<>
Researchers have identified a new class of materials that could address the long-standing conductivity limitations of copper in microchip interconnects. As transistors continue to shrink, copper wires face significant performance degradation once their diameter drops below 40 nanometers, a phenomenon caused by increased electron scattering. According to a study published July 16 in the journal Science, topological materials—specifically Weyl semimetals—may offer a path forward by leveraging exotic quasiparticles to maintain or even enhance electrical flow as wire dimensions decrease.
Interconnects serve as the essential wiring that connects transistors into functional circuits. For decades, copper has been the industry standard due to its high conductivity. However, this effectiveness is tied to the metal’s mean free path—the average distance an electron travels before colliding with an atom. When wires are manufactured below this mean free path, electrons collide with the wire’s boundaries more frequently, causing a sharp rise in resistance. While alternative metals like cobalt and ruthenium have shorter mean free paths—approximately 10 nanometers and 6 nanometers, respectively—they eventually encounter similar physical scaling limits.
The research team, led by scientists investigating topological insulators and semimetals, suggests that the unique structural properties of these materials could bypass traditional scaling hurdles. Topology, a branch of mathematics concerned with properties that remain stable despite deformation, provides the theoretical basis for these materials. In topological insulators, electrons are protected from scattering as they move along the material’s surfaces. Weyl semimetals extend this capability, remaining conductive in their interior while simultaneously hosting protected conductive states on their surfaces.
Performance of Niobium Arsenide Nanowires
To test the potential of these materials, researchers synthesized nanowires of the Weyl semimetal niobium arsenide using a technique known as thermomechanical nanomolding. The study found that these nanowires, measuring between 40 nanometers in width and several micrometers in length, exhibited a decrease in resistivity as their diameter was reduced. Specifically, a 40-nanometer wire showed approximately 70 percent lower resistivity compared to bulk single crystals at room temperature, an improvement attributed to the increased contribution of surface conduction.
Judy Cha, a professor of materials science and engineering at Cornell University, noted the practical nature of these findings. “We do not need superpristine or high-quality samples or superlow temperatures or high vacuum to see the quantum effects,” Cha stated in the research reporting. “Our measurements were done at room temperature in low vacuum or in air. This to me is astonishing.”
While the 40-nanometer niobium arsenide wires outperformed cobalt and ruthenium at the same scale, they did not yet surpass the conductivity of state-of-the-art 10-nanometer copper interconnects. However, the researchers project that if these nanowires are scaled down to approximately 12 nanometers, the surface conduction contribution will dominate, theoretically allowing them to outperform copper. Beyond conductivity, the material demonstrated stability in air and high thermal conductivity, suggesting potential for managing heat dissipation in dense circuit environments.
Pathways to Practical Integration
Despite these promising results, significant barriers remain before Weyl semimetals can replace copper in commercial semiconductor manufacturing. The use of arsenic presents toxicity challenges, and the current thermomechanical nanomolding synthesis technique is not yet compatible with standard CMOS (complementary metal-oxide-semiconductor) or back-end-of-line industrial processing conditions. These manufacturing requirements are critical for the integration of new materials into existing silicon-based production pipelines.
The industry is nevertheless exploring topological materials as potential successors to traditional metals. During the June IEEE/JSAP VLSI Symposium in Honolulu, technology companies presented research on various topological candidates, including molybdenum phosphide from Samsung and cobalt monosilicide from IBM. These presentations suggest that major semiconductor firms are actively monitoring the development of these materials for future nodes.
Cha emphasized that the proof-of-principle work indicates these materials are more than just academic models. Even with structural defects, the enhanced conduction properties of Weyl semimetal nanowires remain observable, making them a “compelling and realistic engineering” option for future hardware development. The next steps for the field involve refining synthesis techniques to ensure compatibility with high-volume manufacturing environments and evaluating the long-term reliability of these materials under operational stress.
Readers interested in the ongoing development of microchip materials can follow updates from the IEEE/JSAP VLSI Symposium or review forthcoming publications in Science regarding the scalability of topological semimetals. Please share your thoughts on the future of semiconductor materials in the comments section below.