Graphene Nucleation Preference at CuO Defects Rather Than Cu2O

Subscriber access provided by University of Sunderland

Surfaces, Interfaces, and Applications

Graphene Nucleation Preference at CuO Defects Rather Than Cu2O on Cu(111): A Combination of DFT Calculation and Experiment Xiucai Sun, Zhen Su, Jing Zhang, Xizheng Liu, Yanlu Li, Fapeng Yu, Xiufeng Cheng, and Xian Zhao ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b13626 • Publication Date (Web): 06 Nov 2018 Downloaded from http://pubs.acs.org on November 7, 2018

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 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 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.

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 36 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

ACS Applied Materials & Interfaces

Graphene Nucleation Preference at CuO Defects Rather Than Cu2O on Cu(111): A Combination of DFT Calculation and Experiment Xiucai Sun,1 Zhen Su, 1 Jing Zhang, 1 Xizheng Liu,2 Yanlu Li, 1* Fapeng Yu, 1* Xiufeng Cheng, 1 Xian Zhao1 1

Institute of Crystal Materials and State Key Laboratory of Crystal Materials, Shandong

University, Jinan, 250100, P.R. China 2

Tianjin Key Laboratory of Advanced Functional Porous Materials, Institute for New Energy

Materials and Low-carbon Technologies, School of Materials Science and Engineering, Tianjin University of Technology, Tianjin 300384, P.R. China * Corresponding authors emails: [email protected], [email protected]

ABSTRACT It is well-known that reducing the nucleation density is an effective way to enhance the growth quality of graphene. In this work, we explore the mechanism of graphene nucleation and growth around CuO defects on a Cu(111) substrate by using density functional theory (DFT) combined with the nudged elastic band (NEB) method. The defect formation mechanism at the initial nucleation stage is also studied. Our calculation results of the C adsorption energy and the reaction barrier of C–C dimer formation illustrate that the initial nucleation of graphene could be promoted by artificially introducing CuO defects on a Cu(111) surface and the nucleation on the clean Cu(111) substrate could thus be suppressed. These conclusions have been verified by graphene growth experiments using a chemical vapor deposition (CVD) method. Further studies

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 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

Page 2 of 36

showed that graphene grown around CuO “seed crystals” could maintain its structural integrity without significantly producing defective carbon rings. This work provides fundamental understanding and theoretical guidance for controllable preparing large dimension and high quality graphene by artificially introducing CuO seeds.

KEYWORDS: Graphene; Nucleation; CuO; Cu2O; DFT


2

Page 3 of 36 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


1. INTRODUCTION Graphene has been extensively studied due to its extraordinary electrical, optical and mechanical properties.1–4 The reliable synthesis of graphene with large dimension and high quality is a prerequisite for realizing its potential uses in various devices. Chemical vapor deposition (CVD) on transition metal substrates (especially Cu) is currently considered to be the most promising method for rapid growth of large area graphene films5,6 as the catalytic action of Cu substrate promotes the decomposition of the carbon source gas (such as CH4) and the nucleation and growth of graphene. 7 Numerous studies have demonstrated that graphene could spontaneously nucleate and reconstruct on the Cu substrate surface to form graphene domains. A large number of defects resulting in non-hexagonal rings would be introduced at the boundaries of these domains 8 which would hinder the migration of electrons and compromise the excellent electrical properties of graphene.9,10 It is widely recognized that reducing the nucleation density is an effective way to grow a high-quality graphene film with fewer boundaries and high carrier mobility.5,11 Extensive attempts have been made to control the nucleation site density on Cu substrates for high quality graphene growth,8,12–14 but the underlying mechanism is far from being understood. As reported recently, feeding oxygen gas during the Cu substrate pretreatment process is a generally applicable method to artificially control the graphene nucleation sites.15–17 Chang et al. suppressed the spontaneous nucleation of graphene by passivating the Cu surface with an oxide layer, and then performed a reduction process to expose a low-density of fresh Cu surface sites as nucleation centers of graphene.18 In addition, Liang et al. have proved that graphene nucleation preferentially occurs at the O-rich sites on a Cu surface by theoretical calculations and experimental methods.19 In our previous study, we found that copper oxide (CuO) defects could


3


Page 4 of 36

be formed on a Cu (001) substrate surface and act as graphene nucleation sites, making the graphene grow around the CuO defects in a regular manner.20 Numerous studies have demonstrated that CuO and Cu2O are the main oxide defects which are formed during the oxidation pretreatment of a Cu substrate,21–25 however, there is still doubt as to which type of copper oxide defect is dominant, and the mechanism of graphene nucleation around copper oxide defects is also not clear yet. Since CuO and Cu2O have different crystal structures and stabilities, these non-periodic defect structures could form different step structures on the Cu surface and thus show distinct influences on graphene nucleation. Therefore, it is essential to identify the surface oxide species and to elucidate the roles of these copper oxide defects on the mechanism of graphene nucleation. Since the hexagonal lattice symmetry of the Cu(111) surface matches well with the honeycomb lattice of graphene, which makes it very advantageous for graphene synthesis,26 we therefore focus on the nucleation mechanism of graphene on Cu(111) substrates. Moreover, non-hexagonal rings such as C5 and C7 are the most common defects formed during graphene growth, and could affect the electric properties of graphene.27 It is therefore of significant interest to elucidate the transformation mechanism between these defects and the perfect graphene structure during the initial nucleation stage to gain a fundamental understanding of the relationship between nucleation and graphene quality. In this work, density functional theory (DFT) was combined with the nudged elastic band (NEB) method to explore the mechanism of graphene nucleation on Cu(111) substrates with CuO and Cu2O defects. The adsorption energies of C and CH active species at different copper oxide surfaces and their atomic steps were compared. The elementary reactions of C aggregation at the surface steps have been studied to evaluate the nucleation barrier. Furthermore, the transformation reactions between the C5, C7 defect rings and the perfect C6 ring were also


