Structure sensitive investigations on Mn-activated perovskites and rare-earth-doped aluminates / von Bastian Henke [original]

Faculty of Natural Sciences

Department of Physics

Structure Sensitive Investigations on Mn-activated

Perovskites and Rare-Earth-doped Aluminates

Dissertation

von Dipl. Phys.

Bastian Henke

Gutachter: PD Dr. Stefan Schweizer

Zweitgutachter: Prof. Dr. Gerhard Wortmann

Abgabe der Dissertation: 19. Oktober 2007

Datum der m¨undlichen Pr¨ufung: 11. Dezember 2007

Abstract

In this work optical, magnetic resonance, and optically-detected magnetic resonance

measurement techniques were used to investigate the Mn-activated fluoroperovskites

RbCdF3and LiBaF3as well as the rare-earth (RE) doped persistent phosphors SrAl2O4

and CaAl2O4.

The photoluminescence (PL) of Mn-activated RbCdF3shows an emission band at

560 nm, which can be attributed to the Mn2+ dopant; the corresponding excitation bands

are between 240 and 520 nm. After x-irradiation an increased Mn2+ emission can be

observed and the excitation spectrum shows an additional intense broad band at 300 nm.

Excitation at 300 nm leads to a decreasing intensity of the Mn2+ emission, whereas

excitation at 240 nm leads to an increase in the PL intensity. Electron paramagnetic

resonance (EPR) shows that the number of Mn2+ ions is reduced upon x-irradiation;

the original Mn2+ level can be restored by optical bleaching.

In LiBaF3it was possible to identify a luminescent oxygen vacancy complex emitting

in the blue spectral region. The structure of that complex was investigated by PL and

PL-detected electron paramagnetic resonance (PL-EPR). At 20 K the oxygen-related

complex shows two luminescence bands peaking at about 430 and 475 nm, when excited

at 220 nm. These bands can be attributed to an excited triplet state (S= 1) with

the z-axis of the fine structure tensor parallel to the <110>direction. This complex is

believed to be next to a Mn2+ impurity on a Ba2+ site and can be described by an oxygen

on a fluorine lattice site with a nearest fluorine vacancy along the <110>direction.

Prior to x-irradiation the PL of Mn-activated LiBaF3shows a Mn2+ emission band

at 710 nm; the corresponding excitation bands are between 210 and 620 nm. After

x-irradiation the PL spectrum shows a new emission peaking at about 610 nm which

is tentatively assigned to a perturbed Mn2+ emission; the intensity of all PL bands

is increased. Structure sensitive investigations on the radiation-induced emission band

iii

were done by PL-EPR at a temperature of 1.5 K. The analysis of the angular dependent

PL-EPR spectra, recorded for different orientations of the magnetic field, yielded that

the 610 nm luminescence band is due to an excited triplet state (S= 1) of a Mn-related

center with the z-axis of the fine structure tensor close to a <110>direction.

Single-crystals of MAl2O4(M=Ca and Sr) persistent phosphors that are nominally

pure or doped with Eu and Nd or Dy, respectively, were investigated. Their recombina-

tion luminescence (RL) and microwave-induced changes in the RL in a high magnetic

field (RL-EPR) were investigated after ultraviolet excitation at low temperatures. Wave-

length dependent RL-EPR measurements indicate that only one intrinsic donor but at

least two different intrinsic acceptors are involved in the recombination process. The

donor-acceptor recombination energy is either emitted directly (undoped samples) or

almost completely transferred to the RE activators (doped samples).

Contents

1 Introduction 1

2 Physical Background 3

2.1 Perovskites ................................... 3

2.1.1 Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1.2 Photoluminescence of Mn2+ ...................... 4

2.2 Experimental .................................. 7

2.3 Magneto-Optical Measurement Techniques . . . . . . . . . . . . . . . . . . 8

2.3.1 Photoluminescence detected Electron Paramagnetic Resonance . . 8

2.3.2 Recombination Luminesence detected Electron Paramagnetic Res-

onance of Weakly Coupled Donor Acceptor Recombination . . . . 9

2.3.3 Experimental Setup . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.4 X-ray Absorption Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . 12

2.4.1 Extended X-ray Absorption Fine Structure Oscillations . . . . . . 12

2.4.2 Experimental Details . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3 The Perovskite RbCdF3:Mn2+ 17

3.1 Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.2 Photoluminescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.3 Magnetic Resonance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.3.1 Electron Paramagnetic Resonance . . . . . . . . . . . . . . . . . . 21

3.3.2 Optically detected Magnetic Resonance . . . . . . . . . . . . . . . 22

3.3.3 Electron Paramagnetic Resonance after x-irradiation . . . . . . . . 23

3.4 Beamline Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.4.1 X-ray Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.4.2 Scintillation Intensity . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.4.3 EXAFS on Mn-doped RbCdF3.................... 27

Contents

3.5 Discussion.................................... 31

4 EXAFS on Eu-doped CaF233

4.1 Advanced Photon Source . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

4.2 HASYLAB ................................... 33

4.3 Discussion.................................... 38

5 The Inverse Perovskite LiBaF341

5.1 Crystal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.2 X-ray Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

5.3 Oxygen Vacancy Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.3.1 Photoluminescence . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

5.3.2 Photoluminescence detected Electron Paramagnetic Resonance . . 44

5.3.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.4 Mn-related Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.4.1 Photoluminescence . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

5.4.2 Photoluminescence detected Electron Paramagnetic Resonance . . 53

5.5 Discussion.................................... 56

6 The Persistent Phosphors MAl2O4(M = Ca, Sr) 59

6.1 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

6.2 Photoluminescence and Recombination Luminescence . . . . . . . . . . . 61

6.3 Electron Paramagnetic Resonance . . . . . . . . . . . . . . . . . . . . . . 66

6.4 Recombination Luminescence detected Electron Paramagnetic Resonance 70

6.5 Discussion.................................... 70

7 Conclusion 75

A BaCl2EXAFS 77

B Miscellaneous 81

B.1 EulerTensor .................................. 81

B.2 Fine Structure Tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

B.3 EPRData.................................... 83

Bibliography 88

1 Introduction

In recent years there has been interest in dosimetric materials where the dose can be

read out via optically stimulated luminescence (OSL) and not as usually via thermally

stimulated luminescence. The most important advantage of OSL active materials is the

possibility to be bleached optically and to be reused.

The use of OSL for personnel dosimetry, is not widely spread and in radiation dosime-

try it has not been extensively reported, mainly because of the lack of a good luminescent

material.

MgS, CaS, SrS, and SrSe doped with different rare-earth (RE) elements such as Ce, Sm,

and Eu were among the first phosphors suggested for OSL dosimetry applications. They

possess a high sensitivity to radiation and a high efficiency under infrared stimulation

at a wavelength around 1 µm, but they show a significant fading of the luminescence at

room temperature (RT). These phosphors also have a very high effective atomic number

and, as a result, they exhibit strong photon energy dependence, which is unacceptable

for use in personnel dosimetry.

Alternative materials can be found in AMeF3perovskites, where A is an alkali metal,

Me is a transition metal, and F is a fluorine ion. These compounds doped with an

activator ion, e.g. Mn2+, are known to emit an enhanced luminescence after irradiation.

This was first observed by Sibley and Koumvakalis in RbMgF3:Mn2+ [1]. It was supposed

that an F-centre (electron trapped at an anion vacancy), in the vicinity of the Mn dopant,

enhances the oscillator strength leading to increased emission intensity. A similar effect

was observed in RbCdF3:Mn2+, where after x-irradiation an increased luminescence and

an additional absorption band around 310 nm occurs; this band can also be observed

in nominally pure RbCdF3[2]. The 310 nm band was tentatively assigned to intrinsic

defects being created upon x-irradiation. The radiation-induced luminescence intensity is

correlated with the x-ray dose; Mn-doped RbCdF3has potential as a radiation dosimeter

material.

Chapter 1. Introduction

Persistent phosphors are compounds which emit a very long lasting afterglow after

excitation mainly in the UV spectral range. Long persistent phosphors have a great

potential in applications such as emergency illumination in planes and trains or as lu-

minous dials. The first materials were ZnS:Cu and alkali sulfides like CaS and SrS. In

1996 rare-earth doped aluminates MAl2O4(M = Ca, Sr) have been found as a new type

of persistent phosphor [3]. Their emission can be attributed to the RE activator Eu2+

and is due to a 5f−4dtransition. Although the single doped systems show already a

phosphorescent behavior, the addition of Dy in the case of SrAl2O4(SAO) and of Nd

in the case of CaAl2O4(CAO) leads to a significant increase in the phosphorescent life-

time. Upon this additional codoping the phosphorescent lifetime is extended to about 16

hours. These two phosphors drew considerable attentions from many scientists because

a 16 hours persistent time could bring the phosphorescence over a whole night.

The complete mechanism of this phosphorescent luminescence is still unknown al-

though several models have been proposed [4, 5]. On the one hand it has been assumed

that only RE traps are involved in the recombination process. On the other hand it was

supposed that intrinsic traps are the source of the traps.

Theoretical and experimental fundamentals of the main experimental methods used

in this work are described in chapter two.

Chapter three deals in the first part with Mn-activated RbCdF3, especially with the

investigation of x-irradiation induced effects. The second part is about two different

luminescent centers in the Mn-activated inverse fluoroperovskite LiBaF3investigated by

photoluminescence detected electron paramagnetic resonance.

The fourth chapter is about the persistent phosphors CAO and SAO. Low temperature

recombination processes leading to phosphorescence in RE doped CAO and SAO single

crystals is studied by detecting the microwave-induced changes in the recombination

luminescence.

Parts of the present work have already been published, a publication list is added at

the end of this thesis.

2 Physical Background

2.1 Perovskites

2.1.1 Crystal Structure

The name perovskite represents a group of crystals having the same structure. The

perovskites presented in this work have an AMeF3structure, where A stands for a

cation, Me for a metal and F for fluorine. The crystal structure of AMeF3perovskites

depends on the tolerance factor t, which gives a relation between the radii of the cation,

the metal, and the fluorine. The tolerance factor is defined by

t=rA+rF

√2·(rMe +rF).(2.1)

A perovskite structure is expected if thas a value within the range 0.76 ≤t≤1.13.

For a tolerance factor of 0.76 ≤t≤0.88 the orthorhombically distorted GdFeO3-type

structure is found, which corresponds to the space group P bnm.

The ideal cubic structure of AMeF3belongs to a tolerance factor range of 0.88 ≤

t≤1.00, space group Pm3m(see Figure 2.1). The A-ions have 12 fluorine neighbors

while the metal ions have six. Thus, the metal ions with six neighbors form a MeF6

octahedron, which is linked to a neighboring MeF6octahedron by sharing the corners

throughout the structure.

In orthorhombic perovskites the Me-F-Me bridge angles (Figure 2.1 left side, horizontal

solid line) deviate significantly from linearity which is the main difference between cubic

and orthorhombic perovskites. This means a rotation of the MeF6octahedron about one

of its fourfold axes. As the tolerance factor is a measure of the misfit between the size

of the MeF6octahedra and that of the A cations, the Me-F-Me bridge angles depend on

t. A smaller tolerance factor means a smaller bridge angle.

Chapter 2. Physical Background

Figure 2.1: Cubic perovskite. The left picture shows how the six fluorine neighbors of the Me

ion form an octahedron.

AMeF3fluorides with a tolerance factor t > 1 adopt a variety of hexagonal structures.

In all these hexagonal structures a face-sharing of the octahedra occurs, which is a kind

of linking.

2.1.2 Photoluminescence of Mn2+

In the following section the electronic structure of Mn2+ will be explained for a better

understanding of its optical properties. Mn2+ has an incompletely filled dshell and

its electron configuration is d5. The energy levels originating from this configuration

have been calculated by Tanabe and Sugano considering the interaction between the d

electrons and the crystal field [6].

The energy levels of a free ion can be found on the left hand side of these Tanabe-

Sugano diagrams, see Figure 2.2. Many of these levels split into two or more levels for

a crystal field 10Dq 6= 0. The ground state is represented by the xaxis. The free ion

levels are marked 2S+1L, where Spresents the total spin quantum number, and Lthe

total orbital angular momentum. The degeneracy of these levels is 2L+ 1 and can be

lifted by the crystal field. Crystal field energy levels are marked 2S+1X, where Xmay

be Afor no degeneracy, Bfor a twofold degeneracy and Tfor a threefold degeneracy.

Chapter 2. Physical Background

Figure 2.2: The energy levels of Mn2+ as a function of the octahedral crystal field 10Dq. The

xaxis represents the ground state.

Subscripts indicate certain symmetry properties and the letter in brackets represents the

energy level of the free ion. For further information the reader is referred to [6, 7].

The Tanabe-Sugano diagram for Mn2+ in an octahedral Ohsurrounding is shown in

Figure 2.2. The ground state is 6A1g(S). Since these energy levels are all within the d

shell, all transitions from the ground state to excited levels are spin and parity forbidden.

However, the Mn2+ transitions can still be observed. It was suggested that the selection

rule is relaxed through a spin-spin interaction and a vibronic mechanism, where the

electronic transitions are coupled with vibrations of suitable symmetry [7].

