Australasian Biotechnology (backfiles) AusBiotech ISSN: 1036-7128 Vol. 8, Num. 5, 1998

Application of Foam Separation Processes for Protein Extraction/Stripping

by Md Monwar Hossain* and Glenn Fenton,

Natural Products Processing, Industrial Research Limited, P O Box 31-310, Lower Hutt, New Zealand
*Author to whom all correspondence should be addressed. Telephone: (64-4-5690000), Fax: (64-4-5690132)

Code Number:AU98036
Sizes of Files:
      Text: 25K
      Graphics: Line drawings (gif) - 97K

The present work examines the foaming characteristics of four well characterised model proteins, in a semi-batch operation. The critical micelle concentration (CMC) value of these proteins was determined by surface tension measurements and was found to be dependent on the physicochemical characteristics of the protein molecule. Protein solutions at initial concentrations higher than CMC value were sparged with air in the separation column. The protein concentration in foam was maximum for b-lactoglobulin, followed by bovine serum albumin (BSA), a-lactalbumin and b-casein. The protein concentration in the residual solution was minimum for BSA, followed by a-lactalbumin, b-casein and b-lactoglobulin.

Separation of BSA was carried out both in semi-batch and continuous operation mode. It was shown that higher enrichment (about twice that of semi-batch) is achievable if the process is operated continuously. There was also slight enhancement in protein recovery with continuous processing.

Foam separation has been studied for recovering high-value components from a large volume of culture media or from natural sources or from industrial waste process solutions. Potential applications reported include heavy metal recovery (Leu et al, 1994), separation of surfactant (Tharapiwattananon, 1996), enzyme purification, and protein extraction (Sarkar etal, 1987) from industrial waste water. The attractiveness of foam-based processes, compared to other separation processes like adsorption, ion exchange, chromatography, precipitation and ultrafiltration, are due to several technical and economic advantages (Gehle and Schügerl, 1984).

The foam separation processes and their potential application have been reviewed by Okamoto and Chou (1979), Wilson and Clark (1987) and Lemlich (1968). The investigations are still on going as waste process streams in many food, pharmaceutical and dairy industries contain small amounts of valuable components, which need to be recovered for economic and environmental reasons. Efficient recovery of these components not only reduces the waste stream BOD, but can also create a potential source of valuable products. Foam separation concentrates the components of a solution by utilising differences in their surface activities. A schematic diagram of a foam separation column is shown in Fig. 1. The process stream is introduced into the column and gas is forced into the solution through a sparger at the bottom of the column. The bubbles are formed and the liquid around them is enriched in the surfaceactive component.

This enriched liquid is carried with the bubbles as they rise through the solution to form a foam or a layer of foam at the top of the liquid. As the foam rises above the foam-liquid interface, some of the entrained liquid drains off due to gravity. When the foam bubbles are collected and allowed to collapse, the liquid is highly concentrated.

In this paper, the foaming characteristics of four model proteins (relevant to a dairy process stream) have been examined in a semi-batch operation. Some of the physicochemical characteristics of these proteins are shown in Table1. The separation performance of Bovine Serum Albumin (BSA) in semi-batch and continuous operational modes have been determined.

Table 1: Properties of proteins and Critical Micelle Concentration (CMC) values

Performance Characteristics

In foam separation two common and important performance measures are enrichment and recovery. Enrichment factor (Ef) is defined as the ratio of foam concentration to that of initial feed.

On the other hand, recovery, R (%), is the fraction of feed protein recovered in the foam. It determines the efficiency of the process and is given by

For production of high purity proteins, both enrichment and recovery are important. In the case of protein stripping out of a process stream, the protein concentration in the residual stream (CR) is an important performance criterion.

The surface tension data at various concentrations were analysed using a modified form of Szyskowski equation (15):

where gS and gW represent the surface tension values of the solute (protein) and water respectively, and A and B are the parameters.

The adsorption of solute from dilute systems at equilibrium can be expressed by the modified Gibbs equation (Atkins, 1990) as below:

where C is the protein concentration in the liquid pool [mg L^-1], Rg is the gas constant [dynes cm K^-1 mole^-1], gs is surface tension of the solute [dynes cm^-1], T is the temperature (K) and G_s is the quantity of absorbed solute at the air-liquid interface [mole m^-2].

