# Research interests

## Many-body methods in electron structure calculations of excited molecules: General

The central topic of our research activities is the development and application of Green's
function or* propagator methods* for the treatment of (generalized) electronic excitations in
molecules, comprising both (neutral) excitation as well as electron removal (ionization)
and electron attachment processes. The propagator methods allow for a direct
determination of excitation (or ionization/electron attachment) energies and spectral
intensities, thus circumventing the necessity of deriving those quantities from separate
computations for the ground and the excited states.

For a long time we have been pursuing a general approach referred to as* algebraic-
diagrammatic construction* (ADC) which allows us to derive in a systematical way higher-
order approximation (ADC(*n*)) schemes for the respective propagators and
their physical information. The ADC approach is based on a specific reformulation of the
respective propagator referred to as *intermediate state representation* (ISR). The
comparison of the (algebraic) ISR representation (ADC form) with the original
diagrammatic perturbation expansion of the propagator through a given order
*n* of perturbation theory allows one to determine successively the ingredients
of the ISR representation, that is, ADC secular matrix elements and (effective) transition
amplitudes. The computational concept of the resulting ADC approximations is simple: It
combines the eigenvalue problem of a hermitian secular (ADC) matrix with perturbation
theory for the matrix elements. The eigenvalues are the excitation energies, while the
transition moments are derived from the corresponding eigenvectors and the effective
transition amplitudes.

The usefulness of the ADC methods is based on three basic properties: the perturbation
expansions obtained for the secular matrix elements and effective amplitudes are
*regular*, that is, they behave essentially like the Rayleigh-Schrödinger
series for the ground-state energy and wave-function, respectively. The explicit
configuration space (dimension) of the ADC secular problem needed for a certain level of
approximation is smaller than those of a comparable configuration interaction (CI)
treatment (*compactness*). The ADC secular equations are
*separable*, that is, local excitations in a system of non-interacting fragments
are strictly decoupled from non-local excitations. As a consequence of the separability
property, the ADC results both for the energies and transition moments are size-consistent
(size-intensive), which is a crucial condition for the applicability of the methods to large
molecules.

The present status of the ADC methods for electronic excitation and ionization is briefly
sketched in the paragraphs below.

************ Sorry, the following part is still under construction* *****************

## Electronic excitations in molecules

An adequate theoretical description of electronic excitations in molecules is a basic
requirement for the understanding of photophysical and photochemical processes. In spite
of the ongoing growth of computer capabilities and impressive methodological advances,
the present state of the art here is much less satisfactory than in the treatment of the ground
state, and it is still an urgent concern to improve existing computational schemes or to
develop new methods.

## Ionization/electron attachment

## Intermediate state representations

## K-shell excitation and ionization of molecules