What an HPLC purity number proves — and what it does not

Reference page · published 2026-08-28

A purity figure on a peptide Certificate of Analysis is almost always an HPLC area percentage: the main peak's share of the total signal the detector recorded when one sample was pushed through one column on one day. It answers a narrow question honestly — how much of what the detector saw was the main peak. It does not tell you what fraction of the vial's weight is peptide, and it does not tell you the main peak is the peptide named on the label. Those are three different measurements, and most certificates report only the first.

This page collects what the published analytical-chemistry literature says about each of the three, with the primary sources linked. None of it is an accusation against any vendor. It is the measurement science, which is dull, well documented, and almost never explained to the people being handed the number.

What an area percentage is actually measuring

High-performance liquid chromatography separates a mixture in time. Components leave the column at different moments, a detector — usually ultraviolet light absorbance — records a trace, and each component shows up as a peak. Area percent is the main peak's area divided by the area of every peak, times one hundred.

That arithmetic contains an assumption: that every impurity absorbs UV light about as strongly as the main compound does. It often does not. A 2025 paper in Analytical Chemistry (PMID 40499007) is devoted to this point for peptides specifically. Its author notes that the impurities peptide synthesis actually generates — from insertion, truncation, deamidation, isomerization, oxidation, and manufacturing processes — may not have equal responses (RRF ≥ 1 or RRF ≤ 1) compared to their main analyte. RRF is the relative response factor: how loudly a given impurity shows up compared with the peptide. The paper's conclusion is blunt about the default: By applying the RRF value by default as 1, impurity estimation results may be overestimated (actual RRF > 1) or underestimated (actual RRF < 1) in peptide therapeutics.

So an area percentage is an estimate whose error bar depends on which impurities are present — and the certificate rarely says whether response factors were measured or assumed. Conflict of interest noted: the author is affiliated with Flamma USA, a contract manufacturer of peptides. The observation stands on the chemistry, but the reader should know who made it.

The number that is usually missing: how much of the vial is peptide

Most research peptides are built by solid-phase peptide synthesis, which relies on trifluoroacetic acid (TFA) both to cleave the finished chain and as an ion-pairing reagent during purification. The consequence is structural, not accidental: peptides come out of that process being obtained as TFA salts (PMID 40872554, ETH Zurich, 2025). The TFA is a counterion — a charged partner sitting alongside the peptide — and it has weight.

How much weight? When metrologists at the National Research Council Canada assigned a purity value to a candidate certified reference material for the peptide angiotensin II, they measured the TFA counterion and found it considered an impurity in this case at nearly 25% by mass. Their final certified value for that material was 691 ± 9 mg/g (coverage factor k = 2) — that is, roughly 69% peptide by weight (PMID 30143839).

That material was not defective. It was good enough to serve as a certified reference standard precisely because the number was measured and stated. The lesson is not that peptides are 69% peptide. It is that chromatographic purity and peptide content by mass are independent numbers, and a high value for one implies nothing about the other.

The ETH Zurich group demonstrated that independence directly. They swapped the TFA counterion for chloride at several acid concentrations and reported that no impact on peptide purity was observed at all HCl concentrations — the purity figure sat still while the material's mass composition changed underneath it. They also found that influences on permeability coefficients depending on peptide sequence and salt form were found, and closed by emphasising the importance of counterion quantification and specification in assays with synthetic peptides (PMID 40872554). Two vials with the same printed purity can hold different amounts of peptide and behave differently in an experiment.

The impurities hardest to see are the ones most chemically similar to the target: chains missing a residue, chains carrying an extra one, chains where a single amino acid has quietly rearranged. They travel through a column at nearly the same speed as the real thing.

How many are there? The Bureau International des Poids et Mesures, working with China's National Institute of Metrology, characterized a synthetic human C-peptide material that served as the study sample for an international comparison of laboratories' ability to assign peptide purity. Using liquid chromatography with high-resolution mass spectrometry, more than 65 structurally related peptide impurities were identified, confirmed and quantified in that one material. Their combined mass fraction was estimated at 83.3 mg/g, with an expanded uncertainty of 3.0 mg/g (k = 2) — about 8% of the material by weight (PMID 29862433).

