The magnets that Tatiana Makarova makes are tiny black cylinders, a tenth of an inch long. If they weren't so small, you might mistake them for ordinary refrigerator magnets. But the stuff they are made from is not ordinary iron: It costs $100 a gram. In fact, Makarova handles that precious black powder in a transparent "glove box"—the kind of thing a biologist might use to contain a virus—precisely because she doesn't want it getting contaminated by some speck of iron-laden dust in her lab. Magnetic iron would not be news. What Makarova has discovered is magnetic carbon. Only four elements in the periodic table—iron, cobalt, nickel, and gadolinium—are naturally ferromagnetic at room temperature, meaning they can be permanently magnetized by exposure to a magnetic field. But the search for nonmetallic magnets—which could be light, cheap, maybe even transparent—has lately become something of a cottage industry. A decade ago, a Japanese lab isolated a metal-free organic compound that became permanently magnetized at a fraction of a degree above absolute zero. Makarova, a Russian physicist working at Umeå University in Sweden, has now found a way to make magnets of pure carbon—to be precise, of buckyballs, the soccer-ball-shaped molecules that consist of 60 carbon atoms each. Her magnets are extremely weak—"They won't stick to your refrigerator," she says—but they do work at room temperature. That's an essential quality if they are ever going to have any practical applications. For the moment, nonmetallic magnets are laboratory curiosities, and the curiosity starts with a fundamental question: Why are they magnetic? No one really knows. Moreover, asking that question inevitably lets you in on a surprising secret: Physicists are also a little fuzzy about those bits of iron alloy attached to your refrigerator. "Only a few people understand or think they understand how a permanent magnet works," says Makarova. "The magnet of everyday life is not a simple thing. It's a quantum-mechanics thing."
Bars of iron (Fe), nickel (Ni), cobalt (Co), or gadolinium (Gd) can be permanently magnetized because of the alignment patterns of their constituent atoms, which act as elementary electromagnets.Metals provided by Alfa Aesar, a Johnson Matthey Company.
Which means it's difficult to grasp, but let's try. To begin, all magnetic fields are generated by moving electrical charges—that much was discovered in the 19th century, before the advent of quantum mechanics. A current flowing through an electrical coil produces a curved magnetic field shaped like that produced by an iron bar magnet. A bar magnet's field, which seems to come from nowhere when you render it visible with iron filings, actually comes from electrons inside the metal that orbit the atoms and also spin on their own axes. It's mostly the electrons' spin that generates the field. But the miracle happens only when two conditions are met—at least according to a theory developed in the 1930s by Werner Heisenberg. First, the individual atoms in the material, and not just the individual electrons, must have magnetic fields of their own. Second, those imperceptibly small atomic fields must somehow line up in one direction to produce a single large field that we can detect. How those two conditions are met is where the physics gets really complicated. Fundamentally, what Heisenberg argued is that a permanently magnetic state is just the lowest-energy way to build certain atoms and solids out of electrically charged particles that attract and repel one another. "The origin of magnetism is the electrostatic interaction," says physicist Michael Coey of Trinity College in Dublin. Consider Heisenberg's first condition for creating a permanently magnetic state. An atom doesn't necessarily have a magnetic field just because its constituent electrons do. Electrons come in two spins, up and down, with corresponding magnetic fields. As they fill the concentric shells of an atom, electrons pair off, with each pair occupying a single part of the shell, a region of space called an orbital. A fundamental principle of quantum mechanics, the Pauli exclusion principle, requires those two paired, spatially identical electrons to have opposite spins—otherwise they would be indistinguishable. Because the paired electrons' spins are opposite, their magnetic fields cancel. Thus an atom can have a net magnetic field only if it has unpaired electrons in one of its outer shells. Unpaired electrons arise inevitably as you move down the periodic table toward larger atoms. After all, two electrons normally repel each other because they are both negatively charged. They pair off in the same orbital only because they are also both attracted to the positively charged atomic nucleus—and they do so only as a last resort. The first electrons to enter an unfilled shell scatter as far from one another as they can, one electron to an orbital. That reduces their electrostatic repulsion and the energy contained in it. The electrons in this configuration all have the same spin. That allows them to stay apart more easily and fall a bit closer to that attractive nucleus—thus lowering the energy of the atom a bit more yet. An iron atom has a strong magnetic field because it has four unpaired electrons in its outer shell, all lined up. An