Chapter 9

THE BEHAVIOR OF
SOLUTIONS
 Chemical Potential of Species
• If we have two phases, α and β, in equilibrium with each other
at constant T and P:
α β
1 1
2 2
3 i 3
. .
. .
k k

dG = dGα + dG β
β β
dG = μαi dnαi+ μi dni
β β β
• Since dni = −dnαi , dG = (μαi −μi )dni
• If dni moles of i, one of the species, passes from α phase to β
phase: dnαi < 0
 β
• If dG<0: μαi > μi . Therefore, passage of i from α to β is
spontaneous.
β
dG=0: μαi = μi . Therefore, α and β phases are in equilibrium
with respect to i.
β
dG>0:μαi < μi . Therefore, passage of i from β to α is
spontaneous.
• For complete chemical equilibrium:
β β β β
μ1α = μ1 , μα2 = μ2 , ......., μαi = μi , .........., μαk = μk
• In the case of pure i, the change in Gibbs free energy by the
addition of 1 mole of i will be:
∆𝐺 = (𝑛 + 1)𝐺𝑖° −𝑛𝐺𝑖° = 𝐺𝑖°
• Therefore:
μ°𝑖 = 𝐺𝑖°
 Chemical Potential of Pure Ideal Gas
• At high enough T and low enough P the interactions among
the atoms or molecules in mixtures of gases are small enough
that they can be neglected, in which case the mixture can be
considered to be a mixture of ideal gases.

• For pure ideal gas i, chemical potential at 1 atm pressure is μ°i .

• If the pressure of the gas is decreased to P, keeping T constant:
P
dG = VdP = RT lnP
1
• Therefore:
∆G = μi − μ°i = RT lnP
 Mixing Process
• Phase: Part of a system that has uniform chemical composition
and physical properties (internal equilibrium).
Solution: Mixture of atoms, ions, or molecules.

• The initial state for mixing process can be viewed as a

collection of containers, each holding an arbitrary quantity of a
pure component in some specified phase form (gas, liquid,
etc.) and at the temperature and pressure of the solution to be
formed.
• This initial state of any particular component is called its
referance state for the formation of solution.
• The mixing process is the formation of a solution from its pure
components at constant temperature and pressure.
 Properties of Ideal Gas Mixtures

Pure A
Pure C
T, P Mixture of
T, P
Mixing A, B, C, D
Pure B Pure D T, P
T, P T, P

The partial pressure of each gas in the mixture is:

Pi = Xi P
The chemical potential experience by an individual component i
during mixing is:
Pi
o Pi
ΔGi = Δμi = μi − μi = Vi dP = RTln = RTlnXi
P P
 Mixtures of Real Gases
RT
• For real gases: Vi ≠
P
RT
• Then we define a function αi = Vi − ,
which reports the
P
deviation of the measured partial molar volume at a given
temperature and pressure from ideality.
• Accordingly,
Pi Pi
RT Pi fi
Δμi = αi + dP = αi dP + RTln = RTln
P P P P P

where fi is the fugacity.

• As the deviation from ideal behavior, αi approaches to zero,
fugacity of component i approaches to its partial pressure.
 Partial Molar Properties
• If a system is multicomponent and open, then in addition to
changes that may result from changing temperature and
pressure of the system, alterations can occur if material is
added to or removed from the system. Therefore,
Q’ = Q’ (T, P, n1, n2, .....nx)
where Q’ is any kind of extensive property (G’, H’,etc.)
• If the system is taken through an arbitrary infinitesimal change
in state, the change in Q’ will be:
𝑥
𝜕𝑄′ 𝜕𝑄′ 𝜕𝑄′
𝑑𝑄′ = 𝑑𝑇 + 𝑑𝑃 + 𝑑𝑛𝑖
𝜕𝑇 𝑃,𝑛 𝜕𝑃 𝑇,𝑛 𝜕𝑛𝑖 𝑇,𝑃,𝑛 ≠𝑛
𝑖 𝑖 𝑖=1 𝑗 𝑖

