THERMODYNAMICS Mathematics review 2020 TUTORIAL 0

Q-0-1 A function f(x,y,z) is called homogeneous of degree n if f(tx,ty,tz) = tnf(x,y,z).

(a) Prove Euler’s theorem on homogeneous functions that states that if f is
homogeneous of degree n, then
∂f ∂f ∂f
x + y +z = nf
∂x ∂y ∂z

Hint: Differentiate f(tx,ty,tz) = tnf(x,y,z) with respect to t, and then let t = 1.

(b) Show that the derivative functions of a homogeneous function of degree n are
homogeneous functions of degree n-1.

0-1 (a) We know f (tx, ty, tz ) = t n f ( x, y, z )

du dv dw
Let u = tx; v = ty and w = tz ⇒ = x; = y and =z
dt dt dt

d d ∂f du ∂f dv ∂f dw
LHS: f (tx, ty, tz ) = f (u , v, w) = + +
dt dt ∂u dt ∂v dt ∂w dt

∂f (u , v, w) ∂f (u , v, w) ∂f (u, v, w)
=x + y +z
∂u ∂v ∂w
∂f (u , v, w) ∂x ∂f (u , v, w) ∂y ∂f (u, v, w) ∂z
= x + y +z
∂x ∂u ∂y ∂u ∂z ∂u
1  ∂f (u , v, w) ∂f (u , v, w) ∂f (u, v, w) 
=  x + y +z 
t ∂x ∂y ∂z 
1  ∂f (tx, ty, tz ) ∂f (tx, ty, tz ) ∂f (tx, ty, tz ) 
=  x + y +z 
t ∂x ∂y ∂z 
1  ∂{t n f ( x, y, z )} ∂{t n f ( x, y, z )} ∂{t n f ( x, y, z )} 
= x + y +z 
t ∂x ∂y ∂z 
 ∂f ∂f ∂f 
= t n −1  x + y +z 
 ∂x ∂y ∂z 

d n
RHS: t f ( x, y, z )  = nt n −1 f ( x, y, z )
dt 

 ∂f ∂f ∂f 
So t n −1  x + y +z  = nt n −1 f ( x, y, z ) .
 ∂x ∂y ∂z 

∂f ∂f ∂f
⇒x + y +z = nf
∂x ∂y ∂z

1
 ∂ ∂
(b) Let=
f x ( x, y , z ) = f ( x, y , z ) f (u, v, w)
∂x ∂u

Now f (u , v, w) = t n f ( x, u , z )

Differentiate wrt x keeping y, z and t constant:

∂ ∂ ∂u
LHS:= f (u , v, w) = f (u , v, w) t f x (u , v, w)
∂x ∂u ∂x

∂ n
And RHS: t f ( x, y, z )  = t n f x ( x, y, z )
∂x 

⇒ f x (tx, ty, tz ) = t n −1 f ( x, y, z )

(c) The answer follows directly from substituting the definitions for T, P and µ i into

 ∂U   ∂U   ∂U 
dS =   dS +   dV +   dN i
 ∂S V , Ni  ∂V  S , Ni  ∂N i  S ,V

Q-0-2 Consider a system comprising two independent subsystems A and B. Note that every
microstate of system A can be chosen in combination with any microstate of system B. The
total number of microstates for the combined system, A plus B, is therefore: W = W A W B . On
the other hand, the entropy S of the system is the sum of the entropies of the subsystems A and
B:
S = SA + SB i.e. S(W) = S(W A W B ) = S(W A ) + S(W B )

Show, by appropriate differentiations, that the entropy function must satisfy the following
differential equation:
W S’’(W) + S’(W) = 0

Answer Q-0-2 (a)

S(W A W B ) = S(W A ) + S(W B )

Differentiate this equation to W A and then to W B :

d d d  dS dW  d  dS  dS d dS
S (=
W)  =   =
W B + WB
dWB dWA dWB  dW dWA  dWB  dW  dW dWB dW
LHS:
2 2
dS d S dW dS d S dS d 2S
= + WB = + WAWB = +W
dW dW 2 dWB dW dW 2 dW dW 2

d d d  dS (WA ) 
RHS: [ S (WA ) + S=
(WB ) ]  =  0
dWB dWA dWB  dWA 

2
Q-0.2-(b) Show that the general solution for this equation is: S (W) = C 1 n W + C2

Evaluate the constants C1 and C2 by considering: (a) the entropy of a perfect crystal; and (b) the
entropy change in the free expansion of two moles (i.e. 2Navo molecules) of a monoatomic ideal
gas at a pressure P expanding from a volume V1 to a volume V2 = 2V1. (Notes: (i) the entropy
of such a gas is given by S = -Nk n P + f(T) where f(T) is a function of temperature only. (ii)
Such a free expansion occurs isothermally.)

dS d 2S d  dS 
0 = +W = W 
dW dW 2
dW  dW 

dS
⇒ = C1
W
dW
⇒=S C1 nW + C2

(i) W = 1 for a monoatomic crystal at T = 0. ⇒ C2 = 0

(ii) Ideal gas expansion occurs isothermally

P2 V
∆S =− Nk n =Nk n 2
P1 V1
⇒ C1 = Nk = R

Q-0-3. Prove:
 ∂U  2  ∂ (P / T ) 
(a)   =T  
 ∂V T  ∂T V
 ∂H  2  ∂ (V / T ) 
(b)   = −T  
 ∂P T  ∂T P
 ∂T   ∂T 
(c) T dS = CV   dP + C P   dV
 ∂P V  ∂V P

Answer Q-0-3

(a) Note that the left hand side (LHS) is equivalent to the energy equation:
∂U   ∂P 
 = T  − P
∂V T  ∂T V

Now expand the right hand side (RHS):

 ∂ (P / T )  
2 1 ∂P P ∂P
T 2  =T   − 2  = T  − P = LHS
 ∂T V  T ∂ T V T  ∂ T V

(b) Notes that the left hand side (LHS) is equivalent to the enthalpy equation:

3
 ∂H   ∂V 
 = V − T 
∂P T  ∂T  P

Now expand the right hand side (RHS):

 ∂ (V / T )  
2 1 ∂V  V  ∂V 
−T2  = −T   − 2  = V − T  = LHS
 ∂T  P  T ∂ T  P T  ∂ T  P

(c) S = S (V,P)

∂S  ∂S 
=dS  dV +  dP
∂V  P ∂P V
∂S  ∂ T  ∂S  ∂T 
=TdS T   dV + T   dP
∂T  P ∂ V  P ∂T V ∂P V
∂T  ∂T 
= CP  dV + CV  dP
∂V P ∂P V

Q-0-4. A simple system satisfies the following equation of state:

P V = RT (1+ B /V )
The energy equation reads:
 ∂P  
dU = C v dT + T  − P  dV
 ∂T V 
(a) Prove that the internal energy of this system is a function of temperature only.
(b) Derive, for the case Cv = constant, a relationship between pressure and volume in a
reversible adiabatic process.

Answer Q-0-4.

(a) The energy equation is only a function of temperature and volume:

RT  B
P = 1 + 
V  V
∂P  RT  B
⇒ T  = 1 +  = P
∂T V V  V
⇒ dU = CV dT + [P − P ] dV = CV dT
This implies that the internal energy U is a function of temperature only.

(b) For an adiabatic process dU = - PdV

4
 RT  B
⇒ CV dT = − P dV = − 1 +  dV
V  V

CV dT 1 B dV B
= −  1 +  dV = − − 2 dV
R T V V  V V
CV  B
d n T = − d n V + d  
R V
Integrate from state 1 to state 2:
CV T V  1 1
n 2 = − n 2 + B  − 
R T1 V1  V2 V1 

T2 R V BR  V1 − V2 
n = − n 2 +  
T1 C v V1 CV  V1V2 

R
T  V  CV BR  V1 − V2 
n 2 + n  2  =  
T1  
V1 CV  V1V2 

R
T2  V2  CV  BR  V1 − V2 
  = exp   
T1  V1   CV  V1V2 

RT  B PV 2
But from: P = 1 +  One can obtain: T =
V  V R (B + V )

2
T2 P V  B + V1
Which implies that: = 2  2 
T1 P1  V1  B + V2

R
 BR  V1 − V2 
2
P2  V2  V  CV B + V1
   2  = exp   
P1  V1   V1  B + V2  CV  V1V2 

R
2+
P2 V  CV  B + V2   BR  V1 − V2 
Substituting ⇒ =  1    exp   
P1  V2   B + V1   CV  V1V2 

Q-0-5. The Helmholtz equation enables us to determine the variation of the internal energy with
volume if the equation of state is known:

5
 ∂U  ∂ P
 = T 2  
∂ V T ∂ T  T V
i. Show that the internal energy of an ideal gas does not depend on its volume
Answer Q-0-5 (i).
∂U   ∂P 
 = T  − P
∂V  T  ∂T V

 ∂P  ∂  RT  RT
T  = T   = = P
 ∂T V ∂T  V V V

∂U 
⇒  = 0
∂V  T

Thus the internal energy is not a function of volume but only of temperature: U = U(T)

ii. Consider a real gas and show that the Helmholtz equation can be integrated to read:
V  ∂ P 
U real (T , V ) = U ideal (T ) + ∫ T 2     dV
∞
 ∂ T  T V 
Answer Q-0-5 (ii).

∂ P
At constant temperature the Helmholtz equation reads: dU = T 2   dT
∂T  T V

Integrate the Helmholtz equation at constant T from V = ∞ to V = V:

U (T ,V ) V ∂ P
∫U (T )
dU = ∫ ∞
T2   dV
∂ T  T V

V  ∂ P 
U real (T , V ) = U ideal (T ) + ∫ ∞
T 2 
 ∂T
 
 T V
 dV

The first term on the right equals the ideal gas internal energy because all gases approach ideal
gas behaviour as density vanishes, i.e. volume become infinite. Furthermore, the internal energy
of ideal gases depends on temperature only.

ii. Use this relation to show that the internal energy of a van der Waals gas is given by:

a
U vdWaals (V , T ) = U ideal (T ) −
V

RT a
Answer Q-0-5 (iii). For the van der Waals gas: P = − 2
V −b V

6
 ∂ P ∂  R a  a
  =  − 2 
= 2 2
∂ T  T V ∂T V − b T V  V T V

V ∂ P V  a  a
∫ ∞
T2   dV =
∂ T  T V ∫ ∞
 2  dV = −
V  V
a
Thus U vdWaals (V , T ) = U ideal (T ) −
V
iv. Give a physical interpretation for the second term on the right. What are the units of the
constant a?

Answer Q-0-5 (iv).

The second term on the right represents the effect of the attractive forces between the molecules.
It is clear that a is a specific energy, i.e. “attraction” energy times unit volume. Typical units
would be kJ.m3/mol2.

7
THERMODYNAMICS (2020) TUTORIAL 1

Q-1.1. A mass m of water at T 1 is isobarically and adiabatically mixed with an equal mass of water
at T 2 . Show that the entropy change of the universe is

(T1 + T2 ) / 2
2mC P n
T1T2

and prove that this is positive by using a geometric construction involving rectangles with sides
T 1 and T 2 .

Answer Q-1.1.

T1 T2 T
m m 2m

T2 + T1
Adiabatic process ⇒q=0 ⇒ T =
2
dq 1 = mC p dT 1 and dq 2 = mC p dT 2
dq 1 + dq 2 = 0 ⇒ dq 1 = -dq 2
mCpd T T dT T
dS1 =
T1
⇒ ∆S1 = ∫ T1
mC p
T
= mC p n
T1

T mC p dT T
Simililarly: ∆S 2 = ∫T2 T
= mC p n
T2

  T + T2   T + T2   (T1 + T2 )2 
∆S = ∆S1 + ∆S 2 = mC p n  1  + n  1  = mC p n  
  2T1   2T2   4T1T2 

But we require that ∆S ≥ 0

(T1 + T2 )2
⇒ > 1
4T1T2

The following geometric Figure shows that this is indeed true.

The area of the four grey blocks represent the

T2 quantity 4 T 1 T 2
The total area of the rectangle represents the
quantity (T 1 + T 2 )2
T1

T2 T1

1
Q-1.2. (a) If increase in entropy (and increase in disorder) is the sine qua non of a spontaneous
process, how is it possible for a liquid, once vaporised, ever to re-condense; or for a solid, once
fused, ever to recrystallize spontaneously?
(b) Let a mole of liquid at its normal freezing point (T F ) be transformed at a finite rate into a mole
of solid at the same temperature. For the total consequent entropy change of this system and its
surroundings, the appropriate expression is

T − T 
∆S = ∆H fus  F 
 TTF 

Where ∆H fus is the molar heat of fusion of the solid and T is the (constant) temperature of the
surroundings. Derive this expression, and indicate why necessarily ∆S > 0 if the solidification
proceeds at a finite rate.

