Dental Constructs and Methods for their Preparation

The present disclosure relates to new glass ceramic materials in particular for use in the preparation or manufacture dental constructs. The present disclosure relates also to dental constructs comprising such glass ceramic materials and to methods of making dental constructs using such glass ceramic materials. Dental constructs to which the present disclosure is applicable include dental restorations, fixed prostheses and the like. More especially, suitable dental constructs include crowns, onlays, inlays, bridges and the like. More especially, dental constructs to which the present disclosure is applicable are all-ceramic constructs, in particular those including an inner part (a

"core") of a first ceramic material and an outer aesthetic layer, in particular a veneering ceramic. The glass ceramic materials of the present disclosure are also suitable in particular as core materials for dental crowns. The present disclosure further relates to glass ceramic materials, in particular for dental constructs which can be processed (e.g. shaped) using CAD/CAM (computer aided design/computer assisted manufacture) procedures.

Introduction and Background

The oral environment presents challenges for materials used for dental constructs. For example, the materials used should preferably:

i. be sufficiently inert and non-toxic in the oral environment so that, for example, the materials have very limited solubility in mouth fluids and have adequate corrosion resistance. Ideally, the solubility must be below

2000μg/mm ² for core material for crowns and below 100μg/mm ² for veneer material. To achieve these parameters, the material should be able to withstand chemical attack in the oral environment, such as from contact with saliva, fluoride applications (toothpaste and mouthwashes) and low pH beverages; ii. be sufficiently strong to resist the forces of mastication; iii. be capable of being formed into shapes and forms compatible with human anatomy, preferably without the use of complex and expensive equipment; iv. have desirable aesthetic qualities, including the ability to colour match to a subject's natural teeth, and a degree of translucency similar to that of natural teeth;

v. be substantially non-moisture absorbing and substantially non-staining; vi. have wear characteristics similar to those of natural teeth.

Glass-ceramics are typically defined as polycrystalline solids prepared by the controlled crystallisation of glasses. Glasses are melted, fabricated to shape, and then converted by heat treatment to a partly (or predominantly) crystalline ceramic. Generally, crystallisation is achieved when the glass is subjected to a carefully regulated heat treatment regime which results in the nucleation and growth of a crystal phase within the glass matrix. A two-stage heat treatment process (often referred to as "ceramming"), comprising a nucleation and crystallisation step, is commonly employed for the preparation of glass-ceramics. The nucleation stage involves heating the glass from room temperature to the nucleation temperature and is used to create a large number of small crystals for efficient nucleation. Having nucleated the glass, it is usual to raise the temperature further in order to permit crystal growth upon the nuclei. Typically, the temperature of the glass is increased at a controlled rate, sufficiently slowly to allow crystal growth to occur without deformation of the glass. The length of time the material is held at the growth temperature depends on the degree of crystallinity that is required in the final glass ceramic material. The glass ceramic material is then cooled back down to room temperature. Generally, a high rate of nucleation and a low crystal growth rate are needed to achieve a fine grained ceramic since many more sites are provided for crystal growth. Proper control of the crystallisation heat treatment is necessary to ensure the nucleation of a sufficient number of crystals and their growth to an effective size.

The properties of glass ceramic materials depend on both their chemical composition and the microstructure. The bulk chemical composition controls the ability to form a glass and its degree of workability. The microstructure of is important in determining many of the mechanical and optical properties (such as fracture toughness and strength) and can promote or diminish the role of the key crystals in the glass-ceramic. Consequently, the microstructure is dependent on the use of a correct and optimal heat treatment schedule. In addition, the crystals that form within the matrix during heat treatment can vary significantly depending on the glass composition. The phase evolution in canasite-based compositions (including fluorcanasite) is complex and small modifications in composition radically change the crystallised product and fundamentally alter the mechanism of nucleation.

Chain silicates, or inosilicates, are polymeric crystals in which single or multiple chains of silica tetrahedra form the structural backbone. In the late 1970's, Beall demonstrated that glass-ceramics based on modified chain silicate compositions (e.g. enstatite, potassium fluorrichterite and canasite) can have high fracture toughness (3-5 MPa m ) and bending strength (200-300MPa) (see for example Beall GH: Chain silicate glass- ceramics. J Non-Cryst Solids, 1991 ; 129:163-173).

