Symposium MS02—Mechanically Coupled and Defect-Enabled Functionality in Atomically Thin Materials

Graphene, the prototypical 2D material, has attracted a considerable degree of attention for device applications since its discovery in 2004. The material is commonly synthesized via chemical vapor deposition (CVD) and can contain a variety of intrinsic and extrinsic defects. To enable applications that utilize the unique properties intrinsic to graphene, these defects will not only need to be controlled during synthesis, but will also need to be engineered to different degrees and patterns after synthesis. The current methods of defect engineering and patterning of graphene are either uncontrollable or result in deleterious side effects (i.e. dangling bonds) that reduce materials performance in fabricated devices.
In this work, we use electron-beam (e-beam) chemistry to engineer defects in CVD graphene by performing a highly controlled and tunable process that is ideal for photonic and electronic device fabrication without the use of photoresist. To enable e-beam chemistry on graphene, radiolysis products were produced by the interaction of the e-beam with water vapor inside a scanning electron microscope (SEM) [1]. These products reacted with the carbon atoms of the graphene at precise locations scanned by the e-beam. E-beam chemistry was performed on CVD graphene transferred using PMMA onto a Si/SiO₂ substrate [1]. The effectiveness of this precise maskless, resistless process in creating defects at precise locations were verified using different characterization techniques. Raman spectroscopy in the defect engineered areas of graphene revealed the presence of large defect band signals at ~ 1350 cm^-1 (D-band) and at ~ 1625 cm^-1 (D’-band) with peak widths < 50 cm^-1. To explore the degree of control in defect engineering exerted via e-beam chemistry, e-beam parameters (e.g. dwell-time, scan-rate, pressure etc.) were varied and their effect on intensity ratios (such as D/G ratio, where G-band corresponds to the graphitic peak at ~ 1590 cm^-1, D/D’ ratio, and G’/G ratio, where G’ is the second order graphitic peak at ~ 2690 cm^-1), peak widths and average distance between defects (L_D) were studied. Raman intensities were also theoretically calculated by considering scattering of incident photon by the phonons, electrons and holes in defect engineered graphene [2]. Comparison of experimental and theoretical analysis of Raman spectra revealed the presence of out-of-plane (sp³-type; L_D > 5 nm and I_D/I_D’ ~ 13) defects with no vacancies on the graphene lattice.
Our study also revealed the high resolution patterning via e-beam chemistry, as we observe negligible D-band intensity near the boundary of the defect engineered area. Atomic force microscopy images show higher phase contrast (as observed for oxidized graphene [1]) on the e-beam patterned regions compared to the non-patterned regions with ~ 100 nm spatial resolution. In addition, the e-beam engineered defects could be healed by annealing graphene at ~ 300 ^oC in air, which resulted in the complete removal of the D-band, thus demonstrating that the e-beam chemistry process is reversible and does not damage (or introduce vacancy in) the graphene lattice itself. E-beam chemistry thus enables us to employ a mask-less, resist-free process to directly engineer reconfigurable defects in graphene.

[1] Islam et al., Appl. Phys. Lett., 111, 103101, 2017; RSC Adv., 2016, p. 42545-42553.
[2] Jiang et al., Carbon 2015, 90, 53-62; J. Phys. Chem. C 2016, 120 (10), 5371-5383.

Two-dimensional transition metal dichalcogenide (2D TMD) layers are highly attractive for emerging stretchable and foldable electronics owing to their extremely small thickness coupled with extraordinary electrical and optical properties. Although intrinsically large strain limits are projected in them, i.e., several times greater than silicon, integrating 2D TMDs in their pristine forms does not realize superior mechanical tolerance greatly demanded in high-end stretchable and foldable devices of unconventional form factors. In this article, we report a versatile and rational strategy to convert 2D TMDs of limited mechanical tolerance to tailored 3D structures with extremely large mechanical stretchability accompanying well-preserved electrical integrity and modulated transport properties. We employed a concept of strain engineering inspired by an ancient paper-cutting art, known as kirigami patterning, and developed 2D TMDs-based kirigami electrical conductors. Specifically, we directly integrated 2D platinum diselenide (2D PtSe₂) layers of controlled carrier transport characteristics on mechanically flexible polyimide (PI) substrates by taking advantage of their low synthesis temperature. The metallic 2D PtSe₂/PI kirigami patterns of optimized dimensions exhibit an extremely large stretchability of ~2000% without compromising their intrinsic electrical conductance. They also present strain-tunable and reversible photo-responsiveness when interfaced with semiconducting carbon nanotubes (CNTs) benefiting from the formation of 2D PtSe₂/CNT Schottky junctions. Moreover, kirigami field-effect-transistors (FETs) employing semiconducting 2D PtSe₂layers exhibit tunable gate responses coupled with mechanical stretching upon electrolytes gating. The exclusive role of the kirigami pattern parameters on resulting mechano-electrical responses was also verified by finite-element modeling (FEM) simulation. These multifunctional 2D materials in unconventional yet tailored 3D forms are believed to offer vast opportunities for emerging electronics and optoelectronics.

Atomically thin semiconductors have gained a lot of interest owing to their unique quantum optical and optoelectronic properties which are strongly dependent on inevitable structural distortions due to strain and defects in their lattice [1-2]. In this work, we study the modification of intrinsic electronic properties using first-principles methods due to the introduction of strain as well as defects in monolayer transition metal dichalcogenides (TMDCs). Our recent theoretical results predict the importance of spin-orbit and electron-phonon interaction on the studied dynamics [3-4]. Here we present the effect of biaxial strain on microscopic interactions in monolayer TMDCs. Our calculations predict a modulation of the spin-orbit splitting (~ 13 meV/%) at the K point of both band edges as a function of strain, where we also find a crossover from a direct to indirect band-gap. Carrier coupling with phonons is also modified with increased strain levels due to changes in intervalley energetics. The resulting dynamics from the static effect of strain were captured via calculation of the wavevector-resolved electron-electron and electron-phonon scattering rates. Along with strain, the formation of atomic defects is unavoidable in 2D materials with currently available growth techniques [1]. Nevertheless, there is a myriad of functionalities in modern optoelectronic and nanophotonic devices that leverage quantum defects including the recent demonstration of single photon emitters in 2D materials [2,5]. Therefore, we present theoretical calculations and analysis of the vibronic structure of a singly charged sulfur vacancy in monolayer TMDCs. Further, we show a pathway to strain and defect engineering for tailoring the valley properties in monolayer TMDCs and enabling sources of quantum emitters that would provide a physical implementation of a localized qubit.
References
[1] D. Edelberg, et al, Nano Letters (2019) doi.org/10.1021/acs.nanolett.9b00985
[2] C. Chakraborty, L. Kinnischtzke, K. Goodfellow, R. Beams, and N. Vamivakas, Nat. Nano. 10, 507 (2015).
[3] C. J. Ciccarino, T. Christensen, R. Sundararaman, and P. Narang, Nano Letters 18, 5709, (2018).
[4] C. J. Ciccarino, C. Chakraborty, D. Englund, and P. Narang, Faraday Discussions 214, 175, (2018).
[5] G. Grosso, H. Moon, B. Lienhard, S. Ali, D. K. Efetov, M. M. Furchi, P. Jarillo-Herrero, M. J. Ford, I. Aharonovich, and D. Englund, Nature Communications 8, 705, (2017).

The strain has been widely used to engineer the electronic properties of 2D transition metal dichalcogenide (MX2) semiconductor. Although there are lots of studies about the strain effects on the band gap, effective mass, excitons etc, much less is known about the mobility. Here using Boltzmann transport theory with the scattering rates determined from first principles, which allows us to accurately calculate the intrinsic (phonon-limited) mobility, we will show how does the isotropic strain modify the electron mobility of 2D MX2 semiconductor. Particularly, we find two distinct behaviors in these materials, which can be explained by a simple physical factor.

Understanding the deformability, flexibility, and bending of two-dimensional (2D) materials is critical for the realization of next-generation electronics and nanomechanical devices. While the mechanics of few-layer graphene have been studied for more than a decade, there is still no consensus on bending stiffness and how it scales with thickness [1-3]. Electron microscopy provides a powerful platform for addressing these challenges by enabling measurements of the conformation and strain of 2D materials at atomic resolution. Using aberration-corrected scanning transmission electron microscopy (STEM), we show that bending in few-layer graphene is dominated by slip and shear between the atomic layers. These results, in combination with density functional theory (DFT) simulations, reveal an unusual, curvature-dependent bending stiffness in few-layer graphene. Unlike in conventional metals where bending produces dislocations which stiffen and embrittle the crystal, we find that slip between the atomic planes of 2D materials dramatically softens few-layer graphene by rendering the layers nearly frictionless when it is curved.

We explore how the bending stiffness of few-layer graphene varies with both thickness and bending angle by systematically examining its conformation over atomically-sharp hexagonal boron nitride (h-BN) steps. These geometries offer direct control over two key variables: number of graphene layers and h-BN step height. Using aberration-corrected STEM images, we extract the bending profiles of few layer graphene over the h-BN steps as a function of the graphene thickness and curvature, then use these values to calculate the bending stiffness of few layer graphene. We find the bending stiffness of few-layer graphene decreases sharply as a function of bending angle, tuning by almost 400% for tri-layer graphene. This softening results from shear, slip, and the onset of superlubricity between atomic layers and corresponds with a gradual change in scaling power of bending stiffness with thickness, from cubic to linear as it is curved. These behaviors are a direct result of changes in atomic registry between the atomic layers; as few-layer graphene is bent, its interlayer interactions transition between two limits: the strong coupling characteristic of Bernal-stacked graphite and the weak, superlubric interactions characteristic of multi-walled carbon nanotubes. Our results indicate that the bending stiffness of few layer graphene can be orders of magnitude smaller than previously thought and provide a new lower limit for the fabrication of ultra-soft, high mobility electronic nanodevices based on 2D materials.

[1] D. Akinwande, et. al., Extreme Mechanics Letters 13, 42-77 (2017).
[2] X. Chen, et. al., Appl. Phys. Lett. 106, 101907 (2015).
[3] D.-B. Zhang, et. al., Phys. Rev. Lett. 106, 255503 (2011).

A challenge and opportunity in nanotechnology is to understand and take advantage of the breakdown in continuum mechanics scaling laws as systems and devices approach atomic length scales. Such challenges are particularly evident in two-dimensional (2D) materials, which represent the ultimate limit of mechanical atomic membranes as well as molecular electronics. For example, after more than a decade of study, there is no consensus on the bending modulus of few-layer graphene, with measured and predicted values ranging over two orders of magnitude, and with different scaling laws[1–5]. However, comparing these studies is challenging because they probe very different and often fixed curvatures or magnitudes of deformation. To unravel the discrepancy, a systematic measurement of bending stiffness versus deformation is needed. The results have practical implications on predicting and designing the stiffness of many 2D mechanical systems like origami/kirigami nanomachines, stretchable electronics from 2D heterostructures, and resonant nanoelectromechanical systems. Moreover, many studies implicitly use the bending modulus to extract important parameters like the interfacial adhesion and friction or to predict the wavelength of 3D rippling in strained membranes.
In this study, we combine atomistic simulation and atomic scale imaging to theoretically and experimentally examine the bending behavior of few-layer graphene. First, we experimentally probe the nanoscale bending by laminating few-layer graphene over atomically sharp steps in boron nitride(h-BN) and imaging the cross-sectional profile using aberration-corrected STEM. Second, we use DFT simulations to examine the bending of few-layer graphene under compression. By measuring the nanoscale curvatures, we extract the simulated and experimental bending modulus while varying both the number of layers and the degree of nanoscale curvature. We further apply this framework to measure the bending stiffness of van der Waals heterostructures, by stacking graphene-MoS₂ heterostructure on the terraced h-BN.
We find remarkable agreement between the theory and experiment and observe an unexpected curvature dependent bending stiffness of few-layer graphene that deviates from continuum scale bending mechanisms. We find that the bending stiffness of few-layer graphene versus curvature corresponds with a gradual change in scaling power with thickness from cubic to linear. We find that the transition in scaling behavior originates from a transition from shear, slip and the onset in superlubricity between the graphene layers at the van der Waals interface, verified by a simple Frenkel-Kontorova model. In contrast to few-layer graphene, the graphene-MoS₂ heterostructure shows an additive bending stiffness for stacking each layer. Thus we evaluate the impact of superlubricity at the van der Waals heterostructure interface under bending deformation. Our results provide a unified model for the bending of 2D materials and show that their multilayers can be orders of magnitude softer than previously thought, among the most flexible electronic materials currently known.

This presentation focuses on the progress on 2D nanomaterials towards greater scientific understanding and advanced engineering applications. In particular the talk will highlight our pioneering work on monolayer memory (atomristors) that can enable various applications including zero-power devices, non-volatile RF switches, and memristors for neuromorphic computing. Non-volatile memory devices based on 2D materials are an application of defects and is a rapidly advancing field with rich physics that can be attributed to sulfur vacancies or metal diffusion. Recent studies based on atomistic modeling and imaging as identified point and cluster defects as primarily responsible for the non-volatile memory effect. Areas with defects typically exhibit memory effect while clean defect-free areas are absent of memory effect. Much work remains to be done to elucidate the underlying phenomenon and realize optimized devices for practical adoption.

