Built-in Potential and Charge Distribution within Single

Feb 4, 2013 - The CdS-PbS NRs show a sharp contrast in measured potential across ... We find the measured built-in potential within an individual NR t...
0 downloads 0 Views 859KB Size
Letter pubs.acs.org/NanoLett

Built-in Potential and Charge Distribution within Single Heterostructured Nanorods Measured by Scanning Kelvin Probe Microscopy Sanjini U. Nanayakkara,† Gilad Cohen,‡ Chun-Sheng Jiang,† Manuel J. Romero,† Klara Maturova,† Mowafak Al-Jassim,† Jao van de Lagemaat,† Yossi Rosenwaks,‡ and Joseph M. Luther*,† †

National Renewable Energy Laboratory, 1617 Cole Blvd., Golden, Colorado, United States School of Electrical Engineering, Tel-Aviv University, Tel-Aviv, Israel



S Supporting Information *

ABSTRACT: The electrostatic potential distribution across single, isolated, colloidal heterostructured nanorods (NRs) with component materials expected to form a p−n junction within each NR has been measured using scanning Kelvin probe microscopy (SKPM). We compare CdS to bicomponent CdS-CdSe, CdS-PbSe, and CdS-PbS NRs prepared via different synthetic approaches to corroborate the SKPM assignments. The CdS-PbS NRs show a sharp contrast in measured potential across the material interface. We find the measured built-in potential within an individual NR to be attenuated by long-range electrostatic forces between the sample substrate, cantilever, and the measuring tip. Surface potential images were deconvoluted to yield built-in potentials ranging from 375 to 510 meV in the heterostructured NRs. We deduce the overall built-in potential as well as the charge distribution across each segment of the heterostructured NRs by combining SKPM data with simulations of the system. KEYWORDS: Charge transfer doping, heterostructured nanomaterials, colloidal nanorods, built-in potential, scanning Kelvin probe microscopy

C

polarization, and dielectric constant23 and illumination effects on semiconductor−metal and semiconductor−semiconductor biphasic materials.24,25 Zaniewski et al. used EFM in conjunction with TEM to extract electrostatic force gradient changes along the long axis of heterostructured CdS-Cu2S NRs. In that work, the sample substrate employed positional markers to correlate EFM and TEM data, from which single NR characteristics were extracted. By using a finite element electrostatic model, the built-in potential was determined from the phase shift values of the EFM signal. Single CdS-Cu2S NRs showed built-in potentials ranging from 100 to 920 mV for various NRs.26 Tang et al. used photolithography and electron beam evaporation to fabricate nanowire solar cells from single nanowires (260 nm in diameter) of CdS-Cu2S and found consistent open circuit voltages of 600 mV.27 However, it remains difficult to characterize sub-10 nm junctions and to build statistics on individual nanostructures to be compared with ensemble measurements that may be found in macroscopic devices or optical techniques.

olloidal nanocrystals containing multiple material phases facilitate synergistic properties of each component to be combined with size-dependent effects present at the nanoscale. Colloidal nanostructures, in general, offer promising attributes for a wide variety of applications in solar photoconversion, sensing, labeling, and electronic devices.1 To realize the full potential of such nanostructures in terms of advanced lightharvesting efficiency, for example, robust characterization techniques are needed to aid in nanostructure design and optimization. Nanocrystal systems often consist of hybrid and multicomponent structures with interfaces that impart advanced functionalities.2−7 It is crucial to be able to characterize such interfaces in a similar manner as is commonly done in bulk and thin film junctions. Here, we demonstrate electrical characterization of heterojunctions within single, isolated nanorods (NRs) using bimodal high-frequency SKPM (also known as scanning surface-potential microscopy), an atomic-force microscopy technique.8−14 We present direct, quantitative measurements of potential changes across a nanoscale heterojunction in a multicomponent NR supported on a surface. Scanning-probe techniques provide a wealth of information for surface science, nanoscience, and technology.9,15−19 Electric force microscopy (EFM) has probed the effects of charge,20−22 © 2013 American Chemical Society

Received: January 2, 2013 Revised: January 22, 2013 Published: February 4, 2013 1278

