Separation of surface active compounds by foam fractionation

The purpose of this paper is to review briefly the theory of separation by foams and to describe a foam fractionation apparatus, which has been found ...
2 downloads 0 Views 2MB Size
R. M. Skomoroskil

Esso Research and Engineering Co. Linden, New Jersey

I

Separation of Surface Active Compounds by Foam Fractionation

Foam fractionation has been used by many investigators to separate mixtures of surface active compounds (1). The separations obtained are surprisingly good considering the relatively simple experimental technique and apparatus used. For example, mixtures of two organic dyes, Patent Blue and Scarlet Red, form violet water solutions whose color intensity depends on the relative concentration of the two dyes. However, after passing air for a short time through the solution to create foam, the residual liquid is red and the foam is blue, indicating that most of the Patent Blue dye has been removed from solution and concentrated in the foam phase. The purpose of this paper is to review briefly the theory of separation by foams and to describe a foam fractionation apparatus, which has been found useful for studing quantitatively the separation of detergents from dilute water solutions. The equipment is not difficult to build or operate and several experiments can he made in 2 or 3 hr. A number of foam fractionation experiments could be profitably incorporated into a laboratory course in analytical or physical chemistry. A surface active molecule contains a polar end ("head"), which is hydrophilic, and a nonpolar, hydrocarbon end ("tail"), which is hydrophobic. A typical example of a surface active molecule is sodium dodecylbenzenesulfonate.

At gas-liquid interfaces, surface active molecules are adsorbed from solution and orient themselves so that their polar "heads" are predominantly in the liquid phase if it is polar, or only their hydrocarbon "tails" are immersed if the liquid is a nonpolar hydrocarbon. This adsorption and orientation of surface active molecules at an air-water interface occurs because the hydrocarbon end of the molecule has little attraction for polar water molecules and will tend to remove itself from solution by accumulating a t the interface (Fig. 1). The amount of surface active compound adsorbed from solution a t an interface can be calculated rigorously using thermodynamics (&, 3). This quantity is given by the Gibbs adsorption isotherm:

where r is the moles of material adsorbed per cm2 of interface surface, c is the concentration (moles/l) of surface active material in the bulk of the solution, r is the surface tension (dynes/cm) of the solution, T is the temperature (OK) and R is the gas constant Present address: Heyden Newport Chemical Corps., Nuodex Products Division, 1075 Magnolia Ave., Elizabeth, New Jeney. 470

/

Journal o f Chemical Education

(8.314 X 10' ergs/'K /mole). Actually, the activity of the surface active compound should be used in the above equation, hut the concentration, c, is used a t low concentrations and it is sufficiently accurate for most calculations. The Gibbs adsorption isotherm indicates that r will he positive if the derivative, dy/dc, is negative. This means the concentration of material at the gas-liquid interface will be greater than in the bulk of the solution if the material is surface active and decreases the surface tension of water. If a procedure could be devised for skimming off the top surface layers at a solution containing surface act.ive molecules, then most of these surface active molecules could be separated from the bulk of the solution and concentrated in a small volume of liquid. Foam fractionation takes advantage of the surface activity of molecules, such as detergents, by providing a practical means for their separation. By creating large liquid-gas interfacial areas, as exists in foams, and then removing this "foam phase" from the surface of the bulk liquid, good separations of surface active molecules from solutions may he obtained.

Figure 1. Schematic representotion of detergent molecvier being a d sorbed ot gar-liquid interface or air bubble pasre, through detergent solution. With roivtionr of sodium dodecylbenrenerulfonote the hydrated polor "heads" reprelent the -SOiNa group and the hydrocarbon "toils" represent the C8H&H2~ group.

A convenient apparatus for studying the foam fractionation of various solutions is shovn in Figure 2. The foam fractionation column is made of glass with a diameter of 2 in. and 3 f t long, although columns of different diameter and length may be used. The column is designed so that the section containing a fritted glass gas disperser, or the column itself, may be replaced. Similarly, the top portion of the column containing foam may be replaced and extensions added t.o lengthen it. The air supply is metered using a rotameter and saturated with water before being introduced a t the bottom of the column. A mercury U-tube is used as a safety valve to prevent excessive

Of course, c, may be obtained by chemical analysis also if a sufficient volume of foam is collected. Calculated values of c, for the three experiments in Table 1 were 780, 400, and 600 ppm. The removal of detergent may be expressed either as per cent, based on concentrations, or as an enrichment ratio, based on the concentration of detergent in the foam and the residual liquid (4). If chemical analysis for detergents or other surface active molecules is not convenient, then foam stability, foam volume, and surface tension measurements may be used to follow the removal by foam fractionation. The data in Table 1 were obtained using the unfoamed solutions only, but they do indicate how these quantities depend on detergent concentration. Thus, foam stability and foam volume increase ~ ~ i detergent th concentration, while surface tension decreases. Hence, foam stability, foam volume, or surface tension may be substituted for chemical analysis in semiquantitative work. For example, calibration curves of surface tension versus detergent concentration may be used t o estimate detergent removal by foam fractionation. Various experiments may be made mith this equiprnmt to study the effect of different variables on foam fractionation. Variables which may be explored are: temperature, aeration time, gas bubble size, gas type, pH of solutions, concentration and type of surface active molecule, and effect of additives on removal (5). It is interesting to note that foam fractionation is being investigated as a potential purification method of waste water effluents (i, 7). The author wishes to thank T. D. Sear1 for help with the analytical work and W. A. Schaefer for constructing the foam fractionation column.

