Factors Influencing Chromatographic Performance and Metrics

Key Insight

Chromatographic column design and performance depend on equilibrium and dispersive factors. Equilibrium factors, such as adsorption equilibrium, dictate solute distribution between phases and account for fast processes like ionic dissociation. Dispersive factors, or 'rate factors,' include mass transfer resistances, kinetic binding resistance, and mobile phase dispersion effects like axial diffusion, hydrodynamic dispersion, and non-uniform flow. Performance is also influenced by three porosities: extra-particle (epsilon, typically 0.3-0.4 for packed beds), intra-particle (epsilon p, ranging from near zero to over 90 percent for gels), and total column porosity (epsilon t). Mobile phase flow, usually laminar (Reynolds number Re << 1), is largely restricted to extra-particle space, defining superficial (u) and interstitial (v) velocities. Solute transport within intra-particle pores typically relies on diffusion, though intra-particle convection can occur with large pores (>100 nanometers), small particles (<20 micrometers), and high flow rates. External factors like column hardware, mixing devices, and detectors also contribute to overall performance.

Column efficiency, a measure of how closely ideal chromatography conditions are approached, is expressed by the height equivalent to a theoretical plate (HETP), H, or the plate number, N. Ideal behavior signifies H approaching 0 or N approaching infinity. Dispersive factors primarily determine H, which varies with solute molecular properties, flow rate, and, to a lesser extent, equilibrium factors. The generalized van Deemter curve empirically relates reduced HETP (h) and reduced velocity (v prime) for porous particles, incorporating axial diffusion (parameter b, typical 2), hydrodynamic dispersion (parameter a, typical 1), and mass transfer effects (parameter c, typical 0.05). In gas chromatography (GC), axial diffusion is dominant, increasing column efficiency with flow rate. For small molecules in high-performance liquid chromatography (HPLC), hydrodynamic dispersion is dominant, making HETP relatively insensitive to flow rate and molecular diffusivity. For proteins and macromolecules in preparative/process scale applications, mass transfer in the stationary phase is typically dominant, causing HETP to increase linearly with flow rate and the square of particle size, and to decrease in proportion to the protein molecular diffusivity.

Chromatographic resolution (Rs) quantifies the separation between two components; Rs ≈ 1.5 indicates complete separation, while Rs = 1 (approximately 98 percent separation) is often considered adequate. For elution chromatography, Rs is related to column efficiency (N), selectivity (alpha, the ratio of Henry's law isotherm slopes), and the average retention factor (k prime). Increasing selectivity or N improves resolution, but larger k prime values have diminishing effects. However, increasing peak separation via more plates generally leads to greater product dilution. Dynamic Binding Capacity (DBC) is a critical metric for capture applications, defining the amount of protein held in the column before the effluent concentration reaches a specified percentage (e.g., 10 percent) of the feed concentration. DBC is influenced by dispersive factors and approaches the Equilibrium Binding Capacity (EBC), which depends solely on thermodynamics, under conditions of infinite column efficiency. Scaling relationships, derived from these performance criteria, aim to maintain constant resolution and DBC by keeping the plate number constant, subject to constraints like allowable column pressure (typically a few bars for process scale) and desired separation time. For protein chromatography where mass transfer is the controlling dispersive factor, scaling requires maintaining a constant value for the expression D0 * L / (U * dp^2). These relationships assume consistent adsorption equilibrium, mass transfer mechanisms, rigid particles, temperature, and viscosity across scales, and are primarily for preliminary estimates, often requiring intermediate-scale experimental validation.