Phase Equilibria

A system is said to be in a state of equilibrium if it shows no tendency to depart from that state either
by energy transfer through the mechanism of heat and work or by mass transfer across the phase
boundary. Since a change of state is caused by a driving force, we can describe a system at equilibrium
as one in which there are no driving forces for energy or mass transfer. That is, for a system in a state
of equilibrium, all forces are in exact balance. It may be noted here that the state of equilibrium is
different from a steady state condition. Under steady state there exist net fluxes for material or energy
transfer across a plane surface placed anywhere in the system. Under equilibrium the net flux is zero.
Transfer of material or energy across phase boundaries occurs till equilibrium is established between
the phases. In our daily experience, we come across a number of processes in which materials are
transferred from one phase to another. During breathing we take oxygen from the air through the lungs
and dissolve it in the blood. During the preparation of tea or coffee we extract the soluble components
in the powder into boiling water. Dilute aqueous solution of alcohol is concentrated by distillation in
which a vapour rich in alcohol is produced from the boiling solution. The phase equilibrium
thermodynamics is of fundamental importance in many branches of science, whether physical or
biological. It is particularly important in chemical engineering, because majority of manufacturing
processes involve transfer of mass between phases either during the preparation of theraw materials or
during the purification of the finished products. Gas–liquid absorption, distillation, liquid–liquid
extraction, leaching, adsorption, etc., are some of the important separation techniques employing mass
transfer between phases. In addition to these, many industrial chemical reactions arecarried out under
conditions where more than one phase exist. A good foundation in phase equilibrium thermodynamics
is essential for the analysis and design of these processes.
In this chapter due emphasis is given to the development of the relationship between the various
properties of the system such as pressure, temperature and composition when a state of equilibrium is
attained between the various phases constituting the system. The temperature–pressure-composition
relationships in multiphase system at equilibrium form the basis for the quantitative treatment of all
separation processes. The two types of phase equilibrium problems that are frequently encountered
are:
1. The determination of composition of phases which exist in equilibrium at a known temperature
and pressure
2. The determination of conditions of temperature and pressure required to obtain equilibrium
between phases of specified compositions.
The present chapter tries to provide solutions to these problems.

8.1 CRITERIA OF PHASE EQUILIBRIUM

Consider a homogeneous closed system in a state of internal equilibrium. The criteria of internal
thermal and mechanical equilibrium are that the temperature and pressure be uniform throughout the
system. For a system to be in thermodynamic equilibrium, additional criteria are to be satisfied.
Consider a closed system consisting of two phases of a binary solution, for example, the vapour and
liquid phases of an alcohol–water solution. The requirement of uniformity of temperature and
pressure does not preclude the possibility of transfer of mass between the phases. If the system is in
thermodynamic equilibrium, mass transfer also should not occur. It means that additional criteria are
necessary for establishing the state of thermodynamic equilibrium.
A system can interact with the surroundings reversibly or irreversibly. In the reversible process, a
state of equilibrium is maintained throughout the process. So it can be treated as a process connecting
a series of equilibrium states. The driving forces are only infinitesimal in magnitude and the process
can be reversed by an infinitesimal change (either increase or decrease) in the potential for the system
or the surroundings. The irreversible process, in contrast, occurs with a finite driving force, and it can
not be reversed by infinitesimal changes in the external conditions. However, all irreversible
processes tend towards a state of equilibrium. We have shown in Chapter 4 under ‘Clausius
inequality’,

In this equation, the equality sign refers to a reversible process which can be treated as a succession
of equilibrium states and the inequality refers to the entropy change for a spontaneous process whose
ultimate result would be an equilibrium state. The first law of thermodynamics expressed
mathematically by Eq. (2.5) can be rewritten as
dQ = dU + dW (8.1)
Substituting Eq. (8.1) into Eq. (4.44), we get
T dS ≥ dU + dW
dU ฀ T dS – dW (8.2)
dW in Eq. (8.2) may be replaced by P dV so that
dU ฀ T dS – P dV (8.3)
Equation (8.3) is valid for cases where external pressure is the only force and the work is, therefore,
the work of expansion only. By this, we exclude other effects like those due to gravitational and
electromagnetic fields and surface and tensile forces. Equation (8.3) can be treated as the combined
statement of the first and second law of thermodynamics applied to a closed system which interact
with its surroundings through heat transfer and work of volume displacement. This equation is utilised
for deriving the criteria of equilibrium under various sets of constraints, each set corresponding to a
physically realistic or commonly encountered situation. These different criteria are discussed now.
Constant U and V. An isolated system does not exchange mass, heat or work with the surroundings.
In Eq. (8.1), d Q = 0, d W = 0 and hence d U = 0. A well-insulated vessel of constant volume would
closely approximate this behaviour. Thus in Eq. (8.3) dU = 0 and dV = 0 so that

The entropy is constant in a reversible process and increases in a spontaneous process occurring in a
system of constant U and V. Since an irreversible process leads the system to an equilibrium state, the
entropy is maximum at equilibrium when no further spontaneous processes are possible.
Constant T and V. Helmholtz free energy is defined by Eq. (6.1).
A = U – TS
Rearranging Eq. (6.1), we get
U = A + TS
dU = dA + T dS + S dT
Substitute this result in Eq. (8.3) and rearrange the resulting expression to the following form
dA ฀ – P dV – S dT (8.5)
Under the restriction of constant temperature and volume, the latter implying no work, the equation
simplifies to

Equation (8.6) means that the spontaneous process occurring at constant temperature and volume is
accompanied by a decrease in the work function and consequently, in a state of thermodynamic
equilibrium under these conditions the Helmholtz free energy or the work function is a minimum.
Constant P and T. Equation (6.6) defines Gibbs free energy as
G = H – TS
Since H = U + PV we can write Eq. (6.6) as
G = U + PV – TS
Taking the differentials
dG = dU + P dV + V dP – T dS – S dT
rearranging these as
dU = dG – P dV – V dP + T dS + S dT
and combining this result with Eq. (8.3), we obtain
dG ฀ V dP – S dT (8.7)
At constant temperature and pressure, Eq. (8.7) reduces to

Equation (8.8) means that the free energy either decreases or remains unaltered depending upon
whether the process is spontaneous or reversible. It implies that for a system in equilibrium at a given
temperature and pressure the free energy must be minimum.
Since most chemical reactions and many physical changes are carried out under conditions of constant
temperature and pressure, Eq. (8.8) is the commonly used criterion of thermodynamic equilibrium. It
also provides a very convenient and simple test for the feasibility of a proposed process. No process
is possible which results in an increase in the Gibbs free energy of the system, because according to
Eq. (8.8) the Gibbs free energy always decreases in a spontaneous process and in the limit of the
reversible process, the free energy doesn’t change at all.
In the equilibrium state, differential variations can occur in the system at constant temperature and
pressure without producing any change in the Gibbs function. Thus, the equality in Eq. (8.8) can be
used as the general criterion of equilibrium or as a thermodynamic statement that characterises the
equilibrium state.
dG = 0 (at constant T and P) (8.9)
To apply this criterion for phase equilibrium problems we need formulate an expression for dG as
function of the number of moles of the components in various phases and set it equal to zero. This
equation along with the mass conservation equations provides the solutions to phase equilibrium
problems.

8.2 CRITERION OF STABILITY

It can be shown that the criterion of equilibrium [Eq. (8.8)] can be used to formulate the criterion of
stability for a binary mixture. When two pure liquids at a given temperature T and pressure P are
mixed together, the resulting mixture should have a lower free energy at the same temperature and
pressure. This is because the mixed state is an equilibrium state or stable state compared to the
unmixed state. The molar free energy of the mixture is thus less than the sum of the molar free energies
of the constituents for all possible concentrations. That is,
G – S xiGi < 0 (8.10)
The left-hand side in the above equation is the free energy change on mixing DG. Therefore,
DG < 0 (8.11)
When the free energy change on mixing DG is plotted against x1 —the mole fraction of constituent 1
in the binary mixture—the resulting curve is one of the two types shown in Fig. 8.1. The upper curve
is for a binary mixture, which is miscible for the entire concentration range. Assume that the points A
and B represent two binary mixtures of composition xA and xB respectively. Points on the dotted line
AB represent the composition as well as DG of the mixture of two phases obtained when solutions
represented by the points A and B are mixed together. Since the line AB is above the solid curve that
represents the free energy of the miscible solution, the free energy of the mixture in the miscible stateis
the minimum and the mixture exists as a single homogeneous phase. However, this argument is notvalid
for the lower curve in Fig. 8.1. The dotted line MN represents the free energy of the two-phasemixture
obtained when two binary mixtures of composition xM and xN, respectively, are mixedtogether. It
lies below the DG curve of the homogeneous solution. Any point on the line MNrepresents the
DG that would result for systems consisting of two phases of mole fraction xM and xN..Thus, when the
system moves from the solid curve to the dotted line MN, there is a decrease in thefree energy. That
is, the system attains stability when it moves from a homogeneous to aheterogeneous state. Therefore,
for mixtures of composition between points M and N, the equilibriumor stable state consists of two
immiscible phases. We see that the second derivative of DG withrespect to x1 is always positive
for stable liquid phase and if it becomes zero or negative, phaseseparation occurs. The criterion of
stability is that at constant temperature and pressure the free energy change on mixing DG, its
first and second derivatives are all continuous functions of theconcentration x and
EXAMPLE 8.1 Show that for a stable liquid phase, the fugacity of each component in a binary
mixture always increases with increase in concentration at constant temperature and pressure.
Solution The excess free energy of mixing was defined in Chapter 7. It was shown there that
DGE = DG – DGid
DGE = GE = RT S xi ln gi

DGid = RT S xi ln xi
Combining these three equations we find that
DG = RT S xi ln (gixi)

= S xi ln (gixi) = x1 ln (g1x1) + x2 ln (g2x2)

Differentiating this with respect to x1,
Differentiating Eq. (8.14) again with respect to x1 and noting that fugacities of pure components are
independent of concentration, we get
Equation (8.12) reveals that the left-hand side of the above equation is greater than zero. Therefore,

The last two equations imply that fugacity of components in a stable solution always increase with
increase in concentration.

8.3 PHASE EQUILIBRIA IN SINGLE-COMPONENT SYSTEMS

Consider the thermodynamic equilibrium in a system consisting of two or more phases of a single
substance. Though the individual phases can exchange mass with each other and are therefore open,
the system as a whole is closed. As an example, we can treat the equilibrium between vapour and
liquid phases of a single substance at a constant temperature and pressure. Applying the criterion of
equilibrium [see Eqs. (8.8) and (8.9)] to this closed system,
dG = 0
 dGa + dGb = 0 (8.17)
where the dGa and dGb are the changes in free energies of the phases ‘a’ and ‘b’ respectively. Since
each phase is open, the change in its free energy may be due to the changes in temperature, pressure
and the number of moles of the components that constitute the phase. Equation (7.35) expresses this
mathematically as
dG = V dP – S dT + dni
Applying this equation to the phases ‘a’ and ‘b’, we can write
dGa = Va dPa – Sa dTa + Ga dna, dGb = Vb dPb – Sb dTb + Gb dnb
At constant temperature and pressure,
dGa = Ga dna, dGb = Gb dnb (8.18)
As the system as a whole is closed,
dna + dnb = 0, or dna = – dnb (8.19)
Substituting Eqs. (8.18) and (8.19) into Eq. (8.17), we get
(Ga – Gb)dna = 0 (8.20)
Equation (8.20) means that
Ga = Gb (8.21)
Whenever two phases of the same substance are in equilibrium under a given temperature and
pressure, the molar free energy is the same in each phase.
We can verify the above result easily by considering the example of boiling water. As long as both
phases are present, an appreciable transfer of material from one phase to the other at constant
temperature and pressure would not disturb the equilibrium. The change in the free energy for the
equilibrium process (or reversible process) of evaporating a mole of liquid water is
DG = DH – TDS (8.22)
As pressure is constant, DH = Q, and the process being reversible, Q = TDS. Equation (8.22) gives
DG = 0
For vaporisation of 1 mol of liquid, DG = GV – GL, where GL and GV are the molar free energy of
water in the liquid and vapour states at the given T and P. Therefore, under equilibrium
GV = GL (8.23)
As the molar free energies are related to the fugacity of the substance by
G = RT ln f + C
Equation (8.21) can be expressed in terms of fugacity of the phases.
f a = fb (8.24)
where fa and fb are the fugacities in phases a and b respectively. It is convenient to work with
fugacities of substances as these have absolute values in contrast to free energies, which are usually
expressed as differences.
The above conclusions can be extended to three phases, which is the maximum number of phases that
can coexist under equilibrium in a system of one component.
EXAMPLE 8.2 Using the criterion of phase equilibrium, show that the change in entropy during
phase changes can be calculated from the latent heat of phase change and the absolute temperature as
DS = DH/T.
Solution Suppose that two phases a and b are in equilibrium. Using the definition of free energy [Eq.
(6.6)],
Ga = Ha – TSa, Gb = Hb – TSb
Here, H and S denote the enthalpy and entropy of the substance. Substituting these results in Eq.
(8.21),
Ha – TSa = Hb – TSb
This equation can be rearranged as,

The left-hand side of the above equation is the entropy change accompanying the phase change of one
mole of the substance (DS), and the numerator on the right-hand side represents the enthalpy change
for the phase change of one mole of the substance or the latent heat of phase change (DH). That is,

EXAMPLE 8.3 Deduce the Clapeyron equation using the criterion of equilibrium, Eq. (8.9).
Solution In Chapter 6 we have derived the Clapeyron equation, Eq. (6.25), using Maxwell’s relations.

