SFC Points to Greener Path for Chromatography

Green chemistry has been reshaping how laboratories operate, and few areas are feeling that pressure more than liquid chromatography (LC), which traditionally leans on solvents like acetonitrile and methanol. Both are effective, but both come with real downsides: acetonitrile is toxic, and prone to price swings, while methanol carries its own environmental and safety concerns. As green chemistry principles become standard practice in laboratories worldwide, researchers are increasingly asking whether these workhorse solvents can be swapped out for greener alternatives, and whether that swap can hold up beyond simple, best-case laboratory conditions.

That's where supercritical fluid chromatography (SFC) enters the conversation. Because SFC relies primarily on supercritical carbon dioxide rather than large volumes of organic solvent, it's already considered one of the eco-friendlier separation techniques available. But there's a catch: most SFC methods still lean on methanol as a co-solvent, which chips away at the technique's green credentials. The real question, then,becomes: Can that methanol be replaced too, and how does SFC compare to conventional liquid chromatography when greener solvents are put to the test?

LCGC International spoke with Karine Faure, one of the researchers behind a new study that tackles this question head-on, comparing six alternative solvents, including ethanol, isopropanol, dimethylcarbonate and the lesser-known bio-based solvent dihydrolevoglucosenone (Cyrene), across both reversed-phase LC and SFC. Using phenolic compounds as a testing ground, Faure’s team set out to see which solvents hold their own against acetonitrile and methanol, and why SFC might be the more promising path forward for sustainable separations. The study has inspired a paper recently published in the Journal of Chromatography A.¹

Given that supercritical CO₂ is the primary eluent, how much does the choice of organic co-solvent (methanol vs. ethanol, isopropanol, dimethyl carbonate (DMC), propylene carbonate, or Cyrene) actually shift supercritical fluid chromatography (SFC)'s overall "eco-sustainability" score once solvent toxicity, energy of production, and required co-solvent percentage are all factored in together?

The most promising alternative to MeOH in SFC is EtOH as a co-solvent. With similar retention times and peak widths, this replacement uses the same volume of co-solvent (up to 50% in the overall mobile phase) but with a solvent that is five times less toxic, a significant advantage for both SFC users and the environment. Additionally, EtOH can be sourced as a bio-based solvent, further enhancing its eco-friendly profile. However, this comes with a drawback: EtOH is approximately five times more expensive than MeOH.

DMC has shown promise as a modifier in chiral SFC—does its performance (miscibility with CO₂, elution strength, peak shape) translate similarly to achiral separations of mid-polarity neutral solutes like phenolics, or are the retention mechanisms too different?

DMC is a highly promising co-solvent for SFC due to its safety and low cost. It maintains the same elution order for phenolic compounds as MeOH, with comparable selectivity. However, its low elution strength prevents the elution of highly retained compounds. To address this limitation, mixtures with EtOH could be considered to enhance its elution power.

Propylene carbonate (PC) has a much higher viscosity than methanol or ethanol—what backpressure and column-temperature adjustments are needed to keep SFC systems operating within safe pressure limits when using it as a co-solvent, and does this erode any green-solvent advantage through higher energy consumption?

The high viscosity of PC generates excessive backpressure, limiting its content to 50% in CO₂ at a flow rate of 1000 mL/min, 40 °C, and 140 bar BPR. Reducing the flow rate to 0.5 mL/min did not resolve this issue. While increasing the temperature would reduce viscosity, it would also significantly increase energy consumption and require reoptimization of the chromatographic method. For example, pure EtOH has a viscosity of 0.89 cP at 40°C, whereas PC only reaches this value at 80 °C, a temperature that is harsh on both columns and analytical systems, eroding its green-solvent advantage.

Cyrene is a relatively novel, more viscous, and chemically reactive (ketone-containing) solvent—has it been characterized for miscibility, chemical stability, and detector compatibility under SFC conditions, and could it undergo degradation or side reactions in the presence of compressed CO₂?

Although Cyrene is increasingly promoted in organic chemistry, its high viscosity (over 14 cP) prevents its use as a pure co-solvent. When mixed with EtOH (10% to 50% Cyrene in EtOH), it became usable as a co-solvent. However, the resulting mobile phase was incompatible with both UV and MS detection systems, preventing further experimentation.

