An Ultra-Compact X-ray Free-electron Laser for Ptychographic Tomography-based Inspection
J. Rosenzweig, UCLA (Principal Investigator)
Z. Huang, SLAC National Accelerator Laboratory (Co-Investigator)
J. Maxson, Cornell University (Co-Investigator)
T. Hodgetts, RadiaBeam Technologies Co-Investigator)
The ultra-compact X-ray free-electron lasers, or UC-XFEL, is a 5th generation light source – a coherent X-ray source based on high gradient accelerator technologies originally pioneered for exploring the high energy physics frontier. The UC-XFEL program has identified a pathway to obtain an X-ray FEL with a footprint and cost (<40 m including X-ray beamlines; <$35M) easily permitting distributed coherent X-ray FEL sources for research, medical, industrial and security applications. Interest is high in the approach, which is based on very high gradient electron RF gun sources and accelerators – both employing an emerging technique using cryogenic operation. In this cryo-RF approach, the sustainable electric fields are increased by over a factor of 2.5 over present technology. Together with the use of very short period magnetic undulators and novel compression methods creating intense femtosecond electron beams., this scheme permits the shrinking the length of the XFEL by an order-of-magnitude, while maintaining a highly capable, high-flux XFEL. The details of a fully realized 1 GeV design, aiming at a 1 nm operating wavelength, have been published in a highly cited article three years ago
The UC-XFEL introduced continues to attract attention, due to the challenges of understanding the intricate technology and experimental approaches needed. Cryo-RF methods stand out for their potential wide range of applications, and a strong development effort to create new, more efficient and less costly designs continues apace in US laboratories. The key enabling part of the UC-XFEL scheme is the cryogenic RF photoinjector, which promises the needed 6D beam brightness, is under intense R&D at UCLA. Parallel, collaborative efforts on cryogenic photoemission and high field operation are currently underway. Likewise, development of the techniques needed for advanced imaging methods based on the UC-XFEL continues vigorously.
A strong science case is emerging, with a particular demand for harder X-rays, to the Angstrom-level. In particular, the collaboration engaging the research under this proposal has been in close contact with representatives of the US semiconductor industry, which is interested in new, high-resolution metrology and inspection methods. The issues behind this are both technical and strategic; with continuing downsizing of chip critical dimensions, dramatically more capable metrology is needed.
Recent impressive investments in the semiconductor sector made through the 2021 CHIPS Act underscore the importance of this work. Every new generation of chips present frontier challenges in the fabrication ecosystem, which are now at a level requiring serious research initiatives. One of the key hurdles to confront in advancing to the next level of chip performance is that of metrology – the need to measure extremely small features in a more robust way. Indeed, after a series of community studies coordinated by NIST, the critical challenge of metrology for next-gen chip manufacturing and validation was emphatically identified, with an urgent call to action described in a series of reports. While the needs for metrology in current chip architectures are adequately addressed by present methods, they will be obsolete soon, as they rely on a time-consuming and destructive process, involving the stepwise removal of chip layers. As such, these methods, and proposed extensions, are not qualified candidates for meeting the future demands of chip metrology. A much finer resolution, non-destructive, and orders-of-magnitude faster approach is demanded.
We have studied such an approach, which is based on the technique of ptychographic laminography. This method has been tested at storage rings, and has given extremely promising results, showing zoom-capable imaging to sub-20 nm resolution, albeit at 60 hours needed per scan. These path-breaking experiments were organized by collaborator A. J. F. Levi at USC. To scale the flux up to level (1E5 times) needed for a paradigm-shifting instruments, we have recently created a design for a very high-flux UC-XFEL operating with 7 keV photons, reaching this flux using a regenerative amplifier. This design is now published in Instruments and, like its predecessor in the soft X-ray regime, it is attracting much attention. The program initiated in this work has been endorsed by Intel scientists, and future funding is being proposed through NIST as well as private investors.
This proposed research in this grant is aimed at a comprehensive study of the conceptual design, with preliminary engineering and costing, of a UC-XFEL instrument which is fully capable of translational application to ptychographic laminography-based chip inspection. This effort is paired with a direct study of the performance and limitations of the UC-XFEL for this application. The success of this initiative may not only have a direct impact on a critical US industry, but also provide a powerful 3D imaging tool to characterize mesoscale materials such as high entropy alloys, aiding national security efforts. Indeed, chip metrology has a key importance for future economic security in this country through both inspection in fabrication, and validation of chips of untrusted provenance. This is amplified by the recent demonstration of a sudden and continuing rise of the critical role enabled by AI/ML in the modern economy, and the attendant needs for a renewed U.S. chip infrastructure.
