Symposium S.EL02 : Advanced Manufacturing of Mixed Dimensional Heterostructures

Antimony selenide solar cells are rapidly approaching 10% efficiency however a number of key materials properties are still not fully understood. In particular the mechanism by which doping is controlled in the material has not been established and there is a high degree of variability in the literature on the level of doping attainable and even carrier type. This talk will discuss the identification of potential sources of doping in Sb₂Se₃, the influence this has on carrier concentration and the subsequent impact for device performance. Through the use of Sb₂Se₃ bulk crystals we will further demonstrate routes to achieve p, i and n doping and methods to transfer this to thin film devices.

The last decade has seen nearly exponential growth in the science and technology of two-dimensional materials. Beyond graphene, there is a huge variety of layered materials that range in properties from insulating to superconducting. Furthermore, heterogeneous stacking and doping of 2D materials also allows for additional band structure engineering. In this talk, I will discuss recent breakthroughs in two-dimensional atomic layer synthesis and properties, with am emphasis on doping 2D semiconductors. Our recent works include development of an understanding of substrate impact on growth and doping of 2D materials to tune them from n- to p-type, and to create 2D magnets. Our work and the work of our collaborators has lead to a better understanding of how substrate not only impacts 2D crystal quality, but also doping efficiency in 2D materials.

The current electronics has been mainly dominated by Si-based devices due to their mature processing system and exceptional cost-effectiveness. However, next generation electronics needs novel functionalities that cannot be realized by Si because of intrinsic limitation of Si. Accordingly, demand for non-Si electronics has been getting substantially high. Unfortunately, current methodology requires extremely high cost for non-Si materials, which impedes the progress in developing the non-Si based electronics. Here, I will discuss about our group’s efforts to address this issue. Our team recently conceived a new crystalline growth, termed as “remote epitaxy”, which can copy/paste crystalline information from substrates remotely through graphene, thus generating single-crystalline films on graphene. As interfacial binding energy is attenuated by inserting graphene at interface, the single-crystalline films can be easily exfoliated from the slippery graphene surface. Also, the graphene-coated substrates can be, in principle, reused infinitely to produce single-crystalline films. Thus, the remote epitaxy can produce non-Si semiconductor films with unprecedented cost efficiency while allowing additional flexible device functionality required for current ubiquitous electronics.
Next, I will discuss about a layer splitting technique which can be a potential solution to overcome the problem in obtaining large-scale and monolayer 2D materials. A 2D material-based heterostructure has been intensively studied because of its unique device functionalities and novel physics. However, it is extremely challenging to secure large-scale and monolayer 2D materials because of following issues: 1) poor scalability for laboratory fabrication processes of 2D heterostructures and 2) lack of well-defined control parameters for kinetics of 2D materials and predictable number of layers of 2D materials. To resolve this issue, we conceived a new approach called “layer-resolved splitting” which obtains multiple monolayer from multilayer 2D materials by controlling interfacial toughness contrast. As this method is versatile and universal, we can, in principle, apply to all 2D materials. We succeeded in having large-scale, monolayer 2D materials through our approach and, thereby 2D heterostructures were demonstrated for functional devices.
Lastly, I would like to discuss opportunities of mixed-dimensional heterostructure demonstrated by remote epitaxy and layer-resolved splitting. As they produce freestanding 3D bulk films and 2D atomic layers, a new type of 3D/2D heterostructures can be realized where a new physics and new device architecture are revealed. Therefore, I believe that a new opportunity will be discovered through the mixed-dimensional heterostructures

New materials and device integration demands motivate the ability to lift off and transfer epitaxial material. We use a Van der Waals-bonded 2D boron nitride (BN)-on-sapphire template to grow GaN by MOCVD then subsequently separate the GaN using a stress-induced lift-off method. This method has potential for wafer-scale heterogenous integration since large crack-free areas can be separated and the bottom epi-layer surface has sub-nanometer roughness. In this method a tensiley-stressed Ni layer deposited on the GaN acts to separate the epi-layer at the 2D BN interface allowing us to lift off whole wafer-sized GaN layers which are crack-free. Although large crack-free areas of GaN can be separated, the residual compressive strain in the GaN can cause cracks if the strain is inhomogeneously relaxed as the layer is transferred to the new substrate. By varying the stress and thickness of the Ni layer we can homogeneously relax some of the GaN strain upon separation and reduce the resulting crack density. We report on our development of this process including modeling and measurement of the GaN strain by Raman spectroscopy throughout the process.

Due to the current global water crisis, wastewater treatment requires considerable attention and development. Dissolved heavy metals in water trigger serious alerts and can have lethal effects on various components of the environment including water resources, soil, plants, animals, and can also be threatening the health of human beings. In this study, we addressed the issue by introducing novel bucky-paper membranes that were fabricated using a combination of biopolymers (i.e., chitosan and carrageenan) and multi-walled carbon nanotubes (MWCNTs). Three dispersions of MWCNTs with chitosan, carrageenan, and chitosan-carrageenan with 0.1% v/w were prepared using vacuum filtration. The removal of six heavy metals (i.e., cobalt, nickel, copper, cadmium, barium, and lead) was investigated in this study. The water permeability and the removal of heavy metals were evaluated using a dead-end (DE) filtration system. Heavy metals removal was studied at pH 7 and under a range of varying applied pressures (1 to 6 bar). At an applied pressure of 1 bar, the removal of lead and copper by the MWCNTs/carrageenan membrane reached 99% and 88%, respectively. However, MWCNTs/carrageenan membrane was found to be fragile. Nevertheless, adding chitosan to carrageenan had significantly improved the mechanical strength of the membrane while sustain the excellent removal properties of the heavy metals. That is, MWCNT/chitosan-carrageenan membrane significantly exceeds MWCNTs/carrageenan membrane in tensile strength, tensile strain and young’s modules by 400%, 17%, and 6%, respectively. On the other hand, the MWCNTs/chitosan membrane showed a high water permeate flux that reached up to 200 L/h.m². Also, the electrical conductivity of all membranes varied from 37 S/CM to 57 S/CM. Additional characterization techniques on the three membranes were conducted in this study as well.

When an electromagnetic wave impinges on a conductor, most of the incident wave is reflected, which makes metals and transparent conducting oxides (TCOs), such as tin-doped indium oxide (ITO), opaque to visible light (429–750 THz) and far-infrared (FIR) (< 20 THz). This incompatibility between optical transparency and electrical conductivity is well-defined fundamental material properties, but this is often not easy to enhance both simultaneously. Opacity due to electrical conductivity is more pronounced in the lower frequency range. This fundamental incompatibility creates a barrier for the realization of enhanced user-interface and device integration. We present a design strategy for preparing megahertz-range transparent conductor and a concept towards ‘device-to-device integration’ enabled by electromagnetic wave transmittance. The approach to the properties of conductors is verified using a conducting polymer, Poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS), whose microstructure is effectively controlled by solution process. The use of a transparent conducting polymer as an electrode enables the fabrication of a fully functional touch-controlled display device and magnetic resonance imaging (MRI)-compatible biomedical monitoring device, which would open up a new paradigm for transparent conductors.

Among all the anode materials, Li and Na have the highest capacity and great potential to increase the energy density of batteries. Unfortunately, dendrite growth in metal anodes is one of the safety concerns in current battery devices. One approach to address this issue is to use molten or liquid-metal electrodes. This concept has been applied to sodium-beta alumina batteries (NBBs), which are based on a liquid-Na anode and β″–Al₂O₃ solid electrolyte. Consequently, NBBs, such as Na-metal halide (ZEBRA) batteries, are among the most promising technologies for large-scale renewable energy storage because of their high theoretical specific energy, high energy efficiency, and good cycle life.

Interestingly, in the initial stages of sodium-beta alumina batteries (NBBs) development, it was not anticipated that the low performance would rise from the Na/β″–Al₂O₃ interface. Liquid Na metal should be an ideal reversible electrode, provided that it can be maintained in contact with the whole operating area of the β″–Al₂O₃ throughout the discharging and charging of the cell. It was expected to be the least problematic component of the cell. In reality, liquid Na does not fully wet the surface of the β″–Al₂O₃. Although early observations revealed that incomplete wetting of the β″–Al₂O₃ surface by Na and consequent interfacial impedance problems were likely in NBBs, little was reported for numerous years. Ideally, full contact with liquid Na should be achieved, which provides a large active interface area and thus leads to high NBB performance at low temperatures.

We propose a simple approach to achieve unprecedented NBB performance with a capillary-induced wetting concept that significantly improves the Na wetting on β″–Al₂O₃. In this study, we adopted sparked reduced graphene oxide (rGO) loaded on the surface of the β″–Al₂O₃ as an ideal “wetting sheet,” compared with previous metal (Li or Na)-ion batteries where sparked rGO layers have been used for such metal reservoirs as an anode. The sparked rGO layers with nanogaps exhibited complete liquid-Na wetting, regardless of the surface energy between the liquid Na and graphene oxide, which originated from the capillary force in the gaps (see the figures). This indicated (i) that the Na nucleation and growth were sufficiently rapid to form liquid Na when the battery was charged and (ii) that the area of the Na+ passway at the Na (anode)/β″–Al₂O₃ (electrolyte) interface was maximized when the battery was discharged. Thus, a high state-of-charge for potential state-of-art low-temperature NBBs was obtained. This cell-stacking architecture is simple and scalable and addresses the fundamental limitations of NBBs by allowing Na wetting at low temperatures.

Van der Waals (vdW) materials are composed of two-dimensional planes that are held together by weak interlayer interaction. VdW crystals have attracted increasing attention in the past couple of decades because of their significantly different properties from three-dimensional materials. However, it is a small class of materials, with fewer than 100 layered compounds catalogues in the Inorganic Crystal Structure Database (ICSD), including graphite, h-BN, transition metal dichalcogenides, metal halides, metal pnictides, and metal oxides. Here, we present a topochemical redox concept to prepare a new binary vdW materials by structure control from ternary parent compounds. The vdW crystals obtained had the same chemical composition as known three-dimensionally bonded compounds, but exhibited layered crystal structures. In our vdW crystal design strategy, we synthesized LiFeAs and CaFe₂As₂ as parent compounds. We then selectively removed Li and Ca from parent materials by topochemical redox to obtain FeAs vdW crystals with various crystal structures. We confirmed crystal structure of FeAs vdW crystal by X-ray diffraction (XRD), transmission electron microscope (TEM), scanning transmission electron microscope (STEM) and X-ray photoemission spectroscopy (XPS) and measured electrical property and mechanical property to identify the difference FeAs vdW crystal with three-dimensional orthorhombic FeAs.

Heteroepitaxy of semiconductor on two-dimensional atomic layered materials (2d-ALMs) has been promising for fabricating transferrable and flexible devices, because the use of 2d-ALMs allows to easily exfoliate overlayer device from host substrate.[1,2] Recently, an emerging epitaxy has been reported, which is the so-called remote epitaxy.[3,4] The remote epitaxy enables to produce single crystalline overlayer on graphene layer because the crystallographic registrations of overlayer can be copied from a underlying single crystalline substrate across graphene layer.[4,5] In the remote epitaxy of microrods (MRs), graphene thickness that is the remote epitaxial gap is critical to determine growth density of MR overlayer because penetrated field strength given from substrate leading to nucleation–growth is attenuated as the graphene thickness increases.[4] This implies that as we use graphene interlayer with spatially different thicknesses can result in different growth regimes of i) remote epitaxy or ii) non-growth, depending on graphene thickness.
We present how the remote epitaxy can be applied to selective-area growth using graphene pattern layer. Position-controlled remote heteroepitaxy was performed by hydrothermal growth of ZnO MRs on intaglio-patterned graphene (IPG)-coated c-plane GaN substrate. The IPG layer consists of two graphene parts of (I) perforated-hole-patterned multilayer graphene (MLG) that prevents the remote epitaxy, which act as a growth mask layer, and (II) single-layer graphene (SLG) penetrating the potential field from underlying GaN substrate to allow remote heteroepitaxy of ZnO through the hole aperture of mask layer, which is embedded between GaN and MLG mask layer.
The hole-aperture SLG area yielded remote heteroepitaxial ZnO MRs, whereas MLG area inhibited growth. Diameter and spacing of ZnO MRs are controlled by changing the hole pattern parameters. Transmission electron microscopy revealed the remote heteroepitaxial relationship between ZnO and GaN across the SLG area. According to density-functional theory calculations, the orbitals of SLG transfer the charge from the underlying GaN to the SLG surface leading to remote epitaxy, but the thick MLG was not capable of charge transfer in a long range. The weak van der Waals adhesion of ZnO/IPG/GaN was applied to exfoliate the ZnO MRs overlayer by a thermal release tape-assisted exfoliation technique and to recycle the original substrate. Our results readily open an opportunity to utilize the remote epitaxy for transfer of arrayed epitaxial structures in the designed size and arrangement for device manufacturing.

Reference
[1] K. Chung et al., Science 330, 655 (2010)
[2] C. H. Lee et al., Adv. Mater. 23, 4614 (2011)
[3] Y. Kim et al., Nature 544, 340 (2017)
[4] J. Jeong et al., Nanoscale, 10, 22970 (2018)
[5] W. Kong et al., Nat. Mater., 17, 999 (2018)

Two-dimensional atomic materials are emerging as epitaxial substrates for transferrable and flexible device application.[1] Graphene has excellent properties for the use as a substrate, such as high electrical conductivity, excellent mechanical strength, and optical transparency.[2,3] Nevertheless, since the difficulties of producing single crystalline thin film and crystallographically aligned nano/microstructures on graphene, due to the use of poly-domain graphene, there remains a challenge to utilize the graphene as a substantially epitaxial substrate. This obstacle can be overcome by the remote epitaxy.[4] The remote epitaxy enables to grow single crystalline overlayer on graphene regardless of the graphene domain because crystallographic registration can be dictated from underlying substrate through graphene. Here, we demonstrate remote homoepitaxy of ZnO microrods (MRs) on different crystal planes of apolar a- and polar c-plane ZnO substrates across graphene using hydrothermal growth method.
Despite of the presence of poly-domain graphene intermediate layer, the ZnO MRs were epitaxially grown on a-plane and c-plane ZnO substrates, which were found to be homogeneous in-plane orientation over the entire surface of graphene-coated ZnO substrates. Such homoepitaxial relationship across graphene between ZnO MR and substrate was revealed through transmission electron microscopic and selected area electron diffraction analyses. The density-functional theory calculations suggested that the charge redistribution occurring near graphene induces the electric dipole formation, so the attracted adatoms lead to nucleation-growth of the remote-epitaxial overlayer. Because of a strong potential field caused by long-range charge transfer given from the substrate, even the use of bi-layer and tri-layer graphene resulted in the remote-epitaxial ZnO MRs. The effect of substrate crystal planes is also theoretically and empirically demonstrated. The ability of the graphene, which can be released from the host substrate without covalent bonds, was adapted to transfer the overlayer MR arrays. After the delamination, the host substrate was reused by repeating the process of remote epitaxy over again. This unconventional epitaxy technique offers an opportunity of the producing well aligned, transferrable and flexible epitaxial nano/microstructure arrays templates for epitaxial electronics and optoelectronics applications and regenerating the substrate for cost-saving device manufacturing.

Reference
[1] K. Chung et al., APL Mater. 2, 092512 (2014)
[2] Y. J. Hong et al., ACS Nano 5, 7576 (2011)
[3] A. M. Munshi et al., Nano Lett. 12, 4570 (2012).
[4] Y. Kim et al., Nature 544, 340 (2017)

III-nitride materials including GaN, InN, AlN, and their ternary alloys have been one of the key materials to realize advanced electronic and optoelectronic devices, such as high electron mobility transistors, light emitting diodes and lasers. The device performance heavily depends on defect density of III-nitride materials. However, the lack of native substrates leads to the challenges in fabricating defect-free III-nitride materials, severely limiting the implementation of variety of new devices utilizing III-nitride materials. Additionally, the attachment of III-nitride thin film to its substrate poses challenges of heterointegration of III-nitrides with other material systems. Separation of III-nitride thin films from the substrate while maintaining the pristine material quality is highly favorable.

We have previously shown the synthesis of GaAs on two dimensional materials followed by the separation of GaAs epitaxial thin film from its substrate [1], by utilizing the process so-called “remote-epitaxy”. Such discovery reveals the possibilities towards the fabrication of defect-free semiconductor materials without the constriction of substrate availability. In this report, we demonstrate the remote-epitaxy process to fabricate thin film GaN [2], as well as its ternary InGaN. The remote-epitaxial materials have been characterized by X-ray diffraction, atomic force microscopy and transmission electron microscopy, rendering device grade material quality. Additionally, the thin film III-nitrides can be peeled off from the substrates, and subsequently bonded to a foreign material surface, including Si and SiO₂. Based on GaN/Si heterostructure, III-nitride based photonic/phononic cavity is demonstrated. This work has demonstrated a promising path for the integration of III-nitride with an arbitrary material system.

Reference:
[1] Kim, Yunjo, et al. "Remote epitaxy through graphene enables two-dimensional material-based layer transfer." Nature 544.7650 (2017): 340.
[2] Kong, Wei, et al. "Polarity governs atomic interaction through two-dimensional materials." Nature materials 17.11 (2018): 999.

III-V compound semiconductors offer outstanding electronic and photonic properties that outperform silicon, but the cost of III-V wafers is extremely expensive. Although reusing original wafers can effectively minimize the cost of wafers, current techniques for wafer recycling add significant costs in fabrication, nullifying the cost savings by reusing the wafer. Remote epitaxy is a newly discovered method that enables single-crystal growth of III-V semiconductor thin films and easy exfoliation of the grown film, thus promising for reusing wafers multiple times. It requires an atomically thin spacer layer, such as graphene, to be present on top of III-V substrates, which allows enough field penetration from substrates through the graphene layer to make epitaxial III-V thin films grown on it maintain single-crystallinity. In addition, the graphene layer provides weak bonding at the III-V/graphene interface, and thus thin films grown on top of graphene can be easily exfoliated, leading to a cost-effective way of wafer reuse and flexible thin film production. However, previous methods of transferring graphene onto III-V wafers, which use polymethyl methacrylate (PMMA) or metal stressor layers to transfer graphene grown on foreign substrates like copper or SiC, introduce defects and damages on graphene and/or substrates during the transfer process. Remote epitaxial films grown on the damaged graphene/substrate suffer from lower crystal quality and imperfect exfoliation, which undermines wafer reusability and device performance.
Here we report the CVD growth of amorphous graphene on III-V wafers at low temperature that enabled improved quality of remote epitaxial films and their perfect exfoliation. By introducing toluene as a carbon source which cracks at a relative low temperature for graphene growth, we show fully covered amorphous graphene on AlGaAs/GaAs substrates despite arsenic’s low decomposition temperature. The surface of graphene-coated AlGaAs/GaAs substrate remains smooth with a RMS roughness of around 3Å. We also demonstrate 100% coverage of single-crystalline GaAs thin films grown on amorphous graphene, with the film’s quality significantly improved compared to the case of transferred graphene. In addition, roughness of the substrate’s surface remains the same after exfoliation of grown GaAs film, and the growth and exfoliation were successfully repeated multiple times, proving the feasibility for wafer recycling. Through this low temperature CVD growth approach and remote epitaxy, we successfully demonstrate wafer-scale flexible thin film exfoliation and recycling of substrates, which will lead to new opportunities in III-V thin film-based electronics and novel heterostructures with reduced cost.

