Trends in Stationary Phases and Column Technologies

Abstract

Chromatography and column technology have undergone significant developments over the past two decades. Highly efficient small sub-2 µm particles and core–shell particle columns have become the new state-of-the-art in liquid chromatography (LC), with applications ranging from pharmaceutical and food analysis to environmental and forensic sciences, and bioanalysis including omics technologies (metabolomics, lipidomics, and proteomics). Peak widths of a few seconds can be achieved and run times with short columns are significantly reduced. Other chromatographic modes besides reversed-phase LC (RPLC), the standard workhorse in (U)HPLC, have found also broader adoption and application in many fields. In short, LC stationary phase technology has become more diverse. A driving force for new developments is new application fields which need new column technologies, like new drug modalities, challenges in environmental and bioanalysis, and new demands to characterize bioparticles like viruses, virus-like particles, microvesicles, and lipid-emulsion formulations. Short monolithic columns are often employed for (bio)particle separations. In many of these application fields, columns with deactivated column hardware surface have shown improved recoveries and reduced peak tailing, which lead to better detection sensitivity. These trends are reflected by the articles submitted to the special collection “Stationary phases and column technologies,” an entirely digital special issue, which was collecting invited articles over the period from April to September 2024. The contents and major ideas of the articles are briefly summarized below.

Multiple papers of this special collection refer to the development of stationary phases for hydrophilic interaction liquid chromatography (HILIC). HILIC stationary phases are typically based on silica with polar-modified surface chemistries. In one work, Zhao and coworkers synthesized a novel zwitterionic polymer grafted silica stationary phase by bonding poly(ethylene maleic anhydride) molecules on the surface of aminopropyl silica via multiple binding sites, followed by ammonolysis of remaining maleic anhydride groups through a nucleophilic substitution reaction with ethylenediamine [1]. The zwitterionic character resulted from carboxylic groups and primary amine groups. Zeta-potential measurements showed a charge reversal at about pH 5.5 from a positive surface charge below to negative surface charge above this isoelectric point of the material. Retention models were evaluated and thermodynamic analysis was employed to get a closer look into the retention mechanism. Nucleosides and nucleic bases, sugars, and Amadori products were successfully separated under HILIC conditions. Polymer particles are less commonly employed as support materials for HILIC columns as they have a hydrophobic character and a tendency to swell in organic solvents. In an article by Yang and coworkers, the high hydrophobicity and chemical inertness of poly(styrene-divinylbenzene) (PS-DVB) microspheres were overcome by grafting of an acrylamide-terminated lysine zwitterionic monomer onto the PS-DVB beads by free-radical polymerization to afford a polar stationary phase suitable for HILIC [2]. The modified beads showed zwitterionic surface chemistry originating from the α-amino group and carboxylic acid functionalities of lysine, which was polymerized and bonded via the ε-amino moiety through an acrylamide monomer. The authors characterized the stationary phase by the Tanaka HILIC test and compared the characteristics with multiple commercial HILIC phases. Column bleeding was also compared to commercial phases and was found to be significantly less than multiple brush-type HILIC phases. The new stationary phase showed a good performance for the HILIC separation of saccharides. Liang, Ke, and coworkers prepared low swelling and high mechanical strength organosilane hybrid polydivinylbenzene microspheres for hydrophilic interaction chromatography applications [3]. The driving force for this work was to overcome limitations of low-bleed HILIC stationary phases based on polymer matrices, which, due to the high content of the organic phase in the HILIC mobile phase, exhibit typically some swelling of the polymer matrix. A seed swelling polymerization approach with polystyrene latices as templates and divinylbenzene, as well as 3-methacryloxypropyltrimethoxysilane as monomers was employed to synthesize monodispersed beads. During this polymerization process of the beads, new crosslink bridges can be formed due to the methoxysilane moieties by condensation reaction, besides the radical polymerization of the DVB monomer. The increase in the degree of crosslinking was found conducive to the improvement of the resistance to swelling as well as to the mechanical strength of the microspheres. HILIC surface characteristics were introduced by a radical-addition reaction of cysteine via thiol-ene click reaction leading to zwitterionic surface. The stationary phase was compared to a formerly prepared Cys-clicked silica phase (ClickXIon) and found comparable in the majority of chromatographic properties. Another study addressed the screening of a number of commercial (silica, amide, and sulfobetaine) HILIC and RP columns to find proper conditions for the analysis of di- and tripeptides adducted with mustard agents originating from their reaction with albumin [4]. The goal was to develop an analytical LC-MS/MS assay that can be used to monitor retrospectively the exposure to mustard agents. Finally, a zwitterionic sulfobetaine type ZIC-HILIC column was selected for the HILIC and RP with C18 phase for RPLC analysis of an adducted tripeptide corresponding to adduct on Cys34 (of human albumin) to allow their complementary analysis in the low ng/mL concentration range as a marker of exposure to mustard agents.

