Carbon-based nanomaterials mediated adsorption and photodegradation of typical organic contaminants in aqueous fulvic acid solution

Carbon-based nanomaterials (CNMs), such as graphene oxide (GO), single-wall carbon nanotube (SWCNT), and multi-wall carbon nanotube (MWCNT), have been widely applied in various industrial and agricultural products and even daily supplies due to their unique chemical and physical properties, including high surface area and adsorption capacity, superb electrical and thermal conductivity, high mechanical strength, cost-effectiveness of mass production, nanoscale size effects, etc. (Zhang et al. 2022; Pérez et al. 2023). For example, both CNTs and magnetic GO have been utilized as active components of adsorbents for removing the various inorganic and organic pollutants from industrial wastewater and landfill leachate (Lerman et al. 2013; Zhang et al. 2016; Li et al. 2020). Meanwhile, during the production, transportation, utilization, and disposal of the CNMs or CNMs-contained products, CNMs are inevitably released into the aquatic environment through domestic sewage, industrial wastewater, and even coastal recreation activities (Park et al. 2017; Sharma et al. 2020; Ding et al. 2022). Therefore, the ecological risk and environmental behavior of CNMs in the aquatic ecosystem have received increasing concern from the scientific community (Qiang et al. 2016; De Marchi et al. 2018; Ko et al. 2019). Previous studies on the biological impacts of CNMs on aquatic ecology have shown that CNMs could inhibit algal growth, cause mechanical damage and oxidative stress, and induce light shielding effects, but promote the production of bioactive substances and prevention of biological contaminants (Jackson et al. 2013; Wang et al. 2019; Huang et al. 2022). Other studies suggested that CNMs may cause incubation delay, cardiac edema, chorionic processes and changes in the protein structure of aquatic animals, as well as remarkably enhance the accumulation of organic contaminants such as perfluoro-octane sulfanilamide (PFOS) in their blood, guts, and muscles (Chen et al. 2015; Qiang et al. 2016). As for the environmental behavior of CNMs in aquatic environments, most previous investigations focused on their aggregation and adsorption of environmental contaminants including heavy metals and organic contaminants (Apul et al. 2013; Jiang et al. 2017a; Ighalo et al. 2022; Rout et al. 2023). However, due to the abundance of dissolved organic matter (DOM) in environmental water, it is necessary to take into account the naturally occurring DOM for comprehensively understanding the environmental behavior of CNMs in a real aquatic environment (Jiang et al. 2017a, 2017b). Many studies demonstrated that the CNMs can easily adsorb DOM by hydrophobic interaction, electrostatic interaction, hydrogen bonding, and π–π interaction (Goodwin et al. 2018; Ren et al. 2018; Zhou et al. 2019). Furthermore, a few reports implied that the CNMs prefer to adsorb the DOM to form CNMs-DOM composites, leading to a significant reduction of the adsorption capacity to coexisting contaminants by blocking the adsorption sites on CNMs’ surface, probably due to the much higher concentrations of DOM than those of trace pollutants in aquatic environments (Lerman et al. 2013; Zhang et al. 2016; Munoz et al. 2021). On the other hand, it is well known that the DOM is the most photochemically active fraction of species mainly contributing to the indirect photodegradation of persistent organic pollutants (POPs) in environmental surface water since they can predominantly absorb sunlight irradiation and produce abundant reactive species such as ¹O₂, ^•OH, and triplet ³DOM* (Yu et al. 2019). However, it is unknown what the impacts of CNMs-DOM composites on the photodegradation of organic contaminants in environmental surface water until now. Thus, investigating the formation of CNMs-DOM composites and their capacities for the adsorption and photodegradation of organic contaminants would be vital to clarifying the effects of CNMs on the migration and transformation of environmental pollutants in real natural surface water.

Herein, the aims of this study were to explore the adsorption and photochemical decomposition of organic contaminants on CNMs in the presence of DOM at naturally occurring concentrations under sunlight irradiation. Fulvic acid (FA) was selected as a surrogate of natural DOM since FA is the most important and abundant component of natural DOM in aquatic environments (Yu et al. 2019; Jiao et al. 2022). Three typical CNMs (SWCNT, MWCNT, and GO) and three typical emerging organic contaminants, including estradiol, dibutyl phthalate (DBP), and tetracycline (TC) (chemical structures are presented in Supplementary Figure S1), were selected to conduct the adsorption and photodegradation experiments. The adsorption kinetics and isotherms of FA on CNMs were determined and the factors affecting the adsorption processes, such as pH and concentration, were evaluated. The prepared CNMs-FA composites were characterized and adsorption and photodegradation processes of the interested organic contaminants on the surface of CNMs-FA composites were investigated.

