盐水溶液中单硼酸盐物种(B(OH)3和B(OH)-4)的拉曼光谱定量分析

盐湖是天然存在的水和盐类共存的复杂体系，卤水中硼酸盐的赋存形态及分布规律较一般水溶液更为复杂，通常随盐类的浓缩富集而发生复杂的聚合、缔合等作用，存在严重的过饱和性，不利于盐湖硼及其他盐类的分离提取。因此，开展盐湖卤水体系中硼酸盐物种分布规律及离子间相互作用机制研究具有重要的实际意义。激光拉曼光谱因具有原位、无损、且水峰干扰小等特点，被广泛应用于硼酸盐溶液结构光谱学研究中，并表现出较大的优越性。近年来，以化学计量学为核心的现代拉曼光谱定量分析技术已成为快速准确获取复杂体系目标物量关系的有效手段，对光谱解析中面临的光谱重叠、背景干扰、基线漂移等问题具有强大的优势，在分析领域中得到了广泛而深刻的应用。基于化学计量学算法，采用拉曼光谱技术探究了三种回归模型(内标法、多元线性回归和偏最小二乘法)对盐水溶液中单硼物种B(OH)₃和B(OH)^-₄的定量分析，并通过外标样进行方法评估。研究表明，多元线性回归和偏最小二乘法对外标样的预测结果更为准确，相对误差均在1%以内，但前者对低硼含量的预测效果更佳。进一步地，根据建立的多元线性回归模型...

Quantitative Analysis of Monoborates (H3BO3 and B(OH)-4) in Aqueous Solution by Raman Spectroscopy

Salt lake is a natural complex system coexisting with water and salts. Borate species in salt lakes and their distributions are complicated than the pure borate solution. Generally, polyborates can be formed in brine by polymerization during the concentration process. Thus, borates in concentrated brine have a severe supersaturation behavior, which cannot favour the salt lake resource separation between boron and other slats. Therefore, the study of the poly borates distributions in the salt lake brine and their transformation mechanisms is of great importance. Laser Raman spectroscopy is characteristic of in-situ, non-destructive and weak water interference and thus has been widely used to determine borate structure in aqueous solutions. Recently, the modern Quantitative Raman technology with Chemometrics has become an efficient method to accurately acquire the number of matters in a complex system. It shows great advantages in solving spectral problems such as spectral overlap, background interference and baseline drifting and has been widely and deeply applied in the analysis field. Based on the Chemometrics, this paper has studied the quantitative analysis of monoboartes in aqueous solutions by Raman technology, with the three regression models as internal standard, multi-linear regression and partial least squares regression. Also, it has evaluated the three models by using the external standard sample. It was found that both multi-linear regression and partial least squares models had a more accurate amount prediction of the sample, with a relative error of less than 1%. However, the former model shows better values at lower boron concentration. Furthermore, based on the multiple linear regression models, we also explored the borate species and its distribution in the oilfield brine in the west of Qaidam Basin by Raman spectroscopy. The results showed that only the boric acid peak at 875 cm-1 was detected in the oilfield brine during the evaporation process. The amount of boric acid predicted by the multiple linear regression models agrees well with the boric acid concentration measured using the titration method. The relative error between them is less than 5%. It indicates that the major form of borate in the oilfield is boric acid, and other borate species can be ignored, which explains why the boric acid solid is the only borate saltthroughout the whole oilfield brine crystallization process. The results of this study could provide fundamental information and theoretical guidancefor the future exploration of the quantitative analysis of the borate speciation in the brine under dynamic environmental conditions.