Analyzing signal attenuation in PFG anomalous diffusion via a non-Gaussian phase distribution approximation approach by fractional derivatives

Abstract: Anomalous diffusion exists widely in polymer and biological systems. Pulsed field gradient (PFG) techniques have been increasingly used to study anomalous diffusion in NMR and MRI. However, the interpretation of PFG anomalous diffusion is complicated. Moreover, there is not an exact signal attenuation expression based on fractional derivatives for PFG anomalous diffusion, which includes the finite gradient pulse width effect. In this paper, a new method, a Mainardi-Luchko-Pagnini (MLP) phase distribution approximation, is proposed to describe PFG fractional diffusion. MLP phase distribution is a non-Gaussian phase distribution. From the fractional diffusion equation based on fractional derivatives in both real space and phase space, the obtained probability distribution function is a MLP distribution. The MLP distribution leads to a Mittag-Leffler function based PFG signal attenuation rather than the exponential or stretched exponential attenuation that is obtained from a Gaussian phase distribution (GPD) under a short gradient pulse approximation. The MLP phase distribution approximation is employed to get a complete signal attenuation expression E{\alpha}(-Dfb*{\alpha},\b{eta}) that includes the finite gradient pulse width effect for all three types of PFG fractional diffusion. The results obtained in this study are in agreement with the results from the literature. These results provide a new, convenient approximation formalism to interpret complex PFG fractional diffusion experiments.