The Central Force Problem: Hydrogen Atom

B. Ramachandran

1 Separation of Variables

The Schrödinger equation for an atomic system with Z protons in the nucleus and one electron “outside”
is

−h̄2 ∇2 − Ze2 1
ψ = Eψ, (1)
2µ (4πε0 ) r
where ε0 is the permittivity of vacuum, which is a constant having the value 8.854187 × 10−12 J−1C2m−1
(4πε0 = 1.112650 × 10−10 J−1C2m−1 ), and µ is the reduced mass of the nucleus-electron system. The
quantity ∇2 appearing above is the Laplacian which, in Cartesian coordinates, is given by

∇2 = ∇ · ∇ = ∂x
∂ ∂ ∂ 2 2 2
2 + ∂y2 + ∂z 2 .

Transforming the Laplacian to spherical polar coordinates, we get


−h̄2 ∇2 = −h̄2 1 ∂ 2∂
r +
1 ∂
sin θ
∂
+
1 ∂2
. (2)
2µ 2µ r2 ∂r ∂r sin θ ∂θ ∂θ sin2 θ ∂φ2
We recognize that

∂ ∂ ∂2
L 2 = −h̄2
1 1
sin θ + .
sin θ ∂θ ∂θ sin2 θ ∂φ2
With this substitution, and making the well-justified assumption that µ  me , the mass of the electron,
Eq. (1) becomes
 
−h̄2  1 ∂ r2 ∂  + L 2 − Ze2 1 − E ψ(r, θ, φ) = 0. (3)
2me r2 ∂r ∂r 2me r2 4πε0 r

The presence of the L 2 operator suggests that the angular part of the solution, i.e., the part that depends
on the angles, are the spherical harmonics Ylm (θ, φ). Therefore, we choose the following function to
attempt a separation of variables:
ψ(r, θ, φ) = R(r)Ylm (θ, φ). (4)
Substituting Eq. (4) into Eq. (3) and simplifying, we get

 
∂ 2 R + 2 ∂R + 2me E + Ze2 1 − l(l + 1) R = 0. (5)
∂r2 r ∂r h̄2 4πε0 r r2
The notation can be further simplified by defining the following dimensionless parameters:

α2 = −2m2e E , β = Ze m2e .
2

h̄ 4πε0 h̄ α

1
Note that α2 will be positive only for negative values of E . This means that the parameter α will
become complex for positive values of E , thereby restricting real solutions to those with negative energy.
Since zero potential energy corresponds to infinite separation of the charges (Ze) and (−e), the negative
energies correspond to the bound states of the atom. Using these definitions, Eq. (5) can be simplified to

∂ 2 R + 2 ∂R − α2 − 2αβ + l(l + 1) R = 0. (6)

∂r 2 r ∂r r 2 r

2 Solving the Radial Equation

We could now attempt a series solution of Eq. (6) as was done in the case of the harmonic oscillator.
However, this would lead to a three-term recursion relationship which is much harder to deal with than a
two-term one. Therefore, we define a new variable ρ = 2αr and express Eq. (6) in terms of this variable
as 
∂ 2T 2 ∂T 1 β l(l + 1)
∂ρ 2 +
ρ ∂ρ
− 4
− ρ
+
ρ2
T = 0, (7)

where T (ρ) is simply R(r) expressed in terms of the new variable.

Let us examine the asymptotic behavior of the solution. As ρ → ∞, Eq. (7) reduces to

∂2T
∂ρ2
− 14 T = 0,
which has the solution T (ρ) = exp(±ρ/2). The solution corresponding to the positive exponent grows
exponentially as ρ increases. This makes it impossible for the solution to be square-integrable. Therefore,
we choose the solution corresponding to the negative exponent. Let us now assume that the general
solution will be of the form
T (ρ) = G(ρ) exp(−ρ/2). (8)
Substituting Eq. (8) into Eq. (7) and simplifying yields

ρ2G (ρ) + (2ρ − ρ2)G (ρ) + {(β − 1)ρ − l(l + 1)} G(ρ) = 0.

