What 98% purity on a peptide COA actually means
"98% pure by HPLC" is the most-quoted number on a peptide Certificate of Analysis. It sounds precise. In practice it's the result of an analyst drawing peaks on a chromatogram, with method choices that can swing the number up or down a few percent. Understanding how the number is produced is the difference between trusting the label and trusting the lab.
How the number is generated
The analyst injects a diluted peptide sample onto an HPLC column. The column separates components by hydrophobicity — target peptide elutes at one time, structurally-similar byproducts at other times. A UV detector at a specific wavelength (commonly 214 or 280 nm) records absorbance as each component passes through. The output is a chromatogram: a graph of detector signal vs time, with peaks where compounds eluted.
Purity is calculated as:
purity % = (target peak area ÷ total peak area) × 100
The "total peak area" is the sum of the target peak plus every impurity peak in the chromatogram that the analyst chose to integrate. Two decisions affect the number: (1) where the analyst draws the baseline under each peak, and (2) which peaks are included as "real" vs dismissed as noise.
What the 2% usually is
For a well-synthesized peptide at 98% purity, the remaining 2% is almost always a mixture of closely-related synthesis byproducts — the molecular-weight neighbors of the target peptide. Common categories:
| Impurity category | What it is | Where it comes from |
|---|---|---|
| Deletion sequences | Peptide with one amino acid missing | Incomplete coupling during solid-phase synthesis |
| Truncation products | Peptide with residues missing from one end | Premature termination during synthesis |
| Deamidation | Asn or Gln converted to Asp/Glu | Storage or synthesis pH/temperature |
| Oxidation | Met or Cys oxidized | Air exposure or oxidative reagents |
| Protecting group remnants | Partially-deprotected peptide | Incomplete final deprotection step |
| Aggregation products | Dimers, trimers | Storage, lyophilization, concentration |
None of these are "the wrong peptide" in a meaningful sense — they are the target peptide with small structural defects. Whether they matter biologically depends on the defect: a deamidated position far from the active site is often silent; a truncation missing the receptor-binding residues is essentially inactive peptide mass that dilutes your dose.
Why 98% isn't always 98%
Three variables make the same number mean different things across COAs:
1. The HPLC method
Different columns (C18 vs C8, different pore sizes), different mobile phases (acid-ammonium acetate vs TFA), different gradient profiles, and different detection wavelengths all change which impurities show up and how well they separate from the target. A shallow gradient with high resolution may show more impurity peaks (lower reported purity) than a fast gradient that co-elutes several small peaks into the main target peak (higher reported purity). A well-documented COA specifies the method; a one-line "98% HPLC" entry does not.
2. Integration choices
Peak integration is not fully automated. The analyst (or the integration software with analyst supervision) decides where each peak begins and ends, and which small bumps are real peaks vs baseline noise. Aggressive baseline-cutting raises the apparent purity number; conservative integration lowers it. A well-drawn chromatogram attached to the COA is the artifact you can audit — a number in a table without the chromatogram, you cannot.
3. Identity confirmation
HPLC purity says "98% of signal came from one compound." Mass spectrometry says "that 98% compound has the molecular weight of the target peptide." Without MS confirmation, a 98% purity result could in principle be 98% of some other peptide that happens to elute at the expected time. LC-MS on the same sample closes this gap. See our COA guide on mass-spec identity.
What thresholds actually matter
| Purity | What it means | Grade |
|---|---|---|
| < 90% | High impurity load; unpredictable biology | Not acceptable for research injection |
| 90–95% | Acceptable for some research; higher byproduct biology concerns | Low-grade research |
| 95–98% | Standard research-grade | Typical |
| 98–99% | High-quality research-grade | Premium research |
| ≥ 99.5% | Pharmacopoeia-level purity | Pharmaceutical-grade |
FAQ
What exactly is the 2% that is not the target peptide?
Almost always structurally-related synthesis byproducts — deletion sequences, truncations, deamidation, oxidation, protecting-group remnants. Rarely random contamination.
Is 98% pure peptide the same across vendors?
No. Different HPLC methods, integration choices, and identity confirmation protocols mean the same number can represent different quality levels.
Does higher purity always mean better results?
Up to a point. 90% → 98% is meaningful. 98% → 99.5% matters less for research but matters for injectable pharmaceutical-grade product.
HPLC vs HPLC-MS?
HPLC alone measures signal proportions from a UV detector. HPLC-MS (LC-MS) identifies each peak by molecular weight, so the purity claim is tied to a confirmed identity.
How do I verify a purity number isn't fabricated?
Look for attached chromatogram, batch number matching the vial, analyst signature, lab name with third-party attribution, documented method, and batch-to-batch consistency.
Related reading
Batch tracking that actually tracks batches
Peptide Protocol logs vendor, batch number, and COA purity for every vial. If effects differ across batches, you'll see it.
Get the iPhone app →Informational and educational only. Not medical advice. Analytical concepts in this article reflect standard pharmaceutical HPLC practice; specific COAs should be evaluated on their own merits.