Mixed Bilayers of Anionic and Nonionic Surfactants on Alumina When nonionic hydrocarbon surfactants with different oxyethylene chain lengths are added to aalumina previously flocculated by addition of an anionic hydrocarbon or an anionic fluorocarbon surfactant, the flocculated alumina redisperses by the formation of mixed bilayers in such a manner that the hydrophobic chains of the nonionic surfactants are in contact with the anionic surfactant-covered alumina and their hydrophilic groups toward aqueous phase. From measurements of the adsorbed amount of surfactants and of steady-state emission spectra of pyrene solubilized in the mixed bilayers, it is found that the mixed bilayers of anionic hydrocarbon-nonionic hydrocarbon surfactants are formed preferentially to those of anionic fluorocarbon-nonionic hydrocarbon surfactants and that the formation of mixed bilayers occurs more easily with decreasing oxyethylene chain length of the nonionic surfactants. Meanparticle size measurements indicate that the redispersion of alumina with mixed bilayers proceeds more easily with increasing oxyethylene chain length of the nonionic surfactants. © 1990AcademicPress,Inc.
Recently, the structural characteristics of the surfactant layer on solids have been elucidated by fluorescence and electron-spin-resonance probing methods. Chandar et al. ( 1) reported from pyrene and dinaphthylpropane fluorescence probes that highly organized surfactant aggregates of sodium dodecyl sulfate at the alumina-water interface are formed by the association of hydrocarbon chains. A similar experiment for Triton X-100-silica was also carried out by Levitz et aL (2). On the other hand, Meguro and Kondo (3) reported that positively charged particles previously flocculated upon addition of an anionic surfactant are redispersed upon addition of another surfactant owning to the formarion of a mixed surfactant-adsorbed layer. Furthermore, Esumi et al. (4-6) studied the properties of mixed surfactant bilayers of anionic and nonionic surfactants on monodispersed ferric hydrosols by measuring the zeta potential, particle size, and adsorbed amount and found that the mixed surfactant bilayer of sodium dodecyl sulfatepoly(oxyethylene)nonylphenyl ether (NP) on the sol is more easily formed than that of lithium perfluoro-l-octanesulfonate (LiFOS)-NP. Our objective in the present investigation is to extend our study of these mixed bilayers on alumina by using nonionic surfactants having different oxyethylene chain lengths. Two different systems were studied: lithium dodecyl sulfate ( LiDS ) - N P and LiFOS-NP. Poly (oxyethylene)nonylphenyl ethers having average oxyethylene chain lengths of 7.5, 10, and 20 were used, which are referred to as NP 7.5, NP 10, and NP 20, respectively. EXPERIMENTAL Materials. LiDS and LiFOS were synthesized in our laboratory and purified (7). NP was obtained from Nikko Chemical Co. Ltd. and used without further purification. The water used in all experiments was purified by passing
through a Milli-Q system until its specific conductivity fell below 10 -7 ~2-1 cm-k The alumina used was a-alumina of 99.995% purity supplied by Showa Denkou Ltd. The specific surface area and mean-particle diameter were 10.1 m2/g and 500 nm, respectively, Methods. The alumina flocculated upon addition of a certain concentration of LiDS or LiFOS at pH 3.5 where the alumina had an optimum positive zeta potential. Then NP was added into the dispersion of flocculated alumina. After reaching adsorption equilibrium, the interaction between surfactants on the alumina was estimated by measuring zeta potential and mean-particle size of alumina. The zeta potential was measured using a Pen Kern Model 500 laser zee meter. The mean particle size of alumina was measured using Malvern Model 700 autosizer which is based on photon correlation spectroscopy. Further, after centrifugation of the suspension, the concentration of LiDS or LiFOS in the supernatant was determined by means of a Shimadzu IP-3A isotachophoretic analyzer. The concentration of NP which has an absorption band at 275 nm was determined by means of a Hitachi 200A UV spectrophotometer. The adsorbed amounts of surfactants on the alumina were determined from the difference in concentrations before and after the adsorption. The suspension prepared was placed in a flask into which pyrene ( 1 × 10 -6 ~ 1 × 10 -7 mole dm -3) dissolved in ethanol had been first introduced and the solvent was then allowed to evaporate. The steady-state emission spectra of pyrene in the suspension and supernatant were obtained using a Hitachi 650-10S fluorescence spectrophotometer. The excitation wave length of pyrene was at 335 nm. In order to keep the ionic strength constant, all the measurements were carried out in 10 mmole dm -3 lithium nitrate solution at 25 °C. The concentration of the alumina was 0.3g/10 cm 3. Nitric acid was used to adjust the pH of the suspension.