A luminescent solar concentrator (LSC) is a photon managing device that can harvest, direct and concentrate solar light to small areas, enabling subsequent coupling to photovoltaic devices (PVs) for enhanced solar energy conversion. However, the intrinsic photon loss through the so-called escape cone of the LSCs significantly limits their light harvesting and concentrating performance. In this talk, I will introduce a facile and low-cost approach for the fabrication of a three-dimensional (3D) macroporous photonic crystal (PC) filter as an efficient photon reflector, which can be coated onto quantum dot (QD) based LSC devices. We demonstrate that by controlling the PC reflection band to match the emission profile of the QD emitters, the light trapping efficiency of the PC coated LSC (PC-LSC) can be significantly improved from 73.3% to 95.1% as compared to the conventional PC-free LSC due to the reduced escape cone photon loss. Both experimental and simulation results show that the enhancement in LSC device performance induced by the PC reflector increases with increasing dimension. Our study sheds light on future design and fabrication of LSC devices with enhanced photon collection and concentrating efficiencies through novel and wavelength-selective photon reflectors.

High-performance semiconducting films with precisely engineered thicknesses and compositions are essential for developing next-generation electronic devices, which are becoming more integrated, complex, and multifunctional. My talk will introduce the novel processes that enable atomic-scale control of the thickness and spatial composition of semiconducting films on the wafer-scale. These processes include: (i) the wafer-scale generation of atomically-thin layered semiconductors such as metal dichalcogenides and metal oxy-dichalcogenides via metal-organic chemical vapor deposition (MOCVD), (ii) the atomic-level engineering of vertical thickness and composition through the layer-by-layer assembly of the layered semiconductors, and (iii) the interlayer engineering of the layered semiconductors for the property control. These capabilities provide a new material platform for both fundamental research and practical applications, including incorporation into existing integrated circuit technology to form hybrid materials and boost electrical and optical functionality.

In this talk, I will discuss about our unique strategy to isolate wafer-scale 2D materials into monolayers and stack them into a heterostructures by using a layer-resolved splitting (LRS) technique [1,2]. This technique enables my group at MIT to explore unprecedented wafer-scale 2D heterodevices. While 2D heterostructures promise interesting futuristic devices, the performance of 2D material-based devices is substantially inferior to that of conventional 3D semiconductor materials. However, 3D materials exist as their bulk form, thus it is challenging to stack them together for heterostructures. Obviously, conformal coating of such single-crystalline bulks on 3D features is impossible. My group at MIT has recently invented a 2D materials-based layer transfer (2DLT) technique that can produce single-crystalline freestanding membranes from any compound materials including III-V, III-N, and complex oxides [3,4]. This technique is based on remote epitaxy of single-crystalline films on graphene followed by peeling from graphene. Stacking of freestanding 3D material membranes will enable unprecedented 3D heterostructures whose performance is expected to be superior to that of 2D heterostructures. I will talk about our group’s effort on 3D heterostructures as well as 2D-3D mixed heterostructured devices [5].

[1] J. Kim et al., “Layer resolved graphene transfer via engineered strain layers”, Science, 342, 833 (2013)
[2] J. Shim, S. Bae, et al, and J. Kim, “Controlled crack propagation for atomic precision handling of wafer-scale two-dimensional materials” Science, 362, 665 (2018)
[3] Y. Kim, et al, and J. Kim, “Remote epitaxy through graphene enables two-dimensional material based layer transfer” Nature, Vol. 544, 340 (2017)
[4] W. Kong, et al, and J. Kim, “Polarity govern atomic interaction through two-dimensional materials”, Nature Materials, Vol. 17, 999 (2018)
[5] S. Bae, et al, and J. Kim, “Integration of bulk materials with two-dimensional materials for physical coupling and applications”, Nature Materials (2019)

Remote epitaxy is a method for growing single-crystalline materials on a graphene-terminated single-crystalline substrate, in which the epitaxial registry occurs between film and substrate rather than film and graphene. This method has been realized for several homoepitaxial systems, such as GaAs, GaN, LiF and ZnO. Yet, the microscopic mechanisms that allow for remote epitaxy remain mysterious. Here we demonstrate remote homoepitaxy of GaSb and comment on the role of defects on the apparent lattice transparency of graphene. In addition, we elucidate the effect of the graphene/substrate interface quality associated with different graphene transfer methods.

We have grown our films via molecular beam epitaxy and used x-ray and electron diffraction, along with transmission electron microscopy to confirm that the single-crystalline film is in-phase with the underlying GaSb (001) substrate. Graphene grown via chemical vapor deposition is cleanly transferred to GaSb, which has been corroborated via Raman spectroscopy and scanning electron microscopy. Remote epitaxy necessitates the absence of the semiconductor substrate’s native oxide, since this amorphous layer would inhibit epitaxial registry to the underlying substrate. Therefore, the question of how the native oxide desorbs when capped with graphene is of critical importance. We have studied this desorption mechanism through in-situ photoemission and reflective high energy electron diffraction. In addition, we have investigated the nucleation selectivity of GaSb on patterned and unpatterned graphene via in-situ scanning tunneling microscopy. The GaSb overlayer can be readily exfoliated, which allows for the substrate to be recycled and for the exfoliated film to be transferred to arbitrary substrates.

Ferroic materials including oxide ferroelectrics, magnets, and multiferroics are of great interest for a range of modern applications. What’s more, these materials stand poised to revolutionize next-generation applications, if we can determine ways to overcome inherent limitations in their function. Stringent device requirements are pushing researchers to exert ever more exacting control over the chemistry, defect structures, interfaces, etc. to elicit the desired function from these materials. From the ever present need to reduce leakage currents and control imprint and fatigue to renewed interest in reducing the voltage of operation of ferroelectric and piezoelectric devices to the sub 100 mV level to the realization of beyond-binary function in the form of multi-state operation that could enable neuromorphic devices, the requirements of next-generation applications are great and varied. Such requirements have thrust materials – including ferroelectrics, piezoelectric, and multiferroics – back into the mix as candidates for many applications.

But, despite considerable research on such materials in the last few decades and advances in our ability to synthesize, control, characterize, and fabricate these materials, the requirements of these applications are already pushing these materials to their limits. In this talk, we will explore a range of materials synthesis and processing approaches that enable researchers to address these requirements. In single-layer films of materials like BaTiO₃, PbTiO₃, BiFeO₃, (1-x)PbMg_1/3Nb_2/3O₃-(x)PbTiO₃, and others, we will explore how the use of exacting chemical control and the ex post facto introduction of certain species (be it constituent elements themselves or specific defects) can effectively “heal” or improve ferroelectric function dramatically – improving leakage and increasing breakdown voltages. Taken further, such on demand introduction of defects can even impart exotic function – such as multi-state switching – as we leverage strong defect-polarization coupling to tune the pinning energy and, in turn, coercive voltages. One can even achieve such control at the nanoscale level, opening the door for the on-demand design of functionality. At the same time, careful synthesis and production of pristine materials and interfaces can also produce thin-film ferroelectrics that begin to approach the switching achieved in single-crystal materials. This, in turn, opens the door for record-breaking small coercive voltages and ultra-low power function and operation. In multi-layered and superlatticed materials, the careful interfacing of materials can produce new combinations of properties – including combined large polarization and large dielectric constant and tunability – in, for examples, PbZr_1-xTi_xO₃ superlattices controlled to have overall compositions near the morphotropic phase boundary (MPB) composition, but being built from layers far from the MPB composition. Finally, we will explore new efforts in the production, utilization, and testing of free-standing versions of ferroelectrics. New selectively etchable layers are now enabling the production of large, free-standing epitaxial stacks. Routes to release, transfer, and measure free-standing films will be introduced and a range of properties measured therein.

Epitaxial growth has been widely used for high-quality growth of crystalline materials. However, epitaxial growth techniques usually require lattice match and similar thermal expansion coefficient between growth material and substrate, which blocks this growth technique from wide applicability. Recently, van der Waals (vdW) epitaxy has been demonstrated that allows to relax the limitation of epitaxial growth. vdW epitaxy on layered or two-dimensional (2D) materials is mediated by weak vdW interactions. In addition to these interactions, a substrate below 2D materials also plays a role to intervene in the epitaxial growth process, leaving room to realize a novel epitaxial growth process: remote epitaxy. Potential fluctuation—the difference between potential energy maxima and minima along the surface of the substrate—is influenced by this remote interaction; the fluctuation is predominant over the vdW force, due to the atomic thickness of 2D materials. The remote interactions can be controlled by modulation of polarities of 2D materials and epitaxy materials. Polar-material-based heterostructure, such as GaAs, InP, GaP and GaN with their substrate sandwiching 2D materials, enables the semiconductor film to be exfoliated from the underlying substrates through the 2D materials. This process is expected to open a novel avenue for the field of non-silicon electronics and photonics, where the ability to re-use graphene-coated substrates allows savings on the high cost of non-silicon substrates.

Demands on flexible lighting apparatus have been rapidly rising for the use in mobile display and wearable lighting device applications. For the device performance and stability, the use of semiconductors is ultimately desirable. Nonetheless, the rigid, brittle characteristics of inorganic semiconductor materials in a film form on wafer have so far precluded the use from the applications. In terms of growth template, graphene, which is flexible and allows delaminate of device overlayer from host substrate due to weakly bound van der Waals bonds to substrate, have begun to be utilized as an emerging epitaxy substrates for transferrable and flexible device applications.[1-3] As for the overlayer, the spatially separate arrays of lighting components are adequate for fabricating the flexible devices, and vertical wire- or rod-based structures are thought to be ideal for the purpose. Without the epitaxial relationship, the rod structures cannot be uni-directionally (or vertically) aligned, which degrades the geometrical feature for conformal electrode formation on top of rod light emitters. Meanwhile, there have recently been an emerging epitaxy technique, the so-called remote epitaxy that the crystallographic registration of single-crystalline wafer is dictated to overlayer across graphene.[4,5] Thus, single crystalline epi-layer is easily obtained for many compound semiconductors via the remote epitaxy. Likewise, once the remote epitaxy is adopted for fabricating the rod-based heterostructures, uni-directionally aligned, spatially separate lighting components can be obtained, which makes the arrays transferable and deformable after transferring on flexible target surface. Hence, the fabrication of rod heterostructures via graphene-mediated remote epitaxy enables flexibility and transferability of the overlayer rod-devices.
Herein, we demonstrate the deformable light-emitting diode (LED) panel via the remote heteroepitaxy of vertical GaN microrod (MR) radial p–n junction arrays embedded with InGaN/GaN multiple quantum-wells on c-Al₂O₃ substrates with a gap of graphene interlayer. The remote epitaxy produced GaN MR arrays with uniform ordering of hexagonal sidewall facet orientation over whole graphene surface, suggesting a specific heteroepitaxial relationship between GaN MRs and c-Al₂O₃. Despite the presence of poly-domain graphene interlayer, the heteroepitaxial relation between GaN MR and substrate was revealed via cross-sectional transmission electron microscopy. Density-functional theoretical calculation demonstrates how/why the remote heteroepitaxy with large lattice mismatch is made possible across graphene. The use of graphene allowed exfoliation and transfer of MR LED overlayer onto flexible conducting copper plate, and the spatially separate MR arrays enabled to deform the LED device in diverse flexural forms, including folded, twisted, and even crumpled deformations. Repetitive bending test over 1,000 cycles between radii of curvatures of ∞ and 10 mm represented the robustness of MR LEDs. Also, the original substrate is recycled, which consistently re-yields the deformable LED panels over again. This study expands the epitaxy technique into flexible devices and substrate recycle.

Reference
[1] K. Chung et al., Science 330, 655 (2010)
[2] C. H. Lee et al., Adv. Mater. 23, 4614 (2011)
[3] Y. J. Hong et al., Nano Lett. 12, 1431 (2012)
[4] Y. Kim et al., Nature 544, 340 (2017)
[5] W. Kong et al., Nat. Mater. 17, 999 (2018)

Tailoring the surface properties of carbon nanotubes (CNTs) through covalent or non-covalent interactions with organic molecules has been highlighted for more than a decade as promising path for the fabrication of functional nanostructures that could be successful in a wide variety of application ranging from nanomedicine to reinforced structural materials¹. However, the efficient functionalization of CNTs, especially multiwalled-CNTs (MWCNTs), still faces many challenges due to the lack of understanding of the interaction mechanisms between MWCNTs and organic molecules in functionalization procedures. In this study, we compare the functionalization of MWCNTs through covalent and non-covalent methods, with the objective of elucidating clear synthetic guidelines for the efficient functionalization of MWCNTs.
To understand the mechanisms of covalent functionalization of MWCNTs we attempted the functionalization of MWCNTs with polystyrene molecules via Prato and nitrene cycloaddition reactions. Through the comparison of TGA and Raman spectroscopy of the products, we showed that the nature of the reaction has no impact on the degree of functionalization of the MWCNTs. In addition, we highlight the use of control experiments using non-reactive polystyrene molecules, that allowed us to demonstrate that the degree of functionalization of MWCNTs was comparable to the products of the covalent functionalization reactions. The independence in the degree of polymer functionalization of the reaction procedure and of the use of non-reactive polystyrene was observed for MWCNTs with different diameter sizes. These experimental evidences demonstrate that covalent bonds between polymers and pristine MWCNTs are not present and that the functionalization occurs through non-covalent forces.
Non-covalent functionalization was then performed in six different solvents to explore the impact of the solvent in the polystyrene-MWCNTs interactions. We observed a characteristic degree of functionalization for each solvent and we propose for the first time a simple model based on surface energy solubility parameter theory that can be used as a synthetic guideline for the design of polymer functionalised MWCNTs². Finally, we demonstrate that this model can be applied for the fabrication and design of nanoparticle/carbon nanotube hybrid structures. We believe that our results and the application of the surface energy based solubility parameter theory as a qualitative model for the design of functionalized MWCNTs will lead to the rational synthesis of a wide range of novel nanostructures.

References
1 D. Tasis, N. Tagmatarchis, A. Bianco, M. Prato, D. Tasis, N. Tagmatarchis, A. Bianco and M. Prato, Chem. Rev., 2006, 106, 1105–1136.
2 P. Quijano Velasco, K. Porfyrakis and N. Grobert, Phys. Chem. Chem. Phys., 2019, 21, 5331–5334.

Combined photonic and electronic systems require diverse devices to be co-integrated on a common platform. This heterogeneous integration is made possible through several separation and transfer methods where the functioning epilayers are essentially released from their growth substrate. The use of 2D layered h-BN as a mechanical release layer has been demonstrated to be a promising technique for the hybrid integration of III-nitride devices. In this talk we will give an overview regarding the epitaxial growth of Nitride-based semiconductor materials by MOVPE including GaN and h-BN, and design device structures for opto-electronic devices. In addition, results on wafer-scale growth and pick-and-place transfer of LED devices via h-BN van der Waals epitaxy, starting from sapphire growth substrates will be presented. Broader range of devices (HEMTs, solar cells, sensors, etc.) and applications (transport, environment, health, etc) will be discussed.

Careful control over chemistry, composition, and morphology are critical in virtually every materials system, and new methods that enable precise control over the rational assembly of constituent building blocks are vital to the future of nanoscale fabrication efforts. Here, we will discuss our recent results published in Nature Communications on the light-driven, solution-based assembly of nanoscale building blocks into periodic, one-dimensional heterostructures, consisting of repeating metal-semiconductor junctions via a photonic nanosoldering process, as well as two other recent examples where local energy transduction is used as a strategy to drive chemistry and control nanoparticle crystal growth. Effects of anisotropic radiation pressure, heat transport, and solvent choice on material transformations and nanostructure assembly will be discussed. We believe that these approaches can be generalized to enable the additive manufacturing of nanoscale structures for numerous applications, and we hope that this demonstration will lead to wider use of optical manipulation techniques for the rational assembly of diverse sets of nanomaterials in a variety of different solvent systems.

Due to high ratio of surface to volume ratio the monocrystalline semiconductor core-shell nanowires (NWs) are very suitable as building blocks a of solar cells, optoelectronic, electronic and environmental sensor devices. They are also perfect model structures for studying nanomechanical behavior of defects appearing during strain relaxation in nanoscale objects. For this purpose strained ternary semiconductors core-shell nanowires were grown by molecular beam epitaxy (MBE) on Si or GaAs substrate using Vapor-Liquid-Solid (VLS) mechanism. Due to high mismatch of the lattice constant of basic compounds ZnTe, CdTe, MgTe, which can reach more than 7%, the growth of nanowires is challenging and their final shape and morphology are complex. By modification of growth conditions (temperature, stoichiometry) nonstandard crystal structure for given compound can be obtained. For example it is possible to obtain hexagonal (wurtzite) structure or polytypic superlattice in CdTe, which usually crystallizes in regular cubic structure. [1]. The core-shell NWs, investigated by transmission electron microscopy (TEM) are strained or partially relaxed depending on the core diameter, lattice mismatch and local thickness of the shell.
The residual strain field and core-shell interface misfit dislocation network is studied for individual nanowires by TEM. The Dislocation Density Tensor mapping is performed using extension of the Geometric Phase Analysis [2] performed on HRTEM and HRSTEM images obtained from different zone axis axes in planar view or/and using perpendicular FIB cross-section of the nanowires. Due to the significant deformations of the NWs (bends and twists) the high-resolution images in zone axes can be obtained only for the several tens of nanometers long fragments of the NWs having few micrometers of total length. Therefore, we use Scanning Electron NanoDiffraction (SEND) to map bending and twisting of the lattice as well as chemical lattice distortions related to the changes in elemental composition. The series of electron diffraction patterns obtained in the STEM mode is analyzed to receive the individual maps of the lattice distortions components, via combination of simulated dynamic diffraction patterns and local cross-correlation with experimental patterns. As a results, it is possible to determine local thickness, projected distortions axial and radial components, local lattice twist(alpha) as well as and tilts (beta and gamma) even for bent NWs. Combination of both approaches of strain measurements allows to determine the projection of the residual strain in such complex 3D geometry as well as establish the dependence between the spatial arrangement of interfacial misfit dislocation network on the asymmetry of the shape of nanowires and shell inhomogeneity.
In the case of the CdTe/ZnTe NWs crystalizing in regular crystal structure form, the core-shell interface dislocation network consisted of 60°dislocations and dissociated Lomer dislocations. We show that lateral and radial spacing between dislocations depends strongly on the local shell thickness as well as the core diameter. These findings are in qualitative agreement with theoretical prediction [3]. In the case of NW with stacking fault , due to section of the [111] plane core stacking faults with shell formed [11-1] stacking fault the static Cottrell-Lomer Lock dislocations was observed with cation core polarity. The difference of the morphology of network of interface misfit dislocations between NW having wurtzite and regular crystal structure are discussed. The initial model of plastic relaxation is proposed.
[1]Kaleta A, Kret S, Sanchez A M, et al . Phys. Pol.A1311399–405
[2] S. Kret, P. Dluzewski, P. Dluzewski, and J.-Y. Laval, Philos. Mag.83, 231 (2003).
[3] Salehzadeh, O., Kavanagh, K.L., Watkins, S.P., J. Appl. Phys. 114, 054301 (2013).
The research was supported by National Science Centre (Poland) by grants No. 2016/21/B/ST5/03411

The fabrication of thin-film semiconductor relies on epitaxy, during which direct bonding is made between the epitaxial layer and the substrate, to ensure highly ordered crystalline lattices in the epitaxial layer as a copy of its substrate. The strong physical bonding between epitaxial layer and substrate implies strict requirements of lattice matching as well as difficulties in heterointegration. We have previously shown that the insertion of a monolayer graphene in between the epitaxial layer and its substrate, in the so-called “remote-epitaxy” process, did not change the crystalline orientation of the epitaxial layer, and the epitaxial alignment between the epitaxial layer and its substrate was maintained as if the monolayer graphene was “transparent” [1]. In this report, we further explore the fundamental mechanisms behind such an epitaxial alignment through two-dimensional materials. And the general rule governs remote-epitaxy will be introduced [2].

