What the analyst sees

The symptoms of insufficient retention are recognisable and fairly consistent:

• The analyte elutes at or very close to the dead time, with a retention factor below roughly 0.5.
• Peak shape degrades as the injection volume increases.
• Response varies strongly between matrices even though the extraction did not change.
• Isobaric matrix components co-elute in the void region, so selectivity depends entirely on the mass spectrometer.
• Recovery from a solid-phase extraction step is poor or irreproducible, because the sorbent has the same retention problem the column does.

A molecular ion visible in the solvent front is not a robust chromatographic measurement. It is a detection without a separation, and it will behave differently in every real sample.

Why reversed phase fails here, in one paragraph

Reversed-phase retention is driven by the analyte partitioning out of a polar mobile phase into a non-polar stationary phase. A molecule that is small, highly hydrated and often permanently charged has no reason to leave the water. Trifluoroacetate, glyphosate, haloacetic acids, sulfamic acid, short-chain sulfonates and many pesticide transformation products fall into exactly this category. These are the substances that European chemicals policy has been grouping under the persistent, mobile and toxic (PMT) and very persistent, very mobile (vPvM) labels — a regulatory push that is one reason the analytical demand has grown so quickly [1].

HILIC and IC solve the problem differently. HILIC uses an acetonitrile-rich mobile phase and a polar stationary phase carrying a water-enriched layer; the analyte partitions into that layer, with additional hydrogen bonding and electrostatic contributions. Suppressed anion-exchange IC retains the analyte by ion exchange against a hydroxide gradient, and the suppressor converts that eluent to water before it reaches the ion source. The consequence matters: IC allows direct injection of an aqueous sample without any organic solvent at all.

Troubleshooting table

The deciding question is charge, not polarity

“Very polar” is too coarse a description to choose a method with. The useful question is what charge the analyte carries in aqueous solution, and whether that charge is permanent.

• Permanent or strong anions — sulfonates, phosphonates, oxyanions, small halogenated carboxylic acids: suppressed anion-exchange IC is the natural fit.
• Weak acids whose ionisation can be controlled by pH: both modes are plausible; HILIC may win on runtime, IC on retention strength.
• Neutral polar compounds — sugars, small amides, many metabolites: HILIC, because there is no charge for ion exchange to act on.
• Bases and zwitterions — amino acids, catecholamines, many drugs: HILIC, usually on a zwitterionic or amide phase.
• Mixed panels spanning several of these: expect to run two methods. A single injection covering all of them is the exception, not the rule.

What IC-MS actually delivers

A recent validated IC-HRMS method for 23 highly polar contaminants — haloacetic acids, pesticide transformation products and other mobile, persistent substances — reported limits of quantification of 0.03–0.37 µg/L over a linear range of 0.01–100 µg/L, with aqueous samples injected directly and no enrichment step [2]. (Peer-reviewed evidence.)

The same work addressed the practical weakness of long-running IC-MS sequences. By using the autosampler to co-inject stable isotope-labelled internal standards, the authors maintained continuous quality control of retention time, signal intensity and mass accuracy. Over four months of measurement, retention-time deviation stayed below 12 seconds and mass deviation below 2 ppm [2]. (Peer-reviewed evidence.) Applied along a drinking-water treatment train, the method confirmed trifluoroacetic acid, sulfamic acid, cyanuric acid and chlorothalonil transformation products in the low microgram-per-litre range.

Two things are worth extracting from that. First, direct injection removes the enrichment step — and with it a major source of analyte loss for compounds that are difficult to retain on any sorbent. Second, co-injected internal standards turn a drift problem into a monitored parameter. That is a transferable idea for any long-running quantitative method, not only IC.

The relevance of these particular analytes is not academic. Trifluoroacetate accumulates irreversibly in the water cycle [3], and pesticide use has been identified as a substantial source of it [4].

What HILIC delivers — and what the word “HILIC” hides

HILIC covers a far wider analyte range than IC because it does not require a charge. A zwitterionic HILIC-HRMS method developed for non-target screening of industrial wastewater reported average relative standard deviations below 6% for intensity and below 1% for retention time across all substances in solvent, influent and effluent (n = 10) [5]. (Peer-reviewed evidence.) That contradicts the common assumption that HILIC is inherently irreproducible — with adequate equilibration and buffer control, it is not.

The real trap is treating HILIC as one selectivity. A four-laboratory comparison of twelve chromatographic methods found that different HILIC phases behaved so differently from one another that their retention-time correlation was very weak, despite carrying the same mode label [6]. Amide, zwitterionic and other polar chemistries combine partitioning with different hydrogen-bonding and electrostatic contributions.

