Using aromatic selectivity to facilitate difficult separations

Typical alkyl-bonded reversed phase columns (C18 and C8) do not always offer the necessary selectivity needed to separate complex mixtures. In many cases, challenging method development procedures such as gradients, high pH mobile phases and ion-pairing reagents may be required to obtain critical separations.

The unique selectivity offered by phenyl phases can provide alternative retention characteristics, resulting in separations not achievable on typical alkyl-bonded phases. This selectivity is due in part to interactions between the p electrons of the phenyl ring of the bonded phases and the p electrons of the sample analyte which are typically from the presence of aromatic groups.

The p-p interaction is a type of electron donor - electron acceptor interaction that in a chromatographic system can occur between the phenyl stationary phase and the sample analyte. This interaction, which is a bit stronger than van der Waals forces and equally as important as other noncovalent interactions such as hydrogen bonding and dipole-dipole interactions, can lead to increased retention for polar aromatic compounds compared to that which is typically observed for alkyl-bonded phases. The net result of such compound specific retention is a change in selectivity between different compounds, which can be used to advantage to perform difficult separations where peaks co-elute on a C18 phase. In this study the retention behaviour between a C18 and Phenyl phase were compared using a mixture of polar and non-polar probes with varying degrees of aromaticity in the methanol and acetonitrile mobile phases.

Materials and methods

Analyses were performed using an Agilent Technologies HP 1100 LC system equipped with a UV detector. The HPLC columns used were a Phenomenex Gemini 5 µm C18 and a Phenomenex Gemini 5 µm C6-Phenyl, 150 x 4.6 mm. All standards were purchased from Sigma Chemicals and solvents were purchased from Fisher Scientific.

Isocratic HPLC runs were performed using either HPLC grade water or 20 mM potassium phosphate pH 2.5, and methanol or acetonitrile were used as the organic modifier for the mobile phase. Column temperature was maintained at 30°C and elution of peaks was monitored by UV (wavelength noted in Figures 1 and 2).

Results and discussion

The work examined selectivity differences between a C18 bonded phase and a phenyl-bonded phase. The effect the organic mobile phase had on these selectivity differences was also explored. Chromatograms of Gemini 5 µ C6 Phenyl comparing methanol and acetonitrile are shown in Figures 1 and 2. (Scott Waite, Krisna Kallury, and Michael McGinley, Phenomenex Inc, Torrance, CA, USA).

In Figure 1 the retention behaviour of a mixture of flavanoids was compared. When using methanol at a concentration of 55%, kaempferol and isorhamnetin were easily baseline resolved whereas when 40% acetonitrile was used, this pair of structurally similar compounds co-eluted.

Figure 2 further demonstrates selectivity differences between methanol and acetonitrile mobile phases on the Gemini C6 Phenyl phase, and how such differences affect the separation of polar compounds typically found in food additives. Despite the similar elution strength of the two solvents (25% acetonitrile and 35% methanol), all of the compounds elute later when methanol is used. This is evidence of additional interactions leading to increasing retention.

These interactions affect different analytes differently. This is shown by the elution order change between saccharine and p-hydroxybenzoic acid (peaks 1 and 2), as well as dehydroacetic acid and methylparaben (peaks 4 and 5).

Figure 3 shows data generated with a mixture of probes (Table 1), comparing the retention (k) when using methanol or acetonitrile. In this evaluation, the comparison of organics in the mobile phase suggest that methanol contributes to an increase in the p-p interactions of the phenyl-phase, thus allowing improved selectivity for a diverse mix of analytes. Aromatic analytes such as sulfamethoxazole and metoprolol exhibited strong retention differences when methanol was used where as nalidixic acid (which is non-aromatic) showed little or no change. Both the methanol and acetonitrile containing mobile phases were adjusted to equal eluotropic strength.

Figure 4 compares the retention behaviour of Gemini 5 µm C6 Phenyl and Gemini 5 µm C18. In this evaluation, methanol and 20 mM potassium phosphate were used as the mobile phase for comparing k (relative retention) of a mixture of polar and non-polar aromatic probes. The data shows that hydrophobic retention mechanisms are able to retain certain compounds such as indomethacin, norpsuedoephedrine and pentylbenzene where as p-p interactions combined with hydrophobic retention offer better selectivity for compounds such as sulfamethoxazole, saccharin and diphenhydramine.

Conclusion

Phenyl phases like Gemini C6-Phenyl offer differences in selectivity when compared to C18 columns. Much of this selectivity difference is attributable to aromatic selectivity (p-p orbital interactions between the phase and analyte); which allows users to potentially separate compounds based on differences in their aromatic structure.

Table 2: Compounds compared on C18 and Phenyl phase.

Further, the use of different organic mobile phases allows users to activate (methanol) or suppress (acetonitrile) such aromatic interactions. Modulating such interactions can selectively change the retention of a specific compound in a mixture, resulting in improved resolution. Such flexibility and utility make Phenyl phases like Gemini C6-Phenyl a powerful method development tool for separations where C18 columns fail to provide the desired separation.