In the past 30 years, transistors on chips have shrunk about a millionfold, while the magnets on those chips have shrunk barely a thousandfold, leaving them the biggest, heaviest, hottest, least-efficient components on many a circuit board.
”They take up the largest real estate in computers. They’re the size of cellphones in cellphones,” says Danny Xiao, chief technology officer at nanomaterials firm Inframat Corp. in Farmington, Conn.
Now a company that’s about to be spun off from Inframat says it can shrink magnets by at least another tenfold, and perhaps even a hundredfold, by making them out of composites loaded with magnetic particles only billionths of a meter wide. If, as seems possible, such nanocomposites can be manufactured directly on chips, engineers could design computers that are smaller and, equally important, cooler than anything available today.
The company, Embedded Nanomagnetics, also in Farmington, says it already has partnerships with an electronics industry giant, a leading cellphone manufacturer, and two aerospace defense contractors, which it declines to name because of nondisclosure agreements. The company expects to license its technology to at least one firm within the next 24 months.
Magnetics are crucial, because they regulate the flow of power, ensuring that each component in a device gets the appropriate voltage, as well as isolating circuitry from potentially harmful spikes in current. Also, by routing incoming and outgoing signals, they allow antennas to serve as both receivers and transmitters.
The smallest magnetic elements in today’s devices are about 2 millimeters in diameter and maybe half a millimeter thick, says George Schaller, a ferrite materials consultant in Bluffton, S.C. John Ings, technical director at Ceramic Magnetics, in Fairfield, N.J., adds that ”there is a real need to make magnetics smaller,” and the government knows it.
A decade ago, the U.S. Defense Advanced Research Projects Agency, or DARPA, formed an industry consortium to shrink magnetics for use in miniaturized phased-array radar. The idea was to put the miniradars inside Navy missiles so that the weapons could detect and track targets without the aid of ship-mounted radar. The project never saw the light of day, Ings says, because the magnetics of the time could be made only at temperatures so high they would damage the chips.
To make a small but powerful magnet, it’s necessary to make its atoms cooperate with maximum efficiency. When the atoms’ individual magnetic fields are arrayed at random, they cancel out, leaving no net magnetic field. However, when atoms with just the right electronic properties interact, their magnetic fields can align in the same direction, a quantum effect known as exchange coupling. In conventional magnetic particles, the magnetic field usually breaks up into smaller volumes, called magnetic domains. The domains tend to work against one other, weakening the field.
The new composites solve this problem by using magnetic particles with diameters of just 20 to 50 nanometers—roughly a tenth a wavelength of visible light. That way, each nanoparticle is smaller than its magnetic domain, so there can therefore be no mutual canceling effect. The nanoparticles are packed into a highly electrically resistive insulating organic polymer matrix, which keeps them from clumping. The polymer also keeps the nanoparticles close enough, at roughly 25 nm apart, to interact by means of exchange coupling and thus to line up their magnetic fields.
The composites are not easy to make, notes materials engineer Pulugurtha Markondeya Raj, of the Georgia Institute of Technology, in Atlanta. Widely used epoxies that are cheap and easy to work with cannot serve as the polymer matrix, because their electrical properties lead to high power loss. Other polymers that are right electrically often are neither strong enough nor tough enough. Still others have the desirable properties, but lose them when loaded with nanoparticles.
Embedded Nanomagnetics plans to get around the manufacturing problem by developing pastes and films that can be deposited at low enough temperatures to spare the electronics on a chip, That way, fabricators could build the magnetics on microchip assembly lines instead of soldering them on afterward, as they do now.
The initial offering, a nickel-zinc-ferrite magnetic nanoparticle-based paste, is due out in six months to a year. The company says the paste can be formed to make devices as strong as the conventional magnets in electronics, but a seventh to a tenth the size. A second-generation product, a thin film incorporating cobalt-silicate nanoparticles, could potentially show a further hundredfold improvement.
Pulugurtha says the composites could ”tremendously benefit” applications in telecommunications, computing, and aerospace. Their possible use in advanced microwave antennas and ultra-high-frequency radio communications has drawn DARPA’s attention. DARPA, the U.S. Air Force, the National Science Foundation, and NASA have together provided Inframat with US $6 million worth of grants over six years to develop the composites.
Other researchers are investigating thin magnetic films and nanomagnetics as well, notably John Xiao at the University of Delaware, in Newark. However, Inframat’s Danny Xiao (no relation to John) says that his company’s patents cover composites made from magnetic nanoparticles isolated by highly electrically resistive insulators.
The Intels of the world could use these miniature magnets to displace a large number of tiny power supplies across a typical motherboard. ”This could mean you could put power supplies where you want them,” Schaller explains. ”By eliminating a large, bulky power supply and its bulky grid, you eliminate a large source of heat, which also could have significant benefits.” For one, we wouldn’t have to lean over such hot laptops all day.
About the Author
CHARLES Q. CHOI is a freelance science writer based in New York City.