1058
H. W. Fox
ness to Sir Eric Rideal for the hospitality and helpfulness in connection with his brief stay a t the Col-
Vol. 61
laid Science Laboratory at the Cambridge University.
FORCE-AREA AND POTENTIALAREA RELATIONS OF MONOLAYERS OF TERMINALLY FLUORINATED OCTADECYLAMINE AND OCTADECANOIC A C I D 1
BY H. W. Fox2 Chemistry Division, U.S. Naval Research Laboratory, Washington 26,D. C. Received January IO, 1057
A study has been made of the force-area and otential-area relations of monolayers of CF3(CH2)&OOH and CF3(CH ),: NH2 spread on water a t tarious pH values. Tge curves obtained for these substances are compared with those for stearic acid and octadecylamine obtained on the same substrates. In contrast to the latter, the fluorinated compounds form films which do not support high pressures, and whose molecular areas a t closest packing are, under some conditions, too large to be accounted for by the extra bulk of the -CFa group. This behavior is attributed to intermolecular repulsions of t h e strong dipoles associated with the -CFs group attached to the hydrocarbon chain. The surface potentials of films of the terminally fluorinated octadecyl derivatives are higher than any yet reported.
Introduction It was of interest to examine the properties of amphipathic molecules whose principal chain was composed of -CH2- groups but which had a terminal -CF3 gsoup. When such molecules (for example, w,w,w-trifluorostearic acid) adsorbed, they might be expected to form a film nearly as resistant to penetration as that formed by stearic acid, for example, while a t the same time exposing a surface of -CF3 groups with the superior non-wetting qualities associated with the latter.3 Accordingly, a t the request of this Laboratory and with Office of Naval Research support, small quantities of CF3(CH2)&OOH (for convenience designated as "trifluorostearic acid") and CFB(CH2)17NH2(" trifluorooctadecylamine") were synthesized a t the Armour Research Foundation. The present paper describes the force-area and potential-area relations of monolayers of these substances on aqueous substrates. For comparison, parallel data are given for stearic acid and octadecylamine on the same substrates. Other papers deal with the adsorption, orientation and wettability of monolayers of these substances5a and their frictional properties.6b Materials and Procedures The trifluorostearic acid was a white crystalline powder melting a t 70.0-70.5' (cf. stearic acid, m a p . 69.4"). The trifluorooctadecylamine was supplied as the hydrochloride, a white crystalliiie powder which had been recrystallized six times from alcohol-acetone mixtures. The stearic acid used for comparison was Eastman white label grade and the oetadeeylamine was of 99.7 mole yo purity prepared by Dr. A. W . Ralston of the Chemical Division of Armour and Company. Both the stearic acid and the octadecylamine gave the correct limiting areas (20.4 A.2 per molecule) as condensed films. The trifluorooctadecylamine hydrochloride was dissolved in benzene containing 10% by volume of methanol; the trifluorostearic arid, stearic acid and octrtdecylamine were dissolved in benzene. The solvents used were tested for freedom from adsorbable impurities by (1) Presented at the 130th Meeting of tlm American Chemical Society, Atlantio City. N. J., September, 1956. (2) Office of Naval Research, Washington 25. D. C. (3) E. F. Hare, E. G.Shafrin and W. A. Zisman, THIS JOURNAL, 58, 236 (1954). (4) G. Gavlin and R. G. Maguire, J . Ow. Chem., 21, 1342 (1956). ( 5 ) (a) E. G . Shafrin and W. A. Zisman, THIS JOURNAL, 61, 1046 (1957). (b) 0. Levine and W. A. Zisman, ibid., 61, 1068 (1957).
