Sensitive label-free Cu2O/Ag fused chemometrics SERS sensor for rapid detection of total arsenic in tea

Arsenic (As) is one of the toxic, persistent, and lethal heavy metalloids and requires rapid, less costly, and sensitive detection methods. This study proposed a label-free cuprous oxide/silver (Cu 2 O/Ag) surface-enhanced Raman scattering (SERS) nanoprobe to detect total As in tea. Different total As spiked tea concentrations were mixed with the Cu 2 O/Ag SERS nanoprobe for the SERS detection. Quantitative models were established for predicting the total As in tea by comparatively applying chemometric algorithms. Amongst the algorithms, competitive adaptive reweighted sampling partial least squares (CARS-PLS) optimized the most effective spectral variables to predict the total As in tea efficiently. The CARS-PLS gave the highest correlation coefficient value (R p  = 0.9935), very low root means square error (RMSEP = 0.0496 μg g −1 ) in the prediction set and recorded the highest RPD value of 8.819. The proposed nanoprobe achieved a lower detection limit (0.00561 μg g −1 ), excellent selectivity, satisfactory reproducibility, and stability. No significant difference was recorded when the performance of the Cu 2 O/Ag total As SERS sensor was compared with the inductively coupled plasma mass spectrometry (ICP-MS) method. Therefore, this developed Cu 2 O/Ag coupled chemometrics SERS sensing method could be used to efficiently determine, quantify, and predict total As in tea to promote monitoring of heavy metal contaminants.