A group of papers focused on new developments of chiral stationary phases (CSPs) for enantiomer separation. In one study, a chiral monolithic capillary column was synthesized by post-modification of an in situ prepared poly(glycidylmethacrylate-co-ethylene dimethacrylate) monolith with L-cysteine as chiral selector via ring-opening by nucleophilic substitution through nucleophilic attack of the epoxy group by the thiol of cysteine [5]. Dansyl amino acids could be successfully separated. Chiral macrocyclic compounds such as chiral crown ethers, cyclodextrins, and cyclofructans turned out to be successful chiral selectors in commercially available columns. Inspired by this class of macrocyclic structures, Yuan and coworkers presented a new CSP based on macrocyclic 1,1’-binaphthyl chiral polyimine [6]. Two molar equivalents each of axially chiral (S)-2,2’-dihydroxy-[1,1’-binaphthalene]-3,3’-dicarboxaldehyde and 1,2-phenylenediamine were reacted to an imine functionalized macrocyclic chiral compound. After introduction of an ether linkage with terminal ene group, it was immobilized to thiol silica via thiol-ene click reaction. The resultant CSP was evaluated in normal-phase and reversed-phase elution modes for its enantiomer separation capability which could be verified by full baseline separation of structurally diverse chiral compounds. Lv, Wang, and coworkers introduced a new composite material as CSP which was obtained by immobilization of reduced porous organic cages with homochiral secondary amine structure on silica gel, followed by coating of amylose tris(3,5-dimethylphenyl carbamate) [7]. The resultant material was carefully characterized by FT-IR (indicating the immobilized carbamate functionalities of the polysaccharide selector), elemental analysis (which confirmed the presence of nitrogen-containing cage and polysaccharide selector structures), nitrogen adsorption–desorption measurements (providing information on the specific surface area of 305 m²/g and pore size of 7.6 nm), thermogravimetric analysis (showing a significant weight loss between 100°C and 800°C of 18.8% which corresponds to the mass of the immobilized organic layer), and scanning electron microscopy (SEM) (showing an increased surface roughness on the silica beads after surface modification). A number of structurally different chiral compounds could be much better resolved on the amylose carbamate-modified cage-like CSP while the cage-like stationary phase without amylose derivative showed weak enantiomer separation performance. New materials may open new avenues for separations and a considerable attention has been recently paid to covalent organic frameworks (COFs), which are porous crystal materials consisting of multiple, organic, small-molecule monomers connected by reversible covalent bonds. Chiral COFs for chiral separation are typically immobilized on silica gel. Such type of composite material has been proposed by Wang, Xie, and coworkers [8]. A chiral core–shell composite material was prepared by stirring 2,4,6-trihydroxybenzene-1,3,5-tricarbaldehyde, 2,5-diaminopyridine, (S)-2-methylpyrrolidine and 3-aminopropylsilica at room temperature for 7 days in dichloromethane (see Figure 1). A comparison of powder x-ray diffraction patterns, FT-IR spectra, and SEM images (Figure 1) of chiral COF material immobilized to silica, aminopropyl silica, and chiral COF material (not immobilized) indicated the successful surface modification. The chiral COF composite material could separate the enantiomers of structurally different chiral compounds under normal-phase conditions. Another article of the collection reviewed the synthesis approaches of COF materials and their utility in separation science with specific focus on biomedical applications [9]. Their exciting features such as chemical diversity from various covalently linked organic building blocks, high surface coverage through their nanoporous structures, and tunable porosity have raised great attention of material and separation scientists. Synthesis and chemical structures like dedicated applications are manifold and the review by Nala, Gajula, and coworkers gives a good overview of some recent trends.