Finding sixty-five of them required high-resolution mass spectrometry and a described calibration hierarchy. A routine area-percent run does not resolve that population; it collapses whatever fraction of it co-elutes into the main peak.

Where mass spectrometry stops

Mass spectrometry is the standard answer to “is it the right molecule,” and it does more than purity alone: it weighs what came off the column. But it weighs. It does not read sequence.

The clearest illustration is leucine and isoleucine. They are different amino acids with different behavior, and they have identical molecular masses — so identical that the literature gives the pair its own placeholder, Xle. A 2026 Analytical Chemistry paper states the problem plainly: the two can exert distinct biofunctional properties when present in the same position in a peptide sequence, yet their identical molecular masses make it challenging to distinguish them by using mass spectrometry (PMID 42423382).

Telling them apart took real work: combining two fragmentation methods, isolating fragment ions at m/z 44.0 and 69.0, and using the ratio of their intensities on an Orbitrap Eclipse Tribrid instrument. The authors present it as a tool for, among other things, accurate identification of Leu/Ile isomeric impurities for drug quality control. Conflict of interest noted: several authors are employed by Bloomage Biotechnology Corporation Limited, a commercial manufacturer.

The practical reading: a matching mass is a necessary check, not a sufficient one. For any sequence containing leucine or isoleucine, a mass number alone cannot confirm which one is where.

The certificate describes a moment, not the vial in front of you

Peptides change on their own. A 2026 study in RSC Medicinal Chemistry tracked deamidation and isoaspartate formation — two spontaneous rearrangements — and found them occurring during peptide analysis, purification and storage (PMID 41541711).

Two of its findings are worth stating exactly. First, on the analysis itself: exposure to acidic conditions particularly in the presence of the additive trifluoroacetic acid, as commonly used during HPLC purification, resulted in substantial direct deamidation by hydrolysis. The purification step can generate the very impurity being counted. Second, on detection: conventional chromatographic methods and standard mass spectrometric analyses often fail to distinguish structurally similar peptides with nearly identical physicochemical properties and masses. The authors also observed isoaspartate forming under mildly basic laboratory conditions such as phosphate-buffered saline, and reported clear site dependence, with C-terminal amides markedly more susceptible in their experiments.

A certificate is a measurement of one aliquot on one date. It does not travel with the vial through shipping, storage and time.

What does the evidence not show?

Four honest limits on everything above.

1. None of this is evidence about any particular vendor. These are metrology and pharmaceutical-development papers, run largely on reference materials. They describe how the measurement behaves. They do not establish what is in anyone's catalogue, ours included.

2. Specialists disagree about how hard to push these numbers. A 2025 review in the Journal of Pharmaceutical and Biomedical Analysis argues the opposite of a maximalist reading: that mass balance should not be required as a specification control and should instead be demonstrated during method development, because using it as a formal acceptance criterion offers little value as a formalized quality acceptance criterion and may create more deviations, non-value added investigations, and potential batch failures (PMID 39442464). More testing is not automatically better science. Conflict of interest noted: all four authors are employees of Eli Lilly and Company, a manufacturer with direct commercial interests in this market.

3. The link between an impurity figure and actual risk is not established. Two scientists at FDA's Center for Drug Evaluation and Research wrote in 2025 that gaps still exist in our understanding of the significance of impurities to the overall peptide immunogenicity risk, that current regulatory guidance on impurity qualification thresholds is sparse, and that the available in vitro and in silico assessment methods present technical and methodological limitations (PMID 40607385). Nobody — no vendor, no directory, not this page — can responsibly convert an impurity percentage into a statement about hazard. The agency's own researchers say the science is not there yet.

4. Even the method comparisons carry caveats their headlines omit. A multi-laboratory study organized by the United States Pharmacopeia compared three ways of quantifying a peptide — HPLC assay, quantitative NMR, and amino acid analysis — across national metrology institutes, regulators and manufacturers, using oxytocin as the test case. HPLC showed the lowest inter-laboratory variability. But the authors state that this variability was calculated without counting the uncertainty associated with the purity assignment of the standard with mass balance (PMID 30640042). The comparison flatters HPLC in part by leaving one of its uncertainty sources outside the calculation. That is disclosed in the paper and routinely dropped when the finding is repeated.

