Subscriber access provided by NEW YORK MED COLL
Article
The Essential Role of Cu Vapor for the Selflimit Graphene via Cu-catalytic CVD Method Hung-Chiao Lin, Yu-Ze Chen, Yi-Chung Wang, and Yu-Lun Chueh J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/jp511251n • Publication Date (Web): 21 Feb 2015 Downloaded from http://pubs.acs.org on February 27, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
The Essential Role of Cu Vapor for the Self-limit Graphene via Cu-catalytic CVD Method Hung-Chiao Lin, Yu-Ze Chen, Yi-Chung Wang, and Yu-Lun Chueh* Department of Material Science and Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan. E-mail:
[email protected]
ABSTRACT- Due to the inconsistent observations, the Cu catalytic decomposition of methane for graphene synthesis are reexamined, i.e., via the surface absorption, decomposition to atomic carbon and segregation. Here, we experimentally show the quantity of ambient Cu vapor is the key factor in graphene synthesis, which influences the dropwise condensations for airborne Cu clusters during growth. The massive carburization in Cu clusters and the calculation of carbon solubility in nano-sized clusters are performed, experimented, and further examined from the growth of diamond-like-carbon films and ball-like diamonds via Cu vapor assisted growth on SiO2. The affinitive interactions between Cu vapor, ambient gases, and solid surface are embodied. By combining the molecular dynamics for the redeposited Cu clusters to surface, the vehicle theory of Cu clusters, which transports the atomic carbon to the surface and completes the graphene growth, is thus proposed as the essential puzzle we considered.
KEYWORDS: Graphene, Cu catalytic CVD, self-limit, dehydrogenation, hydrocarbon radical coupling, Cu sublimation, dropwise condensation, Ostwald ripening process, size effect, carburization
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Introduction The synthesis of graphene by chemical vapor deposition (CVD) technique draws lots of attentions for scientific and industrial usages. Based on the knowledges, known more than 40 years1,2, of catalytic decomposition of hydrocarbons on transition metals, such as Pt3, Ni4, Ru5, Rh6, Ir7, and Cu8, this method is well accomplished by capturing the hydrocarbon precursors via the surface chemical absorption and the following segregation, referring to the supersaturation or the descending carbon solubility of catalyst at lower temperature, will do the growth. Among these catalysts, the self-limit property is found in Ru5, Ir9 and Cu1,8,10–12 attributed to their low carbon solubility13,14 so that few-layer graphene growth is achieved. The ‘bottom-up adlayer’ is further observed in Ru5 and Ir7 which the adlayer could be segregated by lowering temperature in synthesis. Distinctly, similar ‘bottom-up’ configuration is found both at atmospheric (AP) and low pressure (LP) in Cu10,12,15–17, whereas the bottom layer is shaped with the top layer at the same time and its size cannot get bigger17 (or is limited10,12) if the top sheet is big enough. Moreover, via real-time examinations, the segregation of carbon adatoms are found not to be programed by lowering temperature11, and the adatoms show no bulk diffusion during growth18. The extremely low carbon solubility in Cu (around two orders less than Ir at 1000 oC13,14) could interpret so far but also elicits an issue. The ‘dehydrogenation to atomic carbon’ is distinctly unfavorable, observed from the pyrogenation of full methyl19–22 and methylene19 covered Cu, the inevitable intermedia in the decomposition of methane23. Higher species of hydrocarbon gases are formed rather than dehydrogenation process. So, where does the atomic carbon come from for graphene synthesis? Furthermore, in the foregone transport theory of CVD on graphene growth24, the increased gas pressure will intensify the physisorption and sticky probability of gaseous feedstock to catalyst surface, so that the increment of adatom will prosper the catalytic reactions. On the contrary, higher activation energy is required in APCVD than in LPCVD25 attributed to the desorption of carbon atoms, and its ‘symbolized’ hexagonal shape of graphene grain grown in APCVD26 is now sure as the representation of slower growth rate and less growth environment10. We believe there is a crucial mechanism barricaded from the ‘presentation’ and transports the carbon adatoms to catalyst surface during growth, the redeposition of Cu vapor. In this article, we manipulate four different growths to distinguish the importance of ambient Cu vapor in synthesis from beginning. The tremendous increment of carbon solubility in Cu vapor is calculated and we design an experiment to emerge the increasing efficacy of Cu vapor by increasing the Ostwald ripening process during growth. The carburization of Cu vapor, interactions of vapor to catalyst surface and ACS Paragon Plus Environment
Page 2 of 23
Page 3 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
