Inkjet‐Printed rGO/binary Metal Oxide Sensor for Predictive Gas Sensing in a Mixed Environment. (20th March 2022)

• Record Type: Journal Article
• Authors: Ogbeide, Osarenkhoe; Bae, Garam; Yu, Wenbei; Morrin, Ewan; Song, Yungyu; Song, Wooseok; Li, Yu; Su, Bao‐Lian; An, Ki‐Seok; Hasan, Tawfique
• Is Part Of: Advanced functional materials. Volume 32:Number 25(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-03-20
• Subjects: gas prediction -- gas sensors -- graphene -- inkjet printing -- machine learning -- metal oxide -- principal component analysis; Materials -- Periodicals; Chemical vapor deposition -- Periodicals; 620.11
• DOI: 10.1002/adfm.202113348
• Languages: English
• ISSNs: 1616-301X

Abstract

Abstract: Selectivity for specific analytes and high‐temperature operation are key challenges for chemiresistive‐type gas sensors. Complementary hybrid materials, such as reduced graphene oxide (rGO) decorated with metal oxides enables realization of room‐temperature sensors with enhanced sensitivity. However, sensor training to identify target gases and accurate concentration measurement from gas mixtures still remain very challenging. This work proposes hybridization of rGO with CuCoO x binary metal oxide as a sensing material. Highly stable, room‐temperature NO2 sensors with a 50 ppb of detection limit is demonstrated using inkjet printing. A framework is then developed for machine‐intelligent recognition with good visibility to identify specific gases and predict concentration under an interfering atmosphere from a single sensor. Using ten unique parameters extracted from the sensor response, the machine learning‐based classifier provides a decision boundary with 98.1% accuracy, and is able to correctly predict previously unseen NO2 and humidity concentrations in an interfering environment. This approach enables implementation of an intelligent platform for printable, room‐temperature gas sensors in a mixed environment irrespective of ambient humidity. Abstract : A fully inkjet‐printed rGO/binary metal oxide sensor is demonstrated for predictive gas sensing in mixed environments. Using a combination of principal component analysis and machine learning algorithms, the classifier allows a distinguishable decision boundary map for five different gas classes with 98.1% accuracy at ppb concentrations. Furthermore, a projection of regression models onto the classifier plane allows successful prediction of unseen gas concentrations.