CHEM 3322: Physical Chemistry II

Overall Goals:
◦ Use quantum mechanics to understand that molecules have quantized translational,
rotational, vibrational, and electronic energy levels.
◦ In a large sample of molecules (∼ NA ), be able to predict how many molecules are in
each energy level (populations).
◦ Understand the basic principles of spectroscopy using selection rules,the energy levels,
and the populations.
◦ Learn some simple theories of chemical bonding.

Classical Waves
A classical wave is defined as a solution of the classical wave equation:
∂ 2 u(x, t) 1 ∂ 2 u(x, t)
= (1)
∂x2 v 2 ∂t2
where v is the speed of wave propagation.
The wave equation is a partial differential equation. Let us try to solve it using separation
of variables: u(x, t) = X(x)T (t). Specifically, try X(x) = sin ax and T (t) = cos bt because
these are functions whose second derivative is similar to the original function. This works
if v 2 = b2 /a2 . We need ax and bt to be dimensionless; a thus has to be the wavenumber
= k = 2π/λ and b the frequency = ω = 2πν. Then we have v = ω/k.
Now let us use the separation of variables technique more formally, using the following
recipe:
1) substitute the separation of variables form u(x, t) = X(x)T (t) into the partial differ-
ential equation. This gives
1
X 00 T = XT 00 (2)
v2
2) divide both sides by u(x, t) = X(x)T (t), (and rearrange as needed to get the indepen-
dent variables on either side of the equation) giving
X 00 1 T 00
= 2 (3)
X v T
3) argue that the only way this can work is if both sides are equal to a constant. We will
call this constant −c2 .

1
 4) write two separate equations, which are now ordinary differential equations, namely

X 00 = −c2 X and T 00 = −c2 v 2 T (4)

We can use the trigonometric identity sin A cos B − cos A sin B = sin(A − B) to rewrite
the wave form as sin(kx − ωt). Then we can see that the following four functions are all
valid (monochromatic) classical waves:

sin(kx − ωt) , cos(kx − ωt) , ei(kx−ωt) , e−i(kx−ωt) (5)

In 3 spatial dimensions these classical waves generalize to:

sin(k · r − ωt) , cos(k · r − ωt) , ei(k·r−ωt) , e−i(k·r−ωt) (6)

where k = (kx , ky , kz ) is the wave vector which points in the direction of propagation of the
wave, and r = (x, y, z) is the position vector. k has length |k| = 2π/λ with ω defined as
before.
The 3d wave equation is
1 ∂ 2u
∇2 u = (7)
v 2 ∂t2
where ∇2 is the Laplacian operator.
∂2 ∂2 ∂2
∇2 = + + (8)
∂x2 ∂y 2 ∂z 2

Classical Particles
Newton’s 2nd Law:
dv d dp
F = ma = m = (mv) = (9)
dt dt dt
We can express this as dp = F dt

dT = F · dr (10)

Proof:
p2
 
1
dT = d = 2p · dp = v · F dt (11)
2m 2m

2
Then we can use v = dr/dt ⇒ vdt = dr
If F = −∇V where V (x, y, z) is the potential energy function, then dT = −∇V · dr =
−dV (total derivative). Hence dT + dV = 0 ⇒ d(T + V ) = dE = 0 This is the expression for
conservation of energy. Note: Energy is not conserved if the forces felt by an object cannot
be associated with a potential energy function. Examples: (i) friction → forces depend on
velocity, (ii) magnetic forces → forces depend on velocity.
Photons
Planck (1900): “blackbody” radiation
Einstein (1905): photoelectric effect
Light (electromagnetic radiation) behaves like a field of particles called “photons,” each
with energy E = hν
From electromagnetic theory, the momentum should be p = E/c as verified by Compton
(1922) in the scattering of x-rays from electrons. [One way to think about this is to use
Einstein’s formula E = mc2 , which yields p = E/c = mc2 /c = mc which is of the form
“momentum equals mass times velocity” with the velocity being the speed of light.]
Using c = λν we can write p = h/λ
Wave-Particle Duality
de Broglie (1923): if light (a “wave” phenomenon) can behave like particles, then particles
should behave like waves. Wave properties (λ, ν) and particle properties (p, E) are connected
by p = h/λ, E = hν, or, in terms of k and ω, p = ~k, E = ~ω, where ~ ≡ h/2π.
Schroedinger Equation
“Matter waves” in 1d could have the form

