MTMS (AR, 98%) was purchased from Aladdin Industrial Corporation. The absolute ethanol was purchased from Shanghai Titan Scientific Co., Ltd. The n-hexane was purchased from Shanghai Lingfeng Chemical Reagent Co., Ltd. The HCl, N,N-dimethylformamide (DMF) were purchased from Tianjin Comio Chemical Reagent Co., Ltd. Hexadecyltrimethylammonium bromide (CTAB), NaOH were purchased from Sinopharm Chemical Reagent Co., Ltd. Dichloromethane (DCM, ≥ 99.9% purity) was obtained from Merck. n-Hexane (≥ 99.0% purity) was obtained from Macklin Biochemical Co., Ltd. The deionized water was homemade in the laboratory. All the solutions were prepared with deionized water.

Aerogel was prepared by using the classic sol-gel combined with atmospheric pressure drying method. Figure 1 represents the hydrolysis and gelation process of MTMS, along with the schematic workflow for aerogel preparation. Firstly, a mixture of 9 g deionized water, 0.05 g CTAB, and a measured weight of ethanol was stirred for 30 min. Subsequently, 6.8 g MTMS was injected, and the pH of the solution was adjusted to 3.5 using 0.5 M/L HCl solution. The mixture was then continuously stirred at room temperature for 2 h to ensure complete hydrolysis of MTMS. Then, 0.5 M/L NaOH solution was added to the mixed solution until the pH reached 9.5. The mixture was then continuously stirred for 10 minutes, followed by sealing and curing in a drying oven at 50°C for 12h. The wet gel underwent solvent exchange using ethanol (2–3 times the gel volume) to remove water and other impurities. The ethanol was replaced every 12 hours, and the entire solvent exchange process was maintained for 3 days. Finally, the wet gel was subjected to gradient drying in an oven: first at 40°C for 2 hours, then the temperature was raised to 80°C and maintained for another 2 hours, and finally increased to 105°C for a further 2 hours of drying, yielding the desired aerogel product. In this experiment, ethanol was added at different mass quantities: 0 g, 2.3 g, 4.6 g, 6.9 g, 9.2 g, and 11.5 g (with MTMS : ethanol : H₂O molar ratios of 1:(0–5):10). As shown in Fig. 1(c), the prepared aerogels were marked as E₀, E₁, E₂, E₃, E₄, and E₅, corresponding to the increasing ethanol content.

The organic solvent adsorption experiment was conducted on MTMS-based aerogel. A precisely weighed aerogel was immersed in an excess volume of solvent with a sealed glass vial. The sample was then placed in a fume hood maintained at room temperature and allowed to equilibrate for 60 minutes to ensure complete saturation of the adsorption. Then, the samples were promptly isolated (< 15 seconds) and transferred to an analytical balance for immediate gravimetric analysis. to prevent mass loss of adsorbed organic solvent, all measurements completed within 30 seconds of sample removal from the solvent environment. The saturated adsorption capacity of the aerogel for the target organic solvent was quantitatively determined by measuring the mass difference before and after adsorption, as expressed in Eq. 1.

Eq. 1

Where the is the maximum solvent adsorption capacity, and are the mass of aerogel before and after adsorption experiment.

For cyclic adsorption experiments, the monolithic aerogel was mechanically fragmented into 5mm*5mm*5mm cubes or smaller to ensure the adsorption of organic solvent efficiency. The aerogel containing adsorbed organic solvent were subjected to thermal desorption in a oven at 60.0°C ± 0.5°C for 4 hours to ensure the fully desorption of the aerogel. The adsorption-desorption cycle was systematically repeated under identical conditions until reaching the predetermined number of cycles.

Scanning electron microscopy (SEM, Hitachi Regulus 8100) was used to scan the surface topography of silica aerogel. Particle and pore sizes of aerogel were measured with a Brunauer-Emmett-Teller (BET, Quantachrome Instrument Autosorb iQ). Fourier transformation infrared spectra of the aerogel were registered on a FT-IR spectral photometer (FT-IR, Thermo Scientific, IS3). The phase of the sample was analyzed by X-ray diffraction (XRD, Rigaku Miniflex 600) with a scanning range of 5°–80° at 5°/min. The contact angle was tested using a contact angle meter (Rame-hart instrument, USA).

