How to Perform a Separatory Funnel Extraction

Separatory funnel extraction is the most common way chemists physically move a dissolved substance from one liquid into another, exploiting the fact that two immiscible liquids will form distinct layers when mixed and allowed to settle. The technique underpins nearly every organic chemistry workup, environmental water analysis, and natural-product purification you will encounter in a lab setting. Despite its simplicity in concept, getting clean, efficient separations depends on understanding solvent choice, wash sequences, and a handful of practical tricks that are rarely obvious the first time you pick up the glassware.

What Actually Happens Inside the Funnel

A separatory funnel (sometimes called a “sep funnel” or “separating funnel”) is a pear-shaped piece of glass or plastic with a stopcock at the bottom and a stopper at the top. You pour two liquids into it that do not dissolve into each other, such as water and an organic solvent like dichloromethane. When you shake the funnel, you break both liquids into tiny droplets, dramatically increasing the surface area where they meet. Any solute that prefers one liquid over the other migrates across these interfaces. After shaking, you let the funnel sit undisturbed so the droplets coalesce, the two layers re-form, and you drain the denser layer through the stopcock.

The driving force behind the separation is the partition coefficient, a measure of how much a compound “wants” to be in one solvent compared with the other at equilibrium. That preference depends on the compound’s polarity, charge state, and how well each solvent stabilizes it through intermolecular interactions. A compound with a large partition coefficient favoring the organic phase will move almost entirely into that phase in a single shake. A compound with a modest partition coefficient might need several rounds of extraction with fresh solvent to pull most of it across. Partition coefficients can vary dramatically depending on the solvent pair chosen; studies comparing how simple organic acids distribute between water and various organic solvents find order-of-magnitude differences in partitioning when switching from one organic solvent to another, even for the same solute.¹

The Practical Steps, Start to Finish

If you have never done a liquid-liquid extraction before, the sequence looks roughly like this. You dissolve or suspend your mixture in one solvent (often water or an aqueous solution), pour it into the sep funnel, and add the second immiscible solvent. You stopper the funnel, invert it, and immediately open the stopcock briefly to vent any pressure buildup from volatile solvent vapors. Then you close the stopcock and shake or swirl for about 30 seconds to a minute. After shaking, you set the funnel in a ring stand, remove the stopper, and wait for the two layers to separate cleanly. The denser layer settles to the bottom, and you drain it through the stopcock into a flask. You can then repeat the process with fresh portions of the extraction solvent to pull more of your target compound across.

A real-world example from synthetic chemistry shows how involved the washing can get. In a published preparation of a silicate compound, the crude reaction mixture is transferred to a 1,000 mL separatory funnel with pentane as the organic solvent, and water is added. After the layers separate, the aqueous layer is extracted twice more with fresh pentane. The combined organic layers are then washed sequentially with hydrochloric acid, sodium bicarbonate solution, plain water, and finally saturated sodium chloride solution before drying and concentrating to yield the product at 94% recovery.² Each of those washes has a specific purpose: the acid wash removes basic impurities like pyridine, the bicarbonate wash neutralizes residual acid, and the brine wash reduces water dissolved in the organic layer so that the subsequent drying step works efficiently.

Choosing Your Solvent

The extraction solvent needs to meet a few criteria at once. It must be immiscible with the phase your compound starts in, it must dissolve your target compound well, and ideally it should be easy to remove afterward, usually by evaporation. For pulling organic compounds out of water, dichloromethane, ethyl acetate, and diethyl ether are the workhorses of the organic chemistry lab. For plant-derived compounds, solvent choice follows the polarity of the target: water and alcohols capture polar compounds, intermediate-polarity solvents like acetone and dichloromethane capture a broader range, and nonpolar solvents like hexane and chloroform target lipids and waxes.³

Density matters in a very practical way. Dichloromethane is denser than water, so it sinks to the bottom and gets drained first through the stopcock. Diethyl ether and ethyl acetate are lighter than water, so the aqueous layer is the bottom layer you drain off. Accidentally discarding the wrong layer is a rite of passage in introductory organic labs, which is why experienced chemists always keep both layers until the extraction is complete and the product is confirmed.

Volatility and safety are also in the mix. Diethyl ether extracts beautifully and evaporates easily, but it is extremely flammable and can form explosive peroxides on storage. Dichloromethane is not flammable but poses inhalation and environmental concerns. These trade-offs push many labs toward ethyl acetate as a compromise or toward newer green solvents.

Acid-Base Extractions and Selective Washes

One of the most powerful features of separatory funnel work is the ability to selectively move compounds between layers by changing the pH. A carboxylic acid dissolved in an organic solvent is uncharged and stays happily in that layer. But if you wash the organic layer with a basic aqueous solution (sodium bicarbonate or sodium hydroxide), the acid is deprotonated into its charged, water-soluble form and migrates into the aqueous layer. Similarly, an amine in the organic layer can be pulled into water by washing with dilute acid, which protonates the amine and makes it water-soluble. You can then reverse the process, re-adjusting pH to release the compound back into a fresh organic solvent if needed.

