Mathematical Methods in Chemistry (MATH0042) Page 48

Substituting (2.133) into (2.132) (recalling (2.121)) gives

d2 R 1 dR R d2 Θ
Θ + Θ + 2 2 + λ2 RΘ = 0, (2.134)
dr2 r dr r dθ
which can be rearranged into the form

1 d2 R 1 dR 1 d2 Θ
r2 2
+ r + λ 2 2
r = − 2
= ν 2. (2.135)
R dr R dr Θ dθ
Here, we fix the separation constant (ν 2 ) to be positive in order that Θ(θ) is periodic.
Indeed, the ODE for Θ(θ) is then

Θ′′ (θ) + ν 2 Θ(θ) = 0, (2.136)

which has solution

Θ(θ) = D cos(νθ) + E sin(νθ), (2.137)
for arbitrary constants D and E.
The equation for R(r) is the second order ODE with non-constant coefficients

r2 ′′ r ′
R + R + r 2 λ2 = ν 2 , (2.138)
R R
which can be rearranged into the form

r2 R′′ (r) + rR′ (r) + (λ2 r2 − ν 2 )R(r) = 0. (2.139)

Introducing x = λr and y(x) = R(r), (2.139) can be rewritten as

x2 y ′′ (x) + xy ′ (x) + (x2 − ν 2 )y(x) = 0. (2.140)

(2.140) is known as Bessel’s equation. Bessel’s equation is so important in applied

mathematics that we will devote the next section to its detailed study.

2.4.3 Bessel’s equation

Bessel’s equation is:

x2 y ′′ (x) + x y ′ (x) + (x2 − ν 2 ) y(x) = 0. (2.141)

This is a second order homogeneous ODE with non-constant coefficients. Identifying

its coefficients as P (x) = x2 , Q(x) = x and R(x) = (x2 − ν 2 ), we see that the limit

Q(x) x
lim = lim 2 (2.142)
x→0 P (x) x→0 x

does not exist, but the limits

Q(x) x
lim x = lim x 2 = 1, (2.143)
x→0 P (x) x→0 x
Page 49

R(x) (x2 − ν 2 )
lim x2 = lim x2 = −ν 2 (2.144)
x→0 P (x) x→0 x2
both exist. Hence x = 0 is a regular singular point of (2.141). Then, using Frobe-
nius’s method, we look for a solution to (2.141) of the form
∞
X
ρ
y(x) = x an x n . (2.145)
n=0

for some index ρ and some coefficients an , with a0 ̸= 0. (Note that we use ρ instead
of r for the index here, in order to avoid confusion with the radial polar coordinate
r.) Substituting the series (2.145) for y(x) into (2.141), we get
∞
X ∞
X ∞
X ∞
X
n+ρ n+ρ n+ρ+2 2
an (n + ρ)(n + ρ − 1)x + an (n + ρ)x + an x −ν an xn+ρ = 0.
n=0 n=0 n=0 n=0
(2.146)
Redefining the summation index in the third series on the left-hand side gives
∞
X ∞
X ∞
X ∞
X
n+ρ n+ρ n+ρ 2
an (n + ρ)(n + ρ − 1)x + an (n + ρ)x + an−2 x −ν an xn+ρ = 0.
n=0 n=0 n=2 n=0
(2.147)
n+ρ
Hence, for n ≥ 0, equating the coefficient of x on the left-hand side to zero, we
get:
ˆ For n = 0,
ρ(ρ − 1) + ρ − ν 2 a0 = 0,
 
(2.148)
i.e.,
ρ2 − ν 2 a0 = 0.


(2.149)
But since a0 ̸= 0, we must have

ρ2 − ν 2 = 0. (2.150)

This is the indicial equation and it implies that

ρ = ±ν. (2.151)

ˆ For n = 1,
(1 + ρ)ρ + (1 + ρ) − ν 2 a1 = 0,
 
(2.152)
i.e.,
1 + 2ρ + ρ2 − ν 2 a1 = 0.
 
