Towards low-energy consumption artificial synapses from lead halide perovskite
Jeroen de Boer a, Bruno Ehrler a
a AMOLF, Amsterdam, Netherlands, Netherlands
International Conference on Hybrid and Organic Photovoltaics
Proceedings of International Conference on Hybrid and Organic Photovoltaics (HOPV22)
València, Spain, 2022 May 19th - 25th
Organizers: Pablo Docampo, Eva Unger and Elizabeth Gibson
Poster, Jeroen de Boer, 254
Publication date: 20th April 2022

Lead halide perovskites are highly promising materials for a wide range of optoelectronic applications, such as photovoltaics and LEDs. However, perovskite optoelectronic devices typically display large hysteresis effects under mild operating conditions, which is unwanted for these applications. At the same time these large hysteresis effects under mild conditions are highly desirable for artificial synapses, which exploit hysteresis to change the conductivity of the device. Incorporation of these synapses into artificial neural networks has great potential for highly efficient neuromorphic computation. 

Artificial synapses based on lead halide have been demonstrated before[1]. Low energy consumptions were possible for these synapses made of methylammonium lead iodide (MAPbI3), due to the diode behavior of the synapse and the low voltages at which hysteresis occurred. We estimate that downscaling devices based on methylammonium lead iodide (MAPbI3) to device areas of 1 µm2 could already bring the energy consumption of switching events of the synapse to values as low as 1 to 100 fJ, the same range as that of biological synapses. However, the poor chemical stability of MAPbI3, and lead halide perovskites in general, makes downscaling of the device with standard lithography techniques difficult. Here, we investigate a photolithography procedure that is compatible with MAPbI3 to prepare MAPbI3-based artificial synapses with device areas between 1 and 10 000 µm². We show that downscaling of the synapse with this method is indeed a viable strategy to reduce the energy consumption of the device. The goal is to further develop lead halide perovskite artificial synapses with the same energy consumption as biological synapses for ultra-low power physical computing systems.

We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info