Corresponding Author

Bettina V. Lotsch

Max Planck Institute for Solid State Research

References

1.
Ma, R.; Sasaki, T.
Nanosheets of Oxides and Hydroxides: Ultimate 2D Charge-Bearing Functional Crystallites
2.
Osada, M.; Sasaki, T.
Two-Dimensional Dielectric Nanosheets: Novel Nanoelectronics From Nanocrystal Building Blocks
3.
Keller, S.W.; Kim, H.-N.; Mallouk, T.E.
Layer-by-Layer Assembly of Intercalation Compounds and Heterostructures on Surfaces: Toward Molecular "Beaker" Epitaxy

In collaboration with:

C. Scheu (LMU München)

M. Bugnet, G.A. Botton (McMaster University, Hamilton, Ontario, Canada)

Research Group "Nanochemistry"

Artificial solids by design: Assembly and electron microscopy study of nanosheet-derived heterostructures

Authors

C. Ziegler, S. Werner, M. Wörsching, V. Duppel, and B.V. Lotsch

Departments

Research Group "Nanochemistry"

We present the bottom-up fabrication of an artificial superlattice derived from positively charged layered double hydroxide (LDH) and negatively charged perovskite nanosheets sequentially assembled by electrostatic layer-by-layer deposition. In contrast to previously employed bulk methods, we use a combination of HRTEM, STEM and EEL spectroscopy to elucidate the structure and composition of the multilayer stack with high spatial resolution on the subnanometer scale. In addition, we demonstrate that the packing density of the multilayer system can be tuned by changing the LDH dispersing agent.

Introduction

The rational design of solids with tailor-made properties has been a hallmark of "soft chemistry" and a major driving force of modern materials science. In principle, high temperature solid-state synthesis is governed by thermodynamic principles and is therefore inherently constrained with respect to the compositions and structures that can be realized. In contrast, kinetically controlled soft chemistry protocols relying on the use of preformed building blocks and operating at low temperatures open up pathways to unconventional solids with large compositional scope, albeit often at the expense of stability. Driven by the rise of nanochemistry and the ability to sculpture well-defined nanoscale building blocks such as two-dimensional (2D) nanosheets, the modular assembly of preformed nano-objects into hierarchical superlattices has taken shape in recent years. It is anticipated that the design of artificial solids with custom-made properties will be useful in a range of applications, including spintronics, optoelectronics, and catalysis [1,2].

Layer-by-layer assembly of bulk heterostructures

<strong>Fig. 1:</strong> Schematic drawing of the layer-by-layer (LBL) process employed to construct hybrid superlattices: Exfoliation of the layered bulk materials into Mn<sub>2</sub>Al(OH)<sub>6</sub><sup>+</sup> (LDH) and Ca<sub>2</sub>Nb<sub>3</sub>O<sub>10</sub><sup>&minus; </sup>(per) nanosheets is followed by electrostatic LBL assembly with intermediate washing steps in order to achieve 3D (LDH/per)<sub>n</sub> heterostructures immobilized on a substrate. Zoom Image
Fig. 1: Schematic drawing of the layer-by-layer (LBL) process employed to construct hybrid superlattices: Exfoliation of the layered bulk materials into Mn2Al(OH)6+ (LDH) and Ca2Nb3O10(per) nanosheets is followed by electrostatic LBL assembly with intermediate washing steps in order to achieve 3D (LDH/per)n heterostructures immobilized on a substrate. [less]

We present a locally resolved, precise elemental and structural analysis of a heterostructure composed of perovskite (per) and layered double hydroxide (LDH) nanosheets. Both types of nanosheet building blocks were synthesized by intercalation exfoliation protocols. To achieve maximum control over the layer sequence, we used an electrostatic layer-by-layer (LBL) procedure. In LBL, multilayer films form due to electrostatic and hydrophobic forces between positively and negatively charged nanosheets in a self-limiting fashion, thus allowing for a high level of control over the layer sequence at the (sub)nanoscale. The multilayer film was assembled by sequentially adsorbing n cationically charged [Mn2Al(OH)6]+ nanosheets suspended in either formamide or water (LDHfa/wa), and n anionically charged [Ca2Nb3O10] nanosheets suspended in water on a planar Si/SiO2 substrate with intermediate washing steps as shown schematically in Fig. 1.

