Reproducibility Of The LC×LC System Analysis

The reproducibility of the LC×LC system was validated by testing the retention stability of both dimensions as retention time alignment is a crucial part in unknown contaminants identification from environmental samples. The validation were performed by separating EPA 531.1 carbamate pesticide mixture utilizing the LC×LC system, repeating 20 time in different days. The LC×LC data was processed in GC×GC software. In the software, the chromatogram was smoothed first column by 0.1 and second column by 1.0. For Blob detection function, the thresholds were set minimum area as 200, minimum volume as 15000 and minimum peak as 140.0 (relative reference). In the chromatogram, 6 corresponding peaks (A-F) were detected by the above mentioned algorithm
The mixture comprises 21 PAHs (Acenaphthene, Acenaphthylene, Anthracene, Benz[a]anthracene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[ghi]perylene, Benzo[a]pyrene, Chrysene, Dibenz[a,h]anthracene, Dibenzo[a,e]pyrene, 7,12-Dimethylbenz[a]anthracene, Fluoranthene, Fluorene, Indeno[1,2,3-cd]pyrene, 3-Methylcholanthrene, 2-Methylnaphthalene, Naphthalene, Perylene, Phenanthrene, Pyrene), 11 benzene derivatives (Acetophenone, Aniline, Benzoic acid, Benzyl alcohol, 2,4-Dimethylphenol, Hydroquinone, 2-Methylphenol, 3-Methylphenol, 4-Methylphenol, Phenol, Resorcinol), 4 heterocyclic compounds (Dibenzo[a,j]acridine, Dibenzofuran, 2-Picoline, Pyridine) and 2 α, β-unsaturated cyclic ketone (p-Benzoquinone, Isophorone). The solution injected was diluted 10 times from the original standard mixture (in toluene) into acetonitrile and water (50:50, v/v). The LC×LC separation was carried out by a 150mm C18 column as the first dimension and a 50mm Kinetex PFP column as the second dimension, in order to perform fast separation. The contour plot of LC×LC separation was shown in Fig. 3 with peak detection performed by GC×GC software. For peak detection, the template matching function of GC×GC was applied. The monocyclic compounds was eluted first in LC×LC due to their lower Kow value. The UV detection was operated at 290nm to avoid the disturbance from toluene which has the UV cutoff wavelength at 285nm.
For orthogonality measurement in two dimensional chromatography, quite a few metrics have been reported so far, including Pearson’s correlation coefficients [20], mutual information [21], Stoll and Gilar’s box-counting methods [22], and different surface coverage measurement [23-25]. In this study, the orthogonality was estimated by surface coverage and also the very recently reported asterisk methods [26], which the latter algorithm is by calculating the spreading of the peaks throughout the multidimensional separation space. In both methods, the retention time of the two dimensions 1tR and 2tR of all the peaks detected in were normalized by Eq.