High Surface Area Transparent Conducting Oxide Electrodes with a

Jan 6, 2014 - Department of Chemical Engineering, Stanford University, Stanford, California ..... Office, of the U.S. Department of Energy under contr...
0 downloads 0 Views 328KB Size
Download PDF
Article pubs.acs.org/cm

High Surface Area Transparent Conducting Oxide Electrodes with a Customizable Device Architecture Arnold J. Forman, Zhebo Chen, Pongkarn Chakthranont, and Thomas F. Jaramillo* Department of Chemical Engineering, Stanford University, Stanford, California 94305-5025, United States S Supporting Information *

ABSTRACT: Herein, we report the development of optically transparent high surface area electrodes (HSEs) made of indium tin oxide (ITO) using a facile, template-free, scalable wet chemical synthetic approach. The transparent HSEs are electronically conductive and physically robust, with tunable electrochemically accessible roughness factors from 1 through ∼100. These transparent HSEs can serve as a broadly functionalizable scaffold ideally suited to bridge the gap between the need to minimize ionic and electronic transport lengths within a device and the need to achieve high loadings of active material per surface area (e.g. for high capacity as needed for batteries or high optical density as required for electrochromics or photovoltaics). This gap currently stands as a major hurdle for the utilization of many nanomaterials in electronic and optoelectronic devices. The synthetic approach described here for ITO is transferable to other transparent conducting oxide (TCO) materials. This is the first report of a large-pore, tunable, high surface area TCO electrode with excellent optical and conductive properties. KEYWORDS: transparent conducting oxide, high surface area electrode, indium tin oxide, electrochemical capacitance, photoelectrochemical

O

performance metrics such as charge/discharge rates and response time. For any optoelectronic application, including electrochromics, photovoltaics, and photoelectrodes for solar fuel synthesis, among others, HSE scaffolds allow one to achieve increased device optical density with thin films of active material. This feature is particularly helpful for improving the transport and collection of charge carriers in a semiconductor absorber.8c,11 The inherent texturing of the surface imparted by the 3-D HSE support can also potentially improve light management/trapping compared to a 2-D planar device.1d,e,10c,12 All in all, these benefits can potentially lead to improvements in device performance for a wide variety of technologies, motivating the development of a transparent HSE. In order to meet this need, the work presented herein is the first report of a large-pore, tunable, transparent HSE that also possesses excellent optical and conductive properties. Transparent HSEs, including those made of ITO, are characterized by three fundamental performance metrics: (1) surface area, where the figure of merit is roughness factor (RF), (2) visible transparency and light management within the active layer, and (3) electrical conductivity. Additional desirable properties include mechanical strength (cohesion and adhesion), chemical and electrochemical stability, thermal stability, and a facile synthesis amenable to cost-effective, large area

ne means to improve the performance of optoelectronic devices that utilize transparent conducting oxide (TCO) electrodes is by controlling the micro- and nanoscale structure of the TCO substrate upon which active materials are commonly deposited. A TCO platform with an architecture that allows for large interfacial areas and a high loading of active material while maintaining short charge transport distances as well as desirable light management characteristics could boost the performance of many devices1 including electrochromics,2 electrochemical sensors,3 catalysts,4 supercapacitors,5 batteries,6 photovoltaics,1a−c,7 and photoelectrodes.1f,8 For example, in the case of active materials used for surface-based processes such as in the field of electrocatalysis, the large interfacial area of a large-pore transparent conductive high surface area electrode (HSE) can lower local current densities, which minimizes overpotential losses. Meanwhile, the presence of larger pores in such a device structure improves mass transport compared to the tortuous, bottlenecked pore network that is commonly characteristic of high surface area architectures.9 This point is particularly important in applications where device efficiency is coupled to mass transport of liquid or gas phase species,7e,9,10 motivating the ability to tune and optimize HSE pore sizes to enable facile diffusion of reactants and products (including gas bubbles) throughout the structure. For active materials in which both surface and bulk processes can play a role, such as in batteries, supercapacitors, and sensors, larger interfacial areas allow for faster ionic transport both to and through the active material on a geometric area basis, thus improving key © 2014 American Chemical Society

Received: July 28, 2013 Revised: November 27, 2013 Published: January 6, 2014 958

