Ultralow-temperature ultrafast formation of single-crystalline graphene via metal-assisted graphitization of silicon-carbide

Ultralow-temperature ultrafast formation of single-crystalline graphene via metal-assisted graphitization of silicon-carbide | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Ultralow-temperature ultrafast formation of single-crystalline graphene via metal-assisted graphitization of silicon-carbide Hyun Kum, Taehoon Lee, Sungkyu Kim, Hongsik Park, Roy Chung, Se Kim, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5970972/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Non-conventional epitaxial techniques, such as van der Waals epitaxy (vdWE) and remote epitaxy, have attracted substantial attention in the semiconductor research community for their exceptional capability to continuously produce high-quality free-standing films on a single mother wafer without needing surface refurbishment. The successful implementation of these emerging epitaxial techniques crucially hinges on creating a robust uniform two-dimensional (2D) material surface at the wafer-scale and with atomically precise uniformity. The conventional method for fabricating graphene on a silicon carbide (SiC) wafer is through high-temperature graphitization, which produces epitaxial graphene on the surface of the SiC wafer. However, the extremely high temperature needed for silicon sublimation (typically above 1500°C) causes step-bunching of the SiC surface in addition to the growth of uneven graphene at the edges of the step, leading to multilayer graphene stripes and unfavorable surface morphology for epitaxial growth. Here, we fully develop a graphitization technique that allows fast synthesis of single-crystalline graphene at ultra-low temperatures (growth time of less than 1 minute and growth temperature of less than 500°C) at wafer-scale by metal-assisted graphitization (MAG). We found annealing conditions that enable SiC dissociation while avoiding silicide formation, which produces single-crystalline graphene while maintaining atomically smooth surface morphology. The thickness of the graphene layer can be precisely controlled by varying the metal thickness or annealing temperature, allowing the substrate to be utilized for either a remote epitaxial growth substrate or a vdWE growth substrate, depending on the thickness of the graphene. We successfully produce freestanding single-crystalline ultra-wide bandgap (AlN, GaN) films on graphene/SiC via the 2D material-based layer transfer (2DLT) technique. The exfoliated films exhibit high crystallinity and low defect densities. Our results show that low-temperature graphene synthesis via MAG represents a promising route for the commercialization of the 2D-based epitaxy technique, enabling the production of large-scale ultra-wide bandgap free-standing crystalline membranes. Physical sciences/Nanoscience and technology/Graphene/Synthesis of graphene Physical sciences/Materials science/Theory and computation/Electronic structure Physical sciences/Materials science/Nanoscale materials/Synthesis and processing Physical sciences/Materials science/Nanoscale materials/Graphene/Synthesis of graphene Physical sciences/Materials science/Materials for devices/Electronic devices Graphitization Ultra-wide bandgap van der Waals epitaxy Remote Epitaxy 2D-coated substrate Graphene Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Two-dimensional (2D) material-based epitaxy techniques, such as van der Waals epitaxy (vdWE) and remote epitaxy, have recently garnered substantial attention due to their ability to address fundamental challenges inherent in conventional heteroepitaxial techniques caused by lattice and thermal expansion mismatches 1 – 5 . The vdWE and remote epitaxial techniques not only alleviate these issues, but also potentially allow reuse of costly semiconductor substrates for significant cost reduction of electrical devices. These techniques enable precise exfoliation of highly functional single-crystalline membranes and allow heterogeneous integration of dissimilar materials, thereby enabling the integration of distinct electronic and photonic elements onto a single platform 1 . Numerous studies on these epitaxial techniques have been investigated for various semiconductors (Si, Ge, III-V, and III-nitride materials) as well as complex oxides (perovskites, spinels, and garnets) 2 – 4 , 6 – 8 . Successful use of graphitized silicon carbide (SiC) substrates for remote and vdW epitaxy of GaN membranes has also been demonstrated recently 4 , 8 . Single-crystalline SiC is one of the most promising candidates for remote and vdW epitaxial growth due to its ability to form uniform defect-free and atomically flat graphene at wafer-scale through high-temperature graphitization. The graphitization process eliminates the need to transfer graphene onto the host substrate for growth, preventing process-induced damage or residues on the graphene layer that can hinder the growth of high-quality epitaxial films. However, graphitization and subsequent formation of graphene on SiC usually occurs at very high temperatures that exceed 1500°C 9 , 10 . This high-temperature requirement is a technical obstacle as the high temperature not only causes significant step-bunching of the SiC surface but also leads to non-uniform growth of graphene at the edges of the step 9 . Recently, several groups have reported on a new graphitization method of SiC substrates utilizing thin metal films as a catalyst to reduce the temperature of graphitization. This metal-assisted graphitization (MAG) process can be achieved through the deposition of various metals on the SiC substrate, such as nickel (Ni), iron (Fe), ruthenium (Ru), cobalt (Co) and its alloys 11 – 21 . These studies showcase the catalytic effect of metals in breaking the Si-C covalent bond at temperatures below 1100°C. However, a few groups have identified defective interface graphene at the metal-silicide/SiC substrate 15 – 19 , suggesting that liberated carbon atoms diffuse inside the metal-silicide layer in the annealing process, which then precipitates at the silicide/SiC interface to form graphene in the cooling process. Unfortunately, the reaction between the metal and dissociated Si atoms leading to silicide formation appears to be dominant 13 , 16 , which not only deteriorates the uniformity and roughness of graphene 13 , 18 but also causes surface damage to the SiC substrate 15 , 17 . Notably, Lim et al. reported that SiC decomposes at approximately 450°C using Ni as the catalyst, even below the silicide formation temperatures, resulting in the formation of 6 ~ 8 nm multilayer graphene at the metal/SiC interface due to the differences in the solubility of carbon and silicon in Ni 19 . However, the formation mechanism and properties of the interface graphene synthesized by MAG remain unclear due to limited information on surface morphology, domain size, and graphene uniformity as previous studies have predominantly focused on graphene formed on reacted metal surfaces, silicide surfaces, and silicide/SiC interface. Thus, further investigation is needed to bridge this knowledge gap. Here, we demonstrate ultra-fast growth of wafer-scale graphene on a 4-inch SiC substrate below 500°C, with precise control on the number of graphene layers by changing the metal thickness and annealing temperature. The resulting graphene grown by MAG exhibits overall uniform characteristics comparable to graphene synthesized by traditional high-temperature graphitization. Moreover, the III-nitride films grown on MAG-treated SiC via 2D-assisted epitaxy demonstrate quality comparable to those grown on high-temperature graphitized SiC and conventional substrates such as sapphire and SiC via MOCVD. These breakthroughs hold significant promises for advancing semiconductor technologies using SiC as epitaxial substrates. To gain a comprehensive understanding of the MAG mechanism and its applicability, we carried out the catalyst effects of various metal elements such as Ni, Fe, and Ru. Our results offer valuable insights into the general mechanisms of the MAG process, providing a more accessible approach for synthesizing wafer-scale graphene which can be readily used as remote or vdW epitaxial substrates. Results and Discussion Metal-assisted graphitization (MAG) of SiC was first reported by Juang et al. 11 , in which they studied the formation of graphene by depositing a thin Ni layer on SiC and annealing at relatively low temperatures. They found that the metal film acts as a decomposition catalyst above a certain temperature (> 700°C), which is sufficient to fully overcome the activation barrier for silicide formation. The tendency of metal to react with Si reduces the Si-C bonding energy, thereby lowering the temperature required for Si sublimation and subsequent graphitization of the SiC. We carried out a meticulous study by varying the catalyst metal element and annealing temperature as well as performing DFT calculations to fully understand the MAG process on SiC. The process of MAG is illustrated in Fig. 1 a (see Supplementary Fig. 1 for details of the 4-inch wafer MAG process). A thin layer of metal is deposited onto a 4º off-axis 4H-SiC (0001) substrate using sputtering or e-beam evaporation for Ni, Ru, and Fe, followed by annealing in a rapid thermal annealing (RTA) chamber. After cooling, any residual metal film is etched away in its respective etchants. A photograph of the resulting graphitized SiC, prepared with the appropriate metal, is shown in Fig. 1 b. The graphitized sample shows a weak visible-light absorption, as indicated by the ultraviolet-visible (UV-vis) transmittance spectra (see Supplementary Fig. 2). Figure 1 c illustrates a possible mechanism describing how carbon atoms redistribute in the presence of metal catalysts at elevated temperatures, forming a graphene layer at the metal-SiC interface. The MAG process is characterized by three distinct stages. In Stage 1, once sufficient thermal energy is introduced, the metal catalyst facilitates the decomposition of Si–C bonds. In Stage 2, as the annealing temperature increases, further dissociation occurs. The system then stabilizes as newly released C atoms either migrate outward into the metal bulk (i) or remain at the metal/SiC interface to form graphitic carbon or graphene layers (ii or iii). The redistribution of C atoms likely varies significantly depending on the type of metal, possibly due to differences in solubility and chemical affinity. Stable graphene layers are expected to form at the metal/SiC interface as a final product of the interactions (Stage 3), but only when appropriate metals are used. To understand the MAG processes proposed in Fig. 1 c, it is crucial to first examine the microscopic mechanism of metal-assisted graphitization of SiC. Previous studies indicate that the dissociation of SiC is driven by its reaction with metals to form thermodynamically favored metal-silicides 22 – 24 . Ni, Ru, and Fe, which exhibit highly negative enthalpies of silicide formation 25 , enable dissociation at lower temperatures compared to the conventional graphitization process. However, the distinct properties of each metal, such as the solubility of silicon and carbon in the metal and their chemical affinity with the metal, may significantly influence not only the distribution of silicon and carbon within the metal but also the interaction of the metal with SiC or graphene. As a result, significantly different behaviors during Stages 2 and 3 are expected depending on the metal catalyst utilized, which has not been thoroughly investigated. Additionally, most previous MAG studies primarily focused on carbon distribution after the metal had converted into silicide, where precipitated carbon was observed at both the silicide/SiC interface and the silicide surface due to the solubility limitations of carbon during the cooling stage. These approaches, however, overlooked the distribution of C and Si immediately after the Si-C bond dissociation in the initial stage, where the effect of the metal is assumed to be most significant. This lack of knowledge regarding the early stage of carbon redistribution in the MAG process has resulted in outputs unsuitable for device applications due to challenges such as uneven SiC consumption 16 and graphene clustering 13 , both attributed to localized silicide formation 15 , 17 , 18 . To bridge the gap in understanding the microscopic processes at the early stage of the MAG process, as well as to elucidate the metal-specific differences, we employed ab initio molecular-dynamics (AIMD) simulations. We considered three metal elements-Ni, Ru, and Fe-as catalysts and investigated the initial stage of carbon redistribution following Si-C bond dissociation to unveil the dependence of stable graphene formation on metal types (see Supplementary Note 1 and Supplementary Figs. 3 and 4 for detailed simulation methods). Figure 1 d shows the structural evolution of carbon atoms at the metal-SiC interface during AIMD simulations. Interesting observations include: (i) carbon and metal atoms tend to diffuse into each other to a certain extent at the metal-SiC interface, resulting in slight intermixing between the carbon and metal elements (Stage 2); (ii) carbon atoms in the graphene layer maintain their hexagonal arrangement, with the degree of structural order depending on the metal type. The evolution of the number of six-fold rings in the graphene layer and carbon atoms with three-fold coordination indicates that Ni is the most effective in stabilizing the graphene layer (Stages 2 and 3), while significant disordering occurs with the other metals (see Supplementary Video 1–3); and (iii) the van der Waals gap between the metal and graphene layer is relatively well maintained in the Ni-containing model (Stage 3), whereas Fe- or Ru-containing models show a collapse of the gap (Supplementary Fig. 6). These AIMD results highlight the distinctive catalytic role of the Si-dissolved Ni matrix in driving carbon reorganization and graphene-like structure formation. Ni, known for its high silicon solubility 26 and low carbon affinity 22 , facilitates graphene formation at the metal/SiC interface. In contrast, graphene layers become unstable with Ru, (low Si solubility 27 ) and Fe (high carbon affinity 28 ), emphasizing the importance of selecting a catalytic metal with high Si solubility and low carbon chemical affinity. These DFT results align well with our experimental results, demonstrating that only Ni enables the formation of uniform graphene at the metal/SiC interface without any silicide formation. A schematic representation of the metal’s effect on MAG, highlighting the significant differences in outcomes, is illustrated in Fig. 2 a. To experimentally verify the catalyst effect of each metal in reducing the Si-C dissociation barrier energy during the MAG process, we measured X-ray photoelectron spectroscopy (XPS) C 1s and Si 2p spectra after annealing at significantly low temperature (Supplementary Fig. 7). The liberation of C and Si atoms was observed when