Although the properties of canasite are, at least in theory, advantageous, natural deposits of canasite are rare. Fluorcanasite (Ca ₅ K _2-3 Na _3-4 Si _I2 O ₃₀ F ₄ ) is a synthetic double chain silicate glass-ceramic which can be synthesised from glasses close to its stoichiometry. Fluorcanasite displays a combination of high flexural strength and fracture toughness in comparison with currently available resin-bonded glass-ceramic dental restorative systems. In addition, fluorcanasite, unlike many currently known high strength dental ceramics, has a surface that could be bonded with an adhesive composite resin luting agent. However although fluorcanasite is known to have good mechanical properties, a significant disadvantage hitherto has been its high chemical solubility and this has been a key limiting parameter for its use as a dental material.

Resin-bonded all-ceramic dental restorations require for their success in use that a durable and reliable bond is obtained, which must successfully integrate all parts of the restoration into one coherent structure. In prior art systems this bond has typically been created by: (1 ) micromechanical retention by hydrofluoric acid etching and/or grit blasting, and; (2) chemical bonding by a silane coupling agent. Etching the inner surface of a restoration with hydrofluoric acid followed by the application of a silane coupling agent is a well known and recommended method to increase the bond strength. However, Shimada et al in "Micro-shear bond strength of dual-cured resin cement to glass ceramics. Dent Mater, 2002; 18:380-388" reported that hydrofluoric acid etching glass-ceramic can in some instances adversely affect ceramic bonding and may thus not necessary for all clinical applications. The present inventors now suggest that glass-ceramics with a fine crystalline structure such as fluorcanasite may not benefit from hydrofluoric acid etching.

Reasons for seeking to bond components of a dental reconstruction without recourse to a hydrofluoric acid etching step during the procedure may include: (1 ) hydrofluoric acid is a highly toxic chemical, representing a potentially serious health hazard; (2) it has been reported that hydrofluoric acid etching of silica-based ceramics produces insoluble

silica-fluoride salts, which can remain as by-products on the surface (Shimada et al, ibid). If not removed, these by-products can interfere with the bond strength to the resin. However, although elimination of HF from the bonding procedure would be highly advantageous, this would only be possible in practice if the silane bond between the glass ceramic material and the adjacent element of the dental reconstruction can be shown to be adequate.

Various investigations have demonstrated that using adhesive composite resin cements can increase the fracture resistance of glass-ceramic restorations, provide high retention, improve marginal adaptation and prevent microleakage by penetrating surface flaws and irregularities and inhibiting crack propagation (Scherrer SS, de Rijk WG, Belser UC and Meyer JM: Effect of cement film thickness on the fracture resistance of a machinable glass-ceramic; Dent Mater, 1994; 10:172-177; Blatz MB, Sadan A and Kern M: Resin-ceramic bonding: a review of the literature; J Prosthet Dent, 2003; 89:268-274; Filho AM, Vieira LCC, Araujo E and Monteiro S Jr: Effect of different ceramic surface treatments on resin microtensile bond strength. J Prosthodont, 2004; 13:28-35; Burke FJT, Wilson NHF and Watts DC: Fracture resistance of teeth restored with indirect composite resins: the effect of alternative luting procedures, Quintessence Int, 1994; 25:269-275).

Fracture resistance of the ceramic-resin bond is believed to be controlled primarily by the microstructure and surface treatment of the ceramic. Therefore, it is highly desirable that an optimal bonding protocol for securing the ceramic and the resin is developed. Because fluorcanasite is a chain silicate glass-ceramic, the inventors suggest herein that it is possible to achieve a reliable bond using a silane coupling agent and resin cement and furthermore that because of the fine grain, acicular microstructure of fluorcanasite, it may be possible to eliminate the hydrofluoric acid etching stage from the cementation procedure.

As noted above the preparation process for a glass ceramic material may include an initial heating stage to promote crystal nucleation and a second heating stage to promote crystal growth. Typically the materials used for forming the glass ceramic are formed into the desired shape of the dental construct before such a heating procedure and it is therefore desirable that materials are dimensionally stable during such heating stages so that, for example, excessive shrinkage is avoided.

In particular where a dental construct is formed from more than one material (such as an inner core material and an outer aesthetic veneer) it is highly desirable that the coefficients of thermal expansion of the respective materials are matched, in order to avoid cracking or fracturing due to stresses induced by different rates of thermal expansion between the materials.

