A new generation of electron beam monochromators has recently pushed the energy resolution of (scanning) transmission electron microscopes deep into the sub 20meV range [1]. In addition to the obvious increase in resolution which has made exploring the phonon region of the EELS spectrum possible [1], the increased flexibility of these instruments is proving hugely advantageous for materials science investigations. The energy resolution, beam current and electron optics can be adjusted seamlessly and traded off each other as necessary within a greatly increased range, as will be illustrated on a number of systems studied using a Nion UltraSTEM100MC ‘Hermes’ instrument recently installed at the SuperSTEM Laboratory.

This added flexibility was essential in particular in studying the crystal structure of Li- and Mn-rich transition metal oxides, whose pristine state is not yet fully understood in spite of their great potential as high-capacity cathode materials for Li-ion batteries. Only through complementary electron microscopy and spectroscopy techniques at multi-length scale could the structural make-up of Li_1.2(Ni_0.13Mn_0.54Co_0.13)O₂ crystals be described unambiguously. Systematically observing the entire primary particles along multiple zone axes (Fig. 1) reveals that they are consistently made up of a single phase, save for rare localized defects and a thin surface layer on certain crystallographic facets. More specifically, this careful STEM and EELS study at the atomic scale using revealed that the bulk of the oxides consists of randomly stacked domains that correspond to three variants of monoclinic structure. Crucially, it is also shown that only the very surface is composed of a different Co- and/or Ni-rich spinel with antisite defects [2].

Similarly, atomically-resolved monochromated EELS mapping and imaging were used to study the structure and chemistry of promising complex oxide thermoelectrics such as misfit layered cobaltates [3] or Ca-stabilized A-site deficient titanates [4]. For the latter, careful inspection of images and atomically resolved EELS chemical maps in two orthogonal directions suggests that in the Nd_0.6Ca_0.1□_0.3TiO₃ system, Ca predominantly occupies Nd vacancy shared sites, creating locally a higher occupation of the site and thus promotes short range vacancy-cation ordering in both a and b lattice directions. This in turn results in intricately modulated octahedral tilting distortions of the O sublattice, whose signature can be fingerprinted at the atomic scale through the near-edge fine structure of the Ti (and O) electron energy loss edges [5].

[1] O.L. Krivanek et al., Phil. Trans. Roy. Soc. 367 (2009), pp. 3683-3697; T.Miyata et al., Microscopy 5 (2014), pp. 377-382 ; O.L. Krivanek et al., Nature 514 (2014), pp. 209-212.

[2] A.K. Shukla, Q.M. Ramasse, C. Ophus, H. Duncan, F. Hage and G. Chen, Nature Communications 6 (2015), 8711.

[3] J.D. Baran, M. Molinari, N. Kulwongwit, F. Azough, R. Freer, D. Kepapstoglou, Q.M. Ramasse and S.C. Parker, J. Phys. Chem. C 119 (2015), pp. 21818-21827.

[4] F. Azough, D. Kepaptsoglou, Q.M. Ramasse, B. Schaffer and R. Freer, Chemistry of Materials 27 (2015), pp. 497-507.

[5] The SuperSTEM Laboratory is the U.K. National Facility for Aberration-Corrected STEM, supported by the Engineering and Physical Sciences Research Council (EPSRC). 

The aberration-corrected scanning transmission electron microscope (STEM) can provide real space imaging and spectroscopy at atomic resolution with a new level of sensitivity to structure, bonding, elemental valence and even spin state [1]. Coupled with first-principles theory, this represents an unprecedented opportunity to probe the functionality of complex nanoscale systems. Two case studies will be presented utilizing the Nion UltraSTEM: first, determining the origin of the unexpected ferromagnetism in ultrathin, insulating LaCoO_3-x (LCO) films, which is induced by the spontaneous ordering of oxygen vacancies to relieve misfit strain (see Fig. 1) [2]. Second, the origin of the colossal ionic conductivity in Y-stabilized ZrO₂/SrTiO₃ superlattices [3], which is due to spontaneous disordering of the oxygen sublattice in response to biaxial strain coupled with the incompatibility of the  oxygen sublattices at the interfaces (see Fig. 2) [4-6].

A JEOL ARM 200F has recently been installed in the National University of Singapore, equipped with UHR pole piece, ASCOR hexapole aberration corrector, Gatan Quantum ER and OneView camera and Oxford Aztec EDS system. It is installed within a JEOL isolated room with thermal radiation panels, vibration and magnetic field isolation systems. Fig. 3 shows a Ronchigram at 200 kV with near 70 mrad half angle flat phase region, and initial results from Si [110] show good splitting of the dumbbells with information transfer to sub-Angstrom levels, see Fig. 4. Results on oxides will be presented at the meeting [7].

1. J. Gazquez, et al., Nano Lett, 11, 973 (2011).
2. N. Biškup, et al., Phys. Rev. Lett. 112, 087202 (2014).
3. J. Garcia-Barriocanal et al., Science 321, 676 (2008).
4. T. J. Pennycook et al., Phys. Rev. Lett., 104, 115901 (2010).
5. T. J. Pennycook et al., Eur. Phys. J. Appl. Phys. 54, 33507 (2011).
6. Y. Y. Zhang et al., Adv. Mater. Interfaces 1500344 (2015).
7. S. J. Pennycook, et al., ACS Nano, 9, 9437–9440 (2015).

DOE BES Materials Science and Engineering Division, European Research Council Starting Investigator Award STEMOX # 239739, Fundación BBVA, DOE Grant No. DE-FG02- 09ER46554, National University of Singapore and JEOL Asia.

The introduction of two sorts of cations with different valence and size, such as Ba2+ or Sr2+ and Ln3+, in the A-sites of transition metal perovskite oxides has generated numerous remarkable properties such as high Tc superconductivity,  oxygen storage in cobalt based oxides for the realization of solid oxide fuel cell (SOFC) cathodes , CMR in manganates .

The investigation of the system Sm-Ba-Fe-O in air has allowed an oxygen deficient perovskite Sm2-εBa3+εFe5O15-δ (δ=0.75, ε=0.125) to be synthesized. In contrast to the XRPD pattern which gives a cubic symmetry (ap= 3.934Å), the ED patterns of this phase (Fig.1a) show superstructure spots corresponding to c=5ap”. HRTEM study (Fig1b) revealed that this phase is nanoscale ordered with a quintuple tetragonal cell, “ap ´ ap ´ 5ap”. Bearing in mind that one cubic cell corresponds to the formula Sm0.375Ba0.625FeO2.85, these tetragonal nanostructures can be formulated as Sm2-εBa3+εFe5O15-δ (δ~0.75, ε~0.125). They consist of 5 SmO/BaO layers stacked alternately with 5 FeO2 layers along c. The HAADF-STEM image (Fig.1c) of the Sm2-εBa3+εFe5O15-δ structure along the [100] is clearly established from the contrast segregation that the Ba2+ and Sm3+ cations are ordered in (001) layers along the c-axis. One observes rows of bright dots perpendicular to c, which corresponds to three sorts of Sm or Ba cationic layers, judging from their intensity: pure Sm, mixed Ba/Sm and pure Ba layers. Thus the HAADF-STEM image can be interpreted by the following periodic stacking sequence of the A cationic layers along the c axis: “Sm-Ba-Sm/Ba-Sm/Ba-Ba-Sm”.

It appears from the ABF-STEM images along [100] (Fig.1d) and [110] orientations (Fig.1e) of a single Sm2‑εBa3+εFe5O15-δ domain, that the oxygen positions in all the layers are close to the ideal octahedral positions. However, a closer inspection of the images reveals that the oxygen columns in the equatorial positions close to the Sm layer deviate from their ideal octahedral position, and lie closer to the Sm3+ cations yielding a “zigzag” contrast along [100] and [110].

