QM1 - Tutorial 1

The Bohr Atom and Mathematical Introduction

Amir Kahana and Yaakov Yudkin

7 March 2018

Contents
1 Bohr Atom 1
1.1 Energy of a Photon - The Photo Electric Eect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 The Discrete Energy Spectrum of the Hydrogen Atom . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Light Emitted by an Atom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Fourier Transform 4
2.1 Orthogonality of Exponential Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Fourier Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Dirac Delta Function 8

3.1 Denition, Representations and Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2 Fourier Transform of the Dirac Delta Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

1 Bohr Atom
1.1 Energy of a Photon - The Photo Electric Eect
Experimental Observation When light hits a metal it can free electrons from it. This is in itself surprising, but
even more so is the fact that the eect depends on the frequency (or wavelength) of the light - not on the intensity.

Physical Interpretation (Einstein 1905) Light is not continuous. It consists of particles (lets call them photons)
that carry energy and momentum. The energy Eph of a (microscopic) photon depends on the frequency ν of the
(macroscopic) light wave.

Eph = hν

Consider a single photon with energy Eph hitting a metal. The photon will be absorbed by the metal (i.e. it disappears)
while its energy is transferred to an electron in the metal. The equation for conservation of energy is

Eelec = Eph − φ

where φ is the work function

1 of the metal. It is the minimal amount of energy needed to pull an electron out to the
metal. This means that the energy of the photon must be larger than the work function for the photo electric eect to
happen.
Eph ≥ φ
The additional energy that the photon carried becomes kinetic energy of the electron.

1 It is called a function although it does not depend on the parameters of the problem. It depends on the type of metal.

1
1.2 The Discrete Energy Spectrum of the Hydrogen Atom
Incompatibility of Classical Mechanics According to classical physics the atom consists of a heavy positive
nucleus with negative electrons circling it as in a typical Kepler set up. The problem is that such a circular motion of
a charged particle causes electromagnetic radiation. Electromagnetic radiation carries energy away form the electron
causing it to decay into the nucleus. This means that classical matter is unstable.
But matter is stable! Niels Bohr invented a new model for the atom that does not have this problem. His model is
still not correct but it is a good start and shows the idea of quantum mechanics.

Figure 1: Left: The classical atom must decay due to radiation. Right: Bohr assumed discrete allowed radii. The
electron can jump from one radius to another but while it is in one it does not radiate and is thus stable.

The Hydrogen Atom We will derive this theory for the Hydrogen atom. Its nucleus consists of one proton (charge
+e, mass M) and there is one electron (charge −e, mass m) circling it. Our goal is to show that the energy of the
Hydrogen atom can only take distinct discrete values. Having discrete energy values means that a continuous decay
(as in the classical atom) is not possible.

Bohr's Quantization Now we start quantum mechanics. We quantize. What does this mean? It means that a
certain quantity can only take on multiples of some number. Here this quantity is angular momentum. We postulate
that the value of the angular momentum must be a multiple of the number ~2 . Why? No reason. It is a postulate.
We know it works because the consequence of the postulate (discrete energy levels) agrees with experiments.
The postulate states that
L = n~ , n∈N

Discrete Radii The angular momentum can be written as

L = |~r × p~| = mrv

To nd the velocity v we assume classical trajectories: an electron circling a proton with radius r. The velocity v is
then found by setting the Coulomb attraction equal to the centripetal acceleration a = v 2 /r.

v2 e2 1 2 e2 1
m = → v =
r 4π0 r2 4π0 m r
2 The units of ~ are the same as the units angular momentum. Maybe this makes the postulate a little easier to understand. In SI units
the value is
kg · m2
~ = 1.05 · 10−34
s

2
The angular momentum squared can hence be written as

e2 1 e2
L2 = m2 r2 v 2 = m2 r2 = mr
4π0 m r 4π0
2
Setting this equal to L2 = (n~) we obtain

e2 4π0 1 2
mr = n2 ~2 → r = ~2 n
4π0 e2 m
We dene the Bohr radius
4π0 1
aB = ~2 = 0.529Å
e2 m
with which we can write
rn = aB n2
We have found that only specic radii are allowed. We say: the radius is quantized. The little n used in the subscript
of rn is used to show that the radius r is discrete.

r1 = aB , r2 = 4aB , r3 = 9aB , ...

