Acetone versus DMF as the solvent in an acetylene cylinder capacity, carry-over, water sensitivity and why UHP grades use DMF

Dimethylformamide is the solvent in every ultra-high-purity acetylene cylinder KIGL fills, and acetone is the solvent in every industrial one. Ask why and the usual answer is that DMF dissolves more acetylene. It does, a little, but that is not the reason. Per kilogram the two solvents are within 20 % of each other at 1 atm and acetone is marginally ahead at cylinder pressure; the honest capacity gain for DMF is about 18 % in the same cylinder, and it comes from density, not chemistry. The real case for DMF is its vapour pressure, some 65 times lower than acetone's, which means almost no solvent leaves with the gas. This page puts the numbers side by side.

The acetone data are the measurements of Hölemann and Hasselmann (1952–59). The DMF data are mostly unpublished British Oxygen Company measurements as compiled by S. A. Miller in 1965, with the 1 atm column confirmed by Howard, Schoch and Majewski (Petroleum Refiner, 1954) and by Hölemann and Hasselmann's own DMF work. The C.P.I. handbook (Paris, 1968) gives independent coefficients that agree.

Read across the rows and the pattern is clear. On a mass basis the two solvents are close and DMF's small lead at 1 atm vanishes by 15 atm. On a volume basis DMF wins because a litre of DMF weighs 20 % more than a litre of acetone. Since a cylinder is a fixed volume, the practical gain is the per-litre figure: about 18 % more gas in the same shell at the same pressure. That is worth having, and KIGL's ultra-high-purity acetylene benefits from it, but it is not why DMF was chosen.

Why acetone's position in the solvent league table is as modest as it is, and why it survives anyway, is examined in the 300-volumes myth.

The share of the gas leaving the valve that is solvent vapour is set by the solvent's vapour pressure at the cylinder temperature. Acetone boils at 56 °C and has a vapour pressure of about 185 mm Hg at 20 °C; DMF boils at 152 °C and sits at about 2.7 mm Hg, roughly 60 times lower. From the measured partial pressures over acetone solutions, and the same arithmetic for DMF:

An acetone cylinder run down to atmospheric pressure delivers gas that is nearly a quarter solvent vapour, and over half at 40 °C. For welding that is a nuisance and a solvent-economy problem, handled by the residual-pressure rule in acetone carry-over and residual pressure. For a 99.99 % product it is disqualifying: the delivered purity of an acetone cylinder cannot be held constant as it empties, whatever the purity of the gas that went in. A DMF cylinder holds its delivered purity to low residual pressure, which is the whole reason it exists. The grades are described in acetylene grades explained.

Both solvents swell as they load. Acetone adds 1.58–1.63 ml per gram of acetylene dissolved, DMF 1.47 ml per gram at 15 °C, about 8 % less. The pressure relation differs in shape: DMF's volume rises linearly with pressure over 3–22 atm (V = V₀ (1.023 + 0.0431 P) at 15 °C), whereas acetone's has a P² term, so the last few atmospheres of an acetone fill swell the solvent disproportionately. Either way a loaded solvent at 15 atm occupies 1.6 to 1.8 times its unloaded volume, and solvent charge, porous-mass void and filling pressure must be set together, as inside a dissolved acetylene cylinder explains.

The heat of solution is essentially identical: 4.6 kcal/mol for DMF over 25–100 °C against 4.25–4.70 kcal/mol for acetone, which is 684–757 kJ per kilogram of acetylene. Filling a DMF cylinder releases the same heat and needs the same cooling; withdrawing from one chills it just as fast. Nothing about the withdrawal-rate rule changes with the solvent.

It is often said that DMF is far more sensitive to water than acetone. The measurements do not support it. Haiderger, Szebényi and Székely (Magyar Kémiai Folyóirat, 1958) measured the effect of water on acetylene solubility in DMF at 1 atm:

For acetone, Maillard and Rosenthal (Compt. rend. 1952) found that 5 % water by volume costs about 17 % of capacity at 25 °C and 1 atm, and 10 % water about 37 %; DMF at 10 % water loses roughly 26 %. The two solvents are comparably sensitive at low water contents and acetone is worse at high ones. Two real points survive. First, the penalty accelerates, so a cylinder that picks up a few per cent of water from a wet charge, a damp porous mass or a leaking valve loses capacity faster than the water content suggests. Second, DMF is used where the moisture specification of the product is tight, so the same percentage loss costs more. Both solvents must be kept dry; IS 308:2024 limits moisture in industrial dissolved acetylene to 0.25 % by mass.

IS 308:2024, the specification for industrial dissolved acetylene, assumes acetone and sets impurity limits of 0.01 % v/v phosphorus compounds, 0.02 % v/v sulphur compounds, 4 % v/v acetone-insolubles and 0.25 % moisture by mass; details are on the dissolved acetylene standards page. IS 14814:2023 cl 7.2.2 describes the cylinder as holding acetone or other solvent and sets the maximum working pressure at 15 °C, 15.495 atm, for acetone. No Indian standard covers ultra-high-purity acetylene or DMF as a solvent. KIGL therefore labels its UHP grade, minimum 99.99 % v/v in DMF with phosphine and hydrogen sulphide each at 10 ppm or below, as a KIGL specification whose impurity levels sit an order of magnitude inside the IS 308 limits, rather than claiming conformity to a standard that does not address it. The high-purity acetylene grade follows the same logic.