The simulation runs so fast that when I first saw it at a conference I was certain it was a trick. The presenter put about 50 each of four different kinds of electrically neutral atoms inside a three-dimensional volume. The particle positions updated so quickly that I thought it had to be a computer movie, not a real-time simulation. So I decided to challenge the fellow.

In nature even neutral atoms can bond together. The mutual repulsion of the orbital electrons polarizes the atoms, and it turns out there is a range of distances over which these polarized atoms are attracted. So I asked the presenter to add these electrostatic interactions and then slowly decrease the temperature. He did. The heavier atoms began clumping together while the lighter ones kept speeding about, just as they should. He then rapidly brought the temperature to zero. The free atoms settled into small isolated clumps, again just as they should. That made me a believer.

Daniels & Daniels; Source: Molecular Modeling 3-D VIEW shows a cool crystal of krypton with a few helium atoms on its surface.

Daniels & Daniels; Source: Molecular Modeling At higher temperatures, the heliums meander about the surface....

Geologists see this clumping effect because a volcanic rock that cools slowly possesses larger mineral grains than one that cools quickly. Molecular Dynamics makes it possible to study the underlying principles of this process (called annealing) by varying the number and kind of atoms as well as the rate of cooling. By pausing the simulation at each temperature and rotating the virtual container, one can count the clumps and see how many atoms of which type are in each. That suggests an interesting study. Try repeating the experiment a few times and plotting the average size of the clumps versus the cooling rate. You may discover some fundamental facts about annealing that are quite difficult to derive mathematically.

One delightful demo starts with a cubic crystal of 63 krypton atoms. A few added helium atoms quickly bond to the surface. Tweaking the temperature upward causes the helium atoms to walk randomly on the crystal's face. At a little higher temperature the heliums leave the crystal, and if you raise the temperature still further, the crystal will fly apart. These kinds of effects are observed in real crystals. You can do other experiments here as well. Try lowering the temperature and see whether you can get the crystal to re-form. Then plot the time required for the krypton crystal to form versus the number of hydrogen atoms bouncing about. Does the hydrogen interfere with the crystal formation and, if so, why?

You can also explore gas behavior, such as how a gas adjusts to changes in temperature, volume, or number and types of its atoms. The simulation can approximately reproduce the proportionalities that are combined into the well-known ideal gas law. But only approximately. That is because the ideal gas law itself is just an approximation. It holds only if the gas atoms occupy a negligible fraction of the container's volume and if the atoms' kinetic energies are much larger than the interatomic potential energies that tend to make them clump together. As a result, any real gas departs from the ideal gas law at high densities or low temperatures. Molecular Dynamics includes these effects automatically.

Daniels & Daniels; Source: Molecular Modeling ...and at still higher temperatures, the whole thing disintegrates.

My favorite module, "Maxwell-Boltzmann Speed Distribution," lets you discover how few atoms you need before the physicists' mathematical tricks start working. One of the early triumphs of statistical mechanics in the 19th century was its ability to predict the fraction of atoms moving with a particular range of speeds within a gas at a given temperature. The curve of the fraction versus speed has a sharp rise--meaning there are fewer atoms at lower speeds--and a long tail, indicating that some atoms have speeds that are much higher than the average. I placed 100 atoms of helium and argon into the box and watched the distribution of speeds in real time. After just a few collisions, the two curves took on the expected shape. The heavier atoms peaked at a slower speed, as the theory predicts. You might enjoy removing atoms and observing how the distributions deterior

Reprints and Permissions »

Comments

Add Comment

Leave this field empty

Add a Comment

You must sign in or register as a ScientificAmerican.com member to submit a comment.
Click one of the buttons below to register using an existing Social Account.