Autocatalytic synthesis of molecular-bridged silica aerogels with excellent absorption and super elasticity

Abstract

In this study, molecular-bridged silica aerogels (MBSAs) were synthesized via supercritical CO₂ liquid drying of bridged 3-aminopropyltriethoxysilane (APTES) wet gels, which were autocatalytic gelled by the catalyst-free bridging of APTES and 1,4-phthalaldehyde via a Schiff base condensation. The detailed mechanism of this facile, one-pot and autocatalytic approach for the production of MBSAs was revealed by in situ¹H NMR and FTIR measurements, and the critical conditions for the successful autocatalytic gelation of the bridged-APTES were investigated. The densities and average pore diameters were in the range from 0.0535 to 0.0702 g cm⁻³ and from 12.9 to 14.1 nm, respectively. Moreover, the morphologies and thermal stability of the products were characterized by SEM and TGA, respectively. Interestingly, the MBSAs show excellent absorption performance for organic solvents, e.g., hexane, dichloromethane, kerosene, dimethylsulfoxide, tetrahydrofuran, ethanol, ethylsilicate, and N,N-dimethylformamide. Among those, 2437% by mass absorption of dichloromethane was achieved. Moreover, the absorption property of the aerogel was stable, with nearly no deterioration even after 30 absorption–desorption cycles. In addition, the aerogel showed remarkable flexibility, with the highest deformation of 90% and complete recovery after the release of stress, and it can be under through 30 cycles of compression–reversion process with 40% deformation.

---

Introduction

Aerogels are highly porous materials that are synthesized by replacing the liquids within a wet-gel with gas while keeping the networks of the gel intact or without significant collapses.¹ Silica aerogels are one of the most investigated aerogels due to their exclusive properties and versatility in chemistry. For instance, the thermal conductivity of silica aerogels is as low as 10 mW mK⁻¹, which make them a prime candidate for a number of thermal insulation applications such as battery pack insulation on the Mars Rover.² The low density (down to 0.003 g cm⁻³) and high porosity (∼99%) of silica aerogels make them suitable for the collection of hyper-velocity particles in space³ or as absorbents for oil and liquids,^4–6 as catalyst supports,^7–9 and for use in storage media.¹⁰ They also have a low dielectric constant (1.1–2.2)¹¹ and index of refraction (∼1.05).¹² All these properties ensure silica aerogels attract great interests from both the academic and industrial fields.

However, the practical application of aerogels has been restricted due to problems such as fragility and multifarious preparation process. To date, efforts aiming to improve the mechanical properties of silica aerogels have been dedicated via different methods; for example, by constructing a dual network between a nanoscale bacterial cellulose network and a silica gel skeleton, double network aerogels with high strength have been prepared.¹³ Nguyen et al. incorporated an organic linking group into the underlying silica structure to obtain styrene cross-linked silica aerogels, which could nearly recover 100% of their length after compression of the samples to 25%.¹⁴ Meador et al. synthesized cross-linked silica aerogels with organic groups and obtained aerogels with a strength improvement of 2 orders of magnitude over the uncross-linked aerogels.¹⁵ Organic–inorganic composite aerogels produced by linking silica skeletons with organic molecules have been reported by Leventis et al., and the strength of the resulting composite aerogels showed a 2 orders of magnitude increase.¹⁶ However, the abovementioned methods to synthesize flexible aerogels have some disadvantages such as strict synthesis conditions with N₂ protection and the need for a catalyst and high temperature. Moreover, the densities of the aerogels were relatively high due to the incorporation of a large amount of compounds, and furthermore, the compression capacity remains to be improved.

