Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
www.acsami.org
Hydrogen-Assisted Growth of Large-Area Continuous Films of MoS2 on Monolayer Graphene Tongxin Chen, Yingqiu Zhou, Yuewen Sheng, Xiaochen Wang, Si Zhou, and Jamie H. Warner* Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, U.K. S Supporting Information *
ABSTRACT: We show how control over the chemical vapor deposition (CVD) reaction chemistry of molybdenum disulfide (MoS2) by hydrogen addition can enable the direct growth of centimeter-scale continuous films of vertically stacked MoS2 monolayer on graphene under atmospheric pressure conditions. Hydrogen addition enables longer CVD growth times at high temperature by reducing oxidation effects that would otherwise degrade the monolayer graphene. By careful control of nucleation density and growth time, high-quality monolayer MoS2 films could be formed on graphene, realizing all CVD-grown vertically stacked monolayer semimetal/semiconducting interfaces. Photoluminescence spectroscopy shows quenching of MoS2 by the underlying graphene, indicating a good interfacial charge transfer. We utilize the MoS2/graphene vertical stacks as photodetectors, with photoresponsivities reaching 2.4 A/W under 135 μW 532 nm illumination. This approach provides insights into the scalable manufacturing of high-quality two-dimensional electronic and optoelectronic devices. KEYWORDS: MoS2, graphene, CVD, van der Waals heterostructure, photoresponse
■
INTRODUCTION
which can dramatically affect the doping level of TMDs and hence the band gap and light transfer efficiency.22 An ideal solution is to directly synthesize MoS2 on other 2D materials by CVD methods. A few studies have been reported about the realization of direct CVD growth of MoS 2 domains.23,24 However, the MoS2 coverage reported in these studies was relatively low and the size of MoS2 domains synthesized could only reach a submicron scale, which limits their application in device fabrication. Chen et al. reported successful CVD growth of a large-scale bilayer MoS2 film on a graphene substrate, but the photoresponsivity is only 32 mA/ W.25 Besides, most of these synthesis procedures were also carried out at low pressure, which increases the complexity of experimental equipment and processing design. Further work is needed to realize the direct growth of large-area film coverage of monolayer TMD materials on other 2D materials such as graphene and h-BN. In this study, we demonstrate how the direct growth of largearea (1 cm × 1 cm) continuous monolayer MoS2 films on graphene can be realized by the controlled addition of hydrogen to modify the nucleation of domains. We explore the sensitivity of monolayer graphene to decomposition during the reactions
Two-dimensional (2D) materials have received considerable attention since successful synthesis of graphene was first reported.1 Monolayered 2D materials such as hexagonal boron nitride (h-BN)2 and transition metal dichalcogenides (TMDs) including tungsten disulfide,3 molybdenum disulfide (MoS2),4 tungsten diselenide,5 molybdenum diselenide,6 and so forth.7 have been successfully synthesized. Among them, MoS2 has become increasingly investigated because of its semiconductor properties with a direct band gap (1.8 eV), and it has shown a potential in manufacturing the next-generation nanoelectronics, optoelectronics, and flexible devices.8−11 To make such devices, different 2D materials are usually vertically stacked together, giving rise to van der Waals heterostructures that could take the advantage of transparency and flexibility properties of 2D materials.12−16 The successful fabrication of different MoS2 graphene-based heterostructure devices usually involves two steps: individual synthesis of 2D MoS2 by chemical vapor deposition (CVD) growth, exfoliation, or a physical vapor deposition method followed by a layer-by-layer transfer with the aid of a solvent or polymer.17−21 However, scalable fabrication of high-performance devices is hard to achieve by these approaches. The transfer process is not only time-consuming but inevitably introduces imperfections such as cracks, wrinkles, and polymer residuals, © XXXX American Chemical Society
Received: September 29, 2017 Accepted: January 23, 2018
A
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 1. Atmospheric CVD synthesis of MoS2 on the graphene substrate: (a) schematic illustration of the two-furnace CVD system used to synthesize MoS2 on the graphene substrate. (b) Temperature programming process of furnace 1 and furnace 2. (c) Gas flow programming process during the CVD process. (d,e) Photos showing a contrast difference between the graphene films transferred on the SiO2/Si substrate with and without H2, respectively. The embedded figure shows the Raman spectrum of the as-transferred graphene. (f) SEM image of triangular MoS2 domains deposited on graphene. (g) SEM image of the MoS2 film deposited on graphene with the aid of hydrogen.
wafer with an SiO2 surface layer. Graphene was grown by CVD on a Cu catalyst. Sulfur powder and the substrate were placed in the middle of front and back furnaces, respectively, for precise temperature control, while the temperature of MoO3 could be tuned by the distance from the center of furnace 2. The graphene/SiO2 substrate was placed vertically to the direction of gas flow so that the reaction took place uniformly on the substrate. The whole reaction was carried out within two parallel furnaces in an argon atmosphere under an ambient pressure. Similar to the CVD method previously reported, we designed a
and optimize the conditions to enable the growth of continuous MoS2 films within a timescale that allows graphene to survive at the high temperature needed for MoS2 crystallization. The largearea MoS2/graphene vertically stacked layered system is utilized to create arrays of photodetectors with lateral electrodes, yielding high photocurrents and photoresponsivity.