Due to the ever-growing need of small-scale and high-performance semiconductor devices, two-dimensional (2D) materials have emerged as promising candidates for future device integration. Due to their exceptional mechanical properties, strain engineering is effectively used to exploit induced mechanical strain in order to tune the carrier mobility and optoelectronic properties of 2D materials. Although strain in 2D materials has been well studied in recent years, most straining methods rely on stretchable substrates to introduce strain and are thus limited to rather low strain values. Here, we present a novel straining approach enabling the application of ultra-high uniaxial strain on freestanding 2D materials and 2D heterostructures, while simultaneously allowing in-situ electrical and optical characterization.
The micromechanical straining devices patterned from SOI wafers enables ultra-high strain levels up to 15%. Mono and few layer 2D materials are transferred onto the straining device using state of the art mechanical exfoliation and dry viscoelastic stamping techniques minimizing unintentional contaminants. To enable reliable electrical contacts and to avoid slipping of the 2D materials they are pinned down by extended Ti/Au contacts. The applied strain was investigated by in-situ Raman measurements in a back-scattering geometry using a confocal μ-Raman setup. The Raman-active modes of graphene, hBN and MoS₂ are very sensitive to strain-induced shifts and are clearly observable during the experiments, thus proving the viability of this approach.

Graphene is an ideal material for optoelectronic applications. Its photonic properties give several advantages and complementarities over Si photonics. I will show that graphene-based integrated photonics could enable ultrahigh spatial bandwidth density, low power consumption for next generation datacom and telecom applications. Heterostructures based on coupled layers of atomic crystals have a number of properties often unique and very different from those of their individual constituents and of their three dimensional counterparts. I will show how these can be exploited in novel light emitting devices, such as single photon emitters, and tuneable light emitting diodes. The role of defects and strain and non-radiative recombination channels will be discussed.

Van der Waals heterostructures constructed of 2-dimensional (2-D) materials such as single layer transition metal dichalcogenides (TMDs) have sparked wide interest because of their large excitonic binding energy, allowing the exploration of novel quantum optical effects in a solid-state system and new opto-electronic devices. In this talk, I will discuss our results in van der Waals heterostructures formed by stacking together two different TMDs (forming a staggered heterojunction) encapsulated with hexagonal boron nitride (h-BN) with electrical contacts in each layer and a dual gate configuration. Interlayer excitons, with electrons and holes residing in spatially separated quantum wells, have long lifetimes (200nS, 5 orders of magnitude longer than intralayer exciton lifetimes). Because of their repulsive Coulomb interaction, they diffuse across the entire sample (20 mm long) driven by interaction, allowing their manipulation towards condensation. Also, we observed and manipulated long-lived charged interlayer excitons which can be used as carriers for quantum information. Interlayer excitons are important for novel opto-electronic devices such as high-temperature interlayer exciton condensates, valley-tronic devices, near-infrared tunable lasers and light emitting diodes. Our work opens a new frontier to explore novel low-dimensional opto-electronic devices and circuits.

Van der Waals heterostructures are made of stacked atomically thin 2D materials, providing access to novel properties relevant to quantum information sciences. Despite the exciting promise of stacked 2D heterostructures, fabrication remains a laborious low-yield process requiring considerable human expertise and time. We present the development of an automatic robotic exfoliator, meant to automate the repetitive manual tasks of mechanically exfoliating thin flakes of such materials, as well as optically searching for flakes of desired characteristics. We provide statistical mappings between robotic exfoliation motion parameters and flake characteristics, such as thickness, in-plane size, and yield. We present preliminary results demonstrating the exfoliation of graphene, and BSCCO (Bi₂Sr₂CaCu₂O_8+y) flakes. In the near future, we will expand this methodology to a wider range of 2D materials.

Defect and strain engineering of 2D materials is an emerging area of research, where homogeneous or heterogeneous alloying can stabilize exotic electronic phases such as Weyl semimetals or pattern the properties of 2D nanoelectronics and optoelectronics on the nanoscale. Aberration-corrected scanning transmission electron microscopy (STEM) is an important tool to study how atomic defects such as vacancies and substitutional dopants impact the structure and properties of 2D materials. Yet, high-precision characterization of defects in 2D materials remains challenging because they are irradiation sensitive, making it difficult to achieve high resolution and signal-to-noise ratio (SNR) measurements without modifying the intrinsic structure. Here, we apply deep learning techniques based on convolutional neural networks (CNNs) to process large volumes of atomic-resolution images of 2D transition metal dichalcogenides (TMDCs). CNNs have revolutionized image recognition in fields such as medical diagnosis, weather forecasting, and facial recognition; recently, they have been applied to detect defects in atomic-resolution STEM images [1]. In our work we utilize fully convolutional network (FCN), a variant of CNN, to identify defects in monolayer WSe_2-xTe_x. Using the resulting data, we determine local picometer-scale strain fields to understand how single-atom defects interact with one another and the surrounding lattice.

In order to enable pm-scale structural measurements in 2D materials, we use an approach analogous to the single-particle reconstruction methods used in cryo-electron microscopy. First, we construct FCN models to identify and classify defects, including substitutions and single and divacancies. We then generate high quality class-averaged images of each defect type by rigid-registration [2] and averaging hundreds of nominally identical atomic defects. These methods result in final images with high SNR yet low dose acquisition, allowing us to extract the detailed atomic structure of defects such as single vacancies, which are highly sensitive to electron beam irradiation. Finally, we measure the locations of each atomic column using 2D Gaussian fitting and determine the local strain. We show that we are able to measure projected atom-atom separations with precisions of up to 0.38 pm, a 22-fold improvement over measurements using single images—representing the highest reported precisions for experimental strain measurements at single-atom defects. We find that two Te substitutions on top of each other in WSe_2-xTe_x increase the projected W-W nearest neighbor distance by 4.1 pm, while single Se vacancies decrease the W-W distance by 16.3 pm. Picometer-scale precision also enables studies of long-range strain fields around single-atom defects, up to 1 nm away from the defect center. Our experimental strain maps quantitatively match simulations of continuum elastic theory using Eshelby’s inclusion approach [3]. In summary, we have developed measurement techniques using FCN to locate, classify, and measure the local structure of single-atom defects in 2D materials with sub-picometer precision for the first time.

References:
[1] M. Ziatdinov et al., ACS Nano 11 (2017), p. 12742-12752.
[2] B. H. Savitzky, et al., Ultramicroscopy 191 (2018), p. 56-65.
[3] A. L. Kolesnikova, et al., Physics of the solid state 56 (2014), p. 2573-2579.

Acknowledgement:
This work was supported by the AFOSR under award number FA9550-7-1-0213 and carried out in part in the Materials Research Laboratory at UIUC. S.K. and W.Z. would like to acknowledge the support from ONR under grant NAVY N00014-17-1-2973.

The attractiveness of single-photon sources in layered materials stems from their ability to operate at the fundamental limit of single-layer thickness, foreseeing high extraction efficiency and providing the potential to integrate into conventional and scalable high-speed optoelectronic device systems. In this talk, I will discuss recent progress on scalable and deterministic generation of quantum-emitter arrays in layered materials, charge and spin control, and progress towards confinement of interlayer excitons.

Defect-induced sub-bandgap emission from hexagonal boron nitride (hBN) is not only hugely interesting for quantum technology, opening a promising route for the design of next-generation single-photon sources [1], but equally offers a new high-throughput characterisation pathway for large-area h-BN films, an urgent requirement given the recent progress in synthesising h-BN; particularly by chemical vapour deposition (CVD) [2]. We present our latest data employing multidimensional super-resolution fluorescence microscopy and other techniques to simultaneously measure spatial position, intensity, and spectral properties of the emitters in h-BN [3,4]. We specifically focus on CVD grown mono-layer h-BN samples over cm areas, which readily offers scalable device integration routes. Our data reveals a correlation in blinking and spectral diffusion for single emitters in monolayer h-BN, closely reminiscent of the behaviour observed in quantum dots [3]. We explore a range of CVD growth, transfer [2, 5] and encapsulation methods warranting controlled, clean and direct emitting-monolayer interfacing, giving us the ability to vertically place emitters with high accuracy. We report on detailed emitter statistics dependant on process conditions and how surface interactions heavily influence the photodynamics. Drawing on quantum dot literature, we devise approaches to enhance and control h-BN emitter properties. We also compare this optical characterisation with a range of other h-BN characterisation techniques to develop a holistic understanding.

[1] Vogl et al. arXiv:1902.03019 (2019).
[2] Wang, R. et al. ACS Nano 13, 2114 (2019).
[3] Stern, H. L. et al. ACS Nano 13, 4538–4547 (2019).
[4] Comtet, J. et al. Nano Lett. 19, 2516–2523 (2019).
[5] De Fazio et al. arXiv:1904.01405 (2019).

Single-photon sources are elementary building blocks for photonic quantum networks, quantum information processing, and quantum metrology, where the photons are used as “flying” qubits [1]. In this work, we investigate the optical properties of quantum defects in few-layered hexagonal boron nitride interfaced with a phase change material, such as vanadium dioxide (VO₂) on a sapphire substrate. It is well known that VO₂ exhibits an insulator-to-metal phase transition when heated just above room temperature [2] which is accompanied with large change in its electrical and optical properties. By inducing this insulator-to-metal phase transition of VO₂ film of thickness ~100 nm, we control the emission characteristics of point-defects with emission centered (zero-phonon line) in the visible regime. We show that far-field radiation pattern and lifetime of the point-defects undergoes a sharp change around the phase transition temperature. Furthermore, we optimize the thickness of the VO₂ film to enhance to optical response of the defects in the vicinity of the phase transition temperature. Our simulations show a modulation of more than 50% in the lifetime of these defects when VO₂ undergoes the phase transition. Interfacing quantum emitters in atomically thin materials [3] with phase change materials opens new opportunity for manipulating single-photon sources near room temperature.

References:
[1] B. Lounis and M. Orrit, Rep. Prog. Phys. 68, 1129 (2005).
[2] Z. Shao, X. Cao, H. Luo, and P. Jin, NPG Asia Materials 10, 518(2018).
[3] M. Toth and I. Aharonovich, Annu. Rev. Phys. Chem. 70, 123(2019).

Transition-metal dichalcogenides (TMDCs) have been extensively studied for both fundamental science and optoelectronic applications owing to their unique properties, such as the atomic layer thickness, dangling-bond free surface, strong light absorption, and layer-tuned direct/indirect bandgap [1]. Valley polarization of excitons can be realized in monolayer TMDCs by optical pumping [2]. However, valleytronics based on TMDCs typically relies on the delocalized excitons, which have an intrinsically fast recombination lifetime due to the large exciton oscillator strength. This short lifetime has placed a strict limitation on the pseudo-spin manipulation before the valley depolarization. Valley selectivity can be preserved for defect-bound excitons and the selective creation of defect-bound excitons with long lifetime could make significant breakthroughs for the practical valleytronic applications [3]. TMDCs are also promising platforms for solid state quantum light emissions, which have great flexibility in heterogeneous assembly and large extraction efficiency. Even though the position of quantum emitters can now be deterministically controlled by nanoscale strain engineering, the broad energy range of uncontrolled impurity- or defect-induced excitons brings a central obstacle to the development of single photon sources with high purity and indistinguishability based on TMDCs [4]. The lack of control over the presence and spatial distributions of defects in TMDs currently limits their utility for scalable devices.
Controlled defect engineering can be an important means for tailoring the electronic and optical properties of two-dimensional materials. In this work, defect densities ranging many orders of magnitude in single-layer WSe₂ are controllably created by focused ion beam (FIB) irradiation. The influences of defects are systematically characterized by room-temperature Raman scattering/ photoluminescence (PL), low-temperature PL and PL dynamics. Only when the FIB dose is higher than 10¹³ cm^-2 in WSe₂, Raman spectroscopy can discern the defects at room temperature, which shows blue shifts for both A_1g and E_2g¹ peaks. Photoluminescence (PL) intensity of WSe₂ drops with increased FIB doses, while the PL peak position and width show no significant changes for different FIB doses at room temperature. Low-temperature PL further reveals the peak of defect-bound excitons, which redshifts and broadens with increased FIB doses. Similar Raman shifts and PL intensity drops are observed for chemical vapor deposition (CVD) grown WSe₂ after FIB irradiation. PL dynamics reveals a microsecond-long lifetime for these defect-bound excitons of single-layer WSe₂, which is three orders of magnitude longer than that of the defect excitons without FIB treatment. This controllably generated ultra-long lifetime of defect-bound excitons in single-layer WSe₂ can be useful for photo-catalytic reactions, valleytronics and quantum light emissions owing to the longer carrier separation/manipulation time.
[1] Nat. Nanotechnol. 2012, 7, 699-712.
[2] Nat. Nanotechnol. 2012, 7, 490-493.
[3] Phys. Rev. Lett. 2018, 121, 167402.
[4] Nat. Commun. 2017, 8, 15053.

Layered systems of two-dimensional (2D) crystals and heterostructures are widely explored for new physics and devices. In many cases, monolayer or few-layer 2D crystals are transferred to a target substrate including other 2D crystals, and nanometer-scale blisters form spontaneously between the 2D crystal and its substrate. Such nanoblisters are often recognized as an indicator of good adhesion but there is no consensus on the contents inside the blisters. While gas-filled blisters have been modeled and measured by bulge tests, applying such models to spontaneously formed nanoblisters yielded unrealistically low adhesion energy values between the 2D crystal and its substrate. Typically, gas-filled blisters are fully deflated within hours or days. In contrast, we found that the height of the spontaneously formed nanoblisters dropped only by 20%–30% after three months, indicating that liquid instead of gas is likely trapped in them. The blister contents are further analyzed using time-of-flight secondary ion mass spectrometry (ToF-SIMS) showing a mixture of water and organic matter at the 2D material-substrate interface. We therefore developed a simple scaling law and a rigorous theoretical model for liquid-filled nanoblisters, which predicts that the interfacial work of adhesion is related to the fourth power of the aspect ratio of the nanoblister and depends on the surface tension of the liquid. Our model was verified by molecular dynamics simulations, and the adhesion energy values obtained for the measured nanoblisters are in good agreement with those reported in the literature. This model can be applied to estimate the pressure inside the nanoblisters and the work of adhesion for a variety of 2D interfaces, which provides important implications for the fabrication and deformability of 2D heterostructures and devices.

