Mößbauer E↵ect

First page should also contain: E. Dı́az and A. Ho

- Your names
- Group number
- date(s) of experiment 20.12.2014
- date of report submission
- version of report
 Abstract
A Mößbauer experiment was performed, using a 57 Co(Rh) radiation source, with three samples containing
the isotope 57 Fe: natural iron, FeSO4 ⇥ 7H2 O, and K4 [Fe(CN)6 ] ⇥ 3H2 O. The isomeric shifts of the three
samples were determined to be (0.07 ± 0.07)⇥10 7 eV, (0.60 ± 0.04)⇥10 7 eV and (0.09 ± 0.01)⇥10 7 eV,
respectively. The magnetic field at the nucleus and the magnetic moments of the first excited state of 57 Fe
in natural Fe were calculated to be 36 ± 2 T, and ⌥ (0.0526 ± 0.0085) µN and ⌥ (0.1577 ± 0.0254) µN
for mj,e = ± 1/2 and mj,e = ± 3/2 respectively. Additionally, the electrical field gradient at the nucleus for
FeSO4 ⇥ 7H2 O was determined to be (1.66 ± 0.08) ⇥ 10 7 eV. The obtained results generally show good
agreement with the expected values, falling within the determined error ranges and having a percentage
di↵erence within 15%.
Contents

1 Background Theory 2
1.1 Nuclear Shell Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Nuclear Recoil and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Mößbauer Spectrum Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Hyperfine Splitting by Magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.2 Hyperfine Splitting by Electric Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.3 Chemical Isomeric Shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.4 Applications with 57 Fe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2 Experimental Setup and Procedure 13

3 Analysis and Discussion 15

3.1 Experiment with Natural Fe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 Experiment with FeSO4 ⇥ 7H2 O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.3 Experiment with K4 [Fe(CN)6 ] ⇥ 3H2 O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4 Summary 21

1
1. Background Theory

The most crucial information that can be obtained from a sample of unknown material is its chemical
composition and its bonding structure, which can be used in turn to determine its physical and chemical
properties. However, prior to the development of quantum mechanics and nuclear physics, the methods of
extracting this information from a sample proved to be too complex or destructive to the sample itself, and
the discovery of the Mößbauer E↵ect, or recoilless nuclear resonance fluorescence, led to a unique method to
determine these properties within certain elements.
As the name suggests, recoilless nuclear resonance fluorescence is concerned with the transition between two
discrete nuclear energy states, resulting in the absorption and release of a high-energy photon, typically
a gamma ray photon. The principle was to use these quantized nuclear transitions to attenuate a beam
of mono-energetic gamma rays, and thus detect the presence of a particular element within a material,
hence nuclear resonance. However, in order for this resonance to occur, the fluorescence spectrum, or photon
energies emitted from excited nuclei going back to the ground state, must be the same energy as the absorbed
spectrum, such that the emitted photons can then be absorbed by other identical nucleus. This could only be
achieved when the target nuclei are bound to other atoms through chemical bonding and when the incident
gamma ray energy is sufficiently low, hence recoilless. The recoilless interaction is very important, as recoil
gives energy to the target nuclei and takes away energy from the emitted photon, thus e↵ectively changing
the absorption and the emission spectra. Without a recoilless transition, the attenuation of the incident beam
will not be significant enough to be detected. This phenomenon and its importance is discussed in further
detail in Section 1.2.
Once the conditions to observe the Mößbauer E↵ect are satisfied, additional information about the envi-
ronment in which the target nuclei are in can be obtained through observing the shifting and splitting of
the nuclear spectral line. This report attempts to provide an overview on the analysis capabilities of the
Mößbauer E↵ect and show their application, but first, an understanding of some basic concepts of nuclear
physics is required.

1.1 Nuclear Shell Model

Within an atomic nucleus, there exists quantized energy levels similar to the way in which atomic electrons are
quantized. As protons and neutrons are also fermions, like electrons, the physics governing their quantization
remains very similar to that of the electron orbitals in the atomic shell model. However, the potential well
in which the nucleus resides is a result of the strong nuclear force, as opposed to the Coulomb force. The
former is extremely short ranged and causes the nucleus to remain in a tight configuration. Due to smaller
distances between particles, the binding energy of the nucleus is much higher and the quantized energy levels
in the nucleus typically have much higher energies than its electronic equivalent.
Based on this theory, the protons and neutrons within the nucleus, or nucleons, form shells which fill upwards
in energy from the lowest available energy, similar to electrons in their ground state. Also analogous to the
electrons, the nucleons belonging to the outermost shell, called valence nucleons, typically determine the
physical properties of the entire nucleus. Since there exists two distinctly di↵erent particles in the nucleus,
protons and neutrons, a separate shell structure is assigned to each particle, one for protons and a separate
one for neutrons. This treatment is called the nuclear shell model.
However, within this model, di↵erent quantum numbers are required to describe each di↵erent nucleonic state,
due to the spin-orbit interaction, which can be seen as the e↵ect of the spin of a particle on its motion. This
interaction causes the original quantum numbers, the azimuthal quantum number, l, and the spin quantum
number, s, to no longer commute with the Hamiltonian operator, rendering them ine↵ective as quantum
numbers [1]. As such, the quantum numbers used are: the principle quantum number, n, the total angular
momentum quantum number, j, the total angular momentum projection quantum number, mj , and the parity

2
quantum number, p, which can be determined from the first set of quantum numbers as follows [1]:

n=n
j = |l ± s|
(1.1)
mj = ml + ms
l
p = ( 1)

where ml and ms are the projections of the quantum numbers, l and s, respectively. From quantum mechan-
ics, the discrete energy levels of the nucleus, E, can be detemined from the time-independent form of the
Schrödinger Equation, given as:
E =H b = Tb + Vb (1.2)
where Hb is the Hamiltonian operator, Tb is the kinetic energy operator, and Vb = V (r) is the potential
energy operator. However, for a complicated system like a multi-electron atom, the Schrödinger Equation
becomes impossible to solve analytically, but the transition energies have been experimentally determined
and were found to be typically within the kilo- (keV) to mega-electronvolt (MeV) range [2]. Additionally, it
is important to note that this model states, for single atoms, each j energy level is 2j + 1 degenerate.
From this theory, it becomes clear that the transition energies are characteristic properties of the nuclei
themselves and, as such, the photon energies that it absorbs and emits can be used to determine which
nucleus is present in the material. However, in practice, the nuclear resonance phenomenon, in which the
emitted photon from one de-excitation excites another identical nucleus, only occurs under very specific
conditions.

1.2 Nuclear Recoil and Implications

Due to the conservation of momentum and energy, the emission of a photon from the nucleus generates an
impulse on the nucleus, causing it to recoil. This results in a reduction of the energy of the emitted photon,
and a consequent addition to the vibrational energy of the nucleus, according to the following formula:
✓ ◆2
p2 1 E0
Erecoil ⇡ = (1.3)
2M 2M c

where p is the momentum of the emitted photon, M is the mass of the nucleus in question, E0 is the transition
energy of the two nuclear states, and c is the speed of light in vacuum. The approximation holds as it is
expected that the reduction of the photon energy (⇠ 10 2 – 102 eV) will be many orders of magnitude less
than its total energy (⇠ 105 – 107 eV) [3], and the e↵ect of this reduction on E0 is negligible for this calculation.
As a consequence of the recoil, the absorption spectrum is shifted to higher energies, or blue-shifted, and the
emission spectrum to lower energies, or red-shifted.
However, it remains unclear what e↵ect this will have on the nuclear resonance e↵ect, as the shift in energy
may be small enough that it is negligible. In order to determine this, a quantification of the linewidth of
the emission and absorption spectrum is required, which e↵ectively describes the maximum allowed energy
di↵erence between the emitted and absorbed photons in order for the nuclear resonance e↵ect to be observable.
The natural linewidth is a quantity that can be derived from the Heisenberg uncertainty principle for this
transition, as such:
h̄ h̄
E t ) E= = (1.4)
2 2⌧ 2
where ⌧ is the expectation value for the lifetime of the excited nuclear state, which is typically between
10 7 – 10 11 s [4], and is the natural linewidth. From Equation (1.4), the natural linewidth is typically
between 10 8 – 10 4 eV, which is many orders of magnitude smaller than the red- and blue-shifts of the
emission and absorption spectra as a result of nuclear recoil. Although these linewidths are typically Doppler-
broadened as a result of thermal excitations, or phonons, within the source and target, this broadening still
does not increase the linewidth to a value comparable with the shift due to recoil. As a result, this recoil
e↵ectively destroys the nuclear resonance e↵ect and the expected attenuation e↵ect on the incident beam
becomes unobservable.
However, Rudolf Mößbauer discovered that embedding the target nuclei within crystal structures provided
a way to eliminate this recoil entirely and experimentally observe the nuclear resonance e↵ect. Using the
Einstein model for solids, each atom in a crystal is bound in place by its neighbours, similar to a harmonic