Bridged-silsesquioxane aerogels (BSQs) are a family of inorganic–organic hybrid materials derived from bridged silane precursors, which exhibit distinct properties from traditional silica aerogels.^17–19 Variable organic bridges provide BSQs with tuneable chemical compositions and physical properties. For instance, a BSQ from (3-mercaptopropyl)trimethoxysilane (MPTMS) and vinyltri-methoxysilane (VTMS) via a UV-initiated thiolene click reaction was prepared by Wang et al.,²⁰ and showed an excellent mechanical robustness due to the flexible thioether bridge. Because of the stiff bridge, the phenylene-bridged BSQ synthesized by Boday et al. had a flexural strength of 285 kPa, which is nine-fold stronger than that of untreated hexylene-bridged aerogels and twice as strong as a silica aerogel with comparable densities.²¹

Benefiting from the merits of Schiff base chemistry, such as mild reaction conditions and high reaction rate, in this study, a facile method to synthesize molecular-bridged silica aerogels was proposed based on the catalyst-free bridging of APTES by 1,4-phthalaldehyde. These were then autocatalytically polycondensed in various organic solvents to form molecular-bridged silica gels (MBSGs). Compared to the traditional methods of synthesizing silica aerogels,^22,23 which usually involve hydrolysis by an acid catalyst and condensation by a base catalyst, the molecular-bridged gels reported in this study were synthesized without any catalysts (acid, base, or light), and the entire process was carried out under mild conditions. The MBSG were then dried in supercritical carbon oxide to produce a series of MBSAs. The mechanism of the autocatalytic gelation of the molecular-bridged-APTES will be investigated. The structure, morphology, thermal stability, and mechanical property, as well as the absorption performance of the MBSAs will also be reported.

Experimental

Materials

1,4-Phthalaldehyde (98%) was purchased from Aladdin Reagent (Shanghai) and subsequently purified by silica gel column chromatography using CH₂Cl₂/hexane (v/v 1 : 4) as the eluent. APTES was purchased from Aladdin Reagent (Shanghai) and used without further purification. Other reagents were of analytical grade and used as received.

Preparation of the molecular-bridged silica gels (MBSGs) and aerogels (MBSAs)

MBSGs can be formed in various solvents, e.g. ethanol, DMF, and DMSO, and with different monomer concentrations and molar ratios of APTES and 1,4-phthalaldehyde. Taking the reaction in ethanol as an example, 1,4-phthalaldehyde (58.5 mg, 0.436 mmol) was dissolved in ethanol (2.5 ml) by stirring for 10 minutes, then APTES (241 mg, 1.09 mmol) was added dropwise into the previous solution under continuous stirring. Then, the mixture was cured at 40 °C for several days. Solvent exchange was performed at room temperature to remove water and residual chemicals in the wet-gel (4 cycles, 12 h per cycle). Subsequently, supercritical drying (SCD) was performed to dry the wet gels.

Instrumentation

The pore size distributions and the average pore diameters of the aerogels were analyzed by the BJH (Barrett–Joyner–Halenda) nitrogen adsorption and desorption method (ASAP 2020, Micromeritics, Instrument Corp., USA). The surface areas were determined by the Brunauer–Emmett–Teller (BET) method, based on the amount of N₂ adsorbed at pressures of 0.05 < P/P₀ < 0.3. The cumulative pore volume was measured at the point P/P₀ = 0.99. The micro-structural studies of the aerogel were performed using field-emission scanning electron microscopy (Quanta 400 FEG); the sample was coated with Au nano-powder under a current of 30 mA for 2 min. Fourier transform infrared (FTIR) spectrum was obtained on a Nicolet 6700 spectrometer between 4000 and 400 cm⁻¹, using KBr pellets for all the samples. ¹H NMR spectra were obtained on a VNMRS 400 MHz spectrometer (Varian, Inc., Palo Alto, CA), DMSO-d₆ was used as the solvent. Thermogravimetric analysis (TGA) was carried out using a TG 209F1 Libra (NETZSCH) analyzer with a heating rate of 10 °C min⁻¹, from room temperature to 900 °C in a nitrogen atmosphere. The compressive stress–strain measurements were performed on the elastic MBSA using a tensile testing machine (Instron 3365) at a crosshead speed of 5 mm min⁻¹. The contact angle was performed on an OCA20 (Date Physics Company, Germany). Measurement of the mercury intrusion porosimetry (MIP, AutoPore IV9500, Micromeritics Instrument Corp., USA) was carried out to examine the porosity and pore volume of the MBSAs.

