Written and maintained by CASRAI Editorial Board
Last updated
Supercritical fluid chromatography (SFC) uses carbon dioxide held above its critical point (31.1°C, 73.8 bar / 7.39 MPa) as the primary mobile phase, usually blended with a polar organic modifier such as methanol or ethanol. It is not a replacement for HPLC across the board — for a routine reversed-phase achiral assay, HPLC (or UHPLC) remains the simpler, cheaper, better-supported default. SFC pulls ahead in three specific situations: chiral separations, preparative-scale purification where solvent handling dominates cost and cycle time, and methods that need normal-phase-like selectivity without normal-phase HPLC’s practical headaches. This guide covers each case, and the instrument constraints — back-pressure regulation, detector compatibility, sample solubility — that determine whether SFC is actually the right call for a given method.
Why the Mobile Phase Changes the Physics
Supercritical CO2 behaves like a hybrid of a liquid and a gas: liquid-like density (so it dissolves and elutes analytes) with gas-like viscosity and diffusivity. In practice, CO2-based mobile phases run roughly an order of magnitude less viscous than typical HPLC mobile phases, and analyte diffusion coefficients are correspondingly higher. Two consequences follow directly:
- Higher usable flow rates without a proportional back-pressure penalty — SFC methods routinely run several times the flow rate of an equivalent HPLC method on the same particle size without blowing past the system’s pressure ceiling.
- Faster mass transfer between mobile and stationary phase — narrower, more efficient peaks at a given flow rate, which is what makes SFC’s throughput advantage real rather than just a flow-rate trick.
Because pure CO2 is essentially non-polar, most real SFC methods add a polar co-solvent (modifier) — typically methanol, ethanol, or isopropanol, often at 5–40% — to bring the mobile phase’s elution strength into range for the analyte. The modifier percentage is the primary method-development lever, playing the same role %B does in reversed-phase HPLC gradient method development, but starting from a much less polar baseline.
Where SFC Genuinely Beats HPLC
1. Chiral Separations
This is SFC’s strongest, best-established case. Polysaccharide-based chiral stationary phases (CSPs) — amylose and cellulose derivatives, the same phase chemistries used in normal-phase chiral HPLC — generally perform better under SFC conditions than under either normal-phase or reversed-phase HPLC on the same column:
- Faster analyses at comparable or better resolution — SFC’s lower viscosity lets a chiral separation that takes 20–40 minutes in normal-phase HPLC often run in under 10, because higher flow rates are usable without losing plate count.
- Broader, more predictable modifier screening — the same CSP column can be rescreened across modifiers (methanol, ethanol, isopropanol, with or without a basic or acidic additive) in a single automated sequence, which is standard practice in pharma chiral-method development where a screening panel across several columns and modifiers is run before selecting a method.
- Avoids normal-phase HPLC’s hexane dependency — conventional normal-phase chiral HPLC typically runs hexane/heptane with an alcohol modifier; hexane is flammable, has a poor safety and disposal profile relative to CO2, and its higher viscosity limits achievable flow rates on the same column.
This is why chiral SFC has become the default first-pass technique in pharmaceutical enantiopurity work: for a molecule with a defined chiral center, CASRAI would generally point a method developer toward SFC screening before defaulting to normal-phase HPLC, precisely because the same CSP chemistry performs better there.
2. Preparative-Scale Solvent Economics
At analytical scale, solvent volume is a rounding error. At preparative/purification scale — isolating milligrams to kilograms of a target compound — solvent consumption, recovery, and downstream evaporation become the actual cost and time driver, and this is where SFC’s economics separate sharply from prep HPLC:
- The bulk mobile phase leaves as gas. After the back-pressure regulator drops CO2 below its critical pressure, it flashes to gas and vents (or is recovered/recycled on larger systems), carrying almost none of the target compound with it. What’s collected is the compound in a small volume of modifier — not the compound diluted in liters of mobile phase that then has to be rotary-evaporated down, the standard bottleneck in prep HPLC fraction collection.
