Professional Documents
Culture Documents
Textbook Molecular Spectroscopy Experiment and Theory From Molecules To Functional Materials Andrzej Kolezynski Ebook All Chapter PDF
Textbook Molecular Spectroscopy Experiment and Theory From Molecules To Functional Materials Andrzej Kolezynski Ebook All Chapter PDF
https://textbookfull.com/product/functional-molecular-materials-
an-introductory-textbook-first-edition-artizzu/
https://textbookfull.com/product/functional-molecular-materials-
an-introductory-textbook-1st-edition-matteo-atzori/
https://textbookfull.com/product/molecular-symmetry-and-group-
theory-approaches-in-spectroscopy-and-chemical-reactions-1st-
edition-r-c-maurya/
https://textbookfull.com/product/molecular-symmetry-and-group-
theory-approaches-in-spectroscopy-and-chemical-reactions-1st-
edition-r-c-maurya-2/
Molecular Spectroscopy, Second Edition Jeanne L. Mchale
https://textbookfull.com/product/molecular-spectroscopy-second-
edition-jeanne-l-mchale/
https://textbookfull.com/product/scientific-inquiry-in-
mathematics-theory-and-practice-andrzej-sokolowski/
https://textbookfull.com/product/basic-cardiovascular-physiology-
from-molecules-to-translational-medical-science-pasquale-
pagliaro/
https://textbookfull.com/product/molecular-and-laser-
spectroscopy-advances-and-applications-volume-2-1st-edition-v-p-
gupta-editor/
https://textbookfull.com/product/foundations-of-the-theory-of-
parthood-1st-edition-andrzej-pietruszczak/
Challenges and Advances
in Computational Chemistry and Physics 26
Series Editor: Jerzy Leszczynski
Andrzej Koleżyński
Magdalena Król Editors
Molecular
Spectroscopy—
Experiment and
Theory
From Molecules to Functional Materials
Challenges and Advances in Computational
Chemistry and Physics
Volume 26
Series editor
Jerzy Leszczynski
Department of Chemistry and Biochemistry
Jackson State University, Jackson, MS, USA
This book series provides reviews on the most recent developments in computa-
tional chemistry and physics. It covers both the method developments and their
applications. Each volume consists of chapters devoted to the one research area.
The series highlights the most notable advances in applications of the computa-
tional methods. The volumes include nanotechnology, material sciences, molecular
biology, structures and bonding in molecular complexes, and atmospheric
chemistry. The authors are recruited from among the most prominent researchers
in their research areas. As computational chemistry and physics is one of the most
rapidly advancing scientific areas such timely overviews are desired by chemists,
physicists, molecular biologists and material scientists. The books are intended for
graduate students and researchers.
Editors
Molecular Spectroscopy—
Experiment and Theory
From Molecules to Functional Materials
123
Editors
Andrzej Koleżyński Magdalena Król
Faculty of Materials Science and Ceramics Faculty of Materials Science and Ceramics
AGH University of Science and Technology AGH University of Science and Technology
Kraków, Poland Kraków, Poland
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
The idea of this book originated in some discussions we carried out during one
of the conference we had the pleasure to organize, namely XIVth International
Conference on Molecular Spectroscopy, Kraków-Białka Tatrzańska, Poland in
2017, which was the last conference in a series. The history of these conferences
spans over 20 years. In 1991, Prof. Henryk Ratajczak from Wroclaw University,
Poland, the originator of the idea, organized the first one in a series (as a national
conference with international participation) with the intention to have the following
ones every two years. Similar conferences devoted to Vibrational Spectroscopy in
Materials Science were initiated in 1996 by Prof. Mirosław Handke from AGH
University of Science and Technology in Krakow, Poland. For a few years, these
conferences were organized independently and in 2009 were merged into a single
one, organized every two years, alternately by University of Wrocław and AGH
University of Science and Technology. The last conference was organized in 2017
in Białka Tatrzańska and gathered above three hundred participants (almost a third
from abroad) and was devoted to practical application of various spectroscopic
methods to study different types of materials and showed that modern spectroscopy
is an invaluable tool in recent scientific endeavor to understand the structure and
properties of various natural and synthetic materials as well as to create new or to
modify existing ones in order to obtain better suited (functionalized) materials
required by modern industry.