𝜕𝑄′
𝑄𝑖 = where i = 1, 2,.....x
𝜕𝑛𝑖 𝑇,𝑃,𝑛 ≠𝑛
𝑗 𝑖
• Therefore at constant temperature and pressure:
x

dQ′ = Q i dni
i=1

• The partial molar value of Q’, Q i , can also be defined as the

change of Q’ when 1 mol of i is added to a quantity of mixture,
which is sufficienlty large, at constant temperature and
pressure. Accordingly,

Q′ = n1 Q1 + n2 Q 2 + n3 Q 3 + ⋯ + nx Q x

Q′ = Q i ni
i=1
• If we differentiate the equation
x

Q′ = Q i ni
i=1
we will find:
x

dQ′ = (Q i dni + ni dQ i )
i=1
• However, we know that
x

dQ′ = Q i dni therefore ni dQ i = 0

i=1 i=1
• Division of this equation by n, the total number of moles of all
the components of the solution, gives
Xi dQ i = 0 Gibbs-Duhem Equation
i=1
• If we have n1 moles of component 1 (C1) and
n2 moles of component 2 (C2);
Q’ before mixing = n1 Q°1 + n2 Q°2
where Q°i is Q of pure component i
Q’ after mixing = n1 Q1 + n2 Q 2

• Therefore the change in Q caused by mixing, ΔQ′M , will be:

ΔQ′M = (n1 Q1 + n2 Q 2 ) − (n1 Q°1 + n2 Q°2 )
= n1 Q1 − Q°1 + n2 Q 2 − Q°2
= n1 ΔQM 1 + n 2 ΔQ M
2

• If ∆QM is known as a function of composition for a solution at

some temperature and pressure, then it is possible to find ΔQ i
of each component in the mixture.
 Evaluation of Partial Molar Properties from
Total Properties
• For a binary system:
dΔQM = ∆QM
1 dX 1 + ∆Q M
2 dX 2 and
ΔQM = ∆QM X
1 1 + ∆Q M
2 X2
Since X1 + X2 = 1, dX1 = −dX2 :
dΔQM = (∆QM
2 − ∆Q M
1 )dX 2

d∆Q M
∆QM
1 = ∆Q M
2 −
dX2
d∆QM
Therefore ∆QM = ∆QM
2 − X1 + ∆QM
2 X2
dX2

d∆QM d∆QM
∆QM = ∆QM
2 − X1 and ∆QM
2 = ∆QM + X1
dX2 dX2
Graphical Determination of Partial Molar
Properties
𝑄𝐵
𝛥𝑄𝐵𝑀
𝑄𝐵°

𝑄𝑀
𝛥𝑄′𝑀

𝑄𝐴
𝛥𝑄𝐴𝑀
𝑄𝐴°

𝑋𝐵

𝛥𝑄𝑀 = 𝑋𝐴 𝛥𝑄𝐴𝑀 + 𝑋𝐵 𝛥𝑄𝐵𝑀

 Graphical Determination of Partial Molar
Properties
𝑑∆𝑄𝑀
∆𝑄2𝑀 = ∆𝑄 𝑀
+ 𝑋1
𝑑𝑋2
𝑑∆𝐺 𝑀
∆𝐺𝐴𝑀 = ∆𝐺 𝑀 + 𝑋𝐵
𝑑𝑋𝐴

∆𝐺 𝑀 ∆𝐺 𝑀 = 𝑝𝑞 = 𝑜𝑟
𝑋𝐵 = 𝑟𝑞

𝐺𝐵
𝑑∆𝐺 𝑀 𝑟𝑠
=
𝑑𝑋𝐴 𝑟𝑞

𝑟𝑠
𝐺𝐴 ∆𝐺𝐴𝑀 = 𝑜𝑟 + 𝑟𝑞
𝑟𝑞
= 𝑜𝑠
 Chemical Potential in a Condensed Phase

g,st g
i (gas) μi i,j,k,l (gas) μi
Pi° Pi

Pure i μst
i μαi
(condensed state) i,j,k,l (α)

T, P T, P
Standard state

g,st g
μst
i = μi = μ°i + RT lnPi° μαi = μi = μ°i + RT lnPi
𝑃𝑖
μ°i = μst
i − RT lnPi
°
μαi = μst
i + RTln P°i

