Effect of Al 2 O 3 Recombination Barrier Layers Deposited by Atomic Layer Deposition in Solid-State CdS Quantum Dot-Sensitized Solar Cells Katherine E. Roelofs, † Thomas P. Brennan, ‡ Juan C. Dominguez, ‡ Colin D. Bailie, † George Y. Margulis, †,§ Eric T. Hoke, †,§ Michael D. McGehee, † and Stacey F. Bent* ,‡ † Department of Materials Science and Engineering, Stanford University, Stanford, California 94305, United States ‡ Department of Chemical Engineering, Stanford University, Stanford, California 94305, United States § Department of Applied Physics, Stanford University, Stanford, California 94305, United States * S Supporting Information ABSTRACT: Despite the promise of quantum dots (QDs) as a light-absorbing material to replace the dye in dye-sensitized solar cells, quantum dot-sensitized solar cell (QDSSC) efficiencies remain low, due in part to high rates of recombination. In this article, we demonstrate that ultrathin recombination barrier layers of Al 2 O 3 deposited by atomic layer deposition can improve the performance of cadmium sulfide (CdS) quantum dot-sensitized solar cells with spiro-OMeTAD as the solid-state hole transport material. We explored depositing the Al 2 O 3 barrier layers either before or after the QDs, resulting in TiO 2 /Al 2 O 3 /QD and TiO 2 /QD/Al 2 O 3 configurations. The effects of barrier layer configuration and thickness were tracked through current-voltage measurements of device performance and transient photovoltage measurements of electron lifetimes. The Al 2 O 3 layers were found to suppress dark current and increase electron lifetimes with increasing Al 2 O 3 thickness in both configurations. For thin barrier layers, gains in open-circuit voltage and concomitant increases in efficiency were observed, although at greater thicknesses, losses in photocurrent caused net decreases in efficiency. A close comparison of the electron lifetimes in TiO 2 in the TiO 2 /Al 2 O 3 /QD and TiO 2 /QD/Al 2 O 3 configurations suggests that electron transfer from TiO 2 to spiro-OMeTAD is a major source of recombination in ss-QDSSCs, though recombination of TiO 2 electrons with oxidized QDs can also limit electron lifetimes, particularly if the regeneration of oxidized QDs is hindered by a too-thick coating of the barrier layer. ■ INTRODUCTION Dye-sensitized solar cells (DSSCs) offer a compelling low-cost alternative to conventional photovoltaic cells. The DSSC architecture consists of a mesoporous film of a wide-band-gap oxide, such as TiO 2 or ZnO, coated with a monolayer of dye molecules. 1 The pores are filled with a redox electrolyte that regenerates dye molecules that have injected an excited electron into the metal oxide photoanode. DSSCs have recently reached power conversion efficiencies of over 12%, by cosensitization of two donor-π-bridge-acceptor dyes. 2 In the search for new approaches to increase efficiency and device stability, a number of studies have investigated replacing the sensitizing dye with semiconductor quantum dots (QDs), creating quantum dot- sensitized solar cells (QDSSCs). 3-8 The size quantization of QDs 9 allows for precise control over the band gap for optimal absorption and over band offsets for optimal charge transfer. In particular, QDs can be tuned to absorb in the near-IR, which is difficult to achieve with dyes, and QDs can exhibit higher absorption cross sections than organic or metal-organic dyes over a broad spectral range. 10,11 QDs can be grown directly on the mesoporous TiO 2 by chemical bath deposition, 12,13 successive ion layer adsorption and reaction (SILAR), 14,15 electrodeposition, 16 or atomic layer deposition (ALD). 17,18 Commercialization of DSSC technology has generated interest in employing solid-state hole-transport materials (HTMs), such as the commonly used spiro-OMeTAD (2,2′,7,7′-tetrakis-(N,N-di-p-methoxyphenylamine)-9,9′-spiro- bifluorene), to replace the liquid electrolyte. 19 The use of solid- state HTMs avoids the problem of electrolyte leakage and corrosion of metal contacts, and aims to improve long-term stability. Unfortunately, the recombination rate of electrons in TiO 2 with holes in spiro-OMeTAD is higher than the analogous pathway with the standard I - /I 3 - liquid electrolyte. 20 The high recombination rate limits the active layer thickness in solid-state devices to ∼2 μm, due to the consequently low charge carrier diffusion lengths as well as further increases in recombination rate at greater thicknesses from poor pore-filling by spiro-OMeTAD. 21-23 Active layers of TiO 2 coated with dye molecules must reach thicknesses of ∼10 μm to absorb all incident light; thus solid-state DSSC efficiencies are limited by insufficient light absorption. The high absorption cross section of QDs, with the potential to absorb strongly in a limited Received: December 2, 2012 Revised: February 20, 2013 Published: February 25, 2013 Article pubs.acs.org/JPCC © 2013 American Chemical Society 5584 dx.doi.org/10.1021/jp311846r | J. Phys. Chem. C 2013, 117, 5584-5592