Modification of biochar by iron containing adsorption centers as a method to enhance the remediation of perfluorooctanoic (PFOA) and (PFOS) acids from water and soil: a density functional theory study

Abstract

Context

Perfluoroalkyl and polyfluoroalkyl substances (PFAS), with over 15,000 types listed in the US EPA’s CompTox database, are found in everyday items like textiles, packaging, firefighting foams, and medical devices. Their widespread use has led to concerning health effects—including cancers, elevated cholesterol, and fertility issues—with detectable levels present in 98% of Americans.

While perfluorooctanoic (PFOA) and perfluorooctanesulphonic (PFOS) are among the most studied, their environmental behavior and ecological interactions remain poorly understood. Advances in computer-based methods, including chemoinformatics and quantum modeling, now aid in predicting properties and simulating PFAS dynamics.

Biochar (BC), produced via biomass pyrolysis under limited oxygen, is known for its porosity and adsorption capabilities. Magnetic biochar (MBC), enhanced with iron-based compounds, adds the benefit of magnetic separation, making it ideal for water decontamination. This paper explores the use of MBC to remove PFOA and PFOS from the environment, leveraging computational tools to investigate molecular interactions and adsorption properties.

Methods

A doubled crystallographic unit of hematite (Fe₂₄O₃₆) was constructed and fully optimized using density functional theory (DFT) with the M06-2X functional. Geometry optimization used the 6-31G(d,p) basis set, while single-point energies were calculated with 6–311 + + G(d,p). Antiferromagnetic conditions were ensured by setting the total spin to zero (Sz = 0), and triplet instability analysis was performed to evaluate ferromagnetic potential.

To simulate bulk water effects on adsorption, the CPCM solvation model (ε = 78.3) was applied. Harmonic frequency analysis confirmed structural minima, and Gibbs free energies were calculated using Gaussian 16. PFOA and PFOS, with highly negative pKa values (~ –0.1 and <).

Quadratic SCF convergence (scf = qc) addressed numerical challenges, and interaction energies were corrected for basis set superposition error using the counterpoise method. Calculated IR spectra and molecular visualizations were generated with Chemcraft, without applying scaling factors.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00894-025-06491-9.

Introduction

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are a class of chemicals characterized by having at least one fully fluorinated carbon atom in the structure. The history behind their discovery dates back approximately 90 years. In 1934, two scientists from IG Farben, a German company, discovered polychlorotrifluoroethylene (PCTFE). In 1938, Dr. Roy J. Plunkett and his group synthesized polytetrafluoroethylene (PTFE) that we know as Teflon™ [1]. Today, the US EPA’s database, CompTox [2], lists approximately 15,000 different types of PFAS. The size of this group is a consequence of the desired characteristics of stability and resilience. Such stability is explained based on consideration of two factors, namely, the intrinsic power of the atom to attract electronic charge and the atom’s ability to absorb a gain or loss of electronic charge. A recent study [3] demonstrates that these fluorine atom properties can be attributed to the low charge capacity of atomic fluorine, which is linked to its low polarizability. This limits the ability of covalently bonded fluorine to acquire the degree of negative character typically expected based on its high intrinsic electronegativity. This is the quantum-chemically based explanation why PFAS are used in hundreds of products, including stain-resistant textiles, food-handling materials, firefighting foams, construction materials, personal care products, and medical devices.

The enormous commercial value of added/used PFAS has brought dire side effects, such as detrimental health problems, such as various cancers, obesity/increased cholesterol, and decreased fertility. The common exposure occurs through the direct use of commercial products containing PFAS and indirectly through environmental contamination. The typical indirect pathways include drinking water (without removing PFAS), food grown on contaminated soil or in contaminated water, and insufficient water and wastewater treatment (not adjusted for PFAS removal), among others.

The research on health effects related to PFAS concurs with the results of a study by the Centres for Disease Control and Prevention (CDC) conducted between 2000 and 2014. It was found that 98% of Americans have various detectable levels of PFAS in their blood. Currently, investigations and database updates are routinely conducted through community-wide blood testing [4, 5].