3
oscillator. Under this theory, it was found that the energy of this quantum harmonic oscillator is quantized
in an analogous manner as electrons bound to a nucleus. The allowed vibrational energy levels, or phonon
energies, in the quantum harmonic oscillator can be expressed as:
✓ ◆
1
Ephonon = + n h̄!E , n = 0, 1, 2, ... (1.5)
2

where !E is the Einstein frequency of the solid, and is a property of the crystal [5]. Then, provided that
the recoil energy associated with the gamma ray emission is less than one quanta of vibrational energy,
Erecoil < h̄!E ⇡ 10 2 eV [6], then the harmonic oscillator cannot be excited and no energy is lost to
recoil. However, even this was not enough to achieve recoilless emission in most nuclei simply because
the recoil energy was too high. However, Equation (1.3) indicates that the recoil energy is related to the
transition energy. Thus, it could have been possible to find transition energies that were low enough that the
probability that Erecoil < h̄!E was high enough for this e↵ect to be observed, and this indeed was the case
when E0 < 100 keV for room temperatures [4]. Using this fact, the recoilless nuclear resonance fluorescence
was experimentally observed and documented in 1958 by Rudolf Mößbauer, giving this phenomenon its name,
the Mößbauer E↵ect [4].
The most useful consequence of the recoilless nuclear resonance fluorescence e↵ect is that it is capable of
resolving incredibly small energy di↵erences. This is due to natural linewidth of the absorption spectrum,
meaning that small deviations in energy from the resonance peak results in no absorption events, which in
turn, results in no observable attenuation of an incident radiation beam.

1.3 Mößbauer Spectrum Analysis

The Mößbauer E↵ect is not only capable of determining the presence of an element within a sample, but can
also determine the environment in which the target nuclei reside within the crystal. This is due to the fact
that the nucleus, being a charged particle with angular momentum, interacts with external electromagnetic
fields. This interaction e↵ectively changes the potential function, V (r), that the nucleus resides in and,
according to Equation (1.2), the result is that the quantized energy levels are altered. As such, the observed
spectral lines occur at di↵erent energies than the unperturbed case.
However, additional complications arise as the changes in the energy levels due to various is typically within
the micro-electronvolts (µeV) range, meaning that they require an extremely fine energy resolution to re-
solve. It turns out that not every excited nuclear state which meets the recoilless emission condition is also
capable of resolving these small energy di↵erences. This is a result of the Hesienberg uncertainty principle,
Equation (1.4), which e↵ectively restricts the energy resolution of the Mößbauer E↵ect based on the lifetime
of the excited state of the target nucleus.
Thus, in order to select an appropriate nucleus, the expected values for the nuclear energy level changes must
be evaluated. Consequently, in order to perform a quantitative analysis on the energy shifts as a function
of the potential, it is useful to examine the interaction energy between the nucleus and the electromagnetic
field. This examination requires the use of the Taylor expansion of the interaction potential function of the
field in question. The general form of the Taylor expansion of a function around a point, a, is:

f 0 (a) f 00 (a) 2 f (3) (a) 3 f (n) (a) n

f (x) = f (a) + (x a) + (x a) + (x a) + ... + (x a) + ... (1.6)
x=a 1! 2! 3! n!

which is especially useful in the analysis of complex functions around known equilibria points or of physical
configurations which cause the first and/or second terms to become zero.
The expansion can be applied to a 1/r field, such as a three-dimensional electromagnetic potential field,
V (r R), where r is the position at which the field is to be evaluated and R is the position of a point source.
Then, by making the appropriate simplifications, and expanding around R = 0 in Cartesian coordinates, the
result is as follows:
1 1 1 r·R 1 ⇣ 2
⌘
V (r R) / = 1/2
= + 3 + 5 3 (r · R) r2 R2 + ... (1.7)
R=0 |r R| R=0 r (1 + x) x=0 r r 2r

Typically, the first non-zero term of this series expansion determines the behaviour of the potential field and
any successive terms after it are reduced to corrective factors, and can be seen as negligible. This process is
called multipole expansion [7].

4
An additional e↵ect can be observed when there is a change in the electron density around the target nucleus
as a result of chemical bonding with other elements, which causes the entire nuclear energy spectrum to be
shifted to higher energies or lower energies. This phenomenon is called an isomeric shift, and the shift is also
typically within the meV range.

1.3.1 Hyperfine Splitting by Magnetic Fields

The first e↵ect that will be discussed here is the hyperfine splitting of the Mößbauer spectrum from the
presence of a magnetic field. When the nucleus resides within a non-zero magnetic field, the quantized
nuclear energy levels with the quantum number j become split into distinct levels for each quantum number
mj . In other words, the 2j + 1 degeneracy of the energy levels becomes lifted through the breaking of the
symmetry within the nucleus of the target atoms. Each of these newly split energy levels are denoted by
the total angular momentum projection quantum number, mj = j, j + 1, ..., j 1, j. This phenomenon
is called hyperfine splitting as it lifts an additional degeneracy from the fine structure of the shell model,
in which the degeneracy of the principle quantum number, n, is lifted into distinct levels for each quantum
number j. In order to determine the process behind this interaction and the magnitude of the energy shifts,
an understanding of magnetic fields and its application to the nucleus must first be developed.
A magnetic field, B, is typically described by the magnetic vector potential, A. As quantum mechanical
particles typically exhibit static internal magnetic fields, it is convenient to define the field, and the vector
potential describing it, as:
Z
µ0 J (R) 3
B =r⇥A where, A(r) = d R (1.8)
4⇡ V |r R|
where µ0 is the permeability of vacuum (4⇡ ⇥ 10 7 T m A 1 ), J is the current density or current per unit
cross-sectional area, and R is the point of integration inside a closed volume, V . At this point, it is sufficient
and convenient to view the nucleus as a spinning charge distribution such that it has a time-independent
current density due to this motion.
When the nucleus is in the presence of an external, uniform, time-independent magnetic field, B, described
by a magnetic vector potential, A, the interaction energy between them can be expressed as:
Z
1
Um = J (R) · A(r) d3 R (1.9)
2 V

Then, given that the current density distribution of the nucleus is confined such that r R, a multipole
expansion of the potential function, A, can be performed around R = 0, as detailed in Equation (1.7). The
result of this expansion on Equation (1.9), up to the second order term, is as follows:
Z  Z ✓ Z ◆ ✓ ◆
1 µ0 3 µ0 3 1
Um = J (R) · J (R) d R + r· R J (R) d R + O 3 d3 R
2 V 4⇡r V 4⇡r3 V r
Z ✓ ◆
µ0 1 J (R) 3 1
= · (µ ⇥ r) d R + O (1.10)
4⇡ 2 V r3 r3
✓ ◆
1
= µ · (r ⇥ A) + O 3
r
where µ is defined as the magnetic dipole moment of the nucleus. The results of Equation (1.10) come from
the fact that a nucleus has a fixed charge and a time-independent current density. The implications of these
facts can be summarized as [8]:
Z
J (R) d3 R = 0
V
Z Z (1.11)
1
r· R J (R) d3 R = r ⇥ R ⇥ J (R) d3 R = µ ⇥ r
V 2 V
The first portion of Equation (1.11) is consistent with modern electromagnetic theory as it indicates that the
monopole term is zero, or that the sum of all magnetic charges in the nucleus is zero. The second portion,
provided that the current density is non-zero at some position, shows that the interaction energy is dominated
by the dipole term. Then, it is useful to define the magnetic moment of the nucleus, µ, as such:
Z Z
1
µ= M (R) d3 R = R ⇥ J (R) d3 R (1.12)
V 2 V

