Research On Rapid Quantitative Method For Plastic Composition Of Marine Microplastics Mixture

With the production and application of plastics,a large number of microplastics(with a size of 1 um ～ 5 mm)enter the ocean.Microplastics are the migration carrier of organic pollutants,heavy metals and pathogenic bacteria,and transferred and aggregated through the food chain,which can interfere with biological immunity,metabolism,growth and development,etc.These microplastics threaten the safety of the marine environment.Therefore,the microplastic pollution status has become an important monitoring object for marine ecological environmental protection.At present,the detection method of marine microplastics involves the steps of sampling,separation and purification,species identification and quantitative analysis.Among them,the steps of species identification and quantitative analysis mainly rely on the observation of microplastic particles one by one with a microscope,combined with infrared spectroscopy or Raman spectroscopy,to identify and count the species of the mixture components.However,these methods are difficult to identify and count fine particles,and cannot be accurately quantified only by simple counting,and the detection of a single sample requires a large workload and a long period.Therefore,rapid and accurate analysis of microplastics is an important problem to be solved urgently in the field of marine environmental monitoring.In this paper,a rapid quantitative analysis method for the content of microplastics(mixtures)in the marine environment is proposed and established.The main research contents and conclusions are as follows:(1)In view of the problem that the laser irradiation spot of conventional Raman spectrometers is small(less than about 100 um in diameter)and the spectrum collected by single irradiation cannot represent the overall information of the sample,this paper proposed a large spot(2 mm diameter)Raman Spectral collection method.The spectrum collected by this method can better reflect the composition distribution information of microplastic samples,and the total time for collecting a single sample spectrum is reduced from several days(micromolecular spectroscopy method)to about 15 minutes.The spectral collection efficiency is significantly improved.(2)Based on wavelet transform(WT)combined with adaptive iterative re-weighted penalized least squares(air-PLS),a method for eliminating the fluorescence background of microplastic Raman spectra was established,which reduced the relative standard deviation(RSD)of multiple microplastic Raman spectra collected continuously.The RSD of Raman spectra of Polyethylene(PE),polyethylene terephthalate(PET),polypropylene(PP),polyamide(PA),polystyrene(PS),and ethylene-vinyl acetate(EVA)decreased from 11.33%,9.01%,1.30%,9.66%,1.57%,8.00% to 1.43%,0.38%,0.30%,0.74%,0.69% and 1.11%,respectively.The spectral precision after processing can meet the requirements of quantitative analysis of microplastic mixture components.(3)For common microplastic species such as PE,PET,PP,PA,PS and EVA,microplastic mixture samples containing the above six plastic components and their mixtures with biomass such as fish,shrimp,kelp,and shellfish were prepared to simulate microplastic samples in the marine environment.The study found that the characteristic peaks of the Raman spectra of different kinds of microplastics overlap so much that it is impossible to use the unit analysis method to establish their calibration curves.Therefore,the multivariate analysis of partial least squares(PLS)was used to correlate the spectral intensity and the content of microplastic components,and their calibration models were established.The correlation coefficient of calibration(Rc)and the correlation coefficient of validation(Rp)in the microplastic mixture quantitative models were above 0.95,and the standard error of calibration(SEC)and standard error of prediction(SEP)ranged from1 to 5.In the quantitative models of biomass-containing microplastic mixtures,Rc and Rp were above 0.88,and SEC and SEP values ranged from 1 to 8.(4)In view of the difficulty of using PLS modeling method to analyze the content of actual microplastics,a new method for quantitative analysis of the content of microplastics was established in this paper.Among them,a method for extracting spectral components(MESC)was proposed;the spectral signals of the plastic components to be tested were extracted from the mixture spectra.Any one of the characteristic peaks of the obtained spectral signal of the plastic component to be tested can be used to establish a calibration curve.Based on this method,quantitative calibration curves for PE,PP,PET,PS,PA and EVA in microplastic mixtures were established.The linear correlation coefficients were 0.97765,0.99291,0.99184,0.99638,0.99761,0.96259,respectively.Validation samples were used to predict.The predicted values of each microplastic were consistent with the real values,and the maximum error was less than 2.72 wt%.The simulated marine environment microplastic samples were further studied based on MESC.The linear correlation coefficients were 0.94018,0.95181,0.96109,0.99244,0.99858,and 0.97438,respectively.The maximum error of prediction was less than 3.99 wt%.This method reduces the analysis time of a single sample to 15 minutes.The research results show that the use of large-spot laser Raman spectroscopy combined with a new method for quantitative analysis of the content of microplastics has successfully achieved the rapid and accurate determination of the content of marine microplastics,which is of great practical significance for the monitoring of marine microplastic pollution.