In this study, 4,4′-(hexafluoroisopropene) diphthalic acid (PDA)-CD polymers containing β-cyclodextrin (CD) were synthesized for the adsorption of endocrine disrupting chemicals (EDCs) and dyes. It features great adsorption of bisphenol A (BPA), methylene blue (MB) and neutral red (NR). The maximum adsorption capacities of MB, NR and BPA can reach 113.06, 106.8 and 51.74 mg/g, respectively. The tandem adsorption results revealed that adsorptions of dyes and BPA onto PDA-CD polymer were two independent processes: non-polar BPA entrapment by cyclodextrin cavities while dyes were captured by the carboxyl groups and π-π stacking interactions. The adsorption processes performed well in a wide range of pH (4.0–10.0) and were not affected by fulvic acid (FA) and inorganic ions.

  • PDA-CD with abundant cavities, benzene rings and carboxyl groups were successfully fabricated.

  • PDA-CD showed excellent adsorption capacity for dye and BPA.

  • The molecular recognition by CD cavity contributed to the BPA adsorption.

  • Electrostatic attraction and π-π interaction dominated dye adsorption.

Graphical Abstract

Graphical Abstract
Graphical Abstract

In the textile industry, large volumes of water and chemicals are required in processing. According to statistics, about 200 L of water is consumed to manufacture 1 kg of textile (Zhou et al. 2019b). More than 100,000 commercially available dyes with approximately 7,000 hundred tons of dye-stuff are synthesized each year (Robinson et al. 2001). The discharge of textile wastewaters into water bodies normally affects the aquatic organisms and human beings (Verma et al. 2012; Liu et al. 2020). Conventional efficient methods of removing dyes mainly include adsorption technology (Gamoudi & Srasra 2019; Kloster et al. 2019), advanced oxidation (Van et al. 2019; Yang et al. 2019) and membrane separation (Guo et al. 2019; Konca & Çulfaz-Emecen 2019). Normally, the main deficiencies in membrane separation technology consist of high cost, frequent membrane fouling and requirement for different pretreatments (Verma et al. 2012). Intermediates with high toxicity may occur during the advanced oxidation processes and effectiveness fluctuates widely with the type of pollutants present in the textile wastewater (Verma et al. 2012). Adsorption technology has become an extensive approach to remove dyes as it is cost-effective, offers easy operation and is highly efficient (Qadeer 2007). However, many people just pay attention to dye removal from textile wastewater and ignore endocrine disrupting chemicals (EDCs) in effluents (Zhou et al. 2019b). Bisphenol A (BPA) is used as an intermediate chemical in the manufacture of antioxidants and dyes (Xue et al. 2017). Excessive BPA from textile wastes has been detected (Zijian et al. 2015). BPA is a very typical EDC and important raw material for polycarbonate plastics and resins. Thus, the removal of coexisting pollutants in textile effluents becomes a matter of continuing concern.

Recently, use of by-products from agricultural and forestry waste as raw materials to prepare green adsorbents has become a hot topic in the environmental field because of their non-toxicity and easy access (Ozmen et al. 2008). β-cyclodextrin (CD) is a kind of oligosaccharide extracted from starch-containing crops by catalytic enzymes. CD-based materials have been widely used to remove dyes, organic matters, metal ions and active bacteria from water (Chen et al. 2020; Hu et al. 2020). The development trend of CD materials is to construct porous materials with high specific surface area and high porosity to absorb or degrade micropollutants in water (Zhou et al. 2020). The uniqueness of CD lies in its spatial structure. The nature of a hydrophobic cavity at the center of molecular CD allows the formation of a host-guest inclusion complex with guest compounds through a hydrophobic-hydrophobic interaction (Choi et al. 2019; Xu et al. 2020). It is noteworthy that the high density of oxhydryl groups on the exterior are given a great chance to be modified by diverse functional groups, furnishing the CD with extra specific properties, which makes it possible for it to become an adsorption material. But being a high water-soluble cyclodextrin monomer unavoidably limited its application in the field of adsorption. Thus, water-insoluble cyclodextrin polymers have been widely researched by many people (Crini 2014; Wu et al. 2018). A cross-linking agent rich in carboxylic acid, such as citric acid (Huang et al. 2018), thiomalic acid (Zhang et al. 2019) and EDTA (Zhao et al. 2015), has always been the first choice because the carboxyl groups not only perform as cross-linking sites, but also as adsorption sites. And the process is usually handy to operate.

A 4,4′-(hexafluoroisopropene) diphthalic acid (PDA) molecule contains two benzene rings and four carboxyl groups. The ample carboxyl groups can be cross-linked with the hydroxyl groups of the cyclodextrin to form a water-insoluble polymer. In theory, the carboxyl groups and hydrophobic cavity can provide adsorption sites for cationic dyes and BPA, respectively. When organic pollutants with hydrophobic parts or phenyl groups and nitrogen cations are absorbed, people have not figured out whether the hydrophobic cavity and surface groups interfere with each other. Fortunately, based on previous researches (Zhao et al. 2009; Liu et al. 2015), we can conjecture that while CD comes across an amphiphilic guest molecule, its cavity inclines to mastery of the hydrophobic part of the molecule, and the remaining part will interact with the polar groups on the surface (Huang et al. 2018).

