1. Introduction
Perchlorate (ClO
4−) is a typical oxo-anion contaminant in aquatic environments, with high health risks
[1],
[2],
[3]. Exposure to ClO
4− through mouth-intake-drinking water poses a threat to human health mainly via the inhibition of iodine absorption by the thyroid and has been verified to be responsible for various thyroid diseases
[4],
[5],
[6]. According to epidemiological studies
[7],
[8], the incidence of thyroid nodules in China has increased from 20.9% to 50.4%, which could be partially attributed to frequent exposure to ClO
4− in drinking water. Anthropogenic activities, including fireworks, explosives, and rocket fuel manufacturing, are the predominant sources of ClO
4−, making its contamination of surface water difficult to avoid
[9],
[10]. ClO
4− is distributed globally, with concentrations ranging from 0 to 400 μg·L
−1 [3]. Data show that over 100 000 individuals in the United States live in areas where ClO
4−-related health concerns drinking water
[11]. Thus, the United States was one of the first countries to consider ClO
4− contamination and risk control and set a threshold of 15 μg·L
−1 in drinking water
[12]. In China, the permissible limit for ClO
4− in drinking water was set as 70 μg·L
−1 in 2022
[13]. A recent investigation by Zhang et al.
[14] and Zhang et al.
[15] revealed that central-southern China and the central part of Xinjiang Province were risk hotspots for ClO
4− contamination in China, and the maximum ClO
4− concentration in surface water reached 117 μg·L
−1. Therefore, removing ClO
4− from contaminated source water to achieve drinking water limits is an important issue.
For wastewater containing ClO
4− at a concentration of milligram to gram per liter, anion exchange and bioremediation have been reported as potential techniques; however, these techniques fall short in the realm of drinking water because of their high cost and low efficiency
[16],
[17]. Among the traditional treatment techniques for drinking water, Pincus et al.
[18] reported that adsorption is an attractive option for removing oxo-anions such as ClO
4− because of its high removal efficiency, feasible logistic use, and regeneration potential. However, the concentrations of ClO
4− are much lower than those of the background oxo-anions in source water (such as SO
42−, NO
3−, or HCO
3−), which share similar chemical structures with ClO
4− and significantly compete for adsorption sites
[19],
[20],
[21]. Traditional adsorbents such as active carbon, clay, or zeolite have poor performance in removing ClO
4− from source water because the available adsorption sites can be immediately saturated by competing oxo-anions
[22],
[23],
[24],
[25]. Hence, the development of ClO
4−-specific adsorbents with good selectivity and affinity in water is necessary.
As the molecular topology and geometry are similar between ClO
4− and cooccurring oxo-anions, differences in their chemical behaviors, such as electrostatic interactions, surface complexation, steric hinderance, or Lewis acid/base hardness, need to be exploited to fabricate ClO
4−-specific adsorbents
[26],
[27],
[28],
[29]. In previous studies
[30],
[31], various metal organic frameworks and molecularly imprinted polymers were synthesized as sterically selective adsorbents for ClO
4−. Despite the high efficiency of ClO
4− removal, large-scale manufacturing and engineering applications of these novel adsorbents are needed
[32]. Notably, some studies
[33],
[34],
[35] reported that cationic surfactant-modified adsorbents could selectively adsorb ClO
4− over SO
42−, NO
3−, and HCO
3−. Generally, they attribute the selectivity to hydrophobic effects, as ClO
4− has a low hydration energy, and the cationic surfactant has a long carbon chain with hydrophobicity
[36],
[37],
[38]. However, there is still much uncertainty regarding the interactions of raw adsorbents with cationic surfactants, as well as those of cationic surfactants with ClO
4−. For example, a cationic surfactant, also known as long-chain quaternary ammonium, may act as a hydrogen bond donor to form unconventional CH···O bonds with adsorbents or ClO
4− [39],
[40],
[41], which has generally been overlooked in previous studies. Moreover, the steric hinderance effect caused by the interactions of cationic surfactants with adsorbents may play important roles in ClO
4− removal
[42],
[43]. The mechanisms need to be clarified to better facilitate the selective removal of ClO
4−.
