The Purity Number Was Never the Whole Specification
Of Peptides Wholesale Growth Peptides The conventional selection criterion deserves to be stated precisely before it is qualified. A specification of “≥98% by HPLC” is a real commitment, and a supplier who meets it has done competent work. The problem is structural, not dishonest.
That figure is method-dependent. It states that the main peak represents at least 98% of total integrated peak area under a specific column, gradient and detection wavelength — commonly 214 or 220 nm for peptides, where the amide backbone absorbs. Change the column chemistry, shallow the gradient, or move the wavelength, and the same material can legitimately read 1–3% differently. Two competent laboratories reporting 97.4% and 99.1% on aliquots of one lot are not contradicting each other; they are running different methods. This is why the method identifier matters as much as the number, and why a CoA that reports a percentage without naming the method has told you less than it appears to.
Area percent is also a relative measure. It describes the composition of what absorbed at that wavelength, not the mass of peptide in the vial. Lyophilised material carries water, counterions and residual salts; the peptide content of the powder is a separate value, and for quantitative work — preparing a stock solution to a defined molarity, running a calibration curve — it is the value that matters. A vial can be 99% pure by area and still deliver meaningfully less peptide per milligram than the label implies.
None of this makes the purity figure false. It makes it an answer to a narrower question than most buyers assume they are asking.
What a Single HPLC Trace Cannot See
The blind spot of one-dimensional purity testing is co-elution. A species that travels with the main peak under the release method contributes to that peak’s area and disappears from the impurity total, while remaining physically present in the material.
The impurity classes that matter most in synthetic peptides are well catalogued. Sequence-related impurities — deletion, truncation and insertion sequences — arise from incomplete coupling, premature deprotection or repeated coupling of a residue. Stereoisomers arise from epimerisation during synthesis, which is why the EMA guideline on synthetic peptides treats enantiomeric purity as a characterisation question requiring justification if it is not controlled at release. Then come the post-translational and handling-related species: oxidation of Met, Trp or Cys (+16 Da per event), deamidation of Asn or Gln (+1 Da), N-terminal glutamine cyclisation to pyroglutamate (−17 Da), and residual side-chain protecting groups from incomplete deprotection (+58 Da for tBu, +156 Da for Pbf, +86 Da for Trt).
Many of these carry a characteristic mass shift that a mass spectrometer reads directly, even when the species sits under the main HPLC peak and never appears as a separate UV feature. That is the practical reason both techniques belong on a release record: the chromatogram tells you how much of the UV-absorbing material is the main peak, and the spectrum tells you whether a modified species is hiding inside it.
The EMA guideline makes the same point in regulatory language. It expects analytical procedures for identity and peptide-related impurities to account for structural complexity and the risk of co-eluting impurities, and it states that characterisation of purity should be addressed using an orthogonal approach — size-based, charge-based and hydrophobicity-based separation — with the applicant responsible for demonstrating that the purity methods cover the complete impurity profile.
The consequence for a buyer is direct. A peptide can read 98% pure and still carry an oxidised or deamidated species at a level that changes binding, stability or assay performance. The number is not wrong. It simply was never designed to see that species.
Identity and Mass Confirmation: Two Questions, Not One
Purity and identity are routinely conflated in supplier comparisons, and they answer different questions. Purity asks how much of the material is the main species. Identity asks whether that main species is the molecule you ordered.
Mass spectrometry addresses identity by comparing the observed mass against the theoretical mass calculated from the sequence. A match within a few daltons supports the conclusion that the material is consistent with the intended structure; a larger discrepancy is a signal that the compound may not be what the label claims. For constructs above roughly 2 kDa, high-resolution mass spectrometry is the appropriate instrument, since unit resolution becomes insufficient to distinguish closely related compositions.
Intact mass has limits that are worth stating plainly. It confirms that the molecular formula is consistent with the intended sequence. It does not, by itself, establish that the residues are in the right order across a long chain, distinguish a diastereomer from the target, or locate which of several lysines carries a label. Two constructs with the same composition can share an intact mass and differ in behaviour.
