Premium
A nonlinear regression method for the analysis of 1 H T 2 dispersion curves including comments on the role of polydispersity
Author(s) -
Williams M. A. K.,
Keenan R. D.,
Halstead T. K.
Publication year - 1998
Publication title -
magnetic resonance in chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.483
H-Index - 72
eISSN - 1097-458X
pISSN - 0749-1581
DOI - 10.1002/(sici)1097-458x(199803)36:3<163::aid-omr239>3.0.co;2-i
Subject(s) - chemistry , dispersion (optics) , nonlinear regression , curve fitting , nonlinear system , statistical physics , dispersity , interpretation (philosophy) , regression analysis , statistics , optics , mathematics , physics , quantum mechanics , organic chemistry , computer science , programming language
A nonlinear regression method is described in detail for analysing 1 H T 2 (Carr–Purcell–Meiboom–Gill) dispersion curve measurements obtained from water‐rich systems containing molecules possessing labile protons. The method is routine, provides well specified error statistics and does not make assumptions about the accessibility or mobility of the guest molecule site. The approach is validated using data simulated with the Carver–Richards two‐site exchange model and experimental measurements carried out on a series of methanol‐ d 4 –water mixtures of varying pH. Further simulations are described which investigate the role of polydispersity in determining the form of the dispersion curve and the relevance of the parameters extracted by fitting a composite curve to a single Carver–Richards function. The procedure is subsequently used to fit experimental dispersion curves originating from maltodextrin–water systems and the interpretation of the results is shown to be consistent with previously suggested ideas for molecular interactions in polysaccharide systems, involving discrete units of double helical nature, which can further aggregate to develop structure at a supramolecular level. ©1998 John Wiley & Sons, Ltd.