In vitro evaluation of bond strength between dental ceramics and titanium frameworks produced by additive and subtractive methods

Abstract

Background

This study aimed to evaluate the shear bond strength (SBS) of ceramic and titanium substructures produced by different methods.

Methods

After designing 51 disc-shaped samples, three groups were created according to material type and manufacturing method Cobalt-chromium (Co-Cr) group prepared by selective laser melting (GC), titanium group prepared by milling (GTi1) and titanium group prepared by selective laser melting (GTi2). Surface roughness values (Ra, Rz) of six samples from each group were examined and scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX) analyses were applied to one specimen from each group, and failures were examined using a stereomicroscope.

Results

The mean SBS, Ra and Rz values for GC were 21.21 ± 6.15 MPa, 5.79 ± 0.58 µm and 37.54 ± 6.87 µm; GTi1 were 25.80 ± 7.79 MPa, 2.69 ± 0.41 µm and 26.05 ± 7.60 µm; GTi2 were 24.32 ± 7.07 MPa, 5.91 ± 0.51 µm and 44.46 ± 7.78 µm respectively. Although the shear bond strength did not show a significant difference between the groups, the roughness values of the GTi1 group were statistically significantly lower than those of the other groups (p < 0.05). The roughness values of the GTi2 and GC groups did not show a statistically significant difference (p > 0.05).

Conclusions

The ceramic bonding of titanium specimens produced by both methods showed values similar to those of Co-Cr and ceramic bonding used in routine treatments.

Introduction

Metals have been widely used as biomaterials for centuries [1]. Precious metal alloys have been successfully used in dentistry for many years due to their good mechanical properties, excellent biocompatibility and superior ceramic bond strength [2]. However, they have been replaced by base metal alloys due to their high cost. Among the most commonly used base metal alloys are Co-Cr and Ni–Cr alloys. Nickel–chromium alloys contain at least 60% nickel and at least 20% chromium. The addition of chromium to Ni–Cr alloy increases corrosion resistance. They have high tensile strength compared to precious alloys. It can be made thinner than precious alloys and metal ceramic connection is strong. Today, Co-Cr-based alloys are almost exclusively used for the production of metallic frameworks of removable partial dentures and porcelain fused to metal restorations. Despite the strong biomechanical properties of Ni–Cr alloys, the use of these alloys has been abandoned in most countries due to allergic and toxic reactions when exposed to the oral cavity, and they have been replaced by Co-Cr alloys [3, 4].

The high price of precious alloys and concerns about base metal alloys'allergic and carcinogenic properties have increased interest in using titanium (Ti) to construct prosthetic restorations in dentistry [5].

Ti has become a preferred material in prosthodontics due to its advantages, such as high strength, low weight and elastic modulus, low thermal conductivity, excellent corrosion resistance, and lower cost than typical noble metal materials. In addition to its outstanding biocompatibility, it has become an essential material of choice for fixed or removable partial dentures for patients with metal allergies [6, 7].

The oldest and most traditional method of metal fabrication in dentistry is casting [1]. However, casting Ti alloys presents several technical problems, such as shrinkage and porosity. This is due to the extremely low density of Ti (4.5 g/cm³), which reduces its castability [8]. When casting Ti, a major issue is the extremely high melting temperature of the metal. During the casting process, the molten Ti reacts with the components of the mould at high temperatures. Additionally, free oxygen atoms from the air can diffuse into the Ti, forming a hard surface layer known as the"alpha-case layer"on the cast surface. This layer is responsible for a decrease in mechanical properties such as ductility, fatigue resistance, and bond strength to dental porcelain [9]. Due to casting difficulties, Ti has not been widely adopted in metal-ceramic fixed partial dentures (FDP). Computer-aided design and computer-aided manufacturing (CAD-CAM) technologies have provided alternative methods for fabricating Ti substructures for FDPs. Subtractive CAD-CAM systems mill metal frameworks from pre-prepared metal blocks, eliminating Ti castings'shrinkage and porosity problems [10, 11]. The main disadvantages of this method are the high cost, the wear and tear of the milling cutters during production and the fact that complex and delicate details can only be produced to a limited extent, depending on the geometry of the insert. In addition, there is much material wastage due to the scraping of blocks during production [12, 13].