283 0021-9797/90 $3.00 Journal of Colloid and Interface Science, Vol. 134, No. l, January 1990
Copyright © 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
~E JO x
-O- L i F O S - N P ? . 5 (LiFOS) - 0 - Li FOS - NP?.5 ( NP 7.5 ) LiDS -NPT.5 ( L I D S ) 11r- LiDS
- NPT.5 ( N P ? . 5 )
Additive Concn. of NPT.5
/ m m o i - d m -s
"65 ~E O x
-(2)- L i F O S - N P 1 O - 0 - L i F O S - NP10 LiDS - NP10 - A - LiDS - N P I O
(LiFOS) ( NPI0 ) (LIDS) (NPI0)
O 02 Additive Concn. of
' 0.6 0 18 / m m o [ . d m -3
FIG. 1. Adsorbed amount of surfactants on alumina: (a) LiDS-NP 7.5 and LiFOS-NP 7.5 systems; (b) LiDS-NP 10 and LiFOS-NP 10 systems; (c) LiDS-NP 20 and LiFOS-NP 20 systems: each mark shows adsorbed amount of surfactant indicated in bracket in each system
RESULTS AND DISCUSSION When LiDS or LiFOS is adsorbed on positively charged a-alumina, the alumina flocculates. In this study, the optimum flocculation concentration for the alumina was 0.05 mmole dm-3 for LiDS and LiFOS at which the flocculated alumina showed almost a zero-zeta potential value and a maximum mean-particle size. This result implies that as Journal of Colloid and Interface Science, Vol. 134, No. 1, January 1990
the hydrophobic chains of LiDS and LiFOS are oriented outward, the alumina becomes hydrophobic and flocculation takes place. In order to study the interaction between an anionic hydrocarbon (LIDS) or an anionic fluorocarbon surfactant (LiFOS) and nonionic hydrocarbon suffactants (NP) with different oxyethylene chain lengths on the alumina, NP was added to the alumina previously flocculated by addition of LiDS or LiFOS (0.05 mmole dm-S).
~E 5 o
LiFOS LiFOS LiDS LiDS
- NP20 ( LiFOS ) - NP20 ( N P 2 0 ) -NP20(LiDS ) - NP20(NP20)
Concn, of N P 2 0
/ m m o t - d m -3
FIG. 1 - - C o n t i n u e d
Figure 1 shows the adsorbed amounts of LIDS, LiFOS, and NP on the alumina as a function of additive concentration of NP. In the LiDS-NP 7.5 and LiFOS-NP 7.5 systems as shown in Fig. 1 (a), the adsorbed amount of LiDS or LiFOS was nearly constant, while that of NP 7.5 increased with increasing the additive concentration of NP 7.5. These results suggest that LiDS or LiFOS adsorbing on the alumina, even with addition of NP 7.5, is not desorbed due to presumably strong electric attraction between a positively charged alumina and negatively charged hydrophilic group of LiDS or LiFOS. Upon addition of NP 7.5, the adsorption of this surfactant occurs by hydrocarbon to hydrocarbon or hydrocarbon to fluorocarbon chain interaction, resulting in a mixed bilayer with the hydrophilic groups (oxyethylene chain) of NP 7.5 oriented toward the liquid phase (4, 5 ). In addition, the adsorbed amount of
NP 7.5 is greater for the LiDS-NP 7.5 system than that for the LiFOS-NP 7.5 system. This result is understood by the view that the interaction between hydrocarbon and hydrocarbon is more favorable than that between hydrocarbon and fluorocarbon (8). Similarly, Figs. l ( b ) and (c) show that the adsorbed amount of LiDS or LiFOS is nearly constant, whereas that of NP 10 or NP 20 increases; the adsorbed amount of NP 10 or NP 20 in the LiDS-NP 10 or 20 system is also greater than that in the LiFOSNP 10 or 20 system. Further, it is noteworthy that in both systems of LiDS-NP and LiFOS-NP, the adsorbed amounts of NP are decreased with increasing oxyethylene chain length of NP. This is interesting since their isotherms essentially follow the adsorption of a nonionic surfactant
E c 400(
3000 I < .... N o n i o n i c s u r f a c t a n l
-O- LiFOS-NP?-5 LiDS- NP?.5 LiFOS-NP10 LiDS -NP10 -4D-- LiFOS- NP20 -~- LiDS -NP20
o c I000 o ®
Anionic s u r f d c t a n t
FIG. 2. A schematic model of mixed bilayer on alumina.
