Screw Dislocation-Driven Growth of Two-Dimensional Nanoplates

Sep 6, 2011 - Screw Dislocation-Driven Growth of Two-Dimensional Nanoplates .... (32) All minor peaks can be indexed to zinc hydroxide or the pentahyd...
0 downloads 0 Views 4MB Size
Download PDF
LETTER pubs.acs.org/NanoLett

Screw Dislocation-Driven Growth of Two-Dimensional Nanoplates Stephen A. Morin, Audrey Forticaux, Matthew J. Bierman, and Song Jin* Department of Chemistry, University of Wisconsin

Madison, 1101 University Avenue, Madison, Wisconsin 53706, United States

bS Supporting Information ABSTRACT: We report the dislocation-driven growth of twodimensional (2D) nanoplates. They are another type of dislocation-driven nanostructure and could find application in energy storage, catalysis, and nanoelectronics. We first focus on nanoplates of zinc hydroxy sulfate (3Zn(OH)2 3 ZnSO4 3 0.5H2O) synthesized from aqueous solutions. Both powder X-ray and electron diffraction confirm the zinc hydroxy sulfate (ZHS) crystal structure as well as their conversion to zinc oxide (ZnO). Scanning electron, atomic force, and transmission electron microscopy reveal the presence of screw dislocations in the ZHS nanoplates. We further demonstrate the generality of this mechanism through the growth of 2D nanoplates of α-Co(OH)2, Ni(OH)2, and gold that can also follow the dislocation-driven growth mechanism. Finally, we propose a unified scheme general to any crystalline material that explains the growth of nanoplates as well as different dislocation-driven nanomaterial morphologies previously observed through consideration of the relative crystal growth step velocities at the dislocation core versus the outer edges of the growth spiral under various supersaturations. KEYWORDS: Screw dislocation, nanoplate, crystal growth, zinc hydroxy sulfate, metal hydroxide, gold

T

he formation of any crystalline nanomaterial necessarily requires crystal growth and must follow crystal growth theory,1,2 which describes three basic crystal growth mechanisms: spiral growth at screw dislocations (BCF theory),3,4 layer-by-layer growth (LBL),4,5 and dendritic growth. The most favorable growth mechanism is dictated by supersaturation, defined as σ = ln(c/c0) where c and c0 are the precursor and equilibrium concentrations respectively. At the lowest supersaturations, screw dislocations play a vital role in the growth of all crystals. This is because the self-perpetuating step edge provided by screw dislocation growth spirals permits crystal growth to occur at supersaturations below that necessary to create two-dimensional (2D) nuclei for LBL growth. We have shown previously that screw dislocations can drive the growth of one-dimensional (1D) nanomaterials, such as nanowires (NWs)2,6 10 and nanotubes (NTs),8 and dislocation-driven growth is evident in many more 1D materials.11 14 Here we extend dislocation-driven nanomaterial growth to single crystal nanoplates that grow rapidly in two dimensions but very slowly in the third. Despite the drastic differences in morphologies between NWs and nanoplates, dislocation-driven crystal growth is common to both and the subtle kinetics of spiral step propagation associated with dislocation growth can be used to explain each. Two-dimensional nanomaterials are an interesting class of materials whose surface area is dominated by one specific crystallographic plane; therefore they are almost crystallographically isotropic. Referenced in literature by a broad range of names such as nanoflakes,15 nanowalls,16 nanosheets,17,18 nanoplates, or nanoplatelets,19 26 2D inorganic nanomaterials, not to be confused with 2D carbon-based nanomaterials such as graphene,27 are typically nanoscale in one dimension yet microscale in the other r 2011 American Chemical Society

