541 Physikalische Chemie
Refine
Has Fulltext
- yes (10)
Is part of the Bibliography
- yes (10)
Document Type
- Doctoral Thesis (10)
Keywords
- Ultrakurzzeitspektroskopie (10) (remove)
This thesis describes novel concepts for the measurement of the static and dynamic properties of the electronic structure of molecules and nanocrystals in the liquid phase by means of coherent fluorescence-detected spectroscopy in two and three frequency dimensions. These concepts are based on the systematic variation ("phase cycling") of a sequence of multiple time-delayed femtosecond excitation pulses in order to decode a multitude of novel nonlinear signals from the resulting phase-dependent fluorescence signal. These signals represent any permutation of correlations between zero-, one-, two-, and three-quantum coherences. To this end, two new phase-cycling schemes have been developed which can simultaneously resolve and discriminate several nonlinear signals of sixth order, including those of the fourth order of nonlinearity.
By means of the sixth-order signals recorded in this work, static properties of highly excited electronic states in molecules such as their energies, transition dipole moments, and relative displacement of electronic potential surfaces, as well as dynamic properties in terms of their relaxation kinetics, can be ascertained. Furthermore, it was shown that these signals are suitable for the characterization of exciton-exciton correlations in colloidal quantum dots and for the measurement of ultrafast exciton-exciton annihilation in molecular aggregates.
The experiments performed in this thesis mark an important step towards the complete characterization of the nonlinear response of quantum systems. In view of this, the concept of fluorescence-detected multiple-quantum coherence multidimensional spectroscopy introduced here offers a unified, systematic approach.
In virtue of the technical advantages such as the use of a single excitation beam and the absence of nonresonant contributions, the measurement protocols developed here can be directly transferred to other incoherent observables and to sample systems in other states of matter. Furthermore, the approaches presented here can be systematically extended to higher frequency dimensions and higher orders of nonlinearity.
Um den jahrtausendealten Weg der Menschheit vom Papyrus über Buchdruck und siliziumbasierte Halbleiter in Richtung noch leistungsfähigerer Technologien zu gehen und weiterhin Heureka-Momente zu schaffen, bieten Kohlenstoffnanoröhren ein weites Forschungsfeld. Besonders die halbleitenden Charakteristika von SWNTs sowie die Manipulation dieser durch Dotierung bergen viele Möglichkeiten für zukünftige Anwendungen in moderner Elektrotechnologie. Der Weg zu einer industriellen Implementierung von SWNTs in neuartigen optoelektronischen Bauteilen ließe sich durch eine Ausweitung des Wissens bezüglich SWNTs und der dotierungsbasierten Anpassung ihrer Eigenschaften ebnen.
Mit dieser Erkenntniserweiterung als Zielsetzung wurden im Rahmen dieser Dissertation halbleitende, einwandige (6,5)-Kohlenstoffnanoröhren als chiralitätsreine, polymerstabilisierte Proben untersucht. Die ultrakurzzeitaufgelöste Spektroskopie der SWNTs erfolgte an organischen Suspensionen wie auch Dünnschichtfilmen, die je mittels eines gewissen Quantums an Gold(III)-chlorid dotiert worden waren. So konnten die ablaufenden Dynamiken auf einer ps-Zeitskala untersucht werden.
In Kapitel 4 konnte mittels transienter Absorptionsexperimente an redoxchemisch p-dotierter SWNT-Suspensionen zunächst gezeigt werden, dass sich die bei optischer Anregung gebildeten Trionen nicht analog zu Exzitonen diffusiv entlang der Nanoröhre bewegen, sondern lokalisiert vorliegen. Die längere trionischen Zerfallsdauer nach X$_1$- verglichen mit X$_1^+$-resonanter Anregung zeugt außerdem davon, dass das Trion aus dem Exziton gespeist wird. Der Einfluss der Dotierung auf die Zerfallsdynamiken von X$_1$ und X$_1^+$ wurde an SWNT-Dünnschichtfilmen untersucht. Das Photobleichsignal des Exzitons verschiebt hypsochrom und zerfällt schneller mit zunehmender Ladungsträgerdichte durch höherer Gold(III)-chloridkonzentrationen. Dies resultiert aus dem verringerten Abstand zwischen den Ladungsträgern, welche als nichtstrahlende Löschstellen fungieren. Für das X$_1^+$-PB ist ein ähnliches Verhalten zu beobachten. Dabei wird dieses Signal mit weiter steigender Dotierung von einer der H-Bande zuzuordnenden Photoabsorption überlagert. Diese lässt sich in einer starken Sättigung der Dotierung wie auch einer hohen Bandkantenverschiebung begründen.
