%0 Generic
%A Kollenz, Philipp
%A Herrle, Carina
%A Göhringer, Leonard
%A Wickenhäuser, Tom
%A Pernice, Wolfram
%A Klingeler, Rüdiger
%A Deschler, Felix
%C Heidelberg
%D 2025
%F heidok:35906
%R 10.11588/heidok.00035906
%T Excited State Reservoir Computing using Hybrid Perovskite Electrochemically-Gated Luminescent Cells
%U https://archiv.ub.uni-heidelberg.de/volltextserver/35906/
%X Physical reservoir computing aims to increase computational efficiency of machine learning tasks by shifting the computational burden to a physical system. Reservoirs based on ion dynamics are of particular interest due to the non-linear and integrative nature of ion transport. Here, we demonstrate all-optical operation of a physical reservoir based on electrochemical Li-ion doping of lead halide perovskite microcrystals. Optical excitation changes lithium ion insertion kinetics, which in turn modulate the luminescence response. The heterogenous structure of the crystals leads to a large internal state space of the reservoir. The device can be fabricated using solution-based fabrication and operated using LED illumination, reducing fabrication cost. Our proof-of-concept results demonstrate optically excited state dynamics and ion transport as a promising platform for physical reservoir computing.