Analysis of The Effects of Marangoni Stresses PDF

3972
Langmuir 2005, 21, 3972-3980
Analysis of the Effects of Marangoni Stresses on the Microflow in an Evaporating Sessile Droplet
Hua Hu* and Ronald G. Larson
Department of Chemical Engineering, University of Michigan, Ann Arbor, Michigan 48109-2136 Received October 6, 2004. In Final Form: January 25, 2005
We study the effects of Marangoni stresses on the flow in an evaporating sessile droplet, by extending a lubrication analysis and a finite element solution of the flow field in a drying droplet, developed earlier.1 The temperature distribution within the droplet is obtained from a solution of Laplaces equation, where quasi-steadiness and neglect of convection terms in the heat equation can be justified for small, slowly evaporating droplets. The evaporation flux and temperature profiles along the droplet surface are approximated by simple analytical forms and used as boundary conditions to obtain an axisymmetric analytical flow field from the lubrication theory for relatively flat droplets. A finite element algorithm is also developed to solve simultaneously the vapor concentration, and the thermal and flow fields in the droplet, which shows that the lubrication solution with the Marangoni stress is accurate for contact angles as high as 40. From our analysis, we find that surfactant contamination, at a surface concentration as small as 300 molecules/m2, can almost entirely suppress the Marangoni flow in the evaporating droplet.
1. Introduction Marangoni effects, manifested as tears of wine, were observed as early as the 1800s.2 In a wine glass, the evaporation of alcohol generates a surface tension gradient, which produces a traction on the wine surface causing the wine to climb up the side of glass where it forms a thin film. As the wine accumulates, a bulging rim of liquid forms along the top of the film, which eventually pinches into droplets which roll under their own weight, like tears, back into the wine. The Italian physicist Marangoni gave a detailed description of the movement of a liquid surface induced by a surface tension gradient, generated either by a composition or a temperature variation along the free surface. In 1901, Be nard3 discovered the convection cells in a thin liquid film that were later named after him. Later, Block4 performed more careful experiments on a thin liquid film, and Pearson5 gave a detailed theoretical analysis of Be nards observation, concluding that a surface tension gradient, i.e., the Marangoni stress, causes the convection patterns in a thin film. While much experimental and theoretical work has been performed to investigate various surface-tension-driven (i.e., Marangoni) flows in thin films or shallow pools,6-12 there are only a few papers reporting Marangoni flow in an evaporating sessile droplet. Zhang and Yang13 experimentally studied the natural convection in evaporating drops, where they observed a Marangoni flow. Davis and
* Corresponding author. E-mail: huhuadce@umich.edu.
(1) Hu, H.; Larson, R. G. Langmuir 2005, 21, 3963. (2) Scriven, L. E.; Sternling, C. V. Nature 1960, 187, 187. (3) Be nard, H. Ann. Chim. Phys., Ser. 7 1901, 23, 62. (4) Block, M. J. Nature 1956, 178, 650. (5) Pearson, J. R. A. J. Fluid Mech. 1958, 4, 489. (6) Scriven, L. E. Chem. Eng. Sci. 1960, 12, 98. (7) Nield, D. A. J. Fluid Mech. 1964 19, 341. (8) Fanton, X.; Cazabat, A. M. Langmuir 1998, 14, 2554-2561. (9) De Gennes, P. G. Eur. Phys. J. E 2001, 6, 421-424. (10) Stroock, A. D.; Ismagilov, R. F.; Stone, H. A.; Whitesides, G. M. Langmuir 2003, 19, 4358-4362. (11) Mancini, H.; Maza, D. Europhys. Lett. 2004, 66, 812-818. (12) Vogel, M. J.; Miraghaie, R.; Lopez, J. M.; Hirsa, A. H. Langmuir 2004, 20, 5651-5654. (13) Zhang, N. L.; Yang, W. J. Trans. ASME 1992, 104, 656-662.
co-workers14-16 studied both stationary and spreading droplets with consideration of evaporation, Marangoni stresses, and moving contact lines. They applied a lubrication analysis to obtain the evolution equation of the droplet free surface profile as a droplet spreads or resides on a heated substrate. The velocity fields in the droplet were thereafter obtained using the evolution equation for the droplet profile. Savino and co-workers17 numerically solved the axisymmetric steady-state Navier-Stokes equations taking into account the Marangoni stress at the liquid-air interface. They also performed experiments to map the velocity field in the droplets. Their theoretical results were not consistent with the experimental ones due to the errors in using the PIV technique to map the velocity field in a spherical cap droplet, which acts like a lens distorting the real flow field. Meanwhile, many researchers have taken advantage of Marangoni flow in drying droplets to affect the pattern of deposition from the droplet onto the underlying substrate.18-21 Wang and co-workers,18 for example, produced a patterned porous thin film on a substrate. Stebe and co-workers20,21 used a surfactant that during drying develops a concentration gradient along the droplet surface leading to a gradient in surface tension; by varying the initial surfactant concentration, they were able to vary the pattern of particle deposition on the substrate. Marangoni flow plays a key role in coating, thin film deposition, crystal growth, and production of photonic materials. Therefore, a thorough understanding of Marangoni flow in an evaporating droplet will be impor(14) Ehrhard P.; Davis, S. H. J. Fluid Mech. 1991, 229, 365. (15) Anderson D. M.; Davis, S. H. Phys. Fluids 1995, 7, 248. (16) Oron A.; Davis, S. H.; Bankoff, S. G. Rev. Mod. Phys. 1997, 69, 931. (17) Savino R.; Paterna, D.; Favaloro, N. J. Thermophys. Heat Transfer 2002, 16, 562-574. (18) Wang, H. T.; Wang, Zh. B.; Huang, L. M.; Mitra, A.; Yan Y. S. Langmuir 2001, 17, 2572-2574. (19) Maillard, M.; Motte, L.; Pileni, M. P. Adv. Mater. 2001, 13, 200204. (20) Nguyen, V. X.; Stebe, K. J. Phys. Rev. Lett. 2002, 22, 32823285. (21) Truskett, V.; Stebe, K. J. Langmuir 2003, 19, 8271-8279.
10.1021/la0475270 CCC: $30.25 2005 American Chemical Society Published on Web 03/26/2005
Microflow in an Evaporating Droplet
Langmuir, Vol. 21, No. 9, 2005 3973
tant both in fundamental research and in practical applications. Although Davis and co-workers and Savino and coworkers have developed theories for the Marangoni flow in an evaporating droplet, an analytical theory for the locally resolved axisymmetric flow in a slowly evaporating droplet with a pinned contact line is still lacking. Such a theory is needed, for example, to predict particle deposition from a drying droplet in the presence of the Marangoni flow when, as usual, the particles do not rapidly diffuse across the height of the droplet during flow and deposition. In what follows, in section 2, we develop a lubrication theory to describe the velocity field in an evaporating droplet in the presence of Marangoni stresses. To demonstrate the accuracy of the lubrication theory, we also develop a finite element method to solve simultaneously the thermal and flow fields in the evaporating droplet. In section 3, we present the results obtained from these methods and discuss the effect of surface-active contaminants on the Marangoni flow in the evaporating droplet. We finally summarize in section 4. 2. Theory 2.1. Expressions for the Velocity Field with a Thermal Marangoni Stress Boundary Condition. In our companion paper,1 we established the governing equations for an evaporating droplet without considering the thermal transfer due to latent heat of evaporation. Since in many experiments there is strong evidence of the thermal cooling affecting the flow pattern in the drying droplet, we here add the energy equation to the set of governing equations. Therefore, we have the Laplace equation for the vapor concentration distribution
Fcp
+ uT) + kT ) 0 (T t
(5)
where cp is the specific heat and k is the thermal conductivity. Here, we define an inverse Stanton number St-1 ) Fcpu j rR/k, which is a ratio of the convective to the thermal diffusive effects. In typical experiments with water droplets, such as those described in Hu and Larson,23 parameters in this group have the following approximate values: height-averaged radial velocity u j r ) 1 m/s, contact line radius R ) 1 mm, k ) 1.4536 10-3 cal K-1 cm-1 s-1 (from Bird et al.24), Cp ) 1 cal g-1 K-1, and water density F ) 1 g cm-3, which gives St-1 about 0.02. This implies that the rate of the convective heat transfer is much smaller than the rate of the conductive heat transfer, and so we can neglect the convection term in the energy equation. In the finite element analysis presented shortly, we confirm that the heat convection term is negligible. We can also neglect the transient in the energy equation, which we show is valid in Appendix A by estimating the ratio of the relative rates of change of droplet height to that of temperature. Thus, the energy equation (5) simplifies to the Laplace equation
T ) 0
(6)
c ) 0
The flow equations are
(1)
1 (rur) uz + )0 r r z
(2)
(( ) ) ( ( ) )
2ur P 1 (rur) + 2 ) r r r r z 2uz P 1 uz r + 2 ) r r r z z
(3)
(4)
where we have neglected inertial terms, since the Reynolds number Re Fu j rR/ is small (0.003) for weak flow in the slowly evaporating droplet considered here. Here, we will also neglect the buoyancy-driven flow because of the small value of a dimensionless group B Fgh02C/7.1375, which was introduced by Pearson5 to estimate the relative strength of the buoyancy-induced flow compared to that of Marangoni flow. We choose the water density F ) 1 g cm-3; water thermal expansion coefficient22 C ) 2.07 10-4 C-1; the temperature coefficient of surface tension for water, ) -0.1657 dyn cm-1 C-1; g ) 980 cm s-2; and the droplet height h0 ) 0.04 cm. Thus, we obtain B 3 10-4, which shows that the buoyancy-induced flow is very weak compared with the Marangoni flow in the evaporating droplet. We also consider heat transfer in the droplet and the glass coverslip, on which the droplet rests. The energy equation is
We solve eq 6 for a system containing both the droplet and a glass substrate. So the boundary conditions for eq 6 are as follows: 1. On the droplet surface S ) {h(r,t)|r e R}: Jh ) Hw(Jn), where Hw is the latent heat of evaporation of water and Jh is the heat flux. 2. On the glass surface outside the droplet z ) 0: R < r < : Jh ) 0. 3. At the lower boundary of the glass z ) -hg, 0 < r < and in the glass far away from the central axis -hg < z < 0, r f : T ) constant where hg is the thickness of the coverslip. The constant-temperature boundary condition 3 on the underside of the glass coverslip is reasonable in experiments with a water-immersion objective (which we will describe in a future publication), since in that case the gap between the coverslip and the microscope objective is filled with water, which acts as a heat bath. Note that the boundary condition on the free surface of the droplet assumes no heat loss due to either conduction or natural convection in the air above the droplet. These thermal boundary conditions could readily be modified, if necessary to describe other experimental conditions. We now develop a semianalytical lubrication solution for the flow field produced by droplet evaporation under the additional approximation of a relatively flat droplet, h/R , 1. Our procedure is numerically to solve the vapor concentration and heat equations, eqs 1 and 6, to obtain the temperature profile along the liquid-air interface and to obtain the Marangoni stress, which becomes a boundary condition for the momentum equations (3) and (4), used in the lubrication solution as described below. We check the accuracy of this semianalytical solution using a finite element analysis to solve simultaneously the mass balance equation (1), the flow equations (2)-(4), and the heat equation (6). A Marangoni stress, which is a surface-tension gradient on a free liquid surface induced by a temperature or a surfactant-concentration gradient, must be balanced by
(22) Perrys Chemical Engineers Handbook, 7th ed.; McGraw-Hill: New York, 1997. (23) Hu, H.; Larson, R. G. J. Phys. Chem. B 2002, 106, 1334-1344. (24) Bird, R. B.; Stewart, W. E.; Lightfoot, E. N. Transport phenomena; John Wiley & Sons: New York, 1960.
3974
Langmuir, Vol. 21, No. 9, 2005
Hu and Larson
a shear stress in the liquid, which drives a recirculating flow. From a force balance on the surface and assuming that the droplet is almost flat so that the tangential stress along the surface is approximately rz|z)h, we obtain
u z )
3 1 z 2 z 3 [1 + ()(1 - r 2)-()-1] + 41- t 3h 2 h
rz|z)h ) d/dr
(7)
where is the surface tension. Since the temperature on the surface of the evaporating droplet is nonuniformly depressed by evaporative cooling, the surface tension varies along the droplet free surface. In our finite element analysis (see below) we find that the temperature profile along the droplet surface is well fitted by
z 2 3 1 z 3 [(1 - r 2) - (1 - r 2)-()] - 3 h (0, t) 2 21- t 2h 3h h02 z 3 2 -()-1 2 ( J ( )(1 r ) + 1) z + R2 h r 2h02 z 3 J ()(() + 1)(1 - r 2)-()-2 z 2 2 R h r 2h02 z 3 2 -()-1 ( J ( )(1 r ) + 1) h (0, t) R2 h 2 Mah0 Mah0 2 b-2 z 3 (ab r (abr b + + 4(1 - a)) z 2 + 4R 2R h
( )
( ) () }
b + (1 - a)r 2 + c T/T0 ) ar
(8)
2(1 - a)r 2)
()
z 3 h (0, t) (13) h 2
where a, b, and c are the fitting parameters representing the shape of the temperature profile, T0 is the temperature difference between the edge and the top of the droplet, and r r/R. For most liquids, the surface tension decreases with increasing temperature and does so linearly for small temperature variations, i.e.
d/dT )
(9)
where is a property of the liquid; for water, ) -0.1657 dyn cm-1 C-1. From the chain rule, the surface-tension gradient is then given by
d dT d ) R dr dT R dr
(10)
Now, assuming a flat surface, we can write the boundary condition, eq 7, using the Newtonian expression for the shear stress of a viscous liquid, as
ur z
z)h(r,t)
)-
uz Ma (abr b-1 + 2(1 - a)r ) tf r
z)h
(11)
where the thermal Marangoni number is defined as Ma -T0tf/R, which is a ratio of the Marangoni force to the viscous force, and is the solvent viscosity. We apply the new boundary condition (11) within the lubrication theory developed in our companion paper1 and obtain the flow equations with a Marangoni-stress boundary condition
where we have defined the dimensionless variables: u r ) z ) uztf/h0; t ) t/tf; r ) r/R; z ) z/h0; h ) h/h0. tf is urtf/R; u the drying time, R is the contact line radius, h0 is the -J(0,t)/Fh (0,t) is the initial height of the droplet, and J dimensionless mass flux at the top of the droplet surface, where J(0,t) and h (0,t) are given by eqs 11 and 23, respectively, in ref 1. When there is no Marangoni stress along the droplet surface, i.e., the Marangoni number Ma is zero, eqs 12 and 13 reduce to the solutions in our companion paper.1 When the velocity gradient uz/r at the free boundary surface is neglected (by dropping the terms in braces, see the discussion in ref 1), eqs 12 and 13 reduce to the simplified classical expressions for the Marangoni boundary condition. Terms arising from uz/r are normally dropped in a standard lubrication analysis but are included here because of the flow singularity at the droplet edge, as is discussed in our companion paper.1 2.2. General Expressions for the Velocity Field with Marangoni Stresses. In section 2.1, an approximate solution for the full velocity field with the Marangoni stress on the droplet free surface was derived, in which the Marangoni stress was generated by a temperature gradient. However in some situations, such as those examined experimentally by Nguyen and Stebe,20 the Marangoni stress may be generated by a surfactant concentration gradient or by both temperature and surfactant concentration gradients, and the resulting flow affects the solute deposition behavior. For an arbitrary Marangoni stress distribution, we define a function g(r,t), in the Marangoni boundary condition
u r z
z)h(r,t)
d dr
z)h
uz r
z)h
g(r,t)
(14)
u r )
3 1 1 z z 2 [(1 - r 2) - (1 - r 2)-()] 2 - 2 + 81- r h h t r h02h R