Background Art

US 4 386 162 describes glass ceramics wherein the predominant crystal phase is canasite having a composition, in weight percent on the oxide basis of:

EP 0 641 556 describes glass ceramic "biomaterial" which may be used as bone implants or partial replacements. The materials contains F-canasite and F-apatite crystals and is derived from a glass containing:

Glass ceramic materials based on lithium disilicate are also known for use in the fabrication of dental restorations. One example is described in US 6 342 458 which describes a glass ceramic comprising (in wt%):

wherein (a) AI ₂ O ₃ +La ₂ O ₃ amount to 0.1 to 7 wt% and (b) MgO+ ZnO amount to 0.1 to 9 wt%.

Brief Summary of the Disclosure

According to a first aspect of the present disclosure there is provided a glass ceramic material having a predominantly fluorcanasite structure comprising (and preferably consisting essentially of, more preferably consisting exclusively of), subject to any incidental impurities:

The glass ceramic of the disclosure has the general formula:

65SiO ₂ -6Na ₂ O-5.65K ₂ O-12.10.CaO-10CaF ₂ -0.81 ZrO ₂ .

In one preferred embodiment of this aspect of the disclosure, the glass ceramic material comprises essentially in mol%:

Preferred glass ceramic materials have a fracture toughness in the range of from 1 to 6 MPa m ^/2 , more preferably from 2 to 5MPa m ^/2 and especially from 3 to 4.5.MPa m ^/2 .

It is also preferred that the glass-ceramic materials have a solubility in the range of from 2000 to 100 μgcm ^"2 ., more preferably from 1500 to 100 μgcm ^"2 and especially from 1000 to 100 μgcm ^"2 . Solubilities herein are determined in accordance with ISO 6872:1999 Dental Ceramic Standard. This Standard involves subjecting the test ceramic material to 16 hours in 4% acetic acid solution at 80 ⁰ C.

Preferred glass ceramic materials have a total transmittance in the range of 80 to 30%, more preferably from 80 to 50% and especially from 80 to 65%.

According to a second aspect of the disclosure there is provided a dental construct comprising a glass ceramic material of the first aspect of the disclosure.

In particularly preferred embodiments the dental construct is a shaped dental construct such as a dental crown, or a part of a dental crown. In preferred embodiments of this aspect of the disclosure the dental construct is a dental restoration for the posterior region of oral the cavity, preferably a posterior crown. The dental construct may be the core of a dental crown. Dental constructs comprising the glass ceramic material according to the disclosure are especially suitable for securing to an underlying structure (such as a prepared residual portion of a tooth) by resin bonding in particular using a commercial composite resin luting system.

According to a third aspect of the present disclosure there is provided an all ceramic dental construct comprising a core of a glass ceramic material as defined in the first aspect of the invention and a veneering ceramic. Preferably the dental construct comprises a dental restoration for the posterior region of oral the cavity, preferably a posterior crown.

In preferred embodiments the dental construct comprises a silane coupling agent applied to the glass ceramic core and a resin luting agent bonding the core to the veneer.

Preferably the glass ceramic of the core and the veneering ceramic have a substantially matched coefficients of thermal expansion (CTEs). More especially the CTE values of the glass ceramic of the core and the veneering ceramic differ by not more than 2ppm/°C, preferably not more than 1 ppm/°C.

According to a fourth aspect of the disclosure there is provided a method of preparing a dental construct comprising the steps of: providing a glass composition comprising:

ceramming the composition and machining to form the dental construct. It is a particular advantage of the method that only a single machining stage is required, in contrast to some prior art procedures which require machining both before and after the ceramming stage.

In a preferred embodiment of the method the dental construct comprises an all-ceramic dental restoration, and the method further comprises applying a veneering ceramic to the glass ceramic body.

In particularly preferred embodiments the veneering ceramic is applied to the glass ceramic body without any intervening physical surface modification step of the glass ceramic body. More especially the veneering ceramic is applied to the glass ceramic body without any intervening grit blasting step and/or without any intervening chemical etching step, such as a hydrofluoric acid etching step

Preferably the ceramming step includes an initial nucleation stage at a first temperature and a subsequent crystallisation stage at a higher, second, temperature.

Preferably the first temperature is in the range of from about 500 to 650 ⁰ C.

Preferably the second temperature is in the range of from about 800 to 900 ⁰ C.

The inventors suggest that the individual components of the glass ceramic material according to the present disclosure have the following properties and functions:

SiO ₂ acts as a network former. Network formers form glasses when melted and cooled because of their ability to build up continuous three dimensional random networks.