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

The built-in potential is a fundamental characteristic of a material interface possessing a charge-separating junction. Thus it is a key component in a thin-film solar cell as well as novel nanostructures synthesized to increase light-harvesting efficiency. In bulk systems, at thermal equilibrium, it refers to the potential difference across the charge-depleted region at a p−n junction that is ultimately caused by equilibration of the Fermi level in the individual materials with different work functions. Thus the built-in potential can be tuned by varying the carrier concentration (or Fermi level) at each side or both sides of the junction. The carrier concentration is optimized in bulk via incorporation of heterovalent atoms or atomic vacancies, but during growth at the nanoscale, impurities are often not incorporated as desired, yet charge separating interfaces arise due to energy level offsets.28 The properties and the charge distribution of a nanoscale charge-separating interface are largely unknown because the total dimensions are small such that, if the system were described by the same formalism as in bulk, the system would be fully depleted as the probability of containing a single carrier on one side of the junction is close to zero. Hence, the importance of electrical characterization of material interfaces at the nanoscale will facilitate the understanding and control of novel material structures while providing valuable feedback in synthesis of various structures. Our SKPM apparatus uses a commercial AFM housed in an argon-filled glovebox to allow the substrate preparation and measurement to be free of water and oxygen effects. The overall topography and SKPM potential measurements were recorded simultaneously using a single-pass with an external Kelvin Probe Control Unit (Omicron, Kelvin Probe CU) and an external high-frequency lock-in amplifier (Signal Recovery, 7280 DSP) on a Veeco Dimension 5000 AFM and Nanoscope V controller system. Figure 1A shows a schematic of the AFM/ SKPM tip and sample characteristics. In SKPM, electrostatic forces between the tip and sample result from a contact potential difference. The contact potential difference is instantly and automatically nullified by a DC bias applied to the probe tip. This DC bias is recorded at each position to generate a potential image. Topography was measured using tapping mode at the first resonance frequency (50 kHz) of the conductive Pt/Ir coated AFM tip (Nanosensors, PPP-EFM) while an AC bias modulation of 300−400 kHz (2nd resonance of the AFM tip) was added to the tip to measure the electrostatic potential. The feedback laser of the AFM is a 1 mW, 670 nm solid-state laser that impinges directly downward on the cantilever as shown in Figure 1A. The individual NRs are located under the tip and hence under the shadow of the 30 μm wide cantilever; thus the NRs should not be excited by this laser light. CdS NRs were synthesized as described previously using octadecylphosphonic acid and trioctylphosphine oxide for surface passivation.6 TEM (Figure 1B) shows NRs with a minimum distribution in diameter (4.0 ± 0.2 nm) and length (80 ± 30 nm). For the SKPM experiments, very dilute solutions of NRs in hexane were spin-coated onto highly ordered pyrolytic graphite (HOPG) to isolate single NRs over an atomically flat, conductive surface. Figure 1C shows an SEM image of the CdS NRs on HOPG where isolated, individual NRs are highlighted for clarity. The same sample surface (but not the same area) was imaged in AFM, and a topographic image is shown in Figure 1D. This topographic image was recorded with a conducting Pt/Ir-coated tip required for SKPM, and isolated single NRs are visible to about the same

Figure 1. (A) Schematic of AFM/SKPM experimental apparatus. (B) TEM image of CdS NRs. (C) SEM image of CdS NRs spin-coated on a graphite substrate to be measured by AFM/SKPM; isolated individual NRs are circled for clarity. (D) AFM tapping-mode topographical image of the same sample (not the same area) as in C. The AFM tip used for this image is a conducting Pt/Ir-coated tip with a tip apex of 20−30 nm, but due to the dilute sample concentration used, individual, isolated NRs are readily available for characterization.

resolution as in the SEM image. Conductive metal-coated tips are required for the contact potential to be measured. Such metal tips are larger in diameter than conventional AFM tips and generally reduce the spatial resolution of AFM techniques. We compared SEM and AFM images to understand the extent of bundling among NRs adsorbed on the HOPG surface and adjust the concentration of NRs in the deposition solution accordingly. Single NRs such as those circled in Figure 1D were selected for analysis. Using the same sample solution of CdS NRs, we chemically transformed a portion of the CdS NRs to PbS using a cation exchange reaction to form heterojunctions across the long axis of the NR.6 This procedure has been detailed and characterized previously and is useful for these experiments because it allows for a remarkably abrupt junction in which both materials are openly accessible as opposed to core/shell junctions where electrically probing interfaces and interior components is currently difficult. The diameter (and the overall shape) of the structure is based on the initial synthesis as the cation exchange leads to a NR with very little or no topography change across the interface to skew SKPM signals.6,7,29 In this approach, ion exchange likely proceeds through the higherenergy end facets of the NR allowing the original ligand shell to remain on the sides of the structure. The NRs were deposited on freshly peeled HOPG and topography, and SKPM images were then simultaneously recorded. Figure 2 shows the surface topography and the surface potential image of CdS-PbS NRs on the HOPG surface. The topographical image shows NRs of fairly uniform height and length distribution; however the SKPM potential image indicates regions toward the ends of the structures with drastically lower surface potential (appearing darker on the color scale). As we will discuss, the darker regions are due to PbS segments that are introduced during the 1279

