| چکیده انگلیسی مقاله |
ABSTRACT In recent years, one of the main problems of human societies is the contamination of groundwater with heavy metals such as lead, nickel, copper, etc. In fact, the contamination of groundwater resources with heavy metals and the entry of these waters into the food cycle and health industries, etc., causes environmental problems and diseases such as cancer. Factors such as population growth, the creation of large industrial companies, excessive fuel consumption ....... have caused environmental pollution. Heavy metals are among the pollutants that are present in soils, industrial wastewater, mines and rainwater that flows in rivers, etc. Heavy metals enter as a solution and cause pollution of surface water, groundwater and soil and they pollute them, causing the balance of the ecosystems that enter it to be compromised. In this study, the determination of metal ions by nonlinear chemistry-calibration (Artificial neural network) method is introduced. Fourier transform infrared spectroscopy provides a wealth of information on the molecular structure of water samples which can be used using nonlinear regression methods to predict the concentration of elements in water samples. A calibration model was presented for both training data and test data, and the correlation coefficient were acceptable for each element and 0.9879 for nickel, 0.9789 for copper and 0.9948 for lead, respectively. The root mean square error of the prediction for mentioned elements were 1.4511, 1.3626, 1.1302, respectively. Also, using the standard addition method for the heavy metal concentration was determined. And thethe correlation coefficient for the nickel in water related to the two wells of Rashtaghoon and Mahmoudabad were 0.9992 and 0.9966, respectively, and the copper was 0.9977 and 0.9997, the lead was 0.9770 and 0.9997, respectively. Also, the results related to the limit of detection in Rashtaghoon well for the three elements are 1.1, 0.30 and 0.12, respectively and the results of dynamic range are 4.3, 7.9, 0.62, respectively. In Mahmoudabad well, the lowest limit of detection related to lead was 0.12, and nickel and copper were 1.16, 0.62. The results of linear range of nickel, copper and lead were 4.4, 5.8 and 0.7, respectively. Keywords: Heavy metal, Ground water, Non-linear multivariate calibration, Spectrometry References [1] Alejandro C. Olivieri, “Introduction to Multivariate Calibration”, Springer Nature Switzerland AG,2018 [2] Fabricio A. Chiappini , Hector C. Goicoechea, Alejandro C. Olivieri, “Analytical Figures of Merit for Multisensor Arrays”, Chemometrics and Intelligent Laboratory Systems,206,2020 |
