Antoine Wojdyla

We describe an optical metrology tool for the measurement of X-ray gratings, demonstrated for pattern densities up to 600 lines/mm. Reflection gratings illuminated at glancing angles of incidence are the central dispersive elements used in many short-wavelength monochromators, spanning the energy spectrum from the vacuum ultraviolet to tender X-ray. In modern designs, gratings can be hundreds of mm long, and for high performance, their line-placement accuracy requirements are below 50 nm, a small fraction of the grating spatial period. Gratings with uniform and variable line spacing are both important in X-ray optics. We describe line-by-line pattern measurement using a focused 405 nm laser beam and a distance-measuring interferometer, achieving measurement precision as low as 11 nm rms and speeds up to 0.07 mm/s. Periodic variations in speckle intensity reveal the motion of the grating pattern.

We present the development of a digital twin framework for the Advanced Light Source, were digital twins of beamlines can be accessed using API calls and twin experiments can be run using the bluesky experiment framework. We model each optical systems with virtual motors and virtual sensors, and we use a beamline ontology to switch between simulation engines.

The Advanced Light Source Upgrade (ALS-U) will deliver a diffraction-limited storage ring, which places stringent demands on beamline optics. We present a scripted simulation pipeline that predicts beamline performance from measured mirror figure errors and apply it to the COSMIC-U (7.0.1) and MAESTRO-U (7.0.2) insertion-device beamlines. Ray-tracing models built with shadow4 in OASYS 2.0 are exported as Python scripts and turned into parameterized builders, which bring the mirror metrology of the ALS X-Ray Optics Laboratory, retrieved by unique identifier from a Bluesky-Tiled server, and the measured influence functions of the adaptive x-ray optics (AXOs) into one pipeline. With ideal optics, energy scans across every grating agree with analytical estimates of the beam size and photon energy resolving power. Before the AXO correction, including the measured figure errors mainly shifts the horizontal focus away from the exit slit, while the resolving power stays above its design target at every energy; grating figure errors are not yet included. Fitting the AXO voltages directly to the beam at each photon energy keeps the horizontal width within 0.6% of ideal on COSMIC-U and slightly below it on MAESTRO-U, with no actuator above 300V. Including the measured figure errors and the AXO correction, both beamlines are therefore expected to perform close to their design. In general, this approach offers a reproducible way to use as-built optic data in performance predictions during construction, and a starting point for a beamline digital twin.

With diffraction-limited storage rings, the power load on the pre-mirror in grating monochromators can cause deformation that adversely affects the effective photon-energy resolution of the beamline. For variable included angle with variable line-spacing (VIA-VLS) grating monochromators, it is possible to compensate for some deformation effects and achieve nominal resolution by adapting the relative angles of the pre-mirror and grating across the entire energy range. We derive a simplified analytical formulation for grating and mirror positions that can be used for automatically tuning photon energies in simulations and digital twins. We also demonstrate the use of the open-source Rigorous Coupled-Wave Analysis software Reticolo for soft X-ray gratings to compute diffraction efficiency using Python for lamellar, blazed, and arbitrary groove profiles, including higher-order diffraction efficiencies. We provide experimental comparisons from various gratings mea sured at the Advanced Light Source, showing agreement well within 5%. These calculations facilitate grating design and enable more effective prediction of flux and harmonic contamination.

AI agents are increasingly being explored for scientific instrument operation, but their reliability under varied user requests has not been systematically established. Here, we develop an assertion-based evaluation framework for a beamline AI agent. The framework combines structured test-case design with predefined binary assertions to probe agent behavior and potential failure modes. We demonstrate the framework using 44 representative test cases at ALS Beamline 7.0.1.2 and evaluate the reliability of the AI agent when using different LLMs. Preliminary results show substantial performance differences across LLMs and identify benchmark tasks in which the agent still fails. These evaluation results can guide model selection, agent refinement, and continuous regression testing toward more reliable beamline AI agents.