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

showing partial PbS-exchange at only one end. We observe an abrupt (at least within the resolution of the microscope) potential change across 10 nm within the center of the NR, and we attribute this to a difference in charge distribution across the material interface. This resolution is corroborated in other SKPM work but, however, could be improved under ultrahigh vacuum conditions.15,23,30 Based on the fact that during chemical exchange the reaction occurs at the ends (and in many cases, at both ends simultaneously) of the NR,7 the deeper (more negative) potential regions are assigned to PbS.11 For this specific NR, an asymmetric 30−40% PbS-exchange from one end of the CdS NR during the cation exchange reaction is estimated. Figure 3C shows a representative TEM image of a single CdS-PbS NR. The atomically resolved material interface spanning 2−3 atoms is clearly visible in the TEM image and denoted by the arrows. Fast-Fourier transform analysis on the atomic structure was performed for the upper and lower portion of the NR to confirm rock salt and wurtzite lattice structures identifying the PbS and CdS regions, respectively. This interface and assignment is in agreement with interfaces characterized previously made in a similar manner using partial ion exchange reactions.6,7,31 Panel D of Figure 3 shows a progressive transparent overlay between the topography and potential for a region of the substrate containing five individual NRs. CdS and PbS are expected to form a nanoscale chargeseparating heterojunction based on the energy level alignment and relative bulk Fermi levels of each component.32−36 In SKPM, the measured surface potential is due to the static charges in the NR, in equilibrium with the substrate and ligand shell. Since the substrate is constant and conductive and we expect little ligand shell variation, the potential difference across the NR is determined by the charge distribution within the NR. Therefore, the net charge distribution across the NR is primarily due to charge transfer across the nanoscale junction and gives us a direct measure of the built-in potential in individual NRs under dark conditions.

Figure 2. (A) Topographic image of CdS-PbS NRs on HOPG. The sparse distribution clearly separates individual NRs. A step edge of the graphite substrate is seen diagonally across the image. (B) The corresponding SKPM image of the surface taken simultaneously with the topography. The measured potential is shown in color scale where dark regions of the CdS-PbS NRs indicate PbS regions, while orange corresponds to CdS with positive potential in reference to the substrate.

postsynthesis ion-exchange procedure. During the exchange process, the tips of the NRs react, and thus the distribution of the products7 can be seen in the potential image in Figure 2B. A three-dimensional representation (from AFM images) of a single CdS-PbS NR is shown in Figure 3A; the top image shows AFM topography, and the bottom image corresponds to the SKPM potential across the NR. A clear depression in potential is observed on one end of this NR. Line profiles of topography and potential are shown in Figure 3B (over the same NR as Figure 3A) with a schematic representation of a CdS-PbS NR

Figure 3. (A) Three dimensional topographic (top) and potential (bottom) images of a single CdS-PbS asymmetric NR. Both data sets are recorded simultaneously in a single pass. (B) Line profile of the NR shown in A, overlapped to highlight the position of change in potential with respect to the topography of the NR. In this case, we estimate a 30−40% exchange of CdS to PbS on a single side, based on the potential profile. (C) A TEM image of a CdS-PbS NR displaying a sharp interface between wurtzite CdS with lattice planes perpendicular to the z-axis of the NR and rock salt PbS domain with lattice planes diagonal to the z-axis. (D) Series of images of overlaid potential and topography showing where the deep potential segments are located on the NRs, with varying degrees of transparency as indicated above the images. 1280

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

Figure 4. Topographic height (black), SKPM potential (red), and TEM images of (A) CdS NRs, (B) CdSe QD embedded inside a CdS NR, and (C) CdS NR with PbSe nanocrystals grown on the ends of the NR. The center panel images show the topography (top) and SKPM (bottom) images with a white arrow indicating from which NR the line profiles were taken. We do not observe a contrast in potential across the long-axis in any NRs in the samples shown in A and B. In the selected CdS-PbSe NR, one PbSe tip is visible in topography, and the potential drop at both ends is highly indicative that a PbSe segment is present on both ends of the NR, thus creating two charge-separating interfaces.