We present a paradigm for encoding strain into two dimensional materials (2DM) to create and deterministically place single photon emitters (SPEs) in arbitrary locations with nanometer-scale precision. Our material platform consists of a 2DM placed on top of a deformable polymer film. Upon application of sufficient mechanical stress using an atomic force microscope (AFM) tip, the polymer layer plastically deforms. While the AFM tip is in contact with the sample, the AFM tip forces the 2DM to deform with the polymer layer, resulting in tensile strain buildup in the 2DM. When the AFM tip is removed, the adhesive interaction between the polymer and the 2DM prevents the 2DM from relaxing back to its original, strain-free geometry, resulting in a permanent and highly localized strain field in the 2DM. We demonstrate control of indent size by modifying the applied load. Also, we show excellent repeatability of both indent shape and size. We show that SPEs are created and localized at nanoindents in a WSe₂/PMMA structure, as confirmed by antibunching measurements. The SPEs exhibit single photon emission up to 60 K. This quantum calligraphy allows deterministic placement and real time design of arbitrary patterns of SPEs for facile coupling with photonic waveguides, cavities, and plasmonic structures. We further use the AFM to create trenches in the WSe₂/PMMA structure, leading to nominally one-dimensional strain profiles. We find that the linear polarization orientation of SPEs in the trenches correlates with the orientation of the one-dimensional strain profile. This result holds for trenches made at different orientations relative to the WSe₂ crystal orientation, suggesting that the polarization orientation of SPEs in WSe₂ is determined by the strain profile alone. In addition to enabling versatile placement of SPEs, these results present a general methodology for imparting strain into 2DM with nanometer-scale precision, providing an invaluable tool for further investigations and future applications of strain engineering of 2DM and 2DM devices.
Reference:
Rosenberger et al., “Quantum Calligraphy: Writing Single-Photon Emitters in a Two-Dimensional Materials Platform,” ACS Nano, 2019, https://pubs.acs.org/doi/10.1021/acsnano.8b08730.

Characterization of graphene with point defects, impurities or dopants, is important for understanding the material’s response for applications. For example, graphene with physical defects allows improved adsorption of biomarkers for biosensing, while nanoporous graphene can be used for molecular sieving, gas separation, proton transfer, desalination or atomic species transport. In addition, substitutionally doped graphene was found useful for catalysis or nanoelectronics, while patterning graphene with defects provides the means for plasmonic devices. To quantify and characterize defects in graphene, such as generated by defect engineering, prediction of Raman band intensities that distinguish between defect types, is desirable. Herein we describe a method that combines defect potentials calculated by density functional theory with electron-defect matrix elements and Raman intensities predicted within a tight-binding framework. We summarize calculated defect-induced I(D)/I(D')) intensity ratios for oxidized graphene, found to be in good agreement with experimental data, rendering the calculated I(D)/I(D') Raman intensity signature a metric to assist in experimental characterization. Results for B- and N-doped graphene also demonstrated good agreement with experiment. Finally, we report on our recent work on defect engineered graphene by elucidating the Raman intensity characteristics as a function of the concentration of epoxide groups introduced on the graphene surface.

Lattice mismatch in heterostructures traditionally manifests as accumulation of strain in a thin film that has been grown on a bulk substrate. The strain persists until some critical thickness is reached, at which point the lattice relaxes and returns to its unstrained state. In short, only the deposited material bears the strain. Here we investigate the case in which both sides of a heterostructure are atomically thin transition metal dichalcogenide monolayers. We grew WS₂ on MoS₂ by chemical vapor deposition, and characterized the heterobilayer regions using photoluminescence (PL) spectroscopy. Density functional theory calculations predict a change in the optical bandgap of each monolayer under uniform, biaxial strain. The observed shifts in the A exciton peaks, determined from PL spectra, agree well with the numerical predictions. This suggests a mutual equilibrium is reached in which both materials bear some of the strain. Using the relation between strain and band gap, we calculated the strained in-plane lattice parameters based on the A exciton peak positions. Monolayer MoS₂ and WS₂ have in-plane lattice parameters of 3.16 Å and 3.19 Å, respectively, but both of these values become approximately 3.17 Å in the heterobilayer, differing by less than 0.4 pm. Our findings indicate that lattice matching occurs in 2D heterostructures, and should be taken into consideration when designing van der Waals solids for optoelectronic applications.

A low cost, flexible, sensing platform which can detect catechol in biological fluids and environment is of utmost importance in healthcare. In this work, we demonstrate a facile, low-cost approach to fabricate 2D- Tungsten (VI) oxide (WO₃) nanofiber-based strain induced highly sensitive and selective catechol sensor. 2D-WO₃ was synthesized using electrospinning technique followed by the drop casting of annealed nanofibers on to flexible PET substrates. X-Ray Diffraction (XRD) and Raman studies confirmed the formation of Hexagonal phase-WO₃ and O-W-O bending modes respectively while scanning electron microscopy (SEM) studies revealed the uniform distribution of WO₃ nanofibers on PET substrate. The sensor responded to a wide dynamic range of catechol concentrations from 1µM to 400µM, exhibited a sensitivity of 49.29 µA/µM cm² and a limit of detection of 0.52µM which were far more superior than previously reported catechol sensors fabricated using sophisticated fabrication techniques. The sensing response can be ascribed to the oxidation of catechol to benzoquinone in presence of n-type WO₃ nanofibers thus resulting in increased current flow through the device. Interestingly, when the sensor was subjected to compressive strain during catechol sensing, a remarkable increase in sensitivity of 86.34 µA/µM cm² was observed. The limit of detection also further reduced to 42 nM. Upon subjecting the sensor to a strain ranging from 3.14% to 47.6% an exceptional increase in current was observed. This strain induced increase in sensitivity can be attributed to the increase in the density of interconnected WO₃ nanofibers which enhances the potential active sites for catechol interaction. Furthermore, to explore the practical utility of the sensor, it was used to detect catechol in simulated blood samples for which it showed excellent selectivity towards catechol in presence of other interfering analytes like Ascorbic acid, Uric acid, Na+, Ca+, and glucose. The strategy outlined here can be used to fabricate wearable, chemiresistive platforms for detection of various biological analytes useful in point-of-care medical diagnostics.

Understanding phase transitions in constrained crystalline systems such as in thin films on substrates has often been driven by the motivation to tailor the interesting physical properties coming about near the transition. Ferroelectric-paraelectric transitions in perovskites are an example. In all bulk perovskites these transitions are of the first order. There is no consesus in answering the question as to what happens with these transitions in thin films clamped on a substrate. The prevailing opinion is that a first order transition in bulk turns second order though it is in disagreement with results concerning first order transitions in large clamped systems. The apparent reason of this opinion is that in many reported cases the first-order transition in free crystal is relatively weak and would convert into a second order in film form if the system remained homogeneous. In this talk, we report our theoretical analysis of this case both for laterally clamped slabs with free surfaces and for thin films on stiff substrates. The case of clamped objects allows a consistent analytical analysis in 1D case where the system is laterally finite but relatively large. For systems with small lateral sizes we perform numerical simulations. We observe formation of two-phase state that extend to a temperature range more than 10 degrees, likely smearing the transition anomalies. This temperature interval shrinks with diminishing of the lateral size and may disappear for very small sizes. For films on substrates no analytical theory of two-phase formation is possible though it is possible to analyze the loss of elastic stability of non-symmetrical phase when, at heating, the nucleation is inhibited and the homogeneous state conserves even in the region where the two-phase state is more energetically profitable. Although within a different approach this loss of elastic stability in large clamped systems has been predicted earlier but not for films on substrates. There exists a critical thickness for the stability loss and, seemingly, for the two-phase state formation which is fairly small for studied systems. We demonstrate our approach for an isotropic system that is possible to extend albeit rather cumbersome to experimentally studied perovskite systems where the elastic anisotropy might be important.

In this work, we studied the effect of tin doping concentration on the electrical, optical, morphological, and crystallographic characteristics of tin doped indium oxide (ITO) and comparison of interfacial characteristics between indium-gallium-zinc oxide (IGZO) and ITO with different tin doping concentration to improve the properties of ITO electrode which applied in transparent thin film transistors (TFTs). We deposited ITO films on the glass with 4 inch-sintered ceramic target with low (In:Sn=99:1) and high (In:Sn=90:10) concentration of tin doping by magnetron sputtering system at room temperature. To compare the characteristics of ITO films with high and low concentration of tin doping, Hall measurement, UV-vis spectroscopy, filed emission scanning electron microscopy (FE-SEM), atomic force microscopy (AFM), x-ray diffraction (XRD), x-ray photoelectron spectroscopy (XPS), and ultra-violet photoelectron spectroscopy (UPS) were conducted. Both ITO films with high and low concentration of tin doping showed low electrical resistivity about 10^-4 ohm-cm and superior optical transmittance above 85 % in visible range. In detail, ITO film composed 1 wt% tin doping concentration showed lower sheet resistance and higher optical transmittance. As thickness of films increased, transmittance of both films shifted to red side. The highest figure of merit was obtained at the thickness of 150 and 125 nm in the case of low and high tin concentration ITO films, respectively. Through SEM image, both ITO films with high and low tin doping concentration showed crystalline microstructure. XRD results indicated both ITO films are bixbyite-In₂O₃ crystal structure. Therefore, we fabricated the top gate-transparent IGZO TFTs and circular transmission line method (CTLM) devices with ITO electrodes. Top gate-transparent IGZO TFTs consist of 100nm thick of ITO gate electrode, 80nm thick of Al₂O₃ gate insulator layer, 30nm thick of IGZO channel, and 100nm thick of ITO source/drain electrode. The transfer curve of both high and low tin doping concentration ITO electrode TFTs showed traditional n-type transistor characteristic and operated in depletion mode. And also, transistor parameters such as threshold voltage, field effect mobility, subthreshold swing, and on-off ratio were extracted. Saturation mobility and on-off current ratio in 1 wt% tin doped ITO TFT was higher than those in 10 wt% tin doped ITO TFT. In conclusion, we found that low doping level of tin in ITO is more proper to transparent TFT with room temperature process.

A planar distribution of source material and a promoter enabled centimeter-scale growth coverage of molybdenum disulfide (MoS₂) monolayers by conventional ambient pressure chemical vapor deposition (APCVD). The molybdenum source and sodium chloride promoter were spin-coated on select planar substrates like silicon, sapphire and quartz. This technique was found to contribute to stable and consistent growth of MoS₂ monolayers. Results of Raman and photoluminescence spectroscopy attested to the high crystalline quality of the as-grown layers, while results of atomic force microscope verified the monolayer film thickness. The electronic quality of the films was found to depend on the growth substrate; the study enabled by fabrication of field effect transistors with MoS₂ monolayers functioning as the active layer. It was experimentally determined that MoS₂ monolayers grown on sapphire substrates and contacted by two Cr/Au electrodes provides the most effective ohmic contact with very low contact resistance. On the other hand, MoS₂ monolayers grown on silicon and contacted by identical metal contacts yield high contact resistance, inadvertently affecting transport performance of devices with and without gate bias. This work on the effective distribution of source and promotors on the growth substrate for the synthesis of high-quality MoS₂ monolayers sheds light on the growth mechanism and the critical role of promoters on growth. The correlation between electrical performance and growth substrates implicates the role of interface effects and its study provides a path to improve device performance of MoS₂ monolayers.

Friction is inherently a process of energy dissipation where electronic and phononic contributions are considered as two primary channels of dissipative process. Mechanical strain has been demonstrated to be able to reversibly control the phase transition of few-layer MoTe₂ semiconductor (2H) to metal phase (1T’) at room temperature under ambient conditions. Reversibly controlling of the phase transition enabling the monitor of electron concentration in a same region in few-layer MoTe₂, which make it an ideal platform to study the electronic effects in nanotribology.
Herein, the nanoscale friction is measured in situ in few-layer metallic and semiconductor MoTe₂ by atomic force microscopy (AFM) to extract the electronic effects, where the phase transition is controlled by mechanical strain. First of all, mechanically exfoliated MoTe₂ thin film is transferred onto SiO₂/Si substrate pre-coated with 200 nm gold by the wetting transfer method. Due to the intrinsic roughness of the thermal evaporation Au film, a tensile-strain-modulated phase transition occurred when a certain amount of normal force is applied in MoTe₂. Both the semiconductor and metal phase region existed, which has been confirmed by Raman spectra and conductive AFM (CAFM). Then, we use AFM to characterize the nanoscale friction on the few-layer MoTe₂ while in situ record the current by using CAFM to characterize the phase transition. Compared with the semiconductor region, the metal region shows significantly higher current and greater friction. And the current between the metal phase MoTe₂ and the tip is positively correlated with the friction force. Furthermore, we changed bias between conductive tip and substrate, the results show that the friction in metal phase region increase with the increase of the external bias and can be regulated by an external electric field. The experimental results reveal that much higher electron concentration in 1T’-MoTe₂ cause a remarkable increase of friction force implying that friction energy can also be dissipated through the electron channel.
Draw support from the few-layer phase transition material MoTe₂, we are able to extract the electronic contributions in friction energy dissipation avoiding the effects of deformation, surface roughness, and changes in atomic structure. This work facilitates an understanding of electronic effect in nanotribology and provide a new strategy to study the electronic effects in nanotribology.

