3 NUCLEAR REACTIONS 42

3 Nuclear Reactions
If energetic particles encounter matter, nuclear reactions can take place. The energetic particles
can originate from a reactor, accelerator or radioactive source. The first nuclear reaction
experiments were done by Rutherford using alpha particles (He-nuclei).

3.1 Types of Nuclear Reactions

A typical nuclear reaction is represented by

a + X −→ Y + b,

where a is the particle projectile, X is the target, Y and b are the products. Usually, Y is a
heavy product, i.e. not directly observed because it stops in the target, and a and b are nucleons
of light nuclei. But occasionally, b will be a γ-ray for which the reaction is called radioactive
capture. If a is a γ-ray, the reaction is called nuclear photoeffect.

In a compact form, the above nuclear reaction is written as

X(a, b)Y,

which is convenient as it gives a natural way to refer to a general class of reactions with common
properties, e.g. (α, n) and (n, γ) reactions.

Nuclear reactions can be classified in many ways. If the incident and outgoing particles (a & b)
are the same (and correspondingly X & Y the same nucleus), the reaction is a scattering process.
This is elastic scattering if Y and b are in their ground states, and inelastic if either Y or b is in
an excited state, and will generally decay quickly by γ emission.

In other cases, a and b are the same particle, but the reaction causes yet another nucleon to be
ejected separately, to have three particles in the final state. This leads to a knockout reaction.

In a transfer reaction, one or two nucleons get transferred between the projectile and the target.
For example, an incoming deuteron turning into outgoing proton or neutron. In this case, one
nucleon is added to the target X to form Y.

Reactions can also be classified by the mechanism governing the process. For example, “Direct
3 NUCLEAR REACTIONS 43

Reactions”, where only very few nucleons take part in the reaction, the remaining nucleons of
the target being passive. Such reactions might remove or insert a nucleon from a shell-model
state, therefore serving as ways to explore the shell structure of the nuclei.

Then, on the extreme, there is the “Compound Nucleus Mechanism”. In this case, the incoming
and target nuclei merge briefly for a complete sharing of energy before the outgoing nucleon is
ejected.

Between the two extreme kinds of nuclear reactions by mechanism, we have the “Resonance
Reactions”, where the incoming particle forms a quasi-bound state before the outgoing particle
is ejected.

There are techniques for measuring the energies of the outgoing particles. Besides, the direction
of the emission of the outgoing particle can be determined, and the angular distribution can be
observed by counting the number of emission at various angles.

By polarization experiments, the spin orientation of the product nucleus Y can be deduced. We
require an incident beam of polarized nuclei and a spectrometer, for analyzing the polarization
of the outgoing particle b.

3.2 Conservation Laws

The same conservation laws as applied in the study of radioactive decays.

3.2.1 Conservation of Total Energy and Linear Momentum

This is used to relate the energies of the products (unknown but measurable) to the energy of
the projectile (known and controllable). The measured energy of b can be used to deduce the
excitation of states Y, or the mass difference between X and Y.

3.2.2 Conservation of Proton and Neutron Number

This is attributed to low energy of the process, where no meson formation or quark rearrangement
takes place [Note that quarks are elementary particles that combine to form composite particles
called hadrons, the most stable of which are protrons and neutrons (nucleons)]. The weak
3 NUCLEAR REACTIONS 44

interaction is negligible on the timescale of nuclear reactions. Although we still conserve

nucleon/baryon number at high energies, it is at lower energies that we conserve separately
proton number and neutron number.

3.2.3 Conservation of Angular Momentum

This makes it possible to relate the spin assignments of the reacting particles (a & X) and the
orbital angular momentum carried by the outgoing particle, which can be deduced by measuring
its angular distribution.

3.2.4 Conservation of Parity

The net parity before the reaction must equal the net parity after the reaction. By knowing the
orbital angular momentum of the outgoing particle, the (−1)l rule can be used, as well as other
known parities in the reaction to deduce the unknown parities of excited states.

3.2.5 Conservation of Isospin

The interactions of nucleon with its surroundings (e.g. other nucleons) do not (in most cases)
depend on whether the nucleon has spin components ms = ± 12 relative to an arbitrary chosen
z-axis. That is, there is no need to distinguish between a spin-up nucleon and a spin-down
nucleon.