In this paper, an insoluble water PDA-CD was successfully prepared with a carboxyhydroxy esterification crosslinking principle. The use of PDA as a crosslinking agent has the following advantages: (1) rich carboxyl groups can provide more adsorption sites for cationic dyes (Chen et al. 2020); (2) benzene rings of PDA can form a conjugation effect with benzene rings in dyes; (3) the preparation process is simple and convenient. The morphological structure of PDA-CD was analyzed. Cationic dyes such as methylene blue (MB) and neutral red (NR), and BPA were used as target pollutants to further study the adsorption performance. The effects of initial pH of the solution, fulvic acid (FA) and inorganic ions on the adsorption performance were studied.

Materials

4,4′-(hexafluoroisopropene) diphthalic acid (PDA), β-cyclodextrin (CD), bisphenol A (BPA), methylene blue (MB) and neutral red (NR) were purchased from Aladdin Industrial (China). Fulvic acid (FA) was bought from Shanghai Yuanye Bio-Technology (China). KCl and CaCl2 were obtained from Shanghai Lingfeng Chemical Reagent (China). All the chemicals were 98 + % pure. Other chemical agents used in this paper were analytical reagent grade. Water used in this paper was purified using a Milli-Q water purification system from Millipore (USA). All reagents were used without further purification.

Preparation of PDA-CD

In the process of synthesis (Figure 1), potassium dihydrogen phosphate, PDA and CD acted as catalyst, crosslinking agent and matrix, respectively. The active hydroxyl group on CD was esterified with a polyvalent carboxyl group on PDA. The cross-linking reaction is carried out to crosslink a plurality of cyclodextrin monomers together to obtain water-insoluble PDA-CD polymers. According to different molar ratios (PDA: CD = 3:1, PDA: CD = 4:1, PDA: CD = 5:1), an accurately weighed certain amount of β-CD and PDA were added into a 600 mL beaker. The sum of the mass of CD and PDA in the three molar ratios was 1.5 g, and 0.25 g of potassium dihydrogen phosphate (KH2PO4) was added to each beaker. 100 mL of deionized water was used to solubilize the material in the beaker. Then, the beaker was placed in an oven and reacted at 140 °C for 3.5 h. After the reaction was completed, the materials were repeatedly rinsed with deionized water to remove surface residue of the materials, and the washed materials were placed in an oven to be dried to constant weight.

Characterization and adsorption performance test

All PDA-CD polymers were dried prior to inspection to avoid the effect of moisture on the characterization of the materials. Scanning electron microscopy (SEM) measurements were employed with a Hitachi S-3400N scanning electron microscope. Fourier transform infrared (FT-IR) spectra were recorded to distinguish the functional groups of polymers via a Nicolet 5700 spectrometer. The specific surface area of the polymers was calculated using the Brunauer Emmett Teller (BET) method with a Micromeritics TriStar II 3020 surface area and porosity analyzer.

In this paper, three pollutants, MB, NR and BPA were selected as target pollutants. FA, widely used to represent other organics in water, reflects the influence of organics for adsorbents (Yang & Xing 2009).Three different molar ratios of PDA-CD polymers were explored. The removal efficiency was calculated by Equation (1). Three parallel experiments were performed on the same pollutants and pH conditions, and the average value was taken as the final result.
formula
(1)
where C0 (mg/L) and Ce (mg/L) represent the initial pollutants concentration and equilibrium concentration of pollutants.

Effect of pH

Zeta potentials under different pH values of solution were obtained at 25 °C using Zeta sizer Nano ZS (Malvern ZEN3600). The capacity of adsorption was estimated by removal efficiency of pollutants. During the experiment, the initial concentration of the three target pollutant solutions was set to 50 mg/L, and the pH values range of the MB and BPA solutions was set from 2.0 to 10.0. Since NR is an acid-based indicator, it features different colors under different acid-based conditions, which is not conducive to detection. At the same time, it was found that NR solution will produce flocculent precipitation under alkaline conditions, influencing the accuracy of the test results. So the pH values range of the NR solution was set from 2.0 to 6.0. The pH of the solutions was adjusted with HCl and NaOH.

Adsorption isotherms

Different concentrations of pollutants (MB and NR: 10–500 mg/L, 50 mL; BPA: 5–200 mg/L, 50 mL) were prepared in a 150 mL erlenmeyer flask. 50 mg PDA-CD polymers were added into each flask. For better studying the adsorption mechanism between pollutants and polymers, three adsorption isotherm models were selected to fit the experimental data, namely Langmuir, Freundlich and Sips models.

Equation (2) is the calculation formula of the Langmuir model (Chen et al. 2019).
formula
(2)

In the formula, Ce (mg/L) represents the equilibrium concentration of contaminants; qe (mg/g) and qm (mg/g) represent equilibrium as well as theoretical saturated adsorption capacity, respectively; KL represents the Langmuir isotherm constant with respect to enthalpy and adsorption energy and its unit is L/mg.

Equation (3) is the calculation formula of the Freundlich model.
formula
(3)

In the formula, KF ((mg/g) (L/mg)nf) represents the Freundlich capacity constant, indicating the magnitude of the adsorption capacity; 1/nf is the heterogeneity factor (Raza et al. 2019).

Equation (4) is the calculation formula of the Sips model.
formula
(4)

In the formula, KS (L/mg) and ns−1 represents the Langmuir association constant and the Freundlich heterogeneity parameter, respectively.