spreading 0.5 ml. on the film balance. No detectable pressure was developed on reducing the area by a factor of 45. The solutions were applied to the water surface with a calibrated micropipet delivering 0.0489 ml. with a reproducibility of f0.02'%. Concentrations were adjusted to give condensed films occupying about 200 em.* of a total of 1400 available in the film balance. The film balance was a stainless steel trough, 100 em. long and 14 em. wide, coated with white, filtered paraffin, and enclosed in a glass and metal box with exterior controls for manipulating the barriers. The whole assembly was in a constant temperature room held a t 20 i= 0.2" and 50% relative humidity. A few runs were made with the trough equipped with a mica float controlled by a Cenco torsion head with a sensitivity of about 0.1 dyne/cm. The curves obtained in these runs were closely similar to those obtained subsequently on identical systems using the Wilhelmy dipping-plate principle with which most of the data were obtained. A quartz plate with a perimeter of 5.25 cm. and thickness of 0.069 em. was suspended from one stirrup of an analytical balance fitted with a mirror a t the central knifeedge. A scale placed to give an eight-foot optical lever er mitted reading changes in film pressure to 0.03 dyne)& To prevent vertical oscillations of the plate on changing the pressure, the balance was heavily damped, but this damping had no noticeable effect on sensitivity. Surface potentials were obtained with the vibrating electrode method of Zisman.6 A gold-plated brass electrode, 3 cm. in diameter, was attached to the diaphragm of an earphone and was driven at 60 cycles/sec. The resulting signal, amplified using a 60-cycle narrow-band-pass filter to reduce extraneous electrical noise, permitted reading the Volta potential to =!=3millivolts without bringing the electrode undesirably close to the water surface. Reproducibility between different potential-area (AV-A) runs was 510 millivolts indicating that the electrode was adequately stable under the experimental conditions. Potentials were converted to the vertical component of the apparent dipole moment (pn) by the Helmholtz equation AV = 47mp. where area.
7t
is the number of polar molecules adsorbed per unit
Results and Discussion Preliminary experiments on a hydrophil tray showed that neither trifluorostearic acid nor trifluorooctadecylamine formed solid films over a pH range of 2 to 13 a t pressures up to 30 dynes/cm. Additions of barium, ferric and thorium (+IV) ions to the substrate did not solidify the trifluorostearic acid film. At the highest pressures, the trifluoro(6) W. A. Zisman, Rev. Sci. Inalr., 3, 367 (1932).
August, 1957
1069
FORCE-AREA RELATIONS OF MONOLAYZRS OF OCTADECYLAMIN&
substituted compounds appeared to give viscous liquid films. Two reasons may be adduced to account for the failure of these films to solidify. (i) The -CFs group is about 40% bulkier than the cross-sectional area of a -CHZ- chain according to the Stuart-Briegleb atom models,’ thus causing steric hindrance t o optimum adlineation of the principal chains; (ii) intermolecular repulsions due t o strong dipoles associated with the terminal -CF3 groups. The presence of strong dipoles was confirmed by the AV-A curves described below. The force-area (F-A), AV-A, and apparent dipole moment-area (pn-A) curves for trifluorostearic acid and trifluorooctadecylamine a t various pH values are given in Figs. 1, 2 and 3. With each curve of the trifluoro compound, a curve is given for the analogous unsubstituted compound on the identical substrate. Thus direct comparisons can be made between the -CF3 and -CHa terminated molecules. Table I summarizes the properties of the compressed films giving the limiting area (specific molecular area derived from extrapolating to zero pressure the steep part of the F-A curve) when the steep part of the curve was linear; the specific molecular area at the collapse pressure when the steep part of the curve was non-linear; the state of the compressed film according to Harkins’ scale of compressibilities*; and the maximum Volta potential. Acid Films.-The curves given by the stearic acid are identical with those given in many places in the literature for this widely studied substaye. The limiting area on acid substrates is z4.0 Aa2/ molecule and on the basic substrate 20.4 A.2/molecule. The maximum potential decreases from 355 millivolts to zero millivolts with increa.se in pH. On alkaline substrates the acid is, of course, ionized and consequently the dipolar contribution to the Volta potential approaches zero. The films of trifluorostearic acid are greatly expanded relative to those of stearic acid. At p H values between 2.5 and 8.2 they are weak and unstable, collapsing a t relatively small pressures and large specific areas. The potential data show that, in general, the -CF3 group is oriented away from the water substrate. Since the carboxyl group presumably makes the same contribution to stabilizing the films of both stearic acid and trifluorostearic acid, it appears that the films of the latter are unstable as a consequence of the mutual repulsions of the strong dipoles associated with the -CF3 group. That it is not due solely to the size of the -CF3 group is shown by the large areas at which the films collapse a t pH 2.5 and 5.6. A t pH 8.2, the ionized carboxyl group provides a strong anchor for the trifluorostearic acid film, permitting compression to about 26 A.2/molecule before collapse sets in. At this pH, the trifluorostearic acid film is gaseous a t large arcas, exerting a pressure of 2 dynes/cm. even a t 300 A.2/molecule (the molecule lying flat would occupy a maximum area of about 130 Gaseous films exerting considerable pressure at large areas are characteristic of many substances containing more than one hydrophilic group.9 This (7) Chem. Eng. News, 52, 2534 (1954). (8) W. D. Harkins, “The Physical Chemistry of Surface Films,” Reinhold Publ. Corp., New York, N.Y., 1952, p. 135.
