Materials with specifically engineered optical properties are important for the development of photonic technologies, optical sensors and devices that exploit nonlinear optical phenomena. However, developing such materials requires an understanding of how molecules behave in different environments and the ability to predict their properties reliably. Tomáš Hrivnák from IT4Innovations at VSB – Technical University of Ostrava and Heribert Reis from the National Hellenic Research Foundation therefore compared several computational methods for predicting the optical response of molecules, each with a different level of computational cost.

The researchers focused on para-nitroaniline, a molecule with well-known nonlinear optical properties. They investigated its linear and nonlinear optical response in three solvents: cyclohexane, 1,4-dioxane and tetrahydrofuran, which represent environments with different intermolecular interaction characteristics. The demanding calculations were carried out using IT4Innovations supercomputers.

The accuracy of predictions also depends on how the computational model accounts for the molecule's surroundings. While some methods describe the solvent as a continuous medium, others explicitly include individual surrounding molecules and their interactions. The authors compared five computational approaches, covering both implicit continuum models and methods with an explicit description of the surrounding molecules. These ranged from the classical evaluation of the local electric field and a quantum-mechanical description of the entire system to a hybrid combination of explicit and continuum solvent models. They distinguished between three components of the optical response: the response of the molecule itself, the contribution of the reaction field of the surrounding medium, and the resulting effective response of the system.

The results showed that, with an appropriate computational setup, almost all the methods investigated can produce results that are in good agreement with one another, including for the first hyperpolarizability. This quantity characterises the response of a molecule's dipole moment to an electric field at second order. It is important for studying nonlinear optical phenomena such as second-harmonic generation. The rigorous local field (RLF) method proved particularly promising. It enables the rapid evaluation of optical properties across large sets of molecular configurations, with the results directly comparable to experimental measurements. This is particularly valuable when analysing molecular trajectories, which capture how the arrangement of molecules changes over time.

The study also highlights how details of the implementation of individual methods can significantly affect predictions. Important factors include the inclusion of higher-order multipole contributions, the way local electric field distributions are evaluated, the modelling of the effective electric field in the condensed phase, and limitations on the size of systems that can be practically simulated.

This research contributes to the development of computational approaches that make it possible to predict the optical properties of molecules more efficiently and identify promising candidates for further experimental investigation. Faster evaluation of large numbers of molecular configurations can help researchers better assess the influence of the surrounding environment and make more efficient use of computational resources. The findings were published in the scientific journal The Journal of Physical Chemistry B.

 

Research article
Assessment of Explicit and Implicit Solvation Models for Linear and Nonlinear Optical Properties of Para-Nitroaniline in Three Different Solvents
https://doi.org/10.1021/acs.jpcb.6c01326