分子光谱的快检技术解决方案

Strategy of Rapid Detection Based on Molecular Spectroscopy

分子光谱技术操作简便快速、仪器小巧、检测成本低,是理想的快检技术。而其灵敏度低、谱带重叠而导致的光谱干扰问题常常影响其应用。对常量组分分析近红外光谱技术是个理想的选择;对微量和痕量,甚至是超痕量物质分析,采用固相萃取(SPE)结合分子光谱是一种合适的途径(固相萃取光谱,SPES)。多元校正,以及光谱处理和波长选择等化学计量学方法,是解决光谱干扰的有效手段,已经广泛应用于近红外光谱分析。我们提出采样误差分布分析(SEPA)方法,来解决近红外光谱建模难,模型稳健性差等问题,SEPA可以应用于异常样本、光谱处理、交互检验、模型评价、模型转移等方面。

Molecular spectroscopic techniques are competitive tools for rapid detection with advantages of simpleness,rapidness,portability,low-cost.But its low sensitivity and spectroscopic interference limit its application.For macro analysis near infrared spectroscopy(NIR)is a good choice,while for micro or trace,or even extra-trace analysis it is a suitable selection to use solid phase extraction coupled with in situ molecular spectroscopies(Solid Phase Extraction spectroscopy,SPES).Chemometric methods including multivariate calibarion,spectral pretreatment and wavelength selection are good ways to improve spectral selectivity,that have been widely used in NIR.We proposed sampling error profile analysis(SEPA)to resolve problem of difficultly modelling and improve robust of the model.SEPA can be used in outliers screening,spectral pretreatment,cross validation,model estimation and model transfer.