TY - CONF A1 - Radtke, Martin T1 - Machine learning for direct quantification of XRF measurements N2 - In X-ray fluorescence (XRF), a sample is excited with X-rays, and the resulting characteristic radiation is detected to detect elements quantitatively and qualitatively. Quantification is traditionally done in several steps: 1. Normalization of the data 2. Determination of the existing elements 3. Fit of the measured spectrum 4. Calculation of concentrations with fundamental parameters / MC simulations / standard based The problem with standard based procedures is the availability of corresponding standards. The problem with the calculations is that the measured intensities for XRF measurements are matrix-dependent. Calculations must, therefore, be performed iteratively (= time consuming) in order to determine the chemical composition. First experiments with gold samples have shown the feasibility of machine learning based quantification in principle. A large number of compositions were simulated (> 10000) and analyzed with a deep learning network. For first experiments, an ANN (Artificial Neural Network) with 3 hidden layers and 33x33x33 neurons was used. This network learned the mapping of spectra to concentrations using supervised learning by multidimensional regression. The input layer was formed by the normalized spectrum, and the output layer directly yielded the searched values. The applicability for real samples was shown by measurements on certified reference materials. T2 - Denver X-ray Conference CY - Lombard, IL, USA DA - 05.08.2019 KW - Machine learning KW - Artificial intelligence KW - Neural network KW - XRF KW - Synchrotron PY - 2019 AN - OPUS4-48903 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Fittschen, U. A1 - Hampel, S. A1 - Till, H. A1 - Gross, A. A1 - Kulow, Anicó A1 - de Oliveira Guilherme Buzanich, Ana A1 - Reinholz, Uwe A1 - Ingerle, D. A1 - Streli, C. A1 - Radtke, Martin T1 - Improving surface sensitive XRF using ink jet printing and information from the angle dependent signal N2 - Total Reflection X-ray Fluorescence (TXRF) is a small footprint, ressource efficient micro-analytical tool for trace elemental determination. However, depending on the matrix TXRF is also challenging in several ways: the preparation of a representative aliquot maybe difficult for slurries, shading effects and matrix effects may occur and the applicability of an IS maybe hampered by interferences (fitting may be impaired as well) or inhomogeneities. It is therefore crucial to understand and if applicable mitigate the influence of the before mentioned phenomena. We have used the small volume approach using pL droplets to study shading in TXRF previously.Using this approach thin specimens in a favorable geometry are prepared with a well defined morphology; this way minimizing matrix effects and shading. To be used as standard it is also necessary to determine the delivered elemental amounts. Here we will present on the performance of a commercial ink-jet printer cartridge to deliver defined volumes and elemental amounts. The microscopic specimens obtained have been successfully applied to determine relative sensitivities in TXRF and prepare references to study coded apertures in grazing incidence full field micro-XRF.The homogeneous lateral distribution of analyte and IS may be probed by micro-XRF, to obtain information on alikeness of analyte and IS. To study the alikeness of the in depth distribution (film-like or particle-like) angle scans can deliver valuable data. Here we present first results on angle scans using a prototype GIXRF set up and a commercial TXRF instrument. T2 - Denver X-Ray Conference CY - Online meeting DA - 03.08.2020 KW - TXRF KW - Coded Aperture PY - 2020 AN - OPUS4-51913 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Radtke, Martin T1 - Machine learning for direct quantification of XRF measurements N2 - 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. T2 - Seminar series: Artificial Intelligence applied to X-ray / synchrotron techniques CY - Online meeting DA - 24.06.2021 KW - Artificial intelligence KW - Machine learning KW - Synchrotron KW - XRF PY - 2021 AN - OPUS4-54140 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Vogel, Christian A1 - Wittwer, Philipp A1 - Roesch, Philipp A1 - Simon, Franz-Georg T1 - Per- and polyfluoroalkyl substances (PFAS) in sewage sludge and wastewater-based fertilizers and future PFAS remediation N2 - Per- and polyfluoroalkyl substances (PFAS) are a group of anionic, cationic and zwitterionic synthetic products, in which the hydrogen atoms on the carbon skeleton of at least one carbon atom have been completely replaced by fluorine atoms and which include more than 4730 compounds, depending on the definition. As a result of continuous and prolific use, mainly in aviation firefighting foams, thousands of industrial and military installations have been found to contain contaminated soil, groundwater and surface water. As a result of the perpetual use of PFAS containing products, effluents