Specifically, the transparency of two-dimensional materials to the epitaxial relationship is linked to the polarization of related material systems. Although the potential field from non-polar materials is screened by a monolayer of graphene, that from polar materials is strong enough to penetrate through a few layers of graphene. Such field penetration is substantially attenuated by 2D hexagonal boron nitride, which itself has polarization in its atomic bonds. Based on the control of transparency, modulated by the nature of materials as well as interlayer thickness, various types of single-crystalline materials across the periodic table can be epitaxially grown on 2D material-coated substrates. The epitaxial films can subsequently be released as free-standing membranes, which provides unique opportunities for the heterointegration of arbitrary single-crystalline thin films in functional applications.

[1] Kim, Yunjo, et al. "Remote epitaxy through graphene enables two-dimensional material-based layer transfer." Nature 544.7650 (2017): 340.
[2] Kong, Wei, et al. "Polarity governs atomic interaction through two-dimensional materials." Nature materials 17.11 (2018): 999.

The heterogeneous integration of dissimilar materials is a long pursuit of material science community and has defined the material foundation for modern electronics and optoelectronics. The typical material integration approaches usually involve strong chemical bonds and aggressive synthetic conditions and are often limited to materials with strict structure match and processing compatibility. Alternatively, van der Waals integration, in which freestanding building blocks are physically assembled together through weak van der Waals interactions, offers a bond-free material integration strategy without lattice and processing limitations, as exemplified by the recent blossom of 2D van der Waals heterostructures. Here I will discuss the fundamental forces involved in van der Waals integration and generalize this approach for flexible integration of radically different materials to produce artificial heterostructures with minimum interfacial disorder and enable high-performing devices. Recent highlights include the formation of van der Waals metal/semiconductor junctions free of Fermi level pinning to reach the Schottky-Mott limit; the creation of a new class of high-order van der Waals superlattices with highly distinct constituents of atomic or molecular layers; and the development of van der Waals thin film electronics with unprecedented flexibility and stretchability. I will conclude with a brief perspective on exploring such artificial heterostructures as a versatile material platform with electronic structure by design to unlock new physical limits and enable device concepts beyond the reach of the existing materials.

Interdiffusion and interfacial reactions significantly limit the growth, processing, and operation conditions for devices based on functional transition metal / semiconductor interfaces. Conventional approaches require the use of thick (several nanometer) diffusion barriers, which are generally not epitaxial, and due to their finite thickness, are a significant perturbation to the idealized functional interface: for example, half metal / semiconductor interfaces for spintronics. Here, motivated by recent demonstrations of remote epitaxy of semiconductor/graphene/semiconductor heterostructures, we quantify the efficiency of monolayer graphene as a solid-state diffusion barrier in M/graphene/semiconductor heterostructures (M = Fe, Mn).
Through in-situ x-ray photoemission spectroscopy (XPS) we show suppression of the interfacial diffusion and reactivity in these systems by tracking core level shifts as a function of anneal temperature. In samples where M = Fe, we deposit 1nm of elemental Fe on Ge (001) substrates with and without monolayer graphene. The graphene is grown directly on the Ge (001) substrate by CVD to minimize the defects introduced by conventional layer transfer processes. For samples with no graphene barrier, we observe a complicated multi-component Fe 2p core level structure that evolves with increasing anneal temperature, indicative of significant Fe/Ge interfacial reactions. In contrast, for samples with a monolayer graphene barrier, the core level spectra do not change with temperature up to 300°C and there is no evidence for a reacted phase, suggesting that monolayer graphene is an excellent solid-state diffusion barrier. We will discuss similar experiments for M = Mn and for different preparations of graphene on the semiconductor substrate, such as graphene layer transferred to Ge or GaAs. Through these studies, we comment on the optimal synthesis and processing conditions for making graphene diffusion barriers.

Acknowledgements: this work was supported by a CAREER Award from the National Science Foundation and through the University of Wisconsin Materials Research Science and Engineering Center, DMR-1752797 and DMR-1720415, respectively.

Infrared (IR) photodetectors are nowadays in high demand in a wide range of fields such as telecommunication, thermal imaging, remote sensing, night assistance car driving etc. Recently, hybrid phototransistors made of highly efficient light absorbing 0D quantum dots and high mobility 2D materials have introduced a new technology, which dramatically increases the responsivity and gain of the photodetector. In 2012, such low dimensional IR phototransistor based on graphene/PbS QD hybrid was proposed in literature ^{[1, 2]} for the first time. Later in 2017, a high-resolution broadband image sensor based on such hybrid materials was demonstrated ^[3], which is sensitive to ultraviolet, visible and infrared light (300–2000 nm).

We have investigated similar graphene/PbS QD hybrid phototransistors and explored the enhancement of their photo-sensing qualities compared to bare graphene phototransistor in the visible-NIR-SWIR region of optical spectra. Initially, we studied CVD-grown single layer graphene phototransistor in the visible-NIR region, which showed a responsivity of about 10⁴ A/W, associated with the low–doped Si substrate, which acts as the primary material for optical absorption. Such responsivity is 10⁷ orders of magnitude higher ^[4] than conventional graphene phototransistor ^[5]. However, the photo-sensing wavelength range of such device is limited up to 1100 nm depending on the principal absorption material Si. In order to expand the photo-sensing wavelength range up to SWIR region, we have explored the potential of graphene/PbS QD hybrid phototransistors. We synthesized the colloidal PbS QDs absorbing in the NIR as well as SWIR regions and developed a layer-by-layer dip coating with simultaneous ligand exchange procedure to deposit homogeneous PbS QD layers on graphene sheet leading to a well fabricated hybrid phototransistor. We achieved a significantly high responsivity of at 940 nm with irradiation power density of 10^-4 W/m². In the NIR range, the peak responsivity appears to be ∼ 10⁷-10⁸ A/W, whereas in the SWIR range, the observed peak responsivity is ∼ 10⁶ A/W . Moreover, graphene/PbS QD hybrid exhibits high sensitivity at low irradiation power (< 0.05 nW), where bare graphene photodetector is unable to sense such low intensity light in the visible-NIR region.

[1] Konstantatos, Gerasimos, et al. Nature nanotechnology 7.6, 363 (2012).
[2] Sun, Zhenhua, et al. Advanced materials 24.43, 5878-5883 (2012).
[3] S. Goossens et al. Nature Photonics, 11, 366-371 (2017).
[4] Guo, Xitao, et al. Optica 3.10, 1066-1070 (2016).
[5] Xia, Fengnian, et al. Nature nanotechnology 4.12, 839 (2009).

Ideally, electronic heterostructures from dissimilar materials leads to enhanced functionality. Yet, experimentally forming these heterostructures is challenging due to lattice or thermal coefficient of expansion mismatch leading to defect formation or thermally driven atomic diffusion resulting in cross-doping and gradual junction transitions. These challenges may be overcome with the discovery of remote epitaxy and 2D layer transfer [1]. Here, SiC epitaxy is performed on epitaxial graphene as the electrostatic fields from the substrate penetrate the graphene and guide adatom registry. The film is easily peeled away since the graphene is not bonded to either the substrate or epilayer; the epilayer is then van der Waals bonded to a different material enabling new functionality. We will present experimental results on the remote epitaxy of SiC.
There are three necessary steps to create remote epitaxy. The first is to grow epitaxial graphene on SiC, followed by transferring the graphene to a desired substrate (if different from SiC), and finally the growth of the remote epitaxial layer. If the remote epitaxy is to be SiC, which is the focus of this paper, the second step is not necessary. Epitaxial graphene (EG) was first synthesized on 4H- and 6H-SiC in a horizontal hot-wall CVD reactor between 1540 and 1580 °C in 10 slm of Ar and 100 mbar [2]. The growth temperature was dependent upon the offcut of the substrate, where substrates with higher offcuts require a lower growth temperature to ensure 1 ML of EG, which is desired to assist in SiC adatom registry during growth. SiC remote epitaxy was then performed on the EG using silane (2% in H₂) and propane precursors, where the SiC polytype replicated the underlying substrate. In an effort to transfer the remote SiC epi/EG to another substrate such as SiO₂/Si, a metallization step was performed. Thin Ti and/or Ni layers were initially deposited followed by a thicker high stress metal to create strain and aide in removing the remote SiC epi/EG from the SiC substrate [1]. Once transferred, the metal was removed via a metal etch.
In this work, we will discuss the important parameters needed for successful remote SiC epitaxy, such as graphene thickness, process flows, ramping conditions and remote epitaxy growth temperature. The epitaxial morphology characterized by SEM, Nomarski microscopy and Electron Detectio and graphene coverage and transfer evaluated by Raman spectroscopy will be presented.


[1] Kim, et al., Nature 544, 340 (2017).
[2] L.O. Nyakiti, et al., MRS Bulletin 37, 1150 (2017).

Understanding energy level alignment between molecules and two-dimensional materials is conducive to controlling charge and energy transfer in mixed-dimensional heterojunctions. The competition between distinct energy scales, such as the ones associated with electronic correlations, interface dipole, orbital hybridization, and non-local dielectric screening, can lead to significant renormalization of the intrinsic energy levels of each material at these interfaces. The variety of these energy scales also complicates the theoretical description of the level alignment for mixed-heterojunctions accurately.

Here we study the electronic structure and level alignment for mixed 0D-2D heterojunctions containing transition metal phthalocyanines on MoS₂ by first-principles density functional theory (DFT) calculations. By using optimally tuned range-separated hybrid (OT-RSH) functionals, a recently developed method for finding the optimal range-separated parameters for describing short-range and long-range electron exchange effect, we obtain energy level alignment consistent with known experimental results, a significant improvement comparing to standard DFT. This method can be used for other mixed-heterojunctions for better understanding of their electronic and optical properties.

Contemporary human life relies heavily on mobile devices such as smart-phone, lap-top, or tablet PC which process tremendous amount of information every day. In near future, rapid development of information technology would eventually integrate the devices as a wearable and flexible form that will collect and process information ubiquitously. For the fabrication of flexible and stretchable devices with high performance and reliability, monolithic integration of inorganic materials and devices in flexible form must be developed. We introduce hybrid dimensional heterostructures, composed of inorganic nanostructures grown directly on 2-dimensional (2-D) layered materials such as graphene and hexagonal boron nitride (h-BN), as one of the most promising materials for flexible device applications. The inorganic nanostructures in the hybrid nanomaterials exhibit excellent electrical and optical characteristics, including high carrier mobility, radiative recombination rate, and long-term stability. Meanwhile, 2-D layered materials, such as graphene, h-BN, chalcogenides and their hybrids, in the hybrid nanomaterials are flexible, stretchable, and mechanically strong, along with interesting and excellent physical properties. Here I report position- and morphology-controlled growths of ZnO and GaN nanostructures on graphene and h-BN using catalyst-free metal-organic vapor phase epitaxy and describe the methods to fabricate flexible light emitting diodes, transistors and pressure sensors.

Two dimensional (2D) layered materials have strong in-plane covalent bonding and weak van der Walls interaction between the planes, which host a wide variety of interesting physical and chemical phenomena. Here I will present our research over 15 years on nanostructured 2D materials. First, I will show how we synthesize these 2D materials into rich morphologies including nanoplates, nanoribbons, layer-vertical aligned films as well as lateral and vertical heterostructures. Second, we developed liquid-gel gating, solid state ion gating, chemical and electrochemical intercalation to control the physical and chemical property of 2D materials. Third, we demonstrated exciting new exciting application of using 2D materials for batteries and catalysis applications.

Recent advances in MOVPE enabled the high-quality growth of layered h-BN on sapphire substrates, and has resulted in the realization of series of 2D-3D III-nitride based device structures. This 2D layer also enables damage free mechanical lift-off which may drive future flexible and wearable industry with advanced high-efficient devices. Reliable wafer-scale 2D-2D heterostructure based devices have also been fabricated previously. Here we report, demonstration of MOVPE van der Waals epitaxial growth of GaN based nanostructures on 2D h-BN with a focus on large area van der Waals epitaxy and successful transfer to flexible platform. This approach to the growth of III-N nanostructures on h-BN avoids transfer processes and scaling issues seen with other 2D materials, since both the 1D (GaN nanorods) and 2D (h-BN) are grown at the wafer-scale and in one growth run. Further, this process is used to grow vertical core–shell p-GaN/InGaN/n-GaN nano-PIN device structures on wafer-scale 2D h-BN on sapphire and silicon substrates. Interesting results obtained on this futuristic process and mechanisms involved will be discussed.

We engineer electronic and photonic structures to control light-matter interactions including dimension and geometries. These hybrid electronic and photonic structures, especially forming microcavity exciton-polaritons are basis to build solid-state quantum simulators. Quantum simulators are constructed with special-purpose to tackle a certain class of challenging problems for example, investigation of emerging topological quantum materials and their underlying physical mechanisms. Successful demonstration of quantum simulators relies on the controllability of engineered systems. I present our current methods to establish engineered quantum simulators based on solid-state systems with performance status and future perspectives.

When crystals approach 2D, their ferroelectric phase may be destabilized. How their temperature-dependent polarization behaves remains largely unknown. In this work, we attempt to answer this question using 2D pyroelectric oxides. We show that quasi-2D oxides down to a unit cell thickness can be epitaxially grown on perovskite substrate by the molten salt-assisted quasi-van der Waals epitaxy following a screw-dislocation driven mechanism. We experimentally demonstrate switchable in-plane photo-ferroelectricity and thickness-dependent pyroelectricity. Weakly-coupled interface allows us to reveal that electron-phonon renormalization leads to the observed dimensionality effect. Harnessing the dimensionality effect in pyroelectricity via novel epitaxy strategies could promote their applications in uncooled infrared cooling and thermal energy harvesting.

The absorption spectrum for pristine, undoped graphene is constant at 2.3 % across the visible to near-infrared (VIS-NIR) region of the electromagnetic spectrum. When these assumptions are relaxed, graphene can exhibit a strongly nonlinear optical response. Here, we explore the optical properties of graphene that has been integrated with LaAlO₃/SrTiO₃ nanostructures. Nanojunction devices formed at the LaAlO₃/SrTiO₃ interface enable large (~10⁸ V/m) electric fields to be applied to graphene over a scale of ~10 nm. Nanoscale gating of the graphene in this way causes it to interact strongly with light, producing significant broadband THz emission via difference frequency mixing as well as a sum-frequency generated (SFG) response. Unexpectedly sharp spectral features show >99.99% extinction of the response. Extinction features are observed in both the VIS-NIR and SFG response, and can shift in frequency depending on the nanojunction bias and/or linear polarization. These surprisingly strong optical nonlinearities help to probe the fundamental response of nanostructured graphene, and open the way for future exploitation in optical devices.

Complex-oxide materials exhibit a plethora of physical properties desirable for next-generation electronics, photonics, and quantum devices not present in other functional materials such as Si, III-V, and III-N. These properties may be present in bulk form or at a heterointerface, and may be created or enhanced by application of strain. However, to date, only epitaxial methods have been used to create various heterostructures with varying strain states. Unfortunately, epitaxial methods have strict limitations, requiring closely lattice mismatched and similar crystalline structure, which prevent unrestricted manipulation, integration, and utilization of these materials. Thus, the design space for heteroepitaxial systems are limited. Additionally, the range of strain that can be imparted by epitaxy is fixed by pseudomorphic conditions. Finally, epitaxial growth conditions, which typically occurs at high temperatures, preclude integration of materials that are less stable in such environments. This has been a consistent bottleneck preventing universal integration of dissimilar materials to date.
To alleviate these restrictions, we present methods to produce freestanding complex-oxide membranes allowing artificial creation of complex-oxide heterostructures with vastly different crystalline structure, lattice, and orientation, such as perovskite, spinel, and garnet films. The first lift-off method, remote epitaxy, allows epitaxial seeding of the substrate through monolayers of graphene, which can then be easily exfoliated due to the van der Waals nature of graphene. Using this method, we demonstrate freestanding single-crystalline perovskite SrTiO₃ (STO), spinel CoFe₂O₄ (CFO), and garnet Y₃Fe₅O₁₂ (YIG) membranes which can be freely bent and stacked to create artificial heterostructures.
Although this method can be applied to a wide range of materials with ionicity of the atomic bonds, some complex-oxide materials, such as piezoelectric PMN-PT, are grown by sputtering, which damages the graphene due to the harsh plasma ambient. For these materials, we found that by choosing a substrate with a specific lattice misfit, the epitaxial layer can be precisely separated from the substrate using a high stress Ni layer. The mechanism, upon careful TEM analysis, showed that a periodic array of edge dislocations as well as concentrated stress at the epilayer/substrate interface allows for atomically precise separation.
Finally, we demonstrate artificial heterostructures consisting of CFO/PMN-PT as a strain-mediated magnetoelectric composite, where due to the freestanding nature of the device, exhibited an order of magnitude higher magnetoelectric coupling response compared to films clamped to the substrate. Moreover, magnetostatic and magnetoelastic coupling was observed by stacking CFO/YIG membranes together, and more importantly, we found that it is possible to tune the Fermi-level of graphene by sandwiching it between YIG and CFO. We hope such demonstration of artificial 3D-3D and 2D-3D heterostructures will allow to significantly expand the possible permutations of materials and allow experimental realization of structures not possible by epitaxial means alone.

Layered two-dimensional (2D) materials interact primarily via van der Waals bonding, which has created new opportunities for heterostructures that are not constrained by epitaxial growth [1]. However, it is important to acknowledge that van der Waals interactions are not limited to interplanar interactions in 2D materials. In principle, any passivated, dangling bond-free surface interacts with another via non-covalent forces. Consequently, layered 2D materials can be integrated with a diverse range of other materials, including those of different dimensionality, to form mixed-dimensional van der Waals heterostructures [2]. Furthermore, chemical functionalization provides additional opportunities for tailoring the properties of 2D materials [3] and the degree of coupling across heterointerfaces [4]. In order to efficiently explore the vast phase space for mixed-dimensional heterostructures, our laboratory employs solution-based additive assembly [5]. In particular, constituent nanomaterials (e.g., carbon nanotubes, graphene, transition metal dichalcogenides, black phosphorus, boron nitride, and indium selenide) are isolated in solution, and then deposited into thin films with scalable additive manufacturing methods (e.g., inkjet, gravure, and screen printing) [6]. By achieving high levels of nanomaterial monodispersity and printing fidelity [7], a variety of electronic and energy applications can be enhanced including photodetectors [8], optical emitters [9], and lithium-ion batteries [10-12]. Furthermore, by integrating multiple nanomaterials into heterostructures, unprecedented device function is realized including anti-ambipolar transistors [13], gate-tunable photovoltaics [14], and neuromorphic memtransistors [15]. In addition to technological implications for electronic and energy technologies, this talk will explore several fundamental issues including band alignment, doping, trap states, and charge/energy transfer across previously unexplored mixed-dimensional heterointerfaces [16].

[1] X. Liu, et al., Advanced Materials, 30, 1801586 (2018).
[2] D. Jariwala, et al., Nature Materials, 16, 170 (2017).
[3] C. R. Ryder, et al., Nature Chemistry, 8, 597 (2016).
[4] S. H. Amsterdam, et al., ACS Nano, 13, 4183 (2019).
[5] J. Zhu, et al., Advanced Materials, 29, 1603895 (2017).
[6] G. Hu, et al., Chemical Society Reviews, 47, 3265 (2018).
[7] J. Kang, et al., Accounts of Chemical Research, 50, 943 (2017).
[8] J. Kang, et al., Advanced Materials, 30, 1802990 (2018).
[9] C. Husko, et al., Nano Letters, 18, 6515 (2018).
[10] K.-S. Chen, et al., Nano Letters, 17, 2539 (2017).
[11] W. J. Hyun, et al., ACS Nano, 13, 9664 (2019).
[12] A. C. M. de Moraes, et al., Advanced Functional Materials, 29, 1902245 (2019).
[13] V. K. Sangwan, et al., Nano Letters, 18, 1421 (2018).
[14] D. Jariwala, et al., Nano Letters, 16, 497 (2016).
[15] V. K. Sangwan, et al., Nature, 554, 500 (2018).
[16] S. B. Homan, et al., Nano Letters, 17, 164 (2017).