Adsorption kinetics

The 0.1 g PDA-CD were respectively added into 100 mL of MB, NR and BPA solution (pollutant concentration: 50 mg/L), stirring for one hour (250 rpm/min). Water samples were achieved at predefined time intervals (0, 5, 10, 20, 30, 40, 60, 120, 180 min). The adsorption capacity qt of the adsorbent at different time points is calculated by Equation (5).
formula
(5)
where C0 (mg/L) and Ct (mg/L) represent the initial and residual the concentration of the pollutants, whereas V (mL) represents the volume of the pollutant solution and m (mg) means to the total mass of the adsorbent.
All experimental data were fitted with non-linearized pseudo-second-order kinetic model (Acemioglu 2005) as calculated as Equation (6).
formula
(6)
where K (g/mg/min) and qe (mg/g) represent the pseudo-secondary adsorption kinetic constants and the equilibrium adsorption capacity at a reaction time of t, respectively.

Tandem adsorption

According to the adsorption sequences of these three pollutants, the experiments have been divided into two forms: dual (NR/MB, BPA/MB, MB/NR, BPA/NR, NR/BPA, MB/BPA) and triple (NR/BPA/MB, BPA/NR/MB, MB/BPA/NR, BPA/MB/NR, MB/NR/BPA, NR/MB/BPA). Different forms of systems represent different adsorption sequences; for instance, NR/MB means adsorption of MB after the completion of NR adsorption. Specifically, the 0.1 g PDA-CD were added into 100 mL of NR solution (concentration: 50 mg/L), stirring for three hours (250 rpm/min), and concentration of pollutants were measured at the designed intervals (0, 5, 10, 20, 30, 40, 60, 120, 180 min). After 3 h, the adsorbed NR PDA-CD were separated from solution and added into MB solution (concentration: 50 mg/L) to carry on adsorption again. Furthermore, NR/MB/BPA expresses adsorption of BPA after the completion of the NR/MB adsorption. Specifically, the polymers adsorbed on NR and MB were separated from solution and added to a solution of 100 mL of BPA at a concentration of 50 mg/L. Physical and chemical conditions and sampling conditions were the same in all adsorption processes.

Effect of fulvic acid and inorganic ion

Effect of fulvic acid

50 mL (MB: 50 mg/L, NR: 50 mg/L; BPA: 50 mg/L) solutions were prepared respectively in a 150 mL erlenmeyer flask and appropriate amount of humic acid were added to each flask. The concentrations of humic acid in the solutions were designed as 20, 50, 100 and 200 mg/L respectively, and a blank experimental group was set as a control. The dosage of the polymers was set as 1 g/L. The conical flasks were shaken for 24 h at 25°C. After the oscillating adsorption, the precipitate processes were completely allowed to stand for 30 min, and the supernatant in the conical flasks were taken for detection.

Effect of inorganic ion

The presence of inorganic ions may affect the adsorption performance of polymers in water. Therefore two kinds of inorganic ions (K+&Ca2+) have been selected to study the adsorption properties of adsorbent PDA-CD on the three pollutants in the presence of inorganic ions in solution. 50 mL (MB: 50 mg/L, NR: 50 mg/L, BPA: 50 mg/L) solutions were prepared respectively in 150 mL erlenmeyer flasks. K+ and Ca2+ were provided by potassium chloride and calcium chloride, respectively. The concentrations of potassium and calcium ions in each kind of pollutant solution were both set as 20 mg/L and 200 mg/L, along with a blank experimental group as a control.

Analytical method of pollutant concentration

The BPA involved in this paper was detected by L-2000 high performance liquid chromatography. The column of the instrument was a C18 column of Guoan Jielun Company (4.6 × 250 mm, 2.5 μm), the mobile phase used methanol and water (70/30, v/v), the column temperature was controlled at 40°C, and the UV detection wavelength was 278 nm. All test samples should be filtered using a 0.22 μm hydrophilic Teflon needle filter to protect the column before testing. The concentrations of MB and NR, were tested using a Unico 2802 UV-Vis spectrophotometer. According to the full wavelength scanning of the two dye ions, the maximum absorption wavelengths of MB and NR are 665 nm and 530 nm, respectively. All the experimental data in this paper were obtained by three repeated parallel experiments, and the difference between the obtained data was analyzed.

