Refine
Has Fulltext
- yes (22) (remove)
Is part of the Bibliography
- yes (22) (remove)
Year of publication
- 2023 (22) (remove)
Document Type
- Doctoral Thesis (16)
- Journal article (6)
Keywords
- Topologischer Isolator (6)
- ARPES (3)
- Computertomographie (2)
- Fotovoltaik (2)
- Nanometerbereich (2)
- Perovskite (2)
- Perowskit (2)
- Rasterelektronenmikroskopie (2)
- Rastertunnelmikroskop (2)
- Rastertunnelmikroskopie (2)
Institute
- Physikalisches Institut (22) (remove)
Rotationsdriftspektroskopie
(2023)
Die wachsende Verfügbarkeit von magnetischen Nanopartikeln (MNPs) mit funktionalisierten
Partikeloberflächen eröffnet weitreichende Möglichkeiten für chemische, biologische und klinische
Analysemethoden. Durch Funktionalisierung kann eine gezielte Interaktion mit Molekülen bewirkt
werden, die im Allgemeinen auch die Beweglichkeit der MNPs verändern. Methoden zur
Charakterisierung von MNPs wie bspw. AC-Suszeptometrie, Magnetorelaxometrie (MRX) oder
Magnetic Particle Spectroscopy (MPS) können diese Änderung der Beweglichkeit bei MNPs
messen, wenn es sich um MNPs handelt, deren magnetisches Moment im Partikel fixiert ist. Damit
ist mit funktionalisierten MNPs indirekt auch die spezifische Messung von Molekülkonzentrationen
möglich. MNPs können zudem in biokompatibler Form hergestellt werden und sind dadurch auch
als in-vivo Marker einsetzbar. Das 2005 das erste Mal veröffentlichte Magnetic Particle Imaging
(MPI) kann als ein mittels Gradientenfeldern um die räumliche Kodierung erweitertes MPS
betrachtet werden. Dank biokompatibler MNPs handelt es sich dabei um eine in-vivo-taugliche,
nicht-invasive Bildgebungsmethode. Mit funktionalisierten MNPs als Marker ist damit im Prinzip
auch molekulare Bildgebung möglich, die durch Detektion der beteiligten Moleküle (Biomarker)
Stoffwechselprozesse räumlich abbilden kann. Im Vergleich zur Bildgebung von Gewebe- und
Knochenstrukturen lassen sich die diagnostischen Möglichkeiten durch molekulare Bildgebung
erheblich erweitern.
Rotationsdriftspektroskopie (Rotational Drift Spectroscopy, RDS) ist eine in dieser Arbeit
entwickelte Methode für die induktive Messung der Beweglichkeit von MNPs in flüssiger
Suspension. Es verwendet die Rotationsdrift von MNPs in rotierenden magnetischen Feldern als
Grundlage und bietet das Potential die Änderungen der Beweglichkeit von MNPs mit einer
Empfindlichkeit messen zu können, welche potentiell um mehrere Größenordnungen höher sein
kann als mit den oben erwähnten Verfahren. Die vorliegende Arbeit konzentriert sich auf die
Verwendbarkeit dieses Effekts als Spektroskopiemethode. Die Eigenschaften des RDS-Signals sind
jedoch auch als Grundlage für räumliche Kodierung vielversprechend. In weiterführenden Projekten
soll daher auch die Entwicklung von Rotationsdriftbildgebung (Rotating Drift Imaging, RDI) als ein
nicht-invasives Verfahren für molekulare Bildgebung angestrebt werden.
Der Grundgedanke von RDS entlehnt sich aus einem in 2006 veröffentlichten Sensordesign
basierend auf magnetische Mikropartikel in einem schwachen rotierenden Magnetfeld. Das
rotierende Magnetfeld ist dabei so schwach gewählt, dass sich das Partikel aufgrund der viskosen
Reibung nicht mehr synchron mit dem externen Feld drehen kann. Die Frequenz der resultierenden
asynchronen Rotationsdrift liegt unterhalb der Frequenz des externen Rotationsfelds und ist
Abhängig von der viskosen Reibung. Aufgrund dieser Abhängigkeit können Änderungen im
Reibungskoeffizienten des Partikels über Änderungen in der Rotationsdriftfrequenz gemessen
werden.
RDS zielt darauf ab, diese Rotationsdrift bei suspendierten MNPs über deren
makroskopische Magnetisierung messen zu können. Damit wird u.a. auch die nicht-invasive
Messung von MNPs innerhalb opaker biologischer Proben möglich. MNP-Suspensionen sind
großzahlige Nanopartikel-ensembles und können nicht wie ein einzelnes Mikropartikel gemessen
werden. Für die induktive Messung ist vor dem Start eine Ausrichtung aller magnetischen Momente
nötig, da sich deren makroskopische Magnetisierung andernfalls zu Null addiert. Aufgrund von
Rotationsdiffusion bleibt diese Ausrichtung nur eine begrenzte Zeit bestehen, so dass auch die
eigentliche Messung des RDS-Signals nur eine begrenzte Zeit möglich ist. Diese Ausrichtung wurde
in den ersten Experimenten durch einen kurzen Magnetfeldpuls erzeugt. In der Empfangsspule ist
die Induktion durch das Rotationsfeld typischer Weise um mehrere Größenordnungen höher als das
zu erwartende Signal und muss durch einen Tiefpass unterdrückt werden. In diesem Tiefpassfilter
ruft jedoch die Einkopplung des Anfangspulses eine Pulsantwort hervor, die ebenso mehrere
Größenordnungen des zu erwartenden Signals betragen kann und ähnlich langsam wie typische
Signale abklingt. Die Unterdrückung dieser Pulsantwort stellte in den ersten Experimenten die
größte Hürde da. Der erste Aufbau hatte eine Relaisschaltung zur Pulsunterdrückung und resultierte
in einer Totzeit von 3 ms zwischen Anfangspuls und Start der Messung. Aufgrund dieser Totzeit
waren die ersten Messungen auf größere Agglomerate und Sedimente von MNPs beschränkt, da nur
in diesem Fall eine hinreichend lange Zerfallsdauer der Probenmagnetisierung vorlag. Das
Verhalten derartiger Partikelsysteme ist jedoch aufgrund von mechanischer und magnetischer
Interpartikelwechselwirkung vergleichsweise komplex und theoretisch schwer modellierbar. Das
primäre Zielsystem für RDS hingegen, Eindomänenpartikel mit im Partikel fixierter Magnetisierung
und Punktsymmetrie bzgl. des Reibungstensors, erlaubt die Aufstellung einer parametrisierten
Funktion für den Signalverlauf. Es ermöglicht somit aufgrund der besseren Berechenbarkeit eine
solidere Auswertung des RDS-Signals. Um Eindomänenpartikel in wässriger Suspension mit
typischen Partikeldurchmessern um 100 nm messen zu können ist eine Verkürzung der Totzeit auf
mindestens 1/10 erforderlich.
