Preparing a Supersaturated Solution, and Why It Holds

A test tube of clear liquid can hold much more dissolved salt than it should hold. Add one small crystal and the salt comes out all at once. The tube fills with solid in seconds. That is the whole demonstration. It is also the shortest way to see what a supersaturated solution actually is.

The earlier version of this post was a bare laboratory note. It said to fill a test tube with water, add "enough crystals" of sodium thiosulfate, heat gently, cool without shaking, and drop in a crystal. It never said how much, how hot, or why any of it worked.

What a supersaturated solution is

Every salt has a limit. At a given temperature, a fixed amount will dissolve in a given amount of water and no more. That limit is the solubility.

Plot solubility against temperature and you get the solubility curve. It is the boundary between what dissolves and what will not.

A solution sitting exactly on that line is saturated. A solution above it is supersaturated. It is holding more salt than its temperature allows.

Sodium thiosulfate suits this demonstration because its solubility rises steeply with temperature. Dissolve a lot in hot water and cool it, and the limit drops out from under the salt. The salt is now above the curve.

Nothing has pushed it out yet. That is the whole situation.

The numbers that make it supersaturated

Solubility figures for thiosulfate are quoted on different bases, so the arithmetic below follows one source throughout. LibreTexts gives the solubility of sodium thiosulfate as 50 g per 100 g of water at 25 degrees Celsius.

Its worked example dissolves 70 g in 100 g of hot water, then cools to room temperature. Only 50 g can stay dissolved. The other 20 g has to come out.

The GESTIS substance database quotes a different basis. It gives 70.1 g per 100 mL at 20 degrees and 231 g per 100 mL at 100 degrees. Those figures are per 100 mL of solution, not per 100 g of water, and at a different temperature.

The two are not interchangeable, so mixing them produces wrong answers. What the second set does show clearly is the shape: solubility roughly triples between 20 degrees and boiling.

Figure	Basis	Temperature
50 g	per 100 g of water	25 degrees Celsius
70 g dissolved, 50 g stays	per 100 g of water	hot, cooled to 25 degrees Celsius
70.1 g	per 100 mL of solution	20 degrees Celsius
231 g	per 100 mL of solution	100 degrees Celsius

Solubility of sodium thiosulfate, quoted as each source gives it. The two bases are different quantities and should not be combined.

The pentahydrate carries its own water

The material sold as sodium thiosulfate is nearly always the pentahydrate, written Na2S2O3 dot 5H2O. Its molar mass is 248.18 g. Of that, 158.11 g is salt and 90.08 g is water, already locked into the crystal. Water is 36 percent of the mass you weigh out.

The starting material: clear, glassy crystals of sodium thiosulfate, sold as the pentahydrate. Adam Redzikowski, CC BY-SA 3.0, via Wikimedia Commons

This matters for the recipe. To reach the textbook loading of 0.70 g of salt per 1.00 g of water, pentahydrate needs about 183 g for every 100 g of water you add.

The earlier version said only "enough crystals". Add too few and you have a saturated solution. It sits there looking perfectly clear, and the seed crystal does nothing at all.

You would then conclude the solution was not supersaturated. It was. It was simply not loaded enough, and nothing about its appearance tells you which one you have.

Why the tube waits, and what one crystal changes

Supersaturated means above the solubility curve. It does not mean the salt wants to stay there.

A crystal has two costs that pull against each other. Inside, the salt becomes ordered and gives energy back. On the outside, the new surface has to be built and costs energy.

For a small cluster of ions the balance works out as dG(n) equals minus n times the supersaturation, plus a surface term that grows faster than the cluster itself.

At small sizes the surface term wins. Growth is uphill, so the cluster shrinks back and the salt goes into solution again.

Then the curve turns. It rises to a peak and falls away. The peak sits at a critical cluster size, given as n star equals 64 Omega squared times alpha cubed, divided by the supersaturation cubed.

Below the peak a cluster loses. Above the peak, growth pays for itself and the crystal runs away.

Crystal growth under crossed polarizers, field of view about 1.5 mm across. Each crystal is a front with a fixed internal pattern. Lobastov BM, CC BY-SA 4.0, via Wikimedia Commons

The height of that peak is the barrier. It works out to 32 Omega squared times alpha cubed, divided by the supersaturation squared.

The rate follows the barrier in the same way a reaction rate follows an activation energy. It is the exponential of the barrier divided by thermal energy.

