2010 GCEP Report Organic Solar Cells

In this project, we aim at combining the molecular design and device fabrication expertise of Bao, theoretical simulation expertise of Aspuru-Guzik, structural characterization expertise of Toney, and the large distributed computing power of IBM’s World Community Grid (WCG) to rationally design organic semiconductors for solar cells from a completely new angle. Instead of molecular design from intuition, we will combine powerful theoretical tools and various characterization techniques to develop an inverse rational design methodology for novel materials. By doing so, we see a feasible path towards breakthroughs in performance. Such a massive amount of computing resources has not been previously applied to atomic-scale modeling problems in material sciences. For organic semiconductors to find ubiquitous electronics applications, the development of new materials with high mobility and air stability is critical. Achievement 1: Designed and predicted high charge carrier mobility compounds, realized high charge carrier mobility experimentally: Despite the versatility of carbon, exploratory chemical synthesis in the vast chemical space can be hindered by synthetic and characterization difficulties. We show that in silico screening of organic semiconductor materials can lead to the discovery of a new highperformance semiconductor. This work involves the theoretical screening of eight novel derivatives of the dinaphtho[2,3-b:2’,3’-f]thieno[3,2-b]thiophene semiconductor which has a maximum mobility of 8.3 cmVs. Based on the charge transport parameters and the predicted crystal structures, we identified a novel compound expected to demonstrate a two-fold improvement in mobility over the parent molecule. Synthetic and electrical characterisation of the compound is reported with single crystal field-effect transistors, showing a remarkable mobility of 13.7 cmVs. This is one of the very few organic semiconductors with mobility greater than 10 cmVs reported to date. More importantly, this is a significant step towards rationally design organic semiconductors for efficient solar cells. For next year program, we plan to extend the theoretical prediction to exciton diffusion length prediction. Achievement 2: developed a solution processing technique that generated strained organic semiconductor lattice for the first time For organic semiconductors (OSCs), the molecular packing determines the charge transport of the resulting devices. It is desirable to control the molecular packing of small molecular OSCs through facile processing methods in order to tune the electrical properties of OSC devices. We describe the alteration of the 5,12-bis(triisopropylsilylethynyl) pentacene (TIPSE-pentacene) molecular packing by changing the conditions used during solution processing deposition. Our method deposits TIPSE-pentacene thin film in a non-equilibrium state, and the π-π stacking distance between the molecules can be tuned between 3.08 Å to 3.33 Å as a function of these conditions, which in turn significantly affects the electronic properties of TIPSE-pentacene. The charge carrier mobility was increased from 0.3 cm2/Vs to a record high mobility for TIPSE-pentacene at 4.6 cm/Vs. Control of the molecular packing using processing conditions will allow for the development of high performance OSC devices beyond traditional synthetic methods. This is the first time that a π-π stacking distance less than 3.2 Å is realized. Since the charge transport is exponentially dependent on the distance between molecules, such a strained structure is expected to significantly increase the charge transport. This is likely to significantly impact the exciton transport as well, which is the subject of year 2 investigation. Achievement 3: developed and deployed the computational high-throughput screening on the WCG In the spirit of the proof-of-principle studies discussed above we have devised the machinery for the large-scale characterization of OSC candidates. We have developed a molecule generator which has produced a primary library of 10,000,000 molecular motifs of potential interest. The quantum chemistry program package Q-Chem was ported to the WCG and we have so far characterized over 1,600,000 million of these oligomer sequences in more than 22,000,000 first-principles calculations. These calculations utilized 3,000 years of donated CPU time. We have set up the server and storage (including a 90TB hard drive pod which we custom built) for the vast influx of results from the WCG. The setup includes a software infrastructure for process automation. Our data collection is used to build up a massive reference database on organic electronics with an emphasis on photovoltaic applications. We have introduced a preliminary empirical calibration of the computed results to bridge the gap between theory and experiment. A first analysis of the current results points to about 25,000 systems with very favorably aligned energetic levels, ideally suited for high-performance OPVs with power conversion efficiencies of 10% or more. Parallel to the first-principles calculations we have been exploring the use of cheminformatics descriptors and ideas from machine learning, pattern recognition, and drug discovery to rapidly gauge the quality of candidates. We have successfully devised initial empirical models for the relevant performance parameters. In the following year we will extend our data mining and analysis capabilities and we will screen for other relevant features and properties in the materials candidates.