iron crystal made of many atoms has a strong field because the fields of those atoms line up—Heisenberg's second condition for creating a permanently magnetic state. How does this pattern of alignment occur? Again, says Coey, it's a matter of electrons lowering their repulsive energy by spreading out. Only this time, they are spreading beyond a single atom to its neighbors, occupying "holes" available for electrons of that particular energy and spin. By exchanging electrons of the same spin, the neighboring atoms align their magnetic fields. "There's actually no mystery in it," says Coey. In practice, though, calculating the effects of these electron exchanges is such a complicated mess that it has to be done with computers, and even then physicists get only ballpark answers for the properties of a particular material, which leads some of them to wonder whether Heisenberg's theory is a full picture of reality. Some researchers, for instance, suggest that the electrons responsible for a permanently magnetic state might not be attached to the metal atoms at all. They might be the same itinerant electrons that allow the metal to conduct electricity. One thing is certain: Neither Heisenberg's theory nor anyone else's for the moment can explain Tatiana Makarova's magnets. According to Heisenberg, carbon should not be magnetic, and of course the ordinary stuff isn't. Makarova starts with buckyballs, which are extraordinary enough, and then she squeezes them at pressures in the range of a million pounds per square inch and heats them to more than 1,300 degrees Fahrenheit. Under these conditions, the buckyballs fuse into a polymer, a layered latticework similar in structure to graphite, except that the units in a graphite lattice are individual carbon atoms rather than 60-atom buckyballs. Makarova began working with buckyballs in 1994, trying to see if she could get them to superconduct electricity. But one day she noticed that one of her samples stuck to a permanent magnet. "If you see that, you'd say right away there is ferromagnetism there," says Pablo Esquinazi, an Argentine physicist at Leipzig University in Germany. "You don't need complicated equipment." Then again, Esquinazi and a Russian colleague named Yakov Kopelevich have such magnetism-testing equipment, things like SQUIDs (superconducting quantum interference devices). Makarova sent them some samples of her polymer. The SQUID confirmed the evidence of her senses. Because the magnetism of Makarova's polymer is very weak, it still seems likely to some physicists that her samples were contaminated, perhaps with an iron compound such as magnetite. "I suspect there is no intrinsic magnetic carbon," says Coey, who recently found evidence of such contamination in organic compounds from a meteorite that were thought to be magnetic. "Magnetite is everywhere in the air," he adds. Makarova responds that she repeatedly tests two polymer samples in each batch for iron and finds the concentration to be "vanishingly small"; that unpolymerized or depolymerized buckyballs show no signs of magnetism, as you would expect them to if they were contaminated; and that the only samples that are magnetic, in fact, are the ones that have been processed at certain combinations of temperature and pressure. The possibility that her results are due to contamination, Makarova says, "is approximately equal to the possibility that a monkey at a computer will type a Shakespearean sonnet." And anyway, labs in Britain and Japan recently reproduced her results. A more likely possibility, says Makarova, is that once physicists understand magnetism better, magnetic carbon will not seem so outlandish. Which is why she has spent long hours making sample after sample herself (until recently, she had no assistants). "It's not easy," she says. "I'm just working as an engineer, trying to find out where the magnetism comes from."
Buckyball Stick-to-itivenessTatiana Makarova's carbon magnets are made from buckyballs that have been fused into a polymer that has a layered, asymmetrical structure, as seen under an electron microscope, below. What makes the polymer magnetic remains a mystery. One hypothesis is that the 1-million-pounds-per-square-inch pressure needed to make the polymer collapses some of the buckyballs, thereby generating unpaired electrons; another is that the buckyballs remain intact, but unpaired electrons arise at the bonds between them.
A buckyball is a hollow molecular structure formed by 60 atoms of carbon. Graphic by Matt Zang
Access a brief NASA tutorial on magnetism at www-istp.gsfc.nasa.gov/Education/Imagnet.html.
For a short intro to buckyballs, visit www.mpi-stuttgart.mpg.de/andersen/fullerene/intro.html.
The Exploratorium offers a dozen experiments you can do at home to demonstrate magnetism fundamentals: www.exploratorium.edu/snacks/iconmagnetism.html.
The metals that appear on page 62 were obtained from Alfa Aesar, a Johnson Matthey Company, a leading manufacturer and supplier of research chemicals, metals, and other materials. The Alfa Aesar Research Chemicals, Metals, and Materials Catalog offers more than 20,000 products, including high-purity inorganics, organics, pure metals, and alloys. Their complete product line is available on the Web at www.alfa.com. Contact Alfa Aesar at 800-343-0660.
