Despite the plethora of molecular switches, memories and motors developed in recent decades, fabrication of complex molecular machines comparable to living cells or silicon processors is still an elusive goal. Living nature solves this problem by encoding structural information into polymers which self-assemble in a predictable way by matching complementary sequences of hydrogen-bonded side-groups. DNA origami[1], which exploits this principle, is a rare example of scalable nano-fabrication technology. Our ambition is to computationally design photosensitive polymers which self-assemble in vacuum on the surface of ionic crystals, and can therefore be combined with photolithography, high resolution scanning probe microscopy and other tools used in surface science, molecular electronics and chip manufacturing. In order to find molecules with the optimal structure and thermodynamic properties, we explore the configurational space of a broad variety of monomers and polymers including their interaction with different ionic substrates. The key ingredient which makes this exploration feasible is a newly developed simulation software[2] optimized for configuration-sampling of small molecules on ionic substrate which combines GPU accelerated classical force-fields with density functional theory. This software aims to bring methodologies used in fields such as drug-design and ligand-docking into surface-science. Significant speedup is achieved by using grid-projected force-field for the description of interactions with rigid substrate and multiple replicas of the system running in parallel on a single GPU.
References
[1] P.W.K. Rothemund, Nature, 440, 297 (2006).
[2] https://github.com/ProkopHapala/FireCore