201105spring 2001

20.110J / 2.772J / 5.
601J
Thermodynamics of Biomolecular Systems
Instructors: Linda G. Griffith, Kimberly Hamad-Schifferli, Moungi G. Bawendi, Robert W. Field
20.110/5.60 Fall 2005
Lecture #1
page 1
Introduction to Thermodynamics
Thermodynamics:
Describes macroscopic properties of
equilibrium systems
Entirely Empirical
Built on 4 Laws and simple mathematics
0th Law Defines Temperature (T)
1st Law Defines Energy (U)
2nd Law Defines Entropy (S)
3rd Law Gives Numerical Value to Entropy
These laws are UNIVERSALLY VALID, they cannot be circumvented.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
page 2
Definitions:
System: The part of the Universe that we choose to study
Surroundings: The rest of the Universe
Boundary: The surface dividing the System from the
Surroundings
Systems can be:

Open: Mass and Energy can transfer between the System
and the Surroundings
Closed: Energy can transfer between the System and the
Surroundings, but NOT mass
Isolated: Neither Mass nor Energy can transfer between
the System and the Surroundings
Describing systems requires:

A few macroscopic properties: p, T, V, n, m,
Knowledge if System is Homogeneous or Heterogeneous
Knowledge if System is in Equilibrium State
Knowledge of the number of components
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
page 3
Two classes of Properties:

Extensive: Depend on the size of the system
(n,m,V,)
Intensive: Independent of the size of the system
V
(T,p, V = ,)
n
The State of a System at Equilibrium:

Defined by the collection of all macroscopic properties that
are described by State variables (p,n,T,V,)
[INDEPENDENT of the HISTORY of the SYSTEM]
For a one-component System, all that is required is n and 2
variables. All other properties then follow.
V = f (n , p, T)
or
p = g ( n , V, T )
These are Equations of States

E.g. The Ideal gas law: pV = nRT
The Virial expansion:
B (T ) C (T )
pV
=1+
+
+"
RT
V
V 2
The van der Waals equation of state:

p + a V b = RT
V 2
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
Notation:
3 H2 (g, 1 bar, 100 C)
3 moles
(n=3)
gas
p=1 bar
2 Cl2 (g, 5 L, 50 C)
Change of State:
T=100 C
5 Ar (s, 5 bar, 50 K)
(Transformations)
Notation:
3 H2 (g, 5 bar, 100 C) = 3 H2 (g, 1 bar, 50 C)
initial state
final state
2 H2O (, 1 bar, 50 C) = 3 H2O (g, 1 bar, 150 C)
initial state
final state
page 4
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
page 5
Path: Sequence of intermediate states
p (bar)
1
50
100
T (K)
Process: Describes the Path

- Reversible
(always in Equilibrium)
- Irreversible (defines direction of time)
- Adiabatic
(no heat transfer between sys. and surr.)
- Isobaric
(constant pressure)
- Isothermal (constant temperature)
- Constant Volume
-
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
page 6
Thermal Equilibrium (Heat stops flowing)
When a hot object is placed in thermal contact with a cold object,

heat flows from the warmer to the cooler object. This continues
until they are in thermal equilibrium (the heat flow stops). At this
point, both bodies are said to have the same temperature.
This intuitively straightforward idea is formalized in the 0th Law of
thermodynamics and is made practical through the development of
thermometers and temperature scales.
ZEROth LAW of Thermodynamics

If
then
and
are in thermal equilibrium and
and
are in thermal equilibrium,
and
are in thermal equilibrium.
Consequence of the zeroth law:
acts like a thermometer, and

at the same temperature.
, and
are all
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
Operational definition of temperature (t)

Need :
(1)
(2)
(3)
(4)
substance
property that depends on t
reference points
interpolation scheme between ref. pts.
Example: Ideal Gas Thermometer with the Celsius scale.

Based on Boyles Law
lim pV
p 0
= constant = f (t )
for fixed t
depends on t
the substance is a gas

f (t ) is the property
the boiling point (tb = 100C ) and freezing point (tf = 0C ) of
water are the reference points
the interpolation is linear
p 0
f(t)=f(0 C)(1+At)
Experimental result:
A = 0.0036609 = 1/273.15
-273.15
Note:
0 100 C
(t
= 273.15 C ) is special
t = 273.15 C is called the absolute zero,
page 7
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #1
page 8
This suggests defining a new temperature scale (Kelvin)

T (K ) = t ( C ) + 273.15
T = 0K corresponds to absolute zero (t = 273.15 C )

Better reference points used for the Kelvin scale today are
T = 0K (absolute zero) and Ttp = 273.16K (triple point of H2O)
Ideal Gases
Boyles Law and the Kelvin scale
lim pV
p 0
lim pV
tp
p 0
=
T RT
273.16
valid for all gases for p 0
define
the gas constant
An ideal gas obeys the expression pV = RT at all pressures

( the gas molecules do not interact)
lim ( pV )
tp
p 0
= 8.31451 J
R=
273.16
K mol
(gas constant)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Work:
Lecture #1
page 9
w = F A
applied force
Expansion work
distance
pext
pext
F = pext A
w = ( pext A ) A = pext V
convention:
Heat:
Having a - sign here implies w > 0 if V < 0 , that

is, positive work means that the surroundings do
work to the system. If the system does work on the
surroundings ( V > 0 ) then w < 0 .
q
That quantity flowing between the system and the
surroundings that can be used to change the temperature
of the system and/or the surroundings.
Sign convention:
If heat enters the system, then it is

positive.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #2
page 1
Work, Heat, and the First Law

Work:
Epansion work: w = ( pext A ) A = pext V
If pext is not constant, then we have to look at infinitesimal changes
d- means this is not an exact differential
d-w = pext dV
Integral
w = 1 pext dV
2
depends on the path!!!
Other kinds of work

Surface work: d-w = ext dA where ext is the surface tension (J/m2)
and dA is the differential change in area. This is the
work to change surface area.
Elongation work: d-w = fd A where f is the force per unit length and
d is the length differential. This is important for
discussing changing the length of polymers or DNA.
Electrostatic work: d- w =Vde where V is a fixed potential and de is
the change in charge.
Path dependence of w
Example: assume a reversible process so that pext = p

Ar (g, p1, V1)
Compression
Ar (g, p2, V2)
V1 > V2 and p1 < p2
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #2
page 2
p ext= p 1
p ext= p 2
compression
p 1,V1
p 2,V2
initial
final
Two paths:
(1) First V1 V2 at p = p1
then p1 p2 at V = V2
(2)
Ar(g, p1, V1) = Ar(g, p1, V2) = Ar(g, p2, V2)
p
p2
p1
Ar(g, p1, V1) = Ar(g, p2, V1) = Ar(g, p2, V2)
final
(2)
init.
(1)
V1
V2
w(1) = V pext dV V pext dV

V2
V2
First p1 p2 at V = V1
then V1 V2 at p = p2
w(2) = V pext dV V pext dV
= V p1dV = p1 (V2 V1 )
V1
V2
= V p2dV = p2 (V2 V1 )
V2
V2
1
w(1) = p1 (V1 V2 )
w(2) = p2 (V1 V2 )
(Note w > 0, work done to system to compress it)
w(1) w(2) !!!

Note for the closed cycle [path (1)] - [path (2)],
w is not a state function
d-w 0
cannot write w = f(p,V)
closed cycle
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
WORK
Lecture #2
Work (w) is not a function of state.

For a cyclic process, it is possible for d-w 0
state 1
HEAT
page 3
state 2
Heat (q), like w, is a function of path. Not a state function

It is possible to have a change of state
( p 1 , V1 , T 1 ) = (p 2 , V2 , T2 )
or
adiabatically
(without heat transferred)
nonadiabatically.
Historically measured in calories

[1 cal = heat needed to raise 1 g H2O 1C,
from 14.5C to 15.5C]
The modern unit of heat (and work) is the Joule.
1 cal = 4.184 J
Heat Capacity
- connects heat with temperature
q = CpathdT
or
q
C path =
dT path
heat capacity is path dependent
Constant volume:
CV
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #2
Constant pressure:
q=
Equivalence of work and heat
page 4
Cp
C pathdT
path
[Joule (1840s)]
Joule showed that its possible to raise the temperature of H2O
(a) with only heat
T1 T2
(b) with only work
T1 T2
(weight falls &

churns propeller)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #2
page 5
Experimentally it was found that
(d-w +d-q ) = 0
The sum (w + q) is independent of path
This implies that there is a state function whose differential is

w + q
We define it as U, the internal energy or just energy
dU = w + q
For a cyclic process
dU
=0
For a change from state 1 to state 2,

2
U = dU = U2 U1 = q + w
1
does not depend on path
each depends on path individually, but not the sum
For fixed n, we just need to know 2 properties, e.g. (T, V), to fully
describe the system.
U = U (T ,V ) )
So
U is an extensive function (scales with system size).

U =
U
n
is molar energy (intensive function)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #2
page 6
THE FIRST LAW
Mathematical statement:
dU = d-q +d-w
or
U = q + w
or
d-q = d-w
Corollary:
Conservation of energy
Usystem = q + w
Usurroundings = q w
Uuniverse = Usystem + Usurroundings = 0
Clausius statement of 1st Law:

The energy of the universe is conserved.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #3
page 1
EXPANSIONS, ENERGY, ENTHALPY

Isothermal Gas Expansion
( T = 0)
gas (p1, V1, T) = gas (p2, V2, T)

Irreversibly (many ways possible)
(1)
Set pext = 0
p= 0
T
p= 0
p 1,V1
p 2,V2
v2
w(1) = pext dV = 0
V1
(2)
Set pext = p2
p2
T
p2
T
p 2,V2
p 1,V1
v2
w(2) = p2dV = p2 (V2 V1 )

V1
p
p1
p2
V1
-w(2)
V2
Note, work is negative: system expands against surroundings
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
(3)
Lecture #3
page 2
Carry out change in two steps
gas (p1, V1, T) = gas (p3, V3, T) = gas (p2, V2, T)
p1 > p3 > p2
p2
p3
p3
T
p 2,V2
p 3,V3
p 1,V1
v3
v2
V1
V3
w(3) = p3dV p2dV = p3 (V3 V1 ) p2 (V2 V3 )

p
p1
More work delivered to

surroundings in this case.
p3
p2
V1 V3
V2
-w(3)
(4)
Reversible change
p = pext throughout
V
wrev = 2 pdV
V1
p1
p2
V1
V2
rev
For ideal gas:

V
wrev = 2
V1
Maximum work delivered to

surroundings for isothermal gas
expansion is obtained using a
reversible path
V
p
nRT
dV = nRT ln 2 = nRT ln 2
V
V1
p1
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #3
page 3
The Internal Energy U
dU = d-q + d-w
(First Law)
dU = C pathdT pext dV
And U (T ,V
U
U
dU =
dT +
dV
Some frequent constraints:
dU = d-qrev + d-wrev = d-qrev pdV
Reversible
Isolated
d-q = d-w = 0
Adiabatic
d-q = 0
Constant V
(p = pext )
dU = d-w
reversible
-pdV
w = 0 dU = d-qV
Constant V
U
U
dU =
dT +
dV
T V
V T
U
d- qV =
dT
T V
But also
d-qV = CV dT
So
U
= CV
T V
very important result!!
V T
dU = CV dT +
dV
what is this?
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #3
U
)
V T
Joule Free Expansion of a Gas
gas
(to get
vac
gas (p1, T1, V1) = gas (p2, T2, V2)

Since q = w = 0
Recall
V T
V T
q=0
Expansion into Vac.

(pext=0)
w=0
Constant U
dV = 0
dVU = CV dTU
= CV
V T
Joule did this.
Adiabatic
dU or U = 0
dU = CV dT +
page 4
V U
measure in Joule exp't!
V U
T
T
dU = CV dT CV J dV
=
J
V U V U
Joule coefficient
lim
V 0
For Ideal gas
J = 0
dU = CV dT
U(T)
exactly
Always for ideal gas
only depends on T
The internal energy of an ideal gas depends only on temperature

Consequences
U = 0
U = CV dT
For all isothermal expansions or

compressions of ideal gases
For any ideal gas change in state
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #3
H(T,p)
Enthalpy
page 5
H U + pV
Chemical reactions and biological processes usually take place under

constant pressure and with reversible pV work. Enthalpy turns out
to be an especially useful function of state under those conditions.
reversible
gas (p, T1, V1)
gas (p, T2, V2)
const . p
U1
U2
U = q + w = qp p V
U + p V = qp
define as H
U + ( pV ) = qp
H U + pV
Choose
H (T , p )
H
H = q p
What are
T p
H
T p
(U + pV ) = qp
for a reversible constant p process
H
H
dH =
dp
dT +
T p
p T
H
and
?
p T
for a reversible process at constant p (dp = 0)
dH = q p
H
and dH =
dT
T p
H
dT
T p
q p =
but
= Cp
T p
qp = C p dT
also
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
p T
Lecture #3
page 6
Joule-Thomson expansion
adiabatic, q = 0
porous partition (throttle)
gas (p1, T1)
w = pV
1 1 pV
2 2
gas (p2, T2)
U = q + w = pV
1 1 pV
2 2 = ( pV
U + ( pV ) = 0
)
(U + pV ) = 0
H = 0
Joule-Thomson is a constant Enthalpy process.

