Low-temperature sintering of high-alumina ceramics assisted by recycled PV glass waste: Densification, sintering behavior, mechanical, dielectric properties and microstructural characteristics

2.1. Preparation of recycled PV glass waste and raw materials

Recycled PV glass waste was obtained from decommissioned solar panels collected at nearby recycling facilities. To soften the ethylene-vinyl acetate (EVA) interlayer, the front tempered-glass sheets were first mechanically separated from the encapsulated panel structures, followed by thermal delamination at 120°C. The recovered glass was then thoroughly cleaned with ethanol and deionized water to remove residual contaminants and adhesives. After drying for at 100°C for 2 h, the glass fragments were crushed and milled for 8 h at 300 rpm using a high-alumina planetary ball mill with a ball-to-powder ratio of 10:1. The result powder was sieved through a 90 µm stainless-steel mesh to obtain a homogeneous particle fraction suitable for ceramic processing. The recycled PV glass, primarily composed of SiO₂ (∼71 wt%), Na₂O (∼11 wt%), CaO (∼9 wt%), and minor oxides (Al₂O₃, MgO, and Fe₂O₃) was used as a liquid-phase sintering aid in this study [21,22]. High-purity α-AlO₃ powder (≥ 99%, Merck) with an average particle size of 1-3 µm was used as the main ceramic matrix. An aqueous polyvinyl alcohol (PVA, 5 wt%) solution was employed as a temporary binder to improve the cohesion of green bodies during forming and handling.

2.2. Formulating samples and mixing powders

Four ceramic formulations, designated G1, G2, G3, and G4, were prepared with PV glass content of 5, 10, 15, and 20 wt%, respectively as summarized in Table 1. The required amounts of PV glass and alumina powders were wet-mixed with PVA binder in ethanol using a planetary ball mill (Retsch PM 100, Germany) at 200 rpm for 4 h to ensure homogeneous dispersion. After drying at 90°C for 12 h in convection oven (Memmert UF 160, Germany), the slurry was gently ground in an agate mortar to break up agglomerates. The resulting powder mixture was then passed through a 90 µm sieve to ensure a uniform particle size distribution. This procedure was carefully controlled to prevent local enrichment of glassy phases, which could otherwise lead to excessive vitrification or abnormal grain growth during sintering.

2.3. Pre-sintering and compaction procedures

The dried powder mixtures were uniaxially cold-pressed using a hydraulic press (Specac Series Manual Hydraulic Press, Specac Inc.) at a pressure of 10 MPa into rectangular plates measuring 100 × 100 × 3 mm³. To reduce defects and improve green density, the green compacts underwent a controlled binder burnout and hot-pressing treatment prior to final sintering as shown in Figure 1(a). Binder burnout was carried in air by heating to 600°C at a rate of 3°C min⁻¹ and holding for 1 h. Subsequently, as illustrated in Figure 1(b), the samples were hot-pressed using a vacuum hot press (GWL-VSF-RY, Luoyang Juxing Kiln Co., Ltd, China) at 1000°C under an applied pressure of 5 MPa for 30 min. After this stage, the external pressure was fully released before further heating to higher temperatures. This two-step approach minimizes residual stress accumulation, prevents microcracking and springback effects, and promotes uniform densification during the subsequent sintering process. The intermediate hot-pressing step also reduces open porosity and facilitates initial diffusion bonding, thereby enhancing densification efficiency at reduced temperatures.

Close

Figure 1. Sintering regimes of high-alumina ceramics incorporating PV glass waste: (a) dwell-time schedules at 1400°C (2-6 h) for all compositions; (b) intermediate hot-pressing step at 1000°C under 5 MPa for 30 min; and (c) complete heating-cooling profiles, including binder burnout, hot pressing, temperature ramping (5°C min⁻¹) to 1400°C, dwell stages, and furnace cooling.

Export to PPT

2.4. Sintering schedule and thermal profiles

After hot press, the specimens were sintered in a programmable high-temperature furnace. The temperature was increased from 1000°C to 1400°C at a controlled heating rate of 5°C min⁻¹. Upon reaching 1400^oC, the samples was held for different dwell time (2, 3, 4, 5, and 6 h) to systematically investigate the influence of soaking duration on densification behavior. As shown in Figure 1, the complete thermal profiles consist of three main stages: (i) binder burnout at 600°C, (ii) hot pressing at 1000°C under 5 MPa for 30 min, and (iii) final sintering at 1400°C with controlled dwell times. Figure 1(c) illustrates the full heating-cooling path, while Figure 1(a) highlights the dwell-time regimes at the peak temperature. After sintering, the furnace was allowed to cool naturally to room temperature under closed conditions to minimize thermal gradients and avoid thermal shock. The sintered samples were labeled according to PV glass content and dwell time. For subsequent characterization, the specimens were ground and polished using SiC abrasive papers (400-2000 grit), followed by final polishing with a 0.05 µm alumina suspension to obtain smooth surfaces.

2.5. Mechanical and physical testing method

Linear shrinkage, bulk density, apparent porosity, and water absorption were measured in accordance with ASTM C373-18. Mechanical properties of the alumina ceramic samples were evaluated using standardized methods. Flexural strength of bar specimens (3 × 4 × 40 mm³, support span = 30 mm) was measured using a three-point bending test with a crosshead speed of 0.5 mm min⁻¹, via a Shimadzu AG-X Plus 10 kN universal testing machine. Vickers microhardness (HV1) was determined using a Mitutoyo HM-200 tester under a load of 1 kgf for 10 s, with at least five indentations averaged for each sample. Young’s modulus (E) and shear modulus (G) were determined using the ultrasonic pulse-echo method (Olympus Epoch 650) based on longitudinal and transverse wave velocities. Fracture toughness (Kᵢc) was computed using the indentation method based on the Palmqvist crack model. Creep strain was evaluated after exposure at 1000°C for 1 h under a constant load of 10 MPa.

2.6. Electrical properties analysis and structural characteristics

Electrical resistivity and dielectric properties were measured using an LCR meter (Keysight E4980A, USA) over a frequency range of 1 kHz-1 MHz at 25°C. Silver paste electrodes were applied to both surfaces of each polished disc (10 mm × 1 mm) and fired at 500°C for 10 min. Breakdown strength, dielectric constant (ε′), and dielectric loss (tan δ) were measured using a Trek 610E high-voltage test system in accordance with ASTM D149-20.

For structural characterization, XRD patterns were recorded using a Bruker D8 Advance diffractometer with Cu Kα radiation (λ = 1.5406 Å) at 40 kV and 40 mA. The scans were performed over a 2θ range of 5° - 80° () with a step size of 0.02°. Crystallite size and microstrain were estimated using the Scherrer and Williamson-Hall methods. FTIR spectra were obtained using a Nicolet iS50 spectrometer (Thermo Fisher Scientific) in the 400-4000 cm⁻¹ range employing the KBr pellet method. Field-emission scanning electron microscopy (FE-SEM, Hitachi SU8020) in conjunction with energy-dispersive X-ray spectroscopy (Oxford Instruments X-Max 80 mm² EDX) was used to analyze the surface morphology and elemental composition. The grain size distributions were determined from micrographs using ImageJ software.