2
(J ()(1 - r 2)-()-1 + 1)
Mah0h 3z z 2 (abr b-1 + 2(1 - a)r ) (12) 2R 2h h 2
) ( )} ( )
2 3z z 2h h 2
The function g(r,t) combines the surface tension gradient (produced by temperature, surfactant-concentration gradient, or both) and a boundary vertical velocity gradient (produced by evaporation) along the radial direction. Once these are determined, ur/z|z)h(r,t) can be used as a boundary condition to solve the velocity field, giving
u r )
3 1 1 [(1 - r 2) - (1 - r 2)-()] 81- t r t)tfh0h g(Rr ,tf 2R
( (
z 2 z -2 h 2 h 2 3z z (15) 2h h 2
) )
Langmuir, Vol. 21, No. 9, 2005 3975
u z )
3 1 z 2 z 3 [1 + ()(1 - r 2)-()-1] + 2 41- t 3h h z 2 3 1 z 3 [(1 - r 2) - (1 - r 2)-()] h (0, t) + 21- t 2h 2 3h 3 t)tfh0 2 z t)tfh0 2 z g(Rr ,tf g(Rr ,tf 3 3 z + z 4r R 4R h h t)tfh0 z r g(Rr ,tf 3 h (0, t) (16) 2R h 2
where g(Rr ,tf t) is the derivative of g(r,t) with respect to r . For any specific case, if we know the surface tension distribution and boundary vertical velocity gradient along the free surface, then we can obtain g(r,t), which allows us to obtain u r and u z from the above equations. We must note, however, that while the temperature distribution along the droplet surface is dominated by heat diffusion and can therefore be obtained independently of the flow field, surfactant-concentration gradients are controlled by the flow. Hence the Marangoni stress boundary condition produced by surfactants must be obtained simultaneously with the velocity field. 2.3. Finite Element Model. A finite element model corresponding to the governing equations (1) to (6) is
Mc ) J [C(u) + L]T ) F K(T)u ) F(T)
(17) (18) (19)
Figure 1. Contour plots of the temperature fields (in C) in the droplet and the glass coverslip obtained by a finite element analysis of the heat equation, eq 18, at two contact angles. The parameters used for solving the heat eq 18 are as follows: vapor latent heat, Hw ) 541 cal g-1; thermal conductivity of water, kw ) 1.4536 10-3 cal cm-1 s-1 K-1; and thermal conductivity of glass, kg ) 2.2976 10-3 cal cm-1 s-1 K-1. The dimension of the glass substrate is taken to be 1.3 mm in radius and 0.15 mm in thickness. The results are insensitive to increases in the radius of the substrate.
where the rate of mass transfer due to diffusion is represented by the matrix M, which is an assembly over ne e all elements in a mesh, namely, M ) e )1m . The total rate of mass transfer at the boundary is represented by ne e J, where J ) e )1j . The rate of heat transfer per unit temperature due to convection is represented by the matrix C(u), which is an assembly over all elements in a mesh, ne e namely, C(u) ) e )1c(u) . The rate of heat transfer per unit temperature due to conduction is represented by L, ne e where L ) e )1l . The total rate of heat transfer at the ne e boundary is represented by F, where F ) e )1f . The temperature-dependent viscous diffusion term is reprene e sented by the matrix K(T), where K(T) ) e )1k . The total force acting on the boundary is represented by the term ne e F(T), where F(T) ) e )1f . The boundary conditions are described in section 2.1 and the finite element equations (18) to (20) are solved simultaneously. Details of our method of solving these equations can be found in ref 1. In general, if both temperature and surfactant concentration gradients are present along the droplet surface, eq 33 in ref 1 should be used to solve the finite element model (18) to (20). Thus, the Marangoni stress given by eq 33 in ref 1 contains both thermal and surfactantconcentration-gradient Marangoni stresses. The latter stresses can be obtained by incorporating a mass balance equation for the surfactant concentration in the droplet. However, this is beyond the scope of this work and will be considered in future work. 3. Results and Discussions 3.1. Temperature Field. To obtain the Marangoni stress arising from a surface-tension gradient due to evaporative cooling, we first obtain the evaporation flux profile on the droplet free surface from either the analytic formula of Hu and Larson23 or the finite element solution
of eq 17. With this nonuniform evaporation-flux distribution, the latent heat flux can then be calculated and applied as a boundary condition to solve the finite element heat equation. In our finite element analysis, the parameters used for solving the heat eq 18 and the flow eq 19 are as follows: vapor latent heat of evaporation of water Hw ) 541 cal g-1 (from ref 25), thermal conductivity of water kw ) 1.4536 10-3 cal cm-1 s-1 K-1 (from Bird et al.24), thermal conductivity of glass kg ) 2.2976 10-3 cal cm-1 s-1 K-1 (from Bird et al.24), viscosity of water ) 0.01 P, and temperature coefficient of the surface tension of water ) -0.1657 dyn cm-1 K-1 ( is obtained from ref 25). By applying the finite element analysis, the temperature distributions in a droplet with a contact-line radius of 1 mm and initial height 0.364 mm are obtained for 40 and 10 contact angles and plotted in Figure 1. These contour plots show that at the initial contact angle of 40 the temperature increases from the top to the bottom of the droplet, and from the center to the edge of the droplet, while at a contact angle of 10, the temperature decreases from the center to the edge of the droplet. Thus, the thermal field in the droplet changes significantly with evaporation time, and its radial gradient even reverses direction. The reversal of temperature-gradient direction occurs because at early times, the longer conduction path from the bottom of the glass to the top of the droplet makes the temperature lower at the top of the droplet than elsewhere, while at long times the faster rate of evaporation at the droplets edge makes it cooler there. From the temperature fields, the droplet surface temperature profiles at different contact angles are extracted and plotted in Figure 2. When the contact angle decreases
(25) CRC Handbook of Chemistry and Physics, 63rd ed.; CRC Press: Cleveland, OH, 1984.
3976
Hu and Larson
Figure 2. The temperature profiles along the droplet surface at contact angles of 40, 35, 30, 25, 20, 15, and 10. The solid lines are the finite element results, and the dashed lines are the fitting results using eq 8.
below 14, the radial temperature gradient along the surface exhibits a transition from positive to negative slope so that the coldest point along the droplet surface shifts from the center to the edge of the droplet. The surface temperature from the finite element analysis can be fitted accurately by eq 8; see Figure 2. The choice of the functional form given in eq 8 is based on two considerations. First, across the most of the droplet, the temperature distribution can be represented by a simple parabolic function. However, near the contact line the temperature varies rapidly and deviates from the parabolic curve. Therefore, we included a term of higher order in r in eq 8 to correct this deviation. The lack of a term that is first order in radius r is due to the symmetry condition that the first derivative of the temperature along the droplet surface must be zero at r ) 0. Parameters a and b in eq 8 represent the strength of the temperature gradient along the droplet surface. The constant c is just the temperature at top of the droplet, divided by T0, the temperature difference between the edge and the top of the droplet. The effect of the temperature field and the resultant Marangoni stress on the flow will be seen in the following paragraphs. 3.2. Velocity Field. The time-dependent velocity fields coupled with the thermal fields obtained from the finite element analysis are plotted in Figure 3, in which at the initial contact angle of 40 the Marangoni number Ma -T0tf/R ) 841 is large and therefore produces a strong recirculation in the droplet. As the contact angle decreases, the temperature gradient on the droplet surface is attenuated so that the Marangoni number decreases. When the contact angle decreases below a critical value of 14, the recirculation disappears and the velocity vectors all point toward the edge of the droplet, because at this contact angle, the shear stress at the surface changes sign from negative to positive. Thus, the sign of the Marangoni number Ma changes at a contact angle of 14. For the Marangoni-stress boundary condition, we can also obtain the analytical results if we know the parameters a, b, and c in eqs 12 and 13. We therefore fit the finite element temperature distributions at different contact angles depicted by the solid lines in Figure 2 with
Figure 3. The time-dependent velocity fields calculated by the finite element method with the Marangoni-stress boundary condition. Plots a to d are for contact angles 40, 30, 20, and 10, respectively.
Table 1. The Parameters a, b, c, and T0 Obtained by Fitting the Computed Temperature Profile by the Phenomenological Expression, Equation 8
a 40 35 30 25 20 15 10 0.334064 0.286465 0.302891 0.448445 0.487827 0.807235 0.833545 b 8 10 12 16 20 26 8 c 1601.529 2087.617 2532.286 2618.861 4045.307 3681.104 -5342.43 T0 0.015594 0.011965 0.009865 0.00954 0.006177 0.006788 -0.00468 av rel error, % 0.017 0.009 0.015 0.021 0.012 0.013 0.019 max rel error, % 0.11 0.04 0.08 0.25 0.43 0.22 0.44
the empirical expression eq 8, which gives the dashed lines, which lie almost on top of the solid lines. The fitting parameters a, b, and c in eq 8 are listed in Table 1 along with the average and largest relative errors between the fitted and the finite element results. The errors are tiny for all contact angles, confirming the accuracy of the fitting function (8). Hence, with this accurate analytical representation of the temperature profile along the free surface, we can compute the velocity field using the analytical solution equations (12) and (13). As we shall see, the analytical flow field obtained in this way is nearly same as the flow field from the finite element analysis, shown in Figure 3. The streamlines for a contact angle of 40 obtained from the finite element analysis are very similar to those from the analytical solution, as can be seen from the plots in Figure 4. Similar agreement is obtained at a contact angle of 10 (not shown). Furthermore, the ur(z) and uz(z) profiles computed by the analytical solution nearly overlap those computed by the finite element model for contact angles
Langmuir, Vol. 21, No. 9, 2005 3977
Figure 4. Streamline plots of the flow field from the finite element model for the Marangoni-stress boundary condition at contact angle of 40, from (a) FEM solution and (b) analytical solution.
40 and 10ssee Figures 5 and 6sand all contact angles between these two (not shown). The maximum average relative difference between the finite element and the analytical results for radii up to r ) 0.9 is less than 3%. It is surprising that the finite element model and analytical solution agree with each other even better than in the absence of the Marangoni stress; see ref 1. This level of agreement is retained even when the classical lubrication analytic solution is used; i.e., when the terms in braces in eqs 12 and 13, due to the boundary term uz/r|z)h, are dropped. Evidently, the Marangoni stress tends to swamp some of the errors in the classical lubrication approximation, leading to a more accurate solution than in the absence of Marangoni stresses. 3.3. Effects of Maragoni Stress on the Velocity Field. Comparing the velocity profiles of ur(z) and uz(z) in Figure 6 of this paper, with Marangoni stress, to that of Figure 6 in our companion paper1 which neglects the Marangoni force, we find that the differences in the radial velocity profiles ur(z) are particularly pronounced. When the Marangoni number is zero, the gradient ur/z|z)h is close to zero at the free surface. When the Marangoni number is negative (as is the case for a contact angle of 10), there is a positive velocity gradient ur/z|z)h on the droplet free surface, and when the Marangoni number is positive (as is true for a contact angle of 40), this gradient is negative. The axial gradient of the radial velocity profile ur(z) is dominated by the contribution of the Marangoni term to ur/z|z)h, which is much higher than that of uz/r|z)h in the boundary condition (11). The result in Figure 6a shows that a negative Marangoni number promotes motion of the fluid from the center toward the edge of the droplet surface, i.e., a clockwise (or reverse)
Figure 5. (a) Radial and (b) vertical velocities versus vertical position at different radial positions r ) 0.1, 0.2 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 0.9 mm, for a contact angle of 40 and a Marangoni-stress boundary condition. The solid lines are from the finite element method and the dashed lines from the analytic solution in eqs 12 and 13.
circulation. The transition from counterclockwise (positive Ma) to clockwise (negative Ma) recirculation, as the contact angle drops below 14, can be seen in Figures 5 and 6. For large fixed r (near the droplet edge to the right of the stagnation point) the radial velocity changes from positive to negative in sign with increasing z for counterclockwise rotation (Figure 5a) while the opposite occurs for clockwise rotation (Figure 6a). In Figure 6a, the zone of negative ur(z) is tiny, indicating a very weak recirculation. 3.4. Surface-Active Contaminants. The theoretical results in Figure 3 predict a strong recirculation flow in the water droplet. However, in experiments with drying water droplets at most a weak recirculation flow is observed.26 Moreover, a negligible Marangoni flow in water droplets is consistent with the commonly observed coffee ring pattern produced by solute deposition from a drying droplet.26,27 However, our FEM and analytical flow fields predict a very strong thermally driven Marangoni flow in drying water droplets. It has been reported, however, that surface-active contaminants that collect on the free surface can almost entirely suppress Maranogni flow in an evaporating water droplet and that low concentrations of
(26) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827-829. (27) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R. Phys. Rev. E 2000, 62, 756-765.
3978
Hu and Larson
and so resides only at the interface between water and air. Since we have a relatively flat droplet, we can obtain the advection-diffusion equation for the surfactant along the free surface as follows
1 D (u r|z)hr ) ) r r r r r r t
( )
(22)
where is the number of surfactant molecules per unit area; D is dimensionless diffusivity, which is D ) Dtf/R2, and u r|z)h is the radial velocity along the droplet surface. If we assume that the transfer of surfactant along the droplet surface is a quasi-steady-state process, we obtain, from eq 22
r|z)h u ) r D
(23)
For a dilute concentration of surface-active agent, the surface pressure induced by surfactants on the free surface is
S ) -kBT
Using the chain rule, we then have
(24)
dS d dS )R dr R d dr
(25)
Substituting eqs 23 and 24 into eq 25, we obtain the surfactant contribution to the Marangoni stress
Figure 6. The same as Figure 5 except for a contact angle of 10.
these contaminants are almost unavoidable for water surfaces, which are easily contaminated.17,20 We will therefore use our theory to analyze how a small concentration of surface contaminant can affect the Marangoni flow in an evaporating droplet. We consider a simple case, in which a Marangoni stress is induced by a surface tension gradient generated by both temperature and surfactant concentration gradients on the free surface of a relatively flat water droplet. Thus, the total Marangoni stress on the free surface is d/R dr , given by
kBT u dS r|z)h )R dr R D
(26)
Combining eqs 20, 21, and 26 with eq 14, we obtain the axial gradient of the radial velocity on the free surface, which is represented by the function g(r,t)
