How to Test Peptide Purity: HPLC, MS & COA

Testing peptide purity means running the material against three independent questions: what fraction of the sample is the target compound (RP-HPLC, area percent at 214 nm), whether that compound is actually the sequence claimed (mass spectrometry, observed vs. theoretical monoisotopic mass), and what makes up the remainder (water, counterion, residual solvent, closely eluting deletion sequences). For a purchased reference material, the practical answer is: obtain the lot-specific certificate of analysis with the raw chromatogram and MS spectrum attached, verify the lot number on the COA matches the vial label, and — for anything load-bearing — send a retain sample to a third-party lab for $150–$400 per assay. Area percent alone is not purity. That distinction is where most evaluations go wrong.

What "98% Purity" on a COA Actually Measures

The number almost every supplier prints is RP-HPLC area percent: the integrated area of the main peak divided by total integrated peak area, usually at 214 nm (amide bond absorbance) or 220 nm.

This measurement has a specific blind spot. It only sees what absorbs UV at that wavelength and what elutes from the column during the run. A lot that is 98.2% by area percent can easily be 70–80% peptide by mass, because the rest is:

Water — lyophilized peptides typically hold 4–12% residual moisture, higher for hygroscopic sequences left in a humid fill room.
Counterion — TFA from the purification mobile phase binds to basic residues. A peptide with 3 basic residues can carry 8–15% TFA by mass. Acetate salt forms run similar.
Residual solvent — acetonitrile, DMF, or DCM at trace levels, typically under 0.5% but occasionally much higher on a rushed lyophilization.
Non-chromophoric contaminants — salts, silica fines, scavenger residues. They pass through the detector invisibly.

The metric that answers "how much peptide is in this vial" is peptide content (also called net peptide content), determined by amino acid analysis or quantitative nitrogen determination. It costs roughly $120–$250 per lot and it is absent from the majority of COAs in this market. If a COA shows 99% purity and no peptide content figure, you know the fraction of the main peak — not the mass of compound you have.

Area percent tells you the ratio of what eluted. Peptide content tells you what you actually bought. They are different numbers and a good lot reports both.

Reading an HPLC Chromatogram Instead of the Summary Number

Ask for the chromatogram image, not the tabulated result. A PDF with a purity number and no trace is an assertion, not data. On the trace, check:

Run length. A 12-minute gradient on a 20-residue peptide is too short to resolve deletion sequences. Legitimate analytical methods for peptides above 15 residues generally run 25–40 minutes with a gradient of roughly 5–65% B. Short runs compress impurities under the main peak and inflate area percent by 1–4 points.

Baseline placement. Look at where integration starts and stops. If the baseline is drawn above the level of a shoulder on the main peak, that shoulder was excluded from the calculation. Shoulders on the leading or trailing edge are usually des-amino, deletion, or diastereomeric species — chemically closest to the target and hardest to separate.

Peak symmetry. A tailing factor above 1.8 suggests column degradation or secondary interactions, and impurities co-eluting under a tailing peak are undetectable.

Injection details. Sample injection volume, concentration, column dimensions, and detection wavelength should all appear on the trace. Their absence means the method is not reproducible by anyone else, which makes independent confirmation impossible.

Wavelength. 214 nm sees the peptide backbone. 280 nm only sees Trp, Tyr, and Cys — a COA reporting purity solely at 280 nm on a peptide lacking aromatic residues is reporting on almost nothing.

Mass Spectrometry Confirms Identity, Not Purity

HPLC says "one thing came off the column." MS says "that thing weighs what the sequence should weigh." Both are required; neither substitutes for the other.

For peptides under about 5,000 Da, ESI-MS or MALDI-TOF should place the observed monoisotopic mass within ±0.5 Da of theoretical on a standard instrument, and within ±0.1 Da on a high-resolution instrument. Larger peptides are typically judged on average mass with a tolerance near ±1 Da.

Specific mass offsets are diagnostic and worth memorizing:

+16 Da — oxidation, usually Met or Trp. Common in lots stored warm or exposed to air.
−18 Da — dehydration or Asp/Asn cyclization to a succinimide.
+42 Da — acetylation, often unintended capping.
−131 Da, −113 Da, etc. — a deletion sequence missing one residue. The mass gap identifies which one.
+2× peptide mass minus 2 Da — a disulfide-linked dimer, relevant to any Cys-containing sequence.

