Optimization of a robust and reliable FITC labeling process for CE-LIF analysis of pharmaceutical compounds using design of experiments strategy.

Fluorescence, especially laser induced fluorescence (LIF), is a powerful detection technique thanks to its specificity and high sensitivity. The use of fluorescence detection hyphenated to separation technique often requires the labeling of analytes with suitable fluorescent dye, such as FITC for the labeling of molecules presenting amino groups. Nevertheless, the labeling of analytes could be a tedious, time consuming and a non-robust step of the analytical workflow. In this context, the objective of the present work was to propose a robust and reliable FITC labeling process. Primary and secondary amino compounds (i.e. synthetic cathinones) were selected as model compounds because they are representative of a large proportion of pharmaceutical small molecules. Based on prior knowledge, DoE combined with multivariate statistical modeling was performed to optimize the process. Reaction time and pH of reaction buffer were highlighted as the most critical parameters to control the process. The study showed also the benefit of short reaction time to maximize the labeling efficiency. Indeed, optimal condition was defined as reaction time of 32 min with ratio between FITC and analytes of 40.4 and the buffer reaction pH of 9.7. In addition, variance component analysis was integrated to the DoE to estimate the variability of process and to evaluate its applicability for quantitative purpose. These chemometric approaches helped to develop an efficient labeling process able to reach high sensitivity for CE-LIF analysis (i.e. 10 nM) with good precision (i.e. intermediate precision values lower or close to 5 %).