted to appropriate concentration using 2 wt% HNO3 within the calibration curve range. The P, K, Mg, Mn, and Fe were measured by inductively coupled plasma mass spectrometry (ICP-MS) using a Perkin Elmer Nexion 300 ICP-MS system, with Periodic Table mix 1 for ICP (10 mg L−1, 10 wt% HNO3, Millipore Sigma) as the standard. Due to interference from Ar, Ca cannot be measured by ICP-MS. Therefore, Ca was measured by inductively coupled plasma optical emission spectrometer (ICP-OES) using a Perkin Elmer Optima 8300 ICP-OES system. Ca standard (1000 mg L−1, 2 wt% HNO3, Millipore Sigma) was used for the ICP-OES measurement.The soil nitrate-nitrogen serves as an indicator of available nitrogen for plants. The soil nitrate content was measured using IC (Dionex Aquion, 4 × 250 mm IonPac AS23, AERS 500 Carbonate Suppressor). Nitrate standard calibration solutions were prepared by dissolving NaNO3 in ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis) in the concentration range from 0.5 to 15 ppm. The good linearity of the calibration curve demonstrates the effectiveness of this method (Supplementary Fig. 49). To extract nitrate in raw soil, REM soil and calcined soil, 1 g of soil sample was separately added into 10 g of ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis) for the nitrate extraction and sonicated in an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min. Then, the sample was centrifuged (Adams analytical centrifuge, 580 g, 5 min), followed by filtration using PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. Finally, the filtrate was diluted to a concentration within the calibration curve range of 0.5 to 15 ppm.Soil CEC test1.0 g of soil sample was dispersed into 10 mL of 1 M sodium acetate solution in an ultrasonic bath (Cole-Parmer Ultrasonic Cleaner) for 15 min to saturate soil exchange sites with Na+. Then, the sample was washed three times with ethanol (Decon’s Pure 200 Proof, Decon Labs Inc.) to remove the excess Na+. Afterwards, the sample was dispersed into 10 mL of 1 M ammonium acetate in an ultrasonic bath for another 15 min to replace Na+ by NH4+ at exchange sites55. The sample was then centrifuged (Adams analytical centrifuge, 580 g, 5 min), followed by filtration using PES membrane (0.22 µm, Millipore-Sigma) to remove any undissolved particles. The filtrate was diluted to the appropriate concentration using 2 wt% HNO3 within the ICP calibration curve range. The Na+ concentration was measured by the Perkin Elmer Nexion 300 ICP-MS system, with sodium standard solution for ICP (1 g L−1, Millipore Sigma) as the standard. Finally, CEC was calculated by Eq. (4),$${CEC}=\frac{c\left({{{{{\rm{Na}}}}}}^{+}\right)\,\times \,D\,\times \,V}{M\left({{{{{\rm{Na}}}}}}^{+}\right)\,\times \,m\left({{{{\rm{soil}}}}}\right)}$$
(4)
where c(Na+) is the concentration of sodium measured by ICP-MS, D is the dilution factor, V is the volume of ammonium acetate solution to extract Na+ (V = 10 mL), M(Na) is mole mass of sodium (M(Na+) = 23 g mol−1), m(soil) is the mass of soil sample used for CEC test (m(soil) = 1.0 g).Soil organic content testIn all, 1.0 g of soil sample was dispersed into 10 mL of extractant (V1 = 10 mL), composed of 0.5 M NaOH and 0.5 M Na4P2O7. The mixture was shaken on a horizontal shaker (Burrell Scientific Wrist Action, Model 75) for 1 h, and then heated at 95 °C for 30 min, followed by centrifuging (Adams Analytical Centrifuge, 580 g) for 2 min. After filtrating through a PES membrane (0.22 µm, Millipore-Sigma), we obtained solution 1. For the total organic content test, 1 mL of solution 1 (V2 = 1 mL) was mixed with 5 mL of 0.4 M K2Cr2O7 and 15 mL of 2 M H2SO4, and then heated at 95 °C for 30 min to oxidize the organic compounds in the extractant. After cooling to room temperature, the solution was mixed with 78.9 mL ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis) and 0.1 mL of phenanthroline indicator (1.5 wt% phenanthroline and 1 wt% (NH4)2Fe(SO4)2), which was denoted as solution 2. Afterwards, 0.1 M (NH4)2Fe(SO4)2 was gradually added to solution 2 until the solution’s color changed from orange to green and finally to brick red. The consumed volume of (NH4)2Fe(SO4)2 solution was recorded as V3. For the comparison, 0.1 M (NH4)2Fe(SO4)2 was gradually added to a solution with 5 mL of 0.1 M K2Cr2O7, 15 mL of 2 M H2SO4, 74.9 mL of ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis) and 0.1 mL of phenanthroline indicator. The consumed volume of (NH4)2Fe(SO4)2 solution was recorded as V0 when the solution color changed to green. Therefore, the total organic mass content (corg, with the unit of g kg−1) can be calculated from Eq. (5):$${c}_{{org}}=\frac{0.003\times ({V}_{0}-{V}_{3})\times c({{Fe}}^{2+})\times {r}_{o}\times {r}_{c}}{m}\times \frac{{V}_{1}}{{V}_{2}}\times 1000$$
