Multivariate calibration to implement green ETAAS methods when analysing Cu in lubricating oils

Abstract

The quantification of metals in lubricating oils is relevant to trace machinery wearing and to evaluate potential environmental effects related to their disposal. Electrothermal atomic absorption spectrometry (ETAAS) is a common choice to measure metal content and despite efforts being made to reduce the amount of organic materials in the measurement aliquot (e.g., using emulsions), potential interferents still remain there. Therefore, quantification of the analyte is a highly difficult task. In this work a general-purpose methodology based on multivariate partial least squares regression (PLS) is presented to address interferences when difficult organic materials are analysed by ETAAS. It is shown that such a methodology yields powerful quantification models and requires less staff dedication, shorter turnaround times and lower expenses than traditional approaches. Besides, it is totally compatible with green analytical chemistry principles. Further, figures of merit which consider the risk of false negatives and false positives were calculated following the latest ISO and European guidelines: critical level (decision limit), minimum detectable value (detection capability), trueness and precision, multivariate sensitivity and sample-specific confidence interval. They have not been used in the atomic spectrometry field. The case study is about quantifying Cu in lubricating oils as a tracer of machine weathering.