Trilinear analysis of thin-layer chromatography retention of 35 model compounds chromatographed on nine adsorbents with 20 pure solvents

Łukasz Komsta; Robert Skibiński; Natalia Bezpalko; Aleksandra Mielniczek; Barbara Stępkowska

doi:10.1002/jssc.201600705

Trilinear analysis of thin-layer chromatography retention of 35 model compounds chromatographed on nine adsorbents with 20 pure solvents

J Sep Sci. 2016 Nov;39(21):4258-4262. doi: 10.1002/jssc.201600705. Epub 2016 Oct 6.

Authors

Łukasz Komsta¹, Robert Skibiński¹, Natalia Bezpalko¹, Aleksandra Mielniczek¹, Barbara Stępkowska¹

Affiliation

¹ Department of Medicinal Chemistry, Medical University of Lublin, Lublin, Poland.

PMID: 27673481
DOI: 10.1002/jssc.201600705

Abstract

The R_F value dataset of 35 model compounds, chromatographed with 20 pure solvents as the mobile phase each on nine adsorbents: RP2, RP8, RP18, alumina, cellulose, CN, DIOL, NH₂ , and silica, was subjected to trilinear analysis with parallel factor analysis. The two-factor optimal model explained 87% of total information in this complex dataset. The first obtained score (trend) represents two features: the presence of hydrogen bonding and heteroatoms of solute and the mean elution force of the solvent. The second trend represents molecule size, aromaticity, and number of carbons, interconnected with presence of chlorine in mobile phase. The correlation between the scores and molecular descriptors were checked to interpret these trends quantitatively. The scores of adsorbents were slightly intercorrelated, showing NH₂ , alumina, and cellulose as outliers from main adsorbents cloud. The obtained results suggest that molecular size and aromaticity, connected with chlorine atoms in mobile phase, is the second source of retention variability.

Keywords: Chemometrics; Model compounds; Parallel factor analysis.