|M.Sc Student||Gliksman Sagi|
|Subject||Sensing Temperature and Humidity by Monolayer-Capped|
Metal Nanoparticles on Flexible Substrates
|Department||Department of Chemical Engineering||Supervisor||PROF. Hossam Haick|
|Full Thesis text|
Inexpensive and low-power temperature and humidity sensors have been successfully demonstrated. Nevertheless, for many applications that involve new generations of integrated circuits and/or artificial skin (a device that mimics human skin), temperature and/or humidity sensors on flexible substrates are urgently needed. In the current thesis, I present monolayer-capped nanoparticle (MCNP) layers as potentially inexpensive and low-power (<0.5V) operating platforms for temperature and humidity applications. The MCNP-based flexible sensors exhibited excellent sensitivity to temperature (below 1 °C per 1% of change in response) and linearity in the range between 20-43 °C. Using 80% of the results as training set and 20% of the data as blind set I showed ~98% accuracy and ±0.2 °C variance in the performance of flexible sensors, slightly better than the commercially-available solid-state temperature sensors (~95% accuracy; more than ±0.3 °C variance). A modified platform of the same MCNP-based flexible sensors exhibited very good sensitivity to humidity (< 1% relative humidity per 1% of change in response) in the range of 5-94% relative humidity. Using 80% of the results as training set and 20% of the data as blind set showed >85% accuracy and ±1% relative humidity variance in the performance of the flexible sensors. These flexible humidity sensors exhibited faster response time (<2 sec) than that of the commercially-available solid-state humidity sensors (>10 sec). For these devices, a decline in the response was observed above 65% relative humidity, probably because of stiff ionic conduction mechanism that resulted from the water vapor. Based on these separate platforms, a combined prototype that could simultaneously measure temperature and humidity was successfully achieved. The effect of convex and concave bending on the results achieved via the MCNP-based flexible sensors was examined. Blind test of the responses showed that the both the concave and the convex bending modes affected the prediction accuracy of the temperatures minimally. The predictive accuracies in all examined deflection strengths exhibited better results than the commercially-available, solid-state temperature sensors. Blind test of the responses towards various levels of humidity showed no differences between the bending and non-bending modes of the flexible humidity sensor. Altogether, the MCNP-based flexible temperature and humidity sensors provide fast and easy-to-produce platforms that are well-suited for manufacturing a large amount of micro-scale sensors, with high spatial resolution and a well-defined and controllable location on the integrated circuit or artificial skin.