Two drums of solvent can carry the same CAS number, the same GC assay and almost the same boiling point, and still behave differently on the plant floor. The difference usually sits in four lines of the certificate: distillation range, evaporation profile, water figure and trace impurities. A buyer comparing only price per kilogram and one boiling point is choosing a familiar name without knowing what is inside the drum.

A range, not a number

No industrial solvent is a single pure compound in the sense a specification sheet implies. It is a cut. What the sheet should state is a distillation range at a stated pressure — 188 to 192 °C at 760 mmHg, for example — plus assay by GC and non-volatile residue. The initial boiling point says nothing about the last five percent, and the last five percent is what stays behind.

A two percent fraction boiling 20 to 25 °C above the main cut can double the oven time needed to reach a residual-solvent limit, because diffusion of that heavy tail out of a film is slow. In a batch reaction the same tail accumulates in the pot during recovery, and by the sixth campaign the residue colours and the recovered assay drifts.

Two products show why the range matters more than the headline number. Dimethyl sulfoxide boils at 189 °C and is normally specified at 99.9 percent minimum assay with a water limit, because everything below that threshold is water or dimethyl sulfide from decomposition. Sulfolane sits at 285 °C and melts at 27.5 °C — a detail that becomes a logistics problem the first time a drum ships in January without heat tracing.

Evaporation rate is not the inverse of boiling point

Relative evaporation rate is reported against n-butyl acetate as 1.0, and it governs film formation, pot life and residual solvent. Boiling point sets the ceiling but does not define it: molecular weight, hydrogen bonding and heat of vaporisation decide how fast the last part leaves the surface. Water boils at 100 °C and still has a relative rate near 0.3. n-Octane, at 125.7 °C, leaves a plate far faster than propylene glycol methyl ether at 120 °C, because the ether carries a hydroxyl group that holds it in the film and in any hydrogen-bonding resin.

This is why replacing an aromatic hydrocarbon with an oxygenated solvent on the basis of similar boiling points often yields a sticky film and an odour that lingers. Dipropylene glycol has a relative rate below 0.02; it is a retarding agent, not a thinner. Ask for the evaporation band at your drying temperature, not at 25 °C in a fume hood.

Water content: the line that moves after the drum is opened

For moisture-sensitive chemistry the water figure is the specification. N,N-Dimethylformamide is offered at 0.1 percent water maximum, and 0.05 percent for electronic and pharmaceutical use. Both are achievable in a sealed, nitrogen-blanketed drum; neither survives a humid week with the bung loose. Amides and sulfoxides are hygroscopic by design.

Water is not a passive diluent. In DMF it hydrolyses slowly to formic acid and dimethylamine, which is why a drum that smells faintly of fish has already degraded. In a moisture-cure polyurethane it reacts with isocyanate and releases carbon dioxide, and the pinholes in the coating are the visible receipt. In an organometallic step it consumes reagent and shifts the stoichiometry.

Method matters as much as the limit. Coulometric Karl Fischer suits results below roughly 0.1 percent, volumetric titration above it. A sample drawn in a wet syringe returns a number that is precise and wrong. Name the method, the sample point and the maximum on the purchase order.

“Polar” covers three different properties

Dielectric constant, dipole moment and donor number are not interchangeable, and the donor number is the one that changes reaction rates. DMSO has a donor number near 29.8 and DMF near 26.6; both solvate the cation strongly and leave the anion comparatively naked, which is why cyanide displacement runs so much faster in them than in an alcohol. A solvent can be highly polar and a poor anion activator at once.

The property that makes DMSO useful also makes it hazardous: it is an oxidant under the right conditions, and DMSO with sodium hydride at scale has produced well-documented runaway incidents. Where the polarity is needed but the sulfoxide is not, N-Ethyl-2-pyrrolidone gives comparable aprotic behaviour at a boiling point near 218 °C with no sulfur-oxygen bond to donate. N,N-Dimethylpropionamide is the other drop-in candidate, and tetrabutylurea serves as a modifier where a higher-boiling polar component is wanted without adding water or a salt. Regulatory pressure sits alongside the technical case: DMF carries a reproductive-toxicity classification in Europe.

The safety sheet decides how the drum is stored

Dichloromethane has no flash point under normal conditions, often read as safe. Its vapour is about 2.9 times heavier than air, so it collects in pits and reactor bottoms, and it decomposes to phosgene on hot surfaces.

Methanol burns with a flame that is nearly invisible in daylight, while its vapour is heavier than air and its boiling point is 64.7 °C. Static ignition during drum transfer is a recurring cause of fires later attributed to an electrical fault.

Peroxide formation is the slow hazard. Ethers are the classic case, but unsaturated terpenes are worse in some ways: D-limonene autoxidises to hydroperoxides and carvone, so peroxide value belongs on the specification next to inhibitor content. Limonene is also a skin sensitiser with no aromatic content, so it cannot replace xylene in a resin that needs aromatic solvency, whatever the marketing says.

Grade follows the downstream process

Glycol ethers show how grade and isomer distribution change performance. Dipropylene glycol methyl ether and Dipropylene glycol propyl ether share a backbone but not a water sensitivity: the propyl ether is markedly more hydrophobic, which is why it appears in coatings that must resist humidity. Diethylene glycol monoethyl ether is fully water-miscible and slow, useful as a tail solvent and levelling aid in inks, and a liability if the oven cannot drive it off. Commercial DPG is an isomer mixture, and the ratio shifts coupling efficiency from lot to lot.

Ester solvents carry a different question. Dimethyl glutarate is sold as a low-VOC alternative to glycol ethers and performs well, but the ester bond hydrolyses in humid air and in the presence of base; acid value is the specification that tells you whether the drum is still fit for use.

The quotation request should therefore ask for assay and distillation range at a stated pressure; relative evaporation rate at process temperature; water by Karl Fischer with the method named; acidity or total acid number; colour in Pt-Co; peroxide value and inhibitor for ethers and terpenes; residual odour by headspace GC for consumer-facing products; and packaging with nitrogen blanket. Six lines on a specification prevent most batch failures that get blamed on the reactor.