5
where M (R) is the magnetization of the nucleus. The magnetic moment, µ, as expressed here is also called
the first moment of J (R).
By substituting Equation (1.8) into Equation (1.10), the interaction energy between the nucleus and the
magnetic field, or the resulting change in potential energy of the nucleus, can be written as:
✓ ◆
1
Umd = µ · B + O 3 (1.13)
r

where the higher order terms are considered to be negligible. This result shows that the interaction energy
is only appreciable provided that the nucleus has a non-zero magnetic moment. Thus, in order to determine
the magnitude of the splitting e↵ect created by the magnetic field, the magnetic dipole moment, µ, of the
nucleus must be determined. Classically, this value can be derived from a single proton, seen as a rotating
body with angular momentum, L, mass, mp , and charge, +e. From this viewpoint, the magnetic moment of
the proton can be expressed as:
e
µ= L (1.14)
2mp
However, this formulation does not account for the presence of other charge particles or the spin-orbit
interaction, which is required for a full quantum mechanical treatment [9]. As such, a g-factor, gn , sometimes
also called a Landé g-factor, is included to account for these e↵ects. Additionally, due to the fact that
the quantum mechanical angular momentum of the nucleus is dominated by its spin, it is convenient to
express Equation (1.14) by replacing the total angular momentum, L, with the nuclear spin operator or
b Since this spin is both directional and quantized, a further simplification can be made
isospin operator, I.
by choosing a coordinate system such that the axis which the vector quantity is projected onto is parallel to
the external field, typically denoted as z.
The result is the projection of the nuclear spin operator, Ibz , whose expectation value can be described by the
total angular momentum projection quantum number, mj , of the nuclear state it currently occupies. This
expression is as follows: !
D E X X
Ibz = Iz = h̄mj = h̄ p
mj + mj n
(1.15)
protons neutrons

where h̄ is the reduced Planck constant. This equation comes from the fact that the nucleus consists of protons
and neutrons. As both of these particles are classified as fermions, they each have their own shell structure
and for each of them, mj can only have half integer numbers. Additionally, since the values of mj are split
symmetrically into ±mj , the contribution from a coupled pair of nucleons is zero and the overall result can
be seen to simply be the addition of mj values of any unpaired nucleons within the nucleus [1]. While it may
be strange to assign a neutron contribution to an electromagnetic e↵ect, as a neutron is neutrally charged,
let it be sufficient enough to mention that the neutron itself is composed of other subatomic particles with
non-zero charge that collectively have a net charge of zero. These particles can be seen to allow the neutron
to contribute to hyperfine splitting observed with the Mößbauer E↵ect [10]. By adjusting Equation (1.14) for
these quantum e↵ects, the magnetic moment of an atomic nucleus can be expressed, as a quantum operator,
as such:
e b
b = gn
µ I (1.16)
2mp
It is important to note that, although Equation (1.16) was derived from a single proton, the e↵ect of having
multiple protons and neutrons in the nucleus is folded into the g-factor, gn , which then becomes a property
of a given nucleus in a given nuclear state. For this reason, it is useful to define two other physical constants,
the nuclear magneton, µN , and the gyromagnetic ratio, , as:
eh̄ e g n µN
µN = , = gn = (1.17)
2mp 2mp h̄

Now, with the quantum mechanical expression for the nuclear magnetic moment and the proper coordinate
system chosen, the shift in energy of the discretized nuclear energy levels can be calculated by considering
the change in potential energy as a change in the potential energy operator, Vb , in Equation (1.2). Due to
the nature of the potential energy operator, this adjustment is simply the addition of an interaction term to
the Hamiltonian operator, H b i , as such:
⇣ ⌘
E = H b +Hbi (1.18)

6
where, in this case, the interaction term is the magnetic dipole Hamiltonian operator, Hb D , which can be
expressed in terms of quantum operators by combining Equations (1.13), (1.16), (1.17), as such:
bi = H
bD = e b g n µN b
H b·B =
µ gn I ·B = Iz B = Ibz B (1.19)
2mp h̄
Finally, combining Equations (1.18) and (1.19), the energy di↵erence between the perturbed state and the
unperturbed state, ED , which is dependent on the strength of the magnetic field passing through the
b D , as such:
nucleus, can be calculated as the expectation value of H
D E D E
ED = bD
H = Ibz B = h̄mj B (1.20)

where the expectation value of Ibz is given by Equation (1.15).

In some materials, such as ferromagnetic materials, the natural spin orientation of the electrons within
the system are aligned to form a local magnetization, or a magnetic field, e↵ectively causing this e↵ect to
occur without the need to apply an additional magnetic field. This e↵ect then allows information about the
magnetic properties of the material to be extracted from a Mößbauer analysis. In summary, it can be said
that magnetic dipole interaction can only be observed in atoms where the nuclei have a non-zero magnetic
moment, µ / mj 6= 0, and where the electrons generate a non-zero magnetic field, B 6= 0.

1.3.2 Hyperfine Splitting by Electric Fields

The second e↵ect that will be discussed here is the hyperfine splitting of the Mößbauer spectrum from the
presence of an external electric field. Similar to the e↵ect of magnetic fields, electric field gradients also alter
the discrete nuclear energy levels within a nucleus, except in this phenomenon, each quantized nuclear energy
levels with the quantum number j only splits into distinct levels for each quantum number |mj |, and the
±mj levels remain degenerate. This is due to the way in which the electric fields interact with the nucleus,
which follows from the multipole expansion of this field. Thus, in order to determine the process behind this
interaction and the magnitude of the energy shifts, an understanding of electric fields and its application to
the atom must first be developed.
An electric field, E, is typically described by the electric scalar potential, . As it is common that atoms
have some degree of spherical symmetry, it is useful to define the field as a curl-less electric field, with its
corresponding potential, as:
Z
1 ⇢(R) 3
E= r where, (r) = d R (1.21)
4⇡✏0 V |r R|

where ✏0 is the permittivity of vacuum (8.854 ⇥ 10 12 C V 1 m 1 ), ⇢ is the charge density or charge per
unit volume, and R is the point of integration inside a closed volume, V . At this point, it is more useful
to view the electrons in the atom as clouds with charge densities proportional to the square of their orbital
2
wavefunction, | | , such as to obtain a continuous, time-independent charge density function.
When the nucleus is in the presence of an external, time-independent electric field, E, described by an electric
scalar potential , the interaction energy between them can be expressed as:
Z
1
Ue = ⇢(R) (r) d3 R (1.22)
2 V
Then, given that the nuclear charge density distribution is confined such that r R, a multipole expansion
of the potential function, , in Cartesian coordinates can be performed around R = 0, as detailed in
Equation (1.7). The result of this expansion on Equation (1.22), up to the third order term, is as follows:
Z  Z ✓ Z ◆
1 1 3 1 3
Ue = ⇢(R) ⇢(R) d R + r· R ⇢(R) d R
2 V 4⇡✏0 r V 4⇡✏0 r3 V
Z ⇣ ⌘ ✓ ◆
1 1 2 2 2 3 1
+ ⇢(R) 3 (r · R) r R d R+O 4 d3 R
4⇡✏0 r5 2 V r
Z ✓ ◆
1 1 ⇢(R) X 3 1 (1.23)
= 5
Q r r
ij i j d R + O 4
4⇡✏0 2 V r i,j
r
✓ ◆
1X 1
= Qij r2ij + O 4
6 i,j r

7
where Qij is defined as the components of the electric quadrupole moment tensor and the subscripts i and
j both represent the three Cartesian basis vectors, (x, y, z). The results of Equation (1.23) come from the
fact that the atom itself has neutral charge and that the charge distribution in the nucleus is spherically
symmetric. The implications of these facts can be summarized as [8]:
Z
⇢(R) d3 R = qtot = 0
V
Z
r· R ⇢(R) d3 R = r · p = 0 (1.24)
V
Z ⇣ ⌘ X Z X
2
⇢(R) 3 (r · R) r 2 R 2 d3 R = ⇢(R) 3ri Ri rj Rj ri rj R2 ij d3 R = Qij ri rj
V i,j V i,j