3.1. Densification and sintering behavior qualities

Figure 2 illustrates the densification behavior of alumina ceramics incorporating recycled PV glass waste (G1-G4) sintered at 1400°C for different holding times (2-6 h). Both PV-glass content and the dwell time significantly influence sintering shrinkage (Figure 2a), bulk density (Figure 2b), apparent porosity (Figure 2c), and water absorption (Figure 2d). The introduction of a small glassyfraction acts as an effective sintering aid at this relatively low temperature, generating a transient liquid phase that enhances viscous flow and particle rearrangement [23-25]. For all compositions, linear shrinkage gradually increased with dwell times, rising from 1.52% for G1 (5 wt% PVWG, 2 h) to 2.59% for G4 (20 wt% PVWG, 6 h), indicating progressively improved densification kinetics. This behavior is attributed to the partial melting of silicate-rich phases within the PV glass, which fills intergranular voids and increases the particle contact during sintering [25,26].

Close

Figure 2. Effect of PV-glass waste content and sintering holding time at 1400°C on (a) sintering shrinkage, (b) bulk density, (c) apparent porosity, and (d) water absorption of high-alumina ceramics.

Export to PPT

This densification trend is further confirmed by the evolution of bulk density. As shown in Figure 2(b), the density increased from approximately 3.48 g cm⁻³ (G1-2 h) to a maximum of 3.73 g cm⁻³ (G3-4 h), followed by a slight decreased with prolonged holding times. These values correspond to relative densities exceeding ∼92-96%, consistent with the high mechanical strength and hardness obtained for the optimized compositions. The minor density reduction after 4 h can be attributed to pore coalescence or localized viscous softening at grain boundaries, which may limit further densification [26,27]. Notably, G2 (10 wt% PVWG) and G3 (15 wt% PVWG) exhibit the highest density values (∼3.64-3.72 g cm⁻³), indicating that the optimal PV-glass content for achieving maximum densification without excessive glass-phase accumulation lies in the range of 10-15 wt%. At 1400°C, liquid-phase wetting between alumina grains enhances diffusion and particle rearrangement, a mechanism commonly associated with viscous flow-assisted sintering of oxide ceramics [24,28]. However, excessive amorphous content (> 15 wt%) may saturate grain boundaries and hinder densification by trapping closed pore within continuous glassy films [29].

The densification behavior is further supported by the inverse relationship between apparent porosity (Figure 2c) and density as well as shrinkage trend. Porosity decreased markedly from 8.71% (G1-2 h) to 5.05% (G4-4 h), followed by a slight increase at longer dwell times. The reduction in open porosity at intermediate stages reflects effective neck growth and pore elimination driven by capillary forces associated with the transient liquid phase [29,30]. However, extended holding times (> 4 h) may promote pore coalescence or stabilization of closed pores due to the viscous nature of the glassy phase, thereby reducing diffusion efficiency. A critical PV-glass fraction enhances densification up to an optimal level, beyond which microstructural uniformity may deteriorate. This is evidenced by the consistently lower porosity observed for G3 and G4 compared to G1 and G2. Such behavior is consistent with previous studies on glass-added alumina systems, where controlled liquid-phase formation improves sinterability at sub-conventional temperatures [31,32].

Consistent with porosity reduction, water absorption values (Figure 2d) also decreased with increasing dwell time and PV-glass content. Specifically, water absorption decreased from 2.95% (G1-2 h) to 1.71% (G4-6 h), indicating improved densification and effective sealing of surface pores by viscous glass infiltration. Samples containing 15 wt% PV glass exhibited the lowest and most stable absorption (< 1.8%), confirming a dense microstructure with minimal open porosity. However, at higher glass content (20 wt%) a slight increase in water absorption was observed after prolonged sintering, suggesting possible phase segregation or microcrack formation during cooling due to thermal expansion mismatch between the glassy phase and alumina [33,34]. Thus, although glass addition enhances densification, excessive amounts may compromise structural integrity through phase inhomogeneity and residual stress development [34].

A comprehensive understanding of the sintering mechanism can be obtained by correlating shrinkage, density, porosity, and water absorption data. In pure alumina systems, near-theoretical density typically requires temperatures above 1600°C. In contrast, the addition of PV-glass shifts the sintering mechanism toward a transient liquid-phase regime at 1400^oC [35]. Among the glass constituents, SiO₂ plays a dominant role as a network former, enabling viscous flow and structural integrity of the liquid phase, while Na₂O and CaO act as network modifiers that reduce viscosity and enhance ionic diffusion. This synergistic effect lowers the local melting temperature at particle contacts and accelerates densification kinetics. However, excessive glass addition (> 15 wt%) may reduce grain-boundary mobility and promote structural relaxation instead of further densification. Therefore, an optimal PV glass content of 10-15 wt% provides sufficient liquid phase to maximize densification while maintaining mechanical robustness and dielectric reliability.

The results demonstrate that recycled PV glass effectively enhances sintering behavior, serving both as a sustainable raw material and as a fluxing additive. The improved densification achieved at 1400°C indicates that sintering temperatures can be reduced by approximately 200°C compared to conventional alumina processing, without compromising structural compactness. This reduction offers both environmental and the economic benefits by lowering energy consumption and valorizing end-of-life PV glass. The findings support the integration of PV-glass waste into high-performance ceramic systems within a circular-economy framework.

3.2. Mechanical performance and structure-property correlation

Figure 3 presents the mechanical response of alumina ceramics containing recycled PV glass waste sintered at 1400°C, illustrating the effects of PV-glass content and holding time on flexural strength (Figure 3a), fracture toughness (Kᵢc, Figure 3b), Vickers hardness (HV₁, Figure 3c), and creep strain (ε, Figure 3d). The results indicate that the incorporation of an optimal glass fraction (10-15 wt%) enhances the mechanical integrity of the alumina matrix, whereas excessive addition (20 wt%) slightly deteriorates performance due to the formation of continuous amorphous films along grain boundaries. This behavior reflects a balance between two competing mechanisms: liquid-phase-assisted densification and glass-induced grain-boundary softening [33]. The interplay between densification behavior and mechanical strengthening highlights the critical role of microstructural evolution governed by PV-glass content and sintering time at 1400°C.

Close

Figure 3. (a) Flexural strength, (b) fracture toughness (Kᵢc), (c) Vickers hardness (HV₁), and (d) creep strain (ε) of high-alumina ceramics sintered at 1400°C with different PV-glass contents.

Export to PPT

Longer holding time and moderate PV-glass addition show a monotonic increase in flexural strength. The strength rises from 330 MPa (G1-2 h) to 375 MPa (G3-6 h), as shown in Figure 3(a). Across all holding times, the G3 composition (15 wt% PV glass) consistently exhibits superior performance, indicating an optimal balance between controlled glass infiltration and enhanced densification. The observed strength improvement of approximately 13-15% compared to the baseline composition (G1) can be attributed to reduced porosity, improved grain packing, and the formation of glass-bridged necks that promote more uniform stress distribution across grain interfaces. However, the flexural strength plateaued at around 358-364 MPa when the PV-glass content increases to 20 wt% (G4). This reduction is associated with the accumulation of viscous amorphous films, which act as crack initiation sites under load and weaken intergranular regions. This observation is consistent with the densification results (Section 3.1), where at longer soaking times, excessive glass content led to slightly lower bulk density and increased closed porosity.