μαi = μst
i + RTlna i
 Activity
• The activity of i in a solution, with respect to pure i, is the ratio
of the partial pressure of i exerted by the solution to the
saturated vapor pressure of pure i at the same temperature.
𝜕𝑄′
Since Qi =
𝜕ni T,P,n ≠n
j i
and chemical potential is the measure of the rate of energy
change of a system with the change in the amount of ith
component where all other variables being held constant:
μi = Gi
and ∆μi = μi − μoi = RT lnai = RT ln(γi Xi )
where ai is the activity of component i and
γi is the activity coefficient of i
in a solution at a given temperature, pressure and composition.
Both ai and γi are unitless.
• If γi = 1, then ai = Xi , the solution is ideal.
 Raoult’s Law and Henry’s Law
• If small quantity of B (solute) atoms are added to A (solvent)
to form a concentrated solution of A, A atoms, in general, will
experience the same surroundings as in pure state; only a very
small fraction of A atoms will have neighbours of B.
Accordingly, A atoms act as if they were in an ideal solution.
This behavior gives rise to an experimentally observed
limiting law that applies to all solutions, Raoult’s law:
lim 𝑎𝑖 = 𝑋𝑖
𝑋𝑖 →1
• In the same composition range, every B atom is completely
surrounded by A atoms forming a dilute solution of B. In this
range, the average properties of B are proportional to their
concentration. This proportionality is characterized by a
constant that is specific to A-B combination in the system.
This behavior also gives rise to an experimentally observed
limiting law that applies to all solutions, Henry’s law:
lim 𝑎𝑖 = 𝑘𝑖 𝑋𝑖
𝑋𝑖 →0
 𝑋𝐵
1
1
Henry’s law

𝑎𝐴 𝑎𝐵 𝑎𝐴

Henry’s law
Henry’s law

0 𝑋𝐴 1 0 𝑋𝐴 1

• The coefficient, 𝑘𝑖 , called the Henry’s law constant, is the

activity coefficient of solute.
• The value of this constant depends on both solute and solvent
in the system and varies also with temperature and pressure.
• If the A–B bond energy is stronger than the A–A and B–B
bond energies, then, the Henry’s law line lies below the
Raoult’s law line.
 The Behavior of Raoultian Ideal Solutions
Remember that:
• The change in chemical potential of a component during
mixing is:
∆μi = ∆Gi = RTlnai
• The Gibbs free energy of mixing, for the mixing of A and B,
was:
∆GM = XA ΔGAM + XB ΔGBM
• On ther other hand, the components of a Raoultian ideal
solution obey the relation:
ai = X i
Therefore: ΔGAM,id = RTlnXA
∆G M,id = RT(XA lnXA + XB lnXB )
 The Change in Volume for an Ideal Solution
𝜕Gi 𝜕G°i
= Vi and = Vi°
𝜕P T,X 𝜕P T,X

M
𝜕Gi 𝜕Gi −𝜕G°i M
Therefore, = = Vi − Vi° = 𝜕Vi
𝜕P T,X 𝜕P T,X

M,id
Since ∆Gi = RTlnXi ,
Molar Volume
𝑉𝐵°
∆ViM,id =0
and ∆V M,id =0
Accordingly, 𝑉𝐴°
Vi = Vi°
A 𝑿𝑩 B
 The Entropy of Formation of an Ideal
Solution
Since ∆G M,id = RT X A lnXA + XB lnXB :
𝜕∆GM
= −∆SM
𝜕T P,X

Accordingly, ∆S M,id = −R X A lnXA + XB lnXB

∆SAM,id = −RXA lnXA

∆𝐒 𝐌,𝐢𝐝
𝐌,𝐢𝐝
• Find ∆𝐇𝐢 and ∆𝐇 𝐌,𝐢𝐝

A 𝐗𝐁 B
 The Behavior of Nonideal Solutions
∆μi = ∆Gi = RTlnγi Xi = RTlnγi + RTlnXi
xs id
∆Gi ∆Gi
id
• ∆Gi , is the partial molar Gibbs free energy of mixing for an
ideal solution.
xs
• ∆Gi ,is the departure from ideal behavior of component i in
the mixture.