Answer Q-1.2.

(a) Total entropy change of system plus environment must be positive. The system entropy
may decrease provided that the entropy of the environments increases by a larger amount.
Qsystem Q
∆H freezing = ∆H fusion ∆S tot = + environment
(b) T T
Qsystem = − ∆H fus = − Qenviron

− ∆H fus ∆H F  −1 1  T − TF  T −T 
⇒ ∆S tot = + = ∆H fus +  = − ∆H fus   = ∆H fus  F 
TF T  TF T  T TF   T T F 

Freezing only possible if T < T F ⇒ argument is positive

Q-1.3. Two identical systems in rigid boundaries have the same properties and the same initial
state. Adiabatic walls surround the two systems, together, as shown in Figure 1. A refrigerator is
interposed between the two systems. Find:

(a) The work as a function of T 1 and T 2 .

(b) The range of temperatures the second system can assume when the first system reaches T 1 .
(c) The entropy as a function of the work and T 1 .
(d) The minimum work required to change the temperature of one system to T 1 where T 1 < T 2 .

w
adiabatic wall

q2 q1
1 2

Answer Q-1.3.

2
 dq1 = mC p dT1 ⇒ q1 = mC p (T1 − T0 )

dq 2 = mC p dT2 ⇒ q 2 = mC p (T2 − T0 )

w = q 2 + q1 = mC p [T2 + T1 − 2T0 ] (a)

mC p dT T1 T2
dS1 = ∆S1 = mC p n ∆S 2 = mC p n
T T0 T0

 T T   T1 T2 
∆S = ∆S1 + ∆S 2 = mC p n 1 + n 2  = mC p n 2 
 T0 T0   T0 
T1 T2 T02
But ∆S ≥ 0 ⇒ ≥0 i.e. T2 ≥ (b)
T02 T1

w
From (a): T2 = + 2T0 − T1
mC p

T  w 
∆S = mC p n  12  + 2 T0 − T1  (c)
 T0  mC 
 p 

Minimum work when the equal sign holds in equation (b):

 T02   T02 + T12 − 2T0T1 

wmin 
= mC p  + T1 − 2 T0  = mC p 
  = mC p (T0 − T1 )2 / T1 (d)
 T1   T1 

Q-1.4. Show how the efficiency of a Carnot engine can be derived directly from a TS diagram.
Determine and compare the efficiencies of cycles (a) and (b) in the Figure 2 below with the
efficiency of a Carnot cycle. Which one of (a) or (b) is the most efficient?

(a) (b)
T2 T2

T T

T1 T1

S1 S2 S1 S2
S S

Answer Q-1.4.

3
 A isothermal B A B
T2

P adiabatic adiabatic T

D T1
isothermal C
D C

V F S E
Figure: The Carnot Cycle
Consider the right hand side:
T2 a. The area of rectangle ABEF represents
q2
q2

w b. The area of rectangle DCEF represents

q1
q2 c. The work done is given by the
T1 difference, i.e. w = q 2 – q 1 or in other
words, the area defined by the rectangle
ABCD.
Thus we can determine the efficiency as the
ratio of area ABCD to ABEF

Using this principle, the efficiency of cycle (a) is:

½(T2 − T1 )( S 2 − S1 ) (T − T )
ηa = = 2 1
T2 ( S 2 − S1 ) 2 T2

½(T2 − T1 )( S 2 − S1 ) (T2 − T1 )
and for cycle (b): ηb = =
½(T2 − T1 )( S 2 − S1 ) + T1 ( S 2 − S1 ) (T2 + T1 )

Since 2 T 2 > T 2 + T 1 , cycle (b) is more efficient.

4
Q-1.5. (a) The temperature of an ideal gas with constant heat capacity is changed from T 1 to T 2 .
Show that ∆S of the gas is greater if the change in state occurs at constant pressure than if at
constant volume.
(b) The pressure of the gas is now changed from P 1 to P 2 by an isothermal process and by a
constant volume process. Show that the entropy changes in the gas have opposite sign changes for
the two processes.

Answer Q-1.5.
S2 T2 dq
(a) ∆S = ∫ S1
dS = ∫T1 T
Constant pressure process:
dq = C p dT1
T2 C p dT T2
⇒ ∆S ) p = ∫T1 T
= C p n
T1

T2
Constant volume process: ∆S )V = CV n
T1

Now ∆S) P  > ∆S)V  because Cp = CV + R

(b) Isothermal ( T = constant)

q = w
d w = R T d n V = − R T d n P
V P
w = R T n 2 = R T n 1
V1 P2
q w P
∆S ) T = = = R n 1
T T P2

Isochoric (V = constant)
T2 dqV T2 dT T P
∆S )V = ∫
T1 T
= ∫T1
CV
T
= CV n 2 = CV n 2
T1 P1

Thus the entropy changes have opposite signs.

Q-1.6. Figure 3 below represents an imaginary ideal gas-engine cycle. Assuming constant heat
capacities, show that the thermal efficiency is

5
 (V1 /V 2 ) − 1
η =1−γ
(P1 / P2 ) − 1

T2 a
q2 P1

adiabatic
w
P2
q2 c b
T1
V1 V2
Figure 3: Ideal engine cycle
Answer Q-1.6.
w q 2 − q1 q
η= = = 1 − 1
q2 q2 q2

a → b: Adiabatic expansion: q=0

b → c: Isobaric compression: dq = C p dT ⇒ qbc = C p (Tc − Tb )

Cp
qbc = (P2V1 − P2V2 ) = − q1
R
b → c: Isochoric compression: dq = CV dT

CV q1 C p (P2V2 − P2V1 )  (V / V ) − 1
q ca = CV (Ta Tb ) = (P1V1 − P2V1 ) = q 2 ⇒ = = γ 2 1 
R q2 CV (P1V1 − P2V1 )  (P1 / P2 ) − 1

 (V / V ) − 1
Therefore η =1−γ  2 1 
 (P1 / P2 ) − 1

Q-1.7. In the reversible adiabatic expansion of an ideal gas, none of temperature, pressure, or
volume remains constant. Yet we can easily characterise this expansion by imagining a two-step
route that links the same initial and final states between which the system is carried by the adiabatic
expansion. Figure 4 shows such a two-step route, the first leg of which represents a cooling at
constant volume, the second an expansion at constant temperature.

(a) Use this two-step process to show that for the reversible adiabatic expansion of one mole of
dT dV
an ideal gas: Cv = −R .
T V
(b) Define a new parameter γ ≡ C P / CV and show that, on the assumption that Cv is constant,
integration of the last equation leads to:

6
 R / Cv
T2 V  P
=  1 or TV γ −1 = const . or CP / R
= const . or PV γ = const .
T1 V 2  T

P1

adiabatic

2
P2

V1 V2

Figure 4.
Answer Q-1.7.

Basis: 1 Mol of ideal gas

Step 1: Cooling at constant volume ⇒ dw1 = 0

dU 1 = dq1 = CV dT

Step 2: Expansion at constant temperature ⇒ dU2 = 0

⇒ dq 2 = dw2 = PdV

The overall process is adiabatic, i.e. dqtot = dq1 + dq 2 = 0

⇒ CV dT = − PdV

dT dV
Substitute P = RT/V and simplify: CV = −R
T V

T2 V V
Integration yields: CV n = − R n 2 = (C v − C p ) n 2
T1 V1 V1

Where we used the identity for an ideal gas: R = Cp - Cv

1−γ
T V   T1V1γ −1 
Thus: n 2 = n  2  or n  γ −1 
= 0
T1  V1   T2V2 
i.e.
T1V1γ −1 = T2V2γ −1 = constant

The other relationships are obtained by applying the ideal gas law.

7
THERMODYNAMICS (2020) TUTORIAL 2

Q-2.1. Prove:
 ∂U  2  ∂ (P / T ) 
(a)   =T  
 ∂V T  ∂T V
 ∂H  2  ∂ (V / T ) 
(b)   = −T  
 ∂P T  ∂T P

Answer Q-2.1.

(a) Note that the left hand side (LHS) is equivalent to the energy equation:
∂U   ∂P 
 = T  − P
∂V T  ∂T V

Now expand the right hand side (RHS):

 ∂ (P / T )  
2 1 ∂P P ∂P
T 2  =T   − 2  = T  − P = LHS
 ∂T V  T ∂ T V T  ∂ T V

(b) Notes we note that the left hand side (LHS) is equivalent to the enthalpy equation:
∂H   ∂V 
 = V − T 
∂P T  ∂T  P

Now expand the right hand side (RHS):

 ∂ (V / T )  
2 1 ∂V  V  ∂V 
−T2  = −T   − 2  = V − T  = LHS
 ∂T  P  T ∂ T  P T  ∂T P

Q-2.2. (a) Derive the third T dS equation:

 ∂T   ∂T 
T dS = CV   dP + C P   dV
 ∂P V  ∂V P

(b) Show that the T dS equations may be written:

βT
T dS = CV dT + dV
k
T dS = C P dT − Vβ T dP
CV κ CP
=
T dS dP + dV
β βV

1
 Answer Q-2.2.

(a) S = S (V,P)

∂S  ∂S 
=dS  dV +  dP
∂V  P ∂P V
∂S  ∂ T  ∂S  ∂T 
=TdS T   dV + T   dP
∂T  P ∂ V  P ∂T V ∂P V
∂T  ∂T 
= CP  dV + C  dP
∂V P ∂P V
V

(c) Recall:

Heat capacities: ∂U  ∂S  ∂U  ∂S 
CV =  = T  and C P =  = T 
∂T V ∂T V ∂T  P ∂T  P

Isothermal compressibility 1  ∂V 
κ = −  
V  ∂P  T
Volume expansivity 1  ∂V 
β =  
V  ∂T  P

Substitute into TdS equations and the required results are obtained. For example, consider the
3rd TdS equation:

∂T  ∂T 
TdS = C P  dT + CV  dP
∂V P ∂P V

∂T  1
Now  =
∂V P βV

∂V  ∂V 
From V = V(T,P): dV =  dT +  dP
∂T P ∂P  T

∂V  ∂T  ∂P 
Keeping V = constant yields the mathematical identity:    = −1
∂T  P ∂P V ∂V  T

∂T  ∂ V  ∂T  κ
⇒  = − 
  =
∂P V ∂ P  T ∂V  P β
Substitution in 3rd TdS equation yields the required result.

2
Q-2.3. The Joule-Thompson coefficient {(∂P/∂T) H } is of importance in refrigeration
engineering. Show that it can be calculated from PVT and heat capacity data by:

∂P  1   ∂V  
 = − V − T   
∂T  H Cp   ∂T  P 

Answer Q-2.3.

 ∂V  
The enthalpy equation reads: dH = C P dT + V − T   dP
 ∂T  P 

 ∂V   ∂P 
Differentiate with respect to T keeping H = constant: 0 = C P + V − T   
 ∂T  P  ∂T  H

Rearranging provides the required result

Q-2.4. A simple system satisfies the following equation of state:

P V = RT (1+ B /V )

(a) Prove that the internal energy of this system is a function of temperature only.
(b) Derive, for the case C v = constant, a relationship between pressure and volume in a
reversible adiabatic process.

Answer Q-2.4.

(a) The energy equation is only a function of temperature and volume:

 ∂P  
dU = C v dT + T  − P  dV
 ∂T V 
RT  B
P = 1 + 
V  V
∂P  RT  B
⇒ T  = 1 +  = P
∂T V V  V
⇒ dU = CV dT + [P − P ] dV = CV dT
This implies that the internal energy U is a function of temperature only.

3
(b) For an adiabatic process dU = - PdV
RT  B
⇒ CV dT = − P dV = − 1 +  dV
V  V

CV dT 1 B dV B
= −  1 +  dV = − − 2 dV
R T V V  V V
CV  B
d n T = − d n V + d  
R V
Integrate from state 1 to state 2:
CV T V  1 1
n 2 = − n 2 + B  − 
R T1 V1  V2 V1 

T2 R V BR  V1 − V2 
n = − n 2 +  
T1 C v V1 CV  V1V2 
R
T V  CV BR  V1 − V2 
n 2 + n  2  =  
T1  V1  CV  V1V2 
R
T2  V2  CV  BR  V1 − V2 
  = exp   
T1  V1   CV  V1V2 

RT  B PV 2
But from: P = 1 +  One can obtain: T =
V  V R (B + V )

2
T2 P V  B + V1
Which implies that: = 2  2 
T1 P1  V1  B + V2

R
 BR  V1 − V2 
2
P2  V2   V2  CV B + V1
   V  = exp   V V 
P1  V1   1  B + V 2  V  1 2
C 

Substituting
R
2+
P2 V  CV  B + V2   BR  V1 − V2 
⇒ =  1    exp   
P1  V2   B + V1   CV  V1V2 

4
Q-2.6. A substance, for which the isothermal compressibility κ is constant, undergoes an
isothermal, mechanically reversible process from initial state (P 1 , V 1 ) to a final state (P 2 , V 2 )
where V is the molar volume.

