Nanoscale

Nanoscale

Published on 25 June 2015. Downloaded by University of Illinois - Urbana on 27/07/2015 01:02:13.

COMMUNICATION

Cite this: Nanoscale, 2015, 7, 12820 Received 21st May 2015, Accepted 19th June 2015 DOI: 10.1039/c5nr03352g

View Article Online View Journal | View Issue

Self-assembly and continuous growth of hexagonal graphene flakes on liquid Cu† Seong-Yong Cho,a Min-Sik Kim,a Minsu Kim,a Ki-Ju Kim,a Hyun-Mi Kim,a Do-Joong Lee,b Sang-Hoon Leea and Ki-Bum Kim*a

www.rsc.org/nanoscale

Graphene growth on liquid Cu has received great interest, owing to the self-assembly behavior of hexagonal graphene flakes with aligned orientation and to the possibility of forming a single grain of graphene through a commensurate growth of these graphene flakes. Here, we propose and demonstrate a two-step growth process which allows the formation of self-assembled, completely continuous graphene on liquid Cu. After the formation of full coverage on the liquid Cu, grain boundaries were revealed via selective hydrogen etching and the original grain boundaries were clearly resolved. This result indicates that, while the flakes selfassembled with the same orientation, there still remain structural defects, gaps and voids that were not resolved by optical microscopy or scanning electron microscopy. To overcome this limitation, the two-step growth process was employed, consisting of a sequential process of a normal single-layer graphene growth and self-assembly process with a low carbon flux, followed by the final stage of graphene growth at a high degree of supersaturation with a high carbon flux. Continuity of the flakes was verified via hydrogen etching and a NaCl-assisted oxidation process, as well as by measuring the electrical properties of the graphene grown by the two-step process. Two-step growth can provide a continuous graphene layer, but commensurate stitching should be further studied.

1 Introduction Motivated by its superior electrical,1–3 optical4 and mechanical5 properties, various synthesis methods for graphene growth have been proposed for the formation of high-quality, large-scale graphene.6–11 Among those, graphene growth on a solid Cu surface has been extensively studied due to the feasi-

a Department of Materials Science and Engineering, Seoul National University, Seoul 151-742, Korea. E-mail: [email protected]; Fax: +81-2-885-5820; Tel: +82-2-880-7465 b School of Engineering, Brown University, Providence, Rhode Island 02912, USA † Electronic supplementary information (ESI) available. See DOI: 10.1039/ c5nr03352g

12820 | Nanoscale, 2015, 7, 12820–12827

bility of large-scale synthesis with monolayer-dominant graphene growth.12–14 However, during the growth, grain boundaries are inevitably formed between adjacent graphene grains as a result of the nature of graphene synthesis, and those grain boundaries are known to significantly degrade the quality of the graphene.15–18 While various methods have been proposed to grow relatively large-sized graphene domains (up to ∼mm), either by controlling the nucleation density19–25 or by aiming for commensurate growth of graphene domains,26,27 synthesizing high-quality, large domain size graphene within a short period of processing time still remains a great challenge. In this respect, recent publications by Geng et al. provided several intriguing aspects of graphene growth on liquid Cu by chemical vapour deposition (CVD), compared to that on solid Cu.28–31 Firstly, one does not have to worry about various heterogeneous nucleation sites that inevitably form on solid Cu foil as a result of Cu melting. Secondly, it appears that a uniformity of the size of the hexagonal graphene flakes can be easily controlled by tuning the process parameters. Thirdly, the graphene flakes self-assemble in a close-packed manner when the relative size of the graphene flakes is uniform. Each hexagonal graphene flake appears to be aligned to have the same crystallographic orientation relationship. Lastly, and most importantly, they claimed that graphene flakes are merged with each other to possibly form a single grain when the process is continued to obtain full coverage growth, based on observations of optical and scanning electron microscopies (SEM). Despite the limitations of needing to use a high melting point and a good wettability substrate like W or Mo, which does not intermix with Cu, and the drawback of the large thermal expansion mismatch between the liquid Cu and the graphene during cooling,32 this process certainly provides one of the unique opportunities to form a large size single crystal with predominantly monolayer graphene, and warrants further investigation.33 Not only graphene, but also h-BN can be grown on liquid Cu according to a recent report.34 Even though the reports clearly suggest advantages and impacts of using liquid Cu as a growing substrate, understanding on its growth mechanism is still lacking, and there is more room to

