Phase system selectivity and two-dimensional separations in liquid column chromatography

Correlations between the separation selectivity in aqueous and non-aqueous reversed-phase systems and in normal-phase LC systems were investigated for samples containing different numbers of two repeat structural elements. Such samples are best separated in "orthogonal" two-dimensional chromatographic systems, showing selectivity for one type of the repeat structural element only in the first dimension and for the other structural element only in the second dimension. The number of resolved compounds improves as the degree of orthogonality of the separation systems increases with decreasing correlation between the selectivities for the sample structural distribution in the two dimensions. Orthogonal systems with non-correlated selectivities for each repeat structural element provide the highest number of separated peaks and regular arrangement of the peaks over the two-dimensional retention space according to the individual structural element distribution and the best use of the available peak capacity. Fully orthogonal systems are difficult to find in practice. Partially orthogonal system with correlated selectivities for one structural type distribution, but with one system non-distinguishing the distribution for the other structural element are still useful for the two-dimensional separations. The correlations between the selectivities for repeat regular structural increments were employed to evaluate the suitability of phase systems for two-dimensional HPLC separations. The selectivity correlation in various reversed-phase and normal-phase systems was evaluated for two sample types: (1) Various RP columns show significantly inversely correlated selectivities for acyl lengths and numbers of double bonds distribution, but the differences in the double bond selectivity can be used for practical separations of triacylglycerols with the same equivalent carbon numbers. (2) Synthetic EO-PO block (co)oligomers with two-dimensional distribution of oxyethylene and oxypropylene monomer units were separated according to the two distribution types using on-line two-dimensional reversed-phase-normal-phase LC with a C18 column in the first dimension and an aminopropyl silica column in the second dimension.     

Hypercrosslinked polystyrene as a novel type of high-performance liquid chromatography column packing material. Mechanisms of retention

An experimental material, Chromalite 5HGN (Purolite, UK), that represents hypercrosslinked polystyrene as a new type of neutral stationary phase for HPLC was examined. The material contains no functional groups, but is compatible with any kind of nonpolar and highly polar mobile phase, and even with water. It is chemically resistant and thermally stable. When using aqueous organic mobile phases, Chromalite 5HGN works similar to standard C18 reversed-phase packings, but is characterized by much greater hydrophobicity and, sometimes, unusual selectivity. When using nonpolar mobile phases, i.e. under "quasi normal-phase" conditions, the retention is mostly governed by the interactions between pi-electronic systems of the adsorbent and adsorbate. Adding highly polar, even hydrophilic solvents into the mobile phase, leads to a shift of retention times toward the "reversed-phase" kind of chromatography, which gives an additional possibility in fine tuning the column selectivity.     

Chromatographic behaviour of low molar-mass polyesters in normal-phase high-performance liquid chromatography

Summary The normal-phase chromatographic retention behaviour of polyesters on bare silica and on a polymer-based polyamine (PA) column has been studied with a variely of binary mobile phases under isocratic conditions. The dependence of experimental retention data on the degree of polymerization (p) and on mobile phase composition (φ) was characterized by to an approach developed by Jandera et al. The bulky repeating unit and the relatively highly polar end groups of the polyesters both had a large influence on retention behaviour. The two effects in combination explain the molar-mass-independent retention observed experimentally at a particular mobile phase composition for all the mobile phase—stationary phase combinations investigated. These conditions were found to be independent of the type of end group. End group separation on a silica column improves when the polarity of the less polar solvent is increased. End group separation is better on the PA column because of a greater difference between the adsorption energy of the alcohol and acid end groups. Better prediction of retention data on the PA column was achieved by use of an approach which assumes two different types of adsorption site. Results enabled further understanding of retention behaviour in normalphase gradient polymer-elution chromatography (NPGPEC) and explained both the dependence of the order of elution onp and differences between the end-group selectivity of different systems.     