H
dp
p T
dH = C p dT +
H
T
= C p
p T
p H
Define
H
dpH
p T
C p dT =
can measure this
p H
T
T
lim
=
JT Joule-Thomson Coefficient
p 0
p H p H
= C p JT
T
and
dH = C p dT C p JT dp
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #3
U(T),
For an ideal gas:
page 7
pV=nRT
H U (T ) + pV = U (T ) + nRT
H (T )
only
H
depends on T, no p dependence
= JT = 0 for an ideal gas

T
For an ideal gas C p = CV + R

H
,
T p
T V
Cp =
CV =
H = U + pV ,

H
U
=
+R
T
T
p
p

U V
C p = CV +

+R
V T T p
= 0 for ideal gas
C p = CV + R
pV = RT
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #4
page 1
EXPANSIONS, THERMODYNAMIC CYCLES
Reversible Adiabatic Expansion (or compression) of an Ideal Gas

1 mole gas (V1,T1) = 1 mole gas (V2,T2)
q = 0
Ideal gas
adiabatic
From 1st Law
CV dT = pdV
T2
CV
T1
p =RT
V
V dV
dT
= R
V V
T
Define
dU = -pd V
CV
w = -pd V
CvdT = -pdV along path
dT
dV
= R
T
V
R CV
Cp

CV
Reversible
dU = Cv d T
T2 V1
=
T1 V2
Cp
C p CV =R for i.g.
T2 V1 CV
=
T1 V2
T2 V1
=
T1 V2
For monatomic ideal gas:
CV = R
2
5
( > 1 generally)
=
5
3
Cp = R
2
[More generally the heat capacity for a molecular species has a

temperature dependence that can be approximated as
C p (T ) = a + bT + cT 2 with a, b, and c tabulated.]
In an adiabatic expansion (V2 > V1), the gas cools (T2 > T1).
And in an adiabatic compression (V2 < V1), the gas heats up.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #4
pV
T =
R
For an ideal gas (one mole)

pV
page 2
p2 V1
=
p1 V2
pV
1 1 = pV
2 2
is constant along a reversible adiabat
Recall, for an isothermal process T = constant
pV = constant
isotherm
pV=constant
p1
pV=const.
p2
adiabat
V2adiabat < V2isotherm

because the gas
cools during reversible
adiabatic expansion
V2ad V2iso
V1
Irreversible Adiabatic Expansion of an ideal gas against a constant

external pressure
1 mol gas (p1,T1) = 1 mol gas (p2,T2)
adiabatic
Constant pext = p2
Ideal gas
1st Law
Integrating:
Using pV = RT
(pext=p2)
q = 0
w = -p2d V
dU = Cv d T
dU = -p2 dV
Cv d T = - p 2 dV
Cv ( T2 - T 1 ) = - p 2 ( V2 - V1 )
T2 (CV + R ) =T1 CV +
p2
R
p1
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #4
p2 < p1
Note
T2 < T1
(-wrev) > (-wirrev)
Note also
page 3
Again, expansion cools
as expected, less work is

recovered through an
irreversible process
Thermodynamic Cycles
Reversible Ideal Gas processes:

Find U , H , q , w
p
p1
(I)
(T1)
p1
A (isotherm)
p2
p3
(T1)
(rev.
adiabat)
(T2)
V1
(II)
C (const. V)
p2
V2
(T1)
D (const. p )
A
(isotherm)
V1
(T3)
E (const. V)
(T1)
V2
For (I)
1gas(p1,V1,T1)
1gas(p2,V2,T1)
1gas(p3,V2,T2)
There are two paths from initial to final states (A) and (B+C). As far
as functions of states (e.g. U, H) are concerned it doesnt matter
which path is taken.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
[A]
Lecture #4
1 mol gas (p1,V1,T1)
const. T
V
wA = RT1 ln 2
V1
[B]

Adiabat:
Ideal gas:
1st Law:
[C]
UA = 0
Ideal gas isotherm:
page 4
HA = 0
V
qA = RT1 ln 2
V1
rev.adiabat
qB = 0
UB = CV (T2 T1 )
HB = C p (T2 T1 )
w B = CV (T2 T1 )
reversible
const. V
Constant V:
wC = 0
1st Law:
qC = CV (T1 T2 )
Ideal gas:
UC = CV (T1 T2 )
HC = C p (T1 T2 )
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
[A]
Lecture #4
vs.
UA = 0
HA = 0
qB + qC = CV (T1 T2 ) qA
wB + w C = CV (T2 T1 ) w A
UD = CV (T3 T1 )
HD = C p (T3 T1 )
UE = CV (T1 T3 )
HE = C p (T1 T3 )
[E]
[A]
UA = 0
HA = 0
V
qA = RT1 ln 2
V1
V
w A = RT1 ln 2
V1
[B] + [C]
UB + UC = 0 = UA
HB + HC = 0 = HA
V
qA = RT1 ln 2
V1
V
w A = RT1 ln 2
V1
[D]
page 5
vs.
qD = C p (T3 T1 )
wE = 0
wD = R (T3 T1 )
qE = CV (T1 T3 )
[D] + [E]
UD + UE = UA
HD + HE = HA
qD + qE = R (T3 T1 ) qA
wD + w E = R (T3 T1 ) w A
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 1
Thermochemistry
Much of thermochemistry is based on finding easy paths to
calculate changes in enthalpy, i.e. understanding how to work
with thermodynamic cycles.
Goal:
To predict H for every reaction, even if it

cannot be carried out in the laboratory
The heat of reaction
Hrx
is the H for the isothermal
reaction at constant pressure (the complete transfer from reactants

to products, not to some equilibrium state).
Fe2O3(s,T,p) + 3H2(g,T,p) = 2Fe(s,T,p) + 3H2O(l,T,p)
e.g.
Hrx (T , p ) = 2HFe (T , p ) + 3HH O (T , p ) 3HH (T , p ) HFe O (T , p )

2
2 3
[Hrx = H ( products ) H ( reactants )]
We cannot know H values because enthalpy, like energy, is not an

absolute scale. We can only measure differences in enthalpy.
Define a reference scale for enthalpy
H (298.15K, 1 bar) 0
e.g.
For every element in its most stable

form at 1 bar and 298.15K
HH (g) (298.15K ) = 0
2
HC(graphite) (298.15K ) = 0
The means 1 bar
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Hf (298.15K ) :
Lecture #5
page 2
The heat of formation is the heat of reaction
to create 1 mole of that compound from its constituent elements in

their most stable forms.
Example (T = 298.15 K)
H2 (g,T,1 bar) + Br2 (l,T,1 bar) = HBr (g,T,1 bar)
Hf,HBr (T ) = Hrx (T ,1bar ) = HHBr

( g,T )
1
1
HH2 ( g,T ) HBr2 ( l,T )
2
2

0 - elements in most stable forms
We can tabulate Hf (298.15K ) values for all known compounds.

We can calculate Hrx (298.15K ) for any reaction.
e.g. (T=298.15K)
CH4 (g,T,1 bar) + 2O2 (g,T,1 bar) = CO2 (g,T,1 bar) + 2H2O(l,T, 1 bar)
First decompose reactants into elements
Second put elements together to form products
Use Hesss law [An example of a thermodynamic cycle applied

to thermochemistry]
Reactants
H rx
i Hf,i ( reactants )
i
Products
i i Hfi ( products )
,
Elements
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 3
CH4 (g,T,1 bar) = Cgraphite (s,T,1 bar) + 2H2(g,T,p)

HI
2O2 (g,T,1 bar) = 2O2 (g,T,1 bar)
HII
Cgraphite (s,T,1 bar) + O2 (g,T,1 bar) = CO2 (g,T,1 bar) HIII
2H2(g,T,p) + O2 (g,T,1 bar) = 2H2O(l,T, 1 bar)
HIV
____________________________________________
CH4 (g,T,1 bar) + 2O2 (g,T,1 bar) = CO2 (g,T,1 bar) + 2H2O(l,T, 1 bar)
Hrx = HI + HII + HIII + HIV
HI = HC + 2HH HCH = Hf,CH

2
HII = HO HO = 0
2
HIII = HCO HC HO = Hf,CO

2
HIV = 2HH O 2HH HO = 2Hf,H O

2
Hrx = 2Hf,H O + Hf,CO Hf,CH

2
In general,
Hrx = i Hf,i (products ) i Hf,i (reactants )

i
stoichiometric coefficient
H at constant p and for reversible pV process is H = q p
The heat of reaction is the heat flowing into the

reaction from the surroundings
If
Hrx < 0,
qp < 0
heat flows from the reaction to

the surroundings (exothermic)
If
Hrx > 0,
qp > 0
heat flows into the reaction from

the surroundings (endothermic)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 4
Calorimetry
The objective is to measure
Reactants (T1)
Hrx (T1 )
isothermal
constant p
Constant pressure
Products (T1)
(for solutions)
adiabatic
constant p
reaction
calorimeter
React. (T1) + Cal. (T1)
Hrx
Prod. (T1) + Cal. (T1)
constant p
HI
adiabatic
HII
constant p
constant p
I)
HI
II)
HII
adiabatic
constant p
constant p

______________________________________________________
Hrx (T1 )
constant p
Hrx (T1 ) = HI + HII
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 5
(I) Purpose is to measure (T2 - T1)

Adiabatic, const. p
qp = 0
H I = 0
(II) Purpose is to measure heat qp needed to take prod. + cal.

from T2 back to T1.
T1
qp = C p ( Prod . + Cal .) dT = HII

T2
T2
Hrx (T1 ) = T C p ( Prod . + Cal . )dT

1
Constant volume
(when gases involved)
adiabatic
constant V
reaction
calorimeter
Urx
constant V
UI
adiabatic
UII
constant V
constant V

I)
U I
II)
UII
adiabatic
constant V
constant V

______________________________________________________
Urx (T1 )
constant V
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 6
Urx (T1 ) = UI + UII
(I) Purpose is to measure (T2 - T1)

Adiabatic, const. V
qV = 0
UI = 0
(II) Purpose is to measure heat qV needed to take prod. + cal.

from T2 back to T1.
T1
qV = CV ( Prod . + Cal . )dT = UII

T2
T2
Urx (T1 ) = T CV ( Prod . + Cal . )dT

1
Now use
H = U + pV
or
H = U + ( pV
Assume only significant contribution to ( pV
( pV ) = R (nT
Ideal gas
Isothermal
T =T1
)
is from gases.
( pV ) = RT1 ngas
Hrx (T1 ) = Urx (T1 ) + RT1 ngas

T2
Hrx (T1 ) = CV ( Pr od . + Cal . ) dT + RT1 ngas

T1
e.g.
4 HCl(g) + O2(g) = 2 H2O(l) + 2 Cl2(g)
T1 = 298.15 K
Urx (T1 ) = -195.0 kJ
ngas = -3 moles
Hrx (T1 ) = -195.0 kJ + (-3 mol)( 298.15 K)(8.314x10-3 kJ/K-mol)
= -202.43 kJ
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #5
page 7
Temperature dependence of Hrx
Suppose know Hrx at some temperature T1 (e.g. at 298.15 K) and we

want to know it at some other temperature T2.
Generally the difference is small often we assume that there is no
temperature dependence if the difference between T1 and T2 is
small. If the difference between T1 and T2 is large enough, we can
calculate Hrx(T2) from the heat capacities of the reactants and
products (assuming no phase change in any component).
Hrx (T2 )
Reactants (T2)
constant p
Products (T2)
HI
HII
constant p
constant p
Hrx (T1 )
Reactants (T1)
constant p
Products (T1)
Hrx (T2 ) = HI + Hrx (T1 ) + HII

T1
T2
T2
T1
Hrx (T2 ) = Hrx (T1 ) + C p (react .)dT + C p ( prod .)dT

T2
Hrx (T2 ) = Hrx (T1 ) + C p ( prod .) C p (react .) dT

T1
T2
Hrx (T2 ) = Hrx (T1 ) + C pdT

T1
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
page
The Second Law
First Law
U = q + w ,
showed the equivalence of work and heat

dU = 0 for cyclic process q = w
Suggests engine can run in a cycle and convert heat into useful work.
Second Law
Puts restrictions on useful conversion of q to w
Follows from observation of a directionality to natural

or spontaneous processes
Provides a set of principles for

- determining the direction of spontaneous change
- determining equilibrium state of system
Need a definition:
Heat reservoir
Definition: A very large system of uniform

T, which does not change regardless of the
amount of heat added or withdrawn.
Also called a heat bath. Real systems can come close to this
idealization.
Two classical statements of the Second Law:
Kelvin
Clausius
and a Mathematical statement
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
page
I. Kelvin: It is impossible for any system to operate in a cycle that

takes heat from a hot reservoir and converts it to work in the
surroundings without at the same time transferring some heat to a
colder reservoir.
q> 0
w< 0
-w= q
T1 (hot)
q
T1 (hot)
q1
-w
q 1> 0
w< 0
q 2< 0
-w q 1= -w-q 2
-q 2
IMPOSSIBLE!!
T2 (cold)
OK!!
II. Clausius:
It is impossible for any system to operate in a cycle
that takes heat from a cold reservoir and transfers it to a hot
reservoir without at the same time converting some work into heat.
q 2> 0 T1 (hot)
q 1< 0
-q 1
-q 1= q 2
T1 (hot)
q2
q2
IMPOSSIBLE!! T2 (c old)
Alternative Clausius statement:
-q 1
T2 (c old)
q 2> 0
w> 0
q 1< 0
w -q 1= w+ q 2
OK!!
All spontaneous processes are

irreversible.
(e.g. heat flows from hot to cold spontaneously and irreversibly)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
qrev
T
qrev = 0
Mathematical statement:
page
is a state function = dS
and
dS =
qirrev < 0
qrev
T
S ENTROPY
dS
=0
S = S2 S1 =
2 qirrev
qrev
>
1
T
T
irrev
rev
for cycle [1]
[2]
[1]
1 qrev
q irrev
q
+
= irrev < 0
2 T
T
T
2 qirrev
qirrev
S < 0 S >
1
T
T
Kelvin and Clausius statements are specialized to heat engines.

Mathematical statement is very abstract.
Lets Link them through analytical treatment of a heat engine.

The Carnot Cycle
a typical heat engine
All paths are reversible

T1 (hot)
q1
isotherm (T1)
adiabat
adiabat
4
isotherm (T2)
q2
T2 (cold)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
page
12
isothermal expansion at T1 (hot)
23
adiabatic expansion (q = 0)
34
isothermal expansion at T2 (cold) U = q2 + w 2
41
Kelvin:
U = w1
adiabatic compression (q = 0)
Efficiency =
1st Law
U = q1 + w 1
U = w 2
work output to surroundings (w 1 + w1 + w 2 + w 2 )

=
heat in at T1 (hot)
q
v dU
= 0 q1 + q2 = (w1 + w1 + w 2 + w 2 )
Efficiency =
q1 + q2
q
=1+ 2
q1
q1
q2 < 0 Efficiency < 1 (< 100%)

-w = q1 = work output
Note: if the cycle were run in reverse, then q1 < 0, q2 > 0, w > 0.
Its a refrigerator!
Carnot cycle for an ideal gas
12
23
U = 0; q1 = w1 = 1 pdV = RT1 ln 2
V
q = 0; w1 = CV (T2 T1 )
Rev. adiabat
T2 V2
=
T1 V3
V4
V3
U = 0; q2 = w2 = 3 pdV = RT2 ln
41
q = 0; w2 = CV (T1 T2 )
34
Rev. adiabat
T1 V4
=
T2 V1
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
page
q2 T2 ln (V4 V3 )
=
q1 T1 ln (V2 V1 )
1
V1

V4
T V
= 2= 2
T1 V3
or
q1 q2
+
=0
T1 T2
V4 V1
=
V3 V2
qrev
T
q2
T
= 2
q1
T1
=0
this illustrates the link between heat engines to the

mathematical statement of the second law
Efficiency
=1+
q2
T
=1 2
q1
T1
For a heat engine (Kelvin):
100% as T2 0 K
q1 > 0, w < 0, T2 < T1
T T2
q1
Total work out = w = q1 = 1

T1
For a refrigerator (Clausius):
q2 > 0, w > 0, T2 < T1
( w ) < q1
Note: In the limit T2 0 K, (-w) q1, and 100% conversion of

heat into work. 3rd law will state that we cant reach this limit!
Total work in
But
q1
q
= 2
T1
T2
T T
= w = 2 1 q1
T1
T T
w = 1 2 q2
T2
Note: In the limit T2 0 K, w . This means it takes an infinite

amount of work to extract heat from a reservoir at 0 K 0 K
cannot be reached (3rd law).
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
page
The efficiency of any reversible engine has to be the same as the

