Reinventing Silica Modification for Long-Lasting HPLC Columns

Silica is a widely used support material in chromatography due to its high surface area, mechanical stability, and ease of chemical modification. Silica stationary phases are traditionally prepared through silanization, where silanol groups on the silica surface react with organosilanes to create chemically bonded layers, typically with hydrocarbon chains such as C18 or C8 for reversed-phase separations.

A research team from the Department of Chemistry at the University at Buffalo (State University of New York) introduced an alternative surface modification method that avoids silanization, using diazonium chemistry instead. While diazonium-based functionalization is well established for metals and carbon materials, it has rarely been applied to silica for chromatographic purposes. The method involves generating reactive diazonium salts from aryl amines, which then covalently graft phenyl groups onto the silica surface.

LCGC International spoke to Luis A. Colón, SUNY distinguished professor of chemistry at the University at Buffalo (SUNY), and corresponding author of the paper that resulted from the team’s research (1), about the method.

Silica is a widely used support in chromatography. What key properties make it ideal for supporting stationary phases in high-performance liquid chromatography (HPLC) applications?

Silica is the most used support material for HPLC because of its high mechanical stability, high surface area, and simple surface chemistries that allow surface functionalization with a variety of stationary phases. In addition, silica does not shrink or swell in the presence of typical solvents used in HPLC mobile phases.

Most commercial stationary phases rely on silanization chemistry. How does this process work, and what are its main limitations?

Silica surface functionalization is typically achieved through silanization, a process in which an organosilane containing the desired stationary phase moiety reacts with surface silanol groups. This reaction forms organosiloxane linkages (surface-O–Si–C), effectively grafting the organic species onto the silica surface to serve as the stationary phase.

However, these linkages are prone to cleavage under acidic conditions, which limits the column's lifespan when exposed to highly acidic mobile phases and is exacerbated by high temperatures. The efficient separation of compounds, such as proteins, pharmaceuticals, and peptides, often requires operation under acidic conditions and elevated temperatures, both of which accelerate column degradation. As a result, columns must be replaced more frequently, increasing the overall operating costs. Although current strategies to mitigate this issue are in place, such as stationary phases with sterically protected organosiloxane linkages that enhance alkyl density near the surface and the use of hybrid silicas in which organic functionalities are incorporated into the inorganic silica network, there is still considerable potential for developing new phases that incorporate hydrolytic stability. Such advancements would enhance column durability and extend the operational lifespan under low pH conditions.

What makes diazonium chemistry a compelling alternative to silanization for modifying silica surfaces?

The simplicity of the reaction makes it an appealing option for attaching certain species to silica surfaces. Nevertheless, the diazotization reaction of p-phenylenediamine creates a polymer-like layer on the silica material that provides stability to the attachment process. As we demonstrated, the resulting surface layer, which consists of interconnected phenyl groups, remains stable under acidic conditions, leading to more durable modified surfaces.

Can you explain how the diazotization of p-phenylenediamine (p-PDA) enables the covalent grafting of phenylene and amino groups onto core–shell silica particles?

Diazotization refers to the process of converting an aryl amine into an aryl diazonium salt. These diazonium ions are typically generated in situ under reducing conditions and can attach to metals and oxides through radical formation and covalent bonding, a reaction driven by the release of nitrogen gas. The attachment of diazonium ions typically occurs via spontaneous reduction, generating aryl radicals that subsequently form covalent bonds with surfaces, such as metals or metal oxides. While diazonium-based modifications are widely used for metals, metal oxides, and carbon surfaces, their application to silica surfaces, particularly in chromatographic frameworks, is relatively rare if not entirely absent.