There are a lot of books devoted to various issues related to either spectroscopic
methods themselves or their use in the characterization of diverse classes of
materials exhibiting often very different properties. However, the books containing
comprehensive, up-to-date description of theory, experimentation, and simultane-
ous use of different spectroscopic methods to study the properties of a given type of
materials, provided by leading specialists in the field for various classes of materials
are still rare and sought after.
This book is intended to provide such comprehensive description of materials
characterization by means of a wide spectrum of modern spectroscopic methods
and is aimed at scientists interested in theory and practical application of various
spectroscopic methods in their work. It starts with a short introduction to theoretical
v
vi Preface
vii
viii Contents
ix
x Contributors
Andrzej Koleżyński
1.1 Introduction
A. Koleżyński (&)
Faculty of Materials Science and Ceramics, AGH University of Science
and Technology, 30 Mickiewicza Av., 30-059 Krakow, Poland
e-mail: andrzej.kolezynski@agh.edu.pl
Fig. 1.1 Main regions of electromagnetic spectrum with related spectroscopic techniques
1 Computational Methods in Spectroscopy 3
to, e.g., Jahn–Teller effect, or the degree of covalence of bonds formed by the ion
with surrounding atoms or ligands.
With the increase of radiation frequency, the energy of photons increases and
hence the possibility appears of studying phenomena in which energy changes are
higher and correspond to radiation with frequencies of about 1011–1013 Hz, e.g.,
molecular rotations and atomic vibrations in molecules and solids (the domain of IR
and Raman spectroscopy). While rotational spectra can only be obtained for
molecules (those with a permanent dipole moment) and thus provide the infor-
mation on the structure of such molecules (lengths of bonds and angles between
bonds), atomic vibrations are present in both, molecules and solids (amorphous and
crystalline), hence the vibrational spectroscopy (IR and Raman) [6] is a universal
method of studying the local structure of materials (force constants, bonds lengths,
angles, etc.) by analyzing normal modes of vibrations and changes in absorption
(IR) or scattering (Raman) of infrared radiation associated with structural changes
resulting, for example, from phase transitions or changes in chemical composition
(defects, admixtures, etc.).
The next group of spectroscopic methods is UV-Vis [7, 8] techniques related to
electronic transitions between different energy levels and the corresponding radi-
ation range, from near infrared, through visible region, to ultraviolet (about 1014–
1015 Hz). Various types of electronic transitions are possible, both between the low
lying localized levels of the internal shells of individual atoms and the levels of the
outermost valence shells (either localized on atoms or delocalized, forming bands).
Generally, there are four basic types of electron transitions:
• Excitation of an electron from a localized orbital to another localized orbital on
the same atom but with higher energy: The examples are d-d or f-f transitions in
compounds containing transition metals, transitions on outer shells in heavy
metals compounds (e.g., 6s–6p in lead(II) compounds), or transitions related to
defects (trapped electrons or holes, color centers);
• Excitation of an electron from a localized orbital on one atom to a higher energy
orbital localized on an adjacent atom (charge transfer processes, e.g., from
oxygen to chromium in tetrahedral [CrO4]2− anion);
• Excitation of an electron from a localized orbital on an atom to a delocalized
conduction band (e.g., photoconductive materials);
• Electron transitions from valence band to conduction band (e.g., in
semiconductors).
UV-Vis spectroscopy has a very wide range of applications related to the study
of the local structure of materials, because the positions of absorption bands are
very sensitive to the immediate surroundings and the nature of chemical bonding.
For example, by adding a small amount of spectroscopically active elements, such
as transition metals or p-block heavy metals (e.g., Pb2+), a number of information
on the local structure of amorphous materials can be obtained. Materials for lasers
often contain transition metal ions as active species, e.g., Al2O3 doped with Cr3+
(ruby laser) or Nd3+-doped glass (neodymium laser), and UV-Vis spectroscopy
4 A. Koleżyński
allows obtaining detailed knowledge about energy levels and possible transitions—
a key information from the point of view of effective laser working. UV-Vis
spectroscopy is also widely used in analytical chemistry, e.g., to determine the
concentration of ions, molecules, or functional groups responsible for individual
absorption peaks.