The “Sm-Ba-Sm/Ba-Sm/Ba-Ba-Sm” chemical ordering is also confirmed by elemental EELS mapping (Fig.2a,b). The spatially resolved EELS data show that the O-K edge spectra corresponding to the “FeO2” planes (labeled A,B,C) exhibit different intensity ratios of the two pre-peaks to the O-K edge, prepeak1/ prepeak2 at approximately 529/531 eV, depending on the nature of the surrounding “Sm,Ba” layers (Fig.2c). The O-K fine structure in the Sm plane is very similar to that of plane A, whereas those of the Ba and Ba/Sm planes are similar to B and C planes respectively. A first observation is that pre-peak 1 at ~529 eV is less intense for oxygen anions close to Sm cations (SmO layers as well as A layers). According to the literature, the height of this pre-peak is generally rather independent of the rare earth element and should be around the same height as pre-peak 2. Pre-peak 2, related to Fe3d eg – O2p hybridized states seems invariant in the structure, apart from the c plane where it is slightly subdued, accompanied by an increase of pre-peak 1 below 530eV. Pre-peak 1 can be attributed to a charge transfer from the eg to the t2g band of Fe (the eg band is usually empty for Fe3+). This increase of pre-peak 1 related to Fe3d t2g – O2p hybridized states is also visible in the Ba/Sm mixed layers, and can be linked to the presence of oxygen vacancies in those planes. This peak is stronger in the C plane suggesting the presence of more vacancies in this plane.The spatially resolved EELS spectra of the Fe-L2,3 edge are plotted in Fig. 2c. The Fe L3 and L2 “white lines” arise from transitions of 2p3/2 → 3d3/23d5/2 (L3) and 2p1/2 → 3d3/2 (L2) and are known to be sensitive to valency and coordination. Our data shows that the A and B FeO2 planes exhibit very similar Fe-L2,3 edges, with an L3 peak maximum at 709.5 eV, and a pre-peak to L3, even if faint at 708 eV. The energy position of the L3 maximum, together with the shape and positions of the L3 and L2 are then indicative of Fe3+ in an octahedral coordination. All the acquired Fe L3 edges are significantly broadened with respect to the plotted references for 6-fold, 5-fold and 4-fold coordinated Fe3+. This broadening can be explained by a change in coordination of the Fe atoms. Bearing in mind that the measured oxygen stoichiometry is 14.25, instead of 15, this suggests that the iron coordination is mainly 6, i.e. octahedral, but may also be mixed with the presence of some FeO5 pyramids in those layers.

The nanoscale ordering of this perovskite explains its peculiar magnetic properties on the basis of antiferromagnetic interactions with spin blockade at the boundary between the nanodomains. The variation of electrical conductivity and oxygen content of this oxide versus temperature suggest potential SOFC applications.

Emergent phenomena at complex oxide interfaces continue to attract attention as the basis for a variety of next-generation devices, including photovoltaics and spintronics. Tremendous progress has been made toward understanding the role of interfacial defects, cation intermixing, and film stoichiometry in single heterojunction systems; however, the techniques commonly used to study these interfaces, such as X-ray photoelectron and absorption spectroscopies, are either sensitive only to near-surface regions or do not offer depth resolution to probe individual interfaces. Here we explore the induced polarization in superlattices of LaCrO₃ (LCO) and SrTiO₃ (STO) using a combination of aberration-corrected scanning transmission electron microscopy (STEM) and monochromated electron energy loss spectroscopy (STEM-EELS). We show that a correlative approach, utilizing an array of local and non-local probes, is necessary to fully understand the defect-mediated origin of the induced polarization in this system.

We have conducted detailed structural characterization of several LCO-STO superlattices, as shown in Figure 1. We employ high-angle annular dark field imaging (STEM-HAADF) to directly measure the induced ferroelectric polarization in the STO layers. We first acquire a relatively high-speed time series of multiple fast frames (0.4 µs px^-1), which are then aligned using both rigid and non-rigid registration to remove both sample drift and scan distortion [1]. Using this procedure we directly measure the induced polarization with picometer precision, as we have demonstrated elsewhere [2]. Our results reveal that the built-in asymmetric potential across the LCO / STO interfaces is sufficient to induce a sizable polarization, on the order of 40-70 µC cm^-2, in good agreement with ab initio calculations [3].

We next perform detailed characterization of chemical intermixing and local electronic fine structure changes to explore how defects affect the induced polarization. Figure 2 shows the result of monochromated EELS measurements of the Ti L₂₃ edge fine structure, overlaid onto the integrated Ti L₂₃ edge signal. An improved energy resolution of better than 0.120 eV allows us to observe significant Ti intermixing through the superlattice, as well as subtle fine structure changes in the vicinity of the LCO layers not apparent in earlier data. Mapping the Ti L₃ t_2g – e_g crystal field splitting across the film, we find evidence consistent with a slight reduction in Ti valence from 4+ to 3+ in the vicinity of the LCO layers, possibly the consequence of La³⁺ substitution for Sr²⁺ or oxygen vacancies. Measurements of the Ti L₃ t_2g / e_g ratio also point toward such a trend: moving from the STO toward the LCO layers the ratio begins to decrease within the intermixed region, indicating a redistribution of electrons from t_2g to e_g states, suggesting a reduction in valence. In light of these results, our experimental STEM-HAADF measurements and accompanying ab initio calculations indicate that the induced polarization is robust against even sizable chemical intermixing and defect formation.

References

[1] Jones, L. et al, “Smart Align – a new tool for robust non-rigid registration of scanning microscope data”, Advanced Structural & Chemical Imaging 1:8 (2015).

[2] Spurgeon, S. R. et al, “Polarization screening-induced magnetic phase gradients at complex oxide interfaces.” Nat. Commun. 6, 1–11 (2015).

[3] Comes, R.B. et al, “Interface-induced Polarization in SrTiO₃-LaCrO₃ Superlattices.” Adv. Mater. Int. (2016). DOI: 10.1002/admi.201500779

A–site defficient perovskites are of particular interest for a range of engineering applications, such as ionic conductors, batteries and thermoelectric materials. The attraction of these systems lies in particular in their chemically stability over a wide range of compositions and operating conditions. Moreover, they can exhibit different degrees structural ordering, such as cation-vacancy or oxygen-octahedral-tilt-domain ordering [1,2] which can significantly affect their macroscopic physical properties.

In this work we combine state-of-the-art monochromated core loss electron energy loss spectroscopy (EELS) measurements with advanced image analysis to investigate the interplay between A-site cation-vacancy ordering and oxygen octahedral tilting domains in a series of Nd_2/3xTiO₃ double perovskites, exhibiting different degrees of structural ordering as a function of their heat treatment. [3,4] As a result they exhibit vastly different thermoelectric properties related to the presence of domain boundaries which  can suppress the thermal conductivity due to increased phonon scattering. [5,6]

 High-angle annular-dark-field (HAADF) imaging and large scale atomically-resolved EELS maps were used to investigate the chemical ordering of cations on the so-called A site of the perovskite structure (Figure 1), while pm-precision annular bright field (ABF) imaging of these compounds reveals the presence of TiO₆ tilting domains, which we show can be correlated with variations in the A-site occupancy (Figure 2).   Furthermore, advanced image analysis of the electron micrographs was used to measure local distortions in the TiO₆ lattice

These local distortions are further investigated using atomically-resolved monochromated EELS measurements with an energy resolution better than 0.1eV, in a Nion UltraSTEM 100MC. In particular, measurements of the of the Ti L_2,3 edge (Figure 3), reveal local changes of the near-edge fine structure, usually not discernible with conventional EELS measurements. Changes in the Ti L_2,3 pre-peak intensity, as well as subtle local variations in the Ti L₃ e_g/t_g intensity ratios, clearly observable thanks the relatively high beam current still available at this energy resolution, are indicative of local distortions in the oxygen octahedral sublattice.

References

[1] D.M. Smyth, Annu. Rev. Mater. Sci. 15, 329 (1985).

[2] M. Labeau, I.E. Grey, J.C. Joubert, H. Vincent, and M.A. Alario-Franco, Acta Crystallogr. Sect. A 38, 753 (1982).

[3] S. Jackson, et al, J. Electron. Mater. 43 (2014) p.2331.

[4] F. Azough, et al, Chem. Mater 27 (2015) p.497.

[5] K. Koumoto, et al, Annu. Rev. Mater. Res. 40 (2010) p.363.

