1,720,986 research outputs found

    Datasets for Doctoral Thesis 'Structure/Property Relations of Graphene Oxide/Epoxy Nanocomposites'

    No full text
    This dataset supports the thesis entitled: Structure/Property Relations of Graphene Oxide/Epoxy Nanocomposites: Tailoring the Particle Surface Chemistry for Enhanced Electrical and Thermal Performance by Orestis Vryonis, awarded by the University of Southampton, September 2019, https://doi.org/10.5258/SOTON/T0021</span

    Structure/property relations of graphene oxide/epoxy nanocomposites: tailoring the particle surface chemistry for enhanced electrical and thermal performance

    Full text link
    In this study, graphene oxide (GO) of various surface chemistry configurations were characterized and then utilized as epoxy fillers with a main objective of enhancing the electrical and thermal performance of the matrix, without compromising the mechanical properties.The initial step of the study was to distinguish and establish the chemical pathways through which the surface chemistry of highly oxidized GO interacts with the crosslinking reactions of the matrix. For this, GO was produced with acidic oxidation, based upon potassium permanganate (KMnO4) and then characterized via Raman, thermogravimetric analysis (TGA) and X-ray spectroscopy (XPS), which revealed increased graphitic disorder and oxygen-based functionalities decorating the lattice. Afterwards, the GO was dispersed within the epoxy matrix via a solvent-based method, to give nanocomposites containing up to 2 wt.% of GO, a filler content that is sufficient for filler/matrix chemical interactions. The excess of epoxide groups in the system, associated with the GO surface chemistry, was confirmed with Fourier transform infrared spectroscopy (FTIR). These additional moieties react with the hardener consequently, displacing the reaction stoichiometry away from the optimum. The result of this is a change in the macromolecular architecture, which was revealed through the dielectric secondary relaxations. Furthermore, during post-curing (&gt; 100 oC), hydroxyl groups on the GO surface react with residual epoxide groups through etherification reactions, to give a marked increase in the glass transition temperature (Tg). These reactions lead to increased filler/matrix interfacial interactions and contribute to increased tensile performance. In addition, post-curing serves to partially reduce the defect content of the GO lattice which, in turn, slightly increases the electrical conductivity of the system.After establishing the chemical pathways of the GO/epoxy reactions and demonstrating the inefficient features of GO in enhancing the electrical and thermal properties of epoxy, an alternative surface chemistry should be sought. Thus, the second step of this study was to introduce an single-step synthetic route for the production of moderately oxidized GO (mGO), which would: allow enhanced electrical and thermal properties; maintain epoxy compatibility; ensure no adverse influence on the epoxy curing reactions and require potentially simplified material processingstrategies. This route included replacement of the KMnO4 with chromium trioxide (CrO3) as the oxidizing agent. The mGO was then characterized and contrasted with the previously synthesized GO and a commercially available low-oxygen graphitic product (edge-oxidized GO, eGO). Raman spectroscopy, TGA and XPS demonstrated a moderate level of oxidation and a reduced carbon defect content, compared to the GO and the eGO. Subsequently, the eGO and mGO were incorporated into the epoxy via a scalable high-speed mixing method and the respective nanocomposites were contrasted. Transmission and scanning electron microscopy showed a fine dispersion/exfoliation for the mGO and poor compatibility for the eGO which drastically affected the aspect ratio of the respective platelets. It was revealed that the mGO/epoxy interactions include slight perturbation of the epoxy crosslinking, albeit only at high filler contents (&gt; 12 wt.%), while the eGO did not react with the matrix at all. Ultimately, the mGO led to a low electrical percolation threshold (Pt) of ~1 wt.%; a maximum increase in electrical conductivity of about eight orders of magnitude and a maximum thermal conductivity increase of 200% compared to the unfilled epoxy, while the tensile performance of the system was not compromised. Conversely, the eGO/epoxy systems showed poor behaviour, with a Pt of ~10 wt.% and a maximum thermal conductivity increase of 150%, while the tensile performance was rapidly compromised. Those effects were attributed to the fact that mGO displays mildly oxygenated graphitic lattice - not only on the peripheral (as in the case of eGO) but also on the basal plane.Upon the single-step production of moderately oxidized GO surface chemistry, the possibilities of further improvements in terms of electrical and/or thermal performance had to be explored. Thus, the final step of this study was to graft various amino-terminated moieties onto the surface of mGO in an attempt to modify, furtherly, the interfacial interactions with the epoxy matrix. For this, the mGO was functionalised with two bifunctional molecules: poly(propylene glycol) bis(2-aminopropyl ether) of different molar masses (termed d230 and d4000 accordingly) and a trifunctional trimethylolpropane tris[poly(propylene glycol), amine terminated] reagent, termed t440. The grafting process was revealed to be successful via Raman spectroscopy, TGA and XPS, and the resulting functionalised (fGO) systems were termed d230/fGO, d4000/fGO and t440/fGO. It was shown that the grafting included typical epoxide-amine reactions that potentially increase the disorder onto the graphitic lattice, while the elevated temperatures of the process served to slightly reduce the initial mGO oxygen content. Afterwards, the abovementioned three fGO systems were incorporated into the epoxy where it was demonstrated by differential scanning calorimetry (DSC) that the presence of the grafted moieties affected the local fGO/matrix interfacial interactions and slightly perturbed the epoxy curing reactions. X-ray diffraction (XRD) revealed reduced graphitic stacking with increased reagent molecular mass, which eventually led in reduced Pt (0.5 wt.% with the usage of the d4000 reagent). Furthermore, the maximum electrical conductivity of the respective nanocomposites appeared to be slightly increased with increasing reagent molecular weight, an effect also related to the limited platelet stacking. For the same, potentially, reasons the thermal conductivity of the fGO-containing systems was adversely affected at low filler contents