Carnot cycle, this can be shown by running the reversible engine
as a refrigerator, using the work output of a Carnot engine to
drive it so that the total work out is zero, and showing that, if
the efficiency of the reversible engine is higher, then the second
law is broken.
Additionally:
We can approach arbitrarily closely to any cyclic process using a

series of only adiabats and isotherms.
So, for any reversible cycle
This is consistent with the mathematical statement of the second

law, which defines Entropy, a function of state, with
dS =
Note:
qrev
T
S = S2 S1 =
qrev
T
=0
qrev
T
Entropy is a state function, but to calculate S from q

requires a reversible path.
An irreversible Carnot (or any other) cycle is less efficient than a

reversible one.
p
irreversible
isotherm with p ext = p 2
adiabat
adiabat
isotherm (rev.)
12
( w )irrev < ( w )rev wirrev > w rev
U = qirrev + wirrev = qrev + w rev
qirrev < qrev
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
**
Lecture #6
page
An irreversible isothermal expansion requires less heat **

than a reversible one.
irrev = 1 +
also
q2rev
q2rev
<
1
+
= rev
q1irrev
q1rev
qirrev qrev
T
<
(q2 < 0)
qirrev
<0
This leads to the Clausius inequality
v T
qrev
v T = 0
contains
qirrev < 0
v T
Important corollary: The entropy of an isolated system never

decreases
(A): The system is isolated and
irreversibly (spontaneously) changes
from [1] to [2]
(A) irreversible
2
(B) reversible
(B): The system is brought into contact with a heat

reservoir and reversibly brought back from [2] to [1]
qirrev = 0
Path (A):
Clausius
v T
(isolated)
qirrev
T
=0 !
qrev
T
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #6
qrev
T
page
= S1 S2 = S 0
S = S2 S1 0
This gives the direction of spontaneous change!
For isolated systems
2
But!
Ssurroundings
S > 0
Spontaneous, irreversible process
S = 0
Reversible process
S < 0
Impossible
S = S2 S1
independent of path
depends on whether the process is
reversible or irreversible
(a)
Irreversible:
Consider the universe as an isolated system

containing our initial system and its
surroundings.
Suniverse = Ssystem + Ssurroundings > 0
(b)
Ssurr > Ssys
Reversible:
=0
Suniv = Ssys + Ssurr
= Ssys
Ssurr
Suniverse 0 for any change in state (> 0 if irreversible, = 0 if reversible)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 1
Entropy, Reversible and Irreversible Processes and

Disorder
Examples of spontaneous processes
T1
Connect two metal blocks thermally in

an isolated system (U = 0)
T2
Initially
T1 T2
dS = dS1 + dS2 =
q1
T1
q2
T2
= q1
(T2 T1 )
TT
1 2
( q1 = q2 )
dS > 0 for spontaneous process
if T2 >T1
T2 <T1
gas
V
vac.
V
q1 > 0
q1 < 0
in both cases heat flows

from hot to cold as expected
Joule expansion with an ideal gas

1 mol gas (V,T)
U = 0
adiabatic
q=0
1 mol gas (2V,T)
w=0
Need a reversible path to compute S from q! Close the cycle and go

back to the initial state reversibly and isothermally
S = Sbackwards
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 2
qrev 0
1 mol gas (2V,T) = 1 mol gas (V,T)

Sbackwards =
qrev
w
T
V
2
RdV
1
= R ln
V
2
spontaneous
S = R ln2 > 0
IMPORTANT!! To calculate S for the irreversible process, we

needed to find a reversible path so we could determine
qrev and
qrev
T
Mixing of ideal gases at constant T and p
nA A (g, VA, T) + nB A (g, VB, T) = n (A + B) (g, V, T)

nA
VA
nB
VB
spontaneous
mixing
n = nA + nB
V = VA + VB
To calculate Smix , we need to find a reversible path between

the two states.
constant T
A+ B
piston
permeable
to A only
piston
permeable
to B only
Sdemix = Smix
back to initial state
function of state
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 3
For demixing process
U = 0
qrev = w rev = pAdVA + pB dVB
work of compression of each gas
Sdemix =
VA p dV
VB p dV
V
dqrev
V
= A A + B B = nAR ln A + nB R ln B
V
V
T
T
T
V
V
Put in terms of mole fractions

Ideal gas
XA =
VA
V
XB =
XA =
nA
n
XB =
nB
n
VB
V
Sdemix = nR [XA ln X A +XB ln XB ]

Smix = nR [XA ln XA +XB ln XB ]
Since XA , XB < 1
Smix > 0
mixing is always spontaneous
The mixed state is more disordered or random than the

demixed state.
Smixed > Sdemixed
This is a general result:

Entropy is a measure of the disorder of a system
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 4
For an isolated system (or the universe)
S > 0
S = 0
S < 0
Spontaneous, increased randomness

Reversible, no change in disorder
Impossible, order cannot happen in isolation
There is an inexorable drive for the universe to go to a

maximally disordered state.
Microscopic understanding: Boltzmann Equation of Entropy:
S = k ln
Where k is Boltzmanns constant (k=R/NA).
And is the number of equally probable microscopic
arrangements for the system.
This can also be used to calculate S
In the case of the Joule expansion of an ideal gas in volume V
expanding to a volume 2V (as on the first page of these notes):
if we divide the initial volume V into m small cubes, each with
volume v, so that mv=V, the number of ways of placing the N
molecules of ideal gas into these small cubes initially is mN.
After the expansion the number of ways of placing the n
molecules of ideal gas into the now 2m small cubes is (2m)N.
The number of probably microscopic arrangements initially is:

N
N
( m ) , or = C ( m ) (C is a constant)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 5
After the expansion it is:

N
(2m ) , or = C (2m )
So using Boltzmanns equation to calculate S for the

expansion:
N
S = k ln (2m ) k ln m N = kN ln 2 = R ln 2
As we had found above!
More examples of S calculations

In all cases, we must find a reversible path to calculate
(a)
qrev
T
Mixing of ideal gases at constant T and p
nA A (g, VA, T) + nB A (g, VB, T) = n (A + B) (g, V = VA + VB, T)

Smix = nR [X A ln X A +XB ln XB ]
(b) Heating (or cooling) at constant V

A ( T1 , V) = A ( T2 , V )
S =
qrev
T
T2
CV dT
T
if CV is
T -independent
T
CV ln 2
T1
[Note S > 0 if T2 >T1 ]

(c) Reversible phase change at constant T and p
e.g. H2O (l, 100C, 1 bar) = H2O (g, 100C, 1 bar)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #7
page 6
q p = Hvap
Svap (100C ) =
(d)
q pvap
Tb
H vap
(Tb = boiling Temp at 1 bar)
Tb
Irreversible phase change at constant T and p
e.g. H2O (l, -10C, 1 bar) = H2O (s, -10C, 1 bar)

This is spontaneous and irreversible.
\
We need to find a reversible path between the two states

to calculate S.
irreversible
H2O (l, -10C, 1 bar)
H2O (s, -10C, 1 bar)
qrev = C p ( A )dT
H2O (l, 0C, 1 bar)
qrev = C p ( s ) dT
reversible
rev
p
H2O (s, 0C, 1 bar)
= Hfus
Note: Hfus is for the process going from the solid state to the
liquid state, the opposite of what we have above, same for Sfus.
S = Sheating Sfus + Scooling
=
Tfus
S =
Hfus
C p ( A ) dT Hfus
T C p ( s ) dT
+
+
Tfus
T
Tfus
T
S =
Hfus
Tfus
T1
Tfus
T1
+ C p ( A ) C p ( s ) ln
dT
C p ( A ) C p ( s )
T
if Cp values are T-independent
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 1
Fundamental Equations, Equilibrium, Free Energy,

Maxwell Relations
Fundamental Equations relate functions of state to each other using

1st and 2nd Laws
1st law with expansion work: dU = q - pextdV
need to express q in
terms of state variables
because q is path dependent
Use 2nd law: qrev = TdS

For a reversible process pext = p and q = qrev =TdS
So
** dU = TdS pdV **
This fundamental equation only contains state variables

Even though this equation was derived for a reversible process,
the equation is always correct and valid for a closed (no mass
transfer) system, even in the presence of an irreversible
process. This is because U, T, S, p, and V are all functions of
state and independent of path.
AND
The best or natural variables for U are S and V,

** U(S,V) **
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 2
** U(S,V) **
U
From dU = TdS pdV **
=T
S V
= p **
V S
We can write similar equations for enthalpy

H = U + pV
dH = dU + d(pV) = dU + pdV + Vdp

inserting dU = TdS pdV
** dH = TdS + Vdp **
The natural variables for H are then S and p
** H(S,p) **
H
From dH = TdS + Vdp **
=T
S p
= V **
Vp
_______________
We can use these equations to find how S depends on T.
From dU = TdS pdV
S = 1 U = CV
T V T T V T
From dH = TdS + Vdp
S = 1 H = Cp
T p T T p T
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 3
Criteria for Spontaneous Change

The 2nd Law gave the Clausius inequality for spontaneous change
dS > q/Tsurr.
The 1st law gave us
dU = q + w
Putting the two together, assuming only pV work, gives us the following
general criterion for spontaneous change:
** dU + pextdV TsurrdS < 0 **
Equilibrium is when there is no possible change of state that would
satisfy this inequality.
We can now use the general criterion above under specific conditions
Consider first an isolated system (q=w=0, V=0, U=0)
Since dU=0 and dV=0, from the general criterion above, then
(dS)U,V > 0
is the criterion for spontaneity for an isolated system
And equilibrium for an isolated system is then achieved when entropy
is maximized. At maximum entropy, no spontaneous changes can occur.
Consider now S and V constant
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 4
(dU)S,V < 0
is the criterion for spontaneity under constant V and S
At constant S and V, equilibrium is achieved when energy is minimized
Consider now S constant and p=pext constant

dU + pdV < 0 d(U + pV) < 0
=H
(dH)S,pext < 0
So
is the criterion for spontaneity under constant S and constant p=pext.
Consider now H constant and p=pext constant

dU + pdV TsurrdS < 0
but dU + pdV = dH, which is 0 (H is constant)
So (dS)H,p=pext > 0
is the criterion for spontaneity under constant H and constant p=pext.
Now lets begin considering cases that are experimentally more

controllable.
Consider now constant T=Tsurr and constant V
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 5
dU - TdS < 0 d(U - TS) < 0

Define A = U TS, the Helmholtz Free Energy
Then (dA)V,T=Tsurr < 0
is the criterion for spontaneity under constant T=Tsurr and constant V.
For constant V and constant T=Tsurr, equilibrium is achieved when the
Helmholtz free energy is minimized.
We now come to the most important and applicable constraint:
Consider now constant T=Tsurr and constant p=pext.
(dU + pdV TdS) < 0 d(U + pV TS) < 0
Define G = U + pV TS, the Gibbs Free Energy
(can also be written as
G = A + pV
and
G = H TS )
Then (dG)p=pext,T=Tsurr < 0

is the criterion for spontaneity under constant T=Tsurr and constant
p=pext.
At constant p=pext and constant T=Tsurr, equilibrium is achieved when
the Gibbs free energy is minimized.
Consider the process:

A(p,T) = B(p,T)
(keeping p and T constant)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
page 6
Under constant p=pext and T=Tsurr,

G < 0
G = 0
G > 0
A B is spontaneous
A and B are in equilibrium
then B A is spontaneous
Maxwell Relations
With the free energies
Helmholtz free energy
Gibbs free energy
A = U - TS
G = H - TS
weve introduced all our state functions. For closed systems,

U (S ,V
)
H (S , p )
A (T ,V )
G (T , p )
From
and
dU =TdS pdV
dH =TdS +Vdp
dA = SdT pdV
dG = SdT +Vdp
A
A
dT +
T V
V
G
G
dG =
dT +
T p
p
dA =
dV
dp
T
A = S
T V
A = p
V T
= S
T p
=V
p T
Fundamental equations
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #8
The Maxwell relations:
2A
2A
=
V T
T V
page 7
and
2G
2G
=
p T
T p
now allow us to find how S depends on V and p.
S = p
V T T V
S
V
T p
p T
These can be obtained from an equation of state.

We can now also relate U and H to p-V-T data.
p
p
T V
=T
V T
p =T
V T
=T
p T
S
V
+V = V T
T p
p T
For an ideal gas
U and H from equations of state!
pV = nRT
nR p
p
=
=
T
T V V
=0
V T
nR V
V
=
=
p T
T p
=0
p T
This proves that for an ideal gas U(T) and H(T), functions of T only.
We had assumed this was true from Joule and Joule-Thomson
expansion experiments. Now we know it is rigorously true.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 1
Gibbs Free Energy, Multicomponent Systems,

Partial Molar Quantities, and the Chemical
Potential
Comments on
the special role of G(T,p):

If you know G(T,p), you know all other thermodynamic quantitites.
G
,
T p
p T
S =
V =
T p
H = G +TS
H = G T
U = H pV
U = G T
A = U TS
A =G p
S
C p =T
T p
G
G
p
T p
p T
p T
2G
C p = T
2
T p
We can get all the thermodynamic functions from G(T,p).
p-dependence of G(T,p)
From
for liquids, solids, and gases (ideal)
p T
V =
p2
G (T , p2 ) = G (T , p1 ) + Vdp
p1
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
Liquids and solids
V is small
G (T , p2 ) = G (T , p1 ) +V ( p2 p1 ) G (T , p1 )
page 2
G (T
Ideal gases
G (T , p2 ) = G (T , p1 ) +
p2
p1
p
RT
dp = G (T , p1 ) + RT ln 2
p
p1
Take p1 = p o = 1 bar
G (T , p ) = G o (T ) + RT ln
From
p
p0
T p
S =
or G (T , p ) = G o (T ) + RT ln p
(p in bar)
S (T , p ) = S o (T ) R ln p
Multicomponent systems, the chemical equilibrium, partial molar

quantitites.
So far weve worked with fundamental equations for a closed (no
mass change) system with no composition change.
dU =TdS pdV
dA = SdT pdV
dH =TdS +Vdp
dG = SdT +Vdp
How does this change if we allow the composition of the system to

change? Like in a chemical reaction or a biochemical process?
Consider Gibbs free energy of a 2-component system
G (T , p ,n1 ,n2 )
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 3
G
G
G
G
dG =
dT +
dp +
dn1 +
dn2
p T ,n ,n
n1 T , p ,n
n2 T , p ,n
T p ,n ,n
We define

2
1
2
ni T , p ,nj i

1
2
as the chemical potential of species i
i (T , p , nj ) is an intensive variable
This gives a new set of fundamental equations for open systems

(mass can flow in and out, composition can change)
dG = SdT +Vdp + i dni
i
dH =TdS +Vdp + i dni

i
dU =TdS pdV + i dni

i
dA = SdT pdV + i dni

i
G
H
U
A
=
=
=
ni T , p ,nj i ni S , p ,nj i ni S ,V ,nj i ni T ,V ,nj i
i =
At equilibrium, the chemical potential of a species is the same

everywhere in the system
Lets show this in a system that has one component and two parts,
(for example a solid and a liquid phase, or for the case of a cell
placed in salt water, the water in the cell versus the water out of
the cell in the salt water)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 4
Consider moving an infinitesimal amount dn1 of component #1 from
phase a to phase b at constant T,p. Lets write the change in state.

dn1 (T , p ,phase a ) = dn1 (T , p ,phase b )
dG = S dT
+V dp + i dni = 1(b ) 1(a ) dn1

i
1(b ) < 1( a )
dG < 0
spontaneous conversion from (a) to (b)
1( a ) < 1(b )
dG > 0
spontaneous conversion from (b) to (a)
At equilibrium there cannot be any spontaneous processes, so

1( a ) = 1(b ) at equilibrium
e.g. liquid water and ice in equilibrium

ice
water
ice (T , p ) = water (T , p )
at coexistence equilibrium
For the cell in a salt water solution, water (cell ) (T , p ) > water (solution ) (T , p ) and
the cell dies as the water flows from the cell to the solution (this is
what we call osmotic pressure)
The chemical potential and its downhill drive to equilibrium will be

the guiding principle for our study of phase transitions, chemical
reactions, and biochemical processes
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 5
Partial molar quantities

i is the Gibbs free energy per mole of component i, i.e. the
partial molar Gibbs free energy

G
= i = Gi
ni T , p ,nj i
G = n1 1 + n2 2 + " + ni i = ni i = ni Gi
i
Lets prove this, using the fact that G is extensive.

