Thermodynamic Formulations for

Vapor/Liquid Equilibrium
Chapter 13
Excess Gibbs Energy and Activity Coefficient
The Gamma/Phi Formulation of VLE
Raoult’s Law, Modified Raoult’s, Henry’s Law
Models for the Excess Gibbs Energy
Fitting Activity Coefficient Models to VLE
Flash Calculations
 Introduction
• The foundation for VLE calculations was presented in Chapter 10 for pure
species: v l sat
fi = fi = fi
Criterion of vapor/liquid
equilibrium of pure species: & φiv = φil = φisat

∧ ∧v
• For vapor/liquid equilibrium of species in mixture: l
fi = fi
Recalling that, ∧
∧ fi
φi =
xi P

• An analogous equation can be written for the liquid phase, but this phase is
often treated differently

Chemical Engineering Thermodynamics II CHE 341 13-2

 Excess Gibbs Energy and Activity Coefficients
∧
Previously, we defined: G i = µi = Γi (T ) + RT ln f i
For the special case of species i in an ideal solution, we can write:
id ∧id
µiid = Gi = Γi (T ) + RT ln fi
 ∧  ∧
id  fi  fi
Subtracting: G i − G i = RT ln
x f  Define, γi =
 i i
  xi fi
E
Gi = RT ln γ i Activity
∧ coefficient
For ideal solution, G iE = 0 ⇒ γi = 1 or f iid = xi f i of species i

Chemical Engineering Thermodynamics II CHE 341 13-3

 The Gamma/Phi Formulation of VLE
Consider the VLE vessel shown in the drawing, where vapor
mixture exists in equilibrium with the liquid mixture Vapor
• T and P are uniform throughout the vessel. T, P, {yi}

• T, P, {xi} and {yi} are normally given as they can be measured

experimentally. Liquid
∧v ∧ T, P {xi}
• For species i in the vapor phase: fi = yi φi P
∧l
• For species i in the liquid phase: f i = xiγ i f i (i = 1, 2, ….N)
∧v ∧l
Criterion for equilibrium necessitates that: fi = fi
∧
φ P = xiγ i f i
yi i
• The liquid phase fugacity, fi, is given by: f i = φ
sat sat
exp 
(
Vil P − Pi sat  )
i Pi 
 RT 
Substitution, ∧ sat sat
yi φi P = xiγ iφi Pi exp 
Vi
l
P − Pi (
sat
 )
RT 
 

Chemical Engineering Thermodynamics II CHE 341 13-4

  Vil (P − Pi sat )
∧
Define φ i
Φi = exp −
sat 
φ 
i RT 
yi Φ i P Used to calculate γi for a given:
Then, γi =
xi Pi sat T, P, {xi} and {yi} data

Low Pressure VLE

At low P, up to 1 bar, vapor phase usually approximates as an ideal gas
∧
⇒ φ = φisat = 1
i
also the term exp[-Vil(P – Pisat)/RT] ≈ 1.0
∧ sat
Moreover, the ratio φ φ i / i ≈ 1.0 Even at higher P, if differs slightly from 1.0.
Therefore, under these assumptions we can write
yi P Modified Raoult’s Law
γi =
xi Pi sat
This simple equation allows the calculation of γi from the experimental low-
pressure VLE data (N.B. γi = 1 ⇒ yi P = xi Pisat, Raoult’s)

Chemical Engineering Thermodynamics II CHE 341 13-5

 Raoult’s Law
• According to the previous mentioned two assumptions, the mathematical
expression of Raoult’s law is:
xi: liquid-phase mole fraction
sat
yi P = partial pressure yi P = xi Pi yi: vapor-phase mole fraction
sat
of the system Pi : vapor pressure of pure species i at
the temperature of the system

 Raoult’s law provides the equilibrium relationship between the concentration

of species i in liquid phase to that in the vapor phases.
⇒ Raoult’s law can be used to display simple VLE calculations and can serve as
standard of comparison with more complex systems.

 Raoult’s law can be used for the bubble and dew point calculations
 Limitation of Raoults law:
 Applied only to species at which T is below the critical temperature of the
species
 The calculations followed are valid only if Raoult’s law is hold (ideal vapor
and ideal liquid phases).
Chemical Engineering Thermodynamics II CHE 341 13-6
Discussion: “Degrees of Freedom”
Since Pisat = f (T ) only ⇒ Raoults law constitutes a set of N equations in the
variables P, T, {xi} and {yi}.
 Total number of variables = (N-1) yi terms
⇒ 2N independent
= (N-1) xi terms
variables
= 2 (P and T)
 Thus, if N of these variables are specified ⇒ the other N variables can be
determined from N Raoults equations
 In general, when using Raoults law, either T or P are specified and either
{xi} or {yi} composition ⇒ fixing 1+ (N-1) = N variables
 The above observations follows directly the phase rule F = 2 −π + N

 Example: 2-phase (L/V) binary mixture Variables: T, P, x1, y1

F = 2 – 2 +2 = 2 (e.g., specify T, x1 ⇒ P & y1 determined from Raoults)

 Example: 2-phase (L/V) ternary mixture Variables: T, P, x1, x2, y1, y2

F = 2 – 2 +3 = 3 (e.g., specify T, x1, x2 ⇒ P & y1, y2 determined from Raoults)

Chemical Engineering Thermodynamics II CHE 341 13-7

General Types of Phase Equilibria Calculations for Vapor-Liquid Systems

o BUBL P: (Direct) V
{yi} T
Given T and {xi} → P &
{yi} L P
{xi}
o DEW P: (Direct) 90
Constant T (i = 1, 2, …N)
Given T and {yi} → P & 80