Results and discussion

To figure out the optimal gelation conditions of MBSGs, various parameters, including the molar ratio of APTES to 1,4-phthalaldehyde, solvents, temperature, and pH values, were investigated. As shown in Table 1, when APTES was used as the only precursor (entry 1), no gelation occurred in a wide range of pH values, even by the two-step hydrolysis and polycondensation process,²⁴ and there was change for the solvent with water, ethanol, or DMF; this may possibly be due to the large steric hindrance of the alkyl chains.²⁵ Interestingly, by introducing the 1,4-phthalaldehyde as molecular bridges, gels can be formed without adding any catalyst (acid or base) at temperatures ranging from room temperature to 80 °C. This may be due to the formation of hexaalkoxysilanes by covalently bridging two APTES molecules with one 1,4-phthalaldehyde molecule via the Schiff base reaction, therefore gelation occurs based on the newly formed hexaalkoxysilanes (as illustrated in Fig. 1b and as evidence by the in situ¹H NMR and FTIR, which will be discussed in detail later on).²⁶ More interestingly, gels could not be formed even if 1,4-phthalaldehyde was added when the feed molar ratio of 1,4-phthalaldehyde to APTES was 1 : 1.5 (entry 2). This may be explained by the fact that 1,4-phthalaldehyde was in excess in this case, and the newly formed hexaalkoxysilanes cannot be gelled in the presence of 1,4-phthalaldehyde or without any base. However, in the case of entries 3 and 4, in which –NH₂ is in excess or equal to –CHO, gels are formed, possibly due to autocatalytic hydrolysis and condensation of the hexaalkoxysilanes in the presence of APTES.

Table 1 Different gelation conditions of APTES in ethanol

Entry	Molar ratio (APTES : 1,4-phthalaldehyde)	pH	Gelation results^a
a N means gelation did not occur, and G means gels are formed. b pH values ranged from 1 to 11 in an aqueous solution and ethanol, DMF. c No acids or bases were added – the solution was basic due to the presence of APTES.
1	1 : 0	1–11^b	N
2	1.5 : 1	—^c	N
3	2 : 1	—^c	G
4	2.5 : 1	—^c	G

---

Fig. 1 (a) Illustration of the synthetic process of MBSAs, (b) the proposed mechanism for the chemical reactions occurring during the gelation.

Fig. 1 shows a schematic of the synthesis of the MBSAs as well as the proposed reactions and chemical structure of the MBSAs. Based on the previous study discussed above, a series of MBSAs were prepared with different monomer concentrations, while keeping the molar ratio of APTES to 1,4-phthalaldehyde constant at 2.5 : 1 (see Table 2). Combining two classes of monomers in ethanol at 40 °C led to ivory gels, in the density range of 0.0535–0.0702 g cm⁻³. It can be noted that the polymerization was achieved under mild conditions and without use of any catalysts. Furthermore, heating the precursor solution at 40 °C can accelerate the gelation process.

Table 2 Surface area, pore volume, porosity and average pore diameter values for the MBSAs

Density (g cm^−3)	BET surface area (m^2 g^−1)	Average pore diameter (nm)	Pore volume (cm^3 g^−1)	Porosity (%)
a N means a sample with a density of 0.0702 g cm^−3, which is heated at 80 °C after the absorption of hexane.
0.0535	38.7	14.1	10.11	90.1
0.0662	42.2	13.3	10.54	82.2
0.0702	57.9	13.6	10.52	75.7
0.0702^a	45.4	12.9	7.83	74.9

