CHEM 2410:

Physical Chemistry I
REVISION: Chapters 5 ,6 ,12

Course Instructor:
Prof. Jonathan Halpert
 CHEM 2410: Chapter 5
CHAPTER OUTLINE
Thermodynamics of mixtures Phase Diagrams of Binary Systems
1. Partial Molar Quantities 6. Vapor Pressure Diagrams
2. Thermodynamics of Mixing 7. Temp/Composition Diagrams
3. Chemical Potential of Liquids 8. Liquid-Liquid Phase Diagrams
9. Liquid-Solid Phase Diagrams
Properties of Solutions
4. Liquid Mixtures Activities
5. Colligative Properties 10. Solvent Activity
11. Solute Activity
12. Activity of Regular Solutions
13. Activities of Ions in Solution
Required Reading:
“Atkin’s Physical Chemistry”, Atkins, de Paula & Steeler 11th Ed. 2018: Chapter 5
Simple Mixtures
State Properties
Partial Molar Volume
State Properties

(of species in mixtures)

 Partial molar volume

The partial molar volume, VJ of a substance J is

 V  which can be thought of as the change in the system (or sample)

VJ =  

 J  p ,T ,n '
n volume, V, for a change in some substance “J”

such that for a two-component mixture of A and B, V total changes by dV:

 V   V 
dV =    dnA +    dnB dV = VA dnA + VB dnB
 nA  p ,T ,nB  nB  p ,T ,nA

nA nB
only true if both integration steps are at the same
By integrating V =  dV =  VA dnA +  VB dnB = VA nA + VB nB time so that nA/nB (and thus dVA/dVB) is constant
0 0 (as we will prove later on)…the result is obviously
true for the marginal values of VA and VB
Partial Molar Gibbs Energy = μn
State Properties

(of species in mixtures)

 Partial molar Gibbs energy

The partial molar Gibbs energy is the chemical potential of a substance J:

 G  which can be thought of as the change in the total Gibbs

J =  

 J  p ,T ,n '
n energy, G, for a change in substance J

such that for a two component mixture of A and B, G total changes by dG:

 G   G 
dG =   dnA +   dnB dG =  A dnA +  B dnB
 nA  p ,T ,nB  nB  p ,T ,nA
nA nB

G =  dG =  A dnA +   B dnB =  A nA +  B nB (p,T constant) and thus wadd ,max =  A nA +  B nB

0 0
 Partial molar Gibbs energy

The partial molar Gibbs energy is the chemical potential of a substance J:

 G 
J =   n2

 J  p ,T ,n '
n
n1

 G   G 
dG =   dnA +   dnB
 nA  p ,T ,nB  nB  p ,T ,nA

dG =  A dnA +  B dnB G =  A nA +  B nB n1 n2

• the slope of line G (of the system) vs nA gives the chemical potential at each amount of A (or xA = nA/n)
• the chemical potential is strongly dependent on the relative amount, xA in a mixture
Gibbs-Duhem Equation
State Properties
 Gibbs-Duhem Equation (p,T constant)

We derived that for a binary systems G =  A nA +  B nB so for an infinitesimal change in G, dG (p,T const.)

one might expect dG = d (  A nA ) + d (  B nB ) =  A dnA + nA d  A +  B dnB + nB d  B (product rule)

however dG p ,T =  A dnA +  B dnB from our earlier derivation (which was valid!)

since G is a state function, these two functions must be equal, which implies that

nA d  A + nB d  B = 0 this is the Gibbs-Duhem Equation, which can be generalized to n d

J
J J =0

The Gibbs-Duhem Equation, suggests that changes in the chemical potential of any one species relies on
changes in the chemical potential of the other species
nB
nA d  A = −nB d  B d A = − d B see addendum for example
nA this reasoning can be extended to dV = − nB dV
A B
any partial molar quantity, e.g. VJ nA
Gibbs Energy of Mixing: Gases
State Properties
 Example: Gibbs energy of mixing
A container is divided into two equal compartments. One contains 3.0 mol H2(g), the other contains 1.0 mol N2(g),
both at 25 °C. Calculate the Gibbs energy of mixing when the partition is removed, assume perfect behavior.
initial pressure of N2 side is p, pressure of H2 side is 3p (Vi,Ti equal)
final partial pressure of N2 is 0.5p and of H2 is then 1.5p
Gi = nA (  A  + RT ln p ) + nB (  B  + RT ln p )
A B

Gi = ( 3.0 mol )   H 2  + RT ln ( 3 p / bar )  + (1.0 mol )   N2  + RT ln ( p / bar ) 

after

G f = nA (  A  + RT ln p A ) + nB (  B  + RT ln pB )
G f = ( 3.0 mol )   H 2  + RT ln (1.5 p / bar )  + (1.0 mol )   N2  + RT ln ( 0.5 p / bar ) 
change

 mix G = ( 3.0 mol )  RT ln (1.5 p / 3 p )  + (1.0 mol )  RT ln ( 0.5 p /1 p ) 

= − ( 3.0 mol ) RT ln 2 − (1.0 mol ) RT ln 2
= − ( 4.0 mol ) ( 8.3145 J K -1 mol-1 ) ( 298 K )( ln 2 ) = −6870 J ln(1/2) = ln(2)-1 = -ln(2)
Chemical Potential of Mixing: Solvents
State Properties
 Chemical Potential of Solvents in Mixtures
Chemical potential of a pure liquid (relative to std. conditions)

 =  A + RT ln p
*
* * indicate these are pure substances
A A ° indicates p/(1 bar)

Chemical potential of species A, in a mixture can be expressed

 pA 
 A =  A  + RT ln p A  A =  A* + RT ln  * 
 A
p

Raoult’s Law: related partial pressures to liquid compositions (experimentally observed)

chemical potential of a component of an ideal solution

p A = x A p*A  A =  A* + RT ln xA (and the definition of an ideal solution)

vapor pressure exists b.c. the entropy is increased by having molecules in the vapor phase, in a mixture the
entropy of the liquid phase is higher, so less entropic gain to having vapor…thus pA is lower than pA*
Chemical Potential of Mixing: Solutes
State Properties
 Chemical Potential of Solutes in Solvents

A solute is much more dilute, surrounded by solvent molecules and is thus in a very
different environment:

Henry’s Law: partial pressure of a solute in an ideal solution

p B = xB K B (KB is an empirical constant with dimensions of pressure)

or

pB = bB K B B in units of (mol solute/kg of solvent), KB units of (pressure × kg

solvent / mol solute) For practical use, Henry’s Law is often used with molality,
bB, instead of mole fraction (KB found w/ proper units)

