紫外分光光度-局部偏最小二乘回归法同时测定三种人造甜味剂  

Simultaneous Determination of Three Artificial Sweeteners Using UV Spectroscopic Combined with Local Partial Least Squares Regression Method

作　　者：王海燕[1,2] 宋超[1] 张正勇[1,2] 谢伟量 李丽萍[3] 刘军[1,2] 沙敏[1,2] 

机构地区：[1]南京财经大学管理科学与工程学院,南京210046 [2]江苏省质量安全工程研究院,南京210046 [3]南京理工大学机械工程学院,南京210094

出　　处：《化学通报》2016年第8期744-748,共5页Chemistry

基　　金：质检公益性行业科研专项(201410173);国家自然科学基金项目(71433006;61373058)资助

摘　　要：本文报道了一种简便、快速、准确地同时测定3种人造甜味剂安赛蜜、阿斯巴甜和糖精钠的方法。方法基于在p H为3.21的盐酸溶液中对安赛蜜、阿斯巴甜和糖精钠三组分混合溶液进行紫外光度测定,所得重叠光谱数据分别用偏最小二乘回归法（PLSR）、特征峰结合PLSR法和特征峰结合局部偏最小二乘回归法（LPLSR）进行处理。结果表明,选取特征波段的峰值作为自变量,采用4个局部样本做拟合的预报误差最小,总相对偏差仅为3.05%。对果汁样品进行测定,获得了很好的定量分析结果。安赛蜜、阿斯巴甜和糖精钠的定量线性范围分别为1.0～30.0 mg/L、1.0～10.0 mg/L和1.0～10.0 mg/L。Acesulfame-K,aspartame and saccharin sodium are commonly used as food flavoring agents. All of them show strong absorptions in the UV spectral region which are seriously overlapped. This paper reports simultaneous determination of them using UV spectroscopic combined with chemometric methods. Firstly,partial least squares regression（ PLSR） was used,the total relative deviation was 7. 01%. Secondly,PLSR combined with the characteristic peaks were employed,the total relative deviation was 4. 67%. Lastly,local partial least squares regression（ LPLSR） combined with the characteristic peaks were employed for the design of a model. The characteristic peaks of UV spectra of calibrating samples were chose as the independent variable factor,and LPLSR was used in data processing. Compared with PLSR method,the total relative deviation of LPLSR method is smaller,which is only 3. 05%. Finally,the model was used to the concentration prediction of beverage samples with satisfactory results. The linear range of acesulfame-K,aspartame and saccharin sodium is 1. 0 ～ 30. 0 mg / L,1. 0 ～10. 0 mg / L and 1. 0 ～ 10. 0 mg / L respctively.

关 键 词：紫外分光光度法 安赛蜜 阿斯巴甜 糖精钠 局部偏最小二乘回归法 

分 类 号：O657.3[理学—分析化学] TS202.3[理学—化学]