Therefore, in this study, we chose cetyltrimethylammonium bromide (CTAB) and cetylpyridinium chloride (CPC) as cationic surfactants and montmorillonite (MMT) as the raw adsorbent to comprehensively investigate the selective adsorption of ClO4−. The effects of electrostatic attraction, hydrogen bonding, and hydrophobic interactions on ClO4− removal were investigated via experiments, density functional theory (DFT) calculations, and state-of-the-art characterization. Furthermore, the potential of cationic surfactant-modified MMT in engineering applications was evaluated.
2. Materials and method
2.1. Materials
CTAB, CPC, N,N-dimethylhexadecylamine (DMHA), ethyl trimethylammonium iodide (ETY), powdered activated carbon (PAC), kaolin, and sodium perchlorate monohydrate (≥ 98%) were purchased from Aladdin (China), and MMT (1318-93-0) was purchased from Macklin (China).
2.2. Synthesis of cationic surfactant-modified MMT
Predissolved MMT (4 g) and a cationic surfactant (2 g, ∼10 times the cation exchange capacity of MMT) were dissolved in 500 mL of deionized water. After stirring for 24 h at room temperature, the precipitate was collected by centrifugation and washed five times with deionized water. Finally, the solids were dried at 60 °C for 24 h, crushed, and screened to obtain modified MMT with a particle size of 75–150 μm.
2.3. Adsorption experiment of perchlorate
NaClO4 powder was dissolved in deionized water to prepare a NaClO4 stock solution (1000 mg·L−1). Various initial concentrations of NaClO4 solutions were prepared by diluting the stock solution. Manual addition of a set concentration of NaClO4 to an actual water sample was performed to assess the actual treatment capacity of the material. To analyze the impact of the solution pH, experiments were conducted with an initial NaClO4 concentration of 500 μg·L−1. The pH of the solution was initially adjusted to 2 by adding 0.1 mol·L−1 HCl, followed by the addition of 0.1 mol·L−1 NaOH to achieve the desired initial pH value.
After a predetermined time, the solid and liquid phases were separated by centrifugation and filtration, and the liquid phase was filtered through a polyethersulfone syringe filter (0.22 μm) for ion chromatographic determination. All experimental data were averaged in triplicate. Related adsorption modeling and thermodynamic calculations are described in Text S1 in Appendix A.
2.4. Fixed-bed column adsorption experiments
To explore the practical significance of modified MMT for the removal of ClO4−, rapid small-scale column tests (RSSCTs) were conducted. The design and experimental procedures of the chromatographic column are detailed in Text S2 in Appendix A. The initial concentration of ClO4− was controlled at 500 μg·L−1, which is higher than the actual concentration found in ClO4−-contaminated water. Once the fixed-bed columns in the RSSCT were completely saturated with ClO4−, a regeneration step was performed by adjusting the solution pH and employing 0.1 mol·L−1 NaOH as the eluent, which passed through the fixed-bed columns in the opposite direction of the breakthrough experiments. The calculations for the fixed-bed columns are given in Text S3 in Appendix A.
2.5. Characterization and analysis methods
The characterization and analytical methods of the materials can be found in Text S4 in Appendix A.
2.6. DFT Calculations
All calculations were based on DFT and carried out with Gaussian 16 (Gaussian, Inc., USA), Multiwfn
[44], and VMD V1.9.3 (University of Illinois at Urbana–Champaign, USA) software. Detailed information can be found in Text S5 in Appendix A.