In cases where the solvent molecules are very similar to the solute molecules, the local
environment is similar and the solute may obey Raoult’s Law

ideal-dilute solutions
mixture that obeys Raoult’s Law for the solvent, and Henry’s Law for the solute
Thermodynamics of Mixing: Liquids
State Properties
 Gibbs Energy, Entropy, Enthalpy of Solutions

To find the Gibbs energy of mixing (just like for gases) vapor and liquid are at
equilibrium, so u(l)=u(g)
Gi = nA  *A + nB  *B from Raoult’s Law, So we calculate Gmix for
since xA=pA/p*
( ) ( )
Gvapour to get Gliquid
G f = nA  *A + RT ln x A + nB  *B + RT ln xB
 mix G = nA RT ln x A + nB RT ln xB
 mix G = nRT ( x A ln x A + xB ln xB ) Gibbs Energy of mixing to form an ideal solution
(very similar to gas mixing)

The entropy for mixing ideal solutions (same steps as gas mixing)
 mix S = −nR ( x A ln x A + xB ln xB )
the change in volume comes from:
The enthalpy for mixing ideal solutions is thus:
 mix H = 0  G    mix G 
  = V so  V =   =0
 p T ,nA ,nB  p T ,nA ,nB
mix
 Implications for Real Gibbs Energies

 mix G real = nRT ( x A ln x A + xB ln xB +  x A xB )

entropy of mixing enthalpy of mixing

Entropy should always induce spontaneous mixing (ln x < 0) =ξ

If ξ is large enough and positive (endothermic interactions),

then ΔG > 0 and mixing is not spontaneous (const. T, p) and
instead phase separation is favored (immiscible liquids)

Chart shows ΔG for (arbitrary) values of ξ

Colligative Properties
State Properties
 Colligative Properties

Colligative properties: properties that depend on the amount of the solute species present but not its identity.

Colligative properties arise from changes in the chemical potential of the solvent due to entropic effects.
(which is why molecular identity is not important, just the amounts of other species)

Colligative properties include:

• Reduction of partial vapor pressure (previous section)

• Boiling point elevation
• Freezing point depression
• Solubility (not a true colligative property but here we can use the same formalism)
• Osmosis

These properties are explored here for dilute solutions of A and B such that nA>>nB and thus xA>>xB
 Common features of colligative properties

The presence of the solute reduces the chemical potential

 A (l ) =  *A (l ) + RT ln x A

since xA<1 then ln xA<0 so μ(l)<μ*(l)

Assumptions in depicting colligative properties

1. Solutions are dilute
2. Solute non-volatile - solute doesn’t contribute to vapor pressure
3. Solute non-soluble in solid solvent - solid phase is pure solvent

So assume that solute exists only in liquid solution phase

 Boiling Point Elevation
At the boiling point (b.p. or Tb), the chemical potential of the gas and liquid of the solvent A are at equilibrium
(assume at 1 atm, this will not be explicitly written but xA=pA/1 atm)

for solvent A, the liquid has a reduced μ due to RT ln xA<0 but the solid is
 *A ( g ) =  *A (l ) + RT ln x A
unaffected (assumption): thus the gas phase becomes lower in G at a higher T

With the presence of the solute B, the Tb* (at 1 atm) is raised to Tb= Tb* + ∆Tb where

R (Tb* )
2

Tb = KxB and K = xB is the mole fraction of the solute in the solvent
 vap H

For practical use (like with other empirical expressions), the equation is:

Tb = K bbB where b is the molality of B ( molB /kg A )

 Freezing Point Depression

Chemical potential of solid vs liquid at freezing point (fus)

 *A ( s) =  *A (l ) + RT ln x A
By similar arguments, the liquid potential μ(l) is lower so G(l) is still lower than
G(s) even at smaller T…thus the freezing point, where G(l)=G(s), decreases

With the presence of the solute B, the Tf* (at 1 atm) is lowered to Tf= Tf* – ∆Tf where

R (T f * )
2

T f = K xB and K  = xB is the mole fraction of the solute in the solvent

 fus H

For practical use (like with Henry’s law and other empirical expressions), the equation is

T = K f bB where b is the molality of B ( molB /kg A )

 Solubility

chemical potential conditions at equilibrium between variation of solubility

 B* ( s) =  B* (l ) + RT ln xB with temperature
the solid and dissolved solute (liquid phase)
…at equilibrium, this is a saturated solution

for an ideal solute, where Tf is the melting point

of pure solute B

 fus H  1 1 
ln xB =  − 
R  T f T 

=fusH/(RT*)
 Osmosis

From the Greek “to push”, osmosis is the spontaneous passage of pure solvent into a solution separated from
it by semi-permeable membrane that is permeable to solvent but not to the solute

Osmotic pressure: pressure, Π, that must be applied to the solution to

prevent to stop the influx of solvent

The derivation requires the chemical potential be equal on each side of the
membrane at a pressure (p+Π) at which solvent no longer passes through

For dilute solutions, the osmotic pressure is:

van ’t Hoff Formula

nB
 =  B  RT where  B = the molar concentration of the solute B
V
 Composition of the Vapor vs. the Liquid Phase

if y is the mole fraction in the vapor and x is the mole

fraction in the liquid then y=py/p

x A p*A xB pB* y and xA=nA/n in the liquid

yA = and yB = are not necessarily equal
p p
or using the results from the last slide:

x A p*A
yA = and yB = 1 − y A
x A ( p A − pB ) + pB
* * *

pB* p*A
in addition, we can p= to get the vapor pressure from the fraction of A in
rearrange this into y A ( pB* − p*A ) + pB* the vapor (instead of xA in the liquid)
(several steps, algebra!)
Mixed Phase Diagrams: Binary Systems
State Properties
 Counting molecules, finding fractions
We can combine the two diagrams to form a single mixed phase
diagram (for a closed system, e.g. in a piston chamber) vs p (at r.t.)

z is a variable that denotes the total mole fraction in the chamber in

any phase e.g. zA nA,vap
+
nB,vap nvap
x is the mole fraction in the liquid, and y is the mole fraction in the
vapor, e.g. xA and yA zA=nA/n
zB=1–zA
n is the total number of moles of materials, e.g. nA is the moles of A nA,liq
in total + nliq
nB,liq
nvap is the total number of moles in the vapor, and nliq is the total
number of moles in the liquid
nA=nA,vap + nA,liq xA=nA,liq / nliq
So then zA = nvapyA + nliqxA and n = nvap + nliq = nA+nB nvap=na,vap+nB,vap yA=nA,vap/nvap
 Lever Rule

The phases present in a piston chamber: high p low p

The lever rule describes how much of
material is in one phase or another by the
quantities nαLα = nβLβ

