Solute/solvent interaction corrections account for non-ideal freezing point depression.

A new highly accurate curve-fitting technique for looking at freezing-point depression data was proposed by Fullerton et al. (Biochem. Cell Biol., in press). The method involve plotting mass solvent to mass solute ratio (Mw/M(s)) vs. 1/delta T (i.e. the inverse change in freezing point). A measured molecular weight and a solute/solvent interaction parameter (called I value) are inferred from the resultant linear plot. The accuracy of the molecular weight method was first demonstrated with the monomers of ethylene glycol, glycerol, propanol, mannitol, glucose and sucrose to show a mean molecular weight error of 0.02% with root mean square (RMS) error 0.9%. The RMS error (0.9%) is our best estimate of the molecular weight measurement accuracy for the method applied to a monomer. This error is consistent with the experimental precision (approximately 1%) which implies no systematic error. Non-ideality is described with a single constant, I. Polyethylene glycol (PEG) polymers of increasing length (vendor designation 200 to 10,000 Da) were analyzed to show monotonically increasing non-ideality (I values of 0.12 to 3.67) with increasing molecular weight. The measured molecular weights agreed with the end-point titration value for the three smallest polymers (where the number of polymeric units was less than or equal to 7). The method underestimates the vendor molecular weights for longer polymers. This disagreement is assigned to segmental motion (internal entropy) of longer, more flexible, PEG molecules.