Helicity-Defined Analog Conductance States in Chiral Perovskite Synapses.

Optoelectronic synapses encode light through scalar variables (i.e., intensity or wavelength), leaving inputs that differ only in field symmetry indistinguishable at the hardware level. Circularly polarized light introduces helicity as a symmetry-defined degree of freedom; however, the conversion into persistent analog memory remains challenging. This study demonstrates that optical helicity biases the electrically driven formation of persistent analog conductance states in chiral quasi-two-dimensional perovskite synapses. Within the (R/S-PEA)2MAn -1PbnI3 n +1 multi-quantum-well series, the n = 3 composition provided the optimal convergence of chiroptical selectivity, layered structural registry, and transport-supporting structural order, enabling reproducible low-nonlinearity analog weight updates and 72 statistically distinguishable conductance states retained over 106 s. An independently evaluated read-noise-informed framework yielded a read-noise-limited capacity estimate of 72.3 levels, providing a quantitative link between measured read noise and analog-state capacity. In conclusion, optical helicity can be encoded into, retained in, and read from electrically programmed analog conductance states, providing a persistent state variable for neuromorphic computation.