L DEW
{xi}
70

PDew line

P
60

o BUBL T: (Iterative)
50

40
V
xx1 , y y1
30
0.0 0.2 0.4 0.6 0.8 1.0

Given P and {xi} → T & i i

{yi} o DEW P: (Iterative)

Constant P Constant P
95

Given
Bubble
95

90

V DEW
line P and {yi} 90

V line
85
85

80
TBubl TDEW
T

80

75
T
75

L T&
70

65

x y1
70

L
{xi}
0.0 0.2 0.4 0.6 0.8 1.0

x, y 1 x y1
65
0.0 0.2 0.4 0.6 0.8 1.0

xi, yi 1
i i

Chemical Engineering Thermodynamics II CHE 341 13-8

BUBL P Calculations
Given T & {xk} → P and {yk}
From Raoult’s law: k = 1, 2 ⋅ ⋅ ⋅ ⋅⋅, N
P ∑ yk = ∑ xk Pksat
k k

Since P

Then {yk}

Chemical Engineering Thermodynamics II CHE 341 13-9

DEW P Calculations
Given T & {yk} → P and {xk}

Raoult’s law gives:

 yk 
∑ xk = P ∑  
 P sat 
 k 
k k

Since
P

yk P
Then xk = sat
{xk}
Pk
Chemical Engineering Thermodynamics II CHE 341 13-10
BUBL T Calculations
Given P & {xk} → T and {yk}

From BUBL P above: P = ∑ xk Pksat k = 1, 2 ⋅ ⋅ ⋅ ⋅⋅, N

k
xk Pksat
Re-write as: P = Pi ∑
sat
P sat
where i denotes an arbitrary
chosen reference species
k i
P
sat
∴ Pi = (A)
∑ xkα ki
Pksat
where: α ki = sat = f (T ) (B)
Pi
If the vapor pressures are given by Antoine equation, then
 Bk   Bi 
ln α ki =  Ak −  −  Ai −
T + Ck  

T + Ci 
(C)

Chemical Engineering Thermodynamics II CHE 341 13-11
Iterative Solution for BUBL T
1. Start with arbitrary (Select T = ∑ xk Tksat as an initial estimate of T)
2. Calculate αki terms from (C)

3. Calculate Pi sat from (A)

4. Calculate an improved T from Antoine equation for species i:
Bi
T= sat
− Ci General Antoine Equation:
Ai − ln( Pi )
Bi
⇒ Check δT (temperature convergence) ln Pi sat = Ai −
T + Ci
5. Return to step 2 and iterate until convergence Pi sat = f (T ) (A)
6. After convergence of T, calculate yk from Raoults:
T = f ( Pi sat )
xk Pksat
yk = ln α ki = f (T ) (C)
P
Chemical Engineering Thermodynamics II CHE 341 13-12
DEW T Calculations
Given P & {xk} → T and {yk}

From DEW P above: P=

1
 yk 
k = 1, 2 ⋅ ⋅ ⋅ ⋅⋅, N
sat
∑  
 P sat 
 k 
Pi k
P= where i denotes an arbitrary
Re-write as:  Pi sat 
∑ yk  sat 
 Pk 
chosen reference species
k

 yk 
∴ Pi sat = P ∑   (A)
 α ki 
Pksat
where: α ki = sat = f (T ) (B)
Pi

 Bk   Bi 
and ln α ki =  Ak −  −  Ai −
  (C)
 T + Ck   T + Ci 
Chemical Engineering Thermodynamics II CHE 341 13-13
Iterative Solution for DEW T
1. Start with arbitrary (Select T= ∑ yk Tksat as an initial estimate of T)
2. Calculate αki terms from (C)

3. Calculate Pi sat from (A)

4. Calculate an improved T from Antoine equation for species i:
Bi
T= sat
− Ci
Ai − ln( Pi )
⇒ Check δT (temperature convergence)

5. Return to step 2 and iterate until convergence

6. After convergence of T, calculate yk from Raoults:

xk Pksat
yk =
P
Chemical Engineering Thermodynamics II CHE 341 13-14
Example: V/L Equilibrium of binary system
The binary system acetonitrile (1)/nitromethane (2) conform closely to Raoult’s
law. Given the following Antoine equations:
2945.47 sat 2972.64
sat
ln P1 (kPa ) = 14.2724 − o ln P2 ( kPa ) = 14 . 2043 −
T ( C) + 224.0 T (o C) + 209.0
(a) Plot the P-x1-y1 diagram at 75oC. Note: according to phase rule
(b) Plot the T-x1-y1 diagram at 70 kPa. F = 2 −π + N
= 2−2+2 = 2
(a) The P-x1-y1 diagram at T = 75oC 2 pieces of information
are needed (P, T, x, y)
P-x1 calculations BUBL line
From the Antoine equation, at 75oC P1sat
P1sat = 83.21 kPa, P2sat = 41.98 kPa, P
P = p1 + p2 = y1P + y2 P = x1P1sat + x2 P2sat P2sat
p1

