基于BP神经网络的日落黄溶液荧光光谱的研究

Fluorescence spectra of sunset yellow based on BP neural network

对食用合成色素日落黄的荧光光谱进行了研究,发现在310-400nm紫外光的激励下,日落黄溶液发出强荧光,峰值荧光强度随浓度的增加先增强后减弱,且荧光谱峰位置出现明显红移。经分析认为,日落黄溶液能产生荧光是因为分子中偶氮键将一个苯环和一个萘环连接在一起,形成大共轭结构,并且取代基-SO3Na与-OH处于萘环的对位,大大增强了日落黄分子的共轭程度,使其具有强的吸光功能,发出强荧光。另外,结合BP神经网络,通过训练好的网络对4种不同浓度的样本进行浓度预测,结果表明相对误差分别为4.269%,6.078%,4.977%和5.308%,相对标准偏差分别为0.448%,0.375%,0.419%和0.414%。实验表明,该方法具有训练速度快、预测结果准确度高等特点,有望成为一种对食用合成色素进行高效、痕量检测的有效方法。

The fluorescence spectra of synthetic edible pigment of sunset yellow is analyzed, it is found that when the light between 310nm to 400nm is induced, the fluorescence intensity increases and then weakens with the enhancement of the concentration and appears red shift. It is considered sunset yellow solution can produce fluorescence because nitric group links a benzene ring and a naphthyl together to form a conjugate structure locating at the contraposition of naphthyl in place of the group- SO3Na and-OH, it greatly increases the conjugate extent of sunset yellow elements to absorb and give off fluorescence. In addition, By BP neural network, 4 different concemrations of samples is predicted, the relative errors are 4. 269 %, 6. 078 %, 4. 977% and 5. 308%. The RSDs of the results are 0. 4478% ,0. 3752% ,0. 4186% and 0. 4135. The method.is are better than those of others in training speed and accuracy, it is a rapid and effective way to detect trace of synthetic edible pigment.