Optical properties of algogenic organic matter within the growth period of Chlorella sp. and predicting their disinfection by-product formation

Highlights: • Optical properties of AOM were characterized within algal growth phase. • AOM-derived THMs and HAAs within growth phase were also examined. • Correlation tests were conducted between optical properties and DBP formation. • UV absorbance at 254 and 280 nm showed no correlation with THMs and HAAs. • Fluorescent spectra showed a good predictability for DBPs, especially for HAAs. Algogenic organic matter (AOM) in eutrophic waters is a well-known precursor to disinfection by-product (DBP) formation in drinking water. This purpose of this study is (i) to characterize the optical properties of AOM origins, including intra- (IOM) and extra-cellular organic matter (EOM), derived from Chlorella sp. growth as precursors to two major carbonaceous DBPs (C-DBPs), trihalomethanes (THMs) and haloacetic acids (HAAs) and (ii) to correlate these optical properties with THM and HAA formation potential (FP) in order to predict DBP formation. The results show that both EOM and IOM had low UV₂₅₄ and UV₂₈₀ absorbance during their entire growth phase. While IOM chiefly comprised of aromatic proteins and soluble microbial products-like substances (80% of average fluorescent intensity–AFI), EOM spectra were rich in humic- and fulvic-like substances (60% AFI). However, its chemical nature likely differed from terrestrial humics. In DBPFP tests, IOM was a higher-yielding precursor of THMs and HAAs compared to EOM, regardless its growth status. Consequently, C-DBPFP of IOM was always higher than EOM during four growth phases. Results from DBP tests also showed insignificant variation of EOM-derived THMFP and HAAFP during the algal growth phase, while the algal growth status strongly influenced the yields of IOM-derived THMFP and HAAFP. From correlation analysis, our results showed no correlation between UV absorbance with THMFP and HAAFP. Conversely, the regional AFI showed a good correlation with HAAFP and C-DBPFP. Predicting models based on AFI for the formation of HAAs and C-DBPs consequently yielded great predictability for laboratory AOM-containing water samples, with a coefficient of determination R² = 0.879, p < 0.01 and R² = 0.846, p < 0.01. This study indicates a promising application of fluorescent spectra for predicting DBPs derived from algae-rich water sources.