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

Article

fabrication. Historically, research directions for TCO films, including ITO, focused on creating dense, highly transparent planar films for commercial applications in consumer electronics, thermal glass coatings, and other areas in which structured, light scattering films such as the HSE described here would be undesirable. The shifting of recent work toward structured films of active materials reflects a growing need for new high surface area TCO scaffolds. Nevertheless, current literature covering structured TCO HSE materials is limited. The literature on structured ITO can be divided into two types of syntheses: (1) with or (2) without the use of templating/structure-directing agents. Possibly the most crystalline and highest electronic quality material reported among structured ITO films was prepared free of such agents but employs a costly high-vacuum electron-beam synthesis step.7f Other examples of template-free structured ITO films include free-standing fibrous mats13 as well as fiber networks.14 The majority of high surface area ITO syntheses require the addition (and inevitable removal) of templating/structuredirecting agents, which can add significant cost and complexity that may limit the feasibility of large-scale manufacturing. These include the use of block copolymers, polymer spheres or binders, and densifying agents to produce structured ITO thin films and powders with periodic arrays of pores having either mesoscale (2−50 nm)15 or macroscale (>50 nm) diameters,16 as well as disordered ITO particle aggregations.17 There are very few reports of ITO films with demonstrated efficacy as transparent electrodes that also quantify surface area. These include an ordered mesoporous ITO thin film with ∼10 nm pore diameters and a roughness factor of 45 and a disordered nanoparticulate film with a roughness factor of 140.15 These films represent the highest surface area ITO electrodes to date, both with nanoscale porosity. In this article, we report a simple and effective synthesis for producing a new type of HSE, a macroporous HSE, via a low cost, scalable, atmospheric pressure spray deposition of solution-phase precursors. Importantly, templating agents are not needed to introduce porosity, an advantageous feature for improved device performance and low-cost manufacturing. Our synthetic route produces HSEs that simultaneously match superb physical characteristics with high performance electrical and optical characteristics; we produce mechanically durable films that can be >100 μm thick while remaining macroscopically crack-free and that possess high surface area, with tunable RF values up through 100. The films also exhibit low sheet resistance and a structure whose large surface area is accessible electrochemically. The films maintain excellent transparency in the visible region of the spectrum and can diffusely scatter light for enhanced optical device response. This combination of properties has never before been demonstrated in transparent conducting oxide thin film electrodes.

Figure 1. (a) Schematic cross section of ITO HSE and its three main components. (b) Top-down SEM image of ITO HSE. Inset showing crystallinity of ITO adhesion layer.

substrate in a mechanically robust fashion, allowing for conductivity throughout the entire film. The framework of large ITO particles is also responsible for producing macroscopic porosity within the film (Figure 1b), which based purely on geometric considerations should provide an increased RF of ∼3 per additional monolayer, in the theoretical absence of surface texture or porosity (see Supporting Information for details). Below, we describe how greater RFs are achieved in our synthesized HSEs. The superb crystallinity within the large ITO particles allows for excellent transparency and efficient transport of charge across long distances within each particle (micrometers). Careful selection of particle size tunes the final macroporosity of the film, which in turn tunes its light scattering and intrapore mass transport characteristics. The conformal ITO sol-gel layer acts as an electrically conductive adhesive that further increases the surface area due to its own roughness and porosity. Because sol-gel-processed thin films typically result in nanocrystalline compositions with low conductivity due to high granularity,18 it is essential that the large-crystallite, conductive ITO particles serve as a framework for charge transport through the resulting composite film. This open, porous composite structure is also able to efficiently relax internal stresses upon annealing, resulting in macroscopically crack-free films with thicknesses greater than tens of micrometersa difficult feat with current dense micro- or mesostructured film technologies.18 ITO HSEs presented here consist of two layers illustrated in Figure 1a: (1) the aforementioned two-component, high surface area ITO layer consisting of large ITO particles coated with sol-gel derived ITO and (2) a dense, conductive, planar ITO layer on glass upon which the high surface area ITO layer is supported. Thus, the HSE is fabricated with three discrete ITO components: planar ITO, particulate ITO, and the adhesive sol-gel ITO. The underlying planar ITO substrate

■

RESULTS AND DISCUSSION Our design strategy for synthesizing a TCO film with high surface area and excellent transparency and conductivity involves combining two distinct types of ITO: a powder consisting of large ITO particles and an ITO derived from solgel (Figure 1a). The two types of ITO are combined in a hierarchical manner to form a composite film that takes advantage of the best properties of each component. The large, crystalline ITO particles within the composite film serve as the physical framework. The sol-gel-derived ITO conformally coats the large particles, adhering them to one another and to the 959