[6] D.J. Voneshen, et a,l Nat Mater 12 (2013) p.1028.

[7] SuperSTEM is the U.K’s national facility for Aberration Corrected STEM funded by EPSRC.  

The study of novel physical properties appearing when two materials are interfaced has become one of the major fields of research in solid state physics over the last decade. For example, in the strive for novel non-silicon based electronics, the discovery of the formation of a conductive layer right at the interface region between 2 insulators (for example LaAlO₃ and SrTiO₃); a so-called two-dimensional electron gas (2DEG) or two dimensional electron liquid appears. Another important example of such emergent phenomena is the appearance of interface magnetism or superconductivity. As the materials involved in those new physical phenomena are often complex oxides, many factors such as strain, oxygen stoichiometry, cation intermixing, oxygen octahedral coupling.…[1,2] have to be considered when discussing their origin. The exact understanding of those phenomena is a key factor in order to turn these research ideas into working devices and in order to search for the most optimal materials.

In parallel, advances in electron microscopy instrumentation and techniques such as the introduction of aberration-correctors of the probe-forming lens have made it possible to achieve sub-angstrom spatial resolution allowing the study of materials on an atom column by atom column basis. High Angle Annular Dark Field (HAADF) combined with Annular Bright Field (ABF) imaging have made it possible to study and understand respectively the cationic and the oxygen sub-lattices in these materials.

As a first example, in the case of a (La,Sr)MnO₃ (LSMO) film grown on a NdGaO₃ (NGO) substrate, the magnetic easy axis of the LSMO film can be reversed by adding a SrTiO₃ (STO) buffer layer between the substrate and the film. Using the statistical analysis of STEM ABF images, we could reveal that even one single STO layer is enough to remove the transferred octahedral tilt of NGO into the LSMO film reversing its magnetic easy axis from the a to the b axis(see ref 1 and figure 1).

In a second part, The appearance of superconductivity in a superlattice of (Sr,Ca)CuO₂ (SCCO) and BaCuO₂ (BCO) will be investigated. Comparing two different heterostructures where 8 layers of SCCO are sandwiched with STO (no superconductivity) and BCO (superconductivity) the effect of the structure of the different layers (orientation of the CuO₂ in the form of planes or chains) on the appearance of superconductivity will be presented.

[1] Z. Liao, M. Huijben, Z. Zhong, N. Gauquelin, S. Macke, R. Green, S. van Aert, J. Verbeeck, G. Van Tendeloo, K. Held, G. A. Sawatzky, G. Koster & G. Rijnders, Nat. Mater. 15 (2016), p.425-431

[2] N. Gauquelin, E. Benckiser, M.K. Kinyanjui, M. Wu, Y. Lu, G. Christiani, G. Logvenov, H. –U. Habermeier, U. Kaiser, B. Keimer & G.A. Botton, Physical Review B 90 (2014), article 195140

J.V., G.V.T. and S.V.A. acknowledge funding from FWO projects G.0044.13N and G.0368.15N.  N.G. and J.V. acknowledge funding from the European Research Council under the 7th Framework Program (FP7), ERC Starting Grant 278510 VORTEX. All authors acknowledge financial support from the European Union under the Seventh Framework Program under a contract for an Integrated Infrastructure Initiative (Reference No. 312483-ESTEEM2).

Although defects may be seen as detrimental, they play essential roles in many materials. In some complex oxides for instance, they present an opportunity to enhance particular properties, or even engineer new ones. An archetypal example is the high-Tc superconductor YBa₂Cu₃O_7−δ (YBCO), where defects are indispensable to immobilize quantized vortices in the presence of magnetic fields. Therefore, determining the atomic structure of defects is critical to unravel and control their effect on the physical properties. In this talk, we will do precisely that with an unforeseen complex point-defect that leads to an unexpected formation of ferromagnetic clusters embedded within a superconductor. The commonest YBCO defect comprises the doubling of Cu-O chains between Ba-O planes, a ubiquitous intergrowth regardless of the deposition technique. Here, we will unclothe the true nature of this defect using a combination of experiments and theory to provide a complete picture of the structure and chemistry of these intergrowths at the atomic-scale and their effect on the electronic and magnetic properties of YBCO. First, we will show, by means of scanning transmission electron microscopy (STEM), how the system solves the local off-stoichiometry induced by the extra Cu-O chain, removing half of the Cu atoms in selected chains, and the distortions induced by the vacancies. Secondly, we will show using density functional theory (DFT) how the complex structure of these intergrowths affects the electronic properties, and yields a magnetization density that extends into the neighboring Cu-O planes. Finally, we will present X-ray magnetic circular dichroism (XMCD) spectroscopy results, which provide evidence of the theoretically predicted Cu magnetic moments and the presence of a dilute network of magnetic defects within the high-Tc superconducting state, see Figure 1. See for more details “Emerging dilute ferromagnetism in high-Tc superconductors driven by point defect clusters”, Adv. Sci. 2016, 1500295.

Highly adaptable crystal structures of complex oxide materials enable changes in composition and provide the opportunity of fabricating them in different forms e.g., thin films and/or heterostructures. Complex oxide heterostructures show noteworthy electronic properties which are absent in bulk forms, and the phenomena occurring at their interfaces are intriguing. The interest attributed to these structures and interfaces is also related to the wide variety of the functionalities such as ferroelectricity, magnetism, superconductivity etc. [1]. One of the most exciting interface effects is high temperature interfacial superconductivity discovered in heterostructures consisting of non-superconducting layers of insulating La₂CuO₄ (LCO) and metallic La_1.55Sr_0.45CuO₄ (LSCO). Moreover, cuprate bilayers are also attractive in which the critical temperature can be enhanced remarkably in comparison to single phase systems [2].

Atomic-layer-by-layer oxide molecular-beam epitaxy (ALL-oxide MBE) with its unique capabilities allowed us to synthesize bilayers consisting of 3 unit cells of metallic La_1.6M_0.4CuO₄ layer and 3 unit cells of the undoped insulating (I) LCO layer neither of which is superconducting by its own, where M represents the dopant, namely Sr²⁺, Ca²⁺, and Ba²⁺ which have the same ionic charge but different ionic radii. The bilayers are superconducting with a critical temperature of ~17.0 K, ~37.0 K and ~38.5 K for Ca, Sr and Ba dopants, respectively.

For STEM investigations, a JEOL-ARM 200F STEM equipped with a cold field-emission electron source, a probe C_s corrector (DCOR, CEOS GmbH), a 100 mm² JEOL Centurio SDD-type EDX detector, and a Gatan GIF Quantum ERS spectrometer was used. High-angle annular dark-field (HAADF) imaging and electron energy-loss spectroscopy (EELS) were performed at probe semi-convergence angles of 20 mrad and 28 mrad, respectively. Collection angles for HAADF and annular bright-field (ABF) images were 75-310 mrad and 11-24 mrad, respectively.

Atomically resolved ABF and HAADF images were simultaneously acquired. HAADF images unveil not only the perfect interfaces between substrate and the bilayers but also the high structural quality (Fig 1a-b). Atomic resolution EELS and energy-dispersive x-ray spectroscopy (EDXS) analyses were conducted to examine elemental distributions across the interfaces. EDXS analyses and EEL spectrum images (SI) reveal different cation redistribution behaviors and lengths depending on the cation size mismatch with respect to the La ionic radius. In Fig.1c the distribution of La and Sr across the interfaces of a Sr-doped bilayer sample is presented, which reveals relatively homogeneous distribution of Sr and also the redistribution of Sr near the interface.

Through ABF imaging, oxygen positions were determined, interatomic distortions (cation – cation and CuO₆ octahedra) were measured and correlated with the dopant distributions obtained via chemical analyses. In Fig. 2a two of the quantitatively studied images (i.e. HAADF and ABF), are presented as an overlay and the arrows indicate the nominal interfaces. Figure 2b shows the measured interatomic distances which are indicated with arrows. Results for the different dopants will be presented and correlations with high-temperature interfacial superconductivity will be discussed [3].

References

[1] H. Y. Hwang et al., Nature Materials, 11 (2012) p.103.

