Development and characterization of chlorophyll-amended montmorillonite clays for the adsorption and detoxification of benzene

Abstract

After disasters, such as forest fires and oil spills, high levels of benzene (> 1 ppm) can be detected in the water, soil, and air surrounding the disaster site, which poses a significant health risk to human, animal, and plant populations in the area. While remediation methods with activated carbons have been employed, these strategies are limited in their effectiveness due to benzene’s inherent stability and limited retention to most surfaces. To address this problem, calcium and sodium montmorillonite clays were amended with a mixture of chlorophyll (a) and (b); their binding profile and ability to detoxify benzene were characterized using in vitro, in silico, and well-established ecotoxicological (ecotox) bioassay methods. The results of in vitro isothermal analyses indicated that chlorophyll-amended clays showed an improved binding profile in terms of an increased binding affinity (K_f = 668 vs 67), increased binding percentage (52% vs 11%), and decreased rates of desorption (28% vs 100%), compared to the parent clay. In silico simulation studies elucidated the adsorption mechanism and validated that the addition of the chlorophyll to the clays increased the adsorption of benzene through Van der Waals forces (i. e., aromatic π-π stacking and alkyl-π interactions). The sorbents were also assessed for their safety and ability to protect sensitive ecotox organisms (Lemna minor and Caenorhabditis elegans) from the toxicity of benzene. The inclusion of chlorophyll-amended clays in the culture medium significantly reduced benzene toxicity to both organisms, protecting C. elegans by 98–100% from benzene-induced mortality and enhancing the growth rates of L. minor. Isothermal analyses, in silico modeling, and independent bioassays all validated our proof of concept that benzene can be sequestered, tightly bound, and stabilized by chlorophyll-amended montmorillonite clays. These novel sorbents can be utilized during disasters and emergencies to decrease unintentional exposures from contaminated water, soil, and air.

1. Introduction

Benzene is a volatile organic compound that is a natural component of crude oil and gasoline; it also has several applications in industrial processes as a chemical precursor for numerous products, including pesticides, plastics, and dyes (ATSDR, 2007). Benzene can be identified ubiquitously in the air due to emissions from vehicle exhaust and cigarette smoke as well as the volatilization of petroleum products (ATSDR, 2007). Besides air, benzene has been detected in groundwater and soil as a result of industrial operations, mismanagement of waste (Davis et al., 1994; Fayemiwo et al., 2017; Pinedo et al., 2013; Shores et al., 2017; van Afferden et al., 2011), and accidental releases (Suarez and Rifai, 2002). Near industrial sites, a maximum of 734 ppm (mg/L) (Suarez and Rifai, 2002), 3–27 ppm (Davis et al., 1994), and 14 ppm (van Afferden et al., 2011) have been detected in water samples from local monitoring wells. Measurements of soil and sediments from the Netherlands and Thailand also found 1 and 6.5 ppm (mg/kg) benzene in soil, respectively (UNEP, 2011).

The United States Environmental Protection Agency (U.S. EPA) has set a strict maximum contaminant level of 5 ppb (μg/L) for benzene in drinking water (EPA, 2022), but levels up to 9.28 ppm in Ogoniland, Nigeria (UNEP, 2011; Yakubu, 2017) and 200 ppb in Lanzhou City, China (Liu et al., 2020) have been detected. These concentrations of benzene in public drinking water and soil samples demonstrate the severity of benzene pollution in the environment and the need for mitigation. Furthermore, due to mobilization and redistribution of benzene in polluted soil, sediment, and groundwater during natural disasters, like flooding and other emergencies, there is a critical need for safe and field-practical remediation techniques to decrease the risk of benzene exposure in living organisms. Current remediation methods are based on soil and water pumping (“pump and treat”) (Ciampi et al., 2021; Yang et al., 2012), chemical oxidation (Cui et al., 2017; Yang et al., 2020), and air sparging (Ciampi et al., 2021). However, these methods may not be practical due to their high cost (Casasso et al., 2019), generation of large volumes of waste material, and the introduction of potentially toxic and non-biodegradable elements into the environment (Bustillo-Lecompte et al., 2018).

Therefore, considerable efforts have been made to develop effective remediation techniques for soil and groundwater. Among these, adsorption has been shown to be the most common and economical approach. Table 1 shows a summary of common adsorption methods for benzene remediation and their removal efficiency. Adsorption by activated carbon is among the most well-studied tools for mitigating benzene (Li et al., 2020; Lillo-Ródenas et al., 2011; Vikrant et al., 2019); however, it is reported that carbon performs poorly in humid conditions, as binding sites are saturated with water molecules, a characteristic observed even with the application of hydrophobic coatings at high levels of humidity (Oh et al., 2019; Veksha et al., 2009). The variation in adsorption capacity based on sources of carbons (Chiang et al., 2001; Li et al., 2020) and their required extended reaction time to reach adsorption equilibrium (Kutluay et al., 2019; Wang et al., 2015) also contribute to the limitation of carbonaceous sorbents for rapid mitigation response.

Table 1.

Summary of sorbents with their modifications, adsorption capacity, and model used to derive adsorption parameters.

Sorbent	Modification or Amendment	Adsorption Capacity (mg/g)	Reference
Montmorillonite	1-hexyl-3-methylimadazolium chloride	588.2 (L)	(Anjum et al., 2018)
Sodium Montmorillonite	TMA+	425.3 (BC)	(Deng et al., 2020)
Sodium Montmorillonite	Heated: 120 °C; 600 °C	141.2; 67.9 (BC)	(Deng et al., 2017)
Metal-organic-framework (MOF)	MOF-199 (M199); UiO-66 (U6)	94.8; 27.1	(Vikrant et al., 2019)
Activated Carbon	None	93.5	(Vikrant et al., 2019)
Calcium Montmorillonite	Heated: 120 °C; 600 °C	87.1; 77.5 (BC)	(Deng et al., 2017)
Halloysite	None; heated; acid + heat	68.1; 103.6; 204.2 (BC)	(Deng et al., 2019)
Halloysite	None; Heated: 120 °C	62.7; 68.1 (BC)	(Deng et al., 2017)
Kaolinite	Heated: 120 °C	56.7 (BC)	(Deng et al., 2017)
Smectite, illite, chlorite, kaolinite	None	25.5; 17.5; 10.2; 6.7 (DA)	(Lu et al., 2020)
Activated Carbon	None	12	(Shin et al., 2002)
Sepiolite	None; OA; HAD; HDTMA	6.8; 45.8; 104.2; 141.1 (F)	(Varela et al., 2021)
Bentonite	HDTMA; TMPA	0.10; 1.7 (F)	(Onwuka et al., 2020)

