How Peptides Are Made: Synthesis, Purity, and Why Quality Varies

Most people who use peptides never stop to think about how the stuff in the vial got there. But how a peptide is made explains almost everything that matters about its quality: where impurities come from, why two products with the same name can be very different, and why a Certificate of Analysis (a lab report on what’s actually in the product) is worth reading closely. This guide walks through how peptides are made, the ways things tend to go wrong, and how quality is actually measured.

Two ways to make a peptide

There are two main ways to make a peptide, and which one gets used mostly comes down to size.

Chemical synthesis is the everyday workhorse for small and medium peptides. (A peptide is just a short chain of amino acids, the building blocks that make up proteins.) The main method is called solid-phase peptide synthesis (SPPS) — a way of building the chain while it’s attached to a tiny solid bead. It was introduced by Robert Bruce Merrifield in a 1963 paper and earned the 1984 Nobel Prize in Chemistry. The idea is clever. The first amino acid is anchored to a small, solid plastic bead (the “solid phase”), and the chain is built up one piece at a time. Because the growing chain stays stuck to the bead, leftover chemicals and waste can just be rinsed away after each step. The cycle repeats — unlock the open end of the chain, attach the next amino acid, rinse, repeat — until the full sequence is built. At the end, the finished peptide is snipped off the bead and the last protective pieces are removed.

Two recipes lead the field. The original one, called Boc, used strong acid over and over, plus a harsh final step using hydrogen fluoride. The modern standard, Fmoc, is “orthogonal” — meaning its two kinds of protective coverings come off under different conditions, so you can remove one without disturbing the other. A mild base uncovers the open end of the chain, while the acid-sensitive coverings on the side stay put until one final acid step at the end. Fmoc is gentler and handles longer chains better, which is why most synthetic peptides today are made this way.

Recombinant (biological) production is used for bigger peptides and proteins, like insulin. Here, the gene that spells out the peptide is placed into a living host — often E. coli bacteria or yeast — which then makes the peptide for you. Often it comes out as a fusion protein: the peptide you want is joined to a carrier protein that helps the host make and fold it correctly. The carrier is then cut away with an enzyme, and the peptide is folded into shape and cleaned up. Some regulated medical peptides use a mix of both approaches, starting with a recombinant piece and then changing it chemically.

Where impurities come from

This is the part that matters most for quality, and the key point is this: impurities in synthetic peptides are the expected leftovers of the chemistry, not random dirt that fell in. Because the chain is built one step at a time, every step is a chance for something to go slightly wrong.

Sequence errors are the easiest to picture. If a step doesn’t fully finish, you can end up with truncated chains (building stopped too early), deletion chains (a piece in the middle is missing), or insertion chains (an extra piece snuck in). One common factory trick is “capping”: after each step, any chains that failed to react are chemically blocked so they stop growing. This turns hard-to-remove deletion chains into shorter truncated ones, which are easier to filter out later.

Side reactions are trickier to spot. The best-known one is aspartimide formation, a small internal kink that forms at certain spots in the chain (at aspartic acid, one of the amino acids). It is mainly triggered by the base used in the Fmoc step (a chemical called piperidine), though acidic conditions add to it too. It depends heavily on the local sequence: the Asp-Gly pairing is the worst offender, and several other neighboring amino acids (including Ala, Ser, Thr, Cys, Arg, Asp, and Asn) are prone to it as well. Aspartimide formation creates a whole family of unwanted leftovers, including rearranged chains and racemized versions (mirror-image, D-form copies of the molecule), and some of these are hard to separate from the real thing. Other side reactions include broader racemization (more mirror-image forms), oxidation (for example at methionine), deamidation, leftover coverings that didn’t come off, and clumped-together pairs and add-ons.

Non-peptide leftovers round out the picture. Peptides made by SPPS usually come out as trifluoroacetate (TFA) salts — paired with a leftover bit of a chemical called TFA. If that leftover isn’t swapped out, it can throw off lab tests and even change how much actual peptide is in a vial. Depending on the method and how carefully it’s run, products can also carry leftover solvents and chemicals, heavy metals, bacterial endotoxin (a substance from bacteria that can cause reactions), or live microbes — worries that loom larger for recombinant material and for poorly controlled batches.

How purity and identity are measured

Once a peptide is made, it has to be cleaned up and checked. Cleanup usually relies on reversed-phase HPLC (high-performance liquid chromatography — a method that separates a mixture into its parts), sometimes alongside another separation method called ion-exchange chromatography, using UV light near 215 nm to spot the peptide itself.

Two lab tests then answer two different questions:

HPLC tells you how pure the material is. It splits the contents into separate peaks, and purity is reported as the size of the target peak compared to all the peaks together. It’s a relative number based on separation — which is why a trustworthy result shows the actual graph (the chromatogram), not just a percentage.
Mass spectrometry (LC-MS or ESI-MS) tells you what it actually is. By weighing the molecule, it checks whether the weight matches the sequence you expect. One quirk: the technique (electrospray ionization) often gives molecules more than one charge, so a peak at roughly half the expected mass-to-charge value is normal and not a mistake.

