Continuum embeddings in condensed‐matter simulations

Abstract

Continuum models have a long tradition in computational chemistry, where they have provided a compact and efficient way to characterize environment effects in quantum-mechanical simulations of solvated systems. Fattebert and Gygi pioneered the development of continuum dielectric embedding schemes for periodic systems and their seamless extension toward molecular dynamics simulations. Following their work, continuum embedding approaches in condensedmatter simulations have thrived. The possibility to model wet and electrified interfaces, with a reduced computational overhead with respect to isolated systems, is opening new perspectives in the characterization of materials and devices. Important applications of these new techniques are in the field of catalysis, electro-chemistry, electro-catalysis, etc. Here we will address the main physical and computational aspects of continuum embedding schemes recently developed for condensed-matter simulations, underlying their peculiarities and their differences with respect to the quantum-chemistry state-of-the-art.condensed matter, continuum models, electro-chemistry, materials, solvation 1 | INTRODUCTION An embedding liquid environment can substantially affect the electronic properties, geometries and spectroscopic responses of the embedded system. Dealing with this kind of effects is a difficult computational task, due to the multiscale nature of the system. A brute force approach to handle these effects would require including all of the solvent atomistic details in the quantum-mechanical (QM) calculation of the solvated system. The total number of degrees of freedom, in particular the number of electrons that need to be described in a first-principles calculation, can thus increase by orders of magnitude. As a result, a calculation that would be feasible in vacuum can easily become untreatable in solution. Moreover, statistical sampling and averaging of the results over different configurations of the liquid would be required, thus further increasing the computational cost of such an approach. These limitations have motivated the development of incredibly powerful simulation programs, [1][2][3] able to take advantage of the rapid evolution of scientific computing hardware and software, for example, by fully exploiting parallel and hybrid architectures. Regardless of its computational feasibility, the explicit description of liquid environments may also suffer from some well-known limitations of current state-of-the-art ab initio methods, in particular regarding their structural [4][5][6][7] and dielectric [8][9][10] properties.On the other hand, when looking at the properties of a solvated system, it is often the case that the detailed effects caused by solute-solvent interactions are less important than the intrinsic properties of the solute. If this is the case, one can exploit a hierarchical approach, where solvent degrees of freedom are treated with a computationally less expensive technique, but are coupled with an highly accurate description of the solvated system...

Solvent-Aware Interfaces in Continuum Solvation

“…Continuum models of solvation are very popular tools in computational chemistry [1][2][3] and are becoming more and more important in other atomistic simulations fields. 4 They allow to significantly reduce the computational cost of dealing with mobile molecular components, such as the molecules of a solvent or the ionic species in an electrolyte solution. This is done by taking advantage of the statistical nature of liquid systems to implicitly integrate out these environmental degrees of freedom.…”

Excited-state solvation structure of transition metal complexes from molecular dynamics simulations and assessment of partial atomic charge methods

“…There are two commonly used theoretical approaches for considering solvent effects: methods that treat the solvent explicitly and implicit models. In the implicit solvent treatment, also known as continuum solvation models, [18][19][20] the solvent molecules are approximated by a homogeneous medium. The polarizable continuum model (PCM) 18 is a well-known implicit solvent model.…”

Influence of surface restructuring on the activity of SrTiO3 photoelectrodes for photocatalytic hydrogen reduction

“…The SCCS model has been shown to efficiently capture the essential features of liquid-solid interface through its logarithmically definition of the solvation shell: (ρ) = exp[(ζ − sin(2πζ)/2π)ln s ], which involves the smooth switching function ζ(r) = (lnρ max − lnρ)/(lnρ max − lnρ min ) that defines the gradual dielectric transition. ρ max and ρ min denote the thresholds of the electron density that de- It has been discussed recently that the introduction of the volume-dependent energy term is unphysical for a surface system [46,47]. The cavitation energy on the other hand provides a minor improvement in the accuracy of the results [23,47], and would largely cancel out for slab setups [48].…”

“…ρ max and ρ min denote the thresholds of the electron density that de- It has been discussed recently that the introduction of the volume-dependent energy term is unphysical for a surface system [46,47]. The cavitation energy on the other hand provides a minor improvement in the accuracy of the results [23,47], and would largely cancel out for slab setups [48]. Therefore, those terms are not included in this work.…”