That is why the EMA guideline recommends at least two orthogonal methods for identification at release, and why its characterisation table pairs techniques to questions rather than treating them as interchangeable: MS or LC-MS for molecular mass; LC-MS/MS of the intact molecule, LC-MS of enzymatically digested material (peptide mapping) or Edman degradation for sequence confirmation; amino acid analysis for composition; chiral GC-MS or chiral LC for enantiomeric purity; NMR for structural elucidation and higher-order structure where relevant.
The practical implication for a buyer reviewing a CoA is that a single identity technique is a partial answer. The second method should be chosen for the construct: peptide mapping earns its place on a long sequence or a multi-site modification, Biotech Peptide where intact mass cannot resolve placement; chiral control earns its place where epimerisation risk is real. Asking which second method was used, and why, is a more informative question than asking whether identity was confirmed.
Risk-Based Guidance Moved the Bar Upstream
The regulatory shift behind this argument is specific and worth scoping carefully, because it is easy to overstate.
ICH Q14 reached Step 4 on 1 November 2023 and took effect on 14 June 2024. It describes science- and risk-based approaches to developing and maintaining analytical procedures across their lifecycle, and it sets out two routes: a minimal approach — identify the attributes to test, select the technology, run development studies, document the procedure — and an enhanced approach that adds sample property evaluation, a defined analytical target profile, formal risk assessment against prior knowledge, uni- or multivariate experiments, and an analytical procedure control strategy. ICH Q2(R2), signed off the same day, defines the validation evidence that a procedure is fit for its intended purpose.
Three mechanisms in Q14 matter to a buyer evaluating a supplier. The analytical target profile is a prospective statement of the performance criteria a measurement must meet — it forces the question “what does this method need to be able to do?” to be answered before the method is built. Established conditions tie the procedure’s performance to specific parameters that must be controlled. And change management, aligned with ICH Q12, scales the extent of required studies to prior knowledge and the risk of the change: a column swap on a well-understood method needs less bridging than a change to the separation mechanism itself.
The scope limit matters. These guidelines address analytical procedures used in regulatory submissions; Q2(R2) states that its scientific principles can be applied in a phase-appropriate manner to procedures used during clinical development. Neither document mandates a particular package for research-grade material sold outside a submission. What they change is the direction of scrutiny. Assurance now depends on whether a supplier can show that a method was designed for its purpose and controlled through change — not merely that a result was reported on a certificate.
Testing Depth Should Scale With Complexity and Stage
The appropriate analytical package is not a fixed list. It is a function of two variables, and a supplier offering one package for every order has either over-tested the simple constructs or under-tested the hard ones.
The first variable is construct complexity: chain length, hydrophobicity and aggregation propensity, the number and type of modifications, disulfide topology, and the presence of non-natural residues. The second is program stage: screening, lead optimisation, IND-enabling, scale-up. The two interact. A short unmodified peptide in a qualitative binding assay at screening stage needs little beyond identity, purity and content. The same sequence at IND-enabling stage needs method validation status, forced-degradation evidence that the methods are stability-indicating, counterion identity and residual TFA, water content, residual solvents and, where relevant, endotoxin and microbiological testing.
Escalation between those poles follows the construct. A long hydrophobic sequence or a multi-disulfide construct warrants orthogonal purity methods, an aggregation assessment by size-exclusion chromatography, and — where fibrillary aggregation is a concern — a Thioflavin T assay, which the EMA guideline names explicitly for that purpose. A modified construct warrants confirmation that the modification is present and correctly placed, which usually means mass data interpreted against the expected shift rather than intact mass alone.
The specification items themselves can be anchored to the EMA guideline’s non-exhaustive list for control of the active substance: appearance, identification, purity (total and individual impurities), high-molecular-weight impurities by SEC where relevant, assay or content, counterion identity and content, residual TFA, water content, residual solvents, elemental impurities where metal catalysts were used, bacterial endotoxins and microbiological purity. Not every item applies to every order; the list is a menu for a risk assessment, not a checklist to be applied uniformly.
