Application of Silica Aerogel in Composites Protecting Against Thermal Radiation

Textile–polymer composites with aerogel were made from the following materials: -

flame-retardant fabric S (98% Kermel®—meta-aramid fibers/2% conductive fiber, mass per unit area 220 ± 10 g/m², twill weave 3/1 Z, with fluorocarbon finish, manufactured by Andropol S.A. Andrychów, Poland), designated as fabric S,

The tests also involved the uncoated fabric S and fabric S coated with an acrylic–styrene dispersion without aerogel for comparison purposes.

A flame retardant and wetting agent were mixed for 2 min with aerogel in a glass vessel. Then, an appropriate amount of that mixture was added to a vessel containing the selected polymeric dispersion, gently stirring using a mechanical laboratory stirrer with a metal anchor impeller (Biomix, Poland) for 5 min at a minimum rotational speed of 30 rpm. The composition of coating pastes is given in Table 2.

The pastes were applied onto 210 ´ 297 mm fabric samples and spread using a metal roller weighing 0.25 kg. The coated samples were dried for 30 min at 80°C in a Memmert laboratory drier (Germany). Ten sheets were prepared from each composite variants and each sheet was divided into samples for the tests. The samples for tests were collected on a random basis. The thickness and mass per unit area of textile composites were determined (Figure 1). Thickness was measured in 10 replicates in randomly selected locations of each tested material to ensure the uniformity. The tests were conducted according to EN ISO 5084:1996 using a disk micrometer with a diameter of 50.5 mm at a pressure of 1 kPa (Modell Rainbow T, Schröder Prüftechnik, Germany) [26]. Mass per unit area was determined in triplicate for randomly selected samples according to EN 12127:1997 using 100 × 100 mm samples [27]. These tests were carried out to assess the homogeneity of the prepared composites. It was noted that the values of standard deviation for thickness measurement were from 0.01 to 0.06 mm, which was from 1.96 to 8.47% of the measured values. For the mass per unit area measurements, the values of standard deviation were from 0.78 to 38.61 g/m² (0.34–8.58% of the measured values).

Figure 1

Figure 2 shows photographs of sample surface.

Figure 2

Figure 4 shows sample microscopic images of aerogel, the surface of uncoated fabric S, and the fabric S with unmodified and modified coatings. Particle size measurements of CG material and CG incorporated in coatings indicate that the size of aerogel particles was greatly reduced during the technological process, from 397.07 ± 201.78 mm to 7.28 ± 5.17 mm for KCB CG 6% FR 1, which implies a 50-fold reduction in particle size. In the case of the composite KCB CG 6% FR 1, small aerogel particles could be seen only at a magnification of 2,000 × (Figure 4h). Standard deviation values of aerogel size measurements were high and ranged from 50 to 90% of the measured values, despite a huge number of measurements (up to 50). As it was noted in the section Introduction, the aerogel fragmentation is one of the most serious problems associated with textile applications. Aerogel particle size reduction was somewhat less severe in the case of coatings consisting of wetting agent, flame retardant, and aerogel in a polymeric dispersion. In this case, the aerogel particle size was 13.10 ± 9.04 mm for KCB CG 7% FR 2 wa and 10.10 ± 5.17 mm for KCB CG 14% FR 2 wa (the difference is attributable to the greater friction forces during a paste preparation), which corresponds to a 30–40-fold size reduction (Figure 5). The probable reason is that the aerogel was more resistant to crushing in coatings containing a wetting agent due to the fact that the latter decreased the surface tension of the aqueous phase of dispersion, while wetting the hydrophobic surface of aerogel particles and facilitating their adhesion to other components of the mixture. The resulting “shells” provided some protection to aerogel particles, lowering their susceptibility to crushing. It was found that both large agglomerates and small particles of aerogel were located side by side on the surface of the same sample. The authors hypothesized that the reason for this was the way the aerogel particles were attached to the surface of the composite samples, which was also observed in the study by Berardi and Zaidi [10].

Figure 4

Figure 5

Aerogel crushing during a test simulating the real use of footwear material was observed by Kraner et al. [31]. Those researchers studied silica aerogel composites made from a silica aerogel matrix embedded in a noncombustible reinforced fibrous matting material and laminated with a polyether– ester block copolymer membrane. The flexing of the material revealed white areas indicating the presence of silica aerogel dust. Silica aerogel dust (>100 nm) arose from the broken particles of aerogel which migrated from the inner part of the composite to its surface. According to Kraner et al. [31], a lamination was an appropriate method for the preparation of aerogel composites for footwear applications.

