Thanks to a microscope that can see individual molecules, scientists have taken the first snapshots of crystals being born in a solution. Their pictures throw a monkey wrench into the theories about how crystal formation gets started.
If you've ever played with one of those "magic rock" crystal growth kits, you have made a crystal out of a so-called supersaturated solution. This is a fluid in which more of a compound is dissolved than the solution can easily hold. To nudge it out of its supersaturated state, all you need to do is put a seed crystal into the solution; the dissolved substance accretes around it, creating--if you're lucky--a nice, large crystal. In nature, too, crystals grow around the few small crystals that first form in a solution.
Now Peter Velikov and Siu-Tung Yau at the University of Alabama at Huntsville have used an atomic force microscope to take the first images of the birth of the seed crystals, a process called nucleation. As they report in the 3 August issue of Nature, they repeatedly aimed the microscope at the bottom of a 50-microliter jar filled with a supersaturated solution of apoferritin, a sphere-shaped protein. They could see newly formed crystal nuclei gain and lose molecules. This allowed the researchers to determine that the "critical size" of a seed crystal--how big it must be before it triggers the growth of a large crystal--was somewhere between 20 and 50 molecules, depending on how supersaturated the solution was.
But the most interesting feature of these tiny nuclei is their shape. Theory had predicted that they would be spherical, because, like soap bubbles, they are subject to a surface tension that minimizes their surface area. "You really expect something compact," Velikov says. "But these were quite planar, like a raft." It is unclear just why this is so. University of Chicago chemist David Oxtoby finds the result puzzling. "It's a surprise," he says. "It certainly raises questions about classical nucleation theory."