two. We will refer to such nanomaterials as nanoplates. In recent reports, nanoplates have found application in nanoelectronics,22,28 plasmonics,23,24,29 catalysis,18 supercapacitors,19,20,30 and Li ion batteries.15,21 Here we first discuss the low-temperature aqueous synthesis and the dislocation-driven growth of nanoplates of zinc hydroxy sulfate (ZHS, 3Zn(OH)2 3 ZnSO4 3 mH2O),31,32 a layered hydroxide related to ZnO. Even though ZHS nanoplates in various states of hydration have been observed to form under aqueous hydrolytic conditions,17,33,34 the critical role of dislocations in their growth has not been appreciated or revealed. While there are reports of the direct growth of ZnO nanoplates,25,35,36 there are some examples where nanoplates of ZHS have been incorrectly reported as ZnO,37,38 a material with numerous applications in photonics, 39 solar,40 and piezoelectronic41 energy conversion. We show that such ZHS nanoplates can be readily converted to ZnO nanoplates by annealing. The typical explanation put forth for the formation of 2D structures is that surfactants used during synthesis, such as citrate or sulfate, inhibit/slow growth preferentially due to preferential adsorption on certain crystallographic facets leading to 2D growth. While this argument can work for anisotropic crystal structures such as hexagonal wurtzite, there are also examples of nanoplates with highly symmetric structures, such as metal plates,42,43 that cannot be as easily explained in this way. We illustrate how dislocation-driven growth explains 2D crystal morphology in a general scheme independent of crystal structure by providing proof of dislocation-driven nanoplate growth in three other materials, α-Co(OH)2, Ni(OH)2, and gold. We further Received: August 3, 2011 Revised: August 21, 2011 Published: September 06, 2011 4449

dx.doi.org/10.1021/nl202689m | Nano Lett. 2011, 11, 4449–4455

Nano Letters

LETTER

Figure 1. Overview of nanoplate morphology and screw dislocation step edges of ZHS. (A) Low-magnification scanning electron microscope (SEM) image of as synthesized ZHS nanoplates, inset is a representative “hexagonal” nanoplate. (B) Low-magnification contact mode atomic force microscope image of a representative nanoplate (deflection signal shown). (C) A higher-magnification image of the dislocation core. (D) A large nanoplate containing several screw dislocations of different handedness. (E) Bright-field optical image showing the thin film interference pattern generated between the nanoplate and the reflective silicon substrate revealing the radial thickness variations within the nanoplate. (F) SEM image of a nanoplate grown on a silicon substrate with native oxide.

propose a conceptual model based on the relative step velocities44 at the dislocation core versus the outermost regions of the crystal that describes how these distinct morphologies, 1D versus 2D, can both form from dislocation spirals under different supersaturations. In a typical synthesis of ZHS nanoplates, aqueous solutions containing 3 mM ZnSO4, 3 mM Zn(NO3)2, and 6 mM hexamethylenetetramine (HMT) are heated in sealed glass vials at temperatures ranging from 60 to 95 °C in an oven for 1.5 to 2 h. A substrate, typically silicon covered with a 100 nm oxide, is placed in the solution during heating to collect the nanoplate products. Scanning electron microscopy (SEM) revealed that the substrates are densely coated with many nanoplates (Figure 1A) after this reaction procedure. In contrast, in the absence of ZnSO4 ZnO nanorods are the exclusive reaction product.38,45 The majority of nanoplates produced are nearly hexagonal in shape with a typical diameter of 10 20 μm (Figure 1A inset) and thickness of 50 100 nm. When we examined these nanoplates using contact mode atomic force microscopy (AFM), we found they all contain screw dislocations. For the representative hexagonal morphology these dislocations were single source dislocations and the core was located at the very center of the nanoplate (Figure 1B,C, see Supporting Information Figure S1 for step height analysis). Curiously, under simple bright-field optical microscopy, the thin film interference pattern generated between the surface of a nanoplate and the reflective silicon substrate surface reveals the geometry and thickness variation of the nanoplate and the location of the dislocation cores (Figure 1E). Because of the dislocation hillocks, these nanoplates are in fact “pyramids” with extremely large width-to-thickness ratios. For larger and more complex nanoplate morphologies, there can be multiple dislocation centers (Figure 1D) with different handedness (Figure 1D inset). Also, for these larger more defective

