Studying effective column lengths in liquid chromatography of large biomolecules.

Based on their nature, large molecules tend to exhibit on-off elution such that only a small segment of a column bed participates in their separation. We were intrigued to investigate empirical data on this behavior and to apply a simple method to estimate the length of column bed that is needed to produce an effective separation. Models were derived by rearranging the linear solvent strength (LSS) model equations, and data sets from almost 100 different separation conditions were treated to illustrate effects for various types of solutes as separated by reversed phase (RP), ion-pair reversed phase (IP-RP), ion-exchange (IEX), hydrophobic interaction (HIC) and hydrophilic interaction (HILIC) chromatography. By empirically measuring S parameters (S is a solute dependent model parameter, it describes how sensitive is the solute retention to mobile phase composition), and calculating for an exit retention factor of 0.5, we have determined that there is little to no benefit to separating moderately sized solutes (5 - 10 kDa) with a column bed that is longer than 3 cm, particularly when a less than 20 min gradient is desired. Moreover, even shorter columns would be predicted to be adequate for 100 - 150 kDa molecules. Interpretations of this sort have become possible because there is some correlation between a solute's molecular weight and its S parameter. That is, empirical observations on retention behavior are not needed to select appropriate column lengths; molecular weight provides a sufficient approximation. With these insights, we suggest reconsidering the routine use of 5 - 15 cm long columns for >10 kDa biomolecule separations and instead propose that a new focus be placed on 1-2 cm long columns.