ARTICLE pubs.acs.org/Langmuir
Impact Dynamics of Colloidal Quantum Dot Solids Lejun Qi,†,§ Peter H. McMurry,‡ David J. Norris,*,†,|| and Steven L. Girshick*,‡ †
Department of Chemical Engineering and Materials Science and ‡Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota 55455, United States
bS Supporting Information ABSTRACT: We use aerosol techniques to investigate the cohesive and granular properties of solids composed of colloidal semiconductor nanocrystals (quantum dot solids). We form spherical agglomerates of nanocrystals with a nebulizer and direct them toward a carbon substrate at low (∼0.01 m/s) or high (∼100 m/s) velocities. We then study the morphology of the deposit (i.e., the “splat”) after impact. By varying the size of the agglomerate and the spacing between the nanocrystals within it, we observe influences on the mechanical properties of the quantum dot solid. We observe a liquidto-solid transition as the nanocrystals become more densely packed. Agglomerates with weakly interacting nanocrystals exhibit liquidlike splashing and coalescence of overlapping splats. More dense agglomerates exhibit arching and thickening effects, which is behavior typical of granular materials.
C
olloidal semiconductor nanocrystals, or quantum dots, have unique electrical, chemical, and optical properties that can be useful for electronic and optoelectronic devices.1 6 In addition, nanocrystals can be easily deposited as films of close-packed assemblies from colloidal dispersions, which can potentially simplify the fabrication of these devices. However, in most applications the functionality and stability of the nanocrystals strongly depend on the details of the nanocrystal packing. Because nanocrystals are composed of an inorganic semiconductor core surrounded by a layer of organic capping ligands, films typically consist of quantum dots separated by organic molecules. Such materials have been termed nanocrystal solids or quantum dot solids.7 11 A better understanding of their properties is of fundamental interest and could also lead to improvements in potential devices and their fabrication. Although the optoelectronic characteristics of nanocrystal solids continue to be studied heavily, their mechanical behavior has been much less explored. Understanding their mechanical properties is important not only because they influence the structural integrity of devices but also because they exhibit new phenomena due to the nature of the colloidal assembly. For example, depending on the details of the solid, dramatically different mechanical characteristics have been reported in previous static measurements.12 15 Elastic behavior was observed in monolayer sheets of close-packed gold nanocrystals capped by dodecanethiol.15 In contrast, nanoindentation measurements showed that films of CdSe nanocrystals capped by trioctylphosphine oxide (TOPO) and trioctylphosphine (TOP) have viscoplastic properties.13 Furthermore, when the ligands between the nanocrystals were cross-linked, plastic behavior was observed. In particular, ultrathin films of CdSe nanocrystals crumpled like thin foils,12 and r 2011 American Chemical Society
spherical agglomerates of Au and Pt nanocrystals deformed like putty.14 In general, these differences can result from changes in the cohesion between the nanocrystals and the granular nature of the films. However, in these early studies the cohesion and granularity have not been fully explored. Cohesion was controlled only by the presence or absence of cross-linking. More importantly, only one report has discussed the granular nature of nanocrystal solids.13 In that case, the granularity of CdSe nanocrystals capped with TOPO and TOP became apparent only after the removal of the capping ligands. When the ligands are present, they dominate the mechanical behavior under the static or quasi-static conditions that have been utilized. Here we use aerosol techniques to investigate the cohesive and granular properties of CdSe nanocrystal solids. With this approach, we can tune the cohesion between the nanocrystals while simultaneously accessing conditions where the influence of the granularity can be examined. We produce quantum dot nanospheres, which are small pieces of a nanocrystal solid, by aerosolizing a colloidal dispersion of semiconductor nanocrystals. Each nanosphere contains multiple nanocrystals, their organic capping ligands, and possibly other organic residues from the colloidal dispersion.16,17 We then test the mechanical properties of these quantum dot nanospheres by impacting them on carbon substrates at various velocities. Previously, we used aerodynamic lenses to focus and deposit many quantum dot nanospheres as microscale lines and patterns.18 Here we direct Received: July 15, 2011 Revised: September 6, 2011 Published: September 08, 2011 12677
dx.doi.org/10.1021/la202726g | Langmuir 2011, 27, 12677–12683
Langmuir
ARTICLE
Figure 1. (a) Schematic of our in situ aerosol mass-size measurement. The red arrows indicate the trajectories of nanocrystal agglomerates, first through the differential mobility analyzer (DMA) and then through the aerosol particle mass analyzer (APM). (b) Mass mobility relationships for nanocrystal agglomerates from colloidal dispersions of different concentrations. The inset table lists the mass-mobility exponents, Df, obtained from linear fits for the different concentrations. (c) Relationships between the calculated mass density of the nanocrystal agglomerates and their size. The inset table lists the corresponding average mass density of agglomerates aerosolized from nanocrystal dispersions of different concentrations.
