近红外光谱分析技术凭借分析速度快、 样品预处理简单等特点, 被广泛用于农业、 医药、 环境、 石油化工等领域。 为实现高效率的现场检测, 便携式的近红外光谱仪越来越备受关注。 由于近红外光谱分析技术能够对多组分同时进行测定, 因此在该研究中使用便携式近红外光谱仪探索同时对稀溶液中多种重金属离子定量分析的可行性。 使用胺基改性聚合淀粉富集水中的镍离子和铜离子, 然后使用便携式近红外光谱仪直接采集胺基改性聚合淀粉的近红外漫反射光谱, 再借助光谱预处理方法和偏最小二乘回归法建立模型, 最后通过交叉验证和外部验证证明所建立模型的稳定性。 结果表明, 在其他干扰离子存在下, 氨基改性聚合淀粉仍能分别以99.5%和99.8%高效率地富集稀溶液中的镍离子和铜离子。 便携式光谱仪采集的近红外漫反射光谱分别经连续小波变换和多元散射校正结合卷积求导光谱预处理之后, 使用偏最小二乘回归能够得到相关系数分别为0.981 9和0.965 4的稳健模型, 实现了对稀溶液中镍离子和铜离子的同时测定, 并且检测到的镍离子和铜离子的最低浓度均为3.0 mg·L-1。 该方法不仅提高了近红外光谱分析技术的检测灵敏度, 同时也表明了便携式近红外光谱仪对多种重金属离子同时定量分析的可行性, 是一种拓宽近红外光谱分析技术的有益探索。

By virtue of the characters of rapid analysis and simple pretreatment, near infrared reflectance spectroscopy technique is widely used in agriculture, medicine, environment, petrochemical and other fields. To satisfy the rapid on-site identification and analysis, portable near-infrared spectrometers have gained more and more attention. Because near infrared reflectance spectroscopy technique also is a green tool for multi-component analysis, the paper aims at investigating the feasibility for simultaneous quantitative analysis of various heavy metal ions in dilute solution using portable near infrared spectrometer. First, amino modified polymerized starch was used as adsorbent to enrich nickel ions and copper ions in diluted solution. Then, the diffuse reflectance spectra of amino modified polymerized starch samples were measured directly by portable near-infrared spectrometer. Furthermore, with the help of spectral preprocessing methods and partial least-squares regression, quantitative models were built from the near infrared diffuse reflectance spectra of amino modified polymerized starch enriched with heavy metal ions and reference concentrations. At last, the stability of the models was proved through cross validation and external validation. The results show that nickel ions and copper ions in diluted solution can be efficiently enriched by amino modified polymerized starch in the presence of other interfering ions. The adsorption rates for nickel ions and copper ions are 99.5% and 99.8%, respectively. Two robust models can be achieved after spectra processing and partial least squares regression. The spectra processing methods contains Continuous wavelet transform and multiplicative scatter correction combined with Savitzky-Golay. The obtained corresponding correlation coefficients of the two robust models are 0.981 9 and 0.965 4, respectively. Thus, simultaneous quantitative analysis of nickel ions and copper ions in the mixed diluted solution was achieved, and the detectable concentrations of nickel ions and copper ions are both as low as 3.0 mg·L-1. The method not only improves the sensitivity of near infrared reflectance spectroscopy technique, but also demonstrates the feasibility for simultaneous quantitative determination of various heavy metal ions by using portable near infrared spectrometer. Moreover, the method may be a useful exploration to further broaden the application of near infrared reflectance spectroscopy technique.