Degradation of PFOA solutions and PFAS-contaminated groundwater using atmospheric non-thermal plasma treatment Open Access

Per- and polyfluoroalkyl substances (PFASs) are ‘forever chemicals’ found in our surroundings (Buck et al. 2011), which are composed of carbon chains with attached fluorine atoms. They are extremely stable chemically and resistant to biological degradation due to the carbon/fluorine bond. In the past, PFASs were often referred to as per- and polyfluorinated compounds (PFCs). Currently, PFASs are widely favored nomenclature. Perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) are part of a large group of laboratory-made chemicals. PFASs are utilized in a wide range of industrial and consumer products and have the ability to bioaccumulate and undergo biomagnification, which has been documented in multiple studies (Stahl et al. 2011). They can be detected in various sources, including water, plants, animals such as fish, birds, and mammals, as well as in human blood and breast milk. The sources of groundwater PFAS are mostly based on ground pollution sources and enter the groundwater through the aeration zone (Liu et al. 2018). Many publications have highlighted the adverse impact of PFASs on human health, leading to their proposal as a new type of ‘persistent organic contaminants’ (Bell et al. 2021; Carnero et al. 2021; Sun et al. 2023).

It is estimated that there are more than 12,000 species of PFAS (Spyrakis & Dragani 2023) present in the environment, yet common testing methods can identify only a couple dozen (Glüge et al. 2020). For PFOA the tolerable uptake is 1.5 μg/kg bodyweight per day throughout a lifetime of 70 years and for PFOS 0.15 μg/kg. The recommended TDI (tolerable daily intake) value for PFOS is 150 ng/kg BW/day according to the European Food Safety Authority (EFSA) (Stahl et al. 2013).

In this study, we introduced an innovative atmospheric falling-film plasma reactor designed for the efficient degradation of per-and PFASs. The plasma reactor has a hollow stainless-steel electrode that has resistance to corrosion, making it suitable for applications when exposed to harsh environments. The study focused on varying exposure time, selecting appropriate carrier gas, and determining the impact of using a cooling system on the degradation pathway of PFAS. By developing and testing this method, our research makes a significant contribution to reducing PFAS concentrations in contaminated environments. This aligns with global health and safety standards and advances our capabilities in environmental management and pollution control.

PFOA (95%) was purchased from Sigma–Aldrich, Germany, and deionized water was used after Millipore filtration. The PFOA standard solution was prepared at 80 μg/L in the laboratory using deionized water. The study was performed in two trials, trial-1 signified the degradation of PFOA solution and in trail-2, the degradation of PFAS-contaminated groundwater (HYDR.O. Geologen und Ingenieure, Germany) was investigated. The study area consists of three different groundwater locations in North Rhine-Westphalia (Germany) that have been shown to have come into direct contact with PFAS. To obtain the most comprehensive overall picture of the existing pollution, the samples were taken at intervals from one to three different locations in the respective area. The major source of PFAS-contaminated groundwater was extinguishing water and landfills.

At the start of the experiments, the reactor was filled with 5 L of 80 μg/L of PFOA standard solution and then treated for time intervals of 30 and 120 min using a mixture of oxygen and nitrogen gas by either using with or without a cooling system. The initial pH of the PFOA standard solution was around 5.5. Afterwards, using a cooling system for temperature control, a mixture of 60% oxygen and 40% nitrogen at 300 volts was employed to degrade the samples.

The degradation of PFOA standard solution and PFAS-contaminated groundwater after plasma treatment was analyzed using High-Performance Liquid Chromatography coupled with Tandem Mass Spectrometry (HPLC-MS/MS) by Eurofins Umwelt Südwest GmbH. The evaluation of PFAS was conducted using a 1290 Infinity HPLC system from Agilent Technologies (USA). Solvents for the analysis included water and MeOH with 0.1% formic acid. The samples underwent fractionation on a ZORBAX Eclipse Plus C18 separation column (Agilent Technologies) at 40 °C and a flow rate of 0.4 mL/min. The determination of PFAS was an LTQ XL linear ion trap mass spectrometer from Thermo Fisher Scientific (USA), equipped with an ESI source and operating in positive ion mode.

The experimental treatments were analyzed using one-way ANOVA and data were reported as mean and standard deviation (SD) at P < 0.05. SPSS 23 was used for statistical analysis. The electrical power of the DBD reactor using a mixture of oxygen and nitrogen as the carrier gas was calculated using the software LabView 2018.

For the plasma treatment lasting 30 min, the energy yield was found to be 1.25 mg/kWh for the carrier gas mixture of 60% O₂ and 40% N₂, and 2.5 mg/kWh for the mixture of 20% O₂ and 80% N₂. Similarly, for the treatment duration of 120 min, the energy yields were 15.31 mg/kWh for the 60% O₂ and 40% N₂ mixture, and 12.81 mg/kWh for the 20% O₂ and 80% N₂ mixture.

