Thermodynamics Lecture 3 Heat Capacities

To understand heat capacities, first consider a case where all the thermal energy is
translational thermal motion. For an ideal monatomic gas, we can treat each atom as a
point of mass moving at an average velocity v. There are no chemical bonds, so all of U
is associated with translational motion of the atoms. Then, for one mole of gas, U = No x
½ mv2 where No is Avogadro’s number and m is the mass of the atom.

If the gas is contained in a cubic volume V = L3, and atoms moving in the volume
bounce off of the walls of the cube. Assume that the gas is at a low pressure, so the
atoms only hit the walls and do not collide with each other. The atoms exert a force on
the wall when they hit and bounce off in an elastic collision. Before colliding, and atoms
momentum is mv, and after the collision it is –mv. Therefore the momentum change per
collision is 2mv. After each collision, the atom travels back across the volume (distance
L), hits the opposite wall, and comes back again. Therefore the time between collisions
of a particular atom with a particular wall is 2L/v, and the frequency of collisions on that
wall is v/2L. Then the force exerted on the wall by a particular atom is 2mv x v/2L =
mv2/L.

If the velocities of the atoms are random, then 1/3 of them move in each of the x, y, and z
axes. Then the total force on the wall when one mole of atoms are contained in the box is
F = No /3 x mv2/L.
The pressure exerted on that wall is P = F/L2 = No /3 x mv2/L3 = No /3 x mv2/V.

Therefore PV = No /3 x mv2 = 2/3 No x 1/2 mv2 = RT (by the ideal gas law).

The total kinetic energy of the gas molecules in the box is U = No x 1/2 mv2 , so

PV = 2/3 U = RT, or

U = 3/2 RT.

Then the heat capacity at constant volume is Cv = (dU/dT)V = 3/2R

This works pretty well for noble gases, such as He, Ne, Ar, Kr, Xe. It also applies for
metal vapors of low boiling point metals, like Hg, Na, K, Zn, etc., and it applies for metal
vapors used in vapor deposition processes.

Cp is always larger than Cv because of the PV work done when material expands at
constant pressure.

H = U + PV

Cp = (dH/dT)p = (dU/dT)p + P(dV/dT)p

For an ideal monatomic gas, because there are no chemical bonds so all of the energy is
associated with translational motion,

(dU/dT)p = (dU/dT)v Note that this is not necessarily true if the energy is in vibrational
motions. But for ideal gases:

Cp = Cv + P(dV/dT)p = Cv + P(dRT/P /dT)p = Cv + R = 5/2 R

Gases of polyatomic molecules have higher heat capacities than monatomic gases,
because energy can also be stored as rotational and vibrational motion. For diatomic
molecules (such as O2 or N2): Cv ~ 5/2 R + temperature dependent terms. More
complex molecules, such as CO2 or NH3 have higher heat capacities still.

In condensed phases, all of the energy is associated with vibrational motion and
the heat capacities are functions of temperature. At high enough temperatures, the heat
capacities approach a value of 3R, and decrease to 0 at 0 K. This is discussed in Gaskell,
section 6.2. Unless one is working at cryogenic temperatures, it is not usually necessary
to know the variations of heat capacities at very low temperatures. The usual method is
to measure the heat capacity experimentally, and fit it to an empirical function of
temperature. A number of these are given in Gaskell in Table A-2 at the end of the book.
Gaskell uses:

Cp =a + bT + cT-2. with temperature in K

Note that the b term is usually small and the c term is large, since it is divided by
temperature squared. For example for Ag(s)

Cp = 21.30 + 8.54 x 10-3 T + 1.51 x 105 T-2. (J/(mole K)

Note that:

Heat capacities of different phases of the same material are different, as would be
expected because the chemical bonds are different.
Heat capacities are measured over a specific temperature range. If you use it beyond the
measured range you are extrapolating.
In some cases, the accuracy of the data does not require using all three terms of the
polynomial. In some cases an additional term is needed. Note that many computer
programs use 5 or 6 term polynomials, since they are computed automatically anyway.