3. Results and discussion
3.1. Selective adsorption of ClO4−
Batch experiments were first conducted to remove ClO
4− (100 mg·L
−1) from deionized water via five adsorbents, namely, CTAB-MMT, CPC-MMT, MMT, PAC, and kaolin (
Fig. 1(a) and Fig. S1 in Appendix A). As shown, the control experiment using MMT showed negligible ClO
4− removal capability, while the concentrations of ClO
4− drastically decreased over time for the other four sets of experiments. The pseudo-second-order models were employed to fit the adsorption kinetic data to quantitatively compare the adsorption rates, where the second adsorption rate constants were calculated as 0.0306, 0.0153, 0.0030 and 0.0112 min
−1 for CTAB-MMT, CPC-MMT, PAC, and kaolin, respectively (Table S1 in Appendix A). As shown, the cationic surfactant-modified MMT had faster ClO
4− adsorption rates than the other samples did, indicating their greater ability to facilitate ClO
4− diffusion onto the active adsorption sites. The adsorption isotherms of CTAB-MMT, CPC-MMT, PAC, and kaolin at 25 °C were subsequently plotted and compared in
Fig. 1(b). As shown in
Fig. 1(b) and Fig. S2 in Appendix A, the Langmuir model fits better for ClO
4− adsorption on these adsorbents, indicating monolayer adsorption. In addition, the theoretical adsorption capacities (
qm) for ClO
4− at 25 °C were calculated as 48.32, 39.54, 23.25, and 37.41 mg·g
−1 for CTAB-MMT, CPC-MMT, PAC, and kaolin, respectively (Table S2 in Appendix A). Compared with those of PAC and kaolin, greater adsorption capacities for ClO
4− were observed on cationic surfactant-modified MMT, which was attributed to the involvement of cationic surfactants, while the specific mechanisms involved remain uncertain. CTAB/CPC-MMT ranks high among all the adsorbents tested under similar conditions (Table S3 in Appendix A). Additionally, its use of low-cost raw materials and simple synthesis process makes it a highly promising adsorbent.
Furthermore, to investigate the anti-oxo-anions interference capability of the aforementioned adsorbents, source water from the Xiang River was used as the background. The detailed concentrations of typical anions, including SO
42−, Cl
−, NO
3−, PO
43−, HCO
3−, F
−, and Br
−, are listed in Table S4 in Appendix A. As shown in
Fig. 1(c), PAC and kaolin can hardly remove ClO
4− from source water (less than 15%), despite their high efficiency in deionized water, indicating their poor selectivity for ClO
4− removal. For CTAB-MMT and CPC-MMT, the presence of anions slightly inhibited ClO
4− adsorption, where 100% of the ClO
4− (initial concentration of 500 μg·L
−1) was removed from the source water. The results suggested that CTAB-MMT and CPC-MMT showed significantly higher selectivity to ClO
4− than PAC and kaolin. To further confirm the selectivity and affinity of cationic surfactant-modified MMT toward ClO
4−, experiments using individual and mixed oxo-anion solutions were carried out. As shown in
Fig. 1(d), Figs. S3 and S4 in Appendix A, CTAB/CPC-MMT had the highest affinity for ClO
4− (individual solutions) and satisfactory selectivity for ClO
4− (mixed solutions) with selectivity coefficient for the target anion (
βt) values of 3–35.
3.2. Selective adsorption mechanisms of ClO4−
3.2.1. Role of electrostatic attraction
Then, the mechanisms for the increased affinity and selectivity for ClO
4− caused by cationic surfactants are clarified. Electrostatic attraction is the most common adsorption mechanism for oxo-anions by adsorbents. The surface charge of an adsorbent is negative when the solution pH is higher than the isoelectric point, and vice versa. As shown in Figs. S5 and S6 in Appendix A, the isoelectric points of CTAB-MMT, CPC-MMT, PAC, and kaolin were 5.23, 5.67, 4.76, and 7.55, respectively. The impact of the solution pH on ClO
4− removal was subsequently investigated in the pH range of 2.0–11.0 (
Fig. 2). As shown in
Figs. 2(c)–(d), ClO
4− could be removed by PAC and kaolin only at pH values less than their isoelectric points, indicating that ClO
4− was removed mainly
via electrostatic attraction. This also explained the phenomena shown in
Fig. 1(c), where PAC and kaolin showed poor selectivity toward ClO
4− in source water. As shown in
Figs. 2(a)–(b), although electrostatic attraction plays a certain role in ClO
4− removal (pH 2.0–5.0), CTAB/CPC-MMT could maintain a high removal efficiency for ClO
4− at pH values higher than the isoelectric points. Therefore, further investigations are needed to determine the major adsorption mechanism for ClO
4−, which appears to have no pH dependence (pH 5.0–11.0). Previous studies
[34],
[35] in which cationic surfactant-modified adsorbents were used to remove ClO
4− attributed their performance to their hydrophobicity. Gibb and Gibb
[45] provided direct calorimetric and spectroscopic evidence that ClO
4− binds to hydrophobic concave surfaces. Notwithstanding, they also pinpointed that the hydrophobic interactions are weak, which is possibly not the primary driving force for the ClO
4− removal.
