A density functional study of liquid liquid interfaces in partially miscible systems

Similar documents
Experiment 5: Phase diagram for a three-component system (Dated: April 12, 2010)

Mean Field Flory Huggins Lattice Theory

Lecture 4: Thermodynamics of Diffusion: Spinodals

Lecture 3: Models of Solutions

We will study the temperature-pressure diagram of nitrogen, in particular the triple point.

FUNDAMENTALS OF ENGINEERING THERMODYNAMICS

Chapter 8 Maxwell relations and measurable properties

Chapter 10 Phase equilibrium

Phase Diagrams & Thermodynamics

Thermodynamic database of the phase diagrams in copper base alloy systems

7. DYNAMIC LIGHT SCATTERING 7.1 First order temporal autocorrelation function.

vap H = RT 1T 2 = kj mol kpa = 341 K

4. Introduction to Heat & Mass Transfer

ORGANIC LABORATORY TECHNIQUES NEVER distill the distillation flask to dryness as there is a risk of explosion and fire.

PHASE TRANSITIONS IN POLYMERIC AND MICELLAR SYSTEMS

Thermodynamics of Mixing

Name Class Date. In the space provided, write the letter of the term or phrase that best completes each statement or best answers each question.

Lecture 22: Spinodal Decomposition: Part 1: general description and

6 Cloud droplet formation and Köhler theory

Universitätsstrasse 1, D Düsseldorf, Germany 3 Current address: Institut für Festkörperforschung,

Chapter 7 : Simple Mixtures

Chapter 12 - Liquids and Solids

Modern Construction Materials Prof. Ravindra Gettu Department of Civil Engineering Indian Institute of Technology, Madras

Liquid-Liquid Extraction (LLX)

States of Matter CHAPTER 10 REVIEW SECTION 1. Name Date Class. Answer the following questions in the space provided.

Differential Relations for Fluid Flow. Acceleration field of a fluid. The differential equation of mass conservation

Statistical Mechanics, Kinetic Theory Ideal Gas. 8.01t Nov 22, 2004

VAPORIZATION IN MORE DETAIL. Energy needed to escape into gas phase GAS LIQUID. Kinetic energy. Average kinetic energy

Kinetics of Phase Transformations: Nucleation & Growth

Surface tension of strong electrolytes

The Ideal Solution. ChemActivity T15

Ab initio study of gas-phase sulphuric acid hydrates containing 1 to 3 water molecules

Chapter Outline: Phase Transformations in Metals

A New Technique Provides Faster Particle Size Analysis at a Lower Cost Compared to Conventional Methods

Steady Heat Conduction

Molecular Dynamics Simulations

NUMERICAL ANALYSIS OF THE EFFECTS OF WIND ON BUILDING STRUCTURES

Symmetric planar non collinear relative equilibria for the Lennard Jones potential 3 body problem with two equal masses

CHEMICAL ENGINEERING AND CHEMICAL PROCESS TECHNOLOGY - Vol. I - Interphase Mass Transfer - A. Burghardt

Characterization of Porous Glasses by Adsorption: Models, Simulations and Data Inversion

k 2f, k 2r C 2 H 5 + H C 2 H 6

Modeling and Simulations of Cavitating and Bubbly Flows

Interfacial Stress, Interfacial Energy, and Phase Equilibria in Binary Alloys

5. Which temperature is equal to +20 K? 1) 253ºC 2) 293ºC 3) 253 C 4) 293 C

Stability of Evaporating Polymer Films. For: Dr. Roger Bonnecaze Surface Phenomena (ChE 385M)

Phase. Gibbs Phase rule

Unit 3: States of Matter Practice Exam

PARTICLE SIMULATION ON MULTIPLE DUST LAYERS OF COULOMB CLOUD IN CATHODE SHEATH EDGE

Lecture 24 - Surface tension, viscous flow, thermodynamics

Lecture 1: Physical Equilibria The Temperature Dependence of Vapor Pressure

Extension of the OpenFOAM CFD tool set for modelling multiphase flow

inter.noise 2000 The 29th International Congress and Exhibition on Noise Control Engineering August 2000, Nice, FRANCE