(9)

This equation must be valid for all values of ρ. Therefore, at ρ = 0, we have

l(l + 1)G(ρ) = 0; l = 0, 1, 2, ...
Since this equation has to be satisfied for nonzero values of l, this suggests that G(ρ) must have the form
ρs L(ρ), with s > 0. We now expand L(ρ) in a power series as

L(ρ) =

a ρ .
∞

j
j
(10)
j =0

Substituting this expression for G(ρ) into Eq. (9) and collecting terms with the same power of ρ, we get
the following recursion relationship for the coefficients aj :

aj +1 =
l+j +1−β . (11)
aj (j + 1)(j + 2l + 2)

2
 Now, we must verify whether the series will converge as j → ∞. From Eq. (11), it is clear that

aj +1 1
Lim
j →∞ a
=
j
.
j

To understand what such a series might resemble, consider a Taylor expansion of the function eρ :

eρ =

∞
∞
ρj
= bj ρj ,
j =0
j !
j =0
bj +1 j!
Lim
j →∞ bj
=
(j + 1)!
=
1
j +1
 j1 for large j.

3 Truncation of the Series: Quantized Energies

This suggests that L(ρ), and therefore G(ρ), behaves like eρ asymptotically. This, of course, leads to
a non-square-integrable solution, and is also in contradiction to the conclusion from Eq. (8) that T (ρ)
behaves like e−ρ/2 as ρ → ∞. The way out of this difficulty, of course, is to truncate the series at some
value of j (as in the case of the harmonic oscillator), say j = k, so that

ak+1 =
l + k + 1 − β = 0,
ak (k + 1)(k + 2l + 2)
i.e., β = l + k + 1.

Since l + k + 1 must be an integer, having possible values 1, 2, 3, ...∞, we define n = l + k + 1 so

β = n; n = 1, 2, ...∞. Also, since the truncation relationship must be valid for all values of k including
k = 0, we see that n ≥ l + 1 or l ≤ (n − 1). Thus, the series truncation also imposes a limit on the values
l can assume. Now, substituting this into the definition of β , we get

n =
Ze2 me , or
4πε0h̄2α

2 2 2
n2 = − 4Zeπε mh̄e2 2mh̄ E . (12)
0 e

Rearranging, and recognizing that the energies depend on the integer n, we get

En = −Z 2e 2m2e n12 ; n = 1, 2, ...∞.

2 4
(13)
32π ε0 h̄

a0 = 4πε0h̄2
2
We now define a parameter , so that
me e


En =−
1 e2 Z2 . (14)
2 4πε0a0 n2

3
4 Atomic Units

a .
The parameter 0 has units of length (verify this!) and is equal to 0 529177 × 10−10 m. This parameter
has a special place in quantum chemistry because it is the unit of length (called the Bohr) in atomic units
(au), which are a set of units commonly used in quantum chemistry calculations. In these units, the unit
m
of mass is the mass of the electron (i.e., e = 1 au), and h̄ = 1 so that for the case of the Hydrogen
atom (Z = 1), the ground state energy is E1 = −1/2. The unit of energy is the quantity e2 /(4πε0 a0)
called the Hartree and so, we see that the ground state of the hydrogen atom is half of a Hartree below
the ionization limit (i.e., E = 0). From this information, we calculate that

1 Hartree = 27.2115 eV = 4.35979 × 10−18 J,

where “eV” stands for the electron volt, a commonly used unit of energy in quantum chemistry, defined
as 1 eV = e(C) × 1.0 Volt = 1.60219 × 10−19 J.

5 The Rydberg Constant

The energy expression of Eqs. (13) or (14) can be used to calculate the value of the Rydberg constant
appearing in Bohr’s original expression for the frequencies of the lines appearing in the emission spectrum
of the hydrogen atom. Bohr’s expression was
1
1 1
λ
=R
n21 − n22 ; n2 > n1 .
Recognizing that ∆E = hν = hc/λ, we get from Eq. (14),

∆E =
h̄2  Z 2  1 − 1  ; n > n ,
2me a0 n21 n22 2 1
1
=
h  Z 2  1 − 1  .
λ 4π 2cme a0 n21 n22

The value of the Rydberg constant R thus calculated is 109,677 cm−1. Modern experiments have
established the correctness of this value to about 8 significant figures.