individual nanospheres toward a substrate and examine the morphology of the deposit (i.e., the “splat”) after impact. By varying the nanosphere size and impact velocity, we observe influences on the mechanical properties due to the granular nature of the solid. Furthermore, instead of cross-linking, we can tune the cohesion between the nanocrystals by varying the amount of organic material in between. We use nearly monodisperse CdSe nanocrystals prepared by standard techniques.19,20 Our nanocrystals were 4.7 nm in diameter and were capped with TOPO and dispersed in hexane (Chromasolv Plus, Aldrich). The number concentration, n, of the nanocrystals in the dispersion was determined from the optical absorbance at 350 nm.21,22 Aerosols of nanocrystals in the carrier gas were sprayed from colloidal dispersions with a Collison atomizer, as described previously.18 Although this aerosolization process yields some sort of nanocrystal agglomerate, a priori we did not know whether it produced quantum dot nanospheres. Therefore, we first characterized the morphology and mass density of the resulting agglomerates. Figure 1a depicts the experimental setup for tandem measurements of the size and mass of the agglomerates.23,24 Although they are charged after aerosolization due to friction between the solvent and the metallic nozzle of the nebulizer, they were passed through a bipolar neutralizer using a radioactive source (Po-210) to ensure no more than a single charge per agglomerate. The aerosol then enters the differential mobility analyzer (DMA),
which selects particles of a specified mobility. Only agglomerates with a certain mobility diameter exit the DMA. The mass of these agglomerates is then determined by an aerosol particle mass analyzer (APM), located downstream from the DMA, by balancing the electrical and centrifugal forces on the agglomerates. We characterized agglomerates from colloidal dispersions at four different concentrations. All of the nanocrystals originated from the same synthesis. Nanocrystal agglomerates of known mobility diameter (from 70 to 160 nm) were selected with the DMA. For each of these sizes, the mass was measured with the APM. Relationships between the mass (m) and mobility diameter (dm) of the nanocrystal agglomerates are shown in Figure 1b. For agglomerates composed of spherical monodisperse nanocrystals, the relationship between m and dm can be f written as m = kdD m , where k is a constant and Df is the mass mobility exponent, which is an indicator of particle morphology.25 As can be seen in Figure 1b, the m versus dm data for all dispersions tested are well fit by straight lines on a log log plot. The mass mobility exponents of all agglomerates (i.e., the slopes of these lines) are 3.0 ( 0.1 (inset of Figure 1b). The mass mobility exponents for spheres are exactly equal to 3.0, and values for compact nonspherical particles are somewhat smaller than this. For example, the calculated value of Df for cubes is 2.98. In contrast, measured values of Df for highly nonspherical particles such as nanorods or agglomerates range from 0.8 to 2.4.24,26 In the following text, we assume that particles are compact spheres, 12678
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
Langmuir
ARTICLE
Figure 2. Schematic of our experimental setup for low- and high-speed impactions of spherical nanocrystal agglomerates using a parallel plate electrostatic precipitator and an aerodynamic lens system, respectively.