■
RESULTS AND DISCUSSION As depicted in Figure 1a, a two-furnace CVD system was used to synthesize MoS2 on monolayer graphene transferred onto a Si B
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 2. Influence of H2 concentration in the carrying gas on MoS2 and graphene: (a) MoS2 domains grown on the graphene substrate with 15 min nucleation + 15 min growth in a 10 sccm Ar flow; (b) MoS2 domains grown on the graphene substrate with 15 min nucleation + 25 min growth in a 10 sccm Ar flow. Graphene (darker gray) started to be burned away; and (c) MoS2 domains grown on the graphene substrate after 15 min nucleation + 25 min growth in a 7.5 sccm Ar and 2.5 sccm 25% H2/75% Ar gas flow. (d) MoS2 domains grown on the graphene substrate after 15 min nucleation + 25 min growth with a higher H2 concentration. (e) Raman spectrum of graphene before and after CVD synthesis. The black (without H2) and blue (with H2) plots correspond to graphene in (b) and (c), respectively.
multiple-step CVD process with four stages.26 (1) Creation of an S-rich environment in a quartz tube by heating up furnace 1 and furnace 2 to 160 and 200 °C, respectively, for 15 min with a 150 sccm Ar gas flow (Figure 1b,c, section I). (2) Main nucleation stage of MoS2 domains by heating up furnace 1 to 200 °C and furnace 2 to 770 °C which corresponded to the temperature of S and graphene substrate, respectively. The MoO3 precursor was placed 10 cm away from the center of furnace 2, which reached 400 °C when furnace 2 was heated up to 770 °C. The gas flow was set to a relatively high flux to increase the local reactant concentration, which favored the nucleation of MoS2 domains (Figure 1b,c, section II). (3) Slow growth stage by reducing the Ar flow rate to 10 sccm to maintain a slow and stable growth condition which favored the growth of MoS2 domains (Figure 1b,c, section III). (4) When the growth stage ended, the reaction was stopped by cooling down to 700 °C while keeping the Ar flow rate to 10 sccm. Then both furnaces were set to room temperature, and the Ar flow rate was set to 500 sccm for fast cooling. The reason for this multistep CVD setup is to achieve separate control of the nucleation density and then the growth of nuclei into larger domains. In the nucleation stage, the density of MoS2 domains on graphene has a positive correlation with the Ar flow rate, whereas extending the growth time can increase the average MoS2 domain size. By altering the flow rate during nucleation and growth time, we optimized the parameters to grow continuous MoS2 films on graphene. Epitaxial MoS2 triangular domains could be synthesized on the graphene surface using a 50 sccm flow rate during the nucleation period with a 15 min nucleation + 15 min growth CVD process, as shown in Figure 1d,f. However, we noticed that the formation of MoS2 films could
not be easily achieved by altering the parameters within this approach because of the decomposition of graphene. The introduction of hydrogen to the growth enabled faster growth and the formation of large-area continuous MoS2 films could be achieved on graphene, as shown in Figure 1e,g. The region with graphene underneath is about 1 cm × 1 cm, as is pointed out in Figure 1e. Figure 1g is the scanning electron microscopy (SEM) image of the as-synthesized MoS2/graphene heterostructure. The CVD-grown MoS2 film (black region) shows a high continuity that almost completely covers the surface of graphene. The coverage of the monolayer MoS2 could reach up to 96%. The reason MoS2 films could not be formed by simply changing the growth time and flow rate was that the formation of large-area continuous films required a longer growth time, while graphene was observed not to sustain a long-time reaction within our CVD setup. In Figure 2a,b, we present the comparative CVD process with different growth times. We discovered that after an overall 40 min CVD process, graphene started to disappear before the MoS2 films formed. The Raman spectrum of the remaining graphene shows a significant increase of the D peak and a decrease in the 2D peak intensity (Figure 2e), suggesting that graphene was heavily oxidized.27 We speculated that the oxidation of graphene resulted from reactive O released during the decomposition of the MoO3 precursor as well as the decomposition of the SiO2 substrate at high temperature. Therefore, we introduced hydrogen into the carrying gas once the growth temperature reaches 770 °C. Hydrogen serves multiple purposes. First, hydrogen leads to a more efficient reduction of MoO3, aiding the formation of MoS2 by reducing the precursors. Second, it acts as a reducing agent that reacts with reactive O atoms to prevent graphene from C
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 3. SEM images of comparable MoS2 grown on graphene under different conditions. (a−i) SEM images of MoS2 grown on graphene using different nucleation flow rates and growth times. The white dots are three-dimensional (3D) MoS2 particles (Figure S1). Increasing the nucleation flow rate dramatically increased the nucleation density of 3D MoS2 crystals. (j) Growth mechanism of the MoS2 film on graphene.
growth (15 min nucleation + 45 min growth time), the size of the MoS2 domains remained barely unchanged and no continuous MoS2 film was formed, which suggests that there is a limitation in the size of MoS2 domains on the graphene surface, as shown in Figure 3a−c. With the size of individual MoS2 domain restrained in a certain range, the alternative way to grow a continuous layer of the MoS2 film is to increase the nucleation density of MoS2 on graphene so that multiple small MoS2 domains could be linked together. Different nucleation flow rates and growth times were applied to figure out the optimal conditions for the formation of a continuous MoS2 film (Figure 3). We observed that with a high nucleation flow rate, 2D MoS2 domains as well as 3D crystals nucleated, and these 3D crystals grew larger as the growth time increased. To maximize the nucleation density of MoS2 on graphene but minimize the formation of 3D crystals, we tried different flow rates of nucleation stage and growth stage times. The optimal recipe to CVD grow MoS2 films on graphene in our system was a 15 min nucleation stage in 150 sccm Ar followed by a 30 min growth stage in 7.5 sccm Ar + 2.5 sccm 25% hydrogen/75% Ar. A continuous MoS2 film could form on the top of the graphene substrate, while at the same time the formation of 3D MoS2 crystals was limited to a very low level (Figure 4e). It should also be noticed in Figure 3d that upon increasing the nucleation density, the MoS2 domains grown on graphene start to lose their orientation preference and show a
oxidation. Furthermore, at high temperature, hydrogen could help to further clean poly(methyl methacrylate) (PMMA) residuals away from the graphene surface. After introducing H2 into our CVD system, we observed that graphene could be wellpreserved during the long-time CVD process (Figure 2c). Compared with the Raman spectrum of graphene without H2 protection (Figure 2e), the rise in the D peak decreased, indicating that graphene was less damaged in the presence of H2.28 Moreover, we also observed that the concentration of hydrogen was crucial to the formation of MoS2 domains. By applying different hydrogen concentrations, we found that graphene could be well-protected and act as a growth template for MoS2 with the presence of a trace amount of hydrogen gas, whereas with a relatively higher hydrogen concentration, no MoS2 domains formed on the surface of graphene (Figure 2c,d). This is due to the fact that besides being a reducing agent, hydrogen also has an etching effect on MoS2, especially at high temperature with high concentration.29 The optimal carrying gas flow we discovered was a mixture of 7.5 sccm Ar and 2.5 sccm 25% H2/75% Ar. With a trace amount of hydrogen introduced into the CVD system, the CVD growth time could be extended longer than 40 min. However, we noticed that the MoS2 film was not able to grow directly onto the graphene substrate by simply increasing the growth time with the previous CVD process because of oxidization (Figure 2a). Even with an extended reaction time D
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 4. Raman and photoluminescence (PL) studies on MoS2 on Si/SiO2 and graphene substrate: (a) Raman spectra of MoS2 domains grown on SiO2, MoS2 domains transferred on graphene, MoS2 domains grown on graphene, and MoS2 films grown on graphene. The dashed black line indicates the shift of E12g and A1g peaks between the spectra; (b) PL spectra of MoS2 domains grown on SiO2, MoS2 domains transferred on graphene, MoS2 domains grown on graphene, and MoS2 films grown on graphene; and (c−f) fits to the PL peak of MoS2 with subpeaks corresponding to the B excitonic transition (B), neutral excitonic transition A (A), and charged excitonic transition of A (A-).
