AVS2010 Session EN+NS-WeM: Organic Photovoltaics

The energy level alignment between two prototypical conductive polymers, poly(3-hexylthiophene) (P3HT) and poly[2-methoxy-5-(2'-ethyl-hexyloxy)-1,4-phenylene vinylene] (MEH-PPV) and metals was investigated with ultraviolet and x-ray photoemission spectroscopy (XPS, UPS) dependent on the metal substrate work function. In these experiments thin films of the polymer material were grown in several steps on in-vacuum cleaned metal substrates. In between deposition steps the surface was characterized with UPS and XPS without breaking the vacuum. This was enabled by electrospray polymer thin film deposition directly from solution, which allows the growth of clean macro-molecular films in vacuum. This enabled the measurement of the hole injection barriers and interface dipoles unaffected by environmental contamination artifacts. The presented results demonstrate a systematic dependence of the interface dipole on the substrate work function. This indicates that the charge neutrality level-based “induced density of interface states” (IDIS) model also holds for non-reactive conductive polymer/metal interfaces.

This talk reviews recent work on two types of interfaces that are important for organic photovoltaic cells. In the first part, we present the first direct determination via ultra-violet and inverse photoemission spectroscopy (UPS, IPES) of molecular level alignment between donor (D) and acceptor (A) in a bulk heterojunction.[1] We take the example of the interface between poly(3-hexyl thiophene) (P3HT) and [6,6]-phenyl C₆₁-butyric acid methyl ester (PCBM). The P3HT/PCBM blend is a standard, prototypical system for bulk heterojunction organic photovoltaic (OPV) cells. In question here is the relative position of the molecular levels of the donor (D) and acceptor (A) materials in the blend, given that LUMO(D) - LUMO(A) is linked to the charge separation process, and LUMO(A) - HOMO(D) is linked to the open circuit voltage (V_oc) of the OPV cell. A precise measurement of these molecular level offsets provides a firm basis for the accurate modeling of V_oc produced by these cells. The second part of the talk looks at the electronic structure of transition metal oxide films, such as MoO₃ or WO₃, and their role has hole-collecting electrode or central element of a charge recombination layer (CRL) in a tandem solar cell. Recent work has shown that these compounds exhibit exceptionally large electron affinity and work function.[2,3] N-doped by oxygen vacancies, they can act as efficient high work function hole-extractor (via electron injection through their conduction band) on the anode side of the solar cell. Similarly, combined with a low work function interlayer electrode, they form the central element of a CRL in a tandem cell.

[1] Z. Guan, J. Kim, Y.-L. Loo, and A. Kahn, Org. Electr. (submitted)

[2] M. Kröger, S. Hamwi, J. Meyer, T. Riedl, W. Kowalsky, and A. Kahn, Appl. Phys. Lett. 95, 123301 (2009)

[3] J. Meyer, M. Kröger, S. Hamwi, T. Riedl, W. Kowalsky and A. Kahn, Appl. Phys. Lett. (in press, 2010)

There have been many recent advances in improving the efficiency of organic photovoltaics (OPVs) by using new organic active materials and/or employing improved device architectures. However, our understanding of fundamental OPV device operation principles is still incomplete. A new measurement technique for OPVs, synchronous photocurrent measurement, can give insight into the generation and transport characteristics of photogenerated charge carriers. In synchronous photocurrent measurements, a device is illuminated with chopped monochromatic light in addition to a constant white light bias with an intensity close to 1 sun. With the device biased at any given voltage, the current of the OPV is fed into a lock-in amplifier, which extracts the relevant photocurrent response to the monochromatic light with a varying wavelength.

With this technique, we have shown the bias dependence of the photocurrent for various small-molecule device structures. In planar (or bilayer) and planar-mixed organic heterojunctions (HJs), the photocurrent under forward bias is negative (flowing from the cathode to the anode, opposite to the direction of the dark current), up to high forward biases (~1 V), well in excess of the built-in potential. This reveals the surprisingly dominant nature of the diffusion photocurrent in these architectures. However, for mixed HJ cells, the photocurrent reverses direction at a certain forward bias with the reversal of directions for the electric field and the drift current inside the active layer. There exists a strong correlation between the zero-photocurrent voltage and charge generation profile in mixed HJ OPVs.