Final Exam CHM 3410, Dr. Mebel, Fall 2005

μ α =μ β = μ γ = =μ ω μ α =μ β =μ γ = =μ ω Thus for c components, the number of additional constraints is c(p 1) ( ) ( )

Chapter 5: Diffusion. 5.1 Steady-State Diffusion

Why? Intermolecular Forces. Intermolecular Forces. Chapter 12 IM Forces and Liquids. Covalent Bonding Forces for Comparison of Magnitude

Lecture 12: Heterogeneous Nucleation: a surface catalyzed process

Thermodynamics of Adsorption

Phase Equilibria & Phase Diagrams


Iron-Carbon Phase Diagram (a review) see Callister Chapter 9

Phase Transformations in Metals and Alloys

Computational Fluid Dynamics (CFD) and Multiphase Flow Modelling. Associate Professor Britt M. Halvorsen (Dr. Ing) Amaranath S.

Everest. Leaders in Vacuum Booster Technology

Physics 9e/Cutnell. correlated to the. College Board AP Physics 1 Course Objectives

Chapter 8. Phase Diagrams

Interpretation of density profiles and pair correlation functions

How do single crystals differ from polycrystalline samples? Why would one go to the effort of growing a single crystal?

Water: cavity size distribution and hydrogen bonds

So T decreases. 1.- Does the temperature increase or decrease? For 1 mole of the vdw N2 gas:

Nuclear Magnetic Resonance

6. 2. Unit 6: Physical chemistry of spectroscopy, surfaces and chemical and phase equilibria

Objectives: 1. Be able to list the assumptions and methods associated with the Virial Equation of state.

Experiment 12E LIQUID-VAPOR EQUILIBRIUM OF WATER 1

AP Chemistry A. Allan Chapter 8 Notes - Bonding: General Concepts

SURFACE TENSION. Definition

Chemistry 13: States of Matter

REPRESENTATIONS OF CLOUD DROPLET ACTIVATION PROPERTIES FOR SURFACE ACTIVE ORGANIC AEROSOL. Kuopio, Finland

Thermodynamics. Chapter 13 Phase Diagrams. NC State University

Chapter 1: Chemistry: Measurements and Methods

Exp 13 Volumetric Analysis: Acid-Base titration

SOLUBILITY, IONIC STRENGTH AND ACTIVITY COEFFICIENTS

Binary H 2 SO 4 -H 2 O homogeneous nucleation based on kinetic quasi-unary nucleation model: Look-up tables

18th International Congress of Mechanical Engineering

A. Hyll and V. Horák * Department of Mechanical Engineering, Faculty of Military Technology, University of Defence, Brno, Czech Republic

= atm. 760 mm Hg. = atm. d. 767 torr = 767 mm Hg. = 1.01 atm

Study the following diagrams of the States of Matter. Label the names of the Changes of State between the different states.

BIFURCATION PHENOMENA IN THE 1:1 RESONANT HORN FOR THE FORCED VAN DER POL - DUFFING EQUATION

SOLIDIFICATION. (a)formation of stable nuclei. Growth of a stable nucleus. (c) Grain structure

CHAPTER 12. Gases and the Kinetic-Molecular Theory

Formation of solids from solutions and melts

Fundamentals of grain boundaries and grain boundary migration

DIFFUSION IN SOLIDS. Materials often heat treated to improve properties. Atomic diffusion occurs during heat treatment

Orbits of the Lennard-Jones Potential

LIPE- Institut National des Sciences Appliquées Toulouse, France Contact:

Plate waves in phononic crystals slabs

Dimensional analysis is a method for reducing the number and complexity of experimental variables that affect a given physical phenomena.

