| dc.description.abstract |
Over the past decades, transient absorption (TA) spectroscopy has matured into a routine, broadband femto- to microsecond probe for photoinduced dynamics. It is uniquely powerful to disentangle photophysical pathways, including the generation and relaxation of triplet states that govern efficiencies in applications ranging from light-emitting diodes and solar cells to photodynamic therapy and sensing. However, turning TA experiments into quantitative insight often remains a bottleneck since it demands global modeling and substantial coding expertise, while existing workflows too often lack statistical robustness, transparency and flexibility.
This cumulative thesis, based on four research papers, addresses these limitations in TA data handling and modeling by developing a software framework covering the full TA workflow from import and preprocessing to analysis and visualization.
This workflow is then applied to investigate the nature, generation and relaxation of triplets in oxygen-sensitive tungsten iodide clusters and heptacenes.
First, I developed TAPAS (Transient Absorption Processing and Analysis Software) as a fully integrated graphical user interface. Leveraging a modern machine-learning compiler ecosystem, substantial speedups in the computationally intensive fitting engine were achieved when compared to conventional packages. I implemented methods to estimate parameter confidence intervals and correlations via automatic differentiation, effective-sample-size scaling, and Bayesian posterior analysis. These capabilities were validated on multiple TA datasets and complementary measurements of rose bengal, yielding a single, self-consistent kinetic model. The analysis also exposed parameter correlations and weak identifiability in smaller datasets or local fits, which is often not captured by conventional software, resulting in a misleading impression of model robustness. This clearly emphasizes the advantage of the TAPAS engine for jointly fitting multiple datasets and complementary information, as well as the value of the implemented statistical methods for a more reliable assessment of model and parameter uncertainties and identifiability.
Second, I established TA spectroscopy to monitor triplet generation and relaxation of a novel cationic tungsten iodide cluster under varied conditions, correlating the results with time-resolved emission measurement. By globally modeling multiple TA datasets spanning different temporal windows, I characterized the excited-state properties including intersystem crossing (ISC) rates, triplet-emission lifetimes and oxygen quenching. The cationic cluster undergoes ultrafast ISC within about 2 ps into a hot triplet state, which further relaxes to the emissive triplet and then decays on microsecond timescales.
Building on this, in the third study on two general anionic tungsten iodide clusters, I combined TA with time- and temperature-resolved emission spectroscopy and relativistic time-dependent density functional theory to explore the ultrafast triplet-state generation and the unusual broad and temperature-sensitive emission characteristics. Tracking the dynamics from 200 fs to 400 µs shows that both clusters populate triplet states within 6 ps, which subsequently decay either via bimolecular oxygen quenching or phosphorescence. Extracted emission lifetimes and oxygen-quenching rates agree with complementary time-resolved emission measurements. Further, temperature-dependent emission measurements from 4 to 340 K of the clusters dispersed in a polymer matrix reveal systematic shifts in both spectra and lifetimes. Using a group-theoretical model with three distinct emissive spin sublevels yields satisfactory simulations, contrasting earlier reports on molybdenum-based clusters that invoked four sublevels. Relativistic calculations further indicate substantial excited-state geometric relaxation, challenging a purely group-theoretical description and motivating a relativistic model with three thermally accessible excited-state geometries, each possessing three triplet sublevels.
Finally, the photophysics of a novel photochemically stable heptacene derivative were investigated using comprehensive steady-state absorption, emission, and TA measurements. These studies show that the lowest excited state is an electric-dipole-forbidden (dark) transition, in contrast to shorter acenes. In this context, I performed broadband TA experiments which reveal rapid triplet formation and relaxation, plausibly facilitated by this dark singlet state. These results were used to discuss the underlying triplet-generation pathway, considering both ISC and singlet fission as viable mechanisms.
In summary, this thesis delivers an open-source software providing the whole TA processing-, analysis-, modeling- and visualization- workflow with comprehensive statistical treatment essential for complex models and uses this framework to elucidate triplet-state population, configuration and relaxation in tungsten iodide clusters and heptacenes. |
en |
