Session Q01: Growth and Engineering of 2D Materials

Orientated Lateral Growth of Two-Dimensional Materials on C-plane Sapphire

Two-dimensional (2D) semiconducting transition metal dichalcogenides (TMDs) representing the ultimate thickness scaling of channel materials provide a solution to tantalizingly push the limit of technology nodes in the sub-1-nm range. One key challenge with 2D semiconducting TMDs channel materials is the large-scale batch growth on insulating substrates with continuous single crystallinity, spatial homogeneity, and compelling electrical properties. Recent studies have claimed the epitaxy growth of wafer-scale, single-crystal 2D TMDs on C-plane sapphire substrate with deliberately engineered off-cut angles. It has been predominately postulated that exposed step edges break the energy degeneracy of nucleation and thus drive the seamless stitching of mono-oriented flakes. In this talk, I will discuss a more dominant factor that should be considered. The interaction of 2D TMDs grains with the exposed oxygen-aluminum atomic plane establishes an energy-minimized 2D TMD-sapphire configuration. Reconstructing the surfaces of C-plane sapphire substrates to only a single type (symmetry) of atomic planes already guarantees the single-crystal epitaxy of monolayer TMDs without the aid of step edges. Electrical results also evidence the structural uniformity of the monolayers. Our new experimental findings elucidate the long-standing question that curbs the wafer-scale batch epitaxy of 2D TMDs single crystals, an important step toward using 2D materials for future electronics. Experiments extended to other materials like graphydine, and perovskites also support the argument that the interaction with sapphire atomic surfaces is more dominant than the step edge docking.   

Reproducible synthesis of ultrahigh-quality graphene by oxygen-free chemical vapor deposition

Progress in translating advances in growth of graphene films by chemical vapor deposition (CVD) to applications has been hindered by challenges in quality and reproducibility, as well as the lack of a simple model of growth kinetics. Here we show that eliminating trace oxygen leads to fast and highly reproducible CVD graphene growth. We confirm that ppm-level oxygen disrupts growth by etching the graphene edges, and map the boundary between growth and etching in the presence of hydrogen. We demonstrate that trace oxygen causes two major sources of disorder in CVD-grown graphene: pinholes and amorphous carbon deposition. The dependence of growth rate on growth time, temperature, and methane pressure follow straightforward trends, while previously unobserved behavior is seen for hydrogen pressure. Finally, we grow epitaxial graphene on sapphire-supported Cu(111) films. The graphene shows few wrinkles, no evidence of amorphous carbon, and an ultra-low defect density. After dry transfer and device assembly, this graphene shows electrical transport behavior virtually indistinguishable from that of exfoliated graphene, both at room temperature and at low temperature.   

Direct growth of graphene on transition metal oxides via atmospheric pressure chemical vapor deposition

The graphene/transition metal oxides (TMOs) hybrid heterostructures have exhibited the synergetic behaviors, which could not be anticipated in single materials owing to the formation of heterointerface. [1, 2, 3, 4] Typically, the fabrication of hybrid heterostructure predominantly depended on the transfer process of graphene, conventionally grown via chemical vapor deposition (CVD) on metal catalysts, such as Cu, Ni, and Pt with a higher carbon solubility. [5] Consequently, these hybrid heterostructures could not maintain a consistent quality, originated from the typical transfer procedure, including electron doping [6] and mechanical and chemical damage to graphene. [7] The direct growth of graphene onto TMO substrates can avoid these issues, ensuring superior quality and uniformity in the hybrid heterostructures, even within a narrow growth window.

In this study, we directly synthesized graphene on SrTiO₃ (001) substrates at 1100°C, utilizing Ar, H₂, and CH₄ as precursors via atmospheric pressure CVD (APCVD) method. We successfully fabricated full coverage (5x5 mm²) of high-quality graphene on SrTiO₃ substrates with preserving the SrTiO₃ step-terrace structure, confirmed by Raman spectroscopy, atomic force microscopy, and scanning electron microscopy. We found that full coverage of graphene on SrTiO₃ via APCVD necessitates a sufficient deposition temperature and CH₄ flow. Furthermore, we observed that the reduction of the H₂ flow rate improves graphene quality with less residues at the surface. We also conducted the direct fabrication of graphene on other TMO substrates, especially LaAlO₃ and (LaAlO₃)_0.3(SrAl_0.5Ta_0.5O₃)_0.7. Our research offers to standardize the fabrication method for enabling the production of high-quality graphene/TMO hybrid heterostructures using APCVD.

 

 

[1] Kang et al., Adv. Mater, 34, 1803732 (2019).

[2] Park et al., Nano Lett, 3, 1754-1759 (2016).

[3] Kang et al., Adv. Mater, 18, 1700071 (2017).

[4] Shin et al., Adv. Funct. Mater, Early Veiw, 2311287 (2023).

[5] Munoz et al., Chem. Vap. Deposition, 19, 297-322 (2013).

