ณฐมล จินดาพรรณ. Desalination of fish sauce by electrodialysis. Doctoral Degree(Food Engineering). King Mongkut's University of Technology Thonburi. KMUTT Library. : King Mongkut's University of Technology Thonburi, 2011.
Abstract:
The present research project was conducted to develop and study a process for
production of low-salt fish sauce via electrodialysis (ED). The research project was
divided into 4 parts. The first part was dedicated to the design and construction of a
laboratory-scale ED system for removal of salt from fish sauce as well as to the
determination of the effects of applied voltage (6, 7 and 8 V) and residual salt
concentration (2, 6, 10, 14, 18 and 22 % w/w) on the system performance in terms of
the energy consumption, current efficiency, yield and ED time as well as the selected
physicochemical properties, i.e., total soluble solids, density, viscosity, sodium and
potassium ion concentrations, total nitrogen, total amino nitrogen and color, of the EDtreated
fish sauce. The results indicated that an increase in the applied voltage led to an
increase in the rate of salt removal as well as the rates of changes of the electrical
conductivity and total soluble solids of the fish sauce. The energy consumption
increased, whereas the current efficiency and yield decreased significantly with an
increase in the applied voltage and the salt removal level. The total soluble solids,
viscosity, density, sodium and potassium ion concentrations decreased significantly at
higher salt removal levels, but these parameters were not affected by the applied voltage
level. The applied voltage and salt removal level had significant effects on the color of
the ED-treated fish sauce.
Since ED desalination may affect aroma and taste quality of the ED-treated fish sauce,
the second part of the study was performed to determine the effect of ED desalination
on selected aroma compounds, amino acids and sensory quality, in terms ofthe saltiness
intensity and the flavor difference, of the ED-treated fish sauce having various residual
salt concentrations (2, 6, 10, 14, 18 and 22% w/w). The results indicated that the
amounts of trimethylamine, 2,6-dimethylpyrazine, phenols and all carboxylic acids
except for hexanoic acid significantly decreased, whereas benzaldehyde increased
significantly when the salt removal level was higher. The amounts of almost all amino
acids decreased with the increased salt removal level. This led to a significant flavor
difference between the ED-treated fish sauce and the untreated fish sauce when the salt
concentration was reduced to 18 % (w/w) or lower.
Since varying the ED parameters including the applied voltage and residual salt
concentration affects the ED performance and changes in the fish sauce quality, an
ability to predict these changes and search for an optimized ED condition is necessary to
obtain high-quality ED-treated fish sauce that is not much different from the original
fish sauce. The third part of the study was therefore dedicated to development of models
enabling the prediction of the ED performance and changes in quality attributes of the
ED-treated fish sauce using artificial neural networks (ANNs). In addition, the search
for an optimum condition of a process via the use of multi-objective optimization using
genetic algorithm (MOGA) was conducted. Two ANN models were developed to
predict the ED performance and basic characteristics of the ED-treated fish sauce (i.e.,
total nitrogen concentration, sodium and potassium ion concentrations, total soluble
solids, density, viscosity and Triangle E*) as a function of the applied voltage and residual salt
concentration of the treated fish sauce. Another ANN model was developed for the
prediction of the concentrations of total aroma compounds and total amino acids as well
as flavor difference and saltiness intensity of the ED-treated fish sauce as a function of
the residual salt concentration. Subsequently, two sets of the objective functions based
on the developed ANN models were evaluated to search for the optimum ED condition
via the use of MOGA. The results showed that the optimal ANN models were able to
predict the ED performance with R2 = 0.995, fish sauce basic characteristics with R2 =
0.992 and the concentrations of total aroma compounds and total amino acids as well as
flavor difference and saltiness intensity of the ED-treated fish sauce with R2 = 0.999.
Through the use of MOGA the optimum condition of the ED process was the use of an
applied voltage of 6.3 V and the maintenance of the residual salt concentration of the
treated fish sauce of 14.3 % (w/w).
Finally, since it is expected that the ED-treated fish sauce would be highly perishable
due to its lower salt content, in the fourth part of the study the effect of ED pretreatment
prior to spray drying to produce low-salt fish sauce powder, which is a more shelf-stable
and convenient product compared with liquid fish sauce, was investigated. Product
recovery, morphology and selected physicochemical properties, i.e., moisture content,
water activity, hygroscopicity, as well as X-ray diffraction (XRD) patterns of the fish
sauce powder, were determined. The ED-treated fish sauce having various residual salt
concentrations (in the range of approximately 2-25 % w/w) was spray dried at inlet
drying air temperatures of 120-180 C. It was found that the feed salt concentration and
inlet air temperature had significant effects on all observed variables. This is due to the
fact that the feed salt concentration strongly affected water removal from the ED-treated
fish sauce during spray drying, leading to significant alterations of the ED-treated fish
sauce powder properties.