What a chromatogram is showing

A chromatogram is a plot of detector response against time or against volume. The decision it supports is narrow: something the detector can sense left the column in that window. It does not, by itself, name the molecule, prove purity, or tell you which fraction to keep. The modes that produce these traces, from reversed phase to ion exchange, are set out in chromatography methods in life-science labs. Read that page when you are still choosing a chemistry. Stay on this one when the line is already on the screen and you need to know what sentence it allows.

Hardware classes sit in the scientific instruments catalogue. A protein method can be discussed from the custom protein expression and purification reference. That page is an enquiry prompt for the method. The specification itself goes through the quote request.

Response on one axis, separation on the other

IUPAC describes chromatography as a distribution between a stationary phase and a mobile phase that moves in a definite direction. The chromatogram is the record of that experiment as a detector sees it. The horizontal axis is the clock of the pump, or the volume delivered, which is the same story if the flow is steady. The vertical axis is whatever the detector measures: absorbance, fluorescence, refractive index, conductivity, or an ion current. Change the detector and the picture changes, even when the column did the same separation.

A peak is a rise and fall of that response. Its apex is the time or volume where the response was highest. Its area is the integrated response, which becomes a quantity only after you know how this detector answers this analyte inside a calibrated range. Two compounds with different extinction coefficients can share an area and differ in mass. A protein that is poor in aromatic residues can be abundant and nearly quiet at 280 nanometres.

Column volume is a unit, not a verdict

People quote elution in minutes, in millilitres, or in column volumes. A column volume is the geometric bed volume from the bed diameter and the bed height you measured or the insert states. Elution at three column volumes means the peak centre arrived after three bed volumes of mobile phase. That number travels between pumps better than a raw minute, because flow drops out of the comparison. It still belongs to one chemistry. The same protein can leave an ion-exchange resin during a salt rise and leave a size-exclusion bed near the volume that matches its size. Both can be true. Neither is an identity.

Retention time is method-specific. Column lot, organic fraction, buffer pH, temperature, and extra-column volume all sit inside the number. A paper that says the peptide eluted at 18.4 minutes has told you little until the gradient table travels with it.

Co-elution is the ordinary case

Life-science samples are mixtures. Peptides from a digest, host proteins, and related small molecules often leave a column close together. When the column does not resolve them, one peak contains more than one species. A symmetrical line says the band was not badly distorted. It says nothing about how many molecules share the band. A shoulder, a valley that never reaches baseline, or a peak wider than the standard are hints. Perfect overlap can hide even those hints.

If the analyte is a protein, sequence context from UniProt tells you the mass and the features you might later confirm. It does not assign a peak. Reporting an identification from a proteomics experiment follows community practice discussed by HUPO. A labelled apex is not that report.

A way to read the run

Start with the blank acquired under the same programme. Mark every peak that belongs to the blank. Those times are contaminated until you find the source. Then look at a standard injected in this method. If the standard has moved, broadened, or split, stop. Sample peaks collected on a drifting system become tubes with confident labels and the wrong contents.

For the sample, write three observations before any conclusion. Where the response sits relative to the void, the gradient, and the column volume scale. How wide it is beside a standard of similar retention. Whether the collector put that window into one tube or several. On an ion-exchange gradient, note the salt or pH at the detector after you allow for dwell. A peak before the salt rises can be flow-through. A peak during the rise can be bound material releasing. The trace shows the response either way.

Branch when the picture disagrees with the plan. If the standard and the sample are late by the same amount, suspect flow, temperature, or mobile phase, and fix that before you rewrite the sample story. If the standard is on time and the sample is a broad hump, inject less or slow the gradient, then re-read. If you meant to bind the sample and the only response is in the flow-through, the later tubes are empty of that detector's signal. Assay the flow-through before you discard it.

Injection volume belongs in the record. ISO 8655-1 covers piston-operated volumetric apparatus as a class. The volume you intended and the volume the syringe delivered are different facts when a peak is half the area you expected.

Failure modes that look like data

Integration software will baseline a drift and call it a peak if the threshold is low. Lamp noise, a bubble, or a refractive upset in a steep gradient can fill a table with minor components. Read the raw trace at the same scale as the standard before you trust that table.

A peak at the void is often unretained material, a solvent disturbance, or both. On a preparative ion-exchange run, keep the flow-through as its own fraction and test it. The note on affinity and ion-exchange protein purification treats the binding decision. Here the rule is smaller: the trace and the tube stay paired.

If the pump slowed under rising pressure, a late peak is late on the clock and may be on time in volume. Prefer the volume record when pressure moved. A collector that indexes on time will then be out of step with the chemistry.

Research use and chemical safety

These traces support research decisions about separation and collection. They are not a diagnostic result and not a release specification for a medicine. Organic solvents, acids, and buffers are chemical hazards. Follow the ventilation and waste rules your assessment names. A lysate on a column is also a biological material until your institutional rules say otherwise. The WHO laboratory biosafety manual is a public reference for how institutions frame laboratory biosafety. It does not approve a method.

What to put on the enquiry

State the analyte class, whether the record must be analytical or tied to a fraction collector, the detector principle, the column chemistry, and the axis you need (time, volume, or a gradient signal overlaid). Ask which wetted materials are in scope. Send that with the quote request, and point at the scientific instruments catalogue for the hardware class. If the trace is the record of a protein purification, frame the method question on the custom protein expression and purification reference and keep the claim at the level of a discussion. ---