Exciting a system with UV light brings an electron of Mn2+ in a high vibrational level

of the excited state, as shown in Figure 2.3. The electron falls to the lowest vibrational

level of the excited state giving up the excess energy to the surrounding. Another way

to describe this process is to say that the nucleus adjusts its position to the excited

state, so the interatomic distance varies from the distance it has in equilibrium. The

configurational coordinate changes by ∆R. This process is called relaxation. From the

lowest vibrational level the electron can return to the ground state spontaneously under

Chapter 2. Physical Background

Figure 2.3: Configurational coordinate diagram. After absorption the system reaches a high

vibrational level of the excited state. Subsequently it relaxes to the lowest vibrational level from

where the emission occurs.

emission of radiation. From this it can be seen, that the emission has a lower energy

than the excitation of the same excited level. This effect is called Stokes Shift.

Typical photostimulated spectrum and excitation bands from Mn2+ are given in Fig-

ure 2.4. The PL arises from the radiative return to the ground state via the 4T1g(G)→

6A1g(S) transition. The excitation bands are marked with their corresponding energy

level. An interesting feature in this excitation spectrum is the different width of the

excitation maxima. In particular the degenerate states 4A1g(G) and 4Eg(G) have a nar-

row band and the 4T2g(G) and 4T1g(G) states are rather broad. The band-width is due

to coupling with vibrations. Since the crystal field strength varies during the vibration

the Tanabe-Sugano diagrams predict the width of the band. If the level runs parallel to

the ground level a variation of ∆ will not influence the transition energy and a narrow

band can be expected like the ones observed for 4A1g(G) and 4Eg(G). From this reason

these energy levels show for most systems their excitation band at about 400 nm. If

the excited levels has a slope relative to the ground state a variation of ∆ will influence

the transition energy, and a broad transition band can be expected. Because of this,

the 4T1g(G) is most sensitive to changes in ∆ and has significant different values for all

systems.

Chapter 2. Physical Background

Figure 2.4: The dashed curve shows the PL spectrum of Mn2+ in LiBaF3, excited at 255 nm.

The solid curve presents the excitation bands for this transition and the PL was detected at

712 nm.

Another property in these systems is the distance between the Me and the F ions

RMF. The 10Dq parameter strongly depends on RMF by the inverse power law found by

Rodriguez [8]. They showed that the dependence of 10Dq with RMF can be described

in the form

10Dq =K·R−n

MF (2.2)

where Kand nare constants. However, this law is only valid in a small range of RMF

(2.0-2.2 ˚

A).

2.2 Experimental

The luminescence spectrometer is shown in Figure 2.5. Photoluminescence (PL) and x-

ray excited luminescence (XL) spectra were recorded with a single beam spectrometer in

which two 0.22 m double monochromators were available for excitation and luminescence.

Chapter 2. Physical Background

Figure 2.5: Photograph of the luminescence spectrometer: a) Xenon lamp, b) excitation

monochromator, c) cryostat, d) emission monochromator, e) Photomultiplier.

The PL excitation was carried out with a xenon lamp, the x-ray excitation with a

tungsten tube (50-60 kV, 15-30 mA). The luminescence was detected using single photon

counting with a cooled photomultiplier. For the lifetime measurements a nitrogen flash

lamp or an iris in combination with a computer-controlled transient recording board

were used. For low temperature measurements a continuous flow cryostat was used.

The spectra were not corrected for spectral sensitivity of the experimental setup.

2.3 Magneto-Optical Measurement Techniques

2.3.1 Photoluminescence detected Electron Paramagnetic Resonance

With the PL-EPR technique it is possible to detect triplet spin systems. Triplet states

are often found as strongly coupled electrons and holes after band to band excitation,

which may be excitons bound to an impurity. Each of these two particles have a spin

of S= 1/2, are strongly coupled and form a triplet spin S= 1. Figure 2.6 shows the

energy levels of a triplet with fine structure interaction for an external magnetic field ~

Chapter 2. Physical Background

Figure 2.6: Level scheme of a triplet spin state system in an external magnetic field. The

rippled arrows indicate allowed transitions to the ground state.

parallel to the fine-structure axes.

With no applied magnetic field the mS=±1 and mS= 0 levels are split by the

fine-structure constant D. Applying a higher magnetic field leads to a further splitting

of the mS=±1 levels. There are allowed transitions to the singlet ground state for the

|+ 1 >and |−1>levels (indicated by waved arrows) but not for the |0>level. Thus

the |0>level will have a higher population than the |+ 1 >and |−1>energy states.

By applying microwaves it is possible to induce fine-structure EPR transition from |0>

to |+ 1 >and | −1>, indicated by arrows in Figure 2.6. Both transitions lead to an

enhanced PL intensity which can be detected.

2.3.2 Recombination Luminesence detected Electron Paramagnetic

Resonance of Weakly Coupled Donor Acceptor Recombination

Another technique of ODMR is the detection of the recombination luminescence of

weakly coupled donor acceptor pairs. Donors D0can be described by S= 1/2. For

a recombination luminescence donors and acceptors must be at least coupled weakly.

In this case the energy level scheme of Figure 2.7 applies; the spins of the donor and

acceptor can be parallel or antiparallel, which lead to four energy levels. The transition

probabilities for the two triplet spin states into the singlet ground state is rather small

compared to the two singlet-singlet transitions. Thus, there will be more electrons in

the two triplet states. By inducing EPR transitions electrons can be shifted from these

levels into the singlet states and thus the radiative transitions out of these levels are

enhanced.

Chapter 2. Physical Background

Figure 2.7: Level scheme for a weakly coupled donor acceptor pair.

The RL resulting from donor acceptor recombination created by UV light quenches

in high magnetic fields what is due to polarization of the magnetic moments of the

recombination centers. The probability Wfor a radiative recombination is given by

W=W0(1 −P1P2) (2.3)

where W0is the probability of radiative recombination without external magnetic

field; P1and P2are the polarizations of the spins of the donors and acceptors dependent

on the magnetic field given by

Pi=1

2Si

n=0

ne−gnµBB/kT −Si

2Si

n=0

e−gnµBB/kT

2Si

n=0

e−gnµBB/kT

(2.4)

where Siis the spin of the corresponding recombination center; µBis the Bohr mag-

neton; Bis the external magnetic field. The recombination probability calculated for

aS1=S2= 1/2 system is shown in Figure 2.8. It can be seen that the probability

decreases significantly upon decreasing temperature and an increasing magnetic field.

Chapter 2. Physical Background

Figure 2.8: Theoretical dependence for the recombination at two centers with the gfactors

g1=g2= 2 and with the spins S1=S2= 1/2, T=1.5 K [9].

Thus it is essential to record the RL-EPR at low temperatures. Going to higher mag-

netic fields and thus to higher microwave frequencies leads to a higher enhancement of

the RL due to the lower RL seen in higher magnetic fields.

2.3.3 Experimental Setup

A schematic diagram of a magneto-optical spectrometer for PL/RL-EPR measurements

is shown in Figure 2.9. The main part of the spectrometer is an Oxford Instruments bath

cryostat. The sample is placed in the middle of a superconducting magnet, which can

provide magnetic fields up to 4 T. Excitation is carried out with a deuterium lamp and

subsequent band pass or interference filters. The light emitted by the sample is detected

in the integral luminescence with different available edge filters. The microwave field is

coupled into the cavity via a coaxial conductor. The resonator can be varied by adjusting

the bottom of the cavity.

Helium supply in the sample cavity is provided by a needle valve. In case of a helium

filled sample cavity and a closed needle valve temperatures between 4.2 and 1.5 K can

be achieved by vapor pressure reduction of the helium in the sample cavity.

Chapter 2. Physical Background

Figure 2.9: Schematic diagram of a magneto-optical spectrometer.

2.4 X-ray Absorption Spectroscopy

2.4.1 Extended X-ray Absorption Fine Structure Oscillations

In the following section a brief introduction into the spectroscopic use of x-ray absorption

will be given. For further information the reader is referred to [10, 11]. Spectroscopic

methods dealing with x-ray absorption are usually called EXAFS (Extended X-ray Ab-

sorption Fine Structure) or XANES (X-ray Absorption Near Edge Structure). These

investigation origin from the fact, that the absorption edge of elements in a solid shows

a significant fine structure. This fine structure gives information about the vicinity of

the investigated atoms.

An incident x-ray beam can eject an electron from an atom. When this electron has

an energy of more than 40 eV above the edge energy, the electron can be described as

spherical wave. The wavelength λ= 2π/k is described by

k=s2me

¯h2(hν −E0) (2.5)

Chapter 2. Physical Background

Figure 2.10: Scheme of the origin of the EXAFS oscillations (a) for a constructive interference

of the incident and backscattered waves and (b) for a destructive one. Note, the distance between

the absorbing and the scattering atoms is (a) nλ and (b) (2n+ 1)λ/2.

where meis the electron mass, ¯hPlanck’s constant, and E0the binding energy of the

electron. This wave scatters at the surrounding atoms and these backscattered waves

can interfere constructively or destructively with the original wave which influences the

probability for the absorption of an incident x-ray beam. These effects are depicted

schematically in Figure 2.10 for four neighboring atoms.

A typical x-ray absorption spectrum is shown in Figure 2.11 (a). For the analysis

from the raw data a background function has to be subtracted. In addition this new

function has to be normalized that most of the EXAFS part is close to a value of 1,

see Figure 2.11 (b). By subtracting a value E0from the energy scale a pre-stage of

the EXAFS-analysis is achieved. E0is the energy value where the normalized x-ray

Chapter 2. Physical Background

Figure 2.11: (a) X-ray absorption spectrum of MnF2and the corresponding background func-

tion. (b) Normalized x-ray absorption spectrum of MnF2. (c) Pre-stage of the EXAFS analysis.

E0is the energy value where the normalized x-ray absorption reaches a value of 1.

Chapter 2. Physical Background

Figure 2.12: Half thickness of RbCdF3.

absorption has a value of 1. These data can be transferred from energy into k-space by

using Equation 2.5. To enhance these data for higher values of kit is possible to weight

these data with kn(1 ≤n≤3). A Fourier transformation leads to a transformation from

k-space into R-space. The spectra in R-space provide information about the type and

number of atoms in the neighbor shells.

2.4.2 Experimental Details

The linear attenuation coefficient µwas estimated by using the data given by the

National Institute of Standards and Technology [12]. The half thickness is given by

d1/2= ln2/µ and is shown for RbCdF3in Figure 2.12 for a density of 3.586 g/cm3[13].

The absorbance is given by

A= 1 −e−µd.(2.6)

Due to the high absorbance of RbCdF3at the Mn K-edge it is hardly possible to record

EXAFS spectra of single crystalline RbCdF3with absorption technique. In this case a

thin sample (≈50 µm) would be necessary. A different approach would be to crush a

Chapter 2. Physical Background

small amount of the single crystals and press them in cellulose based pellets. But in order

to keep the samples most of the measurements presented in this work were recorded in

fluorescence mode. The experiments were carried out at the 5BM-D beamline of the

Advanced Photon Source (Argonne National Laboratory). For detection a 12 channel

Ge-detector was available.

3 The Perovskite RbCdF3:Mn2+

3.1 Crystal Structure

At RT RbCdF3has a perfect cubic structure (Figure 3.1, left side) with a tolerance

factor of t= 0.86; the lattice parameter is a= 4.399 ˚

A [14]. It shows a structural phase

transition from [15] cubic (O1

h) to tetragonal (D4h

18 ) symmetry at T= 124 K (Figure 3.1,

right side).

Rubidium has an oxidation state of +1 (r= 1.47 ˚

A), cadmium +2 (0.97 ˚

A) and

fluorine −1 (1.33 ˚

A) [16]. Since Mn is a divalent ion it substitutes for a Cd-ion [15] and

is thus surrounded by a fluorine octahedron.

Single crystals of Mn-doped RbCdF3were grown in the Paderborn crystal growth lab-

oratory using the Bridgman method with a vitreous carbon crucible, argon atmosphere

and a stoichiometric mixture of CdF2and RbF powder to which 3, 100, 500, 2000, and

10000 ppm MnF2were added.

3.2 Photoluminescence

RbCdF3:Mn2+ shows a luminescence maximum at 560 nm, which can be attributed

to the Mn2+ impurity. This maximum can be seen in the PL and in the XL spectra

(Figure 3.2 dashed curves). It is due to the radiative 4T1g(G) →6A1g(S) transition of

Mn2+ in an Ohcrystal field symmetry [6, 8].

The PL excitation spectrum recorded before x-irradiation (Figure 3.2 (a), solid curve)

shows several maxima, which are also typical for Mn2+ in an Ohcrystal field. The

excitation bands peak at 310, 332, 351, 396, 428, and 513 nm being transitions from the

6A1g(S) ground level to the excited levels 4T1g(P), 4Eg(D), 4T2g(D), 4A1g(G) / 4Eg(G),

4T2g(G), and 4T1g(G), respectively. The excitation band at 240 nm is probably caused

by oxygen impurities or a double excitation of two Mn2+ ions [17].

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.1: Crystal structure of RbCdF3. a) The cubic and b) the tetragonal phase. The

tetragonal phase exists at temperatures below 124 K. c) The cubic phase in the ab-plane. d)

View on the ab-plane of the tetragonal phase.

After x-irradiation at RT the intensity of the Mn2+ emission is increased significantly

(Figure 3.2 (b), dotted curves). The excitation spectrum (Figure 3.2 (b), solid curve)

ranging from 200 to 550 nm shows an additional intense band peaking at about 300 nm.

The Mn2+-related excitation bands are superimposed by this radiation-induced band;

the transitions to the 4T1g(G), 4T2g(G), and the degenerate 4A1g(G) and 4Eg(G) energy

levels can still be observed as weak peaks.

