In Situ Observations of Free-Standing Graphene-like Mono- and

Oct 8, 2015 - Comparative Study on ZnO Monolayer Doped with Al, Ga and In Atoms as Transparent Electrodes. Dan Sun , Changlong Tan , Xiaohua Tian ...
0 downloads 0 Views 5MB Size
Download PDF
)

)

Huy T. Quang,4†,‡ Alicja Bachmatiuk,§, Arezoo Dianat,^,z Frank Ortmann,^,z Jiong Zhao,4† Jamie H. Warner,4 Ju¨rgen Eckert,¥, Gianaurelio Cunniberti,^,#,z and Mark H. Ru¨mmeli*,§, ,f

ARTICLE

In Situ Observations of Free-Standing Graphene-like Mono- and Bilayer ZnO Membranes †

)

IBS Center for Integrated Nanostructure Physics, Institute for Basic Science, and ‡Departments of Energy Science and Physics, Sungkyunkwan University, Suwon 440-746, Korea, §Institute for Complex Materials, IFW Dresden, P.O. Box D-01171, 01069 Dresden, Germany, Centre of Polymer and Carbon Materials, Polish Academy of Sciences, M. Curie-Sklodowskiej 34, Zabrze 41-819, Poland, ^Institute for Materials Science and Max Bergman Center of Biomaterials, # Center for Advancing Electronics Dresden, and zDresden Center for Computational Materials Science (DCMS), TU Dresden, 01062 Dresden, Germany, 4Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, United Kingdom, ¥Erich Schmid Institute of Materials Science, Austrian Academy of Sciences, Jahnstraße 12, A-8700 Leoben, Austria, Department Materials Physics, Montanuniversität Leoben, Jahnstraße 12, A-8700 Leoben, Austria, and fCollege of Physics Optoelectronics and Energy, Soochow University, 215006 Suzhou, P. R. China

ABSTRACT ZnO in its many forms, such as bulk, thin films,

nanorods, nanobelts, and quantum dots, attracts significant attention because of its exciting optical, electronic, and magnetic properties. For very thin ZnO films, predictions were made that the bulk wurtzite ZnO structure would transit to a layered graphene-like structure. Graphene-like ZnO layers were later confirmed when supported over a metal substrate. However, the existence of free-standing graphenelike ZnO has, to the best of our knowledge, not been demonstrated. In this work, we show experimental evidence for the in situ formation of free-standing graphene-like ZnO mono- and bilayer ZnO membranes suspended in graphene pores. Local electron energy loss spectroscopy confirms the membranes comprise only Zn and O. Image simulations and supporting analysis confirm that the membranes are graphene-like ZnO. Graphene-like ZnO layers are predicted to have a wide band gap and different and exciting properties as compared to other ZnO structures. KEYWORDS: ZnO thin films . graphene-like ZnO . free-standing membrane . TEM

M

etal oxides are a fascinating class of materials with diverse physical, chemical, and electronic properties.1 Metal oxides are widely used in electronic and magnetic devices, in heterogeneous catalysis, and in other applications.2 Among the metal oxides, zinc oxide (ZnO) has emerged as a unique material because of its semiconducting and piezoelectric properties. ZnO is a versatile functional material and can exist with a multitude of morphologies such as nanotubes, nanowires, nanorods, nanobelts, tetrapods, and nanoribbons.3 They have a wide band gap (3.37 eV) and a large excitation binding energy, which lend themselves to a variety of applications such as in transparent electronics, ultraviolet (UV) light emitters, piezoelectric devices, chemical sensors, spin electronics, and catalysis.3 QUANG ET AL.