Multi-walled carbon nanotubes (MWCNTs, >99%), single-walled carbon nanotubes (SWCNTs, >95%), DBP, and furfuryl alcohol (FFA) were obtained from Tokyo Chemical Industry Co., Ltd (Tokyo, Japan). Graphene oxide (GO, >98%), ethinyl estradiol (EE2), 2,4,6-trimethylphenol (TMP, 98%), isopropyl alcohol (IPA, 99.8%), and fulvic acid (FA, 90.0%) were supplied by Aladdin Industrial Co., Ltd (Shanghai, China). TC, sodium hydroxide (NaOH), and phosphoric acid (H₃PO₄) were purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). HPLC-grade methanol (MeOH) and acetonitrile (ACN) were supplied by Supelco (USA). Except where noted, all reagents were of analytical grade and distill-deionized water (18.3 mΩ cm) was applied in all experiments.

Before studying the adsorption and photodegradation of organic contaminants with CNMs-FA composites, the adsorption performance of FA with CNMs, including adsorption capacities, kinetics, and isotherms, was comprehensively studied by static adsorption experiments. The investigations were performed in 100 mL conical flasks containing 100 mL FA solution and 5.0 mg CNMs (dry weight) and the mixture was stirred at 120 rpm in a water bath at 25 ± 1 °C under dark conditions. The pH of the aqueous FA solution was adjusted to 5.0, 7.0, and 9.0 using 0.1M HClO₄ and 0.1M NaOH solutions. Adsorption kinetic experiments were performed in conical flasks containing aqueous FA solutions (initial concentration: 16.0 mg L⁻¹) and pre-dried CNMs (5.0 mg) at pH 7.0 ± 0.1. Additionally, isothermal experiments were conducted at the initial concentrations of the FA solution in the range of 2.0–16.0 mg L⁻¹ at pH 7.0 ± 0.1 and stirred at 120 rpm for 12 h at 25 ± 1 °C. The concentration of FA in sample solutions was measured by using UV–Vis spectroscopy (UV-1900, Shimadzu, Japan) at 285 nm via the calibration curve method. After the adsorption process, the suspensions were centrifuged at 1 × 10⁴ rpm for 15 min and filtered with 0.22 μm polytetrafluoroethylene (PTFE) membrane. Then, the filtrated CNMs-FA composites were washed with distill-deionized water until the residual FA in the eluent could not be detected. Finally, the CNMs-FA composites were dried in a vacuum freeze dryer for 24 h and stored in a desiccator for further utilization.

The original CNMs and prepared CNMs-FA composites were characterized by several techniques to confirm the formation of CNMs-FA and examine their physicochemical properties. A cold-field scanning electron microscope (SEM; SU8010, Hitachi, Japan) was used for morphological observation by uniformly attaching the samples to an aluminum block conductive glue. The infrared spectra were recorded by using a Fourier transform infrared spectroscopy (FTIR; Nicolet iS10, Thermo, USA). The porosity of the samples was determined at liquid nitrogen temperature using a surface area and porosity analyzer (Autosorb-iQ, Quantachrome, USA) after the evacuation of the samples at 150 °C for 12 h.

To study the adsorption of organic contaminants by CNMs-FA composites, 30 mL of EE2, DBP, or TC at a concentration of 5.0 mg L⁻¹ was added into a 50-mL quartz tube. Then, suspensions of 0.6 mL of 1.0 g L⁻¹ CNMs-FA (SWCNT-FA, MWCNT-FA, or GO-FA) were added and stirred at 120 rpm for 12 h in a water bath at 25 ± 1 °C under dark. Then, they were centrifuged at 1 × 10⁴ rpm for 15 min and the concentration of EE2, DBP, or TC in the supernatant was measured by using high-performance liquid chromatography (HPLC) by the calibration curve method. The adsorption efficiencies of studied contaminants on CNMs-FA were calculated by the concentration change of contaminants in samples before and after adsorption.

To examine the photodegradation of organic contaminants in the presence of CNMs-FA composites under solar irradiation, a merry-go-round photochemical chamber reactor equipped with a 500 W Xenon Lamp (XE-JY250, Jiuping Instrument Co., Ltd, Wuxi, China) was used for carrying out all photolysis experiments. The emission spectrum of the Xenon lamp solar simulator is shown in Supplementary Figure S2. The light below 290 nm from the Xenon lamp was filtered with a cutoff filter to simulate sunlight. The intensity of the lamp was set at 5.0 mW cm⁻² to simulate the average intensity of the sunlight. Eight quartz photolysis tubes containing 20 mL of solution were held in the ring of the merry-go-round accessory. A hollow cylindrical lampshade with circulated cooling water (DC-1006, Wuxi Jiuping Co. Ltd, China) was employed to maintain the temperature at 25 ± 1 °C. The ring rotated within the reactor chamber at a speed of 20 rpm to give uniform irradiation to the quartz tubes. Tubes for the control samples were wrapped in aluminum foil to prevent light irradiation. The effect of pH on the photodecomposition of organic contaminants induced by CNMs-FA composites was evaluated. 5.0 mg L⁻¹ of individual organic contaminant (EE2, DBP, and TC) and 200 mg L⁻¹ of CNMs-FA (SWCNT-FA, MWCNT-FA, and GO-FA) suspensions were orthogonally prepared in 30 mL of aqueous solutions at a controlled pH of 7.0 ± 0.1 in quartz tubes. Then, the prepared sample solutions containing CNMs-FA and studied contaminants conducted photolysis experiments under simulated sunlight. Aliquots of samples were withdrawn after irradiating for 0.5, 1, 2, 4, and 8 h and centrifuged for 15 min at 1 × 10⁴ rpm. Following filtration with a 0.22 μm filter, the samples were injected into HPLC to analyze the concentration of the studied contaminants.

