碳纳米管修饰电极-人工神经网络同时检测一氯酚三种异构体  

Simultaneous Determination of o-,m-,p-chlorophenol by MWCNT Modified Glassy Carbon Electrode with the Aid of Artificial Neural Network

作　　者：钱吉兰 朱振中 刘俊康 姜国华 

机构地区：[1]江南大学化学与材料工程学院,江苏无锡214122

出　　处：《分析科学学报》2011年第1期72-76,共5页Journal of Analytical Science

摘　　要：建立了碳纳米管修饰电极-人工神经网络同时检测一氯酚三种异构体的方法。在pH=7.0的磷酸盐缓冲溶液中,邻氯酚、间氯酚和对氯酚在碳纳米管修饰电极上均有一灵敏的不可逆氧化峰。采用微分脉冲溶出伏安法,在优化的实验条件下,邻氯酚、间氯酚和对氯酚的峰电位分别为624 mV、712 mV和632 mV,且浓度分别在2.07～103.66μmol/L、4.62～138.60μmol/L和2.71～108.48μmol/L范围内与峰电流呈良好的线性关系,检出限分别为0.9μmol/L、1.0μmol/L和0.8μmol/L。使用径向基人工神经网络(RBF-ANN)对一氯酚三种异构体混合物的伏安谱图进行分析,总预报误差为7.87%,而偏最小二乘法(PLS)总预报误差为15.57%。结果表明,所建立的方法可同时准确检测一氯酚的三种异构体。A method for simultaneous determination of the mixture of 2-,3-,4-chlorophenol（o-CP,m-CP and p-CP） by multi-walled carbon nanotubes（MWCNT） modified glassy carbon electrode with the aid of artificial neural network was proposed.In 0.1 mol/L phosphate buffer solution（PBS,pH=7.0）,sensitive irreversible oxidation peaks for three target analytes were observed at the MWCNT modified glassy carbon electrode.Under the optimal conditions,the peak potentials of the differential pulse stripping voltammograms of o-CP,m-CP and p-CP were 624 mV,712 mV and 632 mV,respectively.The linear concentrations ranges of o-CP,m-CP and p-CP were 2.07～103.66 μmol/L,4.62～138.60 μmol/L and 2.71～108.48 μmol/L,with detection limits of 0.9 μmol/L,1.0 μmol/L and 0.8 μmol/L,respectively.Furthermore,radial basis function-artificial neural network（RBF-ANN） was used for simultaneous determination of those chlorophenols,and the relative prediction error（RPET） was 7.87%.In addition,the relative prediction error for both single component and total components of RBF-ANN were less than that of the partial least square（PLS） method.Therefore,the proposed method could be used to accurately simultaneous determination of o-CP,m-CP and p-CP.

关 键 词：碳纳米管 径向基人工神经网络 2-氯酚 3氯-酚 4氯-酚 微分脉冲溶出伏安法 

分 类 号：O657.15[理学—分析化学]