sin(kx − ωt), cos(kx − ωt), ei(kx−ωt) (12)

where k and ω are related by

p2 ~2 k 2
E= ⇒ ~ω = (13)
2m 2m

For “ordinary” waves, we had ω = kv and therefore matter waves cannot satisfy the
“ordinary” wave equation. After considering the effect of ∂/∂x and ∂/∂t on the three wave
forms, we deduce that:
~2 k 2
1. a linear wave equation consistent with ~ω = 2m
cannot be found for sin(kx − ωt) or
cos(kx − ωt)

3
 2. for ψ(x, t) = ei(kx−ωt) the equation
∂ψ ~2 ∂ 2 ψ
i~ =− works! (14)
∂t 2m ∂x2
Notice that both the wave equation and its solutions ψ(x, t) are complex. The wave
equation is called the Schroedinger Wave Equation (1926). It is valid for particles in free
space only (no forces acting). In 3d, it has the form
∂ψ ~2 2
i~ =− ∇ψ (15)
∂t 2m
Schroedinger Equation in a Potential
Consider the simple situation below:
(figure here)
There is a constant potential in both regions (i.e. zero force) hence we have a wave of
form ei(kx−ωt) on both sides of the boundary. For continuity at the boundary, ω (i.e. the
frequency) must be the same on both sides. This suggests that ω is related to the total
energy E and not the kinetic energy T because only the total energy E is conserved over
the boundary. Therefore

E (total) = ~ω (both sides) (16)

However, k changes:
In region 1:
p21 ~2 k12
E= ⇒ ~ω = (17)
2m 2m
In region 2:
p22 ~2 k22
E= + Vconst ⇒ ~ω = + Vconst (18)
2m 2m
A wave equation that works in both regions is:
∂ψ ~2 ∂ 2 ψ
i~ =− +Vψ (19)
∂t 2m ∂x2
where 
0 :x<0
V (x) =
 Vconst : x ≥ 0

For a general potential V (x), think about approximating it in a piecewise manner:

4
 (figure here)
and imagine the limit of infinitely small steps. We get

∂ψ ~2 ∂ 2 ψ
i~ =− + V (x)ψ (20)
∂t 2m ∂x2

In 3d, this becomes

∂ψ ~2 2
i~ =− ∇ ψ + V (r)ψ (21)
∂t 2m

and is called the time dependent Schroedinger equation (TDSE). The solutions ψ(x, y, z, t) =
ψ(r, t) are called wavefunctions.
Interpretation of ψ
What is the physical interpretation of ψ?
|ψ|2 = ψ ∗ ψ is a probability density in space. What this means is that the probability of
finding the particle in the volume element dx dy dz at time t is |ψ(x, y, z, t)|2 dx dy dz. And
the probability of finding the particle in the extended region A is
ZZZ
|ψ(x, y, z, t)|2 dx dy dz (22)
A

For this interpretation to make sense we must have

ZZZ
1= |ψ(x, y, z, t)|2 dx dy dz (23)
all space

Wavefunctions ψ satisfying condition (23) are called normalized.

General Solution for ψ(x, t)
Our goal is to solve Eq. (20). This equation is a linear partial differential equation. We
will try the separation of variables approach. Namely, we will assume we can express the
solution ψ(x, t) as a product of two one-variable functions, ψ(x, t) = ψn (x)θn (t). Note that
although ψ and ψn have similar names, they are very different since the former is a function
of two variables while the latter is a function of one variable. To begin, note that

∂ψ dθn ∂ 2ψ d2 ψ n
= ψn and = θn (24)
∂t dt ∂x2 dx2

Substituting into Eq. (20) yields

dθn ~2 d2 ψn
i~ψn =− θn + V ψn θn (25)
dt 2m dx2

5
Now divide both sides by ψ = ψn θn to give

1 dθn ~2 1 d2 ψn
i~ =− +V (26)
θn dt 2m ψn dx2

The left hand side (LHS) is a function of time only, and the right hand side (RHS) is a
function of space only. We have separated the two independent variables to either side of
the equation. How can this work? Only if both sides are constant. Since the dimensions of
the constant are energy, call it En
Then,