This journal is © The Royal Society of Chemistry 2015

View Article Online


Nanoscale

improve the quality of the graphene by amending grain boundaries between the hexagonal graphene flakes. Specifically, we focused on the last stage of graphene growth on the liquid Cu, namely, the continuous growth of each graphene flake. To pursue this purpose, we first grew hexagonal graphene flakes on liquid Cu, as was suggested by Geng et al.,28 and confirmed the self-assembly of those graphene flakes by SEM and transmission electron microscopy (TEM). Then, the continuous growth of graphene flakes on the liquid Cu was investigated by selective hydrogen etching35,36 and NaCl-assisted oxidation of the underlying Cu37 to reveal defect sites. Our results showed that the stitched domain boundaries were well resolved by these processes and indicated that, even in the fully grown graphene on the liquid Cu, there still existed a high density of defects, voids and small gaps. Such results proved that the continuous growth is not perfect by itself. The discontinuous growth of graphene on liquid Cu has already been reported by Fan et al.38 Their work was focused on healing cracks which formed during solidification of the underlying liquid Cu catalyst. To accomplish better continuous growth with minimal defects and grain boundaries, we employed a two-step growth process by additionally introducing growth conditions with a high degree of supersaturation, with increased carbon flux, at the last stage of growth. Surprisingly, samples grown by the two-step process were neither etched by hydrogen nor oxidized by NaCl. The electrical resistance of the graphene grown by the two-step process showed a much lower value than that for graphene grown on conventional solid Cu, and was even lower than that for graphene grown on liquid Cu via a single-step growth process. Intra-grain (within a single grain) and inter-grain (between two grains through the grain boundaries) resistances were also measured and compared by preparing an electrode17 atop the graphene to prove the excellent self-assembly of the graphene flakes. These results comprise the outstanding quality and properties of a full coverage graphene layer grown by the newly proposed two-step process within a short period of growth time.

2

Communication

cooling down the sample to 1050 °C for 5 to 10 minutes (to solidify the underlying Cu). For the two-step growth, once the growth had proceeded on the liquid Cu with the conditions outlined above, the CH4 flux was further increased to 12 sccm, either on the liquid Cu (at 1100 °C) or after solidifying the Cu (at 1050 °C). For comparison, graphene was also grown on solid Cu at 1030 °C using a typical low-pressure CVD process. The CH4 flux was maintained at 0.5 sccm with 12 sccm of H2 flowing, and the CH4 flux was finally increased up to 2 sccm at the final step for full coverage. The grain size of graphene grown on the solid Cu by low-pressure CVD was approximately 50 μm, which is a similar size to the graphene grown on the liquid Cu in this work. Transfer and pattern fabrication The graphene was transferred by a well-known process, of poly methyl methacrylate (PMMA) coating and subsequent etching of underlying W and Cu layers. In particular, W etching was performed by anodic etching, as proposed by Y. Fan et al.38, using bare Cu foil as a cathode in 2 M NaOH solution. The PMMA/graphene/Cu stack was then floated in ammonium persulfate solution (0.1 M) overnight to completely etch the Cu foil. After etching, the floated PMMA/graphene layer was scooped on a TEM grid (Quantifoil, Ted Pella) or on a SiO2 (285 nm)/Si substrate for patterning and for electrical transport measurement. Then, the PMMA layer was removed by boiling acetone or direct heating at 380 °C under Ar and H2 flow. For the preparation of VDP patterns, Ti/Au (5 nm/80 nm) electrode patterns were deposited by e-beam evaporation using a metal shadow mask. Sheet resistance and Hall mobility were measured by the VDP method. For Hall measurements, the magnetic field was 0.6 T (HL 5500PC, BIO-RAD). For transmission line measurement (TLM) patterning after transferring the graphene onto a 285 nm-thick SiO2/Si wafer, photoresist (AZ 5214) was uniformly coated and was patterned using an MA 6-II aligner. Then, the Ti/Au (5 nm/80 nm) electrode was evaporated and a lift-off process was done to leave the desired patterns. For two points of current/voltage measurements, an Agilent 4156C system was used.

Experimental section

Graphene growth Electropolished Cu foils (Alfa Aesar #13382) were put on top of pieces of W foil with 1 inch square (also Alfa Aesar) and loaded into a quartz tube-type furnace (Lindberg, blue). Then, sequentially, the furnace was pumped down to ∼10−3 Torr, hydrogen flowed at 500 sccm until the chamber reached atmospheric pressure, and heating was started. After reaching a target temperature of 1100 °C, which is above the melting point of Cu, CH4 was introduced at 3.5 sccm as a carbon precursor. The growth time was varied from 30 minutes for partial coverage graphene to longer than 2 hours for full coverage graphene. To reveal grain boundaries of the graphene grown on the liquid Cu, hydrogen etching was performed in situ by shutting the CH4 flux off and by flowing only 500 sccm H2, after


3

Results and discussion

Fig. 1(a) shows a schematic picture of the graphene growth on liquid Cu via a single-step process. The self-assembled graphene was grown at 1100 °C with CH4 and H2 flow rates of 3.5 and 500 sccm, respectively. Fig. 1(b) and (c) show representative SEM (b) and optical microscope images (c) showing the formation of hexagonal-shaped graphene flakes and their selfassembly behavior on a partially grown sample. The inset in Fig. 1(c) shows the corresponding Raman spectrum, which clearly indicates graphene monolayer growth after the transfer to a SiO2/Si substrate. The process conditions were tuned to obtain a roughly uniform size of graphene flakes at CH4 and H2 flow rates of 3.5 and 500 sccm, respectively, to match with

