Portland, Ore. -- If you look at the surface-mounted components on a typical printed-circuit board, the large ones are likely to be the capacitors. For reasons that were not previously verified, the smaller a capacitor gets, the less capacitance it exhibits, often necessitating the use of larger capacitors to achieve the desired capacitance.
Now a materials science group at the University of California, Santa Barbara claims to have figured out why, and has concluded that using platinum or gold electrodes can allow capacitor thickness to be reduced by a factor of four.
The keys to solving the mystery are found at the interface, or the region where the capacitor's two metallic electrodes touch the insulating dielectric, said Nicola Spaldin, a professor with the Materials Department at UCSB. Apply an electric field across the dielectric, and you can store charge on the capacitors' electrodes or plates. But the dielectric's insulating ability declines faster than would be assumed for thin-film capacitors. The reason, Spaldin said, is that the interface has a few monolayers that are depolarized. Platinum or gold electrodes will minimize the effect, she said.
"In all real materials you cannot get rid of it completely, but if you use a metal with a lot of free electrons, like platinum, you can get rid of most of the effect," Spaldin said. "So our advice to small-capacitor makers is to start using different metals where electrons are more free to move around." Spaldin said the researchers cracked the case "by using a new method for calculating the properties of materials when an electric field is applied to them."
Previously there were two schools of thought as to why dielectric response falls off so fast for very small capacitors. One was that defects at interfaces play a bigger role as size declines, and that the capacitance reduction could thus be addressed by using higher-quality interface materials.
The other was that something intrinsic to the interface causes the reduction. Since Spaldin's calculations used a computer model, she could measure the parameters of a perfect interface just by inputting zero defects--the ideal interface. In that way she determined the problem to be intrinsic, and she arrived at an optimal solution.
"What we found was that for an ideal interface, you still see a suppression in the dielectric, which is basically the result of the detailed quantum mechanics of the interface," said Spaldin. "When the applied electric field polarizes the dielectric, that puts a charge on either side of the dielectric material. And because of the quantum mechanics of the metal, that charge depolarizes a few atomic layers on each side of the interface, imposing a thin region where the electric field is reversed."
Spaldin experimented with different material formulations to minimize the size of the depolarized layer at the interfaces to the dielectric. She found that the metals with the highest electron mobility, such as platinum and gold, lent some of their electrons to the depolarized region, thereby reducing its effect by a factor of four.
After experimenting with a few formulations using perovskite structures such as strontium titanium oxide (strontium titanate), Spaldin found that the electron mobility of precious metals could enable a fourfold reduction in capacitor thickness. "We found that platinum, which has a very high electron mobility, does much better than more-traditional metals," said Spaldin. "We found strontium titanate to be good, but not as good as platinum, even though a much higher-quality interface can be grown with it."
The next task
The group is moving on to new frontiers, in particular magnetic metal interfaces with dielectrics. By using magnetic metals, such as iron, nickel or cobalt, a magnetic field could be used to achieve further reductions in the width of the depolarized zone at the interface, the researchers believe.
"[With] magnetic metals, you have an interplay between the magnetism or magnetic permeability of the material and the dielectric response or electrical permittivity of the material," said Spaldin. "The new quest is to understand how dielectric depolarization can be minimized at the interface by magnetism. What extra quantum effects will we see there?"
By adjusting that interplay within the interface region, Spaldin hopes to formulate materials in an emerging class, called multiferrics, that combine magnetism with dielectric properties. Multiferric materials, when used as the dielectric, enable a magnetic capacitor to control its electrical behavior with a magnetic field.