In addition, Sarawade et al. [32] reported a reduction in the size of silica aerogel beads exposed to high temperature. SEM and transmission electron microscopic (TEM) images showed that the bead diameter dramatically decreased with increasing temperature. A temperature increase from 200 to 500°C resulted in a considerable rise of aerogel surface weight, with a concomitant reduction in a particle size from 524 to 776 m²/g, while the pore volume and diameter increased only for samples exposed to temperatures of less than 500°C.

Krzemińska et al. [33] reported that the exposure of aerogels to heat led to a reduction in their weight, but only at much higher temperatures. They observed a weight loss of approximately 2% at 180°C (irrespective of treatment time) as compared with approximately 7% at 350°C (a more than threefold increase). In turn, Williams et al. [34] studied polystyrene nanocomposites with different types of silica aerogels: 4 mm granules, ball-milled granules, and 80 µm beads. In cone calorimetry tests with a heat flux of 35 kW/m², they observed disintegration of the mesoporous structure of aerogel, especially as a result of milling, which impaired the beneficial effects of aerogel. Williams et al. concluded that larger mesoporous aerogel particles absorbed the air and decreased heat transfer to the base polymer, delaying or reducing the heat release. Probably, due to the aerogel particle size reduction revealed by SEM, composites containing aerogel coatings exhibited only a slight increase in resistance to thermal radiation.

SEM/EDS analysis revealed variation in Si content depending on the paste composition and aerogel concentration in the KCB CG coating with a flame retardant (see Table 4). The values of standard deviation for average content of Si were 0.6–1.5 wt% and for other elements they were 0.0–4.0 wt%, which corresponded to a maximum of 10% of the measured value.

TG/DTG data for the fabric, coatings, and the resulting composites are presented in Table 5. The values of standard deviation for average temperature values were 0.1–1.1^oC and for average weight losses 0.14–1.34%.

Figure 6 shows TG/DTG curves for different types of coatings and coated fabrics. The thermal degradation of aerogel CG started at 430.0°C with a maximum at 564.4° and 710.2°C. CG aerogel was thermally stable with a weight loss of 11.11% at 850°C (Figure 6a and Table 5).

Figure 6

At the initial stage of heating, the TG curve for the coating KCB CG 0% revealed a slight weight loss in the range of 30–160°C, which corresponded to the water desorption peak in the DTG curve in the same temperature range. The main thermal degradation occurred above 360°C with a maximum of 403.1°C in the DTG curve. The thermal degradation process of coating KCB CG 7% FR2 wa started at the same temperature as for KCB CG 0%. The weight loss at 850°C was comparable with KCB CG 0% and amounted to 90.45%. The initial thermal degradation temperature obtained for the coating KCB CG 14% FR2 wa was higher than for KCB CG 7% FR2 wa by approximately 5°C. The thermal degradation of coating SAL CG 17% started at 333.1°C and proceeded in three steps with maxima at 370.2°, 440.7°, and 472.2°C. In the temperature range of thermal degradation, the weight loss values were 52.12, 15.61, and 9.82%, respectively.

In the initial stage of heating, the TG curve of reference fabric S (Figure 6b) revealed a slight weight loss in the range of 30–120°C, corresponding to the water desorption peak in the DTG curve in the same temperature range. The main thermal degradation occurred above 420°C with a maximum of 471.1°C in the DTG curve. The weight loss at about 850°C was 53.40%.

Fabric S modified with KCB (acrylic–styrene) dispersion without CG aerogel (KCB CG 0%) decomposed in two main steps in the temperature range of 360–613°C. The TG curve showed a weight loss of about 27.57%, corresponding to a peak in the DTG curve with a maximum at 391.3°C, indicating thermal degradation of the KCB coating. The second weight loss of 26.00% corresponded to a peak in the DTG curve with a maximum at 460.4°C, indicating the degradation of the fabric S itself. Interestingly, that peak shifted toward lower temperatures in comparison with the unmodified fabric S. Modification with a KCB coating without CG aerogel increased the weight loss to 65.74% at 850°C.

The degradation of fabric KCB CG 6% FR1 proceeded in three steps. In the first step, the initial thermal degradation temperature was above 150°C with a maximum of 222.3°C, attributable to the decomposition of FR1 and resulting in a weight loss of 13.17%. The second weight loss occurred above 300°C, which corresponded to the KCB CG 6% FR1 coating with the peak temperature shifted toward higher values by about 4°C in comparison with KCB CG 0%. In the third step, the addition of FR1 and 6% of CG did not cause significant changes in the temperature of fabric S in comparison with KCB CG 0%. The weight loss at 850°C increased by about 10% in comparison with KCB CG 0% fabric.

