Near-infrared Spectroscopy For Microanalysis With Functional Enrichment Materials

Near-infrared (NIR) spectroscopy has been proven to be a rapid, non-destructive and cost-effective quantitative method for complex samples. NIR spectrum consists of non-specific and overlapping bands. Therefore, chemometric methods are used to establish calibration models for quantitative analysis. On the other hand, application of the method in microanalysis is badly restricted by the low molar absorptivity of NIR signals. In this thesis, in order to reduce the detection limit, functional materials were prepared for separation and adsorption of micro-compound from aqueous solution. With the aid of enrichment and chemometric methods, the sensitivity of NIR spectroscopy was improved. The contents of this dissertation include:1. A high capacity adsorbent of thiol-functionalized polysilsesquioxane microspheres, named PMPSQ, was prepared for preconcentration of Hg2+and Ag+from aqueous solutions. After adsorbing the analytes onto the adsorbent, near infrared diffuse reflectance spectra (NIRDRS) was measured and partial least squares (PLS) models was established for fast and simultaneous quantitative prediction. Because the interaction of the metal ions with functional group of adsorbent can be reflected in the spectra, the models built with Hg2+and Ag+samples were proven to be efficient enough for precise prediction. The limit of detection for Hg2+and Ag+were0.16and0.15mg L-1, respectively. The wavenumber changes of C-H group are thought to be quantitative basis.2. The possibility to achieve sensitive detection of micro-components by using NIRDRS combined with chemometrics method is studied with two experimental datasets. Layered double hydroxides (LDH) and hydroxyapatite (HAP) were used for adsorption of pesticides and Cr3+from solutions, respectively. Then NIRDRS of adsorbents with analytes were measured directly. The results show that a very high sensitivity can be obtained with the help of chemometric approaches. The maximum absolute prediction error of the method can be as low as8μg for the pesticides and30μg for the Cr3+. The quantity is an equivalence of80ng mL-1and300ng mL-1if100mL solution is used for the analysis. 3. Feasibility for quantitative determination of protein in micro-samples was studied by using NIRDRS. A magnetic materials-magnetic hydroxyapatite (mHAP) was synthesized for enrichment of bovine serum albumin (BSA) from aqueous solutions. After separation under applied magnetic field, mHAP with analyst was dropped onto quartz fiber filter. Then the corresponding NIRDRS were collected and PLS model was established for fast quantitative prediction. The experiment is a tentative exploration for determination BSA with NIRDRS.4. A method for quantitative determination of glutathione (GSH) in solutions was developed by using selective preconcentration and NIRDRS. Maleimido modified SBA-15(MM-SBA-15) was prepared for selective enrichment of GSH. In the experiment, MM-SBA-15with different concentration of GSH was dropped onto quartz fiber filter. Then NIR spectra in diffuse mode were collected and PLS model was established for quantitative prediction. The quantitative analysis result shows that trace GSH can be detected. Therefore, NIRDRS combined with selective preconcentration can provide a feasible way for determination GSH in aqueous solution.