Abstract #18

# 18
Feasibility of front-face fluorescence spectroscopy as a tool to understand protein leak during dairy ultrafiltration.
Yizhou B. Ma*1, Jayendra K. Amamcharla1, 1Food Science Institute, Animal Sciences and Industry, Kansas State University, Manhattan, KS.

Ultrafiltration (UF) is widely used in the dairy industry to separate proteins. However, certain undesirable conditions can cause proteins to leak through the membrane into the permeate. The protein leak can cause economic losses and quality deteriorations in permeate powders during storage. The objective of this study was to evaluate front-face fluorescence spectroscopy (FFFS) to understand and detect the presence of true protein in permeates. For this purpose, 12 whey/milk permeate powders were collected from 2 commercial suppliers. Total protein, true protein, and nonprotein nitrogen of the samples were analyzed. The average nonprotein nitrogen of permeate was 2.4 ± 0.18%, but the true protein varied between 0.09 and1.31%. Based on the true protein to total protein ratio, the samples were grouped as high-true-protein (HTP) and low-true-protein (LTP) permeates. The mean true protein of HTP and LTP were 0.77 and 0.22%, respectively and were significantly different (P < 0.05). The emission-excitation matrix (EEM) of one HTP and one LTP permeates (reconstituted to 5% solution) were collected by a spectrofluorometer at 25°C. The EEM was obtained by recording the emission (λEm) spectra from 300 to 500 nm corresponding excitation wavelengths (λEx) ranging from 220 to 360 nm. Three fluorophores identified were tryptophan, Maillard browning products (MBP), and riboflavin. While the MBP and riboflavin showed similar intensities, HTP permeate indicated high fluorescence intensities in tryptophan region compared with the LTP permeate which was selected to characterize the protein leaks. Tryptophan emission spectra (λEx at 280 nm; λEm 300–500 nm) and excitation spectra (λEm at 340 nm; λEx 200–300 nm) were collected for the 12 permeate solutions (5% wt/wt). Spectral data were analyzed by hierarchical clustering analysis (HCA) with complete linkage. HCA classified LTP permeates in one cluster with 100% accuracy while HTP permeates were split into 2 sub-groups. The results showed that FFFS has potential to differentiate LTP permeates from HTP permeates by tryptophan fluorescence spectra.

Key Words: membrane processing, emission-excitation matrix