ZnO is a material that has a range of crystalline structures, whose configurations include hexagonal wurtzite and cubic zincblende-type.4 Zinc oxide with a hexagonal wurtzite structure is the most thermodynamically stable and is thus the most commonly found form. It has a hexagonal structure which consists of two zinc and two oxygen atoms per unit cell with the lattice parameters a = b = 0.3249 nm and c = 0.5204 nm.4 The cubic zinc blende structure of ZnO can be thought of as an arrangement of two interpenetrating facecentered cubic sublattices, displaced by 1/4 of the body diagonal axis with the lattice constant a = 0.4614 nm. Unlike the wurtzite phase, the zinc blende phase is a metastable state. It can be stabilized by epitaxial growth on cubic substrates, which are also of a zinc blende structure.5 VOL. XXX

’

NO. XX

’

* Address correspondence to [email protected]. Received for review August 26, 2015 and accepted October 8, 2015. Published online 10.1021/acsnano.5b05481 C XXXX American Chemical Society

000–000

’

XXXX

A www.acsnano.org

RESULTS AND DISCUSSION In this work, we demonstrate through in situ observations the formation of suspended free-standing single-atom-thick g-ZnO membranes that form in graphene pores using low-voltage spherical aberrationcorrected transmission microscopy (LVACTEM). Together with TEM morphology and spectroscopic observations, image simulations and density functional theory (DFT) calculations were also conducted to better comprehend the free-standing graphene-like ZnO membranes. The use of graphene pores to serve as platforms for 2D membrane formation was previously shown for the formation of 2D iron; here, we show that the technique can be extended to g-ZnO.17 The formation of ZnO species over graphene was achieved by first placing a small quantity of zinc acetylacetonate powder next a lacey carbon grid with clean transferred graphene on it in a glass tube. The glass tube was then evacuated to below 3  10 6 mbar and then sealed using a flame. The sealed glass tube was subsequently heated at a temperature of 300 C for 12 h during which the zinc acetylacetonate sublimes and decomposes, forming a ZnO nanospecies over the graphene,18 as shown on Figure S1. Analytical studies, discussed later, confirm that the structures comprise Zn and O. TEM examinations showed that ZnO nanoparticles (NPs) had formed on the few-layer graphene. The larger NPs tended to be of the wurtzite phase (e.g., Figure S2, in the Supporting Information). QUANG ET AL.

ARTICLE

The isolation of graphene, a two-dimensional honeycomb lattice of sp2-bonded carbon atoms, and its extraordinary properties6 has triggered the development of new classes of two-dimensional materials, particularly those with an intrinsic band gap. Both theory and experiment have demonstrated that the ZnO wurtzite structure can transform to a planar, graphene-like structure in which Zn and O atoms reside in a trigonalplanar coordination instead of the bulk tetrahedral configuration when ZnO is thinned down to few atomic layers.7,8 Once it transforms to a graphene-like ZnO (g-ZnO) structure, its lattice parameter expands by 1.6% (a = 3.303 Å).8 Monolayer g-ZnO has a direct (wide) band gap of 3.37 eV, which is very promising for switching electronics applications. Thus far, only a few studies have succeeded in synthesizing g-ZnO ranging from one to a few atomic layers in thickness. However, all of these examples required a metal substrate for the g-ZnO to form over. Substrate examples include Ag,8 Pd,9 Pt,10 or Au.11 However, g-ZnO over a substrate might lead to interactions with the substrate, in turn, leading to modified or altered properties. Thus, it is important to demonstrate the feasibility of free-standing g-ZnO. To do this, we exploit the use electron-beam-produced pores in graphene to stabilize g-ZnO. The use of pores in graphene to manipulate and stabilize structures is relatively well-established.12 17

Figure 1. TEM images showing the dynamic activity of ZnO nanocrystals while under electron beam irradiation. (a) ZnO nanocrystal over graphene. (b) ZnO nanostructure becomes amorphous. (c) Self-arrangement to form a g-ZnO single layer; the inset shows the fast Fourier transform of the g-ZnO membrane. (d f) Collapse of the g-ZnO membrane with extended irradiation. All scale bars = 1 nm.

