6.

Phase Diagrams
Gibbs Phase Rule
Formal denition of phase: A phase is a state of mat-
ter that is uniform throughout, not only in chemical
composition but also in physical state.
Denition of number of components: The number of
components is the minimum number of independent
species necessary to dene the composition of all the
phases in the system.
Equivalent denition: The number of components is
the number of distinct chemical species in the system
less the number of constraints on the concentration of
these species.
Constraints: Constraints are conditions that gives
rise to additional relations between the chemical po-
tentials or between intensive variables, for instance,
mole fractions in a given phase. Typical examples
are: (i) Each independent chemical reaction provides
one relation between the chemical potentials, namely
PChem I 6.1
the equilibrium condition

i

i
= 0 (see next Chap-
ter). (The
i
s are the stoichiometric coefcients of
the reacting species, positive for products and nega-
tive for reactants.) Chemical reactions are indepen-
dent if no reaction can be written as a combination
of the others. (ii) Material balance or stoichiometric
relations. (iii) Electroneutrality condition (for ionic so-
lutions).
Degrees of freedom: The number of independent
intensive variables needed to describe a system is
called its degrees of freedom or variance.
Since the chemical potential is an intensive quanti-
ty, we can use n
t ot
= 1 to evaluate its value, and we
have
(T, P, n
1
, . . . , n
N
) =(T, P, x
1
, . . . , x
N
)
i.e., we can characterize the composition of the sys-
tem, or a given phase, by the set of mole fractions
instead of the set of numbers of moles. Further, since

N
i
x
i
= 1, we need only N 1 mole fractions, say
{x
1
, x
2
, . . . , x
N1
}.
PChem I 6.2
Consider a system with C components and phas-
es. (We use , instead of the textbooks P, for the
number of phases to avoid confusion with the pres-
sure variable.) To specify the state of such a system,
we need to know the intensive variables
T, P, {x
(1)
1
, x
(1)
2
, . . . , x
(1)
C1
}, {x
(2)
1
, x
(2)
2
, . . . , x
(2)
C1
},
. . . , {x
()
1
, x
()
2
, . . . , x
()
C1
}
namely 2+(C 1) variables.
On the other hand, at equilibrium the chemical po-
tential of species J must be the same throughout the
system:

(1)
J
=
(2)
J
= =
()
J
for J = 1, . . . ,C. In other words, there are 1 equa-
tions for each component, and there are C compo-
nents, so that we have C(1) equations overall that
establish relations between the 2+(C1) variables.
Thus the degrees of freedom are
F =2+(C 1) C (1)
PChem I 6.3
F =2+C
This is the Gibbs phase rule.
Practice:
How many components are in the following systems:
(a) water, allowing for its ionization
(b) AlCl
3
in water, noting that hydrolysis and precipi-
tation of Al(OH)
3
occur.
In the following, let S denote the number of distinct
chemical species, and R the number of constraints.
(a) The species are H
2
O, H
+
, OH

, i.e., S = 3. The
constraints are:
equilibrium:
H
2
O

H
+
+OH

electroneutrality:
n
H
+
(aq) =n
OH

(aq)
PChem I 6.4
divide by n
tot
(aq)
x
H
+
(aq) =x
OH

(aq)
So, R =2 and C =S R =1
(b) AlCl
3
, Al
3+
, Cl

, Al(OH)
3
, H
2
O, H
+
, OH

so, S =7
equilibria:
AlCl
3
(s)

Al
3+
(aq) +3Cl

(aq)
H
2
O(l)

H
+
(aq) +OH

(aq)
Al
3+
(aq) +3OH

(aq)

Al(OH)
3
(s)
electroneutrality:
n
H
+
(aq) +3n
Al
3+
(aq) =n
OH

(aq) +n
Cl

(aq)
divide by n
tot
(aq)
x
H
+
(aq) +3x
Al
3+
(aq) =x
OH

(aq) +x
Cl

(aq)
PChem I 6.5
So R =3+1 and C =S R =74 =3
one-component systems
F =C +2 =3
3
=1 : F =2 P and T can be freely chosen: area
=2 : F =1 P =P(T) or T =T(P) phase coexistence line
=3 : F =0 P and T are xed; triple point
[Figure: one-component phase diagram, Fig. 6.2]
locating rst-order phase transition points: thermal
analysis, cooling curve
[Figure: cooling curve, Fig. 6.4]
two-component systems
F =4
4
PChem I 6.6
at most 4 phases can coexist; this occurs at a point in
a T-P-x-diagram
vapor pressure diagrams for mixture of volatile liq-
uids
liquid-vapor equilibriumof a mixture of similar liquids,
e.g., toluene and benzene: ideal solution
P
A
=x
A
P

