Präsentation
Filtern
Erscheinungsjahr
Dokumenttyp
- Vortrag (58)
- Posterpräsentation (18)
Referierte Publikation
- nein (76)
Schlagworte
- Synchrotron (76) (entfernen)
Organisationseinheit der BAM
- 6 Materialchemie (31)
- 6.3 Strukturanalytik (30)
- 8 Zerstörungsfreie Prüfung (6)
- 8.5 Röntgenbildgebung (6)
- 9 Komponentensicherheit (3)
- 9.4 Integrität von Schweißverbindungen (3)
- 4 Material und Umwelt (2)
- 4.3 Schadstofftransfer und Umwelttechnologien (1)
- 4.5 Kunst- und Kulturgutanalyse (1)
- 5 Werkstofftechnik (1)
Eingeladener Vortrag
- nein (58)
This contribution provides an overview of the BAMline synchrotron radiation beamline, which specializes in hard X-ray spectroscopy techniques for materials research. The BAMline offers X-ray absorption spectroscopy (XAS), x-ray fluorescence spectroscopy (XRF), and tomography to study materials' electronic structure, chemical composition, and structure. Key capabilities include standard and dispersive XAS for electronic structure, micro-XRF for elemental mapping, coded aperture imaging, and depth-resolved grazing exit XAS. The BAMline enables in situ characterization during materials synthesis and functions for energy, catalysis, corrosion, biology, and cultural heritage applications.
Ongoing developments like the implementation of machine learning techniques for experiment optimization and data analysis will be discussed. For instance, Bayesian optimization is being used to improve beamline alignment and scanning. An outlook to the future, where the BAMline will continue pioneering dynamic and multi-scale characterization, aided by advanced data science methods, to provide unique insights into materials research, will be given.
Our aim is to develop a simple and inexpensive method for full field X-ray fluorescence imaging.We combine an energydispersive array detector with a coded aperture to obtain high resolut ion images. To obtain the information from the recorded image a reconstruction step is necessary. The reconstruction methods we have developed, were tested on simulated data and then applied to experimental data. The first tests were carried out at the BAMline @BESSY II. This method enables the simultaneous detection of multiple elements,which is important e.g. in the field of catalysis.
News from the BAMline
(2022)
Getting more efficient – The use of Bayesian optimization and Gaussian processes at the BAMline
(2022)
For more than 20 years, BAM is operating the BAMline at the synchrotron BESSY II in Berlin Adlershof. During this time, the complexity of the setup and the amount of data generated have multiplied. To increase the effectiveness and in preparation for BESSY III, algorithms from the field of machine learning are increasingly used.
In this paper, several examples in the areas of beamline alignment and measurement time optimization based on Bayesian optimization (BO) with Gaussian processes (GP) are presented. BO is a method for finding the global optimum of a function using a probabilistic model represented by a GP. The advantage of this method is that it can handle high-dimensional problems, does not depend on the initial estimate, and also provides uncertainty estimates.
After a short introduction to BO and GP, the first example is the automatic alignment of our double multilayer monochromator (DMM). To achieve optimal performance, up to three linear and two angular motor positions have to be optimized. To achieve this with a grid scan, at least 100^5 measurement points would be required. Assuming that all positions can be aligned independently, 100*5 points are still necessary. We show that with BO and GP less than 100 points are sufficient to achieve equal or better results.
The second example is the optimization of measurement time in XRF scanning. Here we will show the advantage of the BO GP approach over point-by-point scanning. As can be seen in Fig. 1, the number of points required and thus the measurement time can be reduced by a factor of 50, while the loss in image quality is acceptable. The advantages and limitations of this approach will be discussed.
In this talk an overview about artificial intelligence/machine learning applications @BAMline is given. In the first part, the use of neural networks for the quantification of XRF measurements and the decoding of coded-aperture measurements are shown. Then it is shown how Gaussian processes and Bayesian statistics can be used to achieve an optimal alignment of the set-up and in general for optimization of measurements.
Gold is one of the seven metals already known in antiquity and was used from time immemorial as a medium of exchange and for the production of jewelry because of its luster and rarity. In addition, it is easy to work and largely resistant to chemical influences. Investigations of gold using synchrotron radiation excited X-ray fluorescence analysis are non-destructive and provide information about the chemical elements present in the sample under investigation. The investigations presented here at BAMline focus on questions such as the origin, manufacturing process, and association of gold findings. The different questions are explained by a number of examples ranging from the Viking treasure from Hiddensee to the Nebra Sky Disk and finds from Egypt. The find from Bernstorf is discussed in detail. A Bayesian treatment of the authenticity is shown.
Confocal Micro X-ray fluorescence analysis (μXRF) is a powerful method for the destruction-free investigation of the elemental composition of various samples in three dimensions. Quantitative results rely heavily on the transmission properties of polycapillary optics, used in confocal setups.
Different models for the functional correlation between energy and transmission of a polycapillary half-lens are compared with experimental data, established with a monochromatic confocal laboratory setup [1], as well as Monte-Carlo-Simulations [2]. The effect of the different approaches on quantitative results are compared using well known SRMs, ranging from thick samples to thin layers and structured materials.
Along with this the basic principles for quantitative evaluations of XRF spectra taken with confocal setups are presented, and compared with experimental results.
In this talk I’ll describe the use of artificial neural networks (ANN) for quantifying X-ray fluorescence (XRF) measurements. The main idea of this talk is to give an overview of the process needed to generate a model that can then be applied to a specific problem.
In XRF, a sample is excited with X-rays and the resulting characteristic radiation is detected to determine elements quantitatively and qualitatively. This is traditionally done in several time-consuming steps. I’ll show the possibilities and problems of using a neural network to realise a "one-click" quantification. This includes generating training data using Monte Carlo simulation and augmenting the existing data set with an ANN to generate more data. The search for the optimal hyperparameters, manually and automatically, is also described. For the case presented, we were able to train a network with a mean absolute error of 0.1% by weight for the synthetic data and 0.7% by weight for a set of experimental data obtained with certified reference materials.