Humidity sensor based on PEDOT:PSS/CeO2 composite material for respiratory monitoring of anesthetized patients

A novel PEDOT:PSS/CeO 2 nanocomposite with intimate p-n heterojunction interface was successfully fabricated via hydrothermal and in-situ polymerization methods. • The sensor exhibits outstanding humidity sensing performance: high response (209%), fast response/recovery (5.5 s/16.1 s), wide detection range (0–97% RH). • DFT calculations reveal that the PEDOT:PSS/CeO 2 composite possesses the strongest adsorption energy for H 2 O molecules and significant interfacial charge transfer. • The sensor was integrated into a wearable device and combined with a CNN algorithm, achieving 100% accuracy in real-time recognition of six simulated respiratory patterns. This study prepared a novel poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate)/cerium dioxide (PEDOT:PSS/CeO 2 ) nanocomposite via a hydrothermal method combined with in-situ chemical oxidative polymerization, and applied it in a high-performance resistive humidity sensor. Systematic material characterization confirmed the formation of an intimate heterointerface between CeO 2 and the PEDOT:PSS polymer network within the composite, accompanied by significant interfacial charge interaction and PEDOT chain conformational changes. Benefiting from the abundant oxygen vacancy active sites introduced by CeO 2 and the synergistic p-n heterojunction effect formed with PEDOT:PSS, the fabricated sensor demonstrated outstanding comprehensive performance across a wide relative humidity (RH) range of 0–97%: a high response (209%), excellent linearity (R 2  = 0.9626), rapid response/recovery times (5.5 s/16.1 s), and good long-term stability. Density Functional Theory (DFT) calculations further revealed, at the atomic scale, that the composite possesses the strongest adsorption energy for water molecules and exhibits significant charge transfer, providing theoretical support for its enhanced humidity sensing mechanism. As a proof of concept, the sensor was integrated into a wearable device. Combined with a Convolutional Neural Network (CNN) algorithm, it successfully achieved real-time recognition of six simulated breathing patterns, including apnea, with 99.17% accuracy, demonstrating its great potential for monitoring the respiratory status of anesthetized patients. Download: Download high-res image (86KB) Download: Download full-size image