Nanoscale, 2015, 7, 12820–12827 | 12821

View Article Online


Communication

Fig. 1 Characterization of graphene grown on liquid Cu. (a) Schematic illustration of the graphene growth process on the liquid Cu, (b) SEM image of as-grown graphene islands on the liquid Cu for 30 min of growth time showing a self-assembled growth behavior and (c) optical microscope image of graphene grown on the liquid Cu for 50 min followed by oxidation of the Cu substrate. The inset figure in (c) shows a representative Raman spectrum of a single layer of graphene grown on the liquid Cu after its transfer onto a SiO2/Si substrate.

the results of Geng et al.28 in the aspects of graphene flake shape, size and uniformity. In order to induce self-assembly, the formation of hexagonal graphene flakes and their size uniformity are important factors, while the average size does not play a critical role. Our preliminary experimental results showed that a rigorous selection of the process conditions is required; for instance, when the CH4 flow rate was increased above 3.8 sccm, multiple graphene layers grew, while with a slightly lower CH4 flow rate of 3.3 sccm than the optimum flow, the sizes of the graphene flakes were obviously nonuniform. Further reduction of the CH4 flow rate below 3.0 sccm flow resulted in no graphene growth (ESI Fig. S1†). Certainly, it is important to determine what the driving force of this assembly is. We have noted that the graphene flakes move to the liquid area to assemble during cooling when the overall coverage is low. This result indicates that the capillary force is one of the driving forces for this self-assembly.39 However, when the overall coverage is relatively high, the flakes also selfassemble with each other on liquid Cu, which indicates that van der Waals interactions between the flakes is another driving force. Our interest, then, is how well the adjacent graphene flakes commensurately grow and form a continuous layer when the growth process continues on the liquid Cu. It is well known that grain boundaries are formed by grains merging that have different crystallographic orientations. In the case of graphene, non-hexagon-type carbon arrangements, such as heptagon and pentagon arrangements, are formed in the grain boundaries to accommodate lattice mismatch.15 With a series of TEM images and selected area electron diffraction patterns, we confirmed that an individual hexagonal graphene flake is composed of a single grain and the graphene flake is crystallographically well oriented through the same directions (Fig. S2 and S3†). However, these results are not still sufficient to show

12822 | Nanoscale, 2015, 7, 12820–12827

Nanoscale

Fig. 2 Grain boundaries revealed by hydrogen etching and NaClassisted oxidation. SEM images of graphene islands (a) after 2 h of growth time (as-grown) and (b) the corresponding sample after the hydrogen etching for 5 min. Optical microscopy images of (c) graphene after 2 h of growth time and (d) the corresponding sample after the NaCl-assisted oxidation for 24 h. The scale bar indicates 30 μm.

the commensurate growth and single crystallinity of these flakes. To check the continuity of the hexagonal graphene flakes, we first grew graphene to full coverage for 2 hours, with the optimum conditions addressed above, as is shown in Fig. 2(a). This sample was then cooled to 1050 °C and H2 etching was carried out without breaking the vacuum. The hydrogeninduced etching of graphene has already been reported35,36 and is known to selectively etch defect sites of graphene such as the grain boundaries (Fig. 2(b)). Surprisingly, the original hexagonal flakes are well delineated by hydrogen etching, as are the nucleation centers at the center of the flakes. This result clearly demonstrates that, while the graphene flakes are self-assembled, the flakes do not continuously grow with each other at all and there still remain defect sites which are prone to etching by hydrogen. Similar etching results were obtained with extended growth times, such as 3 and 4 hours. Indeed, closer examination of the fully grown samples often shows small voids (Fig. S4 (a)†) and gaps (Fig. S4 (b)†) between the flakes that are not even filled by extending the growth time up to 4 hours. NaCl-assisted oxidation was also employed to observe grain boundaries, as previously proposed by Ly et al.37 The graphene/Cu/W samples prepared by the single-step process were dipped into NaCl (6 wt%) solution and left for 24 hours. Grain boundaries of the graphene grown on liquid Cu by the single-step process were obviously revealed by the NaCl-assisted oxidation, as shown in Fig. 2(d) (and also Fig. S5†). Contrary to the grain boundaries image with wellarranged boundaries, delineated by H2 etching, as shown in Fig. 2(b), NaCl-assisted oxidation does not give perfect alignment. We believe that this is a difference between in situ H2 etching and ex situ NaCl-assisted oxidation. H2 etching gives a more accurate grain boundaries image since H2 etching could be carried out without cooling Cu, and it can contribute to the


