Session A56: Molecular Crystal Structure Prediction and Polymorphism

Towards in silico design of organic photomechanical materials

Organic photomechanical crystals transform light into mechanical work, producing work densities that are potentially orders of magnitudes higher than those achieved by piezoelectrics or elastomer materials.  However, the complexities of the organic solid state make it difficult to anticipate how the crystal structure will transform upon photomechanical reaction, or, in many cases, to even characterize the transformation experimentally. We have recently developed a computational approach for predicting these crystalline transformations entirely in silico.  It enables us to bridge between the structure of the molecular photochrome and the solid-state photomechanical response, providing an atomistic mechanism for the crystal transformation.  Our approach combines crystal structure prediction, topochemical concepts, solid-state density functional theory, and corrections to overcome the limitations of widely-used density functionals.  Using this approach, we have studied a variety of [4+4] photodimerization reactions in anthracene reactions and ring-opening/closing in diarylethenes, from which we have have demonstrated the large ideal photomechanical work densities and highly anisotropic nature of the crystalline response.  More importantly, we have extracted a series of design principles regarding how molecular structure and crystal packing impact the photomechanical response that can guide the development of improved photomechanical materials.   

Topology, Molecular Simulation, and Machine Learning as Routes to Predicting Structure and Phase Behavior in Molecular Crystals

The different solid structures or polymorphs of atomic and molecular crystals often possess different physical and chemical properties.  Structural differences between organic molecular crystal polymorphs can affect, for example, bioavailability of active pharmaceutical formulations, the lethality of contact insecticides, and diffusive behavior in host-guest systems.  In metallic crystals, structural differences may determine how different phases may be used in electronic device applications.  Crystallization conditions can influence polymorph selection, making an experimentally driven hunt for polymorphs difficult.  These efforts are further complicated when polymorphs initially obtained under a particular experimental protocol “disappear” in favor of another polymorph in subsequent repetitions of the experiment.  Theory and computation can potentially play a vital role in mapping the landscape of crystal polymorphism.  Traditional methods for predicting crystal structures and investigating solid-solid phase transformation behavior face their own challenges, and therefore, new approaches are needed.  In this talk, I will show, by leveraging concepts from mathematics, specifically geometry and topology, and classical and quantum statistical mechanics in combination with techniques of molecular simulation, and machine learning, that new paradigms are emerging in our ability to predict molecular crystal structures and determine kinetics of polymorphic phase transformations and guest-molecule diffusion.   

Structure Prediction and Discovery of Molecular Crystals with Enhanced Electronic Properties

Molecular crystals often exhibit polymorphism, the crystallization of the same molecule in several different structures. Crystal structure may profoundly influence the physical and chemical properties. We combine first principles simulations with machine learning (ML) and optimization algorithms to predict the structure of molecular crystals and discover molecular crystals with enhanced electronic and optical properties. Our crystal structure prediction (CSP) workflow begins by estimating the unit cell volume using a machine learned model based on the volume enclosed by the molecule's packing-accessible surface and molecular topological fragments, which capture the bonding environments of the atoms in the molecule and the types of inter-molecular interactions the molecules may form. Next, the random structure generator, Genarris, performs preliminary configuration space screening by random sampling with physical constraints. Genarris generates structures with a distribution around the target volume in all space groups compatible with the molecular point group symmetry and the requested number of molecules per unit cell, including space groups with molecules occupying special Wyckoff positions. The "raw" pool is clustered and down-selected based on considerations of diversity and energy to form an initial populaiton for the GAtor genetic algorithm (GA). A GA uses the evolutionary principle of survival of the fittest to perform global optimization. GAtor's special features are: A massive parallelization scheme enables effective utilization of high performance computing resources; A variety of breeding operators (crossover and mutation) tailored for molecular crystals provide a balance between exploration and exploitation; Evolutionary niching, performed by using ML for dynamic clustering and then using a cluster-based fitness function, helps overcome initial pool biases and selection biases by steering the GA to under-explored regions of the configuration space; Property-based fitness functions enable inverse design of crystal structures with target properties. In this talk, our CSP workflow will be demonstrated for select cases.   

Digital Chemistry in Action: Insights on Controlling Polymorphism at EMD

Digital chemistry has the potential to significantly accelerate discovery and formulation in our industry. This talk will delve into our crystal structure prediction endeavors, from pharmaceutical formulation to OLED material production. Using a blend of cheminformatics and advanced computational techniques like dispersion-corrected DFT, we efficiently identify polymorphs and solvate forms. While speed often trumps precision in industrial settings, understanding and conveying prediction errors remains paramount. With substantial value generated in the past years and significant integration of digital insights into our workflows, our experiences provide a unique academic-industrial interface worth exploring   

Multimer Embedding Methods: Molecular Crystals and Beyond

In recent years, advances in method developments have been leading to increasingly reliable predictions of molecular crystals. This has been highlighted by the impressive results in the prediction of possible crystal structures for organic molecules and also organic salts. Much of this progress can be attributed to the increased application of density functional theory including dispersion models using periodic boundary conditions.

We assess the available computational approaches for the revised X23b benchmark set of molecular crystals using thermally-expanded structures which were calculated via the quasi-harmonic approximation. The methods evaluated include density functionals in combination with dispersion models, as well as our own methods which are based on multimer embedding schemes. Utilizing the latter approach, we are able to approximate periodic results of hybrid functionals at a fraction of the computational time needed for the canonical methods. These methods can be evaluated not only for energies, but also for geometries, cell volumes, and vibrational properties. The convergence of the embedding scheme with the multimer order (monomer, dimer, trimer) and distances are discussed, whereas trimer interactions are of surprising importance.

Finally, we discuss the applicability including the success of the embedding scheme at the Cambridge Crystallographic Data Centre 7^th crystal structure prediction blind test of organic molecules.