JOURNAL OF COLLOID AND INTERFACE SCI.ENCE 205, 496- 502 (1998)
`ARTICLE NO. CS985721
`
`Comparing the Surface Chemical Properties and the Effect of Salts on
`the Cloud Point of a Conventional Nonionic Surfactant, Octoxynol 9
`(Triton X-1 00), and of Its Oligomer, Tyloxapol (Triton WR-1339)
`
`Hans Schott
`
`School of Pharmacy, Temple University, Philadelphia, Pennsylvania 19140
`
`Received March 3, 1998; revised June 5, 1998
`
`The surface-chemical properties, critical micelle concentra(cid:173)
`tions (CMC), and effect of salts on the cloud points (CP) of
`octoxynol 9 (Triton X-100) and tyloxapol (Triton WR-1339)
`were compared. The latter nonionic surfactant is essentially a
`heptamer of the former. Even though the molecular weight of
`tyloxapol is 7 times larger than that of octoxynol 9, its area per
`molecule adsorbed at the air-water interface is only twice as
`large. This suggests an unusual orientat ion for molecules of
`tyloxa pol at the surface and is in keeping with a plateau that is
`less horizontal and has a somewhat higher surface tension than
`the plateaus of most nonionic surfactants. The CMC of octoxy(cid:173)
`nol 9 was 4.4 times larger than that of tyloxapol. Unexpectedly,
`the CP of dilute aqueous tyloxapol solutions was 2s•c higher
`than that of octoxynol 9 solutions. The salting-out ions Na + ,
`Cl- and so~- lowered the CP of tyloxapol 29% more than that
`of octoxynol 9. However, because the blank tyloxapol solution
`started out with a higher CP value, its CPs in the presence of
`salts were higher than those of octoxynol 9. Pb2+ and Mg2 +
`cations salted both surfactants in, raising their CP, Pb2 + more
`extensively than Mg2+. 0 1998 Academic Press
`Key Words: cloud points of octoxynol 9 and tyloxapol; critical
`micelle concentrations of octoxynol 9 and tyloxapol; octoxynol 9;
`oligomeric nonionic surfactant; salt effects on cloud points of
`octoxynol 9 and tyloxapol; surface tensions of octoxynol 9 and
`tyloxapol; Triton X-100; tyloxapol.
`
`INTRODUCTION
`
`Tyloxapol (Triton WR-1339) is a nonionic surfactant whose
`study is of practical and theoretical interest. Its practical useful(cid:173)
`ness stems from the fact that it is official in the USP employed not
`only as a detergent in preparations for cleaning contact lenses but
`also as a mucolytic agent in preparations for treating pulmonary
`diseases (I , 2). It interacts with plasma lipoproteins (3, 4), which
`rules out its use in injectable preparations.
`Tyloxapol is essentially an oligomer of octoxynol 9 (Tri(cid:173)
`ton X-1 00). Comparison with its monomer is of physico(cid:173)
`chemical importance. The effects of polymerization on the
`solution properties of monomeric surfactants have not been
`investigated beyond their dimers (5) and trimers (6),
`
`whereas tyloxapol is essentially a heptamer of octoxynol 9
`(see below).
`The purpose of the present study is to compare the surface
`activity and the critical micelle concentration (CMC) of the
`two nonionic surfactants and their interaction with electro(cid:173)
`lytes. Such interaction is conveniently investigated by
`changes in cloud point (CP). Extensive data on the effect of
`electrolytes on the CP of octoxynol 9 have been published
`(7). The CP is the lower consolute temperature of nonionic
`surfactant solutions. It is a sensitive indicator of their inter(cid:173)
`action with additives.
`The practical importance of the CP lies in the fact that
`suspensions (8), emulsions (9), and ointments and foams (I 0)
`stabilized with nonionic surfactants become unstable when
`heated in the vicinity of the CP, e.g., during manufacturing,
`steam sterilization, or some end uses. On the other hand, the
`rate of solubilization by non ionic surfactant solutions increases
`near their CP ( 11 ).
`
`EXPERIMENTAL
`
`Materials
`
`Octoxynol 9 NF is p-octylphenol ethoxylated to an average
`p value of 9.5. The octyl moiety is I, I ,3,3-tetramethylbutane;
`i.e., it is an isobutylene dimer:
`
`~CH2CH20),H
`
`Octoxynol9
`
`CgH 17
`
`The molecular weight of octoxynol 9 is ~ 625. Subsequently,
`this compound is referred to simply as octoxynol.
`Ty1oxapol USP is made by treating an excess of octy1phenol
`with fonnaldehyde in the presence of an acid catalyst, which
`causes condensation polymerization via methylene bridges in
`the ortho position. The resulting novolac oligomer is then
`ethoxylated to an average p value of 9.6 ::!:: 0. 1 (I , 2):
`
`0021-9797/98 S25.00
`Copyright <0 1998 by Academic Press
`All rights of reproduction in any form reserved.
`
`496
`
`Innopharma EX1024, Page 1
`
`
`
`

