www.acsnano.org
Controlled Growth of Large-Area Uniform Multilayer Hexagonal Boron Nitride as an Effective 2D Substrate Yuki Uchida,† Sho Nakandakari,† Kenji Kawahara,‡ Shigeto Yamasaki,† Masatoshi Mitsuhara,† and Hiroki Ago*,†,‡ Downloaded via UNIV OF SUSSEX on June 16, 2018 at 09:47:56 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Interdisciplinary Graduate School of Engineering Sciences and ‡Global Innovation Center (GIC), Kyushu University, Fukuoka 816-8580, Japan S Supporting Information *
ABSTRACT: Multilayer hexagonal boron nitride (h-BN) is an ideal insulator for two-dimensional (2D) materials, such as graphene and transition metal dichalcogenides, because h-BN screens out influences from surroundings, allowing one to observe intrinsic physical properties of the 2D materials. However, the synthesis of large and uniform multilayer h-BN is still very challenging because it is difficult to control the segregation process of B and N atoms from metal catalysts during chemical vapor deposition (CVD) growth. Here, we demonstrate CVD growth of multilayer h-BN with high uniformity by using the Ni−Fe alloy film and borazine (B3H6N3) as catalyst and precursor, respectively. Combining Ni and Fe metals tunes the solubilities of B and N atoms and, at the same time, allows one to engineer the metal crystallinity, which stimulates the uniform segregation of multilayer hBN. Furthermore, we demonstrate that triangular WS2 grains grown on the h-BN show photoluminescence stronger than that grown on a bare SiO2 substrate. The PL line width of WS2/h-BN (the minimum and mean widths are 24 and 43 meV, respectively) is much narrower than those of WS2/SiO2 (44 and 67 meV), indicating the effectiveness of our CVD-grown multilayer h-BN as an insulating layer. Large-area, multilayer h-BN realized in this work will provide an excellent platform for developing practical applications of 2D materials. KEYWORDS: hexagonal boron nitride, chemical vapor deposition, multilayer growth, tungsten disulfide, van der Waals heterostructure
R
mobility and/or optical properties of the 2D materials. For example, the carrier mobility of graphene can be increased from 10000 to 30000−50000 cm2/V·s, by introducing multilayer hBN between a graphene channel and SiO2 surface.2 This mobility can be further increased to 350000 cm2/V·s (at 1.6 K) by encapsulating the graphene channel with h-BN flakes.13 The carrier mobility of transition metal dichalcogenides (TMDCs) was also increased by h-BN flakes, reaching 1000−34000 cm2/ V·s at low temperatures.14 Furthermore, optical properties of TMDCs were substantially improved by growing WS2 on multilayer h-BN instead of on the SiO2 surface.12 The photoluminescence (PL) of WS2 becomes much stronger and sharper (full width at half-maximum (fwhm) is 26 meV) on hBN, as compared to that grown on a conventional SiO2 substrate surface (61 meV).12 Such drastic improvement by the h-BN thin film is accounted for by the reduction of the
ecently, two-dimensional (2D) materials have been extensively studied due to their excellent electronic and optical properties, making them promising candidates for next generation electronic and photonic devices.1−5 However, they are very sensitive to surroundings, such as substrate surface and gas adsorption, due to their extremely thin structures, with most of the atoms exposed to the surface. Therefore, it is important to find proper ways to isolate the 2D materials from their surroundings. Hexagonal boron nitride (hBN), consisting of a honeycomb lattice of boron and nitrogen atoms, is an insulating 2D material with a large band gap of 5.9 eV and high chemical and thermal stability.6 Thin films of h-BN have a wide range of potential applications, such as deep ultraviolet light emission source, gas barrier films, and tunneling barriers.7−9 More importantly, a h-BN thin film is widely recognized as an ideal dielectric substrate for bringing out the intrinsic properties of other 2D materials due to its atomically smooth surface and the absence of dangling bonds.2,10−17 The introduction of the h-BN to the interface between a 2D material and substrate surface significantly improves the carrier © XXXX American Chemical Society
Received: April 24, 2018 Accepted: June 4, 2018 Published: June 4, 2018 A
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
Cite This: ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano
Figure 1. (a) Schematic of the CVD growth of multilayer h-BN on a Ni−Fe alloy film catalyst. (b,c) Optical and AFM images of h-BN grown on Ni−Fe film deposited on a spinel(100) substrate measured after the transfer on SiO2/Si substrate. Lower panel of (c) shows the height profile measured along the yellow line in the AFM image. (d) Raman mapping image of E2g intensity of the h-BN film without pixel interpolation. The interval of each measured spot is 0.67 μm for both x- and y-axes. (e) Raman spectra collected at the positions marked in (d). (f,g) Cross-sectional TEM images of the as-grown h-BN on the Ni−Fe/spinel(100).