' o16 o ' 8 ' ,:o ' 0'.2 o 4 Additive Conch. of SurfQctQnt /rnmol .dm-3 FIG. 3. Change in mean-particle size of alumina for systems shown in Fig. 1. Journal of Colloid and Interface Science, Vol. 134,No. 1, January 1990
1.4 1.2 i
/ m r n o t ' d m -3
O O A
NP I 0
FIG. 4. Change in 11, ]3 ratio as a function of additive concentration ofNP: (a) LiDS-NP 7.5 (~, A) and LiFOS-NP 7.5 (©, O) systems; (b) LiDS-NP 10 (A, A) and LiFOS-NP 10 (©, O) systems; (c) LiDS-NP 20 (A, A) and LiFOS-NP 20 (O, O) systems. Open mark; supernatant, closed mark; suspension.
on a hydrophobic surface (9). A schematic model of a mixed bilayer is given in Fig. 2. Similar results were reported on adsorption of nonionic surfactants with different oxyethylene chain lengths on hydrophobic surfaces (10, 11 ). An increase in the length of the oxyethylene chain results in a decrease in the adsorbed amount of NP, presumably because the change in free energy ( - AG) of adsorption is decreased in magnitude and the cross-sectionM area of the molecule at the LIDS- or LiFOS-covered alumina may increase as the number of oxyethylene units is increased (12). The above results indicate that a decrease in the length of the oxyethylene chain of NP allows formation of the mixed bilayer to proceed more easily. The degree of redispersion in the flocculated alumina covered with a LiDS or LiFOS layer upon addition of NP was evaluated from the measurement of particle size. FigJournal of Colloid and Interface Science, Vol. 134, No. l, January 1990
ure 3 shows that the mean-particle size of the alumina suspension decreases with increasing concentration of NP and with increasing length of the oxyethylene chain of NP, for both the LiDS and LiFOS systems. It is also found that the mean-particle size of alumina for the LiDS system decreases with addition of a concentration of NP lower than that for the LiFOS system. Since NP is probably adsorbed on the hydrophobic chain of LiDS or LiFOS while orienting its hydrophilic chain toward the solution as shown in Fig. 2, this adsorption may produce a steric repulsive force due to the randomly coiled oxyethylene chains. This steric repulsion seems to increase when increasing the oxyethylene chain in NP. Similar experiments using three types of iron oxides have been reported by Meguro eta[. (13), and they have obtained a result similar to that of this study. Further, the zeta potentials of the
1.8 .o AA
• 0 Z~
1.2 • I
013 0.2 NP 20 I m m o l - d m - 3
FIG. 4--Continued alumina in a redispersion state were almost zero, although these results are not shown here, indicating that the repulsive forces between the redispersed alumina particles are not the electric forces but the steric repulsions (14). The degree of steric repulsions is thought to increase with increasing oxyethylene chain length. The pyrene fluoresence-fine structure has been found to be markedly dependent on the solvent: the intensity ratio of It/13 (11 first vibronic and/3 third vihronic bands at 373 and 383 nm, respectively) is sensitive to the solvent polarity ( 15 ). This fluorescence technique was applied to the suspension system in order to estimate the polarity of the mixed bilayer on the alumina. Figure 4 (a) shows that the 1~/13value in the LiFOS-alumina system in the absence of NP 7.5 is about 1.8 which is the same value as that in the supernatant; this ratio decreases steeply upon addition of 0.05 mmole dm-3 of NP 7.5 and then becomes constant (about 1.3) while the ratio in the LiDS-alumina system is nearly constant (1.2), even with different additive concentrations of NP 7.5. These results can be interpreted as follows. In the LiFOS-alumina system, pyrene is either solubilized into the high polarity part of LiFOS molecules in the absence of NP 7.5 or exists in water, and with addition of NP 7.5 the pyrene becomes solubilized into a hydrocarbon chain of NP 7.5 forming the