Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] aema_v1_0.tar.gz(4859 Kbytes)
Manuscript Title: The RATIP program for relativistic calculations of atomic transition, ionization and recombination properties
Authors: S. Fritzsche
Program title: RATIP
Catalogue identifier: AEMA_v1_0
Distribution format: tar.gz
Journal reference: Comput. Phys. Commun. 183(2012)1525
Programming language: ANSI standard Fortran 90/95 and subsequent developments.
Computer: PCs and workstations.
Operating system: Suse, Debian and Ubuntu Linux.
RAM: Memory requirements strongly depend on the size of the bound-state wave functions, the property considered as well as the special features selected during the computations.
Word size: All real variables are parametrized by a selected kind parameter and, thus, can easily be adapted to any required precision as supported by the compiler. Presently, the kind parameter is set to double precision (two 32-bit words) in the module rabs_constant.
Keywords: Alignment of ionic states, Angular anisotropy parameters, Atomic, Auger amplitudes and properties, Breit interaction, Configuration interaction, Dirac-Coulomb-Breit Hamiltonian, Einstein coefficients, Gaunt interaction, Hyperfine quenching, Isotope-shift parameters, Large-scale computations, Multiconfiguration Dirac-Fock, Nuclear-size effects, QED estimate, Photoionization amplitudes and cross sections, Radiative and dielectronic recombination, Relativistic, Toolbox for atomic data and wave function manipulations, Transition probability, Transverse interaction.
Classification: 2.1, 2.9.

Subprograms used:
Cat Id Title Reference
ADCU_v1_0 GRASP92 CPC 94(1996)249

Nature of problem:
Ab-initio calculations of atomic properties and data are required in science and technology, not just within the traditional areas of astro and plasma physics but also in several recently emerging research fields. Hereby, often quite different demands arise with regard to the accuracy of the data, the elements of interest as well as their stage of ionization. Therefore, it is desirable to provide a code which is applicable to all elements of the periodic table and which can help incorporate the dominant electron-electron correlation and relativistic effects on equal footings into the computations.

Solution method:
Atomic bound-state wave functions from GRASP92 [1] for different levels and charge states are combined with continuum orbitals to calculate many-electron transition amplitudes and properties as derived from these amplitudes. Three major types of transition amplitudes refer to the electron-electron interaction, based on the Dirac-Coulomb-Breit Hamiltonian, the electron-photon interaction for the coupling of atoms to the radiation field as well as the electron-nucleus (hyperfine) interaction due to the electric and magnetic multipole fields of the nucleus. Apart from the electric-dipole approximation to the electron-photon interaction, this includes also other -- electric and magnetic -- multipole components of the radiation field. All computations are performed within the framework of the multiconfiguration Dirac-Fock (MCDF) method as implemented in GRASP92 [1] and its recent successors [2].

Relativistic calculations of atomic properties are restricted mainly by the size of the wave functions and the (virtual) excitations that can be taken into account with regard to a given set of reference configurations. Further restrictions of the present implementation concern:
  • Despite the relativistic formulation of atomic properties based on Dirac's equation, all calculations are performed within the no-pair approximation; no attempt has been made to incorporate contributions from the negative continuum or radiative corrections beyond some simple estimate of the vacuum polarization and the electron self-energy to the level energies.
  • Continuum orbitals are always generated within a static potential (of the corresponding ionic core) and are utilized to construct distorted waves with well-defined total angular momentum and parity. No continuum (interchannel) interactions are taken into account in the construction of scattering states if one (or more) electrons is in the continuum.
  • As in GRASP92 [1], antisymmetric subshell states with more than two equivalent electrons are supported only for j ≤ 9/2.
  • If wave functions are defined with regard to different configuration lists to represent, for example, the initial and final state of a selected photo- or autoionizing transition, the same order of atomic orbitals (and usually also the same core) has to be used for generating the atomic bound states. The program terminates with an error message if this is not the case.
  • The use of non-orthogonal orbital sets for the representation of initial, intermediate or final atomic states is supported only by a few selected programs, while "orthogonality" is assumed otherwise for the evaluation of the many-electron amplitudes apart from the active electrons.

Unusual features:
The RATIP program is designed as a suite of programs where each of them help calculate one or a few closely related atomic properties, and for a given set of atomic levels. To make use of these programs, it is usually assumed that the wave functions for all bound states have been generated before by means of the GRASP92 [1] or some equivalent code. However, a clear and simple interface is made between the computation of the bound states and their use within the RATIP program [3] by applying only the (standard) input and output files from GRASP92, such as the definition of nuclear parameters (.iso), configuration lists (.csl), radial orbitals (.rwf) and mixing coefficient (.mix) files.

To specify the bound states of interest, most calculations within the RATIP program refer to the level numbers as they (do) occur in GRASP92 for a given configuration basis. Care has been taken that this selection and reference to the atomic levels can be handled flexibly but with some proper tests on the atomic property under consideration. Each program component of RATIP is controlled by an interactive dialog at the beginning of its execution and enables the user to select individual transitions as well as the particular mode of computation. All major results are usually compiled in tables and printed to some summary file, which is specific to each component. The units of energies, rates and cross sections in these tabulations can be specified during the input (from a number of possible choices) if the default is considered not to be appropriate.

Various (modern design) principles of Fortran 90/95 have been applied in developing the RATIP code [4], including the use of modules, the definition of derived data structures, the use of logical flags and the dynamic allocation of all important arrays. Therefore, there are no serious restrictions with regard to the number of open shells, nor to the grid size or the number of atomic transitions that can be calculated within a single run of some component. While some of RATIP's code is common to all programs and is provided by a number of core modules, each component usually refers also to some own(ed) data structures and procedures which are specific to its application.

Running time:
20 minutes on a standard laptop for all test cases.

[1] F. A. Parpia, C. F. Fischer, and I.P. Grant, Comput. Phys. Commun. 94 (1996) 249.
[2] P. Jönsson, X. He, C. Froese Fischer and I. P. Grant, Comput. Phys. Commun. 177 (2007) 597.
[3] S. Fritzsche, J. Elec. Spec. Rel. Phen. 114-116 (2001) 1155.
[4] M. Metcalf and J. Reid, Fortran 90/95 Explained (Oxford University Press, 1996).