Cyclic Ether Contaminant Removal from Water Using Nonporous Adaptive Pillararene Crystals via Host-Guest Complexation at the Solid-Solution Interface

The removal of soluble cyclic ether contaminants, such as dioxane and THF, produced in industrial chemical processes from water is of great importance for environmental protection and human health. Here we report that nonporous adaptive crystals of perethylated pillar[5]arene (EtP5) and pillar[6]arene (EtP6) work as adsorbents for cyclic ether contaminant removal via host-guest complexation at the solid-solution interface. Nonporous EtP6 crystals have the ability to adsorb dioxane from water with the formation of 1:2 host-guest complex crystals, while EtP5 crystals cannot. However, both guest-free EtP5 and EtP6 crystals remove THF from water with EtP5 having a better capacity. This is because EtP5 forms a 1:2 host-guest complex with THF via host-guest complexation at the solid-solution interface while EtP6 forms a 1:1 host-guest complex with THF. EtP6 also shows the ability to selectively remove dioxane from water even in the presence of THF. Moreover, the reversible transitions between nonporous guest-free EtP5 and EtP6 structures and guest-loaded structures make them highly recyclable.


Introduction
1,4-Dioxane, a cyclic ether often simply called dioxane, is primarily used as a solvent in industry as well as in the laboratory and a stabilizer for the transport of halogenated hydrocarbons [1]. Dioxane is also a by-product of the polyester manufacturing process, leading to its subsequent occurrence in industrial wastewater streams [2,3]. Nevertheless, dioxane is also known as a highly stable contaminant and potential carcinogen in water and is becoming a threat for human and animal health [4][5][6]. There has been severe dioxane pollution in history. During 1976During -1985, leakage of dioxane occurred in Ann Arbor, Michigan, and severely damaged the drinking water [7][8][9]. The removal or degradation of dioxane has not been completed until now. Some efforts have been devoted to increasing control, removal, and remediation of dioxane from sources of pollution. Recent methods involve electrolysis and ozonation [2,10], phytoremediation [11], advanced oxidation processes (AOPS) [12][13][14][15], and so on. However, it is still challenging to completely remove dioxane due to its high miscibility with water, low vapor pressure, and nonbiodegradable nature. Moreover, these current methods are complex, highly energy-consuming, and unrecyclable. Thus, the search for new and easy strategies or adsorbents for adsorption and subsequent removal of dioxane from water is of great importance.
Pillar[ ]arenes are a new and important class of macrocyclic hosts [16,17]. They are highly symmetrical and rigid, easy to chemically modify, and possess abundant host-guest properties [18][19][20][21][22][23][24][25][26]. Recently, our group pioneered research on nonporous adaptive crystals (NACs) of pillararenes [27][28][29][30][31][32]. These nonporous crystals with "intrinsic porosity" can capture specific vaporized guests that have noncovalent interactions with them to form new guest-loaded crystal structures, that is, host-guest chemistry at the solid-gas interface. Based on these unique properties, NACs of pillararenes have been successfully applied in the adsorptive separations of hydrocarbons such as styrene purification and xylene isomer separation [28,30]. However, the hostguest chemistry of NACs at the solid-solution interface still remains unexplored. The development of such properties for pillararene NACs may broaden their applications in more areas such as liquid-phase separation and water treatment.  Herein, we found that NACs of pillararenes worked as adsorbents to remove cyclic ether contaminants, such as dioxane and THF, from water via host-guest complexation at the solid-solution interface. Two easily obtained pillararenes, perethylated pillar [5]arene (EtP5) and pillar [6]arene (EtP6), were selected and used as adsorbents. Guest-free EtP6 crystals were found to have the ability to adsorb dioxane from water while EtP5 crystals cannot. Adsorption of dioxane from water led to a structural transition of EtP6 from a guest-free EtP6 structure (EtP6 ) to a dioxane-loaded 1:2 host-guest complex (2(dioxane)@EtP6, Figure 1). However, both guest-free EtP5 and EtP6 crystals removed THF from water via solid-solution host-guest complexation with EtP5 crystals having a better capacity. That is because guestfree EtP5 crystals (EtP5 ) form a 1:2 host-guest complex with THF (2(THF)@EtP5) at the solid-solution interface while EtP6 crystals only form 1:1 host-guest complex with THF (THF@EtP6). EtP6 also shows the ability to selectively remove dioxane from water even in the presence of THF. Upon removal of guests from the host-guest complex crystals, both EtP5 and EtP6 are transformed back to their original guest-free states and can be recycled many times without degradation.

