Elsevier

Journal of Chromatography A

Volume 1408, 21 August 2015, Pages 227-235
Journal of Chromatography A

Development of an achiral supercritical fluid chromatography method with ultraviolet absorbance and mass spectrometric detection for impurity profiling of drug candidates. Part II. Selection of an orthogonal set of stationary phases

https://doi.org/10.1016/j.chroma.2015.07.035Get rights and content

Highlights

  • 23 columns with varied stationary phase chemistry screened.

  • Top columns are rather polar and mixed-polarity stationary phases.

  • Best 11 columns compared in terms of orthogonality.

  • Two versatile columns selected for orthogonal screening.

Abstract

Impurity profiling of organic products that are synthesized as possible drug candidates requires complementary analytical methods to ensure that all impurities are identified. Supercritical fluid chromatography (SFC) is a very useful tool to achieve this objective, as an adequate selection of stationary phases can provide orthogonal separations so as to maximize the chances to see all impurities.

In this series of papers, we have developed a method for achiral SFC-MS profiling of drug candidates, based on a selection of 160 analytes issued from Servier Research Laboratories.

In the first part of this study, focusing on mobile phase selection, a gradient elution with carbon dioxide and methanol comprising 2% water and 20 mM ammonium acetate proved to be the best in terms of chromatographic performance, while also providing good MS response [1].

The objective of this second part was the selection of an orthogonal set of ultra-high performance stationary phases, that was carried out in two steps. Firstly, a reduced set of analytes (20) was used to screen 23 columns. The columns selected were all 1.7–2.5 μm fully porous or 2.6–2.7 μm superficially porous particles, with a variety of stationary phase chemistries. Derringer desirability functions were used to rank the columns according to retention window, column efficiency evaluated with peak width of selected analytes, and the proportion of analytes successfully eluted with good peak shapes. The columns providing the worst performances were thus eliminated and a shorter selection of columns (11) was obtained. Secondly, based on 160 tested analytes, the 11 columns were ranked again. The retention data obtained on these columns were then compared to define a reduced set of the best columns providing the greatest orthogonality, to maximize the chances to see all impurities within a limited number of runs. Two high-performance columns were thus selected: ACQUITY UPC2 HSS C18 SB and Nucleoshell HILIC.

Introduction

Impurity profiling of organic products that are synthesized as possible drug candidates is a significant concern. For this purpose, it is necessary to have complementary high-performance analytical methods to ensure that all impurities are identified. SFC (usually expanded as Supercritical Fluid Chromatography, although the fluid employed is now rarely in the supercritical state) is one such method. SFC makes use of liquid mobile phases comprising a significant portion of carbon dioxide mixed to a co-solvent [2]. CO2 has major advantages over more conventional chromatographic solvents, as it has a low viscosity allowing for high diffusivities of the analytes (hence high efficiencies) and limited pressure drop over packed columns. As a result, high flow rates can be used without strongly affecting efficiency, and columns packed with sub-2 μm particles can be employed with relatively low-pressure pumping systems (400 bar) [3]. Consequently, the recent progresses in stationary phase technology (small particles [4], [5], but also superficially porous particles [6]) has also benefited to SFC.

An interesting feature of SFC is that, in addition to possibly providing an orthogonal method to a reversed-phase HPLC one [5], [6], [7], [8], [9], [10], it can also be orthogonal to itself, when stationary phases are adequately selected [11]. Indeed, all columns that are marketed for HPLC, whether for reversed-phase (RP), normal-phase (NP), hydrophilic interaction (HILIC) or ion-exchange modes, can also be used with mobile phases comprising carbon dioxide [12], [13], [14], [15], [16]. Chemical diversity of the available stationary phases is currently significantly improving, with rising interest of the column manufacturers and research groups to produce original phases dedicated to SFC use [17], [18], [19]. Moreover, while different operating modes in HPLC require different mobile phase composition (for instance, hydro-organic in RP, alkane-alcohol in NP), the same CO2-co-solvent mobile phase may be used with all of them. As a result, two columns with different surface chemistry can be employed with the same operating conditions and provide orthogonal selectivity [4], [20].