---

The proposed mechanism for the gelation of APTES and 1,4-phthalaldehyde deduced from the FTIR and ¹H NMR results is presented in Fig. 1b. The FTIR spectra of the MBSAs are displayed in Fig. 2a. The stretching vibration associated to C=O at 1700 cm⁻¹ can be clearly observed in Fig. 2a-1 (1,4-phthalaldehyde).²⁷ The spectra of MBSAs (Fig. 2a-3 to a-5) all exhibited a strong imine (C=N) stretch at 1640 cm⁻¹, which demonstrated the formation of imine bonds, and that a bridged silica aerogel containing C=N bonds was obtained. On the other hand, the carbonyl (C=O) stretch peak located at 1700 cm⁻¹ cannot be observed, which indicated that the aldehyde groups were all consumed, thus showing the high reactivity between the –CHO and –NH₃ groups. The Si–O–Si stretching vibration peaks derived from APTES can also be observed at 1130 and 1040 cm⁻¹,^28,29 which was similar to that of silica aerogels prepared by the traditional sol–gel polymerization process. The peaks located at 2930–2830 cm⁻¹ were attributed to saturated hydrocarbons on the bridge chains.³⁰ These results suggest that the MBSAs are connected by the molecular bridges due to the formation of imine bonds between APTES and 1,4-phthalaldehyde.

Fig. 2 (a) FTIR spectra: (1) 1,4-phthalaldehyde, (2) 3-aminopropyltriethoxysilane, (3–5) different densities of the MBSAs (from 0.0535 to 0.0702 g cm⁻³), (b) TGA curves of the MBSAs with different densities.

Though the bridged structure can be confirmed by the FTIR results, the detailed process of the gel forming was studied by ¹H NMR in order to further confirm the formation of hexaalkoxysilanes (imine bonds) before gelation. The measurement was carried out in situ soon after dissolving a predetermined amount of APTES to 1,4-phthalaldehyde (2.5 : 1 in molar ratio) in DMSO-d₆ (Fig. 3), in which gelation does not occur (gelation start after 3 h). Compared to the bridged-silsesquioxane precursor, the ¹H NMR spectra of APTES (Fig. 3b) and 1,4-phthalaldehyde (Fig. 3c) were also investigated in DMSO-d₆. As shown in Fig. 3c, the chemical shifts of the protons (a and b) from 1,4-phthalaldehyde appear at 8.09 and 10.12 ppm and that of APETS are clearly identified and indicated in Fig. 3b as c, d, e, f, g, and h. Evidently, the peaks corresponding to 1,4-phthalaldehyde (8.09 and 10.12 ppm) cannot be observed in Fig. 3a, instead a series of peaks appear at 7.77 and 8.34 ppm, which indicated that all the 1,4-phthalaldehyde were consumed. The production of hexaalkoxysilanes was confirmed by the ¹H NMR, as shown in Fig. 3a, and all the signals of the protons can be identified (from a′ to g′), which are different from that of pure APTES and 1,4-phthalaldehyde. However, due to the limited reaction time, a by-product was formed by the reaction of one 1,4-phthalaldehyde and one APTES molecule, the signals can also be indentified in Fig. 3a as indicated by a′′ (7.95 ppm, Ph-H), b′′ (8.42 ppm, imine bonds N=C–H), and c′′ (10.03 ppm, –CHO). Moreover, the signals of APTES can also be observed (c, 3.41 ppm, corresponding to N–CH₂–C–), though most of its signals are overlapped by that of hexaalkoxysilanes, which can be explained by the fact that APTES was in excess according to 1,4-phthalaldehyde. These results indicate that hexaalkoxysilanes was formed immediately after the mixing of APTES and 1,4-phthalaldehyde due to the high reactivity of the –NH₂ and Ph–CHO groups. Based on the ¹H NMR results and the FTIR results, as well as previous studies,²⁶ the mechanism of the gelation of APTES and 1,4-phthalaldehyde was proposed and is illustrated in Fig. 1b.

Fig. 3 ¹H NMR spectra (a) the precursor of MBSA with a density of 0.0535 (g cm⁻³), (b) APTES, (c) 1,4-phthalaldehyde.