`

`

`

`498
`
`HANS SCHOTI
`
`Cloud Point Measurements
`
`The surfactant- salt mixtures used for CP measurements
`were prepared by adding analyzed, concentrated salt solutions
`and water to 15.0% stock solutions of surfactant. All liquids
`were weighed out to the nearest milligram. The final surfactant
`concentration was 2.00% unless specified otherwise. The per(cid:173)
`centage is based on the weight of water present. The mixtures
`were aged ~ 24 h at room temperature in the dark prior to
`measuring their CP.
`CPs were measured visually while the solutions were blan(cid:173)
`keted with nitrogen, as described recently (7). The temperature
`interval between incipient and complete phase separation on
`
`heating was = I °C, as was the interval for the reverse process
`
`on cooling. The CP was taken as the temperature at which the
`immersed portion of the thermometer suddenly became invis(cid:173)
`ible on heating and fully visible on cooling. There was no
`hysteresis, and the six CP values observed on three successive
`heating and cooling cycles agreed within 0.2°C.
`
`RESULTS AND DISCUSSION
`
`Swface Tension and Critical Micelle Concentration
`
`The surface tension versus concentration data for tyloxapol
`and octoxynol are plotted in Fig. I , where the abscissa repre(cid:173)
`sents the natural logarithm of the surfactant concentration
`expressed as giL.
`The first linear segment, which extends from c = 0.004 to
`0.025 giL for tyloxapol, represents saturation adsorption. Its
`regression equation is
`
`y = 27.43 - 3.8905 In c
`
`(n = 6, r = -0.998).
`
`[2]
`
`The intermediate points at c = 0.0455 and 0.0683 giL fall in
`a transition region that may represent premicellar aggregation.
`The regression equation for the saturation adsorption region
`of octoxynol (c ~ 0.05 giL, before the shallow surface tension
`minimum) is
`
`-y = 9.59- 7.6 10 In c
`
`(n = 5, r = -0.999).
`
`[3]
`
`For tyloxapol, the regression equation for the second linear
`segment, which represents the plateau region and begins at c =
`0.075 giL, is
`
`y = 38.14 - 0.6022 ln c
`
`(n = II , r = - 0.977).
`
`[4]
`
`The regression equation for the plateau region of octoxynol
`after the shallow minimum, i.e., at c ~ 0.5 giL, is
`
`'Y = 30.44 - 0.176 In c
`
`(n = 4, r = -0.970).
`
`[5]
`
`For tyloxapol, the CMC is the concentration at which the
`
`two linear segments intersect and where Eqs. [2] and [4] are
`s imultaneous. The 22°C value is 0.0385 giL. The 25°C value
`obtained by replotting Fig. 3 of Ref. (4) is = 0.06 giL. The
`
`agreement between these two values is only fair.
`For octoxynol, the shallow surface tension minimum be(cid:173)
`tween the two linear segments requires a different approach
`( 12). Its CMC was taken as the concentration corresponding to
`the lowest surface tension because the surface tension mini(cid:173)
`mum is caused by traces of a poorly soluble, highly surface(cid:173)
`active fraction of low or zero degree of ethoxylation (13). As
`soon as octoxynol micelles begin to form, they solubilize this
`impurity, re moving it from the air-water interface and thereby
`causing the surface tension to rise. The surface tension mini(cid:173)
`mum is located at 0.16 giL. This value is in good agreement
`with the CMC of 0.18 giL determined by light scattering and
`dye solubilization ( 13).
`The more than fourfold ratio of the CMC values of
`octoxynol to tyloxapol is in keeping with the general obser(cid:173)
`vation that, as the molecular weight of a nonionic surfactant
`increases at constant hydrophilic- lipophilic balance (HLB),
`its CMC: decreases. This ohservation is illustrated hy com(cid:173)
`paring the CMC values at 25°C of two pairs of homoge(cid:173)
`neous polyo xyethylated normal primary alcohols C, EP hav(cid:173)
`in thei r hydrocarbon moiety, p
`ing n carbon a to ms
`oxyethylene units per molecule, and identical HLB values:
`0.072 M for C 6E4 {14) and (9.0 ± 1.9) X 10- 5 M for C 12E8
`(15, 16); 9.9 X 10- 3 M for C8E6 and 2.3 X 10- 6 M for
`C 16E 12 ( 17). Doubling of the surfactants' molecular weight
`decreased their CMC values 800- to 4000-fold.
`As expected, this effect is smaller for surfactants of higher
`molecular weight that are normally distributed, such as octoxy(cid:173)
`nol and tyloxapol. For instance, both C 12E 13_77 (MW = 792.9)
`and C 1 s~o (MW = 115 1.5) have HLB = 15.30. Their 25°C
`CMC values are 9 X 10- 5 M (interpolated) and 2 X 10- 5 M,
`respectively ( 17): a 45% increase in molecular weight reduced
`the CMC 4.5-fold.
`The surface properties of tyloxapol, illustrated in Fig. I,
`have the following three unusual features:
`
`(i) From Eq. [2], dy/d Inc = -3.890 5 dyne/em in the
`saturation adsorption region. According to Eq. [I ], the area per
`tyloxapol molecule in the air-water interface at saturation
`adsorption is 105 A2
`. This is merely twice the 54 A2 area of
`octoxynol calculated from Eq. [3]. The latter value agrees with
`the 55 A 2 area reported for a nonoxynol with the same degree
`of ethoxylation (p = 9.5) (18).
`(i i) The 22°C surface tension of tyloxapol at the CMC,
`40.1 dyne/em, is comparatively high. The 25°C surface
`tension of octoxynol beyond the shallow minimum, 31.5
`dyne/em, is more typical of the plateau surface tension of
`nonionie surfactants.
`(iii) The negative slope of the plateau surface tension region
`oftyloxapol is 3.4 times steeper than the more typical slope of
`octoxynol (compare Eqs. [4] and [5]). The temperature differ-
`
`Innopharma EX1024, Page 3
`
`