F (Freundlich), L (Langmuir), BC (Breakthrough curves), and DA (Dubinin–Astakhov). TMA+ (tetramethylammonium), OA (octylammonium), HAD (hexadecylammonium), HDTMA (hexadecyltrimethylammonium), and TMPA (trimethylphenylammonium). Sorbents with no modifications or amendments show decreased adsorption capacity compared to processed sorbents.

Other than carbon, natural clays modified with organic molecules or processed to increase the lipophilicity, surface area, and binding sites on the clay have also been well-studied for the sorption of benzene. However, the safety and ability of these processed and/or modified clays to tightly bind and detoxify benzene have not been reported. For example, the application of alkylammonium salts has been shown to increase the sorption capacity of clays for benzene (Deng et al., 2020; Sharmasarkar et al., 2000). However, these chemical modifications are made from compounds with known hazards, including eye and skin sensitization, corrosive activity (Sigma-Aldrich, 2021), and toxicity to aquatic species (Brycki et al., 2018; Garcia et al., 2016), leaving the safety of these sorbent strategies unknown. To solve this issue, we used chlorophyll (a) and (b), which are lipophilic and naturally-occurring GRAS agents that are “generally recognized as safe” for human and animal consumption and possess a high affinity for benzene. Chlorophyll (a) and (b) are magnesium-containing tetrapyrrole compounds that play a key role in photosynthesis (Chen, 2014). Chlorophylls have been shown to be effective sorbents for aflatoxin B₁ (Jubert et al., 2009), a potent hepatocarcinogen commonly detected in food and food crops (Phillips et al., 2019; Wang et al., 2017). The addition of dietary chlorophylls showed a reduction in the bioavailability of aflatoxin B₁ in vitro (de Jesús Nava-Ramírez et al., 2021), in animal models (Simonich et al., 2007, 2008), and in human volunteers (Jubert et al., 2009). However, free chlorophyll molecules have a short half-life of 90 min (Bennett, 1981), limiting their application for environmental remediation; but the immobilization and subsequent photo- and thermo-stabilization of chlorophylls onto smectite clays (Ishii et al., 1995; Itoh et al., 1998; Kodera et al., 1992) and silica (Itoh et al., 2002a; b) have been previously reported.

Montmorillonite is a smectite clay with a high surface area, porous structure, and high cation exchange capacity for its sodium and calcium cations, making it a useful tool for developing a targeted sorbent for benzene. Previous research by our laboratory has demonstrated: 1) the safety of calcium montmorillonite (CM) in humans and animals following administration in capsules, food, and flavored water (Phillips et al., 2019; Wang et al., 2020a), 2) the ability of montmorillonite clays to bind environmental contaminants and pollutants (Hearon et al., 2022, 2020; Wang et al., 2021a, 2020a, 2017; Wang et al., 2019a) in soil and water, and 3) the application of CM as a base material for GRAS amendments that can significantly enhance and broaden its binding capacity (Orr et al., 2020; Wang et al., 2021c, 2019c, 2019d; Wang et al., 2021b; Wang and Phillips, 2019e, 2019b). Therefore, in this study, chlorophyll molecules were amended to calcium and sodium montmorillonites to: 1) increase the lipophilicity of the clay, 2) stabilize the chlorophyll from light, and 3) enhance benzene interactions at active chlorophyll sites on clay surfaces.

In this study, we have characterized and optimized the binding of benzene to chlorophyll-amended calcium montmorillonite (CM-CH) and chlorophyll-amended sodium montmorillonite (SM-CH) clays using in vitro adsorption and desorption isotherms and thermodynamic analysis, in silico molecular simulations to characterize binding modes and mechanisms, and two ecotoxicological bioassays (L. minor and C. elegans) to indicate benzene toxicity and validate the safety and efficacy of the sorbent treatments.

2. Materials and methods

2.1. Reagents and materials

High-performance liquid chromatography (HPLC) grade acetonitrile and acetone were purchased from Fisher Scientific (Waltham, MA). Chlorophyll (mixture of types a and b) was purchased from TCI America (Portland, OR). Benzene was purchased from Sigma-Aldrich (St. Louis, MO). CM clay was obtained from TxESI Inc. (Bastrop, TX), which has a cation exchange capacity of 97 cmol/kg, total surface area of 850 m²/g, and an external surface area of approximately 70 m²/g (Grant and Phillips, 1998; Wang et al., 2021a). Sodium montmorillonite (SM) clay was obtained from Halliburton (Houston, TX), with a cation exchange capacity of 75 cmol/kg (Wang et al., 2020a). Medical grade powdered activated carbon with a purity > 99% was obtained from General Carbon Corporation (Paterson, NJ). It is labeled as derived from coconut shell with a surface area of 1100 m²/g surface area and 5% moisture (Generalcarbon.com).

2.2. Synthesis of sorbents

CM and SM clays were collapsed to reduce interlayer spacing by heating samples at 200 °C for 30 min followed by heating at 800 °C for 2 h (Grant and Phillips, 1998). Chlorophyll amended clays (CM-CH and SM-CH) and collapsed CM-CH and SM-CH clays were synthesized using previously described methods (Wang et al., 2017). Briefly, CM or SM suspensions (5% w/w in pH 4.2 water) were mixed vigorously with chlorophyll at 150% cation exchange capacity for 24 h to ensure that the exchangeable sites were oversaturated with chlorophyll. The mixture was centrifuged and washed thoroughly three times with distilled water. Samples were dried in a desiccator, ground, and sieved through a 100 mesh to obtain uniform particle sizes (≤ 149 μm). Hydrophobicity of all sorbents was assessed using established methods (dos Reis et al., 2016). Approximately 300 mg of each sorbent was dried at 100 °C for 24 h and added to 10 mL beakers. Samples were then transferred to a capped jar containing 60 mL of water or n-heptane for 24 h at ambient conditions. Afterwards, the samples were removed from the jar and weighed to determine the mass uptake of n-heptane and water vapor.