This is why the experiment behaves like a trick. Below some supersaturation, nothing happens for as long as you can wait. Past it, everything happens at once.

Notice that the supersaturation is squared in the barrier. Doubling the supersaturation cuts the barrier to a quarter.

That is why the last few degrees of cooling matter more than all the cooling before them. The demonstration is not gentle about it.

The seed crystal, and what it skips

Nothing crossed the barrier. That is the entire reason the tube was still clear.

A crystal you drop in is already past it. It has a face, and that face carries the right arrangement of ions. Salt from the solution can join the pattern instead of having to find it by accident.

This is called seeding, and it is the reason one grain is enough.

The saving comes from the surface term. On a flat face, a growing crystal builds one new layer instead of a closed shell. Less surface for the same amount of crystal. Less barrier.

The same effect explains why the experiment fails when it should work. A scratch in the glass is a surface. A speck of dust is a surface. So is a bubble, or a stirring rod, or the rough inside of a test tube that has been used too many times.

So is shaking. So is warming the tube with your hands while you watch it.

The instruction to cool without shaking is right, but the reason is not fussiness about spills. Every one of those things is a place where a crystal can start.

Heat, and the reason this salt is the easy one

The method works so easily with sodium thiosulfate for a reason that has nothing to do with supersaturation at all.

The pentahydrate melts at 48.3 degrees Celsius. Not decomposes, not burns. Melts, into the water that was already inside the crystals.

The crystal structure of the pentahydrate, built from X-ray data. The water molecules are part of the structure, not coating the outside, which is why the salt melts in its own water of crystallization near 48 degrees Celsius. Ben Mills, public domain, via Wikimedia Commons

So the crystals dissolve in themselves. You do not need to boil anything, and you should not. A water bath is plenty, and boiling is a good way to lose water and change your ratio.

A common way to fail here is to skip the salt entirely and use anhydrous thiosulfate. It has no water of crystallization to melt into, so it behaves like an ordinary salt and the demonstration becomes fussier.

The anhydrous salt, with no water anywhere in the structure. Beside the pentahydrate, the difference explains why the laboratory material is the easy one. Ben Mills, public domain, via Wikimedia Commons

What you see, and what is left

The collapse is fast. The description in the earlier version, "within few minutes", is generous. The visible front usually crosses a test tube in well under a minute.

Crystallization running through a supersaturated sodium acetate solution. The same effect as the thiosulfate tube, caught mid-collapse rather than at the end. MILOS ALEKSANDRIC, CC BY-SA 4.0, via Wikimedia Commons

It gives off heat as it goes. Crystallization releases energy, so the tube warms against your hand. You can put a cold finger on the glass and feel it.

This is the whole basis of the sodium acetate heat pack sold for use on a sore neck. That product is already nucleated, so the heat comes out on demand when you flex a metal disc inside it. The heat of solution of sodium acetate trihydrate is 19.7 kJ per mole.

Growth does not even out afterwards. Two things run at once. New crystals keep forming while there is still salt to spare. Then, as the concentration falls, the large crystals grow at the expense of the small ones, which dissolve back in.

That second process has a name. It is called Ostwald ripening, and it is the normal ending of a phase transformation in a metastable liquid.

It is why the finished mass is never one uniform size. You get a few large crystals surrounded by many small ones, and the small ones are still shrinking when you look at them.

Why the tube is not left nearly full of solid

This is the part the earlier version got backwards. It says the tube ends up full of crystals with "very small amount of solution will be left".

Take the textbook case. Seventy grams of salt in 100 g of water, cooled to 25 degrees. Twenty grams crystallize. Fifty grams stay dissolved in 100 g of water.

That is exactly the solubility at 25 degrees. The final solution is saturated. It sits precisely on the curve.

So the experiment does not end with most of the salt turned to solid. It ends with the salt and water back in their proper ratio, and 20 g of the original 70 g lying on the bottom as crystals.

The demonstration works because you can watch a metastable state end. You cannot watch an equilibrium state.

Where this shows up outside the test tube

Supersaturation is not a classroom trick. It is an industrial problem with a practical answer.

In commercial crystallization, supersaturation is the lever you pull to decide how large the crystals come out. Start many small ones early and you get a fine powder that is hard to filter. Start few and let them grow, and you get large, clean crystals.

The chemistry is identical in both cases. What changes is the nucleation rate, and the nucleation rate is controlled by how the supersaturation is managed as the material cools.

