Phase Formation, Mechanical Strength, and Bioactive Properties of Lithium Disilicate Glass–Ceramics with Different Al2O3 Contents

Owing to its excellent mechanical properties and aesthetic tooth-like appearance, lithium disilicate glass–ceramic is more attractive as a crown for dental restorations. In this study, lithium disilicate glass–ceramics were prepared from SiO2–Li2O–K2O–P2O5–CeO2 glass systems with various Al2O3 contents. The mixed glass was then heat-treated at 600 °C and 800 °C for 2 h to form glass–ceramic samples. Phase formation, microstructure, mechanical properties and bioactivity were investigated. The phase formation analysis confirmed the presence of Li2Si2O5 in all the samples. The glass–ceramic sample with an Al2O3 content of 1 wt% showed rod-like Li2Si2O5 crystals that could contribute to the delay in crack propagation and demonstrated the highest mechanical properties. Surface treatment with hydrofluoric acid followed by a silane-coupling agent provided the highest micro-shear bond strength for all ceramic conditions, with no significant difference between ceramic samples. The biocompatibility tests of the material showed that Al2O3-added lithium disilicate glass–ceramic sample was bioactive, thus activating protein production and stimulating the alkaline phosphatase (ALP) activity of osteoblast-like cells.


Introduction
Glass-ceramics are made of crystalline phase and glass. They are specially formed composite that are created through high-temperature melting, shaping, and heat treatment. Due to their excellent mechanical strength, adjustable thermal expansion, chemical resistance, low dielectric loss, and other properties, glass-ceramics are currently gaining considerable attention in mechanical manufacturing, optics, electronics, aerospace, biomedical, and construction applications [1][2][3][4]. In the dental restoration field, glass-ceramics are widely employed due to their biocompatibility, ease of production, and chemical, physical, and mechanical qualities that are extremely similar to those of human enamel [5]. Various materials such as alumina, zirconia, leucite and lithium disilicate glass-ceramics have been developed [6]. Among these materials, lithium disilicate (Li 2 Si 2 O 5 ) glass-ceramics, which are made of a crystalline ceramic and glass matrix, have drawn the most attention among these materials because of their excellent mechanical characteristics and distinctive translucency [7,8]. There has also been a significant increase in the number of reports on the composition, structure, phase transformations and properties of Li 2 O-SiO 2 -based glass

X-ray Powder Diffraction Analysis (XRD)
Cu-Kα radiation produced at 30 mA and 45 kV was used in the X-ray powder diffraction analysis using an X-ray diffractometer (XRD) (SmartLab, Rigaku, Japan) with a scanning range of 10 • to 60 • and a step size of 0.012 • . Using JCPDS numbers from the ICDD-PDF2 database, the crystalline phases in the controlled heat-treated glass specimens were identified.

Field Emission Scanning Electron Microscopy (FE-SEM)
The microstructure of the glass-ceramic specimens was examined using field emission scanning electron microscopy (FE-SEM) (JSM-6335F, JEOL, Akishima, Tokyo, Japan)An ion sputtering device (JFC-1200, JEOL, Akishima, Tokyo, Japan) was used to coat the fracture surface of the specimens with gold for 10 s.

Mechanical test
Using Vickers hardness tests (Buehler, Karl Frank GMBH Type-38505, Lake Bluff, IL, USA) with a constant load of 1 kg and a dwell time of 15 s, microhardness was used to analyse the mechanical properties of the prepared glass-ceramics [17]. The equation below was used to determine the Vickers hardness (HV): where 1.8544, P, and d denote the constant geometrical factor for the diamond pyramid, the applied load with the unit of kg, and the diagonal length of the indenter impression in µm, respectively. There are universal strength machines that can be used for ISO 6872 [18] with a universal strength machine (AG-X plus, Shimadzu, Japan). The flexural modulus (E) was calculated using the following equation [19]: where L is the support span (mm), m is the gradient (i.e., slope) of the initial straight-line portion of the load deflection curve, (N/mm), and b and d are the width of the test beam and thickness, respectively. A three-point-bending test with the fracture strength (σ) in the unit of MPa and a crosshead speed of 0.01 mm/min, and the test span was 16 mm. The following equation can be used to determine the three-point flexure: In this case, P stands for the breaking load in N, while w, b, and l, respectively, are the width, the thickness of the samples, and the test span in mm.