Preparation of Guest-Free Pillararenes. EtP5 and EtP6
( Figure 1) were synthesized according to previous reports [18,[27][28][29][30][31]. To use EtP5 and EtP6 as adsorbents, guestfree samples of EtP5 and EtP6 were obtained (the detailed method is given in the supplementary file). Powder X-ray diffraction (PXRD) experiments showed that both activated EtP5 and EtP6 were crystalline in the solid state (referred to as EtP5 and EtP6 , respectively). Synchrotron X-ray diffraction experiments were performed to illustrate their single crystal structures. Both EtP5 and EtP6 show rearrangements of the pillar structures and the loss of their cavities ( Figures S6-S9) [30]. Meanwhile, the densely packed arrangement of pillararene units leads to nonporosity of EtP5 and EtP6 as confirmed by N 2 sorption experiments (Figures S10-S11).

Dioxane Removal Experiments.
Despite their nonporosity, we investigated the dioxane adsorption abilities of EtP5 and EtP6 from water, respectively. To do so, dioxane was dissolved in D 2 O (0.600 mL) with a concentration of 0.500 mg mL −1 (5.7 × 10 −3 mmol mL −1 ), and 1.00 mg of waterinsoluble EtP5 and EtP6 crystals were added, respectively. As can be seen from the time-dependent 1 H NMR spectra, the peak related to dioxane barely changed after addition of EtP5 (Figures S14-S15). However, after addition of EtP6 , the peak of dioxane decreased over time and almost completely disappeared after 24 hours (Figure 2(a)). The final concentration of dioxane after adsorption was calculated to be 4.37 × 10 −5 mmol mL −1 , about 130 times lower than the original concentration (Figures 2(b), S12). The adsorption efficiency of dioxane reached 99.2%, indicating the highly efficient adsorption capacity of EtP6 (Figures 2(b), S12).

