Atomistic study of flow enhancement in ceramic nanopores induced by graphitic coatings

Abstract

Carbon nanotubes and graphene are nanostructured carbon allotropes that, as a result of their remarkable physical properties, hold a great potential to become the fundamental building blocks for a wide range of foreseeable functional nanodevices. In particular, these materials have demonstrated extremely low friction to water flow. Hence, attracting a great interest for being used as fluid conduits, integrating complex nanofluidic devices with applications in technological fields such as desalination, chemical separation, nanosensors and Lab-On-a-Chip units. The low friction measured between water and graphitic materials has been associated to the presence of slippage in the fluid-solid interface, which is a consequence of atomic level smoothness of graphitic surfaces and to the generally accepted weak interaction between these surfaces and water. Nevertheless, recent experiments have reported that the effect of airborne contaminants and the presence underlying substrates can alter significantly the wettability of graphene. Therefore, in order to achieve an optimal design of nanofluidic devices, a comprehensive understanding of slippage in a water flow confined between graphitic materials is still lacking. In this thesis, we employ the molecular dynamics technique to study the fluid flow in pores fabricated with graphene layers and carbon nanotubes. In particular, the present work focuses in understanding how the physical properties of graphene, carbon nanotubes and underlying substrates determine the water flow through nanoconduits. Furthermore, an important aspect of the study is to measure the performance of graphitic materials to work as wall coating to reduce the hydrodynamic resistance in nanoconduits. Specifically, this thesis is divided in two main parts. The first part includes an introduction to the research field of nanofluidics and to the technique of molecular dynamics. The second part includes three research works and conclusions. In the first research work, we conduct a study of water flow nanoconfined between parallel graphene layers at high shear rates. We observe that the crystallographic features of graphene influence the transition of the slip boundary condition at high shear rates, resulting in slip lengths that depend on the direction of the flow. In the second work we evaluate the use of graphene layers as wall coatings in silica nanochannels to induce a flow enhancement. Our models reproduce the experimental translucency to wettability of graphene coating on silica surfaces reported by Rafiee et al.. Moreover, we demonstrate that effectively the use of monolayer graphene as coatings induces an important flow enhancement, despite a decrease in the water contact angle and available channel cross section. Finally, we study the water flow enhancement, resultant of the use of carbon nanotubes as coatings on a cylindrical silica pore. In this study, the atomistic model of a silica pore coated by a single walled carbon nanotube is parametrized and characterized. Thereafter, the interactions between water and the pore are calibrated based on two possible scenarios that reproduce the same wettability. In the first scenario, the carbon nanotube is translucenct to the wettability of the underlying silica surface, thus water interacts with the carbon nanotube and the silica pore. The second scenario reproduces a carbon nanotube that is opaque to the interaction between underlying substrate and water, but whose wettability is tuned vii through calibrating the interaction between carbon atoms and water molecules. Furthermore, we evaluate the properties of Poiseuille flow of water in the coated pore. Our results show an important variation in water flow as comparing the two wetting scenarios. We link this difference to the different energy corrugation that they reproduce. Nevertheless, both scenarios present important flow enhancement, demonstrating that the use of single walled carbon nanotubes is advantageous and therefore represents a potential option to reduce hydrodynamic losses in nanoconduits. viii