7. Gases, Liquids, and Solids 7.1 Kinetic Molecular Theory of Matter

Lecture 9 Solving Material Balances Problems Involving Non-Reactive Processes

KINETIC MOLECULAR THEORY OF MATTER

Transcription:

JOURNAL OF CHEMICAL PHYSICS VOLUME 110, NUMBER 12 22 MARCH 1999 A density functional study of liquid liquid interfaces in partially miscible systems Ismo Napari Department of Physics, P.O. Box 900014, University of Helsinki, Finland Ari Laaksonen a) Department of Applied Physics, University of Kuopio, P.O. Box 1627, 70211 Kuopio, Finland Vicente Talanquer Facultad de Química, UNAM. México, 04510, D. F. México David W. Oxtoby The James Franck Institute, The University of Chicago, 5640 S. Ellis Ave., Chicago, Illinois 60637 Received 28 April 1998; accepted 21 December 1998 Liquid liquid interfaces and nucleation in partially miscible Lennard Jones LJ mixtures are considered using density functional theory. We present phase diagrams, interfacial liquid vapor and liquid liquid profiles, and gas liquid as well as liquid liquid surface tensions for two types of mixtures having different mixing rules for the LJ energy parameter. A simple local density approximation does not give oscillatory behavior at the liquid liquid interface, but a more realistic weighted density approximation does show this behavior. Both approaches also give a total density minimum near the interface, comparable to that found in molecular dynamics and integral equation studies. Finally, we calculate the density profiles and free energies for critical nuclei in liquid liquid phase separation. 1999 American Institute of Physics. S0021-9606 99 50612-0 I. INTRODUCTION The structures of planar liquid liquid interfaces have received recent attention in the form of molecular dynamics 1 and integral equation 2 studies. It has been demonstrated that the density profiles at such interfaces exhibit oscillatory structure, and that a strong minimum in total density appears at the interface between two immiscible fluids. In related work, the structures of nanometer-sized clusters composed of species with low mutual solubility have been investigated. These studies are especially interesting due to the experimental observation 3 that two vapors exhibiting a miscibility gap in the bulk liquid phase can form critical liquid nuclei at all compositions. A molecular dynamics study 4 demonstrated density oscillations and a minimum in the total density, qualitatively similar to those seen in bulk interfaces, in cylindrically symmetric clusters having one end enriched in one component and the other end in the other component. A square-gradient density functional approach to critical clusters composed of partially miscible species, 5 on the other hand, was capable of qualitatively explaining the nucleation observations, but neither density oscillations other than those due to surface enrichment at the liquid vapor interface nor the total density minimum in the center of the clusters were seen. In order to further study the nucleation behavior of partially miscible species, we have used density functional theory in both a local density approximation and a nonlocal weighted density approximation. We first look at the bulk a Also at: Department of Physics, P.O. Box 90014, University of Helsinki, Finland. phase diagram and the structure of the planar liquid liquid interface in partially miscible systems; to our knowledge, no papers characterizing liquid liquid interfaces at this level of theory have appeared. We then turn to the structure and free energy of small nucleating clusters of one phase surrounded by a bulk phase with a different composition. Both the local density approximation and the weighted density approximation produce a total density minimum at the interface, but only the latter more accurate approach gives the complex oscillatory structure seen also in simulations. The two methods give quite similar results for the free energy of the planar interface and for the critical nuclei. II. DENSITY FUNCTIONAL APPROACH Consider a simple model for a binary mixture with a Helmholtz free energy given by F i r i dr i r ln i r dr r s r drdr 2 ij r r i, j i r j r i,j 1,2, 1 with 1/kT, where k is Boltzmann s constant and T is the absolute temperature; F i (r) is a functional of i (r), the local average densities of particles of type i. The first term in this equation represents the ideal gas contribution to the free energy of the system, while the second term is the excess 0021-9606/99/110(12)/5906/7/$15.00 5906 1999 American Institute of Physics