modes.31 The E12g mode is related to the in-plane vibration of Mo and S atoms, whereas the A1g mode refers to the out-of-plane vibration of S atoms.32 To determine the layer number and lattice strain of our continuous MoS2 film grown on graphene, we compared the Raman spectrum with the monolayer MoS2 domains grown on graphene, monolayer MoS2 domains grown on SiO2/Si using a similar CVD method, and transferred monolayer MoS2 domains from SiO2/Si to graphene. The characterization of CVD-grown MoS2 domains is presented in Supporting Information S2. The Raman spectra are presented in Figure 4a, with an excitation wavelength of 532 nm. The frequency difference between E12g and A1g peaks could help to determine the layer number of MoS2.32 The fitting result shows that for the MoS2 film grown on graphene, the center of E12g and A1g peaks located at ∼382.4 and 407.0 cm−1 respectively, whereas the two peaks of the monolayer MoS2 domains grown on graphene centered at ∼382.1 and 406.5 cm−1. They both have a similar frequency difference ∼24.5 cm−1, which proves that the continuous MoS2 film produced on graphene was a monolayer.
nonepitaxial growth behavior. These phenomena suggest that the evolution of continuous MoS2 films consists of several steps. First, epitaxial MoS2 domains formed on the surface of the graphene substrate. As the size of these domains increases to their size limit (∼5 μm), the rising lattice mismatch between graphene and MoS2 disfavors further growth of epitaxial MoS2 domains. As a result, secondary MoS2 domains nucleate on the edge of epitaxial MoS2 domains and start a nonepitaxial growth. The secondary domains grow until their edges merge and form the continuous film. Further MoS2 deposited on the MoS2 film leads to the nucleation of the 3D MoS2 particles (Figure 3j). On the basis of the best quality MoS2 film we obtained via controlling the parameter (Figure 3e), we then planned to study the quality of the as-synthesized MoS2 film. Raman and PL spectroscopies are two very useful techniques to determine the layer number as well as to investigate other effects such as lattice strain, doping levels, and interfacial van der Waals interactions between different 2D materials.30 In the Raman spectrum of MoS2, E12g and A1g peaks represent two characteristic vibration E
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 5. PL and Raman measurements of the as-grown MoS2 film on graphene. (a) Raman spectrum of the as-prepared sample; (b,c) Raman maps (200 μm × 200 μm) of the continuous MoS2 film grown on graphene, showing the (b) E12g peak position; (c) A1g peak position; (d) PL spectrum of the as-prepared sample; and (e,f) PL maps of the continuous MoS2 film grown on graphene showing (e) A- peak position and (f) A peak position. (g,h) Average peak position of various peaks in Raman and PL measurements.