annealing the samples at 500°C for Ni and Fe, and at 400°C for Ru, showcasing a significant reduction in the energy required for Si-C bond dissociation. The formation of carbon layers on the metal surface, produced either by dissociated carbon diffusing through grain boundaries 21 or due to the low solubility of carbon in the metal matrix 20 , was further confirmed by Raman spectroscopy, supporting the XPS results (Supplementary Fig. 8). After confirming Si-C bond dissociation in all metal/SiC systems, we investigated the temperature-dependent phase transformations in the metals, including the formation of silicide through solid-state reactions. As shown in Fig. 2 b, no silicide peaks were detected for Ni annealed at 500°C, and Fe annealed at 400°C, even though the liberated Si atoms were observed in the XPS Si 2p spectra (Supplementary Fig. 8). However, at higher temperatures, an upshift in binding energy was observed, 0.3 eV for Ni and 0.5 eV for Fe, indicating the formation of Ni 2 Si and Ni 31 Si 12 mixed phases 29 , and FeSi 30 . For Ru, a shift in binding energy (0.3 eV) was already evident at 350°C, indicating the formation of Ru 2 Si 3 31 . These solid-state reactions of metal/SiC were further supported by X-ray diffraction (XRD) measurements (Supplementary Fig. 9). Due to the overlap of C 1s spectra with Ru 3d spectra in XPS and indistinguishable peaks in XRD scans, Raman spectra were further employed for supporting structural change (Supplementary Fig. 10). We further observed distinct effects of each metal on graphene formed at the metal surface. A distinct 2D peak in Raman spectra was only detected on Ni without the formation of a silicide layer (Supplementary Fig. 11). For Fe, graphene formation occurred only after silicide formation, where liberated carbon atoms were either contained within the metal or distributed at the metal/SiC interface. In contrast, Ru did not form a graphene layer but an amorphous carbon layer below 400°C, despite the silicide reaction having already proceeded, potentially indicating a need for higher temperatures to crystallize the carbon layer into graphene 11 , 16 . Since DFT calculations indicated significant differences in the possibility of graphene formation at the interface of each metal/SiC depending on the metal, further investigations of the metal/SiC interface were conducted. The residual metal film was etched away using its respective etchant after annealing at the minimum temperature that confirmed Si-C dissociation. Raman spectra and the optical image (OM) confirmed that a graphene layer was synthesized on SiC only when Ni was used as the catalyst, as in agreement with DFT results (Supplementary Fig. 11). In contrast, no carbon layer was detected for Ru annealed at 350°C while an uneven amorphous carbon layer was observed for Fe annealed at 400°C, suggesting that the carbon at the metal/SiC interface either diffused outward into the reacted region or was absorbed into the metal layer without any reconstructing the amorphous carbon to graphene, as consistent with DFT results. These results unambiguously verify that the selection of the metal catalyst plays a crucial role in the MAG process. Among the metals studied, only the Ni enabled uniform, wafer-scale graphene formation at the metal/SiC interface without conversion of the metal into silicide or carbide phases during the MAG process. After selecting the appropriate metal (Ni) based on the above results, further experimental studies were conducted to validate the carbon distribution model at both low temperatures (stage 1 in Fig. 1 c) and relatively high temperatures (stage 2 in Fig. 1 c). We further lowered the annealing temperature not only to control the thickness of the graphene but also to explore the relationship between annealing temperature and Si-C bond dissociation. As shown in Fig. 2 d and Supplementary Fig. 12, transmission electron microscopy (TEM) analysis confirmed the formation of 4-layer graphene at the Ni/SiC interface after annealing at 500°C for ~ 1 second, while mono-layer graphene formed at 320°C for ~ 1 second. Both results indicate that slightly thicker graphene was formed at the Ni surface than interface of Ni and SiC, the liberated C atoms tend to diffuse slightly more at the grain boundary in Ni rather than at the interface of Ni and SiC. These results are consistent with the UV-vis transmittance spectra obtained at the macroscopic scale after etching with a respective etchant (Supplementary Fig. 13). Furthermore, when the annealing temperature was increased to 550°C, we observed not only inhomogeneous silicide reactions but also clustering of graphene, as confirmed by atomic force microscopy (AFM), scanning electron microscopy (SEM), and Raman spectroscopy (Supplementary Fig. 14). These findings align with observations reported in the literatures 15 , 20 . Furthermore, we reduced the thickness of Ni from 50 nm to 8 nm to investigate the relationship between metal thickness and graphene thickness. As dissociated Si atoms are homogeneously distributed within the Ni lattice 32 , we hypothesized that reducing the metal thickness would decrease the reactivity between SiC and Ni. This reduction is attributed to the lower gradient-driven diffusion of Si atoms at thinner metal layers, resulting from shorter diffusion pathways or a diminished driving force for Si-C dissociations, thereby leading to thinner graphene on the SiC surface. As shown in Fig. 2 d, the thickness of graphene on the SiC surface decreased to approximately 1 monolayer (ML). These findings align with the UV-vis transmittance spectra measured at the macroscopic scale after etching with the respective etchant (see Supplementary Fig. 15). Overall, the results indicate that the MAG process is primarily temperature-driven, in agreement with the MAG carbon distribution model. Moreover, the thickness of the graphene layer on the SiC substrate can be precisely controlled by adjusting the metal layer thickness or the annealing temperature. Next, we characterized the graphene layers formed on SiC by etching away the Ni layer graphitized at 500°C for ~ 1 second. The presence of graphene was confirmed by C 1s X-ray photoelectron spectroscopy (XPS), as shown in Fig. 3 a. In the spectra, the peak labeled SiC corresponds to the Si-C bonds in the SiC substrate. The components S1 and S2 are attributed to sp 2 - and sp 3 -hybridized carbon atoms, respectively. These peaks indicate the presence of a graphene layer (S1) and a carbon buffer layer (CBL) or sp 3 -defective graphene (S2) 8 . Additionally, the XPS Ni 2p spectra and energy-dispersive X-ray spectroscopy (EDS) confirmed the complete removal of Ni, as it was not detected on the surface (Supplementary Fig. 16). The graphene was also characterized using plan-view TEM (See Experimental details). The TEM image along with the selected area electron diffraction (SAED) pattern reveals a well-aligned honeycomb lattice, providing clear evidence of single-crystalline graphene, as shown in Fig. 3 b. To evaluate the characteristics of graphene on a macroscopic scale, Raman spectroscopy was employed, as shown in Fig. 3 c. The Raman spectra confirmed the presence of a graphene layer through the characteristic D, G, and 2D bands. It also enabled the assessment of graphene thickness and quality by analyzing the 2D band full width at half maximum (FWHM). The FWHM of the 2D band serves as a reliable quantitative measure to distinguish the number of layers, ranging from single-layer graphene (typically, 27.5 ± 3.8 cm − 1 ) to four-layer graphene (typically, 63.1 ± 1.6 cm − 1 ) 33 . To evaluate the macroscopic uniformity of graphene thickness, Raman mapping of the 2D band FWHM was performed, as shown in Fig. 3 d. The results demonstrated consistent values across the sample, with a mean FWHM of 60 cm − 1 and a standard error of 0.059, indicating uniform 4 ML graphene across the entire area. These results are consistent with the TEM images and UV-vis spectra, as shown in Fig. 2 d and Supplementary Fig. 13, respectively. Furthermore, the positions of G and 2D bands provide valuable insights into graphene quality, thickness, strain effects and other characteristics 34 . As shown in Fig. 3 e and Fig. 3 f, the average calculated G band position was 1580 cm − 1 with a standard error of 0.52, while the average calculated 2D band position was 2690 cm − 1 with a standard error of 0.18. These results confirm not only the uniform properties but also the consistent graphene thickness across the entire SiC substrate. Furthermore, complete graphene coverage across the 4-inch SiC substrate was confirmed (see Raman spectra and camera image in Supplementary Fig. 17), demonstrating the capability to produce 4-inch graphene directly on a semiconductor substrate using a very low-temperature, ultra-fast process, that relies solely on RTA and sputtering, which are techniques fully compatible with Si CMOS industry standards. The results of MAG suggest that the SiC substrate can be used as an ideal substrate for producing single-crystalline free-standing high-quality III-nitride films via remote epitaxy and vdWE by controlling the thickness of graphene through modifying the annealing temperature. Remote epitaxy is an advanced epitaxial technique where epitaxial growth occurs on a substrate covered with a 2D material 1 . The partial transparency of the 2D layer to Coulombic interactions allows adatoms to electrostatically interact with the underlying substrate 1 – 3 . At the substrate-epitaxial membrane interface, the 2D material and its van der Waals gap effectively eliminate dislocations or cracks in the epitaxial membrane caused by lattice strain relaxation, which has been a persistent challenge in conventional techniques for achieving high-quality epitaxial devices 2 , 35 , 36 . In contrast, vdWE involves epitaxial growth on surfaces without dangling bonds, such as 2D materials, or on 3D materials with passivated dangling bonds. These slippery interfaces enable strain relaxation, allowing the growth of materials with significant lattice mismatches greater than 60% 1 . Furthermore, in both 2D-based epitaxial techniques, epitaxial membranes are bound to the 2D material via weak van der Waals interactions. This characteristic enables the fabrication of free-standing membranes through 2D material-assisted layer transfer (2DLT) 7 . These approaches are particularly advantageous, as they allow the repeated reuse of expensive substrates while producing multiple single-crystalline membranes. A schematic of the detailed 2D-based epitaxy and 2DLT process is shown in Supplementary Fig. 18. III-nitrides, such as GaN and AlN, are promising candidates for high-temperature logic and power devices as well as light emitting diodes due to their exceptional intrinsic material properties 37 , 38 . These include wide direct bandgaps (3.4 eV for GaN and 6.2 eV for AlN) 38 , 39 , high breakdown electric field (4.9 MV cm − 1 for GaN and 15.4 MV cm − 1 for AlN) 47 , and high electron mobility. These materials are particularly well-suited for 2D-based epitaxial growth on graphitized SiC, as their hexagonal lattice arrangement aligns well with SiC, enabling the formation of single-crystalline membranes via remote epitaxy. Furthermore, the graphitization of SiC produces nearly pristine graphene, whose preserved hexagonal lattice structure enhances the growth of c-plane III-nitrides films, offering significant benefits for vdWE 36 . Finally, the direct synthesis of graphene on SiC eliminates the need for wet-transferred graphene synthesized on metal foils via CVD, which often introduces defects such as wrinkles, holes, interfacial contamination, and organic residues. These defects can disrupt the remote interaction between the substrate and the remote epitaxial film, as well as between the graphene and the van der Waals epitaxial film 2 . Although the advantages of the graphitized SiC template for 2D-based epitaxy are clear, high-quality III-N membrane growth on graphene faces challenges due to its low chemical reactivity 40 . The high surface migration rate of group III metals on slippery graphene prevents nuclei from stabilizing at their original positions, mitigating the formation of high-density boundaries and defects. However, this also results in epitaxial failure due to insufficient nucleation sites 36 . High-quality single-crystalline AlN film growth via vdWE on graphitized SiC has been reported, where plasma treatment was applied to the graphitized SiC to enhance nucleation by introducing defects on the graphene surface 36 , 41 . However, the exfoliation of the resulting layers was not demonstrated, as the focus was on synthesizing crack-free AlN layers that leveraged the stress relaxation benefits of graphene. Since the membrane exfoliation yield via 2DLT improves with a uniform graphene layer covering the entire surface on the substrate, untreated graphene is required to successfully produce freestanding membranes. The successful exfoliation of high-quality single-crystalline GaN on non-defect-induced graphitized SiC via 2D-assisted epitaxy was first reported in 2014, achieved not only by engineering the epitaxial growth strategy on a 2D surface but also by utilizing the periodical step edges of SiC 4 . These step edges, with terrace widths typically ranging from 5 to 10 µm and step heights from 10 to 15 nm 4 , 41 , remain after the step bunching induced by high-temperature processes 42 . These periodic step edges generate uniform fluctuations in electric potential, providing energetically favorable nucleation sites for adatoms and enabling the growth of single-crystalline GaN 4 . In this regard, our MAG-graphitized approach is expected to offer significant advantages since the MAG process is conducted at low temperatures, which avoids step bunching and preserves the naturally periodic small terraces of 4° offcut 4H-SiC. The terraces have widths of under 7.2 nm and step heights of under 0.5 nm 43 , as illustrated in Fig. 4 a. As predicted, we were able to successfully grow high-quality AlN films on both 4 ML graphene/SiC via vdWE and 1 ML graphene/SiC via remote epitaxy. For both samples, electron backscatter diffraction (EBSD) maps with SEM images and XRD scans confirmed a (002) wurtzite orientation across a large area, as shown in Fig. 4 b and Fig. 4 c. These results indicate that single-crystalline AlN can be grown via remote epitaxial seeding from 1 ML graphene/SiC and van der Waals epitaxial seeding from 4 ML graphene/SiC (see the illustrated image in Supplementary Fig. 19). However, the FWHM of AlN (002) peaks in XRD scans broadened with increasing graphene thickness on SiC, suggesting lower seeding efficiency compared to remote epitaxy. A prior study reported the successful exfoliation of high-quality AlN on graphitized SiC grown via 2D-based epitaxy, achieving an AlN (002) FWHM of 3600 arcsec at a thickness of 670 nm 40 . In contrast, our study demonstrated significantly improved quality on MAG-treated SiC, with remote epitaxy achieving an AlN (002) FWHM of 673 arcsec at a thickness of 270 nm and vdWE-grown AlN exhibiting an AlN (002) FWHM of 1890 arcsec at a thickness of 285 nm (see cross-sectional SEM images in Supplementary Fig. 20). To compare the MAG sample with the conventional graphitized sample, we conducted high-temperature graphitization process (see experimental details). Our sample exhibited step bunching along with a graphene layer, consistent with findings from previous research (see Raman spectra with AFM image in Supplementary Fig. 21). We performed AlN growth under identical conditions following graphitization at high temperature. As shown in Fig. 4 d, SEM images reveal that AlN adatoms preferentially nucleate at the SiC step edges, whereas a polycrystalline nature of AlN was observed on the SiC terraces. We concluded that the dramatically enhanced crystallinity of AlN can be attributed to the presence of tightly packed and periodically stepped SiC, as shown in the TEM images in Fig. 2 d and Fig. 2 e. These features provide highly energetically favorable nucleation sites, which are effective for both remote epitaxy 4 and vdWE 36 , 44 . After confirming the successful growth of a single-crystalline AlN layer on MAG-treated SiC, we used it as a buffer layer to grow a single-crystalline GaN. In conventional epitaxy, AlN layers (a-axis: 3.112 Å) are commonly employed as intermediate layers to address the lattice mismatch between GaN (a-axis: 3.189 Å) and SiC (a-axis: 3.073 Å), thereby enhancing the quality of GaN. While graphene as a buffer layer effectively relaxes the lattice strain of GaN thin films, the crystallinity of GaN grown on AlN synthesized via 2D-assisted epitaxy still requires thorough investigation. To evaluate this, we grew GaN on each 2D-assisted epitaxial AlN templates and conducted XRD and EBSD measurements to determine the crystallinity on a macroscopic scale. The EBSD maps with SEM images and XRD scans verified the (002) wurtzite orientation over a large area, indicating single crystallinity of the grown GaN film on both AlN templates, as shown in Fig. 4 e and Fig. 4 f. Notably, the FWHM of GaN (002) on XRD scans ranges from 385 arcsec to 496 arcsec, showing no variation based on the choice of AlN template. These results are comparable not only to those of AlN-buffer-assisted GaN films on conventional substrates such as sapphire or SiC via MOCVD, but also to the remote epitaxial growth of GaN on graphitized SiC (see Supplementary Table 1). After confirming the single crystallinity of the GaN layer, we utilized 2DLT to exfoliate both samples. Following the deposition of the adhesion layer (Ti) and stressor layer (Ni) on the surface of GaN samples, mechanical exfoliation was carried out using thermal release tape (TRT) as a handling layer. The strain energy generated by the Ni stressor guided crack propagation precisely along the AlN/graphene interface, facilitated by the weak van der Waals bonds between the GaN/AlN and the graphitized SiC. As a result, freestanding GaN/AlN membranes were successfully obtained using both MAG-treated templates (see Supplementary Fig. 22). All these results highlight the advantages of MAG-treated SiC templates, suggesting their potential as a future method for producing the freestanding ultra-wide bandgap III-nitrides materials. Our approach not only dramatically reduces the barrier of preparing epitaxial graphene coated single-crystalline substrates, but also significantly enhances the crystallinity of the freestanding single-crystalline membranes for heterogeneous integration. Conclusion In conclusion, we have investigated and identified the most optimal condition for ultralow temperature ultrafast graphitization of SiC. Our findings demonstrate that the metal employed significantly influences the presence of graphene layers on SiC, with Ni being the only catalyst capable of synthesizing uniform graphene on SiC. These findings underscore the importance of selecting appropriate metals to facilitate graphene growth while minimizing undesired reactions, thus contributing to the optimization of graphene synthesis processes for various applications. Our study successfully demonstrates not only the growth of a continuous graphene layer on a 4-inch SiC wafer at low temperature with an ultra-fast process but also the reproducible synthesis of the free-standing single-crystalline membrane through 2D-based epitaxy. This significant breakthrough facilitates the 3D heterogeneous integration of dissimilar materials, paving the way for the seamless integration of distinct electronic and photonic elements on a single wafer. Experimental details Computational details The first-principles calculations were performed using the projected augmented wave (PAW) plane-wave basis, implemented in the Vienna ab initio simulation package (VASP) 45 . An energy cutoff of 520 eV was employed and the atomic positions were optimized using the conjugate gradient scheme without any symmetric restrictions, until the maximum force on each of them was less than 0.01 eV/Å 46 . All atoms were relaxed to their equilibrium positions when the change in energy on each atom between successive steps converged to 1×10 − 6 eV/- atom. The heterostructure was modeled with an 8×8×1 grid for k-point sampling. The generalized gradient approximation (GGA) exchange-correlation (XC) DFT functional Perdew-Burke-Ernzerhof (PBE) was employed for geometrical optimization and electronic structure calculations 47 . The slab models had dangling bonds on the vacuum surface terminated by pseudo-hydrogen atoms with appropriate fractional charges to avoid surface states. To determine the vacuum level, dipole corrections are introduced to compensate for the artificial dipole moment at the open ends (20 Å vacuum space along the c-axis) arising from the periodical boundary condition imposed in these calculations 48 . Ab initio molecular dynamics (AIMD) simulations were performed in supercells using DFT calculations with a gamma-centered k-point. The time step was set to 3 fs. Simulations for 3 ps were run with a time step of 3 fs to study the dynamic graphitization process. The temperature of the simulation system was controlled at 2273 K using the Nosé–Hoover thermostat 49 , 50 . Sample preparation The single-crystalline 4° offcut 4H–SiC (0001) substrates were supplied by Cree, Inc.. To remove organic contaminants, substrates were cleaned sequentially for 5 min in acetone, 5 min Iso Propyl Alcohol in an ultrasound bath, and finally dried by nitrogen gun. Metal deposition and graphene formation/transfer method After preparing the SiC substrate, we deposited Ni, Fe, and Ru onto each SiC substrate to investigate the effect of metal catalysts on the interface graphene layer. Ni was deposited on the substrate using a DC magnetron sputtering apparatus with an Ar plasma at room temperature (JURA deposition apparatus made by Vakuum Servis Ltd., Czech Republic). The sputtering was carried out in an Ar atmosphere (pressure 0.5 mTorr) with DC power 100 W. Resulting deposition rate was 2.5 nm/min. Fe and Ru were deposited on the substrate using an e-beam evaporator (Korea Vacuum Tech., Korea Republic). The base pressure was maintained below 5 ×10 − 7 torr, with a deposition rate of 0.6 Å/s. Following metal deposition, the samples underwent rapid thermal annealing (RTA). The base pressure during RTA was maintained at 7×10 − 3 torr, with 1 torr of N 2 added to prevent metal oxidation. The samples were annealed for 3 minutes to determine the temperature required to decompose the Si-C bond and assess whether it induces a phase change of metals into silicide. X-ray photoelectron spectroscopy (XPS, K-alpha, Thermo Scientific Inc.) and X-ray diffraction (XRD, Rigaku, SmartLab) with Cukα1 (wavelength 1.54051 Å) were used to identify crystalline phases and structural transformations in the metal/SiC structure. Raman spectroscopy (Horiba Jobin Yvon, LabRam Aramis) equipped with a 532 nm wavelength laser was utilized to analyze phase transformations of metals and to confirm the presence of graphene. Following the reaction, the interfacial graphene layer was analyzed using high-resolution transmission electron microscopy (HRTEM, JEOL ARM200F). To further investigate the interface characteristics, the samples were immersed in a 40% w/v Ferric Chloride (FeCl 3 ) solution (for Ni, and Fe) for 5 minutes or a 5% w/v Sodium Hypochlorite (NaOCl) solution (for Ru) for 5 minutes. The samples were then characterized using optical microscopy (OM), scanning electron microscopy (SEM, JEOL, JSM-IT-500HR), Ultraviolet-visible-near-infrared spectrophotometer (UV-vis-NIR, JASCO, V-650), atomic force microscopy (AFM, Park Systems, NX-10), TEM, and Raman spectroscopy. After graphene synthesis, polyvinyl alcohol (PVA) was drop-casted onto the surface and baked at 80°C for 5 minutes to form an adhesion layer. A TRT was then attached, facilitating the detachment of graphene through mechanical exfoliation. The exfoliated graphene was transferred onto an SiO 2 substrate. The TRT was removed by baking the sample at 130°C, and the graphene was revealed by dipping the sample into deionized water. SiC high-temperature graphitization The wafer was loaded into an Aixtron VP508 reactor for graphitization. It was first cleaned in a hydrogen environment for 30 minutes at 1520°C, followed by annealing at 1580°C in a 700 Torr argon ambient for 10 minutes. III - N epitaxial Growth and exfoliation The GaN/AlN hetero-structure was epitaxially grown on a graphene/SiC substrate by a metal-organic chemical vapor deposition (MOCVD) equipped with a vertical showerhead-type chamber from Sysnex Co., Ltd. The MOCVD reactor maintained a stable pressure of 30 Torr with hydrogen as a carrier gas throughout the growth process. Initially, the AlN buffer layer was grown at 1,050°C for 40 mins on 1 ML graphene/SiC and 4 ML graphene/SiC. Subsequently, GaN layer was grown by a single-step growth at 1,100°C for 4 mins on each AlN layer. In this growth process, trimethylgallium, trimethylaluminum, and ammonia were used as the sources of gallium, aluminum, and nitrogen, respectively. The resulting GaN/AlN epilayers were exfoliated using a Ti/Ni stressor stack with a handling layer. A 50 nm thick Ti layer was deposited as an adhesion layer for the Ni stressor layer via e-beam evaporation, followed by the deposition of a 3.5 µm thick Ni stressor layer using DC magnetron sputtering under an argon ambient. After the deposition of Ti/Ni layers, thermal release tape (TRT) was attached as a handling layer. Finally, the TRT/Ti/Ni/epi stack was lifted from the edges, enabling precise and controlled exfoliation of the GaN/AlN epilayers. III - N membrane characterizations The plan-view, cross-sectional, and EBSD images of the grown and exfoliated samples were obtained using a field-emission SEM system (SU8220, Hitachi). X-ray diffraction characterization was carried out using an XRD measurement system with Cu K-α radiation (Empyrean, Malvern Panalytical). Declarations Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgments The team at Yonsei University would like to acknowledge support from the National Research Foundation of Korea (NRF) (grant no. RS-2023-00222070 and grant no. RS-2024-00445081) and LX Semicon. Data availability Data will be made available on request. Credit authorship contribution statement Se H. Kim : Conceptualization, Writing – original draft, Visualization, Methodology, Investigation. Hanjoo Lee : Investigation, Writing – original draft. Dong Gwan Kim : DFT simulation, Writing – original draft. Donghan Kim : Investigation, Visualization, Validation. Seugki Kim : Visualization, Validation. Hyunho Yang : Visualization, Validation. Yunsu Jang : Visualization, Validation. Jangho Yoon : Validation. Hyunsoo Kim : Resources. Seoyong Ha : Resources. ByuongTak Lee : Resources. Jung-Hee Lee : Resources. Roy Byung Kyu Chung : Resources. Hongsik Park : Validation, Resources. Sungkyu Kim : Validation, Resources. Taehoon Lee : Validation, Resources. Hyun S. Kum : Conceptualization, Methodology, Investigation, Resources, Writing – review & editing, Project administration, Supervision. References Kum H et al (2019) Epitaxial growth and layer-transfer techniques for heterogeneous integration of materials for electronic and photonic devices. Nature Electronics vol. 2 439–450 Preprint at https://doi.org/10.1038/s41928-019-0314-2 Kim H et al (2022) Remote epitaxy. Nat Reviews Methods Primers 2 Chang CS et al (2023) Remote epitaxial interaction through graphene. Sci Adv 9:42 Kim J et al (2014) Principle of direct van der Waals epitaxy of single-crystalline films on epitaxial graphene. Nat Commun 5 Narayan J (2013) Recent progress in thin film epitaxy across the misfit scale (2011 Acta Gold Medal Paper). Acta Mater 61:2703–2724 Kum HS et al (2020) Heterogeneous integration of single-crystalline complex-oxide membranes. Nature 578:75–81 Kim Y et al (2017) Remote epitaxy through graphene enables two-dimensional material-based layer transfer. Nature 544:340–343 Qiao K et al (2021) Graphene Buffer Layer on SiC as a Release Layer for High-Quality Freestanding Semiconductor Membranes. Nano Lett 21:4013–4020 Emtsev KV et al (2009) Towards wafer-size graphene layers by atmospheric pressure graphitization of silicon carbide. Nat Mater 8:203–207 Berger C et al (2006) Electronic Confinement and Coherence in Patterned Epitaxial Graphene. Sci (1979) 312:1191–1196 Juang ZY et al (2009) Synthesis of graphene on silicon carbide substrates at low temperature. Carbon N Y 47:2026–2031 Yuan W, Li C, Li D, Yang J, Zeng X (2011) Preparation of single- and few-layer graphene sheets using Co deposition on sic substrate. J Nanomater (2011) Iacopi F et al (2015) A catalytic alloy approach for graphene on epitaxial SiC on silicon wafers. J Mater Res 30:609–616 Røst HI et al (2021) Low-Temperature Growth of Graphene on a Semiconductor. J Phys Chem C 125:4243–4252 Escobedo-Cousin E et al (2012) Trans Tech Publications Ltd,. Local solid phase epitaxy of few-layer graphene on silicon carbide. in Materials Science Forum vols 717–720 629–632 MacHáč P, Fidler T, Cichoň S, Mišková L (2012) Synthesis of graphene on SiC substrate via Ni-silicidation reactions. Thin Solid Films 520:5215–5218 Macháč P, Fidler T, Cichoň S, Jurka V (2013) Synthesis of graphene on Co/SiC structure. J Mater Sci: Mater Electron 24:3793–3799 Escobedo-Cousin E et al (2014) Trans Tech Publications Ltd,. Solid phase growth of graphene on silicon carbide by nickel silicidation: Graphene formation mechanisms. in Materials Science Forum vols 778–780 1162–1165 Lim S et al (2016) Interfacial reactions in Ni/6H-SiC at low temperatures. J Nanosci Nanotechnol 16:10853–10857 Kwon Y, An BS, Yang CW (2018) Direct observation of interfacial reaction of Ni/6H-SiC and carbon redistribution by in situ transmission electron microscopy. Mater Charact 140:259–264 Hähnel A, Ischenko V, Woltersdorf J (2008) Oriented growth of silicide and carbon in SiC-based sandwich structures with nickel. Mater Chem Phys 110:303–310 Chou TC, Joshi A, Wadsworth J (1991) Solid state reactions of SiC with Co, Ni, and Pt. J Mater Res 6:796–809 Tang WM, Zheng ZX, Ding HF, Jin ZH (2002) A Study of the Solid State Reaction between Silicon Carbide and Iron. Mater Chem Phys 74 Wu M, Huang H, Wu Y, Wu X (2024) Mechanism of solid-state diffusion reaction in vacuum between metal (Fe, Ni, and Co) and 4H–SiC. Ceram Int 50:17930–17939 Schlesinger ME (1990) Thermodynamics of Solid Transition-Metal Silicides. Chem Rev 90 https://pubs.acs.org/sharingguidelines Nash BP, Nash A The Ni-Si (Nickel-Silicon) System Equilibrium Diagram Perring L, Bussy F, Gachon JC, Feschotte P (1999) The Ruthenium-Silicon System. J Alloys Compd vol. 284 Mattevi C, Kim H, Chhowalla M (2011) A review of chemical vapour deposition of graphene on copper. J Mater Chem 21:3324–3334 Cao Y, Nyborg L, Jelvestam U (2009) XPS calibration study of thin-film nickel silicides. Surf Interface Anal 41:471–483 Ohtsu N, Oku M, Satoh K, Wagatsuma K (2013) Dependence of core-level XPS spectra on iron silicide phase. Appl Surf Sci 264:219–224 van Vliet S, Troglia A, Olsson E, Bliem R (2023) Identifying silicides via plasmon loss satellites in photoemission of the Ru-Si system. Appl Surf Sci 608 Hoshino Y, Matsumoto S, Nakada T, Kido Y (2004) Interfacial reactions between ultra-thin Ni-layer and clean 6H-SiC(0 0 0 1) surface. Surf Sci 556:78–86 Hao Y et al (2010) Probing layer number and stacking order of few-layer graphene by Raman Spectroscopy. Small 6:195–200 Lee DS et al (2008) Raman spectra of epitaxial graphene on SiC and of epitaxial graphene transferred to SiO2. Nano Lett 8:4320–4325 Bae SH et al (2020) Graphene-assisted spontaneous relaxation towards dislocation-free heteroepitaxy. Nat Nanotechnol 15:272–276 Wang Y et al (2020) Flexible graphene-assisted van der Waals epitaxy growth of crack-free AlN epilayer on SiC by lattice engineering. Appl Surf Sci 520 Pradhan DK et al (2024) Materials for high-temperature digital electronics. Nat Rev Mater. 