Preparation and adsorption properties of a facile solid cucurbit[8]uril-based porous supramolecular assembly

Abstract

Cucurbit[8]uril (Q[8]) was used as the building block for the facile preparation of Q[8]-based supramolecular assemblies in aqueous HCl solutions (A). The Q[8]-based supramolecular assembly was constructed of Q[8] molecules held together through outer surface interactions of cucurbit[n]urils, and hydrogen bonding between latticed water molecules and portal carbonyl oxygens of Q[8]. The porous structure resulted in different adsorption properties for different fluorophore guests (FGs), including dyes and fused-ring compounds, and fluorescence enhancement or colour changes resulted in luminescent materials (FG@As). Moreover, some of the FG@A systems could selectively respond to certain volatile organic compounds, for which they can be used as solid sensors.

Introduction

Cucurbit[n]urils (Q[n]s, n = 5–8, 10) are pumpkin-shaped macrocycles, each consisting of a rigid hydrophobic cavity and two open polar portals lined with carbonyl groups. Recently, our group described the facile preparation of Q[10]-based porous supramolecular assemblies by simply mixing two aqueous HCl solutions containing Q[10] and CdCl₂, respectively. These Q[10]-based assemblies exhibited novel adsorption capacities for certain dyes, and the solid dye@Q[10] assembly displayed intense fluorescence emission. ¹ Over a decade ago, Kim and co-workers described a simple Q[6]-based supramolecular assembly with honeycomb-like channels, permanent porosity, exceptional stability and special capacities for capturing acetylene molecules ² and carbon dioxide. ³ We now know that the outer surface interaction of cucurbit[n]urils, Q[n]s, which is attributed to the positive electro-potential outer surface of Q[n]s, results in the formation of these novel Q[n]-based porous supramolecular assemblies. ⁴ A new field of Q[n]-based supramolecular chemistry in fields of relevance to porous functional materials and light emitters that can be used for detection of explosives, chemical agents, pesticides and other applications has been created. ⁵

It is well known that Q[n]s with an even number of units, such as Q[6], Q[8] and Q[10], are non-functionalized and barely dissolve in neutral water, whereas Q[n]s with odd numbers, such as Q[5] and Q[7], are water soluble. Compared with Q[n]s with odd numbers, Q[n]s with even numbers can more rapidly yield solid Q[n]-based supramolecular assemblies in the absence or presence of inorganic anions or polyaromatic compounds as structure-directing agents.^6–8 For example, we found that Q[10] interacts with metal ions to form one-dimensional coordination polymers in aqueous HCl solutions, which are converted into metal-free Q[10]-based two-dimensional supramolecular assemblies through the outer surface interaction of Q[10]s in aqueous HNO₃ solutions, making Q[10]-based systems suitable as metal-selective materials. ⁹ We also synthesized a series of Q[8]-based porous supramolecular assemblies under different conditions. Among these supramolecular assemblies, two compounds with similar structural features were prepared. One was generated from aqueous HCl solution in the presence of LnCl₃ and CdCl₂, which could form [CdCl₄]^2– anions in aqueous HCl solution (Figure S1a in the Supporting Information), ¹⁰ and the other was generated by crystallization of a ternary mixture of Q[8], CsCl and p-hydroxybenzoic acid in aqueous HCl solution (Figure S1b in the Supporting Information). ¹¹

Ions such as depronated p-hydroxybenzoic acid and [CdCl₄]^2– are so-called structure-directing agents that promote the formation of Q[8]-based three-dimensional (3D) frameworks. However, our recent work revealed that a simple Q[8]-based supramolecular assembly could be obtained from aqueous HCl solutions without using any structure inducer, and the Q[8]-based assembly possesses similar structural features to the two aforementioned Q[8]-based 3D frameworks (Figure S1c in the Supporting Information). Further experiments revealed that all precipitates from HCl solutions, whether microcrystals or single crystals, shared the same structural features, suggesting that the formation of simple Q[8]-based 3D frameworks reflected the intrinsic attributes of the Q[8] molecule.

In the present work, the structural features of this simple Q[8]-based supramolecular assembly (A) were investigated in detail. Given the porous structure of this Q[8]-based 3D framework, 17 selected dyes and fused-ring compounds were tested as fluorophore guests (FGs), and most exhibited fluorescence enhancement or colour changes after being loaded into A, resulting in various solid fluorescent materials (FG@As) that responded to certain volatile organic compounds (VOCs) by a change in fluorescence intensity.