Technological fashion of electronics has recently changed from device miniaturization to flexible device and 3-dimensional heterogeneous integration. For the flexbile devices, organic or polymeric materials have been generally used, but inorganic semiconductor is ultimately desirable in terms of high performance, long lifetime, and harsh environment tolerance. Nonetheless, the rigid, brittle characteristics of semiconductor has limited its utilization in the flexible device applications.
Graphene-inserted growth is promising for tranferrable device fabrication because the surface of graphene has no dangling bonds. The van der Waals interfaces of overlayer/graphene/wafer enable mechanical exfoliation of overlayer from wafer via a facile sticky scotch tape technique.[1] Using the graphene interlayer, there are two epitaxy regimes, which are van der Waals and remote epitaxies,[2,3] that can be classified by the role of graphene. In addition, the growth of wire or rod-shaped semiconductor crystallites is promising for deformable device applications because of the spatially separate geometry.[4]
This talk introduces the aforementioned epitaxies for transferable flexible optoelectronics. We begin discussing the role of graphene or underlying wafer how they dictate the epitaxial relation with overlayer. As the first section, the van der Waals epitaxy of InAs nanowires on graphene is presented. Use of thinnest substrate of single-layer graphene,[5] fabrication of InAs/graphene/InAs double heterostructures using suspended graphene,[6] and position-controlled growth,[3] all of which were achieved through the van der Waals epitaxy, are presented. Then, as the second section, the remote epitaxy is introduced. Hydrothermal remote epitaxy of ZnO microrods and metal–organic vapor-phase remote epitaxy of GaN microrods are presented.[7–9] We also discuss the remote heteroepitaxy of ZnO on GaN and GaN on Al₂O₃ for lattice less- and far-mismatched remote epitaxy, respectively.[8,9] The transfer of GaN heterostructured microrods of p–n junction with InGaN multiple quantum-wells is demonstrated for flexible microrod-light-emitting diode panel whose mechanical property is highly robust against crumpling/folding operations and repeated bending test.[9] Especially, we show the wafer recyclabilty for fabricating the deformable microrod-light-emiting diode panel over again on the same wafer in the remote epitaxy. We fianlly show how the principle of remote epitaxy, in which the limited thickness of graphene can penetrate the bonding feature of underlying wafer to the surface of graphene for epitaxial relationship, can be applied for site-selective growth.[10] The challenges and opportunities of epitaxies on/across graphene are discussed toward future flexible, transferable displays technology.

References
[1] Kim, Cruz, Lee et al. “Remote epitaxy through graphene enables two-dimensional material-based layer transfer” Nature 544, 340 (2017).
[2] Chung et al. “Transferable GaN layers grown on ZnO-coated graphene layers for optoelectronic devices” Science 330, 655 (2010).
[3] Hong and T. Fukui “Controlled van der Waals heteroepitaxy of InAs nanowires on carbon honeycomb lattices” ACS Nano 5, 7576 (2011).
[4] Lee et al. “Flexible Inorganic Nanostructure Light-Emitting Diodes Fabricated on Graphene Films” Adv. Mater. 23, 4614 (2011).
[5] Hong et al. “van der Waals epitaxy of InAs nanowires vertically aligned on single-layer graphene” Nano Lett. 12, 1431 (2012).
[6] Hong et al. “Van der Waals Epitaxial Double Heterostructure: InAs/Single-Layer Graphene/InAs” Adv. Mater. 25, 6847 (2013).
[7] Jeong, Min et al. “Remote homoepitaxy of ZnO microrods across graphene layers” Nanoscale 10, 22970 (2018).
[8] Jeong, Min et al. “Remote heteroepitaxy across graphene: Hydrothermal growth of vertical ZnO microrods on graphene-coated GaN substrate” Appl. Phys. Lett. 113, 233103 (2018).
[9] Jeong, J. Cha, Q. Wang et al. (unpublished)
[10] Jeong et al. (unpublished)

Two-dimensional (2D) materials, such as graphene and transition metal dichalcogenides (TMDs) MX2 (where M = many transition metals; X = S, Se, Te), have revealed new physics and demonstrated potential for practical applications. In recent years, interest in layered van der Waals materials has expanded to quasi-one-dimensional (1D) structures. Unlike the layered MX2 materials that yield 2D layers upon exfoliation, the transition metal trichalcogenides (TMTs) MX3 contain 1D motifs, i.e. atomic threads, that are weakly bound together by van der Waals forces [1-2]. In this invited talk, we will review synthesis methods, properties and possible device applications of this unique family of quasi-1D TMT material systems. As a consequence of their structures, the exfoliation or growth of MX3 crystals results in nanowires and nanoribbons rather than 2D layers. Some of TMTs can be considered as truly 1D materials while others, which have weaker covalent bonds in directions perpendicular to the atomic threads, can be considered as mixed dimensional structures. One can also fabricate heterostructures consisting of 1D and 2D van der Waals materials [1-2]. We have discovered that some of these quasi-1D nanomaterials reveal an exceptionally high current density. For example, quasi-1D TaSe3 nanowires capped with quasi-2D h-BN layers have a breakdown current density exceeding JB~10 MA/cm2, which is larger than what can be sustained by the state-of-the-art Cu interconnects. In a recent contribution, we reported uncapped ZrTe3 nanoribbons with an even more impressive breakdown current density of ~100 MA/cm2, which is more than an order of magnitude larger than the value obtained in DC testing of Cu wires [3]. We have used low-frequency noise (LFN) spectroscopy to investigate carrier recombination in such materials and verify reliability of the van der Waals interconnects [4-5]. It was found that LFN in ZrTe₃ reveals conventional 1/f behavior near room temperature (f is frequency). However, at low temperature it is dominated by the Lorentzian bulges of the generation–recombination noise at low temperatures, which is unusual for metals. Unexpectedly, the corner frequency of the observed Lorentzian peaks revealed a strong sensitivity to the applied bias. This dependence on electric field was explained by the Frenkel–Poole effect in the scenario where the voltage drop happens predominantly on the defects, which block the quasi-1D conduction channels. The obtained results reveal the potential of quasi-1D/quasi-2D materials and heterostructures for applications in future ultimately downscaled interconnects and device technologies.

This work was supported, in part, by the National Science Foundation (NSF) through the DMREF: Collaborative Research (UCR – Stanford): Data Driven Discovery of Synthesis Pathways and Distinguishing Electronic Phenomena of 1D van der Waals Bonded Solids, and by the Semiconductor Research Corporation (SRC) contract 2018-NM-2796: One-Dimensional Single-Crystal van-der-Waals Metals: Ultimately-Downscaled Interconnects with Exceptional Current-Carrying Capacity and Reliability.

[1] M. A. Stolyarov, et al., Breakdown current density in h-BN-capped quasi-1D TaSe₃ metallic nanowires: prospects of interconnect applications, Nanoscale, 8, 34, 157 74 (2016).
[2] G. Liu, et al., Low-frequency electronic noise in quasi-1D TaSe₃ van der Waals nanowires, Nano Letters, 17, 377 (2017).
[3] A. Geremew, et al., Current carrying capacity of quasi-1D ZrTe₃ van der Waals nanoribbons, IEEE Electron Device Letters, 39, 735 (2018).
[4] A. K. Geremew, et al., Unique features of the generation–recombination noise in quasi-1D van der Waals nanoribbons, Nanoscale, 10, 19749 (2018).
[5] T. A. Empante, et al., Low resistivity and high breakdown current density of 10 nm diameter van der Waals TaSe3 nanowires by chemical vapor deposition, Nano Letters, 19, 4355 (2019).

Crystallographic defects, such as dislocation, are well known to strongly affect material’s physical properties. Electron-hole recombination mediated via dislocations (e.g. threading dislocation) is one of the predominant loss mechanisms for the sub-optimum performance in conventional semiconductors devices. However, how dislocation impacts its carrier dynamics in the ‘defects-tolerant’ halide perovskite is largely unknown. Stimulated by the successful pioneer work of remote homoepitaxy of semiconductor GaAs^{1, 2}, researchers have also been achieved (remote heteroepitaxy) in the systems of AlN film on sapphire³, copper film on sapphire⁴, and ZnO film on GaN⁵. In our study, we synthesize epitaxial halide perovskite with controlled dislocation density via a remote heteroepitaxy approach using polar substrates (NaCl and CaF₂) coated with graphene. Density functional theory calculations have revealed the structure and magnitude of the incompletely screened electrostatic potential from the polar substrates, supporting the remote epitaxy in the present case. The regulated film-substrate interactions have further reflected themselves in controlling the wavelengths of the ferroelastic domains. Molecular-dynamics simulations reveal the kinetic process during remote epitaxy. Comparing to the ionic epitaxy with high dislocation density (both misfit and threading), the film grown via remote shows much enhanced photoluminescence intensity and increased carrier lifetime. Our successful demonstration of remote epitaxy in halide perovskite provides an approach to develop free-standing halide perovskite film with reduced dislocation density. More importantly, dislocations and their impacts on carrier dynamics and device performance in halide perovskite have to be recognized and scrutinized.

1. Kim Y, Cruz SS, Lee K, Alawode BO, Choi C, Song Y, et al. Remote epitaxy through graphene enables two-dimensional material-based layer transfer. Nature 2017, 544: 340.
2. Kong W, Li H, Qiao K, Kim Y, Lee K, Nie Y, et al. Polarity governs atomic interaction through two-dimensional materials. Nat Mater 2018, 17: 999–1004.
3. Qi Y, Wang Y, Pang Z, Dou Z, Wei T, Gao P, et al. Fast Growth of Strain-Free AlN on Graphene-Buffered Sapphire. J Am Chem Soc 2018, 140(38): 11935-11941.
4. Lu Z, Sun X, Xie W, Littlejohn A, Wang G-C, Zhang S, et al. Remote epitaxy of copper on sapphire through monolayer graphene buffer. Nanotechnology 2018, 29(44): 445702.
5. Jeong J, Min K-A, Kang BK, Shin DH, Yoo J, Yang WS, et al. Remote heteroepitaxy across graphene: Hydrothermal growth of vertical ZnO microrods on graphene-coated GaN substrate. Appl Phys Lett 2018, 113(23): 233103.

The properties of group III-Nitrides (III-N) such as a large direct bandgap, high melting point, and high breakdown voltage make them very attractive for optoelectronic applications. However, conventional epitaxy on SiC and sapphire substrates results in strained and defective films with consequently poor device performance. In this work, by studying the nucleation of GaN on graphene/SiC by MOVPE, we unambiguously demonstrate the possibility of remote van der Waals epitaxy. By choosing the appropriate growth conditions, GaN crystals can grow either inplane misoriented or fully epitaxial to the substrate. The adhesion forces across the GaN and graphene interface are very weak and the micron-scale nuclei can be easily moved around. The combined use of x-ray diffraction and transmission electron microscopy demonstrate the growth of stress-free and dislocation-free crystals. The high quality of the crystals was further confirmed by photoluminescence measurements. First principles calculations additionally highlighted the importance of the polarity of the underlying substrate. This work lays the first brick towards the synthesis of high quality III-N thin films grown via van der Waals epitaxy.

Developing efficient bifunctional catalysts for overall water splitting that are earth-abundant, cost-effective, and durable is of considerable importance from the practical perspective to mitigate the issues associated with precious metal-based catalysts. In the present study, we introduce a heterostructure comprising perovskite oxides (La_0.5Sr_0.5CoO_3–δ (LSC)) and transition metal dichalcogenides (TMDs, MoSe₂) as an electrochemical catalyst for overall water electrolysis. Interestingly, formation of the heterostructure of LSC and MoSe₂ (LSC&MoSe₂) induces a local phase transition in MoSe₂, 2H to 1T phase, owing to electron transfer from Co to Mo, and the semiconducting MoSe₂ transforms to the metallic phase. In addition, LSC becomes more electrophilic, and Co-O and Co-OH bonds are favored owing to partial oxidation of the Co cation due to the electron transfer. Together with the electrochemically active nature of 1T MoSe₂ and the increased amount of Co-O and Co-OH bonds in LSC, the electrochemical activities are significantly improved for both hydrogen evolution reaction and oxygen evolution reaction. In the overall water splitting operation, LSC&MoSe₂ showed excellent stability at the high current density of 100 mA cm⁻² over 1,000 h, which is exceptionally better than the stability of the state-of-the-art Pt/C || IrO₂ couple.

Van der Waals (vdW) surfaces of 2D materials such as graphene and BN are attractive for growth of GaN and other III-nitride films and device structures. The weak inter-planar vdW bonding between 2D layers allows for easy mechanical separation of III-nitride films from the growth substrate for lift-off, transfer and integration on to arbitrary substrates. This approach of vdW lift-off is suitable for transfer of delicate single devices, like T-gate high electron mobility transistors (HEMT), as well as, wafer sized films. However, to obtain high performance devices, control over nucleation and epitaxy on 2D layers is necessary for high quality films. Additionally, to enable a simple transfer process strain and adhesion between the film and 2D layer must be balanced. In this paper, we describe the growth of high quality GaN and AlGaN/GaN HEMT structures on 2D BN. The effects of nucleation, III-nitride epitaxial growth process and BN morphology will be discussed demonstrating how they effect material quality, properties, strain and adhesion. We will then discuss transfer schemes for individual AlGaN/GaN HEMTs and large area membranes transferred to both flexible polymer substrates and a variety of rigid substrates. With access to GaN and AlGaN/GaN two dimensional electron gas (2DEG) membranes on flexible substrates, we investigate the effects of strain on basic materials properties from compressive to tensile. This is particularly powerful for probing the effects of stain on the electrical properties, like sheet carrier density and mobility, in the polarization induced AlGaN/GaN 2DEG. In transferred devices, we demonstrate exceptional transport properties with mobility > 2,000 cm²/Vs and device performance with G_M of 300 mS/mm and f_T and f_max > 34 GHz and 75 GHz under stain while bending. Lastly, we will cover vdW and adhesive bonding to high and low thermal conductivity substrates and the effects on device performance and self-heating. This paper will highlight the great potential of vdW epitaxy and lift-off for transfer and integration of GaN on to various platforms.

Developing new types of nanoscale electronic devices, such as switches or storage elements, often requires creating both a functional nanostructure with precisely controlled properties and a well-defined connection between the nanostructure and the larger-scale circuit. Achieving sufficient control of nanostructures and their contacts is a key step that generally involves integrating different types of materials as well as bridging between length scales. For example, in Si or Ge-based circuits, metal silicides and germanides are commonly used in low-resistance contacts, while functional nanostructures can involve combinations of semiconductors with insulators and metals. Here we consider nanowires, which are especially favorable geometries for connecting nano and macro length scales, and discuss the opportunities for creating complex nanostructures within individual nanowires. In particular, we focus on silicide and germanide nanocrystals embedded within Si and Ge nanowires.
We start by illustrating how silicide and germanide nanostructures can be formed in silicon nanowires grown by the vapor-solid-liquid (VLS) mechanism. These nanocrystals are initially created by adding the appropriate metal to the liquid droplets that catalyze the nanowire growth. We show that solid silicide or germanide nanocrystals form in the liquid and have freedom to move and rotate until a low-energy interface with the nanowire is found. The crystal attaches to the nanowire, nanowire growth is then continued, and the nanocrystal becomes encapsulated within the nanowire. The process can be repeated to form multiple inclusions that are useful for contact formation as well as for modifying nanowire properties in new ways. In situ environmental TEM clarifies the sequence of phases and structures present at the atomic level during this complex process.
After the nanocrystal-nanowire contact is first made, we find that certain types of silicide and germanide nanocrystals remain attached while others break away without forming a permanent contact. Only the nanocrystals that remain attached to their nanowire can be incorporated by continued growth of the nanowire. We have examined the factors that determine nanocrystal adhesion to the nanowire and suggest that the outcome depends on the symmetry of the contact interface and hence the crystal structures of both materials. This restricts the range of nanocrystal/nanowire combinations possible. To expand the range of materials that can be encapsulated, we explore the use of phase transformations within the nanocrystal. As an example, NiGe does not attach to the growth surface of a Ge nanowire so is difficult to incorporate. But after growing a NiGe nanocrystal we can add Si to transform the nanocrystal to a silicide and this structure does attach and incorporate.
We suggest that the variety of nanostructures with incorporated nanocrystals that can be made using such reaction schemes potentially increases the opportunities for designing specific electronic and contact properties for nanostructured device applications.

Inorganic micro light-emitting diode (micro-LED) arrays are emerging as one of the most promising light emitters for next-generation display technologies and ultrahigh-resolution optogenetic light source arrays. An inorganic micro-LED can exhibit all of the important characteristics of an organic LED, including fast response time and a high contrast ratio, but with significantly improved resolution, brightness, efficiency, and lifetime due to its increased carrier mobilities, radiative recombination rates, and long-term stability. In free-standing and ultrathin form, the applicability of micro-LEDs can be further expanded to include various wearable, medical, and implantable devices, which require conformal contact on human skin or organs with minimal discomfort and stress. By creating an array of free-standing inorganic microstructure devices that are orders of magnitude smaller than the bending radius, mechanical flexibility can be accommodated. However, the inherent rigidity of inorganic materials and difficulty separating inorganic thin films from their single crystal growth substrates represent substantial challenges to the fabrication of freestanding and ultrathin inorganic LEDs. To resolve this problem, growths of inorganic semiconductor nanostructures and thin films on graphene substrates have recently been proposed, since graphene has great scalability and extremely thin layered hexagonal lattice structure as an excellent substrate for GaN growth. Moreover, the inorganic semiconductors prepared on large-area graphene can be transferred easily to or grown on elastic substrates to meet the flexibility demand. Here, we suggest a method of fabricating ultrathin, high-resolution inorganic micro-LED arrays based on individually addressable GaN microdisk LED arrays grown on graphene dots.
Here, we report on the fabrication and EL characteristics of free-standing and ultrathin, individually addressable GaN microdisk LED arrays grown on graphene dots. GaN microdisks were prepared by epitaxial lateral overgrowth on patterned graphene microdots on SiO₂/Si substrates using MOVPE. After preparing the GaN microdisk arrays, p-GaN and InGaN/GaN multiple quantum well, and n-GaN layers were heteroepitaxially grown on the surface of the GaN microdisks. Ultrathin layers composed of GaN microdisk LED arrays on graphene dot were prepared by coating a polyimide layer and lifting-off the entire layers from the substrate. Then, single-walled carbon nanotubes (SWCNTs)/Ni/Au and SWCNTs/Ti/Au multiple electrode lines were formed on the top and bottom surface of GaN microdisk arrays in an aligned manner and crossing each other. The electrical and optical characteristics of the individually addressable GaN microdisk array on graphene dots were investigated by measuring their current-voltage characteristics curves and EL characteristics at various bending conditions. We also confirmed that the ultrathin micro-LED array worked reliably under flexible conditions and continuous operation mode.