During drug discovery, new drug candidates are subjected to numerous in vitro characterization methodologies. Among others, the lipophilicity of new drug candidates is experimentally determined as it can influence the distribution within the body. One concept that has become popular and reached experimental practicality is immobilized artificial membrane (IAM) chromatography, where phospholipid membranes immobilized onto silica mimic biological interactions to simulate the transport across membranes. This chromatographic mode has the advantage of ease of automation, is connected to little compound consumption, and produces minimal waste. In one work from this collection, a new IAM phase prepared by copolymerization of 2-methylacryloxydodecyl phosphatidylserine (MDPS) with ethylenedimethacrylate (EDMA) was prepared in capillary format for nano-high-performance liquid chromatography (nHPLC) to predict the permeability for the blood–brain barrier [10]. This PS-mimicking phospholipid membrane polymer monolithic column showed decent correlations with correlation coefficient of 0.73 between log BB (of compounds with known blood–brain permeability) and the CHI IAM at pH 7.4 as measured with the poly(MPS-co-EDMA) monolith, potentially useful for a first quick estimation of the blood–brain barrier. Kubo and coworkers took a closer look into the retention and selectivity characteristics of fluorous affinity chromatography [11]. A fluoroalkyl-modified silica gel column was synthesized and the retention of polyfluoroalkyl substances was investigated and compared to a corresponding alkyl-silica phase without fluorine substitution. It was found that retention of fluorine compounds increased with their number of fluorine atoms on the fluorine-containing stationary phase, while all nonfluorine compounds were not retained. The binding increments of fluorous affinity were essentially established on the new stationary phase and compared to a commercial fluorous alkyl stationary phase (Fluofix-II). It could be demonstrated that the new fluorinated column showed higher retention of fluorinated compounds compared to the tested commercially available fluorinated column.

Several articles reported the synthesis and application of new ion-exchange or mixed-mode ion-exchange materials. Huang et al. described the synthesis of monodisperse PS-DVB microspheres as stationary phase matrix which was activated by Friedel–Crafts acylation with chloroacetylchloride followed by nucleophilic substitution functionalization using mercaptosuccinic acid as well as further modification of the remaining double bonds on the beads by thiol-ene click reaction with mercaptosuccinic acid to increase the capacity of the new weak cation exchange resins [12]. Six common cations (Li⁺, Na⁺, NH₄⁺, K⁺, Mg²⁺, and Ca²⁺) could be separated within 25 min. Liu and coworkers utilized COF technology to construct a poly(styrene-co-divinylbenzene) based anion-exchanger for the separation of conventional anions, organic acids, and carbohydrates [13]. Thus, poly(styrene-co-divinylbenzene) microspheres were loaded with p-phenylenediamine-1,3,5-triformylphloroglucinol derived nanoparticles constructed by an in situ growth method. The imine groups were then reduced to amino groups and subsequently quaternized. The well-characterized COF-based anion-exchanger microspheres showed good selectivity for the above-mentioned analytes and excellent stability. The same group proposed another chemistry to prepare an organic polymer-based anion exchanger by curing reaction of triglycidyl isocyanate with triethylenetetramine on the surface of PS-DVB microspheres [14]. Since mere PS beads have low chemical reactivity for further surface modification, a crosslinking surface functionalization affording a polymer network on top of the silica beads was found a viable option to generate high capacity anion-exchanger beads. Following the above network formation, the available amino groups on the beads were quaternized and then successfully tested for the separation of inorganic anions, organic acids, and carbohydrates.