Finally, on April 10, 2024, the US EPA announced the National Primary Drinking Water Regulation (NPDWR) for six PFAS, including enforceable maximum contaminant levels (MCLs). The regulated six PFAS in drinking water include perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorohexane sulfonic acid (PFHxS), perfluorononanoic acid (PFNA), and 2,3,3,3-tetrafluoro-2-(heptafluoropropoxy)propanoic acid (HFPO-DA) with individual MCLs, and PFAS mixtures containing at least two or more of PFHxS, PFNA, HFPO-DA, and also perfluorobutanesulfonic acid (PFBS) using a hazard index MCL for the combined and co-occurring levels of these PFAS. In addition, the EPA finalized health-based, non-enforceable maximum contaminant level goals (MCLGs) for these PFAS. The final rules include the completion of initial monitoring (by 2027) of public water systems, implementation (by 2029) of solutions to decrease excessive levels of PFAS, and, after 2029, mandatory notification to the public when violations persist [5]. Within the EU, total PFAS are limited to 0.5 µg/L and levels of 20 individual PFAS to 0.1 µg/L in drinking water under the revised Drinking Water Directive set by European Environmental Agency.

While PFOA and PFOS have been the subject of extensive research, their fate, behavior, and interaction in natural ecosystems remain inadequately understood. Therefore, further research is needed to enhance our understanding of PFAS and develop effective strategies for managing its environmental impact [6]. In light of this statement, computer methods are becoming increasingly helpful [7] (both chemoinformatic (e.g., machine learning) methods—for predicting physicochemical properties important in assessing the environmental fate of PFAS as well as computer chemistry methods allowing for modeling the processes) [8].

Several studies have investigated diverse approaches for PFOA and PFOS remediation [9–11]. Such adsorbents as nanomaterials, clay, biochar (BC), ion exchange resins, polymers, graphene, carbon nanotubes, and minerals have been recognized as effective agents for the removal of PFOA and PFOS from wastewater [11–14]. However, the current technological demands for adsorbents are multifaceted and encompass attributes beyond those provided by mostly natural materials. In essence, modern adsorbents must demonstrate adsorption properties that surpass those of their natural counterparts.

BC represents a carbon-containing product obtained by biomass pyrolysis (e.g., wood waste, agricultural waste) in conditions of limited oxygen availability (see for example [15, 16]). This material is widely used to improve soil fertility, as a carbon storage agent, or as a filtration medium for purifying water due to its high porosity and ability to adsorb various substances. It is possible to distinguish five potential parameters of modification that affect the BC adsorption capacity of PFAS, which are (i) temperature, (ii) pH, (iii) coexisting contaminants, (iv) contact time, and (v) ionic strength.

Magnetic biochar (MBC) has been modified to exhibit the magnetic properties of BC, thereby expanding the scope of its applications. MBC is made by adding to biochar magnetic materials, such as iron or iron oxides [17–19], during or after its production process. One of the key advantages of an MBC is its ability to effectively remove contaminants from water. In addition, after the process, MBC can be easily separated from the water using a magnet.

In this paper, we continue our investigations into the adsorption ability of coal-like materials and their derivatives towards interactions with various environmental contaminants. In initial studies, we developed computational protocols that enable the more accurate prediction of the Gibbs free energy of adsorption than was previously feasible using routine density functional theory (DFT) approximations [20]. The following paper reports an investigation of the nanocomposites formed by graphene oxide and polyvinyl alcohol [10]. Then, we focused on addressing potential solutions to well-known environmental issues, such as the removal of PFOA and PFOS from the environment [21]. Several of our investigations are devoted to the interaction of iron-containing compounds with the species, which are, in fact, environmental pollutants. An investigation of iron-containing compounds is known to be the most challenging task for quantum chemistry because of the number of theoretical and computational problems associated, in particular, with the open-shell electron structure of these systems [22, 23]. The current paper extends the previous study and provides a novel approach to investigations of the interaction of PFOA and PFOS with the Fe₂O₃ component in magnetic biochar.