On the other hand, additive manufacturing technology, which has become popular in recent years, can be preferred by different sectors thanks to its significant advantages and can be applied to various types of products. Additive manufacturing technology, particularly in metal production, is becoming increasingly widespread thanks to its ability to design complex geometries and respond quickly to manufacturing iterations. Adapting lightweight, strong, temperature and corrosion-resistant materials to the technology has enabled the production of high-performance complex components, and these capabilities have attracted the attention of the aerospace industry. The rapid prototyping capabilities of this new technology have also made it the preferred choice for medical and dental applications [13]. Additive manufacturing is applied through melt deposition/mass modelling, stereolithography, selective laser sintering, and selective laser melting (SLM). SLM is an additive manufacturing technology based on the laser melting of multiple layers of powder material to form a three-dimensional structure. It produces substructures for dental restorations [14,15,16].

The long-term success of prosthetic restorations depends on many factors, including the bond strength between the framework material and the ceramic [17]. Few studies have been conducted on the bonding of digitally fabricated Ti-based alloys to ceramics, and there are still questions about Ti-ceramic bonding. This study aimed to investigate the ceramic bonding of Ti alloys obtained by different manufacturing techniques and compare them with Cobalt-chromium (Co-Cr) alloys. The study's null hypothesis is that different manufacturing methods and materials will not affect the framework and ceramic bond strength.

Material and methods

Sample size calculation

The sample size for this study was calculated at the 95% confidence level using G.Power-3.1.9.2. The power analysis, with a power of 0.95 and a standardized effect size of η² = 0.456, indicated that the total sample size should be a minimum of 51, with 17 samples in each group.

Preparing the study groups

51 disc-shaped specimens were desinged with a diameter of 8 mm and a height of 3 mm, and then three groups of 17 specimens were created according to metal type and titanium manufacturing method: Co-Cr group prepared by SLM (GC), Ti group prepared by milling (GTi1) and Ti group prepared by SLM (GTi2).

Fabrication of metal samples

After the samples'designs were made and transferred to STL data. The samples for the GC group were fabricated from Co-Cr alloy powder (Puresphere 43,024, Sentes-BIR) with a laser sintering device (Mysint100 PM, Sisma S.p.A) and the metal discs were separated from the vertical bars and then heat treated and levelled; the samples for the GTi1 group were fabricated by milling Ti blocks (KERA®Ti5-DISC, Eisenbacher) and the samples belonging to the GTi2 group were prepared by using Ti6 Al4 V powder (ERMAK A252, Ermaksan) at 190 W laser power at a working distance of 30 microns in a SLM device (ENAVISION250, Ermaksan Additive) and, after separation from the vertical bars, they were subjected to a heat ageing process at 500 °C for 4 h in an argon gas oven (Carbolite Tube Furnace, CARBOLITE GERO) before proceeding to the levelling process.

Surface preparation of metal samples

The prepared Cr-Co and Ti disc samples were roughened with 110 microns of Al₂O₃ sand at a distance of 10 mm, at a pressure of 1.5 atm, for 10 s, and at an angle of 45 degrees. All surface-prepared samples were then cleaned with an ultrasonic cleaner (BK-3550, BAKU).

Surface roughness measurements

For the surface roughness measurements, five samples were randomly selected from each group, and the surface roughness values were recorded on an optical profilometer (ST400, Nanovea) at a scanning frequency of 2000 Hz with a scanning accuracy of 3 nm.

Porcelain firing on metal specimens

Silicone moulds were prepared for the standard application of porcelain paste on Co-Cr and Ti specimens. The same technician fired porcelain (VITA 3D-Master, VITA) for Co-Cr specimens and low-temperature porcelain (VITA LUMEX AC, VITA) for Ti specimens, according to the manufacturer's recommendations, applying a 4 mm diameter and 3 mm height onto the metal surfaces.

Shear bond strength test

All specimens were embedded in acrylic moulds for the shear bond strength test. A load of 500 N was applied in a universal testing machine (Esetron Smart Robotechnologies,

Mod Dental) at a head speed of 1 mm/min until the porcelain separated from the metal. The force at which the fracture occurred was calculated as MPa.

Scanning Electron Microscope (SEM) and energy-dispersive X-ray spectroscopy (EDX) analysis

After the SBS test, one sample was selected from each group. The surfaces of the samples were coated with carbon, and images were recorded using an SEM device (LEO-1430 VP, Carl Zeiss AG) at 1800 × magnification.

Statistical analyses

The SPSS V25 package was used for statistical analyses. Descriptive statistics of categorical variables in the study were presented as frequency and percentage, and descriptive statistics of continuous variables were presented as mean and standard deviation values. The Shapiro–Wilk test was used to analyse the suitability of continuous variables for normal distribution. The homogeneity of variance was tested using Levene's test. The one-way ANOVA test compares three or more independent groups with a normal distribution. In comparison, the Kruskal–Wallis test evaluates three or more independent groups that are not normally distributed. For all statistical comparisons in the study, comparisons with a p-value below 0.05 were considered statistically significant.

