| Volatile gas sensors play an important role in ensuring human health and safe production in related industries.With the continuous improvement of detection standards,the performance of gas sensors has higher requirements.Especially in complex gas mixtures where multiple gases coexist,accurate detection of various gases still faces great challenges.The design and synthesis of high-performance gas sensing materials and the use of appropriate data analysis methods are the key to solving these problems.In this paper,tin-based and indium-based oxide semiconductor materials commonly used in gas sensing are selected as the research objects.A series of ordered mesoporous tin-based and indium-based oxide gas sensing materials with different mesoscopic structures and doping compositions are synthesized by nanocasting hard template method combined with in-situ doping.The effects of mesoscopic structure and doping composition on gas sensing properties were systematically studied.A program was written to establish a method for quantitatively identifying the concentration of acetone and ethanol mixed gas using BP neural network,including the following research content:1.The ordered mesoporous tin-based and indium-based oxide materials were synthesized using the nano-casting method using silica KIT-6 hard template.The effects of filling ratio and calcination temperature on the mesoscopic structure and sensing performance were systematically investigated,and the optimal filling ratio of 30%and the optimal calcination temperature of 300℃ for ordered mesoporous SnO2 and 400℃ for ordered mesoporous In2O3 were determined.Under the optimal conditions,the response values of SnO2-0.3-300(0.3 is the filling ratio and 300 is the calcination temperature)to 100 ppm acetone and ethanol were 12 and 38;and the response values of In2O3-0.8-400(0.8 is the filling ratio and 400 is the calcination temperature)to 100 ppm acetone and ethanol gas were 34.7 and 63.2.The compositions of tin-based oxide materials were tuned by in situ doping with rare-earth elements Er and Ho,and the effects of doping composition on gas-sensitive performance were systematically investigated.The results showed that rare-earth metal doping significantly improved the gas-sensitive performance of ordered mesoporous tin-based oxide materials.Compared with the undoped materials,the response values of the best doping ratio materials to acetone and ethanol were improved by 5.6 and 2.1 times.3.In order to cope with the problem that multiple gases coexist with each other and are difficult to be accurately identified quantitatively,we wrote our own program to establish a method for quantitative identification of acetone and ethanol gas mixtures using BP neural network,and designed a GUI interface to easily and intuitively display the identification results and error analysis,the average error of the model is less than 10%,which can be more accurate for quantitative identification of two-component gas mixtures. |