(2.153)
But since ρ = ±ν, then this last equation reduces to

(1 + 2ρ)a1 = 0, (2.154)

i.e.,
(1 ± 2ν)a1 = 0 when ρ = ±ν, respectively. (2.155)
This implies that a1 = 0 unless ν = 1/2 and ρ = −ν, or, ν = −1/2 and ρ = ν.
 Mathematical Methods in Chemistry (MATH0042) Page 50

ˆ For n ≥ 2,

(n + ρ)(n + ρ − 1) + (n + ρ) − ν 2 an + an−2 = 0,
 
(2.156)

i.e.,
((n + ρ)2 − ν 2 )an + an−2 = 0. (2.157)
But again, since ρ = ±ν, then this last equation reduces to

(n(n + 2ρ))an + an−2 = 0, (2.158)

i.e.,

(n(n ± 2ν))an + an−2 = 0 when ρ = ±ν, respectively. (2.159)

Now, recalling again (2.151), note that (up to a sign) the difference between the
two allowable values of ρ is 2ν. We have three possible cases:

ˆ 2ν ̸∈ Z. Then Frobenius’s method leads to two linearly independent solu-

tionsPfor y(x) of the form (2.145), i.e., yν (x) = xν ∞ + n
P
n=0 n x and y−ν (x) =
a
−ν ∞ − n + −
x n=0 an x , for some coefficients an and an . Note that it follows from
(2.155) that a± + −
1 = 0 (since, in this case 2ν ̸= ±1). For n ≥ 2, an and an are
determined by the corresponding form of (2.159). Note that (n ± 2ν) ̸= 0,
and since a± ±
1 = 0, it follows from (2.159) that an = 0 for all odd n. Note also
that one of y±ν (x) becomes infinite at x = 0, while the other is finite for all x
except as x → ±∞, e.g., if ν > 0, then y−ν (x) will be infinite at x = 0 due to
the term a0 x−ν in its series expansion.

ˆ ν = 0. In this case the solutions of the indicial equation coincide and so

Frobenius’s method only produces one solution for y(x) - see below. This
solution is finite for all x except as x → ±∞. We shall not discuss the second
linearly independent solution for y(x) in this case, other than to mention that
it becomes infinite at x = 0.

ˆ 2ν ∈ Z, ̸= 0. In this case the solutions of the indicial equation differ by

a non-zero integer. It can be shown that if 2ν ∈ Z but ν ∈ / Z, the two
linearly independent solutions to Bessel’s equation are still both of the form
(2.145), and can be found using Frobenius’s method (see Problem sheet 6 for
the example with ν = 21 ). However, this is not the case if ν ∈ Z; if ν ∈ Z,
there exists only one solution for y(x) of the form (2.145). We shall not discuss
the second linearly independent solution for y(x) in this case, other than to
mention that it becomes infinite at x = 0.

A solution of (2.141) that can be written in the form (2.145) is denoted by Jρ (x)
and referred to as a Bessel function of the first kind of order ρ. For example, J0 (x)
is as follows. With ν = 0, it follows from (2.155) that a1 = 0. Furthermore, (2.159)
becomes
1
an = − 2 an−2 , for n ≥ 2. (2.160)
n
Page 51

Hence, an = 0 for all n odd, while for n = 2k, k ≥ 1, (2.160) gives

a2(k−1) a2(k−2) (−1)k

a2k = − = = . . . = a0 . (2.161)
(2k)2 (2k)2 (2(k − 1))2 22k (k!)2

Hence, substituting these forms for an into (2.145), we get

y(x) = a0 J0 (x) (2.162)

where a0 is an arbitrary constant, and we define

∞
X (−1)n 2n
J0 (x) = 2n (n!)2
x . (2.163)
n=0
2

2.4.4 Back to the 2D wave equation for a circular drum

Let us now return to the 2D wave equation for a circular drum. Recall that we seek
u(r, θ) in the form u(r, θ) = Θ(θ)R(r), and we found that

Θ(θ) = D cos(νθ) + E sin(νθ), (2.164)

for arbitrary constants D and E, while, with x = λr and y(x) = R(r), y(x) satisfies
Bessel’s equation (2.140).
Recall also that we require u(r, θ) to be 2π-periodic in θ (see (2.124)). It follows
that Θ(θ) must be 2π-periodic, and hence (by arguments that we have used previ-
ously) that we must have ν ∈ Z, ≥ 0. Henceforth, we write ν = n, where n ∈ Z, ≥ 0.
Now consider (2.140). With ν = n ∈ Z, ≥ 0, it follows from what we described in
section 2.4.3 that one solution for y(x) is Jn (x), and hence one solution for R(r) is

R(r) = Jn (λr). (2.165)

