Dynamic film formation and the use of retention gaps with direct injection in open‐tubular supercritical fluid chromatography

Abstract

Abstract. Open-tubular supercritical fluid chromatography (SFC) is a useful technique for many analyses, but is somewhat limited, in current practice, with respect to trace analysis of liquid samples. Direct injection of the volumes required for trace analysis with ordinary detectors creates a meter or more length of flooded zone and may lead to unacceptable band broadening. We studied solvent mass transfer on short, uncoated columns for several solvents forming Type I binary mixtures with C02, and found evidence in the solvent peak shapes of distinct liquid and vapor phases, with the liquid present as a dynamically formed film. This study suggests that selecting an injection temperature and pressure to reduce, but not eliminate, the surface tension of the liquidvapor interface, and increasing the mobile phase velocity during injection increase the liquid film thickness and reduce the length of the resulting flooded zone. When a separate uncoated inlet tube (that is, a retention gap) is used to connect the injector with the column, decreasing the inlet tube radius improves the ratio of maximum effective injection volume to inlet tube volume.Just as in GC, dynamically formed films function as pseudo-stationary phases during injection. Without a liquid film, solutes can be transported by injection-solvent-modified mobile phase well beyond the length of a typical film, increasing the difficulty of refocusing the solutes before separation. Refocusing solutes from film-coated flooded zones is easily accomplished using either a solvent effect (specifically, solvent trapping) or phase-ratio focusing.We recommend the use of solvents with phase behavior like toluene and carbon tetrachloride, both of which form films at convenient temperatures and pressures. We discourage the use of solvents with phase behavior like n-pentane which tend to make single-phase, supercritical mixtures with C02 under typical SFC injection conditions.

Injection techniques in SFC

“…Figure 3 shows the critical loci for 14 solvent-CO 2 mixtures [6,7]. The curve for 1-octanol is not connected to the critical point for CO 2 because the mixture is likely Type III [8,9].…”

“…However, its behavior is identical with Type I over the temperature range of the measured data. In addition, anyone with SFC equipment can quickly generate the pressure-temperature points of the critical locus for solvent mixtures not listed using a simple, flow-injection procedure which detects phase separation by its effect on the solvent peak shape [6,7].…”

“…However, an injection of only 100 nl would require almost a 51-mm length of column just to contain the volume of the injected plug. If a separate liquid phase can exist at the temperature and pressure in use, the action of the mobile phase on a liquid plug will spread it into a flooded zone several meters in length [6,[10][11][12][13][14]. Solutes will, of course, be distributed over this entire flooded zone by the liquid solvent.…”

“…This change in spatial peak width occurs whenever two tubes providing different peak velocities are coupled. The difference in peak velocities may result even when tubes with the same stationary phase but different phase ratio are chosen, thus giving us the name phase-ratio focusing [6,[14][15][16]. Focusing will also occur when two columns with different stationary phases are coupled if the second has a slower peak velocity than the first.…”

“…Phase-ratio focusing is often taken to extreme by using an a tube with no stationary phase at all, that is, a retention gap [6,[14][15][16]. Its purposes are to contain flooding that occurs from the injection, to initially retain the solutes, and to accomplish at least a crude separation of the injection solvent from the solutes.…”