PMC

(a) J-V curve of device A, B, and B2 (without solvent annealing). Much higher photocurrent density in reverse bias is observed for device B while the increase of photocurrent density is much slower in device B2. Inset: Relative energy diagram of all the materials used in this manuscript. (b) J-V curve of device B3 (after solvent annealing, hole-dominant device, with MoO₃(5nm)/Al(70nm) electrode) under light exposure. A much higher photocurrent density is observed under reverse bias (while MoO₃/Al electrode is positively biased).

Topography (left) and phase images (right) of device A, B2, and B in 1´m are shown in (a), (b), and (c), respectively. The phase images of devices A and B2 show filbrillar structure while device B represents an island-like surface morphology. The absence of fibrillar structure in the surface morphology of device B is a result of adding NPs coupled with solvent annealing process.

(a) Device structure and chemical structure of PMDTC ligand (not to scale). (b) The J-V curve of devices A and B. Significant photocurrent increase is observed under reverse bias when CdTe NPs are blended into the P3HT:PCBM matrix. (c) The J-V curve of device B in logarithmic scale. Clear diode characteristic is shown by the dark current density curve. (d) Absorption spectrum of films in devices A, B, and C. PMDTC blended with P3HT:PCBM shows no obvious change in absorption. Inset: Absorption spectrum of CdTe NPs (with PMDTC capping ligands) and free PMDTC in DCB.

(a) TEM images of P3HT:PCBM/CdTe NPs (~3.1% (31mg/mL)) film. Scale bar in the major image is 200 nm and is 1´m for the inset. The solution was drop cast on a TEM grid for sample preparation. Highly condensed aggregates of CdTe NPs are observed and thread-like features are clearly visible in the inset. (b) EQE of device B under zero and reverse bias. Higher than 100% EQE is obtained under reverse bias, even at the absorption edge of CdTe NPs (~700 nm). (c) EQE of device A measured under the same conditions.