Strategies for solving matrix effects in the analysis of triphenyltin in sea-water samples by three-way multivariate curve resolution

This study attempts to overcome the influence of the matrix effect in quantitative predictions using second-order multivariate curve resolution. When the sensitivity of the calibration is affected by the presence of the chemical matrix of the sample, predictions using pure standards are expected to be biased. In classical univariate calibration, this drawback is solved by using a standard addition method for calibration. Here, the combination of standard addition with multivariate curve resolution is shown to improve significantly the accuracy of predictions in the presence of matrix effects.