| For a long time,the drug problem has always been a deep-rooted disease that is difficult to eradicate all over the world.Among them,opioids are one of the most abused and harmful drugs.It is of great significance to establish reliable methods of drug detection and identification for the drug control.As a molecular vibrational spectroscopic technique,surface-enhanced Raman spectroscopy(SERS)can enhance the Raman signal of molecules via noble metal substrates,which has the characteristics of fingerprint identification,high sensitivity,rapid response speed,easy operation,etc.These merits make SERS have tremendous application potential in the field of drug detection.Obviously,one of the key elements to achieve efficient SERS detection is high-quality SERS active substrates/chips.In view of this,to develop solid SERS chip and analysis method for accurate identification and efficient detection of opioids,this dissertation carried out a series of research work on the development of solid SERS detection chips and their application in the detection and identification of opioids such as morphine and heroin.The main results are as follows:(1)Polystyrene colloidal spheres template is preparation by the gas-liquid interface selfassembly method,and a gold-coated jagged silicon nanoarray(Au-JSi NA)is fabricated via reactive ion etching of the template and magnetron sputtering for depositing Au.The Au-JSi NA is used as SERS chip to detect morphine in solutions.The results show that trace morphine in solutions could be measured in both dropping and soaking ways adopting Au-JSi NA chip,and the limit of detection is 1×10-4 mg/m L.Meanwhile,the hydrophobic surface of such chip makes it more suitable for detecting morphine in aqueous solutions.Moreover,the surface chemical modification of this Au-JSi NA chip with 3-mercapto-1-propanol or 3-mercaptopropionic acid with appropriate concentration can further increase its SERS performance to morphine.(2)A gold-coated Si nano-cone array(Au-SNCA)is manufactured by further adjusting the conditions of reactive ion etching.Au-SNCA has excellent SERS activity,signal reproducibility and storage stability.Due to the three-dimensional nanostructure can enhance the local electromagnetic field and improve the molecular adsorption capacity,the rapid(5 min)detection of morphine in aqueous samples can be realized using this Au-SNCA as SERS chip,and the limit of detection is 1×10-5 mg/m L,the recoveries of determining morphine in water samples are 90.4%-102.4%.The results of molecular adsorption indicate that the adsorption behavior of morphine molecules on the Au-SNCA chip is consistent with pseudo-second-order kinetic model and Freundlich adsorption mode.Furthermore,this Au-SNCA chip can also be employed to qualitatively identify of morphine and other opioids with similar structure,as well as detect some common non-opioid drugs.(3)A gold/Si O2 composite nanostructure array(Au/Si O2 NSA)is prepared via direct sputtering Au on Si O2 colloidal spheres template obtained by the gas-liquid interface selfassembly.Through optimizing the thickness of Au sediment,the array with equivalent deposition thickness about 165 nm is selected as SERS chip to detect heroin and its main metabolites(6-monoacetylmorphine(6-MAM)and morphine)in solution.Such chip has good SERS activity and signal repeatability,and can realize the determination and recognition of heroin,6-MAM and morphine with the limit of detection is up to 1×10-4 mg/m L.In addition,the differences between Au/Si O2 NSA chip and Au-SNCA chip in production process and detection performance are compared from two aspects of manufacturing process and detection performance,and the characteristics of these two kinds of chips are summarized.(4)Based on the work of SERS detection,pattern recognition is utilized to analyze and process data and establish classification models for the identification of heroin and its metabolites.The results demonstrate that the three kinds of substances can be completely classified by hierarchical cluster analysis,principal component analysis and support vector machine.Simultaneously,heroin,6-MAM and morphine with different concentration can also be quantitatively identified via the established principal component analysis-support vector machine models,and the identification accuracies are 90.9%,84.8% and 70.2%,respectively.Moreover,the results of the classification of the mixture and single components also point out that it is necessary to note the possibility of the Raman spectra of the mixture may be mistaken for the Raman spectra of morphine,resulting in false negative of heroin detection. |