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

Article

serves to move charge laterally through the device structure, though great flexibility exists when choosing the underlying conductive substrate, as it need not be transparent, and reflective substrates may be preferred depending on the application. The substrate may consist of a wide range of metals or degenerately doped conductors that are stable under processing conditions. For the present study, the high surface area ITO layer was deposited onto commercial planar ITO substrates via spray deposition with a commercially available airbrush, though other deposition methods work as well. The spray deposition involves the aforementioned two-part mixture containing ITO powder (9:1 In:Sn ratio, ∼20 μm diameter particles) and ITO sol-gel precursors (9:1 In:Sn) in an acidified ethanolic solvent. The ratio of powder to sol-gel as well as the total amount of precursor mixture deposited onto the substrate readily controls the final RF of the HSE. After allowing films to dry, they are calcined in air for cross-linking and crystallization of the sol, followed by further calcination in N2 to produce oxygen defects for improved electrical conductivity. See Supporting Information for synthesis details. As-made ITO HSEs appear both highly porous and highly crystalline as viewed by SEM (Figure 1b and Supporting Information). The ∼20 μm spherical ITO powder particles create a porous, open network on the micrometer scale, whereas the sol-gel precursors densify into smaller crystallites of ∼50 nm (inset of Figure 1b and Supporting Information Figure S1) and appear as a rough texture that coats and binds the large particles together. Scherrer broadening observed in X-ray diffraction (XRD) of a pure sol-gel derived ITO film deposited onto soda lime glass (Supporting Information Figure S4) confirms the nanocrystalline nature of the ITO in this component of the system. The as-fabricated HSE shows reflections from both the bixbyite structure of ITO and cassiterite structure of SnO2.19 The SnO2 phase impurity also exists in the commercial ITO powder but is not present in the dense planar substrate or sol-gel film (Supporting Information Figure S4) suggesting that the commercial powder is the source of the impurity present in the as-fabricated HSE film. X-ray photoelectron spectroscopy (XPS) studies revealed minimal variation in the In and Sn core electron binding energies (Supporting Information Figure S5a and S5b) between the HSE and its individual ITO components (planar, particulate, and sol-gel), confirming consistent surface oxidation states for the In and Sn across the components. The surface composition also remains consistent, as obtained from integration of the In and Sn peak areas (Supporting Information Figure S5c). Optical properties of the HSEs were studied by UV−vis transmission measurements of three HSEs of increasing RF (labeled HSE 1−3, respectively) and are shown in Figure 2b. This figure also shows several reference samples that help elucidate the optical properties of the ITO particle/sol-gel layer. These include a planar ITO/glass substrate with no particle/sol-gel coating, a blank uncoated quartz substrate (with no ITO of any kind), and a particle/sol-gel layer on quartz (labeled Q3) identical to the particle/sol-gel layer used to make HSE 3. The measurement setups are detailed in the experimental section. It is important to note that these HSE films scatter light very strongly and that a measured loss of transmission is not simply due to light absorption by the ITO (i.e., T ≠ 1 − A, where A is absorptance). Rather, transmission is defined as

Figure 2. (a) Optical photograph of large area (7.5 × 5 cm) ITO HSE shows spatial homogeneity, transparency, and scattering properties. (b) Transmission UV−vis−NIR spectra of HSE devices and select control samples (see text for details).

T = 1 − A − R sp − R d − Ls

(Equation 1)

where Rsp is specular reflectance, Rd is diffuse reflectance, and Ls is a summation of any additional light lost due to scattering that is not traditionally characterized by the other, discrete variables in the equation. This last term, Ls, is a consequence of both sample and experimental instrument geometry and, in our case, is predominantly composed of light scattered into and then along the plane of the glass substrate via internal reflection. The result is that some portion of light is lost through the edges. As expected for the blank quartz, no discrete absorption features are observed (Figure 2b), but the overall transmission is slightly attenuated by Rsp off the substrate surfaces (Supporting Information Figure S6a). The planar ITO substrate exhibits a known absorption onset at wavelengths shorter than ∼400 nm, Fabry−Perot interference from 400 to 650 nm with a corresponding R sp peak (Supporting Information Figure S7a), and an additional broad loss of transmission that increases in magnitude from ∼700 nm toward the IR that is dependent on free carrier concentration (Figure 2b).20 The HSE films on both quartz (sample Q3) and planar ITO substrates (samples HSE 1−3 see Supporting Information for synthesis details) exhibit features similar to the uncoated planar ITO films but with much lower overall transmission. This predominantly wavelength-independent attenuation of transmission across the visible can be explained by three factors: (1) more ITO material is present (films are thicker) in the HSEs and, thus, the total absorption increases (see Supporting Information Figure S7c) compared to flat films, (2) the HSE films induce more light scattering events that drive light off-axis from the transmission UV−vis detector, and (3) this same efficient light scattering within the HSE structure further results in more absorption events for the HSE as compared to the planar film. 960