the transports of carbon feedstock are manipulated and discussed. We finally amend the growth mechanism with the involved redeposition of Cu vapor, the essential puzzle we considered. Experimental Section Chamber preparation All 3-inch-wide quartz tubes are cleaned by the standard clean procedures in the manufactory before they are shipped. Since we purchased, the brand new tubes are wiped by non-woven fabric with acetone, isopropyl alcohol (IPA) and deionized water respectively. After the quartz tube is mounted at our furnace system which is equipped with additional Brewster pump, gate valve (capable to stabilize the modulated pressure in 3 sec), R type thermal coupler, proportional-integral-derivative (PID) controlled 1200 oC heating module, multiple mass flow controllers (MFCs), and vacuum gauge, the chamber is pumped till the pressure is below 1.0×10-3 torr at room temperature and pre-baked at 1000 oC for 1hr with 50 sccm H2 constant flow under 4×10-3 torr. This baked chamber is then qualified as the ‘brand-new chamber’ we used in this article. Each brand new tube will be discarded after we finish the experiments of each single topic to avoid the extradecorated Cu on the inner surface (See the ‘Chamber Remark’ column in Table S1) and each corresponding 8-cm-long quartz holder (configuration in Fig. S11) is deployed at the chamber center and processed with above mentioned procedure as well. The ‘Cu-decorated’ chambers are made from a brand-new chamber by deploying a new 4×4 cm2 Cu foil (25-µm-thick, 99.8%, Alfa Aestar) on a corresponding quartz holder at the chamber center and heated at 1020 oC in 1 hr with 50 sccm H2 constant flow under 4×10-3 torr. After heating, the Cu foil is discarded and we re-bake the chamber again under the same conditions without Cu foil (1020 o
C in 1 hr with 50 sccm H2 under 4×10-3 torr) to smear the distribution of the ‘decorated’ Cu on the inner
surface of the chamber. Sample preparation, graphene growth and experiments Based on the complexity of the experiments for the elusive properties of Cu vapor in this article, we organize the heating sequence as Fig. S1, and the scheme of experiments as Table. S1 including the conditions of inputing gases, pressure, heating temperature, duration, substrate type, chamber type, and the informations of their corresponding usage and adoption in figures. The standard size of foils in this article, both Cu and Ni (25-µm-thick, 99.95%, Alfa Aestar), is 1×1 cm2.
Results and Discussion
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
In our experience of graphene synthesis via Cu-catalytic CVD method, after 3–10 runs at least (the number of operations is negatively correlated to the heating duration and the quantity of consumed Cu foils), we then have quality and higher yield graphene, since we equipped a ‘new’ 3-inch-wide quartz chamber in our CVD system. Dozens of new quartz chambers have been expended with meticulous operations to avoid any possible flaw and then this lesson is ensured. To dispel the mists from this elusive phenomenon, we first prepare the chambers carefully as two different conditions, the brand new and Cu-decorated (see Methods, heating sequence in Fig. S1, and Table S1 for all conditions to various experiments in this article in supplementary material), with annealed Cu foils (Fig. S2) for processes. By starting with a brand new chamber, we perform 30-min growth with one ‘annealed’ Cu foil per time (‘Pure CH4 growth 1’ in Table S1) and then 30-min annealing with one ‘brand new’ Cu foil (‘H2 0.5 hr Annealing 1’ in Table S1) with 3 repeat times to record the gradual changes on the surface of the processed foils (Fig. S3) and on the inner wall of the chamber (Fig. S4). Both the surface conditions of foils and graphene growth are found to be related to the running times (Fig. S3). Till the 3rd growth procedure, few shattered graphene spots are found, as shown in the image of scanning electron microscope (SEM) in Fig. 1a. Another brand new chamber is applied for the similar growth but with a ‘brand new’ Cu foil (providing more Cu evaporation than the annealed foil during heating, supported by Fig. S4) and an addition of 30-min pre-heating (annealing) procedure (‘Conventional growth’ in Table S1). Perceptibly, slightly more graphene grains are found at the 1st growth in Fig. 1b, and the result of the 3rd growth process shows massively increased graphene coverage in Fig. 1c and the conglobate Cu nano-particles are also found sticked at the commissural region of graphene sheets on the foil. Furthermore, in an another new ‘Cu-decorated’ chamber with 16 annealed foils per time, it is applied to repeat the growth in Fig. 1a (see Table S1, ‘Pure CH4 growth 2’) and the growth with lower CH4 flow rate and additional H2 flow (see Table S1, ’CH4 & H2 growth’). The resulted ‘full-graphene-covered’ Cu foils are shown as Fig. 1d and 1e with their corresponding Raman signals in Fig. 1f (Horiba Jobin Yvon HR800 with 514 nm wavelength 50 mW Ar+ laser equipped and 100× objective lens). From the above four growth demonstrations, obviously, we could realize, besides the understandings of the foregone understandings in the catalytic CVD, the quantity of ambient Cu vapor, sublimated from the foil and the condensed Cu on the inner wall of chamber, play an elusive role in graphene synthesis which is still unclarified. To perceive what happened therein, we would like to discuss the activities of Cu vapor during heating. Fig. 2a is the distribution of the equilibrium vapor pressure of Cu in our 3-inch-wide chamber ACS Paragon Plus Environment
Page 4 of 23
Page 5 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