This is finite for all finite r. A second linearly independent solution for R(r) exists,
but is infinite at r = 0; we thus dismiss this solution on physical grounds. It remains
only to satisfy the condition u(a, θ) = 0 (see (2.123)). In order to satisfy this
condition we must have R(a) = 0. Thus, with R(r) given by (2.165), we must have
Jn (λa) = 0. It follows that we must have λ = jn,m /a, where jn,m , m = 1, 2, 3, . . .
are the positive zeros of the function Jn (x) (λ = ω/c > 0). It is known that Jn (x)
has infinitely many positive zeros.
Thus, with λ = jn,m /a and ν = n where n, m ∈ Z, n ≥ 0, m ≥ 1, we have the
following solutions for R(r) to (2.139):
 
jn,m
Rn,m (r) = Jn r , (2.166)
a

and the following solutions for Θ(θ) to (2.136):

Θn (θ) = Dn cos(nθ) + En sin(nθ), (2.167)

 Mathematical Methods in Chemistry (MATH0042) Page 52

where Dn and En are arbitrary constants. Then, multiplying solutions for R(r) and
Θ(θ) that correspond to the same separation constant ν 2 = n2 (recall (2.135)), we
have the following solutions for u(r, θ) to (2.132) with λ = jn,m /a:

un,m (r, θ) = Rn,m (r)Θn (θ)

(2.168)
 
jn,m
= [Dn cos(nθ) + En sin(nθ)] Jn r .
a

We also have the following solutions for T (t) to (2.129) with ω = cλ = cjn,m /a
(recall (2.130))):
   
cjn,m cjn,m
Tn,m (t) = An,m cos t + Bn,m sin t , (2.169)
a a

where An,m and Bn,m are arbitrary constants.

Then, multiplying solutions for u(r, θ) and T (t) that correspond to the same
separation constant γ = −ω 2 = −(cjn,m /a)2 - equivalently, the same values of n and
m - we have the following solutions for f (r, θ, t) to (2.120):

fn,m (r, θ, t) = un,m (r, θ)Tn,m (t)

      
cjn,m cjn,m jn,m
= An,m cos t + Bn,m sin t [Dn cos(nθ) + En sin(nθ)] Jn r
a a a
(2.170)

Hence, summing over all possible values of n and m, we have the following more
general solution for f (r, θ, t) to (2.120):
∞ X
X ∞
f (r, θ, t) = fn,m (r, θ, t). (2.171)
m=1 n=0

2.5 Schrödinger’s equation

Schrödinger’s equation is a PDE that describes the behaviour of quantum mechanical
particles. More specifically, it describes the evolution of the so-called wave function,
which we denote by Ψ(x, t), where x = (x, y, z) denotes a point in 3D space, R3 ,
and t denotes time. Ψ(x, t) is a complex-valued function. |Ψ(x, t)|2 represents a
probability density function, by which we mean that the probability of finding a
particle in a region D of space at a time t is given by the integral
ZZZ
|Ψ(x, t)|2 dxdydz. (2.172)
D

Schrödinger’s equation is:

ℏ2
 
∂Ψ(x, t)
iℏ = − ∆ + V (x, t) Ψ(x, t), (2.173)
∂t 2mp
Page 53

where ℏ = h/2π and h is Planck’s constant, mp is the mass of the particle and
V (x, t) is an associated function known as the potential. Ψ(x, t) must also satisfy
the normalisation condition
ZZZ
|Ψ(x, t)|2 dxdydz = 1, (2.174)
R3

since the probability of finding the particle somewhere in the space is 1.

We are not going to derive Schrödinger’s equation. We are just going to solve it,
or rather, just a particular (relatively simple) case of it.
First, we will assume that the potential V (x, t) is central and independent of
time, i.e., V (x, t) = V (r) depends only on the magnitude r = |x| of x. Now, let us
seek solutions for Ψ(x, t) by separating space and time variables, i.e., solutions of
the form
Ψ(x, t) = ψ(x)f (t). (2.175)
Substituting this form for Ψ(x, t) into (2.173), we get

−ℏ2
iℏψ(x)f ′ (t) = f (t)∆ψ(x) + V (r)ψ(x)f (t).
2mp

Then, dividing through by ψ(x)f (t), we arrive at

f ′ (t) −ℏ2 ∆ψ(x)

iℏ = + V (r).
f (t) 2mp ψ(x)