Stationary and mobile phases in hydrophilic interaction chromatography: a review

Hydrophilic interaction chromatography (HILIC) is valuable alternative to reversed-phase liquid chromatography separations of polar, weakly acidic or basic samples. In principle, this separation mode can be characterized as normal-phase chromatography on polar columns in aqueous-organic mobile phases rich in organic solvents (usually acetonitrile). Highly organic HILIC mobile phases usually enhance ionization in the electrospray ion source of a mass spectrometer, in comparison to mobile phases with higher concentrations of water generally used in reversed-phase (RP) LC separations of polar or ionic compounds, which is another reason for increasing popularity of this technique. Various columns can be used in the HILIC mode for separations of peptides, proteins, oligosaccharides, drugs, metabolites and various natural compounds: bare silica gel, silica-based amino-, amido-, cyano-, carbamate-, diol-, polyol-, zwitterionic sulfobetaine, or poly(2-sulphoethyl aspartamide) and other polar stationary phases chemically bonded on silica gel support, but also ion exchangers or zwitterionic materials showing combined HILIC-ion interaction retention mechanism. Some stationary phases are designed to enhance the mixed-mode retention character. Many polar columns show some contributions of reversed phase (hydrophobic) separation mechanism, depending on the composition of the mobile phase, which can be tuned to suit specific separation problems. Because the separation selectivity in the HILIC mode is complementary to that in reversed-phase and other modes, combinations of the HILIC, RP and other systems are attractive for two-dimensional applications. This review deals with recent advances in the development of HILIC phase separation systems with special attention to the properties of stationary phases. The effects of the mobile phase, of sample structure and of temperature on separation are addressed, too.     

Hydrophilic interaction chromatography using amino and silica columns for the determination of polar pharmaceuticals and impurities

Hydrophilic interaction chromatography (HILIC) is described as a useful alternative to reversed-phase chromatography for applications involving polar compounds. In the HILIC mode, an aqueous-organic mobile phase is used with a polar stationary phase to provide normal-phase retention behavior. Silica and amino columns with aqueous-acetonitrile mobile phases offer potential for use in the HILIC mode. An examination of the retention and separation of several pyrimidines, purines, and amides on silica and amino columns from three manufacturers revealed that mobile phases should contain a buffer or acid for pH control to achieve similar and reproducible results among columns from different sources. Amino columns may also be used in an anion-exchange mode, which provides an advantage for some applications. In some cases, silica can provide different selectivity and better separation than an amino column. Example applications include: low-molecular-mass organic acids and amides as impurities in non-polar drug substances, 5-fluorouracil in 5-fluorocytosine, guanine in acyclovir, and different selectivity for polar basic compounds compared to an ion-pairing system.     

Martin's rule revisited. Its molecular sense and limitations

The linear relation ln k' = Bn + ln A between the retention factor k' in liquid adsorption chromatography (LAC) and the number of repeat units n within a homologous series of oligomers is called Martin's rule. This empirical relation was supported by the retention behavior of the homologous series of different classes of oligomers but had no theoretical justification. In this paper, it is demonstrated that Martin's rule is a consequence of the general theory of liquid chromatography and the molecular sense of coefficients B and A is clarified: B is the Gibbs energy of the repeat unit of the long polymer chain adsorbed at the wall surface, and A is a combination different parameters which characterize the column and the adsorption correlation length H. The theory predicts the deviations from the linear dependence under conditions of weak adsorption between repeat units and stationary phase when H is close to radius of gyration Rg. Experimental data for retention volumes and selectivity of poly(ethylene glycol)s are given for normal and reversed-phase LAC on different columns in acetone-water and methanol-water as mobile phases. These data show excellent agreement between the theory and experiments. It is shown that Martin's rule holds under special conditions, which are theoretically defined by the relation H > Rg/1.5.     

Liquid chromatographic profiles of individual compounds of technical toxaphene

The elution orders of 20 hexa- to nonachlorobornanes and five hexa- to octachlorocamphenes were studied with normal-phase silica and amino phase HPLC, reversed-phase HPLC, as well as gel-permeation chromatography (GPC). Twenty-one compounds of technical toxaphene (CTTs) are commercially available and four were isolated from environmental samples. Structure-activity relationships and chromatographic properties were deduced from the data sets derived on these LC systems. The retention on silica (low-resolution LC and HPLC) increased with the polarity of the CTTs. The elution order of CTTs on amino normal-phase HPLC was, for the most part, the same as on silica normal-phase HPLC. The degree of chlorination determined the elution order of CTTs on C18 RP-HPLC. CTTs eluted from medium-pressure GPC with decreasing molecular size. Chlorobornanes with dichloro substituents on the six-membered ring eluted after the chloroboranes without geminal chlorine atoms on secondary carbons, indicating that these congeners are larger. Altogether, the results increase the knowledge of complex substance class and may serve as a tool in order to gain further standard components.     