In Kapitel 5 wurde die Größe der Exzitonen und Trionen in dotierten SWNT-Dünnschichtfilmen mittels des Phasenraumfüllmodells bestimmt. Dabei lag besonderes Augenmerk auf der Kompensation des PB/PA-Überlapps, dem schnellen Zerfall, einem Ausgleich von Differenzen zwischen Anrege- und Absorptionsspektrum sowie dem Anteil intrinsischer/dotierter Nanorohrsegmente, um korrigierte Größen $\xi_\mathrm{k}$ zu erhalten. Für die Trionengröße wurde zusätzlich der Überlapp der Absorptionsbanden einbezogen, um korrigierte Werte $\xi_{\mathrm{T,k}}$ zu bestimmen. $\xi_\mathrm{k}$ beträgt in der intrinsischen Form 6$\pm$2\,nm und bleibt bis zu einer Ladungsträgerdichte $n_{\mathrm{LT}}<0.10$\,nm$^{-1}$ etwa gleich, anschließend ist ein Absinken bis auf etwa 4\,nm bei $n_{\mathrm{LT}}\approx0.20$\,nm$^{-1}$ zu beobachten. Für diesen Trend ist die Überlagerung von Exziton- und H-Bande verantwortlich, da so der Faktor zur Bestimmung des Anteils intrinsischer Nanorohrsegmente an der SWNT verfälscht wird. Die Abweichung der intrinsischen Größe von den in der Literatur berichteten 13$\pm$3\,nm ist möglicherweise auf Unterschiede in der Probenpräparation zurückzuführen. Für die Trionengröße ergibt sich bei steigender Dotierung ein ähnliches Verhalten: Sie beträgt für $n_{\mathrm{LT}}<0.20$\,nm$^{-1}$ 1.83$\pm$0.47\,nm, was in der Größenordnung in guter Übereinstimmung mit der Literatur ist. Für höhere Dotierungen sinkt $\xi_{\mathrm{T,k}}$ bis auf 0.92$\pm$0.26nm ab. Dies erklärt sich dadurch, dass bei höherer $n_{\mathrm{LT}}$ die H-Bande das Spektrum dominiert, sodass der Einfluss der Absorptionsbandenüberlagerung nicht mehr vollständig durch den entsprechenden Korrekturfaktor kompensiert werden kann.
Kapitel 6 beschäftigte sich anstelle redoxchemischer Dotierung der nanoskaligen Halbleiter mit der (spektro-)elektrochemischen Untersuchung von Vorläufern molekularer Radikale. SWV-Messungen weisen dabei darauf hin, dass die Pyrene Pyr1-Pyr3 entsprechend der Anzahl ihrer Substituenten bei Reduktion Mono-, Bi- beziehungsweise Tetraradikale bilden. Die strukturelle Ähnlichkeit der Moleküle äußert sich in gleichen Reduktionspotentialen wie auch ähnlichen potentialabhängigen Absorptionsspektren. Während nur marginale Unterschiede in den PL-Spektren der neutralen und reduzierten Spezies festgestellt werden konnte, lieferte das zeitkorrelierte Einzelphotonenzählen aufschlussreichere Ergebnisse: So wird die Fluoreszenzlebensdauer stark von der Polarität der Umgegbung beeinflusst - bereits die Zugabe des Leitsalzes führt hier zu Änderungen. Die durchschnittliche Fluoreszenzlebensdauer $\tau_{\mathrm{av}}$ sinkt außerdem mit Reduktion und Radikalbildung; für höhere Emissionswellenlängen ist $\tau_{\mathrm{av}}$ außerdem höher. Insgesamt verdeutlichten die Experimente die gute Abschirmung zwischen Pyrenkern und Naphthalimidsubstituenten der Moleküle sowie die Sensibilität gegenüber dem Medium durch TICT, das Vorhandensein von Bi- und Tetraradikalen kann allerdings nicht vollständig belegt werden, wofür EPR-Messugen notwendig wären.
Nanoröhren, die auf dem Element Kohlenstoff basieren, besitzen ein großes Potential für ihre
Anwendung als neuartige und nachhaltige Materialien im Bereich der Optoelektronik und weiteren
zukunftsweisenden Technologiefeldern. Um jedoch hierfür genutzt werden zu können, ist
ein tiefgreifendes Kenntnis über ihre außergewöhnlichen photophysikalischen Eigenschaften notwendig.
Kohlenstoffnanoröhren sind als eindimensionale Halbleiter sehr vielseitige Materialien.
Jedoch ist der Zusammenhang zwischen ihrer Eignung als Halbleiter und der dafür notwendigen
Dotierung nur sehr unzureichend verstanden.
Die Ziele der vorliegenden Dissertation waren deshalb, ein grundlegendes Verständnis der photophysikalischen
Energietransferprozesse in Nanoröhren zu erlangen und den Einfluss von gezielten
Dotierungen auf diese Prozesse im Hinblick auf ihre Eigenschaften als eindimensionale Halbleiter
detailliert zu untersuchen. Die Grundlage für die Experimente bildeten verschiedene Filme
aus einwandigen (6,5)-Kohlenstoffnanoröhren, die durch ein Polyfluoren-Copolymer in einer
organischen Lösungsmittelumgebung isoliert wurden. Mit Hilfe der Ultrakurzzeitspektroskopie
wurden die auf einer schnellen (ps-ns) Zeitskala ablaufenden photophysikalischen Prozesse an
diesen Filmen unter verschiedenen Bedingungen untersucht und analysiert.
In Kapitel 4 wurde der generelle Energietransfer der Kohlenstoffnanoröhren in Polymermatrizen
im Detail studiert. Hierbei wurden durch Simulationen theoretische dreidimensionale
Verteilungen von Kohlenstoffnanoröhren erzeugt und die nach einem Energietransfer vorliegenden
Polarisationsanisotropien berechnet. Verschiedene Berechnungsansätze ergaben, dass die
Nanorohrdichte ϱSWCNT für ein Massenüberschuss X der Matrix nahezu unabhängig von dem
Röhrenvolumen war und durch ϱSWCNT = X−1 · 40 000 μm−1 angenähert werden konnte. Die
Simulationen lieferten von der Röhrendichte abhängige Gaußverteilungen der zwischen den
Nanoröhren vorliegenden Abständen. Aus den Verteilungen konnte weiterhin der Anteil an Röhren
bestimmt werden, die für einen Energietransfer zur Verfügung stehen. Weitere Simulationen
von Nanorohrverteilungen lieferten die Polarisationsanisotropie in Abhängigkeit von der Anzahl
an durchgeführten Energietransferschritten. Die Ergebnisse aus den Simulationen wurden zur
Interpretation der Ultrakurzzeitmessungen angewandt. Hierbei wurden durch die Variation der
Polymermatrix die zwischen den Nanoröhren vorliegenden Abstände verändert und damit die
Art und Intensität des Energietransfers kontrolliert. In Messungen der transienten Anisotropie
zeigte sich, dass ein Exziton nach seiner Erzeugung zwei depolarisierende Energietransferschritte
durchführte. Die Zerfallsdynamiken des Exzitons gaben auch klare Hinweise auf weitere nicht
depolarisierende Energietransferprozesse, die durch parallel zueinander stehende Übergangsdipolmomente
ermöglicht wurden. Eine Erklärung für dieses Verhalten lieferte die faserige
Struktur der Filme, die sich in Aufnahmen durch das Elektronenmikroskop zeigte.