The photostability of free chlorophyll and CM-CH in various pH conditions was investigated following established methods (Itoh et al., 2002b; Kodera et al., 1992). 4% (w/v) solutions of chlorophyll and CM-CH clay in 4 mL of pH 6.5 and 7.5 water were placed under a 60-watt bulb with an 8 cm distance and stirred constantly. Samples were protected from heating by placing ice and cooled water underneath. CM-CH and chlorophyll solutions were left under the light for 96 h; at 0, 0.5, 1, 2, 4, 6, 24, 48, 72, and 96 h, aliquots were taken from the sample for absorbance measurements using a UV/Visible scanning spectrophotometer (Shimadzu UV-1800, Kyoto, Japan) with scans ranging from 500 to 700 nm. Measurements were analyzed for the shift of maximum adsorption wavelength and change in the magnitude of the absorbance peak.

2.3. Benzene analysis

Benzene concentrations were determined using a HPLC method (Bahrami et al., 2011) on a Waters device equipped with a 1525 binary mixing pump, a 717 plus autosampler, a XBridge C18 (4.6 mm × 50 mm, 3.5 μm) column, and 2487 UV/Visible detector (Waters® Corporation, Milford, MA). The Breeze® software was used to control the system and collect data. Chemical separation was achieved with a mobile phase of 66% acetonitrile and 34% water at 1.0 mL/min flow rate with an injection volume of 25 μL. The retention time for benzene was 1.2 min. A linear standard curve of benzene was derived using a blank solution and a range of 9 benzene concentrations from 0.5 to 30 ppm (r² = 0.999). The limit of detection was 500 ppb.

2.4. Adsorption and desorption isotherms

A 1000 ppm stock solution of benzene was created by diluting pure benzene in 10% ethanol. Sorbents, including CM-CH and SM-CH, parent clays, and activated carbon, were added at a 5 mg/L inclusion rate to solutions with an increasing concentration gradient of benzene ranging from 0 to 30 mg/L. These gradients were achieved by adding a calculated amount of the stock solution and sorbent to a 1.5 mL microcentrifuge tube with 10% acetonitrile solvent to achieve a total volume of 1 mL. Acetonitrile was used in this analysis to overcome the low water solubility of benzene (1.79 g/L) (PubChem, 2022) and to determine the theoretical maximum capacity of the sorbent for benzene. Previous studies have indicated that acetonitrile does not interfere with the interlayer structure of the clays (Wang et al., 2019b). Control samples comprised of solvent only, solvent and sorbent, and a benzene reference solution were prepared fresh for each experiment and analyzed together with test groups. All samples were agitated for 2 h on an orbital shaker set to 1000 rpm (VIBRAX VXR basic, Werke, Germany) at ambient temperature (24 °C). Isothermal analyses were conducted at ambient temperature and pH to assess the performance of the sorbents in environmentally-relevant conditions. For thermodynamic studies, all solutions were mixed at three different temperatures (4 °C, 24 °C, and 37 °C). Samples were centrifuged for 20 min at 2000 g to separate the sorbent bound to benzene from the supernatant. The supernatant was collected and analyzed by HPLC to determine the concentrations of residual benzene.

For desorption, the sorbent-benzene complex from adsorption studies were separated from the supernatant, rinsed with deionized water once, and filled to 1 mL with 10% acetonitrile solution. The tubes were placed on the shaker at 500 rpm for 24 h. Suspensions were centrifuged at 2000 g for 20 min, and the supernatant was analyzed by HPLC. Traces of desorbed benzene were detected at the end of the experiment, and the concentration of remaining bound benzene was determined by the difference between the initial adsorbed amount and the desorbed amount. Standard benzene controls were tested before and after the adsorption and desorption studies to exclude possible loss of benzene by evaporation.

2.5. Data calculations and curve fitting

For adsorption isotherms, the amount of benzene bound to clay was determined from the difference of concentrations between control and test solutions and shown in mmol/kg of sorbent. R code and TableCurve 2D (Systat Software, Inc) were utilized to plot the adsorption data and calculate fitted and upper/lower bound adsorption estimates (Gong, 2020). The adsorption isotherms were plotted by the Freundlich model system using the average values from triplicate analyses. The Freundlich model describes the multilayer adsorption of sorbate onto a heterogeneous surface of the sorbent with theoretically infinite, unsaturable binding sites (Desta, 2013; Reed and Matsumoto, 1993). The Freundlich equation was entered as a user-defined function in TableCurve 2D:

$$\text{Freundlich model}\hspace{1em}q={\text{K}}_{\text{f}}{\text{C}}_{\text{w}}^{\left(1/\text{n}\right)}$$ (1)

Where q is the amount of benzene adsorbed (mmol/kg), K_f is the Freundlich distribution constant for adsorption affinity, C_w is the equilibrium concentration of benzene (mmol/L), and 1/n is the degree of linearity between the solution concentration of benzene and adsorption. n = 1 corresponds to linear adsorption, and n > 1 suggests physisorption (Desta, 2013).

Adsorption parameters were used with the Gibbs free energy and van’t Hoff equations to calculate enthalpy (ΔH) and Gibbs free energy (ΔG):

$$\Delta H=\frac{-\text{R}\text{ln}\left(\frac{K_{d1}}{K_{d2}}\right)}{\left(\frac{1}{T_2}\right)-\left(\frac{1}{T_1}\right)}$$ (2)

$$\Delta \text{G}=\Delta {\text{G}}^{\circ }+{\text{RTlnKe}}^{\circ }$$ (3)

R is the gas constant: 8.314 J/mol/K, T is the absolute temperature, calculated from 273 + t (°C). Adsorption systems that are thermodynamically favored has a negative ΔG°. If ΔG° has a positive value, the adsorption process is not favored, and the adsorption is not significant (Hosseini-Bandegharaei, 2019).

