10:30 〜 10:45
▲ [16a-B508-6] Enhancement of electrical properties and emergence of temperature dependence of PCBM network mediated with Au dopants
キーワード：fullerene derivative, electrical conductivity, doping
Artificial Neural Network that mimics the function of the brain is becoming popular, and it is demonstrated as a device such as voice recognition and image identification. Recently, van der Wiel and coworkers[1] demonstrated nonlinear classification using a system with a disordered dopant atom network in silicon. They clarified the nonlinearity of hopping conduction through an electrically tunable network of boron dopant atoms. However, the nonlinear properties emerged only at low temperatures, and these devices are desired to work at room temperature for future applications. To achieve the nonlinear network at room temperature, organic molecules can be utilized. The conductivity of molecules is limited due to the thermally–activated process, however, it can be regulated with an appropriate dopant. Here, we proposed PCBM/Au system. PCBM ([6,6]-phenyl-C61-butyric acid methyl ester) is a non-conductive fullerene derivative, but soluble and easily fabricated as a film. The comb-type electrode substrate was ultrasonically cleaned with acetone, followed by UV-Ozone treatment. Then the substrate was subject to hydrophobic treatment by using octadecyl trichlorosilane. After that, the substrate was spin-coated by PCBM dissolved in chlorobenzene. Later, the substrate was annealed. Finally, it was deposited under Au under a high vacuum. We found the different amount Au doping showed different temperature dependence. (Fig.1a,b). According to the variable-range hopping model, at 3.4 ML (ML:monolayer) of Au doping, the hopping distance is large, and it shows a temperature dependence. At 31 ML of Au doping, it shows a metallic properties.