Lα Lβ
Temperature vs. Composition: Liquid-Vapor
State Properties
 Temperature Composition Diagrams

Differs from p vs. ZA (liquid is now in lower corner)

Shows composition of phases in equilibrium points

at a single pressure (often 1 atm)

Relates mole fraction ZA to T for both vapor and

liquid phases (between the lines) and provides (xA,
yA, nAα, nAβ) as for pressure-composition diagrams

Distillation is the vaporization and condensation

steps to purify A, which is richer in the vapor phaase
 Theoretical plates

The number of theoretical plates is the

number of steps needed to achieve a
particular mixture, xA

Some mixtures have a larger compositional

spread between the vapor and the liquid

This allows a greater degree of separation

for each plating event

This makes sense, for a component with a

high vapor pressure should represent a Number of theoretical plates, for two curves (a) 3 and (b) 5
relatively higher vapor pressure/vapor
component and thus purify to a higher A plate corresponds to the number of times the mixture boils
degree on each step (plate) and condenses, usually while climbing a distillation tower
 High Boiling Azeotropes

In an azeotrope, the partial pressures of the

two components become equal at some
composition

This is due to intermolecular attractions which

stabilize the liquid phase, raising the boiling
point at the tangent point

Since the vapor is the same composition at this

point as the liquid, the curves are tangential
and the mixture cannot be purified beyond
this point via distillation

During distillation purified vapor can be

removed but will stop once A is depleted in
the liquid which has reached the tangent point
 Low Boiling Azeotropes

For low boiling azeotropes, the curves reach a

minimum at the tangent point

In this mixture the vapor moves towards the

minimum point with reduced purity coming from
the solvent then stop purifying via distillation

Low boiling azeotropes are the most difficult

mixtures for separating via distillation since they
reach (b) in the column itself
 Distillation of Immiscible Liquids
For two immiscible liquids the Gibbs energy
of mixing is positive (non-spontaneous) so
they persist as separate liquids

When heated their partial pressures sum to

give the total pressure (no mixing terms)

p = p*A + pB*

this increases the total pressure on the liquid

more quickly than would occur for a single
composition, permitting boiling (distillation) boiling (spontaneous conversion of the entire liquid to vapor)
at lower temperature occurs when the total vapor pressure equals 1 atm

the fraction recovered is effectively the this is useful for heat sensitive, water insoluble compounds as
partial pressure ratio they can be collected at a lower temperature
Partially Miscible Liquids
State Properties

(L-L Phase Diagrams)

 Liquid-Liquid Phase Diagrams (partly miscible)

Here we can consider temperature-composition phase

diagrams for partly miscible liquids

• In the partly miscible regime, as the mole fraction of B

increases, some of B dissolves in the A phase and
some of A dissolves in the B phase and:

• The amount of B in A phase is xB at a’

• The amount of A in B phase is (1-xB) at a’’

• For points in the two-phase regime, the tie-line is used The upper critical point is the T at which both
to find the relative quantities of the phases by the components are miscible at all compositions
lever rule nαLα = nβLβ
 Low temperature miscibility

triethylamine in water nicotine in water

• Some species, (trimethylamine), form stable complexes at low T, that break apart at high T
• This causes a lower critical temperature (Tlc) above which two phases can form at some compositions
• Other species exhibit both a Tlc and Tuc: first forming stable compounds, later high T makes miscibility spontaneous
Distillation of Partially Miscible Liquids
State Properties

(L-L-V Phase Diagrams)

 Distillation of partially miscible liquids

Above the upper critical temperature, the vapor liquid

line is similar to any other mixture and has one phase

The composition at any point on the isopleths

between the boiling and vapor lines, or between in
the two-phase regime, the tie lines and lever rule are
needed to find the moles n in each mixed phase (P=2)
 Distillation of partially miscible liquids: no Tuc

If there is no upper critical temperature

reachable before boiling, then a complex
phase diagram can be constructed

Distillate vapor boiled from a single phase

liquid at a2 will condense into two phases
below temperature T(e2)

This phase diagram has two 2-phase vapor

regimes bc there is an azeotropic point at e2
Solid-Liquid Phase Diagrams
State Properties
 Binary systems of nearly immiscible solids

In a eutectic system:

• a miscible liquid mixture begins to solidify by

depositing solid A or solid B first

• in Liquid + A, A solidifies and drops out of solution

so the liquid is rich in B (lever rule applies)

• after completely solidifying (<Te2), a two phase

(immiscible) solid is formed

• point e2 is the eutectic point, at which the liquid

solidifies into two phases all at once
Reaction Phase Diagrams
State Properties
 Reacting Systems

• Many binary mixtures will react to create a third

constituent (C=2, product is not independent)

• One example is GaAs, which forms a eutectic with

As and Ga, as in the diagram

• If A>B, the composition is C and A (unreacted),

determined via lever rule from the pure C line (e.g.
a3 )

• If B>A, the composition is C and B (unreacted)

again determine by lever rule from the pure C line

• Pure compound C melts congruently, that is the

composition is the same at all points in the melt
 Incongruent Melting
• Here the compounds K and Na form Na2K (“nak”)
which displays incongruent melting, where the
formula changes in the melt:

E.g. at the melting temperature the Na2K solid does not

exist as a liquid and a eutectic mixture forms

• Along the isopleth Na2K + K (b1-4)

b1: liquid mixture

b2: liquid + solid Na
b3: liquid + solid Na2K
b4: solid Na2K + solid K (some Na)

• The line b3 to a3 is called a peritectic line, at which K Position of the product is noticeable
suffuses into the solid Na and reacts to form Na2K due to the 2:1 ratio needed to form it
 Multiple Products

• Here compounds of Ca/Si form several

compositions, each mixture with a
eutectic point

• at either side of the eutectic points, one

or the other compound is preferentially
deposited

• The eutectic composition for each

mixture corresponds to the minimum
melting temperature 1 liq. phase to 2
solid phases
Ternary Phase Diagrams
State Properties
 Ternary Phase Diagrams
In a ternary phase diagram, there are 3 components:

Each edge of the triangle corresponds to a binary mixture

of the two corner species

An apex line has constant composition ratio of the other

two corner species

Usually these are presented in slices, at fixed T,p so that

the number of free variables is F=3-P+0

• a line denotes an area with different mixtures of phases

• multiple phases may be present in any area, the

number of free variables will determine the other
compositions
Activity
State Properties
 Solvent Activity

So rather than change the form of the expression for chemical potential (based on an ideal solvent), we can
instead alter the ideal ln(xA) to be the real ratio ln(p/p*)

This ratio is called the activity (aA) the real partial pressure ratio, substituted as an effective mole fraction
(instead of the actual mole fraction xA)

