A broad-standard technique for correcting for band broadening in size-exclusion chromatography

Peng Zhang, Paul Mazoyer, Robert G. Gilbert

Band broadening in size-exclusion chromatography (SEC) is always present to some extent. Broadening effects on averages such as the weight- and number average molecular weights (MW¯ and Mn¯ respectively) are minimal with modern SEC systems. However, broadening distorts the shape of the true molecular weight distribution (MWD), which causes problems if one wants to compare the detailed form of the MWD to a model. An addition to current methods for overcoming this problem is presented. One starts with a sufficiently wide range of samples whose exact values of Mn¯ andMW¯ have been measured by non-SEC methods (e.g. by fluorimetry and light scattering, respectively, of the sample without size separation). A true (unbroadened) molecular weight distribution for a sample can be obtained by deconvolution (here using a maximum-entropy algorithm) by fitting SEC data for these samples to these exact Mn¯ and MW¯ values to find the values of the parameters in a sufficiently flexible assumed broadening function. This was modelled using simulated band broadening and subsequent deconvolution, with the broadening parameters least-squares fitted to the "exact" sets of values of Mn¯ and MW¯. The results show that if these Mn¯ and MW¯ values are for a series of broad (not narrow) standards covering a sufficient range of molecular weight, then after deconvolution, a good representation of the original molecular weight distribution used in the simulation is obtained. The method should prove useful for water-soluble polymers, for which it is often difficult to obtain narrow standards of a wide range of molecular weight, as required in a number of well-established methods for correcting for band broadening.