Journal of Materiomics (Jul 2023)

Ferroelectric composite-based piezoelectric energy harvester for self-powered detection of obstructive sleep

  • Swati Panda,
  • Hyoju Shin,
  • Sugato Hajra,
  • Yumi Oh,
  • Wonjeong Oh,
  • Jeonghyeon Lee,
  • P.M. Rajaitha,
  • Basanta Kumar Panigrahi,
  • Jyoti Shukla,
  • Alok Kumar Sahu,
  • Perumal Alagarsamy,
  • Hoe Joon Kim

Journal volume & issue
Vol. 9, no. 4
pp. 609 – 617

Abstract

Read online

Lead-free piezoelectric ceramic is a promising material for energy harvesters, as they have superior electromechanical, ferroelectric, and piezoelectric properties. In addition, piezoelectric ceramics can be blended with polymer to achieve high-flexibility polymer-ceramic composites, providing mechanical robustness and stability. In this context, a new lead-free ferroelectric material, having the chemical formula SrTi2O5 (STO), was synthesized using a high-temperature solid-state reaction. Detailed analyses of the structural, morphological, and electrical properties of the synthesized material were performed. STO crystallizes with orthorhombic symmetry and space group of Cmm2. The frequency and temperature-dependent dielectric parameters were evaluated, and impedance spectroscopy shed light on the charge dynamics. The PDMS-STO composites at different mass fraction of the STO were prepared using a solvent casting route, and a corresponding piezoelectric nanogenerator (PENG) was developed. The electrical output of the different PENG by varying massfractions of STO in PDMS and varying force were investigated. The 15% (in mass) PENG device delivered the highest peak-to-peak voltage, current, and power density of 25 V, 92 nA, and 0.64 μW @ 500 MΩ, respectively. The biomechanical energy harvesting using the PENG device by daily human motions, bending of the device, and attaching the device to laboratory equipment was demonstrated. Later the PENG device was attached to the human throat region, and snoring signals were recorded. A classification model was designed employing the convolutional neural network (CNN) model. Efforts have been laid to differentiate between normal and abnormal snores, which could help the patient with screening and early disease detection, contributing to self-powered healthcare applications.

Keywords