Abstract
Objectives
The aim of this study was to evaluate strengthening mechanisms of yttria-stabilized zirconia (YSZ) thin film coatings as a viable method for improving fracture toughness of all-ceramic dental restorations.
Methods
Bars (2 mm × 2 mm × 15 mm, n = 12) were cut from porcelain (ProCAD, Ivoclar-Vivadent) blocks and wet-polished through 1200-grit using SiC abrasive. A Vickers indenter was used to induce flaws with controlled size and geometry. Depositions were performed via radio frequency magnetron sputtering (5 mT, 25 °C, 30:1 Ar/O 2 gas ratio) with varying powers of substrate bias. Film and flaw properties were characterized by optical microscopy, scanning electron microscopy (SEM), and X-ray diffraction (XRD). Flexural strength was determined by three-point bending. Fracture toughness values were calculated from flaw size and fracture strength.
Results
Data show improvements in fracture strength of up to 57% over unmodified specimens. XRD analysis shows that films deposited with higher substrate bias displayed a high %monoclinic volume fraction (19%) compared to non-biased deposited films (87%), and resulted in increased film stresses and modified YSZ microstructures. SEM analysis shows critical flaw sizes of 67 ± 1 μm leading to fracture toughness improvements of 55% over unmodified specimens.
Significance
Data support surface modification of dental ceramics with YSZ thin film coatings to improve fracture toughness. Increase in construct strength was attributed to increase in compressive film stresses and modified YSZ thin film microstructures. It is believed that this surface modification may lead to significant improvements and overall reliability of all-ceramic dental restorations.
1
Introduction
All-ceramic dental restorations have become extremely popular due to their outstanding esthetic characteristics and biocompatibility . However, these materials, when placed in load bearing applications, often display an inability to resist stress-induced crack propagation . Analysis of clinically failed all-ceramic restorations revealed that the majority of failures were initiated at the surface, where tensile stresses and flaws were found to accelerate construct failure . Additionally, clinicians typically etch or particle air-abrade the inside of the restoration to roughen surfaces for enhanced adhesion, inducing large surface flaws and further promoting premature failure . This study seeks to modify surface stress states and facilitate flaw modification to increase the fracture toughness of dental porcelain.
Fracture toughness improvements have been demonstrated through various surface modification methods; such as, flaw modification, heat treatments, and coatings. Fundamentally, by controlling surface finishing techniques, fracture toughness can be enhanced through modification of flaw size or geometry, resulting in reduced stress concentration in the local area of a critical flaw and therefore increasing overall strength of a specimen . Compressive stress gradients can also be introduced to a specimen through thermal processing, where a surface is cooled at a faster rate than the interior of a specimen, resulting in a compressive surface . Composition gradients and ion implantation techniques have also been used to induce strain into the crystal lattice, resulting in compressive stresses on a specimen surface . These stress-modification mechanisms require crack expansion (or increased tensile stresses), to overcome the induced compressive stresses on the surface of a specimen. As a result, crack propagation is inhibited, resulting in increased fracture toughness.
Thin film coatings have been employed to strengthen substrates; first through critical flaw modification and secondly through inducing compressive surfaces stresses . Ruddell et al. have shown improvements in fracture strength of up to 19% by coating ceramic substrates with less than 10 μm of sputtered metallic thin films . Earlier work reported increases in strength of up to 32% by using a multilayer film (10 μm) structure consisting of alternating layers of yttria-stabilized zirconia (YSZ) (1 μm) and parylene (1 μm) to induce crack deflection . Additionally, it was determined that a 2–3 μm thick YSZ thin film sputtered on porcelain substrates provide the maximum benefit for increased construct fracture strength .
YSZ is a material of particular interest not only due to its biocompatibility, but also for its excellent wear resistance properties . Reports have shown that microstructure and film stress properties of YSZ thin films can be tailored through the control of deposition parameters . By applying a substrate bias during deposition, an interrupted microstructure consisting of lateral microcracks and increased compressive films stresses can be obtained. It is therefore hypothesized that these structures may provide an ideal thin film for increasing the fracture toughness of ceramic substrates.
This study examines the benefits of varying YSZ thin film microstructures and magnitudes of compressive thin film states on substrates that have been fabricated with a controlled flaw size of a given geometry. Analysis of construct fracture strength and measurements of critical flaw size allows for the determination of overall construct fracture toughness. Here we discuss the impact of thin film induced compressive stresses, flaw modification, and crack deflection on increased fracture toughness of modified dental porcelain.
2
Materials and methods
Porcelain substrates were fabricated by cutting 18 × 2 × 2 mm bars from leucite-reinforced feldspathic porcelain blocks (ProCAD, Ivoclar-Vivadent, Schaan, Liechtenstein). All four surfaces of porcelain substrates were polished through 1200 grit (CAMI Grit Designation) SiC abrasive and edges rounded to limit corner/edge failures. Substrates were ultrasonically cleaned in acetone to remove surface debris prior to deposition, then randomly divided into four groups ( n = 30). A Vickers indenter (Mitutoyo Model AAV-500, Aurora, IL) was used to apply a single, controlled flaw into the center of a substrate surface with indent corners perpendicular to substrate edges, in order to normalize failures through control of critical flaw size and geometry. A 0.5 N load was applied with a load time of 15 s, resulting in an indent size of 34–37 μm to simulate critical flaw sizes in clinical surface preparation methods .
