Research On Heavy Metal Concentrations Based On The Molecular Spectroscopy

A lot of heavy metals were taken into the soil and water with improper utilization ofnatural resources. It is a serious threat to human health that the heavy metals were absorbedinto human body with the form of food and water. At present, routine analytical methods usedfor the determination of heavy metals in the food and water are complex, time-consuming andnot environmentally friendly. Molecular spectroscopy analytical technology is efficient，rapid，no-reagents and can be run at low cost, thus, it has extensive application prospect inenvironmental monitoring field. In this research, it was explored a determination technique ofheavy metals detection with Near-infrared and Raman spectroscopy. Near-infrared （NIR）spectroscopy method was used for heavy metal concentrations detection of Cu and Zn inLudwigia prostrata. And Raman spectroscopy method was used for analysing heavy metalcomplexes of Cu, Zn, Cd and Pb in water. Corresponding detection models of heavy metalswere built with chemometric analysis.Mainly research results were as follows:（1） Pot experiment is conducted to study effect of copper and zinc on Ludwigia prostrata.Growth of Ludwigia prostrata after added copper or zinc into soil is less than control group,the symptoms of poisoning are more apparent with the increasing of copper or zincconcentration. Near-infrared （NIR） spectroscopy method was used for Cu and Zn heavy metalconcentrations detection in Ludwigia prostrata powder. The calibration models weredeveloped of Cu and Zn concentrations in Ludwigia prostrata. The study results wereinvolved as follow: when the near infrared spectra of ludwigia prostrata in copper disposalwas pretreated, the best preprocessing methods corresponding to ideal calibration models forleaf, stem and root were standard normal variate transformation （SNV）, MultiplicativeScattering Correction （MSC） and de-trending, respectively. The same preprocessing for thespectra was obtained experimental group by zinc pollution. Partial Least Squares （PLS） modelboth was better than the other （Principal Component Regression PCR and Multiple LinearRegression MLR） for the heavy metal concentrations of Cu or Zn in Ludwigia prostrata. themodels were evaluate by predicting the24unknown samples, in the experimental group bycopper pollution, the correlation coefficients（rp） were0.97、0.96、0.96, respectively, while theroot mean squared error of prediction （RMSEP） were2.96mg/kg、10.89mg/kg、28.15mg/kg,respectively. In the experimental group by zinc pollution, the rpwere0.97、0.98、0.97, respectively, the RMSEP were9.81mg/kg、10.12mg/kg、12.11mg/kg, respectively. Theresults indicate that NIR spectroscopy technique combined with chemometrics can be used forthe quantitative analysis the concentrations of Cu and Zn in Ludwigia prostrata.（2） The coprecipitation of Mn²⁺-phen-SCN-ternary complex systems for copper（Cu），zinc（Zn），cadmium（Cd） and lead（Pb） in the water were studied．The heavy metal complexeswere collected, before Raman spectroscopy method was used for analysing heavy metalcomplexes of Cu, Zn, Cd and Pb in water. The calibration models were developed of theheavy metal concentrations in water. The study results were involved as follow: when theRaman spectra of heavy metal complexes was pretreated, the best preprocessing methodscorresponding to ideal calibration models for the heavy metal concentrations of Cu, Zn, Cdand Pb were Linear Baseline Correcting, first derivative, standard normal variatetransformation （SNV）, Multiplicative Scattering Correction （MSC） and first derivative,respectively. The Raman spectra with pretreatment were used as the inputs of PLS, PCR andMLR to develop prediction models, and the models were evaluate by predicting the28unknown samples. The results showed that PLS model both was better than the other （PCRand MLR） for the heavy metal concentrations of Cu, Zn, Cd and Pb in water. The rpwere0.979、0.964、0.956、0.972, respectively, while the RMSEP were6.587μg/kg、9.046μg/kg、9.998μg/kg、7.751μg/kg, respectively. The results indicate that Raman spectroscopytechnique combined with coprecipitation method can be used for the quantitative analysis theconcentrations of Cu, Zn, Cd and Pb in water.