Discrete Energy Spectrum Also here we start of classically and nd an expression for the energy .
3

1 e2 1
E =T +V = mv 2 −
2 4π0 r
2
1 e 1 e2 1
= −
2 4π0 r 4π0 r
2
e 1
=−
4π0 2r
Plugging in the quantized radii rn we get the quantized energies En .

e2 1 1 13.6eV
En = − =−
4π0 2a0 n2 n2

The quantized energies En are called the energy levels of the states of the Hydrogen atom.
We have shown that the electron cannot decay into the nucleus. The minimal radius is r1 = aB (not r = 0). The
energy of the atom in this minimal state is E1 = −13.6eV . We call this the ground state of the Hydrogen atom.

1.3 Light Emitted by an Atom

Statement of the Problem Consider a Hydrogen atom with energy E5 . How many possible wavelengths can be
emitted by the atom? Find two of them explicitly.

Physical Principle A photon is emitted from an atom when it decays from a high energy level to a low energy level.
The energy of the emitted light (photon) is equal to the energy dierence of the two atomic states.

Eph = Einit − Ef inal

From the photoelectric eect we know that the energy of a photon is Eph = hν and by using the relation λν = c we
get
c hc
λ= =
ν Eph
We conclude that when the atom decays from level n to level m a photon with wave length λm→n is emitted, where

hc
λm→n =
Em − En
3 It is not surprising that we nd |T | = |V |/2. This is just the virial theorem.

3
Figure 2: Emission of photons as the atom decays from level n=5 to the ground state n = 1. This can happen by
emitting one photon with wave length λ5−1 , by emitting two photons or even more.

Dierent Decay Paths One way the atom can decay is directly from level n=5 to the ground state n = 1. Using
the formula derived above we nd that its wave length is

hc
λ5→1 = = 9.5 · 10−8 m = 95nm
E5 − E1
The decay from n=5 can also happen in two stages, e.g. through level n = 3. The atom will start of in n = 5 and
decay to n=3 while emitting a photon with λ5→3 . It will stay there for a certain time (we will not be concerned with
how long) and then decay from n=3 to n = 1 while emitting a photon with λ3→1 . The formula gives
hc hc
λ5→3 = = 1282nm , λ3→1 = = 103nm
E5 − E3 E3 − E1
Notice that λ5→1 6= λ5→3 + λ3→1 . Do you know why? Think about it.

Number of Dierent Photons Two photons are dierent from one another if they have dierent energy. This
means that we are looking for all the dierent wave lengths that can be emitted by the atom.
When the atom is in n=5 it can decay in four possible ways. It can go to n = 4, 3, 2, 1. If the atom managed
to reach the state n=4 it can decay from there in three possible ways and so on. In total, the number of dierent
photons that can be emitted is
4 + 3 + 2 + 1 = 10

Spectral Lines If a gas of hydrogen atoms (so a large number of atoms) is excited to the level n=5 all the atoms
will decay back done to the ground state n=1 within a short time. As this happens 10 dierent types of photons will
come ying out which we can observe as light in dierent colors. These colors are discrete and are thus called spectral
lines. (Search hydrogen spectrum on google and click on images.)

2 Fourier Transform
2.1 Orthogonality of Exponential Functions
Statement of the Problem Consider the function
4
φn (x) dened in the interval [−L/2, L/2]
 
1 2πn
φn (x) = √ exp i x , n∈N
L L
Is fn (x) orthogonal to fm (x) for n 6= m? What happens if n = m?
4 This is actually a set of functions {φn (x)}, one for each n.

4
 Figure 3: All spectral lines for a Hydrogen atom excited to the n=5 energy level.

Denition of Orthogonality The two vectors ~a and ~b are orthogonal if their inner product vanishes.

 
b1
~a? · ~b = a∗1 a∗2 a∗3 b2  = 0


b3

This is true for any dimension. In the same way the two functions f (x) and g(x) are orthogonal if

Z ∞
f ∗ (x)g(x)dx = 0
−∞

In other words a function is a vector with an innite amount of dimensions. It is customary to write the inner product
as Z
hf |gi = f ∗ (x)g(x)dx

The orthogonality condition is thus

hf |gi = 0

Solution We plug the two function φn (x) and φm (x) into the integral hφn |φm i. The borders of integration are
−L/2 → L/2 because this is the interval in which the functions are dened.