heated at 1000 oC (converted from the probed temperature distribution27. See Fig. S5). For the high evaporation rate of Cu heated in vacuum25, the two situations in our chamber will cause the homogeneous nucleation of Cu vapor as the airborne droplets or clusters: (1) The supersaturated vapor quantity28 from the sublimation of the foil and condensed Cu on the inner chamber wall. (2) In aerodynamics, when the kinetic vapor at the higher temperature region move to the lower temperature region, drove by the gradient of vapor concentration, the ‘exceeded’ vapor due to the dropping equilibrium pressure will be coalesced for balancing the state29. The above two cases of the dropwise condensation are airborne and follow the Ostwald ripening process for minimizing the system energy, which the coalesced size is strongly influenced by the molecular weight (MW) of ambient gas and temperature gradient30. From the results in Fig. 1, to determine the different quantities of Cu vapor in the applied chambers, without loading any Cu foil, we utilize the Ni foils as the vapor getter (‘Ni getter experiment’ section in Table S1) to capture the evaporated Cu at LP and AP in a new Cu-decorated chamber and the chamber used in Fig. 1c, processed three runs in the ‘Conventional growth’. The appreciable Cu content in Ni getter is detected only in the Cu-decorated chamber by examinations of X-ray diffraction (XRD) and inductively coupled plasma-mass spectrometer (ICP-MS) (Fig. S6). Furthermore, to distinguish how the ambient Cu vapor from chamber helps the carburization for the foil in the growth, without loading any foil, we lower down the growth temperature to 950 oC and repeat the growth in Fig. 1d by deploying a scratched SiO2 chip at the center of a Cu-decorated chamber (‘Particle carburization’ in Table S1) to catch the carburized Cu clusters at the center zone in the chamber. As shown in Fig. 2b and 2c, serial conglobate Cu clusters are captured with diameter at ~1 µm. Trough the quantitative analysis from the equipped energy dispersive X-ray (EDX) in SEM system (Jeol, JEM 5410), the resulted massive carbon content to the fractional Cu ratio (Fig. S7) suggests us to investigate the suspected answer, the well-known ‘size effect’31. Due to the increasing surface-to-volume ratio in nano size, the subsequently enhanced surface energy reduces the melting temperature of the solid particle remarkably31. Noticeably, in the growth of carbon nanotube via the metal-catalytic CVD method, nano-sized catalyst is affirmed to efficiently capture the carbon feedstock from the hydrocarbon precursors and even the ‘nearly-carbon-insoluble’ transition metal, like Au32,33 and Cu34, are concluded as ‘soluble’32–34 in nano size. For this point of view, the method of computer calculation of phase diagrams (CALPHAD)35 is applied to anatomize the correlation between the size effect and the ability of carburization via the contributed surface free energy in the shrunk Cu particle ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
system. Series discussions and calculations are performed in Fig. S8–S10. The results indicate the melting temperatures of nano-sized Cu and graphite particles are intensely reduced when the size is smaller than ~20 nm (Fig. S9), and the Cu–C binary equilibrium phase diagram at Cu-rich side in bulk and nano-sized particle system, Fig. 2d, is calculated through the combinations of: (1) The most suggested diagram of Cu–C at Cu rich side14 (Fig. S10a and S10b). (2) The reinforced carbon solubility in the ‘solid’ particle (Fig. S10c, and S10d). (3) The readjustment of the critical values at phase transition by means of the fusion changes of the excess surface energy. At our standard synthesis temperature, 1020 oC, the carbon solubility in the 1000 nm ‘solid’ particle is calculated as 7.44823 at.ppm (0.65% more than bulk), solubility at 100 nm is 7.89671 at.ppm (6.71% more), at 50 nm is 8.42677 at.ppm (13.88% more) and at 30 nm is 9.18926 at.ppm (24.18% up). However, by considering the massive carbon content in Cu particles in Fig. S7, the qualitative analysis from this point of view for the ‘ripened size’ at 50–1000 nm cannot represent the reality which the carburization of Cu particle is continual during ripening from the outset in the embryonic stage. Hence, Fig. 2d also shows the results in the sizes at few nano meters with the tremendously descendent melting temperature and ascendent carbon solubility in both liquid and solid phase. At 1020 oC, notably, the carbon mole fraction in 3 nm Cu ‘droplet’ is 40.14747 at.ppm (453.15% more than bulk), and the solubility at 2 nm is 5318.60395 at.ppm (71773.03% more, as shown in the insert in Fig. 2d), which is now comparable to the solubility in the bulk Ni at 1020 oC (~0.59 at.%). That tremendous increment in the embryonic Cu droplets at few nano meters interprets that the superior capacity for capturing ‘atomic’ carbon is endued since the droplets are just born. Since the importances of the quantity of ambient Cu vapor during graphene growth and the tremendous carbon solubility in few nm Cu droplet are performed, how does the Cu particles affect the growth? To answer that, based on the enlightenment from the above calculations, we design an experiment, as shown in Fig. 3a, to help the contribution from the neonatal Cu droplets to be emerged by covering a vaulted quartz tile which the inner height of the arch is 7.4 mm, widish to the mean free path of single Cu atom (~5.8 mm, discussions in Fig. S9), to restrict the moving of Cu vapor slightly under the cap (‘Cap experiment’ in Table S1). For general gaseous molecules in medium vacuum, without any energy changes, the rebound gaseous molecule from the cap will be scattered by another molecule. Consequently, the probability of elastic collision under the arch among gaseous molecules will be tinily increased. However, the graphene growth here wouldn’t be influenced due to the considered unchanged flow condition since the ACS Paragon Plus Environment
Page 6 of 23