J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. 5907 free energy of the hard sphere reference system in a weighted density approximation WDA. 6 The third term introduces the effect of the long-range interactions, ij, between particles ij in a mean field approximation. The excess free energy per particle (r) is assumed to be local in the weighted densities i r dr w i r r i r, 2 with a packing fraction r i 6 ii 3 i r, 3 for the particles of diameter ii, and a total number density s r i i r. 4 The weighting function w i ( r r ) is chosen to be 6 3 w i r 4 d ii r H d ii r, d ii where H(r) is the Heaviside step function, and d ii is the Barker Henderson hard-sphere diameter of particles i. 7 In the particular case in which the weighting function w i ( r r ) is taken to be equal to the Dirac delta ( r r ), Eq. 1 reduces to the corresponding free energy in the so called local density approximation LDA. The equilibrium properties of the fluid can be obtained by minimizing the grand potential 5 i r F i r i i dr i r, 6 at constant chemical potentials i, under appropriate boundary conditions. 8 The requirements 0 i r i 1,2, 7 result in the Euler Lagrange equations i ln i r 1 U i r, 8 with U i r r dr r i r s r w i r r j dr ij r r j r, that can be solved by standard numerical procedures. Equation 6 together with Eq. 8 determine the nature of the coexistence and critical manifolds for the binary mixture, and the density profiles i (r) for inhomogeneous equilibrium states. The surface tension of liquid vapor or liquid liquid interfaces of area A can be calculated by taking the grand potential difference A i r h i, 9 10 where h i is the grand potential of the homogeneous phases at coexistence; similar calculations lead to the work of formation of the critical nuclei * when working with metastable states. In this latter case, we found it convenient to work in a closed system with a fixed number of particles N i, volume V, and temperature T. 9 The solution profiles i (r) for the critical nucleus are then a minimum of the Helmholtz free energy in Eq. 1 when the system is constrained to have a fixed number of particles N i dr i r i 1,2. 11 When these conditions are combined with Eq. 8, the Euler Lagrange equations for i (r) in a closed system result in ln i r ln N i ln dr e U i r 1 1 U i r. 12 III. MODEL AND RESULTS FOR BINARY MIXTURES Our model binary mixture is characterized by the Lennard Jones LJ interaction potential LJ ij r 4 ij 12 ij r 6 ij r i,j 1,2, 13 where ij and ij are characteristic LJ energy and length parameters. We consider only the case of particles of the same size in our calculations ( ii ), with energy parameters 11 22. The values of the energy parameter 12,orof the mixing parameter * ( 11 22 2 12 )/ 11, determine the mixing properties of the mixture of interest at a given temperature T* kt/ 11. 5,10 12 In the context of a conventional Weeks, Chandler, and Andersen WCA perturbation scheme, 13 the reference and perturbative potentials assume the form and ref ij p ij ij LJ r ij r r min, 0 r r min ij r r min, LJ r r r min, ij where r min 2 1/6 is the distance at which LJ ij (r) exhibits its minimum. In the WCA approach, the reference system is replaced by a system of hard spheres with a temperature dependent diameter d(t*). This mapping from LJ to hard spheres in not unique; we follow Zeng and Oxtoby in their study of nucleation of LJ binary mixtures by taking 14 d T* a 1T* b 14 a 2 T* a 3 with a 1 0.56165, a 2 0.60899, a 3 0.92868, and b 0.9718. The Helmholtz excess free energy per particle of the reference hard sphere system (r) is taken to be that of the binary Carnahan Starling forms of Mansoori et al. 15 that in our case reduces to