Scale-up deserves separate mention because it is a risk event in its own right. Splitting batches or pooling sub-batches produced in multiple synthesis cycles is common in solid-phase work, and the EMA guideline expects justification for the practice plus material traceability from synthesis steps through to final active substance. An analytical plan written for a single milligram-scale lot does not automatically cover a pooled kilogram-scale batch. The plan should anticipate the process change rather than react to it after the first comparability question arrives.
What a Connected Data Package Looks Like
The minimum connected data package is the smallest set of records a reviewer needs to reproduce a result or defend a release decision. The operative word is connected: a complete package holds every document, while a connected one makes each document point at the others — lot number to chromatogram, chromatogram to method identifier, method to the specification it was run against, specification to the stability protocol that will test it over time.
A usable CoA carries specific fields, and their absence is diagnosable. It should state the peptide sequence and its version, the N- and C-terminus state (free or capped), the modification site and chemistry, the counterion form (acetate and TFA salts behave differently in cell-based assays, and the default is not always the one you assumed), the lot number, and the method identifiers. For each chromatographic result it should give the column, gradient and detection wavelength, and attach the chromatogram itself rather than a transcribed percentage. For identity it should attach the mass spectrum showing observed against theoretical mass. Purity and peptide content should appear as separate values, each judged against a stated specification, with the release decision attributed to a named analyst.
The integrity expectations behind those fields are public and specific. The ALCOA principles — attributable, legible, contemporaneous, original, accurate — describe what a record must be able to demonstrate, and 21 CFR Part 11 sets the corresponding expectation for electronic records: secure, computer-generated, time-stamped audit trails that record who created, modified or deleted an entry, with prior versions never obscured. A revised CoA issued without a retained prior version is exactly the gap those requirements exist to close.
Change control is the dimension most often missing. A sequence edit or a method change without an immutable version trail is a common cause of a non-reproducible experiment: the lot is real, the data is real, and nobody can prove which molecule was made. ICH Q14’s change-management framing gives a supplier the structure to disclose what changed and what bridging or revalidation followed. Asking for that disclosure is a reasonable question at quotation, not an adversarial one.
The Strongest Counterargument: Testing Costs Money and Time
The objection is fair and should not be waved away: deeper testing raises unit cost and can extend lead time, and for early screening work the marginal value of a full characterisation package is genuinely low. The stage-scaling logic above concedes exactly that point.
The response rests on three grounds. First, the cost of a failed experiment or an unreproducible batch usually exceeds the analytical delta. A month lost to a dose-response curve that will not repeat, or a lead series that has to be re-synthesised because the modification site was never confirmed, is a larger number than the difference between a basic and an extended QC package. Second, testing depth is not testing volume. A well-designed package for a simple construct is cheaper than a bloated one, and the decision rule in this article is explicitly against uniform over-testing. Third, much of the documentation burden is fixed once a supplier has the methods, systems and audit trails in place; the differentiator is capability rather than per-lot effort, which is why suppliers with mature analytical functions can often deliver a fuller package without a proportional price premium.
There is also an honest limit worth stating. Analytical testing characterises the material, not its biological performance. A complete CoA does not establish that a peptide will behave as intended in a given assay, and no certificate substitutes for running the experiment. What the package does is remove a class of variables — wrong identity, hidden modification, unaccounted peptide content — so that when an assay misbehaves, the material is not the first suspect. High Quality Peptides
Where MOL Changes Fits
MOL Changes is a research and development organisation specialising in peptide science, offering custom sequence design through to large-scale production. Synthesis spans solid-phase and microbial fermentation, the modification portfolio covers more than 300 functional groups, scale runs from milligrams to kilograms, and quality control includes HPLC, MS and sterility testing, with production in supplier-described Class 100 cleanroom environments.