Page 7 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Knudsen number for CH4 is ~0.3 in this experiment (Fig. S9). Moreover, for the ‘sticky’ Cu atoms from the foil and inner chamber wall, instead of scattering, the rebound particles at the ‘affected size’ under the canopied region, mostly one-to-few-atoms Cu droplets, will aid the coalescence of Cu droplets due to the Ostwald ripening process so that the airborne Cu clusters will have more supplies from the neonatal Cu droplets than the uncapped region. Therefore, for graphene growth, if the carbon feedstock could be transported from the redeposited Cu particles to bulk Cu, the slight increment of the confined ‘fine’ Cu droplets will help the growth in the capped environment due to their tremendous carbon solubility. Comparing with the results in the 3-min growth at 1020 oC under 8×10-2 torr in Fig. 3b and 3c, the appreciable difference is distinct from the nearly full-covered growth (capped, Fig. 3c) to 84% coverage (uncapped, Fig. 3d). The essentialness of carburization of Cu droplets during growth is now unveiled. Noticeably, under the same condition with an addition of 10 sccm H2 in Fig. 3d and 3e, this contribution is smeared (~2% increment left in graphene coverage) indicating the re-hydrogenation process18 from the additional decomposed hydrogen adatom moderating the carbon content in the neonatal or host Cu cluster, shortening the ‘carrying distance’ of carbon, or moderating the ‘segregated’ carbon at the surface of the ‘ripened clusters’ (discussed in Fig. 6). Another experiment is designed to further understand the carburization of airborne Cu particles by applying the direct growth on a bare SiO2 substrate via the assistance of ambient Cu vapor (‘Growth on SiO2 with Cu vapor supported’ in Table S1). Diamond like carbon (DLC) films with the centered ball-like diamonds36 are grown, as shown in Fig. 4a and 4b. In Fig. 4c, the ball-like structure grows outward from the DLC sheet at the center and some aggregated Cu clusters underneath the circle sheet are found at the fringe, as shown in Fig. 4d. The anatomized studies of these interesting structures are carried out in Fig. S12–18 with the following three conclusions: (1) The growths of DLC films and ball-like structures are positively correlated to the quantity of ambient Cu vapor. (2) The concluded structures of our ball-like diamonds are composed by local SP2 and SP3 bonded structures, which shows distinct diagnostic signals of SP2 and SP3 with their corresponding defective signals in Raman and the SP2-SP3 mixed bonding signal. The SP1 is not probed. The foregone understanding of surface chemical adsorption on bulk Cu cannot well interpret the phenomena in the discussions. (3) Under the same conditions with additional 10 sccm hydrogen flow, no DLC films nor ball-like structures are grown.
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
By combining these three inferences, the inevitable ripening process, and the results in Fig. 2d and 3, the carburization of Cu particles is thus embodied in Fig. 6a and 6b later. For the interaction between Cu vapor and foil surface to complete the entire carbon transport in this article, the theory of Terrace-Step (or ledge)-Kink (TSK) is introduced, as Fig. 5a, which has well interpreted the sublimation, thin film deposition and activities of adatoms on solid surface37. Under this knowledge with the contributions from the ambient gases, an interesting presentation in Fig. 5b, the fractal structures produced under a ‘very usual’ annealing condition, a brand new Cu foil in a Cu-decorated chamber with 150 sccm Ar and 50 sccm H2 introduced at 1020 oC in 30 min and 9.1×10-1 torr (‘Fractal surface 2’ in Table S1), could be well interpreted. The detailed studies are carried out in Fig. S19–S22 with various controls in chamber condition, gases, heating period, temperature, and foil surface. The concluded interpretation of the fractal structures in our environment is sure as the presentation of pure interactions among ambient Cu vapor, gas, and surface, as an evidential presentation for the affinitive interaction between Cu vapor and bulk surface. For the CH4 environment, the interaction between ripened clusters and surface is carried out as well with a control experiment under MW equivalent dose of Ar in Fig. S23. From the discussions in Fig. S3, S22 and S23, the interactions between the ripened clusters and the foil surface among the typical growth gases, H2, Ar, and CH4, are embodied which the molecular collisions from ambient gases of different MW do effect the ripening of cluster during flying and the motions of adatoms (or ad-clusters) on bulk surface. In CH4 environment, the middle MW gas among them, with unobstructive lightest gas, H2, to sweep the deposited carbon, the configurative contributions for the cluster collision to the surface steps and grain boundaries are emerged as the Fig. 5c, which implies the so called ‘step bunching’ phenomenon38–40 in graphene growth is relevant to this. Fig. 5d shows this experiment in our environment (Fig. S24). We found the growth speed is reduced when the steps are all occupied, implying the efficiency of carbon transport is relevant to the step configuration. The mechanisms are discussed in this section. The decomposition of methane to the radicals on bulk Cu is shown in Fig. 6a as (1), for the case of CH4(g)→CH4(s)→CH3(ads)+H(ads) with (g) as gas, (s) as surface, and (ads) as adsorbate. The hydrogen adsorbate tends to diffuse into Cu due to its high H2 solubility18 and will also attack the attached radicals on bulk Cu, as (2) for the case of H2(g)→H2(s)→H(ads)+H(ads) to H(ads)+CH3(ads)→CH4(g). From the observations in the calculation42 and experiments19–22, the pyrolysis of methane sequentially through methyl, methylene and methylidyne to atomic carbon is unfavorable on bulk Cu and the couplings between the radicals to higher species of ACS Paragon Plus Environment
Page 8 of 23
Page 9 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
