Preparation and morphological/chemical characterization of aerogels
Figure 1 shows the preparation process and synthesis mechanism of the transparent cellulose/silica composite aerogels. The cellulose block/film gels were firstly prepared and then impregnated into co-precursors (TEOS/MTES) to prepare cellulose/silica composite aerogels via a two-step sol-gel process followed by supercritical CO2 drying. It is well-known that anions in the ionic dissolve cellulose by forming strong hydrogen bonds with hydroxyl hydrogen atoms in cellulose (Zhang et al. 2017). On the contrary, the cations have little solubilization effect on cellulose (Fig. S5a). As can be seen in Fig. S5b, the co-precursors (TEOS/MTES) undergo hydrolysis reaction, and the hydrolyzed siloxy groups generate hydroxyl groups linking to silicon atoms (Si(OH)4, CH3-Si(OH)3), respectively. Subsequently, the silica sols particles will gradually penetrate into the internal pore structures of the cellulose wet gels, displacing the original ionic solution. After catalyzed by alkaline agent, the hydrolyzed monomers undergo polycondensation and crosslink with each other. Generally, the condensation rate of Si(OH)4 molecules is much higher than that of CH3-Si(OH)3 molecules in the mixed precursor solution, which leads to a preferential condensation of Si(OH)4 molecules to form primary particles (He et al. 2017). In addition, the unreacted hydroxyl groups in co-precursors will further bond with the hydroxyl groups on the surface of the cellulose wet gels, forming the final cellulose/silica composite gels dual network structures. Table S1 shows the summary of the textural properties of aerogels samples with different SiO2 content. The spin-coating process, equipment and representative samples (CAS wet gels and aerogels films) are supplemented in Fig. S2-4. With the increase of silica content, the densities of aerogel samples increase from 0.058 g/cm3 of pure CA to 0.17 g/cm3 of CAS5. Oppositely, the volume shrinkage rates of aerogels samples decrease gradually from 22.9% (CA) to 10.1% (CAS5). Therefore, the feasibility of this study prepared the transparent cellulose/silica composite aerogel has been proved both theoretically and practically.
Figure 2a shows the SEM images of CA and CAS aerogels and the corresponding macro digital photographs. It is evident that both CA and CAS exhibit a uniform three-dimensional porous structure. Meanwhile, all aerogels samples exhibit outstanding integrity and transparency. By contrast, the pure CA aerogels show an intricate three-dimensional network consisting of inter-layered and interwoven multi-scale pore structures. The average pore sizes of CAS aerogels are obviously smaller than that of CA aerogels, mainly due to the introduction of a large number of spherical nanoscale silica aerogels nanoparticles, forming a more uniform composite structure. It is found that with the increase of silica content, the morphology gradually changes from filamentous networks (CA) to pearl-necklace (CAS1-3), then to spherical particles aggregation state (CAS4-5). Furthermore, the evolution of pore structure also has a certain influence on the transparency of the aerogel sample, and the smaller and more uniform the pore size, the better the light transmission (as shown in physical photograph). In order to further observe the micro-morphological combination of cellulose and silica particles, the TEM image shows that nanoscale silica particles loaded on the surface of fibrous micrometer cellulose monomer (as seen in Fig. S6). Therefore, the addition of silica aerogel particles makes the pore structure of the composite aerogel more uniform. Meanwhile, the fibrous cellulose aerogel also plays a role in supporting the network skeleton structure, reducing the shrinkage and improving the flexibility of cellulose/silica composite aerogel. The EDS spectrum of the selected CAS4 is revealed in Fig. 2b. The specific element content distribution is shown in Fig. 2b, which basically corresponds to the molar ratio of CAS4. In addition, the uniform distribution of C, O and Si elements on the surface of the CAS4 further proves that the cellulose micronano fibrous monomers were equably covered by silica aerogel particles.
The N2 adsorption–desorption isotherms and BJH pore size distribution of the as-prepared CA and CAS aerogels are described in Fig. 3. The relevant data are summarized in Table S1. All isotherms are Type IV based on the IUPAC classification, reflecting the characteristic of mesoporous materials. The desorption cycles of the isotherms show a hysteresis loop for the six samples, which is generally caused by the capillary condensation that occurs in the mesopore. The rapid adsorption process in the low-pressure region (0-0.1) is caused by the micropores inside the aerogel matrix (Sehaqui et al. 2011). As shown in Table S1, the specific surface area of CAS samples is significantly higher than that of CA samples, mainly due to the network structures of CAS samples with high porosity. With the increase of silicon content, the specific surface area of CAS samples show an increasing trend (from 660.87 m2/g to 1089.70 m2/g), as shown by the change in volume of N2 adsorbed. From the pore size distribution, the average pore sizes of the CAS samples gradually become smaller from 15.20 nm to 9.89 nm, resulting from the gradually denser skeleton structures of the CAS (as shown in Fig. 2a). It is worth noting that a small number of micropores (2–10 nm) appeared in the CAS5 sample, presumably due to the local agglomeration of particles in the composite aerogel (corresponding to SEM image of CAS5, in Fig. 2a). Moreover, these newly generated micropores may also be one of the reasons for further improving the specific surface area of CAS5 sample. Previous studies have also demonstrated that an appropriate addition of silica or other inorganic nanoparticle can effectively improve the non-uniform macroporous structure of cellulose aerogel, forming a more uniform mesoporous structure of cellulose-based composite aerogel (Cai et al. 2012, Wan and Li 2015).