with very small grain size, and it lacks film uniformity.29−31 It is reported that the catalytic activity of Ni foils toward h-BN growth is strongly dependent on the crystalline planes, thus giving h-BN flakes with a large variation in thickness on the polycrystalline Ni foil.32 Recently, Kim et al. demonstrated the synthesis of multilayer h-BN with high crystallinity by using Fe foil as a catalyst.27 Since Fe metal has relatively high solubilities for both B and N atoms, the dissolved B and N atoms in the Fe metal can form multilayer h-BN via a segregation process. However, there remain difficulties in obtaining a uniform h-BN film using Fe catalyst from two reasons. First, Fe metal changes its crystal structure from body-centered cubic (bcc) to facecentered cubic (fcc) at ∼912 °C. Thus, because the growth temperature (typically 1000−1100 °C) is higher than this transition temperature, the Fe catalyst experiences structural deformation during both the heating up and cooling down steps in the CVD. As the h-BN segregation takes place preferentially at grain boundaries of the metal catalyst, such structural transformation is detrimental for the growth of a uniform h-BN film. In addition, this structural change of the metal catalyst would introduce large strain in the as-grown multilayer h-BN. Second, as the solubilities of B and N atoms in Fe metal are largely different (B, ∼0.1 atom % at 1149 °C; N, ∼8 atom % at 1000 °C),27,33 it is considered that the unbalanced solubility prevents uniform segregation of h-BN. Here, we demonstrate the CVD synthesis of a highly uniform multilayer h-BN sheet by using a Ni−Fe alloy film and borazine (B3N3H6) as catalyst and feedstock, respectively. The Ni metal stabilizes the fcc structure, different from pure Fe films/foils
influences from the underlying substrates (e.g., SiO2/Si substrates) which have charged impurities, optical phonons, and relatively rough surface. Thus, thin films of h-BN are highly desirable as a dielectric layer for various 2D materials. In most of the 2D heterostructure devices, mechanically exfoliated h-BN flakes obtained from bulk h-BN crystals have been used.2,10−16 Their limited size (typical lateral size is ∼10 μm) and inhomogeneous flake thickness hinder further development of 2D materials-based electronic and photonic applications. Therefore, the large-scale synthesis of multilayer h-BN films with high uniformity is strongly required for developing practical applications by utilizing the potential of various 2D materials. Chemical vapor deposition (CVD) utilizing a metal catalyst is expected to be a promising method to produce large-area h-BN films at relatively low cost. In the hBN CVD, like the case of graphene growth, Cu films/foils have been widely used, which dominantly produce monolayer h-BN on the Cu surface.18−23 The growth of monolayer h-BN is explained by the self-limiting growth mechanism due to the very low solubility of N atoms in Cu metal.24 The grain size and growth rate of monolayer h-BN can be increased by utilizing a Cu−Ni alloy catalyst.25,26 However, the monolayer of h-BN is not thick enough to reduce the influences from underlying substrates due to its single-atom thickness. Therefore, recent effort has been paid more on the synthesis of multilayer h-BN films.27−31 The growth of multilayer h-BN on the Cu catalyst is possible, but it relies on gas-phase thermal decomposition of feedstock, mainly ammonia borane (BH3NH3), so that the crystallinity of the h-BN is rather low B
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano
Figure 2. (a−d) Optical images of the transferred h-BN grown on various catalysts. (e−h) Thickness distribution of h-BN determined from the G value of the optical microscope images. The purple shade indicates the optical contrast corresponding to the SiO2 surface. Inset in (f) is an AFM image of triangular h-BN grain measured after the transfer. Scale bar in the inset is 3 μm.
technique.34 This optical image demonstrates the controlled synthesis of the large-area uniform h-BN sheet. To analyze the surface structure, the h-BN surface was measured by an atomic force microscope (AFM), as shown in Figure 1c. The area having a hole was measured to show the presence of multilayer h-BN. This image confirms the formation of a uniform and continuous film, consistent with the optical image. The height of the h-BN film was estimated to be ∼3 nm (bottom panel of Figure 1c), which corresponds to ∼8 h-BN layers. Raman spectroscopy is a powerful tool to analyze various 2D materials because it can identify 2D materials from their specific Raman frequencies. The Raman spectra of the transferred film (Figure 1e) showed a characteristic E2g band of h-BN at 1365− 1370 cm−1, verifying the growth of h-BN on the alloy catalyst. It is known that the E2g intensity increases linearly with increasing h-BN thickness and that for monolayer h-BN the E2g band is very weak and sometimes its intensity is lower than the detection limit.23,35 Thus, the clear E2g peaks seen in Figure 1e indicate the formation of multilayer h-BN. Moreover, the Raman mapping image of the E2g intensity (Figure 1d) collected at the same area shown in Figure 1c displays uniform color contrast, confirming the high uniformity of the h-BN thickness (except for the small blue area). The deviation from the mean E2g intensity, determined by the Raman mapping, is within 10% in the measured area, again supporting the uniformity of the film thickness. This mapping image is quite different from that of the h-BN sample grown on the Fe foil (Figure S2), which shows large variation in the E2g intensity. Therefore, our h-BN has a highly uniform film thickness. As far as we know, the synthesis of such uniform multilayer h-BN by a catalytic CVD method has not been reported so far. The full width at half-maximum (fwhm) of the Raman E2g band is associated with the h-BN crystallinity. As shown in Figure 1e, the h-BN transferred from Ni−Fe/spinel(100) showed a sharp E2g band with an average fwhm value of 17 cm−1 (see Figure S3 for the histogram of the fwhm). This value is much smaller than that of the multilayer h-BN grown on Cu catalyst via noncatalytic decomposition of the feedstock (∼30 cm−1 or larger).29−31 Thus, it is highly likely that our multilayer h-BN growth is based on the catalytic segregation from the metal catalyst. As the fwhm value of exfoliated h-BN flakes from
that show polycrystalline features. Also, the mixture of Ni and Fe metal can tune the solubility of B and N atoms, resulting in the uniform segregation of multilayer h-BN. The obtained hBN has a thickness of 2−5 nm with good uniformity, and Raman spectra indicate that the h-BN is well crystallized due to the catalytic segregation process. To assess the applicability of the multilayer h-BN to a dielectric layer for 2D materials, we investigated the optical property of tungsten disulfide (WS2) grown on the h-BN. The WS2 grains were aligned on the h-BN surface and showed a much stronger and sharper PL peak compared with those directly grown on the SiO2 substrate. Our method can be developed to the wafer-scale production of highquality, multilayer h-BN that will boost various practical applications utilizing 2D materials.