hydrocarbon gases are found spontaneously, as (3) for the case of CH3(ads)+CH2(ads)→C2H4(g)+H(ads). Additionally for the unstable atomic carbon on Cu surface23, the succedent formation of C2 dimer is also found exothermically in Ref. 43. However, in our calculations in Fig. 2d, the atomic decomposition becomes ‘valid’ since the tremendous increment of carbon solubility is endued from the embryonic droplets of Cu, as (4) for the calculated case at 2 nm which is 71773.03% higher (~0.53 at.%) than bulk Cu and similar to bulk Ni (~0.59 at.%) at 1020 oC. Thus, during the ripening process to larger cluster, the solubility in the host cluster will drop rapidly from the released surface energy so that the exceeded inner carbon from the fused droplets will be segregated at the surface, like Fig. 6d for the ‘smallest’ observed case in CH4 at 30 nm which the dramatic drop is from 0.53 at.% to 9.2 at.ppm at 1020 oC (more difference could be expected in larger cluster or ripened size in Ar environment). The massive segregations will appear at the cluster surface, as the left in Fig. 6b which is overstating for better illustration of the growth mechanism for DLC and ball-like diamond in Fig. 4 (discussed in Fig. S17). The moderated segregation on larger cluster surface with H2 is expected (Fig. 3) so that the DLC and ball-like structures are not grown (Fig. 4 and Fig. S17). Furthermore, for the interactions between the metal clusters and their bulk surface, for the deposited melting small particles in the molecular dynamics44–46, the adhesion and the following lateral epitaxial rearrangement on the surface are perceived within picoseconds for Cu46. For the large cluster case44,46, the reflection is happened in the low kinetic impaction case via the deformation at the impaction point and the afterward rebound. The epitaxy for larger cluster case with higher requirement of deposition energy is also viable47. Therefore, for the epitaxial case like the 2 nm ‘droplet’ in Fig. 6c after the deposition, the atomic carbon ‘in’ the droplet is released and diffuse via the surface atomic sites on bulk48,49 during graphene growth (no bulk diffusion18 and programable shallow diffusion11 is probed). The diffusion barrier for both monomer (0.1 eV on Cu(111)) and dimer (0.27 eV) are friendly47 but not to the hydrocarbon radicals20–22. For larger cluster, the carbon ‘on’ the cluster could be also passed to the bulk surface ‘during’ the impaction, the deformation, dragging, or rolling. In other word, the airborne Cu clusters from the embryonic to the ripened sizes, effected by ambient gases for ripening process, are the ‘vehicles’ of atomic carbon during graphene growth. The drifting carbon on bulk surface will be slowed down at the defective or higher diffusion barrier region as the rallying point and then nucleate38,40,41. The higher activation energy25 requirement and the ‘symbolized’ hexagonal shape26 for less environment10 in AP could be thus explained from the suppressed Cu evaporation (Fig. S6d). Moreover, for the relatively high methane partial pressure to H241, the step ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
bunching are shown and could be attributed to the higher collision probability for the incoming clusters (in Fig. 6d, Fig. S23 and Fig. S24). Along with higher H2, this effect will be disappeared41 since the hydrogen could moderate the segregated carbon on cluster surface during ripening in our observation. For the self-limit property, the carbon feedstock from the airborne clusters (not from the surface absorption, shallow diffusion and following segregation) could do the interpretation. Therefore, this unique property turns Cu, the ‘nearlycarbon-insoluble’ metal, to the preferred material for mono-layer graphene growth. For the bottom-up structure10,12,15–17, which the bottom layer is found appeared in the early stage10,12, when the debarking carbon atoms, diffusing via the surface atomic sites, are caught in a defective region and nucleate, the formed graphene sheet will become ‘surface blind’48 to the atomic sites so that there is a certain chance for other debarking carbon atoms to diffuse underneath via the atomic sites and form another graphene. Hence, since the top layer is wider than the bottom, the efficiency for carbon supplies due to the configuration will be distinguished gradually, as shown in Fig. 6e. Through all the evidences and the discussions in this article, this vehicle theory of the airborne Cu clusters which transport the atomic carbon to the bulk Cu surface and do the graphene is thus proposed as the essential puzzle for the graphene synthesis via Cu-catalytic CVD method. This vehicle theory could be also applied to the inconsistent presentations in Ref. 41 and Ref. 50 for the orientation differential growth on Cu by considering the grain dependent binding energy to Cu clusters (Fig. S22) for carbon transport as the future study. Conclusion The inconsistent observations and questions for the related studies on graphene synthesis on Cu are referred first. Four different graphene growths are manipulated to distinguish the importance of the quantity of ambient Cu vapor. The tremendous increment of carbon solubility in Cu droplet is calculated and carried out from the cap experiment by increasing the Ostwald ripening during growth. The carburization of Cu vapor are examined in the growth of DLC and ball-like diamonds on SiO2. and the interactions between Cu vapor and bulk Cu surface, the ripening process to growth gases, and the carbon transport from Cu clusters to bulk surface are also examined. The vehicle theory of the airborne Cu clusters is thus amended as the essential puzzle for the graphene synthesis via Cu-catalytic CVD method.