Hybrid structures composed of conventional semiconductors (3D materials) and atomically thin two-dimensional (2D) materials have offered opportunities of recyclable device manufacturing, novel functionalities based on carrier transfer and exciton transport, and understanding natures of van der Waals gap. However, nucleation of 3D materials on 2D materials has not been understood straightforwardly due to absence of surface dangling bonds on 2D material though interaction the overgrown layer and 2D layer as a substrate has been explained by van der Waals interaction. Recent observation performed by the presenter’s team revealed that surface energy landscape is a key for nucleation of 3D materials on 2D materials. Increase in surface energy on 2D layer enhances nucleation probability of 2D materials significantly.
In the presentation those will be discussed several key aspects of nucleation of 3D materials on 2D materials, such as a few strategies to enhance surface energy on 2D layer, insights of the nucleation strategy on remote epitaxy, and demonstration of devices based on 3D/2D heterostructures.

Mixed-dimensional heterojunctions, such as zero-dimensional organic molecules deposited on two-dimensional transition metal dichalcogenides (TMDs), exhibit unique interfacial effects that modify the properties of the individual constituent layers. To better understand the effect of changing molecular orbital energy on the heterojunction properties, we report here a systematic study of metallophthalocyanine (MPc) – MoS₂ (M = Co, Ni, Cu, Zn, H₂) heterojunctions using optical absorption, Raman spectroscopy, low temperature photoluminescence, variable temperature electrical measurements, and density functional theory. Notable phenomena include the emergence of heterojunction-specific optical absorption transitions, strong Raman enhancement, and quenching of defect emission that depend on phthalocyanine metal center identity. Temperature-dependent electrical noise measurements on field-effect transistors provide further insight into the properties of these mixed-dimensional heterojunctions. These results highlight the tunable nature of metal-organic-TMD van der Waals interfaces and the importance of the organic architecture and electronic structure in designing mixed-dimensional heterojunctions for optical and electronic applications.

In recent years, several transition metal dichalcogenides (TMDs) have been found to exhibit intrinsic ferromagnetic properties near the monolayer limit [1, 2]. However, unlike most ferromagnetic TMDs studied so far which only show weak ferromagnetic properties at extremely low temperature, the monolayer 1T polytype of manganese diselenide (MnSe₂) have been found to display long-range magnetic ordering and high magnetic moments (3μB per unit cell), as well as a high Curie temperature of 250K, tunable up to 375K via 5% biaxial strain [3]. Hence, this material has a high potential for future applications in low energy magnetic switching and non-volatile logic data storage [4]. In this study, two-dimensional heterostructures of manganese (IV) selenide are grown via chemical vapor deposition (CVD) on epitaxial graphene (EG) synthesized on 6H silicon carbide (SiC) substrates.

In our experiments, three different approaches for the CVD growth of the heterostructure have been investigated to grow MnSe₂. In our first approach, powdered manganese (IV) oxide (MnO₂) and selenium powder (Se) reactants were used. In the second method, manganese acetate (Mn(CH₃CO₂)₂) powder was used to improve manganese nucleation on EG. In our final method, δ-phase MnO₂ is electrodeposited on epitaxial graphene using a 0.1M solution of manganese acetate in a three-electrode electrochemical cell. Selenium powder is subsequently used in the selenification of δ-phase MnO₂ using the CVD method.

Characterization via Raman spectroscopy demonstrated the presence of MnSe₂ for all three approaches, which was also confirmed by SEM, AFM and EDX analysis. Although the characteristic A_g Raman peak of MnSe₂ was visible at 269 cm^-1 for the first method, the very low intensity (almost 40% of the control FTA peak of the substrate at 203 cm^-1) of the peak indicated extremely low grain size and yield density. This conclusion was in agreement with the AFM and SEM data, which showed a sparse distribution of small growth particles (around 0.1 μm). Characterization of the second method displayed similar results, except the A_g peak intensity displayed a relative increase of around 30%. The AFM and SEM data also showed a roughly 200% increase in grain size, which can be attributed to the relatively lower sublimation temperature of manganese acetate.

In the Raman data of the final approach, however, both characteristic E_g and A_g peaks of MnSe₂ are clearly visible (at 145 cm^-1 and 267 cm^-1 respectively) with extremely high peak intensity (around 400% of the mentioned substrate control peak), implying significantly higher MnSe₂ yield. The AFM and SEM data also demonstrated significantly high deposition density along graphene step edges. This remarkable improvement can be credited to the electrodeposited MnO₂ promoting manganese adhesion and selenification of the sample. This procedure shows great promise in the synthesis of single-crystal MnSe₂ heterolayer on epitaxial graphene. Samples grown using this mechanism has significant potential application in the field of spintronics on a wafer scale.

Reference:
[1] Y. Ma, Y. Dai, M. Guo, C. Niu, Y. Zhu and B. Huang, ACS Nano, 2012, 6, 1695–1701
[2] Y. Zhou, Z. Wang, P. Yang, X. Zu, L. Yang, X. Sun and F. Gao, ACS Nano, 2012, 6, 9727–9736
[3] M. Kan, S. Adhikari and Q. Sun, Phys. Chem. Chem. Phys. 16 (2014) 4990
[4] D. O’Hara, T. Zhu, A. Trout, A. Ahmed, Y. Luo, C. Lee, M. Brenner, S. Rajan, J. Gupta, D. McComb, and R. Kawakami, Nano Letters 2018 18 (5), 3125-3131

Growth of semiconducting materials on two-dimensional (2D) materials, such as graphene, hexagonal boron nitride (h-BN), and transition metal dichalcogenides, has opened up novel opportunities for controlled functionality, economical device production, and flexible hybrid device manufacturing. A key method of semiconductor growth on 2D materials is van der Waals (vdW) epitaxy. In vdW epitaxy, nucleation of semiconductors on 2D layers is the main bottleneck due to the absence of surface dangling bonds on 2D layers. Induced out-of-plane dipole moments on 2D layers has been introduced to solve the nucleation issue. As demonstrated in 2018, out-of-plane dipole moments can be generated at the interfaces between different 2D materials. However, the screening effects of multiple 2D layers on the out-of-plane dipole have not been thoroughly explored. An atomically thin 2D monolayer does not provide enough screening of potential fluctuation in a substrate. Moreover, stacking multi-layers of atomically thin 2D materials can be still transparent to the potential in a substrate.
We studied the screening effects of graphene multi-layers on out-of-plane dipole moment for nucleation of germanium (Ge) on graphene/h-BN stack. Graphene/h-BN stacking enhances the out-of-plane dipole moment ~1,000 times compared to that of graphene or h-BN alone. The enhanced out-of-plane dipole moment induces a remarkably higher nucleation probability of Ge on graphene/h-BN stack compared to that on graphene. Furthermore, we explored the limitation of stacked layers of multi-graphene/h-BN to elucidate the screening effect of out-of-plane dipole moment on Ge nucleation. VdW epitaxy studies and calculation of out-of-plane dipole moment on the surface of the top graphene layer along with a number of graphene layers stacked on h-BN were systematically conducted. Electron microscopy and thin film field effect transistor characterizations were performed to correlate structural and electrical properties of Ge grown on multi-layer graphene/h-BN. In addition, insight of semiconductor device manufacturing with multi-layers of 2D materials will be discussed.

Interaction of incident light with periodic arrays of three dimensional (3D) metal-dielectric composited nanoarrays provide remarkable opportunities to harness light in a way that cannot be obtained with conventional optics. Diverse types of 3D or quasi-3D plasmonic nanoarrays with tailored feature shapes, sizes and configurations have been explored for a broad range of light-driven sensors and actuators, including imagers, bio-sensors, lasers and antennas. However, their practical applications remain challenged by a lack of effective manufacturing methodology. To bridge this gap, we developed a new concept of methodology that enables physical separation of precisely engineered quasi-3D plasmonic nanoarrays from their donor fabrication wafer and then transfer to a preferred foreign substrate in a defect-free manner that allows the donor wafer can be repeatedly recycled, offering a major cost- and time-saving factor in the manufacturing scheme. Unlike any of existing approaches, the entire process of this method exclusively occurs in distilled (DI) water at room temperature without the need of further chemical, thermal or mechanical treatments, and which thereby can substantially extend the types of receiver substrate to nearly arbitrary materials and structures. This approach provides versatility and modular scalability to arrange various classes of quasi-3D plasmonic nanoarrays in lateral and vertical configurations, offering a unique route to generate heterogeneous material compositions, complex device configurations and tailored functionalities. Comprehensive experimental, computational and theoretical results reported here reveal the essential design features of this method and, taken together with implementation of automated equipment, provide a technical guidance for manufacturable controllability and scalability. Pilot deterministic assembly of specifically engineered quasi-3D plasmonic nanoarrays with an industrial-grade hybrid imager, such as a mid-wavelength infrared type-II superlattice (MWIR-T2SL)-based hybrid pixel detector (HPD) successfully yields the enhancement of the detection performances and functionalities which otherwise cannot be achieved by conventional counterpart systems. The established 3D nanoassembly methodology has a great potential to extend the application of the deterministically assembled plasmonic nanoarchitectures for other types of imagers, antennas and biosensors.

The structural quality and homogeneity of the epitaxial film is critical for the performance of many electronic and optoelectronic devices. The epitaxial films are mainly grown by either vapor-phase epitaxy (VPE) or liquid-phase epitaxy (LPE). The VPE technique is capable to grow multilayer and quantum well structures with controlled deposition process down to atomic scale, so it is favored in industry and developed rapidly during the past decades. However, growth of oxides and nitrides via VPE suffers from defect-rich films containing secondary phases and numerous grain boundaries. LPE, a growth process where a film of crystalline material is deposited from a supersaturated solution onto a single-crystal substrate, yields much higher crystalline quality of epitaxial films than VPE. The supersaturation for LPE is extremely low thus enabling the growth near thermodynamic equilibrium. Therefore, epitaxial films via LPE exhibit many advantages like highest structural perfection, automatic controlled stoichiometry, homogeneous dopant incorporation, flat surfaces and possibility of upscaling and mass production.

In this study, we use a molten salt-assisted LPE technique to prepare a single crystalline epitaxial low-dimensional ferroelectric complex oxide, Dion-Jacobson quasi-2D layered oxide, on LaAlO₃ (LAO) substrates. We present an in-plane polarization switching in this quasi-2D oxide. We find that the epitaxy growth follows a screw-dislocation driven mechanism. X-ray diffraction analysis confirms the epitaxial flakes with high crystalline quality. Our microscopy and second harmonic generation measurements show the presence of ferroelastic domains and lack of inversion symmetry, respectively, in the epitaxial film. Our quasi-2D oxide exhibits a switchable photovoltaic ferroelectric effect where the polarization is modified by above band gap megawatts nanosecond optical pumping. Our high quality of epitaxial quasi-2D oxide would renew attentions and provide a good insight of this traditional LPE technique. In this layered complex oxide, the coupling between the in-plane ferroelectricity and other physical properties, e.g. the switchable photo diode effect and the photon-induced domain switching, may enable versatile applications down to the atomic scale.

As the electronics scaling down, the interfacial properties such as dangling bonds or surficial reconstructions play a critical role in functionality and performance of devices and the crystallinity of the overlayer. In addition, large roughness and pin holes of the insulator layer usually lead to leakage current and degradation in resistive random-access memory due to unrecoverable electrical breakdown. In this study, to eliminate the lattice interaction from the underlying substrate, we present a facile and easy method to achieve van der Waals epitaxy of bismuth iodide (BiI₃) on the silicon wafer by using a self-assembly-monolayer of octadecyltrichlorosilane (OTS) as a buffer layer. As a result, the BiI₃ layer on the silicon wafer possesses considerably flatten surface and high crystallinity, and the resistive-switching devices based on the BiI₃/OTS/Si heterostructure demonstrated excellent resistive switching property with a very high on/off ratio of 10⁹, long term stability for data retention, high endurance to write/erase cycles and the capability of multistate information-storage. The resistive switching behaviors, the roughness, morphology, and crystallinity of the BiI₃ layer are systematically investigated by analyzing the current-voltage characteristics at different temperature, scanning electron microscope, atomic force microscope, X-ray photoemission spectroscopy (XPS) and x-ray diffraction patterns (XRD). Consequently, the resistive switching mechanisms is explained by formation and rupture of a conductive filament consist of metallic bismuth in the insulating layer as a result of the ion migration of iodine that changes the valence charge of BiI₃ under the electrical field. This paper will help the fabrication process in large scale and the integration of BiI₃ based resistive switching devices with the integrated circuit technology.

It has always been a challenge to fabricate devices at the nanoscale and analyse their properties, be it for the lack of equipment or the lack of techniques available. Defect free nanowhiskers attract a lot of attention as a potential candidate for interconnects. A concern regarding the resistivity of nanowires is their increased resistivity, which sometimes is 10 or 100 times larger than their bulk counterparts [1]. With this work, we are presenting a cleaner and a novel method which allows the formation of nanowhisker interconnects with near bulk resistivity facilitating a simple LASER lithography technique.

We adopt a bottom-up approach for device fabrication using what we call “Target Specific Lithography”. The whiskers which are grown by Molecular Beam Epitaxy are single crystals with a high aspect ratio and without any grain boundaries or defects [2]. These whiskers were transferred onto a Si substrate. A layer of resist is applied on this substrate and four contacts are designed in CAD and aligned with high precision so that it lies on top of an individual whisker. The contacts are exposed using a direct LASER writer. 100 nm Niobium is sputtered onto the lithographically patterned substrate. Upon liftoff, Niobium forms the contacts on the whiskers, allowing for the electrical transport measurement.

Electrical transport measurements performed on Au, Ag and Cu nanowhisker interconnects are demonstrated. The lengths of these whiskers are in the range of 18 μm to 40 μm, with an average diameter of 110 nm. The resistivity of gold at room temperature, for example, was found to be 2.20 ± 0.36 μΩcm, which is same as that of the bulk resistivity. Such results show that there is no size effect on the resistivity of the whiskers and that there are no impurities in the interface between the whisker and the contacts. We also have plans to conduct low-temperature measurements to understand the transport behaviour of the nanowhiskers as a function of temperature.

[1] G. D. Marzi, D. Iacopino, A. J. Quinn, and G. Redmond, ‘Probing intrinsic transport properties of single metal nanowires: Direct-write contact formation using a focused ion beam’, Journal of Applied Physics, vol. 96, no. 6, pp. 3458–3462, Sep. 2004.
[2] G. Richter, K. Hillerich, D. S. Gianola, R. Mönig, O. Kraft, and C. A. Volkert, ‘Ultrahigh Strength Single Crystalline Nanowhiskers Grown by Physical Vapor Deposition’, Nano Lett., vol. 9, no. 8, pp. 3048–3052, Aug. 2009.

The oxygen reduction reaction (ORR) is a key electrochemical reaction taking place at the cathode of various energy devices such as fuel cells and metal-air batteries. Heavily-loaded Pt catalysts are used to circumvent intrinsically sluggish kinetics of ORR, yet scarcity and high cost of Pt have triggered a recent drive toward alternative platinum group metal (PGM)-free catalysts, including non-precious metals, metal chalcogenides, metal-free, and single-atom catalysts. The single-iron-site catalyst surrounded by four nitrogen atoms (Fe–N₄) has emerged as one of the most promising electrocatalysts to replace Pt-based catalysts. To immobilize Fe–N₄ species on a conductive carbon support, however, high-temperature annealing for more than several hours is required, which induces severe aggregation of Fe metal ions along with its loose attachment on the carbon support resulting in low ORR activity and low stability. Significant deactivation is observed primarily caused by serious carbon corrosion or metal dissolution, which are common problems of carbon-supported catalysts in general because of intrinsically weak interaction of carbon with the active component. Thus, a high degree of graphitization of carbon supports is vital for the stability of a single atomic Fe–N₄ catalyst.
Carbon is one of the most common support materials for (electro)catalysts due to its high surface area, good conductivity, relative chemical inertness, and low cost. We also note that carbon is among the best microwave absorbers (susceptors) due to its high dielectric properties and short attenuation distance. Upon microwave irradiation, the carbon support can selectively absorb the radiation, keeping the catalyst precursor temperature low, and then heat is migrated from the carbon to the catalyst precursor. Because this selective and directional heat transfer mechanism dramatically reduces the chance of reaction between the catalyst precursors leading to agglomeration, and the entire process occurs in just a few seconds to minutes, the sintering of the catalyst particles is effectively mitigated. Instead, it causes strong attachment of the catalyst to the carbon surface. In this proposed new synthetic strategy, which we term “selective microwave annealing (SMA)”, there is no temperature limit and the reaction atmosphere [oxygen (O₂), nitrogen (N₂), and argon (Ar) gases] can be easily regulated, as in the conventional heating method. Furthermore, it does not require any additional microwave susceptor, such as those used in hybrid microwave annealing systems.
Thus, we applied this SMA technique utilizing the CNT support as a main microwave susceptor to fabricate atomic Fe–N₄ catalysts anchored strongly on CNT supports without Fe agglomeration. Structural characterizations by high-angle annular dark-field transmission electron microscopy (HAADF-STEM) combined with electron energy loss spectroscopy (EELS), X-ray absorption spectroscopy (XAS), and X-ray photoelectron spectroscopy (XPS) revealed that the single Fe atoms were uniformly dispersed in the form of Fe–N₄ and that a few-layer, graphene-like carbon structures were coated on the CNT surface in the selective microwave-annealed catalyst (MA-Fe-N/CNT), while severe agglomeration of Fe metal and amorphous carbon layers were observed for the thermally annealed catalyst (TA-Fe-N/CNT). As a result, the MA-Fe-N/CNT catalyst exhibited outstanding ORR activity with a half-wave potential of 0.92 V vs. reversible hydrogen electrode (V_RHE) and excellent stability both in alkaline and acidic media. The catalytic performance and stability were superior to TA-Fe-N/CNT and even exceeded those of commercial Pt/C catalysts. We also have demonstrated that this SMA technique is generally applicable to other carbon supports, such as reduced graphene oxide and carbon black. Finally, we also have verified a Na-air battery with the as-prepared MA-Fe-N/CNT catalyst operates as effectively as the device with a Pt/C catalyst.

For decades we’ve been hearing about the promise of printing electronics directly onto any surface. However, despite significant progress in the development of inks and printing processes, reports on fully, direct-write printed electronics continue to rely on excessive thermal treatments and/or fabrication processes that are external from the printer. In this talk, recent progress towards print-in-place electronics will be discussed; print-in-place involves loading a substrate into a printer, printing all needed layers, then removing the substrate with electronic devices immediately ready to test. A key component of these print-in-place transistors is the use of inks from various nanomaterials, including 1D carbon nanotubes (CNTs - semiconducting), 2D hexagonal boron nitride (hBN - insulating), and quasi-1D silver nanowires (AgNWs – conducting). Using an aerosol jet printer, these mixed-dimensional inks are printed into functional 1D-2D thin-film transistors (TFTs) without ever removing the substrate from the printer and using a maximum process temperature of 80 C. To achieve this, significant advancements were made to minimize the intermixing of printed layers, drive down sintering temperature, and achieve sufficient thin-film electrical properties. Devices are demonstrated on various substrates, including paper, and evidence of the potential for printing directly onto biological surfaces will be shown. From the versatile printed electronic thin films that have been developed, a diversity of biosensors are being pursued and will be discussed. With continued refinement of the inks and print processes, this print-in-place technique can bring the field of printed electronics closer to where it has been promised to go for so many years: load substrate, press print, remove functional sensors / circuit.