[2] A. Gozar et al., Nature, 455 (2008) p.782.

[3] The research leading to these results has received funding from the European Union Seventh Framework Programme [FP7/2007-2013] under grant agreement n°312483 (ESTEEM2).

Carbon in the sp¹ hybridization (carbyne) is able to form atomic chains that constitute the elementary one-dimensional phase of carbon [1]. Despite many efforts, the synthesis of carbon chains in appreciable quantities remains difficult and even the existence of bulk carbyne is subject of an ongoing controversy. However, the existence of individual carbon chains is undisputed since they have been observed in in-situ TEM studies. The present work allowed, for the first time, to measure the electrical properties of carbon chains by using a dedicated specimen stage for establishing electrical contacts in the TEM.

Carbon chains are unusual conductors that may occur either as metallic cumulene with double bonds or as semiconducting polyyne with alternating single and triple bonds. Now, as it became possible to characterize them electrically, these two electronic configurations can be distinguished. A piezo-driven tip (Nanofactory), integrated into the specimen holder of a TEM, allowed to establish contacts to graphenic material and, by controlled retraction of the electrodes, to unravel chains of carbon atoms (fig. 1a). At the same time, the electrical properties could be measured [2, 3]. By recording current-voltage curves of individual carbon chains, both cumulene and polyyne were identified (fig. 1b). It was found experimentally and by quantum conductance calculations (fig. 1c) that transport through narrow resonant states makes the conductivity much lower than predicted in previous theoretical work. At high applied bias, however, a sudden rise in current occurs, showing the absence of conduction channels at lower energy and their presence at higher energy.

When the 1D system is under strain, the chains exhibit a semiconducting behavior, corresponding to polyyne. Conversely, when the chain is unstrained, an ohmic behavior, corresponding to cumulene, is observed (fig. 1b) [4]. This confirms a recent theoretical prediction, namely that the Peierls distortion, which would stabilize polyyne, is suppressed by zero-point vibrations in an unstrained chain so that cumulene is the stable configuration. In the presence of strain, however, polyyne is favoured by the Peierls instability. Thus, a metal-insulator transition can be induced by adjusting the strain. Furthermore, it is shown that these atomic chains can act as rectifying diodes when they are in a non-symmetric contact configuration, i.e., between a carbon and a metal contact or between two carbon contacts of different type.

Acknowledgements:

Financial support by the LABEX program "Nanostructures in Interaction with their Environment" (NIE) and different projects of the Agence Nationale de Recherche (ANR, France) is gratefully acknowledged.

 References:

1. F. Banhart, Beilstein J. Nanotech. 6, 559 (2015).

2. O. Cretu, A. R. Botello-Mendez, I. Janowska, C. Pham-Huu, J.-C. Charlier, F. Banhart, Nano Lett. 13, 3487 (2013).

3. A. La Torre, F. Ben Romdhane, W. Baaziz, I. Janowska, C. Pham-Huu, S. Begin-Colin, G. Pourroy, F. Banhart, Carbon 77, 906 (2014).

4. A. La Torre, A. Botello-Mendez, W. Baaziz, J.-C. Charlier, F. Banhart, Nature Comm. 6, 6636 (2015).

The characterization of low-dimensional materials such as graphene or transition metal dichalogenides (TMDCs) often requires more than one technique to obtain a thorough understanding of their attributes for specific applications. Graphene and TMDCs both have layered structures and properties that vary significantly with thickness as compared to single layer conformations, making them very interesting for electronics design [1, 2]. Electronic device performance optimization can benefit greatly from knowledge of their crystalline structure and exciton dynamics. The aim of the following analysis is to show how several spectroscopy (Raman/Photoluminescence) and microscopy techniques (AFM/SEM) can, in correlation, provide a more detailed depiction of low-dimensional materials than the constituent measurements could offer in isolation.

Raman spectroscopy, and Raman imaging in particular, has proved to be of great value in differentiating spectra obtained from single, double and multi-layered low-dimensional materials. Raman imaging was also used to evaluate strain, doping, chirality and disorder in graphene and TMDCs [3, 4]. The information acquired with Raman spectroscopy and imaging can be complemented by data from highest resolution microscopy techniques such as Atomic Force Microscopy (AFM), Scanning Near-Field Microscopy (SNOM), or Scanning Electron Microscopy (SEM).

Fig. 1a shows a correlative Raman-SEM image of exfoliated graphene deposited on a non-conductive glass coverslip. The SEM image was acquired in the low vacuum mode using a BSE detector. Different graphene sheets can be clearly visualized. In the center of the SEM image a confocal Raman image was acquired, showing the presence of different number of graphene layers, from one (green) to ten (yellow). The high resolution confocal Raman image (Fig. 1b), acquired from the area marked in Fig. 1a shows the intensity distribution of the G band as a function of layers in the graphene sheet as indicated with 1, 2 and M (multy-) layers. By evaluating the intensity of the D-band, the chirality of the graphene crystal could be determined as highlighted in red color in Fig. 1b. The SNOM image from the same area (Fig. 1c) shows that the transparency of the graphene decreases with increase of number of layers providing direct access to the fine structure of graphene.

In a second example a high resolution SEM image of CVD grown MoS2 is pesented (Fig. 2a). The Raman-SEM image (RISE image)  presented in Fig. 2b shows a correlation of changes in the Raman spectrum at the defects of the crystals.

References:

[1] A. K. Geim , I. V. Grigorieva, Nature 499 (2013), p. 419.

[2] F. H. L. Koppens, T. Nueller, P. Avouris, A.C. Ferrari, M. S. Vitiello, M. Polini, M. Nat. Nano  9 (2014) p. 780.

[3] Y. You, Z. Ni, T.Yu, Z. Shen, Appl. Phys. Letters 39 (2008) p.13112.

[4]  P.K. Chow, R.J. Gedrim, J. Gao, T. M. Lu, B. Yu, H. Terrones, N. Koratka, ACS Nano (in press)

The tunable electrical properties of reduced graphene oxide (RGO) make it an ideal candidate for many applications including energy storage.¹ However, in order to utilize the material in industrially applied systems it is essential to understand the behavior of the material on the nanoscale, especially how naturally occurring phenomenon like wrinkling effects the electronic transport. While there have been many theoretical investigations on the transport behavior, there are much fewer experimental measurements.^{2, 3, 4} Here we used a transmission electron microscope (TEM) with scanning tunneling microscope (STM) probe in-situ holder is used to perform localized electrical measurements on wrinkled, supported RGO flakes. The RGO flakes are deposited onto pre-patterned Au electrodes on SiN membranes. The flakes are 1-3 layers thick and have lateral dimensions on the order of single micrometers. They are deposited through an ammonium laurate (AL) aqueous surfactant. The TEM allows for observation of the local wrinkling structure of the RGO and simultaneously a Nanofactory STM-TEM holder is used to perform localized resistance measurements. The holder allows for manipulation with and positioning of Au STM probes, with diameters less than 100 nm, with sub-nanometer precision. The results show that the overall resistance is low, on the order of single kΩ, if the contact between the probe and the RGO is optimized.  Wrinkles reduce the contact resistance between RGO and the probe. The overall trend for more than 70 measurements is that the total resistance decreases with increasing amount of wrinkles and that the contact resistance dominates the resistance when the probe makes contact with wrinkle free RGO. The sheet resistance and maximal contact resistance for our measurements range from 5-20 kΩ/☐ and 6-20E-7Ω cm², respectively, which falls in line with other scanning probe measurements for graphene and metal contacts.^{5, 6}  Here we use a bottom metal contact for the gold electrodes which eliminates the risk of resist contamination and metal doping of the graphene.⁷ These measurements give evidence that RGO with large amounts of wrinkling on a substrate can transport electrons well and that it is thus not necessary to produce suspended, pristine graphene in order to benefit from its electrical transport properties.

(1) Novoselov, K. S.; Fal'ko, V. I.; Colombo, L.; Gellert, P. R.; Schwab, M. G.; Kim, K. A roadmap for graphene. Nature 2012, 490, 192-200.

