Metal-ceramic bonding is a key technique for the application of ceramics, especially, structural ceramics. Among the several bonding techniques which have been developed, the most available bonding technique is the brazing in which wetting of solid ceramics by metal melt is a basic requirement. Even though many researchers have studied the wetting and bonding phenomena of metal melts to ceramics, many things are still remained as unveiled and the reliable brazing technique for high temperature application is not developed yet. Thus, in this study, the wetting and bounding phenomena of metal melt to ceramics have been studied.
In order to understand the wetting and interfacial reacton phenomena in metal-ceramic system, the kinetics of the reaction induced wetting is formulated in chapter3. It is based on the Yin's approach, but two more assumptions are added. The first is on the quasi-static nature of interfacial reaction and the second is on the driving force for the wetting. Every moment of the wetting is considered as an equilibrium state and modified Naidich's equation on the driving force for wetting is accepted as valid. According to the obtained equation, the contact angle decreases exponentially with time. This result is ascertained from several experimental reports on the wetting. It is considered as resulting from the coupled effects of the drop velocity and the rate of interfacial area increase with the advance of drop. Reduction in the interfacial energy between liquid metal and solid ceramics caused by the interfacial reaction seems to act as the driving force for the wetting. However, the variation of interfacial energy with time does not seem to have a minimum and even the magnitude of the decrease is not directly proportional to that of reaction free energy decrease as Aksay insisted. Rather, it seems to decrease monotonically to a ceratain value as reaction proceed and of its equilibrium value can only be obtained from experiment.
In chapter 4, the wetting and interfacial reaction between $Al/Si_3N_4$ and $Al-Si/Si_3N_4$ was experimented with a view to testify the kinetics of wetting developed at chapter 4. The wetting of $Si_3N_4$ by Al melt starts when the oxide layer of Al is removed and pure Al melt bring into contact with $Si_3N_4$ at 950℃. AIN is formed as a result of the reaction between Al and $Si_3N_4$. Contact angle is decreased exponentially at 1000℃ and 1100℃ as anticipated from the kinetics of wetting obtained at chapter 3. The calculated interfacial energy variation has no minimum with time and its equilibrium value does not show any relation with the magnitude of the reaction free energy of AIN, which supports the conclusion drawn at chapter 3. With the addition of Si to Al, the wettability of Al to $Si_3N_4$ is enhanced and this seems to result from the lower energy of $Si/Si_3N_4$ interface than that of $Al/Si_3N_4$ interface. At 1000℃, the difference of kinetics of wetting between several Al-Si alloys is consider as the result of the difference in viscosity with the Si contents. The bonding strength measured by 4 point bending is undegraded until 20wt% Si was added to pure Al. Thus, it is concluded that Al-Si alloys con serve as a good filler for brazing of $Si_3N_4$ ceramics.
In chapter 5, Cu-$Cu_2O$ eutectic bonding of copper to $Al_2O_3$ has been studied at 1075℃ in $N_2$. In order to elucidate the reaction product at $Cu/Al_2O_3$ interface, the bonding time was prolonged upto 24 hours. $CuAlO_2$ is found at the $Cu/Al_2O_3$ interface as a continuous layer. With the formation of $CuAlO_2$, the bonding strength is enhanced. There seems to be two kinds of $CuAlO_2$ formation mechanism working at $Cu/Al_2O_3$ interface. Until now, $CuAlO_2$, has been known as forming from the solid state reaction between $Cu_2O(S)$ and $Al_2O_3(S)$. However, the formation of $CuAlO_2$ as a continuous layer can't be explained with only this mechanism. Thus, another formation mechanism is conceived in which no solid $Cu_2O$ is needed. This latter mechanism can be expressed as below.
$2Cu(l)+[O]_{Cu(l)}+Al_2O_3=2CuAlO_2$
Here Cu(l) means oxygen free Cu melt and $[O]_{Cu(l)}$ stands for the oxygen dissolved in Cu melt. Thermodynamic calculation shows that this reaction is possible.
