Qualitative and Quantitative Analysis of Three-Dimensional Fluorescence Spectra by Improved Parallel Factor Analysis with Internal Standard Sample Embedding
Hubin Liu,Li Zhang,Guimin Cai,Zhixiang Zhang,Jiachen Li,Longlian Zhao,Junhui Li
DOI: https://doi.org/10.1007/s10895-024-03791-w
2024-06-16
Journal of Fluorescence
Abstract:The strategy of parallel factor analysis, combined with the internal standard method, has been increasingly applied to the qualitative and quantitative analysis of three-dimensional fluorescence spectra of unknown mixed fluorophores. Nevertheless, the disparity in the number of fluorophores included in the internal standard sample set and the number included in test samples may impact the qualitative and quantitative outcomes of parallel factor analysis. In this work, we systematically established the framework of the parallel factor analysis with internal standard sample embedding (ISSE-PARAFAC) strategy. We applied this framework to six datasets representing two scenarios and a real dataset and conducted a detailed discussion on the effects of the disparity between the number of fluorophores in the internal standard sample set and the number in the test set on both qualitative and quantitative results. Additionally, we introduced an enhancement to PARAFAC by aggregating fluorophores with similar emission wavelengths, corresponding to the peaks of emission loadings (spectra) obtained from PARAFAC, as a single fluorophore. This aggregation aimed to mitigate the strong correlation between similar fluorophores. The results imply that the presence of irrelevant fluorophores in the internal standard sample set, whether increased or decreased, does not significantly affect the qualitative and quantitative analysis of target fluorophores in the test set. Moreover, we demonstrated that the improved parallel factor analysis with internal standard sample embedding not only fully decomposes the uncorrelated mixed fluorophores for qualitative analysis but also allows the established linear concentration model for fluorescent components to predict the corresponding fluorophore concentration of test samples, enabling quantitative analysis at the ppm level (mg/L).
chemistry, physical, analytical,biochemical research methods