Das Kapitel 5 beschäftigte sich mit dem Aufbau eines transienten Nahinfrarotspektrometers
und den nötigen experimentellen Umbauten zur Messung der transienten Absorption für energiearme
Signale im Spektralbereich unterhalb von 1.4 eV. Hierzu wurde die Weißlichterzeugung
für die Verwendung von Calciumfluorid umgebaut. Das erzeugte Weißlicht wurde in das aufgebaute
Prismenspektrometer eingekoppelt, um es weitestgehend linear auf einer Energieskala zu dispergieren. Auf diese Weise wurden energiearme Spektralkomponenten nicht auf unverhältnismäßig
viele Pixel verteilt und konnten mit ausreichender Intensität detektiert werden. Die
Lichtdetektion erfolgte mittels zweier Detektorzeilen aus Indiumgalliumarsenid, die das transiente
Signal durch eine direkte Referenzierung stabilisierten. Weiterhin wurde in diesem Kapitel die
Justage und die programmierte Ansteuerung des Systems detailliert beschrieben. Hierbei wurde
auf die Justage der Einkopplung per Freistrahl, die Kalibrierung mittels Bandpassspektren
sowie auf die Aufnahme von Weißlichtspektren und transienten Karten detailliert eingegangen.
An Nanorohrdispersionen durchgeführten Testmessungen zeigten, dass das transiente Nahinfrarotspektrometer
mit direkter Signalreferenzierung einwandfrei funktionierte und daher den
beobachtbaren Spektralbereich auf den Bereich von Energien bis unterhalb von 1 eV erweiterte.
Damit ermöglichte der Aufbau einen Zugang zu der Beobachtung größerer Nanorohrchiralitäten
sowie zu der Untersuchung von energiearmen, spektralen Signaturen von Nanorohrdefekten.
In Kapitel 6 wurde das transiente Nahinfrarotspektrometer genutzt, um das zeitabhängige
Verhalten von redoxchemisch p-dotierten Nanoröhren zu charakterisieren und quantitativ zu
beschreiben. Hierzu wurden die spektralen Eigenschaften von SWCNT-Dünnfilmen als Funktion
eines steigenden Dotierungsgrades durch die Messungen der transienten und linearen Absorption
studiert. In der linearen Absorption im Bereich von 0.9 - 2.5 eV vereinfachte sich das Spektrum
mit ansteigender Dotierung stark und verlor vor allem im Bereich des ersten Subbandes
deutlich an Oszillatorstärke. Bei starker Dotierung verschwanden die Signalbeiträge von X1
und der Phononenseitenbande. Weiterhin bleichte auch die bei mittleren Dotierungsgraden
auftauchende Trionenabsorption aus und ging in die breite Absorptionsbande der H-Bande über.
Das Erscheinen und Verschwinden der trionischen sowie exzitonischen Absorption war ebenfalls
in der transienten Absorption durch zeitgleich auftretende/verschwindende Photobleichsignale
zu erkennen. Sowohl der Zerfall des exzitonischen PB-Signals wie auch des Trions beschleunigte
sich mit einer steigenden Dotierung. Die Zerfallszeit des Exzitons im undotierten Film betrug
6.87 ps und verkürzte sich auf 0.732 ps bei höheren Dotierungsgraden. Die Zerfallszeit des Photobleichens
des Trions reduzierte sich von 2.02 ps auf 0.440 ps. Auffallend war hierbei, dass das
Trion im Vergleich zu dem Exziton exponentiell zerfiel und damit auf eine Lokalisierung dieses
Zustandes hinweist. Bei höheren Dotierungsmittelkonzentrationen tauchte in der transienten
Absorption ein neuer Signalbeitrag auf. Die Existenz dieses Signals konnte auf die H-Bande
zurückgeführt werden und könnte auf einer Verschiebung des linearen Absorptionsspektrums
aufgrund einer Renormalisierung der Bandlücke oder der Sättigung von Ladungsträgern beruhen.
Das Signal zeigte eine klare Abhängigkeit vom Dotierungsgrad des Nanorohrfilmes. So wies es
eine hypsochrome Verschiebung auf, wurde spektral breiter und seine Zerfallsdauer reduzierte
sich von 1.62 ps auf 0.520 ps mit steigendem Dotierungsgrad.
The invention of laser pulse shapers allowed for various quantum control experiments, where a chemical reaction is guided by specifically tailored laser pulses. However, despite of the prominent role of the liquid phase in chemistry, no successful attempt for controlling the selectivity of a bond-fission reaction has yet been reported in this state of matter. Promising candidates for such an experiment are C$_{\infty\mathrm{v}}$-symmetric trihalide anions with two different chemical bonds like $\ce{I2Cl-}$, because these molecules notionally offer the most simplest selectivity-control scenario of breaking either the one or the other bond and they are expected to dissociate under ultraviolet (UV) irradiation like it is known for the most-studied trihalide $\ce{I3-}$.