although our arguments would not be affected if particles were not perfectly spherical. Because the agglomerates are close to compact spheres, one can then calculate the mass density of the agglomerates on the basis of their diameter and mass. These results are shown in Figure 1c. The measured mass densities range from 1.32 to 2.50 g cm 3, depending on the initial nanocrystal concentration in the colloidal dispersion. These values are all much higher than that of pure hexane (0.66 g cm 3 at room temperature), suggesting the nearly complete evaporation of hexane from the agglomerates. Because most of the hexane evaporates after atomization, the agglomerates are composed of inorganic CdSe nanocrystals (with a bulk density of 5.81 g cm 3) in an organic matrix consisting of capping ligands (with a density of 0.83 g cm 3) and other organic residues. For agglomerates from the same dispersion, the mass density remains constant when the agglomerate mobility diameter increases, implying that the volume fraction composed of the inorganic cores is independent of the agglomerate size. However, when the nanocrystal number concentration in the dispersion is increased, the volume fraction of the nanocrystals in the agglomerates goes up as well, resulting in an increase in the agglomerate mass density (inset of Figure 1c). From the measured densities, we estimate that the volume fractions of bare CdSe nanocrystals within the agglomerates are 10, 16, 25, and 34% for our four concentrations. These estimates assume that the remaining volume of the agglomerates is occupied by organic material with a density of 0.83 g cm 3. We note that 34% is essentially what one would expect for randomly close-packed TOPO-capped CdSe nanocrystals. On the basis of the above measurements, we determined that the aerosolized agglomerates were indeed spherical densely packed assemblies of nanocrystals. We then proceeded to impact
and deform these quantum dot nanospheres on hydrophobic carbon substrates (ultrathin carbon type-A transmission electron microscopy grids, Ted Pella), which have a surface roughness measured by atomic force microscopy of ∼0.7 nm. As illustrated in Figure 2, two different methods were employed for impaction. For low-velocity tests, the aerosol bypassed the bipolar neutralizer and was fed instead into a parallel plate electrostatic precipitator operating at 2000 2500 V. Charged agglomerates approached the substrate at a steady drift velocity because of the electric field applied across the plates, resulting in an impact velocity of several centimeters per second. For high-velocity tests, the aerosol was passed through the bipolar neutralizer and then focused by an aerodynamic lens system (a series of centercollimated orifices).18 The nanospheres bombarded the substrate located in a vacuum chamber after they were accelerated to a terminal velocity of several hundred meters per second. Consequently, between these two methods, the impact velocity could be varied by about 4 orders of magnitude. Figure 3a,b presents typical morphologies of splats formed by nanospheres impacting carbon substrates at ∼2 cm s 1 and ∼150 m s 1, respectively. At first glance, no obvious difference in the splat morphologies is seen between the two cases. All of the deposits have a circular shape (details in the Supporting Information), without signs of either fracture or splash, suggesting that the nanocrystal cores within the agglomerates interact cohesively. To verify this hypothesis, the impaction experiments were repeated using nanocrystal dispersions in toluene (Chromasolv Plus, Aldrich), which is less volatile than hexane. The mass density of these agglomerates, measured by DMA-APM, equaled 0.87 g cm 3, which is virtually identical to the density of pure toluene (0.867 g cm 3). This indicates that after aerosolization a droplet is produced that contains nanocrystals dispersed in 12679
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
Langmuir
ARTICLE
Figure 3. Transmission electron microscopy (TEM) images of nanocrystals impacting a carbon substrate. Quantum dot nanospheres formed from hexane dispersions were impacted at (a) ∼2 cm/s and (b) ∼150 m/s. Colloidal droplets formed from toluene dispersions were impacted at (c) ∼1 cm/s and (d) ∼100 m/s.