What to ask for instead of a bigger number

None of this makes purity figures useless. It makes a bare figure incomplete. Five requests turn one number into a record:

The chromatogram, not only the percentage. The trace shows retention time, the main peak's shape, and whether anything is riding on its shoulder. A number in a table shows none of that.
Peptide content by mass, stated separately. Or the counterion content, from which it can be calculated. This is the number the angiotensin II reference material reported as 691 mg/g, and the one a purity percentage cannot substitute for.
Which method produced the figure — and whether response factors were measured or assumed at 1.
Identity evidence beyond a matching mass, for any sequence containing leucine or isoleucine.
The lot and the date, tied to the vial in hand, not to the product in general.

A supplier who reports peptide content separately from chromatographic purity is telling you something a purity percentage structurally cannot. A supplier who cannot produce the chromatogram behind a number has told you something too. In this category the useful question was never “what is the purity number.” It is “what did that number measure, and what went unmeasured.”

Sources

Kumar Kuril A. The Critical Need for Implementing RRF in the Accurate Assessment of Impurities in Peptide Therapeutics. Anal Chem. 2025 Jun 24;97(24):12480–12485. doi:10.1021/acs.analchem.5c02149 · PMID 40499007
Erckes V, Streuli A, Chamera Rendueles L, Krämer SD, Steuer C. Towards a Consensus for the Analysis and Exchange of TFA as a Counterion in Synthetic Peptides and Its Influence on Membrane Permeation. Pharmaceuticals (Basel). 2025 Aug 5;18(8):1163. doi:10.3390/ph18081163 · PMID 40872554
Melanson JE, Thibeault MP, Stocks BB, Leek DM, McRae G, Meija J. Purity assignment for peptide certified reference materials by combining qNMR and LC-MS/MS amino acid analysis results: application to angiotensin II. Anal Bioanal Chem. 2018 Oct;410(26):6719–6731. doi:10.1007/s00216-018-1272-7 · PMID 30143839
Li M, Josephs RD, Daireaux A, Choteau T, Westwood S, Wielgosz RI, Li H. Identification and accurate quantification of structurally related peptide impurities in synthetic human C-peptide by liquid chromatography-high resolution mass spectrometry. Anal Bioanal Chem. 2018 Aug;410(20):5059–5070. doi:10.1007/s00216-018-1155-y · PMID 29862433
Wu P, Zhang T, Wang G, et al. Direct Characterization of Leucine and Isoleucine Residues in Peptides Using High-Resolution Tandem Mass Spectrometry. Anal Chem. 2026 Jul 21;98(28):20912–20925. doi:10.1021/acs.analchem.6c02059 · PMID 42423382
Erckes V, Rendueles LC, Misiek A, Steuer C. Revealing deamidation and isoaspartate formation during peptide analysis, purification and storage by tandem mass spectrometry. RSC Med Chem. 2025 Dec 29;17(2):1144–1154. doi:10.1039/d5md01025j · PMID 41541711
Hetrick EM, Pack BW, Wolfe CN, Zhao M. Mass balance analysis for therapeutic peptides: Case studies, applications, and perspectives. J Pharm Biomed Anal. 2025 Jan 1;252:116501. doi:10.1016/j.jpba.2024.116501 · PMID 39442464
Puig M, Shubow S. Immunogenicity of therapeutic peptide products: bridging the gaps regarding the role of product-related risk factors. Front Immunol. 2025 Jun 18;16:1608401. doi:10.3389/fimmu.2025.1608401 · PMID 40607385
Li C, Bhavaraju S, Thibeault MP, et al. Survey of peptide quantification methods and comparison of their reproducibility: A case study using oxytocin. J Pharm Biomed Anal. 2019 Mar 20;166:105–112. doi:10.1016/j.jpba.2018.12.028 · PMID 30640042