According to Table 1, the results showed the differences between timing, gases used, and cooling system had no significant difference (P > 0.05). However, employing 120 min treatment time showed a lower mean PFOA concentration of 36.33 μg/L and higher reduction efficiency of 54.9%, indicating no statistical significance compared to the 30 min treatment. Comparing different gas compositions, the treatment with O₂ (20%), and N₂ (80%) had the lowest mean PFOA concentration of 29.00 μg/L and the highest reduction efficiency of 63.7%. Likewise, the cooling treatment showed a mean PFOA concentration of 43.00 μg/L and a reduction efficiency of 46.25%, while the treatment without cooling had a mean concentration of 52.00 μg/L and a reduction efficiency of 35.00%. However, the difference in mean PFOA concentration between these two treatments is not statistically significant (P > 0.05).

Table 2 provides a summary of the overall formation of by-products resulting from the plasma treatment of PFOA, highlighting the effects of different contact times. For the byproduct PFBA (perfluorobutanoic acid), at 30 min, the mean concentration was 2.22 μg/L, and the observed difference in mean concentration between 30 and 120 min was not statistically significant (P > 0.05). Conversely, for PFPeA (perfluoropentanoic acid), at 30 min, the mean concentration is 3.62 μg/L. However, at 120 min, the mean concentration notably increases to 6.37 μg/L, indicating a non-significant difference again in concentration over time. Similar patterns can be observed for PFHxA (perfluorohexanoic acid) and PFHpA (perfluoroheptanoic acid), where the mean concentrations increase with longer contact times, although the differences are not statistically significant (P > 0.05). Overall, the table provides insights into the formation of specific by-products of PFOA, with implications for understanding the kinetics and mechanisms of byproduct formation in treatment processes under different contact times.

RWS1, RWS2, and RWS3 are the PFAS-contaminated groundwater samples from three different locations in North-Westphalia, Germany. RWS1 and RWS2 are samples from an airport area while RWS3 is from leachate. In Table 3, after 2 h of experiment, the total PFAS concentration in RWS1 decreased from 3.6 to 3.3 μg/L. Conversely, the total PFAS concentration in both RWS2 and RWS3 increased.

To compare the effects of plasma technology on the degradation of per- and PFASs, three parameters were considered which include the operation time, carrier gas, and with or without cooling system. During trial-1, due to its adsorption properties, the initial concentration of the PFOA standard solution was determined 80 μg/L. PFAS are known to have high adsorption capacities due to their unique properties, such as high surface activity and hydrophobicity (Lei et al. 2023). To address the sorption, Singh et al. (2019) used fixed concentrations of 8.2 and 8.1 mg/L for PFOA and PFOS, respectively. Another study reported adsorption capacities ranging from 0.56 to 90 mg/g for PFOA and from 0.44 to 330 mg/g for PFOS, depending on the adsorbent material used (Yu & Hu 2011). In our study, the major source of PFAS-contaminated groundwater was extinguishing water and landfills. It was observed that the original concentrations of PFAS-contaminated groundwater RWS2 increased by 56.45% and RWS3 increased by 6.67% after treatment in the plasma reactor (Table 3), which indicates the presence of nondetectable precursors. To address this issue, Tsou et al. (2023) suggested using the total oxidizable precursor assay (TOP assay) as a pretreatment. This method involves oxidizing and converting undetectable precursor substances into PFCAs, which can then be accurately measured using HPLC–MS–MS analysis. Singh et al. (2021) removed 60 ± 2% of the TOP (six short-chain perfluoroalkyl acids (PFAAs) and eight PFAA precursors) concentration ranging around 10⁶ ng/L.

In addition to our research showcasing a considerable reduction of 63.75% in PFOA concentration following air plasma treatment, and an energy efficiency of 24.63 mg/kWh, Lewis et al. (2020) reported a higher reduction percentage of 93.1% for PFOS concentration with air plasma treatment. However, their study also demonstrated lower reduction percentages of 65 and 36% for PFOS (100 mg/L) concentration with N₂ and O₂ plasma, respectively. Despite these variations, their reported energy consumption for PFOA removal stands at 47 mg/kWh. Similarly, in our study focusing on groundwater degradation, an 84% reduction in PFOS (1.7 μg/L) was achieved using an 8 W reactor with a plasma gas composition of 60% oxygen and 40% nitrogen over a duration of 120 min (Figure 2). This variability underscores the differences in efficacy and efficiency of plasma treatment methods across various experimental conditions and contaminants.