0.4
28
-
241 I
p H 5.6
O
i > m
w n
-0.46 v)
ti -0.8 -1.2
10 20 30 40 50 60 SQUARE ANGSTROMS PER MOLECULE. Fig. I.--F-A, AV-A and pn-d curves of stearic acid (shaded points) and trifluorostearic acid (open points) spread a6 monolayers on distilled water.
suggests that in contact with water at pH 8.2, the dipole associated with the -CFp group tends to become hydrophilic. In this connection it is worth noting that Henne and Foxlo consider the hydrogens on a carbon atom adjacent to a -CF3 group to be acidic. The -CF3 group is, however, not held in the water firmly but is pushed out to assume a more or less vertical position a t higher pressures as evidenced by the increasing negative potential and calculated apparent dipole moment per molecule as the film is compressed. The film is expanded at the highest pressures it can support as a consequence of two disrupting influences operating together: (i) the mutual repulsions of the dipoles associated with the -CF3 group at the far end of the molecule and (ii) the mutual repulsions of the charged dissociated carboxyl groups in the end buried in the substrate. The potentials associated with trifluorostearic acid films are strongly negative a t every pH of this study. On acid substrates, the maximum negative potential is diminished by the oppositely directed potential due to the un-ionized carboxyl group. At pH 8.2, the contribution of the carboxyl group is essentially zero, so that the potential of the whole molecule is that of the -CFs dipole alone. On this substrate the Volta potential reaches the remarkable value of - 1190 millivolts. It is difficult to compute the vertical component of the apparent dipole moment associated with the -CFa group since the carboxyl group contribution derived from the stearic acid curves may be subtracted only under conditions of packing and orien(9) N. K. Adam, J. F. Danielli and J . B. Harding, PTOC. Roy. SOC. (London),141,491 (1934). (10) A. L. Henne and C. J. Fox, J . A m . Chem. Soc., 76,5750 (1953).
H. W. Fox
1OGO
Vol. 61
TABLE I PROPERTIES OF
FILMS OF
STEARIC
Cornp o u n d
ACID, TRIFLLTOROSTE~4RIC-ACID, OCTADECYLAMINE SPREAD ON WATERAT pH 2.5, 5.6 AND 8.2 Limiting area, A.Z/molecule
Area a t collapse, A. a/molecule
TRIFLUORO~CTADECYLAMINE
AND
State of compressed film
Max. volta potential, mv.
pH 2.5 Hac( CHz)iaCOOH FBC(CHz)iaCOOH H~C(CHZ)LTNH~ F&(CHz)ii"z
24.0
..
..
43
28 34
liquid-condensed liquid-condensed liquid-condensed liquid-condensed
..
28 pH 5.6
HaC(CHz)i&OOH F&( CHz)i&OOH HaC(CHz)irNHz FaC( C H A T N H ~
24.0
H1C( CH2)iaCOOH F~C(CH~)NCOOH HLXCHzhJVHp
20.4
..
.. ..
solid liquid-condensed
38
.......
..
..
*.
-
355 705 635 240 280
- 1000 .. ..
........
pH 8.2 I
,
20.4
0
solid gaseous liquid-condensed liquid-condensed
*.
26 , .
4
.-
- 1190 -
pH 8.2
710 640
4
Av
t
w n
SQUARE ANGSTROM§ PER MOLECULE. Fig. 2,-R'-.4,
AV-.4 and pn--4 curves of stearic acid (shaded points) and trifluorostearic acid (open points) spread as monolayers on dilute sulfuric acid (pH 2.5) and on NaHCOa solution (pH 8.2).
tation identical with those of the trifluorostearic acid films. However, an approximate value may be obtained by comparing the values of pn a t equal molecular areas. When this is done, it is found that pn for the -CFa group in the compressed films is about 1.0 f 0.2 D . Townes, et a1.}I1measured the dipole moment of CH3CF3and found it to be 2.35 D or about twice that estimated from the compressed (11) C. H. Townea, R. G. Shulman and B. P. Dailey, Phys. Rev., 76, 472 (1949).
films of this study. But ikis well known that mutual polarization of neighboring dipoles will lower the measured value. Gerovich and Frumkin12 found 1.26 volts as the difference in surface potential between w-bromopalmitic acid and palmitic acid. They calculated that the potential should be 2.9 volts and attributed the discrepancy between the observed and calculated values to the mutual polarization of the C-Br linkages on close packing. ( 1 2 ) M. Gerovioh and A. Frumkin, J . Chem. Phys., 4, 624 (1936).
