Journal of Chromatography, 506 (1990) 563-578 Elsevier Science. Publishers B.V., Amsterdam - Printed in The Netherlands CHROMSYMP. 1703 Comparison of stationary phases for packed-column supercritical fluid chromatography PETER J. SCHOENMAKERS*, LOUIS G. M. UUNK and HANS-GERD JANSSEN Philips Research Laboratories. P.O. Box 80000, 5600 JA Eindhoven (The Netherlan~) SUMMARY A study of the applicability of different stationary phases for packed-column supercritical fluid chromatography is described. The compatibility of these phases with carbon dioxide and with a number of test solutes was established qualitatively from the observed peak shapes. Retention and selectivity differences between the different columns were studied quantitatively. The stationary phases studied include chemically modified silicas, polysiloxane-coated silicas, modified and unmodified po- rous graphitic carbon and modified and unmodified poly(styrene-divinylbenzene) copolymers. It is concluded that differences in selectivity between different stationary phases are often due to interactions between the solute molecules and active sites on the surface. Such interactions lead to poor peak shapes and are therefore undesirable. For most of the solutes studied, with the notable exception of polyaromatics, a poly (ethylene glycol)-coated carbon phase provided the best results. Owing to the very high retentivity of unmodified carbon, this phase turned out to be very stable. The coating of different types of polymers on solid surfaces may be used to create station- ary phases without residual active adsorption sites, but with considerable differences in selectivity. INTRODUCTION Carbon dioxide is the preferred solvent for supercritical-fluid chromatography (SFC)‘. It has a number of great advantages, including its favourable critical properties (T, z 3 1°C; zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED PC z 73 bar), favourable safety and toxicity characteristics, availability in sufficient purity at low cost and compatibility with a variety of chromatographic detectors. Unfortunately, carbon dioxide is essentially a non-polar solvent2. Non-polar solutes, such as hydrocarbons, can be eluted as sharp, symmet- rical peaks from many different stationary phases. Polar solutes, however, are often eluted as broad, asymmetric peaks, or are not eluted at a113-‘. To improve this situation, more polar mobile phases can be used. Among the possible pure solvents, ammonia is the only realistic choice?. Although ammonia has been used occasionally in SFC6, its routine application is likely to involve major problems. Therefore, it is 0021-9673/90/%03.50 0 1990 Elsevier Science Publishers B.V.