Materials and processes for future microelectronic systems was the focus of a conference held at Pacific Northwest National Laboratory (PNNL), featuring several scientists from the U.S. Department of Energy's (DOE) Princeton Plasma Physics Laboratory (PPPL).
The summer meeting in Richland, Washington, brought together scientists affiliated with the Extreme Lithography & Materials Innovation Center (ELMIC), one of three 2024 Microelectronics Science Research Centers funded by DOE.
ELMIC's research is centered on the science and technology of creating innovative materials with very small dimensions while understanding their underlying physical properties. The center focuses on plasma-based nanofabrication, extreme ultraviolet (EUV) photon sources, 2D-material systems and extreme-scale memory. ELMIC's work includes projects led by PPPL, PNNL, Argonne National Laboratory and Lawrence Livermore National Laboratory.
Yevgeny Raitses, a managing principal research physicist and the lead principal investigator (PI) of the Plasma-enabled 2D Materials for Energy-efficient Microelectronics (PlasMat2D) project - which is part of ELMIC - provided an overview of this project. Igor Kaganovich, a principal research scientist and co-PI of the Plasmat2D project, presented on molecular dynamics modeling. In addition, Barry Rand, co-director of the Plasmat2D project and professor at Princeton University's Department of Electrical and Computer Engineering and Andlinger Center for Energy and the Environment, and jointly appointed at PPPL, ran the industry discussion panel at the meeting. Other presenters from PPPL included Shoaib Khalid, Yerbolat Ussenov and former postdoctoral fellow Stanislav Musikhin, who is now with Applied Materials.
"This conference set the stage not only for the future of the project but also for the future research directions for microelectronics in the U.S.," Raitses said. "Advancing microelectronics requires fundamental science that only National Laboratories and universities can provide. Working with industry, we can bring forward 2D materials and memory scaling solutions that address current challenges in microelectronics progress."
The Lab's expertise in plasma is changing the future of science and industry - advancing next-generation electronics, among other applications. Plasmat2D, which Raitses leads, is key to PPPL's work with industry leaders. In addition to Raitses, Rand, and Kaganovich, PlasMat2D team members include Robert Bruce, Luxherta Buzi, Jane Chang, Vincent Donnelly, David Graves, Eric Joseph, Alice Kunin, Mark Kushner, Saien Xie. PlasMat2D partners with IBM Research's Thomas J. Watson Research Center.
Industry partners and advisers noted progress made by ELMIC teams, both at the conference as well as in the conversations and interviews that followed. Below are excerpts from those panel discussions and interviews, which highlight the importance of National Laboratories and universities, along with continued federal funding support.
Q. What would scientific success for ELMIC look like from your perspective?
Eric Joseph, IBM: "Enabling fundamental advances and insight into the ability to scale EUV costs down and to demonstrate a path to bring transition-metal dichalcogenides (TMD) materials from lab-based research to wafer-scale development. EUV is essential for logic and memory scaling, and TMDs are of high interest for future channel materials. The learning through ELMIC can provide the foundation for key advances, which lead to cost reduction and technology adoption in the future."
Catherine B. Labelle, Intel Foundry: "All of the projects in ELMIC are positioned to have the opportunity to have an impact on the semiconductor industry. EUV is in production today and will need to continue to evolve to enable shrinking dimensions but, at the same time, has to become more cost-efficient. Next-generation devices and products will need new innovations, such as 2D materials and advanced memory. Exactly how they will fit in is still a story to be written, hopefully including by the work of the teams here."
Shahid Rauf, Applied Materials: "As the research in ELMIC is being conducted in academia and National Laboratories, their biggest contributions to furthering the microelectronics technologies over the next five to 10 years would be to contribute to the fundamental science that underpins leading-edge microelectronics. This includes understanding the properties of 2D materials and understanding how they are processed (i.e., deposition, modification and etching). As plasmas are used for many of these processing steps, it is important to understand how these plasmas can be controlled better to deliver the right mix of ions and radicals and ion energies necessary for atomic-scale processing."
Satyavolu Papa Rao, NY Creates: "ELMIC has efforts in all the critical areas - EUV lithography will be important as industry goes down the road of stacked Gate-All-Around Complementary FETs and new materials (including chiral ones) that could serve in ever finer geometry interconnects. Cost, energy efficiency and throughput improvements will be important in mainline computers. As spin qubit architectures grow in maturity, EUV lithographic capabilities will be increasingly harnessed - this application will likely drive further improvements in line-edge roughness and pattern fidelity to design."
Q. For decades, the semiconductor road map was largely defined by transistor scaling. Is continued scaling still the primary objective, or has innovation shifted toward new architectures, advanced packaging and materials?
Joseph: "Critical dimension scaling is reaching the limits. The next generation of scaling is in the 'z' direction. Stacking field-effect transistors, decreasing cell size, continued advances in new materials, heterogeneous integration and the scaling of advanced packaging methods will all be the paths to continued performance improvement."
