UV-Vis quantification of hydroxyl radical concentration and dose using principal component analysis

Ronald S Lankone; Alyssa R Deline; Michael Barclay; D Howard Fairbrother

doi:10.1016/j.talanta.2020.121148

UV-Vis quantification of hydroxyl radical concentration and dose using principal component analysis

Talanta. 2020 Oct 1:218:121148. doi: 10.1016/j.talanta.2020.121148. Epub 2020 May 16.

Authors

Ronald S Lankone¹, Alyssa R Deline¹, Michael Barclay¹, D Howard Fairbrother²

Affiliations

¹ Dept. of Chemistry, Johns Hopkins University, Baltimore, MD, USA.
² Dept. of Chemistry, Johns Hopkins University, Baltimore, MD, USA. Electronic address: howardf@jhu.edu.

PMID: 32797904
DOI: 10.1016/j.talanta.2020.121148

Abstract

Hydroxyl radicals (∙OH) are powerful oxidizing species formed naturally in the environment or artificially produced to destroy contaminants in water treatment facilities. Their short lifetime and high reactivity, however, present a significant challenge to quantifying their concentration in solution. Herein, we developed a novel method to accurately measure the steady-state ∙OH concentration and total ∙OH dose produced during the UV photolysis of hydrogen peroxide (H₂O₂) by monitoring the loss of salicylic acid (SA). This information can be acquired using only benchtop UV-Vis spectroscopy, thus expanding measurement capabilities of resource-limited laboratories by eliminating the need for sophisticated instrumentation. To improve the precision with which the rate of SA loss was measured compared to previous methods, we applied principal component analysis (PCA) to fit the UV-Vis spectra collected during SA exposure to ∙OH. For our experimental conditions consisting of 12 mL solutions composed of ≤ 100 mM H₂O₂ and 0.07 mM SA, the steady-state ∙OH concentration throughout the complete photolysis of H₂O₂ was 1.33 × 10^-11 M ± 1.14 × 10^-12 M. This represents more than a ten-fold improvement in reducing the uncertainty of the measurement, with respect to narrowing the 95 % confidence interval, compared to a previous method that employed matrix analysis to process the spectra. Furthermore, the variance of the measured ∙OH concentrations was reduced by a factor of 100 compared to previous methods. Using PCA, the limit-of-detection and limit-of-quantitation for ∙OH are 5.33 × 10^-13 M and 1.23 × 10^-12 M, respectively. By developing quantitative relationships among ∙OH concentration, H₂O₂ concentration, and UV exposure time, we also show how to calculate the equivalent exposure to ∙OH generated in natural aquatic environments by indirect photolysis. Finally, we use this methodology to demonstrate that the presence of suspended carbonaceous nanoparticles at concentrations as high as 300 ppm does not affect ∙OH concentration.

Keywords: Environment; Hydroxyl radicals; Nanoparticles; Quantification; UV–Vis.