Why Is My Fragrance Oil Separating From the Wax?

Wax is a solvent, and every solvent has a limit

This is the single idea the rest of this page is built on, so it is worth stating precisely. When you add fragrance oil to molten wax you are not making a mixture in the way you make a mixture of sand and gravel. You are making a solution. The oil molecules distribute themselves individually among the wax molecules and stop being a separate substance. That is why a properly made candle has no visible oil in it anywhere — the oil is in there, it is simply dissolved.

Every solution has a saturation limit: the maximum quantity of the dissolved substance the solvent will accept. Below the limit, everything goes in and stays in. At the limit, the solution is saturated. Above the limit, the surplus has nowhere to be. It does not vanish and it does not become part of the wax; it stays a separate liquid, and because oil and wax have different densities and different affinities for the surfaces around them, that separate liquid moves. It travels to the top, to the glass, or into any void it can find. Separation is not the oil going wrong. It is the oil doing exactly what surplus always does.

Three things set where the limit sits, and it is worth being clear that all three matter at once. The first is the wax. Different waxes are formulated differently and hold different quantities. This is exactly what a stated fragrance load is: the wax manufacturer's figure for how much oil that specific wax has been formulated to hold. Luxury Soy Wax Chunks are rated up to 13%. Luxury Soy Wax CSI 464 is rated up to 10%. Those are two different numbers for two soy waxes sold by the same shop, which should be enough to retire the idea that there is a universal soy figure.

The second is the oil. A fragrance oil is not one chemical; it is a blend of dozens, and each of those constituents has its own affinity for wax. An oil weighted toward heavy, waxy, resinous materials tends to sit in wax comfortably. An oil weighted toward small, polar, highly volatile materials tends to sit in it less comfortably. This is why two oils at the same load in the same wax can behave differently, which is covered properly further down.

The third is temperature, and it is the one that causes most of the confusion in candle making, because it is the only one of the three that changes after you have finished pouring. The next section is entirely about it.

Why temperature matters twice

Heat increases how much a solvent will dissolve. That is true of sugar in water and it is true of fragrance oil in wax. Molten wax at pouring temperature is an enthusiastic solvent; the same wax as a solid block at room temperature is a much less enthusiastic one. The saturation limit is not a property of the wax alone. It is a property of the wax at a temperature.

This has a consequence that catches out a great many makers, and it explains the most frustrating version of the complaint. The wax cools, so the limit falls, but the load does not fall with it. The grams of oil you weighed into the pitcher are still in the jar. If you added an amount that the wax could hold at 85°C but cannot hold at 25°C, then somewhere between those two temperatures the solution crossed its saturation point, and from that moment onward the excess is looking for somewhere to go. The mix genuinely was fine when you poured it. It stopped being fine while you were asleep.

Temperature acts a second time, over weeks, in storage. A candle sitting in a workshop in an Indian summer is not held at one temperature — it warms through the afternoon and cools overnight, every day. Each warm phase gives dissolved and undissolved oil more mobility; each cool phase drops the limit again and pushes a little more out. That is why a candle can look flawless on pour day, acceptable at day three, and visibly wrong at day ten. Nothing new was added. The candle simply had more thermal cycles in which to reach equilibrium, and equilibrium at room temperature does not include your surplus.

There is a second temperature effect worth separating out from the first, because makers often confuse them. Adding oil to wax that is too cool does not just risk a lower saturation limit — it risks the oil never dissolving in the first place. Below the specified band the oil disperses as fine droplets that look mixed but never entered solution, and those droplets separate almost immediately. Adding to wax that is too hot costs you the lightest, most volatile constituents of the fragrance, which flash off before the wax reaches the jar. Eco Soy CSI 400 specifies adding fragrance at 80–90°C with two minutes of stirring, and CSI 464 states melt and pour in the same 80–90°C band. Those are published numbers. Use them, and measure with a pen thermometer at ₹354 rather than judging by eye.