exhibit a 100 mV difference in potential across the whole NR structure as compared with the HOPG surface as do the pure CdS NRs. Finally, we measured topography and potential images of CdS NRs that have epitaxial PbSe quantum dots attached to the ends (Figure 4C) synthesized using the procedure described by Kudera et al.5 In the selected CdS-PbSe NR shown in Figure 4C, one PbSe tip is distinguishable in the topographic profile. The observation of the PbSe tip in potential as well as topography, confirms the lateral resolution of the surface potential contrast that was observed throughout this experiment. During the synthesis of PbSe tips onto the CdS NRs, deposition can also occur on both ends, often with one side growing a larger PbSe component.5 In Figure 4C, we clearly see a potential contrast when we observe a PbSe tip that is physically larger than the diameter of the CdS NR and also when the PbSe tip is within the same dimensions as the CdS NR as is shown on the right side of the linescan. Additionally the end with the smaller PbSe crystal (negligible height difference) shows a slightly larger potential change as expected. The potential measurements across pure-phase CdS, heterostructured CdS-PbS, dot-in-rod CdSe-CdS, and sequentially grown CdS-PbSe clearly demonstrate that the potential differences (or lack thereof) are a direct result of charge

The above assignment was substantiated by measuring the potential across three different types and morphologies of NRs based around CdS. Figure 4 is divided into three panels: overlap of topography and SKPM line profiles (left), topography and SKPM images (center panel, top and bottom respectively), and representative TEM images of the samples used (right). Figure 4A shows topography and potential images of pure CdS NRs. The CdS NRs used for this control experiment were a part of the original batch, from which partial exchange to PbS was achieved. As expected, the CdS-only NRs do not display a potential step across its long axis, as no charge redistribution is expected within the pure-phase CdS NRs. Additionally CdS NRs containing a CdSe seed (termed “dotin-rod” or seeded heterostructures in literature) should not display much of a potential change in SKPM even though there is an internal material interface.37−42 CdS and CdSe are both ntype materials in bulk, and the band alignment has been discussed extensively in other publications.2,37,43 The built-in potential between the components in CdSe and CdS NRs is expected to be negligible as the small Fermi level offset between the two materials leads to minimal charge transfer across the interface. As shown in Figure 4B these NRs do not show a measurable potential difference across the NR and additionally 1281

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

redistribution across the NRs and are predictable based on the expected Fermi-level offset across each interface. When the Fermi-level offset is not expected to lead to charge transfer across the interface (as in the pure phase CdS or dot-in-rod CdSe-CdS), a potential change is not observed in the SKPM measurements, but when a Fermi-level offset is expected (as in the case of CdS-PbS or CdS-PbSe), a sharp contrast in potential is observed. To understand both the magnitude of the potential difference between the components and the spatial profile of the potential, we modeled the system using two separate approaches. To gain insight into the nature of the charge distribution, a finite element analysis software package (COMSOL 4.2) was used to simulate the electrostatic potential surrounding the junctions in the CdS-PbS NRs presented in this article. The system was represented by a NR with a diameter of 4.5 nm and a length of 80 nm that is covered with a 0.7 nm organic shell (dielectric constant εs = 3)44 and is adsorbed to a grounded conducting surface. The distance between the NR and the surface is 0.7 nm, which is the approximate length of the ligand shell molecule.44 This scenario best mimics our NRs adsorbed on HOPG. The NR is composed of two segments in contact to mimic an asymmetric NR as shown in Figure 3A−C each with the following material properties, εs = 181 for PbS and εs = 8.3 for CdS.45,46 To obtain equilibrium between the two ends of the NR, charge will transfer across the interface to equilibrate the Fermi level in each material. CdS is usually found n-type with a Fermi level position close to (0.02 eV below)47 the conduction band edge, while PbS has some variability. In quantum dot solar cells,48,49 PbS is usually mildly p-type, and its Fermi level has been measured via photoemission spectroscopy to reside at or slightly below midgap.50 In this situation, electrons are expected to transfer from CdS to PbS as the Fermi levels equilibrate both with each other and the HOPG surface. In the electrostatic model, we added charges (negative charges to PbS and positive charges to CdS) that are distributed in each material component in various ways to visualize the spatial potential shape and the number of charges required to produce a given potential offset. Similar charge transfer schemes across nanoscale interfaces have been demonstrated. Au-PbS core− shell structures where valence band holes are removed from PbS and added to the Au core to balance the Fermi-level offset led to p-type transistor films regardless of the hydrazine chemical treatment that makes pure PbS QD films n-type.51 Very recently, CdS nanowires were determined to become ptype (an oddity for CdS) when coated with MoO3, invoking a similar charge transfer model to equilibrate Fermi levels (electrons transfer away from CdS into the MoO3 leaving mobile holes in the CdS). This charge transfer in CdS is strong enough that it overrules n-type dopants (Ga) in CdS nanowires.52 We investigated whether the charge in each NR accumulates directly at the interface, is distributed uniformly throughout the NR (Figure 5A−B), or is forced to the terminus ends of the NR as would be found if a depletion region were formed. A uniform charge distribution most accurately matches the shape of the potential measured (compare Figure 5B with Figure 3B). This implies that, when the CdS-PbS NRs are in equilibrium with the conducting HOPG surface, the charges within each material are distributed homogenously, thus giving rise to the sharp potential change across the interface. In COMSOL, by adding 10 charges on each side, a ∼300 mV electrostatic potential

