Thermo-Mechanical Characterization of Gypsum–Nano Silica Modified Fired Clay Bricks at Elevated Temperatures

3.1 Characterization of synthesized Silica Nanoparticles

Figure 2 presents the SEM image of the SNPs along with the corresponding particle size histogram. The nanoparticles exhibit a predominantly spherical morphology with nanometric dimensions. The histogram indicates an average particle diameter of 127.1 ± 18.6 nm, demonstrating a narrow size distribution. The calculated dispersion of 11.8% is slightly higher than that reported for pure silica, yet still indicates good size uniformity. The EDS spectrum (Figure 3) confirms the elemental composition of the nanoparticles, with silicon and oxygen as the dominant elements. Quantitative analysis reveals weight percentages of 47.99% for Si and 52.01% for O, and atomic percentages of 32.4% and 67.6%, respectively, confirming a near-stoichiometric SiO₂ composition.

Figure 2. Scanning electron microscopy image SNPs (left) and corresponding particle size histogram (right)

Figure 3. Scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM–EDS) image SNPs

The DLS results of the synthesized SNPs are shown in Figure 4. The size distribution exhibits an average hydrodynamic diameter of 128.5 nm with a standard deviation of 37.79 nm. Although slight differences exist between SEM and DLS values, both measurements fall within the acceptable margin of error. This variation is attributed to the hydrodynamic nature of DLS measurements, which account for surface-bound organic species. Figure 5 shows the zeta (ζ) potential of the SNPs, with a measured surface charge of −32.5 mV. This negative charge arises from hydroxyl (–OH) groups on the silanol-rich nanoparticle surface [14]. Zeta potential values equal to or exceeding ±30 mV indicate good colloidal stability, which is consistent with reported standards for stable dispersions [15]. All measurements were performed in triplicate to ensure reliability.

Figure 4. Dynamic light scattering (DLS) spectra of SNPs

Figure 5. Zeta potential measurement of SNPs

Despite the favorable zeta potential, the SNP suspension exhibited limited colloidal stability, with visible precipitation occurring within a few minutes. Figure 6 illustrates the UV–Vis absorption spectra of SNP solutions at different concentrations. The absorption maximum at 282.1 nm was used to construct the corresponding calibration curve, also shown in Figure 6.

Figure 6. Absorption at 282.1 nm of SNPs (left); Calibration curve constructed from the absorption data (right)

Figure 7. Fourier transform infrared spectroscopy of SNPs and its characteristic absorption peaks

In this study, the XRD test was used to determine the crystal structure, crystal lattice parameters, phase, crystal size of silica, and the elements and compounds contained in the synthesized silica sample. The diffractogram analysis process used Match software to match the obtained data with X-ray diffraction standards for materials [18]. These X-ray diffraction standards are called The Joint Committee on Powder Diffraction Standards (JCPDS/ 29-0085) Based on Figure 8, nanosilica particle with a crystalline structure was obtained. There is Si present at various angles, with the highest intensity peak at 28.33° which corresponds to the silicon phase, has a lattice parameter a = 5.4410 Å, a cubic structure, and a crystal size of 127.6923 nm [18]. Furthermore, SiO₂ is also present at various angles. These peaks correspond to the stishovite phase, with lattice parameters a = 4.1605 Å, b = 4.1294 Å, c = 7.4211 Å, β = 101.375°, and a monoclinic structure. One of these peaks has high intensity, specifically the peak at 31.47° and the peak at 45.20°. At the 31.47° angle, the crystal size is 122.6124 nm, while at the 45.20° angle, the crystal size is 131.1275 nm.

Figure 8. X-ray diffraction of SNP from rice husk

X-Ray Fluorescence (XRF) analysis technique is a technique for analysing a material using a spectrometer equipment emitted by samples from X-ray irradiation. The results of the analysis are given in Table 3.

Table 3. XRF analysis of rice husk–based nano-silica, commercial gypsum, and gypsum–nano silica modified fired clay bricks at elevated temperatures

3.2 Characterization of fired clay brick samples

Thermomechanical Properties: The specimens were evaluated for their physical and mechanical properties in accordance with ASTM-2018 standards after firing [20]. The tests included Linear Shrinkage (LS), Water Absorption (WA), Apparent Porosity (AP), Bulk Density (BD), and Modulus of Rupture (MOR). Four brick compositions were investigated, and the results are summarized in Table 4. For each composition, five specimens were prepared and fired at 900℃, 1000℃, and 1100℃, resulting in a total of 60 samples. The physico-mechanical properties of the gypsum–nano silica modified fired clay bricks are presented in Table 4.