Characterizations and properties of PDA-CD

The FT-IR spectra of the three different molar ratios of PDA-CD and original materials (PDA and CD) are illustrated in Figure 2(d). The CD and PDA-CD of three different molar ratios all exhibited strong absorption peaks with a large span in the range of 3,000 cm−1 to 3,700 cm−1, resulting from stretching vibrations of -OH in β-ring. PDA-CD in three ratios all exhibited new strong absorption peaks at 1,737 cm−1, associated with stretching of C = O groups from carboxyl groups. Beyond that, PDA-CD all presented two new absorption peaks at 1,149 cm−1 and 1,027 cm−1, ascribing to -COC- and CO stretching vibrations of structure of CD (Chen et al. 2020) and revealed fluorine related peaks in the range from 1,214 cm−1 to 1,251 cm−1 (Ren et al. 2004). These results demonstrate the success of CD and PDA crosslinking esterification. For studying the surface morphology and pore structure characteristics of CD polymer, three ratios of polymers were scanned by SEM and BET analysis. The SEM and BET of three molar ratios of synthesis are displayed in Figure 2(a)–2(c). As the proportion of the cross-linking agent PDA in the CD polymer decreased, the surface of the prepared polymers gradually approached smoothness, the crumb-like substances gradually decreased, and cracks appeared. The BET values of three different molar ratios of PDA-CD polymers (b: 3:1, c: 4:1, d: 5:1) were 0.56 m2/g, 0.32 m2/g and 0.41 m2/g, respectively. The surface morphology and specific surface area of the three proportions of PDA-CD polymer show little difference, which could explain that the true crosslinking ratio is not the added ratio, resulting in little difference between the three proportions of the polymer. It is further verified that PDA-CD retained the complete but underdeveloped cavity structure of CD. Therefore, it is speculated that the adsorption capacities of the three ratios of polymer had great differences. Three pollutants (MB, NR and BPA) have been chosen to testify the adsorption properties of PDA-CD polymers. The relative results have been unfolded in Figure 2(e). When the PDA-CD were used in removal of MB, the removal efficiency with polymers in the ratio of 3:1 exceeded about 11% and 18%, in comparison with 4:1 and 5:1, respectively. For NR, the removal efficiency with polymers in three ratios were not too much different. Although the carboxyl groups on all three polymers were active, the amounts differ. Maybe their carboxyl groups were covered with different degrees. In addition, the dispersion of small particles in solution was good, but with the increase of the molar ratio of PDA/CD from 3 to 5, the dispersion of the polymer became worse, which was consistent with the results of SEM, resulting in less carboxyl group exposure. For BPA, polymers in the ratio of 3:1 had a great advantage that the removal efficiency exceeded about 52.6% and 58.4%, compared with the ratios of 4:1 and 5:1, respectively. Considering the result in BET, the oversupply of PDA would format self-polymerization to block the cavity of CD to different degrees. Thus, PDA-CD prepared in the ratio of 3:1 could retain as many carboxyl groups and cavities as possible.

Effect of pH on adsorption

The pH of solution has an important influence on the adsorption capacity. Zeta potential is used to investigate the surface charge of the adsorbent under different pH values and the results are shown in Figure 3(a). When the initial pH value of the solution was within 2.0–10.0, the Zeta potential of the polymers' surface was negative. It is explained that, in the aqueous environment, the -COOH on the surface of the PDA-CD tended to be deprotonated, resulting in –COO groups and H+ ions, contributing to the negatively charged surface of the polymers. With the pH values increased, the deprotonization of carboxyl groups was looming large. So that was why the Zeta potential nearly decreased all the time as the pH values increased. This case unquestionably has boosted formation of electrostatic attraction between polar pollutants and adsorbents. The result of the pH variation during the adsorption process is depicted in Figure 3(b), and this coincided with the variation trend of Zeta potentials. In the processes of adsorbing the MB and NR, the removal efficiencies had a close relationship with the pH of the initial solution. Combined with the experimental results in Figure 3(a), the Zeta potential of polymers showed a steep decrease. When the pH value was from 2.0 to 5.0, removal efficiencies of the MB and NR increased significantly from 44.3% to 92.7% and 31.1% to 80.2%, respectively. While the pH values were from 5.0 to 9.0, the Zeta potentials decreased slowly, but the corresponding removal efficiencies of dyes did not vary much. The differences in removal efficiency of dyes may arise from their different molecular shapes and pka values (Chen et al. 2020).

At low pH level, the protonation of functional groups in virtue of interacting with H+ on PDA-CD contributes to the positive surface charge. The resulting repulsive forces suppressed the interaction of the cationic dyes on the adsorbent and the removal efficiency decreased. With pH value increased, the functional groups were more likely to deprotonate and the cationic dyes were favorably adsorbed by the negatively charged surface of PDA-CD (Zhou et al. 2019a). In contrast with MB and NR, the adsorption of BPA performed well over a broad range of pH values (2.0–10.0) and the removal efficiency stabilized in the range of 33% to 36%. This phenomenon is attributed to the cavity from CD forming host-guest compounds with BPA through hydrophobic interaction (Lee & Kwak 2019). Based on the analysis above, the pH of the solution was adjusted to 6.0 during the next experiments.

Adsorption isotherms

The isothermal adsorption on PDA-CD is exhibited in Figure 4(a)–4(c) and the parameters are demonstrated in Table 1. During the experiment, the initial pH of each pollutant solution was adjusted to 6.0.

The Langmuir model supposes one molecule in thickness and the adsorption sites on materials are assumed as identical and equivalent, without lateral interaction between the adsorbed molecules (Chen et al. 2020). Freundlich isotherm is used to describe non-ideal and reversible adsorption, not limited to monolayer adsorption. From Table 1, it can be found that all the parameters of 1/nf are in the range between 0.1 and 0.5, indicating that all the adsorption processes of three pollutants belong to the favorable level (Crini 2008) and the relationship between the Kf values of the three pollutants follows as: Kf (MB) > Kf (NR) > Kf (BPA). It can be concluded that MB adsorption belonged to highest priority adsorption and BPA the lowest.