A similar trend is observed for fracture toughness (Kᵢc, Figure 3b), which increases with PV-glass content (up to 15 wt%) and sintering time. The G3 sample reaches a maximum toughness of 4.52 MPa·m^1/2 after 6 h, representing an improvement of nearly 15% compared to the 5% glass sample. Two microstructural effects are principally responsible for the enhancement: (i) crack deflection along heterogeneous glass-alumina interfaces and (ii) viscous-phase bridging at crack tips, which dissipates fracture energy. Microstructural evidence (confirmed later by SEM-EDX) suggests that the triple-junction regions between alumina grains are occupied by fine glass films, creating heterogeneous stress fields that favor crack branching rather than catastrophic propagation [29]. Due to the dominance of compliant amorphous networks, which may localize deformation and reduce overall fracture resistance, toughness stabilizes or slightly decreases when the PV-glass content exceeds the optimal level (> 15 wt%). The effect of holding time further indicates that sufficient soaking (4-6 h) allows homogenization of the glassy phase and promotes the formation of stronger glass-alumina bonding interfaces, enhancing stress transfer and delaying fractures initiation [29,33].

In contrast, the hardness evolution (Figure 3c) shows a slightly decreasing trend PV-glass content. After 6 h, Vickers hardness decreases from 1440 HV₁ (G1) to 1372 HV₁ (G3) and 1330 HV₁ (G4). The reduction is consistent with the substitution of rigid α-Al₂O₃ bonding networks by softer SiO₂-rich glassy phases. Nevertheless, the hardness remaining relatively high (> 1300 HV₁) indicating a dense alumina skeleton is preserved even at 1400°C. The presence of glassy films may also enhance fracture tolerance by mitigating brittle cleavage along sharp grain edges, despite slightly lowering local indentation resistance [29,33]. The findings confirm that the observed mechanical strengthening at moderate PV-glass levels is primarily driven by improved densification and intergranular bonding rather than intrinsic phase hardening. A microstructural balance between densification and viscous-phase redistribution at long sintering times is suggested by the G3 and G4 samples’ stable but lower hardness values, whereas the G1 and G2 samples only exhibit slight hardness variation over holding times.

Creep strain (ε) data (Figure 3d) provide further insight into high-temperature deformation behavior. The creep strain increases gradually from 0.061% (G1-2 h) to 0.088% (G4-6 h) with increasing PV-glass content and holding time. This trend reflects the growing fraction of amorphous silicate phase, which facilitates time-dependent viscous deformation under load. The partial grain boundary softening due to glass-phase wetting is linked to the higher creep compliance observed in the G3 and G4 samples. However, the overall creep levels remain within acceptable limits for structural ceramics, indicating that improved densification compensates for the softening effect of the glass phase. Thus, the introduction of PV glass imparts a limited degree of viscoelasticity, which contributes to energy dissipation during stress relaxation and crack propagation, thereby enhancing fracture toughness.

The macroscopic mechanical trends are further supported by elastic moduli derived from experimental data. Both Young’s modulus (E) and shear modulus (G) decrease gradually with increasing PV-glass content as shown in Tables 2, reflecting the incorporation of a more compliant amorphous phase. After 6 h, G decreases from 129 GPa to 121 GPa and 117 GPa, and E decreases from approximately 312 GPa (G1) to 284 GPa (G3) and 274 GPa (G4), respectively. This behavior is consistent with previous studies on glass-modified ceramics, where silicate phases reduce the overall stiffness of the composite. Meanwhile, Poisson’s ratio (v) shows a slight increase, as indicated in Table 2, from 0.255 (G1) to 0.270 (G3) and 0.273 (G4), indicating a marginal increase in lateral deformability. This reflects the growing contribution of viscoelastic deformation mechanisms and intergranular shear accommodation associated with the glassy phase.

The combined evolution of hardness, fracture toughness, and flexural strength highlights the dual role of PV-glass phase in controlling the structure-property relationship. The transient glass phase enhances densification, refines the microstructure, and improves mechanical reliability by sealing pores and strengthening grain boundaries at moderate levels (≤ 15 wt%). In contrast, excessive glass content (> 15 wt%) leads to the formation of continuous amorphous films, while only marginally affecting toughness due to crack-deflection effects. The optimal microstructure therefore consists of a percolated alumina grain network interconnected by discrete glassy bridges, ensuring both rigidity and flaw tolerance. This is consistent with the densification plateau observed at 4-5 h (Figure 2b), where the best balance of strength, toughness, and hardness is achieved. Additionally, prolonged sintering holding time (> 6 h) offers limited benefits and may induce over-softening due to structural relaxation of the glass phase.

These results demonstrate that the mechanical properties of high-alumina ceramics can be effectively tailored at a relatively low sintering temperature of 1400°C through the incorporation of recycled PV-glass waste. The SiO₂-Na₂O-CaO glass system activates liquid-phase sintering, compensating for the inherently slow diffusion kinetics of α-Al₂O₃ and formation of dense, mechanically robust ceramic at reduced energy cost. The optimal processing window (10-15 wt% PV-glass, 4-5h) provides a favorable combination of increased toughness (Kᵢc ≈ 4.5 MPa·m^1/2), acceptable hardness (> 1370 HV₁), and high flexural strength (∼375 MPa), demonstrating the effectiveness of PV glass as both a fluxing and reinforcing additive. This approach not only supports sustainable material utilization but also opens new opportunities for the production of high-performance alumina ceramics at lower temperatures.

3.3. Dielectric properties and electrical behavior of PVWG-modified alumina ceramics

As a function of holding time and glass content, Figure 4 illustrates the changes in (a) electrical resistivity (ρ), (b) dielectric constant (ε′), (c) dielectric loss (tan δ), and (d) dielectric breakdown strength (E₍_b₎) of alumina ceramics containing different proportions of recycled PV glass waste sintered at 1400°C. These characteristics provide critical into the evolution of phase composition, charge transport mechanisms, and microstructural homogeneity induced by liquid-phase sintering. As discussed previously, the incorporation of PV-glass promotes densification and partial amorphization at grain boundaries, significantly influencing the electrical and dielectric responses of the ceramics. The results reveal a delicate balance between improved insulation due to densification and the increased conductivity arising from the glass phase, both governed by composition and thermal treatment.

Close

Figure 4. Dielectric and electrical properties of high-alumina ceramics sintered at 1400°C with varying PV-glass content: (a) Electrical resistivity, (b) Dielectric constant (ε′), (c) Dielectric loss (tan δ), and (d) Dielectric breakdown strength.