Solid state thermionic and thermoelectric modules can operate in power generation mode to convert heat to electricity, refrigeration, active cooling and thermal switching modes. They have potential application in waste heat recovery, solar thermal energy conversion and heat management and power generation of wearable electronics. Thermoelectric efficiency is an increasing function of material figure of merit, ZT. To obtain large ZT, high electronic conductance, large open circuit voltage (i.e. large Seebeck coefficient) and low thermal conductance is needed. The three parameters are interconnected and hence the optimization and improvement of ZT is a challenging task and requires careful control of electrons and phonons. Many of advanced thermoelectric materials are multi-phase compounds wherein thermal and electrical transport are determined not only by the average of the phases but also by the interfaces and the charge transfer between the phases. In this talk I will discuss several cases in which bulk samples are designed in which electronic transport is effectively 1D or 2D, while phonon transport is 3D and I will discuss the advantage of such structures and how thermoelectric figure of merit is enhanced using such strategy. Finally, I will discuss the thermal and electrical transport across 2D van der Waals heterostrcutures. The layers are weakly bonded resulting in extremely low thermal conductance. At the same time, layers could be mixed and matched to fine tune the electronic transport and to optimize the thermionic figure of merit.

In recent years, a van der Waals heterostructure device, stacking 2D architectures atomically with synergistic combinations of nanomaterials have attracted tremendous attention with many potential applications such as chemical sensors in environmental and safety monitoring, and medical diagnostics and biomedical health care systems [1,2]. Atomically thin 2D graphene and transition metal dichalcogenides (TMDs) have an extremely high surface-to-volume ratio which is the most critical parameter for chemical sensing applications. Here we have fabricated a heterostructure of molybdenum disulfide (MoS₂) nanoflower and epitaxial graphene on 6H silicon carbide (SiC) substrate for chemical sensing. We have combined the advantages of high sensitivity and fast response time of graphene with the high surface-to-volume ratio of MoS₂ nanoflower to develop an ultra-high sensitivity chemical sensor with improved selectivity.

Bilayer epitaxial graphene (EG) was synthesized by Si sublimation on 6H SiC, and MoS₂ nanoflowers were grown on graphene/SiC directly using metal-organic chemical vapor deposition (MOCVD). The growth methods of graphene and MoS₂ nanoflowers are discussed in detail elsewhere [3,4]. The structural and optical properties of the samples were investigated by scanning electron microscopy (SEM), atomic force microscope (AFM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, photoluminescence (PL), and cyclic voltammetry (CV). To fabricate the chemical sensor, e-beam lithography and reactive ion etching (RIE, CF₄) were used to prepare a simple pattern. Finally, the metal electrode (Ti/Au = 30/120 nm) was deposited using an e-beam evaporator. The prepared chemical sensor was tested with various gases such as 5 ppm of nitrogen dioxide (NO₂) and ammonia (NH₃), and ~1000 ppm of volatile organic compounds (VOCs) and showed superior chemical response and recovery at room temperature. It is evident that the prepared device is suitable for chemical sensing applications with high sensitivity and selectivity.

References: [1] W. Yang et. al. Inorg. Chem. Front. 3, 433(2016); [2] X. Zhou et. al. Chem. Rev. 115, 7944 (2015); [3] B. K. Daas et. al. J. Appl. Phys. 110, 113114 (2011); [4] J. Choi et. al. Nano Lett. 17, 1756 (2017).

Layered δ-phase MnO₂/graphene heterostructures, separated by van der Waals forces, have attracted significant interest due to their potential to combine the well-defined interlayer spacing of δ-MnO₂ with the high inherent conductivity of graphene[1]. However, the variety of MnO₂ polymorphs and Mn_xO_y structures that can arise during growth makes the synthesis of δ-phase MnO₂ and subsequent heterostructures challenging. Currently, δ-MnO₂ is primarily grown through hydrothermal synthesis from potassium permanganate (KMnO₄)[2]. This method suffers from scalability issues and an inability to deposit high uniformity films with precise layer control. In this work, we present an alternative electrochemical method for the synthesis of δ-MnO₂. Utilizing a 0.1M manganese acetate (Mn(CH₃CO₂)₂) solution, δ-MnO₂ was grown on epitaxial graphene (EG) synthesized on 6H silicon carbide (SiC) substrates in a three-electrode electrochemical cell. The resulting MnO₂ thin film was then characterized using Raman spectroscopy, energy-dispersive X-ray spectroscopy (EDX), atomic force microscopy (AFM), and scanning electron microscopy (SEM) to determine its phase and surface morphology. Scanning electron microscopy revealed the thin film was made up of micro platelets, with the smallest being approximately 70µm x 80µm. Raman spectroscopy confirmed the formation of δ-MnO₂ due to the presence of peaks at 577 cm^-1, 651cm^-1, and 133cm^-1[3]. An additional advantage of the δ-MnO₂/EG/SiC heterostructure is that it is lithography compatible and does not require a transfer process. To demonstrate this, we constructed a gas sensor based on the δ-MnO₂/EG/SiC heterostructure by depositing a metal electrode (Ti/Au = 30/120nm) on the δ-MnO₂ by e-beam evaporation. This gas sensor was then tested with 5ppm nitrogen dioxide (NO₂), 5ppm ammonia (NH₃), and an ~1000ppm Isopropyl alcohol (C₃H₈O)/Methanol (CH₃OH) mix, volatile organic compounds (VOCs), in an environmentally controlled chamber. In preliminary testing at room temperature the device displayed a significant response to NO₂ and NH₃ with ~400ms response time but displayed no response to VOCs. This result displays improved selectivity over sole graphene based chemical sensors [1]. This alternate synthesis method allows for a low temperature, low-cost route to the growth of lithography compatible δ-MnO₂ heterostructures for gas sensing, batteries, and spintronics.
References: [1]N. Joshi et al. Microchim. Acta 185, 213(2018);[2]V. Subramanian, et al. J. Phys. Chem B. 109, 20207-20214 (2005);[3] C. Zhu et al. Mater. Horiz. 4(3), 415-422. (2017)

Two-dimensional (2D) semiconducting transition metal dichalcogenides (TMDs) have been proposed for future high-performance flexible nanoelectronics [1]. However, high-quality chemical vapor deposition (CVD) growth of TMDs requires high temperatures >500°C. Thus, the fabrication process must rely on layer transfer processes for integration with temperature-sensitive flexible substrates.

Most conventional transfer techniques add a thin sacrificial polymer layer on top of the TMD. This stack is released from the growth substrate in an etching solution and picked up by the target substrate, but this technique can lead to polymer residues and damage or wrinkling of the TMD film [2,3]. An alternative method is to spin-coat the flexible substrate on the TMD before transfer, however this has its own limitations because all subsequent patterning would have to be done on the flexible substrate [4].

Here, we demonstrate a novel, alternative approach which allows us to do critical lithography steps on the rigid TMD growth substrate (SiO₂/Si), then transfer the TMD films together with their metal contacts onto the flexible substrate for further integration. This approach results in the shortest channel (sub-100 nm) and highest-performance TMD transistors on flexible substrates to date, using monolayer CVD-grown MoS₂. The “hybrid” fabrication/transfer method can also enable other devices, it can be used with other TMDs like WSe₂ or MoSe₂, and the TMD thickness can range from mono- or bilayers to thicker multilayers.

We first grow the TMDs by CVD on SiO₂/Si substrates [5-7] and perform lithography and lift-off for metal contacts, with contact spacing down to 50 nm. Then, we spin-coat 6 µm flexible polyimide (PI) on top and we release the PI with the embedded TMD and metal contacts in deionized water without any other chemicals. This minimizes damage and contamination of the TMDs, as verified by Raman spectroscopy and photoluminescence. We then flip the PI/TMD stack and deposit Al₂O₃ by atomic layer deposition, then pattern metal gates by lift-off, thus creating staggered TMD transistors.

We have thus realized flexible transistors with CVD-grown MoS₂, WSe₂ and MoSe₂, reaching field-effect mobility >30 cm²V^-1s^-1 for monolayer MoS₂, which is the highest to date on flexible substrates. In sub-100 nm MoS₂ transistors, we also obtain record high on-currents of ~600 µA/µm. Thermal simulations reveal that the intimate metal contacts assist heat spreading during device operation, enabling such high performance despite the low thermal conductivity of the PI substrate. We demonstrate that the devices are stable under tensile bending to a radius of 4 mm, with negligible changes in device characteristics. We also show preliminary data on flexible MoS₂ circuits and flexible vertical devices based on TMDs.

In summary, our novel fabrication method provides a scalable approach for the integration of various types of TMD devices on a single flexible substrate, and the contact patterning before transfer enables nanoscale dimensions, down to sub-100 nm channel lengths. This method also enables vertical device structures, e.g., resistive memory and optoelectronic devices such as photodetectors or solar cells.

1. D. Akinwande et al., Nature Communications 5, 5678 (2014).
2. A. Gurarslan et al., ACS Nano 8, 11522 (2014).
3. T. Zhang et al., ACS Applied Nano Materials 2, 5320 (2019).
4. S.M. Shinde et al., Advanced Functional Materials 28, 1706231 (2018).
5. K.K.H. Smithe et al., 2D Materials 4, 011009 (2017).
6. K.K.H. Smithe et al., ACS Applied Nano Materials 1, 572 (2018).
7. J. Chen et al., ACS Photonics 6, 787 (2019).

Layered two-dimensional (2D) materials interact primarily via van der Waals bonding, which has created new opportunities for heterostructures that are not constrained by epitaxial growth [1]. However, it is important to acknowledge that van der Waals interactions are not limited to interplanar interactions in 2D materials. In principle, any passivated, dangling bond-free surface interacts with another via non-covalent forces. Consequently, layered 2D materials can be integrated with a diverse range of other materials, including those of different dimensionality, to form mixed-dimensional van der Waals heterostructures [2]. Furthermore, chemical functionalization provides additional opportunities for tailoring the properties of 2D materials [3] and the degree of coupling across heterointerfaces [4]. In order to efficiently explore the vast phase space for mixed-dimensional heterostructures, our laboratory employs solution-based additive assembly [5]. In particular, constituent nanomaterials (e.g., carbon nanotubes, graphene, transition metal dichalcogenides, black phosphorus, boron nitride, and indium selenide) are isolated in solution, and then deposited into thin films with scalable additive manufacturing methods (e.g., inkjet, gravure, and screen printing) [6]. By achieving high levels of nanomaterial monodispersity and printing fidelity [7], a variety of electronic and energy applications can be enhanced including photodetectors [8], optical emitters [9], and lithium-ion batteries [10-12]. Furthermore, by integrating multiple nanomaterials into heterostructures, unprecedented device function is realized including anti-ambipolar transistors [13], gate-tunable photovoltaics [14], and neuromorphic memtransistors [15]. In addition to technological implications for electronic and energy technologies, this talk will explore several fundamental issues including band alignment, doping, trap states, and charge/energy transfer across previously unexplored mixed-dimensional heterointerfaces [16].

[1] X. Liu, et al., Advanced Materials, 30, 1801586 (2018).
[2] D. Jariwala, et al., Nature Materials, 16, 170 (2017).
[3] C. R. Ryder, et al., Nature Chemistry, 8, 597 (2016).
[4] S. H. Amsterdam, et al., ACS Nano, 13, 4183 (2019).
[5] J. Zhu, et al., Advanced Materials, 29, 1603895 (2017).
[6] G. Hu, et al., Chemical Society Reviews, 47, 3265 (2018).
[7] J. Kang, et al., Accounts of Chemical Research, 50, 943 (2017).
[8] J. Kang, et al., Advanced Materials, 30, 1802990 (2018).
[9] C. Husko, et al., Nano Letters, 18, 6515 (2018).
[10] K.-S. Chen, et al., Nano Letters, 17, 2539 (2017).
[11] W. J. Hyun, et al., ACS Nano, 13, 9664 (2019).
[12] A. C. M. de Moraes, et al., Advanced Functional Materials, 29, 1902245 (2019).
[13] V. K. Sangwan, et al., Nano Letters, 18, 1421 (2018).
[14] D. Jariwala, et al., Nano Letters, 16, 497 (2016).
[15] V. K. Sangwan, et al., Nature, 554, 500 (2018).
[16] S. B. Homan, et al., Nano Letters, 17, 164 (2017).

Two dimensional (2D) layered materials have strong in-plane covalent bonding and weak van der Walls interaction between the planes, which host a wide variety of interesting physical and chemical phenomena. Here I will present our research over 15 years on nanostructured 2D materials. First, I will show how we synthesize these 2D materials into rich morphologies including nanoplates, nanoribbons, layer-vertical aligned films as well as lateral and vertical heterostructures. Second, we developed liquid-gel gating, solid state ion gating, chemical and electrochemical intercalation to control the physical and chemical property of 2D materials. Third, we demonstrated exciting new exciting application of using 2D materials for batteries and catalysis applications.

Combined photonic and electronic systems require diverse devices to be co-integrated on a common platform. This heterogeneous integration is made possible through several separation and transfer methods where the functioning epilayers are essentially released from their growth substrate. The use of 2D layered h-BN as a mechanical release layer has been demonstrated to be a promising technique for the hybrid integration of III-nitride devices. In this talk we will give an overview regarding the epitaxial growth of Nitride-based semiconductor materials by MOVPE including GaN and h-BN, and design device structures for opto-electronic devices. In addition, results on wafer-scale growth and pick-and-place transfer of LED devices via h-BN van der Waals epitaxy, starting from sapphire growth substrates will be presented. Broader range of devices (HEMTs, solar cells, sensors, etc.) and applications (transport, environment, health, etc) will be discussed.

Recent advances in MOVPE enabled the high-quality growth of layered h-BN on sapphire substrates, and has resulted in the realization of series of 2D-3D III-nitride based device structures. This 2D layer also enables damage free mechanical lift-off which may drive future flexible and wearable industry with advanced high-efficient devices. Reliable wafer-scale 2D-2D heterostructure based devices have also been fabricated previously. Here we report, demonstration of MOVPE van der Waals epitaxial growth of GaN based nanostructures on 2D h-BN with a focus on large area van der Waals epitaxy and successful transfer to flexible platform. This approach to the growth of III-N nanostructures on h-BN avoids transfer processes and scaling issues seen with other 2D materials, since both the 1D (GaN nanorods) and 2D (h-BN) are grown at the wafer-scale and in one growth run. Further, this process is used to grow vertical core–shell p-GaN/InGaN/n-GaN nano-PIN device structures on wafer-scale 2D h-BN on sapphire and silicon substrates. Interesting results obtained on this futuristic process and mechanisms involved will be discussed.

Infrared (IR) photodetectors are nowadays in high demand in a wide range of fields such as telecommunication, thermal imaging, remote sensing, night assistance car driving etc. Recently, hybrid phototransistors made of highly efficient light absorbing 0D quantum dots and high mobility 2D materials have introduced a new technology, which dramatically increases the responsivity and gain of the photodetector. In 2012, such low dimensional IR phototransistor based on graphene/PbS QD hybrid was proposed in literature ^{[1, 2]} for the first time. Later in 2017, a high-resolution broadband image sensor based on such hybrid materials was demonstrated ^[3], which is sensitive to ultraviolet, visible and infrared light (300–2000 nm).

We have investigated similar graphene/PbS QD hybrid phototransistors and explored the enhancement of their photo-sensing qualities compared to bare graphene phototransistor in the visible-NIR-SWIR region of optical spectra. Initially, we studied CVD-grown single layer graphene phototransistor in the visible-NIR region, which showed a responsivity of about 10⁴ A/W, associated with the low–doped Si substrate, which acts as the primary material for optical absorption. Such responsivity is 10⁷ orders of magnitude higher ^[4] than conventional graphene phototransistor ^[5]. However, the photo-sensing wavelength range of such device is limited up to 1100 nm depending on the principal absorption material Si. In order to expand the photo-sensing wavelength range up to SWIR region, we have explored the potential of graphene/PbS QD hybrid phototransistors. We synthesized the colloidal PbS QDs absorbing in the NIR as well as SWIR regions and developed a layer-by-layer dip coating with simultaneous ligand exchange procedure to deposit homogeneous PbS QD layers on graphene sheet leading to a well fabricated hybrid phototransistor. We achieved a significantly high responsivity of at 940 nm with irradiation power density of 10^-4 W/m². In the NIR range, the peak responsivity appears to be ∼ 10⁷-10⁸ A/W, whereas in the SWIR range, the observed peak responsivity is ∼ 10⁶ A/W . Moreover, graphene/PbS QD hybrid exhibits high sensitivity at low irradiation power (< 0.05 nW), where bare graphene photodetector is unable to sense such low intensity light in the visible-NIR region.

[1] Konstantatos, Gerasimos, et al. Nature nanotechnology 7.6, 363 (2012).
[2] Sun, Zhenhua, et al. Advanced materials 24.43, 5878-5883 (2012).
[3] S. Goossens et al. Nature Photonics, 11, 366-371 (2017).
[4] Guo, Xitao, et al. Optica 3.10, 1066-1070 (2016).
[5] Xia, Fengnian, et al. Nature nanotechnology 4.12, 839 (2009).

Hybrid structures composed of conventional semiconductors (3D materials) and atomically thin two-dimensional (2D) materials have offered opportunities of recyclable device manufacturing, novel functionalities based on carrier transfer and exciton transport, and understanding natures of van der Waals gap. However, nucleation of 3D materials on 2D materials has not been understood straightforwardly due to absence of surface dangling bonds on 2D material though interaction the overgrown layer and 2D layer as a substrate has been explained by van der Waals interaction. Recent observation performed by the presenter’s team revealed that surface energy landscape is a key for nucleation of 3D materials on 2D materials. Increase in surface energy on 2D layer enhances nucleation probability of 2D materials significantly.
In the presentation those will be discussed several key aspects of nucleation of 3D materials on 2D materials, such as a few strategies to enhance surface energy on 2D layer, insights of the nucleation strategy on remote epitaxy, and demonstration of devices based on 3D/2D heterostructures.

The heterogeneous integration of dissimilar materials is a long pursuit of material science community and has defined the material foundation for modern electronics and optoelectronics. The typical material integration approaches usually involve strong chemical bonds and aggressive synthetic conditions and are often limited to materials with strict structure match and processing compatibility. Alternatively, van der Waals integration, in which freestanding building blocks are physically assembled together through weak van der Waals interactions, offers a bond-free material integration strategy without lattice and processing limitations, as exemplified by the recent blossom of 2D van der Waals heterostructures. Here I will discuss the fundamental forces involved in van der Waals integration and generalize this approach for flexible integration of radically different materials to produce artificial heterostructures with minimum interfacial disorder and enable high-performing devices. Recent highlights include the formation of van der Waals metal/semiconductor junctions free of Fermi level pinning to reach the Schottky-Mott limit; the creation of a new class of high-order van der Waals superlattices with highly distinct constituents of atomic or molecular layers; and the development of van der Waals thin film electronics with unprecedented flexibility and stretchability. I will conclude with a brief perspective on exploring such artificial heterostructures as a versatile material platform with electronic structure by design to unlock new physical limits and enable device concepts beyond the reach of the existing materials.

Van der Waals (vdW) surfaces of 2D materials such as graphene and BN are attractive for growth of GaN and other III-nitride films and device structures. The weak inter-planar vdW bonding between 2D layers allows for easy mechanical separation of III-nitride films from the growth substrate for lift-off, transfer and integration on to arbitrary substrates. This approach of vdW lift-off is suitable for transfer of delicate single devices, like T-gate high electron mobility transistors (HEMT), as well as, wafer sized films. However, to obtain high performance devices, control over nucleation and epitaxy on 2D layers is necessary for high quality films. Additionally, to enable a simple transfer process strain and adhesion between the film and 2D layer must be balanced. In this paper, we describe the growth of high quality GaN and AlGaN/GaN HEMT structures on 2D BN. The effects of nucleation, III-nitride epitaxial growth process and BN morphology will be discussed demonstrating how they effect material quality, properties, strain and adhesion. We will then discuss transfer schemes for individual AlGaN/GaN HEMTs and large area membranes transferred to both flexible polymer substrates and a variety of rigid substrates. With access to GaN and AlGaN/GaN two dimensional electron gas (2DEG) membranes on flexible substrates, we investigate the effects of strain on basic materials properties from compressive to tensile. This is particularly powerful for probing the effects of stain on the electrical properties, like sheet carrier density and mobility, in the polarization induced AlGaN/GaN 2DEG. In transferred devices, we demonstrate exceptional transport properties with mobility > 2,000 cm²/Vs and device performance with G_M of 300 mS/mm and f_T and f_max > 34 GHz and 75 GHz under stain while bending. Lastly, we will cover vdW and adhesive bonding to high and low thermal conductivity substrates and the effects on device performance and self-heating. This paper will highlight the great potential of vdW epitaxy and lift-off for transfer and integration of GaN on to various platforms.