    Octa-glycidyl POSS: an epoxy filler or co-monomer?

    No full text
    Particle tethering with the matrix permits propertymanipulation depending on the respective architecture andchemistry. However, integration of highly reactive moieties into epoxy networks may detrimentally affect the cure reaction stoichiometry. In this study, octa glycidyl polyhedral oligomeric silsesquioxane (OG POSS) was utilized either as a filler (simple addition into epoxy) or as a co-monomer (stoichiometric additioninto epoxy) to explore related effects. The examined samples were characterized by differential scanning calorimetry, and dielectric spectroscopy. The simple addition of OG POSS resulted in a considerable Tg decrease due to impaired stoichiometry, while the stoichiometric addition moderated this effect; these results are supported by the dielectric relaxation behavior

    Machine learning-enhanced metadata analysis for identifying polymer compositions

    No full text
    In materials science, accurately predicting and identifying polymer compositions from limited experimental data could potentially help in advancing material design, sustainability, and faster development. Traditional methods for analysing aged or complex polymers are often time-consuming and costly, requiring extensive measurements. Machine learning along with meta-data analysis and curation of an un-biased dataset offers a promising alternative, enabling quicker and notably accurate identification of materials or material properties.The aim is to integrate machine learning with metadata analysis to predict polymer-filler combinations from independently measured property values. As an introductory study, a machine learning pipeline using ensemble methods, including KNN, Random Forest and Neural Networks (MLP) with XGBoost as the final estimator combined via stacking, was developed. The model was fine-tuned using Bayesian optimization for efficient hyper-parameter tuning. All individual and combinations of models were evaluated for accuracy and computational efficiency.Evaluation was done on a dataset of various polymer-filler combinations obtained from NanoMine, the performance was assessed using top-k accuracy metrics and cross-validation score, based on the ability to correctly identify likely compositions based on user input. The model predicts the top three polymer-filler combinations with associated confidence levels.<br/

    DC conductivity characteristics of core-shell nanoparticles filled epoxy nanocomposites

