Quantum Mechanics Formulas: by R.L.

Grith@ UCB
Physical Constants
Name Symbol Value Unit
Number 3.14159265358979323846
Number e e 2.71828182845904523536
Eulers constant = lim
n
_
n

k=1
1/k ln(n)
_
= 0.5772156649
Elementary charge e 1.60217733 10
19
C
Gravitational constant G, 6.67259 10
11
m
3
kg
1
s
2
Fine-structure constant = e
2
/2hc
0
1/137
Speed of light in vacuum c 2.99792458 10
8
m/s (def)
Permittivity of the vacuum
0
8.854187 10
12
F/m
Permeability of the vacuum
0
4 10
7
H/m
(4
0
)
1
8.9876 10
9
Nm
2
C
2
Plancks constant h 6.6260755 10
34
Js
Diracs constant = h/2 1.0545727 10
34
Js
Bohr magneton
B
= e/2m
e
9.2741 10
24
Am
2
Bohr radius a
0
0.52918

A
Rydbergs constant Ry 13.595 eV
Electron Compton wavelength
Ce
= h/c 2.2463 10
12
m
Proton Compton wavelength
Cp
= h/m
p
c 1.3214 10
15
m
Reduced mass of the H-atom
H
9.1045755 10
31
kg
Stefan-Boltzmanns constant 5.67032 10
8
Wm
2
K
4
Wiens constant k
W
2.8978 10
3
mK
Molar gasconstant R 8.31441 Jmol
1
K
1
Avogadros constant N
A
6.0221367 10
23
mol
1
Boltzmanns constant k = R/N
A
1.380658 10
23
J/K
Electron mass m
e
9.1093897 10
31
kg
Proton mass m
p
1.6726231 10
27
kg
Neutron mass m
n
1.674954 10
27
kg
Elementary mass unit m
u
=
1
12
m(
12
6
C) 1.6605656 10
27
kg
Nuclear magneton
N
5.0508 10
27
J/T
Diameter of the Sun D

1392 10
6
m
Mass of the Sun M

1.989 10
30
kg
Rotational period of the Sun T

25.38 days
Radius of Earth R
A
6.378 10
6
m
Mass of Earth M
A
5.976 10
24
kg
Rotational period of Earth T
A
23.96 hours
Earth orbital period Tropical year 365.24219879 days
Astronomical unit AU 1.4959787066 10
11
m
Light year lj 9.4605 10
15
m
Parsec pc 3.0857 10
16
m
Hubble constant H (75 25) kms
1
Mpc
1
1
Chapter 1: The wave Function
Schrodinger Equation
i

t
=

2
2m

x
2
+V (1)
Probability
P
ab
=
_
b
a
[(x, t)[
2
dx (2)
For Normalization
1 =
_

[(x, t)[
2
dx, where [(x, t)[
2
=

(3)
once is normalized it stays normalized for
all time
d
dt
_

[(x, t)[
2
dx = 0 (4)
Operators for momentum and position in gen-
eral form
x) =
_

xdx (5)
p) =
_

x
_
dx (6)
Q(x, p)) =
_

Q
_
x,

i

x
_
dx (7)
Expectation value of kinetic energy
T) =

2
2m
_

x
2
dx (8)
de Broglie formula
p =
h

=
2

(9)
uncertainty principle

p


2
(10)
probability current
J(x, t)
i
2m
_

x
_
(11)
Chapter 2: Stationary States
seperation of variables
(x, t) = (x)(t) = (x)e
iEt/
(12)
d
dt
=
iE

(13)
(t) = e
iEt/
(14)
(x, t) = (x)e
iEt/
(15)
Time Independent Schrodinger equation

2
2m
d
2

dx
2
+V = E (16)
Hamiltonian
H(x, p) =
p
2
2m
+V (x) (17)
substitute p = (/i)(/x) and the corre-
sponding Hamiltonian operator is

H =

2
2m
d
2

dx
2
+V (x) = E (18)
H) =

Hdx = E (19)
You can always write the general solution of
the Schrodinger equation as a linear combination
of seperable solutions
(x, 0) =

n=1
c
n

n
(x) (20)
(x, t) =

n=1
c
n

n
(x)e
iEnt/
=

n=1
c
n

n
(x, t)
(21)
The Innite Square Well
V (x) =
_
0, if 0xa
, otherwise
(22)
plugging this into the TISE and we get
d
2

dx
2
= k
2
(23)
k

2mE

(24)
2
(x) = Asin kx (25)
k
n
=
n
a
(26)
E
n
=
n
2

2
2ma
2
(27)
after normalization, the general TISE solu-
tions for the innite square well is

n
(x) =
_
2
a
sin
_
n
a
x
_
(28)
as a collection, the functions
n
(x) have some
interesting properties
1. They are alternately even and odd, with
respect to the center of the well.
2. As you go up in energy, each succesive
state has one more node (zero crossing)
3. They are mutually orthogonal, in the
sense that
_

m
(x)

n
(x)dx = 0 (29)
Kronecker delta
_

m
(x)

n
(x)dx =
mn
(30)

mn
=
_
0, if m=n
1, if m=n
(31)
we say that s are orthonormal
4. They are complete, in the sense that any
other function, f(x), can be expressed as a linear
combination of them:
f(x) =

n=1
c
n

n
(x) =
_
2
a

n=1
c
n
sin
_
n
a
x
_
(32)
the c
n
s can fe found using Fouriers trick
c
n
=
_

n
(x)

f(x)dx (33)
=
_
2
a
_
a
0
sin
_
n
a
x
_
(x, 0)dx (34)
P
n
= [c
n
[
2
(35)

n=1
[c
n
[
2
= 1 (36)
H) =

n=1
[c
n
[
2
E
n
(37)
The harmonic Oscillator
V (x) =
1
2
kx
2
(38)

_
k
m
(39)
V (x) =
1
2
m
2
x
2
(40)
rewriting equation 18 in a more suggestive
form
1
2m
[p
2
+ (mx)
2
] = E (41)
we can use raising and lowering operators to
solve this problem
a
+

1

2m
(ip +mx) (42)
a

2m
(ip +mx) (43)
a

a
+
=
1
2m
[p
2
+ (mx)
2
imw(xp px)]
(44)
commutator
[A, B] AB BA (45)
[x, p]f(x) = if(x) (46)
[x, p] = i (47)
a

a
+
=
1

H +
1
2
(48)
H =
_
a

a
+

1
2
_
(49)
a
+
a

=
1

H
1
2
(50)
[a

, a
+
] = 1 (51)
H =
_
a
+
a

+
1
2
_
(52)
In terms of a

, then the Schrodinger equation

of the harmonic oscillator takes the form

_
a

1
2
_
= E (53)
H(a
+
) = (E + )(a
+
) (54)
H(a

) = (E )(a

) (55)
3
there must be a lowest state and this occurs
when
a

0
= 0 (56)
this means
1

2m
_

d
dx
+mx
_

0
= 0 (57)
the genral solution for the ground state of the
harmonic oscillator is

0
(x) =
_
m

_
e

m
2
x
2
(58)
E
0
=
1
2
(59)
E
n
=
_
n +
1
2
_
(60)

n
= A
n
(a
+
)
n

0
(x) (61)
a
+
a

n
= n
n
, a

a
+

n
= (n + 1)
n
(62)
a
+

n
=

n + 1
n+1
, a

n
=

n
n1
(63)

n
=
1

n!
(a
+
)
n

0
(64)
x =
_

2m
(a
+
+a

) (65)
p = i
_
m
2
(a
+
a

) (66)
The Free particle
V (x) = 0 everywhere
(x, t) = Ae
ik(x
k
2m
t
+Be
ik(x+
k
2m
t)
(67)

k
(x, t) = Ae
i(kx
k
2
2m
t
(68)
k

2mE

, with
_
k>0 travelingtotheright
k<0 travelingtotheleft
(69)
p = k (70)
v
quantum
=
_
E
2m
(71)
v
classical
= 2v
quantum
(72)
(x, t) =
1

2
_

(k)e
i(kx
k
2
2m
t)
dk (73)
to determine (k) we can use the TISE which
is
(x, 0) =
1

2
_

(k)e
ikx
dk (74)
this is a classic problem in Fourier analysis;
the answer is provided by Plancherels theo-
rem
f(x) =
1

2
_

F(k)e
ikx
dk (75)
F(k) =
1

2
_

f(x)e
ikx
dx (76)
so the solution to the generic quantum prob-
lem, for the free particle is equation 74, with
(k) =
1

2
_

(x, 0)e
ikx
dx (77)
The Delta-Function Potential
_
E<0 bound state
E>0 scattering state
(78)
(x)
_
0, if x=0
, if x=0
_
, with
_

