2.2.5 Analyte Recovery
Now that we have presented some of the common extraction techniques, another problem must be pointed out. How is it possible to know all of the analytes were extracted from the sample (i.e. water, urine, soil, fish)? This question becomes more difficult to answer as the sample matrix becomes more complex. For example, how does the chemist quantitatively recover all of the analyte from lake sediment or from food items? These sample matrixes can have analytes contained within every clay particle or biological cell and require the development and testing of rigorous extraction procedures. Fortunately, many of these procedures have been developed and are published by governmental agencies, industry, or research scientists. As a result, incorporation of these procedures into the laboratory is relatively easy. As an aid to determining how well your extraction procedure works, relatively expensive “reference” samples that contain a known amount of analytes can be obtained for a variety of sample matrixes (i.e. fish, sediment, and manufactured goods). A procedure can be validated if the results from your method are statistically equivalent to the known concentration. For many procedures, it is not necessary to have a high recovery (i.e. 98%) but it is necessary to have a known and consistent recovery, even if it is low.
In addition to the potential human errors present in an analysis, instrument detectors can also contain errors due to non-optimum instrumental setting, out-of-date tuning or calibration, and when peaks elute from the column with more than one analyte or in mass spectrometry when more than one reference spectra is identified in the computer search library. This latter situation is common with low concentrations of analytes.
Now that the basic problems and common errors associated with sampling handling and instrumentation have been identified, we will move on to distinctions between gas and liquid chromatography. Gas and liquid chromatography were originally developed due to the existence of two basic different types of analytes: (1) those that are thermally stable (do not degrade at temperatures up to 300 °C) and are volatile at relatively low temperatures (below 300 °C), and (2) for analytes that are not volatile and/or thermally degrade at temperature above room temperature. GC is used for thermally stable and volatile chemicals while HPLC is used for both non-volatile compounds and ones that degrade at high temperatures. Recent advances in the stationary phases on separation columns and mobile phase selection (solvent gradient in HPLC) allow many analytes that were exclusively analyzed by GC to be analyzed by HPLC. For example, GC was the exclusive technique for analyzing mixtures of volatile organic solvents. Yet today, by changing HPLC to a reverse phased system (where the separation column is the nonpolar phase and the solvent is the more polar phase) it can now analyze components of organic solvents. HPLC will be discussed in depth in the next chapter.
GC analysis can also have special concerns. Impurities introduced during sample preparation can result in contamination that may interfere with the analysis of a desired analyte or introduce additional peaks into the chromatogram (the output of a chromatograph). A notable case is a class of compounds known as phthalates that are found in plastics that interfere with the analysis of chlorinated pesticides such as DDT and PCBs in GC analysis with an electron capture detector (ECD). Even with detection by mass spectrometry, the analysis may conclude that these compounds were present in the original samples when in fact it they are laboratory contaminants. As a result, contact with plastics must be avoided regardless of the detector that is used. It is also important to purchase GC grade solvents (at over $150 per four liters) that are certified to contain an extremely low amount of impurities when trace analyses are being conducted.
Some functional groups of analytes, such as in the analysis of Bisphenol A, a known endocrine disruptor present in some plastic bottles, may react with or irreversibly adsorb to the glass surfaces in the GC injector liner and result in the analyst reporting the absence of Bisphenol A in a sample when in fact it was present but lost during the analysis. This can be overcome by deactivating the surfaces with a silanization agent that coats the glass with a non-reactive trimethylsilane group, and allows the analyte(s) to pass through the system to the detector. What and when to worry about these problems, and many others, come with experience and knowledge of the literature.
| Frank's Homepage |
©Dunnivant & Ginsbach, 2008