The inverse design of metasurfaces is made difficult by the highly nonlinear relationship between geometry and electromagnetic response, and conventional optimization is both computationally demanding and prone to local minima. We present a diffusion-based generative framework with two additions aimed at enforcing the electromagnetic specification. The first is a posterior sampling guidance constrained by a spherical Gaussian formulation, which, unlike unconstrained posterior sampling, does not lose accuracy as the array grows, and which allows models trained only on 23 & times; 23-pillar arrays to design metasurfaces of up to 98 & times; 98 pillars in a few minutes. The second is a consistency training loss that evaluates the electromagnetic Figure of Merit during training, which reduces the sensitivity of the model to its diffusion hyper parameters and thus the tuning effort required. All reported accuracies are verified by full-wave FDTD simulation of the generated designs. Across the evaluated range, the proposed method reaches a mean R2 of 0.962 against 0.913 for a phase-retrieval and look-up-table baseline and 0.699 for surrogate-driven gradient descent.
Objective.Interventional radiology procedures can expose patients and physicians to high doses of radiation. Virtual fluoroscopy based on fan-beams helps to reduce this level of irradiation. Here, we demonstrate the benefits of using fan-beams obtained by curved-slit collimation rather than the straight-slit collimation previously proposed.Approach.We study spiral-slits that allow for a localization procedure identical to that with straight-slits. We show that the localization error depends on the shape of the slit and we bound this error. This bound can be used to optimize the slit geometry according to different performance criteria.Main results.Several curved-slit geometries are studied, and their relative performance is evaluated through numerical experiments and preliminary tests on x-ray equipment. These experiments show a systematic improvement in performance with the proposed curved-slit geometries compared with the straight-slit. In particular, a curved-slit geometry with a linear spiral part and an exponential part reduces localization error and solves the problem of the blind central zone with the straight-slit geometry.Significance.We have developed a framework for analyzing the amplification of localization errors in a rotating collimator with symmetrical spiral-slits, as well as criteria for measuring this amplification and bounds indicating their optimality. This framework can be reused to identify or compare slits under conditions of use different from those in our study.
A low-power and area-efficient regulation for negative charge-pump in 18 nm Fully Depleted Silicon on Insulator (FD-SOI) is presented. In FD-SOI, body biasing requires the generation and control of negative well voltages, making an efficient negative charge-pump regulator essential for Adaptive Body-Bias (ABB) systems. In this context, a pseudo-differential 6-bit Capacitive Digital to Analog Converter (CDAC) with 0.5 fF full custom unit capacitors and a double-tail comparator implements a bang-bang control of a negative charge pump, consuming only dynamic power. The capacitor array is analyzed and an optimization methodology is presented to determine design optimum point for this architecture. Silicon measurements over 64 dies in 18 nm advanced node confirm the design methodology presented in this work, with accurate correlation to simulation results. Statistical analysis of the silicon measurements confirms the predicted error variability and the expected current consumption. The regulation current scales nearly linearly with operating frequency, reaching 480 nW of power consumption at 70 kHz frequency measured on silicon at 25°C ambient temperature and 1.8 V nominal supply, while the regulation core occupies only $1632 \mu \mathbf{m}^{2}$, making the solution suitable for lowpower Body-Bias Generator (BBGEN).
Metasurface inverse design is challenged by the intricate relationship between structural parameters and electromagnetic responses, as well as the high dimensionality of the optimization space. Local models, while commonly employed, quickly become infeasible for complex and locally coupled structures. Conventional iterative optimization techniques, on the other hand, are computationally intensive, time-consuming, and susceptible to convergence in local minima. This study explores a versatile generative methodology based on enhanced posterior sampling within the Schrödinger Bridge framework. By decomposing posterior sampling into amplitude and directional contributions, we effectively integrated different kind of posterior sampling. This approach is further supported by refined training strategies to enhance performance and reduce the complexity of hyperparameter optimization. The proposed framework demonstrates exceptional accuracy and robustness, representing a significant advancement in metasurface design. Notably, it enables high-precision inverse design for large-scale configurations of up to 350 × 350 pillar arrays, despite being trained on significantly smaller arrays of 23 × 23 pillars.
Crystalline coatings have emerged as a promising alternative to amorphous mirror coatings for planned upgrades of second-generation detectors such as Advanced Virgo and for future 3G observatories. In this work, we investigate the structural and mechanical properties of epitaxial Cr2O3 thin films grown on c-plane sapphire by molecular beam epitaxy, with the objective of evaluating corundum-structured oxides as a new class of crystalline coating materials for gravitational wave interferometers. We report, for the first time, cryogenic mechanical loss measurements of Cr2O3 coatings and relate the measured dissipation to their crystalline quality. The structural and mechanical properties of the coatings were evaluated by XRD, RHEED, AFM, and GeNS measurements. Our findings show that the highest-quality chromia layers exhibit mechanical losses at 6 K as low as (5 +/- 1) & times; 10-6 rad, while poorer crystalline quality is associated with significantly higher losses. These results demonstrate the potential of corundum-structured oxides as mirror coatings for next-generation interferometers. The wide availability of corundum-structured oxides (X2O3 with X = Al, Ga, Fe, V, Cr, Ti, & mldr;) and their epitaxial compatibility with sapphire, including the possibility of forming solid solutions, further highlights the potential of this largely unexplored materials class.