(x)dx = 1
(79)
_

f(x)(xa)dx = f(a)
_

(xa)dx = f(a)
(80)
thats the most important property of the
delta function: Under the integral sign it serves
to pick out the value of f(x) at the point a.
Lets consider a potential of the form
V (x) = (x) (81)
plugging this into equation 18 in the region
x < 0, V (x) = 0 so
4
d
2

dx
2
=
2mE

2
= k
2
(82)
where
k

2mE

(83)
for bound states E < 0 the general solution is
when x < 0
(x) = Ae
kx
+Be
kx
(84)
the rst term blows up as x so we
must choose A=0
(x) = Be
kx
(85)
for bound states E < 0 the general solution is
when x > 0
(x) = Fe
kx
+Ge
kx
(86)
the second term blows up as x so we
must choose G=0
(x) = Fe
kx
(87)
It remains only to stitch these two functions
together, using the appropriate boundery condi-
tions at x = 0, the standard boundary conditions
for
1. is always continuous 2. d/dx is con-
tinuous except at points where the potential is
innite
in this case the rst boundary condition tells
us that F = B, so
(x) =
_
Be
kx
, (x0)
Be
kx
, (x0)
(88)
the second boundary condition gives

_
d
dx
_
=
2m

2
(0) (89)
Evidently the delat-function well, regardless
of its strength , has exactly one bound state
(x) =

e
m|x|
2
; E =
m
2
2
2
(90)
What about scattering states, with E > 0?For
x < 0 the Schrodinger equation read
d
2

dx
2
=
2mE

2
= k
2
(91)
where
k

2mE

(92)
is real and positive. The general solution is
for x < 0
(x) = Ae
ikx
+Be
ikx
(93)
and this time we cannot rule out either term,
since neither blows up. and similarly for x > 0
(x) = Fe
ikx
+Ge
ikx
(94)
for boundary condition number 1 this implies
F +G = A +B (95)
the derivatives are
_
d/dx=ik(Fe
ikx
+Ge
ikx
), for(x>0), so d/dx|+=ik(FG)
d/dx=ik(Ae
ikx
+Be
ikx
), for(x<0), so d/dx|=ik(AB)
(96)
and hence (d/dx) = ik(F G A + B),
meanwhile (0) = (A+B), so the second bound-
ery condition says
ik(F GA+B) =
2m

2
(A +B) (97)
or more compactly
FG = A(1+2i)B(12i), where
m

2
k
(98)
having imposed both boundary condition we
are left with two equations and four unknowns.
it follows A is the amplitude of a wave coming
from the right, B is the amplitude of the wave
returning to the left, F is the amplitude of the
wave traveling of to the right, and G is the am-
plitude of the wave coming in from the right.
G = 0, for scattering from the left (99)
A = 0, for scattering from the right (100)
for scattering from the left, A is the amplitude
of the incident wave, B is the amplitude of the
reected wave, and F is the amplitude of the
transmitted wave solving equations 95 and 98
for B and F, we nd
B =
i
1 i
A, F =
1
1 i
A (101)
5
The reection coecient is
R
[B[
2
[A[
2
=

2
1 +
2
(102)
and the transmission coecient is
T
[F[
2
[A[
2
=
1
1 +
2
(103)
and the sum should be one
R +T = 1 (104)
R and T are functions of and hence func-
tions of E
R =
1
1 + (2
2
E/m
2
)
(105)
T =
1
1 + (m
2
/2
2
E)
(106)
The Finite Square Well
V (x) =
_
V0, for axa
0, for |x|>a
(107)
This potential admits both bound states and
scattering states. we will look at the bound
states rst. in the region x < a the potential
is zero, so the Schrodinger equation reads
d
2

dx
2
=
2mE

2
= k
2
(108)
where
k

2mE

(109)
the general solution as before is
(x) = Ae
kx
+Be
kx
(110)
the rst term blows up as x so we
must choose A=0
(x) = Be
kx
, for x < a (111)
in the region a < x < a,V (x) = V
0
, and
the Schrodinger equation reads

2
2m
d
2

dx
2
V
0
= E, or
d
2

dx
2
= l
2

(112)
where
l
_
2m(E +V
0
)

(113)
and the general solution is
(x) = C sin(lx) +Dcos(lx), for a < x < a
(114)
when x > a the genral solution is
(x) = Fe
kx
+Ge
kx
(115)
the rst term blows up as x so we
must choose G=0
(x) = Fe
kx
, for x > a (116)
The next step is to impose boundary condi-
tions: and d/dx continuous at a and a.
but we can save a little time by noting that this
potential is an even function, so we can assume
with no loss of generality that the solutions are
either even or odd. since (x) = (x).
for even solutions use Dcos(lx) and for odd
solutions use C sin(lx). I will show the even so-
lutions
The continuity of (x), at x = a, says
Fe
ka
= Dcos(la) (117)
and the continuity of d/dx, says
kFe
ka
= lDsin(la) (118)
deviding equation 118 by 117 , we nd that
k = l tan(la) (119)
This is a formula for the allowed energies,
since k and l are both functions of E. To solve
for E, we rst adopt a nicer notation: Let
z la, and z
0

a

_
2mV
0
(120)
(k
2
+l
2
) =
2mV
0

2
, so ka =
_
z
2
0
z
2
(121)
and now equation 119 read
tan(z) =
_
(z
0
/z)
2
1 (122)
This is a trancendental equation and can olny
be solved numerically or graphically. Two limit-
ing cases are of special interest.
1. Wide, deep well. if z
0
is very large,
the intersections occur just slightly below z
n
=
n/2, with n odd it follows that
6
E
n
+V
0

=
n
2

2
2m(2a)
2
(123)
2. Shallow, narrow well. As z
0
decreases,
there are fewer and fewer bound states, until -
nally (for z
0
< /2,where the lowest odd state
disappears) only one remains. It is interesting
to note that there is always one bound state, no
matter how weak the well becomes.
Now moving on to scattering states where
E > 0
the general solution as before is
(x) = Ae
ikx
+Be
ikx
for (x < a) (124)
where(as usual)
k

2mE

(125)
inside the well, where V (x) = V
0
and the
general solution is
(x) = C sin(lx) +Dcos(lx), for a < x < a
(126)
where as before
l
_
2m(E +V
0
)

(127)
To the right, assuming there is no incoming
wave in this region, we have
(x) = Fe
ikx
(128)
Here A is the incident amplitude, B is the re-
ected amplitude, and F is the transmitted am-
plitude.
There are four boundary conditions: Conti-
nuity of (x) at a says
Ae
ika
+Be
ika
= C sin(la) +Dcos(la) (129)
continuity of d/dx at a gives
ik[Ae
ika
Be
ika
] = l[C cos(la) +Dsin(la)]
(130)
continuity of (x) at +a yields
C sin(la) +Dcos(la) = Fe
ika
(131)
continuity of d/dx at +a gives
l[C cos(la) Dsin(la)] = iFe
ika
(132)
We can use two of these to eliminate C and
D, and solve the remaining two for B and F:
B = i
sin(2la)
2kl
(l
2
k
2
)F (133)
F =
e
2ika
A
cos(2la) i
(k
2
+l
2
2kl
sin(2la)
(134)
The transmission coecient (T = [F[
2
/[A[
2
)
expressed in terms of the original variables, is
given by
T
1
= 1+
V
2
0
4E(E +V
0
)
sin
2
_
2a

_
2m(E +V
0
)
_
(135)
Notice that T = 1 (the well becomes trans-
parent) whenever the sine is zero, which is to
say, when
2a

_
2m(E
n
+V
0
) = n (136)
where n is an integer. The energies for perfect
transmission, then are given by
E
n
+V
0
=
n
2

2
2m(2a)
2
(137)
which happens to be the allowed energies for
the innite square well
Chapter 3: Formalism
The collection of all functions of x constitues
a vector space. To represent a possible physical
state, the wave function must be normalized:
_
[[
2
dx = 1 (138)
The set of all square-integrable functions,
on a specied interval
f(x) such that
_
b
a
[f(x)[
2
dx < (139)
this constitutes a (much smaller) vector space.
mathematicians call it L
2
(a, b); physicist call it
Hilbert Space. In quantum mechanics, then
7
Wave functions live in Hilbert Space
We dene the inner product of two func-
tions, f(x) and g(x), as follows
f[g)
_
b
a
f(x)

g(x)dx (140)
If f and g are both square-integrable (that
is, if they are both in Hilbert space),their inner
product is guaranteed to exist (the integral in
equation 140 converges to a nite number), This
follows from the integral Schwarz inequality