In the case of p-phenylenediamine (p-PDA) and core-shell silica, both amine groups in p-PDA can be diazotized. The resulting diazonium radicals can react with the surface silanol groups on silica, anchoring the moiety via covalent bonding. The experimental XPS data did not show the presence of silicon-carbon bonds; however, signals consistent with silicon-oxygen-carbon linkages suggested that bonding occurred through the surface oxygen atoms. The surface-bound aryl groups and the second diazonium radical can further react with other aryl groups or amines on p-PDA, forming a polymeric-like layer at the surface. Residual amino groups from p-PDA remain within the layer, contributing such a functionality. This interconnected polymeric-like structure contributes to the stability of the surface-modified silica particles.

The thickness of the p-PDA-derived layer was tuned by reaction time and reagent concentration. How does controlling layer thickness impact chromatographic performance and stability?

If the p-PDA-derived layer on the silica-particle surface is allowed to grow too large, it can affect the surface in two ways. First, multiple layers of polymeric-like material can lead to clogging of the pores on the silica particle, reducing the surface area. Second, it introduces excessive material as the stationary phase, which negatively impacts the mass transfer kinetics in the stationary phase, reducing chromatographic performance.

XPS and elemental analysis showed approximately one nitrogen atom per three aromatic rings. What does this indicate about the chemical composition and functionality of the surface layer?

Nitrogen atoms can potentially exist as azo groups bridging two phenyl rings or as amine groups on the surface. Our experiments showed that there is approximately one residual amine group per every 3-4 aromatic rings. This provides a polar moiety at the silica surface, which is otherwise hydrophobic. This is advantageous for the separation of compounds with polar characteristics. Furthermore, the amine functionality at the surface can be used to attach other moieties to the surface to impart new chromatographic selectivity.

Accelerated stability tests demonstrated excellent hydrolytic stability under harsh acidic conditions. What structural features of the diazonium-grafted layer contribute to this robustness?

We attribute the robustness of the grafted layer to the polymeric-like structure at the surface. The linking of the phenyl groups provides stability to the layer.

Preliminary results suggest that the p-PDA-derived phase is promising for peptide separation. What properties make it particularly suited for such applications?

Phenyl type HPLC columns have demonstrated distinct selectivity for certain peptides, suggesting potential similarities with the p-PDA derived phase. Based on this, we anticipated some degree of separation for the peptide mixture used; however, the outcome exceeded our expectations, with nearly all peptides achieving baseline resolution. This phase is well-suited for reverse-phase chromatography, a separation mode commonly used to separate peptides based on hydrophobicity. Additionally, it provides alternative selectivity for peptides containing aromatic residues (such as tyrosine, phenylalanine, and tryptophan-containing peptides) through π–π interactions. The presence of amino groups on the surface can also contribute to selectivity. Under low pH conditions, these groups can become protonated, introducing anionic exchange characteristics. We are eager to further investigate the behavior of the column and its potential for broader peptide applications.

You are exploring the selectivity of the p-PDA phase using the linear solvation energy relationship (LSER) model. How can this approach help predict or understand retention behavior beyond conventional tests like the Tanaka method?

The Tanaka method was originally developed to characterize traditional C18 reversed-phase HPLC columns. However, this test does not account for the potential π–π interactions offered by the p-PDA-derived phase. In contrast, the LSER model provides a systematic framework for describing how solutes interact with the stationary phase, capturing the various contributions to the separation process, such as hydrophobic, dipolar/polarizable, and hydrogen-bonding interactions. Although comprehensive, the original model does not explicitly include parameters to account for π–π interactions, lacking predictive power for aromatic selectivity. To address this, recent modifications to the LSER equation have introduced terms that enhance its ability to differentiate between aromatic and aliphatic solutes, particularly in the phenyl, biphenyl, and pyrene phases. These enhancements will enable a more accurate characterization of the p-PDA-derived phase and its unique selectivity profile.

References

1. Ezzo, J. R.; Salazar, B. L.; Díaz-Santiago, R. J.; Colón, L. A. Silica Surface Modification via Diazotization of p-phenylenediamine: A Stationary Phase for HPLC. Analyst 2025, 150 (21), 4762-4772. DOI: 10.1039/d5an00869g