Contributions to reversed-phase column selectivity: III. Column hydrogen-bond basicity.

Column selectivity in reversed-phase chromatography (RPC) can be described in terms of the hydrophobic-subtraction model, which recognizes five solute-column interactions that together determine solute retention and column selectivity: hydrophobic, steric, hydrogen bonding of an acceptor solute (i.e., a hydrogen-bond base) by a stationary-phase donor group (i.e., a silanol), hydrogen bonding of a donor solute (e.g., a carboxylic acid) by a stationary-phase acceptor group, and ionic. Of these five interactions, hydrogen bonding between donor solutes (acids) and stationary-phase acceptor groups is the least well understood; the present study aims at resolving this uncertainty, so far as possible. Previous work suggests that there are three distinct stationary-phase sites for hydrogen-bond interaction with carboxylic acids, which we will refer to as column basicity I, II, and III. All RPC columns exhibit a selective retention of carboxylic acids (column basicity I) in varying degree. This now appears to involve an interaction of the solute with a pair of vicinal silanols in the stationary phase. For some type-A columns, an additional basic site (column basicity II) is similar to that for column basicity I in primarily affecting the retention of carboxylic acids. The latter site appears to be associated with metal contamination of the silica. Finally, for embedded-polar-group (EPG) columns, the polar group can serve as a proton acceptor (column basicity III) for acids, phenols, and other donor solutes.

Contributions to reversed-phase column selectivity. II. Cation exchange.

The contribution of cation exchange to solute retention for type-B alkylsilica columns (made from high-purity silica) has been examined in terms of the hydrophobic-subtraction (H-S) model of reversed-phase column selectivity. The relative importance of cation exchange in the separation of ionized bases by reversed-phase chromatography (RPC) varies with (a) column acidity (values of the column cation-exchange capacity C), (b) mobile-phase pH and buffer concentration, and (c) the nature of the buffer cation. The effects of each of these separation variables on cation retention were examined. The contribution of cation exchange (and other ionic interactions) to solute retention is represented in the H-S model by properties of the solute (κ') and column (C), respectively. Values of κ' for 87 solutes have been examined as a function of solute molecular structure, and values of C for 167 type-B alkylsilica columns have been related to various column properties: ligand length (e.g., C(8) vs. C(18)) and concentration (µmol/m(2)), pore diameter (nm), and end-capping. These results contribute to a more detailed picture of the retention of cationic solutes in RPC as a function of separation conditions. While previous work suggests that the ionization of type-B alkylsilica columns is generally negligible with mobile-phase pH<7 (as a result of which cation exchange then becomes insignificant), the present study provides evidence for cation exchange (and presumably silanol ionization) at a pH as low as 3 for most columns.

Contributions to reversed-phase column selectivity. I. Steric interaction.

In reversed-phase chromatography (RPC), the restricted retention of "bulky" solutes can occur in one of two ways, giving rise to either "shape selectivity" or "steric interaction." Starting with data for 150 solutes and 167 monomeric type-B alkylsilica columns, the present study examines the steric interaction process further and compares it with shape selectivity. The dependence of column hydrophobicity and steric interaction on column properties (ligand length and concentration, pore diameter, end-capping) was determined and compared. The role of the solute in steric interaction was found to be primarily a function of solute molecular length, with longer solutes giving increased steric interaction. We find that there are several distinct differences in the way shape selectivity and steric interaction are affected by separation conditions and the nature of the sample. Of particular interest, steric interaction exhibits a maximum effect for monomeric C(18) columns, and becomes less important for either a C(1) or C(30) column; shape selectivity appears unimportant for monomeric C(1)-C(18) columns at ambient and higher temperatures, but becomes pronounced for C(30) - as well as polymeric columns with ligands ≥C(8). One hypothesis is that shape selectivity involves the presence or creation of cavities within the stationary phase that can accommodate a retained solute (a primarily enthalpic process), while steric interaction mainly makes greater use of spaces that pre-exist the retention of the solute (a primarily entropic process). The related dependence of hydrophobic interaction on column properties was also examined.

Selecting an "orthogonal" column during high-performance liquid chromatographic method development for samples that may contain non-ionized solutes.