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

Article

Figure 3. (a) Cyclic voltammograms comparing the capacitance charging of four HSE films of increasing RF and planar ITO at 50 mV s−1. (b) The capacitance charging current, Jdl, (extracted from the CVs as the oxidative current at 0.0 V vs SCE) as a function of scan rate is linear for all HSEs and planar ITO; solid lines indicate linear regressions fit to zero. The roughness factors measured by capacitance charging extracted from (c) CV scans and (d) impedance spectroscopy illustrate how desired surface areas can be targeted synthetically on the substrate; high surface area structures can contribute up to tens of percent to the loss in transmission for highly scattering samples (see Supporting Information for more details).

sq−1), and thus the HSE maintains excellent conductive properties as a device. The hierarchical pore size distribution of the ITO HSE should bode well for facile diffusion of reactants and products to and from the electrode surface. The electrochemically active surface area was quantified by comparing the ratio of electrochemical double-layer capacitance-charging current density, Jdl, in cyclic voltammograms (CV) of planar ITO and HSEs collected at scan rates from 5 to 500 mV s−1. Figure 3a shows CVs at 50 mV s−1 comparing planar ITO and HSE 1−4, where higher numbers indicate controllably higher surface area. There are no distinct Faradaic redox features and the symmetrical Jdl for charge/discharge decreases in the anodic direction, likely because this is approaching the potential of zero charge (PZC).21 Qualitatively, this is consistent with ITO’s isoelectric point of ∼8; in the pH 13.6 NaOH electrolyte, the ITO surface should be highly negatively charged and anodic polarization will reduce this charge and the associated Jdl.22 The surface features of these polycrystalline systems can also be complicated by crystal-face dependent surface reconstructions that may vary slightly between the planar ITO and HSE surfaces. The Jdl, which is proportional to surface area and measured for all samples as the oxidative current at 0.0 V vs SCE on the CVs, is plotted as a function of scan rate in Figure 3b. Because Jdl is classically described as a process with no diffusion component, the expected linear relationship between Jdl and scan rate is observed (Figure 3b).21,23 The ratio of HSE to planar ITO capacitance at a given scan rate yields the RF of each HSE and is plotted in Figure 3c, ranging from ∼25−100. These HSE RF values are an average of all scan rates and have

The second factor mentioned (transmission losses due to scattering) is important to note as it is a result of experimental geometry and does not indicate absorption by the ITO. The photograph in Figure 2a visibly confirms the HSE’s high degree of transmission across the visible spectrum. The pale yellowgreen color indicates primary absorption at the violet/UV edge, and the fact that the emblem is not as sharp underneath the HSE is indicative of its aforementioned scattering properties. By simply controlling the amount of material deposited and particle size used during synthesis, these HSEs can be tailored to match the desired optical properties to balance the enhancement from scattering versus the losses from absorption, reflection, and transmission for a given application. To investigate the electrical sheet resistance (Rs) of fully assembled HSEs as well as that of the HSE spray-coated directly onto a quartz substrate (with no planar ITO underlayer), four-point probe measurements were employed. Example current−voltage plots are presented in Supporting Information Figure S8 for bare planar ITO, HSE devices, and particle/sol-gel films on quartz. In general, Rs values for HSEs fabricated on insulating quartz substrates range from 0.2 to 2.0 kΩ sq−1, illustrating that an excellent conductive percolation network exists along the plane of the film even without an underlying conductive substrate. The sheet resistance measured on fully assembled HSEs (planar ITO substrate with a particle/ sol-gel layer on top) is 18.8 ± 0.9 Ω sq−1 for samples with a RF equivalent to sample HSE 2 and varies by no more than ±2 Ω sq−1 for the other HSEs. This sheet resistance is within a factor of 2 of an unmodified planar ITO substrate, (10.8 ± 0.11 Ω 961