The feed solution was prepared batchwise in a container. Reverse osmosis (RO) water was used for all feed samples and the feed temperature was kept in the range 18-20°C. For pH adjustment, of the solution, 0.1M HCl or 0.1M NaOH was added.

Surface Tension Measurements

The surface tension of the protein solutions were measured using a tensiometer and following an ASTM method D971-91 (13). The surface tension of water was measured first and was taken as a reference point. The surface tension of water or the sample was measured by increasing the torque of the ring system. The value at which the ring broke free of the surface was recorded as the surface tension of the sample in dynes/cm.

The foaming experiment was conducted by introducing the feed and compressed air supply into the column. This was continued until no more bubbles were carried over (in batch operation) or until the steady state (continuous operation).

The protein concentrations in the feed, foam and residual solutions were determined by a standard BioRad assay (14) using a dye reagent. The absorbance of the samples were measured at a wave length of 595nm using a Shimadzu UV-160 spectrophotometer. The concentration of the samples were obtained from a standard curve which relates linearly the UV absorbance and concentration.

The surface tension of the proteins ß-casein, BSA, ß-lactoglobulin and a-lactalbumin at various concentrations and pH are shown in Fig. 2. Surface tension of the solutions decreased with concentration and pH more sharply at lower concentrations (i.e. less than 2535mg/L) and at a pH near the isoelectric point. The surface tension attained a constant value at high protein concentrations, which is called the critical micelle concentration (CMC). The value of this CMC is a measure of the affinity of a protein for the gas/liquid interface. The CMC values are listed in Table 1. It is clear that the surface activity of these proteins vary significantly, with BSA being the most surface active protein of these four. As shown in Table1, the CMC value is primarily affected by the molecular weight. As the molecular weight increases more active sites may be available for adsorption, thus lowering the CMC. This value also depends on solution pH, relative to pI. As shown in Fig. 2(c) and (d) the values of surface tension have not reached a constant level, at pH 6.5 for ß-lactoglobulin and 7.5 for a-lactalbumin, respectively. This indicates a larger CMC at pH values higher than pI's of the proteins.

Combining the experimental data of surface tension and using eqns (3) and (4), we obtain the expression for surface adsorption. It is seen that the absorbed concentration at the interface is linearly correlated to the bulk protein concentration:

G_s = kC . . . . . . . . . . .(5)

where k = 7.0 x 10^-8 for ß-casein (pH = 5.9), 3.3 x 10^-8 for BSA (pH = 5.0), 4.1 x 10^-8 for b-lactoglobulin (pH = 6.5), and 8.2 x 10^-8 for a-lactalbumin (pH = 5.1). The smaller values of k for BSA and ß-lactoglobulin suggest that these proteins are likely to be concentrated.

Of the 4 proteins examined in these studies, all displayed different foaming behaviour. The foaming characteristics of the proteins are shown in Table 2. The time required for completion of foaming was about 20-30 minutes. The proteins BSA and ß-lactoglobulin foamed at a superficial velocity (defined as the volumetric flow rate over the cross-sectional area) as low as 0.079 cm/s whereas for the proteins ß-casein and a-lactalbumin, the superficial velocity was about 0.4 cm/s. The percentage volume loss in foam from the starting solution was minimum for a-lactalbumin and increased for ß-casein, BSA and ß-lactoglobulin (maximum).

Table 3: Performance of foam separation of BSA in batch and continuous operations

The protein concentration in the residual solution (C_R) was minimum for BSA, moderate for a-lactalbumin and ß-casein and maximum for ß-lactoglobulin. This suggests that stripping of BSA from any process stream would be greater than any other proteins.

The protein concentration in foam is highest for ß-lactoglobulin (enrichment factor 42), medium for BSA (enrichment factor 33) and low (enrichment factor 11-12) for b-casein and a-lactalbumin.

