Reactive Force Field-Based Molecular Dynamics Simulations on the Thermal Stability of Trimesic Acid on Graphene: Implications for the Design of Supramolecular Networks

In this work, we used the ReaxFF force field to investigate the dynamics of different network structures of trimesic acid (TMA) molecules on graphene as a function of temperature. We considered the so-called honeycomb, filled honeycomb, flower, zigzag, and close-packed TMA motifs. The thermal stability was investigated using molecular dynamics simulations with the constant number of molecules, volume, and temperature and force-biased Monte Carlo calculations up to 650 K. Our simulations provide detailed atomistic insights into the intermolecular and molecule–substrate interactions responsible for the self-assembly or the breakage of the TMA networks at different temperatures. The dynamics of hydrogen bonding were followed by counting the number of hydrogen bonds as well as by analyzing OH radial distribution functions. According to the melting temperatures obtained, the honeycomb structure has a higher stability than the high-coverage zigzag and close-packed structures. Guest TMA molecules within the pores of the honeycomb motif further increase its thermal stability, thus showing the beneficial effect of host–guest interactions. The twisting and rotation of carboxylic groups with increasing temperature are responsible for the breakage of hydrogen bonds, which ultimately leads to the melting of the networks. Partial TMA desorption observed at the onset of network disordering was attributed to the intermolecular vibrational energy transfer between the molecules. For the high-coverage close-packed network and for an island of TMA molecules with a close-packed structure, we observed a phase transition to the honeycomb structure as a consequence of the stronger dimeric −COOH bonding of the latter. The energetics of the formation of the different networks from TMA molecules in the gas phase was also investigated. Intermolecular interactions and TMA–graphene interactions have similar magnitudes. The stability of the different networks cannot be fully understood only based on energetic considerations, and in the case of the dense close-packed structure, MD simulations show how it is rapidly destabilized.