UV-irradiation into the band at 300 nm leads to a bleaching of the luminescence

intensity. Figure 3.3 (a) shows the decay of the Mn2+ luminescence upon bleaching with

a laser diode at 375 nm. The decay is not mono-exponential which indicates that there

are different types of defects involved in the recombination process. UV-irradiation into

the Mn excitation bands should lead to a constant PL signal, which can be observed for

an excitation wavelength of 513 nm in Figure 3.3 (b). Contrary to this, irradiating into

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.2: (a) PL emission (dotted curve) and excitation (solid) spectra of RbCdF3doped

with 1% Mn2+, recorded (a) before and (b) after x-irradiation (tungsten anode, 60 kV, 15 mA,

5 min). In (b) the spectra recorded before x-irradiation are shown for comparison. The PL was

excited at 396 nm; the excitation spectra were detected at 560 nm.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.3: (a) PL intensity recorded for an excitation with a Laser-diode at 375 nm. (b) PL

intensities for an excitation of a RbCdF3doped with 1 % Mn2+ sample with different wavelengths

for a longer period of time. The PL was detected at 560 nm. All spectra were recorded at RT.

the Mn excitation bands at 396 and 420 nm leads to a fading of the PL signal. This

effect is due to the broad new excitation band at 300 nm on whose low energy slope

these bands are. In principle, for these experiments the largest bleaching effect should

be observed for UV-irradiation into the 300 nm band but these measurements were made

on the same sample and was x-irradiated only once.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.4: EPR spectrum of RbCdF3doped with 0.2 % Mn2+ recorded at RT in X-band

(9.35 GHz).

3.3 Magnetic Resonance

3.3.1 Electron Paramagnetic Resonance

A typical EPR spectrum of a RbCdF3:Mn2+ sample can be seen in Figure 3.4. In

principle six groups of EPR resonance lines can be observed.

The ground state of Mn2+ has a half-filled d-shell, i.e., S= 5/2 and L= 0. Since

L= 0 the gvalue should be close to the gvalue of the free electron; the gvalue was

found to be g= 1.99. Additionally the gvalue and the Mn2+ hyperfine interaction

(Mn has a nuclear spin of 5/2) are isotropic. The EPR spectra are dominated by the

Mn2+ hyperfine interaction (Ahf (55Mn)/h =−271 MHz), whereas the superimposed fine

structure splitting due to the crystal field of cubic symmetry is rather small (acf /h =

14 MHz) [18]. The lines are additionally split by the superhyperfine interaction with the

six nearest fluorine neighbors. The number of Mn2+ ions in the crystal can be extracted

from the area under the EPR lines.

In Figure 3.5 the PL intensities and the EPR signal intensities for different doping

levels are shown. Due to the fine structure splitting of the hyperfine groups the spectra

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.5: PL and EPR signal intensity of Mn2+ vs. Mn2+ doping level. The line is a guide

to the eye.

change significantly for different orientations of the magnetic field. Thus the EPR data

points were obtained by recording the spectra for different orientations of the magnetic

field (in angular steps of 5◦); the spectra were then integrated and averaged. It can

be seen, that the Mn2+ luminescence is clearly related to the Mn2+ embedded in the

crystal. Note, that the actual Mn2+ content may deviate from the doping level. It

can be estimated by fitting the temperature dependence of the magnetization to the

Curie-Weiss function. This was done in [19] for another set of Mn-doped RbCdF3.

3.3.2 Optically detected Magnetic Resonance

Figure 3.6 shows a PL-EPR spectrum of a RbCdF3single crystal doped with 0.0003%

Mn. The PL was excited at 1.5 K with the light of a deuterium lamp and a subsequent

212 nm interference filter and was detected in the integral emission using an edge filter

KV 470. The spectrum was recorded in the K-band (27.73 GHz) as microwave-induced

changes in the PL. It shows a broad resonance band at about 865 mT, which means a g

value of g=1.99. In addition, a superimposed structure can be observed which it can be

attributed to the Mn2+ hyperfine structure in RbCdF3[18]. Turning the sample away

from this orientation leads to a decrease of the PL-EPR signal, i.e. this spectrum can

only be observed in one specific orientation.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.6: PL-EPR spectrum of 0.0003% Mn-doped RbCdF3crystal. The spectrum was

recorded as microwave induces changes in the integral PL at 1.5 K applying a microwave fre-

quency of 23.73 GHz. The PL was excited with a deuterium lamp and a subsequent 212 nm

interference filter and detected in the integral luminescence with an edge filter (KV 470).

3.3.3 Electron Paramagnetic Resonance after x-irradiation

Figure 3.7 shows the change in the EPR signal intensity after x-irradiation for different

periods of time. After intense x-irradiation the Mn2+ signal intensity could be reduced

by 15-20% of its original value. As already mentioned above, the EPR data points were

obtained by recording the spectra for different orientations of the magnetic field (in

angular steps of 5◦); the spectra were then integrated and averaged. The error bars are

the root mean square deviation of the average value. As already observed by Dotzler et

al. [19] a saturation effect can be found upon prolonged radiation. We did, however, not

observe any resonance lines from the radiation-induced Fcenters. We assume that the

Mn2+ EPR signal overlaps the (probably broad) Fcenter resonance which can thus not

be detected. The original Mn2+ EPR signal intensity can be restored upon bleaching

with a laser diode (10 mW) at 375 nm (Figure 3.8).

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.7: EPR signal intensity of Mn2+ in 0.0003% Mn-doped RbCdF3vs. x-irradiation

time; the x-irradiation was carried out with a tungsten anode (50 kV, 30 mA). The line is a

guide to the eye.

Figure 3.8: X-band EPR spectra showing one of the Mn2+ hyperfine groups in 0.0003% Mn-

doped RbCdF3. The spectra were recorded after 12 hours x-irradiation (black curve) and after

subsequent bleaching with a laser diode at 375 nm (red curve). The bleaching was carried out in

situ (inside the cavity), i.e., the spectra before and after bleaching were recorded using the same

experimental conditions, in particular, the same orientation of the magnetic field.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.9: Normalized x-ray fluorescence of Mn-doped RbCdF3. The doping levels are 1%

(solid curve), 0.2% (dashed curve), 0.05% (dotted curve), and 0.01% (dash-dotted). As excitation

a monochromatic x-ray beam with an energy of 6640 eV was used. The inset shows the peak

values of the Mn x-ray fluorescence for different doping levels.

3.4 Beamline Experiments

3.4.1 X-ray Fluorescence

X-ray fluorescence spectra of RbCdF3with Mn doping levels of 1%, 0.2%, 0.05%, and

0.01% are shown in Figure 3.9. The strongest fluorescence peak is centred about 3200 eV,

having smaller maxima at 3150, 3285, and 3935 eV. They are due to the Cd Lα1,2, Lβ1,

and Lγ2,3-transitions of the host lattice. The peak at 5880 eV can be attributed to the

K-series of the Mn dopant, whereas the maximum at 6580 eV is the backscatter peak,

which is an inelastic scattering of the exciting x-irradiation. The inset of Figure 3.9

shows the peak values of the Mn x-ray fluorescence versus doping level. It is apparent,

that the value of the peak is related to the Mn doping level.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.10: (a) Increase of the scintillation intensities in Mn-doped RbCdF3upon x-irradiation

with a photon energy of 17 keV. The doping concentrations were 1% (solid curve), 0.2% (dotted),

and 0.01% (dashed). (b) Saturation value of the scintillation intensities versus Mn2+-doping level;

full quares represent the experimental data points, the line is a guide to the eye.

3.4.2 Scintillation Intensity

In the following section the scintillation intensity of Mn-doped RbCdF3is described. The

measurements were performed at the 5BM-C beamline of the Advanced Photon Source

(Argonne National Laboratory). A 6 mm (horizontal) ×4 mm (vertical) monochromatic

x-ray beam is used as excitation light source. The scintillation intensity of the sample is

detected by a cooled charge-coupled device (CCD) camera through a 4X objective lens.

Continuous x-irradiation leads to an increased XL intensity. This effect is shown in

Figure 3.10, where the samples where irradiated with x-rays having an energy of 17 keV.

The different values at 0 s are due to the different doping levels of the samples, i. e.

1% (solid curve), 0.2% (dotted), and 0.01% (dashed). In Figure 3.10 (b) the saturation

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.11: Normalized XANES spectra of 1% Mn-doped RbCdF3single crystal (black curve),

Mn metal foil (red curve), MnF2powder (green curve), and Mn2O3(blue curve). The spectra

for RbCdF3:Mn, Mn foil, and MnF2were recorded in fluorescence mode, the one for Mn2O3was

taken from [20].

value of the scintillation intensities is plotted. For highly-doped samples the saturation

value of the scintillation intensity clearly corresponds to the Mn-doping level. The strong

deviation from linearity for the 0.0003% Mn-doped sample is probably caused by the the

crystal growth process, where possibly too much MnF2was added to the melt.

3.4.3 EXAFS on Mn-doped RbCdF3

EXAFS spectra of different Mn compounds and RbCdF3:Mn2+ are shown in Figure 3.11.

The highest absorption for Mn in RbCdF3is at 6550 eV, what corresponds to the highest

absorption of the divalent Mn in MnF2lying also at 6550 eV. The absorption maxima of

the metallic Mn foil is at 6553 eV and of trivalent Mn in Mn2O3at 6559 eV. A normalized

EXAFS spectrum of a 1% Mn-doped RbCdF3can be seen in Figure 3.12. The XANES

spectra for 0.05%, 0.2%, and 1% doping levels of Mn in RbCdF3are shown in Figure 3.13

for comparison. All spectra show the same structure in their oscillations. There are only

slight differences in the maximal values of the absorption oscillations. Due to the bad

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.12: Normalized EXAFS spectrum of 1% Mn-doped RbCdF3, recorded in fluorescence

mode.

Figure 3.13: Normalized XANES spectra of Mn-doped RbCdF3. The doping levels are 1%

(black curve), 0.2% (green curve), and 0.05% (red curve). The spectra were recorded in fluores-

cence mode.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.14: The solid curve shows the Fourier Transform of the Mn K-edge of Mn-doped

RbCdF3. The inset depicts the normalized k-weighted EXAFS. The solid lines are the experi-

mental data and the dashed line is the simulation.

signal-to-noise ratio for the 0.2% and the 0.05% Mn-doped samples an EXAFS analysis

could only be done for the 1% doped sample. The Fourier Transform of the Mn K-edge

in RbCdF3is depicted in Figure 3.14 (solid curve), with the corresponding normalized

k-weigthed EXAFS in the inset. A calculation is also shown in Figure 3.14 (dashed

curve) for comparison. The calculation was carried out using the Artemis program of

the Athena program package [21].

The most important paths for the calculated spectrum and their contribution to the

total spectrum can be seen in Figure 3.15. Path one which is due to the six F−ions lying

in a <100>direction gives the most significant contribution to the Fourier Transform.

Representatively one of the six ions is marked with 1. The second path is attributed to

the 24 neighboring F−pairs, marked with 2. Its contribution is much smaller and has an

opposite sign with respect to path one, which leads to a dip at 3 ˚

A in the total spectrum

of paths one and two. The third path is due to eight Rb+ions lying in a <111>direction

from the Mn core ion; this leads to the peak at 3.5 ˚

A. The fourth peak at about 4.2 ˚

A is

due to a path from the core ion into a <100>direction to the 6 neighboring Cd2+ ions.

Chapter 3. The Perovskite RbCdF3:Mn2+

Figure 3.15: Fourier Transforms of the different paths as indicated in the model, see top right

corner. The left side shows the Fourier transforms of the single path, the right side depicts the

the sum of the corresponding paths.

Chapter 3. The Perovskite RbCdF3:Mn2+

3.5 Discussion

Mn-doped RbCdF3shows typical Mn2+ related PL emission and excitation spectra. The

maximum at 230 nm is generally assumed to be due to oxygen impurities in the crystal.

Another explanation was proposed by Ferguson [17]: The band at 230 nm could be due

to a simultaneous electronic excitation of a pair of Mn2+ ions. Thus, the band could be

an excitation of 4T1g(G) and 4Eg(G) states.

Due to the charge state of the Mn dopant Mn2+ probably substitutes for divalent Cd.

ENDOR measurements supported this idea [15]. This model was used for the calculation

of EXAFS. It can be seen that the first and most intense shell can be attributed to the

six F−ions forming an octahedron around the Mn dopant. The second and the third

shell are due to Rb and Cd ions, respectively. The experimental spectrum is in good

agreement with the calculated spectrum, in particular for the fluorine neighbors. The

shells attributed to Rb and Cd cannot be seen so clearly.

Losada et al. [2] found in Mn-doped RbCdF3after x-irradiation an absorption band

at 310 nm. They also found an increased Mn2+ luminescence. They suggested that

defects close to Mn2+ were created by x-irradiation. The defects perturb the Mn2+

environment. Because of this, the previously spin and parity forbidden transition get

partially allowed and an enhanced Mn2+ luminescence can be observed. However, the

same absorption band at 310 nm was also found in undoped RbCdF3after x-irradiation.

Lifetime measurements of x-irradiated samples excited at 310 nm also made by Losada

et al. [2] showed the same lifetime value as the unirradiated sample. This shows that

the Mn2+ luminescence found in irradiated crystals comes from unperturbed Mn2+ ions.