For the MB and NR in the three isothermal adsorption models, the correlation coefficient parameter (R2) fitted by the Sips model was 0.961 and 0.950 respectively. The R2 fitted by the Sips model is highest in comparison with the Langmiur and Freundlich models. The theoretical maximum dye (MB, NR) adsorption capacity (qm) obtained by the Sips model can reach 113.06 mg/g and 106.8 mg/g, respectively, which is close to the experimental data (115.3 mg/g, 103.6 mg/g). For BPA, the Sips model also has the highest correlation coefficient of the three isothermal fitting models. The theoretical maximum BPA adsorption capacity (qm) can reach 51.74 mg/g. Regarding the previous study of the adsorption of BPA by cyclodextrin polymer, it was concluded as a single-layer adsorption of pollutants in a 1:1 ratio, which fitted well with the Langmuir model (Wang & Harrison 2018; Lee & Kwak 2019). However, the Sips fitting model obtained in this paper is not contradictory to them. Considering the bias resulting from the transformation, the correlation coefficient may not the only consideration factor in judging the isotherm model (Foo & Hameed 2010). The Sips isotherm is a combination of the Langmuir and Freundlich models, which breaches the restriction of the potentially infinitive pollutants concentration in the Freundlich isotherm model (Meili et al. 2019). Thus, the adsorption processes of dyes and BPA are detected as followed: diffused adsorption at low concentrations of pollutants and a monomolecular adsorption with a saturation value at high concentrations of pollutants (Muntean et al. 2013). According to Ks in Table 1, it can be inferred that MB has higher affinity than NR and BPA for the PDA-CD. The values of ns in all pollutants adsorption are larger than 1, meaning that all processes are heterogeneous in nature. The data in Table 2 have shown some other materials to remove the MB, NR and BPA. By comparing qe,max of other materials, it can be found the adsorption capacity of our material is at high level. Furthermore, cyclodextrin adsorbents generally have a better ability to remove BPA.

Adsorption kinetics

As shown in Figure 5(a), when the effective contact time of MB reached 30 min, the equilibrium uptake achieved 80% and then achieved equilibrium within 180 min. For NR and BPA, the effective contact times were 40 min and 60 min, and the equilibrium adsorption capacity of PDA-CD can reach 39.91 mg/g and 18.35 mg/g after 180 min, respectively. The experimental results in Figure 5(a) also show that, after 30 min, the removal efficiency of MB, NR and BPA by PDA-CD can reach 80.9%, 57.9% and 22.4%, respectively.

In order to further explore the adsorption mechanism of PDA-CD in the adsorption process, pseudo-secondary dynamics models have been performed to fit with the experimental data of adsorption. The results in Figure 5(b) reveal that the kinetic data fitted well with the pseudo-second-order kinetic model, indicating that chemisorption dominates the entire processes of adsorption. Due to the data in Table 3, it can be found that the fitted adsorption capacity (qecal) is close to the experimental results (qeexp). The k is usually used to estimate the rate of adsorption reaction. As shown in Table 3, the adsorption processes of MB, NR and BPA were all fast in the initial stage and then decelerated gradually along with the continuation of the reaction. Furthermore, the adsorption rate of PDA-CD for MB, NR and BPA is: MB > NR > BPA. This would be attributed to limited porosity making the cavity of CD less accessible to BPA comparing with the surface binding sites.

Tandem adsorption

As shown in Figure 5(c), the adsorption capacity of MB by PDA-CD can reach 47.04 mg/g in a single system after 180 min. After adsorption of BPA, the adsorption capacity of MB changed little and maintained at about 47.04 mg/g. The adsorption kinetic curve in Figure 6(a) observed from the single system (MB) and dual system (BPA/MB) overlap almost completely. However, after adsorption of NR, the adsorption capacity of MB decreased to 34.17 mg/g. In two triple systems, MB adsorption capacities were all inhibited. After adsorption of NR/BPA and BPA/NR, the adsorption capacity of MB decreased by 26% and 28% to 33.47 mg/g and 34.07 mg/g, respectively. As shown in Figure 6(a), the trajectory of adsorption curves of dual (BPA/MB) is completely inconsistent with triple (NR/BPA/MB, BPA/NR/MB). It is proved that the adsorption processes of BPA and MB on PDA-CD are relative independently. However, NR on the adsorbent will inhibit the adsorption of MB.

As shown in Figure 5(d), the adsorption capacity of NR by PDA-CD can reach 39.91 mg/g in the single system after 180 min. After adsorption of MB, MB/BPA and BPA/NR, the adsorption capacity of NR fell by 25%, 26% and 28% to 34.17, 33.47 and 34.07 mg/g, respectively. Figure 6(b) shows that the kinetic curves of MB/BPA/NR, BPA/MB/NR and MB/NR basically coincide, and completely disjunction with the kinetic curves of NR. However, BPA on the adsorbent did not affect the adsorption of MB, and its adsorption capacity was maintained at about 40.29 mg/g. The curve of BPA/NR in Figure 6(b) almost coincides with the kinetic curve of NR. MB on the adsorbent inhibited the adsorption of NR but did not affect the adsorption of BPA, which was consistent with the conclusion in Figure 5(c).

As shown in Figure 5(e), the adsorption capacity of BPA remained at about 18 mg/g in all four tandem systems, which was not significantly different from 18.34 mg/g in a single system. The five dynamics curves in Figure 6(c) almost coincide. This obviously shows that the adsorption of dye and BPA are two independent processes.

According to data in Table 3 and Figure 6, all the adsorption curves were in line with the pseudo-second-order kinetic model. This means that the dye or BPA on the adsorbent does not change the chemisorption-dominated processes.

