Publication date: 22nd July 2026
Copper coordination compounds offer a uniquely tuneable platform for autonomous indoor energy systems: their redox potentials, crystallographic dimensionality, and processing routes can all be programmed at the molecular level [1]. This talk traces the complete pathway from photon absorption to on-device neural network inference through four interconnected advances in copper coordination chemistry.
We first introduce morphogenic coordination polymers in which counter-cation selection programs the crystallographic dimensionality of the semiconducting backbone. In mixed-valence Cu(II)/Cu(I) dithiocarbamate–halide systems, switching the bridging halide from bromide to iodide transforms Cu–I valence-band conjugation: the DFT-computed effective hole mass drops from 94me to 6me, conductivity rises over two orders of magnitude to 1 mS cm−1, and Arrhenius pre-exponential factors exceed those of Spiro-MeOTAD by four orders of magnitude, consistent with well-conjugated intra-chain transport limited by inter-chain hopping [2]. Deployed as dopant-free hole conductors in carbon-electrode perovskite solar cells, these polymers reach 13.8% power conversion efficiency with >94% retention over 50 days under ambient humidity.
The same coordination chemistry, in its discrete Cu(tmby)2 redox shuttle form, enables dye-sensitized solar cells (DSCs) exceeding 30% PCE under 1000 lux. Flash infrared annealing now extends these devices to ultralight 12.5 μm polyimide substrates, delivering 5.10% PCE at AM 1.5G and 255 mW g−1 specific power [3]. Monolithic integration of Cu(tmby)2-based DSCs with polyviologen/carbon asymmetric supercapacitors on a shared PEDOT electrode then closes the energy chain: three-terminal photocapacitors deliver 920 mV, 18% overall charging efficiency, and sustain battery-free edge AI for 72 hours, reaching 93% CIFAR-10 accuracy at 0.81 mJ per inference [4].
These results establish that a single family of earth-abundant coordination compounds can carry information from photon to logic gate without a battery, redefining the design space for integrated ambient energy systems.
M.F. acknowledges the Royal Society for a University Research Fellowship (URF\R1\191286; URF\R\241031) and Research Grants (RGS\R1\211321, IES\R3\213090, RF\ERE\210025). This work was supported by the Engineering and Physical Sciences Research Council (EP/V035819/1, EP/W006340/1, EP/S030638/1, EP/V048333/1, EP/W017091/1), the Biotechnology and Biological Sciences Research Council (BB/X00502X/1), and the European Union Horizon 2020 Marie Skłodowska-Curie Individual Fellowship programme (Grant Agreement 101028536). M.F. further acknowledges support through the EPSRC National Edge AI Hub.
