| Sign In to gain access to subscriptions and/or personal tools. |
Comparative Microstructural Study of the Diffusion Zone between NiCr Alloy and Different Dental Ceramics s1,*
1 University of Debrecen, Medical and Health Science Center, Institute of Dental Science, Department of Prosthetic Dentistry, Debrecen, Hungary Nagyerdei krt. 98, H-4012, Hungary; and Correspondence: *corresponding author, hegedus{at}fogaszat.dote.hu
Our knowledge on the bonding mechanisms between the metal and ceramic parts of dental systems is very limited. This work tested the hypothesis that the details of the interface processes can be described in the framework of a chemical diffusion model. The development of interfacial phases was investigated by cross-sectional analytical transmission electron microscopy between a NiCr (Wiron 99) alloy and three different dental ceramics (Carat, Vita VMK 95, and Vision). All systems were investigated at normal firing conditions (suggested by the manufacturer) and at increased firing times as well. The conclusions are based on the results that the formation of a nanocrystalline Cr2O3 layer and amorphous silicon oxide inclusions were detected in the early stage of the firing process in all investigated systems, and that, in the case of Carat and Vision ceramics, formation of complex NiCr and NiCrTi oxides was also observed at longer annealing times.
Key Words: transmission electron microscopy • nickel-chromium alloy • dental ceramics
A detailed knowledge of the bonding mechanisms between metal and ceramic parts in conventional dental systems can assist in the development of new, improved systems and the optimization of recent technologies. The bond strength between metal and ceramic is determined by the properties of the different phases emerging in the diffusion zone during the firing process. The formation and growth of these phases indicate that very complex reduction-oxidation reactions take place in the interface region. These oxide layers play a fundamental role in adherence, i.e., in solid-state diffusion bonding (Klomp, 1987). Time and temperature of firing strongly affect the quality of bonding relative to the different interlayer structures. In these complex systems, diffusion bonding is established not only by mutual interdiffusion but also by simultaneous solid-state reactions. Comparison of the results of microstructural investigations of the diffusion zone and the macroscopic bond-strength tests can clarify the role of various phases in the adherence process. Previously, most of the microstructural studies in the literature were based on optical microscopy (OM), scanning electron microscopy (SEM), and electron probe microanalysis (EPMA). Several studies have been reported on oxidation and ceramic coatings of NiCr alloys (Anusavice et al., 1977a, b; Williams et al., 1978; Ringle et al., 1979; Pask and Tomsia, 1988; Watanabe et al., 1989; Mizutani, 1990; Inoue et al., 1992). The limited space resolution of the abovementioned methods has made it impossibile for an effective investigation of the diffusion zone in the submicron range to be conducted. The use of a cross-sectional transmission electron microscopy (CTEM) technique has made possible, on a finer scale, the study of the reaction layer between metal and glass. The aim of the present work is to carry out a detailed microstructural investigation and to compare the reaction layer developed among three different brands of dental ceramics and NiCr alloy under different firing conditions.
For the measurements, 6 identical samples were cast from WIRON 99 (Bego, Konstantz, Germany) alloy with dimensions of 10 x 10 x 2 mm. (The nominal composition of WIRON 99 in wt% was as follows: 65 Ni, 22.5 Cr, 9.5 Mo, 1 Nb, 1 Si, 0.5 Fe, 0.5 Ce, < 0.02 C.) One side of each sample was ground and polished with 0.3 micron Al2O3 suspension used in the last step. On the polished surface, ceramic opaquers were fired by means of Vision (Wohlwend AG, Eschen, Liechtenstein), Vita VMK 95 (Vita Zahnfabrik, Bad Säckingen, Germany), and Carat (Dentsply DeTrey GmbH, Dreieich, Germany) materials. Regarding the main components, the compositions of the different kinds of ceramics were similar (e.g., VITA VMK, in wt%, 52.4 SiO2, 15.15 Al2O3, 9.9 K2O, 6.58 Na2O, 2.59 TiO2, 5.16 ZrO2, 4.9 SnO2, 0.08 Rb2O, 3.24 B2O3, CO2, and H2O). Two samples were produced from each type of opaquer under different firing conditions. For the first sample, we closly followed the manufacturer's recommendations, while in the case of the second sample, we increased firing time to 10 min (instead of the recommended 1 min [Vision, Vita] and 2 min [Carat]) for opaquer layers 1 and 2, respectively. For CTEM investigations, 1.5 x 1.5 mm plates were cut from the samples. These plates were embedded into an aluminum disc (3 mm in diameter) so that the metal-ceramic interface was perpendicular to the plane of the disc. The discs were ground and polished from both sides and dimpled from one side. The minimum thickness of the sample was about 0.02 mm after dimpling. The final thinning of the sample was performed by low-angle Ar-ion milling with 10 kV ion energy. In the final stage of the milling process, ion energy was reduced to 5 kV, decreasing the possibility that preparation artifacts would be formed. The TEM investigations were performed with the use of a JEOL 2000FX-II microscope (JEOL Ltd., Tokyo, Japan) with an Oxford Link-Isis EDS system (Oxford Instruments Ltd., United Kingdom).
The bright-field image of the NiCr-Carat system after the firing process suggested by the company is shown in Fig. 1  . Two different phases are visible in the diffusion zone. The bright bubble-shaped inclusions (highly electron-transparent phase), marked by ‘A’ in the Fig., grew into the original flat metal surface. The corresponding insert shows the EDS spectra of this area. (In this and in the following EDS spectra, the observable copper signals are scattered from the sample holder. Because of the beryllium window of the detector, the lines of the light elements [Z < 11] are absent from the spectra.) According to the electron diffraction results, this phase has an amorphous structure. Near the directly observed silicon, oxides of light elements such as Li and B may also be present in the inclusions. The amorphous structure and the presence of silicon as the main component indicate that this phase is probably a glassy silicate structure with multiple oxides.