Since the 310 nm absorption band is caused by intrinsic defects (F-centers) but the

emission is due to Mn2+ ions, there is probably an energy transfer between these intrinsic

defects and the manganese. It was suggested by Losada et al. and in recent work from

Dotzler et al. [19] that x-irradiation produces Mn3+ ions and electron trap centers (F-

centers). Upon optical excitation of the radiation-induced absorption band the trapped

electron is released and recombines with a Mn3+ ion in the vicinity, followed by a Mn2+

emission. The bleaching effect is thus related to the destruction of the defect centers

created during x-irradiation.

After x-irradiation the PL excitation spectrum shows that the excitation bands at

297 and 345 nm are more enhanced than the other bands. In case of a perturbed Mn

environment an enhancement of all Mn2+ excitation bands would be expected. But as

Chapter 3. The Perovskite RbCdF3:Mn2+

only two excitation maxima are enhanced, this increase is possibly not associated to

the perturbation produced by defects close to the Mn2+ ions but to the fact that the

absorption band at 310 nm superimposes the excitation spectrum. Therefore, it is likely

that it is due to energy transfer between these intrinsic defects and the manganese ion,

as mentioned above.

The bleaching effect, seen in Figure 3.3, where the bands at 396, 343 and 293 nm

show a decreasing PL intensity, is probably related to the destruction of defect centers

created during X-irradiation [22]. The EPR measurements showed that the number of

Mn2+ ions is reduced by X-irradiation and can be restored by optical bleaching. This

agrees with the model suggested by Dotzler et al. [19] that optically stimulable Mn3+-F

complexes are formed by X-irradiation.

4 EXAFS on Eu-doped CaF2

A comparable problem like Mn2+ in RbCdF3:Mn is Eu2+ in CaF2. The crystal structure

of CaF2is shown in Figure 4.1. The dopant is located in the center of a fluorine cube.

Different vacancies and impurities were discussed for this compound. In the following

section preliminary results of EXAFS measurements made at the Advanced Photon

Source (APS) and at HASYLAB are shown.

4.1 Advanced Photon Source

The x-ray fluorescence spectrum of 2% Eu-doped CaF2is shown in Figure 4.2. It shows

three emission peaks at 3690, 5850, and 6870 eV, which can be attributed to transitions

of the Ca K-series, Eu Lα1,2line and backscattered x-rays (x-rays inelastically scattered

at the sample onto the detector), respectively. The normalized fluorescence Eu LIII-

edge and Eu LII-edge EXAFS spectra are shown in Figure 4.3. The corresponding

k-weighted EXAFS of the Eu LIII-edge (inset) and the Fourier transform spectrum are

shown in Figure 4.4; the analysis was done using the Athena Software [21]. There are

two dominant shells at approximately 1.7 and 3.5 ˚

4.2 HASYLAB

Comparable experiments were also performed at the E4 beamline of HASYLAB in Ham-

burg. In contrast to the previous measurements these spectra were recorded in absorp-

tion. Approximately 10 mg of 0.1% and 2% Eu-doped CaF2were crushed and mixed

with a cellulose powder. These were afterwards pressed in pellets, which were used for

the experiments.

The x-ray absorption spectrum of 2% Eu-doped CaF2can be seen in Figure 4.5 (black

curve). In principle, it shows the same features as the spectra recorded at the APS

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.1: Crystal structure of CaF2. The ionic radii are Ca2+ 0.99 ˚

A, Eu 1.09 ˚

A, and F

1.33 ˚

Figure 4.2: Normalized x-ray fluorescence spectrum of 2% Eu-doped CaF2, recorded for an

excitation energy of 7080 eV.

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.3: (a) Normalized EXAFS spectrum of 2% Eu-doped CaF2at the Eu LIII-edge and

(b) at the Eu LII-edge, recorded in fluorescence mode.

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.4: Fourier transform of the Eu LIII-edge in 2% Eu-doped CaF2. The inset shows the

normalized EXAFS (k) weighted of the Eu LIII-edge in 2% Eu-doped CaF2.

Figure 4.5: Normalized EXAFS spectrum of 2% Eu-doped CaF2at the Eu LIII-edge, recorded

in absorption (black curve). The spectrum recorded in fluorescence mode at the APS is shown

for comparison (red line).

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.6: Normalized EXAFS spectrum of 0.1% Eu-doped CaF2at the Eu LIII-edge, recorded

in absorption.

(red curve), which was further supported by a preliminary analysis of this experiment.

Despite the weaker beam intensity the HASYLAB absorption spectra of the 2% Eu-

doped sample have a similar signal-to-noise ratio as the spectra recorded in fluorescence

mode at the APS.

The x-ray absorption spectrum of 0.1% Eu-doped CaF2is shown in Figure 4.6. Al-

though the signal-to-noise ratio is much weaker than that for the 2% doped sample the

EXAFS spectrum shows similar oscillations as already seen in Figure 4.5 for the 2%

doped sample.

The absorption setup at the E4 beamline at HASYLAB allowed to record the x-ray

absorption of a 0.1% Eu-doped CaF2sample in a relatively good signal-to-noise ratio.

For further experiments, EXAFS spectra should preferably be recorded in absorption

instead of fluorescence mode.

Chapter 4. EXAFS on Eu-doped CaF2

4.3 Discussion

Calculations of the Fourier transform were performed with the Artemis program [21]

package and are shown in Figures 4.7, 4.8, and 4.9. The models used for the calculations

are shown as well; these show only the nearest neighbors of the Eu2+ dopant. The white

circles represent F−ions, the gray ones Eu2+ ions, the black ones O2−ions; the boxes

are F-vacancies.

The calculations for an unperturbed Eu2+ dopant in CaF2, with one F-vacancy and

a combination of one F-vacancy with one O2−impurity are shown in Figure 4.7. The

most significant difference in the spectra can be seen in the first shell at approximately

1.7 ˚

A. An F-vacancy leads to a decrease of this peak, whereas an O2−ion does not

affect the Fourier transform considerably. In Figure 4.8 the possible combinations of

two F-vacancies are shown. In principle there are three ways to arrange an F-vacancy

pair: Along <100>, along <110>and along <111>. All three configurations show the

same Fourier transform, thus it is not important, which two F−ions are missing from a

possibility of six. Compared to the Fourier transform of an Eu-vFcomplex in CaF2the

complex consisting of two F-vacancies shows a much lower peak at 1.7 ˚

A; i.e. the less

F-ions in the vicinity of a Eu2+ ion the smaller the first shell.

In [23] different models of the surrounding of an Eu-dopant in CaF2have been pre-

sented. Three of them and their Fourier transform are shown in Figure 4.9. The figure

illustrates, that the most significant changes in the spectra are due to the number of

missing F−-ions. The height of the first peak decreases with decreasing number of sur-

rounding F−-ions. In contrast an increasing number of substituting O2−-ions leads only

to a slight increase of the second and third shell at 3.5 ˚

A and at 3.9 ˚

A, respectively.

At this point, the analysis indicates that a complex consisting of two fluorine vacancies

in the vicinity of an Eu2+ ion leads to the EXAFS spectrum observed. A statement about

the positions of the fluorine vacancies cannot be made. However, this interpretation is

questionable since recent M¨ossbauer investigations on Eu-doped CaF2[24] showed, that

oxygen neighbors do not play such a dominant role as suggested in [23]1.

1Furthermore this assumption is supported by a work published after completion of this thesis [25].

The Eu-ion seems to be in an cubic surrounding with no pertubation in the vicinity.

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.7: Model and Fourier transform of the Eu LIII-edge in 2% Eu-doped CaF2.

Figure 4.8: Model and Fourier transform of the Eu LIII-edge in 2% Eu-doped CaF2.

Chapter 4. EXAFS on Eu-doped CaF2

Figure 4.9: Model and Fourier transform of the Eu LIII-edge in 2% Eu-doped CaF2.

5 The Inverse Perovskite LiBaF3

5.1 Crystal Structure

LiBaF3is very different from the other crystals discussed earlier. Lithium has an oxida-

tion state of +1 (r= 0.68 ˚

A) and barium of +2 (1.34 ˚

A) [16].

Since LiBaF3has a tolerance factor of 0.94, it has a cubic perovskite structure. It has

no distortions even at low temperatures [26] and the lattice parameter has a value of

a= 3.995 ˚

A [26].

However, the structure of LiBaF3is different from that of RbCdF3. It has a so-called

inverse perovskite structure [27]. The monovalent Li+ion is surrounded by a fluorine

octahedron and Ba2+ is surrounded by 12 fluorines.

Although Mn2+ has the same charge as Ba2+ it does not substitute for the Ba ion.

Yosida [28] showed that the Mn2+ ion substitutes for the Li+ion, which is surrounded

by a fluorine octahedron. This is the same as in the other samples. Mn2+ is located on

the Me site in Figure 2.1 and so it has the same environment as Mn2+ in the other two

crystals.

Single crystals of Mn-doped LiBaF3were grown in the Paderborn crystal growth lab-

oratory using the Bridgman method with a vitreous carbon crucible, argon atmosphere

and a stoichiometric mixture of LiF and BaF2powder to which 2000 molar ppm of MnF2

were added.

5.2 X-ray Fluorescence

The X-ray fluorescence spectrum of 0.2% Mn-doped LiBaF3shows several maxima, see

Figure 5.1. The peaks at 4460, 4840, 5140, and 5540 eV can be allocated to transitions

of the L-series of Bai; i.e. Lα1,2, Lβ1, Lβ2, and Lγ1. The Mn K-series can only be seen as

shoulder of the Ba L-series at 5780 eV. The backscatter peak can be found at 6540 eV.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.1: Normalized x-ray fluorescence of 0.2% Mn-doped LiBaF3. For excitaion a

monochromatic x-ray beam with an energy of 6640 eV was used.

5.3 Oxygen Vacancy Complex

Mn2+ emits at 712 nm and has a characteristic excitation spectrum. After x-irradiation

the magnetic circular dichroism (MCD) of the optical absorption indicated that an F-

type center interacts with the Mn2+ dopant.

In oxygen-doped LiBaF3Shiran and Voronova [29] found a luminescence band peaking

at 420 nm which is caused by the oxygen dopant. Besides the Mn2+ emission we found

in Mn-doped LiBaF3luminescence bands peaking at 423 and 480 nm. These bands

correspond to those observed in oxygen-doped LiBaF3[29] and are tentatively attributed

to oxygen impurities which are difficult to avoid in fluoroperovskites.

5.3.1 Photoluminescence

Figure 5.2 (a) shows the PL emission (solid curve) and excitation (dashed and dotted

curves) spectra of an oxygen-related complex in Mn-doped LiBaF3. In PL, one band is

peaking at 423 nm and a weaker one at 480 nm. The corresponding excitation spectra

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.2: (a) PL emission (solid curve) and excitation (dashed and dotted curves) spectra of

an oxygen-related complex in 0.2% Mn2+-doped LiBaF3recorded at RT. The PL was excited at

220 nm. The excitation spectra were detected at 423 nm (dashed) and 480 nm (dotted). The

increase in the excitation spectra is caused by incident excitation light. (b) PL emission (solid

curve) and excitation (dashed curve) spectra of an oxygen-related complex in LiBaF3:Mn2+

recorded at 20 K. The PL was excited at 220 nm; the excitation spectrum was detected at

475 nm. The excitation spectrum detected at 430 nm is not shown; it is identical to the one

detected at 475 nm.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.3: PL lifetime of an oxygen related complex in 0.2% Mn2+-doped LiBaF3at RT

excited with a nitrogen flash-lamp and detected at 423 nm.

show intense bands at 220 and 398 nm. The radiative lifetime of the 423 nm luminescence

is (2.5±0.3) ms (Figure 5.3). The 423 and 480 nm luminescence bands are not caused

by the Mn dopant.

At 20 K the 423 and 480 nm PL bands do not change significantly. Here both lu-

minescence bands have the same intensity and peak at about 430 and 475 nm. Their

excitation spectra are identical with their maxima at 220 and 412 nm (Figure 5.2 (b)).

5.3.2 Photoluminescence detected Electron Paramagnetic Resonance

The ground state of an O2−impurity is diamagnetic and can thus not be detected by

EPR. However, the excited state of O2−can become paramagnetic if the excited electron

and the remaining unpaired electron have parallel spins, i.e. if the excited (O2−)∗is in

a triplet state.

Figure 5.4 shows a PL-EPR spectrum detected in the oxygen-related luminescence

bands found in Mn2+-doped LiBaF3. The spectrum was recorded in the K-band (23.73

GHz) for a magnetic field orientation parallel to [110]. The PL was excited at 1.5 K

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.4: PL-EPR spectrum of an oxygen-related complex in LiBaF3:Mn2+ for an orientation

of the magnetic field parallel to [110]. The spectrum was recorded as microwave-induced changes

in the integral PL at 1.5 K applying a microwave frequency of 23.73 GHz. The PL was excited

with a deuterium lamp and a subsequent 218 nm interference filter and detected in the integral

luminescence with an edge filter (KV 380).

with a deuterium lamp and a subsequent 218 nm interference filter and detected in the

integral luminescence with a 380 nm edge filter. The angular dependence of these lines

for a rotation of the magnetic field in a {110}plane is presented in Figure 5.5. We

used the spin Hamiltonian of a triplet spin (S= 1) system with an orthorhombic fine

structure (FS) tensor to analyze the angular dependence of the PL-EPR line, i.e.

