Reversed-Phases for LC Deliberately Doped with Positive Charge: Tips and Tricks for Effective Use

Molecules encountered in the early-phase of drug development present a diverse set of challenges to HPLC analysis, as well as purification workflows. Finding conditions that yield adequate resolution of a target product peak can be difficult when limited to the conventional C18 stationary phases. Successful preparative chromatography relies on adequate peak capacity, resolution in the vicinity of the target peak, and increased loading capacity to achieve high chromatographic method productivity. In this article, we discuss both the benefits (improved peak shape/loading) and challenges (excessive interaction) associated with charge-doped reversed-phase (RP) columns for both analytical and preparative separations.

The introduction of commercially available reversed-phase (RP) columns that have stationary phases deliberately doped with positive charge (for example, “charged surface hybrid” columns; these will be referred to as “charge-doped RP phases” hereafter) has resulted in improved chromatographic performance for separating complex mixtures (1,2). Specifically, the elution profiles of basic molecules can be enhanced when using mobile phases with acidic modifiers at low ionic strength. The presence of the controlled concentrations of positively charged groups on the surface of silica particles has been shown to not only improve peak shape but also increase loading capacity. As a result, using such particles can be attractive for purification purposes, where high mass loadability is particularly important. Whereas highly specialized ion-pairing agents may have been used in the past with conventional RP materials, in some cases these are no longer necessary to resolve target peaks, and molecules having substituents with high pK_as can be purified in acidic conditions without concerns about excessive peak tailing. These observations have led to the adoption of the charge-doped RP phases in our laboratories for the purification of small molecule libraries. Since then, large libraries of target molecules have been successfully isolated at the milligram and microgram scale as part of a high-throughput workflow. As part of the continuous assessment of our workflows, we have reviewed the use of charge-doped RP phases for our µg-HTPP to ensure that our platform workflows continue to meet project demands. The review revealed the beneficial attributes of charge-doped RP phases for a majority of samples. Nevertheless, there were important instances where the column chemistry had a negative or unanticipated impact on separation performance. Understanding the causes of these cases can promote more effective uses of the charge-doped RP phases in the future.

Advantages of Charge-Doped RP Phases for Analytical Scale Separations

The advantages of charge-doped RP phases for separating basic analytes can be visualized by comparing the chromatograms obtained for a mixture of standard probe compounds using either a conventional C18 type phase or a charge-doped RP phase (see Figure 1). The peak profiles for the basic analytes metoclopramide and nortriptyline show significant decreases in retention time, as well as improvements in peak symmetry, on the charge-doped RP phase when compared to a conventional C18 phase. Diethylacetamide is neutral under acidic conditions. On the other hand, nortriptyline and metoclopramide are fully protonated under acidic conditions, and thus, a significant improvement in peak shape is observed on the charge-doped RP column because of minimal secondary interactions (3). Figure 1 shows an example of how the selectivity of the charge-doped RP phase is complementary to the selectivity of the C18 phase, in addition to the benefits related to peak shape (4).

Figure 1: Overlay of chromatograms for N,N-diethylacetamide, metoclapramide, and nortriptyline on XBridge BEH C18 (3.0 mm i.d. x 50 mm, 3.5 μm) (red) and Cortecs CSH C18 (3.0 mm i.d. x 50 mm, 2.7 μm) (black) columns under acidic conditions. Mobile phase A: 0.1% formic acid in water. Mobile phase B: 0.1% formic acid in acetonitrile. Flow rate: 1.0 mL/min. Gradient is 5–95% B in 4 min, followed by re-equilibration for 2 min. Injection volume: 1 μL. Detection wavelength: 215 nm. Sample concentration: 1.1 mM (N,N-diethylacetamide), 1.0 mM (metoclopramide), and 1.0 mM (nortriptyline).

Using Charge-Doped RP Columns for Purification of Acidic Compounds

There are instances when a charge-doped RP phase will have undesired effects on the analyte because of undesired electrostatic interactions between stationary phase and analyte. Analytes that are acidic or exhibit a negative charge under chromatographic conditions will adsorb more strongly because of their interactions with the positively charged surface. Figure 2a shows the increased retention of a several acidic components of an unpurified reaction mixture when using a charge-doped RP column, in comparison to a conventional C18 column. The increased tailing observed for the synthesis targets effectively limit the practical loading capacity when moving from analytical to preparative scale separation. It is important to note that when developing a method for purification, the goal is not to resolve every peak in the mixture, but rather to maximize resolution in the vicinity of the peak of interest. The resolution between the main peak and the closely eluting tail peak are quite similar for the charge-doped RP and conventional C18 columns. However, the asymmetry is approximately doubled with the charge-doped RP column, which dramatically decreases the practical loading capacity.

