129A Lecture Notes

Strong Interactions I

1 Four Forces
Becquerel discovered radioactivity at the end of 19th century, and Rutherford
classified different types according to the charge of the emitted particles. -
ray has charge 2, -ray 1, and -ray zero. In todays terminology, the
-ray is the emission of 4 He nucleus (two protons and two nucleons), -ray
is an electron, and -ray a very energetic photon. This was the first point in
the history when three types of interactions manifested themselves: strong,
weak, and electomagnetic.
The -ray is emitted then the nucleus is in an excited state, and decays
into a lower state. This process is very similar to the emission of photons
from excited states of an atom.
The -ray is emitted from large nuclei when it is energetically favored to
release -particle to reduce its Coulomb repulsion within the nucleus. The
parent nucleus of atomic number Z and mass number A changes to Z 2 and
A 4. The dynamics of this decay is quite complicated, but the lifetime is
determined primarily by the Gamov factor, the tunneling probability for the
-particle to go through the Coulomb barrier. The fact that the positively
charged particles can be packed inside small nuclei is already quite puzzling;
that was the original puzzle about the nuclear strong interactions.
The -ray is the only interaction in nuclei that changes the number of
protons and neutrons. The parent nucleus of atomic number Z and mass
number A becomes Z + 1 and A. The net effect is to turn one neutron into a
proton, by the emission of an electron. This is the manifestation of the weak
interactions.
Of course there is another important force in nature: gravity. It plays lit-
tle role in the world of microscopic particles, as its effect is always suppressed
by the Newtons constant GN = (1019 GeV)2 in the natural unit. But it
is important at macroscopic distances and governs the motion of heavenly
bodies as you know well.
In the rest of the lecture notes, we will discuss the strong interactions.

1
2 Proton statistics
How do we know that protons obey Fermi-Dirac statistics? We of course
know that because of the spin-statistic theorem, but this theorem needed to
be established experimentally anyway. We need to know that the proton is
a fermion independent of its spin and the spin-statistics theorem.
For that purpose, we consider molecular band spectrum. A molecular
band spectrum is what appears in the emission lines from a gas of molecules
mostly from vibrational spectra (infrared), but the lines appear to be a
band, i.e. a thick line. Looking more closely, the thick line actually consists
of many many fine lines, which come from rotatinal deexcitations.
Generally, a diatomic molecule has a rotational spectrum due to the rigid
body Hamiltonian
~2
L 2 l(l + 1)
h
H= = . (1)
2I 2I
Here it is assumed that the molecule has a dumb-bell shape and can rotate
in two possible modes.
In the case of hydrogen molecules H2 , two atoms are bosons because
they consist of two fermions (one electron and one proton). Therefore the
total wave function must be symmetric under the exchange of two hydrogen
atoms. A part of the wave function comes from the spin degrees of freedom
of two protons. Depending on S = 0 or S = 1 for two proton spins, the spin
part of the wave function is either anti-symmetric or symmetric. Everything
else being the same between two hydrogen atoms, the anti-symmetry of the
S = 0 spin wave function must be compensated by the rotational wave
function. Using the relative coordinate ~r = ~x1 ~x2 between two protons,
the interchange of two protons will flip the sign of ~r ~r. The rotational
wave function is nothing but spherical harmonics Ylm (~r), which satisfies the
property Ylm (~r) = (1)l (~r). Therefore the interchange of two protons
would result in a sign factor (1)l from the rotational wave function. In
order to compensate the unwanted minus sign for S = 0 case, we need to
take l odd. On ther other hand, for S = 1 case, we need to take l even to
keep the wave function symmetric.
Transitions among rotational levels take place only between the same S,
because the nuclear magneton is too small to cause spin flips in the tran-
sitions. Therefore rotational spectra appear from transitions among even l
states or odd l states, but not among odd and even ls. Because the tran-
sitions are most frequent between two nearest states, the S = 0 case causes

2
spectra for l = 3 to l = 1, l = 5 to l = 3, and so on, and hence

2
h 2
h 2
h
E = ((2n+1)(2n+2)(2n1)2n) = (8n+2) = (10, 18, 26, 34, ).
2I 2I 2I
(2)
On the other hand, the S = 1 case causes spectra

2
h 2
h 2
h
E = ((2n + 2)(2n + 3) 2n(2n + 1)) = (8n + 6) = (6, 14, 22, 30, ).
2I 2I 2I
(3)
Having two series of spectra so far does not prove the statistics. But the point
is that the first spectrum is for S = 0, and hence only one spin orientation,
but the second one is for S = 1, and hence for three possible spin orienta-
tions. In other words, the lines for the second set of spectrum must be three
times stronger than the first set. If protons followed Bose-Einstein statistics
instead, the wave function must changes it sign under the interchange of two
atoms, and hence the S = 0 combination should give the second set of spec-
tra while the S = 1 the first set, and the relative strengh between two sets
reverses.
This method applies in general to any same-nucleus diatomic molecules.
In fact, the statistics of nitrogen nuclei 14 N was determined to be Bose-
Einstein from this type of measurements, and causes a great puzzle. In those
days, people thought that the nuclei consist of protons and electrons, and
hence the nitrogen nucleus was believed to be 14p+7e , and hence a fermion.
This discrepancy (statistics puzzle), together with a continuous spectrum of
nuclear -decay, let Pauli to speculate the existence of neutrinos.