P = 83.21x1 + 41.98(1 − x1 ) p2
∴ P = 41.23 x1 + 41.98 0 x1 1.0
Chemical Engineering Thermodynamics II CHE 341 13-15
P-y1 calculations DEW line
x1P1sat
Use: y1 =
P
Eg. If x1 = 0.6 P = 41.23(0.6) + 41.98 = 66.72 kPa
∴ y1 = (0.6)(83.21)/66.72 = 0.7483
• The results of calculations at 75oC for a number of values of x1 are tabulated
as shown in the following table
100 Bubble
x1 y1 P/kPa T = 75oC line
Saturated
0.0 0.0 41.98 (P2sat) 80 Subcooled liquid
liquid
0.2 0.3313 50.23 Dew
line
P-x1
P
60
0.4 0.5692 58.47 Saturated
vapor
0.6 0.7483 66.72 40
P-y1 Superheated
0.8 0.8880 74.96 vapor
1.0 1.0 83.21 (P1sat) 20
0.0 0.2 0.4 0.6 0.8 1.0

xi, yi

Chemical Engineering Thermodynamics II CHE 341 13-16

(b) The T-x1-y1 diagram at P = 70 kPa.
• In this case T varies as x1 and y1 vary.
• The temperature enters indirectly through Raoult’s law (via Pisat), and an
iterative solution is required.
• For a given P, the temperature range is bounded by T1sat and T2sat
(corresponding to x1 = 0 and x2 = 1, respectively).
• From the Antoine equation for P = 70 kPa: T1sat = 69.84 oC T2sat = 89.58 oC

Construction of the T-x1-y1 diagram x1 = 1 x1 = 0

Step 1: Select values of T between T1sat and T2sat
Step 2: Calculate the corresponding P1sat and P2sat
P − P2sat
Step 3: Evaluate x1 from P= x1P1sat + (1 − x1 ) P2sat ∴ x1 = sat
P1 − Pssat
x1P1sat
Step 4: Evaluate y1 using: y1 =
P
Step 5: Plot the corresponding values of T, x1 and y1 (see sample of calculations,
Table and figure).
Chemical Engineering Thermodynamics II CHE 341 13-17
 Sample of calculations: Take T = 78oC P1sat = 91.76, P2sat = 46.84 kPa
P − P2sat 70 − 46.84
x1 = sat sat
= = 0.515
P1 − Ps 91.76 − 46.84
x1P1sat (0.515)(91.76)
y1 = = = 0.6759
P 70
x1 y1 T/oC 95
P = 70 kPa
Superheated
T-y1
0.0 0.0 89.58 (T2 sat)
90 vapor
T2sat
0.1424 0.2401 86
85

0.3184 0.4741 82 T-x1

80 c
0.5156 0.6759 78
T
c’ Dew line
b’
0.7378 0.8484 74 75 b

1.0 1.0 69.84 (T1sat) 70 Subcooled a T1sat

liquid Bubble line
 Follow the vaporization at constant pressure
65
 b → Bubble point 0.0 0.2 0.4 0.6 0.8 1.0

 c → Dew point xi, yi

 bb’ & cc’ → Tie lines
Chemical Engineering Thermodynamics II CHE 341 13-18
 Calculation of T when liquid phase composition is specified
(BUBL T) for binary systems:
y1P = x1P1sat y2 P = x2 P2sat Iterative procedure
sat  P1sat  Step 1: Guess T (bewen T1sat & T2sat)
P= x1P1sat + x2 P2sat = P2  x1 sat + x2 
 P2  Step 2: Calculate α12 = f(T) from (B)
P Step 3: Calculate P2sat from (A)
whence P2sat = A
x1α12 + x2 Step 4: Calculate T from Antoine
P1sat  y1 / x1  Step 5: Return to 2 until convergence
where α12 = sat ≡
 y2 / x2
: relative volatility 

P2 Example: x1 = 0.6 & P = 70 kPa
 Although P1 and P2 are strong functions of T,
sat sat

α12 is a weak function A temperature of 76.42oC was

Values of α12 can be calculated easily from estimated using the above-
Antoine equation, i.e., mentioned procedure
2945.47 2972.64 @ T = 76.42oC→ P1sat = 87.17 kPa
ln α12 = 0.0681 − + B
T + 224.0 T + 209.4
x1P1sat (0.6)(87.17)
What do we have? Equation B: α12 = f(T) ∴ y1 = = = 0.7472
P 70
Equation A: P2sat = f(α12)
Antoine equation: T = f(P2sat)
Chemical Engineering Thermodynamics II CHE 341 13-19
 Calculation of T when vapor phase composition is specified
(DEW T) for binary systems:
yP y2 P Iterative procedure
x1 = 1sat x2 = sat Step 1: Guess T (bewen T1sat & T2sat)
P1 P2
Step 2: Calculate α12 = f(T) from (B)
 y1 y2 
Adding up: 1 =  sat + sat  P Step 3: Calculate P1sat from (A)
 P1 P2 
Which leads to: Step 4: Calculate T from Antoine
Step 5: Return to 2 until convergence
P1sat = ( y1 + y2α12 )P A

Where again:
2945.47 2972.64
ln α12 = 0.0681 − + B
T + 224.0 T + 209.4

What do we have? Equation B: α12 = f(T)

Equation A: P1sat = f(α12)
Antoine equation: T = f(P1sat)

Chemical Engineering Thermodynamics II CHE 341 10-20

 Henry’s Law
• Used when Pisat of one of the components in the mixture can not be determined .
• Applied for pressure low enough that vapor phase may be assumed as ideal gas.
• For a species present as a very dilute solute in the liquid phase, Henry’s law
states that pi ∝ xi, or
yi P = xi H i Hi: Henry’s constant
Determined experimentally
Example: See Table 10.1
Assuming that carbonated water contains only CO2 and H2O, determine
the compositions of the vapor and liquid phases in a sealed can of “soda”and pressure
exerted on the can at 10oC. Henry’s constant for CO2 in water at 10oC is about 990 bar.
CO2 (1): apply Henry’s law y1P = x1H1
H2O (1): apply Raoult’s law y2 P = x2 P2sat P2sat (10o C) = 0.01227 bar

(Please read the solution in Page 358)

Chemical Engineering Thermodynamics II CHE 341 10-21

 VLE by Modified Raoult’s Law
• To account for deviation from solution ideality in the liquid phase, a factor
is inserted into the Roult’s law, i.e.

