Accepted answer
Send the vial sealed. Then the measurement is straightforward and its uncertainty is dominated by one thing you can ask about in advance.
What the lab does
- Weigh the sealed vial, or not. Better labs record a gross mass by weighing the vial before and after removing the cake, or by difference against an empty matched vial. Many skip this, which means you get content per vial but no gross weight and therefore no peptide-content percentage.
- Reconstitute in a precisely known volume. Typically 1.00 or 2.00 mL of water or dilute acid delivered by a calibrated positive-displacement pipette or, in careful labs, gravimetrically. This is the step you must not do yourself: your pipette is not calibrated and your 2 mL is not 2.00 mL.
- Dissolve completely. Gentle inversion and a wait; acylated analogues are slower to dissolve than plain peptides. Sonication is used but can degrade some peptides, so a good lab avoids it.
- Dilute to the working range of the method — usually a serial dilution into mobile phase A to land somewhere around 0.1 to 0.5 mg/mL where the UV response is linear and not saturating.
- Inject alongside a calibration series of the reference standard, prepared from a standard of known assigned purity, at three to five concentrations bracketing the sample.
- Compute concentration from the regression, then multiply back up through the dilution factor and the reconstitution volume to get mg per vial.
The arithmetic, worked
Say the vial is reconstituted in 2.00 mL, diluted 1 in 10 into mobile phase, and the diluted sample gives a main-peak area of 1,842,600 counts. The standard curve, built from a standard with assigned purity 96.4%, is area = 5,268,000 x C with C in mg/mL and an r-squared of 0.9998.
- Diluted concentration:
1,842,600 / 5,268,000 = 0.3498 mg/mL
- Undiluted:
0.3498 x 10 = 3.498 mg/mL
- Per vial:
3.498 x 2.00 = 6.996 mg, reported as 7.00 mg
Note what the standard's assigned purity did: it is already baked into the slope, because the calibration solutions were prepared by weighing standard powder and correcting the nominal concentration by 96.4%. If that 96.4% is wrong, every number above is wrong by the same proportion.
Realistic uncertainty budget
| Contribution | Typical relative uncertainty | Notes |
|---|
| Reference standard assigned value | 1 to 5% | Under 1% for a certified reference material; up to 5% or worse for an unqualified in-house standard |
| Volumetric reconstitution and dilution | 0.5 to 1.5% | Calibrated glassware and pipettes |
| Injection and detector repeatability | 0.3 to 1.0% | Measured as %RSD across replicate injections; system suitability should cap it at 2% |
| Incomplete recovery from the vial | 0 to 3% | Adsorption, undissolved cake, filter losses; the sloppiest term |
| Peak integration | 0.2 to 1% | Larger if impurities are unresolved from the main peak |
Combining these in quadrature with a good CRM gives roughly 2 to 3% relative — so 7.00 mg plus or minus about 0.2. With a poorly assigned in-house standard, 6 to 8% relative is realistic and the number becomes 7.0 plus or minus 0.5.
To answer your actual question directly: at 2 to 3% uncertainty, a single measurement showing 20% below label is unambiguous. It is seven to ten standard deviations out. You do not need duplicates to be confident the shortfall is real. What you might want duplicates for is to establish that it is characteristic of the batch rather than one bad fill, and for that the right design is two or three vials from the same lot rather than two runs on one vial.
What to ask before you pay
- Which reference standard, from where, with what assigned value and what traceability?
- Will you report gross weight as well as content?
- What are the system suitability results for the run my sample was in?
- Is the reported figure content of target peptide, or net peptide before the purity correction?
Janoshik and Medutest both answer the first and last of those on request; the standard question is the one that separates a 3% measurement from a 7% one. PeptideMeter's published reports include the method conditions, which lets you at least see whether the separation was good enough to integrate cleanly.
edited 14 Jun 2026 by tenth_of_a_unit — fixed an arithmetic slip in the third paragraph
3The "seven to ten standard deviations" point is the answer to every "but could it be measurement error" objection. – loss_on_drying 4 months ago 4Reporting net peptide rather than target content is a real and common ambiguity. Always ask. – stopper_core 6 months ago add a comment