The use of MBC is a promising solution for the easy separation and regeneration of adsorbents after adsorption. To enhance the sorption performance of PFOA and PFOS and facilitate the separation of post-adsorbents by an external magnetic field, a feasible method was used to prepare MBC. The synthesis of MBC involves incorporating various nanoparticles, of approximately 100 nm in size, and micro-sized ferromagnetic metals, into the feedstock materials. The magnetization of biochar can be achieved either before pyrolysis, through pre-modification, or after pyrolysis, through post-modification. To facilitate the easy separation of adsorbents after post-adsorption, magnetic properties can be induced by co-precipitating iron nanoparticles in and around feedstock materials. Furthermore, the incorporation of metal nanoparticles, such as MgO, Fe₂O₃, CaO, La₂O₃, and Al₂O₃, in BC increases the adsorbent’s positive charge, making them ideal for anion sorption. The synthesis and application of MBC have been used successfully to remediate various heavy metal(loid)s and organic contaminants. However, the adsorption of PFOA and PFOS by MBC and their interfacial interaction sorption mechanisms require further verification.⁵

Materials and methods

As we already mentioned, due to the large size, we were not able to model the size of Fe₂O₃ species observed experimentally at a reliable quantum-chemical level. Instead, the crystallographic unit of hematite taken from the study [24] was initially adopted and multiplied twice in the direction of the crystallographic axis a. Obtained in this way, a structure having chemical composition Fe24O36 (see Fig. 1A) has been fully optimized at the density functional level of theory (DFT). Specifically, we employed the M06-2X exchange–correlation functional, which has been demonstrated to be effective in our recent study [22]. Also, the 6-31G(d,p) basis set was used to optimize the geometry and 6–311 + + G(d,p) for single-point calculations. To keep the system to be antiferromagnetic, electronic spin has been assigned to zero (S_z = 0). To verify the ability of the optimized Fe₂₄O₂₆ structure to possess also a ferromagnetic state, the analysis of single determinant DFT wavefunction regarding triplet instability has been performed.

Fig. 1.

Initial (A, B) and optimized (C, D) geometrical structure of Fe₂₄O₃₆ species

Since the separation of the adsorbed substances mostly occurs from the bulk water, the CPCM model, which mimics the influence of the bulk water, was applied with a dielectric permittivity of 78.3 [25]. The Cartesian coordinates of all investigated species are presented in the Supplemental Materials.

Calculations of harmonic vibrational frequencies have verified all the structures. The Gibbs free energies have been calculated as implemented in the Gaussian 16 program package using all calculated parameters at 6-31G(d,p) level, except the total energy, which was calculated at the 6–311 + + G(d,p) level.

It is essential to acknowledge that PFOA and PFOS have significantly acidic character, with pKa values of approximately − 0.1 (values of PFOA and other highly fluorinated carboxylic [26] and much less than zero [27], respectively). Consequently, for PFOA, the amount of PFOA anions will exceed the concentration of its non-dissociated molecular form by over five times. Given the negative pKa values, the presence of the non-dissociated form is expected to be minimal. Therefore, for PFOA, both the anionic and the non-dissociated forms were considered for adsorption, whereas for PFOS, only the anionic form was accounted for in the adsorption analysis.

Due to the numerical unconvergence of the routine SCF procedure implemented in Gaussian 16, quadratic convergence was applied (scf = qc). To correctly calculate interaction energy, a basis set superposition error has been considered as a counterpoise correction.²⁷

All the calculations were performed using the Gaussian 16 program package [28].

IR-vibrational spectrum, as well as obtained results, has been visualized using the Chemcraft program [29]. No scaling factors have been incorporated into computational results.