A
, P
B
=x
B
P

B
x
i
= mole fraction of i in the liquid phase
P =P
A
+P
B
=x
A
P

A
+x
B
P

B
=x
A
P

A
+(1x
A
)P

B
P =P

B
+
_
P

A
P

B
_
x
A
total vapor pressure at xed temperature varies lin-
early with the composition of the liquid phase
PChem I 6.7
phase di-
agram
assumes
A is more
volatile, i.e.,
P

A
>P

B
since A is more volatile, vapor will be richer in A: com-
position of liquid phase and vapor phase in equilibri-
um are not the same
mole fractions in the vapor phase: y
A
, y
B
y
A
=
P
A
P
, y
B
=
P
B
P
y
A
=
x
A
P

A
P

B
+
_
P

A
P

B
_
x
A
=
x
A
P

B
P

A
+
_
1
P

B
P

A
_
x
A
y
A
=
x
A
x
A
+
P

B
P

A
x
B
>x
A
for P

A
>P

B
PChem I 6.8
x
A
=
y
A
P

B
P

_
P

A
P

B
_
y
A
P =
P

A
P

B
P

A
+
_
P

B
P

A
_
y
A
to understand distillation, use a phase diagram that
contains both the liquid phase and the vapor phase
composition
[Figure: vapor pressure phase diagrams, Fig. 6.9, 6.10, & 6.12]
vertical dotted line = isopleth
horizontal lines a
1
-a

1
, etc., = tie lines
relative amounts of liquid and vapor phase: lever rule
n

=n

[Figure: lever rule, Fig. 6.13]

temperature-composition diagrams
distillation of mixtures
PChem I 6.9
[Figure: distillation, Fig. 6.14]
[Figure: fractional distillation, Fig. 6.15]
deviations from ideal solutions
mixing, real solutions
real solutions:
AB
=
AA
and
AB
=
BB
;
i j
= average
interaction energy between particles i and j
G
i
=n
A

A
(l ) +n
B

B
(l )
G
f
=n
A
_

A
(l ) +RT lna
A
_
+
+n
B
_

B
(l ) +RT lna
B
_

mix
G =nRT
_
x
A
lna
A
+x
B
lna
B
_

mix
G =RT
_
x
A
lna
A
+x
B
lna
B
_
a
i
=
i
x
i

mix
G =RT
_
x
A
lnx
A
+x
A
ln
A
+x
B
lnx
B
+x
B
ln
B
_

mix
G =RT
_
x
A
lnx
A
+x
B
lnx
B
_
+
+RT
_
x
A
ln
A
+x
B
ln
B
_

mix
G =
mix
G
ideal
+G
E
G
E
=RT
_
x
A
ln
A
+x
B
ln
B
_
PChem I 6.10
If
AB
>
AA
and
AB
>
BB
, then the A and B molecules
decrease each others escaping tendency. Favorable
interactions between A and B molecules reduce the
vapor pressure of the mixture below the ideal value,
i.e., the AB interactions stabilize the liquid phase =
negative deviations from Raoults law, <1 and G
E
<
0, more favorable to mixing than ideal.
example: acetone and chloroform can form hydro-
gen bonds, a relatively strong intermolecular force
not present in the pure liquids:
[Figure: vapor pressure acetone/chloroform mixture]
negative deviations from ideality may lead to a maxi-
mum in the temperature-composition diagram
[Figure: high-boiling azeotrope, Fig. 6.16]
repeated distillation of the liquid leads to an
azeotrope: evaporation occurs without change in
composition
PChem I 6.11
example: hydrochloric acid/water, azeotropic at 80%
by mass of water, boils unchanged at 108.6

C
If
AB
<
AA
, then the solute breaks up strong inter-
molecular forces in the solvent (nearly always in-
volves molecules containing OH groupswater, al-
cohols, organic acidswhich may form hydrogen
bonds). Unfavorable interactions between A and B
molecules increase the volatility of the liquids, i.e.,
the AB interactions destabilize the liquid phase =
positive deviations from Raoults law, > 1 and G
E
>
0, less favorable to mixing than ideal.
example ethanol/heptane
[Figure: vapor pressure ethanol/heptane mixture]
[Figure: activity of ethanol in heptane]
positive deviations from ideality may lead to a mini-
mum in the temperature-composition diagram
[Figure: low-boiling azeotrope, Fig. 6.17]
fractional distillation of the vapor leads to an
azeotrope
example: ethanol/water, azeotropic at 4% by mass of
water, boils unchanged at 78