4

Page 5 of 36 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


investigated to illustrate the defect self-recovery and formation reactions at the surface steps. Related reactions on the clean Cu(111) surface were also studied for comparison. The oxide species as well as the graphene nucleation and growth around the CuO defects were experimentally verified. 2. CALCULATION AND EXPERIMENTAL SECTIONS 2.1 Computational details All the DFT calculations were performed using the Vienna Ab initio Simulation Package (VASP).28,29 The projector augmented wave (PAW) method was used to describe the interactions between the core electrons and the valence electrons. The exchange correlation potential was described by the Perdew-Burke-Ernzerhof (PBE) form within the generalized gradient approximation (GGA) functional.30 The convergence criterion of total energy was set to be 10-4 eV, and that of the force on each atom was set to be 0.05 eV/Å. For structural optimizations and energy calculations, the electron wave functions were expanded in a plane wave basis set with a kinetic energy cutoff of 520 eV. Monkhorst-Pack k-point grids of 2×2×1 were used. By using the above calculation parameters, the lattice constant of bulk Cu (face-center cubic lattice, space group F3) was calculated to be 3.64 Å, which agrees quite well with the experimental value of 3.61 Å.31 For transition state calculations, the climbing image nudged elastic band (CI-NEB) method was used to determine the energy barriers of various kinetic processes.32,33 The transition states during each elementary reaction were verified by transition state imaginary frequency calculations. Two types of surface models were constructed, including clean Cu(111), CuO(111), and Cu2O(111) surfaces as well as Cu(111) surfaces with CuxO (x=1 or 2) defects. The (111) surface was chosen for monoclinic CuO (space group C2/ ) and cubic Cu2O (space group P3 )


5


Page 6 of 36

according to previous reports of experimental CVD graphene synthesis6,25 and theoretical calculations.31,34 The clean Cu(111) surface was constructed by cutting a face-centered cubic (fcc) Cu crystal along the [111] direction, and then constructing the p(2×2) Cu(111) surface with five Cu layers. A vacuum layer of 15 Å was added along the z direction to construct the slab supercell models. The Cu2O(111) surface was constructed the same way as the Cu(111) surface, while the CuO(111) surface was constructed by using p(4×2) surfaces. We performed convergence tests of the total energies with the number of atomic layers. The results show that the two atomic layers at the bottom could be fixed in their bulk positions while the other three surface atomic layers should be fully relaxed. Furthermore, the models with oxide defects were constructed by putting the non-stoichiometric CuO (Cu7O8) and Cu2O (Cu13O6) clusters as well as the stoichiometric CuO (Cu8O8) and Cu2O (Cu12O6) clusters on p(5x5) Cu(111) surfaces with three-atom-thickness. The lattice matching and the distance between the oxide clusters and Cu(111) surfaces were examined. C and CH were treated as the active carbon species for graphene nucleation and growth on the Cu substrates. The structures with the lowest formation energies are labeled as CuO/Cu(111) and Cu2O/CuO(111); these were used in the adsorption calculations of C active species and various kinetic processes. We use the binding energy (Ebind) to evaluate the binding ability between CuxO clusters and Cu(111) surfaces, which was calculated according to the following equation: Ebind = ECux O/Cu(111) − ECux O − ECu(111)

(1)

where ECux O/Cu(111) is the total energy of a Cu(111) surface with CuO or Cu2O clusters, ECux O and ECu(111) are the total energies of CuxO clusters and the Cu(111) surface. Hence, a negative value of Ebind represents a strong binding ability between defects and surfaces.


6

Page 7 of 36 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 stability of the adsorption geometry can be measured by the adsorption energy (Eads), which is defined as:

Eads = EC/surf − Esurf − EC

(2)

where EC/surf is the total energy of the surfaces (both the clean surfaces and surfaces with CuxO defects) with carbon active species (C or CH), while Esurf and EC are the total energies of the surfaces and the carbon active species respectively. Similarly, a negative value of Eads represents an exothermic process and stable adsorption. During the elementary reaction calculations and transition state search, the reaction barrier Ea and reaction energy ∆E are calculated as: Ea = ETS − EIS

(3)

∆E= EFS − EIS

(4)

where EIS, ETS, and EFS represent the energies of the initial state (IS), transition state (TS), and final state (FS), respectively. 2.2 Growth and characterization of graphene Graphene was grown on 25 µm thick commercial copper foils (Alfa Aesar, 99.8%, NO.46986) by a CVD method. Firstly, the copper substrate was electrochemically polished at a voltage of 1.8 V for 30 min in an electrolyte solution (H3PO4:PEG＝3:1) to make the surface smoother, then rinsed with distilled water. Subsequently, the polished copper substrate was ultrasonically washed in absolute ethanol for 15 min. In the CVD process, high purity Ar gas (5N, 320 sccm) was introduced throughout the whole process as a background atmosphere. Before the graphene growth, the tube furnace was first heated up to 300 °C and kept for 30 min under a low flow of O2 (5N, 5 sccm) to obtain an oxide layer on the copper surface. Thereafter, H2 gas (5N, 21 sccm) was introduced into the furnace for reduction and etching. The tube furnace is continuously


7


Page 8 of 36

heated up to 1000 °C (in 40 min), and the temperature was then slowly increased to 1035 °C (in 30 min). After thermal treatment at 1035 °C for 60 min, the temperature was gradually lowered down to 1000 °C again (this called the temperature gradient annealing process) to achieve a single crystal Cu substrate. The graphene nucleation and growth were conducted on the Cu substrate under a mixture of CH4 (5N, 5 sccm) and H2 (5N, 21 sccm) at 1035 °C for 10 min, after which the temperature was rapidly cooled down below 700 °C. The grown graphene was transferred onto a SiO2/Si substrate by using a PMMA-assisted wet method35 for further characterization. The crystallinity and the orientation of the copper substrate after temperature gradient annealing were analyzed using X-ray diffraction (XRD). The morphology of the grown graphene was observed using a field emission scanning electron microscope (FE-SEM, Hitachi S-4800). The Raman spectra of the as-grown graphene were acquired by using a LabRAM HR800 confocal Raman spectrometer in a backscattering configuration with a 532 nm laser for excitation. Elemental and valence analysis were conducted by using energy dispersive spectroscopy (EDS) and an X-ray photoelectron spectrometer (XPS, ThermoFisher ESCALAB 250) equipped with a monochromatized Al Kα X-ray source under ultrahigh vacuum (< 10-7 Pa). The binding energies were calibrated by using the C 1s peak (284.6 eV) of contaminated carbon as a reference. High-resolution transmission electron microscopy (HRTEM) images and selected-area electron diffraction (SAED) patterns were obtained by a Talos F200X highresolution transmission electron microscope with FEG, where the samples were prepared by using an FIB-SEM system (FIB, Helios NanoLab460HP).


