JOURNAL OF COLLOID ~A_ND INTERFAClg SCIENCE 9,3, 7 3 - 7 9
Mechanisms of Water Transport Through Nonaqueous Liquid Membranes HENRI
L. R O S A N O
Department of Chemistry, The City College of the City University of New York, New Yor~, New York 10031 Received April 29, 1966 AI]STRACT
A If two aqueous solutions are separated by an oil layer (e.g., l-butanol or l-pentanol), transport through the oil layer depends upon the ratios of the activities of water or salt at the two oil-water interfaces. Generally, salts diffuse through an oil membrane to a certain extent, but in a direction opposite to that of water transport. Water transport stops when the chemical potentials of the ions in the two aqueous compartments are equal. The rate of water transport also depends upon the chemical structure of the oil layer, its water content, and the nature of the salt. B Water transport through oil-water interfaces is influenced by additives like polyethoxylated trideeyl alcohols dissolved in the oil membrane. This is probably due to the difference in the states of hydration of the ethylene oxide groups adsorbed at the two interfaces. I t would appear that, in addition to transport of water due to the difference in osmotic pressure at the two interfaces transport of water can also be affected by the degree of hydration of a solute. branes. Special e m p h a s i s is p l a c e d on t h e difference b e t w e e n w a t e r m i g r a t i o n a n d t h e p r e v i o u s l y s t u d i e d ion m i g r a t i o n in t h e presence a n d in t h e a b s e n c e of " c a r r i e r " a d d i t i v e s in t h e m e m b r a n e layer.
T h e m e m b r a n e of a living cell m a y be r e p r e s e n t e d as a n o r i e n t e d lipid leaflet bet w e e n t w o aqueous phases. A stirred nona q u e o u s m e m b r a n e s e p a r a t i n g two a q u e o u s solutions m a y , therefore, b e considered a v a l i d m o d e l for s t u d y i n g t r a n s p o r t phen o m e n a . T h e use of a liquid m e m b r a n e p r o v i d e s t w o m a i n a d v a n t a g e s : (1) two l i q u i d - l i q u i d interfaces (1), a n d (2) t h e p o s s i b i l i t y of a d d i n g biologically a c t i v e m a t e r i a l s to t h e m e m b r a n e . I n p r e v i o u s l y p u b l i s h e d p a p e r s (1) it was shown t h a t salt can diffuse t h r o u g h n o n a q u e o u s liquid m e m b r a u e s (diffusion t r a n s p o r t ) a n d in addition, ions can be e x c h a n g e d in t h e presence of an ionized long-chain c o m p o u n d dissolved in t h e " o i l " m e m b r a n e (carrier t r a n s p o r t ) . T h e p r e s e n t p a p e r r e p o r t s a n i n v e s t i g a t i o n of w a t e r m i g r a t i o n t h r o u g h n o n a q u e o u s liquid m e m -
PROCEDURE T h e a p p a r a t u s is essentially a glasscovered 600 c.c. glass b e a k e r w i t h a glass p a r t i t i o n 5.5 cm. high, d i v i d i n g t h e b e a k e r into t w o c o m p a r t m e n t s (Fig. 1). T h e div i d e r itself p r o j e c t s into t h e oil p h a s e 1.5 cm. a b o v e t h e levels of w a t e r in t h e t w o c o m p a r t m e n t s . I n one c o m p a r t m e n t was p l a c e d 120 e.c. of an aqueous salt solution s a t u r a t e d w i t h an Mcohol. I n t h e second c o m p a r t m e n t of t h e cell was p l a c e d 120 c.c. of w a t e r s a t u r a t e d w i t h t h e s a m e alcohol or 120 c.c. of a q u e o u s salt solution also s a t u r a t e d w i t h t h e s a m e alcohol. T h r e e 73
hundred cubic centimeters of the same alcohol saturated with water was floated over the two aqueous solution surfaces. This system thus permitted transfer of water and salts from one aqueous phase to the other only through the nonaqueous STIRRER
"oil" phase. The entire cell was placed in a constant-temperature water b a t h at 30 ° ± .05°C. T h e "oil" membrane, approximately 3 cm. thick, was stirred b y a m o t o r stirrer operating at a constant speed of 60 r.p.m. At this speed the two interfaces were not disturbed and reproducible results were obtained. T h e oil-water interfacial area in each c o m p a r t m e n t was 26.4 cm}. Two side tubes allowed the change in height of the two oil-water interfaces to be measured. A Gaertner precision cathetometer was used to measure the difference between the two oil-water-interface levels as a function of time. The amount of water in the oil, at equilibrium, was determined b y Hill and Malisoff's method (2). EXPEI~IMENTAL
Water migration in the absence of additives in the oil phase. W a t e r saturated with 1-butanol was placed in one c o m p a r t m e n t while different concentrations of KC1 or K2S04 in water saturated with 1-butanol were placed in the other. T h e oil phase con-
W A T E R BATH
FIG. I. Water transport ceil
 KCl i000 mM/l
I Z00 hours
FIG. 2. W a t e r m i g r a t i o n vs. t i m e c u r v e s f o r a cell c o n s i s t i n g of a s a l t s o l u t i o n / l - b u t a n o l / w a t e r .