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

■

been adjusted to compensate for the actual RF value of the planar ITO (RF = 1.10), which was measured by AFM (Supporting Information Figure S3). Furthermore, the HSEs exhibit a wide potential stability window in both basic (1.0 M NaOH) and acidic (0.5 M H2SO4) electrolytes (Supporting Information Figure S10). The HSE RF values compiled from CVs were corroborated using stepped potential electrochemical impedance spectroscopy (SPEIS). Supporting Information Figure S11 shows representative Nyquist plots at 0.0 V vs SCE for planar ITO and HSE 1. The impedance response of planar ITO could simply be fit to a series RC circuit at all frequencies. In contrast, the impedance response of HSE electrodes was split into two distinct domains. In the low frequency domain, the response could be represented by a series RC circuit. However, the high frequency domain produces a sloped line characteristic of a Warburg impedance that arises from diffusion limitations in a semi-infinite medium. To accurately model this response, the data was fit to a transmission line model previously described by Bisquert, et al.24 From this model, the resulting capacitance at each potential was normalized to the capacitance obtained for planar ITO to determine the RF of each HSE sample (Figure 3d). Very good agreement is found between the SPEISand CV-derived RF values at all potentials, validating the measured RF of approximately 100 for the highest surface area HSE within this set of samples. To our knowledge, this is the first report of a functional ITO HSE electrode with micrometer-scale porosity. The macroporous structure results in a substrate that, compared to existing nanoporous ITO electrodes, may present advantages in terms of mass transport through the pores as well as with loading active materials. The data we present here reveal macroporous HSEs with tunable RFs ranging from 1 to ∼100the only of their kind reported to datevia a scalable spray deposition route. Theoretically, much higher RF films (RF > 100) could be attainable with these methods.

Article

METHODS

Commercial indium tin oxide (ITO) films were purchased on low alkaline earth boro-aluminosilicate glass (4−10 Ω sq−1, cut edges, Delta Technologies Ltd., Stillwater, MN, U. S. A.). Films were precleaned by sonication in acetone and then in isopropanol for 1 h each. ITO powder (9:1, In:Sn, ≥99.99% metals basis, −325 mesh, Sigma Aldrich) was used as received. ITO sol-gel was prepared from anhydrous SnCl4 (Sigma Aldrich), indium(III) acetylacetonate (99.99% metals basis, Sigma Aldrich), ethanol (200 proof), and HCl (concentrated). All chemicals were used as received. For sol-gel preparation, a 12.6 mg mL−1 ethanolic solution of tin(IV) chloride was prepared under argon purging. One milliliter of this solution was added to 200 mg In (III) acetylacetonate, to which 0.05 mL of concentrated HCl was added followed by an additional 0.85 mL of ethanol, and the mixture was stirred for a minimum of 30 min before use. A mixture of ITO powder and sol-gel solution with a ratio of 200 mg of powder to 2 mL of sol-gel was spray deposited via a standard commercial airbrush onto the appropriate substrate. After deposition, films were calcined in a three-step procedure based on the work of Smarsly and co-workers.15 Step 1 is heating to 300 °C at 0.3 °C min−1 in air, step 2 is heating to 450 °C at 10 °C min−1 in air, and step 3 is heating to 450 °C at 10 °C min−1 in N2. Sheet resistance was measured using a Signatone four-point probe (model SP4) connected to a BioLogic VMP3 potentiostat. Tip spacing of 1.0 mm, spring pressure of 45 g, and a large probe-tip radius of 0.5 mm proved effective in reproducibly measuring the highly textured films. All electrochemistry was performed using a BioLogic VMP3 potentiostat with EC-Lab software. Samples were prepared by contacting a Cu wire to the ITO substrate with conductive paint and encasing it in insulating epoxy (Supporting Information Figure S9). Electrodes were photographed, and geometric surface area was measured using Image-J software. This epoxy-encased electrode was fully submerged in electrolyte for testing. Platinum mesh and saturated calomel (SCE) served as the counter and reference electrodes, respectively. Electrochemical capacitance was measured by cyclic voltammetry in aqueous 1.0 M NaOH (99.99% metals basis, Sigma). For each scan rate, three CV cycles were swept, and the third was collected to be processed. No change was observed between the second and third sweep. Stepped potential electrochemical impedance spectroscopy (SPEIS) was performed from −0.25 to 0.25 V vs SCE at 20 mV potential steps with a 5 s rest period between each step. A 20 mV amplitude AC signal was superimposed on each DC potential, and data was collected at 10 logarithmically spaced frequencies per decade from 1 Hz to 100k Hz. X-ray diffraction was performed using Cu Kα radiation on a Philips PANalytical X’Pert Pro diffractometer in parallel beam mode with a programmable divergence slit and PIXcel detector. Scanning electron microscopy was performed using a FEI Magellan with an acceleration voltage of 5 kV. X-ray photoelectron spectroscopy was performed using a PHI VersaProbe XPS Microprobe with binding energies referenced to adventitious carbon at 284.6 eV. Transmission UV−vis spectroscopy was performed using a Cary 6000i spectrophotometer. Diffuse and specular reflectance UV−vis spectroscopy were performed using an Ocean Optics ISP-50-8-R-GT integrating sphere coupled to a Xenon arc lamp for source