The continuous polycrystalline layer with an average grain size of about 50 nm (denoted by ‘B’ in Fig. 1  ) follows the line of the original metal surface. As revealed by the EDS measurements (insert B in Fig. 1), this phase has a very high chromium content. The electron diffraction patterns are in accordance with the structure of the rhombohedral Cr2O3. A selected area diffraction pattern is shown in the insert in Fig. 1.
The morphology, structure, and composition of the observed phases in the NiCr-Vita system after the firing process recommended by the manufacturer are quite similar to those presented above for the NiCr-Carat system. On the other hand, in the case of the NiCr-Vision system, a different structure is visible (Fig. 2
The TEM picture of the NiCr-Carat system after increased firing time is presented in Fig. 3 . Near the well-known glassy inclusions (A) and polycrystalline Cr2O3 layer (B), a new crystalline phase (C) developed. According to the EDS results, it is a complex phase composed of Ti, Cr, and Ni oxides.
In the NiCr-Vita system, only two phases remain, even after elevated firing. The thickness of the Cr2O3 layer was also considerably increased, from 100 to 500 nm. The phase structure of the NiCr-Vision system did not change after the elevated firing process. The extent of the glassy inclusions increased considerably. The potassium content of the crystalline phases was also slightly higher than after the shorter firing process.
Similar growth processes of the phases were observed in all of the investigated systems, in the sense that an amorphous phase was formed at the metal-ceramic interface, producing amorphous inclusions at the metal surface. Silicon oxide was the main component of this phase, as in the case of Carat and Vita ceramics, while in the case of Vision, potassium and nickel oxides were observed after the normal firing process. The extent and composition of the inclusions did not change significantly after increased firing. Furthermore, in all systems and at all firing conditions, a nanocrystalline Cr2O3 layer (with about 50-nm grain size) was commonly observed. In the case of the NiCr-Vision system, a third phase, a cubic Ni-Cr oxide with a lattice parameter of about 0.82 nm, appeared after short firing times. A similar third phase was formed in the NiCr-Carat system at elevated firing, but its structure and composition were different (Ti, Cr, and Ni were observed as main components in this oxide phase). To summarize the common and diverging features of the above processes: It is plausible to assume that, at the start of the firing process, a very thin Cr2O3 layer already exists on the surfaces of the NiCr samples (as is usual in the case of high-chromium alloys). The high surface-segregation ability and high chemical activity of chromium can justify the above supposition. In addition, it was observed (Mizutani, 1990) that the composition of the oxide formed after atmospheric heat treatments was sensitive to the original Cr content. Thus, with about 30-40 wt% Cr, it was mainly the Cr2O3 phase that formed, while at a lower Cr content, not only this phase but also NiO and NiCr2O were observed. In our case, the lower oxygen pressure in the vacuum furnace can explain why only Cr2O3 was formed.
The above nanocrystalline chromium-oxide layer produces a diffusion barrier between the metal and the glass. Only the grain and phase boundaries form fast diffusion paths for the diffusing atoms. At higher temperatures, there is an intermixing by diffusion of the metal atoms of the CrNi phase and the species of the ceramics (Fig. 4a
 is Darken's thermodynamic factor,
 is the activity coefficient and x is the molar fraction of the given species.
and the diffusion flux J (J = -D dc/dx, where dc/dx is the concentration gradient of a given species) can be very high in the case of Cr because of its high chemical activity. In the early stage of the firing process from the metal side, the chromium current is the dominant factor. On the other side, the SiO2 molecules are the most mobile components (their molar weight is the smallest compared with that of other oxides in the ceramics). Obviously, molecules of Li or B oxides can be similarly fast diffusers, but these light elements cannot be detected by our EDS system. Thus, during intermixing, the Cr can reduce (at least partially) the SiO2 (Conforto and Schmid, 2001). The decrease in the dimensions of the SiO2 molecule, due to the partial reduction and the presence of free oxygen in the glass melt, and owing to the introduction of additives such as TiO2 and ZrO2 (Iwamoto et al., 1987), can further enhance the diffusion current of SiO2 and Cr. As a result, silicon oxide bubbles appear (with an approximate composition of SiOx, where 1 < x < 2) between the metal and the Cr2O3 layer (see Fig. 1), and the chromium oxide increases/grows further. Similar considerations can be valid for other oxides present in the ceramics. Although there is almost 1 wt% Si present in the initial NiCr-based alloy, such a small amount of Si alone cannot be responsible for the formation of the bubble-like inclusions (see, e.g., Fig. 1, where they are almost continuously present along the interface). In the later stage of firing, the concentration gradient and consequent diffusion flux of the Cr decrease, which results in the relative enhancement of the Ni flux. This makes the formation of complex oxide phases with relatively high Ni content possible (see Figs. 3 and 4b). It is not clear why potassium instead of silicon was observed in the amorphous inclusions in the case of the NiCr-Vision system. Supposedly, the details of the above-described processes can be strongly affected by the additives of the glass and the exact circumstances of the firing process (e.g., residual atmosphere, water vapor pressure in the furnace).
This investigation was supported in part by DOTE Mecenatura Research Grant 19/99. Received for publication May 14, 2001. Revision received February 27, 2002. Accepted for publication March 18, 2002.
Journal of Dental Research, Vol. 81, No. 5,
334-337 (2002) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