Mixed-mode reversed-phase/anion-exchange type stationary phases were also in the focus of the special collection. Wolter et al. developed a series of mixed-mode stationary phases obtained by either immobilizing triphenylallylsilane or 2-propinyl-triphenylphosphonium bromide onto thiol-silica by thiol-ene click reaction, leading to triphenyl-type and triphenylphosphonium-type mixed-mode phases [15]. The latter stationary phase was oxidized with performic acid affording sulfonic acid moieties of the residual thiol groups on the modified silica surface giving this material a zwitterionic character. Zeta potential measurements clearly showed a distinct surface charge character of the three mixed-mode stationary phases and hence complementarity in retention and selectivity profiles to each other and commercial and in-house benchmarking stationary phases. The triphenylphosphonium SAX stationary phase was found to be particularly interesting for oligonucleotide separations, for example, for siRNA Patisiran. The double-tethered surface bonding of this phase makes it more stable and compatible with biochromatography conditions such as high salt in ion-exchange type separations. The great potential of mixed-mode chromatography in challenging separations was outlined in an article on the separation of glucosinolates, which are a group of secondary metabolites in Brassica vegetables [16]. Due to their anionic form, structural diversities, and coexistence of other phenolic compounds, their separation represented a significant challenge on common RP type columns. Using a commercial BEH C18 AX column versus a conventional C18 column, it could be demonstrated that the optimized method for the mixed-mode column was applicable to the complex Brassica vegetable samples and could, in addition to the 17 well-resolved glucosinolate peaks, also separate 34 peaks for phenolic compounds which therefore could be identified in broccoli microgreen, suggesting the successful, better application scenarios for qualitative analysis in comparison with the single-mode reversed-phase C18 column.

With the increasing importance of biopharmaceuticals and new drug modalities with more complex structures than traditional drugs, new challenges arise for the characterization of their critical quality attributes. LC plays an important role, yet stationary phases often need adjustments in surface chemistries and morphology. Ke and coworkers developed a series of stationary phases for the fractionation of low-molecular-weight heparin (LMWH) with tailored pore size by pseudomorphic synthesis of pore size–tunable mesoporous silica spherical particles [17]. Spherical silica with narrow pore size distributions in the range of 40–80 Å could be obtained by a combination of quaternary ammonium cationic surfactants and different kinds of swelling agents, including polypropylene glycol, 1,3,5-trimethylbenzene, alkanes, and alkanols. Larger pore sizes in the range of 110–200 Å were obtained by using alkyl imidazolium ionic liquid surfactants as pore (size) tailoring agent under mild conditions. The materials were modified with diol chemistry and employed for the fraction of LMWHs by size-exclusion chromatography. Pore size is one of the key factors for the selectivity and performance in SEC. However, nonspecific adsorption may influence SEC performance. Actually, even after decades of optimization of diol-phases for SEC application, nonspecific interactions with silanols or hydrophobic patches on such materials could not be completely offset, as desired in this noninteractive chromatographic mode. However, recently researchers of Waters could make a major step forward with a new surface chemistry which they present in a research article of this special collection. Lauber and coworkers introduced a bridged ethylene polyethylene oxide (BE-PEO) surface chemistry to improve their packing materials for wide-pore size-exclusion chromatography [18]. In their work, a series of different BE-PEO surface-modified SEC materials were synthesized on 1000 Å, 3 µm silica particles by surface modification with a mixture of hydroxy(polyethyleneoxy)propyl]triethoxysilane (HO-PEO) and 1,2-bis(triethoxysilyl)ethane (BTEE) (see Figure 2). These trifunctional silanes can be attached to the surface and also crosslink between each other due to BTEE leading to a crosslinked polymeric network of ligands covalently bonded to the surface. The inertness of the bonded phase could be significantly improved due to fewer unreacted silanols exposed to the surface. The chromatographic performance of BE-PEO silica-packed columns was evaluated by separations of a protein test mixture, a DNA ladder, monoclonal antibody-based therapeutics, and adeno-associated viruses (AAVs).