2.6. Molecular simulations

The model structures for benzene and chlorophyll (a) were extracted from PubChem (PubChem, 2022) and the protein data bank (PDB) (Berman et al., 2000; PDB, 2022), respectively. The model structure for montmorillonite clay was created using CHARMM GUI (Choi et al., 2021; Jo et al., 2008) with Miller indices 001, 0.33333 for the ratio of defect. This was based on our previous work (Wang et al., 2019a), aiming to model a CM-CH clay with a composition of (Si₄)^I-V(Al_1.67Mg_0.33)^VIO₁₀(OH)₂. CM clay was modeled based on SM clay, which was provided as one of the default options by CHARMM-GUI, by exchanging every two Na⁺ ions with one Ca²⁺ion. The montmorillonite clay dimensions in the x-y plane were equal to 50×50 Ų, and the two layers were separated to form a d₀₀₁ spacing of 21 Å in the z-direction (Greenland and Quirk, 1962; Orr et al., 2020; Wang et al., 2017, 2019a, 2021b). Upon modeling, 20 Ca²⁺ ions remained on each of the two layers. During the initial setup of the montmorillonite clay simulations using the INTERFACE FF (Heinz et al., 2013) force field, hydrogens from SiOH groups and aluminate edges were manually removed from the original setup provided by CHARMM-GUI to mimic the system at neutral pH, which was the experimental setup.

Topology and parameters for benzene and CM were obtained using CGENFF (Vanommeslaeghe et al., 2010) and INTERFACE-FF, respectively. These were provided through CHARMM-GUI default input files (Choi et al., 2021; Jo et al., 2008). Topology and parameters for chlorophyll (a) were initially obtained through a combination of CGENFF parameters (Vanommeslaeghe et al., 2010) and published charges for the chlorin ring (Guerra et al., 2015) while maintaining a zero net charge on the molecule. The magnesium atom was introduced and constrained using MMFP harmonic constraints with respect to all four neighboring nitrogen atoms to a distance of 2.107 Å, representing the average distance in the model structure available in the PDB (Berman et al., 2000; PDB, 2022).

Two layers of CM were centered in a cubic (89.25 Å) periodic boundary conditions box and solvated by explicit water molecules in CHARMM (Brooks et al., 2009). We modeled and simulated systems comprised of 24 benzene molecules in complex with parent CM and CM-CH. The CM-CH clay was built based on a short 2 ns independent simulation of 12 chlorophyll molecules in the presence of CM. Four systems of each CM and CM-CH were simulated independently for 35 ns at constant temperature (300 K) and pressure (1 atm) in CHARMM (Brooks et al., 2009). Before running the production, short equilibration runs at constant volume were performed at 200 ps each, during which Ca²⁺ ions were allowed to move away from clay layers where they were initially introduced by CHARMM-GUI (Choi et al., 2021; Jo et al., 2008). The simulation setup was primarily based on CHARMM-GUI (Choi et al., 2021; Jo et al., 2008) and was optimized for the current systems. During the simulations, harmonic constraints were introduced to aluminum and magnesium atoms of the clay with an input force constant of 1 kcal/mol Ų. The large size of the cube allowed the exploration of several binding modes of the molecules within the interlayer and on the exterior surface of the clay.

Upon completion of the quadruplicate simulations, we developed and utilized in-house FORTRAN programs to investigate interactions of benzene molecules with CM and CM-CH clays. In CM-CH, we investigated interactions between benzene and chlorophyll molecules, interactions between chlorophyll molecules and CM clay, and interactions between adjacent chlorophyll molecules due to their observed tendency to form aggregates (Bystrova et al., 1976; Shi et al., 2013; Vladkova, 2000). Chlorophyll molecules were separated into “core” and “tail” sections, where the core was comprised of the chlorin ring and two subsequent carbon groups, and the tail section was composed of the remaining aliphatic chain. In all analyses, two entities were considered to interact when the distance between any pair of their atoms (including hydrogens) was ≤ 3.5 Å. Snapshots of simulations were obtained and visualized using VMD (Humphrey et al., 1996).

2.7. L. minor assay

Lemna minor was purchased from AquaHabit (Chatham, England). The plant population was cultured under cool, white fluorescent lights with 400 ft-c intensity with a 16/8-hour light to dark cycle at 25 °C. Plants were maintained in Steinburg medium, prepared as previously described (Drost et al., 2007). For dosimetry analysis, two individual plants (7–8 fronds total) were exposed to increasing concentrations of benzene (0.1 to 10 ppm) in 12 well-plates for 7 days under a growth light (50 watt, 16/8 hour light-dark cycle) and in a fume hood. Plants were observed daily for changes in frond number and surface area. The surface area of plant groups was measured using ImageJ (NIH, Bethesda, MD). After 7 days of exposure, surviving plants were removed from the dishes and homogenized by a Homogenizer 150 (Fisher Scientific) in 1.5 mL 80% acetonitrile (Drost et al., 2007) and placed in the dark at 4 °C for 48 h to extract chlorophyll. Samples were measured with UV/Visible scanning spectrophotometry (Shimadzu UV-1800, Kyoto, Japan) at 663 nm to quantify total chlorophyll content. For the detoxification study, a 10 ppm benzene solution (prepared in the Steinburg medium) was mixed with 1% or 2% sorbents for 2 h at 1000 rpm, centrifuged, and 4 mL of the supernatant was exposed to plants in glass petri dishes with fitted lids in a fume hood under a growth light for 5 days. Plants were exposed in petri dishes to prevent evaporated benzene from exposure groups contaminating the control and other exposure groups. Plants were exposed to benzene for 5 days because the dosimetry studies indicated noticeable adverse effects on the plants within 5 days. Frond number, surface area, and chlorophyll content were quantified as previously described. Growth rate and inhibition percentage were calculated based on standard OECD guidelines (OECD, 2006).