3.2.2. Unconventional hydrogen bonding interactions (CH···O) for ClO4−
Some studies
[39],
[46] reported that high affinity and selective binding to oxo-anions could be achieved via the synergism of hydrophobic interactions and hydrogen bonding. As conventional hydrogen bonds for oxo-anions are formed as NH···O or OH···O, researchers have usually overlooked the potential role of unconventional CH···O, relating ClO
4− to the head groups of cationic surfactants (
Fig. 3(a)). Recently, unconventional C–H hydrogen bonds (N
+–C–H hydrogen bonds) have been explored and shown to interact with oxo-anions, including ClO
4−, of which Samanta et al.
[46] synthesized tripodal organic cages with Csp3–H bonds to selectively remove ClO
4− from organic solvents/water. These studies inspired us to investigate the role of unconventional CH···O when ClO
4− is removed by cationic surfactant-modified adsorbents.
To confirm the feasibility of C–H in quaternary ammonium as a hydrogen bond donor, the molecular electrostatic potentials of CTAB and CPC were first mapped, as shown in
Fig. 3(b) and Fig. S7(a) in Appendix A. As shown, the carbon atoms in the head groups of CTAB/CPC are the most electron-deficient sites with highly polarized characteristics, indicating that they are the most likely sites to act as hydrogen bond donors in CTAB and CPC molecules. The interaction region indicators were subsequently plotted to reveal the intermolecular interactions between ClO
4− and CTAB/CPC. As presented in
Fig. 3(c) and Fig. S7(b), notable attractions (i.e., hydrogen bonds) were observed between the hydrogen atoms in the head groups of CTAB/CPC and the oxygen atoms in ClO
4−, whereas other atoms in CTAB/CPC could not interact with ClO
4−. These results indicate that quaternary nitrogen (N
+) has a significant influence on the electron-deficient characteristics of its linked carbon atoms. To verify the role of quaternary nitrogen, we chose DMHA as a control reagent, which shares the same carbon chain with CTAB. As shown in
Fig. 3(d), we calculated the interaction energies of CTAB-ClO
4−, CPC-ClO
4−, and DMHA-ClO
4− as 78.21, 76.80, and 4.33 kcal·mol
−1, respectively. These results highlight the importance of quaternary nitrogen groups and suggest that interactions between tertiary amine-linked methyl groups and ClO
4− are too weak to form stable hydrogen bonds. Then, we calculated the bond energies of conventional hydrogen bonds, using CH
3NH
2 and CH
3OH as typical hydrogen bond donors, which are 6.80 and 10.42 kcal·mol
−1, respectively. Thus, the strong affinity and selectivity for ClO
4− were confirmed for the C–H bonds adjacent to the quaternary nitrogen. Compared to the relatively weak hydrophobic interaction, unconventional hydrogen bonding could be the primary driving force for ClO
4− removal. CTAB has more polar C–H bonds influenced by N
+ than CPC does, likely explaining the superior adsorption of CTAB-MMT over CPC-MMT.
3.2.3. Role of the hydrophobic cavity in sheltering hydrogen bonds
Although hydrogen bonds between CTAB/CPC and ClO
4− have been verified as the primary driving force for ClO
4− removal, CTAB/CPC alone cannot remove ClO
4− (Fig. S8 in Appendix A). As the used concentrations of CTAB/CPC are much lower than their critical micelle concentrations, the CTAB/CPC molecules are free in an aqueous solution and exposed to large amounts of H
2O molecules. The competition of H
2O molecules (55.5 mol·L
−1) for hydrogen bonding sites drastically inhibited ClO
4− (μmol·L
−1) removal by CTAB/CPC alone. Hence, it is necessary to protect the hydrogen binding sites from water when removing ClO
4−. Herein, we take CTAB-MMT as an example to discuss its adsorption mechanism. As illustrated in
Fig. 3(a), the head groups (hydrophilic region) of CTAB are bound to the oxygen-containing moieties on the MMT surface (via hydrogen bonds, detailed discussions are carried out in the next section). The long-chain tail groups (hydrophobic region) of CTAB interact with each other to form hydrophobic cavities, allowing only oxo-anions with low hydration energy to enter. Therefore, ClO
4− can enter hydrophobic cavities and bind to sheltered unconventional C–H bonds. To confirm this scenario, we used CTAB-MMT to remove SO
42−, Cl
−, NO
3−, I
−, PF
6−, and BF
4−. As shown in
Fig. 4(a), the removal efficiency of the anions followed the order of ClO
4−, PF
6−, BF
4−, I
−, NO
3−, Cl
−, and SO
42−, which is consistent with the order of the Hofmeister Series
[47],
[48],
[49]. The more hydrophobic the anions are, the greater the degree of removal achieved by the CTAB-MMT.