In order to investigate in this thesis the possibility that the dissociation reaction of such trihalides branches into two different photofragments, the ultrafast photodissociation dynamics of $\ce{I3-}$, $\ce{Br3-}$, $\ce{IBr2-}$ and $\ce{ICl2-}$ (point group D$_{\infty\mathrm{h}}$) as well as of $\ce{I2Br-}$ and $\ce{I2Cl-}$ (point group C$_{\infty\mathrm{v}}$) in dichloromethane solution were measured with broadband transient absorption spectroscopy in magic-angle configuration. The identification of the reaction pathway(s) relies on vibrational wavepacket oscillations, which survive the dissociation process and therefore carry not only informations about the reactant trihalides but also about the fragment dihalides.
These characteristic vibrational wavenumbers were extracted from the measured transient absorption spectra by globally fitting the population dynamics together with the wavepacket dynamics. Until recently, such a combined model function was not available in the well-established fitting tool Glotaran. This made it inevitable to develop a custom implementation of the underlying variable-projection fitting algorithm, for which the computer-algebra software Mathematica was chosen. Mathematica's sophisticated built-in functions allow not only for a high flexibility in constructing arbitrary model functions, but also offer the possibility to automatically calculate the derivative(s) of a model function. This allows the fitting procedure to use the exact Jacobian matrix instead of approximating it with the finite difference method.
Against the expectation, only one of the two thinkable photodissociation channels was found for each of the investigated C$_{\infty\mathrm{v}}$ trihalides. Since the photofragments recombine, their absorption signal as well as the reactant ground state bleach recover. This happens in a biexponential manner, which in the case of $\ce{I3-}$ was interpreted by Ruhman and coworkers with the direct formation of a neutral dihalogen fragment $\ce{I2}$ beside the negatively charged dihalide fragment $\ce{I2-}$. In this thesis, such a direct reaction channel was not found and instead the fast component of the biexponential decay is explained with vibrational excess energy mediating the recombination-preceding electron transfer process $\ce{I2- + I -> I2 + I-}$, while the slow component is attributed to cooled-down fragments.
In addition to the trihalide experiments, the possibility of a magic-angle configuration for polarization-shaping control experiments was theoretically investigated in this thesis by deriving magic-angle conditions for the third-order electric-dipole response signal of arbitrarily polarized laser pulses. Furthermore, the subtleties of anisotropy signals violating the well-known range of \numrange{-0.2}{0.4} were studied.
Coherent Multidimensional Spectroscopy in Molecular Beams and Liquids Using Incoherent Observables
(2018)
The aim of the present work was to implement an experimental approach that enables coherent two-dimensional (2D) electronic spectroscopy of samples in various states of matter. For samples in the liquid phase, a setup was realized that utilizes the sample fluorescence for the acquisition of 2D spectra. Whereas the liquid-phase approach has been established before, coherent 2D spectroscopy on gaseous samples in a molecular beam as developed in this work is in fact a new method. It employs for the first time cations in a time-of-flight mass spectrometer for signal detection and was used to obtain the first ion-selective 2D spectra of a molecular-beam sample. Additionally, a new acquisition concept was developed in this thesis that significantly decreases measurement times in 2D spectroscopy using optimized sparse sampling and a compressed-sensing reconstruction algorithm.
Characteristic for the variant of 2D spectroscopy presented in this work is the usage of a phase-coherent sequence of four laser pulses in a fully collinear geometry for sample excitation. The pulse sequence was generated by a custom-designed pulse shaper that is capable of rapid scanning by changing the pulse parameters such as time delays and phases with the repetition rate of the laser. The sample's response was detected by monitoring incoherent observables that arise from the final-state population, for instance fluorescence or cations. Phase cycling, i.e., signal acquisition with different combinations of the relative phases of the excitation pulses, was applied to extract nonlinear signal contributions from the full signal during data analysis.
Liquid-phase 2D fluorescence spectroscopy was established with the laser dye cresyl violet as a sample molecule, confirming coherent oscillations previously observed in literature that are originating from vibronic coherences in specific regions of the 2D spectrum.
The data set of this experiment was used subsequently to introduce optimized sparse sampling in 2D spectroscopy. An optimization algorithm was implemented in order to find the best sampling pattern while taking only one quarter of the regular time-domain sampling points, thereby reducing the acquisition time by a factor of four. Signal recovery was based on a new and compact representation of 2D spectra using the von Neumann basis, which required about six times less coefficients than the Fourier basis to retain the relevant information. Successful reconstruction was shown by recovering the coherent oscillations in cresyl violet from a reduced data set.
Finally, molecular-beam coherent 2D spectroscopy was introduced with an investigation of ionization pathways in highly-excited nitrogen dioxide, revealing transitions to discrete auto-ionizing states as the dominant contribution to the ion signal. Furthermore, the advantage of the time-of-flight approach to obtain reactant and product 2D spectra simultaneously enabled the observation of distinct differences in the multiphoton-ionization response functions of the nitrogen dioxide cation and the nitrogen oxide ionic fragment.
The developed experimental techniques of this work will facilitate fast acquisition of 2D spectra for samples in various states of matter and permit reliable direct comparison of results. Therefore, they pave the way to study the properties of quantum coherences during photophysical processes or photochemical reactions in different environments.
Time-resolved spectroscopy allows for analyzing light-induced energy conversion and
chromophore–chromophore interactions in molecular systems, which is a prerequisite in
the design of new materials and for improving the efficiency of opto-electronic devices.