`

`COMPARING OCTOXYNOL 9 AND ITS OLIGOMER TYLOXAPOL
`
`499
`
`ence between 22 and 25°C is too small to account for these
`differences between tyloxapol and the typical nonionic surfac(cid:173)
`tant, octoxynol, to any significant extent.
`
`The comparatively small area per molecule of the oligomeric
`tyloxapol indicates an unusual molecular orientation at the
`air-water interface, such as U- or V -shaped instead of extended
`horizontally. The isooctyl chains would fill the inside of the U
`or V, squeezing out much of the water and attracting one
`another (hydrophobic effect), while the polyoxyethylene
`chains would be on the outside of the U or V in randomly
`coiled conformations, surrounded by water and fully hydrated.
`The proposed surface orientation also explains the other two
`unusual features in the surface properties of tyloxapol. The
`relatively high surface tension at the CMC results from a
`reduction in the interfacial area between water and the
`isooctane moieties as the molecules adsorbed at the surface
`bend to assume U or V shapes.
`The third unusual feature, namely, the comparatively
`steep negative slope beyond the CMC, results from a tight(cid:173)
`ening of the U or V shapes. As the bulk surfactant concen(cid:173)
`tration beyond the CMC is increased, the sides of the U- or
`V -shaped molecules are pushed closer together in order to
`make room for the adsorption of additional surfactant mol(cid:173)
`ecules at the air-water interface, in competition with their
`inclusion into micelles. This increases the deviation of the
`surface tension versus log concentration curve beyond the
`CMC from a horizontal plateau. The increased strain on the
`apex is partially offset by the increased hydrophobic attrac(cid:173)
`tion between opposing isooctane chains across the U or V as
`more water is squeezed out from inside.
`Similar unusual features, even more pronounced than those
`of tyloxapol, were reported for the surface properties of
`polyoxyethylene-polyoxypropylene-polyoxyethylene copoly(cid:173)
`mers of low molecular weight (poloxamers or Pluronics) (1 9).
`With molecular weights ranging from 1600 to 8000, their areas
`per molecule at saturation adsorption range from 64 to 146 A2
`.
`Their surface tensions at the inflection points on plots of
`surface tension versus log concentration are even higher
`than that of tyloxapol (50 ± 3 dyne/em compared to 40
`for tyloxapol), and the slopes of the approximately linear
`segments at higher concentrations are even much steeper
`than that of tyloxapol ( -dy/d In c = 13- 16 dyne/em com(cid:173)
`pared to 0.6 for tyloxapol).
`The following conformation was proposed for the polox(cid:173)
`amer chains adsorbed at the surface, based on the fact that
`"increasing the length of the hydrophobic polyoxypropylenc
`segment markedly decreases the area occupied by each mole(cid:173)
`cule. This suggests that the molecules are oriented in the
`surface in a coiled manner, with the polyoxypropylene segment
`out of the aqueous phase and the hydrophilic polyoxyethylene
`groups at both extremities of the molecules anchoring the