G (T , p , n1 , n2 ) = G (T , p , n1 , n2 )
dG
(T , p , n1 , n2 ) = G (T , p ,n1 ,n2 )
d
G
G
( n1 )
( n2 )
+
=G
( n1 ) T , p ,n2 T , p ,n2 ( n2 ) T , p ,n1 T , p ,n1

n1
n2
is arbitrary, we can choose = 1
n1 1 + n2 2 = G
We can define other partial molar quantities similarly.

A
= Ai
n
i T , p ,nj i
A = n1A1 + n2A2 + " + ni Ai = ni Ai

i
partial molar Helmholtz free energy

note what is kept constant
H
= Hi
ni T , p ,n j i
not to be confused with

ni T ,V ,n
H = n1H1 + n2H2 + " + ni Hi = ni Hi

i
= i
j i
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 6
partial molar enthalpy

U
= Ui
ni T , p ,n j i
U = n1U1 + n2U2 + " + niUi = niUi

i
partial molar energy

__________________________________________________
Lets compare of a pure ideal gas to in a mixture of ideal gases.
Chemical potential in a pure (1-component) ideal gas
From
G (T , p ) = G o (T ) + RT ln
p
p0
(T , p ) = o (T ) + RT ln
Chemical potential in a mixture of ideal gases

onsider the equilibrium
pA + pB = ptot
mixed pure
A
A,B
p'A, p'B pA
Fixed Partition
At equilibrium
A ( mix ,T , ptot ) = A ( pure ,T , pA )
Also
pA ( pure ) = pA ( mix ) = ptot XA
Daltons Law
So
p
p0
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #9
page 7
A (mix ,T , ptot ) = A ( pure ,T , ptot XA )

p X
= Ao (T ) + RT ln tot A
p0
= Ao (T ) + RT ln
ptot
+ RT ln XA
p0
A ( pure ,T , ptot )
Note
A ( mix ,T , ptot ) = A ( pure ,T , ptot ) + RT ln XA

XA < 1
A (mix ,T , ptot ) < A ( pure ,T , ptot )
The chemical potential of A in the mixture is always less than the

chemical potential of A when pure, at the same total pressure. This
is at heart a reflection about entropy, the chemical potential of A
in the mixture is less than if it were pure, under the same (T,p)
conditions, because of the underlying (but hidden in this case)
entropy change!
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
page 1
Chemical Equilibrium
Ideal Gases
Question: What is the composition of a reacting mixture of ideal
gases?
e.g.
N2(g, T, p) + 3/2 H2(g, T, p) = NH3(g, T, p)
What are pN , pH , and pNH

2
at equilibrium?
Lets look at a more general case
A A(g, T, p) + B B(g, T, p) = C C(g, T, p) + D D(g, T, p)

The is are the stoichiometric coefficients.
Lets take a mixture of A, B, C, and D with partial pressures
pA = XA p , pB = XA p , pC = XC p , and pD = XD p
Is this mixture in equilibrium?

We can answer by finding G if we allow the reaction to proceed
further.
We know i (T , p ) for an ideal gas in a mixture
and we know that G = ni i
i
G ( ) = C C ( g,T , p ) + D D ( g,T , p ) A A ( g,T , p ) + B B ( g,T , p )
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
page 2
where is an arbitrary small number that allows to let the reaction

proceed just a bit.
We know that
pi
1 bar implied
i ( g,T , p ) = io (T ) + RT ln pi
where io (T ) is the standard chemical potential of species i at 1 bar

and in a pure (not mixed) state.
p C p D
G ( ) = C Co (T ) + D Do (T ) A Ao (T ) + B Bo (T ) + RT ln C A D B
pA pB
G = G + RT lnQ
where
G = C Co (T ) + D Do (T ) A Ao (T ) + B Bo (T )
and
pCC pCD
Q = A B is the reaction quotient
pA pB
G is the standard change in free energy for taking pure reactants
into pure products.

o
o
o
G o = Grxn
= Hrxn
T Srxn
or
o
o
G o = Gform
( reactants )
( products ) Gform
If
G ( ) < 0
then the reaction will proceed spontaneously to form

more products
G ( ) > 0
then the backward reaction is spontaneous
G ( ) = 0
No spontaneous changes
Equilibrium
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
page 3
o
= RT lnQeq
At Equilibrium G ( ) = 0 and this implies Grxn
Define
Qeq = K p
the equilibrium constant
D
C
pCC pCD
XC XC
p
=
A B = p KX
A B
p
p
X
X
A B eq
A B eq
Kp =
and thus
o
Grxn
= RT ln K p , K p = e G
RT
Note from this that K p (T ) is not a function of total pressure p.

It is KX = p K p which is KX ( p ,T ) .
Recall that all pi values are divided by 1 bar, so Kp and KX are
both unitless.
________________________________________________
T = 298 K
p =1 bar
Example: H2(g) + CO2(g) = H2O(g) + CO(g)
H2(g)
CO2(g)
H2O(g)
CO(g)
Initial #
of moles
# moles
at Eq.
a-x
b-x
Total # moles at Eq. = (a x) + (b x) + 2x = a + b

Mole fraction
at Eq.
a x
a +b
b x
a +b
a +b
a +b
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
o
Gform
(kJ/mol)
o
rxn
and
= 28.6 kJ/mol
Kp =
pH O pCO
2
pH pCO
2
page 4
-396.6
Kp = e
XH OXCO
2
XH XCO
2
-228.6
28,600 kJ/mol
(8.314 J/K-mol)(298 K )
-137.2
= e 11.54 = 9.7 x 10 6
x2
=
( a x ) (b x )
Lets take a = 1 mol and b = 2 mol

We need to solve
A)
x2
(1 x )(2 x )
= 9.7 x 10 6
Using approximation method:

K << 1, so we expect x << 1 also.
Assume
1 x 1, 2 x 2
x2
(1 x )(2 x )
x2
2
= 9.7 x 10 6
x 0.0044 mol (indeed << 1)
B)
Exactly:
x2
= K p = 9.7 x 10 6
2
x 3x + 2
x 2 (1 9.7x 10 6 ) + 3x ( 9.7x 10 6 ) 2 ( 9.7x 10 6 ) = 0

x =
3 ( 9.7x 10 6 )
2 ( 9.7x 10 6 )
9 ( 9.7x 10 6 ) + 4 (1 9.7x 10 6 ) 2 ( 9.7x 10 6 )

2
2 (1 9.7x 10 6 )
The - sign gives a nonphysical result (negative x value)

Take the + sign only
x = 0.0044 mol (same)
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
page 5
Effect of total pressure: example

N2O4(g) = 2 NO2(g)
Initial mol #
# at Eq.
n-x
2x
Xis at Eq.
n x
n +x
2x
n +x
Total # moles at Eq. =

n x + 2x = n + x
Kp =
2
pNO
pN O
2 4
Kp = p
2
p 2XNO
pXN O
2 4
4 2
1 2
where = x n is the fraction reacted
K
(1 ) 4 pp = 2
2
2x
4x 2
n + x
=p
=p 2
n x2
n x
n + x
Kp Kp
1 +
=
4p 4p
Kp
4p
1
=
2 =
Kp
4p
1 +
1 +
4p
Kp
4p
= 1 +
K p
1 2
If p increases, decreases
Le Chateliers Principle, for pressure:

An increase in pressure shifts the equilibrium so as to decrease the
total # of moles, reducing the volume.
In the example above, increasing p shifts the equilibrium toward the
reactants.
---------------
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #10
page 6
Another example:
2 NO(g) + O2(g) = 2 NO2(g)
Initial mol #
# at Eq.
2-2x
1-x
2x
Xis at Eq.
2 (1 x )
3x
1x
3x
2x
3x
Kp =
2
pNO
2
pNO
pO
2
p 2XNO
2
p 2XNO
pXO
2
XNO
2
p XNO
XO
K p = 2.3x 1012 at 298 K
Total # moles at Eq.

= 2 2x + 1 x + 2x
=3-x
2
1 x (3 x )
(1 x )
K p >> 1 so we expect x 1 3 - x 2
Kp
p (1 x )3
or
(1 x )
pK p
2
x =1
pK p
13
In this case, if p then x as expected from Le Chateliers principle.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #11
page 1
Equilibrium in Solution
The chemical potential for molecules in solution is given by a formula
that is very similar to that for ideal gases:
A (T , p , cA ) = Ao (T , p ) + RT ln cA = Ao (T , p ) + RT ln [A ]
The precise definition of the standard chemical potential Ao (T , p ) is

now more complicated; it is defined at a given pH, salt concentration,
etc, all solution properties that need to be defined in advance. We
will not go through those and take it as a given that the standard
state is appropriately defined.
Given a standard chemical potential Ao (T , p ) , then the analysis that
we did for the ideal gas follows straight through and we find for a
solution process
A A(g, T, p) + B B(g, T, p) = C C(g, T, p) + D D(g, T, p)

that following the ideal gas analysis in our previous lecture
[C ]C [D ]D
G ( ) = C C (T ) + D D (T ) A A (T ) + B B (T ) + RT ln
[A ] A [B ] B
and the equilibrium constant K comes out through

o
Grxn
= RT ln K ,
Where K = Qeq
K = e G
RT
[C ] [D ] at equilibrium as before, and where the

=
[A ] [B ]
C
concentrations Q are equilibrium concentrations.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #11
page 2
Temperature dependence of K (or Kp)

ln K (T ) =
G o
RT
d ln K
d G o G o
1 d G o
=
dT
dT RT RT 2 RT dT
But at fixed pressure and/or solutions properties (p = 1 bar, pH

constant, etc..)
d G o G o
=
dT
T 1 bar,pH constant, etc...
and from fundamental equation
G
dG = SdT +Vdp
= S
T p
o
o
1
d ln K H (T ) T S (T )
=
+
S o (T
2
dT
RT
RT
d ln K (T ) H o (T
=
dT
RT 2
G o
o
= S (T
T
T2
H o (T
)dT
Integrating:
ln K (T2 ) = ln K (T1 ) +
At constant p:
H o (T ) = H o (T1 ) + C p (T T1 )
T1
T2
ln K (T2 ) = ln K (T1 ) +
T1
RT
H o (T1 ) + C p (T T1 )
RT 2
dT
Over small T ranges, C p (T T1 ) can be assumed small and H o

independent of T.
H o 1 1
H o T2 T1
ln K (T2 ) ln K (T1 ) +
= ln K (T1 ) +
R T1 T2
R TT
1 2
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
If
Lecture #11
H o (T ) < 0
(Exothermic)
page 3
T2 >T1 means K p (T2 ) < K p (T1 )
The equilibrium shifts toward reactants

If
H o (T ) > 0
(Endothermic)
T2 >T1 means K p (T2 ) > K p (T1 )
The equilibrium shifts toward products

This is Le Chateliers principle for Temperature
Example: The Haber process

N2(g, T, p) + 3/2 H2(g, T, p) = NH3(g, T, p)
o
Hrxn
(298 K ) = 46.21 kJ/mol
o
Grxn
(298 K ) = 16.74 kJ/mol
Kp =
pNH
pH3 2 pN1 2
2
=p
XNH
XH3 2XN1 2
2
=e
16,740 J/mol
(8.314 J/K-mol)(298 K )
= 860
For p = 1 bar this is pretty good, lots of product. However, the

reaction at room T is slow (this is kinetics, not thermodynamics).
Raising T to 800 K can speed it up. But since H o (T ) < 0 (exothermic),
Le Chatelier tells us that the equilibrium will shift toward the
reactants.
Indeed:
What to do?
K p ( 800 K ) = 0.007
Note above
KX = p K p
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #11
page 4
Again use Le Chatelier, but with pressure! If we increase p, Eq.

shifts toward products.
Run reaction at high T and high p
For p = 1 bar, T = 800 K, Kp = 0.007

KX =
XNH
XH3 2XN1 2
2
But at p = 100 bar,
= (1 ) K p = 0.007
much better!
K X = (100 ) K p = 0.7
Heterogeneous Equilibria
If a product or reactant is a solid or liquid, it will not appear in the

ratio of partial ps for Kp or in the concentrations if the equilibrium is
in solution. However, it must be used in G.
Why?
Take
A A(s) + B B(g) = C C(l) + D D(g)
The solid and liquid are not mixed they are pure states.
G = C C ( s, pure, p ) + D D ( g, mix, p ) A A ( l , pure, p ) + B B ( g, mix, p )
And for (l) or (s)
C ( pure, p ) o ( pure )
G = C Co + D Do A Ao B Bo + RT ln
(no p-dependence)
pDD
= G o + RT lnQ
B
pB
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #11
pDD
B
pB Eq .
No A or C involved.
Kp =
But we still have

and
page 5
o
Grxn
= C Co + D Do A Ao B Bo
ln K p =
o
Grxn
RT
e.g. the decomposition of limestone

CaCO3 (s) = CaO (s) + CO2 (g)
T = 25C
Calculate equilibrium vapor pressure at room T and elevated T.