This is similar to our use of fugacity, the idealized version of the real pressure, such that equations for
perfect gases worked… by translating real p to f, we could then plug f into the Gibbs equations

In this same way, for any mole fraction, we can determine the activity and plug that into the chemical
potential equation
“gamma”, the activity coefficient, a property of the real solvent

pA
 A =  A* + RT ln ( a A ) where aA  or a A =  A x A so  A =  A* + RT ln ( xA ) + RT ln (  A )
p*A
If the solvent is ideal, then aA = xA and γ = 1 ln(ab)=ln(a) + ln(b)
 Real Solutes: Activity

For an ideal solute then the chemical potential is  B =  B + RT ln ( xB )

So for a real solute where activity a B =  B xB

The chemical potential is therefore  B =  B + RT ln ( aB ) maintains the equation of the ideal

solute but substitutes aB for xB

Unlike the ideal solvent (which is pure), the ideal solute is dilute so γ→1 as xB→0
 Implications of the Margules Equations

If our expression for activity A is Then the partial pressure is

 (1− x A )
2

a A =  A x A = x A exp  (1 − x A )  pA = p  x = p x e
2 * *
  A A A A A

This functions implies that:

• when ξ=0 (ideal solution), Raoult’s Law holds p A = p A x Ae = p A x A

* 0 *

• positive values of ξ, (endothermic mixing, unfavorable interactions) give pA higher than ideal

• negative value of ξ, (exothermic mixing, favorable interactions) give pA lower than ideal
 (1− x A )
2

• plots become linear as x→1, and coincide with Raoult’s Law since e → 1 as (1-x)→0
 
p
• as x→0 equation becomes A = p *
x
A A e K
which is Henry’s Law with A = x A e
Activities of Ions
State Properties
 Activities of Ions

If the chemical potential of the cation M+ is μ+ and that of the anion is μ– then

Gmideal = + + −

 J ideal =  J  + RT ln xJ or  J real =  J  + RT ln aJ

Gmreal = + ideal + −ideal + RT ln  + + RT ln  −

Gmreal = Gmideal + RT ln (  + − ) where all the deviations from ideality are captured in the last term
 Mean Activity Coefficients

Previously Gmreal = Gmideal + RT ln (  + − ) but it is difficult to assign separate activities to the individual ions
Instead we can find the geometric mean of the activities of the two species and find the mean activity coefficient:

  = (  + − ) Gmreal = Gmideal + 2 RT ln (   )
12
so that for an ionic species M+X-

We can generalize this to MpXq which creates p moles of cations and q moles of anions per mole of MX

Now Gmreal = p + ideal + q −ideal + pRT ln  + + qRT ln  −

but if we take the geometric mean to get a mean activity coefficient (general expression)

  = ( + − )
p q 1 ( p+q)
then i = i ideal + RT ln   for each ion, i

then finally
real
G
m = p+ + q− but how to get γ± ? u geom. = ( x y
p
)
q 1 ( p+q)
 Debye-Hückel Limiting Law
Debye-Huckel Theory: the long range of Coulombic (charge-charge) interactions between ions
means that this is likely the dominant factor in the non-ideality of solutions

Charged ions attract each other, so each ion is likely to have a nearby cloud of oppositely charged
ions called the ionic atmosphere: the energy (and thus μ) of the ion is lower due to this interaction
Debye-Huckel
model
The stabilization of ions due to local electrostatic interactions explains why chemists use dilute
solutions to achieve precipitation of ions from solution (less stabilization)

At very low concentration the activity can be derived from the Debye-Huckel Limiting Law: molality of ion, i
a constant related to dielectric of the solvent
1 2  bi 
log   = − A z+ z− I 1 2 A = 0.509 for aq. solution at 25C I=  i  b  
2 i
z

signed total charge +/- ionic strength of the solution signed charge on ion, i
Called the “limiting law” because the activity coefficient becomes more accurate as the solution is more dilute
 CHEM 2410: Chapter 6
CHAPTER OUTLINE
Spontaneous Chemical Reactions Equilibrium Electrochemistry
1. The Gibbs Energy Minimum 5. Half-rxns and electrodes
2. The Description of Equilibrium 6. Varieties of Cells
7. The Cell potential
The Response of Equilibria to Conditions 8. Standard electrode potential
3. Changes in Pressure 9. Applications of Standard Potentials
4. Changes in Temperature

Required Reading:
“Atkin’s Physical Chemistry”, Atkins, de Paula & Steeler 11th Ed. 2018: Chapter 6
Extent of Reaction (ξ)
State Properties
 Equilibrium and Extent of Reaction

The direction of spontaneity for RTP (room temp. and pressure) experiments (decreasing Gibbs) tends toward
equilibrium (as long as there are no kinetic barriers): the point at which macroscale changes in the system stop

• At the equilibrium point, dG = 0 → represents the minimum of the Gibbs energy curve for that system

• Consider the reaction: A B

• A and B are interconverting, but at equilibrium this happens at the same rate

The extent of the reaction: a variable ξ (“xi” or “ksai”, same Greek letter but not the enthalpy variable from Ch.
5) which serves as a metric of how far the reaction goes, with units of quantity (think “moles of the reaction”)

• for a small change in the reaction dξ, a small amount of A (reactant) –dnA becomes B (product) by +dnB

• for a noticeable change Δξ the amount of A, nA,0 → nA,0 – Δξ and B goes to nB,0 → nB,0 + Δξ
Reaction Gibbs Energy (ΔrG)
State Properties
 Reaction Gibbs Energy (ΔrG)

Plot of G vs. ξ: the equilibrium point is the curve minimum

Reaction Gibbs Energy, ∆rG, is in fact defined as the slope of the

curve at any initial point, Gi (where Gf is the equilibrium point)
 G 
rG    so for a small change dG (const. p,T):
   p ,T
dG =  A dnA +  B dnB for reaction A B
dG =  A  ( −d  ) +  B  ( + d )
dG = (  B −  A ) d 
thus ∆rG is the difference between chemical potentials of products minus
reactants at the current composition of the mixture (as one would expect)
 G 
rG =   = B −  A
   p ,T remember: the chemical potential changes with composition, as should ∆rG
 Equilibrium w.r.t. Chemical Potential
 G 
rG =   = B −  A
Consider the reaction A B
   p ,T
Thus when dG < 0 (A→B spontaneous) when:
• ∆ rG < 0
• slope of Gibbs curve is negative
• μA > μB

And dG > 0 (B→A reverse rxn spontaneous) when

• ∆rG > 0 (as the rxn is written)
• slope of Gibbs curve is positive
• μB > μA

And dG = 0 (B→A and A→B equivalent) when

• ∆rG = 0 (as the rxn is written)
• slope of Gibbs curve is 0 thus the chemical potentials live up to their name:
• μB = μA the relative potential for changes in the chemical reaction
Generalizing the Thermodynamic
State Properties

