Fractional Distillation: Principle and Setup

Fractional distillation separates liquids whose boiling points sit close together. Vapor rises through a packed column, condenses, and re-evaporates over and over, and each of those cycles enriches the vapor in the lighter component. One pass through the column does the work that a long series of simple distillations would take. The column is the whole trick.

What is fractional distillation, exactly?

Every distillation rests on one fact: when a liquid mixture boils, the vapor above it is richer in whichever component is more volatile. Boil an ethanol and water mixture and the first vapor off carries more ethanol than the pot did. That single enrichment is simple distillation, start to finish.

One enrichment is rarely enough. When two boiling points sit close together, the vapor leaves only slightly richer than the liquid it came from, and what lands in the receiver is still a mixture. You could redistil that condensate, then redistil the result, climbing toward purity one tedious pass at a time. Nobody has the patience or the material losses for that.

A fractionating column does the same climb automatically, inside one apparatus. Pack a vertical tube with glass beads, Raschig rings, structured wire mesh, or use a Vigreux column with its indented glass spikes, and vapor rising through it keeps hitting cool surfaces. It condenses. The vapor coming up behind it re-boils that condensate. Each condense-and-revaporise cycle is another enrichment step, and chemists count those steps as theoretical plates. A plate is the bookkeeping unit for one full vapor-liquid equilibrium. More plates, sharper cut.

Two consequences matter more than the textbook diagram lets on. Plates are earned with reflux, meaning liquid has to run back down the column against the rising vapor. A column that is not refluxing is an expensive piece of pipe. And plates cost time: the taller and better packed the column, the slower you must draw product off the top to keep those equilibria intact. Rush the take-off and a hundred plates behave like two.

Fractional distillation vs simple distillation: what actually changes?

The hardware difference is one component, the column. The operational difference is much bigger, and it is where most first attempts go wrong.

Look at the head temperature row. That is the row you diagnose a run with. In a simple distillation the still head reading wanders up as the pot composition drifts, which is exactly why the distillate is smeared. In a working fractional distillation the head sits flat while a cut comes over, because the column is holding equilibrium and delivering essentially one component to the condenser. When that reading starts to climb, the next component is breaking through and it is time to swap receivers.

What goes into a fractional distillation apparatus?

A fractional distillation apparatus is six things stacked bottom to top, and every one of them has a way of being set up wrong.

Heat source. A heating mantle or a bath. Bath heating is the gentler option and the one we build into our evaporation equipment, because a jacket of oil or water cannot deliver the hot spots that a bare mantle can. Whatever you use, it needs fine control. A column separates on steady vapor rise, not on raw power.

Boiling flask. Round-bottom, sized so the charge fills a fraction of it. Overfill and you will bump liquid straight up the column and ruin the cut you just spent an hour building. Boiling chips or a stir bar go in here.

Fractionating column. The packed or Vigreux section. Insulate it. An uninsulated column bleeds heat to the room, floods with condensate at the bottom, and starves of vapor at the top, and you get a fraction of the plates the glassware is capable of.

Still head and thermometer. The thermometer bulb goes level with the side-arm opening, not above it and not buried in the neck. Off by a couple of centimetres and the reading tracks nothing useful, and the head-temperature diagnostic above becomes fiction.

Condenser. Water in at the bottom, out at the top, so the jacket stays full. Cooling capacity has to match the vapor load, which is the same problem our customers meet when they size a chiller against an evaporator, and we cover the arithmetic in our guide to matching a vacuum pump and chiller to the evaporator.

Receiver. One flask per cut, or a cow adapter that rotates fresh receivers under the drip without breaking the run. If the separation is worth a column, it is worth not opening the system between fractions.

How do you run a fractional distillation setup?

Assembly is the easy half. The run is where the separation is won, and it goes in five moves.

Charge the flask and bring the heat up slowly. Slowly is not a courtesy, it is the setting. You want vapor creeping up the packing and a visible condensate line advancing with it, not a slug of vapor punching through to the head in one go.

Let the column flood and reflux before you take anything off. The condensate line reaches the still head, the head reading settles, and liquid is visibly running back down through the packing. That is the column at equilibrium. Everything you collect before this point is a mixture, so many operators run at total reflux for a while first and lose nothing by it.

Then read the head. Flat reading means the cut is clean and you can draw off, drop by drop. If the reading sags, you are pulling faster than the column can re-equilibrate: back off the take-off, or add heat, in that order.

Cut on the temperature step, not on the clock. When the head climbs off its plateau, that is the boundary between components. Swap receivers there and label the intermediate as a mixed cut to redistil later, rather than smearing it into either fraction.

Finish the cuts. A collected fraction often still holds solvent that has to come off before the material is usable, and that is a stripping job rather than a fractionation job. It is the point where the column comes off and an evaporator goes on.

Why does vacuum move the whole operating window?

Here is what the theory pages leave out. A liquid boils when its vapor pressure equals the pressure above it, so drop that pressure and the boiling point falls with it. This is not a refinement, it is the difference between a distillation that works and a flask of scorched product.