3 Neutron
Rutherford apparently once speculated that neutron exists, but I do not
know the basis of the speculation. The neutron was discovered by James
Chadwick. For details of the discovery, read Chapter 1 of CahnGoldhaber.
The discovery of neutron immediately solves the statistics puzzle. Now
14
N can be understood as made up of seven protons and seven neutrons,
naturally a boson. Of course it then leaves the question why an electron can
be emitted from the nucleus in a nuclear -decay.

3
4 Nuclei
Nuclei sit at the center of any atoms. Therefore, understanding them is of
central importance to any discussions of microscopic physics. Due to some
reason, however, the nuclear physics had not been taught so much in the
standard physics curriculum. I try to briefly review nuclear physics in a few
lectures. Obviously I cant go into much details, but hope to give you at
least a rough idea on nuclear physics.
As you know, nuclei are composed of protons and neutrons. The number
of protons is the atomic number Z, and the mass number A is approximately
the total number of nucleons, a collective name for protons and neutrons.
Therefore
A=N +Z (4)
where N is the number of neutrons. We know that nuclei are very small. An
empirical formula for the size of the nuclei, which can be measured using the
form factor in elastic electron-nuclei scattering, is

R = r0 A1/3 , r0 = 1.12 fm. (5)

This is a good approximation practically for all nuclei with A > 12. Here,
fm = 1013 cm, or sometimes called also Fermi rather than femto-meter,
and the nuclei are smaller by five orders of magnitude than the atoms. What
the formula means is that the nuclear density is more-or-less constant for any
nuclei, = 1.72 1038 nucleons/cm3 = 0.172 nucleons/fm3 . Of course, the
nuclear density does not drop to zero abruptly. The form factor measurement
is often fitted to the size and the surface thickness, within which the density
smoothly falls from the constant to zero. The result is that the surface
thickness is about t ' 2.4 fm.

4.1 Empirical Mass Formula

Gross properties of nuclei are manifested in the empirical (or Weizsacker)
mass formula. Recall Einsteins relation E = mc2 , which tells us that the to-
tal mass of nuclei has information on its composition as well as its interaction
energies. The empirical mass formula is
B
mnucleus (Z, N ) = Zmp + N mn , (6)
c2

4
Figure 1: From Theoretical Nuclear Physics, by Amos deShalit and Her-
man Feshbach, New York, Wiley, 1974.

5
Figure 2: From Theoretical Nuclear Physics, by Amos deShalit and Her-
man Feshbach, New York, Wiley, 1974.

6
Figure 3: A more realistic shape of nuclei. From Subatomic Zoo, by Hans
Frauenfelder and Ernest M. Henley, Prentice-Hall, Inc., NJ, 1974.

7
Figure 4: Nuclear binding energy is more-or-less independent of its size,
roughly about 8.5 MeV/nucleon. The first few peaks are for 4 He, 12 C, 16 O.
The maximum is for 56 Fe. From Theoretical Nuclear Physics, by Amos
deShalit and Herman Feshbach, New York, Wiley, 1974.

where the last term is the mass deficit due to the binding energy B, and
is given by
(Z N )2 Z2
B = av A as A2/3 asym aC 1/3 + (A). (7)
A A
Among all these terms, the first terms is the most important one, giving
roughly constant binding energy per nucleon. If you neglect all the other
terms, the binding energy is roughly 8.5 MeV/nucleon. But if you fit the
data with all other terms, the number of course comes out differently. We
discuss each of the terms below.
The first term is called the volume term with av = 15.68 MeV, represent-
ing that the total binding energy is roughly proportional to the number of

8
nucleons. This is the dominant term in the formula. Other terms show the
variation of the binding energy as a function of N and Z. The second term is
called the surface term with as = 18.56 MeV, representing that the binding
energy is lost somehow proportional to the surface area. These two terms
can be qualitatively explained by the so-called liquid drop model of nuclei.
You can view a nucleus as a tightly packed drop of nucleons, each feeling
attractive force from its neighbors. The point is that the force comes basi-
cally only from its neighbors due to the short-ranged nature of the nuclear
force responsible for binding nuclei. Because the number of neighbors is
basically the same for any nucleon given the constant nuclear density weve
seen above, the amount of binding energy is proportional to the number of
nucleons, giving rise to the volume term. This is said to be saturation of
nuclear binding, and the nucleons basically dont see nucleons beyond their
neighbors. But those at the surface receive less binding because they do not
have about a half of neighbors. The loss of the binding energy is given by
the surface term.
The symmetry term is less obvious. The empirical fact is that stable nuclei
require more-or-less the same number of protons and neutrons, especially true
for light nuclei. Think about common nuclides: 4 He, 12 C, 14 N, 16 O, etc, with
high natural abundances. This point will be understood in terms of Fermi
gas model of nuclei. By putting in neutrons and protons as free particles
in a Fermi-degenerate gas, protons and neutrons fill up levels independently,
and it is energetically favorable to keep the Fermi energies for protons and
neutrons the same for a given total number of nucleons (mass number). The
symmetry term, with asym = 28.1 MeV, reflects the rise in the energy when
they are not equal with a parabolic approximation around the minimum
Z = N.
The Coulomb term has the obvious meaning of total Coulomb energy
among protons (neutrons are electrically neutral!). Because the number of
protons is Z, and there is Coulomb potential between any pairs of protons
(long-ranged force unlike the nuclear binding force), the energy goes as Z 2 .
The typical distance among them is the nuclear size, given by A1/3 , hence
the dependence Z 2 /A1/3 , with the coefficient aC = 0.717 MeV. It shows that
the Coulomb interaction is actually a very weak interaction compared to
the nuclear force. Of course, the actual size of the Coulomb energy can be
important especially for large nuclei, because it grows like Z 2 (even if you
scale Z and A together, it grows as Z 5/3 ). This tends to prefer smaller Z for
a given A. The competition of the Coulomb term and the symmetry term