Data at 25C:
Substance
CaCO3 (s)
CaO (s)
CO2 (g)
o (kJ/mol)
-1128.8
-604.0
-394.36
Hfo (kJ/mol)
-1206.9
-635.09
-393.51
At equilibrium,
G = ( CaO, s ) + ( CO2 , g) ( CaCO3 , s )

= o ( CaO, s ) + o ( CO2 , g) + RT ln pCO2 o ( CaCO3 , s )
= G o + RT ln K p
where K p = pCO2 (at eq.)
The equilibrium constant includes only the gas, but G o includes

the solids too.
G o (kJ/mol) = -604.0 394.4 (-1128.8) = 130.4 kJ/mol
H o (kJ/mol) = -635.1 393.5 (-1206.9) = 178.3 kJ/mol
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #11
page 6
Equilibrium pressure:
ln K p =
G o
RT
130, 400 J/mol

= 52.50
(8.314 J/K-mol) (298.15 K )
K p = 1.43x 10 23 bar
Nothing there at room T !
Try 1100 K:
H o 1
1
R 1100 K 298 K
178, 300 J/mol 1
1
= 52.50
= 0.17
8.314 J/K-mol 1100 K 298 K
(1100 K ) 0.84 bar
ln pCO2 (1100 K ) ln pCO2 (298 K ) +
pCO
Theres probably some change in Hfo over such a wide T range,

but clearly the equilibrium shifts dramatically.
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #12
page 1
Equilibrium: Application to Drug Design
Based on Rational cytokine design for increased lifetime and

enhanced potency using pH-activated histidine switching, Sarkar,
Lowenhaupt, Horan, Boone, Tidor, and Lauffenburger, Nature
Biotechnology 20, 908 (2002).
The analysis for equilibrium that we have used for reactions involving
breaking and making covalent bonds applies equally well to reactions
such as those involved in ligand-receptor binding, where the ligand
and receptor are proteins
R+L=C
where R is the receptor, L is the ligand, and C is the receptor-ligand
complex. The interactions between these proteins typically involve
multiple non-covalent interactions, including hydrogen bonds,
hydrophobic interactions, and electrostatic interactions. The
equilibrium constant and Gibbs free energy change for the reaction
are related in the usual way
G o = RT ln Ka
where the equilibrium constant Ka is called the association constant
Ka =
[C ]
[R ][L ]
The standard state needed to characterize G o is defined at a set

of specific reference conditions (pH, salt concentration, etc.). By
convention, the reverse process (the dissociation) is used to
characterize the strength of ligand binding through the equilibrium
constant KD, also called the dissociation constant
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #12
KD =
page 2
[R ][L ]
[C ]
The lower the KD, the better the ligand (the tighter the binding).
In an experiment the ligand is typically labeled radioactively (e.g.
with 125I) and added to cells under conditions that prevent the ligand
from being internalized (4C).The ligand is usually in great excess
compared to the number of receptors, so that at equilibrium [L]=[L]0
is a good approximation (the ligand concentration is effectively
unchanged during the process).
If [R]T is the total concentration of receptors, then [R]T=[R]+[C], so
that at equilibrium,
KD =
([R ]
[C ]eq [L ]o
[C ]eq
or
[C ]eq [R ]T [C ]eq
=
KD
KD
[L ]o
The value of KD (and thus G0) can be obtained by measuring the
concentration of complexes formed at various initial ligand
concentrations [L]0 (through the radioactive labeling) by plotting
[C ]eq
[L ]o
as a function of [C ]eq (a Scatchard plot). The slope gives
The fraction of receptors occupied is
KD
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #12
[C ]eq
[R ]T
Note, when [L]o<<KD then

when [L]o>>KD then
page 3
K
1 + D
[L ]o
[C ]eq [L ]o
,
[R ]T KD
[C ]eq
K
1 D
[R ]T
[L ]o
Although free energies for receptor-ligand binding reactions are

generally determined experimentally (through KD), it is possible to
computationally estimate the changes in free energy that accompany
point mutations in one of the amino acids in the ligand. This approach
can be used to design a new and better drug that binds with an
affinity that improves its properties. An example of such a designed
mutated ligand is an improved version of Granulocyte-Colony
Stimulating Factor (GCSF). GCSF is a protein drug that is used to
treat chemotherapy patients and stimulates the growth of white
blood cells.
It is desirable to have GCSF bind tightly to its receptor at the cell
surface (at pH 7.4) as this signals the cell to produce the desired
proteins. But when the complex C is internalized in the cell in
endosomal compartments (pH 5.5), it is desirable for GCSF to fall
off its receptor to be recycled back to the solution to be used again
instead of being degraded within the endosome. Thus a design
principle for an improved mutant GCSF is weaker binding at pH 5.5
(inside cell) than at pH 7.4 (cell surface), or in other words
KD(pH5.5)> KD(pH7.4).
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #12
page 4
For the wild type (WT) GCSF the data gives:
WT
KD(pH 7.4), pM
KD(pH 5.5)/ KD(pH 7.4)
27090
1.70.5
Since G o = RT ln KD , we can get the difference of G o s for the

dissociation reaction at the different pHs
G o ( pH 7.4) G o ( pH 5.5) / RT
WT:
0.530.3 (measured from KD values)
Calculations were performed on several mutants and two showed

appreciable differences in free energies
D110H:
D113H:
8.3 (calculated)
17 (calculated)
These mutant GCSF molecules were synthesized and evaluated for

binding to the GCSF receptor with the following results:
KD(pH 7.4), pM
KD(pH 5.5)/ KD(pH 7.4)
WT
27090
1.70.5
D110H
D113H
370450
320130
4.40.8
6.82.4
20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005
Lecture #12
page 5
The mutants do bind more weakly than does the wild type at low pH,
and thus have the potential to be better drugs (in fact, in animal
trials the mutants have much longer half-lives than the wild type).
Differences in free energies for the mutants can be obtained from
the experimental Ks:
G o ( pH 7.4) G o ( pH 5.5) / RT
WT:
D110H:
D113H:

20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
What is Statistical Mechanics?

Brief Review of probability and combinatorics
S=k ln and S = -k pi ln pi
What is Statistical Mechanics?

Spontaneous processes:
under certain conditions, know what system would like to maximize/minimize:
U (S ,V , ni )
(dU )S ,V ,n
A(T ,V , ni )
(dA)T ,V ,n
H (S , P, ni )
(dH )S , P ,n
G (T , P, ni )
(dG )T , P ,n
0
0
These tell you which way a reaction will go, and when it will stop or reach equilibrium.
Want these thermodynamic properties, now in terms of molecular level picture.
p
Before: ignore molecular properties
Now want to account for it: molecule exists in states, has energy levels (recall 5.111)
If have large number of molecules (NA = 6 1023), it is nearly impossible to track all of the states. Mind
boggling!
How?: use probability to describe molecule instead of tracking each one
pi = probability that molecule will be in ith state. The link between microscopic and macroscopic:
1
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
Boltzmann Equation
= multiplicity
S = k ln
k=Boltzmann constant, = 1.38 10-23 J/K

Another form:
pi = probability that molecule will be in ith state
t = number of states
S = k pi ln pi
i =1
Probabilistic description of entropy!

Then relate these microscopic properties to macroscopic ones: U, S, etc. (Driving forces). Explain:
unfolded protein folded protein
Do not need to know exactly how many molecules are in a state, just the probability that it is.
Ludwig Boltzmann (1844-1906)
Image removed due to copyright reasons.
Approach:
make model to represent problem
describe the states a system can be in
use statistics and probability to average
relate to a thermodynamic property (U, S, etc)
DEMYSTIFYING S
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
Brief Review of Probability and Combinatorics

Combinatorics: counting the number of available states
(conformations, energy levels)
Multiplicity (): Number of possible arrangements
For a set of N distinguishable objects:
= N ( N 1) ( N 2 ) 3 2 1 = N !
Example: 4 mer, A, T, G, C
How many different ways can these 4 nucleotides be arranged?
Write out all possible arrangements:
ACGT
CAGT
GACT
TACG
ACTG
CATG
GATC
TAGC
AGCT
CGAT
GCAT
TCAG
AGTC
CGTA
GCTA
TCGA
ATCG
CTAG
GTAC
TGAC
ATGC
CTGA
GTCA
TGCA
=24 ways
Using equation:
4 available slots: __ __ __ __
= 4 3 2 1 = 4! = 24 possible arrangements
3
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
(remember: 0! =1)
Indistinguishability: What if two of the letters are the same?:
Example: 3mer with A, A, G
Two of the nucleotides are indistinguishable (cant tell them apart)
First: pretend A and A are distinguishable: A1 A2
G A1 A2
G A2 A1
A1 G A2
A2 G A1
A1 A2 G
A2 A1 G
= 6 ways
Then: take away labels. Both columns same. = 3 ways
To account for indistinguishability, divide by NA = number of As
N! 3 2 1
=3
=
N A!
2 1
NA = number of As = 2
General equation:
N objects
t categories
N!
n1!n2!...nt !
ni = number of objects in ith category (degeneracy)

Binomial case: only two possibilities (Heads/Tails, On/off, spin up/spin down, ground state/excited
state)
then, t=2
nH + nT = N
need to specify only nH n
then nT = N-n
N
N!
(n, N ) = =
n n!( N n )!
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
Lets utilize in an example:

Example 1: Entropy of mixing
This example will show that maximum indicates a maximum S. Well choose a model for to
demonstrate the entropy of mixing. This will illustrate the driving force behind mixing based on
entropy.
Lattice model: 8 slots, separated by a wall. Total volume is fixed. 4 black particles and 4 white ones,
which can be placed one per slot in any configuration:
Degree of freedom: is the # of black particles and # white particles on either side of the wall.
N!
8!
=
= 70
nB !nw! 4!4!
Now: count number of ways each configuration is possible for each configuration (number of B on
Left) Use
tot=leftright
A
4B0B
4! 4!
=1
4!0! 0!4!
3B1B
4! 4!
= 16
3!1! 1!3!
2B2B
4! 4!
= 36
2!2! 2!2!
1B3B
4! 4!
= 16
1!3! 3!1!
0B4B
4! 4!
=1
0!4! 4!0!
tot=70 ways
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
Case C has the highest entropy because it has the highest multiplicity. Therefore, this is the state
the system is most likely to be in. This also is confirmed by our intuition of the problem, in that C is the
most disordered case.
S = k ln
This is akin to:
vs
=36
=1
maximizing maximizes S
tendency towards disorder
Probability: likelihood that the system is occupying that state
N = number of total outcomes
nA = number of outcomes of category A
pA =
pA = probability of obtaining outcome of type A
Example: Die
N=6
nodd = 3
6
nA
N
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
nodd 3 1
= =
N
6 2
podd =
Addition Rule:
Probability of outcome A OR outcome B:
p A + pB
Multiplication Rule:
Probability of outcome A AND outcome B
p A pB
If outcomes are collectively exhaustive:
t
=1
i =1
Probability Distributions
Plotting pi vs. i gives you a probability distribution. For example, the probabilities of rolling a die may
look like either case A or case B:
1.00
A
0.75
pi
i =1
0.50
0.25 1/6 1/6 1/6 1/6 1/6 1/6
0.00
1
1.00
pi
= k [ln 6] = 1.79k
6/6
S = k pi ln pi
i =1
0.50
= k [0 ln(0) + 1ln(1) + 0 ln(0) + 0 ln(0) + 0 ln(0) + 0 ln(0)]
0.25
0.00
= k [16 ln ( 16 ) + 16 ln ( 16 ) + 16 ln ( 16 ) + 16 ln( 16 ) + 16 ln( 16 ) + 16 ln( 16 )]
0.75
S = k pi ln pi
0/6
1
0/6 0/6 0/6 0/6

2
=0
20.110J / 2.772J / 5.601J

Lecture 13
5.60/20.110/2.772
Lets use this to calculate S for two different probability distributions. Of the two above, can guess
that the case a) has the higher entropy. More disordered. But lets calculate a number using:
S = k pi ln pi
i =1
Can see that flatter distributions have higher entropy. What does this mean? If you are distributing
particles in energy levels, evenly distributing them leads to the highest entropy.
But what we usually observe is that things tend to have lower energies. Why is that? We have placed
no constraints on the system. But dont worry, we will!
Given a probability distribution, can compute the average values of some property X
< X >=
score p
i
all
states
pi
all
states
For example, if we wanted to

probability distribution:
< score >=
computer the average score for this
1 1 1 1 1 1
= 1 + 2 + 3 + 4 + 5 + 6 = 3.5
6 6 6 6 6 6
We will be using average energy a lot:
< >=
pi
all
states
This is our constraint: that the total energy has to sum up to a particular value.
Stirlings approximation:
S = k ln
then need to do ln N! for large N (N >10)
n
n!
e
ln n! n ln n n
=N! or N!/(n1!n2!.)
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
Why works for large N

Derivation of the Boltzmann Distribution Law
Partition Function
Why works for large N

We have seen that a system will vary its degrees of freedom in order to maximize and thus S. A
system has a higher probability of being in a state due to it being more probable. This allows us to
simply count states and see which one is more likely.
The lattice model of mixing gases had only N=8 particles. Is this approach still justified when we look
at a larger number of particles, like NA? It turns out the most probable state at low N becomes even
more likely at very high N.
Consider: coin flips
nH
4
3
2
1
0
S =kln
4!
N!
=
=1
n!( N n )! 4!0!
4!
==
=4
3!1!
4!
==
=6
2!2!
4!
==
=4
1!3!
4!
==
=1
0!4!
0
1.386k
1.792k
1.386k
0
Then do for N = 10, 100, 1000

max
N=4
max
max
10 n
100
N=1000
becomes increasingly narrower as N. Compare numerically:

1
50
max
N=10
N=100
500
1000
20.110J / 2.772J / 5.601J

Lecture 14
(5,10 ) =
(2,10 ) =
5.60/20.110/2.772
10!
= 252
5!5!
10!
= 45
2!8!
(50,100 ) =
5.6X more likely

(20,100 ) =
100!
= 1 10 29
50!50!
100!
= 5 10 20
20!80!
109X more likely!
Even though the process is totally random: If the number of trials N is large enough, the composition
of the outcomes becomes predictable with great precision.
This allows us to better predict the most probable state!
maximizing = maximizing S
Derivation of the Boltzmann Distribution Law

