Fundamentals and Properties of Size-Exclusion Chromatography Packings and Columns

Apr 01, 2012
Volume 30, Issue 4, pg 46–53

To aid chemists in proper usage of size-exclusion chromatography (SEC) columns, this article reviews the fundamentals of the technique, discusses packing properties, and provides practical approaches for increasing resolution.

This paper is the third of a series of articles on size-exclusion chromatography (SEC) articles (1) published in special column technology supplements to LCGC (2,3) edited by Ron Majors. The fundamentals of SEC are reviewed, with emphasis on column performance parameters that affect SEC resolution. Based on fundamental principles, practical considerations of ways to improve resolution in SEC will be explored. The second installment, which provides a listing of commercial SEC offerings, a comprehensive review of specialized SEC columns, and a prediction of future directions in SEC column technology, will be published in the "Column Watch" column in the July issue of LCGC.

SEC Fundamentals

Unlike all other chromatographic methods (with the exception of gas chromatography analysis with zeolite packings, and the second dimension of 2D gel electrophoresis), SEC is based on an entropic separation mechanism in which the separation is governed by hydrodynamic volume differences among macromolecules in solution. This "size" difference is a function of molecular weight, molecular configuration (architecture), and molecular conformation (shape).

Thermodynamic Model

All chromatographic separations are based on the partitioning of components between two immiscible phases, represented by the thermodynamic partition coefficient

where K t is equal to the concentration ratio of a solute distributed between the two phases under ideal thermodynamic conditions, that is, at equilibrium and infinite solute dilution (4), ΔH o is the change in enthalpy at standard conditions when a solute adsorbs, absorbs, or partitions into the stationary phase, ΔS o is the conformational entropy change at standard conditions when the solute diffuses into the stagnant mobile phase of the pores of the packing, R is the gas constant, T is the absolute temperature of the system, and c 1 and c 2 are the concentrations of the solute in phases 1 or 2 at equilibrium conditions.

In SEC, the mobile phase, packing, and column temperature are chosen so that there is no enthalpic interaction between the solute (macromolecule) and packing; thus ΔH = 0 and equation 1 becomes

In actual practice, SEC conditions are often nonideal because thermodynamically good solvents, rather than theta-solvents, are used as the mobile phase to increase sample solubility. Furthermore, high sample concentrations are often injected because of poor detection sensitivity. Additionally, SEC is a dynamic process and may not be at equilibrium because of high flow rates that are occasionally used to decrease analysis times.

Because of these nonideal conditions, an experimental SEC distribution coefficient, K SEC, is used in place of K t:

where <c>i is the average solute concentration within the pore volume of the packing, and <c>o is the average solute concentration in the interstitial volume of the packed column. As long as we keep ΔH = 0 by the judicious selection of mobile phase composition and chemical composition of the SEC packing, the separation will depend only on the conformational entropy change when a macromolecule diffuses from the interstitial volume into the pores of the packing, and then back out again. The driving force that promotes sample diffusion into and out of the pore volume is simply the sample concentration gradient. Thus, SEC separation depends only on the molecular hydrodynamic volume or size of the sample, with respect to the shape and average pore size of the packing. Because the separation does not depend on enthalpic interactions, gradient elution is not used in SEC, simplifying instrumentation.

Figure 1
The thermodynamic model of SEC can be visualized using Figure 1 for the separation of three sizes of macromolecules. In Figure 1a, we see small MW solutes that freely diffuse into and out of the pores; as a result, they elute at the total elution volume, V t, in which ΔS = 0. (Note that V t = V i + V o where V i is the pore volume and V o is the interstial volume.) In Figure 1b, macromolecules begin to approach the average pore size of the packing. The sample concentration gradient drives macromolecules into the pore volume; ΔS becomes negative and K SEC ≈ 0.5. In Figure 1c, because of this sample's large hydrodynamic volume, it cannot enter the pores on average; thus there is no change in conformational entropy, so Δs≈ 0 and K SEC ≈ 0.

lorem ipsum