Thin film depositions were performed in an r.f. magnetron sputter reactor (CVC Model SC-400, Rochester, NY). A second r.f. power source was capacitively coupled to the substrate and used for depositions requiring a substrate bias. Target material used for sputtering was 99.99% pure zirconia doped with 3 mol% yttria (Plasmaterials, Livermore, CA). Substrates were mounted directly above the sputtering gun (US Gun, San Jose, CA) in a custom fixture at a working distance of 75 mm. At the start of each deposition, a 50 W substrate bias in an argon atmosphere was applied for 5 min to clean organic debris from specimen surfaces and to promote film adhesion. All depositions were performed with a target power of 350 W, atmosphere of 30:1 argon to oxygen, and substrate bias varied between 0 and 100 W. Films were grown to a thickness of 2–3 μm to maximize fracture strength .
X-ray diffraction (XRD) (PANalytical X’Pert PRO MRD HR, Westborough, MA) was used to determine the percentage of tetragonal and monoclinc phases in the YSZ films and thus qualitatively determine trends in film stress . Fracture strength values were determined using a three-point bending fixture, with specimens oriented with films in tension per ASTM Standard C1161 . Flexural testing (Instron Model 5542, Norwood, MA) was conducted with a 10 mm span at a crosshead speed of 0.5 mm/min. Bar dimensions were measured prior to fracture for fracture stress calculations. Peak loads and fracture stresses for each specimen were recorded and single-factor analysis of variance (ANOVA) at a 5% confidence level was performed to evaluate for statistical similarities. For testing controls, a polished sample set with a Vickers indentation was left unmodified by films in order to determine inert strengths of substrates at a given flaw size. Post fracture, scanning electron microscopy (SEM) (Hitachi, S-4700 FE, Tokyo, Japan) and optical microscopy (Sony DXC-390, Exwave HAD, Irvine, California) was used to image specimen fracture surfaces to characterize critical flaws. Fracture strength and critical flaw size measurements then were used to calculate apparent fracture toughness .
2
Materials and methods
Porcelain substrates were fabricated by cutting 18 × 2 × 2 mm bars from leucite-reinforced feldspathic porcelain blocks (ProCAD, Ivoclar-Vivadent, Schaan, Liechtenstein). All four surfaces of porcelain substrates were polished through 1200 grit (CAMI Grit Designation) SiC abrasive and edges rounded to limit corner/edge failures. Substrates were ultrasonically cleaned in acetone to remove surface debris prior to deposition, then randomly divided into four groups ( n = 30). A Vickers indenter (Mitutoyo Model AAV-500, Aurora, IL) was used to apply a single, controlled flaw into the center of a substrate surface with indent corners perpendicular to substrate edges, in order to normalize failures through control of critical flaw size and geometry. A 0.5 N load was applied with a load time of 15 s, resulting in an indent size of 34–37 μm to simulate critical flaw sizes in clinical surface preparation methods .
Thin film depositions were performed in an r.f. magnetron sputter reactor (CVC Model SC-400, Rochester, NY). A second r.f. power source was capacitively coupled to the substrate and used for depositions requiring a substrate bias. Target material used for sputtering was 99.99% pure zirconia doped with 3 mol% yttria (Plasmaterials, Livermore, CA). Substrates were mounted directly above the sputtering gun (US Gun, San Jose, CA) in a custom fixture at a working distance of 75 mm. At the start of each deposition, a 50 W substrate bias in an argon atmosphere was applied for 5 min to clean organic debris from specimen surfaces and to promote film adhesion. All depositions were performed with a target power of 350 W, atmosphere of 30:1 argon to oxygen, and substrate bias varied between 0 and 100 W. Films were grown to a thickness of 2–3 μm to maximize fracture strength .
X-ray diffraction (XRD) (PANalytical X’Pert PRO MRD HR, Westborough, MA) was used to determine the percentage of tetragonal and monoclinc phases in the YSZ films and thus qualitatively determine trends in film stress . Fracture strength values were determined using a three-point bending fixture, with specimens oriented with films in tension per ASTM Standard C1161 . Flexural testing (Instron Model 5542, Norwood, MA) was conducted with a 10 mm span at a crosshead speed of 0.5 mm/min. Bar dimensions were measured prior to fracture for fracture stress calculations. Peak loads and fracture stresses for each specimen were recorded and single-factor analysis of variance (ANOVA) at a 5% confidence level was performed to evaluate for statistical similarities. For testing controls, a polished sample set with a Vickers indentation was left unmodified by films in order to determine inert strengths of substrates at a given flaw size. Post fracture, scanning electron microscopy (SEM) (Hitachi, S-4700 FE, Tokyo, Japan) and optical microscopy (Sony DXC-390, Exwave HAD, Irvine, California) was used to image specimen fracture surfaces to characterize critical flaws. Fracture strength and critical flaw size measurements then were used to calculate apparent fracture toughness .
3
Results
%Monoclinic values were found to increase with substrate bias ( Fig. 1 A ), ranging from 50% (no substrate bias) to 84% (100 W substrate bias). Observable changes in film microstructure were associated with increasing volume fraction of %monoclinic. A representative SEM image of a non-biased film structure ( Fig. 1 B) was shown to have a columnar morphology; whereas, 100 W film ( Fig. 1 C) shows interrupted structure, characterized by the presence of inter-granular ledges.
In order to standardize critical flaw size, a Vickers indent was applied to each surface prior to thin film deposition and characterized. Fig. 2 A is a representative optical image of an indent that shows the damage zone (35 ±2 μm) and critical flaw size (67 ± 1 μm) characterized by the radial cracking from the indent corners. Fig. 2 B shows a critical flaw on the fracture surface of a fractured specimen. A damage zone can be seen in the area immediately under the indent (indicated by small arrows), from which the radial cracks identify the size of critical flaw (indicated by large arrows). Radial crack traces, perpendicular to the fracture surface (beyond the borders of the micrograph), can be observed (indicated by black arrow). Additionally, Fig. 2 C shows a post-fracture specimen indicating failure at the indent (critical flaw) and excellent adhesion between the thin film and porcelain.