The charge independence of nuclear forces means that we do not need to distinguish between
neutrons and protons, thus they should be grouped as nucleons, i.e. members of a common
family. However, electromagnetic interaction can distinguish between protons and neutrons.

The formalism for nuclear reactions may depend on the multiplicity of nucleon states, where
the neutrons and protons are treated as two different states of a single particle (nucleon). The
nucleon is assigned a fictitious spin vector called “Isospin”(often called isotopic spin or isobaric
spin). In the absence of electromagnetic fields, the nucleon has two degenerate spin states,
i.e. isospin-up (arbitrarily assigned to a proton) and isospin-down (assigned to neutron). For a
nucleon with isospin quantum number t = 21 , mt = + 21 (for proton) and mt = − 21 (for neutron).

The isospin obeys the rules for angular momentum, thus we use an isospin vector t of length
3 NUCLEAR REACTIONS 45

p
t(t + 1)~, with 3-axis projections t3 = mt ~.

3.3 Energetics of Nuclear Reactions

In the basic reaction, conservation of total relativistic energy takes the form

mX c2 + TX + ma c2 + Ta = mY c2 + TY + mb c2 + Tb , (57)

where T is the kinetic energy, which we can use the non-relativistic approximation 21 mv 2 at low
energy, and m is the rest mass.

The reaction Q value is defined as the initial mass energy minus the final mass energy.

Q = (minitial − mfinal )c2 = (mX + ma − mY − mb )c2 . (58)

Or, Q value is the excess kinetic energy of the final products;

Q = Tfinal − Tinitial = TY + Tb − TX − Ta . (59)

The Q value may be positive or negative;

1. If Q > 0 implies minitial > mfinal or Tfinal > Tinitial . This is the case of exoergic (exothermic)
reactions, where nuclear mass or binding energy is released as kinetic energy of the final
products.

2. If Q < 0 (minitial < mfinal or Tfinal < Tinitial ), we have endoergic (endothermic) reactions,
where initial kinetic energy is converted into nuclear mass or binding energy.

The changes in mass must be related by

△E = △mc2 ,

which follows from special relativity. Any changes in T of the reacting particles must be balanced
by an equal change in rest energy.

Consider the target nuclei at rest, where thermal energies are negligible on MeV scale of nuclear
reactions. Let the reaction plane be the direction of the incident beam and one of the outgoing
particles. Conservation of momentum along and perpendicular to the plane shows that
3 NUCLEAR REACTIONS 46

Figure 11: Basic reaction geometry.

Figure 12: The inset shows the region of double-valued behaviour near 1.0 MeV.

Pa = Pb cosθ + PY cosξ (60)

0 = Pb sinθ − PY sinξ. (61)

From (59-61), we have three equations in four unknowns, i.e. θ, ξ, Tb and TY . The equations
have no unique solution. If we do not observe particle Y (usually the case), we can eliminate
3 NUCLEAR REACTIONS 47

ξ and TY to find a relationship between θ and Tb , which takes the form

1 1
1 (ma mb Ta ) 2 cosθ ± {ma mb Ta cos2 θ + (mY + mb )[mY + (mY − ma )]Ta } 2
Tb2 = . (62)
mY + mb
A plot of Tb against Ta is shown in Figure 14, for the reaction 3 H(p, n)3 He, Q ≈ -764 MeV.

There are several features of the plot, namely:

1. Except for very small energy region (1.019 - 1.147 MeV), there is a one-to-one correspon-
dence between Tb and θ for a given Ta .

2. There is an absolute minimum value of Ta below which the reaction does not take place.
This is the case for θ < 0, where we have this threshold energy,
mY + mb
Tth = (−Q) . (63)
mY + mb − ma
This always occurs for θ = 0◦ (thus, φ = 0◦ ); the products Y and b move in a common
direction, but they are still separated nuclei. No energy is wasted in giving them momentum
transverse to the beam direction.

For Q > 0, there is no threshold condition, and the reaction will go even for very small
energies.