The criterion of equilibrium provides an alternate route for its derivation. Consider any two phases a
and b of the same substance under equilibrium. Since Ga and Gb are both functions of temperature and
pressure, and these functional relationships are different for different phases, the two phases can
coexist only at such values of the temperature and pressure that Ga = Gb. If the temperature and
pressure are altered infinitesimally without disturbing the equilibrium, the change in the free energy
must be the same in each phase.
dGa = dGb (8.25)
In a phase change there is no work other than the work of expansion, so that
dG = V dP – S dT
Using this in Eq. (8.25),
Va dP – Sa dT = Vb dP – Sb dT (8.26)
V and S are the molar volume and molar entropy of the fluid with the superscript representing the
phase for which the properties correspond to. Equation (8.26) can be rearranged to the following form.

This relation gives the increase in pressure that is necessary to maintain the equilibrium between
phases for a pure substance when the temperature is increased. By using the following simplifications
Eq. (8.29) can be modified to yield the Clausius–Clapeyron equation applicable for vapour–liquid
equilibria.
1. The latent heat of vaporisation is constant and independent of temperature.
2. The molar volume of liquid is negligible compared to that of vapour.
3. The vapour behaves as ideal gas.
The Clausius–Clapeyron equation was derived [Eq. (6.28)] in Chapter 6 and is reproduced below.

where and are the vapour pressures at temperatures T1 and T2 respectively.

8.4 PHASE EQUILIBRIA IN MULTICOMPONENT SYSTEMS

The criterion of equal molar free energy [Eq. (8.21)] is applicable for equilibrium between phases of
a single component. This criterion needs modification when dealing with heterogeneous
multicomponent systems. A heterogeneous closed system is made up of two or more phases with each
phase behaving as open system within the overall closed system. Because each phase consists of two
or more components in different proportions, it is necessary that the criterion of multicomponent
phase equilibrium be developed in terms of partial molar free energies or the chemical potentials of
the components. The criteria of thermal and mechanical equilibrium are, as discussed earlier, the
uniformity of temperature and pressure. For the system to be in equilibrium with respect to mass
transfer, the driving force for mass transfer—the chemical potential—must have uniform values for
each component in all phases. This criterion of internal equilibrium is derived in the following
paragraphs.
Consider a heterogeneous system consisting of p phases indicated by the letters, a, b, g, ..., p. The
various components that constitute the system are 1, 2, 3, ..., C. The symbol denotes the chemical
potential of component ‘i’ in phase ‘k’. Suppose that small amounts of various components are
transferred from one phase to another, the system being in equilibrium and the temperature and
pressure kept constant. Since the system as a whole is closed, the proposed transfer should satisfy the
following criterion.
dG = 0 (at constant T and P) (8.9)
The free energy change in a multicomponent system is given by Eq. (7.35) as
dG = V dP – S dT + S mi dni
At constant temperature and pressure, the above equation becomes
dG = S mi dni
Substitute this into Eq. (8.9) to get
S mi dni = 0 (8.31)
Let denote the increase in the number of moles of component i in phase k. Equation (8.31) may
then be written as

The variations in the number of moles dni are independent of each other. However, they are subject to
the constraints imposed by Eq. (8.33). For all possible variations , Eq. (8.32) is to be satisfied. This
is possible only if
Equation (8.34) means that when a system consisting of several components distributed between
various phases is in thermodynamic equilibrium at a definite temperature and pressure, the chemical
potential of each component is the same in all the phases. If they are different, the component for
which such a difference exists will show a tendency to pass from the region of higher to the region of
lower chemical potential. Thus the equality of chemical potential along with the requirement of
uniformity of temperature and pressure serves as the general criterion of thermodynamic equilibrium
in a closed heterogeneous multicomponent system. In short, we can write

where C is a constant, an alternative and equally general criterion of equilibrium can be written in
terms of fugacities as

Fugacity is a more useful property than chemical potential for defining equilibrium since it can be
expressed in absolute values, whereas chemical potential can be expressed only relative to some
arbitrary reference state. Equation (8.36) is therefore widely used for the solution of phase equilibrium
problems.
EXAMPLE 8.4 Using the criterion of phase equilibrium show that the osmotic pressure over an ideal
solution can be evaluated as

where xA is the mole fraction of solute and VB is the molar volume of the solvent.
Solution Consider a vessel which is divided into two compartments by a semi-permeable membrane.
Pure solvent (say, water) is taken in one of the compartments and a solution (say, sucrose in water) is
taken in the other. Let T be the temperatures on both sides of the membrane and P be the pressure.
While the membrane is impermeable to the flow of the solute, it permits the flow of solvent into the
solution. This phenomenon of a solvent diffusing through a membrane which is permeable to it, but is
impermeable to the solute, is known as osmosis.
Osmosis is caused by the difference in the chemical potentials of the solvent on the two sides of the
membrane. At a given pressure, the chemical potential of a pure solvent is greater than that of the
solvent in the solution. By increasing the pressure at the solution side of the membrane, the chemical
potential of the solvent in the solution can be increased. When the pressure is increased to P฀keeping
the temperature constant, the chemical potential of the solvent in the solution would becomeequal to
that of the pure solvent at pressure P, and the diffusion would stop. If the pressure is increased
above P฀, the direction of diffusion would be reversed. In that event, the solvent would diffuse
from the solution to the pure solvent. This process is known as reverse osmosis. The excess pressure
P฀ – P to be applied over the solution at constant temperature to arrest the process ofosmosis is
known as the osmotic pressure. Thus, osmotic pressure is
Posmotic = P฀ – P
Let the mole fraction of the solutes constituting the solution be represented by xA and the mole
fraction of the solvent be represented by xB. Let denotes the chemical potential of the solvent at
pressure P฀. Equation (7.51) relates the chemical potential of a component in a solution to itsfugacity.
Thus

Combining the preceding two equations, we get

In this equation, is the chemical potential of pure solvent at pressure P฀, fB is its fugacity and
is its fugacity in the solution. Since the solution is ideal, the above equation may be simplified
utilising the Lewis–Randall rule which relates fB and that . Now we get the following
result for the chemical potential of the solvent in the solution at pressure P฀.
Since volume of a liquid is not affected by change in pressure, the integral in this equation can be
easily determined in terms of the molar volume VB. Thus
VB (P฀ – P) = – RT ln xB
Noting the definition of the osmotic pressure, the preceding equation may be written as

8.6 DUHEM’S THEOREM

The state of a system is completely determined when both the intensive as well as the extensive state
of the system is fixed. The phase rule gives the number of independent intensive variables to be
specified to define the intensive state of the system uniquely. It does not deal with the extensive state
of the system. By extensive state we mean the amount (or the mass) of the various phases present and
the total properties of the system. The Duhem’s theorem helps in establishing the extensive state of the
system. It states that for any closed system formed initially from the given masses of prescribed
chemical species, the equilibrium state is completely determined when any two independent variables
are fixed.
In addition to the phase rule variables, the mass of each phase should be known to define the state of
the system completely. Therefore, the total number of variables = p(C – 1) + 2 + p, where the last
term is the number of terms representing the mass of various phases. Since the system is closed, and
is formed from specified amounts of the chemical species present, a material balance equation can be
written for each of the C components present in the system. The total number of equations connecting
the variables will thus be more than the corresponding number in the phase rule by C, the number of
material balance equations. This is equal to C(p – 1) + C. The difference between the number of
variables and the number of equations is therefore two indicating that only two independent variables
need be fixed to define the state of the system completely. These may be either intensive or extensive. If
F < 2, at least one extensive variable must be fixed for complete determination of the system.
Forexample, for water and water vapour in equilibrium, since the number of degrees of freedom is
one, the intensive state of the system is specified by fixing either the pressure or the temperature. But
thetotal properties can be evaluated only if the amount of liquid or vapour is also specified. However,
for a binary mixture, say, water and alcohol, in vapour–liquid equilibrium, since the number of
degrees of freedom is two, no additional specifications are needed to predict the amount of liquid and
vapour present in equilibrium, provided, we know the amounts of the components from which the
system is formed. Duhem’s theorem is applicable to reacting systems as well (see Chapter 9).

8.7 VAPOUR–LIQUID EQUILIBRIA

The vapour-liquid equilibrium (VLE) data are essential for many engineering calculations,
especially in the design and analysis of separation operations such as distillation, absorption, etc.
Thermodynamics provides a system of equations relating the necessary experimental data and the
unknown vapour-liquid equilibrium compositions, temperature and pressure.
The conditions of equilibrium [Eq. (8.36)] require that the fugacity of a component in the liquid phase
be equal to that in the vapour phase. That is,

where, represents the fugacity of component i in the solution and the superscripts V and L represent
the vapour and liquid phases, respectively. Using this equation, the problem of determining the
composition of the liquid and vapour phases in equilibrium is quite simple: it is necessary only to
evaluate the compositions so that the fugacity of each component be the same in both phases. For
example, for a binary mixture of ethanol and water in vapour-liquid equilibrium, at a definite
temperature and pressure, the mole fractions in the liquid and vapour must be such that the fugacity of
ethanol is the same in both phases. That is, . Here, is the fugacity of ethanol in the mixture.To
evaluate quantitatively the equilibrium compositions, the fugacity of a component should be
expressed in terms of its mole fraction in the mixture. Using the definition of activity coefficient, the
fugacity of a substance in the vapour phase can be written in terms of its mole fraction yi in the mixture,
the fugacity of pure i as a vapour at the system temperature and pressure.

If the stable state for i at T and P is not a vapour, evaluating requires the introduction of a
hypothetical state. The use of the concept of fugacity coefficient helps to overcome this difficulty. The
fugacity of a component in a gas mixture can be written as

where is the fugacity coefficient of i in the mixture. The fugacity coefficient may be evaluated
from an equation of state for the mixture.
For the liquid phase, the fugacity of a component can be expressed as the product of its mole fraction
xi in the solution, the activity coefficient gLi and the fugacity of the component in the standard state.

Equations (8.45) and (8.46) are the fundamental relationships for estimating the vapour–liquid
equilibrium by two different approaches—the equations of state approach and the activity coefficient
model approach. Equation (8.45) forms the basis of estimating VLE by theequation of state approach,
which will be discussed under Section 8.12.1.
Equation (8.46) is the fundamental relationship in the study of vapour–liquid equilibrium based on
activity coefficient model for the liquid phase fugacity. The liquid-phase activity coefficients in Eq.
(8.46) can be estimated by any of the models described in the following sections. Estimation of the
vapour–liquid equilibria using activity coefficients is useful when polar components are present in
the system. For most substances at low pressures, can be assumed to be unity. If the pressure is
very high, must be evaluated using an equation of state. Using the activity coefficient approach, the
vapour–liquid equilibrium problems may be attacked, by dividing them into the following grouping
for convenience.

Case 1: Ideal gas-phase, ideal liquid solution. For mixtures of ideal gases, = 1. For ideal
liquid solutions, at low pressures, gL i = 1, and the fugacity is equal to the saturation pressure of
pure liquid at the temperature of interest. Equation (8.46) becomes

Case 2: Low-pressure VLE problems. If the liquid phase is not an ideal solution so that g π 1,
but the pressure is low enough that the assumption of ideal gas behaviour for the gas phase would not
introduce any significant errors in practical calculations, Eq. (8.46) can be simplified as
Case 3: High-pressure VLE problems. In the general case where ideal behaviour cannot be
assumed for the gas and liquid phases, the fugacity coefficient and the activity coefficient

gLishould first be determined for solving vapour–liquid equilibrium problems using Eq. (8.46). These are normally complex functions of
temperature, pressure and compositions and can be written as

The fugacity in the reference state is the fugacity of pure i at the same T, P and state of aggre-
gation as the mixture. To calculate this, it is convenient to determine first the fugacity of pure i in the
liquid state at T under its equilibrium vapour pressure and then apply a correction term for the fact
that . The fugacity of the liquid under its equilibrium vapour pressure is equal to the vapour
pressure times the fugacity coefficient . The fugacity coefficient is the ratio of the fugacity of
the component i under its saturation conditions to the saturation pressure of the substance and tends to
unity if the vapour behaves as an ideal gas. Using Eq. (6.31), we can write

The exponential in the above equation is known as the Poynting correction and it is approximately
unity when pressure is low. Also at low pressures when the gas behaves ideally, and
the above equation reduces to Eq. (8.48).