Research into the mechanisms underlying these acid-base extractions confirms that for acid extraction from water into an organic phase, the identity of the anion strongly influences how much material crosses the interface, consistent with a neutral-species extraction mechanism. In contrast, when bases are extracted, the identity of the metal cation has little effect on the extraction yield, because the dominant mechanism does not require the cation to transfer into the organic phase.⁴ In practical terms, this means that when you are designing an acid wash, the specific acid you choose (hydrochloric vs. sulfuric vs. phosphoric) can matter more than you might expect, because the counterion affects partitioning.

Why Multiple Extractions With Smaller Portions Beat One Large Extraction

A very common beginner question is: if I have 100 mL of extraction solvent, should I use it all in one go, or split it into, say, three portions of about 33 mL each? The answer, reliably, is to split it. The math behind this is straightforward even without equations. Each extraction reaches equilibrium and leaves behind some fraction of the compound in the original layer. That fraction depends on the partition coefficient, the volumes, and not much else. When you drain the first portion and add fresh solvent, the fresh solvent sees only the leftover compound and pulls out the same fraction of what remains. Each successive extraction nibbles away at the remainder, and three smaller bites remove more total compound than a single gulp of the same total volume.

This principle is why the synthetic procedure mentioned earlier extracts the aqueous layer twice more with fresh pentane rather than using one large pentane wash. The improvement can be dramatic: if a single extraction with a given volume recovers 70% of your compound, three extractions each using one-third that volume often recover above 90%. The point of diminishing returns depends on the partition coefficient. For compounds that strongly favor the extraction solvent, even one wash may suffice. For those with modest preference, three to five washes are common.

Dealing With Emulsions

Emulsions are the bane of liquid-liquid extraction. Instead of two clean layers, you get a cloudy, persistent band at the interface where tiny droplets of one phase are suspended in the other and refuse to coalesce. Emulsions tend to form when the two liquids are shaken too vigorously, when surfactant-like compounds are present in the mixture, or when the solutions are very concentrated in dissolved solids.

Industrial extraction research shows that the phase separation time is governed mainly by coalescence behavior: once small droplets rise to the interface, they still need to merge into the bulk layer, and that merging step is typically slower than the droplets’ rise through the liquid. The percentage of interfacial crud (the suspended material at the boundary) decreases with settling time and shows little further change after about 30 minutes.⁵ So patience is the first and simplest remedy.

When waiting is not enough, several tricks help break emulsions:

Add saturated brine: the high salt concentration increases the density and ionic strength of the aqueous layer, helping drive organic droplets out of it.
Gentle swirling: instead of shaking vigorously, invert the funnel slowly and repeatedly, or stir the interface with a glass rod.
Filter through celite: pouring the emulsion through a pad of diatomaceous earth can physically break up the droplet suspension.
Warm slightly: raising the temperature lowers viscosity and surface tension, encouraging coalescence. This only works with solvents that are not too volatile.

Environmental analysts face emulsion problems routinely when extracting surface water samples, which contain natural organic matter and surfactants. A validation study targeting pyrethroid and phenylpyrazole pesticides in emulsion-prone surface water found that after testing several emulsion-breaking techniques with methylene chloride as the extraction solvent, average recoveries of the target pesticides still came in between roughly 88% and 123% for moderately problematic samples and 93% to 117% for severely emulsion-prone samples.⁶ Those ranges are well within acceptable limits for regulatory analysis, showing that emulsions are an annoyance rather than a dealbreaker when proper techniques are applied.

Scaling Up and Continuous Alternatives

A separatory funnel works well for volumes from a few milliliters up to a few liters. Beyond that, shaking a huge glass funnel becomes impractical, and analysts and process chemists turn to continuous extraction devices. Continuous countercurrent extractors pump the two immiscible liquids past each other in opposite directions through a packed or pulsed column, achieving the equivalent of many sequential extractions in a single pass. One system designed for trace organic pollutant analysis in large water samples used a pulsed column of about one meter in length to extract 10-liter water samples with methylene chloride, achieving an average recovery of about 86% for a suite of organic pollutants, which was comparable to performing three sequential batch extractions. The concentration factor reached 10,000 after processing 10 liters, meaning extremely dilute contaminants could be brought to detectable levels.⁷

At the other end of the scale, microfluidic and flow-chemistry setups perform liquid-liquid extractions in channels thinner than a human hair, where the enormous surface-to-volume ratio achieves equilibrium almost instantly. These systems are increasingly used in pharmaceutical process development, where speed and reproducibility matter more than total throughput. They also drastically reduce solvent waste, since the volumes involved are measured in microliters rather than milliliters.