and sewage sludge from wastewater treatment plants (WWTPs) have been observed to be an important pathway for PFAS into the environment. In Germany, phosphorus and other nutrients from sewage sludge and wastewater should be recycled in WWTPs of cities with a large population. However, it is not clear if PFAS contamination from wastewater and sewage sludge end up in novel wastewater-based fertilizers. Normally, PFAS are analyzed using PFAS protocols typically with liquid chromatography tandem mass spectrometry (LC-MS/MS) quantification. To get a better overview of the amount of “total PFAS,” we applied sum parameter methods based on combustion ion chromatography (CIC) to screen the PFAS contaminations in various sewage sludge and wastewater-based fertilizers. Furthermore, current decontamination strategies of PFAS-burdened soils mainly consist of adsorption methods using adsorbents for fixation of PFAS in the ground. A second option is the utilization of a “pump and treat” process, cycling polluted soils through a washing plant leading to the concentration of the pollutants in the fine fraction. Only a subsequent, high-energy consuming pyrolysis process guarantees the total destruction of all fluorinated organic contaminants. Both approaches are cost-intensive and not intended for the direct decomposition of all PFAS contaminants. Hence, there is a great demand for innovative developments and chemical treatment technologies, dealing with new strategies of tackling the PFAS problem. T2 - Seminar of Ben-Gurion University CY - Midreshet Ben-Gurion, Israel DA - 02.11.2022 KW - PFAS KW - Sewage sludge KW - XANES spectroscopy PY - 2022 AN - OPUS4-56166 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Vogel, Christian A1 - Roesch, Philipp A1 - Wittwer, Philipp A1 - Piechotta, Christian A1 - Lisec, Jan A1 - Sommerfeld, Thomas A1 - Kluge, Stephanie A1 - Herzel, Hannes A1 - Simon, Franz-Georg T1 - Per- and Polyfluoroalkyl Substances (PFAS) in Sewage Sludge and Wastewater-based Fertilizers N2 - Per- and polyfluoroalkyl substances (PFAS) are chemicals which were developed to improve humanity’s quality of life. Due to their high chemical stability and resistance to degradation by heat or acids, PFAS were used in a variety of consumer products. The continuous use of PFAS in household products and the discharge of PFAS from industrial plants into the sewer system resulted in the contamination of effluents and sewage sludge from wastewater treatment plants (WWTPs) (Roesch et al. 2022). Since sewage sludge is often used as fertilizer, its application on agricultural soils has been observed as a significant entry path for PFAS into the environment, specifically in our food chain. In Germany the sewage sludge/biosolid application on agricultural land was banned with the amendment of the German Sewage Sludge Ordinance and by 2029 sewage sludge application will be totally prohibited. However, phosphorus (P) from sewage sludge should still be recycled in WWTPs of cities with a population larger than 50,000 residents. To produce high-quality P-fertilizers for a circular economy, PFAS and other pollutants (e.g. pesticides and pharmaceuticals) must be separated from sewage sludge. Due to the strong diversity of industrial PFAS usage it is not clear if a safe application of novel recycled P-fertilizers from WWTPs can be guaranteed. Therefore, we analyzed various sewage sludges and wastewater-based fertilizers. Sewage sludge (SL) samples from various WWTPs in Germany and Switzerland, six sewage sludge ashes (SSA) from Germany, six thermally treated SL and SSA samples with different additives (temperatures: 700-1050 °C), two pyrolyzed SL samples (temperature: 400 °C) and two struvite samples from Germany and Canada were analyzed. The goal was to quantify PFAS in sewage sludges and wastewater-based P-fertilizers with the sum parameter extractable organic fluorine (EOF) by combustion ion chromatography (CIC). The results were compared with data from classical LC-MS/MS target analysis as well as selected samples by HR-MS suspect screening. The EOF values of the SLs mainly range between 154 and 538 µg/kg except for one SL which showed an elevated EOF value of 7209 µg/kg due to high organofluorine contamination. For the SSA samples the EOF values were lower and values between LOQ (approx. 