The FTIR spectra of the as-prepared aerogels samples are described in Fig. 4a. The spectrum of CA is significantly distinct from those of CAS. The two broad peaks at 3410 and 1650 cm− 1 are assigned to the stretching vibrations of hydroxyl groups of cellulose and the presence of absorber H2O, respectively. However, these two bands are less evident in the spectrum of the CAS samples. This is because the fact that the continuous consumptions of the hydroxy groups on the surface of cellulose and silicon co-precursors during the hydrolysis-polycondensation process, resulting in the decrease of the hydrophilic characteristics of the final CAS samples. In addition, the characteristic peaks of cellulose are also observed at 2898, 1347, 1156 and 898 cm− 1, which are related to C-H stretching vibrations, C-H deformations, C-O-C asymmetric stretching and β-glucosidic bonds (Ashori et al. 2012). Compared to CA, there are two new characteristic peaks of 795 and 446 cm⁻¹ in CAS, corresponding to Si-O-Si bending and symmetric stretching vibrations, respectively (Zu et al. 2018). It is worth noting that the intensities of these characteristic peaks become obvious with the increase of silica content. To some extent, the presence and change of silica aerogel in CAS composites can be demonstrated by FTIR spectra. XPS spectroscopy is employed to further observe the chemical composition of the CA and selected CAS4 in Fig. 4(c-f). As shown in Fig. 4c, the C, O, and Si elements signals can be clearly detected from the spectra. Two characteristic peaks, Si-O-Si (532.5 eV) and SiO2 (532.9 eV) are clearly visible in the O1s spectrum. The presence of Si-O-Si groups confirms the occurrence of hydrolysis and polycondensation reactions in TEOS and MTES co-precursors. As for the C1s spectra, two characteristic peaks, C-Si (286.2 eV) and C-C (284.8 eV) are also discernible. The analyses of XPS further confirm the FTIR test results. Figure 4b shows the XRD of CA and representative CAS4. Previous studies have shown that the amount of silica aerogel has no effect on its amorphous state. As shown in Fig. 4b, both CA and CAS4 display the relative broad diffraction peaks at 22.40° (2θ), indicating the presence of amorphous cellulose and silica aerogels particles in as-prepared aerogels samples (Jiao et al. 2018, Zhang et al. 2020). Besides, the introduction of silica aerogel do not change the shape of the original CA diffraction peak, but the diffraction peak of the cellulose/silica composite (CAS4) becomes wider. Therefore, it can be inferred that the CAS samples are comprised of amorphous fibrous cellulose and spherical silica particles with more uniformly dispersed state (as shown in Fig. 2a).
Optical properties of aerogels
The optical performances of CA and CAS samples with different thicknesses are in Fig. 5. As shown in Fig. 5a, all the as-prepared aerogels samples have good light transmittance (physical photos in Fig. 2a). With the increasing of wavelength (400–800 nm), the light transmittance of the aerogel sample exhibits an increasing trend. The CA displays a transmittance of 78.2% at wavelength of 800nm, indicating comparatively high transparency in relation to other cellulose aerogels (Cai et al. 2012, Guzun et al. 2014, Cai et al. 2009). Meanwhile, the curves of transmittance of the CAS samples show a trend of first increasing and then decreasing (from 45.8–80.6%). By contrast, the CAS4 sample exhibits the best transparency properties, which is mainly attributed to its uniform pore structure and the suitable size of secondary particles (Zu et al. 2018, Nakanishi et al. 2020). Similar reports have proved that the silica content has an effect on the transparency of cellulose-silica composite aerogel. For example, Cai et al. synthesized the cellulose-silica composite aerogels with the transmittance of 84% (39wt% of silica aerogel) (Cai et al. 2012). Moreover, the light transmittance of the CAS samples with different thickness was investigated. The highly transparent wet gels and aerogels of CAS samples (circle, 0.5 mm and square, 1 mm) are shown in Fig. 4S. To further demonstrate the influence of thickness on the transmittance of CAS aerogel, the representative CAS4 samples are prepared with gradient thicknesses of 0.5, 1, 2, 3 and 5 mm, and the photos of aerogels are shown in Fig. 5c. Obviously, with the increase of thickness, the transmittance of the CAS4 samples decrease gradually. In contrast, the transparency of the samples (thickness, ༜2 mm) is higher, and the CAS4 of 0.5 mm shows the highest transparency of 91.7%, which is far higher than the requirement of 70% transparency of architectural glass. Compared to the glass-air interface, the minimal refractive index differential of the CAS4-air interface leads to significantly less light reflection enhancing the light transmission for CAS4 within the visible spectral range (Salamati et al. 2020).