- Dramatically less organic solvent purchased and disposed of. A prep HPLC run consumes liters of acetonitrile, methanol, or hexane per purification; the equivalent SFC run replaces most of that volume with CO2, which is inexpensive, non-flammable in its bulk form, and doesn’t carry the same hazardous-waste-disposal cost per liter that spent HPLC mobile phase does.
- Faster fraction work-up. Because the collected fraction is already mostly free of mobile phase, time-to-dry-compound after collection drops substantially compared to a prep-HPLC fraction that still needs full solvent evaporation.
The net effect shows up as fewer purchased solvent drums, a smaller hazardous-waste stream, and shorter turnaround from injection to isolated, dry compound — the reason preparative SFC has become standard equipment in pharma process/medicinal chemistry labs doing repeated small-molecule purification, not just a chiral-specific tool.
3. Normal-Phase-Like Selectivity, Without Normal-Phase HPLC’s Problems
Because supercritical CO2 is non-polar, SFC’s default selectivity resembles normal-phase HPLC — useful for separating compound classes (positional isomers, closely related non-ionizable species) that reversed-phase C18 methods struggle to resolve. SFC gets there without normal-phase HPLC’s two chronic practical problems:
- Water sensitivity and retention drift. True normal-phase HPLC on bare silica is notoriously sensitive to trace water in the mobile phase and column equilibration state, producing retention-time drift between runs and between labs. SFC’s CO2/modifier mobile phase is far less prone to this.
- Long re-equilibration times. Normal-phase HPLC columns often need extended flushing to re-equilibrate after a gradient or a modifier change; SFC systems generally re-equilibrate faster because the mobile phase composition change is a smaller physical disturbance to the column.
In practice, this means a method developer who wants normal-phase-style selectivity for a non-chiral separation (e.g., resolving structurally similar impurities that co-elute on reversed phase) can often get it via SFC with better day-to-day reproducibility than true normal-phase HPLC delivers.
The Instrument Constraints That Decide It
None of the advantages above are free — SFC requires hardware and expertise that a standard HPLC lab doesn’t already have, and these constraints are usually what actually decides whether SFC is the right call for a given lab, not the chemistry alone.
| Constraint | What it means in practice |
|---|---|
| Back-pressure regulator (BPR) | CO2 must stay above its critical pressure through the column and detector or it flashes to gas mid-flow-path. The BPR downstream of the detector maintains that pressure (typically 100–400 bar operating range) and is the single most common SFC-specific maintenance point — precipitated salts or non-volatile buffer components from the modifier can foul it. |
| CO2 supply and pumping | Requires a dedicated CO2 source (typically a bulk cylinder with a dip tube or a CO2 delivery system) and a pump designed to deliver liquefied CO2 reliably — not a standard HPLC pump. This is capital equipment a pure-HPLC lab doesn’t already own. |
| Detector compatibility | UV detection works with a pressure-rated flow cell placed before the BPR. MS coupling is workable but needs a post-BPR interface with make-up solvent flow, since the mass spectrometer’s source operates at atmospheric pressure and CO2 has already flashed to gas by that point. ELSD is straightforward and commonly used for compounds with weak UV chromophores. |
| Sample and modifier solubility | Not every analyte dissolves well in CO2/modifier mixtures — highly polar, ionic, or very high-molecular-weight compounds are a poor fit unless the modifier percentage runs high (at which point some of SFC’s efficiency advantage erodes). Highly aqueous-soluble or ionic analytes are usually still better served by reversed-phase HPLC. |
| Modifier flammability under pressure | Methanol/ethanol modifiers are flammable; venting the depressurized CO2/modifier stream needs proper fume extraction, the same as any other flammable-solvent exhaust in the lab. |
| Narrower vendor and consumables base | Fewer instrument vendors build dedicated SFC/UHPSFC systems than build HPLC systems, and SFC-specific consumables (BPR cartridges, CO2-rated seals) have a smaller supplier pool than HPLC’s. Method transfer between labs is less turnkey than transferring an HPLC method. |
When HPLC Is Still the Right Default
SFC’s advantages are real but situational. Stick with HPLC (reversed-phase, in most cases) when:
- The analyte is achiral and reversed-phase C18/C8 selectivity already resolves it — there’s no reason to add SFC-specific hardware and expertise to a method that reversed-phase HPLC already handles well. See HPLC column selection for choosing the stationary phase.