The first portion of Equation (1.24) comes directly from the condition that the atom as a whole is neutral,
and it states that the monopole term, or the total charge of the system, is zero. The second portion comes
from the condition that the nuclear charge density is spherically symmetric, and it states that the dipole
term, or more specifically, the electric dipole of the nucleus, p, is also zero. The third portion, provided
that the charge density is non-zero at some position, shows that the interaction energy is dominated by the
quadrupole term. Then, it is useful to define the components of the quadrupole tensor of the system, Qij , as:
Z
Qij = ⇢(R) 3Ri Rj R2 ij d3 R (1.25)
V

where ij is the Kronecker delta function, which is equal to 1 when i = j, and 0 otherwise. These components
of the quadrupole tensor, Qij , expressed here are also called the second moments of ⇢(R).
By substituting Equation (1.21) into Equation (1.23), the interaction energy between the nucleus and the
electric field, or the resulting change in potential energy of the nucleus, can be written as:
✓ ◆ ✓ ◆
1 X 1 1 X 1
Ueq = Qij rij · E + O 4 = Qij Vij + O 4 (1.26)
6 i,j=x,y,z r 6 i,j=x,y,z r

where higher order terms are considered to be negligible. Equation (1.26) shows that the interaction energy
is only appreciable provided that the system has a non-zero values within its quadrupole moment tensor. It
is important to note here that this interaction, in contrast to the magnetic hyperfine splitting, is dictated by
the gradient of the electric field, r · E, or EFG, instead of magnitude of the field itself. However, despite this
di↵erence, the treatment of this e↵ect can still proceed analogous to that in Section 1.3.1. The second form
of Equation (1.26) provides a more compact notation and represents the electric field gradient as the second
derivative of a potential function, V , such that rij · E = r2ij V = Vij , according to Equation (1.21).
Then, in order to determine the magnitude of the splitting e↵ect created by the EFG, the tensor components,
Qij , of the nucleus must be determined. The formulation provided in Equation (1.25) describes a classical
situation and must be restated in quantum mechanical terms in order to be applied to the hyperfine EFG
splitting e↵ect. This can be done by utilizing the Wigner-Eckart theorem, which states that tensor operators
acting on angular momentum eigenstates can be expressed as a product of two factors, one which is indepen-
dent of the angular momentum orientation and the other which is a Clebsch-Gordan coefficient, derived from
spherical harmonic equations [8]. Consequently, the quadrupole moment can be expressed as a function of
the nuclear spin, I = (Ix , Iy , Iz ), as follows:
✓ ⇣ ⌘ ◆
eQ 3 bb b b b 2
Qij = I I
i j + I I
j i ij I (1.27)
j (2j 1) h̄2 2
where Q is defined as the quadrupole moment of the nucleus, j is the total angular momentum quantum
number, h̄ is the reduced Planck’s constant, Ii and Ij are the projection of the nuclear spin along the axes
represented by i and j, respectively, and ij is the Kronecker delta function from before.
Further simplifications can be made by realizing that the EFG is generated by external charges, meaning
that the electric potential must satisfy the Laplace’s equation, and by choosing a coordinate system such that
the following statements are true: 0 1
Vxx 0 0
r2ij V = @ 0 Vyy 0 A
0 0 Vzz (1.28)
|Vxx |  |Vyy |  |Vzz |
2
r V = Vxx + Vyy + Vzz = 0

8
Now, with the quantum mechanical expression for the nuclear quadrupole moment tensor and the proper
coordinate system chosen, the remainder of this treatment follows identically to the magnetic hyperfine
splitting theory. The interaction energy can be seen to alter the potential energy, implying that the potential
energy operator, Vb , in Equation (1.2) is also modified by the external electric field gradient. This results
in a simple adjustment of the potential energy operator of the Schrödinger Equation, in the form of an
interaction term to the Hamiltonian operator, as shown in Equation (1.18). In this case, the interaction term
is the electric quadrupole Hamiltonian operator, H b Q , which can be expressed with quantum operators by
substituting Equations (1.27) and (1.28) into Equation (1.26), as such:

bQ = 1
bi = H eQ 3h ⇣ ⌘ ⇣ ⌘ ⇣ ⌘i
H V zz 3Ibz2 Ib 2 + Vyy 3Iby2 Ib 2 + Vxx 3Ibx2 Ib 2
6 j (2j 1) h̄2 2

1 eQ 3Vzz b 2 Vxx Vyy ⇣ b 2 ⌘
= 3Iz Ib 2 + Ix Iby2 (1.29)
6 j (2j 1) h̄2 2 Vzz
e QVzz h ⌘ ⇣ b 2 b 2 ⌘i
= b 2 Ib 2 +
2 3Iz 2
I+ + I
4j (2j 1) h̄
where ⌘ is called the asymmetry parameter, which accounts for any e↵ects that the nuclear charge distribution
has on the EFG and is defined as such:
Vxx Vyy
⌘= (1.30)
Vzz
and Ib+ and Ib are called the raising and lowering operators, which are introduced to replace the remaining
coordinate directions, removing the need to specify them, and are defined as such:

Ib+ = Ibx + iIby and Ib = Ibx iIby (1.31)

Due to the materials examined within this experiment, which uses a 57 Co source to excite 57 Fe nuclei, a
reasonable approximation can be made that ⌘ = 0, though this is generally not the case. As such, only the
first-order quadrupole e↵ects are analysed in this report. Then, the shift in the discrete nuclear energy levels,
as a result of the adjustment to the Hamiltonian shown in Equation (1.29), can be calculated as follows:
D E e QVzz ⇣ D E D E⌘ e QVzz ⇥ 2 ⇤
EQ = bQ
H = b 2 Ib 2
2 3 Iz =
4j (2j 1)
3mj j (j + 1) (1.32)
4j (2j 1) h̄
where EQ is the change in the nuclear energy level as a result of the electric quadrupole interaction with an
external electric field gradient, the expectation values for the spin operators, Ibz and Ib 2 , are given, respectively,
by Equation (1.15) and the following: D E
Ib 2 = h̄2 j (j + 1) (1.33)

It is important to note that Equation (1.32) becomes undefined when j = 0 and j = 1/2, meaning that the
nucleus must have j > 1/2 in order for this phenomenon to occur.
In some materials, the non-cubic structure of the valence electrons of the target nucleus or of surrounding
ions within the system generate an electric field gradient throughout the target nucleus. This breaking of
the cubic symmetry can result from the chemical or physical bonds created with other elements within the
material. This e↵ect then allows information about the crystal lattice structure and bonding characteristics
of the target atom within the material to be extracted using a Mößbauer analysis. In summary, it can
be said that electric quadrupole interaction can only be observed in atoms where the nuclei have a non-zero
quadrupole moment, Q 6= 0, and a nuclear spin such that j > 1/2, and where the electrons generate a non-zero
electric field gradient in an arbitrary orientation, r · E = Vzz 6= 0.

1.3.3 Chemical Isomeric Shifts

The third and final e↵ect that will be discussed here is the isomeric shift of the Mößbauer spectrum, which
results from di↵erences in the charge density distributions between the emitting nucleus and the absorbing
nucleus. Similar to the quadrupole splitting of the discrete nuclear energy levels, this phenomenon reacts
to the electric fields generated by the nucleus and the electrons. However, the result is that the discrete
energy levels are simply shifted to higher or lower energies, that is, without lifting any degeneracies within
the unperturbed nuclear energy levels. As the energy levels of the di↵erent j states are shifted by di↵erent
magnitudes, the final result in the Mößbauer spectrum is a shift of the absorption spectral line to a slightly
di↵erent energy than the source emission line.

9
As this phenomenon also considers electric charges, similar to Section 1.3.2, Equation (1.23) is still applicable
here. However, in this case, the interaction comes from a change in the charge density distribution, implying
that the monopole term is no longer zero. From this, the interaction energy, Uc , can be expressed as such:
Z  Z ✓ ◆
1 1 3 3 1
Uc = ⇢e (R) ⇢n (R) d R d R + O 2 (1.34)
2 V 4⇡✏0 r V r
where ⇢e (R) and ⇢n (R) represent the electronic and nuclear charge density distributions, respectively. How-
ever, as shown in Equation (1.24), this simply evaluates to the total charge density within the volume, qtot ,
and has no significant meaning by itself. As such, a comparison is made for ⇢e between the spectral line of the
source, or emitter, and that of the sample, or absorber, and another for ⇢n between the excited and ground
nuclear energy states in order to obtain useful information from this e↵ect. Applying these mathematically
to Equation (1.34) results in an adjusted formula, as follows:
Z " Z Re # Z " Z Re #
1 1 3 3 1 1
EM = ⇢e (R) ⇢n (R) d R d R ⇢e (R) ⇢n (R) d R d3 R (1.35)
3
2 VS 4⇡✏0 R Rg 2 VT 4⇡✏0 R Rg