    No full text
    Epoxy-based nanocomposites are extensively used for High-Voltage Direct Current (HVDC) insulation. Incorporating inorganic nanoparticles can modulate interfacial traps, but the influence of core-shell architectures remains unclear. Here, we investigate field-dependent DC conductivity in epoxy composites filled with conventional SiO2, Al2O3, TiO2 and SiO2-coated core-shell counterparts (SiO2@SiO2, Al2O3@SiO2, TiO2@SiO2) filled with a total surface area of ≃ 5.82 m2. Samples were fabricated via shear mixing and cured at 80 ◦C (2 h) and 125 ◦C (3 h). J-E measurements (1–10 kV mm−1) were analysed using power-law and Poole-Frenkel/Schottky empirical fits. Currents were fit with stretched-exponential decay to extractrelaxation time (τ ) and stretch exponent (β). Conventional SiO2 and Al2O3 demonstrated sub-Ohmic low-field behaviour and higher transition fields, while TiO2 remained near-Ohmic. Core-shell composites demonstrated Ohmic drift in all cases and raised transition fields to 3.6-4.1 kV mm−1, with significantly accelerated de-trapping (τ ≃ 6.2-6.4 s, β ≃ 0.52-0.54). Findings demonstrate that silica shells result in homogeneous traps, extending the Ohmic regime and deferring space-charge-limitedconduction without compromising core-driven conductivity enhancement

    Effect of Thermal Treatment on the Dielectric Performance of a Silicone Rubber

    No full text
    Poly(dimethylsiloxane) (PDMS) is the most widely known silicone rubber with many applications. Commercial PDMS products, such as the one featured here, might not be optimized for electrical applications, as suppliers tend to keep their products’ compositions secret. PDMS is synthesized by chemical routes that typically yield volatile by-products. Lately it has been demonstrated that a post-curing thermal treatment might be advantageous due to removal of such volatile species. In this study we investigate the effect of post-curing on the DC conductivity, as well as the dielectric performance of a popular, commercially available PDMS product. It is revealed that the elastomer’s performance improves in both aspects upon post-curing. Analytical methods such as thermogravimetric analysis (TGA), as well as Raman spectroscopy show removal of up to ~1.5 wt.% of cyclic oligomeric volatiles, as well as complete depletion of unreacted reactive species, upon post-curing

    Investigation of different commercial boron nitride grades and their effect on loss spectra in epoxy resins and silicone rubbers

    No full text
    Literature on nanodielectrics has shown inconsistent behavior with nominally identical material systems (polymer/filler combinations), sometimes even from the same research groups. This paper compares four different commercial hexagonal boron nitride (hBN) products in two different polymer systems (epoxy resins and silicone rubbers) and compares their dielectric response. It is demonstrated that the dielectric properties of composites containing hBN are affected by the production steps, respectively the supplier of the nanofiller; notably, this effect varies depending on the host polymer, thus trends cannot be translated from material system to the next, even for the filler from the same supplier. Experimental evidence shows that equivalent types of fillers can lead to variations in dielectric losses, both in silicone-based and epoxy-based composites

    An alternative synthesis route to graphene oxide: influence of surface chemistry on charge transport in epoxy-based composites

    No full text
    A synthetic route for the production of graphene oxide is described, in which the commonly used potassium permanganate (KMnO4) is replaced by chromium trioxide (CrO3) as the oxidizing agent. Raman spectroscopy, thermogravimetric analysis and X-ray photoelectron spectroscopy demonstrate that the product is characterized by a reduced level of oxidation and a reduced defect content, compared to conventional graphene oxide (GO). We therefore term the product moderately oxidized graphene oxide (mGO). In comparison with GO, it is shown that when introduced into an epoxy matrix, mGO offers significant potential benefits. These include: excellent compatibility with the epoxy matrix leading to a low percolation threshold for electrical conductivity (~ 0.5 vol%); an associated increase in electrical conductivity of about eight orders of magnitude; no adverse influence on the epoxy curing reactions; potentially simplified material processing strategies
    corecore