Figure 5. (A) Simulations of the electrostatic potential distribution around a single PbS-CdS NRs adsorbed on a grounded conducting surface using COMSOL 4.2. In this simulation we placed 10 charges on each side of the NR (negative charges on PbS and positive charges on CdS). The model displays an electrostatic potential difference of ∼300 mV as a result of the presence of 10 charges on each side of the NR. (B) Line profile across the surface of the modeled NR.

difference was simulated across the junction. Furthermore, the simulation shows a sharp change in electrostatic potential that spanned ∼10 nm, in accordance with the sharp potential step found in the experimental results. Since SKPM relies on electrostatic forces between the probe tip and sample surface, this simulation assists in understanding the fundamental factors that may give rise to the observed results. Charge separation within the NR gives rise to a change in electrostatic potential consistent with the SKPM measurements shown in Figures 2, 3, and 4. We calculate that a NR of this size with 10−15 charges on each side would yield carrier concentration of 2−3 × 1019/ cm3. Using these numbers as the carrier density in a p−n junction, yields a calculated depletion width of 12−19 nm distributed mostly across the PbS due to the higher static dielectric constant. However this assumes that stationary and charged donor or acceptor atoms are present within the depletion region which we believe does not accurately describes the CdS-PbS NR system. Measuring nanoscale objects such as the NRs using SKPM is challenging because the observed SKPM potential is affected (i.e., reduced) by long-range electrostatic interactions between the measuring probe (cantilever and probe tip) and the sample surface.53 By using a boundary-element method to compute the effective point-spread function (PSF) of the measuring probe, the attenuation factors were estimated for each of the scanning conditions.54 During an SKPM measurement, the AFM tip averages electrostatic forces over the total vertical displacement of the cantilever. The PSF describes the response of the imaging system to a point source or object (i.e., the NR) on a flat surface. The SKPM images are then deconvoluted using the effective PSF along with the Wiener filter and to yield an estimated attenuation factor of the measurement (additional details of the deconvolution can be found in the Supporting Information section).55,56 For this study, we estimate that the actual potential across the structure is attenuated in the SKPM measurements by a factor ranging from 2.5 to 3.4 (average of 1282

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

can be deduced from the shape of the measured potential across the NR that the charges are homogenously distributed. By measuring the electrostatic potential across single nanorods of pure compounds, as well as bicomponent nanorods that are known to form a built-in potential, we have established SKPM as a valuable tool in understanding intricacies associated with energy-level arrangement and charge distribution at nanoscale junctions. While no intentional substitutional doping scheme is employed in colloidally grown NRs used in this study, roughly 10−15 electrons are transferred across the CdS-PbS interface to equilibrate the Fermi levels (or chemical potentials) of each component. Once exchanged across the interface, the charges are likely homogeneously distributed in each half of the rod leading to the potential differences measured by SKPM. Further experiments are ongoing to expand this capability as a method to assist in the design and characterization of nanomaterials for use in solar photoconversion applications.