$$\text{Growth}\text{rate}=\frac{\text{Log10}\left(\text{final}\text{frond}\text{no}.\right)-\text{Log10}\left(\text{initial}\text{frond}\text{no}.\right)}{\text{days}}$$ (4)

$$\text{Inhibition}\text{of}\text{growth}\%=100\times \left(1-\frac{\text{frond}\text{no}.\text{in}\text{the}\text{treatment}}{\text{fond}\text{no}.\text{in}\text{the}\text{control}}\right)$$ (5)

2.8. C. elegans assay

Wildtype (Bristol N2) C. elegans and E. coli strains NA22 and OP50–1 were purchased from the Caenorhabditis Genetics Center (CGC, University of Minnesota). Nematodes were grown on 8P media seeded with 8 × 10⁸ cells/mL of E. coli NA22 and were maintained at 18 °C (Wang et al., 2021d). Age synchronized L1 nematodes were obtained by a bleaching and washing process. Briefly, adult populations were washed with a bleaching solution to yield egg cultures; egg cultures were washed with M9 solution. Egg culture solutions were rocked on a rocking platform (VWR, Radnor, PA) at 2.5 rpm in the dark for 18 h at 18 °C.

After the incubation period, the size of the population was estimated by counting the number of nematodes in 3 droplets (2 μL aliquots) of the worm suspension, and the volume containing 5000 nematodes was estimated. For dosimetry studies, 5000 nematodes were transferred to 1.5 mL microcentrifuge tubes with 15 μL of E. coli OP50–1 and increasing concentrations of benzene (10 to 200 ppm) in K-medium complete solution for 24 and 48 h (Boyd et al., 2012). This model was used to simulate a flood scenario when benzene levels can be enhanced and redistributed into flooded water (Xiao et al., 2018). The total volume of nematodes, K-medium, E. coli, and benzene was 1 mL for all experiments. For the detoxification study, a 50 ppm benzene solution (prepared in K-medium complete solution) was mixed with 1% and 2% sorbents at 1000 rpm for 2 h, centrifuged at 2000 g for 20 min, and the supernatants were exposed to the nematodes for 24 and 48 h. For both the dosimetry and detoxification studies, 3 droplets of 10 μL aliquot from each group were used to measure the survival rate. Test samples were observed under an Olympus SZ61 zoom stereomicroscope (Olympus, Waltham, MA) to assess the percentage of live nematodes in each group. Afterwards, C. elegans were washed three times with K medium complete solution, transferred to a nematode growth media and seeded with E. coli OP50–1. Nematodes were left to mature at 18 °C for 48 h before body length quantification. Body length of individual nematodes was measured using the CellSens Entry (standard version 3) software (Olympus, Waltham MA) after nematodes were paralyzed with 25 mM sodium azide (Rangsinth et al., 2019). Relative body lengths were calculated as a percentage of the medium control group, which was adjusted to 100%.

2.9. Statistical analysis

A 2-way t-test was used to determine statistical significance. Blank and benzene controls were included in each experiment. Each experiment was conducted in triplicate. Frond number, surface area, and total chlorophyll content from L. minor and body length data from C. elegans were calculated for averages, standard deviation, and p-value. Results were considered significant at p ≤ 0.05.

3. Results and discussion

3.1. Hydrophobicity and stability

CM and SM are hydrophilic smectite clays, but the addition of organic compounds to the surface has been shown to increase the hydrophobicity of the clay surface (Jaynes and Vance, 1999; Wang et al., 2021a) and enhance the sorbent’s attraction for hydrophobic compounds (Moyo et al., 2014). Chlorophylls are lipophilic (LogP = 11.3 for chlorophyll a) organic compounds, and their intercalation onto intact and collapsed CM and SM clays increased the hydrophobicity of the clay surfaces (Fig. 1A). Specifically, CM-CH and SM-CH had a 62.8% and 89.7% increase in the ratios of absorbed n-heptane to water vapor, respectively. Collapsed CM-CH and SM-CH clays had slightly lower hydrophobicity values compared to the uncollapsed CM-CH and SM-CH, which indicated that some chlorophyll can enter the interlayer space, but basal and external surfaces of the clay were the primary sites for the chlorophyll amendment (95.3% for CM-CH and 83.9% for SM-CH). The increased surface hydrophobicity of CM-CH and SM-CH compared to parent CM and SM clays may contribute to an increased ability to bind hydrophobic compounds, such as benzene (LogP = 2.13).

Fig. 1.

Increased hydrophobicity, measured by the absorption ratio of n-heptane/water, for CM-CH, SM-CH and collapsed derivatives compared to parent CM and SM (A). Absorbance of chlorophyll and CM-CH suspensions in pH 6.5 and 7.5 water after 96 h of light exposure (B). The absorbance reading at the starting point was adjusted to 1. The absorbance of chlorophyll decreased with time at both pH conditions and light exposures, while chlorophyll in the CM-CH clay resisted photodegradation.

Chlorophyll is known to be sensitive to degradation by light and acidic conditions (Chen, 2014), but its conjugation with surfaces of clay (Itoh et al., 1998; Kodera et al., 1992) or silica (Itoh et al., 2002b) materials can increase the photostability of the compound. To test the stability of CM-CH in light and at various pHs, 4% (w/v) suspensions of chlorophyll alone and CM-CH in water (pH 6.5 and 7.5) were subjected to irradiation by light for 96 h and monitored for changes in color and absorbance patterns. These pHs were chosen based on previous studies of chlorophyll degradation in the literature (Koca et al., 2007). As shown in Fig. 1B, the absorbance of the chlorophyll suspensions decreased by 80% and 25% by the end of 96 h of light exposure at pH 6.5 and pH 7.5, respectively. These findings are consistent with previous literature suggesting that the degradation of chlorophyll was more extensive in acidic conditions (Chandra et al., 2021). The CM-CH suspensions at both pH conditions maintained constant absorbance (> 97%) throughout the experiment. Additionally, the characteristic green pigment in chlorophyll suspensions was completely bleached by 72 h of irradiation; meanwhile, CM-CH suspensions retained the dark green color like the starting material, demonstrating the stability of chlorophyll in the CM-CH sorbent. The enhanced stability of chlorophyll molecules when associated with montmorillonite could be due to the spectral shifts of absorbance (Kodera et al., 1992; Micó-Vicent et al., 2021) from promotion of oligomerization of photopigments or the conjugation with the clay surface through chemical processes (Mokaya et al., 1994; Wu and Li, 2009). This result supports the potential application of chlorophyll-amended clays as stable adsorbents for environmental remediation.