Experimental section

Materials

Q[8] was prepared in our laboratory according to a literature method.^12,13 All other chemicals were of analytical grade, obtained from Aladdin and used as received without further purification.

Preparation of Q[8]-based supramolecular assembly A

Solid Q[8] was dissolved in aqueous HCl solutions (3–8 M). Colourless crystals of A appeared following removal of the aqueous HCl solution by evaporation.

Measurements

All ¹H NMR spectra, including those for titration experiments, were recorded at 20 °C in D₂O on a JEOL JNM-ECZ400s spectrometer.

Single-crystal determination

A suitable single crystal (~0.2 × 0.2 × 0.1 mm³) was coated with paraffin oil and mounted on a Bruker SMART Apex II CCD diffractometer equipped with a graphite-monochromated Mo-K_α radiation source (λ = 0.71073 Å, μ  = 0.828 mm⁻¹) operating in the ω-scan mode and fitted with a nitrogen cold stream (–30 °C). Data were corrected for Lorentz and polarization effects using SAINT, and semi-empirical adsorption corrections based on equivalent reflections were also applied using SADABS. The structure was elucidated by direct methods and refined by the full-matrix least-squares method using F² values with the SHELXS-97 and SHELXL-97 packages, respectively.^14,15 All non-hydrogen atoms were refined anisotropically. Carbon-bound hydrogen atoms were introduced at calculated positions and were treated as riding atoms with an isotropic displacement parameter equal to 1.2 times that of the parent atom. Most of the water molecules in compound A were omitted using the SQUEEZE option of the PLATON program, and 40 water molecules were squeezed for compound A. Analytical expressions for neutral-atom scattering factors were employed, and anomalous dispersion corrections were incorporated. Crystallographic data for A have been deposited at the Cambridge Crystallographic Data Centre (CCDC-1859221). These data can be obtained free of charge through http://www.ccdc.cam.ac.uk/data_request/cif, by emailing [e-mail] or by contacting the Cambridge Crystallographic Data Centre (12 Union Road, Cambridge, CB2 1EZ, UK; Fax: [phone]).

Preparation of FG@A solid-state fluorescent materials

A 1.0 g sample of Q[8]-based A was added to 4.0 mL of 5 × 10⁻³ M FG in acetonitrile, stirred for 1 h and filtered, and the solid materials displaying fluorescence of different colours were collected.

Adsorption and response studies of FG@A with VOCs

The required solid assembly (0.5–1.0 g) contained in a tared open glass vial was added to a sealed glass vessel, which was evacuated using a vacuum pump. Pumping was continued until the sample achieved a constant weight. A second open container containing few millilitre of the volatile liquid (methanol, ethanol, diethyl ether, acetonitrile, acetone, dichloromethane, trichloromethane, or tetrachloromethane) was then added and the vessel resealed. The weight change was measured, and the corresponding solid fluorescence spectra were determined at ~1 to 60 min intervals over several hours to obtain the vapour adsorption profiles.

Measurement of fluorescence spectra of solid FG@A assemblies

Fluorescence spectra of solid FG@A assemblies before and after adsorbing volatile compounds were recorded at 25 °C using a Varian Cary Eclipse spectrofluorometer (Varian, Inc., Palo Alto, CA, USA).