8

Page 9 of 36 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


3. RESULTS AND DISCUSSION 3.1 C active species adsorption on Cu(111), CuO(111) and Cu2O(111) surfaces The growth mechanism of graphene on copper substrates has been experimentally demonstrated to be a surface adsorption process,36 and the step-by-step aggregation of carbon active species on the copper surface dominates the initial nucleation of graphene.7 Therefore, in this work, we first investigated the adsorption of C active species on clean surfaces. It has been reported by a large number of research groups that CH4 is widely used as the carbon source gas for graphene CVD growth, and its decomposition could provide abundant C and CH reactive groups as the dominant precursors of graphene growth.37,38 Therefore, we compared the adsorption behaviors of C and CH active species on the clean Cu(111), CuO(111), and Cu2O(111) surfaces. The possible adsorption sites of C and CH active species on the three relaxed clean surfaces are illustrated in Figure S1. Both the surface and sub-surface adsorption sites on the three kinds of surfaces were considered. We also considered both parallel and vertical initial placements of CH on all the possible adsorption sites in order to include the effect of CH steric configuration on the adsorption energies. Finally, the most stable adsorption configuration was obtained by comparing the adsorption energies of C active species on all possible adsorption sites of the surfaces. 3.1.1 C adsorption First, we compare the adsorption energy, Eads, of a C monomer on clean Cu(111), CuO(111), and Cu2O(111) surfaces to find the energetically preferable adsorption site and surface. The corresponding adsorption energies on the considered sites of the three clean surfaces are listed in Table S1. It was found that for a clean Cu(111) surface the most stable surface adsorption of a C monomer occurs at the Br site, which is in agreement with previous calculation results.39


9


Page 10 of 36

However, the adsorption of C on the sub-surface Oct site has ~0.6 eV lower adsorption energy than on the surface Br site, indicating that a C monomer is more likely to diffuse into the sursurface of Cu(111) substrate. In Figure 1, we plot the most and second most stable local configurations for C adsorption on these three surfaces. We can see that the C monomer adsorbed at the Oct site is six-coordinated with Cu–C distances of less than 2.0 Å, while the one adsorbed at the Br site is two-coordinated with a Cu–C distance of 1.80 Å. The higher coordination of the former configuration leads to stronger chemical bonding and thus reduces the adsorption energy. Even so, the closed d-electron shell of Cu results in a low chemical affinity for C so that the solubility of C in copper is actually minimal.7, 40 Therefore, we only consider the surface adsorption sites when constructing the complex step models in the following sections.

Figure 1 (a) The most and (b) second most stable adsorption configurations of C on Cu(111), CuO(111), and Cu2O(111) surfaces. Red, golden, and gray balls represent O, Cu, and C atoms, respectively. The distances between atoms (Å) and the adsorption energies are marked in the figures.


10

Page 11 of 36 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 contrast, as shown in Table S1, the two most stable C adsorption configurations on CuO(111) and Cu2O(111) all occur at the surface sites – specifically the Br4 site of CuO(111) and the Cusuf site of Cu2O(111), indicating that the oxidation of the Cu substrate could reduce the C solid solution process. The lowest adsorption energies of C on CuO(111) and Cu2O(111) were found to be –8.58 and –8.82 eV, respectively, which are all significantly lower (~2 eV) than that on the Cu(111) surface. This indicates that C prefers to adsorb on the oxide surface, especially on Cu2O, rather than the Cu surface. We also found that for C adsorbed on a CuxO surface, it is extremely easy to pull surface O atoms out of the surface by forming a Cu–C–O structure. This, therefore, leads to huge surface relaxation and deformation, resulting in significant reduction of the adsorption energy by ~3 eV, as compared to C adsorption on the Cu(111) surface. 3.1.2 CH adsorption Here we turn to CH adsorption on the clean Cu(111), CuO(111), and Cu2O(111) surfaces. In order to study the effect of CH steric configuration on the adsorption configurations and energies, we put the CH active group in both horizontal and vertical configurations relative to the considered surfaces, and all the adsorption sites were investigated, as in the case of a C monomer. In Figure S2, we give some initial CH adsorption configurations before and after structural relaxation on the three surfaces. We found that the C atom in the CH active group prefers to combine with the Cu substrate during structural optimization, and eventually, the CH group remains in an almost vertical structure with a C-H bond length of ~1.11 ± 0.01 Å. In Figure 2 we plot the two most stable CH adsorption configurations on Cu(111), CuO(111), and Cu2O(111) surfaces. It can be seen that CH could only stably adsorb on the surface sites of the three surfaces due to the steric-hinderance effect, which is different from the sub-surface diffusion of the C monomer. This phenomenon is consistent with the results reported in the


11


Page 12 of 36

literature.38 Similar to the case of C adsorption, CH groups adsorbed on the oxide surfaces have lower adsorption energies than those on the Cu(111) surface. The large difference (~2.5 eV) in the CH adsorption energies between the Cu and oxide surfaces also arises from the formation of C-O valence bonds on the CuO and Cu2O(111) surfaces. Similar to C adsorption, the adsorption of CH also causes slight surface deformation, but it does not induce complete deoxidation on the surface. However, the most stable CH adsorption occurs on the CuO(111) surface which is different from the case for C adsorption (at the Cu2O(111) surface). We summarize the adsorption energies of the most stable C and CH adsorption configurations on Cu(111), CuO(111), and Cu2O(111) surfaces in Table 1. It is noted that for all three surfaces, C adsorption exhibits overall ~1 eV lower adsorption energies than those of CH adsorption, indicating that C adsorption is energetically preferable on any surface with respect to CH adsorption. Therefore, we only consider the C monomer as the active species for graphene nucleation in the following thermodynamic and kinetic calculations.

Figure 2 (a) The most and (b) second most stable adsorption configurations of CH on Cu(111), CuO(111), and Cu2O(111) surfaces. White balls represent H atoms.


12

Page 13 of 36 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


Table 1 The adsorption energies (in eV) of the most stable C and CH adsorption configurations on Cu(111), CuO(111), and Cu2O(111) surfaces, as well as at CuO/Cu(111) and Cu2O/Cu(111) steps. Surface

Adsorption energy C

CH

Cu(111)

−5.93

−5.17

CuO(111)

−8.58

−7.70

Cu2O(111)

−8.82

−7.04

CuO/Cu(111)

−9.27

-

Cu2O/Cu(111)

−7.90

-

3.2 C adsorption at CuxO/Cu (111) steps The CuO/Cu(111) and Cu2O/Cu(111) step models were constructed by placing CuO and Cu2O clusters on clean Cu(111) surfaces. We considered both stoichiometric (Cu8O8 and Cu12O6) and non-stoichiometric (Cu7O8 and Cu13O6) clusters of CuO and Cu2O (Figure S3a) according to their crystal structures in order to obtain the most reasonable step configurations energetically and structurally. By comparing the binding energies of different CuxO clusters on Cu substrates (Figure S3b), we finally chose non-stoichiometric CuxO clusters to build the CuO/Cu(111) and Cu2O/Cu(111) step models. 3.2.1 C adsorption behavior Based on the most stable step configurations, we investigated the C adsorption behavior by comparing the adsorption energies of C on a series of adsorption sites around the CuO/Cu(111) and Cu2O/Cu(111) steps. The considered adsorption sites and their corresponding adsorption energies are shown in Figure S4 and Table S2. It was found that the adsorption energies of C at


