Where Peptide Impurities Actually Come From

Quality & analysis

Peptide impurities are not contamination in the ordinary sense. Nothing fell into the vial. Almost everything that is not the target peptide was manufactured alongside it, by the same chemistry, in the same reaction vessel – and the specific things that go wrong are predictable enough that you can read a sequence and guess which ones will show up. That is the useful part, and it is what a purity percentage on a certificate never tells you.

Peptide impurities are built, not introduced

Solid-phase peptide synthesis assembles a chain one residue at a time on a solid support, with each cycle deprotecting, coupling and washing before the next residue goes on. If you want the mechanics, start with our overview of solid-phase peptide synthesis. What matters here is a consequence of the design: because the chain stays anchored and unreacted material is simply washed away, every failed step leaves a slightly wrong molecule still attached, still growing, and still there at the end.

So the crude material coming off the resin is not the target peptide plus dirt. It is a population of closely related molecules, most of them differing from the target by one residue, one stereocentre, or one rearranged side chain.

The arithmetic that explains most of it

A per-step yield of 99 percent sounds like near-perfection. Over a 31-residue peptide – 30 couplings – it means roughly 74 percent of chains still carry the target sequence. One molecule in four is something else, before any other side reaction has happened at all.

This is not a modern observation. The 1984 Nobel announcement for the invention of the method ran the same calculation over 100 steps: at 90 percent per step the overall yield is 0.003 percent, while at 99.5 percent it is 61 percent. The entire history of the technique is an argument about the second decimal place.

Published crude purities for real sequences land about where the arithmetic predicts. Instrument vendors report optimised crude purities in the low-to-mid 60s to low 80s percent for 17- to 34-residue peptides – and those are best cases, before purification.

Deletions, truncations, and the occasional double insertion

The most common impurity is the simplest: a coupling failed, the next residue went on anyway, and the resulting chain is missing one amino acid. Related to it is the truncated chain, where growth stopped altogether – sometimes deliberately, since capping unreacted chains is standard practice precisely so that failures stop growing rather than producing deletion sequences that are harder to separate.

Capping can also happen by accident, and the mechanism is a nice illustration of how small an input error has to be. Acetic acid contaminating an Fmoc-amino-acid building block causes permanent capping, and because acetic acid has a molecular mass of only 60, negligible amounts terminate a meaningful number of chains. Published guidance calls for levels below 0.02 percent for clean synthesis. It also cannot be detected by RP-HPLC in the building block.

The opposite error exists too. Free amino acid contaminating a building block can result in a residue being incorporated twice, and reagent side reactions can generate dipeptide derivatives that get installed as a unit. So impurity peptides can be one residue short or one residue long.

Aspartimide: nine wrong answers from one right residue

The most serious side reaction in Fmoc chemistry is aspartimide formation, caused by exposing an aspartic-acid-containing sequence to strong base – which happens at every deprotection step. It is described as particularly pernicious because a single site can generate up to nine different byproducts, some of which co-elute with the target peptide.

It is also sequence-predictable. The prone motifs are Asp-Gly, Asp-Asp, Asp-Asn, Asp-Arg, Asp-Thr and Asp-Cys, with Asp-Gly the worst case. Under standard conditions in a model peptide, the measured rate was around 1.65 percent aspartimide per cycle – and cycles repeat. In one test peptide containing Asp-Gly, total impurities reached 44 percent under standard deprotection conditions, falling to 15 percent when an additive was included.

Most of the resulting species separate cleanly on RP-HPLC. One does not: the epimerised alpha-aspartyl peptide is described as very difficult or impossible to resolve. Which brings us to the impurity class that matters most and shows up least.

Epimers: the impurity your chromatogram cannot see

An epimer has the same atoms, the same formula, the same exact mass and often the same retention time as the target. It differs only in the three-dimensional arrangement at one carbon – a D-amino acid where an L belongs. Mass spectrometry confirms identity by mass, so it cannot distinguish them either; see our note on what mass spectrometry actually confirms.