Micro-Shear Bond Strength (µSBS)
All the lithium disilicate glass-ceramic samples with various Al 2 O 3 contents (LSA1, LSA2 and LSA3) were selected and subjected to the µSBS test for the evaluation of their bonding performance. Briefly, 20 ceramic plates (5 × 5 × 2 mm) of each ceramic condition were prepared and embedded in a metal ring using epoxy resin. After polishing with 600, 800, 1000, and 1200 grit silicon carbide paper, the specimens were randomly divided into 4 group according to 4 surface treatments: (1) no surface treatment; (2) silane coating (Monobond N, Ivoclar Vivadent, Liechtenstein); (3) hydrofluoric acid (IPS ® ceramic etching gel, Ivoclar Vivadent, Liechtenstein); and (4) hydrofluoric acid and silane coating. The treated ceramic surface was then cemented with 4 resin cement rods (Multilink ® N, Ivoclar Vivadent, Liechtenstein) by injecting resin cement into 4 plastic tubes (0.8 mm internal diameter and 0.5 mm height). After removing the tubes and storing in water for 24 h, the with the supplementation of a 10% fetal bovine serum (FBS), 2% N-(2-hydroxyethyl) piperazine-N'-(2-ethane sulfonic acid) (HEPES), 1% non-essential amino acid, and 1% penicillin/streptomycin until they reached 80% confluence. Then 1 × 10 4 cells were loaded in the well containing the treated material with a volume of 100 µL and the plate was incubated in a humidified incubator under 37 • C and 5% CO 2 for 24 h. The fresh medium was then replaced and the plate continued to be incubated for 48 and 96 h.

Cell Attachment, Cell Viability, and Cell Proliferation by MTT Assay
Briefly, a 5 mg/mL methyl thiazolyl diphenyl-tetrazolium bromide (MTT) solution was prepared in a sterile phosphate-buffered saline (PBS) solution, pH 7.4 and added into each well at 10 µL. After that, the plate continued to be incubated for 4 h to allow the formation of the formazan crystal by viable cells. The media supernatant was then replaced with 100 µL dimethyl sulfoxide (DMSO). The absorbance at 540 nm was then recorded with a plate reader machine.

Preparation of the Cell Lysate
The celll layer was primarily washed twice with 37 • C sterile PBS prior to the addition of 50 µL cell lysis buffer. The cells were exposed to the lysis buffer at room temperature for 15 min to completely lyse. The lysate was collected and stored in an Eppendorf pipette under −20 • C for further analyses.

Total Protein Determination by Bradford Assay
The total protein content was analysed in 5 µL of the cell lysate by mixing with 100 µL of the working Bradford solution. The complete reaction was allowed to incubate for 15 min at room temperature. The absorbance was recorded with a plate reader at 595 nm. Like the sample, standard bovine serum albumin (BSA) was used to establish the standard curve for the deduction of total protein content in each sample.

Alkaline Phosphatase (ALP) Activity Determination
Briefly, 20 µL of the lysate was mixed with 80 µL of 1 mg/mL p-nitrophenolphosphate in a diethanolamine buffer pH 9.8 for 3 h and the yellow product was determined by with a plate reader to measure the absorbance at 405 nm. Along with the p-nitophenol (pNP) standard, ALP activity could be elucidated and expressed in the specific activity per mg of the total protein.