View Article Online


Nanoscale

visualization of grain boundaries as they form. On the other hand, NaCl-assisted oxidation inevitably accompanies the cooling and solidification of Cu, which can damage the graphene. Also, the severe oxidation environment can also deteriorate the graphene film and NaCl precipitate might form in the 1 M solution. Here, a difference between the CVD growth of graphene and the conventional CVD process of other materials should be noted. The conventional CVD process is the method of continuously depositing films by using decomposition and reaction of gas precursors on a substrate. Thus, a deposition rate is constant with respect to a deposition time at a given process condition. In contrast, CVD graphene growth is a process critically relying on a catalytic effect of the underlying Cu substrate for the decomposition of carbon precursors and the supply/ diffusion of carbon monomers. Thus, as graphene coverage increases, the overall growth rate of graphene decreases due to a limited supply of carbon monomers, which should be generated by the exposed Cu area. This type of growth mode is well explained by two-dimensional Johnson–Mehl–Avrami (JMA)type growth. The question that arises here is whether the limited monomer supply at a final stage of the growth is still enough to fill the gaps between adjacent grains. These narrow gaps between graphene grains should be distinguished from merged neighboring grains (for instance, Fig. 5(a) and (c)) at the initial stage when carbon attachment is facile due to the larger exposed area. The two grains can stitch each other easily. Obviously, to fill the gaps at the last stage, the overall process should be executed at a high degree of supersaturation, namely at a high flow rate of carbon precursors for acquisition of sufficient carbon monomer in-the flux. However, high flow rate conditions can simultaneously result in a high nucleation rate, which is not preferable for the growth of largegrain-sized graphene. To increase the grain size, one has to critically control the degree of supersaturation to minimize the nucleation rate at the initial stage of growth. However, in this case, one often finds out that the small gaps between the grains and little voids are not well filled by the continuous growth even though one successfully grows large-sized graphene grains. Possible mechanisms of void formation during graphene growth have already been reported by a few researchers. For instance, Li et al.40 suggested that carbon in flux should be supersaturated in order to obtain a continuous graphene layer. They claimed that a small amount of carbon monomer is insufficient to continuously drive carbon attachment to the island edges, and thereby a Cu surface is only partially covered with the graphene islands. Kim et al.41 also suggested models describing graphene growth kinetics and a coverage as a function of a growth time, where they described how a final coverage can be expressed as a function of a supersaturation ratio. Additionally, Eres et al.42 found out that voids can exist under a low carbon feeding rate. In our growth conditions, graphene growth was not observed at a CH4 flow rate of less than 3.0 sccm. This means that our growth condition of 3.5 sccm of CH4 is just above the minimum feeding rate for the growth, but is not enough to fill the gaps between gra-


Communication

phene flakes at the final stage of growth. However, a higher carbon flux condition (3.8 and 4 sccm) at the initial stage of graphene growth on the liquid Cu resulted in the size discrepancy of flakes and the formation of multilayer graphene, as shown in Fig. S1.† Based on these arguments, our conjecture here is that narrow gaps still exist between graphene flakes grown on liquid Cu due to the small feeding rate of carbon monomers and the resultant low supersaturation ratio. In this regard, we propose a two-step growth recipe for the formation of continuous, gap-free graphene, as illustrated in Fig. 3(a) (see Fig. S6† for a reaction scheme). Full coverage graphene can also be prepared by using a high CH4 flow rate at the initial stage, but an over-flux condition definitely results in either multilayer formation (ESI, Fig. S1†) or a large size discrepancy in the graphene flakes. For this reason, we have designed a two-step process which uses a low carbon flux at the initial stage for a sufficiently low nucleation rate, while introducing a high carbon flux (high degree of supersaturation) at the final stage for filling the gaps between adjacent graphene flakes. The carbon flux was maintained after cooling to solid to fill the possible gaps which might be formed by thermal stress due to cooling. Cracks were often observed after cooling the sample down to solid. Such cracks are suspected to form because of a thermal stress induced by Cu solidification,43 which inevitably results in volume shrinkage, as identified by SEM images (ESI, Fig. S7†). A recent study also emphasized the crack formation in a graphene layer grown on liquid Cu.38 To overcome the limitation described above, two-step growth for the stitching process was carried out after cooling to solid (1050 °C), which resulted in lower crack formation. Also, crack formation was significantly suppressed by a slow cooling rate and by maintaining a high carbon flux on solid Cu, even though a two-step process was applied on liquid Cu. We

Fig. 3 (a) Schematic illustration of two-step graphene growth on liquid Cu. The CH4 flux was increased for continuous stitching of graphene flakes and further growth was also carried out after cooling to solid. (b) SEM image of a continuous graphene layer by the two-step method after experiencing the hydrogen etching process which still shows a continuous film, and (c) optical microscope image of the graphene layer on the liquid Cu by the two-step growth, with no damage observed after the NaCl-assisted oxidation. The scale bar indicates 30 μm.