Other cross-checks include amino acid analysis, peptide mapping, NMR, and measuring how much counterion, water, and leftover solvent are present. These are the same ideas covered in our guide to reading a COA: purity without identity is empty, because a sample can be 99% pure and still be 99% of the wrong molecule.

What “good” quality control looks like

For pharmaceutical (medical-grade) peptides, there’s a published yardstick. USP General Chapter ⟨1503⟩, “Quality Attributes of Synthetic Peptide Drug Substances” (United States Pharmacopeia, 2021) spells out what to expect for how the peptide is made, the raw materials used, the assay and content, impurities and related compounds, microbes, and bacterial endotoxins. The standard set of lab tests it describes — HPLC/UPLC, LC-MS and LC-MS/MS, amino acid analysis, NMR, and peptide mapping — lines up directly with the impurities described above. A companion chapter, ⟨1504⟩, covers the starting materials used in chemical synthesis. By USP’s own rules, chapters numbered above 1000 (like these) are guidance rather than hard requirements, but they reflect the standard that regulated manufacturers are held to.

Why gray-market quality varies so much

None of these standards apply to peptides sold as “research chemicals.” That’s the heart of the problem, and there’s direct evidence for it.

In a 2008 study, researchers tested the same peptide (obestatin) from five different makers. One product turned out to be a completely different peptide, and about two-thirds of the rest weren’t good enough for reliable lab use — purity below 95%, or single impurities above 1%. That’s what shaky quality looks like in real life: not small differences, but mislabeled products and material that would fail any serious standard.

The “research use only” label doesn’t fix any of this. The US Department of Defense’s Operation Supplement Safety program notes that for peptide-hormone products, “neither the purity nor the potency… can be ensured,” and that the “for research purposes only” label is used even while these products are marketed to consumers and athletes. The label is a legal and marketing shield, not a promise of quality — a point we cover more fully in Research Use Only, Explained.

Regulators have flagged related problems with the compounded and gray-market GLP-1 drugs that now dominate this space. The FDA has raised concerns that some products use salt forms (such as semaglutide sodium or acetate) that are actually different active ingredients than the approved version, that these products get no review for safety or quality before sale, and that problems are likely underreported. The agency has separately warned about dosing errors with compounded injectable semaglutide — reports of doses 5 to 20 times what was intended, with serious harm — and has launched a “Green List” of compliant foreign API manufacturers under an import alert, the point being that a lot of imported material does not meet that bar. The FDA has also sent warning letters to research-peptide sellers offering unapproved tirzepatide without a prescription, and has proposed barring semaglutide, tirzepatide, and liraglutide from the list of substances that outsourcing facilities are allowed to compound from bulk.

Anti-doping rules add one more layer worth stating carefully. As of 1 January 2026, markers of semaglutide and tirzepatide are on WADA’s Monitoring Program — watched, but not banned and carrying no penalty. By contrast, growth hormone secretagogues such as GHRP-2, hexarelin, ipamorelin, and CJC-1295 are banned under Section S2 of the Prohibited List. BPC-157 isn’t named directly but is generally treated as covered by the list’s catch-all wording for non-approved substances. We go deeper on this in Peptides and Anti-Doping.

How this ties back to the COA

Everything above is why a meaningful Certificate of Analysis matters. The COA exists precisely to catch the truncation, deletion, aspartimide, racemization, and counterion problems that synthesis always produces. A useful one lets you confirm five things: that the identity matches the label, that HPLC purity is shown with the actual graph, that mass spectrometry confirms the expected weight, that a batch or lot number ties the document to the exact vial in your hand, and that a named, identifiable lab ran the tests — ideally an independent outside lab, because a self-reported COA can simply be made up. The obestatin study is the reason that last point matters: labels and certificates can be wrong, which is the whole argument for checking with an independent source. See our notes on independent testing and the running ledger for how this plays out across real products.

Bottom line

Peptides are made one of two ways: by building a chain one amino acid at a time on a solid bead (chemical synthesis, used for most peptides), or by growing them in engineered cells (recombinant production, used for bigger molecules). The chemistry creates a predictable set of leftovers — truncations, deletions, aspartimide and racemization byproducts, and leftover counterion and chemicals — which is exactly what cleanup and testing are built to catch. Medical-grade manufacturing is held to published standards like USP ⟨1503⟩. Gray-market “research” peptides are not, and the evidence shows their quality genuinely varies — sometimes to the point of being a completely different compound. That gap is why a real, lab-backed COA, not a marketing claim, is the only honest window you have into what’s actually in a vial.

Sources

New profiles and evidence updates only. No spam, and we never sell or share your email.