8.8 PHASE DIAGRAMS FOR BINARY SOLUTIONS

8.8.1 Constant-pressure Equilibria
Consider a binary system made up of components A and B. Component A is assumed to be more volatile
than B, i.e. the vapour pressure of A is greater than that of B at any given temperature. For a binary
liquid mixture in equilibrium with its vapour, according to the Gibbs Phase rule, the number of
degrees of freedom is two. When the pressure is fixed, only one variable, say liquid phase
composition, can be changed independently and other properties such as the temperature and the
vapour phase compositions get uniquely determined. Vapour–liquid equilibrium data at constant
pressure are usually represented by means of either the temperature-composition diagrams (the
T-x-y diagrams or the boiling point diagrams) or the distribution diagrams (x-y diagrams or
equilibrium curves).
Boiling-point diagram. The boiling point diagrams are plots of temperature as ordinate against
composition of liquid and vapour as abscissa. The composition of liquid is usually indicated by the
mole fraction of more volatile component in the liquid, x, and the composition of the vapour is
indicated by the mole fraction of the more volatile component in the vapour, y. Therefore, the boiling
point diagrams are also called T-x-y diagrams. The upper curve in Fig. 8.2 gives the temperature
versus vapour composition (y), and is known as the ‘dew-point curve’. The lower curve in the figure
is temperature versus liquid composition (x), also called the ‘bubble-point curve’. Below the bubble-
point curve the mixture is subcooled liquid and above the dew-point curve the mixture is superheated
vapour. Between the bubble-point and dew-point curves the mixture cannot exist as a single phase, it
spontaneously separates into saturated liquid and vapour phases that are in equilibrium.

To make these points clearer, consider a mixture whose temperature and composition ( x1) are such that
it is represented by point A in Fig. 8.2. Since the point A lies below the bubble-point curve, the solution
is entirely liquid. The mixture is taken in a closed container and the pressure over the system
is maintained at a constant value by a piston. The mixture is heated slowly so that its temperature
increases along the vertical line passing through point A till point B on the bubble-point curve is
reached. The temperature T1 corresponding to point B, is the bubble point of the original mixture. The
first bubble of the vapour is produced at this temperature and it will have the composition (y1)
represented by point C on the upper curve. The vapour is richer in the more volatile component.
Therefore y1 > x1, and the dew-point curve lies above the bubble-point curve. The mixtures at points
B and C are the liquid and vapour at equilibrium at the system pressure and temperature T1. Since
both are at the same temperature, they can be joined by a horizontal line BC, known as a ‘tie line’.
Further heating will result in the vaporisation of more liquid, and at temperature T2 the system will
consist of saturated liquid represented by point D and saturated vapour represented by point E, which
are in equilibrium. Since the vapour formed is not removed from the system, the overall composition
of the combined mixture of liquid and vapour will be same as x1, the composition of the original
mixture. However, the relative amounts of the liquid and the vapour change as the temperature is
changed.
These relative amounts are given by the ratio in which the point representing the combined mixture (in
this case, point F) divides the tie line DE. By material balance consideration, it can be easily verified
that

If heating is continued, eventually a temperature T3 is reached when almost all liquid is vaporised.
The last drop of liquid getting vaporised at this temperature has a composition denoted by point G and
the equilibrium vapour has the composition at H same as the original mixture. Temperature T3 is the
dew point of the original mixture. The mixture temperature increases along the vertical line HJ on
further heating. On cooling the superheated mixture at point J, the first drop of condensate appears when
the temperature drops to T3, the dew point of the mixture and the composition of the liquid is given
by point G.
We have seen that the mixture at point A has vaporised over a temperature range from T1 (the bubble
point) to T3 (the dew point), unlike a pure substance, which vaporises at a single temperature known
as the boiling point of the substance. For a solution, the term ‘boiling point’ has no meaning, because,
at a given pressure the temperature during vaporisation of a solution varies from the bubble point to
the dew point.
Equilibrium diagram. The vapour–liquid equilibrium data at constant pressure can be represented
on a x versus y plot or an equilibrium distribution diagram. If the vapour composition is taken as the
ordinate and the liquid composition is taken as the abscissa, a tie line such as line BC on the boiling
point diagram gives rise to a point such as point P on the distribution diagram
(Fig. 8.3). Since the vapour is richer in the more volatile component, the curve lies above the diagonal
on which x = y.
A liquid–vapour equilibrium curve very close to the diagonal means that the composition of the
vapour is not much different from the composition of the liquid with which it is in equilibrium; when
the curve coincides with the diagonal, x and y are equal.
Effect of pressure on VLE. On the boiling point diagram the temperatures corresponding to
x = 0 and x = 1 are the boiling points of pure substances B and A respectively. The boiling points of
pure substances increase with pressure. This is true for the bubble and dew points of a mixture of
given composition. Consequently the boiling point diagrams at higher pressures will be above the
boiling point diagrams at lower pressures as shown in Fig. 8.4. Since the relative volatility decreases
as pressure is increased, the closed loop formed by the dew-point and bubble-point curves become
narrow at high pressures. Figure 8.5 indicates the effect of pressure on the distribution diagram. In
Fig. 8.4, P3 is the critical pressure for component A and above this pressure, the looped curves are
shorter.
8.8.2 Constant-temperature Equilibria
Vapour–liquid equilibrium data at constant temperature are represented by means of P-x-y diagrams;
Fig. 8.6 shows a typical P-x-y diagram.
The pressure at x = 0 is the vapour pressure of pure B( ) and the pressure at x = 1 is the vapour
pressure of pure A( ). Since component A is assumed to be more volatile, and therefore,
the P-x-y diagram slopes upwards as shown in the figure. The P-y curve lies below the P-x curve so
that for any given pressure, y > x. A solution lying above the P-x curve is in the liquid region and that
lying below the P-y curve is in the vapour region. In between P-x and P-y curves the solution is a
mixture of saturated liquid and vapour. A horizontal line such as AB connects the liquid and vapour
phases in equilibrium and is therefore, a tie line. Assume that a liquid mixture whose conditions may
be represented by the point C in Fig. 8.6, is taken in a closed container. When the pressure over this
system is reduced at constant temperature, the first bubble of vapour forms at point D, and
vaporisation goes to completion at point E. Further reduction in pressure leads to the production of
superheated vapour represented by point F. The effect of temperature on P-x-y diagram is shown in
Fig. 8.6(b). When the temperature is less than the critical temperature of both components, the looped
curve such as the one shown at the bottom of Fig. 8.6(b) results. The other two curves refer to
temperatures greater than the critical temperature of A.
8.9 VAPOUR–LIQUID EQUILIBRIA IN IDEAL SOLUTIONS
It is possible to determine the vapour–liquid equilibrium (VLE) data of certain systems from the
vapour pressures of pure components constituting the system. If the liquid phase is an ideal solution
and the vapour behaves as an ideal gas, the VLE data can be estimated easily without resorting to
direct experimentation. A solution conforming to the ideal behaviour has the following characteristics,
all interrelated.
1. The components are chemically similar. The average intermolecular forces of attraction and
repulsion in the pure state and in the solution are of approximately the same order of magnitude.
2. There is no volume change on mixing (DV = 0) or the volume of the solution varies linearly with
composition.
3. There is neither absorption nor evolution of heat on mixing the constituents that form an ideal
solution (DH = 0); that is, there is no temperature change on mixing.
4. The components in an ideal solution obey Raoult’s law, which states that the partial pressure in
the vapour in equilibrium with a liquid is directly proportional to the concentration in the liquid.
That is, , where is the partial pressure of component i and xi is its mole fraction inthe
liquid. is the vapour pressure of pure i. This criterion also
implies that the total vapour pressure over an ideal solution is a linear function of its
composition.
For an experimental test of an ideal solution, the last criterion is the safest one to use. For example,
the solution formed by two chemically dissimilar materials like benzene and ethyl alcohol should
definitely be non-ideal. It is found that for an equimolar mixture of benzene and ethyl alcohol, there is
no change in volume during mixing at room temperature. This peculiar behaviour is because of the
fact that when this solution is formed from its constituents, there is increase in volume up to certain
concentration and thereafter the volume decreases as shown in Fig. 8.7. When the solution volume is
plotted against the composition, the curve will intersect the broken line representing the volume of an
ideal solution at a particular concentration represented in the figure by point P.

If the volume were measured for the concentration of the solution corresponding to point P, no change
in volume would be observed. This may be the case for enthalpy change of mixing also at some
particular composition. The conclusion to be drawn is that negligible volume change or temperature
change for one particular composition of the mixture is not a safe criterion of an ideal solution. If
these are to be used as the tests for ideal behaviour, then these tests should be done for more than one
concentration of the solution. In contrast, the criterion that the total vapour pressure over an ideal
solution varies linearly with composition is safe and reliable.
It should be understood that there exists no ideal solution in the strict sense of the word; but actual
mixtures approach ideality as a limit. Ideality requires that the molecules of the constituents are
similar in size, structure and chemical nature; only optical isomers of organic compounds meet these
requirements. Thus a mixture of ortho-, meta- and para-xylene conforms very closely to the ideal
solution behaviour. Practically, adjacent or nearly adjacent members of the homologous series of
organic compounds can be expected to form ideal solutions. Thus mixtures of benzene and toluene,
n-octane and n-hexane, ethyl alcohol and propyl alcohol, acetone and acetonitrile, paraffin
hydrocarbons in paraffin oils, etc., can be treated as ideal solutions in engineering calculations.
Consider an ideal binary solution made up of component 1 and component 2. We have shown in
Chapter 7 that all ideal solutions obey Lewis–Randall rule.
Here is the fugacity of the component i in the liquid and fi is the fugacity of pure i. Using the criterion
of equilibrium and noting that if pressure is not too high, the vapour would not depart too
greatly from ideal gas behaviour, it is possible to write

Equation (8.51) shows that at a given temperature, the total pressure over an ideal solution is a linear
function of composition thus establishing the fourth criteria given above. When the partial pressures
and total pressure are plotted against mole fraction x1, we get according to Eq. (8.50) and Eq. (8.51)
the straight lines shown in Fig. 8.8. The broken lines give the partial pressures and the continuous line
gives the total pressure.
The P-x-y diagram can be easily constructed. At any fixed temperature, the total pressure can be
calculated using Eq. (8.51) for various x values ranging from 0 to 1. The corresponding equilibrium
vapour phase compositions are obtained by applying Dalton’s law according to which the partial
pressure in the vapour is equal to the mole fraction in the vapour (y) times the total pressure (P). That
is

Thus Eq. (8.51) is used to calculate the total pressure at given x and Eq. (8.54) is used to calculate the
corresponding equilibrium vapour phase composition y. The P-x-y diagram can now be plotted as
shown in the Fig. 8.8.
To prepare the T-x-y diagrams at a given total pressure P we can again use Eqs. (8.51) and
(8.54). Assume temperatures lying between the boiling points of pure liquids at the given pressure.
For the temperature assumed, find the vapour pressures of the pure liquids and calculate x from
Eq. (8.51). Use these in Eq. (8.54) and calculate the vapour compositiony. Instead, if we attempt to
find the equilibrium temperature for the solution of known concentration x, the temperature may be
estimated by trial, such that the sum of the partial pressures is equal to the given total pressure. Once
the temperature is thus known, the vapour phase composition is determined as before. The T-x curve
is the lower curve in the figure and is called the bubble-point curve. The T-y curve is the upper curve
and is called the dew-point curve.
The y-x diagram is also prepared from the constant total pressure data. It can be constructed from the
boiling point diagram by drawing horizontal tie lines. The intersections of these lines with the bubble-
point curve give x and the intersections with the dew-point curve give y. Figure 8.10 shows atypical
equilibrium diagram.
There is an approximate method for the construction of the equilibrium diagram, and it is based on the
assumption that the ratio of vapour pressures of the components is independent of temperature. This
assumption may not introduce much error, as it is possible that the vapour pressures of both
components vary with temperature and these variations are to the same extent that their ratio remains
unaltered. Thus

which can be written in the following form.