FT-IR spectra of the aerogels were recorded for the aerogels are shown in Fig. 2(c-d). The broad absorption band centered at 3430 cm^− 1 and 1623 cm^− 1 can be attributed to the stretching vibration and bending vibration of hydroxyl (-OH) groups (Fig. 2(d)). The adsorption peaks at 1133 cm^− 1, 1031 cm^− 1 and 781 cm^− 1 reflects the asymmetric and symmetric vibration of Si-O-Si vibrations. The peak at 1274 cm^− 1 is due to the asymmetric deformations of C-H. As seen from Fig. 2(d), with the increase of the ratio of ethanol, the intensity of the peaks of -OH and C-H decreases, the introduction of ethanol modified the reaction kinetics, facilitating the effective incorporation of Si–OH and –CH₃ groups into the 3D silica network, which enhances the structural coherence and chemical stability of aerogel.

Figure 3 shows the morphology and microstructure of MTMS based aerogel. It can be seen that with the increase of ethanol, the framework and pores of aerogel significantly reduced. The E₀ forms skeletal structures with a thickness of 1–2 µm, featuring pore sizes that typically range from several micrometers to over ten micrometers. The increased ethanol content dilutes the precursor concentration and reduces the system polarity, thereby moderating the MTMS polycondensation kinetics and facilitating the formation of a highly developed nano-porous network structure. The E₃ exhibits a refined skeletal structure (~ 100 nm), characteristic of a typical aerogel 3D nanonetwork. However, a small amount of skeleton stacking is still observed, it may be attributed to the relatively low ethanol content, which results in an excessively rapid polycondensation rate of MTMS. Compared with E₃, the E₅ exhibits a further reduction in skeletal framework size to the nanometers scale, while maintaining comparable pore dimensions. Notably, the skeleton stacking is significantly reduced, resulting in a pore structure that closely with the characteristic architecture reported in literature for aerogels.

Figure 4(a) shows the adsorption capacity of MTMS based aerogel on different organic solvents. The aerogels showed excellent organic solvent adsorption, with capacities reaching several times their own weight. In this section, n-hexane (nonpolar), ethanol (medium-polarity), DMF (highly polar), and dichloromethane (relatively low polarity but with unusual molecular structure) were selected as the target adsorbates. Although the ethanol ratio significantly effects the pore structure of aerogel, its effect on enhancing the adsorption capacity for organic solvents was limited. This phenomenon can be attributed to cooperative adsorption involving both macro- and mesopores within the aerogel framework. It is worth noting that the adsorption performance is highly depends on the properties of organic solvent. The adsorption capacity of aerogel for four organic solvents is manifested as dichloromethane > ethanol > DMF > n-hexane, it can be illustrated by the interaction mechanisms between the solvents and the hydroxyl (-OH) functional groups on the aerogel surface. The aerogel has the highest adsorption capacity for dichloromethane, which might be due to the stronger intermolecular interaction between aerogel and dichloromethane. E₄ demonstrated exceptional adsorption performance, exhibiting a saturated dichloromethane adsorption capacity of 10.59 times its own mass. The adsorption capacity of aerogels for organic solvents progressively increases with higher ethanol content in the precursor. It can be illustrated that ethanol remodels the distribution of active groups on aerogels, strengthening solvent binding at adsorption sites. Moreover, ethanol optimizes the mesoporous architecture to establish three-dimensional interconnected channels conducive to organic solvent diffusion.

Using ethanol as the target adsorbate, the capacity retention rate of the aerogel during multiple adsorption-regeneration cycles was evaluated, with the results presented in Fig. 4(b). The adsorption capacity of the aerogel for ethanol exhibits a progressive decline with increasing cycling numbers. The aerogel exhibits a moderate decline in adsorption capacity during the initial cycling stages. However, as the number of cycles increases, the degradation rate undergoes a marked acceleration. The ethanol ratio in the aerogel precursor exhibits a positive correlation with capacity degradation, the E₅ demonstrates a significant reduction in capacity retention, reaching 90.1% after 50 adsorption-regeneration cycles. This decline may be attributed to the mechanical stress generated during adsorption-desorption cycles, which induces structural damage to the nanoporous framework. This degradation results in the detachment of surface -OH, thereby reducing active sites for organic solvent.