Modified Raoult’s law: yi P = xiγ i Pi sat (i = 1, 2, …. N)

γi is called the activity coefficient γi = f(T, xi)

• Bubble and dew point calculations can be made as before but with a little more
complications
BUBL P: ∑ i
yi = 1 P= ∑i xiγ i Pi sat

1
DEW P: ∑x =1 P=
i i
∑ i
yi / γ i Pi sat

Chemical Engineering Thermodynamics II CHE 341 10-22

 Example 10.3
For the system methanol(1)/methyl acetate(2), the following equations provide a
reasonable correlation for the activity coefficients:
ln γ 1 = Ax22 ln γ 2 = Ax12 A = 2.771 − 0.00523T
In addition the following Antoine equations provide vapor pressures:
3,643.31 2,665.54
ln P1sat (kPa) = 16.59153 − ln P2sat (kPa ) = 14.25326 −
T ( K ) − 33.424 T ( K ) − 53.424
Assuming the validity of modified Raoult’s law, calculate:
(a) P and {yi}, for T = 318.15 K and x1 = 0.25 BUBL P calculation
@ T = 318.15 K P1sat = 44.51 P2sat = 65.64 kPa
A = 2.771 – (0.00523)(318.15) = 1.107
γ1 = exp[(1.107)(0.75)2] = 1.864 γ2 = exp[(1.107)(0.25)2] = 1.072

∴P = ∑i xiγ i Pi sat P = x1γ 1P1sat + x2γ 2 P2sat

P = (0.25)(1.864)(44.51) + (0.75)(1.072)(65.64) = 73.50 kPa
x1γ 1P1sat y2 = 0.718
y1 = y1 = 0.282
P
Chemical Engineering Thermodynamics II CHE 341 10-23
(b) P and {xi}, for T = 318.15 K and y1 = 0.6 DEW P calculation

P1sat = 44.51 P2sat = 65.64 kPa A = = 1.107

1
∴P = x1 is needed to calculate γ1 and γ2
y1 / γ 1P1sat + y2 / γ 2 P2sat
Iterative procedure is required.
Steps
1. For initial values, set γ1 = γ2 = 1.0
2. Calculate P using the current values of γ1 and γ2 using: P = 1
y1P y1 / γ 1P1sat + y2 / γ 2 P2sat
3. Calculate x1 by: x1 = x2 = 1 − x1
γ 1P1sat
Ax Ax
4. Evaluate γ1 and γ2 using current values of x1 and x2: γ 1= e γ 2= e
2 2

2 1

5. Go to step 2 and repeat 2- 4 till convergence (in x1 and P).

When carrying out the above steps, the following final values can be obtained:
P = 62.89 kPa x1 = 0.8169 γ1 = 1.0378 γ2 = 2.0935

Chemical Engineering Thermodynamics II CHE 341 10-24

(c) T and {yi}, for P = 101.33 kPa and x1 = 0.85 BUBL T calculation
This is iterative procedure…
1. Initial estimate of the temperature using Antoine equation, solved for T:
Bi
Ti sat = − Ci
Ai − ln P
Application for P = 101.33 kPa leads to: T1 = 337.71
sat
T2sat = 330.08 K
∴ T = (0.85)(337.71) + (0.15)(330.08) = 336.57 K
2. For the current values of T, calculate values for A, γ1, γ2 and α ≡ P1sat/P2sat
P
3. Find a new value for P1sat using: P1sat =
x1γ 1 + x2γ 2 / α
B1
4. Find a new value for T from the Antoine equation for species 1: T = sat
− C1
A1 − ln P1
5. Return to step ; iterate to convergence on a value for T.
This process yields the following final values:
T = 331.2 K P1sat = 95.24 P2sat = 48.73 kPa
A = 1.0388 γ1 = 1.0236 γ2 = 2.1182

∴ y1 = x1γ1P1sat/P = 0.670 y2 = 1- y1 = 0.33

Chemical Engineering Thermodynamics II CHE 341 10-25
(d) T and {xi}, for P = 101.33 kPa and y1 = 0.4 DEW T calculation
This is iterative procedure…
1. Initial estimate of the temperature as in the previous part
@ P = 101.33 kPa T1sat = 337.71 T2sat = 330.08 K
∴ T = (0.85)(337.71) + (0.15)(330.08) = 336.57 K
Set γ1 = γ2 = 1.0 (because x1 is unknown)
2. For the current values of T, calculate values for A and α ≡ P1sat/P2sat
3. Calculate x1, γ1 and γ2 using: x1 = y1P
Ax
Ax γ 2= e
2

γ 1= e
2

x2 = 1 – x1
1
2

γ 1P1sat
4. Find a new value for P1sat using: y y 
P1sat = P 1 + 2 α 
 γ1 γ 2  B1
5. Find a new value for T from the Antoine equation for species 1: T = sat
− C1
A1 − ln P1
6. Return to step ; iterate to convergence on a value for T.
T = 326.7 K P1sat = 64.63 P2sat = 90.89 kPa
The process yields
the final values: A = 1.0624 γ1 = 1.3629 γ2 = 1.2523
∴ x1 = 0.4602 x2 = 0.5398
Chemical Engineering Thermodynamics II CHE 341 10-26
(e) The azeotropic pressure, and the azeotropic composition, for T = 318.15 K
- First determine whether or not an azeotrope exists at the given temperature
- This calculations is facilitated by the definition of a quantity called the relative
volatility:
y1 / x1
α12 ≡ At an azeotrope: y1 = x1, y2 = x2 α12 = 1.
y 2 / x2
yi γ i Pi sat γ 1 P1sat
In general, = α12 =
xi P γ 2 P2sat
Ax @ x1 = 0, γ1 = eA and γ2 = 1;
γ 1= e
2