`polymer in the aqueous phase" ( 19).
`Such a conformation is compatible with a U or V shape,
`
`albeit an inverted one, where the coiled polyoxypropylene
`segment occupies the apex. Because the hydrophilic and
`hydrophobic moieties of poloxamers are more extensively
`segregated than those of the tyloxapol molecule and because
`the poloxamer chains are far more flexible and capable of
`forming random coils, one would expect their U or V shapes
`to be less distinct and more poorly defined than that of
`tyloxapol.
`The polyoxypropylene apex of the inverted U- or V(cid:173)
`shaped poloxamer molecules may be located above the
`aqueous phase. However, all of the U- or V -shaped tylox(cid:173)
`apol molecules are immersed inside the aqueous phase
`because the pendent polyoxyethylene chains are spaced
`evenly along their backbones. This and the greater flexibility
`of the poloxamer chains and the randomly coiled conforma(cid:173)
`tions of their polyoxyethylene and polyoxypropylene seg(cid:173)
`ments allow for greater compressibility of the surface layer
`at concentrations greater than the CMC, resulting in steeper
`negative slopes on the surface tension versus log concen(cid:173)
`tration plots than that of tyloxapol, which in tum is steeper
`than those of conventional polyoxyethylated nonionic sur(cid:173)
`factants.
`
`Cloud Points in Water
`
`The CPs of 0.50, 2.00, 3.50, and 5.00% tyloxapol solutions
`are 94.3, 93.8, 93.7, and 93.1 °C, respectively. The CPs of2.00
`and 4.00% octoxynol solutions are 65.5 and 65.6°C, respec(cid:173)
`tively.
`The following considerations lead to the prediction that
`tyloxapol should have a lower CP than octoxynol: The CP is
`the critical temperature of aqueous nonionic surfactant solu(cid:173)
`tions. At the 2.0% use level, both surfactants exist almost
`entirely in the form of micelles, whose molecular weights are
`in the range of polymers (see Table 1). Therefore, their solu(cid:173)
`tions should conform to the rules governing the phase equilib(cid:173)
`ria of polymer solutions (20).
`Polyoxyethylated non ionic surfactants, like polyethylene ox(cid:173)
`ides, are more water soluble at lower temperatures. Moreover,
`the micellar molecular weight oftyloxapol at room temperature
`is twice that of octoxynol. Therefore, on heating their aqueous
`solutions, the larger tyloxapol micelles should start to precip(cid:173)
`itate at a lower temperature than the smaller octoxynol mi(cid:173)
`celles. In the case of polyethylene oxides, the CP decreases
`with increasing molecular weight (21).
`However, contrary to the expected behavior, the CP of
`tyloxapol is 28°C higher than the C P of octoxynol. This
`discrepancy between the precipitation temperature of poly(cid:173)
`mers and the CP of nonionic surfactants is ascribed to the
`fact that the molecular weights of dissolved polymer mole(cid:173)
`cules are constant while the micellar molecular weights of
`nonionic surfactants increase with temperature. Apparently,
`as the temperature is increased, the micellar molecular
`weight of octoxynol increases faster than that of tyloxapol,
`
`Innopharma EX1024, Page 4
`
`