[6] Wu et al., RSC Adv, 9, 41447-41452 (2019).

[7] Yoon et al., Sensors. 10, 3944 (2022).   

Highly controllable growth of carbon nanotube-functionalized graphene

The development of graphene with as-grown nanostructures to enhance its electrical and electrochemical properties is vital for various applications, yet remains a challenge. We employ nanoparticles as a catalyst and use ultra-low concentrations of carbon sources for the controlled synthesis of extended carbon nanotubes on the graphene surface via chemical vapor deposition (CVD). Contrary to previously reported non-transparent graphene and carbon nanotube hybrids, the nanotubes synthesized in our study can achieve a height as low as ~100 nm, leading to highly transparent carbon nanotube-functionalized graphene. Based on results from Kelvin probe force microscopy and conductive atomic force microscopy, the graphene substrate and the carbon nanotubes are covalently bonded. Transport and electrochemical measurements reveal that the as-grown carbon nanotube–functionalized graphene has electrical conductivity comparable to bare CVD graphene, with its electrochemical activity being enhanced fivefold.   

2D Semiconductors from Spin-on Molecular Chemistries for Direct Large-Scale Integration

In this talk, I will discuss our recent developments centered on a process to produce fully soluble solutions from a unique molecular chemistry. This enables the creation of wafer-scale monolayers of MoS₂, a prominent 2D semiconductor for logic and memory devices. This molecule can be dissolved in common solvents and spin-coated onto various substrates, including crystalline substrates, printed electrodes, and polymers. It can also be integrated into gate-all-around (GAA) 3D transistor structures for logic devices. Our method of rapidly synthesizing large-area MoS₂ films, particularly in monolayer form via spin-coating, represents a novelty, as it hasn't been demonstrated previously with a single-source chemistry. This advancement is a significant step toward scalable synthesis of wafer-scale 2D semiconductor thin films, eliminating the need for post-growth wafer transfer techniques, a requirement with films grown by chemical vapor deposition (CVD) methods. Our fabrication process highlights the potential of using a single-source chemical precursor to develop monolayer 2D semiconductors that exhibit commendable transport characteristics and optical properties, which enhance with rising crystallization temperatures. Consequently, it holds promise as a semiconductor for microelectronic transistor channels in both back-end-of-line (BEOL) and front-end-of-line (FEOL) architectures.   

Growth of Atomic Layer TMD Quantum Nanoribbons with Controllable Width

Harnessing additional degrees of freedom provided by the number of the layers, the width, and the strain of two-dimensional transition metal dichalcogenide (TMD) materials opens a new perspective for tuning their properties aiming at applications in quantum electronics and photonics. However, direct growth of TMD nanoribbons with controlled widths and numbers of layers, especially for the appealing width range below 30 nm, remains challenge. Here we report a new method for growing single and double atomic layer of MeX₂ nanoribbons (Me=Mo, W; X=S, Se) with width down to sub-10 nm. The nanoribbon growth is sensitive to substrates and occurs via precipitation from pre-deposited seed nanoparticles with properly selected constituents in a chalcogen vapor atmosphere. The width of nanoribbon is determined by the seed nanoparticle’s diameter. The grown nanoribbons demonstrate remarkable elastic robustness with strain up to ~14%. Width-dependent Coulomb blockade oscillations are observed in the transfer characteristics of MoS₂ nanoribbons with width <20 nm at temperatures up to 80 K, attributed to single electron transfer. Moreover, by applying external strains, TMD nanoribbons generate high performance quantum emission of up to ~90% single photon purity, which is indicative of strain-induced localized electronic states. Our new synthesis method provides a general route for width-controllable growth of families of atomic layer quantum nanoribbons, paving a pathway to the synthesis of novel quantum materials.   

Spatially Structuring the Surface Energy of Monolayer Graphene Through Selective Heterointerface Engineering

Selective bottom-up chemical synthesis of low dimensional materials with high spatial resolution has long been a goal of crystal growers. The challenge, however, lies in the spatial modification of the surface energy landscape of a substrate, a crucial factor that promotes the diffusion and accumulation of adatoms and/or molecules along the surface energy gradient, consequently facilitating nucleation in regions of reduced surface energy. Herein, we demonstrate the achievement of a highly controllable surface energy landscape of monolayer graphene on diamond like carbon (DLC) substrate through a heterointerface containing trapped gallium in a uniquely designed spatial structure. The process involves three steps: (i) spatial-selectively Ga⁺ ion implantation into DLC substrate to create “hill” features with a step height of 4 nm; (ii) polymer-free transfer of monolayer graphene on top of Ga⁺-implanted DLC; (iii) in-situ high-vacuum annealing process above 300°C for gallium precipitation at the graphene-DLC heterointerface with low energy electron microscope (LEEM). During the annealing process, both gallium precipitation and gallium-catalyzed reconstruction of the DLC structure contribute to a shift in the local surface work function of graphene, resulting in a decrease in the surface energy of graphene compared to that of pristine graphene. At 300°C, the surface work function difference (ΔWF)  between graphene residing on the “hill” features and unmodulated region is -142 meV, corresponding to a surface energy difference (Δγ)  between the modulated graphene region and unmodulated region of -0.23 mN/mm. However, at 500°C, ΔWF is 240 meV with a Δγ of -14.5 mN/mm. Notably, this difference remains consistent even upon cooling down to 300°C due to the irreversibility of gallium precipitation. In summary, a surface energy landscape of graphene with a high level of complexity can be realized by carefully tuning annealing conditions and the spatial arrangement of “hill” features, facilitating the selective area growth of materials in various nanofabrication processes.   