nanoplates it was common to see off-shooting nanoplate growth that could represent secondary nanoplate growth from an existing nanoplate or delamination of the layered ZHS structure. Moreover, when nanoplates are grown on silicon surfaces with native oxides the dislocation spirals are heavily etched clearly revealing their morphology under standard SEM (Figure 1F). Despite being synthesized under very similar conditions as ZnO NRs or NWs,38,45 the nanoplates synthesized are clearly zinc hydroxy sulfates (ZHS) according to powder X-ray diffraction (PXRD) (Figure 2A). ZHS are layered hydroxides with structures similar to brucite46 that incorporate sulfate and water into their lattice. The stoichiometry is 3Zn(OH)2 3 ZnSO4 3 mH2O where m can equal 5, 4, 3, 1, or 0.5.31,32 As shown by PXRD (Figure 2A), the hemihydrate phase (3Zn(OH)2 3 ZnSO4 3 mH2O with m = 0.5, PDF no. 00-044-0674) is the majority phase produced.32 All minor peaks can be indexed to zinc hydroxide or the pentahydrate form of ZHS where m = 5. Some previous reports have mistakenly identify ZHS (in its various hydration states) as ZnO37,38 and none have made the connection to dislocation-driven growth.17,33,37,38 Interestingly, many metal hydroxides/oxides have been reported to readily form similar nanoplate morphologies16,19 21,47,48 with the most similar being the zinc hydroxy chlorides.49 52 The layers in ZHS are always perpendicular to the c-direction and consist of planes of hexagonally arranged octahedrally coordinated zinc cations bridged by water and sulfate anions (Figure 2B left). While the separation between zinc layers increases with the degree of hydration, the hexagonal geometry of the zinc cations always remains the same.31,32 Furthermore, ZHS have triclinic crystal structures with unit cell parameters that are very close to a hexagonal cell (Supporting Information Table S1).31,32 In particular, the hemihydrate unit cell (a = 8.356 Å, b = 8.356 Å, and c = 7.084 Å and α = 90°, β = 90°, and γ = 120°) is 4450

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455

Nano Letters

LETTER

Figure 2. Structure of ZHS versus ZnO. (A) Powder X-ray diffractograms for ZHS nanoplates (bottom) and resulting porous ZnO nanoplates after annealing (top). All major peaks can be indexed to ZHS hemihydrate (PDF no. 00-044-0674) and ZnO (PDF no. 00-036-1451), respectively. Inset is an SEM image of an annealed nanoplate. (B,C) Structure views of ZHS and ZnO along the a axis (B) and c axis (C). Note that for ZnO, which has the wurtzite structure, the b axis is equivalent to a.

pseudohexagonal.32 As a result, the structure of ZHS has similarities to the wurtzite structure of ZnO. To explore this, side by side structure views of ZHS and ZnO are displayed in Figure 2B, C. Note, because atomic coordinates are not available for the hemihydrate structure, an analogous pentahydrate structure with a larger c-axis lattice constant (11.001 Å versus 7.084 Å, Supporting Information Table S1)32 is used. When viewed along the a axis, the hydration layers and sulfate bridges of ZHS and the difference between the two structures becomes obvious (Figure 2B). However, when viewed down the c axis (Figure 2C), we see that the zinc cations in the basal planes of both structures are arranged hexagonally with nearly identical separations (3.1 and 3.2 Å for ZHS and ZnO, respectively). From this structural similarity and known phase diagrams,53 we reasoned that dehydration and removal of sulfate (in the form of sulfur dioxide and oxygen) via annealing would topotactically convert ZHS to ZnO. PXRD of annealed samples demonstrates and confirms this conversion (Figure 2A, top). Furthermore, thermogravimetric analysis (TGA, Supporting Information Figure S2) supports this dehydration process and the hemihydrate phase assignment. We have further characterized these nanoplates using transmission electron microscopy (TEM) and electron diffraction (ED) (Figure 3). It became immediately obvious that the ZHS nanoplates are not stable under the high energy (200 keV) electron beam of the TEM. The materials were quickly damaged and the single crystal ED patterns, especially under the convergent beam condition, quickly deteriorated to polycrystalline ring patterns that can be indexed to ZnO. Bearing in mind that the structure of ZHS consists of periodic nonconductive hydration layers (Figure 2B) and that the beam direction along the c axis orthogonal to these layers, it is not surprising that electron beam exposure causes

