What an HPLC Purity Percentage Actually Measures
The same vial of material can be reported at 97% or 99.5% depending on the wavelength, the gradient and how long the run was left going. HPLC purity is a real measurement — but the method determines the number nearly as much as the material does, which is why a percentage published without its method is a claim rather than a result.
What the instrument is actually doing
Reversed-phase HPLC separates a mixture by how strongly each component sticks to an oily surface. Everything about how to read the output follows from that one mechanism.
The column
A steel tube packed with silica particles coated in long hydrocarbon chains, usually eighteen carbons — hence C18. This is the oily surface. Hydrophobic molecules cling to it; polar ones do not.
The gradient
Solvent is pumped through, starting mostly water and shifting progressively toward acetonitrile. As it gets more organic, it peels compounds off the column in order of how tightly they were held.
The detector
A UV lamp downstream measures absorbance as each compound elutes. The trace it draws — absorbance against time — is the chromatogram, and every peak is something that came off the column at that moment.
Why this matters for reading the number
Separation is by hydrophobicity, not by identity. Two different molecules that happen to stick to the column equally well come off at the same time and appear as one peak. The instrument has no way to tell that this has happened.
That single fact is behind most of what follows.
How a percentage comes out of a trace
Software integrates the area under every peak, adds them up, and divides. That is the whole calculation, and its assumptions are where the interesting parts live.
Three assumptions buried in that division
That every species absorbs equally. Area normalization treats one unit of area as one unit of substance. At 214 nm this is roughly reasonable for peptides, because absorbance tracks the number of peptide bonds — but a deletion sequence has fewer bonds, absorbs slightly less, and is therefore slightly under-reported.
That everything present has eluted. Anything still on the column when the run stops contributes nothing to the denominator.
That every peak is one thing. The shoulder in the trace above is a second species riding inside the target peak. Integration counts it as target, and the percentage goes up rather than down.
Four choices that move the number without touching the material
None of these is manipulation. They are ordinary method parameters, and every one of them has a defensible reason to be set the way it is. That is precisely why the method has to be published alongside the result.
Detection wavelength
At 214 nm the peptide bond absorbs, so every peptide species shows up. At 280 nm only tryptophan and tyrosine absorb meaningfully — so a peptide containing them looks clean while impurities lacking them are simply invisible.
Gradient steepness
A shallow gradient over thirty minutes pulls closely related species apart. A steep one over five minutes pushes them together, and anything that fails to resolve gets integrated into the main peak.
Run length
Strongly retained impurities elute late. End the run early and they never appear in the denominator at all — the trace looks clean because the instrument stopped watching, not because nothing was there.
Injection load
Overload the column and the target peak broadens and tails, burying small neighbors underneath it. Inject too little and genuine trace impurities fall below the detection limit and are reported as absent.
| Method as stated | What it resolves | Reported result |
|---|---|---|
| 214 nm, 30 min shallow gradient | Deletion sequences, oxidized variants, late eluters | 97.4% |
| 214 nm, 8 min fast gradient | Gross impurities only; near neighbors co-elute | 99.1% |
| 280 nm, 30 min gradient | Only species containing Trp or Tyr | 99.6% |
| 214 nm, run stopped at 12 min | Everything except what elutes after 12 minutes | 99.4% |
Illustrative figures for one hypothetical sample, shown to make the direction and rough scale of each effect clear. The point is not the exact values — it is that the lowest number in that table is the most informative one, and the highest is the least.
The counter-intuitive consequence
A supplier reporting 97.4% with a fully stated thirty-minute method at 214 nm has very likely given you better material, and certainly better information, than one reporting 99.6% with no method at all. A high number obtained by not looking is not a high number.
This is why comparing purity claims between vendors is close to meaningless unless both publish their methods — and most publish neither the method nor the trace.
What the chromatogram carries that the number throws away
Integration compresses a whole trace into one figure. Four things visible in the picture disappear in that compression, and each of them tells you something the percentage cannot.
Peak symmetry
A clean peak is near-symmetrical. Tailing suggests unwanted interactions with the column or a partly degraded stationary phase; fronting often means overload. Neither changes the percentage much, and both say something about how reliable it is.
Shoulders
A bulge on the flank of the main peak is a second species that nearly separated. It is counted as target, so it raises the reported purity. On the trace it is obvious; in the number it is invisible.
Baseline quality
A drifting or noisy baseline makes small peaks hard to distinguish from nothing, which pushes borderline impurities below the integration threshold and out of the denominator.