RESULTS AND DISCUSSION Synthesis of Large and Uniform Multilayer h-BN. Figure 1a illustrates the procedure used to synthesize uniform multilayer h-BN by the CVD process. The multilayer h-BN growth relies on the dissolution of B and N atoms in the metal catalyst and the subsequent segregation as hexagonal BN layers. This dissolution−segregation process induced by the metal catalyst is essential to obtain a multilayer h-BN film with high crystallinity. Multilayer h-BN was synthesized by using the CVD setup shown in Supporting Information, Figure S1. Here, we used borazine as a feedstock because we can avoid the formation of byproducts, such as BN nanoparticles, which are frequently observed when ammonia borane is used as a precursor. In addition, because borazine is liquid with a low boiling point (55 °C), it is much easier to control the supply rate as compared with solid ammonia borane. In this work, we employed spinel (MgAl2O4) and MgO single-crystal substrates with (100) and (111) planes to deposit metal catalyst films. Single-crystalline substrates promote the formation of epitaxial metal films with less grain boundaries compared with conventional metal foils. These substrates also allow the effects of a crystalline plane of a metal catalyst film to be investigated. Figure 1b shows the optical micrograph of a h-BN sheet grown on a Ni−Fe alloy film (Ni-70%) deposited on spinel(100). For the analysis of thickness and uniformity of h-BN, the as-grown h-BN was transferred on a SiO2 (300 nm)/ Si substrate using the standard polymer-mediated transfer C
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano
Figure 3. Crystallographic study of the metal catalysts. Phase maps (a−d) and inverse pole figure maps (e−h) of metal catalysts measured by EBSD. All the scale bars are 100 μm.
bulk crystal is narrower (9 cm−1),7 there is still space to improve the quality of our h-BN film. Figure 1f,g displays cross-sectional transmission electron microscope (TEM) images of the as-grown h-BN on a Ni−Fe/ spinel(100) catalyst. A well-defined layered structure is clearly seen with an interlayer distance of 0.35 nm (Figure 1g). The film thickness determined from Figure 1g is ∼2.4 nm, almost consistent with the AFM height (Figure 1c). In addition, the TEM image shown in Figure 1f indicates uniformity of the hBN film. The wide area image shown in Figure S4 also supports the uniform film thickness. Electron energy loss spectroscopy analysis confirmed the presence of B and N atoms in the specimen (Figure S5). For reliable macroscopic determination of an atomic ratio of B and N atoms, X-ray photoelectron spectroscopy measurement was also carried out for the asgrown h-BN sample. As shown in Figure S6, the peaks from both B and N atoms were clearly observed. The atomic ratio determined from their peak intensities, [B]/[N] = 1:0.86, is close to the ideal stoichiometric ratio of h-BN. Comparison of Different Catalysts. Figure 2a−c compares h-BN films/flakes grown on different metal catalysts supported on spinel(100) and MgO(100) substrates. As already mentioned, the Ni−Fe/spinel(100) produced the continuous multilayer h-BN film, which can also be seen in Figure 2a. However, when we used a Ni−Fe film deposited on the spinel(111) substrate, no clear optical contrast originating in multilayer h-BN was observed, suggesting that the Ni−Fe/ spinel(111) substrate does not give multilayer h-BN (Figure S7). We speculate that this result is related to the crystal structure of the alloy metal catalyst, which will be discussed later. Figure 2b shows the h-BN grown on the Ni−Fe thin film deposited on the MgO(100) substrate instead of spinel(100). Many isolated multilayer h-BN grains were observed whose optical contrast is uniform, suggesting the relatively uniform hBN thickness. The thickness uniformity within one h-BN grain was also confirmed by Raman mapping measurement (Figure S8). The AFM image (Figure 2f inset and Figure S8a,b) shows the formation of triangular h-BN grains with a flat surface whose height is around 4.5 nm, supporting the growth of multilayer h-BN. We observed wrinkles in the AFM image both on the grain surface and outside the grains (see Figure S8a). This suggests the formation of a thin, 1−2 layer h-BN film
covering the Ni−Fe surface in addition to the multilayer h-BN grains. When we used a pure Fe film deposited on spinel(100), the amount of segregated h-BN increased significantly, as shown in Figure 2c. However, the h-BN growth occurs inhomogeneously, deteriorating the uniformity of the h-BN film, as compared to the Ni−Fe film (see Figure 2a). A number of very thick h-BN flakes were observed in addition to multilayer h-BN covering most of the Fe film surface (Figure 2c). The commercial Fe foil (Figure 2d) showed the similar tendency with the Fe/ spinel(100) catalyst, resulting in the nonuniform formation of thick h-BN flakes. These thick h-BN flakes grown on the Fe foil were linearly arranged, suggesting that the h-BN growth occurs via segregation of B and N atoms mainly through grain boundaries existing in the Fe foil. The presence of grain boundaries in the Fe foil is reasonable because the employed Fe foil is polycrystalline and also experiences the phase transformation from bcc to fcc at ∼912 °C that should introduce large strain in the Fe foil. This nonuniform segregation of thick h-BN flakes on Fe foil is consistent with the previous literature.27 The comparison of Figure 2a and 2c signifies that the addition of Ni to Fe catalyst greatly improves the uniformity of the h-BN film thickness by suppressing the formation of thick h-BN flakes. On the other hand, pure Ni films deposited on spinel(100) showed no clear optical contrast on SiO2/Si, indicating the absence of multilayer h-BN (Figure S9). These results can be qualitatively understood based on the solubilities of B and N atoms in Fe and Ni metals (see Supporting Information, Table S1). At the high temperatures close to the CVD temperature, B and N atoms are dissolved in Fe metal with the concentrations of 0.1 and 8%, respectively. On the other hand, almost no N atoms are dissolved in Ni metal, whereas 0.3% of B atoms can be dissolved in Ni. Therefore, we speculate that the introduction of Ni to Fe catalyst reduces the solubility of N atoms and modifies the balance of dissolved B and N atoms, thus enabling the uniform segregation of a multilayer h-BN sheet. Also, as we discuss later, the singlecrystal substrates that are used to support the metal alloy film also assist the uniform segregation of h-BN. The intensity of the G (green) component of optical images is found to well correlate with a thickness of transferred h-BN. D
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano
Figure 4. (a) SEM and (b) AFM images of WS2 grains grown on multilayer h-BN. Lower panel of (b) shows a height profile measured along the yellow line in the AFM image. (c) Representative PL spectra of the WS2 grown on h-BN (red) and SiO2/Si (blue) substrates. (d) Normalized PL spectra. (e) Distributions of the PL line width (total number of collected spectra is >600 for each sample). (f) Relationship between the PL line width and peak position. (g) Curve-fitted PL spectra of WS2 on h-BN (top) and SiO2/Si substrate (bottom). Pink and light blue peaks represent exciton (A) and charged exciton (A−), respectively. The A/(A + A−) ratio is calculated from the peak areas.