Prinzipiell kann diese Problematik durch die Verwendung schneller Halbleiterschalter in
Verbindung mit einer präzise abstimmbaren induktiven Entkopplung des Spulensystems gemindert
werden. Simulationen des RDS-Signals für verschiedene RDS-Sequenzen zeigen jedoch noch zwei
weitere Möglichkeiten auf, die ohne aufwändigen Eingriffe in der Hardware auskommen. Zum
einen kann durch orthogonales Frequenzmischen mit geeignetem Frequenz- und Phasenverhältnis
eine Ausrichtung der magnetischen Momente bewirkt werden. Da die benötigten Frequenzen
vollständig im Sperrband des Tiefpassfilters liegen können, lässt sich damit die Pulsantwort bei
hinreichend „weichem“ Umschalten zwischen der Polarisierungssequenz und der RDS-Sequenz
vollständig vermeiden. Darüber hinaus zeigt sich, dass es bei Anwesenheit eines schwachen
Offsetfelds (< 10 % der Rotationsfeldamplitude) zu einer Ausrichtung der magnetischen Momente
kommt, wenn das magnetische Rotationsfeld seine Richtung ändert und diese Änderung nicht
abrupt erfolgt, sondern das Rotationsfeld übergangsweise in ein linear oszillierendes Feld übergeht.
Hingegen wird die Wirkung des Offsetfelds durch das Rotationsfeld vor und nach dem Wechsel
nahezu vollständig neutralisiert, so dass damit das Störsignale generierende Schalten eines
Offsetfelds ersetzt werden kann. Es ist auf diese Weise nicht möglich, Echosequenzen zu erzeugen,
da hier bei der für Echosequenzen benötigten Richtungsumkehr des Rotationsfelds die zuvor
aufgeprägte Phasenverteilung durch das Offsetfeld zerstört wird und somit anstelle einer
Signalechogenerierung eine neue RDS-Messung gestartet wird. Obwohl es Echosequenzen mit
Anfangspuls erlauben, mehr MNP Parameter zu messen, bietet dieser Ansatz dennoch
entscheidende Vorteile. So ergibt sich eine massive Vereinfachung der Hardware und es sind bei
gleicher Rotationsfrequenz deutlich höhere Wiederholraten möglich.
Die Vermeidung von Schaltvorgängen durch die Verwendung von Offsetfeldern ermöglicht
es, mit dem ursprünglichem Aufbau auch Partikelsysteme zu untersuchen, deren Relaxationszeit
weit unter 3 ms liegt. Hier zeigt sich, dass sich für unterschiedliche Partikelsysteme teils sehr
charakteristische Signalmuster ergeben. Diese lassen sich grob in drei Kategorien einteilen. Die
erste Kategorie sind suspendierte Eindomänenpartikel mit einer nicht vernachlässigbaren
Relaxationszeit. Hier handelt es sich um das bevorzugte Zielsystem für RDS, das durch die
Langevin-Gleichung beschrieben werden kann. Die zweite Kategorie sind Partikelsysteme, bei
denen die Relaxationsdauer vernachlässigbar ist. In diesem Fall kann der Signalverlauf mit der
Langevinfunktion beschrieben werden. Die dritte Kategorie umfasst alle übrigen Partikelsysteme,
insbesondere Suspensionen von MNP-Clustern, die u.a. aufgrund von Interpartikelwechselwirkung
komplexe Signalverläufe ergeben, die sich praktisch nicht berechnen lassen. Spektroskopische
Untersuchungen sind damit dennoch durch das Anlegen entsprechender Referenzdatenbanken
möglich (Fingerprinting). Multiparametrisches RDS, d.h. die Wiederholung der Messung für z.B.
unterschiedliche Amplituden oder unterschiedliche Viskositäten des Suspensionsmediums, erzeugt
aufgrund mehrerer nichtlinearer Abhängigkeiten massive Unterschiede im resultierenden
multidimensionalen Datensatz. Das verspricht die Erreichbarkeit hoher spektroskopischer
Trennschärfen bei geeigneter Partikel- und Sequenzoptimierung.
Die Simulationen und experimentellen Ergebnisse dieser Arbeit zeigen grundsätzliche
Hürden und Möglichkeiten für das ebenfalls in dieser Arbeit eingeführte RDS auf. Es zeigt damit
grundlegende Aspekte auf, die für die Entwicklung von RDS-Hardware und die Optimierung von
MNP-Suspensionen nötig sind. Mit RDS wird in weiterführenden Arbeiten die Entwicklung von
hochempfindlichen Bioassays und die Erweiterung um die räumliche Kodierung angestrebt (RDI),
da der zugrunde liegende Effekt zugleich sehr vielversprechend als Grundlage für molekulare
Bildgebung ist.
To evaluate an iterative learning approach for enhanced performance of robust artificial‐neural‐networks for k‐space interpolation (RAKI), when only a limited amount of training data (auto‐calibration signals [ACS]) are available for accelerated standard 2D imaging.
Methods
In a first step, the RAKI model was tailored for the case of limited training data amount. In the iterative learning approach (termed iterative RAKI [iRAKI]), the tailored RAKI model is initially trained using original and augmented ACS obtained from a linear parallel imaging reconstruction. Subsequently, the RAKI convolution filters are refined iteratively using original and augmented ACS extracted from the previous RAKI reconstruction. Evaluation was carried out on 200 retrospectively undersampled in vivo datasets from the fastMRI neuro database with different contrast settings.
