Abstract:Selectivity is critical for accurately analyzing the biomarkers of human breath, which consist of more than 200 compounds. Herein, we developed a flexible multivariable sensor based on Mxene and laser-induced graphene for detections of VOCs in exhaled breath. Based on the machine learning methods, we demonstrate accurate identifications for different types of VOCs and mixtures using the proposed sensor. Then we demonstrate an accuracy of 89.1% for the prediction of ethanol concentrations in the presence of different concentrations of water and methanol. Furthermore, using breath samples collected from volunteers before and after drinking, blind analysis validated the capacity of the reported sensor to identify whether alcohol with 88.9% accuracy. The high level of identification and concentration prediction shows the potential of the proposed multivariable sensor for detection of biomarker VOCs in human exhaled breath and early diagnosis of the disease.

Flexible Multivariable Sensor Based on Mxene and Laser-Induced Graphene for Detections of Volatile Organic Compounds in Exhaled Breath