(5)
Where c(Fe2+) is the mole concentration of (NH4)2Fe(SO4)2 (c(Fe2+) = 0.1 M), m is the soil mass (m = 1 g), ro and rc is the oxidation factor and the conversion factor from organic carbon to organic compound (ro = 1.1 and rc = 1.724).Considering the insolubility of humic acid in acid solution, 2 M H2SO4 was added to 5 mL of solution 1 (V5 = 5 mL) until the pH reached 1 and it was then left for 30 min to separate humic acid from the soil extractant. After filtering using a sand core funnel (class F), the filter residue was washed by 0.05 M H2SO4 for 5 times. Afterwards, the residue was dissolved by 50 mL 1 wt% NaOH and then diluted to 100 mL using ultrapure water (Millipore Sigma, ACS reagent for ultratrace analysis), which is denoted as Solution 3 (V6 = 100 mL). Similarly, 5 mL of 0.4 M K2Cr2O7 and 15 mL of 2 M H2SO4 were used to oxidize 5 mL Solution 3 (V7 = 5 mL) at 95 °C for 30 min and the residual K2Cr2O7 in solution was titrated by 0.1 M (NH4)2Fe(SO4)2 using the phenanthroline indicator. The consumed volume of (NH4)2Fe(SO4)2 solution was recorded as V8. The humic acid mass content (chumic, with the unit of g kg−1) can be calculated from Eq. (6):$${c}_{{humic}}=\frac{0.003\times ({V}_{0}-{V}_{8})\times c({{Fe}}^{2+})\times {r}_{o}\times {r}_{c}}{m}\times \frac{{V}_{1}}{{V}_{5}}\times \frac{{V}_{6}}{{V}_{7}}\times 1000$$
(6)
The fulvic acid content (cfulvic, with the unit of g kg−1) can be thus calculated by:$${c}_{{fulvic}}\,{=c}_{{org}}{-c}_{{humic}}$$
(7)
Arthropod cultureFor the isopod culture, we performed lab microcosm experiments where 2 adult Armadillidium vulgare, a common isopod species, was added to Petri dishes (35 mm diameter, 10 mm height). Four different kinds of soil treatments were tested for isopods, including (1) raw soil, (2) PFOA-contaminated soil, (3) REM soil, and (4) calcined soil. Isopod culture for each soil sample was replicated 7 times. Before the experiment, 1.5 g of the soil sample was added to the Petri dishes and the isopods were hand-picked from a lab-reared culture. Water and isopod food were added every other day to ensure the high survival of isopods. The isopod microcosm experiment was conducted in a light and humidity incubator at 25 °C and with a 12-h night and day cycle. Isopod survival ratio was measured weekly for 4 weeks.A similar microcosm experiment was employed to evaluate the springtails survival ratio in different kinds of soil samples. Four different soil samples for springtail culture include (1) raw soil, (2) PFOA-contaminated soil, (3) REM soil, and (4) calcined soil. Folsomia candida, cataloged as an Isotomidae family member and a Collembola species, is used as a representative type of springtail for the culture. Springtail culture for each soil sample was replicated 7 times. Before the experiment, 1.5 g of the soil sample was added to the Petri dishes (35 mm diameter, 10 mm height). Because of the small size of springtails ( < 2 mm), it is difficult to place the exact same number of springtails into each Petri dish. Thus, we used an aspirator to collect springtails from a lab culture and added 10-15 springtail adults to each dish to inoculate the Petri dishes with springtails. To create suitable conditions for springtails, 1.5 mL of deionized water and ~10 mg of food (i.e., Baker’s Yeast) were replenished every week to each dish. Springtails survival ratio was measured weekly for 4 weeks.The number of remaining isopods and springtails were counted every week throughout the experiment. This process was continued for 28 days (i.e., 4 total weeks with 3 days of counting). We used the survival ratio after each week as the response variable for both isopods and springtails. To specifically test how the soil treatments influence arthropod survival, we employed generalized linear models (GLMs, ref. 57) with two fixed factors: (1) soil types and (2) culture time. We used GLMs with a logit binomial distribution and a Poisson distribution for isopods and springtails, respectively, because of higher data consistency. GLMs