3. The double-valued situation occurs for incident energies between Tth and the upper limit,
mY
Ta′ = (−Q) . (64)
mY − ma
This also occurs only for θ = 0◦ , and is important only for reactions which involve nuclei
of comparable masses. Thus,
 
m m m
− Tth ∼
a b b
Ta′ = Tth 1− + ... (65)
mY (mY − ma ) mY
If a and b have m 6 4, and if Y is a medium or heavy nucleus, then the range (Ta′ − Tth )
is much smaller than 1% of Tth .

4. There is also a maximum angle θmax at which the double-valued behaviour occurs. The
value is determined by vanishing the argument in the square root of Eq. 62;
(mY + mb )[mY Q + (mY − ma )]Ta
cos2 θmax = − . (66)
ma mb Ta
3 NUCLEAR REACTIONS 48

When Ta = Ta′ , the double-valued behaviour occurs between 0◦ 6 θ 6 90◦ , i.e. θmax = 90◦ .
Near Ta = Tth , it occurs only near θmax ≈ 90◦ .

5. Reactions with Q > 0 have no threshold or double-valued behaviour, where we make the
transition −Q → +Q. The reactions occur down to Ta → 0, and the curves are single-
valued for all θ and Ta .

3
Figure 13: Ta Vs Tb for the reaction 3 He(n, p) H, where no double-valued behaviour exists.

If we measure Tb for a given θ and Ta , we can determine the Q value of the reaction and deduce
the mass relationships among constituents . From Eq. 62, we obtain
     1/2
mb ma ma mb
Q = Tb 1 + − Ta 1 + −2 Ta Tb cos θ. (67)
mY mY mY mY
If the reaction reaches excited state of Y, the Q value equation should include the mass energy
of excited state;

Qex = (mX + ma − m∗Y − mb )c2

= Q◦ − Eex , (68)

where Q◦ is the Q value corresponding to the ground state of Y, i.e m∗Y c2 = mY c2 + Eex , is
the mass energy of the excited state, and Eex is the excitation energy above the ground state.
The largest observed values of Tb correspond to reactions which lead to the ground state, and
we can use Eq. 67 to find Q◦ . Smaller values of Tb correspond to higher excited states, and by
measuring Tb , we can deduce Qex and Eex .
3 NUCLEAR REACTIONS 49

3.4 Reaction Cross Sections

The cross section is a measure of the relative probability for the reaction to occur. Consider
a detector placed to record particle b emitted in a direction (θ, φ) with respect to the beam
direction, defining a solid angle dΩ.

Figure 14: Reaction geometry with outgoing beam in the direction (θ, φ) into a solid angle dΩ.

Let Ia be the current of the incident particles per unit time, and N the target nuclei per unit
area. If the outgoing particles appear at a rate Rb , then the reaction cross section is given by

Rb
σ= . (69)
Ia N

Hence, σ has the dimension of area per nucleus, but it may be very much larger or smaller than
the geometrical area of the disc of the target nucleus seen by the incoming beam. For a typical
nucleus of radius R = 6 fm, the geometrical area = πR2 ≈ 100 fm2 = 1 b. But for neutron
135
capture by Xe, σ = 106 b, while for much more impossible reactions, σ may be measured in
millibarns or microbarns.

The detector occupies only a small solid angle dΩ, so does not observe all the outgoing particles,
i.e. only a small fraction dRb are actually counted. Therefore, only a small fraction of the cross
section dσ will be deduced. Also, the outgoing particles, in general, will not be emitted uniformly
3 NUCLEAR REACTIONS 50

in all directions, but will have an angular distribution that will depend on θ and possibly on φ.
If we represent the angular distribution arbitrarily by r(θ, φ), then
dΩ
dRb = r(θ, φ) , (70)
4π
dΩ
where 4π is introduced to make 4π
a pure fraction. Thus,
dσ r(θ, φ)
= , (71)
dΩ 4πIa N
which is the differential cross section; its measurement gives important information on the
angular distribution of the reaction products, i.e. σ(θ, φ) or σ(θ). The limits of the differential
cross section are barns/steradian. The cross section, σ, can be found from,
Z Z π Z 2π
dσ dσ
σ= dΩ = sinθdθ dφ . (72)
dΩ 0 0 dΩ
dσ
If dΩ
is constant, then  
dσ
σ = 4π , (73)
dΩ
which justifies the inclusion of 4π in Eq. 70.

