The exponential growth of Internet of Things (IoT) devices presents a critical challenge in sustainable energy provision. High-efficiency ambient photovoltaics integrated with artificial intelligence offer a promising solution for self-powered, smart IoT systems. Dye-sensitized solar cells (DSCs) optimized for indoor light harvesting have been developed using novel copper coordination complexes as redox mediators. These mediators regenerate dyes with an overpotential of only 0.1 eV. By employing co-sensitization strategies with dyes such as XY1 and L1, power conversion efficiencies of up to 38% have been achieved under 1000 lux fluorescent light, with open-circuit voltages exceeding 1.0 V.(1) A key innovation is the ”zombie” solar cell concept, where copper complexes form a solid-state hole transport material upon electrolyte evaporation. These devices maintain high efficiency and stability after drying, due to the formation of an amorphous Cu(II/I)(tmby)2 hole transport layer. Raman spectroscopy and impedance analysis have revealed the unique charge transport mechanisms in these systems. Further advancements have been made with the development of dynamic dimer copper coordination redox shuttles. These complexes transition between Cu(I) dimers and Cu(II) monomers during the redox process, enabling a two-electron transfer mechanism that enhances charge transport while minimizing recombination. In addition to traditional molecular complexes, one-dimensional copper coordination polymers have been introduced as hole transport materials. These materials exhibit band-like charge transport with modeled effective hole masses as low as 6me, providing a sustainable alternative to heavily doped organic semiconductors.

These material advancements have been translated into practical devices, integrating DSCs with microcontrollers to create self-powered IoT nodes. On-device machine learning, including convolutional neural networks for image recognition tasks, has been implemented. Long short-term memory (LSTM) networks manage energy use dynamically, adapting computational loads to available light, thus ensuring optimal performance in varying ambient conditions.

This interdisciplinary approach, which spans from molecular design to device engineering and artificial intelligence, demonstrates the critical synergy between materials chemistry and advanced computing in addressing global energy challenges. These findings not only advance ambient photovoltaics but also open new avenues for sustainable, intelligent technologies capable of operating autonomously in low-light environments.