SPECTROSCOPIC AND STRUCTURAL STUDIES OF METALLOPORPHYRINS AS POTENTIAL MATERIALS FOR NONCOHERENT PHOTON UP-CONVERSION BY TRIPLET-TRIPLET ANNIHILATION ENHANCED SOLAR CELLS

Non-coherent photon up-conversion by triplet-triplet annihilation (NCPU-TTA) is a promising approach to improve the incident photon to current conversion efficiency of dye-sensitized solar cell (DSSC). This is because it converts below-band-gap photons, which are normally transmitted through a solar cell, into usable high energy photons. NCPU-TTA involves interaction of two long-lived triplet states of a fluorophore, to produce one molecule in the ground state and one in a higher excited electronic state (i.e. 2T1  S2 + S0). This interaction has been demonstrated to take place, with good efficiency in both solution and solid state, using closed-shell metalloporphyrin (MP) systems - macromolecules that are related in structure to nature’s light harvesting molecules. The implementation of NCPU-TTA in single-threshold cells has been predicted to theoretically increase the maximum efficiency limit from 30% to 40%. Of particular interest to the author is the homomolecular (homo-NCPU-TTA) scheme, where the sensitizer and emitter, are one and the same molecules (dual S-E). However, before homo-NCPU-TTA can be used in DSSCs, there are many potential problems that need to be investigated and overcome. Thus, the goal of this thesis is to contribute to a fundamental understanding of conditions needed to implement homo-NCPU-TTA in DSSCs. This thesis focuses on problems related to excited state energy losses of MPs to other DSSC components and problems related to controlling distance and mutual orientation of MPs in the solid state, both of which crucially affect NCPU-TTA efficiencies. First, work on energy loss by collisional quenching of a model MP’s excited states with iodide is presented. Iodide is a component of a commonly used redox shuttle in DSSCs and is a known effective fluorescence quencher. Iodide quenching of the S1 state can contribute to light to electrical conversion inefficiencies in MP-based DSSCs, while quenching of the S2 state indicates that there is no advantage of adding the MP up-converter as additional component in DSSCs. Investigation of the photophysical interactions between the S1 and S2 excited states of a model MP with iodide ions showed significant quenching of S1 and a small but noticeable quenching of S2 fluorescence. The minimal iodide quenching of the S2 state indicates that no significant loss of efficiency will be introduced by using iodide as component of the electrolyte system in NCPU-TTA enhanced solar cells. Second, minimization of energy losses arising from the mismatch between the photoanode band gap and upconverted photons of the dye is explored. This thesis approached this problem by tuning the band gap of a semiconductor such that its conduction band energy lies between those of the upconverted and the absorbing states of the model dual S-E, zinc tetraphenylporphyrin (ZnTPP), – e.g. between S2 and S1 of ZnTPP. The band gap (Eg) of TiO2, a common photoanode in DSSCs, lies below the S1 and S2 states of ZnTPP. It was expected that ZnTPP will show competing up-conversion (UC) and electron injection processes on TiO2 as photoanode. On the other hand, the Eg of ZrO2 lies above ZnTPP’s S2 state and only UC was expected. The introduction of ZrO2 as a defect in TiO2 allowed for tuning of its band gap to lie between the ZnTPP’s upconverted and prompt S1 excited states. Optimum electron injection efficiency from the ZnTPP S2 state was expected for this system. Contrary to expectations, UC was observed for pure and mixed metal oxide films, regardless of the relative energies of the states. Measurements in this study suggested that ZnTPP was heavily aggregated on all metal oxide surfaces and contributed significantly to the observed UC. Thus, controlling MP dye aggregation was deemed important and explored in the third and fourth projects of this thesis. Langmuir-Blodgett (LB) deposition was used as a facile means of preparing ordered MP assemblies in thin films. The LB technique requires molecules to be amphiphilic to form stable monolayers. Since ZnTPP is hydrophobic, it was synthetically modified to make it surface-active. The resulting compound, referred to as ZnDATPP, exhibited typical solution phase Soret and Q-bands as well as efficient homo-NCPU-TTA observed with ZnTPP. Langmuir films of ZnDATPP showed faceted domains indicative of rigid and crystalline or well-ordered structures. LB films prepared from these monolayers showed UV-Vis spectral shifts consistent with J-aggregation, which was the desired aggregation pattern as studies have shown that J-aggregates can act as energy funnels in photosensitization applications. However, NCPU-TTA was not observed in J-aggregated ZnDATPP films despite the controlled aggregation and close proximity of the molecules. This suggested that both appropriate spatial separation and orientation of MPs were necessary requirements for efficient solid-state homo- NCPU-TTA. In search of the appropriate distance and orientation of MPs, the reticular synthesis strategy of judiciously assembling metal-metalloporphyrin frameworks (MMPFs) was used. MMPFs were made from zinc ions as nodes, and Zn (II) tetrakis(4-carboxyphenyl)porphyrin (ZnTCPP) as a dual S-E strut. This MMPF was referred to as ZMP in this work. In addition, bipyridine (BPY) was used as molecular pillar of ZMP to form a MMPF referred to as ZMPB. Incorporation of these MMPFs into the solid state followed two approaches. First, thin films of MMPFs were prepared using layer-by-layer (LBL) growth. Second, bulk powder MMPFs were prepared and dispersed in polyvinyl alcohol (PVA) films. Spectroscopic and pXRD data suggested that both MMPFs satisfy Dexter ET distances and assumed approximate face-to-face stacking (H-aggregates). Results of prompt and UC emission measurements suggested that BPY reasonably quenched the S2 states of ZMPB. ZMP dispersed in PVA, on the other hand, showed measurable UC emission even in aerated conditions. Observation of UC in only one of the MMPFs in this work, despite both of them having small intermolecular distances and adapting similar stacking patterns, points to a complex interplay of processes that need to be investigated further. Observation of homo-NCPU-TTA in MPs is inherently difficult due to the low quantum yields of the singlet excited states generated from TTA and to the sensitivity of the triplet states to oxygen. Thus, detecting homo-NCPU-TTA for the first time in solid state (and in aerated conditions) shows the great promise of MMPFs as dual S-E molecules and opens them up for further exploration. There remain additional challenges that must be overcome before homo-NCPU-TTA can realistically be implemented in real devices. These are given in Chapter 7. To address some of these challenges, suggestions for future work are also given. These will hopefully serve as guidelines for studies to further improve our understanding of the conditions needed for homo-NCPU-TTA or NCPU-TTA, in general, to be incorporated in DSSCs