To elucidate photo-induced dynamics of complex molecular systems, transient absorption
(TA) and coherent two-dimensional (2D) spectroscopy were employed and combined
with additional experimental techniques, theoretical approaches, and simulation models
in this work.
A systematic series of merocyanines, synthetically varied in the number of chromophores
and subsitution pattern, attached to a benzene unit was investigated in cooperation with
the group of Prof. Dr. Frank Würthner at the University of Würzburg. The global analysis
of several TA experiments, and additional coherent 2D spectroscopy experiments, provided
the basis to elaborate a relaxation scheme which was applicable for all merocyanine
systems under investigation. This relaxation scheme is based on a double minimum on the
excited-state potential energy surface. One of these minima is assigned to an intramolecular
charge-transfer state which is stabilized in the bis- and tris-chromophoric dyes by
chromphore–chromophore interactions, resulting in an increase in excited-state lifetime.
Electro-optical absorption and density functional theory (DFT) calculations revealed a
preferential chromophore orientation which compensates most of the dipole moment of
the individual chromophores. Based on this structural assignment the conformationdependent
exciton energy splitting was calculated. The linear absorption spectra of the
multi-chromophoric merocyanines could be described by a combination of monomeric and
excitonic spectra.
Subsequently, a structurally complex polymeric squaraine dye was studied in collaboration
with the research groups of Prof. Dr. Christoph Lambert and Prof. Dr. Roland Mitric
at the University of Würzburg. This polymer consists of a superposition of zigzag and
helix structures depending on the solvent. High-level DFT calculations confirmed the previous
assignment that zigzag and helix structures can be treated as J- and H-aggregates,
respectively. TA experiments revealed that in dependence on the solvent as well as the
excitation energy, ultrafast energy transfer within the squaraine polymer proceeds from
initially excited helix segments to zigzag segments or vice versa. Additionally, 2D spectroscopy
confirmed the observed sub-picosecond dynamics. In contrast to other conjugated
polymers such as MEH-PPV, which is investigated in the last chapter, ultrafast
energy transfer in squaraine polymers is based on the matching of the density of states
between donor and acceptor segments due to the small reorganization energy in cyanine-like
chromophores.
Finally, the photo-induced dynamics of the aggregated phase of the conjugated polymer
MEH-PPV was investigated in cooperation with the group of Prof. Dr. Anna Köhler at the University of Bayreuth. Our collaborators had previously described the aggregation of MEH-PPV upon cooling by the formation of so-called HJ-aggregates based on exciton
theory. By TA measurements and by making use of an affiliated band analysis distinct
relaxation processes in the excited state and to the ground state were discriminated. By
employing 2D spectroscopy the energy transfer between different conjugated segments
within the aggregated polymer was resolved. The initial exciton relaxation within the
aggregated phase indicates a low exciton mobility, in contrast to the subsequent energy
transfer between different chromophores within several picoseconds.
This work contributes by its systematic study of structure-dependent relaxation dynamics
to the basic understanding of the structure-function relationship within complex
molecular systems. The investigated molecular classes display a high potential to increase
efficiencies of opto-electronic devices, e.g., organic solar cells, by the selective choice of
the molecular morphology.
Within the framework of this thesis, photolysis reactions in the liquid phase were investigated by means of ultrafast optical spectroscopy. Apart from molecular studies dealing with the highly spin-dependent reactivity of diphenylcarbene (DPC) in binary solvent
mixtures and ligand dissociation reactions of so-called CO-releasing molecules (CORMs),
special emphasis was put on the implementation and characterization of methods improving
and extending the signal detection in conventional pump–probe transient absorption setups.
The assumption of DPC being an archetypal triplet-ground-state arylcarbene was recently questioned by matrix-isolation studies at low temperatures. DPC embedded in argon matrices revealed a hitherto unknown reactivity when the carbene environment was modified by small amounts of methanol dopant molecules. To complement these findings with liquid-phase experiments at room temperature, femtosecond pump–probe transient absorption spectroscopy with probing in the visible and ultraviolet regime was employed to unravel primary reaction processes of DPC in solvent mixtures. Supported by quantum chemical simulations conducted by our collaborators, it was shown that a competition between the reaction pathways occurs that not only depends on the solvent molecule near-by but also on its interaction with other solvent molecules. In-depth analysis of the solvation dynamics and the amount of nascent intermediates corroborates the importance of a hydrogen-bonded complex with a protic solvent molecule, in striking analogy to complexes found at cryogenic temperatures.
Probing the transient absorption of molecules in the mid-infrared spectral range benefits from the high chemical specificity of molecules’ vibrational signatures. The technique of chirped-pulse upconversion (CPU) constitutes a promising alternative to standard direct multichannel MCT detection when accessing this spectral detection window. Hence, one chapter of this thesis is dedicated to a direct comparison between both detection methods. By conducting an exemplary pump–probe transient absorption experiment, it became evident, that the additional nonlinear interaction step is responsible for increased noise levels when using CPU. However, a correction procedure capable of removing these additional noise contributions—stemming from the fundamental laser radiation used for upconversion—was successfully tested. Perhaps most importantly for various spectroscopic applications, CPU scored with a significantly extended detection bandwidth owing to the high pixel numbers of modern CCD cameras.