from the monolayer MoS2 could efficiently transfer to graphene and combine rapidly through a nonradioactive recombination, leading to the PL quench.34 Furthermore, we performed the peak fitting on the PL spectra based on the Lorentz peak function to show the relative luminescence quantum efficiency between the samples. As depicted in Figure 4c−f, the PL spectra of MoS2 have three subpeaks which correspond to A, A-, and B peaks according to previous research studies.35 The peaks A and B correspond to the direct transition from the conduction band to the two spin− orbit split valence band at the K point. The peak A-, located at a slightly lower energy than the peak A, is associated with the recombination of negatively charged excitons of trions A-. The trion A- arises from a free electron bounded to a neutral exciton via Coulomb interaction as MoS2 could be easily negatively doped by contamination or contact with a substrate.36 By comparing the integrated A and A- PL peaks of all four samples, we noticed that the PL intensity ratio of A to A- (A/A-) was different. The A/A- ratio indicates the relative populations of neutral and charged A excitons. In the PL spectra of MoS2 grown on SiO2, the PL peak is dominated by the A- peak, revealing a high degree of charged impurities from the SiO2 substrate, in agreement with previous studies.37 On the contrary, for the other three samples of MoS2 on the graphene substrate, the A/A- ratio was much higher. This indicates that graphene could conduct away charged impurities and act as a P-doping agent for MoS2. A higher A/A- ratio indicates a higher P-doping level as well as a stronger interfacial interaction between MoS2 and graphene. It was found that among the three MoS2/graphene samples, MoS2
However, it should also be noticed that the frequency difference of the monolayer MoS2 transferred onto graphene was relatively smaller. The E12g peak showed a significant blue shift to ∼385 cm−1, while the position of the peak remained at ∼406.5 cm−1. We speculate that this phenomenon results from the lattice strain of MoS2, as previous studies reported that the position of the E12g peak could be easily affected by the built-in strain of MoS2, while the A1g peak is less influenced.33 The lattice mismatch between the MoS2 domains and the graphene of the CVD-grown MoS2 introduced a lattice strain in the MoS2 domains and film. On the other hand, the transfer process released a potential strain in the CVD-grown MoS2 domains. This conclusion is also supported by the blue shift of the E12g peak of the MoS2 film grown on graphene compared to the epitaxial MoS2 domains grown on graphene. In previous discussion on the growth mechanism of the MoS2 film on graphene, we discovered that initially MoS2 domains were formed on graphene in an epitaxial pattern, followed by a nonepitaxial growth. The nonepitaxial growth reduced a part of the lattice strain coming from the lattice mismatch, resulting in the blue shift of the E12g mode. We also studied the PL of monolayer MoS2 under different conditions, as depicted in Figure 4b. All PL spectra were normalized against the Raman intensity. For the PL spectra of MoS2 either grown or transferred on the graphene substrate, the PL peak intensity was dramatically quenched compared to that on the SiO2 substrate. The quench indicates that MoS2 displays an effective interlayer coupling and electronic interaction with the graphene substrate. The photoexcited electrons and holes F
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
Figure 6. Photodetector fabricated by all CVD-grown MoS2 film/graphene heterostructure: (a) schematic illustration of the side view and top view of photodetector devices fabricated by the CVD-grown MoS2/graphene film; (b) optic image of the as-fabricated devices. The photo image of the device chip is presented in Figure S3; (c) band diagram explaining the photoelectron transfer; (d) I−V curve of the as-fabricated device under various illuminations; (e) magnitude of photocurrent of the as-fabricated device under various illuminations; (f) photoresponsivity of the as-fabricated device under various illuminations as a function of bias voltage; and (g) statistic distribution of photoresponsivity of the as-fabricated devices under 1 V bias with an incident light intensity of 1165 μW.
domains directly grown on graphene had the highest A/A- ratio, followed by the directly grown MoS2 film on graphene and transferred MoS2. This sequence matches with the Raman study of the built-in strain in MoS2. Moreover, a red shift of A peak could be observed alongside with the decrease of the A/A- ratio. Other research studies suggest that the uniaxial tensile strain could not only influence the Raman spectrum but also reduce the band gap of MoS2 and therefore cause the red shift of A exciton peak, which further supports the observation.38
Besides the change in the A and A- peaks, the change in the B peak is also worth mentioning. Compared to the B peak in MoS2 domains grown on graphene, the B/A ratio of MoS2 domains transferred on graphene and MoS2 film grown on graphene was much higher. In other words, the PL quench is less effective in the B peaks of these two samples. As is mentioned before, the quench results from the nonradioactive recombination of photoexcited electrons and holes. Therefore, we speculate that the increased B/A ratio is due to a weaker coupling between the MoS2 and the graphene layer. In the case of transferred MoS2 on graphene, G
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
A/W with 3.3 × 10−4 A photocurrent under 135 μW illumination at +2 V bias. Photoresponsivity could also be expressed by a function of photogain as
decoupling derives from interlayer impurities introduced during the transfer process, whereas for the CVD-grown MoS2 film on graphene, the oxidation of the graphene underlayer increases the number of defects and surface roughness of graphene, leading to decoupling. However, it is still not clear why the A peak seems to be less influenced than the B peak. As there are very few studies on the factors that may cause the evolution of B peak intensity in the monolayer MoS2, further studies would be worthwhile in the future. In addition, the PL and Raman spectra could help to evaluate the areal uniformity of the as-synthesized continuous MoS2 film on graphene via PL and Raman mappings. Figure 5c,d presents the Raman integrated mappings corresponding to E12g and A1g peak positions, obtained on a typical region of 200 μm × 200 μm. The average position of the E12g peak was 381.66 cm−1, with a standard deviation of 0.12 cm−1. The average position of A1g peak was 406.38 cm−1, with a standard deviation of 0.07 cm−1. Figure 5e,f shows the PL mapping of the peak positions of A and A- on the same region. The average position of the A- peak was 1.825 eV, with a standard deviation of 0.044 eV. The average position of the A peak was 1.867 eV, with a standard deviation of 0.0054 eV. Both mappings show a good uniformity among the measured area, and the peak positions match with the previous discussion, proving a high quality and uniformity of the MoS2/graphene film produced. To characterize the optoelectronic properties of our all CVDgrown MoS2 film/graphene heterostructure, we fabricated a simple photodetector. Figure 6a,b is the schematic illustration and optical image of the photodetector device we fabricated