Mimicking visual memory in stimuli-responsive material

Recent advances in materials science and unconventional computing have enhanced a growing interest in the development of "intelligent matter", materials that integrate sens- ing, memory, and adaptive response within their physical structure. These systems aim to implement computational functions directly at the matter level, enabling distributed, low-power information processing without relying on external control or digital architectures. Within this emerging paradigm, light-responsive materials, especially azobenzene-based polymers, have shown great promise due to their intrinsic light-triggered molecular dynamics. This thesis investigates the azopolymer PAZO as a candidate for implementing elementary visual functionalities directly in matter. Through optical stimulation, PAZO films exhibit birefringence with tunable dynamics, governed by molecular reorientation and relaxation processes. Exploiting this behavior, two biologically inspired tasks are explored: luminance adaptation and color recognition. First, the influence of film thickness on birefringence response is studied to emulate pho- topic and scotopic adaptation mechanisms, analogous to cone and rod cell dynamics. Results show that thickness strongly modulates the amplitude and temporal persistence of the birefringent response, offering a physical basis for light-level adaptation. Second, the material’s relaxation dynamics are studied under stimulation with distinct wavelengths to evaluate its spectral sensitivity. A temporal signature associated with each wavelength suggests the potential for color dis- crimination based purely on birefringence decay profiles. These results set PAZO as a compelling platform for light-driven adaptive behavior, offer- ing a combination of structural tunability, analog and reconfigurable optical responses, and localized spatial addressability enabled by microstructuring techniques.