First-Principle Simulation in Materials Science

It is known that the human knowledge deserves the name of Science depending on the role played there by the number 

We already know that the first-principle models involve no adjustable or experimentally-derived parameters. The simulation technique allows one to solve many actual problems in materials science, solid-state chemistry and metallurgy. Density functional theory was employed successfully in recent years because theorists performed total-energy calculations using the exchange-correlation potentials and showed that they reproduced a variety of ground-state properties only deviating a few percent from experimental data. Thus, the acceptance of local approximations to density functional theory has only emerged after many successful applications to many types of materials and systems. 

The efficiency of these methods may be demonstrated by the following convincing examples. 

---

Abstract SIESTA was developed as an approach to compute the electronic properties and perform atomistic simulations of complex materials from first principles. Very large systems, with an unprecedented number of atoms, can be studied while keeping the computational cost at a reasonable level. The SIESTA code is fi-eely available for the academic community (http //www.uam.es/siesta), and this has made it a widely used tool for the study of materials. It has been applied to a large variety of systems including surfaces, adsorbates, nanotubes, nanoclusters, biological molecules, amorphous semiconductors, ferroelectric films, low-dimensional metals, etc. Here we present a thorough review of the applications in materials science to date. 

Recent dramatic advances in computational techniques and computer power have enabled us to simulate crystalline structures from first-principles by means of the electronic structure calculation of the whole system within the density functional theory. Even liquid and vitreous silica have come to be studied by the ab initio MD method or so-called Car-Parrinello method [59]. Thus the application of the classical MD method is to be shifted to study of dynamics with a larger system size and longer simulation time. For example, the simulation of the oxygen diffusivity mentioned in the previous section needs accumulation of positions of five hundred atoms over 120 ps at each pressure, for which the ab initio MD is too inefficient. On the other hand, a local structural deformation relevant for the diffusion could be simulated with a smaller cell and a shorter time scale. It is obviously fruitful to make proper use ofthese approaches, i.e. the classical MD supported by first-principles cluster calculations and the ab initio MD, in each problem of materials science. 

If computational power permits, first-principles molecular modeling such as ab initio MD can provide the most accurate information without adjustable parameters. However, even the most powerful computer system available today is limited to first-principles solution of molecular information with a certain size. As a result, the hybrid methods such as QM/MM or ONIOM (where different levels of theory for one calculation are involved) are often used (Fig. 1). As an example of future prospects, molecular design and characterization of nanocomposites using computer simulation are also briefly mentioned. Atomistic modeling of clay nanocomposites is a very promising field in clay mineral materials science. 

The fact that the CPMD was a milestone step forward in realistic simulations of materials, at various thermodynamics conditions, can be easily seen by the number of publications in first principles molecular dynamics (FPMD) before and after 1985, i.e. after the original CPMD publication [21]. Indeed, the original Car-Parrinello publication has more than 6500 citations in 2014 (source ISI Web of Science), and, to acknowledge the importance of the method, the international PACS (Physics and Astronomy Classification Scheme) introduced in 1996 a new identification number, 71.15. Pd, to classify Car-Parrinello related publications. Since then, the method has been applied to a wide variety of materials, ranging from solids, to liquids and to biological systems [33]. 

The papers presented in the conference span the spectrum of activity in the science of alloys. The theoretical presentations ranged in content from fundamental studies of electronic structure, to first-principles calculations of phase diagrams, to the effects of charge transfer, to the temperature dependence of short-range order parameters. They encompassed the study of mechanical properties, the properties of dislocations, of phase evolution, and computer simulations. Experimental studies were presented based on a variety of state of the art experimental techniques, from TEM to synchrotron diffraction. The phenomena studied varied from the precipitation of nitrides in steel, to the wetting of interfaces between two different crystal structures, to the ordering of vacancies in carbides. And the materials whose properties were measured ranged from Transition metals, to the Lanthanides, to the Actinide series of compounds and alloys. 

As a useful supplement to the experimental techniques and a powerful tool in the areas of chemistry, physics and materials science, computational simulations and calculations based on first principles electronic structure calculations have gone through fast-growing development in the last couple of decades. Under the help of more and more powerful computing facilities as well as efficient computational theories and algorithms, it is now a standard approach to execute quantum chemical calculations in elucidating the structure-reactivity relationship of the systems ranging from small functional molecules to complicated solid composites such as metal oxides and even supported nanoparticles for catalytic use. 

---