Since, Ax12
2
γ 2= e @ x1 =1, γ1 = 1 and γ2 = eA.
Therefore, in these limits, (α ) P1sat e A P1sat
12 x =0 = sat (α12 ) x =1 = sat A
1
P2 P2 e 1

@ T = 318.15 K: P1sat = 44.51 P2sat = 65.64 kPa A = 1.107

(44.51)e1.107 44.51
(α12 ) x =0 = = 2.052 (α12 ) x =1 = 1.107
= 0.224
1
65.54 1
(65.64)e
Since the value at one limit is greater than 1, whereas the value at the other limit is less than
1, an azeotrope exists, because α12 is a continuous function of x1 and must pass through the
value of 1.0 at some intermediate composition.
Chemical Engineering Thermodynamics II CHE 341 10-27
 γ 1P1sat
Since, for the azeotrope α12 = 1 α12 = sat
=1
γ 1P1
or γ az P sat 65.64
1 2
= = = 1.4747
γ 2az P1sat 44.54
Since, ln γ 1 = Ax22 & ln γ 2 = Ax12
γ1
∴ ln = Ax22 − Ax12 = A[( x1 − 1) 2 − x12 ] = A(1 − 2 x1 )
γ2

∴ ln(1.4747) = 1.107(1 − 2 x1az ) x1az = 0.325

Since at the azeotropic point, x1az = y1az
Modified Raoult’s for species 1 becomes,
P az = γ 1az P1sat = (1.657)(44.51) = 73.76 kPa

Chemical Engineering Thermodynamics II CHE 341 10-28

 VLE from K-Value Correlation
Equilibrium ratio Ki: A measure of the tendency of a given species to partition itself
preferentially between liquid and vapor phases
yi (called a K-value)
Ki ≡
xi
- K-value can be used as a measure of the “lightness” of a constituent species.
- Tendency to present in the vapor phase
- If Ki >1 ⇒ species i exhibits higher concentration in the vapor phase
Pi sat
K-value for Raoult’s law is K i ≡
P
γ i Pi sat
K-value for Modified Raoult’s law is: K i ≡
P
For Bubble point calculations: yi = K i xi , because ∑ y = 1 ∑i Ki xi = 1
i i

For Dew point calculations: xi = yi / K i , because ∑i xi = 1 ∑i yi / K i = 1

 K-value = f(T, P): For mixture of light hydrocarbons and simple molecules,
correlations are available in monographs form (Figs. 10.13,14)
Chemical Engineering Thermodynamics II CHE 341 10-29
DePriester chart for K-values of light hydrocarbons. Low-
temperature range

Chemical Engineering Thermodynamics II CHE 341 10-30

DePriester chart for K-values of light hydrocarbons. High-
temperature range

Chemical Engineering Thermodynamics II CHE 341 10-31

Example 10.4: For a mixture of 10 mol-% methane, 20 mol-% ethane,
and 70 mol-% propane at 50oF, use the K-value method to determine,
(a) The dew point pressure. (b) The bubble point pressure
(a) At the dew point, the given composition are for the vapor phase, yi, and it is
required to determine xi and Pdew. P is needed to determine the Ki value, thus an
iterative procedure is called. This is illustrated as follows:

P = 100 psi P = 150 psi P = 126 psi

Species yi Ki yi/Ki Ki yi/Ki Ki yi/Ki
Methane 0.10 20.0 0.005 13.2 0.008 16.0 0.006
Ethane 0.20 3.25 0.062 2.25 0.089 2.65 0.065
Propane 0.70 0.92 0.761 0.65 1.077 0.762 0.919
∑(yi/Ki) = 0.828 ∑(yi/Ki) = 1.174 ∑(yi/Ki) = 1.00

∴ Pdew = 126 psi

xmethane = 0.006, xEthane = 0.0065, xPropane = 0.919

Chemical Engineering Thermodynamics II CHE 341 10-32

(a) At the bubble point, the given composition are for the liquid phase, xi, and it
is required to determine yi and Pbubble. P is needed to determine the Ki value,
thus again an iterative procedure is called. This is illustrated as follows:

P = 380 psi P = 400 psi P = 385 psi

Species xi Ki Kixi Ki Kixi Ki Kixi
Methane 0.10 5.60 0.560 5.25 0.525 5.49 0.549
Ethane 0.20 1.11 0.222 1.07 0.214 1.10 0.220
Propane 0.70 0.335 0.235 0.32 0.224 0.33 0.231
∑(Kixi) = 1.017 ∑(Kixi) = 0.963 ∑(Kixi) = 1.00

∴ Pbubble= 385 psi

xmethane = 0.549, xEthane = 0.220, xPropane = 0.231

Chemical Engineering Thermodynamics II CHE 341 10-33

 Flash Calculations
Theory: Flash means that a liquid at pressure greater than or equal to its bubble
point pressure “flash” or partially evaporates when the pressure is reduced
below the bubble point pressure producing two-phase system of vapor and
liquid in equilibrium. V, {yi}

