On Lévy (or stable) distributions and the Williams-Watts model of dielectric relaxation (Q1079890)

This paper is concerned with the Lévy, or stable distribution function defined by the Fourier transform \[ Q_{\alpha}(z)=\frac{1}{2\pi}\int^{\infty}_{-\infty}\exp (-izu-| u|^{\alpha})du\quad with\quad 0<\alpha \leq 2. \] When \(\alpha =2\) it becomes the Gauss distribution function and when \(\alpha =1\), the Cauchy distribution. When \(\alpha\) \(\neq 2\) the distribution has a long inverse power tail \(Q_{\alpha}(z)\sim \frac{\Gamma (1+\alpha)\sin \pi \alpha /2}{\pi | z|^{1+\alpha}}.\) In the regime of small \(\alpha\), if \(\alpha\) \(| \log z| \ll 1\), the distribution is mimicked by a log normal distribution. We have derived rapidly converging algorithms for the numerical calculation of \(Q_{\alpha}(z)\) for various \(\alpha\) in the range \(0<\alpha <1\). The function \(Q_{\alpha}(z)\) appears naturally in the Williams-Watts model of dielectric relaxation. In that model one expresses the normalized dielectric parameter as \[ \epsilon_ n(\omega)\equiv \epsilon_ n'(\omega)-i\epsilon_ n''(\omega)=-\int^{\infty}_{0}e^{-i\omega t}[d\phi (t)/dt]dt \] with \(\phi (t)=\exp -(t/\tau)^{\alpha}\). It has been found empirically by various authors that observed dielectric parameters of a wide variety of materials of a broad range of frequencies are fitted remarkably accurately by using this form of \(\phi\) (t). \(\epsilon_ n''(\omega)\) is shown to be directly related to \(Q_{\alpha}(z)\). It is also shown that if the Williams-Watts exponential is expressed as a weighted average of exponential relaxation functions \[ \exp - (t/\tau)^{\alpha}=\int^{\infty}_{0}g(\lambda,\alpha)e^{-\lambda t}dt \] the weight function g(\(\lambda\),\(\alpha)\) is expressible as a stable distribution. Some suggestions are made about physical models that might lead to the Williams-Watts form of \(\phi\) (t).