<strong>Fig. 2:</strong> (a) Bright field HRTEM micrograph of the cross-section of a 100 bilayer (LDH<sub>fa</sub>/per)<sub>100</sub> film obtained by electrostatic LBL assembly. (b) XRD pattern of (LDH/per)<sub>80 </sub>measured in reflection mode. Comparison of STEM cross-section images taken from a (LDH<sub>fa</sub>/per)<sub>100</sub> film assembled with the LDH being dispersed in formamide (c) and a (LDH<sub>aq</sub>/per)<sub>90</sub> film with the LDH dispersed in water (d). Zoom Image
Fig. 2: (a) Bright field HRTEM micrograph of the cross-section of a 100 bilayer (LDHfa/per)100 film obtained by electrostatic LBL assembly. (b) XRD pattern of (LDH/per)80 measured in reflection mode. Comparison of STEM cross-section images taken from a (LDHfa/per)100 film assembled with the LDH being dispersed in formamide (c) and a (LDHaq/per)90 film with the LDH dispersed in water (d). [less]

An overview HRTEM cross-section of a multilayer (LDHfa/per)100 film is shown in Fig. 2(a). The layers are grown in an ordered fashion on the Si/SiO2 substrate, whereas "real structure" effects as opposed to an ideally ordered structure composed of infinitely extended sheets are clearly visible. The layers show a significant degree of flexibility and bending around defects such as sheet terminations, which leads to voids and overlaps in the assembly. According to the alignment of the sheets parallel to the substrate, the XRD pattern of a (LDH/per) multilayer stack shows two (00l) reflections around 5.5 and 11.0° 2θ, pointing to a stacking period of ≈1.6nm (Fig. 2(b)), which is in line with the HRTEM measurements. The number of counted bilayers (100±5) and the effective thickness of the stack (250nm) roughly correspond to that expected for a 100 bilayer film, taking holes and irregularities into account.

In contrast to the (LDHfa/per)100 film, the (LDHaq/per)90 film is less densely stacked and exhibits larger distances between subsequent perovskite layers as depicted in Figs. 2(c),(d). EDX analysis suggests that approximately ten times less Mn2Al(OH)6+ is present in the (LDHaq/per)90 stack compared to the (LDHfa/per)100 film. We attribute the lower amount of LDH in the water-based sample to a proton – TBA exchange equilibrium in the outer [(TBA)1−xHx]+[Ca2Nb3O10] layer. Upon exposure to water, the degree of protonation (given by x) of the exposed perovskite layer increases, thus reducing the apparent layer charge density when assuming only minimal dissociation of the protons as compared to TBA+. As a consequence, the exchange capacity for LDH is reduced. Although the choice of solvent is limited by the dispersibility of the nanosheets, this finding points out an important design criterion in the assembly of nanosheets, which is governed by a subtle interplay between colloidal stability, ionic strength and composition of the Helmholtz double layer, and solution equilibria influenced by the pH.

HR-STEM/EELS analysis

<strong>Fig. 3:</strong> (a) STEM cross-section image of a (LDH<sub>fa</sub>/<sub>per</sub>)<sub>100</sub> film. The inset shows the vertical (vert) and horizontal (horiz) distances between (A) Nb&minus;Nb<sub>horiz</sub>, (B) Nb&minus;Nb<sub>vert</sub>, (C) Ca&minus;Ca<sub>vert</sub>, and (D) Ca&minus;Ca<sub>horiz</sub>. The orange line corresponds to an EELS line-scan, where (b) shows the corresponding summed up EEL spectra, (c) the individual EEL spectra of the Ca-L<sub>2,3</sub> and Nb-M<sub>2,3</sub> edges without shift, and (d) the individual EELspectra of the O-K and Mn-L<sub>2,3</sub> edge with a vertical shift of the blue spectra. In (c) and (d) the spectra from the bright layer (perovskite) are shown in blue and those from the dark layer (LDH) in red. (e) Extracted intensity profiles of Ca-L<sub>2,3</sub>, Mn-L<sub>2,3</sub>, O-K, and the HAADF signal taken along the orange line in (a). Zoom Image
Fig. 3: (a) STEM cross-section image of a (LDHfa/per)100 film. The inset shows the vertical (vert) and horizontal (horiz) distances between (A) Nb−Nbhoriz, (B) Nb−Nbvert, (C) Ca−Cavert, and (D) Ca−Cahoriz. The orange line corresponds to an EELS line-scan, where (b) shows the corresponding summed up EEL spectra, (c) the individual EEL spectra of the Ca-L2,3 and Nb-M2,3 edges without shift, and (d) the individual EELspectra of the O-K and Mn-L2,3 edge with a vertical shift of the blue spectra. In (c) and (d) the spectra from the bright layer (perovskite) are shown in blue and those from the dark layer (LDH) in red. (e) Extracted intensity profiles of Ca-L2,3, Mn-L2,3, O-K, and the HAADF signal taken along the orange line in (a). [less]