Now, the left-hand side here depends only on t whereas the right-hand side depends
only on position x (r = |x|). Therefore, both sides must be equal to a constant (the
separation constant). Let us denote this constant by E. So we have

f ′ (t) −ℏ2 ∆ψ(x)

iℏ = + V (r) = E. (2.176)
f (t) 2mp ψ(x)

(E has the physical meaning of energy.) Thus we obtain one equation for f (t) and
one for ψ(x):
iE
f ′ (t) = − f (t), (2.177)
ℏ
 2 
−ℏ
∆ + V (r) ψ(x) = Eψ(x). (2.178)
2mp
The solution to (2.177) is given by
iE
f (t) = Ce− ℏ t , (2.179)

where C is a constant. Equation (2.178) is known as the time-independent Schrödinger’s

equation. Solving (2.178) is generally difficult; we will do it for the ‘simple’ case of
the hydrogen atom.
 Mathematical Methods in Chemistry (MATH0042) Page 54

2.5.1 The hydrogen atom

For the case of the hydrogen atom, x denotes the position of its single electron
relative to its nucleus (a single proton). The corresponding potential V (r) is given
by
q2
V (r) = − , (2.180)
4πϵ0 r
where −q is the charge of the electron (the charge of the nucleus is +q) and ϵ0 is
the vacuum permittivity. In the case of the hydrogen atom, we in fact replace the
mass mp that appears in (2.173) with
me mn
µ= ,
me + mn
where me is the mass of the electron and mn is the mass of the nucleus.
Let us now consider the associated time-independent Schrödinger’s equation
(2.178). Working in spherical polar coordinates (r, θ, ϕ), let us separate variables
again, and seek solutions for ψ(x) = ψ(r, θ, ϕ) of the form

ψ(r, θ, ϕ) = R(r) Y (θ, ϕ). (2.181)

Substitution of (2.181) into (2.178), with m = µ and the Laplacian expressed in

terms of spherical polar coordinates (recall (2.71)), gives

−ℏ2 R ∂ 2Y 1 ∂ 2Y
  
′′ 2 ′ ∂Y
R Y + RY + 2 + cot θ + + V (r)RY = ERY,
2µ r r ∂θ2 ∂θ sin2 θ ∂ϕ2
which we can write as
∂ 2Y 1 ∂ 2Y
 
2 R ∂Y 2µ
R Y + R′ Y + 2
′′
+ cot θ + − (V (r) − E)RY = 0,
r r ∂θ2 ∂θ sin2 θ ∂ϕ2 ℏ2

and hence (by multiplying by r2 /(RY ))

′′
R′ 2µr2 ∂ 2Y 1 ∂ 2Y
 
2R 1 ∂Y
r + 2r + 2 (E − V (r)) = − + cot θ + . (2.182)
R R ℏ Y ∂θ2 ∂θ sin2 θ ∂ϕ2

It follows from (2.182) that

′′
R′ 2µr2 ∂ 2Y 1 ∂ 2Y
 
2R 1 ∂Y
r +2r + 2 (E−V (r)) = − + cot θ + = λ, (2.183)
R R ℏ Y ∂θ2 ∂θ sin2 θ ∂ϕ2

for some separation constant λ. (2.183) gives the following equation for Y (θ, ϕ):

∂ 2Y ∂Y 1 ∂ 2Y
+ cot θ + + λY = 0. (2.184)
∂θ2 ∂θ sin2 θ ∂ϕ2
But (2.184) is just equation (2.78), which we have already solved. We found that
solutions exist for λ = l(l + 1) for l ∈ Z, ≥ 0, and are given by constant multiples of:

Y (θ, ϕ) = Plm (cos θ)eimϕ (2.185)

Page 55

for m ∈ Z with |m| ≤ l - see (2.97). (These are the so-called spherical harmonics.)
Thus it remains only to solve for R(r). With λ = l(l + 1), l ∈ Z, ≥ 0, and
substituting for V (r) using (2.180), the equation for R(r) that follows from (2.183)
is
q2 l(l + 1)ℏ2
 
2 ′′ ′ 2µ 2
r R (r) + 2rR (r) + 2 Er + r− R(r) = 0. (2.186)
ℏ 4πϵ0 2µ
We can simplify this equation by introducing a new variable ρ that is defined by
2p
ρ= 2µ|E| r. (2.187)
ℏ
(note that E < 0, and so |E| = −E), and the function
!
ℏ
P (ρ) = R p ρ . (2.188)
2 2µ|E|