(2) Guinea, F.; Horovitz, B.; Le Doussal, P. Gauge fields, ripples and wrinkles in graphene layers. Solid State Communications 2009, 149, 1140-1143.

(3) Guo, Y.; Guo, W. Electronic and Field Emission Properties of Wrinkled Graphene. Journal of Physical Chemistry C 2013, 117, 692-696.

(4) Ladak, S.; Ball, J. M.; Moseley, D.; Eda, G.; Branford, W. R.; Chhowalla, M.; Anthopoulos, T. D.; Cohen, L. F. Observation of wrinkle induced potential drops in biased chemically derived graphene thin film networks. Carbon 2013, 64, 35-44.

(5) Yan, L.; Punckt, C.; Aksay, I. A.; Mertin, W.; Bacher, G. Local Voltage Drop in a Single Functionalized Graphene Sheet Characterized by Kelvin Probe Force Microscopy. Nano Letters 2011, 11, 3543-3549.

(6) Robinson, J. A.; LaBella, M.; Zhu, M.; Hollander, M.; Kasarda, R.; Hughes, Z.; Trumbull, K.; Cavalero, R.; Snyder, D. Contacting graphene. Applied Physics Letters 2011, 98.

(7) Babichev, A. V.; Gasumyants, V. E.; Egorov, A. Y.; Vitusevich, S.; Tchernycheva, M. Contact properties to CVD-graphene on GaAs substrates for optoelectronic applications. Nanotechnology 2014, 25.

We recently demonstrated a controllable and reproducible method to obtain suspended monolayer graphene nanoribbons with atomically defined edge shape [1]. Our method exploits the electron-beam of a Scanning Transmission Electron Microscope (accelerated at 300 kV) to create vacancies in the lattice by knock-on damage and pattern graphene in any designed shape. The small beam spot size (0.1 nm) enables close-to-atomic cutting precision, while heating graphene at 900 K during the patterning process avoids formation of beam-induced Carbon deposition and allows self-repair of the graphene lattice.  Self-repair mechanism is essential to obtain well-defined (zig-zag or armchair) edge shape and, if the electron beam dose is lowered, to perform non-destructive imaging of the graphene nanoribbons.

Drawing the electron-beam path with a software script, we were able to obtain reproducible graphene nanoribbons with sub 10 nm width. Using an in-house built microscopy holder equipped with electrical feedthroughs, we performed 2 and 4 wire measurements on several graphene nanoribbons, with different number of layers. Wide nanoribbons (width> 50 nm) exhibited ohmic behaviour, with conductivity linearly proportional to the width. Narrower ribbons (10 nm < width < 50 nm) displayed non-linear current-voltage (IV) relationship at low temperature (4 K, ex-situ), possibly indicating the opening of a band gap. The narrowest ribbon was 1.5 nm wide, with strong non-linear IV also at room temperature.

Electrical measurements were also performed in the high temperature range (300 K – 900 K), from which we concluded that thermal generated carriers give the main contribution to  electrical conductivity above room temperature.

Ackwnoledgements: This work was supported by ERC funding,  project 267922 - NEMinTEM 

References

[1]        Q.Xu, M. Wu, G. F. Schneider, L. Houben, S.K. Malladi, C. Dekker, E. Yucelen, R.E. Dunin-Borkowski, and H.W. Zandbergen, ACS Nano  7 (2), 1566-1572, 2013

        In this work, we perform an in-depth study of the cyclic telescoping of multi-walled carbon nanotubes (MWCNTs) inside a transmission electron microscope. The nanotubes are observed in real time, while simultaneously measuring their electrical properties. To date, examples of in situ TEM nanotube telescoping have remained few in number due to the highly specialized equipment and technical expertise required.

        Experiments were conducted by first ‘sharpening’ a MWCNT by approaching it with a piezo-controlled tungsten tip and applying a voltage pulse. This results in controlled and localized peeling of the outer walls, giving rise to a telescope structure. The exposed core is then contacted with the tungsten tip and reproducibly pulled out then re-inserted, while measuring conductance behavior at each movement step. Figure 1 shows an example of in situ telescoping of a MWCNT in real time.

        Results demonstrate that an electrical current can be maintained for multiple in-and-out cycles (up to 11 cycles have been tested to date), and for telescoping distances of up to 650 nm. This is exemplified in Figure 2. The systems display complex conductance behavior, with of some MWCNT telescopes demonstrating improved conduction (and current) as a function of the number of cycles, while for most these properties diminish.

        Control experiments have been conducted to investigate the effect of the electron beam on the behavior of the MWCNTs, and have indicated that changes in conductance behavior are not strongly influenced by the imaging process under normal beam conditions.

        One potential application for these structures is use as flexible electrical contacts, which allow for relative motion of the two electrodes, while maintaining electrical conduction.

        The authors would like to acknowledge funding as provided by the National Institute for Materials Science (NIMS), through the International Center for Materials Nanoarchitectonics (MANA).

Controlled manipulation of materials on the atomic scale is a continuing challenge in physics and material science. Doping of two-dimensional (2D) materials via low energy ion implantation could open possibilities for fabrication of nanometre-scale patterned devices, as well as for functionalization compatible with large-scale integrated semiconductor technology. High resolution imaging and spectroscopic analysis of these electronic dopants at the atomic scale is fundamental for understanding sites, retention, bonding and resulting effects on Fermi level shifts. This has been made achievable with the advent of aberration corrected transmission electron microscopy (TEM) and scanning TEM, as well as low-loss electron energy loss spectroscopy (EELS).

We show for the first time directly that 2Ds can be doped via ion implantation. Retention is in good agreement with predictions from calculation-based literature values [1], as are initial results of the sites of dopants and their influence on the band structure of surrounding atoms. We present results of N- and B-implantation in graphene [2,3] as well as current progress with ion-implantation in TMDCs.

1. Ahlgren, E. H.; Kotakovski, J.; Krasheninnikov, A. V. Phys. Rev. B, 2011, 83, 115424
2. U. Bangert, W. Pierce D. M. Kepaptsoglou, Q. Ramasse, R. Zan, M. H. Gass, J. A. Van den Berg, C. B. Boothroyd, J. Amani and H. Hofsäss, Nano Letters, 2013, 13, 4902-4907
3. Demie Kepaptsoglou, Trevor P. Hardcastle, Che R. Seabourne, Ursel Bangert, Recep Zan, Julian Alexander Amani, Hans Hofsaess, Rebecca J. Nicholls, Rik M. D. Brydson,  Andrew J. Scott, and Quentin M. Ramasse, A.C.S., 2015, 9, 11398–11407

Charge density waves (CDW) are periodic modulations of charge density in low-dimensional metals observed as a function of temperature, doping and pressure. Due to electron-phonon coupling CDWs, are also accompanied by a periodic lattice distortion (PLD).¹ Quasi two-dimensional (2D) transition metal dichalcogenides metals like 1T/2H-TaSe₂, 1T/2H-TaS₂, 2H-NbSe₂ are some of the materials that exhibit strong CDW distortions. While CDW can be directly probed using Scanning Tunnelling Microscopy (STM), diffraction and imaging techniques such as High Resolution Transmission electron Microscopy (HRTEM) and selected area electron diffraction (SAED) are sensitive to the structural distortions (PLD) accompanying CDW. In an electric field, and in the absence of crystalline impurities and defects, sliding incommensurate CDW show transport properties similar to a superconducting state.^1,2 It is therefore important to understand the effects of commensuration, defects, and impurities on the static and dynamic properties of the CDW state.^2,3 On a transmission electron microscope (TEM), investigating the effects of  defects on the CDW/PLD state can be done in-situ by generating point defects through electron beam irradiation and at the same time monitoring the structure of the CDW/PLD through electron diffraction and atomic resolved HRTEM imaging. 