Several reasons can exist for a change in reversed-phase selectivity, and several separation conditions can be changed for this purpose. In the present investigation, a change in column is considered for the improved separation of non-ionized solutes only. Differences in column selectivity (and the selection of "orthogonal" columns) can be predicted on the basis of the hydrophobic-subtraction model. For ionized solutes, the F(s)-parameter is used to predict orthogonality. For use with non-ionized solutes, it is suggested that the cation-exchange term of the model be dropped [F(s)(-C)] for better predictions. For samples containing both ionized and non-ionized solutes, F(s) and F(s)(-C) should be used together for the best results. Predicted separations involving 64 non-ionized solutes and maximally different columns from a 400-column database were used to validate this procedure.

Anion-exchange behavior of several alkylsilica reversed-phase columns.

Some alkylsilica columns carry a positive charge at low pH, as determined by anion-exchange with nitrate ion. In the present study, the relative positive charge for 14 alkylsilica columns was measured for a mobile-phase pH 3.0. All but 3 of these columns were found to carry a significant positive charge under these conditions. The relative positive charge on these columns was found to correlate approximately with two other column characteristics: relative cation-exchange behavior as measured by the hydrophobic-subtraction model (values of C-2.8), and slow equilibration of the column to changes in the mobile-phase-as evidenced by a slow change in the retention of anionic and cationic solutes with time. The origin of this positive charge may arise from the bonding process, with incorporation of some cationic entity into the stationary phase.

Characterization and applications of reversed-phase column selectivity based on the hydrophobic-subtraction model.

A total of 371 reversed-phase columns have now been characterized in terms of selectivity, based on five solute-column interactions (the hydrophobic-subtraction model). The present study illustrates the use of these data for interpreting peak-tailing and column stability. New insights are also provided concerning column selectivity as a function of ligand and silica type, and the selection of columns for orthogonal separations is re-examined. Some suggestions for the quality control of reversed-phase columns during manufacture are offered.

"Orthogonal" separations for reversed-phase liquid chromatography.

A general procedure is proposed for the rapid development of a reversed-phase liquid chromatographic (RP-LC) separation that is "orthogonal" to a pre-existing ("primary") method for the RP-LC separation of a given sample. The procedure involves a change of the mobile-phase organic solvent (B-solvent), the replacement of the primary column by one of very different selectivity, and (only if necessary) a change in mobile phase pH or the use of a third column. Following the selection of the "orthogonal" B-solvent, column and mobile phase pH, further optimization of peak spacing and resolution can be achieved by varying separation temperature and either isocratic %B or gradient time. The relative "orthogonality" of the primary and "orthogonal" RP-LC methods is then evaluated from plots of retention for one method versus the other. The present procedure was used to develop "orthogonal" methods for nine routine RP-LC methods from six pharmaceutical analysis laboratories. The relative success of this approach can be judged from the results reported here.

Relevance of pi-pi and dipole-dipole interactions for retention on cyano and phenyl columns in reversed-phase liquid chromatography.

Previous work suggests that pi-pi interactions between certain solutes and both phenyl and cyano columns can contribute to sample retention and the selectivity of these two column types versus alkylsilica columns. Recent studies also suggest that dipole-dipole interactions are generally unimportant for retention on cyano columns. The present study presents data for 44 solutes, three columns and two different mobile phases that were selected to further test these conclusions. We find that pi-pi interactions can contribute to retention on both cyano and phenyl columns, while dipole-dipole interactions are likely to be significant for the retention of polar aliphatic solutes on cyano columns. When acetonitrile/water mobile phases are used, both pi-pi and dipole-dipole interactions are suppressed, compared to the use of methanol/water.

Column selectivity in reversed-phase liquid chromatography. VII. Cyanopropyl columns.