Figure 2: (a) Overlay of chromatograms obtained for separations of a crude synthesis sample on XBridge BEH C18 (3.0 mm i.d. x 50 mm, 3.5 μm) (red) and Cortecs CSH C18 (3.0 mm i.d. x 50 mm, 2.7 μm) (black) columns under acidic conditions. Mobile phase A: 0.1% formic acid in water. Mobile phase B: 0.1% formic acid in acetonitrile. Flow rate: 2.0 mL/min. Gradient is 5-95 %B in 1.5 min, followed by re-equilibration for 2 min. Sample concentration: 1.0 mg/mL.; (b) Chromatograms for separations of crude synthesis produce mixtures on CSH PFP (black), CSH C18 (blue), and non-CSH PFP (red) columns under acidic conditions. All columns are 4.6 mm i.d. x 100 mm, 2.5 μm. Mobile phase A: 0.1% formic acid in water. Mobile phase B is acetonitrile. Flow rate: 1.5 mL/min. Gradient is 5–95% B in 8 min, followed by re-equilibration for 2 min.

For the purification of individual samples, the analyst can easily determine if there are acidic moieties present in the target analytes that would discourage the use of charge-doped RP columns. However, when performing the purification of samples in a high-volume/high-throughput environment, manual review of the structures of hundreds of molecules could hinder the potential throughput of the established platform. Analysis of such samples without prior review of the target structures can lead to unexpected results when negatively charged species are present. Figure 2b shows the chromatograms obtained from analysis of negatively charged molecules using either charge-doped or conventional RP phases. In this case, a phosphate moiety was present on the compounds studied and is negatively charged under the mobile phase conditions. Electrostatic interactions between the positive charges on the surface of the charge-doped materials and the negatively charged analytes resulted in excessive peak distortion and increased retention. In the case of the charge-doped pentafluorophenyl (PFP) column, we observe the extreme case where the target compound does not elute at all, even though the method involves running the mobile phase to 95% B solvent. The excessive retention and peak distortion observed in cases like this often leads to losses in resolution, sensitivity, and efficiency. On the other hand, conventional C18 phases (such as non-CSH PFP in Figure 2b) yield better peak characteristics (less tailing and retention), which suggests that the differences in peak shape are because of the presence of a significant population of positively-charged sites on the stationary phase surface.

Effect of Sample Matrix on Basic Compounds when Using a Charge-Doped RP Phase

A large fraction of molecules encountered in the contemporary pharmaceutical landscape contain basic functional groups. Although the synthetic steps to arrive at these molecules are diverse and complex, some reactions are more frequently used than others. One group of reactions involves amide coupling, which are among the most commonly used reactions in pharmaceutical chemistry (5). One route involves deprotonating a carboxylic acid using a base and forming the transitionary ester through the onium salt. The ester is then replaced by the amine to form the final amide bond. Many commercially available coupling agents show unique selectivity and can offer advantages to the coupling in terms of stereochemistry and yield.

The coupling reaction is only a part of the synthesis. The product must be isolated from any remaining starting materials or side products prior to submitting for biological assays. A reaction with no residual starting materials or side products is considered a highly successful reaction from a synthetic perspective. Nevertheless, chromatographic methods may be incapable of resolving the product from the coupling agent. The work-up to remove these unwanted species can be difficult and ineffective depending on the coupling agent used. In the fast-paced, high-throughput research environment, it is often considered more efficient to push reactions to chromatography platforms to isolate target compounds with good purity. However, the purification process can be difficult if the chromatographic behavior of each of the species is not understood prior to injection of the samples.