Results
For limited training data (18 and 22 ACS lines for R = 4 and R = 5, respectively), iRAKI outperforms standard RAKI by reducing residual artifacts and yields better noise suppression when compared to standard parallel imaging, underlined by quantitative reconstruction quality metrics. Additionally, iRAKI shows better performance than both GRAPPA and standard RAKI in case of pre‐scan calibration with varying contrast between training‐ and undersampled data.
Conclusion
RAKI benefits from the iterative learning approach, which preserves the noise suppression feature, but requires less original training data for the accurate reconstruction of standard 2D images thereby improving net acceleration.
Two-dimensional (2D) topological insulators are a new class of materials with properties that are
promising for potential future applications in quantum computers. For example, stanene represents
a possible candidate for a topological insulator made of Sn atoms arranged in a hexagonal
lattice. However, it has a relatively fragile low-energy spectrum and sensitive topology. Therefore,
to experimentally realize stanene in the topologically non-trivial phase, a suitable substrate
that accommodates stanene without compromising these topological properties must be found.
A heterostructure consisting of a SiC substrate with a buffer layer of adsorbed group-III elements
constitutes a possible solution for this problem. In this work, 2D adatom systems of Al and In
were grown epitaxially on SiC(0001) and then investigated structurally and spectroscopically by
scanning tunneling microscopy (STM) and photoelectron spectroscopy.
Al films in the high coverage regime \( (\Theta_{ML}\approx2\) ML\( ) \) exhibit unusually large, triangular- and
rectangular-shaped surface unit cells. Here, the low-energy electron diffraction (LEED)
pattern is brought into accordance with the surface topography derived from STM. Another Al
reconstruction, the quasi-one-dimensional (1D) Al phase, exhibits a striped surface corrugation,
which could be the result of the strain imprinted by the overlayer-substrate lattice mismatch.
It is suggested that Al atoms in different surface areas can occupy hexagonal close-packed and
face-centered cubic lattice sites, respectively, which in turn lead to close-packed transition regions
forming the stripe-like corrugations. On the basis of the well-known herringbone reconstruction
from Au(111), a first structural model is proposed, which fits well to the structural data from
STM. Ultimately, however, thermal treatments of the sample could not generate lower coverage
phases, i.e. in particular, a buffer layer structure.
Strong metallic signatures are found for In high coverage films \( (\Theta_{ML}\approx3\) to \(2\) ML\() \) by
scanning tunneling spectroscopy (STS) and angle-resolved photoelectron spectroscopy (ARPES),
which form a \( (7\times7) \), \( (6\times4\sqrt{3}) \), and \( (4\sqrt{3}\times4\sqrt{3}) \) surface reconstruction. In all these In phases
electrons follow the nearly-free electron model. Similar to the Al films, thermal treatments could
not obtain the buffer layer system.
Surprisingly, in the course of this investigation a triangular In lattice featuring a \( (1\times1) \)
periodicity is observed to host massive Dirac-like bands at \( K/K^{\prime} \) in ARPES. Based on this
strong electronic similarity with graphene at the Brillouin zone boundary, this new structure is
referred to as \textit{indenene}. An extensive theoretical analysis uncovers the emergence of an electronic
honeycomb network based on triangularly arranged In \textit{p} orbitals. Due to strong atomic spin-orbit
coupling and a comparably small substrate-induced in-plane inversion symmetry breaking this
material system is rendered topologically non-trivial. In indenene, the topology is intimately
linked to a bulk observable, i.e., the energy-dependent charge accumulation sequence within the
surface unit cell, which is experimentally exploited in STS to confirm the non-trivial topological
character. The band gap at \( K/K^{\prime} \), a signature of massive Dirac fermions, is estimated by
ARPES to approximately 125 meV. Further investigations by X-ray standing wave, STM, and
LEED confirm the structural properties of indenene. Thus, this thesis presents the growth and
characterization of the novel quantum spin Hall insulator material indenene.
After the discovery of three-dimensional topological insulators (TIs), such as tetradymite chalcogenides Bi$_2$Se$_3$, Bi$_2$Te$_3$ and Sb$_2$Te$_3$ – a new class of quantum materials characterized by their unique surface electronic properties – the solid state community got focused on topological states that are driven by strong electronic correlations and magnetism. An important material class is the magnetic TI (MTI) exhibiting the quantum anomalous Hall (QAH) effect, i.e. a dissipationless quantized edge-state transport in the absence of external magnetic field, originating from the interplay between ferromagnetism and a topologically non-trivial band structure. The unprecedented opportunities offered by these new exotic materials open a new avenue for the development of low-dissipation electronics, spintronics, and quantum computation. However, the major concern with QAH effect is its extremely low onset temperature, limiting its practical application. To resolve this problem, a comprehensive understanding of the microscopic origin of the underlying ferromagnetism is necessary.
V- and Cr-doped (Bi,Sb)$_2$Te$_3$ are the two prototypical systems that have been widely studied as realizations of the QAH state. Finding microscopic differences between the strongly correlated V and Cr impurities would help finding a relevant model of ferromagnetic coupling and eventually provide better control of the QAH effect in these systems. Therefore, this thesis first focuses on the V- and Cr-doped (Bi,Sb)$_2$Te$_3$ systems, to better understand these differences. Exploiting the unique capabilities of x-ray absorption spectroscopy and magnetic circular dichroism (XAS/XMCD), combined with advanced modeling based on multiplet ligand-field theory (MLFT), we provide a detailed microscopic insight into the local electronic and magnetic properties of these systems and determine microscopic parameters crucial for the comparison with theoretical models, which include the $d$-shell filling, spin and orbital magnetic moments. We find a strongly covalent ground state, dominated by the superposition of one and two Te-ligand-hole configurations, with a negligible contribution from a purely ionic 3+ configuration. Our findings indicate the importance of the Te $5p$ states for the ferromagnetism in (Bi, Sb)$_2$Te$_3$ and favor magnetic coupling mechanisms involving $pd$-exchange. Using state-of-the-art density functional theory (DFT) calculations in combination with XMCD and resonant photoelectron spectroscopy (resPES), we reveal the important role of the $3d$ impurity states in mediating magnetic exchange coupling. Our calculations illustrate that the kind and strength of the exchange coupling varies with the impurity $3d$-shell occupation. We find a weakening of ferromagnetic properties upon the increase of doping concentration, as well as with the substitution of Bi at the Sb site. Finally, we qualitatively describe the origin of the induced magnetic moments at the Te and Sb sites in the host lattice and discuss their role in mediating a robust ferromagnetism based on a $pd$-exchange interaction scenario. Our findings reveal important clues to designing higher $T_{\text{C}}$ MTIs.
