The advancement of semiconductor technology has focused on reducing the size of transistors. This is because the performance of semiconductors is determined by how many devices and circuits can fit inside a chip. A typical example is Moore's Law, which states that the number of transistors that can be integrated into a semiconductor integrated circuit doubles every two years. However, as technology continues to advance, we have entered an era where simply reducing transistor size can no longer achieve significant innovation.
On Dec. 13, Professor Kim Jun-seok of Hongik University and Professor Kim Ji-hwan of the Massachusetts Institute of Technology (MIT), along with a joint research team from Samsung Advanced Institute of Technology (SAIT), published a review paper in the international journal Science, stating, "To create next-generation semiconductors, it is necessary to go beyond the limitations of wiring technology, particularly wiring materials, along with transistors." The review highlighted that wiring materials are as crucial to semiconductor performance as reducing the size of transistors.
On the last day of 2024, Dec. 31, Professor Kim Jun-seok, who met at Hongik University, compared the wiring connecting the transistors to the "blood vessels" of semiconductors. Just as human blood vessels gradually thicken from capillaries to arteries, the wiring connecting transistors varies from thin to thick, and making it efficiently thin is essential for increasing the integration density of semiconductors.
Professor Kim noted, "In semiconductor scaling, discussions usually focus only on reducing the size of transistor silicon, but the problem of increased resistivity due to decreasing wiring width hardly gets attention," adding, "This research sheds light on issues related to wiring." Resistivity is a measure of how much a material impedes the movement of electrons. As metal wiring becomes thinner, the resistivity increases, and as resistivity rises, the movement of electrons weakens, resulting in decreased information transmission efficiency.
Currently, copper is primarily used for semiconductor wiring. However, as the wire width narrows, the resistivity of copper increases exponentially. This leads to issues with signal delay and power consumption. Professor Kim explained, "The size of transistors is at the level of 24-25 NANO (nanometers, one-billionth of a meter), and excluding insulators and other components in between, the spacing between wirings is narrower than 10 NANO." He added, "With existing materials like copper, there are limits to improving efficiency as the wiring width decreases."
He emphasized that the development of new wiring materials is essential. Professor Kim pointed out, "The industry is experimentally introducing materials like ruthenium and cobalt, but these are only short-term alternatives," stating that "more innovative materials will be needed in the long term."
This paper is closely related to why Professor Kim moved from the industry to academia. In September of last year, he transitioned from Samsung Advanced Institute of Technology to Hongik University. Based on his experiences at Samsung, he began an academic challenge to explore a new semiconductor paradigm. Professor Kim expressed, "In corporations, there is often a focus on short-term results rather than long-term research," stating that "academia is the best place to explore new paradigms." He plans to begin full-scale experiments this year to research next-generation wiring materials and semiconductor materials.
Professor Kim is confident that the three-dimensional stacking technology that allows another semiconductor to be placed directly on top of a semiconductor, along with wiring technology, will lead next-generation semiconductors. However, he stressed that close cooperation between academia and industry is vital to achieving this.
Professor Kim mentioned, "Although there is ongoing research on wiring materials in academia, some materials have a good resistivity value but poor thermal stability, making them impractical for real use, or due to a lack of understanding of the processes, it is challenging to apply them in practice." He noted, "We need materials that can withstand wiring processes around 400 degrees Celsius while having low resistivity and can endure chemical reactions occurring during the wiring process. To verify this, close collaboration between academia and industry is necessary."
Reference materials
Science (2024), DOI: https://doi.org/10.1126/science.adk6189