Export to PPT

As shown in Figure 4(a), electrical resistivity (ρ) decreases gradually with increasing PV-glass content and sintering holding time. For the baseline sample G1 (5 wt% PVWG), resistivity decreases from 1.73 × 10¹⁴ Ω·m (2 h) to 1.33 × 10¹⁴ Ω·m (6 h). In contrast, G3 (15 wt%) and G4 (20 wt%) exhibit more pronounced reduction, reaching 0.42 × 10¹⁴ Ω·m and 0.19 × 10¹⁴ Ω·m, respectively. This trend indicates that the introduction of an amorphous phase rich in SiO₂, NaO₂, and CaO facilitates the formation of localized conductive pathways along grain boundaries. A portion of alkali and alkaline earth ions (Na⁺, Ca²⁺) diffuse into the alumina interfacial regions during the liquid-phase sintering process, generating charge carriers that enhance ionic and polaronic conduction. Despite this reduction, the resistivity remains sufficiently high (> 10¹³ Ω·m for ≤ 15 wt% glass), confirming the excellent insulating nature of the ceramics. The resistivity decreases with longer holding times (2-6 h) further reflects improved densification and reduced grain boundary potential barrier. However, continuous amorphous films may be encouraged by an excessive amount of glass (> 15 wt%) or prolonged heat exposure, which would raise the possibility of space-charge accumulation and leakage conduction. These results are consistent with the well-known behavior of glass-ceramic composites, in which the resistivity properties are determined by the interaction of intergranular glass viscosity, ion mobility, and structural relaxation.

Both the PV-glass content and the holding time cause a moderate increase in the dielectric constant (ε′), as shown in Figure 4(b). The glass addition improves the overall dielectric polarization capability, as evidenced by values ranging from 9.5-9.9 for G1 to 10.4-10.8 for G4. This increase can be attributed to (i) the presence of amorphous silicate phases, which have a higher ionic polarizability than pure alumina, and (ii) better densification, which reduces the depolarization effect by minimizing pore volume. Moreover, charge accumulation at alumina-glass interfaces causes the interfacial polarization (Maxwell-Wagner-Sillars effect) to intensify with increasing glass content, which adds to the overall permittivity enhancement. In the G3 and G4 samples, where well-distributed glass films produce polarizable boundary layers without resulting in excessive electrical leakage, these effects are especially noticeable. The consistent rise in ε′ over the course of holding time suggests that the microstructure develops toward increased homogeneity and better glassy phase network continuity. Crucially, the obtained ε′ values (9.8-10.8) are either equal to or marginally higher than those of commercial high-alumina dielectrics sintered at ≥ 1600°C [1,2,25], indicating that controlled PV-glass incorporation can achieve effective polarization behavior at a significantly lower processing temperature.

Dielectric loss (tan δ), shown in Figure 4(c), provides insight into energy dissipation mechanisms. For all compositions, the tan δ values stay low (< 7.3 × 10⁻³) and progressively rise with glass content and holding time, in a similar order to that seen for resistivity reduction. The presence of mobile ions (Na⁺, Ca²⁺) in the glass phase that take part in localized hopping conduction under an alternating electric field and improved interfacial polarization relaxation are responsible for this increase. The exceptional dielectric integrity of materials is confirmed by the minimal dielectric loss (< 10⁻²) observed in all samples. This suggests that bulk transport is not dominated by conductive ion migration, which is limited to interfacial zones. The dielectric loss at 1 kHz for G3 (15 wt%) stays between 5.24 × 10⁻³ and 6.35 × 10⁻³, achieving a balance between low energy dissipation and high permittivity. In contrast, the G4 samples exhibit somewhat higher losses (∼6.5-7.2 × 10⁻³) as a result of improved ionic motion and increased glass network connectivity. These findings demonstrate how, depending on its distribution and proportion, the glass phase can act as a dielectric relaxor as well as a densification promoter. Similar behavior has been documented in earlier studies on SiO₂-CaO-Al₂O₃-modified alumina systems, where the optimized glass fraction effectively tunes the dielectric without sacrificing insulation.

The positive impact of moderate PV-glass addition is further supported by the dielectric breakdown strength (E₍_b₎), which is shown in Figure 4(d). E₍_b₎ reaches 27.3 kV/mm for the optimally sintered G3 sample (15 wt% PVWG, 4-5 h), which is an improvement of almost 15% over the baseline G1 composition (24.1 kV/mm). This enhancement arises from the reduction of microstructural defects, such as pores and intergranular voids, which act as preferential sites for electric field concentration and dielectric failure. During sintering, the PV-glass phase promotes viscous flow, sealing any remaining pores and creating a smoother, more uniform microstructure that postpones the start of electrical breakdown. E₍_b₎ for G4 does, however, slightly decrease after the 15 wt% addition, reaching about 26 kV/mm at 6 h. The development of continuous low-resistivity glassy paths and the potential volatilization of alkali constituents, which locally weaken the insulating network, are associated with this behavior. Similar to the densification kinetics, the breakdown strength varies with holding time, reaching its maximum values approximately 4-5 h after the microstructure reaches its maximum homogeneity and minimum porosity. Extended soaking can cause structural coarsening and glass relaxation, which lowers dielectric robustness.

The complex relationship between the electrical polarization-conduction mechanisms and the microstructural evolution in these composites is demonstrated by the combined electrical and dielectric data. The alumina grains control the electrical behavior at low PV-glass levels (5-10 wt%), producing the high resistivity and moderate permittivity characteristic of conventional alumina ceramics. When the glass content reaches 15 wt%, intergranular glass films improve dielectric uniformity and ionic polarization, which promote densification and inhibit conduction caused by defects. Over 15 wt% glass creates interconnected amorphous networks that increase dielectric loss and decrease resistivity. According to these observations, the ideal performance window (10-15 wt% PV-glass, 4-5 h sintering) is associated with a microstructure in which thin, irregular glass layers that function as dielectric bridges rather than conductive channels firmly bond alumina grains. These interfacial microstructures preserve the mechanical robustness outlined in Section 3.2 while promoting high breakdown strength and controlled dielectric loss.

The structure-property correlation framework provides a better understanding of how dielectric properties interact. Porosity and defect concentration have an inverse relationship with the dielectric constant (ε′) and breakdown strength (E₍_b₎), while the amorphous phase fraction and mobile ion concentration have a positive correlation with dielectric loss (tan δ). Up to a critical threshold, densification dominates over conductivity increase, as evidenced by the concurrent observation of higher ε′ and E₍_b₎ with moderate glass addition. The shift from isolated glass pockets to percolative conduction networks is confirmed by the decrease in resistivity and slight increase in tan δ above 15 wt%. In oxide glass–ceramic composites, where the percolation limit usually occurs at 15-20 vol% amorphous phase, such a crossover has been extensively documented. Furthermore, the idea that both sets of properties are controlled by intergranular glass redistribution rather than bulk chemical alteration is supported by the consistency of the mechanical and electrical trends. Therefore, optimizing multifunctional performance requires adjusting the recycled glass phase’s viscosity, chemistry, and spatial distribution.