The density of MBSA increases from 0.0535 g cm⁻³ to 0.0702 g cm⁻³ (see Table 2) as the precursor concentration varies from 120 to 180 mg ml⁻¹. These values are much lower than that of most aerogels dried by SCD (with typical densities in the range of 0.11–0.16 g cm⁻³ according to Martín).³¹ The morphologies of the MBSA frameworks were investigated by SEM, as shown in Fig. 4. It was observed that all the frameworks of the MBSAs are sparsely stacked by irregularly shaped sub-micrometer blocks that are composed of interconnected particles. Moreover, the existence of macropores can also be observed. The formation of the macropores may result from two aspects, as they have also been observed in other studies of flexible silica aerogels:¹⁷ (1) the sol–gel process, which involve the aggregation of nano-sized silica sol to form larger sized particles, resulting in mesopores. Then, the interconnection of the large-sized particles to form the gel networks, resulting in macropores; (2) the relative low solubility of the hexaalkoxysilanes in ethanol tend to form large phase separations, which result in large sized particles and macropores. In the meantime, there is no significant difference in the sizes of particles between the MBSA samples, which indicated that the starting precursor concentration (or density) does not much affect the sizes of the particles, possibly due to the similar reaction conditions.

Fig. 4 FE-SEM images of MBSAs with different densities (g cm⁻³): (a) 0.0535, (b) 0.0662, (c) 0.0702, (d) absorbed hexane after heating at 80 °C (0.0702 g cm⁻³).

The isotherms of nitrogen (N₂) adsorption and desorption of these aerogels are typical type IV isotherms with hysteresis loops according to the ICUPA (as shown in Fig. 5), suggesting that meso and macropores coexist.³² The quantity adsorbed is very low at relative pressures below 0.01, and the desorption branch does not trace the adsorption branch. These results suggest that no micropores exist, but do indicate the presence of mesopores. The specific surface areas of the MBSAs ranged from 38.7 to 57.9 m² g⁻¹, as calculated by the Brunauer–Emmett–Teller (BET) method. However, the specific surface areas seem to be lower than that of the traditional silica aerogels (500–1000 m² g⁻¹).³³ The low surface area values may possibly be due to the large sizes of the aerogel networks, as can be observed in the SEM images shown in Fig. 4, wherein relatively large-sized particle networks have always been observed in flexible silica aerogels. With the increase in density, the BET specific surface area becomes larger (see Table 2); this result may be attributed to the fact that the particles are stacked more closely and as smaller pores are formed with higher concentrations of the precursor.

Fig. 5 N₂ adsorption and desorption isotherms of MBSAs with different densities (g cm⁻³): (a) 0.0535, (b) 0.0662, and (c) 0.0702. (d) Absorbed hexane after heating at 80 °C (0.0702 g cm⁻³).

The thermal stability of the MBSAs was studied by thermogravimetric analysis (TGA) from ambient temperature to 900 °C, as shown in Fig. 2b. It can be observed from the figure that all the products show a small mass loss below 100 °C due to de-intercalation of water absorbed in the aerogel. They then display a steep weight loss from 350 to 600 °C, which can be ascribed to the degradation of hydroxyl and organic bridges (alkyl and phenylene groups). More than 35% of the initial mass of MBSAs is reserved at 900 °C, which corresponds to the residue SiO₂. These results are in accordance with the FTIR, as shown in Fig. 2a. Taking entry 1 as an example, the weight loss at the stage of 350–600 °C is about 46.46%, and 45.76% remaining, thus the mass ratio of the organic part and SiO₂ is about 1.02. In theory the mass ratio between the molecular bridge (legend) and pure SiO₂ (see the chemical structure in Fig. 1b) in the MBSAs is ca. 1.78, and that of pure APTES is 0.97. Considering the mass ratio value (0.97 < 1.02 < 1.78), this could further confirm the existence of the molecular bridge. Moreover, the value of 1.02 is closer to 0.97, indicating that the excess amount of APTES may participate in the reaction of hexaalkoxysilanes and become part of the aerogel network.

