•By in-situ growing bamboo-like CNTs on CB, a developed conductive network was constructed and plentiful CB/CNT interfaces were introduced to improve the electromagnetic wave loss ...ability.•Defect-rich CNTs with open atomic edges can endow nanocarbons with a better balance between good impedance matching and strong loss capacity, causing an effective microwave absorption.•The minimum reflection loss and effective absorption bandwidth of CB-CNT with the CNTs growth time of 40 min reach to −53.6 dB and 4.1 GHz with the matching thickness of 2.7 mm, respectively.
Due to the limited electromagnetic wave (EMW) loss capacity and agglomeration, carbon black (CB) gradually fails to meet the increasingly harsh demanding conditions. Herein, defect-rich bamboo-like carbon nanotubes (CNTs) were grown on CB by the process of chemical vapor deposition. CNTs prepared in situ on CB can assist it to build a developed multilevel conductive network and introduce plentiful CB/CNTs nano-interfaces. What's more, the defects that accompany the growth of CNTs endow CNTs with a moderate conductivity and good impedance matching, thereby causing an effective microwave absorption (MA). Meanwhile, the high-density defects on CNTs can induce dipole polarization to further strengthen the EMW loss ability. The influence of CNTs with different growth time on MA performance has been explored. Profiting from the structural merits, the synthesized CB-CNT with CNTs growth time of 40 min exhibits the optimal absorbing property, which has the minimum reflection loss of -53.6 dB and maximum effective absorption bandwidth of 4.1 GHz with the thickness of 2.7 mm, covering almost the entire X band. The introduction of defect-rich CNTs significantly enhances the EMW loss ability of CB, which provides a rational strategy for the design of high-efficient microwave absorption materials.
Pour la confection des matériaux polymères à base de caoutchouc, le noir de carbone (NC) reste la charge renforçante la plus utilisée. Sa caractérisation morphologique et nano structurale est ...essentielle dans la maitrise des propriétés physico-chimiques qu’il confère aux matériaux auxquels il est mélangé. Les analyses classiques ne permettent d’accéder que de façon indirecte et incomplète à ces propriétés. Cette thèse propose une méthode de caractérisation innovante par le couplage d’un détecteur STEM (Scanning Transmission Electron Microscopy) et d’une chaine d’analyse d’images complètement automatique pour identifier les types de NC. Une étude statistique approfondie d’une centaine de caractéristiques morphologiques et structurales des NC a été réalisée sur les 6000 images STEM acquises. Cette étude a permis d’introduire 7 nouveaux descripteurs et de sélectionner les 37 descripteurs les plus discriminants pour la création du modèle d’identification. Pour rendre le processus le plus automatique possible, un algorithme de segmentation non supervisé a été développé et évalué. Cinq classifieurs ont ensuite été entraînés et comparés sur une base de près de 65000 agrégats. Le modèle le plus adapté s’avère les réseaux de neurones avec une identification des NC avoisinant les 100%. L’identification étant réalisée à partir de projections 2D des agrégats via les images STEM, une confrontation statistique valide la capacité des descripteurs 2D à caractériser la structure tridimensionnelle des NC observée par tomographies électroniques. L'approche complète proposée, depuis le protocole de préparation des échantillons et l'acquisition d'images STEM jusqu'à leur classification en passant par les étapes d'analyse d'images, offre une nouvelle méthode de caractérisation fiable des NC (à l’état natif ou au sein de mélanges élastomères) exploitable en routine.
In the field of rubber material development, CB is the most commonly used reinforcing filler. The characterization of CB morphology and nanostructure is therefore crucial to understand the physicochemical properties induced by the introduction of CB in rubber materials. Classical analytical methods only allow indirect and incomplete access to these properties. This PhD offers an innovative method that allows the automatic identification of CB grades by coupling Scanning Transmission Electron Microscopy (STEM) detector and image processing chain. A thorough statistical investigation over a hundred of morphological and structural characteristics of CB was performed on a set of 6000 STEM images. This study has introduced 7 new features and selected the 37 most discriminating descriptors to create the final model. An unsupervised segmentation algorithm has been developed and evaluated in order to build an automatic process as efficient as possible. Then, five classifiers were trained and compared on a base of nearly 65,000 aggregates. It appears that the most suitable descriptor is the Neuron networks as it gives a perfect recognition. As the recognition model is based on 2D projections of CB aggregates, it is necessary to verify that the chosen descriptors are indeed able to correctly characterize the three dimensional structure of CB. The statistical comparison of the 2D descriptors with 3D descriptors extracted from electronic tomography images has been successful, and therefore demonstrates the relevance of the model. The proposed approach, starting from the sample preparation and STEM acquisitions to their classification and through the image analysis steps, offers a new and innovative method for the reliable characterization of CB. This method can be used routinely on raw CB or CB extracted from vulcanizes rubbers.
