Chemical Engineering Thermodynamics

Solu%on  Thermodynamics:    
                                 Theory  and  Deriva%on  

Mohammad Fadil Abdul Wahab

Single  component  system  !  

   Example:  Steam  Line  

We know that for a single component system, the most

common thermodynamic variables are……

T and P
By now, we know how to determine Tsat, Psat,
H (liquid and ideal/non-ideal gas),
V (liquid and ideal/non-ideal gas) and ∆Hvap.
 Ok….we  also  learned  about  
mixture  before!!!  

H = ∑ yi H i etc.

………we  will  see  that  this  equa%on  is  only  valid  

for  ideal  mixture!!!  
Consider the following systems,

A B
Both systems at the sameT, P and n but with
different compositions. Assuming an ideal mixture,

(H )A = ∑ yiA H i (H )B = ∑ yiB H i
yiA ≠ yiB
so
(H )A ≠ (H )B
So  we  have  new  thermodynamic  variables    
for  
mul%component  system………….  

Composition!! !

(x1, x2, x3) or (y1, y2, y3)!

BUT  typical  chemical  processes  
involve  
a  mul%component  system  
whereby  we  can’t  assume  ideal  
mixture  or  ideal  solu%on  
behavior  
Consider  mixing  1000  cm3  methanol  and  1000  cm3  
water  at  25oC.  

+ =

1000 cm3 MeOH + 1000 cm3 H2O !

! ! ! !≠ 2000 cm3 solution !!!!!!
In! fact,!
! ! !!
1000 cm3 MeOH + 1000 cm3 H2O !
! ! ! != 1970 cm3 solution!
Discuss  with  a  person  next  to  you  why  
the  volume  of  the  solu%on  is  less  than  2000  cm3?  
 Let’s  determine  the  density    
of  the  solu%on  
Assume ideal 1 1
solution ρ= =
V ∑ x iVi
From  above  eqn,  the  calculated  solu%on  molar  
density  is  0.0396  mol/cm3  

The  actual  molar  density  is  0.0410  mol/cm3  that    

could  be  calculated  from  
1 1
ρ= =
V ∑ x iVi Partial Properties
We  just  introduced  a  new  thermodynamic  
property  for  a  mul%component  system……  

Partial properties!! !
11.1  Fundamental  Property  Rela%on    
             (for  variable  composi%on  mixture)    

Defini%on  of  Chemical  Poten%al  

Consider the following system,

- Open system and multicomponent

- Single phase at T and P
- Variable composition (n1, n2, n3 etc)

So Gibbs energy (G) could be written as a function

of T, P, n1, n2, n3 etc.

(nG) = g ( P,T , n1 , n2 ....ni )

 Apply partial differentiation and FPR (Chapter 6),

( )
d nG = ⎢
( ) ⎤⎥
⎡ ∂ nG ⎡ ∂ nG
dP + ⎢
( ) ⎤⎥ ( ) ⎤⎥
⎡ ∂ nG
dT + ∑ ⎢ dni
⎢⎣ ∂ P ⎥⎦T ,n ⎢⎣ ∂T ⎥⎦ P,n i ⎢ ∂ ni
⎣ ⎥⎦ P,T ,n
j

( )
⎡ ∂ nG
d nG = nVdP − nSdT + ∑ ⎢
( ) ⎤⎥ dni
⎣ ∂ ni
i ⎢ ⎥⎦ P,T ,n
j

Let's define µi = ⎢
( ) ⎤⎥
⎡ ∂ nG
≡ Chemical potential for species i in the mixture
⎢⎣ ∂ ni ⎥⎦ P,T ,n
j

This equation forms the basis for the definition of partial properties.

So,

( )
d nG = nVdP − nSdT + ∑ µi dni (11.2)
i
How  do  we  determine  the  phase  of  
a  mul%component  system?  

If it is two phases, how do we determine

the composition of each phase?

Before answering these, we need to establish

the concept of phase equilibria for
multicomponent system.
11.2  The  Chemical  Poten%al  and  Phase  Equilibria  
nα

Consider the following system, nβ

- Closed system and multicomponent

- Two phases in equilibrium.
- Mass transfer occurs if the equilibrium is
disturbed.

Each phase (α and β) is actually an open system,

so eqn 11.2 becomes,
d ( nG) = ( nV ) dP − ( nS ) dT + ∑ µαi dnαi
α α α

d ( nG) = ( nV ) dP − ( nS ) dT + ∑ µ iβ dn iβ
β β β

i
Add the two equations for a VL system
(α is replaced by v and β is replaced by l).
d( nG ) + d( nG ) = ( nV ) dP + ( nV ) dP − ( nS ) dT − ( nS ) dT + ∑ µiv dniv + ∑ µil dnil
v l v l v l

i i

d( nG ) = ( nV )dP − ( nS )dT + ∑ µiv dniv + ∑ µil dnil

For system in equilibrium, i i

( nV ) dP − ( nS ) dT = ( nV ) dP − ( nS ) dT + ∑ µ dn i
v
i
v
i
+ ∑ µil d nil
i

So, ∑ µ dn v
i
v
i
+ ∑ µil d nil = 0 also dnil = −dniv
i i

∑ µ dn
i
v
i
v
i
i i
(
− ∑ µil dniv = ∑ µiv − µil dniv = 0 )
µiv − µil =0 µiv = µil
If apply to system with π phases (i.e. more than two phases)

