Homework 3 (Due Wednesday, September 4th)

1. Show that when the coefficient of thermal expansion, α, is of the form, α = 1/(T + T0 ), where T0 is
a constant, that (∂CP /∂P )T = 0.

2. The adiabatic volume expansivity may be defined as

( )
1 ∂V
αS =
V ∂T S
(a) Show that
CV KT
αS = −
VT α
(b) Find αS for an ideal gas.

3. (a) Show that ( ) ( )

∂U ∂P
=T −P
∂V T ∂T V
(b) Verify this result for an ideal gas by explicitly evaluating the left and right hand sides of the equation.

4. Consider a system with the equation of state

(P + f (V ))(V − bN ) = N kT

where b is a constant.
(a) In general CV = CV (V, T ); show that for this system CV is independent of volume.
(b) Given that the isothermal compressibility for this system is
V − bN
KT =
PV
find f (V ) and show that the equation of state may be written as
N kT − E
P =
V − bN
where E is a constant.

5. A given material has an equation of state

P V = AT 3

where A is a known constant. The internal energy is

U = BT a ln(V /V0 ) + f (T )

where B, a, and V0 are all constants and f (T ) is a function that only depends on the temperature.
Find B and a.

1
 Solutions
1. We first need to rearrange the derivatives as,
( ) ( ( ) )
∂CP ∂ ∂S
= T
∂P T ∂P ∂T P T
∂2S
= T
∂P
( ∂T( ) )
∂ ∂S
= T
∂T ∂P T P

Using the Maxwell relation, ( ) ( )

∂S ∂V
=− = −V α
∂P T ∂T P

then
( ) ()
∂CP ∂
= T (−V α)
∂P ∂T
T
( ) P ( )
∂α ∂V
= −V T − αT
∂T P ∂T P
( )
∂α
= −V T − αT (αV )
∂T P

Finally, we insert our expression for α,

( ) ( )
∂CP V
= T − (1/(T + T0 )) V
2
∂P T (T + T0 )2
= 0

2. We want to rewrite the definition of αS as,

( ) ( ∂S ) −CV
1 ∂V −
αS = = ( ∂T
∂S
)V = ( ∂P
T
)
V ∂T S V ∂V T
V ∂T V

Notice that in the numerator we used the definition of heat capacity while in the denominator we used
a Maxwell relation. The last few steps are,
−CV
CV 1 CV KT
αS = ( ∂P )T ( ∂V ) =− =−
−V ∂V T ∂T P
V T (1/V KT )(αV ) VT α

I must admit that I solved this by working backwards.

(b) For an ideal gas, α = 1/T , KT = 1/P so αS = −CV /P V . Notice that the adiabatic compress-
ibility for an ideal gas is always negative. Since entropy increases with both temperature and volume,
if entropy is fixed then temperature increases when volume is decreased (i.e., an adiabatic compression
leads to an increase of temperature).

2
3. (a) From the first and second laws of thermodynamics,

dU (T, V ) = T dS − P dV

so ( ) ( )
∂U ∂S
=T −P
∂V T ∂V T
Using a Maxwell relation, ( ) ( )
∂S ∂P
=
∂V T ∂T V
which gives the desired result.
(b) For an ideal gas U is independent of volume so (∂U /∂V )T = 0. From the ideal gas law,
(∂P ∂T )V = P/T and inserting this into the right hand side also gives zero.

4. (a) The result we want to establish is that

( )
∂CV
=0
∂V T

Since ( )
∂S
CV = T
∂T V
then
( ) ( ) ( ) ( ) ( )
∂CV ∂ ∂2S
∂S ∂ ∂S
= T =T =T
∂V T ∂V T V ∂T
∂V ∂T ∂T V ∂V T
( ) ( ) ( 2 )
∂ ∂P ∂ P
= T =T
∂T V ∂T V ∂T 2 V

where we used a Maxwell relation to replace entropy with pressure. Finally, applying this general result
to our present equation of state,
( ) ( )
N kT ∂P Nk ∂2P
P = − f (V ); = ; =0
V − bN ∂T V V − bN ∂T 2 V

which establishes our result.

(b) From the definition of the isothermal compressibility,
( )
1 ∂V
KT = −
V ∂P T

Evaluating the derivative ( )

∂P N kT df
=− −
∂V T (V − bN )2 dV
gives
1 V − bn
KT = =
V N kT
(V −bn)2 +V df
dV P V + f (V )V + V (V − bn) dV
df

3
Comparing with the given expression for the compressibility implies
df
f + (V − bn) =0
dV
or
df dV
=−
f V − bN
so
eC
ln f = − ln(V − bN ) + C ⇒ f (V ) =
V − bN
where C is the constant of integration. Finally, putting this expression for f (V ) into the equation of
state allows us to arrange terms such that
N kT N kT eC
P = − f (V ) = −
V − bN V − bN V − bN
which gives the desired result by identifying the constant E = eC .

5. Recall that from dU = T dS − P dV we have

( ) ( ) ( )
∂U ∂S ∂P
=T −P =T −P
∂V T ∂V T ∂T V
where a Maxwell relation is used to convert the derivative on the right hand side from entropy to
pressure (see Exercise 3). Since we’re given U (V, T ) and P (V, T ), we directly evaluate
( )
∂U BT a
=
∂V T V
and ( )
∂P 3AT 2
=
∂T V V
so that { }
BT a 3AT 2 AT 3
=T −
V V V
Comparing terms we see that a = 3 and B = 2A.
Note that if you started from the relation U = T S − P V and write

dU = T dS + SdT − P dV − V dP

it is not correct to write ( ) ( ) ( )

∂U ∂S ∂P
=T −P −V
∂V T ∂V T ∂V T
The last term is not present because we know that SdT = V dP from Gibbs-Duhem so if the SdT term
is absent, so is the V dP term. Another way to see the problem with the above expression is that energy
can only be a function of two independent variables, such as U (S, V ) or U (T, V ). That is, we cannot
independently vary S, P , and V ; again, this is the origin of Gibbs-Duhem.