13


Page 14 of 36

the CuO/Cu(111) step are much lower (0.97-2.36 eV) than those at the Cu2O/Cu(111) step for both of the two most stable configurations, indicating that the introduction of CuO defects is energetically beneficial for C adsorption at the step. From the primary and secondary stable C adsorption configurations in Figure 3, we determine that C adsorption on site I of the CuO/Cu(111) step breaks the original Cu–O bond and forms Cu–C–O and O–C–O structures by combining with a Cu atom and pushing out the top-most O atom, which is similar to the case of C adsorption on the CuO(111) surface. The secondary stable adsorption configuration is different because the introduction of C does not break the Cu–O chemical bond and maintains the original structure of the CuO cluster, which leads to the adsorption energy 1.5 eV higher than the most stable configuration. On the other hand, the C monomer adsorbed at the Cu2O/Cu(111) step showed strong chemical interactions with both the Cu2O cluster and the Cu(111) substrate, which is different from the case of adsorption at the CuO/Cu(111) step. For example, for the most stable adsorption configuration, the C monomer not only interacts with the Cu2O cluster by breaking the Cu–O bond and forming a Cu–C–O structure at site I, but also interacts with the substrate by forming a Cu–C–Cu structure at site II (Figure S3). Although the strong interactions of C with both the Cu2O cluster and Cu substrate could stabilize the structure, the full structural relaxation and energy gain due to the pushing out of an O atom at the CuO/Cu(111) step is beneficial for reducing the adsorption energy and for continuous C adsorption. Therefore, sufficient oxidation is recommended during the oxidation and etching pretreatment process to prepare the copper foil for CVD. This would promote the transformation of Cu2O defects on the Cu substrate to CuO, and thus further promote the adsorption and nucleation of C atoms around these CuO defects. Based on our previous calculation results and conclusions, we will only


14

Page 15 of 36 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


investigate the elementary reactions of C-C dimer formation and defect ring transformation at the CuO/Cu(111) step.

Figure 3 (a) The most and (b) second most stable adsorption configurations of C at CuO/Cu(111) and Cu2O/Cu(111) steps. 3.3 C–C dimer formation reaction In the above sections, we illustrated the types of oxide defects and found the active sites for initial C adsorption according to the calculated adsorption energies. However, graphene nucleation and growth is a continuous C aggregation and reaction process, rather than simple atomic adsorption. Therefore, it is insufficient to understand the preference of C atoms for gathering and nucleation at the initial nucleation stage just by comparing adsorption energies.


15


Page 16 of 36

Here, we take the elementary reaction of two C atoms gathering to form a C–C dimer as an example to investigate the heat of reaction (∆E) and energy barrier (Ea). We will evaluate the nucleation preference by combining the C adsorption energy and the energy barrier of the C aggregation reaction. 3.3.1 C–C dimer formation reaction on Cu(111) surface We first study the C-C dimer formation reaction on a clean Cu(111) surface in order to compare the reaction barriers on the clean Cu(111) surface and the CuO/Cu(111) step. We constructed two individual C adsorbed structures (C*+C*, *indicates adsorbed species) and the C–C dimer structure (C2*) on a Cu(111) surface by adding one C atom (labeled as C2) on the C (C1) of the most stable adsorption configuration, Cu(111)-Br. It was found that when individual C atoms are placed at nearby adsorption sites (such as Fcc, Hcp, and Br sites), a dimer structure is spontaneously formed (activation energy free), and this phenomenon is consistent with the findings of Riikonen et al.39 When the two C atoms are not placed at adjacent sites on the surface, they need to diffuse on the surface and overcome a reaction energy barrier to polymerize as a dimer. Therefore, the energy barrier is an important factor that decides if the elementary reaction from C*+C* to C2* will occur. The optimized C*+C* and C2* dimer adsorption structures are treated as the initial state (IS) and the final state (FS) of the dimer formation reaction, as shown in Figure 4. The IS structure is similar to the bridging-metal (BM) structure proposed by Wu et al.,41 where the surface Cu atom shared by C1 and C2 is pulled upward from the surface. At the same time, C2 penetrates the sub-surface through the hole left behind by the Cu atom, forming a C1–Cu–C2 structure. The distance between the two C atoms is 3.67 Å. In the FS structure, the two C atoms are bonded to the Fcc and Hcp sites of the Cu(111) surface and form a C–C dimer with a chemical bond length of 1.30 Å, which makes the C*+C* adsorption


16

Page 17 of 36 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


configuration more stable. During the reaction from IS to FS, a transition state (TS) is searched. In the TS structure, C1 and C2 become closer compared to in the IS structure, and they adsorb at the nearby Br sites with a distance of 2.16 Å. Moreover, the C2 atom that is inserted into the surface gradually moves back to the Cu surface and the two C atoms maintain a symmetrical structure. The formation of such a TS structure leads to a reaction barrier of 0.77 eV, indicating that the aggregation process of separated C atoms to form a dimer structure is not spontaneous and a relatively high Ea needs to be overcome. This conclusion is consistent with the findings of Riikonen and Li et al.,39,42 and the deviation of our calculated Ea value from the reported ones may come from errors in the theoretical models and calculation parameters.

Figure 4 Potential energy profiles for the formation of a C–C dimer on the Cu(111) surface. 3.3.2 C–C dimer formation reaction at CuO/Cu(111) step By using the same method and based on the most stable CuO/Cu(111) step configurations, we constructed C*+C* (IS) and C2*(FS) structures and investigated the dimer formation reaction


17


Page 18 of 36

behavior. It is noted that the primary and secondary stable CuO/Cu(111) step configurations represent both cases of keeping or not keeping the original CuO lattice structure. In order not to neglect the effect of the detailed CuO structure, the dimer formation reaction was investigated at both CuO/Cu(111) step configurations. The potential energy profile and the corresponding IS, TS, and FS configurations of the reaction from C*+C* to C2* at site I and site II of the CuO/Cu(111) step are shown in Figure 5. Obviously, the formation of a C2* dimer at site I is energy barrier free, indicating that two separately adsorbed C1 and C2 atoms could diffuse and aggregate to form a C2* dimer spontaneously. For the reaction occurring at site II, the formation of a C2* dimer shows an Ea of 0.34 eV, which is 0.43 eV lower than that on the clean Cu(111) surface. It is found that the C2* dimer interacts with the Cu substrate at site II, and this leads to the formation of the Cu–C–C–O structure, which makes the structure more stable and is considered to be the main reason for the large reduction in the Ea as compared to the clean Cu(111) surface. Overall, the introduction of CuO defects on a Cu(111) substrate could significantly reduce the reaction barrier of C2* dimer formation, and the formation even becomes spontaneous at some special sites. Therefore, at the initial nucleation stage, such a large reduction of the reaction barrier could promote graphene nucleation around the CuO defects by suppressing nucleation on the clean Cu(111) substrate surface. In this way, we can artificially introduce a low-density of CuO defects as “seed crystals” on a pretreated Cu(111) substrate to reduce the nucleation density of graphene and achieve controllable growth of graphene around these “seed crystals”.