Modern synchrotron beamlines support increasingly complex experiments, but their operation remains strongly dependent on facility-specific control interfaces and workflows. This reliance creates a significant barrier to entry for new users and limits the portability of experimental procedures across beamlines. Here, we present a minimal, agent-agnostic orchestration architecture for AI-assisted beamline operation, deployed at Advanced Light Source beamline 5.3.1. The approach introduces a capability-based abstraction layer that exposes beamline functionality as a set of predefined operations. Within this framework, AI agents translate natural-language user requests into structured experimental plans composed of these capabilities. All plans are executed through the Bluesky Queue Server , ensuring deterministic operation within existing control and safety constraints without modification of the underlying beamline control system. The framework is defined by three key elements: (i) a portable capability layer that decouples user intent from beamline-specific implementation, (ii) a hybrid interaction framework combining AI-assisted workflow composition with graphical user interface (GUI)-based monitoring and control, and (iii) constrained execution that restricts all actions to predefined capabilities and requires user approval prior to execution. We demonstrate the architecture using grazing-incidence scattering (GISAXS), multi-edge X-ray absorption spectroscopy (XANES), and cross-beamline deployment scenarios. In each case, the system generates physically meaningful experimental plan, adapts to local hardware constraints, and preserves experimental intent across different instruments. These results demonstrate a practical and scalable approach for integrating AI into synchrotron experimentation, enabling intent-driven, portable, and safe beamline operation.

Extreme ultraviolet (EUV) imaging for photomask metrology requires quantitative phase imaging, accurate control of the pupil-fill and wafer-side through-focus behavior. We explore Fourier Ptychographic Microscopy (FPM) as a solution, addressing challenges such as field-dependent aberrations outside of the microscope’s designed field-of-view (FOV). Using binary pseudorandom mask patterns, we achieve 1.8× resolution enhancement beyond the 0.5 λ/NA diffraction limit, improving the resolved feature size from 82 to 46 nm across a 30 μm×30 μm FOV. We investigate various reconstruction algorithms for their robustness under spatially varying aberrations from the zoneplate imaging system. FPM’s phase reconstruction also enables arbitrary illumination synthesis, allowing for the computational emulation of diverse illumination patterns from a single FPM dataset. These results demonstrate FPM’s potential as a powerful computational imaging tool for high-NA EUV mask imaging and metrology.

We demonstrate an experimental approach to determine magneto-optical effects that combines x-ray magnetic circular dichroism (XMCD) with x-ray interferometry, based on the concepts of Young's canonical double slit. By covering one of two slits with a magnetic thin film and employing XMCD, we show that it is possible to determine both the real and the imaginary parts of the complex refractive index by measuring the fringe shifts that occur due to a change in the sample magnetization. Our hybrid spectroscopic-interferometric methodology provides a means to probe changes in the magnetic refractive index in terms of the electron spin moment.

We demonstrate fabrication of an ultra-low blaze angle (LBA) grating designed for time-resolved spectroscopy in the x-ray region. A blaze angle of 0.12° was achieved by the groove depth reduction process providing a perfect saw-tooth shape of the grooves. The 200 lines/mm LBA grating exhibited high diffraction efficiency at a photon energy of 8 KeV. Use in a spectrometer would open up dispersive x-ray absorption spectroscopy across a wide energy range including the K edges of 3d transition metals, with much greater efficiency than traditional crystal spectrometers.

Autonomous methods to align beamlines can decrease the amount of time spent on diagnostics, and also uncover better global optima leading to better beam quality. The alignment of these beamlines is a high-dimensional expensive-to-sample optimization problem involving the simultaneous treatment of many optical elements with correlated and nonlinear dynamics. Bayesian optimization is a strategy of efficient global optimization that has proved successful in similar regimes in a wide variety of beamline alignment applications, though it has typically been implemented for particular beamlines and optimization tasks. In this paper, we present a basic formulation of Bayesian inference and Gaussian process models as they relate to multi-objective Bayesian optimization, as well as the practical challenges presented by beamline alignment. We show that the same general implementation of Bayesian optimization with special consideration for beamline alignment can quickly learn the dynamics of particular beamlines in an online fashion through hyperparameter fitting with no prior information. We present the implementation of a concise software framework for beamline alignment and test it on four different optimization problems for experiments on X-ray beamlines at the National Synchrotron Light Source II and the Advanced Light Source, and an electron beam at the Accelerator Test Facility, along with benchmarking on a simulated digital twin. We discuss new applications of the framework, and the potential for a unified approach to beamline alignment at synchrotron facilities.