Minisymposium

As transistor dimensions have progressively shrunk over the years, X-ray microscopy resolution has also continued to improve to meet the demands of the semiconductor industry. Although electron microscopy can achieve higher imaging resolutions, X-rays offer unique advantages, including non-destructive 3D imaging of fully intact integrated circuit (IC) dies and imaging under operando conditions. In this presentation, I will introduce ongoing innovations of synchrotron-based X-ray microscopy, with a focus on a technique called ptychographic X-ray computed tomography. Our recent work addresses nanometer-scale experimental instabilities and depth-of-field limitations, enabling an unprecedented 4-nanometer resolution achieved during integrated circuit imaging. With the upcoming SLS upgrade, X-ray microscopy performance is expected to dramatically improve. This presentation will discuss the implications of these advancements and the potential of X-ray microscopy for nanoscale metrology of integrated circuits.

4th generation synchrotron sources provide two orders of magnitude more coherent photons, and thus the ability to collect coherent X-ray imaging datasets faster and/or with a higher resolution. Consequently, the increased volume of data requires dedicated tools to fully take advantage of the improved coherent flux. PyNX [https://pynx.esrf.fr] is developed at ESRF - it has been written from the ground up to provide state-of-the-art performance using optimised GPU programming, both in terms of speed and memory requirements (to handle larger datasets). It is used on multiple beamlines notably at ESRF, Soleil, Petra-III, TPS, with scripts for data analysis easily expandable for new instruments (only the data input functions need to be updated). Input/output using the standard CXI format is also supported.

X-ray microscopy via ptychographic imaging has been growing rapidly in the last decade. Where early experiments could only be performed by a few experts groups with their own specialized software implementations. Nowadays ptychographic experiments can be performed at plenty of beamlines and various open source software packages are being used and refined by collaborations across singular groups. With the increased availability and wide spread application of X-ray ptychography the initial challenge of "making it work robustly" has been overcome. New challenges are all connected to every users wish to image more, meaning faster as well as larger samples. On the experimental level these wishes are being address by pushing towards beamlines with more photons that utilize bigger and faster detectors. On the data reconstruction end this push towards more and more extreme data pushes the implementation of the reconstruction algorithms to their limits. Hardware architecture, suitable algorithm design an implementation start to dictate what datasets can and can not be reconstructed. In this talk the problem of "extreme data" in ptychography is dissected, current computation limits are explore and various ideas from the community on tackling these issues are presented and discussed.

X-ray ptychographic tomography is now routinely used at synchrotron facilities around the world, producing nanoscale resolutions in 3D. However, the acquisition times still lag significantly behind other imaging methods. As many synchrotrons upgrade to diffraction limited rings, the community is being presented with a dramatic increase in the coherent flux available. There have many developments in recent years, from multibeam illuminations to fly scanning trajectories. However, translating the increase in flux into increased scientific throughput remains a challenge. The latest advances in high-speed ptychography at I13-1: Coherence of the Diamond Light Source have combined a novel acquisition scheme with the SELUN detector from DECTRIS, which is capable of 120kHz in continuum mode, to achieve ptychography collection rates of over 100 kHz. This is providing sub-second projections and sub-hour 3D ptychographic tomography. We present the latest technical developments as well as their applications in the fields of battery materials and brain imaging.