Nanoscale, 2015, 7, 12820–12827 | 12823

View Article Online


Communication

admit that for better commensurate stitching between graphene islands, two-step growth without cooling to solid is desirable because self-assembly can occur easily on the liquid surface. We believe that continuous graphene with good stitching (hexagon carbon rings) can be grown on liquid Cu by twostep carbon feeding, but continuous graphene on liquid Cu can easily be torn due to the thermal stress between the graphene and the Cu. It might be the reason for the low carrier mobility of the continuous graphene layer, which will be discussed. To examine the continuous growth of graphene, hydrogen etching was applied to two-step grown graphene in this study. Fig. 3(b) shows a continuous graphene layer that was formed by two-step growth and hydrogen etching was tried to reveal grain boundaries (a higher – in Fig. S8†). However, we could not observe an empty area even at higher magnifications. The result shows that grain boundaries were not etched by hydrogen etching at all, contrary to the result for the single-step graphene in Fig. 2(b). We do not believe that this continuous graphene shows no grain boundaries with good stitching because two-step growth was carried out after cooling the Cu catalyst, but continuity of the graphene was achieved. NaClassisted oxidation was also employed to check the continuity of the film. Surprisingly, the sample grown by the two-step growth did not show any remarkable oxidation behavior at the grain boundaries, despite 24 hours of oxidation time, as shown in Fig. 3(c) (see Fig. S9† also for detailed information), which is quite different from the behavior of the single-step sample (Fig. 2(d)). These results clearly suggest that the graphene grown by the two-step recipe in this study is completely continuous and has no gaps between the neighboring graphene flakes at all, which was not achievable from the single-step growth. Not only were the grain boundaries imaged via etching, but the electrical transport properties of the two-step grown graphene were evaluated and compared to those of graphene grown on conventional solid Cu. For the characterization, the graphene grown on the solid Cu was transferred to a 285 nmSiO2/Si substrate using a polymer coating and subsequent wetetching technique.10,11 Meanwhile, in the case of our two-step grown sample on the liquid Cu, an underlying W foil was etched by anodic etching, suggested by Fan et al.38 After transferring the graphene onto a target substrate, a Ti/Au electrode was formed to prepare van der Pauw (VDP) geometry. The detailed VDP sample preparation method is described in the

Nanoscale

ESI, Fig. S6.† The results are shown in Fig. 4(a) and (b). Remarkably, the electrical sheet resistance of the graphene grown on the liquid Cu (two-step process) measured by the VDP configuration (298.1 ± 93.6 Ω sq.−1) was much lower than that for the graphene grown on the solid Cu (700.7 ± 472.2 Ω sq.−1) (Fig. 4(a)). On the other hand, the single-step graphene showed higher resistance (1409.36 ± 273.1 Ω sq.−1) than the two-step sample. This might be from the void area and the reduction of the current path. The carrier mobility was also extracted from Hall measurements, as shown in Fig. 4(b). The hole mobility, which indicates natural p-type doping during transfer of the graphene grown on solid Cu, was 504.9 ± 304.2 cm2 V−1 s−1, probably due to the existence of grain boundaries and random orientation. However, graphene grown on the liquid Cu (two-step process) gave a higher Hall mobility of 1063.4 ± 259.7 cm2 V−1 s−1 (also p-type). The single-step grown graphene showed low carrier mobility, which also indicates degradation of the conducting path due to gaps between islands. Both of these electrical properties clearly show that graphene grown on liquid Cu by the two-step process can have a superior electrical transport behavior than that grown on solid Cu. In the case of the sheet resistance measurement by our van der Pauw geometry, in Fig. 4, the overall film resistance was evaluated through the wafer scale which includes all the possible grain boundaries across a millimeter length. It should not be directly compared to the high mobility values of a single-crystal graphene grain of micrometer scale.12,14,26 Moreover, in order to examine the effects of grain boundaries on the electrical properties of the commensurate, gapfree graphene in this study, the electrode patterns were prepared on top of two adjacent graphene grains using a photolithography and lift-off process, as shown in Fig. 5(a). These grain boundaries were formed in the initial stage of graphene growth, where we do not have to worry about the stitching issue due to the low carbon flux. The large domain size of our graphene (∼100 μm) allowed us to easily fabricate the electrode patterns by photolithography. The representative current–voltage curves are shown in Fig. 5(b) with various lengths of current paths. We did not observe any degradation in the electrical performance due to the grain boundaries (across the electrode #4 in Fig. 5(a)), and the resistance value was only dependent on the length of the current path (distance between the electrodes), as shown in Fig. 5(b). Specifically, there was no abrupt increase of the resistance when current flowed across the grain boundary of the graphene (dotted line (1–5) in Fig. 5(b)). The resistance of the pattern from position a to c can be calculated from the following equation:17 Rac ¼ ρab

Fig. 4 (a) Sheet resistances and (b) Hall mobilities of graphene grown on the solid and the liquid Cu.