Eleven cyanopropyl ("cyano") columns were characterized by means of a relationship developed originally for alkyl-silica columns. Compared to type-B alkyl-silica columns (i.e., made from pure silica), cyano columns are much less hydrophobic (smaller H), less sterically restricted (smaller S*), and have lower hydrogen-bond acidity (smaller A). Because sample retention is generally much weaker on cyano versus other columns (e.g., C8, C18), a change to a cyano column usually requires a significantly weaker mobile phase in order to maintain comparable values of k for both columns. For this reason, practical comparisons of selectivity between cyano and other columns (i.e., involving different mobile phases for each column) must take into account possible changes in separation due to the change in mobile phase, as well as change in the column.

Column selectivity in reversed-phase liquid chromatography. VIII. Phenylalkyl and fluoro-substituted columns.

As reported previously, five solute-column interactions (hydrophobicity, steric resistance, hydrogen-bond acidity and basicity, ionic interaction) quantitatively describe column selectivity for 163 alkyl-silica, polar-group and cyano columns. In the present study, solute retention and column selectivity for 11 phenyl and 5 fluoro-substituted columns were compared with alkyl-silica columns of similar ligand length. It is concluded that two additional solute-column interactions may be significant in affecting retention and selectivity for the latter columns: (a) dispersion interactions of varying strength as a result of significant differences in bonded-phase polarizability or refractive index and (b) pi-pi interactions in the case of phenyl columns and aromatic solutes. These 16 phenyl and fluoro columns were also characterized in terms of hydrophobicity, steric resistance, hydrogen-bond acidity and basicity, and ionic interaction.

A fast, convenient and rugged procedure for characterizing the selectivity of alkyl-silica columns.

Previous work has shown that the selectivity of reversed-phase columns for HPLC can be described by means of five column parameters: H (hydrophobicity), S* (steric resistance), A (hydrogen-bond acidity), B (hydrogen-bond basicity) and C (cation-exchange capacity). Values of H, S*, etc. can be determined by carrying out retention measurements for 18 test solutes under standardized conditions. The reproducibility of the latter procedure has been evaluated by comparison testing in four different laboratories and found acceptable. An alternative 10-solute test procedure which is more reproducible and convenient (but somewhat less accurate), requires only 2-3 h per column.

Choosing an equivalent replacement column for a reversed-phase liquid chromatographic assay procedure.

The column selectivity parameters (H, S*, A, B and C) described in the preceding paper [L.R. Snyder, A. Maule, A. Heebsch, R. Cuellar, S. Paulson, J. Carrano, L. Wrisley C.C. Chan, N. Pearson, J.W. Dolan, J.J. Gilroy, J. Chromatogr. A 1057 (2004) 49-57] can be used to compare columns in terms of selectivity. A detailed procedure for such column comparisons is presented here, and evaluated by its use in finding suitable replacement columns for 12 different routine separations performed in five different pharmaceutical analysis laboratories.

Column selectivity in reversed-phase liquid chromatography. VI. Columns with embedded or end-capping polar groups.

A previous model of column selectivity for reversed-phase liquid chromatography (RP-LC) has been applied to an additional 21 columns with embedded or end-capping polar groups (EPGs). Embedded-polar-group columns exhibit a significantly different selectivity vs. non-EPG, type-B columns, generally showing preferential retention of hydrogen-bond donors, as well as decreased retention for hydrogen-bond acceptors or ionized bases. EPG-columns are also generally less hydrophobic (more polar) than are non-EPG-columns. Interestingly, columns with polar end-capping tend to more closely resemble non-EPG columns, suggesting that the polar group has less effect on column selectivity when used to end-cap the column versus the case of an embedded polar group. Column selectivity data reported here for EPG-columns can be combined with previously reported values for non-EPG columns to provide a database of 154 different columns. This enables a comparison of any two of these columns in terms of selectivity. However, comparisons that involve EPG columns are more approximate.

The hydrophobic-subtraction model of reversed-phase column selectivity.