The same comparison is the strongest available argument against picking a single winner: no individual method exceeded roughly 80% coverage of the test set, and the authors recommended using at least two complementary methods for suspect or non-target screening in water analysis [6]. Runtime is a real consideration in that choice — the SFC methods in that study ran 15 minutes, while the HILIC methods ranged from 25 to 40 minutes [6]. (Peer-reviewed evidence.)

Two decision scenarios

Scenario 1: a targeted panel of small acids in drinking water

You need to quantify trifluoroacetate, haloacetic acids and a handful of pesticide transformation products at low microgram-per-litre levels, with defensible numbers for a regulator.

1. Choose IC. Every analyte is anionic, and direct injection avoids an enrichment step that would recover them poorly.
2. Budget for the suppressor. It is a consumable and a failure point; monitor background conductivity.
3. Co-inject labelled standards rather than adding them only to the sample, so that retention time, intensity and mass accuracy are tracked continuously.
4. Check ionic strength tolerance with a dilution series in the real matrix before validating.
5. Accept the negative-mode restriction and cover anything cationic with a second method.

Scenario 2: broad screening for mobile contaminants

You run non-target screening on surface water and want to stop missing the polar end of chemical space.

1. Keep reversed phase as the baseline. It still covers conventional micropollutants most efficiently and has the most mature libraries.
2. Add one orthogonal method chosen for the gap that matters — HILIC for broad polar coverage, IC if the concern is specifically small anions.
3. Screen two HILIC chemistries before concluding that the mode does not suit your analytes.
4. Tune the data processing per platform. Peak widths differ substantially between modes, and an RP-tuned peak-picking setting will silently discard broad IC peaks.
5. Apply a retention-factor threshold so that void-volume detections are not counted as findings.

Practical starting points

The following are practical starting points, not validated conditions. They are offered as a place to begin screening, and every one of them will need optimisation for a specific analyte panel.

• HILIC screening: compare one amide and one zwitterionic phase under otherwise identical conditions before optimising either.
• HILIC equilibration: allow substantially longer than the reversed-phase habit; verify with a retention marker across a full sequence rather than assuming.
• HILIC diluent: match the acetonitrile-rich starting composition wherever solubility permits, and confirm the sample stays clear after mixing.
• IC injection: begin with a modest volume and establish the tolerance to ionic strength by dilution before pushing sensitivity.
• Both modes: include a labelled internal standard that experiences the same retention mechanism as the analyte.

Column families worth screening include zwitterionic and amide HILIC phases from several manufacturers, and hydroxide-selective anion-exchange columns designed for MS-compatible suppressed operation. Specific part numbers depend on internal diameter, particle size and pressure limits; confirm current specifications on the manufacturer’s own product page before ordering.

Limitations and when to look elsewhere

Neither mode is a universal answer. IC is restricted to ionic analytes and, in the suppressed anion configuration, to negative-mode detection. Its peaks are broader than reversed-phase peaks, which affects both throughput and the software settings needed to detect them. HILIC is more versatile but more sensitive to equilibration, buffer composition and sample diluent, and its selectivity is strongly phase-dependent.

Where both fall short, supercritical fluid chromatography and capillary electrophoresis are the usual alternatives, and the platform comparison cited above found SFC surprisingly capable across a wide polarity range [6]. Enrichment deserves separate attention: concentrating a sample also concentrates the salts, which can shift matrix effects substantially in modes where salts are retained rather than passing through the void.

If quantitative accuracy is the goal and the matrix varies in ionic strength, the method choice is only half the problem. The calibration strategy is the other half.

Action checklist

• Determine the charge state of every analyte at the intended mobile-phase pH.
• Separate the panel into anionic, neutral and cationic groups before choosing a mode.
• Calculate retention factors rather than judging retention by eye.
• Screen at least two HILIC chemistries before rejecting the mode.
• Test IC tolerance to real-sample ionic strength with a dilution series.
• Match the sample diluent to the initial mobile phase in HILIC.
• Co-inject labelled standards to monitor drift over long sequences.
• Adjust peak-picking parameters separately for each chromatographic mode.
• Plan for two complementary methods when the panel is broad.

Bottom line

The choice between IC and HILIC for very polar analytes is not a contest between two techniques. It is a question about your analytes: whether they carry a charge, whether that charge is permanent, and whether you need targeted quantitation or broad coverage. Ion chromatography gives strong, defensible retention for anions and allows direct aqueous injection. HILIC covers a far wider chemical range but demands more care with phase selection, equilibration and sample diluent. For screening work, the published evidence points consistently in one direction: use two complementary separations rather than searching for one method that does everything.

