Ingenieurwissenschaften und zugeordnete Tätigkeiten
Filtern
Erscheinungsjahr
- 2020 (3) (entfernen)
Dokumenttyp
- Posterpräsentation (3) (entfernen)
Sprache
- Englisch (3) (entfernen)
Referierte Publikation
- nein (3)
Schlagworte
- Automatisation (1)
- Convolutional neural networks (1)
- Electron microscopy (1)
- HR-CS-AAS (1)
- Image segmentation (1)
- Imaging techniques (1)
- Iron oxide nanoparticles (1)
- Isotope analysis (1)
- Lithium batteries (1)
- Machine learning (1)
- Nanoparticles (1)
- Reference nanomaterials (1)
- Reliable characterization (1)
- Size and size distribution (1)
- Spectrometry (1)
Organisationseinheit der BAM
- 1 Analytische Chemie; Referenzmaterialien (3) (entfernen)
Background, Motivation and Objective
Nanomaterials are at the core of some of the 21st century’s most promising technologies. In order to utilize and rationally design materials at the nanoscale the reliable characterization of their physico-chemical properties is highly important. Furthermore, the European Commission has taken measures via the REACH Regulations to control the classification of nanomaterials. REACH Annexes which entered into force in January 2020 require manufacturers to register nanomaterials that are traded in larger quantities (at least 1 ton). Every powder or dispersion where 50% (number distribution) of the constituent particles have sizes ≤ 100 nm in at least one dimension are defined as a nanomaterial. This creates a need for both industrial manufacturers and research and analytical service facilities to reliably characterize potential nanomaterials. Currently, BAM is working on developing reference nanoparticles, which shall expand the scarce list of worldwide available nano reference materials certified for particle size distribution, but also targeting other key parameters such as shape, structure (including porosity) and functional properties. Thus, candidate materials are considered to complement the already available spherical and monodisperse silica, Au and polystyrene reference nanoparticles, e.g. iron oxide and titanium oxide, with an average atomic number between those of silica and gold. Particularly for the imaging by electron microscopies, new nanoparticles of well-defined size in the range of 10 nm are decisive for the accurate particle segmentation by setting precise thresholds.
Statement of Contribution/Methods
Synthesis: Highly monodisperse iron oxide nanoparticles can be synthesized in large quantities by thermal decomposition of iron oleate or iron acetylacetonate precursors in high boiling solvents such as octadecene or dioctyl ether in the presence of oleic acid and oleylamine as capping agents.
Scanning Electron Microscope: An SEM of type Supra 40 from Zeiss has been used including the dedicated measurement mode transmission in SEM (STEM-in-SEM) with a superior material contrast for the nanoparticle analysis. The software package ImageJ has been used for the analysis of the STEM-in-SEM images and to determine the particle size distribution.
Dynamic Light scattering (DLS): Particles in suspension were measured in comparison by means of Zetasizer Nano (Malvern Panalytical; cumulants analysis) and NanoFlex (Microtrac; frequency power spectrum).
Results/Discussion
In this study iron oxide nanoparticles synthesized at BAM and pre-characterized by DLS, SEM (including the transmission mode STEM-in-SEM) are presented. The particles are spherical and highly monodisperse with sizes slightly larger than 10 nm.
An alternative method for lithium isotope amount ratio analysis is proposed by combining atomic absorption spectrometry with spectra data analysis by machine leaning. It is based on the well-known isotope shift of around 15 pm for the electronic transition at wavelength 670.7845 nm which can be measured by a high-resolution continuum source atomic absorption spectrometer (HR-CS-AAS). For isotope amount ratio analysis, a scalable three boosting machine learning algorithm (XGBoost) was employed and calibrated with a set of samples with a 6Li isotope amount fraction ranging from 99% to 6%. The absolute Li isotope amount fractions of these calibration samples were previously measured by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) and used as ab-initio data for the machine learning algorithm. Validation of the machine leaning model was performed with two standard reference materials (LSVEC and IRMM-016). The procedure was employed for the isotope amount ratio determination of a set of stock chemicals (Li2CO3, LiNO3, LiCl, LiOH, and LiF) as well as a BAM candidate LiMNC cathode reference material. Achieved uncertainties are one order of magnitude higher than those obtained by MC-ICP-MS. This precision and accuracy is nonetheless sufficient to resolve natural occurring variations in Lithium isotope ratios. Also, the LiMNC material was analyzed by HR-CS-AAS with and without matrix purification. The results are comparable within statistical error.
In this contribution different ways are explored with the aim to generate suitable training data for ‘non-ideal’ samples using various approaches, e.g., computer-generated images or unsupervised learning algorithms such as generative adversarial networks (GANs). We used these data to train simple CNNs to produce segmentation masks of SEM images and tested the trained networks on real SEM images of complex nanoparticle samples. The novel use of CNN for the automated analysis of the size of nanoparticles of complex shape and with a high degree of agglomeration has proved to be a promising tool for the evaluation of particle size distribution on a large number of constituent particles. Further development and validation of the preliminary model, respectively larger training and validation data sets are necessary.