Thermodynamic Properties of Thin Films of Some Dipolar Liquids Adjacent to Fused Silica Surfaces BY K. H. ADLFINGER AND G. PESCHEL Institut fur Physikalische Chemie der Universitat Wurzburg 87 Wurzburg Markus- strai3e 9-1 1 Germany Received 31st March 1970 The disjoining pressure of some dipolar organic liquids (PhN02 PhCN PhCH,CN PhCF3 MeCN) forming thin layers between two fused silica surfaces shows maximum values in temperature ranges which refer to thermodynamic transitions of higher order. Here the molecules in the bulk liquid seem to gain an additional rotational degree of freedom about one axis whereas the molecules in the oriented surface zone still show a rotational restriction about this axis. Thermodynamic quantities for the liquid boundary layers are introduced and discussed.Since the experiments of Deryaguin l* evidence has been accumulated that a solid surface can alter the structure of an adjacent liquid layer 3-5 up to a thickness of about cm. This phenomenon is assumed to be caused by the solid surface inducing a molecular long-range orientation in the vicinal liquid. This orientation implies a decrease of the thermodynamic potential of the surface zone compared with that of the bulk liquid; it gives rise to the disjoining pressure. Thus if two solid plates immersed in a liquid showing long-range orientation approach to distances smaller than about cm under the action of an external force an oppositely directed dis- joining force arises holding the plates apart by a distance which is dependent on the external force. Knowing the geometry of the plates the disjoining pressure can be evaluated.Hitherto only few details are known about the mechanism of the long- range orientation which also alters the visc0sity,~9 ' the thermal conduction,8 and the dielectric constant of liquid surface The most frequently investigated liquid in this respect is water where the dis- joining pressure is of the order of magnitude of lo5 dyn/cm2.10* l1 The most inter- esting effect however is represented by the temperature dependence of the disjoining pressure. The present authors found significant maxima of the disjoining pressure of water at temperatures at which structural transitions of higher order occur.12* l3 These temperatures at about 14 32 45 and 61°C are repeatedly quoted and are discussed by Dro~t-Hansen.'~ By a special experimental device 7 * lo* l1 we succeeded also in determining the viscosity of aqueous boundary layers which show likewise maxima at the characteristic temperatures.Our interpretation agrees with that of Deryaguin,l viz. that small molecular aggregates determine the structure of aqueous boundary layers. The four maxima suggest the existence of four different molecular aggregates being stabitized in different temperature ranges. Applying the theory of viscous flow of Eyring l6 and determin- ing approximate values for the activation energy of viscous flow it can be assumed that the aqueous boundary layers above 14°C consist of small aggregates mainly built up of three or four water molecules. Above 32°C the aggregates are larger and above 89 90 THIN FILMS OF DIPOLAR LIQUIDS 45°C the size of the aggregates seems to be comparable with that of the molecu1ar entities existing above 14°C.Results for the temperature range about 61°C are not yet available because of experimental difficulties. All these effects get smaller with increasing plate distance and vanish at about 1.6 x cm. This value was also found by Deryaguin and coworkers ’* by measuring the shear modulus of the oriented aqueous surface zone. The results for the non-polar liquids benzene and cyclohexane indicate a steep decrease of the disjoining pressure with rising temperature starting from the melting p0int.l’ THERMODYNAMIC RELATIONS FOR ORIENTED SURFACE ZONES Regarding two surface areas each of 1 cm2 and separated by an intermediate oriented liquid layer the free excess energy for the approaching of the surfaces from a distance h+ to a distance h < h f is given by f h (AFEX = -J ndh’.h + II is the disjoining pressure which is dependent on h. h+ is the maximum distance at which the disjoining pressure can no longer be detected ; it is dependent on the sensi- tivity of the experimental apparatus. During the approach of the surfaces an oriented liquid layer with the thickness h f - h gets disordered and is pressed into the outer bulk liquid. A formal case is now considered. In order to press a liquid layer having infinit- esimal