• If γi > 1, then ai > Xi and in the evaluation of chemical

potential, component i acts as if the solution contains more of i
than the mole fraction suggests.
xs
• In this case, ∆Gi is positive and the system is said to exhibit
a “positive departure from ideal behavior’’.
• In general, increasing T of a nonideal solution causes a
decrease in the extent to which its components deviate from
ideal behavior.
• If γi > 1, then an increase in T causes γi to decrease towards
unity, and if γi < 1, an increase in T causes γi to increase
towards unity.
• In non-ideal solutions, ∆HiM ≠ 0.
• According to Gibbs-Helmholtz equation:
M
∆Gi M
𝜕 ∆Hi
T =− 2
𝜕T P,X T
M
Since ∆Gi = RTlnγi + RTlnXi :
M
𝜕 Rlnγi ∆Hi
= − 2
𝜕T P,X T
 1 𝑑T
• Since 𝑑 = − 2:
T T
𝜕 Rlnγi M
= ∆Hi
1
𝜕
T
M
• If γi > 1, ∆Hi > 0:
Thus ∆HM , the molar heat of formation of the solution, is a
positive quantity, which indicates that the mixing process is
endothermic.
M
• If γi < 1, ∆Hi < 0:
Thus ∆HM , the molar heat of formation of the solution, is a
negative quantity, which indicates that such a solution forms
exothermically.
• Exothermic mixing in an A-B binary system occurs when the
A-B bond energy is more negative (stronger) than both the A-
A and B-B bond energies, and this causes a tendency towards
“ordering” in the solution, in which the A atoms attempt to
only have B atoms as nearest neighbors and vice versa.
• Exothermic mixing thus indicates a tendency toward the
formation of a compound between the two components.
• Conversely, endothermic mixing occurs when the A-B bond
energy is less negative than both the A-A and B-B bond
energies, and this causes a tendency toward phase separation
or “clustering” in the solution.
• In both cases the equilibrium configuration of the solution is
reached as a compromise between the enthalpy factors, which,
being determined by the relative magnitudes of the bond
energies, and the entropy factor which attempts to maximize
the randomness of mixing of the atoms in the solution.
k Ni = 0.66
 Application of the Gibbs-Duhem Relation to
the Determination of Activity
• It is often found that the activity of only one component of a
binary solution can be measured experimentally.
• In such cases, the variation of the activity of the other
component can be obtained from an application of the Gibbs-
Duhem equation:
Xi dQ i = 0
i=1
• For a binary A-B solution: ∆GM = RT XA lnaA + XB lnaB
• Therefore: XA dlnaA + XB dlnaB = 0
XB
dlnaA = − dlnaB
XA
 • If the variation of aB with composition is known, then
integration from XA = 1 to XA gives the value of lnaA at XA :
lnaB (XA =XA )
XB
lnaA at XA = − dlnaB
lnaB (XA =1) XA

As an analytical expression
𝐗𝐁 for the variation of the
𝐗𝐀 activity of B with
composition is not usually
computed, this equation is
XA = XA solved by graphical
integration. The value of
lnaA at 𝑋𝐴 = 𝑋𝐴 is equal to
XA = 1
the shaded area under the
curve.
-log aB at X A = X A -log 𝐚𝐁
• Two points are to be noticed in the given figure:
1. As XB → 1, aB → 1 lnaB → 0,
XB
XA → 0 →∞
XA
The curve exhibits a tail to ∞ as XB → 1.
2. As XB → 0, aB → 0 lnaB → −∞, −lnaB → ∞
XB
XA →1 →0
XA
The curve exhibits a tail to ∞ as XB → 0.

• Of these two points, point 2 is more serious since the

calculation of lnaA at any composition involves the evaluation
of the area under a curve, which tails to ∞. This introduces an
uncertainity into the calculations.
• The tail to ∞ can be eliminated by considering activity
coefficients instead of activities in the Gibbs-Duhem
equation.
• For the ideal binary A-B solution:
M,id M,id
X A d∆GA + XB d∆GB =0
XA d RT lnXA + XB d RT lnXB = 0
XA dln XA + XB dln XB = 0
• We already know that:
XA dlnaA + XB dlnaB = 0
• Therefore:
XA dlnγA + XB dlnγB = 0
XB
dlnγA = − dlnγB
XA
• If the variation of γB with composition is known, then
integration from XA = 1 to XA gives the value of lnγA at XA :
lnγB (XA =XA )
XB
lnγA at XA = − dlnγB
lnγB (XA =1) XA
𝐗𝐁
𝐗𝐀

