We describe a methodology for determining the speciation of a metal with no prior knowledge of the sample. Previous investigations use retention time matching between ICP-MS and API-MS to identify metal species, but struggle to distinguish low level analytes. Our approach uses the HPLC-ICP-MS retention time and peak shape (skew, peak width) to identify corresponding analyte peaks on HPLC-API-MS.

Ten uL of sample were injected into a Dionex Summit HPLC w/ C18 column and the effluent was split between a ThermoFinnigan Element2 ICP-MS (25%) monitoring Se-77, and an ABI QSTAR with APCI source (75%) detecting m/z=80-500. Home-written LabVIEW software curve fit peaks on the HPLC-ICP-MS trace and all QSTAR m/z traces over a user defined interval using the following parameters: Area, Retention Time, Peak Width, Skew.

Serial dilutions of SeMet and SeEt, run by PEMMS, were used as test standards to develop transformations to match ICPMS and QSTAR peaks. Peak parameters for the ICP trace and all QSTAR m/z traces were again determined with samples spiked with yeast. PEMMS successfully detected the [M+H]+ and [M-NH3+H]+ ions for 200 ppb SeMet and 80 ppb SeEt, and rejected >20 co-eluting non-selenium containing QSTAR peaks in each sample.

Analysis of several Se containing mixtures revealed that broccoli had the most complex compliment of molecular forms of Se. S-(methylseleno)-cysteine, recently reported in selenized yeast, and Se-methylselenocysteine, which is implicated as an important species for chemoprevention. This strategy is appropriate for the study of any metal-containing compounds that can be separated chromatographically.