No abstract

The influence of coding errors and the matching criterion being used upon the performance of a retrieval system for binary coded spectra has been investigated. A file of 9628 mass spectra, including 773 doublets, originating from an MSDC collection, was used. The spectra were reduced to 120 binary coded m/e values, selected by using the information content as a criterion. It is concluded that the performance of the retrieval primarily depends on the extent of errors occurring in the coded spectra and is hardly affected by the matching criterion being used.

Five species of halophytes were sampled in the salt marshes of the Tagus estuary, dried, ground and digested. They were further extracted with ethanol and the extracts passed through weak and strong cationic ionexchange resins, purified through TLC and submitted to pyrolysis mass spectrometry and HPLC-ICP/MS. Arsenic content and hydrideforming arsenic species were verified, in each step, by GF-AA and HG-QFAA. A high content of arsenic was found in the samples of halophytes studied, both di-and tri-methylated arsenic compounds being present. A considerable fraction of this arsenic content seems to be refractory to hydride generation. Moreover, the arsenic fraction found seems to have the same ion-exchange behaviour as the refractory fractions formerly studied in estuarine water.A partial characterization of these structures by pyrolysis-GC-MS suggests the presence of arsenobetaine and arsenocholine compounds. Furthermore, HPLC-ICP/MS data seem to confirm the presence of these compounds. In addition, the latter hyphenated technique strongly suggests the presence of a number of other organoarsenicals including tetramethylarsonium (TMAs), trimethylarsine oxide (TMAO), cacodylate (DMA) and possibly an arsenosugar-type compound.

No abstract

No abstract