Results and discussion

There are two drawbacks that we can only be considered partially in this work. As mentioned above, the size of Fe₂O₃ composites adsorbed or chemosorbed by the surface of biochar is near 100 nm. Previously, the largest fragment of iron(II) oxide that we were able to model computationally was Fe₁₃O₁₃ [23].

The geometry of Fe₁₃O₁₃ was frozen in this study. Currently, we have significantly extended the size of the iron oxide, investigating the Fe₂₄O₃₆ species. The geometry of this compound has been fully optimized. The reason to consider the interaction of PFOA and PFOS with the optimized structure of Fe₂₄O₃₆ is based on the obvious fact of geometrical relaxation of the Fe₂₄O₃₆ surface from the initial geometry upon the complexation of the investigated species. Initial geometry was the simple superposition of two crystal unit cells of hematite. A relaxation of the surface during the complexation will change surface atoms’ adsorption ability due to the change in their positions (see Fig. 1). Obviously, not exactly such adsorption centers will be formed during the interactions of experimentally studied much large, c.a. 100 nm species; however, we believe in similarities of their active structures.

The next issue is related to choosing the realistic electronic and spin states for the considered species. Previously, to solve this problem for the Fe₁₃O₁₃ nano-particle, we considered it as an associate of 13(FeO) molecules. This assumption allows to assign the initial spin state of Fe₁₃O₁₃ to be equal to S_z = 26. Then, the dependence of the total energy of unoptimized geometry on S_z in the vicinity of 26 value was studied, and the final value of S_z = 28 was assigned. Unfortunately, the size of the considered here system and the necessity to optimize the geometry prevent us from applying the procedure described above. Therefore, we simply took into account the fact that hematite, the most common mineral possessing of a composition of Fe₂O₃, is an antiferromagnetic compound. This is the reason to assign S_z = 0 spin state to the considered Fe₂₄O₃₆ species.

We begin the discussion by considering the changes that have occurred in the initial geometric structure of the Fe₂₄O₃₆ species after the optimization of the geometry (see Fig. 1). To do this, the initial structures (Fig. 1 A and B) have to be compared with the optimized ones (Fig. 1 C and D). The tendency to transform the initial shape of the parallelepiped into a form rather resembling a globule is clearly displayed. During this transformation, the molecular cavity formed during CPCM calculations lost 15% of its volume (from 919 to 786 ų). The formed structure still exhibits certain features of the topology of the initial species, which is especially evident from the comparison of structures B and D (Fig. 1), but with the transition of iron atoms (36, 37, 38, 39 see; Fig. 1A) from the middle layer to the formed layers of the globule (see Fig. 1D).

The values of the coordination numbers of iron and oxygen atoms in the initial and optimized structures are given in Table S1. Analysis of the data displayed in this table showed that during the optimization (relaxation) of the Fe₂₄O₃₆ structure, it is transformed from the parallelepiped like to the globule (see Fig. 1). This is accompanied by an intuitively clear increase in the coordination of iron atoms by oxygen atoms with the dominance of iron atoms coordinated by five oxygen atoms. One may see that five-coordinated iron atoms are located in the centre of the globule, and lower-coordinated ones are at its boundaries. This arrangement of iron atoms prompted us to investigate the interaction of PFOA and PFOS with five-coordinated iron atoms, specifically those at positions 4, 31, and 34 (see Fig. 1). We would also like to mention that in the bulk of the hematite crystal, the coordination number of iron atoms is equal to six [24].

The last issue that we would like to discuss before analyzing the adsorption of PFOA and PFOS is the similarity in magnetic properties of hematite and the investigated structure of Fe₂₄O₃₆. As follows from numerous studies (see for example [30]), hematite is antiferromagnetic below 260 K and exhibits weak ferromagnetic properties between 260 and 950 K. This finding is confirmed by very accurate computations of (Fe₂O₃)_n (n = 1–5) clusters [31]. This investigation basically highlights two facts:

1. Electronic configurations of the clusters (Fe₂O₃)_n (n = 1–5) reveal the appearance of antiferromagnetic and ferromagnetic states.

