Effect of Recasted Material Addition on the Quality of Metal-Ceramic Bond : A Macro-, Micro-, and Nanostudy

Using the recasted alloys in dental prosthetics could affect the quality of the metal-ceramic bond. However, scientists, alloys producers, and prosthetists are still of different opinions..e purpose of this study was to estimate the influence of recasting of the NiCrMo alloy on the metal-ceramic bond quality..e research was carried out on macro-, microand nanoscales using the threepoint bending test procedure and hardness tests as well as atomic force microscopy and SEM analyses. .e SEM analyses showed good integrity of the metal-ceramic bond. .e τb index of all test samples was greater than 45MPa. .e highest values were recorded for the samples made of 50% and 100% of a brand new material. SEM analysis made after the bending test confirmed goodmetal-ceramic bond and exhibited adhesive-cohesive fracture..e largest hardness of metal plates was found for the samples containing 50% of the recycled material. Atomic force microscopy studies showed that the alloy containing 50% of the recycled material was characterized by the highest values of surface roughness parameters.


Introduction
e metal-ceramic dentures are still widely used due to good combination of their durability-high strength of metal substructures and esthetics of porcelain.ey are also more economical than implants.Despite great popularity of precious metal alloys, the nonprecious Co-Cr and Ni-Cr ones are often used interchangeably because of their good mechanical properties and low costs.
Despite questioning the use of nickel alloys as biomaterials, due to the potential harmful effects on human tissues, nickelbase alloys are still widely used to manufacture the substructures of ceramic crowns and bridges.ese alloys have larger elasticity modulus than the gold ones and thus thinner cross section of the alloy can be used to reduce destruction of the healthy tooth during crown manufacturing [1].Moreover, the coefficient of thermal expansion of nickel alloys is compatible with that of thermal expansion of conventional ceramics that are used to produce dental veneers which provides a good metal-ceramic bond [2].
e final quality and reliability of partial dentures is influenced mainly by properties of denture and their design [3], the casting method [4,5], and the use of already melted materials for casting its metal substructures [6].e practice of using recasting materials is a very common way to reduce the cost of prosthetic components manufactured in dental laboratories.However, producer's and researcher's points of view are different in this case.Some manufacturers of dental alloys permit the use of once melted alloys but not less than the 50% addition of the brand new material.ey also require that the material must come from the same batch.Another group of manufacturers do not allow using remelted material (e.g., Heraeus Kulzer Co.) or do not provide any information on the use of dental alloys obtained from recycling.
Recasting of Ni-Cr dental alloys is the current topic in the literature.Researchers examine the effect of using recasted materials on properties such as chemical composition and precipitate formation and microstructure and its influence on mechanical properties, including testing of metal-ceramic bond strength.
Characterization of microstructure and metal-ceramic bond quality is generally made using several methods, for example, scanning electron microscopy (SEM) and atomic force microscopy (AFM).ese methods have also been used to evaluate the surface morphology [7].AFM provides three-dimensional detailed topographical images of surface roughness on a nanometer scale and has been used in dental research [8][9][10][11][12][13][14].Roughness could be defined as a complex role of irregularities or little indentations that characterize a surface and influence on, inter alia, metalceramic bond quality [15].Determination of surface roughness plays an important role in materials used for dental prostheses.
In spite of that the topic of recycling of nickel-based dental alloys is frequently discussed by research institutions, there is still no clear answer whether and how the use of the recycled material affects the quality of the metal-ceramic bond.ere is also a lack of considering this issue from the perspective of different scales, which would allow a more complete identification of the causes of possible differences in the results of individual tests.
e purpose of this study was to investigate the influence of recycled material addition on the quality of the metalceramic bond (for the NiCrMo alloy), basis on tests in the micro-and nanoscales (surface properties) as well as in the macroscale (bond strength).