Table 4. Thermo-mechanical properties of gypsum–nano silica modified fired clay bricks at elevated temperatures

Linear shrinkage varied systematically with nanosilica content and firing temperature. For all mixtures, shrinkage increased with increasing temperature, with the highest values observed at 1100℃. This trend is primarily attributed to enhanced sintering and vitrification of the ceramic bodies at elevated temperatures. Although linear shrinkage is not standardized under Indian Standards (BIS), it remains an important industrial parameter for ceramic classification. Porous ceramics typically exhibit shrinkage values of ~3%, semi-porous ceramics 4–6%, and vitrified ceramics around 8%. Based on these criteria, the studied bricks fall largely within the semi-porous category.

As shown in Table 4, water absorption decreased with increasing firing temperature, reflecting improved densification of the brick matrix. An inverse relationship was also observed between nanosilica content and water absorption, indicating that nanosilica contributes to pore refinement. The obtained water absorption values are suitable for red ceramic applications. Water absorption and bulk density are closely interrelated; higher firing temperatures promote grain growth, stronger interparticle bonding, and pore filling, thereby reducing water uptake [13].

Apparent porosity, which is directly influenced by firing temperature, showed only minor variation across the tested conditions and correlated well with water absorption and density trends [15]. Increasing nanosilica content did not result in increased porosity at any firing temperature. Instead, specimens containing nanosilica exhibited slightly lower apparent porosity than the reference bricks (R (G0–S0)). Although specimens with nanosilica showed marginally lower bulk density than the reference, higher firing temperatures consistently reduced porosity and water absorption while enhancing overall densification.

The Modulus of Rupture (MOR) generally increased with firing temperature and nanosilica addition, reflecting improved microstructural cohesion due to phase transformations in the clay matrix [21]. At temperatures above 900℃, vitrification occurs through the formation of a liquid glassy phase that fills pores and enhances bonding through capillary action. This process is accompanied by shrinkage and depends strongly on firing temperature, duration, and composition. According to Souza Santos, kaolinite-based ceramics reach maximum firing shrinkage near 950℃, while vitrification occurs between 950℃ and 1225℃ due to the release of cristobalite (SiO₂) and subsequent glass formation. Despite this, some nanosilica-containing samples showed slightly reduced MOR, likely due to localized phase changes that limited effective particle bonding. Flexural strength values ranged from 2.05 ± 0.016 MPa to 3.05 ± 0.21 MPa, satisfying the ASTM/BSI minimum requirement of ≥ 2.07 MPa for semi-silica and moderately refractory bricks. Comparable or higher MOR values have been reported for clay bricks and bricks incorporating glass waste, perlite waste, or recycled mullite [21-23].

Compressive strength remains a critical parameter for structural applications. The reference brick exhibited the highest compressive strength (35 MPa). With increasing nanosilica content, compressive strength gradually decreased due to increased pore volume, with the 10% nanosilica specimen approaching the permissible lower limit. However, bricks incorporating both gypsum and nanosilica showed a slower rate of strength reduction, maintaining higher compressive strength than bricks containing nanosilica alone at equivalent replacement levels. Thermal conductivity values measured at room temperature ranged from 0.58 to 0.76 W/m·K, indicating acceptable thermal performance for building applications.

The thermal properties, including thermal conductivity, specific heat, and thermal diffusivity, are presented in Table 4. The incorporation of gypsum alone (G5–S0) reduced density by only 1.63% and increased porosity by 2.82% compared with the reference brick (G0–S0). These minor changes had a negligible effect on thermal conductivity, which is primarily governed by porosity. In this study, thermal conductivity decreased slightly for G5–S0, G5–S5, and G5–S10 relative to G0–S0. The obtained values fall within the typical range reported for conventional bricks (0.39–0.67 W/m·K) and are consistent with literature data for clay–glass waste bricks (0.54–0.58 W/m·K), marble-sludge-based ecological bricks (0.4–0.54 W/m·K), and pure clay bricks (~0.7 W/m·K).