Results and discussion

Description of the crystal structures of compound A

Q[8]-based supramolecular assemblies can be formed easily from Q[8] molecules, or in the presence of inorganic anions or aromatic compounds through the outer surface interaction of cucurbit[n]urils under different conditions.^{10,11,16–19} In the present work, a simple Q[8]-based supramolecular assembly was generated from aqueous HCl solutions (A). In supramolecular assembly A, each Q[8] molecule (the central Q[8]) interacts with four neighbouring Q[8] molecules (Figure 1(a), middle). The portal carbonyls of the central Q[8] were always close to the outer surface of two neighbouring Q[8] (in pink) and in turn, the portal carbonyls of the other two neighbouring Q[8] (in blue) were always close to the outer surface of the central Q[8] and this reciprocal interaction was the main driving force resulting in the formation of Q[8] porous supramolecular assemblies, that is, the outer surface interaction of cucurbit[n]urils was attributed to the positive electro-potential outer surface of cucurbit[n]urils. A close inspection revealed that the detailed interactions mainly include dipole interactions between portal carbonyl oxygens of one Q[8] molecule and positive electro-potential methines and methylenes on the outer surface of a neighbouring Q[8] molecule (Figure 1(a), left and right). In addition, there was hydrogen bonding between the lattice water molecules and portal carbonyl oxygens of neighbouring Q[8] molecules in A (omitted for clarity). The distances between portal carbonyl oxygens of Q[8] molecules and the methine and methylene protons from neighbouring Q[8] molecules were in the range of 2.379–2.677 Å. Figure 1(b)–(e) shows the Q[8] supramolecular assembly with different channels with sections about 21 and 73 Ų, respectively, from different directions. All channels in A were filled with a large number of lattice water molecules that linked Q[8]–Q[8] molecules through hydrogen bonding with portal carbonyl oxygens of Q[8] molecules in A.

Figure 1.

Crystal structures of A (a) showing detailed interactions between the central Q[8] molecule and four neighbouring Q[8] molecules. (b)−(e) Crystal stacking structure in A from different directions.

Powder X-ray diffraction analyses of A and comparison with simulations showed that the bulk of samples essentially consisted of pure crystalline phases (Figure S2 in the Supporting Information). The thermal stability of A was investigated by differential scanning calorimetry (DSC) and thermogravimetry (TG) (Figure S3 in the Supporting Information). For air-dried A, the DSC with TG curves revealed two main stages: water molecules from crystals accounted for ~25% of weight losses (calcd ~36%; from ambient to ~120 °C), and rapid weight losses at ~350 °C, indicating the decomposition of Q[8].

Adsorption properties of Q[8]-based porous supramolecular assemblies

Q[n]-based supramolecular assemblies present adsorbed various dyes and polyaromatic compounds to form solid fluorescent materials.^1,18 In the present case, different channels and spaces were observed in the Q[8]-based porous supramolecular assembly (A). Loading of porous assembly A was tested with 17 luminescent FGs, including anthracene (FG1), rhodamine B (FG2), naphthalene (FG3), naphthol (FG4), 1-naphthaldehyde (FG5), 3,6-diaminoacridine hydrochloride (FG6), levofloxacin (FG7), pyrene (FG8), 9-anthracene-carboxaldehyde (FG9), 4,5-dihydro-3-(4-pyridinyl)-2H-benz[g] (FG10), 2-[4-(dimethylamino)phenyl] (FG11), 1-pyrenecarboxaldehyde (FG12), 2-[(p-(dimethylamino)styryl)]pyridine methiodide (FG13), dansyl chloride (FG14), 8-hydroxyquinaldine (FG15), hydroxycoumarin (FG16) and phenanthrene (FG17). Most of the FG@A products displayed enhanced fluorescence emissions, and/or altered fluorescence colours, with the exception of anthracene (FG1), 9-anthracenecarboxaldehyde (FG9) and 1-pyrenecarboxaldehyde (FG12), which resulted in obvious fluorescence quenching of FG@As. A comparison of the fluorescence spectra of the guest FGs (x = 1−17) and the corresponding luminescent material FGs@A is shown in the figures in the second column in Table S1 in the Supporting Information. For example, loading A with FG8 by immersing A in an acetonitrile solution containing FG8 yielded luminescent material FG8@A that displayed a fluorescence colour change from blue to violet, unlike free FG8 (Figure 2(a)). Meanwhile, loading A with FG12 by immersing A in an acetonitrile solution containing FG12 yielded weakly luminescent material FG12@A that exhibited a ~15-fold more fluorescence quenching than free FG12 (Figure 2(b)). Conversely, loading A with FG15 by immersing A in an acetonitrile solution containing FG15 yielded luminescent material FG15@A that exhibited a ~15-fold increase in fluorescence enhancement compared with free FG15 (Figure 2(c)). The amount of FG8 or other FGs extracted by Q[8]-based assembly A from the solution could be measured using ¹H NMR spectroscopy to compare the integrity before and after extracting the FGs (Figure S4 and Table S2 in the Supporting Information). In general, a change in the fluorescence emission of a compound was a consequence of changes in the molecular structure conformation or the aggregation state. Loading a porous assembly, such as A, with a fluorescent compound could change its conformation or aggregation state, resulting in fluorescence enhancement or quenching. The channels or pores in the Q[8]-based supramolecular assembly in the present work were narrow, which could lead to the transformation of FGs, such as FG2, FG6, FG15, resulting in fluorescence enhancement, or restriction of rotation of the FGs, such as FG1, FG9, FG12, resulting in fluorescence quenching. A more detailed conclusion requires further experimental analysis.

