| Electro-optic materials enable a wide variety of photonics applications such as micro-scale optical sensors, terahertz spectroscopy, photonic computing, quantum key distribution, and high speed data transmission for computing as well as global telecommunications. Organic 2nd-order non-linear optical (ONLO) materials offer several key advantages for photonic devices such as intrinsically higher bandwidth on the order of THz, lower power consumption, and smaller device structures compared to currently used inorganic materials such as lithium niobate. ONLO materials consist of electro-optic chromophores arranged such that overall, acentric dipole order is present in the material. Crucial insight into the acentric ordering of an ensemble of electro-optic chromophores can be provided by computational modeling. Presented in this dissertation is a coarse-graining (CG) Monte Carlo approach, the Level-of-Detail (LoD) method, enabling the systematic determination of CG model parameters with no adjustable parameters from ab initio quantum mechanical calculations and fully-atomistic force fields. The LoD method's ability to correctly represent all-atom behavior is demonstrated on a diverse range of condensed molecular systems relevant to different aspects of the simulation of electro-optic materials such as the accurate simulation of pi-pi interactions, the incorporation of flexible molecular linkers, and the prediction of dielectric behavior. Details of molecular interactions that determine the extent of acentric order are investigated and the observations and conclusions derived in this thesis culminate in a set of design criteria for construction of future molecules by experimentalists. |