mixed bilayer with LiFOS. On the other hand, in the LiDS-alumina system, in the absence of NP 7.5, pyrene is solubilized into the hydrocarbon chains of LIDS; with addition of NP 7.5, the pyrene is located in the mixed hydrocarbon chains of LiDS and NP 7.5. Further, the difference in the It/I3 ratio of the mixed bilayer between LiDS-NP 7.5 and LiFOS-NP 7.5 probably arises from the polarity difference between the hydrocarbon-hydrocarbon chain and fluorocarbon-hydrocarbon chain (16). The result that the It / 13 ratios for both systems are smaller in the suspension than in the supernatant may indicate that the hydrophobic chains between the surfactants are more compact in the
mixed bilayer than in the supernatant. This difference in the compactness would reflect the adsolubilization of water-insoluble compounds in the mixed bilayer and in the mixed micelles (17). Similar results of the It/I3 ratios are observed in NP 10 and NP 20 systems, as shown in Figs. 4 (b) and (c). The added concentration of NP to attain a constant It/13 value decreases with decreasing length of the oxyethylene chain in NP. This result corresponds with the change in the adsorbed amount of NP, i.e., a decrease in the oxyethylene chain length of NP increases the adsorbed amount of NP and enhances the formation of a mixed bilayer. The above results lead to the conclusion that the alumina previously flocculated by addition of LiDS or LiFOS redisperses by further addition of NP with different oxyethylene chain lengths, resulting in the formation of a mixed bilayer. Such mixed bilayers are formed more easily with decrease of the length of the oxyethylene chain in NP, but the redispersion of the alumina occurs more easily with increase of the length of oxyethylene chain in NP due to steric repulsive forces related to the oxyethylene chain. Further, it is found that the formation of a mixed bilayer of LiDS and NP proceeds more easily than that of LiFOS and NP, and that the polarity of the mixed bilayer of the former is much lower than that of the latter.
REFERENCES 1. Chandar, P., Somasundaran, P., and Turro, N. J., J. Colloid Interface Sci. 117, 31 (1987). 2. Levitz, P., Van Damme, H., and Keravis, D., J. Phys. Chem. 88, 2228 (1984). 3. Meguro, K., and Kondo, T., Nippon Kagaku Zasshi 76, 642 (1955). 4. Esumi, K., Ono, Y., Ishizuka, M., and Meguro, K., Colloids Surf 32, 139 (1988). Journal of Colloid and Interface Science, Vol. 134, No. 1, J a n u a r y 1990
5. Esumi, K., Sakamoto, Y., Yoshikawa, K., and Meguro, K., Bull. Chem. Soc. Japan. 61, 1475 (1988). 6. Esumi, K., Sakamoto, Y., Yoshikawa, K., and Meguro, K., Colloids Surf. 36, 1 (1989). 7. Meguro, K., Ueno, M., and Suzuki, T., J. Japan Oil Chem. Soc. 31, 909 (1982). 8. Mukerjee, P., J. Amer. Oil Chem. Soc. 59, 573 (1982). 9. Corkill, J. M., Goodman, J. F., and Tate, J. R., Trans. Faraday Soc. 62, 979 (1966). 10. Abe, R., and Kuno, H., Kolloid-Z. Z. Polym. 181, 70 (1962). 11. Wolf, F., and Wurster, S., Tenside 7, 140 (1970). 12. Rosen, M., "Surfactants and Interfacial Phenomena," p. 49. Wiley-Interscience, 1978. 13. Meguro, K., Tomioka, S., Kawashima, N., and Esumi, K., Prog. Colloid Polym. Sci. 68, 97 (1983). 14. Tadros, Th. F., in "Solid/Liquid Dispersions" (Th. F. Tadros, Ed.), p. 266. Academic Press, New York/London, 1987. 15. Nakajima, A., Bull. Chem. Soc. Japan. 44, 3272 (1971).
Journalof ColloidandInterfaceScience,Vol.134,No. 1,January1990
16. Muto, Y., Esumi, K., Meguro, K., and Zana, R., J. Colloid Interface Sci. 120, 162 (1987). 17. Esumi, K., Sakamoto, Y., Nagahama, T., and Meguro, K., Bull. Chem. Soc. Jpn., 62, 2502 (1989).
KUNIOESUMI1 YUICHI SAKAMOTO
KENJIRO MEGURO Department of Applied Chemistry and Institute of Colloid and Interface Science Science University of Tokyo Kagurazaka Shinjuku-ku, Tokyo 162 Received November 21, 1988; accepted April 27, 1989
i To whom all correspondence should be addressed.