M. Schweizer-Berberich

Gas identification by modulating temperatures of SnO2-based thick film sensors

Abstract A new method is presented to identify the presence of two gases in the ambient atmosphere. The method employs only one SnO2-based gas sensor in a sinusoidal temperature mode to perform the quantitative analysis of a binary gas mixture (CO/NO2) in air.

Grain size control in nanocrystalline In2O3 semiconductor gas sensors

Abstract In2O3 thin films prepared by sol–gel method make it possible to detect low levels (several hundreds ppb) of nitrogen dioxide in air. The possibility of grain size control in indium oxide-sensing layers has been established by using of two preparation methods—electron beam evaporation (EB) and sol–gel technique (SG). SG-prepared samples show smaller particles sizes (down to 5 nm), higher state of agglomeration, higher sensor resistance in air and higher response to NO2 in comparison to EB samples. Sol–gel technique leads to the preparation of polycrystalline indium oxide with particle sizes of about 5–6 nm after calcination at 400°C and 20 nm after calcination at 700°C. The initial state of particle agglomeration in initial indium hydroxide sol (IHS), stabilized with nitric acid, influences the structure and surface morphology of the resulting indium oxide. While the In2O3 layer prepared by using low agglomerated IHS is smooth and porous, In2O3 layers prepared from highly agglomerated IHS consist of two regions—thin layer and crystallite agglomerates in cubic and rectangular parallelepiped form. The last shows the best results in terms of NO2 sensitivity. Sensor resistance and NO2 sensitivity increase with the decrease of the grain sizes in In2O3.

Morphological analysis of nanocrystalline SnO2 for gas sensor applications

Abstract Structural and morphological analysis of nanocrystalline SnO2 for gas sensor applications were performed at different annealing conditions by using nanopowders and thin nanocrystalline layers. The evolution of the grain size and the morphology of Pt doped tin dioxide nanoparticles with increase of annealing temperature from 450 to 1000°C were analyzed by means of transmission electron microscopy (TEM), X-ray diffraction (XRD), and Fourier transform infrared (FTIR) and micro-Raman spectroscopies. TEM shows that the average particle size increases, the size distribution becomes more spread out, and the grain faceting, as a mechanism of energy minimization, is more evident with increasing temperature. Furthermore, the shape of the particles changes with the annealing temperature, which explains the results of the FTIR spectra using the Theory of the Average Dielectric Constant (TADC). As temperature increases, the Raman spectra are modified in agreement with a reduction of the crystalline defect concentration and a grain size increase. The the tin nanocrystalline SnO2 layers, deposited on α-Al2O3 or on thermally oxidezed Si substrates, have been annealed at 700° C for 8 h under different atmospheres, such as oxygen or synthetic air. TEM proves that the annealing atmosphere has a strong influence on the size and size distribution of the nanoparticles in the thin layer. The main differences are found near the layer-substrate interface and are dependent on the annealing atmosphere as well as the nature of the substrate.

Polymer-based sensor arrays and multicomponent analysis for the detection of hazardous oragnic vapours in the environment

Abstract An array of piezoelectric quartz crystals was used to detect volatile organic compounds such as hydrocarbons, chlorinated compounds and alcohols. Steady-state frequency shifts have been used as the input parameters for multicomponent analysis. The coating materials chosen were side-chain-modified polysiloxanes. The results show clearly that these polymers provide excellent reproducibility over months. In addition, the performance of the array in the presence of humidity up to 70% r.h. does not decrease compared with dry air. In the multicomponent analysis, we compared commercially available partial least-squares regression (PLS) and artificial neural network (ANN) software. The neutral network designed for this application was small in order to avoid overfitting. For low-dimensional problems there is no difference between the two evaluation methods, but for complex ternary mixtures and long-term measurements the ANN offers advantages in predictability. Efforts were made to use a reduced set of calibration points, and here PLS presents the possibility of reducing the calibration time by 90% (use of Factorial and Box-Behnken designs) without loss of resolution, whereas the ANN suffers if a small number of training vectors is chosen.

Pulsed mode of operation and artificial neural network evaluation for improving the CO selectivity of SnO2 gas sensors

Abstract Stannic oxide sensors were developed to monitor CO concentrations (10–200 ppm). Cross sensitivities of these sensors against humidity (20–80% r.h.) and methane (400–2000 ppm) can be suppressed by evaluating “mode of operation” data. For this purpose, a sequence of switched operation temperatures introduces kinetic parameters. The latter are evaluated by model-based, or data-based algorithms. For this approach to achieve high selectivity, micromachined sensor dies are particular suitable as they show low thermal inertia.