Microscopic definition of entropy:
t
S = k p j ln p j
i =1
What probability distribution (set of pjs) maximizes S?
S
=0
p j
for all j
constraint that probabilities sum to 1:
=1
j =1
t
dp
=0
j =1
Utilize Lagrange multipliers to solve this problem. We add the constraint to the equation we are
trying to maximize with a multiplier, . Then when we maximize the resulting equation the value of
is determined. i.e., solving the set of equations:
j =1
p j
t
dp j = 0
for all j
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
Plug in definition of S (pj):

t
t

k p j ln p j p j = 0
p j
j
j =1
Take the derivative:
k (ln p j + 1) = 0
ln p j =
pj = e
1
k
Divide pi by 1 to get rid of :
pj
t
pi
i =1
te
1
t
This says: Flattest probability distributions have highest S. This is something we already knew.
Now, what happens when we impose a constraint on the system? i.e., you have a given temperature,
and can sum up to a particular total energy. This is a more realistic problem to solve.
Lets put this into practice with an example.
Simple model to illustrate: 4 bead polymer
-0
compact
open
The polymer can assume multiple configurations. Well label one end atom so that it is distinguishable
from the other atoms in the chain. The polymer is stabilized when in compact configuration by energy
from open state. This is represented by the dashed line. This is a simple model utilized by those
studying protein folding as it can represent the configurations of a protein in the folded and unfolded
states. It represents a polypeptide chain that has only 4 amino acids, and a great simplification of real
proteins in that the chain can assume only a small number of conformations: one compact and four
open.
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
native
denatured
One end bead is labeled so that it is distinguishable from other end.

a microstate
1st excited
state
E=
a macrostate
ground state
E=0
microstate: a possible configuration. snapshot. A measurement averages over several

microstates
macrostate: a collection of microstates with the same energy
Define the E=0 state as the compact form, where the chain is stabilized by some energy relative to
the open state due to the interaction between bead 1 and bead 4.
Degree of freedom: physical conformation of the chain and the energy of each conformation, or
microstate.
What is the probability distribution that minimizes or maximizes a relevant thermodynamic quantity?
What happens if we try do this in real laboratory conditions? (T,V,N) or (T,P,N) controlled.
Lets say we have (T,V,N) constant, making A what we want to minimize.
dA = dU = TdS
at equilibrium:
dA = 0
Goal: Get dU and dS and solve for pj that makes dA = 0.
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
Differentiate with respect to pi

t
dS = k (1 + ln p j )dp j
j =1
Recall definition of an average value:

t
E = pjEj
j =1
dU = d E = ( p j dE j + E j dp j )
j =1
Energy levels do not depend on T. pj, or how they are populated, do.
t
dU = d E = (E j dp j )
j =1
dA = d E TdS = 0
use the constraint:
t
=1
j =1
Which allows us to use the Lagrange Multiplier

t
dp j = 0
j =1
Plug everything back into dA equation:
dA = d E TdS = 0
t
t
= E j dp j T k (1 + ln p j )dp j + dp j = 0
j =1
j =1
j =1
adding 0
group dpj terms:
t
dA = E j + kT (1 + ln p j ) + dp j = 0
j =1
must =0
5
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
ln p*j =
Ej
kT
1
kT
pj*= set of pj that satisfies dA=0
p*j = exp j exp

1
kT
kT
Well eliminate from the equation by using
t
=1
j =1
First, sum both sides:
Ej

exp

=
=
1
1
exp
p
kT
j =1
j =1
kT
t
*
j
E
t

1 exp j
1 = exp
kT j =1
kT
Rearranging the last expression
exp
1
t
Ej
kT
exp
kT
j =1
Plug this back into pj*:
Ej
exp
kT
p *j = t
Ej
exp
j kT
Ej
exp
kT
=
Q
Boltzmann Distribution Law
pj is the probability that the systems is in the Ejth energy level
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
We have defined the denominator as Q, the partition function
1
t
Ej

exp
j =1
kT
We arrived here by finding the probability distribution, or set of pjs, that minimizes the free energy.
What does it say?
When you are trying to maximize entropy, minimize energy: more particles like to have lower
energies. Particles populate relatively low Ej apiece
Probability distributions have an exponential form when you place constraints on them (not flat,
like for the case of no constraints)
Relative populations of two levels:
pi*
(Ei E j )
=
exp
kT
p *j
If j higher than i, then Ei-Ej<0 (negative)

so exp (+)
pi/pj >1, more in ith level.
Note: Particles do not have a preference for the lower energies, there is just a greater number of
ways to arrange the particles so that they distribute the E.
For a given Etot, can arrange particles in several ways to achieve Etot. However, the Boltzmann
Distribution Law says that the left hand situation is much more probable as it has the higher entropy.
many possible
arrangements
few possible
arrangements
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
What is the partition function?

In our derivation of the Boltzmann equation, the partition function, Q, came out.
Ej
Q exp
kT
j =1
Q describes how the particles are partitioned throughout accessible states. It is a number. Note that
Q is temperature dependent!
In simpler terms: Q tells you the number of states that are effectively accessible to the system at a
given temperature.
Qualitatively:
If you have t energy levels:

Et
....
E3
kT
(many states
accessible)
E2
E1
(few states
kT accessible)
E0
Ej
E
E
E
E
= exp 0 + exp 1 + exp 2 + + exp t
Q exp
kT
kT
kT
kT
j =1
kT
t
Ej/kT factor: magnitude of Ej relative to kT is the relevant number.

8
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
Units of kT: [J] (energy)

Lets look at two limits:
a) T (hi temperature) OR Ej0 (small energy spacing)
then Ej/kT0
E
exp j
kT
1
1
p*j = t
=
= = p *j
E (1 + 1 + 1) t
j exp kTj
this means: all states are accessible. Note that
Qt
b)T 0 (low temp) OR Ej (big energy spacing)
then Ej/kT
p *j =0 =
1
= 1 = p *j =0
(1 + 0 + 0)
p *j =rest =
0
= 0 = p *j =rest
(1 + 0 + 0)
this means: only ground state accessible.
Q 1
Now lets do it again for our 4 bead polymer:
E=
E=0
We still need to account for one more thing:

Degeneracy, g of upper the macrostate--there are four microstates.
20.110J / 2.772J / 5.601J

Lecture 14
5.60/20.110/2.772
l max
El
g (E )exp kT
l
l are the levels. gl=0 = 1, gl=1= 4

Q = 1 exp( 0 ) + 4 exp
= 1 + 4 exp
kT
kT
Q(T):
At low T, Q=1 (lowest state accessible)
At high T, Q=5 (all states accessible)
and also pl (T).
p compact
p open
0
kT
1 e
=
Q
4e
=
Q
kT
pcompact =1
popen =4/5
1
=
Q
pl
pcompact =1/5
popen =0
T
This is a unfolding or a denaturation profile for a polymer or protein, etc. Experiments: Fix T,
measure popen vs pcompact.
Why are we so interested in Q? We will re-derive thermodynamic properties in terms of Q. This is
the link between the microscopic and macroscopic descriptions.
Interesting side note: Calculate S of unfolding using S=k ln
Sclosed = k ln 1 = 0
Sopen = k ln 4
S = +
This says that the protein will want to unfold, based only on entropy. However, this model does not
account for things like interaction with the water molecules around the protein, which order around the
chain.
10
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
Q vs. q for distinguishable vs indistinguishable systems

Derivation of Thermodynamic Properties from Q:
U, S, A, , P
Examples
Partition Functions for independent and distinguishable particles

We want to generalize for distinguishable and indistinguishable particles. Lets make it easier on
ourselves by considering only independent subsystems, i.e., the particles do not interact. Then the
energy of the whole system, written as,
E j = i + ( interaction )
i
can be simplified because the 2nd term (interaction) is 0.

This allows us to say:
Q
whole system
= qi
i
1. Distinguishable particles (and independent)

A, B independent of each other, and labeled:
iA
i=1,2,a
mB
m=1,2,b
E j = iA + mB
q A = exp
iA
i =1
Q = exp
j =1
Ej
kT
= exp
qB = exp
and
kT
iA + mB
mB
m=1
i =1 m=1
kT
= exp
iA
i =1 m=1
Because the sums are independent of each other

a
Q = exp
i =1
iA
exp
kT m=1
mB
Generalize for N independent and distinguishable particles:
Q = qN
1
kT
kT
= q Aq B
kT
exp
mB
kT
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
2. Indistinguishable particles (and independent)

Now: no A and B labels!
E j = iA + mB
where i=1,2,t1, m = 1,2,t2.

t
Q = exp
j =1
t1
t2
= exp
kT
iA + mB
i =1 m=1
kT
Now: cannot factor out of sum due to indistinguishability: cant separate sums
WHY?
particle 1
1=10
particle 2
1=167
Cant be distinguished from the situation where

particle 1
1=167
particle 2
1=10
So overcounting is present. Divide by 2!
q2
Q=
2!
In general, for N particles, divide by N!
qN
Q=
N!
Deriving Thermodynamic Properties using Q

All thermodynamic quantities can be calculated from the partition function
The Boltzmann factor and partition function are the two most important quantities for making
statistical mechanical calculations. If we have a model for a material for which we can calculate
the partition function, we know everything there is to know about the thermodynamics of that
model.
Now we will relate our favorite thermodynamic properties to q, the partition function. This is our link
between the microscopic and macroscopic descriptions. Using the convenient dummy variable =
1/kbT to simplify things.
2
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
Deriving Energy, U
t
E = pjEj
j =1
1 t
E
E je j
Q j =1
Use trick
E j
= E j e
so then
1 t
E
E = E je j
Q j =1
=
Substituting this into <E>
ln Q
1 Q
ln Q T
U =< E >=
1
1
=
= 2
kT
T T kT
ln Q
U = kT 2
T
Deriving S:
t
S
= p j ln p j
k
j
Use
pj =
Ej
kT
E j
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
1
S
= e
k
j Q
t
E j
1 E j
ln
Q
kT

kT
Split up
t
1 E j kT 1 t 1 E j kT E j
S

ln + e
= e
k
Q
Q
Q
j
j
kT
2nd term:
1st term:
1 E j kT E j 1 1
j Q e kT = Q kT

t
1
e
j Q
Ej
kT
1
ln
Q
E j
kT
Ej
1 t
U = E j e kT
Q j
So 2nd term is
1
Q ln Q
Q
E j
1 E j kT E j 1 1 t
U
j Q e kT = Q kT j E j e kT = kT

t
= ln Q
Combing both terms:
S = k ln Q +
U
T
We can do this for several other properties!

Helmholtz Free Energy, A
A = U TS
U
= U T k ln Q +
T
= kT ln Q
4
E je
Use
1 t E j
= + e kT ln Q
Q j
=
20.110J / 2.772J / 5.601J

Lecture 15
Chemical potential,
5.60/20.110/2.772
A
=
N T ,V
ln Q
= kT
N T ,V
Pressure, P
A
P =
V T , N
ln Q
P = kT
V T , N
Now have all the thermodynamic properties as a function of Q, the partition function. We can use
these in a couple examples.
An Application Example: Visualizing the complex states of a DNA

molecule
Lets consider the unwinding of a superhelix of DNA as an example of using the Gibbs free energy to
describe the population of states.
Closed superhelical DNA can be unwound by treatment with DNAse, which nicks the DNA. The
break in one chain allows the double helix to twist relative to its axis and relax the supercoiling in
response to thermal fluctuations. The DNA can be healed by treatment with ligase to seal the nick.
When nicked, the DNA will achieve an equilibrium where some of the DNA is completely unwound,
some has one right-handed twist, some has one left-handed twist, some has two right-handed twists,
and so on. When the ligase is added to freeze the fluctuating DNA by fixing the nick, the collection
of DNA molecules is captured in an equilibrium distribution of different configurations. We can use
the connection between the probability of configurations and the free energy to predict this
distribution. (Eisenberg and Crothers)

Please see:
Figure 14-10 in Eisenberg, David S., and Donald M. Crothers. Physical chemistry: with applications to the life sciences.
Menlo Park, CA: Benjamin/Cummings, 1979. ISBN: 080532402X.
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
The frozen collection of DNA molecules with different degrees of superhelicity can be separated by
gel electrophoresis to allow analysis of the relative concentrations of each species:

Please see:
Figure 14-11(a) in Eisenberg, David S., and Donald M. Crothers. Physical chemistry: with applications to the life sciences.
The gel electrophoresis chart shows a clear separation of unique DNA species, occurring at different
concentrations as a function of their superhelicity. (The y-axis represents concentration, while the xaxis represents distance along the gel.) The peaks have been denoted with values of e, measuring
the number of superhelical twists in the DNA present in each peak: = relaxed circular DNA, +1 =
one left-handed superhelical twist, -1 = one right-handed superhelical twist, etc.
How can we predict the relative concentrations observed above? As with all statistical mechanics
calculations, we start with a model: Here, we want a model for how the free energy varies with twist in
the DNA superhelix.
We will start from a very simple model for the twisting energy of the DNA coil (and show that it
correctly predicts the observed distribution of twists). We are all familiar with the simple linear
function known as Hookes law which describes the relationship between the restoring force on a
spring and the displacement of the spring: F = -kx, where k is the spring constant. Twisting DNA is
not a simple spring, but can be thought of as a torsional spring- a coil with a restoring force when a
torque is applied. To remind you, a torque ( ) results when a force acts in a radial manner through
an axis r:
=F r
Both the force F and radius r are vectors. Analogous to the simple linear spring, a torsional spring
feels a torque which is linear to the applied twist:
torsional _ spring = kT
Here kT is a torsional spring constant and is the angle of the twist. If we assume the spring can only
undergo integral numbers of twists, then we could rewrite this as:
torsional_ spring = kT twist

Where twist is simply the angle for one twist of the spring, and is the total number of twists ( =
V
twist). Just as the force on a linear spring is related to a change in potential energy F = kx = ,
x
we can relate the torque on our DNA torsional spring to a change in its free energy:
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
torsional_ spring = kT twist =
G
= kT twist = B
B
G = 2
2
In the equation above, we combine the constants into one stiffness parameter B (B = kTtwist) to
simplify the expression. We are using free energy rather than mechanical potential energy here
because this molecular system (the twisting DNA coil) has internal degrees of freedom (e.g., bonds
among the DNA strands) that could also be affected by supercoiling.
If we ask what is the free energy of one particular DNA molecule i that has some number of twists i,
we have:
B i2
Gi =
2
is the number of superhelical turns; negative for right-hand turns, positive for left-hand turns.
Using our link between the free energy and the probability of observing a state with a particular
energy we have for the twisting DNA:
Pi =
B i2
2RT
all energies
B n2
2RT
n=1
To relate this to our measured quantity (concentration of species I, proportional to the peak in our gel
electrophoresis experiment), we simply recognize:
c i = c o Pi
Where co is the total concentration of DNA. The presence of the squared term in the exponent
means this distribution has a Gaussian shape (the same result we discussed last lecture- except for
this simple model, the entire probability distribution is Gaussian, not just near the peak of the
distribution). Fitting the measured concentration data with a Gaussian curve, we find the theory
predicts the observed distribution of superhelices very well:
20.110J / 2.772J / 5.601J

Lecture 15
5.60/20.110/2.772
Images removed due to copyright reasons.

