【作者】 马超群；

【导师】 陈国庆；

【作者基本信息】 江南大学 ， 光学工程， 2011， 硕士

【摘要】 本文实验研究了3种合成食品色素赤藓红、新红和靛蓝的吸收光谱和荧光光谱特性,对3种色素在不同外界环境条件下其光谱特性的变化进行了实验研究和理论分析。基于光谱特性数据,应用小波变换、导数光谱技术和人工神经网络等方法实现了多组分色素的同时测定。以浓度均为30μg/ml的赤藓红、新红和靛蓝水溶液为实验样品,对其吸收光谱和荧光光谱及相关特性进行了测量,得到了吸收峰值波长、吸光度、荧光的最佳激发波长、发射峰值波长、发射谱宽、荧光量子产率和荧光寿命等光谱特性参数。配制了15种浓度的赤藓红、新红和靛蓝的水溶液,分别测量了在各自的最佳激发波长激发下的荧光光谱,通过曲线拟合得到了荧光峰值强度随溶液浓度的增加先增大后降低的结果,对赤藓红溶液随着溶液浓度增加而出现激发光谱突变及荧光发射峰值波长红移的现象进行了分析,给出了合理的解释。研究了外界环境因素对色素光谱特性的影响:分别以水、甲醇和乙醇为溶剂,测量了不同溶剂的赤藓红溶液的荧光发射峰值波长和荧光量子产率,得到赤藓红的发射峰值波长随着溶剂的极性增大出现红移、随着溶剂形成氢键能力的增大而出现蓝移,荧光量子产率随着溶剂形成氢键能力的增大而减小的结果;测量了不同pH值的赤藓红水溶液的荧光强度,得到其荧光强度在酸性条件下随着pH值的增加而增大、在碱性条件下与溶液pH值无关的结果;测量了不同温度的新红水溶液的荧光强度和荧光寿命,通过曲线拟合得到其荧光峰值强度与温度之间满足( F0 - F) F=ke-E RT+C的关系、其荧光寿命在5℃～90℃的温度范围内随温度的升高而线性递减的结果;用阳光、紫外灯和荧光灯等光源照射靛蓝水溶液,测量了照射不同时间后其吸收光谱和荧光光谱,通过对比不同光源照射下光谱特性的变化,得到靛蓝在紫外光照射下会出现分解,在光照5小时内,吸收光谱变化不大,荧光强度随照射时间而明显增强;对靛蓝水溶液进行持续加热,测量保持85℃不同时间的靛蓝水溶液的吸收光谱和荧光光谱,对比光谱特性变化得到靛蓝在加热条件下会出现分解的结论,并给出了靛蓝溶液的分解机理。运用小波变换、导数荧光光谱和RBF神经网络等方法,建立一种基于荧光光谱的预测模型,实现了对亮蓝和靛蓝的同时测定。首先配制了24种不同浓度的亮蓝和靛蓝混合溶液,以其中19种作为训练样本,以另外5种作为预测样本,分别测量了各个样本的荧光发射光谱。通过小波变换压缩光谱数据,用导数荧光光谱的方法窄化光谱、减少光谱重叠,以求导后得到的小波系数作为输入,训练、建立RBF神经网络进行预测,结果为,测定亮蓝和靛蓝浓度的平均相对误差分别为1.84%和1.26%。该方法可推广应用于对其他多组分合成食品色素或荧光物质进行同时测定。本文的研究结果可为光谱检测精度的提高、合成食品色素的安全使用、毒理学研究等方面的工作提供帮助。本文提出的对多组分色素溶液浓度的预测方法,可为食品安全检测提供一种新的思路。更多还原

【Abstract】 The absorption and fluorescence spectral characteristics of three synthetic food colors Erythrosine, New Red, and Indigotine were studied in this paper, and the variations of their spectral characteristics under different external environmental conditions were analyzed. Based on the spectral characteristic data, we successfully realize simultaneous determination of multicomponent dyes by combining wavelet transformation, derivative fluorescence spectrometry, and artificial neural networks.From the absorption and fluorescence spectra measured experimentally, we obtained some spectral characteristics such as absorption peak wavelength, absorbance, optimal excitation wavelength, emission peak wavelength, emission spectral width, and we get spectral characteristics of fluorescence quantum yield and fluorescence lifetime according to quantum yield and fluorescence decay experiments. Moreover, we measured the fluorescence spectra of respective fifteen concentrations of Erythrosine, New Red, and Indigotine solutions excited by respective optimal excitation wavelengths. The relationships between spectral characteristics of fluorescence peak intensity, emission peak wavelength and concentrations of synthetic food colors solutions can also be obtained.In order to analyze the influence of external environmental conditions on spectral characteristics of synthetic food colors, we carried out a series of work as below: First, we measured the fluorescence emission spectra and fluorescence quantum yields of Erythrosine in water, methanol and ethanol respectively, and obtained the results that the fluorescence emission peaks of Erythrosine exhibited red shift with the increase of solution polarity and blue shift with the increase of hydrogen bond strength, and the fluorescence quantum yields reduced with the increase of hydrogen bond strength; Second, we measured the fluorescence intensities of Erythrosine solutions under different pH value and analyzed the variation law; Then, we measured the fluorescence intensities and fluorescence lifetime of New Red solutions under temperature, and obtained that the relationship between the fluorescence intensities and temperature accorded with the function ( F0 F) F=ke?E RT+C, and the fluorescence lifetime decreased linearly with the increase of temperature; Moreover, we irradiated Indigotine solution by sunshine, ultraviolet lamp, and fluorescent lamp and measured the absorption and fluorescence spectra after different irradiation time, in addition, we heated Indigotine solution under 85℃and measured the absorption and fluorescence spectra after different heating time, in conclusion, we got the mechanism of degradation. The results indicate that the external environmental conditions have significant effects on the spectral characteristics of the three synthetic food colors. Under certain external environmental conditions, synthetic food colors may appear to degrade and degenerate.In this paper, we introduced a method by combining wavelet transformation, derivative fluorescence spectrometry, and radial basis function (RBF) neural network together to establish a predicting model for simultaneous determination of multicomponent dyes. First, we prepared twenty-four different concentrations of mixed Brilliant Blue and Indigotine solutions, taking nineteen of them as training samples, and the other five solutions as predicting samples. And then we measured the fluorescence emission spectra of the samples respectively. Wavelet transformation was used to compress spectral data obtained by experiment, and derivative fluorescence spectrometry was used to narrow spectral width and decrease spectral overlap. We took the derivative wavelet coefficients as inputs to train and establish an RBF neural network model to predict the concentrations of the five predicting samples, the mean relative predicting errors of Brilliant Blue and Indigotine are 1.84% and 1.26% respectively. This method can also be applied to determinate other multicomponent colors and other fluorescent substances simultaneously.The studies on spectral characteristics of the three synthetic food colors can offer a help for improving the accuracy of spectral detection, safe use of synthetic food colors and toxicology research. The method of establishing a model for predicting the concentrations of multicomponent colors solutions can provide a new way for food safety detection.更多还原