Improving the Thermal Conductivity of Copper-Carbon Nanotube (Cu-CNT) Composites using Innovative Coating Techniques

This is a continuation of the work initiated under FY11 CIF project entitled: Copper-Carbon Nanotube (Cu-CNT) Composite for Combustion Chamber Liners in Advanced Rocket Engines. Results from FY11 work (Phase I) have shown that the multiwall CNTs when blended with NARloy-Z (Cu-3%Ag-0.5%Zr) powder do not to break up completely during blending. They tend to form clumps. When consolidated, these clumps produce an undesirable microstructure that is highly segregated at prior particle boundaries. It is proposed to improve the thermal conductivity and mechanical properties of Cu-CNT composite by coating the multiwall CNTs with a few mono-layers of a carbide forming element, Cr and/or Zr. The current technique of embedding CNTs in copper is by blending by ball milling the powder mixture in an attritor followed by sintering at elevated temperatures and pressures (used in Phase I). CNTs are used as is, and are usually present as tangled clumps. This approach has presented two problems described below: 1)The CNTs tend to form clumps and are hard to separate and incorporate into the NARloy-Z powder without destroying them. The microstructure revealed agglomeration of CNT clumps at the prior particle boundaries – see Figure 1. This microstructure feature is believed to be the underlying cause of the degradation of the thermal conductivity in the NARloy-Z-CNT composite. The CNT clumps act as insulators since they are not in intimate contact with the matrix and cannot provide good transfer of the thermal energy through the composite. This effectively reduces the cross sectional area and increases the tortuosity of the conduction path resulting in decreased thermal diffusivity and hence thermal conductivity. 2)There appears to be a lack of good chemical bonding at the CNT-NARloy-Z interface, based on microstructure analyses using Field Emission Scanning Electron Microscopy (FESEM) and Electron Spectroscopy for Chemical Analysis (ESCA). These analyses indicated the presence of ZrO2 in the samples. The amount of ZrC present in the NARloy-Z matrix was extremely small, not sufficient to develop a strong chemical bond between the CNTs and the NARloy-Z matrix. It is planned to address both of these problems in the continuation effort. Innovative techniques will be used to prevent agglomeration of CNT and to make sure that Zr is present at the copper/CNT interface where it is needed for good thermal and metallurgical bonding. More »