_
b
a
f(x)

g(x)dx

_
b
a
[f(x)[
2
dx
_
b
a
[g(x)[
2
dx
(141)
Notice in particular that
g[f) = f[g)

(142)
Moreover, the inner product of f(x) with it-
self,
f[f) =
_
b
a
[f(x)[
2
dx 0 (143)
is real and non-negative; its zero only if
f(x) = 0
A function is said to be normalized if its
inner product with itself is 1; two functions are
orthogonal if their inner product is 0; and a set
of functions, f
n
, is orthonormal if they are
normalized and mutually orthogonal;
f
m
[f
n
) =
mn
(144)
Finally, a set of functions is complete if any
other function (in Hilbert space) can be ex-
pressed as a linear combination of them
f(x) =

n=1
c
n
f
n
(x) (145)
if the functions f
n
(x) are orthonormal, the
coecients are given by Fouriers trick
c
n
= f
n
[f) (146)
Observables
The expectation value of an observable Q(x, p)
can be expressed very neatly in inner-product
notation:
Q) =
_

Qdx = [

Q) (147)
now the outcome of a measurement has got to
be real, and so, a fortiori, is the average of many
measurements:
Q) = Q)

(148)
But the complex conjugate of an inner prod-
uct reverses the order
[

Q) =

Q[) (149)
and this must hold true for any wave function
. Thus operators representing observables have
the verty special property that
f[

Qf) =

Qf[f) for all f(x) (150)

we call such operator hermitian
Observables are represented by hermitian operators
Determinate States

Q = q (151)
This is the eigenvalue equation for the op-
erator

Q; is an eigenfunction of

Q, and q is
the corresponding eigenvalue; thus
Determinate states are eigenfunctions of

Q
Measurement of q on such a state is certain
to yield the eigenvalue q. The collection of all
eigenvalues of an operator is called its spec-
trum. Sometimes two (or more) linearly in-
dependent eigenfunctions share the same eigen-
value; in that case the spectrum is said to be
degenerate.
Eigenfunctions of Hermitian Operators
Discrete Spectra
Mathematically, the normalized eigenfunc-
tions of a hermitian operato have two important
properties.
Theorem 1: Their eigenvalues are real.

Qf = qf (152)
(i.e.,f(x) is an eigenfunction of

Q, with eigen-
value q), and
f[

Qf) =

Qf[f) (153)
then
qf[f) = q f[f) (154)
and q must be real
8
Theorem 2: Eigenfunctions belonging to
distinct eigenvalues are orthogonal.

Qf = qf, and

Qg = q

g (155)
and

Q is hermitian. Then f[

Qg) =

Qf[g)
so
q

f[g) = q f[g) (156)

again the inner product exist because the
eigenfunctions are in Hilbert space by assump-
tion. but q is real so if q

,= q it must be that
f[g) = 0.
Continuous Spectrum
If the spectrum of a hermitian operator is con-
tinuous, the eigenfunctions are not normalizable,
and the proofs of Theorems 1 and 2 fail, because
the inner products may not exist.
Generalized Statistical Interpretation
If you measure an observable Q(x, p) on a par-
ticle in the state (x, t), you are certain to get
one of the eigenvalues of the hermitian operator.
If the spectrum is dicrete, the probability of get-
ting the particular eigenvalue q
n
associated with
the orthonormalized eigenfunction f
n
(x) is
[c
n
[
2
, where c
n
= f
n
[) (157)
If the spectrum is continuous, with real eigen-
values q(z) and associated Dirac orthonormal-
ized eigenfunctions f
z
(x), the probability of get-
ting result in the range dz is
[c(z)[
2
dz, where c(z) = f
z
[) (158)
The eigenfunctions of a hermitian operator
are complete, so that the wave function can be
written as a linear combination of them
(x, t) =

n
c
n
f
n
(x) (159)
Because the eigenfunctions are orthonormal,
the coecients are given by Fouriers trick
c
n
= f
n
[) =
_
f
n
(x)

(x, t)dx (160)

Of course, the total probability (summed over
all possible outcomes) has got to equal one

n
[c
n
[
2
= 1 (161)
Similarly, the expectation value Q should be
the sum over all possible outcomes of the eigen-
value times the probablity of getting that eigen-
value
Q) =

n
q
n
[c
n
[
2
(162)
The momentum space wave function,
(p, t) is essentially the Fourier transform of the
position space wave function (x, t), which by
Plancherels theorem is its inverse Fourier trans-
form.
(p, t) =
1

2
_

e
ipx/
(x, t)dx (163)
(x, t) =
1

2
_

e
ipx/
(p, t)dp (164)
(165)
According to the generalized statistical inter-
pretation, the probability that a measurement of
the momentum would yield a result in the range
dp is
[(p, t)[
2
dp (166)
The Uncertainty Principle
The generalized uncertainty principle is

2
A

2
B

_
1
2i
[

A,

B])
_
2
(167)
The energy-time uncertainty principle
d
dt
Q) =
i

[

H,

Q]) +
_

Q
dt
_
(168)
Chapter 4: Quantum Mechanics in Three
Dimensions
The generalization to three dimensions is
straightforward. Schrodingers equation says
i

t
= H (169)
the Hamiltonian operator H is obtained from the
classical energy
1
2
mv
2
+V =
1
2m
(p
2
x
+p
2
y
+p
2
z
) +V (170)
by the standard prescription (applied now to
y and z, as well as x)
9
p
x


i

x
, p
y


i

y
, p
z


i

z
(171)
or
p

i
(172)
thus
i

t
=

2
2m

2
+V (173)
where

2
x
2
+

2
y
2
+

2
z
2
(174)
is the Laplacian, in cartesian coordinates.
The potential V and the wave function are
now functions of r = (x, y, z) and t.
the normalization condition reads
_
[[
2
d
3
r = 1 (175)
with the integral taken over all space. if the
potential is independent of time, there will be a
complete set of stationary states

n
(r, t) =
n
(r)e
iEnt/
(176)
where the spatial wave function
n
satises
the time-independent Schrodinger

2
2m

2
+V = E (177)
The general solution to the time-dependent
Schrodinger equation is
(r, t) =

c
n

n
(r)e
iEnt/
(178)
Seperation Of variables Typically, the po-
tential is a function only of the distance from
the origin. In that case it is natural to adopt
spherical coordinates, (r, , ). In spherical
coordinates the Laplacian takes the form

2
=
1
r
2

r
_
r
2

r
_
+
1
r
2
sin

_
sin

_
+
1
r
2
sin
2

_

2

2
_
(179)
We begin by looking for solutions that are
seperable by products
(r, , ) = R(r)Y (, ) (180)
,. The Angular Equation
sin

_
sin
Y

_
+

2
Y

2
= l(l + 1) sin
2
Y
(181)
you might recognize this equation, it occurs
in the solution to Laplaces equation in classical
electrodynamics. as always, we try seperation of
variables
Y (, ) = ()() (182)
plugging this in and deviding by gives us
two solutions
1

_
sin
d
d
_
sin
d
d
__
+l(l + 1) sin
2
= m
2
(183)
1

d
2

d
2
= m
2
(184)
The equation is easy
d
2

d
2
= m
2
() = e
im
(185)
Now when advances by 2, we return to the
same point in space, so it is natural to require
that
( + 2) = () (186)
from this it follows that m must be an integer
m = 0, 1, 2, .... (187)
The equation is not so simple, the solution
is
() = AP
m
l
(cos ) (188)
where P
m
l
is the associated Legendre func-
tion, dened by
P
m
l
(x) (1 x
2
)
|m|/2
_
d
dx
_
|m|
P
l
(x) (189)
and P
l
(x) is the lth Legendre polynomial,
dened by the Rodrigues formula
10
P
l
(x)
1
2
l
l!
_
d
dx
_
l
(x
2
1)
l
(190)
Now, the volume element in spherical coordi-
nates is
d
3
r = r
2
sindrdd (191)
so the normalization condition becomes
_
[[
2
r
2
drdd =
_
[R[
2
r
2
dr
_
[Y [
2
sin dd = 1 (192)
It is convenient to normalize R and Y sepa-
rately
_

0
[R[
2
r
2
dr = 1 (193)
_
2
0
_

0
[Y [
2
sindd = 1 (194)
The normalized angular wave functions are
called spherical harmonics
Y
m
l
(, ) =

(2l + 1)
4
(l [m[)!
(l +[m[)!
e
im
P
m
l
(cos )
(195)
where = (1)
m
for m 0 and = 1 for
m 0. As we shall prove later on, they are
automatically orthogonal.
The Radial Equation Notice that the an-
gular part of the wave function, Y (, ), is the
same for all spherically symmetric potentials; the
actual shape of the potential ,V (r), aects only
the radial part of the wave function,R(r), which
is determined by
u(r) rR(r) (196)
using this relationship we can now write the
radial equation as

