(39b) Nanoparticle Structural Organisation and Scaffolding in Organic Media

NANOPARTICLE STRUCTURAL ORGANISATION AND SCAFFOLDING IN ORGANIC MEDIA

To create advanced nanostructured materials and fluid media with unique tailored properties, it is often necessary to form nanocomposites or nanodispersions, which may include homogenous and heterogeneous mixtures of nanoparticles dispersed in a matrix or medium of another compound such as a polymer.  Both nanocomposites and nanofluids have recently started to find important applications due to their exceptional properties. 

Recent and ongoing research on polymer/inorganic nanocomposites has shown dramatic improvement of the performance properties, such as stiffness, strength, electrical, optical and thermal properties over those of unfilled polymers, without compromising their processibility. 

The aim of this study was to examine the structuring of nano-particles within liquid based epoxy resins and their ability to form clusters and scaffolds in which the nanoparticles form a network with potentially interesting end properties.  In particular, we considered nanocluster formation and growth as function of the dispersion and concentration.

A low molecular weight DGEBA epoxy resin was used as a liquid oligomer matrix which could be converted to a solid crosslinked polymer network locking in place the nano-particle structures formed in the liquid phase.  The nanopowder system used was SiO₂ with mean diameter of 12 nm.  Samples with different concentrations of nanoparticles were prepared in the range 1% to 40% by weigh and after mixing the samples were cured and heat treated in specially prepared moulds. Following cure the samples were sectioned to typically 50 to 100nm sections which were studied by transmission electron microscopy (TEM).  The curing process is a chemical reaction to form highly cross-linked, three-dimensional networks.  Curing conditions, rate of epoxy resin to curing agent, type of curing agents are some of the important factors affecting the behaviour of epoxy/nanocomposites.

In order to uncover the morphological changes that take place in different samples, it is necessary to carry out a systematic analysis of the TEM images such as given in Figure 1a.  Evolution of the number of clusters, cluster size and shape were determined for samples across the concentration range.  In order to do this, the objects in the TEM images were distinguished from the background using thresholding image analysis methods.  When the thresholding algorithm developed was applied to the image shown in Figure 1a, the thresholded image shown in Figure 1b was obtained.  The features at the top left and bottom left were lost during thresholding.  If the image is carefully examined, those parts of the image are darker than the rest.  There is a non-uniform background illumination.  This may happen frequently in light microscopes due to non-uniform illumination or varying thicknesses of samples in transmission imaging.

There are a number of techniques that can be employed to estimate the background in an image.  One can use a morphological opening operation with a structuring element to account for the behaviour of the background.  The morphological opening has the effect of removing objects that cannot completely contain the structuring element.  The size of this structuring element is crucial here and might need prior information of the cluster sizes.

Another way to estimate the background is to utilize the surface equation.  By fitting a surface equation through the points or grey levels, one can estimate the background.  The problem is a least squares problem.  Figure 1c shows the estimated background of the Figure 1a. 

            (a)                                (b)                                (c)

            (d)                                (e)                                (f)

Figure 1: a) Original image 2% wt concentration, 20000 x magnifications. b) Original image thresholded. c) Extracted background. d) Background subtracted, non-uniform background corrected. e) Thresholded image. f) Colour labelled nanostructures using cluster analysis.

Fig. 1d shows background or non-uniform corrected image and Fig.1e shows the thresholded image.  The thresholding obtained is now a very good representation of the original image. 

Using this approach it was possible to extract information about percolating nanostructures using cluster analysis.  A cluster analysis algorithm developed has successfully identified nanoclusters and they are shown with different colour labels in Fig. 1f.

                   (a)                                     (b)                                (c)                                         (d)

Figure 2: Original images with 40000 x magnifications with concentration of a) 1%wt, b) 6%, c) 10% , d) 20%

Fig.2 shows TEM images with different concentrations.  They illustrate a very high degree of uniformity in the mixing with dendritic aggregates.  The aggregates appear to sustain a consistent morphology (fibrillar dendritic with a tendency to from loops) whose size and number grow with increasing concentration.  At higher concentrations, it appears that we are close to micro-percolation with extended connectivity occurring over length scales of 0.1 to 0.8 microns which is many more times greater than the nanoparticle size. 

The relationship between length and aspect ratio and number of particles as a function of concentration was assessed and will be discussed.  The general morphology and development of aspect ratio was also explored along side the fractal dimension and radial distribution functions that characterise the longer range structural order in the systems studied.  The cluster analysis mentioned was also extended to evaluate individual clusters as well as inter cluster spacing. 

The structural information obtained for the nano-particle cluster networks will be discussed and related to the differences in the measured macroscopic physical properties of the samples such as tensile strength and thermal conductivity.

References

1-      Becker O., Varley R., Simon G., Polymer, 43, (2002), 4365-4373

2-      Alexandre M., Dubois P., Materials Science and Engineering, 28 (2000) 1-63

3-      Arias M., Frontini P.M., Williams R.J.J, Polymer 44 (2003) 1537?1546

4-      Fujiwara M., Kojima K., Tanaka, Y, Nomura, R, J. Mat. Chem., (2004), 1195-2002

5-      Chiang, C-L., Ma M. C-C. European Polymer Journal 38 (2002) 2219?2224

6-      Ochi M., Takahashi R., Terauchi A., Polymer 42 (2001) 5151-5158