Inversion of simulated positron annihilation lifetime spectra by moving boundary subspaces

Abstract

The retrieval of the density function for the positron annihilation lifetime spectrum is obtained from simulated data using the damped singular value decomposition. Two filters factors were discussed, for noise and noiseless data, and under the L curve criterion. The obtained density function is exact in the absence of noise. The computed areas for these peaks are also in fair agreement when compared with the theoretical results. In this case, a filter factor that takes into account all the singular values was found appropriate. The current study shows that the damped singular value decomposition approach can be used to invert real laboratory data.