Effect of glazing technique and firing on surface roughness and flexural strength of an advanced lithium disilicate

In recent decades, lithium disilicate glass ceramics (LD) have become essential materials in the field of esthetic and prosthetic dentistry [1‐5]. This ceramic provides a balance of strength and esthetics, which allows it to be used in wide indications, including veneers, inlays/onlays, crowns, and bridges for anterior and posterior teeth [6]. Due to its reasonable translucency, porcelain veneering is no longer necessary for LD restorations. Such monolithic restorations reduce the number of manufacturing procedures and eliminate the weak interface between substrate and veneering ceramic, therefore ensuring better mechanical properties and longevity [7, 8].

LD ceramic restorations can be produced by either conventional heat-pressing or computer-aided design and computer-aided manufacturing (CAD/CAM) [9, 10]. The CAD/CAM facility, also known as the milling procedure, includes hard and soft machining; owing to the desirable properties of LD [11], both machining methods are available, making it an ideal chairside restorative material. Hard machining involves milling objects from a fully crystallized ceramic block, which is time-consuming due to the hardness and wear of the milling bur [12]. Soft machining makes it easier to mill an object from a partially crystallized stage, but subsequent thermal processing is needed to achieve full crystallization [13], which is also time-consuming. An advanced lithium disilicate (ALD) was therefore developed to shorten the firing time and the whole fabrication period. “Advanced” nomenclature is not a classification of the material, but named according to the manufacturer due to its different composition from conventional LD, since ALD contains the natural mineral virgilite, which is claimed to have a synergistic effect with the lithium disilicate to improve the mechanical and optical properties of ceramics. Though it requires a slightly longer milling time, only a fast-firing cycle with glaze (4 min 30s) is needed, and such glaze firing is compulsory for matrix and glaze firing [14].

Conventionally, thermal treatment procedures for ceramic restorations are separate steps that consist of crystallization and glaze firing. However, this subsequent glazing has been reported to weaken the biaxial flexural strength of the conventional LD [15]. Recently, one-step firing combining crystallization and glazing has become an alternative to reduce the number of firings. Although it still reduces the biaxial flexural strength of as-crystallized ceramic, it still presents better initial biaxial flexural strength and fatigue strength compared to two-step firing [16]. For ALD, the literature regarding the benefits of glazing in one- or two-step procedure on the mechanical properties of ALD is still unknown.

One of the claims from the manufacturer is that the glaze firing together with the crystallization is necessary to achieve its final strength and to close some defects during the manufacturing procedure. However, the mechanism of strength improvement and repair is not clear. The glazing process consists of glaze material application and firing, while thermal treatment was reported to lead to crack healing [17, 18], and improvement of flexural strength [18]. Thus, it raises the question of whether the mechanical properties of ALD can be promoted by refiring, glaze material, or both. Therefore, this study was aimed at investigating the effects of the number of firings and glaze on the surface roughness and uniaxial flexural strength of ALD in comparison with LD. The null hypotheses consisted that (1) glazing and refiring would not influence the surface roughness of ALD and LD; (2) glazing together with crystallization (one-step) or separated glaze firing (two-step) would not promote different flexural strength values for both ALD and LD; and (3) refiring or the use of glaze material would not affect the flexural strength of ALD and LD.

This study was aimed at determining the influence of the number of firings and the presence of glaze layer on the surface roughness and flexural strength of two lithium disilicate-based ceramics. According to the results, the first hypothesis that glazing and refiring would not influence the surface roughness of ALD and LD was partially rejected since the glaze presence had a negative impact on the mean surface roughness. The purpose of glazing during restoration manufacturing is to improve surface quality and generate a glossy outer layer. However, the glazed surface has a greater tendency for bacterial attachment and oral biofilm formation [21] if the required threshold of 0.2 μm of surface roughness is not respected [22]. However, it is worth mentioning that other interactions can affect the microorganism adhesion, such as surface free energy and communication between existing microorganisms [21]. In this study, both ALD cg and c-g had values close to this standard, while the glazed LD groups showed rougher surfaces. This could be attributed to the surface properties of LD as well as handling. For both materials, the average roughness after glazing was higher compared to c groups. This result is consistent with a previous study [23] in which glazed lithium disilicate presented a rougher surface than polished. In addition, the roughness of one- and two-step glazed ALD was similar, revealing that the glazing protocols do not influence the surface quality of ALD. This was also confirmed by SEM images since ALDcg and ALDc-g showed similar surface morphology. For LD, refiring did not influence the average surface roughness (Ra), but it seems that LDcg and LDc-g have different surface morphology according to SEM images. This could be because two different glaze materials used for LD in different glazing protocols have different properties.

The second hypothesis that glazing together with crystallization (one-step) or separated glaze firing (two-step) would not promote different strength values of both ALD and LD, was rejected. For LD, cg was similar to c, while c-g led to a decrease in the FS. This is in accordance with a previous report, where one-step glazing showed higher fatigue strength of LD than two-step [16]. However, for ALD, the result was the complete opposite. Both one- and two-step glazing improved its flexural strength compared to ALDc, whereas c-g was twice higher than ALDc. It is interesting that ALD material presented an initial strength that was only half of LD but showed the highest strength similar to LD after two-step glazing. So it is of great importance to understand how such glazing influences ALD mechanical properties in order to improve the mechanical performance of restorations. In addition, the influence of glazing on the flexural strength of these lithium disilicate ceramics cannot be explained by the change in surface roughness since the rougher surface could lead to lower strength [24].

