Link to this article: http://journals.cambridge.org/abstract_S096249290400025XHow to cite this article: Claude Le Bris (2005). Computational chemistry from the perspective of numerical analysis.We present the field of computational chemistry from the standpoint of numerical analysis. We introduce the most commonly used models and comment on their applicability. We briefly outline the results of mathematical analysis and then mostly concentrate on the main issues raised by numerical simulations. A special emphasis is laid on recent results in numerical analysis, recent developments of new methods and challenging open issues.
CONTENTS1 Outline 364 2 A short introduction to modelling for molecular simulation 365 3 Discretization of molecular models 376 4 The condensed phase 408 5 Time-dependent problems 411 6 Current and future trends 430 References 431 364 C. Le Bris Computational chemistry 365decade, but a lot remains to be done. We conjecture that a better theoretical knowledge of existing techniques will lead to their enhancement, and will therefore improve their applicability, as demonstrated by the history of scientific computing in the engineering sciences. The purpose of the present article is to overview the numerical difficulties encountered in computational chemistry, to show how they are solved within the current state of the art, and to indicate the needs for further improvements. Of course, whenever they exist, we will indicate the results of numerical analysis that help to give a sound grounding for these techniques. But most often the article will deal with descriptions of techniques rather than with statements of theorems. We hope that this bias will stimulate further works.The article is organized in a rather narrative way, without any ordering of scientific priority and/or importance. We begin with a description of the models and the discretization techniques for small systems in a static picture. Then, we progressively and (hopefully) pedagogically proceed to the modelling and simulation of more difficult situations: larger systems, systems in situ, time-dependent settings. Of course, as the size of the system increases, the models need to be coarse-grained, otherwise they cannot be tackled in practice. Therefore, the article also proceeds from the finest models to the coarsest ones.The present state-of-the-art survey focuses on issues in numerical analysis for a readership familiar with such questions in other settings. A more detailed description may be found in the book of Le Bris, ed. (2003), and also, to a smaller extent, in the proceedings volume by Defranceschi and Le Bris, eds (2000). For readers with a background in chemistry, we refer to two survey articles, Defranceschi and Le Bris (1997) and Defranceschi and Le Bris (1999), for an introduction to the mathematical and numerical analysis. On the other hand, questions related to the mathematical analysis of models are overviewed in Le Bris and Lions (2005).Computational chemistry 367 state of minimal energy) of the system under consider...