Statistical Analysis
ANOVA followed by Tukey's pos-hoc test was used when the ANOVA result showed a significant difference. The p-value was set at 0.05 in all the statistical procedures.  Figure 1 illustrates the effect of the Al 2 O 3 concentration on the thermal parameters of the glass specimens by performing DTA at a heating rate of 10 • C/min while maintaining an air atmosphere throughout the experiment. The letters T g , T C1 , T C2 , and T m indicate the temperatures at which the glass transition, first crystallisation, second crystallisation, and melting temperature occur in the DTA curves. Exothermic peaks indicate that the glass is crystallised, whereas endothermic peaks indicate that crystalline phases dissolve within the glass matrix. In the ranges of 663-668 • C and 825-845 • C, the results for each glass sample showed two exothermic peaks. The Al 2 O 3 content was also found to have a negligible effect on the second crystallisation and melting temperature, which corresponds to a previous study [16]. It is possible that this is because Al 2 O 3 typically makes the glass more viscous, which is caused by the elimination of nonbridging oxygen sites [14]. In addition, T C1 is the temperature at which the lithium metasilicate (Li 2 SiO 3 ) phase begins to crystallise. The second crystallisation temperature, T C2 , denotes the phase transition at which the crystallised LM phase changes into the lithium disilicate (Li 2 Si 2 O 5 ) phase [20][21][22]. According to the DTA results, the heat treatment condition was selected as two stages at 600 • C and 800 • C for 2 h at each temperature. It was speculated that the preliminary glass nucleation would occur at the initial stage at 600 • C and then the crystalline phase of Li 2 Si 2 O 5 would start to occur at the final stage at 800 • C.

Statistical Analysis
ANOVA followed by Tukey's pos-hoc test was used when the ANOVA result showed a significant difference. The p-value was set at 0.05 in all the statistical procedures. Figure 1 illustrates the effect of the Al2O3 concentration on the thermal parameters of the glass specimens by performing DTA at a heating rate of 10 °C/min while maintaining an air atmosphere throughout the experiment. The letters Tg, TC1, TC2, and Tm indicate the temperatures at which the glass transition, first crystallisation, second crystallisation, and melting temperature occur in the DTA curves. Exothermic peaks indicate that the glass is crystallised, whereas endothermic peaks indicate that crystalline phases dissolve within the glass matrix. In the ranges of 663-668 °C and 825-845 °C, the results for each glass sample showed two exothermic peaks. The Al2O3 content was also found to have a negligible effect on the second crystallisation and melting temperature, which corresponds to a previous study [16]. It is possible that this is because Al2O3 typically makes the glass more viscous, which is caused by the elimination of nonbridging oxygen sites [14]. In addition, TC1 is the temperature at which the lithium metasilicate (Li2SiO3) phase begins to crystallise. The second crystallisation temperature, TC2, denotes the phase transition at which the crystallised LM phase changes into the lithium disilicate (Li2Si2O5) phase [20][21][22]. According to the DTA results, the heat treatment condition was selected as two stages at 600 °C and 800 °C for 2 h at each temperature. It was speculated that the preliminary glass nucleation would occur at the initial stage at 600 °C and then the crystalline phase of Li2Si2O5 would start to occur at the final stage at 800 °C.

Phase Formation
The phase formation of the glass-ceramic samples was investigated by forming XRD analysis, which showed the precipitating crystalline phase as a function of the various Al2O3 contents. The XRD results of all the LSA glass-ceramic samples after heat treatment at 600 °C and 800 °C for 2 h are shown in Figure 2. The XRD results revealed that Li2Si2O5 was the main crystalline phase in all the samples. The highest intensity of the Li2Si2O5 peak was obtained when an Al2O3 content of 1%mol was added, and it coexisted with a small

Phase Formation
The phase formation of the glass-ceramic samples was investigated by forming XRD analysis, which showed the precipitating crystalline phase as a function of the various Al 2 O 3 contents. The XRD results of all the LSA glass-ceramic samples after heat treatment at 600 • C and 800 • C for 2 h are shown in Figure 2. The XRD results revealed that Li 2 Si 2 O 5 was the main crystalline phase in all the samples. The highest intensity of the Li 2 Si 2 O 5 peak was obtained when an Al 2 O 3 content of 1%mol was added, and it coexisted with a small amount of cristobalite phase. This finding is consistent with the DTA results at 800 • C; this is close to the crystalline temperature of all the glass samples, especially LSA1, which showed a T c at 825 • C, while LSA2 and LSA3 were at 830 • C and 845 • C, respectively. Therefore, the crystallisation reaction of Li 2 Si 2 O 5 was more complete, and the highest intensity peak was observed. The crystalline phase of Li 2 Si 2 O 5 was transformed in two steps via the following reaction: amount of cristobalite phase. This finding is consistent with the DTA results at 800 °C; this is close to the crystalline temperature of all the glass samples, especially LSA1, which showed a Tc at 825 °C, while LSA2 and LSA3 were at 830 °C and 845 °C, respectively. Therefore, the crystallisation reaction of Li2Si2O5 was more complete, and the highest intensity peak was observed. The crystalline phase of Li2Si2O5 was transformed in two steps via the following reaction: Generally, in this Li2O-SiO2 glass system, the first crystalline phase of Li2SiO3 is formed by a reaction between Li2O and SiO2 (Equation (4)) at a lower temperature of 550-650 °C [22][23][24]. However, when the heat treatment temperature increased, Li2SiO3 became unstable and transformed into the more stable phase of Li2Si2O5, as shown in Equation (5). The amounts of the Li2Si2O5 phase decreased with the increasing Al2O3 content, while the cristobalite phase disappeared and CeO2 started to form in the LSA2 samples. Moreover, the CeO2 phase remained almost unchanged with a further increase in the Al2O3 content in LSA3. This may also explain why the cristobalite phase is metastable and cannot be retained at higher Al2O3 contents. However, the reaction that transformed to Li2Si2O5 was incomplete and showed impurities in the CeO2 phase in the LSA2 and LSA3 samples. This may be attribute to the addition of Al2O3 leading to a decrease in the crystal growth velocity [25], so when the Al2O3 content exceeded the solution limit (LSA2 and LSA3), a foreign CeO2 phase was residual in the glass-ceramic samples, which led to the impurity of the materials. Figure 3 shows the microstructure of the surface glass-ceramic samples. The results illustrated that the surfaces of the LSA1 and LSA2 samples showed the formation of a rodlike Li2Si2O5 phase that was randomly dispersed. However, the morphology of the crystalline phase of the LSA3 sample was different. With the addition of Al2O3 to 1.5 mol%, the grain size that was changed to higher increased, and the number of rod-like shapes gradually decreased. Moreover, the shape of the crystals was slightly transformed to the spherical shape of Li2SiO3, presenting a polymorphic microstructure [26]. Generally, in this Li 2 O-SiO 2 glass system, the first crystalline phase of Li 2 SiO 3 is formed by a reaction between Li 2 O and SiO 2 (Equation (4)) at a lower temperature of 550-650 • C [22][23][24]. However, when the heat treatment temperature increased, Li 2 SiO 3 became unstable and transformed into the more stable phase of Li 2 Si 2 O 5 , as shown in Equation (5).

