Filtern
Dokumenttyp
- Zeitschriftenartikel (3)
- Vortrag (1)
- Posterpräsentation (1)
Sprache
- Englisch (5)
Schlagworte
- Nanoparticles (5) (entfernen)
Organisationseinheit der BAM
- 1 Analytische Chemie; Referenzmaterialien (5) (entfernen)
Eingeladener Vortrag
- nein (1)
The roughness as a property of core–shell (CS) microparticles plays a key role in their functionality. Quantitative evaluation of the roughness of CS microparticles is, however, a challenging task with approaches using electron microscopy images being scarce and showing pronounced differences in terms of methodology and results. This work presents a generalized method for the reliable roughness determination of nonplanar specimens such as CS particles from electron microscopic images, the method being robust and reproducible with a high accuracy. It involves a self-written software package (Python) that analyzes the recorded images, extracts corresponding data, and calculates the roughness based on the deviation of the identified contour. Images of single particles are taken by a dual mode scanning electron microscopy (SEM) setup which permits imaging of the same field-of-view of the sample with high resolution and surface sensitive in SE InLens mode as well as in transmission mode (TSEM). Herein, a new type of polystyrene core–iron oxide shell–silica shell particles is developed to serve as a set of lower micrometer-sized study objects with different surface roughness; the analysis of their images by the semiautomatic workflow is demonstrating that the particles’ profile roughness can be quantitatively obtained.
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.
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.
The mechanism of action of zirconium permanent modifiers on graphite surfaces was investigated in order to understand its influence on the analytical signal in atomic and molecular absorption spectrometry (AAS/MAS). For this, the molecule formation of CaF was studied, which is used for the indirect analytical determination of fluorine in high-resolution continuum source graphite furnace molecular absorption spectrometry (HR-CS-GFMAS). The kinetics of this reaction was established by monitoring its molecular spectrum at different atomisation temperatures. An Arrhenius plot showed a pseudo-first order reaction with respect to fluorine (n = 1). An intermediate state was isolated, and its structure was elucidated by spectroscopic methods: scanning electron microscopy with energy dispersive X-ray spectroscopy (SEMEDX), X-ray photoelectron spectroscopy (XPS), X-ray absorption spectroscopy (XANES and EXAFS), and Raman microspectroscopy. We propose here a mechanism, where ZrO2 acts as a heterogeneous catalyst: after a pyrolytic step, an intermediate state of ZrO(OCaF) is activated, and at higher temperatures, CaF(g) is released from the zirconium-coated graphite surface. No evidence of the formation of zirconium carbide was found. Consequently, as the CaF formation is catalysed by a heterogeneous catalyst, surface modifications with ZrO2 nanoparticles and ZrO xerogels were investigated in order to increase the surface area. Their influence was evaluated in the molecule formation of CaF, CaCl, CaBr, and CaI. Graphite furnace modification with zirconium oxide nanoparticles proves to be the best choice for fluorine analysis with a signal enhancement of more than eleven times with respect a non-coated graphite furnace. However, the influence of zirconium modifications in the analytical signals of Cl, and I is lower than the F signals or even negative in case of the Br. Understanding zirconium modifiers as heterogeneous catalysts offers a new perspective to AAS and MAS, and reveals the potential of surface analytical methods for development of improved permanent modifiers and graphite furnace coatings.