Please see:
Figure 14-11 in Eisenberg, David S., and Donald M. Crothers. Physical chemistry: with applications to the life sciences.
20.110J / 2.772J / 5.601J

Lecture 16
5.60/20.110/2.772
Absolute Entropy
Third Law of thermodynamics
Absolute Entropies
Absolute entropy of an ideal gas
Start with fundamental equation
dU = TdS pdV
dS =
dU + pdV
T
for ideal gas:
dU = CV dT and p =
dS =
nRT
V
CV dT nR
+
dV
T
V
At constant T, dT=0
dST =
pdV
T
dST =
nRdV
V
For an ideal gas, pV = nRT
At constant T
d ( pV ) = d (nRT ) = 0
pdV = Vdp
plugging into dST:
dST =
nRdp
p
This allows us to know how S(p) if T held constant. Integrate!
20.110J / 2.772J / 5.601J

Lecture 16
5.60/20.110/2.772
For an arbitrary pressure p,
S(p,T ) = S( po ,T ) p o
p
nRdp
= S(p o ,T ) nRln o
p
p
where po is some reference pressure which we set at 1 bar.

S(p,T) = So(T) nR lnp
(p in bar)
0
S ( p, T ) = S (T ) R ln p
S as P
But to finish, we still need S o (T ) !

o
Suppose we had S ( 0 K ) (standard molar entropy at 0 Kelvin)
dH = TdS + Vdp
dH = C p dT
for ideal gas
C p dT = TdS + Vdp
dS =
Cp
T
dT
V
dp
T
20.110J / 2.772J / 5.601J

Lecture 16
5.60/20.110/2.772
Cp
S
=
Then using
T p T
o
S (T ) . Integrating over dS eqn, assuming Cp constant over T range:
we should be able to get
T2
dS =
Cp
T1
p2
dT
nR
p p dp
1
So then
T
p
S = C p ln 2 nR ln 2
T1
p1
for p = 1bar
T
= C p ln 2 nR ln p
T1
Given Cp, T1, p1, T2, p2, can calculate S.
We will use T=0K as a reference point.
Consider the following sequence of processes for the substance A:

A(s,0K,1bar) = A(s,Tm,1bar) = A(l,Tm,1bar) = A(l,Tb,1bar) = A(g,Tb,1bar) = A(g,T,1bar)
S (T,1bar) = S o (0K ) +
Tm
So(T)
T b C p ()dT
T C p (g)dT
C p (s)dT H fus
H vap
+
+
+
+
Tm
Tb
T
T
T
Tm
Tb
S =
CpdT
T
Liquid boils, S =
H fus
Solid melts, S =
T
0
H vap
T
20.110J / 2.772J / 5.601J

Lecture 16
5.60/20.110/2.772
Since S0 is positive for each of these processes, the entropy must have its smallest possible value
o
at 0 K. If we take S ( 0 K ) = zero for every pure substance in its crystalline solid state, then we could
calculate the entropy at any other temperature.
This leads us to the Third Law of Thermodynamics:
THIRD LAW:
First expressed as Nernst's Heat Theorem:
Nernst (1905):
As T 0 K , S 0 for all isothermal processes in condensed phases
More general and useful formulation by M. Planck:
Planck (1911):
As T 0 K , S 0 for every chemically homogeneous substance in a perfect crystalline state
Justification:
?
@
It works!
Statistical mechanics (5.62) allows us to calculate the
o
entropy and indeed predicts S ( 0 K ) = 0.

This leads to the following interesting corollary:
It is impossible to decrease the temperature of any system to T = 0 K in a finite number of steps.
How can we rationalize this statement?
Recall the fundamental equation, dU = T dS p dV
dU = Cv dT
so
For 1 mole of ideal gas, P = RT/V

Cv dT = T dS (RT/V) dV
dS = Cv d (ln T) + R d (ln V)
20.110J / 2.772J / 5.601J

Lecture 16
T2,
5.60/20.110/2.772
For a spontaneous adiabatic process which takes the system from T1 to a lower temperature
S = Cv ln (T2/T1) + R ln (V2/V1) 0
but if T2 = 0, Cv ln (T2/T1) equals minus infinity !

Therefore R ln (V2/V1) must be greater than plus infinity, which is impossible. Therefore no
actual process can get you to T2 = 0 K.
But you can get very very close!
In W. Ketterle's experiments on "Bose Einstein Condensates" (recent MIT Nobel Prize in
Physics), atoms are cooled to nanoKelvin temperatures (T = 10-9 K) but not to 0 K !
_______________
Some apparent violations of the third law (but which are not !)
Any disorder at T = 0 K gives rise to S > 0
mixed crystals
If have an unmixed crystal, N atoms in N sites:
N!
=1
N!
S = k ln 1 = 0
But if mixed crystal:
NA of A
NB of B
NA + NB = N
N!
N A! N B !
S = k ln
Use Stirlings approx:
N!
N A! N B !
20.110J / 2.772J / 5.601J

Lecture 16
5.60/20.110/2.772
ln N != N ln N N
S = k ( N ln N N N A ln N A + N A N B ln N B + N B )
= k ( N ln N + N A ln N A + N A + N B ln N B )
Using mole fractions: NA =xAN, NB =xBN
Smix = nR X A ln X A + X B ln X B
> 0 Always !!! Even at T=0K
But a mixed crystal is not a pure substance, so the third law is not violated.
Any impurity or defect in a crystal also causes S > 0 at 0 K
Any orientational or conformational degeneracies such as in a molecular crystal causes S > 0 at 0
K, for example in a carbon monoxide crystal, two orientations are possible:
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
OC
CO
CO
CO
CO
OC
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
Phase Equilibria: One component systems

Phase diagrams
Clapeyron Equation
Phase Rule
Phase Equilibria in a One Component System

Goal: Understand the general phenomenology of phase transitions and phase coexistence
conditions for a single component system.
The Chemical Potential controls phase transitions and phase equilibria, as well as equilibrium in
chemical reactions.
Know that at equilibrium, of each component is the same everywhere in the system
If have multiple phases i, must have same value in every phase. Here well deal with one
component systems only.
Remember our definition of :
G
n
Fundamental equation:
dG = SdT + Vdp
divide by n
d = S dT + V dp
(T,P)

d =
dT + dp
T P
p T
This allows us to say that

= S
T P
and
Why are we interested in ?

Phase with the lowest value of is the most stable phase

= V
p T
20.110J / 2.772J / 5.601J

Lecture 17
For example, consider two phases (liquid and solid) of water at a fixed (T, p).
20.110/5.60/2.772 Fall 2005
If s(T, p) = l(T, p)
then liquid water and ice coexist
If s(T, p) > l(T, p)
then the water is in the liquid phase
If s(T, p) < l(T, p)
then the water is in the solid phase
How does (T)? Slope of line is entropy
= S s
T P
l
= S l
T P
steepest negative
= S g
T P
shallowest negative
Sgas >> Sliq >Ssolid
S
L
Tm
Tb
at a given T, pick the lowest value
Note also: @ T=Tm, solid and liquid phases coexist
s(T, p) = l(T, p)
Now, want to describe phase properties as a function of state variables, (p,T).
What happens if we change p? Our diagram hasnt accounted for that yet.

= V
p T
d = V dp
If decrease p, dp<0, d<0 so lines get shifted down.

Lets make a better diagram that has (p,T) as the variables
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
Phase diagrams
Describe the phase properties as a function of state variables, for example in terms of (T, p).
p
melting
critical point
boiling
triple point
sublimation
phase diagram or an equilibrium diagram
Every point on the diagram represents a state of the system

Important parts of the phase diagram:
1) planar regions (areas in between lines)
For example: @ (p1, T1): solid is the equilibrium phase. T and p can be changed independently in this
region without changing the phase
In the single phase (planar) regions of the diagram, one of the chemical potentials is lower than the
other two. T and p can be changed independently without changing phases.
2) Lines
Indicate coexistence of two phase
For the melting line, for example, solid and liquid coexist
on this line:
s(T, p) = l(T, p)
One equation [s(T, p) = l(T, p)], two variables (T, p). This means that coexistence of two phases is
described by T=f(p) or p=g(T). e.g. a line in the (T, p) phase diagram.
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
3) Triple point
At the triple point, the chemical potential of all three phases are the same solid, liquid and gas
coexist.
s(T, p) = l(T, p) = g(T, p)

Two equations, two variables. This defines a unique point (Tt, pt) in the (T, p) phase diagram.
For H2O,
Tt = 273.16 K and pt = 0.006 bar
(Recall the definition of the temperature scale in Lecture # 2)

4) Critical point
line does not go on forever! It has a distinct end. it stops at critical point (Tc, Pc).
densities equal
gas
liq
T<<Tc
TTc
T>Tc
Above (Tc, Pc), l and g become indistinguishable: single fluid phase for all T > Tc, P > Pc.
Supercritical fluids are finding remarkably practical applications.
Supercritical water (Tc = 375 C, Pc = 221 bar):
organic molecules readily soluble
inorganic salts nearly insoluble
organic compounds oxidized to CO2, N2, mineral salts
Supercritical carbon dioxide (Tc = 31 C, Pc = 75 bar):
reaction solvent, replaces chlorinated and volatile organic compounds
dry cleaning solvent, replaces perchloroethylene
Where are the phase boundaries located?

Can we understand the shape (i.e. slope) of the coexistence lines? For example, can we get an
equation for (dP/dT)coexistence?
Goal: to be able to predict, using state functions, phase transitions and equilibria.
4
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
Let and be two phases (e.g. , are l, s, or g).

On a coexistence curve, (T, p) = (T, p)
Now take
T T + dT
So then
+ d
and
p p + dp
, staying on the coexistence line.
+ d
and
AND d = d
But since
d = dG = S dT + V dp
, having
d = d
implies that S dT + V dp = S dT + V dp on the coexistence line.
S S S
dp
=
dT
V
V
coexist

V
S = entropy change when changing from

V = volume change when changing from
Another way to write this is using
= G = H TS so that
= on the coexistence line implies

or
H TS = H TS
H = T S . Using this then we obtain the two forms of the Clapeyron Equation
(These are always valid)
dp
dT
coexist
= S
or
V
dp
dT
coexist
= H
TV
General: can be applied to any phase transition of a pure substance.

Have a good qualitative feel for S, V when going from liquidgas, gassolid, etc., so we can use
the Clapeyron equation to qualitatively understand the phase diagram.
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
a) lg
S > 0 ,
V >> 0
S
dp
=
> 0 , but small

dT coexist V g
(not steep)
P
l
g
T
b) s g
S >> 0 ,
V >> 0
S
dp
=
> 0 , and

dT coexist V s g
steeper than for l g
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
c) s l
For most substances Vl Vs (almost equal)
and
S > Ss
S
dp
> 0 , and very steep
=

dT coexist V s
g
T
For most substances, raising the pressure above a liquid near the liq to solid coexistence line can
cause it to freeze
Except for one of the most important substances on earth: H2O. In this case V < Vs , so that
dp
<0 .

dT
coexist
Thus increasing the pressure above ice can cause it to melt, and the bottom of the ocean is not ice,
but liquid water at 4C.
Interestingly enough, silicon shows similar behavior (at much higher temperatures).
The Phase Rule

How many intensive variables are needed to describe state of a system?
If have two phases:

only one intensive variable (T or p) is needed to describe state of the system
system has one degree of freedom
20.110J / 2.772J / 5.601J

20.110/5.60/2.772 Fall 2005
Lecture 17
If one phase is present, two variables needed to describe state (T and p)

system has two degrees of freedom
If have three phases(,,):
(T , p) = (T , p)
and also
(T , p) = (T , p)
two equations, two unknowns: T and p are completely determined.

# phases present
degrees of freedom
Can construct a simple rule for a one component system:
F = 3 P
F = degrees of freedom
P = number of phases present
20.110J / 2.772J / 5.601J

20.110/5.60/2.772
Lecture 18
Clausius Clapeyron equation

Condensed phase equilibria
CLAUSIUS-CLAPERYON EQUATION
We derived the Clapeyron Equation last lecture:
S S S
dp
=
=
dT coexist V V V
This is exact!
Lets examine this for equilibrium between solid-liquid and solid-solid phases. For example, fusion:
S fus
dp
=
dT coexist Vfus
use
H fus = T S fus
We can make some assumptions about these types of phase equilibria.
p2
Tm'
H fus dT
T
fus
Tm
dp = V
p1
p2 p1 =
H fus
V fus
Tm'
ln
Tm
assume Hfus, Vfus are independent of p,T
Tm=melting temp @p2, Tm =melting temp @ p1
Tm-Tm~0 (expect small change in Tm)
Tm' Tm
Tm + Tm' Tm
Tm'
= ln1 +
ln = ln
T
T
T
m
m
m
Since fraction is small, then
Tm' Tm' Tm T
=
ln
T
T
m
T
m
Then
20.110J / 2.772J / 5.601J

20.110/5.60/2.772
Lecture 18
H fus Tm
Vfus Tm
gives you Tm, the melting point increase corresponding to a p increase

also says: the coexistence line is approximately linear
Equilibrium between gas/liquid or gas /solid (gas/condensed phase)

substance A:
A(l)= A(g)
or
A(s) = A(g)
Hvap
dp
dT
H
S
=
=
V T V g V c
Hsub
Vc = Vsolid or Vliquid
a) We can approximate
V g V c V g because Vg>>Vc
b) We can assume the gas is ideal
Vg =
RT
p
Clausius Clapeyron Equation
dp
dT
H
H p
RT 2
T Vg
( )
d ln p H
2
dT RT
to get here we made approximations, but very good far from Tc.
Relates the vapor pressure of a liquid to Hvap or vapor pressure of a solid to Hsub.
c) We can make yet another approximation: assume H independent of T. Then can integrate:
p
d ln p = RT
p0
T0
dT
20.110J / 2.772J / 5.601J

20.110/5.60/2.772
Lecture 18
p
H
ln
R
p0
1 1

T T0
This equation allows us to get Hsub from p
p
H H
ln
+
p
RT
RT0
0
ln(p)
-H/R
1/T
20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

0.2
0.175
n(NaOH)
0.15
0.125
0.1
0.075
0.05
pK ~ 4.8
0.025
2
pH
10
12
Titration of acetic acid with a concentrated solution of sodium hydroxide. The

number of moles of NaOH added to a liter of 0.10 M acetic acid represented by n.
Figure by MIT OCW.
20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

3
2.5
NH
2
1.5
1
0.5
2
pH
10
12
14
Average number of hydrogen ions bound by phosphate at 298.

