A vial of research peptide is the output of a chemical process, and the chromatogram that accompanies it is a picture of how that process went. Reading it usefully means knowing what solid-phase peptide synthesis can produce besides the sequence you ordered — because those by-products are not random. They are the predictable residue of steps that did not run to completion, or that ran somewhere else. This guide walks the synthesis from resin to cleavage and names the species that appear on the analytical trace at each stage — and the ones that do not appear at all.

The cycle, and why everything traces back to it

Solid-phase peptide synthesis anchors the C-terminal residue to an insoluble polymer support and extends the chain toward the N-terminus one residue at a time. Each cycle has the same four moves: remove the temporary protecting group from the chain's free α-amino group, wash, couple the next amino acid with its carboxyl activated, wash again. The point of the solid support is that excess reagent is removed by filtration rather than by isolating an intermediate, which is what makes long sequences practical at all. When the chain is complete, the peptide is cleaved from the support and the permanent side-chain protecting groups are removed, usually in the same acid treatment.

The consequence of that architecture is that errors accumulate and cannot be purified out mid-synthesis. A useful way to feel the arithmetic: a 30-residue peptide needs 29 couplings and 29 deprotections. If every one of those 58 steps went to 99% completion, the fraction of chains that saw no failure at all would be 0.9958 ≈ 0.56 — about 56%. That number is illustrative, not a measured yield for any particular synthesis, but it explains why crude SPPS material is a mixture and why purification is doing real work rather than cosmetic work.

Fmoc and Boc: protecting the growing chain

Two protecting-group strategies dominate. In Boc chemistry the temporary α-amino protection is removed with acid at every cycle, and the permanent side-chain groups must therefore survive repeated acid exposure — which pushes the final cleavage to a much stronger acid. In Fmoc chemistry the temporary group is removed with base, typically a secondary amine, so the side-chain groups and the resin linkage can be acid-labile and come off together in a milder final step. The two schemes are mutually orthogonal in exactly this sense: acidic conditions are incompatible with Boc protection and basic conditions are incompatible with Fmoc protection, and safety-catch linkers were developed specifically to exploit that split — stable under the conditions needed for both α-amino and side-chain deprotection until a simple chemical step makes them labile to one of them [2].

Fmoc chemistry is now the default. A 2016 review describes Fmoc SPPS as the method of choice and attributes the availability of very-high-quality building blocks at low cost to the economies of scale arising from multiton-scale production by this route [1]. That matters for impurity work in a practical way: the failure modes described below are mostly Fmoc failure modes, because that is what most material on the market is made by. Boc chemistry has not disappeared — it remains useful where base-driven side reactions are the limiting problem — but the two routes leave different fingerprints, and a chromatogram is easier to interpret when you know which one produced it.

Resin and linker

The linker is the moiety that joins the peptide to the solid support and simultaneously protects the C-terminal carboxyl group [2]. The conditions that release it determine whether you finish with a peptide acid, a peptide amide, or a still-protected fragment. The support itself is not inert scenery. Classical polystyrene beads swell in some solvents and not others, and a chain that folds back on itself or on its neighbors becomes physically inaccessible to reagent. Poly(ethylene glycol)-based supports were developed in response. One all-PEG resin built from primary ether bonds was reported to perform well for hydrophobic, highly structured and poly-arginine sequences relative to polystyrene; in the most striking case reported, stepwise assembly of β-amyloid (1–42) gave crude material of 91% purity, where literature procedures on polystyrene or PEG-polystyrene supports had required convergent approaches, additional time-consuming steps, or both [3].

That comparison is worth holding onto: much of a crude impurity profile is set by choices made before the first amino acid is coupled. Two labs synthesizing the same sequence on different supports do not get the same chromatogram.

Difficult sequences: aggregation on the resin

Some sequences resist synthesis in a way that has nothing to do with reagent quality. The growing chains associate through backbone hydrogen bonding into β-sheet-like structures on the resin, and the α-amino terminus becomes shielded from both the coupling reagent and the deprotection base. A widely used Fmoc/tBu protocol paper states plainly that aggregation of the growing peptide chain is the main cause of failure for peptide chemosynthesis, and that special techniques are required to reduce it — the authors demonstrate depsipeptide and pseudoproline units for a sequence they describe as impossible to obtain using the standard protocol [4]. Those backbone-disrupting building blocks work by breaking up the hydrogen-bonding pattern, not by making coupling chemistry intrinsically faster.

Deletion and truncation sequences

Two related impurity classes arise when a cycle does not go to completion, and the distinction between them is worth keeping straight.