Silicene-graphene heterostructure fabrication via interface epitaxy

Graphene (Gr) remains a material of special interest since its discovery three decades ago, even though the two-dimensional quantum materials (2DQM) research has recently evolved beyond Gr and its analogs (called xenes), as these materials have potential for technological applications, including quantum computing and quantum information. However, xenes other than Gr, such as silicene, antimonene, borophene are generally unstable in atmospheric conditions. In order to protect these reactive xenes, the Gr can be used as a protective, inert coating. In the present study we employed in-situ, real-time low-energy electron microscopy/photoemission electron microscopy (LEEM/PEEM) for the investigation of interface chemistry of silicon allotropes and 2D silica structures grown at the Gr-ruthenium interface. Our results show that the Si epitaxial layer can decouple Gr electronically from the ruthenium substrate. Moreover, these experiments prove that a successful scalable synthesis of Gr/√3-Silicene and Gr/Si/SiO₂ heterostructures can be realized. Tuning the growth conditions also provides means to control the growth of 2DQM at a large scale. The strategy of interface epitaxy is a promising avenue for fabricating  other atomically thin 2D materials with desired structures.   

Intercalation Engineering of 2D vdW Magnet Properties

Van der Waals 2D magnetic materials have emerged as a novel platform that offers unique optoelectronic, magnetic, and quantum properties.¹ Such low-dimensional spin systems have vast potential in applications such as spintronics and nanoscale magnetic devices. Therefore, the ability to engineer the structure and defects with respect to magnetic, optical, and electronic properties is critical.

Here, we create a structural phase transformation on an A-type antiferromagnetic 2D magnet CrSBr via electron beam irradiation inside the transmission electron microscope (TEM).² This structural phase transformation reveals vdW gaps when imaged perpendicular to the original layers. Various transition metals are then deposited, followed by imaging, to study intercalation effects through the vdW gap. These experiments are carried out using an evaporator integrated with a TEM through an ultra high vacuum environment so that oxidation of the metals can be avoided. Density functional theory calculations quantify the electronic and magnetic ground state of the engineered 2D magnet. Lastly, in-situ heating during the intercalation process is done to study how heat drives such process. We believe that property tuning via intercalation at desired locations in a 2D magnet will offer another way to write magnetic/electronic textures for applications like quantum simulators.

References

1. Gibertini, M., et al. Nat. Nanotechnol. 14, 408-129 (2019).

2. Klein, J. P., et al. Nat. Commun. 13, 5420 (2022).   

STM investigations of metal-organic frameworks on Fe-intercalated TaS2

Two-dimensional (2D) metal-organic frameworks (MOFs) offer opportunities for physical property adjustments in heterostructures based on various substrates. While 2D MOFs on metallic surfaces have been widely reported, combinations of MOFs and 2D or quasi 2D emergent crystals are rare but full of potential. Using scanning tunneling microscopy (STM) and spectroscopy (STS) at a low temperature, we investigated in-situ grown PTCDA (perylene-3,4,9,10-tetracarboxylic dianhydride)-based monolayer sheets on the surface of Fe-intercalated TaS₂ (Fe_0.24TaS₂). Depending on post-annealing temperatures, PTCDA molecules form a self-assembly monolayer and different types of Fe-PTCDA frameworks with residual Fe atoms on the Fe_0.24TaS₂ surface. Annealed at a high temperature, a stable honeycomb MOF appears, of which the detailed structure was resolved. STS results unveil the shifting of the PTCDA Fermi level in different MOFs. Our work exhibits controllable interfacial electronic structures between PTCDA and Fe_0.24TaS₂, and suggests fabrications of organic heterostructures on quasi 2D emergent crystals.   

Tuning electronic properties of graphene heterostructures via controlled metal intercalation

The synthesis of heterostructures by intercalating layered materials like epitaxial graphene on SiC (Gr/SiC) is a promising strategy for producing high-quality 2D heterostructures. However, identifying the growth kinetics poses a challenge in achieving predictive synthesis. Here, we systematically study the intercalation of metals (Dy, Gd, Pb) in Gr/SiC as a function of temperature, time, and coverage using surface diffraction (SPA-LEED) and scanning tunneling microscopy (STM). Such a comprehensive approach identifies the intercalation locations throughout the process, which is crucial for the final electronic properties.