Membrane continuous protein hydrolysis

The membrane reactor concept was demonstrated in laboratory scale a decade ago by Butterworth et al. (15) and by Chose and Kostick (16) in studies on the hydrolysis of starch and cellulose, respectively. Later on several publications have appeared describing the analogous, continuous conversion of various proteins into peptides intended for human nutrition (17-22). Among these works only that of laccobucci et al. (18) presents a quantitative model of the membrane reactor in continuous protein hydrolysis, and it is also the only demonstration of the practical feasibility of the concept in pilot plant scale. 

The results given above indicate that there is no obvious advantage of substituting the existing batch process for production of ISSPH by a membrane reactor process. However, this does not in general mean that continuous protein hydrolysis in a membrane reactor will be uneconomical. For example if the substrate is more completely degradable than soy protein (casein might be such a substrate), it is expected that in a small scale plant (where the capital costs would favour the membrane reactor) the membrane reactor process could be very attractive. The production of protein hydrolysates for dietetic and medical use, could well be considered in this context. 

Mannheim, A. and Cheryan, M. 1990. Continuous hydrolysis of milk protein in a membrane reactor. J. FoodSci. 55, 381-385. 

The production method for low molecular protein hydrolysates has been described earlier W. A controlled batch hydrolysis using the pH-stat is performed, and the protein hydrolysate is then recovered by e.g. solids separation. Hyperfiltration may be used for concentration and/or desalination. Instead of using the controlled batch hydrolysis and solids separation processes, the separation of peptides may be performed from an enzyme-substrate reaction mixture under continuous ultrafiltration in a so-called membrane reactor. 

A few experiments have been carried out in the laboratory scale with a one litre hydrolysis vessel, connected to a small impeller pump and a Sartorius laboratory module fitted with DDS GR6-P membranes (0.2 m ). However, the flow resistance in this module was too large, and it was soon concluded that a resonably constant flux was unattainable. Despite these difficulties, the qualitative behaviour of the reactor variables could be predicted from the model and verified experimentally. For example, with decreasing flux DH increased, but the rate of the base consumption decreased, while the protein concentration in the permeate remained quite stable as predicted. The hydrolysate was evaluated and found comparable in quality to ISSPH produced in the batch process. These results have encouraged us to continue the work in pilot plant with the DDS-35 module, where we can expect considerably more favourable flow conditions. The first experiments carried out so far indicate that a reasonable flux in the order of 50 1/m /h (approx. 1 1/m /min.) can be attained but that foaming problems necessitate the construction of pressurized air free reactor. Future studies will therefore be needed to produce a complete experimental verification of the derived model. 

Fig. 11 shows the principles in an ideal, steady state experiment. At t=0 the hydrolysis is started as a batch hydrolysis ( = 0). When the desired DH-value has been reached (DH=DH at t=t ) the membrane reactor is started, i.e. peptides are permeating through the membrane with the volume flux,, and fresh substrate is added continuously to replace the degraded protein. 

A membrane cell recycle reactor with continuous ethanol extraction by dibutyl phthalate increased the productivity fourfold with increased conversion of glucose from 45 to 91%.249 The ethanol was then removed from the dibutyl phthalate with water. It would be better to do this second step with a membrane. In another process, microencapsulated yeast converted glucose to ethanol, which was removed by an oleic acid phase containing a lipase that formed ethyl oleate.250 This could be used as biodiesel fuel. Continuous ultrafiltration has been used to separate the propionic acid produced from glycerol by a Propionibacterium.251 Whey proteins have been hydrolyzed enzymatically and continuously in an ultrafiltration reactor, with improved yields, productivity, and elimination of peptide coproducts.252 Continuous hydrolysis of a starch slurry has been carried out with a-amylase immobilized in a hollow fiber reactor.253 Oils have been hydrolyzed by a lipase immobilized on an aromatic polyamide ultrafiltration membrane with continuous separation of one product through the membrane to shift the equilibrium toward the desired products.254 Such a process could supplant the current energy-intensive industrial one that takes 3-24 h at 150-260X. Lipases have also been used to prepare esters. A lipase-surfactant complex in hexane was used to prepare a wax ester found in whale oil, by the esterification of 1 hexadecanol with palmitic acid in a membrane reactor.255 After 1 h, the yield was 96%. The current industrial process runs at 250°C for up to 20 h. 

Wei JT, Chiang BH. 2009. Bioactive peptide production by hydrolysis of porcine blood proteins in a continuous enzymatic membrane reactor. J Sci Food Agric 89 372-378. 

---