SEM Analysis
The amounts of the Li 2 Si 2 O 5 phase decreased with the increasing Al 2 O 3 content, while the cristobalite phase disappeared and CeO 2 started to form in the LSA2 samples. Moreover, the CeO 2 phase remained almost unchanged with a further increase in the Al 2 O 3 content in LSA3. This may also explain why the cristobalite phase is metastable and cannot be retained at higher Al 2 O 3 contents. However, the reaction that transformed to Li 2 Si 2 O 5 was incomplete and showed impurities in the CeO 2 phase in the LSA2 and LSA3 samples. This may be attribute to the addition of Al 2 O 3 leading to a decrease in the crystal growth velocity [25], so when the Al 2 O 3 content exceeded the solution limit (LSA2 and LSA3), a foreign CeO 2 phase was residual in the glass-ceramic samples, which led to the impurity of the materials. Figure 3 shows the microstructure of the surface glass-ceramic samples. The results illustrated that the surfaces of the LSA1 and LSA2 samples showed the formation of a rod-like Li 2 Si 2 O 5 phase that was randomly dispersed. However, the morphology of the crystalline phase of the LSA3 sample was different. With the addition of Al 2 O 3 to 1.5 mol%, the grain size that was changed to higher increased, and the number of rod-like shapes gradually decreased. Moreover, the shape of the crystals was slightly transformed to the spherical shape of Li 2 SiO 3 , presenting a polymorphic microstructure [26]. These results were confirmed by the fracture surface of the glass-ceramic samples in Figure 3, which shows that in the LSA1 samples, the crystalline shape is rod-like and distributed on the glass matrix of the glass-ceramics. Generally, cracks are always formed at the weaker interfaces of the crystals and propagate through the residual glass matrix. However, in LSA1, the crystal morphology is rod-like. It has a closely packed and multidirectional interlocking microstructure of Li2Si2O5 that deflects and delays crack propagation [27]. Microcracks were formed inside the matrix of glass-ceramics with curves and intersecting paths. This was related to the interlocking structure of Li2Si2O5, which prevented crack propagation in the samples. On the other hand, with increasing Al2O3 content, the density of Li2Si2O5 crystals decreased and became spherical coexisting with the rod shape in LSA3. Notably, increasing the Al2O3 content depressed the reaction of Li2Si2O5 but consumed the Li2SiO3 crystal. These results were confirmed by the fracture surface of the glass-ceramic samples in Figure 3, which shows that in the LSA1 samples, the crystalline shape is rod-like and distributed on the glass matrix of the glass-ceramics. Generally, cracks are always formed at the weaker interfaces of the crystals and propagate through the residual glass matrix. However, in LSA1, the crystal morphology is rod-like. It has a closely packed and multi-directional interlocking microstructure of Li 2 Si 2 O 5 that deflects and delays crack propagation [27]. Microcracks were formed inside the matrix of glass-ceramics with curves and intersecting paths. This was related to the interlocking structure of Li 2 Si 2 O 5 , which prevented crack propagation in the samples. On the other hand, with increasing Al 2 O 3 content, the density of Li 2 Si 2 O 5 crystals decreased and became spherical coexisting with the rod shape in LSA3. Notably, increasing the Al 2 O 3 content depressed the reaction of Li 2 Si 2 O 5 but consumed the Li 2 SiO 3 crystal.

Mechanical Properties
The improved mechanical properties of glass-ceramics are also a consequence of their higher crystallinity, appropriate morphology of precipitated crystals and low porosity [16,26]. The flexural strength, the flexural modulus of elasticity, and microhardness values of the glass-ceramic samples with different Al 2 O 3 are shown in Figures 4 and 5. Similar trends can be observed in all the mechanical properties measured in this study with respect to the Al 2 O 3 contents. The highest mechanical strength values were obtained from the glass-ceramics sample upon adding Al 2 O 3 to 1.0 mol% and slightly decreased as the Al 2 O 3 content was increased to 1.5 mol%. A possible explanation is that LSA1 samples contained the highest crystalline phase of rod-like Li 2 Si 2 O 5 , which formed an interlocking microstructure to prevent cracks within the samples. In addition, the increasing Al 2 O 3 had the effect of decreasing the densification of the crystal, which is related to the amount of glassy phase and its viscosity [28]. In the samples with higher Al 2 O 3 content, the amount of glassy phase was reduced, and the viscosity of the glassy phase increased, so the addition of a crystalline phase could inhibit densification [28]. This is consistent with the work of Fernandes et al. [29] reporting that the glass sample with the lowest Al 2 O 3 increased the densification of the crystal which resulted in an improvement in the mechanical properties of lithium disilicate glass-ceramics.
The mechanical strength values of the specimens in the present study were close to those of the commercial lithium disilicate developed by Ivoclar Vivadent company and other studies, as shown in Table 1  The improved mechanical properties of glass-ceramics are also a consequence of their higher crystallinity, appropriate morphology of precipitated crystals and low porosity [16,26]. The flexural strength, the flexural modulus of elasticity, and microhardness values of the glass-ceramic samples with different Al2O3 are shown in Figures 4 and 5. Similar trends can be observed in all the mechanical properties measured in this study with respect to the Al2O3 contents. The highest mechanical strength values were obtained from the glass-ceramics sample upon adding Al2O3 to 1.0 mol% and slightly decreased as the Al2O3 content was increased to 1.5 mol%. A possible explanation is that LSA1 samples contained the highest crystalline phase of rod-like Li2Si2O5, which formed an interlocking microstructure to prevent cracks within the samples. In addition, the increasing Al2O3 had the effect of decreasing the densification of the crystal, which is related to the amount of glassy phase and its viscosity [28]. In the samples with higher Al2O3 content, the amount of glassy phase was reduced, and the viscosity of the glassy phase increased, so the addition of a crystalline phase could inhibit densification [28]. This is consistent with the work of Fernandes et al. [29] reporting that the glass sample with the lowest Al2O3 increased the densification of the crystal which resulted in an improvement in the mechanical properties of lithium disilicate glass-ceramics.    The mechanical strength values of the specimens in the present study were close to those of the commercial lithium disilicate developed by Ivoclar Vivadent company and other studies, as shown in Table 1. Commercial lithium disilicate has a flexural strength and Vickers hardness values of 4400 and 5800 MPa, respectively. These excellent mechanical properties can be attributed to the development of Li2Si2O5 in the sample.

