Zhang Ling et al. [1] successfully prepared cadmium sulfide polyaniline (CDS PANI) nanorods with hexagonal wurtzite structure by a simple hydrothermal method, and the sensor electrode structure is shown in Fig. 1(a). A schematic diagram of the formation of coordination bonds between CD2+ and PANI amino groups is shown in Figure 1(b). As shown in Figure 1(c), CDS PANI consists of sea urchin-like nanorods with a length of 200 nm and a diameter of 50 nm. CDS Pani thin film sensors have certain response characteristics to low concentrations of formaldehyde, and the CDS Pani sensor has the highest response to formaldehyde gas at 120, as shown in Figure 1 (D) above. Figure 1 (e) shows the response recovery time of the CDS Pani formaldehyde sensor, and the response time of the CDS Pani sensor to 10ppm formaldehyde gas is 25s 30s. As can be seen from Figure 1 (F), the response of the CDS Pani sensor to formaldehyde gas is attributed to the chemical adsorption of oxygen on the surface of the CDS.
Fig.1 CDS Pani gas sensor: (a) Schematic diagram of the gas sensor; (b) Schematic diagram of the formation of coordination bonds between CD2+ and Pani amino groups; (c) surface morphology of three different materials; (d)-(e) Graph of the effect of temperature on the sensor and the recovery time of the sensor response; (h) Schematic diagram of PANI and CDS PANI sensitivity.
Yang Li et al. [2] used an in-situ growth method to prepare vertically arranged SNO2 nanosheets on a substrate, and a tin dioxide polyaniline (SNO2 PANI) composite film sensor was prepared by electrospinning (as shown in Figure 2(a)), and the prepared sensitive film had good contact with the electrode. 05~10.7 ppm ammonia was tested, as shown in Figure 2 (b) and (c), and the gas sensing device had a positive effect on 107ppm ammonia responds to 3700% and the detection limit is 46ppb. As can be seen from Figure 2(c), the SNO2 PANI sensor responds significantly higher to the target gas ammonia than other gases, indicating that the tinoxide-doped polyaniline has excellent selectivity for ammonia. Fig. 2 (d) shows that the response of the Humidity SNO2 Pani sensor has a certain effect.
Fig.2 SNO2 PANI gas sensor: (a) Schematic diagram of the preparation of gas sensor by electrospinning; (b) Sensor response to ammonia at room temperature; (c) Comparison of selectivity tests of sensors for different target gases; (d) Plot of the SNO2 PANI sensor response to humidity.
Krishanu Chatterjee et al. [3] investigated the effects of preparation process changes on the structure, conductivity, and gas sensitivity of polyaniline. Polyaniline doped with 5-sulfosalicylic acid was prepared by four different methods (polymerization, vertical lifting, spin coating, and electrodeposition). Fig. 3 (A) shows the polyaniline transmission SEM (as shown in Fig. 3 (b)) images prepared by four different methods, and it can be seen that the polyaniline nanostructure is determined by the preparation method. The response of polyaniline doped with 5-sulfosalicylic acid to ammonia was different in the above four schemes, and the polyaniline prepared by polymerization method had the highest gas sensing performance. From Fig. 3(c), it is found that the improvement of the sensor's gas sensing performance is due to the large specific surface area of polyaniline prepared based on the in-situ polymerization method.
Fig.3 SSA-doped PANI gas sensor: (A) Transmission electron microscopy of SSA-doped PANI nanostructures; (b) Test charts of sensor response to ammonia gas prepared by four different methods at room temperature; (c) Change in SSA-doped PANI surface area to volume ratio.
In summary, the modification of polyaniline with different nanomaterials can improve the defect of low gas sensitivity of intrinsic polyaniline sensor. Similar to polyaniline, polypyrrole, as one of the commonly used conductive polymer materials, can also be modified by nano-modification to significantly improve the gas sensing performance of the sensor, which provides a new way for the detection of exhaled gas in the human body and becomes an important part of early disease screening and diagnosis. The working principle diagram of the system for early screening and diagnosis of human disease is shown in Figure 4, which comprises the construction of gas sensing array, human disease simulation and multi-dimensional response data acquisition, on this basis, the multi-dimensional data is preprocessed, the intelligent algorithm model is used to learn the training sample data, and then the human body state is diagnosed, so that the early screening and diagnosis of the human disease state is realized.
Fig.4. Working principle diagram of a conductive polymer gas sensing array system.