Pressure V moles V, L, {xi} & {yi} are unknown

reduction {yi} for a given T and P.
Liquid F 1 mole
under L moles
pressure {zi}
{xi}

L, {xi}
- Assume that T, P and the overall composition {zi} are known and that the
system exists in equilibrium
Basis: 1 mole of feed liquid (of mole fraction {zi}.
Material balance: Assuming that no chemical reactions occur
L+V=1
x L + yV = z (i = 1, 2, …., N)
i i i
Chemical Engineering Thermodynamics II CHE 341 10-34
Combining the last two equations to eliminate L gives:
xi (1 − V ) + yiV = zi
zi K i
Substituting xi = yi/Ki, and solving for yi yields: yi = (A)
1 + V ( K i − 1)
Since ∑yi = 1, then summing the previous equation over all species:
zi K i
∑i 1 + V ( K i − 1)
=1 (B)

- In flash calculations, T, P and {zi} are known, the only unknown is V.

- The solution involves trial and error, for N > 2,
- After determining V from (B), {yi} are determined from (A) and {xi} from xi = yi/Ki
- There is no guarantee at the outset that the existing material is in the two phase
region. This has to be verified before attempting to solve the problem
Before proceeding with the flash calculation, make sure that
Pdew < P < Pbubble
This requires BUBL P and DEW P calculations

Chemical Engineering Thermodynamics II CHE 341 10-35

Example 10.5: The system acetone(1)/acetonitrile(2)/nitromethane(3) at 80oC
and 110 kPa has the overall composition, z1 = 0.45, z2 = 0.35, z3 = 0.2. Assuming
that Raoult’s law is appropriate to this system, determine L, V, {xi}, and {yi}.
The vapor pressures of the pure species at 80oC are:
P1sat = 195.75 , P2sat = 97.84, P3sat = 50.32 kPa.

First, do BUBL P with {zi} = {xi}

Pbubl = x1P1sat + x2 P2sat + x3 P3sat
= (0.45)(195.75) + (0.35)(97.84) + (0.20)(50.32) = 132.40 kPa
Second, do DEW P with {zi} = {yi}
1
Pdew = = 101.52 kPa
y1 / P1sat sat
+ y2 / P2 + y3 / P3sat

Pdew < P < Pbubl System is in the two phase region

and flash calculation can be made
Pi sat K1 = 1.7795 K2 = 0.8895
For Raoult’s: K i =
P K3 = 0.4575

Chemical Engineering Thermodynamics II CHE 341 10-36

Substitute in (B): (0.45)(1.7795) (0.35)(1.8895) (0.20)(0.4575)
+ + =1
1 + 0.7795V 1 − 0.1105V 1 − 0.5425V
Solution for V by trial and error gives: V = 0.73664 mol
L = 1 – V = 0.2636 mol

zi K i y1 = 0.5087
yi =
1 + V ( K i − 1) y2 = 0.3389
y3 = 0.1524

x1 = 0.2859
∑ x =∑ y
i i i i =1
xi = yi / K i x2 = 0.3810
x3 = 0.3331

Chemical Engineering Thermodynamics II CHE 341 10-37

Example 10.6: For the system in Example 10.4, what fraction of the system is vapor
when the pressure is 200 psia and what are the compositions of the equilibrium
vapor and liquid phases
Based on the results of Example 10.4: Pdew (126 psia) < P = 200 < Pbubl (385 psia)
The system consists of two phases
With K-values provided T and P, by trial and error find V that satisfies:
(0.1)(10) (0.2)(1.76) (0.7)(0.52)
+ + =1
1 + 9V 1 + 0.76V 1 − 0.48V ⇒ V = 0.273

Methane Ethane Propane

zi K i y1 = 0.289
yi =
1 + V ( K i − 1) y2 = 0.292
y3 = 0.419

x1 = 0.029
∑ x =∑ y
i i i i =1
xi = yi / K i x2 = 0.166
x3 = 0.805
Chemical Engineering Thermodynamics II CHE 341 10-38
 Correlations for Liquid-Phase Activity Coefficients: Practical example

• The table below presents the P-x1-y1 data at 50oC for the binary system Methyl Ethyl
Ketone(1)/Toluene
P/kPa x1 y1 lnγ1 lnγ2 GE/RT GE/x1x2RT
12.3(P2sat) 0.0000 0.0000 0.000 0.000
15.51 0.0895 0.2716 0.266 0.009 0.032 0.389
18.61 0.1981 0.4565 0.172 0.025 0.053 0.342
21.63 0.3193 0.5934 0.108 0.049 0.068 0.312
24.01 0.4232 0.6815 0.069 0.075 0.072 0.297
25.92 0.5119 0.7440 0.043 0.100 0.071 0.283
27.96 0.6096 0.8050 0.023 0.127 0.063 0.267
30.12 0.7135 0.8639 0.010 0.151 0.051 0.248
31.75 0.7934 0.9048 0.003 0.173 0.038 0.234
34.15 0.9102 0.9590 -0.003 0.237 0.019 0.227
36.09(P1sat) 1.0000 1.0000 0.000 0.000

• lnγ1 and lnγ2 can be calculated from γi = yi P/xi Pisat

• Calculate GE/RT from the summability relation: G E / RT = x1 ln γ 1 + x2 ln γ 2
• Calculate values for GE/x1x2RT for each point
• The data in the above table can be presented in graphical forms as shown below.
Chemical Engineering Thermodynamics II CHE 341 12-39
 • In these figures, symbols (circles, squares, etc) represent experimental data, while
solid lines represent various theoretical or empirical equations
• The preceding figures are typical of binary liquid systems for which:
γi ≥ 1.0 or lnγi ≥ 0.0
⇒ Such systems exhibit “positive deviation" from Raoult’s law, where P-x1 data
points all lie above the dashed line representing the linear relation of Raoult’s law