Note the sharp limit: MS cannot distinguish stereoisomers. A peptide containing a D-amino acid where L was specified has an identical mass and often near-identical retention time. Detecting racemization requires chiral amino acid analysis after hydrolysis, roughly $300–$600 per sample, and essentially nobody in this supply chain runs it routinely. If chirality matters to the work, it must be commissioned separately.

What Independent Verification Costs and When It Pays

Third-party analysis is the only way to convert a supplier claim into your own data. Approximate US contract-lab pricing:

RP-HPLC purity, single sample: $150–$300
LC-MS identity confirmation: $200–$400
Amino acid analysis / peptide content: $120–$250
Karl Fischer water determination: $75–$150
Endotoxin (LAL), if relevant to the assay system: $150–$350
Full characterization panel: $600–$1,200 per lot, turnaround typically 5–15 business days

Set that against the material. If a lot costs $40 and the panel costs $800, independent testing costs 20× the material. If you are running a 12-month study consuming $9,000 of one lot, $800 is under 9% of material spend and cheap relative to the cost of discovering a problem after data collection.

A middle path: test the first lot from any new supplier in full, then spot-check one lot per year or on any change in lot number, appearance, or supplier-reported method. That converts a per-lot expense into a supplier qualification expense.

When Testing Peptide Purity Is Not Worth Doing

This is the part most vendor pages omit, so name it plainly.

Skip independent verification entirely when:

The material is for methods development, column conditioning, or system suitability, where an impurity profile changes nothing about the outcome.
Total annual spend on that compound is under roughly $500. The economics never close.
Your assay has an internal readout that would flag a bad lot anyway — a binding assay with a known positive control detects a wrong compound faster and cheaper than an HPLC run.
You need results this week. A 5–15 business day contract lab turnaround kills the schedule, and rushed testing on a compressed timeline produces a number nobody has time to act on.

Independent testing will not rescue you when:

You test one vial from a heterogeneous fill. Purity is measured on the sample injected. Fill-line variation across a 5,000-vial run means one passing vial certifies one vial. Testing one unit of a lot gives you weak evidence about the lot, not proof.
The problem is stability, not synthesis. A lot genuinely 98% at release can drift materially after 18 months at 4 °C, or after two weeks in a shipping container that reached 45 °C. Release data describes the material at release. Ask what the storage condition was between synthesis date and ship date — and note that most suppliers cannot answer that with records.
The impurity is stereochemical. As above: standard panels miss it completely.
The COA is fabricated. COAs are PDFs. Reused chromatograms with edited lot numbers exist in this market. The tell is a trace whose baseline noise pattern is pixel-identical across two different lot documents. Independent testing catches this — reading the COA more carefully does not.

Who should not be buying documented reference material at all: anyone who cannot state which analytical claim actually matters to their work. If you cannot say whether you need 95% area percent, confirmed peptide content, verified counterion form, or endotoxin data, you will pay a premium for documentation you never use, and the premium is real — fully documented lots commonly run 2–4× the price of undocumented material from the same synthesis routes. Buying paperwork you don't read is a worse outcome than buying cheaper material knowingly. Our research use only and quality pages set out the scope of what documentation covers and what it does not.

Building a Supplier Qualification File

Testing one lot proves one lot. What holds up over years is a file per supplier containing:

Lot-specific COAs with attached raw traces, retained for every lot received, not just the first.
Synthesis date and analysis date. A gap over 24 months between synthesis and your receipt means the release data is stale relative to what is in the vial.
Storage temperature claims — typically −20 °C long-term for lyophilized material, with documented excursion tolerance — and whether shipping was temperature-monitored.
Any third-party results you commissioned, filed alongside the supplier's own numbers for the same lot. The delta between them is the single most useful metric you will ever collect on a vendor. A supplier reporting 99.1% where your lab finds 96.4% is showing you a systematic 2–3 point optimism you can then price in.
Retain samples: keep 5–10% of each lot unopened at −20 °C so a question raised 14 months later is answerable.

Comparative write-ups in our library cover how specific suppliers document lots and where the documentation typically thins out.

The Short Version

Ask for the chromatogram and the MS spectrum with lot numbers matching the vial. Check the run length, the baseline, and the detection wavelength. Look for peptide content as a separate number from area percent. Accept that MS confirms mass and not stereochemistry. Test the first lot from any new supplier independently at $600–$1,200, then qualify by exception rather than by every lot. And be honest about the cases where none of this earns its cost — under $500 a year of material, or an assay whose own controls would catch the problem first.

Supplied strictly for in-vitro laboratory research by qualified researchers and institutions. Not a drug, food, cosmetic or dietary supplement. Not for human or veterinary consumption. Not evaluated by the FDA.