It is well known that BPA is a kind of hydrophobic organic micro-pollutant. According to the ‘internal hydrophobic, external hydrophilic’ feature of cavity CD, BPA can be encapsulated by means of hydrophobic interactions (Kono et al. 2013; Huang et al. 2018; Zhou et al. 2019a). Numerous studies have proved that merely one phenyl ring of the BPA is able to get into the cavity of CD, while the other stayed outside (Yang et al. 2008). Weiwei Huang et al. (Huang et al. 2018) argued that coexistence of two or more pollutants in one CD cavity was impossible because of the molecular structure. Meanwhile, the BET characterization indicates that the pore structure of the PDA-CD is underdeveloped and the cyclodextrin cavity provides less adsorption sites than the sites provided by the carboxyl group on the surface. The results of the researches above have further determined our hypothesis that adsorptions of dyes and BPA are mainly attributed to carboxyl groups on PDA and the hydrophobic cavity in CD, respectively.

Effect of fulvic acid and inorganic ion

In total, 3–28% of the organic matter in sewage is humus (Moura et al. 2007). The choice of FA is because it is the most soluble and mobile component of humus in water and would interfere with the adsorbent relying on its rich groups (Yang & Xing 2009). Therefore, in order to study the stability and anti-interference ability of PDA-CD, this section selected FA as a representative of natural organic matter to study the adsorption of natural organic matter on PDA-CD. The relative experimental results are shown in Figure 7(a).

The experimental results show that when the concentration of FA in the solution was in the range of 20 mg/L to 200 mg/L, the removal efficiencies of the three target pollutants (MB, NR, BPA) remained basically alike. The removal efficiencies of MB, NR and BPA were maintained in the range of 92.3% to 93.7%, 79.2% to 80.3%, and 34.7% to 36.7%, respectively. These results indicate that the PDA-CD has superior anti-interference ability and practical application potential. There are two main reasons for this: firstly, the previous analysis and proof that MB and NR were removed from the aqueous solution by binding to the carboxyl group on the surface of the PDA-CD adsorbent material. However, FA did not react with carboxyl groups, and the presence of FA in the solution did not affect the adsorption of MB and NR; secondly, the FA molecule has a large size and cannot enter the hydrophobic cyclodextrin cavity, thus it would not compete with BPA for adsorption sites in the solution. So it can be concluded that the presence of FA in the solution will not affect the PDA-CD adsorption of BPA. In addition to humic acid, there are other complex organic compounds in the textile effluents. The specific organic composition depends on the printing and dyeing process and the type of additives added.

On the other hand, considering various inorganic ions such as K+, Ca2+ and Cl exist in actual textile wastewater, the effect of mineral ions needs to be analyzed. These ions may compete with the contaminant molecules for the adsorption sites available on the adsorbent, resulting in a decrease in the adsorption performance of the adsorbent. As seen from Figure 7(b), when the K+ ion concentration in the solution was 20 mg/L or 200 mg/L, the adsorption capacity of PDA-CD for MB, NR and BPA did not differ much compared with previously.

The removal efficiencies of MB in two concentrations of K+ ion were 94.2% (20 mg/L) and 92.3% (200 mg/L), and NR were 81.6% (20 mg/L) and 80.4% (200 mg/L). The removal efficiencies of BPA were 38.2% (20 mg/L) and 36.4% (200 mg/L). Compared with the blank experiment data, the adsorption capacity did not present too much difference. Similarly, when 20 mg/L and 200 mg/L of Ca2+ ions were present in solution, shown in Figure 7(c), the removal efficiencies of the three contaminants in the solution were hardly affected. Such results indicate that the adsorption performances of the adsorbent PDA-CD to MB, NR and BPA were hardly affected while the inorganic ions existed in the solution.

Proposed adsorption mechanism

For gaining insight into the adsorption mechanism of adsorption performances, the FT-IR before and after adsorption needs to be compared. The results are presented in Figure 8. It was found that the phenolic O-H absorption peak of PDA-CD shifted (3,419 cm−1 → 3,437 cm−1) after adsorbing BPA. It can be reasonably inferred that BPA adsorbed on PDA-CD caused the shift of the phenolic O-H absorption peak. At the same time, the intensity of the C = O (1,737 cm−1) absorption peak on the adsorbed PDA-CD-NR and PDA-CD-MB is weaker than that of PDA-CD, which can predict that the NR molecules and the MB molecules indeed interacted with -COOH on the PDA-CD via electrostatic attraction. Li et al. (2018) concluded that ππ interactions with the benzene rings on β-CDSP and adsorbates contributed significantly to the adsorption of NR. Fu et al. (2015) pointed that the ππ stacking interactions could occur between MB molecules and dopamine according to MB is an ideally planar molecule with abundant aromatic rings. Liu et al. (2019) have achieved a similar conclusion. It is noticed that MB and NR are planar molecules with abundant aromatic rings and a PDA molecule also has two benzene rings. Due to the results in Figure 8, the stretching vibration band of C = C on the aromatic ring is obviously shifted after the dyes were adsorbed (MB:1,612 cm−1 → 1,621 cm−1, NR:1,612 cm−1 → 1,596 cm−1). So it can be confirmed that ππ stacking interactions can happen between MB molecules and PDA-CD, resulting in the shift of the peak.