µiv = µil =µiα = µiβ =...........=µiπ (11.6)

This is the criteria for phase equilibrium for multicomponent system.
 Criteria for Phase Equilibrium for
Multicomponent System

Multiple phase at the same

T and P are in equilibrium
when the chemical potential (µi) of
each species is the same in
all phases

nα µiv
µiv = µil
nβ µil
11.3    PARTIAL  PROPERTIES  
Based on the definition of µi , let’s define a partial molar
property as follows,

Mi = ⎢
( ) ⎤⎥
⎡ ∂ nM
(11.7)
⎢⎣ ∂ ni ⎥⎦ P,T ,n
j

Where M ≡ V ,U , H , S,G etc.

This is a partial property of individual species as they exist

in a solution or mixture.

This property is different from a pure species property

primarily due to molecular interaction of different species in
the solution or mixture
Hence, the chemical potential of species i (μi) is known also
as partial property of Gibbs energy for species i.

µi = ⎢
( ) ⎤⎥
⎡ ∂ nG
= Gi (11.8)
⎢⎣ ∂ ni ⎥⎦ P,T ,n
j
 Example  11.1  
This  example  explains  the  physical  
and  mathema%cal  interpreta%on  of    
par%al  molar  volume  equa%on.  
Let’s familiar with symbols in solution
thermodynamics,

Symbol Type of property Example

Mi Pure-species properties Ui H i Si Gi
Mi Partial properties Ui Hi Si Gi
M Solution or mixture properties UH S G
 In general, for a homogenous mixture,
nM = M (T , P, n1 , n2 , n3 .....)

( )
⎡ ∂ nM ⎤
d nM = ⎢
( ) ⎡ ∂ nM ⎤
⎥ dP + ⎢
( )
⎡ ∂ nM
⎥ dT + ∑ ⎢
( ) ⎤⎥ dni
⎢⎣ ∂ P ⎥⎦T ,n ⎢⎣ ∂T ⎥⎦ P,n ⎣ ∂ ni
i ⎢ ⎥⎦ P,T ,n
j

Could also be written as,

⎡ n∂ M ⎤ ⎡ n∂ M ⎤
( )
d nM = ⎢ ⎥ dP + ⎢ ⎥ dT + ∑ M i dni (11.9)
⎣ ∂ P ⎦T ,x ⎣ ∂T ⎦ P,x i

i.e. constant composition

where d(nM ) = ndM + Mdn

ni
Also, xi = ni = xi n so, dni = xi dn + ndxi
n

Substitute into (11.9)

⎡ n∂ M ⎤ ⎡ n∂ M ⎤
ndM + Mdn = ⎢ ⎥ dP + ⎢ ⎥ dT + ∑ M i (xi dn + ndxi ni )
⎣ ∂ P ⎦T ,x ⎣ ∂T ⎦ P,x i
 Rearrange,
⎡ ⎛ ∂M ⎞ ⎛ ∂M ⎞ ⎤ ⎡ ⎤
⎢ dM − ⎜⎝ ∂ P ⎟⎠ dP − ⎜⎝ ∂T ⎟⎠ dT − ∑ i i ⎥ ⎢ ∑ i i ⎥ dn = 0
M dx n + M − M x
⎢⎣ T ,x P,x i ⎥⎦ ⎣ i ⎦

This is equal to 0 if both terms in the brackets are

zeros, so for the 1st bracket,
⎛ ∂M ⎞ ⎛ ∂M ⎞
dM = ⎜ dP + ⎜ dT + ∑ M i dxi (11.10)
⎝ ∂ P ⎟⎠ T ,x
⎝ ∂T ⎟⎠ P,x i

And the 2nd bracket,

M = ∑ M i xi
i

note: (11.10) similar to (11.9) for n = 1

 M = ∑ xi M i (11.11)
i

or

nM=∑ ni M i (11.12)
i

Eqns 11.11 and 11.12 show that the calculation of

mixture properties of constituent species from
partial properties is by a simple summability
relation!
Eqn 11.11 can be further expanded to give,
M = ∑ xi M i → dM = ∑ xi dM i +∑ M i dxi
i i i

Substitute into 11.10, ,

⎛ ∂M ⎞ ⎛ ∂M ⎞
∑ i i ∑ i i ⎜⎝ ∂ P ⎟⎠
x dM + M dx = dP + ⎜⎝ ∂T ⎟⎠ dT + ∑ M i dxi
i i T ,x P,x i

⎛ ∂M ⎞ ⎛ ∂M ⎞
⎜⎝ ∂ P ⎟⎠ dP + ⎜ ⎟ dT = ∑ xi dM i
T ,x
⎝ ∂T ⎠ P,x i

⎛ ∂M ⎞ ⎛ ∂M ⎞
⎜⎝ ∂ P ⎟⎠ dP + ⎜
⎝ ∂T ⎟⎠ dT − ∑ xi dM i = 0 (11.13)
T ,x P,x i

Eqn 11.13 is a Gibbs/Duhem Eqn that must be

satisfied for any a changes in P,T and Mi in a
homogeneous phase.

