Computational chemistry could only deal with small, isolated molecules. For these, local optimizations from guessed starting structures were sufficient. These starting structures could be taken from chemical intuition or from experiment. This chapter deals with Non-Deterministic Global Optimization (NDGO) approaches. In contrast to deterministic global optimization methods, NDGO approaches have to forgo any attempts to cover all of search space and to find the global optimum with certainty. Interestingly, NDGO methods share the aim of “getting across barriers more quickly than normally” with other areas of method development, mostly pertaining to calculations of free energies. In essence, machine learning combines suitable representations with brute-force big-data interpolation, employing interpolation functions that are intentionally very flexible. Monte Carlo-style “mutation” moves (but also “crossover”-style information exchange moves) in global structure optimization can easily move several atoms too close to each other, resulting in strongly repulsive forces.
This chapter provides the essential background to a novice modeler on the choice of simulation techniques used to model Deep Eutectic Solvents (DESs). It describes methods used to obtain important physical, thermodynamic, transport, and structural properties of bulk DES systems including an evaluation of the strengths and drawbacks of the current simulation models. The chapter discusses future directions for simulating DES-based systems. One of the major thrust areas of ab initio investigations on DESs has been to provide a physical explanation for the observed low melting point in these systems and its effects on their physicochemical properties. Molecular simulations have played a crucial role in conjunction with experimental investigations in elucidating the structure–property relationships of DESs. Classical molecular dynamics simulations that obey Newton's laws of motion use force fields to calculate the potential energy of a system as a function of their atomic coordinates.
This chapter details the different experimental microphase formers and provides a minimal theoretical framework to present the simulation challenges associated with studying model microphase formers. Block copolymers are by far the most studied microphase formers. The chapter focuses on the phenomenological field theory description of the universality of the microphase formation and of the nature of the order-disorder transition. The chapter describes molecular simulation methods that have been specifically designed to achieve equilibrium in the periodic microphase regime. It details the thermodynamic framework and a free energy integration simulation method, followed by a concrete introduction to the ghost particle/cluster switching method. The chapter discusses several classical Monte Carlo algorithms to enhance the efficiency of simulating disordered microphases. It presents three models for which quantitative results have been obtained: a one-dimensional, a lattice, and an off-lattice microphase former. Fine-tuning colloidal suspensions to allow the formation of periodic microphases thus remains an open experimental problem.
