Theoretical and Experimental Studies of Thermal Conductivity in Functionalized Carbon Nanotubes Optimized via Genetic Algorithms

The use of carbon nanotubes (CNTs) to improve the thermal conductivity of composite materials is thwarted by their large thermal boundary resistance. We study how to overcome this Kapitza resistance by functionalizing CNTs with mixed molecular chains. Certain configurations of chains improve the transmission of thermal vibrations through our systems by decreasing phonon mismatch between the CNTs and their surrounding matrix. We explore the space of possible designs using a genetic algorithm (GA) that evolves the molecular sidechain as its DNA and optimize the GA search procedure. We show how different configurations of attached chains affect the samples' $\kappa$ values by considering various permutations of over a dozen molecular units from our chemical library. We vary the composition to maximize $\kappa$. To validate and optimize these designs, we perform molecular dynamics simulations for comparison. We also present experimental results of composites enhanced with functionalized CNTs and make comparisons to the theory. We find that functionalization can substantially improve $\kappa$.