Chemical Engineering Thermodynamics II CHE 341 12-40

 Limiting Behavior on the Preceding Figures
From the VLE data figures, the following can be noted:
• As xi → 1, γi → 1 (or lnγi → 0) Pure species
sat
yi P (1) P
Proof: xlim γ i = lim sat = lim i
sat
=1
i →1 xi →1 xi Pi xi →1 (1) P
i
• As xi → 0 (infinitely dilute solutions), γi → γi∞ A finite limit, not zero
• As xi → 0, GE/RT → 0 Pure species
This is because, G E / RT = x1 ln γ 1 + x2 ln γ 2
lim G E / RT = (0) ln γ 1∞ + (1) ln(1) = 0.0
x1 →0
Similarly, lim G E / RT = (1) ln(1) + (0) ln γ 2∞ = 0.0
x2 → 0
∴ At both limits GE/RT → 0
• At each x1 value, the slopes of lnγ1 and lnγ2 curves are opposite.
At x2 → 0 (x1 → 1) ⇒ Slope of lnγ1 curve = 0 Pure species 1
At x1 → 0 (x2 → 1) ⇒ Slope of lnγ2 curve = 0 Pure species 2

Chemical Engineering Thermodynamics II CHE 341 12-41

• As xi → 0, GE/x1x2RT → lnγi∞
Proof:
Since GE/x1x2RT is indeterminate at x1 = 0 or x1 = 1
⇒ Use de’Lopital rule for the limit
GE G E / RT d [G E / RT ]
lim = lim = lim
x1 →0 x x RT x1 →0 x1 x1 →0 dx1
1 2
d
= lim ( x1 ln γ 1 + x2 ln γ 2 )
x1 →0 dx
1
x 1 + x2 = 1
 d ln γ 1 d ln γ 2 
= lim  x1 + ln γ 1 + x2 − ln γ 2  ⇒ dx1 = - dx2
x1 →0  dx1 dx1 
d ln γ 1 d ln γ 2
According to Gibbs/Duham: x1d ln γ 1 + x2 d ln γ 2 = 0 or 1 x + x 2 =0
dx1 dx1
GE  γ1 
Thus, lim = lim (ln γ 1 − ln γ 2 ) = lim ln  = lnγ1∞
x1 →0 x x RT
1 2
x1 →0 x1 →0 γ2 
GE
Similarly, as x2 → 0 (x1 → 1), lim = ln γ 2∞
x2 →0 x x RT
1 2

Chemical Engineering Thermodynamics II CHE 341 12-42

 The Margules Equations
How it is possible to correlate the measured experimental data by a mathematical
expression for a given γ1 and γ2?
• By inspection of the previous figures, it can be observed that the GE/x1x2RT data
approximates a straight line, i.e,
GE
= A21 x1 + A12 x2
x1 x2 RT
E
A21 and A12 are constants
G
or, = ( A21 x1 + A12 x2 )x1 x2
RT
By applying the basic results,
( E
 ∂ nG / RT  ) and,
(
 ∂ nG E / RT 
ln γ 2 = 
)
ln γ 1 =  ∂n 
∂n   2  P ,T ,n1
 1  P ,T ,n2
one can obtain,
These empirical equations are the
ln γ 1 = [ A12 + 2( A21 − A12 ) x1 ]
x22 Margules equations and are frequently
ln γ 2 = x12 [ A21 + 2( A12 − A21 ) x2 ] used to model and correlate solution
behavior
Chemical Engineering Thermodynamics II CHE 341 12-43
 For infinite dilution
@ x1 = 0 ⇒ lnγ1 = lnγ1∞ = A12
@ x2 = 0 ⇒ lnγ2 = lnγ2∞ = A21
• For methyl ethyl ketone(1)/toluene system(2), it can be found from the previous
figures that,
Intercepts of the straight line joining
A12 = 0.372 and A21 = 0.198
lnγ1∞ and γ2∞
∴ The expression of GE/RT for this specific system can be written as:
GE
= (0.198 x1 + 0.372 x2 )x1 x2
RT
• In this case, we have accomplished a reduction of the VLE data set to a simple
equation which concisely stores all of the information in the data set.
• It will be seen that the correlations for lnγ1, lnγ2 and GE/RT provide good
approximation of the experimental data.

Chemical Engineering Thermodynamics II CHE 341 12-44

 Construction of Correlation for the original P-x1-y1 data
• If A12 and A21 are known, then it is possible to construct correlations for P-x1 and P-y1
yP
Recall the definition of modified Raoult’s law: γ i =
i
sat
x P
i i
⇒ yi P = xiγ i Pi sat = pi
x1γ 1P1sat
Since P = p1 + p2 P = x γ P sat + x γ P sat whence y1 =
1 1 1 2 2 2
x1γ 1P1sat + x2γ 2 P2sat
• The procedure of generating the P-x1 and P-y1 data is as follows:
(i) Determine γ1 and γ2 from the Margules equations for known values
of A12 and A21.
(ii) Determine P1sat and P2sat from tables (or equations).
(iii) Calculate P and y1, for various values of x1, and then plot the results.
⇒ The computed values are shown as solid lines in the preceding figures.
⇒ It is seen that the experimental P-x1-y1 data are adequately correlated to the
Margules equations

Chemical Engineering Thermodynamics II CHE 341 12-45

 Negative Deviation from Raoult’s Law
The chloroform(1)/1,4-dioxane(2) binary system at 50oC exhibit negative deviation
from Raoults law.

