Artificial synapse based on carbon quantum dots dispersed in indigo molecular layer for neuromorphic applications

Abstract

Conventional computers are limited in their performance due to the physical separation of the memory and processing units. To overcome this, parallel computation using artificial synapses has been thought of as a possible replacement in computing architecture. The development of nanoelectronic devices that can show synaptic functionalities is very important. Here, we report the robust synaptic functionalities of carbon quantum dots embedded in two terminal indigo-based organic synapses. The carbon quantum dots (CQDs) are prepared using an easy-to-do process from commercial jaggery. The CQDs have a size range between 3.5 and 4.5 nm with excellent light emission in the green region. CQD+indigo-based devices show extremely stable memory characteristics, with ON and OFF states differing by more than 10 Mohm. Devices show excellent long-term potentiation and long-term depression characteristics, with both synaptic weight updates following a double exponential behavior. The extent of nonlinearity is explained using the nonlinearity factor. The linear increase in memory is established with repeated learning and forgetting (or potentiation and depression) curves. This study gives a robust way to make an artificial synapse work efficiently at room temperature with excellent memory and synaptic behavior.