Rare-earth ions typically exhibit larger magnetic moments than transition-metal ions and thus promise the opening of a wider exchange gap in the Dirac surface states of TIs, which is favorable for the realization of the high-temperature QAH effect. Therefore, we have further focused on Eu-doped Bi$_2$Te$_3$ and scrutinized whether the conditions for formation of a substantial gap in this system are present by combining spectroscopic and bulk characterization methods with theoretical calculations. For all studied Eu doping concentrations, our atomic multiplet analysis of the $M_{4,5}$ x-ray absorption and magnetic circular dichroism spectra reveals a Eu$^{2+}$ valence, unlike most other rare earth elements, and confirms a large magnetic moment. At temperatures below 10 K, bulk magnetometry indicates the onset of antiferromagnetic ordering. This is in good agreement with DFT results, which predict AFM interactions between the Eu impurities due to the direct overlap of the impurity wave functions. Our results support the notion of antiferromagnetism coexisting with topological surface states in rare-earth doped Bi$_2$Te$_3$ and corroborate the potential of such doping to result in an antiferromagnetic TI with exotic quantum properties.
The doping with impurities introduces disorder detrimental for the QAH effect, which may be avoided in stoichiometric, well-ordered magnetic compounds. In the last part of the thesis we have investigated the recently discovered intrinsic magnetic TI (IMTI) MnBi$_6$Te$_{10}$, where we have uncovered robust ferromagnetism with $T_{\text{C}} \approx 12$ K and connected its origin to the Mn/Bi intermixing. Our measurements reveal a magnetically intact surface with a large moment, and with FM properties similar to the bulk, which makes MnBi$_6$Te$_{10}$ a promising candidate for the QAH effect at elevated temperatures. Moreover, using an advanced ab initio MLFT approach we have determined the ground-state properties of Mn and revealed a predominant contribution of the $d^5$ configuration to the ground state, resulting in a $d$-shell electron occupation $n_d = 5.31$ and a large magnetic moment, in excellent agreement with our DFT calculations and the bulk magnetometry data. Our results together with first principle calculations based on the DFT-GGA$+U$, performed by our collaborators, suggest that carefully engineered intermixing plays a crucial role in achieving a robust long-range FM order and therefore could be the key for achieving enhanced QAH effect properties.
We expect our findings to aid better understanding of MTIs, which is essential to help increasing the temperature of the QAH effect, thus facilitating the realization of low-power electronics in the future.
The great progress in organic photovoltaics (OPV) over the past few years has been largely achieved by the development of non‐fullerene acceptors (NFAs), with power conversion efficiencies now approaching 20%. To further improve device performance, loss mechanisms must be identified and minimized. Triplet states are known to adversely affect device performance, since they can form energetically trapped excitons on low‐lying states that are responsible for non‐radiative losses or even device degradation. Halogenation of OPV materials has long been employed to tailor energy levels and to enhance open circuit voltage. Yet, the influence on recombination to triplet excitons has been largely unexplored. Using the complementary spin‐sensitive methods of photoluminescence detected magnetic resonance and transient electron paramagnetic resonance corroborated by transient absorption and quantum‐chemical calculations, exciton pathways in OPV blends are unravelled employing the polymer donors PBDB‐T, PM6, and PM7 together with NFAs Y6 and Y7. All blends reveal triplet excitons on the NFA populated via non‐geminate hole back transfer and, in blends with halogenated donors, also by spin‐orbit coupling driven intersystem crossing. Identifying these triplet formation pathways in all tested solar cell absorber films highlights the untapped potential for improved charge generation to further increase plateauing OPV efficiencies.
In der vorliegenden Arbeit werden die Konzeption und Realisierung eines Computertomographen zur Materialanalyse auf Basis eines Rasterelektronenmikroskops mit einem räumlichen Auflösungsvermögen im Nanometerbereich diskutiert. Durch einen fokussierten Elektronenstrahl, der mit einer Beschleunigungsspannung von 30 kV auf eine mikrostrukturierte Wolframnadel mit einem Spitzenradius von bis zu 50 nm gezielt wird, entsteht ein kleiner Röntgenbrennfleck über den mit geometrischer Vergrößerung hochauflösende Projektionen eines zu untersuchenden Objekts erzeugt werden. Durch Rotation des Testobjekts werden Projektionen aus verschiedenen Blickwinkeln aufgenommen und über einen speziellen Rekonstruktionsalgorithmus zu einem 3-dimensionalen Bild zusammengefügt.
Bei der Beurteilung der Einzelkomponenten des Geräts wird insbesondere auf Struktur, Form und den elektrochemischen Herstellungsprozess der Röntgenquelle eingegangen. Eine ausreichend genaue Positionierung von Messobjekt und Röntgenbrennfleck wird über Piezoachsen realisiert, während die Stabilität des Röntgenbrennflecks über die Elektronenoptik des Rasterelektronenmikroskops und die Form der Quellnadel optimiert wird.
Das räumliche Auflösungsvermögen wird über die Linienspreizfunktion an Materialkanten abgeschätzt. Für eine Wolfram-Block-Quelle ergibt sich dabei ein Auflösungsvermögen von 325 nm – 400 nm in 3D, während der Quellfleck einer Wolframnadel das Auflösungsvermögen der Anlage auf 65 nm – 90 nm in 2D und 170 nm – 300 nm in 3D bei Messungen an einem AlCu29-Testobjekt anhebt. Außerdem werden die Auswirkungen der Phasenkontrastcharakteristik der Röntgenquelle auf die rekonstruierten Bilder nach Anwendung eines Paganin-Filters diskutiert. Dabei zeigt sich, dass durch Anwendung des Filters ein verbessertes Signal-zu-Rausch-Verhältnis auf Kosten der räumlichen Bildauflösung erzielt werden kann.
Eine Vergleichsmessung mit einem kommerziell verfügbaren Röntgenmikroskop zeigt die Stärken des vorgestellten Systems bei Untersuchung von stark absorbierenden Messobjekten. Das kompakte Design erlaubt eine Weiterentwicklung in Richtung eines nanoCT-Moduls als Upgrade Option für Rasterelektronenmikroskope im Gegensatz zu den weitaus teureren bisher verbreiteten nanoCT-Geräten.
