Interplay of organic components in membrane fouling evolution: Statistical evidence from multiple spectroscopic analyses

Membrane fouling is a complex process that can be caused by multiple organic foulants such as polysaccharides, proteins and humic substances. The key mechanisms, especially interactions between coexisting foulant components, have yet to be well understood. This study proposes a combination of spectral and statistical methods to unravel the interactive role of typical organic foulants during microfiltration, taking bovine serum albumin (BSA), humic acid (HA) and sodium alginate (SA) as examples. Results show that the membrane effluent spectra (ultraviolet absorption, fluorescence and circular dichroism) can differentiate the state of foulant interception at the pore adsorption/blocking stage (BSA and HA actively involved) and gel/cake layer stage (SA dominant). Multivariate redundancy analysis revealed interrelations between the spectral characteristics of the effluent and surface fouling layer. Accumulation of proteins and polysaccharides in the fouling layer was detected by infrared spectroscopy while that of humics was more sensitively reflected by Raman spectroscopy, as confirmed by nonnegative linear least square fitting of time-series spectral data during fouling evolution. A variance partitioning analysis nested in multiple linear regression of the infrared or Raman signals offered further statistical evidence for the interplay of organic components by quantifying the single/double/triple-component contributions to the total fouling response. It revealed that ternary BSA × HA × SA interplay is prominent at the initial pore adsorption/blocking stage whereas binary HA × SA interplay prevails at the later gel/cake layer stage. Such role shifting of foulant interactions during the fouling process might provide useful information for a more accurate understanding and control of fouling. The combined spectral and statistical methods, which can effectively trace inter-component relations and link molecular fingerprints to fouling behavior, are particularly promising for exploring fouling mechanisms in complex systems.