An Article In Meteorite-Times Magazine
by Tom Phillips


May 2007 Meteorite Times had a Micro Vision article titled "Space Pearls". This was just a fun name for an unknown feature I noticed in a couple meteorites. Drake Dameräu saw that article and wrote a fantastic explanation that I thought you would all enjoy. Some times other fields of study bring a new perspective to meteorites, perhaps he has done just that.

Tom Phillips

A paper attempting to explain Tom Philip’s “space pearls”, or more specifically, the snowflakes inside them.

About the Author
Drake is not a meteorite expert, but rather a fledgling meteorite collector. He is a material scientist. He has worked for one of the world’s largest defense contractors as a division metallurgist for 15 years. His expertise is in failure analysis, heat-treating and forging, and has a patent pending on a steel alloy. He is a member of ASM and its Heat Treating and Metallographic affiliates. Drake is an ASTM member, and on the ASTM A01, E08, and E28 committees. He is also a member of the Aerospace side of the SAE.

You may be asking how I could possibly have the knowledge to write this paper, given that my expertise is in metallurgy. To that I say; anyone with expertise in any field dealing with “crystal formation during solidification through liquidus”, has the knowledge to write this paper.

Tom Philip’s Space Pearls Revealed

Space Pearl: noun; a chondrule consisting of a dual phase solid solution containing single-phase crystallographic dendritic growths.

What Are Dendrites?

Let’s start with four key definitions:

Unit Cell:
The smallest structural component of a crystal.

A solid composed of unit cells of atoms, ions, or molecules arranged in a pattern, which is repetitive in three dimensions.

The separation, usually from a liquid phase on cooling, of a crystalline phase.

Crystals that are composed of stacked-up unit cells attaching themselves together or "growing" while having the proper conditions to grow in a tree-like branching pattern during crystallization.

The cross-shaped snowflakes in Tom’s Space Pearls are dendrites. Dendrites are crystals with a tree-like pattern that form or “grow” as a liquid begins to solidify. The white material or "phase" surrounding Tom’s snowflakes (called the matrix) is the solvus in the space-pearl alloy. The dendrite is the solute.

Dendrites and crystals grow as a liquid begins to solidify. However, for the dendrites to form, the conditions must be exactly right. If the ratios of elements that make up an alloy, the rate of cooling* and the many other factors are not precise, the dendrites don't get a chance to grow into the perfect little crosses, stars and trees we see in Tom’s pictures.

How do Dendrites Form?
Most elements, alloys and minerals have specific atomic arrangements when in the solid state. These arrangements are called "unit cells". A unit cell will have an exact ratio of atoms, such as five atoms of element "A", and one atom of element "B". (We will call this A+B solution “alloy alpha”) If you have 100 atoms of "A", and only 10 atoms of "B", you will form 10 crystals or unit cells of “alloy alpha”, and it will exist in a matrix of 50 atoms of element "A". The element "A" is called the solvus and the element "B" is called the solute.

Moreover, solutions of two elements usually have a different melting/freezing temperature than either of the two elements by themselves. An example is salt or sugar and water. They both change the freezing and boiling point of water. It is the working ingredient in your radiator and why you throw salt on an icy sidewalk.

Most elements, alloys and minerals have at least three phases: gas, liquid and solid. Alloys and minerals can also have multiple phases in the solid form. These would be different arrangements of the same atoms forming differing unit cells. The key to forming different unit cells and crystals is the rate of cooling. By adding this to the equation, we can have “alloy alpha”, “alloy beta”, “alloy sigma” and “alloy gamma” all forming from the same elements, all forming different dendrites. To this, add a solution containing many elements, and things can really get interesting.

Each alloy unit cell will have a specific atomic arrangement pattern called a lattice. For instance, some have a cubic arrangement (octahedral geometry), like salt. (Salt is an alloy of sodium and chlorine, but you know that.) Therefore, when you look at salt, it will always be cubes. Some unit cells are hexagonal, like the minerals that form many gemstones. An example is amethyst and quartz (SiO2). It “grows” in columns with six sides and has a point at the top.