Transition-metal complexes capable of releasing small molecular messengers upon photoactivation are promising sources of gasotransmitters such as carbon monoxide (CO) or nitric oxide (NO) in biological applications. However, only little is known about the characteristic time scales of ligand dissociation in this class of molecules. For this purpose, two complexes were investigated with femtosecond time resolution: [Mn(CO)3(tpm)]Cl with tpm=tris(2-pyrazolyl)methane, a manganese tricarbonyl complex which has proven to be selective and cytotoxic to cancer cells, and [Mo(CO)2(NO)(iPr3tacn)]PF6 with iPr3tacn=1,4,7-triisopropyl-1,4,7-triazacyclononane, a molybdenum complex containing both carbonyl and nitrosyl ligands. By conducting pump–probe transient absorption measurements in different spectral probing windows supported by quantum chemical calculations and linear absorption spectroscopy, it was shown that both complexes are able to release one CO ligand within the first few picoseconds after UV excitation. The results complement existing studies which focused on the molecules’ ligand-releasing properties upon long-term exposure. The additional information gained on an ultrafast time scale provides a comprehensive understanding of individual reaction steps connected with ligand release in this class of molecules. Hence, the studies might create new incentives to develop modified molecules for specific applications.
Photoinduced processes are nowadays studied with a huge variety of spectroscopic methods. In the liquid phase, transient absorption spectroscopy is probably the most versatile pump–probe technique used to study light-induced molecular phenomena. Optical time-resolved spectroscopy is established in a large number of laboratories and is still further being developed with respect to many technical aspects. Nevertheless, the full potential of shortening the data-acquisition time—necessary for the investigation of rapidly photodegrading samples and observation of macroscopically fast processes—achievable with high-repetition-rate laser systems and shot-to-shot detection was not fully exploited. Especially, shot-to-shot detection is highly beneficial due to the high correlation of subsequent laser pulses.
The development and implementation of 100 kHz broadband shot-to-shot data acquisition was presented in Chapter 3. For an established laser dye as a benchmark system, ultrafast excited-state dynamics were measured for the first time with broadband shot-to-shot detection at 100 kHz. An analysis of both the noise characteristics of the employed laser and the correlation of subsequent pulses quantified the advantage of shot-to-shot data acquisition. In the utilized software environment, the time for measuring a complete data set could be sped up by a factor of three or even higher compared to a laser system working at 1 kHz. So far, the limiting factor is the data processing and the movement of the mechanical delay stage. Nevertheless, the new shot-to-shot detection has the potential to shorten the measurement time up to a factor of 100. The data quality is improved by a factor of three when the hitherto conventional averaging scheme is compared to shot-to-shot acquisition for the same number of laser pulses. The expansion of shot-to-shot data acquisition for high repetition rates will allow studies on sensitive samples as exposure times can strongly be reduced to achieve the same signal-to-noise ratio. In addition, multidimensional spectroscopy can also be extended to high-repetition shot-to-shot readout allowing an efficient recording of data. Therefore, in future experiments, dynamics and couplings in sensitive samples and kinetic processes could be studied in more detail.
Complex photophysical and photochemical phenomena are subject of many fields of research. Many of these multifaceted processes are not yet fully understood. Therefore, a possible approach is the elucidation of single reaction steps with the combination of transient absorption spectroscopy and a suitable, less complex model system. The systematic variation of the model system’s properties and environments, e.g., by chemical substitution or adequate choice of the solvent allows the determination of essential entities and reactivities thereof. Proper knowledge of an individual intermediate step and its determining factors can enhance the understanding of the complete photoreaction process.
The application of transient absorption spectroscopy was shown for the optically-induced electron transfer in a series of donor–acceptor oligomers in Chapter 4. In general, the solvent relaxation times were isolated from the back-electron-transfer dynamics by a global lifetime analysis. For the smallest oligomeric structure where complete charge separation is possible, an ultrafast equilibration leads to charge recombination from the configuration showing the lowest barrier for recombination. The back-electron transfer strongly depends on the utilized solvent. Whereas in dichloromethane the back-electron transfer occurs with the maximum rate in the barrierless optimal region, the dynamics in toluene are governed by a Marcus inverted-region effect. The experimentally observed rates were also estimated by theoretical calculations of the respective barriers. The study did not only successfully unravel charge transfer in the oligomeric systems but also improved the understanding of the electron-transfer properties of larger polymers from an earlier study. Therefore, the combination of length variation and time-resolved spectroscopy is an important step towards the correct prediction of charge-carrier dynamics in macroscopic devices, e.g., for photovoltaics.
The bond dissociation of a carbon-monoxide-releasing molecule in aqueous solution was studied in Chapter 5 as a prototype reaction for the photo-triggered breaking of a bond. It was shown that upon excitation only one carbon-monoxide ligand of the tricarbonyl complex is dissociated. A fraction of the photolyzed molecules restore the intact initial complex by geminate recombination within the temporal resolution of the experiment. However, the recombination could be detected by the hot ground-state infrared absorption of the complex. The detectable dicarbonyl formed upon CO release distributes excess energy from the absorbed photon into low-frequency modes which result in broadened absorption bands like for the recombined tricarbonyl. The free coordination site in the ligand sphere is filled with a solvent water molecule. Despite numerous studies of metal carbonyls studied in alkaneous solutions, the elucidation of the dynamics of a CORM in aqueous solution added another important detail to the photochemistry of this class of compounds. Experiments employing a second ultraviolet pump pulse did not trigger further CO dissociation and hence no formation of a monocarbonyl species; this might either be due to a different release mechanism without a further photochemical step or a strong spectral shift of the dicarbonyl’s absorption. Both reasons could explain why degenerate pump–repump–probe spectroscopy is inefficient. However, further experiments with ultraviolet probe pulses could substantiate whether the intermediate dicarbonyl reacts further photochemically or not. Apart from the model-system character of the CORM for bond dissociation, the study could determine exactly how many CO ligands are initially photolyzed off. Detailed knowledge of the release mechanism will affect the previous use and application as well as the further development of CORMs as therapeutic prodrugs to deliver high local concentrations of CO in cancerous or pathological tissue. Hence, the study of two-photon absorption properties which are important for in vivo applications of CORMs should be the main focus in further spectroscopic experiments.