It is noteworthy that the low density and high porosity aerogels show impressive flexibility. The mechanical behaviors of the MBSAs under compression are shown in Fig. 6. As can be observed, the MBSA (0.0535 g cm⁻³) shows an excellent elastic performance and can withstand at least 50% deformation under compression without any detectable structural fracture, even a 90% deformation is reversible after the release of the stress (Fig. 6a), which is much better than that of reported silica aerogels and is comparable to carbon nanofiber (CNF) aerogels.³⁴Fig. 6b shows the stress–strain curves of compression repeated 30 times at a deformation of 40%. The stress–strain curves are nearly identical, whereas the specimen maintains a 100% recovery to its initial length after each cycle. The size stability of the aerogels presented in the fatigue test suggests an excellent elastic performance, which is especially significant for aerogels with such low density and high porosity.

Fig. 6 (a) σ versus ε curves for an aerogel (0.0535 g cm⁻³) along the loading direction during loading–unloading cycles at ε = 50–90%. (b) Stress–strain curves of a 30-times-repeating compression test of an aerogel (0.0535 g cm⁻³). (c) Digital images showing an aerogel (0.0535 g cm⁻³) recovering its original shape after releasing the stress.

The excellent elasticity of the MBSAs may be possibly attributed to the network structure obtained by the molecular bridge, wherein: (a) the long and rigid organic chains bridges of silica provide the MBSA with flexible 3D networks, which gives the material resistance to a certain extent to compress; (b) rich pore structures provide the possibility of compressing the MBSAs to consume the energy by shutting off the pores without breaking the network;³⁵ and (c) the large percentage of free space due to the high porosity provides the aerogel with significant low density and remarkable flexibility.

By the Schiff reaction, the non-hydrolysable methylene group (–CH₂–CH₂–) can attach to silicon atoms. The attachment of these hydrolytically stable –CH₂–CH₂– groups to the siloxane backbone during the sol–gel method leads to very low solid–liquid interfacial energies, which in turn results in hydrophobic aerogels, as reflected by the contact angle test, where these water droplets can exhibit a contact angle as high as 150.5° (see ESI in Fig. S1d†), and where these surface methylene groups have been confirmed by FTIR (Fig. 2a), and where the peaks at 2930–2830 cm⁻¹ were attributed to saturated hydrocarbons on the bridge chains.

Owing to their surface hydrophobicity, nanoporous structure, and excellent elasticity, MBSAs are an ideal candidate for the separation/extraction of specific substances, such as organic pollutants and oils, as shown in Fig. S1.† This shows that when a small piece of MBSAs was placed in hexane (dye with oil-soluble dyes), the aerogels absorbed the oil quickly, leaving clean water. To investigate the absorption–sorption capacity quantitatively, the weight gain is defined as the weight of absorbed substance per unit weight of the dried MBSAs. A variety of organic solvents, namely, hexane, dichloromethane, dimethylsulfoxide (DMSO), ethylsilicate, kerosene, ethanol, tetrahydrofuran, and N,N-dimethylformamide, were used in this study. As illustrated in Fig. 7a, the adsorption capacity for the organic solvents ranged from 11 to 24 times in weight, and with the increase of density, the highest absorption capacity was reached for dichloromethane (24 times) and the lowest was for hexane (10.6 times). The absorption mechanism is mainly the physical adsorption of organic molecules, which can be stored in the pores of the MBSAs.³⁴ The absorption capacity of the MBSAs for all the organic solvents is considerably higher than the absorption capacity of the bridged-silsesquioxane aerogels (5–12 times their weight in organic liquids),¹⁷ superhydrophobic silica aerogel microspheres (4–11 times),³⁶ and resorcinol–formaldehyde-reinforced silica aerogels (5–7 times).³⁷ The excellent absorption could be attributed to the high pore volume (as shown in Fig. 4), with values as high as 10.52 cm³ g⁻¹. Moreover, it was observed that the aerogel retains its original shape and size, and there is no shrinkage at all in the final aerogel after complete desorption (see ESI in Fig. S2†).

Fig. 7 (a) Absorption capacity of silica aerogel (0.0535 g cm⁻³) for various organic liquids and (b) retention toward hexane through regeneration by heating at 80 °C for 30 cycles.