This eqn shows that the partial properties of species

making up a solution are dependent on one another.
 Par%al  Proper%es  in  Binary  Solu%ons  
 (two-­‐component  system:  i  =1,  2)  
Here we will derive an eqn for partial property of a species
in binary mixtures as in term of solution property (express
a function of composition at constant T and P).
M = x1M1+x2M 2
From eqn 11.11
dM = x1dM1 +M1dx1 + x2 dM 2 +M 2 dx2
dM =x1dM1 +M1dx1 + x2 dM 2 -M 2 dx1
From eqn 11.13 at constant T and P,
⎛ ∂M ⎞ ⎛ ∂M ⎞
⎜⎝ ∂ P ⎟⎠ dP + ⎜ ⎟ dT − ∑ xi dM i = 0 (11.13) x1dM1 +x2 dM 2 =0
T ,x
⎝ ∂T ⎠ P,x i

dM
So, dM = M1dx1 -M 2dx1 = M1 -M 2
dx1
Hence, we could show that,
dM dM
M1 = M + x2 (11.15) M 2 = M − x1 (11.16)
dx1 dx1
Example    11.4  shows  the  applica%on  of  
eqn  11.15  and  11.16  
 Parallel  Rela%on  Between  
Eqn  for  Constant  Composi%on  and  
Par%al  Proper%es  

Eqn for Constant Composition! Eqn for Partial Properties!

Similarly we can write
H = U + PV → H i = U i + PV i
dG = VdP − SdT → dG i = V i dP − S i dT
Ideal-­‐Gas  Mixture  
 Molar  Volume  of  Ideal-­‐Gas  Mixture  

Consider,  
System  A  contain  species  1  at  T,P  
System  B  contain  species  2  at  T,P  
System  C  contain  mixture  1  and  2  at  T,P  

(V ) ig
A
=
RT
P
= V1ig

(V ) ig
B
=
RT
P
= V2ig

(V ) ig
C
=
RT
P
= V ig
 Par%al  Molar  Volume  of  Species  i  
in  an  Ideal-­‐Gas  Mixture  

Applica%on  of  par%al  proper%es  to    

molar  volume,  
⎡ RT ⎤
ig ⎡ ∂(nV ig ) ⎤ ⎢ ∂(n )⎥
RT ⎛ ∂n ⎞ RT
Vi =⎢ = ⎢ P ⎥ = =
⎥ ⎜ ⎟
⎣ ∂ni ⎦T , P,n j ⎢ ∂ni ⎥ P ⎝ ∂ni ⎠ n P
⎢⎣ ⎥⎦T , P,n
j

The last equality infers an independent of composition!!!

So  for  Ideal-­‐Gas  Mixture  

ig
V (T , P) = Vi (T , P) = V i (T , P)
ig ig
Par%al  Pressure  of  Species  i  in  an  
Ideal-­‐Gas  Mixture  
The  pressure  that  a  species  i  would  exert  if  it  
alone  occupied  the  molar  volume  of  the  mixture  
         PiVT,ig = niRT

Divide by ideal gas for mixture, PVT,ig = nTRT ,

P n
i = i = y
So,   P n i
T
yi RT
pi = yi P = ig
V
 Par%al  Molar  Proper%es  of    
Species  i  in  an  Ideal-­‐Gas  Mixture  

Gibb’s  Theorem,  
A  par%al  proper%es  (other  than  par%al  molar  volume)  of  a  
cons%tuent  species  in  an  ideal-­‐gas  mixture  is  equal  to  the  
corresponding  molar  property  of  the  species  as  a  pure  
ideal  gas  at  the  mixture  temperature  but  at  a  pressure  
equal  to  its  par%al  pressure  in  the  mixture.  

ig
M i (T , P) = M iig (T , pi )
Apply  this  theory  to  Enthalpy,  Internal  Energy  and  
Heat  Capacity  

ig
H i (T , P) = H iig (T , pi ) = H iig (T , P)
ig
H i = H iig

Since  Hig  is  independent  of  pressure  

Similarly,
ig ig ig Uig,  Cpig  and  Cvig    are  also    
Ui = U i
ig
C pi = C ig
p,i
C vi = C ig
v,i independent  of  pressure  

So for ideal-gas mixture,

H ig == ∑ yi H iig U ig == ∑ yiU iig C Pig == ∑ yiC P,i
ig
etc
 Property  Change  of  Mixing    
for  ideal  gas  mixture    (ΔHig  and  ΔUig)  

ΔM mix
ig
= M final − M initial = M ig − ∑ yi M iig

but M ig = ∑ yi M iig
+ =
so
Minitial Mfinal
ΔM mix
ig
= 0 for M = H ,U
Note: Both initial and final conditions (T,P) are the same.
For example,
Enthalpy change of mixing for ideal gas mixture,
ΔH mix
ig
=0
 Property  Change  of  Mixing    
for  ideal  gas  mixture    (ΔVig)  

ΔVmixig = V final − Vinitial = V ig − ∑ yiVi ig = V ig − Vi ig ∑ yi

note : Vig = Vi ig = Vjig

so
ΔVmixig = 0

Note: Both initial and final conditions (T,P) are the same.
 Par%al  Molar  Entropy  of    
Species  i  in  an  Ideal-­‐Gas  Mixture  
dT dP
Sig  is  dependent  of  pressure,   dS iig = C Piig −R (6.24)
T P

dP
For isothermal mixing,! dS igi = −R
P

Integrate from pi to P,! P 1

S iig (T , P ) − S iig (T , pi ) = −R ln =-Rln
pi yi
Siig (T , pi ) = S iig (T , P ) − Rlnyi