Although Eq. (8.55) is not exact over a wide range of temperatures, the effect of variation in a is so
small that an average a value can be used in Eq. (8.55) and the whole y-x data required for the
preparation of the equilibrium curve can be evaluated.
EXAMPLE 8.5 Prove that if Raoult’s law is valid for one constituent of a binary solution over the
whole concentration range, it must also apply to the other constituent.
Solution Assume that Raoult’s law is obeyed by component 1 in a binary mixture. Then
As pointed out earlier, Raoult’s law is obeyed by ideal solutions when the vapour phase behaves as
an ideal gas whereas Lewis–Randall rule is obeyed by ideal solutions irrespective of whether the gas
phase is ideal or not. So for component 1, we can write

This equation is sometimes referred to as Duhem–Margules equation. Comparing Eq. (8.56) with Eq.
(8.57) we see that
This is Raoult’s law for component 2. The conclusion to be drawn from the above derivation is that if
Raoult’s law is applicable to one of the constituents of a liquid mixture at all concentrations, it must
be applicable to the other constituent as well.
EXAMPLE 8.6 n-Heptane and toluene form ideal solution. At 373 K, their vapour pressures are 106
and 74 kPa respectively. Determine the composition of the liquid and vapour in equilibrium at 373 K
and 101.3 kPa.
Solution Refer Eq. (8.51). Then,

where x is the mole fraction of heptane in the liquid. On solving this, we get x = 0.853
From Eq. (8.54),
y = 0.853 ฀ 106/101.3 = 0.893
The liquid and the vapour at the given conditions contain respectively 85.3% (mol) and 89.3% (mol)
heptane.
EXAMPLE 8.7 An equimolar solution of benzene and toluene is totally evaporated at a constant
temperature of 363 K. At this temperature, the vapour pressures of benzene and toluene are 135.4 and
54 kPa respectively. What are the pressures at the beginning and at the end of the vaporisation
process?
Solution Put x = 0.5 in Eq. (8.51). Then P = 94.7 kPa. This is the pressure at the beginning of
vaporisation. Equation (8.51) can be written as
Put y = 0.5 in this. Thus, we get P = 77.2 kPa. (This is the pressure at the end of vaporisation).
EXAMPLE 8.8 A mixture of A and B conforms closely to Raoult’s law. The pure component vapour
pressures in kPa at T K are given by

If the bubble point of a certain mixture of A and B is 349 K at a total pressure of 80 kPa, find the
composition of the first vapour that forms.
Solution At 349 K, the vapour pressures of the pure components are:

At a given temperature and pressure, the composition of the liquid and vapour phases in equilibrium
is calculated using Eqs. (8.51) and (8.54), respectively. The composition of the vapour so calculated
at the bubble-point temperature is the composition of the first vapour produced from a liquid on
boiling.
Using Eq. (8.51),
The vapour formed contains 92.5% A.
EXAMPLE 8.9 The vapour pressures of acetone (1) and acetonitrile (2) can be evaluated by the
Antoine equations

where T is in K and P is in kPa. Assuming that the solutions formed by these are ideal, calculate
(a) x1 and y1 at 327 K and 65 kPa
(b) T and y1 at 65 kPa and x1 = 0.4
(c) P and y1 at 327 K and x1 = 0.4
(d) T and x1 at 65 kPa and y1 = 0.4
(e) P and x1 at 327 K and y1 = 0.4
(f) The fraction of the system that is liquid and the composition of the liquid and vapour in
equilibrium at 327 K and 65 kPa when the overall composition of the system is 70 mole per cent
acetone.
Solution (a) From the Antoine equations, at 327 K,

(b) Equation (8.51) can be written as

(c) At 327 K, we have = 85.12 kPa and = 39.31 kPa. Here x1 = 0.4. Using these values in Eq.
(8.51), we get P = 57.63 kPa. Using Eq. (8.54)
y1 = 0.4 ฀ 85.12/57.63 = 0.5908
(d) Equation (8.51) can be written as

Assume a temperature and calculate the vapour pressures using Antoine equations. Substitute the
vapour pressure values in the above equation. See whether the Left-hand side = 0.4. This is repeated
till the left-hand side of the above equation becomes equal to 0.4.
At T = 334 K, = 107.91 kPa and = 51.01 kPa.
(e) At 327 K, we have = 85.12 kPa and = 39.31 kPa. Here y1 = 0.4
Equation (8.54) relates y to x. When P in Eq. (8.54) is eliminated using Eq. (8.51) we get

(f) The composition of the vapour and liquid in equilibrium at P = 65 kPa and T = 327 K were
determined in part (a). They are x1 = 0.5608 and y1 = 0.7344. Let f be the fraction of the mixture that
is liquid. Then an acetone balance gives
1 ฀ 0.7 = f ฀ 0.5608 + (1 – f) ฀ 0.7344
Solving this, we get f = 0.1982. That is, 19.82% (mol) of the given mixture is liquid.
EXAMPLE 8.10 Mixtures of n-Heptane (A) and n-Octane (B) are expected to behave ideally. The
total pressure over the system is 101.3 kPa. Using the vapour pressure data given below,
(a) Construct the boiling point diagram and
(b) The equilibrium diagram and
(c) Deduce an equation for the equilibrium diagram using an arithmetic average a value.

T, K 371.4 378 383 388 393 398.6

PS , kPa 101.3 125.3 140.0 160.0 179.9 205.3
A
PS , kPa 44.4 55.6 64.5 74.8 86.6 101.3
B

Solution Sample calculation: Consider the second set of data. T = 378 K; PSA = 125.3 kPa; PSB =
55.6 kPa.
Using Eq. (8.51),
101.3 = 55.6 + xA(125.3 – 55.6)
Therefore, xA = 0.656.
Using Eq. (8.54), we see
yA = 0.656 ฀ 125.3/101.3 = 0.811
Relative volatility is
a = PS /PS = 125.3/55.6 = 2.25
A B
These calculations are repeated for other temperatures. The results are tabulated below:

T, K 371.4 378 383 388 393 398.6

xA 1.000 0.656 0.487 0.312 0.157 0
yA 1.000 0.811 0.674 0.492 0.279 0
a 2.28 2.25 2.17 2.14 2.08 2.02

(a) Plot of T versus x and y gives the boiling point diagram

(b) Plot of y against x gives the equilibrium diagram
(c)The average of the last row gives a = 2.16. Use this value of a in Eq. (8.55) to get the equation
for the equilibrium curve.

8.10 NON-IDEAL SOLUTIONS

We have seen that the partial pressure of a component in an ideal solution varies linearly with
concentration in the solution. If the solution behaves ideally, the different molecules should be
chemically similar. In that case, the molecules of a particular substance, when brought into solution
with other components, would not experience any difference in the environment surrounding them
from that existed in their pure state. The intermolecular forces in the pure state of the substance and
that in the solution would then be approximately of the same order of magnitude. Therefore, the
fugacity (or the partial pressure) of a substance, which is a measure of the tendency of the substance
to escape from the solution, is not affected by the properties of the other components in the solution. It
depends only on the number of molecules of the substance present, or its concentration. In short, the
components in an ideal solution obey Raoult’s law. But for non-ideal solutions, the partial pressures
do not vary linearly with composition, as shown in Fig. 8.11 for the case of carbon disulphide– acetone
system.
The non-ideal behaviour of liquid mixtures arises due to the dissimilarity among molecules. The
dissimilarities arise from the difference in the molecular structure or from the difference in the
molecular weight. The non-ideal behaviour of light hydrocarbons such as methane, ethylene, etc., in
mixtures of heavier paraffin or crude oil is due to the difference in the molecular weights. In contrast,
it is a type of intermolecular attraction called hydrogen bonding, that is responsible for the non-ideal
behaviour resulting from the difference in the molecular structure. Molecules, which contain atoms
such as oxygen, chlorine, fluorine or nitrogen, tend to be polar. When the electrons in the bonds
between these atoms and hydrogen are not equally shared, a dipole is created. The electrons tend to
be closer to the larger atoms, which become negatively charged compared to hydrogen which
becomes the positive end of the dipole. In a solution of polar substances, the molecules tend to arrange
themselves so that the charge deficiency of the hydrogen atoms is compensated by an intermolecular
bond with a ‘donor’ or negatively charged atom. These hydrogen bonds have energiesof the order of
several kJ/mol. Because of hydrogen bonding, bimolecular complexes between like or unlike
molecules are formed, and even chain-like or three-dimensional aggregates between a large number
of molecules are sometimes formed. The formation or destruction of hydrogen bonding duringmixing
leads to very large heat effects and drastic changes in the thermodynamic properties.
Non-ideal behaviour falls into one of the following two types: positive deviation from ideality and
negative deviation from ideality. The positive deviation from ideality results when the actual partial
pressure of each constituent is greater than it should be if Raoult’s law were obeyed. Solutions in
which intermolecular forces between like molecules are stronger than those between unlike
molecules, show appreciable positive deviation from ideality. On mixing the constituents which form
a solution exhibiting positive deviation from ideality, there is an absorption of heat. This can be
proved easily if we recognise the experimental observation that most solutions tend to exhibit ideal
behaviour as temperature is increased. For a solution showing positive deviation, for each
component is greater than its mole fraction xi, and as temperature is increased it becomes equal to xi ,
because the solution tends to ideality as temperature is increased. It means that for a system of given
composition for which deviation from Raoult’s law is positive, the ratio decreases with
increasing temperature. That is

Comparing Eq. (8.58) with Eq. (8.59) we see that /RT2 < 0, which means . The total
enthalpy of the solution is , whereas the enthalpy of the system before mixing is
S niHi. Since the former is greater than the latter, there is absorption of heat during mixing. Examples
of solutions showing positive deviation from ideality are oxygen–nitrogen, ethanol–ethyl ether,
water–ethanol, carbon disulphide–acetone, benzene–cyclohexane, acetonitrile–benzene,
n-hexane–nitroethane, etc.
For solutions exhibiting negative deviation from ideal behaviour, the partial pressures are less than
those given by Raoult’s law. By a derivation similar to the one presented in the preceding paragraph,
it can be shown that when solutions showing negative deviation are formed from pure constituents
there is evolution of heat. At the molecular level, appreciable negative deviation reflects stronger
intermolecular forces between unlike than between like pairs of molecules. Examples are chloroform–
ethyl ether, chloroform–benzene, hydrochloric acid–water, phenol–cyclohexanol, chloroform–
acetone, etc.
The general nature of the vapour pressure curves showing positive and negative deviation are shown
in Fig. 8.12. Figures 8.12(a) and (b) refer to constant temperature conditions. The uppermost curves
give the total vapour pressure as function of liquid composition. The corresponding curves, as a
function of the vapour composition lie below it, so that the vapour is rich in the more volatile
component.
8.10.1 Azeotropes
Azeotropes are constant boiling mixtures. The word ‘azeotrope’ is derived from Greek word meaning
‘boiling without changing’. When an azeotrope is boiled, the resulting vapour will have the same
composition as the liquid from which it is produced. Whereas, the equilibrium temperature of an
ordinary solution varies from the bubble point to the dew point, the boiling point of an azeotrope
remains constant till the entire liquid is vaporised. The azeotropes are formed by solution showing
large positive or negative deviation from ideality. If the vapour pressures of the constituents of a
solution are very close, then any appreciable positive deviation from ideality will lead to a maximum
in the vapour pressure curve and negative deviations from ideality under the same conditions leads to
a minimum in the vapour pressure curve. Even if an appreciable difference exists in the vapour
pressures of the pure components, the chances for the occurrence of maximum or minimum in the
vapour pressures should not be overruled if the deviation from ideal behaviour is quite high. At the
composition at which there exists a maximum or minimum in the vapour pressure curve, a minimum or
maximum, as the case may be, exists in the boiling point diagrams. The mixture is said to form an
azeotrope at this composition under the given temperature and pressure and it will distil without
change in composition, because the vapour produced has the same composition as the liquid.
Minimum-boiling azeotropes. Solutions showing positive deviation from ideality in certain cases
may lead to the formation of azeotropes of the minimum-boiling type. The P-x-y, T-x-y and x-y curves
for the minimum-boiling azeotropes are shown in Fig. 8.13.
In the boiling point diagram, the liquid and vapour curves are tangent at point M, the point of
azeotropism at this pressure. The temperature at M is the minimum temperature of boiling for the
system. For all mixtures of composition less than M, the equilibrium vapour is richer than the
liquid in the more volatile component. For all mixtures richer than M, the vapour is less rich than the
liquid in the more volatile component. A mixture of composition M boils producing a vapour of
identical compositions and consequently at a constant temperature and without change in composition.
If solutions either at P or Q are boiled in an open vessel with continuous escape of vapours, the
temperature and composition move along the lower curve away from M and towards the pure
substances. Solutions like these cannot be distilled by usual distillation methods. One
of the most important azeotropes in this category is ethanol–water which forms azeotrope at
89.4% (mol) ethanol at 351.4 K and 101.3 kPa. Other examples are benzene–ethanol (341.2 K,
55%benzene), carbon disulphide–acetone (312.5 K, 61% carbon disulphide), isopropyl ether–
isopropyl alcohol (345.1 K, 39.3% alcohol), all at 101.3 kPa.
Maximum-boiling azeotropes. When the total pressure of the system at equilibrium is less than
the ideal value, the system is said to exhibit negative deviation from ideality. When the difference in
vapour pressures of the components is not too great, and in addition, the negative deviations are large,
the curve for total pressure against composition passes through a minimum. This condition results in a
maximum in the boiling temperature and a condition of azeotropism as at point M in Fig. 8.14.
The vapour is leaner in the more volatile component for liquids whose concentration is less than the
azeotropic concentration. Solution on either side of the azeotrope, if boiled in an open vessel with
escape of vapours will ultimately leave a residual liquid of the azeotropic composition in the vessel.
Maximum-boiling azeotropes are less common than the minimum-boiling type. Hydrochloric acid–
water system forms an azeotrope at 11.1% (mol) HCl at 383 K and 101.3 kPa. Other examples are
chloroform–acetone (337.7 K and 65.5% acetone), phenol–cyclohexanol
(455.65 K, 90% phenol), all at 101.3 kPa.
The plot of activity coefficients versus mole fraction in the liquid phase for ideal solutions, minimum-
boiling azeotrope and maximum-boiling azeotrope are given in Fig. 8.15.