9
gives a preferred fraction of protons for a given A, which becomes smaller and
smaller as A increases, consistent with the observed band of stable isotopes.
Finally the last term is called the pairing term. There is a tendency
that nucleons want to be paired between a given state and its time-reversed
state, i.e., the opposite orbital and spin angular momenta. Because of this
property, even-even nuclei (nuclei with even number of protons and even
number of neutrons) have all 0+ ground state. There is a sizable difference
in the binding energies between nuclei with all nucleons paired (even-even
ones) and those with some nuclei unpaired (even-odd, odd-even, and odd-
odd). The pairing term represents the energy difference among them, given
by
3/4

34A
MeV for odd-odd nuclei
(A) = 0 MeV for odd-even nuclei . (8)
3/4


34A MeV for even-even nuclei
Looking at the plot Fig. (4), there are a few anomalously high binding
energies for low A. They are 4 He, 12 C, 16 O. The maximum is for 56 Fe. The
presence of a maximum means that any thermonuclear fusion process, such
as stellar burning, cannot produce nuclei beyond iron. We will come back
to the question how different nuclei and hence elements were born in our
Universe later.
Fig. 5 shows observed nuclides. For small numbers of nucleons, the band
is nearly diagonal, i.e., Z N . As the size grows, the band bends and is
below the diagonal, Z < N . Using the empirical mass formula, the existence
of the band is easy to understand. As you go away from the band, the
symmetry term becomes important and the mass of the nucleus grows. What
it means it that such a nuclide, if exists, decays immediately by ejecting excess
neutrons or protons until the symmetry term becomes small enough to make
it energetically impossible to eject free neutrons or protons. For large nuclei,
Coulomb term is important and smaller number of protons is preferred. That
is why the band bends downwards. Even within the band, the number of
stable nuclides is not so large. All the colored ones decay either by -decay
(N, Z) (N 1, Z + 1)e e or anti--decay (N, Z) (N + 1, Z 1)e+ e
to approach the narrow band of stability moving along 45 line. Unstable
nuclei can also emit an -particle, a unusually tightly bound 4 He nucleus, to
lower the mass number, approaching the maximum binding energy of A = 56.

10
Figure 5: Table of nuclear isotopes, from http://www2.bnl.gov/CoN/. The
horizontal axis is for the number of neutrons N , while the vertical axis the
protons (i.e., atomic number) Z. The black squares represent stable isotopes,
while the others decay either by - or -decays to more stable nuclei. Double
lines are for magic numbers.

11
4.2 Nuclear Force
Protons and neutrons are bound inside nuclei, despite the Coulomb repulsion
among protons. Therefore there must be a different and much stronger force
acting among nucleons to bind them together. This force is called nuclear
force, nuclear binding force, or in more modern settings, the strong interac-
tion. (Here, we are not talking about a strong interaction. This is the name
of the force.) Here are notable properties of the nuclear binding force.

1. It is much stronger than the electromagnetic force. In the empirical

mass formula, we saw that the coefficient of the Coulomb term is more
than an order of magnitude smaller than the other terms in the binding
energy.

2. It is an attractive force, otherwise nucleons wouldnt bind.

3. It is short-ranged, acts only up to 12 fm.

4. It has the saturation property, giving nearly constant B/A ' 8.5 MeV.
This is in stark contrast to the electromagnetic force. For instance,
the ThomasFermi model of atoms gives B = 15.73Z 7/3 eV that grows
with a very high power in the number of particles.

5. The force depends on spin and charge states of the nucleon. To under-
stand nuclei and nucleon-nucleon scattering data, we need not only a
potential V (r) between nucleons in the Hamiltonian but also the spin-
~ (r), and the
spin term ~1 ~2 V (r), the spin-orbit term (~1 + ~2 ) LV
2
tensor term [3(~1 ~r)(~2 ~r) r ~1 ~2 ]V (r).

