Application of the analytical quality by design principles to the development of a qualitative surface-enhanced Raman scattering method: A proof of concept

Analytical quality by design (AQbD) is a systematic approach that allows developing analytical methods in a more science-based way, due to the higher quality of information collected during the development process. However, its application remains limited to the development of separation methods. (i.e., liquid chromatography and capillary electrophoresis). Its application to spectroscopic techniques remains underexplored, despite the potential benefits. The aim of this work was to demonstrate the application of the AQbD principles to surface-enhanced Raman scattering (SERS) method development, detailing each step characterizing this approach. First, crystal violet (CV) was chosen as a model compound and the analytical target profile was defined. Risk assessment was performed to individuate the critical method parameters. The intensity of the CV band at 1175 cm(-1) was selected as critical method attribute to be modeled and maximized. A screening study was conducted to study 11 aggregation agents (AAs) at different concentration levels (range 0.010-1 M) and in two solvent media (aqueous and methanolic). The two AAs exhibiting the strongest signal, one for each medium, were retained. Once these parameters were fixed, an I-optimal design was applied to optimize the entire analysis process. Here, the volumes of nanoparticle suspension and AA solution added to the SERS samples were simultaneously varied. Two method operable design regions, one for each medium, were computed using a Bayesian design space approach. The latter enabled evaluating the failure risk associated to each experimental condition explored and choosing those that best fit the analytical purpose in the experimental domain.

Automated summary

The AQbD principles were applied to the development of SERS method by selecting crystal violet as the model compound and defining the analytical target profile, followed by risk assessment of critical method parameters and a screening study of aggregation agents. An I-optimal design was then implemented to optimize the analysis process.