H=µB·~

B·g·~

S+~

S·D·~

S(5.1)

where gand Dare the gand Dtensors, µBis the Bohr magneton, ~

Bis the magnetic

field vector and ~

Sis the electron spin operator. The orientations of the gand Dtensors

can be described with a set of Euler angles. In the principal axes system the gtensor

is characterized by its three principal values gxx,gyy, and gzz. The Dtensor can be

expressed with the two FS values Dand Ein the principal axes system which have the

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.5: Angular dependence of the PL-EPR lines, recorded for a rotation of the magnetic

field in a {110}plane. The spectra were recorded as microwave-induced changes in the integral

PL at 1.5 K applying a microwave frequency of 23.73 GHz. The PL was excited with a deuterium

lamp and a subsequent 218 nm interference filter and detected in the integral luminescence with

a 380 nm edge filter. The open squares represent the experimental line position; the dashed and

solid (two-fold degenerate) lines are calculated by using the FS parameters of Table 5.3.2. The

lines marked with “∗” consist of two magnetically non-equivalent lines, which cannot be resolved

here.

usual meaning as the axial (D= 3/2·Dzz) and the non-axial (E= 1/2·(Dxx−Dyy)) parts,

respectively (see e.g. [30]). The choice of the zaxis corresponds to the maximal absolute

value of the FS splitting. The calculated EPR angular dependence for a rotation of the

magnetic field in a {110}plane is shown in Figure 5.5 by solid lines. The parameters

are listed in Table 5.3.2. All observed resonances are due to “allowed” (∆mS=±1)

transitions. The principal axes zof the gand the Dtensors are assumed to be collinear

and parallel to a <110>direction. The xand yaxes of the gand Dtensors are rotated

by an angle of 25◦about the zaxis and are thus not aligned along any crystal axes.

In principle, a triplet spin system having orthorhombic symmetry with the principle

axis, z, along a <110>direction should show six magnetically non-equivalent center

Chapter 5. The Inverse Perovskite LiBaF3

orientations, if the xand yaxes of the Dtensor are aligned exactly along the crystal

axes <100>and <110>. In our case neither the xnor the yaxis coincides with a

<100>or a <110>direction. Thus, each of the six orientations mentioned above is

split in two further magnetically non-equivalent center orientations. The spectrum of

each center orientation should be split by the fine structure interaction into two lines

symmetrical about g= 2. Altogether up to 24 resonance lines could appear in the PL-

EPR spectra. The largest splitting of 2Dis for the magnetic field parallel to a <110>

direction. However, for a rotation of the magnetic field in a {110}plane the angular

dependence shows only up to 16 resonance lines, 8 of which are two-fold degenerate (see

Figure 5.5). The number of lines is reduced to 8 for Bk<110>and to 4 for Bk<100>.

Note, that for Bk<110>only 6 lines can be observed in the experimental spectrum

(indicated by bars in Figure 5.4). It was not possible to determine the sign of the FS

parameter D, experimentally by e.g. magnetic circular polarization of emission (MCPE)

measurements.

The highest PL-EPR line intensity was obtained for a photon energy of 2.7 eV (460 nm)

which corresponds to the low temperature luminescence spectrum of Figure 5.2 (b) (solid

curve). The best excitation energy using a set of interference filters was found to be at

approximately 5.7 eV (218 nm) which is in agreement with the excitation spectrum

depicted in Figure 5.2 (b) (dashed curve).

gxx gyy gzz |D/geβe| |E/geβe|ϑ

(mT) (mT) (◦)

1.97 1.98 1.98 207 52 25

±0.01 ±0.01 ±0.01 ±2±2±1

Table 5.1: Parameters of the spin Hamiltonian (equation 5.1) for an oxygen-related lumines-

cence center in Mn-doped LiBaF3. The z-axis of the gand Dtensor, respectively, is parallel to

a<110>direction. The xand yaxes are not aligned along the crystal axes. The xaxis is tilted

in a {110}plane by an angle ϑwith respect to a <100>direction.

Chapter 5. The Inverse Perovskite LiBaF3

5.3.3 Discussion

The analysis of the PL-EPR data shows that the triplet state of the 460 nm luminescence

can be described by gand Dtensors with the zaxes along a <110>direction. The x

and yaxes are not aligned along the crystal axes. The xaxis is tilted in a {110}plane by

an angle of 25◦with respect to a <100>direction. The gvalues as well as the spectral

position of the 460 nm luminescence are very similar to those found for oxygen-vacancy

complexes in other fluoride crystals [31–33]. Thus, we assign our luminescent triplet

state to an oxygen impurity center in LiBaF3. O2−usually substitutes for fluorine and

has an extra negative charge which is usually compensated by an fluorine vacancy nearby

[31–33]. In LiBaF3the O2−

F-vFcomplex is aligned along a <110>direction which is in

agreement with the orientation of the principal axis zof the Dtensor. However, we still

have to find an explanation for the observation that the xand yaxes are not aligned

along the crystal axes, but tilted in a {110}plane. We assume a perturbation close to

the oxygen-vacancy complex, e.g. Mn2+ at a Ba2+ site (Figure 5.6). In this case no

additional charge compensation is necessary. The connecting line between the center

of an unrelaxed O2−

F-vFcomplex and the Ba2+ site nearby makes an angle of 35◦with

respect to a <100>direction. However, relaxations in the O2−

F-vF-Mn2+

Ba complex might

lead to the measured angle of 25◦. Note, that the ionic radius of the Mn2+ ion (0.80 ˚

is much smaller than that of the Ba2+ ion (1.34 ˚

A).

In principle, there would be two further alternatives for an oxygen-related complex

considering the necessary charge compensation: (i) Mn2+ at a Li+site or (ii) a Mn ion

in its possible charge state of Mn3+ at a Ba2+ site. The first possibility can be excluded

because the O2−

F-Mn2+

Li complex would be aligned along a <100>direction. For (ii)

the O2−

F-Mn3+

Ba complex would be aligned along a <110>direction, but in this case we

cannot explain the tilt of the xand yaxes of the Dtensor with respect to the crystal

axes.

5.4 Mn-related Complex

Beside this oxygen vacancy complex having its emission at 420 nm, Mn2+ shows its

characteristic luminescence at 710 nm. This dopant can be observed in EPR, but in

PL-EPR no resonance lines could be observed in this emission band.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.6: Model of the O2−

F-vF-Mn2+

Ba complex in Mn-doped LiBaF3.

5.4.1 Photoluminescence

At RT Mn-doped LiBaF3shows a broad emission at 710 nm, which can be attributed

to the Mn2+ dopant, see Figure 5.7 (a). The corresponding excitation spectrum shows

several maxima between 300 and 620 nm. These are due to the excited states of Mn2+

in an Ohcrystal field symmetry as indicated. The excitation bands at 210 and 255 nm

are probably caused by an oxygen impurity [29] or higher excited states of Mn2+ [17].

While at 20 K the Mn2+ emission is still at 710 nm the Mn2+ excitation bands are

completely superimposed by a broad excitation band at about 510 nm, see Figure 5.7 (b).

This band is temperature dependent, i.e. it increases with decreasing temperature, what

can be seen in Figure 5.8 (a), where the excitation spectra for the 710 nm band is shown

for different temperatures. At about 80 K its intensity starts to increase significantly,

see Figure 5.8 (b).

After x-irradiation a broad emission band at 610 nm can be observed at RT; the band

at 710 nm remains as a small shoulder of the 610 nm band, see Figure 5.9 (a). The

corresponding excitation spectrum shows an intense band at 465 nm and a broad one at

about 530 nm.

At 20 K, RT x-irradiated LiBaF3:Mn has two emission bands: One at 610 nm and

the second at 720 nm, see Figure 5.9 (b). The excitation spectrum of the 610 nm is

comparable to the spectrum recorded at RT. There is an intense band at about 460 nm

and a broad one at about 550 nm with the full width at half maximum, however, smaller

than at RT. In addition, two new bands in the UV spectral range can be observed at 340

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.7: PL emission (dashed curves) and excitation (solid curves) spectra of Mn-doped

LiBaF3. (a) RT: The PL was excited at 255 nm; the PL excitation spectrum was detected at

710 nm. (b) 20 K: The PL was excited at 510 nm; the PL excitation was detected at 710 nm.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.8: (a) Excitation spectra of Mn-doped LiBaF3detected at 710 nm and different

temperatures. The intensity decreases with increasing temperature. (b) shows the PL emission

intensity excited at 510 nm versus temperature.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.9: PL emission (dashed curves) and excitation (solid curves) of Mn-doped LiBaF3

x-irradiated at RT. (a) RT: The PL emission was excited at 460 nm; the PL excitation spectrum

was detected at 610 nm. (b) 20 K: The PL was excited at 460 nm; the PL excitation spectrum

was detected at 610 nm. (c) 20 K: The PL emission was excited at 460 nm; the PL excitation

spectrum was detected at 710 nm.

Chapter 5. The Inverse Perovskite LiBaF3

and at 280 nm. The excitation spectrum of the 720 nm emission band (spectrum not

shown) has an intense excitation peak at 630 nm superimposed on the broad excitation

band at about 510 nm which has already been observed in the non-irradiated sample.

5.4.2 Photoluminescence detected Electron Paramagnetic Resonance

Figure 5.10 shows the PL-EPR spectra of Mn-activated LiBaF3, recorded in the 610 nm

luminescence band after x-irradiation at RT. The spectra were recorded at a microwave

frequency of 23.6 GHz (K-band) for an orientation of the magnetic field parallel to the

<110>, the <111>, and the <100>direction, respectively. The PL was excited at 1.5 K

with a deuterium lamp and a subsequent DUG 11 band-pass filter and detected in the

integral luminescence with a 550 nm edge filter (KV 550).

The angular dependence of the PL-EPR lines indicates that we again deal with a

triplet spin state system (S= 1). The principal axis, z, of the fine structure tensor

is close a <110>direction. But compared to the oxygen vacancy complex the angular

dependence is completely different. But an analysis the same spin Hamiltonian as shown

in Equation 5.1.

The resonances in the magnetic field ranging from 300 to 1600 mT are due to ’al-

lowed’ (∆mS=±1) transitions; the resonances at about 200 mT can be attributed to

’forbidden’ (∆mS=±2) transitions. The forbidden transitions are opposite in sign but

comparable in intensity.

For this rotation of the magnetic field in a (110) plane the angular dependence shows

up to 12 resonance lines, which are reduced to eight lines for a magnetic field parallel

to <110>, indicated by bars is Figure 5.10. For the magnetic field parallel to <111>

three lines can be observed and six lines for the magnetic field parallel to the <100>

direction, i.e. most of the lines are degenerate (Figure 5.10). The largest splitting of the

resonance lines is for a magnetic field parallel to a <110>direction. It was not possible

to determine the sign of the FS parameter, D, experimentally by MCPE measurements.

The PL-EPR spectra show, however, additional resonances, e.g. one of it for a mag-

netic field orientation parallel to a <111>direction (marked with an asterisk in Fig-

ure 5.10). These resonances are tentatively assigned to a second Mn-related lumines-

cence center having the same symmetry but different gand FS parameters. A detailed

analysis of this center is in progress.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.10: PL-EPR spectra of a Mn-related complex in x-irradiated LiBaF3:Mn. The spectra

were recorded at 1.5 K and a microwave frequency of 23.97 GHz for different orientations of the

magnetic field with respect to the crystal axis system. The PL was excited using a deuterium

lamp and a subsequent band-pass filter (DUG 11) and detected in the integral luminescence

using an edge filter (KV 550).

The highest PL-EPR line intensity was obtained for a photon energy of about 2 eV

(610 nm), which corresponds to the low temperature photoluminescence spectrum (Fig-

ure 5.9 (b), dashed curve). The best excitation energy using a band-pass filter was found

to be between 3.1 and 5 eV (250 to 400 nm) which is in agreement with the excitation

spectrum depicted in Figure 5.9 (b), solid curve. Although a significant higher lumines-

cence intensity could be achieved by exciting into the radiation-induced defect band at

about 460 nm, we did not want to have any direct excitation of the Mn2+ bands and

focused thus on the UV excitation bands.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.11: Angular dependence of the PL-EPR lines with forbidden transitions, recorded

for a rotation of the magnetic field in a {110}plane. The spectra were recorded as microwave-

induced changes in the integral PL at 1.5 K applying a microwave frequency of 23.97 GHz. The

PL was excited with a deuterium lamp and a subsequent band-pass (DUG 11, 240 nm < λexc.<

400 nm) and detected in the integral luminescence using an edge filter (KV 550, λem.>550 nm).

The open squares represent the experimental line position; the solid (two-fold degenerate) lines

are calculated by using the FS parameters of Table 5.4.2 (a).

The angular dependence of the PL-EPR lines is shown in Figures 5.11 and 5.12. For

the angular dependence with the forbidden transitions the fit is not very good. There are

significant deviations from the recorded data points. A fit for the angular dependence

without the forbidden transitions is in better agreement with the experimental results.

However, both fits show have the same symmetry an comparable gand FS parameters,

see Table 5.4.2.

Chapter 5. The Inverse Perovskite LiBaF3

Figure 5.12: Angular dependence of the PL-EPR lines without forbidden transitions, recorded

for a rotation of the magnetic field in a {110}plane. The spectra were recorded as microwave-

induced changes in the integral PL at 1.5 K applying a microwave frequency of 23.97 GHz. The

PL was excited with a deuterium lamp and a subsequent band-pass (DUG 11, 240 nm < λexc.<

400 nm) and detected in the integral luminescence using an edge filter (KV 550, λem.>550 nm).

