MOLCAS manual [4] describes standard procedures for regular runs. They are simple and efficient. However, we prefer to follow our own procedures in order to perform the calculations on the excited manifolds of lanthanides in solids. We have developed them in the course of the years of calculations and we feel comfortable with them. More importantly, with them, we have minimized the number of errors we do in the calculations. In our calculations, we have to handle a large number of excited states, and they need to be analyzed and labelled according to their irreducible representations (irreps) in symmetry groups of higher symmetry than the one actually used in the calculations. E.g., it is normal to carry on the calculations of an octahedral system using actual D2h symmetry because of the limitations of MOLCAS, which means that states of different Oh irreps will be together in the same D2h irrep and they need to be identified. Often, it is very hard, if not impossible, to go on to the next step in the calculations without having completed the symmetry analysis in the previous step. Also, we usually need potential energy curves of many states along a geometry variable. And there is a vast amount of information generated during the calculations that has to be very well organized in order to make all the analyses feasible. In summary, we will use in this tutorial a particular structure of subdirectories in the project directory and we will use script shells for analysis of the results of a given step and for input preparation of the next step. This is by no means the only way of doing these calculations, but it is af efficient way that we have been using, developing, and debugging for years. We will stick to it in this document. The reader may find it more cumbersome than necessary in the early stages in the calculations, but may also find it very helpful in the final steps.
We assume we start from a project directory in which we normally define an active cluster, its embedding potential, its symmetry, its basis set, and the restricted or complete active space that will be used. Its name is optional and should be meaningful. The following sub- directories will be used and should be created from the project directory: ./shells, ./inputs, ./inputs/CASPT2, ./inputs/RASSI, ./libs, ./printouts, ./printouts/CASPT2, ./printouts/RASSI, and ./vectors, ./vectors/CASPT2. It is convenient to use also ./results. We will refer to these as the shells, inputs, libs, printouts, vectors, results, etc. directories regardless of the directory we might be located at the time of reference. Depending on the user style of work, other subdirectories can be created, although they will not be used by the scritp shells.
We will use here the project directory names BaF2-Ce and BaF2-Pr for the target calculations described in Sections 3 and 4, respectively; project directories BaF2-Ce-completed and BaF2- Pr-completed will be made available for checking results, for help, or for getting data/files.Â