Main content area

Extraction and quantitative analysis of water by GC/MS for trace-level dimethylsilanediol (DMSD)

Xu, Shihe
Journal of chromatography 2019 v.1600 pp. 1-8
derivatization, detection limit, gas chromatography-mass spectrometry, quantitative analysis, salt content, silicone, soil, solid phase extraction, solvents, water analysis, water solubility
Dimethylsilanediol (DMSD) is related to the most important bifunctional building block for silicone oligomers and polymers, although DMSD itself is not used in any commercial applications. The environmental release of DMSD is linked to the hydrolytic degradation of other silicone materials in soil and water as DMSD is usually one of the major products. Most common extraction and quantification methods are not suitable for the analysis of trace- and ultratrace-levels of DMSD in water. This is because DMSD is highly water soluble and can readily undergo self-condensation when concentrated. In addition, DMSD may also coexist with DMSD precusors such methylsiloxanes in water. In the present study, solid-phase extraction (SPE) in combination with gas chromatography-mass spectrometry (GC/MS) without pre-column derivatization was tested for analyzing water samples for DMSD. It was found that direct analysis by GC/MS can be used for a wide range of concentrations if DMSD was extracted into a dry organic solvent. Isolute® ENV + may be used for such extraction at higher DMSD concentrations, while Supelclean™ ENVI-Carb™ Plus was found to be better for trace and ultratrace analysis. Increased salt content in water can increase its DMSD extraction efficiency, while polarity of the eluting solvents is a determining factor for eluting efficiency. Moisture in the final extract is a detrimental factor for direct GC/MS analysis. With a proper moisture removal procedure and a suitable internal standard, coupling of SPE and direct GC/MS analysis reduces the method detection limits for DMSD to lower ppb levels. Based on field sample analysis, solvent and instrumental background, not instrumental sensitivity, was found to be the limiting factor in lowering the detection limits for this analysis.