The measured d spacings were 2.6 Å (Figure S2b), which match with the wurtzite (0002) orientation.19 Graphene folds near the NP suggest that the NP is embedded in graphene. An image simulation of a ZnO (0002) NP embedded in graphene (Figure S2e) provided for comparison shows a very good match with the experimental data. In general, the larger NPs and embedded NPs were found to be relatively stable upon extended electron beam irradiation (see Figure S2a d). However, in many cases, we observed smaller ZnO nanostructures that upon electron beam irradiation were not stable. An example is provided in Figure 1, where a small NP (ca. 1 2 nm o.d.) changes as it is irradiated going from an ordered crystalline structure (panel a) to an amorphous structure (panel b) and then back to a crystalline structure (panel c). These changes can also be seen in the supporting video. Continued irradiation leads to further dynamic changes to the structure (panels d f). In addition, one can also see a hole forming (panels d f), indicating removal of the supporting graphene which is known to be sputtered away once a defect forms because the knock-on threshold for such dangling bonds is ca. 50 kV, which is well below our electron acceleration measurement voltage of 80 kV.20 This behavior was very reproducible, and closer examination of the NPs prior to their collapse at the edge of a growing graphene pore (e.g., panels e and f) shows that the NPs form as a suspended membrane in the graphene hole. This is confirmed by filtering out the graphene signals (using Fourier masking), as for example shown in Figure S3, in which the ZnO NP signal is filtered out, leaving only the graphene. A hole indicating a graphene pore where the NP resides is clearly observed. Closer examination of the ZnO structures residing on graphene (e.g., Figure S4) or in a graphene pore shows lattice spacings determined from the micrographs directly and average spacings from reflexes in the Fourier domain, between 2.86 and 2.85 Å ((0.02 Å), respectively. These values fit better VOL. XXX

’

NO. XX

’

000–000

’

B

XXXX www.acsnano.org

ARTICLE Figure 2. Characterization of a free-standing g-ZnO monolayer membrane suspended in a graphene pore. (a) TEM image of g-ZnO membrane in graphene hole. (b e) Image simulation of single-layer and bilayer T1, T2, and wurtzite structure of ZnO, respectively. (f) Normalized intensity profiles from the TEM image (pink line) and from the simulated image of single-layer (green line), bilayer T1 (blue line), bilayer T2 (red line), and bilayer wurtzite (yellow line), corresponding to marked profiles in pink, green, blue, red, and yellow line in (a e). (g j) Stick and ball models of ZnO corresponding to images in (b e). All scale bars = 1 nm.

with graphene-like (1010) ZnO (2.86 Å) as compared to the wurtzite (1010) structure (2.81 Å). To better evaluate the graphene-like ZnO (g-ZnO) structures, we conducted image simulations for various g-ZnO structures to confirm the presence of g-ZnO and to determine the number of graphene-like layers from relative intensity measurements.17 Figure 2a shows an experimental micrograph for a monolayer g-ZnO free-standing membrane suspended in a graphene pore. Image simulations for monolayer g-ZnO (panel b), the bilayer T1 and T2 stacking configurations, and the bilayer wurtzite structure (panels c, d, and e, respectively) are provided for visual comparison. The g-ZnO H1, H2, and H3 bilayer stacking configurations are provided in Figure S5 in the Supporting Information. From visual comparisons, the obtained TEM micrograph best fits either monolayer g-ZnO or bilayer T1 g-ZnO. Intensity profiles across Zn and O atoms or columns for these configurations show that the intensity profile from the experimental image is in excellent agreement with monolayer g-ZnO and in poor agreement with all the other configurations. We also examined the effect of the lattice mismatch between the g-ZnO membrane/graphene interface. Figure S6 in the Supporting Information shows the lattice spatial variations across a free-standing g-ZnO monolayer. It clearly shows atom spacing compression at the g-ZnO/graphene interface, and as one moves toward the center on the membrane, the interatomic spacing relaxes to the higher value of 3.3 Å, in agreement with the calculated value.8 QUANG ET AL.