Effect of pressure on azeotropes. The azeotropic composition shifts continuously with change
in pressure or temperature. In some cases, changing the pressure may eliminate azeotropism from the
system. Azeotropism disappears in ethanol–water system at pressures below 9.33 kPa. The table
below illustrates the effect of pressure on this system. The last row gives the mole per cent of alcohol
in the azeotrope.

P, kPa 13.3 20.0 26.6 53.2 101.3 146.6 193.3

T, K 307.4 315.2 321 336 351.3 361 368.5
 %(mol) 99.6 96.2 93.8 91.4 89.43 89.3 89.0
EXAMPLE 8.11 Prove that at the azeotropic composition, the vapour and liquid have the same
composition.
Solution Refer the Duhem–Margules equation, Eq. (8.57).

Assuming the vapour to behave ideally, it can be written as

For a maximum or minimum on the total pressure curve, (dP/dx1) is zero. Then Eq. (8.64)
necessitates that either the terms in the parenthesis should be zero or is zero. If the latter
term is zero the partial pressure would be unaffected by concentration changes, which is not true.
Hence, at the point of azeotropism, we have, the following relation:

Since x2 = 1 – x1 and y2 = 1 – y1, the above result means that at azeotropic condition, x1 = y1 or the
vapour composition and liquid composition are the same.

8.11 VAPOUR–LIQUID EQUILIBRIA (VLE) AT LOW PRESSURES

It has been pointed out that if the pressure is not very high, the fundamental equation relating the
compositions of the vapour and liquid under equilibrium is given by Eq. (8.47) as
 yiP = gixi
The activity coefficients gi vary with composition and temperature at a given pressure. In order to
calculate the relationship between pressure, temperature and composition of the equilibrium phases
for non-ideal solutions at low pressures where the vapour phase is assumed to behave as an ideal
gas, we can utilise Eq. (8.47), provided that the equations relating the activity coefficients to
compositions and temperature are available. Experimental vapour–liquid equilibrium data can be
correlated using such equations and the empirical parameters in them are evaluated. Once an activity
coefficient equation suitable for the given system is identified and the parameters evaluated, it can be
used for vapour–liquid equilibrium calculations through Eq. (8.47). We have seen that the activity
coefficients in a binary solution are not independent, and they are interrelated through the Gibbs–
Duhem equations as

The minimum requirement to be met by an equation for activity coefficient is that it should conform to
the restriction imposed by the above relationship. Among a number of equations between g and x that
are available, as far as phase equilibrium problems are concerned, some equations have got wide
acceptance. They are discussed in the following sections.

8.11.1 Activity Coefficient Equations

Wohl’s three-suffix equations. The relationship between excess free energy and activity
coefficient was discussed in Chapter 7. Most of the equations relating activity coefficient and
concentration of the solution were derived from these excess free energy relationships. Wohl
proposed, statistically, a general method for expressing excess free energy and provided some rough
physical significance to the various parameters appearing in the equations. Wohl’s equation for excess
free energy contained terms for compositions, effective molal volumes and effective volumetric
fraction of the separate constituents of the solution. From these equations, the following empirical
relations for activity coefficient could be written.

Equation (8.65) is known as Wohl’s three-suffix equation . It involves three parameters, A, B and
(q1/q2) which are characteristics of the binary system.
Margules equation. When the term (q1/q2) is unity in Eq. (8.65), we get the following expression,
which is known as the Margules three-suffix equation.
The constant A in the above equation is the terminal value of ln g1 at x1 = 0 and the constant B is the
terminal value of ln g2 at x2 = 0. The three-suffix Margules equation adequately represents the VLE
data of systems like acetone–methanol, acetone–chloroform, chloroform–methanol, etc.
When A = B in Eq. (8.66), the Margules equation takes the following simple form:
(8.67)
Equation (8.67) is called the Margules two-suffix equation. It represents sufficiently and accurately
the activity coefficients of simple liquid mixtures, i.e. mixtures of molecules, which are similar in
size, shape and chemical nature. The constant A may be positive or negative. While in general, the
constant depends on temperature, for many systems it is a weak function of temperature. Vapour–
liquid equilibrium data of argon–oxygen, benzene–cyclohexane, etc., are well represented by the
Margules equation [Eq. (8.67)].
van Laar equation. Let (q1/q2) = (A/B) in Eq. (8.65). The resulting two-parameter equation is
known as the van Laar equation. The van Laar equations can be written as

The constant A is the terminal value of ln g1 at x1 = 0 and B is the terminal value of ln g2

at x2 = 0. When A and B are equal, the van Laar equations simplify to the Margules equation,
Eq. (8.67). The van Laar equation (8.68) may be rearranged to the following forms, which are
very convenient for the evaluation of constants A and B.

Strictly speaking, van Laar equations are applicable only for solutions of relatively simple, preferably
non-polar liquids. But empirically, it has been found that these are applicable for more complex
mixtures. The van Laar equations are widely used for vapour–liquid equilibrium calculations because
of their flexibility and mathematical simplicity. Activity coefficients in benzene– isooctane system, n-
propanol–water system, etc., are accurately represented by the van Laar equations.
The selection of a proper equation for VLE data correlation depends on the molecular complexity o
the system and the precision of the experimental data. When an equation is selected that fits the
experimental data well, the constants for the constant pressure conditions will be different from those
applicable for constant temperature conditions. The effect of pressure on the constants is usually
negligibly small, whereas the effect of temperature is appreciable and cannot be neglected. The van
Laar constants vary with temperatures unless the temperature range involved is small. However, in
vapour–liquid equilibrium calculations, the effect of temperature on the activity coefficient is usually
ignored (Prausnitz, 1985).
The Margules three-suffix equation is suited for symmetrical systems, i.e. where the constants A and B
are nearly the same. The van Laar equations can be used for unsymmetrical solutions, where the ratio
A/B does not exceed 2. Though many systems follow van Laar equations, they cannot represent
maxima or minima in the ln g curve. Margules three-suffix equation should be used in such cases. For
choice of an appropriate equation, a rule of thumb usually employed is this: When the ratio of molar
volumes is close to unity, the Margules equation is preferred. When the ratio is quite different from
unity, as is the case when water is one of the constituents, the van Laar equations are found to be
satisfactory. For example, the chloroform–ethyl alcohol system, which shows a maximum and a
minimum on the ln g curves and whose ratio between pure component molar volumes is 1.38, is
accurately represented by the Margules equation. For n-propanol–water system this ratio is 4.16 and
the van Laar equations are found to represent the behaviour accurately.
It is to be remembered that in equations having only two constants, determination of g1 and g2 at a
single known composition permits the evaluation of the constants and the complete g curve. Equation
(8.47) permits the evaluation of g1 and g2 when it is rearranged to the following form.

The data required are a single set of equilibrium vapour–liquid composition values and the vapour
pressures of the pure components. When an azeotrope is formed, only the azeotropic composition
need be known, because it represents the composition of both the liquid and the vapour phases. The
activity coefficients can be evaluated by putting x = y in Eq. (8.70).

Wilson equation. All the activity coefficient equations discussed so far can be deduced from the
original Wohl’s equation under proper simplifying assumptions. However, there are many equations
that cannot be derived from the Wohl’s general equation. Among such equations, the Wilson equation,
the NRTL equation and the UNIQUAC equation are important from practical point of view. All thes
are based on the concept of local compositions, which are different from the overall mixture
compositions. Based on molecular considerations, Wilson (1964) proposed the following equations
for activity coefficients in a binary mixture.

Wilson equations have two adjustable positive parameters L12 and L21. These are related to the pure
component molar volumes and to the characteristic energy differences by
where V1 and V2 are the molar volumes of pure liquids and l’s are the energies of interaction between
the molecules designated in the subscripts. The differences in the characteristic energies (aij) are
assumed to be temperature independent and this introduces no serious error in practical calculations.
Wilson equation provides a good representation of VLE of a variety of miscible mixtures. It is
particularly suitable for solutions of polar or associating components like alcohols in non-polar
solvents for which the Margules and van Laar equations are generally inadequate.
Wilson equation suffers from two disadvantages, though not serious for many applications. Firstly, it
is not suitable for systems showing maxima or minima on the ln g versus x curves. Secondly, it is not
useful for systems exhibiting limited miscibility. The use of Wilson equation is therefore
recommended only for liquid systems that are completely miscible, or for partially miscible systems
in the region where only one liquid phase exists.
Non-random two-liquid (NRTL) equation. The NRTL model, proposed by Renon and Prausnitz
(1968), also is based on the local composition concept. The activity coefficients are

The constants b12 and b21 are similar to the constants representing characteristic energy
differences appearing in the Wilson equation. These, as well as the constant a12 are independent of
composition and temperature. The parameter a12 is related to the non-randomness in the
mixture. If a1 2 is zero, the mixture is completely random and the NRTL equation reduces to the
Margules equation. It is found from fitting of experimental data that a12 varies from about 0.20
to 0.47. In the absence of the experimental data, the value of a12 is arbitrarily set, a typical
choice being a12 = 0.3. When a12 is arbitrarily fixed, NRTL equation becomes a two-parameter
model.
NRTL equation is applicable to partially miscible as well as totally miscible systems. For moderately
non-ideal systems, it offers no advantage over the van Laar and Margules equations. But for strongly
non-ideal solutions and especially partially miscible systems, the NRTL equations provide a good
representation.
Universal quasi-chemical (UNIQUAC) equation. Abrams and Prausnitz (1975) extended the
quasi-chemical theory of liquid mixtures to solutions containing molecules of different sizes. This
extension is called the UNIQUAC theory. The UNIQUAC model consists of two parts—t
combinatorial part, which describes the prominent entropic contribution and a residual part, which is
due primarily to the intermolecular forces that are responsible for the enthalpy of mixing. The
combinatorial part is determined by the sizes and shape of the molecules and requires only pure-
component data. The residual part depends on the intermolecular forces and involves two adjustable
binary parameters.
The UNIQUAC equations for activity coefficients are

z is the coordination number, r, q and q฀ are pure-component molecular structure constants. The
molecular size and surface area are given by r and q respectively. For fluids other than water or lower
alcohols, q = q฀. For alcohols, the surface of interaction q฀ is smaller than the geometric surface q.
The adjustable binary parameters t12 and t21 are related to the characteristic energies Du as follows.
The UNIQUAC equation satisfies a large number of non-electrolyte mixtures containing non-pola
fluids such as hydrocarbons, alcohols, nitriles, ketones, aldehydes, organic acids, etc., and water,
including partially miscible mixtures. The main advantages of this equation are its wide applicability
and simplicity arising primarily from the fact that there are only two adjustable parameters.
Universal functional activity coefficient (UNIFAC) method. In the UNIFAC method, th
activity coefficients are estimated through group contributions. The liquid is treated as a solution of
different structural groups from which the molecules are formed, rather than a solution of molecules
themselves. This method is based on the UNIQUAC model where the activity coefficient is divide
into two parts—the molecular size contribution (the combinatorial part) and the interaction
contributions (the residual part).

The combinatorial contribution can be estimated from pure-component properties and the size and
shape of the molecules, whereas for the estimation of the second part, group areas and group
contributions are needed. A large number of group interaction parameters are already reported.
UNIFAC has been successfully used for the design of distillation columns, involving even azeotropic
and extractive distillation (Prausnitz et al., 1986).
EXAMPLE 8.12 Liquids A and B form an azeotrope containing 46.1 mole per cent A at 101.3 kPa and
345 K. At 345 K, the vapour pressure of A is 84.8 kPa and that of B is 78.2 kPa. Calculate the van Laar
constants.
Solution Let the material A be component 1 and B be component 2. The activity coefficients at the
azeotropic concentration can be evaluated by Eq. (8.71)
EXAMPLE 8.13 The azeotrope of the ethanol–benzene system has a composition of 44.8% (mol)
ethanol with a boiling point of 341.4 K at 101.3 kPa. At this temperature the vapour pressure of
benzene is 68.9 kPa and the vapour pressure of ethanol is 67.4 kPa. What are the activity coefficients
in a solution containing 10% alcohol?
Solution Let benzene be component 1 and alcohol component 2. For the azeotrope
EXAMPLE 8.14 Water (1)–hydrazine (2) system forms an azeotrope containing 58.5% (mol)
hydrazine at 393 K and 101.3 kPa. Calculate the equilibrium vapour composition for a solution
containing 20% (mol) hydrazine. The relative volatility of water with reference to hydrazine is 1.6
and may be assumed to remain constant in the temperature range involved. The vapour pressure of
hydrazine at 393 K is 124.76 kPa.
Solution The vapour pressure of water at 393 K = 1.6 ฀ vapour pressure of hydrazine at
393 K = 1.6 ฀ 124.76 = 199.62 kPa.
To evaluate the vapour compositions using these equations, we should know the vapour pressure
values at the new equilibrium temperature. Taking the ratio of the last two equations, we get

The composition of the vapour in equilibrium with the liquid containing 20% hydrazine is 5.28%
hydrazine and 94.72% water.
EXAMPLE 8.15 At 318 K and 24.4 kPa, the composition of the system ethanol (1) and toluene (2) at
equilibrium is x1 = 0.3 and y1 = 0.634. The saturation pressure at the given temperature for the pure

components are , respectively. Calculate:

(a) The liquid-phase activity coefficients
(b) The value of GE/RT for the liquid phase
Solution (a) At vapour–liquid equilibrium, the composition of the vapour and liquid phases are
related by Eq. (8.48)

Equation (8.48) may be used to evaluate the activity coefficients.