2
2m
d
2
u
dr
2
+
_
V +

2
2m
l(l + 1)
r
2
_
u = Eu (197)
the normalization condition becomes
_

0
[u[
2
dr = 1 (198)
this is as far as we can go until a specic po-
tential V (r) is provided.
The innite spherical well
V (r) =
_
0, ifra
, ifr>a
(199)
Outside the well, the wave function is zero;
inside the well, the radial equation says
d
2
u
dr
2
=
_
l(l + 1)
r
2
k
2
_
u (200)
where
k

2mE

(201)
as usual. Our problem is to solve this equa-
tion, subject to the boundary condition u(a) =
0. The case l = 0 is easy
d
2
u
dr
2
= k
2
u u(r) = Asin(kr) +Bcos(kr)
(202)
we must choose B = 0 because r 0 the
radial wave function blows up. The boundary
condition then requires sin(ka) = 0, and hence
ka = n, for some integer n. The allowed ener-
gies are evidently
E
n0
=
n
2

2
2ma
2
, (n = 1, 2, 3, ...) (203)
the same as for the one-dimensional innite
square well. The general solution to equation
200 (for an arbitrary integer l) is not as familiar
u(r) = Arj
l
(kr) + Brn
l
(kr) (204)
where j
l
(x) is the spherical Bessel function
pf order l, and n
l
(x) is the spherical Neumann
function of order l. They are dened as follows
j
l
(x) (x)
l
_
1
x
d
dx
_
l
sin x
x
(205)
n
l
(x) (x)
l
_
1
x
d
dx
_
l
cos x
x
(206)
Notice that the Bessel functions are nite at
the origin, but Neumann functions blow up at
the origin. Accordingly, we must have B
l
= 0,
and hence
11
R(r) = Aj
l
(kr) (207)
There remains the boundary condition, R(a) =
0. Evidently k must be chosen such that
j
l
(ka) = 0 (208)
the boundary condition requires that
k =
1
a

nl
(209)
where
nl
is the nth zero of the lth spherical
Bessel function. The allowed energies then, are
given by
E
nl
=

2
2ma
2

2
nl
(210)
and the wave functions are

nlm
(r, , ) = A
nl
(
nl
r/a)Y
m
l
(, ) (211)
with the constant A
nl
to be detemined by nor-
malization.
The Hydrogen Atom
From Coulombs law, the potential energy is
V (r) =
e
2
4
0
1
r
(212)
and the radial equation says

2
2m
d
2
u
dr
2
+
_

e
2
4
0
1
r
+

2
2m
l(l + 1)
r
2
_
u = Eu
(213)
our problem is to solve this equation for u(r),
and determine the allowed energies. Our rst
task is to tidy up the notation. Let
k

2mE

E =
k
2

2
2m
(214)
dividing equation 213 by E gives
1
k
2
d
2
u
dr
2
=
_
1
me
2
2
0

2
k
1
(kr)
+
l(l + 1)
(kr)
2
_
u
(215)
This suggest that we introduce
kr and
0

me
2
2
0

2
k
(216)
so that
d
2
u
d
2
=
_
1

0

+
l(l + 1)

2
_
u (217)
Next we examine the asymptotic form of the
solutions. As , the constant term in the
brackets dominate, so approximately
d
2
u
d
2
= u (218)
The general solution is
u() = Ae

+Be

(219)
but e

blows up ( as ), so B = 0,
evidently
u() Ae

(220)
for large . On the other hand, as 0 the
centrifugal term dominates, approximately then
d
2
u
d
2
=
l(l + 1)

2
u (221)
the general solution is
u() = C
l+1
+D
l
(222)
let D = 0, because
l
blows up as 0,
thus
u() C
l+1
(223)
The next step is to peel o the asymptotic
behaviour, introducing the new function v()
u() =
l+1
e

v() (224)
Finally, we assume the solution, v(), can be
expressed as a power series in
v() =

j=0
c
j

j
(225)
c
j+1
=
_
2(j +l + 1)
0
(j + 1)(j + 2l + 2)
_
c
j
(226)
This recursion formula determines the coef-
cients, and hence the function v(). Now lets
see what the coecients look like for large j (this
corresponds to large , where the higher power
dominate . In this regime the recursion formula
says
12
c
j+1

=
2j
j(j + 1)
c
j
=
2
j + 1
c
j
(227)
suppose for a moment that this were exact.
Then
c
j
=
2
j
j!
c
0
(228)
so
v() = c
0

j=0
2
j
j!

j
(229)
and hence
u() = c
0

l+1
e

(230)
which blows up at large . The series must
terminate. There must occur some maximal in-
teger, j
max
such that
c
jmax+1
= 0 (231)
(and beyond which all coecients vanish au-
tomatically). equation 226 becomes
2(j
max
+l + 1)
0
= 0 (232)
Dening
n j
max
+l + 1 (233)
(the so-called principal quantum num-
ber), we have

0
= n (234)
But
0
determines E
E =
k
2

2
2m
=
me
4
8
2

2
0
(235)
so the allowed energies are
E
n
=
_
m
2
2
_
e
2
4
_
2
_
1
n
2
=
E
1
n
2
n = 1, 2, ..
(236)
This is the famous Bohr Formula. combin-
ing equation 216 and 234 we get
k =
_
me
2
4
0

2
_
1
n
=
1
an
(237)
where
a
4
0

2
me
2
= 0.529x10
10
m (238)
is called the Bohr radius. it follows that
=
r
an
(239)
The spatial wave functions for hydrogen are
labeled by three quantum numbers (n, l, andm)

nlm
(r, , ) = R
nl
(r)Y
m
l
(, ) (240)
and
R
nl
(r) =
1
r

l+1
e

v() (241)
and v() is a polynomial of degree j
max
=
n l 1 in , whose coecients are determined
by the recursion formula
c
j+1
=
2(j +l + 1 n
(j + 1)(j + 2l + 2)
c
j
(242)
The polynomial v() is a function well known
to mathematicians; apart from normalization, it
can be written as
v() = L
2l+1
nl1
(2) (243)
where
L
p
qp
(x) (1)
p
_
d
dx
_
p
L
q
(x) (244)
is an associated Laguerre polynomial,
and
L
q
(x) e
x
_
d
dx
_
q
(e
x
x
q
) (245)
is the qth Laguerre polynomial, the nor-
malized hydrogen wave functions are

nlm
=

_
2
na
_
3
(n l 1)!
2n[(n +l)!]
3
e

r/na
_
2r
na
_
l
_
L
2l+1
nl1
(2r/na)

Y
m
l
(, ) (246)
The spectrum of Hydrogen
An electron may undergo a transition to
some other energy state, either by absorbing en-
ergy or emmiting energy, the dierence in energy
between the initial and nal states are
13
Y

= E
i
E
f
== 13.6eV
_
1
n
2
i

1
n
2
f
_
(247)
Now, according to the Planck formula, the
energy of a photon is proportional to its fre-
quency
E

= hv (248)
Meanwhile, the wavelength is given by =
c/v, so
1

= R
_
1
n
2
f

1
n
2
i
_
(249)
where
R
m
4c
3
_
e
2
4
0
_
2
= 1.097 10
7
m
1
(250)
is know as the Rydberg constant
Angular Momentum
Classically, the angular momentum of a par-
ticle (with respect to the origin) is given by the
formula
L = r p (251)
which is to say
L
x
= yp
z
zp
y
, L
y
= zp
x
xp
z
, L
z
= xp
y
yp
x
(252)
The corresponding quantum operators are
obtained by the standard prescription p
x

i/x etc.
Eigenvalues
The operators L
x
and L
y
do not commute; in
fact
[L
x
, L
y
] = iL
z
, [L
y
, L
z
] = iL
x
, [L
z
, L
x
] = iL
y
(253)
Notice that L
x
, L
y
, and L
z
are incompatible
observables and it would therefore be futile to
look for states that are simultaneously eigenfunc-
tions of L
x
and L
y
. On the other hand, the
square of the total angular momentum
L
2
L
2
x
+L
2
y
+L
2
z
(254)
does commute with L
x
[L
2
, L
x
] = 0, [L
2
, L
y
] = 0, [L
2
, L
z
] = 0 (255)
or, more compactly,
[L
2
, L] = 0 (256)
So L
2
is compatibale with each component of
textbfL, and we can hope to nd simultaneous
eigenstates of L
2
and (say) L
z
L
2
f = f and L
z
f = f (257)
Well use a ladder operator technique, Let
L