The present work aims at developing a rapid screening method for impurity profiling of drug candidates with SFC-ESI-MS. The first part presented in the previous paper focused on the selection of a versatile mobile phase composition to ensure elution of the largest proportion of drug-like compounds with good peak shape and the best possible UV and ESI-MS responses. Several additives introduced in the CO2-methanol mobile phase were thus tested with a wide range of stationary phases to assess their capabilities for successful chromatography and MS detection. Because the method aims at direct applicability in a pharmaceutical company, a large selection (160) of drug candidates provided by Servier Research Laboratories was evaluated. We finally settled our choice on a gradient elution of methanol comprising 2% water and 20 mM ammonium acetate [1].

The second part, described in the present paper, will focus on stationary phase selection to achieve orthogonal methods.

Section snippets

Chemicals, solvents and reagents

160 drug candidates were obtained from Servier Research Laboratories (Suresnes, France) whose structures are confidential. More details about the compounds selected can be found in the first part of this study. Ammonium acetate was obtained from Sigma–Aldrich (Saint-Quentin Fallavier, France) and ultra-pure water was provided by an Elga UHQ system from Veolia (Wissous, France). Solvents used were HPLC-grade methanol (MeOH) and ethanol provided by VWR (Fontenay-sous-Bois, France). Formic acid

Description of the analyte set

The analytes selected for this study were extracted from a library of drug candidates from Servier Research Laboratories. In designing this test set, the purpose was to have a set that would be as representative as possible of the diversity of chemical structures usually encountered in this pharmaceutical company [1]. The resulting group of analytes comprises acids, bases and neutrals, with log P values in the range [−2; 8.0]. For the first screening of 23 columns, a short selection of 20

Conclusions

In the aim of providing an orthogonal set of stationary phases, a selection of 11 efficient columns (out of an initial set of 23) with a variety of stationary phase chemistries were first selected based on the analysis of 20 analytes. Increasing the number of tested analytes to 160 with a thorough evaluation of chromatographic results permitted a ranking of the 11 best columns towards this particular selection of analytes. These 11 columns essentially comprised polar stationary phases and

Acknowledgments

We warmly thank Waters Corporation for the ACQUITY UPC2 system and ACQUITY QDa detector let at our disposal (Mark Baynham, Taraneh Kargar) and for the opportunity to test new stationary phases (Darryl Brousmiche, Jacob Fairchild, Kevin Wyndham, Steve Collier and team). We also thank Thierry Domenger (Thermo), Régis Guyon (Machery-Nagel), Magali Dupin and Marc Jacob (Phenomenex) for the kind gift of columns.

References (23)

Cited by (31)

  • Guilty by dissociation: Part B: evaluation of Supercritical Fluid Chromatography (SFC-UV) for the analysis of regioisomeric diphenidine-derived Novel Psychoactive Substances (NPS)

    2022, Journal of Pharmaceutical and Biomedical Analysis
    Citation Excerpt :

    Two neutral (i.e., cyano and diol ligands) and five basic phases (i.e., weakly basic 2- and 4-ethylpyridine, pyridyl amide and the strongly basic imidazole and diethylamino ligands) were chosen in addition to two acidic silica phases (see Table 1) [21]. Generic gradient SFC conditions which have previously been reported to generate good peak shapes were employed, these utilised CO2 and 10 mM ammonium acetate (pH unadjusted) in MeOH-water (95:5 v/v) respectively [28,30]. A small amount of water was included as this has been shown to yield more reproducible retention times and enhanced peak shapes [30–33].

  • Selection of SFC stationary and mobile phases

    2022, Separation Science and Technology (New York)
  • A technical overview of supercritical fluid chromatography-mass spectrometry (SFC-MS) and its recent applications in pharmaceutical research and development

    2021, Drug Discovery Today: Technologies
    Citation Excerpt :

    The analysis of pharmaceutical compounds were compared by LC-MS and SFC-MS [37]. In recent years, SFC-MS applications have been focused on chiral analysis [38–40], mass directed purification systems [41], two dimensional separation [42,43], and impurity profiling [44–46]. These applications highlighted the advantages of SFC in its superior ability for chiral separations, orthogonality to LC separation, and high efficiency as a purification technique.

View all citing articles on Scopus
View full text