g(r,t) ) -
dT dS d ) + R dr R dr R dr
Ma (abr b-1 + 2(1 - a)r ) tf kBT u r|z)h uz R D r
z)h
(27)
(20)
Substituting eq 27 into the general eqs 15 and 16, we then derive an analytical solution for the flow field in the presence of both temperature and surfactant concentration gradients on the free surface for a flat droplet, namely
where dT/R dr is the Marangoni stress induced by a temperature gradient and dS/R dr is the Marangoni stress induced by a surfactant concentration gradient. If we know the total Marangoni stress in eq 20, we can substitute it into general solutions (15) and (16) to obtain an analytical solution. From eqs 8-10, we can derive the thermal contribution to the Marangoni stress
u r )
dT Ma )(abr b-1 + 2(1 - a)r ) R dr tf
(21)
We now derive the surfactant contribution to the Marangoni stress. First, we make the additional assumption that the surfactant is insoluble in the droplet fluid
z 2 3 1 1 z [(1 - r 2) - (1 - r 2)-()] 2 - 2 + 81- r t h h r h02h 2 z 3z (J ()(1 - r 2)-()-1 + 1) + 2 2h R h 2 Mah0h 2 3z z (abr b-1 + 2(1 - a)r ) + 2R 2h h 2 u r|z)h kBTtfh0h z 2 3z (28) 2 2h D 2R h 2
)}
Langmuir, Vol. 21, No. 9, 2005 3979
u z )
z 3 3 1 z 2 [1 + ()(1 - r 2)-()-1] + 2 41- t 3h h 3 1 z 2 z 3 [(1 - r 2) - (1 - r 2)-()] - 3 h (0, t) 2 21- t 2h 3h h02 z 3 2 -()-1 2 ( J ( )(1 r ) + 1) z R2 h r 2h02 z 3 J ()(() + 1)(1 - r 2)-()-2 z 2 + 2 R h r 2h02 z 3 2 -()-1 ( J ( )(1 r ) + 1) h (0, t) R2 h 2 Mah0 2 b-2 Mah0 z 3 (ab r (abr b + + 4(1 - a)) z 2 + 4R 2 R h u r|z)h kBTtfh0 2 z 3 3 2 z 2(1 - a)r ) 2 h (0, t) z + 2 h D 4r R h kBTtfh0 2 z 3 (u r|z)h) z 2 r 4D R h tfh0 z u r|z)h kBTr 3 h (0, t) (29) 2 D 2R h 2
()
( ) () ( ) { ( ) ( ) () }
) ) )
In eqs 28 and 29, the terms in braces are the contributions from the surfactants on the free surface. If we neglect the effect of surfactant contaminants, i.e., the terms in braces, then eqs 28 and 29 reduce to eqs 15 and 16. To obtain the term u r|z)h in eqs 28 and 29, we set z ) h in eq 28, and solve for u r|z)h to give the radial velocity on the free surface
u r|z)h ) -
h02h 1 Mah0h 1r (J ()(1 - r 2)-()-1 + 1) (abr b-1 + 2 2 R 2 2R 2(1 - a)r )
3 1 1 [(1 - r 2) - (1 - r 2)-()] 81- t r
]/(
1+
kBTtfh0h (30) 2 D 4R
in the absence of either thermal or surfactant Marangoni stress the radial velocity on the free surface is nearly zero but becomes large when thermal Marangoni stresses are present. We now find, however, that a tiny concentration of contaminants can greatly suppress the radial surface velocity that is produced by the thermal Marangoni stress. Hence, the presence of a trace concentration of surface contaminant can largely cancel out the thermal Marangoni effect. In experiments with water droplets, it is very difficult to control the amount of surfactants below the low level that suppresses thermal Marangoni flow, even when the experiments are carried out in a clean room.28-31 Other volatile fluids, whose surfaces are not so easily contaminated, might be expected to show strong Marangoni flows due to evaporative cooling.17 Our theory, which includes Marangoni effects due to both temperature variations and insoluble surface active agents, could be used to predict not only the effects of unintentional surfactant contaminants, but also the effects due to deliberately introduced surfactants, such as those employed at high concentrations by Stebe and co-workers20 to control the deposition patterns of solutes from drying droplets. At high concentrations near surface saturation, the pressure-concentration isotherms depart greatly from the ideal form given by eq 24, and interesting deposition patterns can thereby be induced. In order apply our theory to analyze the particle deposition process in the presence of the surfactants in the droplet, an additional convectivediffusion equation should be introduced to calculate the surfactant concentration distribution. Once the surfactant concentration distribution is obtained along the droplet surface, the Marangoni force can be calculated and the velocity field is then obtained from the general equations (15) and (16). Since the convective flow and surfactant concentration distribution equations are in general coupled, an iterative scheme will be needed to find simultaneous solutions for both the flow and surfactant concentration fields. We expect that our lubrication theory will be of considerable help in explaining and controlling these deposition patterns. 4. Summary We have performed a thorough theoretical study of the effects of Marangoni stress on flow in an evaporating droplet. Marangoni stress due to thermal gradients along the free surface produced by latent heat of evaporation is introduced into the free surface boundary condition, allowing an analytical solution to be obtained for the flow field, using the lubrication approximation. The solution requires specification of the temperature gradient along the droplet surface, which can be obtained by a numerical solution of the thermal field. We developed a finite element model to solve the coupled thermal and velocity fields in the droplet, both to obtain the needed thermal field for the lubrication solution, and to confirm the validity and accuracy of the lubrication theory for the flow field. These solutions show that the heat of vaporization, the nonuniform path lengths for heat conduction, and the nonuniform evaporation rate lead to a nonuniform distribution of temperature along the air-liquid interface and hence a nonuniform surface tension, which drives a thermal Marangoni flow. The lubrication approximation
(28) Ward, C. A.; Stanga, D. Phys. Rev. E 2001, 64, Art. No. 051509. (29) Barnes, G. T.; Hunter, D. S. J. Colloid Interface Sci. 1982, 88, 437-443. (30) Cammenga, H. K.; Schreiber, D.; Rudolph, B. E. J. Colloid Interface Sci. 1983, 92, 181-188. (31) Schreiber, D.; Cmmernga, H. K. Phys. Chem. Chem. Phys. 1981, 85, 909-914.
In above equation, we find that the numerator (in brackets) is the radial velocity on the free surface in the absence of the surfactant. The denominator (in parentheses) represents the suppressing effect of the surfactant contaminants on this radial free surface velocity. We can estimate from this denominator the concentration, , required to suppress the radial velocity by, say, a factor of 100 by setting
kBTtfh0h ) 100 D 4R2
(31)
Using typical values for the parameters for a water droplet (R ) 1 mm, tf ) 360 s, D 1 10-5 cm2 s-1, h0 ) 0.36 mm), we obtain
100 4 0.01 (1 10-5) (4 10-14)(0.036 108)
300 molecules/m2
(about 5 10-22 mol/m2). From this value, we can easily find that for a water droplet with a contact line radius of 1000 m, the total amount of surfactant on the free surface needed to suppress radial flow along the surface by a factor of 100 is about 1 109 molecules. We noted earlier that
3980
Hu and Larson
provides an accurate analytical solution to the timedependent axisymmetric flow field, including the recirculation due to the Marangoni-stress boundary condition, when compared to a finite element solution. We predict that the nonuniformity in heat-conduction path lengths produces a positive Marangoni number at a large contact angle, leading to an inward radial flow along the droplet surface, while the nonuniformity in evaporation rate yields a negative Marangoni number at small contact angles, yielding an outward flow. The general structure of the lubrication solution admits inclusion not only of thermal Marangoni flows but also of flow induced by surfactant concentration gradients along the droplet surface. From our lubrication solution, we learn that a small concentration, as about 300 molecules/m2, of insoluble surfactant on the free surface, significantly reduces the radial flow produced by the thermal Marangoni stresses. This explains why surfaces that are easily contaminated, such as water surfaces, typically do not show thermal Marangoni effects, while droplets with less easily contaminated surfaces do show pronounced thermal Marangoni flows. Appendix A Here we will show that the temperature field in the drying droplet is at quasi-steady state by estimating the ratio of the relative rates of change of droplet height to that of temperature
Ah0FCp
dT dh AFHw dt dt
(A2)
where A is the area of the substrate covered by the droplet and Hw is the heat of vaporization. The decrease in temperature continues until it is limited by conduction of heat from the substrate. Thus, the eventual fluid temperature drop T can be estimated by a quasi-steadystate energy balance, yielding
dh kT FHw h0 dt
(A3)
where the left side is the rate of which heat flows into the droplet from the substrate and the right side is the rate at which latent heat is lost by evaporation. Incorporating these two estimates in eqs A2 and A3 into eq A1 gives
1 dh CpFh0 dh/dt h0 dt ) 1 dT k T dt
(A4)
Noting that dh/dt is approximately the vertical velocity, u j z, and that the ratio of the vertical to the radial velocity is of the order of the drop aspect ratio h0/R, we obtain
( )
(
1 dh h0 dt 1 dT T dt
(A1)
1 dh h0 dt h0 2 -1 = St 1 dT R T dt
()
(A5)
where T is the decrease in temperature of the liquid due to latent heat of vaporization resulting from evaporation. If the above ratio is small, then the droplet temperature equilibrates rapidly compared to the rate at which the droplet evaporates. The initial rate of temperature change can be estimated from a transient adiabatic energy balance, yielding
rR/k. Since h0/R < 1, and St-1 , 1, the where St-1 ) FCpu droplet temperature field reaches steady state quickly compared to the rate of change in droplet height and the temperature field therefore remains at quasi-steady state. Acknowledgment. We thank NASA microgravity research division for supporting this study through Grant NAG3-2134 and NAG3-2708.
LA0475270
Revue
Phys. Appl.
23
(1988)
975-987
JUIN
1988,
975
Classification
Physics Abstracts
68.10
,
Wetting films
N.
and
wetting
of
V. Churaev
Laboratory of thin liquid layers, Department of Surface Phenomena, Institute Academy of Sciences U.S.S.R., Leninsky Prospect 31, 117915 Moscow, U.S.S.R.
Physical Chemistry,
(Reu
le 27 octobre 1987,
accept le
janvier 1988)
Rsum. Une approche macroscopique des phnomnes de mouillage, base sur les isothermes de pression de disjonction des films de mouillage en quilibre avec des mnisques concaves ou des gouttes, permet le calcul des angles de contact de divers liquides sur des substrats solides. Nous prsentons les rsultats de calculs utilisant la thorie des forces de surface longue porte et nous montrons leur accord avec des donnes exprimentales. Nous insistons sur le rle de la zone de transition entre le film de mouillage et le liquide en volume dans ltude des tensions de ligne, de lquilibre mcanique et thermodynamique et de la formation des angles de contact dynamiques. Les transitions de mouillage et de prmouillage sont discutes dans le cadre de cette thorie macroscopique.
2014
Abstract. Macroscopic approach to the wetting phenomena, based on the isotherms of disjoining pressure of wetting films in equilibrium with concave menisci or drops, allows one to calculate the contact angles for different liquids and solid substrates. The results of calculations are presented using the theory of long-range surface forces, and their agreement with experimental data is shown. The role of the transition zone between the wetting films and bulk liquid in the phenomena of line tension, mechanical and thermodynamical equilibrium, and the formation of dynamic contact angles, is demonstrated. Within the framework of the macroscopic approach the phenomena of prewetting and wetting transition are discussed.
2014
1. Introduction.
The phenomenon of wetting consists in the formation of a contact angle between a liquid and a solid substrate. The equilibrium contact angle is determined by the surface forces field and depends on the energy of interaction of the liquid with a solid substrate. A weak interaction causes the partial wetting or nonwetting, whereas a strong one the spreading of the liquid over the surface, i.e. its
complete wetting.
We shall consider
a
macroscopic approach
to
wetting phenomena, based on an analysis of the equilibrium of menisci or drops with wetting films. The wetting films, in distinction from adsorption ones, may be considered as a part of the liquid phase. For such films the known concept of disjoining pressure may be used. The macroscopic approach is restricted to not very large contact angles, and it is valid until the thickness of wetting films exceeds several molecular layers. This allows the wetting phenomenon to be considered within the framework of the theory of long-range surface forces [1]. Reasonable combination with another microscopic
approach taking into account the molecular structure liquids [2-5], permits, as may be supposed, covering the whole range of possible values of contact angles. In distinction from the microscopic approach, the macroscopic one enables one to take into account, not only the molecular forces, but also other components of disjoining pressure : for instance, the double-layer electrostatical forces and structural forces arising due to the overlapping of boundary liquid layer with modified structure. Development of the long-range surface forces theory [1] allowed quantitative evaluations of wetting to be made depending on the physical properties of a solid substrate and the liquid interacting with the
of
latter. This approach was for the first time indicated by Frumkin and Derjaguin [6, 7]. They have substantiated the relationship between the value of contact angle 00 and the disjoining pressure isotherms of wetting films II(h ).
II (h ) isotherms had been studied but insufficiently delayed for a long time the application
That the
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01988002306097500
976
of Frumkin-Derjaguins theory. At present, a considerable progress has been achieved both in the experimental examination of the disjoining pressure isotherms of wetting films [1, 8] and in developing methods for theoretical calculation of different components of disjoining pressure, acting in films. 2.
Theory of
contact
angles.
the radius of action of surface forces, usually on the order of 10- 6 -10- 5 cm. Beyond the zone of influence of surface forces II 0, and equation (1) transforms into Laplace equation for an undisturbed meniscus : y . Ko Pc. At h - ho, the profile of the transition zone becomes all the more gentler sloping (K ~ 0), and for a flat film, when K = 0, equation (1) gives the known condition of equilibrium between the flat wetting film and the meniscus
by
We begin considering the wetting theory with the definition of the concept on the contact angle 00. In the equilibrium state, the contact angle is determined at the intersection point of a continuation of the drop or meniscus profile undisturbed by the surface forces with the substrate (curve 1, Fig. 1). The meniscus contacts with a wetting film, whose equilibrium thickness ho is determined by the equation of H(h) - isotherm. The value of ho corresponds to the disjoining pressure, which is equal to the capillary pressure of an equilibrium meniscus : Pc Ho. A transition zone 2 is formed between the bulk portion of the meniscus having a constant (if neglecting the gravity force) surface curvature, Ko Pc/y (where y is the surface tension), and a flat wetting film. The capillary forces due to the curvature of the liquid layer surface, and the surface forces associated with the long-range substrate field, act simultaneously in that transition
= =
II (ho)
For
Pc.
having the width 2 H, and in the case of the complete wetting (for instance, for the isotherms of the type II A/hn > 0) the extension f of the transition zone in the direction of the axis x is
a
flat slit
approximately equal
to ~~(H.h0)1/2 [10, 11].