August, 1957
1061
FORCE-AREA RELATIONS OF MONOLAYERS OF OCTADECYLAMINE
28 I
0.8 0.4
$ a W D
O
b m
3 0
-0.4> -0.8
SQUARE ANGSTROMS PER MOLECULE. Fig. 3.--F-A, AV-A and p.-A curves of octadecylaniine (shaded points) and trifl.40rooctadecylumille (opcu points) spread as monolayers on dilute sulfuric acid (pH 2.5) and on NaHC03 solution (pH 8.2).
At every pH, pn remained essentially constant for the acid films over the region of molecular areas represented by the steep part of the F-A curves. Because of the change in mutual polarization with change in packing, however, the only conclusion that may be drawn as to the orientation of these films is that they are oriented "normally," Le., the carboxyl group is buried in the substrate and the -CF3 group is oriented more or less vertically away from the surface. Shafrin and ZismanSa indeed found that on adsorbing trifluorostearic acid films from solution on solid platinum, the films were less well oriented than stearic acid films, individual molecules departing as much as 20" from the vertical. This finding is entirely consistent with the results of the film spreading experiments reported here. Amine Films.-The films of octadecylamine give curves (Fig. 3) which resemble closely those seen in the classical surface chemical literature. As with the *acids described above, the films of trifluorooctadecylamine are enormously expanded relative to the unsubstituted analogs. The F-A curve of trifluorooctadecylamine at pH 2.5 contains an interesting example of a second-order transition from a liquid-expanded phase to what Harkins has callei intermediate-liquid phase.8 Thus, a t about 50 A.2/molecule the filmobecomes extremely compressible. At about 35 A.2/molecule, the film passes without discontinuity into a liquid-condensed state. The limiting area is 34 A.2/molecule and the film collapses a t 28 A.2/molecule. The difference in p,, between the films of oc-
tadecylamine and trifluorooctadecylamine a t 34 A.2/molecule is -0.59 D. This value is little more than one-half that found for the corresponding acids a t all pH values or the amines a t pH 8.2. It is difficult to see in detail why this should be so, or why for this particular system an intermediate-liquid phase exists. At pH 8.2, the trifluorooctadecylamine film is saseous, exerting a pressure of 2 dynes/cm. a t 200 A,2/mc$ecule. The film becomes liquid-condensed a t 29 A.2/molecul~and gives an unambiguous limiting area of 28.3 A.2/molecule. As with the analogous acid, the pressure a t large areas indicates that at pH 8.2 the end of the molecule containing the -CF3 group is hydrophilic. But again the -CF3 group is easily pushed out of the substrate and assumes a more or less vertical orientation as evidenced by the potential of -640 millivolts. The difference in pLnbetween the films of octadecylamine and the trifluoro-substituted analog is 0.94 D, in good agreement with the differences found for the acids. On this substrate, presumably, the molecules orient normally, Le., they assume progressively a more vertical orientation with increasing pressure, the -CF3 group pointing away from the substrate a t an orientation similar to that found for the trifluorostearic acid monolayers. Conclusion The substitution of a -CF8 group for a -CH, group a t the end of a long-chain hydrocarbon derivative results in expanding and weakening the films
1062
hl. J, SCHICKAND F. M. Fowiua
Vol. 61
relative to those of the unsubstituted analogs. In trifluoro-substituted compounds suggests that the general, the -CFa group is oriented away from the -CFa group, or perhaps the adjacent -CHz- group, aqueous substrate and contributes to the weaken- is weakly hydrophilic so that these molecules are ing of the film by virtue of its bulk and the mutual attached to the substrate at both ends. This hyrepulsions of the strong dipoles associated with it. drophilic character of the -CFa end of these moleThe limiting areas (where these could be estimated) cules is at a maximum on alkaline substrates, since or the areas at collapse for the condensed films are these are proton-accepting environments. The consistent with the 29 A.2/molecule found t o be the curves of the trifluoro-substituted compounds a t pH cross-sectional area of a -CFr chainla and withothe 8:2 resemble somewhat the curves found by Davindicated projected area of the -CFI group, 28 A.2/ ied4 for films of w-hydroxyheptadecanoic acid on molecule, as measured on the Stuart-Briegleb atom alkaline water. Acknowledgment.-Thanks are due to E. G. models.’ The large expansion undergone by the films of the Shafrin and W. A. Zisman for their interest and assistance during the course of this work. (13) C. H.Arrington, Jr., and G. D.Patterson, Tim JOURNAL,67, (14) J. T . Davies, Trans. Faraday Soc., 49,949 (1953).