6 Atomic Orbitals

It is obvious from Eq. (5) that the solutions of the radial equation depend on the value of l. Now, the
series truncation condition above also brings in a quantum number n. The radial solutions are, therefore,
written as Rnl (r), and including normalization, are given by

Rnl (r) = Nnl Gnl (ρ) e−ρ/2,

 3 
2Z (n − l − 1)!
Nnl = − na0 × ,
2n [(n + l)!]3
G (ρ)
nl = ρ L2 +1
l +1
n
l
(ρ),

4
where the polynomials denoted by Lsr (ρ) are called the associated Laguerre polynomials of degree (r − s),
and are defined as
ds L (ρ),
Lsr (ρ) = dρ s r

where the Lr (ρ) are the Laguerre polynomials given by

Lr (ρ) = eρ d r ρr e−ρ .
r

dρ
Substituting for α in the defintion of ρ, we get ρ = 2Zr/(na0 ). We now write the solutions to the
hydrogen atom problem as

ψnlm (r, θ, φ) = Rnl (r)Ylm (θ, φ), (15)

n = 1, 2, ....∞,
l = 0, 1, ..., (n − 1),
m = 0, ±1, ±2, ..., ±l.

The convention followed in labeling atomic states is to use a letter symbol to denote various values of l,
so that we use
Value of l : 0 1 2 3 4 5
...
Letter symbol: s p d f g h

Therefore, the state |nlm = |210 is referred to as the 2p0 state and so on. Each |nlm state defines
what is known as an atomic orbital.

Let us now examine the nature of the wave functions. We define two types of probability functions
for this purpose. The quantity |ψnlm (r, θ, φ| r2dr sin θdθdφ gives the probability of finding the electron
in a volume element enclosed by the limits (r, r + dr; θ, θ + dθ; φ, φ + dφ). However, when dealing
with a “spherical” species like an atom, it is more useful to average over the angles and get the radial
distribution function, Snl (r), which is obtained as
π π
2
Snl (r) = r |Rnl (r)| dr
2 2
sin θdθ dφ |Ylm (θ, φ)|
0 0
= r2 |Rnl (r)|2 dr.

The final result comes about because the integrals over the angles yield unity because the spherical
harmonics are normalized:
π 2π
sin θdθ dφ |Ylm (θ, φ)| = 1. (16)
0 0

The physical significance of this is that r2 |Rnl (r)|2 dr gives the probability for finding the electron on
a spherical shell of radius r and thickness dr. We will examine the shapes of the radial functions and
the two types of probability distribution functions using class handouts. We will also combine the radial
functions with the spherical harmonics (i.e., generate atomic orbitals) and examine the nodal structure of
these three-dimensional functions.

Below, we tabulate a few solutions of the Schrödinger equation Eq. (1) . For convenience, we define

5
σ = nρ/2 = Zr/a0 .
n l m ψ (r)
 Znlm3/2
Common name
1 0 0 √1π
a e−σ 1s
 Z 3/2 0

2 0 0 √1
4 2π a0
(2 − σ)e−σ/2 2s
 Z 3/2
2 1 0 √1
4 2π a0
σe−σ/2 cos θ 2 zp
 Z 3/2
2 1 ±1 √1
4 2π a0
σe−σ/2 sin θ cos φ 2 xp
 Z 3/2
√ 1
4 2π a0
σe−σ/2 sin θ sin φ 2 yp
 Z 3/2  
3 0 0 √1 27 − 18σ + 2σ2 e−σ/3 3s
81 3π a0
√2  Z 3/2
3 1 0 √
81 π a σ(6 − σ )e−σ/3 cos θ 3pz
√  Z 3/2
0

3 1 ±1 √2
81 π a σ (6 − σ ) e−σ/3 sin θ cos φ 3 xp
√2  Z 3/2
0
√
81 π a
σ (6 − σ)e−σ/3 sin θ sin φ 3py
 Z 3/2
0