Electron spectroscopy methods [9–12] (XPS—X-ray photoelectron spec-
troscopy, UPS—ultraviolet photoelectron spectroscopy, or AES—Auger electron
spectroscopy) measure the kinetic energy of electrons emitted from the material as a
result of exposure to high energy ionizing radiation or high energy particles
bombardment. The main difference between XPS and UPS methods is in the atomic
shells which can be studied: The inner shells electrons can be ejected using higher
energy X-rays (frequencies around 1018 Hz), while valence shells electrons can be
ionized using lower energy ultraviolet light with the frequencies about 1016 Hz. The
kinetic energy of electrons measured by these methods is equal to the difference
between the energy ht of the used incident radiation and the binding energy
(ionization potential) of the emitted electron and depends strongly on the type of
atom from which the electron was emitted, its oxidation state, and its immediate
environment; whereby it is possible to use these techniques to identify the atoms of
a given element in the material, providing the basis for electron spectroscopy for
chemical analysis (ESCA) developed by Siegbahn et al. [13].
The method directly related to the XPS and UPS spectroscopy is the Auger
electron spectroscopy [11, 14], in which emitted electrons are not the electrons
primary ionized by X-ray or ultraviolet radiation, but the product of secondary
processes, i.e., decay of ionized atoms from excited states to lower energy states
associated with emission of a photon, which can be emitted (and measured by XPS
or UPS) or absorbed by the electron from outer shells, resulting in the emission of
such secondary ionized electron (called the Auger electron). Electron spectroscopy
methods are very important and useful methods to determine energy levels in atoms
and molecules. In the case of solids, they are particularly useful for studying the
local structure of the surface of materials, since due to the relatively low energy of
emitted electrons and their strong dissipation in the material, in practice it is pos-
sible to observe only electrons ejected from the surface layer with a thickness of up
to 5 nm.
Even higher energies than those used in electron spectroscopy require processes
of transitions between energy levels of atomic nuclei (around 1019 Hz). The method
that examines such processes is the Mössbauer spectroscopy [15, 16], based on the
Mössbauer effect (i.e., resonant and recoil-free emission and absorption of c-rays by
atomic nuclei in a solid). In Mössbauer spectroscopy, three types of nuclear
interactions are observed:
• Isomer shift (energy shift of the absorption peak resulting from the modification
of the energy levels of the nucleus by changes in the distribution of electron
density in the atom, due to the changes in the local environment of the atom; this
shift may, e.g., correlate with the electron density at atomic nucleus of the outer
s shells electrons);
1 Computational Methods in Spectroscopy 5
• Quadrupole splitting (in the case of nuclei with nuclear spin I > ½, the distri-
bution of positive charge in the nucleus is not spherically symmetrical resulting
in quadrupole moment Q, which in turn results in the split of the nuclear energy
levels and hence the split of the corresponding absorption bands in the
Mössbauer spectrum);
• Magnetic hyperfine Zeeman splitting (when the atomic nucleus with spin I will
be subjected to the magnetic field, then each nuclear energy level splits into
2I + 1 levels).
The source of radiation in this method is a highly monochromatic gamma-ray
beam whose energy is changed using the Doppler effect: The sample is fixed and
the radiation source moves at a constant velocity toward or away from the sample,
and as a result, the energy of radiation incident on the sample increases or decreases
with respect to the primary energy of radiation. In this way, an absorption spectrum
of gamma radiation is created, the analysis of which allows obtaining information
about the local structure (oxidation state, coordination number, nature of binding,
etc.) and precisely measuring—due to a high energy and extremely narrow line
widths of c-rays—small energy changes in nuclei, atoms, and crystals induced by
electrical, magnetic, or gravitational field. The studies of hyperfine splitting,
especially as a function of temperature, allow additionally obtaining information on
the ordering of magnetic materials.