Na ₂ O, K ₂ O, CaO act as network modifiers. Network modifiers are incapable of building up a continuous network and function to partly disrupt the network structure through the introduction of ionic bonds. The metal ions involved tend to form non-directional ionic bonds with oxygen atoms, resulting in the formation of non-bridging oxygens in the structure. The effect of network modifiers is generally to reduce the viscosity of the glass and to increase the coefficient of thermal expansion

CaF ₂ acts as a nucleation agent for the crystallisation of fluorcanasite and plays an important role in controlling types of crystalline phase formed and microstructure of the fluorcanasite. Generally, additional of excess CaF ₂ to the glass ceramic composition promotes improved nucleation and a finer grain size. If the fluorine content is too low nuclei formation and crystal formation may be inhibited. On the other hand, if the fluorine content is too high precipitation of CaF ₂ may occur.

ZrC> ₂ (zirconia) addition is believed to aid CaF ₂ in the crystallisation of fluorcanasite, to produce a glass-ceramic with canasite (rather than frankamenite) as the dominant phase.

Brief Description of the Drawings

Figure 1 shows the mean surface roughness (Ra) measurements of the fluorcanasite and lithium disilicate glass-ceramics following different surface treatments (Example 2);

Figure 2 shows SEM images of fluorcanasite and lithium disilicate demonstrating different surface finishes (Example 2);

Figure 3 shows mean microtensile bond strength and standard deviation for ceramic- composite bond with different surface treatments (Example 2);

Figure 4 is an example of typical SEM micrograph depicting cohesive failure in the luting resin; and

Figure 5 shows an SEM image of adhesive failure at the interface between the luting resin and ceramic.

Example 1

A fluorcanasite based glass ceramic material having the formula shown in the Table 1 was formed by the following method

Batches of the glasses were hand mixed from standard glass-making ingredients were mixed in amounts required for forming a glass ceramic having the composition indicated in Table 1 below. The glasses were melted in a platinum crucible at 1350 ⁰ C for 1 hour static and 1 hour stirred, poured into a pre-heated steel mould and subsequently annealed at 460 ⁰ C for 1 hour. The glass was core drilled into 12mm cylinders before ceramming. A two-stage heat treatment was employed comprising of a 2 hour nucleation hold at 550 ⁰ C and a 2 hour crystallisation hold at 840 ⁰ C. The resulting glass- ceramic blocks were then machined to form dental cores using CAD/CAM technology. No cracking of the ceramic occurred during this milling process and, on subsequent examination, no cracking, defects or marginal chipping was detectable. A veneering

ceramic (50%feldspar/50% high leucite) with matched coefficient of thermal expansion (CTE) was then applied to the core structure and fired in a vacuum furnace at 940 ⁰ C.

Table 1

The resulting glass ceramic material had the properties indicated in Table 2:

Table 2

Comparative Examples

Materials having the composition shown in Table 3 were prepared according to the method of Example 1. Properties of these materials are shown in Table 4.

Table 3

Table 4

Comparative Example 3

Table 5

Comparative Examples 4 and 5

Materials having the composition shown in Table 6 were prepared according to the method of Example 1. Properties of these materials are discussed below.

Table 6:

Comparative Example 4

Differential thermal analysis and X-ray diffraction showed crystallisation to frankamenite and canasite. Chemical solubility was higher than the composition of Example 1 at 2874 ± 608μg/cm ² , which falls outside the range for a dental ceramic core material. Therefore no mechanical property testing was undertaken as the solubility was too high.

Comparative Example 5

This material would not melt to form a clear glass - devitrification occurred with calcium fluoride and hence a cloudy glass was formed.

Example 2 - Evaluation of the effect of surface treatments on the bond strength of fluorcanasite and lithium disilicate glass-ceramics

Summary

Fifteen blocks of an experimental fluorcanasite and a lithium disilicate glass-ceramic (IPS e.max CAD ^® ) were assigned to one of the following three surface treatments:

(1 ) machined with 60μ finish,

(2) machined and grit blasted,

(3) machined and HF etched.

The ceramic blocks were duplicated in composite resin (Spectrum") and cemented together with a resin luting agent (Variolink II ^® ). Thirty microbars per group (1.0 x 1.0 x

20mm) were obtained and subjected to a tensile force at a crosshead speed of 0.5 mm/min using a universal testing machine until failure. The mode of failure was determined using scanning electron microscopy. Each bonding procedure was assessed for durability by storing in water at 100 ⁰ C for 24 hrs. Statistical analyses were performed with ANOVA and Tukey's test (P<0.05). Procedural details are set out below in "Detailed Procedure" and a fuller discussion of the results follows thereafter.