Z
hφn |φm i = φ∗n (x)φm (x)dx

1 L/2
Z    
2πn 2πm
= exp −i x exp i x dx
L −L/2 L L
1 L/2
 
2π(m − n)
Z
= exp i x dx
L −L/2 L

This integral is zero unless n = m in which case it is equal to 1. To see this we can use Euler, namely exp(iθ) =
cos(θ) + i sin(θ)
L/2
    
2π(m − n) 2π(m − n)
Z
1
hφn |φm i = cos x + i sin x dx
L −L/2 L L
The integral of a sinusoidal function over a full cycle vanishes. Hence, we get 0 if n 6= m.

hφn |φm i = 0 , n 6= m

5
If n=m the cosine is 1 and the sine is 0. We thus have for n=m
Z L/2
1
hφn |φm i = dx = 1 , n=m
L −L/2

In summary we have
(
0 n 6= m
hφn |φm i = = δnm
1 n=m

2.2 Fourier Series

Orthogonal Vectors are Basis Vectors A set of orthogonal vectors is very useful since they span a basis. Take
for instance the Cartesian base vectors x̂, ŷ and ẑ . They span the three dimensional space. I.e. any 3D vector ~a can
be written as a linear combination of these three vectors.

~a = a1 x̂ + a2 ŷ + a3 ẑ

The same is true for a set of orthogonal functions. As an example lets consider the set of functions
(r   r  )
2 2πn 2 2πn
sin x , cos x , n∈N
L L L L

which are dened in the interval [−L/2, L/2]. These are the base functions of all periodic functions with periodicity
L. Any periodic function f (x), which satises f (x + L) = f (x), can be written as a linear combination of said base
functions. We know this simply as the Fourier series.

∞   X ∞  
X 2πn 2πn
f (x) = An sin x + Bn cos x
n=0
L n=0
L

Complex Fourier Series Instead of using sinusoidal functions we can use exponential functions. We know from
Euler that they are equivalent. For a periodic function we can thus write

∞  
X 2πn
f (x) = Cn exp i x
n=−∞
L

Notice that the exponential is just φn (x) used above. By replacing the coecients Cn by fn we have

∞   ∞
X1 2πn X
f (x) = fn √ exp i x or f (x) = fn φn (x)
n=−∞ L L n=−∞

Expansion Coecients fn of the Complex Fourier Series We wish to nd the coecients fn of the expansion.
We start from the expansion
∞  
X 2πn
f (x) = fn exp i x
n=−∞
L

and use the orthogonality proved above

Z L/2    
1 2πn 2πm
hφn |φm i = exp −i x exp i x dx = δnm
L −L/2 L L
√
in order to nd the coecients fn . We multiply both sides of the expansion by exp (−i2πmx/L) / L and integrate
over x.
L/2 ∞
1 L/2 X
Z   Z    
1 2πm 2πn 2πm
√ f (x) exp −i x dx = fn exp i x exp −i x dx
L −L/2 L L −L/2 n=−∞ L L

On the right hand side we can do the integral before the sum and recognize the δnm .
Z L/2 ∞     ∞
1 X 2πn 2πm X
fn exp i x exp −i x dx = Cn δnm = fm
L −L/2 n=−∞ L L n=−∞

6
And thus
Z L/2   Z L/2
1 2πm
fm =√ f (x) exp −i x dx or fm = f (x)φ?m (x)
L −L/2 L −L/2

Plugging this in we get

∞   Z L/2  
X 2πn 1 2πn 0
f (x) = exp i x f (x0 ) exp −i x dx0
n=−∞
L L −L/2 L
X∞ Z L/2
= φn (x) f (x0 )φ?m (x0 ) dx0
n=−∞ −L/2

2.3 Fourier Transform

Continuous Limit Lets continue our way towards the Fourier transform. The next step is going to continuity. For
this we dene
2πn
kn =
L
and thus
2π 1 ∆k
∆k = kn+1 − kn = → =
L L 2π
We plug both kn and ∆k in the expansion

∞   Z L/2  
X 2πn 1 2πn 0
f (x) = exp i x f (x0 ) exp −i x dx0
n=−∞
L L −L/2 L
∞
∆k L/2
X Z
= exp (ikn x) f (x0 ) exp (−ikn x0 ) dx0
n=−∞
2π −L/2

and take L → ∞. Note that when doing this kn → k

∆k → dk . Also, the sum over n
and becomes an integral over k.
∞
dk ∞
Z Z
f (x) = exp (ikx) f (x0 ) exp (−ikx0 ) dx0
−∞ 2π −∞

We can now read of the expansion

Z ∞
dk
f (x) = √ eikx f˜(k)
−∞ 2π
with the coecients (which are now a function)

Z ∞
dx
f˜(k) = √ e−ikx f (x)
−∞ 2π

This is the Fourier transform.