Theory to Any Reaction

 General Case of any reaction

Note that for any balanced reaction 0 = 

all elements, J
v J J indicating that the signed moles of substances sum to zero

With this in mind, we can write an expression for the Gibbs reaction energy at any point during the reaction

   G 
dG =   J dnJ =   J v J d =    J v J  d so then rG =   =  vJ J
all species, J all species, J  all species, J     p ,T J

We can then write the chemical potentials as a function of the activities of each species

 
 r G =  v J  J + RT   v J ln aJ  =  r G  + RT  v J ln aJ =  r G  + RT  ln aJ v J ( ) a ln x = ln xa
J  J  J J

 vJ 
and since ln a + ln b +… = ln (a*b*…) then we can write  r G =  r G + RT ln   aJ 

for any reaction
 J 
 Q, the reaction quotient

 
 r G =  r G  + RT ln   aJ v J  from the last slide, so we can write
 J 

( a ) (a )
vp vq
 ...(all products)
Q =  aJ =
vJ p q

( am )  ( an )  ...(all reactants)
vm vn
J

 r G =  r G  + RT ln Q as we expect generally (given the above def. of Q)

 vJ 
furthermore, K can then be defined as K =   aJ  but only when given their values at equilibrium (eq.)
 J eq.
K (dimensionless) in terms of stoichiometric activities is called a thermodynamic equilibrium constant.
Using Activity to Calculate
State Properties

Equilibrium Constants, K
 List of activities for various phases

K (dimensionless) in terms of stoichiometric activities is called a thermodynamic equilibrium constant.

The activity is determined in different

ways for solutes, solvents, gases and
pure liquids/solids (memorize these)

Finally we can calculate the equilibrium constant using the standard

reaction Gibbs Energy (experimentally or using Gibbs formation energies)
 r G  = − RT ln K
Molecular Interpretation of Equilibrium
State Properties
 Molecular Interpretation of Keq
for reaction A→B
As species form, the energy of the system is
distributed among the available states

One expects that lower energy states would be

preferred, and that reaction eq. would tend
towards those species

But entropy also plays a role, and can cause high

energy, but densely packed, states to reduce ΔG

This interplay is captured by:  r G  =  r H  − T  r S 

We can rearrange this using  r G  = − RT ln K

  A dominates, B dominates,
To get K = e −r H RT
e +r S R
enthalpy more important entropy more important
enthalpy contribution entropy contribution
Equilibrium Response to Conditions
State Properties
 Response of Equilibria to Conditions

Equilibrium Constant versus the following conditions:

K is independent of…

Catalysts: may change the rate, but not the energy of the products vs. reactants
Pressure via Inert Gas: relative compositions constant (total composition changes w.r.t. to total moles of gas)
Pressure via Compression (ΔV~1/Δp): K is independent of pressure but compositions may change significantly

K is dependent on…

Temperature: affects standard Gibbs Energy, and T appears in the equation below → must affect ln K and thus K

 
− r G  = + RT ln K where K =   aJ v J 
 J eq.

so we must closely examine changes in (p,T)

Equilibrium Response to Δp
State Properties
 Changes in P via Volume (Compression)

In the case of a compression, the change in volume does effect each partial pressure

For a simple reaction where A dissociates into 2 molecules of B

( pB )
2
ptot must follow the gas law pV=nRT
A 2B K=
( pA ) p o (before reaching a new equilibrium)

therefore to maintain the equilibrium at the new volume, a larger quantity

of A must be produced to maintain the relationship with K

This gives rise to a common restatement of le Chatelier’s Principle: any

perturbation in the system will cause the system to react against it
e.g. to “relieve the pressure”.

In this case, it means xB decreases as this will relieve the change in p

Degree of Dissociation, α
State Properties
 Degree of Dissociation, α
from some starting
Degree of dissociation is the fraction of a compound that has decomposed  = ni − neq point n, composition
ni changes to neq
• Assumes some starting point, n of the material in question

• As the reaction progresses ni is converted to neq which is the endpoint of the reaction

• α is then the amount converted as a percentage of the initial n

• it is often useful to consider αn = ni-neq = Δn the amount by which the reactant changes during the
reaction (neq ≡ nf) and is closely related to Δξ

this is very useful for examining reactions where the final ratio of products to reactants is non-trivial

in particular for reactions where the flask is loaded with reactants, and a noticeable amount of reactant
will be consumed while products form (constrained by the initial mass and the equilibrium constant)
van ‘t Hoff Equation: ΔrH from ln K
State Properties
 Measuring Standard Reaction Enthalpy

Use data below to show the temperature variation of the equilibrium constant of the reaction

Ag 2 CO3 (s) Ag 2 O(s) + CO 2 (g) T/K 350 400 450 500

Keq 3.98×10-4 1.41×10-2 1.86×10-1 1.48×100
d ln K  r H 
= van ’t Hoff Eq.
dT RT 2
d (1/ T )

1 dT dT d ln K 2 r H
remember that = − 2 so = −T 2 and thus  = −T 
dT T d (1/ T ) d (1/ T ) dT RT 2
d ln K r H  d ( − ln K )  r H 
multiply through: =− =
d (1/ T ) R or d (1/ T ) R

so plot data as (–ln K) vs (1/T) and then get linear fit y=mx+b and m=∆H/R
Measuring ln K from ΔrH
State Properties
 Value of K at different temperatures

1 r H 
d ln K =  2
dT
R T

K2 T2
 1 r H   1 T 2  r H   r H  T2
dT  r H   1   1  
K 1 d ln K = T1  R  T 2  dT  = R T1   T 2  dT   R T 1 T 2 = R  −  −  −  
 T2   T1  

allows us to calculate change in K for a change in T

r H  1 1 r H  1 1
ln K 2 − ln K1 = −  −  ln K 2 = ln K1 −  − 
R  T2 T1  R  T2 T1 
 Summary of CHEM2410 to this point

The master equation that combines the 0th, 1st, 2nd, 3rd Laws of Thermodynamics and
chemical equilibrium constant:

rGꝋ = rHꝋ – TrSꝋ = – RT ln(K)

where
0th Law of thermodynamics: T Chapt. 1
1st Law of thermodynamics: H (U) Chapt. 2
2nd & 3rd Laws of Thermodynamics: S Chapt. 3
Combine 0th, 1st, 2nd, and 3rd Laws : G (A) Chapt. 4-5
Chemical equilibrium constant : K (rGꝋ) Chapt. 6
 Basic Concepts for Batteries
State Properties