where VS and VT represent the integration volume within a source nucleus and a target nucleus, respectively,
and Re and Rg represent the nuclear radii of the excited state and the ground state, respectively.
The inner integrals of Equation (1.35), which belongs to the nuclear charge distribution, will be expanded
and evaluated first. From the nuclear shell model, discussed in Section 1.1, the orbital radius of the excited
nuclear energy state can be seen to have a di↵erent quantum expectation value than that of the ground state.
Using this concept, the inner integral can be evaluated over the entire nucleus, in spherical coordinates, as
such: Z Re ✓ ◆ Re
1 3 1 4 3 1 ⌦ ↵ ⌦ 2↵
⇢n (R) d R = ⇡R Ze = Ze Re2 Rg (1.36)
4⇡✏0 r Rg 4⇡✏0 R 3 Rg 3✏ 0

where Ze represents the total charge of the nucleus, as a product of the atomic number of the nucleus and
an electronic charge. It is important to note here that this integral is the same for both the emitter and the
absorber nuclei, as the nuclear properties of the atom do not change as a result of this phenomenon, and can
be factored out of Equation (1.35).
Next, the outer integral of Equation (1.35), which belongs to the electronic charge distribution, will be
expanded and evaluated, using the result obtained in Equation (1.36). As the inner integral is now expressed
in terms that are not dependent on the radius, the outer integral becomes simply the total electronic charge
within the integration volume. Also, since this integration is over the volume of the nucleus, which is much
smaller than the electron shells, the total charge can be approximated as the electronic charge multiplied by
the probability of finding an electron within the nucleus, as such:
Z ✓ Z ◆
1 3 1 2 3 e 2 2
⇢e (R) d R = e | (R)| d R ⇡ 4⇡ | (0)| = 2⇡e | (0)| (1.37)
2 V 2 V 2
2
where | (R)| is the square of the electron wavefunction at the point R, which represents the probability
of finding an electron at that point. It is important to note here that the only electrons which have a
non-zero probability of being within the nucleus are the s-electrons, or electrons inside s-orbitals. As these
orbitals have spherically symmetric probability distributions, combined with the fact that the nucleus over
which the integration is taken is much smaller than the orbital, the wavefunction in Equation (1.37) can be
2
approximated as having a constant value of | (0)| . Then, the volume integral in spherical coordinates yields
the 4⇡ factor.
Finally, substituting Equations (1.36) and (1.37) into Equation (1.35) yields the expression for the chemical
isomeric energy shift, typically denoted with , as follows:
2⇡ ⇣ ⌘ ⌦ ↵ ⌦ ↵
2 2
= EM = Ze2 | T (0)| | S (0)| Re2 Rg2 (1.38)
3✏0
In some materials, the strong electronegativity, or affinity for electrons, of the elements bonded to the target
atoms within the system spatially distort the core electron orbitals, including the s-orbitals. This e↵ect then
allows information about other elements and the additional bonding characteristics of the target atom within
the material to be extracted using a Mößbauer analysis. In summary, it can be said that electric monopole
interaction can only be observed in atoms where there is⌦ a change
↵ ⌦ in↵ nuclear charge density, or expected
nuclear radii, between the excited and the ground state, Re2 6= Rg2 , and where there is a change in the
electronic charge density, or electron probability density, within the nucleus between the emitting nucleus
2 2
(source nucleus) and the absorbing nucleus (target nucleus), | S (0)| 6= | T (0)| .

10
 57
1.4 Applications with Fe
In order to perform a useful analysis using the Mößbauer E↵ect, the excitation scheme of the target nuclei
must be studied. As discussed in Section 1.1, the di↵erence in energies between the discrete nuclear states
are typically within the keV – MeV range, meaning that nuclear transitions emit high energy photons. These
photons are called gamma rays, or gamma radiation, and are a form of electromagnetic radiation with very
high frequencies (> 1018 Hz), i.e. very short wavelengths (< 10 12 m 1 ). The nuclear process in which they
are generated is aptly called gamma decay, and normally occurs after other types of decays, such as alpha
and/or beta decay, provided that the resulting nucleus has more internal energy than its ground state.
Depending on the isotope and element of the source nucleus, the gamma rays emitted by a nuclear transition
have a specific energy, with a very small natural linewidth, as discussed in Section 1.2, which in turn defines
the frequency of the photon. Thus, for identical nuclei, the emission frequency range should match exactly
with the absorption frequency of the other nucleus. This property, apart from being useful for identifying the
isotopes present in a material, is also crucial in the Mößbauer spectroscopy, since the source of the gamma
radiation must contain the exact same isotope as the excited nuclei contained within the target. In this
experiment, 57 Fe, a naturally-occurring isotope of iron, is the intended target nucleus, which then requires
the radioactive source to consist of 57 Co, whose nuclear decay scheme can be seen in Figure 1.1 [11].

57
Figure 1.1: Scheme of the Nuclear Decay of Co

57
Co is an unstable isotope with a half life of about 270 days, which decays into a specific excited state of 57 Fe
with a probability of 99.84%. This process occurs by electron capture, or EC, which is a type of radioactive
decay with the following reaction formula:

p+ + e ! n + ⌫ e (1.39)

in which a nuclear proton, p+ , is converted into a neutron, n, by the absorption of a K-shell electron, e ,
and the emission of an electron neutrino, ⌫e . Due to this process, it is sometimes classified as a form of beta
decay. Electron capture typically occurs when a nucleus is unstable due to having a relative high number of
protons but does not have enough energy to decay by emitting a positron.
The resulting nucleus, or daughter nucleus, is an excited form of 57 Fe, having a nuclear spin corresponding to
j = 5/2, which falls to its ground state, having a nuclear spin corresponding to j = 1/2, through gamma decay.
As shown in Figure 1.1, this process can occur following two di↵erent paths. It can either decay directly to
the stable state, with a probability of 15%, releasing a gamma ray of 136.3 keV, or it can decay first into an
intermediate state, having a nuclear spin corresponding to j = 3/2, and then into its stable state, emitting
two gamma rays, one of 122 keV and then another of 14.4 keV. The latter path has a probability of 85%,
making it the dominant decay path taken by the 57 Co to 57 Fe conversion. It is important to note that the
low-energy 14.4 keV transition, or the transition between the j = 3/2 and j = 1/2 levels, is responsible for the
Mößbauer E↵ect within this nucleus, for reasons detailed in Section 1.2.
Figure 1.2 shows the e↵ect on the j = 3/2 and j = 1/2 nuclear levels of 57 Fe when subjected to an external
magnetic field. The single resonant absorption peak from the degenerate transition becomes split into six
distinct resonant absorption peaks, due to the lifting of the degeneracy of the di↵erent mj states. In this
case, the j = 1/2 level splits into two distinct states, for mj = 1/2, +1/2, and the j = 3/2 level splits into four
distinct states, for mj = 3/2, 1/2, +1/2, +3/2. It is important to note that the quantized angular momentum
selection rule for dipole interactions only allows for transitions in which mj = 0, ±1 [12]. For this reason,
only six peaks are expected, as shown in Figure 1.2, instead of the eight when assuming all possible transitions
are allowed.

11
 Figure 1.2: The E↵ect of Magnetic Dipole Splitting on Nuclear Energy Levels 1/2 and 3/2 [12] [13]

Figure 1.3 shows the e↵ect on the j = 3/2 and j = 1/2 nuclear levels of 57 Fe when subjected to an external
electric field gradient. The single resonant absorption peak from the degenerate transition becomes split into
two distinct resonant absorption peaks, due to the lifting of the degeneracy of the di↵erent mj states. In this
case, the j = 1/2 level remains a single degenerate state and the j = 3/2 level splits into two distinct states,
for mj = ±1/2, ±3/2. It is useful to note that the quantized angular momentum selection rule for quadrupole
interactions only allows for transitions in which mj = 0, ±1 [14]. However, in this particular scenario, this
selection rule is not relevant as both transitions, shown in Figure 1.3, are allowed.

Figure 1.3: The E↵ect of Electric Quadrupole Splitting on Nuclear Energy Levels 1/2 and 3/2 [12] [13]

Both Figures 1.2 and 1.3 show the e↵ect on the discrete nuclear levels of 57 Fe when subjected to a change in
the electron density between the source atom and the target atom. Both the j = 1/2 and the j = 3/2 levels
become shifted in the same direction to either higher or lower energies. However, the states with a higher j
value are shifted to a greater extent, causing the isomeric shift to be measurable through Mößbauer analysis.
It should also be noted that the isomer shift is dependent on di↵erences between the source and the target
nuclei, and as such, the isomer shift is expected to be zero for identical nuclei in source and target. However,
in practice, the exact composition and construction of the source also a↵ect the isomer shift of the Mößbauer
spectra.
Although all of these e↵ects typically only shift the nuclear states by energies on the order of 10 8 eV, this
change can still be resolved by the Mößbauer E↵ect, for reasons discussed in Section 1.2. As the lifetime of
the j = 3/2 state of 57 Fe is relatively long (100 ns) [11], it has an energy resolution of up to 10 9 eV, according
to Equation (1.4), making it an ideal candidate nucleus for Mößbauer spectroscopy.