Machining alone significantly increased the bond strength (MPa) of the fluorcanasite (27.79 ± 6.94) compared to the lithium disilicate (13.57 ± 4.52) (P<0.05). HF etching resulted in the lowest bond strength (8.79 ± 2.06) for the fluorcanasite but the highest for the lithium disilicate (24.76 ± 9.38). In the durability tests, the machined fluorcanasite (15.24 ±5.46) demonstrated significantly higher bond strength than the machined and HF etched lithium disilicate (12.28 ± 3.30).

Thus in accordance with preferred embodiments of the present disclosure when preparing a dental construct, the fluorcanasite glass ceramic should be left as machined, to the exclusion of any physical or chemical treatment which materially affects the physical nature of the glass ceramic surface, more especially to the exclusion of any grit blasting or acid etching treatment. The machined surface is, however, preferably directly treated with a silane coupling agent.

Detailed Procedure

Materials and Methods

Ceramic materials

Two CAD/CAM machinable glass-ceramic core materials were employed in this example; an experimental fluorcanasite glass-ceramic (University of Sheffield) and a commercial lithium disilicate glass-ceramic (e.max CAD, batch number JO8179, Ivoclar Vivadent AG, Schaan, Liechtenstein).

Surface Roughness Profilometry

Four different surface treatments were performed on disc specimens of the fluorcanasite and lithium disilicate glass-ceramic

(i) Polishing to 1 μm finish with 400 to 1200 grit wet silicon carbide paper, then 3 and 1 μm diamond polishing paste using a polishing machine (Buehler Metaserv, UK) (ii) Machined finish using a 60μm diamond bur (Henry Schein, Germany) (iii) Machining and grit blasting with 50μm aluminium oxide particles (MicroEtcher, Danville Engineering, San Ramon, CA)

(iv) Machining and etching with hydrofluoric acid (HF) (Ultradent Porcelain Etch 9.5% Buffered, Ultradent Products, South Jordan, UT) for 1 min, then rinsing and air drying for 1 min.

A surface roughness profile was determined for each of the groups using a profilometer (Mitutoyo Surftest 301 , Mitutoyo America Corp, Aurora, III). A diamond stylus (5μm radius) was used under a constant measuring force of 3.9N. The instrument was calibrated using a standard reference specimen, and then set to travel at a speed of 0.1 mm/s with a range of 600 μm during testing. The roughness of the specimen was analysed by performing two passes of the profilometer, with one pass at a 90 degree angle to the other. Ten recordings per specimen (n=3) in each surface treatment group were obtained.

SEM analysis was performed to ascertain the effects of the different surface treatments on the microstructure of the core materials. The specimens were gold coated with a sputter coater (Evaporation unit, Edwards, UK), mounted on coded brass stubs and examined using scanning electron microscopy (Philips XL-20).

The 1 μm finish produced the smoothest surface profile for both the fluorcanasite and the lithium disilicate glass-ceramic. Machining in conjunction with grit blasting created the roughest surface for the fluorcanasite glass-ceramic whereas with the lithium disilicate glass-ceramic, this was attained by machining alone. The mean surface roughness values and standard deviation are presented in Table 7 and Figure 1. A significant difference was noted between the surface finish of machining and grit blasting between the fluorcanasite and lithium disilicate glass-ceramics. HF etching reduced the surface roughness in comparison to the grit blasted surfaces with both glass-ceramics.

Table 7: Mean surface roughness (Ra) and SD for the different surface treatments.

For both the fluorcanasite and the lithium disilicate glass-ceramics, the 1 μm finish consisted of an extremely low frequency with virtually no amplitude defects. With the machined surface, there was an increased frequency of irregular amplitude defects varying from +13 to -12μm for the lithium disilicate glass-ceramic in comparison to a range of +7 to -7μm for the fluorcanasite glass-ceramic. With the grit blasted surface, a much higher frequency of irregular amplitude defects was found with the fluorcanasite glass-ceramic, with a range of +22 to -22μm. On the other hand, the lithium disilicate glass-ceramic showed a much lower frequency of amplitude defects with a range of +4 to -6μm. The HF etched surface showed similar results for both ceramics. The irregular amplitude defects had a range of +9 to -10μm for the fluorcanasite and +5 to -8μm for the lithium disilicate glass-ceramic.