C
PChem I 6.12
liquid-liquid phase diagrams: partially miscible liquids
[Figure: phase diagram of hexane/nitrobenzene, Fig. 6.19 & 6.20]
upper critical solution temperature or upper conso-
lute temperature T
uc
can be modeled by regular solutions
molecules mix randomly as they do in ideal solutions:
S
E
=0
deviations from ideality due to different strength of
interactions between like and unlike molecules: H
E
=
0
must have H
E
0 as x
A
0 or x
B
0
simplest form that fullls this condition:
H
E
=wx
A
x
B
=RTx
A
x
B
with =w/RT
H
E
symmetric in x
A
and x
B
=1x
A
[Figure: H
E
/nRT vs x
A
, Fig. 5.19]
PChem I 6.13
w
AA
+
BB
2
AB
w is generally weakly temperature and composition
dependent; will be neglected (otherwise we do not
have strictly speaking a regular solution)
G
E
=wx
A
x
B
=RTx
A
x
B
V
E
=
_
_
G
E
P
_
_
T
=0
U
E
=wx
A
x
B
the activity coefcients for regular solutions are given
by the Margules equations:

A
=exp
_
wx
2
B
RT
_
,
B
=exp
_
wx
2
A
RT
_
and thus
a
A
=
A
x
A
=x
A
exp
_
w(1x
A
)
2
RT
_
PChem I 6.14
P
A
=
_
x
A
exp
_
w(1x
A
)
2
RT
__
P

A
[Figure: vapor pressure of solvent for a regular solution, Fig. 5.32]
if w < 0, interactions are stronger between unlike
molecule than like molecules: negative deviation
from Raoults law; more favorable mixing; mixing is
exothermic:
mix
H = H
E
<0
in fact,
mix
G < 0, i.e., mixing is spontaneous for all
compositions,
mix
G has one deep minimum at x
A
=
1/2 =

mix
G
x
2
A
0 for all x
A
=mixing is homogeneous, no phase separation oc-
curs
if w >0, interactions are stronger between like mole-
cule than unlike molecules: positive deviation from
Raoults law; less favorable mixing; mixing is en-
dothermic:
mix
H = H
E
>0;
PChem I 6.15
since
mix
G
ideal
= RT
_
x
A
lnx
A
+x
B
lnx
B
_
0 and G
E
=
wx
A
x
B
0,
mix
G may have a maximum at x
A
= 1/2
and two minima x
I ,I I
A
located symmetrically on either
side of 1/2, i.e.,

mix
G
x
2
A
<0 for a range of x
A
[Figure: Gibbs free energy of mixing for regular solutions, Fig. 5.20]
=phase separation occurs; mixture of composition
x
A
with x
I
A
< x
A
< x
I I
A
separates into phases with com-
position x
I
A
and x
I I
A
, respectively; the proportions of
the two phases are given by the lever rule
since
mix
G
ideal
T and G
E
is temperature indepen-
dent,
mix
G is expected to develop a maximum and
two minima if the interaction term, G
E
, becomes im-
portant enough compared to the ideal mixing term,

mix
G
ideal
, i.e., if T is sufciently small
more precisely, if w >0, then for T <T
uc
(upper con-
solute temperature),
mix
G has two minima, corre-
sponding to two phases, one rich in A and the other
rich in B
PChem I 6.16
if T >T
uc
, then
mix
G has only one minimum
at T = T
uc
,
mix
G has a degenerate or double mini-
mum; this corresponds to a critical point of the sys-
tem
critical point: rst three derivatives of the Gibbs en-
ergy vanish simultaneously, the fourth is positive
determine the upper consolute temperature for a reg-
ular solution

mix
G =RT
_
x
A
lnx
A
+x
B
lnx
B
_
+wx
A
x
B

mix
G =RT
_
x
A
lnx
A
+(1x
A
) ln(1x
A
) +
+x
A
(1x
A
)
_

mix
G
x
A
=RT
_
1 lnx
A
+x
A

1
x
A
+
+(1) ln(1x
A
) +(1x
A
)
1
(1x
A
)
(1)+
+(12x
A
)
_
=0
PChem I 6.17
RT
_
ln
_
x
A
1x
A
_
+(12x
A
)
_
=0

mix
G
x
2
A
=RT
_
1
x
A
+
1
1x
A
2
_
=0

mix
G
x
3
A
=RT
_

1
x
2
A

1
(1x
A
)
2
(1)
_
=0
= x
A,c
=
1
2
2nd derivative = 0 =
c
=2 =
w
RT
c
=T
c
=
w
2R
rst derivative vanishes at x
A
= 1/2 due to the sym-
metry of the model, and it is easily veried that
the fourth derivative is positive at the critical point
(x
A,c
, T
c
)
[Figure: location of phase boundary, Fig. 6.23]
mixture of components that form weak complexes at
low temperatures may have a lower critical solution
PChem I 6.18
temperature or lower consolute temperature T
l c
[Figure: water/triethylamine mixture, Fig. 6.24]
some systems have both upper and lower consolute
temperatures
[Figure: water/nicotine mixture, Fig. 6.25]
PChem I 6.19