EXAMPLE 8.16 The activity coefficients in a mixture of components A and B at 313 K are given by

At 313 K, A and B form an azeotrope containing 49.4 mol percent A at a total pressure of 27 kPa. If the
vapour pressures of pure A and pure B are 25.0 and 24.3 kPa, respectively, calculate the total pressure
of the vapour at temperature 313 K in equilibrium with a liquid mixture containing 12.5 mol percent A.
Solution At the azeotropic composition, Eq. (8.71) is applicable, so that the activity coefficients are:
EXAMPLE 8.17 Using van Laar constants and the vapour pressures of the pure substances how would
you prove whether a given binary system forms an azeotrope or not?
Solution If the mixture does not exhibit azeotropic behaviour, the ratio (y1/x1) will be greater than
(y2/x2) for the entire concentration range 0 to 1. Denoting the ratio of (y1/x1) to (y2/x2) by a, then a
> 1 for 0 < x < 1. However, if the mixture forms an azeotrope, then the value of a will be greater than1
over some concentration range and will be less than 1 over the remaining portion. Since a varies
continuously with x, a should have a value equal to 1 at some x which is the azeotropic composition.
Writing Eq. (8.47) for both components and rearranging the result, we get

A and B are the van Laar constants. If the mixture forms an azeotrope, one of the above values will be
greater than 1 and the other less than 1.
EXAMPLE 8.18 For the binary system methanol (1) and benzene (2), the recommended values of the
Wilson parameters at 341 K are L12 = 0.1751 and L21 = 0.3456. The vapour pressures of pure species
are = 68.75 kPa and = 115.89 kPa. Show that the given system can form an azeotropeat 341 K.
Assume that the vapour behaves like an ideal gas.
Solution Wilson equations [Eq. (8.72)] provide the activity coefficients in a binary mixture as:
Equation (8.48) gives the relationship between activity coefficient and equilibrium phase
compositions as

If the mixture forms an azeotrope, then the value of a will be greater than 1 over some concentration
range and will be less than 1 over the remaining portion. Relative volatilities are calculated at x1 = 0
and x1 = 1. If the system forms an azeotrope, one of these values will be greater than unity and the
other less than unity.
At x1 = 0,

At x2 = 0,
Since the relative volatility at x1 = 0 is greater than unity, and that at x2 = 0 is less than unity, it is clear
that the system forms an azeotrope.
EXAMPLE 8.19 A stream of isopropanol–water mixture is flashed into a separation chamber at 353
K and 91.2 kPa. A particular analysis of the liquid product showed an isopropanol content of 4.7%
(mol), a value that deviated from the norm. It is suspected that an air leak into the separator might
have caused this. Do you agree? The vapour pressures of the pure propanol and water are 91.11 kPa
and 47.36 kPa respectively, and the van Laar constants are A = 2.470 and B = 1.094.
Solution For x1 = 0.047, x2 = 0.953, A = 2.47 and B = 1.094, Eq. (8.68) gives
g1 = 7.388 and g2 = 1.011
The total pressure corresponding to this equilibrium composition is

This is less than the total pressure. This error must have been caused by an air-leak.
EXAMPLE 8.20 Construct the P-x-y diagram for the cyclohexane (1)–benzene (2) system at 313 K
given that at 313 K the vapour pressures are = 24.62 kPa and = 24.41 kPa. The liquid-phase
activity coefficients are given by

Solution Assume x1 = 0.4, then x2 = 0.6. Therefore

ln g1 = 0.458 ฀ 0.62 = 0.16488; g1 = 1.1793.

Similarly,
ln g2 = 0.458 ฀ 0.42 = 0.0733; g2 = 1.0760
The total pressure is determined as:
The above calculation is repeated for various x1 values. The results are tabulated below:

x1 0 0.2 0.4 0.6 0.8 1.0

g1 1.5809 1.3406 1.1793 1.0760 1.0185 1.000

g2 1.000 1.0185 1.0760 1.1793 1.3406 1.5809

P 24.41 26.49 27.37 27.41 26.61 24.62
y1 0 0.2492 0.4243 0.5799 0.7540 1.0

The results are plotted taking P on the y-axis and x1 and y1 on the x-axis.
EXAMPLE 8.21 From vapour–liquid equilibrium measurements for ethanol–benzene system at 318
K and 40.25 kPa it is found that the vapour in equilibrium with a liquid containing 38.4% (mol)
benzene contained 56.6% (mol) benzene. The system forms an azeotrope at 318 K. At this
temperature, the vapour pressures of ethanol and benzene are 22.9 and 29.6 kPa respectively.
Determine the composition and total pressure of the azeotrope. Assume that van Laar equation is
applicable for the system.
Solution Let benzene be component 1 and ethanol component 2. Using Eq. (8.47) the activity
coefficients are determined.
For any liquid composition, the activity coefficients are calculated using these equations. If the
mixture forms an azeotrope at any composition, then as per Eq. (8.71), the following relations also
give the activity coefficients.

Thus, activity coefficients calculated using the van Laar equations should also satisfy the relation
at the azeotropic composition. The azeotropic composition is obtained by trial
assuming values of x. For x1 = 0.6, g1 = 1.3830 and g2 = 1.7806.

These values are so close that it can be assumed that the composition corresponds to an azeotrope.
Thus, the liquid mixture forms an azeotrope containing 60% benzene which boils at 318 K and
approximately 40.86 kPa, (mean of 40.937 and 40.775 kPa).
EXAMPLE 8.22 The activity coefficients in a binary system are given by ln .
Show that if the system forms an azeotrope, then and the azeotropic composition
is given by
At the azeotropic composition, the P-x curve has a maximum or minimum. Therefore, the above result
is equated to zero and after rearrangement, we get

Since the value of x1 lies between 0 and 1 it is essential that . Noting that x1x2 for
a binary mixture is less than or equal to 0.25, the second root leads to A ≥ 2. This condition usually
results in partial miscibility. Thus, condition for homogeneous azeotropy is and A
< 2.
EXAMPLE 8.23 The following values refer to the Wilson parameters for the system acetone(1)–
water(2):
a1 2 = 1225.31 J/mol, a21 = 6051.01 J/mol, V1 = 74.05 ฀ 10–6 m3/mol, V2 = 18.07 ฀ 10–6
m3/mol.
The vapour pressures are given by
where P is in kPa and T is in K. Calculate the equilibrium pressure and composition of
(a) Vapour in equilibrium with a liquid of composition x1 = 0.43 at 349 K.
(b) The liquid in equilibrium with a vapour of concentration y1 = 0.8 at 349 K.
Solution Using the Antoine equations, at 349 K, the vapour pressures are calculated as

(a) Now from Eq. (8.72), we obtain

The vapour composition is found out from the relation y1P = g1x1

(b) Since the liquid composition is not known, the activity coefficients cannot be calculated. Assume
g1 = g2 = 1. The relation Pyi = xi can be written for component 1 and component 2, which on
rearrangement gives
Activity coefficients can now be determined for this composition using Wilson equations.

Now the pressure is recalculated incorporating the activity coefficient values. The equation
This is same as the previous value calculated for P. Therefore, P = 164.48 kPa and x1 = 0.4568.

EXAMPLE 8.24 The system methanol–methyl ethyl ketone forms an azeotrope containing 84.2%
(mol) methanol at 337.5 K and 101.3 kPa. The vapour pressures of the pure species are given by the
Antoine equation

Determine the parameters in the Wilson equation.

Solution The vapour pressures at 337.5 K are calculated using Antoine equations.

L12 is calculated by substituting the known values in the above equation.

8.12 VAPOUR-LIQUID EQUILIBRIA INVOLVING HIGH PRESSURES AND

MULTICOMPONENT SYSTEMS
8.12.1 Equations of State Approach
The equations of state approach is most useful for vapour–liquid equilibria at high pressures. In this
method, the vapour and liquid compositions are related by Eq. (8.45):

The equations of state provide thermodynamic models for evaluating in Eq. (8.45) from
volumetric properties. The fugacity coefficients for the vapour and liquid phases are to be obtained
from appropriate equations of state. Equations of state widely used in engineering calculations were
discussed in Chapter 3. The most generally used equations of state for VLE calculations are given i
Table 8.1.
The fugacity coefficients in Eq. (8.45) can be calculated using a suitable equation of state. The
relationship between fugacity coefficient and the volumetric properties can be written as

where Z is the compressibility factor.

The equations of state approach has the advantage that the standard state fugacity need not be
specified. Another advantage is that the continuity at the critical point is ensured as the properties in
the gas and liquid phases are evaluated using the same model. The equations of state approach is more
suitable to mixtures of nonpolar components. Also, the equations of state are widely used to predict
VLE of light hydrocarbon mixtures.

8.12.2 Vaporisation Equilibrium Constants

For high-pressure vapour–liquid equilibrium calculations, it is convenient to express the phase
equilibrium relations in terms of vaporisation equilibrium constants or K factors. It is defined as the
ratio of mole fraction in the vapour phase y to that in the liquid x or K = (y/x). The equation of state
form of the K-value is obtained from Eq. (8.45) as follows:

In Eq. (8.83a), is the fugacity coefficient in the saturation state, Vi is the molar volume of pure i as

saturated liquid. are evaluated using an equation of state. Assuming the Poynting factor to
be unity, Eq. (8.83a) can be rearranged as

which is nothing but Raoult’s law which is applicable for ideal solution where the vapour behaves as
ideal gas. We see that for ideal solution, the vaporisation equilibrium constant is the ratio of vapour
pressure to total pressure. The K-factors for ideal solutions depend only on the temperature and
pressure and are readily correlated as a function of these two variables.
For mixtures of light hydrocarbons, Eq. (8.83a) can be simplified using two assumptions:
1. Intermolecular forces are weak, so that vapour phase behaves as an ideal solution, so that ,
where fi is the fugacity coefficient of the pure components.
2. Liquid phase behaves as an ideal solution so that =1.0.
The resulting equation is very convenient as it involves properties of pure component only.

Equation (8.83e) reveals that for mixtures of light hydrocarbons, the vaporisation equilibrium
constants are independent of the composition of the liquid and vapour phases in equilibrium
and can be evaluated as pure component properties. Since can be determined from
equations of state or generalised correlations, it is possible to provide correlations for K-values of
substances as functions of temperature and pressure. DePriester nomographs [C.L. DePriester, Chem
Eng. Progr., Symposium Ser. 7, 49 (1953)] provide such correlations for many
hydrocarbons. These nomographs are available in standard references such as Chemical Engineers
Handbook. Figure 8.16 gives the K-factor for light hydrocarbons in the high temperature
range.
8.12.3 Bubble-point Equilibria
The bubble-point temperature is the one at which the first bubble of vapour is produced from the
liquid on heating at constant pressure. At the bubble point the liquid has the same composition as the
original mixture. Therefore, in problems where bubble-point temperature is to be determined, the xi
are known. Assume a temperature and get the Ki values at this temperature. Calculate yi using yi =
Kixi. If the assumed temperature is correct then
S yi = S Kixi = 1 (8.84)
Otherwise, repeat the calculations with another temperature. To find the bubble-point pressure , a
similar procedure as above is adopted by assuming various values of pressure until S Kixi = 1.

8.12.4 Dew-point Equilibria

The dew-point temperature is the one at which the first drop of condensate is formed on cooling a
vapour at constant pressure. The vapour in equilibrium with the liquid at the dew point has the same
composition as the original mixture. In order to find the dew-point temperature, a temperature is
assumed arbitrarily and Ki is determined. Then,

Otherwise, repeat the calculation by assuming another temperature till this equation is satisfied.
Determination of the dew-point pressure involves a similar procedure assuming pressure instead of
temperature.

8.12.5 Flash Vaporisation

The general flash vaporisation problem can be stated as: Given a mixture of known overall
composition zi at temperature T and pressure P, what is the fraction that is vapour (V) and what are the
composition of the liquid and vapour phases in equilibrium? The overall material balance for the
system is
F=V+L (8.86)
where F is the total number of moles of the initial mixture. The component-i balance for the
system is
Fzi = Vyi + Lxi (8.87)
Since yi = Kixi, it can be eliminated from Eq. (8.87) to get the following:
Equation (8.91) can also be utilised in an iterative procedure to estimate T, P or the fraction of the
initial mixture that is vaporised.
EXAMPLE 8.25 A mixture contains 45% (mol) methanol (A), 30% (mol) ethanol (B) and the rest n-
propanol (C). Liquid solution may be assumed to be ideal and perfect gas law is valid for the vapour
phase. Calculate at a total pressure of 101.3 kPa.
(a) The bubble point and the vapour composition
(b) The dew point and the liquid composition.
The vapour pressures of the pure liquids are given below:

Temperature, K 333 343 353 363

PSA, kPa 81.97133.29186.61266.58

PSB, kPa 49.3273.31 106.63166.61

PSC, kPa 39.3262.65 93.30 133.29

Solution The vapour pressures of the components are plotted against temperature so that interpolation
of vapour pressure can be done easily.
(a) If the vapour phase can be treated as an ideal gas and liquid phase, an ideal solution, the K-values
can be written as Ki = yi/xi = . Equation (8.84) can be written as

Now temperatures are assumed till the above equality is satisfied. It is seen that at 344 K,
The bubble-point lies between 344 and 345 K. By interpolation, the bubble-point is obtainedas
344.6 K. At this temperature the vapour pressures are obtained from the P vs T plots. =
137.3 kPa, = 76.20 kPa and = 65.40 kPa.