Then, we compared the removal efficiency for ClO
4− using DMHA-MMT and ETY-MMT, where DMHA has the same hydrophobic group and ETY has the same hydrophilic group as CTAB. As shown, no ClO
4− removal was observed (
Fig. 4(b)), further confirming that efficient ClO
4− removal by CTAB-MMT was attributed to the synergism of hydrophobic and hydrophilic groups. Overall, for ClO
4− removal by CTAB-MMT, the affinity and selectivity are attributed to the hydrophobic cavities, and the adsorption capability is dependent on hydrogen bonding.
3.3. Characterization of the prepared adsorbents
As the selective adsorption mechanisms of ClO
4− by cationic surfactant-modified MMT have been clarified, characterization of CTAB-MMT and CPC-MMT was conducted. The layered porous structures of MMT could be clearly observed from transmission electron microscopy (TEM; Fig. S9 in Appendix A) and scanning electron microscopy (SEM; Fig. S10 in Appendix A) images. As shown in Fig. S11 in Appendix A, dense and abundant CTAB/CPC were uniformly dispersed on the MMT. Then, the increase in contact angle observed in Fig. S12 in Appendix A indicated the augmented hydrophobicity of MMT following modification with CTAB/CPC. The thermogravimetric (TG) analysis presented in Fig. S13 in Appendix A indicates that the mass loss observed in the modified MMT is likely due to the thermal decomposition of surface quaternary ammonium salts. Furthermore, X-ray diffraction (XRD) analysis was conducted, and the characteristic peaks shifted from 6.8° to 1.8° and 1.5° when CTAB and CPC were intercalated, respectively (Fig. S14 in Appendix A). The shifts suggested significant expansions in the interlayer spacing, increasing from 1.46 to 4.7 and 5.1 nm (Fig. S14), which coincided well with the double chain lengths of CTAB and CPC
[50],
[51]. This confirmed the arrangement of CTAB/CPC proposed in
Fig. 3(a).
MMT is a hydrous alumino silicate, with large amounts of Si–O and Al–O bonds exposed on its surface, which could serve as hydrogen bonding acceptors. The potential of CTAB/CPC to act as hydrogen bonding donors has been verified; hence, the interactions of CTAB/CPC with MMT were investigated.
Fig. 5(a) presents the Fourier transform infrared (FT-IR) spectra of CTAB, CTAB-MMT, CPC, and CPC-MMT. The vibration bands of CTAB and CPC at 3029.10 and 3009.04 cm
−1 are due to the stretching of quaternary ammonium groups. When CTAB/CPC was linked with MMT, noticeable blueshifts were observed, with peaks at 3050.07 and 3046.89 cm
−1, indicating the formation of hydrogen bonds between CTAB/CPC and MMT. To further corroborate the existence of hydrogen bonds,
1H nuclear magnetic resonance (
1H NMR) analysis of CTAB, CTAB-MMT, CPC, and CPC-MMT was performed. As shown in
Fig. 5(b), the signals of the protons in the head groups of CTAB and CPC were assigned at 3.49 and 7.14 and shifted drastically, with △
δ values of 1.19 and 1.37, confirming the formation of CH···O hydrogen bonds.