Results

The groups'roughness (Ra, Rz) and SBS values are shown in Table 1. According to the one-way ANOVA test results, there was a statistically significant difference between the Ra values of the groups (df = 2 F = 9.400 p = 0.002). According to the Tukey test used for post-hoc comparison, there was a statistically significant difference between GC and GTi1 (p = 0.043) and between GTi1 and GTi2 (p = 0.002). On the other hand, there is no statistically significant difference between GC and GTi2 groups (p = 0.270). According to the Kruskal Wallis test results, there was a statistically significant difference between the Rz values of the groups (p = 0.003). As a result of the pairwise comparisons, there was a statistically significant difference between GC and GTi1 (p = 0.020) and between GTi1 and GTi2 (p = 0.005), while there was no statistically significant difference between GC and GTi2 groups (p = 1.000). According to the one-way ANOVA test for SBS comparison, there was no statistically significant difference between groups (df = 2 F = 1.887 p = 0.163). The 3D topography of the surfaces of one sample from each group is shown in Fig. 1A-C, and SEM images and EDX analyses of one sample from each group after the SBS test are shown in Fig. 2A-C and Table 2. The failure distributions of the groups are shown in Fig. 3; 82.36% of the samples in the GC group were mixed type, 11.76% were cohesive type, and 5.88% were adhesive type; 94.12% of the samples in the GTi1 group were mixed type, and 5.88% were adhesive type; all samples in the GTi2 group showed mixed type failure.

Table 1 Mean ± standard deviation of Ra, Rz and SBS values and statistical relationships between groups

Fig. 1

3D topography of the surfaces of one sample from control group consisting of Co-Cr specimens (GC) (A), titanium specimens produced by milling (GTi1) (B), titanium specimens produced by the SLM method (GTi2) (C)

Fig. 2

SEM image of one sample from the control group consisting of Co-Cr specimens (GC) (A), titanium specimens produced by milling (GTi1) (B), titanium specimens produced by the SLM method (GTi2) (C)

Table 2 The amount (Weight %) of the elements in the samples of each group

Fig. 3

Failure distribution (%) of groups. GC: Control group consisting of Co-Cr specimens, GTi-1: Titanium specimens produced by milling, GTi-2: Titanium specimens produced by the SLM method

Discussion

In this study, the bond strength of Ti fabricated by different methods with veneer ceramics was investigated to assess the usability of Ti, which has many favorable properties, as a substructure material. As a result of the research, it was determined that the different production techniques did not create a significant difference between Ti and ceramic and the null hypothesis established at the beginning of the study was accepted.

Grades I and II commercially pure titanium Ti (cpTi) are generally used for the fabrication of prosthetic frameworks; the strength of pure Ti is comparable to that of gold alloys and is, therefore, sufficient for FDPs [6]. However, in cases where superior mechanical properties are required, Ti is used in its alloyed form. The Ti-6 Al-4 V alloy is one of the most commonly used titanium alloys in dentistry and contains 6% aluminium and 4% vanadium [18]. The bond strength of Ti-6 Al-4 V and cpTi to porcelain is generally reported to be similar [2, 19, 20]. Therefore, both cpTi and Ti-6 Al-4 V alloys are suitable for fabricating prosthetic structures. However, these two alloys have significant differences in composition and microstructure, and higher porcelain bond strengths have been reported for Ti-6 Al-4 V alloy compared to cpTi [21, 22]. Therefore, the Ti6 Al-4 V alloy was used in this study to evaluate the bonding of metal substrate material to porcelain.

The alpha-case layer formed on Ti surfaces by casting adversely affects the material's mechanical properties [23]. Although significant advances have been made in casting titanium alloys, forming a thick α-sheath cannot be avoided. Therefore, fabricating titanium by casting is complex, and the development of alternative machining methods, such as CAD-CAM, seems to limit the use of titanium casting methods [11]. Therefore, this study analysed a comparison of specimens produced by CAD-CAM instead of casting.

Furthermore, one of the significant causes of porcelain bonding failure in Ti-porcelain systems is the high affinity of Ti for oxygen. However, chemical bonding requires a surface oxide layer of controlled thickness; when Ti is exposed to high temperatures (above 800 °C) during porcelain firing, an uncontrollable oxide layer forms. This extremely thick oxide layer can easily separate from the metal substrate, causing the Ti-ceramic bond to fail. To avoid uncontrolled reactivity of Ti, porcelain firing temperatures must be kept below 800 °C [24]. Therefore, a new generation of low-temperature dental porcelains with adjusted firing temperatures (below 800 °C) has been developed for coating titanium frameworks. Vita Lumex AC, a new generation of low-temperature porcelain, was used in this study.