Stationary-phase effects in gradient high-performance liquid chromatography

The type of the stationary phase for reversed-phase liquid chromatography significantly affects the sample polarity range that can be covered using gradients of organic solvents in water. The polarity range available for gradient separations of samples containing compounds differing in the lipophilic parts of the molecules can be characterized by "gradient lipophilic capacity", Pl, based on the retention of standard compounds with a repeat lipophilic structural unit, such as a methylene group. The gradient lipophilic capacity is also suitable to characterize the separation possibilities of the columns in non-aqueous reversed-phase gradient elution of strongly non-polar compounds, such as triacylglycerols. In the same way, the suitability of various columns for reversed-phase gradient separations of oligomers can be characterized by "gradient oligomer capacity", as demonstrated in the example of oligo(ethylene glycols). To enable a comparison of the properties of stationary phases independent of column efficiency and dimensions, the gradient lipophilic capacity or the gradient oligomer capacity should be normalized for a "standard" column plate number, gradient range and volume (in column hold-up volume units). The gradient lipophilic capacity or the gradient oligomer capacity and the number of compounds that can be resolved during a gradient run decrease as the initial concentration of the strong solvent in the mobile phase increases and (or) the gradient time decreases. These quantities can be used to select a suitable column and to adjust the optimum gradient profile (the initial composition of the mobile phase and the gradient steepness) with respect to the time of analysis and the number of oligomers or other compounds with regular repeat structural groups that can be resolved during the gradient run.     

Gradient elution in normal-phase high-performance liquid chromatographic systems

Gradient elution is widely used for separation of complex samples in reversed-phase HPLC systems, but is less frequently applied in normal-phase HPLC, where it has a notoriously bad reputation for poor reproducibility and unpredictable retention. This behaviour is caused by preferential adsorption of polar solvents used in mixed mobile phases, which may cause significant deviations of the actual gradient profile from the pre-set program. Another important source of irreproducible retention behaviour is gradual deactivation of the adsorbent by adsorption of even traces of water during normal-phase gradient elution. To avoid this phenomenon, carefully dried solvents should be used. Finally, column temperature should be carefully controlled during normal-phase gradient elution if reproducible results are to be obtained. Working with dry solvents at a controlled constant temperature and using a sophisticated gradient-elution chromatograph, reproducibility of the retention data in normal-phase gradient elution better than 2% may be achieved even over several months of column use. The retention data in gradient elution can be calculated accurately if appropriate corrections are adopted for the gradient dwell volume and for the preferential adsorption of the polar solvents using experimental adsorption isotherms. The average error of prediction for the corrected calculated gradient retention data was lower than 2% for a silica gel column and lower than 3% for a bonded nitrile column, which may be suitable for the optimization of separation. Further, a simple approach is suggested for rapid estimation of changes in the retention induced by a change in the gradient profile in normal-phase HPLC. For such a rough estimation, it is not necessary to know the parameters of the dependence of the solute retention factors on the composition of the mobile phase.     

Overpressured layer chromatography in comparison with thin-layer and high-performance liquid chromatography for the determination of coumarins with reference to the composition of the mobile phase

The retention behaviour of fifteen closely related coumarins in normal-phase overpressured layer chromatography (OPLC) was studied with the aim of comparing the retentions with those in normal-phase thin-layer chromatography (TLC) and high-performance liquid chromatography (HPLC) when optimization of the mobile phase was carried out according to the PRISMA system. The mobile phase optimization was carried out on TLC plates in unsaturated chambers. The resulting mobile phases were transposed to off-line, non-equilibrated OPLC and further to HPLC. The retention in TLC was measured at 37 selectivity points and in OPLC and HPLC at 13 points. Capacity factors (k′) and separation factors () were calculated in order to study the retention behaviour in the different systems. Two- and three-dimensional evaluations of k′ against selectivity points showed similar retention behaviours for the coumarins in TLC, OPLC and HPLC. The values for TLC, OPLC and HPLC showed similar patterns in the three-dimensional evaluations. The retention behaviour at different solvent strengths was also examined. According to quadratic regression, k′ showed a dependence on the change in solvent strength. OPLC, which can be considered as a “planar column” technique, and TLC are closely related methods, whereas HPLC shows a different behaviour in the elution process with regard to solvent strength.     

