Related Experiment Video
Updated: Jun 12, 2025
09:42
Fabrication of a Dipole-assisted Solid Phase Extraction Microchip for Trace Metal Analysis in Water Samples
Published on: August 7, 2016
8.7K
Quantification of Lanthanides on a PMMA Microfluidic Device with Three Optical Pathlengths Using PCR of UV-Visible,
Hope E Lackey1,2, Alyssa F Espley1, Savannah M Potter1
1Pacific Northwest National Laboratory, Richland, Washington 99352, United States.
ACS Omega
|September 23, 2024
Summary
This study demonstrates advanced optical spectroscopy in microfluidic devices for precise lanthanide detection. Multivariate and multiblock models significantly improve quantification accuracy, especially for complex samples.
Area of Science:
- Analytical Chemistry
- Spectroscopy
- Microfluidics
Background:
- Microfluidic devices (MFDs) provide efficient platforms for chemical analysis with low sample volumes.
- Optical spectroscopy offers non-destructive monitoring within microfluidic channels, probing analyte speciation and concentration.
- Detecting lanthanides and counterions in solution presents challenges due to spectral overlap and matrix effects.
Purpose of the Study:
- To develop and compare univariate, multivariate, and multiblock optical spectroscopy models for quantifying three lanthanide nitrates in solution using a custom poly(methyl methacrylate) (PMMA) microfluidic device.
- To evaluate the impact of multiple pathlengths and combined spectral data (UV-vis, NIR, Raman) on quantification accuracy.
- To assess the performance of different modeling approaches in terms of root-mean-square error (RMSE), limit of detection (LOD), and residual predictive deviation (RPD).
Main Methods:
- Utilized ultraviolet-visible (UV-vis), near-infrared (NIR) absorbance, and Raman spectroscopy on a PMMA microfluidic device.
- Employed univariate (Beer's Law) and multivariate (Principal Component Regression - PCR) chemometric modeling.
- Investigated single and multiple optical pathlengths, as well as multiblock spectral data fusion (UV-vis, NIR, Raman).
Main Results:
- Univariate modeling provided acceptable quantification for simple signals (e.g., samarium, LOD 5.49 mM).
- Multivariate and multiblock models enhanced quantification for analytes with spectral overlap (e.g., holmium, LOD reduced from 7.21 mM to 3.96 mM).
- Multi-pathlength models maintained accuracy comparable to single-pathlength models, with RPDs ranging from 9.18 to 46.4.
Conclusions:
- Multivariate and multiblock spectroscopic approaches integrated with microfluidic devices offer superior quantification of lanthanides compared to univariate methods.
- Combining multiple spectral data types and pathlengths effectively addresses spectral overlap and matrix interferences.
- The developed methods demonstrate the potential of advanced optical spectroscopy in microfluidic systems for sensitive and accurate chemical analysis.