The open squares represent the experimental line position; the solid (two-fold degenerate) lines

are calculated by using the FS parameters of Table 5.4.2 (b).

5.5 Discussion

The luminescence properties of Mn-activated LiBaF3are complex: At RT, PL emission

and excitation spectra of non-irradiated, Mn-doped LiBaF3show the typical features

of the Mn2+ dopant; at low temperature, however, a new broad excitation appears at

about 510 nm. After RT x-irradiation, the PL emission spectra show a new emission

band at about 610 nm while the PL excitation spectra show radiation induced absorption

bands at 460 nm, 530 nm, and in the ultraviolet spectral range. Many different defects,

intrinsic or radiation-induced, are involved in the luminescent processes.

In [34] it has already been proposed, that an F-type center is responsible for the

radiation-induced 460 nm absorption/excitation band. Based on magnetic circular dichro-

ism of the optical absorption (MCD) and MCD-detected EPR measurements it was

assumed that the F-type center interacts with an S≥1 center.

Chapter 5. The Inverse Perovskite LiBaF3

gxx gyy gzz |D/geβe| |E/geβe|

(mT) (mT)

(a) 2.08 1.97 1.97 640 85

±0.02 ±0.02 ±0.02 ±30 ±10

(b) 2.07 2.04 1.88 605 61

±0.02 ±0.02 ±0.02 ±30 ±10

Table 5.2: Parameters of the spin Hamiltonians (Equation 5.1) for the Mn-related center in Mn-

activated LiBaF3; (a) considering the forbidden transition, (b) without the forbidden transitions.

In both cases the z-axis of the gand the Dtensor, respectively, is close to a <110>direction.

The PL-EPR experiments showed that there is at least one radiation-induced, lu-

minescent spin triplet system in Mn-doped LiBaF3after x-irradiation. This Mn-related

triplet system can be described by a FS tensor with its z-axis close to a <110>direction;

the slight deviation from the <110>direction has a value of θ= 8◦. It might be due to

a perturbation in the vicinity of luminescent spin triplet system. This perturbation has

not been analyzed yet.

Chapter 5. The Inverse Perovskite LiBaF3

6 The Persistent Phosphors MAl2O4

(M = Ca, Sr)

Presently most persistent phosphors are sulfides like ZnS, CdS, CaS, and SrS. More re-

cently the RE doped aluminates MAl2O4(M = Ca, Sr) have been developed to fill this

role [3, 5, 35–40] and these are the subject of investigations described in this chapter.

Both systems emit a purple/green luminescence originating from the optical activator

Eu. Even the single doped systems show phosphorescent behavior. However, the phos-

phorescent decay time can be significantly increased for SrAl2O4(SAO) by Dy co-doping

and for CaAl2O4(CAO) by Nd co-doping, see Figure 6.1. Despite many years of study

the complete mechanism of this persistent phosphorescence is still not understood though

several electron and hole trapping mechanisms have been proposed: RE traps are consid-

ered to have the dominant role in [35, 36, 39] for the recombination process, but intrinsic

electron and hole traps have been put forward as the source of the traps in [5, 38].

The first model for the recombination process was proposed by Matsuzawa et al. [4]

and is described as follows: after excitation of the Eu2+ ion a hole is being released and

then trapped close to the Dy3+ or Nd3+ dopant, respectively. Upon thermal activation

the hole is released from Dy4+ to the valence band and recombines at the Eu+ion,

which leads to the persistent luminescence. This mechanism was further supported by

thermoluminescence measurements and consequently has been widely adopted.

In a recently published paper [41] an alternate mechanism was proposed. A new

level scheme for SAO-ED is shown in Figure 6.2. Here it can be seen, that an excited

Eu ion promotes an electron to the conduction band, which is subsequently trapped

at a Dy ion. Thermally activated release from the Dy2+ ion leads to the persistent

luminescence. However, in this work low temperature recombination luminescence is

primarily investigated and discussed.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.1: Phosphorescence characteristics taken from [4]. The afterglow was measured

at RT after 10 min exposure to 200 lx of D65 light (standard light). A: SrAl2O4:Eu2+, B:

SrAl2O4:Eu2+,Dy3+, C: SrAl2O4:Eu2+,Nd3+, and D: commercially used ZnS:Cu,Co.

6.1 Sample Preparation

CAO and SAO single crystals were grown at the University of Georgia, using the laser

heated pedestal growth (LHPG) method [43]. This method allows the growth of good

quality single crystals in fiber form; the crystalline samples used in this study were

approximately 100 µm in diameter and 1 cm in length. LHPG samples grow in the

same orientation as the seed that is used to grow them; the orientation of the seeds we

used was not known hence the fibers used here were of undetermined orientation. For

the SAO single crystals a stoichiometric mixture of SrCO3and Al2O3was used and the

growth chamber was filled with a 5% H2-N2mixture. The single crystals were doped

with 0.07% Eu (SAO-E) or Dy (SAO-D), and with Eu and Dy (SAO-ED), respectively.

For the CAO single crystals a stoichiometric mixture of CaCO3and Al2O3was used.

They were also doped with 0.07% Eu (CAO-E) or Nd (CAO-N), and with Eu and Nd

(CAO-EN), respectively.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.2: Energy level scheme of SrAl2O4:Eu,Dy [42].

6.2 Photoluminescence and Recombination Luminescence

The PL spectra for CAO and SAO single crystals were recorded at RT and 20 K, re-

spectively. CAO-E shows an intense luminescence peaking at 437 nm at RT and at 20 K

an intense luminescence peaking at 437 nm. While the emission band recorded at RT

is rather broad (0.36 eV) with a small shoulder at about 465 nm, the one recorded at

20 K is narrower (0.21 eV) (see Figure 6.3). CAO-E co-doped with Nd has identical PL

spectra. At RT both samples show a strong RL signal at 437 nm (see Figure 6.4) after

excitation at 350 nm. This luminescence band can be allocated to the 5d-4ftransition of

a Eu2+ ion. The decay time of the co-doped sample is much longer than that of the one

doped only with Eu. Compared to the PL spectra the RL spectra of these samples are

slightly shifted to higher energies, but show basically the same bands. The shoulder at

465 nm could not be observed in the RL. At 20 K the situation is reversed: The sample

only doped with Eu alone has a stronger RL than the co-doped sample. The CAO-EN

crystal shows almost no RL signal.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.3: (a) Normalized PL spectra of CAO-E (dotted curve), CAO-EN (dashed) excited at

350 nm and normalized RL spectrum of CAO-EN (solid) after excitation at 350 nm, recorded at

RT. (b) Normalized PL spectra of CAO-E (dotted) and CAO-EN excited at 350 nm, recorded

at 20 K.

Figure 6.5 shows the PL and RL spectra of SAO-E and SAO-ED, recorded at RT and

20 K, respectively. Only one intense emission peaking at 518 nm can be observed at RT.

At 20 K a second emission band appears at 442 nm. The 518 nm emission is shifted to

524 nm. The RL spectra recorded at RT show the same peak positions as those seen

in the PL spectra. The RL decay of these emissions is comparable to the ones found in

CAO. At RT, both SAO-ED and SAO-E show a significant RL with the RL of SAO-ED

much more intense and much longer (see Figure 6.6). At 20 K there is almost no RL

detectable at 524 nm in SAO-E. But at 442 nm an RL intensity can be observed, which

is more intense than the RL of SAO-E at RT.

Nevertheless, it is possible to induce a recombination luminescence at low temperatures

by irradiating the samples with the integral light of a deuterium lamp for 30-60 minutes.

The RL spectrum of CAO (Figure 6.7 (a), dotted curve) shows luminescence bands at

about 310 and 800 nm. Samples additionally doped with Eu, Nd, or Eu and Nd show

further bands. CAO-E shows an additional band at 440 nm, CAO-N has an additional

band at 880 nm. Figure 6.7 (a), solid curve, shows the RL spectrum for the Eu- and

Nd-doped sample. The band at about 440 nm can be attributed to Eu2+ [36] whereas

the sharp peak at 880 nm can be attributed to Nd3+ [44].

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.4: Normalized RL decay of CAO-E (dotted curves) and CAO-EN (solid) after switch-

ing off the excitation at 350 nm. The decay curves were recorded at 437 nm at RT and 20 K,

respectively.

Figure 6.5: (a) Normalized PL (solid line) excited at 350 nm and RL (dotted) spectra of SAO-

ED after excitation at 350 nm, recorded at RT. (b) Normalized PL spectra of SAO-E (dotted)

and SAO-ED excited at 350 nm, recorded at 20 K.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.6: Normalized RL decay of SAO-E and SAO-ED after switching off the excitation at

350 nm. The decay curves were recorded at 518 nm at RT (solid lines) and at the wavelengths

marked in the figure at 20 K (dotted), respectively. The RL decay of SAO-ED is not shown.

SAO (Figure 6.7 (b), dotted curve) single crystals show a broad luminescence band

centered at about 360 nm and another one at 860 nm. The spectrum is similar to that

already observed for CAO. Several further bands appear in SAO-D crystals at about

485, 585, 680, and 770 nm. These bands can be attributed to Dy3+ [45]. In SAO-E an

additional double-structured band peaking at 435 and 520 nm is found. These bands

are due to Eu2+ [35]. The RL spectrum of SAO-ED (Figure 6.7 (b), solid curve) show

bands caused by Eu2+ and by Dy3+.

RL excitation spectra were recorded at 4.2 K using the magneto-optical spectrometer.

RL was excited with a deuterium lamp and a subsequent set of interference filters; the

luminescence was detected integrally. Figure 6.8 shows the spectra for differently doped

SAO. For SAO and SAO-D the best excitation wavelength is found to be at 200 nm,

whereas for SAO-E and SAO-ED, i.e. additional Eu-doping, it is shifted to about 250 nm.

A similar behavior was observed for CAO single crystals (spectra not shown).

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.7: RL spectra of (a) CAO (dotted curve) and CAO-EN (solid curve), (b) SAO (dotted

curve) and SAO-ED (solid curve), recorded at 20 K after 30 min UV-excitation at 20 K.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.8: RL excitation spectra, recorded at 4.2 K for SAO (full squares), SAO-E (full

circles), SAO-D (open squares), and SAO-ED (open circles) single crystals. The RL was excited

with a deuterium lamp and a subsequent interference filter at the wavelengths indicated; the

luminescence was detected integrally. The symbols represent the experimental data points; the

lines connecting the data points are a guide to the eye.

6.3 Electron Paramagnetic Resonance

Nakamura et al. [46] performed EPR experiments on CAO-ED and SAO-EN powders

in high frequency (180 GHz) spectrometers. They found three sites for the Eu2+-dopant

in SAO and only one site in CAO. The parameters are shown in Table 6.1. Angular

dependent EPR spectra of single crystalline SAO-E and CAO-E taken in a conventional

EPR X-band spectrometer are shown in Figures 6.9 and 6.10. Both crystals show several

Phosphor g-value D(10−4cm−1) E(10−4cm−1) Rel. intensity

SrAl2O4:Eu,Dy 1.989 1402.17 361.76 1.0

1123.59 343.82 1.0

1021.45 250.26 0.25

CaAl2O4:Eu,Nd 1.988 1791.00 46.33 1.0

Table 6.1: Parameters used for the calculation of high frequency EPR spectra of SrAl2O4:Eu,Dy

and CaAl2O4:Eu,Nd at 145 K [46].

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.9: Angular dependent EPR-spectra of CAO-E recorded at 10 K in X-band (9.35 GHz).

The rotation was performed in an arbitrary plane.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.10: Angular dependent EPR-spectra of SAO-E recorded at 10 K in X-band (9.35 GHz).

The rotation was performed in an arbitrary plane.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.11: Calculated EPR powder spectra of CAO-EN for a microwave frequency of (a)

180 GHz and (b) 93 GHz. The parameters can be found in Table 6.1.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

angular dependent lines which cannot yet be assigned. Phosphors doped only with Nd

or Dy or co-doped with Nd or Dy show no EPR signal. For improved resolution it is

essential to use high frequency EPR. Therefore, the RL-EPR measurements which are

described in the following section were not performed in K-band (24 GHz) but in W-

Band (93 GHz). A calculation of powder EPR spectra of CAO-EN at 93 GHz and at

180 GHz are shown in Figure 6.11.

6.4 Recombination Luminescence detected Electron

Paramagnetic Resonance

The RL-EPR spectra of SAO, SAO-D, SAO-E, and SAO-ED (Figure 6.12 lower part)

show a group of lines centered at about 3310 mT (g= 2.02) as well as a structureless

resonance line at about 3390 mT (g= 1.97 for a microwave frequency of 93.7 GHz). Note,

that the high-field side of the 3365 mT line is affected by long spin-lattice relaxation

times. The resonance at 3390 mT shows no angular dependence, whereas the line group

at 3310 mT shows significant changes in structure as a function of orientation with

respect to the magnetic field (see Figure 6.14).

The RL-EPR spectra for CAO, CAO-E, and CAO-EN show similar resonances (Fig-

ure 6.12 upper graphs). The RL of the CAO-N sample was too weak for RL-EPR

measurements. The group of lines at about 3310 mT is comparable to those observed in

SAO and is angular dependent (Figure 6.13). The structureless (isotropic) line is shifted

to 3365 mT (g= 1.99).

