SOME NEW DIRECTIONS IN IMPEDANCE SPECTROSCOPY DATA-ANALYSIS

New developments in two main data analysis areas are discussed: (a) co mplex nonlinear least squares (CNLS) fitting of data and (b) data-tran sforming and optimizing integral tranforms. In the first category, a M onte Carlo study is used to answer the question of which of several di fferent parameterizations of an ambiguous equivalent circuit model lea d to minimum correlation between fitting parameters, a desirable condi tion. In addition, results are briefly discussed which address the que stions of (1) what should be minimized in CNLS fitting? (2) how well c an one discriminate between exact small-signal binary electrolyte resp onse and conventional finite-length diffusion response? and (3) what i s the ultimate precision of parameter estimates obtained in a CNLS fit ? In the second area, new forms of the Kronig-Kramers relations (KKR) are discussed; the accuracy of several different ways of carrying out the numerical quadratures needed in such transforms is compared; and i t is shown how random errors present in complex data are transformed b y the KKR. Then, new transforms are described and illustrated that can replace exponential Fourier and KK transforms and, at the same time, can greatly reduce random error and some kinds of systematic errors in real, imaginary, or complex frequency response data or transient resp onse data without the need for smoothing or filtering parameter choice s.