• The experimental data are shown in Table 12.3, shown in the next slide.
• The data are also presented graphically in Fig. 12.6, next slide.
• It is observed that γ1 & γ2 ≤ 1 (or lnγ1 & lnγ2 ≤ 0) and GE/RT ≤ 0
• In addition, P-x1 data are all below the Raoult’s law (linear plot of P vs. x1)
⇒ As before, the data GE/x1x2RT more or less approximates a straight line with
A12 = - 0.72 and A21 = - 1.27
• Again, the experimental data fairly well correlated by the various equations, as
shown by the solid lines.

Chemical Engineering Thermodynamics II CHE 341 12-46

Chemical Engineering Thermodynamics II CHE 341 12-47
Chemical Engineering Thermodynamics II CHE 341 12-48
 Consistency of Thermodynamic Data
• From the previous figures, it is apparent that the correlations provided by Margules
equations are good but not perfect. There are two reasons beyond this:
1. Margules equations are not perfectly suited to the data (i.e., although a straight
line fit is convenient, it may not be the best approximation)
2. The experimental data themselves may not systematically in error, such that they
do not conform to the Gibbs-Duham equation:
x1d ln γ 1 + x2 d ln γ 2 = 0 (at constant T and P)

There are certain procedure available for verifying the consistency of thermodynamic
data. In this context, it should be noted that:
a) If data are consistent (thermodynamically obeying Gibbs-Duham equation),
then they may be accurate,
b) If data are not consistent, then they must be inaccurate.
• Following are three well known tests for consistency of thermodynamic data. All of
these tests are based on Gibbs-Duham equation
Chemical Engineering Thermodynamics II CHE 341 12-49
 The Slope Test
From the Gibbs-Duham equation, we can write:
d ln γ 1 x d ln γ 2
=− 2
dx1 x1 dx1
This equation is the basis for the following observations:
1. At every value of x1, the lnγ1 and the lnγ2 curves
should have opposite slopes.
2. At x1 = x2 = 0.5, the slopes should be equal and
opposite. 0 x1 1
3. Each curve has zero slope as xi → 1
4. If there is maximum in one curve, there should be a minimum in the other curve
d ln γ 1 d ln γ 2
If =0 ⇒ =0
dx1 dx1
5. If there is no maximum or a minimum, then lnγ curves should be on the same side
of the lnγ = 0 line (above or below).

Chemical Engineering Thermodynamics II CHE 341 12-50

 The Area Test
Integration of the Gibbs-Duham equation:

x1 =1 x2 =1

∫ x d ln γ + ∫ x d ln γ
x1 = 0
1 1
x2 = 0
2 2 =0
A1
A2

Accordingly, this relation implies that the shaded 0 x1 1

areas in the drawing should be equal, i.e.
A1 = A2

Chemical Engineering Thermodynamics II CHE 341 12-51

 The Redlich-Kister Test
GE
Since = x1 ln γ 1 + x2 ln γ 2
RT
GE 
Then d 
 RT  = x1d ln γ 1 + ln γ 1dx1 + x2 d ln γ 2 + ln γ 2 dx2
 
= 0 (from Gibbs-Duham equation)
E
G  γ 
Since dx1 = dx2 ⇒ d  
 RT  = (ln γ 1 − ln γ 2 )dx1 = ln  1
dx1
  γ2 
x1 =1 x1 =1
E
G   γ1 
Integrating, ⇒
∫
x =0
d 

 =
RT  ∫
x =0
ln dx1
 γ2 
1 1 Equal
• As xi → 0, GE/RT → 0 Pure species areas
x1 =1
A1
ln(γ1/γ2)
γ 
⇒
∫
x =0
1
ln 1 dx1 = 0 (A1+A2 = 0)
γ2 
0 x1
A2
1

Chemical Engineering Thermodynamics II CHE 341 12-52

 Models for GE/RT and lnγi
The Van Laar Equations
' '
GE A21 A12
= ' '
x1 x2 RT A12 x1 + A21 x2
Whence:
−2 ' −2
'  
'
'  A12 x1  A21 x2
ln γ 1 = A12 1 + '
 ln γ 2 = A21 1 + '

 A21 x2   A12 x1 
@ x1 = 0 ⇒ lnγ1 = lnγ1∞ = A’12
For infinite dilution
@ x2 = 0 ⇒ lnγ2 = lnγ2∞ = A’21
These are the Van Laar equations, which are often useful when maximum or minimum
values in lnγ1 and lnγ2 occur.
• It should be stressed that none of the above equations have any theoretical
foundations, although they are still useful when fitting experimental VLE data.

Chemical Engineering Thermodynamics II CHE 341 12-53

Wilson Equations
Like the Margules and van Laar equations, the Wilson equation containsjust two
parameters for a binary system (Λ12 and Λ21). It is written:

For infinite dilution, the equations become:

Note that Λ12 and Λ21 must always be

positive numbers.

Chemical Engineering Thermodynamics II CHE 341 12-54

 Other Equations
Many other (more sophisticated) equations exist in the literature. These are based on
molecular thermodynamics which deals with intermolecular forces of particles in
solution. They are termed local-composition models, and are beyond the cope of this
course.
Examples of these models:

• NRTL (Non-Random-Two-Liquid) equations (see pp. 448-449).

• UNIQUAC (UNIversal QUAsi-Chemical) equations (see p. 449, Appendix H)
• UNIFAC equations (see p. 449, Appendix H)

Chemical Engineering Thermodynamics II CHE 341 12-55