Foreword

Y es, there are these strange coincidences. At the time when we are requested to write a short foreword to this special issue of the Journal of Computational Chemistry devoted to problems of the solid state, two different publications which are freely floating through the Internet these days as PDF files have mysteriously materialized on our desks at Aachen and Marburg. The first article of the two is a recent reprint of Enrico Clementi’s classic 1965 manuscript entitled ‘‘Ab initio Computations in Atoms and Molecules’’ which originally appeared in volume 9 of the IBM Journal of Research and Development. In this early article, Clementi summarizes the 1965 status of numerical Hartree-Fock-type calculations carried out on the big blue mainframe computers such as the IBM 7094 model, and he offers results for atoms and very small molecules such as carbon monoxide. To increase the computational precision, Clementi illustrates how to include corrections for relativity and, much more important, electronic correlation, the persistent quantumchemical challenge. One would be more than stunned to find solid-state numerical results in that great article from 1965, and their absence is due to the incredible computational complexity which did not allow such brave enterprise at that time, at least not within the ab initio domain. This message is a bit hard to take for our students who typically cannot imagine that any 1960s many-milliondollars mainframe computer would not have been able to compete with their 2008 cell phones in terms of processor speed (and, also, tawdriness of their operating systems). Three decades later, the entire situation must have radically changed, right? So, let’s have a look into another article dubbed ‘‘Chemical Industry of the Future—Technology Roadmap for Computational Chemistry’’ from the year 1999, sponsored by the Council of Chemical Research (CCR), supported by the US Department of Energy, Office of Industrial Technology, and put together by scientists from various research organizations. Just as expected, the computational power has increased by several orders of magnitude, the computations have become breathtakingly accurate such that computational chemistry eventually finds application in industrial research (e.g., drug and catalyst design, materials design, process simulation, polymer research, lubricants, coatings, and so forth)—and everything has been made possible by (a) smarter theories, (b) less cryptic software, and (c) faster computers. Computational chemistry has succeeded in the end! Isn’t that fantastic? But what happened to the solid state? In this very technology-roadmap white paper, the solid state is covered as follows: ‘‘For solid-state computations, codes based on Hartree-Fock or density functional theory are available but are slow and difficult to use. Chemistry applications in the solid state are currently not well integrated with the solid-state physics community. Calculations for solid state systems cannot yet be done with chemical accuracy, especially for chemically reacting systems.’’ Oops, that doesn’t sound like a success story to us. But is that really so, at least another decade later, in the year 2008? Without any disrespect against the authors of the above quote, all three statements are highly debatable. On the contrary, the typical solid-state computational chemist would emphasize the maturity of his/her tools— let’s not forget that the incredibly powerful approach of density-functional theory (DFT) originated from the solid, simply because Hartree-Fock-related technical difficulties for periodic systems had to be overcome, and nowadays