18

Page 19 of 36 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 Potential energy profiles for the formation of C–C dimer at sites (a) I and (b) II of a CuO/Cu(111) step. To verify our calculations, relevant experiments were conducted and the results are analyzed here. Figure 6(a) shows the XRD pattern of the Cu substrate after temperature gradient annealing. The strong peak appearing at 43.14° corresponds to the Cu(111) plane, indicating that the Cu substrate becomes a single crystal with (111) orientation after annealing. A Cu substrate


19


Page 20 of 36

with a (111) crystal orientation is advantageous for graphene growth because of its higher adaptation ratio of graphene.43 Figure 6(b) shows the SEM images of the copper substrate with CuO nanoparticles (“seed crystals”). In order to confirm the composition of the “seed crystals” the EDS tests were performed on the seed regions before and after graphene growth, results are shown in Figure 6(c). It can be observed that the seed regions all contain a certain amount of oxygen, indicating that the “seed crystals” on the Cu substrate is a kind of copper oxide. It is also noticed that a small amount of Si/SiOx was detected on the substrate after graphene growth, while the Si/SiOx was proved to have no stimulating effect on the graphene nucleation.44 The chemical composite of the copper oxide on the substrate before graphene growth was analyzed by using the HRTEM and SAED (Figures 6(d) and 6(e)), where the Miller indexes of the “seed crystals” were found in accordance with the features of monoclinic CuO. In order to study the chemical stability of the CuO “seed crystals”, the valence of the Cu before and after graphene growth was characterized using the XPS method, and the results are shown in Figure 6(f). It is observed that the XPS spectra are almost the same (the variation of binding energy is 0.2 eV, within the systematic error), where the two strong peaks at 932.9±0.1 eV (Cu2p3/2) and 952.8±0.1 eV (Cu2p1/2) show the characteristics of elemental copper. 45 Although the Cu2p3/2 binding energy of Cu+ is very close to the Cu0,

46

we can still conclude that the peak at 932.9±0.1 eV corresponds to the

elemental Cu, since the fact that the peak-shape shows high symmetry and the main component of the substrate is elemental Cu. In addition, the peaks at 935.4±0.1 eV and 954.9±0.1 eV could be ascribed to the divalent copper (Cu2+).45 Combining with the weak shake-up peak47 near 944±0.1 eV, we can further determine that the “seed crystals” are CuO, which maintains the same before and after graphene growth.


20

Page 21 of 36 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 6 (a) XRD pattern of Cu substrate after temperature gradient annealing. (b) SEM image of seed crystals on a Cu(111) substrate. (c) EDS results of the seed regions before (left) and after (right) graphene growth. (d) HRTEM image of the seed region before graphene growth. (e) The SAED pattern of a seed crystal. (f) The XPS curves for the samples before (left) and after (right) graphene growth.


21


Page 22 of 36

Figure 7(a) presents the SEM image of graphene grown on Cu(111) substrate. As expected, the graphene preferentially nucleates and grows around the CuO “seed crystals”, which well supports our theoretical calculations. Furthermore, the layer number of the grown graphene was characterized using the Raman spectroscopy and HRTEM, and results are presented in Figures 7(b) and 7(c). The intensity ratio of the 2D and G bands (I2D/IG) is 1.3 and the full width at halfmaximum (FWHM) value of the 2D peak is ~41 cm-1, which conforms to the characteristics of bilayer graphene.48 The bilayer graphene is also observed in HRTEM image (Figure 7(c)).

Figure 7 (a) SEM image of graphene grown surrounded the “seed crystals” on Cu(111) substrate. (b) Raman spectrum of the grown graphene. (c) HRTEM image of the graphene grown surrounded the “seed crystals”. 3.4 The transformation between perfect graphene ring and defect rings Typically, due to the aperiodicity of the step structure, graphene grown around steps have a relatively higher defect density than those grown on perfect surfaces. In order to understand the


22

Page 23 of 36 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


influence of CuO defects on the growth quality of graphene, we further explored the transformation mechanism between perfect and defective graphene structures at CuO/Cu(111) steps. As we know, the C5 and C7 defect rings are the most common defect rings in graphene. In order to explore the defect formation and graphene self-recovery mechanisms at the step, we therefore chose to examine C5 and C7 defect rings and investigate the transformation reactions of a defective C5 ring to a perfect C6 ring and a perfect C6 ring to a defective C7 ring at the CuO/Cu(111) step. The related reactions on clean Cu(111) surface were also investigated for comparison. 3.4.1 C5 to C6 transformation reaction We investigated defect self-recovery during the nucleation and growth process by studying the transformation reaction from a C5 defect ring to a perfect C6 graphene ring. The reactions on both the Cu(111) surface and CuO/Cu(111) step were calculated for comparison. First, we found the most stable C5 adsorption configurations on the Cu(111) surface and around the CuO/Cu(111) step (see in Figure S5). It is found that the C atoms in C5 are stably adsorbed at adjacent Br and Hcp sites with C–C bond lengths of ~1.47 Å on the clean Cu(111) surface. In addition, C5 is stably adsorbed at the IX site on the CuO/Cu(111) step (as shown in Figure S4) with C–C bond lengths of ~1.59 Å, and it forms stable Cu–C5–Cu and Cu–C5–O structures with the Cu substrate and CuO cluster. Then we add another C atom based on the above C5 stable adsorption configurations to find the stable IS and FS of the C6 adsorption configurations on the Cu(111) surface and the CuO/Cu(111) step. On the clean Cu(111) surface, similar to the BM structure observed during C–C dimer formation, the stable adsorbed C* easily penetrates into the Cu subsurface and forms a C5–Cu–C BM structure (C5*+C*) with a minimum distance of 3.76 Å between C5 and C, which is considered to be the IS structure of the C5 to C6 reaction (Figure 8a).


