Bryon W. Larson

Tuning the driving force for exciton dissociation in single-walled carbon nanotube heterojunctions

Understanding the kinetics and energetics of interfacial electron transfer in molecular systems is crucial for the development of a broad array of technologies, including photovoltaics, solar fuel systems and energy storage. The Marcus formulation for electron transfer relates the thermodynamic driving force and reorganization energy for charge transfer between a given donor/acceptor pair to the kinetics and yield of electron transfer. Here we investigated the influence of the thermodynamic driving force for photoinduced electron transfer (PET) between single-walled carbon nanotubes (SWCNTs) and fullerene derivatives by employing time-resolved microwave conductivity as a sensitive probe of interfacial exciton dissociation. For the first time, we observed the Marcus inverted region (in which driving force exceeds reorganization energy) and quantified the reorganization energy for PET for a model SWCNT/acceptor system. The small reorganization energies (about 130 meV, most of which probably arises from the fullerene acceptors) are beneficial in minimizing energy loss in photoconversion schemes.

Substituent effects in a series of 1,7-C60(RF)2 compounds (RF = CF3, C2F5, n-C3F7, i-C3F7, n-C4F9, s-C4F9, n-C8F17): electron affinities, reduction potentials and E(LUMO) values are not always correlated

A series of seven structurally-similar compounds with different pairs of RF groups were prepared, characterized spectroscopically, and studied by electrochemical methods (cyclic and square-wave voltammetry), low-temperature anion photoelectron spectroscopy, and DFT calculations (five of the compounds are reported here for the first time). This is the first time that a set of seven RF groups have been compared with respect to their relative effects on E1/2(0/−), electron affinity (EA), and the DFT-calculated LUMO energy. The compounds, 1,7-C60(RF)2 (RF = CF3, C2F5, i-C3F7, n-C3F7, s-C4F9, n-C4F9 and n-C8F21), were found to have statistically different electron affinities (EA), at the ±10 meV level of uncertainty, but virtually identical first reduction potentials, at the ±10 mV level of uncertainty. The lack of a correlation between EA and E1/2(0/−), and between E(LUMO) and E1/2(0/−), for such similar compounds is unprecedented and suggests that explanations for differences in figures of merit for materials and/or devices that are based on equating easily measurable E1/2(0/−) values with EAs or E(LUMO) values should be viewed with caution. The solubilities of the seven compounds in toluene varied by nearly a factor of six, but in an unpredictable way, with the C2F5 and s-C4F9 compounds being the most soluble and the i-C3F7 compound being the least soluble. The effects of the different RF groups on EAs, E(LUMO) values, and solubilities should help fluorine chemists choose the right RF group to design new materials with improved morphological, electronic, optical, and/or magnetic properties.

Scalable slot-die coating of high performance perovskite solar cells

Perovskite based photovoltaic devices hold the promise to greatly reduce the cost of solar energy production; however, this potential depends greatly on the ability to deposit perovskite active layers using large scale deposition methods such as slot-die coating without sacrificing efficiency. Using a perovskite precursor ink with long wet-film processing window, we demonstrate efficient perovskite solar cells based on slot-die coated perovskite layer. We found almost no difference in the photophysical and structural details of perovskite films that were deposited by spin coating to films deposited by slot-die coating. We explored various slot-die coating parameters to determine their effect on the performance of the device metrics. In addition to slot-die coating, we demonstrate the versatility of this wide wet-film processing window by fabricating perovskite solar cells with active layers deposited by spin coating, blade coating, and spray coating that all exhibited similar performance.

Photobleaching dynamics in small molecule vs. polymer organic photovoltaic blends with 1,7-bis-trifluoromethylfullerene

Two organic photovoltaic (OPV) donor materials (one polymer and one small molecule) are synthesized from the same constituent building blocks, namely thiophene units, cyclopentathiophene dione (CTD), and cyclopentadithiophene (CPDT). Photobleaching dynamics of these donor materials are then studied under white light illumination in air with blends of PC70BM and the bis-trifluoromethylfullerene 1,7-C60(CF3)2. For both the polymer and small molecule blends, C60(CF3)2 stabilizes the initial rate of photobleaching by a factor of 15 relative to PC70BM. However, once the small molecule:C60(CF3)2 blend bleaches to ∼80% of its initial optical density, the rate of photobleaching dramatically accelerates, which is not observed in the analogous polymer blend. We probe that phenomenon using time-resolved photoluminescence (TRPL) to measure PL quenching efficiencies at defined intervals during the photobleaching experiments. The data indicates the small molecule donor and C60(CF3)2 acceptor significantly de-mix with time, after which the blend begins to bleach at approximately the same rate as the neat donor sample. The work suggests that perfluoroalkylfullerenes have great potential to stabilize certain OPV active layers toward photodegradation, provided their morphology is stable.

Improving photoconductance of fluorinated donors with fluorinated acceptors

This work investigates the influence of fluorination of both donor and acceptor materials on the generation of free charge carriers in small molecule donor/fullerene acceptor BHJ OPV active layers. A fluorinated and non-fluorinated small molecule analogue were synthesized and their optoelectronic properties characterized. The intrinsic photoconductance of blends of these small molecule donors was investigated using time-resolved microwave conductivity. Blends of the two donor molecules with a traditional non-fluorinated fullerene (PC70BM) as well as a fluorinated fullerene (C60(CF3)2-1) were investigated using 5% and 50% fullerene loading. We demonstrate for the first time that photoconductance in a 50:50 donor:acceptor BHJ blend using a fluorinated fullerene can actually be improved relative to a traditional non-fluorinated fullerene by fluorinating the donor molecule as well.

RAMP-ing the discovery of high-performance organic photovoltaic materials (Conference Presentation)

Organic semiconductors (OSCs) have found numerous applications in thin film electronic devices, including displays, sensors, lighting, and solar cells. OSCs offer the advantage that they can be modified by synthetic chemists to fulfill specific needs, and that they are not limited to being made from single elements or compositional alloys. For example, organic photovoltaics (OPVs) have reached nearly 13% power conversion efficiency (PCE) in small area devices using traditional polymer-fullerene blends, yet non-polymer and non-fullerene composites are now also showing PCEs above 10%. The flexibility offered by synthetic manipulation also presents a challenge: progress towards the discovery of next-generation, high-performance materials can be stifled by the bottleneck of device optimization through process engineering. High-throughput screening techniques that provide high fidelity performance metrics can circumvent this problem and will become important tools to accelerate material development. Here, we introduce such a tool, based on unique microwave conductivity capabilities, and illustrate a cost- and time-effective approach to evaluate the potential of promising new materials, independent of their bulk (i.e., device optimized) thin-film performance. We demonstrate the power of this approach, by correlating figures of merit from our screening tool to the OPV device performance for a library of current state-of-the-art OSCs, based on both polymer and small molecule chemical structure motifs. Finally, in the context of polymeric materials, we highlight the sensitivity of the screening process to physicochemical properties (e.g., molecular weight) and suggest that our tool can be employed for batch-to-batch quality control.