6. It can exchange charge. If you do neutron-proton scattering experi-

ment, you not only see a forward peak but also a backward peak. Note
that a forward peak is analogous to a large impact parameter in the
classical mechanics where there is little deflection (recall Rutherford
scattering), and exists for pretty much any scattering processes. But
a backward peak is quite unusual. The interpretation is that when
the proton appears to be backscattered, it is actually a neutron which
converted to a proton because of the nuclear reaction. In other words,
the neutron is scattered to the forward angle, but has converted to
proton by the scattering and we are fooled to see the proton scattered
backward. This is the charge-exchange reaction.

12
 7. Even though the nuclear force is attractive to bind nucleons, there is
a repulsive core when they approach too closely, around 0.5 fm. They
basically cannot go closer.
8. The nuclear force has charge symmetry, which means that we can
make an overall switch between protons and neutrons without chang-
ing forces among them. For instance, nn and pp scattering are the
same (except for the obvious difference due to the electric charge). For
example, mirror nuclei, which are related by switching protons and
neutrons, have very similar excitation spectra. Examples include 13 C
and 13 N, 17 O and 17 F, etc.
9. A stronger version of the charge symmetry is charge independence.
Not only nn and pp scattering are the same, but also np scattering is
also the same under the same configuration which I specify below
using the concept of isospin.

The last item needs some more explanations. There is a new symmetry
in the nuclear force called isospin, proposed originally by Heisenberg. The
idea is very simple: regard protons and neutrons as identical particles. But of
course, you cant; they are different particles, right? They even have different
masses! Well, the trick is to introduce a new quantum number, isospin, which
takes values +1/2 and 1/2 just like the ordinary spin. We say a proton
is a nucleon with Iz = +1/2, while a neutron with Iz = 1/2. At this
point, it is just semantics. But the important statement is this: the nuclear
force is invariant under the isospin rotation, just like the Hamiltonian of a
ferromagnet is invariant under the rotation of spin. Then you can classify
states according to the isospin quantum numbers because the nuclear force
preserves isospin. But what about the mass difference, then? The point is
that their masses are actually quite similar: mp = 938.3 MeV/c2 and mn =
939.6 MeV/c2 . To the extent that we ignore the small mass difference, we
can treat them identical. Another question is the obvious difference in their
electric charges +|e| and 0. Again, the Coulomb force is not the dominant
force in nuclei, as we have seen in the empirical mass formula. We can
ignore the difference in the electric charge and put it back in as a small
perturbation.
The charge symmetry is a limited example of the isospin invariance. It
corresponds to the overall reversal of all isospins. If you reverse all spins sz ,
that is basically the 180 rotation around the y-axis, and you obtain another

13
 750

600
Potential [MeV]

450 3
P0(np)

300

150 3
P (np)
1 1
P1(np)
0
1
S0(np)
-150
0 0.5 1 1.5 2 2.5 3
Radius [fm]

750
1
F (np)
3
600
Potential [MeV]

450

300
3
F3(np)
150 1
D2(np)

0
3
D (np)
2
-150
0 0.5 1 1.5 2 2.5 3
Radius [fm]

Figure 6: A recent analysis of nucleon-nucleon scattering data to obtain the

nucleon-nucleon potential. Taken from A. Funk, H. V. von Geramb, K. A.
Amos, nucl-th/0105011.

14
Figure 7: Comparison of excitation spectrum of two mirror nuclei, 13 C and
13
N, 17 O and 17 F. From Theoretical Nuclear Physics, by Amos deShalit
and Herman Feshbach, New York, Wiley, 1974.

15
state with degenerate energy. Likewise, if you reverse all isospins, by rotat-
ing the isospin around the isospin y-axis by 180 , you interchange protons
with neutrons, just like interchanging spin up and spin down states. If the
nuclear force is indeed invariant under the isospin rotation, it must also be
invariant under the isospin reversal. Fig. 7) shows that indeed the nuclear
spectra approximately respect this invariance. Of course, isospin is not an
exact symmetry because protons and neutrons have different electric charges.
But the isospin invariance goes even further (charge independence). It says
that the not only the interaction between pp and nn are the same (charge
symmetry), also np is, except that you have to carefully select the config-
uration. Here is what is required. Because proton and neutron both carry
I = 1/2 (and opposite Iz = 1/2), two nucleon states would have both I = 1
and I = 0 components. Both pp and nn states are said to be in the I = 1
state. On the other hand, the np state can either be in the I = 1 or I = 0
states. But the fermion wave function must be anti-symmetric while I = 1
(I = 0) isospin wave function is symmetric (anti-symmetric). Therefore, if
the space and spin wave function of a np state is symmetric (anti-symmetric),
it selects I = 0 (I = 1) isospin wave function. This way, you can separate
purely I = 1 part of the np wave function, and compare the interaction to
that of the nn and pp states. And they are indeed the same up to corrections
from Coulomb interaction. On the other hand, the force in the I = 0 state
can be different. For instance, the only two-nucleon bound state is the deu-
terium, an np state. What is suggests is that the bound state is in the I = 0
state, and anti-symmetric isospin wave function. Then the rest of the wave
function must be symmetric. For a given potential, the S-wave is always
more binding than the P -wave just because it lacks the centrifugal barrier.
Therefore the deuterium is likely to be in the S-wave, a symmetric spatial
wave function. Then the spin wave function must be symmetric, S = 1.
Indeed deuterium does have spin one. A more quantitative test can be seen
in Fig. 8. 21 F, 21 Ar, 21 Na, and 21 Mg all have the mass number 21. Assuming
18
F is in the I = 0 state, all four nuclei can be obtained by adding three
neutrons to it, which can be in either I = 3/2 or I = 1/2 states. The nuclear
excitation spectra show states common only between 21 Ar and 21 Na, which
are in the I = 1/2 state, or states common to all four of them, which are in
the I = 3/2 state. Similarly check can be done among 14 C, 14 N, 14 O, which
show states common to all of them (I = 1) or states special to 14 N (I = 0).

