Electrical conductivity of multi-walled carbon nanotubes-SU8 epoxy composites

Abstract

We have characterized the electrical conductivity of the composite which consists of multi-walled carbon nanotubes dispersed in SU8 epoxy resin. Depending on the processing conditions of the epoxy (ranging from non-polymerized to cross-linked) we obtained tunneling and percolating-like regimes of the electrical conductivity of the composites. We interpret the observed qualitative change of the conductivity behavior in terms of reduced separation between the nanotubes induced by polymerization of the epoxy matrix.Carbon nanotubes, 20 years after their discovery are in the stage where the focus is more and more on their applications. The physical properties which are beneficial for applications are good electrical conductivity (σ), exceptionally high thermal conductivity, and their high mechanical strength. These properties are preferentially used in composites, where carbon nanotube fillers can provide the missing property of the matrix, for example the electrical conductivity. The metallic carbon nanotubes (CNTs) added in concentrations beyond the percolation threshold to the insulating matrix can give an antistatic composite, 1 transparent conducting electrode, 2 or other large area conducting materials. 3 In composite materials, one main issue is the possibility of combining the properties of the filler with those of the matrix. In this respect, SU8 matrix, which is a well-established engineering material for micro-and nano-fabrication, offers wide range opportunities. SU8 is an epoxy-based, negative-tone, UV-sensitive photoresist which besides SU8 resin, contains an organic solvent and a photoinitiator (PI) to provide crosslinking of the oligomers (see a schematic representation of the SU8 epoxy in Fig. 1). 4 One of the great advantages of SU8 is that it can allows fabrication of high aspect ratio three-dimensional structures in a broad range of thicknesses and hence it is often used in nano-and macrometer sized devices. 5 The disadvantages of SU8 are that it is an electrical insulator, it is brittle and it has a low thermal conductivity. It would be desirable not to have all these drawbacks, so numerous fillers materials were used to prepare SU8-based composites. [6][7][8] Here we report the electrical conduction of multi-walled carbon nanotubes-SU8 composite materials (hereafter CNTs-SU8) for a broad concentration range of well-dispersed CNTs. The CNTs-SU8 inks were prepared with and without PI, giving composites with and without cross-linked matrix, respectively. From the theoretical point of view, these two cases are equivalent to study the conduction mechanism of an assembly of CNTs in a solid and in a "liquid" matrix. Interestingly, beside the effects on the conductivity level, the functional dependence of σ on the CNTs' content noticeably differs for these two cases. a) Current address: Powder Technology Laboratory, Ecole Polytechnique Fédérale de Lausanne, CH-1015 Lausanne, Switzerland Multiwalled CNTs were synthesized by the chemical vapor decomposition method (CVD) at 640 • C using Fe-Co catalytic pa...

A percolation-based model for the conductivity of nanofiber composites

“…27 Diamonds represent cases in which the matrix was allowed to remain unpolymerized. The triangles and squares correspond to samples that were in each case crosslinked thermally at 150 • C ("soft baking"); the data represented by the squares connote materials that were additionally subjected to subsequent "hard baking" at 200 o C. The solid lines are calculated from Eqs.…”

“…A model has been developed that unifies the critical path approximation for electrical conductivity [24][25][26][27] with an approach to percolation that captures microstructural variables such as correlations and clustering among the conductive particles. The conductivity is shown to be sensitive to deviations from randomness in the spatial distribution of the particles, and the model addresses (at least partially) the variability that arises from the history of thermal treatment and method of preparation of the composites.…”

“…This critical shell thickness λ(φ) is anticipated to be a monotonically decreasing function of φ that depends upon the distribution over particle shapes and the microstructure of the composite. Within the critical path approximation (CPA), the conductivity σ (φ) due to tunneling amongst such an array of conductive particles may be estimated from: [24][25][26][27] …”

“…The length scale ξ is defined through the relation: g(r) = g 0 e −2r/ξ , where g(r) is the tunneling conductance between a pair of particles for which r represents the minimal distance separating the particle surfaces. [24][25][26][27] It should be noted that the tunneling conductance g(r) is assumed to decline in a continuous and monotonic fashion with the separation between particles, as opposed to in an "all-or-nothing" or step-function like manner. Monte Carlo simulations performed on monodisperse systems of both prolate and oblate spheroids have shown that Eq.…”

Carbon Nanotubes Dispersion in Polymers by Two-Factor Mechanical Method

“…The results were compared to the results of the authors G. Grimaldi et al [24], who studied the electrical conductivity of composites made on the basis of MWCNTs and epoxy SU8. CNTs were dispersed in the SU8 by sonication in the presence of surfactant.…”