From Gibbs Theorem,

(T , P ) = Siig (T , pi ) = S ig (T , P ) − Rlnyi
ig
S i i
 Property  Change  of  Mixing    
for  ideal  gas  mixture    ΔSig  
Hence, for ideal-gas mixture,!
( ) ( ( )
S ig (T , P) = ∑ yi Siig T , pi = ∑ yi S iig T , P − Rlnyi
k
)
S ig (T , P) = ∑ yi S iig -R ∑ yi lnyi (11.26)

So entropy change of mixing for ideal gas mixture,

ΔS mix
ig
= S ig (T , P) − ∑ yi S iig

= − R∑ yi ln yi
1
= R∑ yi ln
yi
 Par%al  Molar  Gibbs  Energy  of    
Species  i  in  an  Ideal-­‐Gas  Mixture  
By  parallel  rela%on,   G ig = H ig − TS ig
ig ig ig
G = H −TS
i i i
So,

ig ig
G = H − T (Siig − R ln yi ) = H iig − T (Siig − R ln yi ) = Giig + RT ln yi
i i

So for ideal-gas mixture,

ig
G = ∑ yi G
ig
i

G ig = ∑ yiGiig + RT ∑ yi ln yi
 Property Change of Mixing 

for ideal gas mixture (ΔGig)!

ΔG mix
ig
= G ig − ∑ yiGiig

= RT ∑ yi ln yi
In  terms  of  variables  (T,P)  for  Gibbs  Energy  ,  

RT
dGiig = Vi ig dP − Siig dT = dP = RTd ln P (for constant T)
P
Integrate
Giig = Γ i (T ) + RT ln P (11.28)
Integration constant at constant T

so
ig
G i = Giig + RT ln yi = Γ i (T ) + RT ln P + RT ln yi
ig
G i = Γ i (T ) + RT ln yi P (11.29)

Hence for ideal-gas mixture,

ig
G = ∑ yi G i = ∑ yi Γ i (T ) + RT ∑ yi ln( yi P)
ig
(11.30)
Summary  of  Property  Change  of  Mixing  for  Ideal-­‐
Gas  Mixture  

ΔM mix
ig
= M final − M initial = M ig − ∑ yi M iig

ΔH mix
ig
=0
ΔU mix
ig
=0
ΔVmixig = 0
1
ΔS mix
ig
= R∑ yi ln
yi
ΔG mix
ig
= RT ∑ yi ln yi
Fugacity  and  Fugacity  Coefficient:  

Pure  Species  
 Fugacity  of  Pure  Species  
The  concept  of  fugacity  is  introduced  when  we  want  to  
apply  an  eqn  similar  to  11.28  (ideal  gas)  to  a  real    
fluid  (liquid  or  gas).    
Gi = Γ i (T ) + RT ln P
ig

This  is  done  by  replacing  P  with  a  property  specific  to  a  

par%cular  species  called  fugacity.  

Gi = Γ i (T ) + RT ln f i (11.31)

f i ≡ fugacity of pure species i (with unit of pressure)

i.e. f i is a pseudopressure.

note : for a gas under an ideal gas condition, f i = P

Residual  Gibb  Energy  and  Fugacity  Coefficient  
of  Pure  Species  

Eqn (11.31) − Eqn (11.28),

fi
Gi − G = RT ln f i − RT ln P = RT ln
i
ig

P
fi
GiR
= RT ln (11.33)
P

let's us defined φi as a fugacity coefficient for species i as follows,

fi
φi = , a dimensionless quantity
P

So for ideal gas, φi is 1

Determina%on  of  Fugacity  
Coefficient   φ i
GiR = RT ln φi (11.33)

or
GR ( Z − 1)dP
G R
RT
= ∫ P
ln φi = i
RT

GiR P dP
Combine with Eqn 6.49 ,
RT
= ∫ 0
(Z i − 1)
P

P dP
ln φi = ∫ 0
(Z i − 1)
P
(11.35)
 Fugacity  Coefficient  of  
Pure  Species  from  Virial  Eqn  
BP
The  2-­‐term  Virial  Equa%on  for  Zi  :   Zi − 1 =
RT

P dP
ln φi = ∫ (Z i − 1)
0 P
Substitute,
P ⎛ B P ⎞ dP B P BP
ln φi = ∫ 0 ⎜ RT ⎟ P = RT
⎝ ⎠
∫ 0
dP =
RT
⎡B P⎤
φi = exp ⎢ ⎥
⎣ RT ⎦
Note : B = Bii
Alterna%vely  for  the  2-­‐term  Virial  Equa%on  for  Zi  :  

B P BPc Pr Pr Pr 1 Pr
Zi − 1 = = = B̂ = (B + ω B )
0

RT RTc Tr Tr Tr Tr
P dP Pr dPr
ln φi = ∫ 0
(Z i − 1)
P
= ∫ 0
(Z i − 1)
Pr
Pr (B0 + ω B1 ) 1 Pr
ln φi = ∫ dPr = (B + ω B )
0
0 Tr Tr

