文献类型：期刊文献

英文题名：Assessment of ability to detect low concentration analyte with near-infrared spectroscopy based on pre-concentration technique

作者：Zhang, Xuan[1,2];Ren, Yulan[3];Du, Yiping[1,2];Li, Xiaoyu[4];Li, Xiuyong[4];Geng, Jinpei[4];Li, Wei[1,2];Iqbal, Jibran[1,2]

机构：[1]E China Univ Sci & Technol, Shanghai Key Lab Funct Mat Chem, Shanghai 200237, Peoples R China;[2]E China Univ Sci & Technol, Res Ctr Anal & Test, Shanghai 200237, Peoples R China;[3]Mudanjiang Normal Univ, Mudanjiang City 157012, Heilongjiang Pr, Peoples R China;[4]Yantai Entry Exit Inspect & Quarantine Bur, Yantai 264000, Peoples R China

年份：2013

卷号：124

起止页码：1

外文期刊名：CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS

收录：;EI(收录号：20242716629076);WOS:【SCI-EXPANDED(收录号:WOS:000319239000001)】；

基金：This work was supported by the National Natural Science Foundation of China (20975039) and General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China (2011IK226).

语种：英文

外文关键词：Near-infrared diffuse reflectance spectroscopy (DR-NIRS); Pre-concentration; Detection limit; Adsorption isotherm; Multivariate calibration; Quantitative determination

摘要：This paper aimed to assess the ability of detecting low concentration analyte in water by near-infrared diffuse reflectance spectroscopy (DR-NIRS) based on pre-concentration technique. HZ818 resin was employed as the adsorbent, on which aqueous carbaryl was adsorbed, and when adsorption equilibrium was achieved, concentration of carbaryl in solution was detected by HPLC for adsorption isotherm fitting, meanwhile DR-NIR spectrum of the adsorbed resin was measured for quantitative analysis by partial least squares (PLS) regression. Langmuir model was chosen for fitting the experimental data of adsorption isotherm with R-2 of 0.9991 and root mean squared error (RMSE) of 03472 mg/g, which was used to develop an equation declaring the relationship between initial concentration of carbaryl in solution and the equilibrium adsorption mass. With the developed equation, the lowest detection concentration was estimated and the effect of solution volume and mass of resin on the lowest detection concentration were investigated also. Furthermore, 'NIR spectrum-carbaryl concentration' calibration models were built for a serial of solutions in different concentration ranges to validate the detection ability of NIR. The results showed that with an acceptable prediction error as low as about 0.8 mg/g carbaryl adsorbed on the resin can be detected by DR-NIR, so that the lowest detection concentration was exhibited by the developed equation to be 1.7 mg/L under fixed solution volume of 1 Land resin mass of 2 g. It revealed that it was possible to quantitatively detect analyte in solution at ppm-level by the combination of DR-NIRS and pre-concentration technique. Crown Copyright (C) 2013 Published by Elsevier B.V. All rights reserved.