Acknowledgement
ACS Paragon Plus Environment
Page 10 of 23
Page 11 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
The research is supported by Ministry of Science and Technology through Grant through grants no, 1022112-M-009-006-MY3, 101-2218-E-007-009-MY3, 102-2633-M-007-002, and the National Tsing Hua University through Grant no. 102N2022E1. Y.L. Chueh greatly appreciates the use of facility at CNMM, National Tsing Hua University through Grant no. 102N2744E1. Supporting information Methods section, experiments details, annealing exam, CALPHAD details, carburization exam, Raman studies on DLC film and ball-like nano diamonds, surface fractal exam, and step bunching experiments. This material is available free of charge via the Internet at http://pubs.acs.org.
Reference: (1) Robertson, J. J. Mater. Chem. 2012, 22 (37), 19858–19862. (2) Wintterlin, J.; Bocquet, M.-L. Surface Science 2009, 603 (10–12), 1841–1852. (3) Sutter, P.; Sadowski, J. T.; Sutter, E. Phys. Rev. B 2009, 80 (24), 245411. (4) Yu, Q.; Lian, J.; Siriponglert, S.; Li, H.; Chen, Y. P.; Pei. S.-S. Appl. Phys. Lett. 2008, 93 (11), 113103. (5) Cui, Y.; Fu, Q.; Bao, X. Phys. Chem. Chem. Phys. 2010, 12 (19), 5053–5057. (6) Liu, M.; Zhang, Y.; Chen, Y.; Gao, Y.; Gao, T.; Ma, D.; Ji, Q.; Zhang, Y.; Li, C.; Liu, Z. ACS Nano 2012, 6 (12), 10581–10589. (7) Nie, S.; Walter, A. L.; Bartelt, N. C.; Starodub, E.; Bostwick, A.; Rotenberg, E.; McCarty, K. F. ACS Nano 2011, 5 (3), 2298–2306. (8) Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D.; Piner, R.; Velamakanni, A.; Jung, I.; Tutuc, E.; Banerjee, S. K.; Colombo, L.; Ruoff, R. S. Science 2009, 324 (5932), 1312–1314. (9) Coraux, J.; N’Diaye, A. T.; Engler, M.; Busse, C.; Wall, D.; Buckanie, N.; Heringdorf, F.-J. M., Gastel. R.; Poelsema, B.; Michely. T. New J. Phys. 2009, 11 (2), 023006. (10) Li, Q.; Chou, H.; Zhong, J.-H.; Liu, J.-Y.; Dolocan, A.; Zhang, J.; Zhou, Y.; Ruoff, R. S.; Chen, S.; Cai, W. Nano Lett. 2013, 13 (2), 486–490. (11) Mofford, J. M.; Nie, S.; McCarty, K. F.; Bartelt, N. C.; Dubon. O. D. Nano Lett. 2010, 10 (12), 4890–4896. (12) Li, X.; Cai, W.; Colombo, L.; Ruoff, R. S. Nano Lett. 2009, 9 (12), 4268–4272. ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(13) Arnoult, W. J.; McLellan, R. B. Scri. Metall. 1972, 6 (10), 1013–1018. (14) López, G. A.; Mittemeijer, E. J. Scripta Mater. 2004, 51 (1), 1–5. (15) Nie, S.; Wu, W.; Xing, S.; Yu, Q.; Bao, J.; Pei, S.-S.; Mcarty, K. F. New J. Phys. 2012, 14 (9), 093028. (16) Wu, B., Geng, D.; Guo, Y.; Huang, L.; Xue, Y.; Zheng, J.; Chen, J.; Yu, G.; Liu, Y.; Jiang, L.; Hu. W. Adv. Mater. 2011, 23 (31), 3522–3525. (17) Wu, W.; Yu, Q.; Peng, P.; Liu, Z.; Bao, J.; Pei. S.-S. Nanotecknology 2012, 23 (3), 035603. (18) Losurdo, M.; Giangregorio, M. M.; Capezzuto, P.; Bruno, G. Phys. Chem. Chem. Phys. 2011, 13 (46), 20836–20843. (19) Chuang, T. J., Chan, Y. L. Chuang, P., Klauser, R. J. Electr. Spectr. Relat. Phenom. 1999, 98–99, 149–173. (20) Chuang, P.; Chan, Y. L.; Chien, S.-H.; Song, K.-J.; Klauser, R.; Chuang, T. J. Langmuir 2002, 18 (12), 4549–4552. (21) Chan, Y. L.; Pai, W. W.; Chuang, T. J. J. Phys. Chem. B 2004, 108 (3), 815–818. (22) Zhai, R.-S.; Chan, Y. L.; Chuang, P.; Hsu, C.-K.; Mukherjee, M.; Chuang, T. J.; Klauser, R. Langmuir 2004, 20 (9), 3623–3631. (23) Zhang, W.; Wu, P.; Li, Z.; Yang J. J. Phys. Chem. C 2011, 115 (36), 17782–17787. (24) Bhaviripudi, S.; Jia, X.; Dresselhaus, M. S.; Kong, J. Nano Lett. 2010, 10 (10), 4128–4133. (25) Vlassiouk, I.; Smirnov, S.; Regmi, M.; Surwade, S. P.; Srivastava. N.; Feenstra, R.; Eres, G.; Parish, C.; Lavrik, N.; Datskos, P.; Dai, S.; Fulvio, P. J. Phys. Chem. C 2013, 117 (37), 18919–18926. (26) Yu, Q.; Jauregui, L. A.; Wu, W.; Colby, R.; Tian, J.; Su, Z.; Cao, H.; Liu, Z.; Pandey, D.; Wei, D.; Chung, T. F.; Peng, P.; Guisinger, N. P.; Stach, E. A.; Bao, J.; Pei, S.