2.1.

Active Layer Thickness and Donor: Acceptor Ratio

In line with our earlier work, a first series of devices with Al top electrode was fabricated. The MD304:PCBM ratio and the active layer thickness were varied, respectively. The largest PCE value of 1.74% was achieved for a PCBM content of ∼70 wt% and for rather small active layer thicknesses of ca. 60 nm; a second maximum with reduced efficiency was found for thicker active layers (ca. 170 nm).¹⁰ The optimized device features a short-circuit current (J_SC) of 6.3 mA cm⁻², an open-circuit voltage (V_OC) of 0.76 V, and a fill factor (FF) of 0.36 (Table 1). Many solution-processed small molecule BHJ solar cells appear to feature smaller FF values compared to polymeric solar cells mostly in the range between 0.3 and 0.5.^{8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21} The highest FF reported for a solution-processed BHJ solar cell containing small molecule donors is 0.61,¹⁴ while for polymer FF >0.7 have been reported.

Table 1

Performance of the investigated MD304:PCBM solar cells with varying top-electrode metals (70 wt% PCBM; d ≈ 60 nm) for AM 1.5 light of 100 mW cm−2 intensity.

As is commonly done for polymer-based devices,²³ we have tried to improve the device performance by adding a number of “secondary solvents” (anisole, nitrobenzene, or tetrahydrofuran) to the chlorobenzene:pyridine mixture; however, this did not enhance the device performance. Also, the commonly used curing protocols, such as heat treatment, solvent soaking, etc., did barely change the device performance. Thus, it can be assumed that the layer morphology obtained from wet deposition of small-molecule inks is much closer to thermodynamic equilibrium than for polymers, where strong annealing effects are commonly observed.

2.2.

OFET Hole Mobility Measurements

Considering the small FF values and the facts that the maximum efficiency is achieved for a rather high PCBM content, at a rather small active layer thickness, and that the efficiency is decreased for thicker active layers (even in the second maximum), the investigated MC:PCBM solar cells seem to suffer from a poor or unbalanced charge-carrier transport. Organic field-transister measurements reveal hole mobilities in the range of 10⁻⁶ cm² V⁻¹ s⁻¹ for layers of pure MD304. For mixed MD304:PCBM layers, the hole mobility decreases by roughly one order of magnitude (5×10⁻⁸ cm² V⁻¹ s⁻¹ at 70 wt% PCBM). These hole mobilities are considerably smaller compared to those obtained for conjugated polymers (e.g., P3HT), measured under the identical conditions and far beyond the electron mobility of PCBM.²⁴ This drawback might be due to the dipolar character of the dyes and could furthermore depend on unfavorable aggregation of the small molecules in the bulk.²⁵ However, the measured hole mobilities of the MC:PCBM devices are still in the range of those obtained for α,α-DH6TDPP:PCBM solar cells (measured in hole-only devices via space-charge-limited current), which are among the most efficient small-molecule solution-processed BHJ solar cells reported thus far.²⁰ Thus, obviously, a deficient carrier mobility is not necessarily detrimental for the solar cell performance.

2.3.

EQE Measurements

This view is further supported by our measurements of the external quantum efficiency (EQE) in our devices (Fig. 1). The data points roughly follow the absorption spectrum of the mixture. At the wavelength of maximum absorption, EQE is ca. 50%, a value that exceeds the EQE of many other solution-processed small-molecule BHJ solar cells, which were reported to be in the range of 30–45%;^{10, 11, 16, 17, 18, 20} values up to 58% were also reported for devices containing a diketopyrrolopyrrole derivative.²¹

2.4.

Permittivity

The origin behind this apparent discrepancy between a decent EQE and a rather poor carrier mobility could be the strong polarity of the MD304 (MC dyes, in general), the molecular dipole moment being ∼14 D.²⁶ Organic materials commonly feature rather low dielectric constants (ɛ_r = 2–3), and as a result, the Coulomb interaction between charges is strong, leading to losses due to recombination. A more polar environment would stabilize charges and thus reduce recombination effects. Thus, we determined the permittivity of our materials by impedance spectroscopy (measurement of the devices’ capacity). For a device PEDOT:PSS/MD304 (47 nm)/Al, an ɛ_r of 3.4 was obtained. This compares to ɛ_r = 2.2 for a heat-treated (i.e., fully aggregated) layer of P3HT. Because ɛ_r of PCBM²⁷ is higher than for MD304 and P3HT, the ɛ_r values of the actual solar cells are higher than for the neat donor compounds. For blends with optimized donor:acceptor ratios, ɛ_r of 4.0 for MD304 with 70 wt% PCBM and 3.6 for P3HT with 40 wt% PCBM were measured. Thus, the exciton binding energy in the MD304 layer is 0.04 eV (10%) smaller than in the P3HT blend (assuming a distance between the charges of 1 nm). It has been shown that in organic solar cells containing materials with high ɛ values, the charge dissociation in the active layer as well as the initial separation distance and the decay rate of the bound electron-hole pairs are improved.²⁸ Overall, assuming similar mobility the performance of the solar cell is enhanced.

2.5.

Variation of the Top Electrode

In order to further optimize the devices, we investigated the influence of the nature of the top electrode (120 nm Al, 6 nm Ca/120 nm Ag, and 6 nm Ba/120 nm Ag) on the solar cell performance (see Table 1). As expected for cathode materials with work functions above the LUMO lowest accepted molecular orbital level of PCBM,²⁹ all cells exhibit similar values for V_OC, whereas J_SC increases with increasing work function ϕ_w of the used metal.³⁰ This might be attributed to a higher built-in voltage (V_bi) in these devices in consequence of the increasing energetic difference between the two electrodes. However, also the different reflection properties of the metals and the resulting electric field distribution inside the active layer have influences on J_SC. The best PCE value of 2.1% under AM1.5 conditions was achieved by using a Ba/Ag cathode.

2.6.

Performance at Reduced Illumination Intensity

The dependency of V_OC, J_SC, FF, and PCE on the power of incident light (P_in) for the optimized MD304 devices is depicted in Fig. 2. Except for high intensities (>150 mW cm⁻²), J_SC increases linearly with P_in. V_OC is nearly constant for the entire intensity range and only drops strongly for light intensities near zero, in agreement with theoretical expectations. Finally, the FF increases with decreasing P_in, which is indicative of reduced recombination effects at low charge carrier concentration (low intensity), again an outcome of the poor carrier mobility. Overall, the efficiency is increased under lower light intensity, reaching 2.7% at an intensity of ∼2 mW cm⁻².