Diffusion path studies relevant for SiC joining

Joining of dissimilar materials such as ceramics to metals is regarded as an enabeling technology for many high temperature applications. For SiC ceramic to metal joints NiCr-braze alloys (Ti-doped) are the most promising candidates to raise the service temperature above the ~400C currently possible by using the CuAgTi-braze technology. Most joints are manufactured and put to service under conditions of "steady state" rather than equilibrium. This steady state is described by the diffusion path. Hence knowledge of the diffusion path between the ceramic and various alloys is a key step to control the interface and thus quality of a joint. The proposed project aims to identify experimentally the diffusion path between SiC and binary as well as selected ternary alloys from the system Cr-Ni-Ti making use of phase diagrams for the quaternary systems Si-Cr-Ni-C, Si-Cr-Ti-C, and Si-Ni-Ti-C. From these diffusion experiments also interdiffusion coefficients and the activation energy for diffusion shall be derived. It will be attempted to model the evolution of the interfacial reaction product scale with regard to sequence of phase appearance and growth rate. This should allow to choose the joining process parameters (alloy composition, joining time and temperature) to obtain prespecified reaction zone structures.

The project "Diffusion Path Studies and Modeling Relevant for Silicon Carbide Joining" provided important new insight into the process of joining in general and of joining silicon carbide (SiC) ceramics with Cr-Ni-alloys in particular. Joining of dissimilar materials like ceramics and metals is regarded as enabeling technology for many high temperature applications. Cr-Ni-alloys are the most promising candidates to raise the service temperature for such joints above the ~400 C currently possible by using CuAgTi-braze technology. The basic assumption is that all strength modifying reactions during making and service life of the joint can be described by diffusion processes. The necessary scientific tools and data bases were significantly broadened by two findings in particular: 1. By making use of previously (FWF project P10736) elucidated thermodynamic descriptions (phase diagrams) of the quaternary system Cr-Ni-Si-C and its subsystems the complete diffusion path for the pair SiC + Cr25Ni75 was clarified. The diffusion path is the sequence of phases formed across the joining zone. For the said pair of partners the diffusion path at 1000 C is: SiC/dNi2 Si+graphite/Cr3 Ni2 SiC/pCr3 Ni5 Si 2 /Cr 25Ni75. The respective interdiffusion coefficients were determined from the time and temperature dependencies of the growth of the individual reaction zones. 2. This diffusion path contains phases (e.g. dNi2 Si) with a very limited homogeneity range (so called "line compounds"), which typically are described as stoichiometric compounds in thermodynamic modeling. This, however, causes the "thermodynamic factor", a function needed for modeling the diffusion behavior, to become infinite. By introducing an "average thermodynamic factor" these problems were resolved and, thus, a tool suitable for for kinetic modeling of "line compounds" was provided. The usefulnes of this modeling technique was demonstrated for the example system Ni-Si, which contains several "line compounds" for which diffusion data are available. Furthermore, a correlation was pointed out, between the self diffusion coefficients and the interatomic distances in intermetallic compounds. If this correlation holds generally, it would provide a very important broadening of the diffusion data base for such compounds.