Multiscale characterisation of particle distribution in nanocomposites across two and three dimensions
File(s)
Author(s)
Fisher John, Sandra
Type
Thesis
Abstract
There are currently a wide range of carbon nanocomposite processing techniques which allow access to a variety of nanoparticle dispersions and distributions. However, in practice, dispersions and distributions within polymer matrices are often inhomogeneous resulting in local variations in microstructure which impact material performance. Understanding the processing-structure-properties relationship is critical in developing the next generation of carbon nanocomposites, as is the need for quantitative methods of characterising the arrangement of fillers in nanocomposite.
In this thesis, new protocols for characterising the dispersion and distribution of particles in nanocomposites are presented; the methodologies address the multiscale and multidimensional nature of the structures. Complementary Raman spectroscopy and optical microscopy techniques, based on the chemically specific vibrational modes and broad optical absorbance, respectively, are used to quantitatively map the local concentrations of particles. The multiscale representation of the distribution is then used to derive an agglomeration index by wavelet transform texture analysis, allowing objective comparison between samples. Providing higher resolution and more rapid data acquisition, the optical microscopy method is extended to 3-dimensional analysis of the particle distribution via array tomography. Covering a region greater than 100 μm, the volumetric characterisation provides clarity to the orientation and size of anisotropic structures on the micron scale. Serial block face scanning electron microscopy extends the 3D analysis down to the few nanometre scale, allowing tracing of the true particle trajectories for analysis of the dispersion and particle-particle interactions. These methods, applied to multi-walled carbon nanotube, graphene nanoplatelet and carbon nanofibre composites, provide a comprehensive model of the particle arrangements across length scales; the implications of this data in process development and simulations are also discussed.
In this thesis, new protocols for characterising the dispersion and distribution of particles in nanocomposites are presented; the methodologies address the multiscale and multidimensional nature of the structures. Complementary Raman spectroscopy and optical microscopy techniques, based on the chemically specific vibrational modes and broad optical absorbance, respectively, are used to quantitatively map the local concentrations of particles. The multiscale representation of the distribution is then used to derive an agglomeration index by wavelet transform texture analysis, allowing objective comparison between samples. Providing higher resolution and more rapid data acquisition, the optical microscopy method is extended to 3-dimensional analysis of the particle distribution via array tomography. Covering a region greater than 100 μm, the volumetric characterisation provides clarity to the orientation and size of anisotropic structures on the micron scale. Serial block face scanning electron microscopy extends the 3D analysis down to the few nanometre scale, allowing tracing of the true particle trajectories for analysis of the dispersion and particle-particle interactions. These methods, applied to multi-walled carbon nanotube, graphene nanoplatelet and carbon nanofibre composites, provide a comprehensive model of the particle arrangements across length scales; the implications of this data in process development and simulations are also discussed.
Version
Open Access
Date Issued
2019-08
Date Awarded
2019-12
Copyright Statement
Creative Commons Attribution NoDerivatives Licence
License URL
Advisor
Shaffer, Milo
Sponsor
European Union
European Commission
Grant Number
642890
Publisher Department
Chemistry
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)