Accumulation and ordering of P3HT oligomers at the liquid-vapor interface with implications for thin-film morphology.

The morphology of semiconducting polymer thin films is known to have a profound effect on their opto-electronic properties. Although considerable efforts have been made to control and understand the processes which influence the structures of these systems, it remains largely unclear what physical factors determine the arrangement of polymer chains in spin-cast films. Here, we investigate the role that the liquid-vapor interfaces in chlorobenzene solutions of poly(3-hexylthiophene) [P3HT] play in the conformational geometries adopted by the polymers. Using all-atom molecular dynamics (MD), and supported by toy-model simulations, we demonstrate that, with increasing concentration, P3HT oligomers in solution exhibit a strong propensity for the liquid-vapor interface. Due to the differential solubility of the backbone and side chains of the oligomers, in the vicinity of this interface, hexyl chains and the thiophene rings, have a clear orientational preference with respect to the liquid surface. At high concentrations, we additionally establish a substantial degree of inter-oligomer alignment and thiophene ring stacking near the interface. Our results broadly concur with the limited existing experimental evidence and we suggest that the interfacial structure can act as a template for film structure. We argue that the differences in solvent affinity of the side chain and backbone moieties are the driving force for the anisotropic orientations of the polymers near the interface. This finer grained description contrasts with the usual monolithic characterization of polymer units. Since this phenomenon can be controlled by concurrent chemical design and the choice of solvents, this work establishes a fabrication principle which may be useful to develop more highly functional polymer films.

Predicting optical spectra for optoelectronic polymers using coarse-grained models and recurrent neural networks.

Coarse-grained modeling of conjugated polymers has become an increasingly popular route to investigate the physics of organic optoelectronic materials. While ultraviolet (UV)-vis spectroscopy remains one of the key experimental methods for the interrogation of these materials, a rigorous bridge between simulated coarse-grained structures and spectroscopy has not been established. Here, we address this challenge by developing a method that can predict spectra of conjugated polymers directly from coarse-grained representations while avoiding repetitive procedures such as ad hoc back-mapping from coarse-grained to atomistic representations followed by spectral computation using quantum chemistry. Our approach is based on a generative deep-learning model: the long-short-term memory recurrent neural network (LSTM-RNN). The latter is suggested by the apparent similarity between natural languages and the mathematical structure of perturbative expansions of, in our case, excited-state energies perturbed by conformational fluctuations. We also use this model to explore the level of sensitivity of spectra to the coarse-grained representation back-mapping protocol. Our approach presents a tool uniquely suited for improving postsimulation analysis protocols, as well as, potentially, for including spectral data as input in the refinement of coarse-grained potentials.

Direct determination of discrete harmonic bath parameters from molecular dynamics simulations.

We present a direct procedure for determining the parameters of a discrete harmonic bath modeling the influence of a complex condensed phase environment on the system of interest. The procedure employs an efficient discretization of the spectral density into modes that correspond to equal fractions of the reorganization energy. The new procedure uses directly the classical correlation function (available from molecular dynamics calculations) as input, avoiding numerical computation of the spectral density by means of a discrete Fourier transform. Convergence is obtained using a shorter time length of the correlation function, leading to significant computational savings. © 2016 Wiley Periodicals, Inc.

Direct Computation of Influence Functional Coefficients from Numerical Correlation Functions.

Influence functional methods provide a powerful approach for simulating the dynamics of a system embedded in a harmonic bath, which may be parametrized to reflect a variety of environments and chemical processes. In this work, we develop a procedure for calculating the coefficients of the discretized influence functional using the classical approximation to the time correlation function, which is usually available through molecular dynamics simulations. This procedure circumvents the calculation of a spectral density via Fourier inversion of the correlation function. When the correlation function is available with high precision, we find that the direct procedure yields results just as accurate as those obtained using the spectral density expressions. However, when statistical noise is present, the direct procedure produces more accurate results. The direct procedure is efficient and easy to implement.