10.1038/s41578-024-00731-9 Zhou C et al (2017) Review—The Current and Emerging Applications of the III-Nitrides. ECS J Solid State Sci Technol 6:Q149–Q156 Gong J et al (2023) Synthesis and Characteristics of Transferrable Single-Crystalline AlN Nanomembranes. Adv Electron Mater 9 Xu Y et al (2017) Growth Model of van der Waals Epitaxy of Films: A Case of AlN Films on Multilayer Graphene/SiC. ACS Appl Mater Interfaces 9:44001–44009 Yu Y et al (2021) Demonstration of epitaxial growth of strain-relaxed GaN films on graphene/SiC substrates for long wavelength light-emitting diodes. Light Sci Appl 10 Avouris P, Dimitrakopoulos C (2012) Graphene: Synthesis and Applications Chen W, Capano MA (2005) Growth and characterization of 4H-SiC epilayers on substrates with different off-cut angles. J Appl Phys 98 Chen Z et al (2019) Improved Epitaxy of AlN Film for Deep-Ultraviolet Light-Emitting Diodes Enabled by Graphene. Adv Mater 31 Kresse G, Furthmü J (1996) Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 54:11169–11186 Blochl PE (1994) Projector augmented-+rave method. Phys Rev B 50:17953–17979 Perdew JP, Burke K, Ernzerhof M (1996) Generalized Gradient Approximation Made Simple. Phys Rev Lett 77:3865–3868 Makov G, Payne MC (1995) Periodic boundary conditions in ab intio calculations. Phys Rev B 51:4014–4022 Nosé S (1984) A unified formulation of the constant temperature molecular dynamics methods. J Chem Phys 81:511–519 Hoover WG (1985) Canonical dynamics: Equilibrium phase-space distributions. Phys Rev (Coll Park) 31 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.pdf Supplementary Information SupplementaryVideo1.mp4 Supplementary Video 1 SupplementaryVideo2.mp4 Supplementary Video 2 SupplementaryVideo3.mp4 Supplementary Video 3 Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5970972","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":417998499,"identity":"7bd1f300-c678-4f36-bacf-a2467b76661a","order_by":0,"name":"Hyun Kum","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAx0lEQVRIiWNgGAWjYBACAx4g8QHMZGwACxClhXEGyVqYeZAFCGox5zljutm2zSZPvr25geFHDYOxeQMBLZa9PWa3c9vSig3OHGxg7DnGYCZzgJDDzvOAtBxO3CCR2MDA28BgI0HIYWAtlkAt8+c/bGD8S5SWs0CHMQK1NNxgbGAG2mJGWMuZY2U3e86lJW44k9hwWOaYhDERWpK33fhRZpM4v/34w4dvamwMZxDSAgaMbBD6AAMDQTtg4A+xCkfBKBgFo2BEAgCC9kDKz6DmIQAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-8009-569X","institution":"Yonsei University","correspondingAuthor":true,"prefix":"","firstName":"Hyun","middleName":"","lastName":"Kum","suffix":""},{"id":417998500,"identity":"88220686-9e14-4327-9760-974ae0d0915d","order_by":1,"name":"Taehoon Lee","email":"","orcid":"https://orcid.org/0000-0001-5553-5123","institution":"Kyungpook national university","correspondingAuthor":false,"prefix":"","firstName":"Taehoon","middleName":"","lastName":"Lee","suffix":""},{"id":417998501,"identity":"e0a914de-c5f5-47d9-99aa-8f96282d0eae","order_by":2,"name":"Sungkyu Kim","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Sungkyu","middleName":"","lastName":"Kim","suffix":""},{"id":417998502,"identity":"68dea77a-930a-4f31-84f1-951b59a5e88c","order_by":3,"name":"Hongsik Park","email":"","orcid":"https://orcid.org/0000-0003-2582-2230","institution":"Kyungpook National University","correspondingAuthor":false,"prefix":"","firstName":"Hongsik","middleName":"","lastName":"Park","suffix":""},{"id":417998503,"identity":"6d9a1cf8-3cd8-419f-8119-04051098afae","order_by":4,"name":"Roy Chung","email":"","orcid":"","institution":"Kyungpook National University","correspondingAuthor":false,"prefix":"","firstName":"Roy","middleName":"","lastName":"Chung","suffix":""},{"id":417998504,"identity":"a97587ce-a884-4d4c-9bc8-c22117520ef2","order_by":5,"name":"Se Kim","email":"","orcid":"","institution":"Yonsei University","correspondingAuthor":false,"prefix":"","firstName":"Se","middleName":"","lastName":"Kim","suffix":""},{"id":417998505,"identity":"bd8c491c-137b-47cc-9bda-e450e2b1dc79","order_by":6,"name":"Han Lee","email":"","orcid":"","institution":"Yonsei University","correspondingAuthor":false,"prefix":"","firstName":"Han","middleName":"","lastName":"Lee","suffix":""},{"id":417998506,"identity":"c72a2f43-6bca-4dd8-bcb1-52d00c77d214","order_by":7,"name":"Dong Kim","email":"","orcid":"","institution":"Kyungpook National University","correspondingAuthor":false,"prefix":"","firstName":"Dong","middleName":"","lastName":"Kim","suffix":""},{"id":417998507,"identity":"6f4ddbe2-af7b-4ec4-a697-c8fc5d3470dc","order_by":8,"name":"Dong Kim","email":"","orcid":"","institution":"Kyungpook National University","correspondingAuthor":false,"prefix":"","firstName":"Dong","middleName":"","lastName":"Kim","suffix":""},{"id":417998508,"identity":"c05438b3-ce92-4958-a2a5-552047a75335","order_by":9,"name":"Seug Kim","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Seug","middleName":"","lastName":"Kim","suffix":""},{"id":417998509,"identity":"a4f801ba-e971-460c-a6e3-932862d28d82","order_by":10,"name":"Hyun Yang","email":"","orcid":"","institution":"Yonsei University","correspondingAuthor":false,"prefix":"","firstName":"Hyun","middleName":"","lastName":"Yang","suffix":""},{"id":417998510,"identity":"a70f6009-f2b2-46ac-b2af-a4f3b1dc5892","order_by":11,"name":"Yun Jang","email":"","orcid":"","institution":"Yonsei University","correspondingAuthor":false,"prefix":"","firstName":"Yun","middleName":"","lastName":"Jang","suffix":""},{"id":417998511,"identity":"0d12ebed-cba4-4c95-a288-5f7dfecfe25c","order_by":12,"name":"Jang Yoon","email":"","orcid":"","institution":"Yonsei University","correspondingAuthor":false,"prefix":"","firstName":"Jang","middleName":"","lastName":"Yoon","suffix":""},{"id":417998512,"identity":"f14b231b-3331-4785-8862-a497c6dec6d1","order_by":13,"name":"Hyun Kim","email":"","orcid":"","institution":"LX Semicon Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Hyun","middleName":"","lastName":"Kim","suffix":""},{"id":417998513,"identity":"d9a1e6c0-6dc9-4dc6-907a-fa8aebcf70e5","order_by":14,"name":"Seo Ha","email":"","orcid":"","institution":"LX Semicon Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Seo","middleName":"","lastName":"Ha","suffix":""},{"id":417998514,"identity":"952dcc29-e638-4e89-9646-38bbb28736fe","order_by":15,"name":"Byoung Lee","email":"","orcid":"","institution":"L\u0026D Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Byoung","middleName":"","lastName":"Lee","suffix":""},{"id":417998515,"identity":"d432a27f-a238-46cf-b2f2-d9a77fc7a71e","order_by":16,"name":"Jung Lee","email":"","orcid":"","institution":"L\u0026D Co., Ltd.","correspondingAuthor":false,"prefix":"","firstName":"Jung","middleName":"","lastName":"Lee","suffix":""}],"badges":[],"createdAt":"2025-02-06 07:25:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5970972/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5970972/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":77387248,"identity":"f6df824c-239a-410a-bc18-4e91d450715e","added_by":"auto","created_at":"2025-02-28 05:39:29","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1342049,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe MAG process and DFT model. a, \u003c/strong\u003eSchematic representation of the MAG process. \u003cstrong\u003eb, \u003c/strong\u003ePhotograph of graphene on a 4-inch wafer SiC (left), with zoomed-in images showing bare SiC (upper right) and 4 ML graphene on MAG-treated SiC (lower right). The darker appearance of the graphene sample indicates reduced visible light transmittance. \u003cstrong\u003ec, \u003c/strong\u003eSchematic illustration of the carbon distribution model during the solid-state reaction in the MAG process. \u003cstrong\u003ed,\u003c/strong\u003e Structural evolution of a carbon layer at the metal-SiC interface during AIMD simulations. The overall 3D view and top view emphasize the differences in the trajectories of carbon atoms at the interface depending on the metal elements. The computed changes in the m-membered rings centered on carbon atoms and coordination numbers clearly reveal that only Ni metal stabilizes a graphene-like structure at the interface, while this stabilization effect is not evident for Fe and Ru metals.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/c3742d982e6d18ce22842809.png"},{"id":77389324,"identity":"9918cca3-0eaf-4e7a-b069-893201686acb","added_by":"auto","created_at":"2025-02-28 06:11:29","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":394981,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMetal effect in the MAG process, and experimental investigations. a, \u003c/strong\u003eSchematic representation of the metal effect in the MAG process, as calculated by DFT calculation and experimental results. The entire interface graphene layer was detected only when using Ni as a catalyst. \u003cstrong\u003eb, \u003c/strong\u003eX-ray spectra\u003cstrong\u003e \u003c/strong\u003eshowing Ni 2p (left), Ru 3d (center), and Fe 2p (right).\u003cstrong\u003e c, \u003c/strong\u003eHR-TEM cross-sectional image ofMAG-treated SiC,annealed at 500°C after depositing a 50 nm Ni layer. \u003cstrong\u003ed, \u003c/strong\u003eHR-TEM cross-sectional image of MAG-treated SiC, annealed at 500°C after depositing an 8 nm Ni layer. The graphene thickness was reduced to 1 layer by lowering the initial metal thickness.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/275da7443b429719ffab1801.png"},{"id":77388443,"identity":"cac6dde6-b9bc-47e9-b92a-377dad1be8c0","added_by":"auto","created_at":"2025-02-28 06:03:29","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":549872,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWafer-scale single-crystalline graphene formed on SiC via the MAG process. a, \u003c/strong\u003eXPS spectra of the C 1s region, showing graphene on the entire SiC. \u003cstrong\u003eb, \u003c/strong\u003ePlan-view TEM image with SAED patterns of graphene transferred onto SiO\u003csub\u003e2\u003c/sub\u003e, verifying the single-crystalline nature of the graphene. \u003cstrong\u003ec, \u003c/strong\u003eRaman spectra showing distinct D, G, and 2D peaks, indicating the presence of graphene on SiC. \u003cstrong\u003ed,\u003c/strong\u003e Raman map of the FWHM of the 2D peak, demonstrating the uniformity of graphene thickness. A total of 256 points were measured within a 15 μm × 15 μm Raman mapping area. \u003cstrong\u003ee,\u003c/strong\u003e Histogram of the G and 2D peak positions, indicating overall uniform graphene quality, and thickness.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/aaa20a804068be6b5bb1779b.png"},{"id":77387245,"identity":"db503021-526a-49cb-984b-0b624c6c73f0","added_by":"auto","created_at":"2025-02-28 05:39:29","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":539091,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFree-standing III-nitride membranes synthesized on MAG-treated graphene/SiC templates. a, \u003c/strong\u003eSchematic illustration of 2D-assisted epitaxy on MAG-treated graphitized SiC and high-temperature graphitized SiC. Unlike high-temperature graphitization, MAG-treated SiC avoids step bunching, offering potential advantages for growing high-quality membranes through 2D-assisted epitaxy. \u003cstrong\u003eb, \u003c/strong\u003eSchematic illustration of the grown film structure (left), SEM image (center), and EBSD map (right), showing the highly single-crystalline nature of the AlN membrane over a large scaleon 1 ML Gr/SiC and 4 ML Gr/SiC templates, respectively. \u003cstrong\u003ec,\u003c/strong\u003e XRD ω-scan of both templates. The FWHM of the AlN (002) peak increases with graphene layerthickness. \u003cstrong\u003ed,\u003c/strong\u003e SEM image of AlN grown on high-temperature graphitized SiC, showing a preference for nucleation on step edges and a polycrystalline nature on terraces. \u003cstrong\u003ee, \u003c/strong\u003eSchematic illustration of the grown film structure (left), SEM image (center), and EBSD map (right), showing the highly single-crystalline nature of the GaN membrane over a large scale on the AlN/1 ML Gr/SiC template.\u003cstrong\u003ef, \u003c/strong\u003eSchematic illustration of the grown film structure (left), SEM image (center) and EBSD map (right), showing the highly single-crystalline nature of the GaN membrane over a large scale on the AlN/4 ML Gr/SiC template. \u003cstrong\u003eg-h,\u003c/strong\u003e XRD ω-scans of GaN membranes grown on AlN/1 ML Gr/SiC and AlN/4 ML Gr/SiC templates, respectively, showing no significant differences. The scale bars in all SEM images and EBSD map are 300 nm.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/8b0efe59a54082e5f6c77482.png"},{"id":77389345,"identity":"68159a7f-95e9-4c62-af9c-31d672e32457","added_by":"auto","created_at":"2025-02-28 06:11:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3647448,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/46173060-4184-49b7-b83d-7ee532398106.pdf"},{"id":77387251,"identity":"23e7f397-61f6-42e7-8f37-8c4c08b58e7c","added_by":"auto","created_at":"2025-02-28 05:39:29","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2928393,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/9e12be7a080c10b41c8c2939.pdf"},{"id":77387254,"identity":"f5784bda-07cb-4850-a27d-2966d28fbee3","added_by":"auto","created_at":"2025-02-28 05:39:30","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":31210883,"visible":true,"origin":"","legend":"Supplementary Video 1","description":"","filename":"SupplementaryVideo1.