● Although the declared technological revolution has been slow in coming, graphene could transform the digital world forever.
● Researchers are exploring other nanomaterials, such as silica, to improve computing performance.
The electronic chip of the future is about to emerge with the secret ingredient graphene. This revolutionary and much dreamed about supermaterial could push the limits of , which asserts that the power of computers grows exponentially. Since 2022, however, certain manufacturers like Nvidia consider Moore’s Law moribund. Only graphene has the potential to push the limits of microelectronics, although its production is notoriously difficult. A team of researchers has found a way to industrialize the production of graphene and to design a modified chip with this material. The team, made up of researchers from the Institute of Materials at the Georgia Institute of Technology, United States, the Institut Néel, France, and the International Center for Nanoparticles and Nanosystems at Tianjin University, China, emphasizes this undertaking’s “colossal technical challenge.”
A conductivity 150 times greater than that of silicon
This new graphene chip could revolutionize the semiconductor industry. Presented in Nature Communications in December 2022, the method studied aims to resolve the difficulties of adopting graphene as a replacement for silicon in the semiconductor industry. In the context of ever more chip-intensive uses, and ever greater and more ubiquitous connectivity needs (5G telecommunications, Internet of Things, AI, sensors, etc.), the capabilities of silicon are reaching their limit. Furthermore, silicon disperses the heat generated by an electric current poorly, which explains why devices like computers heat up.
First isolated in 2004, graphene is a so-called two-dimensional (2D) nanomaterial, composed of six carbon atoms, with the thickness of one atom. Graphene’s electrical conductivity is 150 times greater than that of silicon. In other words, graphene will revolutionize the world of microelectronics.
The importance of these two-dimensional materials lies in a new generation of ultra-small and ultra-fast transistors, said Isabelle Berbezier.
“The importance of these two-dimensional materials lies in a new generation of ultra-small and ultra-fast transistors. Integrating graphene into a conventional transistor could add remarkable properties to this device. And since we already use carbon in transistors, it wouldn’t be a problem to use graphene,” said Berbezier, a researcher at the Institut Matériaux Microélectronique Nanosciences de Provence (IM2NP), under the supervision of the CNRS, Aix-Marseille University and the Université de Toulon.
A necessary paradigm shift
But graphene has a major flaw: it is not a semiconductor, since it is devoid of the “forbidden band” (or “gap”), which makes it impossible to switch an electronic component on or off to handle digital’s 0s and 1s. The production of graphene is also problematic. Although several processes exist, they are hard to integrate into an industrial production line. To permit new materials to replace silicon semiconductors “a paradigm shift in electronics is necessary, while maintaining existing industrial production methods as much as possible,” noted the researchers in their paper in Nature Communications. This is where their research project holds promise.
Exploring other nanomaterials
In France, the IM2NP has recently taken on the subject in a project for the growth of graphene deposited on a layer of pure germanium, obtained by condensing a layer of silicon germanium. “Graphene is based on a small layer of germanium, but integrated on silicon, making it directly usable for the microelectronics industry,” explained Berbezier. Another field of research focuses on other nanomaterials such as silicene. With the thickness of a silicon atom, silicene can function as a semiconductor due to the fact that it is not planar like graphene. In the July 2022 edition of the journal Science, the Massachusetts Institute of Technology unveiled its work on cubic boron arsenide, whose notable property is that it can evacuate heat ten times better than silicon. “The best of all semiconductors?” wondered MIT. “More promises again?” we’re tempted to answer…
This “law,” set down by engineer Gordon E. Moore, cofounder of Intel, in 1965, predicted the evolution of the size and prices of microprocessors. Moore postulated that at equal cost, their complexity would double each year. In 1975, Moore specified that the number of transistors on a microchip would double every two years until reaching the size scale of atoms, which he predicted would happen in 2015.