degradation of ZHS due to charging effects. The minimized beam intensity of select area ED (SAED) allowed for collection of single crystal ED patterns (Figure 3A, inset) that can be indexed to the [001] zone axis pattern (ZAP) of ZHS (the expected zone axis for top down observation of nanoplates along the c axis). Even with SAED, over time the region of observation will degrade (Figure 3A, red circle). High-resolution TEM (HRTEM) images of these damaged regions revealed their polycrystallinity (Figure 3B) and that the individual crystalline grains have lattice fringes with a spacing of ∼2.8 Å, which is the spacing expected for the (100) planes of ZnO. This conversion from single crystal ZHS to polycrystalline ZnO is in agreement with the PXRD data. Moreover, during topotactic conversion of ZHS to ZnO it is expected that the grains of ZnO formed still are oriented with their c axes parallel to the electron beam, thus remaining on the [001] zone axis, which is consistent with our observations. Curiously, many intriguing contrast contours can be observed in the bright-field TEM images of these nanoplates (Figure 3C), especially when the objective aperture is centered over only the zero beam. Such features are not uncommon for materials of thin plate or thin disk morphologies,12,23,42,43 but they are usually characterized exclusively as bend contours43 associated with the thin nanoplates conforming to the uneven TEM grid surface. Undoubtedly, this is the case for most of the contours observed; however, on occasion we observe closed loop contours we call “spider contours” (Figure 3D). They have a different appearance than typical bend contours and may not be so easily explained away as such. It is particularly interesting that some spider contours seem to have open cores (Figure 3D). Similar spider contours have been displayed in some previous nanoplate reports, but the origin of these interesting features was never commented on.42 Upon the 4451

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455

Nano Letters

Figure 3. Material characterization and spider contours observed using TEM. (A) Low-resolution TEM image of a ZHS nanoplate with an inset of the selected area electron diffraction from the area circled in red. Note the damage from the electron beam in this and surrounding areas. (B) High-resolution TEM image of a ZHS nanoplate revealing electron beam conversion to polycrystalline ZnO as indicated by the fast Fourier transform of the area whose ring pattern can be indexed to the (100) planes of ZnO (inset B). (C) Low-resolution image of a nanoplate showing the abundance of contrast contours. (D) “Spider contours” believed to be associated with the dislocations contained within these materials (scale bar in both images is 200 nm).

basis of the knowledge that these nanoplates contain screw dislocations and that the strain associated with such dislocations can perturb the crystal lattice, as with Eshelby twist,54 hollow core dislocations,55 or nanotubes,8 we suggest that these spider contours could result from the strain fields of screw dislocations thus indicating the presence of screw dislocations in the nanoplates. Detailed analysis of these spider contours using dark-field TEM, similar to the process of indexing twist contours in NWs that have lattice twist,9,10 could further elucidate/verify this connection. However, the poor beam stability of ZHS precludes such detailed investigation for these samples. Until now screw dislocations have been shown to drive the growth of ZnO NWs,2,8,9 nanorods (NRs),8 and NTs.2,8 Here we have expanded this to 2D nanoplates of related ZHS. In all cases, the unifying concept is the screw dislocation, however, it may seem unclear initially why or how one type of defect could lead to such disparate morphologies. To explain this we propose an intuitive model that considers the crystal growth step velocities at various positions from the dislocation core (Figure 4A).44 In general, the velocity of steps at the core (vc) should be the same as those at the outer edges of the dislocation hillocks (vo) because they are crystallographically equivalent. However, if during the initial stages of crystal growth, environmental factors, such as impurities or mass transport, cause the velocity of steps created earlier in growth (thus at the outer edge, vo) to be slower relative to newly generated steps closer to the core of the dislocation (vc), two scenarios arise. When vo is slightly less than vc, newly generated steps from the dislocation core would very slowly catch those at the outer edge to create a step pile up and thus a step-free facet, which will lead to low aspect ratio NRs or bulk polyhedral crystals (Figure 4C), particularly when under higher

LETTER

Figure 4. A unified scheme illustrating dislocation-driven growth of various nanomaterial morphologies. (A) Top view and dramatized side view of an idealized nanoplate with one screw dislocation with various step velocities at the dislocation core (vc) and the outer edges (vo). When specific step velocity and supersaturation conditions are met, nanowires or nanotubes (B), nanorods (C), or nanoplates (D) result. (E,F) Idealized nanoplate side views illustrate the relationship between nanoplate slope (p), step height (h), terrace width (λ), and supersaturation (σ).

