Reading a COA: identity, purity and batch

A Certificate of Analysis is, at its core, a laboratory report. It documents the results of analytical tests performed on a sample taken from a specific production batch. In the peptide research context, it is the primary — and often the only — objective evidence available about the quality of what is in a vial. Reading one well is a learnable skill. It requires knowing what three things to look for, in what order, and what each one actually proves. This article covers all three: batch, identity, and purity.

Why the batch number comes first

Before examining any analytical result on a COA, the first question is strictly logistical: does the lot number on the certificate match the lot number on the product?

This matters because a COA is issued for a specific production run, not for a compound in general. Two batches of the same peptide from the same supplier can have meaningfully different analytical profiles — different purity levels, different impurity patterns, potentially different identity results. A certificate issued for batch A tells you nothing verifiable about batch B, even if the compound is nominally the same.

The consequence is straightforward: if a supplier provides only a generic COA — one not linked to a specific lot number, or with a lot number that does not match the product label — that certificate has no evidential value for the batch you are receiving. A clean HPLC result on the wrong batch proves nothing about your batch.

Identity: what LC-MS actually measures

Liquid chromatography–mass spectrometry (LC-MS) is a hyphenated technique: it first separates the components of a sample using liquid chromatography, then passes each separated component into a mass spectrometer that measures molecular mass with high precision. The output relevant to identity is the measured molecular weight of the compound detected in the sample.

For identity confirmation, this measured mass is compared against the theoretical molecular weight of the declared peptide. If the values match within an acceptable analytical tolerance — typically a few parts per million (ppm) or a fraction of a dalton, depending on the instrument — that constitutes confirmation that the correct molecule is present.

This is why LC-MS is so powerful for identity: it is not a visual test, it is not a colour reaction, and it cannot be fooled by adding a label to a vial. The mass spectrometer measures what is actually there at the molecular level. A peptide mislabelled as a different compound, or a different molecule substituted entirely, will produce a mass that does not match the theoretical value of the declared substance.

Reference theoretical molecular weights for common GLP-1 analogues (research context)

Compound	CAS number	Molecular formula	Theoretical MW
Semaglutide	910463-68-2	C₁₈₇H₂₉₁N₄₅O₅₉	4113.58 g/mol
Tirzepatide	2023788-19-2	C₂₂₅H₃₄₈N₄₈O₆₈	4813.45 g/mol
Retatrutide	2381089-83-2	C₂₂₁H₃₄₂N₄₆O₆₈	4731.33 g/mol

If a COA declares retatrutide but reports an LC-MS mass that does not approach 4731.33 g/mol, the certificate contains an internal contradiction that invalidates it. The compounds above are complex, high-molecular-weight peptides; counterfeiting one with a different molecule that happens to produce the same LC-MS mass would require extraordinary effort. In practice, a mass mismatch means something has gone wrong.

What LC-MS does not prove

Identity confirmation by LC-MS is necessary but not sufficient on its own. It confirms the correct molecule is present at detectable levels, but it does not tell you:

• How much of the sample is the target compound versus impurities (that is purity, measured by HPLC)
• Whether the batch is sterile or endotoxin-free
• Whether the stated quantity (e.g. 5 mg per vial) is accurate

A complete COA requires both LC-MS and HPLC. Identity without purity, or purity without identity, is an incomplete picture.

Purity: how HPLC produces the percentage

High-performance liquid chromatography (HPLC) works by pumping a sample through a column packed with a stationary phase. Different components of the sample travel through the column at different speeds, based on their affinity for the stationary phase and the mobile phase (the solvent carrying them). They exit the column at different times — their retention times — and a detector (commonly a UV detector) registers each one as a peak on a chromatogram.

The purity figure is calculated from the area under the peaks: the proportion of the total peak area that belongs to the main peptide peak is the purity percentage. A peptide with 99% HPLC purity means that 99% of the detected signal comes from the target compound, with the remaining 1% distributed across impurity peaks.

The benchmark: ≥98%

The widely accepted quality standard for research peptides is a purity of ≥98% as measured by HPLC. This figure is not arbitrary: at this level of purity, the impurity load is sufficiently low that the analytical character of the sample is dominated by the target compound. Below 98%, the impurity contribution becomes more significant and raises questions about the synthesis and purification process.

Interpreting HPLC purity figures

Purity (HPLC)	Interpretation
≥ 99%	High quality. Expected from a well-synthesised, well-purified batch.
98–99%	Within the accepted standard for research-grade peptides.
95–98%	Below the benchmark. The impurity profile warrants further scrutiny.
< 95% or unstated	Significant red flag. Inadequate purity or an incomplete certificate.

Reading the chromatogram alongside the percentage

A purity percentage without a supporting chromatogram is, in analytical terms, an unsupported assertion. The percentage is derived from the chromatogram; the chromatogram is the underlying evidence. Things to look for:

• One dominant, well-defined main peak. The target peptide should account for the overwhelming majority of the peak area. A chromatogram with several peaks of similar height would not support a 98%+ purity claim.
• Internal consistency. If the declared purity is 99.2% but the chromatogram visually shows prominent secondary peaks, something does not add up.
• Baseline resolution. Peaks should be clearly separated. A broad, unresolved hump in the baseline can indicate a mixture of components that have not been properly separated — a sign of a lower-quality HPLC method.

The independent laboratory requirement

The third structural requirement of a credible COA is that it is issued by a laboratory that is independent of the supplier. This is a prerequisite, not an optional extra. A COA produced by the same entity that is selling the product has no independent evidential value: the seller is, in effect, marking their own work.

What constitutes an independent laboratory? At minimum:

• A name and address that can be independently identified
• No commercial relationship with the supplier that would compromise objectivity
• Ideally, accreditation to ISO/IEC 17025 (the international standard for testing and calibration laboratories)
• A means of cross-referencing the certificate — e.g. a certificate number, a QR code, or a portal login that allows verification

Accreditation bodies in Ireland and the EU publish lists of accredited laboratories. If a COA names a laboratory, it is reasonable to check whether that laboratory is genuinely accredited and identifiable.

Limitations of COA-based verification

Even a complete, well-formed COA from an independent laboratory has limitations that an informed reader should understand:

• The sample analysed may not be identical to the product supplied. The COA is based on a sample taken from the batch; it cannot guarantee every unit in that batch is identical to the sample.
• Sterility and endotoxin levels are separate tests. A COA showing high HPLC purity and correct LC-MS identity says nothing about bacterial endotoxin content, sterility, or fitness for any particular use. These require separate, dedicated testing.
• HPLC methods vary. Not all HPLC methods are equally sensitive. A method that does not detect certain classes of impurity will report a higher purity than a more comprehensive method would.
• A clean COA is not authorisation. Regulatory authorisation (from the HPRA, for instance) is a separate matter from analytical quality. Understanding the distinction is explained further in our article on the HPRA framework and peptide research in Ireland.

These limitations do not diminish the value of COA verification; they contextualise it. A verified COA is the best available analytical evidence about a batch. It is not the only piece of information that matters, and it does not replace professional medical involvement.