The relevant point for this argument is integration rather than any single capability. Modification chemistry, purification method choice and analytical characterisation are planned together, because they are not independent. A dye, PEG chain or lipid changes solubility and LC behaviour, which changes what the purification method must resolve and what the QC plan must confirm. A construct that behaves well on a standard reversed-phase gradient before labelling may not after it. When the modification is designed with the analytical plan in hand, the confirmation that the label is present and correctly placed is built into the release method rather than bolted on after a discrepancy appears.
If you are working with a complex or analytically demanding construct — a long hydrophobic sequence, a multi-disulfide topology, a multi-site modification, or a program moving from screening toward IND-enabling — the productive conversation at quotation is about the specification and the QC plan, not only the purity target. That discussion is where the analytical package becomes part of the service.
A Decision Rule for the Next Supplier Review
The next supplier review can be structured around a short set of questions, Acetyl Hexapeptide 1 Factory each aimed at a specific gap the purity number leaves open.
-
Which methods support the purity figure, and what is the quantitation limit? A percentage without a method identifier is not comparable across suppliers. The quantitation limit determines the smallest impurity the method can honestly report.
-
Is identity confirmed by a second orthogonal technique, and which one for this construct? Intact mass plus a method chosen for the sequence length and modification pattern is a different answer from intact mass alone. Isotope Labeled Peptides Wholesale
-
Are sequence-related impurities addressed, and how are co-eluting species detected? The answer should name the technique — typically mass data interpreted against expected shifts — rather than assert that impurities are controlled.
-
Is peptide content reported separately from total mass? For any quantitative application, this is the number that sets your stock concentration.
-
Does the CoA carry method identifiers, the raw chromatogram and spectrum, and the specification each value is judged against? Transcribed numbers without the underlying data cannot be checked.
-
How are method and sequence changes version-controlled and disclosed? An immutable version trail is what makes a lot traceable to a molecule.
-
Does the testing plan scale with construct complexity and program stage? A supplier who asks about your intended use before quoting a QC package is demonstrating the capability this article has described.
The core claim is simple to restate. The analytical package is part of the service specification, and evaluating it at quotation — not at receipt — is what separates a reliable supply relationship from an expensive surprise.
Frequently Asked Questions
Isn’t a ≥98% HPLC purity specification sufficient for research-grade material?
It depends on the construct and the use. For a short unmodified peptide in a qualitative assay, it may be adequate. The figure is method-dependent and relative, so it does not describe what the remaining percentage contains or how much actual peptide the vial holds. For anything quantitative, modified, or destined for in vivo work, it is a starting point rather than a specification.
If mass confirmation matches the theoretical mass, isn’t identity proven?
A mass match supports identity but does not by itself establish sequence order, isomer status or modification placement. That is why the EMA guideline recommends at least two orthogonal identification methods at release, and why the second method should be selected for the specific construct — peptide mapping for a long or multi-site-modified sequence, chiral control where epimerisation risk is real.
Doesn’t deeper testing just add cost and delay to early-stage work?
It can, and the stage-scaling logic in this article deliberately does not recommend a full package for screening material. The argument is for matching depth to complexity and stage, and for treating the analytical plan as a design decision made at quotation rather than a fixed add-on applied uniformly.
How do ICH Q14 and Q2(R2) apply to a custom peptide order that is not a regulatory submission?
They do not impose requirements on research-grade supply. Their principles — analytical target profile, risk-based development, lifecycle change management, and validation demonstrating fitness for purpose — describe what a mature analytical function looks like and can be applied phase-appropriately. They are useful as a benchmark for supplier capability rather than a compliance threshold.
What single document should I ask for first when evaluating a new supplier?
The CoA for a comparable construct is the highest-yield starting point. Review it for method identifiers, the chromatogram and spectrum rather than a transcribed number, separate purity and content values, and the specification each result is judged against. A supplier unable to produce that for a similar order type is unlikely to produce it for yours.
Disclosure: MOL Changes is a commercial provider of custom peptide synthesis and modification services and has a commercial interest in the quality standards discussed in this article. Standards and guideline references are provided for informational purposes and do not constitute regulatory advice.
A cleanroom classification describes environmental control; product sterility requires separate process and product evidence.