The presence of a second fire cycle increased the strength of ALD. Additionally, if the second firing is for a glaze layer application, it improved even more of the material’s strength. However, only a second firing did not increase the strength of LDc. And the application of glaze in the second firing even decreased its strength. Therefore, the third hypothesis that refiring or the use of glaze material would not affect the strength of ALD and LD was partially rejected. This behavior can be justified because; during the second firing, the glassy phase on the glass ceramics could be partially molten at temperatures above the glass transition temperature (Tg). For the lithium disilicate reinforced glass ceramics, this temperature is around 450°C [25, 26]. With the temperature rising, the glass phase becomes less viscous to fill in and seal surface defects such as machining damages [27]. This also leads to a higher strength of ALDc-g compared to ALDcg since the first has an additional firing cycle. To verify this assumption, a crack healing analysis of ALD and LD was conducted in this study. As shown in Fig. 3, the introduced crack in ALD totally disappeared after the second firing. Therefore, surface damages caused by specimen preparation could be slightly healed in the first firing and further the subsequent firing, decreasing the critical defect size and therefore improving the ceramic strength. However, the crack in LD was only partially closed, and this change in LD failed to affect the average roughness and strength in this study, which can be supported by other studies that show a normal glazing cycle cannot improve the strength of LD [18]. It is possible to hypothesize that the glazing temperature of the two-step glazing might be insufficient for LD to influence the glassy phase of it as for ALD. An extended glaze firing with a longer holding time up to 15 minutes was suggested for LD to improve its mechanical properties [17, 18, 28]. In contrast, the strength improvement of ALD after an additional firing suggested that the initial firing by the manufacturer is insufficient for the material. Thus, longer firing time could be able to increase the strength of ALD as more firing cycles.

The application of glaze at different processing stages showed different effects on the flexural strength of ALD and LD. Comparing the FS of glazed and unglazed LD with one or two firings, the glaze material applied in the first firing did not generate an influence, while the glaze in the second firing decreased the FS. The first firing of LD could be sufficient for surface damage healing reaching its highest strength. This can be supported by the above findings that no polishing scratch was found in LDc, and all the critical defects are volume defects inside the ceramic. Different from LDcg, a number of large pores were detected under SEM at the interface between LD and the glaze layer, which can be the result of impurities induced before the second firing. These large pores act as the stress concentration while loading, weakening the mechanical properties of glazed restoration. Moreover, crystallized LD presented less glass percentage [4] and porosity than uncrystallized LD, and the holding temperature of the second firing was lower than the first firing. These conditions might influence the fusion of the glaze layer to LD substrate, therefore weakening the interface. For ALD, the application of glaze material at both the first and second firing leads to higher strength. It could be attributed that silicate glass, the major component of the glaze, potentially fills surface cracks during firing, collaborating with the crack-healing ability of ALD and compensating for the lack of firing. More evidence is needed to explain the effect of glaze on ALD strength.

In the study of Lubauer et al. [4], crystallized ALD presented a fracture toughness of 1.45 MPa·m^1/2, which was relatively lower than crystallized LD (2.13 MPa · m^1/2). This could partially explain that crystallized ALD without a glaze layer had a lower flexural strength than crystallized LD in the present investigation. However, despite sharing clinical indications [20], both ceramics present different performances, such as ALD demonstrating a better wear behavior than LD [29], and comparable bonding strength before and after aging [30]. Other investigations reported higher mean strength value (374.22 MPa) than this study (282.1 MPa) for ALDcg [31], which could be possibly explained due the difference between biaxial flexural strength and 3-point bending tests. In addition, the manufacturer claims that ALD flexural strength can achieve higher values than 700 MPa. In this study, even the strongest group (ALDc-g) presented 442.3 MPa, which is 63% of the value suggested by the manufacturer, but still superior to the threshold of 300 MPa defined in ISO 6872:2015.

As limitations of this study, in addition to the crack healing effect, it is unknown if the residual stress after glazing and refiring played a role in the different strength results. Also, the effect of successive thermal cycles can possibly lead to a change in crystal contents [16]. Repeated firings cause ceramic materials to be exposed to additional heat treatments. For nonreinforced glass-ceramics, the multiple firing did not affect the flexural strength, hardness, and microstructure in the long-term [32]. While for LD, the repeat firing processes did not affect its flexural strength, the surface hardness and fracture toughness were significantly changed [33]. However, there is a lack of information about that for ALD and low-temperature fused glass ceramics such as glaze. Thus, further studies are advocated to explore the crystal content after different glazing and refiring on the flexural strength of ALD and LD. Moreover, only two firing cycles were evaluated, so it is unknown if ALD can achieve higher strength with more firing cycles or longer firing time. The influence of the number and holding time of firing on the optical properties of ALD is still unknown. Further studies are suggested with a focus on the effect of firing cycles and longer holding time on the mechanical properties as well as the optical properties of ALD. In addition, laboratory studies evaluating fatigue and wear resistance would also improve the literature regarding this new lithium disilicate.

Besides both materials being lithium-disilicate glass ceramics, the glazing technique and firing protocol affected their roughness and flexural strength differently. A two-step crystallization and glazing should be the first choice for ALD, while for LD, glazing is optional and, when necessary, should be applied in one step. Although glazing can generate a homogenous and glossy surface morphology, it increases the surface roughness for both materials.