3.02). When taking this attenuation into account, the potential differences across a 4 nm diameter CdS-PbS NR ranges from 375 to 510 mV. Figure 6 shows an example of the effect of the



ASSOCIATED CONTENT

S Supporting Information *

Additional details regarding the deconvolution of the tip, cantilever, and sample interactions studied in this work. This material is available free of charge via the Internet at http:// pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

Figure 6. (A) Representative TEM image of a single CdS-PbS NR which has two material interfaces as cation exchange has occurred on both sides of the NR. (B) Linescan of the measured and deconvoluted potential across a symmetric PbS-CdS-PbS NR (inset displaying SKPM image from which the linescans were taken) using the Weiner filter and effective PSF. (C) Nonequilibrium alignment of CdS and PbS energy levels with respect to vacuum with appropriate quantum confinement-induced shifts. The estimated Fermi-level positions are drawn to indicate that the bands are aligned in nonequilibrium conditions and are estimated based on macroscopic or bulk values of these materials.47,50 The conduction and valence band levels for CdS57 and PbS58 and reference positions of HOPG59 and the Pt/Ir59 AFM tip are taken from literature.

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Financial support for this work was provided to S.U.N. and J.M.L. by the Center for Advanced Solar Photophysics, and Energy Frontier Research Center funded by the US Department of Energy (DOE), Office of Science, Office of Basic Energy Sciences. C.S.J., M.J.R., and M.A.J. were funded by the DOE Office of Energy Efficiency and Renewable Energy, and J.v.d.L. and K.M. were funded by the Solar Photochemistry Program of the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences of the DOE. DOE funding was provided to NREL through contract DEAC36-08G028308. G.C. and Y.R. were funded by the Israel Science Foundation (grant numbers 498/11). We thank Danielle K. Smith for synthetic assistance, Peter Ciesielski and Bryon Donohoe for TEM assistance, and Matthew C. Beard, Sergei Kalinin, and Peter Graf for helpful discussion.

deconvolution on the potential profile over a single NR with both ends converted to PbS as shown in the TEM image in Figure 6A. Figure 6B shows the measured and unattenuated potential from the SKPM image in the inset. For this size regime, the band gap of the PbS portion is 0.9 eV, and thus the measured built-in potential is roughly half the magnitude of the PbS bandgap. The primary cause of a potential difference across a heterostructured NR is charge transfer across the interface. Due to this charge exchange, an electrical potential difference is established within the NR that is equal in magnitude to the vacuum level shift that is accompanied by the Fermi level alignment. A band alignment drawing is shown in Figure 6C at nonequilibrium. In these NRs, electrons are expected to be transferred from the CdS segment to the PbS segment due to the difference in the Fermi level of each section as drawn and from data in refs 47 and 50. This separation in Fermi levels corresponds well with the range of potential changes measured across NRs, corroborating that our SKPM results present a measurement of built-in potentials. Furthermore, when this heterostructured NR is in contact with the HOPG surface, it



REFERENCES

(1) Milliron, D. J.; Hughes, S. M.; Cui, Y.; Manna, L.; Li, J. B.; Wang, L. W.; Alivisatos, A. P. Nature 2004, 430 (6996), 190−195. (2) Amirav, L.; Alivisatos, A. P. J. Phys. Chem. Lett. 2010, 1 (7), 1051−1054. (3) Dukovic, G.; Merkle, M. G.; Nelson, J. H.; Hughes, S. M.; Alivisatos, A. P. Adv. Mater. 2008, 20 (22), 4306−4311. (4) Halpert, J. E.; Porter, V. J.; Zimmer, J. P.; Bawendi, M. G. J. Am. Chem. Soc. 2006, 128 (39), 12590−12591. (5) Kudera, S.; Carbone, L.; Casula, M. F.; Cingolani, R.; Falqui, A.; Snoeck, E.; Parak, W. J.; Manna, L. Nano Lett. 2005, 5 (3), 445−449. (6) Luther, J. M.; Zheng, H. M.; Sadtler, B.; Alivisatos, A. P. J. Am. Chem. Soc. 2009, 131 (46), 16851−16857. 1283