The existence of nontrivial Berry phases associated with two inequivalent valleys in graphene can provide interesting opportunities for investigating the valley-projected topological states. Examples of such studies include observation of anomalous quantum Hall effect in monolayer graphene, demonstration of topological zero modes in “molecular graphene” assembled by scanning tunneling microscopy, and detection of topological valley transport either in graphene superlattices or at bilayer graphene domain walls. However, all aforementioned experiments involved non-scalable approaches of either mechanically exfoliated flakes or atom-by-atom constructions. Here we report a new approach to manipulating the topological states in monolayer graphene via nanoscale strain engineering. By placing strain-free monolayer graphene on architected nanostructures to induce global inversion symmetry breaking, we demonstrate the development of giant pseudo-magnetic fields, global valley polarization, and periodic one-dimensional topological channels for protected propagation of chiral modes in strained graphene, thus paving a pathway towards realizing scalable graphene-based valleytronics and strain-induced superconductivity.

Equimolar mixtures of zero-dimensional graphene (SkySpring Nanomaterials, 1-5 nm particle size) and zinc ferrite nanoparticles (Alfa Aesar, 50 nm particle size) were exposed to mechanochemical activation by high-energy ball milling for time intervals of 0-12 hours. Their structural and magnetic properties were analyzed by Mossbauer spectroscopy and magnetic measurements. The spectra of zinc ferrite milled without graphene were fitted with one quadrupole-split doublet (quadrupole splitting 0.5 mm/s, isomer shift 0.23 mm/s) and indicated that zinc ferrite was superparamagnetic. The line width of the doublet increased from 0.41 to 0.64 mm/s, which correlates with a reduction in particle size as effect of the ball milling processing performed. When graphene was added to the milling powders, the Mossbauer spectra showed the appearance of another quadrupole doublet, with a quadrupole splitting of 0.84 mm/s and an isomer shift of -0.38 mm/s. Its abundance to the spectrum remained constant to 4.48% while the milling time was increased. This second doublet could be related to carbon atoms occupying neighborhoods in the proximity of iron atoms. Hysteresis loops were recorded in an applied magnetic field of 5 T at a temperature of 5 K. A change in the approach to saturation of the loop was observed, with saturation being achieved for the sample milled for 12 hours with graphene. Zero-field-cooling-field-cooling (ZFC-FC) was performed on all samples between 5-300 K with an applied magnetic field of 200 Oe. Graphene was found to stabilize the magnetic properties of the milled system of powders to a blocking temperature of about 90 K.
NSF-DMR-1002627

Magnetoresistance, the change in electrical resistance under an external magnetic field, has been widely used in magnetic sensing, recording and memory devices. Even though the mass production of single-layer graphene has entered the stage of commercialization, the magnetoresistance of single-layer graphene yielded limited success at room temperature, usually ranging from 60% to 775 % at 9 T which were achieved through various routes, such as decorating with gold nanoparticles, by fluorination and nitrogen doping, or through phonon-mediated effect. In this talk, we demonstrate a colossal magnetoresistance of single-layer graphene, nearly 5,000% at 9 T and 300 K, achieved by forming a graphene/perovskite-oxide heterostructure. For single-layer graphene fabricated on various perovskite oxides, we demonstrated a universal twentyfold enhancement in magnetoresistance compared to previous single-layer graphene devices at the same conditions. Combining scanning tunneling microscopy with theoretical calculations reveals that the colossal magnetoresistance stems from an inhomogeneous charge distribution due to the strong interfacial coupling between graphene and perovskite oxides, which creates a random resistor net-works in graphene. Our results suggest that the perovskite oxides can be an appealing substrate for high performance graphene-based magnetoelectronics.

Scanning transmission electron microscopy (STEM) has been a ubiquitous characterization platform for reconstructing atomic-resolution images from highly scattered electrons or from the direct electron beam. These signals have become the standard for imaging in the electron microscope because of their ease of collection and interpretability. Yet in these imaging modes, only a single contrast value is extracted at each scan position; all other information about the specimen encoded in the electron diffraction pattern is thrown away. Previous attempts to utilize the full electron scattering signal have been limited by electron detectors based on scintillators and charge-coupled devices which have slow acquisition speeds and limited dynamic ranges. Recently, a new generation of high-speed and high-dynamic range detectors have allowed for fast collection of the full electron diffraction pattern at each scan position from which information is no longer limited to conventional imaging techniques - but can extend beyond this to uncover quantitative measurements of thickness, strain, tilt, polarity, atomic fields and the long-range electromagnetic fields. Here, we highlight the electron microscope pixel array detector (EMPAD)¹ whose previous demonstrations have shown broad, cross-disciplinary impact, this includes: (1) the development of the highest resolution electron microscope in the world², (2) quantitative imaging of fields from polarization vortices in ferroelectrics³, and (3) strain and dislocation mapping of 2-dimensional (2D) materials with picometer accuracy⁴.

Here, the high-dynamic range of the EMPAD is extremely advantageous when capturing the full scattering signature of 2D materials. In 2D materials, the incident electron beam tend to dominate the scattering signal such that intensities from diffracted spots are washed out by the incident beam. Using the high-dynamic range and high-speed of the EMPAD, we can map strain in 2D materials with high-precision at micrometers field of view; this method can be used to study lattice distortions to identify where lattice strain and dislocations most commonly occur⁴. With the EMPAD, physical properties of 2D materials can be investigated and directly imaged to provide an interesting playground for exploration where new material physics can be recovered without the complications from multiple scattering.

1 Tate, MW. et al. Microscopy and Microanalysis 22, 237-249 (2016).
2 Jiang, Y. et al. Nature 559, 343 – 349 (2018).
3 Yadav AK,* Nguyen KX*. et al. Nature 565, 468-471(2019).
4 Han, Y, Nguyen, KX. et al. Nano Letters 18, 3746 – 3751 (2018).

Atomically-thin two-dimensional (2D) materials exhibit extraordinary optical, electronic, and mechanical properties, which make them promising material platforms for optoelectronic applications. Monolayer molybdenum disulfide (MoS₂) is one of the widely used semiconducting 2D materials for photosensing owing to its direct bandgap. In addition to its intrinsic material properties, mechanical strain applied to the semiconducting material can modulate intrinsic electronic properties, such as bandgap. In this presentation, we report a mechanically self-assembled, crumpled MoS₂ photosensor with graphene electrical contacts. Graphene was used as electrodes because it is mechanically robust and electrically conductive 2D material forming barrier-free electric contact with MoS₂. Furthermore, by crumping monolayer MoS₂, we introduced strain gradient over crumpled MoS₂ structure, and such strain gradient enables nanoscale bandgap modulation, which induces exciton drift with light illumination. The exciton drift enables higher photoresponsivity in crumpled MoS₂ photosensor, compared to photoresponsivity of flat MoS₂ photosensor. We demonstrate power dependent photoresponsivity of our crumpled MoS₂ photosensor as well as local photocurrent generation with photocurrent mapping. Our approach to mechanical self-assembly of atomically thin materials offers a simple but effective way to enhance photosensitivity and suggests a new way to engineer exciton funneling in 2D material system.

Nano-electro-mechanical resonators (NEMR) consisting of atomically thin two dimensional (2D) materials such as graphene are considered to be capable of highly accurate force measurements. The resonance properties of NEMR consisting of optically opaque 2D materials are highly affected by the light irradiation because of the photothermal effect.[1] A hexagonal boron nitride (h-BN) is transparent in the visible light range. Thus, the photothermal effect on the h-BN NEMR would be suppressed. In addition, h-BN is an emergent material for single photon emitters [2] that are ultrabright and stable under ambient conditions. Very recently, theoretical prediction regarding quantum effects in a mechanically modulated single photon emitter composed of h-BN NEMR has been reported.[3] To discriminate the photothermal effect, the light induced effects on the resonance should be clarified. In this study, we investigate the photothermal effect on resonance properties of the electrically actuated drum type h-BN NEMR.
To compose the h-BN NEMR, a multilayered h-BN flake, which was prepared by mechanical exfoliation of bulk h-BN, was transferred using gel transfer method onto a predetermined pair of drum-shaped electrodes on a patterned n⁺Si / SiO₂ substrate. To form the h-BN NEMR suspended by metal electrodes, the SiO₂ layer underneath the h-BN drum (6 μm in diameter) was etched using buffered HF. The frequency resonance of the h-BN NEMR was measured in vacuum by optical detection method, where a laser with a wavelength of 520 nm was irradiated to the center of the h-BN drum as the probe. Additional laser with a wavelength of 685 nm was irradiated to investigate the photoinduced effect on the resonance.
The h-BN NEMR was successfully oscillated by applying the AC + DC bias voltage between the n⁺Si substrate and the Au electrode. The resonance frequency and Q factor were 26.48 MHz and 115, respectively. The oscillation amplitude increases linearly with increasing the AC or DC voltages. This behavior can be explained by the model based on the electrical actuation induced by the dielectric effect. Under additional light irradiation with a wavelength of 685 nm around the center of the drum, the linearities of the resoance properties were maintained under high intensity of 1 mW. Contrary, in the case of graphene drum resonator, which is optically opaque, the resoance property was affected even under the weak laser irradiation less than 10 mW because of the photothermal effect, which results in the modification of oscillation modes between linear and nonlinear.[1] Thus, the resonance mode of the h-BN NEMR is robust againt the photothermal effect owing to the optical transparacy of h-BN.
References
[1]T. Inoue, Y. Anno, Y. Imakita, K. Takei, T. Arie, and S. Akita, ACS Omega, 2 (2017) 5792.
[2]T. T. Tran, K. Bray, M. J. Ford, M. Toth, and I. Aharonovich, Nat. Nanotechnol., 11 (2015) 37.
[3]M. Abdi and M. B. Plenio, Phys. Rev. Lett., 122 (2019) 023602.

Two-dimensional materials such as graphene and MoS₂ represent the ultimate limit of both nanoelectronic and nanoelectromechanical systems due to their intrinsic molecular scale thickness.[1] While 2D materials exhibit many useful properties, many of the most exciting phenomena and applications arise at the van der Waals interface. Electrically, the van der Waals interface enables the constructing of heterostructures and molecular scale electronics. Mechanically, the van der Waals interface displays superlubricity[2] or solitons[3] depending on whether the interface is aligned. A fascinating question is how the van der Waals interface affects the mechanical properties of 2D membranes. Answering this question is important to incorporating 2D heterostructure electronics into diverse applications such as highly tunable nanoelectromechanical systems from suspended 2D membranes, stretchable electronics from crumpled 2D materials, and origami/kirigami nano-machines.
In this study, we explore the impact of the van der Waals interface by comparing mechanical resonance of electrostatically contacted circular drumhead resonators made from atomic membranes of monolayer graphene to commensurate (Bernal stacked) bilayers, incommensurate (twisted) bilayer, and graphene-MoS₂heterostructures (2D bimorph).
For Bernal stacked bilayer, we observe the creation and destruction of individual solitons manifesting as stochastic jumps in the mechanical resonance frequency tuning. We find individual dislocation creation and destruction of single solitons lead to shifts in membrane stress of < 10 mN/m or an in-plane interlayer slip distance of < 1.42 Å. We observe similar jumps in the few-layer graphene and heterostructure, but not in the twisted bilayer.
For twisted bilayer, amplitude dependent studies reveal that the resonators show a factor of 3 to 5 higher dissipation rate at room temperature. A further increase in amplitude leads to a large peak broadening rather than Duffing nonlinear behaviors observed in monolayer and Bernal stacked bilayer graphene resonators. In addition, the interface dissipation is strongly dependent on temperature, displaying tuning of nonlinear behaviors with a decrease in temperature.
These results show that van der Waals interfaces strongly affect stress and dissipation of many multilayer 2D atomic membranes; an important consideration in engineering 2D nanomechanical devices.





[1]Bunch, J. S.; van der Zande, A. M.; Verbridge, S. S.; Frank, I. W.; Tanenbaum, D. M.; Parpia, J. M.; Craighead, H. G.; McEuen, P. L., Electromechanical resonators from graphene sheets. Science 2007,315(5811), 490-493.
[2]Dienwiebel, M.; Verhoeven, G. S.; Pradeep, N.; Frenken, J. W. M.; Heimberg, J. A.; Zandbergen, H. W., Superlubricity of graphite. Phys Rev Lett 2004,92(12).
[3]Alden, J. S.; Tsen, A. W.; Huang, P. Y.; Hovden, R.; Brown, L.; Park, J.; Muller, D. A.; McEuen, P. L., Strain solitons and topological defects in bilayer graphene. P Natl Acad Sci USA 2013,110(28), 11256-11260.

Monolayers of 2D materials have ultrahigh mechanical strength, which enables the application of very large elastic strains (e.g., ~11% for MoS₂ and 25% for graphene). The ability to sustain such large elastic strain offers unprecedented opportunities to engineer the physicochemical properties by applying mechanical strain. In this talk, I will present two examples on the effect of strain on the properties of MoS₂ layer. The first example shows that elastic tensile strain reduces the bandgap of MoS2. When a gradient strain field is applied to MoS2 monolayer, the created bandgap gradient acts as an efficient funnel of photogenerated excitons that leads to enhanced photoluminescence. The second example shows that when elastic strain is applied to the sulfur vacancy on the basal planes of monolayer 2H-MoS2, the strain modifies the local electronic structure and catalytic activity. The proper combinations of S-vacancy and strain allow us to achieve higher intrinsic activity for hydrogen evolution reaction than the edge site of MoS2.