(a) Starting with the definition of κ, derive an expression that describes the path of the
process in the P-V plane.

(b) Determine an exact expression that gives the isothermal work done on 1 mol of this
constant-κ substance.

1 ∂V 
κ = − 
V ∂P  T

Answer Q-2.6 (a). For an isothermal process, T = constant. Rearranging yields

1 ∂V  V dV P V
−  = κ ⇒ −∫ = − κ ∫ dP ⇒ n = κ ( P − P1 )
V ∂P  T V1 V P1 V1

V = A exp (− κ P )
V
⇒ n = − κ P + κ P1 ⇒ where A = V1 exp (κ P1 )
V1

Answer Q-2.6 (b).

dV  1 ∂V 
dw = − PdV = − P dP = − PV   dP = κ PV dP
dP  V ∂P T

A
dw = κ APe −κP dP = xe − x dx
κ

[( x +1) e ]
P 2
 A −κP  V 
2

= −  e (κP + 1) = −  (κP + 1)

A x2 A
∫
−x − x x2
w = xe dx = −
κ x1 κ x1
κ  P1 κ 1

⇒ w =
V1
(κP1 +1) − V2 (κP2 + 1)
κ κ

Q-2.7. For liquid water the isothermal compressibility is given by:

c
κ =
V ( P + b)
Where c and b are functions of temperature only. If 1 kg of water is compressed isothermally
and reversibly from 1 to 500 bar at 60ºC, how much work is required? At 60ºC, b = 2700 bar
and c = 0,125 cm3/g.

5
Answer Q-2.7. For an isothermal process, T = constant. The work done on the water is:

dw = − P dV

1 ∂V  c
From κ = −  =
V ∂P  T V ( P + b)

c cP
We obtain by rearrangement: dV = − dP ⇒ dw = − PdV = dP
P+b P+b

dP = c [P − b n ( P + b)]1
500 cP
w = −∫
500
Integrate:
1 P+b

500
m3   2700 + 500 
499 − 2700 n  
−4
⇒ w = 1,25.10 = 515,9 J
kg   2700 + 1  1

6
THERMODYNAMICS (2020) TUTORIAL 3

Q-3.1. Formaldehyde may be produced in the reaction between methanol and oxygen:
2 CH 3 OH (l) + O 2 (g) = 2 HCHO (g) + 2 H 2 O (l) ∆H R ° = -326.2 kJ/mol
The standard heat of combustion of hydrogen is:
H2 (g) + ½O2 (g) = H2O (l) ∆HR° = -285.8 kJ/mol
Use these heats of reaction and Hess’s law to determine the standard heat of the direct
decomposition of methanol to form formaldehyde:
CH3OH (l) = HCHO (g) + H2 (g)
Answer Q-3.1.
Rewrite the reactions as follows:
2 CH3OH (l) + O2 (g) = 2 HCHO(g) + 2 H2O (l) ∆HR° = -163.1 kJ/mol
H2O (l) = H2 (g) + ½O2 (g) ∆HR° = 285.8 kJ/mol
Overall: CH3OH (l) = HCHO (g) + H2 (g) ∆HR° = 122.7 kJ/mol

Q-3.2. With reference to the decomposition reaction for calcium carbonate at 298°K:
CaCO3(s) = CaO(s) + CO2(g)
∆H (kcal/mol)
o
f -288.5 -151.9 -94.05
o
S 298 (cal/mol-°K) +22.2 +9.5 +51.06
(a) Determine the equilibrium constant for the reaction at 298°K, and state its units.
(b) On the (good) assumption that ∆HR° for the reaction is substantially constant over the
temperature range concerned, calculate the equilibrium temperature at which the equilibrium
pressure of CO2 becomes equal to 1 atm.

Q-3.2 (a).
∆H Ro = ∆H of ,CaO + ∆H of ,CO2 − ∆H of ,CaCO3 = 42.55 kcal/mol = 178.03 kJ/mol

∆S Ro = S of ,CaO + S of ,CO2 − S of ,CaCO3 = 38.36 kcal/mol.K = 160.50 kJ/mol.K

∆G Ro = ∆H Ro − T ∆S Ro = 130200.7 J/mol

 ∆GRo 
K = exp  −  = 1.5082 x 10-23
 RT 
pCO2
K = ⇒ The equilibrium constant is a dimensionless quantity.
Po

1
Q-3.2 (b). Since ∆HR° is temperature independent, Van’t Hoff’s law allows one to estimate the
effect of temperature on the equilibrium constant. The calculation is simplified if we express
the pressure in atmospheres.
−1
K − ∆H Ro 1 1 1 R K 
n = ( − ) ⇒ T =  − n o 
Ko R T To  To ∆H R
o
K 

pCO2
The reference pressure is 1 atm. We seek the temperature for which K = =1
Po
−1
 1 8,3145 1 
⇒ T =  − n − 23 
= 1109.2 K
 298 178030 1,5082.10 

Q-3.3. Sulfur dioxide from Gypsum. We wish to investigate the possibility of producing SO2
by reducing gypsum, CaSO4, with carbon monoxide according to the following reaction:
CaSO4 (c) + CO (g) = CaO (c) + SO2 (g) + CO2 (g) (I)
However, the following undesirable reaction could occur:
CaSO4 (c) + 4CO (g) = CaS (c) + 4CO2 (g) (II)
There is also the possibility that calcium carbonate will form according to the reaction:
CaO (c) + CO2 (g) = CaCO3 (c) (III)
At 1200ºK the equilibrium constants for these reactions are:
KI = 0.31 KII = 7.92 KIII = 1.4

For the process carried out at 1200ºK and 1 atm pressure, determine:
(a) The minimum ratio of CO2 to CO that will suppress reaction (II);
(b) The maximum mol fraction SO2 which will occur when reaction (II) is just suppressed; and
(c) Whether CaCO3 (c) will be present when reaction (II) is suppressed.

2
Answer Q-3.3.
4
pCO pCO2
(a) K II = 2
4
= 7,92 ⇒ = 1,67757
p CO pCO

If pCO2 > 1,67757 pCO , then the reverse of reaction II will occur.

pCO2 p SO2 pCO 0,31

(b) K I = = 0,31 ⇒ p SO2 = 0,31 ≤ = 0,18479
pCO pCO2 1,67757

)
p SO2
⇒ y SO2 max
= = 0,18479
P
1
(c) K III = = 1,4 ⇒ requires that pCO2 ≥ 0,71429 for CaCO3 to form
pCO2

However, at maximum formation of SO2 we have

pCO2 + pCO = 1 − p SO2 = 0,81521 = 2,67757 pCO

⇒ pCO2 = 0,6955

Since pCO2 < 0.71429 , reaction III will not form solid CaCO3

Q-3.4. Ethylene can be produced from the vapour phase dehydration of ethanol:
C2H5OH = C2H4 + H2O
Calculate the equilibrium conversion at 100°C and 1 atm pressure given that Kp = 3.

Answer Q-3.4.
Choose as basis 1 mol ethanol to react to equilibrium and assume that the vapour is ideal. Let
the conversion of the ethanol be ε:
C2H5OH C2H4 H2O Total mol
Initial mol: 1 0 0 0
At equilibrium: 1-ε ε ε 1+ε
yi (1 - ε)/(1 + ε) ε/(1 + ε) ε/(1 + ε)

Note that pi = yiP and that the process occurs at 1 atm. This simplifies calculations!
p C 2 H 4 p H 2O y C 2 H 4 P y H 2O P y C 2 H 4 y H 2O ε2
Kp = = = = = 3
pC2 H 5OH y C2 H 5OH P y C2 H 5OH 1− ε 2

Solving this quadratic equation yields ε = 0.8660

3
Q-3.5. It is proposed to clean tungsten wire (remove all traces of the oxide WO2) by heating the
wire to 1000 K in a hydrogen atmosphere (1 atm total pressure) in order to take advantage of
the following reaction:
WO2 (c) + 2 H2 (g) = W(c) + 2 H2O (g)
Will this work? If so, what is the maximum water impurity level that can be tolerated in the
hydrogen? Free energies of formation are as follows:
WO2 (c): ∆Gf° = -131 600 + 36.6 T
H2O (g): ∆Gf° = - 58 900 + 13.1 T
Where ∆Gf° is in cal/mol and T is in Kelvin.

Answer Q-3.5.
∆G Ro = 2 ∆G of , H 2O − ∆G of ,WO2 = 2 (−58900 + 13,1T ) − (−131600 + 36,6 T ) = 13800 −10,4 T

∆G Ro (1000 K ) = 3400 cal/mol = 14 226 J/mol

2
 pH O   ∆G Ro 
K =  2  = exp  −  = 0,1807
 pH   RT 
 2 
p H 2O
⇒ = K = 0,4251
pH2

Yes, the cleaning reaction will work provided p H 2O ≤ 0,4251 p H 2

But p H 2O + p H 2 = 1 atm ⇒ p H 2O ≤ 0,2983

4
Q-3.6. 100 kmol/h of a waste gas stream (at ambient temperature and a pressure of >200
atmospheres) with composition 98% H2 and 2% HF (mole basis) is available. The hydrogen
could be sold if the HF content could be reduced to less than 1 ppm (parts per million). You are
a member of the team looking for a process to achieve this. Note that hydrogen fluoride is highly
reactive and is also corrosive, especially in the presence of moisture. Your supervisor asks you
to consider the data given in the Table below and to make recommendations regarding a
possible route to remove the HF impurity.
Q-3.6.1. Explain and define the terms ∆Hfº and ∆Gfº given in the Table.
Q-3.6.2. Write down at least two possible chemical reactions that could be used for the removal
of the HF. Compare their advantages and disadvantages and then select the one you regard as
most promising.
Q-3.6.3. Use thermo-chemistry to determine whether this reaction will actually proceed at room
temperature and pressure.
Q-3.6.4. Determine whether the required purity can be achieved at room temperature and
pressure. Indicate how you would correct for conducting the reaction at high pressures.
Q-3.6.5. Estimate the heat duty of the heat exchanger that would be needed on the assumption
that the gas stream needs to be maintained at room temperature.
Q-3.6.6. Based on the densities given in the Table, would you expect the reaction product to
present a diffusion barrier for further reaction?
Q-3.6.7. Use the data in the Table to estimate the change of the reaction rate constant with
temperature.
Compound State ∆Hfo ∆Gfo So Cpo Density
[kJ/mol] [kJ/mol] [J/mol.°K] [J/mol.°K] [g/cm3]
H2O liquid -286.02 -237.34 69.96 75.34 1.00
gas -241.98 -228.74 188.84 33.60 -
HF gas -271.30 -273.40 173.79 29.14 -
H2 gas 0 0 0 0 8.99x10-5
CaO crystal -635.56 -603.95 38.23 42.16 3.25
CaH2 crystal -186.31 -147.38 41.87 - 1.90
CaF2 crystal -1220.45 -1168.12 68.91 67.07 3.18

5
Standard Enthalpy (∆Hfº) (or Gibbs free energy (∆Gfº)) of formation represents the change in
the enthalpy (Gibbs free energy) when one mol of the given substance is formed isothermally,
and at the given pressure, from the elements, each in its appropriate reference or standard state.

Note that calcium fluoride is a common fluoride mineral in nature. This implies that it is
probably a very stable compound. We therefore seek reactions that will remove the HF and
have CaF2 it as a product:
Reaction I: 2 HF (g) + CaO(s) = CaF2 (g) + H2O(g)
Reaction II: 2 HF (g) + CaH2(s) = CaF2 (g) + 2 H2(g)
Reaction I Reaction II
Advantages CaO is cheap and freely available No contamination of the H2
product
Disadvantages Water contamination may also lead CaH2 is expensive
to corrosion

On balance we choose reaction II as very little HF needs to be removed and because it will not
lead to corrosion.