The thermal degradation of fabric KCB CG 7% FR2 wa started at the similar temperature to that for KCB CG 0% and proceeded in two steps. In step 1, the maximum temperature of KCB CG 7% FR2 wa coating decomposition was higher by 7.6°C. In step 2 (fabric decomposition), the maximum peak temperature shifted upward by about 50°C in comparison with KCB CG 0%. The weight losses recorded during the first and second degradation steps were higher by about 5% and lower by about 18%, respectively, as compared with the fabric KCB CG 0%. The weight loss at 850^oC was 56.03%. The addition of 7% of aerogel, a flame retardant, and a wetting agent improved the thermal stability of KCB CG 7% FR2 wa in comparison with KCB CG 0%. The replacement of FR1 with FR2 decreased the weight loss and increased the degradation temperature of fabric S.

The thermal degradation temperatures obtained for KCB CG 14% FR2 wa (Figure 4f) were the same as those for KCB CG 7% FR2 wa. The degradation of the SAL CG 17% composite started at 341.6°C and proceeded in two steps with maxima at 376.0° and 464.1°C, corresponding to the degradation of the coating and fabric, respectively. Within the thermal degradation temperature range of two steps, the weight loss values amounted to 23.35 and 31.62%, respectively.

The amount of aerogel added to fabric coatings was too low to show in TG/DTG curves (Figure 6).

Measurements of the obtained composites revealed that aerogel application in the fabric coatings slightly improved their resistance to radiation. The RHTI₂₄ of composites with lower aerogel content (6 or 7%) was similar to that of the uncoated fabric (Figure 7). In turn, composites with the higher aerogel content in the coatings (KCB CG 14% FR 2 wa and SAL CG 17%) exhibited an improvement in RHTI₂₄ by approximately 15–25%. The significance of those differences was confirmed by a statistical analysis (p = 0.017). The heat transfer index RHTI₂₄ measurements can be considered as repetitive. The values of standard deviation were 0.20–1.21 s (Figure 7). This represents from 0.36 to 8.34% of the measured index RHTI₂₄.

Figure 7

Dispersions of polychloroprene latex and synthetic resin (SAL CG 17%) led to the better resistance to thermal radiation as compared with acrylic–styrene dispersions (KCB CG 14% FR 2 wa) (Figure 7). The difference amounted to 1.5 s and was statistically significant (p = 0.017). A similar effect was exerted on RHTI₂₄ by both flame retardants. The difference between composites containing coatings with 6 or 7% of aerogel and those containing flame retardants was 1.6 s (Figure 8). Thus, the aqueous solution of ammonium polyphosphate was more useful as it led to higher RHTI₂₄ both at 7 and 14% of aerogel content (approximately 2 s).

Figure 8

It should be noted that the obtained composites were characterized by the resistance to thermal radiation ranging from 13 to 17 s, means that all variants were assigned to the first protection class (7–20 s) according to EN 11612:2015 (see Table 3) [29]. The heat flux of 20 kW/m² used in the present study did not burn or melt the composites with aerogel coatings and the sample surface did not reveal any differences. In turn, Li et al. [35], who exposed glass fiber films with aerogel wrapped in an aluminum foil to heat fluxes of 35 and 60 kW/m², observed that composites became slightly yellowish. The higher heat flux of 60 kW/m² resulted in a partial combustion of composites, leaving behind black charred remains and white silica residues. Song et al. [36] reported a visible flame in their studies of silica aerogels using a cone calorimeter at a heat flux of 50 kW/m². The flame emergence time was 6 s. They observed no flame at or below 40 kW/m². The time, after which large amounts of heat were released, was approximately 80–100 s. Under the present testing conditions, the composites were exposed to a heat flux of 20 kW/m² for 10 s and no flames were observed. As can be seen, the radiant flux intensity is essential.

Chakraborty et al. [22] also obtained a significant improvement of the resistance to radiant heat after the application of silica aerogel to produce aerogel–Nomex nonwoven composite which was used by him as the middle layer in three-layered firefighters clothing.

The protection time of the fabric assembly in protective clothing, outer fabric—Nomex woven, middle layer—aerogel + Nomex nonwoven, inner layer—modacrylic/cotton woven after the exposure to heat flux of ~35.0 kW/m², increased from 40 s for the lowest concentration of the precursor to over 80 s for the highest concentration (15.4 and 46.2 molar ratio, respectively). Thus, there was a twofold increase in the protection time. Chakraborty et al. pointed to the problem of dust shedding during the preparation of the Nomex–aerogel-based fabric assemblies, which was also observed by the authors of this paper.

