Phase separation analysis of bulk heterojunctions in small-molecule organic solar cells using zinc-phthalocyanine and C${}_{60}$

To achieve efficient organic solar cells, donor and acceptor molecules are mixed in the photoactive layer to form a so-called bulk heterojunction. Due to molecular interactions, a certain degree of phase separation between donor and acceptor domains arises, which is necessary to achieve efficient charge extraction within the absorber layer. However, the mechanism that induces the phase separation is not fully understood and gaining detailed information about the molecular arrangement within these blend layers is quite challenging. We show that grazing incidence x-ray diffraction, combined with variable angle spectroscopic ellipsometry is a suitable way to investigate the molecular structure of blend layers in detail, consisting of a mixture of zinc-phthalocyanine (ZnPc) and C${}_{60}$. The degree of phase separation within the blend layer is influenced by substrate heating during the co-evaporation of ZnPc and C${}_{60}$ and by a variation of the mixing ratio. The effect of different blend layer morphologies on optical and electrical device performance is investigated by solar cell characterization and mobility measurements. We find that the molecular arrangement of C${}_{60}$ provides the essential driving force for efficient phase separation. Whereas spherical C${}_{60}$ molecules are able to form crystalline domains when deposited at elevated substrate temperatures, no ZnPc crystallites are observed, although the planar ZnPc molecules are not randomly oriented but standing upright within its domains. Comparing specular and grazing incidence x-ray diffraction, we find that only the latter method is able to detect nanocrystalline C${}_{60}$ in thin films due to its polycrystalline nature and small sized nanocrystallites. Solar cell measurements show an increase in fill factor and external quantum efficiency signal for blends with enhanced phase separation, induced by higher substrate temperatures. However, grazing incidence x-ray diffraction measurements reveal that ZnPc and C${}_{60}$ already form separate domains in unheated ZnPc:C${}_{60}$ blends, which provide fill factors close to 50$\%$ in the corresponding solar cells.

Figure 1

(Top) Specular XRD scan (symmetric θ-2θ-scan mode) in Bragg-Brentano geometry (black line) and out-of-plane GIXRD scan (ω $=$ 0.20° fixed, green/bright line) of a 50-nm thick pristine C${}_{60}$ layer deposited on unheated glass substrate. (Bottom) Out-of-plane GIXRD scan of a 50-nm thick pristine ZnPc layer also deposited on an unheated glass substrate (blue line).

Figure 2

GIXRD pattern of 150-nm thick ZnPc:C${}_{60}$ (1:1) blend layers deposited at different substrate temperatures on glass/5-nm C${}_{60}$ with the corresponding AFM micrographs and resulting AFM roughness. The GIXRD pattern of a 50-nm thick pristine ZnPc and of a 50-nm thick pristine C${}_{60}$ layer is inserted to correlate the Bragg reflections of the ZnPc:C${}_{60}$ blend layers.

Figure 3

GIXRD pattern of 150-nm-thick ZnPc:C${}_{60}$ blend layers deposited at a substrate temperature of 140 °C on glass/5-nm C${}_{60}$ for different blend ratios of 2:1, 1:1, and 1:2. For the corresponding AFM micrographs and resulting AFM roughness the same samples are taken.

Figure 4

Absorption of 150-nm-thick ZnPc:C${}_{60}$ blend layers deposited on glass with varied blend ratio and substrate temperature measured via transmission and reflection measurements.

Figure 5

Anisotropic optical constants as determined by VASE measurements for 100-nm-thick ZnPc:C${}_{60}$ (1:1) blend layers deposited on SiO${}_2$ substrates for substrate temperatures of 30 °C (left) and 100 °C (right). The dashed lines correspond to the out-of-plane and the solid line to the in-plane component of $n$ (green) and $k$ (dark red).

Figure 6

Substrate temperature dependency of electron (red/bright circles) and hole mobility (black squares) of ZnPc:C${}_{60}$ (1:1) blend layers measured by OFET. The 40-nm thick ZnPc:C${}_{60}$ blend layers are deposited on SiO${}_2$ substrate with patterned Au contacts at different substrate temperatures of 25, 100, and 140 °C. The error bars represent the standard deviation of four mobility values evaluated from different OFETs on the same substrate with channel length of 10 and 20 μm.

Figure 7

j-V curves and device stack of ZnPc:C${}_{60}$ bulk heterojunction solar cells using different blend ratios (lower part) and substrate temperature (upper part) during deposition. Solar cells stack: ITO (90 nm)/5-nm C${}_{60}$:W2(hpp)4 (2 wt$\%$)/20-nm C${}_{60}$/60-nm ZnPc:C${}_{60}$ (1:1)/5-nm BF-DPB/35-nm BF-DPB:F6-TCNNQ (10 wt$\%$)/2-nm F6-TCNNQ/100-nm Al.

Figure 8

EQE (left) and absorbance (right) of the ZnPc:C${}_{60}$ bulk heterojunction solar cells shown in Fig. 6. The absorbance of the solar cells is measured in reflectance.