Filtern
Dokumenttyp
- Posterpräsentation (21) (entfernen)
Sprache
- Englisch (21)
Referierte Publikation
- nein (21)
Schlagworte
- Nanoparticles (21) (entfernen)
Organisationseinheit der BAM
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.
Although there are many experimental techniques for measuring particle sizes and size distributions, electron microscopy (EM) is still considered as the gold standard in this field, especially when it comes to particle sizes in the nanorange (1 nm – 100 nm). Furthermore, high-resolution X-ray spectroscopy (EDS) can be applied to individual nanoparticles. To be able to extract accurate information from the EM micrographs and EDS elemental maps that are representative for the material under investigation, one needs to assure the representativity of the particles as sampled on the substrate and their homogeneous spatial distribution, to avoid operator bias when selecting the imaged area. Furthermore, agglomeration should be avoided as far as possible. Several sample preparation techniques exist since a long time, the most common way being suspending the particles in a liquid and depositing them on the grid. However, this procedure includes the drying of larger solvent amounts on the substrate itself, which can affect the spatial distribution of the deposited particles. One possibility to overcome this problem is the use of an electrospray system, where the suspension of particles is sprayed onto the substrate in charged droplets that are so small that they either dry off on the substrate immediately without affecting the position of particles, or even already during their flight time to the substrate. No dedicated commercial instruments are available for the preparation of TEM grids yet, only electrostatic deposition of aerosols on TEM grids has been reported so far. To test the prototype and assess its performance, several materials have been sprayed onto TEM grids and the resulting particle distributions were compared to more traditional sample preparation strategies like the “drop on grid” method. Operation parameters such as the sample flow-rate, capillary – substrate distance, electric field strength and sampling period length have been optimised. It was found that the particles deposited by electrospray generally show a much more homogeneous spatial distribution on the substrate and a substantial increase of the number of single particles, which are much better suited to an automatic image evaluation procedure than the agglomerated particles observed otherwise. The applicability of the technique to a broad range of materials is demonstrated by various examples, but also the influence of the substrate, the choice of the particular TEM grid, on the obtained spatial particle distribution is assessed.
The overall objective of project Improved traceability chain of nanoparticle size measurements is to improve the traceability chain for nanoparticle size measurements. The main impact will be achieved by manifold contributions to standard documents for CEN/TC 352 “Nanotechnologies”, which directly addresses the research needs of CEN, CENELEC and ETSI mandated by EC to develop standards for methods and reference materials to accurately measure the size and size distribution of nanoparticles. This will take place in collaboration with ISO/TC229 ‘Nanotechnologies’, ISO/TC24/SC4 ‘Particle characterization’ and ISO/TC201 ‘Surface analysis’/ SC9 ‘Scanning probe microscopy’.
The synthesis of TiO2 nanoplatelets with fluorine-containing reactants is carried out using Ti (IV) butoxide as precursor and concentrated HF as shape controller, the final product requires a working up in order to eliminate or at least to reduce the amount of residual fluorides, which is realized here by well-defined thermal treatment.
Qualitative investigation of the bulk elemental composition by means of EDX of TiO2 nanoparticles (NPs) has identified fluorine in case of the as-synthesized samples. EDX spectra of thermally treated products exhibit either a fluorine content close to the limit of detection. The latter holds also true for the reference sample, TiO2 NPs of bipyramidal shape and prepared by a different synthesis route. For differentiation whether fluorine is present in the bulk or at the surface of the TiO2 nanoplatelets, top-surface sensitive AES and ToF-SIMS has been applied. Secondary ions of fluorine are detected in ToF-SIMS spectra of all samples, but could be roughly quantified by measurement of same reference sample as for EDX, namely TiO2 nano-bipyramids. This revealed that the amount of fluorine within1 nm depth beneath the surface is reduced in the thermally treated specimen compared to the raw product down to a content about as low as in the reference sample. AES allows analyzing analysis of the first few nanometers from the top-surface of individual NPs by point analysis. An F KLL peak has been detected at the surface of samples of as-prepared TiO2 nanoplatelets under optimized measurement conditions, but was not detectable after their calcination, which is in agreement with ToF-SIMS results. Moreover, high resolution AES on single TiO2 nanoplatelets elucidated that the surface atomic layers surrounding the TiO2 nanopaltelet contain fluorides before thermal treatment of the NPs.