⎡ Pr 0 ⎤
φi = exp ⎢ (B + ω B ) ⎥1

⎣ Tr ⎦
for B0 and B1 , see eqn 3.65 and 3.66
0.422 0.172
B = 0.083 − 1.6
0
and B = 0.139 − 4.2
1

Tr Tr
 Fugacity  Coefficient  of  
Pure  Species  from  Generic  Cubic  EOS  

Combined  eqn  11.33  with  6.66b  (and  apply  for  species  i  ),  
GiR = RT ln φi (11.33)
GiR
= Z i − 1 − ln(Z i − β i ) − qi I i (6.66b)
RT
We get,!

ln φi = Z i − 1 − ln(Z i − β i ) − qi I i (11.37)
 Fugacity  Coefficient  of  
Pure  Species  from  Generalized  Correla%on  
Lee/Kesler  correla%on,  
P dP
ln φi = ∫ (Zi − 1)
0 P
for P = Pc Pr dP = Pc dPr
Pr Pc dPr Pr dPr Pr dPr
ln φi = ∫ (Z − 1) = ∫ (Z − 1) = ∫ (Z + ω Z − 1)
0 1
0 Pc Pr 0 Pr 0 Pr

Pr dPr Pr
1 dPr
ln φi = ∫ (Z − 1) +ω∫ Z
0
0 Pr 0 Pr

ln φi = ln φi0 + ω ln φi1
ln φi = ln φi0 + ω ln φi1

φi = φi0 (φi1)ω

Values for φi0 and φi1 are found from Table E13-E16
 Vapor/Liquid  Equilibrium  for    
Pure  Species  
From  chapter  6  we  know  that  criteria  for  VLE  for    
pure  species  is    Gv=Gl        or      Gv-­‐Gl  =0  

Let’s  apply  eqn  11.31  for  each  phase.   Giv = Γ i (T ) + RT ln f i v

Gil = Γ i (T ) + RT ln f i l
By  difference,  
fi v
Giv − Gil = RT ln l
=0
fi
fi v fi v
ln =0 =1 fi v = fi l = fi sat
fi l fi l Another phase!
Also equilibrium criteria!
fi sat for pure species!!
φisat = φiv = φil = φisat
Pi sat
 Fugacity  of  Pure  Liquid  
Only  for  the  fugacity  of  pure  liquid  at  P  and  T.  Easily  calculated    
by  the  Poyn%ng  factor  equa%on  as  derived  below,  

fi l (Pi sat ) fi v (Pi sat ) fi l (P) fi l (P) fi l (P)

fi (P) =
l
v sat sat l sat
Pi sat
= (1)(φ )
sat
i l sat
Pi sat
= φ Pi
sat
i
sat
l sat
(A)
fi (Pi ) Pi fi (Pi ) fi (Pi ) fi (Pi )

Also
fi l
Gil − Gil,sat = RT ln fi l − RT ln fi l,sat = RT ln l,sat
(B)
fi

also at constant T,

dGi = Vi dP and we integrate from sat liquid to compressed liquid,

Gil − Gil,sat = ∫ V dP
l
i
(C)
Pisat
Combine (B) and (C),

P P
fi l fi l 1
∫ Vi l dP ∫
l
RT ln l,sat
= rearrange, =exp Vi
dP
fi Pi sat fi l,sat RT P sat
i

subs into (A),

P l P
1 V
fi l = φisat Pi sat exp ∫
RT P sat
Vi l dP = φisat Pi sat exp i ∫ dP
RT P sat
i i

note :Vi l is approximately constant and is evaluated at sat liquid, therefore:

Vi l (P − Pi sat )
fi l = φisat Pi sat exp (11.44) fugacity of pure liquid i
RT

The exponential is known as a Poynting factor

 Fugacity    
for  Species  i  in  Solu%on   ( fˆi )
 Revisit, for an ideal gas mixture,
ig
µ = G i = Γ i (T ) + RT ln( yi P)
ig
i
(11.29)

Parallel  to  the  defini%on  of  fugacity  for  pure  species,    

the  introduc%on  of  fugacity  for  solu%on  is  done  by  
replacing  yiP    in  Eqn  11.29  with  fugacity  for  species  in  solu%on   ( fˆi )

So,  for  a  real  mixture  or  solu%on,  

µi = Gi = Γ i (T ) + RT ln fˆi (11.46)

fˆi ≡ fugacity of species i in mixture or solution.

at VLE

µil = µiv

Γ i (T ) + RT ln fˆi l = Γ i (T ) + RT ln fˆi v

So

fˆi l = fˆi v Criteria of VLE for multicomponent system !!!

Also, nGR = nG − nG ig

(
⎡ ∂ nGR
⎢
) ⎤⎥ ( ) ⎤⎥
⎡ ∂ nG
= ⎢
(
⎡ ∂ nG ig
−⎢
) ⎤⎥
⎢ ∂ ni ⎥ ⎢⎣ ∂ ni ⎥⎦P,T ,n ⎢ ∂ ni ⎥
⎣ ⎦P,T ,n j j
⎣ ⎦P,T ,n j

GiR = Gi − Giig
= µi − µiig
= RTlnf̂i − RT ln y i P
f̂i
= RTln
y iP
= RT ln φ̂i (11.51)
 fˆi
where φ̂i =
y iP

fˆi
φ̂i = fugacity coefficient* of species i in solution or mixture.
y iP

fˆi = φ̂i y i P

*mostly applied to gas mixture

Calcula%on  of  Fugacity  Coefficient  
of  Species  i  in  Solu%on   (φ̂i )
 Applica%on  of  Fundamental    
Residual-­‐Property  Rela%on….….  