The exhaled air of the human body is complex and diverse, and it is difficult for a single sensor device to realize the systematic detection of multiple variables, therefore, in order to meet the test of complex gases, the gas sensing array detection technology based on multiple gas sensors is produced. Deng Fanfei et al. [4] used dip coating to prepare four QCM gas sensors, and formed a gas sensor array, so as to achieve non-destructive shelf life of egg samples. Huang Xingyi et al. [5] screened high-performance sensors to form a sensing array according to the different characteristic gases volatilized by bream samples during the continuous decay process, and continuously detected different samples to evaluate the freshness of bream samples. Peng Jing et al. [6] screened out a sensor array composed of nine metal oxide sensors, which realized the qualitative detection of multi-component gases such as acetone, ethanol, and formaldehyde.
For the multivariate and complex nonlinear system composed of human exhaled air, researchers at home and abroad often need to extract and classify data features with the help of experience and appropriate network models. The development of artificial intelligence algorithms, such as artificial neural networks, clustering algorithms, support vector machines, gray systems, fuzzy theory, principal component analysis algorithms, expert systems, etc., has played a huge role in promoting the development of intelligent detection technology. Tang Ju et al. [7] combined the principal component analysis algorithm (PCA) with the radial basis function neural network to construct an output data processing model, so as to achieve accurate identification of sulfur hexafluoride decomposition component gases. The combination of intelligent detection technology and artificial intelligence algorithm has realized the qualitative and quantitative identification of complex systems, and its great value has attracted the attention of researchers at home and abroad. Zhang Haijuan et al. [8] constructed a conductive gas sensing array, which was combined with a principal component analysis algorithm to realize the quantitative identification of six gases, including toluene and ethanol. l.a.Daniel et al. [9] designed a six-input quartz crystal microbalance gas-sensing array, and used an artificial neural network algorithm to process the array signals, so as to realize the quantitative identification of a variety of volatile gases.
Although the intelligent detection technology based on sensor array has made great progress in the detection and identification of multi-component gases, the current research at home and abroad is still in the laboratory stage, especially the detection of exhaled gas from the human body is still in its infancy. The screening of sensing components and the construction of intelligent algorithm models are important topics to achieve real-time detection of exhaled air**. At present, gas sensors are mainly composed of metal oxide sensors, which are not suitable for the normal temperature detection of exhaled gas due to the high working temperature and poor stability of metal oxides, which seriously restricts their development. Most of the intelligent algorithm models are selected as single BP neural networks, clustering algorithms, support vector machines, etc., which have low accuracy, insufficient algorithm learning ability, slow convergence speed, and are easy to fall into local optimal solutions. Therefore, in order to solve the above problems, a new high-performance gas sensor is studied to optimize the sensor array, and the array data is optimized with a new intelligent algorithm to solve the common cross-sensitivity problem of the sensor array, so as to improve the performance of the sensor array intelligent detection system.
References: 1] Zhang L, li x., mu. z, et al. a novel composite of cds nanorods growing on a polyaniline-cd2+ particles surface with excellent formaldehyde gas sensing properties at low temperature[j]. rsc adv., 2018, 8(54): 30747-30754.
2] li. y, ban. h, jiao. m, et al. in situ growth of sno2 nanosheets on a substrate via hydrothermal synthesis assisted by electrospinning and the gas sensing properties of sno2/polyaniline nanocomposites[j]. rsc adv., 2016, 6(78): 74944-74956.
3] krishanu c., palash d., saibal g., morphology dependent ammonia sensing with 5-sulfosalicylic acid doped nanostructured polyaniline synthesized by several routes[j]. sens. actuators b: chem., 2013, 181: 544-550.
4] Deng Fanfei, Wang Jun, Chen Wei, Wei Zhenbo. Preparation of QCM gas sensor array and its detection of egg shelf life[J].Transactions of the CSAE, 2017, 33(20): 292-299
5] Huang Xingyi, Gu Feifei. Screening and Array Construction of Gas Sensor for Bream Freshness[J].Science and Technology of Food Industry, 2013, 19(34): 295-312
6] Peng Jing. Research on the optimization and application of gas sensor array[D].Huazhong University of Science and Technology, 2008
7] Tang Ju, Fan Min, Tan Zhihong, et al. Cross-response processing technology of photoacoustic detection signal of SF6 partial discharge decomposition component[J].High Voltage Engineering, 2013, 2(39): 257-264
8] Zhang Haijuan. VOCs sensor and array technology based on ionic liquids and nanoporous materials[D].Zhejiang University, 2013
9] fernandes d., gomes m. development of an electronic nose to identify and quantify volatile hazardous compounds[j]. talanta, 2008, 77(1): 77-83.