|
QCMP136. EPR for Biological Samples (Version 1.0)
by Frank Neese, Department of Biology, University of
Konstanz, 78434 Konstanz, Germany
The program EPR is a flexible and easy-to-use tool for
simulating complicated EPR spectra which are often met
in practice, especially in the work on biological
samples. Its operation is optimized for speed and data
handling but not for absolute accuracy. Since it is my
belief that a large percentage of the information
contained in biological EPR spectra can be extracted by
means of the first-order solution to the spin-
Hamiltonian, EPR represents a first-order solution to
the spin-Hamiltonian. As there are no first-order
effects from the nuclear Zeeman and the nuclear
quadrupole interaction, these terms are omitted.
A crucial feature of the program is its ability to
simulate the overlapping contributions from several
distinct paramagnets which are non-interacting. This
complication is very often met in practice, and it
seems that there is very little optimized software for
this purpose. A multi-component approach is also
sometimes useful in dealing with systems with zero-
field splittings which are larger than the microwave
energy, because the subspectra from the Kramers
doublets behave approximately like distinct
paramagnets.
Another feature of the program is its graphical user
interface which allows the user to exchange data very
quickly and comfortably with the program. This helps
to remove one of the more boring aspects of simulating
EPR spectra. Another advantage is that simulated and
experimental spectra can be directly compared and the
"objective" goodness of the fit can be judged by
reading the mean-square deviation between experimental
and simulated spectra from the screen.
Since one of the most important questions in simulating
EPR spectra is how to arrive at suitable spin-
Hamiltonian parameters, one should worry about a good
optimization routine. Experience shows that there is
no "golden method" for automatically fitting spin-
Hamiltonian parameters, although the SIMPLEX method is
probably most convincing in view of its stability.
Therefore I have built 5 different fitting routines
into the program, and users should judge for themselves
the most appropriate one for the particular problem.
Of course there will be situations in which none of the
fitting procedures will lead to any meaningful result.
Lines of Code: 47,506
Turbo PASCAL
| 
