The metallic substrates the researchers chose for the porphyrin molecules to assemble on, the close packed single crystal surfaces of copper and silver, are widely used as substrates in surface science. This is due to the densely packed nature of the surfaces, which allow the molecules to exhibit a smooth adsorption environment. Additionally, copper and silver each react differently with porhyrins – the molecule adsorbs more strongly on copper, whereas silver does a better job of keeping the electronic structure of the molecule intact – allowing the researchers to monitor a variety of competing effects for future applications.
In their simulation, porphyrin molecules were placed on a copper or silver slab, which was repeated periodically to simulate an extended surface. After finding the optimal geometry in which the molecules would adsorb on the surface, the researchers altered the size of the metal slab to increase or decrease the distance between molecules, thus simulating different molecular coverages. The computational setup gave them a switch to turn the energy contributions of neighboring molecules on and off, in order to observe the interplay of the individual interactions.
Diller and Reuter, along with colleagues Reinhard Maurer and Moritz Müller, who is first author on the paper, found that the weak long-range van der Waals interactions yielded the largest contribution to the molecule-surface interaction, and showed that the often employed methods to quantify the electronic charges in the system have to be used with caution. Surprisingly, while interactions directly between molecules are negligible, the researcher found indications for surface-mediated molecule-molecule interactions at higher molecular coverages.
“The analysis of the electronic structure and the individual interaction components allows us to better understand the self-assembly of porphine adsorbed on copper and silver, and additionally enables predictions for more complex porphyrine analogues,” Diller said. “These conclusions, however, come without yet considering the effects of atomic motion at finite temperature, which we did not study in this work.”
Page 2 of 2