were fitted using the “lme4” package in R version 4.3.1 software.Theoretical calculationIn the MD simulation, PFOA molecules were mixed with CaO in a supercell, where the periodic boundary condition was applied in all three dimensions. Four different atomic ratios of F over Ca were set by changing the number of PFOA and CaO molecules. DFT method58 was used as implemented in the Vienna Ab-initio Simulation Package (VASP)59 with climbing image nudged elastic band method (CiNEB)60. A plane wave expansion up to 500 eV was employed in combination with an all-electron-like projector augmented wave (PAW) potential61. Exchange-correlation was treated within the generalized gradient approximation (GGA) using the functional parameterized by Perdew-Burke-Ernserhof 62. With the smallest one being 15.0 Å × 15.0 Å × 22.0 Å, all supercells are big enough. Thus, only Γ point was used for the Brillouin zone integration over Monkhorst-Pack type mesh63. For the structure optimization using the conjugate-gradient algorithm as implemented in VASP, both the positions of atoms and the unit cells were fully relaxed, so that the maximum force on each atom was smaller than 0.01 eV Å−1. For modeling of mineralization reaction, the optimized structures were subsequently annealed for 30 ps with the temperature fluctuating at the range of 1500–2500 K in MD simulation. The MD simulation was performed using Nose-Hoover thermostat and NVT ensemble with a time step of 1 fs. Then, we use the number of F-C bonds in the system as a descriptor of the mineralization effect. The number of unbroken F-C bonds were calculated every 20 steps in each of the MD simulation and the results are shown in Fig. 4f and Supplementary Fig. 39. For counting the number of F-C bonds, the cut-off distance was set at 1.55 Å, as compared to the equilibrium F-C bond-length of 1.44 Å.The calculations of the Gibbs free energy change under different temperatures were conducted using the HSC Chemistry 10 software.Other characterizationsSEM images and element analysis by EDS were taken on the FEI Quanta 400 ESEM FEG system under the voltage of 10 kV and the working distance of 10 mm. XRD was performed using the Rigaku SmartLab system with a filtered Cu Kα radiation (λ = 1.5406 Å). The FT-IR spectra were acquired on the Thermo Scientific Nicolet 6700 attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectrometer (Waltham, MA, USA). Raman spectra were obtained on the Renishaw Raman microscope system (laser wavelength: 532 nm, laser power: 5 mW, lens: 50×). XPS spectra were conducted using the PHI Quantera XPS system under the pressure of 5 × 10−9 Torr. The survey spectra were collected with the step of 0.5 eV and the pass energy of 140 eV, and elemental spectra were collected with the step size of 0.1 eV and the pass energy of 26 eV. All XPS spectra were calibrated using the C 1 s peak at 284.8 eV as the reference. TGA was conducted on the Mettler Toledo TGA/DSC 3+ system using a 70 μL pan with the heating rate of 10 °C min−1 and under 100 mL min−1 air flow. The TGA system was connected to a mass spectrometer (PrismaPro Quadrupole, Pfeiffer Vacuum) for the TGA-MS test with 100 mL min−1 nitrogen as the carrier gas and the heating rate of 10 °C min−1. BET measurements were performed on a Quantachrome Autosorb-iQ3-MP/Kr BET surface analyzer at 77 K, where the nitrogen was used as the adsorption/desorption gas.XRF spectra were acquired by a Panalytical Epsilon 4 XRF instrument. Before test, the soil samples were fused into glass beads using lithium metaborate/lithium tetraborate and lithium nitrate as the fluxing agents using a Katanax K2 Prime instrument. Samples were heated in platinum crucibles to 1000 °C for 15 min while being rocked back and forth for dispersion. After fusion, the platinum crucibles containing the samples were poured into the platinum mold to form beads. The fused beads were then automatically fed into the XRF via the sample loader for continued analysis. The SuperQ analytical software used the documented weights of each sample and its flux weight to generate molar quantitative results. Soil pH was measured by a soil pH meter (SOILPHU), where the detector was inserted into 10 g of soil. The average pH values and the standard deviations for each sample were calculated after 5 times of testing.
Electrothermal mineralization of per- and polyfluoroalkyl substances for soil remediation
