Most scientists when faced with the analysis of a sample give little or no attention to how the sample was obtained, i.e., the sampling process. This is not the topic of discussion for this editorial. Suffice it to say one should realize that the results of the analysis of a sample can only be as reliable as the sample is representative. One must remember; to obtain reliable analytical data three components are always involved: 1) the system (a representative sample), which consists of the analytes of interest and the matrix (the part not requiring analysis but which can interfere with the instrumentation), 2) a measuring instrument, and 3) the analyst or observer (a human being!). Well over 50% of the analysis time is spent on sample preparation and, since you have a human being as part of this system, sample preparation is the most error-prone and labor-intensive task in the analytical laboratory.
For this discussion we will assume ALL samples are homogeneous and representative! If the final measurement of analyte concentration is by GC, one should strive to circumvent complete sample matrices; e.g., nonvolatile components and interfering analytes.
One criterion for an analyte is that it must have a vapor pressure of 0.1 Torr at operating conditions in gas chromatography (GC). Thus, it must be able to be vaporized in the system inlet. The sample can be a gas, liquid or in some cases a solid (thermally stable or capable of producing a definite pyrolysis pattern). Most sample preparation techniques for GC are based on variations of extraction theory whereby the analytical chemist may change solvent, temperature, pressure, phases or volumes. Thus, an understanding or comprehension of liquid-vapor, liquid-liquid and liquid-solid equilibria are assumed.
IMPORTANT NOTE: Before you attempt the gas chromatographic analysis of an unknown sample obtain as much information about the sample as possible. Randomly injecting an unknown sample into a gas chromatograph does not reveal complete analytical data about the sample!
Having three states of matter (gas, liquid & solid), we have a series of sample preparation techniques available.
The main sample preparation techniques which can be classified by these equilibria are: Static headspace technique: analytes of interest (volatiles) are equilibrated in a closed vial at a specified temperature & pressure. A gas-tight syringe is used to transfer the headspace sample into the gas chromatographic injection port. Dynamic headspace technique: analytes of interest are swept or purged onto an adsorbent and then thermally desorbed into the gas chromatograph. Solid-phase extraction (SPE): may be used to concentrate analytes from gaseous or liquid samples and often is used to clean up and concentrate liquid extracts. The adsorbed analytes can be eluted with a solvent or thermally desorbed. Solid-phase microextraction (SPME): may be used for both gaseous and liquid samples. A fused-silica polymer coated fiber (e.g., with polydimethylsiloxane) is exposed to the stirred sample. The shielded-fiber is then inserted into the injection port of the gas chromatograph. Distillations: these are predominately macro scale techniques and are rarely employed as sample preparation techniques for GC. Stir bar sorptive extraction (SBSE): this is a dynamic variation of SPME in which a spinning glass-covered magnetic bar (coated with a thick layer of polydimethylsiloxane) is used to sorb analytes of interest, which can be removed by thermal desorption in the gas chromatographic injection port.
The technique of liquid-liquid extraction has lost appeal to the analytical chemist because of (1) the time needed to reach equilibrium, and (2) the volume of solvent needed for quantitative recovery of analytes (the environmental restrictions on waste disposal of used solvents). Quantitative extraction of organic species from aqueous systems requires that the organic moiety (1) is non-polar, (2) does not dissociate in the aqueous phase, and (3) does not dimerize or polymerize in the organic phase. Thus, liquid-liquid extractions have limitations as sample preparation techniques for GC UNLESS the equilibrium between the two phases exhibits a large numerical partition coefficient. If this is the case, one may resort to micro liquid-liquid extractions, where the concentration factor is >1200 times that for the macro-technique [J. Chromatogr. 106,299 (1975) and J. Chromatogr. 177,135 (1979)]. Thus, classical liquid-liquid extractions have been replaced by modern & efficient techniques and are more prevalent in organic synthesis laboratories or for the separations of metal complexes, metal chelates, and/or ion-pairing reagents.
A number of classical liquid-solid equilibria techniques are available (e.g., ion exchange or Soxhlet extraction) but only Soxhlet extractions have application in sample preparation prior to GC. This technique is not commonly used because: (a) a large volume of solvent is needed for the sample extraction, (b) an evaporation step is required to concentrate the sample, (c) lack of thermal stability and volatility of some sample analytes, and (d) interference from contaminants in the extraction thimbles (requires a blank extraction prior to sample extraction).
In the past several years, newer techniques have become available. Accelerated solvent extraction (ASE), sometimes referred to as pressurized liquid extraction (PLE) or pressurized fluid extraction (PFE), may be used for solid and semi-solid samples. Elevated temperatures and pressures used in these techniques cause hydrogen bonds and dipole interactions to be reduced, and surface wetting is increased. Water may be used as the solvent if it is below its critical point. Then, it is known as subcritical water extraction (SWE); which makes it similar to SFE.
A similar sample preparation technique is microwave-assisted extraction (MAE); sometimes referred to as microwave-assisted solvent extraction (MWE). The pressure generated is ca. a few hundred psi; however, the extraction container must be microwave transparent (e.g., PTFE or quartz). The solvent used may be microwave absorbing or non-microwave absorbing. In place of microwaves, ultrasonic vibrations may be used to assure good contact between sample and solvent. This is a fast technique but efficiency is not as high as with other techniques. Low concentrations of analytes in samples require multiple extractions. This sample preparation technique is referred to as ultrasonic extraction (USE).
A technique which became very popular during the 1980s is supercritical fluid extraction (SFE). Supercritical fluids (SFs) are dense gases above their critical temperature & pressure. Thus, SFs possess properties which resemble both liquids and gases. Analytes are more soluble in SFs when they are in their liquid state; thus, analyte melting points and solubility in the SF are important properties to consider. SFEs are fast and very efficient.
Another group of sample preparation techniques are solid-phase extraction (SPE), solid-phase microextraction (SPME) and stir bar sorptive extraction (SBSE). SPE is a technique (invented in the late 1970s) referring to a non-equilibrium exhaustive removal of analytes (semi-volatiles and non-volatiles) from a liquid sample by retention on a solid phase (sorbent) and then subsequent removal of selected analytes by solvent elution. Particulate matter in the sample can interfere with the analysis. Thus, particulate matter may sorb some analytes of interest and cause low analytical recoveries. NOTE: Remove particulates, by filtration, prior to SPE analysis! The efficient use of this technique requires optimization of the sorption and desorption processes.Another group of sample preparation techniques are solid-phase extraction (SPE), solid-phase microextraction (SPME) and stir bar sorptive extraction (SBSE). SPE is a technique (invented in the late 1970s) referring to a non-equilibrium exhaustive removal of analytes (semi-volatiles and non-volatiles) from a liquid sample by retention on a solid phase (sorbent) and then subsequent removal of selected analytes by solvent elution. Particulate matter in the sample can interfere with the analysis. Thus, particulate matter may sorb some analytes of interest and cause low analytical recoveries. NOTE: Remove particulates, by filtration, prior to SPE analysis! The efficient use of this technique requires optimization of the sorption and desorption processes.
An extension of SPE, known as solid-phase microextraction (SPME), came about in 1989; the technique utilizes a liquid or solid stationary phase on a fiber, tube, vessel walls, suspended solids, stirrer, or disk/membrane. The technique may be applied to volatiles, semi-volatiles, and non-volatiles and is an equilibrium technique.
An excellent treatment of sample preparation techniques may be found in Chapters 11 and 15 of Modern Practice of Gas Chromatography, R.L. Grob and E.F. Barry (Eds.), John Wiley & Sons, Hoboken, NJ, 4th ed., 2004.