A

Scattering and absorption cross-sections

Basic definitions
In this section the basic definitions of the cross-section are recalled for processes that involve either
the scattering or absorption of an X-ray photon. The cross-section is an important quantity, as it
is the meeting point of experiment and theory. Although its definition is straightforward, confusion
sometimes arises as there are several, but essentially equivalent, definitions. As illustrated in Fig.
A.1 the definition depends on the situation considered, and in particular on whether or not the cross-
sectional area of the beam is larger or smaller than that of the sample.
We start by considering the scattering event shown in Fig. A.1(a) in which an X-ray beam of
intensity I0 photons per second is incident on a sample, and where the sample is large enough that
it intercepts the entire beam. Our objective is to calculate the number of X-ray photons, Isc , scattered
per second into a detector that subtends a solid angle ∆Ω. If there are N particles in the sample per unit
area seen along the beam direction, then Isc will be proportional to N and to I0 . It will of course also
be proportional to ∆Ω. Most importantly it will depend on how efficiently the particles in the sample
scatter the radiation. This is given by the differential cross-section, (dσ/dΩ), so that we may write
!
dσ
Isc = I0 N ∆Ω
dΩ

Thus the differential cross-section per scattering particle is defined by

!
dσ No. of X-ray photons scattered per second into ∆Ω
= (A.1)
dΩ I0 N∆Ω

Here no restriction has yet been placed on whether or not the scattering event is elastic or inelastic.
The corresponding absorption experiment is simpler to analyse, as the detector is placed directly
in the incident beam, and the change in intensity recorded when a sample is introduced into the beam.

Elements of Modern X-ray Physics, Second Edition. Jens Als-Nielsen and Des McMorrow
© 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd.
344 Scattering and absorption cross-sections

Fig. A.1 (a) A beam is incident on a sample with a larger cross-sectional area. In this case the intensity Isc scattered into a solid
angle ∆Ω is proportional to the incident intensity I0 of the beam, i.e. the number of photons per second. (b) A beam is incident on
a sample with a smaller cross-sectional area. Now the scattered intensity Isc is proportional to the incident flux Φ0 of the beam,
i.e. the number of photons per second per unit area.

The number of absorption events, W4π , per second is proportional to I0 and N as before. The subscript
is used to remind us that the photoelectron liberated from the atom in the absorption process may be
emitted into any direction in 4π steradians. The absorption cross-section, σa , is defined by
W4π = I0 N σa
so that

W4π
σa =
I0 N

Alternatively a different situation could have been imagined in which the incident beam is larger
than the sample, as shown in Fig. A.1(b). It should be clear that in this case it is necessary to consider
 345

the flux of the incident beam, i.e. the number of photons per second per unit area, and not its intensity.
The scattered intensity is now given by
!
dσ
Isc = Φ0 ∆Ω (A.2)
dΩ

where Φ0 is the flux of the incident beam, and the differential cross-section refers to the whole sample.
The absorption cross-section for this geometry is
W4π
σa =
Φ0

These are the definitions of the cross-sections which are operational for analysis of an experiment.
Having defined what exactly is meant by the cross-section the next question is how it may be calculated.
The classical description of the scattering of an electromagnetic wave by a single electron is used in
Chapter 1 and Appendix B to derive the Thomson scattering cross-section. This description is usually
sufficient in many fields such as reflectivity, crystallography, etc. In contrast there is no classical model
of the absorption process, and instead a quantum mechanical approach must be taken.

Quantum mechanical treatment

In a quantum mechanical treatment the scattering process is described by time-dependent perturbation

theory. The interaction between the incident radiation and sample is specified by an Hamiltonian HI ,
which produces transitions between the initial |ii and final | f i states. Here |ii and | f i refer to the
combined states of the X-ray field and sample. The number of transitions, W, per second between |ii
and | f i is given in first-order perturbation theory by Fermi’s Golden Rule as
2π  2
W=  Mi f  ρ(E f ) (A.3)
~
where the matrix element Mi f =h f |HI |ii, and ρ(E f ) is the density of states, defined such that ρ(E f )dE f is
the number of final states with energy in the interval dE f centered around E f . (The correct dimensions
of 1/[time] for W can be confirmed by inspection.)

