Eva Tyteca

Retention modeling and method development in hydrophilic interaction chromatography.

In the present study, the possibility of retention modeling in the HILIC mode was investigated, testing several different literature relationships over a wide range of different analytical conditions (column chemistries and mobile phase pH) and using analytes possessing diverse physico-chemical properties. Furthermore, it was investigated how the retention prediction depends on the number of isocratic or gradient trial or initial scouting runs. The most promising set of scouting runs seems to be a combination of three isocratic runs (95, 90 and 70%ACN) and one gradient run (95 to 65%ACN in 10min), as the average prediction errors were lower than using six equally spaced isocratic runs and because it is common in Method development (MD) to perform at least one scouting gradient run in the screening step to find out the best column, temperature and pH conditions. Overall, the retention predictions were much less accurate in HILIC than what is usually experienced in RPLC. This has severe implications for MD, as it restricts the use of commercial software packages that require the simulation of the retention of every peak in the chromatogram. To overcome this problem, the recently proposed predictive elution window shifting and stretching (PEWS(2)) approach can be used. In this computer-assisted MD strategy, only an (approximate) prediction of the retention of the first and the last peak in the chromatogram is required to conduct a well-targeted trial-and-error search, with suggested search conditions uniformly covering the entire possible search and elution space. This strategy was used to optimize the separation of three representative pharmaceutical mixtures possessing diverse physico-chemical properties (pteridins, saccharides and cocktail of drugs/metabolites). All problems could be successfully handled in less than 2.5h of instrument time (including equilibration).

Towards a chromatographic similarity index to establish localized quantitative structure-retention models for retention prediction: Use of retention factor ratio

Quantitative Structure-Retention Relationships (QSRR) have the potential to speed up the screening phase of chromatographic method development as the initial exploratory experiments are replaced by prediction of analyte retention based solely on the structure of the molecule. The present study offers further proof-of-concept of localized QSRR modelling, in which the retention of any given compound is predicted using only the most chromatographically similar compounds in the available dataset. To this end, each compound in the dataset was sequentially removed from the database and individually utilized as a test analyte. In this study, we propose the retention factor k as the most relevant chromatographic similarity measure and compare it with the Tanimoto index, the most popular similarity measure based on chemical structure. Prediction error was reduced by up to 8 fold when QSRR was based only on chromatographically similar compounds rather than using the entire dataset. The study therefore shows that the design of a practically useful structural similarity index should select the same compounds in the dataset as does the k-similarity filter in order to establish accurate predictive localized QSRR models. While low average prediction errors (Mean Absolute Error (MAE)<0.5min) and slopes of the regression lines through the origin close to 1.00 were obtained using k-similarity searching, the use of the structural Tanimoto similarity index, considered as the gold standard in Quantitative Structure-Activity Relationships (QSAR) studies, generally resulted in much higher prediction errors (MAE>1min) and significant deviations from the reference slope of 1.0. The Tanomoto similarity index therefore appears to have limited general utility in QSRR studies. Future studies therefore aim at designing a more appropriate chromatographic similarity index that can then be applied for unknown compounds (that is, compounds which have not been tested previously on the chromatographic system used, but for which the chemical structures are known).

Use of individual retention modeling for gradient optimization in hydrophilic interaction chromatography: Separation of nucleobases and nucleosides

In this study, the separation of twelve nucleobases and nucleosides was optimized via chromatogram simulation (i.e., prediction of individual retention times and estimation of the peak widths) with the use of an empirical (reversed-phase) non-linear model proposed by Neue and Kuss. Retention time prediction errors of less than 2% were observed for all compounds on different stationary phases. As a single HILIC column could not resolve all peaks, the modeling was extended to coupled-column systems (with different stationary phase chemistries) to increase the separation efficiency and selectivity. The analytical expressions for the gradient retention factor on a coupled column system were derived and accurate retention time predictions were obtained (<2% prediction errors in general). The optimized gradient (predicted by the optimization software) included coupling of an amide and an pentahydroxy functionalized silica stationary phases with a gradient profile from 95 to 85%ACN in 6 min and resulted in almost baseline separation of the twelve nucleobases and nucleosides in less than 7 min. The final separation was obtained in less than 4h of instrument time (including equilibration times) and was fully obtained via computer-based optimization. As such, this study provides an example of a case where individual retention modeling can be used as a way to optimize the gradient conditions in the HILIC mode using a non-linear model such as the Neue and Kuss model.