Theoretical studies on the above-mentioned spectroscopic methods are cur-
rently conducted in practice only with the help of methods based on the for-
malism of quantum mechanics. In practice, only in the case of vibrational
spectroscopy, it is possible to use classical methods (both stationary-molecular
mechanics and dynamical—molecular dynamics), in which atoms are treated as
charged particles, and their interactions are described by means of analytic two
and more center functions, the so-called interatomic potentials. It is then possible
to calculate the total potential energy of such a classical system, hence the forces
acting on individual atoms and their minimization (relaxation of the structure),
and then calculating (for the relaxed structure) the frequencies of normal modes
of vibrations and simulation of vibrational spectra. However, due to the lack of a
correct description of the electron density, such an approach does not allow to
analyze the processes dealt with by other spectroscopic methods mentioned
above, and therefore, in remaining parts of this chapter, classical methods will not
be further discussed.
level of complexity, starting with the simplest atoms, through molecules and
clusters, to amorphous and periodic solids. In the past, experimental studies were
the main and, in some cases, practically the only source of knowledge about the
properties of materials. The development of computer technology and theoretical
methods leads gradually to the increasing use of computer calculations and simu-
lations (the so-called in silico experiments) in materials research, which in many
cases becomes an equivalent (and sometimes also the first, preliminary) tool
allowing understanding of the properties of materials and their modification and
functionalization.
Among the theoretical methods used in material research, the so-called ab initio
methods (methods from the first principles), whose basic idea is to treat
multi-atomic systems as systems of many bodies composed of atomic nuclei and
electrons and to analyze them employing quantum mechanics—the arguably most
important contemporary physical theory, without using any empirical parameters.
In the following few sections, the most important ingredients of quantum
mechanics formalism as well as the density functional theory, rooted in this
formalism (and nowadays the dominant approximation used in theoretical cal-
culations), required for their practical application in modeling of the properties of
complex many-body systems will be presented. Apart from the above-mentioned
“static” methods, i.e., methods based on the solutions of appropriate equations
formulated for stationary states, there are also other, computationally much more
complex and demanding, dynamic (ab initio molecular dynamics allowing to
study the changes of system state with time) and statistical approaches (ab initio
quantum Monte Carlo, allowing going beyond the mean field theory and
independent electrons approximations and direct treatment and description of
complex interactions in many-body systems). However, due to the limited volume
of the chapter and still relatively rare use of these methods, they will not be
presented here and interested readers are referred to the rich literature devoted to
this topic [e.g., 17–21].
The physical theory describing the phenomena occurring in the micro-world, and
therefore also the processes studied by various spectroscopic methods, is quantum
mechanics. The most popular formulation of quantum theory is the formalism of
wave mechanics proposed by Erwin Schrödinger [22–27]. The starting point in
Schrödinger’s formulation is a number of postulates regarding the wave function of
complex values (representing the state of the system), linear Hermitian operators
and their eigenvalues (mathematical representations of observable quantities, the
so-called observables), a recipe enabling finding an operator associated with a given
observable, a description of the measurement process and a way of predicting its
1 Computational Methods in Spectroscopy 7
@Wðr; tÞ ^
ih ¼ HWðr; tÞ ð1:1Þ
@t
where W—wave function describing the state of the system and h—Planck’s
^
constant h, divided by 2p. The Hamiltonian H of the system (the total energy
operator) used in the Schrödinger equation is defined as:
2 2
^ ¼ h @ þ @ þ @ þ Vðr; tÞ ¼
h2 2
2 2
^ ¼ T^ þ V
H r þ Vðr; tÞ ð1:2Þ
2m @x2 @y2 @z2 2m
@Wðr; tÞ h2 2
ih ¼ r Wðr; tÞ þ V ðr; tÞWðr; tÞ ð1:3Þ
@t 2m
In the vast majority of cases, when we model a given system we are interested in
its properties at stationary state (ground or excited), which is why we are usually
interested in solving the so-called time-independent Schrödinger equation (TISE). If
we assume that in the stationary state the external potential is constant, independent
of time V ðr; tÞ ¼ VðrÞ, then we can perform a mathematical operation of separating
the variables and write the function of the state of the system as:
h2 1 1 dTðtÞ
r2 wðrÞ þ VðrÞ ¼ i
h ð1:6Þ
2m wðrÞ TðtÞ dt
true for all r and t, so both sides must be equal to constant E, called the separation
constant. In this way, we obtain two independent equations describing the depen-
dence of the wave function, first on the position:
h2 1 ^
r2 wðrÞ þ VðrÞ ¼ E ) HwðrÞ ¼ EwðrÞ ð1:7Þ
2m wðrÞ
8 A. Koleżyński
1 dT ðtÞ
ih ¼E ð1:8Þ
T ðtÞ dt
The solution of the latter has the form T ðtÞ ¼ AeiEt=h , where E, is energy of
stationary state, one of the eigenvalues of the Hamiltonian operator defined for the
system in time-independent (stationary) state. Finally, the solution to the
Schrödinger equation has the form:
One can easily see that although the potential of V(r) is independent of time, the
wave function oscillates in time with the frequency depending on the energy cor-
responding to the current stationary state of the system (however, at the same time,
the probability distribution Pðx; tÞ ¼ jWðx; tÞj2 ¼ jwðxÞj2 , and therefore also the
distribution of total electron density, is constant in time).