In terms of P (ρ), (2.186) becomes

 2 
2 ′′ ′ ρ
ρ P (ρ) + 2ρP (ρ) + − + γρ − l(l + 1) P (ρ) = 0, (2.189)
4

where
q2
r
µ
γ= . (2.190)
4πϵ0 ℏ 2|E|
(2.189) is a second order homogeneous ODE with polynomial coefficients. Further-
more, one can check that ρ = 0 is a regular singular point of this equation. We
could thus seek solutions to it for P (ρ) of the form
∞
X
β
P (ρ) = ρ αk ρk , (2.191)
k=0

for some index β ∈ R and coefficients αk (with α0 ̸= 0) by using Frobenius’s method.

However, we shall instead proceed as follows.
First, recall that the condition (2.174) must hold. It can be shown that in order
to satisfy this condition, we must have |Ψ(x, t)| → 0 as |x| → ∞, as otherwise the
integral on the left-hand side would be infinite (we omit the full details here, but
roughly speaking, this is because we would be integrating something non-zero over
an infinite volume). Thus we must have R(r) → 0 as r → ∞, or equivalently,

P (ρ) → 0 as ρ → ∞. (2.192)

But as ρ → ∞, so 1/ρ and 1/ρ2 become very small and so (2.189) becomes, approx-
imately,
1
P ′′ (ρ) − P (ρ) = 0. (2.193)
4
(2.193) has solutions P (ρ) = e±ρ/2 . As ρ → ∞, so e−ρ/2 → 0, but eρ/2 → ∞. Thus,
in view of (2.192), we dismiss the possibility that P (ρ) behaves like eρ/2 in this limit,
 Mathematical Methods in Chemistry (MATH0042) Page 56

and instead, seek a solution for P (ρ) (that is valid for all ρ, not just as ρ → ∞) of
the form
P (ρ) = e−ρ/2 G(ρ), (2.194)
for some function G(ρ). Substituting (2.194) into (2.189), we obtain (Exercise:
check this)
ρ2 G′′ (ρ) + (2 − ρ) ρG′ (ρ) + ((γ − 1)ρ − l(l + 1)) G(ρ) = 0. (2.195)
(2.195) is another second order homogeneous ODE with polynomial coefficients.
One can show that ρ = 0 is also a regular singular point of (2.195). Now we will
use Frobenius’s method, and seek a solution for G(ρ) of the form
∞
X
G(ρ) = ρb ak ρk , (2.196)
k=0

for some index b ∈ R and coefficients ak (with a0 ̸= 0). On substituting the form
given by (2.196) for G(ρ) into (2.195) and then considering, in particular, the result-
ing coefficient of ρb on the left-hand side, one finds that the possible values for b are
b = l and b = −(l + 1) (Exercise: check this). But it can also be shown (again, we
omit the details here) that |Ψ(x, t)| must be finite at all finite points x (by a finite
point, we mean a point that is not at infinity). It follows that P (ρ) and hence G(ρ)
must be finite for all finite ρ. However, recalling that l ∈ Z, ≥ 0, one may deduce
that G(ρ) will become infinite at ρ = 0 if b = −(l + 1). Hence we allow only b = l.
It remains to determine the coefficients ak in (2.196). At this point we could
continue by following Frobenius’s method as usual. But instead, let us first introduce
the function H(ρ) defined by
∞
X
H(ρ) = ak ρ k , (2.197)
k=0

so that - recalling (2.196), now with b = l - we have

G(ρ) = ρl H(ρ). (2.198)
Our reason for introducing H(ρ) is that the solutions we are going to find for it
are of another class of special functions, and it is helpful to highlight their presence
here. Substituting (2.198) into (2.195), we get the following equation for H(ρ):
ρH ′′ (ρ) + (2l + 2 − ρ) H ′ (ρ) + (γ − l − 1) H(ρ) = 0. (2.199)
We will now proceed by solving (2.199) for H(ρ) as a power series of the form (2.197)
using the Leibniz-Maclaurin method. Substituting (2.197) into (2.199), we get
∞
X ∞
X ∞
X
ak k(k − 1)ρk−1 + ak k[(2l + 2)ρk−1 − ρk ] + (γ − l − 1) ak ρk = 0,
k=0 k=0 k=0

which we can write as

∞
X X∞ ∞
X ∞
X
k−1 k−1 k−1
ak k(k−1)ρ + ak k(2l+2)ρ − ak−1 (k−1)ρ +(γ−l−1) ak−1 ρk−1 = 0.
k=1 k=1 k=1 k=1
Page 57

