""HomeHeat TransferSolidificationMechanica lPropertiesCatalysisCircular MotionMicrogravity
Solidification MenuSolidification Introduction Crystaline Growth Phase Diagrams Intermetallics and Immiscibles Glossary

Crystalline Growth

What Crystals?

If you talk to many people about crystals they will assume that you want to offer them some new age therapy. In fact crystals are all around us and are vital for modern life. As well as the more obvious crystals such as diamonds and quartz together with other precious and semi-precious stones crystals can be found in all walks of life.

 Etched aluminium ingot showing crystaline structure 

Etched Aluminium Ingot

The above picture shows the crystal structure that forms when molten metal is cast in a cold mould. The crystals that formed on the very outside of the mould, next the the relatively cold wall are small and randomly shaped. Since they grew in all directions at once they are called 'equiaxed'. Deeper in there was a definite temperature gradient and so the crystals that have grown are like long thin columns and so are called ‘columnar’ crystals. The ones that formed in the centre have again grown roughly equally in all directions (equally along all their axes). Even though they have been grown from the same molten material (the melt) and so are chemically the same the individual areas have very different mechanical properties. This could be very dangerous if the casting plays an important role in the safety of the manufactured product. Engineers and metallurgist need to be very aware of the mechanical properties required of a casting and how to treat the metal in order to achieve them.

Think about the structure of the columnar and equiaxed areas of the casting. Which area is likely to have the same mechanical properties in all directions? In what direction is the other area likely to be strongest?

It is very difficult to study how crystals grow when metals cool as it happens. Fortunately there is a safe and easy alternative.



Wash the glass plates in advance and rinse them with de-ionised/distilled water. Place them in an oven or heater cabinet to dry at a temperature of 50 °C. Do not remove them until you are ready to continue.

Place a very small quantity of phenyl salicylate in a boiling tube and heat it gently until it melts. As the melting point is only 42 °C this could best be done in a heat bath.

Hold the glass plates in gloved hands and handle them by the edges. Place the bottom plate on the workbench and poor a few drops of the phenyl salicylate on to it. Immediately lower the upper plate on to the liquid. Surface tension will cause the liquid to spread evenly across the surface.

Allow the plates and liquid to cool and observe what happens.

What should you see?    

If the plates are warm enough and clean then initially nothing should happen. Soon the plates will cool to the melting point of the phenyl salicylate and crystal growth can begin. As the plate should have cooled fairly evenly there will be no cold spots and so crystal growth can start anywhere on the plate, it is however most likely to start at the edge if the liquid touches the cut edge of the glass plate.

The crystals that form will grow into the surrounding liquid. As the crystal grows it organises the molecules in the liquid and so the crystal grows in a regular pattern. However two things can happen to upset this:

The overall result should be that you should see a number of distinct crystals (or grains to a metallurgist).


The initial picture of the metal casting suggests that different conditions can produce different types of crystal growth. Here are a few things to try that should change the way the crystals grow.

Seed Crystals

Crystals cannot form until the liquid has cooled to the melting point. However they still need to have somewhere to start growing. This could be a scratch on the glass plate, an impurity in the solution, a speck of dust, almost anything. However you can force the issue by introducing a seed crystal.

When you lower the top plate make sure that the liquid reaches just beyond one edge of the plate. Then with a pair of tweezers or a metal probe, place a solid crystal of phenyl salicylate up against the liquid. If the plate is still to hot the crystal may melt. If it does then keep adding another small crystal until it survives. Almost immediately the crystal should start to grow and since no other crystals have yet had a chance to form it should end up as the largest crystal on the plate.

Temperature Differences

Despite many materials having a fixed melting point it is possible for the liquid to be cooled below that temperature without solidifying. The greater the temperature is below the melting point (undercooling to a metallurgist) the more likely it is that a particular defect or impurity will trigger crystal growth. These places are called nucleation sites.

Rapid Cooling

Make sure that the liquid phenyl salicylate does not reach the cut edges of the glass plates. Using a ring of plasticine or blue-tac, construct a dam on the upper plate and then place a little crushed ice in it. The ice will melt and so cause the phenyl salicylate to cool rapidly. Many of the potential nucleation sites will trigger crystal growth. The resulting crystals would rapidly run in to each other, preventing further growth. The result should be a large number of very small crystals. You may even need a hand lens in order to see that there are crystals there at all.

Slow Cooling

As soon as the top plate is in place return the plates to the oven or warming cabinet. Close the door and reduce the temperature setting to 40 °C. Assuming that the controls are accurate the temperature will slowly drop to just below the melting point. This may take several hours. Because the cooling process will have been so slow only a small number of potential nucleation sites would have triggered crystal growth. Because of this the crystal would have grown a long way before meeting a competing crystal and so you should see a small number of very large crystals. It is possible that you may even see the extreme case of where only a single crystal had been able to form by the time that the entire sample had solidified.

Differential Temperature

This variant is rather tricky to achieve and needs a little forward planning.

 Cooling Curve: Pure 

Heater and Motor Setup

Cooling Curve: Pure 

Control Circuit Diagram

The heater compartment on the left can be constructed from a glass or ceramic sandwich wound with resistance wire. The thermistor allows the heater to be set to a constant 50 °C and the compartment is covered in (non-flammable) insulation.

The motor and gearing is set to slowly draw the sample plates out from the heater. The outside temperature should be as low as can be easily managed and the metal plate will help with this. The cold metal plate should not touch the heater compartment and if possible there should be a small amount of insulation between the two.

Turn on the heater compartment and allow it to come up to temperature. Prepare the plates as before and then place them in the heater. Attach the motor and gearing and turn it on. Ideally the gearing should be set up so that it takes an hour or so to draw the plates out of the heater compartment. Place a few seed crystals at the open end of the plates.
As the plates are drawn out there will be a very narrow space between the where the temperature is too high for crystals to grow (inside the heater compartment) and where it is far below the melting point (on the metal plate). Due to the seed crystals there will be a number of crystals ready to grow in to this space. No new crystals can form ahead of them and so the existing crystals will continue to grow forwards at the rate at which the plates are drawn out. The result should be that a set of crystals grow as narrow fingers parallel to the direction of motion of the plates. This is exactly the same as is seen at the edges of the first picture.


The implications are that columnar and equiaxed crystals grow under different temperature conditions. If the temperature changes rapidly between two adjacent locations the growth will tend to produce columnar crystals in the direction of the temperature gradient. If the temperature gradient is small then the growth will tend to produce equiaxed crystals.

This explains why the picture of the casting has columnar crystals form at the edges, where there was a large temperature gradient between the cold wall of the mould and the hot molten metal. In the centre of the casting the still molten metal is insulated from the cold outside by the still hot, solid ‘crust’ of metal. Because of this the temperature gradient is very low and so equiaxed crystals will grow.

By understanding how temperature gradients and the speed of cooling (the rate of cooling) influence the crystal structure, metallurgists can adjust the conditions and so the mechanical properties of the casting. It is even possible with very careful manipulation to produce a casting that is made from a single crystal, this is extremely difficult and costly but the result can be incredibly strong. The type of furnace used to achieve this ‘directional solidification’ is called a Bridgeman furnace a small version of which you have built for this experiment.

Top of Page

ESA Logo