Headspace sampling with capillary columns

A headspace sample is in principle a gas sample which has been previously i n contact with a liquid or solid sample from which volatile compounds were released into the gas with subsequent analysis by gas chromatography. Heads pace gas chromatography (HS-GC) is thus a technique of gas extraction and c an be carried out comparable to a solvent extraction as a one-step extracti on (static or equilibrium headspace) or as a continuous extraction (dynamic headspace). If the concentration of the volatile analytes is sufficiently high, a small volume of the gas sample can be injected directly with all co mmon devices, known for gas sampling. The resulting sensitivity depends, ex cept from detector sensitivity, from the capacity of the column for a gas s ample. Packed columns are traditionally preferred for gas analysis. Trace a nalysis on the other hand requires in most cases high resolution also and t hus the application of capillary columns. The critical problem relating to sample introduction for headspace (viz. gas) analysis is a function of colu mn capacity and the initial sample bandwidth. The admissible volume of a ga s sample is limited by beginning of band broadening. Although the capacity of a capillary column for a gas sample depends on its cross section, the in crease of the inner diameter gives a marginal improvement only and is paid off with reduced separation efficiency. Cryogenic trapping of the volatile analytes is a more efficient way and the variety of such techniques can be classified in methods for cryogenic condensation where the volatiles are tr apped just by condensation at very low temperatures followed by instantaneo us evaporation. Alternatively, with a technique, called cryogenic focusing here, the volatile compounds are trapped directly into a coated capillary c olumn but at moderate low temperatures where the liquid phase is still effe ctive as a chromatographic phase. Band concentration here is achieved by de celerated chromatographic migration at the lower column temperature and eve n more efficiently if an additional temperature gradient is included in the system. Cryogenic trapping in both cases, however, has to deal with water, which is often present at a high concentration in the headspace and may bl ock a cooled capillary column by an ice plug. The need to remove the water before cryogenic trapping is more urgent with the dynamic headspace procedu res due to the higher amount of accumulated water compared to static HS-GC and the techniques to remove the water from the headspace sample therefore differ also. Diffusion through semipermeable membranes and condensation in a reflux condenser are mainly applied for the dynamic headspace procedures, while the much smaller water amount in static HS-GC is easier removed by c hemisorption on an inert hygroscopic salt. When comparing and evaluating th e various techniques of headspace sampling, the deciding criterions are not only method sensitivity but equally important is the degree of automation. It improves precision and provides for more effective use of laboratory pe rsonnel, particularly for industrial routine analysis. It also helps to pro cess the high number of samples, necessary to get the many data for method validation to certify an analytical method. (C) 1999 Elsevier Science B.V. All rights reserved.