⎛ nG ⎞ 1 nG
d⎜ ⎟ = d(nG) − dT
⎝ RT ⎠ RT RT 2

⎛ nG ⎞ 1 ⎛ ⎞ n(H − TS)
d⎜ ⎟ =
⎝ RT ⎠ RT ⎝ ⎜ (nV )dP − (nS)dT + ∑ Gi
dni⎟
⎠
−
RT 2
dT

⎛ nG ⎞ nVdP nS Gi dni nH nS
d⎜ ⎟ = − dT + ∑ − dT + dT
⎝ RT ⎠ RT RT RT RT 2
RT

⎛ nG ⎞ nVdP nH Gi
d⎜ ⎟
⎝ RT ⎠
=
RT
−
RT 2
dT + ∑ RT
dni (11.54)

for ideal gas,

⎛ nG ig ⎞ nV ig dP nH ig G igi
d⎜ ⎟ = − 2
dT + ∑ dni (11.54ig)
⎝ RT ⎠ RT RT RT
(11.54 - 11.54ig)

⎛ nGR ⎞ nV R dP nH R GRi
d⎜ ⎟ = − 2
dT + ∑ dni (11.55)
⎝ RT ⎠ RT RT RT

substitute (11.51),

⎛ nGR ⎞ nV R nH R
d⎜ ⎟ = dP − 2
dT + ∑ ln φ̂i dni (11.56)
⎝ RT ⎠ RT RT
So,
⎡ ∂(nGR )⎤
ln φ̂i = ⎢⎢ RT ⎥
⎥ (11.59)
∂ni
⎢⎣ ⎥⎦P,T ,n
j

Note:
ln φ̂i is a partial property of GR
RT
 GR ( Z − 1)dP
RT
= ∫ P
Combine with eqn 6.49 and 11.59 becomes,

⎡ ∂(nZ - n) ⎤
P
dP
ln φ̂i = ∫ ⎢ ⎥
0 ⎣
∂ni ⎦P,T ,n j P

⎡ ∂nZ ∂n ⎤
P
dP
= ∫⎢ − ⎥
0 ⎣
∂ni
∂ni ⎦ P,T ,n
P
j
1
⎡ ∂nZ ⎤
P
⎡ ∂n ⎤ dP
= ∫ (⎢ ⎥ −⎢ ⎥ )
0 ⎣
∂ni ⎦P,T ,n ⎣ ∂ni ⎦P,T ,n P
j j

∫( ) dP
ln φ̂i = Zi − 1 (11.60)
0
P
 Using  Two-­‐term  Virial  EOS    
BP nBP
Z = 1+ nZ = n +
RT RT

⎡ ∂(nZ) ⎤ P ⎡ ∂(nB) ⎤
Z1 = ⎢ ⎥ = 1+ ⎢ ⎥
⎣ ∂n1 ⎦P,T ,n2 RT ⎣ ∂n1 ⎦T ,n2

Substitute into eqn 11.60

P ⎛ P ⎡ ∂(nB) ⎤ ⎞ dP 1
P ⎛ ⎡ ∂(nB) ⎤ ⎞
ln φ̂1 = ∫ ⎜ 1+ ⎢ ⎥ − 1⎟ = ∫ ⎜⎜ ⎢ ∂n ⎥ ⎟⎟dP
0 ⎜
⎝ RT ⎣ ∂n1 ⎦T ,n ⎟⎠ P RT 0 ⎝⎣ 1 ⎦T ,n ⎠
2 2

1 ⎡ ∂(nB) ⎤ P ⎡ ∂(nB) ⎤
ln φ̂1 = ⎢ ⎥ P = ⎢ ⎥
RT ⎣ ∂n1 ⎦T ,n RT ⎣ ∂n1 ⎦T ,n
2 2

How to evaluate?
For mixture,

B = ∑ ∑ y i y j Bij
i j

Consider a binary system (species 1 and 2),

B = y1y1B11 + y1y 2B12 + y 2 y1B21 + y 2 y 2B22

B = y12B11 + 2y1y 2B12 + y 22B22

B = y1(1− y 2 )B11 + 2y1y 2B12 + y 2 (1− y1)B22

Rearrange further,

B = y1B11 − y1y 2B11 + 2y1y 2B12 + y 2B22 − y1y 2B22

= y1B11 + y 2B22 + 2y1y 2B12 − y1y 2B11 − y1y 2B22
= y1B11 + y 2B22 + y1y 2 (2B12 − B11 − B22 )
= y1B11 + y 2B22 + y1y 2δ12

where δ12 = 2B12 − B11 − B22

Multiply by n (note: yi = ni / n),

n2
nB = n1B11 + n2B22 + n1 δ12
n
 n2
nB = n1B11 + n2B22 + n1 δ12
n
⎡ ∂(nB) ⎤ ⎛ ∂n⎞
n1
u v ∂u − u∂v
⎢ ⎥ = B11 + 0 + n2δ12 ⎜ ⎟ note: ∂ =
⎣ ∂n1 ⎦T ,n2 ⎜⎝ ∂n1 ⎟⎠ v v2
⎛ n∂n1 − n1dn ⎞
⎜ n 2 ⎟ ⎛ 1 n1∂n ⎞
= B11 + n2 ⎜ ⎟ δ12 = B11 + n2 ⎜ − 2 ⎟ δ12
⎜ ∂n1 ⎟ ⎝ n n ∂n1 ⎠
⎜⎝ ⎟⎠