supersaturation and there is significant LBL growth. Furthermore, if vo is much less than vc possibly because the steps at the outer edge have more impurity pinning sites, more adsorbed surfactants,56 or slower mass transport kinetics,7 newly generated steps quickly catch those further out, rapidly producing a stepfree facet and thus leading to almost pure 1D growth under low supersaturation conditions (Figure 4B). This is the mode responsible for generating high aspect ratio NWs and NTs of ZnO previously reported.8,9 Here, if it is the attachment of “surfactant” molecules that plays a role in differentiating the growth kinetics, the argument is more subtle than the common narrative of preferential adsorption on certain crystallographic facets, rather the molecules interact with the dislocation steps and change the kinetics of the step advancement.56,57 Finally, if vo is equal to vc, which is more common in general crystal growth, all steps, regardless if they are newly generated steps at the dislocation core or earlier steps at the outer edge of the growth spiral, propagate at the same velocity and a step pile up is never created and thus the steps spread in 2D. Here, this similarity in step velocity is supported by the nearly uniform terrace width (λ, Figure 4E) observed under AFM (Figure 1B,C and Supporting Information Figure S3). This situation in itself does not guarantee the formation of 2D nanoplates because the linear growth rate normal to the crystal facet can still be significant. However, if the growth hillocks have very small slopes (p, Figure 4E), the crystal growth normal to the dislocation spiral is still very slow and the plates spread almost purely in the other two dimensions. We believe this is the case for the nanoplates reported here and perhaps those seen elsewhere.15 25,30,43,47,48 Note that these nanoplates are not truly flat plates as the name may imply, instead they are pyramids with extremely small slopes and thus very large width-to-height ratios, as dramatically illustrated by the side view in Figure 4A and can be clearly seen in Figure 1B,C,E. We further note that according to BCF theory4,44 terrace width (λ) is inversely 4452

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455

Nano Letters

LETTER

Figure 5. Other examples of dislocation-driven nanoplates. SEM images of representative nanoplates of (A) α-Co(OH)2, (B) Ni(OH)2, and (C) gold highlighting the screw dislocation spirals.

proportional to supersaturation (σ) of the system, and thus with the same step height (h) the slope of the nanoplate hillocks (p = h/λ) increases as supersaturation increases, as illustrated in Figure 4E,F. Therefore, the thickness-to-diameter ratio of 2D nanoplates is dependent on the supersaturation of the growth solutions, that is, low supersaturation growth favors the generation of very thin nanoplates with small slopes while higher supersaturations leads to thicker plates of increasing slope. This relationship between terrace width (and thus the slope of dislocation hillocks) and supersaturation has been well documented in crystal growth studies, such as those shown by De Yoreo et al.56 As shown in Supporting Information Figure S4, this expected trend for the ZHS nanoplates described herein was observed, despite the fact that their synthesis, which was always run in a sealed vial that is not replenished with fresh reagents, did not maintain constant supersaturation (Supporting Information Figure S5). Within this unified framework, it is clear that understanding and controlling supersaturation and others factors that influence relative step advancement rates during dislocationdriven growth could eventually enable precise control over the final crystal morphologies generated. So far in this work we have focused on ZHS and previously on ZnO8,9 but the screw dislocation-driven growth of nanoplates is general and also observed for other materials such as gold43 and is also believed to play a fundamental role in the formation of nacre layers in mollusk shells.58 To demonstrate the generality of dislocation-driven 2D crystal growth as illustrated in Figure 4, nanoplates of two layered double hydroxides, α-Co(OH)247 and Ni(OH)2,30,48 and gold,43 which has a face-centered cubic structure and is thus highly symmetric crystallographically, were synthesized in mild aqueous environments following conditions modified from the literature30,43,47,48 (see synthetic details in Supporting Information and PXRD data in Figure S6). Indeed, SEM clearly reveals the screw dislocation spirals of these nanoplates (Figure 5). Comparable to ZHS, the α-Co(OH)2 samples have high densities of nanoplates that all show dislocations, however due to their smaller diameter and wider terrace widths they do not display the same dramatic thin film interference under optical microscopy. Ni(OH)2 nanoplates are equally dense but thinner than ZHS and have a propensity to form flowerlike structures30 that are similar to the off-shooting nanoplates seen in ZHS (Figure 1A). Gold nanoplates are also produced in high density, but their shape and size vary. Consistent with literature,43 gold nanoplates come as triangles and hexagons as well as truncated