thickness and lying in the medium plane between the surfaces into the bulk liquid the free excess energy a(AFE)hdh d(AFE), = - -___- ah has to be brought into the system. This case however is not a practicable one and it is convenient to replace dh by Ah the last quantity being the thickness of a mono- molecular layer.Presuming that Ah<h is always valid the free excess energy to remove the monomolecular oriented liquid layer from the medium plane between the surface areas is a(AFE)h*h. A(AF~), = - - ah (3) The total excess energy to remove this layer can be evaluated from the Gibbs-Helm- holtz relation using d(AFE)h and its temperature derivative. Since the density of the oriented surface zone is-except for water-unknown conversion into molar quantities is only possible if the surface zone density is replaced by the density pe of the bulk 1 i q ~ i d . l ~ This however may imply large errors ; for the extreme case of aqueous surface zones the densities might exceed the bulk value by 20 or even 30 %.I8 The temperature dependence of II was assumed to have the form n = Cexp(-nh) (4) with C and n being constants.K . H . ADLFINGER AND G . PESCHEL 91 Using eqn (4) and taking molar quantities the total excess energy is obtained in the form T If the temperature dependence of A(A U,") is accessible the corresponding excess heat A(AC,",,,) can be evaluated. (5) molar EXPERIMENTAL For the determination of the disjoining pressure a spherically formed and a planar fused silica surface both highly polished totally covered with hydroxyl groups and immersed in the liquid to be investigated were pressed against each other. This system was chosen to avoid the problem that dust particles in the region between the plates appreciably falsify the measurements. For two plane-parallel surfaces this problem seems to be a serious one as Hayward and Isdale l9 have pointed out.In our experimental device the planar fused-silica plate is mounted to the bottom of a container which can be filled with the liquid in question. The spherically formed plate facing the planar one is fastened to one end of a balance which can be operated electro- magnetically. Above the spherically formed plate the measuring pin of a displacement transducer is pressed upon the balance with a definite force. The displacement transducer is connected with an electronic strain gauge measuring bridge connected to a xy-recorder. All deflections of the balance can thus be registered with high acc~racy.~ lo* The distance between the plates can be determined by a rather complicated method which takes sufficient account of the surface roughness derived from the surface profilograms and interferograms of the polished fused silica plates.ll9 2o The whole apparatus is placed in a big container which can be evacuated.In this way contact of the liquid with the humidity of the surrounding air can be avoided. At lower temperatures this precaution becomes important. The distilled and carefully degassed liquid can be sucked into the evacuated container so that the formation of air bubbles can be excluded. The temperature of the liquid in the container can be registered by a special recorder. Mechanical vibrations of the apparatus due to external causes and which disturb the measurements can be reduced by a special suspension. From a theory given by the present authors the parameters n and C in eqn (4) can be evaluated. RESULTS The liquids chosen for investigation are nitrobenzene benzonitrile benzyl cyanide benzotrifluoride and acetonitrile.By applying eqn (4) for different plate distances ranging between crn and plotting II against the temperature fig. 1-3 were obtained. It is still uncertain if for all liquids investigated in this work the oriented surface zone really extends beyond the maximum distances chosen in fig. 1-3 or abruptly turns into the bulk phase at some shorter distance. This problem could not be elucidated previously since our apparatus lacks the sensitivity to yield sufficiently precise results. For the aromatic compounds benzotrifluoride shows the largest disjoining pres- sure acetonitrile an example of an aliphatic dipolar liquid exhibits a still larger effect. The most striking evidence however is the fact that the disjoining pressure seems to assume large values only in distinct temperature ranges lying about 30" above the respective melting point T,, of the compound under consideration.and 5 x 2t a //A- *\ 15 20 25 5 Y 10 15 t "C IC N e -. FIG. 2.