Physical Meaning There are two dierent spaces, the x-space and the k -space. Any function can be represented
in either one of those spaces. The two representations are equivalent because we know exactly how to go from one to
another. The functions we are using for the expansion are plane waves.

1
φk (x) = √ eikx
2π
Given a function in x-space we can expand it in plane waves to get the same function in k -space and vice versa.

Example: Statement of the Problem Find the Fourier transform f˜(k) of a Gaussian

2
f (x) = e−αx

You may use

∞
Z r
−ay 2 π
e dy =
−∞ a

7
Solution Plug in to the formula

Z ∞
dx 2
f˜(k) = √ e−ikx e−αx
−∞ 2π
Z ∞ " 2 #
k2

dx ik
= √ exp −α x + −
−∞ 2π 2α 4α
Z ∞ " 2 #
k2
 
dx ik
= exp − √ exp −α x +
4α −∞ 2π 2α
Switch variables y = x + ik/2α
Z ∞
k2

˜ 1
dy exp −αy 2
 
f (k) = √ exp −
2π 4α −∞
r
k2
 
1
= exp −
2α 4α

Properties of the Fourier Transform A few properties of the Fourier transform f˜(k) = F [f (x)] include

• Linearity
F [af (x) + bg(x)] = aF [f (x)] + bF [g(x)] = af˜(k) + bg̃(k)
• Shift
F [f (x + x0 )] = F [f (x)] eikx0 = f˜(k)eikx0
• Derivative  
d
F f (x) = ikF [f (x)] = ik f˜(k)
dx
 n 
d n n
F f (x) = (ik) F [f (x)] = (ik) f˜(k)
dxn
• Scaling
 
1˜ k
F [f (ax)] = f
a a

3 Dirac Delta Function

3.1 Denition, Representations and Properties
Denition This is the continuous version of δnm . It is written as δ(x − x0 ). By denition

b
X
an δnm = cm a < m < b
n=a
Z b
dxf (x)δ(x − x0 ) = f (x0 ) a < x0 < b
a
while
Z b
dxδ(x − x0 ) = 1
a
The Dirac delta δ(x − x0 ) is only dened inside the integral. It is meaningless without it.

Representations It can be represented in dierent ways. A few examples are

•
1 
δ(x) = lim+
→0 π x 2 + 2
•
1 2
δ(x) = lim √ e−(x/)
→0 + π
•
1 sin(x/)
δ(x) = lim+
→0 π x

8
Example: Proof of a Representation Prove that the following limit is a valid representation of the Dirac delta
function.
1 2
δ(x) = lim+ √ e−(x/)
→0 π

Solution In order to prove this we need to show three properties:

1.
δ(x = 0) → ∞
2.
δ(x 6= 0) = 0
3. Z ∞
δ(x)dx = 1
−∞

Lets do this one by one.

1.
1
δ(x = 0) = lim √ = ∞
→0+ π
2.
1 2 Ω 2
δ(x 6= 0) = lim √ e−(x/) = lim √ e−(Ωx) = 0
→0 + π Ω→∞ π
3.
∞ ∞
1 √ 2
Z Z
1 2
δ(x)dx = lim+ √ e−(x/) dx = lim+ √ π = 1
−∞ →0 π −∞ →0 π

√
Figure 4: Graph of exp(−(x/)2 )/ π with dierent values of . One can see that as  approaches zero the function
becomes more and more δ -like.

Properties of the Dirac Delta Function A few properties include (can be proven under the integral)

• Symmetric
δ(x) = δ(−x)
• Scaling
1
δ(ax) = δ(x)
|a|
• Delta of a function g(x) with roots xi
X 1
δ(g(x)) = δ(x − xi )
i
|g 0 (x i )|

9
3.2 Fourier Transform of the Dirac Delta Function
Statement of the Problem Apply the Fourier transform formula to the Dirac delta function to nd its transform.
Then inverse transform your result to prove the identity

Z ∞
dk ik(x−x0 )
δ(x − x0 ) = e
−∞ 2π

Solution Z ∞
dx 1
F [δ(x − x0 )] = √ e−ikx δ(x − x0 ) = √ e−ikx0
−∞ 2π 2π
Doing the inverse transform we get

δ(x − x0 ) = F −1 [F [δ(x − x0 )]]

 
−1 1 −ikx0
=F √ e
2π
Z ∞
dk ikx 1 −ikx0
= √ e √ e
−∞ 2π 2π
Z ∞
dk ik(x−x0 )
= e
−∞ 2π

Remember this. It will turn out to be very useful!

10