  • A deletion sequence is a chain that failed one coupling but continued to grow. The product is missing an internal residue and is otherwise correct. Its mass differs from the target by exactly the residue mass of the amino acid that never got in.
  • A truncation sequence is a chain that stopped elongating and remains a shortened fragment. Capping — acetylating unreacted amino groups after a coupling step — deliberately converts potential deletions into truncations, on the reasoning that a shorter fragment of different hydrophobicity is easier to separate than a full-length peptide missing one residue.

Chain shortening can also come from a specific chemical dead end rather than from slow coupling, and it produces a species that fits neither definition above cleanly. Diketopiperazine formation, in which the first two residues cyclize and fall off the linker, is a serious side reaction in the solid-phase synthesis of peptide acids with proline or glycine at the C-terminus; it lowers overall yield and, importantly, populates the crude with several deletion peptides lacking the first amino acids. In the same study, resins loaded to higher levels gave more impure crudes, and the authors identified these species by HPLC coupled to electrospray mass spectrometry [6].

Incomplete deprotection produces its own family. A detailed survey of Fmoc procedures reported frequent incomplete deblocking by standard piperidine reactions, double additions of particular amino acids within a single coupling cycle, and — a finding that should temper confidence in in-process checks — failure of the Kaiser ninhydrin test to detect free amino groups of certain amino acids at the N-terminus. Those authors concluded that Fmoc synthesis needed more careful monitoring than standard protocols then recommended, and proposed HPLC plus mass spectrometry as a minimum analytical sequence [5].

Aspartimide: the side reaction that will not go away

Aspartimide formation is the best-characterized and most stubborn base-mediated side reaction in Fmoc SPPS. The backbone nitrogen of the residue following an aspartic acid attacks the protected β-carboxyl side chain, closing a five-membered succinimide ring and expelling the side-chain protecting group. Because ring closure eliminates the elements of water relative to the linear chain, the aspartimide species is 18 Da lighter than the target peptide — a mass shift small enough that it must be resolved, not merely detected.

The consequences are worse than the loss of one clean peak. A 2025 review calls the side reaction a persistent challenge in Fmoc SPPS and surveys suppression strategies across three categories — ester β-carboxyl protecting groups, non-ester β-carboxyl masking groups, and backbone-protecting groups — alongside alternatives to the Fmoc group itself that are cleavable under non-basic conditions; its authors argue that developing compatible and scalable prevention tactics remains essential [7]. That the field is still publishing new masking chemistries is itself the honest summary of where the problem stands.

Three downstream consequences of aspartimide formation deserve specific attention:

  • α- and β-aspartyl peptides. Hydrolysis of an aspartimide-containing peptide — which can happen during HPLC purification itself — yields a mixture of α- and β-aspartyl peptides that cannot be purified easily [10]. These are isomers of the target. They have the same molecular formula and the same mass, so mass spectrometry alone will not distinguish them from the correct product; chromatographic resolution is the practical handle in a routine purity assay.
  • Ring-opened adducts. The aspartimide ring is electrophilic, and a nucleophile that reaches it opens the ring to give an aspartyl derivative. The best-characterized version of this is deliberate rather than accidental: work aimed at finding the optimal conditions for ring opening examined primary and secondary alcohols in combination with bases including diisopropylethylamine, collidine, 4-pyrrolidinopyridine and piperidine, obtained the best results with a primary alcohol plus diisopropylethylamine, and found that secondary alcohols opened the ring only at high temperatures [10]. That study set out to open the ring on purpose; it does not quantify how much ring-opened material an ordinary synthesis carries, and how far this class populates a routine crude is not something the literature reviewed here settles.
  • Chain termination. When the Xaa-Asp-Yaa motif sits at the N-terminus, the aspartimide intermediate can be intercepted by the chain's own free amino group. Investigators at a peptide manufacturer observed significant chain termination from this route and, investigating a linear model peptide alongside an independently synthesized piperazine-2,5-dione derivative, reported that NMR data on the side product present in the crude linear peptide show exclusively the six-membered ring and not the theoretically conceivable seven-membered 1,4-diazepine-2,5-dione. They propose a mechanism involving nucleophilic attack at the α-carbon of the aspartimide carbonyl, and report that hindering aspartimide formation reduces or completely suppresses the termination [9].

Mitigations exist and are partial. Adding HOBt to the deprotection solution, or using piperazine in place of piperidine, reduced aspartimide formation and the aspartic acid racemization that follows it in a microwave-assisted study [8].