ktm wide and at
Thus, for example, for slits H = 10
ho =10- 5
contact
cm, f =10- 4 cm.
zone.
Fig. 1. eo, when

slit.
Formation of an equilibrium contact angle, the liquid is in contact with the surfaces of a flat
The aforesaid method for determination of the angle is inapplicable in two cases. The value of 00 cannot be determined in narrow slits, where the fields of surface forces overlap one another, and where the meniscus has no part of a constant curvature. Here another approach was developed to calculate the equilibrium of the capillary liquid with the wetting films [12]. Special consideration must also be paid to the cases of complete wetting, where a continuation of the meniscus profile does not intersect the substrate (curve 3, Fig. 1), and a contact angle does not form. The macroscopic theory includes consideration of the cases of both partial ( 9 0 > 0 ) and complete wetting. Hereinbelow is given one of the derivations of the corresponding equations of the theory, based on the application of equation (1). For the meniscus in a flat slit H > ho, the curvature of the cylindrical surface of the meniscus is equal to K = h"[1 + (h )2]- 312,where h and h" are the first and the second derivative of the thickness h of a liquid layer with respect to the coordinate x. Substituting this value of K(h ) into equation (1) and using the boundary conditions : h 0 at h = ho, and h ~ oo at h H, we obtain the following solution of
=
equation (1) [13] :
state, we obtain from the of condition of constancy pressure in all the parts of the system [9] :
In the
equilibrium
where h is a local thickness of the transition zone and K(h ) is a local curvature of its surface. This equation is valid for the gently sloping profiles of the transition zone (~h/~x ~ 1 ), when the H (h ) isotherm of a flat film is applicable to each elementary part dx. The extension of the transition zone is determined
Ho H (ho) and denotes the value of the integral. When 00 > 0, the value of Pc may be expressed through cos 0 0. In the case of a flat slit, which is here considered, P c y /r y . cos 0 OIH, where r is
where
= =
=
the radius of curvature of an undisturbed part of meniscus (Fig. 1). Then, instead of equation (2), we obtain :
977
Now
we can
compare this
expression with the known
assume
Young equation
the
ah/ax 1), equation (1) following manner [14] :
may be written in
angle 00 from the specific interphase equation (4) requires knowing energies of the solid substrate at the boundary with the gas phase, ysv, and the liquid, ysL. The Young equation, in distinction from equation (3), does not allow determination of the contact angle, because there are available no methods independent of equation (4), for determining either each of interphase energies or the difference between these. Now, the Young equation is usually employed for solving an inverse problem finding a difference ysv - ysL on the basis of the measured values of 00. The Frumkin-Derjaguins theory enables one to determine the value of ysv for the solid surface coated by a wetting film. Its value is equal to a sum of the interphase energies of two surfaces of a liquid film : one, contacting gas, y, and another, contacting the solid substrate, ysL, and variations in the free
contact
-
To
determine
the
where h dh /dp , h" d2h/dp 2, r Const. is the drop surface curvature radius, and p is the radial coordinate in the substrate plane. In the general case, nonlinear differential equation (5) may be solved only numerically. It can be linearized by using a simplified form of the disjoining pressure isotherm :
= =
=
energy of the film, AG
Jn
h . dII =
..1 + no ho,
where a and t are the parameters of the isotherm, which are characteristic of the slope of its stable part, where 8Hl8h 0, and the radius of action of surface forces, accordingly. The form of simplified isotherm (curve 1, Fig. 2) is similar to the real one (curve 2), when only thin wetting films are stable. The solution of equation (5) gives the following expression for the equilibrium contact angle [14] :
when it is thinning out from ao to the equilibrium thickness ho. As a result, the difference between
where to
is the film thickness at II
0.
and ysL proves to be equal simply to y + Ho ho + d, which just contains equation (3). Equation (3) allows a theoretical determination of the value of 00 in accordance with the known H(h) isotherm of the wetting films of a given liquid on a given substrate. The methods for the experimental and the theoretical determination of H(h) isotherms are presented in reference [8]. Equation (3) has a similar form for cylindrical drops on a flat substrate [13]. This is not surprising, since at H > ho (in this case H denotes the drop height) the equilibrium conditions should not depend on the surface curvature sign beyond the radius of action of surface forces. A difference resides only in that the values of Ho are negative, because the capillary pressure of the drop having a convex surface, has another sign Ho Pc = - y /r, where r is the radius of curvature of an undisturbed portion of the drop. Equation (3) is also applicable at Pc 0, when the surface of the bulk liquid is plane, and its profile is wedge-shaped. In such a case, the second term disappears from the right-hand side of equation (3), since Ho Pc 0. For drops having the spherical surface the calculations are complicated in connection with a necessity of taking into acount the secondary curvature of the drop surface. If we limit ourselves to the case of relatively small contact angles (when it is possible to
=
= = =
Fig. 2. Simplified isotherm of and its real analogue (b).

-
disjoining
pressure
(a)
As appears from this expression, the drop contact angle 00 decreases with its dimensions, which is accompanied by an increase in the negative capillary pressure P c. Such an effect was, in particular, detected experimentally for water drops less than 3 mm in diameter [15]. Under otherwise equal
conditions, an increase in the radius of action of surface forces, t, or a decrease in to causes an

increase in the contact angle values. This is connected with a displacement of the isotherm (6) into the region of negative values of disjoining pressure. The range of the applicability of equation (3) is restricted by the values of 00 > 0, when Ho ho +
978
A -- 0 and cos 0 0 -- 1. In the case where the continuation of the meniscus in a flat slit does not intersect substrates (Fig. 1, curve 3) equation (3) permits determination of the capillary pressure of the meniscus in equilibrium with the film having the thickness ho. Taking into account that in the equilibrium state, Pc Ho, we obtain from equation (2) :
=
equation correlates the capillary pressure of meniscus, Pc, and hence, its curvature radius, r y /Pc, with the slit half-width H and through 4 with the disjoining pressure isotherm. It should be taken into account, however, that ho is a function of Pc. The value of ho can be determined from the isotherm equation, 03A0(h), at Ho Zizi) Po. Equation (8) enables one to determine, instead of the contact angle, another parameter, which may be used to characterize the complete wetting conH - r the diffrence h. ditions ; namely, of h* the the value the smaller The larger (Fig. 1). meniscus curvature radius and the better the liquid wets the solid surface. Transforming equation (8), we obtain the following expression for h. :
This
=
to the wetting films that are stable due to the electrostatic repulsive forces, S 2, and h * 2 ho. In general, the relationship between h. and ho proves to be a function of the slit width H and the form of the H(h) isotherm, including the different components of disjoining pressure. In the case of partial wetting (03B80>0), it is possible to establish, by using equations (3) and (11), a relatively simple functional relationship between the parameter S and the equilibrium contact angle, as formed by the meniscus in a flat slit :
ponds
Only the conditions of partial wetting can be realized for equilibrium drops on a solid substrate. The parameter S is here unacceptable, and the contact angle value remains the only characteristic of wetting. For drops having a small-radius base, the equilibrium contact angle is also influenced by the line tension of the wetting perimeter, which will further be considered (Sect. 8). For the films on curved surfaces, as for example, for those on filaments or in capillaries, a variation in the pressure in a film due to the curvature of its surface adjacent to the gas phase must be taken into account [8].
3. Isotherms of
The values of h* are equal to 0 (curve 4, Fig. 1), when Pc y/H and Ho ho + 4 0. This, according to equation (3), corresponds to cos 90 = 1 and
=
disjoining
pressure of
wetting films.
90=0.
0, h * ho. If the positive values of 4 the values of h * increase, too, exceeding increase, the equilibrium film thickness. In the case of H > ho we may consider that y ~ 0394, and Pc = y /H. Then equation (9) transforms into a simpler form :
At
= =
The macroscopic approach allows one to ascertain what physical phenomenon controls the wetting or the non-wetting of the surface. For this purpose, we consider, as an example the isotherms of disjoining pressure of aqueous wetting films (Fig. 3). Here, curves 1-4 represent the dependences of film thick-
Then the ratio, S h*/ho, may be used as a of characteristic unified wetting, which is suitable in and of the cases complete partial wetting :
=
The value of S 0, corresponding to 90 0, separates the region of complete wetting (S > 0 ) from that of partial wetting (S 0). The higher the positive values of S the better will be the wetting. Thus, in particular, for the isotherms H(h) = A/hn>0, taking into account that Pc=J7o= A/hn0, instead of equation (11) we obtain the follow= =
ing expression :
At n
3, which corresponds
to the
that are stable due to Lifshitz S = 1.5, and h * = 1.5 ho. At n
wetting films dispersion forces,

2, which
corres-
Fig.
3.
Isotherms of the
disjoining
pressure
H(h)
of
water films on the solid surface.
979
h on disjoining pressure, or, which is the same, the capillary pressure of a liquid in equilibrium with the film. Curve 1 relates to water films on a quartz surface. The known experimental data [8] are indicated by points, while a solid line represents a computed isotherm [16], accounting for the effect of three components of disjoining pressure in the filmmolecular 03A0m, electrostatic ne, and structural IIS. The isotherm branches, where an/ah 0, respond to the stable states of the film. The water films on quartz within the range of thicknesses of 60 to 10 nm (curve 1) are unstable, and are not realized. A metastable state of thick 3-films (h > 100 nm ) forms as a bulk water layer is thinning out. The time of their transition into a thermodynamically stable state of thin a-films (h 10 nm ) depends on the closeness of the capillary pressure to the critical one, Pcr (when ahlah 0), and on the surface area of {3films. The larger the surface area the higher the probability of the formation of nuclei of the stable aphase in metastable /3-films. In the framework of the microscopic approach to the wetting theory, the p - a transition is denominated as a pre-wetting transition [5]. The existence of thick {3-films of water is conditioned by the electrostatic repulsion of charged surfaces of the films (ne:> 0 ). As in this case (Ho ho + ..d) > 0, the 03B2 films are completely wetted by water. Hereinbelow (Sect. 9) this case will be illustrated by comparison of the results of the experimental determination of the values of h * with the theoretical ones, calculated with equation (10).
ness on
=
the same sign) increase or approach each other in their values, this causes an increase in the electrostatic repulsive forces. As a result, the whole isotherm may be found within the H > 0 region, which must lead to the complete wetting (curves 3 and 4,
Fig. 3).
Thus, the conditions controlling the wetting of solid surfaces by water may be formulated as follows. those of electrostatic (03A0e) and Two effects can influence the shape of structural (77g) forces the isotherms H(h) of the wetting films of water. The dispersion forces depending on the spectral characteristics of water and the solid substrate, are less sensitive to the composition of an aqueous solution, temperature, and the surface charge. Three factors can influence the structural forces ; namely increasing the electrolyte concentration and raising temperature, which leads to decreasing the structural repulsion ; as well as through adsorption of molecules, which changes the character of the interaction of water molecules with the solid surface.
-
worsening of wetting, which is required, for example, for enhancing the flotation effectiveness, is usually attained through adsorption of ionic surfactants. In this case, it would be of importance that a surfactant would be selectively adsorbed on one of the film surfaces, imparting to it a charge, whose sign is reverse with regard to the charge of another
The
When the meniscus of bulk water is in contact with
a-films, the values of 4 in equation (3) may be
negative in view of the H(h) isotherm partly entering in the region of II 0. The change in the sign of the electrostatic component of disjoining pressure (curve 1) is connected with the different values of electrical potentiels 1/11 and .p2 of the film surfaces. The known tabulated data of electrostatic forces [17]
are used to calculate the lle (h) isotherms. The calculation gives that at h 600 , the repulsive electrostatic forces transform into attraction forces : lle 0. The repulsive forces do appear again as the film thickness decreases further, but these are already associated with the effect of molecular (H. > 0 ) and structural (77g > 0 ) forces. Calculations with equation (3) by using the theoretical isotherm 1 (Fig. 3) lead to the contact angle of water on quartz 90 5 [16], which is close to the experimental data. When the electrostatic repulsive forces are suppressed or the film surfaces have different signs of electrical potentials w, the /3-branch of the isotherm cannot be realized. In this case, the isotherm shifts into the range of II 0 (curve 2, Fig. 3), which causes, in accordance with equation (3), the worsening of wetting. On the contrary, when potentials wi and .p2 (of
=
surface. Thus the forces of electrostatic attraction arise (lle 0 ), which shifts the isotherm into the II 0 region. Adsorption of surfactants may also simultaneously lead to hydrophobization of the solid substrate, which reduces the repulsive structural forces. A high degree of hydrophobization may also reverse the sign of structural forces, too (77g : 0) [18, 19]. At 03A0e 0 and IIS 0, still higher contact angle values may be attained [16]. On the contrary, all the measures causing an increase in the forces of electrostatic and structural repulsion, improve the wetting. This aim is attained either by imparting a high potential of the same sign to the film surfaces and/or through hydrophilization of the substrate, as for example, by increasing the number of centers that are able to form hydrogen bonds with water molecules. Adsorption of nonionic surfactants or polymers leads to an additional effect of the sterical repulsion of adsorption layers. Thus, in each specific case, one can choose the optimum methods for controlling the wetting. A similar program of the theoretical calculation of contact angles on the basis of equation (3) was performed for alkanes on the Teflon surface [20]. In this simpler case, one could restrict oneself to the taking into account of only one molecular component of disjoining pressure llm 0. A good agreement of the calculation results with the experimental values of e o has thus been obtained. The region of the applicability of equation (2) is
980
limited by such thicknesses of wetting films, when these can yet be considered as a liquid phase. With poor wetting ( 00 > 90 ), a two-dimensional adsorption phase is formed on the solid surface, and the thickness of films does not exceed a monolayer. In this case, another expression following from Gibbs equation, correlating the adsorption T with a change in the interphase tension ysv depending on the adsorbate vapour pressure, p, is applicable [21, 22] :
where ps is the pressure of the saturated vapour. Equation (14) transforms into equation (2) with a formal replacement r holv., where vm is the molar volume of liquid, and with the use of a known thermodynamic relationship between the equilibrium vapour pressure over the film p and its disjoining pressure II [1, 8] :
=
The results of calculation of contact angles Fig. 4. 00 for aqueous KCI solutions of different concentrations (curve 1) and at different pH-values (curve 2)..
-
where R is the gas constant, and T is temperature.
4. Effect of concentration of
tants.
electrolyte
and surfac-
tact
difficulty involved in calculating the convalues angle 00 from equations (2), (3), and (14) resides in that at 90 > 0 the isotherms of disjoining pressure 77 (A) or the isotherms of adsorption r (p / p s) are partly disposed in the supersaturation
The main
region (H -- 0, p/ps > 1). The values of cos 0 0 are proportional to the value of d - that is, to the difference between the areas a and b in figure 3 (curve 1). As it is difficult to determine experimentally the isotherms in the supersaturation region, these parts of the isotherms can be determined only theoretically. Such calculations were made on the basis on the theory of surface forces for different
aqueous solutions [16, 23, 24]. In figure 4 are represented the results of calculation with equation (3) of contact angles 00 for the aqueous KCI solutions of different concentrations (curve 1) and different pH values (curve 2), while preserving in the latter case a constant ionic strength 10- 2 mole/1 [23, 24]. In carrying of the solution, 3 out calculation of 03A0e(A), the known dependences on the concentration of quartz-solution potentials t/1 1 and the aqueous solution-air interface t/12 were used. For the dispersion forces nm A/6 7Th3,the value 7.2 x 10-13 erg was adopted. A of constant A known exponential dependence [16] was adopted for the isotherm of structural forces Hs(h), whose parameters were used as adjusting ones in making the theoretical calculation of 00 agree with the experimental data [25]. An increase in the values of
=
9o with the electrolyte concentration (curve 1) may be explained by two causes : i) a reduction in the thickness of the boundary layers of water, which leads to a reduction in the structural repulsive forces ; ii) a decrease in the values of potentials "1 1 and 03C82, which decreases the electrostatic repulsive forces. A decrease in the contact angles as the pH value increases (curve 2), was caused mainly by an increase in the potentials 03C81and "2 of the film surfaces as a result of adsorption of the potential-determining OH-ions. We note that at 3 of 10- 2 mole/1 (curve 2), the dependences of contact angles are described by using the H(h) isotherm, including only two compoan electrosnents of disjoining pressure ; namely an a such molecular one. At tatical, and electrolyte concentration the structural forces are small (03A0s ~ 0), and these could not be taken into account. In figure 5 are presented the results of calculation of the isotherms of disjoining pressure of the wetting films of 10- 3 mole/1 KCI aqueous solution with additions of ionic surfactants. The isotherm indicated by curve 6, is the initial one, for KCI solution without addition of surfactants. In this case, the potentials of the quartz and film surfaces were assumed to be equal to : 03C81 = - 100 mV, and 8 25 mV, respectively. The value of 80 03C82 (as is also shown in Fig. 4) was obtained by calculating with equation (3). In the calculations done, the influence of surfactants was expressed as a variation in the values of potential 03C82 due to the adsorption of surfactant on the film-gas interface. Adsorption of an anionic surfactant increases the negative values of t/12 and leads to an improvement in wetting. Thus, 35 mV the calculated value for example, at 03C82 of 00 reduces to 7, while at 03C82 = - 45 mV it reduces to 5. A further increase in the absolute values of
-
981
The calculations are in good agreement with the results of direct measurements of the contact angles of KCI solutions with additions of anionic sodiumdodecylsulfate and cationic cetyltrimethylammonium bromide [26].
5. Contact
angle hysteresis.
cases a
Fig.
5. 11(h) isotherms calculated for a 10- 3 mole/1 KCI aqueous solution with additions of surfactants at 1/1 - 100 mV 75 Const. ; lP2 = -100 (curve 1) ; 03C82 45 65 (curve 2) ; 03C82 (curve 4) ; (curve 3) ; 03C82 35 (curve 5) ; 112 = - 25 (curve 6) and 03C82 03C82 + 100 mV (curve 7).
-
03C82 (curves 1-3) corresponds to the complete wetting of the quartz surface. Adsorption of a cationic surfactant charged positively the film-gas interface (03C82 + 100 mV). When the substrate remains negatively charged (curve 7), the contact angle increases up to 28 owing to the electrostatic attraction of the film surfaces (1-I,, 0).
=
to form equilibrium contact time. This is connected with angle requires long the retarded mass exchange between the bulk liquid and the thin wetting film [27]. The kinetics of transition into the equilibrium state is controlled by the viscous resistance of films and the diffusion of the , dissolved components, whose equilibrium concentrations in the bulk phase and in a thin film may be different. In view of this, at first a partial mechanical equilibrium can rapidly be established in the bulk part of a drop or meniscus in the absence of both the mechanical and the thermodynamic equilibrium with the film. A possibility of realization of a number of states of the mechanical equilibrium results in the phenomenon of the static hysteresis of the contact angle [13]. In this case the transition zone between the meniscus and the film plays a substantial role (Fig. 6). The state of its mechanical equilibrium, which is determined by fulfillment of the condition (1) breaks up at two values of the capillary pressure PC PA and PC PR, and acan cordingly at two values of the contact angle > and a receding advancing 8 A 8 0, 9 R e o. A plurality of the mechanical equilibrium states can be established in the whole interval between 8 A and 6 R. This interferes with determination of the equilibrium value of the contact angle [28]. The better the that is, the larger the film thickness wetting ho, the quicker occurs the transition into the equilibrium state, and the smaller the difference between the values of 8 A and 9 R.
a
-
In
number of
Fig.
6.
Influence of the transition
zone on
the static
hysteresis
of the contact
angle
in
flat slit :
(a) advancing
meniscus ; (b) receding meniscus.
982
considered earlier that the contact angle hysteresis is due either to the surface roughness or to the presence of the its chemical heterogeneity areas that differ from one another in the values of equilibrium contact angles. Examination of the stability of the transition zone has demonstrated that hysteresis is possible also on a smooth, uniform surface, too. In this case, the values of 6A and B R can also be determined on the basis of the isotherms of disjoining pressure, H(h) [13]. For the S-shaped isotherms (curves 1, Fig. 3) it was shown that the values of 8 A are within the range between 00 and 90 (depending in the n (h) Eq.), whilst the values of 9 R are close to zero, because a thick metastable a-film remains behind the retreating meniscus.
It
was
-
6.
Wetting
transition.
The macroscopic approach enables one also to examine the known phenomenon of the wetting transition, which is at present widely discussed within the framework of the microscopic approach
[5, 29, 30].

