Diffuse Light to Structured information with Hybrid Photovoltaics
Dr Marina Freitag
- Newcastle University
Friday 25 November, 11.30am
Abstract
By 2025 about 75 billion IoT devices will be installed, of which the majority will reside in ambient conditions. It is therefore crucial to find an energy source that yields high efficiencies in this environment. At high efficiencies under ambient light, while being more environmentally friendly, sustainable to produce and to recycle. Dye-sensitized solar cells (DSCs) are known for efficient conversion of ambient light. Fast charge separation in a variety of organic dyes and tuneable energy levels in CuII/I redox systems combined with negligible recombination processes allow DSCs to maintain a high photovoltage under ambient light.1
We tailored dye-sensitized photovoltaic cells based on a copper (II/I) coordination complexes hole transport material for power generation under ambient lighting with an unprecedented conversion efficiency of PCE 38 %, at 1000 lux from a fluorescent lamp using a novel co-sensitization strategy2,3 and electrolyte modifications. Under 1000 lux lighting, 64 cm2 photovoltaic area gives 152 J or 4.41 1020 photons sufficient energy for training and testing of an artificial neural network in less than 24 hours. The implementation of a long short-term memory (LSTM) based energy management on-device predicts changing deployment environments and adapts its computational load accordingly to perpetually operate the energy-harvesting circuit and avoid power losses or brownouts. Ambient light harvesters enable a new generation of self-powered and self-aware IoT device to be powered by a previously untapped energy source.4,5
References