The Fig. 5 illustrates the N₂ adsorption-desorption isotherms of the aerogels after cyclic adsorption-desorption processes with different cycles, and the pore size distributions of the aerogel were obtained by BJH method. Based on the N₂ adsorption-desorption isotherms results, the specific surface area of the samples was represented in Table 1. For E₀ and E₁, the N₂ adsorption-desorption isotherms curves failed to achieve closure, with correspondingly low specific surface areas of 310.3 m²/g and 455.5 m²/g, respectively. This phenomenon can be attributed to the rapid polymerization of the gel framework during synthesis, which formed skeletal structures with dimensions of 1–3 µm and predominantly micron-sized macropores. With the increasing proportion of ethanol in the precursor solution, the specific surface area of aerogel increased significantly and the isotherms of the aerogels could be classified into type Ⅳ. When the relative pressure (P/P0) exceeds 0.5, H3 type hysteresis loop can be observed, indicated that the capillary condensation within the microsphere adsorbent. With increasing adsorption-desorption cycles, the adsorption peak in the high-pressure region of the N₂ adsorption-desorption isotherm for aerogels decreased. It can be illustrated by the structural degradation of the porous network, where repeated adsorption-desorption resulted in the localized collapse or fracture of micro-/mesopore. The reduction of surface area of the aerogels further corroborated this mechanism, demonstrating a strong correlation between structural integrity and adsorption performance.

BJH results further confirms the detrimental effect of adsorption-desorption process on the aerogel pores, particularly in the mesoporous range (2–50 nm). The degradation of aerogel structural integrity, specific surface area, and adsorption capacity exhibits a strong correlation, demonstrating that pore structure integrity is a key factor in adsorption durability performance. Although the skeletal fineness of E₀ is several micrometers thick (as shown in Fig. 3(a)), suggesting that the cyclic adsorption process induces systematic structural damage to the framework of aerogel.

The Fig. 6 represents the photos of aerogel after adsorption- desorption for 50 cycles. Before the cyclic adsorption experiments, the aerogel was fractured into irregular fragments. After 50 cyclic adsorption- desorption cycles, all aerogel samples exhibited pronounced structural disintegration. Clear positive correlation between ethanol content in the precursor solution and aerogel fragmentation severity can be observed. Notably, the E₅ underwent complete disintegration into granular particles, with no monolithic structures can be observed. Although part of aerogels preserved their macroscopic monolithic structure after 50 cycles, systematic morphological evolution was observed. The originally well-defined edges underwent progressive rounding, demonstrating continuous surface smoothing throughout cyclic experiment. Besides the adsorption capacity, the macroscopic structural stability of MTMS-based aerogels has been identified as a critical determinant of cyclic adsorption performance.

The microscopic morphology of the aerogel after 30 and 50 adsorption- desorption cycles is shown in Fig. 7. The cyclic adsorption-desorption experiment did not significantly alter the microporous structure of the aerogels, and all samples retained their original pore size distribution, consistent with those of fresh samples. Due to the skeleton size of E₃ and E₅ are at nanometer scale, the morphological changes were barely detectable under SEM. In contrast, the skeleton of E₀ (about 5 µm) exhibited pronounced fracture during cycling, with fracture density showing a clear dependence on the number of cycles. The fracture of the skeleton explained the fragmentation of the aerogel, consistent with the observations in Fig. 6. Moreover, the fracture of the aerogel skeleton significantly reduced the interconnectedness of the pore network, preventing certain pores from achieving full nitrogen contact during BET analysis. This damage results in a measurable reduction in specific surface area. Since the pore size distribution remains overall stable, the skeletal fractures are likely to concentrate at localized weak regions. Corresponding to the capacity retention rate results, the degree of aerogel particle fragmentation exhibits a positive correlation with the retention rate of adsorption capacity. A structurally uniform aerogel skeleton can withstand multiple adsorption cycles without significant degradation. Therefore, by regulating the sol-gel process of aerogel, prevent skeletal stacking and stress concentration, improve the uniformity of the aerogel framework is the key strategy to improve the cyclic adsorption-desorption durability of aerogels.