An increase in temperature T influences the H(h) isotherms, the consequence of which is transition from the complete wetting to the partial one. For quartz and glass substrates a worsening of the wetting by water is experimentally detected, as T is raised [31-33]. In view of the H(h) isotherms considered in figure 3, the transition of wetting, i.e., transition from isotherms 4 and 3 to isotherms 1 and 2, can result from a number of causes. The most probable is the influence of temperature on the structure of the boundary layers of water. It has experimentally been shown that the raising of temperature from 20 C to 70 C causes gradual disintegration of the special structure of the boundary
Here Il is the Planck constant, subdivided by 2 tut ; 8 (w ) are the frequency dependences of dielectric permeability, where co = i 03BE is the circular frequency along the imaginary frequency axis. The wetting transition corresponds to the change in the sign of the integral in equation (16). At sl > 83 in the frequencies range, making the main contribution to the integral (1016_16 17 rad/s), the values of 03A0m > 0. This is the case of complete wetting. At 03B51 E3, the transition to the partial wetting can occur. For the same liquid and substrate, this is possible when its temperature dependences E (T) are different. The values of e depend on density (the number of molecules per unit volume). Then, at a higher coefficient of thermal expansion of the liquid, than that of substrate, the values of E3 (larger than El at a low temperature) may become less than el as temperature T is raised. In accordance with equation (16), in this case an increase in T should lead to a reversal of the disjoining pressure sign and to the wetting transition. Since in this case the film thickness changes jump-wise, the wetting transition is of the first order. A possibility of the wetting transition ciue to the temperature dependence of the electric potentials of the film surfaces is also not excluded. However, this is possible at the adsorption energy values of potential - determining ions comparable to kT. In this case, as in that of structural forces, the film thickness can change gradually, and the transition may be of the second order.
7.
Dynamic
contact
angles.
of water and its approach to the bulk structure A [34]. decrease in the structural repulsive forces as T is raised, can enhance the extent of isotherm 3 (Fig. 3) entering the range of II 0. This may cause transition from the complete wetting to the partial one. The structural mechanism of transition is confirmed by a thermal reduction in the thickness of afilms [35], as well as by the fact that the higher the substrate hydrophilicity the higher the sensitivity of contact angles to variation in temperature [31]. The cause of wetting transition is quite different for the films of nonpolar liquids on dielectric substrates, when only dispersion forces are acting. As is known [36], the disjoining pressure of such thin films (h 10-20 nm) is determined by a difference in the dielectric function of the substrate 1 and the liquid 3 :
layers
Up to now, we have considered the state of thermodynamic or mechanical equilibrium of the meniscus-film system. When drops or menisci move, the distribution of pressures throughout the transition zone and the film changes, which cause also the curvature of the meniscus to change. If we continue an undisturbed by surface forces meniscus profile
until it intersects the substrate, then the contact angle values determined by that formal method will indicate the dependence on the flow rate v. The dynamic contact angles 8a begin to differ from the static values of 00 and exceed these at v > 10-3 cm/s
[37-39]. The theory of dynamic contact angles has so far been developed only for the case of complete
wetting, and when the meniscus advances at a constant velocity onto an equilibrium wetting film.
The numerical calculations were made on the basis of equation (1) [40]. Assuming that the condition dh/dx 1 holds also for the flowing transition zone, the following expression for the pressure gradient can be obtained :
983
The meniscus profiles Fig. 7. increasing velocity, V Const.

-
of
completely wetting liquid
in
flat
capillary. (a)
at rest ;
(b)
and
(c)
in motion at
an
where an approximation K is used for the surface curvature of liquid layer. In equation (17) P Po - Pc(x) - n(x) is the hydrodynamic pressure in a film. It is equal to the pressure in the gas phase Po minus the local values of the capillary and the disjoining pressure [41]. Expression (17) for dP /dx may now be substituted into the known equation of hydrodynamics of thin layers at v Const. [42] :
= =
d2h/dx2
where 11 is viscosity, h (x) is the local thickness of liquid layer, and ho is the thickness of an equilibrium film. A steady-state solution of the differential equation gives the profile h (x ) of a flowing liquid. In figure 7 are shown the computed profiles of a meniscus moving at different flow rates v through a flat slit [40]. For determination of ed, we have used a
part of the profile of
a
The dependences of the dynamic contact angle Fig. 8. on the meniscus flow rate. (1) H =10-2 cm, 9d =150 ; ho (2) H = 1.25 x 10- 3 cm, ho 74 ; (3) H = 1.25 x 10-5 cm, ho =16 .
-
constant
curvature, which is
directly adjacent
to the flow zone, and which is
therefore found still within the region of the sloping liquid layer dh/dx 1. In figure 8 are pregented computed dependences cos 0 d on the capillary number Ca v 11 / Y. Calculations were made for the isotherms H(h) A/h3 (with A 10-14 erg) and different half-widths of a slit H. As appears from figure 8, an increase in the flow rate causes an increase in the values of 03B8d. Differences of Od from the static value of 00 0, begin to show up at Ca > 10-3. The narrower the slit and accordingly the smaller the the thickness of an equilibrium wetting film ho smaller the meniscus profile is disturbed. Experimental investigations of dynamic contact angles for water on the quartz surface had shown that the values of ad began to exceed the equilibrium values ( 6 0 =10 ) at V 10-3 cm/s, thus attaining a value of 75 to 80 at v > 0.1 cm/s [39]. As appears from figure 9, these data only qualitatively agree with the theoretical ones (full line in Fig. 9). The quantitative discrepancy is associated with the fact
= = =
=
-
v, culs The experimental dependence 8d (v ) (shown by obtained for water in quartz capillaries 30 2013 35 03BCm in diameter. The solid line indicates the results of theoretical simulation [40].
Fig. 9. points)
that the theory has been developed only for the conditions of complete wetting and monotonous H(h) isotherms (curve 4, Fig. 3). For water on
984
quartz, another type of the S-shaped isotherm is characteristic (curve 1, Fig. 3), which substantially changes the shape of the transition zone, and, hence,
the qualitative calculation results. At a slow flow rate v forms the equilibrium contact angle 90 (Fig. 10a). For instance, the spontaneous immiscible displacement of one liquid by another, which better wets the capillary surface, occurs (at v 10-4-10-3 cm/s) at a constant value of and a constant 00 capillary pressure of meniscus In the case of the spontaneous displacement of Pc. tetradecane, dibutylphthalate, and toluene by water from the molecularly smooth quartz capillaries having the radii r 10-20 itm, the values of 00 remained constant and equal to about 83-86 [43]. When displacement takes place under the applied
=
8. Line tension.
pressure gradient (r 10-2-10-1 cm/s),the capillary pressure of the meniscus remains also constant, but equal to Pc = 2 y/r (Fig. 10b). The capillary pressure hinders the displacement, and reduces the flow rate. The condition of the complete wetting of a capillary by the liquid being displaced is dynamic, and is associated with the formation of thick, nonequilibrium films of liquid 2, after the rapidly receding meniscus. When the liquid flow is stopped, these films are ruptured and form slugs of 2 in liquid 1. At a very high flow rate (v -- 1 cm/s), the non-equilibrium films can be ruptured in the course of flow. As a result, for example, the displacing liquid disintegrates into droplets forming an emulsion. The thickness of nonequilibrium films may be evaluated with Derjaguins equation [44] :
The existence of the transition zone between the meniscus and the film leads to one more effect, which in the general case has been predicted by Gibbs namely ; the effect of line tension K [45]. In similarity with the surface tension y, when the transition zone between the liquid and its vapour is replaced by a plane of tension, the transition zone between the meniscus and the film may be replaced by a three-phase contact line. In distinction from y, the values of K may be both positive and negative ; they make the circular wetting perimeter tend to contraction in the first case, and to expansion in the second case. When the transition zone is replaced by the threephase contact line, an additional term, 03BA/r, is introduced in the Young quation :
-
where q is the viscosity of a liquid being displaced. Thus, in displacing dibutylphthalate by water, the
values of thickness h were obtained ranging from 70 to 300 nm ; whilst in displacing water by dibutylphthalate, the values of h varied from 10 to 40 nm, when the displacement rates vary from 10- 2 to
10-1cmls, respectively.
Immiscible displacement of fluid 2 by liquid 1. Fig. 10. (a) during spontaneous imbibition ; (b) rapid displacement under the effect of pressure gradient.
-
where r is the wetting perimeter radius. The line tension effect does show up the more noticeably the smaller the radius r. In the case of drops on the flat substrate, the positive values of K cause an increase in the values of 60, whereas the negative ones their decrease. For the flat films surrounded by a concave meniscus, the influence of the sign of K is inverse, which is just the cause of the introduction of a double sign before the last term in equation (20). For water and aqueous solutions the values of K are of about 10- 6-10- 5 dyne [43]. Thus, the term K /r in the right-hand side of equation (20) becomes noticeable at r 10-4-10-5 cm - that is, for drops and films of a very small radius. The line tension may, in particular, show up to a very noticeable on the initial stage of gas degree in flotation bubbles approaching particles [46], as well as in on the condensation of water on solid surfaces condensate nuclei formation stage [47]. The values of the line tension K for drops on the solid substrate were calculated as a difference between the values of y . cos 0, obtained while neglecting the transition zone and taking it into account [9,14]. Since the line tension arises due to the existence of the transition zone, it is clear that this difference is just associated with the term K /r. The line tension depends on the curvature radius r [9, 14]. This is connected with the dependence of the transition zone profile on the value of r. However, if the influence of curvature on the surface tension y does show up with the surface curvature radius on the order of distances between molecules, then in the case of line tension K the influence of the wetting perimeter curvature is felt at much larger values of r that is, on the order of the radius of action of surface forces.
-
985
analytical expression for K simplified isotherm of disjoining