⎛ n2 n2n1∂n ⎞ ⎛ ∂n ⎞
= B11 + ⎜ − 2 ⎟ δ12 = B11 + ⎜ y 2 − y 2 y1 ⎟ δ12
⎝ n n ∂n1 ⎠ ⎝ ∂n1 ⎠
⎡ ∂(nB) ⎤
⎢ ⎥ = B + (1− y )y δ = B + y 2
δ
⎣ ∂n1 ⎦T ,n2
11 1 2 12 11 2 12
note,

B11 = B1 B22 = B2

B12 =
RTc12
Pc12
( B 0 + ω 12B1 ) Interaction parameter
(=0 if no data provided)

ω1 + ω 2
ω 12 = Tc12 = (Tc1Tc 2 )1/ 2 (1− k12 )
2
3
Zc12RTc12 ⎛V 1/ 3
+V 1/ 3
⎞ Zc1 + Zc 2
Pc12 = Vc12 =⎜ c1 c2
⎟ Zc12 =
Vc12 ⎝ 2 ⎠ 2
So,

P ⎡ ∂(nB) ⎤
ln φ̂1 = ⎢ ⎥
RT ⎣ ∂n1 ⎦T ,n
=
P
RT
(
B11 + y 22δ12 )
2

P
ln φ̂1 = (B11 + y 22δ12 )
RT

similarly

P
ln φ̂2 = (B22 + y12δ12 )
RT
In general, for multicomponent gas mixture,

P ⎛ ⎞
ln φ̂k =
RT ⎜⎝ Bkk +
1
2 ∑ ∑ y i y l (2δ ik − δ il )⎟⎠
i l

where,

δ ik = 2Bik − Bii − Bkk and δ il = 2Bil − Bii − Bll

δ ii = δ ll = 0 and δ ik = δ ki

RTcij ωi + ω j
Bij =
Pcij
( B + ω ij B
0 1
) ω ij =
2
Tcij = (TciTcj )1/ 2 (1− k ij )

3
Zcij RTcij ⎛V 1/ 3
ci
+V 1/ 3
cj
⎞ Zci + Zcj
Pcij = Vcij = ⎜ ⎟ Zcij =
Vcij ⎝ 2 ⎠ 2
 Using  Generic  Cubic  EOS  

bi
ln φ̂i =
b
( )
Z − 1 − ln(Z − β ) − qi I

Refer to chapter 14 if you interested in the details.

(Out of scope for this class)

Ideal-­‐Solu%on    
 ig
We  know  for  an  ideal-­‐gas  mixture,   µ = G = Giig (T , P) + RT ln yi
ig
i i

No intermolecular forces/interaction!
and negligible particle volume (compare to molar volume).!
Only for an ideal gas mixture.!

For an ideal solution, we define! µ iid = Giid = Gi (T,P) + RT ln x i

There exist intermolecular interaction but the various molecules!
have similar size, structure and intermolecular forces.!
Applicable for real gas mixture and liquid solution. !
However, application is most often to liquid solution.!

So for an ideal solution,! G id = ∑xG i i

id
= ∑ x iGi + RT ∑ x i ln x i
Entropy  for  Ideal  Solu%on  
Substitute µi = Gi into Eqn 11.3,
dG=VdP-SdT + ∑ Gi dxi
i

Apply criterion of exactness,

⎡ ∂ (nS) ⎤ ⎛ ∂ Gi ⎞
Si = ⎢ ⎥ = −⎜ ⎟⎠
⎣ ∂ xi ⎦ P,T
j ,x
⎝ ∂ T P, x

For ideal solution,

⎛ ∂ G id
⎞ ⎛ ∂ (Gi (T , P) + RT ln xi ) ⎞
Si = − ⎜
id i
⎟ = −⎜ ⎟⎠
⎝ ∂T ⎠ P, x ⎝ ∂T P, x

⎛ ∂ (Gi (T , P) ⎞ ⎛ R ln xi ∂ T ⎞
S = −⎜
id
⎟ −⎜ ⎟ = Si − R ln xi
⎝ ∂T ⎠ P, x ⎝ ∂T ⎠ P, x
i

S id = ∑ xi (Si − R ln xi ) = ∑ xi Si − R∑ xi ln xi
 Enthalpy  for  Ideal  Solu%on  

H iid = Giid + TSiid

H iid = Gi + RT ln xi + T (Si − R ln xi )
= Gi + TSi
= Hi

So for ideal solution,

H =∑ x i H i
 Molar  Volume  for  Ideal  Solu%on  
dG=VdP-SdT + ∑ Gi dxi
i

Apply criterion of exactness,

⎛ ∂ Gi ⎞ ⎡ ∂V ⎤
⎜⎝ ∂ P ⎟⎠ = ⎢ ∂ x ⎥ = Vi
T ,x ⎣ i ⎦ P,T
j ,x

For ideal solution,

⎛ ∂ Giid ⎞ ⎛ ∂ (Gi (T , P) + RT ln xi ) ⎞
Vi id =⎜ ⎟ = −⎜ ⎟⎠
⎝ ∂P ⎠ T ,x ⎝ ∂P T ,x