Here we report on the interaction of commensurate CDW/PLD with point defects generated by the electron beam in 1T-TaSe₂ and 1T-TaS₂. Due to the CDW/PLD, bulk 1T-TaS₂ and 1T-TaSe₂ are characterized by a commensurate √13 a₀ ×√13 a₀ (a₀= 3.447) superlattice at 180 K and 300 K respectively.¹ Figure 1(a) displays a HRTEM image of 1T-TaSe₂ obtained at 100 K. The HRTEM image was obtained at region consisting of upto 10-15 layers of 1T-TaSe₂ and shows Ta atomic columns. The HRTEM image is also characterized by a bright, dark contrast modulation due to the PLD/CDW. A Fast Fourier Transform (FFT) of the HRTEM image shown in fig 1(b) has spots from the main structure (solid red circles) surrounded by six satellite spots (dotted circle) from the PLD/CDW. HRTEM image analysis involving masking of the satellite spots from the CDW/PLD with a circular mask, followed by inverse FFT helps to visualize the structure of the lattice arising from the commensurate PLD/CDW. The resulting image, shown in fig. 1(c), then displays an intensity image showing a √13 a₀ ×√13 a₀  lattice of the CDW/PLD maxima.

To investigate the effects of electron-beam generated defects on the CDW/PLD, succesive HRTEM images  were obtained from the same sample region over time and at a constant electron dose. Images showing the evolution of PLD/CDW maxima with electron beam irradiation were then obtained from respective HRTEM images using the image analysis procedure elucidated above. Figures 1(d)-(f) show successive HRTEM images obtained over time. The HRTEM images were obtained from a region of upto 15 layers of 1T-TaSe₂ and only show the Ta atomic columns. The PLD/CDW lattices shown in fig.1(g), 1(h), 1(i) correspond to the HRTEM images in figs. 1(d), 1(e), and 1(f) respectively. A radial distribution function (RDF) showing the nearest-neighbor and next-nearest-neighbor periodicities between the CDW maxima can also be calculated. The RDF in figs.1(j), 1(k), 1(l) have been calculated from the CDW/PLD lattices in figs. (g), (h), (i) respectively. The peak with the asterix shows the main periodicity due to the CDW/PLD wave-vector |q_{i =1…6}| =  √13 a₀. Loss of long-range order with increased exposure to the electron beam is characterized by broadening of RDF peaks (see regions marked with dotted rectangle in fig. 1(k) and fig.1 (l)). Based on the characterization of 1T-TaSe₂, and 1T-TaS₂ thin/single layers we show that this loss of long range order is due to interaction of CDW/PLD with S and Se anionic point defects generated by the electron beam.

References and acknowledgments:

^1. J. Wilson, F. D. Salvo, and S. Mahajan, Adv. Phys. 24, 117 (1975).

^2.H. Dai, and C. Lieber, J. Phys. Chem. 97, 2362 (1993)

^3. H. Mutka, Phase Transitions 11, 221 (1988).

Quasi two-dimensional (2D) semiconductor materials are desirable for electronic, photonic, and energy conversion applications and have stimulated extensive research on their synthesis^1-3 and applications^4-6.  One such example is the growth of indium phosphide flag-like nanostructures (Fig. 1) by epitaxial growth on a nanowire template⁷. By the intentional incorporation of defects, control of the nano-flag location along the nanowire was achieved. Long InP nanowires with inherent high defect density in the top area of the flagpole, which pins the catalyst, resulted in top flag geometry (Fig. 1a). Intentional defect creation in the middle section of the flagpole produced middle nanoflags (Fig. 1b). Bottom flags were created by growing defect free primary nanowires under high phosphine flow (Fig. 1c).

Detailed investigation, using aberration corrected high resolution TEM (FEI Titan 80-300), high angle annular dark field STEM and electron diffraction, of the flag-like nano-structures at different stages of the growth provided valuable details about the growth mechanism and the resulting morphology of such nano-structures.

The main conclusion from the TEM observations is that the growth process involves a few stages. The first stage consists of asymmetrical dissolution of the nanowire (NW) tip into the catalyst (Fig. 2a), followed by the Au catalyst unpinning from the top of the nanowire, and its induced migration along the nanowire sidewall. The final position of the catalyst is determined by the position of structural defects (such as stacking faults - Fig. 2b) along the NW (i.e. the "flagpole"). The next stage involves the epitaxial growth of a pure Wurtzite nano-flag on one of the facets of the "flagpole" resulting in a two dimensional nano-membrane on a one dimensional NW sidewall template (Fig. 2c-d).

The comprehensive understanding of the growth process may help improve the synthesis of similar nano-structures. It may allow for better control on the growth process of more complex structures and also help achieving similar nanostructures in a variety of III-V semiconductor material systems with potential applications such as active nanophotonics.

References:

1. Gudiksen, M. S.; Lauhon, L. J.; Wang, J.; Smith, D. C.; Lieber, C. M. Nature 2002, 415 (February), 617–620.
2. Algra, R. E.; Verheijen, M. a; Borgström, M. T.; Feiner, L.-F.; Immink, G.; van Enckevort, W. J. P.; Vlieg, E.; Bakkers, E. P. a M. Nature 2008, 456 (7220), 369–372.
3. Gorji Ghalamestani, S.; Ek, M.; Ganjipour, B.; Thelander, C.; Johansson, J.; Caroff, P.; Dick, K. a. Nano Letters 2012, 12 (111), 4914–4919.
4. Tomioka, K.; Tanaka, T.; Hara, S.; Hiruma, K.; Fukui, T. IEEE J. Sel. Top. Quantum Electron. 2011, 17 (4), 1112–1129.
5. Nakayama, Y.; Pauzauskie, P. J.; Radenovic, A.; Onorato, R. M.; Saykally, R. J.; Liphardt, J.; Yang, P. Nature 2007, 447 (June), 1098–1101.
6. Mourik, V.; Zuo, K.; Frolov, S. M.; Plissard, S. R.; Bakkers, E. P. a. M.; Kouwenhoven, L. P. Science (80-. ). 2012, 336, 1003.
7. A. Kelrich, O. Sorias, Y. Calahorra, Y. Kauffmann, R. Gladstone, S. Cohen, M. Orenstein and D. Ritter, Nano Letters 2016, 10.1021/acs.nanolett.6b00648.

Metallic nanoparticles occupy special places in fundamental research and practical applications, not only because they bridge the gap between atoms and bulk matter but more importantly because some fascinating phenomena appear only at nanoscale level. Research has shown that the properties of nanoparticles are strongly affected by their size, shape, chemical composition as well as their supports.  Hence it is essential to gain full knowledge of their three-dimensional atomic structures, which is still a challenging task for nanoclusters and nanoalloys. 

After a brief overview of the current status of structure research on clusters and nanoalloys, I will discuss what an aberration-corrected scanning transmission electron microscopy (STEM) can offer in this area through an atom counting approach. I will then present two case studies from our recent work: (1) on control and manipulation of the size and atomic structure of Au clusters, formed via gas phase condensation and deposited on amorphous carbon films and MgO(100) surfaces; and (2) on structures of TiO₂(110) supported Au-Rh nanoalloys, prepared through solution based chemical synthesis method, under effects of various thermal treatments.  In both cases, direct evidence of atomistic details of the clusters, nanoalloys as well as nanoparticle-support interfaces will be presented and their possible mechanisms will be discussed, together with computer modeling studies. Our emphasis will be on understanding the relationship between nanoparticle structures and their thermodynamics or kinetics.

Nanoparticle and catalysis research makes extensive use of transmission electron microscopy (TEM). In particular, environmental transmission electron microscopy (ETEM) has attracted considerable attention in recent years. This technique allows us to expose samples to gaseous environments at elevated temperatures in order to investigate local structural changes at the atomic level as the environment changes. Recently, the technique has also been used in nanowire, graphene and electron optical lithography research among others.

Recent developments in TEM instrumentation include monochromation of the electron source and aberration correction of both condenser and objective lenses. These developments have now been introduced onto the ETEM column. The improved spatial resolution and interpretability provided by these additions are beneficial for imaging the surface structure and dynamics of catalyst nanoparticles and provides exciting new possibilities for investigating chemical reactions. In order to take full advantage of this, an understanding of both the interaction of fast electrons with gas molecules and the effect of the presence of gas on high-resolution imaging is necessary.

