About Us

Background

Research Focus

People

Repository

Workshops

Projects

Links

Contact MRI

CCS: Institutes: MRI: Background: Scientific Rationale

• Introduction
• Scientific Rationale
• HPC Rationale
• Thrust of CCS-MRI
• Resources of CCS
• Next Steps
• Enabling National Science Goals

Scientific Rationale

The primary goal of CMS is to develop the capability to reliably predict the properties of real materials and to provide a framework in which complex phenomena can be better understood. In this respect CMS has progressed rapidly over the last two decades and is an integral component - along with experiment and theory - of modern materials research programs. However, it is becoming increasingly clear that, for CMS to achieve its full impact on nano-science and materials research and development, new approaches will have to be developed. This is necessitated by three features of computational materials science that distinguish it from many other branches of computational science:

1. In order to realistically describe real materials it is necessary to simulate phenomena over a vast range of time and length scales. A corollary of which is the use of a wide range of modeling techniques.


2. At each length and time scale the relevant mathematical models used to describe phenomena are under constant development and refinement, driven by new theoretical advances, innovations in the mathematical and computational techniques used to solve the underlying equations, and increased computational capability and capacity.


3. Important codes have been developed by small, often single investigator groups, to satisfy their own scientific interests and for their own use.

Collectively, these features of CMS result in a situation where locally developed monolithic codes are used to solve locally defined problems usually of modest scale, often described as a cottage industry approach. In the past, this modus operandi has served the CMS community well, in that it reflects the fact that CMS science was made up of many small problems rather than one (or a few) overarching problem(s) (c.f. fundamental particle physics). This lack of over-arching problems does not however mean that are not unifying themes that can be clearly identified This is particularly true of the theoretical and computational tools that are needed to address key issues in nanoscience and technology.

Quantum molecular dynamics simulations show that nanotubes initiate breakage by a bond rotation, where a pair of atoms rotates about the center of their bond and converts four hexagons (highlighted in red) into a 5-7-7-5 defect. The barrier for this rotation is very high, which further increases the exceptional strength of nanotubes. [M. Buongiorno Nardelli, B. I. Yakobson and J. Bernholc, Phys. Rev. B 57, R4277 (1998).]

Firstly, the requirements of the nano-scale lead to the realization that CMS is evolving in the direction of problems whose solution demand application of methods covering simultaneously multiple length and time scales, multiple (sometimes new) models, and a breadth of expertise not resident in a single researcher. Thus, there is a particular need to develop interfaces between models so that information can be transferred seamlessly between different length and time scales.

A second feature of nanoscale systems is the increased significance of correlations and disorder attendant to reduced dimensionality. Thus, the familiar paradigms that underpin the first principles electronic structure methods that have revolutionized our understanding of bulk materials begin to fail in lower dimensional systems. Therefore, computationally challenging many-body methods need to be combined with modern electronic structure methods. This is yet another example where expertise resident in different camps of the CMS community is needed to meet the challenging requirements of nano-scale materials science.

When nanosystems are subjected to driving forces it is likely that an additional familiar paradigm that of linear response, will be unequal to the task, in that reduced dimensionally it is much easier to drive systems far from equilibrium. For example, when nano-structured materials are used in electronic devices, current densities become so large that conventional linear response theory no longer applies. Non-equilibrium theories of electronic transport will have to be developed and applied, a task which represents one of the biggest challenges in computational materials science, and which will require expertise from many different areas of condensed matter theory, materials science, chemistry, and mathematics. However, the potential rewards will be enormous. Indeed, recent developments indicate that not only the charge but also the spin of the electron will be actively controlled, heralding the era of spintronics.