In Chapter 6, both abovementioned molecular phenomena—electron transfer and bond dissociation—were studied in combination. The photochemistry of a tetrazolium salt was studied in detail in a variety of different solvents. Being a relatively small molecule, the studied tetrazolium cation shows a multifaceted photochemistry and is therefore a textbook example for the combination of ultrafast molecular phenomena studied in different environments. Within femtoseconds, the tetrazolium ring is opened. The biradicalic species is then reduced via uptake of an electron from the solvent. The formation of the ring-open formazan photoproduct from this point of the reaction sequence on was excluded by experiments with acidic pH value of the solution. The ring-open radical is stabilized by ring-closure. The resulting tetrazolinyl radical was already observed in experiments with microsecond time resolution. However, its formation was observed in real time for the first time in this study. Irradiation of a tetrazoliumsalt solution yields different photoproduct distributions depending on the solvent. However, it was shown that all photoproducts have a tetrazolinyl radical as a common precursor on an ultrafast time scale. In combination with studies from the literature, the complete photochemical conversion of a tetrazolium salt was clarified in this study. Apart from the prototype character of the reaction sequence, the reaction mechanism will have impact on research associated with life science where tetrazolium assays are used on a daily basis without taking into account of photochemical conversion of the indicating tetrazolium ion and its photochemically formed reactive intermediates. On the basis of the tetrazolium-ion photochemistry, the rich photochemistry of the formazan photoproduct, including structural rearrangements and subsequent reformation of the tetrazolium ion, might be the subject of future studies.
This thesis shows a method advancement and application of transient absorption spectroscopy to exemplary molecular model systems. The insights into each respective field did not only enlighten singular aspects, but have to be seen in a much larger context. Understanding complex photoinduced processes bottom-up by learning about their constituting steps—microscopically and on an ultrafast time scale—is an ideal method to approach understanding and prediction of phenomena in large molecular systems like biological or artificial architectures as for example used in photosynthetic light-harvesting and photovoltaics.
This work brings forward successful implementations of ultrafast chirality-sensitive spectroscopic techniques by probing circular dichroism (CD) or optical rotation dispersion (ORD). Furthermore, also first steps towards chiral quantum control, i.e., the selective variation of the chiral properties of molecules with the help of coherent light, are presented.
In the case of CD probing, a setup capable of mirroring an arbitrary polarization state of an ultrashort laser pulse was developed. Hence, by passing a left-circularly polarized laser pulse through this setup a right-circularly polarized laser pulse is generated. These two pulse enantiomers can be utilized as probe pulses in a pump--probe CD experiment. Besides CD spectroscopy, it can be utilized for anisotropy or ellipsometry spectroscopy also. Within this thesis, the approach is used to elucidate the photochemistry of hemoglobin, the oxygen transporting protein in mammalian blood. The oxygen loss can be triggered with laser pulses as well, and the results of the time-resolved CD experiment suggest a cascade-like relaxation, probably through different spin states, of the metallo-porphyrins in hemoglobin.
The ORD probing was realized via the combination of common-path optical heterodyne interferometric polarimetry and accumulative femtosecond spectroscopy. Within this setup, on the one hand the applicability of this approach for ultrafast studies was demonstrated explicitly. On the other hand, the discrimination between an achiral and a racemic solution without prior spatial separation was realized. This was achieved by inducing an enantiomeric excess via polarized femtosecond laser pulses and following its evolution with the developed polarimeter. Hence, chiral selectivity was already achieved with this method which can be turned into chiral control if the polarized laser pulses are optimized to steer an enhancement of the enantiomeric excess.
Furthermore, within this thesis, theoretical prerequisites for anisotropy-free pump--probe experiments with arbitrary polarized laser pulses were derived. Due to the small magnitude of optical chirality-sensitve signals, these results are important for any pump--probe chiral spectroscopy, like the CD probing presented in this thesis. Moreover, since for chiral quantum control the variation of the molecular structure is necessary, the knowledge about rearrangement reactions triggered by photons is necessary. Hence, within this thesis the ultrafast Wolff rearrangement of an α-diazocarbonyl was investigated via ultrafast photofragment ion spectroscopy in the gas phase. Though the compound is not chiral, the knowledge about the exact reaction mechanism is beneficial for future studies of chiral compounds.
Pulse-Sequence Approaches for Multidimensional Electronic Spectroscopy of Ultrafast Photochemistry
(2014)
Observing chemical reactions in real time with femtosecond laser pulses has evolved into a very popular field of research since it provides fascinating insights into the nature of photochemical transformations. Nevertheless, many photochemical reactions are still too complex for which reason the underlying mechanisms and all engaged species cannot be identified thoroughly. In these cases, conventional time-resolved spectroscopy techniques reach their technical limits and advanced approaches are required to follow the conversion of reactants to their products including all reaction intermediates.
The aim of this work was therefore the development of novel methods for ultrafast spectroscopy of photoreactive systems. Though the concept of coherent multidimensional spectroscopy has so far exclusively been used to explore photophysical phenomena, it also offers great potential for the study of photochemical processes due to its capability of extracting spectroscopic information along several frequency dimensions. This allows resolving the photochemical connectivity between various interconvertible molecular species with ultrafast temporal resolution on the basis of their absorption and emission properties as the spectral correlations are explicitly visualized in the detected spectra.
