School of Physics and Electronics, Central South University, Changsha, Hunan 410083, PR China; School of Information Engineering, Shaoyang University, Shaoyang, Hunan 422000, PR China.
School of Physics and Electronics, Central South University, Changsha, Hunan 410083, PR China.
Spectrochim Acta A Mol Biomol Spectrosc. 2019 Dec 5;223:117370. doi: 10.1016/j.saa.2019.117370. Epub 2019 Jul 8.
In the zinc sulfate solution, the concentration ratio of zinc to metal ion impurities can be up to 10, which causes impurity ion signals to be severely masked by the zinc signal. In particular, nickel exhibits a strong nonlinearity. Conventional spectroscopic methods are commonly used to detect multi-component analytes with similar concentrations and require the detection component to be linear to satisfy Beer-Lambert law. In order to solve high concentration ratio and nonlinear problems, a spectrophotometric method combining the extended Kalman filter and derivative methods is proposed to simultaneously determine copper, cobalt and nickel in the zinc sulfate solution by ultraviolet-visible spectroscopy. The derivative method developed by using continuous wavelet transform with a Haar wavelet function was applied to detect copper and cobalt in regions with wavelengths greater than 500nm, in which the absorbance of zinc and nickel changed to a fixed value, where linear regression graphs for copper and cobalt were established at zero-crossing wavelengths. Extended Kalman filter spectrophotometry is a filtering algorithm for nonlinear systems, so it was proposed to iteratively detect nickel concentration. The detection range was found to be 0.5-5mg/L for copper, 0.3-3mg/L for cobalt, and 0.6-6mg/L. The predicted root mean square error was 0.097 for copper, 0.049 for cobalt, and 0.206 for nickel. The average relative deviations of copper, cobalt, and nickel in 10 sets of mixed solutions were 3.19%, 2.23%, and 4.56%, respectively. The spectrophotometric method studied is suitable for real-time detection and control of trace amounts of copper, cobalt, and nickel in purification process of zinc hydrometallurgy, and can be applied to more fields.
在硫酸锌溶液中，锌与金属离子杂质的浓度比可高达 10，这导致杂质离子信号被锌信号严重掩盖。特别是镍表现出强烈的非线性。传统的光谱方法通常用于检测浓度相似的多组分分析物，需要检测元件呈线性以满足比尔-朗伯定律。为了解决高浓度比和非线性问题，提出了一种将扩展卡尔曼滤波器和导数方法相结合的分光光度法，通过紫外可见光谱法同时测定硫酸锌溶液中的铜、钴和镍。采用具有哈尔小波函数的连续小波变换开发的导数方法用于检测波长大于 500nm 的区域中的铜和钴，其中锌和镍的吸光度变为固定值，在零交叉波长处建立了铜和钴的线性回归图。扩展卡尔曼滤波分光光度法是一种用于非线性系统的滤波算法，因此提出了迭代检测镍浓度的方法。发现铜的检测范围为 0.5-5mg/L，钴的检测范围为 0.3-3mg/L，镍的检测范围为 0.6-6mg/L。铜的预测均方根误差为 0.097，钴的预测均方根误差为 0.049，镍的预测均方根误差为 0.206。10 组混合溶液中铜、钴和镍的平均相对偏差分别为 3.19%、2.23%和 4.56%。研究的分光光度法适用于锌湿法冶金净化过程中痕量铜、钴和镍的实时检测和控制，可应用于更多领域。
