Types of Column in HPLC: USP L-Codes & Selection Guide

Written by Darshan Singh, M.Sc. Chemistry, 23+ years in pharmaceutical quality control, quality assurance and drug regulatory affairs. Reviewed against USP General Chapter 621 (harmonised text, official 1 December 2022) and ICH Q2(R2).

Most articles on this topic stop at the eight modes. In a regulated laboratory that is not enough: a monograph names a packing by L-designation, an inspector asks whether your substitution was permitted, and two columns carrying the same L-code can give visibly different chromatograms. This guide covers the modes, the USP L-code system, what USP ⟨621⟩ actually lets you change, and the traps that cost columns and out-of-specification investigations.

How HPLC columns are classified

Column classification only becomes useful when you separate three independent axes. Mixing them is why “what type of column is this?” so often gets a confused answer.

A working description therefore reads: “150 × 4.6 mm, 3.5 µm fully porous C18, USP L1, 100 Å pore” — mechanism, architecture and pharmacopoeial identity in one line. If you inherited a method that names only a brand, reconstruct this line before you attempt any transfer. See our companion guide to the working principle and instrumentation of HPLC for how the column sits inside the wider system.

The USP L-code list: how HPLC columns are classified as per USP

The L-nomenclature was introduced in the Fourth Supplement to USP XIX in 1978 and has grown continuously since. USP maintains a Chromatographic Columns Database listing the columns actually used to validate the procedures published in USP–NF; a designation is assigned a number only when the text referring to it becomes official.[3] Published listings now run past L120.[2]

Choosing an HPLC column: a working selector

Column types in detail

1. Reversed-phase columns

Retention comes from hydrophobic interaction between nonpolar analytes and an alkyl-bonded surface, eluted with an aqueous mobile phase modified with acetonitrile or methanol. It is the default mode for small-molecule pharmaceutical assay and impurity work because it is robust, reproducible and compatible with mass spectrometry.

L1 C18 · L7 C8 · L26 C4 · L11 phenyl · L43 PFP · L60 polar-embedded amide.

Two practical limits govern the mode. First, conventional C18-silica columns are typically recommended only up to about pH 8, above which the silica support begins to dissolve and the bonded phase hydrolyses, costing efficiency and retention.[6] Organic–inorganic hybrid particles were developed specifically to widen that window; check the manufacturer’s certificate for the range claimed for your part number rather than assuming a general figure. Second, highly aqueous mobile phases can cause retention loss — discussed below.

2. Normal-phase columns

Adsorption of polar analytes onto a polar surface, eluted with a nonpolar mobile phase such as hexane or heptane with an alcohol modifier. Bare silica L3, amino L8, diol L20 and cyano L10 are the usual packings. The mode retains real value for positional and geometric isomers and for lipid work, but retention drifts with trace water in the solvent, and the mobile phases are poorly suited to mass spectrometry and to routine QC in a humid climate.

3. HILIC columns

Hydrophilic interaction chromatography retains very polar analytes that reversed phase cannot hold, using a polar stationary phase with a high-organic mobile phase and a water-rich layer at the surface. Bare silica L3, amide L60 and zwitterionic phases are the common choices. HILIC is MS-friendly, which is its main advantage over ion-exchange for polar metabolites, but it is sensitive to the water content of the mobile phase and needs longer re-equilibration than reversed phase.

4. Ion-exchange columns

Charged analytes are retained electrostatically by fixed charges on the packing. Strong anion exchange L14 holds acids and anions; strong cation exchange in hydrogen L17, calcium L19 or lead L34 form appears in specific carbohydrate and organic-acid monographs. Polymeric supports tolerate a far wider pH range than silica. The cost is salt in the eluent, which limits direct MS coupling.

5. Size-exclusion columns

Separation by hydrodynamic size, with larger molecules excluded from the pores and eluting first. Rigid styrene-divinylbenzene L21 and diol-bonded silica L20 are typical. This is the standard technique for aggregate and high-molecular-weight species testing on therapeutic proteins. Resolution is inherently limited, and any secondary interaction with the packing invalidates the size calibration — which is why ionic strength control matters more here than in any other mode.

6. Chiral columns

Enantiomers are resolved through stereoselective interaction with a chiral selector: polysaccharide derivatives (amylose and cellulose carbamates) dominate, with cyclodextrin, protein-based, ligand-exchange and Pirkle-type brush phases used for specific problems. Where a monograph specifies a chiral packing it carries its own L-designation; confirm the exact code in the current monograph rather than assuming it, because chiral designations are assigned individually and are easy to mis-transcribe. Immobilised polysaccharide phases tolerate a much wider solvent range than coated ones, which matters when the sample only dissolves in a solvent that would strip a coated phase.