An example of such complexity is the retention behavior of the hexafluorophosphate (PF6-) ion (as ammonium hexafluorophosphate), present in uranium and phosphonium based coupling agents. Since PF_6- lacks a chromophore, negative ion electrospray mass spectrometry was used to compare the retention behavior of these ions on charge-doped and conventional RP phases. The PF_6- is not retained on a conventional C18 column when using acidic mobile phases additives and acetonitrileto-water gradients, and thus can be easily removed during purification process, as shown in Figure 3a. However, when utilizing the same conditions with a charge-doped RP phase, the retention of the PF_6- ion is much higher and its peak is severely tailed, which complicates the purification process, especially when the PF_6- is coeluted with a target molecule. Figure 3a clearly shows that PF_6- is retained on charge-doped RP column, even in high mobile phase concentrations of organic solvent (~50% acetonitrile in water). This might lead to co-elution with many target analytes. Additionally, the PF6- disrupts normal elution behavior. Previous studies have demonstrated that when PF6- is used as a chaotropic agent in the mobile phase it can have a dramatic effect on retention of basic molecules on charge-doped RP columns (6). In our case, even though the PF_6- is only present in the sample matrix, high concentrations PF_6- can produce similar effects, increasing retention and peak width (Figure 3b). This matrixeffect is especially pronounced in the analysis of poorly retained basic analytes, such as procainamide, while it is absent in the case of neutral analytes (for example, carbamazepine; Figure 3b).

Figure 3: (a) Extracted ion chromatogram (EIC; m/z 145) elution profile of PF_6- on XBridge C18 (3.0 mm i.d. x 50 mm, 3.5 μm) (red trace) and Cortecs C18 (3.0 mm i.d. x 50 mm, 2.7 μm) (black trace) columns under acidic (0.1% formic acid) conditions with a 5–95% acetonitrile in water gradient in 1.5 min at a fl ow rate of 2.0 mL/min; (b) Overlay of 1 µL injections of procainamide and carbamazapine standards with increasing concentration of PF_6- in the samples; control (blue), 2:1 (black), 5:1 (green), 10:1 (red) using Cortecs C18 (3.0 mm i.d. x 50 mm, 2.7 μm) under acidic (0.1% formic acid) conditions with a 5-95% acetonitrile in water gradient in 4 min at a fl ow rate of 1.0 mL/min.

The PF_6- ion can be used to increase the retention of basic compounds through ion-pairing effects. In Figure 4a, we demonstrate the effect of spiking samples with PF6- on the retention of basic analytes. At low injection volumes (1 µL in Figure 4a), there is a small increase in retention time, especially for earlier eluting compounds. When the injection volume is increased to 10 µL (Figure 4b) a breakthrough of the compounds is observed at the column dead time if there is no PF_6- added to the control. When the samples are spiked with 5 or 10 equivalents of PF_6- (that is, a 5:1 or 10:1 mole ratio of PF_6- to analyte), the breakthrough is completely eliminated, as illustrated in Figure 4b. The breakthrough of metoclopramide into the dead volume in the absence of PF_6- was confirmed using the extracted ion chromatograms obtained from mass spectrometric detection (see Figure 4c).

Figure 4: (a) Overlay of 1 L injections of metoclopramide, N,N-diethylacetamide and nortriptyline standards with increasing equivalents of PF_6- in the sample; control (blue), 5:1 (green), 10:1 (red); (b) Overlay of 10 L injections of metoclopramide, N,N-diethylacetamide and nortriptyline standards with increasing equivalents of PF_6- in the sample; control (blue), 5:1 (green), 10:1 (red) using Cortecs C18.; (c) EIC overlay of 10 L injections of metoclopramide confirming breakthrough when there is no PF6- in the sample; (d) Same as (b), but using a XBridge BEH C18 column. Chromatographic conditions: Sample concentrations - 1.1 mM (N,N-diethylacetamide), 1.0 mM (metoclopramide), and 1.0 mM (nortriptyline); Columns - Cortecs C18 (3.0 mm i.d. x 50 mm, 2.7 μm) or XBridge BEH C18 (3.0 mm i.d. x 50 mm, 3.5 μm); Mobile phase B - 0.1% formic acid in acetonitrile; Flow rate - 1.0 mL/min; Gradient is 5-95% in 4 min, followed by re-equilibration for 2 mins. Injection volume - 1 μL; Detection wavelength - 215 nm.

Finally, Figure 4d shows the effect of adding PF_6- to the samples on the retention and peak shape when using a conventional C18 column. Interestingly, no breakthrough is observed in the absence of PF_6-; however, the peak shapes for the basic compounds are not as good when analyzed under the same conditions using the charge-doped RP column. Adding PF_6- to the sample improves the peak shape for metoclopramide, but not nortriptyline. Developing an understanding of these effects increases the likelihood of success when developing methods using these columns. When working at the preparative scale, understanding the effects of the sample matrix on the separation quality is particularly important.