In many applications of nuclear physics, we want to find also b emitted at a certain energy, which
corresponds to a particular energy of the residual nucleus Y. Thus, we must modify the definition
of cross section to give the probability to observe b in the angular range dΩ and in energy range
d2 σ
dEb , giving the so-called doubly differential cross section, dEb dΩ
. For discrete states, there may
be only a single level within the energy range dEb , and the distinction from the differential cross
section becomes less important.

If on the other hand, we do not observe the direction of b, then we measure the differential cross
dσ
section, dE
, where E may represent the ecxitation energy of Y. Another cross section of interest
is the total cross section, σt . For a specific incident particle, the reaction cross sections σ for all
possible outgoing particles are added no matter what their direction or energy. This would tell
the probability for an incident particle to have any reaction at all with the target, and thus be
removed from the beam of incident particles. We deduce this directly by measuring the loss in
intensity of a collimated beam in passing through a certain thickness of target material. The
different measurements and applications are summarized below.

1. Total cross section, σt ; technique attenuation of beam; possible application: shielding.

3 NUCLEAR REACTIONS 51

2. Reaction cross section, σ; technique integrate over all angles and energies of b; possible
application: production of radioisotopes Y in nuclear reaction.

dσ
3. Differential (angular) cross section, dΩ
; technique: observe b at (θ, φ), but integrate over
all energies; possible application: Formation of beam of b particles in a certain direction
(or recoil of Y).

dσ
4. Differential (energy) cross section, dE
; technique: Do not observe b, but excitation of Y
by subsequent γ emission; possible application: study of decay of excited states of Y.

d2 σ
5. Doubly differential cross section, dΩdE
; technique: observe b at (θ, φ) at a specific energy;
possible application: information on excited states of Y by angular distribution of Y.

3.5 Experimental Techniques

Typically, a nuclear reaction study requires a beam of particles, a target, and a direction system.
Beams of charged particles are produced by a variety of accelerators, and neutron beams come
from nuclear reactors and as secondary beams from charged-particle accelerators. To do precise
spectroscopy of the outgoing particle b and the residual nucleus Y, the beam must satisfy the
following:

ˆ It must be highly collimated and focused, to have a precise reference direction to determine
θ and φ, for angular distribution measurements.

ˆ It must have a sharply defined energy, otherwise trying to observe a specific excited state
by finding Qex and Eex , we might find that variation in Ta would give two or more different
values of Eex for the same Tb .

ˆ It must be of high intensity, so that the necessary statistics may be gathered for precise
experiments.

ˆ For timing measurements (such as measuring the lifetimes of excited states of Y), the
beam must be sharply pulsed, to provide a reference signal for the formation of the state,
and the pulses must be separated in time by at least the time resolution of the measuring
3 NUCLEAR REACTIONS 52

apparatus.

ˆ The incident beam intensity should be nearly constant and easily measurable. This is
necessary for determination of the cross section.

ˆ The beam may be polarised or unpolarised, as desired by the experimenter.

ˆ The beam must be transported to the target through a vacuum system, to prevent beam
degradation and production of unwanted products due to collisions with air molecules.

Note that the types of targets vary widely, depending on the goals of the experiment. Thin
metallic foils are often used as targets, but for non-metals (including compounds such as oxides),
the target material is often placed on a thin backing.