Then, for k ≥ 1, we have

ak [k(k − 1) + k(2l + 2)] − ak−1 [(k − 1) − (γ − l − 1)] = 0,

which we can rewrite as

k+l−γ
ak = ak−1 for k ≥ 1 (2.200)
k(k + 2l + 1)

(note that k + 2l + 1 ̸= 0 since l ≥ 0). Now assume that we have fixed l and
γ. Furthermore, assume that γ ̸= (k + l) for all k ≥ 1. Then it follows from
(2.200) that ak ̸= 0 for all k ≥ 1 (a1 = [(1 + l − γ)/(2(l + 1))] a0 ̸= 0, a2 =
[(2 + l − γ)/(2(2l + 3))] a1 ̸= 0, etc). Let us now consider the behaviour of H(ρ)
as ρ → ∞. As ρ → ∞, the behaviour of H(ρ) will be dominated by the terms in
the series in (2.197) that are given by large k (e.g., as ρ → ∞, ρ100 dominates ρ).
These terms will be present since (as we have just shown) ak ̸= 0 for all k ≥ 1. But
for large k, we can approximate the recurrence relation (2.200) by
k ak−1
ak ≈ 2
ak−1 = .
k k
is the relation between the coefficients of the power series expansion of ex ,
But this P
which is ∞ n
n=0 x /n! (recall (1.15)). It follows that, as ρ → ∞,

H(ρ) ≈ eρ .

It then follows from (2.198) and (2.194) that as ρ → ∞,

P (ρ) ≈ ρl e−ρ/2 eρ = ρl eρ/2 .

But this leads to a contradiction of (2.192). Hence, our assumption that γ ̸= (k + l)

for all k ≥ 1 must have been incorrect. Hence we must have

γ = k + l, (2.201)

for some value of k ≥ 1, say k = n − l where n ∈ Z and n ≥ (l + 1) ≥ 1 (recall

that l ∈ Z, ≥ 0). Then γ = n is an integer, and it follows from (2.200) that ak = 0
for all k ≥ (n − l), so the series solution (2.197) for H(ρ) to (2.199) is simply a
polynomial of degree d = n − l − 1. In fact, (2.199) is of the form of the so-called
associated Laguerre equation, and this polynomial solution for H(ρ) is an arbitrary
constant multiple of what is called an associated Laguerre polynomial. The latter is
(following the standard convention) denoted by L2l+1
n−l−1 (ρ) and given by the formula

n−l−1
X [(n + l)!]2
L2l+1
n−l−1 (ρ) = ρk . (2.202)
k=0
k!(n − l − 1 − k)!(2l + 1 + k)!

Thus, recalling (2.194) and (2.198), we have, for l ∈ Z, ≥ 0 and γ = n ∈ Z, ≥

(l + 1), the following solutions for P (ρ) to (2.189):

Pn,l (ρ) = e−ρ/2 ρl L2l+1

n−l−1 (ρ). (2.203)
 Mathematical Methods in Chemistry (MATH0042) Page 58

[Remark: Note that, by recalling again the power series expansion ex = ∞ n

P
n=0 x /n!
(see (1.15)), it is evident that Pn,l (ρ) as given by (2.203) can be written in the
form (2.191). We would have retrieved Pn,l (ρ) in the form (2.191) if we had used
Frobenius’s method to seek solutions for P (ρ). However, given Pn,l (ρ) in the form
(2.191), it is not obvious that one could write it in the form (2.203). Following
the approach that we took, enables one to extract the e−ρ/2 term in this solution
more easily. Note also that Pn,l (ρ) ≈ ce−ρ/2 ρl as ρ → ∞ for some constant c, so
′′
Pn,l (ρ) ≈ 4c e−ρ/2 ρl , and hence in this limit, Pn,l (ρ) satisfies (2.193).]
Now, it follows from (2.188) that
 
2p
R(r) = P 2µ|E|r . (2.204)
ℏ

And it follows from (2.190), now with γ = n, that

r
2p 2p q2 µ
2µ|E| = 2µ
ℏ ℏ 4πϵ0 ℏn 2
2
µq (2.205)
=
2πϵ0 ℏ2 n
2
= ,
nα0
where we introduce
4πϵ0 ℏ2
α0 = . (2.206)
µ q2
Thus (2.204) becomes  
2
R(r) = P r . (2.207)
α0 n
It also follows from (2.190) (now with γ = n) that
2
q2

µ
E = En = − , (2.208)
2 4πϵ0 ℏn

(recall that E < 0, and so |E| = −E).