L
x
iL
y
(258)
The commutator with L
z
is
[L
z
, L

] = (L
x
iL
y
) (259)
so
[L
z
, L

] = L

(260)
And of course,
[L
2
, L

] = 0 (261)
I claim that if f is an eigenfunction of L
2
and
L
z
, so also is L

f equation 261 says

L
2
(L

f) = L

(L
2
f) = L

(f) = (L

f)
(262)
so if L

f is an eigenfunction of L
2
, with the
same eigenvalue , and equation 260 says
L
z
(L

f) = (L
z
L

L
z
) +L

L
z
f = L

+L

(f)
(263)
= ( )(L

f) (264)
so L

f is an eigenfunction of L
z
, with the
new eigenvalue . we call L
+
the raising
operator, because it increases the eigenvalue of
L
z
by , and L

the lowering operator, because

it lowers the eigenvalue by . When using the
raising operator we will eventually reach a state
for which the z-component exceeds the total
L

f
t
= 0 (265)
let l be thje eigenvalue of L
z
at this top rung.
L
z
f
t
= lf
t
; L
2
f
t
= f
t
(266)
14
Now,
L

= L
2
L
2
z
i(iL
z
) (267)
or putting it the other way around
L
2
= L

+L
2
z
L
z
(268)
it follows that
L
2
f
t
= (L

L
+
+L
2
z
+ L
z
)f
t
=
2
l(l + 1)f
t
(269)
and nence
=
2
l(l + 1) (270)
This tells us the eigenvalue of L
2
in terms of
the maximum eigenvalue of L
z
. Evidently the
eigenvalues of L
z
are m, where m goes from l
to l in N integer steps. In particular, it follows
that l = l+N, and hence l = N/2, so l must be
an integer or a half-integer. The eigenfunctions
are characterized by the numbers l and m
L
2
f
m
l
=
2
l(l + 1)f
m
l
; L
z
f
m
l
= mf
m
l
(271)
where
l = 0, 1/2, 1, 3/2, ...; m = l, l + 1, ..., l 1, l
(272)
Eigenfunctions
First of all we need to rewrite L
x
, L
y
, and L
z
in spherical coordinates. Now, L=(/i)(r ) ,
and the gradient, in spherical coordinates, is
= r

r
+

1
r

1
r sin

(273)
meanwhile, r=r r, so
L =

i
_
r( r r)

r
+ ( r

+ ( r

)
1
sin

_
(274)
But ( r r) = 0, ( r

) =

, and ( r

) =
and hence
L =

i
_

1
sin

_
(275)
The unit vectors

and

can be resolved into
their cartesian components

= (cos cos )

i + (cos sin)

j (sin )

k
(276)

= (sin )

i + (cos )

j (277)
Evidently
L
x
=

i
_
sin

cos cot

_
(278)
L
y
=

i
_
+cos

sin cot

_
(279)
and
L
z
=

i

(280)
using the raising and lowering operators we
nd
L
2
=
2
_
1
sin

_
sin

_
+
1
sin
2

2
_
(281)
conclusion: Spherical harmonics are eigen-
functions of L
2
and L
z
. When we solved the
Schrodinger equation by seperation of variables,
we were inadvertently constructing simultaneous
eigenfunctions of the three commuting operators
H, L
2
, and L
z
H = E, L
2
=
2
l(l + 1), L
z
= m
(282)
incidentally, we can use equation 281 to
rewrite the Schrodinger equation
1
2mr
2
_

2

r
_
r
2

r
_
+L
2
_
+V = E
(283)
Spin
In classical mechanics, a rigid object admits
two kinds of angular momentum: orbital (L =
r p), associated with the motion of the center
of mass, and spin (S = I) , associated with
motion about the center of mass. The algebraic
theory of spin is a carbon copy of the theory of
orbital angular momentum, beggining with the
fundamental commutation relations
[S
x
, S
y
] = iS
z
, [S
y
, S
z
] = iS
x
, [S
z
, S
x
] = iS
y
(284)
15
it follows that the eigenvectors of S
2
and S
z
satisfy
S
2
[sm) =
2
s(s + 1)[sm); S
z
[sm) = m[sm)
(285)
and
S

[sm) =
_
s(s + 1) m(m1)[s(m1))
(286)
where S

S
x
iS
y
. But this time the eigen-
functions are not spherical harmonics , and there
is no apriori reason to exclude the half-integer
values of s and m:
s = 0,
1
2
, 1,
3
2
, ...; m = s, s + 1, ..., s 1, s.
(287)
Pi mesons have spin 1/2; photons have spin 1;
deltas have spin 3/2; gravitons have spin 2; and
so on.
Spin 1/2
By far the most important case is s = 1/2, for
this is the spin of the particles that make up ordi-
nary matter( protons, neutrons, and electrons),
as well as quark and leptons. The general state
of a spin-1/2 particle can be expressed as a two-
element column matrix (or spinor);
=
_
a
b
_
= a
+
+b

(288)
with

+
=
_
1
0
_
(289)
representing spin up, and

=
_
0
1
_
(290)
for spin down. Meanwhile, the spin operators
become 22 matrices, which we can work out by
noting their eect on
+
and

.Equation 285
says
S
2

+
=
3
4

+
and S
2

=
3
4

(291)
we can write S
2
in matrix for as
S
2
=
3
4

2
_
1 0
0 1
_
(292)
Similarly
S
z

+
=

2

+
, S
z

(293)
from which it follows that
S
z
=

2
_
1 0
0 1
_
(294)
Meanwhile, equation 286 says
S
+

=
+
, S

+
=

, S
+
= S

= 0
(295)
so
S
+
=
_
0 1
0 0
_
, S

=
_
0 0
1 0
_
(296)
Now S

= S
x
iS
y
, so S
x
= (1/2)(S
+
+S

)
and S
y
= (1/2i)(S
+
S

), and hence
S
x
=

2
_
0 1
1 0
_
, S
y
=

2
_
0 i
i 0
_
(297)
Since S
x
, S
y
, and S
z
all carry a factor of /2,
it is tidier to write S=(/2), where

x

_
0 1
1 0
_
,
y

_
0 i
i 0
_
,
z

_
1 0
0 1
_
(298)
These are the famous Pauli spin matrices.
The eigenspinors of S
z
are (of course)

+
=
_
1
0
_
,
_
eigenvalue +

2
_
(299)

=
_
0
1
_
,
_
eigenvalue

2
_
(300)
If you measure S
z
on a particle in the general
state (equation 288), you could get +/2, with
probability [a[
2
, or /2, with probability [b[
2
.
Since they are the only possibilities
[a[
2
+[b[
2
= 1 (301)
(i.e, the spinor must be normalized). But
what if instead, you chose to measure S
x
? What
are the possible results. According to the gener-
alized statistical interpretation, we nee dto know
the eigenvalues and eigenspinors of S
x
, the char-
acteristic equation is
16

/2
/2

= 0
2
=
_

2
_
2


2
Not suprisingly, the possible values for S
x
are
the same as those for S
z
. The normalized eigen-
spinors of S
x
are

(x)
+
=
_
1

2
1

2
_
,
_
eigenvalue +

2
_
(302)

(x)

=
_
1

2
1

2
_
,
_
eigenvalue

2
_
(303)
As the eigenvectors of a hermitian matrix,
they span the space; the generic spinor (equa-
tion 288) can be expressed as a linear combina-
tion of them
=
_
a +b

2
_

(x)
+
+
_
a b

2
_

(x)

(304)
If you measure S
x
, the probability of getting
+/2 is (1/2)[a +b[
2
, and the probability of get-
ting /2 is (1/2)[a b[
2
.
Electron in a Magnetic Field
A spinning charged particle constitutes a
magnetic dipole. Its magnetic dipole mo-
ment,, is proportional to its spin angular mo-
mentum, S:
= S (305)
the proportionality constant, , is called the
gyromagnetic ratio. When a magnetic dipole
is placed in a magnetic eld B, it experiences a
tourqe, x B which tends to line it up parallel to
the eld. The energy associated with this tourqe
is
H = B (306)
so the Hamiltonian of a spinning charged par-
ticle, at rest in a magnetic eld B is
H = B S (307)
Larmor preccesion: Imagine a particle of
spin 1/2 at rest in a uniform magnetic eld,
which points in the z direction
B = B
0

k (308)
The Hamiltonian, in matrix form, is
H = B
0
S
z
=
B
0

2
_
1 0
0 1
_
(309)
The eigenstates of H are the same as those of
S
z
_

+
, with energy E
+
= (B
0
)/2

, with energy E

= +(B
0
)/2
since the Hamiltonian is time-independent,
the general solution to the time-dependent
Schrodinger equation is
i

t
= H (310)
can be expressed in terms of the stationary
states
(t) = a
+
e
iE+t/
+b

e
iEt/
=
_
ae
iB0t/2
be
iB0t/2
_
(311)
The constants a and b are determined by the
initial conditions
(0) =
_
a
b
_
(312)
With no essential loss of generality I will write
a = cos(/2) and b = sin(/2), Thus
(t) =
_
cos(/2)e
iB0t/2
sin(/2)e
iB0t/2
_
(313)
To get a feel for whats happening here, lets
calculate the expectation value of S, as a func-
tion of time
S
x
) = (t)