Reaction II: 2 HF (g) + CaH2(s) = CaF2 (g) + 2 H2(g)

∆GRo = ∆G of ,CaF2 − ∆G of ,CaH 2 − 2 ∆G of , HF = -473.94 kJ/mol

The Gibbs free energy of reaction at 25°C and 1 atm pressure is large and negative. This implies
that the equilibrium will lie far to the right, i.e. the reaction is very promising.

2
 pH   ∆G Ro 
K =  2  = exp  −  = 1.072 x 1083!
 p HF   RT 
 2 
pH2
⇒ = K = 3.274 x 1041!
p HF

This means that the H2 partial pressure will, for all practical purposes be 1 atm and that the
required purity can be obtained as yHF < 3 x 10-42 << 1 ppm!

6
Reaction II: 2 HF (g) + CaH2(s) = CaF2 (g) + 2 H2(g)
∆H Ro = 2 ∆H of , H 2 + ∆H of ,CaF2 − ∆H of ,CaH 2 − 2 ∆H of , HF = -491.54 kJ/mol

Thus the heat that needs to be removed per mol HF is ∆H Ro / 2 = -245.77 kJ/mol
The heat duty required is:
q = m ∆H R = 0.02 (100 kmol/h) (-245.77 kJ/mol)/(3600 s/h) = -136.5 kW

This is a gas-solid reaction. If the product forming on the surface forms a dense layer, it may
present a diffusion barrier that will slow down the reaction.

The molar masses of the two compounds are: CaH2 : M = 42.1 and CaF2: M = 78.08
The density of CaH2 is 1.9 g/cm3. If there were straight substitution of the hydrogens with
fluorine in the CaH2 lattice the expected density of the CaF2: would be:
ρcalcium fluoride = 1.9 (78.08)/(42.1) = 3.52 g/cm3

The actual density is in fact only 3.25 g/cm3. This implies a volume expansion of the crystal as
it is converted to CaF2. A volume expansion should cause a crumbling of the converted layers
on the surface and will therefore continuously provide fresh surface for further reaction.

If we can assume that the enthalpy of the reaction is more or less temperature independent, then
we can use Van’t Hoff’s law to estimate the effect of temperature on the equilibrium constant.
K2 − ∆H Ro 1 1  − ∆H Ro Tref 
n = ( − ) or K = K ref exp  ( − 1)
K1 R T2 T1  R Tref T 
Choosing T = 298.15 K as reference yields:
 298,15 
K = 1,072 .10 81 exp 198,2839 ( − 1)
 T 

7
Q-3.7. A gas containing only CO2 and H2 in a 1:1 mole ratio is reformed by passing it over a
powerful catalyst at 1000 K and at various pressures. It may be assumed that only the reactions
given in the table below occur, and that equilibrium is attained.

Reaction log10Kp
I C + CO2 = 2CO ?
II C + 2H2O = CO2 + 2H2 0.7
III C + H2O = CO + H2 0.5
IV CO + H2O = CO2 + H2 ?
Pressure measured in atmospheres.

(a) Calculate log10Kp for reactions I and IV.

(b) Over what range of pressures may carbon deposit on the catalyst?
(c) What must the pressure be if exactly one quarter of the carbon in the feed precipitates as
solid carbon?

Reaction Kp log10Kp
p co2
KI =
I C + CO2 = 2CO pCO2 ?

pCO2 pH2 2
K II =
II C + 2H2O = CO2 + 2H2 pH2 2 O 0.7

pCO pH 2
K III =
III C + H2O = CO + H2 pH 2 O 0.5

pCO2 pH 2
K IV =
IV CO + H2O = CO2 + H2 pCO pH 2O ?

8
By inspection, we notice that KI can be expressed in terms of KII and KIII and KIV in terms of KI
and KIII.
2 2
pCO p H2 2  p H2 2O  2
KI :
K III
=   = pCO = K I ⇒ log10KI = 0.30
K II p H2 2O  pCO p H2  pCO2
 2 2 

K III pCO p H 2  pCO2  p p

KIV : =   = CO2 H 2 = K IV ⇒ log10KIV = 0.20
KI p H 2O  pCO
2  pCO p H 2O
 
The important point to note is that the system is fully described when any two of the above
reactions are selected. The other two are then superfluous.

Reactions I, II and III can lead to carbon deposition. Note that all three indicate an increase in
the mol gas if the reactions proceed in the directions indicated. According to Le Chatteliers
principle, this means that carbon formation will be favoured at higher pressures. No carbon has
formed. Thus reaction IV determines the equilibrium composition.

Assume that ε mole each of H2 and CO2 have reacted to form ε mole each of CO and H2O. Use
carbon and hydrogen mass balances to calculate the conversions indicated in the Table below.
Also assume that the vapour is ideal!

CO + H2 O = CO2 + H2 Total
Initial moles 0 0 1 1 2
Equilibrium ε ε 1-ε 1-ε 2

pCO2 pH 2 yCO2 yH 2 (1− ε ) 2 1− ε

At equilibrium: K IV = = = ⇒ K IV =
pCO pH 2 O yCO yH 2 O ε 2
ε

K IV
Thus: ε = = 0.5573
1+ K IV

Note that Reaction IV is not affected by the system pressure.

Since two reactions suffice to describe the system, choose reaction I for carbon deposition
analysis as it is the simplest and therefore the most convenient.

p co2 y co2 P 2 y co2 ε2

KI = = = P = P = 1.9953
pCO2 y CO2 P y CO2 2 (1 − ε )

9
Substituting ε = 0.5573 yields a value for P = 5.69 atm. This is the pressure above which carbon
deposition will occur. Below this pressure no carbon will be present provided equilibrium is
reached.

Basis: 1 mol CO2 and mol 1 H2 in feed (0.25 mol C present at equilibrium)
Once again we only need to consider the following two reactions:
Reaction IV: C + CO2 = 2CO
Reaction IV: CO + H2O = CO2 + H2
Assume that ε mol of H2 has reacted to form ε mol H2O and that δ (δ mol >0.25) mol of the
CO2 has reacted to form 0.25 mol carbon (C) and δ − 0,25 mol CO.

Component CO H2 O CO2 H2 C Total

(vapour)
Initial moles 0 0 1 1 0 2
Equilibrium δ −0,25 ε 1-δ 1-ε 0.25 1.75
O balance δ −0,25 ε 2(1 - δ) - - 2

Note: We need the total mol vapour in order to calculate the mol fractions in the vapour phase!
We have also used, as before, carbon and hydrogen mass balances to calculate the conversions
indicated in the Table above. An oxygen (O) balance:

Oxygen in feed = oxygen in equilibrium mixture ⇒ 2 = 2(1-δ) + ε + δ - 0.25

⇒ δ = ε - 0.25
pCO2 p H 2 y CO2 y H 2 (1 − ε ) (1 − δ ) (1 − ε ) (1,25 − ε )
At equilibrium: K IV = = = =
pCO p H 2O y CO y H 2O (δ − 0,25) ε (ε − 0,5) ε

Rewrite in the form of a quadratic equation, and using KIV = 3.1623 yields:
ε = 0.6748 ⇒ δ = 0.4218
The equilibrium pressure is calculated from Reaction I for carbon deposition.
y co2 (δ − 0,25) 2
KI = P = P = 1.9953
y CO2 1,75 (1 − δ )

Substituting δ = 0.4218 yields a value for the equilibrium pressure P = 65.69 atm.

10
THERMODYNAMICS (2020) TUTORIAL 4

Q-4.1. Note that ice floats on liquid water. Based on this fact, explain whether the application
of pressure will reduce or increase the melting point of ice. Discuss the implications for ice-
skating.

Answer Q-4.1. Consider the transition (“fusion”): H 2 O (solid) = H 2 O (liquid)

dP ∆H f
The Clapeyron equation reads: =
dT T ∆V f

The fact that ice floats on water implies that water has a higher density, i.e. the molar volume
is smaller. Thus ∆V f < 0. Since the heat of fusion is always positive (∆H f > 0), it means that the
right hand side is a negative quantity and hence also dP/dT. The melting point is therefore
decreased by application of pressure. In fact, this very fact makes ice-skating possible. The
pressure exerted by the skate blades on the ice causes it to melt. The resultant liquid film
provides the required lubrication.

Q-4.2. Phosphine exists in three allotropic forms, known as the α, β and γ forms. The α and
β forms are in equilibrium at 49,43 K and the α and γ forms at 30,29 K. Obtain the molar heat
of formation for the γ form changing to the α form at 30,29 K using the following data:
(i) The entropy of the α form at 49,43 K is 34,03 J/(mol.K)
(ii) The entropy change in heating the γ form from 0 K to 30,29 K is 11,22 J/(mol.K)
(iii) The entropy change in heating the α form from 30,29 to 49,43 K is 20,1 J/(mol.K)

Answer Q-4.2.

Τ, Κ
β
Sα = 34,03 J/mol.K
49,43

α
∆Sα = 20,10 J/mol.K

30,29 Sα = 34,03 –20,1

= 13,98 J/mol.K

γ ∆Sγ = 11,22 J/mol.K

Sγ = 0 J/mol.K

1
 At 0 K S γ = 0 J/mol.K (Third law of Thermodynamics). Therefore S γ = 11,22 J/mol.K at
30,29 K

At 30.29 K:
• S α = 34,03 – 20,1 = 13,98
• α in equilibrium with γ-phase
• Thus ∆G α→γ = 0
• ∆H α→γ = T ∆S α→γ = T (S α - S γ ) = 30,29(11,22 - 13,98) = -83,6 J/mol

Q-4.3. In the vicinity of the triple point the vapour pressure of liquid ammonia (in
atmospheres) is represented by
3063
n P = 15.16 −
T
This is the equation of the liquid-vapour boundary curve in a P-T diagram. Similarly, the
vapour pressure of solid ammonia is
3754
n P = 18.70 −
T
(a) What are the temperature and pressure at the triple point?
(b) What are the latent heats of sublimation and vaporization?
(c) What is the latent heat of fusion at the triple point?

Answer Q-4.3.

Solid liquid
n P = 15,16 – 3063/T

Vapour

n P = 18,70 – 3754/T

2
In a narrow temperature range the Clausius-Clapeyron equation can be used for the vapour
pressure vs. temperature curve. For equilibrium between a condensed phase and a vapour with
a much greater molar volume it can be integrated to the form:
n P = C – ∆H/RT

The given vapour pressure equations have this functional form and this implies that:
∆H vap = 3063 R
∆H sub = 3754 R

At the triple point: ∆H sub = ∆H vap + ∆H melt ⇒ ∆H melt = 711 R

In addition, the vapour pressure curves intersect at the triple point. Therefore

n P vap = n P sub
or
15,16 – 3063/T = 18,70 – 3754/T

i.e. T trip = 195,2 K and P trip = 0,588 atm

Q-4.4. Trouton’s rule states that, for many organic compounds, ∆H vap /T b ≈ 88 J/mol.K. Assume
that a particular liquid obeys Trouton’s rule and that its vapour pressure is given by the
Clausius-Clapeyron equation. Calculate the vapour pressure at a temperature equal at one-half
the normal boiling point in atmospheres.

Answer Q-4.4.

The integrated form of the Clausius-Clapeyron equation reads:

P2 ∆H vap 1 1 
n =  − 
P1 R  T1 T2 

Let P 1 = 1 atm and T 1 be the normal boiling point. According to Trouton’s rule we then have
that ∆H vap = 88 T 1 J/mol. With T 2 = T 1 /2 we obtain:

P2 88 T1  1 2
n =  − 
1 atm R  T1 T1 

Solving yields: P 2 = 2,53 x 10-5 atm!

3
THERMODYNAMICS (2020) TUTORIAL 5

Q-5.1. A cleaning solution is to be manufactured from equal masses of acetone (1) and
dichloromethane (2), both at 298 K. If these compounds are mixed adiabatically at a pressure
of 1 atm with the addition of negligible stirring work, what is the temperature of the cleaning
solution formed?

Data: Heat capacities: Cp 1 = 2173 J/kg K & Cp 2 = 1193 J/kg K

Heats of mixing at 1 atm:

T/[K] 293 298 303

∆H/[kJ/kg] 12,468 12,38 12,292

1
Answer Q-5.1.

The process occurs adiabatically and at constant pressure. We show it schematically in the
Figure below.