2. The electronic configurations of the clusters have only small influence on their geometric structure.

To study the possible emergence of ferromagnetic states, we have analyzed the obtained M06-2x/6-31G(d,p) DFT wavefunction on the stability with respect to the UHF solution (ability to form the electronic states containing unpaired electrons). Indeed, we found such instability concerning two double-occupied molecular orbitals of Fe₂₄O₃₆ (see Fig. 1 C and D). The consequence of such RHF wave function instability could be the appearance of so-called broken symmetry ferromagnetic electronic states.

Initially, the molecules of PFOA and PFOS were always oriented perpendicularly to the surface of Fe₂₄O₃₆. As we already mentioned, we considered only the interaction of the five-coordinated iron atom with the carboxylic group of the PFOA and the PFOS molecules. Those five-coordinated iron atoms are located in the centre of the Fe₂₄O₃₆ globule, and one of the oxygen atoms of the carboxylic group obviously interacts with it. The adsorption of PFOA in its undissociated form leads to the dissociation of the carboxyl group, resulting in the formation of a hydrogen bond through the participation of the closest oxygen atom. This case is presented in Fig. 2A. In the case of PFOA and PFOS anions, it is also guessed that the second oxygen atom of the carboxyl group interacts with another iron atom. These cases are indicated in Fig. 2 B and C. We note that both PFOA and PFOS are adsorbed in the so-called skewed form. Such an orientation indicates a particular stabilizing contribution of the electrostatic and dispersion interaction between the surface and the adsorbed molecules. This is in contrast to the interaction of considered species with the surfaces of graphene, graphene oxide, and fluorinated graphene, which exhibit the parallel orientation of PFOA and PFOS regarding the adsorbing surface ²¹. Therefore, we guess that the stabilization contribution is probably smaller than the one which characterizes the interaction of those species with the graphene surface of graphene oxide and fluorinated graphene surface. By making such comparisons, we imply that the structure of the surfaces under discussion can be a simplified model of biochar, which is considered as an effective adsorbent of PFOA and PFOS (see the “Introduction”). Table 1 presents the interaction energy and Gibbs free energy values for all three complexes. Although both values do not contain the correction associated with the basis set superposition error, the range of the analyzed values confidently indicates virtually 100% adsorption of PFOA and PFOS by five-coordinated Fe atoms. Biochar is one of the adsorbents to which hematite (Fe₂O₃) is added. Therefore, it is interesting to compare the adsorption capacity of biochar adsorption centers with the adsorption capacity of five-coordinated Fe atoms generated by us computationally. To facilitate this comparison, we utilized results from our previous study [21] and transferred the calculated values of interaction energies and Gibbs free energies to Table 1. It can be seen that the interaction with the adsorption centers of “pristine” biochar provides almost complete absorption of PFOA and PFAS, but the additional interaction with the adsorption centers of hematite further enhances this adsorption property. This conclusion is in line with the experimental observation that the presence of iron oxides not only makes the easy extraction of biochar from the aqueous bulk more efficient but also enhances its adsorption properties of modified in this way a biochar surface [6].

Fig. 2.

Geometrical structure of adsorbed complexes PFOA (A molecular form; B dissociated form) and C PFOS

Table 1.

Adsorption energies

Adsorption complex	ΔEint	ΔG $\frac{298}{}$	ΔEint(BC)21	ΔG $\frac{298}{}$(BC)21
CF3(CF2)6COOH…Fe24O36	− 53.6	− 36.5	− 10.0 to − 14.8	− 6.3 to − 10.7
CF3(CF2)6COO-…Fe24O36	− 29.7	− 12.6	− 15.2 to − 27.5	− 4.3 to − 7.5
CF3(CF2)7SO3-…Fe24O36	− 129.5	− 107.9	− 18.4 to − 26.8	− 0.4 to − 9.7

Concluding the section related to the discussion of the geometry and energy of adsorption, we would like to note a novel, interesting feature that is not typical for the adsorption of PFOA and PFOS by the surfaces of graphene and graphene oxide. By interacting with the five-coordinated Fe atom, the PFOA is not adsorbed in molecular form. Instead, a proton is transferred to the surface of the hematite to form an ion pair (see Fig. 1 A and B).