A recently developed treatment of reversed-phase column selectivity (the hydrophobic-subtraction model) is reviewed and extended, including its characterization of the selectivity of different column types (e.g., C1-C30, cyano, phenyl, etc.). The application of this model to retention data for various solutes and columns has provided new insights into the nature of different solute-column interactions and their relative importance in affecting sample retention and separation. Reversed-phase columns can be characterized by five selectivity parameters (H, S*, A, B and C), values of which are summarized here for more than 300 different columns. The selection of columns of either equivalent or different selectivity is readily achievable on the basis of their values of H, S*, etc. The development of the hydrophobic-subtraction model, its use in characterizing the selectivity of different reversed-phase liquid chromatography (RP-LC) columns, and its application to various practical problems as described here began in 1998. The original inspiration for this project owes much to Jack Kirkland, who also contributed actively to the initial studies that laid the foundation of this model; he has since provided other important support to this project. Jack and one of the authors (LRS) have enjoyed a strong professional relationship and personal friendship for the past 35 years, and it is the privilege of the authors to dedicate this paper and the work that it represents to Jack. His contributions to HPLC column technology have extended from the mid-1960s into the present century, and it is impossible to conceive of present day HPLC practice without Jack's contributions over the years. In this and other ways, his position as a pioneer and key implementer of HPLC is widely recognized. We wish Jack well in the years to come.

Slow equilibration of reversed-phase columns for the separation of ionized solutes.

Reversed-phase columns that have been stored in buffer-free solvents can exhibit pronounced retention-time drift when buffered, low-pH mobile phases are used with ionized solutes. Whereas non-ionized compounds exhibit constant retention times within 20 min of the beginning of mobile phase flow, the retention of ionized compounds can continue to change (by 20% or more) for several hours. If mobile phase pH is changed from low to high and back again, an even longer time may be required before the column reaches equilibration at low pH. The speed of column equilibration for ionized solutes can vary significantly among different reversed-phase columns and is not affected by flow rate.

Column selectivity in reversed-phase liquid chromatography I. A general quantitative relationship.

Retention factors k have been measured for 67 neutral, acidic and basic solutes of highly diverse molecular structure (size, shape, polarity, hydrogen bonding, pKa, etc.) on 10 different C18 columns (other conditions constant). These data have been combined with k values from a previous study (86 solutes, five different C8 and C18 columns) to develop a six-term equation for the correlation of retention as a function of solute and column. Values of k can be correlated with an accuracy of +/- 1-2% (1 standard deviation). This suggests that all significant contributions to column selectivity have been identified (and can be measured) for individual alkyl-silica columns which do not have an embedded polar group. That is, columns of the latter kind can be quantitatively characterized in terms of selectivity for use in the separation of any sample.

Column selectivity in reversed-phase liquid chromatography II. Effect of a change in conditions.

The isocratic retention of 67 widely-different solutes in reversed-phase liquid chromatography (RP-LC) has been investigated as a function of temperature and mobile phase composition (% B) for three different C18 columns. Similar studies were also carried out in a gradient mode, where temperature, gradient time and solvent type were varied. These results show that changes in retention with these conditions are similar for each of these three columns. This suggests that relative column selectivity as defined by experiments for one set of experimental conditions will be approximately applicable for other conditions, with the exception of changes in mobile phase pH-which can affect values of the column parameter C (a measure of silanol ionization). Column selectivity as a function of pH was explored for several columns.

Column selectivity in reversed-phase liquid chromatography III. The physico-chemical basis of selectivity.

Reversed-phase liquid chromatography (RP-LC) retention data for 23 additional solutes have been acquired to further test and evaluate a general relationship from part I: log alpha = log (k/kref) = eta'H(i) + sigma'S(ii) beta'S(iii) + alpha'B(iv) +kappa'C(v) The physico-chemical origin of terms i-v above is examined here by comparing values of (a) the solute parameters of Eq. (1) (eta', sigma', etc.) vs. solute molecular structure, and (b) the column parameters (H, S, etc.) vs. column properties (ligand length and concentration, pore diameter, end-capping). We conclude that terms i-v correspond, respectively, to hydrophobic (i), steric (ii), hydrogen bonding (iii, iv) and ionic (v) interactions between solute and stationary phase. While steric interaction (term ii) is superficially similar to what previously has been defined as "shape selectivity", the role of the solute and column in determining steric selectivity (term ii) appears more complex than previously proposed for "shape selectivity". Similarly, what has previously been called hydrogen bonding between donor solutes and an acceptor group in the stationary phase (term iv) is very likely an oversimplification.