Beam coherence is one of the defining attributes of Free-Electron Lasers (FELs), enabling their unique ability to produce highly stable and precise light for advanced applications. For Extreme Ultraviolet (EUV) lithography, however, coherence must be managed carefully. Excessive coherence can create interference and speckle effects on wafers, while insufficient coherence can undermine precision and throughput. Research at Lawrence Berkeley National Laboratory (LBNL) and by companies such as EUV Tech has focused on developing the tools and methods needed to control coherence at levels compatible with semiconductor manufacturing. Erik Hosler, an observer of semiconductor technology transitions, highlights that FEL viability depends on translating these research insights into fab-ready solutions. His perspective reflects the reality that coherence control is not an abstract challenge but a practical requirement for lithography adoption.

The interplay between coherence and usability illustrates how FELs differ fundamentally from Laser-Produced Plasma (LPP) systems. Where LPP sources generate relatively incoherent light, FELs produce radiation with exceptionally high temporal and spatial coherence. This strength becomes a double-edged sword in lithography. It allows for precision but risks introducing unwanted artifacts. Managing this balance is where institutions such as LBNL and EUV Tech play a critical role, pioneering methods to measure, tune, and stabilize coherence under realistic conditions. Their work ensures that FEL-generated EUV light can meet the rigorous specifications of high-volume fabs.

Why Coherence Matters in EUV Lithography

Coherence defines how uniform a light wave is across space and time. In FELs, high coherence translates into sharp, precise radiation, but uncontrolled coherence can create practical problems for lithography. Excessive spatial coherence can result in speckle patterns that degrade wafer uniformity, while temporal coherence that is too narrow may limit flexibility with photoresists.

The stakes are high for fabs. Lithography processes demand light that is precise but also stable over billions of pulses. Even minor fluctuations in coherence can translate into defectivity or line-edge roughness, reducing wafer yield. It makes coherence management one of the most important technical considerations for making FELs usable in production environments.

Advances from LBNL

Patrick Naulleau and his colleagues at LBNL have been leading research on coherence in EUV systems. Their work explores how coherence levels can be measured with high precision and how beamline designs can be optimized to mitigate unwanted effects. Using specialized interferometry techniques, they have developed methods to quantify coherence across a range of operating conditions, providing benchmarks that guide FEL design.

LBNL’s research also highlights how coherence interacts with photoresist chemistry. By studying stochastic effects under different coherence regimes, the team has shown how tuning coherence can improve line-edge fidelity. These insights are essential for bridging FEL capabilities with the practical realities of semiconductor lithography.

EUV Tech’s Role in Measurement and Tools

EUV Tech, a company specializing in metrology and characterization, has contributed critical tools for measuring coherence in EUV systems. Its equipment allows researchers and fabs to track coherence levels in real time, ensuring that light sources remain within specification during operation. By providing both hardware and expertise, EUV Tech supports the translation of laboratory findings into fab workflows.

Beyond measurement, EUV Tech is developing diagnostic systems that integrate directly with exposure tools. These systems aim to provide feedback loops that automatically adjust coherence levels, reducing the risk of defects caused by instability. In this way, EUV Tech’s role complements LBNL’s, turning research insights into practical monitoring solutions for fabs.

Techniques for Coherence Tuning

Managing coherence is not only about measurement but also about active control. FELs can be tuned through adjustments in undulator design, seeding techniques, and beamline optics. For example, introducing partial coherence control mechanisms can reduce speckle without sacrificing overall precision. Similarly, bandwidth tuning allows FELs to match coherence levels with resist requirements, improving pattern fidelity.

Seeding strategies are among the most promising methods for refining coherence. In self-seeding, a monochromator filters part of the FEL output and re-injects it into the electron beam to narrow the bandwidth and stabilize coherence. External laser seeding, by contrast, introduces a coherent laser pulse at the start of the undulator to guide emission from the outset. Both approaches enable finer spectral control, reducing stochastic variability in resist exposure and improving uniformity across wafers. These techniques are now being evaluated to determine how well they can meet the demands of semiconductor lithography.

Advanced computational models also play a role. By simulating coherence behavior across beamlines, researchers can predict how changes in FEL settings will influence wafer outcomes. This integration of modeling with experimental validation ensures that coherence tuning strategies are reliable enough for fab deployment.

Industry Perspectives on Coherence Control

Within the semiconductor industry, coherence control is viewed as a necessary step for FEL adoption. Power and stability are important, but without precise coherence management, FELs risk producing results that are unusable in high-volume manufacturing. Industry voices consistently emphasize that coherence tuning must be automated, reliable, and transparent to fab operators.

Erik Hosler notes, “So long as that expectation is met, then Moore’s Law, in a sense, is still alive.” His statement applies directly to the coherence that fabs expect EUV tools to deliver consistent results, and only if coherence is managed properly will FELs meet that expectation. This perspective underscores why coherence management is not a technical side issue but a crucial factor in sustaining scaling.

Toward Fab-Ready Coherence Solutions

The advances of LBNL, EUV Tech, and others show that coherence management is becoming an increasingly mature discipline. Measurement tools, diagnostic systems, and tuning methods now provide a foundation for making FELs compatible with semiconductor lithography. The next step is integrating these approaches into industrial workflows, ensuring they operate seamlessly during continuous production.

Coherence management will likely define the timeline of FEL adoption. As research transitions into commercial prototypes, fabs will demand coherence systems that function automatically and reliably. If LBNL’s insights and EUV Tech’s tools can be translated into fab-ready solutions, FELs will move closer to becoming the backbone of EUV lithography. By solving coherence challenges, the industry can unlock the full potential of FELs, ensuring that advanced nodes are supported by light sources engineered for precision as well as power.