Target analysis of tert-butyldimethylsilyl derivatives of nerve agent hydrolysis products by selectable one-dimensional or two-dimensional gas chromatography-mass spectrometry.

A target analysis method for the sensitive and discriminative determination of the nerve agent hydrolysis products alkyl methylphosphonic acids as their tert-butyldimethylsilyl (TBDMS) derivatives was developed using a combination of selectable one- and two-dimensional (1D/2D) GC-MS, and applied to the analysis of samples with significant interfering matrices. After sample drying, the alkylmethylphosphonic acids and methylphosphonic acid (MPA) were converted to TBDMS derivatives by addition of N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide with heating, and subjected to 1D/2D GC-MS. The apparatus consisted of an initial low thermal mass DB-5 column and a second DB-17 column together with an electron ionization quadrupole mass spectrometer, offering simple and flexible switching between one- and two-dimensional GC-MS analysis in a single GC-MS system. Using 1D/2D GC-MS, analytes that do not co-elute with matrix components can be separated using 1D GC mode alone. Only those parts of the chromatogram that are negatively affected by the co-elution of matrix components need to be transferred and separated with 2D GC. Quantitation can be performed by a combination of both separations and mass spectrometric detection. The TBDMS derivatives of ethyl-, isopropyl-, isobutyl-, pinacolyl-, and cyclohexyl-MPA (cHMPA) and MPA itself were well separated within 3min and determined in 1D GC-MS mode with detection limits of around 10ng/ml of reaction mixture (except for the cHMPA derivative, whose mass spectrum contained noisy background peaks). In 2D-GC-MS mode, where each 0.04min elution window from the 1D GC was subjected to heart-cut (H/C) and transferred to the second column after back-flushing the first column, the peak for the cHMPA TBDMS derivative was isolated and afforded a clean mass spectrum within 6min. The recoveries of all the derivatives on 2D GC from 1D GC were estimated to be over 66%, and the detection limits were around 10ng/ml of reaction mixture. In the presence of urine extract, the target compounds were not detected as separated peaks in 1D GC-MS mode (except for isobutyl-MPA), and quantification based on extracted ion monitoring could not be achieved. However, 2D GC-MS of the H/C fractions of the target derivatives gave single peaks with well-defined mass spectra, and the recoveries of the derivatives were over 70% except for cHMPA (31% at 1.25μg/ml). Phosphonic acids could be detected at less than 60ng/ml. Sulfuric acid and phosphoric acid also negatively affected the determination of alkyl methylphosphonic acid TBDMS derivatives in 1D GC-MS, and the MPA-TBDMS-derivative peak was completely obscured by the large sulfuric-acid-derivative peak. However, under 1D/2D GC-MS conditions, baseline separation of the MPA derivative and sulfuric acid derivative was achieved, enabling highly sensitive MPA detection at 20ng/ml.