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284

Nano Letters

Letter

(7) Sadtler, B.; Demchenko, D. O.; Zheng, H.; Hughes, S. M.; Merkle, M. G.; Dahmen, U.; Wang, L. W.; Alivisatos, A. P. J. Am. Chem. Soc. 2009, 131 (14), 5285−5293. (8) Douheret, O.; Anand, S.; Glatzel, T.; Maknys, K.; Sadewasser, S. Appl. Phys. Lett. 2004, 85 (22), 5245−5247. (9) Garcia, R.; Herruzo, E. T. Nat. Nanotechnol. 2012, 7 (4), 217− 226. (10) Glatzel, T.; Lux-Steiner, M. C.; Strassburg, E.; Boag, A.; Rosenwaks, Y. Scanning Probe Microscopy: Electrical and Electromechanical Phenomena at the Nanoscale; Kalini, S., Gruverman, A., Eds.; Springer Science + Business Media, LLC: New York, 2007; Vol. 1, pp 113−131. (11) Jiang, C. S.; Moutinho, H. R.; Friedman, D. J.; Geisz, J. F.; AlJassim, M. M. J. Appl. Phys. 2003, 93 (12), 10035−10040. (12) Kalinin, S. V.; Bonnell, D. A. Nano Lett. 2004, 4 (4), 555−560. (13) Kikukawa, A.; Hosaka, S.; Imura, R. Appl. Phys. Lett. 1995, 66 (25), 3510−3512. (14) Rodriguez, B. J.; Yang, W. C.; Nemanich, R. J.; Gruverman, A. Appl. Phys. Lett. 2005, 86 (11), 112115(1−3). (15) Balke, N.; Bonnell, D.; Ginger, D. S.; Kemerink, M. MRS Bull. 2012, 37 (7), 633−637. (16) Brukman, M. J.; Bonnell, D. A. Phys. Today 2008, 61 (6), 36− 42. (17) Giessibl, F. J. Rev. Mod. Phys. 2003, 75 (3), 949−983. (18) Melitz, W.; Shen, J.; Kummel, A. C.; Lee, S. Surf. Sci. Rep. 2011, 66 (1), 1−27. (19) Nonnenmacher, M.; Oboyle, M. P.; Wickramasinghe, H. K. Appl. Phys. Lett. 1991, 58 (25), 2921−2923. (20) Cherniavskaya, O.; Chen, L. W.; Brus, L. J. Phys. Chem. B 2004, 108 (16), 4946−4961. (21) Krauss, T. D.; Brus, L. E. Phys. Rev. Lett. 1999, 83 (23), 4840− 4843. (22) Krauss, T. D.; Brus, L. E. Mater. Sci. Eng. B 2000, 69, 289−294. (23) Krishnan, R.; Hahn, M. A.; Yu, Z. H.; Silcox, J.; Fauchet, P. M.; Krauss, T. D. Phys. Rev. Lett. 2004, 92, 216803. (24) Costi, R.; Cohen, G.; Salant, A.; Rabani, E.; Banin, U. Nano Lett. 2009, 9 (5), 2031−2039. (25) Schafer, S.; Reich, A.; Wang, Z.; Kipp, T.; Mews, A. Appl. Phys. Lett. 2012, 100, 022110. (26) Zaniewski, A. M.; Loster, M.; Sadtler, B.; Alivisatos, A. P.; Zettl, A. Phys. Rev. B 2010, 82 (15), 205119. (27) Tang, J. Y.; Huo, Z. Y.; Brittman, S.; Gao, H. W.; Yang, P. D. Nat. Nanotechnol. 2011, 6 (9), 568−572. (28) Norris, D. J.; Efros, A. L.; Erwin, S. C. Science 2008, 319 (5871), 1776−1779. (29) Jain, P. K.; Amirav, L.; Aloni, S.; Alivisatos, A. P. J. Am. Chem. Soc. 2010, 132 (29), 9997−9999. (30) Spadafora, E. J.; Demadrille, R.; Ratier, B.; Grevin, B. Nano Lett. 2010, 10 (9), 3337−3342. (31) Demchenko, D. O.; Robinson, R. D.; Sadtler, B.; Erdonmez, C. K.; Alivisatos, A. P.; Wang, L. W. ACS Nano 2008, 2 (4), 627−636. (32) Garcia, H. M.; Gomez-Daza, O.; Compos, J.; Nair, M. T. S.; Nair, P. K. Proc. Mater. Res. Soc. Symp. 2007, 1012, 451−456. (33) Hernandez-Borja, J.; Vorobiev, Y. V.; Ramirez-Bon, R. Sol. Energy Mater. Sol. Cells 2011, 95 (7), 1882−1888. (34) Mohanta, K.; Pal, A. J. J. Phys. Chem. C 2008, 112 (9), 3232− 3238. (35) Mukherjee, B.; Peterson, A.; Subramanian, V. Chem. Commun. 2012, 48 (18), 2415−2417. (36) Robel, I.; Kuno, M.; Kamat, P. V. J. Am. Chem. Soc. 2007, 129 (14), 4136−4137. (37) Lee, Y. L.; Chi, C. F.; Liau, S. Y. Chem. Mater. 2010, 22 (3), 922−927. (38) Muller, J.; Lupton, J. M.; Lagoudakis, P. G.; Schindler, F.; Koeppe, R.; Rogach, A. L.; Feldmann, J.; Talapin, D. V.; Weller, H. Nano Lett. 2005, 5 (10), 2044−2049. (39) Smith, E. R.; Luther, J. M.; Johnson, J. C. Nano Lett. 2011, 11 (11), 4923−4931.

(40) Steiner, D.; Dorfs, D.; Banin, U.; Della Sala, F.; Manna, L.; Millo, O. Nano Lett. 2008, 8 (9), 2954−2958. (41) Talapin, D. V.; Koeppe, R.; Gotzinger, S.; Kornowski, A.; Lupton, J. M.; Rogach, A. L.; Benson, O.; Feldmann, J.; Weller, H. Nano Lett. 2003, 3 (12), 1677−1681. (42) Talapin, D. V.; Nelson, J. H.; Shevchenko, E. V.; Aloni, S.; Sadtler, B.; Alivisatos, A. P. Nano Lett. 2007, 7 (10), 2951−2959. (43) Borys, N. J.; Walter, M. J.; Huang, J.; Talapin, D. V.; Lupton, J. M. Science 2010, 330 (6009), 1371−1374. (44) Jiang, J.; Krauss, T. D.; Brus, L. E. J. Phys. Chem. B 2000, 104 (50), 11936−11941. (45) Dalven, K. Solid State Physics; Ehrenreich, H., Seitz, F., Turnbull, D., Eds.; Academic Press: New York, 1973; Vol. 28, pp 179−180. (46) Ninomiya, S.; Adachi, S. J. Appl. Phys. 1995, 78 (2), 1183−1190. (47) Brandhurst, H. W. NASA Tech. Note 1969, 1−13. (48) Chen, H. Y.; Hou, J. H.; Dayal, S.; Huo, L. J.; Kopidakis, N.; Beard, M. C.; Luther, J. M. Adv. Energy Mater. 2011, 1 (4), 528−533. (49) Semonin, O. E.; Luther, J. M.; Choi, S.; Chen, H. Y.; Gao, J. B.; Nozik, A. J.; Beard, M. C. Science 2011, 334 (6062), 1530−1533. (50) Gao, J. B.; Perkins, C. L.; Luther, J. M.; Hanna, M. C.; Chen, H. Y.; Semonin, O. E.; Nozik, A. J.; Ellingson, R. J.; Beard, M. C. Nano Lett. 2011, 11 (8), 3263−3266. (51) Lee, J. S.; Shevchenko, E. V.; Talapin, D. V. J. Am. Chem. Soc. 2008, 130 (30), 9673. (52) Li, F.-Z.; Luo, L.-B.; Yang, Q.-D.; Xie, C.; Nie, B.; Jie, J.-S.; Wu, C.-Y.; Wang, L.; Yu, S.-H. Adv. Energy Mater. 2012, DOI: 10.1002/ aenm.201200868. (53) Elias, G.; Glatzel, T.; Meyer, E.; Schwarzman, A.; Boag, A.; Rosenwaks, Y. Beilstein J. Nanotechnol. 2011, 2, 252−260. (54) Strassburg, E.; Boag, A.; Rosenwaks, Y. Rev. Sci. Instrum. 2005, 76, 083705. (55) Gonzalez, R. C.; Woods, R. E. Digital Image Processing, 3rd ed.; Prentice-Hall, Inc. 2006. (56) Machleidt, T.; Sparrer, E.; Kapusi, D.; Franke, K.-H. Meas. Sci. Technol. 2009, 20 (8), 084017(1−6). (57) Van de Walle, C. G.; Neugebauer, J. Nature 2003, 423 (6940), 626−628. (58) Jasieniak, J.; Califano, M.; Watkins, S. E. ACS Nano 2011, 5 (7), 5888−5902. (59) Bohmisch, M.; Burmeister, F.; Rettenberger, A.; Zimmermann, J.; Boneberg, J.; Leiderer, P. J. Phys. Chem. B 1997, 101 (49), 10162− 10165.

1284

dx.doi.org/10.1021/nl4000147 | Nano Lett. 2013, 13, 1278−1284