The details of the optical analysis are described in Figure S10. The distributions of the G value are compared in Figure 2e−h. Under our measurement condition, a h-BN flake with a G value of 220 roughly corresponds to the thickness of 8 nm based on our AFM measurement. As the G value of a 1−2 layer h-BN overlaps with that of the underlying SiO2 substrate (shaded by purple in Figure 2e−h), it is difficult to differentiate 1−2 layer h-BN from the bare SiO2 surface based on the optical contrast. However, we can clearly see the existence of multilayer h-BN from the histograms. For example, in Figure 2f, we can distinguish the multilayer h-BN from 1 to 2 layers based on the two separated peaks. Therefore, the histogram shown in Figure 2e shows that the Ni−Fe/spinel(100) catalyst gives a multilayer h-BN sheet with uniform film thickness. Mechanism of Catalytic Growth of Multilayer h-BN. For crystallographic investigation of metal catalysts, we measured electron back-scatter diffraction (EBSD) after h-BN growth. The top panels of Figure 3 represent the crystalline phase, whereas the bottom panels display the crystalline plane normal to the metal surface. The red color in Figure 3a,b signifies that the Ni−Fe films deposited on both spinel(100) and MgO(100) substrates have fcc structures. The Ni−Fe film was prepared by depositing a Ni film (700 nm thickness) on the single-crystal substrate, followed by depositing a thin Fe film (300 nm). Thus, an EBSD image of the as-deposited film shows a bcc structure originating from the top Fe layer, as the EBSD pattern is generated by the elastic reflection of backscatter electrons near the surface (down to 50 nm). However, as seen in Figure 3a,b, these metal films showed a fcc structure after the h-BN growth. This suggests the alloying of Ni and Fe through interdiffusion during the high-temperature CVD process. The energy-dispersive X-ray spectroscopy (EDX) analysis verifies the formation of the Ni−Fe alloy (Figure S11). The observation of the fcc crystal phase (Figure 3a,b)
implies that Ni stabilizes the fcc structure of Ni−Fe alloy, as reported previously.33 Unexpectedly, after the CVD, the Ni−Fe films deposited on spinel(100) and MgO(100) substrates exhibited fcc(111) and fcc(110) planes, respectively, as seen in Figure 3e,f. Although the mechanism of the formation of these crystal planes is unclear, the EBSD images in Figure 3e,f reveal these Ni−Fe films are highly crystallized with a low density of grain boundaries in the metal alloys. We believe that the high crystallinity of the Ni−Fe film plays an essential role in the segregation of multilayer h-BN with a uniform thickness. As shown in Figure 3c,d, the Fe spinel(100) and Fe foil showed bcc structures after the CVD. This is in stark contrast with the results of the Ni−Fe films (Figure 3a,b). The observed bcc structure in the Fe catalysts is reasonable as Fe metal has a bcc structure at room temperature (the Fe metals should have a fcc structure at the growth temperature, then return to bcc during the cooling down process). The inverse pole figures (Figure 3g,h) of the Fe foil/film indicate the polycrystalline structures having different crystal planes. This is supposed to be related with the above-mentioned phase transition of the Fe metal. We consider that the presence of many grain boundaries and unbalanced solubilities of B and N atoms in the Fe catalysts are the main reasons of the observed inhomogeneous h-BN films with many thick h-BN flakes (see Figure 2c,d). We also investigated the Ni−Fe alloy film with the different metal ratio, 30/70% Ni/Fe, with an Fe concentration higher than that of Ni. The optical images of the transferred h-BN films are shown in Figure S12a,b. The catalysts produced multilayer h-BN, but its uniformity is lower than that observed for the Ni−Fe film with the high Ni concentration (70 % Ni). The EBSD data (Figure S12c−f) indicate that the metal crystallinity is higher than that of the Fe film/foils, but it is lower than that observed for the Ni-rich film (see Figure 3e). E
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano
graphene) can act as nucleation sites, decreasing the TMDC grain size by increasing the nucleation density. It is possible to fully cover h-BN with WS2, as we demonstrated in the MoS2/ graphene system,39 by extending the growth time or increase the concentration of the feedstock. However, increasing the grain size can be a more challenging task considering strong van der Waals coupling between the stacked 2D materials. As shown in Figure 4c, the WS2 grains grown on the surface of h-BN exhibited intense PL. The clear PL signal supports that the WS2 grains are monolayer, reflecting their direct band gap nature. For comparison, we also measured the PL from WS2 grown on the bare SiO2 substrate (AFM images are shown in Figure S15). Notably, we found that the PL intensity of WS2 grains on h-BN is very strong, ∼15 times higher than that grown on SiO2 substrate (Figure 4c). Moreover, as shown in the normalized spectra (Figure 4d), the PL peak on the h-BN is much narrower than that on the SiO2 substrate. The distributions of PL line width are compared in Figure 4e. The mean line width of WS2/h-BN and WS2/SiO2 are 43 and 67 meV, respectively. There is still variation in the line width of WS2 on h-BN, indicating that the quality of the CVD-grown hBN needs to be improved through further optimization of the growth condition. However, we note that the minimum fwhm value of 24 meV (see Figure 4e) obtained for the WS2 grown on our large-area h-BN is comparable or better than that on exfoliated h-BN.12 This suggests large potential of our CVDgrown multilayer h-BN as an effective insulating layer. The PL spectrum of WS2 is known to originate in the emission from excitons and charged excitons (trions) that are generated by photoexcitation.41 Excess negative charge supplied from the SiO2 surface changes excitons to negative trions. In other words, the presence of the trion peak indicates that WS2 is negatively doped from the SiO2 surface. To estimate the doping level of WS2 grains grown on h-BN and SiO2, a ratio of exciton-based emission (r) to total PL intensity was calculated based on
We think that this inferior crystallinity compared with that of the Ni-rich alloy film is one of the reasons that prevents the uniform h-BN segregation, as the grain boundaries stimulate the h-BN formation along the boundaries. In the case of graphene growth, Cu(111) gives a high-quality graphene sheet with a controlled lattice orientation of the graphene lattice.36 Thus, a spinel(111) substrate was expected to give h-BN with better quality than spinel(100). However, the Ni−Fe/spinel(111) produced a small number of isolated h-BN grains without forming a continuous sheet (Figure S7), suggesting that the amounts of dissolved B and N atoms are not large enough to produce a uniform multilayer h-BN sheet. We measured the EBSD for the Ni−Fe/spinel(100) and the Ni−Fe/spine(111) after hydrogen annealing at 1100 °C in order to observe the crystal structure just before the borazine introduction in the CVD process (Figure S13). We found that the former shows the fcc(100) plane, whereas the latter shows fcc(111). Thus, we speculate that the highest packing density of a fcc(111) plane with high chemical stability among other planes makes it difficult for B and N atoms to dissolve into the Ni−Fe film deposited on spinel(111). This will prevent the formation of multilayer h-BN. On the other hand, the fcc(100) surface has a low packing density, allowing B and N atoms to dissolve into the Ni−Fe film on spinel(100). From Figure 3e, the dissolution−segregation process of B and N atoms is considered to change the fcc(100) structure of the Ni−Fe film to the fcc(111) structure. Further study is necessary to understand the growth mechanism of multilayer h-BN in the present alloy catalysts in terms of the behavior of B and N atoms inside the metal catalyst film. Application of Multilayer h-BN to a TMDC Substrate. To demonstrate the feasibility of our multilayer h-BN films as a dielectric substrate for 2D materials, we investigated the optical properties of TMDC grown on the multilayer h-BN sheet. Although the direct synthesis of the TMDC/h-BN heterostructure was achieved using sulfide-resistant catalysts, such as Au and Ni−Mo, the thickness of h-BN was not precisely controlled to obtain excellent optical property of the overgrown TMDC.37,38 After transferring our CVD-grown multilayer hBN sheet on a SiO2/Si substrate, WS2 