Mechanical Properties and Thermal Stability
The mechanical properties of the as-prepared cellulose/silica composite aerogels are explored by compression, tensile, and bending tests. Due to their unique molecular and nanoscale structures, the composite aerogels exhibit excellent mechanical properties, including high compression, substantial tensile toughness, and notable bending elasticity. Figure 6a visually shows the considerable bending elasticity of the composite aerogel and the tested sample could still keep the intact structure after more than 500 times bending tests. The stress-strain curves of the CAS samples are illustrated in Fig. 6b. During compression, all composite aerogels exhibit the typical deformation behavior seen in porous materials. Remarkably, the inorganic core composed of Si-O bonds shows resistance to deformation, allowing the composite aerogels to withstand higher stresses without brittle damage. Simultaneously, the cross-linking of cellulose with organic side chains and the micro- and nanopore structure increases the cross-link density of the aerogels, which also contributes to the strength of the aerogels (Wang et al. 2020). At low strains, aerogels exhibit linear elasticity due to the irreversible collapse of the porous structure, with a plateau of plastic deformation at moderate strains. At strains above about 40%, the sharp increase in stress indicates the presence of a dense state after the complete collapse of the porous structure. The increase in silica significantly boosts compressive strength. Simultaneously, the composite aerogels containing high levels of silica exhibit good toughness, with the ability to endure compression until strains exceed 60%. Figure 6(c) displays the maximum tensile strengths of CA and CAS4, measuring 4.07 MPa and 1.54 MPa, respectively. It is found that the composite aerogels become less flexible with the addition of silica nanoparticles, but still retains a degree of flexibility, making it suitable for applications such as windows and energy-efficient buildings (Abraham et al. 2023). Moreover, as shown in Fig. 6d, the excellent mechanical properties enable the composite aerogels to be easily shaped with a knife, highlighting their practical workability. In addition, the reinforcement mechanisms of silicon source co-precursors on cellulose aerogels are shown in Fig. 6(e ~ g). We propose that the co-precursors (TEOS, MTES), and cellulose create an "organic-inorganic combination" structure, featuring a sturdy inorganic core and flexible organic side chains. During the testing processes of compression, stretching and bending, the inorganic core consisting of Si-O bonds resists deformation, while the organic side chains formed by cellulose exhibit significant flexibility. Therefore, the dual nature allows the cellulose/silica composite aerogels to withstand high stresses without becoming brittle.
Flame retardancy
The flame retardant properties of the various aerogels samples are evaluated through ignition burning tests, as shown in Fig. S7 and Fig. 7(a-b). Figure 7(a-b) shows the comparative state of the CA and CAS4 samples on a burning candle. It is found that the CA sample is easily ignited with a strong flame and disappears after a few seconds, mainly due to the composition of cellulose aerogel is mostly organic matter. Instead, the ignited CAS4 sample self-extinguished in about 5 seconds, maintaining its intact block structure. Thus, the presence of inorganic silica aerogel nanoparticles can significantly improve the flame retardant property of the cellulose-based aerogels. Additionally, as shown in Fig. S7, it can be clearly observed that the higher the content of silica aerogel, the better the flame retardancy of composite aerogel. The flame-retardant process and mechanism of cellulose/silica composite aerogel is further investigated, as shown in Fig. 7c. Because the silica sol is introduced into the macroporous structure of cellulose wet gel, the composite aerogel with a double network structure is finally formed after supercritical CO2 drying process. This dual network structure of the cellulose/silica composite aerogel forms a robust insulation and gas barrier layer during the burning process. The insulating layer slows down the heat transfer within the cellulose aerogel system, thereby retarding the combustion of composite aerogels. Simultaneously, the gas barrier layer efficiently inhibits the volatilization of combustible components, reducing their contact with oxygen and further delaying the effective combustion of composite aerogels. Conversely, the gas barrier layer proficiently impedes the volatilization and escape of combustible components, diminishing their interaction with oxygen and thus extending the combustion delay of composite aerogels. Additionally, the high temperature resistant structure of inorganic silica aerogel supports the formation of a porous carbonization layer, which can be used as an effective barrier layer. Moreover, this porous carbonization layer further retards the heat transfer, the volatilization and escape of combustible components.