Figure 3(a) shows an HR-STEM image where an EELS line-scan was performed along the orange line for the (LDHfa/per)100 film. The bright columns/planes arranged in regular "triplets" originate from the heavy niobium atoms and hence can be related to the perovskite layers composed of three edge-sharing NbO6 octahedra as a fundamental structural motif, while the beam-sensitive LDH resides in the dark regions between the perovskite slabs. 128 EEL spectra were taken along a distance of 16.3nm with a dispersion of 0.5eV/channel. In the range from 169 to 1193eV most major ionization edges of interest are visible, including the Nb-M4,5 with an onset at 205eV and a delayed maximum, the Ca-L2,3 with two white-lines at 346 and 350eV, the Nb-M2,3 at 363 and 378eV with a sharp threshold peak, and the Mn-L2,3 at 640 and 651eV (Fig. 3(b)). The Mn-L2,3 edge is only visible in the dark regions of the stack, whereas the Ca-L2,3 is always present, but with significantly decreased intensity in the dark regions. As already evident from the EDX data, less Mn2Al(OH)6+ than Ca2Nb3O10 is present throughout the sample. Hence, the signal for the Mn-L2,3 edge is weak. Figure 3(c) shows enlarged the normalized signal of Ca-L2,3 and Nb-M2,3, whereas the Mn-L2,3 and the O-K edge are depicted in Fig. 3(d) with the individual spectra of the bright perovskite layer (blue line) and the dark LDH layer (red line), respectively. A change of the O-K near-edge fine structures is visible, indicating a change of the environment of the oxygen atoms when comparing the pristine perovskite with the hybrid superlattice. An intensity profile of the extracted Ca-L2,3, Mn-L2,3 and O-K edges is shown in Fig. 3(e). The maximum of the Mn-L2,3 signal consistently lies in the minimum of the Ca-L2,3 signal, thus unambiguously proving the alternate stacking of individual Mn2Al(OH)6+ nanosheets with Ca2Nb3O10 nanosheets at the nanometer scale. Notably, this finding implies full delamination of the LDH and perovskite sheets in solution or upon interaction with the respective oppositely charged top nanosheet layer.

We have demonstrated the rational synthesis of a complex two component solid composed of Mn2Al(OH)6+ and Ca2Nb3O10 nanosheets by means of "molecular beaker epitaxy" [3], which suggests the feasibility of the rational design of complex solids through judicious combination of nanosheet building blocks. For the first time, we have been able to map out the structure and elemental distribution within the stacks locally resolved by HRTEM, STEM, and EELS analysis and ascertained the presence of alternating, compositionally distinct LDH and perovskite layers of consistent and reproducible minimum thickness, i.e., single sheets. With this study we add evidence to the notion that solids can in fact be "designed" in a rational way by taking advantage of soft chemistry routes to metastable solids, rather than relying on traditional solid-state synthesis which is largely driven by thermodynamic principles.


Reprinted with permission from (Ziegler, C., Werner, S., Bugnet, M., Wörsching, M., Duppel, V., Botton, G.A., Scheu, C., Lotsch, B.V. Chem. Mater., 2013, 25 (24), 4892). Copyright (2013) American Chemical Society.


 
loading content