Einerseits besteht die einfachste Möglichkeit zum Ladungs- und Informationstransport zwischen zwei Punkten in deren direkter Verbindung durch eindimensionale Kanäle. Andererseits besitzen topologische Materialien exotische und äußerst vorteilhafte Eigenschaften, weshalb es nahe liegt, dass schon bald neue Anwendungen aus ihnen realisiert werden. Wenn diese beiden Entwicklungen zusammenkommen, dann ist ein grundlegendes Verständnis von Quanteninterferenz oder Hybridisierungseffekten in eindimensionalen, topologischen Kanälen von fundamentaler Wichtigkeit. Deshalb werden in der vorliegenden Arbeit Wechselwirkungen von eindimensionalen, topologisch geschützten Kantenzuständen, die an ungeradzahligen Stufenkanten auf der (001)–Oberfläche von Pb1−xSnxSe auftreten, untersucht. Aufgrund der lateralen Lokalisierung auf wenige Nanometer um eine Stufenkante herum und der Notwendigkeit zwischen gerad- und ungeradzahligen Stufenkantenhöhen zu unterscheiden, bieten sich die Rastertunnelmikroskopie und -spektroskopie als Methoden an. Die neu entdeckten Kopplungs- bzw. Wechselwirkungseffekte zwischen benachbarten Kantenzuständen treten auf, sobald der Stufe zu Stufe Abstand einen kritischen Wert von dkri ≈ 25nm unterschreitet. Dieses Kriterium kann durch verschiedene räumliche Anordnungen von Stufenkanten erfüllt werden. Infolgedessen werden sich kreuzende, parallel verlaufende und zusammenlaufende Stufenkanten genauer untersucht. Bei letzteren verändert sich entlang der Struktur kontinuierlich der Abstand und damit die Kopplungsstärke zwischen den beiden Randkanälen. Infolgedessen wurden drei Koppelungsregime identifiziert. (I) Ausgehend von einer schwachen Wechselwirkung zeigt der für die Kantenzustände charakteristische Peak im Spektrum zunächst eine Verbreiterung und Verminderung der Intensität. (II) Mit weiter zunehmender Wechselwirkung beginnt sich der Zustand in zwei Peaks aufzuspalten, sodass ab dkri ≈ 15nm an beiden Stufenkanten durchgehen eine Doppelpeak zu beobachten ist . Mit weiter abnehmendem Abstand erreicht die Aufspaltung Werte von einigen 10 meV, während sich die Intensität weiter reduziert. (III) Sobald zwei Stufenkanten weniger als etwa 5nm voneinander getrennt sind, konvergieren aufgrund der schwindenden Intensität und des sinkenden energetischen Abstands der beiden Peaks zu den van Hove Singularitäten die Spektren an den Stufenkanten gegen das Spektrum über einer Terrasse. i Die Aufspaltung verläuft in den Bereichen I und II asymmetrisch, d. h. ein Peak verbleibt ungefähr bei der Ausgangsenergie, während der andere mit zunehmender Kopplung immer weiter weg schiebt. Bezüglich der Asymmetrie kann kein Unterschied festgestellt werden, ob die zusammenlaufenden Stufenkanten eine Insel oder Fehlstelleninsel bilden oder ob die Stufenkanten sogar gänzlich parallel verlaufen. Es zeigt sich keine Präferenz, ob zunächst der niederenergetische oder der hochenergetische Peak schiebt. Erst im Regime starker Kopplung (III) kann beobachtet werden, dass beide Peaks die Ausgangsenergie deutlich verlassen. Im Gegensatz dazu kann bei sich kreuzenden Stufen ein erheblicher Einfluss der Geometrie, in Form des eingeschlossenen Winkels, auf das Spektrum beobachtet werden. Unabhängig vom Winkel existiert am Kreuzungspunkt selbst kein Kantenzustand mehr. Die Zustände an den vier Stufen beginnen, abhängig vom Winkel, etwa 10-15nm vor dem Kreuzungspunkt abzuklingen. Überraschenderweise zeigt sich dabei, dass im Fall rechtwinkliger Stufen gar keine Aufspaltung zu beobachten ist, während bei allen anderen Winkeln ein Doppelpeak festgestellt werden kann. Diese Entdeckung deutet auf Orthogonalität bezüglich einer Quantenzahl bei den beteiligten Kantenzustände hin. Neben einer nur theoretisch vorhergesagten Spinpolarisation kann dieser Effekt auch von dem orbitalem Charakter der beteiligten Dirac–Kegel verursacht sein. Da der topologische Schutz in Pb1−xSnxSe durch Kristallsymmetrien garantiert ist, wird als letzter intrinsischer Effekt der Einfluss von eindimensionalen Defekten auf die Kantenzustände untersucht. Berücksichtigt werden dabei ein nicht näher klassifizierbarer, oberflächennaher Defekt und Schraubversetzungen. In beiden Fällen kann ebenfalls eine Aufspaltung des Kantenzustands in einen Doppelpeak gezeigt werden. Im zweiten Teil dieser Arbeit werden die Grundlagen für eine Wiederverwendung von (Pb,Sn)Se–Oberflächen bei zukünftige Experimenten mit (magnetischen) Adatomen geschaffen. Durch Kombination von Inoenzerstäubung und Tempern wird dabei nicht nur eine gereinigte Oberfläche erzeugt, sondern es kann auch das Ferminiveau gezielt erhöht oder gesenkt werden. Dieser Effekt beruht auf eine Modifikation der Sn– Konzentration und der von ihr kontrollierten Anzahl an Defektelektronen. Als letztes sind erste Messungen an Cu- und Fe–dotierte Proben gezeigt. Durch die Adatome tritt eine n–Dotierung auf, welche den Dirac–Punkt des Systems in Richtung des Ferminiveaus verschiebt. Sobald er dieses erreicht hat kommt es zu Wechselwirkungsphänomenen an freistehenden Stufenkanten. Dies führt zu einer Doppelpeakstruktur mit einer feinen Aufspaltung von wenigen meV. Das Phänomen ist auf ein schmales Energiefenster beschränkt, bei dem die Lage des Dirac–Punkts nur etwa 5 meV (in beide Richtungen) von der des Ferminiveaus abweichen darf.
