Modeling Chromophore Order: A Guide For Improving EO Performance
- PDF / 774,026 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 54 Downloads / 183 Views
Modeling Chromophore Order: A Guide For Improving EO Performance Andreas F Tillack1, Lewis E Johnson1,2, Meghana Rawal1, Larry R Dalton1, Bruce H Robinson1 1 Department of Chemistry, University of Washington, Box 351700, Seattle, WA 98195, U.S.A. 2 Department of Chemistry, Pomona College, 645 N. College Ave, Claremont, CA 91711, U.S.A. ABSTRACT The use of organic nonlinear optical (ONLO) materials in electro-optic (EO) modulators requires that the active molecular components (chromophores) be acentrically oriented. The fundamental molecular constituents are in a condensed, glassy phase. Molecular orientation in such systems is typically achieved by applying a DC poling field to the glassy material. We are developing efficient coarse-grained classical Monte Carlo (MC) methods to simulate the order of such systems. The most challenging aspects of these simulations are convergence to an experimentally relevant equilibrium ensemble and verification of simulation accuracy. We use a variety of molecular descriptions and a variety of MC methods to achieve proper order in the shortest number of computational cycles possible. Herein, we illustrate a few examples of the types of calculations and compare with experimental results with representative amorphous organic materials, including electro-optic chromophores. INTRODUCTION The simulation of non-crystalline condensed matter, such as dendritic and polymeric materials containing ONLO chromophores is extremely challenging. Densities are always around 1 gram/cc, giving a packing fraction on the order of 70%. Obtaining equilibrium ensembles in such condensed systems is extremely difficult. Using MC methods at experimental density with a single step-size typically means that individual molecular moves are often rejected unless the move is so small as to be trivial, requiring very long simulations to reach equilibrium. We will describe how we generate structures that give meaningful moves and generate experimentally verifiable results. Accurately simulating the condensed-phase properties of high-density organic materials requires two major tasks: The first is to construct simple (but accurate) representations to complex molecules to reduce computational complexity. The second is to develop methods that efficiently extend classical MC-based calculations1 in the canonical (NVT) and isothermalisobaric (NPT) ensembles together to obtain correct molecular interactions with maximal control and with efficient (or minimal) number of moves. Coarse-grained (CG) approaches using connected spheres have been reported.2–6 We use CG representations composed of connected ellipsoids to give what we call Level of Detail (LOD) representations of functional groups of atoms in our simulations. This is a natural extension to work on single ellipsoids previously reported by our group.7,8 Ellipsoids provide a wider pallet of molecular shapes, and can be fit around many different molecular units, such as aromatic rings and long conjugated systems. The general deformability of an ellipsoid allows one to u
Data Loading...