-Temperature dependence of the disjoining pressure II for PhCF3 (a) and PhNOz (b) at different plate distances h IO-~CM 0 ; 2x10-6cm A ; 3 x 1 0 - ~ cm V. 2 2 - 4 E 5 FIG. 1 .-Temperature dependence of the disjoining pressure II for PhCN (a) and PhCH,CN (b) at different plate distances t i lo-" cm o ; 2 x cm v ; 4x cm 0. cm A ; 3 x b 35 4 0 t "C K . H . ADLFINGER AND G . PESCHEL 93 - 10 - 5 t “C FIG. 3.-Temperature dependence of the disjoining pressure II for MeCN at different plate distances h 1OV6cm 0; 2~10-~crn A; 3~10-~crn V ; 4~10-~crn 0; 5~10-~crn 0.TABLE PA PARAMETERS n AND C n x 10-6 c x 10-6 liquid temp. (“C) (cm- 1 ) (dyn/cm2) nitro benzene 36.5 38 39.5 40 benzonitrile 16 19.5 22.5 25.5 7.5 10.5 13.5 benzo trifluori de 2.7 4 7 9.5 acetonitrile - 14.5 -11.5 -8 benzyl cyanide 4.5 1.78 1.66 2.74 4.79 1.43 1.01 0.86 0.96 2.00 1.34 1.39 1.66 1.36 1.15 0.87 1.87 1.16 0.91 1.41 2.47 2.21 1.81 1.22 1.59 0.90 0.69 0.70 1.75 2.09 1.85 1.57 3.38 3.10 2.54 2.83 5.73 4.99 8.26 94 THIN FILMS OF DIPOLAR LIQUIDS In fig. 4 A(AU,") obtained by eqn (5) is plotted against the temperature difference T-T for a plate distance of 2 x For all liquids considered in this paper A(AU,E)L exhibits extraordinary properties in the temperature ranges in question ; thus it changes its sign from negative to positive with rising temperature.The corresponding mean values of the molar heat A(AC;,,) derived from the ascending and descending branches of the curves in fig. 4 respectively are given in table 2. cm. T- Tm FIG. 4.-Behaviour of the total excess energy A(AUZ)h in the reduced temperature range PhN02 x ; PhCN ; PhCH,CN 4 ; PhCF3 + ; MeCN A. TABLE 2.-A(ACvF;m)h FOR h = 2x cm liquid for the ascending for the descending branch branch cal mol-1 K- 1 nitro benzene 42 - 41 benzoni trile 2 benzyl cyanide 9 benzotrifluoride 41 acetonitrile 11 The values for A(AUE), and those for A(AC,E,,)h imply large errors which may exceed 50 %. K . H . ADLFINGER AND G . PESCHEL 95 DISCUSSION There is still no satisfactory theory to explain the magnitude of the disjoining pressure found by us and other author^.^ The disjoining pressure calculated on the basis of long-range dispersion forces extending from the solid surfaces is thought to have much smaller values.O 9 In a former paper l7 the present authors introduced the concept of molecular rotation restriction in oriented surface zones. For benzene e.g. one may imagine the molecules to be strongly adsorbed to the surface hydroxyl groups and oriented with the C,-axis perpendicular to the surface rotational movement being only pos- sible about this axis.22 This adsorption effect decreases the intermolecular distance to any vicinal molecular layer compared with the conditions in the bulk liquid in which a benzene molecule can rotate almost freely about all three axes.23 The decreased distance on the other hand gives rise to an excess dispersion interaction causing the disjoining pressure.In this way many molecular layers can be built up but with increasing distance from the surface rotational vibrations about the other two axes will become important and diminish the stabilizing excess interaction. If this concept is correct a disjoining pressure should exhibit a maximum value in a temperature range where the liquid in question is known to have a rotational transi- tion ; because for rotational freedom in the bulk liquid and rotational restriction of one molecular axis e.g. in the surface zone relatively large differences in the chemical potential i.e. A(AFE),, are expected. By raising the temperature the oriented surface zone suffers destruction and the disjoining pressure vanishes.Below the transition range rotational restriction prevails in the surface zone as well as in the bulk liquid and A(AF,"), and IT respectively should become small. Indeed our results seem to confirm the correctness of this concept since the temperature ranges found for the liquids investigated are in accordance with those revealing anomalous bulk properties which can be regarded as evidence for a rota- tional transition in liquid phase.23 Regarding the viscosity of nitrobenzene and benzonitrile respectively the plot of log q against 1 /T (fig. 5 ) shows deviations in the temperature range in question. Moreover the activation energy of flow seems to show an abrupt change in this range which likewise points to an alteration of molecular rotational behaviour. The viscosities were determined by a capillary