Since phenolic biomass-derived compounds show orthogonal separation behavior between reversed-phase liquid chromatography (RPLC) and SFC with a polar stationary phase, how does swapping methanol for a greener co-solvent in the SFC dimension affect that orthogonality—does selectivity remain complementary to RPLC, or does it converge?

In our publication, we demonstrated that replacing methanol with ethanol in the SFC dimension does not affect the elution order. Consequently, SFC using EtOH remains complementary to RPLC using EtOH, preserving the desired orthogonality.

Among ethanol and isopropanol as methanol replacements, how does increasing alkyl chain length/branching affect solute retention, peak efficiency, and enantioresolution in chiral SFC—are there diminishing returns or trade-offs (e.g., slower diffusion, lower polarity range accessible)?

This research did not include chiral experiments. Our preliminary screening indicates that EtOH is a straightforward and effective replacement for MeOH in the separation of phenolic compounds. We observed identical retention times (correlation coefficient R² = 0.9989) and similar peak widths (average 0.95 s for MeOH and 1.02 s for EtOH). For this specific application, there is no justification for continuing to use methanol. These results should encourage SFC users to seriously consider EtOH as a co-solvent candidate for broader applications.

Of course, further fundamental research is needed to explore the kinetic behavior of isopropanol and EtOH in SFC.

Do these alternative co-solvents interact differently with common SFC stationary phases (e.g., polar, diol, or chiral selectors) compared to methanol, potentially requiring re-optimization of phase selection when switching to greener modifiers?

These preliminary experiments were conducted using 1-aminoanthracene and diol stationary phases. Changing the physicochemical characteristics of the solvent is expected to impact the interactions promoted by various stationary phases, and specific studies should be carried out to investigate this further. The possibilities remain open.

Since the current work uses phenolic model compounds, do the elution strength and selectivity patterns observed with these six alternative solvents hold up when applied to structurally more complex or higher-polarity biomass matrices, such as full lignin depolymerization mixtures?

Given the different interactions that large molecules can have with stationary phases compared to smaller ones, we expect the differences between modifiers to be more pronounced. This is currently under investigation.

If both dimensions of an RPLC×SFC system used alternative green solvents (e.g., Cyrene or DMC in SFC paired with ethanol in RPLC), would the resulting 2D peak capacity and orthogonality match or exceed conventional methanol/acetonitrile-based systems?

By selecting a green solvent system that provides selectivity like conventional solvents, it is possible to maintain the same orthogonality, sometimes even slightly higher, as demonstrated in our example with water/DMC in ¹LC and CO₂/EtOH in ²SFC . By nature, RPLC and SFC with polar stationary phases already offer high orthogonality for small neutral molecules, thanks to their distinct interaction types. While replacing conventional solvents with green alternatives is a straightforward win in SFC, the situation is more nuanced in LC, where green solvents tend to increase peak widths, thereby reducing theoretical peak capacity. From a practical point, the separation space provided by 2D mode may accommodate this broadening, but critical separations could still be compromised.

Beyond performance metrics, what practical hurdles—co-solvent cost, supply consistency, instrument seal/gasket compatibility, detector response—currently limit broader adoption of solvents like propylene carbonate or Cyrene in routine SFC method development?

Propylene carbonate and Cyrene are effectively ruled out due to their incompatibility with LC/SFC detection systems. The use of other green solvents also faces practical limitations. In LC, the high viscosity of green mobile phases may restrict analysis speed. While this is not an issue for 2D configurations where LC is operated as a first dimension at very low flow rates, it must be considered for routine LC analysis. Additionally, combining green solvents with water can lead to solubility issues, as seen in the DMC/H₂O system. This can be mitigated by using a bridging solvent such as EtOH, but this introduces additional complexity to the method development.

On the other hand, substituting conventional solvents with green alternatives in SFC is straightforward. There are no solubility or viscosity issues due to their full miscibility with CO₂. However, special attention must be paid to SFC-MS hyphenation, as the need for an adequate make-up solvent may sometimes offset some of the green benefits.

Finally, solvent suppliers should provide cheaper, MS-grade green solvents. By demonstrating that a wide range of applications can accommodate solvent substitution, we may convince them to invest in this direction.

Reference

1. Gregson, M.; Batteau, M.; Faure K. Alternative Solvents in Liquid Chromatography and Supercritical Fluid Chromatography for the Separation of Phenolic Compounds. J Chromatogr A 2026, 1785, 467240. DOI: 10.1016/j.chroma.2026.467240