Written and maintained by CASRAI Editorial Board
Last updated
HILIC — hydrophilic interaction liquid chromatography — exists because reversed-phase columns have a blind spot: very polar, very hydrophilic small molecules (sugars, nucleotides, polar metabolites, many amino acids, small polar drugs) barely retain on C18. They elute at or near the void volume, where matrix and salts also elute, and where quantitation and resolution both suffer. HILIC solves that by inverting the retention mechanism, but it brings three practical traps that catch first-time users hard: equilibration that takes far longer than a reversed-phase method teaches you to expect, sample diluents that split or ghost peaks if you get them wrong, and a mobile-phase composition where retention is far more sensitive to small changes than reversed-phase experience prepares you for.
This guide covers when HILIC is genuinely the right choice over reversed phase, how to pick a stationary phase, and how to develop a method around those three traps rather than discovering them one failed run at a time. For the broader stationary-phase landscape HILIC sits inside, see HPLC column selection: stationary phases & selectivity; for general HPLC fundamentals, see HPLC: columns, mobile phases, and a peak-problem troubleshooting table.
What HILIC actually separates on
A HILIC stationary phase is polar — bare silica, or silica bonded with amide, diol, amino, or zwitterionic (sulfobetaine-type) functional groups — run with a mobile phase that is mostly organic (typically somewhere in the range of roughly 60–95% acetonitrile, with the balance aqueous buffer). At that composition, a water-enriched layer partitions onto the polar stationary-phase surface. Polar analytes partition into that water-enriched layer preferentially over the bulk organic mobile phase; the more hydrophilic the analyte, the stronger the partitioning and the longer the retention. Retention order is broadly the inverse of reversed phase: the compounds that would elute first on C18 tend to retain longest on HILIC, and vice versa.
Because the mobile phase is organic-rich, HILIC also pairs well with electrospray ionization — higher organic content generally improves desolvation and ionization efficiency, which is part of why HILIC shows up so often ahead of LC-MS for polar-metabolite and polar-drug work. See LC-MS explained for how the coupling itself works.
When HILIC beats reversed phase
| Situation | Reversed phase (C18/C8) | HILIC |
|---|---|---|
| Analyte is very polar/hydrophilic (sugars, nucleotides, polar amino acids, many polar drug metabolites) | Poor retention, elutes near void volume with matrix/salts | Strong, tunable retention — this is the case HILIC exists for |
| Analyte solubility is poor in aqueous/low-organic solvent but good in acetonitrile-rich solvent | Forces a compromise mobile phase that hurts either solubility or retention | The organic-rich mobile phase is often close to the analyte’s native solubility profile |
| Downstream detection is ESI-MS and sensitivity is limiting | Aqueous-heavy eluent can suppress ionization for some analyte classes | Organic-rich eluent frequently improves ESI sensitivity |
| Analyte is moderately-to-strongly hydrophobic | Normal, well-behaved retention and selectivity | Little to no retention — wrong mechanism for the analyte |
| Method needs to run near-neutral, dilute-aqueous injections routinely | Native fit | Requires a diluent-matching step (see below) or it will fail |
| Fast re-equilibration between injections/gradients matters (high-throughput screening) | Typically re-equilibrates in a handful of column volumes | Needs materially more column volumes to re-equilibrate — a real throughput cost |
The practical rule: reach for HILIC when reversed phase genuinely cannot retain the analyte class, not as a first-choice default. It solves a specific retention problem at the cost of longer equilibration, tighter diluent control, and (on charged phases) more electrostatic complexity in the retention mechanism.
Choosing a HILIC stationary phase
| Phase | Character | Notes |
|---|---|---|
| Bare silica | Purely polar surface, weakly acidic silanols | Simplest and often most robust starting point; some secondary ion-exchange-like character from residual silanols at low pH |
| Amide-bonded | Neutral, hydrogen-bond-accepting | Lower secondary ionic interaction than bare silica; a common default for basic analytes that would otherwise tail |
| Diol-bonded | Neutral, hydroxyl-rich | Similar in spirit to amide phases; selectivity differs enough to be worth screening as an alternative |
| Zwitterionic (sulfobetaine-type) | Carries both a permanent positive and negative charge on the same ligand | Adds electrostatic selectivity for charged analytes on top of the hydrophilic partitioning mechanism — useful, but makes buffer choice more consequential |
| Amino-bonded | Basic, can act like a weak anion exchanger | Strong retention for acidic analytes (e.g. sugars, sugar phosphates); risk of Schiff-base reactivity with reducing sugars over time — check compatibility before committing |
As with reversed phase, no single HILIC chemistry is universally correct. A short screen — typically bare silica plus one amide or zwitterionic phase — against the actual analyte set, rather than picking on reputation alone, is the efficient path; HILIC selectivity differences between phases are often larger than the equivalent differences between reversed-phase C18 columns from different vendors.