viscometer over only small temperature steps.Our data for nitrobenzene agree with those found in l i t e r a t ~ r e ~ ~ which show the same deviations as cited above. For acetonitrile values of the molar heat C are available.25 The temperature derivative of C against T (fig. 6) indicates an anomaly in the temperature range as found in fig. 3. Work is in progress to obtain accurate evidence for the abrupt change in molecular rotational behaviour of these liquids. For halogenobenzenes we succeeded in showing that in the corresponding transition ranges likewise lying about 30" above the respective melting points the change in rotational behaviour refers to the molecular C2-axis. For these considerations we applied the method of Davies and Matheson 2 3 and cal- culated the rotational volume of a molecule in the gas phase and the space available for the same molecule in the condensed liquid phase at the characteristic temperature.2G It is difficult to explain why the disjoining pressure differs in its magnitude and its temperature dependence for the four aromatic liquids quoted in this paper.How- ever the disjoining pressure is dependent on the structure of the oriented surface zone as well as on the structure of the bulk liquid which is usually unknown. Investi- gations of other dipolar aromatic liquids to be published later revealed a tendency for inolecules with large functional groups (e.g. PhCH2CN) which give rise to steric hindrance in the molecular movement to show marked maxima of the disjoining 96 - 1*7 -1.8 F M 0 - -1.9- -29- THIN FILMS OF DIPOLAR LIQUIDS 3.2 3.4 3.6 I I I I I I I + (1) + - + + + + + (21 + + 1 ++ - + + + + + ++ + + I++$ + + + + + + + + + + + + + 1.0 I I I I I 1 I 2 4 0 2 5 0 260 2 7 0 280 2 9 0 300 T OK FIG.6.-Temperature dependence of dC,/dT for MeCN. The arrow indicates the centre of the transition range found by the disjoining pressure. K . H . ADLFINGER A N D G . PESCHEL 97 pressure whereas molecules with small functional groups (e.g. benzonitrile) exhibit less marked maxima especially at small plate distances. Regarding the excess energies A(A Ug), < RT the orientational molecular packing effect in the surface zone cannot be pronounced except perhaps for those molecules with great rotational restriction. The values of A(AC&), are rather high because of the presence of a transition range.We are indebted to Prof. Dr. G. Briegleb and the Deutsche Forschungsgemeinschaft for support of tlus work. Thanks are due to Mrs. G. NO11 for performing the precise viscosity measurements. B. V. Deryaguin 2. Phys. 1933 84,657. B. V. Deryaguin and E. Obuchov Actaphysicochim. 1936,5 1. J. C. Henniker Rev. Mod. Phys. 1949,21,322. B. V. Deryaguin Disc. FaradQy SOC. 1966,42 109. A. Sheludko Colloid Chemistry (Elsevier Publ. Co. Amsterdam 1966). B. A. Kholodnitskii Vestn. Leningrad Univ. Fiz. Khim. 1968 23 153. ' G. Peschel and K. H. Adlfinger Z. Naturforsch. 1969 Ma 11 13 ; Ber. Bunsenges. phys. Chem. 1970 74 351. M. S. Metsik and 0. S. Aidanova Research in Surface Forces ed. B. V. Deryaguin (Consultants Bureau New York 1966) vol. 2 p. 169. G. Peschel and R. Schnorrer in preparation.lo G. Peschel Z. phys. Chem. (N.F.) 1968 59,27. l 1 K. H. Adlfinger and G. Peschel 2. phys. Chem. (N.F.) in press. l2 G. Peschel and K. H. Adlfinger Naturwiss. 1967,54,614 ; Chem. Labor Betrieb 1970,21,193. l3 G. Peschel and K. H. Adlfinger Naturwiss. 1969,56 558. l4 W. Drost-Hansen Ind. Eng. Chem. 1?69 61 ( l l ) 10. l5 B. V. Deryaguin I. G. Ershova V. K. Simonova and N. V. Churayev Teor. Eksp. Khim. 1968,4 l6 S. Glasstone K. J. Laidler and H. Eyring The Theory ofRate Processes (McGraw-Hill New l7 G. Peschel and K. H. Adlfinger Z. phys. Chem. (N.F.) 1969 63 150. l9 A. T. J. Hayward and J. D. Isdale Brit. J. Appl. Phys. (J. Phys. D) 1969 2 251. 2o K. H. Adlfhger R. Schnorrer and G. Peschel Z. angew. Phys. 1970 29 136. 21 J. Frenkel Kinetic Theory of Liquids (Dover Publ. Inc. New York 1955). 22 A. V. Kiselev and D. P. Pashkus Dokl. Akad. Nauk. S.S.S.R. 1958,120,843 ; D. Michel Z. 23 D. B. Davies and A. J. Matheson Disc. Faraday SOC. 1967 43 216. 24 B. P. Nikolski Handbuch des Chemikers (VEB Verlag Technik Berlin 1956) vol. 1 . 25 W. E. Putnam D. M. McEachern Jr. and J. E. Kilpatrick J. Chem. Phys. 1965 42 749. 26 G. Peschel and R. Schnorrer Ber. Bunsenges. phys. Chem. 1969,73,917. 527. York 1961). * B. V. Deryaguin Z. M. Zorin and N. V. Churayev Dokl. Akad. Nauk. S.S.S. R. 1968 182,811. Naturforsch. 1968 23a 339. SP1-D