using the CVD-grown MoS2/graphene film. To fabricate the device, the MoS2/graphene film was directly transferred onto a SiO2 (300 nm)/p++ Si substrate (University Wafer) with predeposited gold pads, which used the method reported in our previous studies.39 The gold pads directly contacted with the graphene film, acting as electrodes. To study the photoresponse of our device, a source−drain voltage was applied on the gold electrode pair, and I−V measurements were carried on the device under dark and four different illumination circumstances. Figure 6c shows the plots of photocurrent versus applied source−drain voltage. When the laser was off, the I−V plot shows a linear curve, indicative of good contact between the gold electrodes and the MoS2/graphene film. By emitting laser on our photodetector, we observed that the response current was decreased. This phenomenon can be explained by a photoinduced electron transfer from MoS2 to graphene. Previous studies have proved that the graphene film transferred by our method is p-doped.40 When no gate voltage is applied, the drain current in the graphene field-effect transistor (FET) is dominated by holes. When the FET is illuminated with a visible light, photoinduced electrons inject from MoS2 to graphene, which reduces the current41 (Figure 6d,e). Photoresponsivity (R) and photogain (G) are two important figure of merits to evaluate the performance of photodetectors.42,43 Photoresponsivity is a measure of the electrical response of the device and can be defined as R =
Iphoto Plaser
R = ηextGe/hv
where ηext is the external quantum efficiency, e is the elementary charge unit, and v = c/λ is the frequency of the incident light. We assume that ηext = 8.1%, as the absorbance of our MoS2/graphene film is ∼8.1% at 532 nm, measured by absorption spectroscopy, and we assume that the photogain of our device could reach up to 69 under Plaser = 135 μW and Vbias = 2 V. We measured the time-resolved photoresponse of a MoS2/ graphene device at a fixed bias of 1 V with an incident light intensity of 1165 μW, shown in Figure S4. Reproducible highand low-impedance states could be obtained by repeated cycles of on−off states of light irradiation. Also, to evaluate the uniformity of the performance of our devices, we measure 50 devices within an ∼8 mm2 effective MoS2/graphene film area we transferred; among them 48 of them have a sufficient MoS2/ graphene coverage and clean surface to make working devices. The successful rate was 96%. The statistic distribution of the photoresponsivity of our devices is presented in Figure 6g. The average photoresponsivity is about 1.18 A/W, and the standard deviation is 0.51. The result shows an acceptable uniformity in the performance of the as-fabricated devices. It should be made aware that the scattering of the photoresponsivity data is relatively higher than the achieved uniformity. The prime cause of the discrepancy is due to the processing and measurement of our devices. Because these devices were fabricated by a polymerassisted transfer method, contaminations may inevitably be introduced into the system. Besides, during the measurements, the probes need to pierce the MoS2/graphene film to contact with the gold electrodes; this leads to possible contamination between the probe tip and the gold electrode, resulting in high deviation. These photodetectors with a direct CVD-grown MoS2/ graphene film exhibit a high photoresponsivity and photogain in the absence of applied gate potential under an ambient environment, compared to other studies with a similar design of the device. Lin et al.44 demonstrated a responsivity of 79 mA/W when Plight = 40 μW and bias = 1 V on a directly grown MoS2/ graphene/SiC heterostructure. Liu et al.45 observed that on a printable MoS2/graphene heterostructure, a photoresponsivity of 835 mA/W could be achieved under Plight of 0.05 mW/cm2 and bias of 5 V. Our device represents at least 1 order improvement of photoresponsivity. We also noticed that the photoresponsivity of the MoS2/graphene heterostructure could reach up to 107 A/W with a comb-shape drain−source pair.19 This reminds us that the performance of our directly CVD-grown MoS2/graphene heterostructure is likely to be further improved with a modified device design, which is worthy of further studies in future.
■
CONCLUSIONS In summary, we have demonstrated a simple approach to directly synthesize a large area of continuous monolayer MoS2 films on a graphene substrate by a two-step CVD process with the aid of hydrogen under an ambient pressure. We discovered that one important factor that hinders the formation of the MoS2 film on graphene is the oxidation of graphene during the long-time CVD process. Graphene can be effectively protected from being oxidized by adding hydrogen into the carrying gas. A centimeterscale continuous MoS2 film was grown on the graphene substrate
, where
Iphoto is the photocurrent and Plaser is the collected illumination power. In our case, the photoinduced current is negative; therefore, we define the current change as the photocurrent Iphoto. In Figure 6f, we present the calculated R of the device as a function of Vds under different illumination intensities. Under all incident power density, R showed an increasing trend with a larger applied bias. The largest R measured could reach over 2.4 H
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces
sccm Ar + 2.5 sccm 25% H2/75% Ar and kept for 30 min for the slow growth of the MoS2 domains. Then a fast cooling process was applied to stop the reaction by removing the sample from the hot zone of the furnace. Device Fabrication. A JEOL 5500 FS electron beam lithography system was used to pattern metal pad windows into a bilayer PMMA resist. A thermal evaporator was used to deposit Cr/Au (10 nm/80 nm) metal electrode contacts onto a 300 nm SiO2/Si substrate, followed by lift-off in hot acetone. A PMMA/MoS2/graphene film was transferred onto the Si chip with prepatterned metal contact pads and baked overnight at 180 °C to improve contact. Last, PMMA was removed with the aid of acetone. Optoelectronic Characterization of Devices. A LabRam Aramis Raman Spectrometer was used to carry out Raman and PL measurements. The samples were illuminated with a 532 nm laser of 200 μW, through a ×50 objective lens with a spot size of ∼2 μm. A Keithley 2400 source meter was used for I−V and transient photocurrent characterizations of the MoS2/graphene photodetectors. The illumination beam was formed by a 532 nm diode-pumped solidstate laser (Thorlabs, DJ532-40) coupled into a confocal microscope with a spot size of ∼150 μm2 as a light source. The power values of the output laser were measured using an optical power meter (Thorlabs Optics PM100D, ±3% accuracy) placed under illumination before each I−V measurement. All measurements were carried out in air and at room temperature.
with an appropriate nucleation rate, growth time, and hydrogen concentration. The film microstructure and thickness were characterized by PL and Raman measurements, suggesting that the MoS2 film is a monolayer. We also fabricated a simple photodetector based on the MoS2/graphene film, exhibiting high photocurrent, photoresponse reaching up to 2.4 A/W, and photogain reaching up to 69. This approach shows a potential for scalable fabrication of high-performance optoelectronics based on 2D van der Waals heterostructures.
