云南化工2025，Vol.52Issue(5)：117-122,6.DOI:10.3969/j.issn.1004-275X.2025.05.27
基于近红外结合化学计量学构建甘草二氧化硫残留量分析模型
Development of Analytical Models for Sulfur Dioxide Residue of Glycyrrhizae Radix Et Rhizoma Using Near-Infrared Spectroscopy Combined with Chemometrics
摘要
Abstract
[Objective]To achieve accurate,rapid,and non-destructive qualitative and quantitative analysis of sulfur dioxide(SO2)residues of Gly-cyrrhizae Radix Et Rhizoma.[Methods]Based on the differences in near-infrared(NIR)spectra of Glycyrrhizae Radix Et Rhizoma with and with-out SO2 residues,qualitative identification and quantitative models were established using partial least squares discriminant analysis(PLS-DA)and partial least squares regression(PLSR),respectively.The accuracy and stability of the models were evaluated using internal model parameters and external test set sample predictions.[Results]1)The NIR spectra of Glycyrrhizae Radix Et Rhizoma with and without SO2 residues showed differ-ences at wavenumbers 7231.8 cm-1,7162.3 cm-1,7112.2 cm-1,6946.3 cm-1(C—H of methylene,aliphatic hydrocarbons),6047.7 cm-1,5962.8 cm-1(C—H of methyl groups bonded to carbonyl,ketones),5881.8 cm-1(C—H of methyl groups bonded to ether,ethers or alcohols),5797.0 cm-1(C—H of methylene,hydrocarbons,methylene),5287.9 cm-1,5183.7 cm-1,4948.5 cm-1,4855.9 cm-1(HN…O＝C of amides,proteins),4763.3 cm-1(C ＝O—O of polymer(C ＝O and C—O stretching),polysaccharides),4485.6 cm-1,4385.3 cm-1(C—H of starch(·C—H and CH3),polysaccharides),4312.1 cm-1(C—H of branched hydrocarbons RC(CH3)3 or RCH(CH3)3,methylene,hydrocarbons,aliphatic hydrocar-bons),4173.2 cm-1,4049.8 cm-1(C—H of methylene in branched hydrocarbons,proteins,aliphatic hydrocarbons).These differences were asso-ciated with changes in the content of polysaccharides,flavonoids,and coumarins of Glycyrrhizae Radix Et Rhizoma.2)The PLS-DA model for i-dentifying SO2 residues of Glycyrrhizae Radix Et Rhizoma achieved R2X,RY,and Q2 values of 0.999,0.979,and 0.985,respectively.A 200-time cross-validation confirmed that the model did not exhibit overfitting.3)The PLSR model for quantifying SO2 residues of Glycyrrhizae Radix Et Rhi-zoma had a determination coefficient(R2)of 0.973,a calibration standard error of 2.670 mg/kg,and a root mean square error of cross-validation(RMSECV)of 3.042 mg/kg.At a significance level of 0.05,a paired t-test between the model-predicted values and reference method-measured values for 14 external test samples showed no significant difference.The average relative prediction error for the test set samples was 4.68％.[Conclusion]The qualitative and quantitative models established in this study demonstrate good stability and accuracy,providing an accurate,rap-id,and non-destructive method for the detection of SO2 residues of Glycyrrhizae Radix Et Rhizoma.关键词
甘草/二氧化硫残留/近红外/定性分析/定量分析Key words
Glycyrrhizae Radix Et Rhizoma/Sulfur Dioxide Residue/Near Infrared Spectrum/Qualitative Analysis/Quantitative Analysis分类
化学工程引用本文复制引用
张怀辉,马睿,张云蓉,唐小涵,孟从燕,侯英..基于近红外结合化学计量学构建甘草二氧化硫残留量分析模型[J].云南化工,2025,52(5):117-122,6.基金项目
云南省科技厅科技计划项目-农业联合专项(项目编号:202301BD070001-236). （项目编号:202301BD070001-236）
