DMF, DMSO and NMP compared for organic synthesis
Technical article · Eapearl Chemical ·
These three solvents occupy the same shelf in most laboratories and are often treated as alternatives. They dissolve similar things for similar reasons, and then they behave differently in every respect that decides whether a process works.
What they have in common
All three are dipolar aprotic: strongly polar, able to dissolve ionic material, and unable to donate a hydrogen bond. That combination produces the effect these solvents are bought for. A dissolved salt has its cation wrapped in solvent while its anion is comparatively naked, and a naked anion is a much better nucleophile than a solvated one. Nucleophilic substitutions, aromatic displacements, eliminations and many couplings therefore run faster and cleaner here than in protic media.
They also share the properties that make them awkward. Each is miscible with water, each boils far above the everyday solvents, and each is hygroscopic to a degree that matters when a reaction is sensitive to water. Those three facts together determine the work-up, the recovery and much of the cost.
Three molecules, three characters
| DMF | DMSO | NMP | |
|---|---|---|---|
| Name | N,N-dimethylformamide | Dimethyl sulfoxide | N-methyl-2-pyrrolidone |
| CAS | 68-12-2 | 67-68-5 | 872-50-4 |
| Formula | C3H7NO | C2H6OS | C5H9NO |
| Molar mass | 73.09 g/mol | 78.13 g/mol | 99.13 g/mol |
| Polar group | Amide carbonyl | Sulfinyl | Cyclic amide carbonyl |
| Signature weakness | Releases dimethylamine on decomposition | Exothermic decomposition with acid or base | Highest boiling of the three, hardest to remove |
| Typical home | Fibre and polyurethane solutions, substitutions | Oxidations, displacements, dissolution of difficult salts | Engineering polymers, electrode slurries, stripping |
DMF: the amide that will not stay intact
DMF is the longest-established of the three in industrial use and remains the reference solvent for polyacrylonitrile and polyurethane solution processes. Its defining liability is that the amide bond hydrolyses and decomposes, slowly at room conditions and much faster when heated or when acid or base is present, releasing dimethylamine and carbon monoxide.
Dimethylamine is the part that damages chemistry. It is a strong nucleophile and a base, so it competes in exactly the substitution reactions DMF is used for, attacks acid chlorides and anhydrides, and can appear as an unexplained impurity in a product whose synthesis contains no amine. When a reaction in DMF gives a variable yield or an impurity that tracks with solvent age, the amine content of the solvent is the first thing to measure. Buying to a specification that limits free amine, and keeping the solvent dry and cool, addresses most of it.
The related solvent N-methylformamide is a different material with a free amide hydrogen, which makes it protic and changes its behaviour substantially; it is not a substitute for DMF and should not be treated as one.
DMSO: the strongest solvent and the most demanding to heat
DMSO dissolves an unusually wide range of salts and polar organics, and in many displacement reactions it is simply the fastest of the three. It is also the one with a genuine process safety question attached. DMSO decomposes exothermically, and acids, bases and several common contaminants lower the temperature at which that begins. Because the decomposition produces gas and releases heat that accelerates it further, a heated closed system is the dangerous configuration. Any heated DMSO process, particularly one containing acid or base, needs thermal stability data on the actual mixture.
Two further properties shape how it is handled. It freezes a little below ordinary room conditions, so drums in an unheated store can solidify and require managed thawing, during which local overheating must be avoided. And it penetrates skin readily, carrying dissolved material with it, which means that glove selection is a solute question as much as a solvent question and that ordinary laboratory gloves are frequently inadequate. It is also strongly hygroscopic, so an open container quietly acquires water that a moisture-sensitive reaction will find.
NMP: the polymer solvent that will not leave
NMP dissolves engineering polymers that defeat the other two, which is why it dominates polyimide processing, fluoropolymer slurries for electrode manufacture and the stripping of cured coatings. Its cyclic amide structure is more robust than DMF against decomposition to an amine, and in that respect it is the better-behaved amide.
Its cost is at the other end of the process. NMP has the highest boiling point of the three and is fully miscible with water, so removal is genuinely difficult: distillation is energy-intensive and slow, while water washing transfers the problem to a large aqueous stream. In any process where the product must be free of residual solvent, the removal step should be designed and costed before the solvent is chosen, because it will often dominate.
Work-up, recovery and the real cost
- Estimate the removal route before selecting the solvent. Water miscibility plus a high boiling point rules out both of the simple options.
- If you intend to wash the solvent out, size the aqueous stream and establish how it will be treated. This is frequently the step that makes a laboratory route uneconomic at scale.
- If you intend to distil, check what prolonged heat does to your product and, for DMF, what it does to the solvent.
- Decide early whether recovery is intended. Recovered solvent is a different material from fresh, with its own impurity profile, and it needs its own specification and acceptance testing.
- Specify water content and acidity on incoming material and check them. In all three cases these are the parameters that quietly change results.
- Include disposal and treatment cost in the comparison. The cheapest solvent per unit is regularly the most expensive per batch.
Choosing between them
Work through the constraints in order of how hard they are to escape. Start with the regulatory position in every market you supply, because a solvent that cannot be used is not a candidate regardless of how well it performs, and the position around the amide solvents has been an active area for some years. Next consider the safety constraint, which for DMSO means the thermal question and for the amides means occupational exposure control. Then test solvency on the actual solute rather than trusting a general ranking. Only then compare reaction performance, and finally compare the full cost including removal, treatment and any recovery.
Running the order in reverse, which is the natural temptation because reaction yield is the most visible number, produces processes that work beautifully in a flask and cannot be built. Specifications, grade options and samples for dimethylformamide, dimethyl sulfoxide and N-methyl-2-pyrrolidone against a named application are available through our contact page.