An
was
obtained for
pressure
(6) [14] :
where a and t are the parameters of a model isotherm II (h ). At r ~ oo, the values of K tend to a constant value of Ka 2 y t tg 9 0. For small values of (Jo, when a gradual profile of the transition zone is formed, the values of K are negative, and amount to 10- 6 dyne in their order of magnitude, which is in agreement with the experimental data for foam films [48].
= -
9. Transition
zone.
Quite recently, the profile of the transition zone between the meniscus and a film on the solid substrate has for the first time been experimentally investigated, and the correctness of equations (8) and (10) of the theory of complete wetting has been verified [49]. The wetting films were formed on a polished quartz plate 1 (Fig. 11) by making approach to it the meniscus of liquid in a tube 2 about R = 1 mm in radius. The liquid is sucked from the tube through slots 3. The thickness of a film ho in the equilibrium state with the meniscus surrounding it was obtained from the intensity of reflected light. Depending on the capillary pressure of meniscus Pc the radius of the film ro amounts to several scores microns. Simultaneously with measuring thickness, the interference rings from meniscus were photographe. This allowed its profile h (r) to be determined, where r is the radial coordinate. An analysis of experimentally obtained profiles of the meniscus for water and aqueous KCI solutions of low concentration has shown that their continuation that is, the does not intersect the substrate plane The undisturbed takes proplace. complete wetting file was calculated with the Laplace equation for an axisymmetrical meniscus :
-
Formation of an equilibrium wetting film Fig. 11. having the thickness ho and radius ra on the solid substrate 1 in pipe 2, when the liquid is being sucked off through slit 3 under the effect of the capillary pressure.
-
0 (r) is the current angle value, for which ah/ar (Fig. 11). Solution of equation (20) allows the coordinate ro to be determined, at which the theoretical profile of meniscus passes through a minimum. Its position corresponds to the thickness h * (compare Figs. 1 and 11). As has been demonstrated above, the value of S h */ho is a quantitative characteristic of complete wetting.
where
tg e
For the aqueous KCI solutions, Z. M. Zorin has obtained the following values of ho and h * ; namely for a 10- 4 molell solution ho =1025 , and h.=1450; for a 10-3 mole/l solution, ho
=
710 . This leads to the h */ho 590 , and h * values equal to 1.41 and 1.2, respectively. The experimental values of S may be compared with theoretical ones. Large values of the equilibrium thicknesses of films ho and the complete wetting indicate that the long-range electrostatic repulsion forces predominantly act there. Assuming the potentials of quartz surface t/1 1 and film surface t/12, it is possible to calculate from Devereux and de Bruyn tables [17] the isotherm of electrostatic forces 03A0e(h). On the basis of references [50, 51], for 10- 4 molell KCI solution, it is possible to assume 45 mV, while for a that t/11 150 mV, and t/12 10- 3 mole/1 solution, t/11 = - 125 mV, and 112 45 mV. Under the condition of t/1 Const., the calculated isotherms lle (h ) are linearized in logarithmic coordinates with the correlation coefficient being equal to 0.996. This enables one to approximate the /3-part of the isotherm by a power-function, H = Alh. For a 10- 4 mole/1 KCI solution n = 2.87 ; and for a 10- 3 mole/1 KCI solution the value of n was obtained to be equal to 6. Substituting these values of n into equation (12), we obtain the theoretical values of the parameter S, which are equal to S 1.5 for the 10- 4 mole/1 concentration, and to S = 1.2 for the 10- 3 mole/l. These values satisfactorily agree with the experimental ones. Similar results were obtained for water and for a number of aqueous solutions [49]. At a low electrolyte concentration, the extension of the transition zone amounts to about 5-10 03BCm, which allows its investigation by means of optical methods. On the basis of a photogramme of the interference patterns, Z. M. Zorin has calculated the profile of the meniscus and that of the transition zone (curve 1, Fig. 12). Curve 2 gives the profile of the meniscus that has not been disturbed by surface
=
986
12. The profiles of the transition zone between the meniscus and a flat film having the radius r 10 03BCm, both determined experimentally (curve 1) and calculated theoretically (curve 3). Curve 2 indicates the profile of an undisturbed meniscus (10-4 mole/1 NaCI).
Fig.
where x is the tangential coordinate. The width of the slit H, which is equivalent to the round cell, was determined by equalizing the capillary pressure of the cylindrical meniscus in a flat slit (Fig. 1) and that of the meniscus in the tube (Fig. 11). This gives in both cases the equality of the disjoining pressure Ho Pc in the films. The profile of the transition zone calculated with equation (23) (at n 2.87, and A 9.54 x 10-12, if77is expressed in dynelcm2,and h in cm) is indicated by curve 3 in figure 12. Agreement between the experimental and the theoretical profile is a satisfactory one.
=
=
10. Conclusion.
current state of the macroof the scopic theory wetting shows that the disjoining can be successfully used for isotherms pressure theoretical evaluation of contact angles. This approach is limited to solid surfaces which are sufficiently lyophilic that the contact angles did not exceed a value of about 30 2013 40. The object of further investigations will consist in the obtaining of experimental isotherms of disjoining pressure for wetting films of different liquids on different substrates. This will enable one to render more correct the values of the parameters of the isotherms, which will make the results of the calculation more reliable.
Thus, the review of the
forces, as plotted with the use of equation (22). A minimum on this curve determines the layer thickness h *. Then the theoretical profile of the transition zone may also be attempted to be constructed. However, the theory of the transition zone has so far been developed only for a meniscus in a flat slit, and for the isotherms of the type II A/hn [11]. A corresponding equation for the profile of the transition zone h (x ) has the following form :
=
References
V., CHURAEV, N. V. and MULLER, Forces (Nauka, Moscow) 1985 ; New York, London) 1987. [2] SULLIVAN, D. E., J. Chem. Phys. 74 (1981) 2604. [3] TARAZONA, P., EVANS, R., Mol. Phys. 48 (1983)
[1] DERJAGUIN,
V. M.,
B.
[11] NEIMARK,
[12] [13] [14] [15] [16]
[17]
Surface (Plenum Press,
799.
[4]
[5]
VAN
DE
SWOL, F., HENDERSON, J. R., Phys. Rev. Lett. (1984) 1376 ; J. Chem. Soc. Faraday Trans. p. 2,82 (1986) 1685. GENNES, P. G., Rev. Modern Phys. p. 1, 57
53
(1985)
827.
[6] FRUMKIN, A. N., Zh. Fiz. Khim. 12 (1938) 337. [7] DERJAGUIN, B. V., Zh. Fiz. Khim. 14 (1940) 137. [8] DERJAGUIN, B. V., CHURAEV, N. V., Wetting Films
(Nauka, Moscow)
1984.
[9] CHURAEV, N. V., STAROV, V. M., DERJAGUIN, B. V., J. Colloid Interface Sci. 89 (1982) 16. [10] DERJAGUIN, B. V., STAROV, V. M., CHURAEV, N. V., Kolloid Zh. U.S.S.R. 38 (1976) 875.
[18]
[19]
A. V., HEIFEZ, L. I., Kolloid. Zh. U.S.S.R. 43 (1981) 500. DERJAGUIN, B. V., CHURAEV, N. V., J. Colloid Interface Sci. 54 (1976) 157. MARTYNOV, G. A., STAROV, V. M., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 39 (1977) 472. STAROV, V. M., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 42 (1980) 703. GOOD, R. J., KUO, M. N., J. Colloid Interface Sci., 71 (1979) 283. CHURAEV, N. V., DERJAGUIN, B. V., J. Colloid Interface Sci., 103 (1985) 542. DEVEREUX, O. F., DE BRUYN, P. L., Interaction of Plane-Parallel Double Layers (MIT Press, Cambridge Mass.) 1963. ISRAELACHVILI, J. N., PASHLEY, R. M., J. Colloid Interface Sci. 98 (1984) 500. LUZAR, A., BRATKO, D., BLUM, L., J. Chem. Phys. 86 (1987) 2955.
987
[20] HOUGH, [21] [22]
[23]
D. B., WHITE, L. R., Adv. Colloid Interface Sci. 14 (1980) 3. DERJAGUIN, B. V., in Proc. Second Intern. Congr. of Surface Activity (Butterworths, London) 2 (1957) p. 153. ADAMSON, A. W., LING, I., in Contact Angles, Wettability and Adhesion (Amer. Chem. Soc., Wash.) 1964, p. 57. CHURAEV, N. V., in Research in Colloid Science
[36]
E. M., I. E., LIFSHITZ, DZYALOSHINSKII, PITAEVSKII, L. P., Adv. Phys. 10 (1959) 165. [37] HOFFMAN, R. L.,J. Colloid Interface Sci. 50 (1975)
228.
[38] GRIBANOVA, E. V., MOLCHANOVA, L. I., Kolloid. Zh. U.S.S.R. 40 (1978) 30. [39] BEREEZKIN, V. V., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 44 (1982) 417. N. V., V. M., CHURAEV, [40] STAROV,
KHVOROSTJANOV, A. G. , Kolloid. Zh. U.S.S.R.
(FAN, Tashkent) 1987, p. 70. [24] DERJAGUIN, B. V., CHURAEV, N. V., Langmuir 4 (1987). [25] KORNILEV, I. N., ZORIN, Z. M., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 46 (1984) 892. [26] ZORIN, Z. M., ROMANOV, V. P., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 41 (1979) 1066. [27] STAROV, V. M., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. 38 (1976) 100. [28] PENN, L. S., MILLER, B., J. Colloid Interface Sci. 77 (1980) 574. [29] TARAZONA, P., EVANS, R., Surface Sci. 125 (1983)
[30]
298. TELO DO
GAMMA,
687.
M. M.,
EVANS, R., Mol. Phys. 48
(1983) [31] WHALEN, J. W., LAI, K.-Y., J. Colloid Interface Sci. 59 (1977) 483. [32] GRIBANOVA, E. V., Kolloid. Zh. U.S.S.R. 45 (1983) [33] ZORIN,
[34]
422. Z. M., ESIPOVA, N. E., ERSHOV, A. P., in Problem of Shapeformation and Phase Transition (Kalinin University, U.S.S.R.) 1985, p. 3. DERJAGUIN, B. V., CHURAEV, N. V., in Fluid Interfacial Phenomena (Wiley, London) 1986, p. 663.
[35] DERJAGUIN, B. V., ZORIN, Z. M., CHURAEV, N. V., SHISHIN, V. A., in Wetting, Spreading and Adhesion (Acad. Press, London) 1977, p. 201.
(1977) 201. [41] DERJAGUIN, B. V., CHURAEV, N. V., J. Colloid Interface Sci., 66 (1978) 389. [42] BRETHERTON, F. P., J. Fluid Mech. 10 (1961) 166. [43] ABBASOV, M., ZORIN, Z. M., CHURAEV, N. V., Kolloid. Zh. U.S.S.R. (in press). [44] DERJAGUIN, B. V., Acta phys. chim. U.S.S.R. 20 (1945) 349. [45] SCHELUDKO, A., TOSHEV, B. V., PLATIKANOV, D., in The Modern Theory of Capillarity (Khimia, Leningrad) 1980, p. 274. [46] MINGINS, J., SCHELUDKO, A., J. Chem. Soc. Faraday Trans. 75 (1979) 1. A., CHAKAROV, V., TOSHEV, B., J. SCHELUDKO, [47] Colloid Interface Sci., 82 (1981) 83. [48] KOLAROV, T., ZORIN, Z. M., Colloid and Polymer Sci. 257 (1979) 1292. [49] ZORIN, Z. M., PLATIKANOV, D., KOLAROV, T., Colloids and Surfaces 22 (1987) 147. [50] CHURAEV, N. V., SERGEEVA, I. P., SOBOLEV, V. D., DERJAGUIN, B. V., J. Colloid Interface Sci. 84 (1981) 451. [51] EXEROVA, D., ZAKHARIEVA, M., in Surface Forces in Thin Films and Dysperse Systems (Nauka, Moscow) 1972, p. 234.
39
Colloid & Polymer Sci. 256, 784-792 (1978) 1978 Dr. Dietrich Steinkopff Verlag, Darmstadt ISSN 0303-402X / ASTM-Coden: CPMSB (formerly KZZPAF)
Equipe de RechercheC. N. R.S. associ& a l' Universitd Paris V, Paris (France)
Structural disjoining pressure in thin film of liquid crystals I. : Thermodynamics and Frederiksz transition with surface f i e l d s * )
E. Perez, J. E. Proust, and L. Ter-Minassian-Saraga
With 8 figures (Received January 4, 1977)
List of symbols
hT
w', Aw'
K
hs
q0 superscript A f h T V F ,/ dr
: film transition thickness : critical film stability thickness : : : : : : : : : : : : : : : film phase area force per unit area film thickness temperature volume free energy disjoining pressure work performed in varying h capillary pressure specific surface entropy film tension chemical potential surface tension number of molecules per unit surface of film at thickness h film perimeter upper surface lower surface surface free enthalpy free enthalpy specific bulk free enthalpy in film of L.C. specific bulk free enthalpy of L.C. specific excess free enthalpy in thin film of L.C. liquid crystal director angle between director and x, y plan azimutal angle elastic energy density specific bulk free enthalpy of the undistorted film
Au Aa
Fe]
Fs
r)a /s B (h)
An
6 I h*
Pc
s A # a F(h)
: isotropic and anisotropic contribution to the interfacial tension of L.C. : average elastic constant of nematic liquid crystal : non elastic contribution to the film tension : elastic contribution to the film tension : elastic torque : surface torque : classical disjoining pressure : structural disjoining pressure : Van der Waals parameter : optical anisotropy of the liquid crystal : optical retardation : light intensity : reduced thickness
Introduction
T h e p r o p e r t i e s of d i s p e r s e s y s t e m s in l i q u i d s d e p e n d o n t h e n a t u r e a n d state of t h e i r i n t e r faces a n d of t h e t h i n films s e p a r a t i n g t h e p a r t i c l e s of t h e d i s p e r s i o n . A l t h o u g h i n t e r m o l e c u l a r forces a r e s h o r t r a n g e , t h e r e is i n c r e a s i n g belief t h a t n e x t t o t h e s e interfaces, t h e initial p e r t u r b a t i o n of t h e t w o phases in c o n t a c t takes p l a c e o v e r a finite t h i c k n e s s of t h e l i q u i d i n t e r m e d i a t e film. I t has been assumed that the perturbation may struct u r e (1, 2) o r d e s t r u c t u r e (3, 4) t h e l i q u i d . T h e effect of t h e d i s p e r s i o n c o n s t i t u e n t s p o l a r i t i e s has b e e n e m p h a s i z e d , a l t h o u g h r e c e n t exp e r i m e n t s (4, 5) h a v e d e m o n s t r a t e d t h a t this p a r t i c u l a r p r o p e r t y is n o t essential f o r i n t e r facial s t r u c t u r i n g . F o r t h i n films, Derjaguin (1, 2) has p o s t u l a t e d the o c c u r r e n c e o f a s t r u c t u r a l d i s j o i n i n g p r e s sure r e l a t e d to t h e i n t e f f a c i a l s t r u c t u r i n g . T h e p o l a r t h e r m o t r o p i c l i q u i d crystals o r L . C . , w h i c h h a v e a s t r o n g m o l e c u l a r field,
: subscript h: subscript 0: w G : W' : Wb : W(h,z) : 0 ~b Wa Wu : : : : :
*) This paper has been presented at the 50th EUCHEM Conference Chemistry of Interfaces, Collioure (France), April 28-30, 1976. and Vie International Liquid Crystal Conference, KENT, U.S.A. 23-27 August 1976. Part II of this paper has been published in the issue : No 255, 1133--1135 of this journal. W 704
Perez et aL, Structural disjoiningpressure in thin film of liquid crystals, I.
785
are systems appropriate for the study of interfacial structuring. Elastic forces resist any structure perturbation or distortion in bulk so that an interracial structuring will propagate away from the interface. The properties of LC in bulk have been extensively studied (6, 7). A review of the recent results on the effect of surface forces on LC orientation specificity may be found in ref. (7). In the present paper we deal only with the surface forces of physical chemical nature. A direct approach to the study of these forces is the study of wettability or adhesion of the system LC-substrate solid (8, 9, 10, 11) or liquid (12) which provide information on LC substrate interracial tension variation with changes in interracial structure. Finally, surface structuring and orientation in thin films on solid substrates has been considered as an epitaxial growth (2). In the present paper, the thermodynamics of asymmetric nematic liquid crystal, NLC, thin films on water are discussed, using the concept of surface tension anisotropy (13, 14) i.e. variation of N L C interracial tension with molecular orientation at the interface. Derjaguin (1), Sheludko (15) and Everett (16, 17) thermodynamic approaches are used to deduce the appropriate film thermodynamic potential and disjoining pressure. Using the formalism of Jenkins and Barratt (13) and the results of Parsons (14), we discuss the particular equilibrium disjoining pressure for N L C films of various thicknesses for various surface orientations under two alternative assumptions : A) thickness independent surface orientation B) thickness dependent surface orientation. In both cases, it is predicted that very thin films adopt a non-distorted, epitaxial, structure whereas thick films are distorted. However, the transition thickness hT from the undistorted to the distorted N L C film is different in cases (A) and (B). This difference originates specifically in the different behaviour of the disjoining pressure in the two cases (A) and (B). This transition in structure, analogous to the well known Frederiksz transition (18) observed when external electric or magnetic fields are applied to N L C slabs, is directed by surface forces in
the case of asymmetric thin films of N L C on substrates. In the conclusion, we compare our results with those obtained by different methods for different systems.
Thermodynamic potential thin films and structural disjoining pressure of liquid crystals
The system of figure 1 is analogous to that of
Ash, Everett and Radke (16, 17).
The thickness h of fluid corresponds to a force A f exerted on the plates. Work is performed to vary h: dT = - - A f d h . Derjaguin's disjoining pressure (1, 2) is by definition: ~7=--(1/A) (8FISh)A,T,v. As d F = d z , it follows that ~ = f and is measured by the capillary pressure (1, 2) Pc. The variation of the free energy F~ of the film phase, noted ~, is equal to
dF~ = S ~ d T q- A d A - f A d h q- #dn~ - - p d V ~
[1]
where A is the film tension defined by Sheludko (15). Integration at constant T, h, #, A leads to
F ~ = A A q- #n~ - - p V ~ .
[2]
A relevant thermodynamic potential ~ for film phases is therefore:

