SPARTIAN: SPAtially-Resolved Thermodynamic IntegrAtioN
Many computational methods have been developed to calculate chemical potentials of liquids. Most of them rely on the insertion of test molecules into the system, procedure that under high density/concentration conditions becomes unfeasible. The SPARTIAN method, a direct extension of the Hamiltonian adaptive resolution simulations (H-AdResS) framework, circumvents this problem by using different models in different regions of the simulation domain. Atomistic and ideal gas representations of the system coexist in thermodynamic equilibrium such that the chemical potential difference between them can be directly computed. This approach returns accurate results for density ranges that are otherwise unaccessible to standard methods, and provides a powerful tool for the investigation of complex (bio)chemical systems and phenomena. Currently, we are using the SPARTIAN method to compute absolute free energies of various molecular solids and liquids.
Fluctuations, finite-size effects and the thermodynamic limit in computer simulations
Computer simulations exhibit finite-size effects due to periodic boundary conditions and the finite size of the system. In this context, the spatial block analysis method has been originally proposed to extrapolate thermodynamic quantities from finite-size computer simulations. Along the same lines, we have introduced an accurate and efficient method to obtain asymptotic thermodynamic properties, including chemical potentials, from small-sized molecular dynamics simulations. Moreover, the method identifies conditions in which computer simulations can be effectively considered in the thermodynamic limit.
Chemical environment effects on the shape of metal nanoparticles
Semi-empirical interatomic potentials are routinely used to investigate structural and thermodynamic
properties of a great variety of nanometric systems. To study noble and post-transition
metals, in particular, a semi-empirical potential based on the second-moment
approximation (SMA) to the tight-binding Hamiltonian captures the essential physical
information of their many-body character. In addition, SMA can be easily extended to
include extra ingredients that allow for a realistic description of metal systems under a
wide variety of conditions. Examples of such extensions include the study of MgO supported silver nanoparticles and
gold nanoparticles in an implicit-solvent model.