■
EXPERIMENTAL METHODS
CVD Growth of Graphene. Graphene was prepared by a onefurnace CVD procedure on a liquid-phase copper in a horizontal quartz tube. Hydrogen and methane were used as reactants and argon as a carrying gas. This method previously reported to successfully synthesize the highly uniform ultrathin graphene film.46 The whole reaction was carried out within a horizontal furnace. The furnace comprises an inner hot zone in the middle where copper foils supported by tungsten foils were placed in a ceramic boat. The CVD growth was composed of two steps. After flushing by a mixture gas with a flux of argon and 100 sccm 25% hydrogen/75% argon for 30 min, the growth temperature in the middle hot zone was first set to 1090 °C at a ramp rate of 30 °C/s. When the temperature reached 1090 °C, copper melted into the liquid phase and methane was introduced into the system, initiating the nucleation of graphene domains. The nucleation process lasted for 90 min. Afterward, the temperature was reduced to 1060 °C at a ramp rate of 5 °C/s and kept for 30 min after reaching 1060 °C. During this process, graphene domains grew bigger and the edges of graphene domains eventually met up, forming a large area of graphene films. Because of the liquefaction and solidification of the copper substrate, the copper surface got rid of the textures introduced in the copper foil processing and as a result led to an improved uniformity of the graphene film on the surface. Graphene Transfer. Transfer of graphene was achieved by a wet transfer process with the aid of PMMA. The graphene/Cu/W heterostructure was first coated with PMMA to protect graphene during the following processes. The tungsten substrate was then electrochemically etched in a 2 M NaOH solution by using another piece of Cu foil as the anode.40 Afterward, the remaining PMMA/ graphene was separated from the copper substrate by floating on a 2 M ammonium persulfate ((NH4)2S2O8) solution overnight. Finally, the graphene film was transferred onto a SiO2 300 nm/p++ Si substrate, and the protective PMMA coating was removed by acetone after heating to 180 °C for 15 min. CVD Growth of MoS2 Films on Graphene. The growth was carried out in a CVD system using 20 mg of MoO3 powder (≥99.5%, Sigma Aldrich) and 600 mg of S (≥99.5%, Sigma-Aldrich) as the precursor using argon (Ar) as a carrier gas under atmospheric pressure. The S powder was loaded at the central area of furnace 1 in the outer 1 in. quartz tube, whereas MoO3 was placed separately in an inner quartz tube with a smaller diameter of 1 cm at the upstream of furnace 2. The distance from MoO3 to the center of furnace 2 was designed to be 10 cm. The substrate was vertically placed in the center of furnace 2. The growth system was first flushed with 500 sccm of Ar gas for 15 min, followed by a preintroduction of S vapor. The S powder was heated up to ∼180 °C for 15 min under an Ar flow rate of 150 sccm while furnace 2 was maintained at 200 °C at the same time to avoid any deposition of the solid S on the substrate surface. This ensured that the reaction occurs under an S-rich atmosphere, whereby the MoS2 nucleation density is effectively controlled by the flux of MoO3 determined by the carrier gas flow rate. Then, furnace 2 was heated up at a rate of 40 °C min−1 to 770 °C while the MoO3 powder reached an approximate maximum temperature of 400 °C. At the same time, the S powder temperature was programmed to increase from 180 to 200 °C at a rate of 1 °C min−1, which complements the consumption of the S powder to maintain an Srich growth environment. The reaction was first conducted under the temperature of 770 °C for 15 min with 150 sccm Ar to nucleate the MoS2 domains on graphene. After that, the gas flow was changed to 7.5
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b14860. Raman spectrum of MoS2 film/graphene and 3D crystals on the MoS2 film, characterization of the monolayer MoS2 domains on the graphene substrate, optic image of the asfabricated device chip, and transient photocurrent characteristics obtained from the device under +2 V bias and 1165 μW illumination (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Yuewen Sheng: 0000-0003-3067-9520 Jamie H. Warner: 0000-0002-1271-2019 Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Geim, A. K.; Novoselov, K. S. The Rise Of Graphene. Nat. Mater. 