Abstract: Extracting relevant data from real-world experiments is often challenging with intrinsic materials and device property dispersion, such as in organic electronics. However, multivariate data analy-sis can often be a mean to circumvent this and to extract more information when larger datasets are used with learning algorithms instead of physical models. Here, we report on identifying rele-vant information descriptors for organic electrochemical transistors (OECTs) to classify aqueous electrolytes by ionic composition. Applying periodical gate pulses at different voltage magnitudes, we extracted a reduced number of nonredundant descriptors from the rich drain-current dynam-ics, which provide enough information to cluster electrochemical data by principal component analysis between Ca2+-, K+-, and Na+-rich electrolytes. With six current values obtained at the ap-propriate time domain of the device charge/discharge transient, one can identify the cationic iden-tity of a locally probed transient current wit h only a single micrometric device. Applied to OECT-based neural sensors, this analysis demonstrates the capability for a single nonselective de-vice to retrieve the rich ionic identity of neural activity at the scale of each neuron individually when lea rning algorithms are applied to the device physics.

Abstract: Organic electrochemical transistors offer powerful functionalities for biosensors and neuro-inspired electronics, with still much to understand on the time-dependent behavior of this electrochemical device. Here, we report on distributed-element modeling of the impedance of such micro-fabricated device, systematically performed under a large concentration variation for KCl_(aq) and CaCl_2(aq). We propose a new model which takes into account three main deviations to ideality, that wer e systematically observed, caused by both the materials and the device complexity, over large frequency range (1 Hz-1 MHz). More than introducing more freedom degree, the introduction of these non-redundant parameters and the study of their behaviors as function of the electrolyte concentration and applied voltage give a more detailed picture of the OECT working principles. This optimized model can be further useful for improving OECT performances in many applications (e.g. biosensors, neuro-inspired de vices, ...) and circuit simulations.