Highly efficient carbon-nanotube and organic-polymer hybrid photovoltaics
Semiconducting organic polymers are excellent light absorbers, making them ideal for fabricating low-cost and environmentally friendly thin-film photovoltaics.1 However, despite tremendous developments, the power-conversion efficiency of organic photovoltaics is still well below those of other competing thin-film photovoltaic technologies such as copper indium selenide,2 copper monosulfide, and amorphous silicon. A major limitation is the inherently poor charge-transport behavior in organic polymers. Single-walled carbon nanotubes (SWNTs), which are quasi-1D, low-loss conductors, can be integrated with organic polymers to boost their charge transport and, consequently, their overall performance.2 Fabrication of such hybrid photovoltaics requires detailed understanding of the energy levels at the heterojunction between organic polymers and SWNTs.
Photoinduced charge separation at a typical polymer-SWNT junction is believed to be fundamentally governed by the alignment of the highest-occupied and the lowest-unoccupied molecular orbitals (HOMO, LUMO) of the constituents. Accordingly, heterojunctions between the widely used poly(3-hexylthiophene-2,5-diyl) (P3HT) polymer and SWNTs are presumed to function such that semiconducting (s-) SWNTs behave as efficient electron acceptors, while metallic (m-) SWNTs are thought to impede charge transfer by creating recombination centers.3 However, experimental investigation of charge transfer at individual SWNT/P3HT junctions has not been carried out to validate these claims. Furthermore, ensemble SWNT/P3HT photovoltaics fabricated based on these assumptions have thus far resulted in poor device performance. Accurate understanding of the nanoscale charge-transfer behavior at SWNT/P3HT junctions is needed to design novel device architectures that can yield greater power-conversion efficiencies. Therefore, we carried out a comprehensive study of photoinduced charge transfer at individual SWNT-P3HT junctions.4
We grew individual SWNTs using chemical-vapor deposition on quartz substrates, which we electrically contacted using photolithography-patterned source and drain electrodes. A layer of P3HT (65nm) was spin coated onto each SWNT, forming nanoscale P3HT-SWNT junctions, while an indium tin oxide layer (200nm) was deposited onto the P3HT as a transparent top electrode: see Figure 1(a). We carried out transport measurements to separately identify each s- and m-SWNT/P3HT junction. Subsequently, we illuminated each heterojunction using a focused laser beam and measured the resulting photocurrent. At zero bias, we found that both s- and m-SWNT/P3HT heterojunctions generated photocurrent corresponding to hole transfer from P3HT to the nanotubes. This result is clearly contradictory to previous predictions based on HOMO (LUMO)-band alignments of P3HT and SWNT, and demonstrates that both s- and m-SWNT function as efficient hole acceptors to P3HT. We postulate that photoinduced charge transfer is governed by the alignment of the P3HT and SWNT Fermi rather than their HOMO (LUMO) levels. By tuning the s- or m-SWNT Fermi levels using electrostatic gating, we reversed the photocurrent direction at the SWNT/P3HT junctions—see Figure 1(b)—which clearly validated our explanation. In addition, using transport measurements, we discovered that on contact with P3HT, both s- and m-SWNT get significantly hole doped. This creates a drift field at the interface supporting hole transfer from P3HT to the SWNTs.
Individual SWNT/P3HT junction devices can be readily scaled up to fabricate large-area (mm2) bilayer photovoltaics using horizontally aligned SWNT arrays. To obtain the intrinsic conversion efficiency of photons to free carriers, we adopted an effective volume accounting for incident photons within an exciton diffusion length (see inset of Figure 2), and estimated the effective external quantum efficiency of the heterojunction. Impressively, our hybrid photovoltaics exhibited a maximum effective external quantum efficiency of 90% at a wavelength of 405nm (see Figure 2). These bilayer P3HT/SWNT photovoltaics also showed nearly zero bimolecular recombination because of extremely fast hole extraction, promising improved fill factors and open-circuit voltage.
In summary, we discovered a high-efficiency photoinduced hole-transfer mechanism from P3HT to SWNTs, irrespective of the SWNT nature. This charge-transfer process is driven by an electric field formed at the nanoscale junction as a result of differences in the Fermi levels of P3HT and SWNTs. Our results suggest that the interfaces between nanotube and polymer are much more complex than previously assumed. Importantly, these findings also make selective omission of m-SWNTs redundant, promising carbon-nanomaterial-based, low-cost, high-efficiency hybrid photovoltaics. We aim at further developing this technology.
We acknowledge support from the University of Michigan/Shanghai Jiao Tong University collaborative research program in Renewable Energy Science and Technology. This work used the Lurie Nanofabrication Facility (University of Michigan), a member of the National Nanotechnology Infrastructure Network funded by the National Science Foundation.
Zhaohui Zhong received his PhD in chemistry in 2005 from Harvard University. He is an assistant professor of electrical engineering and computer science.
Nanditha Dissanayake received his PhD in electrical engineering in 2008 from the University of Surrey. He is a postdoctoral researcher.