With highly efficient modern UHPLC columns, detrimental effects from extra-column peak dispersion contributions and nonspecific interactions with metal frits and other stainless-steel column and system surfaces become visible, particularly in the form of peak tailing. Also, low recovery due to irreversible adsorption is sometimes a problem and requires preconditioning of the column and system with a representative sample prior to injection of the real study samples. Multiple vendors have introduced solutions to this problem by surface inactivation of column and system hardware with polymer coatings. Two studies of the special collection addressed this issue. In one study, multiple reversed-phase type columns (Accucore C18, CORTECS Shield RP18, Acquity HSS T3, Acquity Premier HSS T3, Accucore 150-C4, Accucore PFP, Synergi Polar RP, and Acclaim WCX-1) were screened for their performance in the analysis of corrosion inhibitors which are nasty molecules from chromatography perspective due to their sticky quaternary ammonium, imidazoline, and phosphoric ester functionalities [19]. From these experiments, it turned out that shielding of stationary phase silanols, for example, by trifunctional grafting of the Acquity HSS T3 stationary phase, and inactivation of column hardware surface, for example, as realized by Acquity Premier HSS T3, appeared to be promising strategies. The latter gave the best performance. Rudaz and coworkers demonstrated that the passivated hardware of a biphenyl stationary phase had clearly significant benefits in steroid analysis, in particular for sulfated and glucuronated steroids [20]. The peak asymmetry of sulfated species could be significantly reduced and was instrumental for the efficient simultaneous analysis of 52 steroids in plasma including their Phase II metabolites.

A number of papers in the special collection were related to challenges in the analysis of biomolecules and bioparticles. Bartolini, Sardella, and coworkers investigated different columns and uncommon nonaqueous reversed-phase (NARP) chromatography for the separation of vitamin D2 and D3 (one additional double bond and methyl group) [21]. The two, although not being isomers, are difficult to separate by common RPLC as the additional double-bond of D3 reduces the retention while the additional methyl group increases retention, leading to similar retention characteristics. The RP type stationary phases were evaluated which specifically differed in their hydrophobicity and silanophilic activity, including a GraceSmart RP C18 column without silanol endcapping, a Robusta RP C18 column with silanol endcapping, and a Waters Xbridge RP C18 column with ethylene-bridged hybrid (BEH) particle technology. The Xbridge C18 stationary phase exhibited the most favorable performance, leading to a resolution of 1.6 with the employed NARP allowing the use of LC-UV for the analysis. Pons and Marcus presented a method for the isolation of urinary extracellular vesicles (EVs) and exosomes, respectively, via hydrophobic interaction chromatography using a nylon-6 capillary-channeled polymer (C-CP) fiber column (see Figure 3) [22]. Exosomes, a subset of EVs in the size range of 30–150 nm, are of significant interest as sample material for biomedical applications such as diagnostic testing and therapeutic delivery. They can be isolated from various biofluids (urine, blood, and saliva) by techniques such as lengthy ultracentrifugation or SEC with limited selectivity and purity. Pons and Marcus suggest step-gradient hydrophobic interaction chromatography method with nylon-6 C-CP fiber column as isolation technique. It provided increased isolation efficiency, higher loading capacity, and more gentle EV elution. Various characterization methods confirmed the quality of the isolated exosomes and the method turned out to be a low-cost and time-efficient (< 20 min) method (Figure 3). In another study of bioparticle separations, Černigoj and coworkers reported on the column reproducibility of a quaternary amine (QA) chromatographic monolithic column, also in the course of upscaling, for AAV viral vector manufacturing using linear potassium chloride (KCl) gradient for elution [23]. A key challenge in this field is the separation of full AAV capsids from undesired nonfunctional (empty, partially filled, etc.) capsids. Since their separations are sensitive to even small changes of the stationary phase, the column reproducibility is of prime importance. In their work, the authors demonstrated the high quality of their monolithic column and excellent repeatability.

This special collection of trends in stationary phases and column technology is a nice representative cross-section of current technologies and innovations from the leading groups on stationary phase synthesis and column technology development. Their contributions are highly appreciated.