12
2. Experimental Setup and Procedure

A basic Mößbauer spectrometer, as shown in Figure 2.1, was assembled to conduct the experiments described
in this report. This device consists of the following [15]:
• a 57
Co(Rh) radiation source;
• a transducer, converting electrical energy to kinetic energy;
• a lead-lined chamber to house the source;
• a sample holder for the absorber materials;
• a radiation detector, consisting of a NaI-scintillation counter and a photomultiplier
• a pre-amplifier for the detector;
• an amplifier;
• a driver for the transducer;
• and a data acquisition module, (CMCA-550).

Figure 2.1: Experimental Setup for Mößbauer Experiment

This experiment examined three (3) di↵erent samples, which were used as the absorber materials: natural
iron (Fe), iron (II) sulfate heptahydrate (FeSO4 ⇥ 7H2 O) and potassium ferrocyanide (K4 [Fe(CN)6 ] ⇥ 3H2 O).
The CMCA-550 is a Multi Channel Analyser, MCA, that can work in two modes: Pulse-Height Analysis
(PHA) mode, and Multi-Channel Scaling (MCS) mode [16]. In PHA mode, the device categorizes pulses
from the detector according to their maximum current, or height, and increments the counter corresponding
to the appropriate channel. The relation between the channel numbers and the pulse height is linear, and for
this reason, when coupled with a NaI-scintillation counter, this mode of operation is useful in measuring the
energy spectrum of the incident radiation. In MCS mode, the device categorizes the raw pulse count from
the detector, typically filtered to be within a range of pulse heights, according to the voltage provided by
a second input. The relation between the channel numbers and the secondary voltage is also linear, and as
such, this mode of operation is useful in measuring e↵ects on the radiation intensity from changes in control
variables, such as transducer voltage.
For the first part of this experiment, the 57 Co source spectrum was acquired by removing any absorbers from
the sample holder, and switching the MCA to PHA mode. The data acquisition time was chosen to be 5
minutes, in order to ensure good statistics on the measured spectrum. This spectrum was used to calibrate

13
the detector, relating pulse height to gamma ray energy. Based on the result of this calibration, an electronic
filter was applied such that the device only counted pulse heights corresponding to an energy range around
14.4 keV. This energy was selected based on the theory of the Mößbauer E↵ect applied to 57 Fe, as discussed
in Section 1.4. Then, the MCA was switched to MCS mode, with the secondary input being the output
of the driver unit controlling the transducer. This ensures that each channel in the MCA corresponds to a
specific source velocity. Afterwards, the attenuation of the incident radiation beam by the three absorbers
was measured, with the driver unit adjusted such that the ranges of velocities were 9 < v < 9 mm s 1 ,
5 < v < 5 mm s 1 , and 3 < v < 3 mm s 1 for the natural iron, FeSO4 ⇥ 7H2 O, and K4 (Fe(CN)6 )⇥3H2 O
samples, respectively. For all measurements taken in this report, data was collected and recorded using the
software Wissoft 2003 [15].

14
3. Analysis and Discussion

Figure 3.1 shows the measured gamma spectrum of the 57 Co source. Lorentzian fits were made for the most
significant peaks in order to convert the x-axis from channel numbers to energy deposited in the detection
volume. As mentioned in Section 1.4, 57 Co decays into the excited state of the isotope 57 Fe, which then
proceeds to its ground state through the emission of a 122 keV gamma photon and a 14.4 keV gamma
photon. These two photon energies can be seen as peaks in Figure 3.1, labelled accordingly. The 136 keV
gamma peak is not observable due to the fact that it is so close to the 122 keV peak and that only 15% of
the decay events in the source result in the release of this photon [11].
Another significant peak is observed at 90.1 keV, which does not belong to the gamma spectrum expected
from 57 Co. This peak can be considered as an escape peak [17], which is specific to the incident radiation
energy in this experiment in combination with the NaI detector. Due to the relatively low energy of the
122 keV gamma photons, they typically do not penetrate far into the detection volume before an interaction
occurs. However, they still have enough energy that this interaction can eject a K-shell electron from the
iodine atoms inside the detector, which then emits a characteristic x-ray as another electron drops to replace
it. As this x-ray has a relatively high probability of leaving the detector without being re-absorbed by
the detection volume, due to its proximity to the surface of the detector, this energy becomes lost. The
K-absorption edge of iodine is 33 keV, which is subtracted from the 122 keV photon, resulting in a total
energy deposition of approximately 89 keV, which corresponds well with the additional peak observed in
the experimental calibration spectrum, shown in Figure 3.1. The 20.2 keV peak corresponds to the K-↵
characteristic x-ray of rhodium (Rh) [18], which is expected due to the Rh content in the 57 Co(Rh) source
used in this experiment.

57
Figure 3.1: Measured Energy Spectrum of Co Source (blue crosses) and Gaussian Fits (red lines)

3.1 Experiment with Natural Fe

Figure 3.2 shows the Mößbauer spectrum acquired from the 57 Fe present in the natural iron sample. The
transducer velocities on the x-axis of the figure can be converted to its corresponding energy shift by combining

15
the Doppler e↵ect, produced when a wave source is in relative motion, and the photon dispersion relation,
as follows: ✓ ◆
c
E = Es (3.1)
c vs
where Es is the energy of the radiation emitted by the source when stationary (14.4 keV), c is the speed of
light and vs is the velocity of the source. Since the shift in energy is usually very small compared to Es ,
it is more common to express these values as the di↵erence in energy, E = E Es . For this experiment,
this transformation means that a transducer velocity of 0.0 mm s 1 in the figure represents the 14.4 keV
Mößbauer line of 57 Fe.

57
Figure 3.2: Mößbauer Spectrum of Fe in Natural Iron

Table 3.1: Experimental Values for Absorption Peaks in Natural Fe Mößbauer Spectrum

Peak Number vs , [mm s 1 ] E [⇥10 7 eV]

1 (leftmost) 5.63 ± 0.07 2.70 ± 0.03
2 3.23 ± 0.07 1.55 ± 0.03
3 0.70 ± 0.07 0.34 ± 0.03
4 0.98 ± 0.07 0.47 ± 0.03
5 3.52 ± 0.07 1.69 ± 0.03
6 (rightmost) 5.91 ± 0.07 2.84 ± 0.03

It is important to note that, due to the nature of the experimental setup, it is unknown whether the lower
channel numbers from the MCA represent negative or positive velocities. For the purposes of this report,
an assumption was made in producing Figure 3.2, and all other similar figures, such that the lower channel
numbers always represent negative velocities. Since most of the Mößbauer e↵ects are measured in terms of
a change in energy, this assumption does not a↵ect the validity of the results given in this report. However,
this does a↵ect the sign of the isomer shift, , and as such, they are reported as an absolute value when
comparing them against expected values.
Then, based on the Mößbauer phenomena, discussed in Section 1.3, and its application to our experimental
nucleus, discussed in Section 1.4, the values for the magnetic dipole splitting (MDS) and isomeric shift (IS)
were extracted from Figure 3.2. These values, given in change in source velocity, | vs |, and transition energy,
|ET |, are shown in Table 3.2.