q) = F ~ q- p V ~
-
n~# = A A = 2 ~ A
= g * -- n#
[3]
where ~ is the surface tension of the film (16, 17) and ~ corresponds to defined values of T, P, #. From [2] and [3] the variation of is:
d ~ = -- S ~ d T q- A d A - - f A d h -- n~d#
[4]
~P TpVlJn -~h-I fluid

Fig. 1. Film of fluid between two surfaces. Uniform temperature T, pressure p, chemical potential #; volume V, n mole number, A surface area, h separation, f force on plates Af.
786
Colloid and Polymer Sdenee, Vol. 256 No. 8 ( 1 9 7 8 )
and the definition of the equilibrium disjoining pressure is deduced: 1 [Oq}]

A -~T,A,~,~
Let the specific bulk free enthalpy in the film W' = W' (h, z) be: W' (h, z) = W(h, z) q- Wb [9]
=f=~.
[5]
A corresponding Gibbs-Duhem equation for films is obtained from [3] and [4] - - s ~ d r - - da - - f d h - V(h)de + hdp = 0 [6] which leads to a second expression for ~7
_
where W~ = 9# is the specific bulk free enthalpy of the very thick and extended domain and W(h, z) is the excess free enthalpy density in the thin film. From [8] and [9] we obtain for G the expression:
h
[ ]
d -d~ -
T,~,e
-- 2
[d l L-N-J
u s = A ~ {e~h + Y g(h,z)dz + wo(h)

T,~,. [7]
+ w~(h) ) .
[10]
when either q~ or A are known, the disjoining pressure ~ may be deduced using [5] or [7]. N e m a t i c liquid crystal (NLC) thin films
The specific film potential # is obtained from [9] and [10]: (q~/A) = A ----[Ge(h)IA] -- ql*h [11]
a) Thermodynamic potential and equilibrium of structure Jenkins and Barratt (13) formalism is used
to find for a domain of NLC of thickness h, cross section A and perimeter 65 as shown in figure 2. The film domain area is large compared to its thickness. Except on the perimeter, where a defect occurs, the molecules are uniformly oriented on each molecular layer of the NLC film. It is assumed that the contribution of the defect free enthalpy to the free enthalpy of the domain is negligible. The film is asymmetric and has an upper surface free enthalpy w~(h) and a lower surface free enthalpy wo(h) dependent on the local molecular orientations which may vary with h. The film free enthalpy is equal to :
A the specific film thermodynamic potential for unit surface A is obtained from [10] and [11]
h
A = f [W(h,z)]dz Jr wo(h) q- wh(h).

o
[12]
When h-~ oo, we have Am = (Wo q- Wh)~. The disjoining pressure is obtained from [5] [7] and [12]:
h
O(w~+ wo)
Oh ~,p,~
= Pc.
[13]
When Pc = O, the equilibrium thickness of the film domain is given by the solution of the equation:
h
G* = y W'dv + y [~,o(h) + w~(h)]dA.

V A
[8]
0
-- [O(w~+w)]T,e,v=O'oh
,"
Wo A
z<
[14]
b) tT,xpression for A and ,1 of N L C

NLC are characterized by the strong mutual alignment of their molecular axis ~ along an axis of uniaxial symmetry. In figure 3, a simplified case (6, 7) is shown. Figure 4a shows an undistorted domain. Figure 4b shows a distorted domain such that
substrate
Fig. 2. D o m a i n of t h i n film o n a substrate : A I = area; V I = v o l u m e . W ' = bulk specific free e n t h a l p y ; w/~, Wo = surface specific free e n t h a l p y
Perez et aL, Structural disjoiningpressure in th& fi& of liquid crystals, I.
787
~I I / /
/ / /
//
/!/
~z = sin O(z)
Fig. 3. Orientation of the uniaxial symmetry axis ~; na = components of ~; 0: azimutal angle
nx,
dynamic potential Aa, the contribution Au being independent of Oh or 0o. The equilibrium value of (dO~dr) or O(z) is obtained by minimizing A a as usual (6, 7), using the equilibrium boundary conditions discussed in references (13, 14). It is obtained: 0 ~ - - ~ z O r d z - - - ~
Oh -- Oo dO, Oh -- Oo
[20]
the angle 0 between h and the x direction varies from the base to the top of the film. This distortion adds an elastic positive contribution to the free enthalpy density of the undistorted film of thickness h. This contribution is very roughly (6, 7):
we = g K
1
At each surface, the equilibrium orientation of NLC molecules is the result of balance of two torques :
--*Pel:
elastic torque opposing disalignment of molecules
[15]
--/'8: surface torque opposing an increase of surface free enthalpy or surface tension. The torques are equal respectively to
Where Kis the average elastic modulus of the NLC. Then, if the specific bulk free enthalpy of the undistorted film is Wn (h, z)
W ( h , z ) = W~(h,z) -- 9/* q- We(h,z).
a) F e l = K
"~z
[16]
b) I s , h = Awh sin 20h; F,, o = Awo sin 200. [211 From [20] and [21] the equilibrium molecular orientation at the film surface are:
Eliminating W ( h , z ) from [12] and [16]

h 0
o
4- f g e ( h , z ) d h + wo(a) wh(h).
[17]
a) b)
K (Oh -- 00)
--
Awh sin 20h
Awh < 0 Awo > 0
According to references [13, 14], the anisotropy of the interracial tension is expressed as follows :
wh = wh + Awh sin20h Wo = Wo + Awo sin20o
K (0h -- 0o) h -- Awo sin 200
[22] From [19] and [20] we deduce:
[18]
where wh (z~/2) < wh (0) and Wo@/2) > Wo(o). From [15], [18] and [17]:
h
A = Au + -~-
(0h -
00) 5
' + Awh sin20h [23]
-k Awo sin20o
A = A~, + Aa = f [W'(h,z) -- 9#]dz + wf,

0 h
, + -x~ + Wo
f
0
0h i i ;
\ dz ] dz + Awo sin20o
~, /
/
Oh So
a
+ Awh sin20h.
[19]
o0
b
A u, A a are respectively the contributions independent or dependent on 0. Distortion or change in Oh or 0o modifies the thermo-
Fig. 4. Thin films of NLC. a) undistorted Oh= Oo= 0; b) distorted 0h # 0o = O
788
Colloidand Polymer Science, Vol. 256 No. 8 (1978)
and from [7] and [24] allowing for [22a] and [22b], the following expression is obtained for the disjoining pressure:
Material and experimental methods

The NLC : 4--4'-pentylcyanobiphenyl (5CB) has been kindly offered by BDH. its purity is better than 99 %. The substrate: distilled water 3 X is swept before depositing the NLC. The films are formed by spreading 2--6 /~l of NLC on various areas: 4, 10, 40 sq cm, of substrate contained in an optical cell at 23 C. The cell is examined between crossed Nicols, using a commercial polarizing microscope and transmitted light. In general, 2 0 X and 100X magnifications were used and 200 X exceptionally. The optical anisotropy of 5CB is 3 n ~ +0,15 (19). The optical retardation for an incident light crossing the film of figure 4a of thickness h is equal to d = h An. The figure 5 represents the experimental setup. A compensating plate L equivalent to a retardation on d0 is placed between polarizer and sample. For a uniform film, with L absent, the intensity of the light crossing the sample
~ = ~ q- ~s = -- ( - ~ b ) T,p,u
-t-2K (Oh --h 200)2 q-
2K_h (Oh --
0o) "-~h'O
[24]
F~ is the classical disjoining pressure (1, 2) of the undistorted NLC or Fa =B(h)h -a. The last terms are standing for r/s, the structural disjoining pressure.
c) Stability and transition critical film thickness

The boundary conditions [22aJ have been discussed in reference (14). It is found that, when 00 = 0 , according to the figures 4a and 4b, the solution of [22a] is a) 0 a = 0 b) Oh 0 h <hT= h >hT-K 2 dwh
K 2dwh"
[25]
Below a transition thickness hT the structure of figure 4a is stable. Above hT the structure of 4b is stable. This thickness hT depends on the ratio (K/Awh) i.e. on the magnitudes of the torques Fel and Fs. From [24] and [25] it follows that the structural transition will modify the disjoining pressure expression as follows: a) ~ - b) ~7 -B (h) K O~ ha q - ~ - h---g h > h T
'g i x
. . . . . . . .
Bha (h)
h < aT
[26]
The structural disjoining pressure is positive and stabilizes the film shown in figure 4b only. From [22a], [25a and 25b] and [26] a final expression is obtained for [26].
B (h)
Aw~ sin~ 2 Oh
[27]
The condition for NLC thin film stability is discussed below. It leads to a critical stability thickness hs ~ hT.
Fig. 5. Experimental device for observation of liquid crystal film; P i polariser; All or Aa_: analyser parallel or perpendicular to the polariser; h: film thickness; ~: liquid crystal director defining the optical axis; // 2 0 = polar angle; : azimutal angle
Perez et al., Structural disjoining pressure in thin film of fiquid crystals, I.
789
parallel or perpendicular to the polarizer P respectively is equal to a) I / / = I o l l (sin2 4)~ sin ( ~ - ~ ) , [28]
0,7
0,6 0,5
Off I E
b) Iz = Io (sin z 2) 2 sin ( - ~ - )
0,2
O,l
1 2 3 ~ 5 6 7 8 h. microns 9
where Io is the incident light intensity (fig. 5). When white light is used, a coloured image is obtained. When polarizer and analyzer are parallel, the zeroth order of the colour sequence is white (silver), It is black for perpendicular polarizer and analyzer. The other colours are determined by (d/k), the respective intensity depending on . When L is inserted (fig. 5), corresponds to the following values of: ~' = d + d o d"=~--d0 for = - - ~ - : k =
7~
Fig. 6. Variation of the average retardation ~ with the film thickness h. I) undistorted. II) distorted films
for - -
4 ~n"
For distorted films (fig. 4b), the average optieal retardation is calculated using the variation of the polar angle of molecular axis 0 (z) in the film given by [20]. For uniform samples, do = d cos~0 (20). We assume that for a distorted sample the average retardation g is equal to:
h
The values of h were deduced from the known volumes v of spread NLC and the area a of the substrate. The error on h is smaller than 15 %. The range of values of h is 0.5--10 #m. The optical retardation 8 measured is smaller than 1 #m. The observed colours correspond to the 1st, 2nd, 3rd orders in Newton colour sequence. They allow for the evaluation of 8 with a satisfactory accuracy ~o as ~ ___N - ~ - where ,to ~ 0,56 #m for white light. The figure 6 represents the results. The slope of I in figure 6 is equal to An =0.15. The slope of H in the same figure is equal to 0.07 and close to (An~2)=0.075. The intercept of H with the ordinate airs is equal to 0.08. The transition I-~II occurs at h=2.2 -t-0,1 m. Interpretation of the results From [29], [20] and [25], we obtain the average retardation
= An j" cos ~ O(z) dz 0 where h is the real film thickness.
[29]
Results
a) Orientation of N L C
mo/ecu/es at the film
interfaces
The thin films studied display many defects of orientation. These are described in reference (12). However, the extension of the uniform domains of the order of 1 mm is large compared to the width of the defect walls or lines which are of the order of 10 #m. We can assume that their contribution to the free enthalpy of the uniform domains is negligible. To find the value of Oh, we measured ~ for various known values of h and used equations [20] and [29].
1 1 Anh = ~ zln h q- ~ 0---7-sin 2Oh.
[30]
Two cases are considered for the distorted films shown in figure 4b. A) Oh = ~/2 independent of h. Then
=7
dnh.
B) 0 < Oh < ~/2 when h varies.
[31]
These cases correspond respectively to weak surface forces or "anchoring" of mole-
790
Colloidand PolymerScience, VoL 256 No. 8 (1978)

