Exciton-harvesting scheme for high-efficiency solar cells

Organic solar cells have the potential to provide cheap, clean energy that can help meet the rapidly growing worldwide demand for them. Power conversion efficiencies over 5% have recently been demonstrated,¹ indicating that the field is well on its way to meeting the 10% target many believe necessary to make this technology commercially successful. But multiple processes must be optimized in order to reach and surpass this target.¹ Better light absorption requires materials designed with smaller bandgaps and high absorption coefficients while maintaining good electrical transport properties. Efficiencies are further limited in many device architectures by the exciton diffusion length, which is typically 6nm or less for many organic materials, owing to exciton trapping.² Excitons, or excited states that form upon light absorption, can become stuck at low-energy sites in the absorbing film rather than migrate to dissociation sites where they split into negative and positive charge carriers (electrons and holes).

One way to overcome small exciton diffusion lengths is to intimately mix the two materials that form a heterojunction by randomly blending them to form a `bulk heterojunction.'¹ In this structure, in which all excitons are created close to a heterojunction, a near-perfect harvest is often possible. Unfortunately, such a disordered structure can have dead ends and disconnected islands, either of the hole- or electron-transporting phases, which leads to carrier recombination. Ordered structures would resolve this issue, but current fabrication techniques generate domains too large to harvest sufficient excitons. Another option is to use materials with high yields of an exciton species known as a triplet. In principle, the triplet can diffuse further and is longer lived because decay is a so-called spin-forbidden process. However, 0.5–0.8eV of energy is typically lost when the triplet forms. Our recent approach, described below, relies on a near-field long-range energy transfer process (Förster resonance energy transfer) to direct exciton transfer to the heterojunction. This allows excitons to be harvested over 25nm from an interface while introducing an energy loss of less than 0.1eV.³

Förster resonance energy transfer traditionally involves coupling two chromophores (molecules with interesting optical properties) known as the donor and the acceptor through the electric dipole field.⁴ The exciton can transfer through the field if there is overlap between the emission spectrum of the donor and the absorption spectrum of the acceptor. Elevated rates of transfer can be achieved with strong spectral overlap, high donor emission efficiencies, and alignment of the chomophores. Efficient transfer between two chromophores typically occurs only over separation distances of 2–5nm. If, however, the single acceptor chromophore is replaced by a 3D array, the rate of transfer increases substantially and can become efficient, well over 25nm from the interface.³

To demonstrate efficient long-range energy transfer, we chose as the donor a luminescent red emitter, Dow Red, and as the acceptor an absorptive low-gap polymer, polyN-hexadecan-2-yloxycarbonyl-2,5-bis(2′-thienyl)-pyrrole-2,1,3-benzothiadiazole), or PTPTB.³ Figure 1 shows photoluminescence (PL) and absorption spectra for Dow Red and PTPTB.

Figure 1. Dow Red photoluminescence (PL) spectrum (solid line) and absorption spectrum for low-gap polymer (PTPTB, dashed line) versus wavelength. Reproduced with permission from Scully et al.³

The material properties enabled us to calculate an effective diffusion length of 27nm. The fraction of excitons harvested was measured by comparing the PL of Dow Red on glass, where no exciton harvesting occurs, with titania, where excitons are harvested only by intrinsic migration and subsequent electron transfer, and, finally, on thin films of the energy acceptor PTPTB. Figure 2 shows the fraction of excitons harvested as a function of Dow Red thickness. The continuous curves are diffusion models assuming 3nm (Dow Red on titania) and 27nm (Dow Red on PTPTB) as effective diffusion lengths, indicating that energy transfer is extremely efficient in this system.

Figure 2. Fraction of excitons harvested versus energy donor (Dow Red) thickness. Squares reflect data for Dow Red on titania, where only exciton diffusion followed by electron transfer occurs, and triangles data for Dow Red on PTPTB, which admits of long-range energy transfer. Lines are fits to a diffusion model for diffusion lengths of 3nm (solid line) and 27nm (dashed line). Reproduced with permission of Scully et al.³ TiO₂: Titanium dioxide.

This scheme can enhance exciton harvesting for use in high-efficiency organic solar cells. Work on devices incorporating long-range energy transfer to increase efficiencies has shown initial success,⁵ and more efficient systems are currently being investigated.