Chemometric decomposition of comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry data employing partial modulation in the negative pulse mode

Abstract

Partial modulation in the negative pulse mode was optimized for comprehensive two-dimensional gas chromatography (GC × GC) coupled to time-of-flight mass spectrometry (TOFMS). With partial modulation in the negative pulse mode, a flow of auxiliary carrier gas is applied through a pulse valve to a T-junction joining the first (1D) and second (2D) dimension columns for nearly all of the modulation period (PM). This results in dilution of the 1D eluate, followed by briefly turning off the pulse valve for a specified pulse width (pw), effectively “injecting” undiluted 1D eluate onto the 2D column. The raw data has the appearance of 2D separations superimposed on top of the 1D separation. While high peak capacity GC × GC data are produced, there are challenges that needed to be addressed regarding the use of chemometrics for analyte decomposition, identification, and quantification. Herein, these data analysis challenges are addressed using multivariate curve resolution - alternating least squares (MCR-ALS). An isothermal separation of a 15-component mixture of similar compounds is obtained in 20 s using a PM = 250 ms and a pw = 6 ms. Various peak overlap situations were purposely produced to facilitate the chemometric method demonstration. The MCR-ALS chromatographic loadings (peak profiles) were used to determine retention times (tR) and width-at-base (Wb) in both separation dimensions. MCR-ALS readily decomposed 1D × 2D regions for analytes with severe 1D overlap if they were fully resolved on 2D. Decomposition was more challenging for analytes severely overlapped on both GC × GC dimensions when their spectra were similar, though ultimately all 15 compounds were successfully decomposed. Additionally, the 2D concentration peak profiles obtained via MCR-ALS are demonstrated to be reliable for identification and quantifiable. The predicted versus prepared concentration values are in good agreement for two representative analytes, with percent deviation values of −5.6% (±2.2%) for 1-hexene, and 1.8% (±3.4%) for 2-pentanone.