5908 J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. FIG. 2. Liquid liquid phase diagrams in the mole fraction temperature plane (x 1l T*). Lower curve: * 0.5. Upper curve: * 1.5. FIG. 1. Liquid vapor phase diagrams in the total density temperature plane ( T*) a * 0.5, b * 1.5. r 4 3 1 2 with 6 d T* 3 1 r 2 r. A. The phase diagrams 15 16 Binary mixtures with * 0 are characterized by a twodimensional liquid liquid phase coexistence manifold in the temperature-activity space (T*,a 1,a 2 ) of the system, with a i e i; this surface is bounded by a line of triple points where the liquids coexist with a vapor. For symmetrical mixtures ( ii, ii / 11 1.0), the triple line ends at an upper critical end point UCEP for low values of * where the liquid phases become critical, and at a tricritical point for higher values of this mixing parameter. 16 Figures 1 and 2 show typical phase diagrams for LJ binary mixtures with * 0.5 and * 1.5. The liquid vapor phase diagrams in Fig. 1 are represented as projections on a total density temperature plane ( 1 2 T*), while liquid liquid phase diagrams in Fig. 2 are shown in a mole fraction temperature plane (x 1l 1 / T*). The liquid vapor coexistence curve for the fluid with * 0.5 in Fig. 1 a shows a liquid vapor critical point at a temperature T* 1.28 and a UCEP at T* 1.07, the latter of which is revealed by a dent in the liquid branch of the curve. Below this temperature, the mixture presents a liquid liquid miscibility gap as shown in the phase diagram in Fig. 2. When the immiscibility of the two components is further increased by increasing *, the dent in the phase diagram of Fig. 1 a leads to a sharp maximum on the liquid side of the phase diagram. At higher values of *, the UCEP and the liquid vapor critical point merge into a tricritical point where the two liquid phases and the vapor become simultaneously critical. This is the case of the mixture with * 1.5 whose liquid vapor and liquid liquid coexistence manifolds end at a tricritical point at T* 1.78 see Figs. 1 b and 2. B. The liquid vapor interface Interfacial density profiles between coexisting phases can be calculated by solving Eq. 8 iteratively to find the equilibrium solutions at a given temperature and composition; the surface tension is then given by the grand potential difference in Eq. 10. We performed these calculations using both a weighted density approximation WDA and a local density approximation LDA in which i(r) i (r) is assumed in calculating the Helmholtz excess free energy per particle (r) in Eq. 15. Results from both approximations will be compared systematically in this paper. In general, the WDA and LDA lead to similar liquid vapor density profiles, and corresponding surface tensions, for all mixtures with * 0 at high temperatures. Partially miscible mixtures tend to exhibit surface enrichment of the

J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. 5909 FIG. 3. Liquid vapor surface tension as a function of temperature T* for equimolar symmetrical mixtures with * 0.5. Results for the WDA and the LDA are depicted in this figure. minority component in the liquid phase in contact with the vapor. Adsorption at the liquid vapor interface reduces the value of the surface tension in comparison to that of the pure liquid under the same conditions. The behavior of the liquid vapor surface tension as a function of temperature T* along the triple line for a mixture with * 0.5 is illustrated in Fig. 3. In this case, the mixture exhibits a UCEP at T* 1.07 at which the rate of decrease of with T* changes drastically. This particular feature disappears in those mixtures whose triple line ends at a tricritical point. Differences between the WDA and LDA results become stronger when the temperature is lowered. The vapor density FIG. 5. Liquid liquid surface tension as a function of temperature T* for mixtures with * 0.5. Results for the WDA and the LDA are depicted in this figure. increases faster at the interface in our WDA calculations and the density profiles become sharper. At very low temperatures the WDA profiles start showing weak density oscillations on the liquid side as depicted in Fig. 4 for a mixture with * 0.5; similar behavior has been observed at the liquid vapor interface in pure fluids. 17 Surface tensions calculated using the LDA are always higher than those from WDA, with an increasing difference up to 10% at lower temperatures see Fig. 3. C. The liquid liquid interface We have also studied the properties of liquid liquid interfaces along the triple line and, at higher pressures (P* 3 P/ 11 ), on the liquid liquid coexistence manifold for several mixtures. FIG. 4. Liquid vapor density profile of the majority component in the liquid phase for a mixture with * 0.5 at the triple point (T* 0.4). LDA profiles correspond to the same thermodynamic conditions. FIG. 6. WDA density profiles for a liquid liquid interface in a binary mixture with * 1.5 at T* 0.7 and P* 1.0.