Using an FEI Titan ETEM equipped with a monochromator and an aberration corrector on the objective lens, we have investigated sintering of supported metal nanoparticles often used in catalysis. A model system consisting of supported gold nanoparticles were prepared by sputter-depositing the metal onto graphene and boron nitride substrates. These samples were imaged under hydrogen at increasing temperatures. Gas was introduced into the environmental cell using digitally controlled mass flow controllers providing accurate and stable control of the pressure in the cell. As the temperature was increased, migrating particles were observed on the support. As they came into contact, a neck was formed between the particles and subsequently, the particles coalesced entirely (Fig. 1). Growth patterns have also been investigated for platinum and palladium nanoparticles supported on silicon oxide substrates. Here, anomalously large particles were observed as the particles were sintered in oxygen atmospheres at temperatures at elevated temparatures. Such large particles have also been observed for industrial catalysts. In this study, we will try to elucidate the mechanisms of metal nanoparticle sintering.

The analytical capabilities of the microscope can be further augmented by adding stimuli such as optical or electrical though the sample holder. Using a holder capable of exposing the sample to light, the redox properties of cuprous oxide have been investigated. Cuprous oxide has been identified as an active catalyst for the water splitting and hydrogen evolution from an ethanol solution. However, Cu2O suffers from photocorrosion. This phenomenon was investigated using controlled atmosphere transmission electron microscopy. Fig. 2 shows how the photoinduced degradation of cuprous oxide to metallic copper under an aqueous atmosphere using bright-field imaging, electron diffraction and electron energy-loss spectroscopy. All three techniques show the transformation from oxide to metal.

Effects of imaging in various elemental as well as di-molecular gases and their effect on imaging and spectroscopy in the environmental transmission electron microscope will also be discussed.

Thin crystalline germanium (c-Ge) and silicon (c-Si) films are widely used in current microelectronic and nanoelectronic devices, such as thin film transistors and highly efficient solar cells. However, Si or Ge films grown from the vapor phase are usually amorphous. Crystallization of such films requires annealing temperatures above 500°C, which are often incompatible with the fabrication processes used in industry. Yet, when a semiconductor film of amorphous germanium (a-Ge) or amorphous silicon (a-Si) is in contact with a metal that forms an eutectic phase diagram (e.g, Au, Ag, Al, Bi, Pd), the crystallization temperature of the amorphous film is reduced significantly. This effect is commonly known as “metal-mediated” or “metal-induced crystallization” (MIC). Despite numerous studies of the MIC effect, the reaction mechanism is still unclear, although the interaction at the interface between metal and semiconductor seems to play a key role in activating the process. In this context, the main motivation of the current work is to understand the mechanism of metal-induced crystallization of semiconductor films in eutectic binaries by studying the interfacial interactions in these systems. Layered films formed by the sequential condensation of components in vacuum are a convenient model system, since the thickness of the film or the size of particles is comparable with the width of the interface boundary in a binary system. By heating such a system in a transmission electron microscope,  the interaction between the metal and the semiconductor film can be observed directly in real time.

We have selected gold-germanium (Au-Ge) films as a model system. Layered Au/Ge films are prepared by the condensation of an amorphous continuous film of Ge of 5 nm, followed by the deposition of a 0.2-0.3 nm Au film on top of the Ge film. KCl crystals covered by an arc-deposited 5 nm carbon layer prior to the deposition of Ge and Au are used as a substrate. The structure and morphology of the films were studied in situ in a FEI TECNAI G2 F20 X-TWIN transmission electron microscope within the 20-400°C range with a Gatan 652 Double Tilt Heating Holder. It has been shown that the process of solid state de-wetting of a 0.2-0.3 nm Au films occurs in the 150-200°C temperature range and results in an a-Ge film uniformly covered with Au nanoparticles with an equivalent diameter of 4.4 nm. The eutectic melting of these Au nanoparticles occurs at temperatures 200-250°C and is accompanied by both an abrupt change of the film morphology and crystallization of the a-Ge film. Long-term annealing of the film at a temperature below the eutectic point does not lead to a change in the mean particle size or to the appearance of c-Ge. Instead, the capillary motion of liquid eutectic particles on Ge surface has been observed (Fig. 1). At higher temperatures, this process intensifies and the motion of the particles becomes visible directly by TEM. This effect leads to the crystallization of a whole surface of amorphous germanium film and causes a drastic change in the Au film morphology. In addition, the formation of metastable fcc Ge phase is found and is shown to drive the capillary motion of eutectic particles. Fig. 2 shows TEM images of the same area of the film taken at 370°C with an interval of 10 seconds between each image. From Fig 2a to Fig 2b, it is evident that the eutectic nanoparticles move, leaving c-Ge behind. However, a more detailed examination of the HRTEM images of the crystalline Ge film shown in Figure 2 exhibit an unexpected feature. In particular, the Fast Fourier Transform (FFT) of region I in Figure 2a, which corresponds to a beam direction of B=[110], shows forbidden reflections for the diamond structure, namely {200} type planes (d=0.283 nm). The appearance of such reflections seems to indicate the crystallization of germanium into a metastable fcc phase, which is then converted into the diamond structure.

In the last decades, radiolytic synthesis routes have exploited the chemical effects of the absorption of high-energy radiation on precursor solutions, to form nanostructures by reproducing a selective reducing/oxidizing environment. Radiation chemical synthesis provides a powerful means to form nuclei which are homogeneously distributed in the whole volume and where the growth rate can be easily controlled.[1] The latter has recently been achieved using liquid cell electron microscopy, with examples of formation kinetics of particles in polar (water)[2] and non-polar (toluene)[3] solvents following a linear relation with applied dose rate. Fine control of particles size with conventional radiation sources typically requires increasing the production of radicals (thus applying “high doses”). On the contrary, the challenge in the scanning transmission electron microscope (STEM), where incident doses are inherently higher by orders of magnitude, is to reduce drastically such production. To illustrate this, we recently proposed the use of non-polar systems (toluene), which are not typically used for radiolytic synthesis outside the STEM, as a solvent that produces much lower amount of radiolytic ionic species upon electron irradiation, as compared to water or other polar solvents such as alcohols (Figure 1 shows an example of

Thus far, most experiments in the liquid cell involved the use of water as a solvent, which explains the large amount of work dedicated to understanding the effects of the electron beam in aqueous conditions. Radiation chemical yields in water are large due to the relatively low bond energies in water molecules, meaning highly reactive oxidizing and reducing radicals and species are created in about an equal amount.[4] Reproducing net reducing conditions for nanoparticle growth can be achieved with the addition of substances that convert primary radicals into free reducing radicals (using OH· scavengers, for instance). The use of organic solvents in the liquid cell allows for tuning the polarity of the medium, giving access to a broader range of synthesis conditions but producing more complex radiolytic products. Here I will discuss, revisiting a number of examples from the literature and presenting our most recent work, general methods for finding more suitable synthesis environments for controlled nanoparticles formation in the liquid cell. Much of the presentation will focus on the solvent radiolysis which is what predominantly dictates the species and yields involved in the chemical processes leading to nanostructure synthesis.[4]

References:

[1] J. Belloni, Catal. Today, 2006, 113, 141-156; S.-H. Choi et al., Colloids Surf., A, 2005, 256, 165-170; J. Belloni et al., New J. Chem., 1998, 22, 1239-1255; M. A. J. Rodgers and Farhataziz, Radiation Chemistry: Principles and Applications, VCH Publishers, New York, N.Y. , 1987.

[2] D. Alloyeau et al., Nano Lett., 2015, 15, 2574-2581; J. H. Park et al., Nano Lett., 2015, 15, 5314-5320.

[3] P. Abellan et al., Langmuir, 2016, 32, 1468-1477.

[4] SuperSTEM is the UK EPSRC National Facility for Aberration-Corrected STEM, supported by the Engineering and Physical Science Research Council (PA). Part of this work was supported by the Chemical Imaging Initiative, under the Laboratory-Directed Research and Development Program at Pacific Northwest National Laboratory (PNNL) and by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, and was performed in part using the facilities located in the William R. Wiley Environmental Molecular Sciences Laboratory, a DOE national scientific user facility sponsored by the DOE’s Office of Biological and Environmental Research and located and the PNNL. PNNL is operated by Battelle for DOE.