Concentration of pH at zero ionic strength.
Figure by MIT OCW.
20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

20.110J / 2.772J / 5.601J

Melting Curves for Short and Long DNA Oligomers

(CT = 1 mM)
f, (fraction dimerized)
5'-ATAGCA-3'
3'-TATCGT-5'
5'-ATAGCAATAGCA-3'
3'-TATCGT TATCGT-5'
1.0
0.8
0.6
0.4
0.2
0.0
275
300
325
T, K
350
Figure by MIT OCW.
375
20.110J / 2.772J / 5.601J

Melting Curve Analysis To Detect Mismatches
Homozygous Wildtype
Fluorescence
3.2
Homozygous Mutant
2.8
2.4
2.0
1.6
0.24
0.20
-dF/dT
0.16
0.12
0.08
0.04
0.00
0.04
52
54
56
58
60
62
64
66
Temperature
Figure by MIT OCW.
68
70
72
74
76
78
20.110J / 2.772J / 5.601J

F1
F2
No mismatch
F1
F2
1 mismatch
Mismatch has very high Go cost
Figure by MIT OCW.
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 11
10/20/2005
1. In class, we considered the example of a polymer comprising 4 monomer units that
can exist in a compact state or an open state. The compact state arises from an
attractive bond between the monomers at the ends of the chain (see fig. 10.1 in Dill). This
bond is characterized by an energy -. If this bond has energy of 588x10-23J, what is the
(Helmholtz) free energy of the system at 310K? at 710K? Will there be more open or
compact chains at 310K? The value of k is 1.38x10-23J/K
Bound or compact:
Unbound or open:
2. Consider a particle that has two states: bonded to a surface, or non-bonded

(released). The non bonded state is higher in energy by an amount of o.
a) Explain how the ability of the particle to bond to the surface contributes to
the heat capacity, and why the heat capacity depends on temperature
b) Compute the heat capacity Cv in units of Nk if T=300K and o=1.2 kcal/mol
(which is about the strength of a weak hydrogen bond in water).
U
Recall: R=1.987 calxmol-1xK-1, Cv =
,
T
N o2
Cv =
kT 2
e kT
(1+ e
o
kT
for a two state system
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 15
11/8/2005
1. Consider equal volumes of two liquids (well refer to them as A and B) in two separate but
identical sealed containers. The rigid containers are held at constant temperature and volume.
a. What is the criterion for equilibrium to be achieved between each liquid and its
vapor in the container?
b. Measurement of the vapor pressures for the two (separate) liquids reveals a vapor
pressure of 16.9 KPa for liquid A and 2.8 KPa for liquid B at 25C. If we assume to first
approximation that the standard state chemical potentials of A and B in the vapor
phases are identical at this temperature, what is the difference in bonding energies
between these two liquids?
=AA BB =?
Express your answer in units of kbT- e.g., 10.7kbT, assume cubic lattice.
2. Osmotic pumps are used to provide slow release of drugs in the body over a long time
period. A cross section of an osmotic pump is shown in the figure. Liquid drug solution is
contained inside a flexible impermeable membrane pouch in the center of the device (drug
region). Any pressure exerted on the membrane forces drug out of the hole shown at the top
of the device. A rigid semipermeable membrane on the outside of the device surrounds a
water-soluble solute that does not pass through either membrane and is present at a
concentration of cs (osmotic region). The rigid semipermeable membrane is permeable to
water. For this problem, you may assume that the concentration of solute in the osmotic
region, cs , remains constant at 0.7 M (this is approximately true as the volume of the drug
region is very small compared to the volume of the osmotic region).
Calculate the net pressure exerted on the drug region (i.e., the pressure on the impermeable
pouch membrane minus the pressure of the external fluid) when the device comes to
equilibrium with the following external fluids:
a. Pure water
b. 0.2 M sucrose, where sucrose does not cross the semipermeable membrane
c. 0.2 M sucrose, where sucrose is able to cross the semipermeable membrane
d. 0.1 M MgCl2, where MgCl2 does not cross the semipermeable membrane.
You may use the dilute solution assumption for these calculations. Assume constant body
temperature=37oC.
20.110J / 2.772J / 5.601J

3. Purified embryonic cells in culture aggregate via cell-cell adhesion into spheres (often
referred to as spheroids), which evidences that their assembly is driven by interfacial tension.
Why?
a. What happens when two different cell types, having different interfacial tensions, are mixed?
b. Which arrangement is the equilibrium state for the system?
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 16
11/10/2005
1. In class we studied the mixing of species A and B, assuming both of them as polar
substances (such that they make hydrogen bonds with each other and also with their own
kind). Now assume that we have a three species system, A, B and C. Assume that for such a
lattice model, N = nA + nB + nC. All sites are filled.
a. Write an expression for entropy of mixing.
b. Using the mean field (Bragg-Williams) approximation, where we assume that the
particles on a lattice are mixed as uniformly and randomly as possible, write an
expression for the energy of mixing (Umix) in terms of the binary interactions
parameters chi (AB, AC, BC).
c. Write an expression for the free energy of mixing (in Helmholtz free energy).
d. Write an expression for the chemical potentials of substances A, B and C.
e. Now assume that the species A is polar, while species B and C are nonpolar and
form very weak inter-species (AC, BC etc) bonds. What do you think will happen to the
free energy of the system, relative to the situation when all species were polar? What
would happen to the entropy of the system?
20.110J / 2.772J / 5.601J

2. (D&B) You have a two-dimensional molecular lock and key in solvent s, as shown in the
figure below. Different parts of each molecule have different chemical characters, A, B, or C.
a. In terms of the different pair interactions, (AB, AC, AS, etc.) write an
expression for the binding constant K (i.e., for association).
b. Which type of pair interaction (AB, AC, BC) will dominate the attraction?
20.110J / 2.772J / 5.601J

3. Lets compare the magnitude of entropic forces that favor unfolding of proteins (gain in
chain conformations) to the entropic forces favoring folding (hydrophobic ordering of water
around exposed nonpolar side chains):
Consider a protein of 200 amino acid units. Lets estimate the entropy gain due to chain
conformations when the chain goes from a compact, unique folded state to a random coil.
Using the lattice model approximation, we showed in lecture that one could estimate the
number of states for a random coil on a 3D lattice as:
= zN
Where N is the number of repeat units (200). If we take the folded state as a unique single
state for the folded protein, we have the following for the entropy change on
denaturation:
Schain freedom
unfolded
= k B ln
folded
Schain freedom = 1.3811023
zN
= k B ln
= k B N ln z
1
J
J
J
(200) ln(6) = 4.95 1021 = 2979
K
K
K mol
Now, from measurements of individual amino acids, we can approximate that the average
entropy loss when a nonpolar amino acid is exposed to water as 100 J/mole K. Using this
value, what fraction f of amino acids in the protein chain would need to be nonpolar in order
to provide enough hydrophobic entropy loss to exactly balance the entropy gain from chain
conformations during unfolding?
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 19
11/22/2005
Force, work and heat in rubber elasticity:
1. A double stranded DNA molecule was stretch 5 um by a force of 2 pN, using optical
tweezers at a temperature of 300 K. Double stranded DNA has a bk = 100 nm and each
nucleotide measures .35 nm.
a) How long is this sequence?
b) How much work was required to pull this molecule?
c) If this process is reversible how much heat was transferred to or from the DNA?
d) If the process if isothermal, what is the entropy change?
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 20
11/29/2005
Practice Test
1.
20.110J / 2.772J / 5.601J

2. Swelling of rubber networks: entropy springs in 3D. Suppose a hydrophilic rubber

network that is initially a solid, cubic block of polymer is isotropically swollen in water. In an
isotropic deformation, the x, y, and z dimensions of the sample change identically; the
fractional deformation in each direction is .
a. Derive an expression for the elastic force resisting swelling of the network in terms of the
number of network chains m, temperature, the chain parameters N and b, and the fractional
deformation of the network, .
b. Show that this restoring force is proportional to the volume of the swollen network to the
1/3 power.
20.110J / 2.772J / 5.601J

3. Measurement of partition coefficients: You add 1 mg of a bile acid with a molecular weight
of 400 to a small vial containing 1 mL of a model oil that represents cell membranes and 1
mL of 0.1M NaCl in water. You observe that all of the bile acid dissolves and determine that
the concentration of the bile acid in the salt water phase is 0.005 mg/mL at equlibrium. What
is the partition coefficient for bile acid, K, into the oil from water?
You add 10 mg and observe that the bile acid does not all dissolve. In this case, the
concentration in the salt water phase is 0.08 mg/mL at equilibrium. What is the concentration
of bile in the oil phase for this case?
20.110J / 2.772J / 5.601J

If you have time, or want to do it instead of 1, do 4:

4. Proteins are polymers, too. We derived in class the end-to-end distance of the freelyjointed chain model of polymers, which makes use of an extremely simple set of assumptions
to determine the polymer chain conformation. The model turns out to be very useful for real
random coil polymers because we can re-scale the polymer chain, viewing a stiff polymer
chain with N segments of length b as an ideal chain that has a smaller number of segments
Nk that have a longer length bk, to account for the tendency of stiff chains to have runs of
bonds that largely point in the same direction in space.
Recently, the persistence length (which is equivalent to twice the Kuhns statistical segment
length bk) of a major glycosaminoglycan polymer component of cartilage was characterized
by a research team in Materials Science and Biological Engineering here at MIT, by
adsorbing the polymers to a substrate and directly imaging the individual (very large, by
polymer standards) macromolecules using atomic force microscopy (L. Ng et al., J. Struct.
Biol. 143, 242-257 (2003)). Two AFM images of the molecules, aggrecans, are shown below.
Aggrecans have a bottle-brush molecular structure, with a linear polymer backbone and side
chains that extend away from the backbone. The two images above are from fetal and adult
cartilage. The following data were collected for these molecules:
Images removed due to copyright reasons.
Please see:
Figure 7A in Ng, L., et al. J Struct Biol 143 (2003): 242-257.
The bond length of these molecules b is 1.2 nm.

a. Calculate the characteristic ratio for adult aggrecan molecules.
b. Calculate the number of effective bonds in the fetal aggrecan molecule, NK.
c. How many real bonds along the aggrecan backbone are included within one Kuhn
segment of the aggrecan polymer?
d. Calculate the ratio of end-to-end distances of fetal to adult aggrecan molecules.
20.110J / 2.772J / 5.601J

20.110/2.772 Fall 2005

Recitations 3-4
Week of Sep 20
1. State functions: An ideal gas is initially at the following conditions: V = 1 L, P = 1 atm,
T = 298K (State 0). It is subjected to the following cycle, where each step between
states is reversible:
Step 1: Isothermal expansion to 2 L (State 1)
Step 2: Adiabatic expansion to 3 L (State 2)
Step 3: Isothermal compression to 1.5 L (State 3)
Step 4: Adiabatic compression to 1 L (State 0)
Calculate the heat, work, and change in internal energy for the gas at each step. For
one complete round of the cycle, what is the net value of q? the net value of w? the
net value of U?
2. Heat of Unfolding of a Protein: Proteins in solution are folded into 3D structures that
are stabilized by many non-covalent bonds between the amino acids. As the temperature
of a protein solution increases, these bonds become less stable and the protein undergoes
a denaturation process where it unfolds and interacts more with water. The enthalpy of
this reaction can be measured by microcalorimetry at constant pressure, where the value
CP, protein = (H/T)P for a dilute solution is determined relative to the solvent and as a
function of temperature. A typical set of experimental data is shown in the figure.
Privalov and coworkers1 have measured
determined CP (T) for the protein lysozyme and
found the data for one of the domains fit the
following form, where Cp is relative to the
solvent
Cp
Cp = CP(Tt) + b (T Tt) + c (T2 Tt2)

In this expression, Tt refers to the temperature at
T
the peak of the transition, and all temperatures are
in K. Determine the enthalpy of denaturation of
lysozyme if Tt = 41.5C, b = 0.3158 J/K2-mol and c = 0.462 x 10-3 J/K3-mol. The
temperature range for the transition is 20-60C.
Y.V. Griko, E. Friere, G. Privalov, H. van Dael, P.L. Privalov, J. Mol. Biol., 252, 447-459
(1995)
20.110J / 2.772J / 5.601J

20.110/2.772/5.601 Fall 2005

Recitation # 17
11/15/2005
1. A recently graduated Biological Engineer was just hired at a big Biotechnology company.
The first task she was assigned was to work on increasing the half-life of a newly developed
therapeutic antibody in the body. She knows that this can be achieved by binding
polyethylene glycol (PEG, OH-(-CH2CH2O-)n-H) to the protein, since this decreases
clearance in the body, without significantly affecting protein activity. The problem is that it is
really hard to control the size of the polymer chain and she doesnt have too much
experience with chemical synthesis of polymers, so she goes and measures the radius of
gyration (Rg) of the synthesized PEG by laser light scattering and finds that Rg = 5nm. The
characteristic ratio (CN) of PEG is 4 in water, and the bond length is 0.3 nm.
a) Calculate the average number of rods (NK), number of monomers (N) and molecular
weight of the PEG synthesized. (Usually, the size of PEG bound to therapeutic
proteins ranges from 5 to 30 kDa)
b) Polystyrene ( -(-CHC6H5CH2-)-n ) has a characteristic ratio of 9.6 in water, and a
bond length of .15 nm. Calculate NK, N and molecular weight for this polymer if its
radius of gyration is 5 nm. Would you bind this to a drug? Why?

201105spring 2001

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

201105spring 2001

Hochgeladen von

Copyright:

Verfügbare Formate

20.110J / 2.772J / 5.

20.110/5.60 Fall 2005

0th Law Defines Temperature (T)

1st Law Defines Energy (U)

2nd Law Defines Entropy (S)

3rd Law Gives Numerical Value to Entropy

These laws are UNIVERSALLY VALID, they cannot be circumvented.

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Systems can be:

Describing systems requires:

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Two classes of Properties:

The State of a System at Equilibrium:

These are Equations of States

The van der Waals equation of state:

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

2 H2O (, 1 bar, 50 C) = 3 H2O (g, 1 bar, 150 C)

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Path: Sequence of intermediate states

Process: Describes the Path

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Thermal Equilibrium (Heat stops flowing)

When a hot object is placed in thermal contact with a cold object,

ZEROth LAW of Thermodynamics

are in thermal equilibrium and

are in thermal equilibrium,

are in thermal equilibrium.

Consequence of the zeroth law:

acts like a thermometer, and

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Operational definition of temperature (t)

Example: Ideal Gas Thermometer with the Celsius scale.

the substance is a gas

t = 273.15 C is called the absolute zero,

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

This suggests defining a new temperature scale (Kelvin)

T = 0K corresponds to absolute zero (t = 273.15 C )

T = 0K (absolute zero) and Ttp = 273.16K (triple point of H2O)

valid for all gases for p 0

the gas constant

An ideal gas obeys the expression pV = RT at all pressures

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Having a - sign here implies w > 0 if V < 0 , that

If heat enters the system, then it is

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Work, Heat, and the First Law

depends on the path!!!

Other kinds of work

Example: assume a reversible process so that pext = p

Ar (g, p2, V2)

V1 > V2 and p1 < p2

20.110J / 2.772J / 5.601J

20.110/5.60 Fall 2005

Ar(g, p1, V1) = Ar(g, p1, V2) = Ar(g, p2, V2)

Ar(g, p1, V1) = Ar(g, p2, V1) = Ar(g, p2, V2)