■

CONCLUSIONS In summary, we have developed high surface area ITO electrodes with high conductivity, controllable roughness factors, and tunable optical properties using a synthetic route applicable to large-scale fabrication without the need for templating agents. Physical and electrical characterization of these electrodes indicates that they are transparent, porous, crystalline, and conductive, with sheet resistances of 19 Ω sq−1. These films can be fabricated with RFs ranging from 1 to >100 (i.e. up to 2 orders of magnitude more surface area than a planar film). When deposited onto a conductive substrate such as planar ITO, these hierarchically porous thin films exhibit a dramatically higher active surface area while maintaining high conductivity and transparency throughout the electrode. The high surface area, transparency, and conductivity are commensurate with requirements for many optoelectronic applications that require enhanced materials performance without enduring resistive losses or sacrificing mechanical or chemical stability. The general synthetic methodology should be amenable to other oxide and TCO materials including fluorine doped tin oxide (FTO), aluminum zinc oxide (AZO), aluminum tin oxide (ATO), and many other well-utilized materials in optical and electronic fields. This development of high surface area transparent conducting oxide electrode enables numerous opportunities to develop a wide range of high performance, low-cost optoelectronic devices. 962

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

Article

(5) (a) Norio, M.; Satomi, O.; Kalakodimi Rajendra, P. Electrochem. Solid-State Lett. 2004, 7 (8), A247−A249. (b) Yoo, J. J.; Balakrishnan, K.; Huang, J.; Meunier, V.; Sumpter, B. G.; Srivastava, A.; Conway, M.; Mohana Reddy, A. L.; Yu, J.; Vajtai, R.; Ajayan, P. M. Nano Lett. 2011, 11 (4), 1423−1427. (6) (a) Yu, Y.; Chen, C. H.; Shi, Y. Adv. Mater. 2007, 19 (7), 993− 997. (b) Park, M.-S.; Wang, G.-X.; Kang, Y.-M.; Wexler, D.; Dou, S.X.; Liu, H.-K. Angew. Chem. 2007, 119 (5), 764−767. (7) (a) Wangperawong, A.; Bent, S. F. Appl. Phys. Lett. 2011, 98 (23), 233106. (b) Fan, Z.; Razavi, H.; Do, J.-w.; Moriwaki, A.; Ergen, O.; Chueh, Y.-L.; Leu, P. W.; Ho, J. C.; Takahashi, T.; Reichertz, L. A.; Neale, S.; Yu, K.; Wu, M.; Ager, J. W.; Javey, A. Nat. Mater. 2009, 8 (8), 648−653. (c) Peng, K.-Q.; Wang, X.; Li, L.; Wu, X.-L.; Lee, S.-T. J. Am. Chem. Soc. 2010, 132 (20), 6872−6873. (d) Putnam, M. C.; Boettcher, S. W.; Kelzenberg, M. D.; Turner-Evans, D. B.; Spurgeon, J. M.; Warren, E. L.; Briggs, R. M.; Lewis, N. S.; Atwater, H. A. Energy Environ. Sci. 2010, 3 (8), 1037−1041. (e) O’Regan, B.; Gratzel, M. Nature 1991, 353 (6346), 737−740. (f) Yu, H. K.; Dong, W. J.; Jung, G. H.; Lee, J.-L. ACS Nano 2011, 5 (10), 8026−8032. (8) (a) Hu, Y.