Optimizing organic photovoltaics using tailored heterojunctions: A photoinduced absorption study of oligothiophenes with low band gaps

A power conversion efficiency of 3.4% with an open-circuit voltage of $1\mathrm{V}$ was recently demonstrated in a thin film solar cell utilizing fullerene ${\mathrm{C}}_{60}$ as acceptor and a new acceptor-substituted oligothiophene with an optical gap of $1.77\mathrm{eV}$ as donor [K. Schulze et al., Adv. Mater. (Weinheim, Ger.) 18, 2872 (2006)]. This prompted us to systematically study the energy- and electron transfer processes at the oligothiophene:fullerene heterojunction for a homologous series of these oligothiophenes. Cyclic voltammetry and ultraviolet photoelectron spectroscopy data show that the heterojunction is modified due to tuning of the highest occupied molecular orbital energy for different oligothiophene chain lengths, while the lowest unoccupied molecular orbital energy remains essentially fixed due to the presence of electron-withdrawing end groups (dicyanovinyl) attached to the oligothiophene. Use of photoinduced absorption (PA) allows the study of the electron transfer process at the heterojunction to ${\mathrm{C}}_{60}$. Quantum-chemical calculations performed at the density functional theory and/or time-dependent density functional theory level and cation absorption spectra of diluted $\mathrm{DCV}n\mathrm{T}$ provide an unambiguous identification of the transitions observed in the PA spectra. Upon increasing the effective energy gap of the donor-acceptor pair by increasing the ionization energy of the donor, photoinduced electron transfer is eventually replaced with energy transfer, which alters the photovoltaic operation conditions. The optimum open-circuit voltage of a solar cell is thus a trade-off between efficient charge separation at the interface and maximized effective gap. It appears that the open-circuit voltages of $1.0–1.1\mathrm{V}$ in our solar cell devices have reached an optimum since higher voltages result in a loss in charge separation efficiency.

Figure 1

Chemical structure of DCV3T $(k=1)$, DCV4T $(m=0)$, DCV5T $(k=2)$, and DCV6T $(m=1)$. Bond lengths (numbers) and dihedral angles are given in Table .

Figure 2

(a) Absorbance [$-{\log }_{10}(I∕I_0)$, where $I$ is the transmitted light and $I_0$ is the incident light] and emission spectra of $\mathrm{DCV}n\mathrm{T}$ thin films and (b) the vertical absorption and fluorescence transition energies (in eV) in thin films (circles) and in $\mathrm{C}{\mathrm{H}}_2{\mathrm{Cl}}_2$ (stars) in comparison to TD-DFT calculated vertical transition energies (squares). Number of thienyl units $n$ corresponds to $N=2n+2$ $(N=2n+1)$ double bonds in the $S_0$ $\left(S_1\right)$ geometry.

Figure 3

Calculated $T_1\to T_n$ (neutral molecule) and $D_0\to D_n$ (radical-cation) absorption spectra.

Figure 4

Absorption spectra of chemically oxidized $\mathrm{DCV}n\mathrm{T}$ in $\mathrm{C}{\mathrm{H}}_2{\mathrm{Cl}}_2$ $\left({10}^{-5}M\right)$ by adding equivalents of thianthrenium pentachloroantimonate (0, 0.5, 1.0, 1.5, and 2.0). The arrows indicate the increase of cation $\left(D^{+}\right)$ and the decrease of neutral molecule absorption with increasing oxidation. (d) For DCV6T the cation $\left(D^{+}\right)$ is oxidized further, if the molar ratio of the oxidant exceed 1 and the absorption of dications $\left(D^{2+}\right)$ appear.

Figure 5

Photoinduced absorption spectra of (a) DCV3T and $\mathrm{DCV}3\mathrm{T}:{\mathrm{C}}_{60}$, (b) DCV4T and $\mathrm{DCV}4\mathrm{T}:{\mathrm{C}}_{60}$, (c) DCV5T and $\mathrm{DCV}5\mathrm{T}:{\mathrm{C}}_{60}$, and (d) DCV6T and $\mathrm{DCV}6\mathrm{T}:{\mathrm{C}}_{60}$. Open circles: absolute value $\mid \xi \mid$ of the spectral difference between neat $(-\Delta T_n∕T_n)$ and blend $(-\Delta T_b∕T_b)$ layer PA [see Eq. (2)]. Spectra are measured at low temperatures $(T=10\mathrm{K})$ and low modulation frequencies $(f=225\mathrm{Hz})$. $D_1$, $D_3$, and $T_4$ indicate the identified cation and triplet transitions; transitions in brackets indicate the expected spectral position.

Figure 6

Energies of the main transitions in the triplet manifold of the neutral molecule and doublet manifold in the radical cation as a function of the molecular size, extracted from (a) photoinduced absorption of thin films, (b) absorption spectra in thianthrenium $\mathrm{Sb}{\mathrm{Cl}}_5$, and (c) TD-DFT calculations.

Figure 7

Modulation frequency dependence of triplet-triplet absorption at $1.51\mathrm{eV}$ (DCV3T), $1.38\mathrm{eV}$ (DCV4T), $1.21\mathrm{eV}$ (DCV5T), and $1.12\mathrm{eV}$ (DCV6T). Rectangles: neat layer of $\mathrm{DCV}n\mathrm{T}$. Triangles: $\mathrm{DCV}n\mathrm{T}:{\mathrm{C}}_{60}$ blend layer. Fit according to Eq. (4) for monomolecular recombination (MR) and Eq. (5) for purely bimolecular recombination (BR). DCV6T cation absorption at $0.78\mathrm{eV}$ (stars) and $1.38\mathrm{eV}$ (circles) for $\mathrm{DCV}6\mathrm{T}:{\mathrm{C}}_{60}$ blend. $T=10\mathrm{K}$, pump intensities correspond to Fig. 5.

Figure 8

$IV$ characteristics of solar cells described in Table with donor material: (A) DCV4T, (B) DCV5T, and (C) DCV6T. Illumination was $130\mathrm{mW}∕{\mathrm{cm}}^2$ simulated sunlight, a spectral mismatch of approximately 0.7 has to be considered.