toluene, rather than a quantum dot nanosphere. When this droplet impacts the substrate under identical experimental conditions, the morphologies are notably different from those of the nanospheres produced from hexane dispersions. For toluene, ring patterns of nanocrystals are formed at low velocity (Figure 3c), and randomly scattered nanocrystals are formed at high velocity (Figure 3d). These morphologies both suggest an absence of cohesion between the nanocrystals. Under lowvelocity impaction (Figure 3c), the droplet spreads on the hydrophobic carbon. The ring patterns of nanocrystals are probably due to the “coffee ring” effect in which capillary flows during drying push the nanocrystals to the periphery of the droplet.27 Under high-velocity impaction (Figure 3d), the droplets appeared to splash and break up into smaller drops.28 After evaporation of the toluene, randomly distributed nanocrystals remain. The contrast between the toluene and hexane results further supports the existence of nanocrystal cohesion within the nanospheres prepared from hexane dispersions and is consistent with previous studies on the effect of the presence or absence of solvent on the pair potential between semiconductor nanocrystals. Recent calculations of the pair potential between semiconductor nanocrystals in a good solvent29 indicate that it is weak— several times smaller than for metal nanoparticles—whereas in the absence of solvent the pair potential is much larger because of the unscreened van der Waals attractive forces between the capping ligands.30
Figure 4. Scanning transmission electron microscopy (STEM) images of deformed quantum dot nanospheres impacting a carbon substrate. Nanospheres aerosolized from more concentrated nanocrystal dispersions (1.24 1014 cm 3) and impacting at (a) ∼150 m/s and (b) ∼2 cm/s. The nanocrystals in all images have an average diameter of 4.7 nm.
As mentioned above, the strength of this cohesion can be varied systematically by changing the concentration of the nanocrystals in the hexane dispersion. Figures 4 and 5 show scanning transmission electron microscopy (STEM) images of nanosphere splats produced under different impact conditions. For STEM, a narrow (∼0.1 nm wide) electron beam was rastered across selected splat patterns and high-angle annular dark-field images were collected. Figure 4a,b shows images of deposits from 12680
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
Langmuir
Figure 5. Scanning transmission electron microscopy (STEM) images of deformed quantum dot nanospheres impacting a carbon substrate. Nanospheres aerosolized from less-concentrated nanocrystal dispersions (4.28 1013 cm 3) and impacting at (a) ∼150 m/s and (b) ∼2 cm/s. The nanocrystals in all images have an average diameter of 4.7 nm.
a relatively concentrated nanocrystal dispersion (n = 1.24 1014 cm 3) impacting at high and low velocities, respectively, and Figure 5a,b shows the corresponding images for the case of a more dilute dispersion (n = 4.28 1013 cm 3). In Figures 4 and 5, the size of the nanosphere increases from left to right in each row. The leftmost images suggest that nanospheres that are sufficiently small form splats consisting of a single monolayer (discussed further below). For larger nanospheres, however, the splat morphology is strongly affected by the concentration of the nanocrystal dispersion and to a lesser extent by the impact velocity. For concentrated dispersions (Figure 4a,b), the nanospheres resist spreading on the substrate, suggesting solidlike behavior and strong cohesion between the nanocrystals. For less-concentrated dispersions (Figure 5a,b), the nanospheres behave more like a viscous liquid during impact, spreading on the substrate to form a pile of nanocrystals with multiple monolayers at the center and a single monolayer at the periphery. In addition, when the nanospheres from less-concentrated dispersions are large and impact at high velocity, they tend to splash, leaving a separate ring around the central deposit (Figure 5a, rightmost image). In this case, randomly distributed branching of nanocrystals occurs at the splat periphery, which is one of the characteristics of droplet impaction (also called “fingering” in droplet impact dynamics28). According to our mass density measurements, the volume fraction of nanocrystal cores in the nanospheres increases with the number concentration of nanocrystals in the colloidal dispersion. Because the CdSe nanocrystals in this study all have the same average size, this increase in volume fraction implies that less organic material is in the solid and thus there is a narrower spacing between neighboring nanocrystals. As the nanocrystals become closer, their cohesion becomes stronger, inducing a liquid-to-solid transition in the impact behavior of the nanospheres. Thus, the cohesion between nanocrystals plays a critical role in determining the rheological properties. Moreover, the cohesion can be modified by simply adjusting the number concentration of the colloidal dispersion that is aerosolized. In STEM images (such as in Figures 4 and 5), both incoherently and elastically scattered electrons are collected. The intensity of each pixel in the image is therefore linearly proportional to the local sample thickness, especially for the small CdSe nanocrystals imaged.31 By scanning across a line, one can then infer the thickness profile of the splats and the corresponding number of nanocrystal monolayers, as illustrated in Figure 6.