Thermal diffusivity decreased with increasing nanosilica content, with G5–S10 exhibiting a lower value than the reference G5–S0. This indicates that heat propagation was approximately 21% slower due to nanosilica incorporation, resulting in improved insulation performance. As reported in the literature, thermal conductivity and insulation capacity are strongly influenced by porosity characteristics, including pore type (open or closed), size, and uniformity of distribution.

Temperature–time curves obtained under direct flame exposure are shown in Figure 9. During the initial 5 minutes, all samples exhibited rapid heating, followed by temperature stabilization on the exposed surface. Peak temperatures of approximately 950℃ were recorded for G5–S0 and G5–S5, while the sample with higher nanosilica content (G5–S10) reached a lower average temperature of about 910℃. On the rear face, temperatures remained below 350℃, corresponding to a thermal gradient of approximately 700℃. Comparable studies on refractory ceramics produced from coal ash waste reported insulation levels of ~500℃ over 25 minutes, while kaolinite–mullite ceramics achieved thermal gradients of ~600℃ during 30-minute tests. Gypsum–nanosilica composites are known to withstand temperatures up to 1700℃ and exhibit good resistance to direct flame exposure, confirming their suitability as thermally insulating and refractory construction materials.

With the thermal conductivity results and the temperature record during direct fire exposure, the heat flow graph for G5-S0 and G5-S10 is shown in Figure 10. From the heat flow results, three stages can be highlighted. The initial stage, which lasts approximately the first five minutes of the test, where the temperature is maintained below 200℃. Subsequently, in stage two, from 5 to 10 minutes, the heat transfer begins to decrease for G5-S0 but remains constant for G5-S10. The last stage, after 10 minutes, is considered to be when temperature stabilization is reached. In the case of G5-S10, the heat transfer during the total test time was lower than that recorded for G5-S0. The heat flow in this study is slightly higher than that reported for ceramics from kaolin-mullite, who obtained heat flows between 2.3 and 3.8 J [24]. In this study, the samples were thicker and also more porous than the samples used.

Figure 10. Heat transfer of samples directly exposed to fire with the thermal conductivity results and the temperature

In Figure 11, the marks left on bricks G5-S0, G5-S5, and G5-S10 by direct contact with fire are observed. For G5-S0, the mark is much larger and darker, which infers that there is a greater degree of deterioration in the piece, while in the mark for samples with gypsum and nanosilica G5-S5 and G5-S10, it is observed that the higher the nanosilica content, the smaller the mark. When the test ended, it was observed that the G5-S0 piece had cracked. The better thermal performance in the samples with gypsum and nanosilica is attributed to the existence of phases such as mullite, anorthite, gehlenite, wollastonite and cristobalite with refractory properties previously found in the XRD for sample G5-S10.

Figure 11. Samples after the flame resistance test

Microstructural and Phase characterization: After completing the compressive strength test, the remaining fragments of the crushed brick samples were taken to observe the morphology or grain size of the samples. The samples were tested using a Scanning Electron Microscope. The samples taken were in the form of fine grains, as the SEM can only accommodate samples measuring 1 mm. The grain size of the brick samples at 1100℃ was determined at 2000x magnification, and the grain size was calculated for comparison. This can be seen in Figure 12. Examination of the figures reveals crystalline structures and pore formations, which were also detected in the XRD analyses.

Figure 12 (G5–S5 and G5–S10) shows spherical nanosilica particles uniformly distributed within the brick matrix, appearing at the microscale. In the fired specimens sintered at 1100℃, the pore sizes are predominantly below 50 μm, with most pores measuring less than 10 μm.

Figure 12. SEM images of produced samples: (a) Reference brick [R (G0–S0)], (b) G5–S0 brick, (c) G5–S5 brick, (d) G5–S10 brick in particular, the brick body with nanosilica addition has a microporous structure due to the hollow spherical shapes of the nanosilica

As the gypsum and nanosilica content increases, a pronounced fluxing effect is observed, which reduces melt viscosity and enhances fluidity during firing. This promotes effective pore filling and a consequent reduction in pore number and size. Moreover, increasing the sintering temperature enhances particle rearrangement and mass transport, resulting in improved densification and further pore refinement.

