EU restrictions on the reproduction-toxic solvent DMF took effect December 12, 2023, forcing peptide manufacturers to replace it in solid-phase peptide synthesis. Binary DMSO/EtOAc mixtures deliver crude purity comparable to DMF on flexible synthesizers, NBP works only at elevated temperatures, and…
EU restrictions on the solvent N,N-dimethylformamide DMF took effect on December 12, 2023. DMF is classified as a reproduction-toxic CMR agent , meaning carcinogenic, mutagenic, or toxic for reproduction, and the regulation removes it from routine peptide manufacturing in Europe. The question that follows is a practical one. Which solvent systems can replace DMF in solid-phase peptide synthesis SPPS without sacrificing purity, and how much of the environmental problem does changing the synthesis solvent actually solve?
The working answer has two parts. Binary mixtures of dimethyl sulfoxide and ethyl acetate DMSO/EtOAc have produced crude purities comparable to or better than DMF in manufacturer tests on modern synthesizers, making them the most immediately usable substitution. The single-solvent alternative abbreviated NBP , written NBPP in one passage of the source material, also works, but only at elevated temperatures, which restricts the sequences it can handle. The second part of the answer is that synthesis solvents are only half of the footprint. Reversed-phase HPLC purification using acetonitrile MeCN and water consumes large solvent volumes at research scale and far larger ones at commercial scale, and MeCN itself is a moderately toxic solvent. A defensible green strategy has to address synthesis and purification together.
The industry context explains the urgency. With more than 50 peptide drugs on the market, roughly 170 in clinical trials, and another 200 in preclinical development, reliance on hazardous legacy synthesis methods has been flagged by the ACS Green Chemistry Institute as a critical need for greener processes PMID 30900880 . A review of the green solid-phase synthesis literature confirms that current methods depend heavily on toxic solvents, DMF, dichloromethane, and N-methyl-2-pyrrolidone among them, and that the field has only begun to map workable alternatives PMID 30681290 . The EU restriction turns that slow-moving research priority into a compliance deadline.
DMF earned its place in SPPS because it does several jobs at once. It dissolves Fmoc -amino acid building blocks and coupling reagents, swells polystyrene resins so the growing chain stays accessible, and keeps the peptide solvated during coupling. That combination is hard to reproduce with any single solvent.
The evaluative framework used across this field is the 12 principles of green chemistry formulated by Paul Anastas and John Warner. The principles give manufacturers criteria for judging whether a substitute is genuinely safer than what it replaces. Four are safety-related and bear directly on solvent choice: principle 3, less hazardous chemical syntheses; principle 4, designing safer chemicals; principle 5, safer solvents and auxiliaries; and principle 12, inherently safer chemistry for accident prevention. Judged against them, DMF fails principle 5 because of its toxicity and principle 12 because of the handling hazards it creates. A replacement is expected to score better on those axes without introducing new problems elsewhere in the framework.
The other principles complete the framework: prevent waste 1 , atom economy 2 , energy efficiency 6 , renewable feedstocks 7 , reduce derivatives 8 , catalysis 9 , design for degradation 10 , and real-time analysis for pollution prevention 11 . Two of them, atom economy 2 and reduced derivatives 8 , come back at the end of any serious attempt to green Fmoc chemistry, because no solvent change can fix the inherent inefficiency of excess activated amino acids or the byproducts of protection and deprotection cycles.
Solvent polarity and viscosity are the two physical properties that decide whether an alternative can function in place of DMF. Polarity determines whether the solvent dissolves the activated amino acid and swells the resin. Viscosity determines how fast reagents diffuse into the polymer matrix and how efficiently washes remove soluble byproducts. A solvent that is too viscous slows mass transfer and extends coupling times. One that swells the resin poorly starves the growing chain. The published work on DMF alternatives treats polarity and viscosity as the primary screening criteria Lopez et al., Org. Process Res. Dev. 2018, 22, 494 .
The regulatory asymmetry is worth stating plainly. The EU restriction binds manufacturers operating in Europe, but DMF is not banned worldwide. European producers must therefore meet a greener standard than competitors elsewhere, and any laboratory supplying European markets must design around the restriction. The restriction is now in force, it targets a solvent with few genuinely equivalent replacements, and it is not going away.
The most consequential substitution data come from a manufacturer application note that tested binary DMSO/EtOAc mixtures against DMF on three model peptides. On the PurePep Chorus synthesizer, the mixtures produced crude purities similar to or superior to the DMF-based protocol, according to Gyros Protein Technologies. The same company reports that its Symphony X parallel synthesizer has run binary DMF-free mixtures successfully. Both results matter because they show the substitution working on instruments whose fluidics had to handle a solvent mixture with very different physical properties from DMF.