Since the solution of the time-dependent part of Schrödinger equation is identical
in each case, in order to obtain full information about the system’s state and its
evolution over time, it is sufficient to find a solution for time-independent
Schrödinger equation, of the general form:
^
Hwðr 1 ; r2 ; . . .; rn ; Ra ; Rb ; . . .; Rm Þ ¼ Ewðr1 ; r2 ; . . .; ri ; Ra ; Rb ; . . .; Rm Þ ð1:10Þ
Except for a few simplest cases, finding an exact, analytical solution of this
equation is impossible, and thus, some simplifications are usually necessary to be
used and solutions are obtained in numerical form.
that since nuclei are much heavier then electrons (approximately 103–105 times,
depending on the number of protons and neutrons making up particular nucleus), they
move much slower than electrons which have a response time of the order of fem-
toseconds and can adjust themselves almost instantaneously to the new atomic con-
figuration (thus—from electrons’ point of view—nuclei seem to be clamped and the
idea goes, we can treat the motions of electrons and nuclei as independent). In this
case, the effective interactions between the atoms take place via electrons, and we
carry out the calculations looking for a solution for the ground state of the electron
cloud, for the assumed, fixed spatial configuration of motionless atoms. The
assumption of the independent movement of the atomic nuclei and electrons allows
separation of the variables describing their positions and writing the multi-electron
wave function as a productof two functions: one dependent only on nuclei positions
(nuclei wave function wN Ra ; Rb ; . . .; Rm ) and the other on positions of electrons
(electronic wave function wel r1 ; r2 ; . . .; rn ; Ra ; Rb ; . . .; Rm ) and current geometry
(nuclei positions):
w r1 ; r2 ; :::; rn ; Ra ; Rb ; :::; Rm ¼ wel r1 ; r2 ; :::; rn ; Ra ; Rb ; :::; Rm wN Ra ; Rb ; :::; Rm
ð1:12Þ
This enables us to solve the electron and nuclear Schrödinger equation sepa-
rately, starting with the one for electrons:
^ el wel ðr; RÞ ¼ Eel wel ðr; RÞ
H ð1:13Þ
where
2 X X X Za e2 X X e2
^ el ¼ h
H r2i þ ð1:14Þ
2me i a i
ria r
j i [ j ij
Once we have the potential energy surface (PES), we can solve the nuclear
Schrödinger equation:
^ N wN ðRÞ ¼ E N wN ðRÞ
H ð1:16Þ
where
X 1
^ N ¼ h
2
H r2 þ U ðRÞ ð1:17Þ
2 a
ma a
10 A. Koleżyński
The solution of this equation allows obtaining information on the quality of the
initially adopted geometry of the studied system (spatial configuration of atoms).
Usually, the assumed geometry deviates from the equilibrium geometry and the
atoms are subjected to the forces that can be easily calculated (as a potential energy
gradient with respect to the atomic position). Information about forces acting on
atoms allows the geometry modification (optimization) by changing the positions of
all atoms according to the direction and magnitude of the respective forces and
repeating the total energy calculations, another forces estimation, geometry modi-
fication according to new forces calculated, and repeating the whole procedure until
the calculated forces on the atoms are smaller than the assumed precision. In
most cases, the relaxation of the structure (minimization of forces on atoms) is a
necessary step in the calculation, because the majority of physical quantities that are
of interest is defined for systems that are in the minimum of potential energy well
(relaxed, equilibrium geometry).