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/dd0060602c09c97b4a2cd2fc.mp4"},{"id":77387256,"identity":"6fcc6e7e-79fc-41c5-8ab7-d0ccdd8d7bb7","added_by":"auto","created_at":"2025-02-28 05:39:30","extension":"mp4","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":31186900,"visible":true,"origin":"","legend":"Supplementary Video 2","description":"","filename":"SupplementaryVideo2.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/0e3a07352814984f556513d1.mp4"},{"id":77387255,"identity":"965af7b5-8c90-4989-834e-64b51f7bfb35","added_by":"auto","created_at":"2025-02-28 05:39:30","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":31534662,"visible":true,"origin":"","legend":"Supplementary Video 3","description":"","filename":"SupplementaryVideo3.mp4","url":"https://assets-eu.researchsquare.com/files/rs-5970972/v1/cdc39584375e099873cb9d36.mp4"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Ultralow-temperature ultrafast formation of single-crystalline graphene via metal-assisted graphitization of silicon-carbide","fulltext":[{"header":"Introduction","content":"\u003cp\u003eTwo-dimensional (2D) material-based epitaxy techniques, such as van der Waals epitaxy (vdWE) and remote epitaxy, have recently garnered substantial attention due to their ability to address fundamental challenges inherent in conventional heteroepitaxial techniques caused by lattice and thermal expansion mismatches\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The vdWE and remote epitaxial techniques not only alleviate these issues, but also potentially allow reuse of costly semiconductor substrates for significant cost reduction of electrical devices. These techniques enable precise exfoliation of highly functional single-crystalline membranes and allow heterogeneous integration of dissimilar materials, thereby enabling the integration of distinct electronic and photonic elements onto a single platform\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Numerous studies on these epitaxial techniques have been investigated for various semiconductors (Si, Ge, III-V, and III-nitride materials) as well as complex oxides (perovskites, spinels, and garnets)\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Successful use of graphitized silicon carbide (SiC) substrates for remote and vdW epitaxy of GaN membranes has also been demonstrated recently\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Single-crystalline SiC is one of the most promising candidates for remote and vdW epitaxial growth due to its ability to form uniform defect-free and atomically flat graphene at wafer-scale through high-temperature graphitization. The graphitization process eliminates the need to transfer graphene onto the host substrate for growth, preventing process-induced damage or residues on the graphene layer that can hinder the growth of high-quality epitaxial films. However, graphitization and subsequent formation of graphene on SiC usually occurs at very high temperatures that exceed 1500\u0026deg;C\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. This high-temperature requirement is a technical obstacle as the high temperature not only causes significant step-bunching of the SiC surface but also leads to non-uniform growth of graphene at the edges of the step\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRecently, several groups have reported on a new graphitization method of SiC substrates utilizing thin metal films as a catalyst to reduce the temperature of graphitization. This metal-assisted graphitization (MAG) process can be achieved through the deposition of various metals on the SiC substrate, such as nickel (Ni), iron (Fe), ruthenium (Ru), cobalt (Co) and its alloys\u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13 CR14 CR15 CR16 CR17 CR18 CR19 CR20\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. These studies showcase the catalytic effect of metals in breaking the Si-C covalent bond at temperatures below 1100\u0026deg;C. However, a few groups have identified defective interface graphene at the metal-silicide/SiC substrate\u003csup\u003e\u003cspan additionalcitationids=\"CR16 CR17 CR18\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, suggesting that liberated carbon atoms diffuse inside the metal-silicide layer in the annealing process, which then precipitates at the silicide/SiC interface to form graphene in the cooling process. Unfortunately, the reaction between the metal and dissociated Si atoms leading to silicide formation appears to be dominant\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, which not only deteriorates the uniformity and roughness of graphene\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e but also causes surface damage to the SiC substrate\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Notably, Lim et al. reported that SiC decomposes at approximately 450\u0026deg;C using Ni as the catalyst, even below the silicide formation temperatures, resulting in the formation of 6\u0026thinsp;~\u0026thinsp;8 nm multilayer graphene at the metal/SiC interface due to the differences in the solubility of carbon and silicon in Ni\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. However, the formation mechanism and properties of the interface graphene synthesized by MAG remain unclear due to limited information on surface morphology, domain size, and graphene uniformity as previous studies have predominantly focused on graphene formed on reacted metal surfaces, silicide surfaces, and silicide/SiC interface. Thus, further investigation is needed to bridge this knowledge gap.\u003c/p\u003e \u003cp\u003eHere, we demonstrate ultra-fast growth of wafer-scale graphene on a 4-inch SiC substrate below 500\u0026deg;C, with precise control on the number of graphene layers by changing the metal thickness and annealing temperature. The resulting graphene grown by MAG exhibits overall uniform characteristics comparable to graphene synthesized by traditional high-temperature graphitization. Moreover, the III-nitride films grown on MAG-treated SiC via 2D-assisted epitaxy demonstrate quality comparable to those grown on high-temperature graphitized SiC and conventional substrates such as sapphire and SiC via MOCVD. These breakthroughs hold significant promises for advancing semiconductor technologies using SiC as epitaxial substrates. To gain a comprehensive understanding of the MAG mechanism and its applicability, we carried out the catalyst effects of various metal elements such as Ni, Fe, and Ru. Our results offer valuable insights into the general mechanisms of the MAG process, providing a more accessible approach for synthesizing wafer-scale graphene which can be readily used as remote or vdW epitaxial substrates.\u003c/p\u003e"},{"header":"Results and Discussion","content":"\u003cp\u003eMetal-assisted graphitization (MAG) of SiC was first reported by Juang et al.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, in which they studied the formation of graphene by depositing a thin Ni layer on SiC and annealing at relatively low temperatures. They found that the metal film acts as a decomposition catalyst above a certain temperature (\u0026gt;\u0026thinsp;700\u0026deg;C), which is sufficient to fully overcome the activation barrier for silicide formation. The tendency of metal to react with Si reduces the Si-C bonding energy, thereby lowering the temperature required for Si sublimation and subsequent graphitization of the SiC. We carried out a meticulous study by varying the catalyst metal element and annealing temperature as well as performing DFT calculations to fully understand the MAG process on SiC. The process of MAG is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea (see Supplementary Fig.\u0026nbsp;1 for details of the 4-inch wafer MAG process). A thin layer of metal is deposited onto a 4\u0026ordm; off-axis 4H-SiC (0001) substrate using sputtering or e-beam evaporation for Ni, Ru, and Fe, followed by annealing in a rapid thermal annealing (RTA) chamber. After cooling, any residual metal film is etched away in its respective etchants. A photograph of the resulting graphitized SiC, prepared with the appropriate metal, is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb. The graphitized sample shows a weak visible-light absorption, as indicated by the ultraviolet-visible (UV-vis) transmittance spectra (see Supplementary Fig.\u0026nbsp;2). Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec illustrates a possible mechanism describing how carbon atoms redistribute in the presence of metal catalysts at elevated temperatures, forming a graphene layer at the metal-SiC interface. The MAG process is characterized by three distinct stages. In Stage 1, once sufficient thermal energy is introduced, the metal catalyst facilitates the decomposition of Si\u0026ndash;C bonds. In Stage 2, as the annealing temperature increases, further dissociation occurs. The system then stabilizes as newly released C atoms either migrate outward into the metal bulk (i) or remain at the metal/SiC interface to form graphitic carbon or graphene layers (ii or iii). The redistribution of C atoms likely varies significantly depending on the type of metal, possibly due to differences in solubility and chemical affinity. Stable graphene layers are expected to form at the metal/SiC interface as a final product of the interactions (Stage 3), but only when appropriate metals are used.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo understand the MAG processes proposed in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec, it is crucial to first examine the microscopic mechanism of metal-assisted graphitization of SiC. Previous studies indicate that the dissociation of SiC is driven by its reaction with metals to form thermodynamically favored metal-silicides\u003csup\u003e\u003cspan additionalcitationids=\"CR23\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Ni, Ru, and Fe, which exhibit highly negative enthalpies of silicide formation\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, enable dissociation at lower temperatures compared to the conventional graphitization process. However, the distinct properties of each metal, such as the solubility of silicon and carbon in the metal and their chemical affinity with the metal, may significantly influence not only the distribution of silicon and carbon within the metal but also the interaction of the metal with SiC or graphene. As a result, significantly different behaviors during Stages 2 and 3 are expected depending on the metal catalyst utilized, which has not been thoroughly investigated. Additionally, most previous MAG studies primarily focused on carbon distribution after the metal had converted into silicide, where precipitated carbon was observed at both the silicide/SiC interface and the silicide surface due to the solubility limitations of carbon during the cooling stage. These approaches, however, overlooked the distribution of C and Si immediately after the Si-C bond dissociation in the initial stage, where the effect of the metal is assumed to be most significant. This lack of knowledge regarding the early stage of carbon redistribution in the MAG process has resulted in outputs unsuitable for device applications due to challenges such as uneven SiC consumption\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e and graphene clustering\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, both attributed to localized silicide formation\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo bridge the gap in understanding the microscopic processes at the early stage of the MAG process, as well as to elucidate the metal-specific differences, we employed \u003cem\u003eab initio\u003c/em\u003e molecular-dynamics (AIMD) simulations. We considered three metal elements-Ni, Ru, and Fe-as catalysts and investigated the initial stage of carbon redistribution following Si-C bond dissociation to unveil the dependence of stable graphene formation on metal types (see Supplementary Note 1 and Supplementary Figs.\u0026nbsp;3 and 4 for detailed simulation methods). Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed shows the structural evolution of carbon atoms at the metal-SiC interface during AIMD simulations. Interesting observations include: (i) carbon and metal atoms tend to diffuse into each other to a certain extent at the metal-SiC interface, resulting in slight intermixing between the carbon and metal elements (Stage 2); (ii) carbon atoms in the graphene layer maintain their hexagonal arrangement, with the degree of structural order depending on the metal type. The evolution of the number of six-fold rings in the graphene layer and carbon atoms with three-fold coordination indicates that Ni is the most effective in stabilizing the graphene layer (Stages 2 and 3), while significant disordering occurs with the other metals (see Supplementary Video 1\u0026ndash;3); and (iii) the van der Waals gap between the metal and graphene layer is relatively well maintained in the Ni-containing model (Stage 3), whereas Fe- or Ru-containing models show a collapse of the gap (Supplementary Fig.\u0026nbsp;6). These AIMD results highlight the distinctive catalytic role of the Si-dissolved Ni matrix in driving carbon reorganization and graphene-like structure formation. Ni, known for its high silicon solubility\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e and low carbon affinity\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, facilitates graphene formation at the metal/SiC interface. In contrast, graphene layers become unstable with Ru, (low Si solubility\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e) and Fe (high carbon affinity\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e), emphasizing the importance of selecting a catalytic metal with high Si solubility and low carbon chemical affinity.\u003c/p\u003e \u003cp\u003eThese DFT results align well with our experimental results, demonstrating that only Ni enables the formation of uniform graphene at the metal/SiC interface without any silicide formation. A schematic representation of the metal\u0026rsquo;s effect on MAG, highlighting the significant differences in outcomes, is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea. To experimentally verify the catalyst effect of each metal in reducing the Si-C dissociation barrier energy during the MAG process, we measured X-ray photoelectron spectroscopy (XPS) C 1s and Si 2p spectra after annealing at significantly low temperature (Supplementary Fig.\u0026nbsp;7). The liberation of C and Si atoms was observed when annealing the samples at 500\u0026deg;C for Ni and Fe, and at 400\u0026deg;C for Ru, showcasing a significant reduction in the energy required for Si-C bond dissociation. The formation of carbon layers on the metal surface, produced either by dissociated carbon diffusing through grain boundaries\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e or due to the low solubility of carbon in the metal matrix\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, was further confirmed by Raman spectroscopy, supporting the XPS results (Supplementary Fig.