Microshear Bond Strength
The cementation of resin cement for ceramic restorations is a crucial process for clinical success [A1]. Various surface treatments have been used before cementation to improve the bond strength of dental ceramics [34,35]. The micro-shear bond strength test was selected in this study to evaluate the bonding performance owing to its benefits of more uniform stress distribution and lack of damage to specimens that are suitable for the glass-ceramic material [36,37].
In addition to the different mechanical properties, there was no significant difference in µSBS among the three group of samples (LSA1, LSA2 and LSA3) of lithium disilicate glass-ceramics. This might be due to the lower effect of the mechanical properties on the bonding ability between the resin cement and the selected ceramics, as shown in Table 2. However, mechanical properties are crucial and must be considered because ceramics with better mechanical properties provide a longer lifetime for all ceramic restorations [38,39].

Microshear Bond Strength
The cementation of resin cement for ceramic restorations is a crucial process for clinical success [A1]. Various surface treatments have been used before cementation to improve the bond strength of dental ceramics [34,35]. The micro-shear bond strength test was selected in this study to evaluate the bonding performance owing to its benefits of more uniform stress distribution and lack of damage to specimens that are suitable for the glass-ceramic material [36,37].
In addition to the different mechanical properties, there was no significant difference in µSBS among the three group of samples (LSA1, LSA2 and LSA3) of lithium disilicate glass-ceramics. This might be due to the lower effect of the mechanical properties on the bonding ability between the resin cement and the selected ceramics, as shown in Table 2. However, mechanical properties are crucial and must be considered because ceramics with better mechanical properties provide a longer lifetime for all ceramic restorations [38,39].
The mean µSBS of all the groups with different surface treatments was significantly higher than that of the control group. This result could be attributed to the effectiveness of the chemical bond created by the silane coupling agent and the micromechanical retention created through hydrofluoric acid etching.
Silane promotes the adhesion between the functional methacrylate group of resin cement and the hydroxyl group of the silica-based ceramic. Additionally, a hydrophobic surface can be created by the application of a silane-coupling agent, which allows resin cement to flow [40][41][42]. Hydrofluoric acid can dissolve the glass matrix and increase the surface area for micromechanical retention between resin cement and the etched glass surface [43][44][45], resulting in increased bond strength. However, the lithium disilicate glass-ceramics used in this study contained Al 2 O 3 , which can neutralise the acidity of hydrofluoric acid and impair the etching ability. Moreover, the penetration of resin cement could be interfered with by the insoluble salt formed owing to the interaction between aluminium ions and hydrofluoric acid [46]. This may be the reason for the low µSBS in the group treated with hydrofluoric acid.
The highest µSBS value was obtained by etching with hydrofluoric acid, followed by the application of a silane coupling agent. This method has been recommended as the most effective surface treatment for glass-ceramics [45,[47][48][49]. This result might be due to the synergistic combination of mechanical and chemical retention, as the increased surface area created by etching provides more surface area for chemical bonds.