-S.; Kleiman-Shwarsctein, A.; Forman, A. J.; Hazen, D.; Park, J.-N.; McFarland, E. W. Chem. Mater. 2008, 20 (12), 3803−3805. (b) Kleiman-Shwarsctein, A.; Hu, Y.-S.; Forman, A. J.; Stucky, G. D.; McFarland, E. W. J. Phys. Chem. C 2008, 112 (40), 15900−15907. (c) Le Formal, F.; Grätzel, M.; Sivula, K. Adv. Funct. Mater. 2010, 20 (7), 1099−1107. (d) Peng, Q.; Kalanyan, B.; Hoertz, P. G.; Miller, A.; Kim, D. H.; Hanson, K.; Alibabaei, L.; Liu, J.; Meyer, T. J.; Parsons, G. N.; Glass, J. T. Nano Lett. 2013, 130328105514002. (9) Ito, S.; Zakeeruddin, S. M.; Humphry-Baker, R.; Liska, P.; Charvet, R.; Comte, P.; Nazeeruddin, M. K.; Péchy, P.; Takata, M.; Miura, H.; Uchida, S.; Grätzel, M. Adv. Mater. 2006, 18 (9), 1202− 1205. (10) (a) Boettcher, S. W.; Warren, E. L.; Putnam, M. C.; Santori, E. A.; Turner-Evans, D.; Kelzenberg, M. D.; Walter, M. G.; McKone, J. R.; Brunschwig, B. S.; Atwater, H. A.; Lewis, N. S. J. Am. Chem. Soc. 2011, 133 (5), 1216−1219. (b) Gratzel, M. Nature 2001, 414 (6861), 338−344. (c) Oh, J.; Deutsch, T. G.; Yuan, H.-C.; Branz, H. M. Energy Environ. Sci. 2011, 4 (5), 1690−1694. (11) Sivula, K.; Formal, F. L.; Gratzel, M. Chem. Mater. 2009, 21 (13), 2862−2867. (12) Green, M. A.; Zhao, J.; Wang, A.; Reece, P. J.; Gal, M. Nature 2001, 412 (6849), 805−808. (13) Aoki, Y.; Huang, J.; Kunitake, T. J. Mater. Chem. 2006, 16 (3), 292−297. (14) Wu, H.; Hu, L.; Carney, T.; Ruan, Z.; Kong, D.; Yu, Z.; Yao, Y.; Cha, J. J.; Zhu, J.; Fan, S.; Cui, Y. J. Am. Chem. Soc. 2010, 133 (1), 27− 29. (15) (a) Hoertz, P. G.; Chen, Z.; Kent, C. A.; Meyer, T. J. Inorg. Chem. 2010, 49, 8179−8181. (b) Fattakhova-Rohlfing, D.; Brezesinski, T.; Rathouský, J.; Feldhoff, A.; Oekermann, T.; Wark, M.; Smarsly, B. M. Adv. Mater. 2006, 18 (22), 2980−2983. (16) Zhang, X. a.; Man, Y.; Wang, J.; Liu, C.; Wu, W. Sci. China, Ser. E: Technol. Sci. 2006, 49 (5), 537−546. (17) Chen, H.; Zhu, L.; Li, W.; Liu, H. Mater. Lett. 2010, 64 (6), 781−784. (18) (a) Sanchez, C. m.; BoissiÃ̈ re, C. d.; Grosso, D.; Laberty, C.; Nicole, L. Chem. Mater. 2008, 20 (3), 682−737. (b) Wan, Y.; Zhao. Chem. Rev. 2007, 107 (7), 2821−2860. (19) Quaas, M.; Eggs, C.; Wulff, H. Thin Solid Films 1998, 332 (1− 2), 277−281. (20) (a) Synowicki, R. A. Thin Solid Films 1998, 313−314, 394−397. (b) Dobrowolski, J. A.; Li, Li; Hilfiker, J. N. Appl. Opt. 1999, 38, 4891−4903. (c) Teixeira, V.; Cui, H. N.; Meng, L. J.; Fortunato, E.; Martins, R. Thin Solid Films 2002, 420−421, 70−75. (21) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications, 2nd ed.; John Wiley & Sons, Inc.: Hoboken, NJ, 2001. (22) Goebbert, C.; Nonninger, R.; Aegerter, M. A.; Schmidt, H. Thin Solid Films 1999, 351, 79−84.

illumination and a spectroradiometer for measurement (See Supporting Information for details).