3.6 Coulom Scattering

The nucleus has a distribution of electric charge. This causes it to scatter a beam of charged
particles. The scattering may be either elastic or ineleastic. The geometry for elastic Coulomb
1
scatteringv(Rutherford scattering) is shown in Figure 15. For unbound orbits in a r2
force, the

Figure 15: Trajectory of a particle undergoing elastic Coulomb scattering, with impact parameter b.

scattered particle follows a hyperbolic path. The particle approaches the target nucleus along a
3 NUCLEAR REACTIONS 53

straight line that would pass a distance b in the absence of the Coulomb force. But as particle
approaches the nucleus, it is scattered through angle θ. Far from the nucleus, the particle has
negligible Coulomb potential energy, i.e. its total energy is the incident kinetic energy,
1
Ta = mv◦2 . (74)
2
At such distances, its angular momentum relative to the nucleus is equal to mv◦ b. In passing
close to the target nucleus, the particle reaches a minimum separation distance rmin, the
absolute minimum value of which occurs in a head-on collision (b=0)for which the particle
comes instanteneously to rest before reversing its motion. At this distance of closest approach
(d), the particle has exchanged its initial kinetic energy for the Coulomb potential energy. Thus,
1 2 Zze2
mv◦ = , (75)
2 4πǫ◦ d
where ze is the charge of the incident particle (projectile), and Ze is that of the target. At
intermediate points,
1 2 1 2 Zze2
mv = mv + . (76)
2 ◦ 2 4πǫ◦ r
The scattering has a cylindrical symmetry about the beam axis, thus the cross section is
independent of φ. Then, considering a ring of annular geometry, so that particles with impact
parameter between b and b + db are scattered into the ring at angles between θ and θ + dθ. Let
the target have n nuclei per unit volume, and thickness x. The number of nuclei per unit area is
nx, and the fraction df of the incident particles passing through the annular ring of area 2πbdb
is
df = nx(2πbdb). (77)

The fraction f with impact parameters less than b is

f = nxπb2 . (78)

If particles scattered with impact parameter b emerge at angle θ, then Eq. 78 also gives the
fraction that are scattered at angles greater than θ. The net linear momentum of the scattered
particles changes in direction only, i.e. far from the scattering, the initial and final linear
momentum are both equal to mv◦ . It is assumed that the target nucleus is massive that it
does not move. The change in momentum vector is of magnitude,

∆P = 2mv◦ sin(θ/2), (79)

3 NUCLEAR REACTIONS 54

in the direction of the bisector of π − θ. ∆P is equal to the net impulse of the Coulomb force in
that direction. From Newton’s second law,
dP
F =
Z Zdt Z
Zze2 dt
∆P = dP = F dt = cosβ, (80)
4πǫ◦ r2
where β is the angle between the bisector and the instanteneous vector ~r locating the particle.
In the initial position (far from the scattering), t = 0, and β has the value -(π/2 − θ/2). In the
finial position (far from the scattering), t = ∞, and β has the value +(π/2 − θ/2).

The instanteneous velocity can be written in terms of the radial and tangential components, i.e.
dr dβ
~v = r̂ + r β̂, (81)
dt dt
where r̂ and β̂ are the unit vectors in the radial and tangential directions, respectively. However,
only the tangential component contributes to the angular momentum about the nucleus. Thus,
dβ
l = |m~r × ~v| = mr 2 . (82)
dt
Far from the nucleus, the angular momentum is mv◦ b. Thus, conservation of angular momentum
gives
dβ
mv◦ b = mr 2
dt
dt dβ
⇒ 2
= (83)
r v◦ b
Eq. 83 into Eq. 80 gives
Z +(π/2−θ/2)
Zze2 Zze2
∆P = cosβdβ = cos(θ/2). (84)
4πǫ◦ v◦ b −(π/2−θ/2) 2πǫ◦ v◦ b
Comparing Eq. 79 and Eq. 84, we obtain
d
b= cot(θ/2). (85)
2
Eq. 85 in Eq. 77 yields
d2
|df | = πnx cot(θ/2) cosec2 (θ/2). (86)
4
And the rate at which particles reach the ring per unit solid angle is
Ia |df |
r(θ, φ) = , (87)
dΩ/4π
3 NUCLEAR REACTIONS 55

where Ia |df | is the number of incident particles that fall between b and b + db. For the ring
geometry, we have dΩ = 2πsinθdθ. The net result is
 2  2
dσ Zze2 1 1
= 4 , (88)
dΩ 4πǫ◦ 4Ta sin (θ/2)

which is the differential cross section for Rutherford scattering.