Cellular Activity
Although the results showed lower cell numbers attached to LSA3, there was no significant difference among the others (Figure 6). The attachment of the cells onto the surface of the glass-ceramic samples were not affected by the observation period (6-24 h), indicating that the glass-ceramics allowed MG-63 cell attachment, regardless of the concentration of Al 2 O 3 . Lithium disilicate glass-ceramics with various Al2O3 were biocompatible with MG-63 cells. Although the MG-63 viability cultured on LSA1 significantly decreased compared with that on LSA2 and LSA3 after 4 days of incubation, this difference was dissipated after 7 days of incubation (Figure 7). The viability of the cells significantly increased over time compared with that at the early time point. This result confirmed the biocompatibility of all types of glasses owing to the ability of MG-63 cell proliferation on all types of glasses. Lithium disilicate glass-ceramics with various Al 2 O 3 were biocompatible with MG-63 cells. Although the MG-63 viability cultured on LSA1 significantly decreased compared with that on LSA2 and LSA3 after 4 days of incubation, this difference was dissipated after 7 days of incubation (Figure 7). The viability of the cells significantly increased over time compared with that at the early time point. This result confirmed the biocompatibility of all types of glasses owing to the ability of MG-63 cell proliferation on all types of glasses.
Lithium disilicate glass-ceramic material with various concentrations of Al 2 O 3 -activated protein was produced by the MG-63 cells. The difference in the total protein content was significantly higher in the MG-63 cells cultured with LSA1 after 4 days of culture; however, these contents were similar in all material types after 7 days of culture ( Figure 8). When compared between time points, the protein content was significantly increased in all the investigated materials. Taken together, the proliferation results confirmed that the glass specimens, regardless of the presence of aluminium, were biocompatible with MG-63 cells.
Lithium disilicate glass-ceramics with various Al2O3 were biocompatible with MG-63 cells. Although the MG-63 viability cultured on LSA1 significantly decreased compared with that on LSA2 and LSA3 after 4 days of incubation, this difference was dissipated after 7 days of incubation (Figure 7). The viability of the cells significantly increased over time compared with that at the early time point. This result confirmed the biocompatibility of all types of glasses owing to the ability of MG-63 cell proliferation on all types of glasses. . The results were from a triplicate (n = 3) experiment and the statistical difference was tested when p < 0.05, which is represented by * (among the material types in each investigation period) and by # (between two investigation periods among the same material types).
Lithium disilicate glass-ceramic material with various concentrations of Al2O3-activated protein was produced by the MG-63 cells. The difference in the total protein content was significantly higher in the MG-63 cells cultured with LSA1 after 4 days of culture; however, these contents were similar in all material types after 7 days of culture ( Figure   8). When compared between time points, the protein content was significantly increased . The results were from a triplicate (n = 3) experiment and the statistical difference was tested when p < 0.05, which is represented by * (among the material types in each investigation period) and by # (between two investigation periods among the same material types).  , as determined with Bradford assay. The results were from a triplicate (n = 3) experiment, and the statistical difference was tested when p < 0.05, which is represented by * (among the material types in each investigation period) and by # (between two investigation periods among the same material types).
In terms of the alkaline phosphatase (ALP) activity of MG-63, the MG-63 cells cultured with lithium disilicate glass-ceramics with various Al2O3 for 4 (D.4) and 7 days (D.7) were assessed for the ALP activity by performing p-nitrophenolphosphate conversion assay, and the results are shown in Figure 9. Regarding the expression of a bone formation marker, the ALP activity was not significantly expressed in the MG-63 cells cultured with different types of lithium disilicate glass-ceramics after 4 and 7 days of investigation. Likewise, the ALP activity expression trend was slightly reduced during prolonged investigation periods in all the tested glass-ceramics. This suggests that lithium disilicate glass-ceramics, regardless of Al2O3, are biologically active, allowing MG-63 cells to normally express the ALP activity, leading to the regulation of bone formation. , as determined with Bradford assay. The results were from a triplicate (n = 3) experiment, and the statistical difference was tested when p < 0.05, which is represented by * (among the material types in each investigation period) and by # (between two investigation periods among the same material types).
In terms of the alkaline phosphatase (ALP) activity of MG-63, the MG-63 cells cultured with lithium disilicate glass-ceramics with various Al 2 O 3 for 4 (D.4) and 7 days (D.7) were assessed for the ALP activity by performing p-nitrophenolphosphate conversion assay, and the results are shown in Figure 9. Regarding the expression of a bone formation marker, the ALP activity was not significantly expressed in the MG-63 cells cultured with different types of lithium disilicate glass-ceramics after 4 and 7 days of investigation. Likewise, the ALP activity expression trend was slightly reduced during prolonged investigation periods in all the tested glass-ceramics. This suggests that lithium disilicate glass-ceramics, regardless of Al 2 O 3 , are biologically active, allowing MG-63 cells to normally express the ALP activity, leading to the regulation of bone formation. tured with lithium disilicate glass-ceramics with various Al2O3 for 4 (D.4) and 7 days (D. 7) were assessed for the ALP activity by performing p-nitrophenolphosphate conversion assay, and the results are shown in Figure 9. Regarding the expression of a bone formation marker, the ALP activity was not significantly expressed in the MG-63 cells cultured with different types of lithium disilicate glass-ceramics after 4 and 7 days of investigation. Likewise, the ALP activity expression trend was slightly reduced during prolonged investigation periods in all the tested glass-ceramics. This suggests that lithium disilicate glass-ceramics, regardless of Al2O3, are biologically active, allowing MG-63 cells to normally express the ALP activity, leading to the regulation of bone formation. IPS e. max is a lithium disilicate glass-ceramic ingot that delivers high-strength and highly aesthetic materials for press technology. Figure 10 shows LSA3 before and after heating at 600/800 • C. This sample can be pressed into the shape of a dental crown. IPS e. max is a lithium disilicate glass-ceramic ingot that delivers high-strength and highly aesthetic materials for press technology. Figure 10 shows LSA3 before and after heating at 600/800 °C. This sample can be pressed into the shape of a dental crown.