The performance of foam separation of BSA at pH 4.5, 5.0, 5.5 and 6.5 in both batch and continuous modes are compared in Table3. Enrichment factor (Ef) increased at least by a factor of two when the separation was carried out in a continuous mode. It remained fairly constant at around 60 for pH within 4.5 - 6.5. In batch separation enrichment increased from 14 to 30 when pH was varied from 4.5 to 5 and remained constant at higher pH.

Percentage of protein recovered (%R) increased when the pH of the foaming solution was changed from 4.5 to 5.0 and then it decreased at higher pH. This decrease was significant for batch separations (e.g. 26% at pH 6.5 from 61% at pH 5.5). The maximum recovery (90%) was obtained at pH 5 and in continuous mode. Also the extent of decrease in protein recovery with pH was less (from 82% at pH 5.5 to 60% at pH 6.5).

The effect of varying superficial air velocity on protein enrichment and recovery at a feed concentration 55mg/L, pH 7 and at a feed flow rate of 2.83L/hr is shown in Figure3. Enrichment decreased and recovery increased with increasing air flow rate to the column. Both enrichment and recovery remained more or less at constant level for a superficial air velocity higher than 0.4cm/s. The main effects of higher air velocity are: increase in interfacial area due to higher gas hold-up and increase in bubble size. The first factor increases mass transfer whereas the second one decreases protein loading at the bubble-liquid interface. The increase in air flow could also allow less liquid to be drained (decreasing enrichment) and enhance the flow rate of foam liquid (increasing protein recovery). At much higher air velocity, all these effects could compensate for each other and give a constant level of enrichment and recovery.

It was observed that the different proteins showed a range of characteristics which are important in understanding foam separation performance. The useful attributes of this study are:

BSA with the largest molecular weight is the most surface active protein

Gas flow requirement to produce stable foam is low for BSA and ß-lactoglobulin (the more surface active proteins)

Enrichment factor is high for BSA and ß-lactoglobulin, it is low for ß-casein and a-lactalbumin

Percentage volume loss in foam is in the same order of magnitude, maximum for b-lactoglobulin and minimum for a-lactalbumin

Separation is superior if performed continuously both in terms of protein enrichment and recovery

ASTM Standards (1993). Designation: D971-91, Vol. 5.01, 297.

Atkins, P.W. (1990). Physical Chemistry, 4th edition, (Oxford University Press, UK), Chap.23.

"Bio-Rad Protein Assay", Life Science Group, 200 Alfred Nobel Drive, CA 94597, USA.

Gehle, R.D. and Schügerl, K., (1984). Protein recovery by continuous flotation. Appl. Microbiol. Biotechnol, 20, 133-138.

Lemlich, R., (1968). Progress in Separation and Purifiction, Volume 1, E.S. Perry, Ed., Interscience, New York, pp 1-56.

Leu, M.-H., Chang, J.-E. and Ko, M.-S. (1994). Removal of heavy metals from a chelated solution with electrolytic foam separation. Sep. Sci. Technol., 29, 2245-2261.

Okamoto, Y. and Chou, E.J. (1979). Foam Separation Processes. Handbook of Separation Techniques for Chemical Engineers, P.A. Schweitzer, Ed., (Wiley, New York), Sec. 2.5.

Sarkar, P., Bhattacharya, P. and Mukherjea, R.N. (1987). Isolation and purification of protease from human placenta by foam fractionation. Biotechnol. Bioeng., 29, 934-940.

Tharapiwattananon, N., Scamehorn, J.F., Osuwan, S., Harwell, J.H. and Haller, K.J. (1996). Surfactant recovery from water using foam fractionation. Sep. Sci. Technol., 31, 1233-1258.

Wilson, J. and Clark, A.N., (1987). Bubble and Foam Separations - Waste Treatment. Handbook of Separation Process Technology, R.W. Rousseau, Ed., (Wiley, NewYork), Chap.17.

This research was supported by the Foundation for Research, Science and Technology (FRST), New Zealand, in the Industrial Separation Technologies Programme CO8515.

Thanks are due to Chiew Wong for doing some of the experiments.

Copyright 1998 Australian Biotechnology Association Ltd.

The following images related to this document are available:

Line drawing images