When crystals approach 2D, their ferroelectric phase may be destabilized. How their temperature-dependent polarization behaves remains largely unknown. In this work, we attempt to answer this question using 2D pyroelectric oxides. We show that quasi-2D oxides down to a unit cell thickness can be epitaxially grown on perovskite substrate by the molten salt-assisted quasi-van der Waals epitaxy following a screw-dislocation driven mechanism. We experimentally demonstrate switchable in-plane photo-ferroelectricity and thickness-dependent pyroelectricity. Weakly-coupled interface allows us to reveal that electron-phonon renormalization leads to the observed dimensionality effect. Harnessing the dimensionality effect in pyroelectricity via novel epitaxy strategies could promote their applications in uncooled infrared cooling and thermal energy harvesting.

Ideally, electronic heterostructures from dissimilar materials leads to enhanced functionality. Yet, experimentally forming these heterostructures is challenging due to lattice or thermal coefficient of expansion mismatch leading to defect formation or thermally driven atomic diffusion resulting in cross-doping and gradual junction transitions. These challenges may be overcome with the discovery of remote epitaxy and 2D layer transfer [1]. Here, SiC epitaxy is performed on epitaxial graphene as the electrostatic fields from the substrate penetrate the graphene and guide adatom registry. The film is easily peeled away since the graphene is not bonded to either the substrate or epilayer; the epilayer is then van der Waals bonded to a different material enabling new functionality. We will present experimental results on the remote epitaxy of SiC.
There are three necessary steps to create remote epitaxy. The first is to grow epitaxial graphene on SiC, followed by transferring the graphene to a desired substrate (if different from SiC), and finally the growth of the remote epitaxial layer. If the remote epitaxy is to be SiC, which is the focus of this paper, the second step is not necessary. Epitaxial graphene (EG) was first synthesized on 4H- and 6H-SiC in a horizontal hot-wall CVD reactor between 1540 and 1580 °C in 10 slm of Ar and 100 mbar [2]. The growth temperature was dependent upon the offcut of the substrate, where substrates with higher offcuts require a lower growth temperature to ensure 1 ML of EG, which is desired to assist in SiC adatom registry during growth. SiC remote epitaxy was then performed on the EG using silane (2% in H₂) and propane precursors, where the SiC polytype replicated the underlying substrate. In an effort to transfer the remote SiC epi/EG to another substrate such as SiO₂/Si, a metallization step was performed. Thin Ti and/or Ni layers were initially deposited followed by a thicker high stress metal to create strain and aide in removing the remote SiC epi/EG from the SiC substrate [1]. Once transferred, the metal was removed via a metal etch.
In this work, we will discuss the important parameters needed for successful remote SiC epitaxy, such as graphene thickness, process flows, ramping conditions and remote epitaxy growth temperature. The epitaxial morphology characterized by SEM, Nomarski microscopy and Electron Detectio and graphene coverage and transfer evaluated by Raman spectroscopy will be presented.


[1] Kim, et al., Nature 544, 340 (2017).
[2] L.O. Nyakiti, et al., MRS Bulletin 37, 1150 (2017).

Mixed-dimensional heterojunctions, such as zero-dimensional organic molecules deposited on two-dimensional transition metal dichalcogenides (TMDs), exhibit unique interfacial effects that modify the properties of the individual constituent layers. To better understand the effect of changing molecular orbital energy on the heterojunction properties, we report here a systematic study of metallophthalocyanine (MPc) – MoS₂ (M = Co, Ni, Cu, Zn, H₂) heterojunctions using optical absorption, Raman spectroscopy, low temperature photoluminescence, variable temperature electrical measurements, and density functional theory. Notable phenomena include the emergence of heterojunction-specific optical absorption transitions, strong Raman enhancement, and quenching of defect emission that depend on phthalocyanine metal center identity. Temperature-dependent electrical noise measurements on field-effect transistors provide further insight into the properties of these mixed-dimensional heterojunctions. These results highlight the tunable nature of metal-organic-TMD van der Waals interfaces and the importance of the organic architecture and electronic structure in designing mixed-dimensional heterojunctions for optical and electronic applications.

For decades we’ve been hearing about the promise of printing electronics directly onto any surface. However, despite significant progress in the development of inks and printing processes, reports on fully, direct-write printed electronics continue to rely on excessive thermal treatments and/or fabrication processes that are external from the printer. In this talk, recent progress towards print-in-place electronics will be discussed; print-in-place involves loading a substrate into a printer, printing all needed layers, then removing the substrate with electronic devices immediately ready to test. A key component of these print-in-place transistors is the use of inks from various nanomaterials, including 1D carbon nanotubes (CNTs - semiconducting), 2D hexagonal boron nitride (hBN - insulating), and quasi-1D silver nanowires (AgNWs – conducting). Using an aerosol jet printer, these mixed-dimensional inks are printed into functional 1D-2D thin-film transistors (TFTs) without ever removing the substrate from the printer and using a maximum process temperature of 80 C. To achieve this, significant advancements were made to minimize the intermixing of printed layers, drive down sintering temperature, and achieve sufficient thin-film electrical properties. Devices are demonstrated on various substrates, including paper, and evidence of the potential for printing directly onto biological surfaces will be shown. From the versatile printed electronic thin films that have been developed, a diversity of biosensors are being pursued and will be discussed. With continued refinement of the inks and print processes, this print-in-place technique can bring the field of printed electronics closer to where it has been promised to go for so many years: load substrate, press print, remove functional sensors / circuit.

New materials and device integration demands motivate the ability to lift off and transfer epitaxial material. We use a Van der Waals-bonded 2D boron nitride (BN)-on-sapphire template to grow GaN by MOCVD then subsequently separate the GaN using a stress-induced lift-off method. This method has potential for wafer-scale heterogenous integration since large crack-free areas can be separated and the bottom epi-layer surface has sub-nanometer roughness. In this method a tensiley-stressed Ni layer deposited on the GaN acts to separate the epi-layer at the 2D BN interface allowing us to lift off whole wafer-sized GaN layers which are crack-free. Although large crack-free areas of GaN can be separated, the residual compressive strain in the GaN can cause cracks if the strain is inhomogeneously relaxed as the layer is transferred to the new substrate. By varying the stress and thickness of the Ni layer we can homogeneously relax some of the GaN strain upon separation and reduce the resulting crack density. We report on our development of this process including modeling and measurement of the GaN strain by Raman spectroscopy throughout the process.

Due to the current global water crisis, wastewater treatment requires considerable attention and development. Dissolved heavy metals in water trigger serious alerts and can have lethal effects on various components of the environment including water resources, soil, plants, animals, and can also be threatening the health of human beings. In this study, we addressed the issue by introducing novel bucky-paper membranes that were fabricated using a combination of biopolymers (i.e., chitosan and carrageenan) and multi-walled carbon nanotubes (MWCNTs). Three dispersions of MWCNTs with chitosan, carrageenan, and chitosan-carrageenan with 0.1% v/w were prepared using vacuum filtration. The removal of six heavy metals (i.e., cobalt, nickel, copper, cadmium, barium, and lead) was investigated in this study. The water permeability and the removal of heavy metals were evaluated using a dead-end (DE) filtration system. Heavy metals removal was studied at pH 7 and under a range of varying applied pressures (1 to 6 bar). At an applied pressure of 1 bar, the removal of lead and copper by the MWCNTs/carrageenan membrane reached 99% and 88%, respectively. However, MWCNTs/carrageenan membrane was found to be fragile. Nevertheless, adding chitosan to carrageenan had significantly improved the mechanical strength of the membrane while sustain the excellent removal properties of the heavy metals. That is, MWCNT/chitosan-carrageenan membrane significantly exceeds MWCNTs/carrageenan membrane in tensile strength, tensile strain and young’s modules by 400%, 17%, and 6%, respectively. On the other hand, the MWCNTs/chitosan membrane showed a high water permeate flux that reached up to 200 L/h.m². Also, the electrical conductivity of all membranes varied from 37 S/CM to 57 S/CM. Additional characterization techniques on the three membranes were conducted in this study as well.

When an electromagnetic wave impinges on a conductor, most of the incident wave is reflected, which makes metals and transparent conducting oxides (TCOs), such as tin-doped indium oxide (ITO), opaque to visible light (429–750 THz) and far-infrared (FIR) (< 20 THz). This incompatibility between optical transparency and electrical conductivity is well-defined fundamental material properties, but this is often not easy to enhance both simultaneously. Opacity due to electrical conductivity is more pronounced in the lower frequency range. This fundamental incompatibility creates a barrier for the realization of enhanced user-interface and device integration. We present a design strategy for preparing megahertz-range transparent conductor and a concept towards ‘device-to-device integration’ enabled by electromagnetic wave transmittance. The approach to the properties of conductors is verified using a conducting polymer, Poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS), whose microstructure is effectively controlled by solution process. The use of a transparent conducting polymer as an electrode enables the fabrication of a fully functional touch-controlled display device and magnetic resonance imaging (MRI)-compatible biomedical monitoring device, which would open up a new paradigm for transparent conductors.

Among all the anode materials, Li and Na have the highest capacity and great potential to increase the energy density of batteries. Unfortunately, dendrite growth in metal anodes is one of the safety concerns in current battery devices. One approach to address this issue is to use molten or liquid-metal electrodes. This concept has been applied to sodium-beta alumina batteries (NBBs), which are based on a liquid-Na anode and β″–Al₂O₃ solid electrolyte. Consequently, NBBs, such as Na-metal halide (ZEBRA) batteries, are among the most promising technologies for large-scale renewable energy storage because of their high theoretical specific energy, high energy efficiency, and good cycle life.

Interestingly, in the initial stages of sodium-beta alumina batteries (NBBs) development, it was not anticipated that the low performance would rise from the Na/β″–Al₂O₃ interface. Liquid Na metal should be an ideal reversible electrode, provided that it can be maintained in contact with the whole operating area of the β″–Al₂O₃ throughout the discharging and charging of the cell. It was expected to be the least problematic component of the cell. In reality, liquid Na does not fully wet the surface of the β″–Al₂O₃. Although early observations revealed that incomplete wetting of the β″–Al₂O₃ surface by Na and consequent interfacial impedance problems were likely in NBBs, little was reported for numerous years. Ideally, full contact with liquid Na should be achieved, which provides a large active interface area and thus leads to high NBB performance at low temperatures.

We propose a simple approach to achieve unprecedented NBB performance with a capillary-induced wetting concept that significantly improves the Na wetting on β″–Al₂O₃. In this study, we adopted sparked reduced graphene oxide (rGO) loaded on the surface of the β″–Al₂O₃ as an ideal “wetting sheet,” compared with previous metal (Li or Na)-ion batteries where sparked rGO layers have been used for such metal reservoirs as an anode. The sparked rGO layers with nanogaps exhibited complete liquid-Na wetting, regardless of the surface energy between the liquid Na and graphene oxide, which originated from the capillary force in the gaps (see the figures). This indicated (i) that the Na nucleation and growth were sufficiently rapid to form liquid Na when the battery was charged and (ii) that the area of the Na+ passway at the Na (anode)/β″–Al₂O₃ (electrolyte) interface was maximized when the battery was discharged. Thus, a high state-of-charge for potential state-of-art low-temperature NBBs was obtained. This cell-stacking architecture is simple and scalable and addresses the fundamental limitations of NBBs by allowing Na wetting at low temperatures.

Van der Waals (vdW) materials are composed of two-dimensional planes that are held together by weak interlayer interaction. VdW crystals have attracted increasing attention in the past couple of decades because of their significantly different properties from three-dimensional materials. However, it is a small class of materials, with fewer than 100 layered compounds catalogues in the Inorganic Crystal Structure Database (ICSD), including graphite, h-BN, transition metal dichalcogenides, metal halides, metal pnictides, and metal oxides. Here, we present a topochemical redox concept to prepare a new binary vdW materials by structure control from ternary parent compounds. The vdW crystals obtained had the same chemical composition as known three-dimensionally bonded compounds, but exhibited layered crystal structures. In our vdW crystal design strategy, we synthesized LiFeAs and CaFe₂As₂ as parent compounds. We then selectively removed Li and Ca from parent materials by topochemical redox to obtain FeAs vdW crystals with various crystal structures. We confirmed crystal structure of FeAs vdW crystal by X-ray diffraction (XRD), transmission electron microscope (TEM), scanning transmission electron microscope (STEM) and X-ray photoemission spectroscopy (XPS) and measured electrical property and mechanical property to identify the difference FeAs vdW crystal with three-dimensional orthorhombic FeAs.

Heteroepitaxy of semiconductor on two-dimensional atomic layered materials (2d-ALMs) has been promising for fabricating transferrable and flexible devices, because the use of 2d-ALMs allows to easily exfoliate overlayer device from host substrate.[1,2] Recently, an emerging epitaxy has been reported, which is the so-called remote epitaxy.[3,4] The remote epitaxy enables to produce single crystalline overlayer on graphene layer because the crystallographic registrations of overlayer can be copied from a underlying single crystalline substrate across graphene layer.[4,5] In the remote epitaxy of microrods (MRs), graphene thickness that is the remote epitaxial gap is critical to determine growth density of MR overlayer because penetrated field strength given from substrate leading to nucleation–growth is attenuated as the graphene thickness increases.[4] This implies that as we use graphene interlayer with spatially different thicknesses can result in different growth regimes of i) remote epitaxy or ii) non-growth, depending on graphene thickness.
We present how the remote epitaxy can be applied to selective-area growth using graphene pattern layer. Position-controlled remote heteroepitaxy was performed by hydrothermal growth of ZnO MRs on intaglio-patterned graphene (IPG)-coated c-plane GaN substrate. The IPG layer consists of two graphene parts of (I) perforated-hole-patterned multilayer graphene (MLG) that prevents the remote epitaxy, which act as a growth mask layer, and (II) single-layer graphene (SLG) penetrating the potential field from underlying GaN substrate to allow remote heteroepitaxy of ZnO through the hole aperture of mask layer, which is embedded between GaN and MLG mask layer.
The hole-aperture SLG area yielded remote heteroepitaxial ZnO MRs, whereas MLG area inhibited growth. Diameter and spacing of ZnO MRs are controlled by changing the hole pattern parameters. Transmission electron microscopy revealed the remote heteroepitaxial relationship between ZnO and GaN across the SLG area. According to density-functional theory calculations, the orbitals of SLG transfer the charge from the underlying GaN to the SLG surface leading to remote epitaxy, but the thick MLG was not capable of charge transfer in a long range. The weak van der Waals adhesion of ZnO/IPG/GaN was applied to exfoliate the ZnO MRs overlayer by a thermal release tape-assisted exfoliation technique and to recycle the original substrate. Our results readily open an opportunity to utilize the remote epitaxy for transfer of arrayed epitaxial structures in the designed size and arrangement for device manufacturing.

Reference
[1] K. Chung et al., Science 330, 655 (2010)
[2] C. H. Lee et al., Adv. Mater. 23, 4614 (2011)
[3] Y. Kim et al., Nature 544, 340 (2017)
[4] J. Jeong et al., Nanoscale, 10, 22970 (2018)
[5] W. Kong et al., Nat. Mater., 17, 999 (2018)

Two-dimensional atomic materials are emerging as epitaxial substrates for transferrable and flexible device application.[1] Graphene has excellent properties for the use as a substrate, such as high electrical conductivity, excellent mechanical strength, and optical transparency.[2,3] Nevertheless, since the difficulties of producing single crystalline thin film and crystallographically aligned nano/microstructures on graphene, due to the use of poly-domain graphene, there remains a challenge to utilize the graphene as a substantially epitaxial substrate. This obstacle can be overcome by the remote epitaxy.[4] The remote epitaxy enables to grow single crystalline overlayer on graphene regardless of the graphene domain because crystallographic registration can be dictated from underlying substrate through graphene. Here, we demonstrate remote homoepitaxy of ZnO microrods (MRs) on different crystal planes of apolar a- and polar c-plane ZnO substrates across graphene using hydrothermal growth method.
Despite of the presence of poly-domain graphene intermediate layer, the ZnO MRs were epitaxially grown on a-plane and c-plane ZnO substrates, which were found to be homogeneous in-plane orientation over the entire surface of graphene-coated ZnO substrates. Such homoepitaxial relationship across graphene between ZnO MR and substrate was revealed through transmission electron microscopic and selected area electron diffraction analyses. The density-functional theory calculations suggested that the charge redistribution occurring near graphene induces the electric dipole formation, so the attracted adatoms lead to nucleation-growth of the remote-epitaxial overlayer. Because of a strong potential field caused by long-range charge transfer given from the substrate, even the use of bi-layer and tri-layer graphene resulted in the remote-epitaxial ZnO MRs. The effect of substrate crystal planes is also theoretically and empirically demonstrated. The ability of the graphene, which can be released from the host substrate without covalent bonds, was adapted to transfer the overlayer MR arrays. After the delamination, the host substrate was reused by repeating the process of remote epitaxy over again. This unconventional epitaxy technique offers an opportunity of the producing well aligned, transferrable and flexible epitaxial nano/microstructure arrays templates for epitaxial electronics and optoelectronics applications and regenerating the substrate for cost-saving device manufacturing.

Reference
[1] K. Chung et al., APL Mater. 2, 092512 (2014)
[2] Y. J. Hong et al., ACS Nano 5, 7576 (2011)
[3] A. M. Munshi et al., Nano Lett. 12, 4570 (2012).
[4] Y. Kim et al., Nature 544, 340 (2017)

III-nitride materials including GaN, InN, AlN, and their ternary alloys have been one of the key materials to realize advanced electronic and optoelectronic devices, such as high electron mobility transistors, light emitting diodes and lasers. The device performance heavily depends on defect density of III-nitride materials. However, the lack of native substrates leads to the challenges in fabricating defect-free III-nitride materials, severely limiting the implementation of variety of new devices utilizing III-nitride materials. Additionally, the attachment of III-nitride thin film to its substrate poses challenges of heterointegration of III-nitrides with other material systems. Separation of III-nitride thin films from the substrate while maintaining the pristine material quality is highly favorable.

We have previously shown the synthesis of GaAs on two dimensional materials followed by the separation of GaAs epitaxial thin film from its substrate [1], by utilizing the process so-called “remote-epitaxy”. Such discovery reveals the possibilities towards the fabrication of defect-free semiconductor materials without the constriction of substrate availability. In this report, we demonstrate the remote-epitaxy process to fabricate thin film GaN [2], as well as its ternary InGaN. The remote-epitaxial materials have been characterized by X-ray diffraction, atomic force microscopy and transmission electron microscopy, rendering device grade material quality. Additionally, the thin film III-nitrides can be peeled off from the substrates, and subsequently bonded to a foreign material surface, including Si and SiO₂. Based on GaN/Si heterostructure, III-nitride based photonic/phononic cavity is demonstrated. This work has demonstrated a promising path for the integration of III-nitride with an arbitrary material system.