This thesis examines the electronic properties of two materials that promise the realization and observation of novel exotic quantum phenomena. For this purpose, angle-resolved photoemission forms the experimental basis for the investigation of the electronic properties. Furthermore, the magnetic order is investigated utilizing X-ray dichroism measurements.
First, the bulk and surface electronic structure of epitaxially grown HgTe in its three-dimensional topological insulator phase is investigated. In this study, synchrotron radiation is used to address the three-dimensional band structure and orbital composition of the bulk states by employing photon-energy-dependent and polarization-dependent measurements, respectively. In addition, the topological surface state is examined on in situ grown samples using a laboratory photon source. The resulting data provide a means to experimentally localize the bulk band inversion in momentum space and to evidence the momentum-dependent change in the orbital character of the inverted bulk states.
Furthermore, a rather new series of van der Waals compounds, (MnBi\(_2\)Te\(_4\))(Bi\(_2\)Te\(_3\))\(_n\), is investigated. First, the magnetic properties of the first two members of the series, MnBi\(_2\)Te\(_4\) and MnBi\(_4\)Te\(_7\), are studied via X-ray absorption-based techniques. The topological surface state on the two terminations of MnBi\(_4\)Te\(_7\) is analyzed using circular dichroic, photon-energy-dependent, and spin-resolved photoemission. The topological state on the (MnBi\(_2\)Te\(_4\))-layer termination shows a free-standing Dirac cone with its Dirac point located in the bulk band gap. In contrast, on the (Bi\(_2\)Te\(_3\))-layer termination the surface state hybridizes with the bulk valences states, forming a spectral weight gap, and exhibits a Dirac point that is buried within the bulk continuum. Lastly, the lack of unambiguous evidence in the literature showing a temperature-dependent mass gap opening in these magnetic topological insulators is discussed through MnBi\(_2\)Te\(_4\).
This PhD thesis addresses the photophysics of selected small organic molecules with the purpose of using them for efficient and even novel light sources. In particular, the studies presented focused on revealing the underlying exciton dynamics and determining the transition rates between different molecular states. It was shown how the specific properties and mechanisms of light emission in fluorescent molecules, molecules with phosphorescence or thermally activated delayed fluorescence (TADF), biradicals, and multichromophores can be utilized to build novel light-emitting devices. The main tool employed here was the analysis of the emitters’ photon statistics, i.e. the analysis of the temporal distribution of emitted photons, during electrical or optical excitation. In the introduction of this work, the working principle of an organic light-emitting diode (OLED) was introduced, while Chapter 2 provided the physical background of the relevant properties of organic molecules and their interaction with light. In particular, the occurrence of discrete energy levels in organic semiconductors and the process of spontaneous light emission were discussed. Furthermore, in this chapter a mathematical formalism was elaborated with the goal to find out what kind of information about the studied molecule can be obtained by analyzing its photon statistics. It was deduced that the intensity correlation function g (2)(t) contains information about the first two factorial moments of the photon statistics and that higher order factorial moments do not contain any additional information about the system under study if the system is always in the same state after the emission of a photon. To conclude the introductory part, Chapter 3 introduced the utilized characterization methods including confocal microscopy of single molecules, time correlated single photon counting and temperature dependent photoluminescence measurements. To provide the background necessary for an understanding of for the following result chapters, in Section 4.1 a closer look was taken at the phenomenon of blinking and photobleaching of individual molecules. For a squaraine-based fluorescent emitter rapid switching between a bright and dark state was observed during photoexcitation. Using literature transition rates between the molecular states, a consistent model was developed that is able to explain the distribution of the residence times of the molecule in the bright and dark states. In particular, an exponential and a power-law probability distribution was measured for the time the molecule resides in tis bright and dark state, respectively. This behavior as well as the change in photoluminescence intensity between the two states was conclusively explained by diffusion of residual oxygen within the sample, which had been prepared in a nitrogen-filled glovebox. For subsequent samples of this work, thin strips of atomic aluminum were deposited on the matrices to serve as oxygen getter material. This not only suppressed the efficiency of photobleaching, but also noticeably prolonged the time prior to photobleaching, which made many of the following investigations possible in the first place. For emitters used in displays, emission properties such as narrow-band luminescence and short fluorescence lifetimes are desired. These properties can be influenced not only by the emitter molecule itself, but also by the interaction with the chosen environment. Therefore, before focusing on the photophysics of individual small organic molecules, Section 4.2 highlighted the interaction of a perylene bisimide-based molecular species with its local environment in a disordered polymethyl methacrylate matrix. In a statistical approach, individual photophysical properties were measured for 32 single molecules and correlations in the variation of the properties were analyzed. This revealed how the local polarity of the molecules’ environment influences their photophysics. In particular, it was shown how an increase in local polarity leads to a red-shifted emission, narrower emission lines, broader vibronic splitting between different emission lines in combination with a smaller Huang-Rhys parameter, and a longer fluorescence lifetime. In the future, these results may help to embed individual chromophores into larger macromolecules to provide the chromophore with the optimal local environment to exhibit the desired emission properties. The next two sections focused on a novel and promising class of chromophores, namely linear coordinated copper complexes, synthesized in the group of Dr. Andreas Steffen at the Institute of Inorganic Chemistry at the University of Würzburg. In copper atoms, the d-orbitals are fully occupied, which prevents undesirable metal-centered d-d⋆ states, which tend to lie low in energy and recombine non-radiatively. Simultaneously, the copper atom provides a flexible coordination geometry, while complexes in their linear form are expected to exhibit the least amount of excited state distortions. Depending on the chosen ligands, these copper complexes can exhibit phosphorescence as well as temperature activated delayed fluorescence. In Section 4.3, a phosphorescent copper complex with a chlorine atom and a 1-(2,6-diisopropylphenyl)-3,3,5,5-tetramethyl-2-pyrrolidine-ylidene- ligand was tested for its suitability as an optically active material in an OLED. For this purpose, an OLED with a polyspirobifluorene-based copolymer matrix and the dopant at a concentration of 20 wt% was electrically excited. Deconvolution of the emission spectrum in contributions from the matrix and the dopant