\u0026nbsp;8). After confirming Si-C bond dissociation in all metal/SiC systems, we investigated the temperature-dependent phase transformations in the metals, including the formation of silicide through solid-state reactions. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, no silicide peaks were detected for Ni annealed at 500\u0026deg;C, and Fe annealed at 400\u0026deg;C, even though the liberated Si atoms were observed in the XPS Si 2p spectra (Supplementary Fig.\u0026nbsp;8). However, at higher temperatures, an upshift in binding energy was observed, 0.3 eV for Ni and 0.5 eV for Fe, indicating the formation of Ni\u003csub\u003e2\u003c/sub\u003eSi and Ni\u003csub\u003e31\u003c/sub\u003eSi\u003csub\u003e12\u003c/sub\u003e mixed phases\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and FeSi\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. For Ru, a shift in binding energy (0.3 eV) was already evident at 350\u0026deg;C, indicating the formation of Ru\u003csub\u003e2\u003c/sub\u003eSi\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e31\u003c/sup\u003e. These solid-state reactions of metal/SiC were further supported by X-ray diffraction (XRD) measurements (Supplementary Fig.\u0026nbsp;9). Due to the overlap of C 1s spectra with Ru 3d spectra in XPS and indistinguishable peaks in XRD scans, Raman spectra were further employed for supporting structural change (Supplementary Fig.\u0026nbsp;10). We further observed distinct effects of each metal on graphene formed at the metal surface. A distinct 2D peak in Raman spectra was only detected on Ni without the formation of a silicide layer (Supplementary Fig.\u0026nbsp;11). For Fe, graphene formation occurred only after silicide formation, where liberated carbon atoms were either contained within the metal or distributed at the metal/SiC interface. In contrast, Ru did not form a graphene layer but an amorphous carbon layer below 400\u0026deg;C, despite the silicide reaction having already proceeded, potentially indicating a need for higher temperatures to crystallize the carbon layer into graphene\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSince DFT calculations indicated significant differences in the possibility of graphene formation at the interface of each metal/SiC depending on the metal, further investigations of the metal/SiC interface were conducted. The residual metal film was etched away using its respective etchant after annealing at the minimum temperature that confirmed Si-C dissociation. Raman spectra and the optical image (OM) confirmed that a graphene layer was synthesized on SiC only when Ni was used as the catalyst, as in agreement with DFT results (Supplementary Fig.\u0026nbsp;11). In contrast, no carbon layer was detected for Ru annealed at 350\u0026deg;C while an uneven amorphous carbon layer was observed for Fe annealed at 400\u0026deg;C, suggesting that the carbon at the metal/SiC interface either diffused outward into the reacted region or was absorbed into the metal layer without any reconstructing the amorphous carbon to graphene, as consistent with DFT results. These results unambiguously verify that the selection of the metal catalyst plays a crucial role in the MAG process. Among the metals studied, only the Ni enabled uniform, wafer-scale graphene formation at the metal/SiC interface without conversion of the metal into silicide or carbide phases during the MAG process.\u003c/p\u003e \u003cp\u003eAfter selecting the appropriate metal (Ni) based on the above results, further experimental studies were conducted to validate the carbon distribution model at both low temperatures (stage 1 in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec) and relatively high temperatures (stage 2 in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). We further lowered the annealing temperature not only to control the thickness of the graphene but also to explore the relationship between annealing temperature and Si-C bond dissociation. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed and Supplementary Fig.\u0026nbsp;12, transmission electron microscopy (TEM) analysis confirmed the formation of 4-layer graphene at the Ni/SiC interface after annealing at 500\u0026deg;C for ~\u0026thinsp;1 second, while mono-layer graphene formed at 320\u0026deg;C for ~\u0026thinsp;1 second. Both results indicate that slightly thicker graphene was formed at the Ni surface than interface of Ni and SiC, the liberated C atoms tend to diffuse slightly more at the grain boundary in Ni rather than at the interface of Ni and SiC. These results are consistent with the UV-vis transmittance spectra obtained at the macroscopic scale after etching with a respective etchant (Supplementary Fig.\u0026nbsp;13). Furthermore, when the annealing temperature was increased to 550\u0026deg;C, we observed not only inhomogeneous silicide reactions but also clustering of graphene, as confirmed by atomic force microscopy (AFM), scanning electron microscopy (SEM), and Raman spectroscopy (Supplementary Fig.\u0026nbsp;14). These findings align with observations reported in the literatures\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Furthermore, we reduced the thickness of Ni from 50 nm to 8 nm to investigate the relationship between metal thickness and graphene thickness. As dissociated Si atoms are homogeneously distributed within the Ni lattice\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, we hypothesized that reducing the metal thickness would decrease the reactivity between SiC and Ni. This reduction is attributed to the lower gradient-driven diffusion of Si atoms at thinner metal layers, resulting from shorter diffusion pathways or a diminished driving force for Si-C dissociations, thereby leading to thinner graphene on the SiC surface. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, the thickness of graphene on the SiC surface decreased to approximately 1 monolayer (ML). These findings align with the UV-vis transmittance spectra measured at the macroscopic scale after etching with the respective etchant (see Supplementary Fig.\u0026nbsp;15). Overall, the results indicate that the MAG process is primarily temperature-driven, in agreement with the MAG carbon distribution model. Moreover, the thickness of the graphene layer on the SiC substrate can be precisely controlled by adjusting the metal layer thickness or the annealing temperature.\u003c/p\u003e \u003cp\u003eNext, we characterized the graphene layers formed on SiC by etching away the Ni layer graphitized at 500\u0026deg;C for ~\u0026thinsp;1 second. The presence of graphene was confirmed by C 1s X-ray photoelectron spectroscopy (XPS), as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea. In the spectra, the peak labeled SiC corresponds to the Si-C bonds in the SiC substrate. The components S1 and S2 are attributed to sp\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e- and sp\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e-hybridized carbon atoms, respectively. These peaks indicate the presence of a graphene layer (S1) and a carbon buffer layer (CBL) or sp\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e-defective graphene (S2)\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Additionally, the XPS Ni 2p spectra and energy-dispersive X-ray spectroscopy (EDS) confirmed the complete removal of Ni, as it was not detected on the surface (Supplementary Fig.\u0026nbsp;16). The graphene was also characterized using plan-view TEM (See Experimental details). The TEM image along with the selected area electron diffraction (SAED) pattern reveals a well-aligned honeycomb lattice, providing clear evidence of single-crystalline graphene, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb. To evaluate the characteristics of graphene on a macroscopic scale, Raman spectroscopy was employed, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec. The Raman spectra confirmed the presence of a graphene layer through the characteristic D, G, and 2D bands. It also enabled the assessment of graphene thickness and quality by analyzing the 2D band full width at half maximum (FWHM). The FWHM of the 2D band serves as a reliable quantitative measure to distinguish the number of layers, ranging from single-layer graphene (typically, 27.5\u0026thinsp;\u0026plusmn;\u0026thinsp;3.8 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) to four-layer graphene (typically, 63.1\u0026thinsp;\u0026plusmn;\u0026thinsp;1.6 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e)\u003csup\u003e33\u003c/sup\u003e. To evaluate the macroscopic uniformity of graphene thickness, Raman mapping of the 2D band FWHM was performed, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ed. The results demonstrated consistent values across the sample, with a mean FWHM of 60 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and a standard error of 0.059, indicating uniform 4 ML graphene across the entire area. These results are consistent with the TEM images and UV-vis spectra, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed and Supplementary Fig.\u0026nbsp;13, respectively. Furthermore, the positions of G and 2D bands provide valuable insights into graphene quality, thickness, strain effects and other characteristics\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee and Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ef, the average calculated G band position was 1580 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e with a standard error of 0.52, while the average calculated 2D band position was 2690 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e with a standard error of 0.18. These results confirm not only the uniform properties but also the consistent graphene thickness across the entire SiC substrate. Furthermore, complete graphene coverage across the 4-inch SiC substrate was confirmed (see Raman spectra and camera image in Supplementary Fig.\u0026nbsp;17), demonstrating the capability to produce 4-inch graphene directly on a semiconductor substrate using a very low-temperature, ultra-fast process, that relies solely on RTA and sputtering, which are techniques fully compatible with Si CMOS industry standards.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe results of MAG suggest that the SiC substrate can be used as an ideal substrate for producing single-crystalline free-standing high-quality III-nitride films via remote epitaxy and vdWE by controlling the thickness of graphene through modifying the annealing temperature. Remote epitaxy is an advanced epitaxial technique where epitaxial growth occurs on a substrate covered with a 2D material\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. The partial transparency of the 2D layer to Coulombic interactions allows adatoms to electrostatically interact with the underlying substrate\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. At the substrate-epitaxial membrane interface, the 2D material and its van der Waals gap effectively eliminate dislocations or cracks in the epitaxial membrane caused by lattice strain relaxation, which has been a persistent challenge in conventional techniques for achieving high-quality epitaxial devices\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. In contrast, vdWE involves epitaxial growth on surfaces without dangling bonds, such as 2D materials, or on 3D materials with passivated dangling bonds. These slippery interfaces enable strain relaxation, allowing the growth of materials with significant lattice mismatches greater than 60%\u003csup\u003e1\u003c/sup\u003e. Furthermore, in both 2D-based epitaxial techniques, epitaxial membranes are bound to the 2D material via weak van der Waals interactions. This characteristic enables the fabrication of free-standing membranes through 2D material-assisted layer transfer (2DLT)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. These approaches are particularly advantageous, as they allow the repeated reuse of expensive substrates while producing multiple single-crystalline membranes. A schematic of the detailed 2D-based epitaxy and 2DLT process is shown in Supplementary Fig.\u0026nbsp;18.\u003c/p\u003e \u003cp\u003eIII-nitrides, such as GaN and AlN, are promising candidates for high-temperature logic and power devices as well as light emitting diodes due to their exceptional intrinsic material properties\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. These include wide direct bandgaps (3.4 eV for GaN and 6.2 eV for AlN)\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, high breakdown electric field (4.9 MV cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for GaN and 15.4 MV cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e for AlN)\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, and high electron mobility. These materials are particularly well-suited for 2D-based epitaxial growth on graphitized SiC, as their hexagonal lattice arrangement aligns well with SiC, enabling the formation of single-crystalline membranes via remote epitaxy. Furthermore, the graphitization of SiC produces nearly pristine graphene, whose preserved hexagonal lattice structure enhances the growth of c-plane III-nitrides films, offering significant benefits for vdWE\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. Finally, the direct synthesis of graphene on SiC eliminates the need for wet-transferred graphene synthesized on metal foils via CVD, which often introduces defects such as wrinkles, holes, interfacial contamination, and organic residues. These defects can disrupt the remote interaction between the substrate and the remote epitaxial film, as well as between the graphene and the van der Waals epitaxial film\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAlthough the advantages of the graphitized SiC template for 2D-based epitaxy are clear, high-quality III-N membrane growth on graphene faces challenges due to its low chemical reactivity\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. The high surface migration rate of group III metals on slippery graphene prevents nuclei from stabilizing at their original positions, mitigating the formation of high-density boundaries and defects. However, this also results in epitaxial failure due to insufficient nucleation sites\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. High-quality single-crystalline AlN film growth via vdWE on graphitized SiC has been reported, where plasma treatment was applied to the graphitized SiC to enhance nucleation by introducing defects on the graphene surface\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. However, the exfoliation of the resulting layers was not demonstrated, as the focus was on synthesizing crack-free AlN layers that leveraged the stress relaxation benefits of graphene. Since the membrane exfoliation yield via 2DLT improves with a uniform graphene layer covering the entire surface on the substrate, untreated graphene is required to successfully produce freestanding membranes. The successful exfoliation of high-quality single-crystalline GaN on non-defect-induced graphitized SiC via 2D-assisted epitaxy was first reported in 2014, achieved not only by engineering the epitaxial growth strategy on a 2D surface but also by utilizing the periodical step edges of SiC\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. These step edges, with terrace widths typically ranging from 5 to 10 \u0026micro;m and step heights from 10 to 15 nm\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, remain after the step bunching induced by high-temperature processes\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. These periodic step edges generate uniform fluctuations in electric potential, providing energetically favorable nucleation sites for adatoms and enabling the growth of single-crystalline GaN\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. In this regard, our MAG-graphitized approach is expected to offer significant advantages since the MAG process is conducted at low temperatures, which avoids step bunching and preserves the naturally periodic small terraces of 4\u0026deg; offcut 4H-SiC. The terraces have widths of under 7.2 nm and step heights of under 0.5 nm\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs predicted, we were able to successfully grow high-quality AlN films on both 4 ML graphene/SiC via vdWE and 1 ML graphene/SiC via remote epitaxy. For both samples, electron backscatter diffraction (EBSD) maps with SEM images and XRD scans confirmed a (002) wurtzite orientation across a large area, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec. These results indicate that single-crystalline AlN can be grown via remote epitaxial seeding from 1 ML graphene/SiC and van der Waals epitaxial seeding from 4 ML graphene/SiC (see the illustrated image in Supplementary Fig.