Where separation shows up, and what each location means

Surplus oil moves, and where it settles is diagnostic. Four locations account for almost everything makers report. Read them as a sequence of severity rather than four unrelated faults — the same mechanism produces all four, and the further down the table you are, the more surplus there is.

One clarification, because it saves arguments. A thin, even sheen appearing on a candle that has been standing in direct afternoon sun is not the same finding as a persistent film on a candle stored in a cool, dark carton. The first may resolve when the candle cools; the second will not. Always assess a candle at a stable room temperature, out of sunlight, with the lid off. If you are unsure whether what you are looking at is separation, sweating or a mixing fault, the sweating troubleshooting guide works through wax, load, mixing and temperature as four separate suspects.

The four inputs you control

You cannot change the chemistry. You can change four things, and only four things matter enough to be worth your attention. Where CSI publishes a figure it is stated below; where CSI does not, the entry says so plainly rather than offering you a number we cannot stand behind.

Oil-to-wax compatibility is real — and you cannot read it off the bottle

Everything above treats the wax as the variable. It is not the only one. Two fragrance oils at the same load in the same wax can behave differently, and one of them can separate while the other does not. That is a genuine effect, not a maker's superstition, and it follows directly from the solubility framing: the saturation limit depends on both substances, so changing the oil changes the limit.

Here is the honest part. You cannot tell from the bottle. Not from the price, not from the fragrance family, not from how the oil smells, not from how thick it looks when you pour it, and not from whether it is described as a designer-inspired or a classic. An oil's composition is a blend of many materials and its behaviour in a particular wax at a particular load is an empirical fact about that pairing. Anyone who tells you a fragrance family is universally safe in soy is guessing. The only way to know is to run the pairing.

This is also the point at which we argue against our own commercial interest, because the commercially convenient advice here would be to sell you a different bottle. CSI stocks 92 fragrance oils, and a page that concluded "your oil is the problem, try another one" would move stock. It would also usually be wrong. In the great majority of separation reports the fix is less oil and more stirring, not a different oil. A maker at the ceiling who drops to 8%, adds inside the published temperature band and stirs a timed two minutes will resolve most separation without changing a single ingredient. Work through those four inputs first. Only when a pairing still separates at a conservative load, correctly added and properly stirred, is it reasonable to conclude that this oil and this wax do not get on — and by then you will have the evidence to say so.

The test: one wax, two oils, two loads, four jars

This is the smallest experiment that produces a usable answer. It isolates the oil by holding the wax constant, and it isolates the load by running two of them, which means a single fourteen-day cycle tells you both whether an oil is the problem and whether coming down solves it.

What separation actually does to the candle

Separation is not only a cosmetic problem, and it is worth being specific about the three consequences, because each one changes what you should do about a batch that has already been poured.

Throw gets weaker, not stronger. This is the outcome that surprises people, because the intuition is that more oil means more scent. Oil that has come out of solution is no longer distributed through the wax body, so it is no longer being carried to the melt pool and released gradually as the candle burns. A film on the surface largely evaporates in the first hour and is gone; oil that has migrated to the glass is not in the pool at all. You paid for fragrance that is not participating in the burn. The candle that separated at 12% will often smell weaker over its life than the same recipe at 8% that stayed in solution.

The wick stops behaving. A wick is engineered to draw molten wax. Feed it a pool that is part free oil and it draws a fuel it was not designed for: the flame guttering, drowning, running high, sooting or refusing to sustain a pool at all. Because the pool never opens out properly, the burn narrows, and that is one of the routes into tunnelling caused by too much fragrance oil. Separation and tunnelling are frequently reported as two problems by the same maker in the same batch; they are usually one.

Finally, separation rarely arrives alone. The same excess of oil, and the same cooling behaviour, produce a family of related defects — surface films, sweating, wet glass, and cavities that open up inside the candle as it sets. If your batch shows separation alongside internal voids, the pour side of the process is worth reading about too: pouring temperature and sinkholes covers where the cooling rate fits in. And if you are still not certain whether you are looking at separation, sweating or simply a mixing failure, work through the four suspects in the sweating troubleshooting guide before you change your recipe.

Frequently asked questions