pressure buildup behind the fritted glass disperser a t high aeration rates. A constant temperature jacket is provided to study the effect of altering temperature on foaming. During an experiment the column is filled with a measnred volume of liquid to a mark about 6 in. from the top, the solution is aerated and the foam collected overhead. Samples of foamed water may be withdrawn periodically from the bottom of the column for analysis. Some typical results which were obtained by foaming solutions of sodium dodecylbenzenesulfonate are shown in Table 1. The detergent concentration before foam fractionation varied from 1.5 to 8 ppm and after foam fractionation it was 1 to 2 ppm. The collapsed foam volume accounted at most for only 1 volume per cent of the original liquid foamed. The detergent concentration in the collapsed foam phase may be calculated using a material balance equation: voeo

=

vrca

+ VLC'

(2)

where V refers to liquid volumes and c refers to detergent concentrations. The subscripts 0, F, and L refer respectively to the original solution, the collapsed foam and the residual, foamed solution. The three liquid volumes are measured and the concentration of detergent before and after foam fractionation is determined by chemical analysis. The equation then can be solved readily for the one unknown quantity, c,.

Literature Cited

(1) BIKERMAN, J. J., "Foams: Theory and Industrial Applications," Reinhold Publishing Carp., S e n York, 1953, chap. 10. (2) ADAMSON, A. W., "Physical Chemistry of Surfaces," Interscience Publishers, New York, 1960. (3) DAYIES,J. T., A N D RIDEAL,E. K., ''lnterfilcid Phenomena," Academic Press, New York, 1961 (4) GADEN, E. L.,AND SCRNEPF,R. W., J. Biochem. Mzcrobzol. Tech. Eno.. .. 1.. l(1959). . (5) KARGER,B. L., A N D ROGERS,L. B., Anal. C'hem, 33, 1165

,A""-,. (19Rll

(6) KLEIN,S. A,, AND MCGAUHEY, P. H., K'dater Pollution Control Fedemtim Journal, 35,100 (1963). (7) ELDIB,I. A,, AND SKOMOROSKI, R. M., Renovation of Waste Water by Foam Fractionation, Report to U S . Public Health Service, Contract No. P H 8662-26 (1962-3).

Figure 2. Schematic d i a g r a m of f o a m fractionation equipment: A, air cylinder; 8, go. rotameter; C, gas saturator; D, mercury U-tube; E, liquid t r a p ; F, fritted g l a s g a r disperser; G, conrtont temperature v o t e r jacket; H , f o a m fradionotion column; I , f o a m receiver; J, t a p for withcolumn. drawing f o a m e d solution; K, t a p for drmining f o a m froction+n

Table 1.

Detergent (ppm) Before After fnnmine fmminz ~~~~~~

1.51 2.43 8.04

0.9~1 1 .R9 2.07

Typical Results with Foaming Solutions of Sodium Dodecylbenzene Sulfonate

Foam oollecded (mi) iVol %/n) , , , 2 4

30

0.08 0.16 1.0

Detergent removed (Vni ,

Enriohment ratio, CFICL .

Foam stability (see] . .

34 22 74

789 215 289

3 1 330

Foam volume (mli 4

5 10

tension st 24T idmelem) . . . . 67.0 618 58.3

For this experiment 3 1 of distilled water plus detergent-dodecylbenzenesulfonate, commercial grade-was foamed in s 4-in. diameter foam fractionation column s t 24°C. Thesir flow rate was 30 X 102co/min (STP) for 3 min. Air was dispersed through a fine fritted glass ' foam collected was disc, nominal maximum pore size 4-5.5 microns. The detergent was analyzed by the methylene blue method. Vol % calculated on the initial volume of the foamed solution, 3 1. The detergent removed was calculated on the basis of detergent cancentmtions in solution. Cr/Cz represents concentration of detergent in foam divided by eonoentration of detergent in residual foamed liquid. The foam stability wasmemured by hand-shakmg50 ml of unfoamed solutmn in a stoppered 100 ml graduated cylinder for 30 sec a t a uniform rate and observing the time required for the foam to eollspse. The foam volume is the maximum ohserved during the foam stability test. Surface tension was measured with a duNuoy ring tenslometer. Volume 40, Number 9, Sepfember 1963

/

471