Self-assembly of TiO2 nanoparticles in multiple layers by layer-by-layer deposition has been selected of different deposition procedures usually applied for fabrication of TiO2 thin films with defined and homogeneous thickness on supports of interest for the large-scale applications. The substrates tested were: conductive (FTO) glass, silica glass and titanium alloy. The selected film fabrication technique consists of the deposition of alternating layers of oppositely charged, i.e. functionalized, TiO2 nanoparticle layers with wash steps in between. The controlled assembly of TiO2 nanoparticles on the supports surface requires both a proper functionalization of the supports to promote the adhesion of the TiO2 film to the substrates and proper functionalization of TiO2 nanoparticles to allow attachment to substrate and subsequent reaction between different NP layers. The current study focusses on the analytical control of the functionalization of the substrates with 3-Aminopropyltriethoxysilane (APTS) and glutaraldehyde (GA) by means of surface sensitive methods, XPS, Auger Electron Spectroscopy (AES) and ToF-SIMS. Chemical composition of surface of functionalized substrates shows differences in the degree and type of modification in dependence on substrate.
Evaluation of particle sizing techniques for implementation of the EC Definition of a nanomaterial
(2017)
Many techniques are available for measuring particle size distribution. For ideal materials (spherical particles, well dispersed) it is possible to evaluate the Performance of these methods. The performance of the analytical instrumentation for the purpose of classifying materials according to EC Definition is unknown. In this work the performance of commercially available particle sizing techniques on representative NanoDefine set of real-world testing materials (RTM) and quality control materials (QCM) for the implementation of the Definition is evaluated.
Electrospray ionization constitutes a promising deposition technique for high-resolution imaging. Particle distribution on TEM grids takes place homogeneously and no losses occur. Suspension must be appropriate (stabilizer may induce artefacts). ESI parameters need to be optimized for each material.
Volume specific surface area (VSSA) by BET: concept and demonstration on industrial materials
(2017)
Volume specific surface area (VSSA) as measured by BET constituites a simple and reliable solution to (most) powders. Porous, coated, polydisperse/multimodal materials are to be treated with care, i.e. doubled by analysis with electron microscopy or more advanced BET analysis (e.g. t-plot)for each material in part.
The potential of electrospray deposition, for the controlled preparation of particles for imaging in electron microscopes, is evaluated on various materials: from mono-modal suspensions of spherical particles to multimodal suspensions and to real-world industrial materials. It is shown that agglomeration is reduced substantially on the sample carrier, compared with conventional sample preparation techniques. For the first time, it is possible to assess the number concentration of a tri-modal polystyrene suspension by electron microscopy, due to the high Deposition efficiency of the electrospray. We discovered that some suspension stabilizing surfactants form artifact particles during electrospraying. These can be avoided by optimizing the sprayed suspension.
In the present paper we show an approach of measuring large numbers of nanoparticles in a single scan TKD. TiO2 anatase nanoparticles (NP) of bipyramidal shape were deposited on standard carbon grid used for TEM. The procedure used promoted formation of NP ‘monolayer’ islands with uniform distribution of NPs on the carbon surface which allowed mapping of large number of nanoparticles in the single island.
Collection of whole map covering ~2800 nanoparticles took nearly 20 minutes. Inverse pole figure color coded map indicates that the NPs are either lying on a {101} facet (within 10° range around perfect {101} parallel to the carbon surface orientation) on the carbon film or are lying on a {100} facet (within 10° range around the perfect {100} parallel to the carbon surface orientation). Very unlikely was the NP orientation standing on a {001} face. The NPs size distribution described as equivalent circle diameter (ECD) has been also evaluated and the mean NP ECD was 59 nm with standard deviation of 15 nm, i.e. in good agreement with electron microscopy or AFM results.
This study shows high potential of the technique for crystalline NPs analysis with respect to geometrical orientation of the particles on the substrate. With known orientation, the 3D dimensional characterisation of such non-spherical NPs becomes possible from 2D projection electron micrographs. Moreover, the NP size distribution can be easily extracted. Superior accuracies down to 1-2 nm are achievable. The approach is applicable also on thin lamellae extracted from particulate (or mesoporous) layers.