(revision of introductory electrochemistry)

 Electrochemical Cell
Two electrodes are in contact with an electrolyte solution in either two compartments or together in one

During operation (Galvanic, flip directions for electrolytic):

• electrons hop from one species (converts ARed → AOx) to the
anode (oxidation)

• they travel through a circuit (wire) to the other electrode

• electrons hop onto another species (BOx to Bred) at the

cathode (reduction)

• Local changes in concentration or ionic charge return to

equilibrium when ions travel through the bulk solution

• the change in potential (voltage) between the two half-

reactions is the total potential (voltage) of the cell
(measured between the two electrodes via a voltmeter)
 Types of Cells
Danielle Cell: ZnSO4/Zn and CuSO4/Cu half reactions, porous connector or salt bridge
Electrolyte Concentration Cell: two compartments with identical solutions at varying concentrations
Electrode Concentration Cell: two electrodes are at different concentrations (partial pressure or Hg amalgam)

(single compartment) The Danielle Cell The Danielle Cell

(porous divider) (two compartments)
How to write cell configurations:
Pt(s)|H2(g)|HCl(aq)|AgCl(s)|Ag(s) soln. of HCl where H+ comes from H2 adsorbed to Pt and Cl- from AgCl
Zn(s)|ZnSO4(aq) ⁞ CuSO4(aq)|Cu(s) a Zn/ZnSO4 and Cu/CuSO4 cell with a porous divider
Zn(s)|ZnSO4(aq) ǁ CuSO4(aq)|Cu(s) the same two half-reactions now each in a separated compartment
Pt(s)|H2(g)|HCl(aq,b1) ǁ HCl(aq,b2)|H2(g)|Pt(s) a concentration cell of two conc. of HCl in two compartments, each H2/Pt
Standard Cell Potentials
State Properties
 List of Standard Potentials (partial)
Meaning the OTHER
e- Ered species is oxidized
Ef

Aox is strongly oxidizing

(reaction proceeds v. spontaneously to products)

oxidation at the anode

Gred  0
Ef
e- Eox
Gred  0

Ared is strongly reducing

relative to SHE (reaction proceeds v. spontaneously to reactants)
reduction at the cathode

Now we have all the parts in place to evaluate cell potentials under standard conditions…
 New Slide

for a primary cell reaction Zn(s) + MnO2(s) ZnO(s) + Mn2O3(s)

[OH-]

1. Assign Oxidation Numbers

Zn = 0, O = -2, Mn = +4 Zn = +2, O = -2, Mn = +3

2. Write half-reactions as unbalanced reductions

ZnO(s) + 2e- Zn(s)
2MnO2(s) + 2e- Mn2O3(s)

3. Balance each half-reaction as reduction (get std. potentials/ID cathode)

ZnO(s) + H2O + 2e- Zn(s) + 2OH- E° = -1.28 V
2MnO2(s)+ H2O + 2e- Mn2O3(s) + 2OH- E° = +0.15 V REDUCTION/CATHODE

4. Balance reactions together with e-

ZnO(s) + Mn2O3(s) E°cell = 0.15 - (-1.28) = +1.43 V
Zn(s) + 2MnO2(s)
[OH-]
Thermodynamic Description
State Properties

of Electrochemical Energy
(Cell Potential + Nernst Equations)
 Electrical Work

Maximum non-expansion work (additional or extra) a system can do is ∆G = wadd, max which is the amount of
additional work done during a reversible process
change in total charges
In electrochemistry wadd=we and in physics dwe =   dQ
electrostatic potential

 dG 
rG =   so  r G  d  = dG since dG = dwe and then  r G  d  = dwe
 d   p ,T
if the reaction advances by dξ then the number of electrons moving is v∙dξ and the total charge is -veNA∙dξ
which if F=eNA then this becomes -vF∙dξ where F=96485 C/mol

dwe =   dQ = Ecell  ( −vFd  ) = −vFEcell  d   r G = −vFEcell Cell Potential

Equation
work of bringing dQ charges across potential φ
 Implications of Cell Potential Eq.

 r G = −vFEcell

ΔG and E have opposite signs so when

ΔG<0 (spontaneous) E>0 (positive voltage)

When ΔG>0 (non-spontaneous) E<0

(negetive voltage), and thus the reaction
occurs in the opposite direction

As the reaction nears equilibrium, the

potential drops to zero
 Nernst Equation:
State Properties

Cell Potential vs. Composition

 Ecell from Reaction Quotient, Q

 r G = −vFEcell and  r G =  r G  + RT ln Q

  r G  RT
−vFE =  r G + RT ln Q E=− − ln Q
vF vF

 G
if we define a standard cell potential E = − r related to standard reaction Gibbs, ∆rG
vF
RT
E = E − ln Q Nernst Equation: relates composition to cell potential (via standard cell potential)
vF
The Nernst Equation: some common examples

Metal/metal ion

Gas/ion
Examples of
common half-cell
potentials and _____
how to use their 1
activities and v to
find E from Eθ Redox

Insoluble salt/metal __
1
Measuring Thermodynamic Variables from
State Properties

The Cell Potential

 Thermodynamic Values from Ecell

In addition to K, and measurements of Q

If the standard formation energy ∆fG(H+,aq) = 0 then the ∆fG of other ions can be determined

 dG  dE   r S 
similarly if   = −S then we can derive =
 dT  p dT vF

     dE 
From this we can also derive  r H =  r G + T  r S = −vF  Ecell − T cell

 dT 
Thus finding cell potentials gives us information on K, ∆rG, ∆rS, and ∆rH
 CHEM 2410: Chapter 12
CHAPTER OUTLINE
Distribution of States
1. The Boltzmann Distribution Derived Functions
2. Molecular Partition Functions 5. Internal Energy and Entropy
6. Thermodynamic Functions
Energy and Entropy in Molecules
3. Molecular Energies
4. Canonical Ensemble

Required Reading:
“Atkin’s Physical Chemistry”, Atkins, de Paula & Steeler 11th Ed. 2018: Chapter 12
Statistical Thermodynamics (Ch.12)
State Properties
 Distribution of States

Consider a system of N molecules.

Although we can measure the system energy, E, we cannot say for certain how it is arranged

Collisions between molecules cause constant changes in the energy for each particular molecule

However we can report the average population of states: for a large system this average is very consistent
even as individual molecules are themselves changing their energy

We make an assumption, the principle of a priori probability: all possibilities of equal energy are equally
likely to be populated…that is the nature of the state (vib, rot, trans etc.) is irrelevant in this assumption
 Configurations and Probability Weighting

If all configurations with are equally probable (later we will constrain them to E), then the probability is the
number of configurations of that type divided by the total number of configurations

For a system with energy E, the system consists of states ε0, ε1, ε2… ε∞ where ε0=0 is the lowest energy state (i.e.
the ground state)

If we add up all the molecules and their energy states we get E the total energy…however since we set ε0=0 we
may need to add a constant U(0) to get U the internal energy (e.g. zero point energy for oscillators)

For a system with N molecules: N0 will be in state ε0 ; N1 will be in state ε1 ; N2 will be in state ε2 ;….etc.