5910 J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. FIG. 7. Density profiles for critical nuclei in a binary mixture with * 1.5 at T r 0.5, P* 1.5, and a x a,2 0.8915, b x a,2 0.9628, c x a,2 0.9847. The liquid liquid binodal is located at x a,2 0.5 and the liquid spinodal at x a,2 0.9999 in this system. LDA profiles correspond to the same thermodynamic conditions. For mixtures with * 0.5, the vapor phase never intrudes between the coexisting liquids but a density minimum at the liquid liqud interface is observed at all temperatures. The surface tension of the planar liquid liquid interface decreases monotonically with T*, vanishing finally at the UCEP as shown in Fig. 5. Both LDA and WDA calculations lead to similar results close to the critical point but larger deviations 5% are observed at low temperatures (T r T*/T c * 0.5) where the WDA liquid liquid density profiles start showing oscillatory structures with periods close to the diameter of the LJ spheres see Fig. 6. As one increases * towards more immiscible mixtures, a drying transition occurs for temperatures higher than a drying temperature T d * where the vapor phase at the triple point intrudes between the coexiting liquids total drying. This is the case for mixtures with * 1.5 that show a tricritical point at T c * 1.78 and a drying transition at T d * 1.15, beyond which total drying is observed at the liquid liquid interface. The drying transition shifts to higher temperatures when the LDA, instead of the WDA, is used to evaluate the liquid vapor and liquid liquid interfacial tensions. In general, the LDA leads to less sharp interfacial profiles but higher surface tensions an increase of 5% to 10% as we move to lower temperatures ; these effects are less pronounced than in liquid vapor interfaces. For mixtures with * 0, the WDA liquid liquid density profiles exhibit oscillatory structures and a total density minimum at the interface Fig. 6 comparable to those found in computer simulations 1 and integral equation studies 2 of systems that have either no attraction or only a weak attraction between unlike particles. The LDA, in contrast, does not lead to any oscillatory density pattern and shows a density gap between coexisting phases that tends to be less deep but broader than that of the WDA. Our study shows that the

J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. 5911 WDA density oscillations become less pronounced as the temperature is increased along the triple line toward the drying transition, while the density gap between bulk liquids gets broader and deeper. Similar results are obtained when reducing the pressure P*, approaching the triple line along the liquid liquid coexistence manifold at a constant temperature. The oscillations and the density are always reduced by increasing the temperature at constant pressure. Oscillations in the density at liquid vapor interfaces are not observed in our binary mixtures except at very low temperatures. Hence, the intrusion of the vapor phase at the drying transition destroys the oscillatory patterns and also changes the effect that immiscibility has on the structure of the density profiles at a given reduced temperature T r T*/T c *. For high pressures P* our results show that the oscillations and the gap become more pronounced when * is increased at fixed T r * and P*. At low pressures, however, the proximity of the triple line and the possibility of crossing a predrying transition invert the trend: density oscillations are better defined in more miscible mixtures where there is no drying transition or where it occurs at higher temperatures. D. Critical nuclei The basic properties of the liquid liquid interfaces in immiscible binary mixtures are reflected in the properties of those nuclei of a stable liquid phase, rich in one component, that start forming from a metastable liquid phase rich in the other. Critical droplets in mixtures with * 0 also exhibit oscillations in their density profiles whose particular features depend on the cluster size. Assuming critical nuclei of spherical symmetry, one can find stable inhomogeneous solutions of Eq. 12 at a constant temperature T* and pressure P*. In this work, we consider the case of metastable liquid states rich in component 1 and follow the properties of the critical nuclei as a function of the activity fraction x a,2 a 2, 17 a 1 a 2 where a i e i and the chemical potentials i are given by Eq. 8. The values of the activity fraction x a,2 are not necessarily close to the values of the mole fractions x 2 2 / in our nonideal mixtures. Figures 7 a 7 c show typical density profiles in a WDA for critical nuclei in an immiscible binary mixture as one moves from the liquid liquid coexistence line towards the liquid spinodal. As can be seen, critical clusters exhibit a layered structure that is more prominent in the smaller clusters. The period of the density oscillations is very close to the diameter of the LJ spheres and these oscillations can induce a deep minimum or a sharp maximum in the density of particles at the center of the droplet depending on its size. LDA density profiles for the corresponding critical clusters are also shown in these figures for comparison. The work of formation of critical clusters * as a function of the activity fraction x a,2 for a binary mixture with * 1.5 at T r 0.5 and P* 1.5 is shown in Fig. 8; results FIG. 8. Work of formation of critical nuclei in a binary mixture with * 1.5 at T r 0.5 and P* 1.5 as a function of the activity fraction x a,2. Results for the WDA and the LDA are depicted in this figure. for the WDA and LDA are included in this figure. The WDA always leads to lower values of * between 5% and 10% than the LDA and to smaller critical sizes under the same conditions. The difference between the results obtained by these two approximations diminishes when T* is increased or P* is decreased, conditions that reduce the size of the oscillations in the density profiles. IV. CONCLUSIONS Density functional theory DFT on different levels of approximation LDA and WDA has been applied to study the properties of planar liquid vapor interfaces, liquid liquid interfaces and critical nuclei of one phase in the other in highly immiscible binary mixtures. Our results for the WDA are comparable with those obtained by molecular dynamics simulations of similar systems. 1 The theory predicts the appearance of oscillatory density profiles on each of the two sides of the liquid liquid interface, and the presence of a density gap between the two liquids. Density oscillations at the interface give a layered structure to critical nuclei in metastable systems. Our approach allows us to explore how the structure and free energy of the interface change as a function of pressure, temperature, and miscibility of the two components. The size of the density oscillations and of the density minimum at the interface are highly affected by the proximity to a drying transition. Oscillations are dampened and the density gap becomes deeper and broader as the vapor phase starts intruding between the immiscible liquid phases. Our results also show that although DFT in a local density approximation LDA fails to reproduce the oscillatory patterns observed in the computer simulations, the deviations in its prediction of interfacial free energies are small except at very low temperatures or very high pressures. In the case of critical nuclei, differences in the calculated barrier heights to nucleation are important only for the very large clusters.

