Detection Of Polycyclic Aromatic Hydrocarbons (pAHs) In Aqueous Solution Using Surface-enhanced Raman Spectroscopy (SERS)

Polycyclic aromatic hydrocarbons (PAHs), as widely existed persistent organicpollutants, are carcinogenic, teratogenetic and mutagenic. Until now, there are more than400kinds of carcinogenic PAHs. It is necessary to develop a fast, sensitive, effective andin-situ method for the detection of PAHs in aqueous solution.In the present work, the systematic studies for the detection of PAHs bysurface-enhanced Raman spectroscopy (SERS), such as the preparation of SERS-activesubstrate, the analysis of PAHs’ SERS spectra and the quantitative analysis of the SERSspectra of PAHs mixture using Partial Least Square-Artificial Neural Network (PLS-ANN)algorithm.There are three main works for the preparation of SERS-active substrate. Firstly, animproved self-assembly method for preparing metal (Au and Ag) colloid film, which wascontrolling the temperature of (3-aminopropyl)trimethoxysilane (APTMS) solution duringthe process of quartz silanization, was developed. The SERS-active substrates with thesurface plasmon resonance close to the excitation wavelength were obtained. And also thedependence of SERS activity on the size of gold nanoparticles was studied. For the671nm excitation wavelength, the optimal average size of gold nanoparticles is about35nm.Secondly, calixarene was firstly layered on the gold colloid film by self-assembly method.The SERS substrates with high activity were prepared by incubating the gold colloid filmsin the1M25,27-dimercaptoacetic acid-26,28-dihydroxy-4-terbutyl calix[4]arene (DMCX)ethanol solution for20min. DMCX on the surface of gold colloid film can not onlyprotect the gold colloid film from the damage of other chemicals in water but also attractmore PAHs molecules to the surface of substrate. So the gold colloid film functionalizedby DMCX has high sensitivity and also can be used repeatedly. It is a promising kind ofSERS-active substrate for the in-situ detection of PAHs. Thirdly, it is the first time thatthe gold colloid solution was used as SERS-active substrate to detect PAHs in aqueous solution. Both changing the pH value of gold colloid solution and adding NaCl solutioninto gold colloid solution can greatly improve the sensitivity of gold colloid solution. Andalso there is a dependence of Raman intensities of PAHs on the size of gold nanoparticles.For the785nm excitation wavelength, the optimal gold colloid solution is that with theaverage gold nanoparticles size of58nm and pH=13. And also the gold colloid solutionwith the average gold nanoparticles size of58nm and0.20ml NaCl (0.3M) solution hashigh sensitivity but a little lower than the gold colloid solution with pH=13.The results of the detection of PAHs applying three kinds of SERS-active substrateswere also presented in this thesis. The bare gold colloid films, which are with the averagegold nanoparticles size of30nm and surface plasmon resonance of650nm or the averagegold nanoparticles size of35nm and surface plasmon resonance of636nm, can obtainRaman signals of anthracene at1nM and pyrene at5nM. The quantitative analysiseswere performed and the results showed the SERS intensities increase with the increasingconcentration of PAHs, which is the foundation of quantitative analysis of PAHs in aqueoussolution. On the surface of gold colloid films functionalized with DMCX, five PAHs(naphthalene, phenanthrene, anthracene, fluoranthene, and pyrene) in aqueous solutionwere detected. Raman signals of phenanthrene at10nM, anthracene at0.5nM,fluoranthene at10nM and pyrene at0.5nM can be acquired except naphthalene cannot bemeasured in the air because of its volatility. And also the dependences of Ramanintensities on the concentrations of PAHs and the vibration modes of five PAHs wereinvestigated. Using the gold colloid solution with the average gold nanoparticles size of58nm and pH=13as SERS-active substrate, the SERS spectra of five PAHs (naphthalene,phenanthrene, anthracene, fluoranthene, and pyrene) and their mixtures were measured.Raman signals from10nM naphthalene,4nM phenanthrene,0.5nM anthracene,1nMfluoranthene, and0.2nM pyrene can be obtained even though using the QE65000Ramansystem which is for in-situ detection so that it is small and has low sensitivity comparingwith laboratory-level Raman system. These results demonstrate this SERS-activesubstrate has very high sensitivity and a great potential application in in-situ detection ofPAHs. In order to quantitative analysis PAHs in aqueous solution, the dependences of Raman intensities on the concentrations of PAHs were studied. The results showed therewere linear relationships between Raman intensities and the concentrations of PAHs in thelow concentration range. Comparing the Raman spectrum of the mixture of five PAHsmeasured with the QE65000Raman spectrometer (spectral resolution is about6cm-1) andmicroscopes Raman spectrometer (spectral resolution is about2cm-1), the result showedthat the spectral resolution of the QE65000Raman spectrometer is good enough for thein-situ detection of PAHs mixture. In this part, the detection of PAHs in seawater wasinvestigated for the preparation the detection of PAHs in marine environment. The resultsshowed the measurements work well in seawater used in my work as well as in distilledwater. Analyzing all the above results of the detection of PAHs, the characteristics ofthese three kinds of SERS-active substrate can be summarized as follows: the bare goldcolloid film has high sensitivity, is easily prepared, but it is also easy to cause thephotochemical reaction of PAHs on its surface so that it is easily damaged and cannot beused repeatedly; the gold colloid film layered by DMCX has high sensitivity, avoids thephotochemical reaction of PAHs so that it can be repeatedly used, and it is a promising kindof SERS-active substrate for the in-situ detection of PAHs; comparing with the gold colloidfilm fuctionalized by DMCX, gold colloid solution (the average gold nanoparticles size of58nm, pH=13) has better sensitivity, better spectrum repeatability, and greatest potential inin-situ detection of PAHs, but this kind of SERS-active substrate has the same defect withthe bare gold colloid film, which is easy to cause the photochemical reaction of PAHs andmakes the structural analysis of PAHs difficult.In order to perform the in-situ quantitative detection of PAHs in water, in the presentwork, partial least squares combined with artificial neural network algorithm (PLS-ANN)was applied to quantitatively analyze the SERS spectra of PAHs mixtures. The SERSspectra of62different concentrations PAHs mixtures were measured in gold colloidsolution (the average gold nanoparticles size of58nm, pH=13) using QE65000Ramanspectrometer. The PLS-ANN procedure was wrote in matlab and all parameters wereoptimized. The SERS spectra of the mixtures of naphthalene, phenanthrene, and pyrenewere quantitatively analyzed using PLS-ANN procedure. The average relative error of naphthalene, phenanthrene, and pyrene is29.4%,27.8%, and21.5%respectively. Thepredicting outcomes are unsatisfactory since the SERS spectra are not enough for the neuralnetwork model. In the near future, more SERS spectra of the PAHs’ mixtures will bemeasured and the further studies will be performed. Analyzing the SERS spectra ofPAHs’ mixtures, it can be known that phenanthrene and pyrene can effect the Ramanintensity of naphthalene, there is no interaction between phenanthrene and pyrene, andnaphthalene has no obvious impact on the Raman intensities of phenanthrene and pyrene.The possible reason for these phenomenons is that there are different adsorption propertiesfor different PAHs molecules on the surface of gold nanoparticles. However, the morespecific reasons need to be further studied.