Professor Dr. Franz Michael Dolg  


Tel: (+49) 221 470 6893 Fax: (+49) 221 470 6896 Email: m.dolg@unikoeln.de Institute Homepage 

Theoretical Chemistry Relativistic quantum chemistry, development of relativistic energyconsistent ab initio pseudopotentials, electronic structure of compounds with heavy and superheavy elements, relativistic effects and electron correlation effects, chemistry of lanthanides and actinides, electronic structure and sizedependent properties of metal clusters, development of quantum chemical ab initio methods for periodic systems, molecular dynamics for amorphous systems A central research topic of the group is relativistic quantum chemistry, i.e., the application of quantum chemical methods for electronic structure calculations to systems with heavy and superheavy atoms. Relativistic effects in atoms (and as a consequence also in molecules) increase roughly with the fourth power of the nuclear charge. For the firstrow transition metals relativistic effects cannot be neglected in quantitative investigations, for heavy and superheavy elements they may even change the qualitative behavior. For example, the nonrelativistic ground state configurations 5f^{2}7s^{2} of _{90}Th and 6d^{10}7s^{1} of _{111}EkaAu are changed due to the relativistic destabilisation of the f respectively dshell to 6d^{2}7s^{2} and 6d^{9}7s^{2}. The computational effort of quantum chemical calculations also increases drastically with the number of electrons. A central topic is therefore the development of relativistic energyconsistent ab initio pseudopotentials (effective core potentials) for all elements of the periodic table (with exception of H and He). The pseudopotential approach allows to incorporate implicitly the major relativistic contributions for the valence electron system with only little loss of accuracy and without explicitly taking the core electrons into account. The valence electron system is treated essentially within a quasinonrelativistic formalism. The pseudopotentials and corresponding optimized valence basis sets developed by the group in cooperation with H. Stoll (University of Stuttgart) are available in the libraries of many quantum chemical standard program packages (e.g. GAUSSIAN, MOLCAS, MOLPRO, TURBOMOLE) as well as some solid state computer codes (e.g. CRYSTAL). New methods of adjustment, which came available a few years ago, are currently used to further improve the accuracy of the approach. Systems with open d or even fshells still pose a considerable challenge to quantum chemical ab initio methods. During the last 15 years we investigated in detail the electronic structure of several lanthanide and actinide compounds. In addition to relativistic effects also electron correlation contributions play a crucial role and often make quantitative investigations quite demanding. Besides the study and prediction of properties of numerous diatomic molecules (e.g. 18 unpaired electrons in the ^{19}S_{g}^{} ground state of Gd_{2}, experimentally confirmed) we also investigated several organometallic compounds and sometimes found surprising results (e.g. a 4f^{1}p^{3} (ca. 80%) + 4f^{0}p^{4} (ca. 20%) ^{1}A_{1g} ground state of cerocene; experimentally confirmed). A new field of research is the development of quantum chemical ab initio methods for periodic systems, i.e., correlated electronic structure calculations for crystalline solids and polymers. Here the accurate treatment of electron correlation effects is the main challenge. Similar to the modern methods for large molecules, this can be achieved in the basis of localized orbitals. However, in contrast to finite large molecules, the correlation contributions to an infinite solid cannot be determined simply within a single calculation, but rather have to be extracted from a series of calculations for appropriate subsystems summed up to the correlation energy per unit cell using an incremental exansion. Using this approach we could perform in collaboration with H. Stoll (University of Stuttgart) the first full configuration interaction calculations for a simple crystalline solid (LiH). Selected publications: 1. "Electronic Structure Calculations for Molecules Containing Lanthanide Atoms" M. Dolg and H. Stoll in: Handbook of Chemistry and Physics of Rare Earths, Vol. 22, Ch. 152, Ed. L. Eyring, Elsevier, Amsterdam, 1996, 607  729. 2. "Lanthanides and Actinides" M. Dolg in: Encyclopedia of Computational Chemistry, Eds. P.v.R. Schleyer, N. L. Allinger, T. Clark, J. Gasteiger, P. A. Kollman, H. F. Schaefer III, P. R. Schreiner, Wiley, Chichester, 1998, 1478  1486. 3. "Effective Core Potentials" M. Dolg in: Modern Methods and Algorithms of Quantum Chemistry, NIC Series 1, Ed. J. Grotendorst, John Neumann Institute for Computing, 2000, 479  508; Neuauflage NIC Series 3, 2000, 507  540. 4. "Relativistic Energyconsistent Pseudopotentials  Recent Developments" H. Stoll, B. Metz, and M. Dolg, J. Comput. Chem. 2002, 23, 767  778. 5. "Quantum Monte Carlo Study of Mercury Clusters" H.J. Flad, F. Schautz, and M. Dolg in: Recent Advances in Quantum Monte Carlo Methods, Part II, Eds. J. B. Lester Jr., S. M. Rothstein, S. Tanaka, World Scientific, New Jersey, 2002, 183  202. Fellowships/Awards/Other Activities: award of the friends of the University of Stuttgart, 1987 