Since such simple product does not fulfill the requirement for electronic wave
function of the system to be antisymmetric (due to Pauli’s Exclusion Principle,
which electrons—as indistinguishable fermions—have to obey) while the linear
combination of different products does, the original Hartree product was replaced
later by antisymmetric-by-definition Slater determinant (with (spin)orbitals as its
elements) of the form:
1 Computational Methods in Spectroscopy 11
/1 ðr1 Þ /1 ðr2 Þ /1 ðrN Þ
1 /2 ðr1 Þ /2 ðr2 Þ /2 ðrN Þ
wel ðr1 ; . . .ri ; . . .; rj ; . . .; rN Þ ¼ pffiffiffiffiffi . .. .. ð1:19Þ
N! .. . .
/ ðr1 Þ /N ðr2 Þ /N ðrN Þ
N
In the general case, the wave function for the N-electron system can be written in
the form of a linear combination of Slater determinants (including in the determi-
nants also partially occupied spinorbitals for excited states, empty in the ground
state):
X
wel ¼ di Ui ¼ d0 U0 þ d1 U1 þ d2 U2 þ ð1:20Þ
i
P
n P
m
^hi where ^hi ¼ T^e þ V
^Ne ¼ h2 r2i Za e 2
ð1:21Þ
2me ria
i¼1 a¼1
X X e2
^ ee ¼
V ð1:22Þ
j
r
i [ j ij
X
n
^eff ðiÞ ¼
V 2^Jj ðiÞ K
^ j ðiÞ ð1:26Þ
j
where
Z
^Jj ð1Þ ¼ 1
uj ð2Þ u ð2Þdr2 ð1:27Þ
r12 j
Z
^ j ð1Þui ð1Þ ¼ 1
K uj ð2Þ u ð2Þdr2 uj ð1Þ ð1:28Þ
r12 i
X
^fi ¼ h r2
2 m
Za e2 ^
þ V eff ðiÞ ð1:29Þ
2me i a¼1
ria
Since the original HF method does not (by definition) take into account electron
correlation effects, in order to get the results in better agreement with experiment, a
few extensions, the so-called post-HF methods, have been proposed in following
years, e.g., CI—Configuration interaction [30–32], MP2 [33, 34], MP3 [35, 36],
and MP4 [37] Møller–Plesset Perturbation Theory, CC—Coupled Cluster Theory
[38, 39], QCI—Quadratic Configuration Interaction [40], CASSCF—Complete
Active Space Self-Consistent Field calculation [41, 42], MRSDCI—
Multi-Reference Single and Double Configuration Interaction [43] or Quantum
Chemistry Composite methods, like Gaussian G1 [44, 45], G2 [46], G3 [47], and
G4 [48], FPD Feller–Peterson–Dixon approach [49, 50], ccCA—correlation con-
sistent Composite Approach [51], or CBS—Complete Basis Set methods [52]. Such
extensions improved considerably the quality of calculations, but at the expense of
significantly higher demand for computer resources and calculation time (in the
worst-case scenario, calculation time increases fivefold with the size of the system,
comparing to original HF method).
X Za e2
Vext ðrÞ ¼ ð1:32Þ
a
j r ra j
1 Computational Methods in Spectroscopy 15
And
X
^ext ¼
V Vext ðri Þ ð1:34Þ
i
Functional F^ is the same for each N-electron system; therefore, the Hamiltonian
and the ground state W are completely defined by N and Vext(r), and the W
eigenstate corresponds to the electron density n(r), expressed as:
Z
nðrÞ ¼ N jWðr; r2 ; . . .; rN j2 dr2 . . .drN ð1:35Þ
and reaches the minimum for the total N-electron density of the ground state in the
external potential Vext(r). The ground state energy of such a system can be calcu-
lated using the variational method, since for each density n′(r) that is not eigenstate
of the external potential Vext(r), Ev[n′(r)] > Ev[n(r)], where the electron density that
minimizes the total energy of the system is the exact density of the ground state
(however, this is true only for the exact functional Ev[n(r)]).
An extraordinary result of DFT is the demonstration of the existence of a uni-
versal functional F[n(r)], independent of external potential, which means that instead
of dealing with the N-electron wave function, we operate on the electron density, i.e.,
three-dimensional spatial coordinates function. The level of complexity of the
problem is thus dramatically reduced and scales linearly with the size of the system
(compare the HF method, where the problem scales with second– third power).