WATER TRANSPORT THROUGH NONAQUEOUS LIQUID MEMBRANES
sisted of 1-butanol saturated with water. The volume of water which migrated from the "water" to the "salt" compartment was plotted as a function of time. See Fig. 2. Note that aRer 120 hours the same volume of water was transported from one aqueous compartment to the other when the initial salt concentration was either 100 mM./1. of KC1 or 20 mM./1, of K2S04. Solubility of water in 1-butanol vs. salt concentrations, l-Butanol was mixed with aqueous salt solutions of various concen-
trations and the amount of water in the 1-butanol phase was plotted a s a function of the salt concentration of the water phase. See Fig. 3. It is interesting to note that the water solubility vs. salt concentration appears to be independent of the nature of the salt, within experimental error, and decreases linearly with increasing salt concentration on a log scale beyond 0.001 molar. Water migration in the presence of additives in the oil phase. Figure 4 shows the
XlO2 ~ Grams H20 p g 1 cc of 1-butanol
@  LICI
FIG. 3. Solubility of water in 1-butanol vs. salt concentration of the water phase.
M/15 NaaSO4/1-Pentanol + i0 millimoles per liter/H20 Acc of water [of
FIG. 4. Volume of water which migrated from the "water" compartment to the "salt" compartment vs. time. For M/15 Na2SO4/1-pentanol + 10 n~Vi./1 of T-(EtO)~/H20 cell.
volume of water transported as a function of time in a cell consisting of 120 c.c. of M/15 Na2SO4 opposed to 120 c.e. of water separated by 300 e.e. of a solution of 1-pentanol (at 30°C.) containing 10 raM./1, of tridecylalcohol ethylene oxide condensate T-(EtO),,. I t can be seen that T-(EtO)3 markedly impeded the water migration, T-(EtO)6 had practically no effect on it, and both T-(EtO)9 and T-(EtO)15 enhanced the water migration. Similar experiments were repeated with 1-butanol containing the same additives. This was done because water has a greater solubility in 1-butanol than in 1-pentanol and so a more precise determination of the water content can be made with it. The change in volume of water transM/15
ported as a function of time is shown in Fig. 5 for three different experiments in which the only variable was the composition of the oil layer. In experiment 1, the oil layer consisted of 1-butanol saturated with water; in experiment 2, it consisted of 1-butanol + 10 raM./1, of T-(EtO)8 saturated with water; and in experiment 3, it consisted of 1-butanol + 10 millimoles per liter of T-(EtO)~5 saturated with water. In all three experiments one aqueous compartment contained water saturated with 1-butanol, and the other contained M/15 Na2S04 saturated with 1-butanol.
Solubility of water in 1-butanol solutions vs. Na2S04 concentration. Water solubility as a function of increasing Na2SO4 concentration was determined for solutions of
Na2SO~/ i Butanol + i0 millimoles per liter of T-(EtO)n/N~o
~ c c of water
25 Experiment No. i
(30Oc) [xperiment No.2
I0 Experiment No. 3
' 50 hours
FIG. 5. Volume of water which migrated from the "water" compartment to the "salt" compartment vs. time for a M/15 Na2SO4/1-butanol +10 raM./1, of T-(EtO)~/H20 cell © Experiment No. 1 : 1-butanol alone. Experiment No. 2: 1-butanol -F 10 mM./1. T-(EtO)3. ~) Experiment No. 3: 1-butanol + 10 raM./1. T-(EtO)I~.