The produced thermal insulating composites with aerogel coatings were characterized by the slightly higher thermal conductivity than the uncoated meta-aramid fiber fabric (Figure 8). The coating of fabric with a polymeric dispersion increased its thermal conductivity from 35.70 ± 0.95 × 10⁻³ W/mK to 42.00 ± 3.29 × 10⁻³ W/mK. Statistical analysis showed significant differences between the unmodified fabric sample, the sample coated with the acrylic–styrene dispersion alone, and composites with aerogel coatings (p = 0.0002 and =0.0003). The highest thermal conductivity (47.94 ± 4.00 × 10⁻³ W/mK) was found for the composite with a coating containing 6% of aerogel and FR 1. The other composites coated with mixtures containing FR 2 exhibited the thermal conductivity of 39 × 10⁻³ W/mK, irrespective of aerogel concentration (Figure 8). The values of standard deviation ranged from 0.95 × 10⁻³ K to 4.00 × 10⁻³ W/mK. This represents an average of about 4% of the measured thermal conductivity.

In the study conducted by Collin et al. [2], a thermal conductivity of 81 × 10⁻³ W/mK was obtained for the outermost layer of firefighter protective clothing made of 0.7 mm thick synthetic fibers. The conductivity of synthetic fibers themselves was typically in the range of 40–80 × 10⁻³ W/mK. As can be seen, there is a substantial difference in the thermal conductivity between the fabric used by Collin et al. and the fabric S coated with polymeric mixtures containing aerogel, even though the fabric S was made of synthetic fibers (meta-aramid polyamide– imide yarns). The overall thermal conductivity of composites may be modified by aerogel, which is characterized by the lowest thermal conductivity found for solids (in this case, only 12 × 10⁻³ W/mK^, according to the data sheet) [23]. Similar results were obtained in the present study—9.8 × 10⁻³ W/mK (Figure 8). Furthermore, the thermal conductivity in the range of 19–23 × 10⁻³ W/mK was reported by the authors of a patent on an aerogel blanket made by impregnating the surface-treated glass felt with silica sol [37]. In the present work, aerogel application in a textile–polymeric composite led to a 4–5-fold increase in the thermal conductivity, reducing its resistance to heat transfer. This may be explained by the compromised structure of aerogel, which affected its properties. It should be noted that Shang et al. [38] observed changes in the thermal conductivity of silica aerogel composites under the influence of high temperature. In that case, the thermal conductivity was 19.1 × 10⁻³ W/mK at the room temperature and 48.9 × 10⁻³ W/mK at 500°C.

It was observed that the thermal resistance of composites with the aerogel coating was similar to that of the uncoated meta-aramid fabric and much lower (by approximately 40%) than that of the fabric coated with polymeric dispersion without aerogel. The thermal resistance of composites with aerogel coatings was approximately 25–29 × 10⁻³ K∙m²/W (Figure 8). The exception was the composite coated with a mixture containing the flame retardant FR 1, the thermal resistance of which was significantly lower at 18.6 ± 4.5 × 10⁻³ Km²/W (p = 0.0117). The thermal resistance measurements have been characterized by the values of standard deviation from 2.60 × 10⁻³ K to 4.77 × 10⁻³ K·m²/W, which corresponded to 3.34–26.27% of the measured value.

Xiong et al. [39] determined the thermal insulation performance of nonwoven aerogel-encapsulated composites using an Alambeta tester. The aerogel composites consisted of three layers: a support layer, a base layer on the bottom, and a face layer on the surface. Aerogel granules were encapsulated in the support layer. Xiong et al. observed that the composites with aerogel-filled voids exhibited a slight decrease in the thermal conductivity (by approximately 5–9 × 10⁻³ W/mK) and a small rise in the thermal resistance (by approximately 20–50 × 10⁻³ Km²/W) [29]. Analysis indicated that nonwoven aerogel-encapsulated composites significantly enhanced the thermal insulation. The effects of aerogel on thermal insulation properties were confirmed in the presented study: the thermal conductivity of composites was higher than that of the fabric, while their thermal resistance was lower or similar, depending on the composition of coating. As can be seen, the coating of textiles could reduce the heat convection and consequently lower the thermal resistance.

In a study investigating nonwoven polyester/polyethylene fabrics treated with aerogel, Venkataraman et al. [40] found the thermal resistance to be a function of fabric thickness. They reported that the thicker samples exhibited the higher thermal resistance, while the observed lower heat losses were associated with the space insulated by the fabric (which may be attributed to aerogel particles present in the fabric). Results of the present study on composites showed that the aerogel coatings are consistent with those findings. Here, it was also observed that the higher the aerogel content, the greater the material thickness (0%—0.59 mm, 7%—0.81 mm, 14%—1.01 mm) (Figure 1) and the higher the thermal resistance (0%—14.9 × 10⁻³ W/mK, 6%—18.6 × 10⁻³ W/mK, 7%—27.7 × 10⁻³ W/mK) (Figure 8). At aerogel content (14%), no additional increment in the thermal resistance was obtained (25.7 × 10⁻³ W/mK) despite a further increase in thickness. Probably, this may be due to the achievement of a certain level of aerogel filling of composite surfaces.