Unfortunately, the exact form of the functional F[n(r)] is unknown and the practical
16 A. Koleżyński
problem faced by DFT (finding an approximate function) results directly from the
complexity of the problem of interacting electrons in many-body system.
Kohn and Sham [59] have reformulated the above problem by replacing the
system of N interacting electrons with a special reference system—the system of
N non-interacting electrons, having the same ground electron density. In the case of
such a reference system, this electron density is equal to the sum over the
one-electron wi(r) orbitals (Kohn–Sham orbitals):
X
N=2
nðrÞ ¼ 2 jwi ðrÞj2 ð1:37Þ
i¼1
1
r2 þ VKS ðrÞ wi ðrÞ ¼ ei wi ðrÞ ð1:38Þ
2
The existence of a uniquely determined VKS(r) potential for the ground state
electron density n(r) is a direct consequence of the first Hohenberg–Kohn theorem.
Therefore, we have a situation in which the problem of finding a universal func-
tional F[n(r)] was replaced by the problem of finding an effective VKS[n(r)] func-
tional. This problem can be solved using the variational principle. Lets write down
the energy in the form:
Z
E ¼ TS ½nðrÞ þ EH ½nðrÞ þ EXC ½nðrÞ þ nðrÞVext ðrÞdr ð1:40Þ
The first part is the kinetic energy of the system of non-interacting electrons:
Z
h2 X
N=2
TS ½nðrÞ ¼ 2 wi ðrÞr2 wi ðrÞdr ð1:41Þ
2m i¼1
and the last term is the classical Coulomb energy of electron interactions with
external potential. The key element in the expression for the energy of the system is
the third term whose exact form is unknown—this is the exchange and correlation
1 Computational Methods in Spectroscopy 17
dF½nðrÞ
l¼ þ Vext ðrÞ ð1:44Þ
dnðrÞ
dTS ½nðrÞ
l¼ þ VKS ðrÞ ð1:45Þ
dnðrÞ
where
and
Z
dEH ½nðrÞ nðrÞ dEXC ½nðrÞ
VH ðrÞ ¼ ¼ dr 0 ; VXC ðrÞ ¼
dnðrÞ jr r 0 j dnðrÞ ð1:47Þ
Both Euler equations are equivalent, which means that the density obtained as a
solution to the variational problem for the KS reference system is identical to the
density obtained for the original system of N interacting electrons and can be in
practice obtained by solving a system of N one-electron Schrödinger equations
(Kohn–Sham equations):
1
r2 þ VKS ðrÞ wi ðrÞ ¼ ei wi ðrÞ ð1:48Þ
2
Another random document with
no related content on Scribd:
also, yet in its absence it is not likely that it kept either valuables
or money at Westminster. But side by side with the state office
was the household office of finance, the Wardrobe, and, though
the wardrobe office was itinerating with the king, it still kept a
“treasury” or storehouse at Westminster, and this, for the sake of
greater safety, had been placed for some years at least within
the precincts of the abbey. From the monastic point of view, it
was doubtless an inconvenience that nearness to the royal
dwelling compelled them to offer their premises for the royal
service. Accordingly, kings not infrequently made demands upon
the abbey to use its buildings. Thus the chapter house became a
frequent place for meetings of parliament, and at a later time it
was used and continued to be used till the nineteenth century, for
the storage of official records. In the same way Edward secured
the crypt underneath the chapter house as one of the
storehouses of his Wardrobe. When the crypt was first used for
this purpose I do not know, but records show us that it was
already in use in 1291, at which date it was newly paved. It was
not the only storehouse of the Wardrobe. There was another
“treasury of the wardrobe” in the Tower of London, but this was
mainly used for bulky articles, arms and armour, cloth, furs,
furniture, and the like. Most of what we should call treasure was
deposited in the Westminster crypt, and we are fortunate in
having still extant a list of the jewels preserved there in 1298, the
time when the court began to establish itself for its five years’
sojourn in the north. In 1303 jewels and plate were still the chief
treasures preserved there. Some money was there also, notably
a store of “gold florins of Florence,” the only gold coins currently
used in England at a time when the national mints limited
themselves to the coinage of silver. But I do not think there could
have been much money, for Edward’s needs were too pressing,
his financial policy too much from hand to mouth, for the crypt at
Westminster to be a hoard of coined money, like the famous
Prussian Kriegsschatz at Spandau, which, we now rejoice to
learn, is becoming rapidly depleted. Whatever its contents,
Edward estimated that their value was £100,000, a sum
equivalent to a year’s revenue of the English state in ordinary
times. Unluckily mediaeval statistics are largely mere guess-work.