\u0026nbsp;19). However, the FWHM of AlN (002) peaks in XRD scans broadened with increasing graphene thickness on SiC, suggesting lower seeding efficiency compared to remote epitaxy. A prior study reported the successful exfoliation of high-quality AlN on graphitized SiC grown via 2D-based epitaxy, achieving an AlN (002) FWHM of 3600 arcsec at a thickness of 670 nm\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. In contrast, our study demonstrated significantly improved quality on MAG-treated SiC, with remote epitaxy achieving an AlN (002) FWHM of 673 arcsec at a thickness of 270 nm and vdWE-grown AlN exhibiting an AlN (002) FWHM of 1890 arcsec at a thickness of 285 nm (see cross-sectional SEM images in Supplementary Fig.\u0026nbsp;20). To compare the MAG sample with the conventional graphitized sample, we conducted high-temperature graphitization process (see experimental details). Our sample exhibited step bunching along with a graphene layer, consistent with findings from previous research (see Raman spectra with AFM image in Supplementary Fig.\u0026nbsp;21). We performed AlN growth under identical conditions following graphitization at high temperature. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed, SEM images reveal that AlN adatoms preferentially nucleate at the SiC step edges, whereas a polycrystalline nature of AlN was observed on the SiC terraces. We concluded that the dramatically enhanced crystallinity of AlN can be attributed to the presence of tightly packed and periodically stepped SiC, as shown in the TEM images in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed and Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee. These features provide highly energetically favorable nucleation sites, which are effective for both remote epitaxy\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e and vdWE\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAfter confirming the successful growth of a single-crystalline AlN layer on MAG-treated SiC, we used it as a buffer layer to grow a single-crystalline GaN. In conventional epitaxy, AlN layers (a-axis: 3.112 \u0026Aring;) are commonly employed as intermediate layers to address the lattice mismatch between GaN (a-axis: 3.189 \u0026Aring;) and SiC (a-axis: 3.073 \u0026Aring;), thereby enhancing the quality of GaN. While graphene as a buffer layer effectively relaxes the lattice strain of GaN thin films, the crystallinity of GaN grown on AlN synthesized via 2D-assisted epitaxy still requires thorough investigation. To evaluate this, we grew GaN on each 2D-assisted epitaxial AlN templates and conducted XRD and EBSD measurements to determine the crystallinity on a macroscopic scale. The EBSD maps with SEM images and XRD scans verified the (002) wurtzite orientation over a large area, indicating single crystallinity of the grown GaN film on both AlN templates, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef. Notably, the FWHM of GaN (002) on XRD scans ranges from 385 arcsec to 496 arcsec, showing no variation based on the choice of AlN template. These results are comparable not only to those of AlN-buffer-assisted GaN films on conventional substrates such as sapphire or SiC via MOCVD, but also to the remote epitaxial growth of GaN on graphitized SiC (see Supplementary Table\u0026nbsp;1). After confirming the single crystallinity of the GaN layer, we utilized 2DLT to exfoliate both samples. Following the deposition of the adhesion layer (Ti) and stressor layer (Ni) on the surface of GaN samples, mechanical exfoliation was carried out using thermal release tape (TRT) as a handling layer. The strain energy generated by the Ni stressor guided crack propagation precisely along the AlN/graphene interface, facilitated by the weak van der Waals bonds between the GaN/AlN and the graphitized SiC. As a result, freestanding GaN/AlN membranes were successfully obtained using both MAG-treated templates (see Supplementary Fig.\u0026nbsp;22). All these results highlight the advantages of MAG-treated SiC templates, suggesting their potential as a future method for producing the freestanding ultra-wide bandgap III-nitrides materials. Our approach not only dramatically reduces the barrier of preparing epitaxial graphene coated single-crystalline substrates, but also significantly enhances the crystallinity of the freestanding single-crystalline membranes for heterogeneous integration.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn conclusion, we have investigated and identified the most optimal condition for ultralow temperature ultrafast graphitization of SiC. Our findings demonstrate that the metal employed significantly influences the presence of graphene layers on SiC, with Ni being the only catalyst capable of synthesizing uniform graphene on SiC. These findings underscore the importance of selecting appropriate metals to facilitate graphene growth while minimizing undesired reactions, thus contributing to the optimization of graphene synthesis processes for various applications. Our study successfully demonstrates not only the growth of a continuous graphene layer on a 4-inch SiC wafer at low temperature with an ultra-fast process but also the reproducible synthesis of the free-standing single-crystalline membrane through 2D-based epitaxy. This significant breakthrough facilitates the 3D heterogeneous integration of dissimilar materials, paving the way for the seamless integration of distinct electronic and photonic elements on a single wafer.\u003c/p\u003e "},{"header":"Experimental details","content":"\u003cp\u003e \u003cb\u003eComputational details\u003c/b\u003e \u003c/p\u003e\u003cp\u003eThe first-principles calculations were performed using the projected augmented wave (PAW) plane-wave basis, implemented in the Vienna ab initio simulation package (VASP)\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. An energy cutoff of 520 eV was employed and the atomic positions were optimized using the conjugate gradient scheme without any symmetric restrictions, until the maximum force on each of them was less than 0.01 eV/Å\u003csup\u003e46\u003c/sup\u003e. All atoms were relaxed to their equilibrium positions when the change in energy on each atom between successive steps converged to 1×10\u003csup\u003e− 6\u003c/sup\u003e eV/- atom. The heterostructure was modeled with an 8×8×1 grid for k-point sampling. The generalized gradient approximation (GGA) exchange-correlation (XC) DFT functional Perdew-Burke-Ernzerhof (PBE) was employed for geometrical optimization and electronic structure calculations\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. The slab models had dangling bonds on the vacuum surface terminated by pseudo-hydrogen atoms with appropriate fractional charges to avoid surface states. To determine the vacuum level, dipole corrections are introduced to compensate for the artificial dipole moment at the open ends (20 Å vacuum space along the c-axis) arising from the periodical boundary condition imposed in these calculations\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. Ab initio molecular dynamics (AIMD) simulations were performed in supercells using DFT calculations with a gamma-centered k-point. The time step was set to 3 fs. Simulations for 3 ps were run with a time step of 3 fs to study the dynamic graphitization process. The temperature of the simulation system was controlled at 2273 K using the Nosé–Hoover thermostat\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e \u003cb\u003eSample preparation\u003c/b\u003e \u003c/p\u003e\u003cp\u003eThe single-crystalline 4° offcut 4H–SiC (0001) substrates were supplied by Cree, Inc.. To remove organic contaminants, substrates were cleaned sequentially for 5 min in acetone, 5 min Iso Propyl Alcohol in an ultrasound bath, and finally dried by nitrogen gun.\u003c/p\u003e\u003cp\u003e \u003cb\u003eMetal deposition and graphene formation/transfer method\u003c/b\u003e \u003c/p\u003e\u003cp\u003eAfter preparing the SiC substrate, we deposited Ni, Fe, and Ru onto each SiC substrate to investigate the effect of metal catalysts on the interface graphene layer. Ni was deposited on the substrate using a DC magnetron sputtering apparatus with an Ar plasma at room temperature (JURA deposition apparatus made by Vakuum Servis Ltd., Czech Republic). The sputtering was carried out in an Ar atmosphere (pressure 0.5 mTorr) with DC power 100 W. Resulting deposition rate was 2.5 nm/min. Fe and Ru were deposited on the substrate using an e-beam evaporator (Korea Vacuum Tech., Korea Republic). The base pressure was maintained below 5 ×10\u003csup\u003e− 7\u003c/sup\u003e torr, with a deposition rate of 0.6 Å/s.\u003c/p\u003e\u003cp\u003eFollowing metal deposition, the samples underwent rapid thermal annealing (RTA). The base pressure during RTA was maintained at 7×10\u003csup\u003e− 3\u003c/sup\u003e torr, with 1 torr of N\u003csub\u003e2\u003c/sub\u003e added to prevent metal oxidation. The samples were annealed for 3 minutes to determine the temperature required to decompose the Si-C bond and assess whether it induces a phase change of metals into silicide. X-ray photoelectron spectroscopy (XPS, K-alpha, Thermo Scientific Inc.) and X-ray diffraction (XRD, Rigaku, SmartLab) with Cukα1 (wavelength 1.54051 Å) were used to identify crystalline phases and structural transformations in the metal/SiC structure. Raman spectroscopy (Horiba Jobin Yvon, LabRam Aramis) equipped with a 532 nm wavelength laser was utilized to analyze phase transformations of metals and to confirm the presence of graphene. Following the reaction, the interfacial graphene layer was analyzed using high-resolution transmission electron microscopy (HRTEM, JEOL ARM200F). To further investigate the interface characteristics, the samples were immersed in a 40% w/v Ferric Chloride (FeCl\u003csub\u003e3\u003c/sub\u003e) solution (for Ni, and Fe) for 5 minutes or a 5% w/v Sodium Hypochlorite (NaOCl) solution (for Ru) for 5 minutes. The samples were then characterized using optical microscopy (OM), scanning electron microscopy (SEM, JEOL, JSM-IT-500HR), Ultraviolet-visible-near-infrared spectrophotometer (UV-vis-NIR, JASCO, V-650), atomic force microscopy (AFM, Park Systems, NX-10), TEM, and Raman spectroscopy.\u003c/p\u003e\u003cp\u003eAfter graphene synthesis, polyvinyl alcohol (PVA) was drop-casted onto the surface and baked at 80°C for 5 minutes to form an adhesion layer. A TRT was then attached, facilitating the detachment of graphene through mechanical exfoliation. The exfoliated graphene was transferred onto an SiO\u003csub\u003e2\u003c/sub\u003e substrate. The TRT was removed by baking the sample at 130°C, and the graphene was revealed by dipping the sample into deionized water.\u003c/p\u003e\u003cp\u003e \u003cb\u003eSiC high-temperature graphitization\u003c/b\u003e \u003c/p\u003e\u003cp\u003eThe wafer was loaded into an Aixtron VP508 reactor for graphitization. It was first cleaned in a hydrogen environment for 30 minutes at 1520°C, followed by annealing at 1580°C in a 700 Torr argon ambient for 10 minutes.\u003c/p\u003e\u003cp\u003e \u003cb\u003eIII\u003c/b\u003e-\u003cb\u003eN epitaxial Growth and exfoliation\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe GaN/AlN hetero-structure was epitaxially grown on a graphene/SiC substrate by a metal-organic chemical vapor deposition (MOCVD) equipped with a vertical showerhead-type chamber from Sysnex Co., Ltd. The MOCVD reactor maintained a stable pressure of 30 Torr with hydrogen as a carrier gas throughout the growth process. Initially, the AlN buffer layer was grown at 1,050°C for 40 mins on 1 ML graphene/SiC and 4 ML graphene/SiC. Subsequently, GaN layer was grown by a single-step growth at 1,100°C for 4 mins on each AlN layer. In this growth process, trimethylgallium, trimethylaluminum, and ammonia were used as the sources of gallium, aluminum, and nitrogen, respectively. The resulting GaN/AlN epilayers were exfoliated using a Ti/Ni stressor stack with a handling layer. A 50 nm thick Ti layer was deposited as an adhesion layer for the Ni stressor layer via e-beam evaporation, followed by the deposition of a 3.5 µm thick Ni stressor layer using DC magnetron sputtering under an argon ambient. After the deposition of Ti/Ni layers, thermal release tape (TRT) was attached as a handling layer. Finally, the TRT/Ti/Ni/epi stack was lifted from the edges, enabling precise and controlled exfoliation of the GaN/AlN epilayers.\u003c/p\u003e\u003cp\u003e \u003cb\u003eIII\u003c/b\u003e-\u003cb\u003eN membrane characterizations\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe plan-view, cross-sectional, and EBSD images of the grown and exfoliated samples were obtained using a field-emission SEM system (SU8220, Hitachi). X-ray diffraction characterization was carried out using an XRD measurement system with Cu K-α radiation (Empyrean, Malvern Panalytical).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eDeclaration of Competing Interest\u003c/h2\u003e \u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAcknowledgments\u003c/h2\u003e \u003cp\u003eThe team at Yonsei University would like to acknowledge support from the National Research Foundation of Korea (NRF) (grant no. RS-2023-00222070 and grant no. RS-2024-00445081) and LX Semicon.