■

ASSOCIATED CONTENT

S Supporting Information *

Roughness factor assumptions, XRD, XPS and detailed optical spectra, 4-point probe measurements, electrode preparation methods, cyclic voltammograms in acid and base, and electrochemical impedance spectra. This material is available free of charge via the Internet at http://pubs.acs.org

■

AUTHOR INFORMATION

Corresponding Author

*T. F. Jaramillo. E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding

This work was supported by the Assistant Secretary for Energy Efficiency and Renewable Energy, Fuel Cell Technologies Office, of the U.S. Department of Energy under contract No. NFT-9-88567-01 and No. AGB-2-11473-01 under Prime Contract No. DE-AC36-08GO28308 from the National Renewable Energy Laboratory (NREL) and through subcontract 7058299 under Prime Contract No. DE-AC0205CH11231 from Lawrence Berkeley National Laboratory (LBNL). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to thank Dr. Erik Garnett for assistance with optical absorption measurements and Benjamin Reinecke, David Abram, Kendra Kuhl, Etosha Cave, and Dr. Peter Christian Kjaergaard Vesborg for helpful discussion.

■

REFERENCES

(1) (a) Schierhorn, M.; Boettcher, S. W.; Kraemer, S.; Stucky, G. D.; Moskovits, M. Nano Lett. 2009, 9 (9), 3262−3267. (b) Kayes, B. M.; Atwater, H. A.; Lewis, N. S. J. Appl. Phys. 2005, 97 (11), 114302. (c) Wyatt, K. M. J. Appl. Phys. 2008, 103 (9), 094515. (d) Yu, Z.; Raman, A.; Fan, S. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (41), 17491−17496. (e) Yablonovitch, E. J. Opt. Soc. Am. 1982, 72 (7), 899−907. (f) Tetreault, N.; Arsenault, E.; Heiniger, L.-P.; Soheilnia, N.; Brillet, J.; Moehl, T.; Zakeeruddin, S. M.; Ozin, G. A.; Gratzel, M. Nano Lett. 2011, 11 (11), 4579−4584. (2) (a) Bach, U.; Corr, D.; Lupo, D.; Pichot, F.; Ryan, M. Adv. Mater. 2002, 14 (11), 845−848. (b) Granqvist, C. G. Sol. Energy Mater. Sol. Cells 2000, 60 (3), 201−262. (3) (a) Baeck, S. H.; Jaramillo, T.; Stucky, G. D.; McFarland, E. W. Nano Lett. 2002, 2 (8), 831−834. (b) Baeck, S. H.; Choi, K. S.; Jaramillo, T. F.; Stucky, G. D.; McFarland, E. W. Adv. Mater. 2003, 15 (15), 1269−1273. (4) (a) Vinodgopal, K.; Hotchandani, S.; Kamat, P. V. J. Phys. Chem. 1993, 97 (35), 9040−9044. (b) Endoh, E.; Otouma, H.; Morimoto, T.; Oda, Y. Int. J. Hydrogen Energy 1987, 12 (7), 473−479. (c) Rausch, S.; Wendt, H. J. Electrochem. Soc. 1996, 143 (9), 2852−2862. (d) Birry, L.; Lasia, A. J. Appl. Electrochem. 2004, 34 (7), 735−749. (e) Chen, Z.; Concepcion, J.; Hull, J.; Hoertz, P.; Meyer, T. J. Dalton Trans. 2010, 39, 6950−6952. (f) Kent, C. A.; Concepcion, J. J.; Dares, C. J.; Torelli, D. A.; Rieth, A. J.; Miller, A. S.; Hoertz, P. G.; Meyer, T. J. J. Am. Chem. Soc. 2013, 135, 8432−8435. 963

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964

Chemistry of Materials

Article

(23) Kissinger, P. T.; Heineman, W. R. Laboratory Techniques in Electroanalytical Chemistry, 2nd ed.; Marcel Dekker, Inc.: New York, 1996. (24) Bisquert, J.; Garcia-Belmonte, G.; Fabregat-Santiago, F.; Ferriols, N. S.; Bogdanoff, P.; Pereira, E. C. J. Phys. Chem. B 2000, 104 (10), 2287−2298.

964

dx.doi.org/10.1021/cm402551m | Chem. Mater. 2014, 26, 958−964