3.2. Isothermal analysis and thermodynamics

Activated carbon has traditionally been implemented in the filtration of benzene pollution, but our experiments showed limited benzene binding to activated coconut carbon surfaces with low affinity (K_f = 0.362) based on the Freundlich model (mean squared error, MSE = 1.31 × 10^–6) (Table 2). This result supports previous work that demonstrated variable binding activity of carbonaceous materials, depending on their sources and types of characterization (Lillo-Ródenas et al., 2011; Oh et al., 2019; Vikrant et al., 2019). Unlike clays that are active when hydrated, carbons have been shown to perform poorly in humid conditions, leading to a decreased adsorption capacity for a wide range of environmental contaminants (Veksha et al., 2009; Xian et al., 2015). Isothermal data for the adsorption of benzene to parent SM and SM-CH (Fig. 2A) and CM and CM-CH (Fig. 2B) display good fit by the Freundlich model for both SM-CH and CM-CH (MSE < 1.70 × 10⁻⁵), and parent clays (MSE < 9.9 × 10⁻⁶). This suggested a multilayer and heterogeneous adsorption of benzene with theoretically infinite binding sites onto the clay surfaces. Importantly, SM-CH and CM-CH showed an increased binding affinity and capacity for benzene, compared to parent SM and CM and activated carbon. While parent clays were shown to bind benzene at 11–19%, there was complete desorption of benzene within 24 h, showing the limitation of parent montmorillonite clays in retaining benzene adsorption on their surfaces. Whereas, SM-CH and CM-CH showed desorption rates less than 30%, indicating an enhanced binding interaction between benzene and the chlorophyll-amended clays (Fig. 3). The increased binding affinity and percentage of benzene sorbed onto chlorophyll-amended clays is likely due to the increased lipophilicity, surface area, and binding sites from the intercalation of chlorophyll into the clay layers and onto the basal surfaces. CM-CH showed a higher adsorption capacity for benzene compared to SM-CH, possibly due to the higher cation exchange capacity of CM clays that allowed for more chlorophyll to be added to clay surfaces, increasing available binding sites for benzene. Adsorption enthalpy (ΔH) and Gibbs free energy (ΔG) were derived for parent and chlorophyll-amended clay interactions with benzene at three different temperatures (4 °C, 24 °C, and 37 °C). Derived thermodynamic values from isothermal analysis of CM-CH (ΔG = −16.06 kJ/mol and ΔH = −24.91 kJ/mol) indicated a spontaneous and favorable physisorption process, which is consistent with the hypothesized contribution of Van der Waals forces for the binding mechanism. Physisorption occurs when sorbate molecules are held to the sorbent surfaces by physical forces; these attractive forces include Van der Waals attractions, hydrogen bonding, and dipole-dipole interactions (Gu et al., 1994). Table 2 summarizes the binding parameters, desorption, and thermodynamic information for activated carbon, CM, SM, CM-CH, and SM-CH clays.

Table 2.

Parameters of Freundlich model of adsorption isotherms.

Sorbent	Adsorption	Desorption	Kf	n	ΔH (kJ/mol)	ΔG (kJ/mol)	MSE
Activated carbon	5%	100%	3.62 × 10−1	1.61	–	13.7	1.31 × 10−6
CM	10%	100%	4.78	1.23	–	− 3.9	1.24 × 10−5
SM	19%	100%	4.68 × 10−2	1.81	–	7.6	9.94 × 10−6
CM-CH	52%	28%	665.60	4.37	−24.9	−16.1	1.70 × 10−5
SM-CH	44%	25%	74.75	1.47	−24.7	−10.7	1.62 × 10−5

K_f (adsorption capacity), n (linearity parameter), ΔH (enthalpy, kJ/mol), ΔG (Gibbs free energy, kJ/mol), and MSE (mean squared error).

Fig. 2.

Plots of adsorption isotherms of benzene with parent SM and SM-CH (A) and CM and CM-CH (B) clays at 25 °C for 2 h. Isotherm plots in the Freundlich model were depicted by an average of adsorption (mmol/kg) (solid shapes), line of best fit (solid line), and upper and lower 95% confidence bands of adsorption (dotted lines). SM-CH and CM-CH showed a short reaction time to reach adsorption equilibrium and an increased adsorption affinity (K_f = 74.75 and 665.60, respectively) compared to parent clays (K_f = 4.68 × 10⁻² and 4.78 for SM and CM, respectively).

Fig. 3.

Plots of benzene adsorption remaining on SM and SM-CH (A) and CM and CM-CH (B) clays after desorption at 24 °C for 24 h, plotted by lines of best fit (solid) and upper and lower 95% confidence intervals (dotted). SM-CH and CM-CH showed a decreased desorption percentage (25–28%) compared to unamended parent clays (100%), indicating an increased ability to retain benzene for extended time periods.

3.3. Molecular simulations

The comparison between the predicted benzene binding propensities for CM and CM-CH is shown in Fig. 4A, indicating the increased benzene binding to CM-CH clay through several mechanisms. The values indicated a relatively low (≈ 20%) propensity of benzene molecules binding to CM or CM-CH through interactions with the parent clay alone (Fig. 4A, red; Fig. 4B). The propensity of benzene molecules to interact directly with clay surfaces mediated by the presence of at least one bound chlorophyll in CM-CH (Fig. 4A, dark blue; Fig. 4C) suggested a minimal contribution of this mechanism to the overall binding percentage. The major binding propensity (≈ 40%) was through direct interactions between one or more bound chlorophyll and benzene molecules. (Fig. 4A, light blue; Fig. 4D–E).

Fig. 4.