Thus, it follows from (2.203) and (2.207) that for n, l ∈ Z with 0 ≤ l ≤ n − 1, we
have the following solutions for R(r) to (2.186) with E = En as given by (2.208):
 
−r/(nα0 ) l 2l+1 2r
Rn,l (r) = cn,l e r Ln−l−1 , (2.209)
nα0
where cn,l is a constant.
Thus, finally, recalling (2.175), (2.179), (2.181), (2.185) and also (2.209), we
have the following solutions for Ψ to Schrödinger’s equation (2.173) for the case of
a hydrogen atom:
   
−iEn r l 2l+1 2r
Ψn,l,m (r, θ, ϕ, t) = Cn,l,m exp t− + imϕ r Ln−l−1 Plm (cos θ),
ℏ nα0 nα0
(2.210)
Page 59

where Cn,l,m is a constant, and n, l, m ∈ Z with 0 ≤ l ≤ n − 1, and |m| ≤ l.

The three numbers n, l, m, are called, the principal, azimuthal and magnetic
quantum numbers, respectively. Note that it is evident from (2.208) that the energy,
E, can only take discrete values, as determined by n. This is referred to as the quan-
tisation of energy levels. Furthermore, for a given n, l can take n values, and for each
of these values of l, m can take 2l + 1 values (namely −l, −l + 1, . . . , −1, 0, 1, . . . , l).
This means that for a given n, there are a total of
n−1 n−1 n−1  
X X X (n − 1)n
(2l + 1) = 2 l+ 1=2 + n = n2 (2.211)
l=0 l=0 l=0
2

distinct solutions (for thePfirst sum, we have used the formula for the sum of a finite
n−1 2
arithmetic progression: l=0 l = (n − 1)n/2), i.e., n different quantum states in
which the electron can be found. Thus, every energy level can accommodate several
different quantum states. This is referred to as the degeneracy of the spectrum.
As an example of the above, consider the case n = 1. Then l = m = 0 and

R1,0 (r) = c1,0 e−r/α0 L10 (2r/α0 ) = ĉ1,0 e−r/α0 , (2.212)

for some constant ĉ1,0 (note that L10 (r) is a constant), and

Ψ1,0,0 (r, θ, ϕ, t) = Ĉ1,0,0 e−iE1 t/ℏ e−r/α0 , (2.213)

for some constant Ĉ1,0,0 (note that P00 (θ) is simply a constant). Note that Ψ1,0,0 (r, θ, ϕ, t)
depends on r but is independent of θ and ϕ. Furthermore, the corresponding proba-
bility density function (for the location of the atom’s electron) is |Ψ1,0,0 (r, θ, ϕ, t)|2 =
2
Ĉ1,0,0 e−2r/α0 (note that |e−iE1 t/ℏ | = 1 for all t). This decreases exponentially with
r. Similarly, forn ≥ 2, Ψn,0,0 (r, θ, ϕ, t) is also independent of θ and ϕ, although
now L2l+1
n−l−1
2r
nα0
is a polynomial (in r) of degree n − 1. Figure 2.4 shows plots of
Rn,0 (r) for n = 1, 2, 3, 4. Also shown are plots of r2 Rn,0
2
(r). These give an indication
of the corresponding probability density functions for the location, x, of the atom’s
electron; the area under the graph of r2 Rn,0 2
(r) for r1 < r < r2 represents (up to a
constant multiple) the corresponding probability that r1 < |x| < r2 - recall (2.172).
(The factor of r2 comes from the element of volume dxdydz, when this is written
in terms of spherical polar coordinates - see Problem Sheet 6). Observe that as n
increases, these plots indicate that the electron tends to move further and further
away from the nucleus.
 Mathematical Methods in Chemistry (MATH0042) Page 60

Figure 2.4: Radial solutions for Schrödinger’s equation for the hydrogen atom.