The concentrations of selected organic contaminants, including EE2, DBP, and TC, were analyzed by an HPLC (LC-20AT, Shimadzu, Tokyo, Japan) coupled with a UV–Vis detector (SPD-20A, Shimadzu, Tokyo, Japan) and a symmetry C₁₈ column (150 × 3.9 mm, 5 μm particle size, Waters, USA) guarded by a 10 mm C₁₈ guard column. The column temperature was maintained at 30 °C. The different mobile phases, flow rates, and detection wavelengths were applied for the analysis of different organic contaminants. For EE2, a mixture of ACN and deionized water (80:20; v/v) at a flow rate of 0.5 mL min⁻¹ was employed as a mobile phase and the detection wavelength was set at 230 nm; for DBP, a mixture of MeOH and deionized water (90:10; v/v) at a flow rate of 1.0 mL min⁻¹ was employed as a mobile phase and the detection wavelength was set at 275 nm; for TC, a mixture of methanol and 0.1% H₃PO₄ (35:65; v/v) at a flow rate of 1.0 mL min⁻¹ was employed as a mobile phase and the detection wavelength was set at 350 nm. The identification and quantification of organic contaminants in samples were performed by matching retention time against that of the standard and the calibration equation method, respectively. The working standard solutions for three organic contaminants were prepared freshly in the concentration range of 0.00–50.0 mg L⁻¹ by diluting the stock standards (1.0 mg mL⁻¹), which were prepared by dissolving 20 mg of the corresponding standards in 20.0 mL of methanol.

The uptake of FA by CNMs varied with solution pH, as shown in Supplementary Figure S3, and the absorption capacity of FA decreased from 84.93 to 30.05 mg g⁻¹, from 48.65 to 5.51 mg g⁻¹, and from 104.9 to 19.2 mg g⁻¹ for SWCNT, MWCNT, and GO, respectively, when pH increased from 5.0 to 9.0 adjusted by 0.1M of HClO₄ and NaOH solutions. This may be attributed to the enhanced hydrolysis of FA with the increase of pH, leading to the dehydrogenation of –OH and –COOH functional groups in FA, which resulted in the decrease of hydrogen-bonding donors, the abatement of hydrophobicity, and the increase of solubility in water (Huang et al. 2000; Lee et al. 2015). In addition, the negative charge associated with the ionized –OH and –COOH functional groups of FA may enhance the electrostatic repulsion between CNMs and FA. This can be confirmed by the observations that the surfaces of all three CNMs are negatively charged based on the results of Zeta potential measurements (Supplementary Figure S3), which were conducted by using a Zeta potential meter (Zetasizer Nano-ZS90, Malvern, UK) at different pHs from 5.0 to 9.0.

As depicted in Supplementary Figure S4(a), the adsorption kinetic curves showing that the uptake of FA quickly increased in the initial 2 h and subsequently slowly changed until the adsorption process reached equilibrium. In addition, the data (Supplementary Table S1) verified that the pseudo-first-order model is more reasonable for kinetic data than the pseudo-second-order model for the adsorption of FA with all three CNMs according to the correlation coefficient (R²) values. As shown in Supplementary Figure S4(b) and presented in Supplementary Table S2, compared with the Freundlich model, the experimental results were more consistent with the Langmuir model according to the correlation coefficients (R²), suggesting that the adsorption process was monolayer adsorption. Furthermore, based on the calculated results via the Langmuir model, the maximum adsorption capacities (q_m) of FA on SWCNT, MWCNT, and GO were 53.86 ± 6.85, 22.52 ± 2.31, and 34.65 ± 5.72 mg g⁻¹, respectively, which are consistent with the data of specific surface areas and pore volumes analysis based on N₂ adsorption and desorption for CNMs (Supplementary Table S3). Taken together, similar kinetic and isotherm models for the adsorption of FA on MWCNT, SWCNT, and GO were obtained, indicating that the similar adsorption mechanisms of the adsorption of FA with different CNMs were complied. The results revealed that the adsorption process of FA with CNMs was dominated by physisorption, which mainly involved hydrophobic interaction, Van der Waals force, π–π, and hydrogen-bonding interactions (Yang & Xing 2009; Jiang et al. 2017a; Zhou et al. 2019).

