First principle and deep learning based switching property prediction of optical bio-molecular switch

Electronic characterization of bio-molecular nanoscale devices is important in the new edge of nanotechnology and the field of nanoelectronics. In this new era, the molecular transmission properties of the Adenine bio-molecular optical switch are predicted using Density Functional Theory along with Non-Equilibrium Green's Function-based First Principle approach and this phenomenon is supported with Machine Learning based algorithms. The photo-induced switching characteristics are observed for the bi-directional bio-molecular switch for both forward and reverse bias conditions. This switching behaviour converts the position of the bio-molecular switch to its straightened and 60° twisted form. The electrical doping process plays an important role in generating p and n regions at the two ends of the switch. In this experiment, gold electrodes are used as the supporting anchor of the Adenine molecules. The HOMO–LUMO gap and I–V characteristics of the bio-molecular switch are analyzed using first principle formalisms and compared with the Machine learning approach. This ML approach helps design the future-generation prediction model for the nano-scale electronic device simulation process. This prediction model can be obtained without knowing the switching characteristics and also without knowing all-time sequence generating waveforms but only observing the waveform prediction for up to a certain period sequence.