3.4. ClO4− removal from actual water via CTAB/CPC-MMT
Cyclic adsorption experiments were conducted to evaluate the regeneration ability of CTAB/CPC-MMT, using surface water from the Xiang River (China) as a background. As shown in
Figs. 6(a) and Fig. S15(a) in Appendix A, the ClO
4− removal efficiency remained at 80.7%/75.8% after 20 cycles, with the initial ClO
4− concentration set at 500 μg·L
−1. Additionally, CTAB/CPC-MMT showed high cyclic stability, with a weight loss of 8.5%/9.2% after 20 cycles of reuse. Taking CTAB-MMT as an example, the SEM images and FT-IR spectra are consistent with the initial results (Figs. S16 and S17 in Appendix A). The results indicated that CTAB/CPC-MMT could exhibit good reusability for long-term use in engineering applications. To further evaluate their potential in application scenarios, fixed-bed column experiments were conducted, and the column parameters are listed in Table S5 in Appendix A. As shown in
Fig. 6(b), with an initial ClO
4− concentration of 500 μg·L
−1, 2900 bed volume (BV) solutions could be effectively treated to below the permissible limit in drinking water (CTAB-MMT: 70 μg·L
−1), and the saturated adsorption capacity (
qf) for ClO
4− was estimated to be 13.34 mg·g
−1 (
Fig. 6(b) and Table S6 in Appendix A), with a value of 6500 BV. In addition, regeneration of the fixed-bed column was carried out, as illustrated in the inset figure of
Fig. 6(b). High desorption efficiency was achieved at 20 BV, with a sharp desorption peak. Almost all the ClO
4− was desorbed via 330 BV NaOH eluents (0.1 mol·L
−1), with an enrichment factor of 10.3. The adsorption and desorption profiles of the CPC-MMT were similar to those of the CTAB-MMT (Fig. S15(b)). Overall, the considerable adsorption and regeneration capabilities of CTAB/CPC-MMT make it a promising adsorbent for use in drinking water treatment plants.
4. Conclusions
In summary, we validated the affinity and selectivity of cationic surfactant-modified MMT toward ClO4−, as well as its high adsorption capacity and fast adsorption kinetics. The wide working pH range along with the isoelectric points excluded the main role of electrostatic attraction in ClO4− adsorption. DFT calculations revealed that the C–H adjacent to quaternary ammonium could act as a hydrogen bonding donor to form unconventional CH···O hydrogen bonds, which have much higher bond energies than conventional bonds. The anion removal efficiency obeyed that of the Hofmeister Series, as did the poor performance of ClO4− removal by short-chain quaternary ammonium, providing evidence that hydrophobic interactions played a synergistic role with unconventional CH···O hydrogen bonds. That is, the hydrophobic cavities sheltered unconventional hydrogen bonds to selectively interact with ClO4−. We also demonstrated the application potential of CTAB/CPC-MMT in real-world scenarios via fixed-bed column experiments.
Currently, numerous studies have focused on the selective adsorption of oxo-anions, which is an emerging necessity in drinking water treatment involving the design, screening, and assessment of adsorbents. The most commonly used mechanisms for selective adsorption are steric interactions, surface complexation, and hard soft acid base principles, leading to the fabrication of state-of-the-art materials. Our present study aims to prepare low-cost selective adsorbents for ClO4− by anchoring unconventional CH···O hydrogen bonds in hydrophobic cavities. The adsorbents can be prepared and used on a large scale. In addition to ClO4−, the as-prepared adsorbents, such as HCrO4−, HSeO3−, and so forth, could show satisfactory performance in terms of selectivity and affinity toward other oxo-anions with low hydration energies. Generally, according to the proposed mechanisms, we can incorporate molecular designs into adsorbents for clean water.
CRediT authorship contribution statement
Jian Ao: Writing – original draft, Software, Resources, Investigation, Formal analysis, Conceptualization. Lingjun Bu: Writing – review & editing, Writing – original draft, Visualization, Software, Resources. Yangtao Wu: Writing – review & editing, Supervision, Software, Data curation. Jinming Luo: Writing – review & editing, Software. Shiqing Zhou: Writing – review & editing, Writing – original draft, Software, Funding acquisition.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The work was financially supported by the National Key Research and Development Program of China (2023YFC3207904).
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2024.12.029.
PDF
(1797KB)