The two glass-ceramics have different microstructures and varying appearances were observed with the surface treatments (Figure 2). For both materials, the 1 μm finish was smooth and polished with minimal surface defects. Machining of the fluorcanasite glass-ceramic opened up the structure and blade-like crystals were easily seen, creating an irregular surface. The lithium disilicate glass-ceramic had a rough appearance after machining but without prominent crystals. The grit blasting treatment had destroyed the crystal structure of the fluorcanasite glass-ceramic and areas were gouged out resulting in crevices and a crater-like appearance. The lithium disilicate glass-ceramic showed an irregular rippled appearance caused by the grit blasting but with no evidence of crevices. HF etching also destroyed the crystal structure of the fluorcanasite glass-ceramic, leaving voids and the appearance of a fragile surface. Surface contamination was also observed. With the lithium disilicate glass-ceramic, large surface irregularities and crevices were clearly visible and the surface also had a rough textured appearance. There was no obvious surface contamination, in comparison to that on the fluorcanasite glass-ceramic.

Microtensile bond strength testing

When a resin-bonded ceramic restoration is placed many factors can play a part in the resin bond and it is imperative that a stable and durable bond is created. Various methods are available for assessment of the bond strength. One of the most common testing methods is the shear bond test. However, this test frequently produces cohesive bulk fracture within the substrate rather than the interface due to the generation of complex stress distributions during testing and may lead to erroneous interpretation of the data. Cohesive failures are rarely seen clinically with bonded restorations. Hence an alternative procedure intended to use purely tensile forces was adopted.

Fifteen 1x1x1 cm blocks were prepared from each of the fluorcanasite and the lithium disilicate glass-ceramics. The specimens were polished with 400-grit through to 1200- grit wet silicon carbide paper using a polishing machine (Buehler Metaserv, UK). Following this, the ceramic blocks were ultrasonically cleaned (Biosonic UC300, Whaledent, Altstatten, Switzerland) in distilled water for 5 min to remove any contamination from the silicon carbide papers. Each ceramic block was duplicated in composite resin (Spectrum TPH, batch no 0506003114, Dentsply DeT rey GmbH,

Konstanz, Germany) with the same dimensions using a mould made of a polysiloxone silicone impression material (Aquasil, Dentsply DeTrey GmbH, Konstanz, Germany). Incremental layers (2mm) of composite resin were condensed into the mould under a standardised load of 4ON, polymerised for 20 seconds (LCU, Bayer Dental, Leverkusen) and repeated until the mould was full. One composite resin block was constructed for each ceramic block.

The 15 blocks of each ceramic were randomly assigned to three groups which received the following surface treatments: (i) Machined finish using a 60μm diamond bur (Henry Schein, Germany). This surface finish was used to simulate the machining process of the CAD/CAM technology.

(ii) Machined finish and grit blasting with 50μm aluminium oxide particles (MicroEtcher, Danville Engineering, San Ramon, CA).

(iii) Machined finish and HF acid (Ultradent Porcelain Etch 9.5% Buffered, Ultradent Products, South Jordan, UT) applied for 1 min, rinsed for 1 min and air dried for 1 min.

A silane coupling agent (Monobond-S, batch no J14325, Ivoclar Vivadent AG, Schaan, Liechtenstein) was then applied to each group with a brush, left undisturbed for 1 min and then dried with an air stream. The ceramic and composite resin blocks were then joined as pairs using a composite resin luting system (Variolink II, batch no J17818, Ivoclar Vivadent AG, Schaan, Liechtenstein) according to the manufacturer's instructions and light polymerised with a standard halogen light (LCU, Bayer Dental, Leverkusen) for four 40 sec periods at right angles to each other. The specimens were stored in distilled water for 24 hr.

Using a water-cooled diamond blade with a slow speed cutting saw (Isomet, Buehler, Lake Bluff, IL, USA), each block was longitudinally cut into a series of 1 mm thick slabs. The sectioning continued until 1 mm remained to keep the specimen in a fixed position. The ceramic-composite block was then rotated 90° and the procedure repeated. Twelve to fourteen bars approximately 1.0 mm ² in cross-section ("microbars") were obtained from each block. The peripheral slices were discarded in case the results could be influenced by either excess or insufficient amount of resin cement at the interface. Specimens were obtained directly from the cutting machine, that is, in a non- trimmed state. Neither polishing nor finishing was performed.