12824 | Nanoscale, 2015, 7, 12820–12827

ðb

dx þ ρbc a W ðxÞ

ðc b

dx : W ðxÞ

ð1Þ

Here, Ra–c refers to a resistance value between positions a and c, ρ means a resistivity (the dimensions are Ω sq.−1 for graphene since it has a two-dimensional structure), and W is the


View Article Online

Nanoscale

Communication


Table 2 2-Point resistance values measured from the TLM patterning between intra- and inter-grain electrodes of two adjacent graphene grains in Fig. 5(c)

Fig. 5 Comparison of commensurate and incommensurate stitching. (a) Optical microscope image of two adjacent graphene islands aligned in the same orientation; Ti/Au electrodes were deposited by the lift-off process, (b) current vs. applied bias curve from electrode #1 to various electrodes, including intra- (1–2, 1–3 and 1–4) and inter-grain (1–5), (c) optical microscope image of two adjacent graphene islands with misaligned orientation and (d) current vs. applied bias curve from electrode #7 to various electrodes, including intra- (7–6 and 7–5) and intergrain (7–4, 7–3, 7–2 and 7–1). The scale bars indicate 50 μm.

Electrodes

Resistance (Ω)

Resistivity (Ω sq.−1)

2–3 (left grain) 4–5 (inter-grain) 5–6 (right grain)

120 847 213

1360 5082 1455

between electrodes 5 and 6 (right grain) was 1455 Ω sq.−1 as shown in (Table 2). However, the resistivity between electrodes 4 and 5 (inter-grain in Fig. 5(c)) was 5082 Ω sq.−1, which shows degradation of the electrical connection through incommensurate stitching in the grain boundaries. Also, Yu et al.17 already made 6 samples to measure the inter-grain resistance of graphene grown on solid Cu, and all the samples showed higher resistance when crossing grain boundaries. Again, these results clearly prove that the aligned graphene grains show negligible degradation of the electrical transport properties across the commensurate grain boundary.

4 Conclusions Table 1 2-Point resistance values measured from the TLM (transfer length method) patterning between intra- and inter-grain electrodes of two adjacent graphene grains in Fig. 5(a)

Electrodes

Resistance (Ω)

Resistivity (Ω sq.−1)

1–2 (left grain) 1–3 (left grain) 1–4 (left grain) 1–5 (inter-grain) 3–5 (inter-grain) 3–6 (inter-grain) 4–6 (right grain) 5–6 (right grain)

307 412 735 1048 840 885 441 356

1458 939 1009 1002 1260 1254 843 1566

width of the graphene. Based on eqn (1), the resistivity values within the left and right grains (intra-grain) were calculated and are listed in Table 1. The resistivity between electrodes 3 and 5 (inter-grain through the grain boundary) was also calculated as 1260 Ω sq.−1, and this value is almost in the same regime as the intra-grain resistivities obtained from the left and right grains. In addition, the resistance value between electrodes 3 and 5 (R3–5) was 840 Ω, which is close to the summation of R3–4 and R4–5, both from the intra-grains, calculated from the resistivity of each grain and the sample geometry. For comparison, on the misaligned grain boundaries as shown in Fig. 5(c), the same electrode pattern was deposited and a current–voltage measurement was carried out. The abrupt increase of resistance was verified near the grain boundary (between electrodes 4 and 5). The resistivity between electrodes 2 and 3 (left grain) was 1360 Ω sq.−1 and the resistivity


Although the growth of hexagonal graphene flakes on liquid Cu was reported to show self-assembly behavior between the flakes, the degree of continuous growth of the graphene flakes at the final stage of growth has not been reported. In this work, we applied hydrogen etching to a fully grown graphene sample to resolve the structural defects in the grain boundaries of the graphene. Our result showed that the hexagonal graphene flakes did not continuously grow under single-step growth conditions because of the limited supply of carbon monomers at the final stage. To induce continuous growth, two-step growth is suggested, by additionally introducing a step with increased CH4 flux (from 3.5 to 12 sccm) to provide a high degree of supersaturation at the final stage of growth. Gaps and voids were completely filled by the two-step growth and continuity between the graphene flakes was confirmed by both NaCl-assisted oxidation and H2 etching. Due to solidification of the liquid Cu catalyst, commensurate stitching was not fully carried out. We believe that the potential grain boundaries might be composed of non-hexagonal carbon rings which are the result of incommensurate stitching. However, our graphene grown on liquid Cu shows significantly improved electrical transport behavior, in terms of both sheet resistance and Hall mobility on a wafer scale. The intra- and inter-grain resistance measurements clearly demonstrated the excellent capability of controlling the commensurate grain boundaries in this study. The demonstration and development of selfassembled, continuous and gap-free graphene, with its improved electrical properties, will accelerate the introduction of the two-step growth recipe on liquid Cu into many other growth studies and applications.