Side view
Top view
Upon addition of another 5.00 mg of EtP6 into the solution, the final concentration of dioxane was calculated to be 4.37 × 10 −6 mmol mL −1 (0.413 mg L −1 ) after 24 hours ( Figure  S13), which is under the discharge limit for 1,4-dioxane of the Korean Ministry of Environment (5.00 mg L −1 ) [8]. This phenomenon indicated that EtP6 instead of EtP5 can remove dioxane from water effectively.
To understand the adsorption mechanism, both EtP5 and EtP6 were filtered from the dioxane aqueous solutions 24 hours after they were immersed. 1 H NMR spectra in CDCl 3 showed that no new peaks appeared for EtP5 , while a dioxane peak appeared for EtP6 (Figures S16, S20). The amount of dioxane can be calculated as two dioxane molecules per EtP6 molecule (mole/EtP6). Moreover, compared with the 1 H NMR spectrum of dioxane in CDCl 3 , the dioxane peak has no chemical shift change in the presence of EtP6 ( Figure S19). This implies that EtP6 does not have hostguest interactions with dioxane in solution due to the presence of CDCl 3 molecules as competitive guests. However, the weak host-guest interactions between EtP6 and dioxane may emerge at the solid-liquid interface because of the absence of competitive guests, thus facilitating EtP6 crystals to capture dioxane from water. Thermogravimetric (TG) analyses also confirmed the results. There is no apparent weight loss below 400 ∘ C for EtP5 after immersion in the dioxane solution, indicating that no dioxane was adsorbed in EtP5 ( Figure  S17). However, an apparent weight loss (13.1%) below 160 ∘ C for EtP6 occurred after being soaked in the dioxane-water solution, which can also be calculated to be two moles/EtP6 ( Figure S21). These results are thus in accordance with NMR. Powder X-ray diffraction (PXRD) experiments were then performed to monitor the structural information. For EtP5 , the PXRD pattern did not change after immersion in the dioxane solution, meaning no structural transitions ( Figure  S18). For EtP6 , the PXRD pattern after immersion in the dioxane-water solution was different from the original one, but with a reservation of several small original peaks ( Figure 2(c), II). Moreover, the PXRD pattern was completely changed after immersion in a higher dioxane-water solution with a concentration of 1.00 mg mL −1 (Figure 2(c), III). These results indicated the occurrence of structural transitions from guest-free EtP6 to a dioxane-loaded new structure after adsorption of dioxane from water.
To reveal the new structure of EtP6, dioxane-loaded EtP6 single crystals were obtained by a solution-growth method and characterized by X-ray crystallography. To our surprise, in the crystal structure of solution-grown dioxaneloaded EtP6 (4(dioxane)@EtP6, Figure 3(a)), four dioxane molecules correspond to one EtP6 molecule with one located in the cavity and three outside the cavity. Meanwhile, the hexagonal shape of EtP6 is deformed to some extent. The deformed hexagonal pillar structure of EtP6 assembles a window-to-window packing mode, leading to the formation of infinite intrinsic 1D channels with dioxane inside and outside the channels (Figure 3(b), right). It should be worth noting that the ratio of dioxane to EtP6 in the single crystal structure is twice of that obtained from EtP6 capturing dioxane from water. Moreover, the PXRD pattern of EtP6 Research 5 after capturing dioxane from water is totally different from the one simulated from the single crystal structure of 4(dioxane)@EtP6 ( Figure S22I). These results implied that after capturing dioxane from water, EtP6 was transformed into a new structure that is unlike the solution-grown dioxaneloaded EtP6 structure. We then focused on a previously reported cyclohexane (CH)-loaded EtP6 crystal structure (2(CH)@EtP6) with two CH molecules per EtP6 molecule [25], the same ratio as EtP6 after capturing dioxane from water. The PXRD pattern of EtP6 after capturing dioxane matched well with that simulated from 2(CH)@EtP6, manifesting their structural similarities ( Figure S22III). Hence, we conclude that after capturing dioxane, EtP6 was transformed into a honeycomb-like structure with two dioxane molecules located in the cavity of one EtP6 molecule.