ARTICLE
Figure 6a shows the STEM image of a splat consisting of a single nanocrystal monolayer. Figure 6b plots the corresponding image intensity profile. Because the intensity is linearly proportional to the thickness, this profile can be interpreted as the geometric cross section of the deposit along the line. Similar data for a thicker deposit are shown in Figure 6c,d. Furthermore, by using the peak intensity of an isolated nanocrystal, as indicated by the rectangles in Figure 6b,d, we can calibrate the thickness of the deposit and determine the total number of nanocrystals in the image (details in the Supporting Information). Thus, STEM can quantify two parameters: the thickness of the splat and the total number of nanocrystals within it. On the basis of data extracted from such images, Figure 7 summarizes the relationship between the thickness of the agglomerate splats (i.e., the number of monolayers) and the number of nanocrystals per nanosphere (Na). This relationship is also plotted as a function of the nanocrystal concentration in the dispersion (n) and the impact velocity. Several trends can be observed because of the relative strength of the cohesive forces between the nanocrystals in the solid. First, for a sufficiently small number Na of nanocrystals per nanosphere, the deposits consist of a single monolayer, regardless of the dispersion concentration or the impact velocity. As Na increases beyond some critical value, a second monolayer forms, and thereafter the deposit thickness increases with Na. The stacking of nanocrystals is analogous to “arching” in granular materials (for example, in sand).32 For solid grains, arching occurs because of the friction between particle contacts. In the case of colloidal quantum dot nanospheres, our results suggest that arching occurs because of the cohesion between the nanocrystals. Second, for deposits that are thicker than a monolayer, Figure 7 indicates that high-velocity impacts tend to produce thicker deposits than low-velocity impacts for nanospheres containing the same number of nanocrystals. Analogous behavior has been reported in several studies of concentrated colloidal particles in the micrometer-size range, where it was observed that at low shear rates the viscosity of the suspension decreases with increasing shear rate but that a transition occurs at a sufficiently high shear rate such that further increases in the shear rate cause the viscosity to increase, resulting in shear thickening.33 This behavior has been attributed to the particles’ spatial organization under flow, which is affected by the balance between hydrodynamic and interparticle forces. Although such interparticle forces are likely to be rather different for quantum dots than for micrometer-sized particles and our study involved inertial impact instead of applied shear, our result that high-velocity impacts produce thicker deposits appears to be at least consistent with the shear-thickening phenomenon. It is known that granular materials can flow like a liquid if an external perturbation is applied with a strength above a certain threshold level.34 Attractive forces between the grains will increase the level of this threshold. Thus, strongly cohesive nanospheres are less likely to flow, as demonstrated in Figure 4a,b and ref 14. In addition to the results above, this phenomenon can also be demonstrated by studying the overlap of two splats, as shown in the TEM images in Figure 8. Figure 8a,b shows deposits from the high-concentration dispersion and from high-velocity and low-velocity impacts, respectively, and Figure 8c,d shows corresponding deposits for the low-concentration dispersion. In the high-n case (strong nanocrystal cohesion), the boundaries of each nanosphere are clearly seen in the overlap region, regardless of the impact velocity, and in the low-n case, the nanocrystals in 12681
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
Langmuir
ARTICLE
Figure 6. (a) Scanning transmission electron microscopy (STEM) image of a single monolayer of nanocrystals impacting a carbon substrate. (b) Brightness intensity for each pixel along the blue line in a. (c) STEM image of a deformed quantum dot nanosphere on a carbon substrate after impaction. (d) Brightness intensity for each pixel along the blue line in c. Individual nanocrystals, indicated by the red boxes, can be used to calibrate the intensity of the STEM signal.