XA = XA

XA = 1
-log γB at X A = X A -log γB at X A =1
-log 𝛄𝐁

Since integrated area under the curve between X𝑁i = X𝑁i and
X𝑁i = 0 is a positive quantity, l𝑜𝑔γFe is everywhere a negative
quantity, and so Fe, like Ni exhibits negative deviation from
Raoult’ law.
• The second tail to ∞ can be eliminated by the introduction of
the α-function, which for the component i is defined as:
lnγi
αi =
(1 − Xi )2
• The α-function always have a finite value since γi → 1 as
Xi → 1.
• For a binary A-B solution:
lnγA lnγB
αA = and αB =
X2B X2A

lnγA = αA XB2 and 2

lnγB = αB XA
• If we differentiate:
2
dlnγB = 2αB XA dXA + XA dαB
XB
• Since dlnγA = − dlnγB ,
XA

dlnγA = −2αB XB dXA − XB XA dαB

• If we integrate this equation from XA = 1 to XA :
XA =XA XA =XA
lnγA XA = − 2αB XB dXA − XB XA dαB
XA =1 XA =1
• Remember that:
d(xy) = ydx + xdy and xdy = d(xy) − ydx

• Therefore, we can say that:

αB (XA =XA ) αB (XA =XA ) αB (XA =XA )

XB XA dαB = d X B X A αB − αB d(XB XA )
αB (XA =1) αB (XA =1) αB (XA =1)

• If we combine two equations:

αB XA =XA αB (XA =XA )
XA =XA
lnγA XA = − 2αB XB dXA − d XB XA αB + αB d(XB XA )
XA =1
αB XA =1 αB (XA =1)
• After several integration, we will find that:
XA =XA
lnγA XA = − XB XA αB − αB dXA
XA =1

• Therefore, lnγA at XA = XA
can be found from
− XB XA αB minus the area
under the plot αB versus XA
from XA = X A to XA = 1.
 The Relationship between Henry’s and
Raoult’s Laws
• Remember that Henry’s law for the solute B in a binary A-B
solution is:
aB = k B X B
dlnaB = dlnX B
XB
• Since we have found that dlnaA = − dlnaB according to Gibbs-
XA
Duhem equation:
XB X B dX B dX B
dlnaA = − dlnX B = − =− = dlnX A
XA XA XB XA
• If we integrate this equation, we will find that:
lnaA = lnXA + lnconstant
aA = constant × X A
• Since ai =1 when X i =1, constant=1.
• Consequently, in the range of composition over which the solute B
obeys Henry’s law, the solvent A obeys Raoult’s law.
 Direct Calculation of the Integral Molar
Gibbs Free Energy of Mixing
M
d∆G
∆GAM = ∆GM + XB
dXA
• If we multiply the equation with dXA and divide by XB2 :
dX A ∆GAM dXA ∆GM + XB d∆GM
2 =
XB XB2
dX A ∆GAM −dXB ∆GM + XB d∆GM
2 =
XB XB2
dXA ∆GAM d∆GM
2 =
XB XB
• If we integrate this equation:
XA =XA M
∆G A
∆GM = XB 2 dX A
XA =0 XB
 XA
lnaA
∆GM = RT XB 2 dX A
0 XB
• For Fe-Ni system, at 1600 °C:
XFe
XNi
0
1 0.8 0.6 0.4 0.2 0
M
lnaNi
∆G = RT X Fe 2 dX Ni
-0.5 0 X Fe
-1

-1.5 ∆GM = 2.303 × 8.314 × 1873 × 0.5 × Area

-2

logaNi -2.5 ln10 R T X Fe

2
XFe
-3
lnaNi
→ −∞ as X Fe → 0
-3.5 X2Fe
-4
The uncertainty caused by the infinite tail as
-4.5
X → 0 is eliminated if the equation is used to
-5
calculate the excess Gibbs free energy.
• If the solution is ideal:
XA lnaA
∆GM = RT XB 0 dXA
X2B

Xi lnXi
∆GM = RT (1 − Xi ) 0 (1−Xi )2
dXi

lnXi
∆GM = RT (1 − Xi ) + ln(1 − Xi )
1−Xi

∆GM = RT Xi lnXi +(1 − Xi ) ln(1 − Xi )