Two things happen when you pull vacuum on a distillation. High-boiling components come over at temperatures the feed can survive. And relative volatility usually shifts, sometimes helpfully, sometimes not, which is why a pair that will not split at atmospheric pressure occasionally splits under vacuum and vice versa. Vacuum is not just about getting the thing to boil.

The stripping stage at the end of a fractionation is where our hardware earns its place. Our 1L to 3L rotary evaporator line takes a collected cut down under vacuum instead of heat, and the numbers set the window you are actually allowed to work in:

Read the pairing of two rows: vacuum up to -0.095MPa against a bath that reaches 200C at plus or minus 1C. Those two together are the operating window. The bath ceiling tells you the highest boiling point you can chase; the vacuum tells you how far below its atmospheric boiling point a component will actually come off; and the plus or minus 1C tells you whether you can park just under a degradation threshold and hold there, which you cannot do on a mantle you are eyeballing.

Note also the 1L collecting flask across all three models. On the RE-301 that pairs with a 3L evaporating flask, so a full 3L charge of volatile-heavy material means emptying the receiver mid-run. Plan the charge around the receiver, not the other way round. And skip the 3L entirely if your cuts never exceed a liter, since the RE-101 does that job on the same 1000W bath and the same 24# ground joint glassware. If the whole working principle behind that flask is new to you, we walk through it in our explainer on how a rotary evaporator works, and the full rotary evaporator category lists the larger formats.

When does fractional distillation stop being the right tool?

Theory pages take you up to the column and stop. The question they never answer is the one that costs money: when is the column the wrong answer?

The mixture azeotropes. Ethanol and water are the standing example. Past a certain composition the vapor and the liquid have the same makeup, and once that happens the column has nothing left to enrich. No amount of packing fixes it, because the limit is thermodynamic rather than mechanical. You change the pressure, add an entrainer, or reach for a non-distillation step.

The feed cannot survive the pot. This is the failure we get called about most. A fractionating column only works if the charge sits at its boiling point long enough for all those plates to do their job, and for a heat-sensitive material that residence time is the problem, not the peak temperature. The compound is fine for a minute at temperature and gone after an hour of it. A column asks for the hour.

The component boils too high to reach. Push the pot hot enough and the material cracks before it ever leaves the flask, and no column saves it.

The last two cases are what molecular distillation equipment exists for. It runs short-path evaporation under high vacuum, so the vapor travels only a short distance from evaporator to condenser and the material’s time at temperature is measured in seconds rather than hours. Our source data for that line is blunt about its fit: vacuum distillation, separation, purification and concentration, aimed at heat-sensitive or high-boiling-point materials requiring gentle separation. That is the trade. You give up the plate count of a column and you buy back thermal survival.

Choosing between them is a materials question before it is an equipment question. Confirm the feed material, vacuum stability, heating temperature, feed rate, condenser support and receiving setup, and the answer usually picks itself. We lay out that sequence in the molecular distillation selection guide, along with what has to be pinned down before a technical sheet means anything.

Frequently asked questions

What is the difference between fractional distillation and simple distillation?

A fractionating column. Simple distillation gives you one vapor-liquid equilibrium, so it only separates liquids with a wide boiling-point gap. Fractional distillation stacks many equilibria up a packed column, one per theoretical plate, which lets it split close-boiling liquids that a simple setup would deliver as a mixture.

What is a theoretical plate in a fractionating column?

It is the bookkeeping unit for one complete vapor-liquid equilibrium: one condensation followed by one re-evaporation. A column with more plates enriches the vapor further before it reaches the still head. Plates only count when the column is refluxing, so an uninsulated or over-drawn column delivers far fewer than its geometry suggests.

Why must a fractional distillation be run slowly?

Because every plate needs time to reach equilibrium. Draw distillate off faster than the column can re-equilibrate and the vapor reaching the head has not been enriched by the full plate count, so your cut is contaminated. The tell is the head temperature sagging off its plateau while you are collecting.

Can fractional distillation separate an azeotrope?

No. At the azeotropic composition the vapor and liquid have identical makeup, so there is nothing for another plate to enrich. The limit is thermodynamic, not a matter of column height. Changing the operating pressure, adding an entrainer, or using a different separation step are the routes around it.

Does fractional distillation work under vacuum?

Yes, and it is standard for high-boiling or heat-sensitive feeds, because reduced pressure lowers the boiling point. Our rotary evaporator line pulls up to -0.095MPa against a bath running from room temperature to 200C at plus or minus 1C, which is the window most stripping and recovery work lives in. When the feed still cannot tolerate the residence time at temperature, short-path molecular distillation is the alternative.

Start with the cut, not the catalogue

Work it in this order. Ask whether the boiling points are close enough to need plates at all, since a wide gap means a column is effort you do not owe. Then ask whether the feed can sit at its boiling point for the length of the run, and if the answer is no, stop designing a column and look at short-path equipment instead. Only then size the glassware: flask volume against the charge, receiver volume against the cut, bath ceiling against the boiling point you are chasing, vacuum against how far you need that boiling point to fall. Our selection guide runs the same sequence across our evaporation and distillation lines. Send us the feed, the boiling points you are trying to split, and the batch size, and we will confirm the configuration against the parameter sheet before anything else happens.