Figure 7. Thickness of the deformed nanospheres versus the average number of nanocrystals within the nanospheres. Data for two different impact velocities and for two different nanocrystal concentrations are plotted.
the overlap region start to flow and rearrange. The nanospheres aerosolized from the more dilute dispersion are able to coalesce and form a more continuous deposit of nanocrystals in both the high- and low-velocity impact cases. These results are consistent with the solid-to-liquid transition discussed above. In summary, we have studied the impact dynamics of aerosolized colloidal quantum dot nanospheres, where each nanosphere consists of an agglomerate of CdSe nanocrystals. Aerosol mass mobility measurements indicate that these agglomerates are close to spherical and that the concentration of the initial colloidal dispersion controls the average spacing between the
Figure 8. Transmission electron microscopy (TEM) images of overlapped nanosphere splats that impacted a carbon substrate. The images were obtained from the following nanocrystal concentrations and impact velocities: (a) 1.24 1014 cm 3 and ∼150 m/s, (b) 1.24 1014 cm 3 and ∼2 cm/s, (c) 4.28 1013 cm 3 and ∼150 m/s, and (d) 4.28 1013 cm 3 and ∼2 cm/s.
nanocrystals. As the spacing between nanocrystals is varied, one would expect their attractive interactions to change. Indeed, experiments conducted with both high- and low-impact velocities of the nanospheres on carbon substrates demonstrate that the impact dynamics are decisively determined by the cohesive interactions in the nanospheres. As the spacing between nanocrystals is reduced, the nanosphere splats reveal a liquid-to-solid 12682
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
Langmuir transition. Mildly cohesive nanospheres that are aerosolized from relatively dilute dispersions exhibit liquidlike behavior, including splashing at high-velocity impaction and the coalescence of overlapping splats. In contrast, strongly cohesive nanospheres that are aerosolized from more concentrated dispersions exhibit arching and thickening, behavior typical of granular materials. In addition to exhibiting these fundamental effects of cohesion and granularity in quantum dot solids, our findings also suggest routes to the aerosol-based printing of colloidal nanocrystals. In particular, the number of monolayers deposited can be controlled by adjusting the concentration of nanocrystals in the colloidal dispersion and, to a lesser extent, the impact velocity.
’ ASSOCIATED CONTENT
bS
Supporting Information. Methods to determine the aspect ratio of the nanocrystal agglomerates, the thickness of the nanocrystal splats, and the number of nanocrystals in each through quantitative analysis of the STEM images. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected] and
[email protected]. Present Addresses §
)
Dow Chemical (China) Co., Ltd., No. 936 Zhangheng Road, Shanghai 201203, China. ETH Z€urich, Universitaetstrasse 6, 8092 Z€urich, Switzerland.
’ ACKNOWLEDGMENT This work was supported by the NSF Nanoscale Interdisciplinary Research Team (NIRT) program (CBET-0506748) and utilized resources at the University of Minnesota Characterization Facility, funded by the NSF through the NNIN program. We thank J. Scheckman for his assistance with the DMA-APM measurements, M. S. Kang for his help with the AFM characterization of the carbon substrate, and U. Ozan for helpful discussions related to STEM imaging.
