Modeling Limits of Detection in Headspace GC/MS

The limits of detection (LOD) for headspace Gas Chromatography-Mass Spectrometry (GC-MS) of a diverse series of compounds were accurately predicted using a simple linear regression model incorporating Henry’s law constants and a polarity descriptor. LODs were predicted with an R2 of 0.9154 and a standard error of the estimate (SEE) of 1184 nmol/L (approximately 17 μg/L) over a range of LODs from 2 μg/L to 585 μg/L using Henry’s law constants and dipole moment.