Table 3.2: Experimental Results for Mößbauer E↵ects in Natural Fe

E↵ect Transition, mj,g ! mj,e | vs |, [mm s 1 ] |ET | [⇥10 7 eV]

± 1/2 ! ± 3/2 5.77 ± 0.14 2.77 ± 0.07
MDS ± 1/2 ! ± 1/2 3.38 ± 0.14 1.62 ± 0.07
± 1/2 ! ⌥ 1/2 0.84 ± 0.14 0.41 ± 0.07
IS — 0.14 ± 0.14 0.07 ± 0.07

16
The isomeric shift was calculated from each symmetric pair of peaks, shown in Figure 3.2, by using the
displacement of their mean from 0.0 mm s 1 . In this case, each pair provided the same mean value, as expected
since the hyperfine splitting of the nuclear energy levels is expected to be symmetric about the energy level
after applying the isomeric shift. This mean velocity was then translated into energy using Equation (3.1),
where the errors were taken to be half of the channel width, determined from the experimental setup. This
is a reasonable assumption due to the extremely good resolution of the Mößbauer lines, as discussed in
Section 1.2. However, this error is on the same order as the measured result for the isomeric shift in natural
iron, as shown in Table 3.2, leading to an ambiguous interpretation of the result. For clarity, as the isomeric
shift was calculated as an average over three values, one for each symmetric pair of peaks in Figure 3.2, it
gives more weight to the fact that an isomeric shift is indeed present, i.e. it is a non-zero value.
It is also important to note that the transition energies, |ET |, are expressed relative to the Mößbauer line
after applying the isomeric shift. Using this system, the symmetry expected from the MDS and electric
quadrupole splitting (EQS) e↵ects are restored and future calculations involving these values become less
complicated.
From the values given in Table 3.2, the magnetic field at the nucleus of natural iron can be calculated using
Equation (1.20), as follows:

ET = ED,e ED,g = g h̄mj,g B e h̄mj,e B (3.2)

where the subscripts e and g represent the excited and ground states, respectively. The gyromagnetic ratio
of the ground state of 57 Fe, g /2⇡, was calculated from the magnetic moment of its ground state, using
Equations (1.16) and (1.17), as such:
g µg 1
= = 1.38 MHz T (3.3)
2⇡ hmj,g

where µg was taken to be 0.0903 µN (µN = 5.0508 ⇥ 10 27 A m2 ) [15]. These values were substituted into
Equation (3.2) and the resulting magnetic field strength, B, was calculated for each pair of transition energies
in Table 3.2, in order to remove the e term, and subsequently averaged. Additionally, using the calculated
values for B for each transition energy, the corresponding gyromagnetic ratios, e /2⇡, were calculated and
subsequently averaged. Then, the magnetic moments of the excited states of 57 Fe, µe , were determined, by
substituting the excited state parameters into Equation (3.3). The results from these procedures are given
in Table 3.3, along with all other relevant results from this experiment.

Table 3.3: Experimental and Expected Values for Mößbauer Analysis of Natural Fe

Property Experimental Value Expected Value Percent Di↵erence

| | (0.07 ± 0.07) ⇥ 10 7 eV 0.06 ⇥ 10 7 eV [19] [20] 17%
B 36 ± 2 T 33.0 T [20] 9%
e /2⇡ 0.8 ± 0.1 MHz T 1 — —
µe (mj,e = ± 1/2) ⌥ (0.0526 ± 0.0085) µN ⌥ 0.0516 µN [21] 2%
µe (mj,e = ± 3/2) ⌥ (0.1577 ± 0.0254) µN ⌥ 0.153 µN [22] 3%

From Table 3.3, it can be seen that the experimental values are consistent with the expected values, insomuch
as they generally fall within the error ranges of the calculated values. It should be noted that natural iron
is not expected to show an isomer shift due to the nature of its metallic bonding, but a finite value for
is measured. This can be attributed to the fact that the radioactive source used for this experiment is
57
Co(Rh) [19]. The rhodium casing of the source actually introduces a minor shift in the 14.4 keV Mößbauer
line, which is undetectable in 57 Co spectrum acquired for calibration, but the extremely fine energy resolution
of the Mößbauer E↵ect is capable of detecting it. The presence of Rh within the source generates an isomer
shift of +5.807 ⇥ 10 9 eV (+0.1209 mm s 1 ) [19], which must be accounted for when analysing all isomeric
shifts observed in this report.
The fact that the Mößbauer spectrum of natural iron exhibits magnetic dipole splitting of the nuclear fine
structure, even without an externally applied magnetic field, implies that there exists an internal magnetic
field inside the crystal. From solid state theory and empirical data, iron is a classified as a ferromagnetic
material. As such, it consists of a large collection of distinct magnetized domains with their magnetization
vector pointing in arbitrary directions. This creates local magnetic fields on the atomic scale, but the
material remains magnetically neutral on the macroscopic scale. However, as metallic bonding creates a
homogeneous electron gas from the valence electrons, no electric fields are generated in the crystal. Therefore,

17
the experimental data is consistent with the expected results, in that the magnetic dipole splitting is observed
but not the quadrupole splitting. As discussed earlier, the observed isomer shift is due to the construction
of the source, specifically related to its Rh content.

3.2 Experiment with FeSO4 ⇥ 7H2 O

Figure 3.3 shows the Mößbauer spectrum acquired from the 57 Fe present in the iron (II) sulfate heptahydrate
(melanterite) sample, with a chemical composition of FeSO4 ⇥ 7H2 O.

57
Figure 3.3: Mößbauer Spectrum of Fe in FeSO4 ⇥ 7H2 O

Table 3.4: Experimental Values for Absorption Peaks in FeSO4 ⇥ 7H2 O Mößbauer Spectrum

Peak Number vs , [mm s 1 ] E [⇥10 7 eV]

1 (left) 2.97 ± 0.04 1.43 ± 0.02
2 (right) 0.47 ± 0.04 0.23 ± 0.02

Then, similarly to the treatment in Section 3.1, the values for the electric quadrupole splitting (EQS) and
isomeric shift (IS) were extracted from Figure 3.3. These values, given in change in source velocity, | vs |,
and transition energy, |ET |, are shown in Table 3.5.

Table 3.5: Experimental Results for Mößbauer E↵ects in FeSO4 ⇥ 7H2 O

E↵ect Transition, mj,g ! mj,e | vs |, [mm s 1 ] |ET | [⇥10 7 eV]

± 1/2 ! ± 3/2 1.72 ± 0.08 0.83 ± 0.04
EQS
± 1/2 ! ± 1/2 1.72 ± 0.08 0.83 ± 0.04
IS — 1.25 ± 0.08 0.60 ± 0.04

From the values given in Table 3.5, the electric field gradient at the locations of the 57 Fe nuclei can be
calculated using Equation (1.32). The value for the electric quadrupole moment of 57 Fe, Q, was taken as
0.29 ⇥ 10 22 mm2 [15]. The result of this calculation is shown in Table, along with all other relevant results
from this experiment.

Table 3.6: Experimental and Expected Values for Mößbauer Analysis of FeSO4 ⇥ 7H2 O

Property Experimental Value Expected Value Percent Di↵erence

| | (0.60 ± 0.04) ⇥ 10 7 eV 0.614 ⇥ 10 7 eV [19] [20] 2%
Q (1.66 ± 0.08) ⇥ 10 7 eV 1.54 ⇥ 10 7 eV [20] 8%
Vzz or r · E (1.14 ± 0.05) ⇥ 1022 V m 2
1.15 ⇥ 1022 V m 2 [23] 1%

18
From Table 3.6, it can be seen that the experimental values are consistent with the expected values, as they
fall within the error ranges of the calculated values. It should be noted that the actual expected value of
the isomer shift in FeSO4 ⇥ 7H2 O is +6.72 ⇥ 10 8 eV (relative to ↵–Fe) [20], but the +5.807 ⇥ 10 9 eV [19]
shift from the Rh content in the source must be subtracted in order to accurately describe the experimental
setup, giving a final expected value of +0.614 ⇥ 10 7 eV.
The fact that the Mößbauer spectrum of FeSO4 ⇥ 7H2 O exhibits electric quadrupole splitting of the nuclear
fine structure, even without an external electric field, implies that there exists an internal electric field gradient
inside the crystal. From crystal structure analysis, the iron (II) sulfate molecule contains iron molecules in
their 2+ oxidation state surrounded by the polar water molecules, H2 O. While the lattice typically forms
cubic symmetry around the iron atom, the presence of the seventh water molecule breaks this symmetry,
creating a local electric field gradient around the iron atom. Additionally, from the electric fields generated
by the polar molecules, it is expected that the K-shell electrons of the iron atom are also a↵ected by these
electric field gradients. However, these polar molecules have no inherent magnetization vector and no internal
magnetic fields should be generated. Therefore, the experimental data is consistent with the expected results,
in that both the quadrupole splitting and an isomeric shift are observed but not the magnetic dipole splitting.

3.3 Experiment with K4 [Fe(CN)6 ] ⇥ 3H2 O

57
Figure 3.4 shows the Mößbauer spectrum acquired from the Fe present in the potassium ferrocyanide
sample, having a chemical formula of K4 [Fe(CN)6 ] ⇥ 3H2 O.