was synthesized by the second CVD step with WO3 and S feedstocks using the setup shown in Figure S14. Figure 4a,b shows scanning electron microscope (SEM) and AFM images of the WS2 grains grown on the h-BN surface. The triangular shape and the measured AFM height (∼0.8 nm) indicate that most of the grains are single-crystalline, monolayer WS2. It is noteworthy that these triangular WS2 grains are aligned in two directions on the multilayer h-BN sheet, indicating the epitaxial growth of WS2. This is consistent with the previous literature that reported the aligned growth of WS2 on mechanically exfoliated h-BN flakes.12 These results are indicative of the high crystallinity and clean surface of our CVD-grown multilayer h-BN. The grain size of WS2 grown on h-BN is relatively small (99%, 20 μm thickness). For the growth of h-BN, a catalyst was placed in the CVD chamber, which is connected to the borazine supply system that is kept at −10 °C. The CVD setup is illustrated in Figure S1. While flowing H2 gas, the furnace was heated to 1100 °C and maintained for 1 h to promote alloying of the metal catalyst as well as to improve the crystallinity and surface flatness. Then, borazine vapor was introduced into the CVD chamber with pressure of ∼30 Pa. The typical reaction time was 30 min. After the reaction, the borazine supply was stopped, and the furnace was cooled slowly while controlling the cooling rate, typically 5 °C/min. The as-grown h-BN was transferred from the metal catalyst to a SiO2/Si substrate by the polymer-mediated transfer method using poly(methyl methacrylate) and iron chloride as support layer and etching solution, respectively. For the growth of monolayer WS2 grains, we performed ambientpressure CVD in the presence of the h-BN/SiO2 substrate using WO3 and S powders as precursors. As illustrated in Figure S14, we employed the double-tube configuration for the WS2 synthesis in order to avoid the reaction of the WO3 source with sulfur before sublimation of WO3. The temperatures of S, WO3, and the substrate were set at 165, 1060, and 800 °C, respectively. The metal catalysts used for the h-BN growth were characterized by EBSD (TSL Solutions, OIM) attached to SEM (Ultra55, Zeiss). An optical microscope (NIKON, Eclipse 300) with a CCD camera (Nikon SD-Fi1) was used to measure the uniformity of the h-BN film with the aid of image analysis software (WinROOF, Mitani Co.). The surface and the thickness of the h-BN films were characterized by AFM (Nanoscope V, Bruker). SEM images and EDX data were measured with HITACHI S-4800 and Bruker FlatQUAD instruments, respectively. Cross-sectional TEM images were measured with a
REFERENCES (1) Geim, A. K.; Novoselov, K. S. The Rise of Graphene. Nat. Mater. 2007, 6, 183−191. (2) Dean, C. R.; Young, A. F.; Meric, I.; Lee, C.; Wang, L.; Sorgenfrei, S.; Watanabe, K.; Taniguchi, T.; Kim, P.; Shepard, K. L.; Hone, J. Boron Nitride Substrates for High-Quality Graphene Electronics. Nat. Nanotechnol. 2010, 5, 722−726. (3) Jariwala, D.; Sangwan, V. K.; Lauhon, L. J.; Marks, T. J.; Hersam, M. C. Emerging Device Applications for Semiconducting TwoDimensional Transition Metal Dichalcogenides. ACS Nano 2014, 8, 1102−1120. (4) Solís-Fernández, P.; Bissett, M.; Ago, H. Synthesis, Structure and Applications of Graphene-Based 2D Heterostructures. Chem. Soc. Rev. 2017, 46, 4572−4613. (5) Li, L.; Yu, Y.; Ye, G. J.; Ge, Q.; Ou, X.; Wu, H.; Feng, D.; Chen, X. H.; Zhang, Y. Black Phosphorus Field-Effect Transistors. Nat. Nanotechnol. 2014, 9, 372−377. (6) Li, L. H.; Cervenka, J.; Watanabe, K.; Taniguchi, T.; Chen, Y. Strong Oxidation Resistance of Atomically Thin Boron Nitride Nanosheets. ACS Nano 2014, 8, 1457−1462. (7) Watanabe, K.; Taniguchi, T.; Kanda, H. Direct-Bandgap Properties and Evidence for Ultraviolet Lasing of Hexagonal Boron Nitride Single Crystal. Nat. Mater. 2004, 3, 404−409. (8) Liu, Z.; Gong, Y.; Zhou, W.; Ma, L.; Yu, J.; Idrobo, J. C.; Jung, J.; MacDonald, A. H.; Vajtai, R.; Lou, J.; Ajayan, P. M. Ultrathin HighTemperature Oxidation-Resistant Coatings of Hexagonal Boron Nitride. Nat. Commun. 2013, 4, 2541. (9) Piquemal-Banci, M.; Galceran, R.; Caneva, S.; Martin, M.-B.; Weatherup, R. S.; Kidambi, P. R.; Bouzehouane, K.; Xavier, S.; Anane, A.; Petroff, F.; Fert, A.; Robertson, J.; Hofmann, S.; Dlubak, B.; Seneor, P. Magnetic Tunnel Junctions with Monolayer Hexagonal Boron Nitride Tunnel Barriers. Appl. Phys. Lett. 2016, 108, 102404. G