23


Page 24 of 36

Besides, the C6* ring formed by aggregating the added C* and the C5* ring is considered to be the FS structure, and all C atoms in the C6* ring occupy the Fcc or Hcp sites of the Cu(111) surface. The stable adsorption configurations of C5* and C6* rings we obtained are well consistent with the structures reported by Wesep and his collaborators.49 From Figure 8a, we can see the reaction barrier from a separated C5*+C* aggregate to C6* is 0.70 eV, which is slightly lower than that of C–C dimer formation on the Cu(111) surface, indicating that the graphene defect C5 rings may be repaired by injecting the carbon source gas, and the processes of graphene nucleation and self-recovery may occur simultaneously. For the CuO/Cu(111) step, when we added one C atom at different sites near C5* to obtain stable C5*+C* (IS) and C6* (FS) configurations, we found that the C* prefers to adsorb on the Cu(111) surface instead of the step with a shortest C*‒C5* distance of 3.62 Å by forming a C5*+C* (IS) configuration (Figure 8b), which is quite different from the C monomer adsorption behavior (preferential adsorption on the CuO/Cu(111) step). In contrast with the case on the Cu(111) surface, no C atom penetrates the sub-surface, and the Cu atoms between C* and C5* are pulled out from the Cu substrate surface by the formation of the C–Cu–C structure. In addition, due to the strong interactions between C5 and Cu, the Cu–C–C–O structure is generated via deoxidation of CuO by the C5*. The dissociated O atom will recombine with CuO to form a C–O–Cu structure with a perfect C6* ring (FS). It was determined that the reaction barrier from C5*+C* to C6* is 0.71 eV, which is almost the same as that on the Cu(111) surface. This illustrates that the possibilities of repairing graphene C5 defects to form a perfect structure are comparable on the clean Cu(111) surface and the CuO/Cu (111) step.


24

Page 25 of 36 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 8 Potential energy profiles for the transformation from C5 to C6 rings on the (a) Cu(111) surface and (b) CuO/Cu(111) step, respectively. 3.4.2 C6 to C7 transformation reaction Using the same method, we investigated defect formation during the nucleation and growth process by studying the transformation reaction from the perfect C6 graphene ring to the C7


25


Page 26 of 36

defect ring. Based on the stable configurations of C6* on Cu(111) and CuO/Cu(111) step shown in Figure S5, we added another C atom at different sites near C6* to find the IS and FS structures for the transformation reactions of C6*+C* to C7*. From Figure 9a, we can see that on the clean Cu(111) surface, the added C atom stably adsorbs at the Br site of the Cu(111) surface near the C6* ring by forming a BM structure of C6–Cu–C in the IS structure. The minimum distance between C6* and C* is 3.67 Å. As the C* moves closer to the Hcp site of the Cu substrate, the C6* ring would obviously rotate relative to the Cu substrate (in the TS structure), and eventually, the C7* defect ring (FS) becomes irregular and bends slightly upward on the Cu(111) surface. From Figure 9b we observe that at the CuO/Cu(111) step, the structural evolution during the reaction from C6*+C* to C7* is similar to that of C5*+C* to C6* in Figure 8b, and both the C– Cu–C and Cu–C–C–O structures were formed. However, no breakage and re-bonding of Cu–O bond were observed during the entire reaction process. The transformation reaction from C6*+C* to C7* on the clean Cu(111) surface needs to overcome a reaction barrier of 0.56 eV, which is lower than that of C5*+C* to C6* by 0.14 eV. This indicates that the formation of defective graphene is easier than self-recovery from defective graphene on the clean Cu(111) surface. When a CuO defect is introduced on the Cu(111) surface, the reaction barrier to form C7* is 0.08 eV lower than that on the clean Cu(111) surface, indicating that more graphene defect structures would be formed at the CuO/Cu(111) steps, which is consistent with our expectations. However, it is worth noting that the reduction in the energy barrier (~0.08 eV) to form a graphene defect structure around CuO defect is an order of magnitude smaller than the reduction in the energy barrier (~0.77 eV) for graphene nucleation at the CuO/Cu(111) step. Overall, the benefits brought out by introducing CuO defects for graphene nucleation far outweigh the disadvantages


26

Page 27 of 36 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


for improving the graphene quality. Therefore, the results and conclusions of our study of defect transformation would not influence the main conclusions drawn in the above sections.

Figure 9 Potential energy profiles for the transformation from C6 to C7 rings on (a) Cu(111) surface and (b) CuO/Cu(111) step, respectively.


27


Page 28 of 36

4. CONCLUSION In this work, we studied the adsorption behavior of C active species on clean Cu(111) substrates and substrates with CuO and Cu2O defects by DFT calculations. The formation reactions of C–C dimers and the transformation reactions between perfect and defective graphene rings at the initial nucleation stage were further studied by combining NEB transition state calculations. It was found that the CuO/Cu(111) step structure could not only significantly reduce the adsorption energy of C by 3.34 eV as compared to that on the clean Cu(111) surface, but also reduce the reaction barrier of C–C dimer formation from 0.77 eV on the clean Cu(111) surface to 0.34 eV and even eliminate the energy barrier at some special step sites. Therefore, the introduction of CuO defects could promote the initial nucleation of graphene around CuO and suppress the nucleation on the clean Cu(111) substrate. Experimentally, the low-density CuO nucleation sites were artificially introduced on Cu(111) substrate as the “seed crystals” by oxidation and hydrogen etching of the copper foil. It was observed that graphene prefers to nucleate and grow centered on CuO “seed crystals,” which verifies the results of our calculations. Taking advantage of this, we can achieve controllable nucleation of graphene and reduce the graphene nucleation density by controlling the CuO seed density. Further calculation results indicate that the step structure would not lead to significant formation of defective graphene. ACKNOWLEDGEMENTS This work was supported by the National Natural Science Foundation of China (Grant Nos. 51502158 and 51202129) and the Foundation 31513020404-1.


28

Page 29 of 36 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


ASSOCIATED CONTENT Supporting Information The supporting information related to this article can be found on the ACS Publication website at DOI:xxxxxx. It includes the following information: Adsorption sites on Cu(111), CuO(111) and Cu2O(111) surfaces (Figure S1); Comparison of the adsorption energies for C and CH (Table S1); CH adsorption configurations (Figure S2); CuxO/Cu(111) step models (Figure S3); Adsorption sites and adsorption energies of C at CuxO/Cu(111) steps (Figure S4, Table S2); Adsorption configurations of C5, C6, and C7 rings (Figure S5) REFERENCES (1) Geim, A. K.; Novoselov, K. S. The Rise of Graphene. Nat. Mater. 2007, 6, 183–191. (2) Mak, K. F.; Sfeir, M. Y.; Wu, Y.; Lui, C. H.; Misewich, J. A.; Heinz, T. F. Measurement of the Optical Conductivity of Grapheme. Phys. Rev. Lett. 2008, 101, 6797–6800. (3) Min, K.; Aluru, N. R. Mechanical Properties of Graphene under Shear Deformation. Appl. Phys. Lett. 2011, 98, 013113. (4) Lee, C.; Wei, X. D.; Kysar, J. W.; Hone, J. Measurement of the Elastic Properties and Intrinsic Strength of Monolayer Graphene. Science 2008, 321, 385–388. (5) Zhou, H. L.; Yu, W. J.; Liu, L. X.; Cheng, R.; Chen, Y.; Huang, X. Q.; Liu, Y.; Wang, Y.; Huang, Y.; Duan, X. F. Chemical Vapour Deposition Growth of Large Single Crystals of Monolayer and Bilayer Graphene. Nat. Commun. 2013, 4, 2096. (6) Gao, L.; Guest, J. R.; Guisinger, N. P. Epitaxial Graphene on Cu(111). Nano Lett. 2010, 10, 3512–3516.


