Abstract
Surface-Enhanced Raman Spectroscopy (SERS) is often used for heavy metal ion detection. However, large variations in signal strength, spectral profile, and nonlinearity of measurements often cause problems that produce varying results. It raises concerns about the reproducibility of the results. Consequently, the manual classification of the SERS spectrum requires carefully controlled experimentation that further hinders the large-scale adaptation. Recent advances in machine learning offer decent opportunities to address these issues. However, well-documented procedures for model development and evaluation, as well as benchmark datasets, are missing. Towards this end, we provide the SERS spectral benchmark dataset of lead(II) nitride (Pb(NO3)2) for a heavy metal ion detection task and evaluate the classification performance of several machine learning models. We also perform a comparative study to find the best combination between the preprocessing methods and the machine learning models. The proposed model can successfully identify the Pb(NO3)2 molecule from SERS measurements of independent test experiments. In particular, the proposed model shows an balanced accuracy for the cross-batch testing task.
Highlights
The contaminants in water are usually complex mixtures of compounds whose detection requires analytical chemistry techniques such as sampling, purification, separation, and quantification using special instruments such as the High-Performance Liquid Chromatography (HPLC) and Gas Chromatography (GC)
The performance of Power Spectrum Density Normalization (PSN) + RBFSVM in the same batch showed an average of 0.946 Balanced Accuracy (BACC)
The performance of the machine learning models according to the SERS preprocessing methods and the reproducibility according to the batch-effect has been rarely discussed
Summary
The contaminants in water are usually complex mixtures of compounds whose detection requires analytical chemistry techniques such as sampling, purification, separation, and quantification using special instruments such as the High-Performance Liquid Chromatography (HPLC) and Gas Chromatography (GC). These analytical chemistry methods exhibit high sensitivity, specificity, and precision, but require expertise because complex instrumentation and analytical procedures are needed. It is not very feasible to apply these technologies that require low-cost, light-weighted portable applications operated by non-professionals To this end, a method for detecting contaminants using the SERS has been actively proposed recently (see, for instance, Bodelon et al [1]). SERS provides greater system design flexibility than Raman spectroscopy, making it suitable for portable applications such as heavy metal detection in water [3]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