WATER TRANSPORT THROUGH NONAQUEOUS LIQUID MEMBRANES
GramsH20 p e r 1 cc of 1 - B u t a n o l solutlonsof[T-(EtO)nJ x 102
14 13 12 ii I0
____~/ / 1
N/15 + .i J
Na2SO ~ concentration i ~-
FIG. 6: Solubility of water in I-butanol solutions of tridecyl alcohol-ethylene oxide condensates vs. Na2S04 concentrations of the water phase.
1-butanol containing 10 mM. of 3-EtOtridecyl alcohol and 15-EtO-trideeylalcohol per liter. See Fig. 6. DISCUSSION
Water transport through an oil membrane. Since water migration is a faster process than salt diffusion, it is assumed at the beginning of a typical experiment (e.g., NaC1 solution/i-butanol/water cell) that: a. No salt diffused through the membrahe. b. The initial volume changes of the aqueous compartments do not affect the salt concentrations. c. The amount of water in oil at the interfaces depends upon the solubility of water in oil, which in turn is a function of the salt concentration of the aqueous phase (see Fig. 3). d. The system is in a steady state. According to the diffusion theory (3) the rate of flow of water through a nonaqueous liquid membrane can be expressed by: AND Jb (W1 -- W2)  Here J is the mass of water (in grams) that has migrated per unit time (in see-
onds), A is the cross-sectional area of the diffusion tube, b is the length of the equivalent diffusion path (in centimeters), D is the diffusion coefficient of the water in oil (in cm2/sec.), and Wi and W~ are the respective concentrations of water at the two oil-water interfaces. In this type of experiment the diffusion layers must be very thin owing to stirring. Sehulman and Teorell (4) estimated the thickness of water bound under a moving monolayer to be 30 ~. Moreover, the calculations of Peterson and Gregor (5) show that the thickness of the unstirred film at the surface of a solid ion-exchange membrane increases from 1 to 30 ~ as stirring is decreased. L. Mibashan-Sharaga (6), after studying the desorption of a laurie acid monolayer spread on a M/100 HC1 solution, estimated a quiet zone of 800 just under the film. With the M/15 Na2SO4/1-butanol/water cell (Fig. 5 experiment No. 1), 30 e,c. of water migrated after 50 hours• Equation  can be solved for b, the length of the diffusion p'ath. When A equals 26•4 em 2, D is of the order of 10-s cm.2/see. (its value in most liquid systems (7)), and wl - w2 is 4 )< 10.2 g. of HsO per cubic centimeter
of 1-butanol (see Fig. 3). The calculated value of b is 636 #. Since there are two interfaces, the thickness of one equivalent diffusion layer at the 1-butanol-water interface is of the order of 300 t~. If the oil layer is not stirred, no noticeable change in level occurs after several days. In this case, the "thickness" of the membrane of our cell would be of the order of V/A ~ 10 cm. (V = volume of the nonaqueous layer). The slopes at the origin of these water migration vs. time curves (Fig. 2) appear to depend on the initial molar eoncentrations of the salt, but the extent of the curvature depends on the salt diffusion. Previous work (1) has shown that K2S04 diffuses at a much slower rate than KC1 through l-butanol, and that it is much less soluble than KC1 in l-butanol. Water migration into the "salt" compartment helps to "dilute" the salt solution and "concentrate" the solution in the "water" compartment. Salt diffusion also tends to "concentrate" the solution in the "water" compartment and "dilute" the solution in the "salt" compartment. Water transport ceases when the solutions in the two compartments have reached equal salt concentrations. For a salt completely insoluble in oil, all the water in the compartment containing water saturated with the alcohol will migrate to the compartment containing the aqueous salt solution. It follows, therefore, that K2S04 causes water to migrate over a longer period of time than KC1 does (Fig. 2). However, Figure 4 shows that an appreciable amount of water is transported when T-(EtO)a is used as an additive to the 1-butanol oil membrane (Expt. 2). Therefore, the water transport in this system cannot be interpreted as resulting from the difference in water concentrations at the interfaces. If it is assumed that the two oil interfaces are saturated, Fig. 6 appears to indicate that the water concentrations at the two interfaces are nearly equal. The experiment with an oil phase consisting of i0 millimoles of T-(EtO)15 per liter in l-butanol saturated with water corresponds to experiment 3 in Fig. 5 and shows that water is transported from the
