Scientists have taken materials-by-design to the next frontier: devices-by-design at the atomic level.

A recent issue of Nature Materials introduces theoretical developments hypothesizing that there is a way to overcome a big problem when one tries to shrink nanocapacitors. The authors of the paper believe  they can reverse the effects of a “dead layer” that is normally found in nanocapcitors, and believe they can reengineer those characteristics into memory storage capabilities that will lead to greatly improved computer memory, transistor density and speed.

Ferroelectric memory can offer powerful speed, density and permanent retention. However, shrinking capacitors to the nanoscale heretofore has lead to its degradation due to the presence of this dead layer at the interface between the thin film and electrodes.

Massimiliano Stengel

Until now, that pesky dead layer has acted as a nonferroelectric material, lowering capacitance. Massimiliano Stengel and his team have mathematically demonstrated that the dead layer can be rendered harmless in ultra thin capacitors to form an “undead layer.” In fact, the group says that by choosing a particular electrode make up, the undead zone actually can be made to improve capacitance.

Prior thinking regarded the dead layer as a defect caused by impurities. Stengel now confirms that this intrinsic effect exists even in a perfect interface. Further research is needed to determine optimum electrode and interface materials. But, their studies show that when using BaTiO3 ferroelectric film, a switch from SrRuO3 to platinum, the dead layers between film and electrode become undead and actually improve the capacitance of BaTiO3. The added capacitance would provide better stability to a nanocapacitor used as an electronic component.

They conclude that the role of these zones depend more on the force constant of the metal oxide bonds. “This result opens exciting new avenues for engineering the electrical properties of thin-film devices by exploiting the rich chemistry of metal-oxide interfaces. By using appropriate growth conditions, it may be well within the realm of possibility (although certainly challenging) to directly control the structure and termination of perovskite/simple metal interfaces, and thus take advantage of the effects described here.”