These results also show the current approach’s impact on sustainability and technical novelty. High-density alumina-based ceramics with competitive dielectric and mechanical performance can be fabricated at a sintering temperature 200-300°C lower than traditional methods thanks to the addition of recycled PV-glass waste. An important development in low-energy ceramic processing is the simultaneous achievement of ε′ ≈ 10, tan δ < 7 × 10⁻³, ρ > 10¹³ Ω·m, and E₍_b₎ ≈ 27 kV/mm at 1400°C. By converting PV waste into electronic ceramics with added value, this result demonstrates how controlled glass-phase engineering not only improves functional performance but also aids circular economy initiatives [16]. Such composites show promise for insulating substrates, dielectric spacers, and high-voltage protective housings where both structural and electrical reliability are necessary due to their balanced electrical behavior and mechanical stability.

3.4. Microstructural characteristics of high alumina ceramic

Examining the microstructural and phase evolution of the alumina-PV glass composite ceramics at the microscopic and atomic scales is essential for understanding the mechanisms governing densification, mechanical strengthening, and dielectric behavior discussed in Sections 3.1-3.3. Since both sintering behavior and functional properties exhibited clear composition- and time-dependent trends, representative samples were selected to capture the contrasting structural states associated with suboptimal and optimal sintering conditions. In this study, particular emphasis was placed on samples sintered for 4 h, as this dwell time corresponds to the maximum densification stage identified in Figure 2, where bulk density reaches its peak and porosity is minimized. Therefore, this condition provides the most meaningful basis for correlating microstructure with macroscopic properties. Among the investigated compositions, two representative samples were selected for detailed microstructural characterization: G1 (5 wt% PVWG, 4 h) and G3 (15 wt% PVWG, 4 h). These compositions correspond to the lower and optimal PV-glass contents, respectively, as demonstrated in Section 3.1. While intermediate compositions (G2 and G4) exhibit similar trends, their structural features fall between these two extremes and do not provide additional mechanistic insights beyond those captured by G1 and G3. Thus, focusing on these two samples enables a clearer interpretation of the role of PV glass in governing microstructural evolution.

The G1 sample represents a low-glass system with limited liquid-phase formation, showing relatively higher porosity, lower dielectric constant (ε′ ≈ 9.6), and moderate mechanical strength (∼335 MPa). In contrast, the G3 sample (15 wt% PVWG) exhibits optimal performance combining high densification (bulk density ≈ 3.73 g·cm⁻³), improved flexural strength (∼375 MPa), and stable dielectric properties (ε′ ≈ 10.2, E_b ≈ 27 kV·mm⁻¹). These contrasting characteristics make G1 and G3 ideal model systems to elucidate how controlled addition of PV-glass-derived phases influences phase composition, crystallographic features, and grain boundary morphology of high-alumina ceramics sintered at 1400°C.

By correlating the results from XRD, FTIR, and SEM-EDX analyses with the macroscopic property trends, the specific roles of PV-glass-derived silicate phases in promoting densification, stabilizing the corundum structure, and tuning dielectric response can be clearly identified. Such a multi-scale structural understanding is essential for validating processing-structure-property relationships and for establishing optimized compositional and thermal parameters for the low-temperature fabrication of sustainable alumina-based ceramic systems.

3.4.1. Crystallographic integrity, microstrain evolution, and defect structures in PV-glass-modified alumina ceramics

Regardless of glass content, the XRD patterns of the raw Al₂O₃ powder, the sintered samples G1 (5 wt% PVWG), and G3 (15 wt% PVWG) at 1400°C for 4 h (Figure 5) show the characteristic reflections of corundum (PDF 00-046-1212), confirming that α-Al₂O₃ remains the dominant crystalline phase after sintering. Despite the presence of reactive silicate melts derived from PV glass waste, the thermal stability of α-Al₂O₃ is evidenced by the retention of all major diffraction peaks, including (012), (104), (110), (113), (024), (018), (214), and (10 10), without detectable phase transformation. The PV glass predominantly segregates into amorphous intergranular regions rather than forming crystalline aluminosilicates, as evidenced by the absence of secondary crystalline silicate phases. This observation is consistent with the FTIR results (Section 3.4.2), which reveal appearance of Si-O-Al vibrations characteristic of amorphous aluminosilicate networks.

Close

Figure 5. XRD patterns of Al₂O₃ raw material and high-alumina ceramic samples G1 and G3 after 4 h of sintering at 1400°C compared with standard corundum diffraction data (PDF#46-1212).

Export to PPT

G1 and G3 exhibit significantly different crystallite sizes, according to peak broadening analysis using the Scherrer equation (Table 3). The average crystallite size decreases from approximately 68 nm in G1 to approximately 46 nm in G3, demonstrating that increasing PV glass content suppresses crystallite growth. This behavior is consistent with the well-known grain-boundary pinning effect of amorphous silicate phases, in which diffusion-controlled crystal growth is inhibited by the presence of a glassy phase. According to SEM-EDX mapping, the 15 wt% PV glass in G3 produces a more continuous intergranular amorphous film, which lessens crystallite coarsening and encourages finer microstructural features. Microstrain (ε) and dislocation density (δ) were computed to obtain a better understanding of defect evolution. With an average crystallite size of 68.043 nm, G1-4h has a δ ≈ 2.16 × 10⁻⁴ nm⁻², while G3-4h has a much higher δ ≈ 4.78 × 10⁻⁴ nm⁻² because of its smaller crystallites. This trend is consistent with classical defect theory, where dislocation density increases with decreasing crystallite size. A more strained lattice for the higher PV glass composition is indicated by the microstrain values, which support this interpretation: ε rises from ≈1.85 × 10⁻³ in G1 to ≈2.44 × 10⁻³ in G3. Chemical substitution (Si⁴⁺ entering near-surface Al³⁺ sites), interfacial stress between alumina grains and the silicate film, and the inhibition of diffusion pathways during sintering are the causes of this strain. These elements work together to produce increased internal lattice distortion in G3, which is consistent with FTIR and SEM data showing a larger Si-rich glass network.

Table 3. XRD data and crystallite size of corundum in samples of G1-4h and G3-4h.

Sample	Diffraction planes	Intensity measured (I)	FWHM (Radian)	2θ (°)	Nanoparticles crystalline size D (nm)	Average NPs crystalline size D (nm)
G1-4h	104	1531.9860	0.0023557	35.37222	61.776	68.043
	116	1510.9470	0.0021561	57.68797	73.405
	113	1176.6750	0.0021519	43.55826	69.379
	012	686.50770	0.0022315	25.82195	63.745
	024	605.19380	0.0021519	52.74757	71.909
G3-4h	104	1835.09700	0.0030898	35.17119	47.073	45.721
	113	1460.14600	0.0034610	43.37270	43.110
	116	1408.44100	0.0031177	57.51999	50.725
	012	966.21640	0.0037764	25.60772	37.651
	024	586.80420	0.0030898	52.57765	50.046

The lattice parameters a and c were calculated using the (104), (116), (113), (012), (024) reflection. The relative comparisons are still significant even though the exact corundum geometry is not recovered by these simplified expressions. The calculated a ≈ 0.263 nm and c ≈ 0.4555 nm for G1-4h increase slightly to a ≈ 0.2642 nm and c ≈ 0.4576 nm for G3-4h (Table 4). This minor expansion suggests the presence of elastic strain at alumina-glass interfaces, consistent with the higher ε observed in G3. Si incorporation at surface sites, residual hydroxyl species in the boundary glass (as indicated by FTIR), and local misfit stresses brought on by amorphous film confinement can all cause this kind of lattice dilatation.