revealed that 60 % of the OLEDs emission was due to the copper complex. It was also shown that the shape of the emission spectrum of the copper complex remains unchanged upon incorporation into the OLED, but is red-shifted by about 233 meV. In Section 4.4, a second copper complex exhibiting thermally activated delayed fluorescence was analyzed. This complex comprised a carbazolate as well as a 2-(2,6- diisopropyl)-phenyl-1,1-diphenyl-isoindol-2-ium-3-ide ligand and was examined in the solid state and at the single-molecule level, where single photon emission was recorded up to an intensity of 78’000 counts per second. The evaluation of the second-order autocorrelation function of the emitted light proved an efficient transition between singlet and triplet excited states on the picosecond time scale. In the solid state, the temperature- dependent fluorescence decay of the complex was analyzed after pulsed photoexcitation in the temperature range between 300 K and 5 K. From these measurements, a small singlet-triplet energy gap of only 65 meV and a triplet sublevel splitting of 3.0 meV were derived. The transition rates between molecular states could also be determined. Here, the fast singlet decay time of τS1 = 9.8ns proved the efficient thermally activated delayed fluorescence process, which was demonstrated for the first time for this new class of copper(I) complexes thus. While the use of thermally activated delayed fluorescence is a potential way to harness otherwise long-living dark triplet states, radicals completely avoid dark triplet states. However, this usually comes with the huge drawback of the molecules being chemically unstable. Therefore, two chemically stable biradical species were synthesized in the framework of the DFG research training school GRK 2112 on Molecular biradicals: structure, properties and reactivity, by Yohei Hattori in the group of Prof. Dr. Christoph Lambert and Rodger Rausch in the group of Prof. Dr. Frank Würthner at the Institute of Organic Chemistry at the University of Würzburg, respectively. In Section 4.5, it was investigated how these molecules can be used in OLEDs. In the first isoindigo based biradical (6,6’-bis(3,5-di-tert-butyl-4-phenoxyl)-1,1’-bis(2- ethylhexyl)-[3,3’-biindolinyl-idene]-2,2’-dione) two tert-butyl moieties kinetically block chemical reactions at the place of the lone electrons and an electron-withdrawing core shifts the electron density into the center of the chromophore. With these properties, it was possible to realize a poly(p-phenylene vinylene) copolymer based OLED doped with the biradical and to observe luminescence during optical as well as electrical excitation. Analyzing shapes of the photo- and electroluminescence spectra at different doping concentrations, Förster resonance energy transfer was determined to be the dominant transition mechanism for excitons from the matrix to the biradical dopants. Likewise, OLEDs could be realized with the second diphenylmethylpyridine based birad- ical (4-(5-(bis(2,4,6-trichlorophenyl)methyl)-4,6-dichloropyridin-2-yl)-N-(4-(5-(bis(2,4,6- -trichlorophenyl)methyl)-4,6-dichloropyridin-2-yl)phenyl)-N-(4-methoxyphenyl)aniline) as dopant. In this biradical, chlorinated diphenylmethyl groups protect the two unpaired electrons. Photo- and electroluminescence spectra showed an emission in the near in- frared spectral range between 750 nm and 1000 nm. Also, Förster resonance energy trans- fer was the dominant energy transfer mechanism with an transfer efficiency close to 100 % even at doping concentrations of only 5 wt%. In addition to demonstrating the working OLEDs based in biradicals, the detection of luminescence of the two biradical species in devices also constitutes an important step toward making use of experimental techniques such as optically detected electron spin resonance, which could provide information about the electronic states of the emitter and their spin manifold during OLED operation. Another class of emitters studied are molecules in which several chromophores are co- valently linked to form a macrocyclic system. The properties of these multichromophores were highlighted in Section 4.6. Here, it was analyzed how the photophysical behavior of the molecules is affected by the covalent linking, which determines the interaction be- tween the chromophores. The first multichromophore, 2,2’-ditetracene, was synthesized by Lena Ross in the group of Prof. Dr. Anke Krüger at the Institute of Organic Chemistry at the University of Würzburg and was analyzed in this work both at the single-molecule level and in its aggregated crystalline form. While the single crystals were purified and grown in a vertical sublimation oven, the samples for the single molecule studies were prepared in matrices of amorphous polymethyl methacrylate and crystalline anthracene. Tetracene was analyzed concurrently to evaluate the effects of covalent linking. In samples where the distance between two molecules is sufficiently large, tetracene and 2,2’-ditracene show matching emission profiles with the only difference in the Franck-Condon factors and a de- creased photoluminescence decay time constant from 14 ns for tetracene to 5 ns for 2,2’- ditracene, which can be attributed to the increased density of the vibrational modes in 2,2’-ditracene. Evaluation of the photon statistics of individual 2,2’-ditracene molecules however showed that the system does not behave as two individual chromophores but as a collective state, preserving the spectral properties of the two tetracene chromophores. Complementary calculations performed by Marian Deutsch in the group of Prof. Dr. Bernd Engels at the Institute of Physical and Theoretical Chemistry at the University of Würzburg helped to understand the processes in the materials and could show that the electronic and vibronic modes of 2,2’-ditracene are superpositions of the modes occurring in tetracene. In contrast, single-crystalline 2,2’-ditetracene behaves significantly different than tetracene, namely exhibiting a red shift in photoluminescence of 150 meV, caused by an altered crys- talline packing that lowers the S1-state energy level. Temperature-dependent photolu- minescence measurements revealed a rich emission pattern from 2,2’-ditetracene single crystals. The mechanisms behind this were unraveled using photoluminescence lifetime density analysis in different spectral regions of the emission spectrum and at different tem- peratures. An excimer state was identified that is located about 5 meV below the S1-state, separated by a 1 meV barrier, and which can decay to the ground state with a time constant of 9 ns. Also, as the S1-state energy level is lowered below the E(S1) ≥ 2 ×E(T1) threshold, singlet fission is suppressed in 2,2’-ditetracene in contrast to tetracene. Therefore, at low temperatures, photoluminescence is enhanced by a factor of 46, which could make 2,2’- ditetracene a useful material for future applications in devices such as OLEDs or lasers. The second multichromophore species, para-xylylene bridged perylene bisimide macrocycles, were synthesized by Peter Spenst in the group of Prof. Dr. Frank Würthner at the Institute of Organic Chemistry at the University of Würzburg, by linking three and four perylene bisimides, respectively. To reveal the exciton dynamics in these macrocycles, highly diluted monomers as well as trimers and tetramers were doped into matrices of polymethyl methacrylate to create thin films in which individual macrocycles could be analyzed. The emission spectra of the macrocycles remained identical to those of the monomers, indicating weak coupling between the chromophores. Single photon emission could be