Labelle: "The center of innovation hasn't shifted rather than broadened. Transistor scaling is still a key component of semiconductor road maps, but it's now complemented by advances enabled by heterogeneous integration, advanced packaging and memory technologies. All of these are needed to advance semiconductor performance."
Rauf: "Raw scaling of transistor dimensions has slowed down significantly in recent years. However, the industry has found alternate ways to enhance performance, reduce cost and manage power consumption. Most memory and logic devices are 3D now, and the growth in the third dimension will likely continue. Heterogeneous integration is leading to further enhancement in microelectronics performance."
Papa Rao: "Heterogeneous integration and advanced packaging are clearly an important path ahead - but it will also drive new materials innovations - for example, thermal interface materials that break through current paradigms."
Q. Given that the focus of ELMIC is EUV science, 2D materials and memory scaling, what advances would have the greatest impact on industry over the next five to 10 years?
Thorsten Lill, formerly Lam Research: "The semiconductor industry is increasingly shifting toward 3D device architectures, where vertical scaling of device density delivers higher performance and more favorable economics. This trend began more than a decade ago with the transition to 3D NAND, which has proven to be one of the industry's most successful scaling innovations. Today, computer chips are undergoing a transformation. Transistors - the tiny electronic switches that power computers - are evolving from flat designs to increasingly complex three-dimensional structures. This allows manufacturers to pack more transistors into smaller spaces, making chips faster, more powerful, and more energy-efficient. A similar shift is also beginning in computer memory (DRAM), where new 3D designs could improve performance while reducing manufacturing costs. This shift does not mean that traditional planar scaling is coming to an end. As these new device architectures are adopted, critical dimensions continue to shrink and remain well below the limits of conventional lithography. At the same time, building increasingly complex 3D structures depends on plasma-based deposition and etch processes to precisely create and shape materials across multiple layers and high-aspect-ratio features. As a result, advanced patterning technologies, including EUV lithography, together with advanced deposition and etch, are essential to enable continued scaling and the successful implementation of increasingly complex 3D structures."
Q. Where should DOE National Laboratories and universities focus more of their microelectronics research efforts?
Labelle: "The role of plasma science and technology in semiconductors cannot be overstated. Therefore, I would focus significant efforts in this field."
Rauf: "I would encourage investments in furthering fundamental research in low temperature plasma-based technologies. Although industry can develop new plasma sources to meet their needs and empirically design the necessary processing technologies, they don't always have the capacity to invest in fundamental plasma diagnostics or have sufficient resources to accelerate innovation in modeling methods. Most microelectronics devices now require processing with atomic-scale precision. Understanding the fundamentals of these technologies requires close collaboration between surface diagnostics, plasma diagnostics and atomic-scale modeling of plasma processes."
Papa Rao: "DOE National Laboratories have the heft to help with the 'valley of death' issues that impede many new innovations (from academia or startups) from reaching commercialization potential."
Q. Why are research consortia like ELMIC important to the semiconductor industry?
Labelle: "These are pre-industry-competitive topics that the industry is coalescing around as leading candidates for future generation semiconductor needs. Consortia are needed to explore the fundamental science behind these topics before industry can invest and adopt them in any meaningful way."
Q. What will make collaborations between industry, universities and National Laboratories successful?
Labelle: "To support these consortia, we need to set realistic goals for them and make sure they are achievable in the time scale of the program(s). Industry expects the consortia to explore the fundamental science behind the topics and deliver meaningful results in a defined time period. Some of the main obstacles for these consortia include (1) having industry relevance and (2) creating centers of real strength/excellence, where the projects are grouped around project synergies/strengths."
Q. What advice would you give early career scientists and engineers who want their research to have a lasting impact on microelectronics?
Joseph: "Continue to be curious and maintain your critical thinking! Your career will take you places you never anticipated, with research shifting the types of materials and challenges we face…at the fundamental, development and production scales (each with their own challenges). Your fundamental critical thinking skills will drive you throughout your career and will be a guide to ask the right questions and learn. Be open to the change."
Labelle: "The wonderful thing about this field is that there are so many ways to contribute and have an impact. As an industry, microelectronics needs the novel innovations that create the groundwork on which semiconductor technologies can be defined; it also needs those ideas to be turned into realistic processes; and finally, those processes need to be refined through volume manufacturing to deliver the products that drive our world today. Think about what excites you and drives you and I'm sure there is a place you can have an impact."
Rauf: "I would strongly encourage scientists in academia and National Labs to make good connections to their peers in industry. Scientists in industry are surrounded by challenging problems that would greatly benefit from fundamental understanding, whether through experimental methods or theoretical means. They should pick some high-value problems and use the tools at their disposal to understand the basic science. Innovations in diagnostics and modeling methods usually come from academia, and they can help develop new tools that can help their industrial peers further advance their technologies."
Papa Rao: "I firmly believe that the most effective researcher is one that is knowledgeable about and expert in multiple areas and not focused on a singular area of expertise to the exclusion of all else."