Spectral dependencies of the RL-EPR spectra are shown in Figure 6.15 for SAO and

CAO. The isotropic line shows no spectral dependence, but the low field group shows

significant changes for SAO as well as for CAO.

6.5 Discussion

CAO-E and CAO-EN shows an emission band at 437 nm, which is attributed to a 5d−4f

transition of a Eu2+ ion. A similar luminescence can be observed for SAO-E and SAO-

ED. At RT a single peak can be observed at 518 nm, which shifts at 20 K to 524 nm. It

is also due to the Eu2+ dopant, which substitutes for a Sr2+ ion. The additional band

appearing at 20 K cannot be attributed to this Eu ion. It is proposed to be due to

interstitial Eu2+ ions [36].

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.12: The upper graphs show the RL-EPR spectra of (a) CAO, (b) CAO-E, and (c)

CAO-EN. The RL-EPR of CAO-N was too weak to be measured. In the lower graphs the RL-

EPR spectra of (a) SAO, (b) SAO-E, (c) SAO-D, and (d) SAO-ED are depicted. The RL-EPR

was detected in the integral RL at 1.5 K in the W-band (93.7 GHz) after UV-excitation with

the light of a deuterium lamp at 4.2 K. The magnetic field dependent background is subtracted.

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.13: Angular dependent RL-EPR spectra of CAO-EN. The RL-EPR was recorded in

the integral RL at 1.5 K in W-band (93 GHz) after UV-excitation with the light of a deuterium

lamp at 4.2 K. The magnetic field dependent background is subtracted; the rotation of the

magnetic field is performed in an arbitrary plane.

The RL which was induced by prior UV-exposure at 4.2 K quenches in high magnetic

fields, but it can be increased significantly by applying microwave radiation (93.7 GHz)

yielding resonance lines for appropriate magnetic fields. We attribute the RL to a recom-

bination of distant donors and acceptors which were generated by the UV-excitation.

The RL spectra show that either the donor-acceptor recombination energy is emitted

directly (undoped samples) or transferred to the rare-earth activators (doped samples).

We do observe luminescence from intrinsic centers as well as from RE dopant ions.

Europium doping leads to a perturbation of the intrinsic defect states; the maxima of

the corresponding RL excitation spectra are shifted to longer wavelengths (Figure 6.8).

The gvalues of the RL-EPR line at 1.97 (SAO) and 1.99 (CAO), respectively, are less

than two, so we can attribute them to donors [47]. The line groups centered at 2.02

are assigned to acceptors [47]. The angular dependence of the donor in CAO and SAO

shows that we are dealing with an isotropic center here, whereas the line group of the

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.14: Angular dependent RL-EPR spectra of SAO-E. The RL-EPR was recorded in the

integral RL at 1.5 K in W-band (93 GHz) after UV-excitation with the light of a deuterium lamp

at 4.2 K. The magnetic field dependent background is subtracted; the rotation of the magnetic

field is performed in an arbitrary plane.

acceptor is clearly angular dependent. Since the gvalues of the donors do not change

upon doping, these traps are intrinsic. The angular dependence of the line group at 2.02

is caused by different center orientations and by a superimposition of resonances lines

of at least two acceptors. We cannot say how the co-doping of Nd or Dy affects the

defect structure of the acceptor since the different patterns of the line group at 2.02 (see

Figures 6.12 and 6.15) are also affected by the different orientation of the magnetic field.

Note that all RL-EPR measurements were performed in an arbitrary orientation of the

magnetic field.

From the RL spectra and the spectral dependent RL-EPR we assume that there are

two different types of intrinsic luminescences in both aluminates. Because the lines at

1.97 and 1.99, respectively, do not change in the spectral dependent RL-EPR it seems

that in both RL processes the same donors are involved. The difference between them

is the type of acceptors. In doped aluminates parts of the recombination energy is

Chapter 6. The Persistent Phosphors MAl2O4(M = Ca, Sr)

Figure 6.15: Spectral dependent RL-EPR spectra of SAO and CAO. (a) SAO, recorded with UV

band-pass filter DUG 11; (b) SAO, recorded with an edge filter KV 550 for the same orientation

of the magnetic field as in (a); (c) CAO, recorded with UV band-pass filter DUG 11; (d) CAO,

recorded with an edge filter KV 550 for the same orientation of the magnetic field as in (c).

The RL-EPR was recorded in the integral RL at 1.5 K in the W-band (93.7 GHz) after UV-

excitation with the light of a deuterium lamp at 4.2 K. The magnetic field dependent background

was subtracted.

transferred to the dopant. Spectral dependent RL-EPR in doped aluminates (not shown

here) showed a comparable behavior to the undoped samples. This indicates that parts of

the intrinsic recombination energy is transferred to the dopant. This is also corroborated

by the RL spectra.

7 Conclusion

The PL emission of Mn-doped RbCdF3at 560 nm could be attributed to the Mn2+

dopant, having the corresponding excitation peaks in the UV and VIS spectral region.

This emission band can be increased significantly upon x-irradiation, where an additional

excitation peak at about 300 nm appears; this excitation band can be bleached optically.

The EPR measurements showed that the number of Mn2+ ions is reduced by x-irradiation

and can be restored by optical bleaching. This agrees with the model that optically

stimulable Mn3+-F complexes are formed by x-irradiation.

Unfortunately RbCdF3shows after x-irradiation no PL-EPR resonance lines. In prin-

ciple it would be possible to analyze the vicinity of the Mn-dopant with EXFAS measure-

ments, but here all Mn ions in the compound are detected not only the optically active

ones. An optical detection of EXFAS would lead to a selective detection of the Mn ions;

i.e. only the optically active ions will be monitored. This technique is usually called

x-ray excited optical luminescence detected x-ray finestructure absorption spectroscopy

(XEOL-XAFS).

The investigations on Mn-activated LiBaF3yielded that there are several luminescent

processes. On the one hand non-irradiated LiBaF3shows a typical Mn2+ luminescence

and its typical excitation bands in the UV and in the VIS spectral region. On the other

hand there is a further luminescence band in the blue region. It turned out to be due

to a perturbed oxygen vacancy complex. An oxygen impurity substitutes for a fluorine

ion with a F−vacancy in a <110>direction. This complex is probably perturbed by a

Mn2+ ion on a Ba2+ site lying about 35◦away from the complex. After x-irradiation the

whole luminescence is covered by a new emission band peaking at 600 nm also showing

PL-EPR resonance lines. From the angular dependence it can be extracted that this

defect is almost parallel to a <110>direction and it is significantly different from the

oxygen vacancy complex.

There are several mechanisms in LiBaF3, which are not understood yet. The new ex-

citation band appearing at low temperatures has not been investigated in detail yet.

Chapter 7. Conclusion

This band is probably due to a Mn-vacancy complex, which is already created during

the growth process. Quenching of a sample slightly below the melting temperature of

LiBaF3should lead to a destruction of this center and the new excitation band should

also disappear. A complete analysis of the PL-EPR spectra recorded after x-irradiation

would yield new information on the defect structure responsible for the luminescence at

600 nm.

The persistent phosphors CAO and SAO show a very strong persistent luminescence

in the visible region originating from a 5d−4ftransition of the Eu2+ activator. They

can be excited in the near UV region. In this work it could be shown that there is

also an RL from CAO and SAO at low temperatures. This luminescence was attributed

to a distant donor acceptor recombination, which were created by UV excitation at

low temperatures. The energy of the recombination process is either transferred to the

rare earth activators (doped samples) or emitted directly (undoped samples). There is

RL coming from intrinsic centers as well as from rare-earth the dopant ions. It was

shown that there is only one type of donors, which could be attributed to intrinsic traps,

whereas there are at least two acceptors. Europium doping leads to a perturbation of

the intrinsic defect states, i.e. the excitation spectra are shifted to higher wavelengths.

A BaCl2EXAFS

The stable phase of BaCl2is the orthorhombic PbCl2structure, space group Pnma(D16

2h)

with lattice parameters a= 7.865 ˚

A, b= 4.731 ˚

A, and b= 9.421 ˚

A [48]. The crystal

structures of different barium halides are shown in Figures A.1. In Figure A.2 the

lattice structure of BaCl2, BaBr2, and BaI2and their corresponding lattice parameters

are depicted. Due to the same charge state of Ba2+ and Eu2+ a Eu2+ dopant should

substitute a barium ion.

The x-ray fluorescence spectrum of 10% Eu-doped BaCl2is shown in Figure A.3. The

most significant peaks at 4469 and 4857 eV can be attributed to the Lα1,2and Lβ3,4

lines of barium. The Lα1,2line of Eu can be found at 5835 eV. The EXAFS spectrum

of the Eu LIII-edge in BaCl2can be seen in Figure A.4, where the white line at 6980 eV

is due to Eu3+. Since in the crystal growth process only a small amount of Eu2+ was

incorporated in the sample the white line of Eu2+ is only a small shoulder at about

6972 eV. The relative intensity of the two white lines is an indication of the Eu2+/Eu3+

ratio. A detailed analysis to obtain quantitative information about the ratio has not

been done yet.

Appendix A. BaCl2EXAFS

Figure A.1: Lattice structure of orthorhombic BaX2projected on the ab,ac, and bc planes,

small white and large gray circles denoting Ba2+ and X−(X = Cl, Br, I) ions, respectively. The

ions discerned by hatched circles lie in a different ab-type mirror plane than the non-hatched

ones. Two types of magnetically non-equivalent Ba sites are labeled 1 and 2. The radii are

scaled to 30% of their original values.

Appendix A. BaCl2EXAFS

Figure A.2: Lattice structure of orthorhombic (a) BaCl2, (b) BaBr2, and (c) BaI2projected

on the ab plane, small white and large gray circles denoting Ba2+ and X−ions, respectively. The

ions discerned by hatched circles lie in a different ab-type mirror plane than the non-hatched

ones. Two types of Ba sites are labeled 1 and 2. The distances indicated are given in ˚

A. The

radii are scaled to 30% of their original values. The principal directions xand yof the crystal

field are indicated.

Appendix A. BaCl2EXAFS

Figure A.3: Normalized X-ray fluorescence spectrum of 10% Eu-doped BaCl2with an excitation

energy of 7080 eV.

Figure A.4: Normalized EXAFS spectrum of 10% Eu-doped BaCl2. The spectrum was recorded

in fluorescence mode. The inset shows the Eu absorption edge in more detail.

B Miscellaneous

B.1 Euler Tensor

The definition of the matrix Eis:

E=





Exx Exy Exz

Exy Eyy Eyz

Exz Eyz Ezz







With its elements:

Exx = cos α·cos β·cos γ−sin α·sin γ

Exy = cos β·cos γ·sin α+ sin γ·cos α

Exz =−cos γ·sin β

Exy =−sin γ·cos α·cos β−cos γ·sin α

Eyy =−sin γ·cos β·sin α+ cos γ·cos α

Eyz = sin β·sin γ

Ezx = sin β·cos α

Ezy = sin β·sin α

Ezz = cos β

These definitions are the same as used in the VISUAL EPR program. To transform a

matrix in a new axis system the formula Mnew =E·M·ETis used. ETis the transpose

of the Euler matrix.

A rotation of the Euler angles of the center orientation is described by Enew =Erot ·

Eold. The new angles are given by:

βnew = arccos (Enew

zz )

Appendix B. Miscellaneous

αnew = arcsin Enew

sin βnew 

γnew = arcsin Enew

sin βnew 

B.2 Fine Structure Tensor

The definition of the fine structure tensor Dis

D=





Dxx Dxy Dxz

Dxy Dyy Dyz

Dxz Dyz Dzz







In the “Grachev” program the elements are defined by:

Dxx = 1/3·(−b0

2+b2

Dyy = 1/3·(−b0

2−b2

Dzz = 2/3·b0

Dxy = 1/3·c2

Dxz = 1/6·b1

Dyz = 1/6·c1

With the inversion:

2= 3/2·Dzz

2= 3 ·Dxz

2= 6 ·Dyz

2= 3 ·Dxx −b0

2= 3 ·Dxy

A further description of the main axis elements is given by:

Dxx =−1/3·D+E

Dyy =−1/3·D−E

Dzz = 2/3·D

Appendix B. Miscellaneous

This yields:

2=D

2= 3 ·E

B.3 EPR Data

Host gA(10−4cm−1) a(10−4cm−1) Reference

LiBaF32.0014 ±0.0004 −88.3±0.1 5.5±0.1 [26]

RbCdF31.99 ±0.01 −90.4±0.3 14 ±0.3 [18]

RbMgF32.001 88.7 [49]

Table B.1: Hyperfine structure parameter for Mn2+ in different host lattices.

Appendix B. Miscellaneous

Bibliography

[1] W. A. Sibley and N. Koumvakalis. Phys. Rev. B 14(1), 35 (1976).

[2] M. A. Losada, R. Alcala, P. J. Alonso, R. Cases, and V. M. Orera. Radiation Effects

Express 2, 39–44 (1988).

[3] Y. Murayama, N. Takeuchi, Y. Aoki, and T. Matsuzawa. Technical report, US

Patent 5,424,006, (1995).

[4] T. Matsuzawa, N. Aoki, Y. Takeuchi, and Y. Murayama. J. Electrochem. Soc. 143,

2670 (1996).

[5] T. Aitasalo, J. H¨ols¨a, H. Jungner, M. Lastusaari, and J. Niittykoski. J. Lumin.

94-95, 59–63 (2001).

