Organotin speciation in environmental matrices by automated on-line hydride generation-programmed temperature vaporization-capillary gas chromatography-mass spectrometry detection

In the present contribution, a new automated on-line hydride generation methodology was developed for dibutyltin and tributyltin speciation at the trace level, using a programmable temperature-vaporizing inlet followed by capillary gas chromatography coupled to mass spectrometry in the selected ion-monitoring mode acquisition (PTV-GC/MS(SIM)). The methodology involves a sequence defined by two running methods, the first one configured for hydride generation with sodium tetrahydroborate as derivatising agent and the second configured for speciation purposes, using a conventional autosampler and data acquisition controlled by the instrument's software. From the method-development experiments, it had been established that injector configuration has a great effect on the speciation of the actual methodology, particularly, the initial inlet temperature (−20 °C; He: 150 ml/min), injection volume (2 μl) and solvent characteristics using the solvent venting mode. Under optimized conditions, a remarkable instrumental performance including very good precision (RSD < 4%), excellent linear dynamic range (up to 50 μg/ml) and limits of detection of 0.12 μg/ml and 9 ng/ml, were obtained for dibutyltin and tributyltin, respectively. The feasibility of the present methodology was validated through assays upon in-house spiked water (2 ng/ml) and a certified reference sediment matrix (Community Bureau of Reference, CRM 462, Nr. 330 dibutyltin: 68 ± 12 ng/g; tributyltin: 54 ± 15 ng/g on dry mass basis), using liquid-liquid extraction (LLE) and solid-phase extraction (SPE) sample enrichment and multiple injections (2 × 5 μl) for sensitivity enhancement. The methodology evidenced high reproducibility, is easy to work-up, sensitive and showed to be a suitable alternative to replace the currently dedicated analytical systems for organotin speciation in environmental matrices at the trace level.