-S.; Chen, Y. P. Nat. Mater. 2011, 10 (6), 443–449. (27) Hersh, H. N. J. Am. Chem. Soc. 1953, 75 (7), 1529–1531. (28) Feder, J.; Russell, K. C.; Lothe, J.; Pound, G. M. Adv. in Phys. 1966, 15 (57), 111–178. (29) Flagan, R. C.; Lunden, M. M. Mater. Sci. Eng. A 1995, 204 (1), 113–124. (30) Granqvist, C. G.; Buhrman, R. A. J. Appl.Phys. 1976, 47 (5), 2200–2219. (31) Buffat, Ph. & Borel, J-P. Phys.Rev. A 1976, 13 (6), 2287–2298. (32) Homma, Y.; Liu, H.; Takagi, D.; Kobayashi, Y. Nano Res. 2009, 2 (10), 793–799. ACS Paragon Plus Environment
Page 12 of 23
Page 13 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
(33) Takagi, D.; Kobayashi, Y.; Hibino, H.; Susuki, S.; Homma, Y. Nano Lett. 2008, 8 (3), 832–835. (34) Takagi, D.; Homma, Y.; Hibino, H.; Suzuki, S.; Kobayashi, Y. Nano Lett. 2006, 6 (12), 2642–2645. (35) Tanaka, T. Mater. Sci. Forum 2010, 653 (1), 55–75. (36) Angus, J. C.; Hayman, C. C. Science 1988, 241 (4868), 913–921. (37) Zhang, Z.; Lagally, M. G. Science 1997, 276 (5311), 377–383. (38) Han, G. H.; Günes, F.; Bae, J. J.; Kim, E. S.; Chae, S. J.; Shin, H.-J.; Choi, J.-Y.; Pribat, D.; Lee, Y. H. Nano Lett. 2011, 11 (10), 4144–41348. (39) Hayashi, K.; Sato, S.; Yokoyama, N. Nanotechnology 2013, 24 (2), 025603. (40) Nie, S.; Wofford, J. M.; Bartelt, N. C.; Dubon, O. D.; McCarty, K. F. Phys. Rev. B 2011, 84 (15), 155425. (41) Hayashi, K.; Sato, S.; Ikeda, M.; Kaneta, C.; Yokoyama, N. J. Am. Chem. Soc. 2012, 134 (30), 12492–12498. (42) Au, C.-T.; Liao, M.-S.; Ng, C.-F. J. Phys. Chem. A 1998, 102 (22), 3959–3969. (43) Gajewski, G.; Pao, C.-W. J. Chem. Phys. 2011, 135 (6), 064707. (44) Yasumatsu, H.; Kondow, T. Rep. Prog. Phys. 2003, 66 (10), 1783–1832. (45) Hussain, F.; Hayat, S. S.; Imran, M.; Ahmad, S. A.; Bouaffa, F. Comput. Mater. Sci. 2012, 65 264– 268. (46) Meinander, K.; Frantz, J.; Nordlund, K.; Keinonen, J. Thin Solid Films. 2003, 425 (1), 297–303. (47) Meinander, K.; Nordlund, K.; Keinonen, J. Nucl. Instr. and Meth. in Phys. Res. B 2006, 242 (1–2), 161–163. (48) Mi, X.; Meunier, V.; Koratkar, N.; Shi, Y. Phys. Rev. B 2012, 85 (15), 155436. (49) Riikonen, S.; Krasheninnikov, A. V.; Halonen, L.; Nieminen, R. M. J. Phys. Chem. C 2012, 116 (9), 5802–5809.
(50) Wood, J. D.; Schmucker, S. W.; Lyons, A. S.; Pop, E.; Lyding, J. W. Nano Lett. 2011, 11 (11), 4547–4554.
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure Captions Fig. 1. Growth demonstrations for the importance of quantity of Cu vapor from the chambers and different handles. (a) SEM image of the 3rd 30-min-grown Cu foil in a new chamber at 1020 oC with 50 sccm CH4 in 8×10-2 torr (sequential records from beginning listed in Fig. S3–4). Till the 3rd growth, petty graphene spots are found at pointed region. (b–c) SEM images of the 1st and 3rd 30-min-grown Cu foil in another new chamber under the same growth condition of a with an addition of 30 min pre-heating process (H2: 50 sccm, without CH4). Comparing to a, slightly more graphene grains are found in the 1st growth (pointed region in b), and the coverage of graphene is massively increased at the 3rd growth in c. (d–e) SEM images of the 30min-grown Cu foil processed in another Cu-decoration chamber with zoomed images as inserts. Under exact the same condition of A, the coverage is full even in the 1st growth in d or with additional H2 in e, as the sample A and B respectively. (f) Raman signals of the sample A and B.