S
x
(t) = (cos(/2)e
iB0t/2
sin(/2)e
iB0t/2
)

2
_
0 1
1 0
__
cos(/2)e
iB0t/2
sin(/2)e
iB0t/2
_
=

2
sincos(B
0
t) (314)
similarly,
S
y
) = (t)

S
y
(t) =

2
sin sin(B
0
t)
(315)
and
S
z
) = (t)

S
y
(t) =

2
cos (316)
17
Evidently S) is tilted at a constant angle
to the z axis, and precesses about the eld at the
Larmor frequency
= B
0
(317)
Just as it would classically
Addition of Angular Momenta
Suppose now that we have twofor e spin-1/2
particlesfor example, the electron and the pro-
ton in the ground state of hydrogen. Each can
have spin up or spin down so there are four pos-
sibilities in all
, , , (318)
where the rst row refers to the electron and
the second row to the proton. Question: What
is the total angular momentum of the atom? Let
S S
(1)
+S
(2)
(319)
Each of these four composite states is an
eigenstate of S
z
the z-components simply add
S
z

2
= (S
(1)
z
+S
(2)
z
)
1

2
= (S
(1)
z
chi
1
)
2
+
1
(S
(2)
z

2
)
= (m
1

1
)
2
+
1
(m
2

2
) = (m
1
+m
2
)
1

2
(320)
So m (the quantum number of the composite
system) is just m
1
+m
2
: m = 1;
: m = 0;
: m = 0;
: m = 1
At rst galnce, this doesnt look right: m is
supposed to advance in integer steps , from s
to +s, so it appears that s = 1 but there is an
extra state with m = 0. One way to untangle
this problem is to apply the lowering operator,
S

= S
(1)

+S
(2)

to the state , using equation

296
S

() = (S
(1)

) + (S
(2)

) (321)
= ( ) + ( ) (322)
= ( + ) (323)
Evidently the three states with s = 1 are (in
the notation [s m)):
_

_
[11) =
[10) =
1

2
( + )
[1 1) =
s = 1 (triplet)
This is called the triplet combination, for the
obvious reason. Meanwhile, the orthogonal state
with m = 0 carries s = 0;
_
[00) =
1

2
( )
_
s = 0 (singlet) (324)
(if you apply the raisong or lowering operator
to this state, youll get zero
I need to prove that the triplet states are
eigenvectors of S
2
with eigenvalue 2
2
, and the
singlet is an eigenvector of S
2
with eigenvalue 0.
Now,
S
2
= (S
(1)
+S
(2)
) (S
(1)
+S
(2)
(325)
= (S
(1)
)
2
+ (S
(2)
)
2
+ 2S
(1)
S
(2)
(326)
Using equations 294 and 297, we have
S
(1)
+S
(2)
() = (S
(1)
x
)(S
(2)
x
)
+ (S
(1)
y
)(S
(2)
y
) + (S
(1)
z
)(S
(2)
z
)
=
_

2

__

2

_
+
_
i
2

__
i
2
)
_
+
_

2
)
__

2
)
_
=

2
4
(2 ) (327)
similarly
S
(1)
+S
(2)
() =

2
4
(2 ) (328)
It follows that
S
(1)
S
(2)
[10) =

2
4
1

2
(2 +2 ) =

2
4
[10)
(329)
and
S
(1)
S
(2)
[00) =

2
4
1

2
(2 2 + ) =
3
2
4
[00)
(330)
Returning to equation 325-326 ( and using
equation 291), we conclude that
18
S
2
[10) =
_
3
2
4
+
3
2
4
+ 2

2
4
_
[10) = 2
2
[10)
(331)
S
2
[00) =
_
3
2
4
+
3
2
4
2
3
2
4
_
[00) = 0 (332)
What we have just done (combining spin 1/2
with spin 1/2 to get spin 1 and spin 0) is the sim-
plest example of a larger problem: if you com-
bine spin s
1
with spin s
2
, what total spins s can
you get? The answer is that you get every spin
from (s
+
s
2
) down to (s
1
s
2
) or (s
2
s
1
), if
s
2
> s
1
in integer steps
s = (s
1
+s
2
), (s
1
+s
2
1), (s
1
+s
2
2), ..., [s
1
s
2
[
(333)
The combined state [sm) with total spin s and
z-component m will be some linear combination
of the composite states [s
1
m
1
)[s
2
m
2
)
[sm) =

m1+m2=m
C
s1s2s
m1m2m
[s
1
m
1
)[s
2
m
2
) (334)
(because the z components add, the only com-
posite states that contribute are those for which
m
1
+ m
2
= m). The triplet combination and
the singlet are special cases of this general form,
with s
1
= s
2
= 1/2 I used the notation = [
1
2
1
2
),
= [
1
2
(
1
2
)). The constants C
s1s2s
m1m2m
are called
Clebsch-Gordan coecients.
Chapter 5: Identical Particles
Two-Particle System
The state of a two-particle system is a func-
tion of the coordinates of particle one (r
1
), the
coordinates of particle two (r
2
), and the time
(r
1
, r
2
, t) (335)
Its time evolution is deterimed (as always) by
the Schrodinger equation:
i

t
= H (336)
Where H is the Hamiltonian for the whole
system
H =

2
2m
1

2
1

2
2m
2

2
2
+V (r
1
, r
2
, t) (337)
The statistical interpretation carries over in
the obvious way
[(r
1
, r
2
, t)[
2
d
3
r
1
d
3
r
2
(338)
is the probability of nding particle 1 in the
volume d
3
r
1
and particle 2 in the volume d
3
r
2
;
evidently must be normalized in such a way
that
_
[(r
1
, r
2
, t)[
2
d
3
r
1
d
3
r
2
= 1 (339)
For time-independent potentials, we obtain a
complete set of solutions by seperation of vari-
ables
(r
1
, r
2
, t) = (r
1
, r
2
)e
iEt/
(340)
where the spatial wave function () satises
the time-independent Schrodinger equation

2
2m
1

2
1


2
2m
2

2
2
+V = E (341)
and E is the total energy of the system
Bosons and Fermions
Suposse particle 1 is in the (one-particle) state

a
(r), and particle 2 is in the state
b
(r). (ignor-
ing spin at the moment). In that case (r
1
, r
2
)
is a simple product
(r
1
, r
2
) =
a
(r)
b
(r) (342)
Quantum mechanics neatly accomidates the
existance of particles that are indistinguishable
in principle: We simply construct a wave func-
tion that is non committal as to which particle
is in which state. there are actually two ways to
do it

(r
1
, r
2
) = A[
a
(r)
b
(r)
b
(r)
a
(r)] (343)
Thus the theory admits to kinds of identical
particles: bosons, for which we use the plus
sign , and fermions, for which we use the minus
sign. Photons and mesons are bosons; protons
and electrons are fermions. It so happebns that
_
all particles with integer spin are bosons, and
all particles with half-integer spin are fermions
It follows, in paticular, that two identical
fermions cannot occupy the same state. For if

a
=
b
, then
19

(r
1
, r
2
) = A[
a
(r)
a
(r)
a
(r)
a
(r)] = 0
(344)
This is the famous Pauli exclusion princi-
ple. I assumed, for the sake of argument, that
one particle was in the state
a
and the other in
state
b
, but there is a more general way to for-
mulate the problem. Let us dene the exchange
operator, P, which interchanges the two parti-
cles
Pf(r
1
, r
2
) = f(r
1
, r
2
) (345)
Clearly P
2
= 1, and it follows that the eigen-
values of P are 1. Now, if the two particles are
identical, the Hamiltonian must treat them the
same: m
1
= m
2
and V (r
1
, r
2
) = (r
2
, r
1
) . It fol-
lows that P and H are compatible observables,
[P, H] = 0 (346)
and hence we can nd a complete set of func-
tions that are simultaneously eigenstates of both.
That is to say, we can nd solutions to the
Schrodinger equation that are either symmetric
(eigenvalue + 1) or antisymmetric (eigenvalue -
1) under exchange
(r
1
, r
2
) = (r
1
, r
2
) (347)
The symmetrization requirement states
that, for identical particles, the wave function is
not merely alllowed, but required to satisfy equa-
tion 347. This is the general statement, of which
equation 343 is a special case.
Exchange Forces
To give some sense of what the symmetriza-
tion requirement does, we can suppose that one
particle is in state
a
(x), and the other is in
state
b
(x), and these two states are orthogonal
and normalized. If the two particles are distin-
guishable, and number 1 is in state
a
, then the
combined wave function is
(x
1
, x
2
) =
a
(x
1
)
b
(x
2
) (348)
if they are identical bosons, the composite
wave function is