Figure 2.

Comparison of the fluorescence spectra of (a) FG8 and solid FG8@A; (b) FG12 and solid FG12@A; and (c) FG15 and solid FG15@A.

Previous work showed that Q[n]-based supramolecular assemblies could adsorb various VOCs,^20,21 and this could influence the fluorescence properties of the resulting FG@A luminescent materials. Thus, the adsorption of 12 common VOCs was tested, including dichloromethane, trichloromethane, tetrachloromethane, methanol, ethanol, acetone, acetonitrile, diethyl ether, benzene, toluene, formaldehyde and tetrahydrofuran. The results revealed that most of the FG@A systems exhibited responses to certain VOCs through fluorescence quenching or enhancement (Figures S5−S21 in the Supporting Information). For example, FG8@A exhibited little or no response to most VOCs, except formaldehyde, toluene, benzene, as demonstrated by obvious fluorescence quenching (Figure 3(a)). It is interesting that loading A with FG14 resulted in a very weak luminescent compound FG14@A, whereas loading FG14@A with most VOCs resulted in moderate fluorescence enhancements, but a significant fluorescence enhancement for formaldehyde, suggesting that FG14@A possessed high selectivity for formaldehyde (Figure 3(b)).

Figure 3.

Fluorescence spectra of (a) FG8@A and (b) FG14@A saturated with 12 VOCs, respectively.

We have noted that among these FG@A systems are those responding to formaldehyde, polychloromethanes and so on, which are in a list of carcinogens released by the world health organization’s international agency for research on cancer; ²² recently, therefore, more detailed investigations on such systems were performed by a fluorescence change approach to the real-time monitoring of the uptake of the selected VOCs. For example, the titration fluorescent spectra of loading FG8@A with formaldehyde show a rapid decrease of fluorescence strength with increasing adsorption time or the amount of absorbed formaldehyde (Figure 4(a)). Figure 4(b) shows the decrease of fluorescence intensity of solid FG8@A at 392 nm along with increasing adsorption time. The fluorescence strength reached a balance after about 30 min, and 50% of the fluorescence intensity was lost. Figure 4(c) shows the adsorption profile, and the adsorption capacity or saturate adsorption of FG8@A for formaldehyde is up to about 9% (wt%) within 6 h. Obviously, the balance time (30 min) is much shorter than the saturate adsorption time (6 h). The response of luminescent materials FG@As for the VOCs was dependent on the properties of the FGs and VOCs, and the detection limit (DL) of the FG@A for VOCs could show the difference of different luminescent materials for different VOCs. Based on the related data collected from intensity/time curves (Figure 4(b)) and the adsorption/time (Figure 4(c)) using the interpolation method, a plot of ΔI versus the adsorption amount of tetrahydrofuran on solid FG8@A can be drawn (Figure 4(d)) and the DLs were calculated by multiplying the standard derivation of 10 measurements in the absence of formaldehyde by three and dividing by the slope of the linear calibration curve over a low concentration. The DL of solid FG8@A for formaldehyde was about 9.18 × 10⁻⁸ mol g⁻¹. Other selected titration experiments of loading FG8@A with toluene and tetrachloromethane were performed as shown in Figures S22 and S23 in the Supporting Information. The DLs of solid FG8@A for toluene and tetrachloromethane were 7.70 × 10⁻⁸ and 1.86 × 10⁻⁷ mol g⁻¹, respectively.

Figure 4.

(a) Titration fluorescence spectra of the loading of FG8@A with formaldehyde; (b) the change in fluorescence intensity of FG8@A with increasing adsorption time; (c) the adsorption profile of the loading of formaldehyde in FG8@A; and (d) a plot of ΔI versus the amount of formaldehyde adsorbed by solid FG8@A.

