As this was an extended abstract with figures, it can not be viewed completely in this database. Introduction Bulk-heterojunction (BHJ) solar cells are composed of a nanoscale co-continuous composite of donor and acceptor phases, facilitating exciton dissociation. A conjugated, light-excitable polymer is most often used as an electron donor, and fullerene derivatives are the most widespread type of electron acceptor due to their high electron affinity and ability to transport charge [1]. An important advantage of such a system is that it can be casted from solvent, facilitating processing. Post-production annealing of such polymer:fullerene bulk heterojunction solar cells is vitally important, not only for fine-tuning the morphology and thus increasing the efficiency, but also for retaining the desired morphology during long-term operation [2-3]. However, knowing the optimal conditions for annealing temperatures and times requires knowledge about thermal transition temperatures and annealing kinetics of the blend systems. Using advanced fast-scanning thermal analysis techniques, the formation of nuclei and growth of crystals during heating or cooling can be reduced or avoided, allowing for the study of the crystallization processed during annealing. In this study, non-isothermal and isothermal crystallization kinetics of the P3HT:PCBM (poly(3-hexyl thiophene: [6,6] -phenyl C61 - butyric acid methyl ester) were studied by Rapid Heating Cooling Calorimetry (RHC) [4] and Fast Scanning Differential Chip Calorimetry (FSDCC) [5]. Methodology P3HT (Merck, Mw=35 000 g mol-1, Mw/Mn = 1.8; regioregularity greater than 98.5%) is mixed with PCBM (Solenne) in a 1:1 ratio and dissolved in chlorobenzene (CB) at a concentration of about 1-2.5 wt %, stirring overnight at 50°C. The solutions were deposited by drop-casting on large glass plates in a glove box under a nitrogen atmosphere to form films with a thickness of 1 µm. After drying in nitrogen atmosphere at room temperature for 50 hours to remove the residual solvent, the remaining solid films were scratched off the glass substrates and collected as a powder for RHC and FSDCC measurements. RHC experiments were performed on a prototype instrument made available by TA Instruments equipped with a liquid nitrogen cooling and specifically designed for operation at high scanning rates. Sample masses in the order of 250 - 300 µg were used in aluminium RHC crucibles. This allows for heating and cooling rates in the order of 2000 K/min. FSDCC was performed on a prototype Fast-Scanning Differential Chip Calorimeter (Rostock Univeristy, Germany) using Xensor Integration XI39399 chips, with an active area of 100 µm by 100 µm, having two on-chip heaters and 6 thermopiles for accurate temperature measurements. The whole chip calorimeter is submerged in a liquid nitrogen vessel, providing effective fast cooling. The highest rates achieved so far with this setup are in the order of 106 Ks-1, for sensor chips with a heated area of about 10 µm by 10 µm. The rates used in this work are limited to 5000 K.s-1. For both devices thermal annealing from the melt and from the glass were simulated by a temperature program as illustrated in figure 1, were the thermal transitions seen during the heating runs after isothermal treatment were used for analysis. Figure 1: Temperature program used to simulate thermal annealing, both from the glass and from the melt. The heating runs after the isothermal segments were used for data analysis. Results & Discussion Controlled fast heating and cooling are imperative for an isothermal crystallization study of rapidly crystallising systems such as P3HT:PCBM.. The fast heating and cooling capacity of advanced fast-scanning thermal analysis techniques enables the reduction of nuclei formation and crystal growth during heating or cooling, by restricting the time available for the blend to crystallise by quickly passing through the crystallisation temperature window that extends from the melting temperature down to the glass transition temperature. The thermal annealing process in P3HT:PCBM blends was studied first by performing RHC isothermal annealing experiments at 110°C for a wide range of isothermal times. All measurements were performed with a heating rate of 500 K.min-1 and ballistic cooling. Ballistic cooling reaches more than 1500 K.min-1 at the beginning of the cooling, which drops down to 750 K.min-1 around 60 °C. Figure 2 shows the subsequent heating curves of the system for increasing annealing times. It is clear that the cold crystallization enthalpies are decreasing and melting enthalpies are increasing by longer annealing times. Besides, by longer annealing at the isothermal annealing temperature, the step in Tg is getting smaller since less amorphous fraction is remaining in the material, and the Tg shifts to higher temperatures. As discussed in previous work, the crystallization of P3HT (having a lower Tg) from the blend is enriching the fraction of PCBM (having a higher Tg) in the remaining amorphous phase, leading to a higher Tg in the subsequent heating. Of course, the Tg is further increased by the close presence of crystalline domains, as is the case for semi-crystalline homopolymers. Another important remark is the clear difference between isothermal crystallisation from the melt and from the glass (without or with cooling before annealing). The crystallization rate is clearly higher when annealing from the glass [6]. Figure 2: RHC heating cycles at 500 K.min-1 for a P3HT:PCBM 1:1 blend after isothermal annealing both from the melt (left) and from the glass (right). This difference between the two possible paths of thermal annealing can be explained by a significant difference between the two in the amount of crystal nuclei. This suggests that the rates available in RHC are not sufficient to avoid most of the nucleation. When similar tests are conducted using FSDCC, using heating and cooling rates of 5000 K.s-1, cold crystallization disappears (see figure 3). There is also no longer a significant difference between isothermal crystallization from the melt and from the glass. Based on these results, it can be concluded that nucleation is mostly avoided by using heating and cooling rates available in FSDCC, and the crystallinity seen results only from the isothermal treatment. This allows for a reliable investigation of the crystallization processes during isothermal annealing, both from the glass and from the melt. Figure 3: FSDCC heating cycles at 5000 K.s-1 for a P3HT:PCBM 1:1 blend after isothermal annealing, both from the melt (left) and from the glass (right). Conclusion The isothermal crystallisation at the annealing temperature of 110 °C was investigated for 1:1 P3HT:PCBM using the advanced fast-scanning thermal analysis techniques RHC and FSDCC. RHC allowed for an in depth study of the isothermal crystallisation of these systems, but does not avoid most of the nucleation, leading to a difference in isothermal crystallisation rates for annealing treatments from either the glass or the melt. This difference, as well as cold crystallisation, disappears when FSDCC is used, owing to the much higher rates of heating and cooling. Because of this it is now possible to obtain only information about the isothermal crystallisation, without unwanted effect of the preceding thermal treatment. References 1. Thompson B.C. and Frechet J.M.J., Angewandte Chemie-International Edition, 2008. 47(1): p. 58. 2. Erb T., Zhokhavets U., Gobsch G., Raleva S., Stuhn B., Schilinsky P., Waldauf C., and Brabec C.J., Advanced Functional Materials, 2005. 15(7): p. 1193. 3. Hoppe H. and Sariciftci N.S., Journal of Materials Chemistry, 2006. 16(1): p. 45. 4. Danley R.L., Caulfield P.A., and Aubuchon S.R., American Laboratory, 2008. 40(1): p. 9. 5. Minakov A.A., van Herwaarden A.W., Wien W., Wurm A., and Schick C., Thermochimica Acta, 2007. 461(1-2): p. 96. 6. Demir F., Van den Brande N., Van Mele B., Bertho S., Vanderzande D., Manca J., and Van Assche G., Journal of Thermal Analysis and Calorimetry, 2011. 105(3): p.845.