Where the impurities sit
Impurities eluting just before or after the target are usually closely related — deletions, oxidations. Ones far away are typically unrelated. The percentage treats both identically; their positions tell you what went wrong in synthesis.
The ceiling nobody mentions
Purity is bounded by what the method can detect. A figure of 99.9% frequently means “nothing else rose above the detection limit,” not “nothing else is present.” Below the limit of detection, material is not measured as zero — it is not measured at all.
This is also why purity quoted to two decimal places tends to overstate its own precision. The repeatability of the assay is generally wider than the last digit implies.
Four questions to put to any percentage
In this order. Each one makes the next worth asking, and a claim that fails the first does not need the other three.
At what wavelength?
214 nm sees every peptide species. 280 nm sees only aromatics. If the wavelength is not stated, the figure could have been produced by a detector that was blind to most of what you care about.
Over what gradient and run time?
A short, steep method resolves less and reports higher. The gradient and total run time say how hard the analysis looked before concluding the material was clean.
Can I see the trace?
The chromatogram shows shoulders, tailing and baseline that the number cannot. It is the underlying evidence, and it either exists or it does not.
What else was measured?
Purity is one dimension. Identity comes from mass spectrometry, and how much peptide is in the vial comes from a separate assay again. Purity alone answers a third of the question.
Three things a purity figure will never tell you
| Question | Answered by HPLC purity? | What answers it |
|---|---|---|
| Is this the right molecule? | No — a pure something is still a something | Mass spectrometry, observed against theoretical mass |
| How much peptide is in the vial? | No — salt and water do not absorb | Amino acid analysis. See purity vs net peptide content |
| Is it free of endotoxin or heavy metals? | No — different assays entirely | LAL testing; ICP-MS |
| How clean is the synthesis? | Yes — this is what it measures | HPLC, with its method stated |
Purity and identity are a pair
A supplier reporting one without the other has answered half the question. Chromatographic purity establishes that a single species dominates; mass spectrometry establishes which species that is. Neither substitutes for the other, and a certificate carrying only one leaves the more important half open. How both appear on a certificate is covered in how to read a peptide COA.
HPLC purity
Why do two labs report different purities for the same material?
Because they almost certainly used different methods. Detection wavelength, gradient steepness, run length and injection load all change the result for identical material, and none of those is standardized across the industry. Two figures produced under different methods are not comparable, which is why the method belongs beside the number on any certificate.
Is 99.9% purity better than 97%?
Not necessarily, and sometimes the reverse. A 97% figure from a thirty-minute gradient at 214 nm reflects a method that looked hard and found things. A 99.9% figure from a short run at 280 nm may reflect a method that could not see most of what was there. Without both methods you are comparing the analyses, not the materials.
What is area percent, and why not mass percent?
Area percent is each peak’s integrated area as a share of all integrated area. It is used because the detector measures absorbance rather than mass, and converting to mass would require a response factor for every species present — including the unknown ones. Area percent is a good approximation for peptides at 214 nm, where absorbance scales with the number of peptide bonds, but it is an approximation.
What does a shoulder on the main peak mean?
A second species eluting so close to the target that the two were not fully separated. Integration typically counts it inside the main peak, so it raises the reported purity rather than lowering it. Shoulders are visible on a chromatogram and invisible in a percentage, which is one of the better arguments for asking to see the trace.
Can HPLC confirm that a compound is what the label says?
No. HPLC separates by hydrophobicity, so it establishes that one species dominates the sample without establishing which species it is. Identity requires mass spectrometry, comparing the observed mass against the theoretical mass calculated from the sequence. A sample can be highly pure and not be the compound named on the vial.
Why is TFA in the mobile phase, and does it affect the result?
Trifluoroacetic acid is added in small amounts as an ion-pairing agent. It sharpens peaks and improves resolution, which is much of why high chromatographic purity is achievable at all. It does not distort the purity figure, but it is the source of the counterion that shows up later in the mass balance of the finished powder.
Batch-documented, domestically shipped
NuMe supplies HPLC-verified, batch-documented research compounds to Canadian laboratories and institutions, shipped domestically so material spends less time in transit.
Explore the CatalogueDisclaimer: This page is provided for educational and informational purposes only. It does not constitute medical advice, and it contains no guidance on the use of any compound in people. All NuMe products are supplied strictly for laboratory research and in-vitro testing. They are not approved for human consumption, veterinary use, or any therapeutic application. Peptides are generally regulated as prescription drugs in Canada. Individuals seeking peptide therapy should consult a licensed healthcare practitioner.
