Phys. 2b 2025, Lecture Notes (Lect.

3 & 4) (1/14-16/2025)

Key Concepts
1. Solving Schrödinger’s Eq & Stationary States
2. Solutions to Infinite Well Potential

Recall:
∂ψ ℏ2 ∂ 2 ψ
iℏ =− +Vψ
∂t 2m ∂x2
∂ ∂2 p̂2
Since p̂ = iℏ , we get p̂2 = −ℏ2 2 , and the first term on the right becomes ψ = T̂ ψ,
∂x ∂x 2m
where T̂ is the kinetic energy operator. Thus Schrodinger’s Eq becomes:
∂ψ  
iℏ = T̂ + V̂ ψ = Ĥψ
∂t

Where Ĥ is called the Hamiltonian or total energy operator.

How to solve it?
Can solve the full Schrödinger Equation (iℏ ∂ψ ∂t
= Ĥψ) easily for a special case:
if Ĥ is not an explicit function of time - i.e. V̂ ̸= V (t).

To do this, we assume separable solns exist, namely Ψn (x, t) = ψn (x)ϕn (t) (see text).
Plugging this in Schrödinger Eq. leads to two separate differential equations (one
time-dependent and one x-dependent. Thus both must be equal to a constant (≡ En ).

1 ∂ϕn 1
→ iℏ = En , and Ĥψn = En
ϕn ∂t ψn
First equation clearly has exponential soln:
iEn t
ϕn (t) = e− ℏ

and second equation is the called the Time-Independent Schrödinger Equation and is also
an Eigenvalue Equation:
Ĥψn = En ψn
where ψn is an eigenfunction on eigenstate and En is an eigenvalue or eigenenergy.
Note:

1. In some cases En are discrete due to Boundary Conditions

2. In other cases En are continuous if no Boundary Conditions

3. These solutions Ψn (x, t) are called stationary states since Ψ∗ Ψ = f (x) ̸= g(t).

4. These solns form a ”complete set” (see Ch3) i.e. any arbitrary soln is a superposition of
∞
X iEn t
ψn ϕn via Ψ(x, t) = cn ψn (x)e− ℏ , where cn are constants.
1

1
EXAMPLE I: Infinite Square Well
Particle of mass m confined in an infinite 1-d potential
V (x) = 0; 0 < x < a
V (x) = ∞; x ≤ 0, x ≥ a

V(x)
Find: En , ψn (x) by solving Ĥψn = En ψn , with

p̂2x −ℏ2 d2
Ĥ = = for 0 < x < a
2m 2m dx2
Ĥ = ∞ for x ≤ 0, x ≥ a

0 a x

For a particle with finite energy, we must have ψn = 0 for x ≥ a and for x ≤ 0.
In addition, for 0 < x < a we need to solve a simple diff. eq.:

−ℏ2 d2 ψn
= En ψn
2m dx2
d2 ψn
 
−2mEn 2mEn
⇒ 2
= 2
ψn = −kn2 ψn with kn2 =
dx ℏ ℏ2
∴ the general solution is ψn (x) = A sin (kn x) + B cos (kn x)
⇒ but we must also satisfy the “Boundary Conditions”: ψn (0) = ψn (a) = 0.
This implies that B = 0 and A sin(kn a) = 0 which leads to
kn a = nπ, n = 1, 2, . . . . Note: n = 0 not useful since then ψ ∗ ψ = 0 so no particle in the box.
Thus we find:

n2 π 2 ℏ2
En = for n = 1, 2, 3, . . .
2ma2
These are the energies of the eigenstates
q for a particle in a box.
2mEn
Substituting this form for kn = ℏ2
in the general solution we get for the eigenstates:
 nπx 
ψn (x) = A sin
a
Note: minimum energy for particle has E > 0 (mmm... interesting = zero-point energy).

R ∞ of ∗A? → can use Normalization of probability:

What about the value
This requires that −∞ ψ ψdx = 1. Thus
Z a  nπx  a
1= |A|2 sin2 dx = |A|2
0 a 2
since
Z nπ
nπ
sin2 (u)du =
0 2

2
 q q
∴ A = a2 ; “up to an overall phase” ⇒ i.e., A = a2 eiα is also OK since A∗ A = a2 ,
but this overall phase α can’t make any difference to what we measure (i.e. the probability
density) so for “convenience” we choose α = 0. (Note: Relative phases in a superpostion are
VERY important - see later).
Thus
q
2 nπx


ψn (x) = a
sin a
, n = 1, 2, 3, . . .

