The Structure of Surface Films. 105 Results. Fig. 1 gives the results on the 18, 20, and 22 carbon nitriles, the last two of which form condensed films only at the temperature of experiment. Fig. 2, on the 17 carbon nitrile, which forms a liquid expanded film and also shows the transition to the condensed state. Fig. 3 exhibits the values for the 14 and 16 carbon nitriles, which form liquid expanded films only. The curves show the smface pressure, F, the surface potential change, Ll V, due to the film, and µ, calculated from the equation LlV = 4miµ (1) plotted against A, the area per molecule, and n, the number of molecules per square centimetre in the film. µ is related to the vertical component of the dipole moment of the molecules in the film ; its meaning will be discussed later in this paper. The surface-pressure area curves agree well with previously published results. The condensed films tend, at zero compression, to an area of 27 sq. A., and their films are moderately compressible, indicating some re-arrangement, or possibly some compression of the polar end groups, on compressing the film. The large size of the end groups is no doubt due to the great amount of space required by the three valencies between carbon and nitrogen -CN which start out from the terminal carbon atom at angles of about 109½ 0 to the bond joining the end carbon to the next one in the chain. The carbon and nitrogen, if singly bound, would not occupy more than the usual area (20·4 sq. A.) for hydrocarbon chains closely packed, so that the area of 27 sq. A. must be ascribed to the bulge formed by the three linkages at the end of the molecule. The surface potential, in the wholly condensed films of the 20 and 22 carbon nitriles, was constant within 10 millivolts, below 27 sq. A., at which area the film first becomes closely packed. From 27 to 25 sq. A. the value of µ was constant, falling slightly on compression to smaller areas. At larger areas than 27 sq. A. the surface was very heterogeneous, the fluctuations amounting to hundreds of millivolts ; this indicates without doubt that there are two surface phases present, namely, coherent islands of condensed film, in equili brium with vapour film. Stearic nitrile appears, from the surface pressure curve alone, to be nearly completely condensed ; below 8 dynes per centimetre, however, the areas .
N. K. Adam and J.B. Harding. 106 are slightly larger for a given pressure than for the longer chain nitriles. Between 26 and 30 sq. A. there are fluctuations of the order 50 mv., i.e., similar to those observed in the transition region between condensed and liquid expanded films with the 17 carbon nitrile. It appears, therefore 30 nxJOM 4-5 3·5 5·0 800 µ-A, F C20,C22 --4-0~ s X ;j rn ai 20 3·0 :j_ 700 .V.J. E 0 u -~ . . i. Q) s $1. rn ~ (sl:): >, ~ A 10 600 + + + ++ F-A, C20,~2·.,,.2- +---i at 18 C. FIG. !.-Condensed nitrile films. that this film is not completely condensed below 8 dynes per centimetre. Even at higher pressures the potential is some 50 mv. less than for the two higher homologues; it appears doubtful therefore if stearic nitrile is com pletely condensed at any pressure. .
The Structure of SU?face Films. 107 The 17 carbon nitrile, margaric, forms an expanded film at 18 · 9° C. at pressures below 5 dynes per centimetre; at this pressure and 34 sq. A. there is the usual sudden change in direction of the surface pressure curve, characteristic 650 a> '9 =X 30 2·0 U) 550 Q) + + :::L rn ~ I .0:: -;, 20 450 E ~ E C: u \ 7 I. ~ Q) + 0. + tf) + Q) + C >, A + + 10 1-------l--l.------1-- - -+= ----1----- -4--- ---+------j3 50 ,/4 + '\. '<I 0 ___J__:____:=:::::::1:====1250 ~--_JL.__ _ __[_ ___ __j__ __ 20 30 40 sq.A 50 FIG. 2.-Margaric nitrile. of the commencement of the transition from expanded to condensed film. The potential curve also shows a fairly abrupt change in direction at this area ; and also, at the same area, marked fluctuations in potential appear, amounting .
N. K. Adam and J.B. Harding. 108 to 50 or 70 mv. At areas between 34 and 46 sq. A. the fluctuations are certainly not more than 10 mv. This is a perfectly clear example of inhomo geneity of the surface in the transition region between expanded and con densed film; the inhomogeneities persist down to 24 or 25 sq. A., so that it seems that the film is not completely condensed even at this area and a pressure of 35 dynes per centimetre. The limiting area, at low compressions, of the expanded film is about 46 sq. A. ; above this area very large fluctuations -n xl011 4·0 3·0 2·0 4·0 500 30 F o-. 79 " ::i . 20 2-0 0 '100 E <l.l .0.0. 0 C.) -> ~ <l.l i::,. E "' C: <l.l C >: >-. i:::i,10 0 300<] 0L------1_ _ -----L.==='===='====-200 20 __3j_0_ __ _J__ __ __'1_j0__ _ 50 sq. A. 60 Fm. 3.-Expanded nitrile films. of potential commence, indicating that there are both liquid expanded and gaseous films present. µ decreases slightly with decreasing area in the expanded film, in the transitional region there is a decided rise in the average value of µ as the area is diminished. The 14 and 16 carbon nitriles, fig. 3, give liquid expanded films; the 16 carbon having a limiting area of 47 sq. A., and the 14 carbon of about 50 sq. A. The potentials were uniform within 10 mv. below this limiting area ; above this area large fluctuations were found, the value of 267 mv. persisting in patches some distance beyond 47 sq. A. with palmitic nitrile; while with .
