Retentivity, selectivity and thermodynamic behavior of polycyclic aromatic hydrocarbons on charge-transfer and hypercrosslinked stationary phases under conditions of normal phase high performance liquid chromatography

Effect of phenyl numbers in polyphenyl ligand on retention properties of aromatic stationary phases

Aromatic phase, as one type of reversed-phase stationary phases, shows complementary selectivity to the n-alkyl counterparts especially for certain challenging separation tasks. However, effect of phenyl numbers in aromatic ligands on retention behaviors has rarely been addressed compared with the alkyl stationary phases. To illustrate the issue, a series of polyphenyl stationary phases were facially prepared via the coupling chemistry of isocyanate with amine, including aniline (π¹), 4-aminobiphenyl (π²), 4-amino-p-terphenyl (π³) and [1,1':4',1'':4'',1'''-quaterphenyl]-4-amine (π⁴), respectively. The chromatographic behaviors of the new stationary phases as well as the traditional C18 were systematically compared in terms of retention mode, hydrophobic and aromatic selectivity, shape selectivity and π-π interaction by various analytes, including alkylbenzenes, polycyclic aromatic hydrocarbons congeners and substituted benzenes with electron-withdrawing groups. Due to the homologous structure of four polyphenyl ligands, the hydrophobic selectivity, aromatic selectivity and shape selectivity of stationary phases increase with phenyl numbers in the bonded polyphenyl ligands, whereas the increment becomes insignificant from U-π³ to U-π⁴. This phenomenon is explained by the insertion degree of analytes in the polyphenyl ligand brushes. Compared with the homemade C18, the polyphenyl phases indicate insignificant changes of shape selectivity with temperature. Notably, the new polyphenyl phases demonstrate the great selective separation towards the electron-deficient compounds through the π-π interaction. These findings make up for the understanding of the retention behavior of aromatic stationary phases.

Normal-phase liquid chromatography retention behavior of polycyclic aromatic hydrocarbon and their methyl-substituted derivatives on an aminopropyl stationary phase

Retention indices for 124 polycyclic aromatic hydrocarbons (PAHs) and 62 methyl-substituted (Me-) PAHs were determined using normal-phase liquid chromatography (NPLC) on a aminopropyl (NH2) stationary phase. PAH retention behavior on the NH2 phase is correlated to the total number of aromatic carbons in the PAH structure. Within an isomer group, non-planar isomers generally elute earlier than planar isomers. MePAHs generally elute slightly later but in the same region as the parent PAHs. Correlations between PAH retention behavior on the NH2 phase and PAH thickness (T) values were investigated to determine the influence of non-planarity for isomeric PAHs with four to seven aromatic rings. Correlation coefficients ranged from r = 0.19 (five-ring peri-condensed molecular mass (MM) 252 Da) to r = -0.99 (five-ring cata-condensed MM 278 Da). In the case of the smaller PAHs (MM ≤ 252 Da), most of the PAHs had a planar structure and provided a low correlation. In the case of larger PAHs (MM ≥ 278 Da), nonplanarity had a significant influence on the retention behavior and good correlation between retention and T was obtained for the MM 278 Da, MM 302 Da, MM 328 Da, and MM 378 Da isomer sets. Graphical abstract NPLC separation of the three-, four-, five-, and six-ring PAH isomers with different number of aromatic carbon atoms and degrees of non-planarity (Thickness, T). The inserted figure plots the number of aromatic carbon atoms vs. the log I value for the 124 parent PAHs.

Study of the slope of the linear relationship between retention and mobile phase composition (Snyder-Soczewiñski model) in normal phase liquid chromatography with bonded and charge-transfer phases

The Snyder model and the Soczewiñski model are compared on classic NPLC bonded phases using literature data, and on the charge transfer 2, 4-dinitroanilinopropyl (DNAP) column using experimentally collected data. Overall, the Snyder model slightly better predicts the n-slope than the Soczewiñski model. However, both models give comparable uncertainty in predicting n-slope for a given compound. The number of aromatic double bonds was the most suitable descriptor for estimating the relative n-slope of PAHs, as it correlated with behavior better than the number of aromatic rings and is simpler to calculate than the solute adsorption area. On the DNAP phase, a modified Soczewiñski model is suggested to allow for the significant contribution of the aromatic rings to the n-slope. For classic NPLC bonded phases and DNAP columns, the contribution of polar group to the n-slope parallels the adsorption energy of each polar group.