Table 4. Structural parameters derived from XRD analysis of the G1 and G3 samples sintered at 1400°C for 4 h.

Samples	Diffraction plane (hkl)	Microstrain, ε (×10⁻3)	Dislocation density, δ (×101⁵ m⁻2)	a (Å)	c (Å)	Cell volume, V (Å3)	Atomic packing fraction (APF)	Positional parameter, (u)
G1-4h	(104)	1.84	0.262	2.930	5.071	37.74	0.51	0.348
	(116)	0.98	0.186	1.839	3.187	9.34	0.51	0.348
	(113)	1.35	0.208	2.399	4.156	20.72	0.51	0.348
	(012)	2.44	0.246	3.980	6.892	94.60	0.51	0.348
	(024)	1.09	0.193	2.002	3.467	12.04	0.51	0.348
	Average values	1.54	0.219	2.630	4.555	34.888	0.51	0.348
G3-4h	(104)	2.43	0.451	2.944	5.101	38.29	0.50	0.351
	(113)	2.18	0.538	2.410	4.173	21.00	0.50	0.351
	(116)	1.42	0.389	1.846	3.197	9.43	0.50	0.351
	(012)	4.16	0.706	4.005	6.938	96.40	0.50	0.351
	(024)	1.56	0.399	2.005	3.472	12.09	0.50	0.351
	Average values	2.35	0.497	2.642	4.576	35.44	0.50	0.351

This interpretation is supported by the calculated cell volumes (V), using V = 0.866 a²c for rhombohedral structure. The volume increases from 0.0377 nm³ in G1 to 0.0384 nm³ in G3. The relative trend, a systematic increase with PV glass addition, is completely consistent with experimental indicators of increased structural disorder and lattice distortion, despite the fact that these values are smaller than the corundum values reported in the literature (V ≈ 0.254 nm³). Additional structural parameters, including Al-O bond length (L), positional parameter (u), and atomic packing fraction (APF), were computed based on standard corundum relations. Compared to reference α-Al₂O₃ values (a₀ = 0.4758 nm, c₀ = 1.299 nm, APF ≈ 0.55, u ≈ 0.352, L ≈ 0.193 nm), both G1 and G3 show slight deviations. A mild decrease in packing efficiency linked to increased lattice strain and amorphous boundary formation is indicated by the calculated APF for G1 (≈0.51), which drops slightly to ≈0.50 for G3. From ≈0.348 in G1 to ≈0.351 in G3, the positional parameter u shows a slight increase, indicating minor disruptions in the symmetry of the Al site. The Al-O bond length also increases from ≈0.190 nm in G1 to ≈0.193 nm in G3, reflecting bond stretching associated with interfacial stresses. The previously reported macroscopic properties are directly affected by these structural evolutions. In line with strengthening mechanisms based on crack-tip shielding and grain-boundary deflection, the superior flexural strength and fracture toughness of G3 sample are a result of its higher defect density and microstrain. In keeping with the defect-dipole and interfacial polarization mechanisms discussed in Section 3.3, the increased lattice distortion and amorphous phase simultaneously lead to a higher dielectric constant but a marginally higher dielectric loss.

Dislocation density (δ) and microstrain (ε), derived from XRD analysis, provide a quantitative link between crystallography and macroscopic characteristics. Grain refinement (smaller D in G3) results in a strong increase in δ because dislocation density scales roughly as 1/𝐷². More internal defects and grain-boundary area per unit volume are indicated by a higher δ; these serve as electrical inhomogeneities (increased interfacial polarization) and mechanical strengthening sites (by preventing dislocation motion and encouraging crack deflection/tip blunting mechanisms). In line with the observed increase in fracture toughness (Kᵢc) for G3, microstrain ε, which reflects lattice distortions and residual stress, contributes to local stress fields that raise the energy needed for crack propagation. In ceramics, the mechanical strength σ can be empirically linked to defect density and inverse grain size through Hall-Petch-like or fracture mechanics relationships: higher intergranular bonding (caused by glassy films) and higher δ increase the effective fracture resistance. Both δ and ε electrically increase the number of polarization centers and widen relaxation spectra, which increases the relative permittivity ε′ (through Maxwell-Wagner interfacial polarization) and marginally increases dielectric loss (tanδ) due to energy dissipation by defect-associated dipoles. All things considered, a balanced increase in δ and ε, as seen for the ideal PVWG content (∼10-15 weight percent), results in a good trade-off: mechanical reinforcement through microstructural toughening and densification, with only a slight dielectric loss penalty, which is consistent with the experimental trends discussed in Sections 3.1-3.3.

The XRD results confirm that recycled PV glass waste significantly influences crystallite size, microstrain, lattice dimensions, and related structural parameters without affecting the stability of the α-Al₂O₃ phase. The mechanistic framework put forth in this study is strongly supported by these results: In order to facilitate simultaneous densification and property optimization, PVWG-derived silicate films control sintering kinetics, limit crystallite coarsening, improve boundary cohesion, and modify defect structures. The best range for achieving high-performance, low-temperature-sintered alumina ceramics is 10-15 wt% PV glass, according to the synergistic agreement of XRD, FTIR, SEM-EDX, mechanical testing, and dielectric measurements.

3.4.2. Vibrational modes and bonding structure evolution in PV-glass-modified alumina ceramics

The FTIR spectra of the raw Al₂O₃ powder, G1 (5 wt% PVWG), and G3 (15 wt% PVWG) samples sintered at 1400°C for 4 h (Figure 6) reveal the evolution of chemical bonding and vibrational modes induced by the incorporation of recycled PV glass waste. Regarding the formation of a secondary silicate-rich intergranular phase and its function in microstructural consolidation, these spectral features directly corroborate the interpretations previously made from densification behavior (Section 3.1), mechanical performance (Section 3.2), and dielectric response (Section 3.3).

Close

Figure 6. FTIR spectra of the raw Al₂O₃ powder and ceramic samples G1 (5 wt% PVWG) G3 (15 wt% PVWG) after sintering at 1400°C for 4 h.

Export to PPT

The FTIR spectrum of raw Al₂O₃ is dominated characteristic corundum lattice vibrations, especially the strong Al-O-Al bending and stretching bands between approximately 500 and 750 cm⁻¹. The strong and well-organized octahedral coordination of Al³⁺ in α-Al₂O₃ is reflected in these features. Crucially, the raw powder shows virtually no discernible bands above 1000 cm⁻¹, indicating that there are no silicates or impurities containing hydroxyls. After sintering, both PVWG-containing samples (G1 and G3) show significant changes to this baseline spectrum following sintering, suggesting that the recycled glass adds new bonding environments and structural motifs that are absent from pure alumina.