Fig. 2. Discussions of the coalescence and the carburization of Cu particles during growth. (a) Distribution of the corresponding Cu vapor pressure at the 1000 oC in our 3-inch-wide furnace chamber. (b–c) SEM images of the Cu particles caught by the scratched SiO2 chip deployed at the chamber center in the Cudecorated chamber after 30 min heated at 950 oC with 50 sccm CH4 under 8×10-2 torr. The diameter of particles is ~1 µm here and each particle shows strong carbon component in the quantitative EDX analysis (Fig. S7). (e) Cu–C binary equilibrium phase diagram at Cu-rich side in bulk and nano-sized particles modified by the surface energy contributed from size effect of nano particle. The inserted diagram is for the 2 nm nano-particle case near the temperature of graphene synthesis.
ACS Paragon Plus Environment
Page 14 of 23
Page 15 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Figure 3. Capping experiment. (a) Schematic drawing of the capping experiment. The capping experiment here is to reveal the growth contribution from the redeposited ultra-fine Cu droplets by slightly limiting the movement to foil under the distance approach from the estimation of mean free path for the sub-nm (0.256–1 nm) sized droplets (detailed experiment configuration in Fig. S11). (b–e) SEM images of the as-3-min-grown Cu foils at 1020 oC in 8×10-2 torr. The juxtaposed uncapped and capped Cu foils under gas conditions of 50 sccm CH4 or with addition of 10 sccm H2 are listed in the first (b, c) and 2nd row (d, e) respectively. The appreciable increasing in graphene coverage is observed at the capped Cu foil with pure CH4 condition in c. In the case of e, with the addition of H2, this contribution is smeared.
Fig. 4. Direct growth on the amorphous SiO2 chip with Cu vapor assisted. (a–d) SEM images of the as-90min-grown SiO2 chips at 1020 oC with 50 sccm CH4 under 9×10-1 torr in the Cu-decorated chamber . Three circular products, from infant to mature, are shown in a. As long as the carbon is deposited at the nucleation site, the lateral growth is preferred in the early stage (in b, only small ball-like nano diamond is grown at the center. The pointed particle in b is a Cu cluster (~1 µm)). The close view in c shows the ball-like nano diamond grows outward from the center of DLC sheet, and the pointed clusters at the edge of cracked DLC film in d are the Cu clusters which are the aggregations from the penetrate Cu vapor through the amorphous DLC film during long time growth and they also reduce the SiO2 as the lumpy surface (detailed studies in Fig. S12–18).
Fig. 5. Discussions of the correlation among Cu vapor, foil surface, and graphene. (a) Schematic drawing of typical atomistic processes in TSK theory. The evaporation and re-deposition of Cu atoms (as 1) are continual during heating. The deposited adatoms will (2) attach to the surface step and diffuse along the step, (3) connect other adatoms to form a 2D or (4) 3D island, or (5) diffuses in a constant distance on Cu surface. (b–c) SEM images for diversiform behaviors for the interaction of Cu vapor, surface and ambient gases. b is the fractal structures (Fig. S19–S22), c is the ad-cluster behavior in methane (Fig. S23), and d is the reexamination in ‘Step bunching’ phenomenon (Fig. S24).
Fig. 6. Schematic drawings of Growth mechanism. (a) Mechanisms for (1) dehydrogenation, (2) hydrogenation, and (3) radical coupling on bulk Cu surface. (4) is for the atomic carbon dehydrogenation in ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2 nm cluster (droplet). (b) Mechanism for the carburization of Cu cluster during the Ostwald ripening process, in the case of the smallest cluster (30 nm) we observed in CH4 (Fig. S3) as the host with 2 nm droplets as the adsorbates under H2 (right) or H2-absent (left) environment. The aggregated segregation in the left is overstated for better illustration for the growth mechanism of DLC film and ball-like diamonds in Fig. 4. The moderated segregations in the right are drawn as dimers. (c) Mechanism for the carbon transportation via Cu clusters to the foil surface, in the case of the 2 nm droplet in CH4 and H2 environment. The insert is the profile for the epitaxial rearrangement after adhesion to bulk surface. (d) The collision of cluster to the flat surface and surface step. (e) The transported atomic carbon to the bottom-up configuration.
Figure 1
ACS Paragon Plus Environment
Page 16 of 23
Page 17 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Figure 2
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 3
ACS Paragon Plus Environment
Page 18 of 23
Page 19 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Figure 4
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 5
ACS Paragon Plus Environment
Page 20 of 23
Page 21 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Figure 6
ACS Paragon Plus Environment
The Journal of Physical Chemistry
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
TOC
ACS Paragon Plus Environment
Page 22 of 23
Page 23 of 23
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
ToC graphics
ACS Paragon Plus Environment
