Measurement uncertainty on an HPLC purity result
What the ± figure on an HPLC purity result means, where it comes from, how it meets a 99% limit, and what a greater-than result tells you.
The "±" figure after an HPLC purity result is the laboratory's measurement uncertainty: the range around the reported value where the true value is expected to lie, usually at about 95% confidence. Alphex GHK-Cu lot 0010, for example, is reported as ">99.80% ± 0.18%". This note explains how figures like these are generally reported and read. A certificate may not spell out every convention behind its numbers, so confirm the exact meaning of any specific result with the laboratory that issued it. Uncertainty matters most when a result sits close to a specification limit, and hardly at all when the result is far from it. This note explains where uncertainty comes from, how to read it, how it interacts with a 99% threshold, and what the ">" sign adds. All Alphex products are for laboratory and in-vitro research only.
What measurement uncertainty is
Every measurement has some spread. Run the same sample twice on the same instrument and the purity figure will differ slightly. Run it on another day, with a new mobile phase batch or a different column, and it will differ a little more. Measurement uncertainty puts a number on that spread. ISO/IEC 17025 requires laboratories to evaluate it. The standard reference for how to do so in analytical chemistry is the Eurachem/CITAC guide Quantifying Uncertainty in Analytical Measurement, now in its third edition.
Most laboratories report an expanded uncertainty. They estimate a standard uncertainty from all the contributing sources, then multiply it by a coverage factor, usually k = 2, which gives a range expected to contain the true value with about 95% confidence. A certificate should say which coverage factor it uses. If it does not, ask the laboratory.
Where the uncertainty in HPLC purity comes from
| Source | Why it adds spread |
|---|---|
| Injection repeatability | Small differences in injected volume and detector noise change peak areas slightly from run to run |
| Integration | Where the baseline is drawn and where peaks are split changes the area assigned to each peak |
| Small impurity peaks near the threshold | Peaks near the integration threshold may or may not be counted |
| Column and mobile phase | Column age and mobile phase batch shift resolution, which changes how well impurities separate from the main peak |
| Response factors | Area percent assumes every species gives the same response per unit mass at the detection wavelength, which is only approximately true |
| Sample preparation | Weighing, dissolution and how long the solution stands before injection |
Background on the calculation itself is in how HPLC purity is calculated from peak area.
How a result like ">99.80% ± 0.18%" is generally read
A result written this way combines two separate pieces of information. The readings below are the usual conventions, not statements made by any particular certificate. If a certificate does not define them, ask the issuing laboratory.
- ">99.80%" is usually read as a result reported as greater than 99.80% rather than as an exact figure. Laboratories commonly report this way when a result lies above an upper reporting limit for the method. Read like that, it tells you the result is above 99.80%, but not by how much.
- "± 0.18%" is usually an expanded uncertainty stated in percentage points. If it is read that way and applied to the reported figure, the bottom of the range is about 99.62%. The coverage factor and whether the figure is absolute or relative are details to confirm with the laboratory.
On that usual reading, the lowest value consistent with the result and its uncertainty is still above a 99% threshold. The GHK-Cu lot 0010 certificate is in the COA library, where you can read the wording for yourself and quote its laboratory ID to the laboratory.
Decision rules: when uncertainty changes the answer
When a result is compared with a limit, the laboratory or the buyer needs a decision rule: how the uncertainty is taken into account when calling a pass or a fail. The Eurachem guide covers this as compliance assessment. Three cases show why it matters.
| Reported result | Range with ± 0.18 | Against a 99% limit |
|---|---|---|
| >99.80% | About 99.62% or higher, on the usual reading | Pass on the usual reading, whatever decision rule is used |
| 99.10% | 98.92% to 99.28% | Pass on the reported value, but the range crosses the limit. A strict rule would not call it a clear pass |
| 98.70% | 98.52% to 98.88% | Fail. The whole range is below the limit |
| 83.48% | 83.30% to 83.66% | Clear fail, more than 15 points below the limit |
The ranges assume the ± figure is read in percentage points, as above. The last row is a real Alphex result: lot 013, published in the library as Rejected. It is published in the Alphex COA library. The middle rows are illustrations. A result in the second row is where a buyer should look closely at the chromatogram rather than at the headline figure alone. The difference between 98% and 99% material is covered in 98% versus 99+% purity.
What uncertainty does not cover
The ± figure describes the precision of the measurement as performed. It does not cover systematic limits of the method itself:
- An impurity that co-elutes under the main peak is counted as the target in every run. No amount of repeat measurement reveals it.
- Species that do not absorb at the detection wavelength, such as water, salts and many counter-ions, are not in the chromatogram at all. Purity is not the same as net peptide content. See net peptide content versus gross vial weight.
- It describes the sample tested, not the rest of the lot. See what "results relate only to the sample tested" means.
ICH Q2(R2), the current guideline on validating analytical procedures, treats specificity, precision and range as separate characteristics of a method for this reason. A precise method is not automatically a specific one.
Questions to ask about any purity result
- Is an uncertainty stated, and with what coverage factor?
- Is the result an exact figure or a "greater than" figure?
- How far is the result from the specification limit, compared with the uncertainty?
- Is the chromatogram published, so you can see whether minor peaks were integrated? See how to read a chromatogram.
Frequently asked questions
Is a smaller ± always better?
It means a more precise measurement, which is useful near a limit. It says nothing about whether the method separates all relevant impurities.
Why can the same ± figure appear on certificates for different lots?
Laboratories commonly evaluate uncertainty for a method and then apply it to each result that method produces, so the same figure can appear on many certificates. How a given laboratory does this is a question for that laboratory.
Where can I check a certificate?
In the COA library, by the code on the vial label, or from the GHK-Cu product page.
Sources and further reading
- Ellison S.L.R., Williams A. (eds). Eurachem/CITAC guide: Quantifying Uncertainty in Analytical Measurement, 3rd edition, 2012: evaluating and reporting expanded uncertainty, and assessing compliance with a limit
- EMA. ICH Q2(R2) Validation of analytical procedures, effective 14 June 2024: specificity, precision and range as validation characteristics for purity procedures
- ILAC. About ILAC: ISO/IEC 17025 accreditation of testing laboratories
Alphex research peptides are supplied for laboratory and in-vitro research by qualified researchers only. They are not authorised for human or veterinary use.
Research use only. Alphex products are supplied exclusively for laboratory and in-vitro research by qualified researchers. They are not authorised for human or veterinary use, and nothing in this note is advice on any such use.
Published 26 October 2026. Alphex reviews research notes when the testing method, the catalogue or UK guidance changes.