Enhancing Materials for Hi-Res Patterning to Advance Microelectronics
August 29, 2019 | Brookhaven National LaboratoryEstimated reading time: 5 minutes
Conventionally, the microelectronics industry has relied upon optical lithography, whose resolution is limited by the wavelength of light that the resist gets exposed to. However, EBL and other nanolithography techniques such as extreme ultraviolet lithography (EUVL) can push this limit because of the very small wavelength of electrons and high-energy ultraviolet light. The main difference between the two techniques is the exposure process.
“In EBL, you need to write all of the area you need to expose line by line, kind of like making a sketch with a pencil,” said Tiwale. “By contrast, in EUVL, you can expose the whole area in one shot, akin to taking a photograph. From this point of view, EBL is great for research purposes, and EUVL is better suited for high-volume manufacturing. We believe that the approach we demonstrated for EBL can be directly applied to EUVL, which companies including Samsung have recently started using to develop manufacturing processes for their 7 nm technology node.”
In this study, the scientists used an atomic layer deposition (ALD) system—a standard piece of nanofabrication equipment for depositing ultrathin films on surfaces—to combine PMMA and aluminum oxide. After placing a substrate coated with a thin film of PMMA into the ALD reaction chamber, they introduced a vapor of an aluminum precursor that diffused through tiny molecular pores inside the PMMA matrix to bind with the chemical species inside the polymer chains. Then, they introduced another precursor (such as water) that reacted with the first precursor to form aluminum oxide inside the PMMA matrix. These steps together constitute one processing cycle.
A schematic showing the process of creating the hybrid organic-inorganic resist through infiltration synthesis, patterning the resist via electron-beam lithography, and etching the pattern into silicon by bombarding the silicon surface with ions of sulfur hexafluoride (SF6).
The team then performed EBL with hybrid resists that had up to eight processing cycles. To characterize the contrast of the resists under different electron doses, the scientists measured the change in resist thickness within the exposed areas. Surface height maps generated with an atomic force microscope (a microscope with an atomically sharp tip for tracking the topography of a surface) and optical measurements obtained through ellipsometry (a technique for determining film thickness based on the change in the polarization of light reflected from a surface) revealed that the thickness changes gradually with a low number of processing cycles but rapidly with additional cycles—i.e., a higher aluminum oxide content.
“The contrast refers to how fast the resist changes after being exposed to the electron beam,” explained Chang-Yong Nam, a materials scientist in the CFN Electronic Nanomaterials Group, who supervised the project and conceived the idea in collaboration with Jiyoung Kim, a professor in the Department of Materials Science and Engineering at the University of Texas at Dallas. “The abrupt change in the height of the exposed regions suggests an increase in the resist contrast for higher numbers of infiltration cycles—almost six times higher than that of the original PMMA resist.”
The scientists also used the hybrid resists to pattern periodic straight lines and “elbows” (intersecting lines) in silicon substrates, and compared the etch rate of the resists with substrates.
Left: A scanning electron microscope (SEM) image of silicon elbow-shaped nanopatterns with different feature sizes (linewidths). Right: A high-magnification SEM image of high-resolution, high-aspect-ratio silicon nanostructures patterned at a pitch resolution (linewidth plus spacewidth, or space between lines) of 500 nm.
“You want silicon to be etched faster than the resist; otherwise the resist starts to degrade,” said Nam. “We found that the etch selectivity of our hybrid resist is higher than that of costly proprietary resists (e.g., ZEP) and techniques that use an intermediate “hard” mask layer such as silicon dioxide to prevent pattern degradation, but which require additional processing steps.”
Page 2 of 3
Subscribe
Stay ahead of the technologies shaping the future of electronics with our latest newsletter, Advanced Electronics Packaging Digest. Get expert insights on advanced packaging, materials, and system-level innovation, delivered straight to your inbox.
Subscribe now to stay informed, competitive, and connected.
Suggested Items
SMTA Ultra HDI Symposium, Day 1: AI at the Core or Out of the Game
04/13/2026 | Marcy LaRont, I-Connect007It was a beautiful 81°F morning in Arizona last Wednesday as I headed to the third annual SMTA Ultra HDI Symposium, focused on AI and ultra high density interconnect technology. Strategically held as part of Arizona’s Tech Week, this year’s conference took place in Avondale in Phoenix's West Valley. The event moved from the cozy offices of the Peoria Sports Complex (which paid homage to baseball’s spring training world) to the larger Avondale Conference Center, highlighting the importance of this area for electronics manufacturing investment.
PCBAA Sets Its Priorities for 2026: Support in the House and Senate
02/03/2026 | Marcy LaRont, I-Connect007David Schild, executive director of the Printed Circuit Board Association of America (PCBAA), leads an organization focused on the domestic manufacture of PCBs and substrates to support America’s critical microelectronics supply. In the continuation of this interview, David talks about the ways his organization is aligning its priorities with Washington, in particular, the Executive Branch. It might feel like a snail’s pace, but progress could mean financial support and more domestic security, even for the smaller companies.
SMT Perspectives & Prospects: Artificial Intelligence Part 6: Data Module 1
10/07/2025 | Dr. Jennie Hwang -- Column: SMT Perspectives and ProspectsData is one of the six pillars of AI infrastructure. It is critical to the performance of artificial intelligence (AI) models. AI data, essential to both the training and inference of Generative AI models, connotes the datasets used to train, validate, and test AI models. Training data provides models with a frame of reference by establishing a baseline against which models can compare new data using pre-trained models for predictions or generating new content.
U.S. Uses Secret Trackers to Trace AI Chips Diverted to China, Sources Say
08/18/2025 | I-Connect007 Editorial TeamTwo sources told Reuters that U.S. authorities have secretly placed location trackers in some advanced chip shipments they see as at high risk of illegal diversion to China. They said the trackers are intended to locate AI chips that are sent to locations restricted by U.S. export laws, but authorities only examine some shipments.
Happy’s Tech Talk #38: Novel Metallization for UHDI
05/07/2025 | Happy Holden -- Column: Happy’s Tech TalkI have been involved in high-density electronics substrates since 1970 when I joined Hewlett-Packard’s RF semiconductor group after college. Figure 1 shows the difference between trace/space lithography for substrates and silicon starting in 1970. My projects involved sapphire circuits for RF devices, but the figure displays the state of PCBs and integrated CMOS circuits and their packaging, not discreet RF devices. Even then, semiconductors were 50X higher density.