AB INITIO THEORY OF POINT DEFECTS IN SiO2

Abstract

It is conceivable that in the near future one will assist to the rapid development of the area of research which is generally named ``scientific computing''. This relatively new field will complement the more classical experimental approach for the solution of scientific problems. Scientific computing should not be confused with the more general areas of theoretical chemistry and theoretical physics. Experiments produce new facts and discoveries, and disclose to humans secrets of nature; the role of theory is to provide a general framework for the explanation of the phenomena observed through a series of mathematical laws. Scientific computing is something in between: it is based on well established theories and formalisms but it is used to provide new facts, in a way which becomes more and more similar to the way experiments are performed. The reason why scientific computing will continuously and rapidly increase its importance in the future is clear. In the past 30 years we have assisted to an exponential increase of the computing power and this has stimulated the creation of new software to solve problems in different scientific areas. So far, the development of the software has been slower than that of the computing power, a trend which could change in the next decade. Thanks to the combined use of advanced software and powerful computers, it is nowadays possible (at least in principle) to ``simulate'' an experiment on a computer before to perform it with a lower cost and a more rapid answer. Of course, the extent of this ``revolution'' will entirely depend on the degree of reliability of the simulation, which in turns is a function of the generality of the underlying theory, of the level of approximations introduced in the simulation, of the accuracy of the software and of the amount of computing power available. In 1929 P.A.M. Dirac stated that a large part of physics and the whole chemistry can be deduced from quantum mechanics [1]. The 1998 Nobel Award to John Pople and Walter Kohn for their contribution to the development of quantum chemistry and density functional theory, respectively, represents a direct recognition of the practical importance of modern ab initio calculations.