7. Affinity and specialty columns

Retention by specific biochemical recognition — Protein A or G for antibody capture and titre, immobilised metal affinity for histidine-tagged proteins, immunoaffinity for targeted extraction, and molecularly imprinted polymers for selective clean-up. Selectivity is unmatched, cost per column is high, and applicability is narrow by design.

8. Mixed-mode columns

Two retention mechanisms on one particle, most often reversed phase plus ion exchange. Adjusting ionic strength and organic content tunes which mechanism dominates, giving selectivity that neither mode reaches alone. Useful for impurity profiles containing both neutral and permanently charged species. Method development is correspondingly harder, and robustness testing under ICH Q2(R2) deserves extra attention because two mechanisms respond to a single parameter change.[7]

Comparison of all HPLC column types

Scroll sideways to see all six columns.

Particle architecture: what actually sets efficiency

Two columns with the same L-code and the same dimensions can differ by a factor of three in plate count, because efficiency is set by particle architecture rather than by bonded chemistry.

Superficially porous (core–shell) particles are the clearest example. Sub-3 µm superficially porous columns rival the efficiency of columns packed with sub-2 µm fully porous particles while generating only about half the back pressure — typically operating at one-half to one-third of the pressure of a fully porous sub-2 µm column.[5] That is what makes them attractive on a conventional 400 bar HPLC: most of the UHPLC efficiency benefit without replacing the instrument. Materials at 1.3 µm have reached minimum plate heights of about 2 µm, corresponding to up to 500,000 plates per metre, but sub-2 µm superficially porous particles only show their benefit beyond roughly 800 bar and need instruments rated to 1000–1300 bar.[5]

If you are budgeting an analytical laboratory rather than choosing a column, the instrument decision sits alongside this one — our comparison of current HPLC instrument models, Indian pricing and service coverage covers the hardware side, and QC instrumentation is a recognised line item in the pharmaceutical plant setup cost calculator.

What USP 621 lets you change — and what it does not

This is the question that decides whether a column swap is an adjustment or a change requiring revalidation. USP General Chapter 621 Chromatography, in the harmonised text official from 1 December 2022, sets out permitted adjustments to a prescribed procedure. Adjustments within these limits are allowed provided system suitability is still met; anything beyond them is a modification.[1]

Common column failures and what actually causes them

Retention loss in highly aqueous mobile phases

Widely described as “phase collapse”, the mechanism is more accurately called phase dewetting. When the organic content falls below about 10%, or reaches zero, the hydrophobic bonded chains self-associate and the mobile phase is expelled from the pores, causing retention shifts and peak distortion. It is reversible by resolvating the phase with a higher organic percentage. Phases with polar embedded groups — amide, urea or carbamate — and those sold with an aqueous-compatible designation resist it.[4] If your method genuinely needs 100% aqueous, specify an L60 polar-embedded amide phase rather than fighting a conventional C18.

Poor peak shape for basic drugs

Residual silanols interacting with protonated bases produce tailing. The corrections are a well-endcapped or polar-embedded phase, adequate buffer strength, and a mobile phase pH set at least two units away from the analyte pK_a so the ionisation state is defined rather than drifting. Hybrid-particle phases give more room to move the pH than conventional silica does.[6]

Rising back pressure

Almost always frit blockage rather than damage to the packed bed. Sub-2 µm columns need 0.3–0.5 µm frits, against 2 µm for 3.5–5 µm columns, so they trap particulates that a conventional column would pass. In-line 0.2–0.45 µm filters, guard columns and disciplined sample preparation address the cause.[4]

Assuming a long re-equilibration is always necessary

The customary ten column volumes is more than most separations need. For non-ionisable solutes in unbuffered eluents, repeatable equilibrium — retention time repeatability better than 0.002 minutes — is reached within approximately two column volumes.[4] Establishing the real figure for your method during development, and writing that into the SOP, recovers meaningful run time across a QC shift. Treat this as a validated method parameter, not a shortcut applied at the bench.

Skipping the guard column

In an accelerated lifetime test, an unprotected column failed at injection 85 through accumulation of milk solids, while a guarded system maintained performance past 160 injections with guard cartridge replacements. Modern guard holders add negligible dead volume.[4] Against the price of an analytical column, this is the cheapest control available. Column care, flushing and storage practice belong in a controlled document — see our SOP for preventive maintenance of HPLC for a worked template covering pump, injector, detector and column.

Frequently asked questions

References

This article is technical and educational content for pharmaceutical professionals. It is not medical advice, laboratory certification or a regulatory submission. Pharmacopoeial texts and Indian statutory instruments are revised frequently: verify every limit, designation and allowable adjustment against the currently official USP–NF, Indian Pharmacopoeia and ICH text before applying it to a validated method. Where a column change may fall outside allowable adjustments, route it through your site change-control procedure.