Predictive elution window stretching and shifting as a generic search strategy for automated method development for liquid chromatography.

We report on the possibilities of a new method development (MD) algorithm that searches the chromatographic parameter space by systematically shifting and stretching the elution window over different parts of the time-axis. In this way, the search automatically focuses on the most promising areas of the solution space. Since only the retention properties of the first and last eluting compounds of the sample need to be (approximately) known, the algorithm can be directly applied to samples with unknown composition, and the proposed solutions are not sensitive to any modeling errors. The search efficiency of the algorithm has been evaluated on an extensive set of random-generated in silico samples covering a broad range of different retention properties. Compared to a pure grid-based search, the algorithm could reduce the number of missed components by 50% and more. The algorithm has also been applied to solve three different real-world separation problems from the pharmaceutical industry. All problems could be successfully solved in a very short time (order of 12 h of instrument time).

Gradient-elution parameters in capillary liquid chromatography for high-speed separations of peptides and intact proteins.

This contribution relates to the assessment of gradient-elution parameters in capillary liquid chromatography affecting the peak widths in the reversed-phase separation of peptides and intact proteins. Gradient separations were performed using both a poly(sytrene-co-divinylbenzene) monolithic column and a microparticulate fused-core column (silica C18, 2.7μm). The applicability of the conventional linear (LSS) and non-linear solvent-strength model (Neue-Kuss) were investigated to describe the retention behaviour of the compounds as a function of the mobile-phase composition. This was performed by using a wide range of gradient conditions, including different gradient slopes (β, ranging from 0.05 to 0.65min(-1)) and mobile-phase compositions (Δϕ, i.e. gradient span). Although the LSS-model provided accurate retention time predictions (<1.3% deviation) of scouting runs with more conventional gradient slopes, the prediction of high-speed separations with a high degree of accuracy (<2%) could only be obtained with the non-linear model. The solvent-strength parameters resulting from the use of both models, as well as the retention factors at the moment of elution (ke), further served as input parameters to assess the influence of the gradient slope on the expected peak-compression effects in gradient mode, with a focus on high-speed separations. The importance of the correct model choice was emphasized in terms of compression; while the LSS-model lead to the conclusion of peak broadening rather than peak sharpening, the use of a more accurate non-linear model showed the existence of peak compression effect. The results presented in this manuscript show the occurrence of gradient-related focusing effects, which appear to be more prevalent for extremely fast separations.

Possibilities of retention modeling and computer assisted method development in supercritical fluid chromatography

The multi-modal retention mechanism in supercritical fluid chromatography (SFC) results in a non-linear dependency of log(k) on the fraction of organic solvent φ and log(φ). In the present study, the possibility of retention modeling for method development purposes in SFC was investigated, considering several non-linear isocratic relationships. Therefore, both isocratic and gradient runs were performed, involving different column chemistries and analytes possessing diverse physico-chemical properties. The isocratic retention data of these compounds could be described accurately using the non-linear retention models typically used in HILIC and reversed-phase LC. The interconversion between isocratic and gradient retention data was found to be less straightforward than in RPLC and HILIC because of pressure effects. The possibility of gradient predictions using gradient scouting runs to estimate the retention parameters was investigated as well, showing that predictions for other gradients with the same starting conditions were acceptable (always below 5%), whereas prediction errors for gradients with a different starting condition were found to be highly dependent on the compound. The second part of the study consisted of the gradient optimization of two pharmaceutical mixtures (one involving atorvastatin and four related impurities, and one involving a 16 components mixture including eight drugs and their main phase I metabolites). This could be done via individual retention modeling based on gradient scouting runs. The best linear gradient was found via a grid search and the best multi-segment gradient via the previously published one-segment-per-component search. The latter improved the resolution between the critical pairs for both mixtures, while still giving accurate prediction errors (using the same starting concentrations as the gradient scouting runs used to build the model). The optimized separations were found in less than 3 h and 8 h of analysis time (including equilibration times), respectively.

