8.

1 Vapor–Liquid Equilibrium (VLE) ◄ 467

► For VLE, LLE, and SLE, relate phase diagrams schematically to the Gibbs
energy of each phase in the mixture using the minimization of Gibbs energy
to determine the equilibrium state of the mixture.
► Calculate the following colligative properties of a dilute solution: boiling-
point elevation, freezing-point depression, and osmotic pressure.
► Fit parameters in binary activity coefficient models on your own and with
ThermoSolver from experimental data with objective functions or by linear
regression, when appropriate.

►8.1 VAPOR–LIQUID EQUILIBRIUM (VLE)

The most common phase equilibria problems chemical engineers encounter involve
vapor–liquid equilibrium (VLE). We can write the general expression for VLE by apply-
ing the defining relations in Chapter 7. At equilibrium, the fugacity of species i in the
vapor and liquid are equal:

f^iv 5 f^il

If we choose to quantify the vapor-phase nonideality using the fugacity coefficient

[Equation (7.5)] and the liquid-phase nonideality using the activity coefficient [Equation
(7.32)],1 we get:

yiw^ vi P 5 xigilfio (8.2)

Here yi represents the mole fraction in the vapor phase, while xi represents the liq-
uid mole fraction. Once we have chosen the appropriate liquid reference state (Lewis/
Randall rule or Henry’s law), we can solve this problem if we have composition depend-
encies of the activity coefficient and the fugacity coefficient. Equation (8.2) is actually
a set of coupled equations, one for each species i. While Equation (8.2) is completely
rigorous (and thus always correct), we have seen that it is not always trivial to calculate
these terms.
Hence, we will begin our exploration of VLE by considering the limiting case of
Equation (8.2), which is valid only under certain circumstances. When we make approxi-
mations to this equation, we must keep in mind that the analysis that follows is valid only
for the specific cases where the approximations apply.

Raoult’s Law (Ideal Gas and Ideal Solution)

Consider the case when we are at low pressure and all the intermolecular forces are
approximately the same. We can treat the vapor as an ideal gas and the liquid as an ideal
solution. If we pick the Lewis/Randall reference state 1 fio 5 fi 2 , our criteria for equilib-
rium can be simplified to:

yiP 5 xifi

1
Alternatively, we can also use a fugacity coefficient for the liquid phase. This approach, which requires an
accurate equation of state for the liquid, will be explored later in this chapter.

c08.indd 467 01/11/12 10:09 AM

 468 ► Chapter 8. Phase Equilibria III: Applications

Applying Equation (7.37) for the pure species fugacity, we get:

yiP 5 xiPisat (8.3)

You will recognize Equation (8.3) as Raoult’s law, which you have undoubtedly seen
before. It directly results from the criteria for equilibrium [Equation (8.1)] under the
special circumstances described above (ideal gas, ideal solution, Lewis/Randall reference
state). This equation is convenient, since the saturation pressure of species i depends
only on the temperature of the system. The relation between Psat i and T is commonly fit
to the Antoine equation. Appendix A.1 provides Antoine equation parameters for several
species.
Equation (8.3) is often rewritten:

yi 5 Kixi

When Raoult’s law applies, we have:

Pisat 1 T only 2
Ki 5
P

Ki is termed the K-value2 of species i and, in this case, depends only on the tempera-
ture and the pressure of the system. Equation (8.6) is frequently used in hydrocarbon
systems, where extensive data for K-values are available. In fact, this approach can be
extended beyond the limiting assumptions of Raoult’s law. If we use the Lewis fugacity
rule to correct for the fugacity of real gases at higher pressure and apply the Poynting
correction to the Lewis/Randall reference state, K remains independent of composi-
tion since both these corrections are based on pure species properties. Thus, K can be
expressed for any species solely as a function of T and P. In some cases K-values are even
extended to real gases, where the ideal gas and ideal solution assumptions discussed
above no longer apply. In such a case, the nonideality of the system is buried in K. Now
K depends on composition and must be found either empirically or through fugacity and
activity coefficients, as described by Equation (8.2).
In order to further explore the implications of Raoult’s law, let’s consider a binary
solution containing species a and species b. We can write the criteria for equilibrium
given by Equation (8.3) for each species:

yaP 5 xaPasat (8.4)

and, ybP 5 xbPbsat (8.5)