Not all of the observed suspended membranes were found to be homogeneous monolayer g-ZnO membranes. Some were found to be a mix of layers (usually this was for membranes >2 nm in diameter). An example is provided in Figure 3, in which intensity profile simulations (including wurtzite ZnO) are compared with the experimental image, indicating a freestanding membrane comprising both monolayer and bilayer g-ZnO. In the case of supported g-ZnO structures over graphene, examples of mixed layer structures were not observed. However, in addition to supported monolayer g-ZnO structures, supported bilayer structures were also observed, as shown in Figure S7 in the Supporting Information. Eighty percent of the bilayer nanocrystalline g-ZnO structures (supported or freestanding) matched with T2 stacking, while the remaining 20% matched with T1 stacking. This is in agreement with calculations by Topsakal et al.,21 which show T2 stacking to be the most energetically favored. We now examine the behavior of free-standing g-ZnO membranes in graphene pores while under electron beam irradiation. As already discussed above (Figure 1), the g-ZnO membranes are dynamic under the electron beam. Figure S8 shows a larger membrane comprising monolayer g-ZnO and bilayer (T2) g-ZnO for different irradiation periods. The g-ZnO membrane is seen to erode and become smaller with extended irradiation. However, the membrane which comprises both mono- and bilayer sections erodes differently, with the bilayer section reducing faster than the VOL. XXX

’

NO. XX

’

000–000

’

C

XXXX www.acsnano.org

ARTICLE Figure 3. Characterization of g-ZnO mono- and bilayer mixed free-standing membrane. (a,b) TEM image and image simulation of the g-ZnO membrane with a single-layer and bilayer in a graphene hole. Inset: Mono- and bilayer with a wurtzite structure. (c) Normalized intensity profiles from the experimental image corresponding to marked profiles with the green line (bi-monolayer), red line (bilayer vacuum), and black line (monolayer vacuum) in (a) and the light blue dashed line (wurtzite) in the inset of (b). (d) Normalized intensity profiles from image simulation corresponding to marked profiles with the pink line (bi-monolayer), dark blue line (bilayer vacuum), yellow line (monolayer vacuum) in (b) and light blue dashed line (wurtzite). All scale bars = 2 nm.

monolayer section. To better comprehend the formation and stability of the g-ZnO membranes, we employ DFT simulations of g-ZnO membranes inside the graphene hole (Figure 4). Three different configurations were investigated, namely, mono-, bi-, and trilayer g-ZnO in the graphene hole with an equal numbers of atoms (C, Zn, O). The calculations for the energy, ΔE, per number of atoms, where ΔE is the energy difference between the monolayer and bilayer and between the bilayer and trilayer, show energy values of 0.44 and 0.13 eV, respectively. This points to the monolayer being the most stable and the progressive layers becoming less favorable. This is in keeping with our experimental observations (no trilayer g-ZnO was observed, and monolayer was more frequently observed than bilayer g-ZnO). This is attributed to the stabilizing effect of the graphene edges of a pore, which is only available for a single g-ZnO layer. In contrast, we find that, without a graphene support, a bilayer structure is energetically more stable than monolayer ZnO, with a decreasing energy difference for larger flake sizes (Figure S9). In the case where both layers are complete (infinite), we estimate a binding energy between the QUANG ET AL.

Figure 4. DFT simulation of free-standing g-ZnO growth inside a graphene hole for mono-, bi-, and trilayer structures. (a) Top view and (b) 3D view of graphene-like ZnO (monolayer) in the graphene hole. (c) Top view and (d) 3D view of bilayer ZnO growth in graphene. (e) Top view and (f) 3D view of trilayer ZnO growth in graphene.

two layers of about 0.1 eV per atom. To gain more insight into possible growth processes, we also conducted calculations to compare g-ZnO lateral growth as compared to layer growth for small membranes in the presence of a dimer Zn O feedstock. The calculations show that the difference in the total energy upon incremental adsorption of Zn O dimers favors adsorption at the edge as opposed to on-top adsorption (basal plane of g-ZnO), as shown in Figure S10 in the Supporting VOL. XXX

’

NO. XX

’

000–000

’

D

XXXX www.acsnano.org

ARTICLE Figure 5. Comparison of EELS spectra of a free-standing g-ZnO membrane and a bulk wurtzite ZnO nanoparticle. (a) Oxygen K-edge and (b) zinc L-edge EELS spectra of the bulk wurtzite ZnO nanoparticle. (c) Oxygen K-edge and (d) zinc L-edge EELS spectra from a free-standing g-ZnO membrane. The inset shows no signal for the C 1s edge (from the membrane).

