On-Site/In Situ Continuous Detecting ppb-Level Metal Ions in Drinking Water Using Block Loop-Gap Resonators and Machine Learning

Sangmin Oh, Imtiaz Hossen, Juan Luglio, Gusphyl A. Justin, James E. Richie, Henry Medeiros, Chung Hoon Lee

Abstract: Microwave measurements and machine learning algorithms are presented to estimate metal ion concentrations in drinking water. A novel block loop gap resonator (BLGR) as a microwave probe is designed and fabricated to estimate Pb ion concentrations in city water as low as 1 ppb with an rms error of 0.18 ppb. No physical contact between the BLGR probe and the water sample allows <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">on-site/in situ</i> continuous detection of ppb-level metal ion concentrations. The <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$S_{11}$ </tex-math></inline-formula> raw data (amplitude and phase) from the BLGR are used to classify and estimate metal ion concentrations using a support vector regression algorithm. The performance of the proposed method to estimate Pb concentrations in the presence of interfering metal ions (Cu <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> , Fe <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3+</sup> , and Zn <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2+</sup> ) is also evaluated, and it is found that the average measurement error remains less than 13%.