Post-processing of EPR spectra by convolution filtering: calculation of a harmonics' series and automatic separation of fast-motion components from spin-label EPR spectra

This communication reports on post-processing of continuous wave EPR spectra by a digital convolution with filter functions that are subjected to differentiation or the Kramers-Kr?nig transform analytically. In case of differentiation, such a procedure improves spectral resolution in the higher harmonics enhancing the relative amplitude of sharp spectral features over the broad lines. At the same time high-frequency noise is suppressed through filtering. These features are illustrated on an example of a Lorentzian filter function that has a principal advantage of adding a known magnitude of homogeneous broadening to the spectral shapes. Such spectral distortion could be easily and accurately accounted for in the consequent least-squares data modeling. Application examples include calculation of higher harmonics from pure absorption echo-detected EPR spectra and resolving small hyperfine coupling that are unnoticeable in conventional first derivative EPR spectra. Another example involves speedy and automatic separation of fast and broad slow-motion components from spin-label EPR spectra without explicit simulation of the slow motion spectrum. The method is illustrated on examples of X-band EPR spectra of partially aggregated membrane peptides.