Component xi Ki = /P yi = Kixi

Methanol 0.45 137.30 1.355 0.610

Ethanol 0.30 76.20 0.752 0.226
Propanol 0.25 65.40 0.646 0.162
S Kixi 0.998

The equilibrium vapour contains 61% methanol, 22.6% ethanol and 16.2% propanol.
(b) Equation (8.85) for the present case becomes

The dew-point temperature is to be determined by trial such that the above relation is satisfied. By
trial, it can be seen that at 347.5 K, = 153.28 kPa, = 85.25 kPa and = 73.31 kPa.

Component yi Ki = /P xi = yi/Ki
Methanol 0.45 153.28 1.5131 0.2974
Ethanol 0.30 85.25 0.8416 0.3565
Propanol 0.25 73.31 0.7237 0.3454
Syi/Ki 0.9993

The values in the last column are the liquid composition at the dew point. Thus, liquid contains 29.7%
methanol, 35.7% ethanol, and 34.5% propanol.
EXAMPLE 8.26 A hydrocarbon mixture contains 25% (mol) propane, 40% (mol) n-butane and 35%
(mol) n-pentane at 1447.14 kPa. Assume ideal solution behaviour and calculate
(a) The bubble-point temperature and composition of the vapour
(b) The dew-point temperature and the composition of the liquid
(c)The temperature and the composition of the liquid and vapour in equilibrium when 45% (mol)
of the initial mixture is vaporised. (The values of Ki can be obtained from Fig. 13.6 of Chemical
Engineer’s Handbook, 5th ed.)
Solution (a) Assume temperature, say 355.4 K, and the Ki values are found out from the nomograph
[Fig. 13.6(b) in Chemical Engineer’s Handbook]. The products of Ki and xi are calculated and their
sum S xiKi is found out. The results for two temperatures 355.4 K and
366.5 K are shown below.

T = 355.4 K T = 366.5 K
Component xi Ki Kixi Ki Kixi

Propane 0.25 2.000 0.500 2.30 0.575

n-Butane 0.40 0.780 0.312 0.90 0.360
n-Pentane 0.35 0.330 0.116 0.40 0.140
S Kixi 0.928 1.075

The bubble-point temperature lies between 355.4 K and 366.5 K. By interpolation, the temperature is
found out to be 361 K. The calculations are carried out at this temperature and the results are as
follows:

Component xi Ki Kixi

Propane 0.25 2.12 0.530

n-Butane 0.40 0.85 0.340
n-Pentane 0.35 0.37 0.130
S Kixi 1.000

Since S xiKi is approximately 1.00, the bubble-point temperature is 361 K. The values in the last
column are the mole fraction of various components in the vapour. At the bubble-point, the vapour
contains 53% propane, 34% butane and 13% pentane.
(b) At the dew-point temperature, S yi/Ki = 1. At 377.6 K, this value is 1.1598 and at 388.8 K it is
0.9677.

T = 377.6 K T = 388.8 K
Component yi Ki yi/Ki Ki yi/Ki
Propane 0.25 2.6 0.0962 2.9 0.0862
n-Butane 0.40 1.1 0.3636 1.3 0.3077
n-Pentane 0.35 0.5 0.7000 0.61 0.5738
Syi/Ki 1.1598 0.9677

By interpolation, the dew-point temperature is found to be 387 K. The calculations for this
temperature is given below.
 Component yi Ki yi/Ki
Propane 0.25 2.85 0.0877
n-Butane 0.40 1.25 0.3200
n-Pentane 0.35 0.59 0.5932
S yi/Ki 1.0009

The last column in the above table is the liquid compositions. The equilibrium liquid at the dew point
contains 8.77% propane, 32.0% butane and 59.32% pentane.
(c) In the following calculations, temperature is assumed so as to satisfy Eq. (8.91). For a basis o
100 mol of the initial mixture, F = 100 mol, V = 45 mol and L = 55 mol. Equation (8.91) becomes

T = 366.5 K T = 377.6 K
Component zi Ki zi/[1 + L/(VKi)] Ki zi/[1 + L/(VKi)]

Propane 0.25 2.30 0.1632 2.6 0.1701

n-Butane 0.40 0.90 0.1696 1.1 0.1895
n-Pentane 0.35 0.40 0.0863 0.5 0.1016
S zi/[1 + L/(VKi)] 0.4191 0.4612

From the calculations given above, we see that the equilibrium temperature is between 366.5 K and
377.6 K. By interpolation, T = 374.6 K.
T = 374.6 K
Component zi Ki zi/[1 + L/(VKi)]
Propane 0.25 2.50 0.1679
n-Butane 0.40 1.08 0.1876
n-Pentane 0.35 0.48 0.0987
S zi/[1 + L/(VKi)] 0.4542

Comparing Eqs. (8.90) and (8.91), we can see that

These are calculated using the values in the last column. Corresponding xi values are found out using
the material balance [Eq. (8.87)].
Fzi = Vyi + Lxi
The results of the calculation are given below:
 Component yi xi

Propane 0.3697 0.1521

n-Butane 0.4130 0.3894
n-Pentane 0.2173 0.4586

8.13 CONSISTENCY TESTS FOR VLE DATA

Many practical cases like distillation calculations are dependent on vapour–liquid equilibrium data
and such data should be reasonably accurate if the results are to be reliable. As the VLE
measurements are prone to inaccuracies, it is essential that we have some means for checking the
consistency of the measured results. Thermodynamics provides tests for consistency of experimental
VLE data. Almost all these tests are based on the Gibbs–Duhem equations written in terms of activit
coefficients [Eq. (7.101)].

8.13.1 Using Slope of ln g Curves

The Gibbs–Duhem equation in terms of activity coefficient [Eq. (7.101)] provides a very simple tes
for thermodynamic consistency.

Plot the logarithm of the activity coefficients against mole fraction x1 of component 1 in a binary
solution as shown in Fig 8.17(a) and measure the slopes of the tangents drawn to the resulting curves
at any selected composition x1.
Equation (8.92) tells us that if the Gibbs–Duhem equation is to be satisfied, both slopes must have
opposite sign. Otherwise, the data are inconsistent. If the slopes are of opposite sign, substitute the
values in Eq. (8.92) and if it is satisfied reasonably well, then the data is consistent at the selected
composition. For a complete test, the slopes determined at other compositions are substituted into Eq.
(8.92) to see whether the equality is satisfied or not.
In addition to the above observations, we can make the following generalisations with the help of Eq.
(7.101).
1. If one of the ln g curves has a maximum (or minimum) at certain concentration, the other curve
should have a minimum (or maximum) at the same composition.
2. If there is no maximum or minimum point, then both curves must be positive or both must be
negative over the entire range. Or in other words, if one component has g values always greater
than unity and has no maximum, the g values of the other component must likewise be greater
than unity. This is a consequence of the fact that Raoult’s law is to be obeyed by the component
 as its mole fraction tends to unity.
From the above discussion, it is clear that Fig. 8.17(b) represents plots of consistent data whereas Figs.
8.17(c) and (d) are plots of thermodynamically inconsistent data. In Fig. 8.17(c), though there is
maximum on one curve and minimum on the other, these are shown at different compositions. In Fig.
8.17(d), the slopes have the same sign and the data are thermodynamically inaccurate.

8.13.2 Using Data at the Mid-point

For testing the thermodynamic consistency of VLE data, the integrated forms of the Gibbs–Duhem
equation like the van Laar or the Margules equation are sometimes found very convenient. Consider
the van Laar equation Eq. (8.68). Put x1 = x2 = 0.5 and we see that
 Case 1. Assume that A = B. Then ln g1 = B/4 and ln g2 = A/4
Case 2. Assume that A = 2B. Then ln g1 = (2/9)B and ln g2 = (2/9)A
Case 3. Assume that A = 3B. Then ln g1 = (3/16)B and ln g2 = (3/16)A
In all cases cited above, the ln g1 at the mid-point (i.e. at x1 = x2 = 0.5) is approximately one-fourth
the van Laar constant B and ln g2 approximately one-fourth the constant A. We have already seen that
the van Laar constant A is ln g1 as x1 tends to zero and B is ln g2 as x1 tends to one. Now we have a
rough check on the consistency of VLE data for the mid-point. The lng value at this point will be
approximately one-fourth the terminal value of the other ln g curve. That is the curve, which is highest
at the end-point will be lowest at the mid-point and vice versa, as shown in Figs. 8.18(a) and (b).
The mid-point values are approximately one-fourth the terminal values of the other curves in both the
figures. However, Fig. 8.18(c) reveals an inconsistent data, as the curve which is highest at the end-
point is also the highest at the mid-point.

8.13.3 Redlich–Kister Method

This method, also known as the zero area method can be applied to test the consistency of
experimental data when the activity coefficient values over the entire concentration range is available.
It is based on the excess free energy of mixing which is the difference between the free energy of
mixing of a real solution and that of an ideal solution. Referring to Eqs. (7.134) and (7.145)the excess
free energy of mixing can be written as
DGE = RT S xi ln gi (8.94)
For a binary solution, it can be written as
DGE = RT(x1 ln g1 + x2 ln g2)
Differentiating this with respect to x1, we get
From the experimental values of activity coefficients, ln (g1/g2) values are calculated and plotted
against x1 taken on the x-axis. The net area of the diagram should equal zero if the data are
thermodynamically consistent. That is, the area above the x-axis will be equal to the area below it as
shown in Fig. 8.19.

8.13.4 Using the Coexistence Equation

The coexistence equation can be used for testing the consistency of vapour–liquid equilibrium data. If
the vapour in equilibrium with a binary liquid mixture behaves as an ideal gas, Eq. (8.47) can be used
to describe the equilibrium. Rearranging Eq. (8.47), the activity coefficients can be written as
Substitute this in the Gibbs–Duhem Equations [Eq. (8.92)] written in the following form

Equation (8.99) is known as the coexistence equation. It can be used to calculate any one of the three
variables P, x or y if experimentally measured values of the other two variables are available. If all
the three variables are experimentally determined, then Eq. (8.99) can be used to test the consistency
of the measured data.

8.13.5 Using the Partial Pressure Data

At low pressures, the fugacity of a gas equals the pressure and therefore, the Gibbs–Duhem equation
in terms of fugacity [Eq. (7.99)] can be rewritten as,

or

The partial pressures of both components are plotted against mole fraction x1 as in Fig. 8.20. The
slopes, are determined at any selected composition. and
are calculated. Then, according to Eq. (8.100), the absolute values of these
quantities should be the same if the data are thermodynamically consistent.
EXAMPLE 8.27 The following results were obtained by experimental VLE measurements on
the system, ethanol (1)–benzene (2) at 101.3 kPa. Test whether the data are thermodynamically
consistent or not.
x1 0.003 0.449 0.700 0.900
y1 0.432 0.449 0.520 0.719
, kPa 65.31 63.98 66.64 81.31
, kPa 68.64 68.64 69.31 72.24

Solution Assuming that the gas phase behaves ideally, the activity coefficients are calculated as

The values so calculated are listed below:

x1 0.003 0.449 0.700 0.900
g1 223.3 1.58 1.13 0.99

g2 0.841 1.475 2.34 2.92

ln(g1/g2 ) 5.58 0.068 – 0.734 – 1.082

ln (g1/g2) values are plotted against x1. The net area is found out. Since this is not equal to zero, the
given experimental measurements do not satisfy the Redlich–Kister criterion [Eq. (8.97)] for
consistency.

8.14 CALCULATION OF ACTIVITY COEFFICIENTS USING

GIBBS–DUHEM EQUATION
The activity coefficients in a binary solution are related to each other according to the Gibbs–Duhem
equations as given below. See Eq. (7.101).

The integral can be evaluated analytically if g2 is expressed as a mathematical function of composition.

To find the integral graphically, plot a graph with ( x2/x1) on the y-axis and ln g2 on the x-axis. The
area under the curve from ln g2 at x2 = 0 to ln g2 at any desired composition x2 gives the value
– ln g1 for the component 1.
EXAMPLE 8.28 The activity coefficients for component 1 in a binary solution can be represented by
, where a, b and c are concentration independent parameters. Derive an
expression for ln g2.
Solution Equation (8.101) may be written as

For the present case,

EXAMPLE 8.29 The following data gives the composition versus total pressure for the system
chloroform (1)–ethyl alcohol (2) at 328 K are:

Vapour pressures of chloroform and acetone at 328 K are 82.35 and 37.30 kPa respectively. Estimate
the constants in the Margules equation [Eq. (8.66)].
Solution The Margules equations are:
(Note: Equations (8.103) and (8.104) are known as Carlson and Colburn relations for activity
coefficient.)