16
Figure 8: Comparison of excitation spectrum of four nuclei with the same
mass number, showing states with I = 1/2 and I = 3/2 multiplet struc-
ture. From Theoretical Nuclear Physics, by Amos deShalit and Herman
Feshbach, New York, Wiley, 1974.

17
4.3 Yukawa Theory and Two-nucleon System
Given the properties of the nuclear force described in the previous section,
what, after all, is it? I briefly go through the explanations in a quasi-historic
way, but this is by no means rigorous or exhaustive. But hopefully I can
give you an idea on how we came up with the current understanding, namely
Quantum ChromoDynamics (QCD).
The obvious oddity with the nuclear force was its short-rangedness. Peo-
ple knew gravity and electromagnetism; both of them are long-ranged, with
their potential decreasing as 1/r. On the other hand, the nuclear force is
practically zero beyond a few fm. As we will discuss in the Quantization of
Radiation Field, the electromagnetic interaction is described by photons in
the fully quantum theory. Likewise, the nuclear force must also involve a par-
ticle that is responsible for the force. Such a particle is often called a force
carrier. The idea of the force carrier is simple: quantum mechanics allows
you to borrow energy E violating its conservation law as long as you give
it back within time t h /E allowed by the uncertainty principle. Take
the case of an electromagnetic reaction, say electron proton scattering. An
electron cannot emit a photon by itself because that would violate energy
and momentum conservation. But it can do so by borrowing energy as
long as the created photon is absorbed by the proton within t allowed by
the uncertainty principle. Then the virtual photon has propagated from
the electron to the proton, causing a scattering process, because of its kick
when emitted by the electron and when absorbed by the proton. Since the
photon is a massless particle with E = cp, its energy can be arbitrarily small
for small momenta, and hence t can be arbitrarily long. The distance the
virtual photon can propagate can also be arbitrarily long d = ct. This
is why the electromagnetic interaction is long-ranged. If, on the other hand,
the force carrier had a finite mass m, there is a minimum energy required
to create the force carrier particle Emin = mc2 . Therefore the time to pay
back the debt is limited: t = h /mc2 . The distance the force carrier can go
within the allowed time limit is then also limited: d = ct = h /mc. There-
fore the force carrier cannot go beyond this distance and the force becomes
short-ranged. This distance determined by the mass of the particle is called
Compton wavelength. Yukawa suggested back in 30s that the force carrier
of the nuclear force must therefore be massive. Judging from the range of the
nuclear force of about two fm, he suggested that the force carrier must weigh
about 200 times electron, or 100 MeV/c2 . The short-rangedness is then an

18
immediate consequence of the finite mass.
The presence of the charge exchange reaction suggests that the force
carrier is (or at least can be) electrically charged. This particle is called
charged pion or + in the modern terminology. The charge exchange
reaction, producing the backward peak in the np scattering is caused by the
following process. When the neutron comes close to the proton, the neutron
emits the force carrier , and it becomes a proton (!). Even though (from
the neutron point of view) she is still going pretty much straight ahead, we
see the proton coming along the original direction of the neutron, namely the
backscattered proton. The emitted is then absorbed within the time
allowed by the uncertainty principle and the proton becomes a neutron.
By 40s there was discovered a particle that weighs 200 times electron in
cosmic rays (or more precisely, 105.7 MeV/c2 ). This of course raised hope
that the discovered particle may be the force carrier for the nuclear force.
After intensive research, however, especially that carried out by Italians hid-
ing (literally) underground in Rome under Nazis occupation in 1945, it was
shown that the new particle does not show any sign to feel the nuclear force.
This particle is what is now called muon . Indeed, underground is a good
place to study muons! Later on people speculated that there may be two new
particles weighing 200 times electron, and this is indeed what happened. By
going to higher altitudes on the Andes in cosmic ray studies, people have
found that the charged pions exist in cosmic rays, which quickly (within
about 108 sec) decay to muons which live longer (about 106 sec) and reach
the surface of the Earth. (Of course their life is stretched by the relativistic
time dilation effect. Otherwise we didnt have a chance to detect them even
on the Andes.) Only at higher altitudes, pions had chance to enter the de-
tector (photographic films). Later on, a neutral pion 0 was also discovered
that decays into two photons. They are later determined to have no spin and
odd parity. Once found, it seemed to confirm Yukawas suggestion. Read
Chapter 2 of CahnGoldhaber for more details.
The potential between nucleons caused by the exchange of a virtual
pion was calculated to have the following form