The ring-open merocyanine form of the photochromic compound 6-nitro BIPS was studied in Chap. 4 of this work. Merocyanines and their associated ring-closed spiropyrans are promising candidates for future applications as, for instance, molecular electronics or optical data storage due to their unique property of being switchable between two stable congurations via light illumination. Transient absorption with sub-50 fs temporal resolution and broadband probing was employed to characterize the photodynamics of this system with variable excitation wavelengths. Using global data analysis, it could be inferred that two different merocyanine isomers with differing excited-state lifetimes exist in solution. These isomers differ in the cis/trans conguration in the last bond of the methine bridge. The minority of isomers exist in the all-trans conguration (TTT) while the isomer with a cis conguration of the third dihedral angle (TTC) is dominant. A characteristic band, detected after long pump-probe delays, was attributed to the unidirectional cis->trans photoisomerization reaction of the TTC to the TTT form. The quantum yield of the reaction was estimated to be (18+-4) %. In addition, pronounced coherent vibrational wave-packet oscillations were observed and it was concluded that these signatures are related to the product formation.
Coherent two-dimensional electronic spectroscopy was successfully implemented using a partially collinear pump-probe beam geometry in combination with a femtosecond pulse shaper. The use of a whitelight probe continuum enabled us to probe contributions far-off the diagonal over the complete visible range. By properly adjusting the relative phase between the first two laser pulses with the pulse shaper, the principle of phase-cycling was explained and it was demonstrated that the measurement can be carried out in the so-called "rotating frame" in which the observed frequencies detected during the coherence time are shifted to lower values. It was shown that these concepts allow the extraction of the desired background-free photon echo while the amount of necessary data points is highly reduced.
In order to put our proposal of multidimensional spectroscopy of photoreactive systems into practice, third-order two- and three-dimensional spectroscopy was then employed for an in-depth analysis of a photoreactive process, in which the photoisomerization of 6-nitro BIPS served as a model system. The measured two-dimensional spectra revealed the cis->trans photoisomerization after long population times. By collecting a large data set of two-dimensional spectra for short population times and by applying a Fourier transform along the population time axis, the third-order three-dimensional spectrum was obtained. The novelty of this approach compared to coherent two-dimensional spectroscopy is the introduction of a third axis associated with the vibrational frequencies of the molecular system. In this way, the formation of the reaction product was evidenced and it was shown that the product is formed in its first excited singlet state within 200 fs after excitation. This method hence visualizes the photochemical connections between different reactive molecular species in an intuitive manner and further exposes the normal modes connecting reactant and product. Such conclusions cannot be drawn with conventional third-order techniques such as transient absorption since they are
not capable of capturing the full third-order response, but only a subset of it. The reaction mechanism and the role of the observed vibrational modes were uncovered by comparing the experimental data with the results of high-level quantum-chemical calculations performed by our collaborators in the group of Prof. B. Engels from the
theoretical chemistry department at the University of Würzburg. Specific calculated molecular normal modes could be assigned to the experimentally observed vibrational frequencies and potential energy surfaces of the electronic ground state and of the first excited state were computed. The technique implemented in this chapter is general and is applicable for the time-resolved analysis of a wide range of chemical reaction networks.
In the first part of Chap. 5, coherent two-dimensional spectroscopy was employed to track the reaction paths of the related 6,8-dinitro BIPS after S1 excitation. Several differences to the photochemical properties of 6-nitro BIPS were found. From the 2D spectra, the cis-trans isomerization between the two merocyanine isomers could be excluded as a major reaction path for this compound. To explore the dynamics after reexcitation to higher-lying electronic states, pump-repump-probe spectroscopy was implemented and the formation of a new species, a radical cation, was observed. To identify the precursor isomer, triggered-exchange two-dimensional spectroscopy, a fifth-order technique previously only available in the infrared regime for vibrational transitions, was implemented for the first time for electronic excitations in the visible. This approach combines the properties of the pump-repump-probe technique with the potential of coherent two-dimensional spectroscopy. It correlates the absorption frequency of a reactive molecular species with the emission signatures of the product formed from this species after an additional absorption of a photon. Using this method, it was unambiguously proven that only the TTC isomer reacts to the radical cation thus forming the precursor species of the reaction. Electronic triggered-exchange two-dimensional spectroscopy is hence another improved technology for time-resolved spectroscopy with applications in the study of multistep photoreactions and higher-lying electronic states. While in the two preceding chapters third- and fifth-order experiments were discussed that neglect the vectorial character of light-matter interactions, Chap. 6 focused on a novel theoretical formalism enabling the description of light fields optimized for polarization-sensitive higher-order nonlinearities. This formalism is based on the von Neumann time-frequency representation of shaped femtosecond laser pulses which permits the definition of multipulse sequences on a discrete time-frequency lattice. Hence, not only the temporal spacing between subpulses is adjustable, but also the center frequencies may be adapted such that they fit the experimental requirements. This method was generalized to the description of pulse sequences with time-varying polarization states. It was shown that by using this description, the polarization ellipticity, orientation angle, relative phase and intensity, and the time-frequency location of each subpulse is explicitly controllable. The accuracy of the transformations from Fourier space to von Neumann domain and vice versa was demonstrated. Moreover, a strict accordance between the von Neumann polarization parameters with the conventional parameters in time domain was found for well separated subpulses. A potential future application of this approach is polarization-sensitive multidimensional spectroscopy in which hidden cross peaks may be isolated by defining the pulses in the von Neumann picture with suitable polarization sequences. This method could also be used in quantum control experiments in which the polarization of the light field is used as a major control knob.
This thesis summarizes our efforts to open the field of femtochemistry to the concept of coherent multidimensional electronic spectroscopy. Making use of femtosecond pulse shaping, sub-50 fs temporal resolution, broadband spectral probing, higher-order nonlinearities, and new types of laser pulse descriptions, the presented methods might stimulate further future advancements in this research area.