Back to qct and beyond

Are quasiclassical trajectories (QCT) still useful for molecular dynamics applied to chemical kinetics modeling? The question is far from being purely academic. Nowadays astrophysical, technological and environmental challenges require complex modellisation of molecular processes (1), with hundreds or thousands of elementary processes involved. Model accuracy strictly depends on quantitative knowledge of all these detailed input data (2). In principle, dealing with molecules in collisions, a quantum mechanical (QM) method should be used for accurate calculations of processes such as vibrational energy transfer, dissociation, etc. Computational efficient codes exist to perform accurate QM calculations, exploiting the power of modern highly parallel supercomputers. So, why to worry about less accurate methods such as QCT? The point is that in real life there are also many limitations in QM calculations. Normally these limitations are overcome by means of approximations that can drastically change the reliability of results, in a way that is not easy to ascertain. As a consequence, it is possible to find examples of approximate QM calculations much worse than an accurately calculated QCT result. The limitations of QCT, on the contrary, are easier to know. By accurately study these limitations (3,4), one can "extract" the correct contributions of QCT and then find the missing complementary contributions in other semiclassical methods, exploiting a criterion based on the behavior of trajectories along the collision. This fusion of quasi- and semi-classical methods, which is expected to maximize both computational efficiency and accuracy, may be performed at the level of cross sections or, better but harder, at the level of trajectories. Some partial results and considerations about the potential applications will be presented.