Organic solar cells are made of thin layers of interpenetrating structures from two different conducting organic materials and are increasingly popular because they are both potentially cheaper to make than those currently in use and can be "painted" or printed onto a variety of surfaces, including flexible films made from the same material as most soda bottles. A large community is exploring a number of promising material combinations (polymer/fullerene, polymer/inorganic, all-polymer, and dye-sensitized cells), seeking a deeper understanding of their fundamental structure, operation, and limitations. A team of researchers from North Carolina State University and the UK has now found, through microscopy and resonant scattering and reflectivity studies at ALS Beamlines 6.3.2 and 5.3.2, that the low rate of energy conversion in model all-polymer solar cells is caused by domains that are too large and interfaces that are not sharp enough. This insight will lead to new approaches to all-polymer device technology that will help realize the intrinsic potential of these materials.

Schematic of the layers in a two-polymer solar cell. An incoming photon from the sun creates an exciton (an electron–hole pair) that must travel to the interface, where charge separation can occur, producing a useful voltage.

Solution-processed organic solar cells are attracting substantial, world-wide attention due to their potential as a low-cost photovoltaic technology. The active thin layer in such solar cells has to be simultaneously thick enough (~150–200 nm) to absorb most of the light and have internal structures small enough (~10 nm) for that captured energy—known as an exciton—to be able to travel to the site of charge separation and electricity generation. One of the field's fundamental challenges is to create such structures in a controlled fashion, and advances often depend on trial-and-error methods. Furthermore, conventional tools are often unable to visualize the morphology or characterize details of the important interface between electron donor and acceptor materials where the charge separation of the exciton takes place. Substantial progress has been made in understanding all-polymer devices using advanced characterization tools at the ALS, including x-ray microscopy, resonant soft x-ray scattering, soft x-ray reflectivity, photoluminescence quenching, device characterization, and Monte Carlo studies.

For the morphology studies, model blends of the conjugated polymers PFB and F8BT were annealed at different temperatures. The resulting phase separation was studied using resonant soft x-ray scattering and scanning transmission x-ray microscopy. The data showed a wide variety of domain sizes with rather impure compositions. The dominant domain sizes were much larger than the exciton diffusion length for all processing conditions, including annealing at intermediate temperatures yielding maximum device efficiency. Intermixing within these large domains means that not all excitons are lost to recombination; however, this hierarchy of phase separation prevents the creation of "ideal" morphologies of pure domains ~10 nm in size by the casting and annealing approach. Thus, even in the morphology that maximized device performance, some excitons must travel too far and are lost, while some domains are still too intermixed, hindering charge separation. This contributes to the poor performance of solar-cell devices made from blends of these polymers and might be indicative of all polymer devices in general. Other all-polymer-based solar cells are being investigated to see if their low efficiencies are due to this same structural problem.

Pair distance distribution function for PFB/F8BT blends annealed as indicated. The zero crossing relates to the domain size and has been determined to be as follows: As spun, ~77 nm; 140°C, ~71 nm; 160°C, ~89 nm; 180°C: ~110 nm; 200°C, ~260 nm. In all cases, the domains are larger than the exciton diffusion length, which is ~10 nm, contributing to the relatively low power-conversion efficiency of all-polymer devices.

The influence of interface structure on the performance of devices was also investigated using PFB/F8BT bilayer heterojunctions. Resonant soft x-ray reflectivity provided quantitative information on the roughness of the interface as a function of annealing protocol. Only the cold laminated bilayers were sharp, and even mild annealing caused the interface to roughen. The measured roughness was an important input parameter for Monte Carlo simulations that were used to reproduce the experimental device current–voltage characteristics. Nonequilibrium, sharp interfaces are optimal for charge separation and optimal device performance. The challenge will now be to reliably create small, interpenetrating networks with sharp interfaces in order to produce high-efficiency devices. New fabrication methods will have to be found.

Current–voltage characteristics for devices that were cold laminated (black) or annealed at 140°C (red). Resonant soft x-ray reflectivity measured a significant increase in interfacial roughness from 0.68 nm for the cold laminated to 2.6 nm for the annealed device.

Cheng Wang (left) and Hongping Yan (right), two of the papers' authors, at the soft x-ray scattering beamline.

Research conducted by H. Yan, S. Swaraj, B. Watts, and H. Ade (North Carolina State University); C. Wang (ALS); J. Lüning (Stanford Synchrotron Radiation Lightsource); I. Hwang, C.R. McNeill, and N.C. Greenham (University of Cambridge, UK); and C. Groves (Durham University, UK).

Research funding: U.S. Department of Energy (DOE), Office of Basic Energy Sciences (BES); Engineering and Physical Sciences Research Council (UK); materials supplied by Cambridge Display Technology Ltd. Operation of the ALS is supported by DOE BES.

Publications about this research: S. Swaraj, C. Wang, H. Yan, B. Watts, J. Lüning, C.R. McNeill, and H. Ade, "Nanomorphology of bulk heterojunction photovoltaic thin films probed with resonant soft x-ray scattering," Nano Lett. 10, 2863 (2010); H. Yan, S. Swaraj, C. Wang, I. Hwang, N.C. Greenham, C. Groves, H. Ade, and C.R. McNeill, "Influence of annealing and interfacial roughness on the performance of bilayer donor/acceptor polymer photovoltaic devices," Adv. Funct. Mater. 20, 4329 (2010).

ALS Science Highlight #218

ALSNews Vol. 314