Organic photovoltaic materials have long-term promise for large area devices on flexible substrates produced by low-cost processes such as inkjet printing. Control of order and morphology at the nanometre scale during processing steps such as thermal annealing is crucial for optimised device efficiency. In this study, the structural changes during in situ annealing were characterised by time-resolved grazing incidence X-ray diffraction.
Photovoltaic panels based on crystalline semiconductors such as silicon are relatively expensive to produce and process for large-area applications. Organic photovoltaic materials offer cheap, large area processing on flexible substrates and are particularly attractive for their versatility. Their applications include diffuse lighting and everyday objects such as rucksacks, which are already being produced with integrated organic solar panels for charging hand-held devices. Organic photovoltaic efficiencies have increased dramatically over the last decade, promising large-scale, wider applications . Most organic photovoltaic devices are based on thin films comprising an electron accepting component (such as a fullerene derivative) and an electron donating component (usually a conjugated polymer) between two electronically different electrodes. Excitons (electron-hole pairs), generated by absorption of light, have a diffusion length of ~10 nm and hence need to be generated close to a donor-acceptor interface, where the free energy gained on charge transfer from the photoexcited donor molecule to an acceptor molecule can overcome the electrostatic attraction between electron and hole (Figure 1a). A second requirement is a continuous network of donor and acceptor paths for the carriers to percolate to their respective electrode (Figure 1b). These two requirements involve a delicate balance of intimate mixing of the two components, to ensure exciton dissociation, and local segregation to form the bicontinuous network for carrier conduction to the electrodes.
Figure 1. a) Schematic of the energy lineup in an organic photovoltaic: charge separation as electrons are transferred from the LUMO (Lowest Unoccupied Molecular Orbital, corresponding to the conduction band edge of an inorganic semiconductor) of the donor to that of the acceptor; b) the donor (dark blue) and acceptor (pale blue) bicontinuous percolated paths of the carriers to the electrodes.
The best studied material system is the combination of regioregular poly-3-hexylthiophene (P3HT) with the methanofullerene, phenyl C61 butyric acid ester (PCBM). A wide range of experimental studies have demonstrated that variations in polymer regioregularity, blend film composition, solvent, deposition conditions, and post deposition treatments such as thermal and vapour annealing lead to variations in organic photovoltaic device performance, apparently as a result of variations in blend film morphology [2-4]. Such studies also demonstrate that improvements in solar cell performance are usually associated with increases in the polymer crystallinity. An AFM image of the P3HT/PCBM blend is shown in Figure 2, in which the bicontinuous nature is clearly seen.
Figure 2. a) The molecular structure of P3HT and PCBM. b) An atomic force microscopy image (1×1 µm) of the nanoscale morphology of P3HT:PCBM. The nanofibrillar P3HT can be seen, the darker regions consisting of amorphous P3HT and PCBM.
Grazing incidence X-ray diffraction (GIXRD) is a prime tool for investigating the structure of thin films. We have used it to study the changes in P3HT:PCBM film structure during a thermal annealing process, which is used in solar cell optimisation. When P3HT:PCBM solar cells are thermally annealed, they are usually put on a pre-heated hot plate at a temperature ranging from 120°C to 155°C for 4 to 120 minutes. After thermal annealing, devices are moved to a metallic surface at room temperature for fast cooling. These conditions were simulated in our experiments by optimisation of temperature control for rapid heating to 140°C with <1°C overshoot. The cooling transient is slower than that for heating due to the absence of a cooling stage.
Figure 3 shows the GIXRD pattern for P3HT:PCBM (50 wt% PCBM) taken with an area detector on beamline BM28 (XMaS). By monitoring the position and the full width at half maximum (FWHM) of the (h00) P3HT peaks during annealing, changes in the lattice spacing and ordering of P3HT along the alkyl stacking direction can be tracked (Figure 4). The thermal expansion during heating and cooling is extremely anisotropic (thermal expansion coefficients deduced from the heating ramp is αT ≈ 5.3×10-4 K-1 and 4.3×10-5 K-1 along the  direction and  directions respectively). The P3HT domain size, arising from the nanofibril width for the (100) peak, increases rapidly during the anneal and stabilises within around 3 minutes at the anneal temperature. Parallel device measurements show that maximum increase in photocurrent generation also occurs within these first few minutes of annealing. Maximum performance is achieved on a longer time scale than P3HT crystallisation, and may be associated with fullerene diffusion. Further experiments are underway to investigate the effect of the substrate on the anneal behaviour and to probe the depth dependence of the structure.
Figure 3. GIXRD images before and after annealing of the P3HT/PCBM blend (SiO2, Pedot, P3HT:PCBM).
Figure 4. The P3HT lattice constant and domain size as a function of time during in situ annealing, measured with a time step of 15 seconds. The domain size has been determined with the Debye-Scherrer equation for the (100) peak. The red curve shows the sample temperature.
Principal publication and authors
T. Agostinelli (a), S. Lilliu (b), J.G. Labram (a), M. Campoy-Quiles (a), M. Hampton (b), E. Pires (b), J. Rawle (c), O. Bikondoa (d), D.D.C. Bradley (a), T.D. Anthopoulos (a), J. Nelson (a), and J.E. Macdonald (b), Real-Time Investigation of Crystallization and Phase-Segregation Dynamics in P3HT:PCBM Solar Cells During Thermal Annealing, Advanced Functional Materials 21, 1701-1708 (2011); DOI 10.1002/adfm.201002076.
(a) Imperial College, London (UK)
(b) Cardiff University, Cardiff (UK)
(c) Diamond Light Source (UK)
(d) XMaS, ESRF (France)
 G. Dennler, M.C. Scharber, C.J. Brabec, Adv. Mat. 21, 1323 (2009).
 W.L. Ma, C.Y. Yang, X. Gong, K. Lee, A.J. Heeger, Adv. Funct. Mat. 15, 1617 (2005).
 Y. Kim, S.A. Choulis, J. Nelson, D.D.C. Bradley, S. Cook, J.R. Durrant, J. Mater. Sci. 40, 1371 (2005).
 Y. Kim, S. Cook, S.M. Tuladhar, S.A. Choulis, J. Nelson, J.R. Durrant, D.D.C. Bradley, M. Giles, I. McCulloch, C.S. Ha, M. Ree, Nature Materials 5, 197 (2006).