Counter-Current Chromatography Enables Use of Green Solvents for Productive Peptide Purification Processes

Counter-Current Chromatography Enables Use of Green Solvents for Productive Peptide Purification Processes

Rosella Prestia, Damian Hauri, Mattia Sponchioni, Sebastian Vogg and Thomas Müller-Späth

Separations 2026, 13(7), 195 · Published 4 July 2026
Special Issue: Advanced Separation Media and Technologies for Biomolecules
DOI: 10.3390/separations13070195 · Open access, CC BY

Read the full article Download PDF Supplementary material

Competing interests, as disclosed by the authors: D.H., S.V., T.M.-S. and — for the duration of the work — R.P. are employees of ChromaCon AG, which commercializes the technology described in the paper. The research received no external funding.

Citation (BibTeX)

@article{prestia2026countercurrent,
  title   = {Counter-Current Chromatography Enables Use of Green Solvents
             for Productive Peptide Purification Processes},
  author  = {Prestia, Rosella and Hauri, Damian and Sponchioni, Mattia
             and Vogg, Sebastian and M{\"u}ller-Sp{\"a}th, Thomas},
  journal = {Separations},
  volume  = {13},
  number  = {7},
  pages   = {195},
  year    = {2026},
  doi     = {10.3390/separations13070195}
}

Preparative reversed-phase liquid chromatography is one of the most resource-intensive stages of synthetic peptide manufacturing, and production processes commonly rely on acetonitrile (ACN) with trifluoroacetic acid (TFA). That eluent system is chosen mainly for its high chromatographic resolution — and it is exactly the combination that environmental and safety pressure is now working to remove.

Greener alternatives are well characterized. Ethanol is low-toxicity and bio-derived; dimethyl carbonate (DMC) is an emerging bio-based substitute for acetonitrile; acetic acid is a typical candidate for a green counterion in place of TFA. What has blocked their adoption is not chemistry but process economics: in conventional batch chromatography, they lose too much product.

This highlight summarizes the study, its data, and its limitations, and links to the full open-access article.

1. Why This Study Matters: The Solvent Problem in Peptide Purification

Peptide therapeutics are in a demand surge. More than 80 peptide drugs have received regulatory approval since 2023, with over 200 candidates in clinical development. GLP-1 receptor agonists have become the dominant class in the pharmaceutical market, and the global anti-obesity segment is projected to exceed $100 billion by 2030. That demand has exposed manufacturing bottlenecks in synthesis, purification, and fill-finish.

Purification carries a disproportionate share of the burden. Solid-phase peptide synthesis already runs at a process mass intensity of roughly 3,000–15,000 kg of material per kg of active ingredient for peptides in the 1,000–5,000 Da range. Downstream chromatography then adds several hundred to several thousand litres of eluent per kilogram of purified peptide, of which 10–40% is organic modifier.

Three problems with the standard eluent system

• TFA is a PFAS Trifluoroacetic acid belongs to the per- and polyfluoroalkyl substances and is highly persistent in the environment — the paper describes it as a “forever chemical” and notes its potential toxicity risks. The conclusions point to increasing regulatory and environmental pressure on PFAS in pharmaceutical manufacturing.
• Acetonitrile is flammable and toxic Replacing the organic modifier is therefore one half of the problem, alongside replacing the acid. Ethanol and dimethyl carbonate are the substitutes tested here.
• The dilution prevents recycling Because the organic modifier is so heavily diluted, the eluent cannot practically be recycled. The paper links that volume directly to waste generation, environmental impact, and limits on process throughput.

The obstacle to switching has been consistent and well documented: replacing ACN/TFA with green alternatives reduces chromatographic resolution, and reduced resolution in a batch process means lower yield and lower productivity. Until now, no industrially scalable process had combined high purity, high yield, and a low environmental footprint.

2. The Study at a Glance

Researchers at ChromaCon AG and Politecnico di Milano purified two therapeutic peptides using eluent systems containing no acetonitrile and no TFA, first as conventional batch processes and then as continuous multi-column counter-current solvent gradient purification (MCSGP) processes. Method development held the column, the 10 CV gradient volume and the 10 g/L_resin loading fixed, so that the batch comparisons isolate the eluent system; the MCSGP processes were then built from those batch methods.

The two model peptides

Both molecules are conventionally purified with ACN and TFA, and both were run against an optimized ACN/TFA batch process as the reference. Together they bracket the range of therapeutic peptides — a heavily engineered modern construct and a long-established sequence — while sharing the same downstream challenges.

3. Where the Yield Goes

The clearest way to understand the result is to look at how the product pool is defined in each of the three processes. The schematic below redraws the Tirzepatide elution profiles reported in the study; the shaded area is the material actually collected.

4. The Green Solvent Penalty, Quantified

Both peptides were first optimized as conventional batch processes using a common gradient-narrowing procedure, so the green and conventional batches are directly comparable. The results reproduce, with numbers, the trade-off that has kept green eluents out of production.

