LOD and LOQ in HPLC: What Trace Peptide Impurity Limits Mean
When a lab signs off on a large research order, the purity percentage gets all the attention. The quieter numbers behind it, the LOD and LOQ in HPLC analysis, decide which trace impurities could ever have been counted in the first place. A chromatogram cannot report a peak it cannot see, and it cannot put a trustworthy number on a peak it can barely see. For QC staff and procurement leads comparing lots across many vials, understanding those two limits (and a third, the reporting threshold) turns a single headline figure into something you can actually question.
Why absence and invisibility look the same on paper
Picture two vials, each with a certificate showing 99.1%. In the first, the remaining 0.9% is a handful of well-resolved minor peaks. In the second, the same 0.9% is reported, but a scattering of tiny species sat below the level the method was set up to notice, and they never entered the sum. Both documents look identical. Nothing on either page is false.
The difference lives in method parameters that rarely reach a customer-facing certificate. Three of them matter most:
- Limit of detection (LOD): the lowest amount of a substance the method can distinguish from background noise with reasonable confidence.
- Limit of quantitation (LOQ): the lowest amount the method can measure with acceptable precision and accuracy.
- Reporting threshold: a level chosen by the laboratory below which peaks are ignored when area percentages are calculated.
The first two are properties of the method and the instrument. The third is a policy decision. Mixing them up is where most misreadings begin.
LOD and LOQ in HPLC: how the limits are set
The most common working definition uses signal-to-noise. A peak roughly three times taller than the surrounding baseline noise is taken as detectable. A peak roughly ten times taller is taken as quantifiable. Validation guidance also allows an equivalent statistical route: LOD estimated as 3.3 times the standard deviation of the response divided by the calibration slope, and LOQ as 10 times the same ratio.
Whichever route is used, the result creates three zones on any chromatogram:
| Zone | What the analyst can say | What appears on a report |
|---|---|---|
| Below LOD | Nothing; the signal cannot be separated from noise | Usually “not detected”, if anything |
| Between LOD and LOQ | A species is present, but its amount is not reliable | Often “detected, below LOQ” or omitted |
| Above LOQ | The species is present and its amount can be stated | An area percentage, if above the reporting threshold |
The middle band deserves attention. A peak there is real. The instrument has seen it. It simply cannot be given a number that would stand up to a repeat injection, so many reports leave it out altogether.
The reporting threshold is a choice, not a measurement
Separately from what the detector can resolve, each method specifies a cut-off below which integrated peaks are disregarded. A threshold of 0.05% and a threshold of 0.1% are both defensible. They will not give the same answer.
Consider a sample containing fifteen minor species at 0.06% each. With a 0.05% threshold, all fifteen are summed, and together they remove about 0.9% from the purity figure. With a 0.1% threshold, none of them count, and the reported purity rises by that same 0.9% without any change to the material. The calculation is correct in both cases. It is simply answering a slightly different question.
This is one of several reasons two competent laboratories can report different purities for identical material. For a buyer tracking many lots over a year, it also means a change of testing lab, or a change of method at the same lab, can shift the trend line with no change in the product. Ask for the threshold to be stated, and keep a note of it beside each lot in your records.
What moves the limits up or down
LOD and LOQ are not fixed constants of a compound. They depend on the day, the column and the settings. The main drivers are:
Detection wavelength and chromophore
At around 214 nm, absorbance is dominated by the peptide backbone, so most peptide-related impurities respond similarly to the main compound. At 280 nm, response depends on tryptophan and tyrosine content. An impurity missing those residues can fall far below its true detection limit at 280 nm while the parent peak looks strong.
Baseline behaviour
Noise sets the denominator of every signal-to-noise estimate. A tired lamp, an unstable pump or a steep gradient all lift baseline noise. Gradient runs also show a rising baseline as organic content climbs, so late-eluting regions are typically less sensitive than the middle of a run.
Amount injected
Loading more sample raises small peaks above the limits. Push too far, though, and the main peak broadens and flattens, swallowing neighbours and creating co-elution. There is a practical optimum, and it differs from sequence to sequence.
Integration settings
Peak-width and slope-sensitivity parameters in the data system decide whether a small bump is integrated at all. Two analysts with different settings can report different impurity counts from the same raw file.
Reading “not detected” with the right caveats
A result of “not detected” is shorthand for a longer statement: not seen by this method, at this wavelength, at this load, on this column, above this limit. Without the limit, the phrase cannot be checked or compared.
Some impurities also sit outside what a standard reversed-phase method can detect at any sensitivity. A diastereomer formed by racemisation during synthesis has the same mass and nearly the same retention as the parent, so it tends to hide inside the main peak. Lowering the LOD does nothing for a species the column never separates. Chiral methods exist for that purpose, and they are a separate test.
Practical questions for a multi-vial order
For a lab buying in volume, a few requests turn a purity figure into something traceable across lots:
- The detection wavelength used for the purity calculation.
- The reporting threshold applied to minor peaks.
- The chromatogram itself, so small peaks near the baseline can be seen rather than inferred.
- Whether LOD and LOQ were established for the method, and roughly where they sit.
- Confirmation that the same method and threshold were used for earlier lots you hold.
Record the answers in your inventory log alongside the lot number and receipt date. When a later lot reads half a percent higher or lower, you will be able to tell whether the material changed or the method did.
Bulk Peptides products are sent for third-party HPLC and purity testing. Certificates are published for some products, and each vial can be matched to its certificate by cap and crimp colour, which makes lot-level record keeping simpler on large receipts.
Keeping detection limits in perspective
None of this makes a purity figure unreliable. It makes it conditional. A figure produced under stated limits, at a stated wavelength, with the trace available, is a strong piece of evidence. The same figure with none of that context is a weaker one, and knowing the difference is part of routine incoming QC.
Detection limits also speak only to chemical composition. They say nothing about sterility, endotoxin or fitness for any purpose beyond analytical and in-vitro study.
Everything Bulk Peptides sells is intended strictly for laboratory research. It is not for human or animal use of any kind.