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano (10) Gannett, W.; Regan, W.; Watanabe, K.; Taniguchi, T.; Crommie, M. F.; Zettl, A. Boron Nitride Substrates for High Mobility Chemical Vapor Deposited Graphene. Appl. Phys. Lett. 2011, 98, 242105. (11) Lee, G.-H.; Yu, Y.-J.; Cui, X.; Petrone, N.; Lee, C.-H.; Choi, M. S.; Lee, D.-Y.; Lee, C.; Yoo, W. J.; Watanabe, K.; Taniguchi, T.; Nuckolls, C.; Kim, P.; Hone, J. Flexible and Transparent MoS2 FieldEffect Transistors on Hexagonal Boron Nitride-Graphene Heterostructures. ACS Nano 2013, 7, 7931−7936. (12) Okada, M.; Sawazaki, T.; Watanabe, K.; Taniguch, T.; Hibino, H.; Shinohara, H.; Kitaura, R. Direct Chemical Vapor Deposition Growth of WS2 Atomic Layers on Hexagonal Boron Nitride. ACS Nano 2014, 8, 8273−8277. (13) Banszerus, L.; Schmitz, M.; Engels, S.; Dauber, J.; Oellers, M.; Haupt, F.; Watanabe, K.; Taniguchi, T.; Beschoten, B.; Stampfer, C. Ultrahigh-Mobility Graphene Devices from Chemical Vapor Deposition on Reusable Copper. Sci. Adv. 2015, 1, e1500222. (14) Cui, X.; Lee, G.-H.; Kim, Y. D.; Arefe, G.; Huang, P. Y.; Lee, C.H.; Chenet, D. A.; Zhang, X.; Wang, L.; Ye, F.; Pizzocchero, F.; Jessen, B. S.; Watanabe, K.; Taniguchi, T.; Muller, D. A.; Low, T.; Kim, P.; Hone, J. Multi-Terminal Transport Measurements of MoS2 Using a van Der Waals Heterostructure Device Platform. Nat. Nanotechnol. 2015, 10, 534−540. (15) Calado, V. E.; Goswami, S.; Nanda, G.; Diez, M.; Akhmerov, A. R.; Watanabe, K.; Taniguchi, T.; Klapwijk, T. M.; Vandersypen, L. M. K. Ballistic Josephson Junctions in Edge-Contacted Graphene. Nat. Nanotechnol. 2015, 10, 761−764. (16) Xu, S.; Wu, Z.; Lu, H.; Han, Y.; Long, G.; Chen, X.; Han, T.; Ye, W.; Wu, Y.; Lin, J.; Shen, J.; Cai, Y.; He, Y.; Zhang, F.; Lortz, R.; Cheng, C.; Wang, N. Universal Low-Temperature Ohmic Contacts for Quantum Transport in Transition Metal Dichalcogenides. 2D Mater. 2016, 3, 021007. (17) Iqbal, M. W.; Iqbal, M. Z.; Khan, M. F.; Shehzad, M. A.; Seo, Y.; Park, J. H.; Hwang, C.; Eom, J. High-Mobility and Air-Stable SingleLayer WS2 Field-Effect Transistors Sandwiched between Chemical Vapor Deposition-Grown Hexagonal BN Films. Sci. Rep. 2015, 5, 10699. (18) Kim, K. K.; Hsu, A.; Jia, X.; Kim, S. M.; Shi, Y.; Hofmann, M.; Nezich, D.; Rodriguez-Nieva, J. F.; Dresselhaus, M.; Palacios, T.; Kong, J. Synthesis of Monolayer Hexagonal Boron Nitride on Cu Foil Using Chemical Vapor Deposition. Nano Lett. 2012, 12, 161−166. (19) Guo, N.; Wei, J.; Fan, L.; Jia, Y.; Liang, D.; Zhu, H.; Wang, K.; Wu, D. Controllable Growth of Triangular Hexagonal Boron Nitride Domains on Copper Foils by an Improved Low-Pressure Chemical Vapor Deposition Method. Nanotechnology 2012, 23, 415605. (20) Tay, R. Y.; Griep, M. H.; Mallick, G.; Tsang, S. H.; Singh, R. S.; Tumlin, T.; Teo, E. H. T.; Karna, S. P. Growth of Large SingleCrystalline Two-Dimensional Boron Nitride Hexagons on Electropolished Copper. Nano Lett. 2014, 14, 839−846. (21) Wu, Q.; Park, J.-H.; Park, S.; Jung, S. J.; Suh, H.; Park, N.; Wongwiriyapan, W.; Lee, S.; Lee, Y. H.; Song, Y. J. Single Crystalline Film of Hexagonal Boron Nitride Atomic Monolayer by Controlling Nucleation Seeds and Domains. Sci. Rep. 2015, 5, 16159. (22) Sinha, S.; Takabayashi, Y.; Shinohara, H.; Kitaura, R. Simple Fabrication of Air-Stable Black Phosphorus Heterostructures with Large-Area hBN Sheets Grown by Chemical Vapor Deposition Method. 2D Mater. 2016, 3, 035010. (23) Uchida, Y.; Iwaizako, T.; Mizuno, S.; Tsuji, M.; Ago, H. Epitaxial Chemical Vapour Deposition Growth of Monolayer Hexagonal Boron Nitride on a Cu(111)/Sapphire Substrate. Phys. Chem. Chem. Phys. 2017, 19, 8230−8235. (24) Kidambi, P. R.; Blume, R.; Kling, J.; Wagner, J. B.; Baehtz, C.; Weatherup, R. S.; Schloegl, R.; Bayer, B. C.; Hofmann, S. In Situ Observations during Chemical Vapor Deposition of Hexagonal Boron Nitride on Polycrystalline Copper. Chem. Mater. 2014, 26, 6380− 6392. (25) Lu, G.; Wu, T.; Yuan, Q.; Wang, H.; Wang, H.; Ding, F.; Xie, X.; Jiang, M. Synthesis of Large Single-Crystal Hexagonal Boron Nitride Grains on Cu-Ni Alloy. Nat. Commun. 2015, 6, 6160.