ARTICLE
(12) Lin, Y.; Skaff, H.; Boker, A.; Dinsmore, A. D.; Emrick, T.; Russell, T. P. J. Am. Chem. Soc. 2003, 125, 12690. (13) Lee, D.; Jia, S. G.; Banerjee, S.; Bevk, J.; Herman, I. P.; Kysar, J. W. Phys. Rev. Lett. 2007, 98. (14) Klajn, R.; Bishop, K. J. M.; Fialkowski, M.; Paszewski, M.; Campbell, C. J.; Gray, T. P.; Grzybowski, B. A. Science 2007, 316, 261. (15) Mueggenburg, K. E.; Lin, X. M.; Goldsmith, R. H.; Jaeger, H. M. Nat. Mater. 2007, 6, 656. (16) Kim, D. J.; Koo, K. K. Cryst. Growth Des. 2009, 9, 1153. (17) Chu, M. Q.; Wu, F.; Zhang, Q.; Liu, T. T.; Yu, Y.; Ji, A. L.; Xu, K. Y.; Feng, Z. H.; Zhu, J. Nanoscale 2010, 2, 542. (18) Qi, L. J.; McMurry, P. H.; Norris, D. J.; Girshick, S. L. Aerosol Sci. Technol. 2010, 44, 55. (19) Peng, Z. A.; Peng, X. J. Am. Chem. Soc. 2001, 123, 183. (20) Reiss, P.; Bleuse, J.; Pron, A. Nano Lett. 2002, 2, 781. (21) Klimov, V. I.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Phys. Rev. B 1999, 60, 13740. (22) Leatherdale, C. A.; Woo, W. K.; Mikulec, F. V.; Bawendi, M. G. J. Phys. Chem. B 2002, 106, 7619. (23) Park, K.; Kittelson, D. B.; McMurry, P. H. Aerosol Sci. Technol. 2004, 38, 881. (24) Scheckman, J. H.; McMurry, P. H.; Pratsinis, S. E. Langmuir 2009, 25, 8248. (25) Mandelbrot, B. B., The Fractal Geometry of Nature; Freeman: San Francisco, 1982. (26) Kim, S. H.; Mulholland, G. W.; Zachariah, M. R. Carbon 2009, 47, 1297. (27) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Nature 1997, 389, 827. (28) Yarin, A. L. Annu. Rev. Fluid Mech. 2006, 38, 159. (29) Evers, W. H.; De Nijs, B.; Filion, L.; Castillo, S.; Dijkstra, M.; Vanmaekelbergh, D. Nano Lett. 2010, 10, 4235. (30) Landman, U.; Luedtke, W. D. Faraday Discuss. 2004, 125, 1. (31) Liu, J. Y. J. Electron Microsc. 2005, 54, 251. (32) Duran, J. Sands, Powders, and Grains: An Introduction to the Physics of Granular Materials; Springer-Verlag: New York, 2000. (33) Hebraud, P.; Lootens, D. Mod. Phys. Lett. B 2005, 19, 613. (34) Jaeger, H. M.; Nagel, S. R.; Behringer, R. P. Rev. Mod. Phys. 1996, 68, 1259.
’ REFERENCES (1) Colvin, V. L.; Schlamp, M. C.; Alivisatos, A. P. Nature 1994, 370, 354. (2) Coe, S.; Woo, W. K.; Bawendi, M.; Bulovic, V. Nature 2002, 420, 800. (3) Gur, I.; Fromer, N. A.; Geier, M. L.; Alivisatos, A. P. Science 2005, 310, 462. (4) Talapin, D. V.; Murray, C. B. Science 2005, 310, 86. (5) Robel, I.; Subramanian, V.; Kuno, M.; Kamat, P. V. J. Am. Chem. Soc. 2006, 128, 2385. (6) B€oberl, M.; Kovalenko, M. V.; Gamerith, S.; List, E. J. W.; Heiss, W. Adv. Mater. 2007, 19, 3574. (7) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. Rev. Mater. Sci. 2000, 30, 545. (8) Houtepen, A. J.; Vanmaekelbergh, D. J. Phys. Chem. B 2005, 109, 19634. (9) Chandler, R. E.; Houtepen, A. J.; Nelson, J.; Vanmaekelbergh, D. Phys. Rev. B 2007, 75. (10) Houtepen, A. J.; Kockmann, D.; Vanmaekelbergh, D. Nano Lett. 2008, 8, 3516. (11) Soreni-Harari, M.; Mocatta, D.; Zimin, M.; Gannot, Y.; Banin, U.; Tessler, N. Adv. Funct. Mater. 2010, 20, 1005. 12683
dx.doi.org/10.1021/la202726g |Langmuir 2011, 27, 12677–12683