We then obtain an expression for the total system pressure by adding together Equations
(8.4) and (8.5):
yaP 1 ybP 5 P 5 xaPasat 1 1 1 2 xa 2 Pbsat (8.6)

Plugging Equation (8.6) into (8.4) gives:

xaPasat
ya 5 (8.7)
xaPasat 1 1 1 2 xa 2 Pbsat
2
The K-value is the analogous quantity to phase equilibrium that the equilibrium constant is to chemical
reaction equilibrium (Chapter 9).

c08.indd 468 01/11/12 10:09 AM

 8.1 Vapor–Liquid Equilibrium (VLE) ◄ 469

Equations (8.6) and (8.7) can be generalized to a system with m components. In this
case, the sum of the partial pressures gives:

m
P 5 xaPsat
a 1 xbPb c 1 xiPi c 1 xmPm 5 a xiPi
sat sat sat sat (8.8)
i51

and the mole fraction of species i is given by:

xiPsat
i
yi 5 m
(8.9)
sat
a xiPi
i51

Equations (8.8) and (8.9) apply to multicomponent mixtures where the vapor phase is
assumed to be an ideal gas and the liquid phase, an ideal solution.
Bubble Point and Dew Point Calculations
Four common types of vapor–liquid equilibria calculations are illustrated by a grid in Fig-
ure 8.1. In a bubble-point calculation, the liquid-phase mole fractions of the system are
specified and the vapor mole fractions are solved for. The solution represents the com-
position of the first bubble of vapor that forms when energy is supplied to a saturated
liquid. Conversely, in a dew-point calculation, the liquid mole fractions are determined
given the vapor mole fractions. This case corresponds to the composition of the first
drop of dew that forms from a saturated vapor. Bubble- and dew-point calculations are
represented by the two columns in Figure 8.1. In addition to knowing the composition,
the value of either the temperature or the pressure needs to be specified to constrain the
state of the system. The former case is represented by the first row in Figure 8.2, while
the latter case is represented by the second row. Hence, the grid in Figure 8.2 represents
four typical combinations of independent and dependant variables found in VLE prob-
lems. They are defined by the quadrants I, II, III, and IV for reference in Examples 8.2,
8.3, and 8.5. When faced with such a calculation, it is helpful to recognize the independ-
ent and dependant variables appropriately. For binary systems that follow Raoult’s law, it
is possible to solve for the vapor and liquid mole fractions when T and P are known (see
Problem 8.22). However, in mixtures containing three or more components, specifying
only T and P underconstrains the problem (i.e., there are multiple possible solutions).

Bubble Point Dew Point

(xi known) (yi known)

II I
T known

Find yi, P Find xi, P

Pisat set Pisat set

III IV
P known

Find yi, T Find xi, T

Pisat unknown Pisat unknown

Figure 8.1 Grid of common VLE calculations.

c08.indd 469 01/11/12 10:09 AM

 470 ► Chapter 8. Phase Equilibria III: Applications

EXAMPLE 8.1 Consider a system with liquid containing 30% n-pentane (1), 30% cyclohexane (2), 20% n-hexane
Bubble-Point (3), and 20% n-heptane (4) at 1 bar. Determine the temperature at which this liquid develops
Calculation with P the first bubble of vapor. What is the vapor composition?
Known
SOLUTION This problem corresponds to quadrant III in the grid of Figure 8.1. Since the com-
ponents in this system are chemically similar, we will assume an ideal solution. Additionally,
at 1 bar, we may assume an ideal gas; thus, we can write the criterion for phase equilibrium in
terms of Raoult’s law for each of the components, that is:

yiP 5 xiPisat (E8.1A)

Since the sum of the partial pressures equals the system pressure, we get:

P 5 a yiP 5 a xiPisat 5 x1P1sat 1 x2P2sat 1 x3P3sat 1 x4P4sat (E8.1B)