N =  Ni E =   i Ni
i i

N is conserved E is conserved
 State Configurations
x! is “x factorial” where
We can write configurations as populations in each state: x! = 1×2×3×4….(x-2) ×(n-1) ×n
population of state #3
{N, 0, 0, 0,….} all ground state with ε0
{N-1,1,0,0,….} one in state 1, the rest in the ground state (state 0)
{N-2, 1, 1, 0, 0, ….} one in state 2, one in state 1, the rest in the ground state….etc…

There are more ways to create the latter configurations (i.e. by swapping molecules between the ground state)

{N, 0, 0, 0,….} {N-1, 1, 0, 0,….} {N-2, 1, 1, 0,….} {N0, N1, N2, N3,….}
N!
1 way to obtain it N ways to obtain it ½N(N-1) ways to obtain it N 0 ! N1 ! N 2 ! N 3 !... ways to obtain it

N!
In the last example, the general case, we can define W as the weight W=
(# distinguishable permutations) N 0 ! N1 ! N 2 ! N 3 !...
 Molecular Partition Functions

Partition function for states Partition function for energy levels

N i e −  i N i gi e −  i
= e −  i
=q =  i =q
g e −  i

N q i N q i

the Boltzmann Equation for states gives the probability of a particular state i, which has an energy of εi and a
partition function which is a sum of the other states I

the Boltzmann Equation for energy levels gives the probability of an energy level, which is the state probability
times a simple degeneracy factor gi (since P ~ ε/T only) and partition function, a sum over all energy levels i
Types of Partition Functions
State Properties
 Types of Partition Functions

Partition functions vary only in expressions for εi, which is determined by the energy levels for a particular type of
state (translational, rotational, vibrational, electronic, spin, etc.)

The underlying statistics (and distributions) are the same.

Example: if the energy levels are evenly spaced then for state n, εn = nε = [0, ε, 2ε, 3ε, 4ε…] with no degeneracy

q =  gi e −  i = e −  ( 0) + e −  ( ) + e −  ( 2 ) + e −  (3 ) + e −  ( 4 ) + ....
i

+ (e ) + (e ) + (e )
−  −  2 −  3 −  4
q = 1+ e + ....

1 1
1 + x + x 2 + x 3 + x 4 + .... = for x  1 q
1− x 1 − e − 
an infinite ladder
 Fractional population in a 2-state system

2-state system: a system consisting of 2 states, state 0 with ε0=0, and state 1 with ε1=ε

state 1: i=1 ; population N1 ; energy ε1=ε

state 0: i=0, population N0 ; energy ε0=0

q =  e −  i = e −  0 + e − 1 = 1 + e − 
i

 1
1 + e −  for i = 0
N i e −  i e −  i 
= = =
N q 1 + e −   e − 
 for i = 1
1 + e − 
 Translational Contribution: x,y,z
The same analysis holds for all directions (x,y,z) such that:  n ,n
x y , nz
=  nx +  n y +  nz

And then (as for any independent energy states we can separate summands):

( )=  −   −   −  
       
−  nx ,n y ,nz −   TX + YT + ZT −  −  − 
q = = =
T T T T T T
T
e e e e e e e e 
X Y Z X Y Z

nx , n y , nz X ,Y , Z X Y Z  X  Y  Z 
volume, V=XYZ
 X  Y  Z  V h
so that qT =         = 3 = β=1/kT
      ( 2 mkT )
12
k= 1.380×10−23 J∙K-1
Boltzmann constant
If q is the total number of accessible states, then qT/N is the number of translational states per molecule
So for a large value, V/NΛ >> 1.
Since V/N is the volume per molecule, and d=(V/N)1/3 is avg distance between molecules
so d>> Λ for this to be true…

The distance between molecules must be much larger than the thermal wavelength for the integration to be valid
 The Rotational Contribution: linear rotor

As we saw earlier, the rotational energy levels for a linear rotor are:  J = hcBJ ( J + 1) for integers J  0
speed of light in a vac. rotational constant (in cm-1)
c= 299 792 458 m / s “characteristic frequency”
where no adjustment is needed since E(0)=0, and the degeneracy is g J = ( 2 J + 1)
 
this means that q =  gJ e
R
J
−  J
=  ( 2 J + 1) e −  hcBJ ( J +1) (for linear molecules except A2 symmetrical)
J =0 J =0
discussed later

for many molecules, if kT is larger than the separation between energy levels then:

 
qJR =  ( 2 J + 1) e −  hcBJ ( J +1)   ( 2 J + 1) e −  hcBJ ( J +1) dJ
J =0 0

which can be x =  hcBJ ( J + 1) 1


1 kT
 hcB 0
−x
integrated with q R
e dx = so qJR 
the substitution ( 2 J + 1) dJ = dx  hcB J
 hcB hcB
 Rotational Contribution: non-linear rotors
Rearranging terms from the last page T dependence

12
1     kT    
32 12
qR =  q =
 hcB  hc A 
R
  2 
 hc   AB 
constants Rotational constant terms

In the case of a non-symmetric top (with frequencies A,B,C), similar arguments lead to

 kT    
32 12

q =
R
  
 hc   ABC 

Rotational constant terms (one for each of three unequal rotational axes)
 Characteristic Rotational Temperature

In the derivation, the major assumption (integration instead of summation) is valid when changes for e-βε from J
to J+1 are small (e.g. lots of populated states), so as to be nearly continuous

Determining when that is valid, requires introducing a characteristic rotational temperature: at a “high”
temperature T>>θR these (continuous) equations are valid

It can be shown that  R hcB

linear − rotor 
k
kT
So that at high temperature the partition function is just T  =
R
(derived earlier)
hcB
We can use the value of θR to assess the molecule:
• Molecules with large moments of inertia (small rot constants ~ low θR) have large partition functions
• This implies small spacing between energy levels, and thus many higher states are populated
• High θR indicates that energy levels are widely spaced, and that most are empty (summation required!)
 Symmetry Number

One further complication to rotational levels, is that rotations which produce indistinguishable states (in
highly symmetric molecules) are forbidden:

An example is 1H2 where the nuclear spin states are either paired or parallel….
in this case J must be even (para) or odd (ortho) for the total molecular wavefunction to be non-zero

