© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 2313 www.advmat.de www.MaterialsViews.com COMMUNICATION wileyonlinelibrary.com Adv. Mater. 2011, 23, 2313–2319 Michele Guide, Xuan-Dung Dang, and Thuc-Quyen Nguyen* Nanoscale Characterization of Tetrabenzoporphyrin and Fullerene-Based Solar Cells by Photoconductive Atomic Force Microscopy M. E. Guide, Dr. X.-D. Dang, Prof. T.-Q. Nguyen Mitsubishi Chemical Center for Advanced Materials Department of Chemistry and Biochemistry University of California Santa Barbara, CA 93106, USA E-mail: quyen@chem.ucsb.edu DOI: 10.1002/adma.201003644 Organic photovoltaic devices (OPVs) have garnered signifi- cant interest in the scientific community due to their poten- tial applications as low-cost, light-weight and flexible energy sources. [1–3] The power conversion efficiency (PCE) of such devices is predominantly influenced by the optical and elec- tronic properties of the donor and acceptor materials and the morphology of the donor and acceptor blend. [4–7] The sizes of donor and acceptor domains are often on the scale of nanometers; therefore, an understanding of how varia- tions in local morphological features and charge generation at the nanoscale influence device performance is of the utmost importance. An array of microscopy techniques offering various types of local physical and chemical information has been utilized to characterize OPVs, including scanning electron microscopy, [8–9] Raman spectroscopy, [10–12] atomic force microscopy, [13–17] and transmission electron microscopy. [18–22] Recently, conducting atomic force microscopy (C-AFM) and photoconducting atomic force microscopy (PC-AFM) have been used to probe the nano- scale properties of OPVs. [23–33] C-AFM/PC-AFM map out sur- face current/photocurrent and topography simultaneously; thus, these techniques provide some of the most pertinent information to understanding OPV device performance at the nanoscale. In PC-AFM, a scanning probe microscope sits atop an inverted optical microscope through which a light source is coupled via a fiber optic ( Figure 1 c). In this way, a sample or a device can be characterized in a setup analogous to that used in bulk photovoltaic device testing, with illumination through the indium tin oxide (ITO)-coated glass side. Using this tech- nique, it is possible to observe differences in morphology, probe phase separation, identify donor-acceptor domains, obtain rela- tive charge carrier mobilities, map out current/photocurrent, measure current density–voltage curves and extract relative external quantum efficiencies. [23–33] In this work, we examine the nanoscale morphology and con- ductivity/photoconductivity of a recently developed OPV system based on tetrabenzoporphryin (BP) and either [6,6]-phenyl-C 61 - butyric acid methyl ester (PCBM) or [6,6]-phenyl-C 61 -butyric acid n-butyl ester (PCBNB) using C-AFM and PC-AFM (Figure 1). [9] The precursor to BP, 1,4:8,11:15,18:22,25-tetraethano-29 H,31 H- tetrabenzo[ b,g,l,q]porphyrin (CP), is solution processable from organic solvents, and the resulting film can be converted to BP (p-layer) by thermal annealing. Once the conversion takes place, the BP film is polycrystalline and insoluble in common organic solvents. This allows for facile solution processing and annealing of a subsequent layer comprising BP:PCBM or BP:PCBNB (i-layer), followed by the deposition of a thin PCBM or PCBNB layer (n-layer). The resulting p/i/n device architecture is quite complex ( Figure 2 a). It is important to understand the individual properties of each layer, particularly with respect to morphology as a function of acceptor structure, and correlate these features with bulk performance. We first examine the bulk device characteristics of the p/i/n devices for both acceptor systems. The device fabrication proce- dure and architecture is shown in Figure 2a; more details can be found in the experimental section. The current density–voltage ( J–V) characteristics at AM 1.5 G, 100 mW cm −2 are shown in Figure 2b. The device fabricated using PCBNB as an acceptor consistently exhibits higher PCEs than the device with PCBM (2.8% versus 1.5%), owing to an increase in open-circuit voltage ( V OC ) from 0.44 to 0.60 V, short-circuit current ( J SC ) from 5.66 to 7.22 mA cm −2 , and fill-factor (FF) from 0.61 to 0.64. The incident-photon-to-current efficiency (IPCE) spectra of the two devices are shown in Figure 2c. The IPCE of the BP:PCBNB device reaches 40% from 300 nm to 700 nm. For the BP:PCBM device, the IPCE is roughly 35% from 300–400 nm; however, it drops to less than 30% in the 400–700 nm region. The published values for the lowest unoccupied molecular orbital (LUMO) energies of both PCBM and PCBNB are nearly identical (see Figure 1b), [34,35] suggesting that morphological features could be responsible for the differences in device performance. From bulk measurements, it remains unclear why the devices with PCBNB perform better than devices with PCBM; thus, we applied C-AFM and PC-AFM to examine the nanostructure and electronic properties of each layer of the p/i/n system. In PC-AFM measurements (Figure 1c), a conductive AFM probe scans a sample surface in contact mode and collects either injected current at a fixed bias (defined as being applied to the substrate) in the dark or photo-induced current at 0 V (short circuit condition) under illumination ( ∼30 W cm −2 , or 300 suns). For nanoscale IPCE measurements, white light is passed through a band pass filter wheel before being coupled into the inverted optical microscope, resulting in near- monochromatic light with a light intensity of about 3 suns. [27] An important difference between the PC-AFM and the bulk measurements is that a Au-coated silicon tip is used as the top electrode instead of aluminum. We use gold because it is more