Lattice misalignments between the two layers of a bilayer graphene are found to provide peculiar electromechanical properties of the graphene. An example is the unconventional superconductivity in magic-angle graphene superlattices (Cao, et al. Nature, 556, 2018), and another the localization of electric polarization at the flexoelectric crinkle kinks (Kothari, et al. Roy Soc Proc, 474, 2018). The lattice misalignments are typically represented by the super-lattice moire-interference patterns for translational and rotational misalignments. Electronic force-field interactions, e.g. van der Waals and Coulomb interactions, between the atoms typically relax the super-lattice configuration from the geometrical moire interference pattern. In particular, non-uniform translational misalignments in translation lead to formation of van der Waals dislocations and flexoelectric crinkles. On the other hand, non-uniform misalignments in rotation generate in-plane disclinations and out-of-plane wrinklons.
Here, we present studies of the dislocation, crinkle, disclination and wrinklon characteristics of a bilayer graphene. The characteristics could be investigated with an atomic-lattice interferometry (ALI), which is carried out with a dual-tip AFM interferometer (DT-AFMI). The optical coupling system of the DT-AFMI enables sensing differences between two simultaneous dual-tip scans of sample and reference graphene lattices, which in turn impart the moire fringes of the interference between the sample and reference lattices. The ALI moire fringes indeed reveal the relaxation of the super-lattice configuration. We compare the relaxations with those evaluated by density functional theory (DFT) calculations. Then, presented are correlations between the lattice moire patterns and the electromechanical properties of the bilayer graphene. Moreover, looked after are studies on stabilities of the van der Waals dislocations, the flexoelectric crinkles, the in-plane disclinations and the out-of-plane wrinklons in the bilayer graphene.

Abstract: Two-dimensional materials constitute an exciting platform for investigation of both fundamental phenomena and electronic applications. However, many 2D materials are not air stable, making their studies and integration challenging. Here we discuss routes to stabilize such materials, using few-layer black phosphorus, a 2D semiconductor, and chromium triiodide, a 2D magnetic insulator, as examples. The chemical degradation pathways of both materials are accelerated by light. While simple encapsulation by Al₂O₃, PMMA and hexagonal BN (hBN) are sufficient to ensure air stability for few-layer black phosphorus devices, it only leads to modest reduction in degradation rate for CrI₃. We find that minimizing exposure of light markedly improves stability of CrI₃, and CrI₃ sheets sandwiched between hBN layers are air-stable for >10 days. By monitoring the transfer characteristics of CrI₃/graphene heterostructure over the course of degradation, we show that the aquachromium solution hole-dopes graphene. Finally, I will also discuss our latest results on transport measurements of twisted bilayer graphene devices that show both superconductivity and insulating behaviors.

Two-dimensional materials, such as graphene, are promising valleytronic candidates due to the K and K' valleys at the Dirac points. All-electronic control is particularly desirable for device applications. Many proposed setups exploit strain-induced pseudomagnetic fields which act oppositely in the K and K' valleys. This is illustrated for graphene nanobubbles, which can filter or split a charge current into its different valley components [1]. Experimental approaches in this direction are advancing, but the most promising signatures of valley-dependent phenomena have instead emerged from graphene/hexagonal boron nitride heterostructures.

Large non-local resistance signals here have been interpreted in terms of a valley Hall effect (VHE) driven by a bulk Berry curvature [2], which in turn emerges from a gap-opening sublattice-asymmetric potential (mass term). A complete understanding of such measurements in terms of either bulk [3]- or edge-driven [4] mechanisms is very much an open question.

Here I demonstrate the emergence of a valley-splitting bulk transport mechanism in the absence of a global band gap [5]. This phenomenon requires instead the presence of local spatial regions (dots) with finite mass. An exact analytic solution to the scattering problem is derived for simple dots, which establishes a strong valley dependence for scattering at low energies. Tight-binding simulations confirm that this behaviour is robust for a wide range of mass distributions, and gives rise to a valley Hall conductivity in periodic systems.

Our findings provide an alternative mechanism for the generation of bulk valley currents and suggest that valley-dependent scattering may give rise to analogous effects to those expected for Berry-phase induced deflections. They also suggest more reliable guidelines for valley engineering, in contrast to proposals which require atomically precise edges or strain profiles.

References:
[1] M. Settnes et al, Physical review letters 117, 276801 (2016).
[2] R. Gorbachev et al., Science 346, 448 (2014).
[3] Y. D. Lensky et al, Physical Review Letters 114, 256601 (2015).
[4] J. M. Marmolejo-Tejada et al, Journal of Physics: Materials 1 (1), 015006 (2018).
[5] T. Aktor et al, in preparation (2019).

Interlayer rotation in van der Waals structures of 2D materials couples strongly to electronic properties and, therefore, has significant technological implications. Nevertheless, controlling the rotation of an individual 2D material flake remains a challenge in the development of rotation-tunable electronics. Here we reveal a general moiré-driven mechanism that governs interlayer rotation, relying on the concept of interlayer or van der Waals (vdW) dislocations. We present a theory to explain the connection between the moiré pattern and the arrangement of vdW dislocations, which can be understood to minimize the interfacial energy in the system. As a demonstration of this theory, we consider the arrangement of vdW dislocations due to growth or rotation of finite-sized MoS₂ flakes on larger crystalline substrates. We compare this to the classic critical thickness problem in thin film mechanics, and explain the small, stable rotation angles in finite 2D material flakes on crystalline substrates that are commonly observed experimentally. We show that by applying strain, and thereby controlling the moiré pattern, it is possible to select the amount of interlayer rotation between flake and substrate. The approach provides a powerful tool for the on-demand design of rotation-tunable electronics.

Recent studies have established synthesis protocols to enable twisting of bilayer graphene to change the stacking order, thereby controlling the resulting moiré pattern. Due to the moiré, electronic transport properties of the twisted bilayer graphene are found to be strongly-dependent on the twist angle, thus yielding a potential new class of low-dimensional carbon electronics. However, being driven by weak van der Waals interactions at the interfaces (~1/10^th of room-temperature KT), these structures are susceptible to nontrivial thermal fluctuations which may severely affect their device level performance. Using large-scale molecular dynamics simulations, we uncover a size dependent thermal stability of Moiré patterns in free-standing twisted bi-layers. We show that small twisted graphene flakes on the order of 10nm in size can rapidly re-orient themselves into stable AB stackings at room temperature, while twisted graphene flakes on the order of 100nm and larger remain thermally-stable even at temperatures exceeding 1000^o C. This size-effect on the thermal stability and kinetic behavior of twisted bilayer graphene is related to the incomplete moiré periodicity of the graphene flakes and is explained using a dislocation-based framework. For thermally-stable twisted graphene flakes, we show that the different stacking arrangements present in the moiré pattern can be used to bias the hydrogen-plasma etching rates. Specifically, the AA and SP regions in the bilayer graphene have much lower barrier energies, compared to the perfect AB stacking, for ion transmission through the top graphene layer for etching of the bottom layer. These results suggest that selective patterning of twisted graphene can be achieved, which is of significant importance in the top-down fabrication of porous graphene.

As atomically thin direct-gap semiconductors with strong light-matter interaction, monolayer transition metal dichalcogenides (TMDs) are promising for a broad range of optical and optoelectronic applications. In TMDs, strong electron-hole interactions due to quantum confinement and reduced dielectric screening produce strongly bound composite quasiparticles, including both neutral (e.g., excitons and biexcitons) and charged species (e.g., trions and charged-biexcitons). Of particular interest are trions, three-body charge-exciton bound states, which simultaneously carry charge and excitation energy, and inherit the robust valley polarization of the parent excitonic states. As charged quasiparticles, trions can be manipulated by electric fields. Furthermore, trions in monolayer TMDs have much stronger binding energy, compared to those in traditional semi-conducting materials of GaN, InN, and GaAs, an important factor to prevent dissociation under applied fields. These unique properties offer new opportunities to manipulate charge and excitonic energy in 2D TMDs, which can be exploited for optoelectronic applications including light emitting and photodetecting devices where directional energy or charge transport is critical.
To date, the promising utilities of trions have been hindered by their short lifetime (∼15 ps), limiting drift length of trions under applied electric fields. The lifetime of trion in TMDs is governed by fast non-radiative decay processes such as trion-phonon interaction and defect-mediated non-radiative recombination. Also, the low trion quantum yield (QY) in monolayer TMDs is another obstacle to realization of trion-based optoelectronic devices. QY of monolayer TMDs is reported to be 0.1 - 8 %. The enhancement of the low QY has been demonstrated through the superacid treatment or the plasmonic cavity coupling techniques. However, the superacid treatment, which passivates charge sources, suppresses trion generation. Also, the cavity coupling approach enhances the spontaneous emission rate by the Purcell effect, thereby preventing the achievement of long trion lifetime. Therefore, the alternative approach to achieve both high trion quantum yield and long lifetime is to improve materials quality, since the low quantum yield and the short lifetime are attributed to the fast, non-radiative recombination rate of trion species in current TMD monolayers. In other words, the currently low QY and trion lifetime primarily arise from mediocre quality of TMDs with point defects concentrations on the order of 10¹²- 10¹³cm².
Here we report near-unity emission quantum yield, ∼89.2 %, by using monolayer MoSe₂ with a significantly lower defect density. Importantly, this produces the photoluminescence (PL) intensity which is mainly contributed to the trion species (>98%). The high trion PL quantum yield in the high-quality monolayer MoSe2 is attributed to greatly suppressed non-radiative decay rate, leading to the long lifetime of trion (∼250 ps) approaching intrinsic radiative lifetime. Moreover, we propose that band-trap impact ionization in the high-quality monolayer MoSe₂ contributes to generation of ultrabright trion light emission over a wide range of doping concentrations. Our experimental results indicate that point defects as charge sources promote efficient trion generation in high quality MoSe₂. The high trion quantum yield and the long trion lifetime in the high-quality MoSe₂ provide the platform to explore new physics, including trion valley Hall effect and Moire excitons in heterostructures, demanding both high quantum yield and long lifetime. The superior optical properties of MoSe₂ are also expected to enhance the performance of the optoelectronic devices including light-emitters and photodetectors. Furthermore, the secondary carrier induced by band-trap impact ionization is expected to improve performance of optoelectronic devices, including solar cell and avalanche photodiodes.

Selectively functionalized graphene can realize spatially-defined properties that are highly desirable for atom-thin devices. Because curvature tunes the surface reactivity of graphene, patterning graphene into regions of different local curvatures can achieve domains with different levels of functionalization. Previous buckling methods were limited by the range of tunability over curvature and control of the orientation at the microscale. This presentation describes a scalable approach to achieve spatially selective graphene fluorination using multiscale wrinkles. Graphene wrinkles were formed by relieving the strain in thermoplastic polystyrene substrates conformally coated with fluoropolymer and graphene skin layers. Chemical reactivity of a fluorination process could be tuned by changing the local curvature of the graphene nanostructures. Patterned areas of graphene nanowrinkles and crumples followed by a single-process plasma reaction resulted in substrates with regions having different fluorination levels. We also demonstrated the conductivity of the functionalized graphene nanostructures could be locally tuned as a function of feature size without affecting the mechanical properties.

Defects play an important role in tuning the electronic and optical functionality of 2D materials that are useful for biosensing and nanoelectronics, or in providing the basis for single-photon emission for quantum information processing. We show that patterned graphene with vacancies may offer a route to bandgap generation, and potentially improvement of the I_ON/I_OFF ratio in a field effect transistor (FET) by nanohole passivation, e.g. through hydrogenation. FETs based on patterned graphene with small pores could have a high level of performance similar to graphene nanoribbons, however with the added benefit of no width confinement. In this context, we report on the theoretical characterization of Raman spectroscopy defect-induced intensities in graphene, which is difficult to achieve experimentally. Defective graphene oxide nanostructures are investigated in comparison to experiment, where we find that the Raman I(D)/I(D’) intensity ratio decreases with increase in the nanohole size and the number of adsorbed oxygens, explaining the decrease of this characterization signature with increase of the exposure time to oxygen plasma. Predicted Raman spectroscopy intensity ratios are also confirmed by measurements for graphene oxide quantum dots. In addition, we discuss the optical spectra of monolayer transition metal dichalcogenides upon introduction of defects. The tunability in the optical response of mechanically deformed WSe₂ in comparison point defects is described. Finally, Raman intensity calculations demonstrate that a comparison between pristine and 2D WSe₂ with a single vacancy can provide a fingerprint for defect characterization.

Two-dimensional (2D) materials, such as Graphene, hBN and MoS₂, are promising candidates in a number of advanced functional and structural applications, owing to their exceptional electrical, thermal and mechanical properties. Understanding mechanical properties of 2D materials is critically important for their reliable integration into future electronic, composite and energy storage applications. However, ithas been a significant challenge to quantitatively measure mechanical responses of 2D materials, due to technical difficulties in the nanomechanical testing of atomically thin membranes. In this talk, we will report our recent effort to determine the engineering relevant fracture toughness of graphene with pre-existing defects, rather than the intrinsic strength that governs the uniform breaking of atomic bonds in perfect graphene.Our combined experiment and modeling verify the applicability of the classic Griffith theory of brittle fracture to graphene. In another example, we systematically characterize chemical vapour deposition-grown MoS₂by photoluminescence spectroscopy and mapping and demonstrate non-uniform strain in single-crystalline monolayer MoS₂and strain-induced bandgap engineering.Strategies on how to improve the fracture resistance in graphene, and the implications of the effects of defects and strain on mechanical and other functional properties of other 2D atomic layers will be discussed.