+
(x
1
, x
2
) =
1

2
[
a
(x
1
)
b
(x
2
) +
b
(x
1
)
a
(x
2
)]
(349)
and if they are identical fermions

(x
1
, x
2
) =
1

2
[
a
(x
1
)
b
(x
2
)
b
(x
1
)
a
(x
2
)]
(350)
Lets calculate the expectation value of the
square of the separation distance between the
two particles
(x
1
x
2
)
2
) = x
2
1
) +x
2
2
) 2x
1
x
2
) (351)
Case 1: Distinguishable particles. For
the wave function in equation 348
x
2
1
) =
_
x
2
1
[
a
(x
1
)[
2
dx
1
_
[
b
(x
2
)[
2
dx
2
= x
2
)
a
(352)
and
x
2
2
) =
_
[
a
(x
1
)[
2
dx
1
_
x
2
2
[
b
(x
2
)[
2
dx
2
= x
2
)
b
(353)
and
x
1
x
2
) =
_
x
1
[
a
(x
1
)[
2
dx
1
_
x
2
[
b
(x
2
)[
2
dx
2
= x)
a
x)
b
(354)
in this case then
(x
1
x
2
)
2
)
d
= x
2
)
a
+x
2
)
b
2x)
a
x)
b
(355)
Case 2: Identical particles. For the wave
function in equations 349 and 350
x
2
1
) =
1
2
(
_
x
2
1
[
a
(x
1
)[
2
dx
1
_
[
b
(x
2
)[
2
dx
2
+
_
x
2
1
[
b
(x
1
)[
2
dx
1
_
[
a
(x
2
)[
2
dx
2

_
x
2
1

a
(x
1
)

b
(x
1
)dx
1
_

b
(x
2
)

a
(x
2
)dx
2

_
x
2
1

b
(x
1
)

a
(x
1
)dx
1
_

a
(x
2
)

b
(x
2
)dx
2
)
=
1
2
_
x
2
)
a
+x
2
)
b
0 0

=
1
2
_
x
2
)
a
+x
2
)
b
_
(356)
similarly
x
2
2
) =
1
2
_
x
2
)
b
+x
2
)
a
_
(357)
20
(naturally, x
2
2
) = x
2
1
), since you cant tell
them apart.) But
x
1
x
2
) =
1
2
(
_
x
1
[
a
(x
1
)[
2
dx
1
_
x
2
[
b
(x
2
)[
2
dx
2
+
_
x
1
[
b
(x
1
)[
2
dx
1
_
x
2
[
a
(x
2
)[
2
dx
2

_
x
1

a
(x
1
)

b
(x
1
)dx
1
_
x
2

b
(x
2
)

a
(x
2
)dx
2

_
x
1

b
(x
1
)

a
(x
1
)dx
1
_
x
2

a
(x
2
)

b
(x
2
)dx
2
)
=
1
2
(x)
a
x)
b
+x)
b
x)
a
x)
ab
x)
ba
x)
ba
x)
ab
)
= x)
a
x)
b
[x)
ab
[
2
(358)
where
psi
a
(x
1
)[
2
dx
1
_
x
2
2
[
b
(x
2
)[
2
dx
2
= x
2
)
b
(359)
and
x
1
x
2
) =
_
x
1
[
a
(x
1
)[
2
dx
1
_
x
2
[
b
(x
2
)[
2
dx
2
= x)
a
x)
b
(360)
in this case then
(x
1
x
2
)
2
)
d
= x
2
)
a
+x
2
)
b
2x)
a
x)
b
(361)
Case 2: Identical particles. For the wave
function in equations 349 and 350
x
2
1
) =
1
2
(
_
x
2
1
[
a
(x
1
)[
2
dx
1
_
[
b
(x
2
)[
2
dx
2
+
_
x
2
1
[
b
(x
1
)[
2
dx
1
_
[
a
(x
2
)[
2
dx
2

_
x
2
1

a
(x
1
)

b
(x
1
)dx
1
_

b
(x
2
)

a
(x
2
)dx
2

_
x
2
1

b
(x
1
)

a
(x
1
)dx
1
_

a
(x
2
)

b
(x
2
)dx
2
)
=
1
2
_
x
2
)
a
+x
2
)
b
0 0

=
1
2
_
x
2
)
a
+x
2
)
b
_
(362)
similarly
x
2
2
) =
1
2
_
x
2
)
b
+x
2
)
a
_
(363)
(naturally, x
2
2
) = x
2
1
), since you cant tell
them apart.) But
x
1
x
2
) =
1
2
(
_
x
1
[
a
(x
1
)[
2
dx
1
_
x
2
[
b
(x
2
)[
2
dx
2
+
_
x
1
[
b
(x
1
)[
2
dx
1
_
x
2
[
a
(x
2
)[
2
dx
2

_
x
1

a
(x
1
)

b
(x
1
)dx
1
_
x
2

b
(x
2
)

a
(x
2
)dx
2

_
x
1

b
(x
1
)

a
(x
1
)dx
1
_
x
2

a
(x
2
)

b
(x
2
)dx
2
)
=
1
2
(x)
a
x)
b
+x)
b
x)
a
x)
ab
x)
ba
x)
ba
x)
ab
)
= x)
a
x)
b
[x)
ab
[
2
(364)
where
x)
ab

_
x
a
(x)

b
(x)dx (365)
Evidently
(x
1
x
2
)
2
)

= x
2
)
a
+x
2
)
b
2x)
a
x)
b
2[x)
ab
[
2
(366)
comparing equations 355 and 360, we see that
the dierence resides in the nal term
(x)
2
) = (x)
2
)
d
2[x)
ab
[
2
(367)
Atoms
A neutral atom, of atomic number Z, consists
of a heavy nucleus, with electric charge Ze, sor-
rounded by Z electrons (mass m and charge e).
The Hamiltonian for this system is
H =
Z

j=1
_

2
2m

2
j

_
1
4
0
_
Ze
2
r
j
_
+
1
2
_
1
4
0
_
Z

j=k
e
2
[r
j
r
k
[
(368)
The term in the curly brackets represents the
kinetic plus potential of the j
th
electron, in the
electric eld of the nucleus; the secon sum (which
runs over all values of j and k except j = k) is the
potential associated with the multual repulsion
of the electrons. (the factor of 1/2 in front cor-
responds that the summation counts each pair
twice. The problem is to solve Schrodingers
equation
21
H = E (369)
for the wave function (r
1
, r
2
, ...r
Z
). Because
electrons are identical fermions, however, not all
solutions are acceptable;only those for which the
complete state (position and spin)
(r
1
, r
2
, ...r
Z
)(s
1
, s
2
, ...s
Z
) (370)
is antisymmetric with respect to interchange
of any two electrons. In particular, no two elec-
trons can occupy the same state
Helium
After hydrogen, the simplest atom is helium
(Z = 2). The Hamiltonian,
H =
_

2
2m

2
1

1
4
0
2e
2
r
1
_
+
_

2
2m

2
2

1
4
0
2e
2
r
2
_
+
1
4
0
e
2
[r
1
r
2
[
(371)
consists of two hydrogenic Hamiltonians (with
nuclear charge 2e), one for each electron 1 and 2,
together with the nal term describing the repul-
sion of the two electrons. It is this last term that
causes all the the trouble. If we simply ignore it,
the Schrodinger equation sperates, and the so-
lutions can be written as products of hydrogen
wave functions
(r
1
, r
2
) =
nlm
(r
1
)
n

m
(r
2
) (372)
only with half the Bohr radius and four times
the Bohr energies. The total energy would be
E = 4(E
n
+E
n
) (373)
where E
n
= 13.6/n
2
eV. In particular, the
ground state would be

0
(r
1
, r
2
) =
100
(r
1
)
100
(r
2
) =
8
a
3
e
2(r1+r2)/a
(374)
and its energy would be
E
0
= 8(13.6 eV) = 109 eV (375)
Solids
The Free Electron Gas
Suppose the object in question is a rectangu-
lar solid, with dimensions l
x
, l
y
, l
z
, and imagine
that an electron inside experiences no force at
all, except at the impenetrable walls
V (x, y, z) =
_
0, if0 < x < l
x
, 0 < y < l
y
, and0 < z < l
z
, otherwise
The Schrodinger equation,

