中红外衰减全反射光谱对水体硝态氮快速原位监测的关键技术研究

Nitrate nitrogen is the main component of nitrogen pollutants in natural water bodies. Accurate in-situ identification of nitrate nitrogen and its source analysis are important issues that need to be addressed in the control of nitrogen pollution in water bodies. The traditional analysis and determination of nitrate nitrogen in water is mainly accomplished by laboratory chemical analysis, which has a long analysis period and can not meet the urgent need of water mass information collection. Spectrometric analysis is a fast in-situ nondestructive analysis method, which provides a new and important means for the analysis of nitrate nitrogen in water. And mid-infrared attenuated total reflectance spectroscopy (FTIR-ATR) has the advantages of real-time, in-situ, fast rapid in the analysis of spatial and temporal distribution of water pollutants. In this project, taking nitrate nitrogen in water as the research object, rapid in-situ characterization of nitrate nitrogen in water by FTIR-ATR was studied. The response characteristics and mechanism of spectra to nitrate nitrogen under different water conditions were analyzed. A quantitative analysis model was established by combining the mathematical method of multivariate calibration. The quantitative relationship between FTIR-ATR and nitrate nitrogen response in water was established scientifically. The influencing factors and principles of FTIR-ATR in nitrate nitrogen identification in water were expounded systematically. By means of FTIR-ATR indoor simulation study on response characteristics of nitrogen isotope labeled nitrate nitrogen, the dynamic change process of nitrate nitrogen in water and its source and composition were traced. Then it is applied in the analysis of nitrate nitrogen in natural water body and its source analysis, which provide scientific basis for the control and traceability of nitrogen pollution in water body.

硝态氮是水体氮污染物的主要组成部分，其快速原位准确识别及其源解析是水体氮污染控制亟待解决的重要问题。传统水体硝态氮的分析测定主要是借助实验室化学分析完成，其分析周期较长，无法满足水体海量信息采集的迫切需求。光谱分析方法是一种原位快速无损分析方法，为水体硝态氮的分析提供了新的重要手段。本项目利用中红外衰减全反射光谱（FTIR-ATR）在水体污染物时空分布分析方面具有实时、原位、快速等优势，重点剖析在不同水体条件下，光谱对硝态氮的响应特征与机制，结合多元校正的数学方法构建定量分析模型，科学建立FTIR-ATR在水体硝态氮响应中的定量关系，系统阐明FTIR-ATR在水体硝态氮识别中的影响因子与作用原理。通过FTIR-ATR对氮同位素标记硝态氮响应特征的室内模拟研究，示踪水体硝态氮的动态变化过程及其来源与组成。继而在自然水体硝态氮分析及其源解析中进行应用，以期为水体氮污染的控制与溯源提供科学依据。

硝态氮（NO3--N）是水体氮污染物的主要组成部分，其快速原位准确识别及其源解析是水体氮污染控制亟待解决的重要问题。传统水体硝态氮的分析测定主要是借助实验室化学分析完成，其分析周期较长，无法满足水体海量信息采集的迫切需求。光谱分析方法是一种原位快速无损分析方法，为水体硝态氮的分析提供了新的重要手段。本项目以傅里叶变换红外衰减全反射（FTIR-ATR）为技术研究手段，考虑水体的时空变化，重点研究不同水体条件和来源下，中红外光谱对NO3--N响应的特征与机制。结果表明，由于水分的强烈干扰，硝酸盐（1200 ~ 1500 cm-1）的典型吸收与常规的扣水算法并不明显吻合，创新提出了反卷积算法拟合可以有效地减少水干扰，提取有用的光谱信息。光谱第一主成分解释了95 %以上的方差与硝酸盐含量呈线性关系，高浓度组偏最小二乘（PLSR）模型的相关系数（R2）为0.9578，预测集的标准差与校正集的标准差之比（RPD）为4.22，具有良好的预测性能。对于低浓度组模型，R2和RPD分别为0.9865和3.15，预测能力也显著提高。水体碳酸根（CO32-）含量对NO3- -N的光谱特征吸收峰强度有一定的影响，当CO32-浓度达到200 mg/L时，其PLS模型预测水平将大大削弱。南京6个自然水体中硝态氮的时空变异较大，同一季节不同水体中碳酸根的浓度存在一定差异。基于最佳主成分数，采用偏最小二乘法对春夏秋冬四季水体反卷积后1200 ~ 1500 cm-1特征波段的光谱进行建模，表明4个模型的整体预测能力良好，其检测限（LOD）分别为0.3476、0.4039、0.9098和0.4497 mg/L。FTIR-ATR光谱结合高斯反卷积预处理和化学计量学算法可同时为特定类型工业废水中硝酸盐含量的原位监测提供一种快速有效的方法，其中PLSR和SVR模型对电镀废水硝酸盐的预测具有良好的准确性和鲁棒性，而对农药废水和印染废水的硝酸盐预测还需要进一步优化。SVR模型对冶金废水具有较好的预测效果。本研究为FTIR- ATR光谱在水体中硝酸盐的快速原位表征提供了一定的理论依据，也为FTIR- ATR光谱测定水体中硝酸盐的影响机制研究提供了有力参考。