1 g 2 m2 er
" ! #
2 3 3
V = m c (~
1 ~
2 ) (~
1 ~
2 ) + 1 + + S 12 . (9)
3h c 4m2N r (r)2 r

h with m with the small difference between m = 139.6 MeV/c2

Here = m c/
and m0 = 135.0 MeV ignored in the same spirit as we ignore the proton-

19
neutron mass difference and call it mN . The factor
1
S12 = 2
[3(~1 ~r)(~2 ~r) (~1 ~2 )r2 ] (10)
r
is the form for the phenomenologically required tensor force. The matrices
~ = 2I~ are the analogs of Pauli matrices for the isospin. The important
point with the potential is that it is indeed invariant under the rotation of
the isospin space because of the form (~1 ~2 ).
The OPE (one-pion-exchange) exchange Eq. (9) works well in the two-
nucleon system. We have seen that there is only one bound state in two-
nucleon system, namely deuteron, with I = 0, L = 0, S = 1. Let us see if
this is consistent with the OPE potential. We focus on the s-wave (L = 0)
which doesnt have the centrifugal barrier and presumably binds the most.
When I = 1 ((~1 ~2 ) = +1), the Fermi statistics requires S = 0 (~2 = ~1
and hence (~1 ~2 ) = 3). Then the tensor force is proportional to
1 1
S12 = 2
[3(~1 ~r)(~2 ~r) (~1 ~2 )r2 ] = 2 [3(~1 ~r)(~1 ~r) + 3r2 ]. (11)
r r
At the lowest order in the potential in perturbation theory, using the fact
that the s-wave is isotropic, we find hri rj i = 31 hr2 i, and hence the tensor
force vanishes identically. Therefore, the OPE potential is
r
g 2 m2 2e
V = m c . (12)
c 4m2N
h r

This potential is attractive, of finite range, and may or may not have a bound
state depending on the size of the coupling g 2 / hc and m . For the actual
values, there is no bound state.
On the other hand, for the I = 0, L = 0, S = 1 case, we have (~1 ~2 ) = 3
and (~1 ~2 ) = +1. Let us take Sz = +1 state as an example. Then the
tensor force does not vanish, and its expectation value is proportional to
1
hS = 1, Sz = +1| 2
[3(~1 ~r)(~2 ~r) (~1 ~2 )r2 ]|S = 1, Sz = +1i
r
1
= hS = 1, Sz = +1| 2 [3(1z z)(2z z) r2 ]|S = 1, Sz = +1i
r
2z 2 x2 y 2
= . (13)
r2

20
Therefore, the OPE potential is
g 2 m2 2z 2 x2 y 2 er
" #
2
V = m c 1 + . (14)
c 4m2N
h r2 r
The coefficient of the potential is the same as the I = 1 case,
q except that
5
there is an addition of the quadrupole moment r Y2 = 16 (2z 2 x2
2 0

y 2 ). If the quadrupole moment is positive, which means a cigar-like shape,

as opposed to negative, which means a pancake like shape, the quadrupole
moment adds to the attractive force and can lead to a bound state even
if the I = 1 case doesnt. Experimentally, the quadrupole moment of the
deuteron is confirmed and has the value Q(d) = 2.78 1027 cm2 . The
deuteron indeed has a cigar-like shape where the spins are lined up along the
elongated direction.
In order for the quadrupole moment to be non-vanishing, however, a pure
S-wave would not do the job because it is completely isotropic. However,
the state with total J = 1 with S = 1 can also arise from L = 2. In fact,
the deteron has a mixture of L = 2 state that is responsible for the finite
quadrupole moment.

5 New complications
Now the world looked simple: there are protons and neutrons in nuclei,
bound together by the force mediated by the exchange of pions. But the
world wasnt so simple after all. The first little problem is that the coupling
needed for the pion-nucleon coupling was extremely big. The analog of the
fine-structure constant was
g2
' 15. (15)
h
c
Clearly, the perturbation theory which expands systematically in powers of
g2h c is very badly behaved. Therefore the nuclei are very strongly coupled
system and theoretically very hard to deal with.
The problem starts when you want to probe shorter distances. The first
sign of the problem is the hard core in the nucleon-nucleon potential. The
one-pion-exchange potential does not give you that. Then what about two-
pion exchange? Remember the large coupling constant: the two-pion ex-
change is actually bigger than the one-pion exchange in general. Fortunately,
the two-pion exchange would be suppressed beyond the distance h /(2m c),