Lets look at one more, but a bit more closely. Sugar. The table sugar crystal structure is monoclinic hemihedral and has three elements - C12 H22 O11. You know that sugar is easily dissolved in water, but you can only dissolve a limited amount. If you attempt to add too much sugar to a glass of cold water, you end up with a glass of sugar water and undissolved sugar. This is because water can only dissolve a specific amount of sugar, at a specific temperature and pressure. The water is considered saturated. If the temperature is raised, more sugar can be dissolved and the solution of sugar water then becomes supersaturated. When the solution is cooled, the sugar will precipitate back out of solution by growing crystals. Given that sugar unit cells are cubic, the crystals grow as cubes, not the tree branch patterns. Cubes are not that exciting but it is the same process. It is how rock candy is made.

Water dissolved in air does this as well, and we get snowflakes. Snowflakes are dendrites.

Dendrites in my world
I see dendrites quite often. Because they are easily broken during metal processing and forming, they are usually only observed in cast microstructures. I see them in copper, aluminum and iron alloys. As a rule, larger dendrites are seen in material that has cooled slower*. Below are pictures from a recent report of mine. The dendrites can be seen growing into a void in a high carbon steel. Since the material was cooled slowly*, the dendrites are large, and are accompanied by a Widmanstätten microstructure.


Pictures by Author

Photomacrograph of void in steel showing dendrites.
Original magnification – 20X
Over a dozen dendrites can be seen in this view. This steel alloy was cast at 50ºF over the liquid temperature of the alloy. Due to shrinkage during solidification, a void was formed, allowing the dendrites to form without interruption or disturbance.


Selected dendrite etched with 2% Nital.
Original magnification – 50X
Compared to the Space Pearls dendrites, this is monstrous. Etching has revealed that this dendrite is made of hundreds of individual smaller crystals, whereas the Space Pearls dendrites are made of individual crystals.


Photomicrograph of steel showing Widmanstätten microstructure near dendrites etched with 2% Nital.
Original magnification – 200X
The matrix seen in this view consists of ferrite, pearlite and alloy carbides. The Widmanstätten structure seen (technically known as “Widmanstätten side plate morphology”) is made of α-ferrite and is formed during cooling.

The Formation of Space Pearls:
There were at least two elements that existed, each having a different solidification temperature. There was much more of one element(s) than the other, and one was able to dissolve in the other. They were liquid. As the liquid began to cool, a specific ratio of the elements began to solidify first. They were allowed to cool and grow without interruption or disturbance. Since there are only a few dendrites in each pearl, there was very little of the material that formed them. Once all the solute had solidified into dendrites, the solvus around them solidified. The result was "Space Pearls".

I do not know the chemical composition of Tom’s space pearls. This leaves me at a severe disadvantage in the aforementioned formation statement. EDS spectra can be performed on the dendrites and the matrix using an SEM to determine their chemical composition. From there, it may be possible to create a phase diagram and mathematically describe their formation.

*Slow cooling is a relative term. When discussing manufacturing cooling conditions, the term “slow-cool” is in 10’s of hours and a rapid cool is measured in seconds. Relative to the cooling and condensing of our solar system, rates of cooling are orders of magnitude different.

Drake Dameräu
Plant Metallurgist
Laboratory Manager
General Dynamics
Ordnance and Tactical Systems
Scranton Army Ammunition Plant

Further Study

There is a great site explaining snowflakes and dendrites at  Have a look at “Snowflake Physics”.

Short digital movies on growing dendrites to help you visualize the growth process. 

NASA experiments on dendrite growth in space on STS 87.

Neat pictures of dendrites.

Tom Phillips can be reached by email at:

Meteorite Photos and Images
The Tom Phillips Microscopic Meteorite Photography and Gallery