Applications
-
Mixed noise and posterior estimation with conditional deepGEM
This work develops an expectation-maximization algorithm for jointly estimating posterior distributions and mixed (additive and multiplicative Gaussian) noise parameters in Bayesian inverse problems. The authors apply their method to real-world applications in nanometrology, specifically EUV scatterometry for characterizing nanostructures. JCMsuite was used to simulate the complex optical forward model (solving Maxwell's equations) for a line grating with an oxide layer, generating the data necessary to train and validate their proposed deep learning framework.
P. Hagemann, et al. Mixed noise and posterior estimation with conditional deepGEM. Mach. Learn.: Sci. Technol. 5, 035001 (2024).
2024 DOI Publication link
Optical Metrology and Sensing, software benchmarks, Advanced Finite Element Methods, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Bayesian Target-Vector Optimization for Efficient ParameterReconstruction
In this paper, a Bayesian target-vector optimization scheme, specialized for parameter reconstruction problems with hundreds of observations is presented. The performance is compared to established methods for an optical metrology problem and two least-square problems.
M. Plock, et al. Bayesian Target-Vector Optimization for Efficient ParameterReconstruction. Advanced Theory and Simulations, 5, 2200112 (2022).
2022 DOI Publication link
Optical Metrology and Sensing, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Bayesian optimization of metal grating back reflectors for multijunction solar cells
A triple-junction solar cell with a metal grating back reflector is accurately simulated using JCMsuite's finite-element solver. Based on the simulations the parameters of the metal grating are optimized with JCMsuite's Analysis and Optimization Toolkit to maximize the efficiency of the solar cell.
P. Tillmann, et al. Optimizing metal grating back reflectors for III-V-on-silicon multijunction solar cells. Opt. Express 29, 22517 (2021).
2021 DOI Publication link
Photovoltaics, diffractive optics, plasmonics, Light Scattering Computation, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Shape- and element-sensitive reconstruction of periodic nanostructures with grazing incidence X-ray fluorescence analysis and machine learning
The angular resolved fluorescence signal from a grazing incidence X-ray illumination of periodic nanostructures is used to reconstruct its geometry parameters. The parameter reconstruction using JCMsuite is based on a finite-element model of the scattering and fluorescence process as well as an efficient Bayesian minimization of the disagreement between the simulated and the measured fluorescence signal.
A. Andrle, et al. Shape- and element-sensitive reconstruction of periodic nanostructures with grazing incidence X-ray fluorescence analysis and machine learning. Nanomaterials, 11, 7 (2021).
2021 DOI Publication link
Optical Metrology and Sensing, Light Scattering Computation, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Investigating surface structures by EUV scattering
An exploration of soft X-ray and EUV-scatterometry from grazing to near normal incidence is presented. Measurements are performed on e-beam written silicon gratings. The reconstructed geometrical line shape models are statistically validated by applying a Markov-Chain Monte Carlo sampling technique. Experimental data and simulation results provide insight into the potential of EUV scatterometry.
V. Soltwisch, et al. Investigating surface structures by EUV scattering. Proc. SPIE 10143, 101430P (2017).
2017 DOI
Optical Metrology and Sensing, Optical and EUV Lithography, Advanced Finite Element Methods, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Quantifying parameter uncertainties in optical scatterometry using Bayesian inversion
A Newton-like method is presented to solve inverse problems and to quantify parameter uncertainties. FEM, including direct computation of partial derivatives, is used to solve the forward-problem.
M. Hammerschmidt, et al. Quantifying parameter uncertainties in optical scatterometry using Bayesian inversion. Proc. SPIE 10330, 1033004 (2017).
2017 DOI Publication link
Optical Metrology and Sensing, other fields, Advanced Finite Element Methods, Light Scattering Computation, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods
-
Efficient Bayesian inversion for shape reconstruction of lithography masks
In order to quantify the uncertainties of reconstructed geometry parameters, a fast-to-evaluate surrogate for the forward model (a polynomial chaos expansion) is introduced. The surrogate allows, e.g., for determining the probability distribution of the geometry parameters given measurement data, and for a global sensitivity analysis of the measurement process. All methods are implemented in JCMsuite's analysis and optimization toolbox.
N. Frachmin, et al. Efficient Bayesian inversion for shape reconstruction of lithography masks. Journal of Micro/Nanolithography, MEMS, and MOEMS, 19(2), 024001 (2020).
2010 DOI Publication link
Optical Metrology and Sensing, Optical and EUV Lithography, Light Scattering Computation, Optimization and Parameter Retrieval Methods, Uncertainty Quantification Methods