e Alloy and Casting
Method.Specimens for all tests were made of the Heraenium NA alloy (Heraeus Kulzer GmbH, Germany), whose chemical compositions were as follows (by weight): Ni-59.3%Cr-24% Mo-10% Fe-1.5%,Mn-1.5%, and Ta-1.5% Si-1.5% Nb-1.0%, according to the manufacturer's data [16].e abovementioned alloy is used for casting crowns and bridges.e castings were made using a disposable alloy with a starting composition of 100%, 50%, and 0% of the brand new material made up to 100% by the recasted material according to Table 1.All materials came from the same batch.Casting and dental ceramic coating were carried out in a professional prosthetic laboratory according to the procedures applied to manufacture dental crowns and bridges.
e vacuum pressure casting method (Nautilus, Bego Co.) and the ceramic crucibles (Nautilus) were used).
e specimens were purified from the investment material using first the coarse (250 µm) and then the precision (100 µm) abrasive cleaning (sandblasting).Sandblasting time for all samples was equal (15 s for each particle size).After sandblasting and cleaning the casting channels, from specimens "100" were used to manufacture the second and third groups of test specimens (foundry cones were not reused).

Samples and Porcelain
Coating Method.Shapes and dimensions of the samples were in agreement with the ISO 9693 standard [17].e samples had the shape of rectangular plates with dimensions of 25 × 3 × 0.5 mm.ey were cast from patterns cutout from wax sheets of 0.5 mm thick.To make foundry molds, the Bellavest SH investment bond (Bego Co.) was applied.e 10 minutes oxidation program was applied for all tested samples (1223.15K).
e castings were coated with d.Sign porcelain (opaque, dentin, glaze, and Ivoclar Vivadent) into the rectangular area (8 × 3 mm) in the center of one side of each metal specimen according to the manufacturer's specification.e samples were checked for dimensional correctness and the presence of potential external casting defects.Defected specimens were removed from the tests.e image of prepared specimens is presented in Figure 1(a).

SEM and AFM Structure Characterization.
Characterization of specimen's surface was made by SEM and atomic force microscopy (NTEGRA Prima, NT-MDT, Moscow, Russia).Images of ceramic and metal surfaces were obtained using the scanning electron microscopy Nova NanoSEM 450 (FEI, Eindhoven, Netherlands) before and after the bending test.For the preparation for SEM observations sputtering of the samples with the gold layer of thickness about 7 nm was applied.Additionally, chemical composition analyses using the EDS method were carried out.Analyses were made using the Octane Pro EDS detector (EDAX) with the excitation voltage of 15 kV.Each specimen was analyzed three times, and the signal was detected from the area of about 600 × 800 µm.e metal surface roughness parameters (R q is the root-mean-square deviation of the roughness profile, R a is the arithmetical mean deviation of the roughness profile, R p is the maximum peak height of the roughness profile, R v is the maximum valley depth of the roughness profile, and R z is the maximum height of roughness profile) were determined with NT-MDT software application [13].In order to determine surface roughness of the alloy, the areas of 10 × 10 μm were scanned with 512 × 512 data points in the semicontact mode using a silicon cantilever (NSG30; NT-MDT) 125 μm long, 4.0 μm thick with a tip radius of 10 nm, and the average resonance frequency of 300 kHz.e scan rate was 1 Hz.Regions of interest were defined under the optical microscope.For each of the samples, three different areas, not closely located, were chosen.Obtained height images were used for determination of surface roughness parameters.All parameters were calculated for 15 lines per one image, and e roughness parameters were checked for normal distribution and variances homogeneity (Levene's test).As the data were not normally distributed, the Kruskall-Wallis analysis of variance was applied to assess significant effects of recasted NiCrMo alloy addition on the determined parameters.Multiple comparisons of mean ranks were used to determine significant differences among the experimental groups (at the significance level of 0.05).