At a thickness called h, of mechanical stability limit, Oh = x/4, the structural disjoining pressure is maximum and equal to F].
cules at the film free surface. The substrate anchors strongly the molecules. We eliminate sin 2 Oh from [30] using [22] and [25] and obtain:
Aw~
~s
8 =
K (~/4) 2
_
Anh ( -- 2 1
K) 2b AWh
An = ~ (h + hT)
2K
(hs)2
[34]
[32]
in which the definition [25] of hT has been used. The plot of 8 vs. h according to [31] or assumption 1 should provide a line with zero intercept at the origin and slope equal to (An~2). The analogous plot according to [32] or assumption 2 should provide a line with a positive intercept equal to (An/2)hT. For undistorted films or Oh =0o = 0 (see fig. 4a) the retardation becomes:
From [34] and Aw=4.5 10 -6 Jm-2 we find hs = 1,78 #m and r/] = 1 Jm -a. For 0 <Oh < ~ / 4 and h <h,, according to the mechanical stability condition and to [34], the stability criterium.
&
d~ - --
2 Aw~
K
d0h
cos 2 0 h [35]
sin 2 0h d~ > 0.
The distorted film of figure 4b verifies [35] for -~- < Oh < -~- i.e. for h > h8 = 1.8 #m while according to [25] distorted structures may persist in the range h>hT=l.1 #m. The lower limit of distorted films is determined by their mechanical stability, rather than by the stability of their structure (conditions [26]).
:g 9~
= Anh.
[33]
The slope of the plot g vs h is equal to An. In the figure 6 we show this plotting. It is seen that: a) For h < 2.2 #m, the films verify the equation [33] (line I). b) For h >2.2 #m the results verify the equation [32] corresponding to the distorted films with Oh varying with h. From the intercept it is obtained for the structural transition thickness: hT =1.1 #m. From the definition [25] of hT, we find for the anisotropy factor of surface tension of 5CB, Awh =4.5 10 -6 Jm-e, for K = 10-'1N. Therefore the expression of the surface tension of 5CB may be written: [lSbis) wh = w ~ - 4,5 10-6 sin 2 0~ [Jm-2]. The anisotropic contribution is negligible compared to the first isotropic contribution of the order of 38, 10-aJm -z. However, the observed structural t~ansition thickness hT is smaller than the discontinuity in a observed at h=2.2 #m (fig. 6). This disagreement may be due to our method o averaging the orientations in equation [29]. The structural disjoining pressure exists only for h > 2.2 #m. It has a maximum value for Oh = ~/4, according to its expression [27] and assuming
that ~/a < < ~]s.
erqs
5.
crn32341",///
t-
~/4
Oh
zr/2
Fig. 7. Variation of the structural disjoining pressure ~s with film molecular orientation 0h at the free surface of film. I) Stable distorted film; unstable mechanically. II) Mechanically stable distorted films
231,5678 h IJm Fig. 8. Variation of structural disjoining pressure ~Ts of 5CB with the reduced thickness h* = (h/he)
Perez et aL, Structural disjoining pressure in thin film of liquid crystals, L
791
The "Frederiksz" transition in asymmetric thin films may have to conform two sets of conditions: mechanical [35] and thermodynamical or physicochemical: [22] and [25]. In figure 8, the results ~7 s shown in figure 7 have been plotted h* =h/hr. For 1 <h* < 1.64. The film of figure 4b and corresponding to a 0h changing with h would be unstable according to [34] unless the first in [24] has a significant contribution. However, if Oh =~z/2, according to [26a] in this region the positive contribution of ~]s enhances the film stability. Therefore the validity of assumption 1 and 2 may be checked by operating in the range 1 < h * <1.64. In this range which is of the order of 1 #m for 5CB, letting [B(h)] __ 10-z0 J. The dispersive contribution would be of the order of 10-2 Jm-2 and much smaller than the values shown in the figures 7 and 8, If the results display a decrease in Ws for 1 < h * < 1 , 6 4 , the orientation of molecules changes with film thickness and the interfacial tension or the enthalpy verifies equation (18bis). If the limit of film stability and film distortion structure disappear at the same film thickness h*, the orientation of the surface molecules Oh is fixed. These two cases corresponding to assumptions A and B above are named weak and strong anchorages respectively.
D i s c u s s i o n and conclusion
tension of NLC equal to Awh=4,5 10 -6 Jm-2 independent of film thickness and structure. Therefore, we deduce that Awh is relevant to local, short range surfaceforces only. As two of us (25) have found that 5CB has no polar contribution to its surface tension, we may tentatively conclude that balance between bulk elastic and surface orientation forces is achieved only when surface polar forces are small. When the short range polar forces are present and much stronger than elastic ones, they fix the surface molecular orientation [9, 25, 26]. They are much larger than the longrange Van der Waals interaction energies film-substrate according to recent theoretic estimations [27, 28, 29] without exception
[24].
Surface molecular orientation in asymmetric NLC films is also very important for thin film mechanical stability [30]. The disjoining pressure expression as a function of the angle of molecular orientation at the film surface has been deduced. For the NLC film studied, 5CB, characterized by a small interracial anisotropy, lower disjoining pressures and equilibrium thicker films can be predicted, contrary to asymmetric NLC films strongly anchored at both surfaces.
Acknowledgement
The authors t h a n k A . Sheludko, E. Manev, J. B. Ivanov, 19. G. de Gennes, D. Langevin, M. Kleman and A . Regner for interesting discussions.
A Frederiksz transition is found in the absence of external electrical or magnetic macroscopic fields which in our case are replaced by surface short range forces acting at the boundary of asymmetric thin films of 5CB formed at the surface of water. Evidence is found that the interracial tension anisotropy at the film-water interface is stronger than at the film vapour interface. At the last interface, the weaker anisotropy is balanced by the bulk elastic anisotropy. Then the molecular orientation at this interface is dependent on film thickness beyond a predicted transition thickness h T = l , 1 #m and observed one at hT =2,2 #m. This variation of molecular orientation allows asymmetric equilibrium distorted NLC films to be formed on water. It corresponds to an anisotropic factor contribution to the free surface
Note added in proof

Subsequent studies of the structural disjoining pressure u n d e r different conditions - small area films have s h o w n that the surface molecules of L. C. may be anchored more strongly than in the present experiments.
Summclry
T h i n films (0,2 -- 10 #m) of a nematic liquid crystal N L C : 4 - - 4 ' - p e n t y l c y a n o b i p h e n y l (5CB) have been spread o n large surfaces of water. The orientation of the molecules has been examined between crossed polarizer and analyzer and by studying the average retardation of white l i g h t as a function of film thickness. A t h r -----2.2/~m a Frederiksz transition takes place owing to the balance of elastic and surface forces acting o n the film and determining the molecular orientations in bulk at the surface as a function of films thickness. F r o m this variation and hT, an anisotropy of surface tension of 5CB of the order of 10 -6 J m -2 is found. This weak anisotropy
792
Colloid and Polymer Science, I/o/. 256 No. 8 (I978) 16) Everett, D. H., C. J. Radke, Absorption at Interfaces, ACS Symposium, Series 8, Ed. K. L. Mitlal, 1, (1975). 17) Ash, S. G., D. H. Everett, C. J. Radke, J. Chem. Soc. Faraday Trans. II, 69, 1256 (1973). 18) Frederiksz, V., V. Zwetkoff, Acta Physicochemica U. S. S. R., 3, 895 (1935). 19) Hareng, M., Personal communication. 20) Born, M., E. Wolf, Principles of Optics, 695699 (Oxford 1964). 21) Langevin, D., A . M. Bouchiat, J. Phys., 33, C 1-77 (1972). 22) Langevin, D., A . M. Bouchiat, Mol. Cryst. & Liq. Cryst., 22, 317 (1973). 23) Sicart, jr., J. Phys., 37, L.-25 (1976). 24) Ryschenkow, G., M. Kleman, J. Chem. Phys., 64, 404 (1976). 25) Perez, E., J. E. Proust, 5CB CR Acad. Sci., 282, 559 (1976). 26) Kleman, M., C. E. Williams, Philos. Mag., 28, 725 (1976). 27) Richmond, P., L. A . While, Mol. Cryst. Liq. Cryst., 27, 217 (1973). 28) de Gennes P. G., C. R., Hfibd. Sdan. Acad. Sci., 271, 469 (1970). 29) Dubois-Violette, E., P. G. de Gennes, J. Phys., 36, (1975). L-255.
determines also the variation of the structural disjoining pressure and the film stability which is discussed at length. A thermodynamic potential for this films is suggested. Reference 1) Derjaguin, B. B., Disc. Faraday Soc. 18, 26 (1954). 2) Derjaguin, B. V., AT. V. Churaev, J. Colloid Interface Sci. 49, 249 (1974). 3) Adamson, A. IV., J. Colloid Interface Sci. 27, 180 (1968). 4) Everett, D. H., G. H. Findenegg, J. Chem. Thermodynamics, 1, 573 (1969). 5) Ash. S. G., G. H. Findenegg, Special Disc. Faraday Soc., 1, 105 (1970). 6) de Gennes, P. G., The Physics of Liquid Crystals (Oxford 1974). 7) Priestley, E. B., P. J. Wojtowicz, Ping Sheng, Introduction to Liquid Crystals (Oxford 1975). 8) Proust, J. E., L. Ter-Minassian-Saraga, C. R. Acad. Sci., Paris, 279, Serie C-615 (1974). 9) Proust, jr. E., L. Ter-Minassian-Saraga, J. Phys. Colloq., 36, Cl-77 (1975). 10) Proust, J. E., L. Ter-Minassian-Saraga, E. Guyon, Solid State Comm., 11, 1227 (1972). 11) Proust, J. E., L. Ter-Minassian-Saraga, Colloid & Polymer Sci. 254, 492-496 (1976). 12) Proust,J. E., E. Perez, L. Ter-Minassian-Saraga, Kolloid Zeits. & Zeits. ftir Polymere (presented 1976). 13) Jenkins, J. T., P. J. Barratf, Quart. J. Mich. Appl. Math., 27, 111 (1974). 14) Parsons, J. D., Mol. Cryst. Liq. Cryst., 31, 79 (1975). 15) Sheludko, A., Adv. Colloid Interface Sci., 1, 391 (1967).
Authors' address : Dr. L. Ter-Minassian-Saraga Equipe de Recherche C. N. R. S. associde ~t l'Universit~ Paris V, U. E.R. Biom&dicales 81, Paris (France)

Analysis of The Effects of Marangoni Stresses PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analysis of The Effects of Marangoni Stresses PDF

Uploaded by

Copyright:

Available Formats

3972

Langmuir 2005, 21, 3972-3980

Microflow in an Evaporating Droplet

Langmuir, Vol. 21, No. 9, 2005 3973

2ur P 1 (rur) + 2 ) r r r r z 2uz P 1 uz r + 2 ) r r r z z

Langmuir, Vol. 21, No. 9, 2005

3 1 z 2 z 3 [1 + ()(1 - r 2)-()-1] + 41- t 3h 2 h

uz Ma (abr b-1 + 2(1 - a)r ) tf r

3 1 1 z z 2 [(1 - r 2) - (1 - r 2)-()] 2 - 2 + 81- r h h t r h02h R

Mah0h 3z z 2 (abr b-1 + 2(1 - a)r ) (12) 2R 2h h 2

3 1 1 [(1 - r 2) - (1 - r 2)-()] 81- t r t)tfh0h g(Rr ,tf 2R

Microflow in an Evaporating Droplet

Langmuir, Vol. 21, No. 9, 2005 3975

Mc ) J [C(u) + L]T ) F K(T)u ) F(T)

(17) (18) (19)

Langmuir, Vol. 21, No. 9, 2005

Microflow in an Evaporating Droplet

Langmuir, Vol. 21, No. 9, 2005 3977

Langmuir, Vol. 21, No. 9, 2005

Ma (abr b-1 + 2(1 - a)r ) tf kBT u r|z)h uz R D r

dT Ma )(abr b-1 + 2(1 - a)r ) R dr tf

Microflow in an Evaporating Droplet

Langmuir, Vol. 21, No. 9, 2005 3979

h02h 1 Mah0h 1r (J ()(1 - r 2)-()-1 + 1) (abr b-1 + 2 2 R 2 2R 2(1 - a)r )

3 1 1 [(1 - r 2) - (1 - r 2)-()] 81- t r

kBTtfh0h ) 100 D 4R2

100 4 0.01 (1 10-5) (4 10-14)(0.036 108)

Langmuir, Vol. 21, No. 9, 2005

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/rphysap:01988002306097500

to ~~(H.h0)1/2 [10, 11].

Thus, for example, for slits H = 10

cm, f =10- 4 cm.

Fig. 1. eo, when

equation (1) [13] :

expression with the known

ah/ax 1), equation (1) following manner [14] :

energy of the film, AG

is the film thickness at II

Fig. 2. Simplified isotherm of and its real analogue (b).

conditions, an increase in the radius of action of surface forces, t, or a decrease in to causes an

that are stable due to Lifshitz S = 1.5, and h * = 1.5 ho. At n

wetting films dispersion forces,

water films on the solid surface.

When the meniscus of bulk water is in contact with

a-films, the values of 4 in equation (3) may be

where R is the gas constant, and T is temperature.

Influence of the transition

meniscus ; (b) receding meniscus.

[5, 29, 30].

The meniscus profiles Fig. 7. increasing velocity, V Const.

completely wetting liquid

to the flow zone, and which is

analytical expression for K simplified isotherm of disjoining

Thus, the review of the

Surface (Plenum Press,

[20] HOUGH, [21] [22]

EVANS, R., Mol. Phys. 48

Equipe de RechercheC. N. R.S. associ& a l' Universitd Paris V, Paris (France)

: subscript h: subscript 0: w G : W' : Wb : W(h,z) : 0 ~b Wa Wu : : : : :

Perez et aL, Structural disjoiningpressure in thin film of liquid crystals, I.

A relevant thermodynamic potential ~ for film phases is therefore:

~P TpVlJn -~h-I fluid