Environmental and Analytical Applications

Outside the synthetic chemistry lab, separatory funnel extraction is a workhorse of environmental monitoring. Regulatory methods for testing drinking water, wastewater, and surface water for organic pollutants almost universally include a liquid-liquid extraction step, often codified in official protocols from agencies like the U.S. Environmental Protection Agency. The general approach involves collecting a water sample, adjusting its pH to ensure the target analytes are in their extractable (usually uncharged) form, shaking with an organic solvent like dichloromethane or hexane in a separatory funnel, and then analyzing the organic extract by gas chromatography or mass spectrometry.

The challenge in environmental work is that real-world water samples are nothing like the clean aqueous phases in a teaching lab. They contain humic acids, suspended sediment, microbial by-products, and agricultural surfactants, all of which promote emulsions and interfere with clean separations. Analysts have developed robust protocols that incorporate salting-out, pH adjustment, and specific emulsion-breaking steps to maintain reliable recoveries even in difficult matrices. These protocols routinely achieve recoveries that satisfy regulatory quality-control requirements, as the pesticide study mentioned earlier demonstrates.

Green Solvents and Changing Practices

Traditional extraction solvents like dichloromethane and chloroform work well, but they are toxic, environmentally persistent, and increasingly subject to regulatory scrutiny. The push toward greener chemistry has prompted researchers to evaluate bio-derived and less hazardous solvents for liquid-liquid extraction. A review of green organic solvents used in extraction and liquid membrane processes found that greener alternatives, including mixtures of conventional and green solvents as well as purely green options, performed comparably to traditional solvents in terms of extraction efficiency for wastewater treatment and resource recovery applications.⁸

Ionic liquids, deep eutectic solvents, and supercritical carbon dioxide are among the alternatives gaining traction. Ionic liquids are salts that are liquid at or near room temperature, and they can be tailored by swapping their component ions to dissolve specific classes of compounds. Deep eutectic solvents are mixtures of two or more solid compounds that become liquid when combined, often involving a hydrogen-bond donor like a sugar alcohol and a salt like choline chloride. Both classes offer tunable polarity, extremely low vapor pressure (meaning almost no evaporative loss or inhalation hazard), and in many cases biodegradability. The trade-off is cost and, sometimes, higher viscosity that slows phase separation.

Supercritical CO₂ extraction sidesteps immiscible liquid layers entirely by using carbon dioxide pressurized above its critical point, where it behaves as a hybrid between a gas and a liquid with excellent dissolving power for many organic compounds. Releasing the pressure returns the CO₂ to gas, leaving the extracted material behind with no solvent residue at all. This is already standard practice for decaffeinating coffee and extracting essential oils, and it is expanding into pharmaceutical and analytical chemistry. For bench-scale work, though, the high-pressure equipment required makes it far less accessible than a simple glass sep funnel.

Common Mistakes and How to Avoid Them

Even experienced chemists occasionally run into trouble with separatory funnel extraction. A few recurring pitfalls are worth knowing about:

Forgetting to vent: volatile solvents like ether and dichloromethane build pressure quickly in a closed funnel. If you shake without venting, the stopper can pop off and spray your mixture across the fume hood. Invert the funnel, open the stopcock to release gas, close it, and only then shake.
Not knowing which layer is which: if you are unsure whether the top or bottom layer is organic, add a drop of water to the funnel. Whichever layer it merges into is the aqueous layer. Keeping both layers until the end of the procedure is cheap insurance against a mistaken discard.
Over-shaking: aggressive shaking creates fine emulsions that take forever to separate. A firm but controlled series of inversions, with periodic venting, works better than violent agitation.
Skipping the drying step: organic solvents always dissolve a small amount of water after contact with an aqueous layer. Passing the organic layer through a drying agent like anhydrous sodium sulfate or magnesium sulfate removes this water before you concentrate your extract, preventing bumping and water-related side reactions later.

When a Separatory Funnel Is Not the Right Tool

Liquid-liquid extraction in a sep funnel is versatile, but it has limits. Compounds that partition poorly into any available organic solvent, highly polar biomolecules like proteins or nucleic acids, for example, often cannot be extracted this way. For those, chromatographic techniques or precipitation methods are more appropriate. Very large-scale industrial separations use mixer-settler units, centrifugal contactors, or continuous extraction columns rather than batch funnels. And when the target analyte is present at extremely low concentrations, solid-phase extraction (passing the sample through a cartridge packed with adsorbent) often outperforms liquid-liquid extraction in both sensitivity and solvent consumption, which is why it has largely replaced sep-funnel work for clinical and some environmental assays.

Still, the separatory funnel endures because it is cheap, immediate, and extraordinarily flexible. You can set one up in minutes with no electricity, no pumps, and no specialized training. For synthetic organic chemists running reactions at the gram scale, it remains the default purification step between the flask and the rotary evaporator, just as it has been for well over a century.