[6] G. Blasse and B. C. Grabmaier. Luminescent Materials. Springer Verlag, Berlin

Heidelberg, (1994).

[7] B. Henderson and G. F. Imbusch. Optical Spectroscopy of Inorganic Solids. Claren-

don Press, Oxford, (1989).

[8] F. Rodriquez and M. Moreno. J. Chem. Phys. 84(2), 692–697 (1986).

[9] P. G. Baranov, V. A. Vetrov, and N. G. Romanov. Sov. Phys. Solid State 25(5),

784–788 (1983).

[10] Boon K. Teo. EXAFS: Basic Principles and Data Analysis. Springer-Verlag, (1986).

[11] J. Dumschat. Hochaufl¨osende R¨ontgenabsorptionsspektroskopie an vierwertigen

Seltene-Erdsystemen. PhD thesis, University of Paderborn, (2000).

[12] http://physics.nist.gov/PhysRefData/FFast/html/form.html (Retrieved 2007).

Bibliography

[13] T. Rewaj, M. Krupski, J. Kuriata, and J. Y. Buzar´e. J. Phys.: Condens. Matter

4(49), 9909–9918 (1992).

[14] M. Rousseau, J. Y. Gesland, J. Julliard, J. Nouet, J. Zarembowitch, and A. Zarem-

bowitch. Phys. Rev. B 12(4), 1579–1590 (1975).

[15] P. Studzinski and J. M. Spaeth. Radiation Effects 73, 207–213 (1983).

[16] D’Ans-Lax. Taschenbuch f¨ur Chemiker und Physiker. Springer-Verlag, (1970).

[17] J. Ferguson. Australian Journal of Chemistry 21, 307–321 (1968).

[18] P. Studzinski. Paramagnetische Ionen zur Untersuchung struktureller

Phasen¨uberg¨ange - Eine Endor Untersuchung. PhD thesis, Universit¨at Paderborn,

(1985).

[19] C. Dotzler, G. V. M. Williams, A. Edgar, S. Schweizer, B. Henke, J.-M. Spaeth,

A. Bittar, J. Hamlin, and C. Dunford. J. Appl. Phys. 100, 033102 (2006).

[20] http://cars9.uchicago.edu/∼newville/ModelLib/search.htm (Retrieved 2007).

[21] http://cars9.uchicago.edu/∼ravel/software/aboutathena.html (Retrieved 2007).

[22] M. C. Marco de Lucas, F. Rodriguez, M. Moreno, and A. Tressaud. J. Phys.:

Condens. Matter 6, 6353–6366 (1994).

[23] R. L. Lambe and D. Schroeer. Phys. Rev. Lett. 36(1), 45 (1976).

[24] J. Selling. Barium Halide Nanocrystals in Fluorozirconate based Glass Ceramics for

Scintillation Application. PhD thesis, University of Paderborn, (2007).

[25] V. M¨ollmann, J. Selling, B. Henke, S. Schweizer, P. Keil, V. Lavin, and G. Wort-

mann. Annual HASYLAB Report 2007 submitted (2007).

[26] T. Yosida, H. Aoki, H. Takeuchi, M. Arakawa, and K. Horai. J. Phys. Soc. Jpn.

46(5), 1661–1662 (1979).

[27] W. L. W. Ludekens and A. J. E. Welch. Acta Crystallographica 5, 841 (1952).

[28] T. Yosida. J. Phys. Soc. Jpn. 49(1), 127–135 (1980).

[29] N. Shiran and V. Voronova. J. Lumin. 87-89, 561–563 (2000).

Bibliography

[30] A. Abragam and B. Bleany. Electron Paramagnetic Resonance of Transition Ions.

Dover Publications, New York, (1986).

[31] F. K. Koschnick, Th. Hangleiter, K. S. Song, and J.-M. Spaeth. J. Phys.: Condens.

Matter 7, 6925–6937 (1995).

[32] S. Schweizer, U. Rogulis, K. S. Song, and J.-M. Spaeth. J. Phys.: Condens. Matter

12, 6237–6243 (2000).

[33] U. Rogulis, S. Schweizer, and J.-M. Spaeth. J. Phys.: Condens. Matter 14(28),

6949–6956 (2002).

[34] B. Henke, M. Secu, U. Rogulis, S. Schweizer, and J.-M. Spaeth. Phys. Status Solidi

C2(1), 380–383 (2005).

[35] W. Jia, H. Yuan, L. Lu, H. Liu, and W. M. Yen. J. Lumin. 76&77, 424–428 (1998).

[36] W. Jia, H. Yuan, L. Lu, H. Liu, and W. M. Yen. J. Cryst. Growth 200, 179–184

(1999).

[37] D. Jia, S. Meltzer, W. M. Yen, W. Jia, and X. Wang. Appl. Phys. Lett. 80(9),

1535–1537 (2002).

[38] T. Aitasalo, P. Deren, J. H¨ols¨a, H. Jungner, J.-C. Krupa, M. Lastusaari, J. Leg-

endziewicz, J. Niittykoski, and W. Strek. Journal of Solid State Chemistry 171,

114–122 (2003).

[39] Y. Lin, Z. Tang, Z. Zhang, and C. Nan. Journal of the European Ceramic Society

23, 175–178 (2003).

[40] A. Nag and T. R. N. Kutty. Materials Research Bulletin 39, 331–342 (2004).

[41] P. Dorenbos. Journal of the Electrochemical Society 152(7), H107–H110 (2005).

[42] P. Dorenbos. Phys. Status Solidi B 242, R7–R9 (2005).

[43] W. M. Yen. Preparation of Single Crystal Fibers in: Insulating Materials for Op-

toelectronics. World Scientific, Singapore, (1995).

[44] A. A. Kaminskii. Laser Crystal, volume 14 of Optical Series. Springer-Verlag,

Berlin, 2nd edition, (1981).

[45] H. M. Crosswhite and G. H. Dieke. J. Chem. Phys. 35(5), 1535–1548 (1961).

Bibliography

[46] T. Nakamura, T. Matsuzawa, C. C. Rowlands, V. Beltr´an-L´opez, G. M. Smith, and

P. C. Riedi. J. Chem. Soc., Faraday Trans. 94, 3009–3012 (1998).

[47] J.-M. Spaeth and H. Overhof. Point Defects in Semiconductors and Insulators.

Springer, Berlin, (2003).

[48] E. B. Brackett, T. E. Brackett, and R. L. Sass. J. Phys. Chem. 67(10), 2132 (1963).

[49] J. M. Dance and N. Kerkouri. C. R. Acad. Sc. Paris 289(C), 425–428 (1979).

List of Publications

[1] B. Henke, M. Secu, U. Rogulis, S. Schweizer, and J.-M. Spaeth.

Optical and magneto-optical studies of Mn-activated LiBaF3

phys. stat. sol. c 2(1), 380–383 (2005).

[2] S. Schweizer, B. Henke, U. Rogulis, and W. M. Yen.

Magneto-optical resonance investigations on Eu and Nd/Dy-codoped CaAl2O4and

SrAl2O4single crystals

in Science and Technology of Dielectrics in Emerging Fields and Persistent Phos-

phors (PV 2005-13), edited by K. W¨orhoff, D. Misra, P. Mascher, K. Sundaram,

W. M. Yen, and J. Capobianco (The Electrochemical Society, Quebec City, Canada,

2005), Part II, p. 191.

[3] C. Dotzler, G. V. M. Williams, A. Edgar, S. Schweizer, B. Henke, J.-M. Spaeth,

A. Bittar, J. Hamlin, and C. Dunford.

The effect of x-ray, γ-ray, and UV radiations on the optical properties of RbCdF3:Mn2+

J. Appl. Phys. 100, 033102 (2006).

[4] B. Henke, U. Rogulis, and S. Schweizer.

Luminescent oxygen-vacancy complex in Mn-doped LiBaF3investigated by opti-

cally detected magnetic resonance

Proceedings of the Eighth International Conference on Inorganic Scintillators and

their Use in Scientific and Industrial Applications, Alushta, Crimea, Ukraine,

September 19-23, 2005, ed. by A. Gektin and B. Grinyov (National Academy

of Sciences of Ukraine, Ukraine - Kharkov, 2006), 70–73 (2006).

[5] B. Henke, U. Rogulis, and S. Schweizer.

Optically detected magnetic resonance investigation of a luminescent oxygen-vacancy

complex in Mn-doped LiBaF3

J. Phys.: Condens. Matter 18(5), 1577–1583 (2006).

Bibliography

[6] J. A. Johnson, S. Schweizer, B. Henke, G. Chen, J. Woodford, P. J. Newman, and

D. R. MacFarlane.

Eu-activated fluorochlorozirconate glass-ceramic scintillators

J. Appl. Phys. 100, 034701 (2006).

[7] S. Schweizer, B. Henke, S. K¨oneke, J. A. Johnson, G. Chen, and J. Woodford.

Energy-dependent scintillation intensity of fluorozirconate-based glass-ceramic x-

ray detectors

Proc. of SPIE 6142, 61422Y–1 (2006).

[8] G. V. M. Williams, S. Schweizer, B. Henke, C. Dunford, and A. Edgar.

X-ray and UV induced photo-luminescence from RbCdF3:Mn2+

Curr. Appl. Phys. 6, 351–354 (2006).

[9] S. Schweizer, B. Henke, U. Rogulis, and W. M. Yen.

Recombination processes in undoped and rare-earth doped MAl2O4(M=Ca, Sr)

persistent phosphors investigated by optically detected magnetic resonance

Appl. Phys. Lett. 90, 051902 (2007).

[10] B. Henke, S. Schweizer, and U. Rogulis.

Optical and electron paramagnetic resonance studies on radiation defects in Mn-

activated RbCdF3

phys. stat. sol. c 4, No. 3, 677-682 (2007).

[11] S. Schweizer, B. Henke, U. Rogulis, and W. M. Yen.

Recombination processes in rare-earth doped MAl2O4(M = Ca, Sr) persistent phos-

phors investigated by optically-detected magnetic resonance

phys. stat. sol. a 204, 677-682 (2007).

[12] B. Henke, S. Schweizer, J. A. Johnson, and D. T. Keane.

Zr and Ba edge phenomena in the scintillation intensity of fluorozirconate-based

glass-ceramic X-ray detectors

J. Sychrotron Rad. 14, 252-256 (2007).

[13] B. Henke, U. Rogulis, and S. Schweizer.

Structure sensitive investigations on luminescence centres in Mn-activated LiBaF3

dosimeters

Radiat. Meas., (submitted September 2007).

Bibliography

In Preparation

[1] B. Henke, U. Rogulis, Mike and S. Schweizer.

Structure sensitive investigations of a x-ray induced defect in Mn-activated LiBaF3

J. Phys.: Condens. Matter.

[2] B. Henke, J. A. Johnson, and S. Schweizer.

EXAFS investigations on the Mn-doped fluoroperovskite RbCdF3

J. Phys.: Condens. Matter.

[3] B. Ahrens, B. Henke, J. A. Johnson, P. T. Miclea and S. Schweizer.

Enhanced up-converted fluorescence in fluorozirconate based glass ceramics for high

efficiency solar cells

Proc. of SPIE.

Bibliography

Danksagung

Abschließend m¨ochte ich mich bei den Menschen bedanken, die mir w¨ahrend meiner

Promotion zur Seite gestanden haben:

•PD Dr. Stefan Schweizer f¨ur die M¨oglichkeit in seiner Arbeitsgruppe zu

promovieren und sein stets f¨orderndes Engagement.

•Prof. Dr. G. Wortmann f¨ur viele wertvolle Diskussionen.

•Meinen beiden Kollegen Julia Selling und Bernd Ahrens f¨ur die Hilfe in allen

m¨oglichen Lebenslagen.

•Prof. Dr. Uldis Rogulis f¨ur die nicht immer unproblematische Einf¨uhrung in

die ODMR Spektroskopie.

•Dr. Jacqueline A. Johnson, die meine offensive Auslegung der englischen

Grammatik in die richtigen Bahnen gelenkt hat.

•Dr. Mihail Secu f¨ur viele wissenschaftliche Anregungen.

•Den Mitarbeitern des Kristallzuchtlabors D. Niggemeier und R. Winterberg.

•J. Pauli f¨ur die problemlose Versorgung mit fl¨ussigem Helium.

Bibliography

•Den Mitarbeitern der APS Denis T. Keane und Qing Ma f¨ur die freundliche

und kompetente Hilfe bei den Experimenten an der APS1sowie Dariusz Zajac

als Mitarbeiter des HASYLABs.

•Allen namentlich nicht erw¨ahnten Institutsmitgliedern f¨ur die gute Zusammenar-

beit und das hervorragende Arbeitsklima.

Auch danke ich meiner Freundin Sarah Schuller und meinen Eltern, die mich

zu jeder Zeit nach besten Kr¨aften unterst¨utzt haben.

1This work was performed at the DuPont-Northwestern-Dow Collaborative Access Team (DND-CAT)

Synchrotron Research Center located at Sector 5 of the Advanced Photon Source. DND-CAT is

supported by the E.I. DuPont de Nemours & Co., The Dow Chemical Company, the U.S. National

Science Foundation through Grant DMR-9304725 and the State of Illinois through the Department

of Commerce and the Board of Higher Education Grant IBHE HECA NWU 96.

Use of the Advanced Photon Source was supported by the U. S. Department of Energy, Office of

Science, Office of Basic Energy Sciences, under Contract No. W-31-109-Eng-38.