TG and DSC analyses are further carried out in an air environment to study the effect of silica aerogels nanoparticles on the thermal oxidation and thermal stability of CA and CAS samples, as shown in Fig. 4(g, h). Comparatively speaking, the CAS samples show higher thermal stability than that of the CA sample. The first stage from room temperature to 100°C, the CAS samples show a weight loss of approximately 5%, due to the evaporation of physically adsorbed H2O and residual solvent. At this stage, the mass loss of CA is slightly larger (about 10%), mainly because of the presence of micron-sized pores. In general, because of the nano-sized porous structure and high porosity, H2O, CO2 and solvent adsorbed in the porous structure of the sample cannot be removed completely during the supercritical CO2 drying process. A significant mass loss of all the aerogels samples occurs at 250–350°C, which corresponded to the continuous thermal decomposition of organic components of cellulose polymer. Immediately, another significant mass loss occurs in the temperature range of 350–500°C, probably due to the further consumption of organic groups in composite aerogel, such as -CH3 groups. For the last stage (above 500°C), all the TG curves of the samples tend to smooth and stabilization, while the corresponding final residual mass of 1.63%, 48.14%, 64.43%, 68.67%, 71.78% and 73.41%, respectively. Specifically, the Tonset (the onset temperature during the decomposition of the samples) has changed to a higher temperature with the increasing of silica content. The Tonset for the CA is 278°C, which continuously increases to maximum temperature of 331°C for the CAS5 sample. Additionally, the CA aerogel displays a significant secondary degradation stage, completely degrading at approximately 500°C. This degradation behavior aligns with prior reports on cellulose and cellulose aerogels (Aggarwal et al. 1997, Yuan et al. 2016). Compared to CA, all the CAS samples show a higher Tonset (325ཞ331°C) and residual mass (48.14ཞ73.41%), which indicate that silica nanoparticles can effectively improve the thermal stability of cellulose/silica composites aerogels. Moreover, inorganic porous silica aerogel used as an efficient heat-resistant barrier can delay the thermal decomposition rate of cellulose polymers and further improve the flame-retardant property of composite aerogels, as shown in Fig. 7c. Therefore, the as-prepared CAS with excellent thermal stability and heat-insulating property is a promising candidate for a variety of applications in high temperature environments.
Insulation performance
Lightweight nanoporous aerogels are known for their excellent optical and thermal insulation properties, and have great application potential in energy-saving buildings. As shown in Fig. 8a, the transparent aerogel glass windows used in the building can not only be efficiently penetrated by the bright sunlight, but also maintain the balance of indoor temperature for a long time, providing people with a comfortable living environment all year round. The light-transmitting, heat-insulating aerogel glass can significantly reduce energy consumption for heating in winter and cooling in summer, which is fully in line with the concept of global green ecology. Figure 8c displays an aerogel glass window model placed in an outdoor environment to further investigate the real effect of heat insulation using two miniature thermometers. We clearly found that after prolonged exposure to intense sunlight, the temperature on the thermometer placed outside quickly rose from the initial 30°C to 49°C. Oppositely, the temperature inside the model showed a slow rising trend and eventually stabilized at 37°C. The specific comparative temperatures changes over time are shown in Fig. 8b. Additionally, Fig. 4i shows the thermal conductivities of the CAS samples. With the increase of silica aerogel content, the thermal conductivity of CAS shows a significant decreasing trend. Particularly, the CAS4 sample possesses the lowest thermal conductivity of 0.0188 W·m− 1·K− 1, mainly due to the uniform mesoporous structure of the composite aerogel. Moreover, the comparative cold surface temperatures of the CA and CAS4 samples were measured on a heating platform set at 130°C by using an infrared thermal imager, as shown in Fig. 8(d-f). By contrast, the selected CAS4 sample (3mm) shows better heat-insulating property at ambient temperature of 22°C, with the cold surface temperature rising from 35.7°C to 51.4°C in 5 min, while the CA sample (3 mm) increases from 43.8°C to 72.7°C. The corresponding curves of cold surface temperatures with the test time are shown in Fig. S9. In order to make the as-prepared composite aerogel used in different scenarios, the cold surface temperatures of CAS4 samples with different thicknesses and shapes were also tested, and the results showed that the samples still had good thermal insulation properties, as shown in Fig. 8f. Some previous studies have shown that the practical applications of the cellulose-based aerogels depend on their comprehensive properties (Chen et al. 2021, Cai et al. 2018, Fu et al. 2018, Kunjalukkal Padmanabhan et al. 2021, Demilecamps et al. 2015, Yuan et al. 2017). The comparison of comprehensive properties of the as-prepared CAS with other representative aerogels materials are further carried out, and the detailed dates are in Fig. S10 and Table S2. The CAS4 aerogels exhibit a combination of high strength, low density, insulation properties, specific surface area, thermal stability, and transparency. This overall performance, coupled with the cost-effectiveness of cellulose, opens the door to a wide range of potential applications.