e Bending Test and Hardness Measurements.
After the preliminary bending tests, using the power test analysis from the Statistica package, the number of samples needed for testing was determined.e test power was equal to 0.8, the value of the significance level was α ≤ 0.05, and the measurement error was equal to 5%.
Twelve samples were selected from each batch for the bending test, according to the ISO 9693-1:2012 standard [17].
e three-point bending test was performed on a Instron 8801 machine equipped with a 0.1 N accuracy load cell.e distance between the supports was 20 mm, and the diameter of the rollers supporting the sample was ⌀1 mm.
e test was conducted at the crosshead rate of 1.5 mm/min.e measurements were made up to a significant reduction in the value of the force appeared.is indicated the destruction of the metal-porcelain connection.e maximum recorded force was taken as the fracture force (F fail ).e bond strength was calculated as the bonding compatibility index τ b [17] using the following equation: e fail force Ffail was multiplied by the coefficient k 1 , which depends onthe thickness of the metal substrate, dM d M ((0.5 ± 0.05) mm), and the Young's modulus, E M and the Young's modulus value, EM, of the used metallic materials.Young's modulus was provided by the alloy manufacturer (for Heraenium Na, the value was equal to 222 GPa).Hardness measurements (on the metal) were performed using the Vickers hardness tester FM-800 (Future-Tech Corp.).e applied load was 4.9 N which corresponds to the scale HV0.5.e duration of action of force was 10 s.

Results and Discussion
Microstructural (SEM) analysis showed a good integrity of the metal-ceramic bonding in each case of tested specimens, regardless of the content of recycled material.Metal surface is characterized by a homogeneous structure, without any cracks and porosity (Figures 1(b) and 1(c)).On the ceramic surface, cracks were not observed either (Figure 1(b)).In a lateral view, the presence of microporosity in the ceramic layer was observed in most specimens (Figure 2).However, this fact cannot be  associated with the content of the recycled material in metal castings.It would rather be due to precision of porcelain application on the sample surface.It is noteworthy that the precise of the application of the ceramic layer was the same in each case.
e results of chemical composition analyses (EDS) are presented in Figure 3 and Table 2 and contain the mean values.
ere were no signi cant di erences observed in the amount of metallic elements for each type of the specimen.e di erences were observed only for the carbon  .However, due to the EDS method limitations, the quantitative results were not presented.e results of surface roughness analysis are shown in Table 3. e highest values of all calculated roughness parameters were observed for the specimens marked 50, whereas the 0 specimens were characterized by the smallest surface roughness (Table 3). is is also shown in Figure 4, which presents typical 2D and 3D topography images of investigated alloys.e statistical analysis showed significant effects of recasted NiCrMo alloy addition on R q , R a , R p , R v , and R z parameters (P < 0.05).Despite the fact that the roughness was significantly the largest for specimen 50, its average value was still small (R q below 0.5 μm).
e low roughness is very important to avoid bacterial adhesion (size of about 1 μm) on the alloy surface when parts of the crown, not coated with the porcelain veneer, may remain in contact with the oral environment [18][19][20][21].
e τ b index of all tested samples, determined at the three-point bending test, was higher than 45 MPa (Table 4).
is indicates good connection of the pair metal-dental ceramic.According to Craig [22], a proper metal-ceramic bond exists when the τ b index reaches over 25 MPa.e highest values were recorded for the samples cast at 50% and 100% of the brand new material (approximately 48-49 MPa).As the data did not meet the normal distribution conditions, the Kruskall-Wallis analysis of variance was applied to estimate significant effects of recasted NiCrMo alloy addition on the bonding compatibility index.e test did not show significant differences between values of τ b obtained for casting composed of 100%, 50%, and 0% of the brand new material (Table 4).
e macro-and microscopic observations of the samples after the bending test are shown in Figure 5.
e crack appearing between the dental ceramic and metal substrate is shown in Figure 5(a).It is visible only in the SEM imaging.e shape of the cracks takes the lamellar form with the angle between the crack and the metal substrate of about 10 degrees.
In almost all cases, the type of damage was mixed, adhesivecohesive.e next image (Figure 5(b)) shows the surface of the metal substrate after separation of the porcelain from the metal.e beginning of the cracking in the metal-ceramic joint was observed on the edge of the porcelain layer (maximum stresses).e presence of the thick residue ceramic layer at the ends of the layer area can be observed.
e hardness test showed that the highest values were found for the samples containing 50% of the recycled material (Table 5).
e lowest hardness values characterized castings made entirely from the recycled material, but the differences between samples 0 and 100 were not statistically significant.e ANOVA analysis demonstrated that there were statistically significant differences between the hardness of samples 100 and 50 as well 50 and 0 (Table 5).
An explanation of different hardness and surface roughness is required due to the same parameters of the sandblasting process.Sandblasting is an abrasive process which effectiveness and results are dependent on the mechanical properties of the material.In turn, the mechanical properties are related to, except chemical composition, the microstructure.e microstructure determines the susceptibility for abrasion.Materials exhibiting ductility and high strength are less susceptible for abrasion.Materials which are hard and brittle exhibit higher susceptibility for the erosion process.Microstructure of prosthetic alloys consists of solid solution of alloying elements in nickel with precipitations of intermetallic phases [3].e amount of precipitates, related to materials used and the manufacturing process, corresponds with the hardness.e presence of precipitates causes brittleness and as a result higher roughness.It is confirmed by the appearance of the surface obtained from AFM (Figure 4), where steep slopes with sharp edges are visible.
e results are consistent with the similar study for the Remanium CS+ alloy (from Dentarum Co.) obtained by Walczak [23,24], but they are not consistent with the Ucar's et al. research [25] for the RemaniumCSe alloy (the same manufacturer).ese two alloys (Remanium CS+ and   [25] showed that the samples were evaluated visually (correctness of execution).However, in this study, the measurements of metal substrate hardness and the evaluation of surface roughness were not presented.Also, higher hardness values of metal framework cause greater 6 Advances in Materials Science and Engineering rigidity of the metal-ceramics system and result in quality differences of the metal-ceramic bond.A significant adverse effect of recasting NiCr alloys on the strength of the metalceramic bond and different points of view were presented by Madani et al. e authors reported the largest adhesion of dental porcelain on the castings made of the brand new material [26].e study of multiple castings of nickelchromium alloys was also conducted by Mirkowić [27].He noticed the occurrence of inconsiderable linear decrease in the elastic modulus up to the sixth generation of castings.