29


Page 30 of 36

(7) Li, X. S.; Cai, W. W.; Colombo, L.; Ruoff, R. S. Evolution of Graphene Growth on Ni and Cu by Carbon Isotope Labeling. Nano Lett. 2009, 9, 4268–4272. (8) Li, X. S.; Magnuson, C. W.; Venugopal, A.; Tromp, R. M.; Hannon, J. B.; Vogel, E. M.; Colombo, L. G.; Ruoff, R. S. Large-Area Graphene Single Crystals Grown by Low-Pressure Chemical Vapor Deposition of Methane on Copper. J. Am. Chem. Soc. 2011, 133, 2816–2819. (9) Li, J.; Wang, X. Y.; Liu, X. R.; Jin, Z.; Wang, D.; Wan, L. J. Facile Growth of CentimeterSized Single-Crystal Graphene on Copper Foil at Atmospheric Pressure. J. Mater. Chem. C 2015, 3, 3530–3535. (10) Cummings, A. W.; Duong, D. L.; Nguyen, V. L.; Tuan, D. V.; Kotakoski, J.; Vargas, J. E. B.; Lee, Y. H.; Roche, S. Charge Transport in Polycrystalline Graphene: Challenges and Opportunities. Adv. Mater. 2014, 26, 5079–5094. (11) Yang, M.; Sasaki, S.; Suzuki, K.; Miura, H. Control of the Nucleation and Quality of Graphene Grown by Low-Pressure Chemical Vapor Deposition with Acetylene. Appl. Surf. Sci. 2016, 336, 219–226. (12) Wang, H.; Wang, G. Z.; Bao, P. F.; Yang, S. L.; Zhu, W.; Xie, X.; Zhang, W. J. Controllable Synthesis of Submillimeter Single-Crystal Monolayer Graphene Domains on Copper Foils by Suppressing Nucleation. J. Am. Chem. Soc. 2012, 134, 3627–3630. (13) Yan, Z.; Lin, J.; Peng, Z. W.; Sun, Z. Z.; Zhu, Y.; Li, L.; Xiang, C. S.; Samuel, E. L.; Kittrell, C.; Tour, J. M. Toward the Synthesis of Wafer-Scale Single-Crystal Graphene on Copper Foils. ACS Nano 2012, 6, 9110–9117. (14) Zhang, Y. H.; Chen, Z. Y.; Wang, B.; Wu, Y. W.; Jin, Z.; Liu, X. Y.; Yu, G. H. Controllable Growth of Millimeter-Size Graphene Domains on Cu Foil. Mater. Lett. 2013, 96, 149–151.


30

Page 31 of 36 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


(15) Xu, X. Z.; Zhang, Z. H.; Qiu, L.; Zhuang, J. N.; Zhang, L.; Wang, H.; Liao, C. N.; Song, H. D.; Qiao, R. X.; Gao, P.; Hu, Z. H.; Liao, L.; Liao, Z. M.; Yu, D. P.; Wang, E.; Ding, F.; Peng, H. L.; Liu, K. H. Ultrafast Growth of Single-Crystal Graphene Assisted by a Continuous Oxygen Supply. Nat. Nanotechnol. 2016, 11, 930–935. (16) Hao, Y. F.; Bharathi, M. S.; Wang, L.; Liu, Y. Y.; Chen, H.; Nie, S.; Wang, X. H.; Chou, H.; Tan, C.; Fallahazad, B.; Ramanarayan, H.; Magnuson, C. W.; Tutuc, E.; Yakobson, B. I.; McCarty, K. F.; Zhang, Y. W.; Kim, P.; Hone, J.; Colombo, L.; Ruoff, R. S. The Role of Surface Oxygen in the Growth of Large Single-Crystal Graphene on Copper. Science 2013, 342, 720– 723. (17) Chen, J. W.; Cui, M. Q.; Wu, G. X.; Wang, T. Y.; Mbengue, J. M.; Li, Y. F.; Li, M. C. Fast Growth of Large Single-Crystalline Graphene Assisted by Sequential Double Oxygen Passivation. Carbon 2017, 116, 133–138. (18) Chang, R. J.; Lee, C. H.; Lee, M. K.; Chen, C. W.; Wen, C. Y. Effects of Surface Oxidation of Cu Substrates on the Growth Kinetics of Graphene by Chemical Vapor Deposition. Nanoscale 2017, 9, 2324–2329. (19) Liang, T.; He, G.; Huang, G. W.; Kong, Y. H.; Fu, W. F.; Chen, H. Z.; Wang, Q.; Iwai, H.; Fujita, D.; Liu, Y. C.; Xu, M. S. Graphene Nucleation Preferentially at Oxygen-Rich Cu Sites Rather Than on Pure Cu Surface. Adv. Mater. 2015, 27, 6404–6410. (20) Zhang, J.; Yang, Z. Y.; Sun, L.; Yu, F. P.; Cheng, X. F.; Liu, X. Z.; Zhao, X. Preparation of Bilayer Graphene Utilizing CuO as Nucleation Sites by CVD Method. J. Mater. Sci.: Mater. El. 2018, 29, 4495–4502. (21) Hsieh, Y. P.; Chu, Y. H.; Tsai, H. G.; Hofmann, M. Reducing the Graphene Grain Density in Three Steps. Nanotechnology 2016, 27, 105602.