In this paper, CD, a kind of oligosaccharide extracted from starch-containing crops, was cross-linked with PDA in three different molar ratios (PDA: CD = 3:1, PDA: CD = 4:1, PDA: CD = 5:1) through simple and friendly synthesis procedures. The molar ratio of 3:1 (PDA: CD = 3:1) was the most optimal choice by comparing yields, water absorption volumes and adsorption performances. The PDA-CD showed good adsorption performances in an extensive pH range of 4.0 to 10.0. Besides, the kinetics and isotherms studies showed that the adsorption processes abided by the pseudo-second-order kinetic model and the Sips model, respectively. The maximum adsorption capacities can reach 51.74 mg/g for BPA, 113.06 mg/g for MB and 106.8 mg/g for NR. Through tandem adsorption, it is further proved that the adsorption processes of dyes and BPA are relatively independent. Moreover, PDA-CD had a strong anti-interference ability. The FA and inorganic ions (K+, Ca2+) had little effect on the adsorption performance. Considering the actual situation of textile wastewaters, this kind of polymer will be a good choice for treatment in the future.

This work was supported by the National Natural Science Foundation of China (51778230), Shanghai Rising-Star Program (17QA1401000), and the Fundamental Research Funds for the Central Universities for their financial support (222201718003).

Acemioglu
B.
2005
Batch kinetic study of sorption of methylene blue by perlite
.
Chemical Engineering Journal
106
(
1
),
73
81
.
Chen
W.
Lin
Q.
Cheng
S.
Wu
M.
Tian
Y.
Ni
K.
Bai
Y.
Ma
H.
2019
Synthesis and adsorption properties of amphoteric adsorbent HAx/CMC-yAl
.
Separation and Purification Technology
221
,
338
348
.
Crini
G.
2014
Review: a history of cyclodextrins
.
Chemical Reviews
114
(
21
),
10940
10975
.
Foo
K. Y.
Hameed
B. H.
2010
Insights into the modeling of adsorption isotherm systems
.
Chemical Engineering Journal
156
(
1
),
2
10
.
Hu
X.
Xu
G.
Zhang
H.
Li
M.
Tu
Y.
Xie
X.
Zhu
Y.
Jiang
L.
Zhu
X.
Ji
X.
Li
Y.
Li
A.
2020
Multifunctional β-cyclodextrin polymer for simultaneous removal of natural organic matter and organic micropollutants and detrimental microorganisms from water
.
ACS Applied Materials & Interfaces
12
(
10
),
12165
12175
.
Huang
W.
Hu
Y.
Li
Y.
Zhou
Y.
Niu
D.
Lei
Z.
Zhang
Z.
2018
Citric acid-crosslinked β-cyclodextrin for simultaneous removal of bisphenol A, methylene blue and copper: the roles of cavity and surface functional groups
.
Journal of the Taiwan Institute of Chemical Engineers
82
,
189
197
.
Liu
L.
Gao
Z. Y.
Su
X. P.
Chen
X.
Jiang
L.
Yao
J. M.
2015
Adsorption removal of dyes from single and binary solutions using a cellulose-based bioadsorbent
.
ACS Sustainable Chemistry and Engineering
3
(
3
),
432
442
.
Liu
R. J.
Fu
H. X.
Lu
Y. M.
Yin
H. Q.
Yu
L. Q.
Ma
L. Y.
Han
J.
2016b
Adsorption optimization and mechanism of neutral red onto magnetic Ni0.5Zn0.5Fe2O4/SiO2 nanocomposites
.
Journal of Nanoscience and Nanotechnology
16
(
8
),
8252
8262
.
Liu
Q.
Li
Y.
Chen
H.
Lu
J.
Yu
G.
Möslang
M.
Zhou
Y.
2020
Superior adsorption capacity of functionalised straw adsorbent for dyes and heavy-metal ions
.
Journal of Hazardous Materials
382
,
121040
.
Meili
L.
Lins
P. V. S.
Costa
M. T.
Almeida
R. L.
Abud
A. K. S.
Soletti
J. I.
Dotto
G. L.
Tanabe
E. H.
Sellaoui
L.
Carvalho
S. H. V.
Erto
A.
2019
Adsorption of methylene blue on agroindustrial wastes: experimental investigation and phenomenological modelling
.
Progress in Biophysics and Molecular Biology
141
,
60
71
.
Muntean
S. G.
Paska
O.
Coseri
S.
Simu
G. M.
Grad
M. E.
Ilia
G.
2013
Evaluation of a functionalized copolymer as adsorbent on direct dyes removal process: kinetics and equilibrium studies
.
Journal of Applied Polymer Science
127
(
6
),
4409
4421
.
Ndagijimana
P.
Liu
X.
Li
Z.
Yu
G.
Wang
Y.
2019
Optimized synthesis of a core-shell structure activated carbon and its adsorption performance for Bisphenol A
.
Science of The Total Environment
689
,
457
468
.
Qadeer
R.
2007
Adsorption behavior of ruthenium ions on activated charcoal from nirtic acid medium
.
Colloids and Surfaces A: Physicochemical and Engineering Aspects
293
(
1–3
),
217
223
.
Ren
L.
Fu
W.
Luo
Y.
Lu
H.
Jia
D.
Shen
J.
Pang
B.
Ko
T.-M.
2004