triangles, however only a small percentage of these plates clearly show dislocation spirals. Moreover, the step heights, previously estimated to be ∼100 Å,43 and thus the burgers vectors of the dislocations observed in gold plates are very large in comparison to the d111 spacing of 2.35 Å for gold. The surface of gold nanoplates is the (111) plane and this should be the elementary burgers vector. Suito and Uyeda43 suggested that in an acidic media small gold particles are not stable and therefore growth can proceed via more favorable large crystal growth units, which could explain the observation of such large screw dislocation spiral step heights. The question of whether or not all the gold nanoplates grow via screw dislocation as shown in Figure 5C is still open, but we believe empirical proof of this may just be limited by the ease and ability to resolve the dislocation step edges on all gold nanoplates, which could be as small as 2.35 Å. Indeed, ZHS and α-Co(OH)2 have lattice parameters larger than 7 Å32,47 along the vertical growth direction (along c axis), which lead to large elementary burgers vectors and therefore increase the likelihood and ease to observe dislocation spirals in these systems. Furthermore, the tendency for some materials to have large burgers vectors and bunched step heights make the observation of dislocation spirals easier such as the case of ZHS. Similar to gold, β-Co(OH)2 has a small lattice constant c of 4.65 Å,47 which is the growth axis, and we did not readily observe screw dislocations (Supporting Information Figure S7A). The βCo(OH)2 case is particularly interesting as the screw dislocationdriven growth of nanowires has been reported for this system,13 supporting the suggestion that the nanoplates also grow according to this mechanism similar to what is observed for the α phase. Here we have shown that the growth of nanoplates of ZHS, αCo(OH)2, Ni(OH)2, and gold can be dislocation-driven. Additionally, the direct product of hydrolysis of Zn(NO3)2 in the presence of sulfate is not ZnO but nanoplates of ZHS, which can be converted topotactically through annealing to ZnO. SEM, and also AFM for ZHS, confirm that the nanoplates often contain at least one screw dislocation that propagates 2D crystal growth. A peculiar type of TEM image contour, the spider contour, is observed for ZHS and suggested to be caused by dislocation strain in nanoplates. This explanation of 2D growth is rooted in classic crystal growth theory and is more general than the previous ones that attribute anisotropic growth to surfactants slowing/prohibiting growth on certain crystal faces because it considers the role of step velocities of dislocation growth spirals. We propose a simple model considering the relative step velocities 4453

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455

Nano Letters at the dislocation core versus the outer edges of the growth spiral under various supersaturations to explain the different dislocation-driven morphologies, NWs/NTs, NRs, and nanoplates, of crystalline materials in general. Understanding and controlling the factors that cause these differences in step velocities could enable precise morphological engineering of dislocation-driven nanomaterials in the future.

’ ASSOCIATED CONTENT

bS

Supporting Information. Experimental details, additional AFM images, ICPAES data and TGA trace of the ZHS nanoplates, PXRD patterns of the α-Co(OH)2, β-Co(OH)2, and Ni(OH)2 nanoplates, and SEM images of β-Co(OH)2 and gold nanoplates. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT This research is supported by the UW-Madison NSEC (NSF DMR 0832760). S.A.M. was also partially supported by a 3M Graduate Research Fellowship. S.J. thanks NSF (DMR-0548232 and DMR-1106184) and Sloan Research Fellowship for support. We also thank Professor Younan Xia for helpful discussions about gold nanoplates. ’ REFERENCES (1) Xia, Y.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed. 2009, 48, 60–103. (2) Jin, S.; Bierman, M. J.; Morin, S. A. J. Phys. Chem. Lett. 2010, 1, 1472–1480. (3) Burton, W. K.; Cabrera, N.; Frank, C. Philos. Trans. Roy. Soc. London, Ser. A 1951, 243, 299–358. (4) Markov, I. V. Crystal Growth For Beginners: Fundamentals of Nucleation, Crystal Growth, and Epitaxy, 1st ed.; World Scientific Publishing Co. Pte. Ltd.: Singapore, 1995. (5) Gibbs, J. W.; Bumstead, H. A.; Longley, W. R.; Van Name, R. G. On the Equilibrium of Heterogeneous Substances, Collected Works; Longmans, Green and Co.: New York, 1928. (6) Bierman, M. J.; Lau, Y. K. A.; Kvit, A. V.; Schmitt, A. L.; Jin, S. Science 2008, 320, 1060–1063. (7) Lau, Y. K. A.; Chernak, D. J.; Bierman, M. J.; Jin, S. J. Am. Chem. Soc. 2009, 131, 16461–16471. (8) Morin, S. A.; Bierman, M. J.; Tong, J.; Jin, S. Science 2010, 328, 476–480. (9) Morin, S. A.; Jin, S. Nano Lett. 2010, 10, 3459–3463. (10) Meng, F.; Morin, S. A.; Jin, S. J. Am. Chem. Soc. 2011, 133, 8408–8411. (11) Cherns, D.; Meshi, L.; Griffiths, I.; Khongphetsak, S.; Novikov, S. V.; Farley, N. R. S.; Campion, R. P.; Foxon, C. T. Appl. Phys. Lett. 2008, 93, 111911. (12) Jacobs, B. W.; Crimp, M. A.; McElroy, K.; Ayres, V. M. Nano Lett. 2008, 8, 4353–4358. (13) Li, Y.; Wu, Y. Chem. Mater. 2010, 22, 5537–5542. (14) Zhu, J.; Peng, H. L.; Marshall, A. F.; Barnett, D. M.; Nix, W. D.; Cui, Y. Nat. Nanotechnol. 2008, 3, 477–481. (15) Reddy, M. V.; Yu, T.; Sow, C. H.; Shen, Z. X.; Lim, C. T.; Rao, G. V. S.; Chowdari, B. V. R. Adv. Funct. Mater. 2007, 17, 2792–2799. (16) Yu, T.; Zhu, Y. W.; Xu, X. J.; Shen, Z. X.; Chen, P.; Lim, C. T.; Thong, J. T. L.; Sow, C. H. Adv. Mater. 2005, 17, 1595–1599.