The formation of aluminosilicate linkages is confirmed for the G1 sample (5 wt% PVWG) by the appearance of weak but distinct Si-O-Al and Si-O stretching modes in the ∼980-1100 cm⁻¹ range. These vibrations, which are consistent with the introduction of a small amount of amorphous glassy phase, are caused by SiO₄ tetrahedra interacting with AlO₆ units at grain boundaries. These modes are consistent with still-visible porosity of the G1 sample (Figure 2) and the moderate increase in densification and mechanical strength that was noted in that material, suggesting that 5 wt% PVWG only produces a thin, irregular silicate network. Similarly, the H-O-H bending near ∼1600 cm⁻¹ and the small O-H absorption near ∼3400-3500 cm⁻¹ indicate that a small number of hydroxyl groups are still trapped in the residual glassy phase, which is typical of recycled soda-lime glass precursors.

The FTIR spectrum of G3 sample (15 wt% PVWG), on the other hand, shows significantly stronger and more defined Si-O-Al and Si-O stretching bands in the 950-1050 cm⁻¹ region, along with more distinct shoulders on either side of the alumina lattice peaks. This significant rise in silicate-associated vibrations suggests that the amorphous phase derived from PVWG is more abundant and structurally integrated into the alumina framework. A continuous aluminosilicate network is implied by the strong coupling between [SiO₄]^4- and [AlO₆]^9- units, which is consistent with the remarkably high densification efficiency and reduced open porosity reported for G3 in Section 3.1. These bonding features also support the improved mechanical performance of G3 (Section 3.2), as intergranular aluminosilicate films are known to enhance grain bonding, promote crack deflection, and suppress intergranular decohesion.

Furthermore, the broader and more intense O-H absorption bands observed in G3 compared to G1 indicate a higher residual hydroxyl content within the intergranular glass phase. This finding supports the notion that alkali-rich silicate species that have a propensity to incorporate moisture are introduced by PV glass decomposition, as evidenced by the increased Na and Si found by EDX. These glassy hydroxyl species might also play a role in dielectric behavior of G3 sample (Section 3.3), specifically the controlled rise in dielectric loss and the moderate increase in ε′, which reflect increased dipolar activity at grain boundaries.

The FTIR results clearly demonstrate that recycled PV glass waste not only acts as a fluxing agent but also induces significant chemical reconfiguration of the alumina matrix through the formation of Si-O-Al networks. A composition-dependent shift from isolated silicate clusters, which are insufficient for strong densification, to a coherent intergranular film, which is ideal for mechanical and dielectric performance, is highlighted by the contrast between G1 and G3. The superior performance window found for alumina ceramics containing approximately 10-15 wt% PVWG is explained at the molecular level by these bonding changes, which also directly support the microstructural evidence from SEM–EDX (Section 3.4.3).

3.4.3. PVWG-Modified alumina ceramics: microstructural development and elemental partitioning

High-resolution SEM imaging and EDX elemental mapping were employed to further elucidate the microstructural evolution of alumina-PV glass composites sintered at 1400°C in order to help clarify the structural causes of the densification and property variations described in Sections 3.1-3.3. Representative samples G1 (5 wt% PVWG, 4 h) and G3 (15 wt% PVWG, 4 h) were selected to compare low and optimal glass contents under identical densification conditions. Figures 7-9 show significant variations in grain morphology, intergranular bonding, and elemental distribution that together support the macroscale mechanical and dielectric behaviors.

Close

Figure 7. SEM micrographs of high-alumina ceramics sintered for 4 h at 1400°C demonstrating how PV glass content affects packing density and grain morphology at 10,000x and 5,000x magnification, (a-b) Sample G1 (5 wt % PVWG) shows partially bonded grains and residual intergranular pores; (c-d) Sample G3 (15 wt % PVWG) shows densely packed equiaxed grains, improved boundary cohesion, and reduced porosity, all of which are signs of liquid-phase-assisted densification.

Export to PPT

Close

Figure 8. EDX characterization of the G1 alumina ceramic (5 wt % PVWG) sintered for 4 h at 1400°C (a) SEM morphology; (b–f) elemental maps of O-K, Na-K, Al-K, and Si-K demonstrating a restricted and irregular distribution of silicates; (g) EDX spectrum and quantitative composition table demonstrating the predominance of Al and O with trace Si and Na coming from the recycled PV glass phase.

Export to PPT

Close

Figure 9. EDX characterization of the G1 alumina ceramic (5 wt % PVWG) sintered for 4 h at 1400°C (a) SEM morphology; (b–f) elemental maps of O-K, Na-K, Al-K, and Si-K demonstrating a restricted and irregular distribution of silicates; (g) EDX spectrum and quantitative composition table demonstrating the predominance of Al and O with trace Si and Na coming from the recycled PV glass phase.

Export to PPT

Both G1 and G3 display the distinctive equiaxed grain morphology of sintered α-Al₂O₃ at 10,000× magnification (Figures 7a and c); nevertheless, notable differences in grain compactness and boundary quality are visible. It should be noted that the grain sizes observed in SEM are in the micrometer range and represent agglomerated structures composed of numerous nanoscale crystallites formed during sintering. In which, each nanoscale crystallite has the sizes, calculated from the XRD pattern (∼45-68 nm) represent crystalline scattering domains with a uniform lattice structure. In line with its higher measured porosity (∼6.5%) and lower shrinkage (<2%), the G1 microstructure exhibits more loosely packed grains with discernible intergranular voids and partially developed necks. On the other hand, the G3 sample exhibits minimal residual porosity, smoother, well-bonded interfaces, and densely interlocked grains. These characteristics confirm that 15 weight percent PV glass facilitates improved liquid-phase sintering, which accelerates particle rearrangement and encourages viscous flow at grain boundaries. These processes are directly responsible for the previously reported superior densification (bulk density ≈ 3.73 g·cm⁻³) and mechanical strengthening (σ ≈ 371-375 MPa).

At lower magnification of 5,000× (Figures 7b and d), G3 shows a more homogeneous grain-size distribution without abnormal growth or microcracking. It is important to emphasize that the micrometer-sized grains observed in SEM consist of aggregates of nanoscale crystallites identified by XRD analysis, highlighting the hierarchical nature of the microstructure. Maintaining high fracture toughness (Kᵢc ≈ 4.4-4.5 MPa·m^1/2) requires such microstructural homogeneity because the lack of heterogeneities inhibits preferential crack propagation pathways. However, G1 sample has lower strength (∼330-335 MPa) and toughness (∼3.7-3.8 MPa·m^1/2) are probably caused by microstructural discontinuities and residual micro-voids, which is consistent with limited liquid-phase formation at only 5 weight percent PV glass.

EDX analyses (Figures 8 and 9) provide further insight into elemental. The spectra of both samples are dominated by Al and O, indicating that the α-Al₂O₃ matrix is the main crystalline phase. However, there are noticeable variations in the spatial distribution and presence of Si and Na, which are important indicators of the PV glass fraction. In G1, Si is present at low level (≈2 at%) distributed discontinuously, indicating limited formation of silicate boundary films. This is consistent with the moderate dielectric properties (ε′ ≈ 9.6-9.8, tan δ < 4.5×10⁻³) and the incomplete densification seen in the SEM images.

