Some of the variables that affect the accuracy of the determination of polychlorinated biphenyls (PCBs) in soils are the linear dynamic range, calibration range and detector drift of the gas chromatographic instrum ent. The linear dynamic range is especially narrow for PCBs. In the interlaboratory study on which this series of papers is based, a large number of laboratories submitted results that were strongly biased by analysts who did not take account properly of the linear dynamic range of the instrument. There were three types of errors: calibrating the instrument within the linear dynamic range but analysing samples above and below this range; calibrating the instrument outside the linear dynamic range; and allowing excessive drift in the detector. Each of these errors produced a characteristic pattern of biased results. The net result was a large interlaboratory variance and biases.
The various factors in the separation of polychlorinated biphenyls from soil for determination by gas chromatography were compared using data from an interlaboratory study, including extraction equipment, solvents and clean-up procedures. These were compared for accuracy and precision using five soils spiked with Aroclor 1260 at concentrations covering a range of four orders of magnitude. Results from laboratories using Soxhlet extraction were significantly more accurate than those obtained using sonication, especially at higher concentrations, but with equal precision. Results from laboratories using non-polar solvents showed significantly lower accuracy than results with more polar solvents, with equal precision. Results from laboratories using Florisil column clean-up were significantly more accurate and precise at lower concentrations than those from laboratories using no clean-up procedure.
The determination of polychlorinated biphenyls in soils is one of the most widely performed tests among environmental laboratories. It is a very difficult test and is subject to a great deal of inter-laboratory variance and bias. A new inter-laboratory study, using five soils spiked at four different concentrations with Aroclor 1260, was conducted with two groups of laboratories numbering 20 and 129, respectively. The results of this study are compared with the results of the other studies but show that the bias is concentration dependent. A number of definite patterns of recovery are noted, indicating that the large variance and bias was not due to random errors but to wide-spread systematic errors.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.