A supersaturated solution is also metastable, and metastable states do not last forever. Modern work describes the crystal embryo forming inside a dense liquid droplet that appears first, rather than forming directly in the water.

That two-step picture fixes an old problem. Rates predicted by the older theory came out many orders of magnitude above what anyone measures. The dense droplet lowers the barrier before the crystal has to form at all.

At high enough supersaturation the barrier vanishes entirely and nucleation becomes spontaneous. Below that point there is a real barrier, and above it there is none. The change between the two is sharp rather than gradual.

Key Terms in Plain English

These are the words in this article that would send you to a dictionary. Each one is given here the way it is actually used above.

Term	What it means
Saturated solution	Holds as much dissolved salt as it can at that temperature, and no more.
Supersaturated solution	Holds more than that limit, which means it is above the solubility curve.
Solubility	The most that will dissolve in a given amount of water at a given temperature.
Solubility curve	A graph of solubility against temperature. It is the line a solution sits on when it is saturated.
Nucleation	The moment a small cluster of dissolved ions becomes stable enough to grow into a real crystal.
Seed crystal	A crystal you add on purpose, so that growth starts from an existing pattern instead of by accident.
Metastable state	A state that will last, but is not the true equilibrium and will end on its own given time or a nudge.
Ostwald ripening	Large crystals growing at the expense of small ones, which dissolve back into the solution.

The words in this article that would send you to a dictionary, each given the way it is used above.

Key Takeaways

• A supersaturated solution holds more dissolved salt than its temperature allows. It is above the solubility curve, and it is waiting to change.
• It holds that state because every crystal has to climb an energy hill first, and until one climbs, nothing happens.
• The hill is only a few tens of nanometres tall, but the rate falls off as its exponential, so the effect looks like a switch rather than a slope.
• A seed crystal skips the hill, because it already has a face with the right pattern on it.
• Cooling the last few degrees matters more than all the cooling before them, because the supersaturation enters the barrier squared.
• Use the pentahydrate. It carries 36 percent of its weight as water, and it melts in that water at 48.3 degrees, which is why it is the easy salt to show with.
• "Enough crystals" is not a quantity. Too few and you have a saturated solution that looks identical and does nothing when seeded.
• The demonstration does not end nearly solid. It ends saturated: 70 g in 100 g of water leaves 50 g dissolved and 20 g as solid.
• Watch for dust, scratches, bubbles and shaking while it cools. Each one is a surface where a crystal can start.
• The same physics sets crystal size in industry. Manage the supersaturation and you manage the product.

Frequently Asked Questions

Why does it not just crystallize on its own?

Because the first crystals are the hardest ones to make. Every crystal has to start from a cluster that costs more energy as a surface than it gains as bulk material.

Below that point the cluster falls apart. It takes a rare fluctuation to get past it, and until one happens the solution sits there.

How much salt should I use?

For a small test tube, about 18 g of the pentahydrate to 10 g of water. That gives roughly 0.70 g of salt per gram of water, well above the saturation ratio of about 0.50 at room temperature.

Can I use table salt?

You can get the effect, but it is far weaker. Sodium chloride dissolves about 36 g per 100 g of water at 20 degrees.

Its solubility curve is nearly flat. Heating it changes very little, so cooling it back down loads it only slightly past saturation.

Thiosulfate roughly triples across the same range. That steep curve is the entire reason this demonstration works.

How long can I leave it before it crystallizes by itself?

Long enough that nobody reports a limit, provided the tube is clean and nothing disturbs it. The induction time is a real measured quantity that depends on how clean the glass is and how far past saturation you went.

Is it dangerous?

Sodium thiosulfate has low toxicity, and this is a warm water bath with a salt in it. Gloves and eye protection are sensible, and you should not boil the solution, because boiling drives off water and changes the ratio you were trying to set.

Does this explain why sugar frosting forms on jam?

Yes. Supersaturated sugar solution is what you are fighting when you boil jam, and it is why a spoonful left to cool can turn to glass. A dust speck or a scratch on the pan is enough to start it.

Key references
• LibreTexts, saturated and supersaturated solutions
• Vekilov 2010, nucleation, Cryst Growth Des
• Alexandrov and Alexandrova 2020, metastable liquids
• LibreTexts, how temperature influences solubility
• Sodium thiosulfate: physical data and GESTIS solubility
• University of Washington, supersaturation of sodium acetate

Images: Wikimedia Commons, with the author and license named in each caption. Solubility figures, melting points and molar masses follow the sources above.