1 dθn dθn iEn

LHS becomes i~ = En ⇒ =− θn (27)
θn dt dt ~

This ordinary differential equation has the simple solution

θn (t) = e−iEn t/~ (28)

and,

~2 d2 ψn
RHS becomes − + V ψn = En ψn (29)
2m dx2

with solutions depending on the potential energy V (x). This equation has many solutions
for different energy values En (hence the n subscript, to label the different solutions). This
equation is called the time independent Schroedinger equation (TISE).
Putting everything back together, the general solution for ψ(x, t) is
X X
ψ(x, t) = cn ψn (x)θn (t) = cn ψn (x)e−iEn t/~ (30)
n n

with linear combination coefficients cn ∈ C

Stationary States
Consider the special case where all the coefficients are zero except for c1 = 1. Then

ψ(x, t) = ψ1 (x)eiE1 t/~ (31)

The probability density in this special case is |ψ(x, t)|2 = |ψ1 (x)|2 , namely time independent.
Hence the probability density is literally “stationary” or static. In electromagnetic theory,
a static charge density cannot emit radiation. This suggests that these are the stationary,
non-radiating, states imagined by Bohr (1913). To summarize, the stationary (or “allowed”)

6
states of a particle under the influence of the potential energy V are given by the time
independent Schroedinger equation

~2 d2 ψn
− + V ψn = En ψn (32)
2m dx2
where the constant En is the energy of the particle in that state (will prove later).
Particle in a Finite Box: 1d
Consider the following potential energy function:

 V :x<0 region A
 0


V (x) = 0 :0≤x<L region B


 V0 : x ≥ L

region C
Our goal is to find the stationary (non-radiating, stable, allowed) states for a quantum
mechanical particle experiencing V (x), whose energy E satisfies 0 ≤ E < V0 . To do this we
must solve the TISE which is Eq. (29). We will solve it for each region separately because
V (x) is defined in a piecewise manner, and then we will combine the pieces to assemble the
entire solution.
Regions A and C

~ 2 d2 ψ n
− + V0 ψn = En ψn (33)
2m dx2
Rearranging gives

d2 ψn 2m 2m
2
= − 2 (En − V0 )ψn = 2 (V0 − En )ψn (34)
dx ~ ~
The constant prefactor multiplying ψn on the RHS is positive and so we can define c2 =
2m(V0 − En )/~2 with c ∈ R. This is a 2nd order linear ordinary differential equation so we
are guaranteed to cover all possible solutions if we find two linearly independent solutions
and form their linear combinations. Clearly two linearly independent solutions are:

ψn+ (x) = ecx and ψn− (x) = e−cx (35)

We eliminate the ψn− solution in region A and the ψn+ solution in region C on physical
grounds.
Region B

7
 ~2 d2 ψn
− = En ψn (36)
2m dx2
Rearranging gives
d2 ψn 2mEn
2
= − 2 ψn (37)
dx ~
The constant prefactor multiplying ψn on the RHS is negative. We will define a2 = 2mEn /~2
with a ∈ R. Two linearly independent solutions are:

ψnsin (x) = sin(ax) and ψncos (x) = cos(ax) (38)

Combine the pieces

Actually, we can multiply the given wavefunctions by an arbitrary constant, so that in
fact the solutions we have found can be expressed as:
Region A: ψA = Aecx
Region B: ψB = B sin(ax) + C cos(ax)
Region C: ψC = De−cx
Thus we have 4 unknowns (A, B, C, and D). To combine the pieces, we must ensure
continuity and differentiability across the regions, which gives us 4 equations:
ψA (x = 0) = ψB (x = 0)
ψA0 (x = 0) = ψB0 (x = 0)
ψB (x = L) = ψC (x = L)
ψB0 (x = L) = ψC0 (x = L)
Particle in a Box: 1d
Now consider the limiting case V0 → ∞. The value of c in Eq. (35) approaches ∞ which
makes the solutions identically zero in regions A and C. Therefore the region B solution
ψB = B sin(ax) + C cos(ax) is subject to the “boundary conditions” ψB (x = 0) = 0 and
ψB (x = L) = 0. The x = 0 boundary condition forces C = 0, so that we are left with
√
ψB = B sin(ax). The x = L boundary condition is then B sin(aL) = 0 = B sin( 2mEn L/~).
This is satisfied for
p
2mEn L/~ = nπ, n ∈ Z (39)

