Valve-based comprehensive two-dimensional gas chromatography with quadrupole mass spectrometry detection using a porous layer open tubular column in the first dimension

Abstract

A comprehensive two-dimensional gas chromatography instrument with quadrupole mass spectrometry (GC×GC-qMS) is developed and evaluated, by retrofitting a commercial GC-qMS instrument with a high temperature diaphragm valve (HT-DV) modulator. A porous layer open tubular (PLOT) column in the first dimension (1D), and a more polar wall coated open tubular (WCOT) column in the second dimension (2D) were used. The HT-DV can modulate analytes of any volatility and the combination of the PLOT and WCOT columns provides ample separation selectivity for samples containing an extended range of volatility from highly volatile analytes to analytes of more moderate volatility. In addition to instrument development, we describe data visualization considerations when upgrading from GC-qMS to GC×GC-qMS. Unlike time-of-flight mass spectrometry (TOFMS), the mass spectrum cycle time for qMS (defined as the sum of scan time and interscan delay) is a linear function of the range of mass channels (m/z) analyzed. If the cycle time does not divide evenly into the modulation period, PM, visual distortions will appear in the GC×GC chromatogram. We present two approaches to address this GC×GC chromatogram distortion issue. The first method does not require post-run computations; however, care must be taken in selecting an appropriate mass range to minimize the visual distortion issue. A half of the modulation period of total shift in the 2D separations with a PM of 2 s was determined tolerable for a 15 min separation, with 44 mass ranges (and their accompanying scan rate) deemed suitable out of 350 possible mass ranges. Also, a computational “resampling” method, which can be used to correct any qMS scan rate used was implemented. Figures-of-merit for the HT-DV modulated GC×GC-qMS instrument were also examined. Reproducible peak height (average RSDheight = 9.2%, s = 3.5%) and areas (average RSDarea = 9.3%, s = 2.8%) were obtained with low LODs (average LOD = 4.2 mg/L, s = 2.5 mg/L).