The Structure of Surface Films. 109 myristic nitrile, though it was more difficult to obtain coherent islands of 230 mv. of any considerable extent, the potential fluctuated between about this value and practically nothing. There is less lateral adhesion between the molecules in films of myristic nitrile than with palmitic nitrile. The surface pressure curves of Adam and Jessop* are not conclusive as to whether myristic nitrile is a liquid expanded (coherent) film or a vapour expanded film; it is now clear, from the magnitude of the fluctuations observed just above 50 sq. A., that the film of myristic nitrile is just below the critical point at which the liquid expanded film ceases to exist, not just above that point, as was surmised from the earlier measurements of surface pressure. A re-determination of the surface pressure curve at large areas now indicates that there is a region of ± constant surface vapour pressure, extending from 50 to 450 50 sq. A., ± having a pressure of O· 39 O• 02 dynes per centimetre. µ. decreases slightly with decreasing area, just as with the expanded part of the film of margaric nitrile. Discussion of Results. (1) There is a marked discontinuity in the surface potential curve with margaric nitrile, where the expanded film ceases and the transition region to condensed film commences ; at the same time the value of µ. begins to rise sharply, and the film is markedly inhomogeneous in the transition region. In the previous work with myristic acid we failed to notice any sharp change at this point, or fluctuations in the transition region except when there was a probability of collapse of the film having occurred. On the other hand, Schulman and Rideal, and later Schulman and Hughes, found small fluctuations in the transition region with myristic acid and believed this transition to be heterogeneous. Since, with myristic acid, the value of µ. does not change appreciably on passing from the expanded to the condensed film, the smallness of the fluctuations in the transition region does not necessarily mean that the film is homogeneous. With margaric nitrile the :film is clearly inhomogeneous in the transition region. The surface pressure is never constant during a transition from expanded to condensed films; a great many such transitions have now been investigated. In the only other type of transition in the films, where definite evidence of two types of co-existent film has been found, namely, the transition from coherent to vapour films, the surface pressure is constant. Reasoning by analogy with heterogeneous equilibria in three dimensions, and regarding the material of the * 'Proc. Roy. Soc.,' A, vol. 110, p. 428 (1926). .
N. K. Adam and J. B. Harding. llO spread film as the only component present, one might expect a two-dimensional analogue of the phase rule to hold ; i.e., there should be a constant surface pressure wherever there are two co-existent types of film. Such an analogy has the wealmess, however, that the relative numbers of molecules of water and of the spread material in the film are varying during compression; it appears incorrect to apply a two-dimensional analogue of the phase rule in its simplest form, considering the spread material as the only component, and the surface pressure as the analogue of vapour pressure. (2) Interpretation ofµ.-µ, calculated from equation (1), is often referred to as the " vertical component " of the dipole moment of the film per molecule. If the film could legitimately. be considered as a parallel plate condenser, having the upper ends of the molecular dipoles as the upper, positively charged sheet, and the lower ends as the lower (negative) sheet, and if the dielectric constant of the space between these sheets could be ta.ken as unity, then µ would be the vertical component of the dipole moment of the molecules in the film. An average value of the total dipole moment of the nitrile group is about 35 X 10-19 e.s.u.*; µ varies between 2·4 and 4·6 X 10-19_ Thus the maximum value of µ is only about one-seventh of the total dipole moment of the nitrile group ; it is always found that µ is much less than the dipole moment of the polar groups in a film on water. For the nitriles, µ increases considerably from the expanded to the condensed states, and it seems probable that this increase is due to the dipoles of the nitrile groups becoming oriented more nearly per pendicular to the surface. Below about 27 sq. A., however, when the films are ~ondensed, µ is nearly constant; the slight fall at areas less than 25 being possibly not real, but due to errors in the calculation caused by neglecting ~ollapse of the films under high pressure. If we assume that the dipoles are t perpendicular to the surface in the condensed films, where µ is nearly constant and a maximum, the dielectric constant in the films is seen to be about 7. This is of the order one-eleventh of the dielectric constant of water, a reasonable value, perhaps, when it is considered that the " dielectric constant " of the principal space intervening between the ends of the dipoles, namely, the sub stance of the molecules forming the film, could not appear in this calculation ; and the greater part of the dielectric constant of the film must be due to a re orientation of water molecules lying close to and between film molecules, under the influence of the oriented dipoles of the film molecules. * Smyth, "Dielectric Constant and Molecular Structure," Appendix I. t This assumption cannot be substantiated at present, but if the dipoles are not quite vertical, there will still probably be quite a large dielectric constant in the films. .
The Structure of Surface Films. lll The difficulties in a quantitative interpretation of the surface potential in terms of the dipoles of the film molecules and their orientation have been well indicated by Frumkin and Williams*; we feel that there is some justification for supposing that, as µ increases, the axis of the dipoles usually becomes more nearly perpendicular to the surface, but a thorough quantitative treatment of the subject seems lacking at present. One of us (J. B. H.) thanks the Department of Scientific and Industrial Research for a grant. Summary. Surface potential measurements on monomolecular films of long chain nitriles indicate a maxim.um value of µ, calculated from the simple Helmholtz equation, only about one-seventh of the dipole moment of the nitrile group as obtained by other methods. In the nitrile group there is no possibility of modification of the value of the dipole by internal rotation of the different parts of the polar groups, hence the effective " dielectric constant " of the sur roundings of the film molecules is probably of the order 7. The transition between condensed and expanded films is unquestionably heterogeneous with the nitriles. * ' Proc. Nat. Acacl. Soi., Wash.,' vol. 15, p. 400 (1929). .