The sections in this article are Introduction Hydrocarbons Aliphatic Hydrocarbons Aromatics Olefins Amines and Amides Hydroxyls Alcohols Phenols Multiple Hydroxyl Compounds Carboxylic Acids Hydrogen Peroxides Water Liquid Water, Ice, and Water Vapor Water in Various Solvents Water in Other Matrices, Including Glasses Ionic Species in Water Carbonyl Compounds Miscellaneous Acknowledgments
The concentrations of methanol and ethanol in carbon tetrachloride solutions have been determined using near infrared (NIR) spectra in the region between 1100 and 2500 nm. Spectral non-linearities due to the effects of concentration and temperature on hydrogen bonding were observed. The feasibility of obtaining accurate predictions of MeOH concentrations was assessed by applying multilinear regression (MLR) techniques, full spectra statistical methods and artificial neural networks (ANN) to their NIR spectra. Principal Component Regression (PCR), Partial Least Squares (PLS) and ANN produced calibration models with decreasing relative standard errors of prediction (RSEP) of 0.02%, 0.012% and 0.008%, respectively, compared with MLR methods based on selected wavelengths which yielded a RSEP of 0.04%. Modelling of ethanol spectra gave similar results. We conclude that using full spectra chemometric techniques and ANN methods can accurately predict alcohol concentrations in spite of non-linearities.
Three 100-compound spectra libraries have been used to evaluate artificial neural network classifications of functional groups. A near-IR gas-phase library was used to compare neural network classifications with those obtained by two-dimensional principal component analysis (PCA) score plots and by the use of the Mahalanobis distance metric based on multidimensional (score) vectors. The neural network using a radial basis function algorithm was able to correctly classify all aromatic and nonaromatic samples in a test set of 40 samples from the 100-compound library; PCA score plots were successful in separating ∼92% of the 100-compound library into aromatic and nonaromatic classes, whereas the Mahalanobis distance metric could not separate the in-class vs out-of-class aromatics in the library. Using principal component scores as input to the neural network training with 40 randomly selected samples, validating with 20 randomly selected samples, and testing with 40 randomly selected samples were performed in less than 5 s and produced perfect classifications. The neural network algorithm incorporating the radial basis function was then used to compare the information available in a near-IR spectral library of condensed-phase molecules with spectra of identical (or very similar) compounds in a mid-IR library. Results with the radial basis function were very good for both libraries, with classifications >85% in all cases. The near-IR library produced better results for aromatics (95 vs 88%), identical or very similar results for OH's (98%), alkyls (>85%), and halogens (98%), and poorer results for carbonyls (85 vs 98%). Better mid-IR results for carbonyls were anticipated due to the sharp band for carbonyl-containing compounds in the fingerprint region; however, the improved results for aromatics in the near-IR were not anticipated.
A mathematical technique for the identification of components in the near-infrared spectra of liquid mixtures without any prior chemical information is demonstrated. Originally, the technique was developed for searching mid-infrared spectral libraries. It utilizes principal component analysis to generate an orthonormal reference library and to compute the projections or scores of a mixture spectrum onto the principal space spanned by the orthonormal set. Both library and mixture spectra are analyzed and processed in Fourier domain to enhance the searching performance. A calibration matrix is calculated from library scores and is used to predict the mixture composition. Five liquid mixtures were correctly identified with the use of the calibration algorithm, whereas only one mixture was correctly characterized with a straight dot-product metric. The predictions were verified with the use of an adaptive filter to remove each of the resulting components from the library and the mixture spectra. In addition, a similarity index between the original mixture spectrum and a regenerated mixture spectrum is used as a final confirmation of the predictions. The effects of random noise on the searching method were also examined, and further enhancements of searching performance are suggested for identifying poor-quality mixture spectra.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.