Scattering
To evaluate the differential cross-section for scattering the number of transitions per second into the
solid angle ∆Ω needs to be found, and as we are mostly interested in elastic scattering, the restriction
E f = Ei needs to be placed.
Here we follow the standard method to calculate the density of states ρ(E f ), where it is assumed
that the total system (X-rays + sample) occupies a box of volume V. Periodic boundary conditions are
applied to the X-ray wavefunctions, resulting in a uniform density of states in wavevector of V/(2π)3.
By definition ρ(E f )dE f is the number of states with energy between E f and E f + dE f , which is equal
to the number of states with wavevectors between k f and k f + dk f . We can therefore write
 V 
ρ(E f )dE f = dk f
8π3
346 Scattering and absorption cross-sections

Fig. A.2 Illustration of the quantum mechanical derivation of (a) the differential scattering and (b) the absorption cross-sections.
For the differential scattering cross-section, one must integrate over the values of k f accessible within the solid angle element
∆Ω. For the photoelectric absorption cross-section all directions of the expelled electron must be integrated over. In both cases,
energy conservation is ensured by introducing a delta function in the integrand.
 347

or
 V  dk
f
ρ(E f ) = (A.4)
8π3 dE f
The differential cross-section can then be calculated from Eq. (A.2) and (A.3) as
!
dσ W∆Ω
=
dΩ Φ0 ∆Ω

where W∆Ω ≡ Isc is the number of transitions per second into ∆Ω. The restriction on elastic scattering
events is introduced by including a delta function δ(E f − Ei ) in Eq. (A.3), and then integrating over all
E f , which yields
Z 
2π  M 2 ρ(E ) δ(E − E ) dE
W∆Ω = (A.5)
~ if f f i f

According to Eq. (A.4)

 V   dk 
 f 
ρ(E f ) = 3
k2f   ∆Ω
8π dE f 

where the differential volume element dk f has been replaced by k2f dk f ∆Ω as indicated in Fig. A.2. The
expression for W∆Ω may be simplified considerably. Since E f = ~k f c, it follows that
 
 dk f 
k2f   = 1 E2 (A.6)
dE f  ~3 c3 f

With the flux given by Φ0 = c/V, we obtain

!  2 Z
dσ V 1
= |Mi f |2 E2f δ(E f − Ei ) dE f (A.7)
dΩ 2π ~4 c4

Absorption
Calculation of the absorption cross-section proceeds along similar lines, with two main differences.
The first is that the condition on the δ function that appears in Eq. (A.5) is altered, since the process is
no longer elastic. In an absorption process the incident photon expels an electron from an atom with
binding energy Eb . The difference between the energy of the incident photon, E = ~ω, and the binding
energy of the electron, Eb , is the kinetic energy of the photoelectron, Epe = ~2 q2 /2m. The second
difference is that there is no restriction on the direction of q, the wavevector of the photoelectron, so
that instead of integrating over the states in ∆Ω, as was the case for the scattering cross-section, it is
now necessary to integrate over the entire solid angle of 4π. Thus the absorption cross-section is
W4π
σa =
Φ0

with Z
2π  2
W4π =  Mi f  ρ(Epe ) δ(Epe − (E − Eb ))dEpe
~
348 Scattering and absorption cross-sections

The density of states for the photoelectron is evaluated using the same box normalization introduced
for the density of X-ray states in the case of scattering discussed above. The result for the photoelecton
is  V  dq !
ρ(Epe ) = 2
8π3 dEpe
where the factor of 2 allows for the two possible spin states of the electron. The absorption cross-section
is evaluated by replacing the volume element dq in the above by q2 sin θdqdθdϕ, with the integral taken
over the entire solid angle of 4π, as indicated in Fig. A.2(b). It follows that the absorption cross-section
is
Z 
2π V 2  M 2 δ(E − (E − E )) q2 sin θdqdθdϕ
σa = (A.8)
~c 4π 3 if pe b

where once again the incident flux is given by Φ0 = c/V.

It remains of course to calculate the matrix elements Mi f of the interaction Hamiltonian: this is
described in Appendix C for the Thomson scattering cross-section, and in Chapter 7 for the absorption
cross-section.

Further reading
Quantum Mechanics, A.I.M. Rae (Adam Hilger, 1986)
Quantum Mechanics, F. Mandl (John Wiley & Sons,1992)