Frequently asked questions

Can HILIC retain small anions such as trifluoroacetate?

Some zwitterionic phases retain them to a degree, because electrostatic interactions contribute to HILIC retention. But retention is generally weaker and more condition-dependent than on a dedicated anion-exchange column. For a targeted method at low concentrations, IC is the more robust choice.

Does IC-MS require special instrumentation?

It requires an ion chromatograph with a suppressor that converts the hydroxide eluent to water before the source. The mass spectrometer itself is a conventional LC-MS instrument. The suppressor is the component that makes the coupling possible and is also the part that needs monitoring.

Why does my HILIC method drift when the reversed-phase method never did?

HILIC retention depends on a water layer held on the stationary-phase surface. That layer takes longer to establish than a reversed-phase equilibrium and is sensitive to buffer concentration and temperature. Most drift complaints trace back to equilibration that was adequate for reversed phase but not for HILIC.

Is one method ever enough for polar screening?

The published comparison found that no single chromatographic method covered more than roughly 80% of a broad test set, and the authors explicitly recommended at least two complementary methods [6]. For targeted quantitation of a narrow, chemically similar panel, one well-chosen method is often sufficient.

Should I enrich the sample first?

For very polar analytes, enrichment is often the weakest step — the same properties that prevent chromatographic retention also prevent sorbent retention. Direct injection, where sensitivity allows it, avoids that loss entirely and is one of the practical arguments for IC.

Related reading on LC–MS Guide

• How HILIC retention works: five mechanisms behind the peak — why one HILIC phase is not another.
• RP-LC vs HILIC vs SFC vs IC-HRMS: why no single method sees everything — the four-laboratory comparison behind several of the arguments above.
• Sample solvent and gradient mismatch in HPLC — the most common cause of HILIC peak distortion.
• Ion suppression in LC–MS: causes, diagnosis and solutions — what happens when salts and matrix reach the source together.

References

1. Mohr T, Schliebner I, Neumann M, Oules L, Arp HPH, Hale SE. Progress in European chemicals policy to support the protection of the environment and human health from persistent, mobile and toxic and very persistent and very mobile substances. Environmental Sciences Europe 2024. doi:10.1186/s12302-024-00932-7
2. Flottmann J, Schmidt TC, Schips T, Schmidt R, Bader T, Winzenbacher R, Seitz W. Ion chromatography high-resolution mass spectrometry — quality assurance by co-injection of internal standards. Analytical and Bioanalytical Chemistry 2026;418:3397–3411. doi:10.1007/s00216-026-06450-0
3. Arp HPH, Gredelj A, Glüge J, Scheringer M, Cousins IT. The global threat from the irreversible accumulation of trifluoroacetic acid (TFA). Environmental Science & Technology 2024;58(45):19925–19935. doi:10.1021/acs.est.4c06189
4. Joerss H, Freeling F, van Leeuwen S, et al. Pesticides can be a substantial source of trifluoroacetate (TFA) to water resources. Environment International 2024;193:109061. doi:10.1016/j.envint.2024.109061
5. Armin R, Wachendorf J, Weber M, Schmidt TC. Enhanced industrial wastewater monitoring: method development for non-target screening of highly polar substances using ZIC-HILIC-HRMS. Analytical and Bioanalytical Chemistry 2025;417(1):167–181. doi:10.1007/s00216-024-05635-9
6. Zweigle J, Schlüsener M, Flottmann J, Bader T, Vidkjær NH, Bollmann UE, Christensen JH, Tisler S. Not one method to rule them all: a comparative study of chromatographic platforms (RP-LC-, HILIC-, SFC-, and IC-HRMS) for water analysis. Analytical Chemistry 2025;97(45):25099–25110. doi:10.1021/acs.analchem.5c04114
7. Pasquini L, Lardy-Fontan S, Rosin C. Pesticide transformation products: a potential new source of interest for drinking water. Environmental Science and Pollution Research 2025;32(8):4619–4635. doi:10.1007/s11356-025-35979-3

Editorial note. All numerical performance figures above are attributed to the cited studies and apply only to the conditions those authors reported. Values labelled as practical starting points are not validated conditions and must be optimised for a specific analyte panel and matrix. The three figures are original LC–MS Guide illustrations; no published figure has been reproduced. LC–MS Guide has not independently reproduced any of these methods.