Thermal degradation of the polyimide synthesized from 4,4′-(hexafluoroisopropylidene)diphthalic dianhydride and 4,4′-diaminodiphenylmethane
.
Journal of Applied Polymer Science
91
(
4
),
2295
2301
.
Senthamarai
C.
Kumar
P. S.
Priyadharshini
M.
Vijayalakshmi
P.
Kumar
V. V.
Baskaralingam
P.
Thiruvengadaravi
K. V.
Sivanesan
S.
2013
Adsorption behavior of methylene blue dye onto surface modified Strychnos potatorum seeds
.
Environmental Progress & Sustainable Energy
32
(
3
),
624
632
.
Simsek
G.
Imamoglu
M.
2015
Investigation of equilibrium, kinetic and thermodynamic of methylene blue adsorption onto dehydrated hazelnut husk carbon
.
Desalination and Water Treatment
54
(
6
),
1747
1753
.
Van
H. T.
Nguyen
L. H.
Hoang
T. K.
Tran
T. P.
Vo
A. T.
Pham
T. T.
Nguyen
X. C.
2019
Using FeO-constituted iron slag wastes as heterogeneous catalyst for Fenton and ozonation processes to degrade Reactive Red 24 from aqueous solution
.
Separation and Purification Technology
224
,
431
442
.
Verma
A. K.
Dash
R. R.
Bhunia
P.
2012
A review on chemical coagulation/flocculation technologies for removal of colour from textile wastewaters
.
Journal of Environmental Management
93
(
1
),
154
168
.
Wang
J. Z.
Harrison
M.
2018
Removal of organic micro-pollutants from water by -cyclodextrin triazine polymers
.
Journal of Inclusion Phenomena and Macrocyclic Chemistry
92
(
3–4
),
347
356
.
Wu
D.
Hu
L.
Wang
Y.
Wei
Q.
Yan
L.
Yan
T.
Li
Y.
Du
B.
2018
EDTA modified β-cyclodextrin/chitosan for rapid removal of Pb(II) and acid red from aqueous solution
.
Journal of Colloid and Interface Science
523
,
56
64
.
Xu
H.
Zhou
C.
Jian
C.
Yang
S.
Liu
M.
Huang
X.
Gao
W.
Wu
H.
2020
Salt/current-triggered stabilization of β-cyclodextrins encapsulated host-guest low-molecular-weight gels
.
Chinese Chemical Letters
31
(
2
),
369
372
.
Xue
J.
Liu
W.
Kannan
K.
2017
Bisphenols, benzophenones, and bisphenol A diglycidyl ethers in textiles and infant clothing
.
Environmental Science & Technology
51
(
9
),
5279
5286
.
Yang
K.
Xing
B.
2009
Adsorption of fulvic acid by carbon nanotubes from water
.
Environmental Pollution
157
(
4
),
1095
1100
.
Yuan
X.
Zhuo
S.-P.
Xing
W.
Cui
H.-Y.
Dai
X.-D.
Liu
X.-M.
Yan
Z.-F.
2007
Aqueous dye adsorption on ordered mesoporous carbons
.
Journal of Colloid and Interface Science
310
(
1
),
83
89
.
Zhang
M.
Zhu
L.
He
C.
Xu
X.
Duan
Z.
Liu
S.
Song
M.
Song
S.
Shi
J.
Li
Y. e.
Cao
G.
2019
Adsorption performance and mechanisms of Pb(II), Cd(II), and Mn(II) removal by a β-cyclodextrin derivative
.
Environmental Science and Pollution Research
26
(
5
),
5094
5110
.
Zhao
D.
Zhao
L.
Zhu
C. S.
Shen
X.
Zhang
X.
Sha
B.
2009
Comparative study of polymer containing β-cyclodextrin and -COOH for adsorption toward aniline, 1-naphthylamine and methylene blue
.
Journal of Hazardous Materials
171
(
1–3
),
241
246
.
Zhao
F.
Repo
E.
Yin
D.
Meng
Y.
Jafari
S.
Sillanpää
M.
2015
EDTA-cross-linked β-cyclodextrin: an environmentally friendly bifunctional adsorbent for simultaneous adsorption of metals and cationic dyes
.
Environmental Science & Technology
49
(
17
),
10570
10580
.
Zhou
Y.
Chen
L.
Lu
P.
Tang
X.
Lu
J.
2011
Removal of bisphenol A from aqueous solution using modified fibric peat as a novel biosorbent
.
Separation and Purification Technology
81
(
2
),
184
190
.
Zhou
Y.
Cheng
G.
Chen
K.
Lu
J.
Lei
J.
Pu
S.
2019a
Adsorptive removal of bisphenol A, chloroxylenol, and carbamazepine from water using a novel β-cyclodextrin polymer
.
Ecotoxicology and Environmental Safety
170
,
278
285
.
Zhou
Y.
Lu
J.
Zhou
Y.
Liu
Y.
2019b
Recent advances for dyes removal using novel adsorbents: a review
.
Environmental Pollution
252
,
352
365
.
Zhou
Y.
He
J.
Lu
J.
Liu
Y.
Zhou
Y.
2020
Enhanced removal of bisphenol A by cyclodextrin in photocatalytic systems: degradation intermediates and toxicity evaluation
.
Chinese Chemical Letters
.
https://doi.org/10.1016/j.cclet.2020.02.008
.
Zijian
Z.
Qingwei
G.
Zhencheng
X.
Li
W.
Kai
C.
2015
Distribution and removal of endocrine-disrupting chemicals in industrial wastewater treatment
.
Environmental Engineering Science
32
(
3
),
203
211
.
Zou
W.
Bai
H.
Gao
S.
2012
Competitive adsorption of neutral Red and Cu2+ onto pyrolytic char: isotherm and kinetic study
.
Journal of Chemical & Engineering Data
57
(
10
),
2792
2801
.