There are four possible spin combinations, one para and three ortho, so the average (per molecule) partition
function will be
1  kT    
32 12

qR =  non-linear
  hc   ABC 
1 3
q R
J , avg =  ( 2 J + 1) e −  hcBJ ( J +1) +  ( 2 J + 1) e −  hcBJ ( J +1)
4 J =even 4 J =odd 1 kT
qJR =  linear
which approximates to (since term under summation is identical) under high T  hcB
1
qJR,avg = 
2 J = all
( 2 J + 1) e −  hcBJ ( J +1)
so the term ½=1/σ where σ=2 is the symmetry number
for a symmetric diatomic (linear) molecule
 Vibrational Contribution to Partition Function
For a harmonic oscillator, any given normal mode will have an infinite ladder of states so the partition function
is the same as that given earlier:


1
q =  e −  v 
V

1 − e −  hcv
v
v =0

for a polyatomic molecule, there are 3N-6 normal modes (3N-5 for linear molecule) all of which are treated
independently such that:

qvV1 ,v2 ,v3 ... = qvV1 qvV2 qvV3 qvV4 .... By the same “separation” arguments made previously
 Characteristic Vibrational Temperture

Many common modes have a high vibrational frequency v, such that  hcv  1 at low temperatures

In this case, the ground state, v=0 is primarily populated

Some modes may have a low vibrational frequency such that  hcv  1 , and in this case further
approximations can be made since e-x=1+x+x2/2+x3/3… for x <<1 (also vibrational state is active at r.t.)

1
q =  e −  v 
V
q  V 1

1
1 − e −  hcv 1 − (1 −  hcv...)  hcv
v v
v =0

kT
so for low frequency modes: qvV 
hcv

hcv
so the vibrational characteristic temperature   V
V
such that approximations are valid for T
k
 Electronic Contribution to Partition Function
Electronic energy differences (ΔE) from the ground state are very large so for most cases qE=1

Exceptions occur for degenerate ground states, where qE≈gE×1= gE such as alkali metals which have spin
degeneracy of 2 and thus qE=2

Another example is when atoms/molecules have degenerate ground and low lying degenerate excited states

E.g. NO is a molecule with degenerate ground states 1 2 (unpaired spin) and low energy degenerate excited
2

states the 2 3 2 (also unpaired spin) at an energy ∆E=ε above the ground state (εgnd=0 so εexc= ε)

q E =  e −  i = 2 + 2e − 
E
 Average Energy
State Properties

via the Partition Function

 Average Energy Per Molecule

If the total energy is E, we can then describe the average energy per molecule (average energy state) as E/N

E 1 N N i e −  i
E =   i Ni  = =    i Ni =  i i where =
via the
i N N i N N q Boltmann Eq.
i
(if εgs=0)
 i e −  1
 = =   i e −  i
i

so then
q q i at constant volume, e.g. no change in state energies with change in T
i

if we notice that  i e −  = −i

d
(
d −  i
e ) then we can make a substitution Average energy ONLY in
terms of partition function

1 
 =  −
d e −  i
( )  = − 1  d 1 dq d ln q  d ln q 
q i  d  q d
e −  i
=− 
q d
=−
d
 =  gs − 
 
 i  d V
constant for assumption about εgs
Ensembles
State Properties
 Canonical Ensembles
The microcanonical system (1 member, N molecules) is the focus of our study to this point (isolated, same V,E,N)

It is important to remember that ensembles are imaginary replications of the system, each version may be slightly
different with variations based on probability of configurations (the number of members can be infinite)

The number of members with E=Ei (energy of a member of

the ensemble, i) in an canonical ensemble is denoted Ñi

So to recap:
Ñ is the number of members (replicas of system, not N),
Ẽ is total energy of all members,
Ŵ is the weight of the configuration (Ñ1, Ñ2, Ñ3….) with (E1, E2, E3….)
a canonical ensemble
by the same arguments that gave us the weight of configurations for a single system of states
N! N i e −  Ei
where Q =  e −  Ei
Q is the canonical
W= =
N 0 ! N1 ! N 2 ! N 3 !.... N Q partition function
i
 Energy Density of States

The canonical distribution function gives the probability of occurrence of members in a single state i of energy Ei

In a large ensemble it is likely that there are many members with energy near (but not quite equal to) Ei, we can
define the energy density of states as the number of states in an energy range divided by the width of the range

# states per molecule (or per volume)

in interval [E, E+dE]

dΩ(E)/dE

= #/J or #/cm-3/J
Working with the Canonical Ensemble
State Properties
 Independent Molecules Revisited

In any system, the total energy of N molecules in state i will be Ei=εi(1)+εi(2)+εi(3)… where εi(1) is the energy of
molecule #1 in configuration state i, etc.

Q =  e −  Ei =  e
−   i (1) + i ( 2 ) + i ( 3).... i ( N ) 
The canonical partition function is then
i i
Rather than cycle through each configuration state i, for independent molecules we can rewrite the index as
cycling through all the states of each molecule (i.e. each molecule’s partition function independently evaluated)

 −  J   −  J   −  J   −  J 
Q = e      = ( ) q =  e −  i
N
e e  .... e  q remember that
 j 1  j 2  j 3  j N i

this is true as long as each molecule is distinguishable from the others so that Q = qN
qN
if they are indistinguishable, then Q must be divided by indistinguishable permutations (there are N!) so Q =
N!
same molecule/atom, able to switch positions at random, like in a gas
 Partition Functions and Cases
Partition functions thus far:

A single independent molecule q =  e −  i

i

Molecules/atoms are distinguishable and independent Q = qN

e.g. perfect crystalline solid

qN
Molecules are indistinguishable and independent Q=
e.g. perfect gas N!

Molecules are interacting (indist. or dist.) use Q (but we did not derive any expression)
e.g. imperfect gas, liquids, some solids
 Mean Energy of a Canonical Ensemble
q assumed that molecules were independent, non-interacting and isolated (i.e. N single molecular partition functions)

Q accepts they may be interacting (and is thus more likely to be true for real gases and condensed phases)

Total energy is Ẽ and Ñ is the number of members, so average energy per system is <E>= Ẽ/ Ñ

NNi i ee−−EEi i
Pi = == Q ==  e−  Ei
where Q
where
NN QQ ii

so it follows that the average can be calculated in the same manner as <ε> (average energy per molecule)

1 1  dQ   d ln Q  expectation value for

E =  PE =  E e −  Ei
E =−   = −   system energy in a
Q  d  V 
i i i
i Q i  d V closed system

probability of the system having energy Ei this is the average energy of a system at
thermal equilibrium with the surroundings
HW Questions
State Properties