Comparison of benzene binding propensity (%) between CM and CM-CH clays (A). Red: binding to clay alone; dark blue: binding to clay through chlorophyll; light blue: binding to single or aggregated anchored chlorophyll. Corresponding examples are shown in cartoons, labeled and colored in boxes (B-E). Cartoons show examples of interactions and do not cover all possibilities of interactions within each category. Brown: clay, denoted as “CM”; black: benzene; green: chlorophyll, denoted as “CH”. Detailed descriptions of the interactions depicted in each case (B-E) is provided in the text.

Our investigation aimed to additionally obtain insights into the benzene-chlorophyll interactions by dividing chlorophyll molecules into core and tail sections. Within the simulations, benzene molecules primarily interacted with the aliphatic tail (45.5 ± 2.5%), where the tail was directly bound to the clay or was part of an aggregate of bound chlorophyll molecules. Benzene was also observed to interact with the core section only or simultaneously with the core and tail at 27.7 ± 3.2% and 26.8 ± 3.7%, respectively. It is interesting to note that chlorophyll molecules were mostly bound to the clay by both core and tail (76.1 ± 2.0%), and less likely bound to the clay through core only (19.1 ± 3.4%); the binding through the aliphatic tail only was negligible (4.8 ± 2.3%).

The enhanced binding of benzene onto CM-CH was mostly due to their interaction with 1 or more chlorophyll molecules found in large aggregates, which primarily occurred on the external surfaces of the clay. A summary of the observed clay-bound chlorophyll aggregates and the corresponding number of interacting benzene molecules in 4 independent simulations is provided in Table 3. Aggregates of chlorophylls, comprised of ≥ 2 molecules where at least one is bound to the clay, can be observed interacting with multiple benzene molecules simultaneously. For instance, at 20 ns of the first simulation, an aggregate of 8 chlorophyll molecules at the upper basal surface interacted with 10 benzene molecules, an aggregate of 2 chlorophyll molecules at the lower basal surface interacted with 2 benzene molecules, and 2 individual chlorophyll molecules interacted with 2 benzene molecules in the interlayer space (Fig. 5A). Similar aggregates of 4–10 chlorophyll molecules were formed and interacted with benzene molecules in the subsequent simulations. The largest cluster was observed in the second simulation, composed of an aggregate of 10 chlorophyll molecules that interacted with 7 benzene molecules. Fig. 5B presents an example of a dimer of chlorophyll, demonstrating the three diverse binding possibilities between the clay, chlorophyll, and benzene molecules. These snapshots highlight the importance of both the hydrophobic tail and the core to the binding mechanism and the ability of chlorophyll molecules (either individually or as part of an aggregate) to interact with several benzene molecules.

Table 3.

Information on clay-bound chlorophyll aggregates at 35 ns of simulation time and benzene molecules within these aggregates.

Simulation	Number of Aggregates	Number of Chlorophylls (Number of Benzene molecules)
1	2	8(12), 2(1)
2	1	10(7)
3	3	5(9), 4(7), 2(0)
4	4	4(2), 2(5), 2(3), 2(0)

The table lists the total number of chlorophyll aggregates (second column) bound to the clay that were encountered at the end of each simulation trajectory (first column). A clay-bound aggregate is comprised of at least 2 chlorophyll molecules where at least one is bound to the clay. The number of chlorophylls in each aggregate with the corresponding number of benzene molecules interacting with chlorophyll(s) in each aggregate are provided (third column); the number of benzene molecules is provided in parenthesis.

Fig. 5.

Snapshots of simulations showing diverse interactions between benzene and anchored chlorophyll molecules (A). Several interactions between benzene molecules and single or aggregated chlorophyll molecules. Clay is shown in the vdW (van der Waals) representation, and chlorophyll and benzene molecules are shown in green and black, respectively. Snapshots of simulations showing a bound chlorophyll dimer interacting with three benzene molecules (B). From left to right, benzene interacts with hydrophobic tail (pink), tail and core simultanously (purple), and core (blue) of chlorophyll.

3.4. L. minor assay

Lemna minor is a floating, aquatic plant that is commonly utilized as an ecotoxicological bioassay for toxicity testing due to its simplicity, short generation time, and sensitivity to a wide variety of chemicals (Aliferis et al., 2009; EPA, 2009). As it is a floating species, it may be sensitive to surface-restricted or light non-aqueous phase liquids that do not readily dissolve in water, such as benzene and other oil products (Taraldsen and Norberg-King, 1990). In this study, plants were exposed to increasing concentrations of benzene (0.1 to 10 ppm) for 7 days and then assessed for changes in overall surface area, frond number, chlorophyll content, and growth metrics (Fig. 6). Plants exposed to 1–10 ppm benzene all had a significantly decreased surface area (Fig. 6A) and frond number (Fig. 6B) compared to control plants (p ≤ 0.05). While 1 ppm benzene plants had a normal number of fronds, the overall size and surface area of these leaflets were smaller than those of control plants. Exposure to 1–10 ppm of benzene caused a reduction of chlorophyll content in the plants, which corresponded to an observable discoloration of plants, especially with exposure to higher than 2.5 ppm benzene. The overall change in frond number was used to derive the growth rates and growth inhibition percentages. High benzene exposures (for example, 5 ppm) caused a decline in growth rates (1.05) and increased inhibition percentages (33%), compared to the blank medium control (Fig. 6D). All assessed response endpoints showed high correlation and indicated that L. minor is sensitive to benzene with significant adverse effects from exposure to as low as 1 ppm benzene, making it a useful model for assessing benzene toxicity.

Fig. 6.

Dose-dependent toxicity of benzene exposure on L. minor surface area (A), frond number (B), chlorophyll content (C), inhibition percentage (D, bar graph), and growth rate (D, line). Data is depicted as the mean response ± standard deviation. * = p ≤ 0.05; ** = p ≤ 0.01, compared to blank medium control.