This technique can also determine the relative contributions of field-induced exciton dissociation and donor-acceptor interface exciton dissociation in planar HJ cells. Traditionally, exciton dissociation in planar HJ cells is assumed to occur almost exclusively at the donor-acceptor interface; field-induced dissociation is taken as a negligible contributor to the photocurrent in these devices. However, as the thickness of an active layer (either donor or acceptor) increases, field-induced dissociation becomes more important. The field-induced contribution increases as the average location of exciton generation moves greater than one exciton diffusion length away from the interface, as demonstrated by using optical field simulations.

We have further used photocurrent measurements as an instrument in analyzing the recombination behavior in planar organic HJs utilizing different materials. Synchronous photocurrent measurement is a useful technique in determining the photocarrier behavior in organic HJ solar cells.

Photocurrent production in organic photovoltaic structures differs fundamentally from current generation in inorganic semiconductor solar cells. Dissociation of excitons formed by optical absorption in organic materials requires heterointerfaces between electron donor and acceptor components. The extent to which molecular architecture, particularly along the donor-acceptor interface, impacts electronic level alignment and charge separation is of fundamental interest. In this work, we prepare well-defined molecular interfaces by the physical vapor deposition of select donor (MPc, Pn) and acceptor (C₆₀) components under UHV conditions. We determine the detailed structure of the donor-acceptor interface with Scanning Tunneling Microscopy and establish a correlation with electron band alignment (PES) and exciton dynamics (2PPES).

For technologically relevant interfaces between C₆₀ and donors such as pentacene (Pn) or phthalocyanines (Pc), distinct structures/molecular orientations can be selectively engineered by organic MBE through deposition sequence and flux. For the case of C₆₀ and Pn, “co-facial” C₆₀-Pn interfaces are formed by C₆₀ deposition on crystalline Pn bilayer films supported by Ag(111), whereas “edge-on” C₆₀-Pn interfaces result from Pn deposition on hexagonal close-packed C₆₀ monolayers supported by Ag(111). Such “edge-on” interfaces expand into large dendritic islands, as per reported “thin-film” phases, and support C₆₀ cluster formation under subsequent C₆₀ deposition. We show how electronic level alignments critical to V_oc and charge separation efficiency are impacted by these structural changes, and extend this information to other small-molecule cases, ZnPc:C₆₀ and perfluroinated ZnPc, as time permits.

Finally, for interfaces between CuPc and C₆₀, we will present the first studies of charge separation at well-characterized organic donor-acceptor interfaces using TR-2PPE. By pumping the CuPc Q-band at 1.65eV, a time-delayed UV pulse then probes the excited state population. We identify dominant relaxation processes on timescales from 100fs to >100ps. By varying the CuPc film thickness, we observe significantly enhanced charge transfer of the singlet exciton at the interface with C₆₀. Following the population dynamics as a function of energy also provides evidence for recombination from charge transfer states back to the low-lying CuPc triplet.

This work has been supported in part by the NSF under the UMD MRSEC (DMR0520471) and the Surface & Analytical Chemistry Program (CHE0750203).

The electronic structure of the interfaces in dye-sensitized solar cell structures was investigated using x-ray and ultraviolet photoemission spectroscopy (XPS, UPS). Electrospray thin film deposition in high vacuum was used to build the interfaces of interest directly in vacuum without exposure to the ambient. Electrospray enables the fabrication of clean, essentially uncontaminated thin films of organic molecules and nanoparticles directly in vacuum.

The experiments focused on the investigation of the indium tin oxide (ITO)/nanocrystalline TiO₂ interface, as well as the characterization of the TiO₂/RuL₂(NCS)₂ [cis-bis(4,4’-dicarboxy-2,2’-bipyridine)–bis(isothio-cyanato)-ruthenium(II)] (“N3”, a prototypical dye used in many currently pursued device structures)-dye interface. Both TiO₂ and N3 films were built up in several steps. After each step, characterization by XPS and UPS was performed. The resulting sequence of spectra allowed the determination of charge injection barriers and interface dipoles at the ITO/TiO₂ and TiO₂/N3 interfaces. Our experiments revealed a strong influence of the UV radiation during UPS measurements on the band line-up at these interfaces. This was revealed though low intensity x-ray photoemission spectroscopy (LIXPS) measurements, which allow the measurement of the work function prior to UV exposure. These results suggest that even low-level UV radiation, such as encountered in an encapsulated solar cell, may lead to cell degradation over time due to a re-alignment of the electronic structure with detrimental effect on charge transport.