Tirzepatide with ethanol/acetic acid

Under ACN/TFA, the highest purity appeared along the tail of the main peak and a wide collection window satisfied the 89.0% specification, recovering 81.9% at 5.3 g/L_resin/h. Under ethanol/acetic acid the peak took on a double-peak shape, with high purity only at the front and progressive co-elution with closely related impurities after it. The usable window narrowed sharply and yield fell to 43.2%, with productivity down to 2.9 g/L_resin/h.

Tetracosactide with dimethyl carbonate/acetic acid

The ACN/TFA batch produced a sharp product peak with fractions above 95.0% purity and a maximum in-specification yield of 82.8%. The DMC/AcOH batch produced a very narrow collection window followed by a long tail that was high in purity but low in product concentration. The pool consisted of two fractions, giving 92.3% pool purity at 35.3% yield — a 2.3-fold reduction against the conventional batch — and productivity of 1.6 g/L_resin/h.

The penalties compound rather than trade off. Lower recovery — and, for Tetracosactide, a longer process time from the added counterion exchange — reduced productivity by a factor of two and three respectively, and solvent consumption moved the wrong way at the same time — roughly doubling for Tirzepatide and rising more than three-fold for Tetracosactide. Reducing the loading density would recover some yield, but at the cost of still lower productivity and still more solvent. In batch operation there is no setting that delivers all three.

5. How MCSGP Closes the Gap

MCSGP — multi-column counter-current solvent gradient purification — is a continuous, twin-column form of gradient chromatography. It does not improve the resolution of the separation. It changes what happens to the fractions that resolution fails to resolve.

The operating principle

• Two columns, one repeating switch While one column is eluting, the other is being loaded. The roles reverse at each switch, so neither column sits idle and the process runs continuously.
• The overlap regions are recycled, not discarded The elution is separated into weakly adsorbing impurities, the product fraction, and strongly adsorbing species. The two overlap regions between them — product co-eluting with weak impurities, and product co-eluting with strong impurities — are returned to the loading column instead of being discarded.
• Inline dilution keeps recycled material on the column For Tetracosactide, a 4.5-fold inline dilution was used to ensure the recycled material re-adsorbed completely on the downstream column. For Tirzepatide, a 4-fold dilution served a second purpose: preventing precipitation or gelation during the collection and recycling phases.
• The process converges to a cyclic steady state Recycled material accumulates until the internal concentration profile stabilizes. Here that took three cycles; from cycle four onward the chromatograms overlaid almost exactly, and cycles 4 and 5 were used for the performance calculations.

Holding the cut points: UV-based dynamic control

The switch points were set from the live UV signal rather than from elapsed time, using AutoPeak process control. For the Tirzepatide process, a first threshold at 15 mAU opened weak-side recycling, a second at 200 mAU started product collection, and a fall to 75% of the peak maximum — plus a 20-second delay — began strong-side recycling. Because the cut points track the actual peak, the process maintained reproducible cyclic steady-state operation despite variations in the elution profiles.

Applied to the green eluent systems, that mechanism recovered what the batch processes had been discarding. The ethanol/acetic acid MCSGP process reached 88.1% yield at the 89.0% purity specification and 6.3 g/L_resin/h — higher yield and higher productivity than the conventional ACN/TFA batch, using 3.4 L/g of solvent instead of 4.3 L/g. The DMC/AcOH MCSGP process for Tetracosactide, which used weak-side recycling only, reached 93.8% yield at 94.0% purity, exceeding the conventional batch on both.

6. Full Results: Process Comparison

The table below reproduces the study’s process comparison. For the batch processes, the pool with the maximum achievable yield within specification is reported; for MCSGP, the values are averages of steady-state cycles 4 and 5.

Swipe the table sideways to see all four columns.

Source: Table 3, Prestia et al., Separations 2026, 13, 195 (CC BY). Green values improve on the conventional ACN/TFA batch benchmark; red values fall short of it.

Experimental conditions

7. What This Means for Peptide Manufacturing

Sustainability and process performance have generally been treated as a trade-off in peptide downstream processing. The practical contribution of this study is to separate them into two independent levers — the eluent chemistry and the operating mode — and to show that the second can pay for the first.

For Tirzepatide, the green continuous process was simply the better process on every measured axis: higher yield, higher productivity, and lower solvent consumption than the conventional batch, while eliminating both TFA and acetonitrile. For that molecule at least, removing a PFAS from the purification step did not have to be traded against process performance. Because eluent consumption is directly associated with the overall waste of the step, and the dilution of the organic solvent prevents recycling, lower consumption means less waste.

The authors also argue that the scalability and operational robustness of MCSGP with AutoPeak support implementation at manufacturing scale, as previously demonstrated for complex biopharmaceutical products such as oligonucleotides, and that the technique has been characterized for process qualification. In that context they identify solvent viscosity and the associated pressure drop as the main additional consideration when using green eluents.

8. Key Terms in This Study (Glossary)

9. Frequently Asked Questions