Green extraction: SC-CO2, DES, ultrasound

Classical extraction chooses a bulk solvent for brute solvency and pays for it twice — in the solvent itself and in the selectivity it lacks. The three methods collected here make the dissolving force itself a designable variable. That is their common shape, and each pays a different bill for it: supercritical CO2 pays in pressure, deep eutectic solvents in viscosity and solvent recovery, ultrasound in energy distribution. None is automatically greener; each is greener exactly where its bill is affordable.

SC-CO2: solvency from density

Above the critical point — 31.1 °C and 7.38 MPa for carbon dioxide — liquid and gas become one phase, and its density varies continuously with pressure. Solvent power tracks density: near the critical point a modest pressure change alters solubility dramatically, and at higher pressure the fluid approaches a conventional liquid solvent for nonpolar solutes. This tunability is the selectivity mechanism: raising pressure stepwise extracts the feed in fractions, like a distillation in reverse. The physics helps everywhere else too — gas-like diffusivity and zero surface tension carry the fluid into pores that liquid solvents wet badly. But CO2 is nonpolar; lipids, aromas and pigments dissolve, and polar solutes need an entrainer such as ethanol, at which point part of the solvent advantage is spent. The bill is pressure itself: thick-walled vessels, batch extraction cycles, compression energy — and the thermodynamic subtlety that solubility near the critical point is not monotonic in pressure, so the same feed can extract differently on either side of a crossover.

DES: solvency from hydrogen bonding

A deep eutectic solvent mixes a hydrogen-bond donor (urea, glycerol, an organic acid) with an acceptor (typically the quaternary ammonium salt choline chloride); the melt collapses to a eutectic far below either component’s melting point — the classic choline chloride–urea 1:2 mixture stays liquid near 12 °C while urea melts at 133 °C and the salt around 302 °C. The liquid is a dense hydrogen-bond network that dissolves phenolics, flavonoids and other polar natural products, and it is cheap, non-volatile and designable: donor, acceptor and molar ratio are solvent-design variables. The bill has two parts. Viscosity runs orders of magnitude above water, so diffusion is slow and the solvent must be thinned with water or warmed. And non-volatility, sold as a virtue, is a recovery problem: what evaporates off a hexane extract cannot be evaporated off a DES, so the solute must be back-extracted, precipitated with antisolvent, or adsorbed — the step the green ledger is apt to forget.

Ultrasound: not a solvent, an amplifier

Acoustic cavitation nucleates microbubbles in a liquid under an intense pressure wave; the bubbles collapse adiabatically, producing microsecond hot spots and, at solid surfaces, microjets that tear cell walls open and slam solvent into the matrix. Diffusion paths shorten and extraction times fall from hours to minutes — but ultrasound dissolves nothing itself; it intensifies whatever liquid is already there. Its bill is uniformity and scale: cavitation lives near the radiating surface, so a large vessel is mostly dead field unless energy is distributed through flow cells or multiple transducers, and transducer surfaces erode under their own mechanism.

The common shape

All three buy a tunable force and pay in a currency set by physics: CO2 in pressure and a limited polarity range, DES in viscosity and the difficulty of getting the solute back out, ultrasound in energy uniformity. Choosing among them is choosing which bill the product can carry — and, in the same spirit as the purification logic in ../downstream-purification-bioseparation/, it is the molecule’s properties, not the method’s fashionable label, that decide.