66
C.B. Elias and J.B. Joshi
phenomenon. This renewing is achieved by different methods including generating high shear at the surface by rapid flow or by vibration of a porous plate
above the membrane. These methods however may also damage the proteins.
Charm and Lai [39] have compared four different types of membranes used for
the filtration of suspensions and solutions. Suspension of protein micelles
(casein), cell debris as well as enzyme solutions have been studied. The ultrafiltration equipment included in the study included a vibrating porous plate with
plate openings (0.3 cm at the bottom and 0.9 cm at the top) accommodating
a 47 mm membrane filter which could be emptied by periodically flushing out
the contents with air pressure. The second system was a mounted ultrafiltration
module with a channelled membrane support, which allowed a high shear rate
(10000s -1) at the membrane surface and laminar flow in the rectangular
channels with a 150mm diameter membrane. A rectangular ultrafiltration
module with turbulent flow, 0.035 m z membrane area and hold up volume of 71
and a vibrating plate filter accomodating 90 mm membrane with hold up of
0.01121 were the other systems included in their study. The study also involved
different types of membranes viz. 0.22 ~tm and 0.1 gm diameter isotropic microporous membranes, anisotropic diffusive membranes with molecular weight no
greater than 35 000 and anisotropic microporous membranes with molecular
weight no greater than 30 000. The proteins studied were also subjected to shear
stress in a viscometer at comparable levels. They observed that for casein
suspension, the filtration rates were highest in the vibrating plate filter. The
vibration increased the filtration rate by 1.5-fold and prevented rapid clogging of the membrane. Catalase enzyme showed a 7-40% loss in activity in
the different filtration systems whereas the rennet enzyme showed no loss
in activity. However, the rennet enzyme showed a decrease in activity when
treated to shear in a viscometer but recovered its activity on keeping post
shear treatment, indicating that the loss in activity was reversible. The loss
of activity in the catalase enzyme was dependent on the shear stress and time
of exposure.
Bowski and Ryu [74] showed that the recirculation of invertase enzyme
through a thin channel ultrafiltration module did not cause any loss in activity.
Fink and Rodwell [75] observed loss in horse liver alcohol dehydrogenase when
recirculated through a hollow fiber reactor. The apparently different susceptibility of enzymes to shear may be due to differences in their molecular weight and
tertiary structures. Narendranathan and Dunnill [76] investigated loss in activity of yeast alcohol dehydrogenase when recirculated through ultrafiltration
modules consisting of bundles of hollow fibres. Two types of hollow fiber
capillaries were studied, porous and nonporous capillaries. The results with the
non-porous capillaries showed that there was an initial loss of about 20%
protein but subsequently the protein concentration remained constant. The
specific activity of the enzyme however, was close to 100% of the initial activity
throughout the experiment. For the different types of porous capillary membranes tested there was a similar initial loss of protein ranging from 24-37% but
there was no loss in the final specific activity of the enzyme. The maximum shear
C.B. Elias and J.B. Joshi
phenomenon. This renewing is achieved by different methods including generating high shear at the surface by rapid flow or by vibration of a porous plate
above the membrane. These methods however may also damage the proteins.
Charm and Lai [39] have compared four different types of membranes used for
the filtration of suspensions and solutions. Suspension of protein micelles
(casein), cell debris as well as enzyme solutions have been studied. The ultrafiltration equipment included in the study included a vibrating porous plate with
plate openings (0.3 cm at the bottom and 0.9 cm at the top) accommodating
a 47 mm membrane filter which could be emptied by periodically flushing out
the contents with air pressure. The second system was a mounted ultrafiltration
module with a channelled membrane support, which allowed a high shear rate
(10000s -1) at the membrane surface and laminar flow in the rectangular
channels with a 150mm diameter membrane. A rectangular ultrafiltration
module with turbulent flow, 0.035 m z membrane area and hold up volume of 71
and a vibrating plate filter accomodating 90 mm membrane with hold up of
0.01121 were the other systems included in their study. The study also involved
different types of membranes viz. 0.22 ~tm and 0.1 gm diameter isotropic microporous membranes, anisotropic diffusive membranes with molecular weight no
greater than 35 000 and anisotropic microporous membranes with molecular
weight no greater than 30 000. The proteins studied were also subjected to shear
stress in a viscometer at comparable levels. They observed that for casein
suspension, the filtration rates were highest in the vibrating plate filter. The
vibration increased the filtration rate by 1.5-fold and prevented rapid clogging of the membrane. Catalase enzyme showed a 7-40% loss in activity in
the different filtration systems whereas the rennet enzyme showed no loss
in activity. However, the rennet enzyme showed a decrease in activity when
treated to shear in a viscometer but recovered its activity on keeping post
shear treatment, indicating that the loss in activity was reversible. The loss
of activity in the catalase enzyme was dependent on the shear stress and time
of exposure.
Bowski and Ryu [74] showed that the recirculation of invertase enzyme
through a thin channel ultrafiltration module did not cause any loss in activity.
Fink and Rodwell [75] observed loss in horse liver alcohol dehydrogenase when
recirculated through a hollow fiber reactor. The apparently different susceptibility of enzymes to shear may be due to differences in their molecular weight and
tertiary structures. Narendranathan and Dunnill [76] investigated loss in activity of yeast alcohol dehydrogenase when recirculated through ultrafiltration
modules consisting of bundles of hollow fibres. Two types of hollow fiber
capillaries were studied, porous and nonporous capillaries. The results with the
non-porous capillaries showed that there was an initial loss of about 20%
protein but subsequently the protein concentration remained constant. The
specific activity of the enzyme however, was close to 100% of the initial activity
throughout the experiment. For the different types of porous capillary membranes tested there was a similar initial loss of protein ranging from 24-37% but
there was no loss in the final specific activity of the enzyme. The maximum shear