Energy-dispersive X-ray spectroscopy (EDS) analysis (Figure 13) of the binder regions reveals distinct elemental distributions for all four compositions fired at 1100℃. In the reference sample, silicon and oxygen peaks dominate due to the silicate-rich clay matrix. In contrast, the G5–S0 specimen exhibits a pronounced calcium peak accompanied by a reduced silicon signal, indicating the presence of CaSO₄ derived from gypsum addition. For nanosilica-doped samples (Si 5% and 10%) combined with 5% gypsum, higher silicon concentrations are detected alongside strong calcium signals. This confirms the coexistence of CaSO₄ and nanosilica phases and the formation of a composite microstructure.

Figure 13. EDS analysis of the produced samples: (a) Reference brick [R (G0–S0)], (b) G5–S0 brick, (c) G5–S5 brick, (d) G5–S10 brick

These microstructural and compositional observations provide a clear basis for correlating the physicochemical characteristics with the enhanced mechanical performance observed in the gypsum–nanosilica modified brick samples.

Suction: Regarding the suction test, the results obtained from the laboratory tests (Table 5) were analysed and processed in order to obtain the variance, standard deviation and coefficient of variation for the standard brick and each of the bricks with substitution percentages in order to obtain statistical data with better understanding and comparison. Regarding the evidence, and based on what was mentioned above with respect to ASTM C62/C216, it shows that the standard bricks have the lowest average suction index (14.14 gr/200cm^₂/min) compared to the other controls analysed, being the bricks with the addition of 10% of nanosilica who presented the highest suction index (81.835 gr/200cm²/min) which represents the accelerated process of water suction during settling. Percentages of addition similar to those in this research resulted in significant improvements in compressive strength. According to the National Building Code of India (NBC) [25], bricks containing gypsum and nanosilica significantly outperformed standard bricks in terms of suction characteristics, with a value of 14.138 gr/200cm²/min. These bricks were categorized as type I bricks.

Table 5. Suction of masonry units

XRD characterization of the fired samples: XRD analyses were used to determine the crystalline phases formed in relation to sintering temperatures and increasing gypsum (0%, 5%) and nanosilica (0%, 5%, 10%) content. Figure 14 shows the XRD graphs of the experimental samples sintered at 1100℃. Examination of the figure reveals that with increasing gypsum and nanosilica addition, the formation of Ca-containing phases begins. In the sample coded A (anorthite: (CaO·Al₂O₃·2SiO₂), M (mullite: 3Al₂O₃ 2SiO₂), C (gehlenite (2CaO·Al₂O₃·SiO₂), W (wollastonite: CaO·SiO₂) phases are present. From the results it was concluded that the prepare brick samples shows chemical compatibility between the clay matrix, gypsum and nanosilica doping. During firing, gypsum experienced dehydration and decomposition, providing CaO that responded with aluminosilicate phases in the clay matrix [26]. At high temperatures, expected chemical reactions as shown below:

xAl₂O₃⋅SiO_2$+$yCaO$+$SiO_{2$\to $}yCaO⋅xAl₂ O₃⋅SiO₂

The existence of nanosilica improves silica availability and helps liquid-phase sintering, leading to enhanced compression, density and pore refinement. 

Figure 14. XRD graphs of the produced materials (A: anorthite, M: mullite, C: cristobalite, W: wollastonite)

FT IR analysis of the brick samples: The FTIR spectra (Figure 15) reveal pronounced structural modifications in the brick samples as a function of nanosilica (G) and gypsum (S) concentrations. The broad absorption band in the 3620–3300 cm⁻¹ region, along with the peak at approximately 1650 cm⁻¹, corresponds to O–H stretching and H–O–H bending vibrations, respectively, indicating increased hydration and chemically bound water in the G5–S5 and G5–S10 samples [27]. From the image was identified that there a high intense band between 1100 to 1000 cm^-1, indicating the Si–O–Si / Si–O–Al Stretching band, by the additive nanosilica increases the silicate frame work in the composition, the peak around 1100 cm^-1 suggesting improved structural polymerization. The incorporation of nanosilica enhances the silicate framework, as evidenced by the intensification and slight shift of the Si–O–Si stretching bands (1130–1030 cm⁻¹), suggesting improved structural polymerization [28]. Other peaks and correspond functional groups are listed in the Table 6. From the data it was confirmed the additive may promote the development of complex grouping and modifying the geochemical fingerprint to prepare brick sample.