Anisotropic Magnetic Spin Interactions of Transition Metal …
59
experimental data. This statement, however, only holds for the GGA functionals. The
B3LYP and PBE0 hybrid functionals severely underestimated the isotropic coupling
a iso of
14 N NO and
14 N His93 by ~10 MHz. The best values from BP86 (34, 17, 13 MHz)
and PBE (36, 18, 13 MHz) calculations for
14 N NO ,
14 N His93 and
1 H His93 agree very
well with experimental values of 38, 18 and 9 MHz, respectively [86], and allow
an assignment of the rhombic ‘R-form’ EPR spectral properties to an active site
conformation resembling ‘model 2’.
5 Conclusion
The advancement of spectroscopic techniques to investigate larger and more complex
chemical and biological systems affords detailed insight into the electronic structure
and spatial arrangement of ligands and the protein environment. Electron spin resonance spectroscopy, due to the Zeeman splitting in an external magnetic field, yields
the electronic g-tensor. Its principal values provide initial insight into the type of paramagnetic species (organic radical, spin and oxidation state of a transition metal) but
the determination of the g-tensor principal axes and their orientations in the molecular
structure reveal additional information about their directionality. Very often, the computation of EPR parameters solely focusses on reproducing experimental g-values
or an isotropic g iso . Despite recent progress in the development of computational
approaches to include electron correlation and spin–orbit coupling as accurately as
possible, it is not possible to generate computational results that perfectly match
experiment. A systematic deviation between experimental and calculated g-tensor
principal values is still apparent but its magnitude and whether calculations are overor underestimating experiment cannot be stated in general. Additional confidence
into computational results can be gained when also the orientation of the magnetic
axes is compared with experiment. Current computational methods are able to reproduce well the orientation of magnetic principal axes for g-tensors, hyperfine tensors
and ZFS tensors in a molecular geometry.
Additional interactions between the electron spin and the nuclear spin are a local
probe of the distribution of the unpaired electron spin. The isotropic hyperfine coupling interaction yields information about the (non-uniform) spin density distribution; the scalar a iso is a direct probe of the unpaired spin density at the nucleus and
of conformational states of ligands and amino acids close to the spin center. The
directional hyperfine interactions also probe the orientation and distance of nearby
interacting partners, for example protons. This information is complementary to Xray crystallography. In large systems with multiple paramagnetic sites or large spatial
extension, the analysis and interpretation of electron-nuclear interactions is difficult
and often not unambiguous. Current computational methods are able to give detailed
insight into the unpaired electron spin distribution and produce hyperfine couplings
(isotropic and dipolar) in very good agreement with experiment and thus assist the
analysis and interpretation of magnetic resonance spectra from frozen solution or
crystalline samples.
59
experimental data. This statement, however, only holds for the GGA functionals. The
B3LYP and PBE0 hybrid functionals severely underestimated the isotropic coupling
a iso of
14 N NO and
14 N His93 by ~10 MHz. The best values from BP86 (34, 17, 13 MHz)
and PBE (36, 18, 13 MHz) calculations for
14 N NO ,
14 N His93 and
1 H His93 agree very
well with experimental values of 38, 18 and 9 MHz, respectively [86], and allow
an assignment of the rhombic ‘R-form’ EPR spectral properties to an active site
conformation resembling ‘model 2’.
5 Conclusion
The advancement of spectroscopic techniques to investigate larger and more complex
chemical and biological systems affords detailed insight into the electronic structure
and spatial arrangement of ligands and the protein environment. Electron spin resonance spectroscopy, due to the Zeeman splitting in an external magnetic field, yields
the electronic g-tensor. Its principal values provide initial insight into the type of paramagnetic species (organic radical, spin and oxidation state of a transition metal) but
the determination of the g-tensor principal axes and their orientations in the molecular
structure reveal additional information about their directionality. Very often, the computation of EPR parameters solely focusses on reproducing experimental g-values
or an isotropic g iso . Despite recent progress in the development of computational
approaches to include electron correlation and spin–orbit coupling as accurately as
possible, it is not possible to generate computational results that perfectly match
experiment. A systematic deviation between experimental and calculated g-tensor
principal values is still apparent but its magnitude and whether calculations are overor underestimating experiment cannot be stated in general. Additional confidence
into computational results can be gained when also the orientation of the magnetic
axes is compared with experiment. Current computational methods are able to reproduce well the orientation of magnetic principal axes for g-tensors, hyperfine tensors
and ZFS tensors in a molecular geometry.
Additional interactions between the electron spin and the nuclear spin are a local
probe of the distribution of the unpaired electron spin. The isotropic hyperfine coupling interaction yields information about the (non-uniform) spin density distribution; the scalar a iso is a direct probe of the unpaired spin density at the nucleus and
of conformational states of ligands and amino acids close to the spin center. The
directional hyperfine interactions also probe the orientation and distance of nearby
interacting partners, for example protons. This information is complementary to Xray crystallography. In large systems with multiple paramagnetic sites or large spatial
extension, the analysis and interpretation of electron-nuclear interactions is difficult
and often not unambiguous. Current computational methods are able to give detailed
insight into the unpaired electron spin distribution and produce hyperfine couplings
(isotropic and dipolar) in very good agreement with experiment and thus assist the
analysis and interpretation of magnetic resonance spectra from frozen solution or
crystalline samples.