ABSTRACT
Ethylene-bridged azobenzene (diazocine) has been shown to have superior photochemical properties. So far, however, experimental and theoretical quantum yields did not match, not even qualitatively. Here, a large-scale QM/MM surface-hopping study of this molecule is presented. For both photo-isomerization directions, surprisingly prominent solvent effects are found and analyzed by detailed comparisons against gas phase dynamics and experimental data. By taking explicit n-hexane solvent into account, the quantum yields change markedly and approach the experimental values, where the same solvent was used.
Subject(s)
Azo Compounds/chemistry , Quantum Theory , Thermodynamics , Ethylenes/chemistry , Molecular Structure , Surface PropertiesABSTRACT
Using QM/MM methods, we have simulated the action of a simple molecular machine, a cilium. It consists of a platform for surface mounting, a photochemical motor unit, and a tail-like effector that amplifies the small-scale conformational change of the motor unit into a larger-scale beating motion usable for molecular transport. In this proof-of-principle application, we show that the techniques used here make it possible to perform such simulations within reasonable real time, if the device action is sufficiently fast. Additionally, we show that this molecular device actually works as intended for one isomerization direction. For the other direction, results are inconclusive, possibly because the total propagation times we can afford are too short to capture the complete event.
ABSTRACT
We demonstrate the possibility to design molecules for specific tasks, using a fully automatic global optimization setup employing genetic algorithms. As an example, we tune the two excitation wavelengths of a molecular switch backbone to arbitrarily pre-set values, by an automatic optimization of the substituent pattern.