KEYWORDS: Mid infrared, Linear regression, Near infrared, Bioalcohols, Spectroscopy, Statistical modeling, Near infrared spectroscopy, Light absorption, Laser spectroscopy, Data modeling, Calibration, Attenuated total reflectance
The strong absorption of water in the mid-infrared (MIR) causes difficulties in performing combined near-infrared (NIR) and MIR spectroscopy on aqueous samples using a single instrument. Combining spectra from different instruments can result in unwanted spectral variations, which can influence the prediction models and mitigate the advantages of the combination approaches. In this work, absorption spectra were collected in the NIR (1300nm-2500nm) and MIR (2500nm-3800nm) region by combining a single high-brightness broadband supercontinuum (SC) laser spanning from ~1-4.0 μm with attenuated total reflectance (ATR) and a transmission cuvette in a single-path configuration to provide a uniform spectral response across the NIR and MIR regions. The measured NIR- and MIR- spectra were assessed based on their ability to predict varying concentrations of ethanol, sucrose, and L-proline in aqueous solutions. The NIR-based partial least square regression (PLSR) model gave higher prediction accuracy for sucrose (R2 = 0.95) as compared to both ethanol and L-proline (R2 = 0.75 and R2 = 0.57 respectively). On the other hand, the MIR-based model enhances the prediction accuracy of ethanol (R2 = 1.00) and L-proline (R2 = 0.62) while demonstrating no significant change in prediction accuracy for sucrose (R2 = 0.96). The prediction models based on the combined NIR-MIR spectra performed similar but slightly worse than the MIR-only models for ethanol and L-proline (R2 = 0.97 and R2 = 0.54 respectively), while for sucrose, it was slightly improved (R2 = 0.99).
We demonstrate a high repetition rate (3 MHz) Mid-IR supercontinuum (SC) source spanning whose spectrum spanning 1000-4200 nm using a cascade of different nonlinear fibers. Multi-tone absorption spectroscopy measurements are subsequently carried out using this source and a scanning spectrometer probing various concentrations and a combination of different analytes. We further explore a novel algorithm for rearranging the absorption in the IR-region and the NIR region for three-dimensional modeling. We show this method of analyzing the data is robust, that is being able to predict newly added samples of slightly different nature without having to the recalibrate the model.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.