Information. This is in agreement with our experimental observations, where we preferentially observe free-standing g-ZnO as compared to bilayer g-ZnO. Finally, we look at the analytical results of the membranes studied above. Energy-dispersive X-ray (EDX) spectroscopy (Figure S11) confirms the presence of Zn and O. However, to confirm that g-ZnO membranes are formed solely of Zn and O, we perform local electron energy loss spectroscopy (EELS). The EELS studies confirmed the presence of Zn and O by the presence of the O K-edge and Zn L1,2,3-edge. Moreover, no carbon was found for the free-standing membranes, further confirming the presence of g-ZnO (see Figure 5). In addition, the O K-edges and Zn L1,2,3-edges differ from those measured for bulk wurtzite nanostructures, which are in agreement with published values in the literature.22 In the case of the O K-edge, a broader peak at ca. 540 eV is observed. No significant difference in the rising edge is observed. For the Zn L1,2,3-edge, the rising edge of the L3 peak occurs at a lower energy as compared to the bulk spectrum. In addition, the L2-edge is downshifted. These differences in the Zn and O core level spectra from

METHOD Sample Preparation. A single-layer graphene was grown over polished high-purity copper foil (99.99%) using atmospheric pressure chemical vapor deposition with methane (99.999%) as the feedstock. To transfer graphene, the as-produced sample

QUANG ET AL.

the g-ZnO membranes (as compared to bulk ZnO) are attributed to differences in bonding configurations. CONCLUSION In summary, we show experimental evidence for the formation of suspended free-standing single- and bilayer graphene-like ZnO membranes in graphene pores and on the surface of graphene. The g-ZnO structures form in situ under electron beam irradiation. Local EELS investigations confirm that the free-standing structures are composed solely of Zn and O. Slight shifts in the core level edges were found, and this is attributed to differences in the bonding configurations between g-ZnO and bulk wurtzite ZnO. DFT calculations indicate that monolayer g-ZnO forming in a graphene pore is favored over bilayer and trilayer g-ZnO, in agreement with our experimental observations. In the cases where we did observe bilayer g-ZnO membranes, T2 stacking configurations were the dominant stacking order, in agreement with previous theoretical predictions.21 These studies provide new insight into graphene-like ZnO mono- and bilayer membranes and g-ZnO as a whole.

was coated with poly(methyl methacrylate) (PMMA). It was then floated on a copper etchant (CE-100, Transene) for 30 min. After being rinsed in deionized water a few times, the PMMA/ graphene was then transferred onto a standard lacey carbon TEM grid. After being dried, the PMMA was removed by

VOL. XXX

’

NO. XX

’

000–000

’

E

XXXX www.acsnano.org

Conflict of Interest: The authors declare no competing financial interest. Supporting Information Available: The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.5b05481. Additional sample preparation procedure, TEM, DFT calculation, and EDX in Figures S1 S11 (PDF) Membrane generation (MP4) Acknowledgment. This work was supported by the Institute for Basic Science (IBS-R011-D1) and in part by BK21-plus through the Ministry of Education, Korea. A.B. thanks the National Science Centre for the financial support within the frames of the Sonata Program (Grant Agreement 2014/13/D/ST5/02853). A.D. gratefully thanks the financial support from the Free State of Saxony of Germany (SMWK) and computational resources by the Center for Information Services and High Performance Computing (ZIH) of the TU Dresden. Computational resources from the Center for Information Services and High Performance Computing (ZIH) of the TU Dresden are gratefully acknowledged.