More importantly, the MBSAs can be regenerated simply by high temperature evaporation, which is a key criterion for oil/cleanup application. As shown in Fig. 7b, there is nearly no deterioration of the adsorption capacity after thirty adsorption–regeneration cycles of hexane (the sample was heated up to 80 °C for hexane desorption). Moreover, the micro-structure had no significant change after heating the sample at 80 °C (Fig. 4d) as compared to Fig. 4c. Furthermore, nitrogen sorption and pore size distribution (Fig. 5d) analyses revealed that the specific surface and the average pore size slightly decreased to 45.4 m² g⁻¹ and 12.9 nm, respectively, further confirming that the structure of the aerogel was not destroyed, thus making it recyclable many times. The results clearly show the excellent recyclability of the MBSAs as organic solvent absorbents. The large mass uptake and good reusability demonstrated the MBSAs as good candidates for collecting organic pollutants or oil.

Conclusion

Molecular-bridged silica aerogels (MBSAs) have been successfully prepared via Schiff base chemistry, derived from 1,4-phthalaldehyde and APTES in the absence of any catalysts. The gels were formed by the polycondensation of bridged-APTES under mild conditions. ¹H NMR and FTIR results revealed the mechanism of the MBSGs' formation, which was first by forming a bridged-APTES by imine bonds, followed by hydrolysis and polycondensation of the hexaalkoxysilanes. All the process was confirmed to be catalyst-free and autocatalytically gelled by the APTES precursors. The obtained aerogel show good mechanical property, with an impressive 90% deformation and 30 times compression cycles at a deformation of 40%. It is worth noting that the self cross-linked silica aerogels have an absorption capability as high as 11–24 times their weight for organic liquids or oil, as well as excellent recyclability, and can therefore be used for oil/chemical cleanup.

Acknowledgements

This study was financially supported by the National Natural Science Foundation of China (21373024, 21404117, 51572285), the Natural Science Foundation of Jiangsu Province (BK20151234), the Innovation Program of the Beijing Institute of Technology and the 100 Talents Program of the Chinese Academy of Sciences.