Mixture at Tf
3
Real path

Heating
at constant
pressure

Pure (1) & (2)

at 298 K Mixture at 298 K
1 mixing at 298 K 2

The process may be considered to be mechanically reversible. It is then convenient to assume

that it occurs in two steps as indicated above. Since the process occurs adiabatically:

∆H 13 = ∆H 12 + ∆H 23 = 0

At 298 K: ∆H 12 = (m 1 + m 2 ) ∆H mix

Tf
∆H 23 = ∫298
(m1 + m2 ) C p dT ≈ (m1 + m2 ) C p (T f − 298)

Now, for any property change on mixing: ∆M = M - Σx i M i

This also holds for the heat capacity. Thus
∆Cp = Cp - Σx i Cp i
Rearranging
Cp = ∆Cp + Σx i Cp i

∂ ∆H  − 12 292 − (−12468)
Now ∆C p =  ≈ ≈ 17,6 J / kg
∂T  P , x 303 − 293

Rearrange and solve for T f :

∆H mix
T f = 298 − = 305,3 K
∂∆H
x1Cp1 + x 2 Cp 2 +
∂T

2
Q-5.2. Liquid phase non-ideal behaviour of a particular system is described by the Van Laar
expression for the excess Gibbs free energy:

GE A12 A21
=
x1 x 2 RT A21 x1 + A12 x 2

(i) Show that the activity coefficient for component 1 is given by:
2
 A12 x 2 
n γ 1 = A21  
 A12 x1 + A21 x 2 

Answer Q-5.2.

(i) For convenience we set A 12 = A 1 and A 21 = A 2 .

GE A1 A2 x1 x 2 nG E A1 A2 n1 n2
= and =
RT A2 x1 + A1 x 2 RT A2 n1 + A1 n2

∂  nG E  ∂  A1 A2 n1 n2   n ( A n + A1 n2 ) − n1 n2 A2 
n γ 1 =   T , P ,n2 =   T , P ,n2 = A1 A2  2 2 1 
∂n1  RT  ∂n1  A2 n1 + A1 n2   ( A2 n1 + A1 n2 ) 2 
2
 A1 n2 n2  A12 x 22  A1 x 2 
n γ 1 = A1 A2  2 
= A = A2  
 ( A2 n1 + A1 n2 )  ( A2 x1 + A1 x 2 )  A2 x1 + A1 x 2 
2 2

2
 A2 x1 
Similarly: n γ 2 = A1  
 A2 x1 + A1 x 2 

(ii) Show that the Van Laar activity coefficients can be written in the following linear form:

n γ 1 = c1 + c 2 n γ 2

Express the constants c 1 and c 2 in terms of A 12 and A 21 .

3
 n γ 1 A1 x 2 n γ 2 A2 x1
(ii) Note that = and =
A2 A2 x1 + A1 x 2 A1 A2 x1 + A1 x 2

n γ 1 n γ 2 A2 x1 A1 x 2
Adding these two equations: + = + = 1
A2 A1 A2 x1 + A1 x 2 A2 x1 + A1 x 2

A1
Rearranging yields the required relationship: n γ 1 = A1 − n γ 2
A2

(iii) Use the above expression to determine whether the Van Laar activity coefficient model fits
the VLE data for the acetone (1) – cyclohexane (2) system obtained at 45°C and shown in the
Table below. If it does, determine the coefficients A 12 en A 21 .

P/mmHg 399.3 488.2 518.8 537.1 546.8 556.2 563.2 560.9 548.9
x1 0.094 0.197 0.2810 0.384 0.491 0.598 0.7 0.771 0.895
y1 0.458 0.587 0.629 0.658 0.68 0.704 0.734 0.759 0.851
Note:
P 1 sat = 512,38 mmHg and P 2 sat = 224,9 mmHg

4
(iii) The data provided was evaluated using the Excel spreadsheet programme. The results are
given in the Tables and Figure below. The figure shows that a reasonably straight line is
obtained when the square root of n γ 1 is plotted against the square root of n γ 2 . The Van Laar
model constants were calculated from the slope and intercept of the least squares fitted straight
line fitted through the data. It is clear that

P/mmHg x1 y1 γ1 γ2 (ln γ 1 )0.5 (ln γ 2 )0.6

224.90 0.000 0.000
399.3 0.094 0.458 3.797 1.062 1.155 0.246
488.2 0.197 0.587 2.839 1.116 1.022 0.332
518.8 0.281 0.629 2.266 1.190 0.905 0.417
537.1 0.384 0.658 1.796 1.326 0.765 0.531
546.8 0.491 0.680 1.478 1.529 0.625 0.651
556.2 0.598 0.704 1.278 1.821 0.495 0.774
563.2 0.700 0.734 1.153 2.220 0.377 0.893
560.9 0.771 0.759 1.078 2.625 0.273 0.982
548.9 0.895 0.851 1.019 3.463 0.136 1.115
512.38 1.000 1.000
slope -1.152
Intercept 1.3995
A 21 =ln γ 1 ∞ = 1.4766
A 12 =ln γ 2 ∞ = 1.9587

A1
c1 = A1 and c2 = −
A2

5
 1.2

Slope = -1,1517
1.0 Intercept = 1,3995

0.8
Sqrt(ln γ1)

0.6

0.4

0.2

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

Sqrt(ln γ2)

Q-5.3. The excess Gibbs free energy for a binary mixture is defined as

GE
= x1 n γ1 + x2 n γ 2
RT
(a) Show that the infinite dilution activity coefficient is given by

GE
n γ1∞ = lim
x1 → 0 x1 x2 RT

Hint: Use the Gibbs-Duhem relationship: x1 d n γ 1 + x 2 d n γ 2 = 0

Answer Q-5.3.

GE n γ 1 n γ 2 n γ 2
(a) lim = lim + lim = ln γ 1∞ + lim
x1 →0 x1 x 2 RT x1 →0 x2 x1 →0 x1 x1 →0 x1

n γ 2
So we just have to prove that lim =0
x1 →0 x1

This limit is an indeterminate form of the type 0/0. Such limits can be determined by applying
L’Hospitals rule: Differentiate both the numerator and the denominator with respect to x 1 :

6
  dn γ 2 
 
n γ 2  dx1  dn γ 2
lim = lim = lim
x1 →0 x1 x1 →0  dx1  x1 →0 dx1
 
 dx1 

d n γ 2 x d n γ 1
From the Gibbs-Duhem equation one obtains: = − 1
dx1 x 2 dx1

n γ 2  x dn γ 2 
⇒ lim = lim  − 1  = 0
x1 →0 x1 →0
x1  x 2 dx1 

d n γ 1
Because is finite in the limit where x 1 → 0.
dx1

GE
⇒ n γ 1∞ = lim
x1 →0 x1 x 2 RT

(a) Determine n γ 1 for the Bruin model

GE − x1 x 2 n (Λ 12 Λ 21 )
=
RT (x1Λ12 + x2 )(x1 + x2 Λ 21 )
(b) Bruin Model:

GE − n (Λ 12 Λ 21 ) n (Λ 12 Λ 21 )
n γ ∞
= lim = lim =−
1
x1 →0 x1 x 2 RT x1 → 0 (x1Λ12 + x2 )(x1 + x2 Λ 21 ) Λ 21

Q-5.4. It has been stated [WH Prahl, Chem. Eng., 118 (Oct. 22, 1979)] that binary vapor-liquid
equilibrium data can be well fitted by relating the relative volatility a to the liquid mol fraction
by the following expression:

n α12 = a + b x 1 + c x 1 z

where x 1 is the liquid mole fraction of the more volatile component (P 1 ° > P2°)
in a binary mixture and a, b, c, and z are empirically determined parameters.

The relative volatility is defined as:

y1 x1
α12 =
y2 x 2
In fitting this equation (α12 vs. x1) to isothermal vapor-liquid equilibrium data, does the Gibbs-
Duhem equation allow all these parameters (a, b, c, and z) to be independently determined?
Explain.

7
Answer Q-5.4.

The vapour-liquid equilibrium relationship is: Φ i yi P = γ i xi Pi *

y i γ i Pi *
This can be rearranged to: =
xi Φ i P

γ   Φ 2 P1* 
For a binary system: α 12 =  1   
* 
γ2   Φ 1 P2 

γ   Φ P* 
n α 12 = n  1  + n  2 1* 
γ2   Φ 1 P2 

γ   Φ P* 
n  1  = n α 12 + n  1 2* 
γ2   Φ 2 P1 

The Redlich-Kister consistency test states:

1 γ 
∫ 0
n  1
γ2
 dx1 = 0

Thus

1 γ  1   Φ 1 P2* 
∫ n  1  dx1 = ∫ a + b x1 + c x1 + n   dx1
z
* 
0
γ2  0
 Φ
 2 1 
P

1
  Φ P*   b c Z +1 
=  a + n  1 2*   x1 + x12 + x1 
  Φ 2 P1   2 z +1  0

 Φ P*  b c
= a + n  1 2*  + + = 0
Φ
 2 1 
P 2 z + 1

This equation only contains the adjustable parameters a,b,c and z. Thus the Redlich-Kister
equation imposes a constraint and these parameters are not independent: only three of them are
truly adjustable.

Note also that an additional requirement is that z > 1 otherwise the integral will be undefined.

8
Q-5.5. In 1895 Margules suggested that the activity coefficients γA and γB of a binary solution
could, at any given temperature, be presented by a power series of the form:
1 1
n γ A = α 1 x B + α 2 x B2 + α 3 x B3 + ........
2 3

1 1
n γ B = β 1 x A + β 2 x A2 + β 3 x 3A + ........
2 3

Consider the situation where the variation in the activity coefficients can be represented by
series truncated after the quadratic terms.

5.5.1. Show that if these equations are to hold over the entire composition range, then
α1 = β1 = 0.
5.5.2. Show that in addition thermodynamic consistency also requires that α2 = β2 also holds.

5.5.3. The activity coefficient for component B in a binary mixture was determined
experimentally and fits the equation:
1 1
n γ B = x A2 − x 3A
2 3
Determine the limiting activity coefficients ( n γ B∞ and n γ A∞ ) for both components.

Answer Q-5.5: Series truncated after the quadratic terms.

1 1
Thus we have n γ A = α 1 x B +α 2 x B2 and n γ B = β1 x A + β 2 x A2
2 2
The Gibbs-Duhem equation states: x A dn γ A + x B dn γ B = 0
(
For our model we have: dn γ A = α 1 + α 2 x B dx B ) and
( )
dn γ B = − β1 + β 2 x A dx B
Substituting into Gibbs-Duhem and simplifying:

x A (α 1 + α 2 x B )dx B − x B (β1 + β 2 x A )dx B = 0

or
− β1 + (α 1 + β1 + α 2 − β 2 ) x A − (α 2 − β 2 ) x A2 = 0
This equation must hold for all xA. This implies that the coefficients of this quadratic equation
must all vanish. Therefore:
• β1 = 0
• α2 = β2
• α1 = 0

This answers the first two questions.

9
 1 2 1 3
Limiting activity coefficients for system A-B with: n γ B = xA − xA
2 3

1 1 1
It is immediately clear that n γ B∞ = − =
2 3 6

We derived in the text the equation:

x A =1 γ B∞  xB 
∫ x A =0
d n γ A = − ∫ γ B =1

 xA
 dn γ B = n γ A∞


Now ( )
dn γ B = x A − x A2 dx A = x A (1 − x A ) dx A = x A x B dx A

Substitute and integrate:

1
x A =1 xB =0 x3  1
n γ ∫ ∫
∞
A = − x dx A =
2
B
2
x dx B
B = B =
x A =0 x B =1 3 0 3

10
THERMODYNAMICS (2020) TUTORIAL 6

Q-6.1. The following relationship between fugacity and pressure has been proposed:
f = P + αP2 with α = α(T) only
(a) Find the equation of state (EOS) conforming to this equation.
(b) Is the EOS realistic at low pressures?
(c) What is the predicted value of the compressibility factor at high pressures?
(d) Express the coefficient α in terms of a limiting value for the variation of the
compressibility factor z with pressure.

Answer Q-6.1.

f  1 dP
The fugacity coefficient is defined as: n ϕ = n   = ∫ ( z −1)
P 0 P
f α
n ϕ = n   = n (1 + αP ) = ∫
1 1 dP
dP = ∫ ( z −1)
0 1 + αP
P 0 P
z −1 α
⇒ =
P 1 + αP
α
⇒ z = 1 +
1 + αP
A realistic EOS would reduce to the ideal gas law at low densities, i.e. very low pressures and
should predict a bounded asymptote as P tends to infinity:

• Low P limit: lim z = 1 (Agrees with ideal gas law)

P →0

• High P limit: lim z = 2 (Bounded!)

P →∞

This shows that the EOS is realistic in theory.

