Stable, Fast, and Tunable Bilayer vOECTs via a Commodity Polymers Blending Approach for Specialized Spiking Neurons

The human brain deploys a complex network of interconnected spiking neurons featuring a variety of electrophysiological characteristics to perform complex classification and recognition tasks ^1,2,3. Although machine learning algorithms implemented on traditional von Neumann electronics have emerged for diverse computing applications, they lag behind biological calculators in terms of performance and power consumption¹. This “technological gap” stems from the brain’s massively parallel event-based processing of information, via interconnected neuronal networks, and the high specificity of its circuital elements, via neuronal diversity, which neuromorphic devices attempt to mimic^1,4. Advancements in organic neuromorphic devices over the last few years has led to several circuits that mimic neuronal spiking behaviors, but these systems are limited by complexity, large device footprints, high power consumption, while showing limited customizability ^{1,4,5,6,7,8,9}.
Replicating the vast diversity of neurons with a scalable fabrication strategy is a critical step towards the development of large scale organic artificial neural networks that can approach the computational capabilities and functionality of the brain¹. In my past work, neuromorphic systems have been developed via a novel bilayer vertical organic electrochemical transistor (vOECT), which was implemented in Hodgkin-Huxley spiking circuits, enabling reduced size footprint, low powered neurons, with a full control of the biomimetic neuron characteristics, including spiking threshold, refractory period, and frequency. However, the library of materials compatible with the bilayer vOECT fabrication process is restricted due to the lack of stable, fast n-type organic mixed ionic and electronic conductors (OMIECs), thus limiting the library of specialized and unique neuron types ^4,7,8,9.
Here, an alternative strategy is offered by blending an insulating polymer (PMMA) with various n-type OMIECs, including BBL and gNDI-gT2, leading to improved OMIEC characteristics. The components intermix in the solution state enabling the creation of an ion-percolation framework during film formation that facilitates ion-flux in the solid state leading to orders of magnitude improvements in device stability. Indeed, this improved ion penetration reduces the swelling-induced increases in contact resistance, thereby improving response time and stability. Furthermore, via the addition of insulating components, we reduce the amount of semiconducting material within a vOECT active layer, which reduces the device response time, without the subsequent reduction in thickness that can lead to electrical shorts between the top and bottom contacts. To demonstrate the benefits of this increased performance, these devices are leveraged to create low powered, small, fast Hodgkin-Huxley spiking neurons, where neuronal customization and specialization is achieved both through this blending and the bilayer strategy.
The blending approach formulates a critical tool for the fabrication of fast, stable, large scale organic artificial neural networks that are not only significantly reduced in size, due to the vertical structure, but also customizable in neuron characteristics, thereby providing a fundamental step towards diverse bio-mimetic and interconnected networks of neurons.

1. van de Burgt, Y. et al. Nat Electron, 1, 386–397. (2018)
2. Subkhankulova, T. et al. Front Mol Neurosci. 3, 10. (2010)
3. Ascoli, G. et al. Nat Rev Neurosci 9, 557–568 (2008)
4. Harikesh, P. C. et al. Nat. Mater. 22, 242–248. (2023)
5. Sebastian, A. et al. Nat. Commun. 10, 4199. (2019)
6. Huang, W. et al. Nature 613, 496–502. (2023)
7. Harikesh, P. C et al. Nat Commun. 13, 901. (2022)
8. Sarkar, T. et al. Nat Electron. 5, 774–783. (2022)
9. Laswick, Z., et al. Nat Commun 15, 6309 (2024)