LETTER

(17) Gao, X. D.; Li, X. M.; Yu, W. D.; Peng, F.; Zhang, C. Y. Mater. Res. Bull. 2006, 41, 608–611. (18) Huang, X.; Tang, S.; Mu, X.; Dai, Y.; Chen, G.; Zhou, Z.; Ruan, F.; Yang, Z.; Zheng, N. Nat. Nanotechnol. 2011, 6, 28–32. (19) Wang, H. L.; Casalongue, H. S.; Liang, Y. Y.; Dai, H. J. J. Am. Chem. Soc. 2010, 132, 7472–7477. (20) Zou, G. F.; Li, H.; Zhang, D. W.; Xiong, K.; Dong, C.; Qian, Y. T. J. Phys. Chem. B 2006, 110, 1632–1637. (21) Li, Y. G.; Tan, B.; Wu, Y. Y. Chem. Mater. 2008, 20, 567–576. (22) Kong, D. S.; Dang, W. H.; Cha, J. J.; Li, H.; Meister, S.; Peng, H. L.; Liu, Z. F.; Cui, Y. Nano Lett. 2010, 10, 2245–2250. (23) Chen, S. H.; Carroll, D. L. J. Phys. Chem. B 2004, 108, 5500– 5506. (24) Sun, X. P.; Dong, S. J.; Wang, E. Angew. Chem., Int. Ed. 2004, 43, 6360–6363. (25) Xu, F.; Yuan, Z. Y.; Du, G. H.; Halasa, M.; Su, B. L. Appl. Phys. A 2007, 86, 181–185. (26) Masuda, Y.; Nagahama, D.; Itahara, H.; Tani, T.; Seo, W. S.; Koumoto, K. J. Mater. Chem. 2003, 13, 1094–1099. (27) Geim, A. K. Science 2009, 324, 1530–1534. (28) Hong, S. S.; Kundhikanjana, W.; Cha, J. J.; Lai, K. J.; Kong, D. S.; Meister, S.; Kelly, M. A.; Shen, Z. X.; Cui, Y. Nano Lett. 2010, 10, 3118–3122. (29) Yin, Y. D.; Zhang, Q. A.; Hu, Y. X.; Guo, S. R.; Goebl, J. Nano Lett. 2010, 10, 5037–5042. (30) Jiang, H.; Zhao, T.; Li, C. Z.; Ma, J. J. Mater. Chem. 2011, 21, 3818–3823. (31) Bear, I. J.; Grey, I. E.; Madsen, I. C.; Newnham, I. E.; Rogers, L. J. Acta Crystallogr., Sect. B 1986, 42, 32–39. (32) Bear, I. J.; Grey, I. E.; Newnham, I. E.; Rogers, L. J. Aust. J. Chem. 1987, 40, 539–556. (33) Xue, L. H.; Mei, X. T.; Zhang, W. X.; Yuan, L. X.; Hu, X. L.; Huang, Y. H.; Yanagisawa, K. Sens. Actuators, B 2010, 147, 495–501. (34) Wang, L. D.; Liu, G. C.; Zou, L. J.; Xue, D. F. J. Alloys Compd. 