Quantification of dopamine in biological samples by surface-enhanced Raman spectroscopy: Comparison of different calibration models

In this contribution, an attempt was made to quantify dopamine (DA) in complex biological samples (i.e., plasma and urine) using surface-enhanced Raman spectroscopy. Silver nanoparticles modified with iron-nitrilotriacetic acid (DA-selective probe) and 4-mercaptobenzoic acid (internal standard) were used as enhancing substrate. The performance of different calibration models for the quantification of DA in plasma and urine samples spiked with DA was evaluated and compared. The calibration models investigated were univariate ratiometric model based on the intensity ratio between the characteristic SERS peaks of DA and 4-mercaptobenzoic acid, PLS models built on the raw, standardized or pre-processed SERS measurements, and multiplicative effects model for surface-enhanced Raman spectroscopy (MEM SERS ). Experimental results showed that among the models considered in this contribution, only MEM SERS achieved quite accurate and precise concentration predictions for DA in the spiked plasma and urine samples with recovery rates varying within the range of 91.9–112%, surprisingly better than the corresponding values of the quantitative results obtained by LC-MS/MS. This work provided further evidence of the effectiveness of MEM SERS in solving the problem of poor accuracy of quantitative SERS assays.