Out-of-plane deformation patterns, such as buckling, wrinkling, scrolling, and folding, formed by multilayer van der Waals materials have recently seen a surge of interest. One crucial parameter governing these deformations is bending rigidity, on which significant controversy still exists despite extensive research for more than a decade. Here, we report direct measurements of bending rigidity of multilayer graphene, molybdenum disulfide (MoS₂), and hexagonal boron nitride (hBN) based on pressurized bubbles. By controlling the sample thickness and bubbling deflection, we observe plate-like responses of the multilayers and extract both their Young's modulus and bending rigidity following a nonlinear plate theory. The measured Young's moduli show good agreement with those reported in literature (E_graphene>E_hBN>E_MoS2), but the bending rigidity follows an opposite trend: (D_graphene<D_hBN<D_MoS2) for multilayers with comparable thickness, in contrast to the classical plate theory, which is attributed to the interlayer shear effect in the van der Waals materials.

Due to the directionality of Ti₃C₂T_x and Ti₂CT_x MXene, the fracture initiation property varies with loading directions. Understanding this rarely explored anisotropic fracture behavior of MXene is essential for material design and engineering application. In this work, a series of molecular dynamics (MD) modeling is conducted on mono-layer Ti₃C₂ and Ti₂C MXene. Pre-cracks of both zigzag and armchair directions are created by removing atoms. For both cases, pre-cracked MXene is stretched to the crack initiation point and then energy minimization is conducted with canonical ensemble, where the stress distribution and J-integral are calculated and compared. Both mode-I and mixed mode scenarios with ranging loading directions are conducted to explore the combined effect of structural and loading directionality. Also, mode-I crack initiation scenario of graphene is conducted and compared with the Ti₃C₂T_x and Ti₂CT_x MXene to explore the thickness effect. Results show that the thickness variation leads to difference of stress distribution around the crack tip, thus influences the crack initiation property.

Transition metal dichalcogenides (TMDs) and their alloys have attracted great interest in recent years due to their potential electronic, photonic and energy applications. However, due to the challenges on anisotropic thermal conductivity measurements, the thermal conductivities of layered TMDs and their alloys remain largely unexplored despite their critical roles in the reliability and functionality of TMD-enabled devices. Motivated by this perspective, we employed frequency-dependent time-domain thermoreflectance (TDTR) with variable spot sizes to systematically and accurately measure the anisotropic thermal conductivities of a series of TMD crystals, including MoS₂, MoSe₂, WS₂ and WSe₂, and TMD alloys, WSe_2(1-x)Te_2x, in both the in-plane direction (parallel to the basal planes) and the cross-plane direction (along the c-axis). In both the TMD crystals and the WSe_2(1-x)Te_2x alloys, the cross-plane thermal conductivity was consistently observed to be dependent on the heating frequency (modulation frequency of the pump laser). A two-channel thermal model was used to analyze the experimental data and the frequency dependence was attributed to the non-equilibrium thermal resistance between different groups of phonons in the substrate. In WSe_2(1-x)Te_2x alloys, a clear discontinuity in both the cross-plane and the in-plane thermal conductivities is observed as x increases from 0.4 to 0.6, due to the phase transition from 2H to Td phase in the layered alloys. The temperature dependence of thermal conductivity for the TMD alloys was also found to be weaker compared to the pristine 2H WSe₂ and Td WTe₂ due to the atomic disorder. Our measurements serve as an important baseline to understand the thermal conductivity of TMDs and other layered materials.

Interfaces provide an effective opportunity to enable novel functionalities and avenues where intriguing physics occurs. While pristine interfaces have been mostly studied, random interfaces have thus far been less explored. In this work, we investigate the effects of correlated interfacial disorder on vibrational energy transport, using graphene-hexagonal boron nitride (G-hBN) interfaces as our model system. More than Poisson point process (white noise, uncorrelated) type of randomness, we studied different types of correlated disorder with varying correlation functionals. When disorder is present, in contrast to the conventional scattering theory (e.g. Rayleigh’s theory), our non-equilibrium Green’s function (NEGF) results show surprising enhancement of energy transmission for a wide spectrum of vibrational modes. To underpin the transport physics, we send probing wavepackets across the interfaces with (q, w) modal resolution. These wavepacket simulations demonstrate an identical frequency regime of transmission enhancement predicted by NEGF analysis, and reveal the effects of mode conversion absent for pristine interfaces due to structural disorder. Beyond the above single-particle picture, we calculate modal correlation matrix based on Green-Kubo modal analysis, and show the effects of interfacial disorder on modal correlation, as well as their contribution to overall interface conductance, both diagonal and off-diagonal. This work could help understand the physics of phonon transport across random interfaces, and would be interesting to practical interface designs, such as thermoelectric and acoustic applications.

Graphene is viewed as a promising material with unique properties. Thanks to their excellent thermal conductivity, graphene and related materials can for example be used as composite filler materials. There is, however, a crucial lack of information regarding the connection between transport properties and materials morphology. For computing the thermal transport properties, traditional methods based on systematic enumeration of finite differences scale poorly with system size, especially in simulations with low symmetry and for systems with defects. Here, we combine advanced regression techniques for the extraction of force constants with the T-matrix approach [1] to systematically explore the impact of defects of varying dimensionality on the thermal conductivity. We employ the HIPHIVE [2] package for the extraction of force constants, which exploits crystal symmetries and advanced regression algorithms to dramatically reduces the number of references calculations. As the latter are commonly carried out using computationally expensive first-principles methods, this reduces the computational effort by orders of magnitude.
[1] Katre A, Carrete J, Dongre B, Madsen GKH, Mingo N, Phys. Rev. Lett. 119, 075902 (2017)
[2] Eriksson F, Fransson E, Erhart P, Adv. Theory Simul. 1800184 (2019)

MoSe₂ as one of the promising two-dimensional transition metal dichalcogenides (TMDCs) recently emerged as a promising alternative of graphene for nano-electronic and opto-electronic devices due to its unique transport properties. The inefficient heat removal due to the low thermal conductivity of monolayer MoSe₂ can cause critical challenges for its devices. The defects due to the imperfection of growth processes, such as vacancies significantly reduced thermal conductivity and aggravate the thermal management challenges. Very few studies has focused on exploring the effect of defects on the thermal transport in monolayer MoSe₂. In addition, a fully parameter-free defect model is needed to examine and estimate the thermal properties considering the influence of defects; this requires improvement in the existing empirical models. The first-principles Density Functional Theory (DFT) and the phonon Boltzmann transport equations based simulations are used to study the phonon transport properties of pristine and defective monolayer MoSe₂ with Se vacancies. A model, which is fully parameter-free, is developed using first-principles approach. The impact of three different types of Se vacancies on the thermal conductivity of monolayer MoSe₂ is studied. The findings indicate how the vacancy-induced phonon states and different phonon scatterings are influenced by different types of Se vacancies. The existing Se vacancies redistribute the phonon DOSs by lowering the frequency cutoff and generating defect-induced phonon states, which soften the phonon modes in the dispersion relations and result in a lower phonon group velocity. Furthermore, an empirical model to estimate the effect of Se vacancies is developed to compare with the DFT results. It’s found that the empirical defect model overestimates the influence of point vacancy at low temperature. Additional terms are introduced to the modified empirical model to consider defects providing a better match with the parameter-free DFT simulation. The results from this study will help us understand the mechanism of phonon transport under the influence of different types of point defects. This work provides insights and directions for validating and improving the empirical defect model, which can be used for estimating the material properties considering the influence of defects.

Graphene atop (or sandwiched in) h-BN is, by far, the most investigated 2D material heterostructure. Their weak van der Waals interlayer interaction enables the investigation of exquisite physical properties. On a different path, the present work indicates a van der Waals-to-covalent interaction transition for such heterostructure under pressure. Insulating and conductive states appear depending on the number of graphene layers and pressure. While it benefits from weak van der Waals interaction for fabrication, the present study both induces and investigates the effects of strong interlayer interaction. As a consequence, it shows the emergence of different properties from their parent 2D materials with eventual intricate competing mechanisms. More specifically, scanning probe microscopy and ab initio calculations reveal modifications on the electronic and structural properties of graphene/h-BN heterostructures induced by compression. Using AFM and EFM techniques, with charge injection being made in the heterostructures at different pressures, the charge injection efficiency monotonically decreases with increasing pressure for monolayer-graphene (MLG)+BN heterostructures, indicative of a conductor-insulator electronic modification. Bilayer-graphene (BLG)+BN and trilayer-graphene (TLG)+BN heterostructures show a non-monotonic behavior of charge injection versus pressure, indicative of competing electronic structure modifications. First-principle calculations of these systems indicate a pressure-induced van der Waals-to-covalent interlayer transition, where such interlayer covalent binding, in the presence of water molecules, results in a disordered insulating structure for the MLG+BN case, while it leads to an ordered conducting structure for both BLG+BN and TLG+BN heterostructures.
Besides introducing a distinct perspective on the study of 2D materials and their van der Waals heterostructures, the present work stresses the importance of interlayer interaction in any device based on such heterostructures. For example, flexible optoelectronic devices are expected to suffer a significant strain variation, which may lead to different interlayer interaction regimes, ultimately affecting device performance. In other words, the present work shows that there is a strong role for interlayer interaction in the (initially) weakly-bound van der Waals heterostructures.

Strain is an effective tool to tune physical properties in a wide range of materials. In transition metal dichalcogenides (TMDs), a family of two-dimensional (2D) van der Waals (vdW) solids, strain can alter the interlayer distance, as well as bond strength, length and angle between the transition metal and chalcogen atoms, modifying the interatomic orbital coupling, interlayer wavefunction overlap and valence band splitting. In traditional mechanical bending/stretching experiments, the 2D materials sit on a flexible substrate and strain is determined by the elongation or radius of curvature of the substrate. Any slippage across the sample/substrate interface or imperfect strain transfer across layers can introduce large uncertainties. Moreover, strains generated in stretching/bending experiments are typically only less than 4%. Hydrostatic pressure created in a diamond anvil cell (DAC) can generate compressive strain as high as 30%, without introducing any damage to the samples. DACs have been extensively used in the geophysics field to simulate the high-pressure environment in planetary interiors. Pressure in a DAC is determined by monitoring the fluorescence peak of a ruby crystal placed adjacent to the sample, with an accuracy better than 1 GPa.
Previous studies demonstrated that with about 9% cross-plane compressive strain, molybdenum disulfide (MoS₂) exhibits a semiconductor to metal (S-M) transition, with an electrical conductivity enhancement from 0.03 S/m to 18 S/m. Extreme strain should also have a profound impact on phonon transport properties, which can affect the thermal transport in MoS₂. In this talk, I will present our recent work about tuning thermal transport of MoS₂ under extreme strain. A DAC device is integrated into our recently developed picosecond transient thermoreflectance (ps-TTR) system to measure the strain-tuned cross-plane thermal conductivity (κ) in bulk MoS₂ up to ~19 GPa (over 9% cross-plane strain). We observed roughly a 7x increase of κ, from 3.5 W m^-1K^-1 at ambient pressure to about 25 W m^-1K^-1 at 19 GPa. First-principles calculations and electrical conductivity measurements suggest that this drastic change arises mainly from the substantially strengthened interlayer force and heavily modified phonon dispersions along the cross plane direction. The group velocities of coherent longitudinal acoustic phonons (LAP), measured with coherent phonon spectroscopy (CPS), increase by a factor of 1.6 at 19 GPa due to phonon hardening, while their lifetimes decrease due to the phonon unbundling effect. These results suggest possible parallel tuning of structural, thermal and electrical properties of vdW solids with strain in multi-physics devices.

2D materials have unique structural mechanics [Nano Letters 15 (2015) 1302]
that makes their defects and phase transformations dramatically different from
those of 3D materials. I will focus on light-structure [Nano Letters 18 (2018) 7794] and electron-structure [Science Advances 5 (2019) eaav2252] interactions in this talk, with a range of behaviors such as ferroelasticity [Nature Comm 7 (2016) 10843], ferroelectricity and electronic topology change [Science
346 (2014) 1344] in one to few-layer transition metal dichalcogenides. Design of
ultrafast responsive 2D materials [Nano Letters 18 (2018) 7794] and engineering single-atom dynamics with electron beam irradiation [Science Advances 5 (2019) eaav2252] are demonstrated.

We have utilized data mining approaches to elucidate over 1000 2D materials and several hundred 3D materials consisting of van der Waals bonded 1D subcomponents, or molecular wires. We find that hundreds of these 2D materials have the potential to exhibit observable piezoelectric effects, representing a new class of piezoelectrics. A further class of layered materials consists of naturally occurring vertical hetero structures, i.e. . bulk crystals that consist of stacks of chemically dissimilar van der Waals bonded layers like a 2-D super lattice. We further combine this data set with physics-based machine learning to discover the chemical composition of an additional 1000 materials that are likely to exhibit layered and two-dimensional phases but have yet to be synthesized. This includes two materials our calculations indicate can exist in distinct structures with different band gaps, expanding the short list of two-dimensional phase change materials. We find our model performs five times better than practitioners in the field at identifying layered materials and is comparable or better than professional solid-state chemists. Finally, we find that semi-supervised learning can offer benefits for materials design where labels for some of the materials are unknown.

Heterogeneous interfaces between two dissimilar materials are an essential building block for modern semiconductor devices. The 2-dimensional (2D) van der Waals (vdW) materials and their heterostructures provide a new opportunity to realize atomically sharp interfaces in the ultimate quantum limit for the electronic and optoelectronic processes. By assembling atomic layers of vdW materials, such as hexa boronitride, transition metal chalcogenide and graphene, we can construct atomically thin novel quantum structures. We demonstrate the enhanced electronic optoelectronic performances in the vdW heterostructures, suggesting that these a few atom thick interfaces may provide a fundamental platform to realize novel physical phenomena. In this presentation, we will discuss two topics associate with the commensuration and incommensuration between the vdW layers. Many exotic physical phenomena occur associated with the incommensurability of the moiré superstructures; the fractal energy spectrum of Hofstadter butterfly and recently discovered Mott insulating and unconventional superconducting behavior of the ‘magic’ twist angle bilayer graphene have demonstrated the wealth of the nontrivial topology of electronic band structures. We find that the vdW interaction energy that favors interlayer commensurability competes against the intralayer elastic lattice distortion to form a quasi-periodic domain structure, inducing profound changes in electronic structure. Particularly, we show quantitative analysis of the engineered atomic-scale reconstruction completely controlled by the twist angle between two graphene layers and anomalous electron transport occurring in the network of topologically protected propagation modes along the domain boundaries.