∂z  ∂  1+ 2α P  2 α (1 + α P) − α (1 + 2 α P) α
 =   = =
∂ P T ∂ P  1+ α P  (1 + α P) 2
(1 + α P) 2
∂z 
⇒ α = lim 
P →0 ∂ P  T

1
Q-6.2.

(a) Show that if the volumetric behaviour of a gas is described by the truncated virial
expression
BP BP
z = 1+ then n ϕ =
RT RT
(b) For a binary mixture B = B 11 y 1 + 2B 12 y 1 y 2 + B 22 y 2 noting that n ϕ̂ i is a partial molal
2 2

quantity, i.e. that

∂ (n n ϕ )
n ϕˆi =
∂ ni

Show that: n ϕ = ∑ y i n ϕˆ i and n ϕˆ1 =

BP
RT
(B11 + δ 12 y 22 )
i

where δ 12 = 2B 12 – B 11 – B 12 .

Answer Q-6.2.

Part (a) deals with the fugacity of a pure gas and part (b) with that of a gas mixture. We will
solve for the latter first and obtain the answer for the pure gas by simply setting the mol fraction
for one of the components equal to unity.
Recall that in general: dµ i = RT d n fˆi
And that for an ideal gas: dµ i )IG = RT d n ( y i P)
Integrate both expressions and subtract:
 fˆ 
µ i − µ i , IG = Γ(T ) + RT n fˆi − (Γ(T ) + RT n ( y i P) ) = RT n  i  = RT n ϕˆ i
 
 yi P 
(Note that G(T) is a temperature dependent integration constant)
Reduced properties are defined as the property value relative to the corresponding property
value for the ideal gas. Thus:
G R = G − G IG = ∑ µ i y i − ∑ µ i , IG y i = ∑ (RT n ϕˆ i ) y i
This shows that RT n ϕ̂ i is the partial molal quantity corresponding to the reduced Gibbs free
∂
energy G R . Therefore RT n ϕ̂ i =
∂ ni
(n G R )T , P ,n j ≠ ni

GR P VR P dP
Now we know that
RT
= ∫0 RT
dP = ∫ 0
( z −1)
P

∂  nG R  ∂ P dP P dP
n ϕ̂ i = 
∂ ni

 RT  T , P ,n j ≠ ni ∂ ni
= ∫ 0
( n z − n)
P
= ∫ 0
( z i −1)
P
∂
where zi = (n z )T , P ,n j ≠ ni
∂ ni

2
 BP
The simplest expression for z is the one given by the truncated virial EOS: z = 1 +
RT
∂
zi = (n z )T , P ,n j ≠ ni = ∂  n + n B P  = 1+
P ∂
(n B )T , P ,n j ≠ ni
∂ ni ∂ ni  RT  T , P ,n j RT ∂ ni
Substituting leads, for a binary system, to:

1 P ∂ (n B) P ∂
∫
n ϕˆ1 = dP = (n B )T , P ,n2
RT 0 ∂ n1 RT ∂ n1
From statistical thermodynamics we know that B is a quadratic function of composition:
Bmix = ∑∑ y i y j Bij
i j

For a binary mixture the expression is:

B = B11 y12 + 2 B12 y1 y 2 + B22 y 22
= B11 y1 (1 − y 2 ) + 2 B12 y1 y 2 + B22 y 2 (1 − y1 )
= B11 y1 + δ 12 y1 y 2 + B22 y 2
Where δ 12 = 2 B12 − B11 − B22
Now n B = n1 B11 + n2 B22 + n1 n2δ 12
∂ (n B) n δ nn δ
= B11 + 2 12 − 1 22 12 = B11 + δ 12 ( y 2 − y1 y 2 ) = B11 + δ 12 y 2 (1 − y1 ) = B11 + y 22δ 12
∂ n1 n n

⇒ n ϕˆ1 =
P
RT
(
B11 + y 22 δ 12 )
B11 P BP
If the gas is pure component (1), then y 2 = 0 and we obtain: n ϕ 1 = or n ϕ =
RT RT

3
Q-6.3. Calculate the vapour phase corrections (Φ i ) for the chloroform (1) + ethanol (2)
system at temperature T = 55°C, pressure P = 63,76 kPa and y 1 = 0,558 using the truncated
virial equation of state and given the following data:

• B 11 = -963 cm3/mol; B 12 = 52 cm3/mol; en/and B 22 = -1523 cm3/mol.

• At 55°C: P 1 sat = 82,37 kPa and P 2 sat = 37,31 kPa.

Answer Q-6.3.

ϕˆ1  V1 L ( P − P1sat ) 
The vapour phase correction is Φ1 = sat exp − 
ϕ1  RT 
The saturated liquid volumes can be estimated from the critical properties (obtainable
fromAppendix B in Smith and Van Ness) and using the Rackett equation: V1sat = Vc ,1 zc(1,1−T
0 , 2857
r ,1 )

Other equations required are:

n ϕ1sat =
B11P1sat
RT
and n ϕˆ1 =
P
RT
(
B11 + y 22 δ 12 )
The calculations are summarized in the Table below.

Component Tc, K V c , cm3/mol zc Tr V sat , cm3/mol

chloroform (1) 536.4 239 0.293 0.612 93.66
ethanol (2) 513.9 167 0.240 0.639 57.45

Component B ii , cm3/mol δ ij , cm3/mol P i sat ϕ 1 sat ϕ̂ i Φi

chloroform (1) -963 52 82.37 0.9713 0.9780 1.0075
ethanol (2) -1523 52 37.31 0.9794 0.9654 0.9852

4
Q-6.4. Consider the solubility of a solid (2) in a gas (1) as a function of pressure P at some fixed
temperature T. Assuming that the vapour phase nonideal behaviour is described by the virial
equation truncated after B, show that there is a minimum solubility of (2) in the gas given by:

− 2 B12 )P2sat
(B11
y 2 )min = − c P )at minimum = −
RT
and
RT B11 + 2B12
Where c is a constant depending only on the units used. Note: Neglect ϕ 2 sat and the Poynting
correction for the solid!

Answer Q-6.4.

Consider the solubility of a pure solid (2) in gas (1). The equilibrium condition can be
expressed as: fˆ2vap = fˆ2solid

Now fˆ2vap = y 2 P exp 

 P
(B22 + y12 δ 12 )

 RT 
ˆf solid = γ x f o = f o ≈ P sat (Because the solid is a pure crystalline substance)
2 2 2 2 2 2

 P
( 
)
sat
P
⇒ y2 = 2
exp − B22 + y12 δ12 
P  RT 
Usually the solubility of the solid would be very low. So, to a first approximation, y 1 = 1.
Now δ 12 = 2 B12 − B11 − B22 and therefore B22 + δ12 = 2 B12 − B11
 P
Thus we finally have: y2 =
P2sat
exp − (2 B12 − B11 )
P  RT 
exp[− b P ]
a
This can be simplified to: y2 =
P
Where, at constant temperature, a = P 2 sat and b = (2B 12 – B 11 )/RT are constants.
To find the extremum we differentiate with respect to pressure and set the result equal to zero:
= − 2 exp[− b P ] − exp [− b P ] = 0
d y2 a ab
dP P P
e ( B11 − 2 B12 ) sat
y 2 )min =
1 RT
Minimum solubility at: P = − = and P2
b B11 − 2 B12 RT
(where e = 2,7128…)

5
Q-6.5. A gas-phase dimerization 2 A = B occurs in the presence of a non-volatile solvent.
Starting amounts are one mole each of A and the solvent S. Vapour-liquid equilibrium constants
are K A = 5 and K B = 0,5. The chemical equilibrium constant is K y = 0,08. Find:
(i) The fractional conversion of A
(ii) The amounts and composition of the two phases
Hint: The fractional conversion is expected to be in the range 0,3 to 0,5.

Answer Q-6.5. Basis: 1 mol A in feed

Reaction summary: 2A = B
Initial mol 1 0 Mol
Mol at equilibrium: 1 – 2ε ε

The following equations describe the situation:

Mass balances: Vapour
• Vapour phase: y A + y B = 1 yA and yB
• Vapour phase: x A + x B + x S = 1
y
Chemical equilibrium: K y = B2 = 0,08
yA Liquid
Vapour-pliquid equilibrium: xA, xB and xS
yA = K Ax A = 5 x A
y B = K B x B = 0,5 x B

The vapour compositions can be determined immediately combining the mass balance and the
chemical equilibrium expressions. This yields the following quadratic equation:
0,08 y A2 + y A = 1
Solving gives as only positive root: y A = 0.9307
From the vapour mass balance: y B = 1 - y A = 0,0693
The VLE equations allow us to determine the liquid mol fractions for components A and B:
x A = y A /5 = 0,1861 and x B = 2 y B = 0,1386

The liquid phase mass balance yields x S : x S = 1 - x A + x B = 0,6753

But this corresponds to 1 mol of solvent S. Therefore N liquid = 1/0,6753 = 1,4809.
Thus N A,liquid = x A N liquid = 0,2757 mol and N B,liquid = x B N liquid = 0,2052 mol
Recall that there us a total of ε mol B present in the reactor. Consider the vapour phase:
N A,vapour = 1 - 2ε - N A,liquid and N B,vapour = ε - N B,liquid

6
Thus N vapour = N A,vapour + N B,vapour = 1 - ε - N A,liquid - N B,liquid
N B , vapour ε − N B ,liquid
yB = = = 0,0693
N vapour 1 − ε − N A,liquid − N B ,liquid

Solving for ε yields: ε = 0,2256

The conversion of A equals twice the value of ε, i.e. the conversion is 45,1%

Q-6.6. Derive the following expression for the fugacity of a van der Waals fluid

b 2a  a (V − b ) 
n ϕ = − − n  1 − 
V − b RTV  RTV 2 

dz dP dV
Hint: First derive the expression = +
z P V

Q-6.6. Derive the following expression for the fugacity of a van der Waals fluid

b 2a  a (V − b ) 
n ϕ = − − n  1 − 
V − b RTV  RTV 2 

The solution is given in the scanned document

7
THERMODYNAMICS (2020) TUTORIAL 7

Q-7.1. Justify the following rule of thumb: For an equimolar liquid mixture, the vapour
composition of a binary system, in subcritical isothermal VLE, is approximately:
P sat
y1 = sat 1 sat for x 1 = x 2 = 0,5
P1 + P2
Note that this result is trivially true if Raoult’s law applies. Show that it could be approximately
true, even for non-ideal systems.

Answer Q-7.1. Assume ideal vapour and the Porter model for the activity coefficients:
nγ 1 = Ax12 and nγ 1 = Ax 22
y1 P = γ 1 x1 P1 y 2 P = γ 2 x 2 P2
sat sat
and
y1 γ 1 x1 P1 sat
P
sat

y1 = = = sat 1
y1 + y 2 γ 1 x1 P1 + γ 2 x 2 P2
sat sat
P1 + ϕ P2
sat

γ 2 x2
2
x 2 e Ax 1

where ϕ = = = 1 at x1 = x 2 = 0,5 !
γ 1 x1
2
x1e Ax 2

Q-7.2. Sulphur hexafluoride as dielectric. Vapour sulphur hexafluoride, at pressures of about

1600 kPa, is used as a dielectric in large primary circuit breakers for electric transmission
systems. As liquids, SF 6 and water are essentially immiscible, and it is therefore necessary to
specify a low enough moisture content in the vapour SF 6 so that if condensation occurs in cold
weather a liquid-water phase will not form first in the system. For a preliminary determination,
assume the vapour phase an ideal gas and use the equations for vapour pressure given in Table
1.
(a) Sketch the phase diagram for the H 2 O(1)/SF 6 (2) system at 1600 kPa and indicate all the
phases present. Calculate and show all the important temperatures!
(b) Estimate the maximum allowable moisture content of the SF 6 vapour that will prevent the
formation of liquid water.

Table 2: Vapour Pressure Correlations

(1) H 2 O: n P 1 * = 19,1478 – 5363,70/T With P in kPa and T in Kelvin.
(2) SF 6 : n P 2 * = 14,6511 – 2048,97/T with P in kPa and T in Kelvin.

1
Answer Q-7.2. Sulfur hexafluoride as dielectric.

9.0

Water (1) - SF6 (2)

8.9

Vapour

8.8 II
T,oC Vapour +
liquid H2O
I
Vapour
8.7
+
liquid SF6

8.6

Liquid SF6 + liquid H2O

8.5
0.0000 0.0002 0.0004 0.0006 0.0008 0.0010
y1

1 z
Non-miscible liquid phases:
(a) First find T b(1) and T b(2) at 1600 kPa
5363,70
nP1* = 19,1478 − ⇒ T b(1) = 182,56°C
T
2048,97
nP2* = 14,6511 − ⇒ T b(2) = 8,56°C
T
(b) T azeo defined by P = P1*,azeo + P2*,azeo = 1600 kPa (because immiscible liquids)
Guess T azeo until true ⇒ P1*,azeo ≅ 1,112 kPa & P2*,azeo ≅ 1598,98
T azeo = 281,69 K = 8,53°C
P* 1,112
y azeo = 1, azeo = = 0,0007 mol fraction
Ptot 1600
Water content must be kept lower than 0,07 mol% to avoid condensation of H 2 O above
8,53°C. See Figure below for the phase diagram.