Each microbar was glued with a cyanoacrylate adhesive (Zapit, CA, USA) to a jig designed to transmit purely tensile forces when mounted on a universal loading machine (Lloyds LRX tensometer, Lloyds Instruments, UK). Bending forces were avoided by gluing specimens in the most parallel position possible and in contact with the jig. The tensile load (100N) was applied at a crosshead speed of 0.5 mm/min until failure. The load at failure in Newtons was recorded, and the fragments of the specimen were carefully removed from the fixture with a scalpel blade. The cross- sectional area at the site of fracture was measured to the nearest 0.01 mm with a digital calliper (Mitutoyo, Tokyo, Japan) in order to calculate the bond strength at failure in MPa.

The microtensile bond strength values were calculated using the formula σ = L/A where L is the load at failure (Newtons) and A is the adhesive area (mm ² ) and

expressed in MPa. Thirty microbars from each group were tested for microtensile bond strength.

The fractured surfaces were examined by optical microscopy (Bausch and Lomb, USA) and scanning electron microscopy (Philips XL-20) to determine the type of failure, which was classified as adhesive, cohesive or mixed within any of the substrates or interfaces. The fractured surfaces were gold coated (Evaporation unit, Edwards, UK) prior to examination.

The mean bond strength values and standard deviations for the microtensile bond strength testing are presented in Table 8 and illustrated in Figures 3. No pre-test failures occurred.

Table 8: Mean microtensile bond strength data for ceramic-composite bond with different surface treatments. Groups with different superscript letters indicate significant differences (P<0.05).

The highest microtensile bond strength of 27.59 MPa was achieved with the machined surface finish of the fluorcanasite glass-ceramic in comparison to only 13.57 MPa with the machined finish of the lithium disilicate glass-ceramic (P<0.05). On the other hand, HF etching produced the highest microtensile bond strength for the lithium disilicate glass-ceramic (24.76MPa) but a very poor result for the fluorcanasite glass-ceramic (8.79MPa) (P<0.05). There was no significant difference between the machined finish fluorcanasite glass-ceramic and the lithium disilicate glass-ceramic with HF etched finish. Grit blasting afforded only mediocre bond strengths of 19.45 and 18.57MPa for the fluorcanasite and lithium disilicate glass-ceramics respectively, which were not statistically significantly different.

SEM observations of the fractured surfaces of the bonded specimens revealed that the mode of failures was either cohesive fracture in the luting resin or adhesive fracture at the interface between the ceramic and luting resin. There was no cohesive failure in the composite or ceramic.

The machined finish fluorcanasite and the HF etched lithium disilicate glass-ceramics exhibited predominately cohesive failures whereas the HF acid etched fluorcanasite and the machined finish lithium disilicate glass-ceramics exhibited predominately adhesive failures. With grit blasting, a mixture of cohesive and adhesive failures occurred with both the fluorcanasite and lithium disilicate glass-ceramics. In general, as the microtensile bond strength of the glass-ceramics increased, there was a decrease in number of adhesive failures at the interface between the ceramic and luting resin. Figure 4 illustrates a representative SEM image of a specimen showing cohesive failure in the luting resin and Figure 5 shows adhesive failure at the interface between the luting resin and ceramic.

Testing of durability of the bond

The surface treatment methods which afforded the highest microtensile bond strengths were also assessed for bond durability. Thirty microbars of the fluorcanasite and lithium disilicate glass-ceramic, with surface treatments of machined and machined plus HF etching respectively, were fabricated as described before. The specimens were then stored in distilled water at 100 ⁰ C (boiling water) for 24 hours using an extraction apparatus. After a 30 min drying period, the microbars were subjected to the same microtensile bond strength test.

The data were analysed using ANOVA with Tukey's multiple comparison tests. The software programme used was SPSS for Windows, version 14.0, SPSS Inc, Chicago, III. The results were considered significant for P < 0.05.

The mean bond strength values, standard deviations and for the microtensile bond strength testing following 24 hours in boiling water are presented in Table 9. The fluorcanasite glass-ceramic achieved a significantly higher mean microtensile bond strength value of 15.23 MPa in comparison to 12.29 MPa for the lithium disilicate glass- ceramic (P<0.05). With the fluorcanasite glass-ceramic, the mode of failure was cohesive in the luting resin whereas with the lithium disilicate glass-ceramic, 75% of the failures were cohesive in the luting resin and 25% were classified as adhesive failure at the interface between the ceramic and the luting resin (Table 9).

