Multivariate Curve Resolution Applied to Infrared Reflection Measurements of Soil Contaminated with an Organophosphorus Analyte
Multivariate curve resolution (MCR) is a powerful technique for extracting chemical information from measured spectra on complex mixtures. The difficulty with applying MCR to soil reflectance measurements is that light scattering artifacts can contribute much more variance to the measurements than the analyte(s) of interest. Two methods were integrated into a MCR decomposition to account for light scattering effects. Firstly, an extended mixture model using pure analyte spectra augmented with scattering ‘spectra’ was used for the measured spectra. And secondly, second derivative preprocessed spectra, which have higher selectivity than the unprocessed spectra, were included in a second block as a part of the decomposition. The conventional alternating least squares (ALS) algorithm was modified to simultaneously decompose the measured and second derivative spectra in a two-block decomposition. Equality constraints were also included to incorporate information about sampling conditions. The result was an MCR decomposition that provided interpretable spectra from soil reflectance measurements.
- Research Organization:
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States). Environmental Molecular Sciences Lab. (EMSL)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- AC05-76RL01830
- OSTI ID:
- 889036
- Report Number(s):
- PNNL-SA-43768; APSPA4; 1626; NN2001000; TRN: US200619%%333
- Journal Information:
- Applied Spectroscopy, 60(7):713-722, Vol. 60, Issue 7; ISSN 0003-7028
- Country of Publication:
- United States
- Language:
- English
Similar Records
Application of Extended Inverse Scatter Correction to Mid-Infrared Reflectance Spectra of Soil
Multivariate analysis of X-ray, ion and electron spectral images: from surface to 3D materials characterization.