How much appears is a function of conditions, not fate. Measured D-cysteine content during coupling of Trt-protected cysteine under base-mediated activation rose from 8.0 percent at room temperature to 10.9 percent at 50 degrees C and 26.6 percent at 80 degrees C. Histidine behaves similarly with preactivation time, rising from about 1 percent with no preactivation to 7.8 percent after five minutes.

Better protecting groups cut both figures dramatically – the same cysteine experiment fell to 0.4 to 1.3 percent across the same temperature range with a different protecting group. The point is not that cysteine is doomed; it is that two vials with the same stated purity can come from very different process decisions.

What survives purification

Crude peptide is purified, usually by preparative reversed-phase HPLC, and this removes most of the deletion and truncation species. It does not remove what it cannot separate. Our explainer on HPLC purity covers the method; the table below is the map of what it can and cannot resolve.

The counterion is an impurity by weight, not by peak

Trifluoroacetic acid is used both to cleave the peptide from the resin and as an ion-pairing reagent during purification, so peptides are routinely obtained as TFA salts. The counterion pairs with positively charged groups – the N-terminus and the arginine, lysine and histidine side chains – and it does not simply wash out.

The quantities are not trivial. In a 2025 analysis, calculated theoretical trifluoroacetate mass fractions across a set of peptides ran from 22.7 percent for a doubly charged hexapeptide to 35.2 percent for a nine-charge cell-penetrating peptide. Measured values were consistent and sometimes higher: angiotensin I straight from purification contained 0.333 mg of trifluoroacetate per mg of peptide salt, and the authors noted that synthetic peptides appear to carry excess TFA beyond what charge alone predicts, because repeated lyophilisation does not break the ion pairing. Three lyophilisation cycles with plain water still left 21.5 percent.

None of that shows up as a chromatographic peak, because it is not a peptide. It shows up on the balance. That is the distinction between purity and net peptide content, and it is why the salt form belongs on a certificate. The same paper noted that counterions are often unreported in the published literature, let alone in commerce.

What to do with this when reading a certificate

• Ask which method produced the number. Purity is method-dependent by definition. A percentage with no gradient, column or detection wavelength is a claim, not a measurement.
• Treat purity and content as two questions. One is what fraction of the peptide-like material is the target. The other is what fraction of the powder is peptide at all.
• Read the sequence for red flags. Asp-Gly, multiple arginines, cysteine and histidine all have documented, specific failure modes.
• Do not treat length as the difficulty metric. A widely quoted 50-residue ceiling for synthesis has been called meaningless in practice by working chemists, because many much shorter sequences are extremely problematic. Sequence beats length.
• Remember the regulatory floor is much lower. Regulators work with peptide impurity identification thresholds far below what a research certificate reports. A 98 percent purity claim leaves 2 percent unaccounted for, and in a regulated setting that space would need naming.

Our guides to batch-matched certificates and reading a COA go through the document itself line by line.

Frequently asked questions

If a peptide is 98 percent pure, is 2 percent of the vial junk?

Not exactly. The 2 percent refers to other peptide-related species detected by that method – mostly close relatives of the target, such as deletion sequences. It says nothing about counterion or water content, which are measured separately and can account for a much larger share of the powder’s mass.

Why can’t HPLC detect a D-amino acid impurity?

Because an epimer is chemically identical in composition and mass; only its three-dimensional arrangement differs. It frequently co-elutes with the target on a standard reversed-phase method, and it has the same mass on a mass spectrometer. Detecting it requires methods designed for the purpose.

Does a longer peptide always have more impurities?

On average yes, because stepwise yields compound. But sequence matters more than length. A short sequence containing Asp-Gly and a hydrophobic stretch can be harder to make cleanly than a longer, better-behaved one.

Can the manufacturer remove the TFA?

Only by an additional ion-exchange step to a different salt form. Published guidance states TFA cannot be completely removed by lyophilisation alone because of the strength of the salt pairing, and repeated lyophilisation with water was shown to leave over 20 percent in place.

References

Informational only – not medical advice · 21+. VialHelp is an educational resource and does not sell or recommend peptides or suppliers.