These data come from the manufacturer's own application note, and that provenance sets their limits. Three model peptides is a thin basis for generalization, and the source does not disclose the identities or the sequences of those peptides. Long, aggregating, or otherwise difficult sequences are exactly where solvent effects bite hardest, and the public record does not establish that DMSO/EtOAc handles them as well as DMF. The claim of similar or superior crude purity should be read as applying to the tested peptides, not as a blanket license for all of peptide chemistry.
The temperature constraint on the solvent abbreviated NBP is the finding that matters most for instrument planning. NBP works as a DMF replacement, but only when the synthesis is run at elevated temperatures. That qualification rules it out for temperature-sensitive sequences and for coupling steps that degrade when heated. It also means that adopting NBP requires a synthesizer with reliable temperature control, not just solvent compatibility. The abbreviations NBP and NBPP appear without expansion in the source material, a documentation gap for anyone trying to source the solvent.
Instrument design emerges from these data as a variable in its own right. The application note argues that instrument design and flexibility are essential for adopting new solvents and solvent mixtures, and the physical reasoning supports the claim. DMSO/EtOAc mixtures require pumps and valves that tolerate viscous DMSO, lines that resist EtOAc, programmable mixing so the ratio can be tuned, and enough wash volume to clear the system between steps. A synthesizer built around DMF's low viscosity and ambient-temperature operation will struggle with all of these. Solvent choice and instrument choice are coupled decisions, not independent ones.
A green synthesis that ends in a conventional HPLC purification step has only moved the problem downstream. Reversed-phase HPLC using MeCN/water eluents is a major ecological contributor to peptide manufacturing for two compounding reasons. MeCN is moderately toxic under green chemistry principle 5, safer solvents. And the volumes are enormous: preparative HPLC at research scale consumes large quantities of solvent, and at commercial scale the waste stream grows with every column load. Principle 1, prevent waste, is the criterion that HPLC fails.
Catch-and-release purification , also called orthogonal purification, is the alternative receiving the most attention. The peptide is bound to a functionalized support and released in a concentrated volume, avoiding the large gradient volumes of reversed-phase HPLC and cutting both solvent consumption and waste generation. The manufacturer's version of this approach is the PurePep EasyClean, and the broader concept appears in the published discussion of DMF-free synthesis cited above. The mechanism is genuinely different from HPLC, and the direction of the environmental benefit is clear.
What is not clear is the size of the benefit. The evidence reviewed here does not quantify liters of solvent saved per gram of peptide for catch-and-release versus HPLC. That absence matters, because manufacturers need numbers to build a waste-reduction case. The analytical side of peptide chemistry shows the same pattern: progress toward greener methods for absolute quantification exists, but challenges in sample preparation and chromatography remain PMID 25864956 , and analytical HPLC is a small fraction of the preparative solvent bill.
The holistic point is that crude purity and final purity are different targets. A DMF-free synthesis can deliver high crude purity, and some aqueous synthesis studies report single-peak HPLC chromatograms PMID 35969667 , suggesting that greener chemistry can reduce the purification burden. But high crude purity does not guarantee that purification can be skipped for a given application. Synthetic route and purification method must be designed as one system, with the quality specification of the final product as the fixed point.
The manufacturer data on DMSO/EtOAc and NBP sit alongside a peer-reviewed literature that is smaller but instructive. An industrial-perspective review of greening peptide therapeutics focuses on replacing and reducing problematic solvents at manufacturing scale and surveys emerging production technologies PMID 35516773 . That review is consistent with the direction of the vendor claims, which raises confidence that the substitution is chemically plausible even if the specific three-sequence results are not independently confirmed.
More radical alternatives exist at the research frontier. One experimental study demonstrated water-based solid-phase peptide synthesis without hydroxy side chain protection, using the coupling agent DMTMM under microwave heating. The researchers made a laminin-related peptide and acyl carrier protein 65-74 ; for the laminin-related peptide they reported no significant O-acylation or racemization, as judged by a single-peak HPLC chromatogram PMID 35969667 . Two model peptides is again a narrow demonstration, but the result shows where this research program ends: not swapping one dipolar aprotic solvent for another, but eliminating the toxic solvent class entirely.
Coupling chemistry is evolving in the same direction. A review of the past decade of racemization-free coupling reagents catalogues alternatives, including ynamides and tantalum or niobium alkoxides, that lower cost and toxicity without introducing racemization PMID 38775347 . These reagents address a different green-chemistry gap than solvents do. Solvent policy drives the switch away from DMF; reagent design determines how much waste the activation step itself generates.