"water" to the "salt" compartment. If the mechanism was solely due to a difference in water concentrations at the two interfaces, the water should have been transported in the opposite direction; i.e., from the "salt" compartment to the "water" compartment, since Fig. 6 shows that WI W2 is negative for l-butanol containing T-(EtO)js. This suggests that the additive was somehow involved in the water transport. It is well established (8) that ethylene oxide condensates are hydrated to a lesser extent in an aqueous NasSO4 solution than they are in pure water. Therefore, it is probable that, during the process of adsorption-desorption at the two oil-water interfaces, the T-(EtO)~ transports a net amount of water from the "water" compartment to the "salt" compartment. By analogy, it is plausible to say that, in addition to osmosis, carrier transport of water in a living membrane is affected by differences in the states of hydration of the faces of the membrane. Furthermore, it is this ability of the T-(EtO)~ to become hydrated that explains why more water can dissolve in l-butanol solutions containing additives than in pure l-butanol (as is evident from a comparison of Fig. 3 with Fig. 6). Another possible explanation is that the additive emulsifies the water in the oil at the more dilute interface. The water globules are then transported (by stirring) to the other interface where they burst (possibly owing to the presence of salt). However, this explanation does not account for the fact that, although W1 - W2 can be negative or positive, the water is always transported from the "water" to the "salt" compartment in the presence of the additives studied. It was previously observed that there are two transport phenomena taking place in opposite directions: water migration and salt migration. With time, the chemical potential of the salt in both compartments equalizes (the volumes being different). As detailed in a previous publication (1 (b)), the migration of different salts through different nonaqueous liquid membranes fits a flux equation in which two parameters
WATER TRANSPORT THROUGH NONAQUEOUS LIQUID MEMBRANES appear to control the rate: (1) the partition coefficient and (2) the rate constant. The rate constant permits the determination of the activation energies for NaC1 and KC1 migrating across w a t e r / l - b u t a n o l / w a t e r interfaces. T h e energies of activation for NaC1 and KC1 migration through 1-butanol are 6 and 5.2 kcal./mole, respectively. Fick's law had previously been rejected on the assumption t h a t the diffusion layers were v e r y thin. The present results on the water transport through l-butanol indicate t h a t this interpretation must be modified. T h e activation energies measured are practically the same as one would obtain b y writing the diffusion coefficient in the f o r m D = Doe -•/Rr and assuming a constant diffusion gradient for the salt under the steady state and as a function of temperature. Diffusion, as the rate-controlling step, would help to explain the differences in rates of transport of NaC1 and I4C1, although their distribution coefficients are the same. Finally, water transport through l-butanol (experiment No. 1, Fig. 5) vs. temperature (15°C.-41°C.) has permitted us to calculate an energy of activation of water
transport of 2.2 kcal./mole, which is in agreement with our conclusions. ACKNOWLEDGMENTS The author expresses his thanks to J. Anderson and L. Smith for their technical assistance, and to the National Science Foundation Undergraduate Research Program at the City College of the City University of New York for financial assistanee rendered. REFERENCES 1. (a) ROSANO, H. L., SCHULMAN, J. i . , AND WEISBUCH, J. B., Ann. N. Y. Aead. Sci. 9, 457 (1961). (b) ROSANO, H. L., DUBY, P., .aND SCHULMAN,J. H., J. Phys. Chem. 65, 1704 (1961). 2. HI~L, A. E., M~LISOFF, W. M., J. Am. Chem. Chem. Soc. 48, 918-827, (1926). 3. ROSANO, H. L., AND LA2g[ER, V. K. or. Phys. Chem. 60, 350 (1956). 4. SCttULMAN, J. H., AND TEORELL, W., Trans. Faraday Soc. 34, 1338 (1938). 5. PETERSON, P., aND GREGOR, H. or. Electrochem. Soc. 106, 1051-1061 (1959).
6. MIBASHAN-SHARAGA,L., Comptes Rendus J. Chimie Physique. Tome 47 (1950) No. 5-6, p. 26-27. 7. GLASSTONE, S., LAIDLER, K. J., AND EYRING,
H., "The Theory of Rate Processes" pp. 526527. McGraw-Hill, New York, 1941. 8. ScI~ICK,M., or. Colloid Sei. 17,801 (1962).