Molecular Theory of Liquid Adsorption Chromatography

Abstract

A statistical-mechanical theory, based on a lattice model, has been developed to address the molecular mechanism of retention and selectivity in both normal-phase and reversedphase liquid adsorption chromatography. The model is a natural “competitive-equilibrium” one, where possible contributions from solvent-solvent and solute-solvent interactions, and, hence, from solution nonideality, are not neglected. Homogeneous and heterogeneous adsorbent surfaces, single-solvent and binary mixed-solvent mobile phases, and solute molecules of different size and shape are treated. Practical applications of the theory are presented to demonstrate its utility and significance.

For homogeneous adsorbents and neat solvents, the molecular energetics of retention and selectivity are examined, with special emphasis on the effects of solute size and shape, and, relatedly, the modes of solute adsorption. Separations of geometrical isomers and homologous series in real and simulated chromatographic processes are investigated, confirming predictions of the theory and the important role of solvent-solvent and solute-solvent interactions in reversed-phase systems. The implications of a more general retention equation for microscopically heterogeneous adsorbents are discussed. The dependence of capacity ratio on mobile-phase composition for binary solvents is analyzed in some detail. An often important contribution arising from solution nonideality is predicted theoretically. This is shown to be consistent with experimental results on normal-phase and reversed-phase systems.     

LC Enantioseparation of β-Amino Acids on a Crown Ether-Based Stationary Phase

A pentaproline-based chiral stationary phase was prepared and the selectivity of the column was evaluated with 194 racemic compounds in three mobile phase modes: normal-phase mode, polar organic mode and reversed-phase mode. 94 racemates out of 194 were separated and the normal-phase mode proved to be the separation mode of broadest applicability. The column is stable in all common organic solvents and no degradation in column performance was observed in any mode even after more than 1,000 injections. A brief sample loading test was performed on the 250 mm × 4.6 mm column and 13.2 mg of α-methyl-9-anthracenemethanol was baseline separated. Retention behavior in the normal-phase mode and the effect of analyte structure on retention and enantioselectivity are discussed.     

Evaluation of Pentaproline-Based Chiral Stationary Phase by LC

A pentaproline-based chiral stationary phase was prepared and the selectivity of the column was evaluated with 194 racemic compounds in three mobile phase modes: normal-phase mode, polar organic mode and reversed-phase mode. 94 racemates out of 194 were separated and the normal-phase mode proved to be the separation mode of broadest applicability. The column is stable in all common organic solvents and no degradation in column performance was observed in any mode even after more than 1,000 injections. A brief sample loading test was performed on the 250 mm × 4.6 mm column and 13.2 mg of α-methyl-9-anthracenemethanol was baseline separated. Retention behavior in the normal-phase mode and the effect of analyte structure on retention and enantioselectivity are discussed.

     

High throughput screening or drug impurity profiles using gradients in normal phase chromatography

Summary Developmental drug substances and a variety of precursors and process-related impurities were separated using mobile phase gradients under normal phase conditions. These separation conditions produced efficient separations in a fraction of the time necessary for comparable reversed-phase separations. The use of elevated temperatures, polar modifiers and moderate levels of acidic and basic additives allowed highly polar analytes to be eluted with high efficiency and selectivity using fast flow rates. Reproducibility for retention time and peak areas was demonstrated to be very good despite very short equilibration times.     

A unified classification of stationary phases for packed column supercritical fluid chromatography

The use of supercritical fluids as chromatographic mobile phases allows to obtain rapid separations with high efficiency on packed columns, which could favour the replacement of numerous HPLC methods by supercritical fluid chromatography (SFC) ones. Moreover, despite some unexpected chromatographic behaviours, general retention rules are now well understood, and mainly depend on the nature of the stationary phase. The use of polar stationary phases improves the retention of polar compounds, when C18-bonded silica favours the retention of hydrocarbonaceous compounds. In this sense, reversed-phase and normal-phase chromatography can be achieved in SFC, as in HPLC. However, these two domains are clearly separated in HPLC due to the opposite polarity of the mobile phases used for each method. In SFC, the same mobile phase can be used with both polar and non-polar stationary phases. Consequently, the need for a novel classification of stationary phases in SFC appears, allowing a unification of the classical reversed- and normal-phase domains. In this objective, the paper presents the development of a five-dimensional classification based on retention data for 94-111 solutes, using 28 commercially available columns representative of three major types of stationary phases. This classification diagram is based on a linear solvation energy relationship, on the use of solvation vectors and the calculation of similarity factors between the different chromatographic systems. This classification will be of great help in the choice of the well-suited stationary phase, either in regards of a particular separation or to improve the coupling of columns with complementary properties.     