The treatment of Rutherford scattering above is based entirely on classical mechanics, i.e. no
quantum considerations. The uncertainty principle, however, renders doubtful any treatment
based on fixed trajectory and particle orbits. Any attempt to locate a particle with an impact
parameter of arbitrarily small uncertainty would introduce a large uncertainty in the transverse
momentum, and thus in the scattering angle. Consider an uncertainty ∆b in the impact
parameter. There is a corresponding uncertainty ∆P in the transverse momentum of order
~/∆b. The above derivations are valid if ∆b ≪ b, and ∆P ≪ Ptransverse . Then,

b∆Ptransverse ≫ ∆b∆P & ~

b∆Ptransverse
≫ 1. (89)
~

Two extreme cases to consider are:

1. For 90◦ scattering, assuming 8 MeV incident alpha particle, b = d/2 = 14 fm and ∆Ptransverse
= 250 MeV/c. The ratio in Eq. 89 is ∼ 18, which is far from quantum limit.

2. For small-angle scattering (θ ≈ 1◦ ), b = 1600 fm and ∆Ptransverse ≈ mv◦ tanθ ≈ 4 MeV/c,

giving the ratio in Eq. 89 is ∼ 32, also far from quantum limit.

However, far from the 1/r 2 force, the classical calculation as ~ß0 gives the same result as quantum
calculation with no factor ~. As the energy of the incident particle increases, a point is reached
where the distance of closest approach decreases to the nuclear radius. Thus, the projectile and
the target feel each other’s nuclear force. Rutherford’s formula fails at such a point to account for
the cross section. The cross section contains both Coulomb and nuclear parts. The internuclear
separation at which the Rutherford’s formula fails gives a measure of the nuclear radius.
3 NUCLEAR REACTIONS 56

3.6.1 Optical Model

This is a simple model used to account for elastic scattering in the presence of absorptive
effect. The calculation resembles that of light incident on a somewhat opaque glass sphere. The
scattering is represented in terms of a complex potential U(r) given by

U(r) = V (r) + iW (r), (90)

where V and W are real functions selected to give the potential its proper radial dependence.
The real part V (r) is responsible for the elastic scattering. It describes the ordinary nuclear
reaction between the target and the projectile. The imaginary part W (r) is responsible for the
absorption. To demonstrate this, consider a square-well form of U(r), i.e

 −V − iW , for r < R
◦ ◦
U(r) = (91)
 0, for r > R.
p
The outgoing wave takes the form eikr /r, where the wave number, k = 2m(E + V◦ + iW◦ )/~2 ,
and is complex in nature, i.e .
k = kr + iki .

Thus, the wave function behaves like eikr ·e−ikr /r. The wave is therefore exponentially attenuated
as it passes through the nucleus.

For W > 0, we have a loss in intensity from Eq. 91. If it is assumed that W◦ ≪ E + V◦ , i.e. the
absorption is relatively weak, then using the binomial theorem,
r s  
2m(E + V◦ ) W◦ 2m 1
k= +i . (92)
~2 2 ~2 E + V◦

The usual shell-model potential has a depth V◦ ∼ 40 MeV, and taking E = 10 MeV for a typical
low-energy projectile, the distance over which the intensity is attenuated by e−1 (∼ mean free
path) is r
1 1 ~2 (E + V◦ )
d= = . (93)
2ki W◦ 2m
For d to be at most the order of the nuclear radius (∼ 3 fm), then W◦ ≈ 11 MeV. Thus, for
the usual case where absorption is relatively weaker than elastic scattering, |V | ∼ 40 MeV and
|W | ∼ 10 MeV. A more detailed form of the potential is
−V◦
V (r) = , (94)
1 + e(r−R)/a
3 NUCLEAR REACTIONS 57

where the constants V◦ , R and a are adjusted to give the best fits with the scattering data. The
absorptive part W (r) at low-energies must have a different form.

Tightly bound nucleons in the nuclear interior cannot participate in the absorption of incident
nucleons (because of the exclusion principle). Thus, only the valence nucleons (near the surface)
can absorb the relatively low energy carried by the incident particle. Then W (r) is always chosen
as ∝ dV /dr, which is large only near the surface, as shown in Figure 16.