Against the full framework, two gaps stand out. Principle 2, atom economy, remains a weakness of Fmoc chemistry, which typically uses excess activated amino acid to push couplings to completion. Principle 8, reduce derivatives, is equally unresolved, because Fmoc synthesis depends on protection and deprotection cycles that generate byproducts at every step. No solvent substitution fixes either gap. Improved atom economy and reduced derivatization remain open process-chemistry problems rather than solved ones.
The actionable conclusions from this evidence are as follows. Evaluate binary DMSO/EtOAc mixtures as the primary DMF replacement; they have the strongest reported support, they run on standard flexible synthesizers, and their components are widely available. Audit the instrument before committing to a solvent. A synthesizer must handle the viscosity of DMSO, the chemistry of EtOAc, programmable mixing of binary systems, and, if NBP is under consideration, reliable elevated-temperature operation. Treat NBP as a conditional option that applies only where the synthesis tolerates heat.
Design the purification step together with the synthesis. Measure the current liters of MeCN and water per gram of purified peptide as the baseline, then test catch-and-release purification against it. The qualitative waste reduction is clear in the literature; the quantitative case has to be made with local numbers. Do not assume that improved crude purity removes the need for purification. Application-specific purity requirements decide that, not the crude HPLC trace.
| Solvent system | Reported performance | Temperature requirement | Evidence base | Principal limitation |
| --- | --- | --- | --- | --- |
| DMF | Reference standard | Ambient | Legacy practice | CMR-restricted in the EU since December 12, 2023 |
| DMSO/EtOAc mixtures | Crude purity similar to or better than DMF | Ambient | Manufacturer application note, three model peptides | Sequences undisclosed, not independently replicated |
| NBP NBPP | Functions as a DMF replacement | Elevated temperatures required | Manufacturer studies | Unsuitable for temperature-sensitive sequences, abbreviation undefined |
| Water with DMTMM and microwave heating | Successful syntheses; single-peak HPLC for the laminin-related peptide | Microwave heating, specialized equipment | Peer-reviewed study, two peptides PMID 35969667 | Only two peptides demonstrated |
The table makes a point that is easy to miss in the regulatory noise: every available alternative rests on a small evidence base. The strongest, DMSO/EtOAc, rests on three undisclosed peptides. The peer-reviewed water-based alternative rests on two. Until comparative studies report crude purity, yield, and racemization data across a panel of difficult sequences, adoption decisions will carry more uncertainty than they should. The manufacturer reports that DMSO/EtOAc and NBP are already being used on flexible instrument platforms in academic and industry settings; that adoption is real, but it does not substitute for independent comparative data. Monitoring the process-chemistry literature is therefore part of the job. Water-based synthesis PMID 35969667 , new coupling reagents PMID 38775347 , and catch-and-release purification are all active research fronts, and manufacturers who build solvent flexibility into their instruments now will be able to adopt better systems as they appear.
Several questions remain open, and they are worth listing because a researcher or buyer should know where the uncertainty sits. The identities of the three model peptides in the DMSO/EtOAc application note are undisclosed, so nothing can be said about transfer to aggregating, long, or hydrophobic sequences. The abbreviations NBP and NBPP are used without definition, which is anomalous for a regulation-driven substitution. The quantitative solvent and waste savings of catch-and-release purification relative to HPLC are not established in the public record. And the deeper green-chemistry goals, atom economy and reduced derivatives in Fmoc chemistry, remain unmet by any solvent change.
The limits of the vendor evidence deserve emphasis on their own. The central substitution claims, similar-or-better crude purity for DMSO/EtOAc and the temperature ceiling for NBP, come from a single manufacturer's application note and studies. They are reported results, not peer-reviewed comparative trials, and no independent replication is publicly available. That does not make them wrong. The direction is consistent with the broader published push toward replacing toxic solvents in peptide manufacturing PMID 35516773; PMID 30681290 . But the strength of the evidence matters when a manufacturer stakes a validated process on it.
What the EU restriction has changed is the default. Before December 12, 2023, DMF was the path of least resistance. After that date, in Europe, the question is no longer whether to leave DMF but which route to take, and the honest answer is that the route depends on sequence class, instrument capability, and purification load. DMSO/EtOAc mixtures are the most immediately usable option. Catch-and-release purification is the most direct way to cut the downstream solvent bill. Neither is a complete solution. The unresolved chemistry of atom economy and derivatization will require process innovation after the solvent question is settled, and the peptide industry's shift to greener synthesis will be judged by whether that second wave of work happens.
Related reading: Fluorescent Labeling of BeKm-1 Retains Specificity and Affinity, AIEX-SEC with WorkBeads 40Q and Macro SEC for Virus Purification, Choosing a peptide synthesizer: 5 factors every lab should weigh, Five Reasons Crude Peptides Are Not Enough for Screening.