Although FTIR studies are among the most common in environmental chemistry, we did not find a large number of publications devoted to studying the interaction between iron oxides and PFOA and PFAS. Perhaps one of the detailed studies is [6] (see also [32]), in which the IR bands have been assigned. Besides, the fragment of the spectrum characterizing the region of vibrations of the atoms Fe and O is a wide, insufficiently resolved region, it is proposed to consider that the peaks between 550 cm⁻¹ to 700 cm belong to the Fe–O functional groups, the peaks at 636 cm⁻¹ and 559 cm⁻¹ represent the vibration of Fe atoms located in tetrahedral and octahedral positions. The peaks at 1616 cm⁻¹ correspond to the bending vibration of moisture content on the bare iron oxide nanoparticles. The peaks at 3413 cm⁻¹ correspond to the hydroxyl functional groups on the surface of the iron nanoparticles (OH⁻). The peak band at 1000–1400 cm⁻¹ corresponds to the vibrations of the -CF₃ and -CF₂- groups that originate from organic fluorine, indicating that the peaks at 1384 cm⁻¹ and 1245 cm⁻¹ represent -CF₂- and -CF₃ bending due to the adsorption of PFOS.

Although the resolutions of our computationally generated spectra appear more detailed (see Fig. 3), we are unable to provide such a detailed interpretation. The data presented in Table 2 show that all the observed bands include several vibrations. We have chosen the most intensive one to assign the band.

Fig. 3.

Computationally generated vibrational spectra of Fe₂₄O₃₆ and adsorbed complexes

Table 2.

Assignment of vibrational bends

Band (cm−1)	Assignment	Band (cm−1)	Assignment
Fe24O36		Fe24O36...HOOCC7F15
386.0	Asymmetric stretching of low coordinated O atoms	451.9	Collective motion of selected Fe and O atoms
484.2	Collective motion of selected low coordinated O atoms	484.9	Collective motion of selected low coordinated low coordinated O atoms
532.8	Collective motion of selected low coordinated O atoms	564.0	Stretching motion of selected low coordinated O atoms
570.8	Collective motion of selected low coordinated O atoms	751.9	Mix collective motion of selected O atoms and stretching vibration of C-F bonds
624.1	Stretching motion of low coordinated selected O atoms	814.5	Collective motion of selected low coordinated O atoms
721.0	Stretching motion of low coordinated selected O atoms	1210.1	Stretching motion of selected C-F bonds
749.3	Stretching motion of low coordinated elected O atoms	1282.6	Stretching motion of selected C-F bonds
768.7	Mix of bending and stretching motion of selected O atoms	1780.3	Asymmetric stretching vibration of C-O bond
812.0	Mix of bending and stretching motion of selected O atoms	3557.0	Stretching vibration of O–H bond
835.9	Mix of bending motion of selected O atoms
849.1	Mix of bending and stretching motion of selected Fe and low coordinated O atoms
915.4	Stretching motion of selected Fe and low coordinated O atoms
945.5	Stretching motion of selected low coordinated O atoms
973.3	Stretching motion of selected low coordinated O atoms
1211.1	Stretching motion of selected C-F bonds
1780.3	Stretching motion of C-O bonds
3557.0	Stretching motion of O–H bonds
Fe24O36...CO2C7F$\frac{-}{15}$		Fe24O36....SO3C8F$\frac{-}{17}$
470.6	Collective motion of selected low coordinated O atoms	404.1	Collective motion of selected low coordinated O atoms
538.6	Collective motion of selected low coordinated O atoms	485.0	Collective motion of selected low coordinated O atoms
660.0	Stretching motion of selected C-F bonds and bonding motion of C-O bonds	532.1	Collective motion of selected low coordinated O atoms
716.7	Collective motion of selected low coordinated O atoms	591.4	Collective motion of selected Fe and O atoms
760.0	Collective motion of selected low coordinated O atoms	626.4	Collective motion of selected low coordinated O atoms
808.9	Collective motion of selected low coordinated O atoms	722.1	Collective motion of selected low coordinated O atoms
870.4	Stretching motion of selected low coordinated O atoms	767.9	Collective motion of selected low coordinated O atoms
972.4	Stretching motion of selected low coordinated O atoms	811.0	Collective motion of selected low coordinated O atoms
12,120	Stretching motion of selected C-F bonds	840.6	Collective motion of selected low coordinated O atoms
1283.2	Stretching motion of selected C-F bonds	1189.6	Asymmetric stretching vibration of S–O bond
1754.2	Stretching motion of C-O bonds	1219.2	Stretching motion of selected C-F bonds
		1291.9	Stretching motion of selected C-F bonds
		1360.6	Stretching motion of selected C–C bonds
			Bending motion of C–C bonds

