When HILIC retention drifts, equilibration is not the only suspect. Learn how solvent bottle materials, buffer chemistry and salt leaching affect reproducibility.
Why a seemingly stable HILIC method moves overnight
A HILIC method may give tidy peaks during the first few injections and then shift steadily, even when the pump pressure and gradient are unchanged. The instinctive response is to add more re-equilibration time. That often helps, but it is not a complete diagnostic strategy. HILIC depends on a water-enriched layer associated with the stationary phase, as well as surface adsorption and electrostatic interactions. Small changes in the ionic environment can therefore affect retention more strongly than in many reversed-phase methods. See our guide to HILIC retention mechanisms for the fundamentals.
The surprising solvent-bottle study
Serafimov and colleagues compared borosilicate-glass and perfluoroalkoxy (PFA) solvent reservoirs in HILIC. For their tested metabolites, average retention-time RSD over 30 injections in 12 hours fell from 8.4% with borosilicate bottles to 0.14% with PFA bottles. That is a result from their specific experimental setup, not a promise for every column, solvent or laboratory. The authors hypothesised that ions released from glass altered the surface-associated water layer and electrostatic screening. The finding is important precisely because a new column or longer equilibration would not necessarily solve a changing reservoir composition.
Distinguish drift from poor injection precision
Plot retention time against injection number, not only the mean and standard deviation. A monotonic shift across a batch suggests a changing system state; random scatter points more toward injection variability, gradient mixing or unstable temperature. Include at least two compounds with different charge states if your method permits it. Drift that moves cations and anions in opposite directions may implicate changing electrostatics. A shift confined to early peaks could indicate an injection solvent mismatch, which we explain in our sample-solvent guide.
A practical, controlled troubleshooting experiment
First, prepare fresh mobile phases from the same stock reagents, record preparation order, buffer concentration and water source, and document the exact bottle material. After full conditioning, run a fixed standard mix repeatedly without sample matrix. If possible, compare compatible glass and inert-reservoir systems while holding all other parameters constant. Change only one factor at a time: reservoir material, buffer batch, pH preparation method, injection solvent or equilibration volume. Never substitute a bottle material without checking chemical compatibility and the laboratory safety procedure.
Do not forget instrument and column causes
Temperature variation, inadequate initial-gradient restoration, nonrepeatable proportioning-valve operation, column ageing and sample carryover can also change retention. Record pressure traces and gradient dwell volume; compare a fresh quality-control preparation with an old one. An apparent retention shift may even come from peak-picking errors when the true peak broadens or coelutes. Use raw chromatograms for the diagnosis rather than relying entirely on processed peak tables.
Take-home lesson
A stable solvent recipe is not necessarily a chemically stable mobile phase throughout a long analytical sequence. Container materials should be part of the method documentation and deviation investigation. The Serafimov result is an unusually compelling reason to include the solvent reservoir in a HILIC robustness study. For a complete development workflow, see our HILIC method guide.
Lab-ready checklist
- Inject the same mixed standard at least ten times after documented equilibration.
- Graph retention time versus injection index, separated by analyte class.
- Check fresh buffer, reservoir composition, inlet frits and preparation sequence.
- Repeat one-factor-at-a-time tests before replacing the column.
Frequently asked questions
Can glass bottles always cause HILIC drift?
No. The published effect was method-dependent; many HILIC methods work reproducibly with glass. It is a controllable variable worth testing, not a universal diagnosis.
Does longer re-equilibration fix ion leaching?
Not necessarily. Equilibration addresses column conditioning; continuous changes in mobile-phase composition need a different intervention.