Reversed-phase high-performance thin-layer chromatography of mono-, di- and trisubstituted bile acid methyl esters

A comparative study is reported on separation of series of mono-, di-, and trisubstituted methyl 5β-cholanates, which differ only in the position and stereochemistry of hydroxyl or keto groups at position and stereochemistry of hydroxyl or keto groups at positions C-3, C-7, and/or C-12, by reversed-phase [with chemically-bonded (C-18) silica gel] and normal-phase (silica gel) high-performance thin-layer chromatography (HPTLC). Methnol (or acetonitrile)/water systems were employed as mobile phase. Reversed-phase HPTLC found to be particulary effective for separation of the stereoisomers of di- and trisubstituted compounds whereas the less polar monosubstituted isomers are well resolved in normal-phase HPTLC.     

Two-dimensional liquid chromatography normal-phase and reversed-phase separation of (co)oligomers

Many samples contain compounds with various numbers of two or more regular structural groups. Such "multidimensional" samples (according to the Giddings' notation) are best separated in orthogonal chromatographic systems with different selectivities for the individual repeat structural groups, described by separation factors. Correlations between the repeat group selectivities characterize the degree of orthogonality and suitability of chromatographic systems for two-dimensional (2D) separations of two-dimensional samples. The range of the structural units in that can be resolved in a given time can be predicted on the basis of a model describing the repeat group selectivity in the first- and second-dimension systems. Two-dimensional liquid chromatographic system combining reversed-phase (RP) mode in the first dimension and normal-phase (NP) mode in the second dimension were studied with respect to the possibilities of in-line fraction transfer between the two modes. Hydrophilic interaction liquid chromatography (HILIC) with an aminopropyl silica column (APS) is more resistant than classical non-aqueous NP systems against adsorbent desactivation with aqueous solvents transferred in the fractions from the first, RP dimension to the second dimension. Hence, HILIC is useful as a second-dimension separation system for comprehensive RP-NP LCxLC. A comprehensive 2D RP-NP HPLC method was developed for comprehensive 2D separation of ethylene oxide-propylene oxide (EO-PO) (co)oligomers. The first-dimension RP system employed a 120 min gradient of acetonitrile in water on a C18 microbore column at the flow-rate of 10 microL/min. In the second dimension, isocratic HILIC NP with ethanol-dichloromethane-water mobile phase on an aminopropyl silica column at 0.5 mL/min was used. Ten microliter fractions were transferred from the RP to the HILIC NP system at 1 min switching valve cycle frequency.     

Deformation of gradient shape as a result of preferential adsorption of solvents in mixed mobile phases

Gradient elution has been studied in typical normal and reversed-phase systems. Deformations of gradient profiles have been evidenced as a result of preferential adsorption of modifiers of the mobile phase. This phenomenon was pronounced in the normal-phase system, for which gradient profiles deviated significantly from those programmed. This influenced the retention and shapes of band profiles of the eluting solute. Hence, in order to predict gradient propagation correctly the adsorption equilibrium of modifiers has been quantified. Moreover, at low modifier content, deformations of band profiles of the solute has been registered as a result of the competitive adsorption in the system solute-modifier. This effect has been predicted by a competitive adsorption model. For the reversed-phase systems the influence of the modifier adsorption on gradient propagation was insignificant for typical mobile phases investigated. Therefore, the work has been focused on gradient predictions in the normal-phase system.     

Column selectivity for two-dimensional liquid chromatography

Selectivity of phase system is of primary concern when designing a 2-D separation, as it affects the 2-D system orthogonality and consequently the peak capacity controlling the number of peaks that can be separated in the available 2-D retention space limited by the time of analysis. Possibilities for characterization of LC phase system selectivity with respect to different polar and nonpolar structural units are compared, with special attention to multidimensional samples with various types of repeat groups, such as homopolymers, (co)polymers, fatty acid esters with various acyl lengths and number and position of double bonds, etc. Possibilities of the 2-D LC separations of these and other sample types, including pharmaceuticals, natural phenolic compounds, biopolymers, etc., using various combinations of separation modes are reviewed. Rules for design of comprehensive 2-D LC x LC systems are discussed, with respect to mobile phase compatibility in the two systems and modulation techniques suppressing band broadening connected with the sample fraction transfer from the first to the second dimension. Pitfalls connected with online connection of normal-phase and RP LC systems and their possible practical solutions are addressed and illustrated by practical examples.