Stochastic Methods for Molecular Chiroptical Properties

Much like hands whose mirror images cannot be superposed, chiral molecules have a three-dimensional "handedness". The two "hands" of the compound can exhibit dramatically different chemical behavior when reacting in a chiral environment, such as the human metabolism. One "hand" can be effective against certain symptoms, the other "hand" can be inactive or have grim side effects. Between 1957 and 1962 the two "hands" of thalidomide were prescribed to pregnant women in Europe against morning sickness. Though effective, one of the "hands" was found to cause birth defects and the drug soon withdrawn from commerce. The case of thalidomide reinforced the need for methodologies to distinguish between the different "hands" of a chiral molecule as this is critical for the design of new drugs or materials with innovative properties. The intrinsic dissymmetry of chiral molecules leads to peculiar interactions with light sources. Quantifying molecular chiroptical properties provides an invaluable tool to shed light on the structure of chiral compounds. Despite impressive leaps in experimental technologies,spectroscopy alone is in many cases unable to provide the ultimate and conclusive answer on the handedness of a given compound without input from computational methods. It is still an open challenge to devise a computational methodology that is accurate and efficient for large chiral molecular systems both in the gas and the liquid phase. The FLAKS project has opened new frontiers in the faithful computational prediction of molecular electronic structure. Robust computational methodologies have been implemented using cutting-edge Monte Carlo algorithms. The latter leverage random number generators and are well-suited to harness current state-of-the-art computer hardware. Using diagrammatic techniques, a new formulation of the highly accurate Fock-space coupled cluster Monte Carlo algorithm has been put forth. This new formulation reduces RAM and CPU cost of highly accurate computations. FLAKS is a major step towards predictive computations of chiroptical properties, helping the rational design of chiral molecules for biological,pharmaceutical and materials applications.

The project has improved, on a fundamental level, the formulation and implementation of stochastic methods for the solution of the coupled cluster equations. The project has initiated a new, fruitful direction for the application of Monte Carlo methods in high-level correlated wavefunction theories. The approach developed is already useful for the study of complex electronic structure problems, but their full impact has still to be realized. The methods can potentially be applied on large molecular systems, efficiently exploiting existing and upcoming high-performance computing infrastructure. It is expected that the method will have an increasing impact in the quantum chemical community, as it allows highly accurate calculations at a fraction of the computational cost of current methodologies.

Chirality is the unique property possessed by molecules with a three-dimensional ``handedness'' so that their mirror images cannot be superposed. Chiral molecules are used in a wide range of applications where the two enantiomers ("hands") of the compound can exhibit dramatically different chemical behavior when reacting in a chiral environment, such as the human metabolism. The drug thalidomide is an infamous example. Prescribed in Europe against morning sickness to pregnant women between 1957 and 1962, it was withdrawn soon thereafter when numerous cases of birth defects were reported. Clearly, the ability to distinguish between the different "hands" - known as assigning the absolute configuration of the compound - has a critical impact on the design of improved drugs or materials with innovative properties. These species have peculiar responses to electromagnetic radiation and this can be exploited to shed light on their molecular structure. Despite tremendous technical advances in spectroscopic technologies, experiments alone can in many cases not provide the ultimate and conclusive answer on the question of absolute configurations without theoretical input. It is still an open challenge to devise a computational methodology that is accurate and efficient for sizable molecular systems both in the gas and the liquid phase. The intrinsic complexity engendered by the solution is an additional challenge to reliable computations of molecular chiroptical properties. The FLAKS project will open new frontiers in the computational prediction of molecular chiroptical properties, by providing the community with a faithful, accurate and efficient protocol. I will achieve this goal by combining robust and well-established formulations of chiroptical properties with parallel and scalable stochastic solution strategies. The open-source computational software tool developed within FLAKS will be a major step towards predictive computations of chiroptical properties.