In contrast, G3 exhibits significantly higher Si content (≈9-10 at%, Figure 9) with a more uniform distribution, forming continuous nanoscale silicate networks along grain boundaries. The presence of Na and Ca as network modifiers further enhances the formation of a viscous liquid phase, promoting densification and intergranular bonding. These microstructural features explain the improved densification and multifunctional properties of G3. The improvement in mechanical performance and the controlled increase in dielectric permittivity (ε′ ≈ 10.1-10.3) without excessive dielectric loss (tan δ ≈ 5-6×10⁻³) can be explained by the dual functions of these boundary films as densifying agents and dielectric modifiers. The thin and well-distributed silicate films act as both densification promoters and dielectric modifiers, enhancing mechanical strength and maintaining stable dielectric behavior without forming conductive pathways, which is consistent with its comparatively high breakdown strength (> 26 kV·mm⁻¹). However, excessive glass addition (e.g., G4) may lead to over-thickened boundary films, phase segregation, or microcrack formation during cooling, which can degrade mechanical integrity.

The relationships between process, structure, and property that were established during the course of the study are validated by the SEM-EDX results. The change from G1 to G3 shows how gradually increasing the amount of recycled PV glass changes the ceramic microstructure by creating regulated silicate boundary phases that improve densification, smooth out grain bonding, and adjust dielectric characteristics. These results validate the viability of employing PV glass waste as a sustainable sintering aid for high-performance alumina ceramics, in addition to clarifying the mechanisms underlying the enhanced multifunctional properties. In contrast, the absence of microcracking or excessive grain growth indicates that the amount of PV glass stays below the threshold that would otherwise introduce harmful glass pools or low-resistance phases, even though the increased glass content in G3 greatly strengthens microstructural cohesion. This finding is consistent with previous mechanical data that demonstrated that performance starts to deteriorate at 20 wt% PV glass (G4), most likely as a result of excessive amorphous phase accumulation that reduces the grain–grain load transfer capacity. The conclusion that 10-15 wt% PV glass defines an ideal composition for low-temperature sintering at 1400°C, balancing boundary fluidity, structural integrity, and dielectric stability, is thus strongly supported by the microstructural evidence.

3.5. Discussion and correlation between sintering behavior, microstructure, and functional properties

The performance of PV-glass-modified high-alumina ceramics sintered at 1400°C is governed by a closely related structure-property-processing relationship, as shown by the combined results from Sections 3.1-3.4. The densification behavior (Section 3.1) provides a fundamental framework for interpreting mechanical reinforcement, dielectric response, and microstructural evolution across the G1-G4 compositions. Through the creation of a transient silicate-rich liquid phase, PV glass consistently encourages a sintering mechanism that is more effective. Significant increases in sintering shrinkage and bulk density result from this, especially in compositions containing 10-15 wt% PVWG, as it encourages particle rearrangement, improves mass transport, and supports pore filling. For these compositions, the simultaneous decrease in open porosity and water absorption suggests a more efficient grain boundary sealing process, producing ceramics with enhanced structural compactness. However, at longer holding times, the 20 wt% PVWG sample (G4) shows decreasing densification, indicating liquid phase oversaturation that prevents capillary-driven consolidation. These observations highlight that PV-glass-assisted liquid-phase sintering is beneficial only within an optimal compositional range.

The densification trends are a strongly correlated with mechanical performance (Section 3.2). The G3 composition (15 wt% PVWG) achieves the best balance between liquid-phase content and the integrity of alumina skeleton, resulting in maximum flexural strength, fracture toughness, and near-optimal Vickers hardness. A group of complementary mechanisms are responsible for the improvements: (i) improved packing and finer grain distribution as a result of viscous flow during sintering; (ii) the development of thin, glassy intergranular films that prevent catastrophic cracks from spreading; and (iii) the decrease in critical flaw size made possible by decreased porosity. Together, these factors increase crack-growth resistance (reflected in K_IC) and crack-initiation stress (reflected in flexural strength). In the meantime, too much glass in G4 confirms the negative effects of too many amorphous phases by softening the matrix, decreasing stiffness (E and G), and increasing creep strain. XRD-derived microstructural parameters further support the mechanical trends: higher dislocation density (δ), increased microstrain (ε), and decreased crystallite size (D) in G3 compared to G1 indicate enhanced lattice distortion and grain boundary relaxation, which are known to promote crack deflection and energy dissipation during mechanical loading. Thus, the combined effects of densification kinetics, grain-boundary chemistry, and lattice-scale modifications induced by recycled PV glass directly govern the observed mechanical behavior.

Microstructure and PV-glass content have a similarly strong influence on the dielectric properties (Section 3.3). Moderate PV-glass additions (5-15 wt%) yield high breakdown strength (>26 kV/mm), low dielectric loss (tan δ ≈ (4-7) × 10⁻³), and stable dielectric constants (ε′ ≈ 9.8-10.5). These characteristics are indicative of well-densified microstructures with reduced pore volume, uniform grain boundaries, and restricted ionic mobility. As porosity decreases and intergranular pathways become more uniform, polarization mechanisms shift toward more stable intrinsic contributions dominated by the corundum lattice. In contrast, the G4 sample deviates from this ideal once more. This is because a higher glass content lowers breakdown strength, decreases resistivity, and increases structural disorder, all of which are consistent with more continuous silicate networks that promote ionic conduction. With stronger hydroxyl-related features and improved Si-O-Al linkages in G3 and particularly G4, the FTIR analysis confirms this interpretation and suggests the presence of amorphous or partially polymerized silicate units. The correlation between these chemical features and dielectric behavior further suggests that optimal electrical reliability is achieved only when the glass content remains below the percolation threshold that would otherwise enable continuous charge transport along grain boundaries.

These correlations are finally confirmed by microstructural evidence from SEM-EDX and XRD (Section 3.4). The densely packed, equiaxed grains observed in G3, together with the uniform distribution of Si and Na, indicate that PV glass forms ultrathin intergranular films without generating detrimental secondary phases. The absence of intergranular crystallization or coarse glass pools confirms that the liquid-phase sintering is properly regulated at 15 weight percent. G4 should exhibit excessive amorphous segregation (though not examined here), whereas G1 shows cleaner but less bonded grain boundaries consistent with limited viscous flow. The FTIR evidence of new chemical environments is supported by the lattice parameter expansion in G3 (a→4.786 Å, c→13.055 Å), decreased APF, and increased bond length (L = 1.863 Å), which all together indicate lattice relaxation brought on by Si incorporation. The mechanical and electrical results are unified by these minor lattice changes, which also improve crack deflection and dielectric polarization stability. The results demonstrate that optimal material performance at 1400°C is achieved through a carefully balanced combination of controlled glass-mediated microstructural evolution, efficient densification, and the resulting improvements in mechanical robustness and dielectric stability. The composition range of 10-15 wt% PVWG is therefore identified as the most effective engineering window for achieving high-performance alumina ceramics under reduced-temperature sintering conditions.