21
and hence the long-distance behavior is still valid with the one-pion exchange.
But at shorter distances, more and more pion exchanges, or higher orders in
g 2 /
hc are increasingly important and the perturbation theory is clearly not
working. In general, this simple picture of the world starts faltering as you
go to shorter distances, or equivalently, higher momentum transfers.1
The hell really broke loose when people discovered many more particles
that participate in the nuclear force, collectively called hadrons. There are
many more mesons, a general version of pions that are bosons, have integer
spins. There are also many more baryons, a general version of nucleons
that are fermions, have half-odd integer spins. These particles appear in
the collision of nucleons and mesons as resonances. Are they all elementary
particles? People believed for a long time that they are, because they are of
the same family as protons and neutrons, which people firmly believed were
elementary.2
One set of new particles was the resonances, new particles that are
extremely short-lived. They are produced in collisions of hadrons, such as
pion and proton, and decay within 1023 sec or so. The resonances appear as
peaks in cross sections, i.e., event rates in scattering experiments. The width
of the peak is given by the inverse of the lifetime , = h / . For more
discussions of this issue and the discovery story of resonances, see Chapter 4
of CahnGoldhaber.
Another set of new particles was the strange particles. They were
produced by the strong interaction, and therefore are expected to decay also
by strong interactions; then the lifetime must be as short as 1023 sec as seen
in hadron resonances. However, these new particles were long-lived, with
1
Another sign of the problem is that the magnetic moments of proton and neutron are
anomalous: gp /2 = 2.79 and gn /2 = 1.91 as opposed to the Diracs values: gp /2 = 1,
gn /2 = 0. One can try to explain the numbers by the quantum fluctuation of pions in
the vacuum, as we do in the Lecture Note QED, but again the non-convergence of
perturbation series makes it impossible to draw a reliable conclusion.
2
The word meson means intermediate mass, which was appropriate for pions whose
mass is between proton and electron. The electron was light, and was called a lepton,
meaning light particle, while the proton was called a baryon, meaning heavy particle
in Greek. All these names turned out to be misnomers. Eventually mesons were found
that are heavier than proton as resonances in hadronic collisions. Even a lepton heavier
than proton was discovered later, the tau-lepton. But we are stuck with these names. As
we will see later, the better characterization of these particles is that baryons are bound
states of three quarks, mesons of a quark and anti-quark, and leptons are elementary as
far as we can tell.

22
lifetimes of about 1010 sec suggesting that the decay is due to the weak
interaction. See Chapter 3 of CahnGoldhaber. The idea of strangeness
was proposed by Gell-Mann and Nishijima.

6 Conservation Laws
In order to put some order in the growing mess of particles, the concept of
conservation laws played an important role.
Recall that conserved quantities appear in physical systems because of
certain symmetries, as shown by Emilie Noether. For example, the momen-
tum conservation is a consequence of the translational invariance in space,
while the energy in time. The angular momentum is conserved when the
system is rotationally invariant.
In quantum mechanics, any physical observable is a hermitian operator.
Let us call a conserved hermitian operator O. Then we can define a unitarity
operator U = eiO . It is easy to check that it is unitarity. The interesting
point is that this unitarity operator actually reproduces the symmetry of the
system. For example, take the momentum p in one dimension for simplicity.
Then you can define a unitarity operator U (a) = eiap/h . One way to see what
this operator does is to let it act on a wave function (x). Recall that the
momentum operator is a derivative in the position space, p = hi . Then,

U (a)(x) = eiap/h (x)

n
1 i
X 
= ap (x)
n=0 n! h

1
(a)n (x)
X
=
n=0 n!

X an n
= (x)
n=0 n!
= (x + a). (16)

The last equality is nothing but the Taylor expansion of the function (x+a)
around x. Therefore, the unitarity operator U (a) causes translation in space.
One can also check what it does on operators. The way a unitarity operator
acts on the position operator is

U (a) xU (a) = eiap/h xeiap/h

23

X 1 i i i
= [ ap, [ ap, [ [ ap, x] ]]
n! h
n=0 h
h

= x + a. (17)

Again it causes translation in space.

This observation leads to a general idea. Noethers theorem says that a
symmetry of the system leads to a conserved quantity. On the other hand, a
conserved quantity generates the symmetry. You exponentiate the conserved
quantity to define a unitarity operator, which causes the symmetry opera-
tion. This way, the symmetry and the conserved quantity are in one-to-one
correspondence in quantum mechanics.
There are, however, conservation laws that do not appear in classical
mechanics but are important in quantum mechanics. One of them is the
parity. The parity inverts space, ~x ~x. The unitarity operator that
causes this symmetry is therefore

P ~xP = ~x. (18)

On the other hand, doing parity twice is nothing, and hence P 2 = 1. Com-
bined with the unitarity P P = 1, we find P = P , and hence the parity
operator is both unitarity and hermitian. The momentum is also odd un-
der parity, P p~P = ~p. On the other hand, the angular momentum is even,
because L~ = ~x p~ changes the sign twice. Similarly, the spin angular mo-
mentum (and hence the total angular momentum) is also even under parity.
The parity is not very useful in classical mechanics because it changes
a trajectory to its parity conjugate, a different trajectory. In quantum me-
chanics, however, the wave function is spread out in space, and we can talk
about parity eigenstates. Given P 2 = 1, its eigenvalue can only be 1. For
example, a wave function of definite orbital angular momentum is given by
the spherical harmonics Ylm (, ). Under parity, and + .
A general property of the spherical harmonics is

Ylm ( , + ) = (1)l Ylm (, ). (19)

