The analysis of dielectric relaxation phenomena with the inverse Fourier transformation

1993

Abstract A method to determine the distribution of relaxation times directly from dielectric loss spectra is presented. The method is based upon a deconvolution procedure: the Fourier transform of the loss factor is divided by sech( π 2 f ) and then, via an inverse Fourier transformation, transformed into the time domain. Limitations and possible improvements of the method are discussed. It is shown that the present method is able to reveal local relaxation processes not perceptible in the loss factor spectrum. With stimulated noise-free data, the resolution of the method is one third of a decade on a logarithmic relaxation timescale.