Enhanced Target Identity Confirmation In 8270D Using Deconvoluted Spectral Matching

Overcoming Legacy Obstacles with Innovative Approaches
Oral Presentation

Presented by D. Walker
Prepared by

Contact Information: bruce_quimby@agilent.com; 302-636-3667


Analyte identification in methods like EPA 8270D relies on the measured ratio of response at a compound’s target ion to the response of (typically) 2 qualifier ions in the retention time window for that compound. Identity can be further confirmed by comparing a baseline subtracted apex spectrum of the analyte to a reference spectrum. These approaches work well when there are no chromatographic interferences with the analyte. Many samples, however, contain significant levels of matrix compounds that can interfere with both processes above. In these cases, analyte identification becomes more challenging.

This talk describes adding deconvoluted spectral matching to the data analysis method for 8270D. Spectral deconvolution is a long-used software approach to removing the ions of co-eluting compounds from the spectrum of an analyte. In deconvolution, ion chromatograms are extracted at all masses in the scan range. Ions with peaks having the same shape and RT are grouped into components. The response of Ions present in multiple overlapping peaks are apportioned to each peak using a process similar to that in chromatographic integrators. Spectra are then constructed from the components. The deconvolution process greatly reduces or eliminates interfering ions in the analyte spectra.

Deconvolution is used in two ways. First, in data analysis quantitation software, the deconvoluted spectrum is presented with the reference spectrum and a library match score. These can be displayed with the extracted ion chromatograms of the target and qualifier ions. This is very useful when reviewing high matrix samples in deciding if an analyte is present or absent. The second use is for identifying compounds not associated with the calibration standards. In this case, the deconvoluted spectra of all peaks in the chromatogram can be searched against libraries like NIST.