Applicability of linear and nonlinear retention-time models for reversed-phase liquid chromatography separations of small molecules, peptides, and intact proteins.

The applicability and predictive properties of the linear solvent strength model and two nonlinear retention-time models, i.e., the quadratic model and the Neue model, were assessed for the separation of small molecules (phenol derivatives), peptides, and intact proteins. Retention-time measurements were conducted in isocratic mode and gradient mode applying different gradient times and elution-strength combinations. The quadratic model provided the most accurate retention-factor predictions for small molecules (average absolute prediction error of 1.5%) and peptides separations (with a prediction error of 2.3%). An advantage of the Neue model is that it can provide accurate predictions based on only three gradient scouting runs, making tedious isocratic retention-time measurements obsolete. For peptides, the use of gradient scouting runs in combination with the Neue model resulted in better prediction errors (<2.2%) compared to the use of isocratic runs. The applicability of the quadratic model is limited due to a complex combination of error and exponential functions. For protein separations, only a small elution window could be applied, which is due to the strong effect of the content of organic modifier on retention. Hence, the linear retention-time behavior of intact proteins is well described by the linear solvent strength model. Prediction errors using gradient scouting runs were significantly lower (2.2%) than when using isocratic scouting runs (3.2%).

Computer-assisted multi-segment gradient optimization in ion chromatography

This study reports simulation and optimization of ion chromatography separations using multi-segment gradient elution. First, an analytical expression for the gradient retention factor under these complex elution profiles was derived. This allows a rapid retention time prediction calculations under different gradient conditions, during computer-assisted method development. Next, these analytical expressions were implemented in an in-house written Matlab(®) routine that searches for the optimal (multi-segment) gradient conditions, either via a four-segment grid search or via the recently proposed one-segment-per-component search, in which the slope is adjusted after the elution of each individual component. Evaluation of the retention time simulation and optimization approaches was performed on a mixture of 18 inorganic anions and different subsets with varying number of compounds. The two considered multi-segment gradient optimization searches resulted in similar proposed gradient profiles, and corresponding chromatograms. Moreover, the resultant chromatograms were clearly superior to the chromatograms obtained from the best simple linear gradient profiles, found via a fine grid search. The proposed approach is useful for automated method development in ion chromatography in which complex elution profiles are often used to increase the separation power.

Supercritical fluid chromatography: a promising alternative to current bioanalytical techniques

During the last years, chemistry was involved in the worldwide effort toward environmental problems leading to the birth of green chemistry. In this context, green analytical tools were developed as modern Supercritical Fluid Chromatography in the field of separative techniques. This chromatographic technique knew resurgence a few years ago, thanks to its high efficiency, fastness and robustness of new generation equipment. These advantages and its easy hyphenation to MS fulfill the requirements of bioanalysis regarding separation capacity and high throughput. In the present paper, the technical aspects focused on bioanalysis specifications will be detailed followed by a critical review of bioanalytical supercritical fluid chromatography methods published in the literature.

Effect of gradient steepness on the kinetic performance limits and peak compression for reversed-phase gradient separations of small molecules

The effect of gradient steepness on the kinetic performance limits and peak compression effects has been assessed in gradient mode for the separation of phenol derivatives using columns packed with 2.6μm core-shell particles. The effect of mobile-phase velocity on peak capacity was measured on a column with fixed length while maintaining the retention factor at the moment of elution and the peak-compression factor constant. Next, the performance limits were determined at the maximum system pressure of 100MPa while varying the gradient steepness. For the separation of small molecules applying a linear gradient with a broad span, the best performance limits in terms of peak capacity and analysis time were obtained applying a gradient-time-to-column-dead-time (tG/t0) ratio of 12. The magnitude of the peak-compression factor was assessed by comparing the isocratic performance with that in gradient mode applying different gradient times. Therefore, the retention factors for different analytes were determined in gradient mode and the mobile-phase composition in isocratic mode was tuned such that the difference in retention factor was smaller than 2%. Peak-compression factors were quantitatively determined between 0.95 and 0.65 depending on gradient steepness and the gradient retention factor.