These are the eigenstates (aka stationary states or energy eigenfunctions) for a particle in box
Some pictures of these states:

We can define average values (aka “expectation values”) for observables like average position
< x > and average momentum < px > via e.g.
Z ∞
< px >= ψ ∗ pˆx ψdx
−∞

then for these stationary states we have:

a
⟨x⟩ = 2
for all n. Is this obvious?
Ra
⟨p⟩ = 0 for all n since 0 sin nπx nπx



a
cos a
dx = 0

3
Key Concepts
1. Overview of Schrödinger Eq Solns when Ĥ ̸= f (t)
2. Quantum Simple Harmonic Oscillator (QSHO)

Last time we solved Infinite Square Well

Note:
1. For eigenstates of infinte well what about n >>> 1? In general ψn∗ (x)ψn (x) will have n
maxima and probability of finding the particle is ≃ uniform inside the box, ≃ classical
E.g. n = 50:
22
50
2222222
2 50

This is example of Bohr’s Correspondence Principle = Bohr’s C.P.

Quantum systems behave like classical systems when n >> 1 and ℏ is unimportant or absent
Consider 10 gram mass in 7 cm box with v ∼ 10 cm/s
1 n2 π 2 ℏ2
E = mv 2 ≃ 5 · 10−5 joules ≃ ⇒ n ≃ 1029
2 2ma2
....
2. The energy spectrum for the infinite square well: 16 E_1

Now if n >>> 1, the energies should be nearly

E 9 E_1
continuous (classical) according to Bohr’s
4 E_1
Correspondence Principle.
E_1
But apparently the level spacing is increasing Increasing energy level
spacing
as n increases:

⇒ ∆En = En+1 − En = [(n + 1)2 − n2 ]E1 = (2n + 1)E1

∆En 2 1


What’s up?... Actually fractional energy spacing: En
= n
+ n2
⇒ 0 if n >> 1
Thus the level spacing is a tiny tiny fraction of the energy value - way below the
ability to resolve discrete lines experimentally.

4
 3. ψn are discrete, infinite set of functions → can be used to represent any function that
satisfies f (0) = f (a) = 0 ⇒ Fourier’s Theorem.

4. Superposition state is not a stationary state:

Let’s look at time evolution [Ψ∗ (x, t)Ψ(x, t)] for a state that is a superposition of the first
three eigenstates of the infinite square well. In particular:

−itE1 −itE2 −itE3

Ψ(x, t) = c1 ψ1 e ℏ + c2 ψ2 e ℏ + c3 ψ3 e ℏ

with c1 = 0.648, c2 = 0.648, c3 = 0.40. See video demo on webpage link.

Where does this weird behavior come from?? → consider simple superposition of ψ1 & ψ2 :
with x1 = πx
a
, x2 = 2πx
a
and Ψ(x, t) = sin(x1 )e−iE1 t/ℏ + sin(x2 )e−iE2 t/ℏ .
Then probability distribution of finding particle at x (prob. density) is given by:

Ψ∗ Ψ = sin2 (x1 ) + sin2 (x2 ) + sin(x1 )sin(x2 )e−i(E2 −E1 )t/ℏ + sin(x1 )sin(x2 )ei(E2 −E1 )t/ℏ

= sin2 (x1 ) + sin2 (x2 ) + 2sin(x1 )sin(x2 )cos[(E2 − E1 )t/ℏ]

and we get an extra time-dependent term [since 2cos(u) = eiu + e−iu ]. Of course this
time-dependent term vanishes when we integrate over x because sin(x1 ) and sin(x2 ) are
orthogonal.

Summary of stationary state solns. for Ĥ

For Ĥ ̸= H(t), eigenstates of Ĥ are stationary states and can form a complete set of
orthonormal functions. ”Normal” means normalized, ”ortho” means orthogonal (see below).

Now recalling that an arbitrary solution of Schrödinger’s Eq can be written:

∞
X iEn t
Ψ(x, t) = cn ψn (x)e− ℏ

1
Z ∞
we can show that Ψ∗ (x, t)Ψ(x, t)dx = 1:
−∞
This is a product of sums, each with ∞ # of terms · · ·
Picking some specific terms to calculate:

Z ∞ Z ∞
∗
 ∗ ∗ iE1 t/ℏ
c1 ψ1 e−iE1 t/ℏ + · · · + c∗1 ψ1∗ eiE1 t/ℏ c2 ψ2 e−iE2 t/ℏ + · · · dx

Ψ (x, t)Ψ(x, t)dx = c1 ψ 1 e
−∞ −∞

Z ∞ Z ∞ Z ∞
= 2
|c1 | ψ1∗ ψ1 dx + 2
|c2 | ψ2∗ ψ2 dx + ··· + c∗1 c2 ψ1∗ ψ2 ei(E1 −E2 )t/ℏ dx + · · ·
−∞ −∞ −∞
Z ∞ Z ∞
But normalization gives ψ1∗ ψ1 dx = 1 while orthogonality gives ψ1∗ ψ2 dx = 0
−∞ −∞

