Dynamic Molecular Switches Based on HATNA Derivatives for Application in Neuromorphic Computing

Currently, about 5%–15% of the world’s energy is spent in some form of data manipulation and this fraction is expected to rapidly increase. Between 20 and 30 MW are estimated for von Neumann type computers to emulate same level of complexity as our brains. Our brain works with 20 W.¹ Neuromorphic computing architecture focuses on building electronic neural processing systems that directly emulate real neurons and synapses. Currently, most approaches to neuromorphic computing rely on high energy consuming silicon materials.²An attractive energy efficient solution is given by molecular switches which can potentially simulate synaptic plasticity. However, up to now, molecular switches can only switch between fixed on and off states, hence showing a static rather than dynamic behaviour. The key factor to reproduce synaptic plasticity is coupling different processes each characterised with its own time constant, analogues of action potential coupled to slow diffusion of Ca²⁺ and neurotransmitters in synapses. 5,6,11,12,17,18- hexaazatrinaphthylene (HATNA) in a tunnelling junction is a promising candidate because it can couple a fast electron transfer to slow proton transfer.³ Synthetic modification of HATNA with diphenylacetylene (DPA) moieties can potentially facilitate tunneling electron transfer rates in the junction while at the same time enhance packing density of molecules on the electrode.^4,5 In this poster we give an overview of the synthesis route to achieve the dynamic switch HATNA-DPA and initial monolayer composition studies.<br/><br/><br/><br/>1. Christensen, D. v <i>et al.</i> 2022 roadmap on neuromorphic computing and engineering. <i>Neuromorphic Computing and Engineering </i><b>2</b>, 022501 (2022).<br/>2. Upadhyay, N. K. <i>et al.</i> Emerging Memory Devices for Neuromorphic Computing. <i>Advanced Materials Technologies</i> <b>4</b>, 1800589 (2019).<br/>3. Nijhuis, C. <i>et al.</i> Dynamic Molecular Switches Drive Negative Memristance Emulating Synaptic Behaviour. (2022) doi:10.21203/RS.3.RS-1156230/V1.<br/>4. Zhang, Z. <i>et al.</i> Energy-Level Alignment and Orbital-Selective Femtosecond Charge Transfer Dynamics of Redox-Active Molecules on Au, Ag, and Pt Metal Surfaces. <i>J. Phys. Chem. C</i> <b>125</b>, 18474–18482 (2021).<br/>5. Yuan, L. <i>et al.</i> Transition from direct to inverted charge transport Marcus regions in molecular junctions via molecular orbital gating. <i>Nature Nanotechnology 2018 13:4</i> <b>13</b>, 322–329 (2018).