The buckling of graphene via strain relief of elastomeric substrates can engineer its physical properties such as electrical conductivity and plasmon resonances without using lithographic processes. Because the wavelength and orientation of resulting crumples can change in response to applied strain, graphene functionalities can be dynamically modulated by stretching or bending the substrate. Graphene crumples, however, have been limited by crack formation under tensile strain, particularly in delaminated regions. Furthermore, area-specific tuning of structural parameters and hence properties is not possible because only textures with globally uniform wavelength and orientation can be produced. This presentation will describe a conformal wrinkling strategy that creates crack-free graphene nanostructures with locally different wavelengths on an elastomeric platform. By sandwiching fluoropolymer skin layers with spatially varying thicknesses between graphene and the substrate, we formed multiscale graphene wrinkles with predetermined wavelengths after strain relief. Because delamination was suppressed with the presence of the skin layer, the wrinkle orientation could be switched under cycles of stretching and releasing without significant cracking—even beyond the intrinsic fracture limit of graphene. Through mechanics modeling, we revealed that the fluoropolymer layer mediated structural evolution of the graphene wrinkles without cracking via conformal adhesion. With exquisite control over wrinkle topography, our crack-free, multiscale wrinkling strategy will be useful for optoelectronics and plasmonics based on graphene and other two-dimensional nanomaterials.

I will review our group’s work on the molecular transport properties of angstrom-scale channels fabricated by der Waals assembly of 2D crystals. These channels can be viewed as if individual atomic planes were extracted from a bulk crystal leaving behind two edge dislocations with an empty space in between. Gas, water, ion and proton transport have been studied in capillaries down to one atom in height.

Absence of surface dangling bond on atomically thin two-dimensional materials provides epitaxy researchers with novel opportunities of materials combination and technique to overcome materials compatibility issues. Crystalline 3D semiconductor thin films, such as silicon (Si) and germanium (Ge), on 2D materials was grown on various 2D materials by chemical vapor deposition. Study of controlling nucleation of conventional 3D semiconductors on 2D materials offers a novel strategy of materials growth. Moreover, van der Waals heterostructures composed of 2D and 3D materials exhibit formation of novel band structures via charge transfer. The presentation shows a few general aspects of formation of 2D/3D heterostructures and their physical properties different from those of individual components. The unique characteristics of 2D/3D heterostructures implying potentials of semiconductor device applications will be also discussed.

Direct growth of transition metal dichalcogenides (TMDs) such as WS₂ and WSe₂ on graphene and hexagonal boron nitride (hBN) is of interest for the fabrication of 2D heterostructures as an alternative to exfoliation and 2D layer stacking. Commensurability between the TMD lattice and graphene or hBN is favorable for epitaxy, however, it is difficult to control nucleation and wetting of the TMD on pristine van der Waals surfaces due to the low surface energy and lack of chemical bonding. In addition, mirror twins and inversion domains can form due to rotational symmetry which introduces grain boundaries in coalesced TMD films. However, the presence of defects in hBN and graphene can dramatically alter the energy landscape providing an alternative route to control nucleation and epitaxy.

Our recent studies have focused on the epitaxial growth of WSe₂, WS₂ and related TMD monolayer films by gas source chemical vapor deposition (CVD) on hBN and graphene. The CVD process is carried out in a cold-wall reactor using metal hexacarbonyls and hydride chalcogen precursors in a hydrogen carrier gas. In the case of growth on epitaxial graphene on SiC, nucleation of WS₂ and WSe₂ occurs primarily at wrinkles and step edges resulting in different WS₂ rotational orientations. In the case of WSe₂ growth on hBN, however, single atom vacancies in the hBN were found to act as sites for metal atom trapping which facilitates the nucleation of WSe₂ domains on the surface. In this case, the TMD nucleation density can be controlled by manipulating the density of surface atom vacancies which can be achieved through plasma irradiation and annealing in NH₃. In addition, the metal atoms break the surface symmetry which leads to a preferred orientation (~95%) for WSe₂ domains on hBN. Through careful control of nucleation and extended lateral growth time, fully coalesced WSe₂ monolayer films on hBN were achieved which exhibit optical and transport properties superior to comparable films grown on sapphire substrates. The results demonstrate the important role of defects in nucleation and epitaxial growth of 2D heterostructures.

Strain engineering is widely used for tuning the properties of materials and thereby realizing high performance electronic and optoelectronic devices. In bulk three-dimensional materials, the strain-controlled tunability is, however, limited by low values of ‘failure strain’ which triggers formation of lattice defects. On the other hand, two-dimensional (2D) materials such as graphene and layered transition metal dichalcogenides (TMDs) can tolerate much greater strain, and thus are perfect materials for practically implementing the ideas of strain engineering. [1] Among many different methods to introduce strain in TMDs, mismatch between coefficients of thermal expansion (CTE) of a growth substrate and a TMD layer is one of the most promising method mainly due to two reasons: (1) large-area strained TMDs can be directly prepared through bottom-up synthesis such as chemical vapor deposition (CVD) or molecular beam epitaxy (MBE); (2) the induced strain is maintained intrinsically without any external force or treatment since the as-grown TMD is anchored to the substrate at the fabrication stage. Previous reports on CTE mismatch of CVD-grown TMDs have been only able to achieve tensile strain or low compressive strain (<0.2%). [2] While tensile strain normally decreases the band gap of TMDs, compressive strain tends to increase the band gap. Moreover, recent theoretical and experimental studies [3, 4] have shown that a large compressive strain can be used to stabilize a nominally unstable 1T TMD phase, which is attractive for applications including hydrogen evolution and low-contact-resistance electronic devices. [5]

Here, we use single crystal quartz substrates, which have large CTE compared with MoS₂, to achieve compressive strain up to 1%. The strong built-in strain is indicated by the increased photoluminescence and optical absorption energies as well as ‘stiffening’ of characteristic A_1g and E¹_2g Raman modes. Owing to the different CTE of single crystal quartz along the a- and c-axes, MoS₂ grown on different facets experiences different levels of strain. Specifically, the layers grown on the c-plane exhibit higher strain compared to those prepared on the x-plane which is direct consequence of the higher CTE. Additionally, the MoS₂ layers grown on the AT-cut quartz are highly aligned due to the strongly anisotropic nature of the underlying quartz surface. [6]

The ability to tune electronic properties of TMD layers by engineered strain can be especially useful for controlling interfacial interactions in complex hetero-structures comprising TMD materials as electronically and/or optically active components. Presently, we are exploring this capability for manipulating charge and energy flows in multi-dimensional materials assembled from TMDs and colloidal quantum dots.


1. Deng, S., Sumant, A. V., Berry, V. Strain engineering in two-dimensional nanomaterials beyond graphene. Nano Today 22, 14-35 (2018)
2. Ahn, G. H., Amani, M, Rasool, H., Lien D.-H., Mastandrea, J. P., Ager III, J. W., Dubey, M., Chrzan, D. C., Minor, A. M., Javey, A. Strain-engineered growth of two-dimensional materials. Nat. Commun. 8, 608 (2017)
3. Li, X., Wu, M., Xu, B., Liu, R., Ouyang, C. Compressive strain induced dynamical stability of monolayer 1T-MX₂ (M=Mo, W;X=S, Se). Mater. Res. Express 4, 115018 (2017)
4. Shang, B., Cui, X., Jiao, L., Qi, K., Wang, Y., Fan, J., Yue, Y., Wang, H., Bao, Q., Fan, X., Wei, S., Song, W., Cheng, Z., Guo, S., Zheng, W. Lattice-mismatch-induced ultrastable 1T-phase MoS₂–Pd/Au for plasmon-enhanced hydrogen evolution. Nano Lett. 19, 2758-2764 (2019)
5. Kappera, R., Voiry, D., Yalcin, S. E., Branch, B., Gupta, G., Mohite, A. D., Chhowalla, M. Phase-engineered low-resistance contacts for ultrathin MoS₂ transistors. Nat. Mater. 13, 1128-1134 (2014)
6. Wu, L., Yang, W., Wang, G. Mechanism of substrate-induced anisotropic growth of monolayer WS2 by kinetic Monte Carlo simulations. NPJ 2D Mater. Appl. 3, 6 (2019)

Petroleum coke (PC), a cheap and readily available hydrocarbon feedstock has great potential as a starting material for synthesis of high value graphitic materials. Graphene produced by traditional graphitization of carbon materials at up to 3000 °C [1], and subsequent exfoliation, comes at a large energy cost. By utilising chemical vapour deposition (CVD), graphitic materials can be produced at far lower temperatures, thus providing a large cost and energy saving. Direct exfoliation of high quality coke (carbon content > 96 wt%) has been reported [2], however, lower quality coke requires modification prior to exfoliation.
In this work we present evidence for the synthesis of graphene directly from low quality PC powder by CVD growth from defective graphitic edges. Subsequent exfoliation of the coke materials after treatment allows for the isolation of high quality graphene flakes. We investigate low temperature plasma treatment of the coke powders in order to yield functionalised graphitic materials, with potentially useful properties. We examine the strain behaviour of the PC-derived graphene and their plasma functionalised analogues in order to understand the role of functional defects upon the properties of the graphene. Finally, we investigate the potential of PC for carbon nanotube (CNT) production by seeding coke powder with transition metal nanoparticles, thus synthesising graphene/CNT hybrid powders.
[1] A. Oberlin, Carbon, 22, 521-541 (1984)
[2] U. Sierra, P. Álvarez, C. Blanco, M. Granda, R. Santamaría and R. Menéndez, Carbon, 93, 812-818 (2015)

Green Chemistry protocols have been recently applied to the design of nanomaterials, being the preparation of gold and silver nanoparticles successful examples of this trend. However, the preparation of graphene still offer challenging opportunities for the development of green technologies, involving the use of natural sources, safe solvents and energy efficient processes. Moreover, the implementation of the concept safe by design also applies to the preparation of nontoxic graphene products.
During this talk we will try to give answers to these questions: Can we design an aqueous soluble graphene material, suitable for being employed in biomedical studies, without changing graphene chemical characteristics? Can this design be optimized for Green Chemistry principles?
Following our previous experience in the preparation of graphene materials, we will describe the use of a mechanochemical treatment, in solvent-free conditions, using carbohydrates to exfoliate graphite in a simple and green approach. Among glucose, fructose and saccharose, the former shows the best behavior in terms of exfoliation, generating graphene materials with a relatively low number of defects. Once exfoliated, glucose can be washed and recovered, just by using water, and the final graphene material can be thoroughly characterized. Moreover, the addition of molar equivalents of water to the ball milling treatment, allows the formation of glucose-graphene cocrystals. Analogously to what happens with drugs, glucose-graphene cocrystals display higher performance in terms of stability in aqueous dispersions. In addition, we will show preliminary toxicological tests illustrating the safety of the designed material

Here, we present an effective fabrication method of graphene liquid crystalline fibers to attain high mechanical and electrical properties simultaneously inspired by mussel adhesive polydopamine (PDA). Two-step defect engineering was designed relying on bioinspired surface polymerization within GO dispersion and subsequent solution infiltration of PDA to improve the intrinsic limitation of graphene fibers arising from structural defects of graphene layers during wet-spinning process. For a straightforward understanding of the PDA-induced defect engineering mechanism, interfacial adhesion between GO flakes was carefully analyzed by AFM pull-off test. In addition, PDA could be converted into N-doped graphitic domain within the fiber structure by following pyrolysis resulting in mechanically strong fibers without scarifying electrical conductivity. This bioinspired graphene-based carbon fiber retains a great potential for a wide range of applications such as flexible electronics, multifunctional textiles, and wearable sensors.

Industrial processes such as natural gas separation, hydrogen production, etc. rely heavily on gas separation.¹ For instance, hydrogen reforming from methane uses several techniques such as metallic, silica, zeolite, and polymer membranes for gas separation.¹ The thickness of these membranes can range from tens of nanometers to microns.¹ The efficiency of these membranes depends heavily on permeance through the membrane as well as gas selectivity. Ideally, a membrane would have pore sizes close to the kinetic diameter of the desired gas, while being as thin as possible to increase permeance.¹ 2D materials with high impermeability, high aspect ratio, high strength, chemical inertness and atomic thickness may be able to fulfill this need.

Graphene has been shown to be an extremely impermeable barrier, largely due to its unique structure. Graphene is composed of C with sp² bonds in plane with π orbitals perpendicular to the surface, which form a delocalized cloud repelling other molecules.² It was shown that mechanically exfoliated, pristine graphene is impermeable to He.³ Thus, any permeability must be induced by adding pores to the system. Adding uniformly distributed pores of a specific size has been a challenge, and has not yet been demonstrated with an adequate level of control. Other groups have focused on graphene synthesized by chemical vapor deposition (CVD) due to its scalability compared to mechanically cleaved graphene; however, transferred CVD graphene often has holes from the transfer process.⁴ These holes seriously degraded the selectivity of the graphene composite membrane so that it was only selective to molecules with diameters of > 1 nm, and smaller diameter gases were able to pass freely.⁴ Groups have tried to alter transfer methods, plug the induced holes, or stack graphene sheets, but these transfer-induced defects make it difficult to study the effects of intrinsic defects in CVD graphene, such as grain boundaries and point defects.