verified for monomers as well as macrocycles, as exciton-exciton annihilation processes suppress the simultaneous emission of two photons from one macrocycle. Nevertheless, the proof of the occurrence of a doubly excited state was obtained by excitation power dependent photon statistics measurements. The formalism developed in the theory part of this thesis for calculating the photon statistics of multichromophore systems was used here to find a theoretical model that matches the experimental results. The main features of this model are a doubly excited state, fast singlet-singlet annihilation, and an efficient transition from the doubly excited state to a dark triplet state. The occurrence of triplet-triplet annihilation was demonstrated in a subsequent experiment in which the macrocycles were excited at a laser intensity well above the saturation intensity of the monomer species. In contrast to the monomers, the trimers and tetramers exhibited neither a complete dark state nor saturation of photoluminescence. Both processes, efficient singlet-singlet and triplet-triplet annihilation make perylene bisimide macrocycles exceptionally bright single photon emitters. These advantages were utilized to realize a room temperature electrically driven fluorescent single photon source. For this purpose, OLEDs were fabricated using polyvinylcarbazole and 2-tert-butylphenyl-5-biphenyl-1,3,4-oxadiazol blends as a host material for perylene bisimide trimers. Photon antibunching could be observed in both optically and electrically driven devices, representing the first demonstration of electrically driven single photon sources using fluorescent emitters at room temperature. As expected from the previous optical experiments, the electroluminescence of the molecules was exceptionally bright, emitting about 105 photons per second, which could be seen even by eye under the microscope. Finally, in the last section 4.7 of this thesis, two additional measurement schemes were proposed as an alternative to the measurement of the second-order correlation function g (2)(t) of single molecules, which only provides information about the first two factorial moments of the molecules’ photon statistics. In the first scheme, the g (3)(t) function was measured with three photodiodes, which is a consequential extension of the Hanbury Brown and Twiss measurement with two photodiodes. It was demonstrated how measuring the g (3)(t) function is able to identify interfering emitters with non-Poisson statistics in the experiment. The second setup was designed with an electro-optic modulator that repeatedly gen- erates photoexcitation in the form of a step function. The recording of luminescence transients for different excitation intensities yields the same results as the correspond- ing g (2)-functions measured on single emitters, both in their shape and in their depen- dence on excitation power. To demonstrate this concept, the TADF emitter TXO-TPA (2- [4-(diphenylamino)phenyl]-10,10-dioxide-9H-thioxanthen-9-one) was doped at a concen- tration of 10−4 wt% in a mCP (1,3-Bis(N-carbazolyl)benzene) matrix. This concentration was low enough that TXO-TPA molecules did not interact with each other, but an ensem- ble of molecules was still present in the detection volume. The intramolecular transition rates between singlet and triplet states of TXO-TPA could be derived with an error of at most 5 %. Other experimental techniques designed to obtain this information require ei- ther lengthy measurements on single molecules, where sample preparation is also often a challenge, or temperature-dependent fluorescence lifetime measurements, which require a cryostat, which in turn places constraints on the sample design used. In future, this ap- proach could establish a powerful method to study external factors influencing molecular transition rates. Overall, this thesis has introduced new molecular materials, revealed their photophys- ical properties, and demonstrated how they can be used to fabricate efficient and even novel light sources.
The fact that photovoltaics is a key technology for climate-neutral energy production can be taken as a given. The question to what extent perovskite will be used for photovoltaic technologies has not yet been fully answered. From a photophysical point of view, however, it has the potential to make a useful contribution to the energy sector. However, it remains to be seen whether perovskite-based modules will be able to compete with established technologies in terms of durability and cost efficiency. The additional aspect of ionic migration poses an additional challenge. In the present work, primarily the interaction between ionic redistribution, capacitive properties and recombination dynamics was investigated. This was done using impedance spectroscopy, OCVD and IV characteristics as well as extensive numerical drift-diffusion simulations. The combination of experimental and numerical methods proved to be very fruitful. A suitable model for the description of solar cells with respect to mobile ions was introduced in chapter 4.4. The formal mathematical description of the model was transferred by a non-dimensionalization and suitable numerically solvable form. The implementation took place in the Julia language. By intelligent use of structural properties of the sparse systems of equations, automatic differentiation and the use of efficient integration methods, the simulation tool is not only remarkably fast in finding the solution, but also scales quasi-linearly with the grid resolution. The software package was released under an open source license. In conventional semiconductor diodes, capacitance measurements are often used to determine the space charge density. In the first experimental chapter 5, it is shown that although this is also possible for the ionic migration present in perovskites, it cannot be directly understood as doping related, since the space charge distribution strongly depends on the preconditions and can be manipulated by an externally applied voltage. The exact form of this behavior depends on the perovskite composition. This shows, among other things, that experimental results can only be interpreted within the framework of conventional semiconductors to a very limited extent. Nevertheless, the built-in 99 potential of the solar cell can be determined if the experiments are carried out properly. A statement concerning the type and charge of the mobile ions is not possible without further effort, while their number can be determined. The simulations were applied to experimental data in chapter 6. Thus, it could be shown that mobile ions make a significant contribution to the OCVD of perovskite solar cells. j-V characteristics and OCVD transients measured as a function of temperature and illumination intensities could be quantitatively modeled simultaneously using a single global set of parameters. By the simulations it was further possible to derive a simple experimental procedure to determine the concentration and the diffusivity of the mobile ions. The possibility of describing different experiments in a uniform temperaturedependent manner strongly supports the model of mobile ions in perovskites. In summary, this work has made an important contribution to the elucidation of ionic contributions to the (photo)electrical properties of perovskite solar cells. Established experimental techniques for conventional semiconductors have been reinterpreted with respect to ionic mass transport and new methods have been proposed to draw conclusions on the properties for ionic transport. As a result, the published simulation tools can be used for a number of further studies.