On the other hand, the titration fluorescent spectra of solid FG14@A loading with formaldehyde shows a rapid fluorescence enhancement with an increase of the adsorption time (Figure 5(a)). Figure 5(b) shows the change of fluorescence intensity of solid FG14@A with increasing adsorption time, and the fluorescence intensity of solid FG14@A at 470 nm is increased eightfold after 0.5 h, Figure 5(c) shows the adsorption profile of formaldehyde on FG14@A, and the adsorption capacity of FG14@A for formaldehyde is up to 11% (wt%) within 3 h. Based on the data from Figure 5(b) and (c), a plot of ΔI versus the adsorption amount of formaldehyde on solid FG14@A can be drawn (Figure 5(d)) and the DL of solid FG14@A for formaldehyde was thereby determined to be about 2.9 × 10⁻⁸ mol g⁻¹.

Figure 5.

(a) Titration fluorescence spectra of the loading of FG14@A with formaldehyde; (b) the change in fluorescence intensity of FG14@A with increasing adsorption time; (c) the adsorption profile of the loading of formaldehyde by solid FG14@A; and (d) a plot of ΔI versus the amount of formaldehyde adsorbed by solid FG14@A.

In general, a change in the fluorescence emission of a FG is a consequence of changes in molecular structure, conformation or the aggregation state of the guest. Loading of an FG into a porous assembly such as the Q[n]-based porous supramolecular assemblies could alter its conformation or aggregation state, resulting in enhancement or quenching of guest fluorescence. Thus, we loaded A with FG to yield novel luminescent materials. Most of the resulting solid FG@A systems responded to common VOCs, suggesting that the adsorbed VOCs influence the molecular structure, conformation, aggregation state or charge distribution of the loaded FG through intermolecular interactions between the loaded FG, the adsorbed VOC and Q[8] molecules in the assembly A. When the adsorbed VOCs were removed, the fluorescence of solid FG@A was recovered. Analysis of the lifetime of the adsorption capacity or fluorescence emission revealed that the FG8 or FG14@A/VOC systems exhibited reversible adsorption and were relatively stable (Figures S24 and S25 in the Supporting Information). Comparison of the fluorescence changes among selected FG@A systems following adsorption and desorption of selected volatile compounds was performed (Figure S26 in the Supporting Information), but more detailed structural information is needed to elucidate the detailed mechanism, and we are currently attempting to prepare single crystals of FG@A systems.

Out of curiosity, we investigated the adsorption of a neat FG for the volatile compounds, and the results were negative as expected (Figure S27 in the Supporting Information). Thus, the volatile compounds were unable to induce an obvious fluorescence change in the selected FGs.

Conclusion

In summary, we selected Q[8] as a building block and prepared a Q[8]-based porous supramolecular assembly by simple crystallization in aqueous HCl solutions (A). Crystal structure analysis showed that the Q[8]-based assembly contained different porous structural features and were held together by the outer surface interaction of cucurbit[n]urils, in addition to hydrogen bonding between lattice water molecules and portal carbonyl oxygens of Q[8]. Furthermore, the Q[8]-based porous supramolecular assembly (A) exhibited different adsorption properties for various FGs (x = 1−17) and could be used to fabricate solid fluorescent materials (FG@As) with different coloured fluorescence emissions. Moreover, the solid fluorescent materials also adsorbed different common VOCs, accompanied by fluorescence enhancement or quenching. Thus, it is common that loading Q[n]-based porous supramolecular assemblies with various FGs yield solid fluorescent sensors that respond to common VOCs due to different Q[n]s and structural features. This could launch a new field of Q[n]-based supramolecular chemistry to generate porous materials and light emitters that could be used as solid sensors for VOCs. More extensive investigations of these Q[n]-based supramolecular assemblies and their functions are currently underway.

Supplemental Material

Supplementary_File-CHL-19-0072 – Supplemental material for Preparation and adsorption properties of a facile solid cucurbit[8]uril-based porous supramolecular assembly

Supplemental material, Supplementary_File-CHL-19-0072 for Preparation and adsorption properties of a facile solid cucurbit[8]uril-based porous supramolecular assembly by Chunchun Chen, Weiwei Ge, Kai Chen, Zhu Tao, Yunqian Zhang and Qian Jiang Zhu in Journal of Chemical Research