What Controls Kinetics in Organic Mixed Conductors for Neuromorphic Computing and Beyond?

University Of Washington, Seattle WA

Investigators

Abstract

Non-technical Description Plastics that conduct both electrons and ions are important for many applications, from sensors that measure brain activity to batteries that store energy. One emerging application for these materials is in computers that can mimic the learning ability of the human brain at the hardware level. Such biologically inspired, or neuromorphic, computers could potentially replicate the learning process and perform key tasks faster, and with lower energy consumption, than today’s computers. Just as neurons in the brain can “learn” over time through repeated activation, devices based on these conducting plastics can change how easily they conduct electrons based on an input. One key limitation in this field is understanding how ions and electrons move through such materials together as a function of time. For example, it is not clear why plastics can change very slowly to reach a high-conductance state, yet they exhibit a rapid change when turned off to a low-conductance state. This project investigates these properties using a range of different techniques that measure how the electrical device performance is affected by the solid structure of the materials, the device geometry, and the chemical properties of the system. The scientific knowledge from this project will enable better understanding of how polymers and other materials can be designed for better next-generation computing devices. The project also extends the principal investigator’s role in education by developing new outreach materials and by supporting local organizations that help first-generation college students to achieve scientific careers. Technical Description The scientific goal of this project is to better understand the structure/function relationships that govern the performance of organic semiconductors in neuromorphic computing devices. Organic mixed ionic-electronic conductors (OMIECs), typically conjugated polymers, are well-suited to these systems because they can efficiently accommodate ions, resulting in tunable changes in conductance state. This property makes them amenable to applications where controlled “learning” via a voltage-induced conductance change is desired, as in hardware-based artificial neural networks. However, it is currently unclear how different chemical and morphological properties of OMIECs control ion transport kinetics, hysteresis, and non-linear response. A successful neuromorphic device should be able to change conductance quickly with long-lived state retention, and either linear or highly non-linear response depending on the application. This project explores the interconnected factors of kinetics, non-linearity, and geometric scaling by 1) investigating kinetics of ion injection and expulsion using different polymer and counterion combinations; 2) probing non-linearity due to active layer and gate electrode composition; and 3) testing how kinetics and non-linear responses in OMIECs translate to transport measurements in transistors to relate geometric scaling in the device architecture with neuromorphic function. To accomplish these goals, this project combines spectroelectrochemistry, electrical scanning probe microscopy, time-resolved optical microscopy, and electrochemical transistor device measurements to provide insight into how characteristic length scales in OMIECs and local structure affect the measure transport properties and neuromorphic device functionality. The research activity here provides important insight into how the various chemical and morphological factors are interrelated, while also providing guidance for the rational design of better conjugated polymers and other materials for neuromorphic applications. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.