Michael McGehee, Department of Materials Science and Engineering, Stanford University

This project will involve making efficient photovoltaic (PV) cells with semiconducting polymers that could be deposited in reel-to-reel coaters. Careful analysis and optimization of each process that occurs in bulk heterojunction PV cells will be carried out and devices based on ordered interpenetrating networks of organic and inorganic semiconductors will be created. Specifically this research will lead to devices that will efficiently split excitons and carry charge to electrodes, that will have improved packing of the molecules in the organic semiconductor to enhance its ability to carry charge, and that will have a modified organic-inorganic interface to prevent recombination of electrons and holes. It is anticipated that charge recombination in the cells will be almost completely eliminated and energy conversion efficiencies in the range of 10-15% will be obtained.

Figure 1: Schematic energy diagram of a bulk heterojunction PV cell

Currently the best commercially available PV cells are made of crystalline silicon and have an energy conversion efficiency of 12%. The cost of these cells is $3 per Watt of power generated under solar AM 1.5 conditions. These costs need to be reduced by an order of magnitude to around $0.3 per Watt for PV cells to be competitive with other energy generation systems and be manufactured on a large scale. A revolutionary breakthrough in reducing the costs of PV cells may be achieved if the semiconductor were deposited from solution onto large flexible substrates in reel-to-reel coating machines similar to those used to make food packaging. Manufacturing costs would be much lower because reel-to-reel coaters use very little energy and have an exceptionally high throughput. Installation costs would be lower because lightweight flexible PV cells could be handled more easily than heavy silicon panels.

The processes involved in operating a bulk heterojunction PV device are shown in Figure 1. To optimize performance of these cells, the desirable processes (1. light absorption, 2. exciton diffusion, 3. forward electron transfer, and 4. charge transport) should be maximized, while the undesirable recombination processes (5. geminate recombination and 6. back electron transfer) should be limited. This can be achieved by improving charge carrier mobility and slowing down the rate of back electron transfer so that photogenerated charge carriers can escape from the film before recombination occurs, while maintaining a thick enough film to allow most of the light to be absorbed.

In this study, well-ordered interpenetrating networks of organic and inorganic semiconductors will be made to create bulk heterojunction PV devices instead of the disordered blends that have been used in the past. In these structures, electrons and holes will have straight pathways to electrodes. Consequently these charge carriers will travel the shortest possible distance and will never be trapped at dead ends. To make the interpenetrating nanostructures that are needed, nanoporous films or nanowire brushes will be fabricated using an inorganic semiconductor and the pores or brush will be filled in with a semiconducting polymer. There are many combinations of inorganic semiconductors and semiconducting polymers that have suitable energy levels for light absorption and electron transfer. The approach used requires addressing three different issues:

• Patterning of nanostructured films: The techniques used to pattern the inorganic films will include titania/block copolymer co-assembly (see Figure 2), growth of nanowires off of arrays of catalysts to create a nanowire brush, and stamping of sol-gel oxide precursors. In each case, the size of the polymer region will be approximately 20-nm so that all of the excitons in the polymer can diffuse to the polymer/inorganic semiconductor interface. The films will be approximately 300-nm thick so that they can effectively absorb light.

Only two layers of unit cells are shown. A typical film has more than 100 layers. The hydrophobic blocks are shown in green. The hydrophilic blocks are shown in blue. After a film is made, pores are created by burning out the block copolymer.

• Improving hole transport in organic semiconductors: Hole mobilities in pore structures may change dramatically from values in pristine films and may directly impact the ability to extract the holes in the polymer out of the PV cells before they recombine with electrons in the electron-accepting semiconductor. Several approaches will be used for filling nanoporous films and nanowire brushes with organic semiconductors. These include: modification of the chemical groups on the polymer chains and the inorganic semiconductor surface to optimize the interactions between the organic and inorganic materials; use of techniques to electropolymerize conjugated monomers directly in the pores which may lead to higher absorption cross-sections and better contact for improved charge transport; and use of discotic liquid crystals. With each approach, charge carrier mobility measurements will be made, and the molecular packing will be characterized in order to determine in a rational way, how to maximize hole mobility.
• Optimizing electron transfer: Electron transfer plays a critical role in bulk heterojunction PV cells. The forward transfer from the polymer to the electron accepting material must occur before the electron and hole in the exciton undergo geminate recombination, and back electron transfer from the electron acceptor to the polymer must be inhibited. Photoluminescence quenching will be used to determine the effectiveness of forward electron transfer for all of the organic-inorganic combinations developed and time-resolved photoinduced absorption to measure the rate of back electron transfer. Retardation of back electron transfer will be achieved by coating the inorganic surface with a thin, wide-energy-gap coating with appropriate energy levels.

By following these approaches, it is anticipated that the small prototype devices developed in this effort may be scaled up to allow manufacture of stable low-cost PV cells using reel-to-reel coaters with energy conversion efficiencies of 10-15% at a cost of around an order of magnitude cheaper than presently available.