namely the energy is restricted to a discrete spectrum:

n2 π 2 ~ 2
En = n∈Z (40)
2mL2

8
Another way to say this is that the energy is “quantized”, with n the quantum number.
Note that since the potential energy in region B is zero, the energy we are talking about
here is kinetic energy.
Substituting Eq. (40) into ψB yields
 nπx 
ψn = B sin (41)
L
where we have labeled the wavefunction with its quantum number. The constant B is then
determined by requiring normalization:
Z L Z L
2 nπx
  p
2 2
1= |ψn (x)| dx = B sin dx = B 2 (L/2) ⇒ B = 2/L (42)
0 0 L

Particle in a Box: 3d
Denote the edge lengths of the box as a, b, and c in the x, y, and z directions, respectively.
Inside the box we have (since V = 0)
~2 2
− ∇ ψ = Eψ (43)
2m
with boundary conditions ψ(0, y, z) = 0; ψ(a, y, z) = 0; ψ(x, 0, z) = 0; ψ(x, b, z) = 0;
ψ(x, y, 0) = 0; ψ(x, y, c) = 0. This is a linear partial differential equation so we will try the
separation of variables technique. We postulate that we can express

ψ(x, y, z) = f (x)g(y)h(z) (44)

Then ∇2 ψ = f 00 (x)g(y)h(z) + f (x)g 00 (y)h(z) + f (x)g(y)h00 (z) = −2mE/~2 f (x)g(y)h(z).

Divide through by ψ = f (x)g(y)h(z) to get
f 00 (x) g 00 (y) h00 (z) 2mE
+ + =− 2 (45)
f (x) g(y) h(z) ~
For this to work each individual term must be constant. If we call these constants −kx , −ky ,
and −kz then we have

kx + ky + kz = 2mE/~2 (46)

and

f 00 (x) = −kx f (x) (47)

g 00 (y) = −ky g(y) (48)

h00 (z) = −kz h(z) (49)

9
 √
Consider the x-equation: the solution is f (x) = A sin( kx x) since the cosine solution is
not allowed by the x = 0 boundary condition. To satisfy the boundary condition at x = a
√
we must have kx a = nx π, nx ∈ Z, or kx = n2x π 2 /a2 . Similarly, ky = n2y π 2 /b2 and
kz = n2z π 2 /c2 . Put these into Eq. (46) to get
nx n2y n2z
 2 
2 2mE
+ + π = (50)
a2 b2 c2 ~2
or
π 2 ~2 n2x n2y n2z
 
E= + 2 + 2 (51)
2m a2 b c
where, for reasons similar to the 1d case, nx , ny , and nz ∈ N = {1, 2, 3, . . .}. The
normalized stationary states are
8 1/2  n πx 
x
 n πy 
y
 n πz 
z
ψnx ,ny ,nz (x, y, z) = sin sin sin (52)
abc a b c
If the box is cubic then a = b = c = L and E becomes
π 2 ~2
Enx ,ny ,nz = (n2x + n2y + n2z ) (53)
2mL2
Now we have degenerate energy levels, namely different states with the same energy.

nx ny nz n2x + n2y + n2z ∝ E (54)

1 1 1 3 (55)



 2 1 1 6

triple degeneracy 1 2 1 6 (56)


1

1 2 6

1 2 2 9



triple degeneracy 2 1 2 9 (57)


1

2 2 9



 3 1 1 11

triple degeneracy 1 3 1 11 (58)


1

1 3 11

2 2 2 12 (59)
.. .. .. ..
. . . . (60)

