Non-Target Chemical Analysis Using Liquid Chromatography, Differential Ion Mobility and Tandem Mass Spectrometry
Identification of trace unknown analytes in complex samples remains a significant challenge for analytical chemistry. Mass spectrometry (MS) and analytical separations techniques can now be used to develop and support a new analytical strategy called non-target analysis which aims to provide comprehensive identification and quantification of all detectable chemical species in a complex sample. This thesis addresses challenges currently limiting the utility of this non-target approach by developing analytical methods for acquiring MS data suitable for identification of trace unknowns and investigating current tools available for unknown identification from MS spectral data. Liquid chromatography (LC) - MS, a widely used technique in trace analysis, was used to develop an analytical method capable of simultaneously acquiring high resolution MS and tandem mass spectrometry (MS/MS) data for hundreds of metabolites in urine. An emerging separation technique called high field asymmetric waveform ion mobility spectrometry (FAIMS) was also investigated, as an alternative to LC, for the identification of non-target analytes in urine. Modifications were carried out to the FAIMS-MS source interface allowing for transmission of small metabolite ions from FAIMS to MS. The challenge of direct electrospray (ESI) in urine analysis using ESI-FAIMS-MS was addressed by using sample dilution and extending MS data acquisition time using FAIMS. This allowed for higher quality MS data to be acquired for low abundance urinary metabolites than was possible by LC-MS and the complete elimination of ionization suppression in dilute urine samples. Insight gained into ESI suppression in complex samples allowed for two methods of semi-quantification to be proposed for non-target analytes in complex samples without using unavailable chemical standards. To address the challenge of unknown identification, faced throughout this thesis, an integrated approach was implemented to identify metabolites based only on spectral data without the usual requirement of availability of chemical standards. This approach combined spectral libraries, literature reports on ion chemistry and de novo identification based on gas phase ion chemistry with a detailed fragmentation study on nucleic acid bases, notably protonated uracil. Together, the instrumental methods and approaches to data analysis described allowed for the identification of 110 abundant chemical species detected in urine.