Residual Solvents In Synthetic Peptides, And How They Are Limited

Solid-phase synthesis runs in dimethylformamide or N-methylpyrrolidone. Fmoc removal runs in piperidine. Cleavage from resin runs in trifluoroacetic acid, often with scavengers. Purification runs in acetonitrile and water with an acid modifier, and the collected fractions are frozen and dried. Every one of those liquids has an opportunity to stay behind, either trapped in the solid or bound to the peptide as a salt.

Residual solvent testing separates the volatile fraction from the powder and quantifies it. The standard approach is static headspace sampling: a weighed portion of powder is dissolved or suspended in a diluent, sealed in a vial, equilibrated at a fixed temperature, and a measured volume of the vapour above it is drawn onto a gas chromatography column. Flame ionisation detection gives the response, and quantitation is against a prepared standard of the same solvents. USP General Chapter 467 and European Pharmacopoeia 2.4.24 both describe this procedure.

The result is a set of numbers in parts per million by mass, one per solvent looked for. A solvent that was never used in the route will not be reported, because headspace gas chromatography only finds what the method was set up to separate and calibrate.

What ICH Q3C actually limits

ICH Q3C sorts solvents into three classes by toxicological concern. Class 1 solvents are to be avoided. Class 2 solvents are restricted. Class 3 solvents are regarded as low concern and carry a general ceiling. The guideline expresses limits two ways, and the one that is checkable against a certificate is Option 1, a fixed concentration ceiling in parts per million.

The Class 3 ceiling of 5000 ppm is 0.5% by mass. That is not a trivial quantity, and it is one reason a purity figure and a mass balance can disagree. A powder can be well within every solvent limit and still be several percent something other than peptide once water and counter-ion are counted.

Trifluoroacetate is not a residual solvent

Trifluoroacetic acid is the most abundant non-peptide species in most crude synthetic peptides, and it is the one that residual solvent testing is worst at seeing. After cleavage and acidic purification the peptide leaves the process as a trifluoroacetate salt, with the anion paired to every basic side chain. It is not a trapped volatile sitting in the headspace. It is an ionic component of the solid.

ICH Q3C does not classify it, so there is no Q3C limit to compare a number against. Measuring it takes a different method: ion chromatography with conductivity detection, capillary electrophoresis, or fluorine-19 nuclear magnetic resonance against an internal fluorinated standard. Reported figures for peptides purified as the trifluoroacetate salt commonly fall in the range of 5% to 15% by mass for peptides with several basic residues, determined by ion chromatography. A salt exchange step, typically into acetate or hydrochloride, is what lowers it.

The failure mode: occlusion inside the cake

Static headspace gas chromatography measures the vapour in equilibrium with the sample, and that equilibrium assumes the solvent can reach the vapour phase within the equilibration time. Solvent physically trapped inside a glassy solid cannot. A peptide dried from an acetonitrile and water mixture that collapsed during freeze-drying can hold acetonitrile in sealed pockets within the collapsed mass, and a short equilibration on an undissolved suspension will under-report it.

It shows up as a solvent result that rises when the same lot is retested with full dissolution, a longer equilibration, or a higher equilibration temperature. Method development for residual solvents therefore has to demonstrate recovery from the actual matrix, by spiking a known quantity into the powder and showing it comes back, rather than only from a solution standard. ICH Q2 covers that requirement for specificity and accuracy.

• Dissolution before equilibration, not suspension, wherever the peptide and the diluent allow it.
• Spike recovery in the real matrix, reported as a percentage, not calibration against solvent standards alone.
• Equilibration temperature and time stated with the result, because the number moves with both.
• A stated list of which solvents the method looks for, since anything outside the calibration is invisible to it.

What the number does and does not tell

Read together these five say something. Read alone, each one is easy to overstate. A residual solvent panel is a statement about volatiles and about nothing else on that list, and the choice between electrospray and MALDI ionisation changes what the identity line can resolve without changing any of it.

Where our own record stops

Aurum publishes chromatographic purity by reverse-phase HPLC and identity by mass spectrometry. Residual solvent testing is not among the specifications Aurum publishes, and residual solvent panel is not among the specifications Aurum publishes, a trifluoroacetate content figure or a salt form determination for the material we sell. Where a lot was purified as the trifluoroacetate salt, that counter-ion is present in the powder and is not accounted for anywhere in the figures we publish.

Stating that plainly is more useful than implying a specification exists. A missing measurement is a known gap. A figure of unknown provenance is not.

Common questions

Does a high purity figure imply low residual solvent?

No. Reverse-phase HPLC area percent at 214 nm compares peptide-related peaks against each other. Acetonitrile and water are the mobile phase; they are not impurity peaks in that measurement and contribute nothing to the percentage.

Why is acetonitrile limited more tightly than ethanol?

Because ICH Q3C places acetonitrile in Class 2 on toxicological grounds and ethanol in Class 3. The Option 1 limits, 410 ppm against 5000 ppm, follow from that classification rather than from how hard each is to remove.

Can freeze-drying remove all the acetonitrile?

Not reliably. Acetonitrile and water form a system whose composition shifts during freezing, and solvent held in a collapsed or partially collapsed solid has no open path to the vapour phase. Secondary drying reduces it; nothing guarantees it reaches zero.

Is trifluoroacetate detectable by mass spectrometry?

Its presence can be seen in negative-ion mode, but the peptide mass measurement itself is made on the peptide ion and is unaffected by which anion was paired with it. Quantifying the counter-ion needs ion chromatography or fluorine-19 nuclear magnetic resonance.

Do Class 3 solvents need to be measured at all?

ICH Q3C allows Class 3 solvents to be controlled by a loss-on-drying limit rather than individual quantitation, provided the total stays under the general ceiling. Which approach was taken is part of the specification and is worth stating rather than assuming.