In the detoxification study, exposure to 10 ppm of benzene was selected as the toxin control since it significantly reduced the surface area, frond number, chlorophyll content, and growth rates of the plants. For surface area, the 2% CM-CH inclusion in the medium showed the greatest protection efficacy, followed by 1% CM-CH, 2% SM-CH, and 1% SM-CH inclusions (Fig. 7A). For frond number, the 2% CM-CH also displayed the highest protection efficacy, followed by the 2% and 1% SM-CH and the 1% CM-CH (Fig. 7B). All sorbents had similar performance on the protection of chlorophyll content against 10 ppm benzene (Fig. 7C). While the 1% CM-CH and SM-CH showed higher growth rates and lower inhibition percentages than the 10 ppm benzene group, the 2% CM-CH group showed the highest growth rate (0.69) and lowest inhibition percentage (−6%) (Fig. 7D).

Fig. 7.

Protection of 1% or 2% CM-CH and SM-CH against benzene toxicity on plant surface area (A), frond number (B), chlorophyll content (C), inhibition percentage (D, bar graph), and growth rate (D, line) on day 7 of exposure. Data is depicted as the mean response ± standard deviation. * = p ≤ 0.05; ** = p ≤ 0.01, compared to blank medium control.

The inclusion of all sorbents effectively protected the plants from benzene toxicity in all outcomes, but the 2% CM-CH clay showed the greatest protection for exposed plants. This finding is supported by higher surface area, frond number, and growth rate than benzene control, other clay treatments, and even the blank medium control. The inclusion of these clays alone showed no adverse alterations in morphology or growth of the plants; instead, the inclusion of the sorbents promoted the growth of the plants, likely because L. minor is a photosynthetic plant that is dependent on chlorophyll molecules for growth, and the potentially detached chlorophyll molecules from the chlorophyll-amended clays may provide added benefits to plant health and growth (Gou et al., 2020).

3.5. C. elegans assay

C. elegans is a microscopic soil nematode that is commonly used in toxicity studies due to its ease of cultivation, short generation time, and sensitivity to a wide variety of environmental contaminants (Boyd et al., 2012; Hunt, 2017). A previous study reported that nematodes are sensitive to benzene at low exposure levels, causing adverse effects on fertility and locomotion (Eom et al., 2014). Preliminary studies with 1–2% CM-CH and SM-CH alone in culture medium showed no adverse effect on the nematodes, which supports the safety of these clays. In this study, nematodes were first exposed to increasing concentrations of benzene (50 – 200 ppm) for 24 and 48 h and assessed for changes in body length and overall survival rate each day. As shown in Fig. 8A, nematodes exposed to benzene showed an overall decrease in body length on both days. Specifically, at 24 h, the benzene exposure groups showed similarly reduced body lengths (≈ 80% control, p ≤ 0.001); while by 48 h, the change in body length displayed a clear dose-dependent relationship. Additionally, the survival rate after 48-hour exposure of benzene decreased in all groups, with the 50 ppm benzene group displaying 86% survival (Fig. 8B). Therefore, the 50 ppm benzene group was utilized in the detoxification studies.

Fig. 8.

Dose- and time-dependent toxicity on C. elegans body length after 24 and 48 h of exposure (A) and survival rate after 48 h of exposure (B). Protection against benzene toxicity by SM-CH and CM-CH on body length after 24 and 48 h of exposure (C) and survival rate after 48 h of exposure (D). Data is depicted as the mean response ± standard deviation. * = p ≤ 0.05 ** = p ≤ 0.01, *** = p ≤ 0.001, compared to blank medium control.

In the detoxification studies, 50 ppm of benzene was shown repeatedly to be toxic to C. elegans, reducing their overall body length at 24 and 48-hour time points and reducing survival rates. Both CM-CH and SM-CH at 1% and 2% inclusion rates in the culture medium significantly protected the nematodes in terms of body length from benzene exposure for 24 and 48 h, with the 2% CM-CH sorbent performing the best (98% protection at 24 h, and 100% protection at 48 h) (Fig. 8C). Specifically, the body length in 2% CM-CH showed no significant difference from the blank medium control, showing high protection efficacy of CM-CH. This result is consistent with the earlier L. minor assay where CM-CH at high inclusion (2%) was the most effective sorbent for the detoxification of benzene. All sorbents also significantly increased the survival rates of nematodes exposed to benzene for 48 h and provided 98–100% protection (Fig. 8D). No sorbents increased the overall body length of nematodes more than the control, likely because chlorophyll has only been reported to enhance oxidative stress response and life span, not growth parameters in C. elegans (Duangjan et al., 2019; Wang and Wink, 2016). A low level of sorbent material (1–2%) was included in both bioassay studies, which delivered dose-dependent protection against benzene toxicity. Therefore, higher sorbent inclusion is expected to result in higher detoxification efficacy for benzene.

4. Conclusions

In the present study, we assessed the ability of novel chlorophyll-amended sodium and calcium montmorillonites to adsorb and detoxify benzene. These multicomponent clays were assessed for their binding capabilities using in vitro, in silico, and bioassay methods in living organisms. Our in vitro isothermal analyses suggested that CM-CH and SM-CH clays had notably increased adsorption percentage, affinities and free energy of sorption for benzene with decreased dissociation compared to parent clays and activated carbon. The thermodynamic results indicated chemosorption of benzene binding onto CM-CH and SM-CH clays. In silico methods indicated that chlorophyll increased the adsorption of benzene by promoting interactions between benzene and the clay surfaces and by binding benzene directly. Results from living organisms L. minor and C. elegans indicated that the sorption and detoxification of benzene by these amended clays was substantial, and that they significantly protected these organisms from benzene toxicity. Together, the in vitro and in silico results provided a unique insight into the binding mechanisms for benzene. The bioassay studies predicted the toxicity of benzene and validated the efficacy of CM-CH and SM-CH. Importantly, combined results indicated that novel chlorophyll-amended clays may be useful materials to remediate benzene polluted water, soil, and air following natural disasters and spills, thus reducing unintended exposures to humans, animals, and plants.

Further work is needed to assess the ability of chlorophyll-amended montmorillonites to decrease the availability of benzene in mixtures (BTEX). Additionally, air containing high levels of benzene is another important exposure pathway for humans and animals; investigations into the ability of the chlorophyll-amended montmorillonites to reduce benzene and mixtures of BTEX from contaminated air are ongoing in our laboratory.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Data Availability

Data will be made available on request.