Conclusions
The aim of this study was to investigate the effect of Al 2 O 3 at the three different contents of 0.5, 1.0 and 1.5 mol% on the mechanical properties of the modified Li 2 O-SiO 2 glass systems with CeO 2 . The Al2O3 content played a major role in phase formation changes in glass-ceramic samples. With increasing the content of Al2O3, the relative content of the Li2Si2O5 phase decreased. The data showed that the LSA1 had the highest crystallinity of Li2Si2O5 as well as the highest mechanical strength correlated with the homogeneous crystallisation of interlocked rod-like Li2Si2O5. In spite of the differences in the mechanical strength, of the samples, all the ceramic conditions showed the same micro-shear bond strength for each surface treatment. An MTT assay was used to test the cellular adhesion and biocompatibility of lithium disilicate glass-ceramic sample for all three contents of Al2O3. Even though the results were insignificant among the three samples, can be reason- Figure 10. The LSA3 sample before heating (left), the LSA3 sample heated at 600/800 • C and the completely pressed object (right).

Conclusions
The aim of this study was to investigate the effect of Al 2 O 3 at the three different contents of 0.5, 1.0 and 1.5 mol% on the mechanical properties of the modified Li 2 O-SiO 2 glass systems with CeO 2 . The Al 2 O 3 content played a major role in phase formation changes in glass-ceramic samples. With increasing the content of Al 2 O 3 , the relative content of the Li 2 Si 2 O 5 phase decreased. The data showed that the LSA1 had the highest crystallinity of Li 2 Si 2 O 5 as well as the highest mechanical strength correlated with the homogeneous crystallisation of interlocked rod-like Li 2 Si 2 O 5 . In spite of the differences in the mechanical strength, of the samples, all the ceramic conditions showed the same micro-shear bond strength for each surface treatment. An MTT assay was used to test the cellular adhesion and biocompatibility of lithium disilicate glass-ceramic sample for all three contents of Al 2 O 3 . Even though the results were insignificant among the three samples, can be reasonably claimed that these samples are suitable for dental applications.