Assessment of the accuracy of DFT (Density Functional Theory) for the photochromic behavior of dihydroazulene (DHA)

Abstract

Efficient utilization of the sun as a renewable and clean energy source is one of the greatest goals and challenges of this century due to the increasing demand for energy and its environmental impact. Photoactive molecules that can store the sun's energy in the form of chemical bonds have been of interest to harness the sun's energy since the 1970s. However, all of the photoactive systems studied have problems with degradation making them impractical. Recently, the Grossman Group used computation to show that nanotemplating of the azobenzene photoactivesystem improves problems with degradation. We believe that this could be a platform technology for other photoactive systems like azobenzene. We would like to use high throughput screenings to identify other promising photoactive molecules. We would like to use Density Functional Theory (DFT) calculations for these studies, since DFT is the least computationally intense Quantum Mechanical model used on large chemical systems. For photosystems like azobenzene, nombomadiene, and diruthenium fulvalene, DFT predictions have been found to match well with experimental predictions, suggesting that DFT can be used to confidently predict properties of these fuels. However, for dihydroazulene(DHA) photoactive predictions using different DFT functionals do not match with each other and experiment. Our analysis suggests that lack of error cancelation due to a drastic change in the conjugation in DHA as compared to VHF might account for the variation in predictions based on different DFT functionals. It was also found that the DFT functional, [psi]B97X-D, makes similar predictions as the more computationally intense post Hartree-Fock methods by including couple cluster terms that better capture weak interactions.