(26) Lu, G.; Wu, T.; Yang, P.; Yang, Y.; Jin, Z.; Chen, W.; Jia, S.; Wang, H.; Zhang, G.; Sun, J.; Ajayan, P. M.; Lou, J.; Xie, X.; Jiang, M. Synthesis of High-Quality Graphene and Hexagonal Boron Nitride Monolayer In-Plane Heterostructure on Cu-Ni Alloy. Adv. Sci. 2017, 4, 1700076. (27) Kim, S. M.; Hsu, A.; Park, M. H.; Chae, S. H.; Yun, S. J.; Lee, J. S.; Cho, D.-H.; Fang, W.; Lee, C.; Palacios, T.; Dresselhaus, M.; Kim, K. K.; Lee, Y. H.; Kong, J. Synthesis of Large-Area Multilayer Hexagonal Boron Nitride for High Material Performance. Nat. Commun. 2015, 6, 8662. (28) Caneva, S.; Weatherup, R. S.; Bayer, B. C.; Blume, R.; CabreroVilatela, A.; Braeuninger-Weimer, P.; Martin, M.-B.; Wang, R.; Baehtz, C.; Schloegl, R.; Meyer, J. C.; Hofmann, S. Controlling Catalyst Bulk Reservoir Effects for Monolayer Hexagonal Boron Nitride CVD. Nano Lett. 2016, 16, 1250−1261. (29) Koepke, J. C.; Wood, J. D.; Chen, Y.; Schmucker, S. W.; Liu, X.; Chang, N. N.; Nienhaus, L.; Do, J. W.; Carrion, E. A.; Hewaparakrama, J.; Rangarajan, A.; Datye, I.; Mehta, R.; Haasch, R. T.; Gruebele, M.; Girolami, G. S.; Pop, E.; Lyding, J. W. Role of Pressure in the Growth of Hexagonal Boron Nitride Thin Films from Ammonia-Borane. Chem. Mater. 2016, 28, 4169−4179. (30) Jang, S. K.; Youn, J.; Song, Y. J.; Lee, S. Synthesis and Characterization of Hexagonal Boron Nitride as a Gate Dielectric. Sci. Rep. 2016, 6, 30449. (31) Lin, W.-H.; Brar, V. W.; Jariwala, D.; Sherrott, M. C.; Tseng, W.S.; Wu, C.-I.; Yeh, N.-C.; Atwater, H. A. Atomic-Scale Structural and Chemical Characterization of Hexagonal Boron Nitride Layers Synthesized at the Wafer-Scale with Monolayer Thickness Control. Chem. Mater. 2017, 29, 4700−4707. (32) Cho, H.; Park, S.; Won, D.-I.; Kang, S. O.; Pyo, S.-S.; Kim, D.-I.; Kim, S. M.; Kim, H. C.; Kim, M. J. Growth Kinetics of White Graphene (h-BN) on a Planarised Ni Foil Surface. Sci. Rep. 2015, 5, 11985. (33) Dudrová, E.; Selecká, M.; Burea, R.; Kabátová, M. Effect of Boron Addition on Microstructure and Properties of Sintered Fe1.5Mo Powder Materials. ISIJ Int. 1997, 37, 59−64. (34) Ago, H.; Ohta, Y.; Hibino, H.; Yoshimura, D.; Takizawa, R.; Uchida, Y.; Tsuji, M.; Okajima, T.; Mitani, H.; Mizuno, S. Growth Dynamics of Single-Layer Graphene on Epitaxial Cu Surfaces. Chem. Mater. 2015, 27, 5377−5385. (35) Gorbachev, R. V.; Riaz, I.; Nair, R. R.; Jalil, R.; Britnell, L.; Belle, B. D.; Hill, E. W.; Novoselov, K. S.; Watanabe, K.; Taniguchi, T.; Geim, A. K.; Blake, P. Hunting for Monolayer Boron Nitride: Optical and Raman Signatures. Small 2011, 7, 465−468. (36) Ago, H.; Kawahara, K.; Ogawa, Y.; Tanoue, S.; Bissett, M. A.; Tsuji, M.; Sakaguchi, H.; Koch, R. J.; Fromm, F.; Seyller, T.; Komatsu, K.; Tsukagoshi, K. Epitaxial Growth and Electronic Properties of Large Hexagonal Graphene Domains on Cu(111) Thin Film. Appl. Phys. Express 2013, 6, 075101. (37) Fu, L.; Sun, Y.; Wu, N.; Mendes, R. G.; Chen, L.; Xu, Z.; Zhang, T.; Rümmeli, M. H.; Rellinghaus, B.; Pohl, D.; Zhuang, L.; Fu, L. Direct Growth of MoS2/h-BN Heterostructures via a Sulfide-Resistant Alloy. ACS Nano 2016, 10, 2063−2070. (38) Zhang, Z.; Ji, X.; Shi, J.; Zhou, X.; Zhang, S.; Hou, Y.; Qi, Y.; Fang, Q.; Ji, Q.; Zhang, Y.; Hong, M.; Yang, P.; Liu, X.; Zhang, Q.; Liao, L.; Jin, C.; Liu, Z.; Zhang, Y. Direct Chemical Vapor Deposition Growth and Band-Gap Characterization of MoS2/h-BN van der Waals Heterostructures on Au Foils. ACS Nano 2017, 11, 4328−4336. (39) Ago, H.; Endo, H.; Solís Fernández, P.; Takizawa, R.; Ohta, Y.; Fujita, Y.; Yamamoto, K.; Tsuji, M. Controlled van der Waals Epitaxy of Monolayer MoS2 Triangular Domains on Graphene. ACS Appl. Mater. Interfaces 2015, 7, 5265−5273. (40) Ago, H.; Fukamachi, S.; Endo, H.; Solís Fernández, P.; Mohamad Yunus, R.; Uchida, Y.; Panchal, V.; Kazakova, O.; Tsuji, M. Visualization of Grain Structure and Boundaries of Polycrystalline Graphene and Two-Dimensional Materials by Epitaxial Growth of Transition Metal Dichalcogenides. ACS Nano 2016, 10, 3233−3240. (41) Zhu, B.; Chen, X.; Cui, X. Exciton Binding Energy of Monolayer WS2. Sci. Rep. 2015, 5, 09218. H
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX
Article
ACS Nano (42) Wang, Y.; Cong, C.; Yang, W.; Shang, J.; Peimyoo, N.; Chen, Y.; Kang, J.; Wang, J.; Huang, W.; Yu, T. Strain-Induced Direct−indirect Bandgap Transition and Phonon Modulation in Monolayer WS2. Nano Res. 2015, 8, 2562−2572. (43) Kobayashi, Y.; Sasaki, S.; Mori, S.; Hibino, H.; Liu, Z.; Watanabe, K.; Taniguchi, T.; Suenaga, K.; Maniwa, Y.; Miyata, Y. Growth and Optical Properties of High-Quality Monolayer WS2 on Graphite. ACS Nano 2015, 9, 4056−4063.
I
DOI: 10.1021/acsnano.8b03055 ACS Nano XXXX, XXX, XXX−XXX