Reference:
[1] Kim, Yunjo, et al. "Remote epitaxy through graphene enables two-dimensional material-based layer transfer." Nature 544.7650 (2017): 340.
[2] Kong, Wei, et al. "Polarity governs atomic interaction through two-dimensional materials." Nature materials 17.11 (2018): 999.

III-V compound semiconductors offer outstanding electronic and photonic properties that outperform silicon, but the cost of III-V wafers is extremely expensive. Although reusing original wafers can effectively minimize the cost of wafers, current techniques for wafer recycling add significant costs in fabrication, nullifying the cost savings by reusing the wafer. Remote epitaxy is a newly discovered method that enables single-crystal growth of III-V semiconductor thin films and easy exfoliation of the grown film, thus promising for reusing wafers multiple times. It requires an atomically thin spacer layer, such as graphene, to be present on top of III-V substrates, which allows enough field penetration from substrates through the graphene layer to make epitaxial III-V thin films grown on it maintain single-crystallinity. In addition, the graphene layer provides weak bonding at the III-V/graphene interface, and thus thin films grown on top of graphene can be easily exfoliated, leading to a cost-effective way of wafer reuse and flexible thin film production. However, previous methods of transferring graphene onto III-V wafers, which use polymethyl methacrylate (PMMA) or metal stressor layers to transfer graphene grown on foreign substrates like copper or SiC, introduce defects and damages on graphene and/or substrates during the transfer process. Remote epitaxial films grown on the damaged graphene/substrate suffer from lower crystal quality and imperfect exfoliation, which undermines wafer reusability and device performance.
Here we report the CVD growth of amorphous graphene on III-V wafers at low temperature that enabled improved quality of remote epitaxial films and their perfect exfoliation. By introducing toluene as a carbon source which cracks at a relative low temperature for graphene growth, we show fully covered amorphous graphene on AlGaAs/GaAs substrates despite arsenic’s low decomposition temperature. The surface of graphene-coated AlGaAs/GaAs substrate remains smooth with a RMS roughness of around 3Å. We also demonstrate 100% coverage of single-crystalline GaAs thin films grown on amorphous graphene, with the film’s quality significantly improved compared to the case of transferred graphene. In addition, roughness of the substrate’s surface remains the same after exfoliation of grown GaAs film, and the growth and exfoliation were successfully repeated multiple times, proving the feasibility for wafer recycling. Through this low temperature CVD growth approach and remote epitaxy, we successfully demonstrate wafer-scale flexible thin film exfoliation and recycling of substrates, which will lead to new opportunities in III-V thin film-based electronics and novel heterostructures with reduced cost.

We investigate nanoparticle film structures as components for daytime radiative cooling, using a generalized effective medium theory which is capable of recovering the Maxwell-Garnett, Bruggeman, and Coherent Potential mixing formulas. We show that, in all cases, nanoparticle fill fraction is a degree of freedom which can be used to improve free space coupling and spectrally tune the emissivity and absorptivity curves so as to optimize radiative emission within the atmospheric transmission window (8 – 14 μm). Based on this theory, we design different radiative cooling structures from composites of SiO₂ and Si₃N₄ nanoparticle films, which were chosen as the example emissive materials because of their strong absorption peaks within the atmospheric transmission window. Our results show that two-layer nanoparticle films of these materials outperform all thin-film analogs and are sufficient to achieve cooling performances comparable to leading reports in literature. Further research will explore how the low thermal conductivity of nanoparticle films could be used to further enhance daytime radiative cooling performance. This research supports the idea that simple nanoparticle films couple provide a viable method for designing radiative cooling structures, provided scalable methods are identified for nanoparticle fabrication.

The history of coloration has been proceeded by trials and development of many technologies for selectively splitting the light. Structural coloration, among them, has grown up as a potential candidate alternating traditional pigments/dyes coloration since their great durability, sustainable production, and fine color tuning by controlling materials and dimension. As this property, structural color facilitates the creation of security features for anti-counterfeiting. However, conventionally, most of them have used metal/dielectric materials to make a resonating system at a specific wavelength and polarization condition which demands complex fabrication processes such as e-beam lithography or ion beam lithography technology. Even though this system can be suitable to an ultra-high definition structural color display, it is limited to only a few millimeter scales or less. To apply into a realistic application such as banknote, luxury products, or customer goods, it is necessary to make it over centimeter scale with a flexible medium.

In this work, to break through these significant flexibility/scalability limitations, we developed polarization distinguishable covert display with ultrathin porous nanocolumns (PNCs) on a metal film by glancing angle deposition (GLAD) method. Particularly, the unique property of this covert display showed powerful functions which provide a hurdle to the inadvertent viewing of stored optical information without compromising their aesthetic. This structure with lossy PNCs on a metal film makes a non-trivial phase shift at the interfaces causing short reflection path and finally shows strong resonance [1]. Furthermore, the PNCs fabricated with GLAD produce anisotropic media and has different effective complex refractive index causing different resonating condition depending on polarization direction [2]. Using this concept, we designed polarization distinguishable color filters with a variety of materials and dimensions with rigorous coupling wave analysis (RCWA) method. In this process, we obtained an enlarged color range occupying over 80% of standard RGB area. In this color response, some showed a small color difference (ΔE), the other showed a large color difference following polarization angle. Additionally, we confirmed this result is originated from human eye’s spectral responses according to between tristimulus value and reflectance spectra having a single minimum. Using these results, we designed optical data label (i.e., QR code) which shows inadvertent viewing in the naked eye (under unpolarized light), on the other hand, this label can be selectively detected by polarized light. For practical application, we designed this label into daily consumer goods with the color matching between labels and products to avoid compromising aesthetics of products and successfully demonstrated onto wine bottleneck, food packaging with confirming the authentication process. As a multi-functional color display, we demonstrated a colorimetric detection. Due to the low complex refractive index of porous medium, PNCs are sensitive to subtle external environment change. Consequently, PNCs reveal a sensitive color change according to background refractive index or additional layer thickness change. Using this unique property, we presented colorimetric detection with external change for volatile organic compounds (VOCs) and humidity.

[1] M. A. Kats, R. Blanchard, P. Genevet and F. Capasso, “Nanometre optical coatings based on strong interference effects in highly absorbing media,” Nat. Mater. 12 (2012): 20–24.
[2] Y. J. Yoo, J. H. Lim, G. J. Lee, K. I. Jang, and Y. M. Song, "Ultra-thin films with highly absorbent porous media fine-tunable for coloration and enhanced color purity," Nanoscale 9 (2017): 2986-2991.

The last decade has seen nearly exponential growth in the science and technology of two-dimensional materials. Beyond graphene, there is a huge variety of layered materials that range in properties from insulating to superconducting. Furthermore, heterogeneous stacking and doping of 2D materials also allows for additional band structure engineering. In this talk, I will discuss recent breakthroughs in two-dimensional atomic layer synthesis and properties, with am emphasis on doping 2D semiconductors. Our recent works include development of an understanding of substrate impact on growth and doping of 2D materials to tune them from n- to p-type, and to create 2D magnets. Our work and the work of our collaborators has lead to a better understanding of how substrate not only impacts 2D crystal quality, but also doping efficiency in 2D materials.

Technological fashion of electronics has recently changed from device miniaturization to flexible device and 3-dimensional heterogeneous integration. For the flexbile devices, organic or polymeric materials have been generally used, but inorganic semiconductor is ultimately desirable in terms of high performance, long lifetime, and harsh environment tolerance. Nonetheless, the rigid, brittle characteristics of semiconductor has limited its utilization in the flexible device applications.
Graphene-inserted growth is promising for tranferrable device fabrication because the surface of graphene has no dangling bonds. The van der Waals interfaces of overlayer/graphene/wafer enable mechanical exfoliation of overlayer from wafer via a facile sticky scotch tape technique.[1] Using the graphene interlayer, there are two epitaxy regimes, which are van der Waals and remote epitaxies,[2,3] that can be classified by the role of graphene. In addition, the growth of wire or rod-shaped semiconductor crystallites is promising for deformable device applications because of the spatially separate geometry.[4]
This talk introduces the aforementioned epitaxies for transferable flexible optoelectronics. We begin discussing the role of graphene or underlying wafer how they dictate the epitaxial relation with overlayer. As the first section, the van der Waals epitaxy of InAs nanowires on graphene is presented. Use of thinnest substrate of single-layer graphene,[5] fabrication of InAs/graphene/InAs double heterostructures using suspended graphene,[6] and position-controlled growth,[3] all of which were achieved through the van der Waals epitaxy, are presented. Then, as the second section, the remote epitaxy is introduced. Hydrothermal remote epitaxy of ZnO microrods and metal–organic vapor-phase remote epitaxy of GaN microrods are presented.[7–9] We also discuss the remote heteroepitaxy of ZnO on GaN and GaN on Al₂O₃ for lattice less- and far-mismatched remote epitaxy, respectively.[8,9] The transfer of GaN heterostructured microrods of p–n junction with InGaN multiple quantum-wells is demonstrated for flexible microrod-light-emitting diode panel whose mechanical property is highly robust against crumpling/folding operations and repeated bending test.[9] Especially, we show the wafer recyclabilty for fabricating the deformable microrod-light-emiting diode panel over again on the same wafer in the remote epitaxy. We fianlly show how the principle of remote epitaxy, in which the limited thickness of graphene can penetrate the bonding feature of underlying wafer to the surface of graphene for epitaxial relationship, can be applied for site-selective growth.[10] The challenges and opportunities of epitaxies on/across graphene are discussed toward future flexible, transferable displays technology.

References
[1] Kim, Cruz, Lee et al. “Remote epitaxy through graphene enables two-dimensional material-based layer transfer” Nature 544, 340 (2017).
[2] Chung et al. “Transferable GaN layers grown on ZnO-coated graphene layers for optoelectronic devices” Science 330, 655 (2010).
[3] Hong and T. Fukui “Controlled van der Waals heteroepitaxy of InAs nanowires on carbon honeycomb lattices” ACS Nano 5, 7576 (2011).
[4] Lee et al. “Flexible Inorganic Nanostructure Light-Emitting Diodes Fabricated on Graphene Films” Adv. Mater. 23, 4614 (2011).
[5] Hong et al. “van der Waals epitaxy of InAs nanowires vertically aligned on single-layer graphene” Nano Lett. 12, 1431 (2012).
[6] Hong et al. “Van der Waals Epitaxial Double Heterostructure: InAs/Single-Layer Graphene/InAs” Adv. Mater. 25, 6847 (2013).
[7] Jeong, Min et al. “Remote homoepitaxy of ZnO microrods across graphene layers” Nanoscale 10, 22970 (2018).
[8] Jeong, Min et al. “Remote heteroepitaxy across graphene: Hydrothermal growth of vertical ZnO microrods on graphene-coated GaN substrate” Appl. Phys. Lett. 113, 233103 (2018).
[9] Jeong, J. Cha, Q. Wang et al. (unpublished)
[10] Jeong et al. (unpublished)

The fabrication of thin-film semiconductor relies on epitaxy, during which direct bonding is made between the epitaxial layer and the substrate, to ensure highly ordered crystalline lattices in the epitaxial layer as a copy of its substrate. The strong physical bonding between epitaxial layer and substrate implies strict requirements of lattice matching as well as difficulties in heterointegration. We have previously shown that the insertion of a monolayer graphene in between the epitaxial layer and its substrate, in the so-called “remote-epitaxy” process, did not change the crystalline orientation of the epitaxial layer, and the epitaxial alignment between the epitaxial layer and its substrate was maintained as if the monolayer graphene was “transparent” [1]. In this report, we further explore the fundamental mechanisms behind such an epitaxial alignment through two-dimensional materials. And the general rule governs remote-epitaxy will be introduced [2].

Specifically, the transparency of two-dimensional materials to the epitaxial relationship is linked to the polarization of related material systems. Although the potential field from non-polar materials is screened by a monolayer of graphene, that from polar materials is strong enough to penetrate through a few layers of graphene. Such field penetration is substantially attenuated by 2D hexagonal boron nitride, which itself has polarization in its atomic bonds. Based on the control of transparency, modulated by the nature of materials as well as interlayer thickness, various types of single-crystalline materials across the periodic table can be epitaxially grown on 2D material-coated substrates. The epitaxial films can subsequently be released as free-standing membranes, which provides unique opportunities for the heterointegration of arbitrary single-crystalline thin films in functional applications.

[1] Kim, Yunjo, et al. "Remote epitaxy through graphene enables two-dimensional material-based layer transfer." Nature 544.7650 (2017): 340.
[2] Kong, Wei, et al. "Polarity governs atomic interaction through two-dimensional materials." Nature materials 17.11 (2018): 999.

Crystallographic defects, such as dislocation, are well known to strongly affect material’s physical properties. Electron-hole recombination mediated via dislocations (e.g. threading dislocation) is one of the predominant loss mechanisms for the sub-optimum performance in conventional semiconductors devices. However, how dislocation impacts its carrier dynamics in the ‘defects-tolerant’ halide perovskite is largely unknown. Stimulated by the successful pioneer work of remote homoepitaxy of semiconductor GaAs^{1, 2}, researchers have also been achieved (remote heteroepitaxy) in the systems of AlN film on sapphire³, copper film on sapphire⁴, and ZnO film on GaN⁵. In our study, we synthesize epitaxial halide perovskite with controlled dislocation density via a remote heteroepitaxy approach using polar substrates (NaCl and CaF₂) coated with graphene. Density functional theory calculations have revealed the structure and magnitude of the incompletely screened electrostatic potential from the polar substrates, supporting the remote epitaxy in the present case. The regulated film-substrate interactions have further reflected themselves in controlling the wavelengths of the ferroelastic domains. Molecular-dynamics simulations reveal the kinetic process during remote epitaxy. Comparing to the ionic epitaxy with high dislocation density (both misfit and threading), the film grown via remote shows much enhanced photoluminescence intensity and increased carrier lifetime. Our successful demonstration of remote epitaxy in halide perovskite provides an approach to develop free-standing halide perovskite film with reduced dislocation density. More importantly, dislocations and their impacts on carrier dynamics and device performance in halide perovskite have to be recognized and scrutinized.

1. Kim Y, Cruz SS, Lee K, Alawode BO, Choi C, Song Y, et al. Remote epitaxy through graphene enables two-dimensional material-based layer transfer. Nature 2017, 544: 340.
2. Kong W, Li H, Qiao K, Kim Y, Lee K, Nie Y, et al. Polarity governs atomic interaction through two-dimensional materials. Nat Mater 2018, 17: 999–1004.
3. Qi Y, Wang Y, Pang Z, Dou Z, Wei T, Gao P, et al. Fast Growth of Strain-Free AlN on Graphene-Buffered Sapphire. J Am Chem Soc 2018, 140(38): 11935-11941.
4. Lu Z, Sun X, Xie W, Littlejohn A, Wang G-C, Zhang S, et al. Remote epitaxy of copper on sapphire through monolayer graphene buffer. Nanotechnology 2018, 29(44): 445702.
5. Jeong J, Min K-A, Kang BK, Shin DH, Yoo J, Yang WS, et al. Remote heteroepitaxy across graphene: Hydrothermal growth of vertical ZnO microrods on graphene-coated GaN substrate. Appl Phys Lett 2018, 113(23): 233103.

Understanding energy level alignment between molecules and two-dimensional materials is conducive to controlling charge and energy transfer in mixed-dimensional heterojunctions. The competition between distinct energy scales, such as the ones associated with electronic correlations, interface dipole, orbital hybridization, and non-local dielectric screening, can lead to significant renormalization of the intrinsic energy levels of each material at these interfaces. The variety of these energy scales also complicates the theoretical description of the level alignment for mixed-heterojunctions accurately.

Here we study the electronic structure and level alignment for mixed 0D-2D heterojunctions containing transition metal phthalocyanines on MoS₂ by first-principles density functional theory (DFT) calculations. By using optimally tuned range-separated hybrid (OT-RSH) functionals, a recently developed method for finding the optimal range-separated parameters for describing short-range and long-range electron exchange effect, we obtain energy level alignment consistent with known experimental results, a significant improvement comparing to standard DFT. This method can be used for other mixed-heterojunctions for better understanding of their electronic and optical properties.

The current electronics has been mainly dominated by Si-based devices due to their mature processing system and exceptional cost-effectiveness. However, next generation electronics needs novel functionalities that cannot be realized by Si because of intrinsic limitation of Si. Accordingly, demand for non-Si electronics has been getting substantially high. Unfortunately, current methodology requires extremely high cost for non-Si materials, which impedes the progress in developing the non-Si based electronics. Here, I will discuss about our group’s efforts to address this issue. Our team recently conceived a new crystalline growth, termed as “remote epitaxy”, which can copy/paste crystalline information from substrates remotely through graphene, thus generating single-crystalline films on graphene. As interfacial binding energy is attenuated by inserting graphene at interface, the single-crystalline films can be easily exfoliated from the slippery graphene surface. Also, the graphene-coated substrates can be, in principle, reused infinitely to produce single-crystalline films. Thus, the remote epitaxy can produce non-Si semiconductor films with unprecedented cost efficiency while allowing additional flexible device functionality required for current ubiquitous electronics.
Next, I will discuss about a layer splitting technique which can be a potential solution to overcome the problem in obtaining large-scale and monolayer 2D materials. A 2D material-based heterostructure has been intensively studied because of its unique device functionalities and novel physics. However, it is extremely challenging to secure large-scale and monolayer 2D materials because of following issues: 1) poor scalability for laboratory fabrication processes of 2D heterostructures and 2) lack of well-defined control parameters for kinetics of 2D materials and predictable number of layers of 2D materials. To resolve this issue, we conceived a new approach called “layer-resolved splitting” which obtains multiple monolayer from multilayer 2D materials by controlling interfacial toughness contrast. As this method is versatile and universal, we can, in principle, apply to all 2D materials. We succeeded in having large-scale, monolayer 2D materials through our approach and, thereby 2D heterostructures were demonstrated for functional devices.
Lastly, I would like to discuss opportunities of mixed-dimensional heterostructure demonstrated by remote epitaxy and layer-resolved splitting. As they produce freestanding 3D bulk films and 2D atomic layers, a new type of 3D/2D heterostructures can be realized where a new physics and new device architecture are revealed. Therefore, I believe that a new opportunity will be discovered through the mixed-dimensional heterostructures

Remote epitaxy is a method for growing single-crystalline materials on a graphene-terminated single-crystalline substrate, in which the epitaxial registry occurs between film and substrate rather than film and graphene. This method has been realized for several homoepitaxial systems, such as GaAs, GaN, LiF and ZnO. Yet, the microscopic mechanisms that allow for remote epitaxy remain mysterious. Here we demonstrate remote homoepitaxy of GaSb and comment on the role of defects on the apparent lattice transparency of graphene. In addition, we elucidate the effect of the graphene/substrate interface quality associated with different graphene transfer methods.

We have grown our films via molecular beam epitaxy and used x-ray and electron diffraction, along with transmission electron microscopy to confirm that the single-crystalline film is in-phase with the underlying GaSb (001) substrate. Graphene grown via chemical vapor deposition is cleanly transferred to GaSb, which has been corroborated via Raman spectroscopy and scanning electron microscopy. Remote epitaxy necessitates the absence of the semiconductor substrate’s native oxide, since this amorphous layer would inhibit epitaxial registry to the underlying substrate. Therefore, the question of how the native oxide desorbs when capped with graphene is of critical importance. We have studied this desorption mechanism through in-situ photoemission and reflective high energy electron diffraction. In addition, we have investigated the nucleation selectivity of GaSb on patterned and unpatterned graphene via in-situ scanning tunneling microscopy. The GaSb overlayer can be readily exfoliated, which allows for the substrate to be recycled and for the exfoliated film to be transferred to arbitrary substrates.