Nanoscale, 2015, 7, 12820–12827 | 12825

View Article Online

Communication


Acknowledgements This research was supported by the Pioneer Research Center Program through the National Research Foundation of Korea funded by the Ministry of Science, ICT & Future Planning (2012-0009563). This research was also supported by the Global Frontier Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2014M3A6B206301).

Notes and references 1 S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V. Morozov and A. K. Geim, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 10451. 2 K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666. 3 A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183. 4 R. R. Nair, P. Blake, A. N. Grigorenko, K. S. Novoselov, T. J. Booth, T. Stauber, N. M. Peres and A. K. Geim, Science, 2008, 320, 1308. 5 C. Lee, X. Wei, J. W. Kysar and J. Hone, Science, 2008, 321, 385. 6 C. Berger, Z. M. Song, X. B. Li, X. S. Wu, N. Brown, C. Naud, D. Mayo, T. B. Li, J. Hass, A. N. Marchenkov, E. H. Conrad, P. N. First and W. A. de Heer, Science, 2006, 312, 1191. 7 W. A. de Heer, C. Berger, X. S. Wu, P. N. First, E. H. Conrad, X. B. Li, T. B. Li, M. Sprinkle, J. Hass, M. L. Sadowski, M. Potemski and G. Martinez, Solid State Commun., 2007, 143, 92. 8 J. N. Coleman, M. Lotya, A. O’Neill, S. D. Bergin, P. J. King, U. Khan, K. Young, A. Gaucher, S. De, R. J. Smith, I. V. Shvets, S. K. Arora, G. Stanton, H.-Y. Kim, K. Lee, G. T. Kim, G. S. Duesberg, T. Hallam, J. J. Boland, J. J. Wang, J. F. Donegan, J. C. Grunlan, G. Moriarty, A. Shmeliov, R. J. Nicholls, J. M. Perkins, E. M. Grieveson, K. Theuwissen, D. W. McComb, P. D. Nellist and V. Nicolosi, Science, 2011, 331, 568. 9 Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Sun, S. De, I. T. McGovern, B. Holland, M. Byrne, Y. K. Gun’Ko, J. J. Boland, P. Niraj, G. Duesberg, S. Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A. C. Ferrari and J. N. Coleman, Nat. Nanotechnol., 2008, 3, 563. 10 K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J. H. Ahn, P. Kim, J. Y. Choi and B. H. Hong, Nature, 2009, 457, 706. 11 Y. Lee, S. Bae, H. Jang, S. Jang, S. E. Zhu, S. H. Sim, Y. I. Song, B. H. Hong and J. H. Ahn, Nano Lett., 2010, 10, 490. 12 X. Li, C. W. Magnuson, A. Venugopal, R. M. Tromp, J. B. Hannon, E. M. Vogel, L. Colombo and R. S. Ruoff, J. Am. Chem. Soc., 2011, 133, 2816.