Nano assemblies are two- or three-dimensional (3D) collections of nanoparticles. The properties of the assemblies are determined by the number of particles, their position, shape and chemical nature as well as the bonding between them. If we are able to determine these parameters in 3D, we will be able to provide the necessary input for predicting the properties and we can guide the synthesis and development of new assemblies or superstructures. A detailed structural characterization is therefore of utmost importance.

Electron tomography is a versatile and powerful tool that has been increasingly used in the field of materials science. Also for the investigation of nanoassemblies, electron tomography has been of high value as illustrated in Figure 1 [1-3]. However, to extract quantitative information, optimization of the electron tomography experiment is required, especially for large assemblies or assemblies consisting of more than one type of nanoparticle. Here, we will show how advanced electron microscopy,  in combination with novel reconstruction algorithms, enables us to determine the number of particles, their stacking, interparticle distances and outer morphology. This will be illustrated for very large assemblies consisting of spherical nanoparticles [4], assemblies that contain anisotropic particles (Figure 2) and binary assemblies (Figure 3) consisting of particles with different sizes or different compositions [5,6].

Going a step further is the investigation of self assembly and oriented attachment at the atomic scale. Using annular bright field scanning transmission electron microscopy, the possible effect of polarity can be investigated, whereas high angle annular dark field scanning transmission electron microscopy in combination with advanced computational techniques enables the investigation of the interfacial planes in two-dimensional superlattices from nanocrystals [7]. Finally, we will discuss the use of exit wave reconstructions to investigate the effect of surface ligands on self assembly.

[1] S. Bals, B. Goris, L.M. Liz-Marzán, G. Van Tendeloo, Angewandte Chemie 53 (2014) 10600

[2] T. Altantzis, B. Goris, A. Sánchez-Iglesias, M. Grzelczak, L.M. Liz-Marzán, S. Bals, Particle and Particle Systems Characterization 30 (2013) 84

[3] M.P. Boneschanscher, W. Evers, J.J.Geuchies, T. Altantzis, B. Goris, F.T. Rabouw, S.A.P. van Rossum, H.S.J. van der Zant, L.D.A. Siebbeles, G. Van Tendeloo,I. Swart, J. Hilhorst, A. Pethukov, S. Bals, D. Vanmaekelbergh, Science 344 (2014) 1377

[4] D. Zanaga, F. Bleichrodt, T. Altantzis, N. Winckelmans, W.J. Palenstijn, J. Sijbers, B. de Nijs, M.A. van Huis, A. Sánchez-Iglesias, L.M. Liz-Marzán, A. van Blaaderen, K.J. Batenburg, S. Bals, G. Van Tendeloo, Nanoscale 8 (2016) 292

[5] M. Grzelczak, A. Sanchez-Iglesias, L. M. Liz-Marzán, Soft Matter 9 (2013) 9094-9098

[6] T. Altantzis, Z. Yang, S. Bals, G. Van Tendeloo, M.-P. Pileni, Chemistry of Materials 28 (2016) 716

[7] E. Javon, M. Gaceur, W. Dachraoui, O. Margeat, J. Ackermann, M. Ilenia Saba, P. Delugas, A. Mattoni, S.Bals, G. Van Tendeloo, ACS Nano 9 (2015) 3685

Acknowledgements

S.B., D.Z. and N.C. acknowledge financial support from European Research Council (ERC Starting Grant # 335078-COLOURATOMS). The research leading to these results has received funding from the European Union Seventh Framework Programme under Grant Agreements 312483 (ESTEEM2). 

Understanding processes leading to the formation of nanoparticles is of utmost importance for tailoring their size, shape and spatial arrangements, which are crucial for the nanoparticles’ use in heterogeneous catalysis. As the nanoparticle growth involve processes at or across atomic surfaces, observation made in situ at high-spatial resolution would be beneficial for elucidating the nanoparticle growth mechanism. In recent years, transmission electron microscopy (TEM) has become a powerful technique for studying nanoparticles. With TEM, individual nanoparticles can be observed at atomic-resolution and in reactive gas environments [1]. This capability opens up new possibilities for uncovering dynamics of nanoparticles immersed in gas environments. Here, we employ such in situ TEM capabilities to study the growth of Cu nanoparticles supported on SiO₂ by the reduction of a homogeneous precursor consisting of Cu phyllosilicate platelets (Fig. 1) [2, 3]. The homogeneous precursor represents an important class of industrial catalyst precursor material for which local TEM observations can directly be related to the catalyst at a macro-scale.

By monitoring the individual nanoparticles over time, quantitative information about the nucleation time and size evolution of the Cu nanoparticles were obtained from TEM image series (Fig. 2). To employ such quantitative data in a kinetic description of the reduction process, it is imperative to reduce the influence of the electron beam on the process to a negligible level. To achieve that, we developed a low dose-rate TEM imaging procedure in which several regions of the sample was monitored in parallel in a dose-fragmentated way [1,4,5,6].  This strategy allows for a direct evaluating of the role of the electron dose-rate, accumulated dose and dose history, which were all found to play a role on the growth process (Fig. 1C). Based on this evaluation, the role of the electron beam was quantitatively assessed and used to pinpoint illumination conditions that allowed the observation of the inherent thermal reduction process.

In view of this detailed analysis, quantitative data were extracted from the time-resolved TEM image series to describe the thermally induced growth of the Cu nanoparticles supported on SiO₂ (Fig 1B). By comparing the quantitative TEM data of the process with kinetic models, it was found that the growth process was best characterized as an autocatalytic reaction with either diffusion- or reaction-limited growth of the nanoparticles. This finding is significant because the autocatalytic reaction limits probability for secondary nucleation and because the platelike precursor structure restricts the diffusion. This describes a way to synthesize nanoparticles with well-defined sizes, which in turn offers a more stable catalyst[7]. In this way, in situ observations made by electron microscopy provide mechanistic and kinetic insights that can guide the formation of metallic nanoparticles for catalytic applications in a rational way.

References: 

[1] S. Helveg, J. Catal. 328, 102 (2015)

[2] R. van den Berg et. al., J. Am. Chem. Soc. 138, 3433 (2016)

[3] R. van den Berg et. al., Catal. Today (2015)

[4] C. Holse et. al., J. Phys. Chem. 119, 2804 (2015)

[5] J. R. Jinschek and S. Helveg, Micron 43, 1156 (2012)

[6] S. Helveg et. al., Micron 68, 176 (2015)

[7] S. B. Simonsen et. al., J. Am. Chem. Soc. 132, 7968 (2010)

Boron nitride (BN) nanostructures exhibit excellent mechanical properties. For example, the in situ tensile tests on individual BN nanotubes (BNNTs), which were conducted in situ in a transmission electron microscope (TEM) column, demonstrated the strength and Young’s modulus of ~33 GPa and ~1.3 TPa, respectively [1].Such superb mechanical properties of BN nanostructures make them very attractive materials as a reinforcement phase in lightweight composites. Besides nanotubes, nano-BN can be obtained in the form of spherical nanoparticles (BNNPs) with different morphologies – solid or hollow, with smooth or rough surfaces. High electrical and chemical resistance, thermal stability and biocompatibility of BNNPs were reported but their mechanical properties have not been detailed yet.

BNNPs in the present study were synthesized by CVD technique. By changing technological parameters of CVD process particles of various morphologies were obtained. For each type of individual BNNPs in situ compression tests were accomplished in a TEM column. As-synthesized BNNPs were dispersed in acetone and the suspension was placed onto a Si wafer. The truncated cone-like shape indenter with a diameter of top circle of 1 μm was used. Elastic moduli were determined for all tested particles. The obtained results (Fig. 1) show that the high strength enables the hollow spherical BNNPs to withstand a considerable compressive deformation before failure (up to 50% of the hollow particle diameter). The structures also demonstrated a high percentage of elastic recovery.

1. X.L. Wei, M.S. Wang, Y.Bando, D.Golberg, “Tensile tests on individual multiwalled boron nitride nanotubes,” Adv. Mater. 22 (2010), 4895-4899.