La métrologie en chimie. Quelques principes et beaucoup d'incertitudes

Abstract

The increasing need for reliable analytical measurements in chemistry is a consequence of the impact of results on socio-economic or legal decisions. Lack of reliability of analytical results, partly originates from the absence of a structured metrology system in chemistry. Ideally, such a metrology system should be able to allow to generate results traceable to the International System of units and to evaluate reliably the uncertainty. Most chemical measurements request long and complex procedures leading to physical or/and chemical treatments of samples forbidding a direct link to the kilogram or the mole. Only few analytical methods can be proposed as 'primary methods' linking results directly to the fundamental units and often they cannot be used on real samples. To perform their task analytical chemists use secondary methods, validated and calibrated with reference materials, certified if possible. The way used to obtain secondary traceable methods, to develop and use reference materials, and to evaluate the uncertainty of measurement results are described and discussed.