2
Q-7.3. An α-olefin stream at 40°C contains 50 mole % C-5, 30 mole % C-6 and 20 mole % C-
7. With only the data in the Table available estimate:
(a) The bubble point and dew point pressures at a temperature of 40°C.
(b) Calculate the mole ratio of vapour to liquid and the composition of the equilibrium
Clearly state all assumptions made!

Table 1: Vapour Pressure Correlations with P in kPa and T in Kelvin.

1-heptene ln P 7 * = 16,134 – 4217/T
1-hexene ln P 6 * = 15,838 – 3768/T
1-pentene ln P 5 * = 15,200 – 3209/T

Answer Q-7.3.

z 5 = 0,5; z 6 = 0,3; z 7 = 0,2; Only vapour pressure data available ⇒ assume Raoult’s law
P*
applies: y1 P1 = xi Pi * ⇒ K i = i
P
z
(a) Dew Point: L = 0, V = 1 ⇒Σ i = 1
Ki
At 40°C: P7* = 14,40; P6* = 44,94 and P5* = 141,53
z z z
∑ ki = ∑ P Pi* = P∑ Pi* = 1 ⇒ Pdew = 41,5 kPa
i i i

Bubble Point: V = 0, L = 1, Σ K i z i = 1 ⇒ P bubble = Σ zi Pi* = 87,13 kPa

(b) For the flash at 60 kPa, use the flash equation to determine the compositions:
( K i − 1) zi
F = ∑ = 0
1 + ( K i −1)V
Find the root of function F using the Newton-Raphsom algorithm: V n+1 = V n - F n / Fn'
0,6794 0,0753 0,152
F = - -
1 + 1,3588V 1 − 0,251 V 1 − 0,76 V
dF ( K i − 1) 2 zi 0,942317 0,0189 0,11552
−
dV
= ∑ 1 + ( K −1)V 2 = [1 + 1,3588V ]2 +
[1 − 0,251V ] 2
+
(1 − 0,76V )2
i

Pbubble − P
Initial guess: V≅ = 0,59456
Pbubble − Pdew
zi
i Ki zi xi = yi = Ki xi
1 + ( K i − 1)V
5 2,3588 0,5 0,2736 0,6454
6 0,6749 0,3 0,3541 0,2652
7 0,240 0,2 0,3723 0,0893

With V 0 = 0,5946 V 1 = 0,60897 V 2 = 0,60887

⇒ Converges to better than 4 decimals by 2 iteration!
nd

3
 V 0,60887
V/L = = = 1,5567 →
1− L 0,3911

Q-7.4. Pure liquid species 2 and 3 are for all practical purposes immiscible. Liquid 1 is soluble
in liquids 2 and 3. One mole of each of these liquids is shaken together to form an equilibrium
mix of two phases. What are the mole fractions of 1 in the α and β phases given that:
α β
GE  α α GE 
  = 0,4 x1 x 2   = 0,8 x1β x3β
 RT   RT 

Answer Q-7.4.

GE
For = Ax1 x 2 the activity coefficients can be obtained as follows:
RT
nn
GE = nG E = A RT 1 2
n1 + n2
∂ GE  An2 An1 n2
nγ 1 = ( )  = − = Ax 2 - Ax 1 x 2 = Ax 2 (1 - x 1 ) = Ax 22
∂n1 RT  P T n n1 + n2 n1 + n2
1 1 2

Phase equilibrium:
a1α = a1β or γ 1α x1α = γ 1β x1β
[
x1α exp 0,4 x 2α
2
]= x 1
β
[
exp 0,8 x3β
2
]
There are 3 moles in total. Let phase α consist of M moles:
M −1 1
⇒ x1α = and x 2α =
M M
⇒ Phase β consist of 3 – M moles
2−M 1
x1β = and x3β =
3−M 3−M
M −1  0,4  2−M  0,8 
⇒ exp  2  = exp  2 
M M  3−M  (3 − M ) 
⇒ Solve for M by plotting LHS + RHS vs M
⇒ M ≈ 1,59 ⇒ x1α = 0,371 x1β = 0,291

4
Q-7.5. For the completely miscible binary system benzene (1) / chloroform (2) the VLE K-
values at 343 K and 1 bar are K 1 = 0,719, K 2 = 1,31. Calculate the equilibrium compositions
of the liquid and vapour phases and the fraction of the total system, which is liquid for a total
system composition of 40 mole % benzene.

Answer Q-7.5.

y1 = K1 x1 and y 2 = K 2 x 2 i.e. 1 – y 1 = K 2 (1 – x 1 ) or 1 – K 1 x 1 = K 2 - K 2 x 1
K 2 −1
⇒ x1 = = 0,5245
K 2 − K1
K ( K −1)
⇒ y1 = 1 2 = 0,3771
K 2 − K1
Basis: 1 mol total
Overall mol balance: L+V=1
Component 1 mol balance: x 1 L + y 1 V = z 1
x 1 L + y 1 (1 – L) = z 1
z −y
L = 1 1 = 0,1551
x1 − y1

5
Q-7.6. Toluene (1) and water (2) are essentially immiscible as liquids. Determine the dew point
temperatures and the compositions of the first drop of liquid formed when a vapour mixture of
these two species with mole fractions z 1 = 0,2 and z 1 = 0,7 are cooled at a constant pressure of
101,325 kPa. What is the bubble point temperature and the composition of the last drop of
vapour in each case? Use a Temperature-composition phase diagram to illustrate your
arguments.

Table 1: Vapour Pressure Correlations

Toluene (1) ln P 1 * = 14,0098 – 3103,01/(219,79 + T) P in kPa en T in ºC
Water (2) ln P 2 * = 16,2620 – 3799,89/(226,35 + T) P in kPa and T in ºC

Answer Q-7.6.

The two processes are indicated on the Figure below. The calculations are done as follows:
b
Step 1: Calculate pure component boiling points from vapour correlations: n P = a −
c+t
P = 101,325 kPa ⇒ Tb ) H 2O = 100.0°C and Tb ) toluene = 110,6°C
Step 2: Calculate T azeo , the point where P = P1∗ + P2∗
101,325 = exp(14,0098 – 3103,01/(219,79 + T)) + exp(16,2620 – 3799,89/(226,35 + T))
⇒ T azeo = 84,3°C (bubble point for both cases)
P∗
P1∗ = 45 kPa P2∗ = 56,33 kPa ⇒ z1, azeo = ∗ 1 ∗ = 0,4441
P1 + P2
This is composition of the vapour when we boil a mixture of the two liquids at room
temperature. It is also the composition of the last bubble of vapour when we condense a mixed
vapour.

(a) The composition z 1 = 0,2 lies to the left of azeotrope, i.e. region in I. In this region
pure water is in equilibrium with vapour. Assume ideal vapour. Then
y 2 P = γ 2 x 2 P2∗ = P2∗ (When x 2 = 1, it follows that γ 2 = 1 too)
P2∗ P2∗
⇒ Dew point temperature given by z 1 = 0,2 = 1 - = 1 - ⇒ P2∗ = 81,06
P 101,325
⇒ T dew = 93,86°C
(b) The composition toluene z 1 = 0,7 lies to the right of azeotrope. Now pure toluene is in
equilibrium with vapour. The dew point is determined from y1 P = P1∗
Thus dew point is where P1∗ = y 1 P = 70,93
⇒ T dew = 98,53°C

6
 115

Toluene (1) - water (2)

110
Vapour
105

Dew points
100
T,oC
II
95 Vapour + liquid
Azeotrope toluene
I
Vapour +
90 liquid H2O

85

80
0.0 0.2 0.4 0.6 0.8 1.0
y1
z1,azeo

Q-7.7. Two Perfume formulations A and B both contain only two fragrance components each.
In each case the perfume is supplied as an equimolar mixture of the two constituents. At skin
temperatures both these mixtures form azeotropes at x 1 = 0,60 with the reduced vapour versus
liquid compositions as shown in Figure 1. Assume that the perceived smell of the perfumes
varies with the composition of the released vapour and that the most agreeable smell coincides
with the azeotropic composition. Discuss the relative merits of these perfume formulations
assuming that their smells and release rates are comparable at equivalent liquid compositions.
(Note: The reduced vapour mole fraction is defined as the one calculated by disregarding the
presence of air molecules)

7
Answer Q-7.7.

When applied to the skin perfume B will change its smell whereas perfume A will stabilise at
the azeotrope. Treating the problem as a batch distillation reveals this.
In case B the vapour is enriched with component 1 with respect to the liquid composition.
Therefore the liquid becomes progressively depleted of component 1.
In case A the reverse holds and the smell will finally stabilize at composition corresponding to
azeotrope. We conclude that formulation A is best.

1 1
Perfume A Perfume B

Reduced vapour mole fraction y1

Reduced vapour mole fraction y1

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Liquid fraction x1 Liquid fraction x1

Figure 1: Vapour-liquid equilibrium for perfume formulations.

8
 THERMODYNAMICS (2020) TUTORIAL 8

Q-8.1. At 101.325 kPa total pressure, ethyl acetate and ethyl alcohol form an azeotrope that
boils at 71.8°C and that contains 53.9 mole % ethyl acetate. Assuming that the Van Laar model
for activity coefficients is applicable, estimate the azeotropic composition and the total pressure
at a boiling point of 56.3°C. Use the vapour pressure data given below:

Component 71.8°C 56.3°C

Ethyl acetate (1) 84.92 kPa 48.53 kPa
Ethyl alcohol (2) 77.90 kPa 39.67 kPa

Answer Q-8.1. The Van Laar excess Gibbs free energy model is

GE A1 A2
= g=
E
x1 x2 (1)
RT A1 x1 + A2 x2
∂ ∂ n1n2 A1 A2 n A A ( A n + A2 n2 ) − n1n2 A1 A2 A1
=n γ 1 = ng E ) = 2 1 2 1 1
∂n1 ∂n1 A1n1 + A2 n2 ( A1n1 + A2 n2 )
T , P , n2 2

2
A1 A1 A2 x1 x2 + A2 A1 A2 x2 x2 − x1 x2 A1 A1 A2  A2 x2 
= A1  
( A1 x1 + A2 x2 )  A1 x1 + A2 x2 
2

2
 A2 x2 
n γ 1 = A1   (2)
 A1 x1 + A2 x2 
2
 A1 x1 
n γ 2 = A2   (3)
 A1 x1 + A2 x2 
−2
 Ax 
γ 1 A1 1 + 1 1 
n= (4)
 A2 x2 
−2
 Ax 
γ 2 A2 1 + 2 2 
n= (5)
 A1 x1 
Ratio eqn (1) and eqn (2):
2
n γ 1 A1  A2 x2  A2 x22
= =  
n γ 2 A2  A1 x1  A1 x12
x1n γ 1 A2 x2
= (6)
x2 n γ 2 A1 x1
Substitute eqn (6) in eqn (5)
−2 −2
 Ax   x n γ 1 
n γ 2 =
A2 1 + 2 2  =
A2 1 + 1 
 A1 x1   x2 n γ 2 
2
 x n γ 1 
=A2 n γ 2 1 + 1  (7)
 x2 n γ 2 

1
 By symmetry:
2
 x n γ 2 
=A1 n γ 1 1 + 2 
 x1n γ 1 

Note that we take ethyl acetate as component (1) as it has the highest vapour pressure.

The vapour-liquid equilibrium relationship is: Φ i yi P = γ i xi Pi *

If we assume that the vapour is ideal then Φ = 1. At the azeotrope we also have x 1 = y 1 .

γ 1 P2sat
Thus γ 1 = P / P1* and γ 2 = P / P2* and therefore also =
γ 2 P1sat

Substituting and solving for A 12 and A 21 yields: A 1 = 0.9123 and A 2 = 0.8377.

Now assume that the activity coefficients are temperature independent.

γ 1 P2sat
Guess the azeotrope x 1 until = for the new vapour pressures:
γ 2 P1sat
Result: x 1,azeo = 0.74
=
Calculate total azeotrope pressure using: P γ 1 x1 P1sat + γ 2 x2 P2sat

Result: P = 62.27 kPa

Check the azeotrope vapour compositions y1 = γ 1 x1 P1sat P

Check if y 1 = x 1
If not repeat. (Note: This problem is easily solved using the Excel solver!)

