The FTIR-ATR Spectral Analysis Of Mineral Oil In Water

The progress of science and technology promoted the rapid development ofpetrochemical industry, and the country increased the investment in oil and petrochemicalindustry, but people ignored the concerns for the environment under the background ofabundant material resources. Water quality detection technology is facing theunprecedented challenge for that the mining enterprises in the actual operation processproduced a large number of sewage with mineral oil which serious threatens to thesurrounding water quality security and the stability of water circulation system. Therefore,rapid and accurate detection of the types and content of mineral oil in water pollution hasimportant significance for the timely screening and control of pollution sources.This article is based on the Fourier transform infrared(FTIR) spectroscopy, whichcombinines with the attenuated total reflection(ATR) technology, and the related modeloptimization method is used as technical means, which realises the building of mineral oilcomponents in index model.Due to the complexity of mineral oil composition, the acquisition of spectrumcontains abnormal spectral data when measure the FTIR-ATR spectrum of mineral oil inwater directly. The method of principal component analysis(PCA) combined with markovdistance(Mahalanobis) is used to eliminate singular samples which contains abnormal data.The PLS respectively combine with random method, the concentration gradient and SPXYmethod to establish regression model, then use regression image fitting with acombination of assessment indicators to select the preferred method of sample divided.Sample collection has been divided on the use of chemometric methods for modeling,namely the use of PLS, IBPLS modeling analysis, to get the optimization model algorithm.Order to further optimize the model, SG smoothing method can be used for spectralpretreatment, to improve the SNR of the model, and establish the model which can detectoil diesel and petrol in the water rapidly at the same time.