⎡ ∂ Gi (T , P) ⎤
Vi id
=⎢ ⎥ = Vi
⎣ ∂P ⎦T , x
V id = ∑ xiVi
Non-­‐Ideal  Solu%on  
 Excess  Property   (M E )

In  previous  discussion,  the  real  solu%on  is  evaluated  with  

reference  to  ideal  gas  mixture  by  using  the  defini%on  of  
residual  property.    Hence  the  following  equa%on  is  generally  
and  easily  apply  for  a  gas  mixture.  

fˆi
φ̂i =
y iP

Real liquid solution is, however, much more convenient to

be evaluated with reference to its departure from an
ideal solution (not ideal-gas) behavior.!
Hence,  we  define  an  excess  property  as,  

       The  difference  between  actual  property  of  a  solu%on  and  the  

property  of  an  ideal  solu%on  at  the  same  T,  P  and  xi  

M E = M − M id

So,
M = M E + M id

for example: H = H E + H id

Now, let’s apply this definition to Gibb’s Energy.

G E = G − G id
nG E = nG − nG id

⎢
(
⎡ ∂ nG E ) ⎤
⎥ =⎢
( )
⎡ ∂ nG ⎤
⎥
(
⎡ ∂ nG id
−⎢
) ⎤⎥
⎢ ∂ni ⎥ ∂ni ⎥⎦ P,T ,n ⎢⎣ ∂ni ⎥
⎣ ⎦ P,T ,n j ⎢⎣ j
⎦ P,T ,n j

GiE = Gi − Giid

from 11.46, Gi = Γ i (T ) + RT ln fˆi

also we know, Giid = Gi + RT ln xi

= Γ i (T ) + RT ln f i + RT ln xi
so,

GiE = Γ i (T ) + RT ln fˆi − (Γ i (T ) + RT ln f i + RT ln xi )

GiE = RT ln fˆi − RT ln f i − RT ln xi

fˆi
G = RT (ln
i
E
)
xi f i
 Ac%vity  Coefficient  of  Species  i  
in  Solu%on   γ
i

The previous equation could be written as,

GiE = RT (ln γ i )i

where
fˆi
γi = an activity coefficient of species i
xi f i
or
fˆi = γ i xi f i

For a liquid solution!

 Ideal  Solu%on  
(Lewis/Randall  rule)  
For  ideal  solu%on,  
GiE = 0
so
fˆi id
RT (ln ) = RT (ln γ i ) = 0
xi f i
therefore,

fˆi id
γi = =1
xi f i
fˆi id = xi f i Lewis/Randall rule

also divide by Pxi

fˆi id xi f
=
Pxi Pxi
φ̂iid = φi
 Fundamental  Excess-­‐Property  Rela%on  
⎛ nG ⎞ 1 nG
d⎜ ⎟ = d(nG) − dT
⎝ RT ⎠ RT RT 2

⎛ nG ⎞ 1 ⎛ ⎞ n(H − TS)
d⎜ =
⎝ RT ⎟⎠ RT ⎜⎝ (nV )dP − (nS)dT + ∑ i i ⎟⎠ − RT 2 dT
G dn

⎛ nG ⎞ nVdP nSdT Gi dni nH nS

d⎜ ⎟
⎝ RT ⎠
=
RT
−
RT
+ ∑ RT
−
RT 2
dT +
RT
dT

⎛ nG ⎞ nVdP nH Gi
d⎜ ⎟ = − dT + ∑ dni (11.54)
⎝ RT ⎠ RT RT 2
RT

for ideal solution,

⎛ nG id ⎞ nV id dP nH id G idi
d⎜ ⎟ = − 2
dT + ∑ dni (11.54id)
⎝ RT ⎠ RT RT RT
 Fundamental  Excess-­‐Property  Rela%on  
(11.54) - (11.54id)

⎛ nGE ⎞ nV E dP nH E GEi
d⎜ ⎟ = − 2
dT + ∑ dni (11.89)
⎝ RT ⎠ RT RT RT

⎛ nGE ⎞ nV E dP nH E
d⎜ ⎟ = − 2
dT + ∑ ln γ i dni (11.93)
⎝ RT ⎠ RT RT
So,
⎡ ∂(nGE )⎤
ln γ i = ⎢⎢ RT ⎥ (11.96)
∂ni ⎥
⎢⎣ ⎥⎦P,T ,n
j

Therefore,
ln γ i is a partial property of GE
RT
 Consider a multicomponent system in VLE, the!
fugacity of species i for each phase,!

ˆ
fi = φ̂i y i P
v
For  vapor  mixture  
For liquid solution! fˆi l = γ i xi f i
VLE criteria,
fˆi v = fˆi l
so φ̂i yi P = γ i xi f i

This is the VLE relation that relates the composition of

vapor phase and that of liquid phase. See Chapter 10 for
application of this relation. Chapter 12 for correlation for γ i