Analyzing the vibrations of the Fe₂₄O₃₆ globule, we are mostly able to observe only the collective motion of oxygen atoms inside the Fe₂₄O₂₆ species. An example of the vibration representing such collective motion is displayed in Fig. S1, where the displacement vectors of the most contributing vibration to the band with a peak at 749 cm⁻¹. Also, we have not identified where well-resolved vibrations of just five coordinated Fe atoms are involved. This is probably because of the size of the iron-oxygen species considered to be too small compared to the size of the experimentally fixed species. However, the assignment of the bands related to stretching C-F, C-O, and O–H motion is clearly seen from the data presented in Table 2.

Conclusions

In light of the increasing threat posed by PFAS to the environment and humans as well as insufficient knowledge about the spread and presence of these compounds, it is necessary to search for new solutions to limit their occurrence as well as methods of their capture from the environment. In our study, we investigate the magnetic biochar as a solution for the separation and regeneration of adsorbents after adsorption. To perform computational analysis of the interactions of PFOA and PFOS with Fe₂O₃, a component of magnetic biochar, we designed a simplified model with the iron oxide stoichiometry of Fe₂₄O₃₆. Analysis of the interaction of five-coordinated Fe(III) ions of this model with CF₃(CF₂)₆COOH, CF₃(CF₂)₆COO⁻), and CF₃(CF₂)₇SO₃⁻⁻suggests the skewed configuration, which those species possess during the adsorption. Additionally, it is worth noting that the molecular form of adsorption for CF3(CF2)6COOH has not been observed. Instead, the study revealed the proton transfer with the formation of a surface ion pair.

Analysis of the interaction energies has concluded that PFOA and PFOS interact more strongly with the adsorption surfaces formed by coordinated iron ions than with the pristine carbon and oxidized carbon surfaces of biochar.

The computationally generated IR spectra of adsorbed species are not fully comparable with the experimental ones due to the small size of the Fe₂₄O₃₆ species considered in the calculations. They do not exhibit well-resolved vibrations that can identify five coordinated Fe(III) ions. However, they allow one to observe the collective vibrations of oxygen atoms in the Fe24O36 particle and assign C-F, C-O, and Fe–O-H stretching vibrations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution

L.G., T.P., J.L. conceptualised the manuscript, L.G., D.L., and A.S. wrote the manuscript, N.B. wrote the manuscript and prepared figures, T.P. and J.L. supervised the manuscript.

Funding

This work was supported by the US Army Engineer Research and Development Center (ERDC), grant number W912HZ-23–2-0006 and the European Union’s Horizon 2020 research and innovation programme via the PROMISCES project under grant agreement No. 101036449.

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Competing intersts

The authors declare no competing interests.