Estudo de sinapses artificiais em transistores neuromórficos de porta eletrolítica utilizando mel como eletrólito

• Ano de defesa: 2025
• Autor(a) principal: Oliveira, Wesley Astudillo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Curitiba; Brasil; Programa de Pós-Graduação em Física e Astronomia; UTFPR
• Palavras-chave em Português: Transistores - Aspectos ambientais; Redes neurais (Computação); Inteligência artificial; Semicondutores; Eletrólitos; Transistors - Environmental aspects; Neural networks (Computer science); Artificial intelligence; Semiconductors; Electrolytes; CNPQ::CIENCIAS EXATAS E DA TERRA::FISICA; Física
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/38944

The advances in conventional electronics have progressed significantly since the advent of the invention of transistors. However, the characteristic pace of progress in this field has been challenged, as limitations related to energy efficiency, heat dissipation, and physical constraints become increasingly difficult barriers to overcome, compromising both the continuity of miniaturization and the improvement of the performance of electronic devices. In this context, neuromorphic computing emerges as a promising alternative, as it is capable of reproducing learning, storage, and data processing mechanisms in a single physical component, based on principles inspired by the functioning of the human brain, combining efficiency with parallel information processing. In this scenario, electrolyte-gated transistors (EGTs) stand out as a class of devices capable of performing such functions when applied as neuromorphic transistors. Among them, organic electrolyte-gated transistors show great potential because they can operate at low voltages without suffering significant losses in electrical conductivity. Along this line of research, there is also a pursuit to develop devices that are sustainable in their entirety or focused on just one of their layers. This set of requirements formed the basis for the present research, where an electrolyte-gated transistor was developed and studied, focusing on the application and electrical characterization as a neuromorphic transistor, using bee honey as a sustainable electrolyte. For this purpose, measurements were taken that included transfer curves, output curves, and transient curves (output current as a function of time). The analyses of the transient curves were successful due to the properties of the honey electrolyte, which is capable of generating different electrical conduction mechanisms and ionic coupling, in addition to enabling the fabrication of low environmental impact devices, also being a low-cost option. As main results, regarding the transfer curve, it was observed that the output current showed a direct dependence on both the intensity of VDS (drain-source voltage) and the sweep rate applied to VG (gate voltage), making it possible to achieve charge transport closer to the steady state when the device operated in the linear regime (VDS = −1.0,V ) with a sweep rate of 200, mV/s. The behavior in the saturation regime (VDS = −1.8,V ) was also analyzed, a condition in which device stabilization proved more difficult. Regarding the neuromorphic character, the ability to emulate short-term synaptic plasticity behaviors and, in specific cases, long-term ones was confirmed. It is concluded that, despite the limiting factor associated with prolonged response times, it was possible to explore the neuromorphic properties of the EGT with honey as an electrolyte, consolidating its potential for use in applications that require sensing and sustainability.