Summary

Charge-doped RP column technologies are an important asset to any chromatographic screening and purification toolbox. Understanding the advantages and disadvantages of these stationary phases help us make better decisions about when to use the columns, and to be aware of situations where we may encounter problems. A major potential advantage of these phases is that they can provide improved peak shape and loading capacity for many basic analytes. However, the positive charge deliberately added to the stationary phase, which leads to the aforementioned benefit, can also interact very strongly with negatively charged species, which can lead to poor peak shape and prohibitively high retention. It is also important to keep in mind that the components of the sample matrix can have a significant effect of separations when using charge-doped RP phases. Hexafluorophosphate ions present in samples from amide coupling reactions can be strongly retained and potentially are coeluted with target compounds of interest, which is not desirable in the context of purification workflows. In situations where it is acceptable for PF_6- to be coeluted with the purified peak, a significant improvement in retention and loading capacity resulting from ion-pairing effects can be realized with charged-doped RP phases. Understanding these effects can facilitate development of effective methods using these columns and provide clues for troubleshooting problems when they arise.

Acknowledgments

The authors would like to thank Drs. Margaret Chu-Moyer and Troy Handlovic for their valuable feedback and insightful suggestions in reviewing this work.

References

(1) Halsey, H.; Hall, J.; Huang, C.; et al. Challenges with Retention and Recovery of Impurities Containing Acidic Moieties During Analytical UHPLC Method Development and Validation for Gefapixant Freebase. J. Pharm. Biomed. Anal. 2024, 237, 115772. DOI: 10.1016/j.jpba.2023.115772

(2) Fountain, K. J; Iraneta, P. C.; Morrison, D. Practical Applications of Charged Surface Hybrid (CSH) Technology, in: Application Notes, Waters Corporation website. https://www.waters.com/nextgen/us/en/library/application-notes/2010/practical-applications-charged-surface-hybrid-csh-technology.html?
srsltid=AfmBOoq39FMNgbo9dpG55iMaD1H_dOkIOjFPTlbd-
NxEtut1HS1foUNTJ (accessed 2025-04-09)

(3) Stoll, D. R. The Evolution of LC Troubleshooting: Strategies for Improving Peak Tailing. LCGC North Am. 2023, 41 (10), 404–408. DOI: 10.56530/lcgc.na.br4774h5

(4) Åsberg, D.; Langborg Weinmann, A.; Leek, T.; et al. The Importance of Ion-Pairing in Peptide Purification by Reversed-Phase Liquid Chromatography. J. Chromatog. A 2017, 1496, 80–91. DOI: 10.1016/j.chroma.2017.03.041

(5) Albericio, F. Developments in Peptide and Amide Synthesis. Curr. Opin. Chem. Biol. 2004, 8 (3), 211–221. DOI: 10.1016/j.cbpa.2004.03.002

(6) Pan, L.; LoBrutto, R.; Kazakevich, Y. V.; et al. Influence of Inorganic Mobile Phase Additives on the Retention, Efficiency and Peak Symmetry of Protonated Basic Compounds in Reversed-Phase Liquid Chromatography. J. Chromatog. A 2004, 1049 (1), 63–73. DOI: 10.1016/j.chroma.2004.07.019

About the Authors

Kevin Crossley is a scientist in the Synthetic Separations group at Amgen with 23 years of pharmaceutical industry experience in purification and chromatography. He began his career with largescale natural product isolations and joined Amgen in 2008, supporting programs from FIH to commercial with chiral and achiral purifications. Since 2019, he has contributed to advancing the Discovery pipeline and purification capabilities. Kevin is a recipient of multiple green chemistry awards, holds patents, and has authored several publications.

Wesley W. Barnhart is a principal scientist in the Synthetic Separations group within the Research and Development department at Amgen. His expertise is focused on analytical and preparative separations of achiral and chiral small molecules, oligonucleotides, and mass-directed purification of microgram levels of pharmaceutical compounds.

Dwight R. Stoll is the editor of “LC Troubleshooting”. Stoll is a professor and the co-chair of chemistry at Gustavus Adolphus College in St. Peter, Minnesota, USA. His primary research focus is on the development of 2D-LC for both targeted and untargeted analyses. He has authored or coauthored more than 75 peer-reviewed publications and four book chapters in separation science and more than 100 conference presentations. He is also a member of LCGC’s editorial advisory board. Direct correspondence to: amatheson@mjhlifesciences.com