3 2 0 √
1
81 6π a0
σ 2e−σ/3 (3 cos2 θ − 1) 3dz 2
√2  Z 3/2 2 −σ/3
3 2 ±1 √
81 π a σe sin θ cos θ cos φ 3 xzd

√2 Z 3/2
0
√
81 π a
σ2 e−σ/3 sin θ cos θ sin φ 3 yzd
 Z 3/2
0

3 2 ±2 1√
81 π a σ 2e−σ/3 sin2 θ cos 2φ d
3 x2 −y2
 Z0 3/2
1√
81 π a σ2 e−σ/3 sin2 θ sin 2φ 3dxy
0

7 Most Probable and Average Values of r

The most probable and average distance of the electron
chemistry and physics. The most probable distance of
position of the highest maximum in Snl (r). Recall that
maximum or minimum. For example, in the case of the
and get
d 2 −2σ
dr
σ e = (2σ
from the nucleus is of considerable interest in
the electron from the nucleus is given by the
the derivative of a function goes to zero at its
1s orbital, we set the derivative of S10(r) = 0
− 2)e−2σ = 0, (17)

where we have divided both sides by the multiplicative constants present in S10(r) to eliminate them.
Since the exponential term in the derivative is non-negative for all real values of σ, we conclude that the
most probable value of r corresponds to σ = 1, or rmp = a0 /Z . For the hydrogen atom, this is exactly
equal to the Bohr radius, which is the radius of the ground state orbit in Bohr’s planetary model of the

6
hydrogen atom.The average value of r is given by the expectation value
∞ π 2π ∗ ∗ (r, θ, φ)r2 sin θdrdθdφ
rnl = ψnlm (r, θ, φ)rψnlm
0 0 0
∞ π 2π
=
∗ 2
Rnl (r)rRnl (r)r dr m 2
Yl (θ, φ)|
| sin θdθdφ
0∞ 0 0
= r3 |R (r)|2 dr
nl
0

=
n a0 1 + 1 1
2
− l(ln+2 1) .


Z 2

The final result is obtained by a somewhat lengthy and clever use of the method of integration by parts.1

8 Hydrogenic atoms in a magnetic field

Note that the energy of a one-electron atom, given in Eq. (14), is independent of quantum numbers l
and m. This means that states with the same value of n but different values of l and m have the same
energy. Let us examine what happens to these states when a magnetic field is applied along the Z -axis.

Consider a circular loop of conducting wire of radius r through which a charge Q is moving with
velocity v. Since current is the charge flow per unit time, we get

I = 2Qv
πr .
The value of the magnetic dipole |µ| associated with this current is given by IA, where A is the area
enclosed by the loop, i.e., A = πr2 . Therefore,

Qv
| µ
| =
2πr
× πr2 = 12 Qvr = Qrp
2m
.

Therefore, the magnetic dipole vector is given by

µ = 2Qm r × p = 2Qm L.
For an electron,
µ = 2−me L, (18)
e

eh̄  
and
µ
| | =− l (l + 1) = µB l(l + 1), (19)
2me
where the quantity µB is a constant called the Bohr magneton.

Now let us consider a magnetic field B

applied to the atom. The potential energy of interaction
B
between the field and the magnetic dipole µ is given by

VB = −µ · B (20)
1
Yep. You guessed it. Homework assignment!

7
From Eq. (18), we get VB = e/(2me )L · B. If we assume that the field is oriented along the Z -direction,
i.e., the components of the field along the x and y directions are zero, then,

VB = 2m e iL + jL + k L  · k B
x y z
e
e
= BLz .
2me
In operator terminology, therefore,
B = eB 
V Lz . (21)
2me
Therefore, the Schrödinger equation for a hydrogen atom in a magnetic field is
 
 + VB Rnl (r)Y m (θ, φ) = (En + µB Bm)Rnl (r)Y m (θ, φ),
H l l (22)

where the Hamiltonian operator H  corresponds to the Schrodinger equation we started with in Eq. (1),
En is given by Eq. (14), and µB is the Bohr magneton defined in Eq. (19). Therefore, in a magnetic
field, states with nonzero values of m will aquire different energies.