ISO 9693–1:2012 recommends a bond strength of 25 MPa for the bonding of substrate and veneer material in metal-ceramic systems. In this study, the GTi1 group met this standard with a bond strength of 25.26 MPa. However, despite initially high expectations for milled Ti-ceramic systems, an increased risk of ceramic failure has been observed [25,26,27,28,29]. A study published in 2013 examining the mechanical complications of milled Ti-ceramic single crowns after six years of clinical observation reported a low success rate (67.8%) for milled Ti-supported restorations. The researchers stated that this may be related to non-optimised substructure design and that improvements in CAD software may increase the success of milled titanium-supported restorations [30]. This study found the highest bond strength in the milled group, possibly because designs and fabrications are more optimised in parallel with the developments in CAD-CAM systems today.

Although the GTi2 group was slightly lower than GTi1, it was close to the ISO standard. In addition, there was no significant difference in ceramic bond strength between the Ti samples prepared by both methods and the Co-Cr samples in the control group. Co-Cr substructure ceramics have been used in prosthetic treatments for many years. In this context, the ceramic bonding of Ti specimens prepared by either method is similar to that used in routine treatment. In addition, SEM and stereomicroscope images showed that the fracture type was mainly mixed, and the presence of silicon atoms involved in the main structure of the ceramic on the fracture surfaces in EDX analysis indicated a strong bond between titanium and ceramic. On the other hand, the SBS of the GC group obtained by the SLM method in this study was found to be below ISO standards. The manufacturing techniques of Co-Cr substructures influence their bond strength to ceramics. Altuntaş and Güleryüz compared the bond strength of Co-Cr substructures produced by different techniques. They reported that the oxide layer formed between the substructures produced by the SLM method and the ceramic, which is required for bonding, is more uncertain compared to the casting and milling methods and that the samples obtained by the SLM method have a lower bond strength [31]. The low SBS of the GC group is thought to be due to the manufacturing process's failure to properly form the oxide layer, which provides the chemical bond between the Co-Cr samples and the ceramic.

In this study, although there were significant differences between the surface roughness values, the SBS of the groups were similar. The surface roughness value was lowest in the GTi1 group, with the highest bond strength. In addition, although the roughness of the Co-Cr substructures of the control group was higher than that of the GTi1 group, the bond strength was lower. The absence of ceramic fractures or metal deformation in metal-ceramic restorations depends on the mechanical compatibility of the two materials. There are four main mechanisms of mechanical compatibility: chemical bonding, the occurrence of mechanical interlocks, van der Waals forces and compressive forces. The first and most important of these bonding mechanisms is chemical bonding, which is responsible for forming an intermediate oxide layer between the two materials [32]. Although a positive correlation between roughness values and SBS would be expected, the results indicate that chemical bonding is more effective than mechanical interlocking in the bond strength between metal and ceramic. Additionally, recent findings by YANG et al. highlight the influence of microstructure on metal-ceramic bonding in SLM-manufactured titanium alloy crowns and bridges, suggesting that careful control of processing parameters may further improve bond strength in these systems [33].

This study has some limitations. First, disc-shaped specimens were used instead of more complex restorative shapes, which may not fully represent clinical scenarios. Additionally, the in vitro design did not simulate actual clinical conditions. The sample size per group was also relatively small, and only one type of surface treatment (Al₂O₃ sandblasting) was utilized.

Furthermore, another significant limitation is that aging or thermocycling was not incorporated in this study. These factors can significantly impact bond strength over time; therefore, their omission may affect the generalizability of the results.

Future research could address these limitations by using larger sample sizes, examining various surface conditioning methods, and conducting more detailed microstructural analyses, as well as including aging and thermocycling to better understand the long-term performance of the bond strength.

Conclusions

Within the limitations of this study, the following conclusions can be drawn;

• 1. The GTi1 group exhibited the highest shear bond strength. While this suggests favorable bonding characteristics, it is important to note that the differences among all groups were not statistically significant, indicating that all methods may provide comparable bonding performance.

• 2. The GTi2 group showed the highest surface roughness value compared to the other groups. Higher surface roughness may enhance adherence and could be an important consideration in clinical applications.

• 3. The ceramic bonding of titanium specimens produced by both additive and subtractive methods demonstrated values similar to those of Co-Cr and ceramic bonding typically used in routine treatments. This finding suggests that both methods are equally effective in achieving reliable ceramic bonding with titanium specimens.