2
2m

2
= E (376)
separates, in cartesian coordinates: (x, y, z) =
X(x)Y (y)Z(z), with

2
2m
dX
2
dx
2
= E
x
X (377)

2
2m
dY
2
dy
2
= E
y
Y (378)

2
2m
dZ
2
dz
2
= E
z
Z (379)
and E = E
x
+E
y
+E
z
. Letting
k
x

2mE
x

, k
y

_
2mE
y

, k
z

2mE
z

(380)
we obtain the general solutions
X(x) = A
x
sin(k
x
x) +B
x
cos(k
x
x) (381)
Y (y) = A
y
sin(k
y
y) +B
x
cos(k
y
y) (382)
Z(z) = A
z
sin(k
z
z) +B
x
cos(k
z
z) (383)
(384)
The boundery conditions require X(0) =
Y (0) = Z(0) = 0, so B
x
= B
y
= B + z = 0,
and X(l
x
) = Y (l
y
) = Z(l
z
) = 0, so that
k
x
l
x
= n
x
, k
y
l
y
= n
y
, k
z
l
z
= n
z
(385)
where each n is a positive integer
n
x
= 1, 2, 3, .., n
y
= 1, 2, 3, .., n
z
= 1, 2, 3, ..
(386)
The normalized wave functions are

nxnynz
=

8
l
x
l
y
l
z
sin
_
n
x

l
x
x
_
sin
_
n
y

l
y
y
_
sin
_
n
z

l
z
z
_
(387)
22
x)
ab

_
x
a
(x)

b
(x)dx (388)
Evidently
(x
1
x
2
)
2
)

= x
2
)
a
+x
2
)
b
2x)
a
x)
b
2[x)
ab
[
2
(389)
comparing equations 355 and 360, we see that
the dierence resides in the nal term
(x)
2
) = (x)
2
)
d
2[x)
ab
[
2
(390)
Atoms
A neutral atom, of atomic number Z, consists
of a heavy nucleus, with electric charge Ze, sor-
rounded by Z electrons (mass m and charge e).
The Hamiltonian for this system is
H =
Z

j=1
_

2
2m

2
j

_
1
4
0
_
Ze
2
r
j
_
+
1
2
_
1
4
0
_
Z

j=k
e
2
[r
j
r
k
[
(391)
The term in the curly brackets represents the
kinetic plus potential of the j
th
electron, in the
electric eld of the nucleus; the secon sum (which
runs over all values of j and k except j = k) is the
potential associated with the multual repulsion
of the electrons. (the factor of 1/2 in front cor-
responds that the summation counts each pair
twice. The problem is to solve Schrodingers
equation
H = E (392)
for the wave function (r
1
, r
2
, ...r
Z
). Because
electrons are identical fermions, however, not all
solutions are acceptable;only those for which the
complete state (position and spin)
(r
1
, r
2
, ...r
Z
)(s
1
, s
2
, ...s
Z
) (393)
is antisymmetric with respect to interchange
of any two electrons. In particular, no two elec-
trons can occupy the same state
Helium
After hydrogen, the simplest atom is helium
(Z = 2). The Hamiltonian,
H =
_

2
2m

2
1

1
4
0
2e
2
r
1
_
+
_

2
2m

2
2

1
4
0
2e
2
r
2
_
+
1
4
0
e
2
[r
1
r
2
[
(394)
consists of two hydrogenic Hamiltonians (with
nuclear charge 2e), one for each electron 1 and 2,
together with the nal term describing the repul-
sion of the two electrons. It is this last term that
causes all the the trouble. If we simply ignore it,
the Schrodinger equation sperates, and the so-
lutions can be written as products of hydrogen
wave functions
(r
1
, r
2
) =
nlm
(r
1
)
n

m
(r
2
) (395)
only with half the Bohr radius and four times
the Bohr energies. The total energy would be
E = 4(E
n
+E
n
) (396)
where E
n
= 13.6/n
2
eV. In particular, the
ground state would be

0
(r
1
, r
2
) =
100
(r
1
)
100
(r
2
) =
8
a
3
e
2(r1+r2)/a
(397)
and its energy would be
E
0
= 8(13.6 eV) = 109 eV (398)
Solids
The Free Electron Gas
Suppose the object in question is a rectangu-
lar solid, with dimensions l
x
, l
y
, l
z
, and imagine
that an electron inside experiences no force at
all, except at the impenetrable walls
V (x, y, z) =
_
0, if 0 < x < l
x
, 0 < y < l
y
, and 0 < z < l
z
, otherwise
The Schrodinger equation,

2
2m

2
= E (399)
separates, in cartesian coordinates: (x, y, z) =
X(x)Y (y)Z(z), with
23

2
2m
dX
2
dx
2
= E
x
X (400)

2
2m
dY
2
dy
2
= E
y
Y (401)

2
2m
dZ
2
dz
2
= E
z
Z (402)
and E = E
x
+E
y
+E
z
. Letting
k
x

2mE
x

, k
y

_
2mE
y

, k
z

2mE
z

(403)
we obtain the general solutions
X(x) = A
x
sin(k
x
x) +B
x
cos(k
x
x) (404)
Y (y) = A
y
sin(k
y
y) +B
x
cos(k
y
y) (405)
Z(z) = A
z
sin(k
z
z) +B
x
cos(k
z
z) (406)
The boundery conditions require X(0) =
Y (0) = Z(0) = 0, so B
x
= B
y
= B + z = 0,
and X(l
x
) = Y (l
y
) = Z(l
z
) = 0, so that
k
x
l
x
= n
x
, k
y
l
y
= n
y
, k
z
l
z
= n
z
(407)
where each n is a positive integer
n
x
= 1, 2, 3, .., n
y
= 1, 2, 3, .., n
z
= 1, 2, 3, ..
(408)
The normalized wave functions are

nxnynz
=

8
l
x
l
y
l
z
sin
_
n
x

l
x
x
_
sin
_
n
y

l
y
y
_
sin
_
n
z

l
z
z
_
(409)
and the allowed energies are
E
nxnynz
=

2

2
2m
_
n
2
x
l
2
x
+
n
2
x
l
2
x
+
n
2
x
l
2
x
_
=

2
k
2
2m
(410)
where k is the magnitude of the wave vector,
k (k
x
, k
y
, k
z
).
If you imagine a three-dimensional space,
with axes k
x
, k
y
, k
z
, and planes drawn in at k
x
=
(/l
x
), (2/l
x
), (3/l
x
), ..., at k
y
= (/l
y
), (2/l
y
), (3/l
y
), ...,
and at k
z
= (/l
z
), (2/l
z
), (3/l
z
), .... Each
block in this grid, and hence each state, occu-
pies a volume

3
l
x
l
y
l
z
=

3
V
(411)
of k-space, where V = l
x
l
y
l
z
is the volume of
the object itself. Suppose the sample contains N
atoms , and each contribute q electrons. They
will ll up one octant of a sphere in k-space,
whose radius ,k
F
is determined by the fact that
each pair of electrons require a volume
3
/V
1
8
_
4
3
k
3
F
_
=
Nq
2
_

3
V
_
(412)
thus
k
F
= (3
2
)
1/3
(413)
where

Nq
V
(414)
is the free electron density ( the number of
free electrons per unti volume)
The boundery seperating occupied and unoc-
cupied states, in k-space, is called the Fermi
surface. The corresponding energy is called the
Fermi energy, E
F
; for a free electron gas,
E
F
=

2
2m
(3
2
)
2/3
(415)
The total energy of the electron gas can be
calculated as follows; Each of these states carries
an energy
2
k
2
/2m, so the energy of the shell is
dE =

2
k
2
2m
V

2
k
2
dk (416)
and hence the total energy is
E
tot
=

2
V
2m
2
_
kF
0
k
4
dk =

2
k
5
F
V
10
2
m
=

2
(3
2
Nq)
5/3
10
2
m
V
2/3
(417)
This quantum mechanical energy plays a role
rather analogous to the internal thermal energy
(U) of an ordinary gas. In particular, it exerts a
pressure on the walls, for if the box expands by
an amount dV , the total energy decreases
dE
tot
=
2
3

2
(3
2
Nq)
5/3
10
2
m
V
5/3
dV
=
2
3
E
tot
dV
V
(418)
24
and this shows up as work done on the out-
side (dW = PdV ) by the quantum pressure P.
Evidently
P =
2
3
E
tot
V
=
2
3

2
k
5
F
10
2
m
=
(3
2
)
2/3

2
5m

5/3
(419)
this is sometimes called degeneracy pres-
sure strictly due to quantum eects.
25