31


Page 32 of 36

(22) Wang, S. N.; Hibino, H.; Suzuki, S.; Yamamoto, H. Atmospheric Pressure Chemical Vapor Deposition Growth of Millimeter-Scale Single-Crystalline Graphene on the Copper Surface with a Native Oxide Layer. Chem. Mater. 2016, 28, 4893–4900. (23) Reckinger, N.; Hooijdonk, E. V.; Joucken, F.; Tyurnina, A. V.; Lucas, S.; Colomer, J. F. Anomalous Moiré Pattern of Graphene Investigated by Scanning Tunneling Microscopy: Evidence of Graphene Growth on Oxidized Cu(111). Nano Res. 2014, 7, 154–162. (24) Chuang, M. C.; Woon, W. Y. Nucleation and Growth Dynamics of Graphene on Oxygen Exposed Copper Substrate. Carbon 2016, 103, 384–390. (25) Gottardi, S.; Müller, K.; Bignardi, L.; Moreno-López, J. C.; Pham, T. A.; Ivashenko, O.; Yablonskikh, M.; Barinov, A.; Björk, J.; Rudolf, P.; Stöhr, M. Comparing Graphene Growth on Cu(111) Versus Oxidized Cu(111). Nano Lett. 2015, 15, 917–922. (26) Reckinger, N.; Tang, X. H.; Joucken, F.; Lajaunie, L.; Arenal, R.; Dubois, E.; Hackens, B.; Henrard, L.; Colomer, J. F. Oxidation-Assisted Graphene Heteroepitaxy on Copper Foil. Nanoscale 2016, 8, 18751–18759. (27) Banhart, F.; Kotakoski, J.; Krasheninnikov, A. V. Structural Defects in Graphene. ACS Nano 2010, 5, 26–41. (28) Blöchl, P. E. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953–17979. (29) Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, 11169–11186. (30) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865–3868. (31) Mi, X.; Meunier, V.; Koratkar, N.; Shi, Y. F. Facet-Insensitive Graphene Growth on Copper. Phys. Rev. B 2012, 85, 155436.


32

Page 33 of 36 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


(32) Henkelman, G.; Uberuaga, B. P.; Jónsson, H. A Climbing Image Nudged Elastic Band Method for Finding Saddle Points and Minimum Energy Paths. J. Chem. Phys. 2000, 113, 9901– 9904. (33) Mills, G.; Jónsson, H.; Schenter, G. K. Reversible Work Transition State Theory: Application to Dissociative Adsorption of Hydrogen. Surf. Sci. 1995, 324, 305–337. (34) Chen, H.; Zhu, W. G.; Zhang, Z. Y. Contrasting Behavior of Carbon Nucleation in the Initial Stages of Graphene Epitaxial Growth on Stepped Metal Surfaces. Phys. Rev. Lett. 2010, 104, 186101. (35) Li, X. S.; Zhu, Y. W.; Cai, W. W.; Borysiak, M.; Han, B. Y.; Chen, D.; Piner, R. D.; Colombo, L.; Ruoff, R. S. Transfer of Large-Area Graphene Films for High-Performance Transparent Conductive Electrodes. Nano Lett. 2009, 9, 4359–4363. (36) Hwang, C. Y.; Yoo, K.; Kim, S. J.; Seo, E. K.; Yu, H.; Biró, L. P. Initial Stage of Graphene Growth on a Cu Substrate. J. Phys. Chem. C 2011, 115, 22369–22374. (37) Shu, H. B.; Tao, X. M.; Ding, F. What Are the Active Carbon Species During Graphene Chemical Vapor Deposition Growth? Nanoscale 2015, 7, 1627–1634. (38) Wang, X. L.; Yuan, Q. H.; Li, J.; Ding, F. The Transition Metal Surface Dependent Methane Decomposition in Graphene Chemical Vapor Deposition Growth. Nanoscale 2017, 9, 11584–11589. (39) Riikonen, S.; Krasheninnikov, A. V.; Halonen, L.; Nieminen, R. M. The Role of Stable and Mobile Carbon Adspecies in Copper-Promoted Graphene Growth. J. Phys. Chem. C 2012, 116, 5802–5809. (40) Mattevi, C.; Kim, H.; Chhowalla, M. A Review of Chemical Vapour Deposition of Graphene on Copper. J. Mater. Chem. 2011, 21, 3324–3334.


33


Page 34 of 36

(41) Wu, P.; Zhang, W. H.; Li, Z. Y.; Yang, J. L.; Hou, J. G. Communication: Coalescence of Carbon Atoms on Cu(111) Surface: Emergence of a Stable Bridging-Metal Structure Motif. J. Chem. Phys. 2010, 133, 071101. (42) Li, Y. F.; Li, M. C.; Wang, T.; Bai, F.; Yu, Y. X. DFT Study on the Atomic-Scale Nucleation Path of Graphene Growth on the Cu(111) Surface. J. Phys. Chem. C 2014, 16, 5213– 5220. (43) Ogawa, Y.; Hu, B. S.; Orofeo, C. M.; Tsuji, M.; Ikeda, K.; Mizuno, S.; Hibino, H.; Ago, H. Domain Structure and Boundary in Single-Layer Graphene Grown on Cu(111) and Cu(100) Films. J. Phys. Chem. Lett. 2016, 3, 219–226. (44) Ge, X. M.; Zhang, Y. H.; Chen, L. X.; Zheng, Y. H.; Chen, Z. Y.; Liang, Y. J.; Hu, S. K.; Li, J.; Sui, Y. P.; Yu, G. H.; Jin, Z.; Liu, X. L. Mechanism of SiOx particles formation during CVD graphene growth on Cu substrates. Carbon 2018, 139, 989-998 (45) Poulston, S.; Parlett, P. M.; Stone, P.; Bowker, M. Surface Oxidation and Reduction of CuO and Cu2O Studied Using XPS and XAES. Surf. Interface Anal. 2015, 24, 811–820. (46) Li, Z. H.; Huang, W.; Zuo, Z. J.; Song, Y. J. XPS Study on CuZnAl Catalysts Prepared by Different Methods for Direct Synthesis of Dimethyl Ether. Chinese J. Catal. 2009, 30, 171–177. (47) Biesinger, M. C.; Lau, L. W. M.; Gerson, A. R.; Smart, R. St. C. Resolving Surface Chemical States in XPS Analysis of First Row Transition Metals, Oxides and Hydroxides: Sc, Ti, V, Cu and Zn. Appl. Surf. Sci. 2010, 257, 887–898. (48) Chen, S. S.; Cai, W. W.; Piner, R. D.; Suk, J. W.; Wu, Y. P.; Ren, Y. J.; Kang, J. Y.; Ruoff, R. S. Synthesis and Characterization of Large-Area Graphene and Graphite Films on Commercial Cu–Ni Alloy Foils. Nano Lett. 2011, 11, 3519–3525.


34

Page 35 of 36 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


(49) Wesep, R. G. V.; Chen, H.; Zhu, W. G.; Zhang, Z. Y. Communication: Stable Carbon Nanoarches in the Initial Stages of Epitaxial Growth of Graphene on Cu(111). J. Chem. Phys. 2011, 134, 171105.


35


Page 36 of 36

Table of Contents


36

Graphene Nucleation Preference at CuO Defects Rather Than Cu2O

Recommend Documents