The saturation pressure can be found according to Antoine’s equation:

Bi
ln Pisat 3 bar 4 5 Ai 2 (E8.1C)
T K 4 1 Ci
3

where the coefficients for parameters Ai, Bi, and Ci are reported in Table E8.1A.
Substitution of Equation (E8.1C) into (E8.1B) results in one equation with one
unknown—T. This equation can be solved by trial and error, by Excel using “solver,” or by
other graphical or numerical techniques to give:

T 5 333 3 K 4

Saturation pressures and vapor-phase mole fractions for each of the species at this temperature
can be calculated according to Equations (E8.1C) and (E8.1A), respectively. Their values are
presented in Table E8.1B.
We see that proportionately much more of the lighter n-pentane goes into the vapor while
very little of the heavier n-heptane does. This result forms the basis for separation by distillation.

TABLE E8.1A Antoine Coefficients

Species n-C5H12 C6H12 n-C6H14 n-C7H16

Ai 9.2131 9.1325 9.2164 9.2535
Bi 2477.07 2766.63 2697.55 2911.32
Ci 239.94 250.50 248.78 256.51

TABLE E8.1B Saturation Pressures and Mole Fractions at T 5 333 [K]

Species n-C5H12 C6H12 n-C6H14 n-C7H16

Psat at 333 3 K 4 2.13 bar 0.514 bar 0.757 bar 0.218 bar
yi 0.639 0.154 0.151 0.056

c08.indd 470 01/11/12 10:09 AM

 8.1 Vapor–Liquid Equilibrium (VLE) ◄ 471

EXAMPLE 8.2 Consider a system with vapor containing 30% n-pentane (1), 30% cyclohexane (2), 20% n-hexane
Dew-Point (3), and 20% n-heptane (4) at 1 bar. Determine the temperature at which this vapor develops the
Calculation with P first drop of liquid. What is the liquid composition?
Known
SOLUTION This problem is the dew-point analog to Example 8.1. It corresponds to quadrant
IV in the grid of Figure 8.1. We can write Raoult’s law for each species as:

yiP
xi 5 (E8.2A)
Pisat

Since the sum of the liquid mole fractions equals 1:

yiP y1P y2P y3P y4P
1 5 a sat 5 sat 1 sat 1 sat 1 sat (E8.2B)
Pi P1 P2 P3 P4

Again, the saturation pressures can be written in terms of Antoine’s equation:

Bi
ln Pisat 3 bar 4 5 Ai 2 (E8.2C)
T K 4 1 Ci
3

where the coefficients for parameters A, B, and C are reported in Table E8.1A.
Substitution of Equation (E8.2C) into (E8.2B) gives one equation with one unknown—T.
The temperature is solved to be:
T 5 349 3 K 4

Saturation pressures and liquid-phase mole fractions for each of the species at this temperature
can be calculated according to Equations (E8.2C) and (E8.2A), respectively. Their values are
presented in Table E8.2.
In this case, proportionately much more of the heavier n-heptane is condensed while little
of the lighter n-pentane is, again helping us separate by distillation.
TABLE E8.2 Saturation Pressures and Mole Fractions at T 5 349 [K]

Species n-C5H12 C6H12 n-C6H14 n-C7H16

Psat at 349 3 K 4 3.296 bar 0.870 bar 1.256 bar 0.494 bar
xi 0.091 0.345 0.159 0.405

EXAMPLE 8.3 A compressed liquid feed stream containing an equimolar mixture of n-pentane and n-hexane
Isothermal Flash VLE flows into a flash unit as shown in Figure E8.3 at flow rate F. At steady state, 33.3% of the feed
Calculation stream is vaporized and leaves the drum as a vapor stream with flow rate V. The rest leaves as
liquid with flow rate L. If the flash temperature is 20°C, what is the pressure required? What
are the composition of the liquid and vapor exit streams?

SOLUTION A mass balance on component a gives:

xa,feedF 5 yaV 1 xaL (E8.3A)

(Continued)

c08.indd 471 01/11/12 10:09 AM