Strategies to avoid VOC cross-sensitivity of SnO2-based CO sensors☆

Abstract CO gas sensors have to fulfil the requirements of standards like the UL2034 or BS7860. These standards require negligible interference in cross sensitivity tests of CO gas sensors. Besides their CO sensitivity, tin oxide gas sensors are also cross sensitive to volatile organic compounds like long chain hydrocarbons, alcohols or ethyl acetate. To fulfil the standards, charcoal filters are used combined with a micromachined SnO2 sensor. The latter is operated in a pulsed mode. Optimized size and material parameters of the charcoal filters protect the sensor from long chain hydrocarbons and alcohols. Small reducing molecules can be distinguished from CO by mode of operation and a suitable multi component analysis. Sensors with a combination of both methodologies fulfil the standards. The sensors are characterized by electrical measurements as required by the standards. The filters are characterized by SEM, EDX, BET and TG/DSC.

Application of neural-network systems to the dynamic response of polymer-based sensor arrays

The conventional calibration method for sensor arrays uses steady-state signals that depend on the gas concentration. This method can be time consuming if many concentrations and compositions of a multicomponent mixture are required. Good experimental design may reduce the necessary effort so that the number of calibration experiments is minimized. Dynamic measurements may significantly reduce the time of each calibration experiment. In the present approach a random walk through the domain of the gas concentrations is chosen with each step of the walk adjusted for a short time only. The sensor array consists of six polymer (polysiloxanes with functional groups)-coated bulk acoustic wave (BAW) devices. The concentration domain is defined by a binary mixture of n-octane and toluene (150 to 800 ppm). Neural networks evaluate both qualitative and quantitative information from the sensor response. In particular, the extensions of feed-forward nets towards recurrent or time-delay structures can be used to solve problems related to dynamic evaluations (e.g., no steady-state signal, parameter drift). These network architectures with different numbers of hidden neurons are applied to evaluate the data from the BAW device array. The networks are trained with back-propagation-like training algorithms and are validated with arbitrary gas mixtures.

Filters for tin dioxide CO gas sensors to pass the UL2034 standard

Abstract CO gas sensors have to fulfil the requirements of standards like the UL2034 or BS7860. Unfortunately, tin oxide gas sensors are also highly sensitive to volatile organic compounds (VOCs) like long chain hydrocarbons, alcohols or ethyl acetate. These standards require negligible interference in cross-sensitivity tests of CO sensors. To fulfil the standards, charcoal filters were used combined with a micromachined SnO 2 sensor. The latter was operated in a pulsed mode. Sensors with optimized size and material parameters of the charcoal filters are shown to fulfil the UL2034. The sensors are characterized by electrical measurements as required by these standards. The filters are characterized by SEM, EDX, BET and TG/DSC.

Evaluation Of Dynamic Sensor Signals By Artificial Neural Networks

Random walks have been used to calibrate neural nets for evaluating dynamic data (non-equilibrium signals) as they occur under realistic environmental conditions. The prediction of test data even at varied exposure times is possible for organic volatiles with low RMS errors in the range of 30 ppm. The neural nets consisted of feed-forward and recurrent nets. The results show that a careful selection of the training data is of major importance. In contrast the improvements by the recurrent nets or more sophisticated network architectures as compared with feed-forward nets were only small.

Wie menschlich sind elektronische Nasen

Das Verständnis für die komplexen Geruchsund Geschmacksempfindungen über die menschliche Nase erfordert eine Charakterisierung der über Rezeptoren ausgelösten Informationskaskade und der Datenverarbeitung im Gehirn. Einzelkomponenten dieses komplexen Zusammenspiels können für den Aufbau technischer Sensorsysteme nicht direkt übernommen oder in ihrer Funktion imitiert werden. Technische Lösungen zum Aufbau elektronischer Nasen müssen daher auf völlig anderen Funktionsprinzipien beruhen. In einem modularen Aufbau werden dabei Signale verschiedener Einzelsensoren über Mustererkennung ausgewertet und nachfolgend mit subjektiv erfaßten Geruchsempfindungen oder mit objektiven chemischen Analysen korreliert. Typische Beispiele charakterisieren die heutigen Möglichkeiten und Grenzen dieses Ansatzes.