XA ∆HM
∆HM = XB 0 2
A
dX A
XB

XA ∆SM
∆SM = XB 0 A
dXA
X2B
 Regular Solutions
• So far, we have discussed two types of solutions:
1. Ideal (Raoultian): 𝑎𝑖 = 𝑋𝑖 , ∆HiM = 0, ∆ViM = 0, ∆SiM = 𝑅𝑙𝑛𝑋𝑖
2. Nonideal: 𝑎𝑖 ≠ 𝑋𝑖 , ∆HiM ≠ 0

• Attempts to classify nonideal solutions have involved the

development of equations that describe the behavior of
hypothetical solutions. The simplest of these mathematical
models is the regular solutions.
• Hildebrand defined a regular solution as one in which relative
molar enthalpy of mixing is different from zero while relative
molar entropy is the same as that for an ideal solution:
∆HiM ≠ 0 and ∆SiM = 𝑅𝑙𝑛𝑋𝑖
• The properties of a regular solution are best examined by
means of the concept of excess functions. The excess value of
an extensive thermodynamic solution property is simply the
difference between its actual value and the value that it would
have if the solution were ideal.
G = Gid + Gxs
• If we substract the Gibbs free energy of the unmixed
components:
∆𝐺 𝑀 = ∆G M,id + Gxs
• Therefore:
Gxs = ∆𝐺 𝑀 − ∆G M,id = ∆𝐻 𝑀 − ∆HM,id − 𝑇 ∆𝑆 𝑀 − ∆SM,id
• Remember that for ideal solutions ∆HM,𝑖𝑑 = 0 and for regular
solutions, ∆𝑆 𝑀 = ∆SM,id .
• Therefore:
Gxs = ∆𝐻 𝑀
• Remember that:
∆G M,id = RT(XA lnXA + XB lnXB )
∆GM = RT(XA lnaA + XB lnaB )
∆GM = RT XA lnXA + XB lnXB + RT(XA lnγA + XB lnγB )

• Therefore:
xs xs xs
G = RT XA lnγA + XB lnγB = XA GA + XB GB

• In 1929, Hildebrand showed that in the binary A-B solution:

RTlnγA = α′ XB2
2
RTlnγB = α′ XA

• Since for nonideal solutions:

XA =XA
lnγA = − XB XA αB − αB dXA
XA =1
• If the value of α for one component, say, component B, is
independent of composition:
XA =XA
lnγA = − XB XA αB − αB dXA
XA =1

lnγA = − XB X A αB − αB XA − 1 = − XB X A αB + αB XB
lnγA = αB XB 1 − XA = αB XB2
• But we already know that:
lnγA = αA XB2
• Therefore:
αA = αB = α =constant
• Since RTlnγA = α′ XB2 ,
α′
α=
RT
• Therefore:
Gxs = RT XA lnγA + XB lnγB = RT 𝛼XA XB2 + 𝛼XB XA
2

Gxs = RT𝛼XA XB = 𝛼 ′ XA XB
• Since
𝜕𝐺 𝑥𝑠
= −𝑆 𝑥𝑠 = 0
𝜕𝑇 𝑃,𝑋

Gxs , and hence Hxs , are independent from temperature.

• Since Gxs is constant at any given composition:
xs
GA = R𝑇1 γA 𝑇1 = R𝑇2 γA 𝑇2 = 𝛼 ′ XB2
𝑙𝑛γA 𝑇1 𝑇1
=
𝑙𝑛γA 𝑇2 𝑇2
• This equation is of considerable practical use in converting
activity data for a regular solution at one temperature to
activity data at another temperature.
• The molar excess Gibbs free energy can be obtained from the
knowledge of the activity coefficient dependence of one
component on composition:
XA =XA
lnγA
∆Gxs = RT XB 2 dX A
XA =0 XB

• To sum up, for regular solutions:

xs
GA = 𝛼 ′ XB2
xs
∆HAM = GA = 𝛼 ′ XB2
∆SAM = −𝑅𝑙𝑛XA
∆GAM = 𝛼 ′ XB2 + 𝑅𝑇𝑙𝑛XA

• Therefore, if 𝛼 ′ >0, positive deviation from ideal behavior.