Tetrahydrofuran Removal Experiments.
Tetrahydrofuran (THF), another cyclic ether pollutant with a smaller molecular size, is also encountered in many chemical processes [33]. THF can react readily with oxygen to produce an unstable hydroperoxide. Distillation of peroxide containing THF increases the peroxide concentration, resulting in a serious risk of explosion. THF also forms an azeotrope with water and the mixture of THF-water needs separation during the manufacture of THF [34,35]. Although EtP5 cannot remove dioxane from water presumably due to size effect of host-guest complexation at the solid-liquid interface, its potential in the removal of THF was explored. Upon addition of EtP5 crystals (1.00 mg) to 0.600 mL of D 2 O with a THF concentration of 0.500 mg mL −1 (6.90 × 10 −3 mmol mL −1 ), the time-dependent 1 H NMR spectra showed that the peaks of THF decreased over time and almost completely disappeared after 24 hours (Figures S23-S24). The final concentration of THF after adsorption was calculated to be 1.42 × 10 −4 mmol mL −1 , about 49 times lower than the original concentration (Figure 4(a)). Interestingly, EtP6 also showed the ability to remove THF from water (Supplementary Figure  26). However, the final concentration of THF after treatment with EtP6 was 1.43 × 10 −3 mmol mL −1 , much higher than that with EtP5 (Figure 4(a)). Upon addition of another 1.00 mg of EtP5 or EtP6 into the respective solutions, the final concentrations of THF after 24 hours were calculated to be 7.15 × 10 −6 mmol mL −1 and 1.02 × 10 −4 mmol mL −1 ( Figures  S25, S28), respectively. These results indicate that although both EtP5 and EtP6 can remove THF from water, the efficiency of EtP5 (98.0%) was much higher than that of EtP6 (79.5%).
After filtration from the THF-water solution, both EtP5 and EtP6 crystals were characterized by 1 H NMR, TGA, and PXRD. 1 H NMR of both crystals dissolved in CDCl 3 showed clear peaks related to THF (Figures S29, S32). The molar ratios of THF to EtP5 and EtP6 were calculated to be 2:1 and 1:1, respectively. This suggests the reason why EtP5 had a better performance in the THF removal. Similar to the case in the dioxane removal, the THF peaks in the presence of either EtP5 or EtP6 have no chemical shift changes compared with those of single THF in CDCl 3 , indicating the absence of host-guest interactions of THF with either EtP5 or EtP6 in solution (Figures S30-S33). Thus, the weak host-guest interactions that happen at the solid-liquid interface without competitive guests may be the driving force for EtP5 and EtP6 crystals to capture THF in water. TG analyses also showed similar results to that obtained by NMR. The weight loss below 120 ∘ C can also be calculated as 2 and 1 THF molecules per host molecule, respectively (Figures S31-S34). PXRD experiments showed that both EtP5 and EtP6 underwent structural changes after immersion in the THF-water solution. The PXRD pattern of EtP5 was completely changed to a new one and matched the pattern simulated from the single crystal structure of THFloaded EtP5 (2(THF)@EtP5, Figure 4(c)) [28], indicating the structural transition from EtP5 to 2(THF)@EtP5 after adsorption of THF from water. Interestingly, the PXRD pattern of EtP6 after adsorption of THF became similar to that of EtP6 after immersion in the dioxane-water solution (Figure 4(d)), manifesting their structural similarities. Thus, it can be deduced that the THF-loaded EtP6 (THF@EtP6) is also a honeycomb-like structure but with 1:1 rather than 1:2 host-guest complex.

Selective Removal of Dioxane in the Presence of THF.
Since EtP6 can remove dioxane and THF individually from water, we wondered whether it could selectively remove THF or dioxane from an aqueous solution containing both THF and dioxane. Upon addition of EtP6 (5.00 mg) to a THF/dioxane/D 2 O mixture (both the weight concentrations of THF and dioxane were 0.500 mg mL −1 ; the total volume of the mixture was 0.600 mL), the time-dependent 1 H NMR spectra ( Figure S35) showed that the concentration of dioxane decreased over time while THF almost remained the same. After 24 hours, the final concentration of dioxane was calculated to be 0.018 mg mL −1 while the concentration of THF remained as high as 0.480 mg mL −1 ( Figure 5). These results implied that EtP6 can remove dioxane from water even in the presence of THF with high selectivity.

Recyclability.
One shortcoming of common adsorbents is the decreased performance over time due to fouling. In practical use, an adsorbent must be recycled without any degradation. Upon heating to completely remove dioxane guests from 2(dioxane)@EtP6, the PXRD pattern showed that the desolvated 2(dioxane)@EtP6 was transformed back to EtP6 (Figure S43, II and III). Similar phenomena were also observed for 2(THF)@EtP5 and THF@EtP6. PXRD experiments confirmed the complete removal of THF from 2(THF)@EtP5 and THF@EtP6 ( Figures S38, S43), respectively. Furthermore, the recovered EtP5 and EtP6 remove THF and dioxane from water again, respectively, without degradation after recycling five times ( Figure 6). Thus, we can conclude that reversible host-guest complexation at the solid-liquid interface contributes to the recyclability of pillararene crystals.

