J.M. Gorham, W.A. Osborn, J.W. Woodcock, K.C.K. Scott, J.M. Heddleston, A.R. Hight Walker, J.W. Gilman
Keywords: mult-walled carbon nanotubes, MWCNT, nanofillers, nanoparticle release
Summary:Multi-walled carbon nanotubes (MWCNT) and other carbon nanofillers are used in different applications to enhance the materials and electrical properties of consumer products as polymer composites. As a result, concern has arisen regarding the potential for nanoparticle release during the life cycle nanocomposites. To address this concern, an approach to (A) detect and (B) characterize dispersion properties of MWCNT at the polymer composite’s surface by imaging X-ray photoelectron spectroscopy (XPS) was developed1. Detection and quantification of MWCNT in a polymer composite by XPS is typically complicated by problems with separating carbon photoelectrons from filler and matrix components of the composite. Typically, the carbon photoelectrons emitted from those materials are comparable making them challenging to separate2. Recent studies have demonstrated that for MWCNT composite systems, the photoelectrons associated with polymers differentially charge and separate from the MWCNT photoelectrons making qualitative to semi-quantitative identification of MWCNT by XPS in traditional spectroscopy mode possible3-4. In attempts to determine the cause of the observed differential charging, we probed the micrometer scale dispersion and electrical properties of MWCNT in an epoxy matrix using parallel, or hyperspectral, XPS imaging1. The current presentation will demonstrate the feasibility of detecting MWCNT –rich regions in 1%, 4%, and 5% MWCNT-epoxy composites with dimensions as small as 10 μm. For each sample, a series of images were taken over evenly spaced intervals to obtain images over the entire C (1s) binding energy scale. Using code written at NIST, the series of images were processed into conductive, MWCNT-rich and insulating epoxy rich regions of interest (ROI) in order to extract spectra from each ROI. The extracted spectra confirmed that the conductive ROI was predominantly composed of MWCNT. The extracted spectra suggested the nonconductive ROI was more ‘composite’ like, with the MWCNT contribution impacting the degree of charging. SEM and Raman imaging were employed as orthogonal techniques to verify this approach. Results demonstrate that XPS imaging can separate MWCNT signals from epoxy, providing information on the degree of heterogeneity of MWCNT dispersion within the composite as well as its electrical properties.