EXAMPLE 8.30 The following table gives the partial pressure of acetone versus liquid composition
for acetone (1)–water (2) system at 333 K.

x1 0 0.033 0.117 0.318 0.554 0.736 1.000

1, kPa 0 25.33 59.05 78.37 89.58 94.77 114.63

The vapour pressure of water at 333 K is 19.91 kPa. Calculate the partial pressure of water in the
vapour phase.
Solution Equation (8.100) is the Gibbs–Duhem equation in terms of partial pressures.
Using the given data calculate x1/[(1 – x1) ] and this is plotted against . The area under the curve
between the limits 0 and gives the integral in Eq. (8.106) from which the partial pressure can be
calculated. The results are given below:

x1 0 0.033 0.117 0.318 0.554 0.736 1.000

, kPa 19.91 19.31 18.27 16.99 15.42 13.90 0

8.15 VLE FOR SYSTEMS OF LIMITED MISCIBILITY

8.15.1 Partially Miscible Systems
Figure 8.21 shows the temperature-composition diagram for a partially miscible system at constant
total pressure. Points A and B indicate the boiling points of pure liquids A and B at this pressure. A
binary system consisting of two liquid phases and a vapour phase is univariant according to the phase
rule. By fixing the pressure, the system is completely defined. The states of the three phases in
equilibrium lie on the horizontal line at T*, the three-phase equilibrium temperature. Points C and D
represent the saturated liquid phases and point E the vapour in equilibrium with these liquids.
Below T* the system is entirely liquid. It may exist as a homogeneous system, or, as a heterogeneous
system consisting of two saturated liquid phases, depending upon its overall composition. If the
overall composition of the mixture lies within the region bounded by the curves FC and GD below
temperature T*, the mixture cannot exist as a single phase. It separates into two saturated liquid
phases, a A-rich phase (LA) represented by the curve GD and a B-rich phase (LB) represented by the
curve FC. The compositions of these phases are determined by the intersection of a horizontal line
corresponding to the temperature of the system and the curves GD and FC. Consider, for example, a
liquid whose combined concentration and temperature are such that it is represented by the point M in
Fig. 8.21(a). Since M lies in the two-liquid region (LB –LA), the mixture separates into a B-rich phase
represented by point P, and a A-rich phase represented by point Q.
The mutual solubility of A and B increases with increase in temperature as shown by the curves FC and
GD in Fig. 8.21(a). The change in the solubility of liquid A in liquid B with temperature is along FC
and that of B in A is along GD. To the right of the curve GD and to the left of the curve CF the mixture
exists as homogeneous liquid, if the temperature is below the three-phase temperature. The higher the
three-phase temperature the more the mutual solubility of liquids A and B. Suppose that themixture at
M in Fig. 8.21(a) is heated at constant pressure. Its temperature increases till T* is reached. At T*, the
vapour of the composition corresponding to point E is formed. If additional quantities of liquid A or B
are introduced into the system at this point, the relative amounts of the three phases in equilibrium will
change, but the composition of these phases represented by points C, D and E must remain the same.
The temperature T* remains constant until one of the liquid phases disappears. At temperatures above
T*, the system can exist either as two phases (liquid and vapour) or as a single phase depending on
the overall composition. The liquid phase that disappears depends on whether the combined mixture
composition is to the left or right of point E. For the mixture at M,
the A-rich phase disappears. The system now consists of a saturated liquid (LB) in equilibrium with
saturated vapour (V). When the temperature is increased the overall composition must lie on the
vertical line MN, the relative amounts and composition of the liquid and the vapour phases change
accordingly. Above the temperature corresponding to point R, the mixture is entirely vapour. If the
initial mixture were to the right of point E, the liquid in equilibrium with the vapour would have been
LA instead of LB.
The three-phase equilibrium temperature increases with pressure as shown in Fig. 8.22 where
temperature-composition diagrams are plotted at various pressures. The curves FC and GD
eventually merge to a single point and the two liquid phases become identical. The temperature at
which this occurs is known as the upper critical solution temperature (UCST). For pressures abov
this critical condition, the three-phase equilibrium conditions do not exist.

8.15.2 Immiscible Systems

The phase diagram for a completely immiscible system is shown in Fig. 8.23. This diagram is a special
case of Fig. 8.21 occurring when the two liquid phases LA and LB are the pure liquids A and B
respectively. Consider a binary mixture of two immiscible liquids having an overall composition
represented by point M. When this mixture is heated its temperature increases along the line MN. For
all temperatures up to point N, each component exerts its full vapour pressure as its partial pressure
in the vapour phase. When the temperature T* corresponding to point N is reached, the sum of the
vapour pressures becomes equal to the surrounding pressure and the system cannot exist entirely in
the liquid state and vapour is produced. T* is therefore known as the three-phase temperature.
The composition of the vapour in equilibrium with the pure liquids at the three-phase temperature is
given by point E. Since there are three phases and two components present, according to the phase
rule, the number of degrees of freedom is one as in the case of partially miscible systems. It means
that, when pressure is fixed the system is completely defined. The three-phase temperature and the
equilibrium vapour composition get automatically fixed by specifying the pressure in such a way that
the sum of the vapour pressures equals the surrounding pressure.

On further addition of heat, the temperature remains constant at T* and more vapour of the same
composition as given by point E is formed. This continues till one of the components disappears from
the liquid and the system becomes a two-phase mixture either LA – V or LB – V depending upon the
initial composition.
Now let us consider the cooling of a vapour of initial composition and temperature indicated by point
J. When the temperature is lowered to that corresponding to point K the vapour pressure of pure liquid
B will be equal to the partial pressure of B in the vapour. Pure liquid B gets condensed and thevapour
composition changes along the line KE. When the temperature T* is reached, the partial pressure of A
in the vapour will be equal to the vapour pressure of A and at this condition, pure liquids A and B and
the vapour are present in equilibrium. Further cooling results in the elimination ofthe vapour phase and
the system now consists of two immiscible liquids. If the initial vapour were at point P, pure liquid A
would have condensed out first, instead of B.
EXAMPLE 8.31 A high boiling organic liquid is purified from non-volatile impurities by allowing it
to mix with steam directly at a total pressure of 93.30 kPa. The vapour pressure data are given as
follows:
 Temperature, K 353 373
Vapour pressure of water, kPa 47.98 101.3
Vapour pressure of liquid, kPa 2.67 5.33

Assume that water and the organic liquid are immiscible and the impurities do not affect the
vaporisation characteristics. The vapour pressures vary linearly with temperature. Calculate under
three-phase equilibrium
(a) The equilibrium temperature and
(b) The composition of the resulting vapour.
Solution (a) At 353 K, sum of the vapour pressures is 50.65 kPa and at 373 K it is 106.63 kPa. Sinc
the vapour pressures vary linearly, the temperature at which the sum of vapour pressures is 93.3 kPa
is obtained by interpolation.

(b) At 368.2 K, the vapour pressure of water is 88.50 kPa and that of the liquid is 4.80 kPa. Since a
three-phase equilibrium, the partial pressure is equal to the vapour pressure, the ratio of mole fractions
of the components will be same as the ratio of vapour pressures. Let y be the mole fraction of water in
the vapour. Then

The vapour contains 94.86% (mol) water vapour.

EXAMPLE 8.32 Assuming that benzene is immiscible with water, prepare a temperature-
composition diagram for benzene (1)–water (2) system at 101.3 kPa using the following vapour
pressure data:

T, K 323 333 343 348 353 363 373

, kPa 12.40 19.86 31.06 37.99 47.32 70.11 101.3
, kPa 35.85 51.85 72.91 85.31 100.50 135.42 179.14

The boiling point of pure benzene at 101.3 kPa is 353.1 K.

Solution The three-phase temperature is first found out. At T*, is calculated

for each given temperature, and this is plotted against temperature. T* is the tempe-
rature at which P is equal to 101.3 kPa. This is found out to be 342 K. The horizontal line CD
in Fig. 8.24 is drawn at this temperature. The vapour pressures at this temperature are =
71.18 kPa and = 30.12 kPa. The mole fraction of benzene in the vapour represented by point E in
Fig. 8.24 is 71.18/101.3 = 0.70.
The dew-point curve BE is plotted by choosing a temperature lying between 373 K (boiling point of
water) and 342 K (the three-phase temperature). For example, take T = 353 K. The partial pressure of
water at the dew point is equal to the vapour pressure. For the dew-point temperature of 353 K, the
partial pressure of water is
47.32 = (1 – y) ฀ P
where y is the mole fraction of benzene in the vapour. We get, y = 0.5329.
This calculation is repeated for various temperature and the entire curve BE is drawn. For getting the
curve AE the procedure is the same. Here temperatures are assumed between 342 K and 353.1 K, the
latter being the boiling point of pure benzene. On the curve AE, the partial pressure of benzene in the
vapour equals its vapour pressure. For example, for a dew-point temperature of 348 K,
85.31 = P ฀ y, or y = 0.8422
The following table gives the results of a few such calculations:

T, K 342 348 353 353.1 363 373

y (curve AE) 0.70 0.84 0.99 1.00 – –
y (curve BE) 0.70 – 0.53 – 0.31 0

8.16 LIQUID–LIQUID EQUILIBRIUM DIAGRAMS

8.16.1 Binary Liquid–Liquid Equilibria
When two liquids are only partially miscible, the equilibrium can be represented on rectangular
coordinates as shown in Fig. 8.25. A dome-shaped region is formed by the mutual solubility
curves and within the dome the mixture exists as two phases. The compositions of the equilibrium
phases lie at the ends of the horizontal line at a given temperature. For example, the mixture M in Fig.
8.25 will separate into two equilibrium phases A and B. The relative amounts of the phases are given
by the inverse lever rule.

The point P gives the critical solution temperature. Outside the dome the mixture is homogeneous.

8.16.2 Ternary Equilibrium Diagrams

Liquid–liquid equilibria involving three components are important in the analysis of extraction
operations. The extraction process involves bringing a binary mixture of components A and C into
intimate contact with a solvent B. The solvent B is either partially soluble in liquid A or is immiscible
with it. The component C gets distributed in different proportions between the two insoluble phases
known as the ‘raffinate’ and the ‘extract’. The A-rich phase is known as the raffinate and the B-rich
phase is known as the extract. When the solvent added is only partially miscible with A, the extract and
raffinate phases contain three components. The ternary liquid–liquid equilibrium diagrams are usually
represented on equilateral triangular coordinates. On the equilateral triangle the length of thealtitude
is allowed to represent 100% composition and the length of the perpendiculars from any point to the
bases represent the percentages of the three components. The apexes of the triangle represent the pure
componen ts A, B, and C and points on the sides represent binary mixtures.
Figure 8.26(a) shows the equilibrium diagram of type-I systems in which one pair is partially soluble.
The pairs A-C and B-C are miscible in all proportions and the pair A-B is miscible only partially.
Examples are water (A)–chloroform (B)–acetone (C), water (A)–benzene (B)–acetic acid (C), water
(A)–methyl isobutyl ketone (B)–acetone (C), etc. Liquid C dissolves completely in A and B whereas A
and B dissolve only to a limited extent in each other. In the region below the mutual solubility curves
the two liquid phases exist under equilibrium. The compositions of the equilibrium phases are
obtained at the ends of the tie line passing through the point representing the overall composition of
the mixture. For example, the mixture, whose combined composition is represented by point M separates
into a raffinate R and an extract E at equilibrium. Thus, RE is a tie line for the system. The weight
fraction of C in the raffinate is denoted by xR, and that in the extract is denoted by yE. Severaltie lines
can be drawn and each gives rise to a set of equilibrium xR and yE values, which can be used to plot
the equilibrium diagram shown in Fig. 8.26(b).

The curve DRPF is the binodal solubility curve, which shows the change in the solubility of A-rich
and B-rich phases upon addition of C at a fixed temperature. Any mixture outside this curve will be a
homogeneous solution of a single liquid phase. There is one point on the binodal curve P, which will
represent the last of the tie lines where the A-rich and B-rich phases become identical. It is known as
the plait point. With increase in temperature the mutual solubilities of A and B increase and as a result
the heterogeneous area shrinks. Above the critical solution temperature of the binary A-B they dissolve
completely and the heterogeneous area vanishes completely. Extraction is not possible under this
condition.
The ternary equilibrium diagram for type II systems is shown in Fig. 8.27. In this type of systems, tw
pairs are partially soluble. Examples of type II systems are chlorobenzene (A)–water (B)–Methylethyl
ketone (C) , n-heptane (A)–aniline (B)–methylcyclohexane(C), etc. Here A and C are completely
miscible while A-B and B-C pairs show only limited solubility. Points D and F represent the mutual
solubility of A and B and points H and G those of B and C at the prevailing temperature. Curves DRH
and FEG are the ternary solubility curves. Mixtures such as at M inside the heterogeneous area form
two liquid phases in equilibrium at E and R. As temperature is increased the mutual solubilities
increase and above the critical solution temperature of the binary pair B-C, the system becomes
identical to type I system.