A Farewell to All Six Sides of Ice and Snow
This is the season of sorrow for my friend Carl Benson. He roams around the Geophysical Institute with a wistful frown upon his face, baiting the innocent into asking him what's wrong.
"The ice is melting," he'll say. "The snow is almost gone. Isn't it terrible?"
He's been saying this during every breakup for years, and no one is entirely sure that he's kidding. Benson is the glaciologist who leads the annual institute Christmas party singing of "Ice is Nice," so there's just no telling.
Nevertheless, in honor of Benson's annual rue, this column is dedicated to a very basic question about ice, one posed and answered in a recent issue of the British publication New Scientist: Why do snowflakes have six sides?
I admit it's not a question that much concerned me. For years, I simply accepted that snowflakes have six sides the way I accepted that peanut-butter sandwiches have insides and outsides. Scientists don't think that way, which is why they're scientists and I'm not.
Scientifically minded people have been puzzling over why snowflakes have six sides for a long time. Way back in 1611, the famous astronomer Johannes Kepler took time off from peering at the heavens to write an essay "On the six-cornered snowflake." Kepler spun a theory based on geometry. He knew that the most efficient way to pack spheres is in a hexagonal array. He believed that snowflakes were made up of globules of condensed and frozen moisture. He put his knowledge (which is true) together with his belief (which wasn't true) and came up with a reasonable but incorrect explanation: the symmetry of snowflakes arises from geometrical efficiency, because they are well packed arrays of tiny spheres.
Though others soon appreciated the crystalline nature of snow, it took a little more than three centuries to correct Kepler fully. Thanks to the X-ray diffraction technique developed by another Teutonic scientist, Max von Laue, the arrangement of water molecules in an ice crystal was worked out by 1929. And the shape of a snowflake is determined first at the molecular level, with the atoms composing the water molecule.
Whether it's in the form of a vapor, a liquid, or a solid, each molecule of water contains two atoms of hydrogen joined to one atom of oxygen. Oxygen has a more powerful hold on the electrons than the two types of atoms share to form their water-making bond. The electrical effect of oxygen's tighter grip is that the oxygen atom becomes slightly more negative while the hydrogen becomes slightly more positive. Opposite charges attract, so a positive hydrogen atom of one water molecule tends to stick to the oxygen atom of a neighboring molecule.
The more energetic the water, the more the molecules jostle and move relative to their neighbors, with little stability in the links between them. As the mass of water molecules loses heat energy---that is, as the water cools down toward the freezing point---these hydrogen bonds between molecules break less often.
Upon freezing, everything seizes up. Von Laue's X-ray patterns from seized-up water crystals showed that Kepler hadn't been far wrong. Each water molecule is surrounded tetrahedrally by four others to which it has hydrogen bonds. The oxygen atoms are arranged hexagonally in layers. This is what provides the underlying sixfold symmetry of the crystal lattice, the structure that grows to become the visible snowflake or platelet of water-borne ice.
The growth that takes place on this molecular skeleton maintains the initial hexagon not because of some geometric efficiency, as Kepler surmised, but because of energy efficiency. When a new layer forms on its sides, the growing crystal needs to spend less energy than for adding a layer on its top or bottom. The side faces advance more quickly, and the crystal grows into a hexagonal plate.
So, Benson's views notwithstanding, I have an appropriate comment for the departure of all the little six-sided crystals that cluttered the landscape all winter: I'm glad, glad, glad, glad, glad, glad to see you go!