12826 | Nanoscale, 2015, 7, 12820–12827

Nanoscale

13 X. S. Li, W. W. Cai, J. H. An, S. Kim, J. Nah, D. X. Yang, R. Piner, A. Velamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo and R. S. Ruoff, Science, 2009, 324, 1312. 14 X. Li, W. Cai, L. Colombo and R. S. Ruoff, Nano Lett., 2009, 9, 4268. 15 P. Y. Huang, C. S. Ruiz-Vargas, A. M. van der Zande, W. S. Whitney, M. P. Levendorf, J. W. Kevek, S. Garg, J. S. Alden, C. J. Hustedt, Y. Zhu, J. Park, P. L. McEuen and D. A. Muller, Nature, 2011, 469, 389. 16 K. Kim, Z. Lee, W. Regan, C. Kisielowski, M. F. Crommie and A. Zettl, ACS Nano, 2011, 5, 2142. 17 Q. Yu, L. A. Jauregui, W. Wu, R. Colby, J. Tian, Z. Su, H. Cao, Z. Liu, D. Pandey, D. Wei, T. F. Chung, P. Peng, N. P. Guisinger, E. A. Stach, J. Bao, S.-S. Pei and Y. P. Chen, Nat. Mater., 2011, 10, 443. 18 S. Y. Cho, K. J. Kim, H. M. Kim, D. J. Lee, M. H. Lee and K. B. Kim, RSC Adv., 2013, 3, 26376. 19 H. Zhou, W. J. Yu, L. Liu, R. Cheng, Y. Chen, X. Huang, Y. Liu, Y. Wang, Y. Huang and X. Duan, Nat. Commun., 2013, 4, 2096. 20 Y. Hao, M. S. Bharathi, L. Wang, Y. Liu, H. Chen, S. Nie, X. Wang, H. Chou, C. Tan, B. Fallahazad, H. Ramanarayan, C. W. Magnuson, E. Tutuc, B. I. Yakobson, K. F. McCarty, Y. W. Zhang, P. Kim, J. Hone, L. Colombo and R. S. Ruoff, Science, 2013, 342, 720. 21 L. Gan and Z. Luo, ACS Nano, 2013, 7, 9480. 22 Z. Yan, J. Lin, Z. Peng, Z. Sun, Y. Zhu, L. Li, C. Xiang, E. L. Samuel, C. Kittrell and J. M. Tour, ACS Nano, 2012, 6, 9110. 23 G. H. Han, F. Güneş, J. J. Bae, E. S. Kim, S. J. Chae, H.-J. Shin, J.-Y. Choi, D. Pribat and Y. H. Lee, Nano Lett., 2011, 11, 4144. 24 Z. Luo, Y. Lu, D. W. Singer, M. E. Berck, L. A. Somers, B. R. Goldsmith and A. T. C. Johnson, Chem. Mater., 2011, 23, 1441. 25 H. Wang, G. Wang, P. Bao, S. Yang, W. Zhu, X. Xie and W.-J. Zhang, J. Am. Chem. Soc., 2012, 134, 3627. 26 A. W. Tsen, L. Brown, M. P. Levendorf, F. Ghahari, P. Y. Huang, R. W. Havener, C. S. Ruiz-Vargas, D. A. Muller, P. Kim and J. Park, Science, 2012, 336, 1143. 27 V. L. Nguyen, B. G. Shin, D. L. Duong, S. T. Kim, D. Perello, Y. J. Lim, Q. H. Yuan, F. Ding, H. Y. Jeong, H. S. Shin, S. M. Lee, S. H. Chae, Q. A. Vu, S. H. Lee and Y. H. Lee, Adv. Mater., 2015, 27, 1376. 28 D. Geng, B. Wu, Y. Guo, L. Huang, Y. Xue, J. Chen, G. Yu, L. Jiang, W. Hu and Y. Liu, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 7992. 29 D. Geng, B. Luo, J. Xu, Y. Guo, B. Wu, W. Hu, Y. Liu and G. Yu, Adv. Funct. Mater., 2014, 24, 1664. 30 B. Wu, D. Geng, Z. Xu, Y. Guo, L. Huang, Y. Xue, J. Chen, G. Yu and Y. Liu, NPG Asia Mater., 2013, 5, e36. 31 D. Geng, L. Meng, B. Chen, E. Gao, W. Yan, H. Yan, B. Luo, J. Xu, H. Wang, Z. Mao, Z. Xu, L. He, Z. Zhang, L. Peng and G. Yu, Adv. Mater., 2014, 26, 6423. 32 J. A. Cahill and A. D. Kirshenbaum, J. Phys. Chem., 1962, 66, 1080.


View Article Online


Nanoscale

33 Y. A. Wu, Y. Fan, S. Speller, G. L. Creeth, J. T. Sadowski, K. He, A. W. Robertson, C. S. Allen and J. H. Warner, ACS Nano, 2012, 6, 5010. 34 M. H. Khan, Z. Huang, F. Xiao, G. Casillas, Z. Chen, P. J. Molino and H. K. Liu, Sci. Rep., 2015, 5, 7743. 35 I. Vlassiouk, M. Regmi, P. Fulvio, S. Dai, P. Datskos, G. Eres and S. Smirnov, ACS Nano, 2011, 5, 6069. 36 Y. Zhang, Z. Li, P. Kim, L. Zhang and C. Zhou, ACS Nano, 2011, 6, 126. 37 T. H. Ly, D. L. Duong, Q. H. Ta, F. Yao, Q. A. Vu, H. Y. Jeong, S. H. Chae and Y. H. Lee, Adv. Funct. Mater., 2013, 23, 5183. 38 Y. Fan, K. He, H. Tan, S. Speller and J. H. Warner, Chem. Mater., 2014, 26, 4984.


Communication

39 N. Bowden, F. Arias, T. Deng and M. Whitesides, Langmuir, 2001, 17, 1757. 40 X. Li, C. W. Magnuson, A. Venugopal, J. An, J. W. Suk, B. Han, M. Borysiak, W. Cai, A. Velamakanni, Y. Zhu, L. Fu, E. M. Vogel, E. Voelkl, L. Colombo and R. S. Ruoff, Nano Lett., 2010, 10, 4328. 41 H. Kim, C. Mattevi, M. R. Calvo, J. C. Oberg, L. Artiglia, S. Agnoli, C. F. Hirjibehedin, M. Chhowalla and E. Saiz, ACS Nano, 2012, 6, 3614. 42 G. Eres, M. Regmi, C. M. Rouleau, J. Chen, I. N. Ivanov, A. A. Puretzky and D. B. Geohegan, ACS Nano, 2014, 8, 5657. 43 J. J. Hoyt, M. Asta and A. Karma, Mater. Sci. Eng., R, 2003, 41, 121.

Nanoscale, 2015, 7, 12820–12827 | 12827