But the amount of the guess at least suggests the feeling that
the value of the treasures stored in the crypt was very
considerable.
The crypt under the chapter house is one of the most
interesting portions of the abbey buildings at Westminster. It is
little known because it is not, I think, generally shown to visitors.
I am indebted to the kindness of my friend, Bishop Ryle, the
present dean, for an opportunity of making a special inspection of
it. It is delightfully complete, and delightfully unrestored. The chief
new thing about it seems the pavement, but the dean’s well-
informed verger told me that it was within living memory that this
pavement had replaced the flooring of 1291. Numerous windows
give a fair amount of light to the apartment; though the enormous
thickness of the walls, some thirteen feet, it was said, prevent the
light being very abundant, even on a bright day. The central
column, the lower part of the great pillar from which radiates the
high soaring vaults of the chapter house above, alone breaks the
present emptiness of the crypt. Considerable portions of the
column are cut away to form a series of neatly made recesses,
and there are recesses within these recesses, which suggest in
themselves careful devices for secreting valuables, for it would be
easy to conceal them by the simple expedient of inserting a
stone here and there where the masonry had been cut away, and
so suggesting to the unwary an unbroken column. I should not
like to say that these curious store-places already existed in
1303; but there is no reason why they should not. Certainly they
fit in admirably with the use of the crypt as a treasury.
One other point we must also remember about the
dispositions of this crypt. There is only one access to it, and that
is neither from the chapter house above nor from the adjacent
cloister, but from the church itself. A low, vaulted passage is
entered by a door at the south-east corner of the south transept
of the abbey, now for many centuries the special burial place for
poets, eminent and otherwise. This passage descends by a flight
of steep steps to the crypt itself, and the flight originally seems, I
am told—doubtless as another precaution against robbery—to
have been a broken one suggesting that a steep drop,
presumably spanned by a short ladder, further barred access to
the crypt. We must remember, too, that this sole access to the
treasury was within a few feet of the sacristy of the abbey. The
sacristy was the chapel to the south of the south transept, and
communicating with it where the sacrist kept the precious vessels
appropriated to the service of the altar. Altogether it looks as if
the crypt were originally intended as a storehouse for such
church treasure as the sacrist did not need for his immediate
purposes. From this use it was diverted, as we have seen, to the
keeping of the royal treasures. Nowadays the sacristy is called
the chapel of St. Faith and is used for purposes of private
devotion. We must not forget the close connexion in our period of
the sacristy and the crypt. The connexion becomes significant
when we remember that among Pudlicott’s monastic boon
companions at the palace-keeper’s lodge was the sacrist of the
abbey, Adam of Warfield.
Pudlicott had made up his mind to steal the king’s treasure.
The practical problem was how to get access to it. If we examine
the evidence collected at the enquiry, we find that there are two
discrepant accounts as to how the robber effected his purpose.
The one is warranted by the testimony of a large number of
sworn juries of reputable citizens of every ward in the city of
London, of burgesses of Westminster, and of the good men of
every hundred in the adjacent shires of Middlesex and Surrey. It
is—like much truthful evidence—rather vague, but its general
tendency is, while recognizing that Pudlicott is the prime offender,
to make various monks and palace officers his accomplices. Of
the latter category William of the Palace seems to have been the
most active, while of the many monks Adam Warfield the sacrist
was the most generally denounced. But the proved share of both
Adam and William was based largely on the discovery of stolen
property in their possession. The evidence of the juries suggests
theories as to how the crime may have been perpetrated; it does
not make the methods of the culprits clear and palpable. But it
suggests that masons and carpenters were called in, so that
some breaking in of the structure was attempted, and in particular
it suggests that the churchyard was the thoroughfare through
which the robbers removed their booty.
I. Note on Authorities.