2007, 6, 183−191. (2) Wang, X.; Hooper, T. N.; Kumar, A.; Priest, I. K.; Sheng, Y.; Samuels, T. O. M.; Wang, S.; Robertson, A. W.; Pacios, M.; Bhaskaran, H.; Weller, A. S.; Warner, J. H. Oligomeric Aminoborane Precursors For The Chemical Vapour Deposition Growth Of Few-Layer Hexagonal Boron Nitride. CrystEngComm 2017, 19, 285−294. (3) Rong, Y.; Fan, Y.; Leen Koh, A.; Robertson, A. W.; He, K.; Wang, S.; Tan, H.; Sinclair, R.; Warner, J. H. Controlling Sulphur Precursor Addition For Large Single Crystal Domains Of WS2. Nanoscale 2014, 6, 12096−12103. (4) Ramakrishna Matte, H. S. S.; Gomathi, A.; Manna, A. K.; Late, D. J.; Datta, R.; Pati, S. K.; Rao, C. N. R. MoS2 And WS2 Analogues Of Graphene. Angew. Chem. 2010, 122, 4153−4156. (5) Huang, J.-K.; Pu, J.; Hsu, C.-L.; Chiu, M.-H.; Juang, Z.-Y.; Chang, Y.-H.; Chang, W.-H.; Iwasa, Y.; Takenobu, T.; Li, L.-J. Large-Area Synthesis Of Highly Crystalline WSe2 monolayers And Device Applications. ACS Nano 2014, 8, 923−930. I
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces (6) Wang, X.; Gong, Y.; Shi, G.; Chow, W. L.; Keyshar, K.; Ye, G.; Vajtai, R.; Lou, J.; Liu, Z.; Ringe, E.; Tay, B. K.; Ajayan, P. M. Chemical Vapor Deposition Growth of Crystalline Monolayer MoSe2. ACS Nano 2014, 8, 5125−5131. (7) Mas-Ballesté, R.; Gómez-Navarro, C.; Gómez-Herrero, J.; Zamora, F. 2D Materials: To Graphene and Beyond. Nanoscale 2011, 3, 20−30. (8) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Electronics and Optoelectronics of Two-Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol. 2012, 7, 699−712. (9) Li, H.; Yin, Z.; He, Q.; Li, H.; Huang, X.; Lu, G.; Fam, D. W. H.; Tok, A. l. Y.; Zhang, Q.; Zhang, H. Fabrication of Single- And Multilayer MoS2 Film-Based Field-Effect Transistors For Sensing NO At Room Temperature. Small 2011, 8, 63−67. (10) Amani, M.; Chin, M. L.; Birdwell, A. G.; O’Regan, T. P.; Najmaei, S.; Liu, Z.; Ajayan, P. M.; Lou, J.; Dubey, M. Electrical Performance Of Monolayer MoS2 Field-Effect Transistors Prepared By Chemical Vapor Deposition. Appl. Phys. Lett. 2013, 102, 193107. (11) Lopez-Sanchez, O.; Lembke, D.; Kayci, M.; Radenovic, A.; Kis, A. Ultrasensitive Photodetectors Based On Monolayer MoS2. Nat. Nanotechnol. 2013, 8, 497−501. (12) Geim, A. K.; Grigorieva, I. V. van der Waals Heterostructures. Nature 2013, 499, 419−425. (13) Sheng, Y.; Xu, W.; Wang, X.; He, Z.; Rong, Y.; Warner, J. H. Mixed Multilayered Vertical Heterostructures Utilizing Strained Monolayer WS2. Nanoscale 2016, 8, 2639−2647. (14) Tongay, S.; Fan, W.; Kang, J.; Park, J.; Koldemir, U.; Suh, J.; Narang, D. S.; Liu, K.; Ji, J.; Li, J.; Sinclair, R.; Wu, J. Tuning Interlayer Coupling In Large-Area Heterostructures With CVD-Grown MoS2 and WS2 monolayers. Nano Lett. 2014, 14, 3185−3190. (15) Tsai, M.-L.; Su, S.-H.; Chang, J.-K.; Tsai, D.-S.; Chen, C.-H.; Wu, C.-I.; Li, L.-J.; Chen, L.-J.; He, J.-H. Monolayer MoS2 Heterojunction Solar Cells. ACS Nano 2014, 8, 8317−8322. (16) Georgiou, T.; Jalil, R.; Belle, B. D.; Britnell, L.; Gorbachev, R. V.; Morozov, S. V.; Kim, Y.-J.; Gholinia, A.; Haigh, S. J.; Makarovsky, O.; Eaves, L.; Ponomarenko, L. A. Vertical Field-Effect Transistor Based On Graphene−WS2 Heterostructures For Flexible And Transparent Electronics. Nat. Nanotechnol. 2012, 8, 100−103. (17) Bertolazzi, S.; Krasnozhon, D.; Kis, A. Nonvolatile Memory Cells Based On MoS2/Graphene Heterostructures. ACS Nano 2013, 7, 3246−3252. (18) 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.; Taneguchi, T.; Nuckolls, C.; Kim, P.; Hone, J. Flexible And Transparent MoS2 field-Effect Transistors On Hexagonal Boron Nitride-Graphene Heterostructures. ACS Nano 2013, 7, 7931−7936. (19) Zhang, W.; Chuu, C.-P.; Huang, J.-K.; Chen, C.-H.; Tsai, M.-L.; Chang, Y.-H.; Liang, C.-T.; Chen, Y.-Z.; Chueh, Y.-L.; He, J.-H.; Chou, M.-Y.; Li, L.-J. Ultrahigh-Gain Photodetectors Based On Atomically Thin Graphene- MoS2 Heterostructures. Sci. Rep. 2014, 4, 3826. (20) Roy, K.; Padmanabhan, M.; Goswami, S.; Sai, T. P.; Ramalingam, G.; Raghavan, S.; Ghosh, A. Graphene−MoS2 Hybrid Structures For Multifunctional Photoresponsive Memory Devices. Nat. Nanotechnol. 2013, 8, 826−830. (21) Ling, Z. P.; Yang, R.; Chai, J. W.; Wang, S. J.; Leong, W. S.; Tong, Y.; Lei, D.; Zhou, Q.; Gong, X.; Chi, D. Z.; Ang, K.-W. Large-Scale TwoDimensional MoS2 Photodetectors By Magnetron Sputtering. Opt. Express 2015, 23, 13580. (22) Schmidt, H.; Wang, S.; Chu, L.; Toh, M.; Kumar, R.; Zhao, W.; Neto, A. H. C.; Martin, J.; Adam, S.; Ö zyilmaz, B.; Eda, G. Transport Properties of Monolayer MoS2 Grown by Chemical Vapor Deposition. Nano Lett. 2014, 14, 1909−1913. (23) Shi, Y.; Zhou, W.; Lu, A.-Y.; Fang, W.; Lee, Y.-H.; Hsu, A. L.; Kim, S. M.; Kim, K. K.; Yang, H. Y.; Li, L.-J.; Idrobo, J.-C.; Kong, J. van der Waals Epitaxy Of MoS2 layers Using Graphene As Growth Templates. Nano Lett. 2012, 12, 2784−2791. (24) 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.
