PARAFAC and FRI Preferred 3D Fluorescence Extraction Time of Dissolved Organic Matter

The judgment and assessment of preprocessing scheme for 3D fluorescence spectroscopy Optimized still depend on the total extraction volume. Since the release of the DOM during extraction with the discrepancy between components, the studies might be more reasonable if we had considered the extraction of DOM components. This paper aims to clarify the differences in the extraction of DOM components and total and whether it is necessary to set the separate extraction time. Two common extraction methods, oscillating and centrifuging, were used. The time was set as the upper limit of the high-speed centrifuge and the constant temperature shaking box. PARAFAC and FRI were used to characterize the DOM compositional characteristics at different extraction times. According to the correlation and principal component analysis, it was found that there were different interactions between the components, and the optimal preprocessing time was screened out on the condition that the regional integral value and Fmax reached the maximum. The humus-like components were strongly positively correlated with the total (p<0. 05), and the protein-like components were less correlated with the total using FRI analysis. The humic acid-like components were strongly positively correlated with the total. The fulvic acid-like fractions were weakly correlated with the total amount using PARAFAC analysis (p<0. 05). There were differences in the extraction of total and components, which required separate pre-treatment time settings. The choice of extraction time is premised on the economy and stability of the distribution, the reby choosing the maximum of fluorescence regional integration or fluorescence intensity score (F-max). The results showed that the optimal extraction times for entirety, humic acid-like, fulvic-like, complexine-like, tryptophan-like substances and microbial metabolites were 12, 21, 12, 12, 12, and 12 h, respectively. PARAFAC analysis showed that the optimal extraction times for entirety, C1, C2, and C3 were 12, 21, 12, and 39 h, respectively. In summary, the two extraction methods have their advantages. The same extraction time can be used for the components and the overall time under the centrifugal treatment. The regional integral and Fmax values under the shaking treatment showed more oscillation than the centrifugation, and the extraction effect is relatively stable. The research results can provide a data basis and guidance for optimizing the preprocessing scheme when measuring DOM by three-dimensional fluorescence spectroscopy.