Table 9: Mean microtensile bond strength data for ceramic-composite bond following 24 hours in boiling water. Groups with different superscript letters indicate significant differences (P<0.05).

Example 3 - Thermal Shocking

Discs (12mm x 1 mm) and standardised crowns (n=10) were fabricated from the fluorcanasite composition as shown in Table 1 and, by way of comparison from lithium disilicate glass-ceramic (Ivoclar, Vivadent AG, Schaan, Liechtenstein). The fluorcanasite discs were core drilled and sectioned using a rotary diamond cutting machine (LECO VC-50, LECO Corporation, Michigan, USA) with a diamond wafering blade (Buehler, Illinois, USA) from the glass, cerammed at 550 ⁰ C for 2 hrs and 840 ⁰ C for 2 hrs and subsequently veneered with the a veneering ceramic having a suitably matched CTE. The lithium disilicate discs were similarly core drilled, sectioned and

heated treated according to the manufacturer's instructions as set out in section Table 10 below. These discs were then veneered with the fluoroapatite ceramic at 750 ⁰ C. Digital callipers and a silicone mould were used to standardise these bilayered discs and ensure the thickness of the veneering ceramic was 0.7mm.

Table 10

For the crowns, a maxillary upper first molar tooth (Frasaco, Tettnang, Germany) was prepared with 1.0mm axial, 1.5mm occlusal reduction and a 90° shoulder preparation with a rounded internal line angle. An impression of this preparation was taken using polysiloxone impression material (Aquasil, Dentsply DeTrey GmbH, Konstanz, Germany) and then the impression was cast using a CAD/CAM compatible refractory die material (CamBase, Sirona Dental Systems GmbH, Bensheim, Germany). The resulting model was scanned using the inEos and Cerec inLab system (Sirona Dental Systems GmbH, Bensheim, Germany) and a coping was designed using the software programme (inLab V2.90). All of the copings were milled from the same design and subsequently veneered with the appropriate matched veneering ceramic. The thickness of the veneering ceramic was standardised by the use of digital callipers and silicone moulds. The cooling protocols used following firing were also standardised.

Ten specimens of each group were placed inside an oven (Vecstar ECF2, Chesterfield, UK), which had been preheated to 90 ⁰ C. After a 30 min hold to allow the samples to reach temperature equilibrium, they were removed from the oven and quenched in ice cold water. The samples were then dried, returned to the oven, then reheated to 90 ⁰ C for 30 mins and subsequently cooled to room temperature to allow for inspection. The specimens were inspected for crazing using light microscopy (Wild M3Z, Heerbrugg, Switzerland) at 4Ox magnification with fibre optic transillumination (Intralux 4000, Switzerland). If crazing was observed, this constituted a failure at δT = 90 ⁰ C. If no failure was observed, the specimens were tested again at increasing temperature increments of 10 ⁰ C until failure.

Statistical analysis was undertaken using one-way ANOVA with Tukey's multiple comparison tests (SPSS for Windows, version 14.0, SPSS Inc, Chicago, III). The results were considered significant for P<0.05.

For each ceramic system, the mean δT value, standard deviation and δT range are shown in Table 1 1. The δT value represents the temperature difference required to produce a failure in the ceramic.

Table 1 1 : δT values that resulted in failure.

The fluorcanasite glass-ceramic system, in both disc and crown form, greatly outperformed the lithium disilicate glass-ceramic system (P<0.05). The fluorcanasite glass-ceramic discs had δT values ranging from 230-270 ⁰ C while the lithium disilicate glass-ceramic ranged from 190-240 ⁰ C. With the crown system, the difference was even more noticeable, with a δT values of 370-450 ⁰ C for the fluorcanasite glass-ceramic but only 190-230 ⁰ C for the lithium disilicate glass-ceramic system. The δT values of fluorcanasite glass-ceramic crowns were significantly higher than the other 3 groups (P<0.05). There was no statistically significance difference between the two lithium disilicate glass-ceramic groups.

Thus it can be seen that the glass ceramic materials of the present disclosure has excellent properties of strength, colour, insolubility and low thermal expansion and, more especially, translucency which make the materials especially suitable for use as dental constructs such as dental crowns and the like. The comparative examples show

that even small compositional differences can result in significant loss of desired properties.

Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of the words, for example "comprising" and "comprises", means "including but not limited to", and is not intended to (and does not) exclude other moieties, additives, components, integers or steps.

Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the disclosure are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith.