Shell Quantum Dots Formed

Jan 17, 2012 - A Straightforward Method For Interpreting XPS Data From Core–Shell Nanoparticles. Alexander G. Shard. The Journal of Physical Chemist...
0 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/JPCC

Probing the Structure of Colloidal Core/Shell Quantum Dots Formed by Cation Exchange Keith A. Abel,† Paul A. FitzGerald,‡ Ting-Yu Wang,§ Tom Z. Regier,∥ Mati Raudsepp,⊥ Simon P. Ringer,§ Gregory G. Warr,‡ and Frank C. J. M. van Veggel*,† †

Department of Chemistry, University of Victoria, Victoria, British Columbia, Canada V8W 3V6 School of Chemistry, The University of Sydney, New South Wales 2006, Australia § Australian Centre for Microscopy and Microanalysis, The University of Sydney, New South Wales 2006, Australia ∥ Canadian Light Source, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 0X4 ⊥ Department of Earth and Ocean Sciences, The University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z4 ‡

S Supporting Information *

ABSTRACT: Cation-exchange reactions have greatly expanded the types of nanoparticle compositions and structures that can be prepared. For instance, cation-exchange reactions can be utilized for preparation of core/shell quantum dots with improved (photo)stability and photoluminescence quantum yield. Understanding the structure of these nanomaterials is imperative for explaining their observed properties and for their further development. Core/shell quantum dots formed by cation exchange are particularly challenging to characterize because shell growth does not lead to an increase in overall particle size that can easily be characterized by standard transmission electron microscopy (TEM). Here, we report on the direct observation of the PbSe/CdSe core/shell structure (formed by cation exchange) using high-angle annular dark field (HAADF) imaging and energy-filtered TEM (EF-TEM). These results are further confirmed by energy-dependent X-ray photoemission spectroscopy (XPS) data that show increasing Pb/Cd signal with increasing X-ray photon energies. High-resolution XPS at varying X-ray photon energies was also used to examine chemical speciation and reveal greater complexity in both the PbSe coreonly and the PbSe/CdSe core/shell structures than previously reported. Finally, small-angle X-ray scattering (SAXS) and smallangle neutron scattering (SANS) methods are combined to provide further inorganic and organic structural information. All experiments agree within error, and the results are summarized as final structural models for the core and core/shell particles.

■

INTRODUCTION Colloidal semiconductor quantum dots (QDs) are an important class of materials with a multitude of potential applications1 ranging from bioimaging2,3 and solar energy conversion4,5 to quantum computing.6,7 Despite the tremendous progress and efforts by researchers around the world, understanding the composition of nanomaterials at the atomic level is still a very big challenge, especially for QDs, with sizes typically below ∼10 nm (and often below 5 nm). This problem is compounded for core/shell QDs, where thin shells and small lattice mismatches can add to the structural characterization challenges. Herein, the morphology of ∼5 nm colloidal PbSe/ CdSe core/shell QDs formed with a cation-exchange reaction is examined. By combining advanced high-resolution microscopy techniques with synchrotron photoemission experiments as well as X-ray and neutron scattering methods a detailed examination of both the inorganic and the organic structural composition of these interesting heterostructured nanomaterials is presented for the first time. All of the experimental techniques utilized agree within error, and the following key conclusions are drawn from this study: (1) there is evidence for © 2012 American Chemical Society

structural and/or chemical reconstruction at the surface of the PbSe QDs; (2) the PbSe/CdSe QDs have a core/shell structure (neglecting the oleate shell) with a purely PbSe core, an outer shell of pure CdSe, and a sharp interface of (strained) Pb1−xCdxSe material; (3) the outer surfactant layer for PbSe is composed of primarily Pb−oleates with some trioctylphosphine ligands (∼1.7 nm thick), while for PbSe/CdSe it is entirely composed of Cd−oleates (∼1.8 nm thick); (4) the anionic framework is conserved during the cation exchange; (5) the increase in sample heterogeneity after cation exchange is at least partly associated with a difference in exchange rate for different size PbSe QDs within the initial QD ensemble; (6) lifetime measurements support a reduction in emission oscillator strength for the core/shell QDs. Cation-exchange reactions in a wide range of nanocrystals were first reported by Alivisatos and co-workers in 2004.8 The authors demonstrated very convincingly that complete and fully Received: November 23, 2011 Revised: January 17, 2012 Published: January 17, 2012 3968

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

OPO (10 Hz repetition rate, 5 ns pulse width, tunable from 400 to 2000 nm) and measured with a liquid nitrogen-cooled Hamamatsu R5509 photomultiplier tube (PMT). Emission decay curves were recorded with a Tektronix Oscilloscope (TDS 3012B), and lifetime fits were performed with Origin software. X-ray diffraction (XRD) data in Figure 2 were collected over a range 3−80° 2θ (step size 0.04°2θ, counting time 1 s/step, sample spinner 50 rpm) with Co Kα radiation on a Bruker D8 Focus Bragg−Brentano diffractometer equipped with an Fe monochromator foil, 0.6 mm (0.3°) divergence slit, incidentand diffracted-beam Soller slits, and a LynxEye detector. The long fine-focus Co X-ray tube was operated at 35 kV and 40 mA using a takeoff angle of 6°. Mineral identification was done using the International Centre for Diffraction Database PDF-4 and Search-Match software by Siemens (Bruker). XRD data were refined with Rietveld program Topas 4.2 (Bruker AXS). Concentrated samples were drop casted onto a zero-diffraction quartz plate. Figure S16, Supporting Information, XRD data was acquired with a Rigaku Miniflex Diffractometer. Samples were drop casted onto a zero-background holder and then allowed to dry. Cr radiation was used at 30 kV and 15 mA with a Kβ filter, a 4.2° scattering slit, and a 0.3 mm receiving slit. XRD data were collected in the 20−140° (2θ) range. Microscopy images were performed with a JEOL JEM2100FS with a built-in HAADF detector and Omega filter. The microscope was operated at 200 keV. Acquisition and data analysis was performed with a digital micrograph by Gatan Inc. Samples were deposited on thin, carbon-coated copper grids for imaging. No plasma cleaning was performed. For EF-TEM imaging, the TEM has a unique built-in Omega filter, which provides better image quality than traditional filters. The energies acquired were the lead O2,3 edge at 86 eV and the Se M4,5 edge at 57 eV. It was not possible to measure reliably the Cd signal (M4,5 edge at 404 eV). The jump-ratio method was used for elemental mapping. X-ray photoelectron spectroscopy (XPS) was carried out at the Canadian Light Source, Inc. in Saskatoon, Saskatchewan, using the spherical grating monochromator (SGM) undulator beamline, 11ID-1, and a Scienta 100 hemispherical analyzer equipped with a channel plate detection system. The particles were washed extensively with ethanol to remove excess ligands that can lead to charging. Particles were drop casted onto Au foil on a sample holder and allowed to dry, followed by immediate mounting into the vacuum sample chamber. No charging effects were observed at low particle concentration. For energy-dependent measurements spectra were measured with pass energy of 100 eV and an energy step size of 100 meV. The beamline exit setting was set to 100 μm. For the highresolution photoemission scans the pass energy was set to 50 eV and energy step size of 50 meV. The beamline exit setting was set to 25 μm. For the energy-dependent measurements all photoemission spectra were fitted and subtracted with polynomial background fits using Origin data analysis and graphing software (version 7.5). Three different background fits were performed for each peak, and the peaks were integrated after subtraction of the background fit. Each measured area was normalized to the photon flux, number of scans, and absorption cross-section. The areas were averaged after correction, and the standard deviation was calculated. After propagating the error to the final ratio (Pb/Cd), the % error (standard deviation/ratio) of each point was calculated and then averaged as a whole (over all

reversible exchange of cations in nanocrystals can occur. Furthermore, they showed by examining high-resolution transmission electron microscopy (HR-TEM) images that above a critical size the anion sublattice can remain intact and the shapes of the initial nanocrystals can thus be retained. This seminal work has greatly expanded the range of nanoscale materials that can potentially be developed.9−12 An example of this is the work performed by Hollingsworth and co-workers, who were able to use the basic idea of cation exchange to form heterostructures.13 The controlled exchange of Pb2+ cations at the surface of PbSe QDs for Cd2+ cations results in PbSe/CdSe QDs with improved photostability and photoluminescence (PL) quantum efficiencies. Unlike the conventional epitaxial growth of a shell material onto a core at temperatures typically above 150 °C (e.g., CdSe/ZnS QDs),14 the cation-exchange reaction can be performed at more modest temperatures (∼80 °C), which allow the optical properties of the PbSe QDs to remain intact. However, due to the good lattice match, Hollingsworth and co-workers had difficulty seeing the interface between the rock salt PbSe core (lattice constant = 6.124 Å) and zinc blende CdSe shell (lattice constant = 6.077 Å) material domains, even in HR-TEM images. As such, they were not able to draw definitive conclusions on the sharpness of the PbSe/CdSe inorganic interface or whether a true core/ shell structure was formed (i.e., complete/uniform shell formation). PbS/CdS13 and PbTe/CdTe15 heterostructured QDs have also been prepared, and because of their small bulk band gaps and large exciton Bohr radii these lead chalcogenidebased materials are well suited to applications in the nearinfrared.16 Despite the aforementioned advantages, a major drawback of the cation-exchange core/shell procedure is the overall increase in full-width at half-maximum (fwhm) of the photoluminescence (PL) peak that suggests an increase of core size and shell thickness dispersions.13,17 Recent work has also shown a significant (ca. factor of 4) reduction of the oscillator strength of the band gap emission for PbSe/CdSe QDs compared to the PbSe core precursor material18 and evidence for a quasi-type II band alignment in these materials.19 Understanding the internal structure of these materials is of vital importance for further development and understanding their electrical and optical properties. For instance, a gradient of elements present (at the interface or in the shell) versus a sharp inorganic interface could explain some of the observed optical properties of these materials.18 Additionally, advancing the techniques and methods required to characterize heterostructured materials with dimensions below ∼10 nm is of fundamental interest to the scientific community. For example, the characterization techniques discussed herein could potentially provide further structural insight into other QD types, such as PbSe/PbS core/ shell QDs,20,21 and help to elucidate the optical properties in those cases as well.

■

EXPERIMENTAL SECTION Synthesis. The syntheses of PbSe core QDs and PbSe/ CdSe core/shell QDs were reported previously.17 Temperature and reaction time were carefully controlled to yield appropriate QD size and shell thickness. The transition of the first exciton peak (λabs), as measured by absorption spectroscopy, is provided where applicable. Absorption spectra were recorded on a Perkin-Elmer LAMBDA 1050 UV/vis/NIR spectrophotometer. PL lifetime measurements were performed with a pulsed Nd:YAG pumped 3969

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

kinetic energies) to give an average error of ca. 5%. This value was then used as the % error for each point of the Pb/Cd ratio at varying kinetic energies. For the high-resolution energy spectra, photoemission spectra were measured concurrently with scans of the C 1s peak (binding energy of 284.5 eV). The C 1s peak was then used as a reference point in determining the binding energies for all peaks at all energies. Binding energies were compared to ref 22. XPS curve fitting was performed with XPSpeak41 software. Gaussian−Lorentzian line shapes were used for fitting with a Tougaard background. IMFP values were approximated using the TPP-2 M equation and based on the bulk values (i.e., bulk densities and band gaps) of PbSe and CdSe.23 Small-angle X-ray scattering (SAXS) was performed on a point-collimated Anton Paar SAXSess through 1 mm capillaries. Data was collected on high-sensitivity 62 × 66 mm2 imaging plates and radially averaged to produce 1D scattering with q = 0.02−0.4 Å−1. The beam spread from the collimation system introduces a polydispersity of Δq/q = 0.063 and is accounted for in the fitting by convolution of the fit models with a Schulz distribution24

Figure 1. Absorbance spectra of PbSe core (solid) and PbSe/CdSe core/shell (dash) QDs in solution (tetrachloroethylene). Cationexchange reaction was performed at 85 °C for 25 min.

core-only QDs and the resulting PbSe/CdSe core/shell QDs in solution. A clear blue shift is present, consistent with a cationexchange reaction that produces a smaller core but an overall equivalent size as confirmed by standard TEM.17 X-ray diffraction (XRD) confirmed the cubic crystal structure for both core and core/shell materials. Rietveld refinement was utilized to fit the measured XRD pattern as shown in Figure 2. However, because of the small crystallite size that results in a significantly broadened pattern, accurate core size and shell thickness could not be determined for the core/shell particles. Therefore, further characterization of the core/shell QDs is required. Although the core size and shell thickness may vary slightly for the QDs investigated herein, we focus on QDs with shell thicknesses in the range from ∼0.5 to ∼1 nm and total particle sizes of ∼5 nm. The transition of the first exciton peak (λabs), as measured by absorption spectroscopy, is provided where applicable. Imaging with HR-TEM, HAADF, and EF-TEM. Hens and co-workers were the first to examine in detail the core/shell structure of PbTe/CdTe core/shell QDs formed by cation exchange.15 Although a very important step forward, in their case, only some of the particles show clear visualization of the core and shell lattice. Due to the small lattice mismatch between rock salt PbTe (lattice constant of 6.462 Å) and zinc blende CdTe (lattice constant of 6.480 Å), they could only resolve the core/shell lattice for specific orientations of the particles, because in HR-TEM visualization of the crystal lattice depends on crystal orientation and the focus. Although they were only able to present a small number of particles showing a core/shell structure, the authors concluded that the cationexchange process is anisotropic, resulting in sample heterogeneity. These observations are qualitatively consistent with our own work examining the temperature-dependent PL of PbSe/ CdSe QDs.17 In that work, based on a detailed analysis of the temperature-dependent shift and variation of the emission line width from 4.5 to 295 K, it has been determined that the Gaussian-like distribution of QD sizes broadens after cation exchange. This observation requires further structural analysis of the QDs to help understand and explain its origin. For the PbSe/CdSe QDs investigated here, the close lattice match between core and shell phases introduces significant challenges in generating clear discriminating contrast by standard HR-TEM, as has been pointed out previously.15 In

z

S(R ) =

⎛ (z + 1)R ⎞ (z + 1)z + 1 ⎛ R ⎞ ⎜ ⎟ exp⎜ − ⎟ R ave Γ(z + 1) ⎝ R ave ⎠ R ave ⎠ ⎝

where R is the radius, Rave is the average radius, z = 1/p2 − 1, p is the polydispersity, and Γ is the gamma function. This also accounts for any polydispersity of the particles, which we found to be small (i.e., < 6%). At the low polydispersities in these experiments the Schulz distribution tends to a normal (or Gaussian) distribution, where Mattousi et al. have shown that the choice of distribution is not critical for such small polydispersities.25 Small-angle neutron scattering (SANS) experiments were performed on the Quokka beamline at the Australian Nuclear and Science Technology Organisation (ANSTO).26 Measurements were performed at 25 °C using neutrons with an average wavelength of 4.95 Å and a wavelength spread of Δλ/λ = 0.0655. Scattering was collected from 1.0 mm thick samples onto a 1.0 × 1.0 m2 2D detector with 192 × 192 elements at two distances: 1.3 and 8.8 m, with the detector offset by 30 cm at 1.3 m to give a combined q range of 0.0084−0.62 Å−1. Raw SANS data were reduced to 1D data and fit in Igor Pro 6.12A using the reduction and fitting procedures provided by NIST.27 Scattering length densities were calculated using the NCNR SLD calculator.28,29 Scattering length densities (SLDs) for both SAXS and SANS were calculated using the NIST online calculator,29 which calculates the SLD, ρ, using ρ = (∑inbi)/(vm), where bi is the coherent scattering length for atom i and vm is the molecular volume calculated from the known density. For neutrons, the coherent scattering lengths, bi, are obtained from tabulated values,28 and for X-rays bi = Zre, where Z is the atomic number of the ith atom and re = 2.81 × 10−15 m is the classical radius of the electron. SLDs used for fitting are given in Table S1 (Supporting Information).

■

RESULTS The core/shell synthesis utilized in this study involves two steps. PbSe core QDs are first grown using previously reported hot-injection methods.30,31 The PbSe/CdSe core/shell QDs are then formed by cation exchange as reported previously.13,17 Figure 1 shows a typical absorption spectrum of the initial PbSe 3970

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

Figure 2. XRD Rietveld refinement plots of (a) PbSe core (λabs = 1700 nm) and (b) PbSe/CdSe core/shell (λabs = 1320 nm) QDs. Dotted lines are the observed intensity at each step; green lines are the calculated patterns; solid gray lines are the difference between observed and calculated intensities. Red (PbSe) and blue (CdSe) lines are the individual diffraction pattern contributions. For the sample shown in a, a crystallite size of ∼5 nm was calculated. For the sample in b, crystallite sizes of ∼3 nm were calculated for each of the PbSe Fm3m̅ (42%) and CdSe F4̅3m (58%) contributions. Cationexchange reaction was performed at 110 °C for 25 min.

Figure 3. HR-TEM images of PbSe/CdSe core/shell QDs (λabs = 1150 nm). Circled particles show clear contrast between core and shell materials. Initial PbSe core precursor QDs had λabs = 1590 nm. Cationexchange reaction performed at 105 °C for 25 min.

particular, at a given HR-TEM focus, only certain particles will exhibit contrast that reveals a core/shell structure, and this will be highly dependent on their orientation with respect to the direction of the electron beam. Many other such nanoparticles will exhibit a uniform contrast with no indication of core/shell interfaces. This makes it difficult to interpret whether a core/ shell structure has formed in all particles, as Hens and coworkers pointed out in their HR-TEM study.15 Moreover, HRTEM does not provide elemental identification, making it difficult to draw definitive conclusions on core and shell dimensions, as well as alloying at the PbSe/CdSe inorganic interface. A few examples of HR-TEM images are provided in Figure 3 (lower magnification images are shown in Figure S1 of the Supporting Information), comparable to the results presented by Hens and co-workers. For the circled particles, both core and shell structures are observed, but for the remaining particles no conclusion can be drawn because the particles are not in an orientation amenable to reveal contrast from the core/shell interface. This has obvious disadvantages when trying to understand and characterize these materials, particularly because there is no increase in particle size after

cation exchange. It is also noteworthy that the dimensions of the core may be underestimated using this technique because the image contrast is the result of phase interference of the electron beam. Delocalization of the electron beam at the interface between the core and the shell materials results in a false contrast around the interface. Delocalization can easily occur when using a highly coherent electron beam, such as the field emission gun TEM used here. On the other hand, the information from TEM is specific and localized to a particular nanoparticle. These considerations warrant further characterization with complementary techniques to explore whether a complete core/shell had formed on all particles or if only a fraction of them possess the core/shell structure. For the above reasons, high-angle annular dark field (HAADF) imaging in scanning TEM (STEM) mode was utilized. Although in many cases the STEM technique poses more challenges with beam damage to the sample than standard TEM, the method is highly sensitive to the atomic number (Z) of the materials, scaling proportionally to ∼Z2.32 For this reason, HAADF imaging is also referred to as Z3971

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

contrast imaging. For the materials discussed herein, Pb has Z = 82 and Cd has Z = 48. This makes it possible to visualize core and shell materials as shown in Figure 4. From the HAADF

Figure 4. HAADF image of PbSe/CdSe core/shell QDs. Both core (bright) and shell (dark) contrast are visible in the image. QDs are the same batch as those presented in Figure 3.

images shown (same batch of QDs as in Figure 3) it was possible to estimate a core size of 2.7 ± 0.3 nm and a shell thickness of ∼1.2 nm (overall particle size 5.1 ± 0.4 nm). Figure S2 (Supporting Information) shows histograms of the size distributions for the core and overall particle size. The increase in relative standard deviation of the core size from the HAADF images is indicative of an increase in heterogeneity of the sample caused by the cation-exchange process. The sharpness of the interface however is difficult to discern with accuracy, and an aberration-corrected microscope may help further in this regard.32,33 HAADF images alone provide qualitative information only. Sensitivity to particle thickness can influence the contrast in the image.34 For this reason, the particles were examined further by energy-filtered TEM (EF-TEM). In EF-TEM imaging, only electrons of particular kinetic energies are used to form the image. Inelastic scattering of electrons and the accompanied loss of energy are characteristic of the element in the sample; thus, the formed image in EF-TEM is element specific.35 EFTEM imaging also allows for a larger number of particle statistics by directly imaging elements in the sample, as compared to individual particle line scans performed by electron energy-loss spectroscopy (EELS) or by energydispersive X-ray spectroscopy (EDS). Figure 5 shows the EF-TEM image for Pb and the corresponding bright-field image for the core/shell QDs (same QDs shown in Figures 3 and 4). EF-TEM imaging was performed on the Pb O2,3 edge at 86 eV and the Se M4,5 edge at 57 eV. It was not possible to measure reliably the Cd signal (M4,5 edge at 404 eV) because of low signal intensities at higher energies. Detailed statistics revealed a Pb core size of 3.3 ± 0.4 nm (EF-image) and an overall particle size of 5.2 ± 0.5 nm (bright-field image), comparable to the results obtained from the HAADF images. Histograms of the size distributions are shown in Figure S3 of the Supporting Information. Again, a comparable relative increase in core size standard deviation was measured, confirming an increase in heterogeneity after cation

Figure 5. (a) Pb map by EF-TEM imaging of PbSe/CdSe core/shell QDs. (b) Bright-field image of the same core/shell particles measured af ter the EF-TEM image. QDs are the same batch as those presented in Figures 3 and 4.

exchange. The Se image (not shown) confirmed the same particle size for both the EF-TEM image and the bright-field image. We note that EF-TEM imaging requires longer scan times than standard TEM imaging; thus, higher magnifications resulted in significant beam damage to the particles. The magnification utilized is a compromise between resolution and avoiding beam damage. To improve signal-to-noise in the image, 20 particles from each of the Pb map and bright-field images were summed together, rather than imaging at higher magnifications. This result is shown and detailed in Figure S4 of the Supporting Information and makes clear the distinct particle size difference between the particles in the Pb-only image and those in the bright-field image. Energy-Dependent XPS. Angle-resolved XPS (AR-XPS) is a proven technique for nanometer depth profiling of samples by taking advantage of the angle dependence of the inelastic mean free path (IMFP) of the emitted photoelectrons.36 However, AR-XPS is unable to depth profile nanomaterials with spherical structures. In contrast, energy-dependent XPS is able to depth profile because the IMFP also depends on the kinetic energy of 3972

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

the photoelectrons.37 Thus, by gradually increasing the X-ray photon energy, measurements at increasing sampling depth can be performed. Energy-dependent XPS cannot be performed on a standard XPS instrument because of the monochromatic Xray source and therefore requires the use of continuous synchrotron radiation (where energy selection is performed with a monochromator). XPS spectra were measured at varying photon energies using tunable synchrotron radiation (Canadian Light Source, Inc.) to confirm the core/shell structure measured by HAADF and EFTEM imaging. XPS allows for direct observation of both Pb and Cd signals that was not possible in EF-TEM (the Cd signal was too weak in that case), as well as sampling of a much larger number of particles. For a core/shell-type structured nanomaterial, the photoelectrons from the core will be screened by the shell material at low kinetic energies. A detailed analysis of the photoemission peaks from Pb and Cd allows for determination of a core/shell structure or alloyed (Pb1−xCdxSe) material. A uniformly alloyed material is expected to have a constant Pb/ Cd ratio at different kinetic energies. Figure S5 (Supporting Information) shows examples of Pb and Cd photoemission peaks at similar kinetic energies with polynomial background fits. After background subtraction, the XPS peaks were integrated and normalized for their absorption cross sections38,39 and for the photon flux. From a detailed analysis of the photoemission peaks, Figure S6 (Supporting Information) was made by plotting the Pb/Cd ratio as a function of kinetic energy. The increasing Pb/Cd ratio with increasing photoelectron kinetic energy strongly suggests that Pb is in the core and Cd is on the surface because the Pb signal is screened at low kinetic energy, thus confirming the results obtained by the HAADF and EF-TEM imaging techniques. Due to the small particle size and small lattice mismatch between PbSe and CdSe materials, the microscopy techniques described in the previous section are unable to probe directly the interface between the core and shell materials. The composition, structure, and nature of the interface between core and shell materials have a strong influence on the electronic properties of nanomaterials.40,41 Recent 3D atomic characterization work on larger rod-shaped PbSe/CdSe core/ shell heteronanocrystals using HAADF-STEM has attempted to probe the interfacial layer between core and shell materials.33 In that work, the authors found evidence for interfacial mixing of Pb and Cd atoms. XPS is highly sensitive to chemical speciation (changes in chemical state) and combined with the high-energy resolution and the high brilliance of the SGM beamline (Canadian Light Source, Inc.) allows for depth sensitivity down to a few Angstroms.42,43 By varying the X-ray photon energy and scanning with high-energy resolution it was possible to vary the surface sensitivity to probe variations in chemical speciation throughout the nanoparticle.44 Any change in chemical speciation is accompanied by a small shift in binding energy. From a detailed analysis of the high-resolution XPS photoemission spectra at different photon energies, a structural model was generated for each of the PbSe core and PbSe/CdSe core/shell QDs. The conclusions drawn from this study suggest that the QDs are structurally much more complicated than previously assumed in the literature. Pb 4f, Cd 3d, and Se 3d photoemission peaks were measured at different photon energies for both the PbSe core-only and the PbSe/CdSe core/shell QDs at high-energy resolution (HRXPS) (∼100 meV resolution). HR-XPS spectra for Cd 3d and Se 3d photoemission peaks are shown in Figures S7 and S8

(see Supporting Information), respectively. The Cd 3d photoemission peaks (Cd 3d5/2 binding energy of 405.3 eV) match with those for CdSe, confirming that the measured Cd signal is primarily associated with CdSe. No Cd signal intensity was measured for the PbSe QDs, as expected. The Se 3d photoemission peaks from the PbSe QD sample correspond to PbSe material (Se 3d5/2 binding energy of 54.1 eV). The Se 3d3/2 and 3d5/2 peaks from the PbSe/CdSe core/shell sample are broadened and overlap, making it hard to resolve them due to a Se 3d signal that originates from both PbSe and CdSe materials. More informative are the Pb 4f photoemission peaks shown in Figure 6. At low photon energies (i.e., surface sensitive) the

Figure 6. HR-XPS spectra of Pb 4f photoemission peaks at two different photon energies. The asterisked (**) peaks are attributed to Pb−oleates and PbO on the surface of the PbSe QDs. Calculated IMFP values are ∼0.9 and ∼2.0 nm for the 450 and 975 eV photon energies, respectively. Cation-exchange reaction performed at 80 °C for 25 min. PbSe/CdSe QDs λabs = 1276 nm (initial PbSe QDs λabs = 1513 nm).

PbSe QDs have measured photoemission peaks that show a broad shoulder on the high binding energy side (Pb 4f7/2 binding energy of ∼138.9 eV). The broad shoulder is attributed to surface-bound PbO and Pb−oleate (the peaks are too convoluted to distinguish). At higher photon energies (bulk sensitive) the PbO and Pb−oleate peaks reduce in relative intensity quickly, confirming that they are located on the surface of the QDs. Surface PbO could have important implications on the optical properties of these materials, as 3973

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

discussed previously.17 It is important to note that the core/ shell QDs do not show this shoulder, confirming the absence of Pb−oleates and that the shell protects the core against (photo)oxidation (consistent with the increased photostability). At low photon energies the presence of phosphorus (binding energy of ∼133.6 eV) on the PbSe QDs was also observed and confirms that there is some trioctylphosphine (TOP) bound to the surface of the QDs as an additional secondary ligand (most of the surface-bound ligands are oleate molecules). TOP likely coordinates to Se sites on the surface of the particles and is in a significantly lower amount (confirmed by NMR45 and the rapid decrease in XPS intensity with increasing X-ray photon energy) compared to oleates that primarily coordinate to surface Pb2+ sites. At low photon energies the phosphorus peak is missing for the core/shell QDs because TOP was not utilized in the cation-exchange reaction. Mass balance (large excess of oleates) apparently removes most of the TOP from the surface of the final core/shell QDs. The main Pb 4f intensity signal is assigned as PbSe (Pb 4f7/2 binding energy of 137.4 eV) near the core of the QD for both core and core/shell materials. The shoulder at low binding energies on the Pb 4f peaks is of particular note. For the PbSe QDs (4f7/2 binding energy of ∼136.6 eV) the shoulder decreases with increasing X-ray photon energy, suggesting it is a surface-bound species. Previous energy-dependent XPS work on PbSe QDs by Sarma and co-workers (but with a much lower energy resolution) attributed this shoulder to a nonstoichiometric Pb1−xSe shell, which is consistent with the work presented here.46 However, it should be noted that reconstruction of the PbSe material near the surface of the QDs could also account for the additional shoulder (and the measured shift in binding energy). Shifts in binding energy associated with interface reconstruction have been reported for thin films.47 Therefore, we can only conclude that the PbSe material near the surface (denoted as PbSe* from here on to distinguish it from PbSe in the core) is in some form that is different from the PbSe core. Sarma and co-workers also reported evidence for a final outer layer of purely Se ions that are coordinated to a TOP organic capping layer. This will be discussed further below, but an outer shell of Se ions cannot account for the large excess of Pb in these samples (as confirmed by elemental analysis) and the low amount of TOP measured by NMR.45 No evidence for excess Se ions as an outer shell was found in our work. For the core/shell QDs, the relative intensity of the shoulder decreases with increasing X-ray photon energy as well but more so than the shoulder for the PbSe core-only QDs over the same energy range. Figure S9 of the Supporting Information shows an example of our attempts to fit the HR-XPS spectra (see Experimental Section for details). From the curve fits, the area of the shoulder peak was compared to the main PbSe peak at four different photon energies for both the PbSe core-only and PbSe/CdSe core/shell QDs (see Figure S10, Supporting Information). This result suggests that this layer is thinner in the core/shell QDs than in the PbSe QDs. The peak position of the shoulder is also shifted to slightly lower binding energy (4f7/2 binding energy of ∼136.4 eV), consistent with a new chemical species. This is also consistent with shell formation because the CdSe shell is much thicker than the initial PbSe* shell, so there can no longer be any PbSe* material present in the core/shell QDs after cation exchange. Thus, the shoulder in the core/shell XPS spectra likely originates from a newly

formed species. The interface between the PbSe core and CdSe shell is therefore assumed to be composed of a very thin interfacial layer of Pb1−xCdxSe material (i.e., interfacial mixing of Pb and Cd atoms), which is consistent with both the shift in binding energy and the rapid decrease in relative intensity with increasing X-ray photon energy. This is consistent with recent 3D atomic characterization work on larger rod-shaped PbSe/ CdSe core/shell heteronanocrystals using HAADF-STEM.33 However, it should be noted that it is not necessary that the interface be composed of Pb1−xCdxSe material; interfacial strain of PbSe could also lead to the shift in binding energy (discussed further below). SAXS and SANS. Small angle scattering techniques probe the size, shape, and structure of samples with length scales in the range of 0.530 nm.48 Small angle X-ray scattering (SAXS) and small angle neutron scattering (SANS) are two complementary techniques which cover roughly the same size range but provide information about different aspects of a system based on the different contrast of X-rays versus neutrons. In this work, SAXS is used to probe the inorganic component of the particles (where the oleate stabilizer is essentially invisible to the X-rays) while SANS is used to probe the organic part (where hydrogenous matter, such as the oleates, provide good neutron contrast). Examples are shown in Figure 7, where the SANS and SAXS data for individual samples are plotted on the same graph for ease of comparison. The scattering intensity, I(q), as a function of the scattering vector, q (q = ((4π)/λ)sin(θ/2), where λ is the wavelength and θ is the scattering angle), is modeled as I(q) = NPP(q), where NP is the number density of particles and P(q) is the particle form factor (modified to include the scattering length density difference, see below). The SAXS data was fit using a singlesphere model, where P(q) is given by P sphere (q) = [(3VΔρ)((sin(qr) − qr cos(qr))/(qr)3)]2, where r is the particle radius, V is the particle volume, and Δρ is the scattering length density difference (see Experimental Section).49 The SANS data was fit using an inorganic-core/ organic-shell sphere model, where P(q) is given by PCSsph(q) = [(3V c Δρ shell−core )((sin(qr c ) − qr c cos(qr c ))/(qr c ) 3 ) + (3VsΔpsolv−shell)((sin(qrs) − qrs cos(qrs))/(qrs)3)]2, where rc is the core radius, rs = rc + shell thickness, Δρi is the scattering length density difference (see Experimental Section), and Vi is the volume of the core (i = c) or the volume of the total particle including the shell (i = s).49 Instrumental smearing of the data and particle polydispersity were accounted for by convolution of the model scattering with an appropriate distribution function (see Experimental Section). SAXS data for the QDs were fit with a single-sphere model to obtain the inorganic particle diameter, e.g., data in Figure 7a gave a diameter of 5.3 nm, and this value was then fixed in the core/(organic)shell SANS fits (discussed below). The SAXS fits could not separate the PbSe/CdSe core/shell structure and interface due to the similarity in their scattering length densities (i.e., the core and shell materials look almost the same to the Xrays, see Figure S11 in Supporting Information). The polydispersities of the PbSe particles (e.g., Figure 7a) are indistinguishable from the SAXS beam spread, indicating near monodispersity. The polydispersities for the PbSe/CdSe core/ shell particles (e.g., Figure 7b) are similarly small and demonstrate that addition of a thin CdSe shell by cation exchange does not produce a significant increase in overall particle polydispersity. This is in contrast with growing a shell by epitaxial growth methods; for example, Dabbousi et al. found 3974

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

on a TEM grid do not necessarily represent how the ligands behave in dispersion. The SANS data for these particles were initially fit as single core/shell particles, i.e., inorganic core and oleate shell, but failed to account for the shoulder at q ≈ 0.1 Å−1 (see dashed curve in Figure 7a). Both the shape of this shoulder and the position where it hits background are consistent with an extra population of small particles with a diameter of ∼2.6 nm. Further, the presence of this shoulder in the SANS but not in the SAXS demonstrates that these extra particles are organic (i.e., not a second population of PbSe QDs which would be visible in the X-ray scattering experiments). This suggests the presence of free oleate/oleic acid in the toluene solution, and this was confirmed with additional SANS measurements on a pure oleic acid sample that reproduced the shoulder scattering as shown in Figure S12 of the Supporting Information. Thus, the SANS fits were modified to account for these additional aggregates, that is, the data was fit to core/shell spheres (i.e., oleate-coated PbSe particles) + small oleic acid aggregates. From a detailed analysis of the SANS fits, the oleate coating was found to have a shell thickness of ∼1.8 nm. These results are consistent with the cis double bond in the oleate that bends the ligand partially (the cis double bond oleate length is estimated to be ∼1.9 nm). On average, the PbSe QDs have a slightly smaller oleate shell thickness (1.7 ± 0.2 nm) than the PbSe/CdSe QDs (1.8 ± 0.1 nm), which suggests that they may not have the same average surface density of oleates. Complete Cation Exchange. For core/shell QDs prepared by cation exchange, control over shell thickness is achieved primarily by reaction temperature and time (Cd− oleate is held in excess). An example of shell thickness control by temperature is shown in Figure S13 (Supporting Information). Formation of CdSe QDs by complete cation exchange can occur at elevated temperatures (∼180 °C) under the conditions17 utilized here. Previous work by Hollingsworth and co-workers reported that the exchange is a self-limiting process with a terminal shell thickness of ∼1.5 nm; however, they primarily investigated reaction times at a single temperature (100 °C) with varying reaction times. The absorption spectrum (shown in Figure S14 (Supporting Information)) of completely exchanged QDs (from PbSe to CdSe) has distinct excitonic peaks in the visible region that are indicative of a highquality sample. The XRD pattern of CdSe QDs formed by cation exchange is also shown in Figure S14 (Supporting Information). The pattern matches with the bulk zinc blende CdSe structure. Retention of the cubic structure, rather than formation of the more common wurtzite (hexagonal) CdSe structure, strongly suggests that the anion sublattice is maintained with fast exchange of Pb2+ cations for Cd2+. This is further supported by TEM images that confirm the same particle size before and after cation exchange, regardless of reaction temperature (from 30 to 180 °C), shown in Figure S15 (Supporting Information). PL Lifetimes. In recent work, Hens and co-workers argued that PbSe/CdSe QDs have a lower lying emission state with reduced oscillator strength, as compared to the initial PbSe QDs (but the absorption states are equivalent).18 Figure 8 are the PL lifetime measurements of typical PbSe and PbSe/CdSe QDs presented here. The increase in PL lifetime (a factor of ∼2.5) is consistent with the work of Hens and co-workers.18 The nature of the interface (i.e., the interfacial mixing of Pb and Cd atoms) between the PbSe core and CdSe shell materials will certainly have a strong influence on the electronic properties of

Figure 7. (a) SAXS (circles) and SANS (squares) data for PbSe QDs (λabs = 1475 nm) with fits (solid lines). Dashed line is the fit prior to including oleic acid aggregates. (b) SAXS (circles) and SANS (squares) data for PbSe/CdSe core/shell QDs (λabs = 1280 nm) with fits (solid lines). Initial PbSe core precursor QDs had λabs = 1588 nm. Cation-exchange reaction performed at 85 °C for 25 min.

a significant increase in particle polydispersity when they grew ZnS shells of only 5.3 monolayers directly onto a CdSe core.14 Measured trends in total inorganic particle size from SAXS are consistent with the microscopy results discussed above. On average, SAXS measured a slightly larger particle size (from ∼0.1 to 0.5 nm larger) for both PbSe and PbSe/CdSe, which could be the result of excess Pb−oleate (or Cd−oleate for the core/shell QDs) on the surface of the particles. It might also be possible that the particles are not perfectly spherical, and so the SAXS model requires a slightly larger diameter to account for the difference. Most of the techniques (i.e., HR-TEM, STEM, EF-TEM, and SAXS) described above probe the inorganic composition of the nanoparticles. Structural information about the organic surfactant layer is virtually nonexistent in the literature for lead chalcogenide QDs. It is important to understand the structure of the organic capping layer as it plays an important role in device applications, whereby the QDs are often integrated with polymer blends (e.g., in solar cells).50 SANS was used to probe the outer organic shell and determine the thickness of the surface ligand layer. Conventional techniques that are used to probe particle size (such as TEM) are unable to provide information about the organic layer, and therefore, SANS is a unique technique that allows direct measurements to be performed while the QDs are dispersed in solvent. Estimates of ligand length based on interparticle distances of QDs dried 3975

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

thickness were roughly approximated from the PbSe*/PbSe XPS spectra peak intensities and are consistent within error with the other characterization techniques. Although not shown in Scheme 1, there is also evidence for PbO and TOP on the surface of the QDs (possibly only on certain facets). The surface species PbSe* and PbO may be responsible for the reduced PL yield observed for PbSe QDs, as compared to the PbSe/CdSe QDs.17 The excess Pb2+ cations on the surface are consistent with elemental analysis, which consistently showed excess lead present in these samples.17 The excess positive charge must be compensated for by oleate ligands (not TOP). There has been some recent debate in the literature regarding the actual state of protonation of the capping molecules.51 A previous NMR study argued that the primary capping ligands for PbSe QDs are oleic acid molecules coordinated to excess neutral Pb atoms.45 However, in that study no Fourier transform infrared (FTIR) measurements were carried out. A FTIR spectrum of the PbSe QDs studied herein is provided in Figure S16 (Supporting Information) and strongly supports Pb−oleates bound to the surface of the PbSe QDs, in agreement with a recent independent study.52 We therefore conclude that PbSe QDs are terminated with excess Pb2+ bound to an oleate capping layer. These results also contradict the work of Sarma and co-workers, who suggested that the PbSe QDs are terminated with Se ions and a passivating layer of entirely TOP molecules (not oleates).46 The SAXS and SANS results are also consistent with the proposed structure and prove that the oleate ligands are not fully extended most likely because of their cis double bond (thickness of 1.7 ± 0.2 nm). These new results on the organic capping layer should prove useful for theoretical modeling studies.51 We note that variations in QD synthetic preparation methods as well as changes in QD size may have a strong influence on the final QD structure and on surface reconstruction. Scheme 2 is a schematic representation of a typical PbSe/ CdSe core/shell QD. The thickness of the outer CdSe shell (and likely the interface) is controlled by the reaction temperature and time. HAADF images, EF-TEM imaging, and energy-dependent XPS all confirm that Pb is located in the core of the particle and Cd is in the shell, thus supporting a core/shell structure and not a uniformly alloyed material. These results confirm that the PbSe/CdSe QDs are indeed all true core/shell QDs, with an overall increase in core size (and shell thickness) dispersion caused by the cation-exchange reaction. The increase in size dispersion measured by HAADF and EFTEM imaging seems to be associated with the initial size distribution of the PbSe core precursor material. For instance, it was consistently found that batches of smaller PbSe QDs blue shifted more than batches with larger particle sizes under the same reaction conditions (Figure S17, Supporting Information). Although variations can exist in colloidal solution reactions, this result suggests that the surface free energy of the particles plays a very important role in the kinetics of the cation-exchange reaction. It is therefore reasonable to assume that smaller QDs within an ensemble of QDs exchange Pb2+ for Cd2+ at a faster rate than the larger QDs, and thus, the resulting core/shell QDs will have a greater overall core size (and shell thickness) dispersion. It is also noteworthy that the QDs presented here typically show complete and uniform core/shell structures, despite the increase in core size (and shell thickness) dispersion. This is contradictory to previous works15,19 that show incomplete shell growth on some particles as well as faster cation exchange on

Figure 8. Photoluminescence lifetime measurements of PbSe coreonly and PbSe/CdSe core/shell QDs in solution (tetrachloroethylene). Lifetimes of 1.0 ± 0.1 and 2.6 ± 0.3 μs were measured for PbSe and PbSe/CdSe QDs, respectively. Additional fast component (254 ± 25 ns) was measured for the core/shell QDs.

these materials and could explain the observed increase in PL lifetimes (e.g., new sub-band gap states).40,41

■

DISCUSSION Scheme 1 is a schematic representation of a typical PbSe QD that is consistent with the results obtained by high-resolution Scheme 1. Schematic of a Typical PbSe QDa

a

The QD has a central PbSe core and a thin PbSe* shell. PbSe* is structurally (and/or chemically) different than the PbSe core. The outer layer is composed of excess Pb2+ cations coordinated to oleate ligands. Although not shown, there is also evidence for PbO and TOP ligands present on the surface. Dimensions vary with reaction conditions (and QD size) and are approximations based on the described characterization techniques.

XPS as well as microscopy, SANS, and elemental analysis. Both the PbSe core and the outer PbSe* layer were confirmed by the XPS measurements. The total particle size was taken from HRTEM images, while the PbSe core size and outer PbSe* layer 3976

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

certain facets of the nanocrystal. Furthermore, the particle dimensions (i.e., the PbSe* and Pb1−xCdxSe thickness) calculated from the XPS data can be taken only as rough approximations due to the large number of variables in the calculation, and should be treated with care. Despite this, we identified by direct measurement the interfacial mixing and/or strain at the interface of the PbSe/CdSe QDs as well as surface reconstruction on the PbSe core-only QDs, which will prove useful in future simulations. These estimated dimensions need confirmation by other techniques. For instance, an aberrationcorrected microscope will be necessary to identify the exact thickness on each facet of the particles and to correlate that data with the reaction conditions. It is our hope that the new characterization results presented here will be used in simulations of the electronic structure of these materials and will help to elucidate the nature of the measured optical properties as well as the measured surface and interface reconstruction.

Scheme 2. Schematic of a Typical PbSe/CdSe Core/Shell QDa

■

CONCLUSIONS A detailed investigation of the morphology of core/shell QD materials formed by cation exchange has been presented. HRTEM, HAADF, and EF-TEM images support a core/shell structure for the PbSe/CdSe QDs. These results were further substantiated by energy-dependent XPS, which showed an increasing Pb/Cd ratio with increasing kinetic energy, consistent with Cd present on the surface of the QDs (and Pb in the core). From a detailed analysis of the high-resolution X-ray photoemission peaks, it is shown that PbSe QDs likely have some PbO present on their surface but a CdSe shell helps protect the QDs from oxidation. Furthermore, there is strong evidence for a complex internal structure for both PbSe and PbSe/CdSe QDs that includes a shell for the PbSe QDs (of surface PbSe* that is chemically and/or structurally different from the PbSe in the core) and a sharp Pb1−xCdxSe (or strained PbSe) interface for the PbSe/CdSe core/shell QDs. SAXS data further confirmed the particle sizes measured by the microscopy techniques (but with greater particle statistics) but could not confirm the structural complexity measured by XPS because of reduced resolution. For the first time, the organic surfactant layer was structurally investigated for lead chalcogenide QDs using SANS methods. SANS analysis has shown that there is indeed a well-solvated oleate ligand shell on the surface of the PbSe and PbSe/CdSe QDs with a thickness of ∼1.8 nm. All of the experimental results were combined to provide structural models for the PbSe and PbSe/CdSe QDs that include both inorganic and organic materials, with greater complexity than previously reported. Simulations that take into consideration the structural information obtained by these methods could help explain the origin of the observed optical properties of these materials and the measured increase in PL lifetimes.

a

The QD has a central PbSe core and an outer purely CdSe shell. There is evidence for a sharp interface that is composed of PbSe structurally and/or chemically different from the inner core (labeled as Pb1−xCdxSe). The outer layer is composed of excess Cd2+ cations coordinated to oleate ligands. There are no TOP ligands or PbO present. Dimensions vary with reaction conditions and are approximations based on the characterization techniques.

certain facets.15,33 A small number of the particles measured by HAADF imaging (shown in Figure 4) have a higher contrast region (corresponding to PbSe) that extends to the edge of the particle, which suggests that some particles may show preferred growth on certain planes, in agreement with previous works. The higher contrast could, in principle, also be in part due to strain. Without further measurements and simulations it is impossible to draw firmer conclusions. However, this effect seems to be reduced to a significant extent in our core/shell materials because the majority of the particles have complete and almost uniform shells. Small differences in reaction conditions likely play an important role in shell formation and suggest that further control over shell growth may be possible by careful control over the reaction conditions. Although the interface between the core and the shell is challenging to image directly, high-resolution X-ray photoemission scans suggest that the interface is sharp (due to a rapid decrease in signal with increasing X-ray photon energy) and composed of a PbSe material that is structurally and/or chemically different than the core (likely as Pb1−xCdxSe or strained PbSe). The Pb1−xCdxSe interface thickness is roughly approximated from the Pb1−xCdxSe/PbSe XPS spectra peak intensities, while the total particle size and CdSe shell thickness were taken from the microscopy techniques described above. Elemental analysis shows excess cadmium in the samples, which is consistent with surface-bound Cd2+ cations. The surface is assumed to be capped with Cd−oleates, as no signal from TOP was measured by XPS. SAXS and SANS results demonstrate that the oleate ligands have a thickness of 1.8 ± 0.1 nm and are consistent with the cis double-bond oleate length (∼1.9 nm). It is important to note that for Scheme 1 and 2 the schematic representations show uniform PbSe* and Pb1−xCdxSe thickness, but there could be more or less of these materials on

■

ASSOCIATED CONTENT

S Supporting Information *

Further materials characterization and supporting figures described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. 3977

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978

The Journal of Physical Chemistry C

Article

Notes

(22) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-Ray Photoelectron Spectroscopy; Physical Electronics, Inc.: Chanhassen, 1995. (23) Tanuma, S.; Powell, C. J.; Penn, D. R. Surf. Interface Anal. 1991, 17, 927−939. (24) Schulz, G. V. Z. Phys. Chem. 1935, 43, 25. (25) Mattoussi, H.; Cumming, A. W.; Murray, C. B.; Bawendi, M. G.; Ober, R. Phys. Rev. B 1998, 58, 7850−7863. (26) Gilbert, E. P.; Schulz, J. C.; Noakes, T. J. Physica B 2006, 385− 386, 1180−1182. (27) Kline, S. R. J. Appl. Crystallogr. 2006, 39, 895−900. (28) Sears, V. F. Neutron News 3 1992, 3, 26−37. (29) Munter, A. Scattering Length Density Calculator. http://www. ncnr.nist.gov/resources/sldcalc.html. (30) Murray, C. B.; Sun, S. H.; Gaschler, W.; Doyle, H.; Betley, T. A.; Kagan, C. R. IBM J. Res. Dev. 2001, 45, 47−56. (31) Yu, W. W.; Falkner, J. C.; Shih, B. S.; Colvin, V. L. Chem. Mater. 2004, 16, 3318−3322. (32) Sanchez, S. I.; Small, M. W.; Sivaramakrishnan, S.; Wen, J.-G.; Zuo, J.-M.; Nuzzo, R. G. Anal. Chem. 2010, 82, 2599−2607. (33) Bals, S.; Casavola, M.; van Huis, M. A.; Van Aert, S.; Batenburg, K. J.; Van Tendeloo, G.; Vanmaekelbergh, D. Nano Lett. 2011, 11, 3420−3424. (34) Yu, Z.; Guo, L.; Du, H.; Krauss, T.; Silcox, J. Nano Lett. 2005, 5, 565−570. (35) Williams, D. B.; Carter, C. B. Transmission Electron Microscopy; Springer Science+Business Media: New York, 2009. (36) Briggs, D.; Sheah, M. P. Practical Surface Analysis: Auger and Xray photoelectron spectroscopy, 2nd ed.; John Wiley: New York, 1990; Vol. 1. (37) Borchert, H.; Haubold, S.; Haase, M.; Weller, H.; McGinley, C.; Riedler, M.; Möller, T. Nano Lett. 2001, 2, 151−154. (38) Yeh, J. J. Atomic Calculation of Photoionization Cross-Sections and Asymmetry Parameters; Gordon and Breach Science Publishers: Langhorne, PE, 1993. (39) Yeh, J. J.; Lindau, I. At. Data. Nucl. Data Tables 1985, 32, 1− 155. (40) Chen, X.; Lou, Y.; Samia, A. C.; Burda, C. Nano Lett. 2003, 3, 799−803. (41) Yang, S.; Prendergast, D.; Neaton, J. B. Nano Lett. 2010, 10, 3156−3162. (42) Regier, T.; Paulsen, J.; Wright, G.; Coulthard, I.; Tan, K.; Sham, T. K.; Blyth, R. I. R. AIP Conf. Proc. 2007, 879, 473−476. (43) Regier, T.; Krochak, J.; Sham, T. K.; Hu, Y. F.; Thompson, J.; Blyth, R. I. R. Nucl. Instrum. Methods Phys. Res., Sect. A 2007, 582, 93− 95. (44) Huang, K.; Demadrille, R.; Silly, M. G.; Sirotti, F.; Reiss, P.; Renault, O. ACS Nano 2010, 4, 4799−4805. (45) Moreels, I.; Fritzinger, B.; Martins, J. C.; Hens, Z. J. Am. Chem. Soc. 2008, 130, 15081−15086. (46) Sapra, S.; Nanda, J.; Pietryga, J. M.; Hollingsworth, J. A.; Sarma, D. D. J. Phys. Chem. B 2006, 110, 15244−15250. (47) Wadati, H.; Hawthorn, D. G.; Geck, J.; Regier, T. Z.; Blyth, R. I. R.; Higuchi, T.; Hotta, Y.; Hikita, Y.; Hwang, H. Y.; Sawatzky, G. A. Appl. Phys. Lett. 2009, 95, 023115. (48) Zemb, T.; Lindner, P. Neutron, X-ray and Light Scattering; Elsevier: Amsterdam, 1990. (49) Guinier, A.; Fournet, G. Small-Angle Scattering of X-rays; John Wiley and Sons: New York, 1955. (50) Nann, T.; Skinner, W. M. ACS Nano 2011, 5, 5291−5295. (51) Argeri, M.; Fraccarollo, A.; Grassi, F.; Marchese, L.; Cossi, M. J. Phys. Chem. C 2011, 115, 11382−11389. (52) Sliem, M. A.; Chemseddine, A.; Bloeck, U.; Fischer, R. A. CrystEngComm 2011, 13, 483−488.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Natural Sciences and Engineering Research Council (NSERC) of Canada, the Canada Foundation for Innovation (CFI), the British Columbia Knowledge Development Fund (BCKDF), and an NRCNSERC-BDC Nanotechnology Initiative Research Fund. Research described in this paper was performed at the Canadian Light Source, which is supported by NSERC, NRC, the Canadian Institutes of Health Research, the Province of Saskatchewan, Western Economic Diversification Canada, and the University of Saskatchewan. Research was also performed at the Australian Nuclear and Science Technology Organisation (ANSTO). The authors gratefully acknowledge the support of staff and facilities of the Australian Microscopy & Microanalysis Research Facility (AMMRF) node at the University of Sydney. F.C.J.M.v.V. is thankful for an ‘International Visiting Research Fellowship’ from the University of Sydney.

■

REFERENCES

(1) Talapin, D. V.; Lee, J.-S.; Kovalenko, M. V.; Shevchenko, E. V. Chem. Rev. 2009, 110, 389−458. (2) Ruan, G.; Winter, J. O. Nano Lett. 2011, 11, 941−945. (3) Singhal, S.; Nie, S. M.; Wang, M. D. Annu. Rev. Med. 2010, 61, 359−373. (4) Nozik, A. J. Nano Lett. 2010, 10, 2735−2741. (5) Sargent, E. H. Nat. Photonics 2009, 3, 325−331. (6) Ladd, T. D.; Jelezko, F.; Laflamme, R.; Nakamura, Y.; Monroe, C.; O’Brien, J. L. Nature 2010, 464, 45−53. (7) Yao, P. J.; Rao, V.; Hughes, S. Laser Photonics Rev. 2010, 4, 499− 516. (8) Son, D. H.; Hughes, S. M.; Yin, Y.; Alivisatos, A. P. Science 2004, 306, 1009−1012. (9) Jain, P. K.; Amirav, L.; Aloni, S.; Alivisatos, A. P. J. Am. Chem. Soc. 2010, 132, 9997−9999. (10) Luther, J. M.; Zheng, H.; Sadtler, B.; Alivisatos, A. P. J. Am. Chem. Soc. 2009, 131, 16851−16857. (11) Robinson, R. D.; Sadtler, B.; Demchenko, D. O.; Erdonmez, C. K.; Wang, L. W.; Alivisatos, A. P. Science 2007, 317, 355−358. (12) 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, 5285−5293. (13) Pietryga, J. M.; Werder, D. J.; Williams, D. J.; Casson, J. L.; Schaller, R. D.; Klimov, V. I.; Hollingsworth, J. A. J. Am. Chem. Soc. 2008, 130, 4879−4885. (14) Dabbousi, B. O.; Rodriguez-Viejo, J.; Mikulec, F. V.; Heine, J. R.; Mattoussi, H.; Ober, R.; Jensen, K. F.; Bawendi, M. G. J. Phys. Chem. B 1997, 101, 9463−9475. (15) Lambert, K.; De Geyter, B.; Moreels, I.; Hens, Z. Chem. Mater. 2009, 21, 778−780. (16) Wise, F. W. Acc. Chem. Res. 2000, 33, 773−780. (17) Abel, K. A.; Qiao, H.; Young, J. F.; van Veggel, F. C. J. M. J. Phys. Chem. Lett. 2010, 1, 2334−2338. (18) De Geyter, B.; Justo, Y.; Moreels, I.; Lambert, K.; Smet, P. F.; Van Thourhout, D.; Houtepen, A. J.; Grodzinska, D.; de Mello Donega, C.; Meijerink, A.; et al. ACS Nano 2010, 5, 58−66. (19) Swart, I.; Sun, Z.; Vanmaekelbergh, D.; Liljeroth, P. Nano Lett. 2010, 10, 1931−1935. (20) Maikov, G. I.; Vaxenburg, R.; Sashchiuk, A.; Lifshitz, E. ACS Nano 2010, 4, 6547−6556. (21) Trinh, M. T.; Polak, L.; Schins, J. M.; Houtepen, A. J.; Vaxenburg, R.; Maikov, G. I.; Grinbom, G.; Midgett, A. G.; Luther, J. M.; Beard, M. C.; et al. Nano Lett. 2011, 11, 1623−1629. 3978

dx.doi.org/10.1021/jp2112928 | J. Phys. Chem. C 2012, 116, 3968−3978