Highly stretchable carbon-based triboelectric nanogenerator textiles woven using coaxial fibres via direct ink writing

In recent years, triboelectric nanogenerators have seen great progress in the Internet of Things, wearable electronics, virtual reality and artificial intelligence technologies. However, current triboelectric nanogenerator devices generally face challenges of complex preparation processes and structures, hindering their application in large-area wearable textiles. In this study, we develop highly stretchable triboelectric fibres using direct ink writing as building blocks for further woven textiles. This consists of a conductive core and insulating sheath, namely composites of multi-wall carbon nanotube and fluorinated ethylene propylene particles, both doped in silicone rubber. The fibre enjoys outstanding electrical performance, as well as excellent mechanical stretchability. Particularly, a single-centimetre-long fibre, upon slight patting, can surprisingly produce an open-circuit voltage of 8 V, a short-circuit current of 110 nA and 2.5 nC short-circuit transferred charges, or equivalent lighting of four to five commercial light-emitting diodes alone, greatly outperforming current wearable devices. These fibres also can endure an extraordinary strain of up to ∼874 %. The fibres are then woven into wearable gloves and socks using traditional weaving techniques, which can realise motion pattern identification and recognition with classification accuracy as high as ∼96 % using convolutional neural networks. This creates application prospects in wearable electronics, providing a scalable and affordable method for weaving daily wearable textiles and favouring large-area, wearable, self-powered sensors and artificial intelligence technologies. • Coaxial TENG fibres fabricated through direct ink writing provide facile and scalable production of smart wearable textiles. • A single-centimetre-long fibre, can produce an open-circuit voltage of 8 V, lighting of four to five light-emitting diodes. • Fibres can withstand an extraordinary strain of ∼874 % and recognise movement by machine learning when woven into wearables.