10
 This is the simplest example of a general rule in quantum mechanics: symmetry often
implies degeneracy.
Particle on a Ring
Consider a ring of radius R, with V = 0 on the ring and V = ∞ off the ring. By analogy
to the particle in a box problem, the wavefunction is identically zero off the ring. Therefore
we only need to find the solution ψn (x) on the ring.
The TISE with V = 0 is
~2 ∂ 2 ψn (x)
− = En ψn (x) (61)
2m ∂x2
where x is the distance along the circumference of the ring (called arclength). It is more
natural to use θ for this problem where x = Rθ. Changing variables gives the TISE as
~2 ∂ 2 ψn (θ) ∂ 2 ψn (θ) 2mR2 En
− = E ψ
n n (θ) or = − ψn (θ) (62)
2mR2 ∂θ2 ∂θ2 ~2
The boundary conditions are not obvious for this problem. It turns out they they consist
of enforcing the periodicity of the angular variable; namely ψn (θ = 0) = ψn (θ = 2π),
ψn0 (θ = 0) = ψn0 (θ = 2π), ψn00 (θ = 0) = ψn00 (θ = 2π), . . . Both real (sin and cos) and complex
(e+i and e−i ) solutions are possible. We will use the real representation. If we define
c2 = 2mR2 En /~2 then the sin solution is ψnsin (θ) = A sin(cθ) with the boundary condition
2πc = 2nπ, n ∈ Z. Thus

ψnsin = An sin(nθ), n ∈ Z > 0 (63)

and
n2 ~2
En = (64)
2mR2
Likewise, the cos solution is ψncos (θ) = A cos(cθ) with the boundary condition 2πc =
2nπ, n ∈ Z. Thus

ψncos = An cos(nθ), n ∈ Z ≥ 0 (65)

and
n2 ~2
En = (66)
2mR2
Negative n values are redundant because only |ψ|2 has physical meaning. But n = 0 gives a
valid solution for cos, so we need to be careful not to miss this.

11
 Summary: The ground state is just ψ0cos = 1 with E0 = 0. The excited states have
energies

n2 ~2
En = (n ∈ Z > 0) (67)
2mR2

and are doubly degenerate with pairs of states, ψnsin = sin(nθ) and ψncos = cos(nθ) for each
energy level (ignoring the normalization constants). This is a useful conceptual model for
cyclic π-electron systems.
Harmonic Oscillator
Consider the TISE with V (x) = 12 kx2 . k is called the force constant because in Newtonian
mechanics the force is given by F = −kx (Hooke’s Law spring). The classical oscillation
p
frequency is ω = k/m where ω = 2πν. In terms of ω, the potential energy is V (x) =
1
2
mω 2 x2 . Expressed in this form, the Schroedinger equation to solve is

~2 ∂ 2 ψn 1
− + mω 2 x2 ψn = En ψn (68)
2m ∂x2 2

∂ 2 ψn m2 ω 2 x2 2mEn
 
⇒ = − ψn (69)
∂x2 ~2 ~2
2
Let us try ψ0 (x) = e−αx for the ground state. Taking two derivatives, we obtain (4α2 x2 −
2
2α)e−αx . Comparing with the RHS of Eq. (69) we see that this guess can work if we make
sure that

m2 ω 2 2mE0
4α2 = 2
and 2α = (70)
~ ~2

which means that

mω
α= (71)
2~

and E0 = ~2 α/m or

1 1
E0 = ~ω = hν (72)
2 2

This result is significant because of the zero point energy. The ground state has zero nodes.
We know that we must add nodes for the excited states. One way to introduce more nodes
is to decrease the wavelength of sin and cos functions. But we don’t have a trigonometric
solution here. Another common way to introduce more nodes is to use the fact that an nth

12
degree polynomial has n roots. We can think of the ground state as including the zeroth
degree polynomial x0 = 1. Along these lines, we can guess that the first excited state looks
2 2
like ψ1 (x) = xe−αx . Taking two derivatives, we obtain (4α2 x2 − 6α)xe−αx . Comparing
with the RHS of Eq. (69) we see that this guess can work if we make sure that

m2 ω 2 2mE1
4α2 = 2
and 6α = (73)
~ ~2

which means that

mω
α= (74)
2~

3
E1 = ~ω (75)
2

In general, the nth harmonic oscillator stationary state has energy

En = (n + 1/2)~ω, n ∈ Z ≥ 0 (76)

and wavefunction

2
ψn (x) = (xn + Axn−2 + Bxn−4 + · · · )e−αx (77)

with the final term in the polynomial being x1 = x or x0 = 1 depending on whether n is

odd or even, and with α = mω/2~.

13