Mohammadmahdi Kiaee
Sensors capable of detecting and classifying volatile organic compounds (VOC) have been gaining more attention by the advent of internet-of-things (IoT) enabled devices and integration of various sensing elements into hand-held and portable devices. The recent applications of VOC sensors call for detectors that operate with low power consumption as well as fabrication techniques that allow integration of sensing elements with devices. Hence, in this dissertation, the application of chemiresistive sensing elements based on polymer nanocomposites, which operate at near room temperature, for low-power detection of VOCs was investigated. As a versatile fabrication method, drop-on-demand inkjet printing (DoD IJP) was used to transfer the sensing material onto the sensor platform.
A group of commercially available polymers with representative solubility parameters were selected to detect a wide range of chemically diverse analytes. A highly conductive high-structure carbon black (CB) was used as the conductive filler. The inkjet inks containing different polymers and CB were then formulated via solution mixing.
Due to various constraints on the physical properties of inkjet inks, a systematic process for ink formulation had to be established, especially when formulating inks that contained multiple components. For this aim, a systematic approach consisting of successive characterization steps, including assessment of rheology, evaporation profile, and CB particle size distribution was proposed. At the end of this step, inkjet inks containing different polymer composites were formulated, allowing stable droplet jetting and deposition of uniform sensory films.
The effect of ink formulation on the sensing performance of the printed sensors was investigated. It was demonstrated that by changing CB concentration in the composite, the sensor sensitivity varied. Lower CB loading resulted in better sensitivity (55% higher sensitivity to pentane for 3.3 vol% compared to 8.3 vol% CB loading) but inferior baseline noise (45% increase in baseline noise for 3.3 vol% compared to 8.3 vol% CB loading). By optimizing the CB concentration, high sensitivity with an improved limit of detection (LoD) was obtained as demonstrated by comparing the LoD of composites containing 3.3, 5.5, and 8.3 vol% CB to pentane, which was approximately 80, 20, and 90~ppm, respectively.
The concentration dependence of the sensor response upon exposure to chemically diverse analytes, namely water, ethanol, acetone, pentane, and heptane, was studied. It was shown that the sensor response to each analyte was in line with the analyte-polymer affinity expected from their respective solubility parameters. In the studied concentration range, all sensors showed a linear response to the concentration change of those analytes to which they were sensitive, allowing to extract the sensor sensitivity and expected LoD. The sensor LoD values appeared to be in the range of few to tens of ppm, depending on the analyte and the printing parameters. By performing the principal component analysis (PCA), it was demonstrated that using three principal components would allow the classification of the tested analytes.
Overall, the methods, materials, and results of this work demonstrated the potential of DoD IJP for the fabrication of versatile and sensitive low-temperature VOC detectors, paving the way to fully printed low-cost and low-power electronic nose devices.
EPFL2021