`

`

`

`

`

`502
`
`HANS SCHOTI
`
`II. Carroll, B. J., O' Rourke, B. G. C., and Ward, A. J. 1., J. Pharm. Phar-
`macal. 34, 287 (1982).
`12. Schott, H., and Han, S. K., J. Phann. Sci. 65, 975 (1976).
`13. Schott, H., J. Colloid Interface Sci. 173, 265 (1995).
`14. Schubert, K.-V., Strey, R., and Kahlweit, M., J. Colloid b11e1jace Sci. 14 1,
`21 (1991).
`15. Meguro, K., Takasawa, Y., Kawahashi, N., Tabata, Y., and Ueno, M., J.
`Colloid lnte1jace Sci. 83, 50 (1981).
`16. Rosen, M. J., Cohen, A. W., Dahanayake, M., and Hua, X.-Y., J. Phys.
`Chem. 86, 54 1 (1982).
`17. Becher, P., in "Nonion ic Surfactants" (M. J. Schick, Ed.), Chap. 15.
`Dekker, ew York, 1967.
`18. Hsiao, L., Dunning, H. N ., and Lorenz, P. B., J. Phys. Chem. 60, 657 (1956).
`
`., Luong, T. T., Florence, A. T., Paris, J., Vaution, C., Seiller,
`19. Prasad, K.
`M., and Puisieux, F., J. Colloid Interface Sci. 69, 225 (1979).
`20. Billmeyer, F. W., "Textbook of Polymer Science," 3rd. ed., Chap. 7D.
`Wiley, New York, 1984.
`21. Kjellander, R., and Florin, E., J. Chen1. Soc. Faraday Trans. I 77,2053
`(1981).
`22. Schott, H., and Han, S. K., J. Pharm. Sci. 65, 979 (1976).
`23. Schott, H., and Royce, A. E., J. Pharm. Sci. 73, 793 (1984).
`24. Schott, H., Royce, A. E., and Han, S. K., J. Colloid Interface Sci. 98, 196
`( 1984).
`25. Schott, H., Colloids Swf 11, 51 ( 1984).
`26. Schott, H., J. Colloid lnte1jace Sci. 43, 150 (1973).
`27. Schott, H., J. Colloid lnte1jace Sci. 192,458 (1997).
`
`Innopharma EX1024, Page 7
`
`
`
`