5912 J. Chem. Phys., Vol. 110, No. 12, 22 March 1999 Napari et al. ACKNOWLEDGMENTS This work was supported by the National Science Foundation through grant CHE 98-00012 and by the Academy of Finland Project 36594. Support to V. Talanquer from the Facultad de Química and DGAPA Project IN100597 at UNAM is also gratefully acknowledged. 1 S. Toxvaerd and J. Stecki, J. Chem. Phys. 102, 7163 1995 ; 103, 4352 1995. 2 S. Iatsevitch and F. Forstmann, J. Chem. Phys. 107, 6925 1997. 3 R. Strey, Y. Viisanen, and P. E. Wagner, J. Chem. Phys. 103, 4333 1995. 4 A. S. Clarke, R. Kapral, and G. N. Patey, J. Chem. Phys. 101, 2432 1994. 5 V. Talanquer and D. W. Oxtoby, J. Chem. Phys. 104, 1993 1996. 6 P. Tarazona, Mol. Phys. 52, 81 1984. 7 J. A. Barker and D. Henderson, J. Chem. Phys. 47, 4714 1967. 8 R. Evans, Adv. Phys. 28, 143 1979. 9 V. Talanquer and D. W. Oxtoby, J. Chem. Phys. 100, 5190 1994. 10 V. Talanquer and D. W. Oxtoby, J. Chem. Phys. 102, 2156 1995. 11 A. Laaksonen and D. W. Oxtoby, J. Chem. Phys. 102, 5803 1995. 12 A. Laaksonen, J. Chem. Phys. 106, 7268 1997. 13 J. D. Weeks, D. Chandler, and H. C. Andersen, J. Chem. Phys. 54, 5237 1971. 14 X. C. Zeng and D. W. Oxtoby, J. Chem. Phys. 95, 5940 1991. 15 G. A. Mansoori, N. F. Carnahan, K. E. Starling, and T. W. Leland, J. Chem. Phys. 54, 1523 1971. 16 P. H. van Konynenburg and R. K. Scott, Philos. Trans. R. Soc. London, Ser. A 298, 495 1980. 17 R. Evans, R. J. F. Leote de Carvalho, J. R. Henderson, and D. C. Hoyle, J. Chem. Phys. 100, 591 1994.