2010, 493, 471–475. (35) Cao, B. Q.; Cai, W. P. J. Phys. Chem. C 2008, 112, 680–685. (36) Cao, X. L.; Zeng, H. B.; Wang, M.; Xu, X. J.; Fang, M.; Ji, S. L.; Zhang, L. D. J. Phys. Chem. C 2008, 112, 5267–5270. (37) Chu, D. W.; Hamada, T.; Kato, K.; Masuda, Y. Phys. Status Solidi A 2009, 206, 718–723. (38) Govender, K.; Boyle, D. S.; Kenway, P. B.; O’Brien, P. J. Mater. Chem. 2004, 14, 2575–2591. (39) Huang, M. H.; Mao, S.; Feick, H.; Yan, H. Q.; Wu, Y. Y.; Kind, H.; Weber, E.; Russo, R.; Yang, P. D. Science 2001, 292, 1897–1899. (40) Law, M.; Greene, L. E.; Johnson, J. C.; Saykally, R.; Yang, P. D. Nat. Mater. 2005, 4, 455–459. (41) Wang, Z. L.; Song, J. H. Science 2006, 312, 242–246. (42) Viswanath, B.; Kundu, P.; Mukherjee, B.; Ravishankar, N. Nanotechnology 2008, 19, 1–7. (43) Suito, E.; Uyeda, N. Bull. Inst. Chem. Res., Kyoto Univ. 1965, 42, 511–541. (44) Teng, H. H.; Dove, P. M.; De Yoreo, J. J. Geochim. Cosmochim. Acta 2000, 64, 2255–2266. (45) Morin, S. A.; Amos, F. F.; Jin, S. J. Am. Chem. Soc. 2007, 129, 13776–13777. (46) Evans, D. G.; Slade, R. C. T. Struct. Bonding (Berlin, Ger.) 2006, 119, 1–87. (47) Liu, Z. P.; Ma, R. Z.; Osada, M.; Takada, K.; Sasaki, T. J. Am. Chem. Soc. 2005, 127, 13869–13874. (48) Wang, H. L.; Robinson, J. T.; Diankov, G.; Dai, H. J. J. Am. Chem. Soc. 2010, 132, 3270–3271. (49) Wang, F.; Liu, R.; Pan, A.; Cao, L.; Cheng, K.; Xue, B.; Wang, G.; Meng, Q.; Li, J.; Li, Q.; Wang, Y.; Wang, T.; Zou, B. Mater. Lett. 2007, 61, 2000–2003. (50) Wang, X. B.; Cai, W. P.; Lin, Y. X.; Wang, G. Z.; Liang, C. H. J. Mater. Chem. 2010, 20, 8582–8590. (51) Zhang, W. X.; Yanagisawa, K. Chem. Mater. 2007, 19, 2329– 2334. 4454

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455

Nano Letters

LETTER

(52) Illy, B. N.; Ingham, B.; Ryan, M. P. Cryst. Growth Des. 2010, 10, 1189–1193. (53) Goodwin, F. E. In Kirk-Othmer encyclopedia of chemical technology, 5th ed.; Kroschwitz, J., Ed.; J. Wiley: Hoboken, NJ, 2007; Vol. 26, pp 617 619. (54) Eshelby, J. D. J. Appl. Phys. 1953, 24, 176–179. (55) Frank, F. C. Acta Crystallogr. 1951, 4, 497–501. (56) Teng, H. H.; Dove, P. M.; Orme, C. A.; De Yoreo, J. J. Science 1998, 282, 724–727. (57) De Yoreo, J. J.; Dove, P. M. Science 2004, 306, 1301–1302. (58) Yao, N.; Epstein, A.; Akey, A. J. Mater. Res. 2006, 21, 1939– 1946.

4455

dx.doi.org/10.1021/nl202689m |Nano Lett. 2011, 11, 4449–4455