Low fill factors can be caused by significant series resistance losses related to contact and interface layers. Also recombination within the active layer, especially for large thicknesses can decrease the fill factor. Can you tell us your device architecture including the layer thicknesses?
By the way, I think a short-circuit current desnity of 13 mA/cm2 is not realistic for the P3HT:PC60BM system. You should increase the accurancy of your JV measurements.
13 mA/cm2 is too high to and not achieveable with P3HT:PCBM. So I believe you make some errors during your measurement that lead to overestimation of the Jsc. A possible error source is a not well calibrated solar simulator which provides intensities larger than 100 mW/cm2 and/or high spectral mismatch. Another prominent error source is the determination of the active area of the cells, usually defined by the overlap of the ITO and metal electrodes. This area has to be measured carefully for example using an optical microscope.
Do you have an EQE setup in your group? An easy method to check the accurancy of your JV measurements is to measure an EQE spectrum of your device and integrate it with the AM1.5G standard spectrum to obtain the expectable photocurrent density.
Under the assumption it's 0.62V you achieved, and 13 mA/cm2 with 47% FF.
0.62V is still on the high side for a P3HT / PCBM efficient device so you probably still have some morphology issues, which would correlate with low FF
the first things I'd check would be :
- Quality of P3HT (high regioregularity > 95%, and Mn/Mw)
- Annealing process of P3HT/PCBM
for the 13mA/cm2, that's simply not possible and looking at the 2 other factors I guess you're not anywhere close (probably more on the 7-8mA/cm2). this could come from a too high intensity or none matching AM1.5 spectrum, or from an illumination of your full substrate (no mask) and possibly contribution even from outside the area of the overlap of the 2 electrodes.