Mobile-phase design: why small changes move retention a lot
This is the trap in the guide’s title, and it is the one that most surprises chromatographers coming from reversed phase. In reversed phase, retention is a comparatively gentle function of %organic — a few percentage points rarely upend a separation. In HILIC, retention is driven by the thickness and composition of the water-enriched layer on the stationary phase, and that layer’s character changes sharply with small shifts in aqueous content. A change of just a few percent in %organic can shift retention times substantially and can reorder elution — the opposite sensitivity profile from what reversed-phase experience trains you to expect, and a common cause of “the method worked yesterday” failures when a mobile phase is remade slightly off-target.
Two further variables compound this:
- Buffer concentration. Increasing aqueous buffer strength generally increases the water content of the adsorbed layer and tends to reduce retention for the hydrophilic-partitioning mechanism, while simultaneously changing the electrostatic screening that governs retention on charged (amino, zwitterionic) phases. The two effects can pull in different directions depending on the analyte’s charge state, which is why buffer concentration needs to be screened deliberately rather than assumed.
- pH. pH sets both the analyte’s ionization state and, on bare silica or amino phases, the stationary phase’s own charge character. A pH that looks fine on paper for analyte solubility can still produce poor peak shape if it puts the analyte and the stationary-phase surface into an unfavourable electrostatic relationship.
Practically: hold %organic, buffer concentration, and pH as three separate variables in a small, deliberate screening design rather than adjusting them ad hoc. Use a volatile buffer (ammonium formate or ammonium acetate at a modest concentration is a common, MS-compatible starting point) unless the detector specifically requires otherwise, since HILIC pairs with LC-MS often enough that buffer volatility is worth defaulting to even for a UV-only method that might later move to MS.
The equilibration trap
Because retention depends on a water-enriched layer that has to build up on the stationary-phase surface, a HILIC column re-equilibrates far more slowly than a reversed-phase column does — commonly requiring on the order of two to three times as many column volumes, and sometimes more, before retention times stabilize after a mobile-phase change or a gradient run. Three consequences follow directly:
- Isocratic startup takes longer. Don’t inject the first sample the moment the pressure trace looks flat — pressure equilibrates much faster than the adsorbed-layer chemistry does. Confirm retention-time stability by running two or three consecutive blank or standard injections and checking they overlay before trusting the method.
- Gradient methods drift. A HILIC gradient that returns to starting conditions too briefly before the next injection will show retention-time creep across a sequence, because the column hasn’t actually re-equilibrated even though the pump has returned to the nominal starting %B. Build a genuinely longer re-equilibration segment into the method than a reversed-phase gradient would need, and verify it empirically rather than copying a reversed-phase rule of thumb.
- Throughput cost is real and should be budgeted up front. If a method needs to run many samples per day, the longer re-equilibration is a genuine capacity constraint, not a rounding error — account for it when deciding whether HILIC’s retention benefit is worth the throughput trade against a reversed-phase or ion-pairing alternative.
The sample-diluent trap
The mobile phase is mostly organic; if the injected sample is dissolved in a diluent that is mostly aqueous (or otherwise has a lower organic content than the mobile phase at the point of injection), the local solvent strength at the injection zone spikes relative to the mobile phase. On a HILIC column that shows up as band broadening, fronting, or split/double peaks — and it gets worse, not better, as injection volume increases, because a larger volume of the mismatched diluent has more disruptive effect on the local equilibrium at the head of the column.
The practical fix is straightforward once it’s understood as a diluent-matching problem rather than an injection-volume problem: dissolve (or reconstitute) the sample in a diluent whose organic content is at or above the mobile phase’s starting organic composition, not simply “in water” or “in the same aqueous buffer used elsewhere in the assay.” When the native sample matrix is aqueous — a biological fluid, an aqueous extraction — a solvent-exchange or evaporate-and-reconstitute step into an organic-rich diluent is often unavoidable; solid-phase extraction is a common way to do that exchange cleanly, see SPE cartridge selection and method optimization for the wash/elution mechanics. Where the sample simply cannot tolerate a fully organic-rich diluent, keeping injection volume small and diluting the sample in a diluent that at least approaches the mobile-phase organic content usually mitigates the effect even when it can’t eliminate it outright — verify empirically rather than assuming a partial match is good enough.