REFERENCES AND NOTES 1. Henrich, V. E.; Cox, P. A. The Surface Science of Metal Oxides; Cambridge University Press: Cambridge, 1996. 2. Umar, A. Metal Oxide Nanostructures and Their Applications; American Scientific Publishers: Valencia, CA, 2010. 3. Wang, Z. L. Zinc Oxide Nanostructures: Growth, Properties and Applications. J. Phys.: Condens. Matter 2004, 16, R829– R858. 4. Morkoc, H.; Ozgur, U. Zinc Oxide: Fundamentals, Materials and Device Technology; Wiley-VCH: Weinheim, Germany, 2009. 5. Klingshirn, C. F.; Waag, A.; Hoffmann, A.; Geurts, J. Zinc Oxide: From Fundamental Properties towards Novel Applications; Springer: Heidelberg, 2010.

QUANG ET AL.

6. Warner, J. H.; Schäffel, F.; Rümmeli, M. H.; Bachmatiuk, A. Graphene Fundamentals and Emergent Applications; Elsevier: Oxford, 2013. 7. Claeyssens, F.; Freeman, C. L.; Allan, N. L.; Sun, Y.; Ashfold, M. N. R.; Harding, J. H. Growth of ZnO Thin Films-Experiment and Theory. J. Mater. Chem. 2005, 15, 139–148. 8. Tusche, C.; Meyerheim, H. L.; Kirschner, J. Observation of Depolarized ZnO(0001) Monolayers: Formation of Unreconstructed Planar Sheets. Phys. Rev. Lett. 2007, 99, 026102. 9. Weirum, G.; Barcaro, G.; Fortunelli, A.; Weber, F.; Schennach, R.; Surnev, S.; Netzer, F. P. Growth and Surface Structure of Zinc Oxide Layers on a Pd(111) Surface. J. Phys. Chem. C 2010, 114, 15432–15439. 10. Liu, B. H.; McBriarty, M. E.; Bedzyk, M. J.; Shaikhutdinov, S.; Freund, H. J. Structural Transformations of Zinc Oxide Layers on Pt(111). J. Phys. Chem. C 2014, 118, 28725– 28729. 11. Deng, X.; Yao, K.; Sun, K.; Li, W.-X.; Lee, J.; Matranga, C. Growth of Single- and Bilayer ZnO on Au(111) and Interaction with Copper. J. Phys. Chem. C 2013, 117, 11211– 11218. 12. Yang, Z.; Yin, L.; Lee, J.; Ren, W.; Cheng, H.-M.; Ye, H.; Pantelides, S. T.; Pennycook, S. J.; Chisholm, M. F. Direct Observation of Atomic Dynamics and Silicon Doping at a Topological Defect in Graphene. Angew. Chem. 2014, 126, 9054–9058. 13. Lee, J.; Zhou, W.; Pennycook, S. J.; Idrobo, J.-C.; Pantelides, S. T. Direct visualization of reversible dynamics in a Si6 cluster embedded in a graphene pore. Nat. Commun. 2013, 4, 1650. 14. Guo, J.; Lee, J.; Contescu, C. I.; Gallego, N. C.; Pantelides, S. T.; Pennycook, S. J.; Moyer, B. A.; Chisholm, M. F. Crown ethers in graphene. Nat. Commun. 2014, 5, 5389. 15. Lee, J.; Yang, Z.; Zhou, W.; Pennycook, S. J.; Pantelides, S. T.; Chisholm, M. F. Stabilization of graphene nanopore. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 7522–7526. 16. Zan, R.; Ramasse, Q. M.; Bangert, U.; Novoselov, K. S. Graphene Reknits Its Holes. Nano Lett. 2012, 12, 3936– 3940. 17. Zhao, J.; Deng, Q.; Bachmatiuk, A.; Sandeep, G.; Popov, A.; Eckert, J.; Rümmeli, M. H. Free-Standing Single-Atom-Thick Iron Membranes Suspended in Graphene Pores. Science 2014, 343, 1228–1232. 18. Music, S.; Saric, A.; Popovic, S. Formation of Nanosize ZnO Particles by Thermal Decomposition of Zinc Acetylacetonate Monohydrate. Ceram. Int. 2010, 36, 1117–1123. 19. Ge, M. Y.; Wu, H. P.; Niu, L.; Liu, J. F.; Chen, S. Y.; Shen, P. Y.; Zeng, Y. W.; Wang, Y. W.; Zhang, G. Q.; Jiang, J. Z. Nanostructured ZnO: From Monodisperse Nanoparticles to Nanorods. J. Cryst. Growth 2007, 305, 162–166. 20. Warner, J. H.; Rummeli, M. H.; Ge, L.; Gemming, T.; Montanari, B.; Harrison, N. M.; Buchner, B.; Briggs, G. A. D. Structural Transformations in Graphene Studied with High Spatial and Temporal Resolution. Nat. Nanotechnol. 2009, 4, 500–504. 21. Topsakal, M.; Cahangirov, S.; Bekaroglu, E.; Ciraci, S. FirstPrinciples Study of Zinc Oxide Honeycomb Structures. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 235119. 22. Ding, Y.; Wang, Z. L. Electron Energy-Loss Spectroscopy Study of ZnO Nanobelts. J. Electron Microsc. 2005, 54, 287– 291. 23. VandeVondele, J.; Krack, M.; Mohamed, F.; Parrinello, M.; Chassaing, Th.; Hutter, J. Comput. Phys. Commun. 2005, 167, 103. 24. Goedecker, S.; Teter, M.; Hutter, J. Separable dual-space Gaussian pseudopotentials. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 1703. 25. Grimme, S. Semiempirical GGA-Type Density Functional Constructed With a Long-Range Dispersion Correction. J. Comput. Chem. 2006, 27, 1787–1799.

