Flexoelectricity to miniaturize chips

Flexoelectricity to miniaturize chips

The revolution in information technology has traditionally been synonymous with chips pack more in less space to increase computing capacity. The famous Moore’s Law, which states that the number of transistors per chip will double every two years, is the framework within which this technological race that has maintained a frenetic pace of improvement has long been developed.

It now appears that this law is approaching its physical limits, so it becomes important alternative strategy to improve the chips that go beyond increasing the number of transistors. Nicknamed more than Moore (Moore’s), researchers around the world are trying to add new features to chips integrating smart materials still pervasive and indispensable basis of silicon.

Among the so-called smart materials, piezoelectric out for its ability to convert mechanical deformation in voltage (which can generate energy or charge a battery) or, on the contrary, change shape when a voltage is applied (which can be applied For example, in the design of piezoelectric fans chilled circuit).

Despite its interest, the integration of piezoelectricity with silicon-based technology is extremely complex. The number of piezoelectric materials is limited and more efficient are ferroelectric piezoelectric materials based on lead, with significant environmental implications. Moreover, their properties are strongly affected by temperature changes, making it difficult to use them in the context of a typical computer components that can reach temperatures up to 150 degrees Celsius.

There is, however, another type of electromechanical property that allows mimic the functions of a piezoelectric bending material rather than tightening. This feature is called flexoelectricity and although almost half a century ago is known, has generally been ignored because its effects are relatively weak, almost imperceptible at the macroscale.

When we studied this property at the nanoscale, however, flexoelectricity may be as or more important than piezoelectricity: just keep in mind that doubling a thick material requires a lot of energy, but something very small bend much easier.

In addition, the flexoelectricity offers other interesting properties: it is a universal property of all dielectrics, which means we can avoid the use of toxic lead-based materials, and is a more linear and independent ownership temperature piezoelectricity of a ferroelectric.

Within all this context, researchers from the Catalan Institute of Nanoscience and Nanotechnology (ICN2), a center of excellence Severo Ochoa located in the Bellaterra Campus of the Autonomous University of Barcelona (Spain), in collaboration with Cornell University (United States) and the University of Twente (Netherlands) have now succeeded in producing the first flexoelectric microelectromechanical system (MEMS) integrated into silicon.

They have found that in the nanoscale the advantageous characteristics are maintained flexoelectricity and also the results of the first prototype (levers that bend in response to a voltage) and are comparable to the most advanced micro-piezoelectric levers. If this were not enough, the universality of flexoelectricity implies that much of dielectric materials currently used in transistor technology are already flexoelectricos.

This work opens the door to the integration of intelligent materials and electromechanical capabilities on existing technologies. These results were published this week in the journal Nature Nanotechnology.

The project, headed by researcher and professor Umesh Bhaskar Gustau Catalan ICREA, the Nanoelectronics Group Oxides of ICN2 in Barcelona, was funded through a European Research Council (ERC) Consolidator Grant and Spanish draft National Plan Excellence Research, along with national scholarships and Cornell groups Twente.

Related Posts

No Comments

Leave a Reply