Differentiation of Aroclors in environmental samples using linear discrimination
Environmental samples suspected of containing polychlorinated biphenyls (PCBs) and analyzed by Environmental Protection Agency (EPA) methods frequently contain non-PCB components, such as phthalates, polynuclear aromatic hydrocarbons, or organochlorine pesticides. The presence of these interferences can often obscure the gas chromatographic (GC) patterns and cause problems in differentiating the Aroclor types by visual inspection. Because EPA methods require the identification of Aroclor types in order to determine which Aroclor standard should be used for the quantitation calibration, and possibly, to trace the source of PCB occurrences, electron capture negative ion chemical ionization (ECNICI) mass spectrometry was used to provide additional parameters for discriminating PCB congeners from the interferences. In this study, eight Aroclors (i.e., 1016, 1232, 1242, 1248, 1254, 1260, 1262, and 1268) were analyzed by ECNICI for a range of concentration levels. Using selective ion display, the ion abundances of the prominent peaks in the isotopic clusters of molecular ions were measured for eight PCB homologs (Cl{sub 2}-Cl{sub 9}) within their GC retention time windows for each sample. Corresponding relative ion abundances from eight Aroclor standards were used as classification training sets. These training sets were used to develop a classification algorithm using standard linear discriminant analysis to classify the PCBs present in an unknown sample(s) as a specific Aroclor. This technique employed a sequential application of two sets of linear discriminant functions to successfully identify known EPA quality control and environmental samples. However, classifications of Aroclor mixtures were less successful.
- Research Organization:
- Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
- Sponsoring Organization:
- USDOE, Washington, DC (United States)
- DOE Contract Number:
- AC05-84OR21400
- OSTI ID:
- 110686
- Report Number(s):
- CONF-9408107-7; ON: DE96000248; TRN: 95:007262
- Resource Relation:
- Conference: American Statistical Association, Toronto (Canada), 14-18 Aug 1994; Other Information: PBD: [1995]
- Country of Publication:
- United States
- Language:
- English
Similar Records
PCB field screening using the D TECH{trademark} immunoassay field test
PCB congener analysis with Hall electrolytic conductivity detection