(25) Chen, C.; Feng, Z.; Feng, Y.; Yue, Y.; Qin, C.; Zhang, D.; Feng, W. Large-Scale Synthesis Of A Uniform Film Of Bilayer MoS2 on Graphene For 2D Heterostructure Phototransistors. ACS Appl. Mater. Interfaces 2016, 8, 19004−19011. (26) Wang, S.; Pacios, M.; Bhaskaran, H.; Warner, J. H. Substrate Control For Large Area Continuous Films Of Monolayer MoS2 by Atmospheric Pressure Chemical Vapor Deposition. Nanotechnology 2016, 27, 085604. (27) Feng, H.; Wang, X.; Wu, D. Fabrication Of Spirocyclic Phosphazene Epoxy-Based Nanocomposites With Graphene Via Exfoliation Of Graphite Platelets And Thermal Curing For Enhancement Of Mechanical And Conductive Properties. Ind. Eng. Chem. Res. 2013, 52, 10160−10171. (28) Park, C.-S.; Zhao, Y.; Kim, H.; Shon, Y.; Kim, E. K. Resistivity Peaks And Magnetic Properties Of An Annealed Graphene. Chem. Commun. 2014, 50, 12930−12932. (29) Sheng, Y.; Tan, H.; Wang, X.; Warner, J. H. Hydrogen Addition For Centimeter-Sized Monolayer Tungsten Disulfide Continuous Films By Ambient Pressure Chemical Vapor Deposition. Chem. Mater. 2017, 29, 4904−4911. (30) Splendiani, A.; Sun, L.; Zhang, Y.; Li, T.; Kim, J.; Chim, C.-Y.; Galli, G.; Wang, F. Emerging Photoluminescence In Monolayer MoS2. Nano Lett. 2010, 10, 1271−1275. (31) Yang, M.; Cheng, X.; Li, Y.; Ren, Y.; Liu, M.; Qi, Z. Anharmonicity Of Monolayer Mos2, Mose2, And Wse2: A Raman Study Under High Pressure And Elevated Temperature. Appl. Phys. Lett. 2017, 110, 093108. (32) Buscema, M.; Steele, G. A.; van der Zant, H. S. J.; CastellanosGomez, A. The Effect Of The Substrate On The Raman And Photoluminescence Emission Of Single-Layer MoS2. Nano Res. 2014, 7, 561−571. (33) Yang, L.; Cui, X.; Zhang, J.; Wang, K.; Shen, M.; Zeng, S.; Dayeh, S. A.; Feng, L.; Xiang, B. Lattice Strain Effects On The Optical Properties Of Mos2 Nanosheets. Sci. Rep. 2014, 4, 5649. (34) Wan, W.; Li, X.; Li, X.; Xu, B.; Zhan, L.; Zhao, Z.; Zhang, P.; Wu, S. Q.; Zhu, Z.-z.; Huang, H.; Zhou, Y.; Cai, W. Interlayer Coupling Of A Direct van der Waals Epitaxial MoS2/Graphene Heterostructure. RSC Adv. 2016, 6, 323−330. (35) Mak, K. F.; He, K.; Lee, C.; Lee, G. H.; Hone, J.; Heinz, T. F.; Shan, J. Tightly Bound Trions In Monolayer MoS2. Nat. Mater. 2012, 12, 207−211. (36) Tongay, S.; Suh, J.; Ataca, C.; Fan, W.; Luce, A.; Kang, J. S.; Liu, J.; Ko, C.; Raghunathanan, R.; Zhou, J.; Ogletree, F.; Li, J.; Grossman, J. C.; Wu, J. Defects Activated Photoluminescence In Two-Dimensional Semiconductors: Interplay Between Bound, Charged And Free Excitons. Sci. Rep. 2013, 3, 2657. (37) Wang, S.; Wang, X.; Warner, J. H. All Chemical Vapor Deposition Growth Of Mos2:H-BN Vertical van der Waals Heterostructures. ACS Nano 2015, 9, 5246−5254. (38) Conley, H. J.; Wang, B.; Ziegler, J. I.; Haglund, R. F.; Pantelides, S. T.; Bolotin, K. I. Bandgap Engineering Of Strained Monolayer And Bilayer MoS2. Nano Lett. 2013, 13, 3626−3630. (39) Tan, H.; Fan, Y.; Zhou, Y.; Chen, Q.; Xu, W.; Warner, J. H. Ultrathin 2D Photodetectors Utilizing Chemical Vapor Deposition Grown WS2 with Graphene Electrodes. ACS Nano 2016, 10, 7866− 7873. (40) Tan, H.; Fan, Y.; Rong, Y.; Porter, B.; Lau, C. S.; Zhou, Y.; He, Z.; Wang, S.; Bhaskaran, H.; Warner, J. H. Doping Graphene Transistors Using Vertical Stacked Monolayer WS2 heterostructures Grown By Chemical Vapor Deposition. ACS Appl. Mater. Interfaces 2016, 8, 1644− 1652. (41) Yunus, R. M.; Endo, H.; Tsuji, M.; Ago, H. Vertical Heterostructures Of MoS2 and Graphene Nanoribbons Grown By Two-Step Chemical Vapor Deposition For High-Gain Photodetectors. Phys. Chem. Chem. Phys. 2015, 17, 25210−25215. (42) Kim, C. O.; Kim, S.; Shin, D. H.; Kang, S. S.; Kim, J. M.; Jang, C. W.; Joo, S. S.; Lee, J. S.; Kim, J. H.; Choi, S.-H.; Hwang, E. High Photoresponsivity In An All-Graphene P−N Vertical Junction Photodetector. Nat. Commun. 2014, 5, 3249. J
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces (43) Tsai, D.-S.; Liu, K.-K.; Lien, D.-H.; Tsai, M.-L.; Kang, C.-F.; Lin, C.-A.; Li, L.-J.; He, J.-H. Few-Layer MoS2 with High Broadband Photogain And Fast Optical Switching For Use In Harsh Environments. ACS Nano 2013, 7, 3905−3911. (44) Lin, Y.-C.; Lu, N.; Perea-Lopez, N.; Li, J.; Lin, Z.; Peng, X.; Lee, C.-H.; Sun, C.; Calderin, L.; Browning, P. N.; Bresnehan, M. S.; Kim, M. J.; Mayer, T. S.; Terrones, M.; Robinson, J. A. Direct Synthesis Of van der Waals Solids. ACS Nano 2014, 8, 3715−3723. (45) Liu, Q.; Cook, B.; Gong, M.; Gong, Y.; Ewing, D.; Casper, M.; Stramel, A.; Wu, J. Printable Transfer-Free And Wafer-Size MoS2/ Graphene van der Waals Heterostructures For High-Performance Photodetection. ACS Appl. Mater. Interfaces 2017, 9, 12728−12733. (46) Fan, Y.; He, K.; Tan, H.; Speller, S.; Warner, J. H. Crack-Free Growth And Transfer Of Continuous Monolayer Graphene Grown On Melted Copper. Chem. Mater. 2014, 26, 4984−4991.
K
DOI: 10.1021/acsami.7b14860 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX