Reciprocal carrier collection in organic photovoltaics

Buffer layers between the acceptor and cathode can perform several functions in organic photovoltaic devices, such as providing exciton blocking, protection of active layers against damage from cathode deposition, and optical spacing to maximize the electric field in the active device region. Here, we study electron collection by replacing the common buffer layer, bathocuproine, with a series of six, substituted tris(β-diketonato)Ru(III) analogues in the structure: indium-tin-oxide/copper phthalocyanine/C${}_{60}$/buffer/Ag. These buffer layers enable collection of photogenerated electrons by transporting holes from the cathode to the ${\mathrm{C}}_{60}$/buffer interface, followed by recombination with photogenerated electrons in the acceptor. We use a model for free-polaron and polaron-pair dynamics to describe device operation and the observed inflection in the current-voltage characteristics. The device characteristics are understood in terms of hole transfer from the highest occupied molecular orbital energy levels of several Ru-complexes to the acceptor.

Figure 1

(a) Schematic of the reciprocal carrier collection process. (b) Diagram of the model for reciprocal carrier collection. Exciton flux $J_x$ reaches the donor/acceptor interface (D/A) and creates polaron pairs (PPs) at density ${\zeta }_{\mathrm{DA}}$. These PPs are in dynamic equilibrium with electrons with density $n_{\mathrm{DA}}$ in the acceptor LUMO and holes with density $p_{\mathrm{DA}}$ in the donor HOMO. Similarly, at the reciprocal acceptor/buffer interface (A/B), there are PPs (density ${\zeta }_{\mathrm{AB}}$) and carriers $n_{\mathrm{AB}}$ and $p_{\mathrm{AB}}$ in the acceptor LUMO and buffer HOMO, respectively. The interfaces are coupled by $n_{\mathrm{DA}}$ and $n_{\mathrm{AB}}$ via current density ($J$) continutity and whose magnitudes are determined by the voltage drop across the acceptor ${\delta }_A$(V${}_a$-V${}_{\mathrm{bi}})$. Similarly, $p_{\mathrm{DA}}$ and $p_{\mathrm{AB}}$ are determined by respective density of states, injection barriers (ϕ${}_{h,D}$ and ${\varphi }_{h,B}$), and voltage drops (${\delta }_D$ and ${\delta }_B$).

Figure 2

Characteristic current density-voltage (J-V) curves calculated from the model in text re-create the inflection behavior seen in OPV devices with a charge extraction barrier. For these calculations, we assume $J_{\mathrm{ph}}$ $=$$5\mathrm{\hspace{0.5em}}\mathrm{mA}$ is the photocurrent, and $J_{\mathit{sD}}$ $=$ 10 μA and n $=$ 3 are the reverse saturation current density and ideality factor of the forward donor/acceptor junction, respectively. Solid and broken lines correspond to different saturation currents of the reversed acceptor/buffer junction spanning $J$${}_s$${}_{\mathrm{AB}}$ $=$ 12.5, 2.5, 0.5, 0.2, and 0.04 mA/cm${}^2$. Symbols show the reverse bias slope generated by shunt resistances of 300 Ω (triangles) and 1 kΩ (squares) in the equivalent circuit model, shown in the inset.

Figure 3

(a) “Ruthenium Blue” process for synthesizing the Ru-based complexes. (b) Molecular structure of the ligand bonded to the Ru core for each buffer layer. Numbers given to each material are ordered by decreasing voltage at the inflection.

Figure 4

Current density vs. voltage (J-V) characteristics under approximately 1 sun AM1.5G illumination of ITO/copper-phthalocyanine (40 nm)/C${}_{60}$ (40 nm)/buffer (10 nm)/Ag (100 nm) devices. Data are shown by symbols and lines. Slight differences in photocurrent magnitude from device to device are attributed to small variations in light intensity and device efficiency.

Figure 5

Current density vs. voltage (J-V) characteristics under approximately 1 sun AM1.5G illumination of ITO/copper-phthalocyanine (40 nm)/C${}_{60}$ (40 nm)/buffer (20 nm)/Ag (100 nm) devices. Data are shown by symbols, and fits to the model in text up to breakdown at ∼0.5 V are shown by solid lines. Inset: Fit (line) of the measured reverse bias characteristic (symbols) ignoring the voltage dependence of the polaron-pair dissociation rate.

Figure 6

The J-V characteristics in Fig. 5 replotted on a semi-log scale. Fits to the data using model in text are shown by lines.

Figure 7

(a) Energy levels measured by a combination of ultraviolet photoelectron spectroscopy (for the highest occupied molecular orbital, HOMO) and cyclic voltammetry (for the lowest unoccupied MO, or LUMO) for 5-nm-thick RuL${}_3$ buffer layers on ${\mathrm{C}}_{60}$ (5 nm)/ITO. (b) Energy-level alignment at the acceptor/buffer interface obtained assuming vacuum-level alignment.

Figure 8

Saturation current, $J_s$${}_{\mathrm{AB}}$, and polaron pair recombination rate, $k_{r,\mathrm{AB}}$, at the acceptor/buffer layer junction compared to the interface energy gap $\Delta E_{\mathrm{HL}}$ $=E_{\mathrm{LUMO},A}$ $-E_{\mathrm{HOMO},B}$ for the different buffer compositions. Here, $J_s$${}_{\mathrm{AB}}$ is obtained by fitting the current density vs. voltage characteristics to the model in text. $k_{r,}$${}_{\mathrm{AB}}$ is estimated from $J_s$${}_{\mathrm{AB}}$ using (6) as described in the text.