VOL. XXX

’

NO. XX

’

000–000

’

ARTICLE

exposure to acetone vapor. Both the graphene/TEM grid and the zinc acetylacetonate, Zn(acac)2, powder were load inside a glass tube (one side sealed), as shown in Figure S1. The glass tube was pumped down to 10 6 mbar by a turbo-pump and then sealed at the open end to maintain vacuum. It was then annealed at 300 C for 12 h. Characterization. A FEI Titan3 80-300 transmission electron microscope with an image CEOS spherical (Cs) aberration corrector operating with an acceleration voltage of 80 kV was used. The samples were measured at room temperature. The vacuum during measurement was ca. 10 7 mbar. All image simulations used parameters equivalent to those used in the TEM experiments. The scanning transmission electron microscopy, EELS, and EDX experiments were conducted with a JEOL (ARM) 200F TEM with a probe Cs corrector operating with an acceleration voltage of 80 kV. Low-pass filtering has been applied to the micrographs to reduce noise. The filtering does not affect the final resolution of the images. Image Simulation. All image simulations used parameters equivalent to those used in the TEM experiments. The multislice high-resolution TEM image simulations were performed using JEMS software. For the simulations, an accelerating voltage of 80 kV with an energy spread of 0.2 eV was used. The chromatic aberration Cc was set to 1 mm, and the spherical aberration, Cs, to 1 μm. A defocus of between 2 and 3 nm and a defocus spread of 2 nm were implemented. These values are consistent with the experimental conditions used. DFT Calculation. DFT calculations were performed within a combined plane-wave and atomic orbital approach as implemented in the cp2k code (www.cp2k.org).23 A Perdew Burke Ernzerhof exchange-correlation functional and its corresponding norm-conserving pseudopotential GTH were used.24 In addition, we used a DZVP (double-ζ for valence electrons plus polarization functions) basis set complemented with a planewave basis with an energy cutoff of 350 Ry. Dispersion corrections were further taken into account through the standard Grimme parametrization.25 The convergence criteria for both geometry and energy calculations were set to 1  10 7 Hartree for the SCF energy and 9  10 4 Hartree Å 1.

F

XXXX www.acsnano.org