Bias, limit of detection, and limit of quantitation for the ASTM D2425 method updated in 2019

• Published in: Journal of Chromatography A Vol. 1614; p. 460705
• Main Authors: Ikonen, K. Elias, Wehde, Katherine E., Khalida, Habeb, Kenttämaa, Hilkka I.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 15.03.2020
• Subjects: Gas Chromatography-Mass Spectrometry - methods; Hydrocarbons - analysis; Limit of Detection; Reproducibility of Results
• ISSN: 0021-9673; 1873-3778
• DOI: 10.1016/j.chroma.2019.460705

•The LOD was 0.004 wt% and the LOQ was 0.001 wt% for renewable jet fuel.•The LOD was 0.01 wt% and the LOQ was 0.02 wt% for gas oil.•Both sample sets were positively biased for paraffin content.•Both sample sets were negatively biased for alkylbenzene content.•The ASTM D2425 method may not work well for renewable jet fuel samples. An update for ASTM D2425, “Standard Test Method for Hydrocarbon Types in Middle Distillates by Mass Spectrometry,” was recently released to expand its scope by adding modern mass spectrometric instrumentation and a new sample type, synthesized hydrocarbons. While the method update is a proper step towards modernization, it still lacks validation parameters. The description of the updated standard method contains precision estimations but no consideration for a possible bias, meaning that the method accuracy was not evaluated. The 2019 update does not either provide the lowest concentration of an analyte that can be detected and quantitatively determined, ‘limit of detection’ (LOD) and ‘limit of quantification’ (LOQ). In the present study, the LOD, LOQ, and bias were evaluated as an intra-laboratory study. The LOD and LOQ determination was based on the adjusted standard deviation of the least abundant hydrocarbon type in a sample, and were 0.004 and 0.01 w/w% in renewable jet fuel (RJF) and 0.01 and 0.02 w/w% in gas oil (GO), respectively. The bias was determined by measuring relative hydrocarbon recovery percentages for spiked samples. The renewable jet fuel was spiked with cycloparaffins and alkylbenzenes and the recovery percentages were 140% and 63%, respectively. The gas oil was spiked with paraffins and alkylbenzenes and the recovery percentages were 117% and 90%, respectively. Both sample sets were positively biased for paraffin content (recovery percentage over 100%) and negatively biased for alkylbenzene content (recovery percentage less than 100%). Based on above results, this method may not work well for RJF samples.