摘要
该文针对传统重金属检测方法速度慢、通量低等问题,发展了表面富集扫描激光诱导击穿光谱(SES-LIBS)检测新方法,显著提升了液态奶中重金属的检测灵敏度和通量。SES-LIBS方法利用金属置换反应原理,将液态奶中游离的痕量重金属离子置换并富集到活性金属基底的特定区域表面,进而高效采集该区域表面的LIBS信号。SES-LIBS方法在有效避免液体离子猝灭效应的同时,显著提升了痕量重金属离子的检测灵敏度。为克服样品采集和基质的干扰,采用重加权特征光谱驱动的自编码孪生多网络算法(RCSD-ASMN)进行SES-LIBS信号解析,从复杂、变动的LIBS信号中准确提取出待测组分的光谱特征信息。SES-LIBS方法可同时检测Cd、Cu等多种重金属元素,检出限分别为0.11、0.13 mg/kg,R^(2)均不低于0.97。实验结果证明,SES-LIBS技术能有效克服不同品牌液态奶基底和重金属元素的交叉干扰,具备良好的检测精度和线性度,为液态样品中重金属的高通量检测提供了一种新手段。
Heavy metal elements represent a significant threat to the quality safety of dairy products,making their rapid detection a focal point in food safety research.This paper introduces a novel detec⁃tion method,surface enhanced scanning laser induced breakdown spectrum(SES-LIBS),for im⁃proving the detection sensitivity and throughput of heavy metals in liquid milk.The SES-LIBS uti⁃lized the principle of metal displacement reactions to enrich trace heavy metal ions in specific areas above an active metal substrate.Subsequently,LIBS scanned the specific surface area of the metal substrate to collect trace heavy metal signals in liquid milk.The SES-LIBS was capable of avoiding the plasma cancellation with the improvement of detection sensitivity for trace heavy metal ions.To overcome both interference of sample collection and matrix in SES-LIBS signals,the paper devel⁃oped a reweighted characteristic spectrum driven auto-encoder siamese multitasking network(RCSDASMN),which accurately extracted spectral information of the interested analytes from complex and fluctuant LIBS signals.The results demonstrated that the SES-LIBS was capable of detecting multiple heavy metal elements such as Cd and Cu simultaneously,with limitation of detection of 0.11 mg/kg and 0.13 mg/kg,respectively.The linearity(R^(2))of the detected elements is not less than 0.97.These results revealed that the SES-LIBS methodology supressed cross-interference from different liq⁃uid milk substrates and heavy metal elements,presenting excellent detection accuracy and linearity.This would definitely provide a novel approach for high-throughput detection of heavy metals in liquid milk samples,which may well extend to other liquid systems.
作者
黄志轩
何天伦
郭祥
陈达
HUANG Zhi-xuan;HE Tian-lun;GUO Xiang;CHEN Da(School of Precision Instrument and Opto-Electronics Engineering,Tianjin University,Tianjin 300072,China;Tianjin Engineering Research Center of Civil Aviation Energy Environment and Green Development,Civil Aviation University of China,Tianjin 300300,China)
出处
《分析测试学报》
CAS
CSCD
北大核心
2024年第7期1032-1038,共7页
Journal of Instrumental Analysis
基金
国家自然科学基金资助(21973111)。
关键词
液态奶
重金属
表面富集扫描激光诱导击穿光谱
重加权光谱
自编码孪生多任务网络
liquid milk
heavy metals
surface enhanced scanning laser induced breakdown spectrum
reweighted characteristic spectrum
auto-encoder siamese multitasking network
