Anisotropic porous ceramic material with hierarchical architecture for thermal insulation

Hollow silica spheres (D₅₀ = 5.268 μm, figure S1(a) (https://stacks.iop.org/BB/17/015002/mmedia)) were purchased from Sigma-Aldrich Tech. Co., Ltd, USA. Solid silica spheres (D₅₀ = 4.872 μm, figure S1(b)) were purchased from Aladdin Chemistry Co., Ltd., China. Hydroxypropyl cellulose (HPC, Mw = 100 000) was purchased from Shanghai Macklin Biochemical Co., Ltd, China. Polydimethylsiloxane (PDMS, Sylgard 184) was purchased from Dow corning, USA.

The hollow spheres were received with a wide size distribution. After dispersing all the spheres in water, the ultra large ones were removed by a separatory funnel. The relatively smaller spheres were dried in an oven before use. The treated hollow or solid spheres were mixed with deionized water to form suspensions of silica spheres. HPC was added as an additive (10 wt.% of the silica spheres). Before freezing, the suspensions were mixed for 24 h and vacuumed to remove air bubbles.

For making samples with the lamellar porous structure, the suspension was poured into the square Teflon tube with a PDMS wedge and then directionally frozen by a cryogenic ethyl alcohol bath at −90 °C. When preparing samples with the random porous structure, the suspension was poured into the square Teflon box. After sealing with Teflon, the suspension was immersed in liquid nitrogen. After the suspension was entirely frozen, the sample was tapped out of the mold and freeze-dried for more than 48 h at −60 °C with a freeze dryer under 0.05 mbar pressure (Labconco 8811, Kansas City, USA). The freeze-dried sample was transferred into a muffle furnace (KSL-1700X, Hefei KeJing Materials Technology Co., Ltd, China). The samples made with hollow and solid spheres were sintered under 600 °C for 3 h and 1200 °C for 2 h respectively to obtain similar final porosity.

The size of the hollow and solid silica spheres was measured with a Coulter LS 13320 (Beckman Coulter, Fullerton, CA, USA). The morphology of the porous silica ceramics was observed by scanning electron microscopy (SEM, S-3500, Hitachi, Tokyo, Japan) at an acceleration voltage of 5 kV. The volume (V) was measured by Vernier caliper, and the weight (m) was measured by analytical balance. The density (ρ) equals weight (m) divided by volume (V). The porosity (P) equals (1 − ρ/ρ_s), where ρ_s is the density of solid silica [25]. For each parameter, more than 5 samples were measured.

The thermal conductivity was measured with a Hot Disk TPS 2500 S instrument in the transient anisotropic mode. The specific heat capacity was also measured by the Hot Disk TPS 2500 S instrument. All samples were cut and polished into blocks of roughly 15 × 15 × 7 mm. At least five samples were tested for each condition. As for the statistic thermal insulation, the samples were heated up by a hot stage (INSTEC mK2000, USA). Infrared images were captured using a Fotric 226 thermal imager.

Compressive properties of the porous silica ceramics were measured in the perpendicular and parallel directions by an Instron 5944 testing system at a displacement rate of 1 mm min⁻¹. The Young's modulus was the slope of the stress–strain curve in the elastic region. All samples were cut and polished into blocks of roughly 15 × 15 × 7 mm. At least five samples were tested for each condition.

The fabrication route is illustrated in figure 1. A bidirectional freeze-casting method was utilized [26–28]. Briefly, hollow silica spheres were dispersed in water with HPC as a dispersant. The hollow silica spheres were chosen here for their low density (∼1.1 g cm⁻³) and excellent thermal insulation property. The suspension with 15 vol% of hollow silica spheres was then poured into a plastic mold, separated by a low thermally conductive PDMS wedge (figure 1(a)). When cooling down the cold source, temperature gradients were generated on both the horizontal (ΔT_H) and vertical (ΔT_V) directions. With ice crystals grew preferentially along the temperature gradients, hollow silica spheres were expelled into the space between the ice crystals to replicate the lamellar ice pattern. The frozen sample was then freeze-dried and sintered at 600 °C for 3 h. The sintering condition was carefully chosen to remove the polymer additives and sinter the boundary between the spheres without destroying their hollow structure (figure S2). The linear shrinkage of the material after sintering is around 20%. The optical image in figure 1(b) shows a large-size porous silica ceramic (roughly 90 × 60 × 7 mm, see also figure S3 for the detailed fabrication process), indicating possible scale-up fabrication which is essential for practical applications.

