Mobile Phase

8.3.1 The Influence of the Organic Modifier on RM

Mobile phases represent mixtures of water (or buffer) and an organic solvent (modifier) of various composition. The application of binary mixtures is mandatory even on those stationary phases, allowing the use of pure water. Solutes with intermediate or high lipophilicity would exhibit marginal migration distances in pure water, counteracting their exact measurement.

The most commonly used organic modifiers are methanol, acetone, and acetoni-trile, while the use of dioxane or tetahydrofuran is less frequent. Braumann [27] recommends methanol as the modifier of choice due to the quite pronounced similarity of its physico-chemical properties to water. It provides both strong hydrogen-bond donor and acceptor properties so that the addition to an aqueous mobile phase over a wide range of volume fractions will change the ordering of water molecules only in a limited manner. Our own experience supports the above recommendation (Fig. 1).

From a theoretical point of view, the extrapolation of Ru values, obtained with different solvent systems, to /?Mw (i.e., modifier-free conditions) should yield identical results for a given solute. This was experimentally proven by Biagi et al. [15] on silicone oil impregnated silica-gel plates. From their experimental findings the latter authors conclude, that it is solely the surface tension of the modifier which impacts the Rm values [15]. On the other hand, Cserhati [28] using paraffin-coated silica-gel showed substantial differences in RMv/ depending on the modifier used (methanol, acetone, oi acetonitrile).

Retention on stationary phases with free silanol groups, such as ODS, will also be due to silanophilic hydrogen bond and dipole interactions, which vary with the modifier used [29, 30]. In these cases the electronic and H-bonding properties of the modifier will influence the RM values in addition to surface tension.

8.3.2 The Influence of Solvent pH and Ionic Strength on RSi

The pronounced impact of ionization on partitioning processes is undisputed. As far as TLC is concerned, Biagi et al. [31] using silicone oil-impregnated plates and Cserhati [28] using paraffin-impregnated silica-gel showed that RM depends on pH. Also de Voogt and coworkers [32] call for proper pH buffering in the case of ionizable solutes. In most cases investigators avoid pK correction by using experimental pH values which surmount the pH of the test compounds by at least 2 log units.

Several investigators have observed deviations from the generally applied rules for pK correction in TLC. For some organic acids Wilson [33] detected no variation in RF related to pH (2-11). Negligible impact of salt concentration and pH (3-12) on retention was reported by Cserhati et al. [34] for some peptides and amino acids, using paraffin-impregnated silica-gel as the stationary phase. Kovacs-Hadady and Szilagyi [35], using tricaprylmethylammonium-impregnated silica-gel plates, also found no effect of ionic strength and pH (2.4-9.4) on retention of minoxidil and its intermediates. Varying the pH betweenl and 13 does not affect the retention of weak acids and bases on RP18 silica-gel, as reported by Dross et al. [11]. According to these authors stronger bases (pK 8.0-10.7) exhibit no pH-related variation in mobile phases with high water content (75 %), whereas some pH-related effects are seen in the case of reduced water content (40%). This latter finding is attributed by the authors to polar adsorption, which becomes more prominent in solvents with low water content. The pH-dependence of polar adsorption was shown by Cserhati and Szogyi [36] as well as Dross et al. [24], while Dingenen and Pluym [37] speculated on the influence of silanol dissociation in this context.

Above-described observations might be explained on the basis of concepts of Hor-vath et al. [38] on solvophobic interactions with octadecyl silica-gel. According to these authors it is solely the "hydrocarbonaceous moiety" of the solute "that can enter into hydrophobic interactions with the octadecyl chains of the stationary phase". Thus, the polar groups of the solute would remain outside the lipidic part of the stationary phase; correspondingly, ionization should not influence the chromatographic process. These conditions contrast profoundly with partitioning processes due to shaking out into an organic phase.

Polar interactions, which take place between the solute and SiOH groups of the plate material, are labeled silanophilic effects [39, 40], An effect of ionization on partitioning behavior is therefore expected in the case of strong bases in modifier systems containing high amounts of methanol.

Whether the effects of ionization under the experimental conditions of HPLC and TLC are similar or not remains to be clarified. Even in the case of identical stationary phases, an essential difference between TLC and HPLC is the fact that for HPLC the columns have to be equilibrated before the runs. Correspondingly, the stationary phase of the column is adjusted to the pH of the solvent, while in the case of a TLC-plate a pH gradient is formed.

0 0

Post a comment