Premium
Evaluation of light‐scattering detectors for size exclusion chromatography. II. Light‐scattering equation selection
Author(s) -
Jeng L.,
Balke S. T.
Publication year - 1993
Publication title -
journal of applied polymer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.575
H-Index - 166
eISSN - 1097-4628
pISSN - 0021-8995
DOI - 10.1002/app.1993.070490805
Subject(s) - radius of gyration , extrapolation , scattering , light scattering , chemistry , analytical chemistry (journal) , detector , computational physics , optics , physics , statistics , mathematics , chromatography , polymer , organic chemistry
A new multiangle laser light‐scattering (MALLS) detector for size exclusion chromatography promises simultaneous measurement of both weight‐average molecular weight ( M w ) and radius of gyration ( r g ) at each retention volume across the chromatogram. However, there are a variety of ways of interpreting the raw data to provide these results. This study examines variations of three different rearrangements of the basic light‐scattering equation. Data from a room temperature analysis of polystyrene and a high‐temperature analysis of polyethylene were used. The degree of fit of each equation to the data and the precision of the M w and r g values are evaluated. To define precision, joint confidence regions (JCRs) were calculated and compared to simple confidence intervals based upon standard deviations in order to see the effect of interdependence of M w and r g . Results showed that the Debye equation was superior to the inverse Debye equation (similar to the Zimm plot) for the interpretation of MALLS data. The effect of the quantity of data included in the regression model was also assessed. Use of only the most precise four detector angles was compared to use of a full set of 15 angles. Precision of weight‐average molecular weight values was found to be improved as the detector angle decreased because of the shortened extrapolation to zero angle. Precision at room temperature was much superior to that at high temperature. Use of simple confidence intervals was shown to provide only a fair approximation to the more accurate JCR. The “natural scatter” of data shown by the JCR generally shows the same trend as do plots in the literature of M w vs. r g . Thus, it is concluded that JCRs should be more often calculated in light‐scattering studies in order to distinguish random scatter from meaningful correlations of these values. © 1993 John Wiley & Sons, Inc.