Development and characterisation of metal oxide gas sensors
There is an increasing interest among researchers of environmental sensors to improve the functionality and portability of chemical sensor devices, while applying simple materials and innovative techniques. Nanotechnology, which is at the cross-road of science, technology and arts, has provided the platform for this multidisciplinary technological development. The research described in this thesis involves the design, fabrication and characterisation of chemical sensors for the analysis of trace volatile hydrocarbons. The objective of this study is the systematic investigation of the sensing-dependence of the composition of veritable materials used in the preparation of nanocomposites, characterisation of their nanostructures, and development of their sensing mechanism based on their surface-gas interaction behaviour.
This research involved the preparation of five chemical sensors using 100:0, 75:25 50:50, 25:75 and 0:100 molar ratios of tin dioxide and zinc oxide. These sensors were labelled as SnO2, 𝕊3ℤ1, 𝕊ℤ, 𝕊1ℤ3 and ZnO sensors, respectively. The samples were prepared using the radio frequency (RF) magnetron sputtering under the same conditions. A similar set of samples were annealed. Both as-fabricated and annealed samples were characterised using field emission - scanning electron microscope (FE-SEM), energy dispersive X-ray spectroscopy (EDS) and atomic force microscope (AFM). Nanostructural analysis revealed the nanocrystalline images to have minor hillocks on a relatively dense film surface. The unannealed samples exhibited more rounded protrusions than the annealed samples. The grain heights of the as-fabricated samples were higher than the annealed samples, while there was reduction in surface roughness as a result of annealing. The grain size was observed to increase from pure SnO2 and ZnO samples toward the 𝕊ℤ samples. Also, the 𝕊ℤ samples were observed to reflect the lowest surface roughness parameters, while the 𝕊1ℤ3 samples showed the highest surface roughness values.
The sensor signals, usually quantised in raw form, were smoothed using the Savitzky-Golay filter, before characterisation of the sensitivity of the sensors. Experimental investigation proved that gas sensitivity increased with increasing gas concentration and increasing temperature for all sensors. The best sensitivities were displayed by 𝕊1ℤ3, followed by 𝕊3ℤ1 sensor devices, while ZnO was more sensitive than SnO2. This behaviour was attributed to the high photocatalytic activity of pure ZnO and coupled SnO2–ZnO nanocomposites than pure SnO2.
This fact was collaborated with the results of the thermodynamic analysis of each sensor. For both methanol and ethanol, the activation energy of SnO2 was higher than that of ZnO, while the activation energy of 𝕊¬1ℤ3 sensor devices was the lowest. It was observed that ethanol was more sensitive than methanol, indicating a possibility for good selectivity of the sensors. Statistical analysis confirmed that sensor type, gas concentration and temperature influenced respective sensor sensitivity; but the effects were varied depending on the sensing conditions and sensor types.
With the development and simulation of modified chemisorption and linear models, excellent sensitivity behaviours were observed at extended concentration range. These results collaborated the facts that 𝕊1ℤ3 and 𝕊3ℤ1 sensors were the best sensors. At higher concentrations however, the 𝕊ℤ sensors were observed to improve in sensitivity. With these behaviours, different sensing mechanisms were proposed for each chemical sensor. It is proposed that this result is a very significant contribution to the state-of-knowledge in the domain of scientific endeavour.