Discussion
In summary, we found that nonporous adaptive pillararene crystals, EtP5 and EtP6 , can be used as adsorbents to   remove cyclic ethers from water via host-guest complexation at the solid-solution interface. EtP6 crystals have the ability to adsorb dioxane from water while EtP5 crystals cannot. Adsorption of dioxane leads to a structural transition of EtP6 from EtP6 to a 1:2 host-guest complex 2(dioxane)@EtP6. However, both EtP5 and EtP6 crystals remove THF from water via host-guest complexation at the solid-solution interface with EtP5 having a better capacity. This is due to the formation of a 1:2 host-guest complex of EtP5 with THF (2(THF)@EtP5) rather than the 1:1 host-guest complex of EtP6 with THF (THF@EtP6). EtP6 also shows the ability to selectively remove dioxane from water even in the presence of THF. Compared with current methods to remove dioxane and THF, this approach via host-guest recognition at the solid-liquid interface has several advantages such as the simple and cheap synthesis of pillararenes, solution-processability, and high thermal and chemical stability. Moreover, the reversible transformations between nonporous guest-free structures and guest-loaded structures make pillararenes highly recyclable. Future work will try to expand the applications of pillararene crystals via host-guest complexation at the solid-solution interface such as liquid-phase separation. Other types of hosts with the potential to encapsulate guests at the solid-solution interface are worth exploring for more unique applications.

Materials. p-Diethoxybenzene was purchased from JK
Chemicals and used as received. All other chemicals, including tetrahydrofuran (THF) and 1,4-dioxane, were purchased from Sigma-Aldrich and used as received. EtP5 and EtP6 were synthesized as described previously [18]. Desolvated crystalline EtP5 (EtP5 ) was recrystallized from acetone and dried under vacuum at 100 ∘ C overnight. Desolvated crystalline EtP6 (EtP6 ) was recrystallized from acetone and dried under vacuum at 140 ∘ C overnight.

Solution NMR.
Solution 1 H NMR spectra were recorded at 400.13 MHz using a Bruker Avance 400 NMR spectrometer.

Thermogravimetric
Analysis. TGA analysis was carried out using a Q5000IR analyzer (TA instruments) with an automated vertical overhead thermobalance. The samples were heated at the rate of 10 ∘ C/min using N 2 as the protective gas.

Powder X-Ray Diffraction.
PXRD data before and after vapor sorption were collected in a Rigaku Ultimate-IV X-ray diffractometer operating at 40 kV/30 mA using the Cu K line ( = 1.5418Å). Data were measured over the range of 5−40 ∘ in 5 ∘ /min steps over 7 min.

Single Crystal Growth.
Single crystals of dioxaneloaded EtP6 were grown by a slow evaporation method: 5 mg of dry EtP6 powder was put in a small vial where 2 mL of 1,4dioxane was added. The resultant transparent solution was allowed to evaporate slowly to give nice colorless crystals in 2 to 4 days.

Single Crystal X-Ray Diffraction.
Single crystal X-ray data sets were measured on a Rigaku MicroMax-007 HF rotating anode diffractometer (Mo-K radiation, = 0.71073 A, Kappa 4-circle goniometer, Rigaku Saturn724+ detector). Unless stated, solvated single crystals, isolated from the crystallization solvent, were immersed in a protective oil, mounted on a MiTeGen loop, and flash-cooled under a dry nitrogen gas flow. Empirical absorption corrections, using the multiscan method, were performed with the program SADABS [36]. Structures were solved with SHELXD [37] or SHELXT [38] or by direct methods using SHELXS [39], refined by full-matrix least squares on |F| 2 by SHELXL [40], and interfaced through the programme OLEX2 [41]. Unless stated, all non-H-atoms were refined anisotropically, and all H-atoms were fixed in geometrically estimated positions and refined using the riding model.

Data Availability
All data needed to evaluate the conclusions in the paper are present in the paper and the Supplementary Materials. Additional data related to this paper may be requested from the authors.