Figure 16: Optical model functions V (r) and W (r), with model parameters V◦ 40 MeV, a = 0.523 fm, A = 64,
and R = 1.25A1/3 .

At high energies, the inner nucleons can also participate in absorption, and W (r) = dV /dr looks
more like V (r). In modern optical potentials, a spin-orbit term is included, which also peaks
near the nuclear surface as the spin density of the inner nucleons vanishes. So, a Coulomb term
must be included if the incident particle is charge (like alpha particle). For a chosen potential,
the Schrödinger equation can be solved, and equating the boundary conditions at r = R, we can
obtain the scattering amplitudes, which can be used to calculate cross sections with experiments.
3 NUCLEAR REACTIONS 58

3.7 Compound-Nucleus Reactions

If an incident particle enters a target nucleus with an impact parameter small compared with
the nuclear radius, it will have a high probability of interacting with one of the nucleons of the
target. The struck nucleon and the incident particle can each make successive collisions with
other nucleons, and after some time of interactions, the incident energy is shared among many
nucleons of the target projectile system.

As more or less random collisions occur, there is a statistical distribution in energies, and a
small probability for a single nucleon to gain enough energy to escape from the nucleus. Such
reactions have intermediate state that lies just before the emission of the outgoing particle (s),
but after the absorption of the incident particle. The intermediate state, called the compound
nucleus, can be represented in the general reaction equation by

a + X −→ C∗ −→ Y + b, (95)

where C∗ represents the compound nucleus. We can then consider the reaction that proceeds
through the compound nucleus to be a two-step process, involving the formation and subsequent
decay of the compound nucleus.

Essential to the compound-nucleus model of nuclear reactions is the assumption that the relative
probability for decay into any specific set of final products is independent of the means of
formation of the compound nucleus. The decay probability depends only on the total energy of
the system, i.e. the decays are governed primarily by statistical rules. Thus, a given compound
nucleus can decay in a variety of ways. For example consider the compound nucleus 64
Zn∗ ,
which can be formed and decays in several ways; Experimentally, the decay of 64
Zn∗ is shown

to be independent of how it was originally formed. Thus, we expect the relative cross sections
63
Cu(p,n)63 Zn and 60 Ni(α,n)63 Zn to be the same at incident energies that give the same excitation
energy of 64
Zn∗ .
3 NUCLEAR REACTIONS 59

The compound-nucleus model operates best for low incident energies (∼ 10−20 MeV), where the
incident particle (projectile) has little chance of escaping from the nucleus. It also works well for
medium weight and heavy nuclei, where the nuclear interior is large enough to absorb the incident
energy. Because of random interactions among the nucleons, the outgoing particle is expected
to be emitted with nearly isotropic angular distribution, i.e. the same in all directions. In cases
where a heavy ion is the incident particle, large amounts of angular momentum is transferred to
the compound nucleus, and to extract the angular momentum, the emitted particles tend to be
emitted at right angles to the angular momentum. With light projectiles, this effect is negligible.

3.8 Direct Reactions

These are at the extreme opposite from the compound-nucleus reactions. Here, the incident
particle interacts primarily at the surface of the target nucleus, hence the reactions are also
called peripheral processes. These reactions are most likely to involve one nucleon or very few
valence nucleons near the surface of the target nucleus.

It may be possible to have both direct and compound-nucleus reactions, but how can
their contributions be distinguished in a reaction? Two principal differences that can be
experimentally observed are:

1. Direct reactions occur very rapidly, i.e. in a time of order of 10−22 s, while compound-
nucleus reactions take longer ∼ 10−16 − 10−18 s. The additional time is for the distribution
and re-concentration of the incident energy.

2. The angular distributions of the outgoing particles in direct reactions are more sharply
parked than in the compound-nucleus reactions.

Inelastic scattering could take place either through a direct reaction or compound nucleus. This
depends largely on the energy of the incident particle. An example of a reaction that occurs
by either mechanism is deuteron stripping reaction, which is an example of a transfer reaction
in which a single protron is transferred from the projectile to the target nucleus. The deuteron
stripping reaction (d,p) may be more likely to go by direct process, since the evaporation of
protons from the compound nucleus is inhibited by the compound barrier.