Final answer: P = 54,64 kPa and x 1,azeo = 0,4647

Q-8.2. The system benzene (1)-acetonitrile (2) forms an azeotrope at x 1 = 0,5370 at 45°C and
279 mm Hg. Assume an ideal vapour phase and estimate the Van Laar constants from this data

2
point. Compare predicted values with the following data at 45°C. The Antoine constants for
are:

Acetonitrile: A = 7,33986 B = 1482,290 C = 250,523

Benzene: A = 6.87987 B = 1196.761 C = 219.161
P,mmHg x1 y1
216.4 0.0167 0.0484
260.9 0.2047 0.3241
277.6 0.4320 0.4854
278.4 0.5855 0.5639
268.6 0.7779 0.6982
234.4 0.9753 0.9318

Answer Q-8.2. This problem is similar to Question 7.1 with respect to the determination of the
Van Laar constants: We find A 12 = 1,0291 and A 21 = 0,9671. The Figures below plot the
experimental data and the curve predicted from the fit to the azeotropic point. The calculations
were done in Excel.

Composition Pressure

1 300
Vapour fraction Benzene

250
0.8
200
0.6
P 150
0.4
100

0.2 50

0
0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Liquid fraction Benzene
Liquid fraction Benzene

The agreement between experiment and prediction is good.

3
 Activity coefficients

3.5

Activity coefficients
2.5

1.5

0.5

0
0 0.2 0.4 0.6 0.8 1
Liquid fraction of Benzene

Q-8.3. Two compounds A and B are known to form an ideal solution. A vapour mixture
containing 50 mole % of each compound is initially at 100°C and 1 atmosphere pressure. The
vapour pressures of the components at 100°C are: P A * = 1,20 atm; P B * = 1,40. This mixture is
compressed isothermally until condensation occurs. At what pressure does condensation occur
and what is the composition of the first drop of liquid that forms?

Answer Q-8.3. Assume ideal vapour. The equilibrium condition is: y A P = x A P A *

At the dew point y A = y B = 0,5 and therefore x A P A * = (1 - x A )P B *

The total pressure is P = x A P A * + (1 - x A )P B *

PB* 2 PA* PB*

Solving yields: xA = * and P = *
PA + PB* PA + PB*

Substituting the component values gives x A = 0,4615 and P = 1,292 atm

4
Q-8.4. In the system P + Q, activity coefficients can be expressed as:
n γ P = 0,5x Q 2 and n γ Q = 0,5x P 2

The vapour pressures of P and Q at 80°C are P P ° = 120 kPa and PQ° = 80 kPa. Is there an
azeotrope in this system at 80°C, and if so, what is the azeotrope pressure and composition?

Answer Q-8.4. Assume ideal vapour behaviour. If there is an azeotrope, then

yP = xP ⇒ γP = P/PP* = exp (0,5 xQ2)

yQ = xQ ⇒ γQ = P/PQ*= exp (0,5 xP2)

n (γP/ γQ) = n (γP) - n (γQ) = 0,5 [xQ2 - xP2] = 0,5 [xQ - xP] = 0,5 [1 - 2xP] = n (PQ*/ PP*)

⇒ xP = 0,5 - n (PQ*/ PP*) = 0,9055 and thus xP = 0,0945

P = xP γP PP* + (1 - xA) γQ PB* = 120,5 kPa

Yes, there is an azeotrope!

Q-8.5. Vapour-liquid equilibrium data is wanted for a particular binary system (known here as
the AB system). No data can be found in the literature. However, in checking the azeotropic
data compilation, we find that the system has been studied at 1 atm but that no azeotrope exists.
Some members of our design group claim that the nonexistence of an azeotrope implies an ideal
system, but others disagree. What do you think? If your answer is no, would it be possible to
estimate the error which might be involved in approximating the system at 1 atm as an ideal
system? The following vapour pressure data are available:

Table 1: Vapour Pressures

T /[°C] PA° / [atm] PB° / [atm]
80 1,00 0,704
100 1,42 1,00

Answer Q-8.5. The absence of an azeotrope does not necessarily imply an ideal solution.
However, it does place a limit on the magnitude of deviations. If there were an azeotrope we
could write:
γB P*
= A* = 1,42
γA PB

(Note that the vapour pressure ratio appears to be independent of temperature!)

5
(a) First consider positive deviations from Raoult’s law. In this case both the activity
coefficients are greater or equal than unity over the whole composition range. The maximum
possible value for γB = 1,42 and it occurs at xA = 1 (where γA = 1). For simplicity, choose the
Porter model for the activity coefficients: n γB = A xA2 and n γA = A xB2

Thus A = n (1,42) = 0,351

We can use this value to construct the y-x-T phase diagram at 1 atm.

(b) Negative deviations from Raoult’s law. Now the activity coefficients are less or equal
than unity over the whole composition range. The lowest possible value for γA = 1/1,42 and it
occurs at xA = 1 (where γA = 1). Again using the Porter model we now find

A = n (1/1,42) = -0,351

The Figure below shows the predicted equilibrium curves for the two cases together with the
ideal solution case.

1
A- B
0.8

0.6 1 atm
yA

0.4

Positive deviations
0.2
Negative deviations
Ideal
0
0 0.2 0.4 0.6 0.8 1

xA

6
Answer Q-8.6 Regular Solution theory
Consider a regular binary mixture in equilibrium with an ideal vapour. The excess Gibbs free
energy for the solution is given by Porters model:
E
G
= Ω x1 x 2
RT

(a) Show that the activity coefficients are given by: n γ1 = Ωx22 and n γ2 = Ωx12

∂  nGE  ∂  Ω n1 n2   n (n + n ) − n n  n22
n γ 1 =   =   = Ω  2 1 2 2 1 2  = Ω = Ω x 22
∂ n1  RT  n2 ∂ n1  n1 + n 2  n2  (n1 + n2 )  (n1 + n2 ) 2

Similar for n γ2: n γ 2 = Ω x12

.

(b) Express the relative volatility, of species 1 relative to species 2, in terms of the vapour
pressures and composition.

The vapour-liquid equilibrium relationships (assuming ideal vapour behaviour) are:

y1P = γ 1 x1P1* and y2 P = γ 2 x2 P2*

y1 x1 γ P* P *  exp (Ωx 22 ) 
Now α 12 = = 1 1* = 1*  
y 2 x2 γ 2 P2 P2  exp (Ωx12 ) 

(c) Find the limiting relative volatility in the limit of infinite dilution of component 2.

P1*
lim α 12 = exp (− Ω )
x1 →1 P2*

(d) Determine an expression for the Henry’s Law constant.

Henry’s law may be expressed as: y1 = k1x1

It holds in the limit of infinite dilution of component (1), i.e. in that case k1 assumes a constant
value. That means:
y
k1 = lim 1
x1 →0 x
1

( )
y1 P = γ 1 x1 P1* = x1 P1* exp Ω x 22

exp (Ω x ) =
* *
y P P
Thus lim 1 = lim 1 2
2exp (Ω ) =
1
k1
x1 →0 x x1 →0
P P
1

7
(e) Show that, if an azeotrope exists, the azeotrope composition can be expressed in terms of
Ω and the vapour pressures of the components:
The vapour-liquid equilibrium relationships (assuming ideal vapour behaviour) are:
y1P = γ 1 x1P1* and y2 P = γ 2 x2 P2*

At the azeotrope composition: y1 = x1 and y2 = x2

Thus: γ 1 = P / P1* and γ 2 = P / P2*

The Porter activity coefficient model states that:
n γ 1 = Ω x22 and n γ 2 = Ω x12
Combining these results yields:
 P  P
Ω x12 = n  *  and Ω x22 = n  * 
 P2   P1 

 P* 
Subtracting: ( )
Ω x12 − x22 = Ω ( x1 + x2 )( x1 − x2 ) = Ω (2 x1 −1) = n  1* 
 P2 
Rearranging, we obtain for the azeotrope composition:

1 1  P1* 
x1, azeo = y1, azeo = 1 + n  *  (1)
2 Ω  P2 

(f) Determine an expression for the azeotrope pressure in terms of the vapour pressures of
the components.
Ω   P1*   
2
* 2 *
[ ] 1
Pazeo = γ 2 P2 = P2 exp Ω x1 = P2 exp  1 + n  *   
*

Ω  P2   
(2)
 4 
 

(g) Show that azeotropes always occur at a maximum or a minimum in the P-x1 curve.

[ ]
P = x1γ 1 P1* + x 2 γ 2 P2* = x1 P1* exp Ω (1 − x1 ) 2 + (1 − x1 ) P2* exp Ω x12[ ]
dP
dx1
[ ] [ ] [ ] [
= P1* exp Ω (1 − x1 ) 2 − P2* exp Ω x12 − 2 Ω x1 (1 − x1 ) P1* exp Ω (1 − x1 ) 2 + 2 Ω x1 (1 − x1 ) P2* exp Ω x12 ]
dP
dx1
[ ] [ ]
= P1* exp Ω x 2 ) 2 (1 − 2 Ω x1 x 2 ) − P2* exp Ω x12 (1 − 2 Ω x1 x 2 )

dP
dx1
( )
= γ 1 P1* − γ 2 P2* (1 − 2 Ω x1 x 2 ) = 0

8
Because γ 1 P1* = γ 2 P2* at the azeotrope.

Thus the system pressure will be at a maximum or a minimum at the azeotrope point.
(h) Express the azeotrope composition in terms of the azeotrope pressure and the vapour
pressures of the components only.

P  P 
n  azeo
*
 = Ω x12 and n  azeo
*
 = Ω (1 − x1 ) 2
 P2   P1 

Divide and simplify

(
n Pazeo / P2*
=
) Ω x12
(
n Pazeo / P1* )
Ω (1 − x1 ) 2

(
n Pazeo / P1* ) =
1 − x1
=
1
−1
(
n Pazeo / P2* ) x1 x1

⇒
1
= 1 +
(
n Pazeo / P1* )
x1,azeo (
n Pazeo / P2* )
 P* 
(i) Show that the condition for an azeotrope to exist can be expressed as: Ω > n  1* 
 P2 

For an azeotrope to exist: 0 < x1,azeo < 1

1 1  P1*  1  P1* 
⇒ 0 < 1 + n  *  < 1 ⇒ 0 < 1+ n   < 2
2 Ω  P2  Ω  P2* 

1  P1* 
⇒ −1 < n   < 1
Ω  P2* 

Our component assignment convention specifies that component 1 is the more volatile
(
component. Thus we know that P1* / P2* >1 , i.e. n P1* / P2* > 0 . )
 P* 
Therefore if Ω > 0 then: − Ω < n  1*  < Ω
 P2 

 P* 
This implies that: Ω > n  1* 
 P2 

 P* 
Similarly, when Ω < 0 then: − Ω > n  1*  > Ω
 P2 

9
  P* 
This (again) implies that: Ω > n  1* 
 P2 

(j) Explain the physical implication for the system if Ω > 2 RT.

When Ω > 2 RT, the liquid phase will split into two separate phases. The equilibrium conditions
are:
γ 1,α x1,α = γ 1, β x1, β and γ 2,α x2,α = γ 2, β x2, β

(k) In practice the temperature dependence of Ω can be approximated as Ω = ω/RT. Consider

such an azeotropic binary system composed of components 1 and 2. We want to determine
the effect of the system pressure on the azeotropic composition. Assume that the vapour
pressures, in the neighborhood of the azeotropic point, are given approximately by:

n P1* = A1 − B1 / T and n P2* = A2 − B2 / T

Prove that the shift in the azeotrope point with temperature (and by implication pressure), is
given by:
x1 = a T + b

Where a and b are constants. Give the constants a and b above in terms of A1, A2, B1, B2 and
Ω.

At the azeotrope composition: y1 = x1 and y2 = x2

Thus the equilibrium relationships reduce to: γ 1 = P / P1* and γ 2 = P / P2*

γ2 P*
Thus = 1* and
γ1 P2

γ   P*  B − B2
n  2  = n  1*  = n P1* − n P2* = A1 − A2 − 1
 γ1   P2  T

γ  α 2 α
n  2  = n γ 2 − n γ 1 = [
x1 − (1 − x1 ) 2 = ]
[2 x1 −1]
 γ1  RT RT

 B1 − B2  1 R (B1 − B2 )
x1,azeo =
RT
 A − A −  + =
RT
( A − A ) − +
1
2α 2α 2α
1 2 1 2
 T  2 2

1 R (B1 − B2 )
Thus a =
R
( A1 − A2 ) and b = −
2α 2 2α

10
11