Making Decisions by Analytical Chemistry

It has been long recognized that results of analytical chemistry are not flawless, owing to the fact that professional laboratories and research laboratories analysing the same type of samples by the same type of instruments are likely to obtain significantly different results. The European Commission maintains a program of interlaboratory testing that collects information on numerous samples, which provides the statistical documentation of the disagreements. Surprisingly large variations were observed among results on samples and methodologies that are considered uncomplicated. These discrepancies are very unfortunate because erroneous conclusions may arise from an otherwise meticulous and dedicated effort of research staff. This may eventually lead to unreliable conclusions thus jeopardizing investigations of environmental monitoring, climate changes, food safety, clinical chemistry, forensics and other fields of science where analytical chemistry is the key instrument of decision making. In order to elucidate the potential origin of the statistical variations found among laboratories, a major program was undertaken including several analytical technologies where the purpose was to correlate variations in intralaboratory results to variations of interlaboratory investigations. Results from analysis of iron by UV/VIS spectrophotometry and by GF-AAS are presented, and the statistical evaluation of data shows that the standard deviations approaches the results predicted by the Horwitz curve. In addition, the standard deviations obtained by merging all results of interlaboratory testing under the auspices of the European Commission, considering a few prerequisites, also show are remarkable resemblance to the results obtained in a single laboratory, which contradicts the level of accuracy published in research literature. The possible deviations are suspected to originate from long-term variations of detection systems of instrumental analysis, and the impact on these findings on future measurements of analytical chemistry is discussed.