Conclusions
e study showed very good metal-porcelain bonding for all kinds of investigated castings.
e best quality of the metal-ceramic bond was recorded for the material containing 50% of the brand new alloy; however, the differences between the groups were not statistically significant.e SEM analysis conducted after the bending test confirmed good metalceramic connection and exhibited the adhesive-cohesive fracture.
e results of the bond strength are also confirmed by the microanalyses of the surface (SEM).Analysis of the surface morphology (AFM) revealed that the highest values of calculated roughness parameters were obtained for the material containing 50% of the brand new alloy.e metal-ceramic bond does not deteriorate quality when the material from recasting is used.Any small differences are due to those in hardness of the metal substrate and/or roughness.
ey do not affect the final quality of bonding.Further researches are planned to obtain required strength and corrosion resistance of the recasted material and improve the metal-ceramic bond.

Figure 1 :
Figure 1: e exemplary groups (SEM images): (a) specimens used for the bending test; (b) metal-ceramic bond surface (view from the top); (c) metal-ceramic bond surface (lateral view) .

Table 5 :
e mean values of hardness (±SD) for each group of specimenswith dissimilar indices are significantly different at P < 0.05.e letters a and b indicate the existence of significant differences between the mean values of hardness.

Table 1 :
Marking of the specimens.
finally, a set of 540 data per experimental group was taken for further analysis.

Table 3 :
e mean values of surface roughness parameters (with SD values) and Kruskal-Wallis H K-W statistics results calculated for each parameter.q , root-mean-square deviation of the roughness profile; R a , arithmetical mean deviation of the profile; R p , maximum roughness profile peak height; R v , maximum roughness profile valley depth; R z , ten-point height of irregularities; the mean values with dissimilar letters in the same column are significantly different at P < 0.05.e letters a, b, and c indicate the existence of significant differences between the mean values of surface roughness parameters. R
RemaniumCSe) have almost identical chemical composition.Ucar et al.