A practical HILIC method-development workflow
| Step | What to do | What it settles |
|---|---|---|
| 1. Confirm HILIC is the right mechanism | Check the analyte retains poorly (elutes near void) on a standard reversed-phase screen first | Avoids building a HILIC method for an analyte reversed phase would have handled more simply |
| 2. Screen 2–3 stationary phases | Bare silica plus one amide or zwitterionic phase, same generic gradient, same analyte set | Which chemistry gives usable retention and peak shape for this analyte class |
| 3. Screen %organic | Run a shallow gradient or a small set of isocratic points across the practical range for the chosen phase | The organic-content window where retention and resolution are acceptable |
| 4. Screen buffer concentration and pH independently | Vary one at a time against a fixed %organic | Which combination gives stable, sharp peaks without over- or under-retaining charged analytes |
| 5. Fix the diluent | Match sample diluent organic content to the mobile phase’s starting composition; test injection-volume sensitivity | Whether peak shape holds at the injection volume the assay actually needs |
| 6. Set and verify equilibration time empirically | Run consecutive standard injections after each mobile-phase or gradient change until retention times overlay | The real re-equilibration time for this specific column/mobile-phase combination, not a borrowed rule of thumb |
| 7. Confirm system suitability and validate | Resolution, tailing factor, plate count, %RSD across the working range | Whether the method is fit for its intended use — see ICH Q2(R2) analytical procedure validation for which tests apply |
Troubleshooting
| Symptom | Likely cause | What to check |
|---|---|---|
| Retention times drift across a sequence | Incomplete re-equilibration between injections or gradient cycles | Lengthen the equilibration segment; confirm with consecutive-injection overlay, not just pressure stability |
| Split or fronting peaks, worse at higher injection volume | Sample diluent has lower organic content than the mobile phase at the injection point | Match diluent organic content to mobile-phase starting composition; reduce injection volume as a partial mitigation only |
| Retention collapses after remaking mobile phase | Small %organic error during preparation, or a buffer concentration/pH drift | Verify %organic, buffer concentration and pH against the validated method exactly — HILIC retention is unusually sensitive to all three |
| Poor peak shape for basic analytes specifically | Secondary ionic interaction with residual silanols (bare silica) or an unfavourable electrostatic relationship at the working pH | Try an amide or zwitterionic phase instead of bare silica; re-screen pH |
| Loss of retention for charged analytes as buffer strength increases | Increasing ionic strength is screening the electrostatic contribution to retention (charged phases) and/or increasing the adsorbed water layer | Re-screen buffer concentration deliberately rather than assuming more buffer is simply “more robust” |
Frequently asked questions
Is HILIC the same as normal-phase chromatography?
They’re related but not identical. Both use a polar stationary phase with a comparatively non-polar (or less polar) mobile phase, but normal phase classically runs with a fully non-aqueous, non-polar mobile phase, while HILIC deliberately includes an aqueous component whose partitioning into a water-enriched surface layer is the actual retention mechanism. HILIC is often described as a variant that sits between normal phase and reversed phase.
Can I run HILIC and reversed-phase methods on the same LC system without dedicated hardware?
Generally yes — HILIC doesn’t require different pumps or detectors, only a dedicated column (don’t alternate reversed-phase and HILIC runs on the same physical column) and enough re-equilibration time built into the method. Some labs do keep a dedicated HILIC system to avoid solvent-changeover downtime when it’s run frequently.
Why does my HILIC method behave so differently from the reversed-phase method it replaced?
Because the retention mechanism inverted, not just the mobile-phase composition. Elution order is broadly reversed, sensitivity to %organic is much higher, re-equilibration is slower, and sample diluent now has to be organic-rich instead of aqueous. Treat it as a genuinely different method to develop and validate, not a mobile-phase swap on an existing reversed-phase method.
HILIC is a real answer to a real problem — polar analytes reversed phase can’t retain — not a general-purpose alternative to C18. Confirm the analyte actually needs it, screen phase and mobile-phase composition deliberately, and build the diluent match and equilibration time into the method from the start rather than discovering both as failures.
