Stacked organic solar cells increase efficiency

Among renewable energy sources, photovoltaic cells are our most promising means of electrical power generation. Although modern solar cells are still more costly in terms of price-per-kilowatt than traditional energy sources, those based on organic materials (as opposed to silicon) offer lower-cost option because they use very little material and can be produced quickly and efficiently.^1,2 However, a major drawback of organic solar cells remains: in contrast with silicon or other inorganic materials, they have a narrow absorption band that limits solar spectrum wavelength absorption.

Standard organic solar cells are typically composed of two different materials, one acting as an electron donor and the other as an electron acceptor. Most modern architectures use C₆₀, or Buckminster fullerenes, as acceptors due to their uniquely efficient exciton dissociation in combination with a wide range of donor materials.^3,4 The efficiency of these architectures is limited primarily by the narrow absorption width of the donor material. More efficient structures can be constructed by stacking two cells comprised of different donor materials on top of each other, thereby increasing the absorption width of the overall solar cell (see Figure 1). Different thin metal layers have been applied to connect the two cells^5,6 and, in this work, the comparative efficiency of stacked solar cells connected with three different metals has been evaluated.⁷

Figure 1. A stack of two individual cells comprising different donor materials. The gray layers are metals, and the other layers are organic materials.

Although the exact working principle of the stacked solar cells is still under investigation, we suggest the following behavior starting with a model for single planar cells: the carrier generation takes place at the interface between the donor and the acceptor material. This local generation induces a high carrier concentration gradient away from the interface, leading to a diffusion current. A drift current generated from an applied external bias can neutralize this. Since a steeper carrier profile needs a higher applied bias for current neutralization, a higher open-circuit voltage can be reached. Applied to the stacked solar cell, this model suggests that if the carrier concentration at the connection between the two cells is low, high diffusion currents occur inside the two separate solar cells. A higher applied external bias, compared to a single cell, is needed to neutralize the diffusion currents and increase the open-circuit voltage. The carrier concentration can be kept low by introducing an efficient recombination center, which in this case takes the form of thin metal layers.

The combination of materials used in this work were pentacene and C₆₀, both of which are characterized by high mobilities.^8,9 A single solar cell with a planar pentacene/C₆₀ active bilayer was made as reference cell. PEDOT:PSS polymer was used as a smoothing layer on top of the transparent contact made of indium tin oxide (ITO). Pentacene grown on the polymer layer formed large crystals, while the pentacene growth on bare ITO was much rougher (see Figure 2).¹⁰ The polymer layer reduced the efficiency of the completed solar cell (see Figure 3).

Figure 2. Atomic force microscopy images of a 100nm-thick pentacene layer grown on indium tin oxide (ITO, left) and on a polymer layer (right). Maximum height differences are 40nm (left) and 30nm (right).

Figure 3. I-V curves characterize the performance of a single cell by plotting current versus voltage. The power conversion efficiencies for the cell with the polymer layer was 0.55% and without the polymer layer was 0.58%

To study the recombination center, aluminum, gold, and silver were used as metallic layers in stacked solar cells (see Figure 4). As the thickness of the layers is typically thinner than the wavelengths of the incident light, reflection and transmission at each interface causes complicated light distribution inside the devices. The amount of photon absorption in a typical stacked solar cell is lower in the first cell compared to the second cell, limiting the overall current (see Figure 5). An improved layer structure could optimize the light intensity and increase the cell's current.

Figure 4. I-V curves characterize performance of stacked solar cells comprising different layers. T1, T3, and T5 are stacked planar heterojunction cells containing a polymer layer on top of the ITO: T2 and T6 do not have a polymer layer. T1 and T2 contain an aluminum layer, T3 contains a gold layer, and T5 and T6 contain silver layers. The overall current of a stacked cell is lower than the current of a single cell due to the non-ideal light distribution inside the structure.

Figure 5. Absorbed photons as a function of the wavelength and position inside the organic layer structure due to optical interference. The incident light is broadband normalized light, coming in from the left.

Stacked cells containing a layer of aluminum tended to produce lower open-circuit voltage (V_oc) than the reference single cell. Since an efficient recombination center should increase V_oc, we suspect oxidation of the aluminum, introducing a barrier for carrier extraction, as suggested by significant resistance around the V_oc value. Stacked cells containing a layer of gold resulted in increased V_oc, but due to the high diffusivity of gold particles, the yield was low. Finally, stacked cells containing a layer of silver produced nearly double the open-circuit voltage of a single cell.

In conclusion, two stacked organic solar cells exhibit increased overall spectral sensitivity. Gold and silver show promise as thin-metal films used as recombination centers between stacked cells. Future work should aim to optimize cell structure for current matching between the stacked cells.