What a 99% purity number on a peptide COA actually measures

Peptide HPLC purity is the most quoted line on a certificate and the one most often misread. This article covers how the figure is calculated, what falls outside it, which impurities it is designed to catch and which companion tests close the gaps. For the other fields on the page, see reading a certificate of analysis line by line.

How peptide HPLC purity is calculated

In reversed-phase HPLC, the sample runs through a C18 column while the mobile phase shifts from mostly water toward more acetonitrile. Molecules leave the column at different times depending on how strongly they bind, and a UV detector records absorbance as they pass. Peptide bonds absorb strongly in the far ultraviolet, so peptide detection generally runs at 210 to 220 nm. Many methods read at 214 nm, inside that band.

Software draws a baseline under each peak and integrates its area. The purity figure is the main peak’s area divided by the total area of all integrated peaks, times 100. The numbers below are illustrative only and describe no real batch.

Peak	Area (illustrative units)	Area percent
Impurity A	40	0.40
Main peak	9,900	99.00
Impurity B	35	0.35
Impurity C	25	0.25
All integrated peaks	10,000	100.00

The calculation assumes every species gives a comparable signal, which is only approximately true. At 214 nm, tryptophan absorbs about 30 times more than a peptide bond, and phenylalanine, tyrosine and histidine about six times more. An impurity missing a tryptophan absorbs less per molecule than the full peptide, so its peak understates its amount.

What the purity number cannot see

A chromatogram counts only what absorbs at the chosen wavelength, elutes inside the run and separates from its neighbors. Three kinds of material fail one or more of those tests.

• Co-eluting peptides. An impurity that leaves the column with the main peptide adds to the main peak’s area. In one LC-HRMS study of a synthetic peptide, UV purity at 214 nm was 94.7% while the mass spectrometry figure was 74.8%, because several peptides sat under the main UV peak.
• Water. Residual moisture in a lyophilized powder has no UV peak and needs a separate test, such as Karl Fischer titration.
• Counterions and salts. Synthetic peptides are isolated as salts. The counterion and any inorganic ions add weight but sit outside the purity calculation.

A mass balance study of synthetic glucagon measured trifluoroacetic acid at about 103 mg per gram and water at about 50 mg per gram, and placed the glucagon content at about 896 mg per gram once every impurity was subtracted. In a 2025 study of trifluoroacetate counterions, the counterion reached up to 35% of total weight in some peptide salts, and exchanging it for chloride had no impact on measured purity.

Impurities the method is designed to catch

Most synthetic peptides today are made by solid phase peptide synthesis, where any coupling or deprotection step can fall short. A review of related impurities in peptide medicines sorts the results into synthesis-related and degradation-related families. RP-HPLC is meant to separate these from the target.

• Deletion and insertion sequences. Chains missing a residue after inefficient Fmoc deprotection, or carrying an extra one from excess amino acid reagent. In the LC-HRMS study above, several co-eluted with the main peak.
• Incomplete deprotection. Side chain protecting groups that were never removed stay attached, such as leftover Fmoc or t-butyl groups.
• Racemization. Fmoc deprotection can convert a residue from the L to the D form, giving a diastereomer. The HPLC chapter cited earlier shows L and D peptide pairs that separated at one column temperature and co-eluted at higher and lower ones.
• Oxidation. Oxidized side chains appear in the review and among the co-eluting impurities in the LC-HRMS study.
• Deamidation. Asparagine residues deamidate without any enzyme through a cyclic succinimide intermediate, leaving aspartic acid and isoaspartic acid in about a 1 to 3 ratio. The review lists succinimide formation among peptide degradation routes.

A certificate describes a batch on its test date. Deamidation proceeds spontaneously, so conditions after release matter too; see cold chain in transit and the first hour after a box arrives.

Why identity needs mass spectrometry

A UV peak carries two facts: when it eluted and how much light it absorbed. Neither names the molecule. A wrong sequence or a different peptide altogether can produce one clean peak and a high area percent.

Mass spectrometry measures the molecular mass of what is in the peak and compares it with the mass calculated from the intended sequence. Tandem mass spectrometry fragments the peptide as well; in the LC-HRMS study, fragment ions confirmed the sequence and both terminal modifications.

The FDA guidance on validating analytical procedures puts the principle generally: a procedure that lacks specificity should be compensated by other procedures, and where one does not discriminate enough, a combination of two or more is recommended. HPLC for purity with mass spectrometry for identity is that combination.

Mass spectrometry has its own blind spot. An L residue and a D residue weigh the same, so a diastereomer passes a mass check. That part of the job falls to chromatographic separation.

Method details that belong next to the number

A purity value without its method cannot be compared with any other. Change the column, gradient, temperature or wavelength and the same powder can give a different figure, because separated peaks merge or merged peaks come apart.

• Column and mobile phase. The stationary phase, such as C18, the solvents and the gradient program. Trifluoroacetic acid is widely used as the ion-pairing reagent, one reason peptides end up as trifluoroacetate salts.
• Wavelength. A far UV setting such as 214 nm sees every peptide bond. A reading at 250 to 290 nm depends on aromatic side chains, which not every peptide carries.
• Column temperature. Temperature shifts selectivity, as the L and D example shows.
• Integration settings. Where baselines were drawn, the smallest peak counted and the length of the run. Dropping small peaks from the total raises the result.
• Resolution and the chromatogram. The validation guidance names resolution of the two closest eluting components as one way to show specificity. An attached chromatogram lets a reader see it.

What to check on a certificate

A useful certificate lets a reader rebuild the claim from what is printed.

• The testing laboratory named, with its own signed report
• A batch number that matches the vial
• The test date
• The method stated: column, gradient and wavelength
• A chromatogram with integrated peaks, not only a number
• Identity by mass spectrometry, with observed and expected mass
• Water and counterion results wherever net peptide content is claimed

Imperial Biolab batches are tested by an accredited third party laboratory in the United States, by RP-HPLC on a C18 column with UV detection at 214 nm and identity by LC-MS. The batch number on each vial, and a QR code on the label, lead to that exact batch’s certificate, with the laboratory’s own signed report. Certificates stay in the certificate library after a batch sells out. Questions about a certificate can go through the contact page.

Frequently asked questions