Therefore, the spherical harmonics is an eigenstate of the parity operator

with eigenvalues (1)l .
Why is the parity useful? Here is an example. If you consider electric
dipole transitions of an atom, the consideration of parity allows you to decide

24
which transitions are possible and which arent. The matrix element of the
transition is given by hn0 l0 m0 |~x|nlmi. Using the parity operator,

hn0 l0 m0 |~x|nlmi = hn0 l0 m0 |P P ~xP P |nlmi

= (hn0 l0 m0 |P )(P ~xP )(P |nlmi)
0
= hn0 l0 m0 |(1)l (~x)(1)l |nlmi)
0
= (1)l+l +1 hn0 l0 m0 |~x|nlmi. (20)

Therefore, the matrix element does not vanish only when l and l0 differ by
an odd integer. This simple consideration alone tells you a great deal about
how excited states of the hydrogen atom decay. For example, 2p state decays
to 1s, while 3p to either 2s or 1s. On the other hand, 3d state can decay
only to 2p, because this is the only state with lower energy and odd l. In the
same token, 3s state decays only to 2p. An interesting one is the 2s state. It
cannot decay anywhere! It turns out that the only way the 2s state decays
is by emitting two photons instead of one. It first decays to 2p state and
a photon. But this state is virtual because the 2p state has the same energy
as the 2s state while the photon carries away a finite energy. This virtual 2p
state then decays to 1s by emitting another photon. This way, the 2s state
decays to 1s with two photons with a long (order second) lifetime, compared
to typical lifetime of excited states (order 10 nanoseconds).
Similarly, we assign parity eigenvalues to particles. For instane, all pions
were determined to carry odd parity. The parity operator acts as P |(~p)i =
|pi(~p)i, and the state at rest p~ = 0 is an eigenstate of the parity operator.
Spinless particles are in general called scalar, and when they have odd parity,
they are called pseudoscalars specifically.
An important difference between other conserved quantities and the par-
ity is that the parity is a multiplicative quantum number while the momen-
tum, energy, and angular momentum are additive quantum numbers. When
you have two particles of momenta p~1 and p~2 , the total mometum is their
sum, P~ = p~1 + p~2 . On the other hand, when you have two particles of definite
parity eigenvalues, the total parity is the product of the two. For instance, a
state with two pions at rest have positive parity.
There are many more quantum numbers in particle physics. When the
positron was discovered, it raised a very naive question: why cant a proton
decay into a positron and a photon p e+ ? This process would conserve
momentum, energy, angular momentum, electric charge, and even parity.

25
Even if the lifetime is as long as 1016 years,3 Maurice Goldhaber pointed out
that we receive too large radiological dose because of protons decaying in
our body. Stuckelberg (and later Wigner) decided that there must be a con-
served quantity called baryon number, which forbids this decay. Because
the proton is a baryon, you assign a baryon number +1. But the positron is
a lepton, and has baryon number zero. Nor the photon has a baryon num-
ber. Then the initial state has baryon number one, while the final state has
baryon number zero. Because the baryon number is conserved, this decay
is forbidden. You see that this conservation law was made up to explain
the absence of this process. But even up to now, the proton has never been
seen to decay. Todays experiments have set a lower limit on proton decay
of about 1033 years. Many theories such as grand-unified theories however
predict that the proton does decay eventually. In any case, in all known four
forces, the baryon number is conserved. We assign the baryon number one
to other baryons, neutron, , , , , etc.
Similarly, when the muon was discovered, it also raised the question: why
doesnt a muon decay as e . This led Inoue and Sakata propose yet
another quantum number: lepton family number. You assign an electron
number one to electron, but muon has no electron number. Conserversely,
the muon has the muon number one, while the electron has no muon number.
Then the initial state, muon, has one muon number and zero electron number,
while the final state has zero muon number and one electron number. As long
as these quantum numbers are conserved, this decay is forbidden. The recent
discovery of neutrino oscillation, however, implies that the lepton family
number is not conservedin nature. We will come back to this point later.
Therefore, the lepton family number is an approximate conservation law, not
an exact one. Nonetheless, most reactions do conserve the lepton family
number and hence it is useful.
There are other approximate conservation laws. We already discussed
isospin. The electromagnetism violates isospin because particles in the same
isospin multiplet (proton and neutron) have different electric charges. How-
ever, the nuclear strong force does respect isospin as seen in nuclear levels.
The strangeness was introduced as yet another approximate conserva-
tion law. The purpose of it was to explain why certain particles, such as
K-mesons and 0 -baryon, are stable over the typical time scale of strong in-
teractions. If they carry a new quantum number, namely strangeness, and if
3
For a comparison, the age of our Universe is about 1.4 1010 years.

26
it is conserved under the strong interactions, they cannot decay because they
are the lightest meson and baryon, respectively, with non-zero strangeness.
Indeed, the reaction such as 0 p K + 0 is possible by the strong interac-
tion because K + carries strangeness +1 while 0 1. On the other hand,
K + cannot decay by the strong interaction because only ligher hadron of the
same charge is + which carries zero strangeness. It can decay only via weak
interaction which violates strangeness and allows K + + 0 . Strangeness
is an additive quantum number just like the electric charge.

27