Evaluation of a data-processing method for target and non-target screening using comprehensive two-dimensional gas chromatography coupled with high-resolution time-of-flight mass spectrometry for environmental samples

Abstract

We evaluated the possibility of semi-automatic target and non-target screening by comprehensive two-dimensional gas chromatography coupled with high-resolution time-of-flight mass spectrometry (GC×GC-HRToFMS) by using estuarine sediment samples. Two-way screening consisting of compound-based (target screening) and data-based (non-target screening) analyses was performed by using two original software programs, including mass-spectral deconvolution software based on non-negative matrix factorization, the NIST Mass Spectral Library, and an in-house accurate mass spectral library with a retention index (RI) for the 1st column. Target screening detected 47 organohalogen compounds, including polychlorinated biphenyls, polybrominated diphenyl ethers and organochlorine pesticides, in the sediment sample. Among about 54,000 peaks detected in the non-target screening, 53.6% showed match factors (MF) of ≥ 700 in an automatic mass spectral library search using the peak-top mass spectra. Both mass error of molecular ions and RI error were calculated automatically for the first-hit compounds in the library search; 0.03% of peaks passed all criteria, namely MF ≥ 700, mass error ≤± 20 ppm, RI error ≤ ± 20, and intensity of molecular ion ≥ 10,000. Two compounds—a phosphorus flame retardant and a brominated phenol—were tentatively identified in the non-target screening process. The advantages of semi-automatic GC×GC-HRToFMS data processing with the two original software programs is its simplicity, high reproducibility, and shortened time for processing a large volume data. The results suggest that an In-source HiRes search—i.e. library matching using accurate mass spectra—is useful and could be applied for non-target screening using high-resolution MS in the future.