BET N₂ adsorption–desorption isotherm. The specific surface areas (SSAs) and pore volumes (PVs) of the studied CNMs and the corresponding CNMs-FA composites were measured by the N₂ adsorption data collected for each sample at liquid nitrogen temperature (Bel Japan, Inc.) and further calculated using the standard Brunauer–Emmett–Teller (BET) method, respectively. As presented in Figure 1(g) and Supplementary Table S3, both SSA_BET and PV of CNMs-FA were significantly reduced by 9.8–23.8% and 7.0–23.7% compared to the original corresponding CNMs, respectively, indicating that the adsorption sites on the surface of CNMs were occupied and the porous structure of CNMs might be clogged due to the adsorption of FA.

FTIR. The FTIR is one of the most powerful techniques to characterize the presence of functional groups and chemical bonds in organic compounds and synthesized materials. In this study, the FTIR spectra of SWCNT, SWCNT-FA, MWCNT, MWCNT-FA, GO, GO-FA, and FA were recorded as shown in Figure 1(h) and 1(i). The SWCNT and MWCNT exhibit typical C = C stretching vibration at ∼1,600 cm⁻¹. Meanwhile, the C = C peaks of SWCNT-FA and MWCNT-FA show red shifts of 1612 → 1639 and 1608 → 1627, respectively, which implies CNTs probably adsorbed FA via π–π interaction (Ahmed et al. 2018; Tokgöz et al. 2020). For GO, the bands at 3,407, 1,726, 1,396, 1,225, and 1,051 cm⁻¹ are attributed to O–H, =C–O, carboxyl C–OH, epoxy CO–OC, and alkoxy CO–O stretching vibrations, respectively, corresponding to the sp³ structure of GO (Hartono et al. 2009). The band at 1,620 cm⁻¹ is assigned to the C = C stretching vibration, corresponding to the sp² structure of GO (Chen et al. 2021). The intensity of the C = C peak significantly increases in GO-FA than that in the original GO, indicating that the sp² structure was enhanced due to the FA coating on GO. Also, the C = C peak of GO-FA shows a red shift of about 50 cm⁻¹ compared to that of FA. The reason might be that the dual-band characteristics of the aromatic ring of FA can be weakened by π–π interaction (Chen et al. 2021). Additionally, O–H flexural vibration peak shifts from 3,407 to 3,379 cm⁻¹ and C–O flexural vibration moves from 1,053 to 1,072 cm⁻¹ after absorbing FA, which suggests hydrogen bonding between FA and GO (Jiang et al. 2017b; Zhou et al. 2019). These observations strongly confirmed that the FA was adsorbed on CNMs and the dominant interactions are π–π interaction and hydrogen bonding.

It is well known that photodegradation, including direct photodegradation and indirect photosensitive degradation, is one of the most important naturally occurring transformation pathways of contaminants and plays a remarkable role in removing organic contaminants in environmental surface water (Yu et al. 2019). Numerous studies have illustrated that the naturally occurring DOM can significantly contribute to the photosensitive decomposition of environmental pollutants due to the strong oxidation or reduction capability of the photogenerated triplet ³DOM*, hydrate electrons (e_aq⁻), and reactive oxygen species (ROS), which are able to react with pollutants via direct oxidation, energy transfer, and/or electron transfer (Jiao et al. 2022). To verify whether the CNMs-FA can promote the photodegradation of pollutants in natural surface water, the photodegradation of selected contaminants, including EE2, DBP, and TC, was examined in the presence of CNMs-FA composites. Since the CNMs-FA composites are able to significantly adsorb the target organic contaminants, as discussed in Section 3.2, the photodegradation experiments were performed after pre-adsorption of organic contaminants on CNMs-FA for 12 h under dark.

The current study demonstrated that the environmental behavior of CNMs in the real aquatic environment mainly involves adsorbing naturally occurring DOM to form CNMs-DOM composites, which are able to adsorb and further promote the photodegradation of trace organic contaminants in surface water under sunlight irradiation. This study presented new data for the comprehensive elucidation of CNMs-DOM formation and their associated environmental impacts on the organic contaminants in environmental water. It would shed light on the investigation of the environmental behavior, environmental implications and applications of CNMs to meet the increasing demand for sustainable nanotechnology development.

We thank technicians from the Department of Chemistry at South-Central Minzu University for their operation of FTIR and BET experimental tests for this work. The work was jointly supported by the Research Funding (YZZ18018) and the Fundamental Research Funds for the Central Universities (CZT20017; KTZ20043), South-Central Minzu University.

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

The authors declare there is no conflict.