References

1. S. S. Kistler, Nature, 1931, 127, 741.
2. S. M. Jones, J. Sol-Gel Sci. Technol., 2006, 40, 351–357.
3. P. Tsou, J. Non-Cryst. Solids, 1995, 186, 415–427.
4. S. Yun, H. Luo and Y. Gao, J. Mater. Chem. A, 2014, 2, 14542–14549.
5. X. Yang and E. D. Cranston, Chem. Mater., 2014, 26, 6016–6025.
6. Y. Wang, S. Yadav, T. Heinlein, V. Konjik, H. Breitzke, G. Buntkowsky, J. J. Schneider and K. Zhang, RSC Adv., 2014, 4, 21553–21558.
7. S. Martínez, M. Moreno-Mañas, A. Vallribera, U. Schubert, A. Roig and E. Molins, New J. Chem., 2006, 30, 1093–1097.
8. Y. Zhang, X. Wu, Y. Fu, W. Shen, X. Zeng and W. Ding, J. Mater. Res., 2014, 29, 2863–2870.
9. J. S. Yoon, Y. Lee, J. Ryu, Y.-A. Kim, E. D. Park, J.-W. Choi, J.-M. Ha, D. J. Suh and H. Lee, Appl. Catal., B, 2013, 142, 668–676.
10. H. El Rassy, A. Perrard and A. C. Pierre, J. Mol. Catal. B: Enzym., 2004, 30, 137–150.
11. A. Jain, S. Rogojevic, S. Ponoth, N. Agarwal, I. Matthew, W. Gill, P. Persans, M. Tomozawa, J. Plawsky and E. Simonyi, Thin Solid Films, 2001, 398, 513–522.
12. J. Fricke, R. Caps, D. Büttner, U. Heinemann and E. Hümmer, J. Non-Cryst. Solids, 1987, 95, 1167–1174.
13. H. Sai, L. Xing, J. Xiang, L. Cui, J. Jiao, C. Zhao, Z. Li and F. Li, J. Mater. Chem. A, 2013, 1, 7963.
14. B. N. Nguyen, M. A. B. Meador, M. E. Tousley, B. Shonkwiler, L. McCorkle, D. A. Scheiman and A. Palczer, ACS Appl. Mater. Interfaces, 2009, 1, 621–630.
15. M. A. Meador, A. S. Weber, A. Hindi, M. Naumenko, L. McCorkle, D. Quade, S. L. Vivod, G. L. Gould, S. White and K. Deshpande, ACS Appl. Mater. Interfaces, 2009, 1, 894–906.
16. M. A. B. Meador, E. F. Fabrizio, F. Ilhan, A. Dass, G. H. Zhang, P. Vassilaras, J. C. Johnston and N. Leventis, Chem. Mater., 2005, 17, 1085–1098.
17. S. Yun, H. Luo and Y. Gao, J. Mater. Chem. A, 2015, 3, 3390–3398.
18. Z. Wang, D. Wang, Z. Qian, J. Guo, H. Dong, N. Zhao and J. Xu, ACS Appl. Mater. Interfaces, 2015, 7, 2016–2024.
19. Z. Wang, Z. Dai, N. Zhao, X. Zhang, H. Dong and J. Xu, J. Sol-Gel Sci. Technol., 2015, 75, 519–529.
20. Z. Wang, Z. Dai, J. Wu, N. Zhao and J. Xu, Adv. Mater., 2013, 25, 4494–4497.
21. D. J. Boday, R. J. Stover, B. Muriithi and D. A. Loy, J. Mater. Sci., 2011, 46, 6371–6377.
22. L. Zhu, Y. Li, J. Wang and J. Shen, J. Appl. Phys., 1995, 77, 2801–2803.
23. G. Liu, B. Zhou, A. Du, J. Shen and G. Wu, Colloids Surf., A, 2013, 436, 763–774.
24. W. J. Malfait, R. Verel and M. M. Koebel, J. Phys. Chem. C, 2014, 118, 25545–25554.
25. K. Kanamori and K. Nakanishi, Chem. Soc. Rev., 2011, 40, 754–770.
26. J. Robert, J. Mater. Chem., 2000, 10, 1113–1120.
27. C. Volcke, R. Gandhiraman, V. Gubala, J. Raj, T. Cummins, G. Fonder, R. Nooney, Z. Mekhalif, G. Herzog and S. Daniels, Biosens. Bioelectron., 2010, 25, 1875–1880.
28. M. Pantoja, N. Encinas, J. Abenojar and M. A. Martínez, Appl. Surf. Sci., 2013, 280, 850–857.
29. N. Encinas, B. Díaz-Benito, J. Abenojar and M. A. Martínez, Surf. Coat. Technol., 2010, 205, 396–402.
30. F. Alba-Elías, E. Sainz-García, A. González-Marcos and J. Ordieres-Meré, Thin Solid Films, 2013, 540, 125–134.
31. L. Martín, J. O. Ossó, S. Ricart, A. Roig, O. García and R. Sastre, J. Mater. Chem., 2008, 18, 207–213.
32. M. Kruk and M. Jaroniec, Chem. Mater., 2001, 13, 3169–3183.
33. J. Wang, Y. Wei, W. N. He and X. T. Zhang, RSC Adv., 2014, 4, 51146–51155.
34. Z. Y. Wu, C. Li, H. W. Liang, J. F. Chen and S. H. Yu, Angew. Chem., Int. Ed., 2013, 52, 2925–2929.
35. W. N. He, G. Y. Li, S. Q. Zhang, Y. Wei, J. Wang, Q. W. Li and X. T. Zhang, ACS Nano, 2015, 9, 4244–4251.
36. S. Yun, H. Luo and Y. Gao, RSC Adv., 2014, 4, 4535–4542.
37. S. Yun, H. Luo and Y. Gao, J. Mater. Chem. A, 2014, 2, 14542–14549.

---

Footnotes

† Electronic supplementary information (ESI) available: Photo images of adsorption–desorption of hexane, selective oil adsorption, and contact angle test, etc. See DOI: 10.1039/c5ra19776g

‡ These authors contribute equally to this work.

---