5
 Thus all of the cross terms vanish leaving us with:
Z ∞ ∞
X
Ψ∗ (x, t)Ψ(x, t)dx = |cn |2 = 1
−∞ n

Overall we have developed a general path to solving Quantum Problems:

1. Solve Ĥψn = En ψn (i.e. find the eigenstates ψn and associated eigenenergies En of Ĥ)

2. Given arbitrary initial state Ψ(x, 0), express this in terms of a superposition of eigenstates
P
ψn , e.g. Ψ(x, 0) = an ψn (x) → we’ll see how this is always possible next week
−itĤ
cn ψn (x)e−iEn t/ℏ
P
3. Use ψ(x, t) = e ℏ Ψ(x, 0) to evolve wavefunction in time. Thus: Ψ(x, t) =

1D Quantum Simple Harmonic Oscillator (QSHO):

V(x)
ˆ Any V (x) with a minimum looks like
a SHO, at least near the minimum
V(x) = 1/2 kx 2
ˆ It’s a useful 1st guess for bound systems
(e.g., p, n in nucleus, quarks in p, n)

ˆ It almost looks like square well except that

x
V → ∞ only when x → ±∞
∴ particle in SHO can have finite probability all the way to x → ±∞

For Quantum Mechanics Solution ⇒ find Energy eigenvalues and eigenfunctions!

Given the Hamiltonian:

p̂2x 1
ĤSHO = + kx̂2
2m 2
then the Eigenvalue Equation is:

−ℏ2 d2 ψn (x) k 2
ĤSHO ψn (x) = + x ψn (x) = En ψn (x)
2m dx2 2
The En are discrete
q because the solutions must be bounded (since V → ∞ as x → ±∞).
k
and ξ = mω x, then (dx)2 = mω (dξ)2 and the Differential
p 0 ℏ
Now let ω0 = m ℏ 0
Equation becomes:

−ℏ2  mω0  d2 ψn (ξ) mωo2

 
ℏ
2
+ ξ 2 ψn (ξ) = En ψn (ξ)
2m ℏ dξ 2 mω0
simplifying:

6
 d2 ψn (ξ)
−ℏω0 + ℏω0 ξ 2 ψn (ξ) = 2En ψn (ξ)
dξ 2
rewriting gives our final, simple, Diff. Eq:

d2 ψn
 
2 2En
= ξ − ψn
dξ 2 ℏω0
⇒ Likewise we can “guess” an approximate solution if x → 0 (e.g. ξ → 0)
d2 ψn
which simplifies the Diff Eq to: = −K 2 ψn (ξ)
dξ 2
which has the soln., for K = constant: ψn = sin(Kξ) or cos(Kξ).
⇒ Likewise we can “guess” an approximate solution if x → ∞
d2 ψn
which simplifies the Diff Eq to: = ξ 2 ψn (ξ)
dξ 2
2
which has the soln.: ψn = e−ξ /2 .
How do we know this works? ⇒ Check it! · · ·

 
d d  −ξ2 /2  d  2
 d  2
 2 2
e = −(2ξ/2)e−ξ /2 = −ξe−ξ /2 = −ξ(−2ξ/2)e−ξ /2 − e−ξ /2
dx dx dx dx
2 2
= ξ 2 − 1 e−ξ /2 ≃ ξ 2 e−ξ /2 = ξ 2 ψn since ξ → ∞ Q.E.D.

Thus we can guess that the stationary states (energy eigenstates) are
standing waves (e.g. sin/cos) near the origin that vanish as |x| → ∞:
V(x)

V(x) = 1/2 kx 2

φ(x) 0 as x x

Now, from the text which does all of the Math, we can write the exact solns
for both the eigenstates and energy eigenvalues:
 mω 1/4 1 −ξ2
0
ψn (ξ) = √ Hn (ξ)e 2 , with n = 0, 1, 2, ...
πℏ 2n n!

7
where Hn (ξ) are the so-called Hermite Polynomials (see text) with e.g.
H0 = 1, H1 = 2ξ, H2 = 4ξ 2 − 2, ...
,→ They are a complete set of orthogonal functions well-known to mathematicians.

And finally the energy eigenvalues are

En = (n + 12 )ℏω0 , n = 0, 1, 2, . . .

....
E
and the spectrum of energies is uniform 5 hω0 /2
(e.g. energy spacing is equal) - see Figure ⇒ 3 hω0 /2
with ∆En ≡ En+1 − En = ℏω0 !
hω0 /2

Equally spaced energy levels