243 (1953).
FOAM STABILIZING ADDITIVES FOR SYNTHETIC DETERGENTS. INTERACTION OF ADDITIVES AND DETERGENTS IN MIXED MICELLES BY M. J. SCHICK AND F. M. FOWKES Shell Development Company, Emeryville, California Receiued January 10, 1967
Addition of a small amount of certain non-ionic surface active agents to anionic detergents has been shown to enhance foam stability. The causes of the specificity and the structural relationship of effective combinations of additives and detergents were investigated by a study of their mixed micelles. The effect of additives on the critical micelle concentration of various detergents was determined in salt-free solutions a t 55” by the dye titration method. The additives which lowered the critical micelle concentration the most were in general the most effective for enhancing foam stability; these were compounds having an unbranched paraffin chain equal in length to the detergent, and with highly hydrophilic non-ionic polar groups at one end. It is concluded that the more effective foam stabilizing additives are solubilized into the palisade layers of detergent micelles or into surface films whereas the less effective additives are solubilized into the interior of the micelles. The incorporation of non-ionic additives into the palisade layer of mixed micelles lowers the activity of the detergent molecules and increases the attractive van der Waals forces between the molecules so that the micelles form a t lower detergent concentrations. Furthermore, i t is suggested that an analogy exists between the mode of packing in the palisade layers of the micelles and the surface layers of the foam lamellae.
Introduction An important aspect of the formulation of heavy duty synthetic anionic detergents is the addition of a small amount of a non-ionic surface-active agent which enhances foam stability and detergency of detergent solutions. Such additive-detergent combinations are lauryl alcohol with sodium lauryl sulfate, or the ethanolamide of lauric acid with sodium alkyl benzene sulfonates. The effect of foam stabilizing additives on the critical concentration for micelle formation (C.M.C.) in detergent solutions is used to observe the interaction of additives with detergents in the mixed micelles. The lowering of the C.M.C. by additives in salt-free detergent solutions has been correlated with foam stabilizing capacity of these additives in built detergent solutions. On the basis of this correlation general conclusions regarding the optimum structural relationship between foam stabilizing additive and detergent are presented. Experimental The detergents and additives used in this investigation were synthesized in this Laboratory. Certain dyes are known to change sharply in color a t the C.M.C. of detergent solutions.’ This fact is the basis for a convenient method for determining the critical micelle concentration by a titration whose end-point may be determined visually. (1) M. L. Corrin and W. D. Harkins, J . A m . Chem. Soc., 69, 679 (1947).
Pinacyanol dye (Eastman Kodak Co.) was used throughout this investigation. The disappearance of micelles, ;.e., the C.M.C., was determined by a distinct color change of the pinacyanol solution from blue to violet. The titrations were carried out in 100-ml. volumetric flasks, and the colors of sample and blank were compared through equal thickness of solution in the flask necks against a fluorescent light. The detergent was dissolved in aqueous pinacyanol solution and titrated with aqueous pinacyanol solution. Only freshly prepared dye solutions were useful, because the color faded on prolonged standing. The C.M.C. values of sodium dodecane sulfate were independent of dye concentration in the region 3 X 10-4 to 192 X 10-4 g./l. However, a pinacyanol concentration of 48 X 10-4 g./l. ( 1 X mole/l.) appears to render optimum conditions for visual observation. With the purpose of testing the accuracy of the dye method, a number of C.M.C. determinations of pure detergents and detergent8 in combination with additives have been repeated with the well established conductivity method. This comparison is illustrated in Table I. It is apparent that the dye method yields C.M.C. values which are 10 to 15y0 lower than those obtained by the conductivity method.2 However, the simplicity of the dye titration method appeared more suitable to perform the majority of the measurements. The foam stability tests were performed with 25 ml. of detergent solution contained in 100-ml. Pyrex glass-stoppered, graduated cylinders along with 2 sq. in. of standard General Dyestuff Corporation “Soil Cloth” cut into eight pieces of equal size. Solutions were made from a concentrated stock solution of detergent and sodium sulfate. Hard water (to give 350 p.p.m. Ca as carbonate) and sodium tripolyphosphate were added just prior to testing. The dilute solutions were added to a weighed amount of additive (2) P. Mukerjee and K. J. Mysels, ibid.. 77,2937 (1955).