\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eData will be made available on request.\u003c/p\u003e\n\u003ch3\u003eCredit authorship contribution statement\u003c/h3\u003e\n\u003cp\u003e \u003cb\u003eSe H. Kim\u003c/b\u003e: Conceptualization, Writing – original draft, Visualization, Methodology, Investigation. \u003cb\u003eHanjoo Lee\u003c/b\u003e: Investigation, Writing – original draft. \u003cb\u003eDong Gwan Kim\u003c/b\u003e: DFT simulation, Writing – original draft. \u003cb\u003eDonghan Kim\u003c/b\u003e: Investigation, Visualization, Validation. \u003cb\u003eSeugki Kim\u003c/b\u003e: Visualization, Validation. \u003cb\u003eHyunho Yang\u003c/b\u003e: Visualization, Validation. \u003cb\u003eYunsu Jang\u003c/b\u003e: Visualization, Validation. \u003cb\u003eJangho Yoon\u003c/b\u003e: Validation. \u003cb\u003eHyunsoo Kim\u003c/b\u003e: Resources. \u003cb\u003eSeoyong Ha\u003c/b\u003e: Resources. \u003cb\u003eByuongTak Lee\u003c/b\u003e: Resources. \u003cb\u003eJung-Hee Lee\u003c/b\u003e: Resources. \u003cb\u003eRoy Byung Kyu Chung\u003c/b\u003e: Resources. \u003cb\u003eHongsik Park\u003c/b\u003e: Validation, Resources. \u003cb\u003eSungkyu Kim\u003c/b\u003e: Validation, Resources. \u003cb\u003eTaehoon Lee\u003c/b\u003e: Validation, Resources. \u003cb\u003eHyun S. Kum\u003c/b\u003e: Conceptualization, Methodology, Investigation, Resources, Writing – review \u0026amp; editing, Project administration, Supervision.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eKum H et al (2019) Epitaxial growth and layer-transfer techniques for heterogeneous integration of materials for electronic and photonic devices. Nature Electronics vol. 2 439\u0026ndash;450 Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41928-019-0314-2\u003c/span\u003e\u003cspan address=\"10.1038/s41928-019-0314-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim H et al (2022) Remote epitaxy. Nat Reviews Methods Primers 2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang CS et al (2023) Remote epitaxial interaction through graphene. Sci Adv 9:42\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim J et al (2014) Principle of direct van der Waals epitaxy of single-crystalline films on epitaxial graphene. Nat Commun 5\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNarayan J (2013) Recent progress in thin film epitaxy across the misfit scale (2011 Acta Gold Medal Paper). Acta Mater 61:2703\u0026ndash;2724\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKum HS et al (2020) Heterogeneous integration of single-crystalline complex-oxide membranes. Nature 578:75\u0026ndash;81\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim Y et al (2017) Remote epitaxy through graphene enables two-dimensional material-based layer transfer. Nature 544:340\u0026ndash;343\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiao K et al (2021) Graphene Buffer Layer on SiC as a Release Layer for High-Quality Freestanding Semiconductor Membranes. Nano Lett 21:4013\u0026ndash;4020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEmtsev KV et al (2009) Towards wafer-size graphene layers by atmospheric pressure graphitization of silicon carbide. Nat Mater 8:203\u0026ndash;207\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerger C et al (2006) Electronic Confinement and Coherence in Patterned Epitaxial Graphene. Sci (1979) 312:1191\u0026ndash;1196\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJuang ZY et al (2009) Synthesis of graphene on silicon carbide substrates at low temperature. Carbon N Y 47:2026\u0026ndash;2031\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYuan W, Li C, Li D, Yang J, Zeng X (2011) Preparation of single- and few-layer graphene sheets using Co deposition on sic substrate. \u003cem\u003eJ Nanomater\u003c/em\u003e (2011)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIacopi F et al (2015) A catalytic alloy approach for graphene on epitaxial SiC on silicon wafers. J Mater Res 30:609\u0026ndash;616\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eR\u0026oslash;st HI et al (2021) Low-Temperature Growth of Graphene on a Semiconductor. J Phys Chem C 125:4243\u0026ndash;4252\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEscobedo-Cousin E et al (2012) Trans Tech Publications Ltd,. Local solid phase epitaxy of few-layer graphene on silicon carbide. in Materials Science Forum vols 717\u0026ndash;720 629\u0026ndash;632\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacH\u0026aacute;č P, Fidler T, Cichoň S, Miškov\u0026aacute; L (2012) Synthesis of graphene on SiC substrate via Ni-silicidation reactions. Thin Solid Films 520:5215\u0026ndash;5218\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMach\u0026aacute;č P, Fidler T, Cichoň S, Jurka V (2013) Synthesis of graphene on Co/SiC structure. J Mater Sci: Mater Electron 24:3793\u0026ndash;3799\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEscobedo-Cousin E et al (2014) Trans Tech Publications Ltd,. Solid phase growth of graphene on silicon carbide by nickel silicidation: Graphene formation mechanisms. in Materials Science Forum vols 778\u0026ndash;780 1162\u0026ndash;1165\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim S et al (2016) Interfacial reactions in Ni/6H-SiC at low temperatures. J Nanosci Nanotechnol 16:10853\u0026ndash;10857\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwon Y, An BS, Yang CW (2018) Direct observation of interfacial reaction of Ni/6H-SiC and carbon redistribution by in situ transmission electron microscopy. Mater Charact 140:259\u0026ndash;264\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026auml;hnel A, Ischenko V, Woltersdorf J (2008) Oriented growth of silicide and carbon in SiC-based sandwich structures with nickel. Mater Chem Phys 110:303\u0026ndash;310\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChou TC, Joshi A, Wadsworth J (1991) Solid state reactions of SiC with Co, Ni, and Pt. J Mater Res 6:796\u0026ndash;809\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang WM, Zheng ZX, Ding HF, Jin ZH (2002) A Study of the Solid State Reaction between Silicon Carbide and Iron. Mater Chem Phys 74\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu M, Huang H, Wu Y, Wu X (2024) Mechanism of solid-state diffusion reaction in vacuum between metal (Fe, Ni, and Co) and 4H\u0026ndash;SiC. Ceram Int 50:17930\u0026ndash;17939\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchlesinger ME (1990) Thermodynamics of Solid Transition-Metal Silicides. Chem Rev 90 \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pubs.acs.org/sharingguidelines\u003c/span\u003e\u003cspan address=\"https://pubs.acs.org/sharingguidelines\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNash BP, Nash A The Ni-Si (Nickel-Silicon) System Equilibrium Diagram\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerring L, Bussy F, Gachon JC, Feschotte P (1999) The Ruthenium-Silicon System. J Alloys Compd vol. 284\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMattevi C, Kim H, Chhowalla M (2011) A review of chemical vapour deposition of graphene on copper. J Mater Chem 21:3324\u0026ndash;3334\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao Y, Nyborg L, Jelvestam U (2009) XPS calibration study of thin-film nickel silicides. Surf Interface Anal 41:471\u0026ndash;483\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhtsu N, Oku M, Satoh K, Wagatsuma K (2013) Dependence of core-level XPS spectra on iron silicide phase. Appl Surf Sci 264:219\u0026ndash;224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Vliet S, Troglia A, Olsson E, Bliem R (2023) Identifying silicides via plasmon loss satellites in photoemission of the Ru-Si system. Appl Surf Sci 608\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoshino Y, Matsumoto S, Nakada T, Kido Y (2004) Interfacial reactions between ultra-thin Ni-layer and clean 6H-SiC(0 0 0 1) surface. Surf Sci 556:78\u0026ndash;86\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao Y et al (2010) Probing layer number and stacking order of few-layer graphene by Raman Spectroscopy. Small 6:195\u0026ndash;200\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee DS et al (2008) Raman spectra of epitaxial graphene on SiC and of epitaxial graphene transferred to SiO2. Nano Lett 8:4320\u0026ndash;4325\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBae SH et al (2020) Graphene-assisted spontaneous relaxation towards dislocation-free heteroepitaxy. Nat Nanotechnol 15:272\u0026ndash;276\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y et al (2020) Flexible graphene-assisted van der Waals epitaxy growth of crack-free AlN epilayer on SiC by lattice engineering. Appl Surf Sci 520\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePradhan DK et al (2024) Materials for high-temperature digital electronics. Nat Rev Mater. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41578-024-00731-9\u003c/span\u003e\u003cspan address=\"10.1038/s41578-024-00731-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou C et al (2017) Review\u0026mdash;The Current and Emerging Applications of the III-Nitrides. ECS J Solid State Sci Technol 6:Q149\u0026ndash;Q156\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGong J et al (2023) Synthesis and Characteristics of Transferrable Single-Crystalline AlN Nanomembranes. Adv Electron Mater 9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu Y et al (2017) Growth Model of van der Waals Epitaxy of Films: A Case of AlN Films on Multilayer Graphene/SiC. ACS Appl Mater Interfaces 9:44001\u0026ndash;44009\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu Y et al (2021) Demonstration of epitaxial growth of strain-relaxed GaN films on graphene/SiC substrates for long wavelength light-emitting diodes. Light Sci Appl 10\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAvouris P, Dimitrakopoulos C (2012) Graphene: Synthesis and Applications\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen W, Capano MA (2005) Growth and characterization of 4H-SiC epilayers on substrates with different off-cut angles. J Appl Phys 98\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Z et al (2019) Improved Epitaxy of AlN Film for Deep-Ultraviolet Light-Emitting Diodes Enabled by Graphene. Adv Mater 31\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKresse G, Furthm\u0026uuml; J (1996) Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 54:11169\u0026ndash;11186\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlochl PE (1994) Projector augmented-+rave method. Phys Rev B 50:17953\u0026ndash;17979\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerdew JP, Burke K, Ernzerhof M (1996) Generalized Gradient Approximation Made Simple. Phys Rev Lett 77:3865\u0026ndash;3868\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMakov G, Payne MC (1995) Periodic boundary conditions in ab intio calculations. Phys Rev B 51:4014\u0026ndash;4022\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNos\u0026eacute; S (1984) A unified formulation of the constant temperature molecular dynamics methods. J Chem Phys 81:511\u0026ndash;519\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoover WG (1985) Canonical dynamics: Equilibrium phase-space distributions. Phys Rev (Coll Park) 31\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Graphitization, Ultra-wide bandgap, van der Waals epitaxy, Remote Epitaxy, 2D-coated substrate, Graphene","lastPublishedDoi":"10.21203/rs.3.rs-5970972/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5970972/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNon-conventional epitaxial techniques, such as van der Waals epitaxy (vdWE) and remote epitaxy, have attracted substantial attention in the semiconductor research community for their exceptional capability to continuously produce high-quality free-standing films on a single mother wafer without needing surface refurbishment. The successful implementation of these emerging epitaxial techniques crucially hinges on creating a robust uniform two-dimensional (2D) material surface at the wafer-scale and with atomically precise uniformity. The conventional method for fabricating graphene on a silicon carbide (SiC) wafer is through high-temperature graphitization, which produces epitaxial graphene on the surface of the SiC wafer. However, the extremely high temperature needed for silicon sublimation (typically above 1500\u0026deg;C) causes step-bunching of the SiC surface in addition to the growth of uneven graphene at the edges of the step, leading to multilayer graphene stripes and unfavorable surface morphology for epitaxial growth. Here, we fully develop a graphitization technique that allows fast synthesis of single-crystalline graphene at ultra-low temperatures (growth time of less than 1 minute and growth temperature of less than 500\u0026deg;C) at wafer-scale by metal-assisted graphitization (MAG). We found annealing conditions that enable SiC dissociation while avoiding silicide formation, which produces single-crystalline graphene while maintaining atomically smooth surface morphology. The thickness of the graphene layer can be precisely controlled by varying the metal thickness or annealing temperature, allowing the substrate to be utilized for either a remote epitaxial growth substrate or a vdWE growth substrate, depending on the thickness of the graphene. We successfully produce freestanding single-crystalline ultra-wide bandgap (AlN, GaN) films on graphene/SiC via the 2D material-based layer transfer (2DLT) technique. The exfoliated films exhibit high crystallinity and low defect densities. Our results show that low-temperature graphene synthesis via MAG represents a promising route for the commercialization of the 2D-based epitaxy technique, enabling the production of large-scale ultra-wide bandgap free-standing crystalline membranes.\u003c/p\u003e","manuscriptTitle":"Ultralow-temperature ultrafast formation of single-crystalline graphene via metal-assisted graphitization of silicon-carbide","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-02-28 05:39:24","doi":"10.21203/rs.3.rs-5970972/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"fe517e24-cbdd-4f79-b2ec-5a16595d048f","owner":[],"postedDate":"February 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":44566660,"name":"Physical sciences/Nanoscience and technology/Graphene/Synthesis of graphene"},{"id":44566661,"name":"Physical sciences/Materials science/Theory and computation/Electronic structure"},{"id":44566662,"name":"Physical sciences/Materials science/Nanoscale materials/Synthesis and processing"},{"id":44566663,"name":"Physical sciences/Materials science/Nanoscale materials/Graphene/Synthesis of graphene"},{"id":44566664,"name":"Physical sciences/Materials science/Materials for devices/Electronic devices"}],"tags":[],"updatedAt":"2025-02-28T05:39:25+00:00","versionOfRecord":[],"versionCreatedAt":"2025-02-28 05:39:24","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5970972","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5970972","identity":"rs-5970972","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}