57
Figure 3.4: Mößbauer Spectrum of Fe in K4 [Fe(CN)6 ] ⇥ 3H2 O

Table 3.7: Experimental Values for Absorption Peaks in K4 [Fe(CN)6 ] ⇥ 3H2 O Mößbauer Spectrum

Peak Number vs , [mm s 1 ] E [⇥10 7 eV]

1 0.18 ± 0.02 0.09 ± 0.01

Then, similarly to the treatment in Section 3.1, the value for the isomeric shift (IS) was extracted from
Figure 3.4. These values, given in change in source velocity, | vs |, and transition energy, |ET |, are shown in
Table 3.8.

Table 3.8: Experimental Results for Mößbauer E↵ects in K4 [Fe(CN)6 ] ⇥ 3H2 O

E↵ect Transition, mj,g ! mj,e | vs |, [mm s 1 ] |ET | [⇥10 7 eV]

IS — 0.18 ± 0.02 0.09 ± 0.01

From the values given in Table 3.8, the isomeric shift of the 57 Fe within K4 [Fe(CN)6 ]⇥3H2 O can be extracted,
as shown in Table 3.9.

19
 Table 3.9: Experimental and Expected Values for Mößbauer Analysis of K4 [Fe(CN)6 ] ⇥ 3H2 O

Property Experimental Value Expected Value Percent Di↵erence

| | (0.09 ± 0.01) ⇥ 10 7 eV 0.078 ⇥ 10 7 eV [19] 15%

From Table 3.6, it can be seen that the experimental values are consistent with the expected values, as they
fall within the error ranges of the calculated values. It should be noted that the actual expected value of the
isomer shift in K4 [Fe(CN)6 ] ⇥ 3H2 O is 2.02 ⇥ 10 9 eV (relative to ↵–Fe) [19], but the +5.807 ⇥ 10 9 eV [19]
shift from the Rh content in the source must be subtracted in order to accurately describe the experimental
setup, giving a final expected value of 0.078 ⇥ 10 7 eV.
The fact that the Mößbauer spectrum of K4 [Fe(CN)6 ] ⇥ 3H2 O exhibits an isomeric shift but neither form of
hyperfine splitting implies that an internal uniform electric field is distorting the K-shell electron densities
within the iron atoms in this crystal. From crystal structure analysis, potassium ferrocyanide has cubic
symmetry and the ferrocyanide complex consists of a central iron atom in the 2+ oxidation state bonded to
six cyanide ligands in octahedral symmetry. The potassium ions are then positioned with cubic symmetry
around this structure. Due to the cyanide ligands, the water molecules tend to reside closer to the potas-
sium atoms than to the iron atoms within this crystal, meaning that they do not contribute to forming an
electric field gradient as is the case with FeSO4 ⇥ 7H2 O. The symmetrical positive charge distribution of
the potassium ions symmetrically reduce the electron density within the iron nucleus as the electrons are
attracted outwards. Additionally, this crystal does not inherently have any magnetization and no internal
magnetic field is generated. Therefore, the experimental data is consistent with the expected results, in that
an isomeric shift is observed but not the magnetic dipole splitting nor the quadrupole splitting.

20
4. Summary

The purpose of this experiment was to describe and demonstrate the utility of the Mößbauer E↵ect, or
recoilless nuclear resonance fluorescence e↵ect, in crystal structure analysis and material characterization.
The 57 Fe nucleus was shown to be an e↵ective target for the Mößbauer analysis, due to the presence of a long-
lived, low-energy nuclear transition as well as a relatively long-lived and easily obtained radioactive parent
nucleus, 57 Co. Using a 57 Co(Rh) source, three material containing 57 Fe were analysed for any magnetic
dipole splitting, electric quadrupole splitting, and isomer shift e↵ects: natural iron (Fe), iron (II) sulfate
heptahydrate (FeSO4 ⇥ 7H2 O), and potassium ferrocyanide (K4 [Fe(CN)6 ] ⇥ 3H2 O). The experimental results
were found to generally agree with the expected values to within 15%, for values with magnitudes of less than
10 6 eV, highlighting the extremely fine energy resolution of the Mößbauer E↵ect on measuring nuclear state
transition energies. All findings in this report are validated by solid-state theory and the crystal structure
analysis of these well-documented materials.
Overall, this report e↵ectively demonstrates the usefulness of the Mößbauer E↵ect for determining the physical
and chemical properties of specific elements within crystal structures, as well as the electric and magnetic
properties of the bulk material containing these elements. For this reason, the discovery of the Mößbauer
E↵ect has contributed greatly to the development and applicability of quantum and solid-state theory, and
continues to be used in modern scientific research as a technique of classifying unknown materials, especially
where traditional sample collection is not feasible.

21
Bibliography

[1] B. R. Martin, Nuclear and Particle Physics: An Introduction. John Wiley and Sons Inc., 2006.
[2] W. R. Leo, Techniques for Nuclear and Particle Physics Experiments. Springer-Verlag, 2nd ed., 1994.
[3] B. Brandsden and C. Joachain, Physics of Atoms and Molecules. Prentice Hall, 2nd ed., 2003.
[4] R. Mössbauer, The Rudolf Mössbauer Story. Springer-Verlag, 2012.
[5] Y. Chen and D. Yang, Mössbauer E↵ect in Lattice Dynamics: Experimental Techniques and Applications.
Wiley-VCH Verlag GmbH, 2007.
[6] D. Rogers, Einstein’s Other Theory: The Planck-Bose-Einstein Theory of Heat Capacity. Princeton
University Press, 2005.
[7] D. J. Griffiths, Introduction to Electrodynamics. Prentice-Hall Inc., 3rd ed., 1999.
[8] C. P. Slichter, Principles of Magnetic Resonance. Springer-Verlag, 3rd ed., 1990.
[9] R. B. Singh, Introduction to Modern Physics. New Age International, 2002.
[10] J. T. Cremer, Neutron and X-ray Optics. Elsevier Inc., 1st ed., 2013.
[11] H. Z. Berlin, “57 fe mössbauer spectroscopy.” http://www.helmholtz-berlin.de/forschung/oe/ee/
solare-brennstoffe/analytische-methoden/57fe-moessbauerspektroskopie_en.html.
[12] J. D. King, “Development of a mössbauer spectrometer,” tech. rep., Department of Physics and
Astronomy, Union College, 2006. http://ohm.bu.edu/~pbohn/REPAIR__Advanced_Lab/Mossbauer/
Reference/Mossbauer%20Setups.pdf.
[13] D. J. Bland, “A mössbauer spectroscopy and magnetometry study of magnetic multilayers and oxides,”
2003. http://www.cmp.liv.ac.uk/frink/thesis/thesis/thesis.html.
[14] B. R. Suits, Handbook of Applied Solid State Spectroscopy, vol. XIX. Springer, 2006.
[15] B. Gompf, Mößbauer E↵ect. Advanced Physics Lab. Universität Stuttgart.
[16] Wissenschaftliche Elektronik GmbH, Data Acquisition Module CMCA-550 Instruction Manual, 2.2 ed.
[17] E. Bassaldi, A. von Kienlin, et al., “Ground-based calibration and characterization of the fermi gamma-
ray burst monitor detectors,” tech. rep., Max-Planck-Institut für extraterrestrische Physik, 2008.
[18] Kaye and Laby, “Tables of physical and chemical constants.” http://www.kayelaby.npl.co.uk/
atomic_and_nuclear_physics/4_2/4_2_1.html.
[19] G. Belozerski, Mössbauer Studies of Surface Layers. Elsevier, 1993.
[20] J. Duarte and S. Campbell, “Mössbauer spectroscopy,” tech. rep., Massachusetts Institute of Technology,
MA 02142, 2009. http://web.mit.edu/woodson/Public/8.14finalpapers/Duarte_mossbauer.pdf.
[21] D. V. Perepelitsa, “Mössbauer spectroscopy of 57 fe,” tech. rep., MIT Department of Physics, 2007.
http://web.mit.edu/dvp/www/Work/8.14/dvp-mossbauer-paper.pdf.
[22] P. B. B. Srivastava, Fundamentals of Nuclear Physics. Rastogi Publications, 1st ed., 2006.
[23] D. Henderson et al., eds., Physical Chemistry: An Advanced Treatise - Molecular Properties, vol. 4.
Academic Press, 1970.

22