Computational Design of Nanofabrication of Molecular Computers

Text

Molecular electronics holds the potential for the ultimate miniaturization of computing and other machines. However, its widespread application is hindered by the lack of scalable nanofabrication methods. Our goal is to develop scalable bottom-up fabrication methods for molecular electronics and nanotechnology that would bridge bio-mimetic bottom-up approaches like DNA origami with state-of-the-art top-down photolithography. A central focus of our project is the design of programmable polymer templates that function similarly to DNA but, unlike DNA, are photosensitive and operate in anhydrous environment of the surface of ionic crystals. These polymers encode structural information that drives deterministic assembly of molecular components, while photo-chemical reactions ensure the covalent bonding of these pre-assembled structures. Ultimately, such nanofabrication method should allow for atomically precise assembly of molecular electronic and photonic components below the resolution limits of photolithography, while also enabling industrial-scale fabrication of large-scale structures with billions of components, paving the way to molecular nano-chips and computers.

Assembly of end groups under UV light
Description
Assembly of molecular electronics components by non-covalent interaction between endgroups, followed by creation of covalent bonds in photochemical reaction induced by UV light [ACS Nano 18 (2024) 9969].  

Publications

FireCore: Integrated simulation software for on-surface chemistry

For efficient design of molecular assemblers, templated synthesis and simulation of on-surface chemical processes in general, we are developing FireCore, an integrated multi-scale simulation platform specifically dedicated to molecular surface science. For this purpose, FireCore includes several spetialized or optimized methods which makes it more efficient than general-purpose software (e.g. LAMMPS) in this specific domain:

  • Efficient molecular-substrate interaction modeling using fast grid-projected force fields (GridFF), which significantly outperform traditional pairwise classical force fields in speed (and under some conditions can also improve accuracy).
  • GPU-accelerated classical force fields, optimized for paralel simulations of many instances of simple molecules (>1000 molecular replicas) on a single GPU, allowing high-throughput configuration sampling, including global optimization and free-energy sampling.
  • Integration of classical force fields with density-functional (tight-binding DFT/DFTB) package Fireball, enabling efficient QM/MM modeling of chemical reactions.
  • High-resolution AFM simulation modules that allow for easy interpretation of experimental images by rapid exploration of simulated images of different molecular configurations on the surface.
Theme is contributed to by