60 µg/kg) and 121 µg/kg could be detected. For the pyrolyzed SLs no EOF values above the LOQ were detected. Moreover, the two wastewater-based struvite fertilizers contain 96 and 112 µg/kg EOF, respectively. In contrast to the EOF values, the sum of PFAS target values were relatively low for all SLs. Additional applied PFAS HR-MS suspect screening aimed to tentatively identify PFAS that could contribute to the hitherto unknown part of the EOF value. The majority of the detected fluorinated compounds are legacy PFAS such as short- and long-chain perfluorocarboxylic acids (PFCA), perfluorosulfonic acids (PFSA), polyfluoroalkyl phosphate esters (PAPs) and perfluorophosphonic acids (PFPA). Moreover, fluorinated pesticides, pharmaceutical as well as aromatic compounds were also identified, which are all included in the EOF parameter. Our research revealed that the current PFAS limit of 100 µg/kg for the sum of PFOS + PFOA in the German Fertilizer Ordinance is no longer up to date. Since the number of known PFAS already exceeds 10,000, the ordinance limit should be updated accordingly. Recent regulations and restrictions on using long-chain PFAS (≥C8) have resulted in a significant shift in the industry towards (ultra-)short-chain alternatives, and other, partly unknown, emerging PFAS. Ultimately, also fluorinated pesticides and pharmaceuticals, which end up as ultrashort PFAS in the WWTPs, have to be considered as possible pollutants in fertilizers from wastewater, too. T2 - Dioxin Konferenz CY - Maastricht, Netherlands DA - 10.09.2023 KW - Combustion Ion Chromatography KW - Per- and Polyfluoroalkyl substances (PFAS) KW - Sewage sludge KW - Fertilizer PY - 2023 AN - OPUS4-58345 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Vogel, Christian A1 - Roesch, Philipp A1 - Wittwer, Philipp A1 - Lisec, Jan A1 - Borca, C. A1 - Huthwelker, T. A1 - Simon, Franz-Georg T1 - Combining DGT and combustion ion chromatography (CIC) as screening tool for per- and polyfluoroalkyl substances (PFAS) contamination in wastewater-based fertilizers N2 - Per- and polyfluoroalkyl substances (PFAS) are a large group of organofluorine surfactants used in the formulations of thousands of consumer goods. The continuous use of PFAS in household products and the discharge of PFAS from industrial plants into the sewer system have been resulted in contaminated effluents and sewage sludge from wastewater treatment plants (WWTPs) which became an important pathway for PFAS into the environment. Because sewage sludge is often used as fertilizer its application on agricultural soils has been observed as significant input path for PFAS into our food chain. To produce high-quality phosphorus fertilizers for a circular economy from sewage sludge, PFAS and other pollutants (e.g. pesticides and pharmaceuticals) must be separated from sewage sludge. Normally, PFAS are analyzed using PFAS protocols typically with time-consuming extraction steps and LC-MS/MS target quantification. However, for screening of PFAS contaminations in wastewater-based fertilizers also the DGT technique can be used for the PFAS extraction. Afterwards, combustion ion chromatography (CIC) can be applied to analyze the “total” amount of PFAS on the DGT binding layer. The DGT method was less sensitive and only comparable to the extractable organic fluorine (EOF) method values of the fertilizers in samples with >150 µg/kg, because of different diffusion properties for various PFAS, but also kinetic exchange limitations. However, the DGT approach has the advantage that almost no sample preparation is necessary. Moreover, the PFAS adsorption on the DGT binding layer was investigated via surface sensitive spectroscopical methods, such as Fourier-transform infrared (FT-IR) and fluorine K-edge X-ray absorption near-edge structure (XANES) spectroscopy. T2 - DGT Konferenz CY - Paris, France DA - 11.10.2023 KW - Combustion Ion Chromatography KW - Per- and Polyfluoroalkyl substances (PFAS) KW - Sewage sludge PY - 2023 AN - OPUS4-58575 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Oster, Simon A1 - Becker, Tina A1 - Breese, Philipp Peter A1 - Scheuschner, Nils A1 - Maierhofer, Christiane A1 - Ulbricht, Alexander A1 - Frisch, Tobias A1 - Mohr, Gunther A1 - Altenburg, Simon T1 - Defect prediction in laser powder bed fusion based on thermographic features utilizing convolutional neural networks N2 - The appearance of irregularities such as keyhole porosity is a major challenge for the production of metal parts by laser powder bed fusion (PBF-LB/M). The utilization of thermographic in-situ monitoring is a promising approach to extract the thermal history which is closely related to the formation of irregularities. In this study, we investigate the utilization of convolutional neural networks to predict keyhole porosity based on thermographic features. Here, the porosity information calculated from an x-ray micro computed tomography scan is used as reference. Feature engineering is performed to enable the model to learn the complex physical characteristics of the porosity formation. The model is examined with regard to the choice of hyperparameters, the significance of thermal features and characteristics of the data acquisition. Based on the results, future demands on irregularity prediction in PBF-LB/M are derived. T2 - GIMC SIMAI YOUNG 2022 CY - Pavia, Italy DA - 29.09.2022 KW - Laser Powder Bed Fusion KW - Thermography KW - In-situ Monitoring KW - Convolutional Neural Networks PY - 2022 AN - OPUS4-56331 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Oster, Simon A1 - Becker, Tina A1 - Breese, Philipp Peter A1 - Scheuschner, Nils A1 - Maierhofer, Christiane A1 - Ulbricht, Alexander A1 - Fritsch, Tobias A1 - Mohr, Gunther A1 - Altenburg, Simon T1 - Defect prediction on the Base of Thermographic features in Laser Powder Bed Fusion Utilizing Machine Learning Algorithms N2 - Avoiding the formation of defects such as keyhole pores is a major challenge for the production of metal parts by Laser Powder Bed Fusion (LPBF). The use of in-situ monitoring by thermographic cameras is a promising approach to detect defects, however the data is hard to analyze by conventional algorithms. Therefore, we investigate the use of Machine Learning (ML) in this study, as it is a suitable tool to model complex processes with many influencing factors. A ML model for defect prediction is created based on features extracted from process thermograms. The porosity information calculated from an x-ray Micro Computed Tomography (µCT) scan is used as reference. Physical characteristics of the keyhole pore formation are incorporated into the model to increase the prediction accuracy. Based on the prediction result, the quality of the input data is inferred and future demands on in-situ monitoring of LPBF processes are derived. T2 - Additive Manufacturing Benchmarks 2022 CY - Bethesda, MA, USA DA - 14.08.2022 KW - Laser Powder Bed Fusion KW - Thermography KW - In-situ Monitoring KW - Machine Learning KW - Defect prediction PY - 2022 AN - OPUS4-55591 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Oster, Simon A1 - Becker, Tina A1 - Breese, Philipp Peter A1 - Scheuschner, Nils A1 - Altenburg, Simon T1 - Machine Learning based defect detection in Laser Powder Bed Fusion utilizing thermographic feature data N2 - The formation of irregularities such as keyhole porosity pose a major challenge to the manufacturing of metal parts by laser powder bed fusion (PBF-LB/M). In-situ thermography as a process monitoring technique shows promising potential in this manner since it is able to extract the thermal history of the part which is closely related to the formation of irregularities. In this study, we investigate the utilization of machine learning algorithms to detect keyhole porosity on the base of thermographic features. Here, as a referential technique, x-ray micro computed tomography is utilized to determine the part's porosity. An enhanced preprocessing workflow inspired by the physics of the keyhole irregularity formation is presented in combination with a customized model architecture. Furthermore, experiments were performed to clarify the role of important parameters of the preprocessing workflow for the task of defect detection . Based on the results, future demands on irregularity prediction in PBF-LB/M are derived. T2 - International Conference on NDE 4.0 CY - Berlin, Germany DA - 24.10.2022 KW - Laser Powder Bed Fusion KW - Thermography KW - In-situ Monitoring KW - Convolutional Neural Networks PY - 2022 AN - OPUS4-56332 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Geburtig, Anja T1 - Combined impact of UV radiation and nitric acid (HNO3) on HDPE jerrycans – comparison of outdoor and lab test N2 - A damaging action of HNO3-55% only occurs in combination with its decomposition into nitrous gases, which can be caused by UV radiation. In a laboratory test, transparent HDPE jerrycans have been exposed to both UV radiation and 55 wt-% nitric acid solution at (41 ± 2)°C, for up to 20 days. For comparison, UV radiant exposure (21 days) and nitric acid exposure (up to 6 weeks) were performed separately, at nearly equal temperatures. The respective damages are compared with FTIR spectroscopy in ATR and HT-gel permeation chromatography (GPC) on a molecular level and with hydraulic internal pressure testing as a component test. For the used jerrycans, relevant oxidation can only be found after the combined exposure. The gradual increase in oxidative damage shows the good reproducibility of the lab exposure. The decomposition of nitric acid into nitrous gases by UV radiation – as well as the jerrycan oxidation – is also observed at lower HNO3 concentration (28 wt- %). Similar results are obtained after outdoor tests. Again, the damage occurs only after combined exposure, in contrast to the exposures to UV only and to HNO3 only, which were conducted in parallel. Outdoor exposures are most readily accepted as they represent possible end-use conditions. However, the reproducibility of these exposures is poor due to the large temporal variations in weather. There are also several safety risks, which is why the number of replicates remains limited. Since the outdoor and lab exposure tests show the same qualitative results, it is appropriate to conduct systematic studies in the laboratory. After 6 days of lab exposure, the oxidation damage is rated as critical, which corresponds to about 1/10 year in Central Europe, according to the UV radiant exposure. It should be noted that this amount can also occur in two sunny weeks. T2 - 24th iapri World Packaging Conference CY - Valencia, Spain DA - 17.06.2024 KW - UV exposure KW - Nitric acid KW - Polyethylene KW - Oxidation PY - 2024 AN - OPUS4-60416 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -