Development of semiconductor metal oxide-based chemical sensors for pattern recognition of biomarkers in exhaled breath

Abstract

Breath analysis is getting intensive attention due to non-invasive and simple diagnostic method. A number of diseases such as halitosis, diabetes, and lung cancer can be diagnosed by detection of sub-ppm level of biomarkers in exhaled breath. As an emerging type of breath analyzing material, semiconductor metal oxides (SMOs) have been intensively studied. The advantages of SMO-based sensing layers are provided including high potential for miniaturization by integration with portable devices, cost effectiveness, and possible for mass production. As a sensing performance point of view, SMO sensors exhibit fast responding speed, which can realize real-time diagnosis by on-line breath analysis. Even though the SMO-based sensors are promising for breath analyzer, several issues should be addressed for clinical application. The major issues of the SMO sensors for diagnostic application are sensitivity and selectivity considering that SMO sensors react with a number of analyte species without selectivity. In addition, the concentration of biomarker molecules in the exhaled breath normally in the range of sub-ppm (part per million) level, which is challenging for SMO sensors to detect with high sensitivity. The advances in nanostructure synthetic routes can address these issues because the analyte molecules mainly react on the surface of SMO layers. In this thesis, diverse SMO sensing composites were proposed using electrospinning approach, which is facile and versatile technique to obtain one-dimensional (1D) SMO nanostructure with large surface area as well as high porosity. In addition, noble metallic nanoparticles and graphene-based catalytic materials were functionalized on the SMO sensing layers to improve sensitive and selective properties. Therefore, a number of sensing composites were prepared to establish sensor libraries. These sensor libraries were investigated to understand the sensing performances for breath pattern recognition using principle component analysis. In addition, clinical demonstration was implemented to discriminate between healthy person and patient with diabetes using the proposed sensing materials. The result revealed that the breath analysis can clearly discrim-inate healthy person and diabetes patient. Moreover, breath sensing module that can be integrated with mobile devices was developed for application in portable and real-time diagnosis. This thesis demonstrated high potential of SMO-based sensing layer functionalized with diverse catalysts for application in breath analysis.