Conductive Carbon Nanotube Composite Microprobes

Because of their outstanding electrical and mechanical properties, including high stiffness and mechanical resilience, ballistic electron transport at micrometer scales, and high current-carrying capacity, carbon nanotubes (CNTs) could enable a new class of electrical elements ranging from transistors to interconnects. However, realizing the properties of individual CNTs in assemblies of CNTs has been a formidable challenge. Realistic applications of CNTs at the micrometer scale must employ thousands or millions of CNTs in a parallel fashion, yet, the understanding of the electromechanical behavior is still not mature enough. Most of the research so far has been either for sensing or actuation applications, whereas the potential benefits of CNTs in probing applications, where both mechanical integrity and electrical conduction is critical, have not been investigated as yet. Electromechanical probing applications continuously require smaller pitches, faster manufacturing, and lower electrical resistance. Achieving low-pitch structures without compromising elastic properties is a challenge with conventional techniques, such as in microelectromechanical systems (MEMS)-based structures. CNT-based structures present a scalable manufacturing approach, in which thousands of probes can be fabricated in very short production times and by means of a full wafer batch fabrication mode, reducing the fabrication steps currently necessary with conventional methods, all of which can have significant cost benefits. Growth of vertically aligned CNTs (VA-CNTs) by thermal chemical vapor deposition (CVD) has created films and microstructures containing large numbers of CNTs aligned in parallel. Unfortunately, these high-temperature synthesis processes are typically optimized for growth from metal nanoclusters on silicon and ceramic substrates; whereas devices utilizing these CNTs will require a wider variety of substrates, including metals and plastics that cannot withstand the harsh processing conditions for high-yield CNT growth. To our knowledge, growth of VA-CNTs by thermal CVD with simultaneous ohmic contact on electrically conductive substrates has not yet been achieved and will remain a major challenge because of the apparent necessity of a buffer layer, such as Al2O3, SiO2, or MgO, for high-yield VA-CNT growth. Furthermore, weak mechanical adhesion to the substrate and low bulk density often prevent robust electrical and mechanical integration of CNTs under as-grown conditions. An emerging approach for device integration of VA-CNTs is to transfer the VA-CNTs from the growth substrate to a second “device” substrate. Full films of VA-CNTs have been transferred with a main emphasis on field-emission applications; however, the nature of interconnection to the CNTs and the lengthwise electrical and mechanical properties of the transferred films have not been assessed. CNTs have been previously embedded in an otherwise insulating matrix to form conducting composites, but these are not aligned CNTs. Most of the work on CNT-based composites presented in the literature focuses on using CNTs as a reinforcement, or as a filler, in a polymeric matrix by dispersing and perhaps subsequently aligning singleor multiwalled CNTs (SWNTs or MWNTs, respectively) in the matrix. According to Thostenson et al., the most influential factors for the mechanical properties of the hybrid composites are: CNT length, matrix–CNT adhesion, and, in particular, alignment and dispersion of the CNTs within the matrix. Alignment and dispersion are critical factors that are difficult to control experimentally. Furthermore, poor adhesion to the matrix and stress concentrations compromise the effectiveness of the CNTs as reinforcement. Unlike approaches that disperse bulk, randomly oriented CNTs in a matrix, this work focuses on the growth of aligned CNTs from patterned catalyst areas, and then their transfer onto a conductive substrate; a process which simultaneously results in fast, robust, and repeatable composite structure formation, and avoids the above-mentioned shortcomings. Demonstrations of transfer processes of VA-CNTs in the literature have been on bulk quantities of carbon nanotubes and more specifically on aligned CNT films mostly aimed for field-emission applications, where no electrical measurements are presented through the thickness of the transferred film. To the best of our knowledge, the only electromechanical characterization (excluding actuation demonstrations) of transferred VA-CNTs presented in the literature is where the resistance perpendicular to the array is measured as the C O M M U N IC A IO N