- The analyte is highly polar, ionic, or aqueous-soluble — these are a poor match for a CO2-based mobile phase even with a high modifier percentage.
- The method needs to run on infrastructure the lab already has and validate against an existing compendial or regulatory method written for HPLC — method-transfer and revalidation cost can outweigh SFC’s throughput gain for a single method.
- The lab doesn’t already have SFC capital equipment and the volume of chiral or preparative work doesn’t justify the acquisition — a contract lab or CRO with existing SFC capacity may be the more practical route for occasional chiral method development. See HPLC system cost for the capital-cost comparison point HPLC starts from.
SFC vs HPLC at a Glance
| Dimension | SFC | HPLC |
|---|---|---|
| Primary mobile phase | Supercritical/subcritical CO2 + polar modifier | Aqueous buffer + organic modifier (reversed-phase) or organic solvent (normal-phase) |
| Chiral separation performance | Generally faster, better resolution on polysaccharide CSPs | Workable (normal-phase chiral HPLC) but typically slower, hexane-dependent |
| Preparative solvent/waste cost | Substantially lower — bulk mobile phase vents as gas | Higher — full solvent volume must be evaporated/disposed |
| Normal-phase-style selectivity | Yes, without water-sensitivity/re-equilibration issues | Yes on bare silica, but retention drift-prone |
| Best fit for highly polar/ionic analytes | Poor without heavy modifier | Strong (reversed-phase) |
| Capital equipment | Dedicated SFC/UHPSFC system, CO2 supply, BPR | Standard HPLC/UHPLC system, widely available |
| MS coupling | Workable, needs post-BPR make-up flow interface | Native fit — see LC-MS coupling |
| Vendor/consumables base | Narrower | Broad, mature |
Frequently Asked Questions
Can SFC fully replace HPLC in a lab?
No, and that’s not really the comparison — most labs that run SFC keep HPLC as the default for routine achiral, aqueous-compatible methods and bring SFC in specifically for chiral work and preparative purification, where it has a real, mechanism-level advantage rather than just being a novelty technique.
Does SFC require special sample preparation?
Sample solubility in the modifier (not in CO2 itself, since CO2 is delivered as the bulk mobile phase and the modifier is what typically dissolves the analyte) is the main consideration — otherwise sample prep is comparable to HPLC. Samples are usually dissolved in a solvent miscible with the modifier being used.
Is SFC considered a greener technique than HPLC?
Directionally, yes, for the reason covered above: CO2 is the dominant share of the mobile phase, non-flammable in bulk, and doesn’t require the same solvent purchase and hazardous-waste-disposal volume that an equivalent HPLC method does — particularly at preparative scale. It isn’t solvent-free, since a polar modifier is still required for most real methods.
What CO2 purity does SFC require?
SFC generally requires high-purity, chromatography-grade CO2 — trace impurities in lower-grade CO2 supply can affect baseline stability and detector response, so instrument vendors specify a minimum purity grade for their systems; check the specific system’s manual rather than assuming standard industrial CO2 is adequate.
For the reversed-phase and normal-phase HPLC fundamentals this comparison assumes, see HPLC: columns, mobile phases, and troubleshooting and HPLC vs UPLC vs UHPLC. For column chemistry generally, see HPLC column selection and, for a different separation mechanism entirely, size exclusion chromatography.