Racemization and epimerization

A stereochemically inverted residue produces a diastereomer of the target: same mass, different retention time, and often only slightly different. In a model 20-mer containing all 20 natural amino acids assembled under microwave heating, cysteine, histidine and aspartic acid were the residues susceptible to racemization; lowering the coupling temperature from 80 °C to 50 °C limited racemization of histidine and cysteine, and racemization during coupling was found to be limited to the activated ester state up to 80 °C, so those two residues could be coupled conventionally while the rest of the chain was assembled under microwave without deleterious effect [8]. Use of the hindered base collidine in the coupling also minimized formation of D-cysteine in that work.

Like the aspartyl isomers, epimers are invisible to a mass-only assay. If a certificate reports purity by LC-UV and identity by mass, neither measurement is sensitive to this class of impurity unless the chromatographic method happens to resolve it.

Side reactions from activation and from cleavage

Two more classes are worth naming because they are sequence-specific and therefore predictable.

Asparagine with an unprotected side chain can be dehydrated at the side-chain amide during carboxyl activation, giving β-cyanoalanine in the product; work comparing activation methods found varying amounts of this by-product and obtained homogeneous peptide either by using side-chain-protected asparagine derivatives with BOP activation or by coupling the unprotected residue as its preformed pentafluorophenyl ester [11]. The same study documents the other end of the problem: the Mbh and Tmob side-chain protecting groups generate carbonium ions during acidolytic cleavage that readily alkylate tryptophan residues, and alkylation could be reduced by adding scavengers to the trifluoroacetic acid during cleavage [11]. This is why cleavage cocktails contain scavengers at all, and why a tryptophan-containing sequence deserves a closer look at the trailing edge of its main peak.

What the chromatogram shows and what it hides

The honest position is that a single reversed-phase run with UV detection is a screening tool, not a full accounting. Work at the Bureau International des Poids et Mesures with China's National Institute of Metrology, on a synthetic human C-peptide material that served as the study material for an international comparison of laboratories' peptide purity mass fraction assignments, used liquid chromatography with high-resolution mass spectrometry to identify, confirm and quantify more than 65 structurally related peptide impurities, estimating their total mass fraction at 83.3 mg/g with an expanded uncertainty of 3.0 mg/g (k = 2); the authors note that every approach to assigning peptide purity requires a correction for structurally related impurities in order to avoid bias [12]. Sixty-five species is not what a routine purity chromatogram resolves. It is what is there.

Practical implications for reading a document you have been handed:

  • A stated purity figure is method-dependent. It is the area percent that one gradient on one column with one detector was able to separate. It is not a count of molecules.
  • Isomeric impurities — aspartyl isomers, epimers — carry no mass signature. Confirming the target mass by MS does not exclude them.
  • Deletion sequences do carry a mass signature, and a mass spectrum covering the region below the target mass is a more informative document than a single reported m/z.
  • Partially protected peptides survive incomplete final deprotection and are typically more hydrophobic than the target, so they elute later.
  • The impurity profile is characteristic of the route. Two lots of the same sequence made on different resins with different deprotection cocktails are not expected to have the same minor peaks.

What is measured and what is convention

Much of what follows the synthesis is convention with thin primary literature behind it. The 95% and 98% purity thresholds that appear on research-grade certificates are commercial and regulatory conventions, not values derived from a published study of what a given assay requires. Capping with acetic anhydride is standard practice and its rationale is chemically sound, but the claim that it improves final purified purity for a given sequence is one to expect demonstrated rather than assumed. Nothing in the literature reviewed here establishes a threshold below which an impurity profile stops mattering — that judgment belongs to the assay, not to the certificate.

What is well established is narrower and more useful: aspartimide formation in Fmoc chemistry is base-mediated [7], and its extent depends on the beta-carboxyl protecting group, on the acid or base used, and on the peptide sequence [10]; racemization at cysteine and histidine is temperature-dependent under microwave conditions [8]; diketopiperazine formation is a serious side reaction for peptide acids carrying proline or glycine at the C-terminus, and higher resin loading gives more impure crudes [6]; aggregation is the dominant cause of synthesis failure for difficult sequences [4]; and one synthetic peptide material, examined by liquid chromatography with high-resolution mass spectrometry, was found to contain more than 65 structurally related peptide impurities [12].

Once material arrives, the impurity profile it was born with is fixed; the remaining variable is what handling does to it, which is covered in our guide to peptide reconstitution and handling. For converting a labeled mass into a molar concentration, you need the compound's molecular weight; terms used above are defined in the research glossary.

All material discussed here is for laboratory research use only.