Reactive wetting of solid-liquid interfaces in the Au-Sn alloy system
Mechanical Properties of Metallic Materials and their Microstructural Origins
Final Report Abstract
On the experimental side, we successfully measured density, surface tension, and viscosity for the entire Al-Au system as functions of temperature and composition. Combining density and viscosity data, we developed a potential based on the embedded atom model and performed MD-simulations. This way, we not only studied the dynamics in the system, we were also able, for the first time, to propose a mechanism on how the strong negative excess volumes might be generated that are typically observed in Al-based liquid alloys. In addition, long capillary experiments are carried out in order to determine inter-diffusion coefficients and wetting experiments are performed. The outcome of the latter, however, is more complicated than we originally thought. Experiment and simulation were performed on different length scales. It remains a task for future studies to find out in which way both results can be related to each other. On the simulation side, we have investigated the complex reactive wetting process in the binary Al-Au system which involves phase transitions and fluid dynamics, based on the concept of a modeling ternary Al- Au-X phase diagram. The reactive wetting in the Al-Au system has been studied with and without considering the formation of the intermetallic Al2Au phases. It has been found that the dynamic contact angle as well as the equilibrium contact angle both are affected by the formation of the intermetallic phase. According to the simulation results, an increase of the liquid-intermetallic interfacial tension can suppress the growth of the intermetallic phase. The change of the surface tension can be achieved, for example, by changing the temperature or the concentration. In the phase-field simulations, we have also explored the effect of capillary flow on the reactive wetting by solving the Navier-Stokes equation. Two dimensionless numbers Re and C which depend on the viscosity and density have been introduced for computing the capillary flow. The experimental data for the viscosity and density are used to estimate these two dimensionless numbers. The two microstructures both include the effect of capillary flow with C=10000, which is estimated by using the viscosity and density at a temperature of 1100 K. Increasing the capillary flow by increasing C to 10000 leads to a significant inhibition effect on the growth of the intermetallic phase. The increase of the dimensionless number C can be achieved by decreasing the temperature, based on the experimental data for the viscosity and density. The liquid-vapor surface tension has been experimentally measured. The surface tensions of liquid-solid, liquid-intermetallic, solid-vapor, solid-intermetallic which are not feasibly determined by experiments are estimated by a nearest bond-breaking model. The surface tension from our DLR-partner as well as the surface tensions from the nearest bond-breaking model have been used in the phase-field model to simulate the reactive wetting. Four different liquid-intermetallic interfacial tensions have been used to simulate the microstructures of the reactive wetting. It has been found that an increase of the liquid-intermetallic interfacial tension gives rise to a suppressing effect on the growth of the intermetallic phase and a slower kinetics of the reactive wetting. These four different interfacial tensions correspond to the values at different temperatures and concentrations. In addition, we also have pointed out that the contact angle in the wetting phenomenon is not only determined by the surface tensions but also affected by the curvature. The reason is that the contact line owes an excess energy and this excess energy gives rise to a line tension which has a contribution to the force balance at the triple junctions. When the intermetallic phase is not present, the line tension only occurs at the S-L-V contact line. When the intermetallic phase is present, the line tension exists at the S-l-V and the L-I-V contact lines. Theoretical estimations as well as experimental determinations of the excess energy at a triple contact line are demanded in future in order to more accurately determine the dynamic as well as the equilibrium contact angle in the reactive wetting process. In all, in the phase-field simulations, the reactive wetting in the Al-Au system has been systematically explored with and without the formation of the intermetallic Al2Au phase. In particular, the investigation of the effect of the capillary flow on the reactive wetting benefits from the experimental data for the viscosity and density.
Publications
- Acta Mater. 95 (2015) 65
F. Wang, B. Nestler
(See online at https://doi.org/10.1016/j.actamat.2015.05.002) - Phys. Rev. B 92 (2015) 184201
H. L. Peng, Th. Voigtmann, G. Kolland, H. Kobatake, J. Brillo
(See online at https://doi.org/10.1103/PhysRevB.92.184201) - “Surface tension of liqud Al-Au binary alloys”, International Conference: High Temperature Capillarity, Karlsruhe, Germany, 2015
J. Brillo, G. Kolland
- J. Mater. Sci. 51 (2016) 4888
J. Brillo, G. Kolland
(See online at https://doi.org/10.1007/s10853-016-9794-x) - “Surface tension of binary and ternary alloys”, 12th International Workshop on Subsecond Thermophysics, Krakau, Poland, 2016
J. Brillo
- Acta Mater. 146 (2018) 106
F. Wang, A. Reiter, M. Kellner, J. Brillo, M. Selzer, B. Nestler
(See online at https://doi.org/10.1016/j.actamat.2017.12.015)