Zoom In Zoom Out Reset image size

As shown in figure 1(c), the as-prepared anisotropic porous silica ceramic displays a hierarchical architecture. Long-range aligned lamellar layers are observed in the cross-section parallel to the cooling stage. The orientation of the lamellar layers is quantitatively measured based on the SEM image (figure 1(d)), showing excellent alignment of the layers. In addition, each lamellar layer is composed of densely packed hollow silica spheres (figures 1(e) and (f)). The hollow feature of the spheres after sintering fairly changes comparing to the as-received sphere (figure 1(g)).

In order to study the effect of the anisotropic hierarchical architecture on the thermal insulation and mechanical properties, silica ceramics with random porous structures made of hollow and solid silica spheres were fabricated for comparison. These three kinds of samples are illustrated in figure 2(a) as 'lamellar hollow', 'random hollow', and 'random solid', respectively. Solid silica spheres with similar size to the hollow spheres were chosen for proper comparison. The porosity of three samples after sintering were kept similar (∼81%). The random porous structure and lamellar layers are both remained after sintering (figure S2).

Zoom In Zoom Out Reset image size

The effect of the anisotropic hierarchical structure on the thermal insulation properties was investigated. Three kinds of samples were polished into blocks (roughly 15 × 15 × 7 mm) and placed on a hot stage to compare their thermal insulation property with the lamellar layers aligning parallel to the stage (figures 2(a) and (b)). A series of infrared images were taken when the stage was heated from 50 to 250 °C (figures 2(b) and S4). All the images were taken when the sample surface temperature was stabilized. The temperatures of the sample surface and the stage were then measured based on the infrared images (figure S4). The corresponding absolute temperature differences (|∆T|) between the sample surface and the stage were also calculated and summarized in figure 2(c). At the whole temperature range, the anisotropic porous silica ceramic ('lamellar hollow') shows a larger temperature difference, indicating a better thermal insulation property.

Quantitatively, the room temperature thermal conductivity of the 'lamellar hollow', 'random hollow', and 'random solid' porous silica ceramics in the perpendicular direction to the lamellar layers was measured as 29.7 ± 1.7, 60.7 ± 5.7, and 99.7 ± 2.8 mW m⁻¹ K⁻¹, respectively (figure 2(d)). The porous silica ceramic made from hollow spheres shows better thermal insulation than that made from solid spheres because of the entrapped air in the hollow structure. The superior thermal insulation in the 'lamellar hollow' silica ceramic could be attributed to its long-range aligned lamellar structure, although it was made from the same number of hollow spheres as in the 'random hollow' silica ceramic. Besides, upon heating from 20 to 150 °C, the thermal conductivity of the 'lamellar hollow' silica ceramic increased slightly from 29.7 ± 1.7 mW m⁻¹ K⁻¹ to 30.9 ± 1.3 mW m⁻¹ K⁻¹, indicating its excellent thermal insulation property under high temperature (figure S5).

We note that attempts have been made to develop porous thermal insulating ceramics by the traditional freeze-casting technique [29, 30]. However, the resulting scaffolds exhibited limited anisotropic thermal conductivity. Interestingly, by applying the bidirectional freeze-casting technique, the 'lamellar hollow' porous silica ceramic shows higher anisotropic thermal conductivity than the above porous ceramics as well as the 'random hollow' and 'random solid' ceramics (figure 2(e) and table S1).

To further study the relationship between the thermal insulation property and the anisotropic hierarchical architecture, porous silica ceramics with lamellar and random structures ranging in a wide distribution of porosity were fabricated for the thermal conductivity test. The room-temperature thermal conductivity of the porous silica ceramics was correlated with their porosity in the perpendicular direction (figure 2(f)). For the whole range of porosity (71%–83%), the silica ceramics with the lamellar porous structure show a better thermal insulation than those with the random porous structure.

We depict the heat transfer process of silica ceramics with lamellar and random porous structures in figure 2(g). Theoretically, the thermal conductivity of a cellular material can be thought as the sum of thermal conduction, convection, and radiation [31]. Convection is important only when the pore size reached millimeter scale. As the pore sizes of the silica ceramics are as small as several microns, the convection can be suppressed. As radiation is less effective at low temperature, we assume that the heat radiation at room temperature is almost the same for two structures. The lamellar layers extend the heat conduction path, so the heat conduction of the lamellar sample is much lower than the random sample, resulting in a lower thermal conductivity of the lamellar sample than the random sample.

