Peak Area, Retention Time, And What A Chromatogram Cannot Show
A chromatogram is a plot of detector signal against time. Retention time is where a compound came off the column, peak area is how much signal it produced while doing so, and area percent is one peak's share of the total integrated signal in that run. Each of those three quantities is narrower than it looks, and the gap between what they measure and what they are read as is where most misreadings of an HPLC chromatogram peak area happen.
Retention time is a property of the separation, not of the molecule. The same peptide on a different column chemistry, at a different gradient slope, at a different column temperature or with a different mobile phase additive will elute at a different time. Retention time is reproducible within a single method and only within it, which is why a purity record is worth little without the method that produced it.
Peak area is the integral of detector response over the width of the peak. With ultraviolet detection that response depends on the extinction coefficient of whatever is passing the detector at the wavelength in use. Two compounds present in identical amounts produce identical areas only if they happen to absorb identically, which for a peptide and its impurities is rarely true.
Area percent, the figure normally quoted as purity, divides one peak area by the summed area of all integrated peaks in the run. It is therefore a closed accounting of what eluted and absorbed, and it says nothing about anything outside that set.
The system suitability numbers that make a run readable
Before any area is trusted, a chromatogram is checked against a small set of shape and separation metrics. These are the numbers that decide whether the integration is even meaningful.
| Metric | How it is calculated | Typical requirement |
|---|---|---|
| Resolution, R | difference in retention times divided by average peak width | at least 2.0 for baseline separation |
| Tailing factor, T | peak width at 5% height divided by twice the front half-width | 0.8 to 1.5 for a symmetrical peak |
| Plate count, N | from retention time and peak width, a measure of column efficiency | method-specific, monitored for decline over column life |
| Signal-to-noise ratio | peak height divided by baseline noise amplitude | about 3 to detect, about 10 to quantify |
| Replicate area precision | relative standard deviation across repeat runs of one solution | commonly 2.0% or below for the main peak |
A tailing factor above the accepted range distorts area on its own, because the integrator has to decide where a long trailing edge stops and the baseline starts. A resolution figure below 2.0 between the main peak and its nearest neighbour means part of one peak is being counted as the other, whichever integration rule is applied.
The failure mode: an impurity hiding under the main peak
The single most consequential thing a chromatogram can get wrong is co-elution. A closely related impurity, a deamidated species, an oxidised residue or a diastereomer, can differ from the parent peptide in retention by less than the peak width. The integrator then sees one peak, reports one area, and the purity figure comes back high while the sample is not what the number implies.
It shows up as subtle asymmetry: a shoulder on the leading or trailing edge, a tailing factor that has drifted from the historical value for that method, or a peak that is measurably wider than the reference standard run alongside it. On a diode array detector it shows up as a spectrum that changes across the peak, since a pure peak has the same spectral shape at its front, apex and tail.
Confirming it takes an orthogonal separation. A second column chemistry, a shallower gradient, a different mobile phase pH, or a mass detector across the same run will separate what one gradient merged. Mass detection is the usual answer because it distinguishes on a different physical property entirely.
Four things absent from every chromatogram
- Anything that did not elute. Material retained on the column past the end of the gradient, or precipitated before the column, contributes no area and is simply invisible.
- Anything transparent at the detection wavelength. Inorganic salts and many counterions pass the detector without absorbing.
- Water. Residual moisture in a lyophilized solid is not a chromatographic peak and requires a separate determination.
- Anything microbiological. A chromatogram is a chemical separation and carries no information about bioburden, sterility or endotoxin.
The first two set the honest ceiling on area percent as a purity statement. A figure of 99.1% by area at 214 nm means 99.1% of the absorbing, eluting material in that run, and the phrase carries all of its own qualifiers.
Where our own published figures stop
Material listed by Aurum is independently assayed for purity by reverse-phase HPLC, and identity is assayed by mass spectrometry. Those two records are what the published figures rest on. The purity figure is HPLC purity and is not net peptide content, which is a different determination and is not among the specifications Aurum publishes.
Sterility, endotoxin and pyrogen testing are not performed and are not claimed. A chromatogram would not support such a claim in any case, and the point of saying it beside the purity figure is that the two are often read as though one implied the other.
Common questions
Does a bigger peak mean more material?
Within one compound and one method, yes, and the relationship is linear across the validated range. Between two different compounds, no, because the detector response factors differ.
Is peak height as good as peak area?
Height is less sensitive to baseline choice and more sensitive to peak shape. Area is the convention for purity because it is preserved when a peak broadens, while height falls.
Why did retention time shift between two runs of the same method?
Column temperature, mobile phase composition drift, column age and system dwell volume all move retention. A small consistent shift across a sequence of runs usually points at the column or the temperature control rather than at the sample.
What does a negative peak mean?
Usually that the sample solvent absorbs less than the mobile phase at that wavelength, so the baseline dips as the sample band passes. It is a solvent artefact rather than a compound.
Can two peaks be the same compound?
Yes. Slowly interconverting conformers, proline cis and trans isomers, and partially resolved diastereomers all produce more than one peak from one nominal sequence. Identity by mass across both peaks is how that case is recognised.
References
- 01United States Pharmacopeia General Chapter <621> Chromatography. USP–NF.
- 02United States Pharmacopeia General Chapter <1225> Validation of Compendial Procedures. USP–NF.
- 03European Directorate for the Quality of Medicines General Chapter 2.2.46: Chromatographic separation techniques. European Pharmacopoeia.
- 04International Council for Harmonisation Q2(R2) Validation of Analytical Procedures. ICH, 2023.
- 05International Council for Harmonisation Q6B Specifications: Test Procedures and Acceptance Criteria for Biotechnological/Biological Products. ICH, 1999.
- 06Snyder LR, Kirkland JJ, Dolan JW Introduction to Modern Liquid Chromatography. Wiley, 3rd edition, 2010.
Every citation links out to the paper on PubMed. Identifiers are omitted deliberately rather than reproduced from memory, so where we do not hold a verified PMID or DOI the link is a PubMed search for that exact title — it resolves to the paper without anything being invented.
FOR RESEARCH USE ONLY · NOT INTENDED FOR HUMAN CONSUMPTION. This article describes compounds and the research literature in which they appear. Nothing here is a recommendation, protocol, or statement of effect.