[Readers Insight] HPLC Column Selection: Core to Method Development (Part II)

In this first half of the article, we discussed how HPLC columns are classified by base material and bonded phase. But still, how should we choose the correct column for an analysis? Below are some common questions chromatographers often encounter, answered in a practical Q&A format.

A: About 70% of pharmaceutical compounds contain basic functional groups. In earlier discussions, we talked about identifying the acidity or basicity of compounds, which helps guide the selection of both mobile phase and stationary phase.

For basic compounds, hybrid silica columns are generally preferred because they reduce peak tailing caused by residual silanol (Si–OH) groups. Endcapped, polar-embedded, or positively charged columns can further improve peak symmetry.

In contrast, when using non-endcapped columns, basic analytes tend to exhibit significant tailing (typically a right-skewed peak) under standard reversed-phase conditions. In such cases, peak shape can be improved by adding ionic modifiers such as potassium hexafluorophosphate (KPF₆) or sodium perchlorate (NaClO₄) to the mobile phase.

For acidic compounds, analysis is usually performed under acidic conditions (pH < 3), where silanol groups are not ionized. This allows for a broader column selection range. Some non-endcapped C18 columns from certain manufacturers can even withstand pH levels as low as 0.5.

Once again, there are no absolute “good” or “bad” columns—only differences in selectivity. Column selection is never unique; a compound may be successfully separated using different stationary phases. The goal of a method developer is to pursue the optimal balance—shorter run times, simpler operation, and high reproducibility. There is no “best” method, only a “better” one, and continuous refinement leads to better separations.

A: The C18 chain is hydrophobic and does not mix well with water. In a pure aqueous system, the C18 chains can collapse or “fold,” creating a phase separation from the water. To overcome this, manufacturers have developed several approaches: reducing bonding density of the C18 chains, applying polar endcapping, or embedding polar groups within the bonded phase. These modifications allow such C18 columns to remain stable and usable under 100% aqueous conditions.

A: The specifications listed in product datasheets provide basic parameters—silica type (hybrid, core–shell, fully porous), bonded phase, column dimensions (length, diameter, particle size), carbon load, surface area, pore size, pH tolerance, pressure limit, and temperature range.

While these parameters may be similar, however, the silica source varies among manufacturers, significantly affecting selectivity. Moreover, labeled values often represent averages rather than absolutes. For instance, a column labeled as having 5 μm particles actually includes a normal distribution centered at 5 μm—the narrower the distribution, the more uniform the packing and the higher the column efficiency. Similarly, a 100 Å pore size rating is also a distribution, and smaller pores can influence compound retention and selectivity. Finally, differences in packing techniques between manufacturers can lead to notable variations in column efficiency and chromatographic performance.

For positional aromatic isomers, use aromatic bonded phases like pentafluorophenyl (PFP) columns. As discussed previously, fluorine is a strong electron-withdrawing group that lowers the electron density of the aromatic ring. Differences in the electron cloud density of the analyte’s aromatic rings at different substitution positions cause variations in π–π interaction strength, enabling the separation of these isomers. PFP columns also provide dipole–dipole interaction effects.

For diastereomers and cis/trans isomers, since their structures are otherwise identical to the main component, they cannot be separated through chemical interaction differences alone. Instead, try using embedded polar group columns or C8 phases to exploit subtle steric differences, which can improve resolution.

A: Add a chelating agent such as EDTA (ethylenediaminetetraacetic acid) or methylenediphosphonic acid (medronic acid) to the mobile phase. Additionally, use an inert-treated column, such as one with a PEEK-lined or diamond-like carbon (DLC) coating, to minimize metal adsorption.

A: A gradual rise in column pressure under normal use typically indicates contamination at the column inlet. You may perform a reverse flush to clear the blockage. However, keep in mind that columns are consumables and cannot be regenerated indefinitely.

If pressure increases within a single sequence, check whether the chromatographic conditions (pH, flow rate, solvent compatibility) exceed the column’s tolerance. If conditions are appropriate, consider system blockage or poor sample solubility in the mobile phase as possible causes.

A: Yes. Modern columns are well-packed with uniform beds and fitted with small-pore frits on both ends, so reverse flushing is generally safe. However, never connect the column to the detector during backflushing, as contaminants may be released and foul the detector cell.

(A side note: Welch's column manuals indicate whether a column is recommended to be reversed-flushed.)

A: Each manufacturer designs columns for a slightly different optimal fitting length. Standard stainless-steel fittings have fixed lengths, which may not always match perfectly. If your system operates below 200 bar, it’s recommended to use PEEK fittings at the column inlet. If a metal fitting gets stuck, gently wiggle and pull it out using needle-nose pliers while holding the tubing near the tip.

Column selection is a complex science in HPLC, and there is no dominate universal rule for it. What truly matters is understanding the general principles and using them as a guide to choose the most suitable column through practical experimentation. By doing so, you can design robust, reproducible, and efficient HPLC methods suited to your analytical goals.