The effect of the anisotropic hierarchical architecture on the mechanical properties was also investigated (figures 3(a)–(c) and S6). The compressive strength and the Young's modulus of the 'lamellar hollow' samples in the parallel direction are 1.90 ± 0.15 MPa and 87.5 ± 11.0 MPa, respectively, superior to those tested in the perpendicular direction (0.67 ± 0.09 MPa and 29.6 ± 2.7 MPa, table S2). The compressive strength and the Young's modulus of the 'lamellar hollow' samples in the parallel direction are higher than those of the 'random hollow' and 'random solid' samples (table S2). Besides, the mechanical properties of the 'lamellar hollow' silica ceramics show higher anisotropy than those of the 'random hollow' and 'random solid' samples (figures 3(b) and (c)). In addition, silica ceramics with lamellar and random porous structures ranging in a wide distribution of porosity were fabricated for further compression test in the parallel direction. The compressive strength is correlated with their porosity (figure 3(d)). For the whole range of porosity (71%–83%), the silica ceramics with the lamellar porous structure show higher compressive strength than those with the random porous structure. These results indicate that the anisotropic hierarchical structure can simultaneously improve both the mechanical and thermal insulation properties.

Zoom In Zoom Out Reset image size

The thermal conductivity, compressive strength, and working temperature of the anisotropic porous silica ceramics are compared with other thermal insulation materials (figures 3(e) and (f)) [32–41]. Polymer foams and polymer/ceramic hybrid foams show similar thermal conductivity as our porous silica ceramics, but the moderate mechanical properties and inflammability have severely hindered their practical applications [32–35]. Porous ceramics with similar porosity show higher working temperature and higher mechanical properties, but they are usually less thermal insulating and much more expensive than our porous silica ceramics [36–41]. Our anisotropic porous silica ceramics show a good combination of low thermal conductivity and high mechanical properties, making it a promising candidate for thermal insulation, especially in high temperature conditions.

The fire-resistant property of the anisotropic porous silica ceramic (roughly 30 × 30 × 12 mm; porosity: ∼81%) was demonstrated by exposing it to an alcohol lamp, which generated a high temperature as 600 °C. As shown in movie S1, the porous silica ceramic was fire-resistant under 600 °C. As many building fires caused even higher temperature, the fire-resistant property of the porous silica ceramic was also explored by exposing it a butane blowlamp (10 cm from the material surface), which generated a temperature as high as 1300 °C (figure 4(a)). While burning, the back-side temperature was recorded by an infrared camera in a distance of 50 cm (figure 4(b) and movie S2). As shown in figure 4(c), the temperature is much lower than 1300 °C after burning for 10 min, showing a good high temperature thermal insulation property. When cooling down to the room temperature, the front surface of the porous silica ceramic cracked while the back side remained unchanged (figure 4(d)). The structural change after firing is also shown by the SEM images in figure 4(e). The hollow silica spheres are melted and densified on the front side while unaffected on the back side. These results show that the porous silica ceramic is thermal insulating and fire-resistant under high temperature, which means the porous silica ceramic can be applied in a wide range of temperatures.

Zoom In Zoom Out Reset image size

To demonstrate the potential application of our material, a large piece of anisotropic porous silica ceramic (roughly 90 × 60 × 7 mm) was fabricated and put on a 3D printed mini-house as the roof. Two thermocouples were used to record the roof and indoor temperatures, respectively (figure 5). The initial roof and indoor temperatures were both around 18 °C. When the roof was quickly heated to 48 °C by a solar simulator with a light intensity of 1000 W m⁻², the indoor temperature fairly changed, resulting in a large temperature difference (∆T = 28 °C, figure 5(b)). Upon cooling by liquid nitrogen (around 1 cm above the roof), both the roof and indoor temperature dropped down significantly and stabled at −14 and 9 °C, respectively (∆T = 23 °C, figure 5(d)). These demonstrate that the anisotropic porous silica ceramic can stabilize thermal condition against environmental temperature fluctuation by insulating heat exchange.

Zoom In Zoom Out Reset image size