Molecular weights and molecular-weight distributions from ultracentrifugation of nonideal solutions

Ultracentrifugation, membrane osmometry and capillary viscometry experiments have been performed on two dextran samples, which have molecular-weight distributions (MWDs) similar to those of dextrans used as blood plasma extenders. The manufacturer reported values of M n and M w, determined by end group analysis and by light scattering, respectively. Our values of M n, determined by osmometry, and M w, calculated from ultracentrifugal and viscometry experiments, agreed quite well with the manufacturer's results. Good agreement was obtained with values of M w and B LS (the light scattering second virial coefficient) obtained from sedimentation equilibrium experiments at different speeds using sector or nonsector-shaped centerpieces. Several ways of obtaining M w, M z and B Ls from sedimentation equilibrium experiments are presented. We have also shown how to obtain the speed-dependent term of the sedimentation equilibrium second virial coefficient. Both B Ls and the speed-dependent nonideal terms could be used to correct the sedimentation equilibrium data, so that ideal values of d In c/d( r 2) or d c/d( r 2) could be estimated and used to obtain the MWDs of the dextran samples. Both Donnelly's and. Scholte's methods were used with the sedimentation equilibrium data. With both methods, unimodal MWDs were encountered, Which gave good agreement with the manufacturer's MWDs, obtained by a combination of analytical gel chromatography and light scattering, Uncoreected sedimentation equuibrium data gave MWDs quite different from the manufacturer's results. The MWD calculated from the differential distribution of sedimentation coefficients also gave a unimodal MWD, but this MWD did not give as good agreement with the sedimentation equilibrium results or with the manufacturer's results.