Abstract
It is of great significance to prepare a temperature and pressure (T-P) dual-mode electronic skin (DMES) with a controllable porous structure and use it to achieve material cognition. In this work, a quasi-periodic porous structure-based T-P DMES was proposed, exhibiting excellent performance in material cognition. The quasi-periodic porous structure of the electronic skin was prepared through the interaction of confined two-dimensional bubbles and polydimethylsiloxane (PDMS). After attaching graphene, PEDOT: PSS, and Bi(2)Te(3) to the porous PDMS, the obtained electronic skin can precisely detect and distinguish T-P stimuli without crosstalk. Benefiting from the quasi-periodic porous structure, the temperature and pressure sensing performance of the electronic skin can be precisely constructed and optimized by changing the size of the porous structure, which is impossible for electronic skin with a random porous structure. By analyzing the thermoelectric and piezoresistive signals of the electronic skin and combining them with a convolutional neural network, the electronic skin can identify 33 different materials, including materials with similar softness or thermal conductivity, cotton fabrics with different textures, and even diverse alloys, with an accuracy of 97.64%. The proposed T-P DMES significantly outperforms both existing electronic skins and human skin in terms of material cognition.