In this work, single layer graphene was synthesized by CVD on a Cu foil catalyst. H₂ is permeable in many metals, including Cu. Thus, the permeability of CVD grown graphene can be measured directly while still on the catalyst using a gas driven technique. The difference in permeability of the graphene-coated Cu from the bare Cu can be attributed to the permeability of the graphene. The graphene-coated samples consistently showed lower permeability compared to the bare Cu sample. Furthermore, since pristine graphene has already shown impermeability to gases, any permeability through the graphene should be through intrinsic defects like grain boundaries and point defects in the graphene. Graphene grain sizes with varying orders of magnitude were synthesized, and the H₂ permeabilities measured. This work is able to systematically measure and model the effects of grain size by measuring gas permeation of graphene that is not transferred (so that cracks are not induced), and the only permeation must be through graphene defects. Thus, graphene with different grain sizes can be engineered to create membranes with different permeabilities to H₂, without needing to induce nanopores or fix transfer-induced cracks.

The project depicted is supported by the Savannah River National Lab (SRNL) Laboratory Directed Research and Development program (LDRD-2017-00028). SRNL is managed and operated by Savannah River Nuclear Solutions, LLC under contract no. DE-AC09-08SR22470 with the U. S. Government. This work was performed in part at the Georgia Tech Institute for Electronics and Nanotechnology, a member of the National Nanotechnology Coordinated Infrastructure, which is supported by the National Science Foundation (Grant ECCS-1542174).

1. Jiang, D., et al., Nano letters 2009, 9 (12), 4019-4024.
2. Berry, V., Carbon 2013, 62, 1-10.
3. Bunch, J., et al., Nano letters 2008, 8 (8), 2458-2462.
4. O’Hern, S., et al., ACS nano 2012, 6 (11), 10130-10138.

Graphene is a highly desirable material for a variety of applications due to its inherent material properties. However, realising the full potential of these properties has proved challenging due in part to difficulties in uniformly incorporating it within other media. One route towards tailoring the properties of graphene to allow better interaction with other materials and thereby improving the ultimate product performance is through functionalisation. For example, in nanocomposites the incorporation of functionalised graphene can lead to changes in tensile strength, optical transparency and electrical and thermal conductivity. However, often the resulting structure of the functionalised graphene material, and the location and level of the chemical species added via different functionalisation processes is not known or understood.

In this study we characterise a commercially available powder containing few-layer graphene (FLG) flakes, and the subsequent composite products, after both plasma and chemical functionalisation of the FLG with amine and -OH groups. Alongside confocal Raman spectroscopy, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM) results, tip-enhanced Raman spectroscopy (TERS) and secondary ion mass spectrometry (ToF-SIMS and NanoSIMS) were used to show chemical changes in the FLG material at both the nanometre and micron scale and monitor the dispersion of the flakes within the composite for the various functionalisation processes. [1] These are the first reported TERS measurements on commercially available graphene materials and show the location of the defects associated with the attachment of functional groups to the FLG. Nanoscale TERS images of the D-peak intensity show changes within flakes [2] and variations between the different types of functionalisation, indicating that it is possible to discriminate between functionalisation taking place at the edge of the flakes or within the basal plane.

The plasma functionalisation process was observed to produce flakes with higher levels of functionalisation than chemical functionalisation. The amine functionalisation via plasma process created defects in the basal plane of the flakes, whereas the –OH functionalization appeared to have functional groups present predominantly at the edges of the flakes. ToF-SIMS depth-profiling results of nanocomposites subsequently produced with polyurethane suggested that the plasma –OH material agglomerated the least within the polymer matrix. However, the non-functionalised material produced the largest improvement in mechanical properties for the nanocomposite.

[1] Structural, chemical and electrical characterisation of conductive graphene-polymer composite films, B. Brennan, S.J. Spencer, N.A. Belsey, T. Faris, H. Cronin, S.R.P. Silva, T. Sainsbury, I.S. Gilmore, Z. Stoeva, A.J. Pollard; Appl. Surf. Sci., 403, 403-412 (2017).

[2] Probing individual point defects in graphene via near-field Raman scattering S. Mignuzzi, N. Kumar, B. Brennan, I. S. Gilmore, D. Richards, A. J. Pollard, and D. Roy; Nanoscale, 7, 19413-19418 (2015).

Metal chalcogenides that can be isolated as 2D semiconducting layers form a diverse family of materials with unique properties. Defects that are found or can be created in these materials are an important route for controlling their chemical reactivity. This talk will highlight our recent work on several defect-mediated chemical modifications of 2D metal chalcogenides. First, the covalent functionalization of the basal planes of the transition metal dichalcogenides (TMDCs) MoS₂, MoSe₂, WS₂, WSe₂, and the pnictogen chalcogenides (PCs) Bi₂S₃, and Sb₂S₃ using aryl diazonium salts occurs at initial point defect sites and edges, but propagates across the surface in a chain-like morphology. This functionalization scheme is studied in detail using density functional theory (DFT) and kinetic modeling for MoS₂, and the diazonium salts can also be modified to enable tethering of proteins to the MoS₂ surface. Second, MoO₃ and WO₃ nanoscrolls are formed by transformation from 2D MoS₂ and WS₂ nanosheets, respectively, via atmospheric plasma treatment which occurs due to the combination of different radicals in the plasma that simultaneously form defects and cracks in the 2D layers and oxidize them, so that the strain of the altered crystal structure is relieved by rolling into nanoscroll geometries. Finally, self-assembled monolayers of fullerene molecules on the surface of WSe₂ are studied using atomic resolution scanning tunneling microscopy to show that the molecules exhibit a 2x2 rotational superstructure due to the interplay of molecule-molecule, molecule-substrate, van der Waals, and Coulomb interactions to form a charge redistribution network with long-range ordering.

Flexoelectricity is the strain gradient (direct effect) or electric field gradient (converse effect) induced electromechanical response in a dielectric material. Unlike piezoelectricity, flexoelectricity is ubiquitous in materials, not limited by inversion asymmetric crystal structure. However, flexoelectricity is limited in bulk material, because any significant amount of polarization would require large amount of strain gradient which is limited by mechanical fracture. Thanks to the atomically thin dimension, 2D materials can withstand large amount of strain gradient without mechanical failure. Therefore, flexoelectricity has enabled new opportunities in micro or nano electro-mechanical systems. However, a study of the mechanism of this universal yet underappreciated flexoelectricity in 2D materials has been limited. In this work, we investigated the converse flexoelectric response of 2D transitional metal dichalcogenides (TMDCs) by piezoresponse force microscopy (PFM). PFM is widely used technique to quantify effective piezoelectric coefficients, providing a direct correlation between applied electric field and resulting deformation in materials. Our results showed that the effective out-of-plane piezoelectric coefficient (d₃₃) rises from mono-layer to bi-layer for MoS₂ and WSe₂, then saturates at a slightly reduced magnitude for tri-layer and quad-layer structures. We have correlated the d₃₃ trend observed in our TMDC layers with materials adhesion to the substrate, flexural rigidity of the 2D materials and possible charge transfer between the substrate and the materials. We believe that the converse flexoelectric effect observed in 2D TMDCs arises from a combination of the effects of flexural rigidity of the atomically thin materials and charge transfer between the material and substrate. We further report that in-plane tensile strain reduced d₃₃ by restricting the out-of-plane deformation. Our findings can help understanding the fundamentals of flexoelectric behavior in 2D materials and pave the way for next generation actuators and energy harvesting devices.

A defining feature of emergent phenomena in correlated systems is the competition and cooperation between ground states, which can be sensitively tuned by the lattice. This has been explored using hydrostatic pressure, epitaxial lattice-mismatch in films, and uniaxial strain. Manganites provide canonical examples where various metallic and insulating phases, and associated magnetic and orbital ordering, are delicately balanced. Extending the range of lattice control would enhance the ability to access and investigate new phases. Recent advances in the thin film epitaxy of a water-soluble layer enable the synthesis of many types of complex oxides and heterostructures as macroscale freestanding 2D nanomembranes [1] with considerable mechanical flexibility. Here we stabilize extreme tensile strain in nanoscale La_0.7Ca_0.3MnO₃ membranes, exceeding 8% uniaxially and 5% biaxially [2], and study their magnetotransport properties down to cryogenic temperatures. These macroscopic strain states are found to correspond to microscopic lattice parameter changes, i.e. elastic deformation. The systematic control of biaxial strain reveals a rich phase diagram of competing phases, including a dramatic shift in the metal-to-insulator and magnetic transition, as well as a new insulating ground state that can be extinguished by magnetic field. Electronic structure calculations indicate that the insulator consists of charge-ordered Mn⁴⁺ and Mn³⁺ with staggered strain-enhanced Jahn-Teller distortions within the membrane plane. This highly-tunable strained membrane approach provides a broad opportunity to design and manipulate correlated states in 2D quantum materials.

[1] D. Lu, D. J. Baek, S. S. Hong, L. F. Kourkoutis, Y. Hikita, & H. Y. Hwang, Nature Materials 15, 1255 (2016).
[2] S. S. Hong, M. Gu, M. Verma, D. Lu, V. Harbola, A. Vailionis, Y. Hikita, R. Pentcheva, J. M. Rondinelli, & H. Y. Hwang, submitted.

Two-dimensional (2D) transition metal carbides (MXenes) have attracted a great interest as a relatively new and large class of materials with unique electronic and optical properties. Among 20 different kinds of MXenes, Ti₃C₂T_x and Ti₂CT_x are most widely investigated. Understanding of adhesion between MXenes and MXenes or other 2D materials is critically important for 2D layered nanostructure composites fabrication, which are promising candidates for conducting high-performance energy conversion and environmental remediation. Herein, through atomic force microscopy (AFM) experiment on Ti₃C₂T_x-Ti₃C₂T_x, Ti₃C₂T_x-Ti₂CT_x, Ti₃C₂T_x- 1-5 layer graphene interfaces using a Ti₃C₂T_x functionalized spherical tip, adhesion energy was measured respectively. To obtain adhesion energy more accurately, the Maugis-Dugdale theory which considers surface roughness was applied to convert the AFM measured adhesion force to adhesion energy. Our preliminary showed that the adhesion energy between Ti₃C₂T_x and Ti₃C₂T_x or Ti₂CT_x MXenes are in the same level of the adhesion energy between Ti₃C₂T_x and graphene. In addition, no layer dependency was observed with the MXene adhesion measurements.

Ubiquitous periodic structures formed at the nanometer scale on the surface of exfoliated graphite has been interpreted as ripples, stripe domains, 2D molecular patterns, corrugations, and wrinkled networks. The true origin for the formation of these periodic structures on single layer solids and the ways to control their formation and organization are fundamental to their potential utility in band gap engineering in graphene and single layer semiconductors such as MoS2 and WSe2 towards their utility, e.g., as strain sensors. Using both graphite and MoS2, we show here that the nature of the buckling is indeed due to strain induced sub-nm scale periodic buckling and that its pattern can be controlled on atomically flat surfaces using the self-assembled solid-binding dodecapeptides that form 2D self-organized pattern in aqueous solutions forming ordered peptide nanowires commensurate and aligned with the underlying crystalline substrate. We attribute nanobuckling to be due to out-of-plane rippling effect caused by compressive strain within the top single layer introduced during mechanical exfoliation. The electronic behavior of the nanobuckled regions have been examined by scanning tunneling microscopy and spectroscopy show that the tunneling current varies with the peaks and valleys in the buckled region consistent with the tension and compressive stresses that we estimated by computational modeling using density functional theory. The phenomenon is observed both in graphite (semimetal) and MoS2 (semiconductor) may be a strong indication that nanobuckling is a universal phenomenon likely to be present also in other 2D layered solid systems. The results of the conductive atomic force microscopy (C-AFM) was also used to interrogate the electrical properties of the buckled samples of MoS2 showing changes in the gap energy that can be correlated with the compressed and expanded density of states in the top layer. While the study of new physical phenomena is enabled in the nanobuckling by the self-assembled peptides, they also facilitate controlled pattern formation in a wide range of 2D solids in designing new devices by controlling the local chemical and band-gap engineering by local strain. Periodic buckling patterns guided by self-organized peptides may pave a simplistic way for tailoring the surface nanostructures and electronic properties of single atomic layer solids towards designing new devices based on periodic local surface chemical potential and band gap by periodic surface-confined strain. The research is supported by NSF-DMREF program through the grant DMR-1629071 as part of the Materials Genome Initiative.

Graphene has been the promising material for flexible optoelectronic devices because of its outstanding material properties. Many research groups have shown potentials of graphene for various flexible devices such as touch-screen sensors, organic light-emitting diodes and organic photovoltaic devices. However, Graphene still faces some hindrances to overcome for the success of real commercialization.
In this talk, from the commercial perspective, we'd like to share our opinion on most probable application fields of graphene for OLED displays.
Specifically, we will highlight the application of graphene as a transparent electrode for OLED displays because of its extraordinary optical, mechanical and electrical material properties. In addition, we will also emphasize superior moisture barrier properties of graphene which have potentials for the application of OLED encapsulation. Graphene's unique mechanical flexibility combined with barrier property will be able to replace inorganic materials which have been widely used in OLED encapsulations. we will also briefly mention other possible application fields of graphene for OLED display.
For your information, we will also introduce the outline of LG display which has a global No.1 display market share. Finally, we will suggest some requirements and specifications to realize the commercialization of graphene for OLED displays. In my view, this presentation will be a chance to discuss the practical use of graphene for OLED displays.