Furthermore, a comprehensive review of plasma-based PFAS removal processes in the literature reveals a diverse array of methodologies and outcomes (Table 4). Studies employing different discharge types, carrier gases, and energy inputs have demonstrated significant reductions in PFAS concentrations. For instance, Saleem et al. (2020) achieved a 49% removal of PFOA, while Nau-Hix et al. (2021) attained a 97% removal of perfluorooctane sulfonic acid (PFOS) using DBD with helium gas. Cheng et al. (2022) achieved a remarkable 99% removal of PFOA with argon gas. Additionally, novel plasma configurations such as plasma jets and gliding arc plasma (GAP) have shown promising results in PFAS removal. Singh et al. (2019) pointed out a 90% removal of PFOS and PFOA using a custom-designed plasma reactor, while Aziz et al. (2021) reported a 95% removal of PFOS using pulsed corona discharge with oxygen gas. Palma et al. (2021) demonstrated >99% removal of PFOS using pulsed streamer discharge with air. These studies collectively highlight the versatility and efficacy of plasma-based technologies in PFAS remediation, offering valuable insights for future research and application in environmental remediation efforts.

Regarding energy requirements, the NTP technologies are very promising, they consume the same energy level (1–100 kWh/cm³) of photo-fenton and pure electrolytic AOPs. In contrast, UV-based photocatalysis, ultrasound, and microwave-based AOPs have much higher energy consumption (>100 kWh/cm³) (Miklos et al. 2018). In our study, PFOA standard solution showcased a degradation efficiency of 63.75%, when used O₂ (20%) and N₂ (80%) and 8 W (24.63 mg/kWh) after 2 h (Table 1). In accordance, Saleem et al. (2020) achieved 80% efficiency in degrading PFOA (41.1 ppm) using a 19 W (87.4 mg/kWh) pulsed discharge for 30 min. Likewise, Palma et al. (2021), got a 46% efficiency in PFOA degradation using 300 W (6 mg/kWh) over a duration of 1 h. While the energy efficiency of our process falls within a competitive range compared to other reported methods, further optimization may be warranted to enhance removal efficiency and reduce energy consumption. Continued research and development in plasma-based technologies hold promise for addressing PFAS contamination in environmental and industrial settings effectively.

In our study, the individual PFAS compounds exhibited a specific sequence that was consistent across various treatments and largely aligned with their respective concentrations. Specifically, we observed a degradation pattern of PFHpA > PFHxA > PFPeA > PFBA, which is indicative of a degradation pathway proposed by Singh et al. (2019). This pathway suggests that the plasma treatment gradually shortens the carbon–fluorine (CF) chains of PFAS through reactive species. A similar pattern was also observed in the study by Singh et al. (2021), where PFHpA, PFHxA, PFPeA, and PFBA were among the typical degradation products occurring during the plasma treatment of perfluorooctanoic acid (PFOA).

Furthermore, Jovicic et al. (2018) explored non-equilibrium plasma discharges, particularly the reverse vortex GAP, which were investigated for PFAS removal using air, nitrogen (N₂), and pure oxygen (O₂) gases. Among these, air achieved the most rapid destruction of the tested compounds, with high percentages of PFNA, PFOS, and 8:2 FtS destroyed within a short duration of treatment. These findings suggest that the composition of the carrier gas significantly influences the degradation efficiency and formation of by-products. The observation that nitrogen and oxygen gas were less effective than air in the destruction of the three tested PFAS compounds implies the involvement of both reactive oxygen species (ROS) and reactive nitrogen species (RNS) in the degradation mechanism (Lewis et al. 2020).

In this study, the degradation of the PFOA standard solution using a coaxial DBD falling-film reactor revealed insights into the effectiveness of treated parameters. A simple air plasma setting up yielded a 63.75% reduction of PFOA concentration with an energy efficiency of 24.63 mg/kWh. This setup exhibits higher efficiency compared to using a carrier gas mix of 60% O₂ and 40% N₂. The plasma treatment of PFAS-contaminated groundwater samples yielded diverse outcomes with a decrease in concentration observed in one sample and increased in others. This variability suggests, the presence of undetectable precursors or incomplete degradation processes. Further optimization and refinement of treatment protocols are warranted to enhance removal efficiency and address the complexities associated with diverse PFAS compositions and environmental matrices. Clarifying the underlying mechanisms driving treatment outcomes, particularly regarding the formation and fate of by-products, is crucial for advancing plasma-based remediation strategies. Collaborative interdisciplinary efforts are essential to develop holistic and sustainable solutions for addressing PFAS contamination and safeguarding environmental and human health.

We gratefully acknowledge funding from the German Federal Ministry of Education and Research (BMBF) for the ‘AtWaPlas’ project (Promotional reference 02WQ1601B) under the ‘KMU-innovativ: Ressourceneffizienz und Klimaschutz’ funding measure. The authors also express gratitude to Fraunhofer IGB, Stuttgart, for the invaluable opportunity to collaborate on this research.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.