Pure SnO ₂ is widely used to detect VOC gases due to its high stability. Unfortunately, there is a problem of poor selectivity, especially for gases such as ketones and alcohols, which have a similar high sensitivity. In this paper, dynamic temperature modulation technology was used to solve this problem. Rectangular wave and sine wave, typical voltage jump waveform and voltage smooth waveform, were used as heating voltage waveform. It was introduced in detail that the optimization of dynamic temperature modulation parameters based on response time and power consumption. According to the different heating waveform parameters, we obtained the dynamic response data of six groups of sensors to five concentration gradients of acetone, butanone, n-propyl alcohol and isopropyl alcohol. It was used to evaluate the performance of heating waveform parameters by support vector machine (SVM) and principal component analysis (PCA) combined with K-nearest neighbor (KNN) algorithm. Both data analysis methods showed that the recognition effect was better when the SnO ₂ sensor was heated by sine wave. In order to reduce power consumption, the heating waveform was determined as 0-6 V sine wave. Then the mechanism was analyzed in dynamic temperature modulation of the sensor. The voltage smooth waveform reached the stepless response of multiple temperature points by traversing the temperature range. More characteristics can be produced by experiencing more types proportion of adsorbed oxygen species within the cycle. It provided a new research idea for semiconductor gas sensor dynamic temperature modulation technology.

Semiconductor gas sensors have been widely applied to Volatile Organic Compounds (VOCs) detection. However, the poor selectivity in the actual complex gas environment has become a bottleneck to restrict the development of semiconductor gas sensors. In order to solve this problem, SnO ₂ gas sensor is prepared and temperature modulation measurement is carried out. In this paper, the method of zigzag-rectangular wave temperature modulation is first proposed, which improved the selectivity of the sensor and significantly increased the characteristic peak. Support vector machine (SVM) is used for pattern recognition, which proved the superior of the SVM in small sample size. The results demonstrate that the combination of zigzag-rectangular wave temperature modulation and SVM pattern recognition method can effectively improve the selectivity of the gas sensor.

Alcohols detection is essential for the health of chemical and food production personnel. Metal oxide semiconductor (MOS) gas sensors are widely used, however, sensitive and selectivity of traditional MOS gas sensor is still a bottleneck issue. Recently, ABO ₃ perovskite-type MOS has attracted much attention due to its abundant and controllable physicochemical properties. In this article, macroporous perovskite-type LaFeO ₃ (LFO) microspheres were prepared by simple one step hydrothermal process. The gas sensing test results show that the LFO gas sensor is highly sensitive and selective to alcohols, especially methanol. For 100 ppm methanol, the response value can reach 120 at 215 °C. The excellent gas sensing properties are attributed to the macroporous characteristics (>100 nm), which facilitates the diffusion of methanol molecules inside the nanospheres, enhancing the contact between the gas molecules and the material and improving the gas sensing performance. In addition, the adsorption characteristics and electron exchange processes of eight gases on the material surface were investigated based on first principles analysis. The results show that the hydroxyl group is the main contributor to the stable adsorption of alcohols gas on the material surface. The perovskite-type LFO is a very promising sensitive material for alcohols detection.

In this paper, the highly sensitive n-butanol sensors with high selectivity have been successfully fabricated based on spindle shaped α-Fe $_{2}$ O $_{3}$ nanorods modified by Au nanoparticles (Au NPs). Spindle shaped α-Fe $_{2}$ O $_{3}$ nanorods were synthesized by an improved facile hydrothermal method. Uniform Au NPs were deposited on the surface of the as-synthesized α-Fe $_{2}$ O $_{3}$ nanorods by a simple impregnation method. The achieved α-Fe $_{2}$ O $_{3}$ and Au/α-Fe $_{2}$ O $_{3}$ nanorods were analyzed by means of XPS, TEM, EDS, XRD and SEM. The sensing properties of both original and Au-modified spindle shaped α-Fe $_{2}$ O $_{3}$ samples were investigated and discussed. The experimental results indicate that the sensors based on Au/α-Fe $_{2}$ O $_{3}$ nanorods exhibit better responses and selectivity to n-butanol gas than the original α-Fe $_{2}$ O $_{3}$ , and the highest response towards 100 ppm n-butanol gas was enhanced from 51 to 177 at the optimal operating temperature of 250°C.