Quantitative determination of polymer and mineral content in paper coatings by infrared spectroscopy. Improvements by non-linear treatments

被引：15

作者：

Dupuy, N ^{[1
]}

Ruckebush, C ^{[1
]}

Duponchel, L ^{[1
]}

BeurdeleySaudou, P ^{[1
]}

Amram, B ^{[1
]}

Huvenne, JP ^{[1
]}

Legrand, P ^{[1
]}

机构：

[1] RHONE POULENC RECH,F-93308 AUBERVILLIERS,FRANCE

来源：

ANALYTICA CHIMICA ACTA | 1996年 / 335卷 / 1-2期

关键词：

paper coatings; infrared spectrometry; attenuated total reflectance; partial least squares; chemometrics; neural networks;

D O I：

10.1016/S0003-2670(96)00328-5

中图分类号：

O65 [分析化学];

学科分类号：

070302 ; 081704 ;

摘要：

The quality of a printed image on coated papers, to a great extent, depends on the coating surface properties (nature of the components, relative abundance and so on). The potential of infrared spectroscopy as a rapid analytical technique for the quantitative analysis of the coating is demonstrated. Experimentally, spectra are recorded by Attenuated total reflectance (ATR) with Fourier transform infrared spectroscopy, Partial least squares (PLS) regression is used as the multivariate linear calibration method for predicting calcium carbonate, styrene and butadiene in the coatings. The effect of non-linear treatments is established by the improvements which occurred after the multiplicative scatter correction or the application of an artificial neural network on the PLS scores, especially near the concentration limits of each compound.

引用

页码：79 / 85

页数：7

共 22 条

[11] PARTIAL LEAST-SQUARES REGRESSION - A TUTORIAL [J].