基于荧光分析法和APTLD相结合的多环芳烃的检测

为准确进行浓度检测,用Savitzky-Golay(SG)多项式曲面平滑法去除三维荧光光谱数据的冗余信息,分别采用平行因子法(PARAFAC)算法和交替惩罚三线性分解(APTLD)算法对光谱数据进行分解。设计多环芳烃类污染物的检测实验,分析了芴(FLU)、苊(ANA)及两者混合溶液的荧光光谱特性。FLU溶液在λ_(ex)/λ_(em)=302/322 nm处存在一个明显的荧光峰,并且存在连续侧峰。ANA溶液存在两个荧光峰,分别为λ_(ex)/λ_(em)=290/322 nm和λ_(ex)/λ_(em)=290/336 nm。在激发波长200～370 nm扫描范围和发射波长240～390 nm扫描范围内,FLU和ANA荧光光谱重叠严重。结果表明,两种算法均能分辨出FLU和ANA,并取得了很高的回收率,但APTLD算法的检测效果更好。

Determination of Polycyclic Aromatic Hydrocarbons Combining Fluorescence Analysis with APTLD

In order to detect the concentration accurately, Savitzky-Golay polynomial smoothing method is proposed to remove the redundant information of the three-dimensional fluorescence spectral data. The parallel factor analysis(PARAFAC) and alternating penalty trilinear decomposition(APTLD) are respectively adopted to decompose the spectrum data. The detection experiment of polycyclic aromatic hydrocarbons is designed in this paper. Fluorescence spectrum properties of ANA, NAP and their mixture are analyzed. We found that the main characteristic florescence peak of FLU exists in λ_ex/λ_em=302/322 nm, and there is a continuous side peaks. The two characteristic florescence peaks which exist in ANA are λ_ex/λ_em=290/322 nm and λ_ex/λ_em=290/336 nm. When the excitation wavelength is 200~370 nm and the emission wavelength is 240~390 nm, the fluorescence spectra of FLU and ANA overlap seriously.The experimental results show that both algorithms can all distinguish FLU and ANA, both of them have a high average recovery rate. However, the APTLD algorithm has a better detection result.